Purpose: Test the hypothesis that in BRAF-mutated melanomas, clinical responses to selumetinib, a MEK inhibitor, will be restricted to tumors in which the PI3K/AKT pathway is not activated.
Experimental design: We conducted a phase II trial in patients with melanoma whose tumors harbored a BRAF mutation. Patients were stratified by phosphorylated-AKT (pAKT) expression (high vs. low) and treated with selumetinib 75 mg per os twice daily. Pretreatment tumors were also analyzed for genetic changes in 230 genes of interest using an exon-capture approach.
Results: The high pAKT cohort was closed after no responses were seen in the first 10 patients. The incidence of low pAKT melanoma tumors was low (∼25% of melanomas tested) and this cohort was eventually closed because of poor accrual. However, among the five patients with melanoma accrued in the low pAKT cohort, there was one partial response (PR). Two other patients had near PRs before undergoing surgical resection of residual disease (one patient) or discontinuation of treatment due to toxicity (one patient). Among the two nonresponding, low pAKT patients with melanoma, co-mutations in MAP2K1, NF1, and/or EGFR were detected.
Conclusions: Tumor regression was seen in three of five patients with BRAF-mutated, low pAKT melanomas; no responses were seen in the high pAKT cohort. These results provide rationale for co-targeting MEK and PI3K/AKT in patients with BRAF mutant melanoma whose tumors express high pAKT. However, the complexity of genetic changes in melanoma indicates that additional genetic information will be needed for optimal selection of patients likely to respond to MEK inhibitors.