Activating kinase mutations in melanoma

The treatment of patients with metastatic melanoma has been revolutionized by the discovery of crucial activating kinase mutations, making melanoma susceptible to targeted kinase inhibition. The result is that melanoma, once a notoriously treatment resistant malignancy, has become the poster child for the promise of personalized cancer therapy. We have sought to identify what the incidence of various kinase mutations are in a group of 192 melanoma patients, sampled from a population that harbors one of the highest rates of melanoma in the world, and what this may mean for their prognosis and treatment. In our cohort of patients with advanced melanoma, mutations in BRAF, NRAS, KIT and KRAS occurred with frequencies of 24.5%, 20.8%, 4.7% and 2.6%, respectively. We show that BRAF-mutant melanomas have a specific clinical profile. Furthermore, we reveal that BRAF and NRAS mutant melanomas show significant differences in their respective rates of metastatic spread and overall survival. Our mutation screen also revealed two relatively common juxtamembranous MET single nucleotide polymorphisms to be significantly over-represented in our melanoma patient cohort. As the identification of some of these clinically relevant mutations will become routine in the workup of patients with advanced melanoma, we turned our attention to the search of a phenotype/genotype association that may be employed by anatomical pathologists to triage cases for molecular analysis. Using m-RNA extracted from formalin-fixed paraffin embedded melanoma tissue, we identified a number of genes differentially expressed in BRAF, NRAS and KIT-mutant melanomas. The levels of these differentially expressed genes were then assessed using immunohistochemical staining and quantitative scoring on a matching melanoma tissue microarray. We show a strong correlation in the expression data and immunohistochemical staining for some of the genes identified. Some of these differences in protein staining were also shown to be capable of having a good predictive capacity for identifying cases likely to harbor KIT mutations. Quantitative cytomorphological analysis identified a BRAF-mutant melanoma specific morphological phenotype, characterized by larger cellular volumes, more rounded shapes and lower degrees of cellular pleomorphism. In conclusion, we show that incorporating discriminatory clinical; cytomorphological and immunohistochemical data can generate a decision tree algorithm with fair to good predictive power for the detection of KIT and BRAF-mutant melanomas.