Distinguishing between nodular and superficial spreading melanoma using specific microRNA alterations.

8540 Background: Distinct molecular alterations in acral, mucosal and uveal melanoma subtypes have been recently reported and have prompted the development of tailored treatment approaches. Superficial spreading melanoma (SSM) and nodular melanoma (NM), the 2 most common types of cutaneous melanoma (70% and 20% respectively), are currently perceived as sequential phases of linear progression from radial to vertical growth. However, clinicopathological data from our group and others suggest that SSM and NM might be two biologically different entities. In this study, we used microRNA profiling to identify the signature that distinguishes SSM from NM. We further determined the association between decreased expression of SSM-specific microRNAs and loss of the corresponding genomic loci. METHODS miRCURY LNA microRNA Arrays were used to profile RNA isolated from 82 micro and macrodissected primary melanomas (26 SSM and 56 NM) and 9 congenital nevi. Differentially expressed microRNAs were identified using T test (p<0.05) and SAM (FDR<10%) analyses. The genomic loci corresponding to specific microRNAs were examined in 4 "radial growth phase-SSM like" and 6 "vertical growth phase-NM like" melanoma cell lines, as well as in an independent validation cohort of 25 primary melanomas. RESULTS We identified a signature of 134 microRNAs that distinguishes SSM from NM (59 microRNAs showed higher expression in SSM compared to NM and 75 microRNAs showed lower expression in SSM compared to NM). 31 of the microRNAs that were expressed at lower level in SSM compared to NM showed lower expression compared to nevi as well (p<0.05). Downregulation of 7 known tumor-suppressive microRNAs (let-7g, miR-15a, miR-16, miR-138, miR-181a, miR-191 and miR-933) was associated with selective loss of the corresponding genomic loci in SSM but not NM in the melanoma cell lines and in the independent validation cohort (p<0.05). CONCLUSIONS Our data support a new molecular classification model in which SSM and NM are molecularly distinct phenotypes as opposed to sequential phases of melanoma progression. New therapeutic strategies taking into account the subtype-specific alterations that we have identified might improve the outcome of melanoma patients.