Primary melanomas can be both deadly and unpredictable;a subset of thin melanomas can behave very aggressively while a subset of more advanced melanomas can show indolent behavior. We do not understand, and certainly cannot predict, this biological behavior. Molecular analysis, including transcriptional profiling (TP), has revealed unexpected heterogeneity in other cancers, informing management and outcome. Recently, a degree of molecular heterogeneity has been discovered in cutaneous melanoma, including tumor BRAF* and NRAS* status and germline MCIR alleles;however, transcriptional profiling of most primary melanomas has been impossible for technical reasons. Since 2001, Harvard Skin SPORE investigators have performed TP on >200 melanomas, including 31 rare frozen primary melanomas, several cell lines, and many fresh or frozen metastatic lesions and short term cultures. Strikingly, bio-informatic analysis has revealed two distinct classes of melanoma defined by TP signatures: one characterized by higher expression of MITF and related genes, and the other by higher expression of immune, inflammatory, and growth factor genes (IIG). All melanomas have segregated with high confidence into one of these two classes define by TP signature. In parallel, SPORE investigators have helped translate a novel platform forTP, termed DASL, that permits analysis of formalin-fixed paraffin-embedded tissue (FFPE). Using the DASL platform, recent experiments have validated these two signatures in FFPE primary melanomas;accordingly, large numbers of primary melanomas are now available for TP. For Project 5, Harvard Skin SPORE investigators have identified and recruited several unique cohorts with attendant FFPE primary melanomas (collectively, n>1300), including Nurses Health Study 1&2 and Health Professionals Follow up study (with environmental risk exposure data), the Harvard Familial Melanoma Registry (with substantial genetic data), and two outcome studies assessing the utility of melastatin (MLSN) as a prognostic biomarker. In Project 5, investigators will correlate the MITF/IIG TP signatures of primary melanomas with 1) genetic variables, including somatic (NRASVBRAF*) and germline (MCIR), 2) risk and environmental exposure variables (NHS1.2 and HPFS), and 3) prognosis and outcome measures (MLSN 1&2). The Broad Institute, Core B, and Core C will facilitate the activities of Project 5 investigators in the generation, analysis and interpretation of the data. Over the next five years, it is expected that Project 5 will deliver a higher level of molecular genetic analysis of clinically annotated melanomas than has ever been available before, with the goal of informing prevention, diagnosis, management, and outcome strategies in this deadly human cancer.
Showing the most recent 10 out of 132 publications
