You are here

Article Text

Article menu
Basic tumor immunology
Type I MET inhibitors cooperate with PD-1 blockade to promote rejection of hepatocellular carcinoma
  1. Ricardo DeAzevedo1,
  2. Madeline Steiner1,2,
  3. Broderick X Turner1,2,
  4. Arthur Liu1,2,
  5. Sherwin Newton1,
  6. Joanna Schmidt3,
  7. Rachel Fleming3,
  8. Angelica Tolentino3,
  9. Ahmed O Kaseb4 and
  10. Michael A Curran1
  1. 1Department of Immunology, The University of Texas MD Anderson Cancer Center, Houston, Texas, USA
  2. 2Immunology Program, The University of Texas MD Anderson UTHealth Graduate School of Biomedical Sciences, Houston, Texas, USA
  3. 3Ionis Pharmaceuticals, Carlsbad, California, USA
  4. 4Department of Gastrointestinal (GI) Medical Oncology, Division of Cancer Medicine, The University of Texas MD Anderson Cancer Center, Houston, Texas, USA
  1. Correspondence to Dr Michael A Curran; mcurran{at}mdanderson.org

Abstract

Blockade of the immune checkpoints programmed death-1 (PD-1) and cytotoxic lymphocyte antigen 4 has improved outcomes for patients with hepatocellular carcinoma (HCC), yet most still fail to achieve objective clinical benefit. MET plays key roles in both HCC tumorigenesis and immunosuppressive conditioning; however, inhibition of MET causes upregulation of PD-ligand 1 (PD-L1) suggesting the use of these inhibitors in the context of PD-1 blockade. We sought to investigate across the Hepa1-6, HCA-1 and diethylnitrosamine (DEN) models of HCC whether the combination of more specific type I versus more pleiotropic type II MET inhibitors would confer superior outcomes in combination with PD-1 blockade. While MET inhibition demonstrated cooperativity with αPD-1 across all three models, the type I MET inhibitor capmatinib showed optimal activity in combination and statistically significantly outperformed the combination with the type II inhibitor cabozantinib in the αPD-1 refractory DEN model. In both HCA-1 and DEN HCC, the capmatinib and αPD-1 combination enhanced CD8 T cell frequency and activation state while limiting intratumoral myeloid immune suppression. In vitro studies of antigen-specific T cell activation reveal significantly less inhibition of effector cytokine production and proliferation by capmatinib versus by type II or type III MET inhibitors. These findings suggest significant potential for clinical HCC combination studies of type I MET inhibitors and PD-1 blockade where prior trials using type II inhibitors have yielded limited benefit.

  • Immunotherapy
  • Hepatocellular Carcinoma
  • Combination Therapy
  • Immune Checkpoint Inhibitor

Data availability statement

Data are available upon reasonable request. Data were generated by the authors and included in the article. Raw data supporting the findings in this study can be obtained from the author upon reasonable request.

http://creativecommons.org/licenses/by-nc/4.0/

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/.

https://doi.org/10.1136/jitc-2024-009690

Statistics from Altmetric.com

    Request Permissions

    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.

    Introduction

    Until recently hepatocellular carcinoma (HCC) carried a dismal prognosis with systemic therapy limited to modestly beneficial kinase inhibitors such as sorafenib. Introduction of programmed death 1 (PD-1), PD-ligand 1 (PD-L1) and cytotoxic lymphocyte antigen 4 immune checkpoint blockade (ICB) therapies alone or in combination with vascular endothelial growth factor kinase inhibition have significantly improved this therapeutic landscape with more than 20% of patients achieving objective responses. Unfortunately, this still leaves the majority of metastatic patients with HCC lacking any substantially effective therapy.

    c-Met (MET), the receptor for hepatocyte growth factor (HGF), is commonly overexpressed in HCC driving tumor cell proliferation and angiogenesis and correlating with poor survival outcomes.1 2 Three classes of MET inhibitor referred to as type I or selective and ATP-competitive, type III or selective and non-competitive, and type II or poly-kinase and competitive have been developed with multiple members of each class now under clinical study.3 4 While MET inhibition can suppress HCC tumor xenograft growth in immune-deficient animals,1 these drugs have not had a significant impact on clinical disease as monotherapies. A prior study demonstrated that MET inhibition resulted in PD-L1 upregulation which could attenuate therapeutic benefit through increased immune suppression.5 In PD-1 blockade-sensitive models of HCC, these studies also demonstrated the potential for anti-PD-1 and type II or III MET inhibitors to cooperate in promoting tumor rejection. Despite the promise of these findings, clinical studies combining the Type II MET inhibitor cabozantinib and PD-1 or PD-L1 blockade have not provided significant improvement over existing FDA-approved therapies.6 7

    We investigated whether distinct classes of MET inhibitors might exhibit differential levels of cooperativity with PD-1 blockade, as well as to reveal the underlying cellular mechanisms by which each combination improved anti-HCC tumor immunity. We chose to focus our comparison on the highly selective type I MET inhibitor capmatinib, which was more effective than the type III inhibitor tivantinib with PD-1 in the prior study,5 versus the broadly acting type II inhibitor cabozantinib which has been selected for most clinical combination trials. In the syngeneic ICB-sensitive Hepa1-6 and HCA-1 HCC models, both MET inhibitors improved anti-PD-1 responsiveness with a trend toward superiority of the type I capmatinib combination driven by enhanced T cell infiltration and function coupled with more pro-inflammatory myeloid phenotypes. In the in vivo passaged diethylnitrosamine (DEN) HCC model which is anti-PD-1 refractory; however, we observed significant superiority of the type I MET inhibitor capmatinib and anti-PD-1 combination over the type II cabozantinib and PD-1 blockade dual therapy. Again, we found unique advantages of the capmatinib and anti-PD-1 combination in improving the anti-HCC CD8 T cell response while diminishing myeloid suppression, particularly at the level of granulocytic myeloid-derived suppressor cells (GrMDSC). This finding was particularly compelling in the context of a prior manuscript showing that MET attenuates ICB benefit via recruitment of neutrophils that convert to GrMDSC in the tumor,8 and another documenting increased neutrophil infiltration with the cabozantinib and anti-PD-1 combination therapy in HCC.9 In vitro, we also find direct attenuation of CD8 T cell effector cytokine production and Granzyme B by MET inhibitors that is least apparent with capmatinib. These findings provide a compelling rationale for combining type I MET inhibition and PD-1 checkpoint blockade in clinical HCC.

    Results

    Cooperativity between type I MET inhibition and PD-1 blockade in ICB-sensitive HCC

    Consistent with previously published findings for type I and III MET inhibitors performed in Hepa1-6 (mouse) and Hep3B (human), we found that all three classes of MET inhibitors upregulated PD-L1 on mouse and human HCC tumor lines in vitro (online supplemental figure 1).5 The combination doses of each drug that we use for HCC therapy with PD-1 blockade were based on those previously published for capmatinib5 and cabozantinib.9 The C57BL/6 Hepa1-6 HCC model is known to be ICB sensitive with PD-1 blockade alone capable of rejecting 80% of 7 day pre-implanted tumors in our hands. While some have found less checkpoint sensitivity, grown in our facility using male Jackson Labs C57BL6 mice, we found our results were more aligned with those of Jin et al showing among the highest levels of αPD-1 sensitivity for Hepa1-6.10 Even in this context, we found that combining the type I c-Met inhibitor capmatinib with the anti-PD-1 antibody RMP1-14 provided a statistically significant improvement with 100% of mice tumor-free (p=0.0323) versus no advantage for the type II cabozantinib combination (figure 1A). Subcutaneous HCA-1 tumors in C3H/HEJ mice are still immunotherapy sensitive, but to a lesser degree than is observed in Hepa1-6.11 While we found 50% tumor-free animals with PD-1 blockade alone, the addition of cabozantinib (73%, p=0.0452) or capmatinib (97%, p<0.0001) significantly improved HCA-1 HCC clearance while neither inhibitor alone showed efficacy (figure 1B). In this setting, the higher efficacy of the type I MET inhibitor combination with αPD-1 versus the type II inhibitor was statistically significant (p=0.014).

    Supplemental material

    Figure 1
    Figure 1

    PD-1 blockade and c-Met inhibitor combination therapy in αPD-1 sensitive HCC models. (A) 5×106 Hepa1-6 cells were implanted s.c. on the right flank of male C57BL6 mice. On days 5, 9, 13, 17 and 21 mice received either 10 mg/kg capmatinib, 30 mg/kg cabozantinib or vehicle by oral gavage. On days 7, 11, 15, and 19, mice received 250 ug anti-PD-1 RMP1-14 i.p. or isotype control antibody. Two repeats of 10 mice/group. (B) 5×106 HCA-1 cells were implanted s.c. on the right flank of male C3H/HeJ mice. Three repeats of 10 mice/group. Mice were followed for survival (tumor size≥1000 mm3) and tumor growth (calipers). Log-rank (Mantel-Cox) test. *p adj<0.05, **p adj< 0.01, ***p adj<0.001, ****p adj<0.0001; data are mean±SEM. HCC, hepatocellular carcinoma; i.p., intraperitoneally; PD-1, programmed death 1; s.c., subcutaneously; TF, tumor free.

    Combination capmatinib and αPD-1 enhances both T and myeloid cell antitumor immunity

    In order to gain insight into the observed advantage of capmatinib versus cabozantinib in combination with PD-1 blockade, we treated HCA-1 tumors with 5 doses of the indicated MET inhibitor and/or 4 doses of PD-1 antibody and then analyzed their tumor immune microenvironment (TIME) the day after the last treatment. Compared with PD-1 blockade alone, adding capmatinib increased the frequency of CD8 and CD4 effector T cells in the TIME while concomitantly decreasing the CD4 FoxP3+regulatory T cell population (Treg) (figure 2A, online supplemental figure 2). In both Hepa1-6 and HCA-1, prior studies with PD-1 blockade have shown an essential role for CD8 T cells in tumor regression while CD4 T cells have appeared less required.12–15 CD4+FoxP3− frequencies did not change with therapy; however, only the capmatinib combination increased active CD4+FoxP3−GranzymeB+frequency (online supplemental figure 3A). Infiltrating CD8 T cells appeared more active based on the level of CD44 expression and less exhausted based on substantially lower PD-1 expression compared with anti-PD-1 alone (figure 2B). While the cabozantinib combination displayed none of these advantages over anti-PD-1 alone, we did find that both c-Met inhibitors improved CD8 proliferation (Ki67) and memory phenotype frequency (CD44+, CD62L+) over αPD-1 alone with equal efficiency (figure 2C). Only the capmatinib and αPD-1 combination demonstrated significant pro-inflammatory enhancement of the myeloid stroma, however, including elevated cDC1 frequency (CD11c+, CD8α+) and reduced Arginase expression in the granulocyte compartment (CD11b+, Ly6G+, Ly6C−) (figure 2D, online supplemental figure 4). Overall, type I c-Met inhibition appears to enhance the benefits of PD-1 blockade to both T cell and myeloid antitumor immunity, while the benefits of less specific type II inhibitors are limited to modest improvement in CD8 phenotype.

    Figure 2
    Figure 2

    Combination αPD-1 and c-Met inhibition reshapes the HCA-1 TIME. 5×106 HCA-1 cells were implanted s.c. on the right flank of male C3H/HeJ mice in 30% Matrigel (Corning). On days 5, 9, 13, 17 and 21 mice received either 10 mg/kg capmatinib, 30 mg/kg cabozantinib or vehicle by oral gavage. On days 7, 11, 15, and 19, mice received 250 ug anti-PD-1 RMP1-14 i.p. or isotype control antibody. Tumors were extracted and analyzed by flow cytometry on day 22. Two repeats of 10 mice/group. (A) T cell subset frequencies (B, C) T cell phenotypes and (D) myeloid cell frequencies and phenotypes are shown following analysis in GraphPad Prism using Student’s t-test. *p adj<0.05, **p adj< 0.01, ***p adj<0.001, ****p adj<0.0001. GrMDSC, granulocytic myeloid-derived suppressor cells; i.p., intraperitoneally; PD-1, programmed death 1; s.c., subcutaneously; TIME, tumor immune microenvironment; Treg, regulatory T cell population.

    Type I c-Met inhibition overcomes αPD-1 resistance in the DEN HCC model

    While the Hepa1-6 and HCA-1 models of HCC are known to be sensitive to PD-1 blockade, spontaneous liver tumors arising from administration of DEN are known to demonstrate limited to absent αPD-1 sensitivity depending on diet.16 17 To facilitate reproducible experiments at a large scale, we used an in vivo passage DEN HCC tumor system in which slices of isolated tumor are implanted subcutaneously (s.c.) on the flank preserving the tumor and its microenvironment during engraftment that was developed by the Tolentino Laboratory at Ionis. In this system, mice were treated when measurable tumors emerged with 10 doses of c-Met inhibitor and 5 doses of the PD-1 antibody RMP1-14. In this system, we observed no significant benefit to c-Met inhibition alone or with PD-1 blockade. In contrast, the combination of PD-1 blockade and the type I c-Met inhibitor capmatinib cured 45% of mice with a mean overall survival (OS)of 173 days versus 20% and 88 days for aPD-1 alone and 20% and 113 days for the cabozantinib combination (figure 3). Notably, the capmatinib combination demonstrated significant superiority to the cabozantinib combination in this αPD-1 refractory setting (p=0.03).

    Figure 3
    Figure 3

    PD-1 blockade and c-Met inhibitor combination therapy in αPD-1 resistant DEN HCC model. 2 mm x 2 mm slices of in vivo passaged DEN HCC tumor were implanted s.c. on the right flank of male C57BL6 mice. On days 1–10 following tumor palpation (∼45 days) mice received either 10 mg/kg capmatinib, 30 mg/kg cabozantinib or vehicle by oral gavage. On days 2, 4, 6, 8, and 10 post-palpation, mice received 250 ug anti-PD-1 RMP1-14 i.p. or isotype control antibody. Two repeats of 10 mice/group. Mice were followed for survival (tumor size≥1000 mm3) and tumor growth (calipers). Log-rank (Mantel-Cox) test. *p adj<0.05, **p adj< 0.01, ***p adj<0.001, ****p adj<0.0001; data are mean±SEM. DEN, diethylnitrosamine; HCC, hepatocellular carcinoma; i.p., intraperitoneally; PD-1, programmed death 1; s.c., subcutaneously; MFI, mean fluorescence intensity.

    Pro-inflammatory remodeling of the DEN HCC time by capmatinib with PD-1 blockade

    We repeated the DEN HCC tumor therapy experiments as above, this time analyzing the composition of the TIME after therapy using multiparameter flow cytometry. The capmatinib and αPD-1 combination enhanced absolute immune infiltration of these DEN HCC tumors, and, similar to HCA-1, increased CD8 frequency while decreasing Treg frequency (figure 4A). While overall CD4+FoxP3− frequency did not change for either c-Met inhibitor combination compared with αPD-1 alone, only the capmatinib combination increased Granzyme B activation in the CD4 compartment (online supplemental figure 3B). Major histocompatibility complex II expression could be strongly induced in both Hepa1-6 and DEN tumor cells, while in HCA-1 it was constitutively expressed at a low level regardless of interferon-γ (IFNγ) stimulation (online supplemental figure 3C). In the DEN model, only the capmatinib combination with αPD-1 increased CD8 T cell proliferation, cytotoxic potential, memory fraction and decreased PD-1 expression (figure 4B). While both c-Met inhibitors helped drive down monocytic MDSC frequency generally and of the Arginase+ subset, only capmatinib combined with αPD-1 reduced GrMDSC frequency, and that of the Arginase+subset, to marginal levels (figure 4C). M2 macrophage frequency also declined precipitously when capmatinib was added to αPD-1 indicating a broad reversal of suppressive myeloid polarization. Within both the cDC1 subset (CD8α+) and dendritic cell (DC)2 subsets of DCs, PD-L1 expression decreased and only the combination with capmatinib also drove elevated costimulatory CD86 expression by the DC2 compartment (figure 4D). Reflecting the HCA-1 HCC data, these findings show a unique potential for the type I c-Met inhibitor capmatinib to mediate pro-inflammatory remodeling of the DEN HCC TIME that is not replicated by the type II inhibitor cabozantinib.

    Figure 4
    Figure 4

    Capmatinib and PD-1 blockade augment antitumor immunity in the DEN TIME. 2 mm x 2 mm slices of in vivo passaged DEN HCC tumor were implanted s.c. on the right flank of male C57BL6 mice. On days 1–10 following tumor palpation (∼45 days) mice received either 10 mg/kg capmatinib, 30 mg/kg cabozantinib or vehicle by oral gavage. On days 2, 4, 6, 8, and 10 post-palpation, mice received 250 ug anti-PD-1 RMP1-14 i.p. or isotype control antibody. Two repeats of 10 mice/group. Tumors were extracted and analyzed by flow cytometry on day 11 post-treatment. Two repeats of 10 mice/group. (A) Immune cell prevalence (B) T cell phenotypes, (C) MDSC frequencies and (D) myeloid cell phenotypes are shown following analysis in GraphPad Prism using Student’s t-test. (E) OT-I CD8 T cells were activated with SIINFEKL peptide for 72 hours and cytokine secretion was measured using the BD Cytometric Bead Array and (F) proliferation was measured using the CellTrace CFSE Cell Proliferation Kit (Thermo) and Granzyme B was assessed by flow cytometry following fixation and staining with the GB11 anti-mouse/human Granzyme B antibody. *p adj<0.05, **p adj< 0.01, ***p adj<0.001, ****p adj<0.0001. DEN, diethylnitrosamine; GrMDSC, granulocytic myeloid-derived suppressor cells; HCC, hepatocellular carcinoma; IFNγ, interferon-γ; i.p., intraperitoneally; MoMDSC, monocytic myeloid-derived suppressor cells; PD-1, programmed death 1; PD-L1, PD-ligand 1; s.c., subcutaneously; TIME, tumor immune microenvironment; TNF, tumor necrosis factor, Treg, regulatory T cell population; TF, tumor free; MFI, mean fluorescence intensity.

    In addition to the more pro-inflammatory environment preserved by capamatinib, we sought to investigate whether direct effects on the T cell compartment might be contributing to the apparent advantage of type I MET inhibition in the context of ICB. We activated Ovalbumin-specific OT-I CD8 T cells in vitro for 72 hours with their cognate SIINFEKL peptide and then measured effector cytokine production in the presence of each type of MET inhibitor. In this setting, IFNγ and tumor necrosis factor α production were significantly inhibited by either type II or III MET inhibitors, but only marginally so by capmatinib (figure 4E). Similarly, antigen-driven T-cell proliferation was largely unaffected by capmatinib but increasingly inhibited by ascending doses of either tivantinib or cabozantinib (figure 4F). CD8 T cell cytotoxic potential measured as Granzyme B production was also preserved with capmatinib but significantly diminished even at the lowest concentrations of tivantinib or cabozantinib (figure 4F). Interestingly, capmatinib has been shown to increase perforin production which has been inhibited through HGF/c-Met signaling in human CD8 T cells suggesting the potential for an overall gain in antitumor cytotoxic potential when coupled with our Granzyme B observations.18 This data demonstrates that the improved cellular features of the T cell compartment observed here are likely accompanied by increased functionality in the presence of capmatinib rather than cabozantinib.

    Discussion

    Given the pivotal role of c-Met signaling in both HCC tumorigenesis and microenvironmental immune suppression,1 8 19 significant potential exists for c-Met inhibition to cooperate with PD-1 ICB to achieve synergistic, or at least cooperative, antitumor activity. While prior publications using PD-1 sensitive models of HCC suggested this might be possible through combination with the type II inhibitor cabozantinib,9 clinical outcomes for this combination were at best inferior to other HCC therapeutic combinations (ie, atezolizumab and bevacizumab),7 and, at worst, no better than PD-1 blockade alone.6 The reality is that combinations with type II MET inhibitors and checkpoint blockade in human HCC have fallen short of the therapeutic benefit of current FDA-approved options.20 Here, we demonstrate a unique advantage for the highly c-Met specific inhibitor capmatinib in potentiating anti-PD-1 activity against HCC across three pre-clinical models including the immune checkpoint-resistant DEN model.

    Capmatinib augments both effector T cell frequency and pro-inflammatory phenotype versus PD-1 blockade alone or versus the cabozantinib combination, and, uniquely, also reduces suppressive myeloid stromal phenotype while increasing the activation state of tumor resident DCs. Accompanying in vitro studies show that type II and III MET inhibitors also compromise T cell effector cytokine production and cytotoxic potential, whereas the type I inhibitor capmatinib has only minimal impact. Even in the PD-1 refractory DEN HCC model, the type I c-Met inhibitor capmatinib combination cures nearly half of animals demonstrating true therapeutic synergy. The failure of the type II inhibitor cabozantinib to replicate these outcomes may be due to its effects on the granulocyte compartment. A prior study showed that c-Met activation can mediate PD-1 resistance through mobilization of neutrophils that become GrMDSC in the TIME and suppress T cell responses.8 Neutrophils can adopt both protumor and antitumor phenotypes in the tumor microenvironment; however, MET can cooperate to activate STAT3 which promotes GrMDSC differentiation directly and indirectly through driving interleukin (IL)-6 secretion.21 In this context, it has been shown that cabozantinib alone and in combination with PD-1 blockade also augments neutrophil infiltration into HCC tumors.9 In contrast, we find decreased frequencies of immunosuppressive, Arginase 1+GrMDSC when capmatinib and PD-1 antibody are used together suggesting selective blockade of signals promoting immunosuppressive neutrophil polarization. An important caveat of these mouse experiments is that they were performed entirely in male mice as all three HCC tumor models used were originally derived from male mice and cannot be used in female mice without engendering potentially confounding anti-Y immune responses. Thus these studies would not be capable of revealing sex-specific differences in therapeutic response.

    A recently reported study (NCT02323126) in non-small cell lung cancer (NSCLC) met the primary endpoint of prolonged 6-month progression free survival in both the high-MET expressing, and, significantly, also the low-MET expressing subgroups.22 This significant finding shows the potential of PD-1 blockade with capmatinib in combination to surpass the clinical efficacy of either drug alone. In contrast to this finding, however, a recent study (NCT02795429) of spartalizumab and capmatinib failed to show any advantage for the combination in HCC.23 It is noteworthy that in this study, spartalizumab, a PD-1 antibody never approved by the FDA, delivered monotherapy efficacy of only 10% which is about half of the expected objective response rate for similar FDA-approved drugs in this patient cohort.24 Also, the optimal 600 mg two times per day capsule dose of capmatinib was not used with 400 mg two times per day tablets used instead.25 Further, this trial was performed in already sorafenib-resistant patients which confounds both the subsequent response to MET inhibition and reduces the efficacy of PD-1 blockade.24 26 27 The setting of this trial is also one that is no longer relevant in HCC as any PD-1 blockade and MET inhibitor combination would be considered a front-line therapy in the modern treatment paradigm. Overall, this study emphasizes the importance of conducting a type I MET inhibitor (eg, capmatinib, savolitinib) study in combination with an FDA-approved PD-1 inhibitor with demonstrated efficacy in HCC so that the benefit now seen in NSCLC can be realized in HCC.

    Of course, a relevant consideration for any combination of MET inhibition and ICB in HCC is toxicity. Across our three mouse models of HCC treated with MET inhibition and PD-1 blockade we did observe gross signs of toxicity in terms of body weight changes>10%, behavioral changes (eg, poor feeding, hunching) or changes in appearance. The clinical data on these combinations is most important, however, in assessing the suitability of this therapy for patients with HCC. For the capmatinib and nivolumab combination in NSCLC, in which∼one-third of patients had liver metastases,22 there were no drug-associated deaths, but adverse events were common with approximately one-third of patients discontinuing one or both drugs due to toxicity. While this is high, the investigators found the toxicity manageable to the point that there was a highly significant therapeutic benefit across the cohort. Again, arguing against the viability of spartalizumab combinations versus those with FDA-approved PD-1 drugs, a similar NSCLC study of capmatinib with spartalizumab was discontinued due to the high rate of adverse events.28 Overall, the NSCLC experience suggests that, while the capmatinib and nivolumab combination would produce a high rate of adverse events, those should be of a manageable grade and frequency.

    These findings provide a compelling rationale for the combination of type I c-Met inhibitors, which are already FDA-approved for patients with lung cancer with c-Met mutations, and PD-1 pathway blockade in new clinical studies for the therapy of patients with HCC. The activity of this combination in both the PD-1 sensitive and refractory HCC mouse models suggests the potential utility of this combination in both first-line and second-line, post-αPD-(L)1 failure, settings.

    Methods

    Animals

    Six to 8-week-old C57BL/6 and C3H/HeJ mice were purchased from The Jackson Laboratory (Bar Harbor, ME). All procedures were conducted in accordance with the guidelines established by the University of Texas MD Anderson Cancer Center Institutional Animal Care and Use Committee and under protocol 00001378-RN02.

    Cell lines and reagents

    The murine HCC cell line Hepa1-6 was purchased from American Type Culture Collection (CRL-1830). HCA-1 was a kind gift from Dr Dan Duda’s laboratory where it was created. The DEN HCC 1B2-2 in vivo serial passage line was obtained from Dr Angelica Tolentino at Ionis and maintained in male C57BL6 mice. The tumors are carried subcutaneously in male C57BL6 mice and when they reach greater than 500 mm3, they are resected, cut into 2 mm x 2 mm slices and implanted subcutaneously into 4–6 week old naïve C57BL6 male mice. Tumors generally become palpable 30–45 days after slice implantation.

    Therapeutic antibodies

    Therapeutic antibodies targeting PD-1 (RMP1-14 at 250 µg per dose) were purchased from Leinco at<1 endotoxin unit/mg of lipopolysaccharide.

    Tumor challenge and therapeutic antibody administration

    Mice received a primary tumor challenge of 1.×106 Hepa1-6 (C57BL6) or HCA-1 (C3H) cells s.c. on the right flank delivered in phosphate-buffered saline (and 30% matrigel for TIME analysis studies). Mice received either 10 mg/kg capmatinib or 30 mg/kg cabozantinib by oral gavage on days 5, 9, 13, 17, and 21 and/or 250 ug of αPD-1 clone RMP1-14 (Leinco) intraperitoneally on days 7, 11, 15 and 19. For survival mice were followed with caliper measurements and euthanized once tumors exceeded 1000 mm3. For analysis, tumors were harvested on day 22. For the DEN model, tumors were isolated from the flanks of DEN tumor-bearing C57BL6 mice and cut into (what size) slices ex vivo and then implanted s.c. on the flank of naïve 4–6 week old mice. DEN mice were treated when tumors became palpable (∼ day 45) with 10 days of c-Met inhibitor (concentrations as above) and/or αPD-1 on days 2, 4, 6, 8, 10 at 250 ug/injection.

    Cell isolation

    At various time points throughout the experiment, post–s.c.-tumor challenge, mice were sacrificed, and various tissues were harvested to characterize cell-mediated antitumor responses. Briefly, tumors and solid tissues were digested in X-Vivo 15 (Lonza) supplemented with Collagenase H (Sigma) and DNase (Roche) and incubated at 37°C, 5% CO2 for 30 min before being filtered through a 70 µm cell strainer. T cells were then enriched through density gradient separation over Histopaque 1119 (Sigma).

    Flow cytometry analysis

    Tumors were digested in X-Vivo 15 (Lonza) supplemented with Collagenase H (Sigma) and DNase (Roche) and incubated at 37°C, 5% CO2 for 30 min before being filtered through a 70 µm cell strainer. T cells and myeloid cells were then enriched through density gradient separation over Histopaque 1119 (Sigma) at 1000 g. Samples were fixed using the Foxp3/Transcription Factor Staining Buffer Set (Thermo Fisher) and then stained with 24 antibodies (online supplemental table 1) at a time from BioLegend, BD Biosciences, and Thermo Fisher. Flow data were collected on a five-laser, 18-color BD Biosciences LSR II or a 5-laser 30-color BD Symphony A3 cytometer and analyzed using FlowJo V.7.6.5 or 10.X (Tree star).

    OT-I T cell cytokine production

    Splenic OT-I T cells were harvested from C57BL/6-Tg(TcraTcrb)1100Mjb/J (#003831) mice and activated for 72 hours in RPMI with 10% fetal bovine serum (FBS), 1% penicillin-streptomycin, 1% L-glutamine, 50 u/mL murine IL-2 (R&D Systems), 1 ug/mL OVA257-264 (SIINFEKL) peptide, 50 µM β-mercaptoethanol, and varied concentrations of either capmatinib, cabozantinib, or tivantinib, respectively, in a 96-well plate. Secreted cytokines in the supernatant were measured using the CBA Mouse-Th1/Th2/Th17-Cytokine-Kit (BD Biosciences).

    OT-I T cell proliferation assay

    Splenic OT-I T cells were harvested from C57BL/6-Tg(TcraTcrb)1100Mjb/J (#003831) mice and fluorescently labeled using the CellTrace CFSE Cell Proliferation Kit (Thermo Fisher). Labeled cells were then activated for 72 hours in RPMI with 10% FBS, 1% penicillin-streptomycin, 1% L-glutamine, 50 u/mL murine IL-2 (R&D Systems), 1 ug/mL OVA257-264 (SIINFEKL) peptide, 50 µM β-mercaptoethanol, and varied concentrations of either capmatinib, cabozantinib, or tivantinib respectively in a 96-well plate. After 72 hours, CD3+CD8+ cells were analyzed for CFSE fluorescence by flow cytometry on a BD Biosciences LSR II.

    Statistics

    All statistics were calculated using GraphPad Prism software. Statistical significance was determined using a Student’s t-test for comparisons between treatment groups or a paired Student’s t-test for comparison of cell populations within the same treatment group unless otherwise indicated. Graphs show mean±SEM unless otherwise indicated. P values of<0.05 were considered significant.

    Data availability statement

    Data are available upon reasonable request. Data were generated by the authors and included in the article. Raw data supporting the findings in this study can be obtained from the author upon reasonable request.

    Ethics statements

    Patient consent for publication

    Ethics approval

    Not applicable.

    References

    1. ↵
      1. Gherardi E,
      2. Birchmeier W,
      3. Birchmeier C, et al
      . Targeting MET in cancer: rationale and progress. Nat Rev Cancer 2012;12:89–103. doi:10.1038/nrc3205
      OpenUrlCrossRefPubMedWeb of Science
    2. ↵
      1. Goyal L,
      2. Muzumdar MD,
      3. Zhu AX
      . Targeting the HGF/c-MET pathway in hepatocellular carcinoma. Clin Cancer Res 2013;19:2310–8. doi:10.1158/1078-0432.CCR-12-2791
      OpenUrlAbstract/FREE Full Text
    3. ↵
      1. Fogli S,
      2. Tabbò F,
      3. Capuano A, et al
      . The expanding family of c-Met inhibitors in solid tumors: a comparative analysis of their pharmacologic and clinical differences. Crit Rev Oncol Hematol 2022;172:103602. doi:10.1016/j.critrevonc.2022.103602
      OpenUrl
    4. ↵
      1. Hong L,
      2. Zhang J,
      3. Heymach JV, et al
      . Current and future treatment options for MET exon 14 skipping alterations in non-small cell lung cancer. Ther Adv Med Oncol 2021;13:1758835921992976. doi:10.1177/1758835921992976
    5. ↵
      1. Li H,
      2. Li C-W,
      3. Li X, et al
      . MET Inhibitors Promote Liver Tumor Evasion of the Immune Response by Stabilizing PDL1. Gastroenterology 2019;156:1849–61. doi:10.1053/j.gastro.2019.01.252
      OpenUrlCrossRef
    6. ↵
      1. Yau T,
      2. Zagonel V,
      3. Santoro A, et al
      . Nivolumab Plus Cabozantinib With or Without Ipilimumab for Advanced Hepatocellular Carcinoma: Results From Cohort 6 of the CheckMate 040 Trial. J Clin Oncol 2023;41:1747–57. doi:10.1200/JCO.22.00972
      OpenUrl
    7. ↵
      1. Kelley RK,
      2. Rimassa L,
      3. Cheng A-L, et al
      . Cabozantinib plus atezolizumab versus sorafenib for advanced hepatocellular carcinoma (COSMIC-312): a multicentre, open-label, randomised, phase 3 trial. Lancet Oncol 2022;23:995–1008. doi:10.1016/S1470-2045(22)00326-6
      OpenUrlCrossRefPubMed
    8. ↵
      1. Glodde N,
      2. Bald T,
      3. van den Boorn-Konijnenberg D, et al
      . Reactive Neutrophil Responses Dependent on the Receptor Tyrosine Kinase c-MET Limit Cancer Immunotherapy. Immunity 2017;47:789–802. doi:10.1016/j.immuni.2017.09.012
      OpenUrl
    9. ↵
      1. Esteban-Fabró R,
      2. Willoughby CE,
      3. Piqué-Gili M, et al
      . Cabozantinib Enhances Anti-PD1 Activity and Elicits a Neutrophil-Based Immune Response in Hepatocellular Carcinoma. Clin Cancer Res 2022;28:2449–60. doi:10.1158/1078-0432.CCR-21-2517
      OpenUrl
    10. ↵
      1. Jin Y,
      2. An X,
      3. Mao B, et al
      . Different syngeneic tumors show distinctive intrinsic tumor-immunity and mechanisms of actions (MOA) of anti-PD-1 treatment. Sci Rep 2022;12:3278. doi:10.1038/s41598-022-07153-z
    11. ↵
      1. Chen Y,
      2. Huang Y,
      3. Reiberger T, et al
      . Differential effects of sorafenib on liver versus tumor fibrosis mediated by stromal-derived factor 1 alpha/C-X-C receptor type 4 axis and myeloid differentiation antigen-positive myeloid cell infiltration in mice. Hepatology 2014;59:1435–47. doi:10.1002/hep.26790
      OpenUrlCrossRefPubMed
    12. ↵
      1. Kim K-J,
      2. Kim J-H,
      3. Lee SJ, et al
      . Radiation improves antitumor effect of immune checkpoint inhibitor in murine hepatocellular carcinoma model. Oncotarget 2017;8:41242–55. doi:10.18632/oncotarget.17168
      OpenUrl
    13. ↵
      1. Qiu L,
      2. Ji H,
      3. Wang K, et al
      . TLR3 activation enhances abscopal effect of radiotherapy in HCC by promoting tumor ferroptosis. EMBO Mol Med 2024;16:1193–219. doi:10.1038/s44321-024-00068-4
      OpenUrl
    14. ↵
      1. Hu B,
      2. Yu M,
      3. Ma X, et al
      . IFNα Potentiates Anti–PD-1 Efficacy by Remodeling Glucose Metabolism in the Hepatocellular Carcinoma Microenvironment. Cancer Discov 2022;12:1718–41. doi:10.1158/2159-8290.CD-21-1022
      OpenUrl
    15. ↵
      1. Ishii K,
      2. Shimizu M,
      3. Kogo H, et al
      . A combination of check-point blockade and α-galactosylceramide elicits long-lasting suppressive effects on murine hepatoma cell growth in vivo. Immunobiology 2020;225:151860. doi:10.1016/j.imbio.2019.10.009
    16. ↵
      1. Tolba R,
      2. Kraus T,
      3. Liedtke C, et al
      . Diethylnitrosamine (DEN)-induced carcinogenic liver injury in mice. Lab Anim 2015;49:59–69. doi:10.1177/0023677215570086
      OpenUrlCrossRefPubMed
    17. ↵
      1. Leslie J,
      2. Mackey JBG,
      3. Jamieson T, et al
      . CXCR2 inhibition enables NASH-HCC immunotherapy. Gut 2022;71:2093–106. doi:10.1136/gutjnl-2021-326259
      OpenUrlAbstract/FREE Full Text
    18. ↵
      1. Kubo Y,
      2. Fukushima S,
      3. Inamori Y, et al
      . Serum concentrations of HGF are correlated with response to anti-PD-1 antibody therapy in patients with metastatic melanoma. J Dermatol Sci 2019;93:33–40. doi:10.1016/j.jdermsci.2018.10.001
      OpenUrl
    19. ↵
      1. Agarwalla P,
      2. Barnard Z,
      3. Fecci P, et al
      . Sequential immunotherapy by vaccination with GM-CSF-expressing glioma cells and CTLA-4 blockade effectively treats established murine intracranial tumors. J Immunother 2012;35:385–9. doi:10.1097/CJI.0b013e3182562d59
      OpenUrlCrossRefWeb of Science
    20. ↵
      1. Sangro B,
      2. Sarobe P,
      3. Hervás-Stubbs S, et al
      . Advances in immunotherapy for hepatocellular carcinoma. Nat Rev Gastroenterol Hepatol 2021;18:525–43. doi:10.1038/s41575-021-00438-0
      OpenUrlCrossRefPubMed
    21. ↵
      1. Giese MA,
      2. Hind LE,
      3. Huttenlocher A
      . Neutrophil plasticity in the tumor microenvironment. Blood 2019;133:2159–67. doi:10.1182/blood-2018-11-844548
      OpenUrlAbstract/FREE Full Text
    22. ↵
      1. Felip E,
      2. Metro G,
      3. Tan DSW, et al
      . Capmatinib plus nivolumab in pretreated patients with EGFR wild-type advanced non-small cell lung cancer. Lung Cancer (Auckl) 2024;192:107820. doi:10.1016/j.lungcan.2024.107820
      OpenUrl
    23. ↵
      1. Santoro A,
      2. Assenat E,
      3. Yau T, et al
      . A phase Ib/II trial of capmatinib plus spartalizumab vs. spartalizumab alone in patients with pretreated hepatocellular carcinoma. JHEP Rep 2024;6:101021. doi:10.1016/j.jhepr.2024.101021
    24. ↵
      1. Zhu AX,
      2. Finn RS,
      3. Edeline J, et al
      . Pembrolizumab in patients with advanced hepatocellular carcinoma previously treated with sorafenib (KEYNOTE-224): a non-randomised, open-label phase 2 trial. Lancet Oncol 2018;19:940–52. doi:10.1016/S1470-2045(18)30351-6
      OpenUrlCrossRefPubMed
    25. ↵
      1. Qin S,
      2. Chan SL,
      3. Sukeepaisarnjaroen W, et al
      . A phase II study of the efficacy and safety of the MET inhibitor capmatinib (INC280) in patients with advanced hepatocellular carcinoma. Ther Adv Med Oncol 2019;11:1758835919889001. doi:10.1177/1758835919889001
    26. ↵
      1. Tang W,
      2. Chen Z,
      3. Zhang W, et al
      . The mechanisms of sorafenib resistance in hepatocellular carcinoma: theoretical basis and therapeutic aspects. Sig Transduct Target Ther 2020;5:87. doi:10.1038/s41392-020-0187-x
      OpenUrl
    27. ↵
      1. Khalaf AM,
      2. Fuentes D,
      3. Morshid AI, et al
      . . JHC 2018;Volume 5:61–73. 10.2147/JHC.S156701
    28. ↵
      1. Wolf J,
      2. Heist R,
      3. Kim TM, et al
      . 994P Efficacy and safety of capmatinib plus spartalizumab in treatment-naïve patients with advanced NSCLC harboring MET exon 14 skipping mutation. Ann Oncol 2022;33:S1007–8. doi:10.1016/j.annonc.2022.07.1120
      OpenUrl

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Footnotes

    • Contributors Conceptualization: MAC; methodology: RAD, MS, MAC; investigation: RAD, MS, BXT, AL, SN, and MAC; formal analysis: RAD, MS and MAC; resources, financial support, or coordinated research activity: JS, RF, AT, AOK and MAC; supervision: MAC; writing—original draft: MAC; writing—review and editing: AOK and MAC. Study guarantor: MAC.

    • Funding This work was supported by The University of Texas MD Anderson Cancer Center SPORE in Hepatocellular Carcinoma 1 P50 CA217674-01A1. SN was supported by the UT MD Anderson Summer Undergraduate Experience Program.

    • Competing interests MAC reports grants and personal fees from ImmunoGenesis, personal fees from Alligator Bioscience, ImmunOs, OncoResponse, Kineta, Xencor, Agenus, AstraZeneca outside the submitted work; In addition, MAC has a patent “Dual specificity antibodies which bind both PD-L1 and PD-L2 and prevent their binding to PD-1” with royalties paid by ImmunoGenesis.

    • 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.