Axitinib in Combination With Toripalimab, a Humanized Immunoglobulin G4 Monoclonal Antibody Against Programmed Cell Death-1, in Patients With Metastatic Mucosal Melanoma: An Open-Label Phase IB Trial

Abstract

Purpose: Metastatic mucosal melanoma responds poorly to anti-programmed cell death-1 (PD-1) monotherapy. Vascular endothelial growth factor (VEGF) has been shown to play an important immunosuppressive role in the tumor microenvironment. The combination of VEGF inhibition and PD-1 blockade provides therapeutic opportunities for patients refractory to either therapy alone.

Patients and methods: We conducted a single-center, phase IB trial evaluating the safety and preliminary efficacy of toripalimab, a humanized immunoglobulin G₄ monoclonal antibody against PD-1 in combination with the VEGF receptor inhibitor axitinib in patients with advanced melanoma, including patients with chemotherapy-naïve mucosal melanomas (88%). Patients received toripalimab at 1 or 3 mg/kg via intravenous infusion every 2 weeks, in combination with axitinib 5 mg orally twice a day, in a dose-escalation and cohort-expansion study until confirmed disease progression, unacceptable toxicity, or voluntary withdrawal. The primary objective was safety. Secondary objectives included efficacy, pharmacokinetics, pharmacodynamics, immunogenicity, and tumor tissue biomarkers.

Results: Thirty-three patients were enrolled. No dose-limiting toxicities were observed. Ninety-seven percent of patients experienced treatment-related adverse events (TRAEs). The most common TRAEs were mild (grade 1 or 2) and included diarrhea, proteinuria, hand and foot syndrome, fatigue, AST or ALT elevation, hypertension, hypo- or hyperthyroidism, and rash. Grade 3 or greater TRAEs occurred in 39.4% of patients. By the cutoff date, among 29 patients with chemotherapy-naïve mucosal melanoma, 14 patients (48.3%; 95% CI, 29.4% to 67.5%) achieved objective response, and the median progression-free survival time was 7.5 months (95% CI, 3.7 months to not reached) per Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1.

Conclusion: The combination of toripalimab plus axitinib was tolerable and showed promising antitumor activity in patients with treatment-naïve metastatic mucosal melanoma. Patients enrolled in this study were all Asian, and this combination therapy must be validated in a randomized phase III trial that includes a non-Asian population before it can become a standard of care.

FIG 1.

(A) Maximal change of tumor size from baseline assessed by investigator per Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1 (N = 33). The length of the bar represents maximal decrease or minimal increase in target lesion(s). (B) Change in individual tumor burden over time from baseline assessed by investigator per RECIST version 1.1 (N = 33). (C) Exposure and duration of response per RECIST version 1.1 (N = 33). (*) Patient with target lesion(s) reduction of more than 30% but with new lesion(s) or progression of nontarget lesion(s). (†) Unconfirmed partial response classified as stable disease. Mbp, million base pairs; PD-L1, programmed death ligand-1; TMB, tumor mutational burden.

FIG 2.

Progression-free survival by (A) Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1 and (B) Immune-Related RECIST (irRECIST) and (C) overall survival of 29 patients with chemotherapy-naïve mucosal melanoma. Probability of survival is shown at indicated time points. Censored patients are marked with a vertical line in the graph. Numbers of patients at risk at indicated time points are shown below the x-axis. NR, not reached.

FIG 3.

Clinical response in relation to tumor programmed death ligand-1 (PD-L1) expression and tumor mutational burden (TMB) in patients with chemotherapy-naïve mucosal melanoma. (A) PD-L1–positive status was defined as the presence of membrane staining of any intensity in 1% or more of tumor cells or immune cells by SP263 immunohistochemistry (IHC) staining. TMB was calculated by summing up somatic mutations within the coding regions by whole-exon sequencing. A TMB of 6 mutations per million base pairs (Mbp) was the cutoff value. (B) Progression-free survival per Immune-Related Response Evaluation Criteria in Solid Tumors (irRECIST) of PD-L1–positive and PD-L1–negative patients. (C) Progression-free survival per irRECIST of patients with TMB of 6 mutations/Mbp or greater and patients with TMB of less than 6 mutations/Mbp. PD-L1–positive status was defined as the presence of membrane staining of any intensity in 1% or more of tumor cells or immune cells by SP263 IHC staining. Probability of survival is shown at indicated time points. Censored patients are marked with a vertical line in the graph. Numbers of patients at risk at indicated time points are shown below the x-axis. HR, hazard ratio; NR, not reached.

FIG 4.

Inflammatory signature and correlation with clinical response. RNA sequencing and expression profiling of tumor biopsies were acquired from 24 patients with chemotherapy-naïve mucosal melanoma. The expression signatures of six inflammation-related genes (IL-6, CXCL1, CXCL2, CXCL3, CXCL8, and PTGS2), six angiogenesis-related genes (VEGFA, KDR, ESM1, PECAM1, ANGPTL4, and CD34), or 12 selected immune- or angiogenesis-related genes (CD274/PD-L1, CXCR6, CD27, CXCL9, IDO1, TIGIT, PDCD1LG2/PD-L2, LAG3, ANGPTL5, ANGPTL6, CD34, and KDR) were evaluated. To obtain a single score for the signature for each sample, the mean expression of the genes composing the signature was calculated. (A) The scores between patients with objective response (partial response [PR]; n = 13) per Immune-Related Response Evaluation Criteria in Solid Tumors (irRECIST) and patients with stable disease (SD) or progressive disease (PD; n = 11) were compared (P < .001). (B) The scores between patients achieving clinical benefit (PR or SD; n = 21) and patients with PD (n = 3) were compared (P < .001). (C) Associations between tumor mutational burden (TMB) and inflammation and angiogenesis gene expression profiles were also evaluated by Spearman correlation. Median TMB and median gene expression profiling scores are shown as dashed lines in the figure. r and P values of the correlations are provided. Mbp, million base pairs.

FIG A1.

Design of phase IB combination study of toripalimab plus axitinib in treating patients with mucosal melanoma and clinical efficacy summary. DLT, dose-limiting toxicity; IV, intravenous; PD, progressive disease; PR, partial response; SD, stable disease.

FIG A2.

CONSORT diagram for phase IB combination study of toripalimab with axitinib in treating patients with mucosal melanoma.

FIG A3.

Serum concentration of toripalimab over 56 days in the combination study.

FIG A4.

Kaplan-Meier plots of median overall survival of (A) Patients who are PD-L1 positive versus PD-L1 negative. (B) TMB ≥ 6 mutations/Mbp versus TMB < 6 mutations/Mbp patients. HR, hazard ratio; Mbp, million base pairs; NR, not reached; PD-L1, programmed death ligand-1; TMB, tumor mutational burden.

FIG A5.

The correlation of clinical response with inflammation or angiogenesis signature panel scores. McDermott et al. inflammation panel: IL-6, CXCL1, CXCL2, CXCL3, CXCL8, and PTGS2; McDermott et al. angiogenesis panel: VEGFA, KDR, ESM1, PECAM1, ANGPTL4, and CD34; Cristescu et al. inflammation panel: IDO1, CXCL10, CXCL9, HLADRA, STAT1, and IFN-g. Responders are those with complete or partial response; nonresponders are those with stable or progressive disease.

FIG A6.

Genetic alternations and frequencies identified by whole exome sequencing (WES) from 31 available patients. Patients were grouped by clinical responses. PD, progressive disease; PR, partial response; SD, stable disease.