Immunotherapy prolongs survival in only a subset of melanoma patients, highlighting the need to better understand the driver tumor microenvironment. We conducted bioinformatic analyses of 703 transcriptomes to probe the immune landscape of primary cutaneous melanomas in a population-ascertained cohort. We identified and validated 6 immunologically distinct subgroups, with the largest having the lowest immune scores and the poorest survival. This poor-prognosis subgroup exhibited expression profiles consistent with β-catenin–mediated failure to recruit CD141+ DCs. A second subgroup displayed an equally bad prognosis when histopathological factors were adjusted for, while 4 others maintained comparable survival profiles. The 6 subgroups were replicated in The Cancer Genome Atlas (TCGA) melanomas, where β-catenin signaling was also associated with low immune scores predominantly related to hypomethylation. The survival benefit of high immune scores was strongest in patients with double-WT tumors for BRAF and NRAS, less strong in BRAF-V600 mutants, and absent in NRAS (codons 12, 13, 61) mutants. In summary, we report evidence for a β-catenin–mediated immune evasion in 42% of melanoma primaries overall and in 73% of those with the worst outcome. We further report evidence for an interaction between oncogenic mutations and host response to melanoma, suggesting that patient stratification will improve immunotherapeutic outcomes.

Purpose Gene expression studies in melanoma have been few because tumors are small and cryopreservation is rarely possible. The purpose of this study was to evaluate the Illumina DASL Array Human Cancer Panel for gene expression studies in formalin-fixed melanoma primary tumors, and to identify prognostic biomarkers. Experimental Design Primary tumors from two studies were sampled using a tissue microarray needle. Study 1: 254 tumors from a melanoma cohort recruited 2000-2006. Study 2: 218 tumors from a case-control study of patients undergoing sentinel node biopsy. Results RNA was obtained from 76% of blocks. 1.4% of samples failed analysis (transcripts from less than 250 of the 502 genes on the DASL chip detected). Increasing age of the block and increased melanin in the tumor were associated with reduced number of genes detected. The gene whose expression was most differentially expressed in association with relapse free survival in study 1 was osteopontin (SPP1, p=2.11×10−6) and supportive evidence for this was obtained in study 2 used as a validation set (p=0.006) (unadjusted data). Osteopontin level in study 1 remained a significant predictor of relapse free survival when data were adjusted for age, sex, tumor site and histological predictors of relapse. Genes whose expression correlated most strongly with osteopontin were PBX1, BIRC5 (Survivin) and HLF. Conclusion Expression data were obtained from 74% of primary melanomas and provided confirmatory evidence that osteopontin expression is a prognostic biomarker. These results suggest that predictive biomarker studies may be possible using stored blocks from mature clinical trials.

Development and validation of robust molecular biomarkers has so far been limited in melanoma research. In this paper we used a large population-based cohort to replicate two published gene signatures for melanoma classification. We assessed the signatures prognostic value and explored their biological significance by correlating them with factors known to be associated with survival (vitamin D) or etiological routes (nevi, sun sensitivity and telomere length). Genomewide microarray gene expressions were profiled in 300 archived tumors (224 primaries, 76 secondaries). The two gene signatures classified up to 96% of our samples and showed strong correlation with melanoma specific survival (P=3×10−4), Breslow thickness (P=5×10−10), ulceration (P=9.x10−8) and mitotic rate (P=3×10−7), adding prognostic value over AJCC stage (adjusted hazard ratio 1.79, 95%CI 1.13-2.83), as previously reported. Furthermore, molecular subtypes were associated with season-adjusted serum vitamin D at diagnosis (P=0.04) and genetically predicted telomere length (P=0.03). Specifically, molecular high-grade tumors were more frequent in patients with lower vitamin D levels whereas high immune tumors came from patients with predicted shorter telomeres. Our data confirm the utility of molecular biomarkers in melanoma prognostic estimation using tiny archived specimens and shed light on biological mechanisms likely to impact on cancer initiation and progression.