Predominance of triple wild-type and IGF2R mutations in mucosal melanomas

Primary mucosal melanoma (MM) is a rare subtype of melanoma that arises from melanocytes in the mucosa. MM has not been well profiled for mutations and its etiology is not well understood, rendering current treatment strategies unsuccessful. Hence, we investigated mutational landscape for MM to understand its etiology and to clarify mutations that are potentially relevant for MM treatment.

These findings demonstrated a potential etiologic factor and driver mutation for MM and strongly suggested that MM initiation or progression involves distinct molecular-mechanisms from CM. This study also identified mutational signatures that are clinically relevant for MM treatment.

Forty one MM and 48 cutaneous melanoma (CM) tissues were profiled for mutations using targeted deep next-generation sequencing (NGS) for 89 cancer-related genes. A total of 997 mutations within exons were analyzed for their mutational spectrum and prevalence of mutation, and 685 non-synonymous variants were investigated to identify mutations in individual genes and pathways. PD-L1 expression from 21 MM and 18 CM were assessed by immunohistochemistry.

Mutational spectrum analysis revealed a lower frequency of UV-induced DNA damage in MM than in CM (p = 0.001), while tobacco exposure was indicated as a potential etiologic factor for MM. In accordance with low UV damage signatures, MM demonstrated an overall lower number of mutations compared to CM (6.5 mutations/Mb vs 14.8 mutations/Mb, p = 0.001), and less PD-L1 expression (p = 0.003). Compared to CM, which showed frequent mutations in known driver genes (BRAF 50.0%, NRAS 29.2%), MM displayed lower mutation frequencies (BRAF; 12.2%, p < 0.001, NRAS; 17.1%), and was significantly more enriched for triple wild-type (no mutations in BRAF, RAS, or NF1, 70.7% vs 25.0%, p < 0.001), IGF2R mutation (31.7% vs 6.3%, p = 0.002), and KIT mutation (9.8% vs 0%, p = 0.042). Of clinical relevance, presence of DCC mutations was significantly associated with poorer overall survival in MM (log-rank test, p = 0.02). Furthermore, mutational spectrum analysis distinguished primary anorectal MM from CM metastasized to the bowel (spectrum analysis p < 0.001, number of mutations p = 0.002). Conclusions: These findings demonstrated a potential etiologic factor and driver mutation for MM and strongly suggested that MM initiation or progression involves distinct molecular-mechanisms from CM. This study also identified mutational signatures that are clinically relevant for MM treatment.