Outcomes of non-treatment naive melanoma patients with central nervous system relapse.

Abstract

9557 Background: Melanoma has a high probability of central nervous system (CNS) spread. Although first line nivolumab and ipilimumab resulted in 56% response rate and 29.2 months median overall survival (OS) in patients with melanoma brain metastases (MBM), there is paucity of data for patients who develop MBM after prior systemic therapy. As this subgroup is often underrepresented in clinical trials, we aimed to evaluate the OS of non-treatment naïve patients who develop MBM and identify factors related to survival. Methods: In this single-center, retrospective study, consecutive melanoma patients with > 90 days from exposure to either immune checkpoint inhibitor (ICI), targeted therapy (TT), or chemotherapy, to CNS relapse were included. OS was defined as the time between CNS relapse and death by any cause. The Log-Rank method was used to calculate OS. Cox regression analysis was used to identify differences between subgroups. Variables with a p value < 0.1 were included in a multivariate model. A p value < 0.05 was considered statistically significant. Results: Between 2012 and 2018, 135 patients were identified. Median age was 57 (29-92) years, 92 (68%) were male, and median number of prior systemic therapies was 2 (1-6). One-hundred and nine (81%) patients had cutaneous melanoma; acral lentiginous melanoma (ALM) comprised 11 (8%) patients. Molecular studies were available for 123 patients, of whom 61 (50%) were BRAF V600 mutant. Eighty-nine (66%) patients had prior ICI, of whom 33 (37%) had prior exposure to both anti-PD1 and anti-CTLA-4, either as monotherapy or combination. Amongst the BRAF V600 mutant population, 48 (79%) had prior TT. Radiotherapy was given to 112 patients, of whom 55 (49%) had SRS. Median follow-up was 41 (95% CI 30-51) months. Median OS was 6.4 (95% CI 5.3-7.5) months. Patients with ALM, > 3 MBM, ECOG 2-4 and active treatment at CNS relapse (< 30 days from last dose of treatment to MBM diagnosis) were at increased risk of death, whilst subsequent treatment with ICI was related to better survival (Table). On multivariate analyses, age ( p = 0.007), subtype ( p = 0.04), number of MBM ( p = 0.01), active treatment at CNS relapse ( p < 0.001) and subsequent ICI ( p = 0.002) remained statistically significant. Exploratory analyses suggested subsequent treatment with anti-PD1 + anti-CTLA-4 (n = 42) compared favourably to subsequent anti-CTLA-4 only (n = 21) (13 x 7 months, p = 0.004), and was independent of prior ICI. Conclusions: Previously treated melanoma patients who develop MBM have a poor prognosis, but subsequent ICI therapy seems to be associated with better OS. Further clinical investigation to identify optimal anti-PD1-based therapies is warranted for non-treatment naïve patients who develop MBM.[Table: see text]