Major histocpmpatibilty complex class I (MHC I) antigen presentation allows CD8+ T cells to detect and eliminate cancerous or virally infected cells. The MHC I pathway is not essential for cell growth and viability and consequently cancers and viruses can evade control by CD8+ T cells by inactivating antigen presentation. In cancers, two common ways for this evasion are the loss of either the MHC I light chain (ß2M) or the cytosol-to-endoplasmic reticulum (ER) peptide transporter (TAP). ß2M-null cells are generally thought to lack the MHC I pathway because the MHC I heavy chain by itself lacks the proper conformation for peptide display. TAP-null cells are thought to have severely defective MHC I antigen presentation because they are incapable of supplying peptides from the cytosol to MHC I molecules in the ER. However, we have found that highly reactive memory CD8+ T cells could still recognize cells that completely lacked ß2M or TAP. This was at least in part because in TAPnull cells, the Sec62 component of the Sec61 translocon supported the transfer of cytosolic peptides into the ER. In ß2M-negative cells, free MHC I heavy chains were able to bind peptides and assume a conformation that was sufficiently recognized by CD8+ T cells. This process required ER chaperones and the peptide-loading complex. We found that these mechanisms supported antigen presentation at a level that was sufficient for memory CD8+ T cells to kill melanoma cells both in vitro and in tumor-bearing mice. The implications for tumor immunotherapy are discussed.

Immune checkpoint inhibitors (ICIs) have revolutionized the treatment of unresectable hepatocellular carcinoma (HCC), but their impressive efficacy is seen in just a fraction of patients. One key mechanism of immunotherapy resistance is the paucity of dendritic cells (DCs) in liver malignancies. Here, we tested combination blockade of programmed death receptor 1 (PD1) and CXCR4, a receptor for CXCL12, a pleiotropic factor that mediates immunosuppression in tumors. Using orthotopic grafted and autochthonous HCC models with underlying liver damage, we evaluated treatment feasibility and efficacy. In addition, we examined the effects of treatment using immunofluorescence, flow cytometric analysis of DCs in vivo and in vitro, and RNA-sequencing. Combination anti-CXCR4 and anti-PD1therapy was safe and significantly inhibited tumor growth and prolonged survival in all murine preclinical models of HCC tested. The combination treatment successfully reprogrammed antigen-presenting cells, revealing the potential role of conventional type 1 DCs (cDC1s) in the HCC microenvironment. Moreover, DC reprogramming enhanced anticancer immunity by facilitating CD8+ T-cell accumulation and activation in the HCC tissue. The effectiveness of anti-CXCR4/PD1 therapy was compromised entirely in Batf3-KO mice deficient in cDC1s. Thus, combined CXCR4/PD1 blockade can reprogram intra-tumoral cDC1s and holds the potential to potentiate antitumor immune response against HCC.

Anti-PD-1, trastuzumab, and chemotherapy are used in the treatment of patients with advanced HER2-positive esophagogastric adenocarcinoma (EGA), but long-term survival remains limited. Herein, we report extended follow-up data from the INTEGA trial (NCT03409848), which investigated the efficacy of the anti-PD-1 nivolumab, trastuzumab, and FOLFOX chemotherapy (FOLFOX arm) in comparison to a chemotherapy-free regimen involving nivolumab, trastuzumab, and the anti-CTLA-4 ipilimumab (Ipi arm) in the first-line setting for advanced disease. The 12-month overall survival (OS) showed no statistical difference between the arms, with 57% OS (95% CI: 41%-71%) in the Ipi arm and 70% OS (95% CI: 54%-82%) in the FOLFOX arm. Crossing of the survival curves indicated a potential long-term benefit for some patients within the Ipi arm, but early progressors in the Ipi arm underlined the need for biomarker-guided strategies to optimize treatment selection. To this end, metabolomic and cytokine analysis demonstrated elevated levels of normetanephrine, cortisol, and interleukin 6 (IL-6) in immunotherapy-unresponsive patients in the Ipi arm, suggesting a role for systemic inflammatory stress in modulating antitumor immune responses. Patients with this signature also showed an increased neutrophil-to-lymphocyte ratio (NLR) that persisted in the Ipi arm, but not in the FOLFOX arm, and strongly correlated with survival. Furthermore, a low NLR characterized patients benefiting from immune- and targeted therapy without the need for additional chemotherapy. This data suggests that patient selection based on inflammatory stress-driven immune changes could help to customize first-line treatment in patients with advanced HER2-positive EGA to potentially improve long-term survival.

Epithelial ovarian cancer (EOC) is the most common type of ovarian cancer with a low rate of response to immunotherapy such as immune checkpoint blockade (ICB) therapy. Here, we report that nucleus accumbens-associated protein 1 (NAC1), a putative driver of EOC, has a critical role in immune evasion. We showed in murine ovarian cancer models that depleting or inhibiting tumoral NAC1 reduced the recruitment and immunosuppressive function of myeloid-derived suppressor cells (MDSCs) in the tumor microenvironment (TME), led to significant increases of cytotoxic tumor-infiltrating CD8+ T cells, and promoted antitumor immunity and suppressed tumor progression. We further showed that tumoral NAC1 directly enhanced the transcription of CXCL16, by binding to CXCR6, thereby promoting MDSC recruitment to the tumor. Moreover, lipid C20:1T produced by NAC1-expressing tumor cells fueled oxidative metabolism of MDSCs and promoted their immune-suppressive function. We also showed that NIC3, a small molecule inhibitor of NAC1, was able to sensitize mice-bearing NAC1-expressing ovarian tumors to anti-PD-1 therapy. Our study reveals a critical role for NAC1 in controlling tumor infiltration of MDSCs and in modulating the efficacy of ICB therapy. Thus, targeting of NAC1 may be exploited to sensitize ovarian cancer to immunotherapy.

Natural killer (NK) cell tumor infiltration is associated with good prognosis in patients with metastatic castration-resistant prostate cancer (mCRPC). NK cells recognize and kill targets by a process called natural cytotoxicity. We hypothesized that promoting an antigen-specific synapse with co-activation may enhance NK cell function in mCRPC. We describe a Tri-specific Killer Engager (TriKE) construct that engages with the activating receptor CD16 on NK cells, prostate-specific membrane antigen (PSMA) on mCRPC cells, and has an interleukin (IL)-15 moiety that is essential for NK cell survival, proliferation, and priming. We show that the PSMA TriKE specifically binds to PSMA-expressing cells and significantly enhances expansion, degranulation and cytokine production of NK cells derived from healthy donors or prostate cancer patients. Bystander killing of PSMA-negative was also achieved with PSMA TriKE treatment when co-cultured with PSMA-positive cells, suggesting potential PSMA TriKE benefit in controlling tumor antigen escape. When tested under physiologic conditions recapitulating the mCRPC tumor microenvironment (TME), NK cells treated with PSMA TriKE and prolonged exposure to hypoxia or MDSCs maintained their potent function while IL-15 treated NK cells showed greatly impaired cytotoxicity. Finally, in vivo testing of PSMA TriKE showed improved tumor control and survival of mice as compared to IL-15 and untreated control groups. In conclusion, PSMA TriKE demonstrates potential as a new therapy for advanced prostate cancer by providing additional signals to NK cells to maximize their anti-tumor potential in prostate cancer, especially in the setting of a hostile TME.

Neoantigen-targeted therapy holds an array of benefits for cancer immunotherapy, but the identification of peptide targets with tumor rejection capacity remains a limitation. To better define the criteria dictating tumor rejection potential, we examined the capacity of high-magnitude T cell responses induced towards several distinct neoantigen targets to regress MC38 tumors. Surprisingly, despite their demonstrated immunogenicity, vaccine-induced T-cell responses were unable to regress established MC38 tumors or prevent tumor engraftment in a prophylactic setting. However, T cells were functional with robust killing capacity towards neoantigen peptide-loaded cells. Furthermore, tumor cell killing was rescued in proportion to the expression level or saturation of target peptide-loaded MHCs on the cell surface. Overall, this study demonstrates a pivotal role for target protein expression levels in modulating the tumor rejection capacity of neoantigens. Thus, inclusion of this metric, in addition to immunogenicity analysis, may benefit antigen prediction techniques to ensure the full anti-tumor effect of cancer vaccines.

The survival rate of glioma patients has not significantly increased in recent years despite aggressive treatment and advances in immunotherapy. The limited response to treatments is partially attributed to the immunosuppressive tumor microenvironment, where regulatory T cells (Tregs) play a pivotal role in immunological tolerance. In this study, we investigated the impact of complement factor H (FH) on Tregs within the glioma microenvironment and found that FH is an ICOS ligand. The binding of FH to this immune checkpoint molecule promoted the survival and function of Tregs and induced the secretion of TGF-beta (TGF-β) and IL-10, while also suppressing T-cell proliferation. We further demonstrated that cancer cells in human and mouse gliomas directly produce FH. Database investigations revealed that upregulation of FH expression was associated with the presence of Tregs and correlated with worse prognosis for glioma patients. We confirmed the effect of FH on glioma development in a mouse model, where FH knockdown was associated with decrease in number of ICOS+ Tregs and demonstrated a tendency of prolonged survival (p=0.064). Since the accumulation of Tregs represents a promising prognostic and therapeutic target, evaluating FH expression should be considered when assessing the effectiveness of and resistance to immunotherapies against glioma.

Resistance to immune checkpoint inhibitors (ICIs) is common, even in tumors with T cell infiltration. We thus investigated consequences of ICI-induced T cell infiltration in the microenvironment of resistant tumors. T cells and neutrophil numbers increased in ICI-resistant tumors following treatment, in contrast to ICI-responsive tumors. Resistant tumors were distinguished by high expression of IL-1 Receptor 1 (IL1R1), enabling a synergistic response to IL-1 and TNFα to induce G-CSF, CXCL1, and CXCL2 via NF-κB signaling, supporting immunosuppressive neutrophil accumulation in tumor. Perturbation of this inflammatory resistance circuit sensitized tumors to ICIs. Paradoxically, T cells drove this resistance circuit via TNF both in vitro and in vivo. Evidence of this inflammatory resistance circuit and its impact also translated to human cancers. These data support a mechanism of ICI resistance, wherein treatment-induced T cell activity can drive resistance in tumors responsive to IL-1 and TNFα, with important therapeutic implications.

B-cell maturation antigen (BCMA) chimeric antigen receptor (CAR) T cell therapy has been approved for the treatment of relapsed and refractory multiple myeloma (RRMM); however, whether patients have long-term response has yet to be established. We investigated the feasibility of CBG-002, an CD27-armored BCMA CAR T therapy, to improve clinical efficacy in patients with RRMM. We present preclinical data showing the activity of CBG-002 against myeloma and results from a phase I clinical trial (NCT04706936) evaluating its safety and efficacy in patients with RRMM. The primary endpoint was safety, as assessed by grade 3 or 4 adverse events(AEs). Key secondary endpoints were overall response rate (ORR), duration of response (DOR), progression-free survival (PFS) and overall survival (OS). A total of 11 patients were enrolled and received CBG-002 therapy. Nine patients developed grade 1 or 2 cytokine release syndrome (CRS), while no patients experienced grade 3 or higher CRS or immune effector cell-associated neurotoxicity syndrome. Other grade 3 or higher AEs included neutropenia (72.7%), thrombocytopenia (45.5%) and anemia (36.4%). At a median follow-up of 16.7 months, the ORR was 81.8%, including a stringent complete response/complete response rate of 45.5%, very good partial response rate of 18.2%, and partial response rate of 18.2%, with a median DOR of 8.9 (range 1.8-21.9) months. The median OS was not reached, and the median PFS was 8.5 (2.7-22.9) months. In this phase I study, CBG-002, a CD27-armored BCMA CAR T therapy, demonstrated safety and clinical efficacy in patients with RRMM.

Glutamine is a major energy source for tumor cells and blocking glutamine metabolism is being investigated as a promising strategy for cancer therapy. However, the antitumor effect of glutamine blockade in bladder cancer remains unclear, necessitating further investigation. Here, we demonstrated that glutamine metabolism was involved in the malignant progression of bladder cancer. Treatment with the glutamine antagonist 6-Diazo-5-oxo-L-norleucine (DON) inhibited the growth of bladder cancer cells in vitro in several ways. In addition, we observed inhibition of tumor growth in bladder cancer-bearing mice using JHU083, a prodrug that was designed to prevent DON-induced toxicity. However, the antitumor immune effect of T cells changed from activation to inhibition as the administrated time extended. We found that both in vitro treatment with DON and in vivo prolonged administration of JHU083 led to the upregulation of PD-L1 in bladder cancer cells. Mechanistically, glutamine blockade up-regulated PD-L1 expression in bladder cancer cells by accumulating ROS, subsequently activating the EGFR/ERK/C-Jun signaling pathway. Combination treatment of JHU083 and gefitinib reversed the up-regulation of PD-L1 in bladder cancer cells induced by prolonged glutamine blockade, resulting in the alleviation of T-cell immunosuppression and a significant improvement in therapeutic outcome. These preclinical findings show promise for glutamine metabolism targeting as a viable therapeutic strategy for bladder cancer, with the potential for further enhancement through combined treatment with gefitinib.

γδ T cells have recently raised great interest as effector cells in cancer immunotherapy because of their HLA-independent mode of action and their broad tumor reactivity. To translate the application of γδ T cells into clinically effective immunotherapies, specific tumor targeting and/or boosting of γδ T-cell activation in vivo seem to be a critical step. In this issue, Le Floch and colleagues report a new strategy for enabling γδ T cells to be specifically activated to kill acute lymphoblastic leukemia cells and solid tumor cells using agonistic BTN2A1 antibodies. See related article by Le Floch et al., p. XX .

The histone methyltransferase enhancer of zeste homolog 2 (EZH2) plays important roles in T-cell differentiation, proliferation and function. Previous studies have demonstrated that genetic deletion of EZH2 in CD8+ or total T cells impairs their antiviral and antitumor activity, cytokine production and ability to expand upon rechallenge. Contrary to the detrimental role of deleting T cell-intrinsic EZH2, here we have demonstrated that transient inhibition of EZH2 in T cells prior to the phenotypic onset of exhaustion with a clinically approved inhibitor, Tazemetostat, delayed their dysfunctional progression and preserved T-cell stemness and polyfunctionality but had no negative impact on cell proliferation. Tazemetostat induced T-cell epigenetic reprogramming and increased the expression of the self-renewal T-cell transcription factor TCF1 by reducing H3K27 methylation at its promoter preferentially in rapidly dividing T cells. In a murine melanoma model, T cells depleted of EZH2 induced poor tumor control, whereas adoptively transferred T cells pretreated with tazemetostat exhibited superior antitumor immunity, especially when used in combination with anti-PD-1 blockade. Collectively, these data highlight the potential of transient epigenetic reprogramming by EZH2 inhibition to enhance adoptive T-cell immunotherapy.

The T cell antigen coupler (TAC) is a chimeric receptor that facilitates tumor antigen-specific activation of T cells by co-opting the endogenous T cell receptor complex in the absence of tonic signaling. Previous data demonstrates that TAC affords T cells with the ability to induce durable and safe anti-tumor responses in preclinical models of hematological and solid tumors. Here, we describe the preclinical pharmacology and safety of an autologous Claudin 18.2 (CLDN18.2)-directed TAC T cell therapy, TAC01-CLDN18.2, in preparation for a Phase I/II clinical study in subjects with CLDN18.2-positive solid tumors. Following a screen of putative TAC constructs, the specificity, activity, and cytotoxicity of TAC T cells expressing the final CLDN18.2-TAC receptor were evaluated in vitro and in vivo using gastric, gastroesophageal, and pancreatic tumor models as well as human cells derived from normal tissues. CLDN18.2-specific activity and cytotoxicity of CLDN18.2-TAC T cells were observed in coculture with various 2D tumor cultures naturally expressing CLDN18.2 as well as tumor spheroids. These effects occurred in models with low antigen levels and was positively associated with increasing CLDN18.2 expression. CLDN18.2-TAC T cells effectively eradicated established tumor xenografts in mice in the absence of observed off-target or on-target/off-tumor effects, elicited durable efficacy in recursive killing and tumor rechallenge experiments, and remained unreactive in coculture with human cells representing vital organs. Thus, the data demonstrate that CLDN18.2-TAC T cells can induce a specific and long-lasting anti-tumor response in various CLDN18.2-positive solid tumor models without notable TAC-dependent toxicities, supporting the clinical development of TAC01-CLDN18.2.

Neutrophils are the primary myeloid cells that are recruited to inflamed tissues, and they are key players during colitis, being also present within the tumor microenvironment during the initiation and growth of colon cancer. Neutrophils fundamentally serve to protect the host against microorganism invasion, but during cancer development, they can become protumoral and lead to tumor initiation, growth, and eventually, metastasis-hence, playing a dichotomic role for the host. Protumoral neutrophils in cancer patients can be immunosuppressive and serve as markers for disease progression but their characteristics are not fully defined. In this review, we explore the current knowledge on how neutrophils in the gut fluctuate between an inflammatory or immunosuppressive state and how they contribute to tumor development. We describe neutrophils' antitumoral and protumoral effects during inflammatory bowel diseases and highlight their capacity to provoke the advent of inflammation-driven colorectal cancer. We present the functional ambivalence of the neutrophil populations within the colon tumor microenvironment, which can be potentially exploited to establish therapies that will prevent, or even reverse, inflammation-dependent colon cancer incidence in high-risk patients.

Accumulating evidence indicates that the gut microbiome influences cancer progression and therapy. We recently showed that progressive changes in gut microbial diversity and composition are closely associated with tobacco-associated lung adenocarcinoma (LUAD) in a human-relevant mouse model. Furthermore, we demonstrated that the loss of the antimicrobial protein Lcn2 in these mice, exacerbates pro-tumor inflammatory phenotypes while further reducing microbial diversity. Yet, how gut microbiome alterations impinge on LUAD development remains poorly understood. Here, we investigated the role of gut microbiome changes in LUAD development using fecal microbiota transfer and delineated a pathway by which gut microbiome alterations incurred by loss of Lcn2 fostered the proliferation of pro-inflammatory bacteria of the genus Alistipes, triggering gut inflammation. This inflammation propagated systemically, exerting immunosuppression within the tumor microenvironment, augmenting tumor growth through an IL-6-dependent mechanism and dampening response to immunotherapy. Corroborating our preclinical findings, we found that patients with LUAD with a higher relative abundance of Alistipes species in the gut showed diminished response to neoadjuvant immunotherapy. These insights reveal the role of microbiome-induced inflammation in LUAD and present new potential targets for interception and therapy.

T cells expressing PD-1 in the peripheral blood (PB) of patients with tumors possess therapeutic potential; however, the immunosuppressive, PD1-triggered signaling pathway and limited proliferative capacity of PD-1+ T cells present challenges to their therapeutic application. Here, we observed no discernible distinction between PD-1+ and PD-1- T cells in terms of clonal overlap. However, CD8+PD-1+ T cells from PB and tumor tissues exhibited tighter clustering based on clone size. Single-cell RNA sequencing analysis showed that PD-1+ T cells from PB highly expressed cytotoxicity-related genes and were enriched for T cell activation-related pathways compared with PD-1- T cells from PB or tumor tissues. Consistent with this, PB-derived PD-1+ T cells exhibited strong cytotoxicity towards autologous tumor cells and tumor cell lines. To augment PD-1+ T-cell activity against solid tumors in vivo, we introduced a PD-1/CD28 fusion receptor combined with a CD19 chimeric antigen receptor (CAR) into PD-1+ T cells, which were then expanded in vitro. The modified PD-1+ T cells exhibited superior proliferation and antitumor abilities in vitro. In addition, four patients with cancer were infused with autologous PD-1/CD28-CD19-CAR PD-1+ T cells. None of these patients experienced severe side effects and one patient with melanoma achieved a complete response that was maintained for 6.7 months. The three other patients had stable disease. Collectively, these results suggested that cell therapy with modified PB-derived PD-1+ T cells is both safe and effective, and it may constitute a promising treatment strategy for cancer patients.

Vγ9Vδ2 T cells are potent but elusive cytotoxic effectors. Butyrophilin subfamily 2 member A1 (BTN2A1) is a surface protein that has recently been shown to bind the Vγ9 chain of the γδ T-cell receptor (TCR) but its precise role in modulating Vγ9Vδ2 T-cell functions remains unknown. Here, we show that 107G3B5, a monoclonal BTN2A1 agonist antibody, was able to significantly enhance Vγ9Vδ2 T-cell functions against hematological or solid cell lines and against primary cells from adult acute lymphoblastic leukemia patients. New computer vision strategies applied to holotomographic microscopy videos showed that 107G3B5 enhanced the interaction between Vγ9Vδ2 T cells and target cells in a quantitative and qualitative manner. In addition, we found that Vγ9Vδ2 T cells activated by 107G3B5 induced caspase 3/7 activation in tumor cells, thereby triggering tumor cell death by pyroptosis. Together, these data demonstrate that targeting BTN2A1 with 107G3B5 enhances the Vγ9Vδ2 T-cell antitumor response by triggering the pyroptosis-induced immunogenic cell death. These new pyroptosis-based therapies have great potential to stimulate the immune system to fight cancer, especially "cold" tumors.

Despite recent advances in immunotherapy with immune checkpoint inhibitors (ICI), many patients with non-small cell lung cancer (NSCLC) fail to respond or develop resistance after an initial response. In situ vaccination (ISV) with engineered viruses has emerged as a promising antigen-agnostic strategy that can both condition the tumor microenvironment (TME) and augment anti-tumor T cell responses to overcome immune resistance. We engineered a live attenuated viral vaccine, Hyper-Interferon Sensitive virus (HIS), by conducting a genome-wide functional screening and introducing eight interferon (IFN)-sensitive mutations in the influenza genome. Compared to wild-type (WT) influenza, HIS replication was attenuated in immunocompetent hosts, enhancing its potential as a safe option for cancer therapy. HIS ISV elicited robust yet transient type I IFN responses in murine NSCLCs, leading to an enrichment of polyfunctional effector Th1 CD4 and cytotoxic CD8 T cells into the tumor. HIS ISV demonstrated enhanced anti-tumor efficacy compared to WT in multiple syngeneic murine models of NSCLC with distinct driver mutations and varying mutational burden. This efficacy was dependent on host type 1 IFN responses and T lymphocytes. HIS ISV overcame resistance to anti-PD-1 in LKB-1 deficient murine NSCLC, resulting in improved overall survival and enduring systemic tumor-specific immunity. These studies provide compelling evidence to support further clinical evaluation of HIS as a novel 'off-the-shelf' ISV strategy for patients with NSCLC refractory to ICI.

Radiotherapy (RT) triggers an immune response that contributes to anti-tumor effects. Induction of interferon beta (IFN-β) is a key event in this immunogenicity of RT. We have previously shown that TRIM33, a chromatin reader, restrains IFN-β expression in Toll-like receptor-activated myeloid cells. Here, we explored whether deleting Trim33 in myeloid cells might improve the radio-induced immune response, and subsequent efficiency of RT. We first established that Trim33-/- bone marrow-derived macrophages showed increased expression of IFN-β in response to direct irradiation, or to treatment with irradiated cancer cells, further supporting our hypothesis. We then tested the efficiency of a single dose RT in three subcutaneous and one orthotopic tumor models. In all situations, myeloid deletion of Trim33 led to a significantly improved response after RT, leading to a complete and durable response in most of the treated mice bearing orthotopic oral tumors. This effect required the IFN-I pathway, and the presence of CD8+ T lymphocytes, but not NK cells. In addition, cured mice were capable of rejecting a secondary tumor challenge, demonstrating an in situ vaccination effect. We conclude that deleting Trim33 in myeloid cells improves RT efficiency, through a mechanism involving the IFN-I pathway and the immune response. Our work suggests that myeloid Trim33 is a host factor affecting the tumor response to RT, thus representing a new potential therapeutic target for modifying RT responses.

Ovarian cancers and microsatellite stable (MSS) colorectal cancers (CRC) are insensitive to anti-PD1 immunotherapy, and new immunotherapeutic approaches are needed. Preclinical data suggests a relationship between immunotherapy resistance and elevated angiopoietin 2 levels. We performed a phase 1 dose-escalation study of pembrolizumab and the angiopoietin 1/2 inhibitor trebananib (NCT03239145). This multicenter trial enrolled patients with metastatic ovarian cancer or MSS CRC. Trebananib was administered intravenously weekly for 12 weeks with 200 mg intravenous pembrolizumab every 3 weeks. The toxicity profile of this combination was manageable, and the protocol-defined highest dose level (trebananib 30 mg/kg weekly plus pembrolizumab 200 mg every 3 weeks) was declared the maximum tolerated dose. The objective response rate for all patients was 7.3% (90% confidence interval: 2.5-15.9%). Three patients with MSS CRC had durable responses for ≥3 years. One responding patient's CRC harbored a POLE mutation. The other two responding patients had left-sided CRCs with no baseline liver metastases, and genomic analysis revealed that they both had KRAS wild-type, ERBB2 amplified tumors. After development of acquired resistance, biopsy of one patient's KRAS wild-type, ERBB2 amplified tumor showed a substantial decline in tumor-associated T cells and an increase in immunosuppressive intratumoral macrophages. Future studies are needed to carefully assess whether clinicogenomic features, such as lack of liver metastases, ERBB2 amplification, and left-sided tumors, can predict increased sensitivity to PD1 immunotherapy combinations.