ESMO Congress 2026

The STARBOARD readout concerns a first-line triplet strategy in unresectable or metastatic cutaneous melanoma with BRAF V600 mutations: the BRAF inhibitor encorafenib plus the MEK inhibitor binimetinib, combined with the anti-PD-1 pembrolizumab. It should be noted that the background references provided with this material bear no direct relevance to melanoma or this drug class and contain no information on previously published pivotal Phase III trials of this category; the historical evidence benchmark within this material is therefore undisclosed and to be confirmed, and all comparisons below are limited to figures explicitly stated in the on-site posts.

What was actually reported at the meeting was the STARBOARD Phase III trial (n=220), presented by Dirk Schadendorf at ESMO26: encorafenib + binimetinib + pembrolizumab versus pembrolizumab + placebo in first-line BRAF V600E/K unresectable or metastatic cutaneous melanoma. The core figures can only be quoted as given: ORR 70.0% vs 32.7% (one-sided p<0.0001); mPFS 33.0 vs 3.4 months, HR 0.33 (90% CI 0.24–0.46); OS still immature, HR 0.71 (90% CI 0.48–1.05), with no demonstrated benefit; grade 3-4 TEAEs of any cause 70.0% vs 39.1%, with no treatment-related fatal AEs; discontinuation of encorafenib/binimetinib was 43%. At 31 months of follow-up there were still too few OS events, and OS is immature following early trial closure.

Relative to existing knowledge, the advance of this readout is concentrated in efficacy: the triplet more than doubled ORR, lifted mPFS from 3.4 to 33.0 months with HR 0.33, and attendees called it an impressive PFS. Alongside the main readout, the posts also raised two matters distinct from the efficacy numbers, which should be viewed separately: first, the control choice — fit BRAF-mutant patients would currently receive ipilimumab-nivolumab, so whether pembrolizumab monotherapy was the right control was questioned; second, with too few OS events at 31 months of follow-up, the trend favors the triplet (HR 0.71) but is not yet sufficient to demonstrate a survival benefit — as the saying goes, STARBOARD has not reached the OS shore yet.

As for limitations: OS has not been read out, events are few, and the trial closed early; grade 3-4 TEAEs reached 70.0% (vs 39.1%) with a 43% discontinuation rate for encorafenib/binimetinib, a safety burden that cannot be ignored; and the control-arm design limits generalizability of the results. Subsequent OS data and follow-up updates remain to be confirmed.