Correspondence: Xin Wang (jimmyxinwang@filtricine.com), Xiyan Li (xiyanli@filtricine.com)

Background

Cancer cells require outside supply of some non-essential amino acids (NEAAs) to survive. Dietary deprivation of select (NEAAs) can slow down the growth of multiple solid tumors in mice, creating a new non-drug strategy in cancer treatment. However, deprivation of NEAAs could negatively impact the immune activation, an essential process for immunotherapy, because fast cell proliferation poses a higher demand for building blocks such as NEAAs. It is not clear whether dietary NEAA deprivation could be combined with immunotherapy for better safety-efficacy profiles.

Methods

In this study, we tested the effects of NEAA-deprived diets and checkpoint inhibitor anti-PD-1 and anti-PD-L1 in colon cancer using syngeneic mouse model (Balb/c) bearing tumors of mouse colorectal cancer cell line CT-26. Three diets were tested, including a natural rodent diet Teklad ENVIGO Global 16% Protein Rodent Diet (control 1), a formulated NEAA-complete diet COMPLETE (control 2, using amino acid mix in place of protein), and a formulated NEAA-deprived diet FTN203 (treatment, using amino acid mix in place of protein). Both COMPLETE and FTN203 have the same nutritional structures, contain 17% w/w protein equivalent, and are isocaloric. After tumor size-based randomization, these diets were provided to mice ad libitum throughout the whole test. Each of these diets was used alone or combined with anti-PD-1 antibody (i.p., twice per week for 2 weeks) or anti-PD-L1 antibody (i.v., twice per week for 2 weeks).

Results

We found 1) On day 24 post tumor implantation, NEAA-deprived diet FTN203 significantly reduced tumor growth when used alone, compared to the group fed with Teklad ENVIGO (by 81%, P=0.0054, unpaired t-test after Welch correction) and COMPLETE (by 81%, P=0.013), respectively; 2) The efficacy of FTN203 is comparable with that of anti-PD-1 or anti-PD-L1 in tumor growth and median survival; 3) FTN203 did not negate the efficacy of anti-PD-1 or anti-PD-L1 immunotherapy antibody when combined; 4) FTN203 significantly improved the efficacy of anti-PD-1 by further reducing the tumor growth (by 80% on day 26, P=0.046) and increasing the median survival (by 5 days or 14%, Log-rank test P= 0.031), against the combo of COMPLETE and anti-PD-1; 5) None of the mono or combo treatments caused body weight loss.

Conclusions

Our data supports the use of dietary NEAA deprivation to improve the efficacy of anti-PD-1 or anti-PD-L1 immunotherapy for colorectal cancer without noticeable side effects. With further development, dietary NEAA deprivation may become the promising foundation for a broad spectrum of cancer therapies.

Ethics Approval

The study CA-XLI-6 was approved by the CRO's Ethics Board under IACUC approval number 19-015.9.

Correspondence: Christopher Larson (clarson@epicentrx.com)

Background

In the Phase 3 CORRECT study, which led to the approval of the multi-kinase inhibitor, Regorafenib, in 3rd/4th line metastatic colorectal cancer, the OS was 6.4 months and the PFS was 1.9 months compared to an OS of 5.0 months and a PFS of 1.7 months for placebo. However, Regorafenib is very poorly tolerated with a Grade 3/4 drug related adverse event rate of 54%, mostly due to hand-foot skin reactions, fatigue and diarrhea, resulting in frequent dose reductions and discontinuations and a general reluctance among GI oncologists to administer it. RRx-001 is a minimally toxic macrophage repolarizing agent in Phase 3 clinical trials that is associated with a reduced side effect profile from these chemotherapy agents. Recent studies have demonstrated the inhibitory impact of M2 macrophages on the activity of tyrosine kinases suggesting that the repolarization of macrophages by RRx-001 may enhance the activity of TKIs.

Methods

These experiments determined whether combination therapy with RRx-001 and regorafenib not only enhanced anticancer activity in vitro with HCT-116 and HCT-15 colorectal cell lines and in vivo with HCT 116 and HCT 15 xenografts but also attenuated the toxicity of regorafenib in these two xenografts.

Results

The results from these experiments demonstrate that 1) RRx-001 + regorafenib is more effective than either agent alone both in vitro and in vivo and that 2) the addition of RRx-001 to regorafenib attenuates the toxicity of regorafenib in vivo.

Conclusions

A clinical trial is planned to investigate the translational potential of the RRx-001 + regorafenib combination. Future experiments will determine whether RRx-001 also enhances the activity and decreases the toxicity of other tyrosine kinase inhibitors such as sorafenib, sunitinib, dasatinib, imatinib, lapatinib, and cabozantinib, all of which possess similar efficacy and safety profiles, not only in colorectal cancer but also other tumor types.

Correspondence: Akseli Hemminki (akseli.hemminki@helsinki.fi)

Background

Treatment of metastatic solid tumors remains a challenge in oncology, especially in the context of locally given therapies. Some responses in non-treated metastases were described in the middle of the last century after using local radiotherapy in breast cancer patients. This phenomenon was named abscopal effect[1]. Nowadays, this effect is known to be elicited by systemic adaptive immune responses against tumor cells, following local treatment. Of note, abscopal effect remains somewhat limited in the clinic mainly due to the suppressive microenvironment surrounding cancer cells. The systemic use of immune checkpoint inhibitors, such as anti-CTLA-4 (anti cytotoxic T-lymphocyte-associated protein 4) and anti-PD-1 (anti programmed cell-death protein 1), has shown promise in the treatment of various tumor types but only a minority of patients respond[2]. Oncolytic virotherapy is an interesting strategy that has re-emerged in the last few decades. It is able to prime the host immune system against tumor epitopes, generating anti-tumor immunity and thus theoretically complementing anti-PD-1 and anti-PDL-1 therapy in an appealing manner[3,4]. The challenge with oncolytic immunotherapy is that not all metastases can usually be injected. Therefore, it is of importance to study if local treatment can induce distant responses. While our previous work has shown synergy between oncolytic adenovirus and anti-PD-1, here we sought to establish if local adenovirus injection can impact also non-injected tumors in pre-clinical models in the context of systemic anti-PD-1 therapy.

Methods

We utilized a murine melanoma model (B16.OVA) where each mouse had bilateral tumors. All mice received anti-PD-1 systemically, but only one tumor received local virotherapy with the non-replicative adenovirus. Both tumors were then studied and compared for possible systemic effects following local treatment, including immune responses and anti-tumor efficacy.

Results

Improvement in the overall survival was seen in the group receiving both therapies (anti-PD-1 plus virus) compared to the monotherapy groups. In particular, treatment with anti-PD-1 plus virus was the only group that presented a complete tumor regression of both tumors. Likewise, this therapeutic group presented significantly better tumors control in non-injected tumors than the other treatment groups. Further information, including full tumor growth and immunological mechanism-of-action data will be presented.

Conclusions

These two cancer immunotherapies seem to be a promising approach that could increase survival upon clinical translation. It is of importance that systemic anti-tumor effects result from local injection of adenovirus coding TNF- α and IL-2, even in the absence of virus replication.

Acknowledgements

We thank the Biomedicum FACS Core facility and the Animal the Finnish Centre for Laboratory Animal Pathology (FCLAP) at the University of Helsinki, for their expert assistance. This is study was supported by Jane and Aatos Erkko Foundation, HUCH Research Funds (EVO), Sigrid Juselius Foundation, Finnish Cancer Organizations, University of Helsinki, The Finnish Society of Sciences and Letters, and TILT Biotherapeutics Ltd.

Ethics Approval

All experiments described here have been approved by The Gene Technology Board of Finland, by the Experimental Animal Committee of the University of Helsinki and the Provincial Government of Southern Finland, under the license number ESAVI/7755/04.10.07/2016.

References

1. Mole RH. Whole Body Irradiation—Radiobiology or Medicine? Br J Radiol. 1953; 305: 234-241.

2. Demaria S, Kawashima N, Yang AM, Devitt ML, Babb JS, Allison JP, Formenti SC. Immune-Mediated Inhibition of Metastases after Treatment with Local Radiation and CTLA-4 Blockade in a Mouse Model of Breast Cancer. Clin Cancer Res. 2005; 15:728-34.

3. Havunen R, Santos JM, Sorsa S, Rantapero T, Lumen D, Siurala M, Airaksinen AJ, Cervera- Carrascon V, Tähtinen S, Kanerva A, and Hemminki A. Abscopal Effect in Non-injected Tumors Achieved with Cytokine-Armed Oncolytic Adenovirus. Mol Ther Oncolytics. 2018; 11: 109–121.

4. Cervera-Carrascon V, Siurala M, Santos JM, Havunen R, Tähtinen S, Karell P, Sorsa S, Kanerva A, Hemminki A.TNFa and IL-2 armed adenoviruses enable complete responses by anti-PD-1 checkpoint blockade. Oncoimmunology. 2018; 9;7(5):e1412902. doi: 10.1080/2162402X.2017.1412902. eCollection 2018.

Correspondence: William Redmond (William.Redmond@providence.org)

Background

Metabolic dysregulation is a hallmark of cancer; as such, many tumors exhibit auxotrophy for various nutrients as they are unable to meet the demand for these metabolites through endogenous production. Arginine auxotrophic tumors, for instance, are highly sensitive to depletion of systemic arginine. Pegzilarginase, a human arginase 1 enzyme engineered to have superior stability and enzymatic activity relative to the native human arginase 1 enzyme, depletes systemic arginine by converting it to ornithine and urea. Therapeutic administration of pegzilarginase in the setting of arginine auxotrophic tumors exerts direct anti-tumor activity by starving the tumor of exogenous arginine. Pegzilarginase monotherapy has shown significant anti-tumor activity in various arginine auxotrophic human and murine tumor models [1]. We hypothesized that in addition to direct effects, pegzilarginase treatment indirectly augments anti-tumor immunity through increased antigen presentation, thus making pegzilarginase a prime candidate for combination therapy with immuno-oncology (I-O) agents.

Methods

Tumor-bearing mice (CT26; MC38; MCA-205) received pegzilarginase (3 mg/kg; 1x/wk; ip) in combination with aPD-L1 (10mg/kg; 2x/wk; ip), aCTLA-4 (10mg/kg; 2x/wk; ip), or agonist aOX40 mAb (250ug, 2x; ip). The activation status of CD4+, CD8+, and NK cells in the tumor (7, 10, and 17 days post-treatment) was evaluated by flow cytometry. Effects on immune subsets were evaluated by flow cytometry and single cell RNA sequencing of sorted CD45+ cells from tumors 3 days post-treatment. Data represents the result of 2-3 independent experiments (n=9-18/group). Statistical significance was determined using a 1-way ANOVA with a p-value cut-off of 0.05.

Results

Pegzilarginase monotherapy increased MHC expression on antigen presenting cells (APCs) and the frequency of intratumoral CD8+ T cells relative to controls (p<0.01). Further, administration of pegzilarginase in combination with I-O agents, including aPD-L1 (p<0.01) and aOX40 (p<0.05), resulted in increased therapeutic benefit compared to I-O agents alone, including an increase in complete tumor regression. Combination therapy evinced the greatest number of activated intratumoral CD8+ T cells and elevated systemic levels of IFN-gamma.

Conclusions

Combination pegzilarginase/immunotherapy induces robust anti-tumor immunity characterized by increased intratumoral CD8+ T cells and M1-polarized macrophages. Our data suggests two potential mechanisms of synergy between pegzilarginase and I-O agents: 1) Increased intratumoral MHC expression on APCs and tumor antigen presentation; and 2) Increased presence of M1-like anti-tumor macrophages. These data support the clinical evaluation of T cell agonists and/or checkpoint inhibitors in conjunction with pegzilarginase for the treatment of patients with cancer.

Ethics Approval

All mice were maintained under specific pathogen-free conditions in the Providence Portland Medical Center animal facility. Experimental procedures were performed according to the National Institutes of Health Guide for the Care and Use of Laboratory Animals and in accordance with the Earle A. Chiles Research Institute Institutional Animal Care and Use Committee (Animal Welfare Assurance No. A3913–01).

References

1. Agnello G et al. Depleting blood arginine with AEB1102 (Pegzilarginase) exerts additive anti-tumor and synergistic survival benefits when combined with immunomodulators of the PD-1 pathway. SITC Annual Meeting, 2017, #275.

Correspondence: David Zaharoff (dazaharo@ncsu.edu)

Background

Clinically, cryoablation is used to treat various types of cancers including certain prostate, liver and kidney tumors. Because it liberates massive amounts of tumor antigen, cryoablation can potentially serve as an in situ vaccine that is patient- and tumor-specific. However, numerous studies report that cryoablation alone rarely provides abscopal immunity [1]. Therefore, we hypothesized that direct injection of an immunomodulator (IM) following cryoablation (cryo) can induce tumor-specific immunity, controlling both treated and untreated tumors.

Methods

MB49 cells were purchased from ATCC. Female C57BL/6 mice were purchased from Jackson Laboratory. Primary tumor implantation, 3x105 cells were injected subcutaneously (s.c.) in the right flank. Rechallenge, 3x105 cells (s.c.) were implanted in the left flank of cured mice. Bilateral model, 3x105 and 1.5x105 cells were s.c. injected in the right and left flanks respectively on the same day. Tumor volume calculation, 0.5*a*b2 given perpendicular long (a) and short (b) dimensions. Tumors, 200-500 mm3 ,were cryoablated with three cycles of freeze/thaw using Visual-ICETM Cryoablation System (Galil Medical/BTG). Unless noted otherwise, an IM was directly injected into the tumor bed two days post cryo. R848 (Resiquimod) 10 μg/mouse, DMXAA 500 μg/mouse, and Interleukin-12 (IL-12) 1 μg/mouse in 1.5% (w/v) chitosan acetate (CS) dissolved in dPBS.

Results

Cryoablation of a single primary tumor resulted in a tumor-free survival rate of 40%. The addition of R848 or DMXAA did not significantly enhance survival, while adding CS/IL-12 resulted in a 90% survival rate (Figure 1). Cured mice were rechallenged with live MB49 cells and then given no further treatment. As of day 70 post rechallenge, the number of tumor free mice out of total survivors were cryo alone (3/4), cryo + CS/IL-12 (5/7), cryo + R848 (2/5), and cryo + DMXAA (1/4). In the bilateral model, the median survival of the cryo + CS/IL-12 group was significantly longer (p = 0.0016) at 49.5 days compared to 26 days for cryo alone (Figure 2). Furthermore, one mouse treated with cryo + CS/IL-12 experienced a complete response up to 90 days post tumor implantation.

Conclusions

IL-12 formulated with chitosan and delivered after cryoablation is superior in treating not only a primary tumor, but also in inducing an abscopal effect on a distant, untreated tumor. Of note, neither DMXAA nor R848 was formulated in a delivery vehicle, as we have yet to optimize these agonists. This work supports further study of cryoablation and IL-12 for in situ vaccination and cancer immunotherapy.

Acknowledgements

This work is supported by BTG, the NC State University Provost’s Fellowship, and startup funds provided by the College of Engineering at NC State University.

Ethics Approval

The study was approved by the NC State University's Ethics Board, approval number 17-052.

References

1. Abdo J, Cornell DL, Mittal SK, Agrawal DK. Immunotherapy Plus Cryotherapy: Potential Augmented Abscopal Effect for Advanced Cancers. Front Oncol. 2018;8:85.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1

(abstract P505). Survival of single primary tumor model

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2

(abstract P505) Survival of bilateral model

Correspondence: Lisa Ridnour (ridnourl@mail.nih.gov)

Background

The presence of tumor infiltrating lymphocytes in breast cancer (BC) is associated with improved survival. More recently a study found that increased CD8 cells and Th17 cells are specifically associated with triple negative breast cancer (TNBC) patients, a highly aggressive subclass of breast cancer. However, they undergo functional reprogramming in the tumor micro-environment (TME) evident from decreased IFN-γ and granzyme B. These immune escape mechanisms contribute to inability of the immune system to control tumor progression. Thus modulation of TME is necessary to effectively target the tumor.

Radiation therapy (RT) is commonly used in cancer patients including BC. Focal radiation limits systemic side effects and acts as immune modulator. Currently there is no approved immunotherapy available for BC patients, early data from several ongoing clinical trials show activity in various subclasses of BC including TNBC. It is reported that PD-L1 is high in 20% of TNBCs and COX2 may be involved in its regulation in tumor-infiltrating myeloid cells. This led us to hypothesize that modulation of inflammation associated biomarkers in the TME would increase the efficacy of RT and immunotherapy by amplifying anti-tumor immunity.

Methods

We investigated the effect of nitric oxide synthase2 (NOS2) or cycloxygenase2 (COX2) inhibition using commercially available inhibitor on radiation and αPD-L1 induced tumor growth delay and lung metastases in murine model of TNBC using 4T1 cell line implanted in flank of Balbc mice. Change in immune cell populations in the TME was investigated using confocal microscopy, CO-Detection by indEXing (CODEX) technology and flow-cytometry. TME associated metabolic changes was investigated using seahorse bioanalyzer and LC-MS/MS. We also measured the levels of inflammation associated cytokines in serum using LEGENDplex.

Results

We found that RT induced inflammation associated biomarkers NOS2 and COX2 in the TME, specifically in the tumor cells. We have shown that co-expression of pro-inflammatory enzymes NOS2 and COX2 is a powerful prognostic indicator of poor outcome (HR=21) among ER– patients which in turn drive major oncogenic pathways. We demonstrated that co-treatment with COX2 inhibitor led to tumor growth delay and reduced metastases. COX2 inhibition along with RT induced proliferation of cytotoxic CD8, dendritic cells and pro-inflammatory macrophages while Treg activity was reduced. Furthermore, CODEX showed correlation between decreased macrophage population and increased CD8 infiltration suggesting role of macrophages in inhibiting CD8 infiltration.

Conclusions

We demonstrated that co-treatment of COX2 inhibitor with radiation and immunotherapy effectively activated the immune system by changing the TME to support tumor clearance.

Correspondence: Sandra Demaria (szd3005@med.cornell.edu)

Background

We have shown in both murine tumor models and patients that focal tumor radiotherapy synergizes with CTLA-4 blockade to induce systemically effective anti-tumor immunity in otherwise resistant tumors [1,2,3]. Mechanistically, radiation-induced viral mimicry is critical for priming of anti-tumor CD8+ T cells, and is triggered by cytosolic DNA that stimulates cancer-cell intrinsic interferon type I release [2,3]. The exonuclease TREX1 abrogates this response by clearing cytosolic DNA thus acting as a central regulator of RT-induced anti-tumor immunity and we demonstrated that Trex1 gene expression is upregulated by radiation doses of >12Gy per fractions in most carcinoma cells [2]. Thus, we investigated the mechanisms responsible for inducing Trex1 expression in irradiated cancer cells.

Methods

The mouse mammary carcinoma TSA cells were irradiated in vitro with 8 or 20 Gy, followed by chromatin immunoprecipitation (ChIP) assay to determine if the transcription factor cFos was bound to Trex1 promoter region 4 hours after irradiation. A doxycycline-inducible shRNA was used to knockdown cFos expression in TSA cells.

Results

The transcription factor c-Fos was found to be phosphorylated in TSA cells treated with a radiation dose that induces TREX1 upregulation (20Gy), but not in 8Gy-treated cells. c-Fos has been previously implicated in the induction of Trex1 in response to UV irradiation [4]. Therefore, it was a good candidate as a regulator of Trex1 in 20 Gy-treated TSA cells. Consistently, a ChIP assay demonstrated that 20 Gy but not 8Gy irradiation induced cFos binding to Trex1 promoter. shRNA-mediated c-Fos knockdown abrogated Trex1 upregulation induced by after 20 Gy and restored cytosolic DNA accumulation and type I interferon pathway activation.

Conclusions

Overall, these data identify cFos as the transcription factor that regulates Trex1 gene expression in a radiation dose-dependent way. We are currently investigating the signals upstream of cFos. Results of this work will identify potential actionable targets to increase radiation-induced anti-tumor immunity.

Acknowledgements

Grant support: NIH R01CA201246

References

1. Demaria et al., 2005 Immune-mediated inhibition of metastases after treatment with local radiation and CTLA-4 blockade in a mouse model of breast cancer. Clin Cancer Res.

2. Vanpouille-Box, C. et al. 2017. DNA exonuclease Trex1 regulates radiotherapy-induced tumour immunogenicity. Nat Commun

3. Formenti, S. C. et al. 2018. Radiotherapy induces responses of lung cancer to CTLA-4 blockade. Nat Med 24: 1845-1851.

4. Christmann M. et al.,2010 Three prime exonuclease I (TREX1) is Fos/AP-1 regulated by genotoxic stress and protects against ultraviolet light and benzo(a)pyrene-induced DNA damage. Nucleic Acids Res. 38:6418-32.

Correspondence: James Welsh (jwelsh@mdanderson.org)

Background

Radiotherapy has been routinely used to treat lung cancer alone or in combination with chemotherapy or surgery. Despite the tremendous benefits of radiotherapy in localized tumor control, it is not very effective for treating metastasized lung cancers. In addition, the application of radiotherapy is limited by its damage to the surrounding healthy tissues.

Methods

In this study, we combined radiotherapy with checkpoint inhibitor immunotherapy and a radio-enhancing nanoparticle NBTXR3 (R3) to extend the application of radiotherapy to treating metastasized lung cancer in mice.

Results

In both the 344SQ-anti-PD-1 sensitive and 344SQ-anti-PD-1 resistant lung cancer models, we found that the combination of anti-PD1 and radiotherapy-induced robust immune responses, which resulted in significantly better control of both the local tumor (implanted on the right leg of the mice) and the metastasis (implanted on the left leg of the mice) than either immunotherapy or radiotherapy alone. In addition, we found that R3 significantly enhanced the efficacy of combined radiotherapy and immunotherapy in the two lung tumor models to control the local tumor and metastasis. The results also showed that with the supplement of R3, radiation plus anti-PD-1 achieved similar control of the metastasis under 3x8Gy compared to 3X12Gy, indicating that the treatment benefits acquired at a high dosage of radiation can now be achieved at lower dosages with the presence of R3. We also found that the triple combination of R3, radiotherapy, and anti-PD-1 significantly reduced the number of metastases in lungs compared to other combinations of these therapies. The immunoprofiling of the tumors after treatment showed that triple combination therapy reduced the population of regulatory T cells and M2 macrophages, resulting in favorable conditions for immune-mediated cancer cell killing.

Conclusions

In conclusion, R3 considerably enhances the the ability of the combination of radiotherapy and immunotherapy to control both the local tumor and the metastasis in both the 344SQ-anti-PD-1 sensitive and 344SQ-anti-PD-1 resistant lung cancer models. In addition, R3 allows a lower dosage of radiation to achieve similar treatment outcome compared to a higher dosage of radiation.

Correspondence: Marcin Kortylewski (mkortylewski@coh.org)

Background

Therapeutic and immunogenic effects of radiation therapy (RT) are often constrained by tolerogenic activity of the tumor microenvironment. We recently found that combined modality therapy (CMT), using RT with concurrent cisplatin or cetuximab treatment, triggers potentially tolerogenic STAT3 signaling in circulating myeloid immune cells in patients with stage III-IV head and neck squamous cell carcinoma (HNSCC). Here, we extended these studies to tumor-containing lymph nodes biopsies collected from the same patients before and after two-week CMT. The clinical observations were later verified in feasibility studies testing the effect of STAT3 targeting on sensitivity to radiation therapy in HPV-positive and HPV-negative HNSCC tumor models in mice.

Methods

The STAT3 activity in immune populations infiltrating HNSCC tumors was analyzed in patients' derived fixed or frozen fine-needle biopsies. We used the multicolor immunofluorescent analysis and also gene expression analysis using Nanostring assay. To assess the effect of targeting STAT3 in the identified tolerogenic myeloid cell populations, we used TLR9-targeted STAT3 antisense oligonucleotides (CpG-STAT3ASO) in combination with local tumor irradiation.

Results

Using the multicolor immunofluorescent analysis, we found a significant increase in the percentage of CD163+ M2 macrophages with activated STAT3 and a concurrent reduction in a number of CD8+ T-cells in tumors post-treatment. Similarly, the Nanostring analysis of frozen tumor-containing lymph nodes biopsies detected significant increase in the percentage M2 macrophages (CD163, CD206), with the upregulation of Th2 (IL4, IL5) and immunosuppressive regulators (ARG1, IL10, LIF, TGFB1). These results suggested that STAT3 activation in tumor-associated myeloid cells can dampen immunogenic effects of CMT. To test this hypothesis, we combined single-dose localized RT with intratumoral injections of myeloid cell-specific STAT3 antisense oligonucleotide (CpG-STAT3ASO) against three mouse models of HPV-positive (mEERL) or HPV-negative (MOC2 and mEERL/sh) head and neck cancers. The combination of both strategies resulted in a strong synergistic effect with frequent eradication of mEERL and MOC2 tumors in at least half of treated mice. Regardless of the HPV status, effects of RT/CpG-STAT3ASO therapy were associated with reduced numbers of tumor-associated M2 macrophages and increased percentage of CD8+ T-cells and/or CD8:Treg ratio in MOC2 and mEERL tumors. In addition, we observed recruitment of M1 macrophages with enhanced expression of MHC class II and costimulatory molecules in tumor-draining lymph nodes after RT/CpG-STAT3ASO treatments.

Conclusions

Our clinical observations and animal study results suggest that myeloid cell-targeted inhibition of STAT3 together with CpG-mediated immunostimulation dramatically augment the outcome of radiation therapy against HPV-positive and HPV-negative HNSCC.

Ethics Approval

The clinical protocol including the relevant informed consent form was approved by the institutional review board at City of Hope (IRB-14255) and the study was conducted in accordance with the amended Declaration of Helsinki and the International Conference on Harmonization Guidelines.

Correspondence: Maryland Franklin (mfranklin@mibioresearch.com)

Background

Despite the success of immune checkpoint blockade in treating many types of cancers, there remain many that are insensitive to these treatments. We have previously shown that focal radiation can sensitize the immunologically “cold” 4T1 tumor model to immune checkpoint inhibition. Combination treatment with radiation and anti-CTLA4 antibody resulted in decreased numbers of infiltrating regulatory T cells as well as B cells in the tumors. While numbers of CD8+ T cells were not increased in tumors that received combination treatment, the cells displayed a more activated phenotype. We hypothesized that combination treatment decreased numbers of regulatory cell types and led to increased cytotoxic activity of CD8+ T cells, Natural Killer (NK) cells, and Natural Killer T (NKT) cells in the tumor microenvironment.

Methods

We orthotopically implanted 4T1 mammary carcinoma cells into BALB/c mice and treated with anti-mCTLA4 antibody, focal radiation, or both. Tumors were analyzed by flow cytometry, utilizing the MI-CompLeukocyte™ panels with absolute cell counting, as well as the MI-T Effector/Memory™ panel and a custom intracellular cytokine panel.

Results

The 4T1 tumor model is typically unresponsive to immune checkpoint inhibition. As expected, we saw a synergistic effect of combination treatment on tumor burden compared with radiation alone (77.1% vs. 42.3% TGI, respectively), suggesting that radiation primes the tumors to be more responsive to checkpoint inhibition. Our analysis showed that there was a 2.4-fold increase in the percentage of CD8+ T effector memory cells with a corresponding 9.4-fold decrease in CD8+ T central memory cells. The percentage of T effector memory cells was inversely correlated with tumor volume in the combination group, suggesting that this cell type could be directly responsible for the decreased tumor burden. Consistent with the increased effector memory phenotype, the percentage of CD8+ T cells secreting Granzyme B was increased 2-fold with combination treatment compared to radiation alone. Combination treatment also increased the percentage of NKT cells that secreted Interferon-γ, suggesting that the increased inflammatory cell phenotype is a broader phenomenon and not specific to CD8+ T cells.

Conclusions

Taken together, these data suggest that synergistic treatment with anti-mCTLA4 and focal radiation can induce the tumor immune microenvironment to change to a more inflammatory, and less suppressive, state than either treatment alone. Reduction in numbers of regulatory T cells and B cells resulted in increased percentages CD8+ T cells and NKT cells with a phenotype consistent with improved anti-tumor activity.

Ethics Approval

This study was approved by the IACUC at MI Bioresearch under Protocol MI05 - Subcutaneous Tumor Models for Evaluation of Test Agents.

Correspondence: Zachary Morris (zmorris@humonc.wisc.edu)

Background

Radiation therapy (RT) to enhances the efficacy of immune checkpoint inhibition (ICI) in multiple immunologically “cold” preclinical tumor models[1,2]. However, the optimal approach to delivering immunostimulatory RT has not been established. We have developed a novel strategy to deliver immunostimulatory low dose (LD) RT to all tumor sites in the setting of metastatic disease by using targeted radionuclide therapy (TRT), with a tumor-selective alkyl-phosphocholine (90Y-NM600). Here we evaluate the mechanistic basis for an observed cooperative interaction between LDTRT and ICI in an immunologically cold murine melanoma model that does not respond to ICI alone.

Methods

Mice were engrafted with B78 melanoma tumors, 80-90 mm3, and were randomized to the following treatment groups: vehicle only (VO), anti-CTLA4 (C4), 90Y-NM600 TRT (2.5 Gy tumor dose), external beam RT (EBRT, 2.5 or 12 Gy), or combination TRT+C4. To compare LDRT delivered via EBRT or TRT or to high dose EBRT, tumors were harvested 1, 7, and 14 days after RT. We performed flow cytometry (FC) and qPCR on these specimens to evaluate immune cell infiltrates and gene expression. To determine treatment effects, tumors treated with VO, C4, TRT, or TRT+C4 were harvested on Day 25 after initiation of therapy. Immune infiltrates were quantified by FC in these samples, cytokine levels via multiplex ELISA, and T cell clonal analysis via T cell receptor (TCR) sequencing.

Results

Comparison of RT dose effects demonstrated increased ratios of CD8+ T cells to CD4+FoxP3+ regulatory T cells in the TRT group, compared to the EBRT groups at Day 1, as well as increased myeloid cell (CD11b+) and NK cell infiltrate with TRT. Gene expression analysis showed peak apoptosis markers on Day 1, peak type I interferon signaling (ISG) and chemokine expression on Day 7. High dose EBRT had significantly greater ISG expression than LD-RT (p <0.05). Following combined treatment with TRT + C4, tumors showed significantly greater CD8+ T cells and γδ T cells, compared to single agent or untreated control groups. Multiplex ELISA showed increased concentrations of multiple cytokines with the notable exception of IL-10 in tumors from mice receiving TRT+C4 (Figure 1). In contrast to prior studies with high dose EBRT, LDTRT stimulated increased TCR clonal expansion but did not increase TCR diversity when combined with C4 ICI.

Conclusions

Our results demonstrate a cooperative therapeutic interaction between TRT+ICI in a murine tumor model that is functionally immunologically cold and does not respond to ICI alone.

Acknowledgements

Research material support for this work was provided by Bristol Myers Squibb. This work was supported by the National Institute of Health under the following awards U01CA233102, the University of Wisconsin Carbone Cancer Center Support Grant P30 CA014520, RSNA Fellow Award, ASCO YIA, and the Bentson Translational Research Fellowship.

References

1. Twyman-saint victor C, Rech AJ, Maity A, et al. Radiation and dual checkpoint blockade activate non-redundant immune mechanisms in cancer. Nature. 2015;520(7547):373-7.

2. Patel, RB, Czapar, AE, Fiering, S et al. Radiation therapy combined with cowpea mosaic virus nanoparticle in situ vaccination initiates immune-mediated tumor regression. ACS Omega 2018. 3(4):3702-3707

Ethics Approval

The study was approved by the University of Wisconsin institutional animal care and use committee protocol 005670.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P511).

TRT + ICI turns an immunological “cold” tumor “hot”

Correspondence: Gabriel Helmlinger (gabriel.helmlinger@astrazeneca.com)

Background

Combination of radiation with immuno-therapy had shown promising results in preclinical experiments as well as in clinical trials. Information on clinical efficacy of immune checkpoint inhibitors (ICI) combined with radiotherapy (RT) is available from numerous retrospective analyses, mainly focused on the treatment of brain metastases. The interpretation of retrospective clinical meta data, however, may be complicated due to variability in the design of studies and confounding factors. The aim of the present research was to systematically review and perform a meta-analysis comparing efficacy profiles of mono-RT vs ICI-RT combination in the treatment of brain metastases. The analysis also aimed at identifying factors affecting treatment outcomes.

Methods

The meta-analysis was performed using the Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) guideline. Studies with at least one ICI-RT arm and including patients with brain metastases were curated via a systematic literature search. Information on 1-year overall survival (OS) and 1-year local control (LC) was extracted and analyzed; additionally, if studies included mono-RT arms, risk ratios (RR) for the aforementioned endpoints were calculated and analyzed given mono-RT as an active control. Random-effects meta-regression models were tested to evaluate the impact of different factors, such as combination treatment sequencing or type of ICI and RT on outcomes.

Results

In total, information on OS and LC was reported in, respectively, 26 and 6 studies, with 50 and 13 arms, featuring 2528 and 525 patients. The corresponding information for mono-RT arms was available from, respectively, 12 and 4 studies with 2021 and 463 patients enrolled. Higher 1-year was observed in ICI-RT treatments vs mono-RT, with corresponding incidence rates were 59% [95% CI: 54-64%] vs 32% [95% CI: 25-40%] (P

Concurrent treatment was associated with higher 1-year OS comparing to sequential regimen with incidence rates 69% [95% CI: 60-78%] vs 52% [95% CI: 45-58%] (P=0.039); efficacy of ICI-RT treatment was not affected by ICI and RT type.

Conclusions

RT and ICI combinations were associated with improved 1-year OS vs mono-RT. The highest benefit in outcomes was determined for cohorts with concurrent ICI-RT treatment regimen.

Correspondence: Lorenzo Galluzzi (log3001@med.cornell.edu)

Background

Hormone receptor+ (HR+) breast cancer (BC) is the most frequent cause of BC death [1], and immunotherapy with immune checkpoint blockers (ICBs) in HR+ BC patients has been disappointing [2]. Conversely, recent results from MONALEESA 7 trial demonstrate a robust survival advantage for the combination of cyclin dependent kinase 4 (CDK4)/CDK6 inhibitors and hormonotherapy over hormonotherapy alone [3]. However, not all patients benefit from hormonotherapy + CDK4/CDK6 inhibition [4]. This implies that additional strategies are needed to improve disease outcome in this patient population. In this context, focal radiation therapy (RT) stands out as a promising therapeutic partner for several reasons, including (1) both RT and CDK4/CDK6 inhibitors have cytostatic/cytotoxic effects; and (2) both RT and CDK4/CDK6 inhibitors can have immunostimulatory effects [5,6]. We therefore decided to investigate the potential synergy between RT, and CDK4/CDK6 inhibitors, with a particular focus on administration schedule, ultimately aiming to inform the design of a clinical trial testing the combination of these therapeutic agents in patients with HR+HER2- BC.

Methods

The synergy between RT and palbociclib (a highly selective CDK4/CDK6 inhibitor approved for the treatment of BC in humans) was tested: (1) in vitro, in 2 human BC cell lines with differential sensitivity to palbociclib and RT (MCF7 and MBA-MD-231 cells), to evaluate the effects on cell cycle distribution, cell death, cell senescence and IFN secretion; (2) in vivo, in a unique endogenous model of mammary carcinogenesis that recapitulate multiple immunobiological features of human HR+HER2- BC.

Results

We identified the lowest doses of RT and palbociclib that mediate robust short-term cytostatic/cytotoxic effects in MDA-MB-231 and MCF7 cells. Optimized dose combinations and sequencing experiments pointed to RT followed by palbociclib as the approach with superior therapeutic potential. This largely reflect the ability of palbociclib to block in G1 cells escaping the G2/M arrest mediated by RT. Both RT and palbociclib were active in endogenous carcinomas driven in mice by slow-release progesterone pellets combined with an oral carcinogen, but failed to mediate disease eradication. Ongoing experiments are elucidating the impact of combination and sequencing in vivo.

Conclusions

Our preliminary results suggest that RT should precede CDK4/6 inhibition to achieve superior disease control. Altogether, these experiments will determine an optimal approach to combine CDK4/CDK6 inhibitors with RT in the context of endocrine therapy that has an elevated potential for clinical translation.

Acknowledgements

GP is supported by Fondazione Umberto Veronesi. ABM supported by a Breakthrough Level 2 grant from the US Department of Defense (DoD), Breast Cancer Research Program (BRCP) [#BC180476P1. LG is supported by a Breakthrough Level 2 grant from the US Department of Defense (DoD), Breast Cancer Research Program (BRCP) [#BC180476P1], by a startup grant from the Dept. of Radiation Oncology at Weill Cornell Medicine (New York, US), by industrial collaborations with Lytix (Oslo, Norway) and Phosplatin (New York, US), and by donations from Phosplatin (New York, US), the Luke Heller TECPR2 Foundation (Boston, US) and Sotio a.s. (Prague, Czech Republic).

References

1. Pan H, Gray R, Braybrooke J, Davies C, Taylor C, McGale P, Peto R, Pritchard KI, Bergh J, Dowsett M, Hayes DF; EBCTCG, 20-Year Risks of Breast-Cancer Recurrence after Stopping Endocrine Therapy at 5 Years. N Engl J Med. 2017; 377(19):1836-1846.

2. Rugo HS, Cortes J, Awada A, O'Shaughnessy J, Twelves C, Im SA, Hannah A, Lu L, Sy S, Caygill K, Zajchowski DA, Davis DW, Tagliaferri M, Hoch U, Perez EA, Change in Topoisomerase 1-Positive Circulating Tumor Cells Affects Overall Survival in Patients with Advanced Breast Cancer after Treatment with Etirinotecan Pegol. Clin Cancer Res. 2018;24(14):3348-3357.

3. Im SA, Lu YS, Bardia A, Harbeck N, Colleoni M, Franke F, Chow L, Sohn J, Lee KS, Campos-Gomez S, Villanueva-Vazquez R, Jung KH, Chakravartty A, Hughes G, Gounaris I, Rodriguez-Lorenc K, Taran T, Hurvitz S, Tripathy D, Overall Survival with Ribociclib plus Endocrine Therapy in Breast Cancer. N Engl J Med. 2019;381(4):307-316.

4. Pandey K, An HJ, Kim SK, Lee SA, Kim S, Lim SM, Kim GM, Sohn J, Moon YW, Molecular mechanisms of resistance to CDK4/6 inhibitors in breast cancer: A review. Int J Cancer. 2019;145(5):1179-1188.

5. Ameratunga M, Kipps E, Okines AFC, Lopez JS, To Cycle or Fight-CDK4/6 Inhibitors at the Crossroads of Anticancer Immunity. Clin Cancer Res. 2019;25(1):21-28.

6. Sahu AD, S Lee J, Wang Z, Zhang G, Iglesias-Bartolome R, Tian T, Wei Z, Miao B, Nair NU, Ponomarova O, Friedman AA, Amzallag A, Moll T, Kasumova G, Greninger P, Egan RK, Damon LJ, Frederick DT, Jerby-Arnon L, Wagner A, Cheng K, Park SG, Robinson W, Gardner K, Boland G, Hannenhalli S, Herlyn M, Benes C, Flaherty K, Luo J, Gutkind JS, Ruppin E, Genome-wide prediction of synthetic rescue mediators of resistance to targeted and immunotherapy. Mol Syst Biol. 2019;15(3):e8323.

Ethics Approval

The animal study was approved by Weill Cornell Medicine's IACUC; approval number 2019-0022.

Correspondence: Douglas McNeel (dm3@medicine.wisc.edu)

Background

Prostate cancer responds poorly to checkpoint blockade due to its immunologically “cold” microenvironment with low CD8+ T cell infiltration along with relatively high numbers of myeloid-derived suppressor cells (MDSCs) and regulatory CD4+ T cells (Tregs) [1–3]. Our lab has studied DNA vaccines as a simple method of increasing antigen-specific CD8+ T cell infiltration [4,5]. To further increase CD8+ T cell infiltration and deplete suppressive populations, external beam radiation (EBRT) can be used [6]. However, it is often infeasible to irradiate all metastases using EBRT. Molecularly targeted radionuclide therapy (MTRT) is a method of delivering radioisotopes radiation selectively to all sites of disease [7]. The effects of MTRT’s effects on the tumor immune microenvironment are not yet well described. In this study, we examined the effect of 90Y-NM600 MTRT on immune populations within murine prostate tumors and assessed whether MTRT combined with antigen-specific vaccination can improve anti-tumor response.

Methods

6-week old male FVB mice (n=3 per group) were implanted with subcutaneous (s.c.) MyC-CaP murine prostate tumors then given a single intravenous injection of either 50 (3.1 Gy) or 250 (15.5 Gy) μCi of 90Y-NM600. Mice were euthanized at day 0, 1, 4, 7, 14, and 21 post MTRT administration. Tumors were harvested and examined by multiparametric flow cytometry and immunohistochemistry for effects on immune populations. In a second study, mice (n=5 per group) were implanted with s.c. MyC-CaP tumors then given five weekly treatments of either DNA vaccine or vehicle as well as a single administration of 125 μCi (7.8 Gy) of 90Y-NM600. These mice were followed for tumor growth.

Results

CD8+ T cell tumor infiltration increased by 13% at Day 14 compared to Day 0 (p=.039) following MTRT administration in both dose groups while GR-1+ cells (including MDSCs) transiently decreased by 31% at Day 4 (p=.007). Recovering CD8+ T cells showed high expression of the checkpoints PD-1, CTLA-4, and LAG-3. CD4+ T cells were depleted following MTRT administration and did not recover by Day 21. MTRT combined with vaccination resulted in decreased tumor growth compared to MTRT alone.

Conclusions

Our data indicates that MTRT disrupts the suppressive tumor microenvironment of prostate tumors by depleting suppressive populations and increasing CD8+ T cell infiltration. High expression of exhaustion markers on CD8+ T cells suggests that combination of MTRT and vaccination with checkpoint blockade may further improve anti-tumor response.

Acknowledgements

UW-Madison Medical Scientist Training Program: GM008692

UW-Madison Institute for Clinical and Translational Research Predoctoral TL1 Program: TR002375

University of Wisconsin Carbone Cancer Center Support Grant P30 CA014520

Archeus Technologies for kindly providing the NM600

References

1. Danaher P, Warren S, Lu R, Samayoa J, Sullivan A, Pekker I, et al. Pan-cancer adaptive immune resistance as defined by the Tumor Inflammation Signature (TIS): results from The Cancer Genome Atlas (TCGA). J Immunother Cancer 2018;6:63. doi:10.1186/s40425-018-0367-1.

2. Parker KH, Beury DW, Ostrand-Rosenberg S. Myeloid-Derived Suppressor Cells: Critical Cells Driving Immune Suppression in the Tumor Microenvironment. Adv Cancer Res 2015;128:95–139. doi:10.1016/bs.acr.2015.04.002.

3. Flammiger A, Weisbach L, Huland H, Tennstedt P, Simon R, Minner S, et al. High tissue density of FOXP3+ T cells is associated with clinical outcome in prostate cancer. Eur J Cancer Oxf Engl 1990 2013;49:1273–9. doi:10.1016/j.ejca.2012.11.035.

4. Olson BM, Johnson LE, McNeel DG. The androgen receptor: a biologically relevant vaccine target for the treatment of prostate cancer. Cancer Immunol Immunother CII 2013;62:585–96. doi:10.1007/s00262-012-1363-9.

5. Olson BM, Gamat M, Seliski J, Sawicki T, Jeffery J, Ellis L, et al. Prostate Cancer Cells Express More Androgen Receptor (AR) Following Androgen Deprivation, Improving Recognition by AR-Specific T Cells. Cancer Immunol Res 2017;5:1074–85. doi:10.1158/2326-6066.CIR-16-0390.

6. Demaria S, Bhardwaj N, McBride WH, Formenti SC. Combining radiotherapy and immunotherapy: a revived partnership. Int J Radiat Oncol Biol Phys 2005;63:655–66. doi:10.1016/j.ijrobp.2005.06.032.

7. Grudzinski JJ, Hernandez R, Marsh I, Patel R, Aluicio-Sarduy E, Engle J, et al. Preclinical Characterization of 86/90Y-NM600 in a variety of murine and human cancer tumor models. J Nucl Med 2019:jnumed.118.224808. doi:10.2967/jnumed.118.224808.

Correspondence: Sandra Demaria (szd3005@med.cornell.edu)

Background

Focal radiotherapy (RT) in combination with CTLA-4 blockade activates T-cells that reject tumors in both mice and patients unresponsive to CTLA-4 blockade alone [1, 2]. In metastatic non-small cell lung cancer patients dynamic changes in the blood TCR repertoire were strongly associated with response [1]. In the 4T1 mouse model of triple negative breast cancer we found higher intratumoral CD8+/CD4+ ratio and CD8+ T cell clonality in tumors of mice treated with RT+CTLA-4 blockade compared to each treatment alone [3]. However, the functional state of expanded intratumoral T cells remained unknown. The purpose of this study was to identify the intratumoral T cell landscape associated with response to RT+CTLA-4 blockade [4].

Methods

For single cell sequencing (SC-seq) and bulk TCR sequencing (TCR-seq) experiments, mice were inoculated in one or both flanks with 50,000 4T1 cells, respectively. To obtain pre-treatment samples in each individual mouse, when applicable, one tumor was removed 2 days before treatment with RT (3x8 Gy) and/or anti-CTLA-4 antibody (9H10 Ab, 3x200 μg i.p.), and the treated tumor was resected 1 day after last antibody administration. TCR-seq was performed using the iRepertoire platform, and 5’ gene expression, VDJ, and feature barcoding SC-seq was performed on sorted CD3+CD4+ and CD3+CD8+ T cells using the 10X Genomics chromium system.

Results

In tumors treated with RT alone or in combination with anti-CTLA-4, both the TCRα and TCRβ repertoires increased in clonality, whereas a smaller increase in TCRβ clonality was found after anti-CTLA-4 monotherapy. Increased divergence between pre- and post-treatment samples was found after RT alone or in combination with anti-CTLA-4 but only correlated with treatment effect in mice treated with the combination (Pearson correlation coefficients: TCRα = 0.85, TCRβ = 0.93). Compared to untreated tumors, tumors treated with RT+anti-CTLA-4 were enriched with an oligoclonal population of Ifng+Cd69+Tnf+Cd8+ and Ifng+Cd4+ T cells, and had decreased proportions of both Foxp3+Ox40+Cd25+Cd4+ (Tregs) and Pd1+Gzmb+Cd8+ (likely terminally differentiated/exhausted) T cells. In contrast, tumors treated with RT alone were enriched with T cells exhibiting the terminally differentiated/exhausted phenotype, which may explain why the TCR repertoire divergence between baseline and post-treatment tumors correlated with treatment effect only for mice treated with RT+anti-CTLA-4.

Conclusions

Overall, these results suggest that RT promotes the differentiation of the intratumoral TCR repertoire but in the absence of CTLA-4 blockade it does not drive the expansion and functional differentiation of T cell phenotypes necessary for immune-mediated tumor rejection.

Acknowledgements

Grant support: R01CA198533

References

1. Formenti SC, Rudqvist NP, Golden E, Cooper B, Wennerberg E, Lhuillier C, et al. Radiotherapy induces responses of lung cancer to CTLA-4 blockade. Nat Med. 2018;24(12):1845-51.

2. Vanpouille-Box C, Demaria S, Formenti SC, Galluzzi L. Cytosolic DNA Sensing in Organismal Tumor Control. Cancer Cell. 2018;34(3):361-78.

3. Rudqvist NP, Pilones KA, Lhuillier C, Wennerberg E, Sidhom JW, Emerson RO, et al. Radiotherapy and CTLA-4. Blockade Shape the TCR Repertoire of Tumor-Infiltrating T Cells. Cancer Immunol Res. 2018;6(2):139-50.

Correspondence: Sandra Demaria (szd3005@med.cornell.edu)

Background

Radiation therapy (RT) used at immunogenic doses (8GyX3) leads to cytosolic DNA accumulation which activates interferon type I (IFN-I) production via cGAS/STING pathway in cancer cells [1-3]. RT-induced IFN-I is critical for activation of systemically effective anti-tumor immune responses in combination with immune-checkpoint blockade therapy [1]. We have recently demonstrated that tumor-derived exosomes (TEX) produced by irradiated (8GyX3) TSA murine carcinoma cells (RT-TEX) contain more dsDNA compared to TEX derived from untreated cancer cells (UT-TEX). Interestingly, the dsDNA transferred by RT-TEX but not UT-TEX induces upregulation of costimulatory molecules CD40, CD80 and CD86, and STING-dependent IFN-I pathway activation in recipient dendritic cells. Furthermore, in vivo, RT-TEX elicit tumor-specific CD8+T-cell responses and protect mice from tumor development [4]. Here, we tested the hypothesis that not only quantitative but also qualitative differences may exist between the dsDNA cargo of RT-TEX and UT-TEX.

Methods

TEX were isolated 48 hours after last dose of radiation or mock treatment using ultracentrifugation. Internal DNA from TEX and from the parent TSA cells cytosolic fraction was analyzed by (1) semi-quantitative PCR using primers for mitochondrial 12S and 16S ribosomal subunits, NADH dehydrogenase subunit 1 (ND1) and genomic 18S ribosomal subunit genes; (2) DNA length was determined using Agilent Bioanalyzer; (3) DNA sequencing was performed using the Illumina HiSeq4000 system (paired end 50).

Results

Genomic and mitochondrial DNA were identified in both TEX and TSA cytosolic fraction regardless of RT. DNA length analysis demonstrated an enrichment of DNA fragments between 60 and 250 bp in RT-TEX compared to UT-TEX, and in the cytosolic fraction of irradiated TSA cells. These results are especially intriguing in light of the recent report that cGAS is optimally activated by dsDNA in this length range [5]. Whole-genome sequencing is ongoing to determine the genomic versus mitochondrial origin of the differentially represented DNA fragments.

Conclusions

Our data suggest that the immunostimulatory DNA cargo of RT-TEX has unique features that reflect molecular changes occurring in radiation-treated cells. The impact of these differences on the DNA-mediated activation of IFN-I pathway in innate immune cells that uptake TEX is under investigation. Identification of a “DNA signature” associated with TEX ability to activate cGAS could provide a blood-based biomarker for the immunogenic tumor response to radiotherapy.

Acknowledgements

Supported by American-Italian Cancer Foundation Fellowship, The Chemotherapy Foundation, R01CA201246 and Cornell University Seed Grants for Intercampus Collaborative Projects

References

1.Vanpouille-Box C, Alard A, Aryankalayil MJ, Sarfraz Y, Diamond JM, Schneider RJ, Inghirami G, Coleman CN, Formenti SC, Demaria S. DNA exonuclease Trex1 regulates radiotherapy-induced tumour immunogenicity. Nat Commun. 2017 Jun 9;8:15618

2.Mackenzie KJ, Carroll P, Martin CA, Murina O, Fluteau A, Simpson DJ, Olova N, Sutcliffe H, Rainger JK, Leitch A, Osborn RT, Wheeler AP, Nowotny M, Gilbert N, Chandra T, Reijns MAM, Jackson AP. cGAS surveillance of micronuclei links genome instability to innate immunity. Nature. 2017 Aug 24;548(7668):461-465

3.Harding SM, Benci JL, Irianto J, Discher DE, Minn AJ, Greenberg RA. Mitotic progression following DNA damage enables pattern recognition within micronuclei. Nature. 2017 Aug 24;548(7668):466-470

4.Diamond JM, Vanpouille-Box C, Spada S, Rudqvist NP, Chapman JR, Ueberheide BM, Pilones KA, Sarfraz Y, Formenti SC, Demaria S. Exosomes Shuttle TREX1-Sensitive IFN-Stimulatory dsDNA from Irradiated Cancer Cells to DCs. Cancer Immunol Res. 2018 Aug;6(8):910-920

5.Du M, Chen ZJ. DNA-induced liquid phase condensation of cGAS activates innate immune signaling. Science. 2018 Aug 17;361(6403):704-709. doi:10.1126/science.aat1022

Correspondence: Thomas Andresen (tlan@dtu.dk)

Background

TLR7 agonists are potent immune activators that are highly attractive for anti-cancer immunotherapy with potential to act synergistically with other cancer treatments such as vaccines, radiotherapy, and chemotherapy. The limited tolerability of systemically administered TLR7 agonists makes administration problematic at therapeutically active concentrations. Drug delivery systems may increase tolerability by optimizing biodistribution and cellular uptake thereby allowing for administration of doses that are therapeutic at the target site.

The current study investigated lipid nanoparticles containing a TLR7 agonist (1V270) in combination with local radiotherapy (to induce antigen release) for an in situ vaccination effect in murine cancer models.

Methods

Biophysical properties of the nanoparticles were characterized and anti-tumor efficacy was investigated in three murine syngeneic subcutaneous cancer models (CT26, MC38, and B16-F10). Treatments were initiated when tumors reached a mean size of ~100 mm3. Nanoparticles were given intravenously as multiple doses in combination with fractionated radiotherapy of the tumor region. The optimal formulation and dosing schedule were based on in vivo efficacy studies. Tolerability of the treatment was evaluated based on body weight, hematology, and blood biochemistry as well as cytokine levels in plasma measured 2h and 6h post injection. Antibody responses against the nanoparticles were also investigated. The effect on the tumor microenvironment was evaluated by flow cytometry.

Results

Nanoparticles containing 1V270 displayed a synergistic effect with local radiotherapy and all (8/8) mice carrying CT26 tumors displayed complete tumor rejection. In the MC38 model, tumor growth delay was observed and 3/9 mice demonstrated complete tumor rejection. No anti-tumor effect was observed in the poorly immunogenic B16-F10 model. Immunologic memory against tumor antigens was confirmed by tumor-rechallenge with 80-100% tumor-rejection rate in both CT26 and MC38. Improved tolerability, compared to other lipid nanoparticles, was demonstrated. The tumor microenvironment analysis demonstrated that the TLR7 agonist formulation induced infiltration of a large number of innate immune cells and supported the formation of activated tumor-specific CD8 T cells.

Conclusions

The novel immunotherapeutic nanoparticle formulation induced a potent anti-tumor immune response; resulting in very potent efficacy in vivo against immunogenic tumors. Furthermore, we demonstrated improved tolerability in mice, which will be further investigated in larger animals. These results are highly encouraging for advancing potent immune stimulating therapy into the clinic.

Ethics Approval

The study was approved by the Danish Animal Experiments Inspectorate.

Correspondence: Sandra Demaria (szd3005@med.cornell.edu)

Background

Immune checkpoint inhibition targeting PD-1/PD-L1 has shown promise in breast cancer but is largely limited to triple negative breast cancer (TNBC).[1] This limitation is primarily due to inherently low levels of tumour infiltrating lymphocytes (TILs), particularly in HR+ disease.[2] Focal radiation therapy (RT) has been shown to generate anti-tumour T cells and increase TILs using mouse models of breast cancer (BC).[3,4] Mechanistically, radiation-induced increase in cytosolic DNA leads to activation of cGAS/STING pathway and production of IFN-훃, which is essential for priming of anti-tumour CD8+ T cells.[5] This process is under the control of TREX1, and dependent on RT dose and fractionation.[5] Inter-tumor variability in optimal RT dose for activating the IFN-I pathway and heterogeneity in immunological response amongst BC subtypes highlight the importance of precision use of RT.[6] We are testing the hypothesis that improved in vitro and in vivo assays allow testing of the pro-immunogenic response to radiation in individual BC patients.

Methods

Spheroids were established from human adenocarcinoma cell lines MCF-7 (ER+,PR+,HER2-) and MDA-MB-231 (TNBC) by seeding 1,000 cells in a mixture of 2% Matrigel in DMEM. Patient-derived tumour xenograft (PDTX) was dissociated and organoids formed using 66% Matrigel in media. Single radiation doses between 2Gy and 20Gy as well as fractionated doses of 8Gyx3 were delivered, samples harvested at 24 hours post-treatment and RNA extracted. RT-qPCR was used to interrogate expression of key IFN-I response genes (IFNβ1, MX1, CXCL10, IFNAR1) as well as TREX1.

Results

Spheroids and organoids were successfully established using human cell lines and PDTX, respectively (Figure 1). IFN-I pathway activation by single radiation doses ranging from 6 to 10Gy was detected in MCF-7, MDA-MB-231 and in PDTX, whereas single doses of 15 to 20Gy enhanced the expression of TREX1 and dampened IFN-I activation. Similar results were obtained in an in vivo irradiated PDTX and its in vitro-irradiated organoids. We are currently establishing and testing additional pairs of organoids and PDTX from the same BC tumor to determine the most time and cost-efficient platform.

Conclusions

Developing a platform that utilizes pre-treatment tumor tissue from a patient to determine the functional response to different RT doses and aid in the selection of the best combination treatment with anti-PD-1/PD-L1 is a critical step towards precision use of RT to enhance responses to immunotherapy. Our data suggest that organoids and PDTX could both be used to test the functional response of an individual patient tumor to various RT doses.

Acknowledgements

The authors which to acknowledge financial support from the Breast Cancer Research Foundation (award BCRF-18-053).

References

1. Emens LA, Breast Cancer Immunotherapy: Facts and Hopes. Clin Cancer Res. 2018; 24(3):511-520.

2. Denkert C, von Minckwitz G, Darb-Esfahani S, Lederer B, Heppner B, Weber KE et al. Tumour-infiltrating lymphocytes and prognosis in different subtypes of breast cancer: a pooled analysis of 3771 patients treated with neoadjuvant therapy. Lancet Oncol. 2018;19(1):40-50.

3. Demaria S, Kawashima N, Yang AM, Devitt ML, Babb JS, Allison JP et al. Immune-Mediated Inhibition of Metastases after Treatment with Local Radiation and CTLA-4 Blockade in a Mouse Model of Breast Cancer. Clin Cancer Res. 2005;11(21):728-734.

4. Demaria S, Golden E, Formenti SC. Role of Local Radiation Therapy in Cancer Immunotherapy. JAMA Oncol. 2015;1(9):1325-1332.

5. Vanpouille-Box C, Alard A, Aryankalayil M, Sarfraz Y, Diamond J, Schneider R et al. DNA exonuclease Trex1 regulates radiotherapy-induced tumour immunogenicity. Nat Commun. 2017;8:15618.

6. Vanpouille-Box C, Formenti, SC, Demaria S. Toward Precision Radiotherapy for Use with Immune Checkpoint Blockers. Clin Cancer Res. 2018;24(2):259-265.

Ethics Approval

All animal procedures were approved by the Institutional Animal Care and Use Committee (IACUC) at Weill Cornell Medicine.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P518).

Patient-Derived Breast Cancer Organoids

Correspondence: Thomas Andresen (tlan@dtu.dk)

Background

Polarizing the tumor microenvironment (TME) towards an inflamed and immunosupportive type is highly attractive for anti-cancer therapies. However, critical challenges are recognized for optimally modulating the TME. Firstly, immune activating drugs are poorly tolerated and secondly, the plastic and reactive TME must be continuously manipulated. Moreover, multiple pathways must be stimulated to recruit and activate immune cells. Here we present a well-tolerated multi-drug intratumoral delivery technology (Immunogel), which has the potential to transform the TME and support immunological anti-cancer activity by ensuring a local sustained drug release. The Immunogel can easily be injected with small gauge needle technology, which has been demonstrated in clinical cancer patients using advanced high-precision image-guided technologies.

Methods

The Immunogel was formulated to provide sustained release of a TLR7/8 agonist (NT03-gel) and combined sustained release of a TLR7/8 agonist and TGFβ-inhibitor (NT04C-gel). Therapeutic efficacy and immune-modulating properties were evaluated in syngeneic cancer models. Drug release kinetics and tolerability was demonstrated in preclinical rodents and laboratory canines.

Results

The NT03-gel demonstrated a sustained intratumoral drug release of 31% after 24 hours and 81% after 7 days. Combination of NT03-gel and immunogenic cell death inducing chemo- or radiotherapy in the CT26 cancer model induced tumor rejection in >60% compared to 0% of treatment controls. TME analysis demonstrated that NT03-gel in combination with chemotherapy reduced intratumoral levels of immunosuppressive myeloid subsets and improved the cytotoxic T cell (cT) to regulatory T cell ratio more than nine-fold compared to controls. Additionally, a high increase of strongly activated CD137+ cTs and intratumoral infiltration of central memory cTs was identified. NT03-gel and radiotherapy increased intratumoral levels of T cell attracting cytokines and chemokines. Encouragingly, NT03-gel significantly enhanced anti-PD-1 therapy, indicating that the immune modulating properties of NT03-gel are attractive for improving immune-checkpoint therapy. The potential of multitarget therapy was demonstrated for NT04C-gel, which in combination with radiotherapy induced 100% tumor rejection. Furthermore, 18FDG PET/CT and lung metastasis scoring demonstrated that NT04C-gel in combination with low-dose radiotherapy of the primary tumor significantly reduced lung metastasis in the metastatic 4T1 cancer model. Importantly, the NT03-gel and NT04C-gel were well tolerated in laboratory beagles during a three-week evaluation period.

Conclusions

By enabling intratumoral sustained drug release, the Immunogel has the ability to polarize the TME towards an inflamed immunosupportive state, which is key for improving anti-cancer immunotherapies.

Ethics Approval

All experimental procedures involving animals were approved by The Danish Animal Experiments Inspectorate.

Correspondence: Oluwakayode Adejoro (oluwakayode.adejoro@merck.com)

Background

To date, pembrolizumab plus axitinib (P+A) and nivolumab plus ipilimumab (N+I) are the only immunotherapy combination therapies demonstrating superior overall survival (OS) compared to sunitinib in Phase III clinical trials in subjects diagnosed with advanced/metastatic renal cell carcinoma (mRCC). No head-to-head trial has compared the efficacy of P+A versus N+I in mRCC.

Methods

An anchored matching-adjusted indirect comparison (MAIC) was conducted using individual patient-level data (IPD) from KEYNOTE-426 and pseudo-IPD (digitized KM curves) in the published data from CHECKMATE-214. To adjust for differences in baseline potential effect modifiers between these trials, data for subjects from the KEYNOTE-426 were re-weighted to match the baseline characteristics reported in CHECKMATE-214. Progression-free survival (PFS) and OS were compared. The analyses were conducted in the intention-to-treat (ITT) populations and repeated in subjects with intermediate and poor risk disease (by International Metastatic Renal Cell Carcinoma Database Consortium [IMDC] criteria).

Results

After matching, P+A resulted in a lower hazard of disease progression or death than N+I (HR=0.81; 95% CI 0.63, 1.03; p-value = 0.090), and a lower hazard of death (HR=0.71; 95% CI 0.49, 1.03; p-value = 0.069) than N+I in the ITT populations. In the intermediate and poor risk populations, P+A resulted in a lower hazard of disease progression or death (HR=0.84; 95% CI 0.63, 1.12; p-value = 0.238) and a lower hazard of death (HR=0.75; 95% CI 0.50, 1.11; p-value = 0.148) than N+I.

Conclusions

Following the adjustment of cross-trial differences, P+A demonstrated lower hazards in the PFS and OS compared to N+I in both the ITT, and intermediate and poor risk populations. In the absence of direct comparison and sufficiently powered studies, these analyses provide valuable insights on the relative efficacy of P+A versus N+I for patients, physicians, and payers.

Correspondence: Howard Kaufman (Howard.Kaufman@replimune.com)

Background

Talimogene laherparepvec (T-VEC) is an oncolytic HSV-1 approved for the treatment of melanoma. We previously showed MEK inhibition enhances T-VEC-mediated immunogenic cell death in human melanoma cell lines in a BRAF mutation-independent manner. In this study we sought to understand how MEK inhibition and T-VEC promote host anti-tumor immunity.

Methods

Immunocompetent C57/BL6 mice bearing D4M3A melanoma tumors were treated with an adapted T-VEC encoding murine GM-CSF (mT-VEC;106 PFU) given by intra-tumoral administration biweekly with or without MEK inhibitor trametinib (0.1 mg/kg) by oral gavage for 2 weeks. Tumor growth was measured by calipers and, in some experiments, animals were monitored for survival or collection of tumor specimens for biomarker assessment. In independent experiments, treatment was evaluated in mice following depletion of specific immune cell subsets (CD8+ T cells, CD4+ T cells, macrophages and Batf3+ DCs). Flow cytometry analysis was performed on tumor-infiltrating lymphocytes using various markers of immune activation and analyzed by FloJo software (v 10.4). RNA was collected from tumors and subjected to Nanostring gene expression analysis. Statistical comparisons between groups were determined using Student’s t test and Kaplan-Meier method was used to estimate survival. P

Results

In an immunocompetent murine HSV-1-sensitive D4M3A melanoma model we observed a significant decrease in tumor growth and increased survival following the addition of trametinib to T-VEC therapy. Therapeutic activity was associated with recruitment of CD8+ T cells into the tumor microenvironment and was dependent on both CD8+ T cells and Batf3+ DCs, as anti-tumor activity was lost when CD8+ T cells, but not CD4+ T cells or macrophages were depleted, and in treated Batf3-/- mice. Characterization of the CD8+ T cells revealed early recruitment of HSV-specific CD8+ T cells followed by melanoma antigen-specific (gp100 and TRP2) CD8+ T cell responses. Gene expression analysis demonstrated that combination therapy induced an immune-inflamed signature supportive of lymphocyte recruitment and activation within the local tumor microenvironment. Combination therapy enhanced PD-L1 expression providing rationale for addition of PD-1 blockade in vivo. In this murine model, triple combination therapy with trametinib, T-VEC and anti-PD-1 was associated with significant improvement in therapeutic effectiveness and resulted in 80% overall survival. Finally, re-challenge with D4M3A in surviving mice demonstrated complete protection suggesting development of long-term anti-tumor memory.

Conclusions

Oncolytic virus immunotherapy is improved by MEK inhibition and appears to sensitize tumors to immune checkpoint blockade. These studies support clinical translation of this combination approach into the clinic.

Ethics Approval

The study was approved by Rutgers University IACUC.

Correspondence: Jacques Moisan (jacques.moisan@emdserono.com)

Background

Indoleamine 2,3-dioxygenase (IDO1) and tryptophan 2,3-dioxygenase (TDO2) are key immunoregulatory enzymes that catabolize the essential amino acid tryptophan (Trp) to kynurenine (Kyn), an immunosuppressive metabolite. Kyn accumulation exerts immunosuppressive effects by preventing effector T-cell differentiation and enhancing regulatory T-cell (Treg) activity, thereby creating and sustaining a hostile tumor microenvironment and providing a mechanism of tumor escape from anti–PD-1/PD-L1 blockade. Preclinical data show IDO1 inhibitors have single-agent antitumor activity. In human cancers, tumor expression of IDO1 and TDO2 is distinct and does not completely overlap. Therefore, dual inhibition of IDO1 and TDO2 could provide a broader antitumor effect in a wider range of tumors. The preclinical rationale and testing of M4112, a highly selective, dual inhibitor of IDO1/TDO2 is presented and shows enhancement of immune function and antitumor activity in various in vitro and in vivo models.

Methods

One-way, mixed lymphocyte reaction (MLR) assays were performed on human peripheral blood mononuclear cells and mature dendritic cells treated with different concentrations of M4112 in combination with avelumab, a fully human anti–PD-L1 IgG1 antibody, to measure interferon-gamma (IFN-gamma) levels [1]. Naive mice or CT26-KSA and MC38 colon carcinoma mouse models were used for pharmacokinetic and pharmacodynamic and antitumor activity studies. Measurements of M4112, Kyn, and Trp levels in the plasma, liver, and tumor were performed after treatment with M4112 alone or in combination with avelumab. The CANScript tumor explant platform, which predicts potential antitumor effects using an M score [2], was used on 20 fresh human head and neck squamous cell carcinoma (HNSCC) tumor samples. Supernatants were measured for Kyn, Trp, and cytokine levels after 72 hours of M4112 or avelumab treatment alone and in combination.

Results

In combination with avelumab, M4112 treatment increased IFN-gamma production, indicating enhanced T-cell activation. Kyn levels and the Kyn/Trp ratio were also reduced in mouse and human tumor explant models. Treatment with M4112 in combination with avelumab led to reduced tumor volumes in syngeneic tumor models. In addition, M4112 exhibited antitumor effects in 4 (20%) and 7 (35%) of 20 HNSCC tumor explants as a single agent and in combination with avelumab, respectively.

Conclusions

The dual-selective IDO1/TDO2 inhibitor, M4112, in combination with avelumab enhanced antitumor activity in mouse and human tumor models and normalized systemic Kyn/Trp levels. Thus, M4112 represents a promising therapeutic agent targeting both the IDO1 and TDO2 immunosuppressive pathways.

References

1. Hamilton G, Rath B. Avelumab: combining immune checkpoint inhibition and antibody-dependent cytotoxicity. Expert Opin Biol Ther. 2017;17(4):515-523.

2. Majumder B, et al. Predicting clinical response to anticancer drugs using an ex vivo platform that captures tumour heterogeneity. Nat Commun. 2015;6:6169.

Correspondence: Esther Chang (change@georgetown.edu)

Background

Glioblastomas are among the most lethal of cancers [1], and antibodies against the programmed cell death protein 1 (anti-PD-1) represent a promising immunotherapeutic strategy. However, significant numbers of glioblastoma patients fail to respond to anti-PD-1 [2,3]. Recent studies suggest that p53 participates in immune regulation [4,5], and we have shown that SGT-53, a novel nanomedicine for TP53 gene therapy, can enhance both innate and adaptive immune responses against a variety of tumor types [6,7]. SGT-53 is a liposome encapsulating the normal TP53 gene that targets tumors via an antibody fragment recognizing the transferrin receptor. SGT-53 has completed Phase I & Ib trials [8.9] is now in Phase II clinical trials. Not only does systemically administered SGT-53 target cancer cells with specificity, but it can also actively ferry its payload across the endothelial cells that constitute the blood-brain barrier (BBB) [10]. We therefore hypothesized that adding SGT-53 to anti-PD-1 would be more effective against glioblastomas than checkpoint blockade monotherapy .

Methods

We evaluated whether SGT-53 could augment anti-PD-1 therapy for murine glioblastomas (GL261) grown subcutaneously or intracranially in syngeneic mice. We examined a number of biological markers of immune responses and monitored expression of immune-relevant genes after mice bearing glioblastomas were treated with SGT-53 or anti-PD-1 or the combination of these therapeutic agents.

Results

Adding SGT-53 to anti-PD-1 immunotherapy resulted in increased intratumoral infiltration of immune effector cells and altered expression of a number of genes related to immune responses. SGT-53 triggered changes in both innate and adaptive immune responses and caused immunogenic changes that included elevated tumor expression of CRT, FAS, PD-L1, ICAM1 and MHC I. SGT-53 also markedly modified the previously immunosuppressive microenvironment of tumors. Importantly, the combination treatment restored T-cell effector function and increased CTLs within tumor (Figure 1), which was correlated with inhibition of subcutaneous and intracranial tumor growth (Figures 2 & 3, respectively) as well as a survival benefit in mice bearing intracranial glioblastomas.

Conclusions

Tumor-targeted p53 gene therapy via SGT-53 was able to augment anti-PD-1 immune checkpoint blockade and convert mouse syngeneic glioblastoma tumors that were unresponsive (immunologically “cold” tumors) into tumors that were more responsive to anti-PD-1 immunotherapy. These findings suggest that SGT-53 has potential to enhance the efficacy of checkpoint blockade and thereby provide for improved outcomes for patients with glioblastomas. Our observations provide compelling motivation to test the combination of SGT-53 and a checkpoint inhibitor in the context of a clinical trial for glioblastoma.

Acknowledgements

This work was supported by the NCI grant (5R01CA132012-02 to E.H.C.); and a research grant from SynerGene Therapeutics, Inc. This study was conducted at the Lombardi Cancer Center Shared Resource facilities that are partially supported by NCI (P30-CA051008). Some of the data to be presented at SITC have recently been published [11].

References

1. Wen PY, Kesari S. Malignant gliomas in adults. N Engl J Med 2008;359:492–507.

2. Lim M, Xia Y, Bettegowda C, et al. Current state of immunotherapy for glioblastoma. Nat Rev Clin Oncol 2018;15:422–42.

3. Kim ES, Kim JE, Patel MA, et al. Immune check-point modulators: an emerging antiglioma armamentarium. J Immunol Res 2016;2016:4683607.

4. Menendez D, Shatz M, Resnick MA. Interactions between the tumor suppressor p53 and immune responses. Curr Opin Oncol 2013;25:85–92.

5. Cui Y, Guo G. Immunomodulatory function of the tumor suppressor p53 in host immune response and the tumor microenvironment. Int J Mol Sci 2016;17:E1942.

6. Moore EC, Sun L, Clavijo PE, et al. Nanocomplex-based TP53 gene therapy promotes anti-tumor immunity through TP53- and STING-dependent mechanisms. Oncoimmunology 2018;7: e1404216.

7. Kim SS, Harford JB, Moghe M, et al. Combination with SGT-53 overcomes tumor resistance to a checkpoint inhibitor. Oncoimmunology 2018;7: e1484982.

8. Senzer N, Nemunaitis J, Nemunaitis D, et al. Phase I study of a systemically delivered p53 nanoparticle in advanced solid tumors. Mol Ther 2013;21:1096–103.

9. Pirollo KF, Nemunaitis J, Leung PK, et al. Safety and efficacy in advanced solid tumors of a targeted nanocomplex carrying the p53 gene used in combination with docetaxel: a phase 1b study. Mol Ther 2016;24:1697–706.

10. Kim SS, Rait A, Kim E, et al. A nanoparticle carrying the p53 gene targets tumors including cancer stem cells, sensitizes glioblastoma to chemotherapy and improves survival. ACS Nano 2014;8:5494–514.

11. Kim SS, Harford JB, Moghe M, Slaughter T, Doherty C, Chang EH. A tumor-targeting nanomedicine carrying the p53 gene crosses the blood–brain barrier and enhances anti-PD-1 immunotherapy in mouse models of glioblastoma. Intl J Cancer (2019) Epub 2019/06/27.

Ethics Approval

All animal studies was performed under an IACUC-approved protocol in compliance with the Animal Welfare Act, PHS Policy, and other Federal statutes and regulations relating to animals in the AAALAC-accredited animal facility at Georgetown University, which adheres to principles stated in the Guide for the Care and Use of Laboratory Animals, National Research Council .

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P523).

See text for description

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2 (abstract P523).

See text for description

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3 (abstract P523).

See text for description

Correspondence: Julia Schuler (julia.schueler@crl.com)

Background

Around 10% of breast cancer cases are attributed to genetic disorders like mutations in BRCA-1/2 genes. Targeted therapy of BRCA-deficient cancers has been achieved using poly(ADP-ribose) polymerase (PARP) inhibitors, which block BRCA-independent DNA repair.

Methods

To study the effect of this common mutation on tumor biology in general and specifically the immunological landscape of the tumor microenvironment, we knocked out the BRCA1 gene in the murine EMT6 breast cancer cell line. Subsequently, we analyzed the tumor infiltrating lymphocyte (TIL) population and the cytokine profiles in serum of tumor bearing animals.

Results

The percentage of TILs (=CD45+ cells) was similar in both lines (33.0% vs 29.6%, BRCA1-/- vs BRCA1wt ). Nevertheless, a more detailed analysis of the TIL subpopulations revealed higher percentages for gMDSC and M2 macrophages in the EMT6/BRCA1-/- line. The analysis of 23 cytokines showed a significant down regulation of 22 cytokines in the EMT6/BRCA1-/- bearing animals. Solely, G-CSF was significantly upregulated in the mutated cell line (multiple t-test, p< 0.001). With the objective to elucidate more in detail the biology behind the differences in cytokine levels and TIL distribution, we investigated the influence of different PARP and checkpoint (CP) inhibitors on both parameters. Both lines showed distinct TIL and cytokine profiles under treatment. In the EMT6/BRCA1-/- the TIL population decreased under PARP inhibitor monotherapy (Talazoparib) but increased significantly under treatment with CP inhibitors anti CTLA-4 and anti PD-1 in monotherapy as well as in combination with Talazoparib (Kruskall-Wallis test, p< 0.005). In contrast, the TIL frequency in the EMT6/BRCA1wt model was similar in all treatment arms. Differences in the TIL composition were mainly detected in the NK and gMDSC cell fraction which were upregulated under treatment including Talazoparib specifically in the EMT6/BRCA1-/- model. The TIL analysis in the EMT6/BRCA1wt tumors showed no differences in the individual subpopulations although the observed antitumoral activity of anti CTLA-4 as well as anti PD-1 in this model (tumor reduction of >60% as compared to untreated control). In line with the enhanced NK cell numbers under therapy was the fact that serum levels of RANTES and IL-2 were enhanced under treatment in EMT6/BRCA1-/- bearing animals.

Conclusions

Further studies using microdialysis as read-out will give additional information on the influence of the BRCA1 gene knockout and the resulting modulation of the tumor microenvironment and will help to elucidate the tumor biology and clinical relevance behind these findings.

Ethics Approval

The study was approved by the local authorities (Regierungspräsidium Freiburg; Germany) approval number G17-78

Correspondence: Jieqiong Liu (liujieqiong@163.com)

Background

Cancer immunotherapy using immune checkpoint inhibitors has significantly improved the clinical outcomes of patients with multiple aggressive and advanced-stage cancers. Despite its enormous successes, the overall response rates of immune checkpoint inhibitor as a monotherapy remain suboptimal, especially in treating breast cancers. In order to improve the efficacy of cancer immunotherapies, there is an increased interest in combining immune checkpoint inhibitors with targeted agents to enhance antitumor effects. Previous studies have shown that anti-angiogenic drugs that suppress tumor vasculatures can improve the therapeutic effect of immune checkpoint blockades in preclinical models of human cancer. However, the exact mechanism of action of such synergism is unclear and how the two treatment modalities should be administered in terms of dosing and sequencing to produce the optimal response remains unknown. In the present study, we investigate the antitumor effect of combined blockade of vascular endothelial growth factor receptors 2 (anti-VEGFR2) and programmed death (PD-1) in murine breast cancer models as well as in a phase II trial.

Methods

We tested the combination of anti-PD-1 with different dose of VEGFR2-targeting antibody in syngeneic breast cancer mouse models. Tumor infiltrated immune cell subsets were profiled by flow cytometry. A cytokine array was carried out to identify mechanism of the synergy between the two modalities. The efficacy of this combination was further evaluated in advanced triple-negative breast cancer (TNBC) patients.

Results

Antibody targeting VEGFR2 sensitizes breast tumors to PD-1 blockade in a dose-dependent manner. Although both conventional and low-dose anti-VEGFR2 antibody treatments normalize tumor vessels, low-dose VEGFR2 blockade results in more robust immune cells infiltration and activation, and promotes the secretion of osteopontin (OPN) by CD8+ T cells. OPN subsequently induces tumor cell production of TGF-ß, which in turn upregulates PD-1 expression on immune cells. In advanced TNBC patients, combined low-dose anti-VEGFR2 and anti-PD-1 treamtnets demonstrated excellent tolerability and efficacy with a doubling of progression free survival as compared to historical anti-PD-1 trials. Tissue analyses further revealed that higher OPN and TGF-β expressions were correlated with improved clinical responses in patients.

Conclusions

Taken together, these results demonstrate that a dose-dependent synergism exists between anti-angiogenic therapy and immune checkpoint blockade, thus providing important insights into identifying the optimal strategies to combine immunotherapy with molecular targeted agents.

Trial Registration

NCT03394287

Ethics Approval

All patient-related procedures were performed with the approval of the Internal Review and the Ethics Boards of the Sun Yat-sen Memorial Hospital. All animal studies were reviewed and approved by the Institutional Animal Care and Use Committee of UT Southwestern Medical Center and Sun Yat-sen University.

Correspondence: Emily Kinsey (enkinsey@gmail.com)

Background

The treatment landscape has drastically changed in mRCC, moving from anti-VEGF therapies to an immunotherapeutic approach in the first line setting for IMDC intermediate or poor risk mRCC. Combining anti-VEGF and anti-PD-1 or anti-PDL-1 therapies have shown survival improvements in mRCC, leading to approvals for first-line axitinib-pembrolizumab and axitinib-avelumab [1,2]. Cabozantinib and nivolumab (cabo/nivo) is also a safe option in phase 1 trials with some durable responses [3]. We evaluated the outcomes of patients who received cabo/nivo for mRCC refractory to immunotherapy alone.

Methods

A retrospective analysis was performed of patients with mRCC treated with ipilimumab and nivolumab (ipi/nivo) and subsequently with cabo/nivo at Duke Cancer Center between September 2017 and February 2019. Patient outcomes were collected including demographic information, treatment details, responses, and frequency of adverse events. The cohort of patients treated with the cabo/nivo combination is presented here.

Results

Eighty-six patients were treated with ipi/nivo for mRCC and of these patients, 34 patients also received cabozantinib, either alone or in combination with nivolumab. Nine patients received cabozantinib prior to ipi/nivo, and 14 patients received cabozantinib after ipi/nivo, and 10 patients received combination therapy with cabo/nivo. Of the ten patients who received combination cabo/nivo, 2 were favorable risk, 6 were intermediate risk, and 2 were poor risk. One patient had progressive disease, 2 patients had stable disease, 3 patients had a partial response (50% responses), and 4 patients were unevaluable due to insufficient follow up. Eight of 10 patients had treatment ongoing at the time of data collection.

Conclusions

Progression on immunotherapy alone did not appear to confer resistance to cabo/nivo treatment for five of the six patients who had a disease response assessment. Phase III studies COSMIC-313 and Alliance A031704 (PDIGREE) are ongoing to evaluate cabozantinib in combination or in sequence to ipilimumab immunotherapy.

References

1. Motzer RJ, Penkov K, Haanen J, Rini B, Albiges L, Campbell MT, et al. Avelumab plus Axitinib versus Sunitinib for Advanced Renal-Cell Carcinoma. N Engl J Med [Internet]. 2019;380(12):1103–15.

2. Rini BI, Plimack ER, Stus V, Gafanov R, Hawkins R, Nosov D, et al. Pembrolizumab plus Axitinib versus Sunitinib for Advanced Renal-Cell Carcinoma. N Engl J Med [Internet]. 2019;NEJMoa1816714.

3. Nadal R, Mortazavi A, Stein M, Pal SK, Davarpanah N, Parnes HL, et al. Final results of a phase I study of cabozantinib (cabo) plus nivolumab (nivo) and cabonivo plus ipilimumab (Ipi) in patients (pts) with metastatic urothelial carcinoma (mUC) and other genitourinary (GU) malignancies. Ann Oncol. 2017;28(supplement 5):846O.

Ethics Approval

This study was approved by the Duke IRB (#Pro00101984)

Correspondence: Michael Hollingsworth (mahollin@unmc.edu)

Background

Pancreatic ductal adenocarcinoma (PDAC) is a lethal disease, accounting for 338,000 new cases worldwide with a 5-year survival rate of less than 10%. Despite progress with new treatments, PDAC remains indomitable. New therapeutic strategies that are tumor-specific, relieve tumor-associated immunosuppression and modulate tumor stroma for the efficient delivery of chemotherapeutics are urgently required. Abundant evidence suggests a reciprocal relationship between allergies and pancreatic cancer. Distinct immune surveillance in allergic individuals might enhance immune protection against cancer [1-2]. We investigated a novel immunotherapeutic strategy to trigger a tumor antigen-specific allergic phenotype in a pre-clinical model of pancreatic cancer.

Methods

Using a combination of humanized anti-MUC1.IgE, anti-PD-L1, and PolyICLC, we investigated the efficacy of this unique triple combination in a pre-clinical model of pancreatic cancer using transgenic mice expressing human MUC1 and FcεRI (hMUC1/hFcεRI). IgE antibody against prostate-specific antigen (anti-PSA.IgE) and anti-MUC1.IgG was employed as controls. In parallel, transgenic mice for only hMUC1 or hFcεRIα were utilized as additional controls. We also performed immuno-depletion studies and analysis of tumor-infiltrating lymphocyte (TIL) from treated mice to investigate the role of innate (NK cells) and adaptive immune players (CD8 T cells) in providing protective benefits of derived from the combination therapy. In parallel, we performed a phospho-proteomic screen to elucidate the mechanism through which our proposed therapy harnesses NK cell-mediated anti-tumor immune response against pancreatic tumors.

Results

We noted that combined immunotherapy with a humanized IgE directed towards a tumor antigen expressed on pancreatic adenocarcinoma (MUC1), PolyICLC, and anti-PD-L1 induces antigen-specific cellular immune responses that mediate rejection of tumor cells in a pre-clinical model of pancreatic cancer. Our initial mechanistic studies suggest that effector mechanisms for anti-MUC1.IgE includes induction of ADCC by NK cells and activation of antigen-specific CD8 cytotoxic T lymphocytes (CTLs). Furthermore, we noted that the anti-MUC1. IgE-combination treatment promotes intra-tumoral NK cell activation and increased CD103+ DCs proportions inside the tumor, which were reversed upon NK cell depletion. Also, we noted the activation of SMAD1, but not SMAD3, pathways in NK cells post-treatment with anti-MUC1.IgE-based combination.

Conclusions

Taken together, this is the first study to show that specific stimulation of the IgE/FcεRIα axis in combination with PolyICLC and anti-PD-L1 can activate CD8 T cell and NK cell effector pathways and provide long-lasting tumor-protective benefits against pancreatic cancer. Our study provides preliminary evidence for the clinical applicability and rapid translation of an anti-MUC1.IgE based combination therapy against pancreatic cancer.

References

1. Gomez-Rubio P, Zock Jp, RAVA M et al. Reduced risk of pancreatic cancer associated with asthma and nasal allergies. Gut. 2017; 66 (2): 314-322.

2. Huang BZ, Le Marchand L, et al. Atopic allergic conditions and pancreatic cancer risk: Results from the Multiethnic Cohort Study. Int J Cancer. 2018;142 (10):2019-2027.

Ethics Approval

All animal studies were performed in accordance with the Institutional Animal Care and Use Committee guidelines (Approved IACUC protocol no: 15-123-02FC).

Correspondence: Sripad Ram (sripad.ram@gmail.com), Cathy Zhang (Cathy.Zhang@pfizer.com)

Background

The role of B cells in tumor immunity is poorly understood due to their conflicting roles that augment and suppress anti-tumor responses. For example, B cells can augment T-cell responses against tumor challenge as antigen presenting cells, while B cells can also blunt anti-tumor responses by developing neutralizing antibodies against therapeutic antibodies. In the context of targeted immunotherapy using antibody-based therapeutics, an important question arises as to whether B cells are necessary for eliciting an anti-tumor immune response.

Methods

We use two different clones of anti-PD-L1 antibody, 10F9G2 and MIH5, which are rat IgG2b and rat IgG2a, respectively, in conjunction with a mouse anti-mouse 4-1BB antibody. CT26 tumors were implanted subcutaneously in BALB/c mice or genetically B-cell deficient JH-/- mice. Binding affinity measurements were performed on BIACore. For flow cytometry, tumor, spleen and lymph nodes were harvested at different times after the last dose, dissociated and assayed for different immune cell markers. For single-cell RNASeq, mouse PBMCs were isolated and analyzed using the 10x Genomics platform.

Results

In vivo and in vitro pharmacokinetic studies revealed that MIH5 clone had better bioavailability than the 10F9G2 clone which is attributed to its higher binding affinity to FcRn. Combination treatment with the MIH5 clone resulted in superior antitumor efficacy when compared to using the 10F9G2 clone in wild type and JH-/- mice. However, for a given combination, JH-/- mice showed consistently poorer efficacy than wild type mice. More specifically, in the absence of B cells CD8+ T-cell expansion in the tumor was diminished and memory T-cell development was systemically impaired. We also report that B cells are required for reciprocal crosstalk activation between 4-1BB and PD-L1 signaling pathways which could augment the synergistic response of combination therapy. Results from single cell RNASeq suggest that B cells promote T-cell mediated tumor immunity through the production of cytokines that increase antigen processing and presentation, and support T-cell activation, proliferation and differentiation.

Conclusions

Our results reveal an underappreciated role of B cells in T-cell targeted cancer immunotherapy. They also underscore the importance of serum persistence of therapeutic antibodies in mediating anti-tumor immunity.

Ethics Approval

All experimental animal procedures complied with the Guide for the Care and Use of Laboratory Animals (Institute for Laboratory Animal Research, 2011) and were approved by the Pfizer Global Research and Development Institutional Animal Care and Use Committee (IACUC)

Correspondence: Jae Hyung Park (jhpark1@skku.edu)

Background

Cytotoxicity T cells play central roles in destroying cancer cells in cancer immunotherapy, which is similar to immunological rejection [1]. However, in many cancer types, cancer cells are able to avoid recognition of cytotoxic T cells because they present self-antigen on the cell surface [2]. In this study, we designed polymeric conjugate which can cause foreign-antigen presentation by delivering foreign antigen to the cancer cells for recognition of cancer cell by cytotoxic T cells.

Methods

We conjugated ovalbumin (OVA), a model foreign antigen, to biocompatible hyaluronic acid (HA) via simple reductive amination method. The surface of HA was coated with polyethylene glycol (PEG) containing tumor microenvironment-sensitive linker. The polymeric conjugate was then used to demonstrate in vitro and in vivo foreign antigen presentation and cancer cell death by cytotoxic T cells.

Results

The OVA contents in the polymeric conjugate were 13.2 wt.% determined by BCA assay. Cellular uptake study revealed that the polymeric conjugate can deliver antigen into the cancer cells and cause antigen presentation as well. In the tumor-bearing mice model, the polymeric conjugate has been shown to accumulate at significant levels in tumor tissue, thus leading to effective in vivo antigen presentation.

Conclusions

In this study, we synthesized model foreign antigen-bearing polymeric conjugate. The polymeric conjugate showed a significant anticancer effect in tumor-bearing mice model. Therefore, it is suggested that the foreign antigen delivery system might be an alternative for cancer immunotherapy.

References

1. Friedl P, den Boer AT, Gunzer M. Tuning immune responses: diversity and adaptation of the immunological synapse. Nat Rev Immunol. 2005;5:532-545.

2. Garcia-Lora A, Algarra I, Garrido F. MHC class I antigens, immune surveillance, and tumor immune escape. J Cell Physiol. 2003;195:346-55.

Correspondence: Taha Merghoub (merghouT@mskcc.org), Jedd Wolchok (wolchokj@mskcc.org)

Background

Although immune checkpoint blockade (ICB) may result in clinical benefit for a subset of cancer patients, multiple mechanisms of resistance can impair optimal response. The catabolism of Tryptophan (Trp) by the enzymes IDO or TDO is a frequent phenomenon that plays a suppressive role in tumor immunity. Accumulation of the Trp metabolite L-Kynurenine (Kyn) was shown to act as agonist of the aryl hydrocarbon receptor (AHR) and that its activation promotes immunological tolerance.

Methods

Here by using pre-clinical models of melanoma and analysis of clinical samples we sought to characterize the mechanisms of immune suppression associated with the AHR pathway and to evaluate its potential as therapeutic target.

Results

RNAseq analysis of human cancers revealed an up-regulation of Kyn-degrading enzymes in patients responsive to PD-1 blockade. Also, we observed a correlation between the expressions of Kyn-AHR pathway genes with a TGF-β immune signature and markers associated with immunotherapy resistance (PD-1, FOXP3, CD206). We found that the overexpression of IDO or TDO in a pre-clinical model of melanoma (B16-F10) promoted CD8+ T cell dysfunction and accumulation of M2-TAMs and Tregs with an activation of the AHR pathway and enhanced suppressive function. Moreover, IDO-Kyn-AHR-mediated tumor progression was dependent on an interplay between Tregs and TAMs, which could be reversed by selective inhibition of the AHR. Treatment of IDO/TDO-expressing but not wild-type tumor models with a selective AHR antagonist delayed tumor growth, and its efficacy was further improved with PD-1 blockadeRNAseq analysis of human cancers revealed an up-regulation of Kyn-degrading enzymes in patients responsive to PD-1 blockade. Also, we observed a correlation between the expressions of Kyn-AHR pathway genes with a TGF-β immune signature and markers associated with immunotherapy resistance (PD-1, FOXP3, CD206). We found that the overexpression of IDO or TDO in a pre-clinical model of melanoma (B16-F10) promoted CD8+ T cell dysfunction and accumulation of M2-TAMs and Tregs with an activation of the AHR pathway and enhanced suppressive function. Moreover, IDO-Kyn-AHR-mediated tumor progression was dependent on an interplay between Tregs and TAMs, which could be reversed by selective inhibition of the AHR. Treatment of IDO/TDO-expressing but not wild-type tumor models with a selective AHR antagonist delayed tumor growth, and its efficacy was further improved with PD-1 blockade

Conclusions

In summary, our findings demonstrate that targeting the Kyn pathway through AHR inhibition could overcome key suppressive mechanisms and sensitize otherwise immune resistant tumors to PD-1 blockade.

Ethics Approval

All mice were maintained in microisolator cages and treated in accordance with the NIH and American Association of Laboratory Animal Care regulations. All mouse procedures and experiments for this study were approved by the MSKCC Institutional Animal Care and Use Committee.

Correspondence: Brent Hanks (brent.hanks@duke.edu)

Background

An in-depth understanding of immune escape mechanisms in cancer are likely to lead to innovative advances in immunotherapeutic strategies. However, our understanding of these mechanisms driving both primary and secondary immunotherapy resistance remains incomplete. While many groups have described the role of NOD-, LRR- and pyrin domain-containing protein-3 (NLRP3) as a sensor for pathogen-derived danger signals by antigen-presenting cells in the innate immune system, relatively little is known about the contribution of NLRP3 to tumorigenesis and its role in modulating tumor responses to immunotherapy has not been explored

Methods

Several pre-clinical syngeneic and transgenic tumor models, an autologous humanized mouse tumor model, and clinical plasma and tumor specimens derived from advanced melanoma patients were utilized to investigate the genetic and cellular alterations of tumors escaping anti-PD-1 antibody immunotherapy.

Results

Using several pre-clinical tumor models, we have determined that granulocytic myeloid-derived suppressor cells (PMN-MDSCs) accumulate within progressing tumors undergoing treatment with anti-PD-1 antibody (ab) immunotherapy. This PMN-MDSC recruitment effect was found to be due to an autocrine Wnt5a-dependent upregulation of CXCR2 chemokines in tumor tissues. After additional study, we determined that this pathway was induced by HSP70-TLR4 stimulation and occurred in response to cytolytic CD8+ T cell activation. Using genetic and pharmacologic approaches we have found that this tumor-intrinsic signaling cascade is triggered by a PD-L1-dependent activation of the NLRP3 inflammasome via a STAT3-PKR signaling axis. This mechanism requires CD8+ T cell activation and IFN-gamma expression. However, treatment with chemotherapy or anti-CTLA-4 ab immunotherapy failed to induce a similar effect. Genetic ablation and pharmacologic inhibition of NLRP3 suppresses PMN-MDSC recruitment and enhances the efficacy of ant-PD-1 ab immunotherapy in an autochthonous model of BRAF(V600E) melanoma. We then sought to validate this pathway in clinical specimens where we found elevations in various myeloid markers and CXCR2 chemokines in tumor tissues as well as an increase in plasma HSP70 levels in non-responding melanoma patients progressing through anti-PD-1 ab immunotherapy.

Conclusions

This work reveals a tumor-intrinsic PD-L1-NLRP3 inflammasome signaling pathway triggered by CD8+ T cell activation that ultimately drives adaptive resistance to anti-PD-1 immunotherapy by promoting the recruitment of PMN-MDSCs to the tumor bed. The identification of the key PD-L1-STAT3-PKR-NLRP3 signaling axis responsible for this mechanism provides insight into promising strategies for augmenting the efficacy of checkpoint inhibitor immunotherapy as well as for monitoring the management of cancer patients undergoing these therapies. NLRP3 inhibitors represent promising agents for future anti-PD-1 combination clinical trials.

Ethics Approval

This study utilized clinical specimens derived from ongoing tissue acquisition protocols approved by the Institutional Review Boards at Duke University Medical Center (IRB approval number Pro00059349) and Vanderbilt University Medical Center (IRB approval number 100178).

Correspondence: Lyse Norian (lnorian@uab.edu)

Background

Associations between modifiable factors (e.g., adiposity) and immunotherapeutic efficacy remain uncertain. Preclinical studies suggest increased adiposity improves response rates to immune checkpoint blockade; however, few studies control for diet effects between chow-fed lean and high-fat diet-induced obese (DIO) mice. We found previously that DIO reduces the efficacy of an immunotherapy consisting of replication-deficient adenovirus (Ad) encoding tumor-necrosis factor-related apoptosis-inducing ligand (TRAIL) (AdT) plus class B oligonucleotides (CpG1826) in renal tumor-bearing mice [1]. To eliminate the compounding effects of diet and explore if therapy response can be improved in DIO mice, the current study evaluated response rates to AdT/CpG combined with anti-cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) in diet-matched obese-resistant (OBR) versus DIO mice.

Methods

BALB/c mice were administered a high-fat diet for 20-weeks. Following diet administration, mice were characterized as either OBR or DIO, as previously described [2]. DIO mice possess hallmarks of obesity, including increases in systemic glucose, insulin, and leptin; whereas, OBR mice display phenotypic characteristics similar to chow-fed, lean controls. Mice were injected orthotopically with Renca-Luc renal cancer cells into the left kidney. At day seven post-tumor challenge, mice were randomized to receive saline or AdT/CpG injected intra-tumorally, followed by isotype control (n=7-9/group) or anti-CTLA-4 therapy (n=12/group). At day 21 post-tumor challenge, immunogenetic profiling of whole-tumors was performed via nanoString and tumor-infiltrating lymphocytes were characterized by flow cytometry.

Results

Both OBR and DIO therapy-treated mice had significant reductions in tumor weight compared to therapy-free controls (66% reduction for OBR; p=0.001 versus 55% reduction for DIO; p=0.007). However, therapy was effective in 75% of OBR mice compared to 50% of DIO mice. Over 30 differentially expressed genes were shared between OBR and DIO therapy-treated responders compared to therapy-free controls. A third of these “responder” genes were related to T cell function or migration (e.g., Cd8a, CD274, Icos, Icosl, Zap70, Cxcr3, Xcr1, Il16). Cellular analyses demonstrated that OBR therapy-treated responder mice had a greater percentage of tumor-infiltrating activated (CD44+) CD8+ T cells compared to DIO therapy-treated responder mice (p=0.004); however, CD8+ T cell IFNγ production and PD-1 expression were comparable between these groups.

Conclusions

When controlling for diet, increased adiposity reduced the response rate to combinatorial anti-CTLA-4 based therapy. However, both OBR and DIO therapy-treated responder mice shared immunogenetic and T cell profiles independent of weight status. To improve treatment strategies, future studies are needed to understand the mechanistic drivers of the differential response rates to immunotherapy observed with obesity.

Acknowledgements

Research reported in this abstract was supported by The National Institutes of Health (NIH) Grant #5R01CA181088 to LAN. WJT is supported by The Research Training Program in Basic and Translational Oncology (T32CA183926), The University of Alabama at Birmingham School of Medicine Department of Medicine, Division of Hematology and Oncology, Birmingham, AL. RMO is supported by The Cancer Prevention and Control Program (T32CA047888), The University of Alabama at Birmingham, Birmingham, AL. JTG is supported by The Training Program in Cell, Molecular, and Developmental Biology (T32GM008111), The University of Alabama at Birmingham, Birmingham, AL.

References

1. James, B.R., Tomanek-Chalkley, A., Askeland, E.J., Kucaba, T., Griffith, T.S., Norian, L.A. Diet-Induced Obesity Alters Dendritic Cell Function in the Presence and Absence of Tumor Growth. J Immunol. 2012; 189:1311-1321.

2. Boi, S.K., Buchta, C.M., Pearson, N.A, Francis, M.B., Meyerholz, D.K., Grobe, J.L., Norian, L.A. Obesity Alters Immune and Metabolic Profiles: New Insight from Obese-Resistant Mice on High-Fat Diet. Obesity. 2016; 24(10):2140-2149.

Ethics Approval

All animal procedures were approved by the Institutional Animal Care and Use Committee (IACUC) at the University of Alabama at Birmingham, approval number 20233.

Correspondence: Sonia Iyer (iyers@wi.mit.edu), Robert Weinberg (weinberg@wi.mit.edu)

Background

The cornerstone of the existing treatment of high-grade serous ovarian cancer (HGSOC) is DNA-damaging chemotherapy; however, practically all patients eventually develop the progressive disease and the 5-year survival is only 40%. Immunotherapy would seem to be an attractive alternative treatment to chemotherapy, yet existing immunotherapies perform poorly in ovarian cancer, with only ~10% of patients responding to checkpoint-blockade. Why this is the case remains poorly understood and there is a pressing need to understand the underlying biology of immune evasion in ovarian cancer. Unfortunately, the preclinical tools required to explore the relationship between the types of DNA damage repair deficiencies and immune evasion have been lacking. Hence, we have modeled the biology of ovarian cancer using patient-relevant mutational landscapes in an immune-proficient, syngeneic-mouse model to help us identify the contribution of common driver mutations to the immune repertoire in the tumor microenvironment, and thus to responses of HGSOC tumors to immunotherapy.

Methods

We hypothesize that the immune composition and gene expression signatures of the resulting tumors will vary based on the combination of genetic alterations and the DNA repair proficiency of the transformed cells. To this end, we have engineered novel syngeneic mouse models from murine fallopian tube epithelium using CRISPR/Cas9 technology. These tumors capture the most common combinations of co-occurring mutations observed in patients. These models can identify the contribution of common driver mutations to the heterotypic interactions between cancer and stromal/immune compartments and examine how DNA repair proficiency contributes to immunogenicity.

Results

To validate the DNA repair proficiency of the transformed cells, we measured Rad51 nuclear focus formation after ionizing radiation (IR) and PARP-inhibitor and DNA-damaging-agent sensitivity. The HR-deficient cell lines had significantly fewer Rad51 nuclear foci and were more sensitive to PARP-inhibition in comparison to HR-proficient cells. Initial immune /stromal analysis using flow cytometry, scRNAseq transcriptomic and immunofluorescence analysis revealed substantial differences in the myeloid and T-cell regulatory compartments between HR-proficient and -deficient primary and metastatic tumors and within the ascitic fluid. Preliminary results also suggest that inhibition of the DNA damage response, checkpoint kinase 1 in combination with immune checkpoint inhibitors, potentiates antitumor effects and augments cytotoxic T-cell infiltration.

Conclusions

These results reveal how common mutational drivers determine the microenvironment of the tumor and its response to treatment. Understanding the genetic basis of these complex cellular interactions will be critical to better tailor combinations of existing targeted treatments and immunotherapies in ovarian cancer to fight this devastating disease.

Correspondence: Lisa Blum (lblum@pcyc.com)

Background

Characterization of heme malignancies, particularly in response to therapies, requires detailed monitoring and analysis of changes to patient immune subsets. New high-dimensional cytometry technologies have expanded the depth of data that can be collected, and several clustering algorithms have been described to improve the consistency and efficiency of associated data analysis. However, few studies have described a pipeline that is user friendly to immunologists and compared such results to validated data generated with manual gating by experts in the field. Here, we detail a flow analysis pipeline, ASAP, and demonstrate its precision and accuracy on data from a published study of over 200 clinical samples [1].

Methods

ASAP was built on several existing R packages (flowCore, flowPeaks, ggcyto, uwot) and makes use of marker enrichment modeling (MEM) scores for cell subset classification [2]. The pipeline performs data formatting and manages intermediate inputs and outputs to provide a seamless user experience. Starting with 3 input files defining panel information, preprocessing settings, and fcs files, ASAP performs preprocessing which includes data transformation, applying compensation, and removal of debris, doublets, and dead cells. Processed files are then clustered by cell markers and assigned to subsets based on MEM scores. ASAP automates the tabulation and visualization of user-defined and novel cell types. Results produced by ASAP were benchmarked with results from manual gating of a 16-color panel [1].

Results

Results produced by ASAP are highly concordant with those produced by expert manual gating, r2=0.98 (Figure 1). This high correlation is observed for major immune subsets with strong marker expression, nearly independent of cell frequency, ranging from 1%–90%. Analytical variability from 14 repeated acquisitions of the same human peripheral blood mononuclear cell sample on multiple days was low (interquartile range <10% median frequencies) and within the expected inter-experiment variability (Figure 2). The typical runtime of ASAP is less than 30 minutes for a study of 20 samples at 200,000 events/sample, and less than 5 hours for a large study of 150 samples at 3 million events/sample. In comparison, the manual analysis and quality control of the same dataset would require approximately 500 hours.

Conclusions

ASAP enables the automated analysis of flow cytometry data from large studies with precision, accuracy, and reproducibility comparable to the gold standard of expert manual gating. The automation of such a pipeline allows high consistency of gating and immune cell classification, which facilitates objective evaluation of patient response over time.

Acknowledgements

Funding source: Pharmacyclics LLC, an AbbVie Company

References

1. Solman I, You H, Taylor M, Dubovsky JA, Kipps TJ, Burger JA, Lal I, Chang BY, James DF, Hill JS. Ibrutinib vs chlorambucil: immunophenotypic and quantitative impacts on circulating immune cells in chronic lymphocytic leukemia (CLL). J Clin Oncol. 2017;35:7524-7524.

2. Diggins KE, Greenplate AR, Leelatian N, Wogsland CE, Irish JM. Characterizing cell subsets using marker enrichment modeling. Nat Methods. 2017;14:275-278.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P535).

ASAP concordance with manual gating

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2 (abstract P535).

Reproducibility of ASAP for a repeat-run sample

Correspondence: Ayesha Butt (ayeshabutt21@outlook.com)

Background

Orbital tumour board is a multidisciplinary approach towards ophthalmic cancer management, where relevant experts collaborate to manage patients holistically. Studies corroborate the notion that tumor boards positively affect patient outcomes. However, there is a scarcity of literature on the influence of orbital tumor boards globally and about tumor boards in general in Pakistan. There is no existing literature about orbital tumor boards in Pakistan. This study aims to assess the impact of tumor boards in managing ophthalmic cancers.

Methods

A retrospective review of data was carried out on cases that were presented in the orbital tumor board of Aga Khan University Hospital, the first orbital tumour board in Pakistan, from its commencement in August 2017 to May 2018.

Results

Out of a total of 80 patients in the study, 40%(n=32)were female and 60%(n=48) were male. The age of the patients ranged from 4 months to 78 years, with mean of 30 years. For analysis we divided the patients in three groups according to the nature of the diagnosis. Group 1 included those who had malignant tumours 40%(n=32), group 2 were those with benign tumours 30%(n=24) and group 3 was the non tumour group who were eventually diagnosed with other conditions 30%(n=24). Overall, 31.25% (n= 25) patients had a change in management plan, which was recorded, which included 10 patients in the malignant group, 9 in the benign group and 6 in the non tumour group.

Conclusions

A systematic review of literature has concluded that 54% of studies stated a change in management plans of at least 10% of patients post tumor board meetings. In our cohort, 31.25% of the patients had their treatment modified after tumour board discussion. The study shows that the management of ophthalmic cancers is influenced positively by tumour board discussions.

Correspondence: Logan Rhea (logan.rhea@inova.org)

Background

Nivolumab is a fully human immunoglobulin-G4 (IgG4) monoclonal antibody that selectively inhibits PD-1 activity, and can have various side effects. Drug-induced subacute cutaneous lupus erythematosus (SCLE), a unique drug-induced lupus erythematosus (DILE) syndrome, and is rare. Here, we report a case of SCLE after adjuvant nivolumab for resected stage IV melanoma.

Methods

A 77-year-old man with history of Stage I melanoma of the forehead, developed solitary pulmonary metastasis after 7 years which was resected with free margins. He began treatment with adjuvant nivolumab 480 mg every 28 days. Approximately two weeks after his 4th dose, he developed a pruritic rash limited to the trunk (Figure 1). This rash was believed to be a non-specific dermatitis related to nivolumab, and was treated with 6-day course of methylprednisolone. Although his symptoms initially improved, the rash progressed and he developed papules on the arms (Figure 2), along with polycyclic patches on the extensor forearms bilaterally (Figure 3).

Results

Skin biopsies were obtained, revealing interface dermatitis with focal epidermal pallor and necrosis, along with occasional eosinophils. The histologic differential included drug-induced erythema multiforme or DILE. Laboratory testing noted elevated erythrocyte sedimentation rate, C-reactive protein, positive antinuclear antibody (ANA), elevated anti-Ro and elevated anti-La antibody titers. Classic drug-induced lupus and systemic lupus erythematosus markers were negative.

Conclusions

Our case illustrates the development of DILE with SCLE morphology in a patient without previous history of autoimmune disease with nivolumab. DILE includes several lupus-like syndromes that develop with certain medication use, and resolve once discontinued. Histopathologic and autoantibody formation differ among DILE syndromes, though virtually all patients will have a positive ANA. The diagnosis is clinical. Chemotherapeutic agents are scarcely documented to cause DILE[1].

SCLE is a unique cutaneous variant of DILE. Skin lesions are generally annular, polycyclic or papulosquamous, and typically involve the upper trunk and upper extremity extensor surfaces[1]. Autoantibody production in SCLE is typical, and associated with high anti-Ro antibodies titers[2]. Anti-Ro antibodies are neither specific for Sjogrens, nor SCLE, additionally seen in malignancy and autoimmune liver injury. Symptoms may occur at any point during therapy. Treatment, like all DILE, is discontinuation of offending medication. Additional supportive treatments include photoprotection and topical corticosteroids. Lesions are expected to resolve in weeks, without scarring[3].

Our patient’s diagnosis was made clinically, with support of pathology and autoantibody panels. Nivolumab has been documented to cause SCLE in literature at least twice[4,5], however previous patients developed also other immune-related side effects.

References

1. Lamond NW, Younis T, Purdy K, Dorreen MS. Drug-induced subacute cutaneous lupus erythematosus associated with nab-paclitaxel therapy. Curr Oncol. 2013;20(5):e484-7.

2. Lowe G, Henderson CL, Grau RH, Hansen CB, Sontheimer RD. A systematic review of drug-induced subacute cutaneous lupus erythematosus. Br J Dermatol. 2011;164:465–72.

3. Srivastava M, Rencic A, Diglio G, et al. Drug-induced, Ro/ssapositive cutaneous lupus erythematosus. Arch Dermatol. 2003;139:45–9.

4. Liu, Rose C et al. “Subacute cutaneous lupus erythematosus induced by nivolumab.” The Australasian journal of dermatology. 2018;59(2): e152-e154 .

5. Zitouni N, Arnault JP, Dadban A, Attencourt C, Lok C, Chaby G. Subacute cutaneous lupus erythematosus induced by nivolumab. Melanoma Research. 2019; 29(2):212-215.

Consent

Consent was obtained from the patient for publication of this abstract.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P537).

See text for description

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2 (abstract P537).

See text for description

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3 (abstract P537).

See text for description

Correspondence: Ayesha Butt (ayeshabutt21@outlook.com)

Background

Retinoblastoma is the most common intraocular malignancy of childhood, accounting for 3 - 4% of all pediatric neoplasms. Mortality associated with retinoblastoma in the developed world is 3% to 5% whereas in the developing world it is 40%–70%. There is scarcity of research about retinoblastoma in Pakistan. We aim to study the presentation of retinoblastoma and impact of multidisciplinary management and tumor board discussions on its treatment at a tertiary care hospital in Pakistan

Methods

Retrospective review of retinoblastoma cases from May 2016 to April 2018 at Aga Khan University Hospital.

Results

A total of 35 eyes of 26 patients were studied with 9 presenting with bilateral retinoblastoma. 57.7% (n=15) of the population were males. The age range was 1 - 84 months (median: 14.5 months). The children presented from 3 different countries with most (76.9%, n=20) belonging to medium, upper medium or upper socioeconomic class.

The most prevalent symptom was leucocoria (38.5%, n=10) followed by decreased visual response (15.4%, n=4), squint (15.4%, n=4) and positive family history (15.4%, n=4). Most of the eyes (57.1%, n=20) were Group E on presentation, followed by Group D (20%, n=7) and 17.1% (n=6) Group B. One eye each presented with Group C and A. All the eyes that presented with Group A or B were of patients who had bilateral disease with advanced stage in the other eye. 54.3% (n=19) of the eyes were treated with a combination therapy including enucleation, focal lasers and systemic chemotherapy, 20% (n=7) with only focal lasers and 17% (n=6) with only enucleation. 42.3% (n=11) of the children presented with heritable traits. One patient (3.8%) did not survive

Conclusions

At our center we see Retinoblastoma in children with diverse background and majority presented with advanced disease. Multimodal therapy improved outcome. Awareness needs to be created for early detection and referral especially in the developing world.

Correspondence: KINJAL PARIKH (kparikh@medscape.net)

Background

The evolving melanoma treatment algorithm has highlighted the role of immune checkpoint inhibitors in treating patients with this cancer. From the first drug approval in 2011 through today, many changes have occurred in treatment approaches. Selecting the most appropriate therapy with implementation of monitoring strategies for managing adverse events is vital. The goal of this study was to determine if participation in an educational activity can improve the proficiency of oncologists on the application of checkpoint inhibitors in adjuvant and metastatic melanoma.

Methods

An online continuing medical education (CME)-certified interactive, text-based activity developed by Medscape Education Oncology and the Society for Immunotherapy of Cancer included 2 patient cases, which served as the foundation for interactive questions. Educational design included a “test, then teach” approach to elicit cognitive dissonance, with evidence-based feedback provided following each learner response. A repeated pairs pre-/post-assessment study design with 3 case-based questions and 1 confidence question was used. A chi-square test assessed differences from pre- to post-assessment. P values .26 is extensive). The activity was launched online 2/25/19 and data were collected through 5/8/19.

Results

Participation in education resulted in statistically significant improvements and an extensive educational effect for oncologists (n=62; P < .001; V =.342). An average of 41% correctly responded to pre-assessment questions, increasing to 75% post-assessment. 40% of oncologists reported greater confidence prescribing immunotherapy in melanoma. Significant improvements in competency were observed in the following areas:

• Selection of an evidence-based regimen for patients with melanoma in the adjuvant setting (24% vs 73%, P < 0.001, V = 0.484, Figure 1)

• Identification of the most appropriate regimen for patients with metastatic melanoma following molecular testing (47% vs 74%, P <.01, V = 0.280, Figure 2)

• Implementing team-based strategies for monitoring and managing irAEs (52% vs 77%, P <.01, V = 0.269, Figure 3)

Conclusions

This online interactive, case-based, CME-certified educational activity resulted in significant gains in oncologist competency in identifying and selecting appropriate treatments for patients with melanoma in the adjuvant and metastatic setting, and in implementing monitoring and management strategies (Figure 4). These results demonstrate the effectiveness of on-demand education in improving the translation of clinical knowledge into practice scenarios while improving clinician confidence.

Acknowledgements

Supported by an educational grant from: Bristol-Myers Squibb Company, Incyte Education, and Merck & Co, Inc.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P539).

Selecting Evidence-Based Regimens

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2 (abstract P539).

Competency with Molecular Profiling Treatment Selection

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3 (abstract P539).

Competency Implementing Team-Based Strategies for irAEs

• Download figure
• Open in new tab
• Download powerpoint

Fig. 4 (abstract P539).

Confidence Prescribing Immunotherapy in Melanoma

Correspondence: Candice Yong (candice.yong@astrazeneca.com), Brian Seal (brian.seal@astrazeneca.com)

Background

While platinum-based chemotherapy doublets were the first-line (1L) standard of care for metastatic non-small cell lung cancer (mNSCLC), the treatment landscape has recently shifted with the approval of immune checkpoint inhibitors. With the availability of these newer treatment options, it is important to elucidate the risk:benefit decision-making with regards to efficacy and toxicities that underlies oncologists’ real-world practice in mNSCLC. This study aimed to quantify preferences for attributes associated with immunotherapy and chemotherapy for mNSCLC among US oncologists.

Methods

Oncologists were recruited via a healthcare research panel to complete a cross-sectional online survey. Preferences were assessed in two separate discrete choice experiments (DCEs) where oncologists chose their preferred option in a series of tasks showing two alternative treatment profiles. The profiles varied based on six attributes identified in qualitative interviews with oncologists (Table 1). In both DCEs, the attributes were based on 1L treatment of patients with PD-L1 tumor cell (TC) >=1% and PD-L1 TC >=50% expression, respectively. A Hierarchical Bayes model was applied to estimate preference weights for each level of the treatment profile attributes.

Results

Oncologists (N=216) with an average of 16 years in practice participated; 69% were in community settings, and 31% were in academic settings. Attribute preferences were similar for patients with PD-L1 TC >=1% and PD-L1 TC >=50% (Figures 1a-1b). Improvements in OS were more important relative to improvements in any other attributes evaluated. Reducing the risk of neutropenia was second most important. The preference data showed a willingness to make trade-offs between efficacy and toxicity. For example, for both PD-L1 patient groups, oncologists would require an increase in OS by 5 months to accept an increase in risk of neutropenia from <1% to 25% and an increase in OS by 4 months to accept an increase in risk of pneumonitis from <1% to 5% (all grades) and <1% to 3% (Grade 3/4).

Conclusions

This study provides key insights into the trade-offs that oncologists are willing to make between efficacy and toxicities when selecting 1L treatment for patients with mNSCLC. Results suggest that improvements in OS were highly valued by oncologists and required to offset increases in the risk of toxicities, particularly neutropenia and pneumonitis. These findings were observed irrespective of patient PD-L1 status, which supports the consistency of oncologist preferences despite the differences in OS that have been observed for patients with PD-L1 TC >=1% or PD-L1 TC >=50% treated with immunotherapy regimens.

Acknowledgements

The authors acknowledge Oliver Will, PhD, and Bernadette Hallissey, PhD, for their assistance with data analysis

Ethics Approval

This study was reviewed by Pearl Institutional Review Board (Indianapolis, IN, US) and determined to be exempt.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P540).

See text for description

Correspondence: Claire Saxton (csaxton@cancersupportcommunity.org)

Background

Providing education and support for caregivers of cancer patients improves physical, psychosocial, and quality of life outcomes for both caregivers and patients.[1-4] As the use of immunotherapy continues to expand, it is important that caregivers have access to education and community resources to support their loved ones in making informed decisions, recognizing side effects, and getting optimal care. This analysis explores gains from Cancer Support Community’s Frankly Speaking about Cancer (FSAC): Immunotherapy program, a national evidence-based educational program created for cancer patients and caregivers.

Methods

One thousand two hundred thirty-two adults attending FSAC: Immunotherapy psychoeducational workshops nationwide from 2014-2018 completed a post-workshop survey. Participants rated their pre and post-workshop immunotherapy knowledge (“How knowledgeable were/are you about immunotherapy before/after this workshop?” 1= Not at all, 5= Very much) as well as their level of confidence speaking to their doctor (“Before/After this workshop I spoke with/feel more confident in speaking with my doctor about immunotherapy:” 1=Strongly disagree, 5=Strongly agree). 23% of respondents were non-professional caregivers; the remainder included cancer patients (66%), health care professionals (8%), and others (3%). This analysis focuses on 282 non-professional caregivers who self-reported background characteristics and workshop-specific outcomes (81% response rate). Descriptive analyses and paired samples t-tests were conducted to assess workshop outcomes.

Results

Of the 282 caregivers, 54% were spouses/partners, 31% family, and 15% friends. Caregivers were primarily White (72%) and female (70%); average age was 55 (s.d.= 17.6 years; range: 19-85 years). 56% of caregivers reported this was their first informational workshop. While 46% reported that immunotherapy treatment was an option for their loved one, only 9% reported speaking with their doctor about immunotherapy and its potential side effects. Post- workshop, 47% caregivers reported ‘high’ or ‘very high’ immunotherapy knowledge vs. pre-workshop (11%), a statistically significant difference in perceived knowledge (t(210)=24.05, p<.01). Caregivers also reported an increase in confidence in discussing immunotherapy and its side effects with their doctors, with 71% reporting confidence (agree/strongly agree) in speaking to their doctor about immunotherapy post-workshop vs. 24% pre-workshop (t(83)=5.8, p<.01), and 64% feeling more confident in talking with their doctor about immunotherapy’s side effects post-workshop vs. 32% pre-workshop (t(98)=8.9, p<.01). The workshop was well-received, with 87% recommending the workshop to others.

Conclusions

Results suggest that the FSAC: Immunotherapy program increases caregivers’ perceived knowledge about immunotherapy and their perceived confidence in talking to their doctor about immunotherapy. Our results highlight the potential benefits of immunotherapy education and support for cancer caregivers.

Acknowledgements

These workshops and analysis were funded in part by generous support from AstraZeneca, Bristol-Myers Squibb, and Merck.

References

1. Northouse L, Williams AL, Given B, McCorkle R. Psychosocial care for family caregivers of patients with cancer. J Clin Oncol. 2012;30(11):1227-1234.

2. Harvey A, Amsellem M, Suarez R. Informational and emotional support utilization and needs of lung cancer patients and caregivers: Results from a national education program. J Canc Educ 2014;29(Suppl 1).

3. Harvey A, Amsellem M, Suarez R. Information and emotional support utilization among cancer caregivers: results from a national sample of education program attendees. Psycho-Oncology 2015;24(Suppl. 2):160.

4. Saxton C, Amsellem M, Suarez R. Caregiver participation in a psychoeducational cancer support program: results from a national sample. J Canc Educ 2016;31(Suppl 1).

Ethics Approval

This study was conducted under IRB-exempt protocols [category 45 CFR 46.101(b) 2].

Correspondence: Thomas Duhen (thomas.duhen@providence.org)

Background

Microsatellite stable colorectal cancers are known to harbor low mutation burden, a limited immune infiltration of the tumor and thereby respond poorly to immunotherapy.

Methods

Here we describe a case of microsatellite stable metastatic colon cancer with an outstanding anti-cancer response. Briefly, the patient was a 53-year-old man with a T4bN2aM1 metastatic colon cancer diagnosed in 2012. The primary tumor was resected at time of diagnosis and the patient underwent several cycles of chemotherapy from 2012 to 2017 in an effort to treat the multiple lesions present in the liver considered unresectable at time of surgery. In 2018, the patient’s CEA levels began to rise and he underwent a left hepatectomy to remove a slow growing mass.

Results

Analysis of the liver resection by flow cytometry revealed a large number of immune cells infiltrating the tumor, a phenotype distinct from the other colorectal liver metastases (CRLMs) we analyzed to date. Most importantly, more than 60% of the infiltrating immune cells were CD8 T cells, one third of which co-expressed the markers CD39 and CD103, a phenotype characteristic of tumor-reactive T cells. Those CD39+CD103+ CD8 T cells expressed high levels of PD-1 and CTLA-4 and were highly proliferative in vivo as demonstrated by Ki-67 staining. Using multiplex IHC, we confirmed that this pattern was unique to the recurring CRLM as very few CD8 T cells were found in the primary tumor. Sequencing of the CDR3 region of the TRB genes expressed by the CD8 T cells isolated from the CRLM supported the flow cytometry data, with the presence of one unique clonotype accounting for one-third of all the CDR3 sequences in the CD8 T cells, indicative of a strong local expansion. We are now performing whole-exome and RNA sequencing on both tissues in order to identify the mutations recognized by those tumor-infiltrating CD8 T cells and determine if those mutations were already present in the primary tumor or emerged during the metastatic transformation process.

Conclusions

This case shows that even cancers with low tumor mutation burden can lead to strong immune responses overtime and this knowledge should be considered to establish patient’s treatment plans in the future.

Ethics Approval

All subjects signed written informed consent approved by the Providence Portland Medical Center Institutional Review Board (IRB protocol no. 06-108A).

Consent

Consent was obtained for publication of this abstract.

Correspondence: Peter Lee (plee@coh.org)

Background

Estrogen receptor positive (ER+) breast cancer tumors are generally poorly immune infiltrated tumors as compared to triple negative breast cancer (TNBC). In TNBC the presence of tumor infiltrating lymphocytes (TILs) is predictive of response to chemotherapy and associates favorably with patient survival. However, in ER+ breast cancer the relationship between T cell infiltration and disease is less clear. Exhausted CD8+ T cells has been described in murine models of chronic disease and recently several human carcinomas. Here we profile human breast tumors for the functional, phenotypic, and transcriptional profile of exhausted CD8+ T cells and their correlation with survival in patients.

Methods

Fresh surgical tumor specimens were obtained from consented patients. Single cell suspensions were analyzed by flow cytometry for immune phenotyping and functional assessment of cytokine production by T cells. Single cell sorted T cells were subjected for whole transcriptome RNA sequencing and bulk sorted T cells were submitted for T cell receptor repertoire sequencing by Adaptive Biotechnologies. Transcriptional profiles generated from single cell sequencing were utilized for signature score analysis of outcomes in public TCGA and Metabric data sets.

Results

Exhausted CD8+ T cells were identified in both ER+ and TNBC. Functional and phenotypic analysis identified T cells to have a reduced capacity for IFNγ, TNFα, and IL-2 production. Further phenotyping of exhausted CD8+ T cells revealed a loss of expression of both the IL-7 receptor alpha (CD127) and KLRG1, suggesting terminal differentiation of these cells. Single cell and T cell receptor sequencing revealed distinct transcriptional and clonal signatures of exhausted CD8+ T cells. T cell receptor analysis demonstrated increased oligoclonality of exhausted CD8+ T cells as compared to other CD8+ TILs. Finally we demonstrate that a transcriptional signature of exhausted CD8+ denotes relapse-free survival in ER+ breast cancer patients.

Conclusions

This work is the first to demonstrate a T cell subset beneficial to survival outcomes of ER+ breast cancer patients. Immunotherapeutic interventions to expand exhausted CD8+ T cell populations may offer improved overall survival in breast cancer patients.

Ethics Approval

Fresh tissues were obtained from patients who gave institutional review board (IRB)-approved written informed consent prior to inclusion in the study (City of Hope IRB 05091, IRB 07047, and IRB 14346).

Correspondence: Mohammad Farhad (farhad@ohsu.edu)

Background

Background: CD8 T cells are a critical component of anti-tumor immunity and enhancing their function can promote tumor elimination. The levels of extracellular galectin-3 (Gal-3), a carbohydrate binding protein, are increased during tumor progression in many cancers and higher levels of Gal-3 are associated with a poor prognosis. Interestingly, intracellularly, Gal-3 functions through protein-protein interactions and is upregulated in T cells following T cell receptor (TCR) stimulation in the presence of IL-2. However, the function and mechanisms by which intracellular Gal-3 regulates CD8 T cell responses are poorly understood.

Methods

Wild type (WT) or Gal-3-/- OT-I CD8 T cells were adoptively transferred into WT mice and then stimulated with soluble ovalbumin protein along with an agonist anti-OX40 mAb or rIL-2/IL-2 mAb complexes (IL-2c). The frequency, phenotype, and effector/memory status of the adoptively transferred CD8 T cells was determined in the peripheral blood (days 7 and 14) and spleen (day 29) by flow cytometry.

Results

In the current study, we demonstrate that Gal-3-deficient CD8 T cells exhibited no defects in early (36 hrs) activation or proliferation. In contrast, Gal-3-/- CD8 T cells exhibited decreased survival and a reduced capacity to develop into memory cells following stimulation with cognate antigen plus agonist anti-OX40 mAb or IL-2 in vivo. Decreased survival of Gal-3-/- T cells was associated with increased apoptosis (Annexin V) and occurred in a cell-intrinsic manner. Additional studies are underway to evaluate the extent to which Gal-3 regulates T cell-mediated anti-tumor immunity.

Conclusions

Together, these data implicate intracellular Gal-3 as a critical mediator of OX40-mediated CD8 T cell survival and memory formation following antigen exposure.

Correspondence: Craig Slingluff (cls8h@virginia.edu)

Background

T-cell retention in epithelial tumors is mediated by transmembrane proteins (CD49a, CD49b, CD103) known as retention integrins (RI)[1]. Expression of RI correlates with cytotoxicity of CD8+ T-cells in human epithelial tumors [2-3] and with immunosuppression by regulatory T-cells in murine models [4]. Recently, we reported that RI expression by CD8+ cells was modified by exposure to IL2, IL4, IL10, and TGFβ, as well as by T-cell receptor (TCR) stimulation [5]. TGFβ has also been shown to upregulate CD103 in human CD4+ memory cells [6] but investigation into RI expression by CD4+ cells has otherwise been sparse. We hypothesized that expression of RI by human CD4+ cells would be modulated by cytokines and TGFβ in a pattern similar to CD8+ cells.

Methods

Peripheral blood mononuclear cells isolated from five human donors were stimulated with mouse anti-CD3 and anti-CD28 for 24 hours. Cells were then cultured with TGBβ (5 ng/mL), IL2 (1000 CU/mL), IL4 (20 ng/mL), or IL10 (100 ng/ml) for six days. Integrin expression was quantified by flow cytometry. Data generated by flow cytometry was analyzed using FlowJo software (Tree Star), and results were compared using Student’s two-tailed T test.

Results

As shown in Table 1, TGFβ upregulated expression of CD49a and CD103 in TCR stimulated (TCR+) CD4+ cells relative to TCR+ controls (p = 0.02,

Conclusions

When comparing the expression of RI by CD4+ cells to our prior findings in CD8+ cells[5], we noted both common themes and significant differences. TGFβ upregulates CD49a and CD103 on both TCR+CD8+ and TCR+CD4+ cells. In contrast, the effects of IL2 and IL4 differ based on cell type. IL2 upregulates CD49a in TCRnegCD8+ cells but downregulates CD49a in TCR+CD4+ cells. IL4 decreased CD49a expression by TCR+ CD8+ cells, but decreased CD49a expression only in TCRnegCD4+. This heterogenous response to cytokine stimulation emphasizes the complexity of RI regulation and raises the question of whether regulation of RI expression is CD4+ subtype specific. This data supports the need for further investigation into the regulation of RI expression by CD4+Th1, CD4+Th2, and CD4+T regulatory cell populations.

References

1. Konkel J, Zhang D, Zanvit P, Chia C, Zangarle-Murray T, Jin W, Wang S, Chen W. Transforming Growth Factor-beta Signaling in Regulatory T cells Controls T Helper-17 Cells and Tissue Specific Immune Responses. Immunity. 2017; 18:660-674.

2. Boutet M, Gautheir L, Leclerc M, Gros G, de Montpreville V, Theret N, Donnadieu E, Mami- Chouaib F. TGFbeta Signaling Intersects with CD103 Integrin Signaling to Promote T-Lymphocyte Accumulation and Antitumor Activity in the Lung Tumor Microenvironment. Cancer Res. 2016; 76(7):1757-69.

3. Ling K, Dulphy N, Bahl P, Salio M, Maskell K, Piris J, Warren B, George B, Mortensen N, Cerundolo V. Modulation of CD103 Expression on Human Colon Carcinoma-Specific CTL. J Immunology. 2007; 178:2908-2915.

4. Anz D, Mueller W, Golic M, Kunz W, Rapp M, Koelzer V, Ellermeir J, Ellwart J, Schnurr M, Bourquin C, Endres S. CD103 is a hallmark of tumor infiltrating regulatory T cells. International Journal of Cancer. 2011; 129: 2417-2426.

5. Melssen M, Olson W, Wages N, Capaldo B, Mauldin I, Mahmutovic A, Hutchison C, Melief C, Bullock T, Engelhard V, Slingluff C. Formation and phenotypic characterization of CD49a, CD49b, CD103 expressing CD8 T cell populations in human metastatic melanoma. Oncoimmunology. 2018; 7(10): 10.1080/2162402X.2018.1490855

6. Watanabe R, Gehad A, Yang C, Scott L, Teague J, Schlapbach C, Elco C, Huang V, Matos T, Kupper T, Clark R. Human skin is protected by four functionally and phenotypically discrete populations of resident and recirculating memory T cells. Sci Transl Med. 2015; 7:279ra39

Ethics Approval

This study was approved by the UVA IRB-HSR (IRB 10598).

Correspondence: Edwin Parra (erparra@mdanderson.org)

Background

Myeloid-derived suppressor cells (MDSCs) are a heterogeneous population of cells that consists of myeloid progenitor cells and immature granulocytes or polymorphonucleares (PMN), immature macrophages (M), and immature dendritic cells (DCs). MDSCs play a critical role in tumor-associated immunosuppressive function, which is important in the effective immunotherapies for cancer. The aim of this study was analyze of full range of MDSCs to investigate their role in the microenvironment of non-small cell lung carcinomas (NSCLCs).

Methods

TMAs containing 247 patients with stage II/III NSCLCs (adenocarcinomas, ADC=162; squamous cell carcinomas, SCC=85) were analyzed using a multiplex immunofluorescence panel contained CD68, CD11b, Arg-1, CD66b, CD33, CD14, and a pancytokeratin (Figure 1). Densities of different phenotypes and neighbor distances were correlated with clinicopathologic features and outcomes.

Results

Using the median density of MDSCs we identified overall higher densities of those cells in SCC than ADC, with significant P value of total granulocytic (CD66b+CD11b+, P=0.016) cells, PMN-MDSCs (CD11b+CD66b+CD33+, P=0.021), and M-MDSCs (CD11b+Agr1+CD14+CD33+, P=0.009, Figure 1 and Table 1). In ADCs the densities of M2 macrophages (CD68+Arg1+CD11b+) were higher in solid pattern than other histologic patterns (acinar, lepidic, papillary, P=0.038); and the total macrophages CD68+ were higher in stage III than in stage I and II (P=0.075), Figure 2. Current smokers was showed higher densities of PMN-MDSC than never-smokers (P=0.031). In addition, EGFR mutant tumors had fewer M2 macrophages, total granulocytic and M-MDSCs than WT patient tumors (P=0.009, P=0.009 and, P=0.043; respectively, Figure 2). A 10μm distant radius from malignant cells compared with the other distant analyzed showed higher densities of CD68+ and CD68+CD11b+ myeloid cells in smaller and EGFR mutant tumors than in larger (P=0.024 and P=0.028; respectively) and WT (P=0.039 and P=0.034; respectively) tumors in ADC. Similarly, higher densities of PMN-MDSCs were observed in stage II tumors than stage I and III in SCC (P=0.029) at this distant. Patients with ADC and a high number of total granulocytic cells (P=0.081) showed worse prognosis than patients with lower total granulocytic cells.

Conclusions

While SCCs showed the highest densities of MDSCs as compared to ADCs, our data suggest MDSCs are related with clinicopathologic characteristics that can increase the risk of tumor progression, such as solid pattern, advanced stage, smoking status and prognosis. Additional studies need to be done to understand much better the role of MDSCs in lung cancer.

Ethics Approval

The study received approval from the MD Anderson Cancer Center Institutional Ethics Review Board. Tthe Lung Cancer Moon Shot projects through the National Cancer Institute’s Cancer Center Support Grant (P30CA016672), and a grant from the Cancer Prevention and Research Institute of Texas (MIRA-RP160688).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P546).

See text for description

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2 (abstract P546).

See text for description

Correspondence: Tullia Bruno (tbruno@pitt.edu)

Background

Current FDA-approved immunotherapies aim to reinvigorate CD8+ T cells, but the contribution of the humoral arm of the immune response in human cancer remains poorly understood [1,2]. B cells can mediate anti-tumor immunity by presenting antigen , producing tumor-specific antibodies and immunomodulatory cytokines [3-5]. Head and neck squamous cell carcinoma (HNSCC) can be induced by human papillomavirus (HPV+) and carcinogens (HPV-), and the immune infiltrate in the two etiologies has been reported to be distinct, in particular, more B cells in HPV+ HNSCC patients [6]. Further, the presence of B cells in HNSCC correlates with increased patient survival [6]. Our study seeks to differentiate B cell phenotype and function in HPV+ and HPV- HNSCC and identify putative immunotherapeutic targets.

Methods

We utilized a multi-level approach to categorize B cells in HNSCC patients. Single cell RNA sequencing (scRNAseq) was performed on CD45+ tumor infiltrating lymphocytes (TIL) and peripheral blood leukocytes (PBL) from HPV+ and HPV- HNSCC patients. Data was analyzed using the Seurat R toolkit in R studio. HNSCC TIL and PBL were stained via multiparameter flow cytometry panel using the Cytek Aurora to allow for unbiased analysis of B cell subsets via computational spectral unmixing. Paraffin embedded slides from HNSCC primary tumors were utilized for single parameter immunohistochemistry of CD20 (n=25 each for HPV + and HPV - disease), and multispectral immunofluorescence was performed on matched tumors from scRNAseq for CD20, CD4, CD8, CD68, PanCK, DAPI and Foxp3.

Results

We demonstrated distinct trajectories for B cells in HPV+ and HPV- disease. B cell signatures in HPV- HNSCC patients were predominantly memory B cells and plasma cells, while the signatures in HPV+ HNSCC were naïve and germinal center B cells. We quantified B cells and CD4 T cells in tertiary lymphoid structures (TLS), and the presence of germinal center-rich TLS was associated with HPV+ disease. GC-rich TLS within the tumor bed of HPV+ patients correlated with increased overall survival. Independent of HPV status, we have identified an atypical B cell signature increased in the HNSCC TME compared to healthy and inflamed tonsil tissue with no cancer. This atypical B cell subpopulation is also elevated in the periphery of patients with advanced, metastatic disease.

Conclusions

Characterization of B cell phenotype and function in HNSCC is important for devising new therapeutic options for patients. Development of therapeutics to enhance B cell responses in the TME should be prioritized as a compliment to T-cell mediated therapies.

References

1.Economopoulou, P., Kotsantis, I. & Psyrri, A. The promise of immunotherapy in head and neck squamous cell carcinoma: combinatorial immunotherapy approaches. ESMO Open 1, e000122 (2016).

2. Yuen, G. J., Demissie, E. & Pillai, S. B lymphocytes and cancer: a love-hate relationship. Trends Cancer 2, 747–757 (2016).

3. León, B., Ballesteros-Tato, A., Misra, R. S., Wojciechowski, W. & Lund, F. E. Unraveling effector functions of B cells during infection: the hidden world beyond antibody production. Infect Disord Drug Targets 12, 213–221 (2012).

4. Bruno, T. C. et al. Antigen-Presenting Intratumoral B Cells Affect CD4+ TIL Phenotypes in Non-Small Cell Lung Cancer Patients. Cancer Immunol Res 5, 898–907 (2017).

5. Germain, C. et al. Presence of B cells in tertiary lymphoid structures is associated with a protective immunity in patients with lung cancer. Am. J. Respir. Crit. Care Med. 189, 832–844 (2014).

6. Wood, O. et al. Gene expression analysis of TIL rich HPV-driven head and neck tumors reveals a distinct B-cell signature when compared to HPV independent tumors. Oncotarget 7, 56781–56797 (2016).

Ethics Approval

This study was approved by the local Institutional Review Board under protocol UPCI 99-069, and patients provided informed consent.

Correspondence: Anil Shanker (ashanker@mmc.edu)

Background

The interaction between the central nervous system and immune system underscores an intricate neuro-immune network.

Methods

We were intrigued by the expression of glutamate receptors (GluRs), particularly ionotropic GluR3 (AMPAR subtype 3, GluA3) and metabotropic GluR1 on T-lymphocytes, which were upregulated following TCR activation. We blocked group I metabotropic GluRs and AMPA by specific antagonists during T-cell stimulation and analyzed various parameters of T-cell functions in vitro and tumor-bearing mice.

Results

We found a delayed CD8+T-cell activation exhibited by downregulated CD25, CD69 and CD44 molecules, resulting in impaired proliferation and cytolytic activity against orthotopic mammary tumors expressing a cognate low-avidity antigen in vivo or specific peptide-loaded target cells in vitro. While GluR+CD8+T-cell frequency ranges about 30% of total CD8+T-cells, treatment with glutamate receptor antagonists profoundly affects T-cell functions. These effects are mediated by transient reduction in phosphorylation of molecules such as Lck/Akt as well as calcium, indicating interaction between TCR and GluR.

Conclusions

Data suggest a costimulatory effect of GluRs on T-cell function. Therefore, optimized glutamate agonists may overcome immunosuppression in tumor microenvironment and boost antitumor T-cell therapy.

Correspondence: Rosanna Wustrack (rwustrack@gmail.com)

Background

Undifferentiated Pleomorphic Sarcoma (UPS) is the most common STS subtype in older adults. Despite advancements in systemic treatment for other solid tumors, 5- and 10-survival has remained stagnant for advanced stage UPS. With the development of immune-based therapeutic strategies, it is becoming increasingly clear that the microenvironment plays a critical role in the surveillance and potential elimination of nascent tumor cells [1]. In STS for example, subtype specific effects of checkpoint inhibition were observed in patients with UPS, suggesting that UPS derived antigens modulate tumor-specific adaptive immune responses [2]. While lymphocyte subtypes have been quantified in large panels of STS patients, these studies are hampered by the diverse heterogeneity and relatively small numbers of UPS patients analyzed [3]. We aimed to identify tumor infiltrating immune cell subset and tumor specific antigens in UPS using a robust and well-validated multi-parameter flow cytometry platform and determine if there was any correlation between the immune phenotype and other recognized prognostic factors such as tumor size and depth.

Methods

All patients undergoing resection of grade 2 or 3 UPS were eligible for inclusion. We performed single cell multi-parameter flow cytometry using 15 immune-based markers.

tSNE plots were used to demonstrate the relative immune landscape within UPS tumors. The relative frequency of tumor infiltrating lymphocytes were then correlated with clinical outcomes data.

Results

From January 1, 2017, to July 31, 2019, 20 resected UPS specimens were collected and successfully immunoprofiled using well-validated multi-parameter flow based cytometry. We identified an increased frequency of CD8+ tumor infiltrating lymphocytes in smaller tumors (8cm) tumors representing 25.82% and 15.01% respectively (Figure 2).

Conclusions

We were able to successfully identify the tumor infiltrating immune cell subsets in 20 freshly resected extremity undifferentiated pleomorphic sarcoma specimens. Lymphocytes represented a greater proportion of the immune cell population in smaller tumors compared to larger tumors; specifically, CD8+ T cells were seen at a lower frequency in the tumors greater than 8cm compared to those less than 8cm. As patients with high grade tumors greater than 8cm historically have had worse outcomes, this data shows that the host immune environment may play a role in these clinical outcomes and could be a future target for sarcoma immunotherapy (Figure 3).

References

1. Ribatti D. The concept of immune surveillance against tumors. The first theories. Oncotarget. 2017;8(4):7175–7180. doi:10.18632/oncotarget.12739.

2. Tawbi HA, Burgess M, Bolejack V, et al. Pembrolizumab in advanced soft-tissue sarcoma and bone sarcoma (SARC028): a multicentre, two-cohort, single-arm, open-label, phase 2 trial. Lancet Oncology. 2017;18(11):1493–1501. doi:10.1016/S1470-2045(17)30624-1.

3. D'Angelo SP, Shoushtari AN, Agaram NP, et al. Prevalence of tumor-infiltrating lymphocytes and PD-L1 expression in the soft tissue sarcoma microenvironment. 2015;46(3):357–365. doi:10.1016/j.humpath.2014.11.001.

Ethics Approval

The study was approved by the UCSF Committee on Human Subject Research, approval number 15-16384

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P549).

See text for description

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2 (abstract P549).

See text for description

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3 (abstract P549).

See text for description

Correspondence: Micah Benson (mbenson@ksqtx.com)

Background

Although immunotherapy with PD-1/PD-L1 antagonists has significantly advanced patient care, the majority of cancer patients currently do not benefit from checkpoint inhibitor therapies. To identify novel targets for the treatment of PD-1 insensitive cancers, we developed a novel Immune-CRISPRomics platform that enabled genome-wide CRISPR/Cas9 screens in primary T cells in an in vivo setting.

Methods

In vivo genome-wide CRISPR/Cas9 screens were conducted by introducing sgRNA libraries into Cas9-expressing TCR-Tg T cells, followed by adoptive transfer into tumor-bearing hosts. Targets enabling enhanced TCR-Tg T cell function in the context of the tumor microenvironment were identified and validated by follow-on efficacy studies.

Results

Our CRISPRomics platform identified clinically active molecules, such as PD-1, and also predicted recent clinical failures. In addition, we identified multiple targets that enhanced anti-tumor T-cell function similar to or better than PD-1 as a monotherapy. The anti-tumor activity of targets was assessed across a collection of PD-1 sensitive and PD-1 refractory syngeneic tumor models. One of the targets identified, IO-7, was found to possess robust activity across multiple PD-1 refractory models. We found that inhibition of IO-7 leads to long-term T-cell memory and prevents tumor growth upon re-challenge. Moreover, mechanistic follow-up studies demonstrate that IO-7 inhibition leads to an expansion of central memory T-cell subsets, which have been implicated in driving the durable clinical response of checkpoint inhibitors. We then extended these findings further with an in-vivo double CRISPR screen to examine more than 2,000 pairwise combinations of active sgRNAs in-vivo. These combination studies identified CD8 T-cell gene pairs that when inactivated gave anti-tumor activity that exceeded any single gene inactivation from the primary screen and greatly exceeded the efficacy of PD-1 monotherapy.

Conclusions

We describe an Immune-CRISPRomics platform which 1) identified multiple functional targets with therapeutic promise and 2) functionally screened combinations of these targets to reveal the most potent T-cell target combinations for the treatment of PD-1 resistant solid tumors.

Correspondence: Ivan Borrello (iborrell@jhmi.edu)

Background

In multiple myeloma (MM), marrow infiltrating lymphocytes (MILs) have shown increased anti-tumor reactivity and proliferative capacity compared to their peripheral blood counterparts. Despite the recent identification of a core signature for tissue resident memory T cells (TRM), a definitive bone marrow-resident T cell phenotype has yet to be identified.

Methods

Here, we used multiparametric flow cytometry to perform single-cell analysis on CD8+ T cells from MM patients to identify those subsets that are enriched in the bone marrow (BM) compared to peripheral blood (PB).

Results

The activation and tissue retention marker CD69 identified a population of CD8+ MILs that is virtually absent in PB and displays peculiar phenotypic and functional features compared to its CD69-negative counterpart. Despite the virtual absence of CD103 expression, CD69+ MILs displayed increased levels of CXCR6 and variable levels of CD101, thereby sharing resemblance with the canonical TRM phenotype. Interestingly, CD101 has been associated with both tissue residency and exhaustion. As such, its variable expression may depend on MM burden. CD69+ MILs consistently displayed high levels of the bone marrow homing chemokine receptor, CXCR4 and low/absent expression of the follicular marker, CXCR5. The expression pattern of these extracellular markers further characterizes the BM-resident phenotype and suggests a unique role for the BM microenvironment in anti-tumor immunity. Analysis of exhaustion and senescence markers clearly separated the CD69+ MILs population from both PBL and CD69- T cells because of intermediate expression of exhaustion markers such as TIGIT, TIM3, and PD1, increased CD27, and markedly reduced CD57 levels. This unique phenotype is associated with preserved cytotoxic activity and proliferative capacity. In human T cells, TCF1 and T-bet have a reciprocal gradient of expression. Stem-like features are associated with increased TCF1 and reduced T-bet levels. Interestingly, CD69+ MILs were enriched for TCF1+ and T-bet low T cells suggesting the presence of a stem-like subset in the BM of MM patients. As a whole, CD69+ MILs are distinct from their PB and CD69- counterparts as they uniformly displayed higher levels of TCF1, CD27, CXCR6 and high CXCR4 levels and reduced expression of both CD57 and T-bet.

Conclusions

In conclusion, we identified a subset of MILs that appears to display a stem-like/resident phenotype. Studies are ongoing to determine whether this population can be exploited and/or enriched to improve the efficacy of adoptive T cell therapy with in MM, especially in the post-transplant setting where activated MILs therapy has already been proven feasible.

Correspondence: S. Rhiannon Jenkinson (rhiannon.jenkinson@crl.com)

Background

The tumor microenvironment (TME) is a complex network, consisting of the tumor, blood vessels, stromal and immune cells, and soluble factors. The immune system plays an important role in combating tumor growth and various tractable immuno-oncology targets have been identified. We have developed a range of primary human assays to model the TME in vitro, this provides a multi-cellular platform in which to study the mode of action of novel immune-modulators and to identify biomarkers before moving into the clinic. This platform models multiple aspects of anti-tumor immune responses and has been validated with standard of care therapeutics. The assays include: T cell or NK cell-mediated tumor killing, myeloid/macrophage assays, exhausted T cells, Th1/Th17/iTreg differentiation and regulatory T cell suppression assays.

Methods

The ability of therapeutics to reverse or inhibit T cell exhaustion was performed using primary human cells. T cell-mediated tumor killing using Pembrolizumab and Ipilimumab (PI) as controls, or 3D tumour cell spheroid ADCC assays (Trastuzumab control) were performed using an IncuCyte. Myeloid/macrophage cells were differentiated from monocytes in tumor-conditioned media (TCM), phenotyped and tested for their phagocytic capability using anti-CD47 as a control. The TME is associated with increased Treg numbers and low frequency of Th1 or Th17 cells. Many therapeutics aim to shift the balance away from Treg, towards Th1 or Th17 cells. Naïve CD4+ T cells were therefore differentiated into iTreg, in the presence or absence of a USP7 inhibitor to examine if iTreg differentiation altered.

Results

PI enhanced tumour cell killing in a 2D co-culture assay. Trastuzumab drove enhanced ADCC-mediated killing in a 3D spheroid format. TCM drove the generation of macrophage which had a phenotype similar to tumour associated macrophage, CD25lo, CD127lo, CD184hi, CD80lo, CD163hi, CD68lo and MHCIIlo, produced IL-10 and VEGF and were suppressive. Reduced iTreg generation, without significant alteration in Th1/17 frequency was observed in the presence of the benchmark compound and resulting iTreg were less suppressive compared to non-treated cells. nTreg suppression assays were performed and the USP7 inhibitor had a partial effect.

Conclusions

We present an immuno-oncology platform to model the TME in human cells which enables a rapid assessment of the immunomodulatory capacity of therapeutics. The 3D assays represent a more complex system in which to support translational drug discovery, sitting alongside T cell-mediated tumor killing, myeloid/macrophage assays, Th1/Th17/ iTreg differentiation and nTreg assays. Together these assays help to define diverse aspects of the immune response in the TME.

Ethics Approval

The study was approved by HRA NRES South West, Bristol (UK), approval number 15/SW/0029.

Correspondence: Ana Micaela Carnaz Simoes (ana.micaela.carnaz.simoes@regionh.dk)

Background

Platelets (PLTs) are well-known players during cancer progression. For several cancers, an increased number of circulating PLTs correlates with poor prognosis. Not only can these cells modulate angiogenesis and directly bind cancer cells to facilitate the metastatic process, but they can also protect cancer cells from immune attack by mechanisms that are poorly understood [1,2]. Studies in the autoimmunity field, have shown that PLTs form aggregates with T cells, downregulating T cell activation, proliferation and interferon-ɣ production [3,4]. Nevertheless, no similar studies have been conducted in the context of cancer.

Methods

We investigated the presence of circulating PLT-lymphocytes aggregates in solid and haematological cancers. Furthermore, we assessed the effect of PLT binding on T cell proliferation and phenotype. To that purpose, cryopreserved peripheral blood mononuclear cells were analyzed by multicolor flow cytometry and fluorescent microscopy for PLT-T cell aggregates, as well as CD4 and CD8 subpopulations. Lastly, to evaluate the effect PLT binding on T cell anti-tumor reactivity, in vitro cytotoxic response was continuously monitored for 40 hours, using the xCELLigence technology.

Results

Our preliminary results show that, compared to healthy donors, cancer patients have an increased number of PLT-lymphocyte aggregates, specially CD8+ T cells. Furthermore, PLTs seem to bind preferentially to a specific T cell phenotype and modulate T cell tumor reactivity. Data will be further discussed in detail at the meeting.

Conclusions

These results suggest that binding of PLTs have an impact on T cell activity, which could greatly influence the immune response against tumors. However, further studies are needed to confirm our preliminary results.

References

1. Borsig L. The role of platelet activation in tumor metastasis. Expert Rev Anticancer Ther. 2008;8(8):1247-1255. doi:10.1586/14737140.8.8.1247.

2. Bambace NM, Holmes CE. The platelet contribution to cancer progression. J Thromb Haemost. 2011;9(2):237-249. doi:10.1111/j.1538-7836.2010.04131.x.

3. Zamora C, Canto E, Nieto JC, et al. Functional consequences of platelet binding to T lymphocytes in inflammation. J Leukoc Biol. 2013;94(3):521-529. doi:10.1189/jlb.0213074.

4. Zamora C, Cantó E, Nieto JC, et al. Binding of Platelets to Lymphocytes: A Potential Anti-Inflammatory Therapy in Rheumatoid Arthritis. J Immunol. 2017;198(8):3099-3108. doi:10.4049/jimmunol.1601708.

Correspondence: Justin Kline (jkline@medicine.bsd.uchicago.edu)

Background

We have previously demonstrated that high-affinity, leukemia-specific CD8+ T cells are efficiently deleted in leukemia-bearing hosts in a manner that depends on cross-presentation of the antigen by splenic Batf3-lineage dendritic cells [1 ,2].

Methods

Here, we generated a T cell receptor transgenic mouse strain (Tg101) that expresses TCR-α and -β chains from a CD8+ T cell clone specific for a native H-2Kb-restricted antigen on C1498 leukemia cells.

Results

In naïve mice, Tg101 thymocytes predominantly develop along the CD8+ lineage, and exist as naïve CD8+ T cells in peripheral lymphoid organs. Tg101 T cells do not recognize other H-2Kb tumor lines in vitro, suggesting that their cognate antigen may indeed be leukemia-specific. In leukemia-bearing mice, high-affinity 2C CD8+ T cells specific for the H-2Kb-restricted model antigen, SIYRYYGL, rapidly encounter antigen (day 3-4) via its cross-presentation by splenic CD8alpha+ DCs, proliferate briefly, and are largely deleted by day 14. Conversely, antigen encounter by Tg101 CD8+ T cells in mice with C1498 leukemia is delayed (day 7-8), and primarily occurs in the liver - a prominent site of leukemia progression. Subsequently, Tg101 T cells expand, upregulate expression of co-inhibitory receptors (PD-1, LAG-3, TIM-3 and TIGIT), adopt a transcriptional program enriched for upregulation of known genes associated with T cell anergy/exhaustion (Cblb, Egr2/3, Ctla4 Pcdc1, Tox) and produce effector cytokines poorly, all consistent with the acquisition of a dysfunctional state. Antigen-experienced Tg101 cells isolated from leukemia bearing mice fail to expand upon transfer into secondary C1498 cell-challenged mice, suggesting that the dysfunctional state induced in Tg101 T cells becomes irreversible at some point. Surprisingly, antigen encounter by Tg101 T cells in leukemia-bearing mice occurs independently of CD11c+ cells, suggesting that cross-presentation of the Tg101 antigen is not required for in vivo proliferation. Ex vivo cross-presentation assays also fail to demonstrate detectable cross-presentation of the Tg101 antigen by DCs subsets isolated from leukemia-bearing mice. Rather, in vivo Tg101 T cell proliferation is strikingly blunted in mice harboring H-2Kb-/- C1498 AML cells, suggesting that the antigen is directly presented to Tg101 T cells.

Conclusions

Collectively, our results reveal that CD8+ T cell tolerance is effectively generated in leukemia-bearing animals. However, the nature of the leukemia antigen and the context in which it is presented are critical factors in dictating the underlying mechanism of tolerance that ensues. A search for the Tg101 antigen and investigation of the epigenetic landscape of naïve versus tolerized Tg101 T cells are ongoing.

Acknowledgements

Janet D. Rowley Discovery Fund

References

1.Zhang L, Chen X, Liu X, Kline DE, Teague RM, Gajewski TF, Kline J. CD40 ligation reverses T cell tolerance in acute myeloid leukemia. Journal of Clinical Investigation. 2013 May 123: 1999-2010.

2. Kline DE, MacNabb BW, Chen X, Chan WC, Fosco D, Kline J. CD8α+ Dendritic Cells Dictate Leukemia-Specific CD8+ T Cell Fates. J Immunol. 2018 Dec 15;201(12):3759-3769

Ethics Approval

The study was approved by The University of Chicago Ethics Board, approval number ACUP 71945.

Correspondence: Dario Vignali (dvignali@pitt.edu)

Background

Regulatory T cells (Tregs) maintain peripheral tolerance;[1] however, in cancer, Tregs dampen anti-tumor immunity, contributing to disease progression.[2,3] Neuropilin-1 (NRP1) is required for intratumoral Treg stability and Treg-specific knockout of NRP1 attenuates tumor growth.[4] Transcriptional analyses revealed that intratumoral NRP1-deficient Foxp3+ Tregs develop an effector phenotype, characterized by interferon-gamma (IFNγ) production and decreased suppression of conventional T cells.[5] Whereas NRP1 is constitutively expressed on mouse Tregs, expression on human Tregs is activation-driven and thus may be modulated by immune processes. Furthermore, the role of NRP1 in maintaining human Treg stability amidst proinflammatory signals is not known. We hypothesize that (1) surface NRP1 expression marks highly suppressive human Tregs, (2) NRP1 expression is driven by proinflammatory signals in the tumor microenvironment, and (3) NRP1 ligation is required for maximal suppressive function.

Methods

Phenotypic profiling of peripheral blood (PBL) and tumor infiltrating lymphocytes (TILs) from head and neck squamous cell carcinoma (HNSCC), melanoma, non-small cell lung, ovarian, and colorectal cancer patients was conducted by flow cytometry. Treg function was evaluated in vitro by a micro-scale suppression assay, which measures the ability of Tregs to suppress CD8+ T cell proliferation. Paired with our phenotyping, we cultured Tregs under various stimulatory conditions to query drivers of NRP1 expression. These included numerous cytokine conditions, T cell receptor stimulation, co-culture with antigen presenting cells, as well as blockade of specific costimulation/inhibitory receptors.

Results

NRP1+ Tregs are greatly enriched in cancer patient PBL and TIL across numerous malignancies, and the degree of enrichment negatively correlates with disease-free survival in HNSCC. NRP1+ Tregs upregulate a module of functional receptors that mark highly suppressive intratumoral Tregs, including TIGIT, ICOS, and CCR8, as well as a network of TNFRSF members. Correspondingly, NRP1+ Tregs suppress cytotoxic T cell proliferation to a greater degree than their NRP1- counterparts, which can be inhibited using blocking antibodies. NRP1+ Tregs are enriched in vitro in low interleukin-2 (IL-2) conditions as well as upon T cell activation. Proinflammatory cytokine exposure, such as IFNγ, increases NRP1 expression in a subset of cancer patient samples.

Conclusions

NRP1+ Tregs constitute a more suppressive human Treg subset based on their enhanced regulatory phenotype and function. Given the increased expression of proliferation and survival markers, especially under destabilizing conditions, NRP1 confers a survival advantage to these Tregs, allowing them to persist and function amidst such signals. Therefore, destabilizing intratumoral Tregs with NRP1 blockade may complement other T cell therapies such as anti-PD1 blockade.

References

1. Vignali, D. A. A., Collison, L. W. & Workman, C. J. How regulatory T cells work. Nat. Rev. Immunol. 8, 523–532 (2008).

2. Tanaka, A. & Sakaguchi, S. Regulatory T cells in cancer immunotherapy. Cell Res. 27, 109–118 (2017).

3. Liu, C., Workman, C. J. & Vignali, D. A. A. Targeting regulatory T cells in tumors. FEBS J. 283, 2731–2748 (2016).

4. Delgoffe, G. M., Woo, S.-R., Turnis, M. E., Gravano, D. M., Guy, C., Overacre, A. E., Bettini, M. L., Vogel, P., Finkelstein, D., Bonnevier, J., Workman, C. J. & Vignali, D. A. A. Stability and function of regulatory T cells is maintained by a neuropilin-1–semaphorin-4a axis. Nature 501, 252–256 (2013).

5. Overacre-Delgoffe, A. E., Chikina, M., Dadey, R. E., Yano, H., Brunazzi, E. A., Shayan, G., Horne, W., Moskovitz, J. M., Kolls, J. K., Sander, C., Shuai, Y., Normolle, D. P., Kirkwood, J. M., Ferris, R. L., Delgoffe, G. M., Bruno, T. C., Workman, C. J. & Vignali, D. A. A. Interferon-γ Drives T reg Fragility to Promote Anti-tumor Immunity. Cell 169, 1130–1141.e11 (2017).

Ethics Approval

This study was approved by the local Institutional Review Board under protocol UPCI 99-069, and patients provided informed consent.

Correspondence: Fabio Conforti (fabio.conforti@ieo.it)

Background

We previously demonstrated that patients’ sex is a variable that influences the response to anticancer immunotherapy: in RCTs comparing immunotherapy-containing regimens to standard treatments in several solid tumors, men obtained larger survival benefit than women when treated with anti-CTLA4 or anti-PD-1 drugs as monotherapy, whereas women experienced larger benefit from anti-PD-1 or PD-L1 combined with chemotherapy.[1,2]

We hypothesized that such heterogeneity of response is due to differences in the molecular mechanisms that govern anticancer immune-response in men and women.

Methods

We analyzed 7 public datasets containing data on genome-wide RNA expression analysis of NSCLC samples.[3-9]

We used gene-expression data to estimate the abundance of 64 different cell-types in the microenvironment of each tumor sample, using single sample gene-set enrichment analysis (ssGSEA) according to a previously validated computational algorithm.[10,11]

The Enrichment Scores (ES) of the gene-sets specifically associated with each of the 64 different cell-types were first calculated in tumors of men and women and then compared using a multivariable linear model that adjusted for other covariates including patients’ age, stage at diagnosis, tumor-histotype and smoking status.

Finally, we performed a meta-analysis of the adjusted ES sex-related differences obtained in each single dataset using a random-effects model.

The false discovery rate (FDR) was used to correct for multiple comparisons.

Results

2590 tumors were analyzed: 1529 tumors (60%) were from men and 1047 (40%) from women. (table1) The majority of patients (95%) had stage I-III tumors, and did not receive immunotherapy.

Figure. 1 shows results obtained in each dataset as well as the pooled meta-analytic results. (Figure 1)

In the pooled analysis, the immune cell-types found enriched in the tumor microenvironment of women at a FDR cut-off

1)Dendritic cells;

2)CD4+ T-cells (including naive and central memory T-cells);

3)B-cells (including Memory and Class-switched memory B-cells),

4)Eosinophils;

5)Mast-cells;

6)Granulocyte-monocyte progenitor cells (GMP);

Immune-cells found enriched in the tumors of women at FDR cut-off

1)Regulatory T-cells;

2)Gamma-delta T-cells;

3)Natural killer T-cells;

4)Macrophages (both M1 and M2);

5) CD4+ effector memory T-cells;

The only immune cell-type found enriched in the tumor infiltrates of men at FDR

Conclusions

We found relevant sex-based differences in the cell-type composition of the immune-infiltrate of NSCLC.

The higher levels of immune-suppressive cells, such as GMP, Macrophage-M2 and Regulatory T-cells observed in tumors of women may explain lower efficacy of anti-CTLA4 or anti-PD1 given as monotherapy, and the need to add chemotherapy to counteract their inhibitory effects.

References

1. Conforti F, Pala L, Bagnardi V, et al. Cancer immunotherapy efficacy and patients' sex: a systematic review and meta-analysis. Lancet Oncol. 2018; 19:737-746.

2. Conforti F, Pala L, Bagnardi V, et al. A. Sex-based heterogeneity in response to lung cancer immunotherapy: a systematic review and meta-analysis. J Natl Cancer Inst. 2019 May 20. pii: djz094. doi: 10.1093/jnci/djz094. [Epub ahead of print].

3. Cancer Genome Atlas Research Network. Comprehensive molecular profiling of lung adenocarcinoma. Nature 2014; 511:543-50.

4. Cancer Genome Atlas Research Network. Comprehensive genomic characterization of squamous cell lung cancers. Nature. 2012; 489:519-25.

5. Schabath MB, Welsh EA, Fulp WJ, et al. Differential association of STK11 and TP53 with KRAS mutation-associated gene expression, proliferation and immune surveillance in lung adenocarcinoma. Oncogene 2016; 35:3209-16.

6. Shedden K, Taylor JM, Enkemann SA, et al. Gene expression-based survival prediction in lung adenocarcinoma: a multi-site, blinded validation study. Nat Med. 2008; 14:822-7.

7. Rousseaux S, Debernardi A, Jacquiau B, et al. Ectopic activation of germline and placental genes identifies aggressive metastasis-prone lung cancers. Sci Transl Med. 2013;5:186ra66.

8. Sato M1, Larsen JE, Lee W, et al. Human lung epithelial cells progressed to malignancy through specific oncogenic manipulations Mol Cancer Res. 2013;11:638-50.

9. Noro R, Ishigame T, Walsh N, et al. A Two-Gene Prognostic Classifier for Early-Stage Lung Squamous Cell Carcinoma in Multiple Large-Scale and Geographically Diverse Cohorts. J Thorac Oncol. 2017; 12:65-76.

10. Aran D, Hu Z, and Butte A, xCell: digitally portraying the tissue cellular heterogeneity landscape. Genome Biology. 2017; 18:220.

11. Subramanian A, Tamayo P, Mootha V, et al. Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles. PNAS 2005; 102:15545-15550.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P556).

See text for description

Correspondence: Matthew Walters (mwalters@arcusbio.com)

Background

The tumor microenvironment (TME) generates high levels of adenosine which binds to A2aR and A2bR receptors on immune cells to inhibit their activity. We have previously shown that AB928, a dual A2aR/A2bR antagonist, blocks the immunosuppressive effects of adenosine in human cultured cells and mouse syngeneic tumors. Herein, we describe studies to assess the contribution of A2bR to adenosine-mediated immunosuppression.

Methods

CD14+ monocytes were differentiated with IL-4/GM-CSF for 6 days to generate monocyte-derived dendritic cells (moDC). Cells were taken for NanoString analysis or placed in a mixed lymphocyte reaction (MLR) with CD4+ T cells. PBMCs were activated with CD2/3/28 beads +/- NECA or CGS-21680 for 24 hours and supernatants run on a CXCL5 ELISA.

Results

Using publicly-available gene expression databases, we identified that T cells and other non-myeloid cells predominantly express the A2a adenosine receptor. In contrast, most myeloid cells express both A2aR and A2bR with cross-presenting CLEC9A+ dendritic cells (DCs) having the highest levels of A2bR. This pattern of adenosine receptor expression makes DCs a prime target for dual adenosine receptor antagonism in the TME. Experiments in vitro showed that moDC have the highest levels of A2bR expression compared to other cultured cells such as monocytes and macrophages. As expected, adenosine-differentiated moDC showed a decreased ability to stimulate IFN-γ secretion in a MLR. Notably, AB928 attenuated IFN-γ suppression by adenosine and showed significantly greater rescue than a comparable A2aR-specific antagonist. NanoString gene expression profiling identified 39 genes (>2.0 fold change, p <0.05) regulated by adenosine in moDC. These gene expression changes were largely rescued with AB928 but not significantly changed with the A2aR-specific antagonist. Importantly, stimulation of primary human DCs with the dual A2aR/A2bR agonist NECA reproduced several of the adenosine-driven gene expression changes observed in moDC.

To further highlight the importance of A2bR, NECA or the A2aR-specific agonist CGS-21680 were added to human PBMC cultures. Both agonists significantly upregulated the ligand CXCL5 with equivalent rescue by AB928 and the A2aR-specific antagonist when stimulated by CGS-21680. However, upon NECA stimulation, AB928 showed significantly greater inhibition of CXCL5 compared to the A2aR-specific antagonist (86% vs 72% of control, p<0.01), demonstrating that A2bR signaling in PBMCs significantly contributes to gene expression changes.

Conclusions

Collectively, these studies demonstrate an important role for A2bR in adenosine-mediated immunosuppression and provide a mechanistic rationale for stimulation of anti-tumor immune responses with the dual adenosine receptor antagonist AB928, which is currently undergoing evaluation in several Phase 1/1b trials.

Correspondence: Robert Ferris (ferrisrl@upmc.edu)

Background

Immunotherapy has significantly advanced HNSCC treatment. However, mechanisms of resistance to immunotherapy mediated by intrinsic and extrinsic factors produced in the Tumor microenvironment (TME) formed by stroma, cancer and immune cells leads to a relatively low response rate of only 15-20%. The emerging technology of single-cell RNA sequencing can be employed to investigate heterogeneous cell populations within the TME by interrogating the transcriptome of individual cells. This allows for transcriptomic mapping of the HNSCC TME at the single cell level.

Methods

17 HNSCC specimens were processed for scRNAseq with matched peripheral blood leukocytes (PBL) from treatment-naïve HPV+ and HPV- patients. Following manual and enzymatic tumor dissociations, cell suspensions were sorted into CD45+ (tumor infiltrating leukocytes: TIL) and CD45- (tumor and associated stromal) cells. 10x Genomics 3’ single cell libraries were sequenced on a NextSeq500 (Illumina). Cellular transcriptomes from all samples were aggregated and normalized using the CellRanger pipeline. Downstream bioinformatic analyses were performed using the Scanpy package.

Results

We identified 30 different TME cell clusters, with 22 clusters formed by immune cells (PBL and TIL) while CD45- non-immune cells formed 8 clusters. In the immune cell compartment, therapeutically and prognostically relevant cell types were identified (cytotoxic T cells, NK cells, regulatory T cells). Sub-clustering revealed important cell states such as activation, senescence and exhaustion in these immune cell types. Inflamed tumors (n=8) showed a marked increase in the number of tumor-infiltrating CD8+ T cells (43%) versus non-inflamed tumors (17%). Immune checkpoint receptor levels were elevated in CD8+ T cells from inflamed tumors. Novel gene expression signatures representing IFNγ, IFNα and previously described allograft rejection were uniquely associated with these CD8+ T cells, while TNF, apoptosis and hypoxia signaling were predominant in CD8+ T cells from non-inflamed tumors. Modeling putative interactions between immune and fibroblasts, endothelial cells and cancer cells identified potential therapeutic targets to manipulate the immune-stroma-cancer interaction in the TME.

Conclusions

Heterogeneous cell populations of immune, stroma and cancer cells can be found in the TME of HNSCC. Transcriptomic characterization using single-cell RNA sequencing allows to identify diverse cell states as well as to model putative cell-to-cell interactions.

Ethics Approval

The study was approved by the University of Pittsburgh’s Institutional Review Board (#99-069).

Correspondence: John Ligon (jligon1@jhmi.edu)

Background

Patients with metastatic osteosarcoma (OS) have 5 year overall survival <25% [1]. We aim to understand the immune architecture of the tumor microenvironment (TME) of OS, with the goal of harnessing the immune system as a major therapeutic strategy for the treatment of patients with OS.

Methods

Immunohistochemistry (IHC) slides from 66 formalin-fixed paraffin-embedded OS tissue blocks were digitally analyzed. Laser capture microdissection and RNA extraction was performed on 13 OS specimens, and gene expression profiles were obtained utilizing RNA seq and analyzed. Tumor infiltrating lymphocytes (TILs) were isolated from 25 freshly obtained OS specimens and analyzed by multiparameter flow cytometry (MFC).

Results

Digital image analysis of IHC demonstrated significantly higher infiltrating immune infiltrating cells in the pulmonary metastases compared to primary bone tumors, particularly at the tumor-normal lung interface. There is increased expression of multiple immune checkpoint molecules at the interface, including programmed cell death 1 (PD-1), T cell immunoglobulin mucin-3 (TIM-3) and lymphocyte activation gene 3 (LAG-3) (Figure 1). Gene expression profiling revealed increased CD8 T cell and resting CD4 memory T cell signature at the interface compared to the tumor interior, and a strong M2 macrophage signature in both regions. Genes related to neutrophil and myeloid cell chemotaxis and known to be associated with polymorphonuclear myeloid-derived suppressor cells (PMN-MDSCs) were shown to be highly expressed in the interface region (Figure 2). MFC revealed that TILs isolated from pulmonary metastases had higher expression of PD-1 and other immune checkpoint molecules, and these TILs were more capable of producing the effector molecules IFN-gamma and granzyme B (Figure 3).

Conclusions

Pulmonary metastatic OS lesions are more highly infiltrated with immune cells than primary bone lesions, particularly at the interface. These immune cells are capable of producing cytotoxic cytokines and proteases by MFC, suggesting that these TILs may be tumor-specific and capable of acting against the tumor. The inability of these immune cells to penetrate further into the tumor interior may represent an adaptive immune response by the tumor through a combination of repellent cytokines and inhibition of TILs via upregulation of multiple immune checkpoint molecules and recruitment of PMN-MDSCs. Treatment with a combination of therapies targeting both inhibitory immune checkpoint molecules and the immunosuppressive myeloid compartment may unleash these immune cells and allow them to destroy metastatic OS tumor cells.

References

1. Mirabello L, Troisi RJ, Savage SA. Osteosarcoma incidence and survival rates from 1973 to 2004: Data from the surveillance, epidemiology, and end results program. Cancer. 2009;115(7):1531-1543.

Ethics Approval

This study was approved by Johns Hopkins University's Ethics Board, approval number FWA00005752.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P559).

Pulmonary metastases have high infiltration of immune cells

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2 (abstract P559).

Gene expression analysis of interface compared to interior

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3 (abstract P559).

Metastasis TILs express higher checkpoints and cytokines

Correspondence: Séverine Mouraud (severine.mouraud@gustaveroussy.fr)

Background

We reported that PDL-1 expression on circulating CD8+ T cells was a predictor of bad prognosis and of resistance to CTLA4 blockade in melanoma [1]. PD-L1 expression is currently assessed by IHC stainings and implemented into clinical practice for selecting cancer patients eligible to anti-PD(L)1 immunotherapies. However, PD-L1 IHC stainings have a low sensitivity compared to flow cytometry and fall short in identifying the very nature of PD-L1 expressing cells in the TME. We aimed at better defining the PD-1/PD-L1 expression in the Tumor Micro Environment (TME) by using flow cytometry.

Methods

We analyzed 132 fresh tumors by flow cytometry: primitive tumor samples of Urothelial Carcinoma (UC; n=45), liposarcoma (n=19), Ovarian Cancers (OC; n=24), Kidney carcinoma (n=17), Head and Neck Squamous Cell Carcinoma (HNSCC; n=11), Thyroid (n=8) and liver metastases of Colorectal Cancers (CRC; n=8). We evaluated membrane expression of PD-1, PD-L1, TIM-3, LAG-3, MHC-I and MHC-II in the TME using a BD Fortessa X20 and analyzed the data with the Kaluza software.

Results

We found a high variability of tumor infiltrating leukocytes across individuals (from 0% to 90% of CD45+ cells) and across tumor types (median range from 4% for CRC metastasis to 20% for OC and NSCLC primitive tumor samples). However, the level of T-cell infiltrates (CD3+) (TILs) was consistently around 60% across histologies with an equal proportion of CD4+ and CD8+ T-cells (except for UC and HNSCC with a CD4/CD8 ratio= 2). Unsupervised analysis of immunophenotyping data could not cluster patients by histological type or by the expression of TIM-3, LAG-3, MHC-I or MHC-II on T cells. More surprisingly, we found that cancer patients could be grouped by clusters with mutually exclusive patterns of PD-1 and PD-L1 expression on T-cells. Indeed, patients were either PD-1POS/PD-L1NEG (18% of patients), PD-1NEG/PD-L1POS (45% of patients), or PD-1NEG/PD-L1NEG (30% of patients) on their tumor-infiltrating T-cells with an identical phenotype between CD4+ and CD8+ T-cells.

Conclusions

Patients' TIL are either PD-1POS or PD-L1POS or PD-1NEG/PD-L1NEG. This pattern is uniform across all TILs and mutually exclusive across patients, independently from their tumor histology. This result could have a major implication in terms of anti-tumor activity of anti-PD(L)1 therapies and patient selection for clinical trials and routine practice.

References

1. Jacquelot N, Roberti MP, Enot DP, Predictors of responses to immune checkpoint blockade in advanced melanoma. Nat Commun. 2017; 8(1):592.

Ethics Approval

All samples were collected in accordance with French Law (MR-004).

Correspondence: Victor Engelhard (vhe@virginia.edu)

Background

Integrins CD49a and CD49b can mediate retention of lymphocytes in peripheral tissues by binding to collagens type IV and type I, respectively. Their expression is upregulated on CD8+ tumor infiltrating lymphocytes (TIL) compared to circulating lymphocytes and density of CD49a+ TIL is associated with improved patient outcome. Little is known about how expression of these integrins is regulated in vivo and the functional capacity of the cells expressing them. We hypothesized that CD49a and CD49b expression evolves over time as T cells differentiate in the tumor microenvironment (TME) and that integrin expression is associated with T cell function, by directly supporting effector function or localization.

Methods

To address these hypotheses, CD8 TIL from an implantable breast carcinoma model, BRPKp110, were for expression of CD49a, CD49b and functional markers. Where indicated, mice were treated daily with 5ug/mouse FTY720. For transfer experiments, OT1 cells were activated in vivo with ovalbumin, polyIC and anti-CD40 vaccination. After 5d, CD49b+ cells were purified and transferred into tumor-bearing mice.

Results

In early stage tumors (d14), T cells were predominantly CD49b single positive (SP) or CD49aCD49b double positive (DP). Later (d23), CD49b SP cells largely disappeared, and CD49a SP cells appeared, while DP cells remained unchanged. After treatment with FTY720, to block further T cell infiltration, the switch from CD49b SP to CD49a SP cells was even more pronounced, suggesting the change is not due to newly infiltrating CD49a SP cells. Regardless of timepoint, CD49b SP cells were antigen-responsive, whereas DP and CD49a SP cells had features of exhaustion. To test whether the switch in integrin expression depends on antigen engagement or environmental factors, we transferred activated CD49b SP OT1 cells into OVA-negative BRPKp110 tumor-bearing mice. In these tumors, CD49b SP cells upregulated CD49a, followed by downregulation of CD49b (Figure 1). Preliminary in vitro experiments showed that restimulation of T cells in presence of CD49a/CD49b ligands, collagen type I or IV, strongly inhibit IFNγ expression, suggesting a potentially direct negative role for CD49a or CD49b in anti-tumor immunity.

Conclusions

Together, our data suggest that in CD49b SP CD8 TIL evolve over time to upregulate CD49a, followed by downregulation of CD49b, as TIL gain an exhausted phenotype. Furthermore, the switch in expression is dependent on the microenvironment, not antigen-specific differentiation. Future experiments will assess which environmental factors play a role in CD49a upregulation, and how processes of integrin expression relate to the antigen-dependent generation of exhaustion.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P561).

CD49a and CD49b expression on transferred OT1 cells

Correspondence: Daniel Olive (daniel.olive@imcheck.fr)

Background

Anti-tumoral response of Vg9Vd2 T cells requires sensing of phosphoantigens accumulated in malignant cells through binding of butyrophilin 3A(BTN3A). Moreover, an unknown partner located in human Chr6 was shown to be mandatory to BTN3A-mediated Vg9Vd2 T cell activation in murine models. Here, we identified butyrophilin 2A (BTN2A), which is located to Chr6, as a requirement for BTN3A-mediated Vg9Vd2 T cell cytotoxicity against cancer cells.

Methods

CRISPR-Cas9-mediated inactivation of BTN2A1/2A2 isoforms was performed in Daudi, K562 and HEK-293T cells. Vg9Vd2 T cells expanded from healthy PBMCs were co-cultured with wild-type or BTN2AKO cells +/- BrHPP(1 μM), HMBPP(0.1 μM) or zoledronate(45 μM), or anti-BTN2A mAb, and Vg9Vd2 T cell degranulation (%CD106αβ+ cells), and intracellular TNFα and IFNγ assessed after 4h. Mouse T cell hybridoma 53/4 expressing TCRVg9Vd2-MOP were co-cultured overnight with NIH3T3 murine fibroblasts transfected with BTN3A- and/or BTN2A-encoding plasmids +/-HMBPP(10 μM), or increasing doses of HMBPP or anti-BTN3 20.1 mAb. BTN2A transcript expression in normal vs. tumoral tissue was analyzed using GEPIA tool. Anti-BTN2A mAb staining was performed on human samples of primary AML, cervical and pancreatic carcinoma and assessed by flow cytometry.

Results

Degranulation and intracellular IFNg/TNFa (n=6) were abolished in Vg9Vd2 T cells co-cultured with BTN2AKO Daudi, K562 and HEK-293T cells compared to wild-type, in all conditions tested including anti-BTN3 20.1. Murine cells do not express no BTN2A1 or BTN3A orthologs and are unable to activate human Vg9Vd2 T cells. Ectopic expression of BTN2A and BTN3A combination but neither BTN2A or BTN3A alone in murine NIH3T3 cells, allows triggering of IL-2 secretion in mouse 53/4-TCRVg9Vd2-MOP reporter cells in presence of HMBPP or 20.1 mAb in dose-dependent manner. Anti-BTN2A mAb was able to suppress Vg9Vd2 T cell degranulation/cytokine secretion against cancer cell lines and activation of mouse 53/4-TCRVg9Vd2-MOP reporter by BTN2A/BTN3A-expressing NIH3T3 in a dose-dependent manner. BTN2A transcript was significantly up-regulated in pancreatic, ovarian and cervical carcinoma vs. normal tissue. Extracellular BTN2A protein was detected in primary hematological and solid tumors.

Conclusions

Here, we show that BTN2A is mandatory for BTN3A-mediated function in human Vg9Vd2 T cells. Moreover, concomitant BTN2A and BTN3A expression empowers murine T cells with activation through Vg9Vd2 TCR, opening new roads for mouse models of Vg9Vd2 T cell anti-tumoral responses. We describe an anti-BTN2A able to suppress Vg9Vd2 T cell function, and we show BTN2A expression in primary tumors. These results are relevant for understanding Vg9Vd2 T cell antitumoral immunity triggered by phosphoantigens and amino-bisphosphonates.

Correspondence: Sean Downing (sean.downing@ultivue.com)

Background

Immune cell infiltration and exhaustion has been a key question within the field of immunotherapy research. In particular, regulatory T cells (T-regs) has been identified as one of the cell types responsible for modulation of the immune response. Spatial resolution through multiplex immunofluorescence allows researchers to identify cell types only identifiable through multiple-marker expression and their behavior and relationships with other cells. In this work, we set out to identify three main cell types and their interactions in the tumor microenvironment of non-small cell lung cancer samples: cytotoxic T cells, T-regs, and tumor cells. We used the UltiMapper I/O T-reg panel (CD4, CD8, FoxP3, pan-Cytokeratin) for characterization of cell phenotypes and additional testing to assess the reproducibility of the assay for downstream studies.

Methods

Two sets of 5 serial sections from four different tissue sets (one human tonsil and three different NSCLCs) were stained with the UltiMapper I/O T-reg panel (CD4, CD8, FoxP3, pan-Cytokeratin) on the Leica Biosystems BOND RX autostainer. Whole-slide images were acquired using the ZEISS AxioScan.Z1, without the need for linear unmixing. Analysis was performed using Indica Labs HALO® software Indica Labs. Samples were analyzed for immunophenotyping and spatial mapping. Coefficients of variation (CVs) were calculated based on resulting data. All image acquisition and analysis was performed using the same analysis template for each sample type.

Results

Detection of single-marker and multi-marker phenotypes were identified. T-regs were identified by CD4+/FoxP3+. Cytotoxic T cells were identified by CD8+. Tumor regions were mapped with pan-Cytokeratin+. Cell population comparisons and proximity analysis was used to compare T-helper:cytotoxic T cell, T-reg:cytotoxic T cell, and cytotoxic T cell:tumor cell ratios. Overall CV’s were less than 15% across all markers across serial sections. This included the total cell counts and average signal intensity.

Conclusions

The results presented here indicate that the UltiMapper I/O T-reg panel offers robust and reproducible multiplex immunofluorescence data for the detection and characterization of immune cell infiltration and T-cell regulation. Histological standards for coefficients of variation in IHC based assays are typically

Correspondence: Haydn Kissick (haydn.kissick@emory.edu )

Background

CD8 T cells are a critical part of the immune response to tumors. With the increased use of immunotherapies to treat many cancers understanding what leads to highly CD8 T cell infiltrated tumors is crucial in the improvement of these therapies. Especially since the number of CD8 T cells within the tumor has been shown to predict the response to anti-PD1 blockade and survival. Recent work in CD8 T cell immunology described how CD8 T cells respond to chronic diseases, finding two subsets of CD8 T cells within tumors. One is a stem-like CD8 T cell and the other is an effector CD8 T cell with cytotoxic capabilities [1]. Understanding how tumor-specific CD8 T cells activate and differentiate to produce an effective CD8 T cell response with both stem-like CD8 T cells and effector CD8 T cells is crucial to determining why certain tumors are highly infiltrated by CD8 T cells while others are not.

Methods

To understand how tumor-specific CD8 T cells respond to prostate cancer we have made a prostate cancer model which expresses the viral LCMV glycoprotein (GP) which acts a tumor-specific antigen. Using this model, we have used this model to study tumor-specific CD8 T cell activation by adoptively transferring LCMV GP specific TCR transgenic P14 CD8 T cells into TRAMPC1-LCMV-GP bearing mice.

Results

We have data showing that to have a highly infiltrated there must be a population of stem-like CD8 T cells which are capable to differentiating into effector CD8 T cells and supporting the CD8 T cell response against the tumor. We have found when tumor-specific CD8 T cells are activated they acquire an undifferentiated but activated program, upregulating CD44, PD1 but retaining high TCF1 and CD62L expression (Fig 1, C,E). They also upregulate effector molecules such as CD43 slower and completely fail to upregulate CX3CR1 (Fig 1, C). These undifferentiated activated CD8 T cells do not acquire a typical effector program that is seen in an acute viral infection such as LCMV Armstrong (Fig 1, A,B), and proliferate slower than P14 CD8 T cells activated in LCMV Armstrong (Fig 1, D).

Conclusions

Tumor-specific CD8 T cells do not acquire an effector program after activation and instead gain an undifferentiated but activated phenotype. Based on this we hypothesize that tumor-specific CD8 T cells are activated in the TDLN and differentiate to become the stem-like CD8 T cells within the tumor, promoting the anti-tumor CD8 response.

References

1. Im, S. J., et al. "Defining Cd8+ T Cells That Provide the Proliferative Burst after Pd-1 Therapy." Nature 537.7620 (2016): 417-21. Print.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P564).

CD8 T cell activation in viral infection and tumor model

Correspondence: Douglas McNeel (dm3@medicine.wisc.edu)

Background

DNA vaccines, relative to other vaccine strategies, are an attractive approach for cancer treatment given their safety, easy manipulation, scalability, stability and economical manufacturing. While a DNA vaccine has been approved for canine melanoma, early phase clinical studies have been generally disappointing [1]. Further studies to understand their mechanism of action, in order to improve their immunogenicity, are therefore needed. We have previously found that passive uptake of DNA by dendritic cells (DC) and macrophages led to degradation, while passive uptake of DNA by B cells led to transcription of encoded genes and ultimately activation of antigen-specific CD8+ T cells.

Methods

We used splenic B cells from wild type C57/Bl6 mice, DCs from Flt3L-treated splenocytes from C57/Bl6 mice, CD8 T-cells from OT-1 mice carrying the T-cell receptor specific for the SIINFEKL peptide epitope, and DNA encoding ovalbumin. Antigen-presenting cells (APCs) were treated with different cytokines for in vitro activation and to evaluate antigen presentation following passive uptake of DNA. The activation and proliferation of antigen-specific CD8+ T-cells was then analyzed by flow cytometry and IFN-γ release by ELISA.

Results

We found that B cells are the only APC subtype that could take up DNA and transcribe it to mRNA, but they could not present the antigen themselves to CD8+ T cells. B cells require direct contact with DCs to cross present the antigen, and only co-culture of B with DC led to T-cell activation and proliferation. We also found that CD4+ T cells, including antigen-non-specific CD4+ T cells, act as helper cells to promote this presentation, and this function could be replaced by directly providing IL-4.

Conclusions

B cells serve as primary antigen-presenting cells for DNA vaccine, but their ability to subsequently activate CD8+ T cells is mediated through DC. Further evaluation of this mechanism may lead to novel vaccine delivery approaches that could improve the immunogenicity of DNA vaccines.

References

1. Wahren B. & Liu M. (2014) DNA Vaccines: recent developments and the future. Vaccines 2, 785

Correspondence: Seema Kashyap (dr_skashyap@hotmail.com)

Background

Inflammation in Uveal Melanoma (UM) is linked to a bad prognosis. Infiltration with an inflammatory infiltrate increases with disease progression but does not seem to inhibit metastasis. Genetic predictors for metastatic tumor behavior is the loss of BRCA1-associated protein 1 (BAP1) expression. NF-κB is a principal coordinator of innate immunity and inflammation and has emerged as an essential endogenous tumor promoter. We hypothesize that genetic changes not only influence the immunological microenvironment but also drive metastasis in UM and that NC-NFκB proteins are the consequence of a highly-inflammatory profile.

Methods

In our study, seventy-five patients are recruited for a period of 4 years (2013-2017). Based on the expression of CD3 (infiltrating lymphocytes) and CD68 (infiltrating macrophages), we divided our study cohort into two categories: UM with inflammation and UM without inflammation. Expression of BAP-1 and NC-NFκB proteins (RelB & p52/NFκB2) was evaluated using immunohistochemistry and western blotting. Real-time PCR was performed on 60 frozen tumor samples. To detect the presence of p52/RelB heterodimer by Co-immunoprecipitation was performed on five each case of the inflammation and non-inflammation group of UM. BAP-1 sequencing was performed on 10 cases.

Results

In the inflammation group, activation of NC-NFκB proteins found in 82% and 64% of cases while the loss of BAP-1 was observed in 82% of cases. Loss of BAP-1 protein along with activation of NC-NFκB proteins was seen in the 70% of cases of the inflammation group. Loss of BAP-1 along with activation of C-NFκB proteins was statically significant with inflammatory factors such as CD34+ (p=0.036), IL-6 (p=0.012), LBD>15mm (p=0.031) and epithelioid cell type (p=0.027). Our western blotting results supporting the immunohistochemistry results. In the inflammation group fold-change value of RelB(5.21) & NFκB2 (4.65) genes was reduced to 2.85 (RelB) & 2.34 (NFκB2) gene in the non-inflammation group. Mutation of BAP-1 was more frequently seen in the inflammation group than the non-inflammation group. Loss of BAP-1, along with the activation of NC-NFκB proteins, was associated with reduced metastasis-free survival and overall survival (p

Conclusions

Our study showed that in an inflammation group loss of BAP-1 showed the synergistic role with the activation of NC-NFκB proteins and are the poor prognostic indicator of overall and metastasis-free survival. By predicting those patients who have a poor prognosis might help to differentiate which patients require more aggressive local treatments and which of those systemic treatments.

Ethics Approval

The study was approved by the institutional ethics committee, All India Institute of Medical Sciences (IESC/T-417/2015).

Consent

Written informed consent was obtained from the patient for publication of this abstract and any accompanying images. A copy of the written consent is available for review by the Editor of this journal

Correspondence: Sean Downing (sean.downing@ultivue.com)

Background

Myeloid-derived suppressor cells (MDSCs) are developed to protect the host, and in the context of immuno-oncology these cells protect tumor cells through suppression of immune cell activity. MDSCs are not present in healthy environments, and only rise during pathological conditions associated with inflammation. Two main types of MDSCs have been studied in depth: monocytic MDSCs (M-MDSCs) and polymorphonuclear MDSCs (PMN-MDSCs). M-MDSCs are an intermediate state along the path of differentiation from myeloid cells to monocytes. PMN-MDSCs are an intermediate state along the path of differentiation from myeloid cells to neutrophils. Much of the research into MDSCs has been collected through flow cytometry because of the ability to immunophenotype single cells with multiple markers. Within tissue, research has been limited because of the current multiplex immunohistochemistry technologies’ inability to detect co-expression of multiple markers on single cells and measure a range of expression for dynamic markers. Here we assess the UltiMapper I/O MDSC panel (CD11b, CD14, CD15, HLA-DR) for characterization in a colorectal cancer cohort and for reproducibility studies.

Methods

A set of tissue microarrays (TMAs) containing multiple colorectal carcinoma samples was stained with the UltiMapper I/O MDSC kit. Staining was conducted on the Leica Biosystems BOND RX autostainer. Multiplex images were acquired with the ZEISS Axio Scan.Z1 slide scanner. TMAs were analyzed for cell phenotyping and spatial distribution using Indica Labs HALO software. Coefficients of variation (CVs) were calculated based on resulting data.

Results

Detection of co-expression of multiple markers in the panel and range of expression for HLA-DR was identified. M-MDSC were identified by CD11b+/CD14+/CD15-/HLA-DR- cells. PMN-MDSCs are CD11b+/CD14-/CD15+/HLA-DR-/low. Overall CV’s were less than 15% across all markers across serial sections. This included total cell counts and average signal intensity. Differences in densities of MDSCs were measured in various colorectal carcinoma samples.

Conclusions

Given the robustness and high reproducibility of the UltiMapper I/O MDSC panel, the identification and assessment of MDSC cell types in tissue is reliably attainable. This allow researchers to apply this technology to cohorts of colorectal cancer to assess the utility of measuring MDSCs in the context of immunotherapy-related studies and the overall immunosuppressive landscape.

Correspondence: Greg Delgoffe (delgoffeg@upmc.edu)

Background

Blockade of co-inhibitory ‘checkpoint’ molecules, PD-1 and CTLA-4, has induced impressive clinical responses in advanced tumors; yet only in a subset of patients. Limited success with checkpoint blockade therapy suggests other cell extrinsic or intrinsic mechanisms may be dampening an effective immune response. Cytotoxic CD8+ T cells (CTL) encountering chronic antigen and metabolic restriction can differentiate to a terminally exhausted state (Texh), marked by hyporesponsiveness and metabolic, epigenetic, and transcriptional dysfunction. While enrichment of this population in tumor is a negative prognostic factor, it remains unclear whether Texh are simply non-functional or instead possess tolerogenic or suppressive properties. Transcriptional profiling of tumor-infiltrating CD8+PD-1int (progenitor exhausted) CTL versus CD8+PD-1hiTIM-3+ terminally differentiated Texh reveals that exhausted cells express a pattern of genes associated with immune suppression. We hypothesize that Texh potentiate the suppressive microenvironment of solid tumor by autoregulation and inhibition of local immune responses.

Methods

Texh and were isolated from B16 melanoma by expression of inhibitory receptors (IR) and assayed in tandem in microsuppression assay with progenitor exhausted and CD4+FoxP3+ regulatory T cells. Murine melanoma clones with inhibited oxidative metabolism were generated by CRISPR-Cas9 and validated for ablated mitochondrial respiration by extracellular flux analysis. Enforced expression of CD39 in effector T cells was attained by murine retroviral vector delivery. Deletion of CD39 was accomplished through tamoxifen-inducible Cre recombinase expression in CD8+ T cells (E8icre-ERT2Entpd1flox/flox).

Results

When sorted directly from tumor, terminally exhausted (PD-1hiTim3+) Texh, but not PD-1 intermediate, progenitor exhausted CD8+ T cells, induce marked suppression of T cell effector responses, comparable to suppression from Foxp3+ Treg cells sorted from the same environment. Suppressive activity correlated with exposure to hypoxia, and exhausted T cells sorted from tumors engineered to be less hypoxic had significantly less suppressive capacity. Our data suggest hypoxia induces HIF1a-dependnet expression of the cell surface ectonucleotidase, CD39, which depletes extracellular ATP, contributes to generation of immunosuppressive adenosine, and has been previously associated with terminal exhaustion. CD39 overexpression on effector T cells can confer a diminished phenotype, and preliminary data using CD8-specific CD39 deletion suggests it confers the regulatory phenotype in tumor-infiltrating exhausted T cells.

Conclusions

Our data support a model that as CTL progress to terminal exhaustion, hypoxic exposure upregulates CD39, providing Texh a mechanism to suppress proinflammatory processes. These findings suggest Texh are not solely dysfunctional but are rather deleterious to anti-tumor immunity and may need to be drastically reprogrammed or deleted in order to alleviate immunosuppressive functions.

Correspondence: Marij Schoenmaekers-Welters (M.J.P.Schoenmaekers-Welters@lumc.nl)

Background

Apart from anogenital cancers, high-risk human papillomavirus type 16 (HPV16) can induce oropharyngeal squamous cell carcinomas (OPSCC). Interestingly, patients with HPV16-induced OPSCC respond better to standard (chemo)radiotherapy compared to HPV-negative OPSCC patients. We previously demonstrated that intratumoral HPV16-specific type 1 T-cell responses contributes to this improved outcome. To further decipher the impact of tumor immune contexture on standard therapy in HPV16-negative and HPV16+ OPSCC we performed in-depth high-dimensional studies.

Methods

HPV16 status was determined by GP5+/6+ PCR and p16ink4a staining. Archived formalin-fixed paraffin-embedded tumor tissues were used for multicolor immunofluorescent (IF) staining. Prior to therapy tumor samples were dispersed into single cells and analyzed by CyTOF, flow cytometry and single cell RNAseq. HPV16-specific T-cell reactivity was determined by proliferation and cytokine release using IL-2-expanded tumor-infiltrating lymphocyte (TIL).

Results

Intratumoral HPV16-specific T cells were detected in 5/9 HPV16+ and in 0/4 HPV16- OPSCC. The immune cell composition of HPV16+ tumors differed between T-cell non-reactive (IR-) and reactive (IR+) OPSCC as determined by CyTOF analysis. HPV16+IR- OPSCC were strongly infiltrated by B cells while HPV16+IR+ contained mainly T cells with a highly activated (CD38+, HLA-DR+ and/or PD1+) effector memory phenotype and more CD4+CD161+ and CD8+CD103+ T cells. Analysis of 75 HPV16+ OPSCC patients in the TCGA database demonstrated that high expression of CD4 (log-rank p=0.003), CD8 (log-rank p=0.007) or CD161 (log-rank p=0.0007), but not CD103, was associated with better overall survival. Importantly, flow cytometry analysis showed that these intratumoral CD161+ T cells produced higher levels of pro-inflammatory cytokines than their CD161-negative counterparts (p<0.0001).

Conclusions

HPV16+ OPSCC patients respond better to standard therapy due to infiltration of highly activated and functional HPV-specific effector memory T cells that express CD161 and/or Tbet into the tumor. These type 1-oriented T-cell inflamed OPSCC were concomitantly infiltrated with Tbet+Tregs, associated with improved survival but may hamper full control of tumor growth.

Acknowledgements

This study was financially supported by grants from the Dutch Cancer Society (2014–6696 and 2016–10726).

Ethics Approval

This study was approved by the Leiden University Medical Center Ethics Board (P07-112 and P08-197).

Correspondence: Mark White (mark.white@berkeleylights.com)

Background

Human T cells represent a highly heterogenous population that can recognize a wide variety of antigens through their T cell receptor. Dissecting how diverse populations of single T cells respond to antigenic stimulus is key to understanding immunity and disease. Berkeley Lights has developed a novel platform to assay the interaction of 1000s of individual T cells and antigen-presenting cells in an automated fashion in 24hrs. The T Cell Phenotype and Functional Analytics workflow enables simultaneous detections of T cell surface markers like CD137 and IFNɣ secretion at the single-cell resolution. Following analysis, individual clones can be selected and recovered alive for downstream expansion or genomic analysis.

Methods

On the Lightning, BLI’s proprietary light technology was used to load IFNɣ capture beads into NanoPens with volumes less than 1 nL. Next, single antigen-specific T cells were loaded and followed by antigenic peptide-pulsed T2 cells. In a small subset of the chip, T2 cells were loaded that were pulsed with an irrelevant peptide to serve as a negative control. Following overnight incubation, chips were assayed for IFNɣ on the capture beads and CD137 upregulation on T cells.

Results

Linking complex cellular phenotypes to genotype can be applied to a diverse range of cell interactions and applications including CAR construct screening and validation as well as TCR and antigen discovery and validation. In addition, the functional analysis of primary samples like TILs with low cell numbers based on multiple parameters is now possible

Correspondence: Amy Rosenberg (amy.rosenberg@FDA.hhs.gov)

Background

Colorectal cancers (CRCs) have very low response rates to current immunotherapeutic approaches with the exception of the microsatellite instability high (MSI-H) phenotype. To investigate the basis for such poor responsiveness, we evaluated expression levels of 6 well-known immune checkpoint genes (ICPs), as well as potential novel immune checkpoint related genes (ICPRGs).

Methods

The biological connectivity of 6 known ICPs with IFNγ and its downstream genes was examined by NGS in 79 stringently procured and processed CRC and their normal, patient matched intestinal epithelium pairs. Identification of novel IFNγ-induced molecules with potential ICPRG activity was sought.

Results

In our study, the 6 “classical” ICPs were statistically upregulated and correlated with IFNγ and 183 additional immune related genes in IFNγ positive (FPKM>1) tumors (Figure 1). By ICP coexpression analysis, we identified 3 IFNγ-induced genes (IFI30, GBP1 and GBP4) as potential novel ICPRGs. These 3 genes were statistically upregulated compared to normal tissue, expressed at relatively high abundance (average FPKM = 115) across 79 CRC pairs compared to the abundance of the 6 ICPs (average FPK = 12; Table 1 ) and correlated with expression of IFNγ and 172 immune-related genes in IFNγ positive tumors (Figure 2). Further evaluation of the TCGA database revealed commonality of IFNγ dependent expression of 6 ICPs and the 3 potential ICPRGs in 638 CRCs, 103 skin cutaneous melanomas (SKCMs), 1105 breast cancers (BCs), 184 esophageal cancers (ESCs), 416 stomach cancers (STCs) and 501 lung squamous carcinomas (LUSC).

Conclusions

The expression of higher abundance and novel ICPRG genes, including IFI30, GBP1 and GBP4 requires further evaluation of protein expression levels in tumors because these genes have the potential to be druggable targets for immunotherapy of CRC as well as other tumors.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P571).

ICP coexpression network

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2 (abstract P571).

ICPRG coexpression netweork

Correspondence:Arnold Zea (azea@lsuhsc.edu)

Background

In mammals, NO is generated by nitric oxide synthases (NOS 1, 2, 3). All type of NOS catabolizes L-arginine (L-Arg) to produce L-citrulline and nitric oxide (NO) the later known to have anti-tumor activity. L-Arg metabolizes arginase (ARG2) to synthesize polyamines necessary for tumor growth. NOS2 is commonly up-regulated by inflammatory mediators, and it produces NO as long as the molecule is intact and its substrate L-Arg is available. Because ARG2 and NOS2 compete for L-Arg, the metabolism of this amino acid will become a crucial component in modulating cell growth, cell death and resistance in renal cell carcinoma. We hypothesize that L-Arg availability in renal carcinoma cells expressing ARG2 regulates the activity and expression of NOS2 protein as a mechanism that induces resistance against IFNgamma-induced tumor activity.

Methods

300,000 Renca cells per well were plated in 6 well plates in RPMI media containing 1,000; 2,000; 4,000 and 6,000 micro-molar of L-Arg and culture for 48h. Then, the cells were stimulated with100U/ml of IFNgamma and cultured for an additional 24 h. At this time, supernatants and cells were harvested and tested for NOS2 protein by immunoblot and NOS2 activity by Greiss Assay (nitrite formation). Supernatants were collected every day during the length of the experiments, to determine L-Arg levels by High-Performance Liquid Chromatography (HPLC). As a control for the experiments, we used CL-19 renal carcinoma cells responsive IFNgamma stimuli.

Results

Here, we report that decreased availability of L-Arg in Renca cells blocked the expression of NOS2 protein and NO-induction after stimulation with IFNgamma. Furthermore, the addition of L-Arg to the cell cultures reverted the process. INFgamma receptor, NOS2 gene expression, ARG2 protein expression did not change under the different experimental conditions.

Conclusions

The lack of IFNgamma-induced responses could likely be due to the tumor’s ability to develop strategies of tumor escape and define a distinct mechanism by which L-Arg can regulate the activity of its associated NOS2 enzyme. Our plan, include testing if, in this phenomena, the inhibition of NOS2 expression and activity by L-Arg depletion occurs via inhibition of translation of NOS2 mRNA.

Acknowledgements

This work is supported by the Stanley S. Scott Cancer Center Funds, and NIH-R25GM121189, NCI-1P20CA202920-01AI grants.

Correspondence: Jessica Thaxton (Thaxton@musc.edu)

Background

The requisites for translation in T cells are poorly understood and how translation shapes the anti-tumor efficacy of T cells is unknown.

Methods

We used multiple innovative tools to dissect how translation shapes T cell lineages and impacts tumor growth. LC-MS/MS proteomics, cytokine conditioning, immunblotting, and a novel FACS-based assay of mRNA translation were employed to elucidate the molecular components that control translation in T cells. A novel in vitro tumor microenvironment assay was developed to measure the impact of the microenvironment on T cell translation. Adoptive transfer of T cells to tumor-bearing mice was employed to elucidate how translation is impacted in T cells in tumors, and fresh tumor infiltrating lymphocytes (TILs) from cancer patients were assessed to measure translation in human TILs. Pharmacological modulation of translation was used to test that remodeling protein synthesis is a new strategy to augment T cell-mediated tumor control.

Results

Here we demonstrate that memory T cells are primed by metabolic energy sensor AMP-activated protein kinase (AMPK) to undergo diminished translation relative to effector T cells. However, we show that memory T cells exhibit a remarkable capacity to enhance translation in tumors that effector T cells are unable to duplicate. Study of modulation of translation for applications in cancer immunotherapy revealed that translation is suppressed in CD8 TILs in multiple human tumor types, and that direct ex vivo pharmacological inhibition of the translation elongation step primes powerful T cell anti-tumor immunity by remodeling metabolic properties of T cells.

Conclusions

Our work elucidates the new finding that paths to translation shape CD8 T cell anti-tumor capability. We show that metabolic energy sensors shape protein synthesis in T cells and control anti-tumor immunity, and we demonstrate that direct modulation of translation is a powerful and unique strategy to shape the efficacy of T cells to combat tumor growth.

Correspondence: Buvana Ravashankar (bravashankar@rapt.com)

Background

The tumor microenvironment (TME) is characterized by deficiencies in oxygen and key nutrients, such as glucose and amino acids, resulting in an overall immune suppressive environment. Stromal cells and myeloid-derived suppressor cells (MDSC) within the tumor create a nutrient-poor environment that inhibits immune function and supports tumor growth. GCN2 (general control nonderepressible 2), a stress response kinase, plays a key role in sensing and modulating the response to amino acid deprivation. GCN2 activation in T cells leads to an induction of the integrated stress response pathway and subsequently to T cell anergy and apoptosis.

Methods

Culturing primary mouse or human immune cells under low nutrient conditions activates the GCN2 pathway limiting T cell proliferation and function. Treatment of these nutrient-deprived T cells with GCN2i resulted in rescue of CD8+ T cell proliferation and effector functions as measured by flow cytometry. In addition, GCN2 inhibition in MDSC alone fully reversed CD33+MDSC-induced T cell suppression and effector functions. Using murine syngeneic tumor models, we demonstrate that the pharmacologic inhibition of GCN2 in-vivo leads to an observed anti-tumor effect.

Results

Furthermore, GCN2 inhibition induced tumor specific immunity by increased antigen-specific T cell frequency and increased cytokine production. Our GCN2i is currently being evaluated to further elucidate the immune contribution in the tumor microenvironment.

Conclusions

The GCN2/eIF2α pathway is activated in immune cells during amino acid deprivation, and this induces a functional suppression of the immune response. Our results demonstrate that inhibition of GCN2 is an attractive approach for relieving T cell suppression and promoting anti-tumor activity, demonstrating GCN2 as a promising therapeutic target for the treatment of cancer.

Ethics Approval

The study was approved by RAPT Therapeutics Ethics Board, approval number FL0002

Correspondence: Paolo Vignali (pdv7@pitt.edu), Greg Delgoffe (gdelgoffe@pitt.edu)

Background

Regulatory T (Treg) cells are vital for maintaining immune homeostasis and preventing autoimmunity but represent a major barrier to robust cancer immunity as the tumor microenvironment (TME) actively recruits, activates, and promotes their differentiation. Tumor cells have deregulated cellular metabolism leading to a metabolite-depleted, hypoxic, and acidic TME. Highly glycolytic tumor infiltrating CD8 T cells are in direct competition with the tumor for glucose and oxygen which impairs their effector function. In contrast, Treg cells maintain their high suppressive function within the TME. Further, studies of in vitro induced and directly ex vivo Treg cells reveal a distinct metabolic profile compared to effector T cells. Thus, it may be that the altered metabolic landscape of the TME and the increased activity of intratumoral Treg cells are linked.

Methods

Fluorescent glucose tracers, extracellular flux analysis, isotopic flux analysis, and RNA sequencing were used to determine the metabolic profile of murine lymph node and B16 tumor resident Treg cells. Mice harboring a Treg specific deletion of the lactate transporter Slc16a1 (MCT1) were used to determine the in vivo significance of Treg cell lactate uptake.

Results

Here we show that Treg cells, especially tumor derived, do not robustly engage in glycolysis or consume glucose. Transcriptomic and functional analysis of glucose consuming Treg cells revealed a diminished Treg cell signature and reduced suppressive capacity. While both glucose low and high consuming Treg cells expressed glycolysis pathway genes, low glucose consuming Treg cells had enriched expression of the lactate transporter Slc16a1 (MCT1) and lactate hydrogenous a (Ldha). Isotopic flux analysis revealed Treg cells take up lactate to generate TCA cycle intermediates and phosphoenolpyruvate (PEP) which we show is critical for intratumoral Treg cell proliferation. Using B16 melanoma bearing mice harboring a Treg specific deletion of lactate transporter MCT1, we demonstrate MCT1 is dispensable for peripheral Treg cell function but required intratumorally resulting in slowed tumor growth and prolonged survival. Analysis of tumor infiltrate demonstrated increased proliferation and cytokine production by intratumoral effector T cells and decreased suppressive capacity of MCT1-deficient intratumoral Treg cells.

Conclusions

Taken together these data support a model in which Treg cells are metabolically flexible to utilize ‘alternative’ metabolites to support their function in nutrient poor environments like that of the tumor. Targeting lactate metabolism via MCT1 directed antibodies or inhibitors or lowering tumor lactate levels provides a promising therapeutic strategy to impair intratumoral Treg suppressive capacity and enhance anti-tumor immunity.

Ethics Approval

Animal work was done in accordance with the Institutional Animal Care and Use Committee of the University of Pittsburgh (protocol #17071235).

Correspondence: Todd Triplett (todd.triplett@gmail.com)

Background

The alteration of epigenetic status by HDACs, regulates chromatin structure and transcription by controlling protein access to DNA through the removal of acetyl groups. In addition, HDACs are now known to regulate non-histone proteins as well. Histone deacetylase inhibitors (HDACi) are a class of agents that modulate the acetylation status of histones and non-histone proteins, that are used clinically to treat cancer. Recently, HDACi have been shown to rely on CD8+ T-cells during monotherapy and enhance anti-tumor immune responses when given with immunotherapy. However, this is counterintuitive as majority of early studies have found HDACi are largely immunosuppressive. While, prior HDACi studies on immune function has primarily been on bulk T-cells, various T-cell subsets exist and play diverse roles in immune responses, including naive T-cell activation and immunological memory responses, which may be differentially impacted. Understanding the role of HDACi on T-cell subsets will lead to important biological insights and may aid in the development of future drug design.

Methods

Normal PBMCs isolated from healthy individuals were subjected to HDAC Inhibitors with varying specificities: Panobinostat, Entinostat, Vorinostat and OKI-005 and stimulated with PMA/Ionomycin and protein export was blocked with BFA. Cells were fixed and permeabilized after 18 hours and stained with markers: CD4, CD8, CD95, CD31, and CCR7 to delineate naïve and memory subtypes.

Results

To determine HDACi effects on T-cells, I evaluated cytokine production on bulk CD4 and CD8 T-cells, and observed a decrease in IL-2, consistent with previous literature that HDACi is negatively impacting T-cells overall. However, TNFa also significantly increased, showing HDACi effects are not only due to cytotoxicity, but cytokines are being differentially affected. Interestingly, the cytokine change effect was across both CD4 and CD8 T-cells and when delineated into naïve versus memory subsets, the naïve T-cells were much more impacted than memory cells, exhibiting more drastic decreases of IL-2 and increases in TNFa. Naïve T-cells also exhibited higher acetylation levels of total acetylated lysine and di-acetyl H3 than memory cells 24 hours after drug treatment.

Conclusions

Naïve T-cells comprise a large portion of T-cells from PBMCs, which may explain why previous studies with HDACi show immunosuppressive effects. However, anti-tumor immunity largely relies on tumor infiltrating lymphocytes, which consist of memory T-cells. Thus, memory T-cells’ reduced susceptibility to HDACi could explain enhanced anti-tumor immune responses seen previously. Further investigation into the effect of HDACi on distinct T-cell subsets could aid in rationally designing future immunotherapy strategies in cancer.

Acknowledgements

This work was funded by the CPRIT SCHOLAR IN CANCER RESEARCH Grant #RR160093 awarded to Dr. S. Gail Eckhardt. Special thanks to Anna Capasso, Todd Triplett, Carla Van Den Berg, the personnel of the Developmental Therapeutics Lab for their guidance and support and OnKURE Therapeutics.

Correspondence: Julie Carman (julie.carman@bms.com)

Background

Immunotherapy has transformed cancer treatment, providing durable responses for a subset of patients. Improved responses were recently observed for combinations of checkpoint inhibitors with chemotherapy regimens as compared to either therapy alone [1]. Chemotherapeutics are thought to enhance anti-tumor responses by increasing tumor antigens via immunogenic cell death and by modulating the tumor microenvironment. In addition to killing tumor cells, these agents may also directly impact immune cell function. In order to understand the impact of different classes of chemotherapeutic agents on immune cell function, a panel of compounds was profiled using several in vitro functional assays of human cells.

Methods

T cells and monocytes were isolated from the whole blood of 6-8 healthy donors. Monocytes were differentiated in vitro to immature dendritic cells by culture with GM-CSF and IL-4 for 7 days. T-cell proliferation was assessed by stimulating isolated T cells with antibodies to CD3 and CD28 for 3 days. Proliferation was measured by uptake of 3H-thymidine. T-cell cytokine production was assessed in an allogeneic mixed lymphocyte culture. Immature dendritic cells were cultured with T cells from HLA-mismatched donors for 5 days. IFNγ in the supernatant was assessed using ELISA. Dendritic-cell maturation was assessed by culturing immature dendritic cells with CD40L overnight. Surface markers (CD80, CD86, CD83, CD32, CD40, HLA-DR, PD-L1) were assessed using flow cytometry.

Results

Consistent with their anti-proliferative activity, several classes of chemotherapy agents potently inhibited anti-CD3 plus anti-CD28–driven T-cell proliferation. These included microtubule targeting agents (paclitaxel, docetaxel, eribulin), gemcitabine, and doxorubicin. Alkylating agents (cisplatin, carboplatin, oxaliplatin) and CDK4/6 inhibitors (abemaciclib, palbociclib) had moderate anti-proliferative activity. PARP inhibitors (rucaparib, olaparib), pemetrexed, and fluorouracil each had weak activity. Although pemetrexed and fluorouracil had weak anti-proliferative activity, they did inhibit IFNγ production in the allogeneic mixed lymphocyte culture. Gemcitabine, doxorubicin, CDK4/6 inhibitors, and some of the microtubule targeting agents (paclitaxel, docetaxel) also had strong activity in this assay. Alkylating agents weakly inhibited IFNγ production. In the dendritic-cell maturation assay, eribilin and doxorubicin were the only compounds among those tested that impacted expression of maturation and activation markers.

Conclusions

Many chemotherapeutic agents inhibit T-cell proliferation and IFNγ production, with IC50 values in the range of concentrations achieved at therapeutic exposures. These data suggest that several of these agents could also negatively impact T-cell–driven anti-tumor responses. This understanding may help guide treatment selection and scheduling of optimal combination therapeutics with checkpoint inhibitors to broaden anti-tumor responses.

Acknowledgements

Bristol-Myers Squibb.

References

1. Gandhi L, Rodríguez-Abreu D, Gadgeel S, et al. Pembrolizumab plus chemotherapy in metastatic non-small-cell lung cancer. N Engl J Med. 2018;378:2078-2092.

Ethics Approval

The protocol was approved by site institutional review boards or independent ethics committees and conducted according to Good Clinical Practice guidelines, per the International Conference on Harmonisation. Volunteers provided written informed consent based on Declaration of Helsinki principles.

Correspondence: Hye Seung Lee (mw9195@snubh.org), Jae-Sung Kim (jskim@snubh.org)

Background

The classification of tumor microenvironments based on the presence of tumor infiltrating lymphocytes (TILs) and programmed cell death ligand-1 (PD-L1) expression has been used to predict the efficacy of immune checkpoint blockade in several cancer types. Few studies have evaluated the predictive and prognostic role of CD8 and PD-L1 coexpression in locally advanced rectal cancer.

Methods

CD8+ TIL density and tumor PD-L1 expression were assessed by immunohistochemistry using pretreatment biopsies from one-hundred and eighty nine consecutive patients with locally advanced rectal cancer who had preoperative chemoradiotherapy (CRT). Response to CRT was determined based on Dworak tumor regression grade (TRG); tumors with complete (TRG 4) or near-complete (TRG 3) regression were grouped as good responders, and those with TRG 0 to TRG 2 as poor responders.

Results

High CD8 TILs showed borderline significant association with better disease-free survival (DFS). Although PD-L1 expression alone was not correlated with DFS, the combination of PD-L1-positive and CD8 low group identified a subtype with poorer DFS. In multivariate analysis, this classification based on combination of CD8 and PD-L1 was independent prognostic factors even when considering the clinical parameters (p =0.007). Among pathologic node-positive patients, high CD8 TILs compared to low were significantly associated with higher rate of good responders (29.2% vs 5.6%, p = 0.023).

Conclusions

A combination of tumor PD-L1 expression and low CD8+ TIL density was significantly associated with unfavorable survival in locally advanced rectal cancer patients who underwent preoperative CRT. It is proposed that PD-L1 expression in combination with CD8+ TIL density could be a useful predictive and prognostic biomarker in these patients.

Ethics Approval

The study was approved by Seoul National University Bundang Hospital's Ethics Board, approval number B-1707-406-306.

Correspondence: Olivier Elemento (ole2001@med.cornell.edu), Himisha Beltran (hip9004@med.cornell.edu)

Background

Prostate cancer (PCa) shows a modest clinical benefit from immunotherapy. Neuroendocrine prostate adenocarcinoma (NEPC) is a histological subtype of advanced PCa that arises clonally from castrate resistant prostate adenocarcinoma (CRPC) as a mechanism of resistance, but shares pathological, clinical, and molecular features with small cell lung carcinoma (SCLC). We studied the immunogenomics landscape of NEPC compared to other prostate malignancies and SCLC to identify mechanisms of immune evasion and potential immunological targets in NEPC.

Methods

We evaluated RNA-seq data from a 190 patient cohort comprised of benign prostate (n=29), localized prostate adenocarcinoma (PCa; n=68), hormone-naïve metastatic prostate adenocarcinoma (mPCa; n=11), CRPC (n=54), NEPC (n=25) with follow-up data, and SCLC (n=28) [1]. Whole exome sequencing available for 234 cases was used to quantify the mutations and copy number load in the cohort. Unsupervised hierarchal clustering followed by a consensus clustering approach was applied to the expression profiles to determine T-cell immune status for each sample i.e., inflamed or depleted.

Results

Immune cluster analysis revealed a predominantly T-cell depleted immune status across prostate subtypes, where in all de novo NEPC cases belonged exclusively to the immune depleted cluster. Poor overall survival (OS) was associated with the inflamed compared to the depleted immune cluster (HR=2.62; p= 0.002; adjusted for subtype). Significantly higher genomic copy number gains were found in T-cell depleted compared to inflamed cases (p= 0.001); however, no difference in total mutation load was seen. Further analysis of NEPC showed a relatively lower expression of cytokines, cytolytic activity, checkpoint, antigen processing machinery, and HLA presentation genes in NEPC (p<0.001).

Conclusions

NEPC is characterized by a T-cell depleted tumor microenvironment similar to other metastatic prostate cancer subtypes but with higher PDL1 expression, which is comparable to SCLC. The inverse correlation between OS and infiltration, as well as the novel expression changes in cytokines and checkpoint markers support further investigation into the potential immunological targets for NEPC.

References

1. Rudin CM, Durinck S, Stawiski EW, Poirier JT, Modrusan Z, Shames DS, Bergbower EA, Guan Y, Shin J, Guillory J, et al., Comprehensive genomic analysis identifies SOX2 as a frequently amplified gene in small-cell lung cancer. Nat Genet. 2012;44:1111-1116.

Correspondence: Davide Bedognetti (dbedognetti@sidra.org)

Background

Network analysis in biology has led to discovery of molecular and genetic interactions, biomarkers of disease, and therapeutic targets. Large-scale biomedical data availability offers a unique opportunity to assess the effect and impact of key cellular determinants for an observed phenotype. In this work, we consider the phenotype associated with the expression of the immunologic constant of rejection (ICR) gene set. This gene set consists of 4 major pathways observed in immune rejection of cancerous tissue, th1 signaling, chemokine production, effector function and immune regulation Higher expression of these genes is associated with tumor infiltrating lymphocytes and an active immune engagement, often correlated with prolonged survival. [1–6]

Methods

Consensus clustering of tumor samples based on ICR gene expression divides them into three classes, ICR high, medium and low. The ICR high phenotype is associated with better survival in a subset of cancer types which we refer to as ICR enabled (N=8). In other cancer types we observe the opposite. (ICR Disabled, N=4) [7]

Reverse engineering the gene regulatory networks (TF-target) for each cancer subtype from transcriptional profiles (RNA sequences) of tumors obtained via the cancer genome atlas (TCGA) was performed using a machine learning technique called RGBM. We then perform functional gene set enrichment analysis to discover differential active master regulators.

Results

Our aim is to identify key transcription factors (TF) i.e. proteins binding in the regulatory regions of the DNA for specific target genes and regulating their expression, which are differentially activate between ICR High and ICR Low samples which are molecular signatures of hot or cold immune response respectively. We unravel common differentially active TFs or master regulators for these phenotypes over the 12 selected cancer subtypes, where the ICR can be shown to have an impact on survival. We found genes such as IRF1, STAT1, CXC10, CTLA4 to be common master regulators which are positively active for the ICR High phenotype and a validation of our proposed approach as these genes are part of the ICR gene signature. Master regulators, L3MBTL1, HDAC11, SALL2 etc., are positively active for the ICR Low phenotype across the 12 cancers.

Conclusions

Downstream analysis of the master regulators specific to the ICR Low phenotype resulted in identification of the NOTCH signaling pathway, transcriptional regulation of oncogene TP53 and several other cancer related signaling pathways that need to be investigated further to understand mechanisms of poor immune response (ICR Low).

References

1. Wang, E., Worschech, A. & Marincola, F. M. The immunologic constant of rejection. Trends Immunol 29, 256–62 (2008).

2. Spivey, T. L. et al. Gene expression profiling in acute allograft rejection: challenging the immunologic constant of rejection hypothesis. J Transl Med 9, 174 (2011).

3. Bedognetti, D., Tomei, S., Hendrickx, W., Marincola, F. M. & Wang, E. Toward the Identification of Genetic Determinants of Responsiveness to Cancer Immunotherapy. in Developments in T Cell Based Cancer Immunotherapies (eds. Ascierto, P. A., Stroncek, D. F. & Wang, E.) 99–127 (Springer International Publishing, 2015). doi:10.1007/978-3-319-21167-1_5

4. Bedognetti, D., Hendrickx, W., Ceccarelli, M., Miller, L. D. & Seliger, B. Disentangling the relationship between tumor genetic programs and immune responsiveness. Curr. Opin. Immunol. 39, 150–158 (2016).

5. Hendrickx, W. et al. Identification of genetic determinants of breast cancer immune phenotypes by integrative genome-scale analysis. OncoImmunology 0, 00–00 (2017).

6. Bertucci, F. et al. The immunologic constant of rejection classification refines the prognostic value of conventional prognostic signatures in breast cancer. Br. J. Cancer (2018). doi:10.1038/s41416-018-0309-1

7. Roelands, J. et al. Genomic landscape of tumor-host interactions with differential prognostic and predictive connotations. bioRxiv 546069 (2019). doi:10.1101/546069

Correspondence: Taigo Kato (kato@uro.med.osaka-u.ac.jp)

Background

Cryoablation is one of treatment modalities for kidney cancer and is expected to induce strong local and systemic T cell-mediated immune reactions. So far, in some cases, these immune reactions cause the elimination of distant tumor (abscopal effect). However, the molecular mechanisms of the immune response, particularly for T cells that play fundamental roles in attacking cancer cells, still remain unclear.

Methods

In this study, we collected tumor tissues and blood samples from 22 kidney cancer patients, before cryoablation and at 3 months after cryoablation. We applied a next generation sequencing approach to characterize T cell receptor beta (TCRB) repertoires using RNAs isolated from samples.

Results

TCRB repertoire analysis revealed expansion of certain T cell clones in tumor tissues by cryoablation. We also found that proportions of abundant TCRB clonotypes (frequency of ≥ 1%) were significantly increased in the post-cryoablation tissues than those of pre-cryoablation tumors. Interestingly, some of these TCRB clonotypes were increased in peripheral blood. Expression analysis of immune-related genes in the tissues of pre- and post-cryoablation showed significantly elevated transcriptional levels of CD8, CD4, GZMA, and CD11c with higher CD8/FOXP3 ratio in the post-cryoablation tissues.

Conclusions

Our findings revealed that cryoablation could induce strong immune reactions in tumors with oligoclonal expansion of anti-tumor T cells, which circulate systemically and induce abscopal effect in distant tumor.

Correspondence: S. Rhiannon Jenkinson (rhiannon.jenkinson@crl.com)

Background

In order to develop therapeutics which drive the immune system to target tumour cells and eliminate tumour growth, sophisticated in vitro and in vivo models are required. We have developed models which enable us to combine data from human immune cell in vitro assays and murine pharmacodynamic (PD) and syngeneic tumour models to evaluate which pathways a therapeutic is hitting and whether it is effective at inhibiting tumour growth in vivo. We show here the powerful combination of using TCR transgenic T cell adoptive transfer and syngeneic tumour models with flow cytometry and Nanostring profiling of gene expression within the tumour microenvironment (TME) to determine the effect of therapeutic intervention.

Methods

To determine the effect of checkpoint inhibitors, mice bearing a defined population of ovalbumin (OVA)-specific T cells were challenged either with antigen (PD model), or an OVA expressing tumour (EG7 or MC38). In the PD model, activation, proliferation and differentiation of these cells into CTL were assessed. Using the OVA-expressing tumour models, comprehensive TIL analysis was carried out to determine how changes to T cell activation impacted on the wider TME and whether a model tumour vaccine could improve T cell responses. Nanostring analysis of gene expression using the murine immune-oncology 360 panel enabled a deeper analysis of the TME and acted as a powerful tool for the identification of potential PD biomarkers. Tumour killing assays were also performed with human cell lines and PBMC using clinical antibodies targeting the same pathways.

Results

In the PD model, anti-PD-L1 caused robust upregulation of CD25, IFNγ and granzyme B by antigen-specific CD8+ T cells. Anti-CTLA-4 exerted a similar but less potent effect. Immunisation with peptide increased T cell infiltration in the syngeneic model and this correlated with enhanced T cell functional capacity. Nanostring analysis of tumours saw raised CTL, NK cell and neutrophil gene expression within tumours from peptide immunised mice. Clinical therapeutics targeting PD1 or CTLA-4 drove enhanced tumour killing using human cell-based assays.

Conclusions

These models allow testing of novel therapeutics alongside benchmark reagents. The PD model provides a screening platform to assess enhancement of T cell function before moving into syngeneic models. Nanostring gene expression profiling may direct which in vitro assays should be used to map the immunological mechanisms underlying immune modulation. The use of in vitro and in vivo PD/efficacy models better enable assessment of novel cancer immune therapeutics and allow early identification of potential biomarkers.

Ethics Approval

The study was approved by HRA NRES South West, Bristol (UK), approval number 15/SW/0029.

Correspondence: Jordi Petriz (jpetriz@carrerasresearch.org)

Background

Multiple myeloma (MM) is characterized by accumulation of malignant plasma cells in the bone marrow (BM) with an associated immunosuppressive BM microenvironment. Programmed Death-Ligand 1 (PD-L1) has been shown to suppress immune responses through the interaction with Programmed Death-1 (PD-1). In MM, PD-L1 is overexpressed in MM plasma cells and in Myeloid Derived Suppressor Cells (MDSCs). PD-1 is expressed in T-cells. The interaction between PD-L1 and PD-1 decreases TCR-mediated proliferation and cytokine production. PD-L1 plays an important role in tumor immune evasion and drug resistance, and is considered a therapeutic target. However, some MM individuals do not respond to treatments with PD-L1 or PD-1 inhibitors. The aim of this study was to design and evaluate a direct functional screening assay to identify MM MDSCs PD-L1+ using flow cytometry, and its potential use in MM management.

Methods

Samples were labeled with PE-PD-L1, APC-CD11b, PE-Cy7-CD33, and FITC-HLA-DR (Invitrogen™ eBioscience) to detect MDSCs using no-lyse no-wash methods. Vybrant™ DyeCycle™ Violet Stain (Thermo Fisher) was used to discriminate nucleated cells from erythrocytes and debris. Samples were analyzed using Attune™ NxT Flow Cytometer (Thermo Fisher). PD-L1 expression was studied in n=35 MM subjects, with and without bone marrow stimulation with PMA (Merck), for 10 minutes at 37ºC. PD-L1 cell surface expression was compared with cytoplasmic expression (n=11 subjects). Kinetics of PD-L1 expression were also studied over time. Competitive experiments in the presence of Durvalumab (0 ng/μL to 250 ng/μL) were used to study its interaction with PD-L1.

Results

PD-L1 expression was found dramatically increased after stimulation (n=33 subjects, 94.3%) ranging from 2 to 650 times. PD-L1 cytoplasmic levels were found to be undetectable. PD-L1 expression was found to be higher after 1 to 5 minute stimulation, with a progressive decrease up to 1 hour. Co-incubation with Durvalumab resulted in different inhibitory PD-L1 immunofluorescent profiles.

Conclusions

PD-L1 reactivity appears to result from complex interactions that can only be detected with minimal sample perturbation. Since PD-L1 is not found at the cytoplasmic level, PD-L1 may reveal steric changes in response to stimulation, even for a short period of time. These changes may be associated with a PD-L1 immunoregulatory mechanism that may affect therapies targeting the PD-1/PD-L1 checkpoint. Assessment of PD-L1 folding may help to develop better treatment strategies or to predict therapy resistance. No-lyse no-wash methodologies in combination with functional assays show promise as an emerging strategy to model conformational changes in the target site.

Ethics Approval

All procedures were performed in accordance with the internal protocols of the laboratory, which were authorized by the HGTiP’s Ethical Committee, in accordance with current Spanish legislation, by the Departament de Medi Ambient i Habitatge (file #1899) of the Autonomous Government of Catalonia (Generalitat de Catalunya).

Correspondence: Xiao-Jun Ma (xiao-jun.ma@bio-techne.com)

Background

Immunotherapy has proven to be a powerful anti-tumor therapy by harnessing the body’s own immune system to target and kill tumor cells. However, immunotherapy is not successful in all cancer patients due to both intrinsic non-responsiveness and adaptive resistance. Developing predictive biomarkers and understanding mechanisms of resistance are major goals of the immuno-oncology community. The extracellular matrix (ECM), an important factor for promoting tumor growth and metastasis, may also act as a physical barrier to prevent immune cell infiltration and promote tumor immune escape. Components of the ECM such as COL11A1, COL4A1, and LOXL2 have been shown to be associated with cancer progression. Furthermore, new data suggests that TGFβ activation leads to up-regulation of ECM genes in cancer-associated fibroblasts and immune suppression. However, it remains poorly understood which cells in the tumor microenvironment (TME) are the sources of ECM gene expression and how they are related to tumor infiltrating cytotoxic T lymphocytes (CTLs).

Methods

In this study, we employed a highly sensitive and specific RNAscope in situ hybridization (ISH) duplex assay to directly visualize the tissue distribution of cells expressing COL4A1, COL11A1, LOXL2, and TGFB1 in relation to tumor infiltrating CTLs in non-small cell lung carcinoma (NSCLC). NSCLC tissue microarrays (TMAs) consisting of 63 independent patient FFPE tumor samples were analyzed using this ISH assay with the following probe combinations: Hs-CD8/Hs-IFNG, Hs-CD4/Hs-FOXP3, Hs-LOXL2/Hs-COL4A1, and Hs-TGFB1/Hs-COL11A1.

Results

We observed COL4A1 expression in both tumor and tumor-associated stromal cells in different samples. In contrast, COL11A1 was only expressed in tumor-associated stromal cells. Interestingly, high COL4A1 expression was associated with high CD8+ T cell infiltration, whereas high COL11A1 expression was associated with poor CD8+ T cell infiltration. In addition, tumor expression of TGFB1 was positively correlated with COL11A1 expression. These data depict a complex landscape of ECM gene expression and their relationship to T cell infiltration in the tumor and TME.

Conclusions

Taken together, these results demonstrate that the RNAscope assay provides a powerful approach to directly examine the interactions between tumor, ECM, and T cell immune infiltration and offers advantages over immunohistochemistry (IHC) for identifying the cellular sources of secreted proteins such as ECM components in the TME.

Correspondence: Marianne Stanford (mstanford@imv-inc.com)

Background

DPX-Survivac is a novel lipid-based formulation designed to elicit de novo cytotoxic T cell response to survivin-expressing tumours. DeCidE1 trial is assessing the clinical activity of DPX-Survivac in combination with low dose intermittent cyclophosphamide with or without epacadostat in advanced ovarian cancer patients. To understand the underlying mechanism of action of this immunotherapy, we performed immune-profiling of PBMC and tumour samples from this study.

Methods

Pre-treatment and longitudinally collected on-treatment PBMCs were used to assess survivin-specific T cell response by ex vivo IFN-γ ELISPOT and in vitro MHC multimer staining. Immune-phenotyping was performed on a subset of samples using multi-parametric flow cytometry and tSNE analyses. Pre- and on-treatment tumour biopsies were subjected to whole-transcriptome RNA-sequencing to assess changes in immune infiltration. TCR repertoires were assessed in tumours by TCRb ImmunoSEQ analysis.

Results

Treatment with DPX-Survivac induced systemic survivin-specific T cell response in nearly all evaluable subjects as assessed by ELISPOT. On-treatment enrichment in systemic survivin-specific T cells was also detected after in vitro expansion and tetramer analysis, confirming activated T cells are functional and proliferative. Immunophenotyping of PBMC did not show substantial increases in the expression of immunosuppressive markers (CTLA-4, PD-1, B7-H3) within T cell population in subjects that responded to the treatment, further suggesting that induced T cells remained active over time. RNA-profiling of immune cells within the tumour micro environment of subjects demonstrating clinical response revealed on-treatment overexpression of genes related to T cell activation and cytolysis as well as enrichment of B cell and NK cell specific signatures, suggesting the strong potential of treatment to induce tumor-infiltration and activation of cytotoxic cells. Infiltration of T cells into tumor post treatment did not correlate with highest levels of systemic survivin-specific T cells, suggesting migration of circulating activated T cells into tumours. In contrast to tumor infiltration analyses, minimal changes in circulating B cell populations were observed on-treatment. Analysis of TCRβ repertoire in pre- and on-treatment tumours demonstrated that DPX-Survivac therapy can promote proliferation and tumor-infiltration of new T cell clones.

Conclusions

DPX-Survivac combinational therapy induces robust and sustained survivin-specific responses and promotes T cell infiltration of tumours, without a loss in functionality. The infiltration immune cells beyond T cells has been demonstrated in tumor tissue yet was not consistently detected in PBMCs; a finding that emphasizes the need for immune-profiling of both blood and tumor to gain a full understanding of the mechanism of action of novel immunotherapies and combinations.

Ethics Approval

The study was approved by Dalhousie University's research ethics board, approval number 2018-4659

Correspondence: Daniel Hargboel Madsen (daniel.hargboel.madsen@regionh.dk)

Background

Tumor progression is accompanied by extensive remodeling of the surrounding extracellular matrix leading to the formation of a tumor-specific ECM, which is often more collagen-rich and of increased stiffness. High collagen-density within tumors is known to promote tumor progression, however, it is unknown if these pro-tumorigenic effects involve modulation of T cell activity. To investigate if a high-density tumor-specific ECM could influence the ability of T cells to kill cancer cells, we studied how T cells respond to 3D culture in different collagen densities.

Methods

T cells cultured in 3D conditions surrounded by a high or low collagen density were imaged using confocal fluorescent microscopy, and the effects of the collagen density on T cell survival, proliferation and differentiation were examined using flow cytometry. Similarly, cancer cell proliferation in 3D was also evaluated. Using immunohistochemistry, triple-negative breast cancer specimens were analyzed for the number of infiltrating CD8+ T cells and for the collagen density. The influence of collagen density on T cell cytotoxicity was examined using chromium-51 release assay. Whole-transcriptome analyses were used to study in detail the effects of collagen density on T cells and computational analysis was used to identify transcription factors involved in the collagen density-induced gene regulation. Observed changes were confirmed using qRT-PCR analysis.

Results

Using 3D collagen gels of varying density, we identified that T cells respond to their ECM environment and that T cell proliferation is significantly reduced in a high-density matrix, while cancer cell proliferation is unaffected. Consistently, the number of infiltrating CD8+ T cells in mammary tumors with high collagen-density was reduced, indicating it can play a role in regulating T cell abundance in human breast cancer. Additionally, whole-transcriptome analysis of 3D-cultured T cells revealed distinct transcriptional profiles depending on the surrounding collagen-density. Specifically, a high-density matrix induced downregulation of cytotoxic activity markers and upregulation of regulatory T cell markers. Importantly, these transcriptional changes were accompanied by an impaired ability of tumor-infiltrating T cells cultured in a high-density matrix to kill autologous melanoma cells.

Conclusions

Our study identifies a new immune modulatory mechanism, which could be essential for suppression of T cell activity in the tumor microenvironment.

Correspondence: Chris Langsdorf (chris.langsdorf@thermofisher.com)

Background

With continued growth in development and approval of biologic drugs and cell therapies, comes a need for more robust and reliable cell-based assays and analysis systems. Here we describe new methods to assess the specific killing of target cells by antibody-mediated complement-dependent cytotoxicity (CDC), antibody-dependent cellular cytotoxicity (ADCC) by natural killer cells, T cell killing, and antibody-dependent cellular phagocytosis (ADCP).

Methods

These methods are demonstrated with single-cell analysis models compatible with flow cytometry and image-based analysis. Additionally, we have transitioned these assays into more biologically-relevant models such as tumor spheroids and whole blood. T and NK cell penetration and killing of breast and lung cancer spheroids was evaluated using quantitative confocal microscopy. Ramos Burkitt's lymphoma cells were opsonized with Rituximab, spiked into human whole blood, and complement- and NK-dependent killing evaluated using flow cytometry. Ramos cells opsonized with Blinatumomab were spiked into human whole blood and bispecific-dependent T cell killing evaluated with flow cytometry.

Results

T and NK cells were imaged at single-cell resolution as they penetrated and killed tumor spheroids in an antibody-dependent manner. Whole blood analysis with flow cytometry was able to clearly distinguish target cells from white blood cells, and clearly indicated the percent of target cells killed in an antibody-dependent fashion.

Conclusions

Penetration and potency of immune effector cells can be evaluated using whole-spheroid imaging. Flow cytometry provides a robust method to directly measure T, NK, and complement-dependent killing of cancer cells in human whole blood.

Correspondence: Kimberly Schluns (kschluns@mdanderson.org)

Background

Loss of Interleukin (IL)-15 expression by colon tumors strongly correlates with classification as microsatellite stable, decreased tumor infiltrating T cells, increased metastasis, and poor responses to immunotherapy. To better understand how IL-15 expression regulates tumor-associated immune cells and demonstrate the importance of IL-15 expressed by tumor cells, we developed a mouse model system to examine this by generating MC-38 tumor cells deficient in IL-15 using Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR)/Cas9 system. The MC-38 cells are ideal for this as the tumor cells are the dominant source of sIL-15 complexes in the tumor microenvironment (TME) [1].

Methods

After one round of CRISPR/Cas9 deletion of IL-15, we have generated MC-38 cells that lack one IL-15 allele (i.e IL-15+/- MC-38 cells) and express decreased levels of soluble IL-15 complexes. A second round of deletion has led to complete deletion of IL-15 in MC-38 cells (i.e. IL-15-/- MC-38). IL-15+/- and IL-15-/- MC-38 cells and wildtype (Wt) MC-38 cells were implanted subcutaneously into Wt mice and tumor growth measure overtime. Additionally, immune cells in tumors were analyzed at earlier time points. RNAseq of tumor-associated myeloid cells was conducted in Wt MC-38 tumors treated with neutralizing IL-15 antibody.

Results

Surprisingly, deletion of one IL-15 allele was sufficient to affect tumor growth as IL-15+/- MC-38 tumors grew faster than Wt MC-38 tumors but had similar growth kinetics in vitro. IL-15+/- MC-38 tumors harbored less CD8 T cells and NK cells than Wt tumors whereas numbers of CD4 T cells were similar. Interestingly, these tumors also had increased CD11b+Ly6G+ and CD11b+Ly6ChiLy6G- myeloid cells (putative MDSCs). Implanted IL-15-/- MC-38 cells also progressed faster than Wt MC-38 cells. Gene expression analysis of myeloid cells showed that blocking IL-15 in tumors increased expression of CXCL4, CD206, Arginase 1, and tryptophan 2,3-dioxygenase and decreased IL-12β and IL-6.

Conclusions

These results demonstrate that even partially decreasing the levels of IL-15 in the TME negatively impacts the numbers of CD8 T cells and NK cells and promotes differentiation of putative MDSCs. These results suggest that IL-15 expressed by the tumor cells is a key attribute creating a microenvironment that supports cytolytic lymphocytes by not only acting directly on lymphocytes but also indirectly by skewing the myeloid cells towards pro-inflammatory activities. Overall, the IL-15-/- MC-38 cells is a model that can be used to identify the mechanisms contributing to the poor therapeutic responses observed in human colon carcinomas.

References

1. Santana Carrero, R.M. et al. IL-15 is a component of the inflammatory milieu in the tumor microenvironment promoting antitumor responses. 2019. Proc Natl Acad Sci U S A. 116(2):599-608.

Ethics Approval

All animal procedures were conducted in accordance with the animal care and use protocols (00000851-RN02) approved by the IACUC at the UT MD Anderson Cancer Center.

Correspondence: Amin Osmani (amin.osmani@immunXperts.com)

Background

During the last years, significant advancement has been made in the clinical application of cancer immunotherapies. Molecules directed against immune checkpoints and other agonists show great promise for treatment of a variety of malignancies. Next to CTLA-4 and PD-1 blockade, a wide range of therapeutics with the potential to reverse the tumor-induced suppression are under development.

Methods

Mixed lymphocyte reaction assays using both innate cells and lymphoid cells mimic a real physiological T cell response and are widely used for the potency screening of candidate therapeutics. The use of different allogenic donor combinations can provide additional information on the profile of the responding population. Other T cell assays such as antigen-specific CMV recall activation assays or SEB activation assays can be used to evaluate the ability of test molecules to promote T cell responses.

Next to these activation assays, the antigen-specificity within a naïve or pre-exposed T cell population can be evaluated by using T cell enrichment assays, which can be set-up using different formats depending on the origin of test molecules. Additionally, it might also be beneficial to generate peptide-specific T cell pools or clones to be used for functional testing e.g. in the context of tumor-associated antigens (TAA).

Results

All these in vitro T cell assays are optimized and fine-tuned for the screening of certain types of molecules. An important factor for sensitive and reproducible assays and consistent results is the quality of the primary immune cells. Therefore, all PBMC donor preparations are quality controlled and HLA-typed, and optimized procedures are used to generate functional subpopulations such as dendritic cells and T cell subpopulations.

Conclusions

Early evaluation of the effectiveness of candidate therapeutics and combination therapies can be assessed using mouse models and in vitro bioassays with mouse or human immune cells.

Correspondence: Amin Osmani (amin.osmani@immunXperts.com)

Background

The increasing interest in the tumour microenvironment leads to focus on new bioassays to represent all the players of the cancer immune response. Some of these players like Tumour Associated Macrophages (TAM), Myeloid Derived Suppressor Cells (MDSC) regulatory T cells (Treg) play an important role by downregulating the anti-tumour response. Their regulation mechanisms constitute an important target for new therapeutics. In order to study these mechanisms in a human model, suppressive bioassays, mimicking the suppressive action of these cells on T cells activations, were developed.

Methods

Macrophages possess important active and regulatory functions in both innate and adaptive immune responses. Classical activated macrophages, also classified as M1-like macrophages, comprise immune effector cells with an acute inflammatory phenotype while the alternatively activated M2-like macrophages have suppressive and healing capacities. Tumor associated macrophages (TAMs) are present at high densities in solid tumors and share many characteristics with so called M2 macrophages.

Results

Although distinguished classification and in vitro generation and polarization of M1- and M2-like macrophages is challenging, in vitro assays can be a first step to screen the effect of the test molecules on the phenotype and function of the macrophages. For example, macrophage precursors display extraordinary plasticity in response to exogenous and endogenous stimuli which can lead them to M2-polarized macrophages or towards the M1-activated status. Using in vitro polarization and functional macrophage assays, one can screen molecules with the potential to influence M1 and M2 like macrophage generation and polarization. Next to that, the effect of the test molecules on the function of the macrophages can be evaluated using a macrophage suppressive assay. Here the ability of the molecules to reverse the stimulating effect of the M1-macropahges or suppressive effect of the M2-macrophages can be determined.

Myeloid-derived suppressor cells (MDSC) and regulatory T cells (Treg) can also be found in the tumour microenvironment and present a highly suppressive phenotype. Their role in relation to cancer development and progression has shown to be of great importance. Therefore, the ability of molecules to reverse the suppressive function of the MDSC or Treg on immune effector cells can be evaluated in vitro using these cell-type specific suppressive bioassays.

Conclusions

The use of the bioassays contributes to a better understanding of the tumour microenvironment and the steps needed to generate an anti-tumour response by the immune system will help to assess the functional potential of new drugs, design clinical trials and ultimately discover relevant biomarkers.

Correspondence: Kristina Young (kristina.young@providence.org)

Background

Transforming growth factor beta (TGFβ) is a multipotent cytokine capable of promoting tumors in part through immunosuppression. With the development of clinically relevant immunotherapy, inhibition of TGFβ is a widely explored therapeutic strategy to enhance the efficacy of both traditional cytotoxic therapies and immunotherapies.

Methods

Based on our previous work, we initiated a Phase II trial of a TGFβ type I receptor (TGFβRI or ALK5) inhibitor in combination with neoadjuvant chemoradiation in patients with locally advanced rectal adenocarcinoma (NCT02688712). We sought to better understand the mechanism of action of TGFβRI inhibition in this setting using preclinical CT26 and MC38 colorectal cancer models and cell type-conditional Alk5-deficient transgenic mice. The ALK5 inhibitor was administered P.O. b.i.d. in two-1 week cycles with drug respite of 1 week between cycles. 5Gy x 5 consecutive daily fractions of CT image guided radiation was delivered between ALK5 inhibitor cycles with concurrent 5-FU (25mg/kg i.p. t.i.w.). To interrogate the actions of TGFβR1 signaling on specific cell types, we evaluated MC38 tumor growth and response to radiation in three separate conditional Alk5 knock-out mice under control of cell-type specific promoters: CD8Cre-ALK5flox/flox, LysMCre-ALK5flox/flox, and Foxp3Cre-ALK5flox/flox to ablate ALK5 expression in CD8+ T cells, macrophages and regulatory T cells, respectively.

Results

No improvement in tumor growth or radiation response was seen in animals with Treg or macrophage specific ALK5 deletion. However, CD8Cre-ALK5flox/flox (ALK5ΔCD8) mice largely rejected MC38 tumors in a CD8+ T cell dependent manner and demonstrated increased sensitivity to radiation therapy associated with superior CD8+ T cell effector phenotypes, and reduced threshold for TCR stimulated activation. Alk5-deficient CD8+ T cells had diminished SMAD2/3 occupancy at the CXCR3 promoter, increased CXCR3 expression and enhanced migration to CXCL10. Importantly, neutralization of CXCR3 in vivo blocked the rejection of tumors in ALK5ΔCD8 mice.

Conclusions

These results indicate that TGFβ signaling represses CXCR3-dependent trafficking of CD8+ T cells into tumors and can be reversed to improve tumor immunity alone and in combination with radiation therapy.

Trial Registration

NCT02688712

Ethics Approval

This clinical study was approved by Providence Health & Services IRB, approval number 15-050. The animal modeling was performed under the approval of our institution's IACUC Protocol #37, in our OLAW certified animal facility (Assurance #D16-00526).

Correspondence: Kerry Campbell (kerry.campbell@fccc.edu)

Background

Rituximab is widely used to treat B-cell non-Hodgkin lymphomas (NHL). The antibody depletes B cells in patients by several mechanisms, including antibody-dependent cellular cytotoxicity (ADCC) by NK cells and monocytes, complement-mediated cell death, and signaling apoptosis. We tested degranulation responses by NK cells in peripheral blood from B-cell NHL patients before, during and after rituximab treatment.

Methods

Seventy-five patients with B-cell NHL [follicular lymphoma or marginal zone lymphoma] were monitored for immune phenotype and NK cell function in peripheral blood, as well as 54 healthy volunteers. All blood donors provided informed consent in accordance with our Institutional Review Board. Thirteen NHL patients were treated with rituximab and provided blood samples immediately before the first treatment, during the treatment period and three months after treatment. Peripheral blood mononuclear cells (PBMC) were isolated on Ficoll. Degranulation assays were performed by exposing PBMC to 721.221 EBV-transformed B cell line (natural cytotoxicity) for two hours and staining for LAMP-1 (CD107a). Degranulation was quantified by flow cytometry as percentage of LAMP-1+ CD56dim (mature, cytolytic) NK cells.

Results

NK cell degranulation responses were not significantly different between healthy donors or pretreatment samples from NHL patients. Within the patients treated with rituximab, a significant increase in the degranulation response of CD56dim NK cells was observed between the pretreatment and during treatment samples when assayed under natural cytotoxicity conditions (p

Conclusions

Our results provide evidence that a treatment course of rituximab can potentiate degranulation activity of peripheral blood NK cells under natural cytotoxicity conditions for at least 3 months after the end of therapy. At the same time NK cells become more resistant to activation-induced death during these natural cytotoxicity responses, which may contribute to the improved functional response.

Acknowledgements

We thank the FCCC Biosample Repository for coordinating patient consenting and sample acquisition and support from the FCCC Keystone Program, FCCC Boards of Associates, and NIH grants CA083859 (KSC) and CA06927 (FCCC).

Ethics Approval

The study was approved by the Fox Chase Cancer Center Ethics Board, approval numbers 99-802 and 99-839

Correspondence: Jesus Banuelos (jbanuelos@arcusbio.com)

Background

Elevated levels of myeloid-derived suppressor cells (MDSCs) have been associated with a blunted response to PD-(L)1 inhibitors in vivo and suppressive myeloid cells have been shown to inhibit T cell responses in vitro. Suppressive myeloid cells mediate their immunosuppressive effects via multiple mechanisms, including expression of immune checkpoint protein ligands, activation of the PI3Kγ pathway, and release of arginase-1 (ARG1), an arginine-depleting enzyme. The significant role of myeloid cells in dampening anti-tumor immunity makes them an attractive target for immunotherapy. To that end, we have developed potent and selective inhibitors against PI3Kγ and ARG1.

Methods

Recombinant and endogenous ARG1 from human granulocytic lysates were used to determine AB474 potency. The ability of AB474 (ARG1 inhibitor), to rescue recombinant ARG1-mediated inhibition of T cell activation was assessed in human T cells isolated from healthy donors. The ability of PI3Kγ inhibition to stimulate immune cells was assessed in CD14+ monocytes or monocyte-derived M1 polarized macrophages and confirmed using mixed lymphocyte reactions. Anti-tumor activity was determined using syngeneic tumor models

Results

Using sorted immune cells from healthy donor PBMC, we determined that PI3Kγ expression is highest in monocytes, whereas PI3Kδ expression is higher in lymphocytes. Inhibition of PI3Kγ significantly (p

Elevated levels of ARG1 were observed in cancer patients compared to healthy donors. In particular, ARG1 was highly expressed in polymorphonuclear-MDSCs while lower levels were seen in monocytic-MDSCs. AB474 inhibited the activity of both recombinant and endogenous ARG1 with an IC50 of 15 nM and 20 nM, respectively. Consistent with this observation, AB474 reversed ARG1-mediated suppression of human CD8+ T cell activation. Additionally, single agent AB474 significantly (p

Conclusions

Myeloid cells are associated with blunted responses to checkpoint protein blockade. High expression of both PI3Kγ and ARG1 within suppressive immune subsets contributes to the dampening of pro-inflammatory responses. Inhibiting these targets can reduce immunosuppression, enhance anti-tumor responses and act in concert with immune checkpoint blockade and other immunotherapies.

Correspondence: Jian Chen (jchen@celetrix.com)

Background

The interest in the therapeutic potentials of the innate immune system is growing as we gained more knowledge of role of the innate immune system in fighting cancer. Natural killer (NK) cells and macrophages are capable of attacking tumor cells directly and dendritic cells are important for antigen presentation. Macrophages are also important for modulation of tumor microenvironment that presents both obstacles and opportunities to fighting tumor cells. Gene expression or gene editing in the innate immune cells are not only important for studying gene functions in these cells, but also important for potential clinical applications.

Methods

Using our novel electroporation technology, we have found that the innate immunity cells can all be modified by protein expression plasmid vectors or CRISPR-Cas9 mediated gene editing.

Results

Primary NK cells, primary bone marrow monocytes and peripheral blood monocytes need relatively high voltage for electroporation. Primary mouse peritoneal macrophage needs lower voltage and expanded NK cells as well as mature dendritic cells need lowest voltage. Using the same electroporation buffer, we were able to successfully electroporate these different cells using different voltages. The transfection efficiency for plasmids can reach 80% and Cas9 ribonucleorprotein (RNP) with gRNA can also be electroporated successfully.

Conclusions

Our capability to modify all of the major types of the innate immunity cells by electroporation is a significant step that could contribute to harnessing the power of the innate immunity system for immunotherapy of cancer.

Correspondence: Kanae Gamo (kanae.gamo@fimecs.com)

Background

IRAK-M has an important role in tightly controlling innate immune responsiveness to preserve immune homeostasis by acting as a negative feedback regulator of TLR/IL-1R signaling pathway. Targeting IRAK-M, which expression is restricted to myeloid cells, would be potentially limiting adverse events against non-target tissues. From supporting evidence for the role of IRAK-M in innate immunosuppressive capacity of myeloid cells in tumor microenvironment, we have generated compounds targeting IRAK-M as an effective cancer-immunotherapy strategy by converting tumors to more inflamed state. Since IRAK-M is a pseudokinase which lacks kinase activity, it is considered 'undruggable’ that could not be targeted pharmacologically by the conventional small molecule. Therefore, we have developed heterobifunctional degrader molecules comprising IRAK-M-binding moiety linked to proprietary E3 binders to eliminate IRAK-M protein.

Methods

Optimization of IRAK-M degraders have conducted by applying RaPPIDS(TM) which is a proprietary divergent degrader synthetic platform and identified multiple preclinical candidates within a year. We measured the activity of degradation mechanisms in vitro by analyzing degradation rates, protein turnover and proteasome dependence. To assess the effect of IRAK-M degrader on myeloid-derived suppressor cells (MDSCs), we performed MDSC suppression assays in a co-culture model with CD8+ T cells. One of the drug candidates was profiled extensively in different syngeneic models. Subsets of tumors and spleen from treated mice were analyzed ten days following dosing initiation for myeloid cell populations and phenotype.

Results

We demonstrated a dose- and time-dependent degradation of IRAK-M protein in THP1 human monocytic leukemia cells with degrader treatment in a proteasome-dependent manner. MDSC suppression assays revealed that IRAK-M degrader could release the suppressive function of MDSC on both IFNγ production by CD8+ T cells as well as T cell proliferation. When examined in mouse models, IRAK-M degrader showed significant anti-tumor activity in several syngeneic models at tolerated doses and schedules. In the 4T1-HA model, we found increase of LPS-induced TNFα production in white blood cells from compound injected mice associated with degradation of IRAK-M protein as pharmacodynamic effect. FACS analysis showed that IRAK-M degradation translated into increased infiltration of M1-like activated macrophage into tumors and spleen. Advanced profiling of the drug candidate is ongoing in IND-enabling studies.

Conclusions

In summary, we demonstrated IRAK-M degrader as a novel and promising immuno-modulatory therapeutic modality and the great potential of shifting the balance between tolerance and immunity by releasing immunosuppressive network activity, leading to a more favorable tumor microenvironment to enhance host immunity.

Correspondence: Young Kim (y2.kim@vumc.org)

Background

Tumor infiltrating myeloid cells, such as Myeloid Derived Suppressor Cells (MDSCs) and Tumor Associated Macrophages (TAMs), are important mediators of immune evasion in cancer, and are associated with worse overall survival and progression free survival in various types of solid tumors. However their procarcinogenic functions in the tumor microenvironment are unclear. Here, we propose that the inflammasome signaling pathway is a key T cell independent mechanism of myeloid-mediated tumor growth.

Methods

In this work, we use various adoptive transfer models to generate bone marrow specific inflammasome knockout mice and utilized a B16 melanoma model to assess changes in tumor growth.

Results

We first performed single cell sequencing of Head and Neck Squamous Cell Carcinoma (HNSCC) tumors, and we found IL-1β expression is restricted to the tumor infiltrating myeloid cell populations. From our working hypothesis that myeloid inflammasome signaling can promote T-cell independent tumor growth, we initially showed that caspase 1 is important for promoting tumor growth in a myeloid specific manner. However, the mechanisms by which myeloid-intrinsic inflammasome activation promotes T-cell independent tumor growth are not well understood. Here we utilize a B16 melanoma tumor growth model to show that NLRP3 dependent caspase 1 activation is driving tumor growth in vivo, while AIM2 has no effect on tumor growth. In addition, we have found that circulating T cells in tumor bearing inflammasome knockout mice have higher expression levels of the checkpoint molecule PD-1. We utilized a C57Bl/6 mouse model of B16 melanoma to demonstrate that combination of inflammasome blockade and checkpoint blockade further reduces tumor burden in vivo.

Conclusions

This work will have broad implications across all solid tumor types and will allow for the understanding of how myeloid-specific production of IL-1β promotes tumor growth. In addition, the proposed work will contribute to our basic knowledge of inflammasome activation within the tumor microenvironment.

Correspondence: Catalina Lee-Chang (catalina.leechang@northwestern.edu)

Background

Recent clinical trials in brain tumor immunotherapy have gathered tremendous attention while at the same time establishing a favourable safety profile and preliminary efficacy. Immunotherapy is a promising approach in other malignancies but needs to be tailored specifically to glioblastoma (GBM) given the highly immunosuppressive tumor microenvironment [1, 2]. A hallmark of GBM microenvironment is the massive infiltration of bone marrow-derived tumor-associated myeloid cells (TAMCs) and regulatory T-cells (Tregs) [3-7]. We recently identified the existence of perivascular regulatory B-cells (Bregs) that blocked the activation of cytotoxic CD8+ T-cells in a subset of GBM patients. Even though tremendous progress has been made to understand GBM-mediated immunosuppression [8-10], there is still limited knowledge regarding the mechanisms by which the tumor selects, maintains and promotes immunosuppressive cells in detriment of anti-tumor immunity. A deeper comprehension of these process will propel the rational design of successful immunotherapies for GBM. Recent studies in our laboratory where tumor-infiltrating myeloid cells were specifically depleted using antibody-targeting drug delivery via nanoparticles in preclinical settings, pointed out to the almost total dependency of Tregs and Bregs on TAMCs to maintain their cellular abundance and immunosuppressive functions. The data presented here highlight the biological relevance of TAMC-derived microvesicle in controlling remotely regulatory lymphocytes.

Methods

Microvesicles were isolated by ultracentrifugation. Microvesicle uptake by lymphocytes was assessed by flow cytometry using fluorescent lipophilic dye.

Results

TAMC can control both Bregs and Tregs via microvesicle production. Microvesicle uptake by Tregs and Bregs was assessed using CD163+ TAMCs from GBM tumors. Proteomics and immunophenotype analysis of TAMC-derived microvesicles revealed that inactive TGFb and PD-L1 are transported as “cargos”. Our data showed that microvesicle-TGFb controls Tregs and Bregs cellular abundance. On the other hand, microvesicle-PDL1 are transfer and recycled via endocytosis by target cells, conferring them with immunosuppressive properties.

Conclusions

Glioma TAMCs utilize microvesicles to remotely promote Tregs and Bregs survival and immunosuppressive functions. This data suggest TAMC-mediated immunosuppression extends beyond the tumor vicinity limits.

References

1. Vega EA, Graner MW, Sampson JH. Combating immunosuppression in glioma. Future Oncol 2008;4(3):433-42 doi 10.2217/14796694.4.3.433.

2. Wainwright DA, Nigam P, Thaci B, Dey M, Lesniak MS. Recent developments on immunotherapy for brain cancer. Expert Opin Emerg Drugs 2012;17(2):181-202 doi 10.1517/14728214.2012.679929.

3. Wintterle S, Schreiner B, Mitsdoerffer M, Schneider D, Chen L, Meyermann R, et al. Expression of the B7-related molecule B7-H1 by glioma cells: a potential mechanism of immune paralysis. Cancer Res 2003;63(21):7462-7.

4. Raychaudhuri B, Rayman P, Ireland J, Ko J, Rini B, Borden EC, et al. Myeloid-derived suppressor cell accumulation and function in patients with newly diagnosed glioblastoma. Neuro Oncol 2011;13(6):591-9 doi 10.1093/neuonc/nor042.

5. Fecci PE, Mitchell DA, Whitesides JF, Xie W, Friedman AH, Archer GE, et al. Increased regulatory T-cell fraction amidst a diminished CD4 compartment explains cellular immune defects in patients with malignant glioma. Cancer Res 2006;66(6):3294-302.

6. El Andaloussi A, Lesniak MS. An increase in CD4+CD25+FOXP3+ regulatory T cells in tumor-infiltrating lymphocytes of human glioblastoma multiforme. Neuro Oncol 2006;8(3):234-43 doi 10.1215/15228517-2006-006.

7. El Andaloussi A, Han Y, Lesniak MS. Prolongation of survival following depletion of CD4+CD25+ regulatory T cells in mice with experimental brain tumors. J Neurosurg 2006;105(3):430-7 doi 10.3171/jns.2006.105.3.430.

8. Nduom EK, Weller M, Heimberger AB. Immunosuppressive mechanisms in glioblastoma. Neuro Oncol 2015;17 Suppl 7:vii9-vii14 doi 10.1093/neuonc/nov151.

9. Kamran N, Kadiyala P, Saxena M, Candolfi M, Li Y, Moreno-Ayala MA, et al. Immunosuppressive Myeloid Cells' Blockade in the Glioma Microenvironment Enhances the Efficacy of Immune-Stimulatory Gene Therapy. Mol Ther 2017;25(1):232-48 doi 10.1016/j.ymthe.2016.10.003.

10. Kamran N, Chandran M, Lowenstein PR, Castro MG. Immature myeloid cells in the tumor microenvironment: Implications for immunotherapy. Clin Immunol 2016 doi 10.1016/j.clim.2016.10.008.

Ethics Approval

IRB N° STU00202003, AICUC protocol N° IS00002459

Correspondence: Jacques Fieschi (jacques.fieschi@haliodx.com)

Background

Among the various immune cells recruited to the tumor microenvironment (TME), tumor-associated macrophages (TAMs) are particularly abundant and play a central role at all stages of cancer progression. TAMs represent a heterogeneous population in terms of cell morphology, functions, and tissue localization. Activated M1 macrophages express pro-inflammatory cytokines and drive potent antitumoral Th1 responses. Alternatively, activated M2 macrophages induce Th2 cells and enhance cancer progression and metastasis. The specific polarization of TAM may therefore skew the TME toward tumor rejection (M1) or favor immune suppression (M2). An additional level of complexity is conferred by their ability to shift between M1 and M2 subtypes (plasticity), depending on microenvironmental stimuli. Albeit the extensive literature on TAMs protumoural function and poor prognosis, some studies have shown that these cells possess an antitumour activity and their increase correlates with patient’s survival. Noteworthy, emerging studies suggest that TAMs are important in tumor metastases formation by regulating epithelial mesenchymal transition of colorectal cancer (CRC) cells. Until now, the most commonly used marker to identify human macrophages was the pan-macrophage marker CD68. The latter overlooks the prognostic potential of macrophage subsets with different functions in the TME. Therefore, there is a strong and urgent need to dissect TAMs heterogeneity in tumors using more refined approaches such as multiplex technology.

Methods

Here we assessed the presence, abundance and localization of TAMs subsets within the CRC microenvironment by using the Brightplex® solution, an automated sequential chromogenic multiplex assay. A unique combination of biomarkers (CD11b, CD68, CD86, CD163, CD204, CD206, and CD64) was developed to assess M1 and M2 TAMs heterogeneity and plasticity on a single FFPE tumor tissue section by preserving tissue’s spatial contexture.

Results

Briefly, a tissue section was sequentially stained, digitized, unstained and re-stained with antibodies targeting the seven markers. Images of the whole slide were then analyzed by digital pathology and the detection of positive cells was performed for each marker independently. In addition, tissue segmentation tools were used to assess TAMs densities in parenchyma, tumor stroma and invasive margin regions. This new panel of the Immunoscore® Suppressor Cells family enhances the ability to unravel the tumor microenvironment.

Conclusions

Deciphering the heterogeneity and plasticity of TAMs populations in CRC could help to identify new reliable prognostic markers to improve patient’s stratification and design more individualized therapeutic approaches.

Correspondence: Arnon Rosenthal (ar@alector.com)

Background

The SIRPα-CD47 axis represents a myeloid checkpoint in cancer. Recent clinical data with CD47-targeted therapies that competitively block the interaction with SIRPα show clinical responses in some patients with hematologic and solid cancers when combined with anti-tumor antigen antibodies. However, known limitations associated with targeting CD47, such as clearance of red blood cells (RBCs) and disruption of the SIRPγ-CD47 interaction important for T cell activation, likely hinder the anti-tumor immune response necessary for robust clinical benefit. A therapeutic advantage may be realized by targeting SIRPα and promoting immunostimulatory pathways that drive anti-tumor immunity. Here, we describe the discovery and characterization of a SIRPα-specific antibody, AL008, a first-in-class inhibitor of SIRPα that simultaneously antagonizes SIRPα signaling and stimulates Fcγ receptor (FcγR) activation.

Methods

Multiple functional assays were performed to ascertain the mechanism of action of AL008 and to benchmark activity against anti-CD47/anti-SIRPα antibodies in the clinic. For example, in vitro phagocytosis assays determined whether AL008 promoted engulfment of various tumor cell lines by human macrophages. Mixed leukocyte reactions (MLR) assessed the effect of AL008 on human dendritic cell-mediated T cell function. Furthermore, AL008 was administered to non-human primates to obtain pharmacokinetic and immunotoxicology profiles. Anti-tumor activity and pharmacodynamic effects of AL008 were examined in tumor-bearing mouse models expressing human SIRPα and human CD47.

Results

AL008 is a non-competitive inhibitor of SIRPα that triggers SIRPα internalization and degradation from the cell surface and simultaneously activates FcγR. Unlike other competitive antagonist antibodies to SIRPα, which require combination with opsonizing antibodies to promote tumor cell phagocytosis, AL008 stimulates tumor cell engulfment by macrophages in the absence of opsonizing antibody due to intrinsic FcγR activation. Because of its restricted binding specificity towards SIRPα, AL008 promotes dendritic cell-mediated stimulation of T cell proliferation, which contrasts with anti-CD47 antibodies and other anti-SIRPα/γ cross-reactive antibodies. Despite FcγR engagement by AL008, in vitro ADCC assays and in vivo studies in non-human primates verify that AL008 does not deplete myeloid cells or RBCs, which anticipates a significantly improved safety profile in patients. Lastly, inhibition of tumor growth by AL008 in mouse models coincides with downregulation of SIRPα and induction of activation markers on tumor-associated macrophages.

Conclusions

AL008 potentiates the anti-tumor effector functions of macrophages and dendritic cells in vitro, as well as in vivo models. This dual mechanism of AL008, which couples the reversal of inhibitory signal with immunostimulation, provides a novel therapeutic strategy for targeting myeloid cells for immune activation.

Ethics Approval

All in vivo studies were reviewed and approved by Alector Institutional Animal Care and Use Committee (IACUC).

Correspondence: Dariusz Stepniak (dariusz.stepniak@thermofisher.com)

Background

V-set and Immunoglobulin Domain Containing 4 (VSIG4), also known as Complement Receptor of the Immunoglobulin superfamily (CRIg), is a cell surface receptor structurally related to the B7 family of immune regulatory proteins. Expression of VSIG4 on macrophages renders them less responsive to LPS. In addition, VSIG4 promotes immune tolerance by attenuating early T cell activation and supporting the induction and maintenance of Foxp3 in T cells. Expression of VSIG4 on tumor-infiltrating macrophages suggests that it may be implicated in immune evasion.

Methods

Using our new VSIG4 (CRIg) antibody (clone NLA14), we demonstrate that Large Peritoneal Macrophages (LPM) in mouse consist of two distinct subsets, the VSIG4+ LPM and VSIG4- LPM.

Results

VSIG4+ LPM co-express Arginase 1 more frequently than the VSIG4- LPM, particularly in C57BL/6 mice. Further comparative analysis of the two LPM populations using a panel of macrophage specific antibodies revealed that VSIG4+ and VSIG4- LPM show a similar pattern of expression that is distinct from Small Peritoneal Macrophages (SPM). This phenotypic diversity within the LPM most likely reflects differences in function and activation status. In vitro cultured macrophages show a spontaneous decrease in the expression of VSIG4, and addition of proinflammatory factors, including IFN-γ and LPS, further reduces VSIG4 expression.

Conclusions

Characterization of VSIG4+ LPM might provide an opportunity to better understand the immunoregulatory role of VSIG4 in tumor infiltrating macrophages.

Correspondence: Brian Gastman (gastmab@ccf.org)

Background

IL-17A is a key pro-inflammatory cytokine indicated in multiple pathologies including tumor progression via suppression of anti-tumor immunity. While IL-17A is a signature cytokine of CD4+ T helper cells (Th17 cells), IL-17A is also produced by other cell types, including CD8+ T cells and type 3 innate lymphoid cells (ILC3s) in the mucosal surfaces. It remains unclear what are the key cellular sources of IL-17A in cutaneous squamous cell carcinoma.

Methods

Human cutaneous squamous cell carcinoma were collected from patients undergoing surgery to remove skin tumors that were diagnosed as SCCs by a dermatopathologist. Specimen collection and processing was approved by the Institutional Review Board of the Cleveland Clinic Foundation. For flow cytometry analysis, fresh tumors were cut into small pieces and dissociated with collagenase IV/hyaluronidase, followed by staining with antibodies against CD45 (BioLegend, clone 2D1,368529), Linage (BD, lin3, 643510) and IL-17A (Bioligand, Clone BL168, 512319). For immunohisotochemistry/fluorescence staining, tissues were fixed with 10% formalin overnight and then kept in 70% ethanol at 4 °C until processed into paraffin tissue blocks by Imaging Core at Cleveland Clinic Lerner Research Institute.

Results

We detect CD45+Lin-(CD3-CD14-CD19-CD2-)IL-17A+ cells in the cutaneous squamous cell carcinomas (cSCCs) by flow cytometry of the cell suspensions prepared from tumor tissues. We found CD3-IL-17+ cells in tumor tissue of skin cSCCs by immunohistochemistry staining of serial sections of SCCs from both immunocompetent and immunocompromised patients (e.g. transplant patients on iatrogenic long-term immunosuppressive therapy). In some cases, the CD3-IL-17+ cells consist of over 90% of the total IL-17+ cells in the tumor tissue. Furthermore, these CD3-IL-17+ cells are negatively stained for SMA, CD11b and CD19, suggesting that they are unlikely to be fibroblast, myeloid cells or B cells.

Conclusions

Taken together, we found a population of lineage negative IL-17A producing cells present in the cSCCs, which share the “CD45+Lin-” features with innate lymphoid cells. We are actively identifying this cell population. This study suggests that IL-17A can be produced by immune cell populations other than T cells in skin SCCs.

Correspondence: Robert Ferris (ferrisrl@upmc.edu)

Background

Resistance to the current generation of immunotherapies is mediated by complex relations between stromal, cancer and immune cells found within the tumor microenvironment (TME). Development of more efficacious drugs is predicated on improved understanding of these multi-spatial interactions. With emergence of new immune checkpoint receptor (ICR)-targeting therapies, a better understanding of topological expression of immune checkpoint ligand (ICL) on suppressive cell types in the TME may allow for improved strategies to treat cancer patients.

Methods

Single cell RNA sequencing (scRNAseq) was performed from head and neck squamous cell carcinoma (HNSCC) specimens (n=17) with matched blood from treatment-naïve patients. Immune and non-immune cells were enriched from tumor cell suspensions. Novel transcriptomic cell-to-cell interactions were predicted between heterogeneous cell populations. Histologic inflammation was corroborated with scRNAseq and multi-plex flow cytometry.

Results

Major cell type clusters (immune, epithelial, fibroblast and endothelial cells) were identified. Expression patterns for PD-1, TIGIT, LAG-3 and TIM-3 ligands were evaluated on these suppressive TME cell types. While LGALS9/galectin-9 levels were expressed highly on tumor-associated macrophages (TAMs) from both inflamed and non-inflamed tumors, CD274/PD-L1 was only upregulated on TAMs from inflamed lesions. By modeling receptor-ligand interactions between CD8+ T cells and the rest of the major TME cell types, CD8+ T cells were predicted to form more ICR-ICL interactions with TAMs than with any other cell type. Flow cytometry confirmed the abundance of PD-L1 and galectin-9 on TAMs.

Conclusions

Our data suggest that in the setting of HNSCC, TAMs are the major contributors of ICL in the HNSCC TME. Strategies that selective target this immunosuppressive population may be necessary to break tolerance to PD-1-targeting therapies.

Ethics Approval

The study was approved by the University of Pittsburgh’s Institutional Review Board (#99-069).

Correspondence: Cassian Yee (cyee@mdanderson.org)

Background

Recent reports suggest that tissue resident memory cells (T_RM) may play an important role in anti-tumor immunity. Specifically, the accumulation of CD103+ tumor infiltrating lymphocytes (TIL_RM) is associated with favorable prognosis [1,2]. However, relatively little is known regarding T_RM differentiation and endogenous T_RM are difficult to isolate, impeding their basic study and translational application. Thus a means to generate T_RMin vitro would be desirable. TGF-β is known to be critical in establishing T_RM populations and attempts to identify other factors have focused on cytokines [3,4]. Oxygen tension is another factor that distinguishes the circulation from peripheral tissues; thus we hypothesized that hypoxia may contribute to a TGF-β1-induced T_RM phenotype in human CD8+ T-cells.

Methods

Naïve CD8+ T-cells from human peripheral blood were activated in normal cell culture conditions (20% O2) or hypoxia (2% O2) and then cultured in the presence of rhTGF-β1. Cells were then analyzed for expression of T_RM-associated genes via quantitative real-time PCR (qPCR) and T_RM-associated surface markers via flow cytometry. The individual contributions of hypoxia and TGF-β1 to induction of T_RM phenotype cells were assessed by performing in vitro differentiation experiments as described above with or without the addition of rhTGF-β1. CD69-CD103- cells generated in 20% O2, CD69+CD103- cells generated in 2% O2, and CD69+CD103+ cells generated in 2% O2 + TGF-β1 were sorted before RNA-sequencing (n=3).

Results

We demonstrate that differentiation of human CD8+ T-cells in hypoxia and TGF-β1 in vitro led to the development of a T_RM phenotype, characterized by an increase in CD69+CD103+ cells expressing CD49a, CD101, and PD-1 – hallmarks of human T_RM (n=7, p_RM phenotype cells than the additive effects of either condition alone (n=4, p_RM and TIL_RM gene signatures in CD69+CD103+ cells generated in hypoxia and TGF-β1 (Figure 1).

Conclusions

Our findings identify a previously unreported cue for T_RM differentiation and enable a facile means of generating human T_RM phenotype cells in vitro with the potential for application to multiple existing Adoptive Cell Therapy (ACT) modalities. As hypoxia and TGF-β are common features of the tumor microenvironment our results indicate that these conditions may contribute to generation of CD103+ TIL in vivo.

Acknowledgements

This work was supported by the Parker Institute for Cancer Immunotherapy (PICI); C.Y. is a member of PICI

References

1. Smazynski J, Webb JR. Resident Memory-Like Tumor-Infiltrating Lymphocytes (TILRM): Latest Players in the Immuno-Oncology Repertoire. Front Immunol. 2018;9:1741. doi: 10.3389/fimmu.2018.01741.

2. Kim Y, Shin Y, Kang GH. Prognostic significance of CD103+ immune cells in solid tumor: a systemic review and meta-analysis. Sci Rep. 2019;9(1):3808. doi: 10.1038/s41598-019-40527-4.

3. Casey KA, Fraser KA, Schenkel JM, Moran A, Abt MC, Beura LK, et al. Antigen-independent differentiation and maintenance of effector-like resident memory T cells in tissues. J Immunol. 2012;188(10):4866-75. doi: 10.4049/jimmunol.1200402.

4. Skon CN, Lee JY, Anderson KG, Masopust D, Hogquist KA, Jameson SC. Transcriptional downregulation of S1pr1 is required for the establishment of resident memory CD8+ T cells. Nat Immunol. 2013;14(12):1285-93. doi: 10.1038/ni.2745.

Ethics Approval

This study was approved by University of Texas MD Anderson Cancer Center’s Institutional Review Board, approval number PA14-0105.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P604).

In vitro induction of tissue resident memory phenotype cells

Correspondence: Saurabh Chhabra (schabra@mcw.edu)

Background

Severe cytokine release syndrome (CRS) results in an increased incidence of infections after T-cell engaging therapies. Several retrospective, single-center studies have shown infectious complications with haploidentical hematopoietic cell transplantation (haploHCT) but no data correlating infections with CRS that develops after haploHCT is available. [1-10]

Methods

We evaluated 78 consecutive adult haploHCT recipients at our center for the development of CRS, graded by Lee et al. criteria, and examined the incidence of infectious complications in correlation with CRS severity. All patients received haploHCT for hematologic malignancies between April 2012 and April 2018, using post-transplant cyclophosphamide (PTCy) and tacrolimus/MMF for GVHD prophylaxis. The incidence of infections was examined in two separate time periods in relation to the day of stem cell infusion (day 0): day 0-100 and day 101-180. The study cohort was further stratified by CRS grades into 2 groups: 53% of patients in cohort A, CRS grade 0-1 (n=41) and 47% in cohort B, CRS grade ≥2 (n=37).

Results

Overall, 81% patients developed infections (63/78) within 180 days. Fewer patients in cohort A developed an infection (71% vs. 92%; P=.018). 58% of these infections were of viral origin (104 episodes) and nearly half of those were CMV reactivations (46%). 38% infections were of bacterial origin (70 episodes) and 5% fungal. On multivariate analysis, higher-grade CRS (CRS grade ≥2) was independently associated with an increased frequency of bacterial infections (RR 2.25; 95% CI, 1.12 to 4.53; P=.023) and there was also evidence of higher overall infections (RR 1.6; 95% CI, 0.98 to 2.63; P=.057). Increasing frequency of overall and bacterial infections, including BSI and CLABSI, was observed as CRS became more severe (higher in grade 3 vs grade 2). CMV recipient positivity was independently associated with an increased frequency of overall infections (RR 1.98; 95% CI, 1.23 to 3.18; P=.0048) as well as an expectedly greater viral infections (RR 2.77; 95% CI, 1.60 to 4.78; P=.0003). 10 patients died of infection/sepsis-related causes and the cumulative incidence of infection-related mortality was higher in cohort B (Figures 1-3).

Conclusions

These findings suggest that the severity of CRS developing after haploHCT using PTCy-platform is associated with an increased frequency of bacterial infections, including BSI and CLABSI, and may lead to increased mortality. Pathogenesis of CRS after haploHCT remain an unmet need and potential mechanisms that may confer an increased frequency of infections are also discussed.

References

1. Marks DI, Abid MB. A Stem Cell Donor for Every Adult Requiring an Allograft for Acute Lymphoblastic Leukemia? Biol Blood Marrow Transplant. 2017;23(2):182-183.

2. Bashey A, Zhang X, Sizemore CA, et al. T-cell-replete HLA-haploidentical hematopoietic transplantation for hematologic malignancies using post-transplantation cyclophosphamide results in outcomes equivalent to those of contemporaneous HLA-matched related and unrelated donor transplantation. J Clin Oncol. 2013;31(10):1310-1316.

3. Abboud R, Keller J, Slade M, et al. Severe Cytokine-Release Syndrome after T Cell-Replete Peripheral Blood Haploidentical Donor Transplantation Is Associated with Poor Survival and Anti-IL-6 Therapy Is Safe and Well Tolerated. Biol Blood Marrow Transplant. 2016;22(10):1851-1860.

4. Lee DW, Gardner R, Porter DL, et al. Current concepts in the diagnosis and management of cytokine release syndrome. Blood. 2014;124(2):188-195.

5. Park JH, Romero FA, Taur Y, et al. Cytokine Release Syndrome Grade as a Predictive Marker for Infections in Patients With Relapsed or Refractory B-Cell Acute Lymphoblastic Leukemia Treated With Chimeric Antigen Receptor T Cells. Clin Infect Dis. 2018;67(4):533-540.

6. Hill JA, Li D, Hay KA, et al. Infectious complications of CD19-targeted chimeric antigen receptor-modified T-cell immunotherapy. Blood. 2018;131(1):121-130.

7. Cho C, Perales MA. Rapid identification of cytokine release syndrome after haploidentical PBSC transplantation and successful therapy with tocilizumab. Bone Marrow Transplant. 2016;51(12):1620-1621.

8. Slade M, Goldsmith S, Romee R, et al. Epidemiology of infections following haploidentical peripheral blood hematopoietic cell transplantation. Transpl Infect Dis. 2017;19(1).

9. Crocchiolo R, Bramanti S, Vai A, et al. Infections after T-replete haploidentical transplantation and high-dose cyclophosphamide as graft-versus-host disease prophylaxis. Transpl Infect Dis. 2015;17(2):242-249.

10. Fayard A, Daguenet E, Blaise D, et al. Evaluation of infectious complications after haploidentical hematopoietic stem cell transplantation with post-transplant cyclophosphamide following reduced-intensity and myeloablative conditioning: a study on behalf of the Francophone Society of Stem Cell Transplantation and Cellular Therapy (SFGM-TC). Bone Marrow Transplant. 2019.

Ethics Approval

The study was approved by Froedtert & the Medical College of Wisconsin Institutution‘s Ethics Board, approval number PRO00032170.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P605).

See text for description

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2 (abstract P605).

See text for description

Correspondence:Lisa Butterfield (lbutterfield@parkerici.org)

Background

Metastatic melanoma patients successfully treated with combination immunotherapy using checkpoint blockade often display pre-existing antitumor immunity at time of initial treatment. Cancer vaccination can be used to promote anti-tumor T cell responses in patients lacking such immunity.

Methods

To generate optimal anti-tumor immune responses in melanoma patients, we previously developed a dendritic cell (DC)-based adenovirus-engineered vaccine encoding three common melanoma-associated antigens: Tyrosinase, MART-1, and MAGE-A6. This vaccine was used to treat 35 patients in a Phase 1 clinical trial. Immature DC were matured using IFN + LPS, and then transduced with the recombinant adenovirus yielding the vaccine. Patients received three intradermal injections of vaccine over one month. In addition to standard phenotyping and cytokine testing, a human genome RNA microarray was used to transcriptionally profile the DC vaccines produced for each patient. Cryopreserved patient and healthy donor monocytes were also examined in functional assays.

Results

Patient (matured and adenovirally-transduced) monocyte-derived DC expressed reduced levels of cell surface ICOSL (a costimulatory molecule) vs. DC generated from healthy donors (HD). Microarray analysis of patient cells revealed that ICOSLG mRNA gene expression was increased post-maturation and viral transduction, and differently expressed between HD and patient cells.

ICOSL surface protein showed a positive trend with the ability of adenovirally-transduced DC to elicit specific CD8+ T cell responses in vitro to Tyrosinase and MART-1. Using a CMV p65 antigen-based model, we observed reduced activation of naïve, CMV-specific CD8+ and CD4+ T cells in seronegative donors when primed using antigen-loaded autologous DC treated with anti-ICOSL antibody vs. an IgG control.

Microarray analysis further suggested that (matured and adenoviral-transduced) patient DC exhibit dysregulated NFĸB signaling vs. DC isolated from HD, and non-canonical NFκB signaling known to transactivate ICOSL. Pharmacologic inhibition of NFĸB signaling in DC from HD also resulted in decreased expression of surface ICOSL. We also observed that patient vaccine DC expression of NLRP2 (a natural NFĸB inhibitor), negatively correlated with clinical outcome.

Conclusions

Reduced levels of ICOSL on patient vaccine DC may limit their ability to prime anti-tumor immune responses in vitro (and in vivo). DC-expressed ICOSL appears important for specific T cell priming and is at least partially regulated by intrinsic canonical NFĸB signaling pathway. These data identify critical DC stimulatory pathways and inform the next generation DC-based vaccines capable of promoting superior anti-tumor benefits in advanced-stage melanoma patients.

Acknowledgements

This study utilized the UPMC Hillman Cancer Center’s Immunologic Monitoring and Cellular Products Laboratory shared facility, supported in part by award P30 CA047904.

This work is funded by the NIH SPORE IN SKIN CANCER Grant 4P50CA121973-09, the Melanoma Skin Cancer T32 Award 129297 (08/1/2017 – 07/31/2017), and the CTSI TL1 Pre-doctoral Scholar Fellowship, TL1 TR001858 (10/01/2018 – 09/30/2019).

Trial Registration

NCT01622933

Ethics Approval

The study was approved by the University of Pittsburgh's Ethics Board, approval number IRB12010416.

Correspondence: Jinpu Yu (jinpu_yu@hotmail.com)

Background

Early stage myeloid-derived suppressor cells (eMDSCs) are a newly identified immunosuppressive cell subtype, but less is known about the generation of eMDSCs yet. In this study we elucidated that tumor-derived IL-6 manipulated the development and function of eMDSCs in vitro and vivo.

Methods

We constructed IL-6 knockdown 4T1 mammary tumor-bearing BALB/C models and myeloid lineage specific SOCS3 knockout C57BL/6 model. The system of eMDSCs induction was established, in which bone marrow cells were cocultured with 4T1 cells to induce eMDSCs in vitro.

Results

We identified a subpopulation of CD11b+Gr-1-F4/80-MHCII- eMDSCs in situ in IL-6 knockdown 4T1 mammary tumor-bearing mice models, which displayed more potent suppression on T cell immunity than conventional CD11b+Gr-1+ MDSCs. The proportion of eMDSCs in tumor tissues was highly correlated with the tumor size and the numbers of lung metastatic nodules. Bone marrow cells were co-cultured with 4T1 cells to induce eMDSCs in vitro, but IL-6R Ab treatment significantly decreased the proportion of eMDSCs. Tumor-derived IL-6 inhibited the expression of SOCS3 and persistently activated the JAK/STAT3 signaling pathways, which further caused the differentiation arrest of myeloid linkage and comparable immunosuppressive capacity of eMDSCs. The conditional myeloid SOCS3 knockout C57BL/6 (SOCS3mKO)mice was constructed to study SOCS3-related amplification of eMDSCs. The eMDSCs increased significantly in SOCS3mKO mice compared to SOCS3fl/fl mice, in which myeloid differentiation arrest was observed. The miRNA array screening indicated miR155-5p was the predominant regulator of SOCS3-related differentiation arrest in eMDSCs. We demonstrated that miR155-5p suppressed transcription factors C/EBPβ and PU.1, and thus promoted the development of eMDSCs. Transfecting miR155-5p inhibitor in myeloid precursors could efficiently reduce the number of eMDSCs and recover myeloid differentiation in vitro.

Conclusions

Tumor derived IL-6 activated the JAK/STAT3/ miR155-5p pathway in myeloid precursors by inducing SOCS3 deficiency, which consequently downregulated C/EBPβ and PU.1, induced myeloid differentiation arrest, and promoted the development of eMDSCs.

Acknowledgements

This work was supported by National Natural Science Foundation of China (grant numbers: 81872143, 81702280, 81472473, 81272360) and Key Project of Tianjin Health and Family Planning Commission (16KG126).

Correspondence: Karin Enell Smith (kae@alligatorbioscience.com)

Background

ATOR-1017 is a Fcγ-receptor (FcγR) crosslinking dependent agonistic IgG4 antibody targeting the costimulatory receptor 4 1BB. 4-1BB is highly expressed on tumor infiltrating CD8+ T effector cells (T effs) in several cancer indications. By binding to 4-1BB, ATOR-1017 can enhance the activity of tumor reactive T effs and NK cells within the tumor and induce a potent anti-tumor attack. 4-1BB is an attractive target for immunotherapy, even though 4-1BB targeting antibodies in clinical development so far have been suffering from toxicity or poor efficacy. Therefore, ATOR-1017 was designed to improve tolerability and to enhance efficacy.

Methods

Human 4-1BB knock-in transgenic mice with established murine colon carcinoma MC38 tumors were used to demonstrate anti-tumor efficacy and immune cell infiltration with FACS after systemic treatment with ATOR-1017. In vitro, immune activation (IFN-γ production) of purified human immune cells from healthy donors with ATOR-1017 was analyzed using FcγR expressing cells for crosslinking in the presence of human IgG.

Results

ATOR-1017 reduced tumor growth and improved survival dose dependently in human 4-1BB knock-in transgenic mice with established tumors. No tumor growth was detected after tumor re-challenge in the complete responders previously treated with ATOR-1017, but in the naïve untreated control group only. These results demonstrated that ATOR-1017 induced a long-lasting immunological memory. The immunological response induced by ATOR-1017, leading to an efficient tumor eradication, consisted of a significant immune cell infiltration and a high T eff to T regulatory cell ratio within the tumor specifically but not in the spleens. In vitro it was demonstrated that immune activation with ATOR-1017 was blocked dose dependently in the presence of human IgG, due to saturation of the FcγRs which blocked crosslinking of ATOR-1017.

Conclusions

ATOR-1017 induced a potent tumor-directed immune response, leading to an efficient tumor eradication and survival and development of a long-lasting immunological memory. ATOR-1017 is FcγR crosslinking dependent with an activation profile that is expected to minimize the risk for inducing systemic immune activation and toxicity, by directing the immune response to the tumor tissue and tumor draining lymph nodes where 4-1BB as well as certain FcγRs are highly expressed. The FcγR crosslinking-dependency will also reduce the risk for immune activation in the circulation, due to the high concentration of endogenous circulating IgG which will compete with ATOR-1017 for binding to the FcγRs [1,2]. A first-in-human phase I study of intravenously administered ATOR-1017 is planned for 2019.

References

1. Preithner S, Elm S, Lippold S et al. High concentrations of therapeutic IgG1 antibodies are needed to compensate for inhibition of antibody-dependent cellular cytotoxicity by excess endogenous immunoglobulin G. Mol Immunol 2006; 43:1183-1193.

4. Eigenmann MJ, Karlsen TV, Krippendorff BF et al. Interstitial IgG antibody pharmacokinetics assessed by combined in vivo- and physiologically-based pharmacokinetic modelling approaches. J Physiol 2017; 595:7311-7330

Ethics Approval

The study was approved by Institutional Animal Care and Use Committee (IACUC) of China, approval number 10278L001.

Correspondence: Ugur Sahin (ugur.sahin@biontech.de), Maria Jure-Kunkel (mjk@genmab.com)

Background

PD-1/PD-L1-targeting monoclonal antibodies (mAbs) have shown impressive responses in a wide range of tumors, although only in a subset of patients, leaving room for improvement. 4-1BB agonistic mAbs are potent T-cell co-stimulators and showed promising results in pre-clinical cancer models. However, no 4-1BB agonist has advanced beyond early phase II studies, due to dose-limiting hepatitis or lack of monotherapy efficacy. Fc-silenced bispecific antibodies (bsAbs) targeting PD-L1 and 4-1BB may overcome these limitations by combining immune checkpoint inhibition with conditional agonistic activity.

Methods

Fc-silenced PD-L1x4-1BB bsAbs (anti-mouse surrogate: mbsAb-PD-L1x4-1BB; anti-human: DuoBody®-PD-L1x4-1BB [GEN1046]) were obtained by controlled Fab-arm exchange of PD-L1 and 4-1BB mAbs. Activity of the mouse-specific mbsAb-PD-L1x4-1BB was tested in vivo in an ovalbumin-specific OT-1 T-cell adoptive transfer and vaccination model in C57BL/6 mice, and in efficacy studies in CT26 and MC38 tumor-bearing BALB/c and C57BL/6 mice, respectively. Tolerability was assessed in C57BL/6 mice and compared to treatment with a 4-1BB agonistic mAb. Induction of T-cell proliferation and cytokine production by DuoBody-PD-L1x4-1BB was analyzed in primary human T-cell assays in vitro and using patient-derived tumor-infiltrating lymphocytes (TILs) ex vivo. The non-clinical safety profile of DuoBody-PD-L1x4-1BB was assessed in cynomolgus monkeys.

Results

Surrogate mbsAb-PD-L1x4-1BB induced significant expansion of adoptively transferred OT-1 T-cells upon ovalbumin vaccination in proof-of-concept studies in vivo. In a therapeutic setting, mbsAb-PD-L1x4-1BB induced long-term tumor remission in up to 90% of CT26 and MC38 tumor-bearing mice, associated with expansion of tumor-specific CD8+ T-cells. mbsAb-PD-L1x4-1BB was well tolerated, while an Fc-active mouse 4-1BB agonistic mAb, alone or with a PD-L1-specific mAb, induced a significant increase in serum transaminase levels and liver immune infiltrates.

DuoBody-PD-L1x4-1BB enhanced T-cell proliferation and pro-inflammatory cytokine production in vitro, and was more potent than checkpoint blockade alone or the combination of Fc-silenced PD-L1- and 4-1BB-specific mAbs. 4-1BB agonist activity of DuoBody-PD-L1x4-1BB was strictly conditional and dependent on simultaneous binding of both targets. DuoBody-PD-L1x4-1BB increased TIL expansion ex vivo. DuoBody-PD-L1x4-1BB showed a favorable safety profile in cynomolgus monkeys at doses up to 30 mg/kg (1q3wx2).

Conclusions

Dual targeting of PD-L1 and 4-1BB with an Fc-silenced bsAb combines immune checkpoint blockade and conditional T-cell co-stimulation in one molecule. This distinctive mechanism of action results in potent anti-tumor activity without inducing hepatotoxicity in mice and enhances proliferation of human primary T-cells and TILs in vitro and ex vivo. The clinical safety of DuoBody-PD-L1x4-1BB is currently being assessed in patients with solid cancers in a first-in-human trial (NCT03917381).

Ethics Approval

Cynomolgus monkey studies were performed at Charles River Laboratories, Tranent, UK, in accordance with the EU legislations described in Directive 2010/63/EU. This study was licensed by UK Home Office under the Animals (Scientific Procedures) Act 1986 (approval number PBAD559F8, Toxicology of Pharmaceuticals, Protocol 15) and overseen by Genmab B.V.

All mice studies were performed by BioNTech SE at its research facilities in Germany, and the mice were housed in accordance with German federal and state policies on animal research. All experiments were approved by the regulatory authorities for animal welfare in Germany.

The use of tumor tissue resections was approved by BioNTech SE’s Ethics Board, approval number 837.309.12 (8410-F).

Correspondence: Sribalaji Lakshmikanthan (sribalajilakshmikanthan@gmail.com)

Background

Claudin18.2 (CLDN18.2), a gastric epithelial junction protein, has been discovered as a potential therapeutic target for stomach cancers. The aim of our study was to elucidate mechanism of action for ABP150, a CLDN18.2 specific CD3 engager. Specifically, we assessed the ability of ABP150 to be redirected to CLDN18.2 tumors followed by T-cell dependent killing.

Methods

CLDN18.2 specific clones were obtained by Abpro’s proprietary platform [1]. The clone selection criteria involved Claudin-isoform specific binding and off-target screening; HTS-flow cytometry based cross-reactive studies using ≥5000 other proteins expressed on HEK293T cell system. The bispecific T-cell engager for ABP150 antibody was engineered as described [1]. The built constructs were transiently transfected into HEK-293S cells grown in FreeStyle 293 Expression Medium followed by purification using Protein A IgG Purification Kit. Purified ABP150 was subjected to Size Exclusion Chromatography (SEC-HPLC) to assess its monomeric state. Relative binding of ABP150 to CLDN18.2 positive tumors and CD3 expressing Jurkat-T cell line was performed using flow cytometry. Following manufacturer’s protocol, T-cell activation and T-cell Dependent Cytotoxicity (TDCC) were examined by T Cell Activation Bioassay (NFAT) and CellTiter-Glo luminescent cell viability assay kits respectively.

Results

Out of 7 hybridomas, 2 clones, ABP150-1 and ABP150-2 were finalized based on relative binding affinity and specificity to CLDN18.2. Flow cytometric experiments demonstrated dose-dependent binding to CLDN18.2 positive cells exclusively. No binding to CLDN18.1 isoform was observed. Their ability to bind to T-cells were comparable to ABP100 [2, 3] suggesting functional monovalent binding to CD3. In the off-target screening experiments, both clones showed no cross-reactivity to extracellular membrane proteins. SEC-HPLC analysis displayed ≥95% of monomeric protein with no aggregations. In T-cell activation bioassay, both ABP150 clones presented a dose dependent T cell activation. Further, TDCC experiments resulted in dose dependent tumor killing only in the presence of CLDN18.2 expressing tumors. In absence of CLDN18.2 tumors, no toxicity was observed.

Conclusions

Lead CLDN18.2 clones, ABP150-1 and ABP150-2 as T-cell engagers showed CLDN18.2 specific binding and functionally weak CD3 binding (in the absence of target) resulting in strong TDCC response only in the presence of CLDN18.2 tumors. Taken together, we conclude that these two clones are expected to efficiently clear CLDN18.2 tumors without cytokine release syndrome.

References

1. Lakshmikanthan, S., et al., Novel tetravalent bispecific T-cell engaging antibodies for cancer immunotherapy. Journal of Clinical Oncology, 2019. 37, no. 15_suppl.

2. Boudot, A., et al., ABP-100: A Tetravalent Bispecific T-cell Engaging Antibody for HER2+ Solid Tumors. Annals of Oncology, 2018. Volume 29 Supplement 8.

3. Lopez-Albaitero, A., et al., Overcoming resistance to HER2-targeted therapy with a novel HER2/CD3 bispecific antibody. Oncoimmunology, 2017. 6(3): p. e1267891.

Correspondence: Paula Hammond (hammond@mit.edu)

Background

Interleukin-12 (IL-12) is a particularly compelling target for cytokine therapy, as it showed promise in preclinical tests. However, it failed in the clinic to high schedule dependent toxicities, including two patient deaths. An alternative approach is to deliver IL-12 in a nanoparticle (NP) delivery vehicle. Yet, cytokines pose a difficult challenge for NP delivery, as they require efficiently loading labile proteins, maintaining activity on cytokine membrane receptors from a NP that is often internalized, and achieving high association with tumors to prevent systemic leakage. Layer-by-layer (LbL) is a simple electrostatic method of layering polymers onto materials to alter the properties of the carrier. Through incorporation of different polymers, LbL offers the opportunity to address the challenges of cytokine NP delivery.

Methods

A single chain variant of IL-12 was loaded onto liposomal nanoparticles via Ni:HIS interactions and layered via the LbL process (Figure 1A). Particles were assessed for material properties in vitro including biological activity via their ability to stimulate an interferon-y response. Particles were further tested in vivo for their toxicity and efficacy in multiple tumor models.

Results

LbL IL-12 NPs with poly-L-arginine and poly-L-glutamic acid layers (PLE-IL-12-NPs) were able to efficiently load and deliver active cytokine. PLE-IL-12-NPs showed drastically reduced toxicity compared to carrier-free IL-12. Healthy mice treated with PLE-IL-12-NPs showed no differences in weight to PBS and unloaded NPs, while carrier-free IL-12 treated mice lost 10% of their body weight by the fifth dose when treated daily. Furthermore, PLE-IL-12-NP treated mice showed significantly less systemic cytokine response when compared to carrier-free IL-12. Importantly, PLE-IL-12-NPs also showed no loss in anti-tumor activity at these same doses when delivered against both subcutaneous MC38 and the more difficult to treat HM-1 ovarian cancer cell line. Finally, at significantly higher doses, carrier-free IL-12 again showed significant toxicity while the PLE-IL-12-NPs showed no differences to the controls and lead to pronounced antitumor efficacy in the MC38 model (Figure 1B,C).

Conclusions

PLE-IL-12-NPs show promise as an IL-12 therapy. Most importantly, PLE-IL-12-NPs were able to significantly reduce the toxicity of IL-12 therapy while maintaining its efficacy. Of noted importance, PLE-IL-12-NPs were also able to show efficacy in ovarian cancer, which has been refractory to many immunotherapies in the past. Finally, although the above results demonstrate activity with IL-12, the described particle uses no IL-12 specific methods and can likely be applied across many other cytokines or combinations

Ethics Approval

These experiments were approved by the Massachusetts Institute of Technology Committee on Animal Care (CAC).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P611).

PLE-IL-12-NPs reduce toxicity and maintain efficacy

Correspondence: Marcos Milla (mmilla@synthorx.com)

Background

Upon TCR stimulation via tumor-specific antigen presentation, CD8+ cells upregulate the IL-10 receptor (IL-10R) and become responsive to this cytokine, leading to localized T effector cell (Teff) proliferation and tumor infiltration along with the release of IFN and cytolytic products. The presence of 13 lysine residues in IL-10 and participation of its N-terminal region in IL-10R1 and R2 engagement presents a challenge to conventional PEG strategies. To harness IL-10 potential as an immuno-oncology therapeutic, the generation of IL-10 variants with extended half-life is desired.

Methods

The Synthorx synthetic biology platform expands the genetic alphabet (EGA) by one base pair that generates new codons for site-specific incorporation of novel amino acids at any position within the polypeptide, which enables the design of biotherapeutics with improved drug properties. We have applied this platform to rapidly screen and identify residues within the IL-10 amino acid sequence for optimized PEG bioconjugation distal from the IL-10R binding sites, producing a series of single IL-10 variants that were then pegylated. Biochemical characterization of those pegylated variants by SPR and SEC-MALS together with in vitro and ex vivo potency profiling led to the identification of homogeneous Synthorin IL-10 dimers with extended in vivo half-life and minimal interference with IL-10 receptor engagement and activation.

Results

Analytical characterization of purified IL-10 Synthorins showed that they are homogeneous, and dimeric pegylated variants. Among them, THOR-1003 and THOR-1009 showed comparable in vitro potency to native IL-10 for proliferation of the human MC/9 cell line, and in a human recombinant cell assay directly measuring receptor activation, underscoring the importance of choosing the right site for pegylation on the agonist activity of IL-10. Other sites tested resulted in compounds exhibiting up to 1,000-fold decrease in potency in cell-based assays. THOR-1003 and THOR-1009 have been selected for further testing, focusing on the release of IFN and cytolytic enzymes from primary human CD8+ T cells, following stimulation with antigen-presenting cells displaying a CMV peptide epitope. Pharmacokinetics and pharmacodynamics profiling of these molecules in vivo is ongoing.

Conclusions

We have demonstrated that our synthetic biology platform can generate functional, pegylated Synthorin IL-10 variants that retain the folding and structural stability of native IL-10. Two of those Synthorins, THOR-1003 and THOR-1009, exhibit physiochemical properties similar to native IL-10, and functional properties consistent with a fully active agonist that can be dosed in vivo for durable pharmacodynamic effects.

Correspondence: Marcos Milla (mmilla@synthorx.com)

Background

IL-15, distinctly from IL-2, induces proliferation and survival of CD8+ T memory (Tmem) cells, activates natural killer (NK) cells, and enhances antibody-driven cellular toxicity without the expansion of immunosuppressive regulatory CD4+ T cells (Treg). However, its high potency and resulting toxicity, combined with its short half-life, has limited its potential as an immunotherapeutic agent. We hypothesize that extension of the half-life of IL-15 could provide persistence of signaling for CD8+ Tmem expansion, while reducing potential Cmax-associated toxicities. Further modulation of the interaction with specific receptor chains (the private IL-15Rα and/or IL-2/15Rβ) by pegylation might enable tuning of drug exposure (AUC) for preferential stimulation of different cell subsets and therapeutic effects.

Methods

We applied the Synthorx Expanded Genetic Alphabet platform to the rational design of pegylated IL-15 compounds with extended half-life and differential interaction with the IL-15Rα and IL-2/15Rβ. In vitro biochemistry and cell bioassays and ex vivo profiling in mouse and human immune cells were used to select initial IL-15 Synthorin candidates for in vivo PK/PD studies.

Results

THOR-924 is a pegylated IL-15 Synthorin with similar binding affinity to IL-15Rα and IL-2/15Rβ, and comparable signaling potency, relative to native IL-15. PK/PD responses in mice, demonstrated an extended half-life with sustained induction of pSTAT5 and Ki67 resulting in strong proliferation of peripheral CD8+ Tmem and NK cells, without proliferation of Treg cells. 90-100% of peripheral CD8+ T cells were positive for both pSTAT5 and Ki67 after a single IV dose of 0.3 mg/kg. Additional IL-15 Synthorins, THOR-918 and THOR-908, showed overall reduced potency or specific blockade of IL-15 interaction with IL-15Rα, respectively, by SPR. This selective interaction accounts for the differential in vitro and ex vivo potency of these compounds compared to native IL-15 and THOR-924. Ongoing in vivo studies will shed light on the potential pharmacologic advantages of such activity-modulated IL-15 Synthorins.

Conclusions

IL-15 Synthorins are biochemically differentiated forms of IL-15 with extended half-life and optimized pharmacologic parameters for balanced engagement of IL-15R vs IL-2/15R inducing proliferation and cytotoxicity of CD8 T and NK cells in vivo without significant activation of immune-suppressive Treg cells or toxicities associated with dosing of native IL-15. They offer the potential to fulfill the promise of IL-15 as a clinical agent for therapeutic induction of immune responses in the treatment of solid tumors.

Correspondence: Brian Champion (brian.champion@psioxus.com)

Background

T-SIGn virus vectors for cancer gene therapy are transgene-modified variants of enadenotucirev (EnAd), an Ad11p/Ad3 chimeric group B adenovirus, which retain all the functional and epithelial tumor-selectivity properties of EnAd while also mediating the selective expression of transgenes by infected tumor cells. We have created T-SIGn viruses expressing a single chain variant of human IL-12 as the primary transgene modality; production of IL-12 by virus-infected tumor cells could be an effective way of driving T-cell and NK cell activity within the tumor microenvironment to promote effective anti-tumor immunity.

Methods

T-SIGn virus constructs were generated to express either the p35 and p40 chains of human IL-12 separately or as a linked single chain molecule using a short linker sequence. To enable tumor-selective expression of IL-12, transgenes were placed under the control of the virus major late promoter (MLP) which is activated following initiation of genome replication in permissive cells. Tumor cell lines and primary cells derived by gentle disaggregation of surgically excised tumor samples from carcinoma patients were used to evaluate the production and activities of the different IL-12 viruses.

Results

ELISA assays measuring levels of the full IL-12 p70 molecule or the specific p40 chain in infected tumor cell supernatants showed that the virus encoding separate IL-12 chains produced p40 very effectively but very little p70 was detected. By contrast the virus encoding the single chain variant effectively produced p70 at levels >100ng/ml. Using a HEK-Blue IL-12 signalling reporter assay, this p70 protein was shown to be functionally active and thus folded correctly. Primary cells derived from excised patient tumor samples were also evaluated and shown to produce IL-12 when treated with the single chain IL-12 encoding virus. The functional activity of virus encoded IL-12 was then further demonstrated by its ability to enhance the production of IFNγ by T-cells and NK cells using PBMCs and PBMC-derived T-cell cultures as well as cultures of primary tumor cells.

Conclusions

In this study, we have shown that a functionally active single chain variant of human IL-12 can be effectively produced by tumor cells infected with a T-SIGn virus encoding this transgene payload. Ongoing studies are evaluating several different cytokines and chemokines encoded together with the single chain IL-12 transgene for their ability to enhance the activities of IL-12, either directly or indirectly via their ability to recruit IL-12 responsive immune cells in order to optimize the potential for clinical activity.

Correspondence: Christian Klein (christian.klein.ck1@roche.com)

Background

Interleukin-2 (IL-2) has been the first effective cancer immunotherapy to treat metastatic melanoma and renal cell carcinoma for more than three decades. Although only a small fraction of patients benefits from such a treatment, the anti-tumor responses to IL-2 therapy tend to be complete, durable and the responding patients seem to be cured. Unfortunately, the applicability of IL-2 in the clinic is limited by its toxicity (i.e. capillary leak syndrome), unfavourable pharmacokinetic and bio-distribution in the lymphatic system rather than in the tumour, and the detrimental activation/expansion of regulatory T cells (Tregs) through IL2Ra (CD25) binding.

Methods

To overcome these limitations, we developed a novel immunocytokine that is constituted of a high affinity bivalent blocking anti-PD-1 antibody and a heterodimeric silent Fc-part fused to a monomeric IL-2 variant (IL2v).

Results

IL2v binding to CD25 is abolished, and therefore the binding to Tregs and endothelial cells is reduced. The anti-PD-1 antibody, having roughly 50-fold higher affinity than IL-2v, dictates the preferential cis-delivery of IL-2v to PD-1 positive T cells, such as antigen-experienced tumour-reactive T cells. In addition, the blocking of the PD-1 inhibitory pathway synergizes with the IL-2 signalling cascade translating into more potent T cell activation and effector functions than untargeted IL-2v. Due to reduced expression levels of PD-1 on Tregs compared to tumour-reactive T cells, PD1-IL2v binds preferentially to the latter rescuing them from Treg-mediated suppression. All these properties enabled PD1-IL2v to eradicate aggressive tumours and provide a significant survival benefit over combination of the parental molecules in the tested mouse models.

Conclusions

PD1-IL2v is a differentiated immune cell-targeted IL-2v which promotes an effective and long-term anti-tumor immune-response by delivering IL-2v to PD-1+ tumor-specific T cells

Correspondence: Lei Fang (lei.fang@i-mabbiopharma.com)

Background

Lymphocyte counts in the peripheral immune system and tumor are correlated with positive clinical outcomes in PD-(L)1 immunotherapy. Interleukin 7 (IL7) is an immune homeostasis cytokine for T cells. Treatment with recombinant human IL7 preferentially expands recent thymic emigrants, naive and central memory T cells, but spares regulatory T cells. Here we constructed a series of immunocytokine L1I7 comprised of a PDL1 blocking mAb fused with an IL7 molecule with varying degree of attenuated potency. The aim of the study was to identify the right molecule that achieves enhanced biodistribution at the tumor site and an optimal balance between synergistic anti-tumor activity and an acceptable safety profile.

Methods

PDL1 binding affinity and antagonist function of L1I7 were evaluated by BiaCore and PD1/PDL1 cell-based functional assay, respectively. IL7 receptor (IL7R) binding and internalization were evaluated in human primary CD4+ T cells by flow cytometry (FACS). IL7 potency was evaluated by p-STAT5 signaling assay and human CD4+ T cell proliferation assay. The in vitro synergistic effect of L1I7 was evaluated in the human mixed lymphocyte reaction (MLR) assay. The in vivo tracking assay was conducted by injecting fluorescence-labeled L1I7 into HCC827-transplanted CD34+ hematopoietic stem cells humanized mice. The in vivo efficacy of L1I7 was investigated in B16F10 melanoma syngeneic mice model with surrogate L1I7. Absolute numbers of tumor-infiltrating CD4+ T and CD8+ T cells were analyzed by FACS.

Results

L1I7 series of immunocytokine maintained PDL1 binding and antagonist function when compared with PDL1 mAb. As compared with wildtype L1I7, L1I7 variants showed attenuated IL7 activity as evidenced by reduced IL7R binding/internalization, p-STAT5 activation and CD4+ T cell proliferation. Both wildtype and the attenuated variants were more effective in promoting CD4 T cell activation than anti-PDL1 or IL7 monotreatment in MLR. Similar with PDL1 mAb, L1I7 was enriched in the tumor site whereas IL7 was widely distributed in in vivo tracking experiment. In an anti-PDL1-resistant B16F10 tumor model, L1I7 showed superior efficacy as compared with either anti-PDL1, IL7 alone or in combination, which was correlated with an increased number of tumor-infiltrating CD4+ T and CD8+ T cells.

Conclusions

The novel immunocytokine L1I7 is designed to target poor and non-responders to PD-(L)1 treatment and even PD-(L)1 treatment relapsed patients. L1I7 represents potentially a best-in-class molecule with enhanced local immunostimulation achieved by concurrent PD-(L)1 blockade and IL7 stimulation. L1I7 is undergoing preclinical development with an aim to enter clinical studies in 2021.

Correspondence: Tian Zhao (tzhao@onconanomed.com)

Background

Immunotherapy has become a powerful strategy for cancer treatment. Immunomodulators such as interleukin-2 (IL-2) can induce anti-tumor immune responses, but their clinical applications are limited by unfavorable pharmacokinetic properties that can elicit serious dose-limiting toxicities (e.g. vascular leak syndrome). [1] We have developed an ultra-pH sensitive nanoparticle platform that targets the acidic tumor microenvironment (TME) with high specificity. At normal physiological pH, the nanoparticles exist as intact micelles, while at acidic tumor pH, they dissociate into single unimers, releasing the payload. The feasibility of selectively targeting the acidic TME by this platform has been successfully validated by imaging of different types of tumors in a Phase 1 clinical trial. [2-4] Using this platform, we have encapsulated and delivered IL-2 to the TME to improve its pharmacokinetic properties.

Methods

We have developed an ultra-pH sensitive nanoparticle (ONM-400) with similar pH sensitivity to ONM-100, which is being tested in a Phase 2 clinical trial for intraoperative imaging of tumors. Multiple strategies have been exploited to encapsulate IL-2 including mixing, single emulsion, acid/base titration, double emulsion/solvent evaporation, and covalent conjugation. The ONM-400•IL-2 complexes were purified by ultracentrifugation and characterized by dynamic light scattering and transmission electron microscopy. Encapsulation efficiency (EE) and drug loading (DL) were measured for each formulation using ELISA or gel electrophoresis. In addition, IL-2 was labeled with a near-infrared dye (IRDye® 800CW) to study the in vivo tumor accumulation and biodistribution after intravenous injection in mice bearing head and neck tumors using a LI-COR Pearl® camera.

Results

ONM-400 nanoparticles were able to encapsulate IL-2 with high loading. The EE and DL ranged from 81.8% to 35.7% and 8.2% to 3.5%, respectively. In vivo imaging of ONM-400•IL-2 showed significantly higher tumor accumulation and less bladder signal compared to free IL-2 at the same dose after IV injection (Fig. 1A). ONM-400•IL-2 provided longer and increased IL-2 exposure in tumor than free IL-2 which decreases overtime. Quantitative analysis of the tumor signal at different time points indicated an alteration of pharmacokinetics for IL-2 after encapsulation by the ONM-400 nanoparticle (Fig. 1B). Biodistribution at 24h further revealed a lower accumulation in kidney but higher accumulation in tumor for ONM-400•IL-2 versus free IL-2.

Conclusions

The ONM-400 nanoparticles can efficiently encapsulate IL-2 and alter its pharmacokinetics in vivo. These results suggest a great potential for using the ultra-pH sensitive nanoparticle technology to delivery immunomodulators to improve their efficacy and safety profile.

References

1. Riley R, June C, Langer R, Mitchell M. Delivery technologies for cancer immunotherapy. Nat Rev Drug Discov. 2019; 18: 175-196.

2. Steinkamp P, Voskuil F, Koller M, van der Vegt B, Doff J, Zhao T, Hartung J, Jayalakshmi Y, Sumer B, Gao J, Witjes M, van Dam G. Image guided surgery for tumor agnostic detection of solid tumors using the pH activated micellar imaging agent ONM-100: The Shine study. Cancer Res. 2019; 79 (13 Supplement): CT018.

3. Voskuil F, Steinkamp P, Koller M, van der Vegt B, Doff J, Zhao T, Hartung J, Jayalakshmi Y, Sumer B, Gao J, Witjes M, van Dam G. Image-guided surgery for tumor agnostic detection of solid tumors using the pH-activated micellar imaging agent ONM-100. J Clin Oncol. 2019; 37 (15 Supplement): 3068.

4. Steinkamp P, Voskuil F, Koller M, van der Vegt B, Doff J, Zhao T, Hartung J, Jayalakshmi Y, Sumer B, Gao J, Witjes M, van Dam G. Image guided surgery for tumor detection in breast cancer using the PH activated micellar tracer ONM-100: The SHINE study. Cancer Res. 2019; 79 (4 Supplement): P2-14-29.

Ethics Approval

All animal experiments were reviewed and approved by, and performed in accordance with, the UT Southwestern Institutional Animal Care and Use Committee under Animal Protocol Number: 2015-101065.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P617).

See text for description

Correspondence: William Dower (bdower@medikine.com), Ronald Barrett (rbarrett@medikine.com)

Background

Efforts to modify IL-2 for immuno-oncology applications have focused on alterations that reduce interaction with the Rα subunit of the receptor complex via mutation, chemical modification, or complexation with antibodies or the Rα-ectodomain. IL-2Rβγc-agonists may also incorporate features to reduce side effects associated with peak drug levels, and to extend duration of action.

Methods

Peptides were selected from recombinant peptide libraries to identify molecules binding simultaneously to the β and γc subunits of the IL-2 receptor. Identified peptides were synthesized by both chemical and recombinant means, and evaluated for properties of IL-2Rβγc activation in IL-2 responsive cells lines and human lymphocytes.

Results

We report the discovery and optimization of small synthetic peptides, unrelated to IL-2, that simultaneously bind IL-2Rβ and γc subunits to induce IL-2R signaling. The peptides do not bind IL-2Rα and are therefore IL-2Rβγc-selective agonists with MW less than 5000D. Synthetic compound MK1169 activates the JAK-STAT (pSTAT5), MAPK (pERK1/2), PI3K (pAKT) pathways, and induces proliferation in cell lines NK92 (Rαβγc) and TF-1β (Rβγc). Agonism is attributable to direct activation of the IL-2Rβγc receptor: (1) conversion of IL-2Rβ-deficient parental TF-1 cells to the IL-2-responsive TF-1β line, by transfection of Rβ, imparts MK1169 responsiveness, and (2) neutralizing antibodies against IL-2 and IL-15 do not inhibit the agonist activity of MK1169. MK1169 exhibits potency similar to that of IL-2 in cell lines expressing Rβγc, but not Rα. MK1169 induces pSTAT5 in human lymphocyte subpopulations with potency bias similar to that of IL-2 variants that are attenuated in Rα binding. MK1169-treated human PBMCs cause PD-L1 up-regulation, and killing of tumor cell lines. MK1169 retains activity when fused to IgG and Fc molecules, and when pegylated; providing options for pharmacokinetic (PK) enhancement. We have also shown that fusions of MK1169 to pembrolizumab and cemiplimab exhibit full activity in both PD-1 inhibition and IL-2Rβγc agonist activity, indicating the feasibility of incorporating MK1169 into fusion proteins with dual pharmacology. The synthetic peptide, and peptide fused to the C-terminus of Fc exhibit promising in vitro serum stability, and the in silico predicted potential for immunogenicity is very low.

Conclusions

PK-enhanced MK1169 is an attractive alternative to engineered IL-2 variants. Dual pharmacology therapeutics incorporating MK1169 also offer promise for immuno-oncology. To our knowledge, this work is the first demonstration of small peptide agonists of a heterodimeric cytokine receptor.

Ethics Approval

The use of human PBMC in this study was authorized under Minimal Risk Research Related Activities at Stanford Blood Center (SQL 79075)

Correspondence: Takahiro Miyazaki (tmiyazaki@nektar.com)

Background

NKTR-255 is a polymer-conjugated IL-15 pathway agonist that retains binding affinity to the alpha subunit of the IL-15 receptor and exhibits reduced clearance. Sustained IL-15 pathway activation leads to expansion and activation of NK cells potentially enabling combinatorial immunotherapies including tumor targeting antibodies that mediate tumor killing by antibody dependent cellular cytotoxicity (ADCC). Here we investigate NKTR-255 induced enhancement of NK cell ADCC dependent tumor cell killing in vitro and in vivo in the tumor environment when combined with tumor-targeting monoclonal antibodies in solid tumors.

Methods

NKTR-255 dose dependent enhancement of NK cell cytotoxicity as a single agent or in combination with cetuximab was evaluated in vitro in co-cultures of purified human NK cells with HCT-116 or FaDu tumor cells. Target cell killing was evaluated by flow cytometry based viability scoring. In vivo tumor growth inhibition was measured in xenograft models with diverse histologies including colorectal (Colo25, HT29, HCT-116), lung (H1975), epidermal (A431), ovarian (SKOV3) and gastric (NCI-N87). Nude or SCID mice in Balb/c background were subcutaneously inoculated on the flank with tumor cells. NKTR-255 was administered intravenously weekly at 0.3mg/kg. Cetuximab or trastuzumab were administered at dose levels and schedules optimized for each tumor type with a once or twice a week schedule. Treatment was initiated at established ~100-150mm3 tumor volumes. Tumor growth inhibition was assessed by tumor size caliper measurements.

Results

NKTR-255 pre-treatment of NK cells led to enhanced cytotoxic activity against HCT-116 and FaDu cells in vitro. Additional enhancement of NK cell dependent killing of tumor cells was observed in the presence cetuximab antibody in the co-culture assay demonstrating NKTR-255 dependent enhancement of NK cell ADCC. In agreement with in vitro activity, NKTR-255 combination treatment with cetuximab or trastuzumab showed significant tumor growth inhibition in multiple models. Notably tumor models resistant to cetuximab single agent treatment (HT-29 and HCT-116) showed significant growth inhibition in combination treatment with NKTR-255. Trastuzumab targeted tumors showed significant enhancement of antibody mediated tumor growth inhibition in combination with NKTR-255 including 100% complete responses in the SKOV3 model.

Conclusions

NKTR-255 treatment of NK cells enhanced ADCC activity against tumor cells in vitro and in vivo efficacy of ADCC inducing antibodies in human solid tumor xenograft models. Furthermore, antibody single agent treatment resistant tumor models showed tumor growth inhibition in combination treatment with NKTR-255 suggesting potential for increasing response rates of therapeutic antibodies with ADCC mechanism of action.

Ethics Approval

All procedures in this study were reviewed and approved by Nektar Therapetuics Institutional Animal Care and Use Committee.

Correspondence: Ugur Sahin (ugur.sahin@biontech.de)

Background

Interleukin-2 (IL2) is known to support effector T cell differentiation, proliferation and survival. A particular challenge of IL2 for cancer treatment is the preferential stimulation of regulatory T cells (Tregs) constitutively expressing the high-affinity IL2 receptor alpha chain (CD25). Tregs can undermine the antitumoral activity of tumor-specific effector T cells only transiently expressing CD25. Interleukin-7 (IL7) is a homeostatic cytokine that plays an important role in lymphopoiesis, survival and memory formation, and can increase the expression of CD25 on antigen-specific T cells. Prompted by their complementary effects, we investigated the potential synergy between these two cytokines on the T cell level and on antitumoral activity.

Methods

We designed nucleoside-modified, RNA encoded, extended half-life IL2 and IL7, formulated as nanoparticles for intravenous delivery. The potency of IL2 and IL7 to enhance antitumoral activity of a liposome-formulated RNA vaccine [1] was determined for both cytokines separately and in combination. Therapeutic success and changes in blood T cell subset composition were determined in subcutaneous, syngeneic tumor-bearing mice.

Results

Using nanoparticle formulated, nucleoside-modified RNA coding for extended half-life IL2 and IL7, we demonstrated that IL2 and IL7 each increases the number and frequency of antigen-specific CD8+ T cells in combination with an antigen-specific RNA vaccine compared to the vaccine alone. While the effect on expansion of antigen-specific T cells of IL2 is strong and immediate, IL7 requires repetitive cycles to generate its maximal effect. In addition, IL7 turned out to expand CD8+ T cells with specificities for targets other than the vaccine-encoded antigen better than IL2. As expected, IL2 strongly expanded the number and frequency of Tregs. In contrast, IL7 reduced the fraction of Tregs among CD4+ T cells. The combination of IL2 with IL7 proved to be powerful in overcoming their individual shortcomings: The combination boosted the number of antigen-specific CD8+ T cells beyond the levels of IL2 alone, supported the expansion of CD8+ T cells not specific for the vaccine-encoded target, and further improved the ratio of antigen-specific CD8+ T cells over Tregs. These findings translated into superior antitumoral activity of the combination in subcutaneous mouse tumor models.

Conclusions

Based on the complementary mode of action and the promising preclinical data, clinical evaluation will be pursued.

References

1. Kranz, L. M., Diken, M., et al. Systemic RNA delivery to dendritic cells exploits antiviral defence for cancer immunotherapy. Nature 534, 396–401 (2016).

Ethics Approval

Experimental group sizes were approved by the regulatory authorities for animal welfare after being defined to balance statistical power, feasibility and ethical aspects. All mice were kept in accordance with federal and state policies on animal research at BioNTech SE.

Correspondence:Chris Ullman (chris.ullman@cts-bioscience.com), IL Minn (iminn1@jhmi.edu)

Background

Therapies through systemic injection of recombinant immunomodulatory cytokines have not been successful due to low therapeutic index. Here we describe the use of a polycation-encapsulated DNA plasmid to locally deliver therapeutic cytokine genes under the control of the cancer-selective rat progression elevated gene 3 (PEG-3) promoter1,2 to tumors in the lungs of diseased mice. [1-3]

Methods

A cGMP grade polycationic carrier, in vivo-jetPEI® was chosen to formulate the plasmid DNA construct to aid rapid transition to clinical studies. Clinically compatible vectors were created that featured a CpG-free backbone containing a kanamycin resistant gene, scaffold/matrix attachment regions and CpG-free payload genes. Plasmids were then packaged into size-controlled nanoparticles with in vivo-jetPEI® using a scalable flash nanocomplexation (FNC) manufacturing method3 for production of lyophilized, shelf-stable nanoparticles with an excellent reproducibility under quality assurance and quality control standards.

Results

These well-defined pDNA/lPEI nanoparticles, delivered systemically, were well-tolerated in mice and accomplished significant reduction in levels of acute inflammatory cytokines compared to an equivalent CpG-containing plasmid. In vivo transfection of metastatic lesions in the lung by nanoparticles expressing firefly luciferase from the PEG-3 promoter could be detected bioluminescent imaging. Furthermore, intravenous delivery of nanoparticles expressing murine single-chain interleukin-12, under the control of PEG-3 promoter, significantly improved the survival of mice in two different models of lung cancer with no marked symptoms of systemic toxicity (Figure 1).

Conclusions

Increased survival in both an orthotopic and a metastatic model of lung cancer using clinically relevant nanoparticles raises the possibility of translation to human therapy.

Acknowledgements

This study is partially supported by a research contract from Cancer Targeting Systems, Inc. (CTS) and an NIH grant R01EB018358.

References

1. Bhang, H. E. et al. Tumor-specific imaging through progression elevated gene-3 promoter-driven gene expression. Nat. Med, 2011; 17(1), 123-129.

2. Minn, I. et al. Molecular-genetic imaging of cancer. Adv, Can. Res. 2014; 124:131-69.

3. J. L. Santos et al., Continuous production of discrete plasmid DNA-polycation nanoparticles using flash nanocomplexation. Small, 2016; 12(45), 6214-6222.

Ethics Approval

All animal experiments were performed in accordance with protocols approved by Animal Care and Use Committees of the Johns Hopkins University and of the Pennsylvania State University.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P621).

Expression of mIL-12 provides survival benefits

Correspondence: Takahiro Miyazaki (tmiyazaki@nektar.com)

Background

IL-15 is a cytokine that activates and provides survival benefit to T and NK cells and has great potential for the treatment of cancer. Exploiting the therapeutic value of native IL-15 has been challenging due to its unfavorable pharmacokinetic properties and poor tolerability. Several IL-15-based proteins aimed to overcome these problems are currently in development. NKTR-255 is a polymer-conjugated native IL-15 that is designed to retain a binding affinity to IL-15 receptor alpha (IL-15Ralpha), while the most clinically advanced IL-15-based protein is an IL-15 superagonist (IL-15 mutein/IL-15Ralpha-Fc complex). We investigate the pharmacological properties of NKTR-255 and the IL-15 superagonist to explore differentiations between the IL-15Ralpha dependent and independent cytokines.

Methods

For cellular receptor binding, IL-15Ralpha-overexpressing HEK-293 cells were incubated with rhIL-15, NKTR-255 or IL-15 superagonist to measure surface binding of the cytokines by flow cytometry. The biological activity of cell surface bound cytokines was measured as pSTAT5 induction following co-culture with human PBMCs. Surface expression of IL-15Rs was measured by flow cytometry at 20 min or 2 hr following incubation with each of the cytokines. In functional assays, human PBMCs were incubated with each of the cytokines for 1, 2 or 5 days for surface marker characterization, cytokine production or proliferation assay, respectively.

Results

NKTR-255 binds to surface IL-15Ralpha on the receptor-overexpressing HEK-293 cells. Maximal binding was similar to that of rhIL-15, while IL-15 superagonist showed negligible binding. In concordance, the HEK-293 cells incubated with NKTR-255 induced pSTAT5 on human PBMCs, while the HEK-293 cells incubated with IL-15 superagonist failed to induce pSTAT5. Similarly to rhIL-15, incubation of NKTR-255 with human whole blood reduced surface staining of IL-15Ralpha on T and NK cells; in contrast IL-15 superagonist reduced IL-2Rbeta staining. Each of the cytokines dose-dependently induced proliferation of both NK cells and T cells; the EC50 value of NKTR-255 was approximately 20-fold higher than the EC50 values of both rhIL-15 and IL-15 superagonist. Interestingly, NKTR-255-treated NK cells showed more rounds of division than IL-15 superagonist-treated NK cells at the highest concentration. The maximal responses of NKTR-255 for induction of surface CD107a expression in NK cells and IFNgamma production were also higher than those of IL-15 superagonist.

Conclusions

NKTR-255 is a novel IL-15Ralpha dependent molecule that provides enhanced PK and PD properties relative to the native IL-15 cytokine while IL-15 superagonist is a direct engager to IL-2Rbeta complex. These results indicate the potential of NKTR-255 pharmacology to capture the full spectrum of native IL-15 biology.

Correspondence:Alexander Rakhmilevich (rakhmil@humonc.wisc.edu), Dan Spiegelman (dspiegelman2@wisc.edu), Ravi Patel (rpatel55@wisc.edu), Claire Baniel (cbaniel@gmail.com), Amy Erbe (aerbe@wisc.edu), Jacquelyn Hank (hank@humonc.wisc.edu), Deborah Charych dcharych@nektar.com), Willem Overwijk (wwoverwijk@gmail.com), Zachary Morris(zmorris@humonc.wisc.edu), Paul Sondel (pmsondel@humonc.wisc.edu)

Background

Bempegaldesleukin (NKTR-214) is a first-in-class, CD122-preferential interleukin-2 (IL2) pathway agonist being investigated for its potential ability to leverage the clinically validated IL2 pathway and selectively stimulate an immune response, without overactivating the immune system. We have previously shown a synergistic interaction between local low dose radiation therapy (RT) and NKTR-214 in mice bearing moderate sized melanoma tumors. This combination is less effective, however, in the presence of larger tumors and/or systemic disease. Checkpoint blockade effectively disables immune inhibitory functions, thereby enabling a stronger anti-tumor immune response. We hypothesized the addition of checkpoint blockade to [RT]+[NKTR-214] would help overcome the immune suppression in difficult models, allowing for a stronger anti-tumor effect.

Methods

To generate a model of bulky unresectable disease, C57BL/6 mice were inoculated with B78 melanoma cells on the right flank and tumors were allowed to grow for ~7 weeks (average tumor size ~800mm3). Mice were randomized and treated with 12 Gy external beam RT to the flank tumor and intravenous (IV) NKTR-214 or [RT]+[NKTR-214]+[α-CTLA-4]. To create a model of heterogeneous and disseminated metastatic disease, C57BL/6 mice were inoculated with B78 cells on the right flank and tumors were allowed to grow for ~4 weeks (average tumor size ~185mm3). On the day the treatments began, mice were randomized and injected IV with B16 Luciferase+ melanoma cells. B16 is parental to B78 melanoma, sharing some common antigens. Mice were treated with: 1) PBS, 2) [NKTR-214]+[α-CTLA-4], 3) [RT]+[α-CTLA-4], 4) [RT]+[NKTR-214]+[α-CTLA-4], or 5) [RT]+[NKTR-214]+[α-CTLA-4]+[T-cell depleting antibodies (α-CD4+α-CD8)]. Tumor growth and survival were monitored. Metastatic disease burden was evaluated by luciferase imaging.

Results

In very large B78 tumors (~800mm3), [RT]+[NKTR-214]+[α-CTLA-4] caused greater tumor regression than [RT] +[NTKR-214]. In a model of disseminated metastatic disease, [RT]+[α-CTLA-4] slowed primary tumor progression but had little effect on metastatic disease, demonstrated by the group’s similar survival kinetics to the PBS control. Adding [NKTR-214] to [RT]+[α-CTLA-4] not only caused primary tumor regression but also resulted in greater overall survival than all other groups in the B78 primary/B16 metastasis study. Control of the lung metastasis by [RT]+[NKTR-214]+[α-CTLA-4] was T cell-mediated. Though [NKTR-214]+[α-CTLA-4] demonstrated relatively weak control of the primary tumor, this combination slowed metastatic tumor burden displaying a prolongation of survival compared to vehicle and [RT]+[α-CTLA-4].

Conclusions

Our preliminary results suggest that [RT]+[NKTR-214]+[α-CTLA-4] has the ability to mount a T cell dependent anti-tumor response capable of regressing large unresectable tumors and disseminated, heterogeneous metastatic disease in murine melanoma models.

Correspondence: Paula Hammond (hammond@mit.edu)

Background

Ovarian cancer represents an area of great unmet clinical need, with the 5-year survival rate for high grade ovarian cancer at just 25% for those with advanced disease and with only ~ 10% of patients responding to checkpoint inhibition or other immune therapeutics [1]. Thus, there remains a pressing need to explore alternative strategies to potentiate immunotherapies in ovarian cancer. One attractive class of therapeutics that could play such a role are pro-inflammatory cytokines. However, many proinflammatory cytokines have been under-utilized in ovarian cancer and other solid tumors, largely due to toxicity concerns. In this study we develop a nanoparticle to deliver interleukin-12 (IL-12), one of the most potent and toxic cytokines, to an orthotopic model of murine ovarian cancer. The layer-by-layer (LbL) method was used to modify liposomes bearing IL-12 with a polymeric surface chemistry that covers the cytokine to limit its accessibility outside of the tumor environment, enhances ovarian cancer targeting, and prevents nanoparticle internalization.

Methods

IL-12 nanoparticles were constructed as shown in Figure 1A. Briefly, HIS-tagged, single-chain IL-12 was loaded onto the surface of liposomes bearing Ni headgroups. The liposomes were then layered with positively charged poly-L-arginine followed by poly-L-glutamine (PLE). These constructs (PLE-IL12-NP) were fully characterized and tested in vitro. For in vivo testing, female B6C3F1 mice were implanted intraperitoneally with 1MM HM-1 ovarian cancer cells. Beginning 7 days after implantation, mice were treated intraperitoneally 1x/day for five days with PLE-IL12-NP (5 μg). Mice were tracked for tumor burden via luciferase expression, survival, and signs of toxicity.

Results

Incubation with splenocytes confirmed that PLE-IL-12-NPs retained the activity of IL-12. Toxicity studies in mice showed that at 5 μg the PLE-IL-12-NPs reduced toxicity in terms of weight loss (Figure 2B) and systemic cytokine production (data not shown). Indeed, the weight of mice treated PLE-IL-12-NPs was indistinguishable from vehicle treated mice. Antitumor studies against established metastatic ovarian tumors showed that PLE-IL-12-NPs were just as efficacious as soluble IL-12 (Figure 2C) extending survival over both vehicle and unlayered cytokine-bearing liposomes.

Conclusions

PLE-IL-12-NPs offer a simple approach to ameliorate the toxicity of IL-12. Indeed, this approach offers promise for the systemic delivery of potent cytokines to hard-to-treat tumors such as ovarian cancer. Ongoing studies are exploring the biodistribution following additional routes of administration, the immunological mechanisms of these responses, and the combination of this approach with checkpoint inhibition.

References

1. Odunsi K. Immunotherapy in ovarian cancer. Ann Oncol. 2017;28:1-7

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P624).

Layer-by-layer nanoparticles for IL-12 delivery

Correspondence: Matthew McCourt (matthew.mccourt@kymab.com)

Background

IL-2 therapy has been approved for the use in certain cancers since 1989. The major limitation with this therapy is tolerability, with many patients experiencing grade 3 or higher adverse events. Approaches to improve this have focused on engineering molecules with selectivity towards the dimeric (CD122 and CD132) IL-2 receptor (IL-2R). However, this form of the receptor complex is widely expressed on peripheral naïve T cells as well as memory T cells and NK cells, potentially leading to unproductive stimulation of irrelevant cell populations and an increased possibility of systemic toxicity. KY1043 is a highly differentiated immunocytokine consisting of a neutralising anti-PD-L1 antibody, fused via its light chains to an attenuated IL-2, in which the balance of signaling has been adjusted to favour the trimeric form of the IL-2R, which contains CD25. We hypothesised that reduced binding to circulating cells may improve targeting to PD-L1+ cells in the tumour or lymphoid tissues while also targeting IL-2 to antigen-experienced T cells that express CD25 in the tumour microenvironment.

Methods

Tumour reactive HLA-A2+ CD4+ and CD8+ T cells were generated by co-culture with growth-arrested A375 cells, supplemented with IL-2 and IL-7. After 20 days, these tumour-directed T cells were cultured with fresh tumour cells in the presence of either KY1043, a non-targeted immunocytokine or the parental anti-PD-L1 antibody. Tumour cell growth was measured in real-time using the Incucyte S3™ system. In vivo, human PD-L1 knock-in (hPD-L1) mice bearing established MC38 hPD-L1 tumours were treated with KY1043.

Results

KY1043 induced T-cell-mediated killing of A375 tumour cells in vitro, with a higher potency compared to non-targeted immunocytokine or anti-PD-L1 alone, and comparable to that of anti-PD-L1 and equimolar amounts of recombinant native IL-2. In vivo, treatment with KY1043 led to a dose-dependent complete regression of hPD-L1-expressing MC38 tumours in hPD-L1 knock-in mice, with no significant adverse effects. Pharmacodynamic analysis confirmed proliferation of both effector (Teff) and regulatory (Treg) T cells in the periphery, with preferential expansion of Teff cells in the tumour. Immunological memory was demonstrated, as treated mice were resistant to re-challenge with MC38 but not to other tumours.

Conclusions

KY1043 induces potent T cell activation and can direct highly effective tumour killing in vitro and in vivo. These results challenge the dogma that selective binding to dimeric (CD122/CD132) IL-2R is required for tumour control and demonstrate that a CD25 directed immunocytokine may provide an advantageous therapeutic-index for the treatment of cancer.

Acknowledgements

We thank the Technology and Animal Sciences teams for support of knock-in mice and cell line generation, and the Antibody Discovery team for identification of the parental anti-PD-L1 antibody

Ethics Approval

This project was reviewed and approved by the Greater Manchester West Local Research Ethics Committee (LREC no. 2/NW/0553)

Correspondence: Ugur Sahin (ugur.sahin@biontech.de)

Background

Interleukin-2 (IL2) supports the differentiation, proliferation, survival and effector functions of T cells [1–4]. Recombinant IL2 has been used for decades in the treatment of late stage malignant melanoma [5] and renal cell cancer [6,7]. Most patients with complete responses after IL2 treatment remain regression-free for more than 25 years after initial treatment, but overall response rates are low [8,9,7,5]. Recombinant IL2 treatment is complicated by its very short half-life and requirement of high and frequent dosing, which in turn potentiates side effects [10]. Furthermore, a particular challenge of IL2 for cancer treatment is the preferential stimulation of immunosuppressive regulatory T cells (Tregs), which can dampen anti-tumor immune responses and may counteract the beneficial effects of activated effector T cells.

Methods

We designed a nucleoside-modified RNA encoded IL2 variant with the goal of an extended half-life, reduced Treg bias and potent stimulation of effector T cells and NK cells. Activation and proliferation of human, cynomolgus or mouse T cells was characterized in vitro. IL2 variant serum levels after intravenous RNA nanoparticle administration were investigated in mice and cynomolgus monkeys. Subsequently, immune subset composition as well as therapeutic activity as monotherapy or in combination with cancer immunotherapies was determined in multiple syngeneic mouse tumor models.

Results

Characterization of the IL2 variant in vitro validated potent proliferation of CD8+ T cells, NK cells and a reduced Treg bias. Analysis of IL2 variant protein levels in blood of mice or non-human primates upon injection of RNA nanoparticles demonstrated drug exposure for several days. The IL2 variant mediated a substantial increase in murine CD8+ effector T cells and NK cell while Treg expansion was hardly detected in vivo. In syngeneic tumor-bearing mice, a spontaneous expansion of cancer (neo-epitope) specific T cells and a strong increase of the CD8+ T cell to Treg ratio translated into potent tumor control. When combined with T-cell vaccination anti-PD-1/PD-L1 checkpoint blockade, the IL2 variant significantly increased tumor specific T-cell expansion, further boosting anti-tumor efficacy.

Conclusions

We demonstrate substantial therapeutic activity of a RNA encoded, nanoparticle formulated IL2 variant in monotherapy or in combination with cancer immunotherapy. Patient recruitment for an open-label, clinical phase I/II, first-in-human trial is anticipated for the first half of 2020.

References

1. Gillis S, Smith KA: Long term culture of tumour-specific cytotoxic T cells. Nature 1977, 268:154–6.

2. Blattman JN, Grayson JM, Wherry EJ, Kaech SM, Smith KA, Ahmed R: Therapeutic use of IL-2 to enhance antiviral T-cell responses in vivo. Nat Med 2003, 9:540–7.

3. Bamford RN, Grant AJ, Burton JD, Peters C, Kurys G, Goldman CK, Brennan J, Roessler E, Waldmann TA: The interleukin (IL) 2 receptor beta chain is shared by IL-2 and a cytokine, provisionally designated IL-T, that stimulates T-cell proliferation and the induction of lymphokine-activated killer cells. Proc Natl Acad Sci U S A 1994, 91:4940–4.

4. Kamimura D, Bevan MJ: Naive CD8+ T cells differentiate into protective memory-like cells after IL-2 anti IL-2 complex treatment in vivo. J Exp Med 2007, 204:1803–12.

5. Atkins MB, Lotze MT, Dutcher JP, Fisher RI, Weiss G, Margolin K, Abrams J, Sznol M, Parkinson D, Hawkins M, et al.: High-dose recombinant interleukin 2 therapy for patients with metastatic melanoma: analysis of 270 patients treated between 1985 and 1993. J Clin Oncol 1999, 17:2105–16.

6. Rosenberg SA: Interleukin 2 for patients with renal cancer. Nat Clin Pract Oncol 2007, 4:497.

7. Klapper JA, Downey SG, Smith FO, Yang JC, Hughes MS, Kammula US, Sherry RM, Royal RE, Steinberg SM, Rosenberg S: High-dose interleukin-2 for the treatment of metastatic renal cell carcinoma : a retrospective analysis of response and survival in patients treated in the surgery branch at the National Cancer Institute between 1986 and 2006. Cancer 2008, 113:293–301.

8. Rosenberg SA: Raising the bar: the curative potential of human cancer immunotherapy. Sci Transl Med 2012, 4:127ps8.

9. Rosenberg SA, Yang JC, White DE, Steinberg SM: Durability of complete responses in patients with metastatic cancer treated with high-dose interleukin-2: identification of the antigens mediating response. Ann Surg 1998, 228:307–19.

10. Lotze MT, Chang AE, Seipp CA, Simpson C, Vetto JT, Rosenberg SA: High-dose recombinant interleukin 2 in the treatment of patients with disseminated cancer. Responses, treatment-related morbidity, and histologic findings. JAMA 1986, 256:3117–24.

Ethics Approval

Experimental group sizes were approved by the regulatory authorities for animal welfare after being defined to balance statistical power, feasibility and ethical aspects. All mice were kept in accordance with federal and state policies on animal research at BioNTech SE.

Correspondence: Claudia Zylberberg (czylberberg@akronbiotech.com)

Background

Recombinant human interleukin-2 (IL-2) promotes the proliferation of activated T cells and has long been used in vivo to stimulate and maintain the growth of effector T cells and increasingly, to expand T cells in vitro (including genetically engineered CAR-T cells) for adoptive cell therapy in patients with cancer. IL-2 was approved by the FDA for treatment of metastatic renal cancer in 1992 and for treatment of metastatic melanoma in 1998. However, its administration in the approved dose has been associated with serious side effects, including vascular toxicity. There have recently been efforts to address these issues through modifications to the molecule, through the development of fusion proteins, and through the development of low-dose formulations. Akron Biotechnology manufactures IL-2 for in vitro use. The product is embedded in the manufacturing process for several CAR T therapies. There are opportunities to leverage this IL-2 to treat patients suffering from acute and chronic graft versus host disease (GVHD) following allogeneic hematopoietic stem cell transplantation, as well as a range of other autoimmune diseases.

Methods

CyTOF (cytometry by time-of-flight mass spectrometry): antibody panel designed by Dana Farber Cancer Institute. There are three categories of target proteins: surface proteins, intracellular markers and signaling molecules. Different markers were checked after stimulation with Akron IL-2 or Proleukin® (15 minutes) and verified with lymphocyte subset and intrinsic signaling cascade.

Results

The research focused on three major T cell subset populations (Treg, Tcon and CD8 T cells). Upon stimulation with Akron IL-2 and Proleukin®, CD25, CD122, CD132, CTLA-4, PD-1, PD-L1 markers were checked for individual T cell subsets, observed via t-SNE plot. The data indicated that the antibody panel setting was accurate and reliable; each marker was expressed in the expected T cell subset. For example: CD25 was expressed exclusively in Tregs. The expression of pSTAT5, pSTAT3, and pSTAT1 was identical in Akron IL-2 and Proleukin®, controlling for exposure and dosage (Figure 1 includes summary of pSTAT5 results). These results confirm that Akron IL-2 and Proleukin® are identical in their ability to boost adaptive immune response.

Conclusions

The toxicity associated with high dose IL-2 has led many to seek alternatives that can achieve positive therapeutic outcomes while minimizing negative side effects. This study demonstrates that Akron's IL-2 is functionally similar to Proleukin®, validating its utility for the production of adoptive immunotherapies, and potentially for treatment of patients suffering from autoimmune diseases.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P627).

Summary of pSTAT5 Expression after Stimulation

Correspondence: Samir Khleif (snk48@georgetown.edu)

Background

Tumor microenvironment (TME) often harbors epigenetic alterations mediated by histone acetyltransferases (HATs) and histone deacetylases (HDACs)[1, 2]. The epigenetic dysregulation of T-cells and enhanced numbers of immunosuppressive cells in the TME are associated with decreased anti-tumor effects. Hence, targeting the epigenetic modifications using modulators such as HDAC inhibitors (HDACi) provides the basis for a potential role for these agents in cancer immunotherapy. Earlier, we found that HDACi, entinostat reprograms the TME by increasing the number of central memory T-cells and enhancing the functionality of effector cells, resulting in enhanced anti-tumor responses. However, the mechanisms by which HADCi enhances the effector functions of CD8 cells are not well understood.

Methods

In the TC-1 mouse tumor model, entinostat (3mg/kg) was administered with a tumor-specific vaccine (E7-peptide; 3 doses one-week apart). Tumor growth and mice survival were recorded. Three-days after the second immunization, immune-responses were determined in the tumors. Mitochondrial mass of CD8 cells was estimated by incorporation of MitoFM dye, while glucose uptake rates were determined by 2-NBDG incorporation.

Results

We found that entinostat, in presence of optimal CD8 cell priming using tumor specific E7-peptide vaccine, significantly delayed the tumor growth and prolonged the mice survival. Immunologically, entinostat significantly enhanced the numbers of total and antigen-specific CD8+ T cells in the TME. Importantly, we observed a significant increase in the numbers of activated to total and antigen-specific granzyme B+ CD8+ T-cells in the TME when combined with the vaccine. The HDAC inhibition also augmented IFN-gamma production in CD8+ T cells. It is now widely appreciated that T-cell differentiation and effector functions are coupled to metabolic reprogramming. Therefore, using in-vitro assay systems, we checked the status of CD8+ T cells treated with the HDAC inhibitor. We found that HDAC inhibition in CD8 cells during antigen specific activation resulted in significantly increased mitochondrial mass. Activated CD8+ T effector cells are known to rely on glucose metabolism for their energy requirements. Interestingly, the glucose uptake rates of entinostat treated CD8 cells were not different than the non-treated cells indicating that glucose was not utilized differentially in entinostat treated cells. We are further in process of delineating the mechanisms of entinostat mediated enhancement of CD8 cells metabolism.

Conclusions

Overall, our study strongly supports the key role of epigenetic factors in regulating CD8+ T-cell effector and/or memory generation, and metabolic fitness. However, the epigenetic changes that accompany these functional T-cell stages in relation to metabolism remain to be understood.

References

1. Zhang, L. and P. Romero, Metabolic Control of CD8(+) T Cell Fate Decisions and Antitumor Immunity. Trends Mol Med, 2018. 24(1): p. 30-48.

2. McCaw, T.R., et al., Manipulating the epigenetic framework of T cells with histone deacetylase inhibitors for more robust and durable anti-tumor responses. The Journal of Immunology, 2018. 200(1 Supplement): p. 57.41-57.41.

Correspondence: Walter Storkus (storkuswj@upmc.edu)

Background

Tertiary lymphoid structures (TLS) are non-encapsulated immune cell aggregates that form at sites of chronic inflammation, including tumors in some cases. Recent studies have shown that the presence of TLS in human tumors is an indicator of positive clinical outcome. However, due to dysregulated angiogenesis, many tumors have poorly organized and leaky vasculature that impedes entry of immune effector cells into tumors and consequently TLS formation. Recently, pre-clinical studies have shown that low doses (well below maximum tolerate dose) of antiangiogenic agents normalize the tumor vasculature, leading us to hypothesize that treating tumors with low doses of these vascular normalizing (VN) therapies will improve immune cell infiltration and TLS formation within the tumor microenvironment (TME).

Methods

To test this hypothesis, tumor-bearing mice were treated intratumorally with VN agents. Five days post-treatment, tumors were digested into single cell suspensions and RNA was isolated and used for RT-PCR. Transcript levels of TLS-promoting factors (CCL19, CCL21, CXCL13) and markers of vascular normalization (HIF1A, GLUT1) and inflammation/immune cell infiltration (CXCL10, VCAM1, CD8A) were measured and compared to PBS treated controls. Additionally, primary cultures of murine T cells and DCs and the (BRAFV600E PTEN-/-) BPR melanoma cell line were treated in vitro with VN agents. Flow cytometry was then used to evaluate cell phenotype/function, while RT-PCR was used to evaluate transcript levels of TLS-promoting factors.

Results

We observed that the VN agents Dasatinib, Aduro, Bevacizumab, and agonist anti-TNFR1 antibody each induced global changes in the TME that are potentially supportive of immune cell infiltration and TLS formation. These changes include increases in transcript levels of CCL19, CCL21, and VCAM1. In vitro, Dasatinib induced DCs and BPR melanoma cells to express higher levels of co-stimulatory receptors, MHC class I and II, and TLS-promoting chemokines. However, while treatment with Aduro was able to enhance T cell activation, treatment with Dasatinib inhibited T cell activation, as measured by CD25 and CD69 expression.

Conclusions

VN agents induce global changes in immune cell infiltration and TLS-promoting factors in the TME. When these changes are evaluated on specific cell types, some agents were found to increase DC and T cell activation, while others have an inhibitory effect. This knowledge can be used to determine optimal combination immunotherapy designs in the cancer setting.

Correspondence: Christian Steinkuhler (c.steinkuhler@italfarmaco.com), Gianluca Fossati (g.fossati@italfarmaco.com)

Background

The thorough understanding of the immune homeostatic mechanisms has led to the development of target-based immunotherapies. Still, the immunosuppressive tumor microenvironment limits the effectiveness of immunotherapy in several patients. To overcome this obstacle, new targets and combination therapies are the aim of numerous ongoing clinical trials. Histone deacetylase inhibitors have pleiotropic and promising activities as stimulators of anti-tumor immune response alone or in combination with immune checkpoint inhibitors and activators. HDAC6 orchestrates key processes in both innate and adaptive immune cells. HDAC6 null mice are viable and show few phenotypic alterations, indicating that HDAC6 inhibition is potentially safe in humans. We studied the effect of our potent and selective HDAC6 inhibitor, ITF3756, in a model of murine colon carcinoma in vivo and in a model of exhaustion in healthy human T cells in vitro.

Methods

BALB/c mice were injected s.c. with 1x106 CT26 cells and treated with anti-PD1 (i.p) gemcitabine (i.p) ITF3756, (os) or combinations. CD4 and CD8 T cells were depleted using specific antibodies given at the beginning of the experiment. Depletion was checked by flow cytometry. Human CD3+ cells were stimulated three times with anti-CD3/CD28 beads, during 7 days with ITF3756 added at the time of the first two stimulations. During the process, the expression of exhaustion, memory and effector T cells markers were analyzed by flow cytometry.

Results

In vivo, ITF3756 was as effective as an anti PD-1 antibody and their combination additionally reduced tumor growth. Spleen T cells of animals treated with ITF3756 responded ex vivo to immunogenic CT26-derived peptides, suggesting that the antitumor activity of the molecule was immune mediated. Indeed, ITF3756 was completely ineffective in SCID/beige mice bearing CT26 tumor. Selective in vivo depletion of CD4 and CD8 T cells indicated that the activity of ITF3756 strictly depends on both T cell subsets with a greater contribution of CD8 T cells.

In a model of in vitro human T cell exhaustion, ITF3756 countered the exhaustion process as evidenced by the reduced expression of PD-1 and Lag3. Furthermore, ITF3756 increased the CD8 T central memory and reduced the effector memory subsets.

Conclusions

This study shows in vivo and in vitro mechanisms whereby ITF3756 increases the immune response and constitutes a robust basis for the rational use of selective HDAC6 inhibitors either alone or in combination for the effective induction of antitumor immune response.

Ethics Approval

All the animal procedures (including housing, health monitoring, restrain, dosing, etc) and ethical revision were performed according to the current Italian legislation (Legislative Decree March 4th, 2014 n. 26) enforcing the 2010/63/UE Directive on the protection of animals used for biomedical research

Correspondence: Sang-Soo Kim (sangsoo.kim@georgetown.edu

Background

Lung cancer is the world's most common malignancy (>2 million new cases & ~1.7 million deaths annually) [1]. Current lung cancer treatments are woefully inadequate, and immunotherapy has had limited success with but 10-20% of patients with advanced non-small-cell lung cancer (NSCLC) showing durable responses [2]. This has led to a focus on combining immunotherapy with other treatment modalities, and we are exploring combining gene therapy and immunotherapy. scL-RB94, is a cationic liposome encapsulating the RB94 gene and using a single-chain antibody fragment recognizing the transferrin receptor to target tumors (Figure 1). RB94 protein is a potent tumor suppressor [3], and the scL delivery system targets tumors with exquisite specificity [4,5]. In a Phase I trial of scL-RB94 (aka SGT-94), we observed that systemically administered scL-RB94 delivers its payload to small bladder cancer metastases in the lungs resulting in tumor-specific expression of RB94 protein [6]. This ability of scL-RB94 to deliver the RB94 gene to tumor cells in the lungs prompted us to explore use of this investigational agent against NSCLC.

Methods

Because RB94 is active against human but not mouse tumor cells, we have utilized human NSCLC cell lines (H292 & H358) and subcutaneous xenografts of these cells in athymic mice. NSCLC tumor growth was monitored with and without scL-RB94 added to immunotherapy. Expression of markers for innate immunity were also assessed.

Results

Intravenous injection of scL-RB94 markedly inhibited growth of NSCLC xenografts (Figure 2) by inducing immunogenic cell death (ICD) (Figure 3). We observed enhanced tumor infiltration by NK cells (Figure 4) and elevated expression of NK activation ligands. scL-RB94 altered polarization of macrophages toward the M1 phenotype (Figure 5). HLA/A and TAP1/2 gene expression were upregulated (Figure 6), and expression of PD-L1 on the tumor cells was significantly elevated (Figure 7).

Conclusions

scL-RB94 renders NSCLC tumors more immunologically "hot" by triggering ICD and enhancing innate immunity via pleiotropic effects on multiple immune-relevant genes (Figure 8, Box A & B). We are currently exploring the effects of scL-RB94 on adaptive immune responses and in relieving tumor immunosuppression (Figure 8, Box C & D). We hypothesize that combining scL-RB94 with a checkpoint inhibitor will improve response rates compared to immunotherapy alone. With scL-RB94, it may also be possible to reduce the dose of the checkpoint inhibitor to mitigate toxicity. Our longer-term goal is to improve outcomes for NSCLC patients through more effective treatment that combines gene therapy and immunotherapy.

References

1. IARC Global Cancer Observatory web site (http://gco.iarc.fr/).

2. Moya-Horno I, Viteri S, Karachaliou N, Rosell R. Combination of immunotherapy with targeted therapies in advanced non-small cell lung cancer (NSCLC). Therap Advances Med Oncol. 2018;10:1758834017745012. Epub 2018/02/01.

3. Xu HJ, Xu K, Zhou Y, Li J, Benedict WF, Hu SX. Enhanced tumor cell growth suppression by an N-terminal truncated retinoblastoma protein. Proc Natl Acad Sci USA. 1994; 91:9837-9841.

4. Pirollo KF, Dagata J, Wang P, Freedman M, Vladar A, Fricke S, Ileva L, Zhou Q, Chang EH. A tumor-targeted nanodelivery system to improve early MRI detection of cancer. Molecular Imaging. 2006; 5: 41-52.

5. Pirollo KF, Zon G, Rati A, Zhou Q, Yu W, Hogrefe R, and Chang EH. Tumor targeting nanoimmunoliposome complex for short interfering RNA delivery. Human Gene Therapy. 2006; 17:117-124.

6. Siefker-Radtke A, Zhang XQ, Guo CC, Shen Y, Pirollo KF, Sabir S, et al. A Phase l Study of a Tumor-targeted Systemic Nanodelivery System, SGT-94, in Genitourinary Cancers. Molecr Therapy. 2016; 24:1484-1491.

Ethics Approval

All animal studies was performed under an IACUC-approved protocol in compliance with the Animal Welfare Act, PHS Policy, and other Federal statutes and regulations relating to animals in the AAALAC-accredited animal facility at Georgetown University, which adheres to principles stated in the Guide for the Care and Use of Laboratory Animals, National Research Council .

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P631).

See text for description

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2 (abstract P631).

See text for description

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3 (abstract P631).

See text for description

• Download figure
• Open in new tab
• Download powerpoint

Fig. 4 (abstract P631).

See text for description

• Download figure
• Open in new tab
• Download powerpoint

Fig. 5 (abstract P631).

See text for description

• Download figure
• Open in new tab
• Download powerpoint

Fig. 6 (abstract P631).

See text for description

• Download figure
• Open in new tab
• Download powerpoint

Fig. 7 (abstract P631).

See text for description

• Download figure
• Open in new tab
• Download powerpoint

Fig. 8 (abstract P631).

See text for description

Correspondence: Evan Lind (linde@ohsu.edu)

Background

Acute myeloid leukemia (AML) remains difficult to treat due to mutational heterogeneity and the development of resistance to treatment. Targeted agents, such as MEK inhibitors, may be incorporated into treatment, however, the impact of MEK inhibitors on the immune microenvironment in AML is not well understood. A greater understanding of the implications of MEK inhibition on immune responses may lead to greater understanding of immune evasion and more rational combinations with immunotherapies

Methods

AML murine model

Mice expressing FLT3-ITD under the endogenous FLT3 promotor were crossed to mice with the Tet2 gene flanked by LoxP sites and CRE recombinase under control of the LysM promotor. All mouse experiments were performed in accordance with IACUC protocol #IP00000907.

Primary patient samples

Peripheral blood and bone marrow samples were collected from patients with AML. Peripheral blood was collected from healthy donors. All participants gave informed consent to participate (OHSU eIRB# 4422).

Murine proliferation assays

Mononuclear cells were stained with CellTrace CFSE and plated with plate-bound anti-CD3. The mononuclear cells were treated with anti-PD1 and/or trametinib.

Primary patient sample proliferation assays

Mononuclear cells were labeled with CellTrace Violet. Cells were plated with anti-CD3. Cells were treated with combination of anti-PD1 and/or trametinib for 5 days.

Results

In the mouse model, trametinib increases T cell viability and restores T cell proliferation. Importantly, we report greater proliferation in the CD8+CD44+ effector subpopulation and impaired activation of CD8+CD62L+ naïve cells. Transcriptome analysis revealed trametinib sensitive samples have an inflammatory gene expression profile, and we also observe increased PD-L1 expression on trametinib sensitive samples. Finally, we demonstrate trametinib consistently reduces PD-L1 and PD-L2 expression in a dose-dependent manner on the myeloid population.

Conclusions

There are many clinical trials studying immunotherapies in AML. However, immunotherapies may fail as monotherapies in clinical trials if there is strong persisting immunosuppression in the tumor microenvironment or due to a lower mutational burden. Previous work in other cancers have demonstrated benefits of using MEK inhibition in combination with immunotherapies, including checkpoint inhibitors, adoptive T cell therapy, and viral therapy. Subpopulations of AML patients may have improved response to immunotherapies when used in combination with targeted agents. For this reason, we believe our data is useful in beginning to determine immune evasion pathways, such as MEK signaling, in AML in order to target these pathways with the combination of targeted inhibitors and immune based therapies, such as checkpoint inhibitors, bi-specifics, vaccines, or CAR T cell therapies.

Acknowledgements

This work was supported by grant U54CA224019 (E.F.L., S.K.M.) from the National Cancer Institute, grant U41HG003751 (G.W.) from the National Human Genome Research Institute, and by the Oregon Clinical and Translational Research Institute (OCTRI) grant TL1TR002371 (K.B.M.) from the National Center for Advancing Translational Sciences (NCATS) at the National Institutes of Health (NIH). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

Ethics Approval

All mouse experiments were performed in accordance with IACUC protocol #IP00000907. All participants gave informed consent to participate in the study, which is approved by the Institutional Review Board at OHSU (eIRB# 4422).

Correspondence: Rangaprasad Sarangarajan (rangaprasad.sarangarajan@berghealth.com)

Background

Regulation of mitochondrial metabolism is crucial for immune cell differentiation and function. Therapeutic agents influencing mitochondrial metabolism can potentially impact immune-mediated tumor elimination to affect efficacy outcomes. BPM31510 is a Coenzyme Q10-containing lipid nanodispersion in clinical development for treatment of solid tumors. In a phase II study, patients treated with BPM31510 demonstrated signs of pseudo-progression suggesting immune-modulating properties associated with BPM31510 that was further investigated in a preclinical model.

Methods

A MC38 syngeneic mouse model was utilized to characterize tumor infiltrating lymphocytes by flow cytometry. Peripheral blood mononuclear cells (PBMCs) isolated from healthy donor leukopaks were treated ex vivo with increasing concentrations of BPM31510 to assess effects on immune cells and cytokine secretion. Phenotypic characterization of cells was assessed by flow cytometry using cell surface receptor staining CD3/CD8/CD4/PD-1/CTLA-4. Proliferation was assessed via EdU incorporation and cytolytic potential was assessed by intracellular staining for CD107a/b, both in combination with T cell markers. Extracellular cytokines IL-2, IFN-γ, and IL-10 were measured via ELISA.

Results

Consistent with the hypothesis, in a syngeneic mouse model, BPM31510 increased cytotoxic T cell frequency and cytotoxic T cell/regulatory T cell ratio in tumors. Ex vivo, treatment of PBMCs with BPM31510 lead to an increased frequency of viable CD3+ cells. Proliferation measurements by EdU-incorporation indicated enhanced cytotoxic T cell proliferation, and likewise, BPM31510 increased cytolytic potential of activated cytotoxic T cells, as indicated by measurement of plasma membrane-exposed lysosomal-associated membrane protein 1 (CD107a). Treatment of healthy donor PBMCs with BPM31510 decreased cell surface expression of inhibitory receptors PD-1 and CTLA-4 in cytotoxic and helper T cells. To investigate if BPM31510 regulates the cytokine milieu, cytokine secretion profiles of activated PBMCs treated with BPM31510 were established. Analysis of supernatants by multi-analyte ELISA platform revealed effector cytokines IL-2 and IFN-γ to be dose-dependently increased while IL-10, a key immune-regulatory cytokine, was decreased. Analysis of monocyte depleted PBMCs compared to monocytes alone demonstrated that the observed changes in IL-2 and IFN-γ secretion occur only when both lymphocytes and monocytes are present; however, monocytes alone were identified as the significant source for IL-10.

Conclusions

Collectively, the results identify and characterize immune-modulatory activity of BPM31510, particularly in the T cell compartment, specifically on the regulation of T cell exhaustion and immunosuppression in supporting an anti-cancer effect. The data support the utility of BPM31510 in ‘immunologically cold’ tumor types or in combination with immune checkpoint blockade strategies.

Correspondence: Scott Williams (scott.williams@petermac.org)

Background

Localized prostate cancer (PCa) treated with high dose rate brachytherapy (HDRBT) has an excellent cure rate in low, but not in high grade group tumours. Localized PCa immune response to radiation damage is unknown. This study explored this question in order to build a rationale for combining HDRBT with immune checkpoint blockade to treat patients with localized higher grade score PCa and improve patient outcomes.

Methods

In this study we used ultrasound guided core biopsies collected from 24 men prior to, and 14 days post-HDRBT (10Gy). Serial FFPE sections were stained by H&E or OPAL, using two 7-plex panels to describe the density and tissue location of T cells (CD8+, CD4+ and CD4+ FoxP3+), B cells, macrophages, dendritic cells, PDL1+ and HMWCK+ cells. The H&E sections were examined by a uro-pathologist. RNA was extracted for RNAseq (3’ Lexogen) and Nanostring gene expression profiling (GEP) using the pan cancer immune gene set.

Results

HDRBT altered the immune context of PCa. We observed a significant increase post-HDRBT in CD4+FoxP3+ T cells and PDL1- macrophages in both the tumor and tumor stroma (Figure 1). All other immune subsets remained unchanged. All patient samples were grouped into one of three tumor inflammation signature (TIS) [1] clusters (1) high, (2) intermediate and (3) low TIS, Fig 2. Clusters one and two had increased expression of 14/16 and 8/16 TIS genes respectively; cluster three had low expression of all TIS genes. A pre-existing TIS signature was present in 8/24 patients; of the remaining patients 13/16 were converted from a low TIS to intermediate-high TIS gene signature by HDRBT. However, 3/24 PCa tumors did not develop a TIS gene signature post-HDRBT, these tumors had a significantly decreased levels of baseline type I and II IFN response genes, and decreased TNFalpha signalling compared to the 13/16 TIS HDRBT ‘responders’. Finally, we showed that patients with Gleason grade group score (GGG) 4-5 had a pre-existing TIS gene signature level which was not altered by HDRBT (Figure 2); they also exhibited elevated TGFbeta levels pre-HRDBT, that were increased further post-HDRBT.

Conclusions

Our study showed PCa has a heterogeneous immune context including a pre-existing TIS, which is further increased by HDRBT. GGG 4-5 patients expressed high TIS levels post-HDRBT, indicating that anti-PD-1/PDL1 antibodies may be effective in these patients; however TGFbeta remains an additional hurdle for radio-immunotherapy efficacy in these patients.

References

1. Ayers, M., et al., IFN-γ–related mRNA profile predicts clinical response to PD-1 blockade. The Journal of Clinical Investigation, 2017. 127(8): p. 2930-2940.

Ethics Approval

This study was approved by the Peter MacCallum Cancer Centre Human Research Ethics Committee, approval number 14/182

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P634).

Radiotherapy induced immune changes in prostate cancer

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2 (abstract P634).

Radiotherapy induced PCa immune gene expression changes

Correspondence: Victoria Villaflor (victoria.villaflor@nm.org)

Background

Beta-blockers have been associated with anti-tumorigenic effects, potentially by reducing adrenergic-mediated stress responses. Preclinical studies have additionally shown that beta-blockade may enhance the efficacy of cancer immunotherapy. We investigated lung cancer patients who concomitantly used beta-blockers and immune checkpoint inhibitors, with the hypothesis that beta-blockade would positively impact clinical outcomes.

Methods

We retrospectively reviewed the health records of 104 patients who were treated at Northwestern University between January 2014 through August 2018 with immune checkpoint inhibitors for non-small cell lung cancer (NSCLC). Comparisons of overall survival (OS) and progression-free survival (PFS) were performed using Kaplan-Meier analysis with log-rank test, and a multivariate analysis was performed with a Cox regression model.

Results

Among 104 patients treated with immune checkpoint inhibitors for NSCLC, 27 of them were concomitantly prescribed beta-blockers. Use of beta-blockers was associated with increased PFS, with hazard ratio (HR) of 0.57 and 95% confidence interval (CI) of 0.35-0.93. There was not a significant increase in overall survival (OS) among patients who took beta-blockers (HR 0.63, 95% CI 0.35-1.12). In a regression model, beta blockers were identified as predictive of PFS, as were non-squamous histology, tumor PD-L1 positivity, and lower line of treatment. A subgroup analysis of patients with brain metastases (n=37) suggested greater overall survival among patients taking beta blockers (HR 0.17, 95% CI 0.06-0.54).

Conclusions

Beta-blocker use of at least 1 year was associated with improved PFS among patients treated with immune checkpoint inhibitors for NSCLC in our cohort of patients. This was a small study and these findings should be further validated in prospective clinical studies.

Correspondence: Timothy Thompson (timthomp@mdanderson.org)

Background

In prostate cancer (PCa), androgen receptor (AR) signaling regulates the cell cycle and mitigates replication stress via upregulation of CDC6 and the TopB1-ATR-Chk1 pathway, the key axis in replication stress response (RSR)[1]. Preclinical studies and early phase clinical trials of DDR inhibitors, including PARP inhibitors, have shown synergy with immune checkpoint therapy, whereby DDR inhibitors lead to S-phase specific damage, accumulation of cytosolic DNA and activation of innate immune pathways[2]. DDR inhibitors have rapidly expanded, now including inhibitors of ATR[2]. However, the impact of these novel, potent, highly-specific inhibitors of ATR on immunocompetent PCa preclinical models has not been fully investigated. Furthermore, dynamic changes in DDR and innate immune gene expression profiles that result from various targeted therapies have not been reported. In this study we use preclinical, immunocompetent models of advanced PCa to investigate the efficacy, transcriptional targets, induction of immunostimulatory cytosolic DNA, and synergy with anti-PD-L1 therapy associated with ATR inhibition.

Methods

We used PCa cell lines derived from ras+myc—induced mouse PCa tumors (RM-1, RM-9 and RM-1-BM) to analyze ATRi (BAY1895344 or VX-970) effects on ATR-driven survival/proliferation, DNA damage, cGAS-STING signaling, and chemokine expression in comparison to olaparib. RM-1-BM xenograft model was used to assess the efficacy and safety of single agent ATR inhibitor, single-agent anti-PD-L1, and the combination in vivo.

Results

Treatment of the PCa cell lines with ATRis (BAY1895344 or VX-970) or a PARP inhibitor (olaparib) demonstrated dose-dependent (0.125-8μM) growth suppression and cytosolic DNA accumulation, while ATRis showed greater growth suppression than olaparib. ATRi and olaparib treatment generated similar levels of DNA damage (γH2AX); however, ATRi treatments resulted in significantly increased CCL5 and CXCL10 levels compared to olaparib (2-10 fold, P< 0.05), with BAY1895344 demonstrating superior induction of chemokine expression (P< 0.05). cGAS, STING, TBK1 or IRF3 siRNA knockdown significantly suppressed all of these responses (P< 0.05). The combination of BAY1895344 with anti-PD-L1 was safe and resulted in significantly greater tumor response than either single agent in RM-1-BM xenograft models.

Conclusions

ATR inhibition induces S-phase specific DNA damage, accumulation of cytosolic DNA, activation of downstream cGAS-STING signaling, and suppression of cancer growth in immunocompetent PCa models. ATRi had greater impact on innate immune pathway activation than olaparib, and the combination of ATRi with anti-PD-L1 was safe and resulted in significant efficacy in our model system. These results will directly inform biomarker-directed clinical trials of ATR inhibitors in combination with immune checkpoint blockade for patients with advanced solid cancers.

Acknowledgements

We would like to acknowledge members of the Bayer-MD Anderson Cancer Center Alliance, including but not limited to Dr. Antje Wengner, Dr. Dominik Mumberg.

References

1. Karanika S, Karantanos T, Li L, et al. Targeting DNA Damage Response in Prostate Cancer by Inhibiting Androgen Receptor-CDC6-ATR-Chk1 Signaling. Cell Rep. 2017; 18(8):1970-1981.

2. Pilie PG, Tang C, Mills GB, Yap TA. State-of-the-art strategies for targeting the DNA damage response in cancer.. Nat Rev Clin Oncol. 2019; 16(2):81-104.

Ethics Approval

The study, including animal work, was approved by MD Anderson Cancer Center's board, approval number #885 and #889.

Correspondence: Charles Drake (cgd2139@cumc.columbia.edu)

Background

Radiation-induced immunostimulation is of great interest as a strategy to reinvigorate or augment an immunotherapy-induced tumor-specific response. Evidence supports photon therapy (PhRT) as a tumor immunomodulator, pre-clinical data suggest that carbon-ion therapy (CiRT) may be more potent. Here we evaluate the immunomodulatory effects of CiRT, compared to biologically equivalent doses (BEDs) of PhRT in an orthotopic breast cancer mouse model.

Methods

Using the orthotopic 4T1 breast cancer mouse model, we evaluated the immunomodulatory effects of very low (0.4 Gy), low (1 Gy) and high (4 Gy) doses of CiRT (290 MeV/u, SOBP) delivered to mammary tumors at the Heavy Ion Medical Accelerator in Chiba (HIMAC), Japan compared to BEDs (RBE 2) of PhRT (SARRP, Xstrahl; 220 kV). Primary tumors were treated 7-days post implantation and harvested 72-hours later for flow cytometric characterization and protein cytokine analysis.

Results

At low doses, CiRT was lymphocyte sparing effect compared to PhRT. Only low and high-dose CiRT resulted in a significant (p

Conclusions

We demonstrate that at low doses, lymphocytes differ in their sensitivity to CiRT compared to PhRT -- CiRT being more lymphocyte-sparing. Very high-dose CiRT induced a more potent and mechanistically diverse of pro-inflammatory cytokine response, compared to PhRT. Combining less lymphotoxicity at low dose with more potent pro-inflammatory effects at high doses, CiRT may be a more optimal synergistic tool to augment an immunotherapy-induced, tumor-specific response.

Correspondence: Darren Woodside (dwoodside@texasheart.org)

Background

The use of immune stimulatory agents as systemic drugs to augment adjuvantation can be problematic due to elevated risks of toxicity and non-specific immune responses. Integrins αLβ2/ICAM-1 and α4β1/VCAM-1 are essential for antigen-specific immune responses. [1,2] In particular, these integrins stabilize receptor-mediated cell adhesion between Antigen Presenting Cells and naïve CD4+ T cells, providing a co-stimulatory signal required for effective antigen presentation.[3] 7HP349 is a first-in-concept, orally bioavailable, positive allosteric activator of these integrins.

Methods

To determine if systemic administration of 7HP349 in combination with known vaccines would increase adjuvantation, 7HP349 was co-administered with T. cruzi TcG2 and TcG4 vaccines against Chagas disease, inactivated vaccine against Influenza, and BCG vaccines against M. tuberculosis (Mtb) in mice. Moreover, 7HP349 as monotherapy in an immunogenic murine CT26 carcinoma model, and in combination with aCTLA-4/GVAX in a non-immunogenic murine B16 melanoma model.

Results

In the CT26 model, 7HP349 monotherapy significantly delayed tumor growth (Table 1). In the B16 tumor model, 7HP349 in combination with GVAX/aCTLA4 significantly increased long-term (>120 day) survival by 2.7-fold over GVAX/aCTLA4. In combination with T. cruzi TcG2/TcG4 vaccines, 7HP349 significantly decreased parasite burden in myocardial (~9.5-fold) and skeletal muscle (~5.3-fold) in challenged mice and increased splenic frequency of polyfunctional CD4+ and CD8+ T cells expressing IFN-gamma, perforin or granzyme B. In addition, in vitro stimulation with Tc antigens of splenocytes from vaccinated animals resulted in maximal release of TNFα, IFN-γ, IL-1β and IL-6 cytokines when 7HP349 was co-administered with vaccine. 7HP349 increased influenza H1N1 seroconversion and hemagglutinin inhibition titers by ~30 fold when co-administered with inactivated influenza vaccine. 7HP349 addition to a prime-boost immunization regimen with BCG reduced pulmonary mycobacterial burden and inflammation after Mtb challenge compared to vaccine alone. At the 0.5 and 1 mg/kg doses, the therapeutic plasma 7HP349 concentration range of AUC of 50-100 ng.h/mL, appears to be sufficient to engage αLβ2/ICAM-1 and α4β1/VCAM-1 and increase tumor immunity, checkpoint blockade, and T cruzi, BCG and H1N1 vaccination. 7HP349 exposures of >1,500 ng.h/mL for 28 days in rats and dogs in cGLP toxicology studies showed no adverse findings

Conclusions

7HP349 as systemic drug may facilitate immune priming against a variety of antigens by stabilizing αLβ2/ICAM-1 and α4β1/VCAM-1 interactions, with a high safety margin. 7HP349 oral drug product is IND ready. Phase I studies as monotherapy and combined with vaccine for influenza, or with immune checkpoint inhibitors in refractory solid tumors are planned.

References

1. Davignon D, Martz E, Reynolds T, Kürzinger K and Springer TA. Monoclonal antibody to a novel lymphocyte function-associated antigen (LFA-1): mechanism of blockade of T lymphocyte-mediated killing and effects on other T and B lymphocyte functions. The Journal of Immunology. 1981;127:590-595.

2. Shimizu Y, van Seventer GA, Horgan KJ and Shaw S. Costimulation of proliferative responses of resting CD4+ T cells by the interaction of VLA-4 and VLA-5 with fibronectin or VLA-6 with laminin. The Journal of Immunology. 1990;145:59-67.

3. Woodside DG, Teague TK and McIntyre BW. Specific inhibition of T lymphocyte coactivation by triggering integrin beta 1 reveals convergence of beta 1, beta 2, and beta 7 signaling pathways. The Journal of Immunology. 1996;157:700-706.

Correspondence: Paul Sondel (pmsondel@humonc.wisc.edu)

Background

An important component of research is ensuring rigor and reproducibility of work using animal models. We have identified a variable influencing the response of a syngeneic B78 melanoma flank tumor model to treatment with an in situ vaccine (ISV) regimen [1]. This study was prompted after two experimenters performing virtually identical studies were noted to obtain different results. Upon further evaluation, we observed different efficacy of ISV against B78 tumors when the tumors were implanted at different depths. Though the baseline in vivo immunogenicity of tumors can depend on depth of implantation, response to ISV as a function of tumor location, particularly in immunologically “cold” tumors, has not been thoroughly investigated. The goal of this study was to evaluate the difference in growth kinetics and response to ISV between identically sized melanoma tumors implanted at different depths in the skin.

Methods

We injected C57BL/6 mice with syngeneic B78 melanoma cells in the flank region. Half these mice were injected intradermally (ID), and half were injected subcutaneously (SQ). When tumors reached 190-230mm3, they were grouped into a ‘wave’ and treated according to our previously published in situ vaccine regimen: 12Gy local external beam radiation and intratumoral injections of hu14.18-IL2 immunocytokine [1]. Some mice from each wave were randomly assigned to be ‘untreated.’ Tumor volume and survival were monitored twice weekly for the duration of the experiment.

Results

Physical examination demonstrated that ID-implanted tumors were mobile upon palpation, while SQ-implanted tumors became fixed to the underlying fascia. Histologic examination identified a critical fascial layer, the panniculus carnosus, which separated ID and SQ tumors. SQ tumors reached the target tumor volume significantly faster compared to ID tumors. The majority of ID tumors exhibited either partial response or complete response to ISV, whereas the majority of SQ tumors did not. Direct comparison of ID vs. SQ tumors that reached target volume simultaneously further emphasized this difference in response to treatment. Immunohistochemistry evaluation of separately grown tumors demonstrated greater immune infiltrate following ISV in ID compared to SQ tumors.

Conclusions

We demonstrate that the physical fixed vs. mobile characterization of the tumors may be one method of ensuring implanted tumors are in the same tissue plane prior to initiation of treatment. Moreover, after controlling for strain of mouse, type, size, and growth rate of tumor, we demonstrate that depth of tumor implantation still has a dramatic effect on response to ISV.

References

1. Morris Z, Guy E, Werner L, Carlson P, Heinze C, Kler J, Busche S, Jaquish A, Sriramaneni R, Carmichael L, Loibner H, Gillies S, Korman A, Erbe A, Hank J, Rakhmilevich A, Harari P, Sondel P. Tumor-Specific Inhibition of In Situ Vaccination by Distant Untreated Tumor Sites. Cancer Immunol Res. 2018 Jul;6(7):825-834.

Correspondence: Regina Heidenreich (regina.heidenreich@curevac.com)

Background

Intratumoral activation of innate immune signaling pathways is a promising approach to overcome the immunosuppressive tumor microenvironment and induce or restore anti-tumor immunity. CV8102, a non-coding RNA complexed with a cationic peptide that signals via TLR-7/-8 and RIG-I [1], has already been shown to exhibit immunomodulatory properties after intradermal injection [2]. Here, we investigated whether intratumoral injection of CV8102 alone or in combination with systemic anti-PD-1 treatment can also modulate the tumor microenvironment and enhance anti-tumoral responses.

Methods

Mice were injected subcutaneously with the syngeneic colon carcinoma cell line CT26 or the B cell lymphoma cell line A20 and established tumors were treated intratumorally with CV8102 alone or in combination with systemic anti-PD-1 antibodies. Anti-tumoral efficacy was determined in tumor bearing mice, which were treated twice weekly for three weeks. Additionally, cytokine concentrations, immune cell composition or gene expression within the tumor microenvironment, the draining lymph nodes or serum were analyzed at different time points post injection.

Results

Intratumoral administration of CV8102 led to a dose-dependent anti-tumoral response in syngeneic mouse models. Combination of local CV8102 delivery with systemic checkpoint inhibitors (CPIs) further improved the anti-tumoral response and enabled checkpoint inhibition efficacy in anti-PD-1 refractory tumor models. Gene expression profiling revealed activation of the innate immune compartment within the tumor microenvironment, reflected by up-regulation of genes involved in Type I interferon and anti-viral responses as well as RIG-I like, NOD-like and TNF signaling pathways. Accordingly, cytokine/chemokine analysis showed increased intratumoral concentrations of IFN-α/β, IL-6, TNF, MCP1, MIP1α/β and RANTES. Such effects were observed upon CV8102 single agent treatment and were more pronounced upon combination with systemic anti-PD-1 treatment. On the cellular level, intratumoral CV8102 monotherapy caused an increase in numbers and activation of innate immune cells like monocytes, neutrophils and NK cells within tumors and draining lymph nodes. Besides further enhancement of innate immunity, combination therapy synergistically enabled activation of the adaptive immune system and led to the infiltration of CD8+ T cells into the tumor microenvironment.

Conclusions

Our results demonstrate that intratumoral CV8102 delivery is a promising approach for local cancer immunotherapy. By modulating the tumor microenvironment CV8102 induces potent anti-tumoral immune responses and improves CPI efficacy in mice. An ongoing phase I trial is investigating intratumoral CV8102 either as a monotherapy or in combination with systemic anti-PD-1 treatment in various tumor entities (see abstract by Eigentler et al.).

References

1. Ziegler A, Soldner C, Lienenklaus S, Spanier J, Trittel S, Riese P, Kramps T, Weiss S, Heidenreich R, Jasny E, Guzmán CA, Kallen KJ, Fotin-Mleczek M, Kalinke U. A New RNA-Based Adjuvant Enhances Virus-Specific Vaccine Responses by Locally Triggering TLR- and RLH-Dependent Effects. J Immunol. 2017;198(4):1595-1605. doi: 10.4049/jimmunol.1601129

2. Heidenreich R, Jasny E, Kowalczyk A, Lutz J, Probst J, Baumhof P, Scheel B, Voss S, Kallen KJ, Fotin-Mleczek M. Int J Cancer. 2015 Jul 15;137(2):372-84. doi: 10.1002/ijc.29402

Ethics Approval

The study was approved by the Regierungspraesidium Tuebingen, Germany under Cur 07/18.

Correspondence: John Schiller (schillej@mail.nih.gov)

Background

The discovery that the tumor immune microenvironment can be used as biomarker of cancer progression and the advent of immunotherapies are revolutionizing cancer treatments. Notwithstanding these successes, there is still an urgent need for new therapeutics able to increase the proportion of clinical responses. Intratumoral delivery of immunotherapeutic agents constitutes a potential strategy to convert poorly immunogenic tumors into immunogenic tumors and to overcome acquired resistance to current immunotherapies.

Methods

Here we investigate a tumor antigen-agnostic intratumoral approach that harnesses preexisting antiviral T cell responses. To implement such approach, Human cytomegalovirus (hCMV) would be an ideal candidate as it is highly prevalent in human populations and induces polyfunctional T-cell responses that expand with age. In our experimental model, mice persistently infected with murine cytomegalovirus (mCMV) were challenged with TC-1 tumor cells expressing human papillomavirus E6 and E7 oncogenes or MC38 colon adenocarcinoma tumor cells. We investigated the effects of redirecting pre-existing mCMV-specific T cells by repeated intratumoral injection using minimal MHC class I and/or class II restricted mCMV peptide epitopes combined with polyI:C, a Toll-like receptor 3 and RIG-I-like receptor agonist.

Results

Intratumoral injection of mCMV MHC-I and MHC-II restricted peptides led to the expansion of CD8 and CD4 mCMV-specific T cells in situ and in blood. Intratumoral injection of mCMV MHC-I restricted peptides provoked arrest of TC-1 and MC38 tumor growth that sometimes resulted in complete tumor rejection. Interestingly, intratumoral injection of mCMV MHC-II restricted peptides with poly(I:C), without MHC-I peptides, also substantially delayed tumor growth. In addition, injection of mCMV MHC-II restricted peptides alone or in combination with mCMV MHC-I restricted peptides increased epitope spreading to MHC-I restricted tumor-derived epitopes, promoted clearance of established subcutaneous tumors, and conferred long-term protection against tumor re-challenge. Finally, analysis of the tumor microenvironment using whole tumor RNA gene expression profiling, flow cytometry and confocal microscopy revealed that intratumoral injection of mCMV MHC-I or MHC-II restricted epitopes caused local activation of adaptive and innate immune components.

Conclusions

Our results provide a proof-of-concept for a new generation of tumor antigen-agnostic immunotherapies based on pre-existing antiviral T cells that is potentially applicable to multiple tumor types. The approach induced rapid tumor regression, profound changes in the tumor immune microenvironment, epitope spreading, and long-term antitumor immunity. We are now using our model to decipher the contribution of CD4+ T cell to epitope spreading against tumor-derived epitopes and to evaluate this approach in spontaneous tumor models.

References

1. Aznar MA, Tinari N, Rullán AJ, Sánchez-Paulete AR, Rodriguez-Ruiz ME, Melero I. Intratumoral Delivery of Immunotherapy-Act Locally, Think Globally. J Immunol.2017 Jan 1;198(1):31-39.

2. Frank MJ, Reagan PM, Bartlett NL, Gordon LI, Friedberg JW, Czerwinski DK, Long SR, Hoppe RT, Janssen R, Candia AF, Coffman RL, Levy R. In Situ Vaccination with a TLR9 Agonist and Local Low-Dose Radiation Induces Systemic Responses in Untreated Indolent Lymphoma. Cancer Discov. 2018 Oct;8(10):1258-1269.

3. Ribas A, Medina T, Kummar S, Amin A, Kalbasi A, Drabick JJ, Barve M, Daniels GA, Wong DJ, Schmidt EV, Candia AF, Coffman RL, Leung ACF, Janssen RS. SD-101 in Combination with Pembrolizumab in Advanced Melanoma: Results of a Phase Ib, Multicenter Study. Cancer Discov. 2018 Oct;8(10):1250-1257.

Correspondence: Karl Wittrup (wittrup@mit.edu)

Background

Agonist antibodies that target co-stimulatory receptors, including targets such as CD28, CD40, CD137, and OX40, have demonstrated impressive efficacy in preclinical cancer models. Broadly, drugging these receptors leads to immune cell proliferation, cytokine production, and enhanced effector function. In the clinic, however, the development of these antibody drugs has been hampered by dose-limiting toxicities which have limited their utility and prevented approval of any co-stimulatory receptor targeting agonist antibodies. Novel strategies are required to dose these drugs in a safe manner while still retaining their anti-tumor activity.

Methods

To contain the activity of agonist antibodies to the tumor microenvironment (TME), we have fused a collagen binding domain (CBD) to the Fc region of several agonist antibodies targeting a suite of co-stimulatory receptors. We have recently validated this collagen anchoring approach to improve the therapeutic index of cytokine therapies [1]. Fluorescent imaging was used to characterize intratumoral retention and the biodistribution of intratumorally injected antibodies. Survival studies were carried out to determine the efficacy of antibodies with and without CBDs in murine subcutaneous tumor models. We measured toxicity by quantifying serum protein levels as well as with histology in treated mice.

Results

Imaging studies revealed that antibody-CBD fusions were retained significantly longer in the TME compared to their non-CBD counterparts and experienced little to no detectable systemic dissemination. Intratumorally administered antibody-CBD fusions displayed comparable, or in some cases, improved efficacy when benchmarked against free antibody. These fusion proteins also caused markedly fewer signs of toxicity in murine models.

Conclusions

Intratumorally administered antibody-CBD fusions represent a new class of therapeutics with improved therapeutic indices compared to agonist antibodies that have already entered the clinic. Because intratumoral retention is dependent on collagen and not a tumor specific target, these drugs are tumor-agnostic and have the potential to impact large patient populations.

References

1. Momin et al., Anchoring of intratumorally administered cytokines to collagen safely potentiates systemic cancer immunotherapy. Sci Transl Med. 2019; 11; eaaw2614

Correspondence: Wilson Meng (meng@duq.edu)

Background

Intra-tumoral immunotherapy has emerged as an important modality of delivering anti-tumor immunomodulators. Rich in tumor-associated antigens and antigen-presenting cells, the tumor microenvironment is an apt locale for reprogramming the immune responses from a suppressive phenotype to one that is activating insofar anti-tumor cytotoxic T cell responses can be generated locally and systemically.

Herein we report a small peptide that can potentially serve as an adjuvant for enhancing antigen presentation in the tumor microenvironment. Our group discovered the peptide YTYQGKL, called pTNF, as a ligand of tumor necrosis factor receptor II (TNFRII) by screening phage display libraries on dendritic cells (DCs) [1]. The peptide binds to DCs in a concentration-dependant, TNFalpha-reversible manner, albeit at micromolar binding affinity. Subsequently, other groups have confirmed the DC-targeting property of pTNF[2,3]. In addition, it has been shown that pTNF upregulates costimulatory molecules on DCs [4,5]. This led us to hypothesize that pTNF can be used as an immune adjuvant in the context of cancer immunotherapy. To transform the low affinity ligand pTNF into a high avidity adjuvant, the peptide is appended with [AEAEAKAK]2, a self-assembling peptide (SAP) that drives oligomerization (Fig. 1a).

Methods

To test the capacity of pTNF_EAK in activating antigen presenting cells, BALB/c mice were injected with 100 mcg of the peptide in each footpad. After 48 hours, the draining lymph nodes were collected and analyzed using flow cytometry. The cells obtained from draining lymph nodes were restimulated ex vivo with a mouse renal adenocarcinoma (RENCA) cell line for 48hours. The cells were analyzed using flow cytometry and ELISA, quantifying interferon-gamma (IFN gamma) production upon in vitro restimulation.

Results

Preliminary results indicated that pTNF_EAK exhibited self-assembling characteristics based on Congo red staining (data not shown). Flow cytometric analyses showed upregulated CD40 on cells in pTNF_EAK treated draining lymph nodes compared to control footpads injected with saline (Fig. 1b). Upon restimulation in vitro with RENCA cells, pTNF_EAK treated lymph node-derived cells produced higher levels of (IFN gamma) compared to control cells (data not shown).

Conclusions

pTNF_EAK appears to upregulate the functions of antigen-presenting cells in draining lymph nodes in vivo.

References

1. Chamarthy SP, Jia L, Kovacs JR, Anderson KR, Shen H, Firestine SM, Meng WS. Gene delivery to dendritic cells facilitated by a tumor necrosis factor alpha-competing peptide. Molecular immunology. 2004 Jul 1;41(8):741-9

2. Margaroni M, Agallou M, Kontonikola K, Karidi K, Kammona O, Kiparissides C, Gaitanaki C, Karagouni E. PLGA nanoparticles modified with a TNFα mimicking peptide, soluble Leishmania antigens and MPLA induce T cell priming in vitro via dendritic cell functional differentiation. European Journal of Pharmaceutics and Biopharmaceutics. 2016 Aug 1;105:18-31.

3. Herringson TP, Altin JG. Convenient targeting of stealth siRNA-lipoplexes to cells with chelator lipid-anchored molecules. Journal of Controlled Release. 2009 Nov 3;139(3):229-38.

4. Caux C, Dezutter-Dambuyant C, Schmitt D, Banchereau J. GM-CSF and TNF-α cooperate in the generation of dendritic Langerhans cells. Nature. 1992 Nov;360(6401):258.

5. Caux C, Massacrier C, Vanbervliet B, Dubois B, Van Kooten C, Durand I, Banchereau J. Activation of human dendritic cells through CD40 cross-linking. Journal of Experimental Medicine. 1994 Oct 1;180(4):1263-72.

Ethics Approval

The study was approved by Duquesne University's IACUC, protocol number: 1806-04

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P645).

See text for description

Correspondence: Samir Khleif (snk48@georgetown.edu)

Background

Exhaustion and short persistence are factors that limit the efficacy of endogenously generated or adoptively transferred effector T-cells. Metabolism in T-cells has been found to be a significant factor that contributes to their anti-cancer potential. Interestingly, tumor-induced metabolic insufficiency in immune cells renders these cells exhausted and hence ineffective in controlling the tumors. Hence the mechanisms that would prevent the metabolic degradation in immune cells would have strong anti-cancer potential. Renewable stem-cell-memory CD8+T-cells (Tscm) are metabolically fit cells that persist longer and produce stronger effector functions. Here, we determined the role of MAP kinase pathway in the generation of Tscm CD8 cells and their metabolic status.

Methods

Mice bearing TC-1 tumors (a mouse lung epithelial cell-line expressing human papillomavirus-specific E7-peptide) were treated with MEKi (selumetinib) in conjunction with tumor-specific E7-peptide vaccine, followed by determination of tumor growth rates and mice survival, and estimation of immune responses in the TME. Also, using in-vitro and in-vivo assay systems, we delineated the kinetics of MEKi-mediated Tscm generation and the molecular mechanisms of metabolic enhancement of CD8 cells.

Results

We found that MEKi treatment of tumor-bearing mice in conjugation with tumor-specific E7-vaccine resulted in a significant reduction in tumor growth and enhancement of mice survival. This treatment resulted in a strong increase in the effector functions and prevention of exhaustion in effector cells. We observed a significant increase in the number of Sca1+CD95+CD62L+CD44-CD8+ T-cells (Tscm cells) in the TME and under in-vitro conditions. Interestingly, we found that MEKi suppressed the cell-cycle progression in antigen-activated CD8 cells that led to increased mitochondrial biogenesis and fatty-acid oxidation (FAO) by modulating ERK1/2/CyclinD1/PGC1α/SIRT3 signaling pathway. FAO was crucial for MEKi-mediated induction of Tscm cells as its pharmacological inhibition or SIRT3KO prevented the Tscm generation. Further, ex-vivo MEKi-treated CD8+T-cells showed strong cell-activation, high antigen recall-responses and prolonged-survival, generating robust T-cells that significantly enhanced the efficacy of adoptive cell therapy (ACT) that was critically dependent upon MEKi-induced Tscm cells. Together, we show that MEKi-induced Tscm generation occurs as a result of a crosstalk between cell activation, cell-cycle progression and differentiation, and metabolic machinery.

Conclusions

Our data provide a novel strategy to reprogram CD8 cells into a self-regenerating stem cell phenotype with higher metabolic fitness and superior effector functions. This data also provide an effective approach to produce superior T-cells for in-vitro treatments such as ACT. Furthermore, the findings from this study provide a strategy for enhancement of treatment approaches using inhibitory molecules.

Correspondence: Howard Kaufman (Howard.Kaufman@replimune.com)

Background

Talimogene laherparepvec (T-VEC) is an oncolytic herpes simplex virus, type 1 (HSV-1) encoding granulocyte-macrophage colony stimulating factor (GM-CSF) and is approved for the treatment of melanoma. To date, there are no predictive biomarkers to identify patients likely to respond to T-VEC treatment, which is a high priority. Thus, we sought to identify intracellular factors associated with host anti-tumor immunity following T-VEC treatment.

Methods

Human melanoma cell lines were plated in 96-well plates (104 cells per well) and treated with T-VEC (MOI 0.001-1.0). At various time points, cell viability was assessed by standard MTS assay. To evaluate specific intracellular factors, knock out cell lines were generated by using CRISPR-Cas9 technology and assessed for oncolytic activity in vitro. For in vivo experiments, B6 mice bearing the HSV-1-sensitive D4M3A (3 x 105) melanoma bilateral flank tumor model was used. Mice were treated by injection in the right flank only with a T-VEC (6 x 106 PFU) once tumors reached 9-12 mm2 three times a week for two weeks. Tumor growth was measured by calipers daily for tumor growth and survival studies. In separate experiments, tumors were harvested at specified timepoints and flow cytometry analysis was performed to characterize immune cell populations and RNA was extracted for Nanostring gene expression analysis. Statistical comparisons between treatment groups were determined using the Student’s t test with and the Kaplan-Meier method was used to estimate survival. P

Results

We initially interrogated multiple human melanoma cell lines and found differential sensitivity to T-VEC-mediated oncolysis. Characterization of the cell lines suggested that STING expression, but not PKR or cGAS, was inversely associated with T-VEC-mediated cell lysis. This was confirmed using CRSPR-Cas9-generated STING knockout cells which reversed T-VEC resistance. Induction of host immunity was confirmed in vivo using a STING deficient melanoma tumor model, which is resistant to PD-1 blockade, and demonstrated significant delay in tumor growth and improved survival following treatment with T-VEC. Therapeutic responses were associated with recruitment of viral- and tumor antigen-specific CD8+ T cells and induction of a pro-inflammatory gene signature in both injected and un-injected tumors.

Conclusions

Intracellular STING was inversely associated with T-VEC-mediated lysis and suggests that STING may be a predictive biomarker for T-VEC responses. These data also support a role for T-VEC in restoring host immune responses in STING-deficient melanoma.

Ethics Approval

This study was approved by the Rutgers University IACUC.

Correspondence: Greg Delgoffe (delgoffeg@upmc.edu)

Background

Oncolytic virotherapy is an attractive immunotherapeutic approach as it can induce robust immune infiltration in immunologically cold tumors. While there is one OV approved for use by the FDA and many others in clinical trials, little is understood of their immunologic mechanism of action. The effect of OV treatment on the existing immune infiltrate in the tumor is especially understudied. Seven days after one intratumoral treatment of oncolytic vaccinia (vv) in a murine head and neck squamous cell carcinoma model we observed an influx of CD8 T cells into the tumor, including a large population of effector-memory like PD-1int cells. However, the suppressive PD-1hi Tim-3+ exhausted T cells (Texh) and Foxp3+ regulatory T cells (Treg) were missing. We hypothesized that vvDD may have the capacity to infect these inhibitory populations and lead to their selective deletion.

Methods

Infection and flow cytometric analyses of T cell populations were performed using both in vitro and in vivo modeling systems. An oncolytic Western Reserve strain of vaccinia virus (vv) was used at a multiplicity of infection of 10 for all in vitro experiments and at 2.5E6 PFU per mouse for in vivo experiments. T-reg specific BCL2 over-expressing mice were generated by crossing Rosa26-BCL2-IRES-GFP mice to Foxp3-YFP-cre mice.

Results

Analysis of the immune infiltrate of treated tumor-bearing mice revealed that Texh and Treg in the tumor were actively infected by virus. This may lead directly to the death of these infected T cells. This phenomenon was recapitulated in vitro under nutrient limited conditions. This was further investigated in vivo using a mouse model where the anti-apoptotic protein BCL2 was selectively overexpressed in Tregs. In this model the death of Tregs is prevented in the tumor. These mice were unable to clear tumor after vv treatment while all wild-type animals did.

Conclusions

These data suggest that the mechanism of action of oncolytic vaccinia also includes the direct ability of the virus to infect immune populations in the tumor. Infection of regulatory T cells in the tumor leads to their death. By preventing the death of these cells, treatment efficacy is lost, thus highlighting the importance of this mechanism during oncolytic vaccinia treatment.

Correspondence: Tae-Ho Hwang (thhwang@pusan.ac.kr)

Background

Oncolytic virus (OV) is an emerging class of cancer treatment which has received increasing attention. The pharmacodynamics (safety and efficacy) of OV shows large inter-individual variability―the clinical outcomes of OV treatments can range from complete response to possibly treatment-related early death in different patients [1,2]. The innate immune system as well as adaptive immune system play critical roles in governing the OV treatment outcomes. This study aimed to investigate the therapeutic potential for the interaction between circulating myeloid cells, oncolytic vaccinia virus (OVV), and cancer cells; and to optimize the treatment outcomes of OTS-400 (OVV series with modified HSV-tk insertion).

Methods

The functional outcome of interaction between circulating immune cells, OVV, and cancer cells was evaluated quantitatively and qualitatively using a trans-well co-culture system and analyzed by fluorescence-activated cell sorting. After identification of subtypes of OV-disfavoring myeloid cells (OV-DMCs), candidate chemicals, including clinically used drugs, were screened to find the best modulators for OV-DMCs affecting clinical outcome of OTS-400 series in >3,000 tumor-bearing animals.

Results

Time to 90% lethality after intranasal administration of Western Reserve (107 pfu) vaccinia virus was significantly prolonged by inhibition of OV-DMCs, 30 days+; by stimulation, 7 days; and by no modulation, 11 days (Log-rank p-value

Conclusions

OV-DMCs play a critical role in the pharmacodynamics of OVV therapy in a variety of preclinical tumor models. Targeting OV-DMCs may significantly improve the treatment outcome for the upcoming clinical trials of the combination therapy of OTS-400 and PD-L1 inhibitor scheduled in early 2020.

Acknowledgements

This study was supported by a grant from the National R&D Program for Cancer Control, Ministry of Health and Welfare, Republic of Korea (HA16C0013) and National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIP) (NRF-2015R1A5A2009656).

References

1.Park BH, Hwang T, Liu TC, Sze DY, Kim JS, Kwon HC, Oh SY, Han SY, Yoon JH, Hong SH et al: Use of a targeted oncolytic poxvirus, JX-594, in patients with refractory primary or metastatic liver cancer: a phase I trial. Lancet Oncol 2008, 9(6):533-542.

2.Breitbach CJ, Burke J, Jonker D, Stephenson J, Haas AR, Chow LQ, Nieva J, Hwang TH, Moon A, Patt R et al: Intravenous delivery of a multi-mechanistic cancer-targeted oncolytic poxvirus in humans. Nature 2011, 477(7362):99-102.

Correspondence: Juan Fueyo (jfueyo@mdanderson.org)

Background

Immune checkpoint blockade has revolutionized cancer therapy; however the therapeutic benefit is limited to only a subset of patients with immunogenic (“hot”) tumors and is compromised by immune-related adverse events. We have reported that intratumoral injection of oncolytic adenovirus Delta-24-RGDOX, an oncolytic adenovirus expressing immune co-stimulator OX40 ligand (OX40L), induces efficacious anti-glioma immunity and immune memory in syngeneic glioma mouse models. We hypothesized that localized treatment with the virus is effective against disseminated melanomas, including intracranial melanomas.

Methods

We tested the therapeutic effect of injecting Delta-24-RGDOX into primary subcutaneous (s.c.) tumors derived from luciferase-expressing B16-Red-FLuc cells in s.c./s.c. and s.c./intracranial (i.c.) melanoma models in C57BL/6 mice. The growth of treated s.c. and untreated disseminated s.c. and i.c. tumors, luciferase-expressing ovalbumin-specific (OT-I/Luc) T cells within the tumor were monitored via bioluminescence imaging. Cells were profiled for surface markers using flow cytometry.

Results

In both s.c./s.c. and s.c./i.c. models, three injections of Delta-24-RGDOX significantly inhibited the growth of both the virus-injected s.c. tumor and untreated distant s.c. or i.c. tumor, thereby prolonging survival. The surviving mice were protected from rechallenging with the same tumor cells. The virus treatment increased the presence of T cells and the frequency of effector T cells in the virus-injected tumor and mediated the same changes in T cells from peripheral blood, tumor-draining lymph nodes (TDLNs), spleens, and brain hemispheres with untreated tumor. Moreover, Delta-24-RGDOX decreased the frequency of exhausted T cells and regulatory T cells in the virus-injected and untreated i.c. tumors. Consequently, the virus promoted the in situ expansion of tumor-specific T cells and their migration to tumors expressing the target antigen.

Conclusions

Intratumoral injection of Delta-24-RGDOX induced an in situ antovaccination of the treated melanoma, resulting in systemic immune activity against the disseminated tumors. This is the first report demonstrating that localized treatment with an oncolytic virus in the subcutaneous tumor is able to reject intracranial tumors, suggesting Delta-24-RGDOX could be applied to patients with brain melanoma metastasis. Furthermore, given the immunogenicity, cancer-selectivity and localized application of the virus, Delta-24-RGDOX is expected to be efficacious in patients with immunosuppressive (“cold”) tumors as well and have an improved safety profile in cancer patients compared to immune checkpoint blockade treatment strategies.

Correspondence: Eric Quemeneur (quemeneur@transgene.fr)

Background

Vaccinia virus (VACV) has been intensively used as oncolytic virus for the treatment of various types of cancers over the last years. Targeted gene deletion have enabled the selection of new VACV variants that retain tumor-specific replication, and oncolytic potential, but are safer for the surrounding healthy tissues. Recent examples of such deleted VACVs are TG6002, deltaJ2R(TK-)deltaI4L(RR-)-Fcu1 (Foloppe et al. 2019), and DeltaJ2R(TK-)DeltaF1L (Pelin et al. 2019). It is also well established that VACV secretes various factors interfering with major immune pathways, largely contributing to immune evasion for the virus. However, to our knowledge, none of these factors has already been used in the field of cancer immunotherapy.

Results

We here report the identification of M2L as a strong binder to both CD80, and CD86 co-stimulatory receptors. This binding antagonizes the interaction with their natural ligands CD28, and CTLA-4. M2L can also stabilize the interaction between CD80, and PD-L1 in vitro. We characterized M2L as a secreted homo-oligomeric protein (8 x 35 kDa), and could determine that apparent affinities are in the same range as natural ligands. It proved as active as CTLA-4, as an inhibitor for lymphocyte activation in a MLR assay.

M2L properties thus make it a potential new immuno-suppressive drug. Interestingly, M2L is largely conserved within the poxvirus family, and we could demonstrate that its ortholog from myxomavirus can also interact with CD86.

Expecting to reinforce the immunogenic properties of VACV, we engineered a triple-deleted Vaccinia virus (TD, DeltaJ2R, DeltaI4L, DeltaM2L). Oncolytic activity was not affected by the deletion of M2L, as assessed both in the regular mouse tumor xenograft models (HCT116), or in the syngeneic models (B16F10). We could demonstrate a better tolerance for the TD variant in a humanized model, where the CD80/CD86 pathway might be prominent in the neutralizing response.

Conclusions

Finally, the TD oncolytic backbone might be interesting for the development of our invirIO™ platform. Recent results with triple-deleted VACV expressing immune engagers will be presented.

References

1. Kleinpeter et al. By binding CD80 and CD86, the Vaccinia virus M2 protein blocks their interactions with both CD28 and CTLA4 and potentiates CD80 binding to PD-L1. J. Virol. 2019, 93: e00207-19

2. Foloppe et al. The enhanced tumor specificity of TG6002, an armed oncolytic Vaccinia virus deleted in two genes involved in nucleotide metabolism. Mol Ther Oncolytics 2019, 14: 1-14

3. Pelin et al. Deletion of apoptosis inhibitor F1L in Vaccinia virus increases safety and oncolysis for cancer therapy. Mol Ther Oncolytics 2019 (in press)

Correspondence: Michael Warren (mwarren@surfaceoncology.com)

Background

The purine nucleoside adenosine dampens innate and adaptive immune responses under various inflammatory conditions [1]. In contrast, high levels of extracellular adenosine triphosphate (ATP), generated as a result of tissue damage or immunogenic cell death, can initiate proinflammatory responses [2]. Extracellular adenosine accumulates in cancerous tissues through the degradation of ATP and constitutes an important mechanism of tumor immune escape, induction of angiogenesis, and metastasis [3]. Reduction of ATP and subsequent increase of extracellular adenosine production in the tumor microenvironment (TME) depend on the concerted enzymatic activity of two ectonucleotidases, CD39 (ENTPD1) and CD73 (NT5E), which convert ATP to adenosine monophosphate (AMP), and AMP to adenosine, respectively. Targeting this enzymatic pathway by inhibition or loss of CD39 has been shown to have antitumor effects in multiple studies [4-7]. The preclinical characterization of SRF617, a fully human CD39 antibody that potently inhibits CD39-mediated conversion of ATP to AMP, is described.

Methods

A series of studies were conducted to examine the CD39 targeting properties of SRF617. Biolayer interferometry (BLI) and flow cytometry-based cellular binding assays were used to examine the affinity of SRF617 to both recombinant and cellular CD39. The malachite green phosphate detection assay was used to assess the enzymatic inhibitory properties of SRF617 on recombinant and cellular CD39. In vitro functional assays to observe SRF617 modulation of dendritic cell (DC) maturation and T-cell proliferation were performed. Antitumor activity of SRF617 was assessed in a MOLP-8 xenograft model with multiple pharmacodynamic endpoints including immune cell infiltration, target occupancy, and in situ CD39 activity.

Results

SRF617 displays single-digit nanomolar inhibition of CD39 enzymatic activity and binds tightly to CD39 on cells. In vitro, SRF617 enhances DC maturation and T-cell proliferation under high ATP conditions. In vivo, SRF617 has significant single-agent antitumor activity in the MOLP-8 xenograft model. In addition, SRF617 elevates ATP levels systemically in treated mice and increases tumor macrophage infiltration in MOLP-8 xenograft tumors. An in situ CD39 enzymatic assay demonstrates that SRF617 inhibits CD39 activity in the MOLP-8 tumor xenograft model and correlates with target occupancy.

Conclusions

These studies demonstrate that SRF617 blocks CD39 enzymatic activity, thereby reversing the anti-inflammatory/immunosuppressive TME caused by reduced extracellular ATP and the generation of adenosine. These findings support future clinical studies of SRF617 as an immune-modulating medicine in patients with cancer.

References

1. Antonioli L, Blandizzi C, Pacher P, Hasko G. Immunity, inflammation and cancer: a leading role for adenosine. Nat Rev Cancer. 2013;13(12):842-57.

2. Di Virgilio F, Sarti AC, Falzoni S, De Marchi E, Adinolfi E. Extracellular ATP and P2 purinergic signalling in the tumour microenvironment. Nat Rev Cancer. 2018;18(10):601-18.

3. Spychala J. Tumor-promoting functions of adenosine. Pharmacol Ther. 2000;87(23):161 73.

4. Feng L, Sun X, Csizmadia E, et al. Vascular CD39/ENTPD1 directly promotes tumor cell growth by scavenging extracellular adenosine triphosphate. Neoplasia. 2011;13(3):206-16.

5. Jackson SW, Hoshi T, Wu Y, et al. Disordered purinergic signaling inhibits pathological angiogenesis in cd39/Entpd1-null mice. Am J Pathol. 2007;171(4):1395-404.

6. Sun X, Wu Y, Gao W, et al. CD39/ENTPD1 expression by CD4+Foxp3+ regulatory T cells promotes hepatic metastatic tumor growth in mice. Gastroenterology. 2010;139(3):1030-40.2000;87(2 3):161-73.

7. Zhang H, Vijayan D, Li XY, et al. The role of NK cells and CD39 in the immunological control of tumor metastases. Oncoimmunology. 2019;8(6):e1593809.

Ethics Approval

Mice were used in compliance with protocols approved by the IACUC of Mispro Biotech Services, Cambridge, MA (#2017-03-21SUR-1).

Correspondence: Paul Sondel(pmsondel@humonc.wisc.edu)

Background

We are investigating low dose mTRT as an immunomodulator in combination with in situ tumor vaccines (ISV). Most shared resource flow cytometry facilities do not analyze radioactive samples. To study the immunomodulatory effect of mTRT, we evaluated different fixation and cryopreservation protocols allowing radioactive tumor samples to be stained, frozen and safely stored until radioactivity decayed to background, enabling flow analyses. Our aim is to identify a protocol which most accurately reflects the tumor immune microenvironment (TME) of a freshly analyzed sample.

Methods

We generated syngeneic B78 melanoma flank tumors in C57BL/6 mice. Cohorts of mice received: 1) no treatment; 2) our previously reported ISV regimen (12Gy local external beam radiation and intratumoral injections of hu14.18-IL2 immunocytokine [1,2]); or 3) a combination of ISV and 50μCi of 90Y-NM600 mTRT [3]. We then harvested, dissociated and labeled tumor infiltrating immune cells according to one of four protocols: 1) freshly labeled cells; 2) freezing cells before staining; 3) freezing cells just before fixation; or 4) freezing cells after staining and fixation. Cells were labeled with panels specific for adaptive and innate immune cells.

Results

All three staining protocols involving cryopreservation performed similarly to freshly stained cells. Protocols 2 and 3 resulted in increased PD1 expression compared to fresh cells, suggesting cryopreservation and thawing of unfixed cells impacts the detection of the activation state of the lymphocytes. Protocol 4 most closely resembled the labeling of “fresh” cells. In comparing untreated tumors to those treated with ISV, we analyzed half of each sample “fresh” and half using staining protocol #4. We observed the same immune stimulatory effect of in situ vaccination using both protocols, confirming that freezing cells after labeling and fixing can be used to evaluate changes in the TME. Lastly, we used protocol #4 on a time course of weekly harvested tumors following mTRT alone, in situ vaccine alone, or combination mTRT+ISV. We demonstrated greater immune infiltrate and an increased CD8/Treg ratio when combining mTRT with ISV than either alone.

Conclusions

We demonstrate that fixation and cryopreservation after labeling closely reproduces the analyses of freshly dissociated tumors. We confirm this by comparing tumors treated with several immunotherapies using both the traditional and modified protocols. Lastly, we use this protocol with radioactive specimens to demonstrate that low dose mTRT followed by ISV results in greater immune infiltrate and activation with this combination compared to ISV or mTRT alone.

Acknowledgements

Thank you to Archeus Technologies of Madison, WI for their kind gift of the NM600 mTRT agent.

References

1. Morris Z, Guy E, Werner L, Carlson P, Heinze C, Kler J, Busche S, Jaquish A, Sriramaneni R, Carmichael L, Loibner H, Gillies S, Korman A, Erbe A, Hank J, Rakhmilevich A, Harari P, Sondel P. Tumor-Specific Inhibition of In Situ Vaccination by Distant Untreated Tumor Sites. Cancer Immunol Res. 2018 Jul;6(7):825-834.

2. Morris Z, Guy E, Francis D, Gressett M, Werner L, Carmichael L, Yang R, Armstrong E, Huang S, Navid F, Gillies S, Korman A, Hank J, Rakhmilevich A, Harari P, Sondel P. In Situ Tumor Vaccination by Combining Local Radiation and Tumor-Specific Antibody or Immunocytokine Treatments. Cancer Res. 2016 Jul 1;76(13):3929-41.

3. Grudzinski J, Hernandez R, Marsh I, Patel R, Aluicio-Sarduy E, Engle J, Morris Z, Bednarz B, Weichert J. Preclinical Characterization of 86/90Y-NM600 in a variety of murine and human cancer tumor models. J Nucl Med. 2019 Apr 6.

Correspondence: Vesselin Tomov (vesselin.tomov@uphs.upenn.edu)

Background

Melanoma is currently the fifth most common cancer in men and sixth most common in women with an estimate of 7230 deaths from the disease in 2019 in the United States. Treatment with the immune checkpoint inhibitors anti-CTLA-4 and anti-PD-1 has tremendously improved outcomes for patients with metastatic disease, but although durable response can be observed in over 50% of the patients treated with monotherapy, resistance is still present. Studies suggest that combining therapies (e.g. immune checkpoint inhibitors, radiotherapy, and/or agonistic antibodies) are more effective than monotherapy. 4-1BB (CD137) is expressed on various cell populations including T cells, dendritic cells and macrophages. 4-1BB agonists have been proved to be effective in multiple murine tumor models and have had modest responses in early clinical trials. We hypothesize that 4-1BB agonists augment the anti-PD-1 anti-tumor immune response resulting in improved overall survival and decreased tumor burden in melanoma resistant to pre-treatment with a combinational therapy of anti-CTLA-4 and radiation

Methods

499 melanoma cells, derived from unirradiated B16-F10 tumors resistant to treatment with anti-CTLA-4 and radiation, were implanted on both flanks of C57BL6 mice. Treatment with 4-1BB agonists and/or anti-PD1 was given on day 10, 13 and 16 post-implantation. The anti-tumor response was assessed by both overall survival and tumor burden. Tumors were harvested on day 17-20 for protein, transcriptomic and flow cytometric immune profiling

Results

Dual immunotherapy with 4-1BB agonist and anti-PD-1 significantly increases overall survival, whereas there is no benefit from treatment with either antibody alone. With combinational treatment, complete remission was achieved in 45.8%, however, there is still resistance to immunotherapy with response and nonresponse. Investigating mechanisms of response and non-response to dual immunotherapy with 4-1BB and anti-PD-1, we identified cells from the innate and adaptive immune system to play important roles in the anti-tumor response. Our proteome array analysis revealed several proteins of the myeloid and lymphoid compartment to be upregulated in the non-responders. Transcriptomic profiling by Nanostring IO 360 pan cancer panel in the non-responders and responders to immunotherapy corroborated our findings and detected important genes of the myeloid and lymphoid compartment especially macrophages and CD8+ T cells respectively

Conclusions

Overcoming resistance to immunotherapy crucially depends on cells of the innate and adaptive immune system especially the macrophages and CD8+ T cells. Investigation of response mechanisms revealed potential targets that might be effective amplifiers of the anti-tumor response to overcome resistance after immunotherapy.

Correspondence: Gregory Driessens (gregory.driessens@iteostherapeutics.com)

Background

Extracellular adenosine, acting predominantly through the A2A receptor (A2AR), mediates immunosuppression which includes suppression of Th1 responses and cell-mediated cytotoxicity as well as increasing the activity of Tregs and MDSC. We have developed EOS100850/1, a novel, non-brain penetrant and highly selective inhibitor of A2AR with sub-nanomolar Ki. In vivo, EOS100850/1 significantly delays tumor growth in different syngeneic models when combined with aCTLA-4 or aPD-L1. In vitro, we have demonstrated that EOS100850/1 maintains its potency in high adenosine environment (>10uM) using human PDX models. However, these PDX were engrafted on to immunodeficient mice and therefore any possible contribution of the immune infiltrating cells for generation of adenosine was not taken into account. Thus, the first aim of this study was to determine the adenosine levels in immunocompetent tumor models. The second aim was to confirm that EOS100850/1 can mediate antitumor efficacy in vivo regardless of the ambient adenosine concentration in the tumor microenvironment (TME).

Methods

MCA205, CT26, A20, and EMT6 tumor cells were subcutaneously implanted into immunocompetent syngeneic mice. When tumors reached 100-200mm3 and again at 600-700mm3 tumor extracellular fluid was collected by microdialysis and the adenosine concentration was measured by LC/MS. In a second study, the antitumor effect of EOS100850/1 administered orally at 0.6mg/kg QD was evaluated in models with different level of adenosine concentration (MCA205 and CT26) in combination with oxaliplatin or doxorubicin dosed at 10 or 5 mg/kg, respectively.

Results

The adenosine concentrations in individual tumors ranged from 2 to 50uM across different models, while adenosine levels in the non-tumor region were always less than 1uM. For most tumors, size did not affect the concentration of adenosine, except for CT26, where there was a 2 fold increase when tumors grew from 100-200mm3 to 600-700mm3. MCA205 tumors displayed the highest adenosine concentration with an average of 33uM while CT26 was considered as intermediate with an average of 6uM. EOS100850/1, in combination with chemotherapy, demonstrated similar tumor growth inhibition in both models, regardless of the concentration of adenosine.

Conclusions

These data demonstrate that the adenosine concentrations vary in the tumor microenvironment of different syngeneic tumor models and that EOS100850/1 is able to inhibit the A2AR induced immunosuppression and therefore tumor growth in combination with chemotherapy regardless of the adenosine concentration in the TME.

Acknowledgements

We would like to thank the CRO Explicyte for performing the microdialysis experiments.

Ethics Approval

The study was approved by La S.A. Biopole ULB Charleroi, approval number BUC 2016-02

Correspondence: Thomas Andresen (tlan@dtu.dk)

Background

Liposomal drug delivery systems have historically focused on tumor accumulating properties. This is generally achieved by securing a long-circulating half-life of liposomes that then accumulate by the enhanced permeation and retention (EPR) effect. Several important therapeutic targets beyond the cancer cells have been identified in the tumor microenvironment. Pro-tumorigenic myeloid populations in the tumor microenvironment are potent suppressors of anti-cancer cytotoxic T-cells and their depletion improve therapeutic outcome [1,2]. Liposomal drug delivery systems that, not only induce immunogenic cell death of cancer cells but also eliminate potent immunosuppressive populations are therefore highly attractive.

Methods

Weakly cationic PEGylated Oxaliplatin (L-OHP) liposomes (R2) were formulated (POPC/cholesterol/DOTAP/chol-PEG/Chol, (52.5/36/7.5/4)) (size 120 nm, polydispersity 0.1, charge +10 mV). Stealth liposomes were formulated (DSPC/cholesterol/DSPE-PEG2000, (55/40/5)) (size 106, polydispersity 0.02, charge -10). Therapeutic efficacy was investigated in the syngeneic CT26 colorectal cancer cell-line and tumor model. In vivo therapeutic efficacy was performed and the tumor microenvironment was analyzed using flow cytometry. Liposome biodistribution and pharmacokinetics were investigated using rhodamine B labeled stealth and R2 liposomes. L-OHP levels in the tumors were measured using inductively coupled plasma mass spectrometry (ICPMS). Real-time in vivo biodistribution was evaluated by PET/CT using copper-64-radiolabeled stealth and R2 liposomes according to a previously described technology [3].

Results

The in vitro half-maximal inhibitory concentration (IC50) of CT26 cancer cells was comparable between R2 liposomes and free L-OHP (2.2 and 2.0 μM). Due to in vitro stability, IC50 value for stealth L-OHP liposomes could not be achieved even at the highest concentration evaluated. PET/CT scans and ICPMS demonstrated that the R2 liposomes had a shorter circulating half-life and significant lower tumor accumulation compared to stealth liposomes, but higher than free L-OHP. However, the R2 L-OHP liposomes displayed an impressive anti-cancer efficacy with >60% of treated CT26 tumor-bearing mice displaying a complete response (CR), while no CR’s were observed in the L-OHP stealth liposome treated group. Advanced flow cytometric analysis of the tumor microenvironment identified that both R2 and stealth liposomes accumulated in central immunosuppressive myeloid populations but only R2 L-OHP liposomes were able to eliminate these compared to stealth and free L-OHP.

Conclusions

Here we report the potent anti-cancer effect of weakly cationic PEGylated L-OHP liposomes that depletes suppressive populations and eradicates CT26 tumors. The data demonstrate how advanced drug delivery systems may provide important immune-modulating properties in addition to direct cancer-cell cytotoxicity.

References

1. Parker KH, Beury DW, Ostrand-Rosenberg S. Myeloid-Derived Suppressor Cells: Critical Cells Driving Immune Suppression in the Tumor Microenvironment. In: Advances in Cancer Research. Vol 128. NIH Public Access; 2015:95-139.

2. Iclozan C, Antonia S, Chiappori A, Chen D-T, Gabrilovich D. Therapeutic regulation of myeloid-derived suppressor cells and immune response to cancer vaccine in patients with extensive stage small cell lung cancer. Cancer Immunol Immunother. 2013;62(5):909-918.

3. Petersen AL, Binderup T, Rasmussen P, et al. 64Cu loaded liposomes as positron emission tomography imaging agents. Biomaterials. 2011;32(9):2334-2341.

Ethics Approval

The study has been approved by The Danish Animal Experiments Inspectorate.

Correspondence: Meir Azulay (meir@neotx.com)

Background

Tumor-targeted superantigens (TTS) such as Naptumomab Estafenatox (Nap) are fusion proteins that consist of genetically engineered Superantigens (Sag) linked to Fragment antigen binding (Fab) moieties directed to tumor-associated antigens. Unlike CD3-based T cell redirection approaches (e.g. BiTEs) which bind and activate all T cells, TTS only bind and activate subsets of T cells that contain certain TCR β variable (TRBV) regions, e.g. TRBV 7-9 [1] and are thus defined as STR. We previously reported the synergistic anti-tumor effect of combining CPI with our lead STR compound, Nap (5T4 targeted Sag) or its murine surrogate protein [2]. Here, we present new pre-clinical data showing that STR not only enhances the anti-tumor effect of CPIs, but also stimulates the overall immune response that could lead to long term immunity against the tumor.

Methods

The combination of Nap with PD-L1 inhibitor (durvalumab) was tested in vitro against high (MDA-MB 231) and low (RKO) 5T4-expressing cancer cell lines in the presence of human PBMCs. For the in vivo studies, mice bearing EpCAM transfected MC38 tumors were treated with TTS (consisting of a Fab against EpCAM), an anti-PD-1 antibody, or the combination. Tumor growth and survival were assessed and tumor recurrence following re-challenge was evaluated.

Results

Combination of Nap with durvalumab had synergistic anti-tumor effect against both high and low 5T4-expressing cancer cell lines. Concomitant treatment of MC38-EpCAM tumor bearing mice with TTS and anti-PD-1 achieved complete tumor rejection in 4 of 10 mice and significantly prolonged survival and delayed outgrowth of tumors compared to the monotherapies. All cured mice rejected re-challenge with MC38-EpCAM and parental MC38 tumors, indicating long-term memory responses.

Conclusions

Our studies show that combination of CPI with STR overcomes the limited effect of CPI monotherapy regardless of tumor antigen expression level. In addition, our in vivo studies demonstrate that the combination of STR with CPI may lead to long term durable responses not possible in most patients receiving single agent CPI therapy. Moreover, the ability of these “cured mice” to reject tumor re-challenge suggests that STRs cause release of secondary antigens that prime subsequent immune responses. Taken together, our data suggests that combining anti-PD(L)1 with STR may be a promising therapeutic strategy for patients with solid tumors. Clinical phase 1/2 trial was recently initiated investigating the combination of Nap with durvalumab in subjects with selected advanced or metastatic solid tumors [NCT03983954].

Trial Registration

NCT03983954

References

1. Hedlund G, Eriksson H, Sundstedt A, et. al. The Tumor Targeted Superantigen ABR-217620 Selectively Engages TRBV7-9 and Exploits TCR-pMHC Affinity Mimicry in Mediating T Cell Cytotoxicity. PLoS One. 2013; 8(10): e79082

2. Azulay M, Lifshits S, Friedmann A, et al. Naptumomab Estafenatox induces T cell recognition, turning anti-PD-1 unresponsive "cold" tumors into "hot" responsive tumors. Cancer Research. Jul 2018, 78 (13 Supplement) abstract # 2712 AACR Annual Meeting 2018; Chicago, IL; DOI: 10.1158/1538-7445.AM2018-2712

Ethics Approval

The study was approved by of the Institutional Animal Care and Use Committee (IACUC) of Tel Aviv University, approval number 01-18-008

Correspondence: Efstathios Karathanasis (stathis@case.edu)

Background

The generation of a robust antigen-presenting cell (APC) response at the site of a tumor is central to effective cancer immunotherapy. Systemic administration enables widespread access to the microvasculature of a primary tumor and metastases. We encapsulated two powerful synergistic immune agonists within a single ~50-nm liposomal immuno-nanoparticle (immuno-NP) and exploited systemic delivery to achieve efficient deposition within the APC-rich tumor/metastases perivasculature. We elected to co-deliver an agonist of the stimulator of interferon genes (STING) pathway and a Toll-like receptor 4 (TLR4) agonist on the same immuno-NP to mediate their co-uptake by the same APCs to ensure functional synergistic production of Type I inteferon β (IFNβ) by these target cells. Here, we investigated the treatment efficacy of immuno-NPs in reducing primary tumor burden and preventing metastasis in an orthotopic murine model of triple-negative breast cancer (TNBC) and established the mechanistic basis for this anti-tumor immunity.

Methods

Female BalbC mice were inoculated orthotopically in the mammary fat pad with 5x10^5 4T1 TNBC cells. Tumor burden was monitored longitudinally by bioluminescence measurements of luciferase-expressing tumor cells and physical caliper measurements. Cytokines, immune cell populations, and immuno-NP deposition was measured by ELISA, flow cytometry, and confocal microscopy, respectively. One-/two-way ANOVA and Student’s t-test were used for statistical analysis.

Results

Immuno-NPs carrying both STING/TLR4 agonists mediated a robust synergistic production of IFNβ from macrophages that was 10-fold increased from NPs carrying either agonist alone (Fig. 1a, with transmission electron micrograph of immuno-NPs shown in inset). Tumor-bearing mice treated with immuno-NPs had elevated levels of tumor dendritic cells (DCs), macrophages, and natural killer (NK) cells, compared to controls that included mice treated with clinically-approved immune checkpoint inhibitors anti-PD1 and anti-CTLA4 (Fig. 1b). Primary tumor sizes were 50-60% reduced in mice treated with immuno-NPs compared to controls (Fig. 1c) and metastasis was prevented (Fig. 1d). Significantly, treatment with immuno-NPs outperformed treatment with immune checkpoint inhibitors.

Conclusions

Nanoparticle-mediated co-delivery of synergistic STING/TLR4 agonists drives a significant therapeutic outcome in the treatment of TNBC and prevention of metastasis in the 4T1 orthotopic model. Systemic delivery enables accumulation of immuno-NPs within the APC-rich perivasculature of the tumor/metastases, which is pivotal for the mediation of a significant anti-tumor immune response.

Acknowledgements

This work partially supported by grants from the National Cancer Institute (R01CA177716, U01CA198892), Alex’s Lemonade Stand Foundation, and the Angie Fowler AYA Cancer Research Initiative.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P658).

See text for description

Correspondence: Betty Chang (bchang@angarustx.com)

Background

Ectonucleotide pyrophosphatase 1 (ENPP1) was identified as the extracellular hydrolase for cGAMP, the natural ligand for STING [1,2]. Multiple lines of evidence suggest cancer cells produce soluble extracellular cGAMP, resulting in activation of CD14+ PBMCs, and subsequent production of IFNb in cell culture. Inhibition of ENPP1 and ionizing radiation further increase extracellular cGAMP concentrations and anti-cancer immunity in mice. Genetic knockout of ENPP1, in breast cancer cells or in the host, delayed tumor progression, demonstrating antitumor activity of ENPP1 deficiency. ENPP1 downregulation correlated with survival advantage in breast cancer patients. Here, systemic delivery of extracellular ENPP1 antagonists were evaluated in syngeneic mice tumor models alone and in combination with ionizing radiation and/or immune checkpoint inhibitors.

Methods

Novel SMI to ENPP1 were synthesized and characterized. ANG-1084 and ANG-1623, were identified as leads, and tested for potency, selectivity, stability, safety and PK. ANG-1623 was delivered at 50 mg/kg/d systemically for 7-28 days in models of pancreatic adenocarcinoma, neuroblastoma, orthotopic TNBC, and colon carcinoma.

Results

ANG-1084 and 1623 have IC50 of <0.5 nM against human ENPP1 with cGAMP as substrate at pH7.6, and are selective against hENPP2/3, PDEs, CEREP44 and stable in human and mouse microsomes. ANG-1623 had no hits in a 468-kinase-screen at 1 uM, and its protein binding was 66% and 55%, in human and mice, respectively. ANG-1084 and 1623 did not have CYP liability and hERG IC50 was >25 uM. In addition, both were superior in potency, selectivity, and hERG to the published ENPP1i QS-1. ANG-1623 at 50 mg/kg/day had single agent activity in the Panc02 model, with 63% TGI (p<0.001) and 81% TGI when combined with radiation (p<0.001, 5/15 tumor free). ANG-1623 in combination with radiation and/or checkpoint inhibitors had significant anti-tumor effects in Neuro-2a, orthotopic 4T1 or MC38 models.

Conclusions

Novel selective and potent small molecule inhibitors were designed to inhibit extracellular ENPP1. Lead molecule ANG-1623 has desirable properties of a drug and was subsequently tested in syngeneic mice tumor models: 4T1, Panc02, Neuro-2a, and MC38, where it was administered subcutaneously for 7 to 28 days. The drug was well tolerated and demonstrated significant anti-tumor activity alone or in combination with radiotherapy and/or anti-PD-L1. Since ENPP1 antagonists aim to preserve extracellular cGAMP produced by tumors either naturally or induced by ionizing radiation, and are delivered systemically, they may be promising alternatives to synthetic CDNs that are restricted to intratumoral injections of surface tumors, and pose challenges in optimal dose determinations.

References

1.Li, L et al., Hydrolysis of 2’3’-cGAMP by ENPP1 and design of nonhydrolyzable analogs. Nat Chem Biol 2014;10(12)1043-8.

2.Carozza, JA et al., 2’3’-cGAMP is an immunotransmitter produced by cancer cells and hydrolyzed by ENPP1. 2019; bioRxiv539312 doi: https://doi.org/10.1101/539312

Correspondence: Anthony Desbien (TDESBIEN@ADURO.COM)

Background

The stimulator of interferon genes (STING)-pathway is a critical component of the adaptive immune response against tumors. Aduro Biotech is developing the synthetic cyclic dinucleotide (CDN) STING agonist, ADU-S100 (MIW815), as a cancer treatment.

Methods

A dose-response analysis of intratumorally administered STING agonists in syngeneic mouse tumor models was performed to determine immunogenicity and tumor control.

Results

Data show that low to mid-range doses caused injected tumor collapse and the activation and expansion of tumor-specific CD8 T cells. In contrast, overstimulation of the STING pathway with high-doses of ADU-S100 or more potent STING agonists failed to induce tumor-specific CD8 T cells and caused disruption of tumor draining lymph nodes. Blocking tumor necrosis factor-alpha (TNF-alpha) signaling restored induction of tumor-specific CD8 T cells during overstimulation of STING and preserved lymph node integrity.

Conclusions

These data demonstrate that the induction of anti-tumor CD8 T cells is inhibited by TNF-alpha during overstimulation of the STING pathway, suggesting that intervention in TNF-alpha signaling may broaden the therapeutic window of STING agonists.

Ethics Approval

All animals were used according to protocols approved by Institutional Animal Use Committee of

Aduro Biotech.

Correspondence: Monika Dobrzańska (monika.dobrzanska@selvita.com)

Background

Stimulator of Interferon Genes (STING) is a major player in the initiation of robust innate immune activation leading to initiation and enhancement of a tumor-specific adaptive immunity. Several clinical and pre-clinical studies employ modified cyclic dinucleotides, natural STING ligands. Yet, relative instability and chemical nature limit their use as systemic immuno-therapeutics. Herein, we present potent and selective non-nucleotide, non-macrocyclic, small molecule direct STING agonists, structurally unrelated to known chemotypes with potential for systemic administration.

Methods

Binding to recombinant STING protein was examined using FTS, MST, FP and crystallography studies. Phenotypic screen was performed in THP-1 Dual reporter cells. Human macrophages (HMDM) and dendritic cells (HMDC) were differentiated from monocytes (obtained from PBMC) in the presence of GM-CSF and GM-CSF/IL-4 for HMDM and HMDC, respectively. Human T cells were isolated from PBMC, activated with plate-bound anti-CD3/anti-CD28 and exposed to compounds followed by viability and proliferation assessment using flow cytometry. Mouse bone marrow-derived dendritic cells (BMDC) were obtained from C57BL/6 or STING Gt mice and differentiated with mIL-4 and mGM-CSF. Cell maturation marker expression and the presence of phosphorylated forms of the STING pathway proteins were assessed by flow cytometry and Western Blotting. BALB/c mice were injected with compounds and the cytokine release was measured in the plasma. Additionally, mice were inoculated with CT26 cells and the compound was administered followed by the regular tumor growth monitoring.

Results

Selvita’s agonists bind to all tested recombinant STING proteins: human, mouse, rat and monkey. Selvita’s compounds induce cytokine responses and upregulate DC maturation markers on BMDC in low nM range only in the presence of STING, confirming their high selectivity. They trigger pro-inflammatory cytokine release from human PBMC and HMDC and induce dendritic cell maturation regardless of the STING haplotype. In vitro repolarization of immuno-suppressive M2d macrophages into pro-inflammatory M1-like phenotype was demonstrated. The compounds have fine-tunable ADME properties with particularly good solubility, permeability, human plasma stability and plasma protein binding. Systemic in vivo administration led to robust upregulation of STING-dependent pro-inflammatory cytokines suggesting immune activation which translated into efficacy in vivo in subcutaneous CT26 model.

Conclusions

Selvita’s STING agonists activate STING-dependent signaling in both mouse and human immune cells promoting anti-tumor immunity. Treatment with Selvita’s STING agonists led to tumor growth inhibition by engagement of the immune cells. The compounds show good selectivity and in vitro ADME properties enabling development for systemic administration as a single agent or in combinations with immunotherapies or targeted agents.

Correspondence: James Mulé (james.mule@moffitt.org)

Background

STING pathway is a major innate immune sensing mechanism for the detection of immunogenic tumors. We have recently shown that STING signaling is heterogeneously regulated across human melanoma cell lines. While intact activation of STING signaling in a subset of melanoma cell lines enhances their antigenicity, defects in STING pathway limit their sensitivity to lysis by human melanoma tumor infiltrating lymphocytes (TIL) [1]. In addition, we have reported a high incidence of STING gene methylation in a variety of tumors including melanoma [2]. Based on these findings, we hypothesized that reconstitution of STING expression through DNA demethylation would rescue STING signaling in STING-defective melanoma cell lines and could therefore promote antitumor T-cell activity.

Methods

Using methylation microarray analysis, we identified human melanoma cell lines with aberrant STING promoter methylation across a panel of 16 cell lines and subjected them to treatment with the DNA methyltransferase inhibitor 5-aza-2’-deoxycytidine (5AZADC). We next evaluated the induction of STING expression by immunoblot. We assessed phosphorylation of IRF3 and induction of IFN-beta and CXCL10 in 5AZADC-treated melanoma cells following their stimulation with the STING agonist 2’3’-cGAMP. We also co-cultured 5AZADC-treated melanoma cell lines with their HLA-matched human melanoma TIL in the presence or absence of 2’3’-cGAMP and assessed TIL production of IFN-γ.

Results

We identified 6 STING-negative human melanoma cell lines with high levels of STING promoter methylation. Immunoblot analysis revealed inhibiting DNA methylation by 5AZADC treatment can partially or completely restore STING expression in all 6 cell lines. Stimulation with 2’3’-cGAMP resulted in phosphorylation of IRF3 and induction of CXCL10 (~1200 pg/ml, p < 0.01) and IFN-beta (~900 pg/ml, p < 0.0001) in 3 of 6 5AZADC-treated cell lines. In co-culture studies 5AZADC-treated melanoma cells stimulated up to a 4-fold increase in TIL production of IFN-γ compared to untreated controls in the presence of 2’3’-cGAMP (510 to 2100 pg/ml, p < 0.05).

Conclusions

We provide evidence that DNA methylation is a major contributor to the suppression of STING signaling in melanoma. Reconstitution of STING expression through DNA demethylation can restore STING signaling in STING-defective melanoma cell lines and promote antitumor T-cell activity. Collectively, these observations argue that targeting epigenetic loss of STING in melanomas should be considered as a strategy to improve the efficacy of clinical interventions using STING agonists.

Acknowledgements

NCI P50 CA168536, Cindy and Jon Gruden Fund, Chris Sullivan Fund, V Foundation, Dr. Miriam and Sheldon G. Adelson Medical Research Foundation.

References

1. Falahat R, et al. Tumor cell-intrinsic STING signaling impacts antigenicity of melanoma and can promote antitumor T-cell activity. Cancer Immunology Research, 2019.

2. Konno H, et al. Suppression of STING signaling through epigenetic silencing and missense mutation impedes DNA damage mediated cytokine production. Oncogene, 2018; 37(15): 2037.

Correspondence: Joseph Krueger (jkrueger@invicro.com)

Background

The cGAS-STING (cyclic GMP-AMP synthase-stimulator of interferon genes) pathway is becoming an important point of focus in the expanding landscape of next-generation immunotherapies. STING agonists can induce systemic anti-tumor immunity and thus may be also synergistic with other immunotherapies (i.e. immune checkpoint blockade, cancer vaccines, and CAR-T cell therapy). Importantly, a functional cGAS-STING pathway may be essential for response to oncolytic viral therapy. However, advanced stages of solid cancers (i.e. melanoma and colon adenocarcinoma) frequently show loss of STING or cGAS and would not be expected to respond to STING agonists. Since STING agonists have now entered clinical trials, there is an emerging need for predicting patient response which can accurately classify patients according to STING or cGAS expression. Here, we show application of a novel immunohistochemistry (IHC)-based approach using fluorescent nanoparticles (Quanticell™) to deliver highly sensitive and quantitative analysis of STING and cGAS within the tumor contexture.

Methods

cGAS and STING expression were simultaneously detected in melanoma tissues using our novel fluorescent nanoparticle IHC approach (Quanticell™). Serial sections were stained with chromogenic IHC to correlate between the expression of cGAS/STING and the patterns of tumor-infiltrating immune cells.

Results

Our results show that the Quanticell assay format enables a more objective and quantitative scoring of target proteins compared to standard IHC assay formats. The ability to multiplex cGAS and STING enables simultaneous quantitative analysis of these two critical markers to characterize distinct patient phenotypes with tumors associated with diverse immune contexture.

Conclusions

We show that our technique quantitatively measures cGAS and STING expression in cancer tissue biopsies with high sensitivity. The multiplexed IHC format has immediate value for predictive and prognostic application in STING agonist-based single/checkpoint inhibitor combination therapies.

Acknowledgements

Hemi Dimant, Ankit Gandhi, Stephan Collins, Mark Lawson.

Correspondence: Brendon Monahan (brendon@cancercrc.com)

Background

STING is a ubiquitously expressed innate immune sensor which is essential for the production of host defence-related proteins including type I interferons and proinflammatory cytokines that promote the recruitment and effector activity of innate and adaptive immune cells [1]. Targeted activation of STING and the associated release of type I interferons can enhance the sensitivity of antigen presenting cells to neoantigens and promote their ability to prime T cell-based anti-cancer immune responses [1,2]. This can lead to the establishment of T cell inflamed tumor microenvironments that are more susceptible to the anti-tumor effects of checkpoint blockade therapy [3]. The Cancer Therapeutics CRC (CTx) is developing direct small molecule STING agonists that can be administered systemically as an adjunct to immune checkpoint antibody therapy.

Methods

A phenotypic screen was run using an IRF3-luciferase reporter cell line looking for activation of Type I Interferon genes when screened against CTx small molecule libraries. Validated hits were deconvoluted by STING SPR to confirm direct STING binding. A mature medicinal chemistry program is underway with an established screening cascade including in vitro target engagement and primary cell functional assays, structural biology, in vivo PK/PD, and efficacy assays to determine mechanism(s) of action and demonstrate therapeutic index.

Results

CTx has developed potent drug-like small molecule STING agonist compounds that show direct STING binding. Cellular activity is dependent upon STING (inactive in STING-/- cells), independent of cGAS (active in cGAS-/- cells), and can be blocked by TBK1 inhibitors. Compounds show activity in primary mouse and human cells across a range of STING variants. Dose-dependent in vivo PD biomarker responses, including increased plasma IFN-β levels, were seen when administered intravenously (I.V) or intratumorally (I.T). In vivo, single agent efficacy of CTx STING agonists has been tested in multiple syngeneic mouse models of solid cancer. Treatment via I.T or I.V administration was well tolerated and evoked therapeutically beneficial immune-mediated anti-tumor responses. Successful induction of immunological memory, in response to STING agonist treatment was demonstrated in tumor-rechallenge experiments. In less responsive models, where intermediate efficacy was observed, increased tumor growth control could be achieved by co-treating mice with our STING agonist and anti-PD-1 antibody therapy.

Conclusions

CTx has a mature STING agonist program that has successfully produced direct STING activators that can promote primary DC function and elicit immune-mediated anti-tumor responses in mouse models of solid cancer.

References

1. Bose, D. cGAS/STING Pathway in Cancer: Jekyll and Hyde Story of Cancer Immune Response. Int J Mol Sci. 2017. 18:2456.

2. Vanpouille-Box, C. et al. Cytosolic DNA Sensing in Organismal Tumor Control. Cancer Cell. 2018. 34:361-378.

3. Woo, S.R., Corrales, L. & Gajewski, T.F. The STING pathway and the T cell-inflamed tumor microenvironment. Trends Immunol. 2015. 36:250-256.

Correspondence: Thomas Gajewski (tgajewsk@medicine.bsd.uchicago.edu)

Background

Previous work from our laboratory has shown that tumors activate the STING pathway in antigen-presenting cells. STING signaling leads to IFN-β production which is required for T cell priming through dendritic cells. While the best defined transcription factor for IFN-β gene expression is IRF3, other transcriptional regulators also contribute, including NF-κB. Enhancing STING activation and downstream IFN-β production has therapeutic potential, and STING agonists are currently being evaluated in clinical trials. However, innate immune activation in response to a pathogen rarely occurs by activating only one signaling pathway, and targeting multiple pathways may lead to more robust activation. Activating multiple innate immune pathways similar to a natural infection could improve the therapeutic efficacy of STING agonists.

Methods

In vitro experiments were performed using macrophages and bone marrow-derived dendritic cells stimulated with the STING agonist DMXAA alone or in combination with several TLR agonists. Cells were subjected to several downstream analyses, including qPCR, Western blotting, and immunofluorescent imaging to visualize transcription factor localization.

Results

Each innate immune agonist was titrated to identify the dose inducing maximal IFN-β expression. DMXAA induced greater peak IFN-β expression than any individual TLR agonist. After testing several combinations, we found that LPS + DMXAA induced significantly greater IFN-β transcription than the sum of either agonist alone. To explain this synergy, we assayed each step of STING pathway signaling. Initial activation leads to STING aggregation, and LPS did not increase STING aggregation beyond that induced by DMXAA. Additionally, LPS did not increase IRF3 phosphorylation nor IRF3 nuclear translocation in combination with DMXAA. Rather, LPS increased the phosphorylation and nuclear translocation of the NF-κB subunit p65. These results suggest that the synergy in IFN-β production is achieved through IRF3 activation downstream of STING signaling and NF-κB activation downstream of TLR4 signaling. Additional mechanistic experiments are ongoing.

Conclusions

TLR activation by LPS synergized with STING signaling to augment IFN-β production beyond the level of either agonist alone. Mechanistically, STING pathway activation preferentially signals through IRF3 whereas TLR4 pathway activation preferentially signals through NF-κB. Combining activation of these pathways could improve the innate immune response in a way that is therapeutically beneficial.

Correspondence: Sriram Sathyanarayanan (sriram.sathy@codiakbio.com)

Background

Activation of the STING pathway by direct intra-tumoral administration of synthetic cyclic dinucleotide (CDN) agonists has shown anti-tumor activity in animal models and early stage human clinical trials. At higher doses CDNs induce local tissue damage due to non-selective uptake into cells at the injection site, including ablation of immune effector cells (CD8+ T cells and Antigen Presenting Cells [APC])[1]. CDN administration has been shown to result in a bell-shaped dose response curve, wherein anti-tumor effects are lost at high doses [1]. We have developed a novel engineered exosome therapeutic candidate, exoSTING™, which selectively activates the STING pathway in tumor-resident APCs demonstrating greater potency than free CDN and without immune ablation.

Methods

Exosomes were engineered to overexpress PTGFRN, an exosome surface protein, and loaded ex vivo with a proprietary CDN. In vivo pharmacodynamics (PD) and anti-tumor responses were demonstrated in a B16F10 syngeneic mouse model and ex vivo histo-culture of human tumor explants (head and neck squamous cell carcinoma [HNSCC] and non-small cell lung cancer [NSCLC] patients) was utilized to evaluate PD after intra-tumoral administration.

Results

ExoSTING™ was preferentially taken up by differentiated macrophages and dendritic cells. In vitro proliferation studies showed a 100-fold increase in potency of macrophage activation by exoSTING™ treatment compared to free CDN. In contrast, T-cells failed to take up exosomes and were not activated by exoSTING™. We have shown that intra-tumoral administration of exoSTING™ in B16F10 tumors preserves the viability of T-cells and APCs, reduces collateral tissue damage, and eventually recruits more T-cells into tumors, whereas free CDN induces immune cell ablation. NanoString gene expression analysis showed increased immune cell-related gene signatures in tumors after dosing with exoSTING™ but not free CDN exposed tumors. ExoSTING™ produced a systemic tumor-specific T-cell response as demonstrated by growth inhibition of non-injected distal tumors, which was further enhanced by combination with anti-PD1 checkpoint blockade. Activity of exoSTING™ was also examined in ex vivo human tumor histo-culture. ExoSTING™ induced IFNβ mRNA expression in a dose-dependent manner and induced more IFNβ, CXCL9, CXCL10, and IFNγ mRNA expression than equivalent amounts of free CDN.

Conclusions

ExoSTING™ is an engineered exosome therapeutic candidate that specifically targets the STING pathway in APCs in the tumor microenvironment, resulting in enhanced PD effects, avoidance of T cell uptake and enhanced anti-tumor immune responses compared to free CDN. Human clinical testing is planned to begin in the first half of 2020.

References

1. Sivick KE, Desbien AL, Glickman LH, et al. Magnitude of Therapeutic STING Activation Determines CD8+ T Cell-Mediated Anti-tumor Immunity. Cell Rep. 2018;25(11):3074-3085.

Ethics Approval

The study was approved by Codiak BioSciences' internal IACUC committee, approval number CB2017-001.

Correspondence: Darrell Irvine (djirvine@mit.edu)

Background

Cyclic dinucleotides (CDN) – agonists of Stimulator of IFN genes – can initiate potent anti-tumor immunity due in part to activation of antigen presenting cells [1]. Recent studies have highlighted the role of CDN dose in governing the involvement of CD8+ T cells in anti-tumor immune response [2]. Intratumoral CDN co-therapy with systemic anti-PD-1 checkpoint blockade in bilateral 4T1 mammary carcinoma flank tumors results in regression of injected and distal tumors [2]. However, dual flank models largely ignore the role of the orthotopic tumor microenvironment, the natural process of metastatic development, and the impact of surgical resection – a predominant intervention in breast cancer [3].

Methods

We employed a surgical resection model of metastatic 4T1 mammary carcinoma to examine the effects of surgery and lung metastasis on CDN efficacy. 4T1-luciferase cells were inoculated in the mammary fat pad on Day 0, palpable tumors were then treated with immunotherapy starting on Day 6, any remaining primary tumor was surgically resected on Day 15, and mice were monitored for metastases by luciferase imaging. Intratumoral bisphosphorotioate 2’3’ c-di-AMP (CDN) was combined with systemic administration of mouse serum albumin IL-2 fusion protein (MSA-IL2) and anti-PD-1 followed by primary tumor growth measurements and monitoring for overall survival. AH1 tetramer staining of PBMC was utilized to monitor antigen-specific CD8+ T cells.

Results

In mice bearing established 4T1-luciferase tumors, administration of three CDN doses results in no cures if surgical resection is not performed. When administered prior to surgical resection CDN monotherapy yields a 20% cure rate and enhanced median overall survival compared to untreated mice (median survival 44.5 days vs 38 days, p=0.0026). Combination of CDN with MSA-IL2 and anti-PD-1 further improves survival, with 60% of mice surviving long-term. Using a tdTomato-expressing tumor line we find increased numbers of tdTomato+ CD103+ DCs in the tumor draining lymph node of CDN + MSA-IL2 + anti-PD-1-treated mice compared to CDN alone (p=0.016), indicative of enhanced antigen uptake. After surgical resection, triple combination-treated mice displayed higher levels of AH1-specific CD8+ T cells in the blood compared to CDN or CDN + anti-PD-1 treated mice. Upon rechallenge, AH1-specific T cells remain elevated in the combination group; however, complete rejection was not observed.

Conclusions

Our findings suggest that CDN monotherapy is less effective in the recurrent, orthotopic model of 4T1 mammary carcinoma than in flank tumors, but combination with systemic immunotherapy can improve immune control of both primary and metastatic disease.

Acknowledgements

The research reported was supported by the Howard Hughes Medical Institute, the Marble Center for Cancer Nanomedicine, and the NIGMS/NIH Interdepartmental Biotechnology Training Program (Grant# T32-GM008334).

References

1. Corrales L et al. Direct activation of STING in the tumor microenvironment leads to potent and systemic tumor regression and immunity. Cell Reports. 2015; 11:1018-1030.

2. Sivick K et al. Magnitude of therapeutic STING activation determines CD8+ T cell-Mediated anti-tumor immunity. Cell Reports. 2018; 25:3074-3085.

3. Al-Sahaf O et al. Surgical injury enhances the expression of genes that mediate breast cancer metastasis to the lung. Ann Surg. 2010; 252:1037-1043.

Ethics Approval

All mouse experiments were approved by MIT’s Committee on Animal Care, protocol# 0717-076-20.

Correspondence: Zezhou Wang (joe@trilliumtherapeutics.com)

Background

Stimulator of interferon genes (STING) is a master regulator of interferon (IFN)-mediated immune responses, and the cGAS-STING pathway plays an essential role in detection of aberrant cytosolic DNA fragments. The cascade of type I IFNs and pro-inflammatory cytokines that are produced following STING activation serve to bridge the innate and adaptive immune systems, and trigger potent anti-tumor immunity. STING is thus a promising therapeutic target in cancer and several STING agonists are currently in clinical development, though most are suitable only for intratumoral (i.t.) rather than systemic administration. We have previously shown that TTI-10001, a novel non-cyclic dinucleotide (CDN) small molecule STING agonist, potently induces the STING pathway in vitro and results in robust anti-tumor activity in vivo following i.t. administration. Here we report that TTI-10001 is well tolerated and demonstrates potent anti-tumor monotherapy activity in mice after oral or intravenous (i.v.) dosing.

Methods

TTI-10001 was administrated as a single dose (oral, i.v., or i.t.) to MC38 tumor-bearing mice and plasma or tumor samples were collected for pharmacokinetic (PK)/pharmacodynamic (PD) analysis. Repeat administration (oral, i.v., or i.t.) of TTI-10001 was used for maximum tolerated dose (MTD) studies or efficacy studies in single or dual flank MC38 tumor-bearing mice. Off-target effects were assessed using the SafetyScreen44™ panel.

Results

Mice dosed with TTI-10001 demonstrated favorable drug exposure with excellent oral bioavailability (85%) and a long tumor retention time (T1/2: ~ 4-8h) following all three routes of administration. Treatment with TTI-10001 induced dose-dependent expression of pro-inflammatory cytokines in tumors including interferon beta, tumor necrosis factor α, and interleukin 6. No significant off-target effects were observed in vitro. TTI-10001 was well tolerated in mice with no weight loss or overt morbidity after repeated systemic dosing. Potent, dose-dependent anti-tumor efficacy in the MC38 syngeneic mouse tumor model was observed when TTI-10001 was administered at doses below the MTD using all three modes of delivery.

Conclusions

TTI-10001, a novel non-CDN small molecule STING agonist, is well tolerated in vivo and displays an excellent PK/PD profile that results in potent anti-tumor activity by i.t., oral, and i.v. routes of administration. These data highlight the potential of TTI-10001 to achieve best-in-class status among next generation STING agonists.

Ethics Approval

All mouse experiments were approved by the University of Toronto animal care committee in accordance with the regulations of the Canadian Council on Animal Care (University of Toronto approved protocol # 20011874).

Correspondence: Arthur Sands (asands@nurix-inc.com)

Background

E3 ubiquitin ligases play critical roles in directing cellular protein fate by controlling the specificity of ubiquitin conjugation to substrate proteins and targeting them for cellular localization or degradation by the ubiquitin proteasome system. The E3 ubiquitin ligase CBL-B is expressed in immune cell lineages and negatively regulates activity of the T-cell receptor (TCR) through substrate proteins that impose a requirement for a costimulatory signal to mount a productive immune response upon TCR engagement. Mice deficient in Cbl-b or lacking CBL-B ligase activity demonstrate a tumor rejection phenotype mediated by CD8+ T cells (1,2). CD4+ and CD8+ T cells from mice deficient in Cbl-b have 5 to 10-fold enhanced secretion of IL-2 and IFN γ when stimulated ex vivo. Cbl-b deficient mice also demonstrate enhanced NK cell function (3). These data provide a genetic rationale for the development of a small molecule inhibitor of CBL-B ligase activity for use in patients with tumor-mediated immune suppression of effector T cells.

Methods

We have identified a series of small molecule inhibitors of CBL-B activity with biochemical potency at low nanomolar concentrations. CBL-B inhibitors enhanced in vitro T cell activation measured by cytokine secretion at low nanomolar concentrations in primary human and mouse T cells stimulated with anti-CD3/anti-CD28 or anti-CD3 alone.

Results

Furthermore, the compounds increased antigen recall responses in human PBMCs. CBL-B inhibitors also amplified T cell responses in models of anergy and exhaustion, with greater effects observed when combined with a checkpoint inhibitor. The CBL-B inhibitors also enhanced cytokine response in primary human NK cells measured by TNFα and IFNγ secretion. Oral dosing of an optimized CBL-B inhibitor enhanced anti-CD3 stimulated T cell activation in mouse CD4+ and CD8+ T cells, demonstrating a dose proportional pharmacodynamic effect.

Conclusions

These data support the continued advancement of small molecule oral CBL-B inhibitors for future development in immuno-oncology.

References

1. Loeser S, Loser K, Bijker MS, Rangachari M, van der Burg SH, Wada T Beissert S, Melief CJ, Penninger JM. Spontaneous tumor rejection by cbl-b-deficient CD8+ T cells. J Exp Med, 2007;204(4):879-91.

2. Paolino M, Thien CB, Gruber T, Hinterleitner R, Baier G, Langdon WY, Penninger JM. Essential role of E3 ubitquitin ligase activity in Cbl-b-regulated T cell functions.J Immunol. 2011 Feb 15;186(4):2138-47.

3. Paolino M, Choidas A, Wallner S, Pranjic B, Uribesalgo I, Loeser S, Jamieson AM, Langdon WY, Ikeda F, Fededa JP, Cronin SJ, Nitsch R, Schultz-Fademrecht C, Eickhoff J, Menninger S, Unger A, Torka R, Gruber T, Hinterleitner R, Baier G, Wolf D, Ullrich A, Klebl BM, Penninger JM. The E3 ligase Cbl-b and TAM receptors regulate cancer metastasis via natural killer cells. Nature. 2014 Mar 27;507(7493):508-12.

Correspondence: James Appleman (jappleman@primmunerx.com)

Background

While ICPIs (immune checkpoint inhibitors) have fundamentally changed the practice of cancer therapy for tumors arising from many different tissues, ways to increase both response rate and durability are critically needed. Combining stimulators of innate immunity with activators of adaptive immunity should lead to better treatment outcomes. We have therefore created a series of novel orally administered, systemically acting Toll-like receptor 7 (TLR7) agonists for long-term combination therapy with enhancers of adaptive immunity like ICPIs. Our molecules incorporate a set of atypical properties critical to achieving this product profile [1].

Methods

Key assays in our testing cascade include:

• Cellular reporter assays and human PBMC assays to evaluate selectivity, potency and other characteristic aspects of cellular pharmacology (e.g. targeted cytokine and chemokine profiles)

• Characterization in cynomolgus monkeys

---Appropriate PK with active TLR7 agonist

---Efficient oral delivery into systemic circulation with prodrugs of the TLR7 agonist

---Targeted pharmacodynamic response at relevant oral dose (gating for candidate selection)

Additional descriptive assays include:

• Evaluation of potency and cellular pharmacology in monkey PBMCs

• In vitro predictors of pharmacokinetics and safety

• Pharmacodynamic response in rodents

• Antitumor activity in syngeneic rodent tumor models

Results

While TLR7 agonists having appropriate potency and specificity were discovered early in our program, systemic oral delivery even with an optimized prodrug approach was relatively poor. Modification of the lead TLR7 agonist itself increased efficiency of oral delivery by the prodrug from 14% for Compound A to 99% for Compound C (Figure 1) while also modestly increasing potency in PBMC assays. Surprisingly, Compound C is significantly less potent in monkey plasmacytoid dendritic cells (PDCs) than in human PDCs. Consequently doses of the prodrug of Compound C eliciting our targeted degree of immune induction in cynomolgus monkeys are higher than anticipated. In contrast, Compound B has equivalent potency in human and monkey PBMC assays. While either of these compounds may be suitable for development, we continue to investigate additional compounds that retain the preferred features found in Compounds B and C while having greater potency in cellular assays.

Conclusions

From our original starting point - a relatively weak TLR7 agonist with no oral bioavailability - we have invented a novel series of molecules that are designed to be dosed QOD continuously over a 24-month period to appropriately engage innate immunity at a level that is well-tolerated by the patient while increasing treatment response rate and durability.

References

1. Appleman J, Webber, S. Discovery of a series of novel toll-like receptor 7 agonists for systemic immunotherapy of cancer [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2019. Mar 29 - Apr 3; Atlanta, GA. Abstract # 3262.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P670).

Efficient Oral Delivery of Novel TLR7 Agonists

Correspondence: Austin Boesch (aboesch@torquetx.com)

Background

TLR7 agonists boost immune responses in the tumor microenvironment (TME), primarily through dendritic cell (DC) engagement, enhancement of antigen presentation and T cell co-stimulation. However, multiple TLR agonists have displayed unfavorable PK/PD profiles and considerable toxicities upon systemic administration. Instead, we designed a T cell mediated delivery system of TLR7 agonists that target TME and the lymphatic system to maximize efficacy while avoiding systemic toxicities. Torque’s Deep-Primed™ T cell technology enhances T cell function through tethering of immune modulators to the cell before adoptive cell transfer (ACT) and uses Torque’s multi-targeted T cells (MTC) platform that targets multiple tumor antigens. By transporting payload to antigen-expressing tissues, Deep-Primed™ MTC’s focus the effect of the immunomodulators on these tissues. Here, we show that Deep TLR Primed T cells delivering TLR7 agonists induce potent immune cell activation in the TME and elicit exquisite anti-tumor efficacy without overt toxicity.

Methods

TLR7 agonists and several liposomal formulations were screened for optimal T cell tethering and release, measured by HPLC. Formulations with desired characteristics were tethered to murine PMEL CD8 T cells specific for the B16-F10 melanoma antigen gp100 to generate Deep TLR Primed T cells. Following ACT into immunocompetent syngeneic tumor-bearing animals, the T cell product was evaluated for efficacy and immune cell activation.

Results

ACT with Deep TLR Primed PMEL T cells inhibited tumor growth significantly more than ACT of PMEL T cells alone or combined with systemically delivered TLR agonists. Deep TLR Primed cells displayed higher proliferation but lower PD-1 expression in blood and tumors and caused a trend towards an increase in CD8+ DC, plasmacytoid DC, and MDSC in the TME. Despite their remarkable efficacy, ACT of Deep TLR Primed PMEL T cells caused no treatment-specific weight loss and elicited substantially lower levels of plasma IFNg and IL-12 than systemic TLR7 agonist delivery.

Conclusions

Deep TLR Primed T cells display superior efficacy, PD, and safety compared to T cells alone or co-administered with systemic TLR agonists. Torque’s Deep-Primed™ technology allows the delivery of small molecules to the TME, with controlled doses. Compared to intratumoral delivery, agonist delivery via Deep-Primed™ tumor antigen-specific autologous T cells can target a wider variety of tumors, including distant metastases. It has the potential to improve agonist PK profile through sustained delivery over time in TME and draining lymph nodes, with limited systemic exposure.

Correspondence: Douglas McNeel (dm3@medicine.wisc.edu)

Background

Expression of T-cell checkpoint receptors occurs following T-cell activation during exhaustion, and ligation of these receptors can impair T-cell function and anti-tumor immunity [1]. We previously found that T cells activated with cognate antigen upregulate the expression of PD-1, and the expression of PD-1 can be attenuated by the presence of specific Toll-like receptor (TLR) agonists [2,3]. This effect was mediated by IL-12 secretion from professional antigen presenting cells and resulted in CD8+ T cells with greater anti-tumor activity. In the current report, we sought to determine whether multiple TLR agonists can affect the expression of T-cell checkpoint receptors.

Methods

Splenocytes were isolated from OT-1 mice and stimulated with chicken ovalbumin peptide (SIINFEKL) in the presence of different combinations of TLR agonists for 4 days. Cells were collected daily and evaluated for expression of a marker of activation (4-1BB) and T-cell checkpoint receptors (PD-1, CTLA-4, CD160, CD244, LAG-3, TIM-3, TIGIT and VISTA) by flow cytometry. Purified dendritic cells were stimulated with these combinations of TLR agonists and evaluated for IL-12 release by ELISA.

Results

Activation of CD8+ T cells in the presence of specific TLR ligands resulted in decreases in expression of PD-1 and/or CD160 and increases in CTLA-4. These changes in T-cell checkpoint receptor expression were modestly affected when TLR ligands were used in combination. Similarly, stimulation of DC with several TLR agonists resulted in secretion of IL-12 that was modestly increased by combination treatment.

Conclusions

Multiple TLR agonists can modulate the expression of T-cell checkpoint receptors, notably PD-1 and CTLA-4. These data may provide rational combinations of TLR ligands that could be investigated as vaccine adjuvants, alone or in combination with T-cell checkpoint blockade, to improve the efficacy of anti-tumor vaccines.

References

1. Jin HT, Anderson AC, Tan WG, et al. Cooperation of Tim-3 and PD-1 in CD8 T-cell exhaustion during chronic viral infection. Proc Natl Acad Sci USA. 2010;107(33):14733-8.

2. Zahm CD, Colluru VT, Mcneel DG. Vaccination with High-Affinity Epitopes Impairs Antitumor Efficacy by Increasing PD-1 Expression on CD8 T Cells. Cancer Immunol Res. 2017;5(8):630-641.

3. Zahm CD, Colluru VT, Mcilwain SJ, Ong IM, Mcneel DG. TLR Stimulation during T-cell Activation Lowers PD-1 Expression on CD8 T Cells. Cancer Immunol Res. 2018;6(11):1364-1374.

Correspondence: Jae Hyung Park (jhpark1@skku.edu)

Background

Although immune checkpoint blockade therapy (ICT) contributed remarkably to elevated survival rate and durable tumor remission for the past decade, several limitations of ICT such as that only partial patients have corresponded to those therapeutic interventions have appeared. Therefore, recently, the need for combination therapies have been emerged to improve the response rate of ICT and elicit robust antitumor immunity. In the tumor microenvironment, multiple immunosuppressive immune cells (e.g. Treg, tumor-associated macrophage (TAM), MDSC) reside and they hamper antitumor immunity. Specifically, TAMs have been intensively studied to turn immunosuppressive microenvironment into an immunosupportive one, since they have plasticity in phenotypes (i.e. immunosupportive M1 macrophages and immunosuppressive M2 macrophages). In the tumor microenvironment, the dominant phenotype of TAMs are M2-like TAM, and they hinder the effector function of cytotoxic T lymphocytes (CTL) and NK cells. In this respect, we developed the polymeric nanocarrier for the targeted repolarization of M2-like TAMs toward tumoricidal M1-like TAMs to enhance the tumor targetability, reduce systemic adverse effects, and reinvigorate antitumor immunity.[1-3]

Methods

For in vitro studies, bone marrow-derived macrophages (BMDMs) were prepared as previously reported. To conduct in vivo studies, CT26-bearing mice were prepared and the tumor-infiltrating leukocytes were separated using MACS for further flow cytometric analysis.

Results

The polymeric nanocarrier (TR-NP) to target M2-like TAMs was successfully prepared by the layer-by-layer assembly, resulting in 145 nm of hydrodynamic diameter. TR-NPs targeting IL-4R alpha selectively demonstrated the enhanced uptake by M2-polarized BMDMs and effectively deliver the TLR agonists. As the main purpose of this study is to repolarize M2-like TAMs toward M1-like TAMs, we carried out in vitro flow cytometric analysis for macrophage markers. Upon treatment of TR-NPs, the expression levels of M2 markers (CD206, Arginase) decreased and the expression levels of M1 markers (iNOS, CD86, MHC II) increased, compared with untreated macrophages or free TLR agonist. Furthermore, we performed in vivo antitumor efficacy for combination therapy with PD-L1 antibody to assess the improvement of ICT. Consequently, TR-NP treatment showed an increased population of M1-like TAMs, CD8+ T cell in the tumor tissue, leading to effective combination therapy.

Conclusions

In this study, we designed the polymer nanocarrier that targets TAMs and repolarizes them toward M1-like macrophages. The TR-NPs successfully altered M2-like macrophages to anti-tumoral ones, thereby inhibiting the cancer cell proliferation. Consequently, TR-NP treatment provided immune-activating microenvironment by TAM repolarization and CTL infiltration in vivo, suggesting the potential as an immune adjuvant for combination therapy with ICT.

Acknowledgements

This work was financially supported by the Basic Science Research Programs (2018R1D1A1B07047950) of the National Research Foundation (NRF), Republic of Korea.

References

1. Pitt JM, Vetzou M, Daillere R, Roberti MP, Yamazaki T, Routy B, Lepage P, Boneca IG, Chamaillard M, Kroemer G, Zitvogel L. Resistance Mechanisms to Immune-Checkpoint Blockade in Cancer: Tumor-Intrinsic and -Extrinsic Factors. Immunity 2016;44:1255-1269.

2. van der Burg SH, Arens R, Ossendorp F, van Hall T, Melief CJ. Vaccines for established cancer: overcoming the challenges posed by immune evasion. Nat Rev Cancer. 2016;16:219-233.

3. Mantovani A, Marchesi F, Malesci A, Laghi L, Allavena P. Tumour-associated macrophages as treatment targets in oncology. Nat Rev Clin Ocol. 2017;14:399-416.

Ethics Approval

The study was approved by Sungkyunkwan University‘s Ethics Board, approval number SKKUIACUC2019-01-07-1.

Correspondence: Jeffrey Hubbell (jhubbell@uchicago.edu)

Background

Difficultly translating therapeutic vaccination success from murine models to the clinical treatment of cancer has highlighted the need for vaccination strategies that may be more broadly applicable. To address this problem, we developed an in situ vaccination strategy to adjuvant tumor cells directly, utilizing the tumor itself as the antigen source in initiating tumor-reactive cellular responses. Our engineered vaccine is comprised of a novel polymeric TLR7-agonist (pTLR7) [1] chemically linked to tumor-binding antibodies. We hypothesized that following intratumoral administration, our pTLR7-(tumor-binding antibody) conjugates (pTLR7-tAbs) would bind to surface molecules expressed by tumors, thereby increasing persistence of the adjuvant within the tumor. Given the ease of conjugation of pTLR7 to various antibodies, our in situ pTLR7-tAb vaccination platform is tunable in specificity from highly tumor-specific to widely applicable, depending on the choice of antibody used. For generalizability between tumor models, we selected an anti-CD47 antibody to create anti-CD47-pTLR7 conjugates for testing our vaccination.

Methods

Efficacy studies were conducted in poorly immunogenic orthotopic murine tumor models of EMT6 breast cancer and B16F10 melanoma. Biodistribution and intratumoral retention studies were conducted using an in vivo imaging system (IVIS). Intratumoral T cell and antigen presenting cell numbers, frequencies, and activation phenotype were assessed via flow cytometry.

Results

Tumor retention studies tracking fluorescently labeled pTLR7-tAbs showed pTLR7-tAbs are retained within the tumor microenvironment (TME) 3-fold longer than isotype control antibody-pTLR7 constructs. Interestingly, following vaccination, pTLR7-tAb treated tumors showed increased numbers and activation of intratumoral antigen presenting cells (APCs). Assessment of tumor infiltrating lymphocytes (TIL) showed pTLR7-tAb vaccination stimulated tumor antigen TRP2-specific T cell responses in the B16F10 melanoma model and converted these immunologically ‘cold’ tumors into T cell inflamed tumors, with increased numbers of CD4+ and CD8+ T cells. Strikingly, we observed complete remission (CR) in 50% of ‘cold’ immune-excluded EMT6 breast cancer tumors and delayed tumor progression of B16F10 melanoma after treatment with pTLR7-tAbs (p

Conclusions

Together, these data support the use of our pTLR7-tAb platform for in situ vaccination and as a therapeutic tool for enhancing anticancer immunity.

References

1. Wilson DS, Hirosue S, Raczy M, Bonilla-Ramirez L, Jeanbart L, Wang R, Kwissa M, Franetich JF, Broggi MAS, Diaceri G, Quaglia-Thermes X, Mazier D, Swartz MA, Hubbell JA. Antigens reversibly conjugated to a polymeric glyco-adjuvant induce protective humoral and cellular immunity. Nat Mater. 2019; 18:175–185.

Ethics Approval

All studies with animals were carried out in accordance with procedures approved by the Institutional Animal Care and Use Committee at the University of Chicago.

Correspondence: Jason Williams (drwilliams@cancerimmunebio.com)

Background

Depending on the origin of the primary lesion, patients with distant metastases of aggressive solid tumors have traditionally been deemed incurable. A complete abscopal effect, or cure, is a rare exception rather than the rule. This is secondary to cancer stem cells providing resilience to treatment and propensity to recur after remission [1]. Harnessing the potential of the immune system to recognize and destroy malignant cells harboring tumor antigens may allow stage 4 cancer to be completely eliminated from the body [2, 3]. This report describes a case of diffuse liver metastases from advanced breast cancer; eliminated immunologically by intratumoral injection of immunostimulatory pharmacotherapy. α-OX40 refers to a stimulatory monoclonal antibody with a dual mechanism of action on T cells designed to potentiate long-term immunity. Activating the OX40 protein triggers a sustained immune response in effector T cells while also potentiating the demise of regulatory T cells. Intratumoral α-OX40 was shown to work synergistically with intratumoral CpG, a toll-like receptor 9 agonist, in a mouse model of breast cancer [4]. To our knowledge, this is the first report of an α-OX40 antibody administered intratumorally, in vivo, in a human patient.

Methods

Two liver metastases were treated with combination immunotherapy suspended in montanide, an oil adjuvant. A combination of α-OX40, CpG, ipilimumab, and ketolorac was injected into two accessible, dominant, hepatic lesions under computed tomography guidance with general anesthesia. A CT scan was performed 6 weeks post-procedure, and a PET/CT was performed 12 weeks post-procedure. Tumor markers CA125, CA27.29, and CA15.3 were monitored weekly for 12 weeks post treatment.

Results

No suspicious uptake was noted anywhere in the body on the 12-week post-procedure PET/CT. Complete abrogation of all diffuse liver metastases was documented at that time. At 12 weeks post treatment tumor markers CA125 decreased by 98% to 35 from a baseline of 1767, CA27.29 decreased 93% to 112 from a baseline of 1511, and CA15.3 decreased 93% to 56 from a baseline of 821.

Conclusions

Intratumoral injection allows a much higher proportion of immunotherapy to directly interact with the tumor microenvironment compared to intravenous injection. With intratumoral dosing of multimodal immunotherapy an order of magnitude less than intravenous dosing, a patient with stage 4 breast cancer was able to achieve complete immunologic tumor elimination at 12 weeks follow-up. This durable abscopal effect remains complete 20 weeks post-treatment.

References

1. Kreso A, Dick JE. Evolution of the cancer stem cell model. Cell Stem Cell. 2014 Mar 06,;14(3):275-91.

2. Rosenberg MA, Williams J. Image guided cryoablation of cancer with intra-tumoral injection of anti-CTLA-4 and PD-1 immune check-point inhibitors. Journal for ImmunoTherapy of Cancer. 2015;3(2):P142.

3. Soule E, Bandyk M, Matteo J. Percutaneous ablative cryoimmunotherapy for micrometastaic abscopal effect: No complications. Cryobiology. 2018 06;82:22-6.

4. sagiv-Barfi I, Czerwinski DK, Levy R. In Situ Vaccination with a TLR9 Agonist and Anti-OX40 Antibody Leads to Tumor Regression and Induces Abscopal Responses in Murine Lymphoma. Blood. 2016;128(22):1847.

Consent

Written informed consent was obtained from the patient for publication of this abstract and any accompanying images. A copy of the written consent is available for review by the Editor of this journal.

Correspondence: Juha Punnonen (jpn@ascendispharma.com)

Background

Local delivery of pattern recognition receptor agonists (PRRAs) to the tumor microenvironment (TME) stimulates innate immune sensors such as toll-like receptors (TLR) and can enhance antigen uptake and presentation, induce proinflammatory immune cell recruitment, and reverse tumor-associated immunosuppression (1, 2). Local delivery of PRRAs, such as STING or TLR agonists, has shown encouraging preclinical and clinical anti-tumor benefit (3-5). However, current approaches to deliver PRRAs to the TME suffer from the lack of local retention in the TME and rapid drug clearance, thus limiting anti-tumor benefit, promoting systemic treatment-related adverse events (e.g. cytokine storm), and necessitating frequent and often impractical dosing regimens. Additionally, systemic toxicity associated with current PRRA treatments may limit combination therapies (2, 6).

Methods

Given the need for safe and effective intratumoral therapies, we developed a sustained-release TransCon™ (refers to “transient conjugation”) TLR7/8 Agonist prodrug by conjugating resiquimod to a hydrogel via a TransCon linker to provide consistent intratumoral release of unmodified resiquimod.

Results

In vivo pharmacokinetics in rodents, following local injection of TransCon TLR7/8 Agonist, showed long-term resiquimod release over several weeks with minimal systemic exposure compared to an equivalent dose of unconjugated resiquimod. Intratumoral injection of TransCon TLR7/8 Agonist also demonstrated prolonged release. Furthermore, in a syngeneic tumor model, a single intratumoral injection of TransCon TLR7/8 Agonist mediated significant tumor growth inhibition and was associated with significantly lower systemic proinflammatory cytokine induction when compared to an equivalent dose of unconjugated resiquimod. Intratumoral delivery of TransCon TLR7/8 Agonist was well tolerated and mediated tumor growth inhibition in a dose-dependent fashion. Finally, in a bilateral syngeneic tumor model, TransCon™ TLR7/8 Agonist mediated significant tumor growth inhibition in both injected and non-injected tumors as a monotherapy and when combined with systemic IL-2 treatment.

Conclusions

These data provide strong evidence that a single dose of TransCon TLR7/8 Agonist can mediate long term local release of unmodified resiquimod with minimal systemic exposure and systemic pro-inflammatory induction compared to an equivalent dose of unconjugated resiquimod. Moreover, TransConTLR7/8 has potent anti-tumor effects as a monotherapy and in combination with cytokine therapy, suggesting further evaluation as a monotherapeutic and in combination with other immunotherapies is warranted. TransCon TLR7/8 Agonist represents a potentially novel PRRA therapy class that may overcome the shortcomings of existing PRRA treatments by providing a potent anti-tumoral response and reducing adverse events related to systemic drug exposure.

References

1. Aznar MA, Tinari N, Rullan AJ, Sanchez-Paulete AR, Rodriguez-Ruiz ME, Melero I. Intratumoral Delivery of Immunotherapy-Act Locally, Think Globally. J Immunol.2017;198:31-39.

2. Marabelle A, Tselikas L, de Baere T, Houot R. Intratumoral immunotherapy: using the tumor as the remedy. Ann Oncol.2017;28:xii33-xii43.

3. Rook AH, et al. Topical resiquimod can induce disease regression and enhance T-cell effector functions in cutaneous T-cell lymphoma. Blood.2015;126:1452-1461.

4. Clark CM, Furniss M, Mackay-Wiggan JM. Basal cell carcinoma: an evidence-based treatment update. Am J Clin Dermatol.2014;15:197-216.

5. Singh M, et al. Effective innate and adaptive antimelanoma immunity through localized TLR7/8 activation. J Immunol.2014;193:4722-4731.

6. Marabelle A, et al. Starting the fight in the tumor: expert recommendations for the development of human intratumoral immunotherapy (HIT-IT). Ann Oncol.2018;29:2163-2174.

Correspondence: Enikő Tőke (eniko.toke@treosbio.com)

Background

Analysis of current data with cancer vaccines suggests that the lack of efficacy is likely due to their two primary design challenges: 1. Vaccines have little chance of destroying heterogeneous tumor cells since they rarely induce polyclonal T-cell responses; 2. Even when polyclonal T-cell responses have been successfully induced, they are often directed against tumors where the target is absent (not expressed). Recent mutated neoantigen-based vaccines (MNeoV) aim to solve this latter issue, however only about 10-20% of selected epitopes proved to induce CD8+ T-cell responses in patients. In addition, development of MNeoV for commercial use is challenging. To overcome these limitations, we developed PASCal for improved selection of peptides (epitopes) that induce T-cell responses targeted against heterogeneous tumor cells.

Methods

PASCal operates by 3 moduls: (1) validated epitope database containing 108 true HLA-epitope pairs (2) Expression frequency-based shared tumor antigen database established for 19 indications based on >96,000 tumor biopsies. (3) Validated algorithm for the identification of immunogenic peptides by the selection of personal epitopes (PEPIs) binding to multiple autologous HLA alleles.[1,2] Using PASCal, a library of 3,286 immunogenic 20mer peptides derived from 184 antigens associated with 19 cancer indications -based on 16,000 subjects’ HLA genotype (both class I&II alleles) was compiled. Personal vaccines were selected and tested for 3 HLA-genotyped metastatic cancer patients (with ovarian-, breast- and colorectal cancer). Immunogenicity of the vaccines was tested by IFN-γ ELISPOT. Signed informed consent were obtained from each patient, including allowance of publication.

Results

Personal cancer vaccines were selected to fulfill the following criteria: 12 immunogenic peptides derived from 12 different tumor-specific antigens frequently expressed in the patient’s disease type, with the expected number of expressed antigens on the patient’s tumor cells of at least 3 (by statistical estimation). CD8+ T-cell responses were induced by 97%, CD4+ T-cell responses by 85% of peptides, confirming aimed polyclonal T-cell responses. Long-lasting CD8+ T-cell responses were detected ex vivo 4.5 months, in vitro 14 months after last vaccination. Pre-existing T-cell reactivities were detected against at least 25% of vaccine antigens demonstrating their presence in the patients’ tumor, confirming the success of vaccine design strategy aiming to induce polyclonal T-cell responses against at least 3 antigens expressed by the tumor.

Conclusions

PEPIs outperform reported immunogenicity of mutated neoantigen-based personal vaccines and induced unprecedented immune responses in cancer patients. The „off-the-shelf” personalized approach with PASCal enables commercially scalable vaccine development, without need for tumor biopsy and on-demand manufacturing.

References

1. Hubbard JM, et al. J. Clin. Oncol., 37, 2019 (suppl; abstr 3557)

2. Toke ER, et al, J. Clin. Oncol., 37, 2019 (suppl; abstr e14295)

Ethics Approval

This study was performed in accordance to the Declaration of Helsinki (amendment 37). Signed informed consent were obtained from each patient, including allowance of publication.

Consent

Signed informed consent were obtained from each patient, including allowance of publication.

Correspondence: Jessica Flechtner (jessica.flechtner@genocea.com)

Background

Neoantigens are mutated tumor antigens that make attractive targets for immunotherapy as they are recognized as foreign by the immune system. While considered as the optimal strategy for targeting immune tumor killing while minimizing risks of autoimmunity, identification of true neoantigens has been elusive due to poor predictive value of in silico selection. ATLAS™ is a T cell response profiling platform that does not use predictive methods, but instead employs high-throughput ex vivo screening of putative neoantigens using autologous antigen presenting cells and T cells. Antigens are characterized as stimulatory or inhibitory by significant up- or downregulation of inflammatory T cell cytokine secretion relative to controls. Neoantigen identification using an autologous cell assay allows differentiation of desired as well as potentially unwanted antigen-specific T cell responses that are biologically-relevant to the patient.

Methods

ATLAS screening of the B16F10 melanoma mutanome was performed, consisting of whole exome sequencing, non-synonymous mutation library construction, and screening of splenic CD8+ T cells from tumor-bearing C57BL/6 mice. Stimulatory or inhibitory neoantigens that modulated secretion of IFNγ and/or TNFα were identified and corresponding synthetic long peptides were evaluated in a therapeutic B16F10 tumor challenge model. Vaccine immune responses were measured by ELISpot, flow cytometry and immunohistochemistry, and animals were monitored for tumor growth and survival.

Results

ATLAS screening of B16F10 melanoma identified 66 stimulatory and 57 inhibitory neoantigens, the majority of which were not predicted by peptide binding affinity. Strikingly, when an adjuvanted antigen cocktail containing a top stimulatory neoantigen was therapeutically administered as monotherapy into tumor-bearing mice, tumor outgrowth was either significantly delayed or completely abrogated in all mice (n=12). Conversely, therapeutic immunization with adjuvanted inhibitory neoantigen peptides resulted in no benefit and in some cases led to a significant increase in tumor growth kinetics representative of tumor hyperprogression. Stimulatory antigen vaccination induced significant antigen-specific inflammatory T cell responses, whereas T cell responses to inhibitory antigens were variable. Immunohistochemistry analysis of hyperprogressing tumors indicate reduced infiltration of CD8+ T cells.

Conclusions

Currently, the only method to reliably and comprehensively identify neoantigen targets and rationally omit potentially deleterious antigens is ATLAS screening using autologous cells. These studies demonstrate proof of concept for ATLAS-based neoantigen vaccine selection and establish a model for further mechanistic study. In a recent interim analysis, 91% of vaccine neoantigens selected as stimulatory by ATLAS, elicited T cell responses in patients participating in the GEN-009 Phase 1/2 clinical trial of a personalized cancer vaccine (NCT03633110).

Ethics Approval

Animals were housed within a USDA registered, AAALAC accredited and OLAW assured (#A4591-01) animal facility in accordance with the Institutional Animal Care and Use Committee (IACUC).

Correspondence: Douglas McNeel (dm3@medicine.wisc.edu)

Background

Androgen deprivation therapy is the primary treatment for recurrent and metastatic prostate cancer. Androgen deprivation can also modulate the immune system and cause effects such as T cell infiltration into the prostate, affect antigen processing and/or presentation and possibly affect T cell homeostasis. In previous studies, we have shown that immunizing tumor-bearing mice using a DNA vaccine against the androgen receptor significantly slowed down tumor growth compared to control. Furthermore, combining androgen deprivation with vaccination against the AR further augments the anti-tumor response compared to the AR vaccination alone. In this study, we hypothesized that the timing of DNA vaccination with respect to androgen deprivation treatment, may impact the efficacy of the anti-tumor effect.

Methods

We implanted the androgen-responsive mouse prostate tumor line MycCaP into 6 week old male FVB mice. The tumors were allowed to grow for up to 32 days, then mice were treated with androgen deprivation therapy (degarelix). Vaccines consisted of either the AR ligand binding domain within the pTVG4 backbone (pTVG-AR), or the pTVG4 backbone alone (control). Immunizations were either started one week after tumor implantation (Vx→ADT), or one day following androgen deprivation (ADT→Vx).

Results

Mice treated with AR vaccine before androgen deprivation (AR →ADT) had slower tumor growth compared to the control plasmid (pTVG4 →ADT) but not when vaccinated after ADT (ADT→Vx). Furthermore, when immunizations were given before androgen deprivation and continued after androgen deprivation, the emergence of castration resistance was further delayed.

Conclusions

These results have implications on the scheduling of androgen deprivation with immunotherapy for the treatment of prostate cancer, and correspond to similar results observed in human clinical trials with anti-tumor vaccines.

Correspondence: Kevin Pojasek (kpojasek@svhealthinvestors.com)

Background

Recent advances in immunotherapy have confirmed that adaptive immune responses can recognize and eliminate cancer cells. Past approaches using tumor-associated antigens have elicited poor immune responses and the cancer vaccine field has shifted to single nucleotide variant-derived neoantigens to avoid T-cell depletion by central tolerance. These highly personalized vaccines are beginning to show promise. We hypothesized that highly immunogenic target antigens exist within the cancer genome and become aberrantly expressed due to the epigenetic changes that accompany neoplasia. We utilized markers of endogenous retroviruses (ERVs), which are particularly abundant in repressed regions of DNA, to search for novel, high-value melanoma antigens, to circumvent the need for personalization.

Methods

We created a de novo pan-cancer transcriptome assembly with RNA-seq reads from 31 cancer types obtained from TCGA. Transcript sequences were subsequently filtered to identify those containing ERV elements. Next, they were subject to differential expression analysis selecting those expressed in the melanoma patients compared to normal tissues (utilizing GTEx data).

To discover bona fide antigens encoded by our melanoma-specific transcripts, we interrogated the ORFs from these transcripts using public and independently-generated mass spectrometry-based immunopeptidomics data from melanoma samples. HLA-bound peptides from these analyses matching the known proteome were removed, and those peptides solely mapping to our predicted ORF sequences were used to identify novel transcript-derived antigens specifically expressed and presented in primary melanoma tissue.

Confirmation of tumor cell-specific expression of antigen-encoding transcripts was carried out using RNAScope®. The ability of detected epitopes from our discovered antigens to elicit an adaptive immune response was assessed by characterization of antigen-specific T-cell responses from naïve donors.

Results

Our de novo assembly revealed the presence of approximately 100 melanoma-specific transcripts encoding over 2,000 potential antigens (ORFs). Interrogation of these ORFs against MS immunopeptidomic datasets mapped HLA-bound peptides to dozens of ORFs, demonstrating presentation of our novel antigens in multiple melanoma patient tissues.

RNAScope® revealed melanoma-specificity at the transcript level, with little to no transcript expression identified across normal tissues. Assessment of immunogenicity in naïve subject T cells revealed strongly reactive T cells that were able to kill peptide-pulsed APCs, indicating a lack of central-tolerance deletion of T cells specific for these ERV-derived peptides.

Conclusions

We have identified a number of novel melanoma-specific antigens that are shared among patients. T cells reactive for these antigens can be detected in naïve subjects, and thus these antigens show promise as candidates for development of off-the-shelf cancer vaccine-based immunotherapies.

Acknowledgements

The results presented here are based in part upon data generated by The Cancer Genome Atlas (TCGA) Research Network (http://cancergenome.nih.gov/); and the Genotype-Tissue Expression (GTEx) Project (supported by the Common Fund of the Office of the Director of the National Institutes of Health, and by NCI, NHGRI, NHLBI, NIDA, NIMH, and NINDS). Primary patient material utilized in this study was provided by the Cambridge University Hospitals NHS Trust Human Research Tissue Bank.

Ethics Approval

All work involving the use of human tissue was approved by the NHS Health Research Authority South Central - Oxford A Research Ethics Committee (reference number 04/Q1604/21), and North West - Haydock Research Ethics Committee (reference number 19/NW/0216).

Correspondence: Gary Kohanbash (gary.kohanbash2@chp.edu)

Background

Diffuse intrinsic pontine glioma (DIPG) is one of the most lethal pediatric brain tumors. DIPG is insensitive to chemotherapy and surgically inaccessible, creating an urgent need for novel therapeutic approaches. We have been evaluating peptide vaccine immunotherapies that target glioma-associated antigens (GAAs). Enhancing the expression of these immunogenic GAAs and MHC I on tumor cells may promote immune-mediated tumor recognition and killing following peptide vaccine immunotherapy. DNA methyltransferase (DNMT) inhibitors have been shown to augment the expression of MHC, tumor antigens, and other immunosensitizing molecules. Guadecitabine (SGI-110), a next-generation DNMT inhibitor, has been developed to prolong tumor cell exposure to its active metabolite, decitabine. In this study, we investigated whether SGI-110 can immunosensitize glioma cells to peptide vaccine immunotherapy by enhancing their surface expression of MHC I and a GAA, EphA2. [1-5]

Methods

We developed a novel C57BL/6-syngeneic DIPG model by culturing cells from a Sleeping Beauty de novo DIPG induced in a neonatal mouse using a K27M-mutated histone 3.3 plasmid and other oncogenic plasmids. Murine SB-DIPG-11 and human DIPG cell lines were treated with SGI-110 for 5 days. For vaccination, mice were intradermally injected with 100μg each of HBV core128-140 (MHC-II, lab restricted) and the GAA peptides hgp10025-33, mEphA2671-679, and mEphA2682-689, emulsified in 40μg of poly(I:C) and IFA. Control mice received HBV core128-140, also emulsified in poly(I:C) and IFA. Five days after tumor injection, mice received 2.0 mg/kg/day of SGI-110 for 6 days.

Results

Flow cytometry analysis showed that SB-DIPG-11 cells express both MHC I (H-2Kb/H-2Db) and EphA2 on their surface. In vitro, treatment of SB-DIPG-11 or C57BL/6-syngeneic GL261 cells with SGI-110 resulted in a dose-dependent increase of MHC I and EphA2 surface expression. In vivo, subcutaneous administration of SGI-110 in combination with an EphA2-targeted peptide vaccine significantly prolonged the survival of mice bearing orthotopic DIPG-like tumors compared with the control treatment. Stable EphA2 knockdown reverted the survival benefit of SGI-110 in combination with peptide vaccine. Analysis of dissociated tumor tissue by flow cytometry showed elevated levels of MHC I and EphA2 in mice receiving SGI-110 or SGI-110 and peptide vaccine.

Conclusions

Compared to control, combination therapy proved the most effective, significantly prolonging the survival of tumor-bearing mice. Based on these data, we have begun evaluating whether subcutaneous or intracerebroventricular administration of the SGI-110 is more effective in combination with peptide vaccine. Overall, SGI-110 may sensitize children with DIPG to peptide vaccine immunotherapy.

Acknowledgements

his research was supported by funding from the Pediatric Brain Tumor Foundation’s Project “All In” for DIPG, St. Baldrick’s Foundation, the National Cancer Institute, the National Institute for Biomedical Imaging and Bioengineering, the Pittsburgh Foundation’s Copeland Fund, and UPMC Children’s Hospital of Pittsburgh’s Scientific Program.

References

1. Hargrave, D., Bartels, U., Bouffet, E. Diffuse brainstem glioma in children: critical review of clinical trials. Lancet Oncol 2006; 7: 241-248.

2. Long, W., Yi, Y., Chen, S., Cao, Q., Zhao, W., Liu, Q. Potential New Therapies for Pediatric Diffuse Intrinsic Pontine Glioma. Front Pharmacol

2017; 8: 1-13.

3. Griffiths, E.A., Choy, G., Redkar, S., Taverna, P., Azab, M., Karpf, A.R. SGI-110: DNA Methyltransferase Inhibitor Oncolytic. Drugs Future 2013;

38: 535-543.

4. Srivastava, P., et al. Immunomodulatory action of the DNA methyltransferase inhibitor SGI-110 in epithelial ovarian cancer cells and xenografts.

Epigenetics 2015; 10: 237-246.

5. Wiesner, S.M., et al. De Novo Induction of Genetically Engineered Brain Tumors In Mice Using Plasmid DNA. Cancer Res 2009; 69: 431-439.

Ethics Approval

The study was approved by University of Pittsburgh Ethics Board.

Correspondence: Piotr Stepniak (piotr.stepniak@ardigen.com)

Background

The strong involvement of neoantigens-driven CD8+ T-cells is considered a substantial element of immune-mediated tumor rejection. Neoantigens are, in fact, not subject to central immune tolerance and may constitute ideal agents for eliciting tumor-specific immune responses. For these reasons, neoantigens are rapidly gaining interest [1] as central components of personalized cancer vaccines (PCVs), as shown by the number of PCV clinical trials. One of the challenges in the design of PCVs is that only a small fraction of potential neoantigens arising from somatic mutations are immunogenic [2]. Several aspects including highly individual antigenic landscape, restricted number of targetable neoantigens per tumor, complex tumor subclonal structure, private neoantigen-specific T-cell repertoire, presence of inhibitory neoantigens, immune-escape mechanisms, etc. make the prediction of effective neoantigens a particularly challenging task.

Standard methods for the selection of neoantigens are typically based on peptide-HLA (pHLA) binding prediction algorithms. Advances in pHLA elution and mass spectroscopy [3] make it possible to identify pHLAs presented on the cell surface, which drives the development of algorithms for the prediction of the natural neoantigen presentation [4, 5]. Combining aspects such as pHLA binding affinity, neoepitope presentation, gene expression, etc. for the prediction of neoepitopes’ immunogenicity is essential for the design of therapeutic cancer vaccines and adoptive cell therapies.

Methods

We perform a benchmarking study of the available methods for vaccine design. To this end, we carry out a retrospective analysis on selected datasets [2] comprising WES and RNA-seq measurements, as well as validation of CD8+ response. We benchmark our approach - ArdImmuneVax - against other available methods.

Results

For each of the vaccine design tools, we assess the enrichment of immunogenic neoepitopes within the suggested vaccine compositions (e.g., top 20 peptides) using early-enrichment metrics [4]. We analyze the results in an indication-specific fashion in order to assess whether there are significant differences among cancer types. Finally, for each vaccine design method, we assess additional important properties of the resulting target neoepitopes, such as toxicity, tolerance, and peptides’ manufacturability.

Conclusions

The presented work provides insights into the vaccine design process by quantitatively and qualitatively comparing available tools for the selection of both cancer vaccine composition and targets for adoptive cell therapies.

References

1. Garcia-Garijo A, Fajardo CA, Gros A, Determinants for Neoantigen Identification. Front. Immunol. 2019;10:1392.

2. Parkhurst MR, et al., Unique Neoantigens Arise from Somatic Mutations in Patients with Gastrointestinal Cancers. Cancer Discov. 2019;9:1-14

3. Abelin JG, et al., Mass Spectrometry Profiling of HLA-Associated Peptidomes in Mono-allelic Cells Enables More Accurate Epitope Prediction. Immunity. 2017;46(2):315-326

4. Bulik-Sullivan B, et al., Deep learning using tumor HLA peptide mass spectrometry datasets improves neoantigen identification. Nat. Biotechnol. 2019;37:55-63

5. Zeng H, Gifford DK, DeepLigand: accurate prediction of MHC class I ligands using peptide embedding. Bioinformatics. 2019;35:i278-i283.

Correspondence: Craig Slingluff (CLS8H@hscmail.mcc.virginia.edu)

Background

Immunogenicity of cancer vaccines is impacted by adjuvants and schedule, but systematic assessments of their effects have not been performed. Montanide ISA-51, an incomplete Freund’s adjuvant (IFA), is used in many vaccine trials, but concerns have been raised about negative effects in murine studies [1, 2]. However, we found in humans that IFA enhances systemic immune responses over a toll-like receptor (TLR) agonist alone [3] and that repeat vaccination at one site (same site vaccination, SSV) creates tertiary lymphoid structures (TLS) in the vaccine site microenvironment (VSME) [4]. We hypothesized that one vaccine with peptides +IFA +pICLC or SSV x 3 with peptides in IFA will create an immunogenic milieu locally at the VSME, with activated and mature dendritic cells (DC), TLS-associated chemokines, and a Th1-dominant VSME.

Methods

Biopsies of the VSME were obtained from participants on two clinical trials (MEL48, NCT00705640; MEL58, NCT01585350) who were immunized with multiple melanoma peptides (MELITAC 12.1) in adjuvants comprised of IFA and/or the toll-like receptor (TLR)3-agonist pICLC. Biopsies were obtained either a week after one vaccine or a week after SSV x 3. Controls included normal skin and skin injected with IFA without peptides. Gene expression analysis was performed by RNAseq. Multiparameter immunofluorescence histology (mIF) was performed for selected markers.

Results

VSME samples were evaluated from 27 patients. One vaccine with peptides in pICLC + IFA enhanced expression of CD80, CD83, CD86 (p<0.01), CD40 and CD40L (p<0.0001) over normal skin; these effects were significantly enhanced for SSV with peptides + IFA (Figure 1). CD40L protein expression by CD4 T cells was confirmed by mIF (Figure 2). Vaccines containing pICLC increased expression of TBX21 (T-bet) but did not decrease GATA3 over normal skin, whereas SSV with peptides in IFA dramatically enhanced TBX21 and decreased GATA3, with high expression of IFNγ and STAT1 (Figure 3). SSV with peptides in IFA also dramatically reduced arginase-1 (ARG1) expression and enhanced expression of TLR adapter molecules TICAM-1 (TRIF) and MYD88 (Figure 4). Furthermore, SSV with IFA and peptides also enhanced expression of chemokines associated with TLS formation (not shown).

Conclusions

These findings suggest that SSV with peptides in IFA enhances CD40L expression by CD4 T cells, supports a Th1 microenvironment, with accumulation of activated and mature DC. Increased expression of TLR adaptor proteins after SSV with peptides in IFA might implicate effects of the skin microbiome. Reduced ARG1 may reflect diminished suppressive myeloid activity in the VSME.

Trial Registration

Samples from two clinical trials were used, including MEL48, NCT00705640; MEL58, NCT01585350.

References

1. Hailemichael, Y., et al., Cancer vaccine formulation dictates synergy with CTLA-4 and PD-L1 checkpoint blockade therapy. J Clin Invest, 2018. 128(4): p. 1338-1354.

2. Hailemichael, Y., et al., Persistent antigen at vaccination sites induces tumor-specific CD8(+) T cell sequestration, dysfunction and deletion. Nat. Med, 2013. 19(4): p. 465-472.

3. Melssen, M.M., et al., A multipeptide vaccine plus toll-like receptor agonists LPS or polyICLC in combination with incomplete Freund's adjuvant in melanoma patients. J Immunother Cancer, 2019. 7(1): p. 163.

4. Harris, R.C., et al., The vaccine-site microenvironment induced by injection of incomplete Freund's adjuvant, with or without melanoma peptides. J Immunother, 2012. 35(1): p. 78-88.

Ethics Approval

Patients were studied following informed consent, and with Institutional Board Review (IRB) (HSR-IRB 13498 and 15781, respectively) and FDA approval.

Consent

Written informed consent was obtained from patients for publication of this abstract and any accompanying images. A copy of the written consent is available for review by the Editor of this journal

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P683).

See text for description

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2 (abstract P683).

See text for description

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3 (abstract P683).

See text for description

• Download figure
• Open in new tab
• Download powerpoint

Fig. 4 (abstract P683).

See text for description

Correspondence: Gad Berdugo (gberdugo@epivaxonco.com)

Background

Recent advances in whole exome sequencing have produced a renewed interest in therapeutic cancer vaccines targeting tumor-specific neoantigens. A central challenge to the development of successful therapeutic cancer vaccines is the identification of neoantigen sequences that elicit effector T cell responses able to control tumor growth while avoiding the induction of inhibitory immune pathways. In the study reported here, we present data highlighting the efficiency of neoantigen selection using the innovative Ancer™ screening platform and demonstrate antigen-specific responses and tumor control in a mouse tumor model.

Methods

The CT26 mutanome was evaluated with Ancer™, an innovative and automated neoantigen prediction platform that combines proprietary machine learning-based MHC class I and MHC class II neo-epitope identification tools with removal of inhibitory regulatory T cell (Treg) epitopes for optimal personalized cancer vaccine design. MHC class I- and MHC class II-restricted CT26 neoantigens, devoid of putative Treg epitopes, were ranked according to their immunogenic potential and tumor expression level. The immunogenicity and efficacy of the 20 highest Ancer-ranked neoantigens, delivered subcutaneously as a peptide pool with poly-ICLC (Oncovir) as adjuvant was evaluated in naïve and CT26 tumor-bearing BALB/c mice.

Results

T cell responses to individual peptide neoantigens were determined using ELISpot assay for IFNγ production. While around 20% of neoantigen sequences identified in studies using commonly available predictive algorithms are immunogenic, we found that up to 65% (13 out of 20) of the computationally predicted peptide sequences included in the CT26-Ancer™ vaccine were immunogenic (Figure 1). Responding T cells were both CD8+ and CD4+ and were shown to be multi-functional by flow cytometric analysis for IFNγ, TNFα and IL-2 production.

Tumor control was demonstrated by immunizing BALB/c mice implanted with syngeneic CT26 tumor cells with the CT26-Ancer™ vaccine. Average tumor volumes in CT26-Ancer™ immunized mice were significantly reduced by 45% (unpaired t test, p = 0.0156) and 38% (unpaired t test, p = 0.0291) at days 21 and 25 post implantation, respectively, compared to vehicle (poly-ICLC) immunized tumor-bearing animals (Figure 2).

Conclusions

The results presented here highlight the efficiency of the Ancer™ platform in selecting highly immunogenic neoantigen targets and that these responses are able to control tumor growth in a peptide formulation. Using advanced in silico methods to select antigenic targets for use in therapeutic cancer vaccines promises to improve the efficacy of tumor immunotherapy. While confined to mice, these studies further validate our personalized cancer vaccine process for future human clinical studies.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P684).

See text for description

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2 (abstract P684).

See text for description

Correspondence: Thomas Paul (Thomas.A.Paul@pfizer.com)

Background

A central challenge to elicit effective anti-tumor T cell responses remains limited expression of productive tumor antigens or antigen presentation machinery. During immune editing, cancers utilize epigenetic silencing of cancer neo-antigens as well as components of the antigen presentation machinery to escape immunosurveillance[1]. As a result, therapeutic approaches that have the potential to reactive epigenetically silenced genes that enhance tumor immunogenicity are attractive drug combinations with checkpoint blockade[2].

EZH2 is an epigenetic modifier responsible for H3K27 histone methylation associated with the silencing of gene transcription. In cancer, dysregulation of EZH2 results in the redistribution of H3K27me3 to the promoters of genes involved in the regulation of cell cycle, cell survival, and cellular differentiation. Emerging studies have provided evidence for an additional role of EZH2 dysregulation in the transcriptional repression of genes involved in tumor immune pathways.

Methods

ChIP-seq, RNA-seq, and exome-seq was conducted on murine tumor cell lines treated with the EZH2 inhibitor PF-06821497. Peptides with mutations that are upregulated by EZH2 inhibitor treatment were computationally selected based on their predicted binding affinity to the corresponding MHC class I and class II expressed in each cell line. Presentation of peptide candidates were further validated by immunopeptidomics. Predicted peptides were synthesized and tested in pools for their ability to elicit antigen-specific T cell responses.

Results

In this study, we have identified a common mechanism of EZH2-mediated epigenetic suppression of genes involved in tumor immune recognition. Using epigenomic profiling, we demonstrate the involvement of EZH2 in the repression of genes involved in tumor antigen presentation including MHC class I and class II and numerous genes harboring mutations with potential to serve as neoantigens. Inhibition of EZH2 combined with IFN stimulation results in reduced H3K27me3 and transcriptional up-regulation of the MHC class II transactivator-CIITA and MHC class II genes. Moreover, by employing computational prediction tools, we observed a cluster of genes with somatic mutations that are transcriptionally re-activated by EZH2 inhibition. The re-expression of these genes is predicted to represent over 100 potential neoantigens with high binding affinity to MHC class I and class II molecules.

Conclusions

Our data illustrate the potential for targeting EZH2 to increase the immunogenicity of tumors through relieving epigenetic suppression of tumor antigen presentation machinery and silenced neoantigens. As EZH2 inhibitors are currently under clinical investigation, targeting EZH2 represents a promising combination approach with checkpoint inhibitors to elicit responses in poorly immunogenic tumors.

References

1. Schreiber RD, Old LJ, Smyth MJ. Cancer immunoediting: integrating immunity's roles in cancer suppression and promotion. Science. 2011; 331:1565-70.

2. Havel JJ, Chowell D, Chan TA. The evolving landscape of biomarkers for checkpoint inhibitor immunotherapy. Nat Rev Cancer. 2019; 19(3):133-150.

Correspondence: Shizuko Sei (seis@mail.nih.gov)

Background

The NCI PREVENT Cancer Preclinical Drug Development Program (PREVENT) is a peer-reviewed program that supports the preclinical development of innovative cancer preventive interventions towards clinical applications (https://prevention.cancer.gov/major-programs/prevent-cancer-preclinical). One of the Program’s current focus areas is cancer immunoprevention. Recent advances in cancer immunotherapies have clearly demonstrated that the immune system can mount effective antitumor immune responses, if tumor-associated immunosuppression is overcome, for example, by immune checkpoint blockade. It is conceivable that cancer vaccines can arm the host with antitumor immunity that intercepts, arrests, and/or reverses tumorigenesis if delivered before or early in the tumorigenic process. PREVENT seeks to develop effective vaccines for precision cancer immunoprevention in high-risk groups in partnership with principal investigators (PIs) with innovative ideas.

Methods

PREVENT utilizes NCI contract resources to operate its agent development pipeline, which consists of three key stages: early proof of concept phase, secondary testing phase, and advanced preclinical development phase. Competitively selected contractor pools with necessary technical expertise are in place to implement approved projects under NCI oversight. Data and materials generated through PREVENT are returned to the applicant PIs for further agent development. A wide array of organ-specific preclinical cancer models is available to assess agent efficacy and toxicity.

Results

Since the Program’s inception, a total of 21 cancer preventive vaccine projects have been selected for support. Of those, two have progressed to the clinical trial stage with two additional projects in the advanced development phase. The PREVENT immunoprevention portfolio consists of a variety of vaccination approaches that target oncogenic drivers, early tumor associated antigens, and infectious agents known to induce cancers. These vaccines are developed for clinically identifiable high-risk cohorts, such as genetically predisposed populations (individuals with Lynch syndrome, BRCA1/2 mutations, and familial adenomatous polyposis, etc.) and those diagnosed with pre-cancerous lesions.

Conclusions

While the expanded knowledge base in tumor immunology has uncovered the potential hurdles for developing effective cancer immunoprevention strategies, it has also helped better understand the key determinants, including the selection of optimal target antigens, vaccine delivery strategies and schedules, immune biomarkers for vaccine immunogenicity, immune durability, and immune correlates of antitumor protection. Learning from and building on the success of prophylactic human papillomavirus vaccines in preventing cervical cancer, PREVENT offers partnership opportunities to researchers interested in developing novel cancer preventive vaccines for oncogenic pathogen-induced as well as cancers not linked to pathogens toward clinical applications.

Correspondence: Satwinder K. Singh (s.singh@dcprime.com)

Background

Induction of optimal anti-tumour immune response is challenging as tumour antigens are self-antigens and antigen-specific receptors against self-antigens are negatively selected in the body. Therefore, anti-tumour immune responses that are present in the body harbour low affinity antigen-specific receptors that will not optimally eradicate tumours. The new insights on tumour-specific antigens (TSA) has led to personalised vaccines. However, not every tumour type has high mutation burden resulting in TSA and thus these tumour types cannot rely on these strategies. Also TSA are patient-specific and the broad application in cancer patients is limited. In this study we present a novel vaccination strategy which uses foreign (non-self) antigens to mount robust immune responses enabling solid tumour eradication. This strategy is based on two key elements: to induce immunity against a foreign antigen, and to mark the tumour as a target for the induced immune response with the same foreign antigen. To proof the concept, we selected keyhole limpet hemocyanin (KLH) as a foreign immunogenic neo-antigen which will serve both as vaccine composition and tumour marker to elicit effective anti-tumour immune response.

Methods

As solid tumour models, A375 melanoma or U87-MG glioblastoma cells were subcutaneously engrafted into CD34 humanised NCG. Mice received intraperitoneal (i.p.) vaccination with KLH. The tumour marking was done intratumourally (i.t.) by using DCOne® mDC loaded with KLH protein. Tumour reduction and immune responses induced against KLH were evaluated.

Results

We observed reduced tumour growth in both A375 and U87-MG tumour models in KLH vaccinated mice where KLH served as vaccine and i.t. injection of DCOne® mDC loaded with KLH protein for tumour marking compared to the mice injected with PBS. We observed that both in U87-MG and A375 mice, there was no significant difference regarding anti-KLH IgM concentration between treated groups and the PBS control group over time. Noteworthy, the mice from the KLH/KLH loaded DCOne® treated group produced significantly more anti-KLH IgG than the PBS control group at sacrifice.

Conclusions

These findings support the new concept that tumour independent antigens can be used to develop vaccines against solid tumours. Importantly, DCOne® cells can be used as a carrier for intratumoural delivery of foreign proteins.

Correspondence: Teri Heiland (THeiland@immunomix.com)

Background

Cancer vaccines have traditionally succeeded in preventing viral-induced malignancies, such as papilloma virus-associated cervical cancer and hepatitis B-associated hepatocellular carcinoma. Conceptually, tumor-specific protein coding mutations (neoantigens) are ideal targets for cancer immunotherapy. Neoantigens are not expressed in healthy tissues and therefore are exempt from central tolerance and can potentially be recognized by T cells to facilitate tumor rejection. However, every patient’s tumor possesses a unique set of mutations that must first be screened for optimal target selection, presenting a challenge for neoantigen vaccine development. We present here the merging of two innovative and complementary approaches for identifying neoantigens and delivering personalized cancer vaccines.

Methods

The CT26 mutanome was evaluated with Ancer™, an innovative and automated neoepitope prediction platform that combines proprietary machine learning-based MHC class I and MHC class II neoepitope identification tools with removal of inhibitory regulatory T cell (Treg) epitopes for optimal personalized cancer vaccine design. MHC class I- and MHC class II-restricted CT26 neoepitopes, devoid of putative Treg epitopes, were ranked according to their immunogenic potential and tumor expression level. The twenty most highly Ancer-ranked neoepitopes were subsequently introduced as a “string of beads” DNA vaccine into the UNITE (UNiversal Intracellular Targeted Expression) platform (poly-neoepitope UNITE vaccine). The UNITE platform is based in part on lysosomal targeting technology which results in enhanced antigen presentation and a balanced T cell response.

Results

We report that prophylactic vaccination with poly-neoepitope UNITE vaccine successfully induced IFNγ-producing Th1 cells, with complete rejection of CT26 tumors observed in 50% of mice. Mice rejecting tumors were also protected from rechallenge with CT26, demonstrating that effective antigen-specific memory was induced in these animals. In therapeutic vaccination studies, we observed CT26 tumor growth inhibition in 46% of animals immunized with the poly-neoepitope UNITE vaccine, as well as significantly prolonged survival, as compared to animals immunized with the control vector.

Conclusions

Therefore, targeting multiple mutations encoding the best set of CD4 and CD8 neoepitopes, as predicted by Ancer™, and using the UNITE platform may solve critical problems in current cancer immunotherapy development. Follow-up studies will include evaluation of the poly-neoepitope UNITE vaccine in combination with checkpoint inhibitors.

Correspondence: Peter DeMuth (Pete.DeMuth@Elicio.com)

Background

Subunit vaccines targeting tumor antigens have shown limited capacity for expanding cytotoxic T-cells against tumors in the clinic. Especially in the case of KRAS-driven tumors, responses elicited by conventional vaccines have been exceedingly weak. For molecular immunogens including peptides and oligonucleotides, inefficient delivery to immune cells residing in the lymphatics is a significant challenge limiting their ability to induce cancer-directed immune responses of sufficient strength and functionality to impact tumors. Improving the targeting of immunogens to lymph nodes (LN), where resident immune cells potently orchestrate immunity, can substantially amplify their ability to induce effective tumor-directed immunity. Here, we demonstrate such an approach for significantly enhancing mKRAS-directed T-cell responses by precisely targeting antigens and adjuvants directly to the draining LN through a simple one-step conjugation to albumin-binding lipids. These amphiphilic conjugates (“Amphiphiles”) then ‘hitch-hike’ on albumin into the LNs where they elicit strong immune responses. LN accumulation of structurally optimized amphiphiles in mice is greatly improved over soluble equivalents.

Methods

C57BL/6J mice received two or more doses of benchmark or amphiphile-modified vaccines, comprised of mKRAS peptide and CpG adjuvant, subcutaneously injected into the tail base in two-week intervals. Immunological readouts were performed 7 days post dosing. For ELISpot analysis of IFNγ and Granzyme B production and flowcytometric bead array analysis of Th1/2 cytokines, splenocytes were harvested and re-stimulated with antigen overnight. In vivo, cytolytic capabilities of antigen-specific T-cells were evaluated by pulsing CFSE-stained splenocytes from naïve mice with mKRAS antigen and injecting these cells intravenously into immunized mice. Recovery of CFSE-labeled target cells from immunized mice was performed 24h later and analyzed flowcytometrically.

Results

We show robust immune responses that yield strong activation against all common mutations in the mKRAS protein compared to low or undetectable responses generated by soluble or benchmark treatments. Further, this response is composed of CD4+ as well as CD8+ T-cells resulting in the production of high levels of TH1-associated cytokines upon re-stimulation with mKRAS-specific peptides in vitro. In vivo, robust cytolytic function towards mKRAS-presenting targets can be measured in T-cells.

Conclusions

By targeting immunogens directly and precisely to the LNs, the Amphiphile platform can significantly amplify the potency of subunit vaccines. In the case of mKRAS, substantially improved cytolytic immune responses represent a promising therapeutic strategy for targeting mKRAS-driven tumor growth and survival in a large fraction of human tumors. Furthermore, this platform technology is simple, rapid and scalable for broad clinical application.

Correspondence: Liang Deng (dengl@mskcc.org)

Background

Modified vaccinia virus Ankara (MVA) is a highly attenuated vaccinia strain that belongs to the poxvirus family. MVA is a promising vaccine vector for infectious diseases and cancers due to its immunogenicity, safety, and a large capacity to accommodate insertions of transgenes. However, MVA has certain disadvantages that need to be overcome to improve its immunogenicity and vaccine efficacy. Those include suboptimal induction of type I IFN in dendritic cell (DC) subsets, presumably due to the expression of a viral inhibitory protein(s) for the cGAS/STING-mediated cytosolic DNA-sensing pathway, as well as the induction of cell death in infected DCs [1,2].

Methods

To identify vaccinia viral inhibitors of the cGAS/STING pathway, we performed an unbiased screen of 70 vaccinia viral early genes in HEK293 T cells using a dual-luciferase system. We discovered that the vaccinia E5R gene encodes a dominant inhibitor of cGAS.

Results

Whereas MVA infection of bone marrow-derived DCs (BMDCs) induces cGAS degradation in a proteasome-dependent manner, recombinant MVA with deletion of the E5R gene (MVA∆E5R) infection fails to do so. MVA∆E5R infection of BMDCs induces much higher levels of cyclic GMP-AMP (cGAMP) production, IFNB gene expression, and IFN-beta protein secretion compared with live MVA or Heat-inactivated MVA. In addition to BMDCs, MVA∆E5R infection of plasmacytoid DCs (pDCs) and CD103+ DCs also activates the cGAS/STING/IFNB pathway very efficiently. Furthermore, MVA∆E5R infection of DCs triggers DC maturation and induces antigen cross-presentation by DCs in a cGAS-dependent manner. MVA∆E5R-infected DCs are much less prone to cell death compared with MVA-infected DCs, and they appear activated and healthy for more than five days in vitro cell culture. Using MVA∆E5R virus expressing mcherry under vaccinia synthetic early-late promoter, infected DCs appear to have robust transgene (mcherry) expression. We engineered a recombinant MVA∆E5R expressing chicken ovalbumin (OVA) as a model antigen. Vaccination with MVA∆E5R-OVA either through skin scarification or via intradermal injection generates stronger OVA-specific CD8+ T cell responses compared with MVA-OVA.

Conclusions

In this study, we discovered that the vaccinia E5R gene encodes a dominant cGAS inhibitor. Expression of E5 protein by MVA leads to proteasome-dependent degradation of cGAS. Deletion of the E5R gene from the MVA genome dramatically improves the immunogenicity of MVA, likely through stronger type I IFN induction, more robust DC maturation, and marked reduction of cell death compared with the parental virus.

References

1. Dai, P., W. Wang, H. Cao, F. Avogadri, L. Dai, I. Drexler, J. A. Joyce, X. D. Li, Z. Chen, T. Merghoub, S. Shuman, and L. Deng. (2014). Modified vaccinia virus Ankara triggers type I IFN production in murine conventional dendritic cells via a cGAS/STING-mediated cytosolic DNA-sensing pathway. PLoS Pathogen 10: e1003989. (PMCID: 3990710)

2. Dai, P., W. Wang, N. Yang, C. Serna-Tamayo, J.M. Ricca, D. Zamarin, S. Shuman, T. Merghoub, J.D. Wolchok, and L. Deng. (2017). Intratumoral delivery of inactivated modified vaccinia virus Ankara (iMVA) induces systemic antitumor immunity via STING and Batf3-dependent DCs. Science Immunology 2, eaal1713: 1-15.(PMCID: PMC5559204)

Ethics Approval

The study was approved by the Committee on the Ethics of Animal Experiments of Sloan Kettering Cancer Institute, approval number 19-01-002.

Correspondence: Loise Francisco-Anderson (loise@evelobio.com)

Background

Systemic immunity is regulated by interactions of commensal bacteria with immune cells in the gut. Enrichment of specific intestinal microbes has been shown to enhance the anti-tumor response to PD-1 blockade in both murine models and cancer patients. Here we report that oral administration of a monoclonal microbial, EDP1503, induces systemic anti-tumor immunity via the direct activation of antigen presenting cells (APCs) and natural killer (NK) cells with subsequent induction of T cell immunity.

Methods

The clinical efficacy of EDP1503 was investigated in subcutaneously implanted isograft tumor models as well an intravenous lung metastasis model. Mechanism of action was dissected by ex vivo analysis of the tumor microenvironment (TME) and gut-draining lymph nodes, and further interrogated by in vitro functional studies with murine and human cells.

Results

EDP1503 increases expression of costimulatory molecules on CD11c+ dendritic cells (DCs) within the mesenteric LNs with an accompanying increase in proinflammatory CD103+ DCs within tumor draining lymph nodes. Direct stimulation of DCs with EDP1503 results in the upregulation of the proinflammatory cytokines TNFa, IL-1B and IL-12. In addition, EDP1503 amplifies both myeloid and lymphocyte responses via production of DC-derived growth factors, M1 polarization of macrophages, and production of the lymphocyte-recruiting chemokines, CXCL9 and CXCL10. Mechanistically, EDP1503 triggers specific pattern recognition receptors, resulting in a proinflammatory signature from APCs and enhanced cytolytic activity by NK cells. Dissection of the tumor microenvironment reveals increased activated Ki67+ NK cells and CD8+ T cells producing IFNg. Moreover, in vivo, treatment of mice with EDP1503 results in decreased tumor volume, delayed tumor growth and reduction in the number of metastatic foci in a CD8T and NK dependent manner. The prominent anti-tumor effects of EDP1503 are further augmented by combination with anti-PD-1 neutralizing Abs.

Conclusions

Together, these data clearly demonstrate the ability of an orally delivered non-colonizing monoclonal microbe to enhance innate and adaptive anti-tumor immunity and substantiates the rationale for ongoing clinical trials. EDP1503 is currently in Phase 1b/2 studies (NCT03775850; NCT03595683) with enrollment open at multiple sites.

Correspondence: Pawel Kalinski (Pawel.Kalinski@roswellpark.org)

Background

Bacillus Calmette-Guérin (BCG) is commonly used in the treatment of non-muscle invasive urothelial cancer (bladder cancer; BlCa), but its benefit is typically transient and limited to a fraction of patients. We observed that the impact of BCG on human BlCa tissue involves undesirable induction of immunosuppressive factors and regulatory T(reg) cell- and myeloid-derived suppressor cell (MDSC)-attracting chemokines. Prompted by these observations, we evaluated the molecular pathways regulating the patterns of BCG-induced inflammation in BlCa tumor microenvironments (TME), and the feasibility of their modulation to selectively enhance local attraction and function of the desirable cytotoxic CD8+ T cells (CTLs) and counteract local MDSCs and Treg accumulation.

Methods

Human monocyte-derived macrophages and freshly resected human bladder cancer specimens were cultured ex vivo in the absence or presence of BCG, indomethacin (COX-1/2 blocker), celecoxib (selective COX-2 blocker) or selective EP4 blocker (ARY-007). The expression of inflammatory mediators and chemokine genes, secretion of chemokines and their cellular sources were analyzed by quantitative PCR, ELISAs and immunofluorescence microscopy. Migration of CTLs was evaluated ex vivo, using 24-transwell plates. Data were analyzed by Student’s t-test or Wilcoxon test. Spearman’s correlation is used for correlative analysis by using z-score downloaded from cBioportal.

Results

We observed that the BCG-driven-induction of MDSC- and Treg-attracting chemokines in human macrophages and whole BlCa explants was correlated with its induction of COX-2, the key enzyme controlling the PGE2 biosynthesis. Blockade of COX-2 activity using selective or non-selective COX-2 inhibitors (celecoxib, indomethacin) or interference with EP4-driven PGE2 signaling (using selective EP4 blocker, ARY-007) strongly elevated the production of CTL attractants (CXCL9, CXCL10 and CCL5), but suppressed the induction of MDSC/Treg attractants (CCL22, CXCL8, CXCL12), and several immunosuppressive mediators including COX-2, IL-10, NOS2 and IDO. Our functional studies demonstrated that the combination of BCG with PGE2 antagonism selectively enhances CTL attraction.

Conclusions

We observed that PGE2 represents a nodal point controlling the balance between desirable/immunostimulatory and undesirable/immunosuppressive components of BCG activity and that interference with PGE2 synthesis or it’s signaling improves the magnitude and character of the BCG-orchestrated immune response in BlCa TME. Our data suggest that PGE2 antagonism may be used to enhance the therapeutic effects of BCG treatment.

Correspondence: Daniel Peeper (d.peeper@nki.nl)

Background

Immune checkpoint blocking therapies (ICB) such as anti-PD1 antibodies have revolutionized the treatment of melanoma and lung cancer, leading to clinical benefit in approximately 50% and 19% of patients, respectively [1,2]. However, since ICB resistance remains a major challenge, novel treatment options are needed to combat these immunotherapy-refractory tumor fractions. We and others previously reported that the receptor tyrosine kinase AXL is a critical resistance marker to targeted therapy in melanoma and lung cancer [3,4], and that AXL-positive cell fractions can be effectively eliminated by an AXL-targeting antibody-drug conjugate (enapotamab vedotin, EnaV [5]). AXL may also mark immunotherapy-resistant tumor cell fractions [6], and antibody-drug conjugates may induce an inflammatory response, which could reinvigorate the immune response and may synergize with immune therapies [7]. Here we explored the potential of EnaV to cooperate with tumor-specific T cells and ICB.

Methods

We established a cohort of melanoma and lung cancer cell lines HLA- and antigen-matched to tumor-specific T cells using the MART-1 TCR system [8, 9]. These cell lines were treated with EnaV and matched cytotoxic T cells in vitro to evaluate their sensitivity. Furthermore, we tested whether EnaV could induce an inflammatory response in vitro and in vivo. Finally, we assessed any cooperative anti-tumor activity of EnaV, T cells, and anti-PD1 in human xenografts of melanoma and lung cancer.

Results

Several AXL-positive melanoma and lung cancer models showed a dose-dependent cooperative sensitivity to combined treatment with EnaV and tumor-specific T cells. EnaV had a profound effect on tumor models that were refractory to either tumor-specific T cells and/or ICB. Mechanistically, RNA and proteomic profiling revealed that EnaV treatment induced an inflammatory response in tumors, including markers of immunogenic cell death. Combining EnaV with tumor-specific T cells and anti-PD1 proved superior to either treatment alone in melanoma and lung cancer models. This effect was also seen in a model which showed no response to anti-PD1 treatment alone, suggesting that EnaV can create a de novo sensitization to anti-PD1.

Conclusions

Our findings demonstrate that EnaV and immunotherapy cooperatively exert a strong anti-tumor response in therapy-refractory melanoma and lung cancer models. Importantly, even in models fully refractory to either T cells and/or anti-PD1, EnaV treatment restores sensitivity to promote tumor elimination. Together, these results indicate that targeting AXL-positive, immunotherapy-resistant tumor fractions with EnaV enhances ICB sensitivity, thus warranting further investigations into this combination in a clinical setting.

References

1. Larkin J, Chiarion Sileni V, Gonzalez R, Grob J-J, Cowey CL, Lao CD, et al. Combined Nivolumab and Ipilimumab or Monotherapy in Untreated Melanoma. N Engl J Med. Massachusetts Medical Society; 2015 Jul 2;373(1):23–34.

2. Garon EB, Rizvi NA, Hui R, Leighl N, Balmanoukian AS, Eder JP, et al. Pembrolizumab for the treatment of non-small-cell lung cancer. N Engl J Med. Massachusetts Medical Society; 2015 May 21;372(21):2018–28.

3. Müller J, Krijgsman O, Tsoi J, Robert L, Hugo W, Song C, et al. Low MITF/AXL ratio predicts early resistance to multiple targeted drugs in melanoma. Nat Commun. Nature Publishing Group; 2014;5:5712.

4. Zhang Z, Lee JC, Lin L, Olivas V, Au V, LaFramboise T, et al. Activation of the AXL kinase causes resistance to EGFR-targeted therapy in lung cancer. Nat Genet. Nature Publishing Group; 2012 Aug;44(8):852–60.

5. Boshuizen J, Koopman LA, Krijgsman O, Shahrabi A, van den Heuvel EG, Ligtenberg MA, et al. Cooperative targeting of melanoma heterogeneity with an AXL antibody-drug conjugate and BRAF/MEK inhibitors. Nat Med. Nature Publishing Group; 2018 Feb;24(2):203–12.

6. Aguilera TA, Rafat M, Castellini L, Shehade H, Kariolis MS, Hui AB-Y, et al. Reprogramming the immunological microenvironment through radiation and targeting Axl. Nat Commun. Nature Publishing Group; 2016 Dec 16;7:1–14.

7. Müller P, Kreuzaler M, Khan T, Thommen DS, Martin K, Glatz K, et al. Trastuzumab emtansine (T-DM1) renders HER2+ breast cancer highly susceptible to CTLA-4/PD-1 blockade. Sci Transl Med. American Association for the Advancement of Science; 2015 Nov 25;7(315):315ra188–8.

8. Jorritsma A, Gomez-Eerland R, Dokter M, van de Kasteele W, Zoet YM, Doxiadis IIN, et al. Selecting highly affine and well-expressed TCRs for gene therapy of melanoma. Blood. 2007 Nov 15;110(10):3564–72.

9. Vredevoogd DW, Kuilman T, Ligtenberg MA, Boshuizen J, Stecker KE, de Bruijn B, et al. Augmenting Immunotherapy Impact by Lowering Tumor TNF Cytotoxicity Threshold. Cell. Elsevier Inc; 2019 Jul 9;:1–31.

Ethics Approval

The collection and use of human tissue was approved by the Medical Ethical Review Board of the Antoni van Leeuwenhoek. Animal experiments were approved by the animal experimental committee of the institute and performed according to Dutch law.

Correspondence: Jeffrey Hubbell (jhubbell@uchicago.edu)

Background

Cancer immunotherapies using checkpoint inhibitor antibodies (CPI) show considerable antitumor response in the clinic. However, a substantial number of patients suffered from treatment-related adverse events. A number of patients receiving combination CPI therapy could not continue with the therapy due to adverse events. Immunotherapies serve to activate immune responses, and as such, side-effects include the symptoms of systemic lymphocyte activation and auto-immune disease induction, which typically results from drug action in healthy organs. One strategy to address this problem is through localized drug delivery systems, thereby restricting their actions to the disease site. Here, we report two technologies of tumor localized CPI antibodies via conjugation to an extracellular matrix promiscuous binding peptide derived from placenta growth factor-2 (PlGF-2123-144) or a collagen-binding domain (CBD) derived from the blood protein von Willebrand factor (VWF) A3 domain. PlGF-2123-144 conjugated CPI retains at the injection site long-term; thus PlGF-2123-144-CPI is effective when injected peri-tumor. CBD-CPI harnesses the specific exposure of tumor stroma collagen to blood stream due to the abnormal leakiness of the tumor vasculature; thus CBD-CPI is effective when it is injected intravenously.

Methods

PlGF-2123-144 peptide was synthesized chemically. CBD recombinant proteins were produced in HEK-293F cells and purified by ÄKTA system. We have used a chemical cross linker to conjugate PlGF-2123-144 and CBD to CPI antibody. B16F10 and BrafPten melanoma, MMTV-PyMT breast cancer, CT26 models were used. 100μg each of CPI(anti-CTLA4+anti-PD-L1) was administered.

Results

We show that PlGF-2123-144-conjugated CPI localized to the injection site for long time, and limiting the CPI exposure to the body and toxicity including diabetes induction risk (AB). Also, PlGF-2123-144-CPI was more efficacious than unmodified CPI, injected at same dose and same way, when injected near tumor (C). PlGF-2123-144-CPI increased intratumoral effector CD8+ T cell number compares to CPI. Next we have tested CBD conjugated CPI, injected intravenously. Intravenously administered CBD protein accumulated mainly in tumors in an orthotopic MMTV-PyMT breast tumor model (D). CBD-conjugation to CPI decreased systemic toxicity, such as T cell infiltration into the liver (E). CBD-CPI significantly suppressed tumor growth compared to their unmodified forms in B16F10 melanoma, CT26 colon carcinoma and MMTV-PyMT breast cancer models (F). CBD-CPI increased tumor-infiltrating CD8+T cells; increases in the ratio of effector CD8+ T cells to T regulatory cells were observed.

Conclusions

Our data suggest that the tumor-matrix binding localized CPI technology can improve both safety and efficacy of CPI therapy, with high translational promise.

References

1.Ishihara et al., Sci Transl Med 9, eaan0401 (2017).

2. Ishihara, J., et al., Sci Transl Med, 11. Eaau3259 (2019).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1 (abstract P694).

See text for description

Correspondence: Raghida Bukhalid (rbukhalid@mersana.com), Timothy Lowinger (tlowinger@mersana.com)

Background

STimulator of Interferon Genes (STING) has emerged as an innate immune pathway capable of inducing anti-tumor immune activity through activation of antigen presenting cells and production of type I interferon, leading to T-cell priming and activation. In murine models, both intratumoral and intravenous administration of STING agonists have been shown to induce tumor regressions and generate immunological memory; clinical studies are underway. We hypothesize that a STING antibody-drug conjugate (ADC) – in which the STING agonist is conjugated to an antibody – will exhibit several advantages, including targeted delivery of STING agonist to desired cell types, an optimal pharmacokinetic profile, the convenience of systemic administration, and ultimately an improved therapeutic index.

Methods

We generated novel STING ADCs by developing a conjugation platform which consists of a STING agonist incorporated into a chemical scaffold for bioconjugation, designed to provide optimal drug-like properties. We studied the activity and mechanism of action of STING ADCs in monoculture and co-culture in vitro systems, as well as anti-tumor activity and pharmacodynamics in mice. Herein we report data with STING ADCs generated from antibodies to two therapeutic targets.

Results

The STING ADCs achieved a 100-fold higher potency than the corresponding free agonist in in vitro assays using THP-1 reporter cells. Similarly, in tumor cell/PBMC co-cultures, the STING ADCs exhibited 100-fold higher tumor cell-killing activity compared to free agonist, demonstrating the benefits of targeted delivery. Intravenous administration of STING ADCs regressed tumors in mouse models after a single dose of 1 mg/kg ADC in a target-dependent manner; in the same models, intravenously administered free agonist at a 50-fold higher dose resulted in only modest tumor growth delay. Importantly, a single intravenous administration of the STING ADC led to a significant increase in immune cell infiltration and tumor cell death within 72 hours and a significant increase in tumor-localized inflammatory cytokines within 12 hours, while levels of systemic cytokines remained relatively low.

Conclusions

We have developed a STING ADC platform and demonstrated target-dependent anti-tumor immune responses both in vitro and in vivo for two therapeutic targets.

Ethics Approval

Animal studies were conducted following the recommendations of the Guide for Care and Use of Laboratory Animals and in compliance with the Institutional Animal Care and Use Committees of Translational Drug Discovery and Charles River Discovery Services.

Correspondence: Alan Epstein (aepstein@usc.edu)

Background

While checkpoint inhibitors have found stunning success as single agent therapies, their clinical efficacy only extends to a subset of patients across multiple cancers. Even among likely responder preclinical models, response rates are often triggered by an early, unknown stochastic event. Ultimately, we do not know why some tumors respond and some do not. Previous work from our laboratory devised a strategy for dramatically improving systemic administration of an immune agonist of Toll-like receptor 9 (TLR9) known as CpG Oligodexoynucleotide (ODN) whose anti-tumor efficacy has been traditionally limited to intra- and peri-tumoral injections.[1,2] Through antibody-guided delivery of conjugated CpG ODN to the tumor microenvironment (TME), TLR9 activation promotes dendritic cell maturation thus upregulating cross-presentation of TME antigens.

Methods

Here we describe our latest multi-targeted approach which tackles immune suppression by overcoming mechanisms of acquired resistance to immunotherapy and synergizing therapeutic responses. This novel antibody immunoconjugate (AIC) targets murine TLR9 by tethering ODN1826 to a checkpoint inhibitor antibody against programmed death ligand 1 (αPD-L1) thereby stimulating and disinhibiting the immune system. Our studies of αPD-L1/ODN1826 showed superior in vivo efficacy over their free agent counterparts when delivered by intraperitoneal injection to D2F2 tumor-bearing BALB/cJ mice. All animals were treated humanely and in accordance with the guidelines of the USC’s Institutional Animal Care and Use Committee.

Results

Tumor growth inhibition was calculated by using the geometric mean of relative tumor volumes. The AIC demonstrated a statistically significant delay in tumor growth compared to single agents ODN1826 (-∆71.6%, p≤0.001) and αPD-L1 (p≤0.001) as well as compared to combination therapy ODN1826 + αPD-L1 (-∆51.8%, p≤0.01).

Conclusions

By transitioning to bioluminescence imaging techniques of tumor cells, we overcome the limitations of traditional tumor volume measurement by caliper and capture a fuller picture of both early therapeutic response events and metastatic disease. Additionally, subsequent deep immunophenotyping by Mass CyTOF combined with recent advancements in Digital Spatial Profiling, we can establish a more comprehensive method of tumor immune profiling. Elucidating the immune status of murine tumor models is a critical factor for predicting therapeutic response and understanding the mechanisms underlying immune suppression and activation.

Acknowledgements

This work was supported by Cancer Therapeutics, Inc., Los Angeles, CA.

References

1. Li, Z. et al. Generation of tumor-targeted antibody-CpG conjugates. J. Immunol. Methods 389, 45–51 (2013).

2. Jang, J. K. et al. Systemic delivery of chTNT-3/CpG immunoconjugates for immunotherapy in murine solid tumor models. Cancer Immunol. Immunother. 65, 511–23 (2016).

Ethics Approval

All animals were treated humanely and in accordance with the guidelines of the USC’s Institutional Animal Care and Use Committee under protocol 20265.

Correspondence: Francesca Zammarchi (francesca.zammarchi@adctherapeutics.com)

Background

Regulatory T cells (Tregs) play an important role in the establishment and progression of tumors and are considered a major obstacle to tumor eradication by immunotherapies [1]. Moreover, the intra-tumoral balance between Tregs and effector T cells (Teffs) appears to influence the outcome of immunotherapies [2], and poor prognosis in solid tumors is often associated with high tumor infiltration by Tregs and a low ratio of Teffs/Tregs [3].

Methods

Sur301 is an antibody-drug conjugate (ADC) composed of PC61, a rat monoclonal antibody directed against mouse CD25, conjugated to a pyrrolobenzodiazepine (PBD) dimer via a protease-cleavable linker, with a drug-to-antibody ratio of 2 [4]. We have previously shown that sur301 has potent and durable anti-tumor activity in vivo against CD25-negative immunogenic solid tumors with infiltrating CD25-positive Tregs and, when used at a sub-optimal dose, its activity is further enhanced in combination with PD-1 blockade [4]. Sur301 anti-tumor activity, either alone or combined with an anti-PD-1 antibody, was significantly reduced in the absence of CD8+ T cells, and when tumor-free survivors were re-challenged, they did not develop new tumors indicating that sur301 was able to induce tumor-specifi¬c protective immunity [4].

Results

In order to characterize the mode of action of sur301, we analysed the immunophenotype of Tregs and Teffs isolated from tumors, blood and spleens of mice bearing established MC38 tumors following a single dose of sur301 either alone or in combination with an anti-PD-1 antibody. Additionally, we performed a T cells dynamic study in non-tumor bearing immunocompetent mice over a period of 21 days. Spleen, lymph node, thymus and blood were analysed following a single dose of sur301 or an isotype control ADC. In both studies, sur301 resulted in strong but specific Tregs depletion, while Teffs were not affected by sur301. Interestingly, the substantial Tregs depletion observed in spleen, lymph node and blood of non-tumor bearing mice was accompanied by a temporary but significant elevation of thymic Tregs.

Conclusions

In conclusion, these new data show that sur301 is able to mediate potent Tregs depletion without affecting Teffs. Together with our previous data [4], these results suggest that sur301’s mode of action is at least in part mediated by Tregs depletion while Teffs are not affected. Translation of these pre-clinical data in the clinic is currently being investigated in a phase I trial evaluating the efficacy of camidanlumab tesirine (ADCT-301), a PBD-based ADC targeting human CD25, in patients with selected advanced solid tumors (NCT03621982).

References

1. Sasidharan Nair, V. and E. Elkord, Immune checkpoint inhibitors in cancer therapy: a focus on T-regulatory cells. Immunol Cell Biol, 2018. 96(1): p. 21-33.

2. Menetrier-Caux, C., et al., Targeting regulatory T cells. Target Oncol, 2012. 7(1): p. 15-28.

3. Arce Vargas, F., et al., Fc-Optimized Anti-CD25 Depletes Tumor-Infiltrating Regulatory T Cells and Synergizes with PD-1 Blockade to Eradicate Established Tumors. Immunity, 2017. 46(4): p. 577-586.

4. Zammarchi, F., et al., A CD25-targeted pyrrolobenzodiazepine dimer-based antibody-drug conjugate shows potent anti-tumor activity in pre-clinical models of solid tumors either alone

Correspondence: Yanzhang Wei (ywei@clemson.edu)

Background

Major histocompatibility complex class I chain-related protein A (MICA) is a stress-induced ligand for the activating receptor NKG2D on natural killer (NK) cells. It is expressed on the cell surface of most cancer types at the early stages, but usually down-regulated or shed in advanced cancers leading to immune evasion [1,2]. Prolactin is an essential protein hormone for normal reproduction and maintenance of pregnancy, and contributes to pathogenesis of gynecologic malignancies [3]. The prolactin receptor (PRLR) was found over-expressed in many cancers, especially breast cancer [4]. G129R is a mutant of human prolactin with a single amino acid substitution. It still binds to PRLR but inhibits the growth of breast cancer cells [5]. We aim to create a chimeric protein with MICA and G129R, and use it to bridge breast cancer cells and NK cells. The MICA portion in the chimeric protein binds to the NKG2D receptor and activate NK cells, and the G129R portion targets to the PRLR on the breast cancer cells and leads NK cells to them.

Methods

A chimeric gene was created by cloning and fusing the cDNA of the extracellular domain of MICA and the cDNA of G129R into a plasmid, and then transfected into HEK293 cells to produce MICA-G129R protein. The MICA-extracellular-domain only protein and the G129R only protein were also generated as controls. When the proteins were added into the co-culture of NK cells (NK-92 cell line) and breast cancer cells (T47D cell line), the cytotoxicity was measured using a lactate dehydrogenase (LDH)-based assay, the granzyme B released by NK-92 cells was detected using ELISA, and the apoptosis of the T47D cells was examined using a caspase 3/7 luminescence assay.

Results

We successfully generated HEK293 stable clones to produce MICA-G129R, MICA and G129R proteins. The proteins were all confirmed using Western blot. The chimeric protein MICA-G129R was proved to be able to bind to NK-92 cells and T47D cells using immunofluorescent staining. The chimeric protein MICA-G129R significantly increased the cytotoxicity of the NK-92 cells to T47D cells in their co-culture. The granzyme B secretion of NK-92 cells was induced, and the caspase3/7 in the T47D cells were elevated in the co-culture of NK-92 and T47D with the presence of MICA-G129R.

Conclusions

The chimeric protein MIC-G129R can bridge NK cells and breast cancer cells, and induce the cytotoxicity of NK cells to kill the breast cancer cells.

References

1. Salih HR, Rammensee H-G, Steinle A. Cutting edge: down-regulation of MICA on human tumors by proteolytic shedding. J Immunol. 2002;169(8):4098-4102.

2. Duan S, Guo W, Xu Z, et al. Natural killer group 2D receptor and its ligands in cancer immune escape. Mol Cancer. 2019;18(1):29.

3. Bachelot A, Binart N. Reproductive role of prolactin. Reproduction. 2007;133(2):361-369.

4. Goffin V. Prolactin receptor targeting in breast and prostate cancers: New insights into an old challenge. Pharmacol Ther. 2017;179:111-126.

5. Chen WY, Ramamoorthy P, Chen N, Sticca R, Wagner TE. A human prolactin antagonist, hPRL-G129R, inhibits breast cancer cell proliferation through induction of apoptosis. Clin Cancer Res. 1999;5(11):3583-3593.

Correspondence: Ayelet Chajut (ayelet@kahr-medical.com)

Background

Checkpoint inhibitors have transformed cancer treatment in certain patient subgroups. To improve response rates, considerable effort is now being invested developing novel immune-oncology (IO) drugs and exploring IO target combinations.

DSP107 is a bi-functional, trimeric protein consisting of the fused extracellular domains of human SIRPα and 4-1BBL that uniquely combines activation of both innate and adaptive immunity in a single product. Interaction of CD47, frequently overexpressed on cancer cells, with SIRPα on phagocytes transmits a “don’t eat me” signal. Binding of DSP107 to CD47 on tumor cells blocks the inhibitory signal delivered to phagocytes and cross-presents trimerized 4-1BBL to activated tumor antigen-specific T-cells that express the 4-1BB costimulatory receptor. Cross-presentation and trimerization of the 4-1BBL are required for optimal 4-1BB activation.

Methods

DSP107 was manufactured in CHO-S cells. Binding and affinity to CD47 and 4-1BB were evaluated by Biacore, FACS and ELISA. T-cell activation, proliferation, IFN-γ secretion, cytotoxicity and phagocytosis were confirmed in vitro using lymphoma, leukemia and carcinoma cell lines. In-vivo efficacy was tested against hCD47-MC38 tumors in h41BB-KI-C57BL/6 mice and against SUDHL4 (lymphoma) in humanized NSG mice. Safety was evaluated in a non-GLP, repeat-dose study in cynomolgus monkeys.

Results

DSP107 binds to CD47 and 4-1BB with high affinity (1.2 nM and 0.7 nM, respectively). Cell based reporter assays showed CD47 binding dependency for signaling through 4-1BB. Using in vitro assays, DSP107 amplified T-cell activation and proliferation, and augmented T-cell mediated elimination of cancer cells (0.5-5 nM). DSP107 also augmented macrophage-mediated phagocytosis of tumor cells as monotherapy and in combination with targeted monoclonal antibodies (6-12 nM). DSP107 led to statistically significant tumor growth inhibition (TGI) as monotherapy (lymphoma model; p

Conclusions

DSP107 is a bifunctional, fusion protein with a unique mechanism of action. In vivo, DSP107 was efficacious as monotherapy and in combination with anti PD-L1. Doses ≤ 50 mg/kg were safe and well tolerated in monkeys. IND-enabling studies are ongoing and clinical studies will commence in Q2/2020.

Ethics Approval

With reference to the study in h41BB-KI-C57BL/6 mice, the protocol and any amendment(s) or procedures involving the care and use of animals were reviewed and approved by the Institutional Animal Care and Use Committee (IACUC) of CrownBio prior to execution. During the study, the care and use of animals were conducted in accordance with the regulations of the Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC).

With reference to the lymphoma model in humanized NSG mice, the study was approved by the Hebrew University Ethical Board, approval number MD-19-15821-5.

The protocol for the cynomolgus monkey study (Covance Study 8402826) was reviewed and approved by the Establishment Animal Welfare and Ethical Review Body. This study also complies with the permissions and conditions contained within Project Licence P98E2EBFD "Safety Testing of Medicinal Products Using Non-Human Primates" as authorised by the UK Home Office.