Background IGV-001 is a personalized, autologous cancer cell-based immunotherapy conceived to deliver a tumor-derived antigenic payload in the context of immunostimulatory signals to patients with glioblastoma (GBM). IGV-001 consists of patient-derived GBM cells treated with an antisense oligodeoxynucleotide against insulin-like growth factor 1 receptor (IGF1R) and placed in proprietary biodiffusion chambers (BDCs). The BDCs are then exposed to 5–6 Gy radiation and implanted at abdominal sites for ~48 hours. IGV-001 has previously been shown to be generally safe with promising clinical activity in newly diagnosed GBM patients.
Methods Mouse (m) or human (h) variants of IGV-001 were prepared using GL261 mouse GBM cells or human GBM cells, respectively. BDCs containing vehicle or mIGV-001 were implanted in the flanks of C57BL/6 albino female mice in preventative and therapeutic experiments, optionally in combination with a programmed cell death 1 (PD-1) blocker. Bioactivity of the general approach was also measured against hepatocellular carcinoma Hepa 1–6 cells. Mice were followed for the growth of subsequently implanted or pre-existing tumors and survival. Draining lymph nodes from mice receiving mIGV-001 were immunophenotyped. mIGV-001 and hIGV-001 were analyzed for extracellular ATP and high mobility group box 1 (HMGB1) as indicators of immunogenic cell death (ICD), along with flow cytometric analysis of viability, surface calreticulin, and reactive oxygen species. Stress and cell death-related pathways were analyzed by immunoblotting.
Results IGV-001 causes oxidative and endoplasmic reticulum stress in GL261 cells, resulting in a cytotoxic response that enables the release of antigenic material and immunostimulatory, ICD-associated molecules including ATP and HMGB1 from BDCs. Immunophenotyping confirmed that IGV-001 increases the percentage of dendritic cells, as well as effector, and effector memory T cells in BDC-draining lymph nodes. Consistent with these observations, preventative IGV-001 limited tumor progression and extended overall survival in mice intracranially challenged with GL261 cells, a benefit that was associated with an increase in tumor-specific T cells with effector features. Similar findings were obtained in the Hepa 1–6 model. Moreover, therapeutically administered IGV-001 combined with PD-1 delayed progression in GBM-bearing mice.
Conclusions These results support treatment with IGV-001 to induce clinically relevant ICD-driven anticancer immune responses in patients with GBM.
- Immunogenicity, Vaccine
- Brain Neoplasms
Data availability statement
Data are available upon reasonable request.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
WHAT IS ALREADY KNOWN ON THIS TOPIC
IGV-001, a biologic-device combination product, has been shown to be safe and mediate potential efficacy in a Phase I clinical trial enrolling patients with glioblastoma (GBM).
WHAT THIS STUDY ADDS
We demonstrate that IGV-001 can induce oxidative and endoplasmic reticulum stress on GBM cells in the product, resulting in immunogenic cell death and consequent antitumor immunity.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
This study supports the use of IGV-001 as a therapeutic modality in GBM, while providing data on its mechanism of action.
Glioblastoma (GBM) is the most prevalent and lethal primary brain malignancy in adults.1 Standard of care (SOC) for GBM generally consists in maximal safe surgical resection followed by adjuvant radiation therapy (RT) plus temozolomide, followed by temozolomide-based maintenance.2 GBM is well known for its elevated spatial and temporal intertumoral and intratumoral heterogeneity,3 4 pointing to a suboptimal clonal control by cytotoxic T lymphocytes (CTLs) and hence phenotypic clonal divergence.5 Consistent with this observation, the GBM tumor microenvironment (TME) is generally enriched with immunosuppressive cells that hinder CTL functionality,4 6 7 contributing to a pronounced resistance to immune checkpoint inhibitors (ICIs) and other immunotherapeutic agents. Indeed, nivolumab, an ICI targeting programmed cell death 1 (PDCD1, best known as PD-1), failed to improve outcomes as compared with vascular endothelial growth factor A (VEGFA) neutralization with bevacizumab, which is approved for the treatment of GBM,8 in patients with recurrent GBM.9 Chimeric antigen receptor (CAR)-expressing T cells targeting a GBM-associated variant of epidermal growth factor receptor (EGFR) also did not improve the survival of patients with GBM.10 Similarly, various therapeutic cancer vaccines that have been or are being clinically evaluated in patients with GBM have yet to provide compelling therapeutic benefits to this patient population.11–15
IGV-001 is an autologous cancer cell-based immunotherapeutic approach conceived to deliver an antigenic payload in the context of immunostimulatory molecules to patients with GBM. Specifically, IGV-001 consists of autologous GBM cells that are (1) incubated with an antisense oligodeoxynucleotide (IMV-001) targeting insulin-like growth factor 1 receptor (IGF1R), (2) placed in proprietary biodiffusion chambers (BDCs) of 0.1 µm pore size together with additional IMV-001, (3) exposed to radiation in a single dose of 5–6 Gy, and (4) implanted in patients at abdominal sites (ie, at distance from the immunosuppressive GBM TME and draining lymph nodes).16 IGV-001 is advantageous in that it comprises an unselected population of cancer cells, thereby including a broad antigenic signature of each tumor. Moreover, the BDC pore size allows for the selective release of relatively small subcellular tumor particles, which are taken up efficiently by immunostimulatory dendritic cells (DCs)17 18 while preventing contact between dying cancer cells and macrophages, which generally promote immunosuppression.19 20 IGV-001 has recently been evaluated in patients with newly diagnosed GBM as part of a Phase 1b clinical trial (NCT02507583). Specifically, IGV-001 was manufactured shortly after tumor resection (during patient hospitalization) and administered 4–6 weeks prior to SOC adjuvant RT plus temozolomide-based chemotherapy. In this context, median overall survival (OS) of ‘Stupp-eligible’ patients (n=10) receiving the highest IGV-001 exposure was 38.2 months as compared with 16.2 months in recent patients receiving SOC therapy (p=0.044).16 Importantly, these findings led to the initiation of a follow-up Phase 2b study (NCT04485949) and prompted preclinical exploration to unravel the mechanism of action of IGV-001.
Importantly, cancer cells can respond to stress (including mechanical and RT-elicited stress) in a variety of ways, ranging from the activation of survival pathways that support the restoration of cellular homeostasis to the initiation of molecular cascades precipitating regulated cell death (RCD).21 While RCD is obviously desirable in the context of cancer therapy, multiple signal transduction pathways that often accompany RCD actively suppress the initiation of cancer-specific immune responses, potentially limiting the overall efficacy of treatment.22 Conversely, specific therapeutic approaches including various chemotherapies23 and targeted anticancer agents,24 RT (at least when used according to precise dose and fractionation schedules),25 as well as photodynamic therapy26 can all elicit an immunologically proficient variant of RCD commonly known as ‘immunogenic cell death’ (ICD), which culminates with the activation of tumor-targeting immunity in support of treatment efficacy.22
Using both murine and human GBM cell lines as well as patient-derived GBM cells, here we demonstrate that IGV-001 rapidly (<48 hours) elicits several ICD-associated immunostimulatory signals in support of tumor-targeting immunity.
Materials and methods
Luciferase-expressing GL261 mouse GBM cells (GL261-luc2) were obtained from PerkinElmer (by Labcorp). The GL261 cell line was purchased from DSMZ (Leibniz Institute, DE; ACC 802). JAWS II (CRL-11904) mouse immature DC, T98G (CRL-1690) and U-87 MG (HTB-14; herein referred to as U87) human GBM cell lines were purchased from the American Type Culture Collection (ATCC). Culture media was prepared according to provider’s recommendations with Roswell Park Memorial Institute (RPMI)-1640 or Eagle’s Minimum Essential Medium and supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin. Cultures were maintained at 37°C under 5% CO2 according to the provider’s instructions.
Primary GBM cell isolation
Clinical samples were obtained with patient consent under Thomas Jefferson University’s Clinical Study Protocol #TJU-14 379–102, Version #102.0. GBM tumors were collected under Jefferson’s approved tissue procurement protocol Control #21C.030 and shipped in normal saline in temperature-controlled (2–8°C) containers. Resected tumor material was minced and dissociated at 37°C using a proprietary enzyme cocktail. After dissociation, the cell pellet was red blood cell(RBC) lysed, passed through a 40 µm cell strainer, resuspended in saline, and counted. Culture medium for experiments involving patient-derived GBM cells used Human Plasma-Like Media (Thermo Fisher Scientific, California, USA) supplemented with 5% human serum and 1% penicillin/streptomycin.
IGV-001 formulation and culture
For the preparation of mIGV-001 (GL261-luc2 or GL261) and hIGV-001 (T98G or U87), cell lines were grown to ~80% confluency in T75 tissue culture flasks and then treated with IMV-001 (Nitto Denko Avecia, Massachusetts, USA) for 3 hours. After treatment, cells were harvested, washed with complete media, resuspended in 0.9% saline (SteriCare Solutions, Texas, USA) with additional IMV-001 and counted. The cell suspension (365 µL) was loaded into pre-made BDCs [0.1 μm-pore size, polyvinylidene difluoride (PVDF) membrane] at a density of 5×105–1×106 cells/BDC and then irradiated with 5–6 Gy (225 kV) X-rays using a MultiRad irradiator (Precision X-Ray Irradiation, CT). IGV-001 was prepared with patient-derived tumor cells as previously described, with modifications.16 For in vitro experiments, BDCs were placed into 12-well tissue culture plates containing 750 µL of complete media, covered with 200 µL more media and incubated at 37°C under 5% CO2 atmosphere for 24–48 hours. Due to passive diffusion, 200 µL of media was added to the top of each BDC every 24 hours to maintain moisture across both membranes and supply the cells with a minimal flow of nutrients. For cell death inhibition studies, GL261 cells pretreated for 3 hours with 25 µM zVAD-fmk, 30 µM necrostatin-1 (Nec-1), 1 µM ferrostatin-1 (Fer-1), or 1–2 mM N-acetyl cysteine (NAC) (Sigma Aldrich, Missouri, USA). Following manufacturing, mIGV-001 was further co-incubated with inhibitors at the same concentrations for 24 hours.
Orthotopic GL261-luc2 mouse glioma model and lymph node immune response in vivo
In vivo murine studies were designed by Imvax, and executed by Labcorp in compliance with the National Institutes of Health and Labcorp’s Animal Care and Use Committee or at Rutgers, the State University of New Jersey at an Association for Assessment and Accreditation of Laboratory Animal Care International (AAALAC)-accredited animal facility, under an Institutional Animal Care and Use Committee (IACUC)-approved protocol. Female C57BL/6 albino mice or male C57BL/6 mice received either BDCs filled with phosphate buffered saline (PBS) (control mice) or formulated mIGV-001 (1×106 GL261-luc2 cells/BDC, one BDC/mouse). BDCs were subcutaneously implanted in the flank for 48 hours followed by explantation and wound closing. On day 26 after BDC explantation, mice received an intracranial inoculation of 1×106 GL261-luc2 cells to generate orthotopic tumors. T-cell subsets were depleted using anti-CD4 (clone GK1.5) or anti-CD8 antibody (clone 2.43) (Bio X Cell; 500 µg/dose/animal on days 0–2, and 200 µg/dose/animal on days 5, 8, 11, 14, 17, 20, 23, 26, all administered intraperitoneally). Tumor burden was evaluated using weekly bioluminescent imaging (BLI) of injected D-luciferin conversion with an IVIS Spectrum imager (PerkinElmer, Missouri, USA). Mice were euthanized before study termination when ethical abortion criteria were reached (eg, body weight loss ≥20%, signs of sickness, distended cranium, severely impaired movement, severe respiratory distress or loss of righting reflex). Lymph nodes (draining and contralateral) from mice that received mIGV-001 (GL261-luc2) were immunophenotyped by flow cytometry. To enhance the response, 26 days after explantation, mice received a subcutaneous (SC) boost with 1×107 GL261-luc2 cells/0.2 mL and lymph node collection was performed 7 days later. Additional methods for GL261 therapeutic setting and hepatocellular carcinoma models can be found in online supplemental methods and materials section.
Serum cytokine analysis
Blood serum samples were isolated at end of life and circulating levels of interferon gamma (IFN-G, best known as IFN-γ), interleukin-10 (IL-10), and interleukin-6 (IL-6) were measured by Luminex analysis using a MILLIPLEX MAP mouse CD8+ T-cell premixed kit on a MAGPIX Luminex xMAP reader (Sigma Aldrich, Missouri, USA) as per the manufacturer’s protocol. Serum samples were also taken to coincide with tumor burden BLI measurements. Circulating levels of IL-6 were also quantitated using a V-PLEX Proinflammatory Panel I (Meso Scale Discovery). Circulating levels of IL-6 evaluated on days 1, 14, and 28 days were correlated with BLI measurements on days 6, 13, and 27. At each time point and for each mouse receiving mIGV-001 (n=34), IL-6 levels were assigned to “Progressed” or “Stable” if the natural logarithm Ln(BLI total flux)>20 or Ln(BLI total flux)<20, respectively.
Tumor-specific T-cell assays
Peripheral blood mononuclear cells (PBMCs) collected on day 1 post-tumor challenge or splenocytes collected at end-of-life from control mice and surviving mIGV-001 treated mice were used in FluoroSpot or ELISpot assays to measure IFN-γ production by T cells. PBMCs were stimulated using tumor lysate-pulsed DCs and splenocytes were incubated with major histocompatibility complex (MHC)-I H-2Db and H-2Kb GL261-specific peptides27 (CHI Scientific). IFN-γ was measured using a murine IFN-γ capture FluoroSpot kit (CTL) or mouse IFN-γ ELISpot Kit (R&D Systems). IFN-γ counts were evaluated in an ImmunoSpot S6 UNIVERSAL Analyzer (CTL) or S6 Ultimate-V Analyzer (CTL).
ATP and high mobility group box 1 release
Supernatants from mIGV-001 and hIGV-001 BDCs were isolated at indicated time points, clarified by centrifugation, and stored at −80°C for later analysis. Similarly, IGV-001 BDC contents from a Phase 1b clinical trial16 were stored at −80°C on BDC explantation. Extracellular high mobility group box 1 (HMGB1) was quantified by ELISA kit (Tecan, Switzerland) and extracellular ATP (eATP) was quantified using luciferase-luciferin RealTime-Glo eATP Assay (Promega, Wisconsin, USA), as per manufacturer’s protocols. Measurements were recorded on a Cytation 5 Cell Imaging Multimode Reader (BioTek Instruments, Vermont, USA). Saline-loaded BDCs without cells were used as a vehicle control for both HMGB1 and eATP analyses. ATP release was also assessed after the IGV-001 manufacturing process.
Cell death and reactive oxygen species (ROS) quantification by flow cytometry mIGV-001 and hIGV-001 or patient-derived IGV-001 cells were retrieved from BDCs after 0, 24 or 48 hours and viability was assessed via annexin V and 7-AAD staining (BioLegend, California, USA), NucView 488 staining (Biotium, California, USA) for caspase 3–7 activity, and MitoView 633 for mitochondrial membrane potential, as per manufacturer’s recommendations. Recovered cells were washed and resuspended in either binding buffer containing annexin V PeCy7 conjugate with 7-AAD or PBS plus 5% FBS containing NucView-488 or MitoView 633 substrates and incubated for 15 min at room temperature or 37°C in the dark. Stained cells were diluted in an appropriate buffer and then analyzed on a MACSQuaroont Analyzer 16 flow cytometer (Miltenyi Biotec, Germany). Cell counting via trypan blue exclusion was also used. Flow cytometric data analysis was performed using the Kaluza Analysis Software (Beckman Coulter, California, USA) package. mIGV-001 and hIGV-001 cells harvested from BDCs after 24 hours with or without NAC treatment were washed twice in PBS plus 5% FBS and ROS was determined via ROS Detection Assay Kit (Abcam, UK) using the 2’,7’-dichlorodihydrofluorescein diacetate (H2DCFDA) oxidation system as per manufacturer’s protocol. Data was collected on a MACSQuant Analyzer 16 flow cytometer and analysis was performed using the Kaluza Analysis Software package. An aliquot of stained cells was further counter-stained with 4’, 6-diamidino-2-phenylindole (DAPI) and then imaged on an EVOS Cell Imaging system (Thermo Fisher Scientific, California, USA), maintaining consistent exposure time and light intensity for each time point.
Analysis of IGV-001 immediately after manufacturing was used as control to determine flow cytometry gating strategies for further time points.
Cell pellets from mIGV-001 were harvested at 0, 2, and 24 hours post preparation and lysed in a RIPA buffer containing 1× protease and phosphatase inhibitors (Thermo Fisher Scientific, California, USA). 1.5× protease and phosphatase inhibitors were subsequently used to maintain phosphorylation status of target proteins. Lysates were clarified by centrifugation and protein concentration determined by Pierce Rapid Gold BCA Assay (Thermo Fisher Scientific, California, USA). 30–50 µg of protein per well was loaded onto 4–20% SurePAGE Bis-Tris gels (GenScript, New Jersey, USA) and separated at 120 V for 1–1.5 hours. Proteins were transferred to either PVDF or nitrocellulose (NC) membranes using iBlot 2 system (Thermo Fisher Scientific, California, USA) and blocked in Tris-buffered saline with 0.1% Tween 20 (TBST) containing 5% non-fat milk or TBST containing 5% bovine serum albumin (BSA) (for phosphoproteins). Membranes were washed and incubated overnight at 4°C with primary antibodies diluted to 1:500-1:1000 in 5% BSA-TBST followed by a 1-hour incubation at room temperature with horseradish peroxidase (HRP)-conjugated secondary antibodies diluted at 1:2500 in 5% milk-TBST. After washing, membranes were developed with an Immobilon Forte Western HRP substrate (EMD-Millipore, Massachusetts, USA) and imaged on an iBright 1500 Imaging System (Thermo Fisher Scientific, California, USA). For phosphorylated proteins, membranes were first incubated with primary antibodies against specific phosphopeptides; imaged, stripped using Pierce Restore stripping buffer (Thermo Fisher Scientific, California, USA), incubated with primary antibodies against the relevant total protein, and reimaged. Membranes were stripped following image acquisition of target proteins and reprobed for glyceraldehyde 3-phosphate dehydrogenase (GAPDH) as lysate loading control. Images were exported and densitometry performed using image analysis software ImageJ (National Institutes of Health). Target proteins were normalized to GAPDH as loading control and data presented as fold change from 0 hour. All antibodies used were purchased from Cell Signaling Technologies (Danvers, Massachusetts, USA).
Appropriate statistical analyses were performed in GraphPad Prism (V.9.4). Significance was determined using analysis of variance followed by t-criteria, Mann-Whitney, non-parametric unpaired t-test or independently by Student’s t-test with or without Welch’s correction, as appropriate. Survival studies were analyzed using the log-rank test. A p<0.05 was considered statistically significant.
IGV-001 elicits TH1-polarized tumor-targeting immune responses that control murine GBMs in both preventative and therapeutic settings
To assess the ability of IGV-001 to elicit ICD, we manufactured a mouse (m) version of the product based on luciferase-expressing GL261 (GL261-luc2) cells and used it in gold-standard preventative vaccination assays.28 Specifically, mIGV-001 or PBS-loaded BDCs were implanted SC in the flank of immunocompetent C57BL/6 albino mice, left in place for 48 hours and then explanted. Twenty-six days later, mice were challenged intracranially with live GL261-luc2 cells (figure 1A). Representative bioluminescence images show development of progressive GBM in mice receiving PBS-loaded BDCs (figure 1B). Conversely, earlier bioluminescence (BLI) readouts (until day 28) demonstrate tumor regression in 47% of mice receiving mIGV-001 in the preventative setting (figure 1B,C). Mice in the control group had a median OS of 21 days versus 53.5 in mIGV-001-exposed mice (figure 1D), pointing to elicitation of protective GBM-targeting immunity. Consistent with this observation, depletion of CD8+ and more so CD4+ T cells limited ability of mIGV-001 to induce preventative immunity in this model (figure 1E). A similar preventative activity was demonstrated harnessing luciferase-expressing mouse hepatocellular carcinoma (HCC) Hepa 1-6(Z1)-luc (hereinafter Hepa 1-6) cells for product manufacturing (online supplemental figure 1A,B). Of note, all mice successfully rejecting orthotopically implanted Hepa 1-6 cells on vaccination, remained fully protected against a second SC inoculation of Hepa 1-6 cells >100 days after initial challenge (online supplemental figure 1C). Importantly, while neither mIGV-001 nor a PD-1 blocker significantly delayed progression in mice with established GBM, their combination exhibited at least some degree of activity (figure 1F).
Blood samples from mice vaccinated with PBS-loaded BDCs or mIGV-001 were profiled for multiple TH1 and TH2 cytokines at end-of-life. Consistent with the activation of a TH1 response, IFN-γ was elevated while both IL-6 and IL-10 were reduced in sera of mice receiving mIGV-001, as compared with their control counterparts (figure 2A–C). Moreover, circulating levels of IL-6 were higher in mice that received mIGV-001 but developed progressive GBM as compared with mice actively controlling disease establishment and progression (figure 2D). Finally, T cells isolated from mice receiving mIGV-001 efficiently responded with IFN-γ secretion (as assessed in ELISpot assays) to GL261-luc2 cell lysates (figure 2E) or known GL261-derived MHC class I immunogenic peptides27 (figure 2F), as compared with T cells from mice receiving PBS-loaded BDCs or naïve mice, respectively.
Taken together, these findings suggest that IGV-001 elicits TH1-polarized immunity against GBM cells, which is associated with disease control in mice.
IGV-001-draining lymph nodes are enriched in mature antigen-presenting cells and lymphoid effector cells
To elucidate cellular mechanisms involved in preventative GBM-targeting immunity elicited by IGV-001, we used multiparametric flow cytometry to compare mIGV-001-draining lymph nodes (IDLNs) to contralateral lymph nodes (CLNs). IDLNs contained a similar fraction of CD45+CD11b+ myeloid cells as compared with CLNs (figure 3A) but were enriched in CD11c+MHC-II+ DCs (figure 3A), which are involved in antigen presentation.18 In agreement with these results, IDLNs also contained an increased proportion of CD8+ and effector CD4+FOXP3− T cells as compared with CLNs (figure 3B), and both these cellular compartments exhibited a more pronounced polarization towards an effector memory (CD62L−CD44+) over central memory (CD62L+CD44+) phenotype (figure 3C,D). Finally, both CD8+ and effector CD4+ T cells found in IDLNs were characterized by increased levels of PD-1, as compared with their CLN-resident counterparts (figure 3E).
To extend these findings to an in vitro model of immune activation, we used immortalized immature murine DCs (JAWSII cells) to evaluate the expression of maturation markers on co-culture with mIGV-001. Specifically, JAWSII cells were pretreated overnight with 100 ng/mL lipopolysaccharide (LPS, a broad DC activatory stimulus) and then either left untreated or exposed for 24 hours to mIGV-001 or mIGV-001 prepared with paraformaldehyde (PFA)-fixed GL261 cells (which prevents cells from delivering antigenic and immunostimulatory payloads to DCs). In line with in vivo data (figure 3A), mIGV-001 (but not its PFA-fixed variant) efficiently boosted LPS-driven upregulation of DC activation markers, including CD80, CD86 and MHC-II by JAWSII cells (online supplemental figure 2). Additionally, particle analysis confirmed that during 48 hours in vitro incubation, a human (h) variant of IGV-001 prepared with T98G human GBM cells produced particles that can efficiently diffuse through the BDC membranes (<0.1 µm) (online supplemental figure 3A). Computational diffusion models generated under static and dynamic conditions demonstrate that regardless of internal concentration, particles of 25–90 nm can exit BDCs. Under static conditions, particle equilibrium inside and outside BDCs could be achieved after 40 hours, 80 hours, and 260 hours for 25 nm, 50 nm and 90 nm particles, respectively (online supplemental figure 3B). Under dynamic conditions, equilibrium was predicted in less than 5 hours, irrespective of particle size (online supplemental figure 3C).
These findings indicate that tumor particles generated by IGV-001 support DC activation in IDLNs, culminating with nodal T-cell expansion and acquisition of effector functions.
IGV-001 generates ICD-associated danger signals in both mouse and human systems
To characterize immunostimulatory factors underlying the ability of IGV-001 to drive anticancer immunity, we used both mIGV-001 and hIGV-001 prepared with human T98G or U87 GBM cells. First, we employed multiparametric flow cytometry to assess cytotoxic responses in mIGV-001 and hIGV-001, and found a statistically significant, time-dependent decrease in recoverable viable cells (figure 4A). Importantly, similar results were obtained with hIGV-001 prepared with patient-derived primary GBM cells and maintained in vitro (figure 4B), as well as with both mIGV-001 and hIGV-001 implanted SC for 48 hours in immunocompetent mice and then explanted (figure 4C). Besides radiation-elicited cellular stress, the demise of GBM cells manipulated to manufacture IGV-001 involves lack of substrate attachment (Online supplemental figure 4A) as well as reduced glucose and protein availability (online supplemental figure 4B-E) within BDCs.
In vitro, IGV-001-associated RCD was not accompanied by increased exposure of the “eat-me” signal calreticulin22 on the surface of GBM cells (online supplemental figure 5). Conversely, IGV-001 elicited RCD along with release of the ICD-associated endogenous immunostimulatory molecule HMGB122 in both mouse and human systems, both within BDCs and in the surrounding culture medium (figure 5A,B). Importantly, similar results were obtained not only with mIGV-001 and hIGV-001 implanted in C57BL/6 mice, as described above (figure 5C), but also in supernatants of BDCs explanted from patients involved in clinical IGV-001 testing (NCT02507583)16 (figure 5D), which had detectable levels above controls (ie, in vitro cultured saline BDCs and surrounding media) and reported circulating levels of HMGB1 in healthy individuals.29 eATP, another immunostimulatory signal associated with ICD,22 could also be detected within mIGV-001 and hIGV-001 BDCs explanted 48 hours after establishment in C57BL/6 mice, respectively (figure 5E). Similarly, eATP was measurable upon mIGV-001 and hIGV-001 manufacturing in vitro (figure 5F), but less so within BDCs (figure 5G–I), potentially reflecting a steep gradient linked to expression of various ectonucleotidases by GBM cells.30 31
These data demonstrate the ability of IGV-001 to elicit RCD along with the emission of ICD-associated immunostimulatory signals in both human and mouse GBM models.
IGV-001 imposes oxidative and endoplasmic reticulum stress onto GBM cells
Next, we investigated the implication of stress response pathways that have previously been linked to the emission of ICD-associated danger signals, such as the integrated stress response (ISR),22 in immunostimulatory effects of IGV-001. The conditions experienced by GL261 cells in mIGV-001 were found to elicit ROS overproduction that could be scavenged, at least partially, by the antioxidant NAC (figure 6A,B). Similar results were observed with hIGV-001 (figure 6B). ROS as elicited by mIGV-001 manufacturing also turned out to contribute, at least kinetically, to mIGV-001-driven cytotoxicity, as demonstrated by cytofluorometric assessment of RCD in the optional presence of NAC (figure 6C). In the human system, though, ROS scavenging failed to alter cytotoxic responses of GBM cells to hIGV-001 preparation (figure 6C). mIGV-001 resulted in inactivating phosphorylation of eukaryotic translation initiation factor 2 subunit alpha (best known as eIF2α) (figure 6D), a conserved marker of ISR initiation,32 coupled with upregulation of ISR effector molecules including activating transcription factor 4 and DNA damage inducible transcript 3 (DDIT3, best known as CHOP)33 (figure 6E). Moreover, consistent with establishment of oxidative and endoplasmic reticulum stress, we detected rapid phosphorylation of ROS-sensitive kinase mitogen-activated protein kinase 8 (MAPK8, best known as JNK) in GL261 cells manufactured to generate mIGV-001 (figure 6F).
These findings identify oxidative stress responses and ISR as ICD-associated pathways triggered by IGV-001.
Molecular RCD mechanisms triggered by IGV-001
Next, we looked at various biomarkers of cell death pathways that have been previously associated with immunogenicity, including apoptotic and necroptotic cell death.22 Due to the upregulation of CHOP and the phosphorylation of JNK detected in GL261 cells employed for mIGV-001 manufacturing (figure 6E,F), we initially focused on apoptotic signaling. Using flow cytometry on staining with dedicated fluorescent markers, we detected a considerable increase in the activity of apoptotic ‘executioners’ caspase 3 and caspase 7 24 hours after product manufacturing (figure 7A). This was paralleled by an increase in cells with dissipated mitochondrial transmembrane potential (Δψm) as well as by the downregulation of the antiapoptotic protein BCL2 apoptosis regulator (BCL2) (figure 7B,C), strongly implicating mitochondrial apoptosis in the cytotoxic effects of mIGV-001.21 Interestingly, however, the necroptotic signal transducer mixed lineage kinase domain like pseudokinase (MLKL) also appeared to undergo activating phosphorylation in the context of mIGV-001 (figure 7D), potentially suggesting the concomitant activation of multiple RCD pathway by this drug-device product. At least in part supporting this result, pharmacological inhibition of apoptosis with the pan-caspase blocker zVAD-fmk, necroptosis with the receptor interacting serine/threonine kinase 1 (RIPK1) inhibitor Nec-1, or ferroptosis (another regulated variant of necrosis) with Fer-1 did not alter the kinetic of RCD as imposed on GL261 cells by mIGV-001 manufacturing (figure 7E).
These findings suggest that IGV-001 immunogenicity may originate from the activation of a multipronged RCD program downstream of failing adaptation to stress.
Taken together, the findings reported herein demonstrate that IGV-001 elicits immunogenic cellular stress culminating with bona fide ICD and the activation of therapeutically relevant, TH1-polarized immune responses against GBM. These data are in line with immunological correlates measured in patients with newly diagnosed GBM receiving IGV-001 as part of a recently reported Phase 1b clinical trial.16 Specifically, peak elevations of circulating IFN-γ, IL-2, and IL-12, which altogether reflect DC maturation culminating with T-cell activation/expansion,18 were observed shortly after IGV-001 implantation to patients with GBM.16 While these changes did not persist systemically over time, most likely owing to the immunosuppressive nature of SOC therapy delivered to all patients involved in this study,34 they suggest that IGV-001-driven ICD may also be operational in patients with GBM.
Interestingly, the ability of IGV-001-driven ICD to protect mice from a subsequent challenge with GBM cells was compromised on depletion of CD8+ or CD4+ T cells, which is in line not only with the need of both these cell compartments for ICD to mount optimal tumor-targeting responses,22 but also with a previously described key role of both CD8+ and CD4+ T cells in the control of mouse GBM GL261 cells.35 However, while we were able to directly assess CD8+ T-cell reactivity in the periphery of mice receiving IGV-001 on stimulation with GL261 cell-specific MHC class I-restricted peptides, similar MHC class II-restricted reagents are unavailable, which complicated direct assessment. Thus, whether CD4+ T cells contribute to GBM-targeting immunity as elicited by IGV-001 purely as ‘helpers’ for CD8+T cells or also by mediating direct antitumor effects (as reported in multiple other mouse tumor models)36 remains to be formally validated.
Our findings suggest that IGV-001-elicited immune responses exhibit signs of TH1 polarization, including decreased circulating IL-6. Importantly, IL-6 has previously been associated with tumor progression in multiple human solid tumors, including GBM.37 Moreover, tumor-derived IL-6 has previously been found to upregulate the immunosuppressive molecule CD274 (best known as programmed cell death ligand-1; PD-L1) on GBM-infiltrating myeloid cells, culminating with accelerated GBM progression.38 Consistent with these findings, IL-6 blockage limited intratumoral PD-L1 expression and improved the survival of GL261-bearing mice.38 These findings suggest that at least part of the therapeutic effects of IGV-001 may result from reductions in IL-6.
Importantly, GBM generally displays scarce infiltration by immune cells (ie, it is a so-called “cold” tumor) and hence is virtually insensitive to ICIs9 and other immunotherapeutic agents.10–12 IGV-001 immunotherapy may be advantageous in this respect as BDCs are implanted at abdominal acceptor sites between the rectus muscle and sheath,39 which are minimally affected by the immunosuppressive intracranial microenvironment established by progressing GBM.40
One of the caveats of this study is predominant reliance on preventative rather than therapeutic experimental settings. This largely reflects our focus on bona fide ICD, which can primarily be discerned from RCD-independent immunostimulation in preventative vaccination assays.28 Indeed, therapeutic models based on systemic interventions cannot readily discriminate against ICD over a direct immunostimulatory effect of the treatment that does not involve RCD.28 Moreover, established GL261 GBMs are highly aggressive, often enabling insufficient time for the establishment of therapeutically relevant anticancer immunity in therapeutic settings (ie, when treatment is initiated in the presence of detectable tumors). That said, while neither IGV-001 nor a PD-1 blocker employed as standalone agents delayed progression in GBM-bearing mice, the combinatorial regimen elicited at least some degree of tumor control. Whether such a control can be boosted with additional immunotherapeutic agents with proven activity in murine GBM such as tumor necrosis factor receptor superfamily member 4 (TNFRSF4, best known as OX40) agonists41 remains to be formally elucidated. Irrespective of this caveat, IGV-001 has already shown potential activity in the clinic.16
Another limitation of our studies emerges from the use of a C57BL/6 mouse background, as imposed by the GL261 model to preserve a syngeneic system.42 C57BL/6 mice are indeed intrinsically prone to mount TH1-polarized responses as compared with BALB/c mice and other mouse strains,43 which may not necessarily reflect the natural polarization of tumor-targeting immunity in humans. However, all well-characterized transplantable mouse models of GBMs are syngeneic to C57BL/6 mice,42 and surgical resection of transgene-driven GBMs (most of which are anyway available in the C57BL/6 mouse background)42 for individual IGV-001 manufacturing is not technically practical. Novel transplantable mouse models of GBMs syngeneic to BALB/c mice are needed to extend our findings to a different mouse strain.
Finally, it is plausible (although it remains to be formally investigated) that cancer cell type may influence, at least to some degree, the ability of IGV-001 to elicit tumor-targeting immunity downstream of ICD. For instance, it is conceivable that cancer cells with intrinsic defects in ICD-associated stress responses, such as the ISR or autophagy (which underlies ATP release during ICD),44 may engage suboptimal immunostimulation on IGV-001 administration. Along similar lines, the impact of tumor mutational burden (TMB), which has been associated with sensitivity to ICIs in multiple cohorts of patients with cancer,45 on the immunogenicity of IGV-001 remains to be fully assessed. Whether assessing TMB and/or functional markers of a functional ISR or proficient autophagy may identify patients at increased likelihood to respond to IGV-001 remains to be investigated.
Despite these and other caveats, our data formally establish IGV-001 as a bona fide ICD inducer in both mouse and human model systems, warranting further preclinical and clinical investigation into IGV-001 in combination with immunotherapy. Moreover, our findings with HCC cells support the utility of this biologic-device platform for the treatment of solid tumors for which malignant cells can be readily isolated, other than GBM. Work is underway to explore and develop a therapeutic platform with broader applications based on the mechanism of action of IGV-001.
Data availability statement
Data are available upon reasonable request.
Patient consent for publication
There are two institutional review board protocols involved with this study. Please see below for their titles, control# and Board# (153 and 2405). (1) Phase I Study in Humans Evaluating the Safety of Rectus Sheath Implantation of Diffusion Chambers Encapsulating Autologous Malignant Glioma Cells Treated with Insulin Growth Factor Receptor-1 Antisense Oligodeoxynucelotide (IGF-1 R/AS DON) in 32 Patients with Newly Diagnosed Malignant Glioma (Private Foundation) Control #14G.96Document(s) Approved: Clinical Study Protocol 102.0, Study Number: Protocol #TJU-14379-102, Version #102.0 dated 02 Jan 2014; Informed Consent Version V1.0In accordance with Federal-Wide Assurance #00002109 to the US Department of Health land Human Services, this study was approved by Board #153. (2) Analysis of cell count, viability, and immunogenicity of de-identified discarded newly diagnosed glioblastoma tissue and more broadly other newly diagnosed solid tumor tissues from the operating room. Control #21C.030In accordance with Federal-Wide Assurance #00002109 to the US Department of Health and Human Services, this study was approved by Board #2405. Participants gave informed consent to participate in the study before taking part.
The authors would like to acknowledge John Limongelli and Drs Raul Perez-Olle and Belen Carrillo-Rivas for their critical review of this manuscript. We would like to thank Ms Mayrimar Torres for her outstanding laboratory management. Finally, we would like to recognize the Preclinical Oncology group at Labcorp Drug Development, the In Vivo Pharmacology group at Reaction Biology, Germany, and the In Vivo Research Services staff at Rutgers University, for their assistance with the execution of in vivo experiments.
CU and KK contributed equally.
Contributors CC: Experimental planning, data curation, formal analysis, and writing (original draft). CU: Experimental planning, data curation, formal analysis, and writing (original draft). KK: Data curation, formal analysis. EAE: Data curation. AZ: Data curation. DWA: Writing (review and editing). CS: Statistical analyses. LG: Writing (review and editing). MAE: Investigation, writing (review and editing) and guarantor. JZ: Investigation, experimental planning, conceptualization, formal analysis, writing (original draft, review, and editing).
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests CC, CU, KK, EAE, AZ, DWA, MAE, and JZ are/were employed by Imvax. CS is a paid consultant of Imvax. LG is/has been holding research contracts with Lytix Biopharma, Promontory and Onxeo, has received consulting/advisory honoraria from Boehringer Ingelheim, AstraZeneca, OmniSEQ, Onxeo, The Longevity Labs, Inzen, Imvax, Sotio, Promontory, Noxopharm, EduCom, and the Luke Heller TECPR2 Foundation, and holds Promontory stock options.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.