Phase 2 (BRF113220 [NCT01072175]) and phase 3 (COMBI-d [NCT01584648], COMBI-v [NCT01597908]) trials provide a large dataset for evaluating D + T in patients (pts) with BRAF V600E/K–mutant melanoma. Pooled analyses showed that D + T efficacy across studies (N = 617) is consistent with that in individual trials (response, 67%; median progression-free survival [PFS], 11.1 mo [95% CI, 9.7-12.9]; median overall survival [OS], 25.6 mo [95% CI, 23.1-34.3]) and identified factors associated with these outcomes. Here we describe additional analyses to determine factors that predict durable outcomes (≥ 24 mo) in the pooled D + T population.
The pooled population included randomized pts treated with D 150 mg twice daily + T 2 mg once daily in BRF113220 (Part C; cutoff Jan 2015), COMBI-d (cutoff Jan 2015), and COMBI-v (cutoff Mar 2015). Baseline factors (Table) were analyzed by regression tree analyses to identify predictors of D + T treatment (Tx), PFS, or OS lasting ≥ 24 mo.
Of 617 pts treated with D + T in BRF113220 (n = 54), COMBI-d (n = 211), and COMBI-v (n = 352), 165 (27%) received it for ≥ 24 mo. Long-term PFS (n = 472) and OS (n = 456) analyses excluded pts censored prior to 24 mo. A total of 85 pts (18%) had PFS ≥ 24 mo and 186 (41%) had OS ≥ 24 mo (Table). Regression tree analyses identified baseline lactate dehydrogenase (LDH) as the most predictive factor for durable benefit. Pts with normal vs elevated LDH had improved PFS (median: 12.1 vs 5.5 mo; 2-y rate: 26% vs 5%) and OS (median: 31.6 vs 10.8 mo; 2-y rate: 56% vs 15%). At data cutoff, 172 pts (28%) remained on D + T, including 140 (81%) with Tx ≥ 24 mo and 69 (40%) with PFS ≥ 24 mo.