Too much skin in the game? A paradigm shift in our understanding of vulvar and vaginal melanomas as distinct tumor types compared with cutaneous melanomas

Abstract

<h3>Objectives:</h3> Due to the rarity of vulvar/vaginal melanoma (VVM), current therapeutic strategies mimics that of cutaneous melanoma (CM) without scientific rationale. Currently, immuno-oncology (IO) is a front-line treatment option for advanced melanoma, however data are limited for IO outcomes in advanced VVM. Considering 5-year survival for vulvar (58%) and vaginal (27%) melanoma is significantly inferior to cutaneous melanoma (81%), this calls into question the assumed similarity of these malignancies. As such, our goal is to compare molecular profiles of VVM with CM and explore the significance of IO agents on survival. <h3>Methods:</h3> Samples were analyzed using next-generation sequencing (NextSeq, 592 Genes and WES, NovaSEQ), IHC and WTS (NovaSeq) (Caris Life Sciences, Phoenix, AZ). PD-L1 expression was tested by IHC using 28-8 (Agilent) and SP-142 (Spring Biosciences) (positive cut-off ≥1%). MSI was tested by FA, IHC and NGS. TMB was measured by totaling somatic mutations per tumor (TMB-high cut-off ≥10 mutations per MB). Immune cell fraction was calculated by QuantiSeq (Finotello 2019, Genome Medicine). Survival was extracted from insurance claims data and calculated from time of IO treatment to last contact using Kaplan-Meier survival curves. Statistical significance was determined using chi-square and Wilcoxon rank sum test and adjusted for multiple comparisons. <h3>Results:</h3> Molecular analysis was performed on 171 VVM and 5255 CM between 1998 and 2020. Median age for VVM and CM was 65 and 63, respectively. A total of 114 (66.7%) VVM and 3538 (67.3%) CM were metastatic at time of diagnosis. Immunogenicity of VVM was significantly lower than CM, demonstrated by an absence of high tumor mutation burden (0% vs 48%) and a decrease in PD-L1 expression (34.1% vs 45.2%) (Fig 1A). Adaptive immune gene expression was lower in VVM compared to CM (Fig 1B). By QuantiSeq, the cell fractions for type I macrophages and C8+ T-cells were significantly lower in VVM compared to CM (Fig 1C). Median survival was shorter for VVM than for CM (19 vs 37 months, p=.058; Fig 1D). VVM also demonstrated significantly (p<.01) less frequent BRAF mutations (8.4% vs 35.8%), more frequent KIT mutations (13.2% vs 2.8%), KIT amplifications (14.7% vs 1.5%), ATRX mutations (28.4% vs 3.8%), and SF3B1 mutations (27.8% vs 1.6%). NRAS mutations were similar (14.6% vs 19.5%). Alteration in pathways involving DNA damage (16.4% vs 5.2%) and mRNA splicing (28.9% vs 2.8%) were more common in VVM, while alterations in cell cycle (7.2% vs 18.8%) and chromatin remodeling (6.3% vs 21.4%) were less common (Fig 1E). <h3>Conclusions:</h3> VVM represents a distinct molecular profile from CM with a less favorable immune phenotype demonstrated by absence of TMB-high, lower rates of PD-L1 positivity, and lower adaptive immune gene expression and cell fractions of effector T-cells and immune promoting macrophages. Compared with CM, patients with VVM were found to have significantly worse survival when treated with IO therapy. Though IO has been a mainstay of treatment in recent years, these findings suggest that new therapeutic strategies are needed.