Impact of systemic therapy sequencing on overall survival for patients with advanced BRAF-mutated melanoma.

Abstract

9552 Background: Both immune checkpoint inhibitors (ICI) and BRAF targeted therapy (TT) are effective treatments for patients with advanced BRAF-mutated melanoma. However, the choice of first-line (1L) therapy is at the discretion of treating oncologists without clear guidance from current available data or established guidelines. Utilizing prospectively collected data from the Canadian Melanoma Research Network (CMRN) database, we provide real-world evidence to highlight the impact of sequencing these therapies. Methods: Prospective data from 9 cancer centres in Canada was retrieved from the CMRN database for patients with unresectable/metastatic melanoma, with BRAF targetable subtypes, who received at least one-cycle of 1L palliative-intent ICI or TT, and at least 1-year of follow-up. We categorized patients into 2 groups: 1L BRAF±MEK inhibitors with/without subsequent PD-1±CTLA-4 inhibitors (1L-TT), or vice versa (1L-ICI). The primary study outcome was overall survival (OS). Survival outcomes were analyzed through Kaplan-Meier methods, and multivariable Cox analysis was utilized to account for potential confounders. Results: Our study (N=235) included 152 and 83 patients in 1L-TT and 1L-ICI groups, respectively. Combined BRAF-MEK inhibitors accounted for 59% of the 1L-TT group, whereas single-agent IO accounted for 66% of the 1L-ICI group. There were 93 patients who received second-line (2L) therapy, with a non-significant trend of 1L-TT group receiving more 2L therapy compared to 1L-ICI group (65% vs. 43%, P=0.404). Neither treatment group showed significant differences in median time on 1L therapy (P=0.645) or 2L therapy (P=0.686). The 1L-ICI group was associated with a favourable median overall survival (OS) compared to 1L-TT group (19.3 vs. 10.0 months, P=0.031). Specifically, the ICI only group had the highest median OS, followed by TT-ICI sequence, ICI-TT sequence, and TT only groups respectively (not reached vs. 38.3 vs. 16.9 vs. 6.1 months, P<0.001). However, this OS benefit (HR 0.89, 95% 0.51-1.53, P=0.644) was non-significant upon controlling for confounders such as baseline metastatic sites >2 (HR 2.07, 95%CI 1.24-3.46, P=0.006) and ECOG ≥2 (HR 3.47, 95%CI 2.02-5.97, P<0.001) in multivariable Cox analysis. Conclusions: There was no significant difference in OS between 1L-TT and 1L-IO groups. Rather, OS is driven mostly by the patient’s clinical status and tumour-associated features. Our study provides real-world evidence in an understudied area. Further studies are needed to validate our findings to inform guideline development.[Table: see text]