Introduction: Understanding the molecular underpinnings of vulvar squamous cell carcinoma (vSCC) has strong potential for clinical impact, given the unmet need for this rare cancer. vSCCs have two main etiologies, including human papillomavirus (HPV) infection or skin conditions such as lichen sclerosis. Clinically, few treatment options with limited efficacy are available for recurrent vSCCs and large whole transcriptome disease subtyping studies are limited due to the rarity of this cancer type. Here, we present a whole transcriptome subtyping study of 244 vSCC clinical samples with DNA genomic profiling. Furthermore, we compared vSCCs to other SCC disease subtypes (pan-SCC), including lung, esophageal, skin, urothelial, anal, and head and neck (HN). Our goal was to gain novel insights into the molecular landscape of vSCCs to inform new routes of therapy.

Methods: Deidentified records of SCC patients were selected from the Tempus clinico-genomic database. The vSCC cohort was predominantly composed of, but not exclusively, naive or pre-treatment primary tissue. The pan-SCC cohort was limited to randomly sampled primary pre-treatment female samples with paired RNA and DNA sequencing. To perform vSCC subtyping, we used bulk RNA-seq and clustered the samples using FastPG-CC, an internally developed consensus-clustering extension of FastPG, a graph-based parallelized clustering tool. We next performed differential expression, pathway analysis, and cell type abundance estimation.

Results: Unbiased gene expression clustering analysis identified three vSCCs clusters; an HPV+ cluster (HPV+, 32% of vSCC samples), HPV- cluster (HPV-, 33%), and an intermediate cluster which contained both HPV+ and HPV- samples (Mix, 35%). The HPV- cluster was enriched for TP53 and TERT promoter mutations, EMT and tumor microenvironment gene expression-derived pathways, and was most similar to HPV- oral HN tumors. The HPV+ cluster was enriched with PIK3CA mutations, cell cycle gene expression and was most similar to other HPV+ SCC samples, including cervical, anal, and HN. The Mix cluster was enriched for neutrophils, fatty acid metabolism and was most similar to skin SCC.

Conclusions: We performed a multi-modal molecular subtyping analysis on a large cohort of vSCCs and identified three molecular subtypes, which enabled us to identify heterogeneity within vSCC, and observe that a subset of vSCCs share features with SCC tumor types for which there are emerging therapy options. Thus, this approach identifies a subset vSCCs that may be responsive to existing SCC therapeutics.

VIEW THE PUBLICATION

VIEW THE POSTER