The melanocytic transcriptomic state is independently associated with poor overall survival in patients with metastatic melanoma.

Abstract

Melanomas display distinct transcriptomic states, but it remains unclear how they associate with clinical outcomes. We performed digital spatial RNA profiling (DSP-RNA) of metastatic tumors from patients to investigate how transcriptomic states correlate with melanoma specific survival (MSS) and acral melanoma (AM). We performed DSP-RNA across a tissue microarray constructed from 111 patients with in-transit metastatic melanoma (ITM) diagnosed from 1990 to 2020. Data quality control, noise correction, and normalization yielded high-quality profiles from 105 patients, including 30 (36%) who received immune checkpoint inhibitors and 20 (24%) with AM. We performed principal component (PC) analysis and correlated the results with published gene signatures: The PC1 axis differentiated transitory from undifferentiated melanoma, PC2 reflected immune cell infiltration, PC3 corresponded to stromal cells and neural crest-like melanoma, and PC4 associated with melanocytic melanoma. Across a cohort of treatment-naïve ITM, high expression of the melanocytic state conferred a median MSS difference of 7.72 years (melanocytic "high" = 5.16 years versus "low" = 12.88 years, log-rank <i>P</i> = 0.0061) and independently associated with poor survival in multivariate analysis. AMs showed higher melanocytic state gene expression compared to nonacral. Findings were validated in external datasets, supporting that the melanocytic state predicts poor prognosis. The melanocytic state is associated with poor prognosis and may be enriched in AM, implying that identifying patients with melanocytic melanoma may be important for therapeutic decisions. Unlike other gene expression predictors proposed for prognostic stratification, the melanocytic state characterizes a biological subtype, suggesting that it may have specific therapeutic vulnerabilities.