Genomic and molecular profiling predicts response to temozolomide in melanoma.

PURPOSE Despite objective response rates of only approximately 13%, temozolomide remains one of the most effective single chemotherapy agents against metastatic melanoma, second only to dacarbazine, the current standard of care for systemic treatment of melanoma. The goal of this study was to identify molecular and/or genetic markers that correlate with, and could be used to predict, response to temozolomide-based treatment regimens and that reflect the intrinsic properties of a patient's tumor. EXPERIMENTAL DESIGN Using a panel of 26 human melanoma-derived cell lines, we determined in vitro temozolomide sensitivity, O(6)-methylguanine-DNA methyltransferase (MGMT) activity, MGMT protein expression and promoter methylation status, and mismatch repair proficiency, as well as the expression profile of 38,000 genes using an oligonucleotide-based microarray platform. RESULTS The results showed a broad spectrum of temozolomide sensitivity across the panel of cell lines, with IC(50) values ranging from 100 micromol/L to 1 mmol/L. There was a significant correlation between measured temozolomide sensitivity and a gene expression signature-derived prediction of temozolomide sensitivity (P < 0.005). Notably, MGMT alone showed a significant correlation with temozolomide sensitivity (MGMT activity, P < 0.0001; MGMT expression, P <or= 0.0001). The promoter methylation status of the MGMT gene, however, was not consistent with MGMT gene expression or temozolomide sensitivity. CONCLUSIONS These results show that melanoma resistance to temozolomide is conferred predominantly by MGMT activity and suggest that MGMT expression could potentially be a useful tool for predicting the response of melanoma patients to temozolomide therapy.