Abstract
PURPOSE: We present the results of a post hoc tumor tissue analysis from the phase 3 MILO/ENGOT-ov11 study (NCT01849874). PATIENTS AND METHODS: Mutation/copy-number analysis was performed on tissue obtained pre-randomization. The Kaplan-Meier method was used to estimate progression-free survival (PFS). Unbiased univariate analysis, Cox regression, and binary logistic regression were used to test associations between mutation status and outcomes, including PFS and binary response by local RECIST 1.1. RESULTS: MILO/ENGOT-ov11 enrolled 341 patients, ranging in age from 22 to 79, from June, 2013 to April, 2016. Patients were randomized 2:1 to binimetinib or physician's choice of chemotherapy (PCC). The most commonly altered gene was KRAS (33%). In 135 patients treated with binimetinib with response rate (RR) data, other detected MAPK pathway alterations included: NRAS (n = 11, 8.1%), BRAF V600E (n = 8, 5.9%), RAF1 (n = 2, 1.5%), and NF1 (n = 7, 5.2%). In those with and without MAPK pathway alterations, the RRs with binimetinib were 41% and 13%, respectively. PFS was significantly longer in patients with, compared with those without, MAPK pathway alterations treated with binimetinib [HR, 0.5; 95% confidence interval (CI) 0.31-0.79]. There was a nonsignificant trend toward PFS improvement in PCC-treated patients with MAPK pathway alterations compared with those without (HR, 0.82; 95% CI, 0.43-1.59). CONCLUSIONS: Although this hypothesis-generating analysis is limited by multiple testing, higher RRs and longer PFS were seen in patients with low-grade serous ovarian cancer (LGSOC) treated with binimetinib, and to a lesser extent in those treated with PCC, who harbored MAPK pathway alterations. Somatic tumor testing should be routinely considered in patients with LGSOC and used as a future stratification factor.