Abstract 5168: Analysis of biomarker and outcomes data: 7-year update from COLUMBUS

Abstract Background: In the randomized, 2-part, phase 3 COLUMBUS study (NCT01909453), encorafenib (enco) + binimetinib (bini)—an approved regimen in the US, EU, and other countries—improved 7-year PFS and OS vs vemurafenib (vemu) in patients (pts) with BRAF V600E/K-mutant locally advanced unresectable or metastatic melanoma. We investigated genetic and transcriptional correlates of outcomes in an exploratory biomarker analysis of COLUMBUS Parts 1 and 2. Methods: In Part 1, 577 pts were randomized 1:1:1 to enco 450 mg QD + bini 45 mg BID, enco 300 mg QD, or vemu 960 mg BID; in Part 2, 344 pts were randomized 3:1 to enco 300 mg QD + bini 45 mg BID or enco 300 mg QD. Baseline tumor samples were analyzed using the ACE ImmunoID NeXT whole exome sequencing/whole transcriptome sequencing (WES/WTS) assays (Personalis). PFS and OS were analyzed (cutoff: Jan 13, 2023) by treatment arm and presence of specific genetic or transcriptomic alterations. Results: Among treated pts, 668 had WES or WTS data. K-means clustering identified 3 major transcriptional clusters: (1) microphthalmia-associated transcription factor (MITF)-low (invasive dedifferentiated tumor cell phenotype), (2) MITF-high (proliferative melanocytic phenotype), and (3) immune-high. Cluster assignments were significantly associated with biopsy site (P<0.001), with cluster 1 enriched in skin biopsies and cluster 3 in lymph node metastases; biopsy site-independent characteristics were also captured. Improved PFS with enco + bini vs vemu was most pronounced in cluster 3 (Table 1). ERBB2 expression and PI3KCA pathway mutation, previously found to be associated with reduced survival benefit with enco + bini vs vemu, were enriched in clusters 1 and 2 vs cluster 3. Conclusions: Enco + bini showed benefit across molecular subtypes in BRAF V600E/K-mutant melanoma. The contribution of bini was most pronounced in the immune-high cluster. Improved outcomes in this cluster support ongoing trials of combinations with immune checkpoint inhibitors, such as STARBOARD and PORTSIDE. Table 1. Hazard ratios of PFS and OS for enco + bini or enco vs vemu by cluster HR (95% CI) vs vemu PFS OS Enco + bini Enco Enco + bini Enco Overall n=255 vs 100 n=159 vs 100 n=255 vs 100 n=159 vs 100 0.47 (0.35, 0.63) 0.78 (0.57, 1.07) 0.66 (0.50, 0.86) 0.85 (0.63, 1.13) Clusters Cluster 1: MITF-low n=72 vs 34 n=44 vs 34 n=72 vs 34 n=44 vs 34 0.74 (0.44, 1.26) 0.92 (0.51, 1.66) 0.63 (0.40, 1.00) 0.57 (0.34, 0.96) Cluster 2: MITF-high n=120 vs 46 n=76 vs 46 n=120 vs 46 n=76 vs 46 0.48 (0.32, 0.74) 0.75 (0.48, 1.17) 0.66 (0.45, 0.97) 0.86 (0.56, 1.32) Cluster 3: Immune-high n=63 vs 20 n=39 vs 20 n=63 vs 20 n=39 vs 20 0.23 (0.12, 0.44) 0.64 (0.33, 1.21) 0.80 (0.40, 1.59) 1.55 (0.76, 3.15) Numbers are given as the number of pts included in the treatment arm vs the number of pts in the comparator arm (vemu-treated). Statistical significance (P<0.05) is shown by bold font. Bini, binimetinib; enco, encorafenib; MITF, microphthalmia-associated transcription factor; pts, patients; vemu, vemurafenib. Citation Format: Reinhard Dummer, Keith Flaherty, Shibing Deng, Tao Xie, Phineas Hamilton, Nuzhat Pathan, Caroline Robert, Ana Arance, Jan Willem de Groot, Claus Garbe, Helen J. Gogas, Ralf Gutzmer, Ivana Krajsová, Gabriella Liszkay, Carmen Loquai, Mario Mandala, Dirk Schadendorf, Naoya Yamazaki, Paolo A. Ascierto, Alessandra di Pietro, Craig B. Davis. Analysis of biomarker and outcomes data: 7-year update from COLUMBUS [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 5168.