Purpose: Combination PD-1 and CTLA-4 inhibitor therapy has dramatically improved the survival of patients with advanced melanoma but is also associated with significant immune-related toxicities. This study sought to identify circulating cytokine biomarkers of treatment response and immune-related toxicity. Experimental Design: The expression of 65 cytokines was profiled longitudinally in 98 patients with melanoma treated with PD-1 inhibitors, alone or in combination with anti-CTLA-4, and in an independent validation cohort of 49 patients treated with combination anti-PD-1 and anti-CTLA-4. Cytokine expression was correlated with RECIST response and immune-related toxicity, defined as toxicity that warranted permanent discontinuation of treatment and administration of high-dose steroids. Results: Eleven cytokines were significantly upregulated in patients with severe immune-related toxicities at baseline (PRE) and early during treatment (EDT). The expression of these 11 cytokines was integrated into a single toxicity score, the CYTOX (cytokine toxicity) score, and the predictive utility of this score was confirmed in the discovery and validation cohorts. The AUC for the CYTOX score in the validation cohort was 0.68 at PRE [95% confidence interval (CI), 0.51–0.84; P = 0.037] and 0.70 at EDT (95% CI, 0.55–0.85; P = 0.017) using ROC analysis. Conclusions: The CYTOX score is predictive of severe immune-related toxicity in patients with melanoma treated with combination anti-CTLA-4 and anti-PD-1 immunotherapy. This score, which includes proinflammatory cytokines such as IL1a, IL2, and IFNα2, may help in the early management of severe, potentially life-threatening immune-related toxicity. See related commentary by Johnson and Balko, p. 1452
Transcriptomic signatures designed to predict melanoma patient responses to PD-1 blockade have been reported but rarely validated. We now show that intra-patient heterogeneity of tumor responses to PD-1 inhibition limit the predictive performance of these signatures. We reasoned that resistance mechanisms will reflect the tumor microenvironment, and thus we examined PD-1 inhibitor resistance relative to T-cell activity in 94 melanoma tumors collected at baseline and at time of PD-1 inhibitor progression. Tumors were analyzed using RNA sequencing and flow cytometry, and validated functionally. These analyses confirm that major histocompatibility complex (MHC) class I downregulation is a hallmark of resistance to PD-1 inhibitors and is associated with the MITFlow/AXLhigh de-differentiated phenotype and cancer-associated fibroblast signatures. We demonstrate that TGFß drives the treatment resistant phenotype (MITFlow/AXLhigh) and contributes to MHC class I downregulation in melanoma. Combinations of anti-PD-1 with drugs that target the TGFß signaling pathway and/or which reverse melanoma de-differentiation may be effective future therapeutic strategies.
Liver metastasis from colorectal cancer is a leading cause of cancer mortality. Myeloid cells play pivotal roles in the metastatic process, but their prometastatic functions in liver metastasis remain incompletely understood. To investigate their role, we simulated liver metastasis in C57BL/6 mice through intrasplenic inoculation of MC38 colon carcinoma cells. Among the heterogeneous myeloid infiltrate, we identified a distinct population of CD11b/Gr1mid cells different from other myeloid populations previously associated with liver metastasis. These cells increased in number dramatically during establishment of liver metastases and were recruited from bone marrow by tumor-derived CCL2. Liver metastasis of Lewis lung carcinoma cells followed this pattern but this mechanism is not universal as liver colonization by B16F1 melanoma cells did not recruit similar subsets. Inhibition of CCL2 signaling and absence of its cognate receptor CCR2 reduced CD11b/Gr1mid recruitment and decreased tumor burden. Depletion of the CD11b/Gr1mid subset in a transgenic CD11b-diphtheria toxin receptor mouse model markedly reduced tumor cell proliferation. There was no evidence for involvement of an adaptive immune response in the prometastatic effects of CD11b/Gr1mid cells. Additionally, an analogous myeloid subset was found in liver metastases of some colorectal cancer patients. Conclusion: Collectively, our findings highlight the importance of myeloid cells—in this case a selective CD11b/Gr1mid subset—in sustaining development of colorectal cancer liver metastasis and identify a potential target for antimetastatic therapy.
Pancreatic ductal adenocarcinoma (PDAC) is considered a non‐immunogenic tumor, and immune checkpoint inhibitor monotherapy lacks efficacy in this disease. Radiotherapy (RT) can stimulate the immune system. Here, we show that treatment of KPC and Pan02 murine PDAC cells with RT and gemcitabine upregulated PD‐L1 expression in a JAK/Stat1‐dependent manner. In vitro , PD‐L1 inhibition did not alter radio‐ and chemosensitivity. In vivo , addition of anti‐PD‐L1 to high (12, 5 × 3, 20 Gy) but not low (6, 5 × 2 Gy) RT doses significantly improved tumor response in KPC and Pan02 allografts. Radiosensitization after PD‐L1 blockade was associated with reduced CD11b + Gr1 + myeloid cell infiltration and enhanced CD45 + CD8 + T‐cell infiltration with concomitant upregulation of T‐cell activation markers including CD69, CD44, and FasL, and increased CD8:Treg ratio. Depletion of CD8 + T cells abrogated radiosensitization by anti‐PD‐L1. Blockade of PD‐L1 further augmented the effect of high RT doses (12 Gy) in preventing development of liver metastases. Exploring multiple mathematical models reveals a mechanism able to explain the observed synergy between RT and anti‐PD‐L1 therapy. Our findings provide a rationale for testing the use of immune checkpoint inhibitors with RT in PDAC.