C018 | OVERALL SURVIVAL AND DURATION OF TREATMENT IN PATIENTS WITH ADVANCED SYSTEMIC MASTOCYTOSIS RECEIVING AVAPRITINIB VERSUS MIDOSTAURIN OR BEST AVAILABLE THERAPY IN A REAL-WORLD SETTING

Introduction In this retrospective study of the efficacy of avapritinib, a selective KIT D816V inhibitor, in adults with advanced systemic mastocytosis (AdvSM), we compared overall survival (OS) and duration of treatment (DOT) in patients treated with either avapritinib in the PATHFINDER (NCT03580655) study or best available therapy (BAT) in a real-world (RW) retrospective chart review study (NCT04695431). Methods Median follow-up for PATHFINDER data was 26.3 months. Avapritinib in PATHFINDER was compared in the first-line (1L) setting to RW midostaurin and in later lines (2L+) to all BAT used in RW clinical practice. Inverse probability of treatment weighting (IPTW) was used to adjust for baseline differences between cohorts. IPTW-weighted Cox proportional hazards regression models were used to compare outcomes between cohorts. Results Baseline characteristics are in the Table. In 1L IPTW-weighted analyses, median (95% confidence interval [CI]) OS was not reached (NR) (not estimable [NE], NE) and 32.2 months (20.0, 44.6) and median DOT (95% CI) was 41.3 months (33.9, 41.3) and 13.0 months (7.5, 25.5) in the avapritinib (n=38) and midostaurin cohorts (n=58), respectively. In 1L IPTW-weighted Cox analyses, OS (hazard ratio [HR] [95% CI]: 0.19 [0.06, 0.57]; P=0.003) and DOT (HR [95% CI]: 0.37 [0.19, 0.70]; P=0.002) were significantly longer in avapritinib- vs midostaurin-treated patients. 2L+ BAT patients contributed data on 104 lines of therapy. Common 2L+ agents were midostaurin (46.1%), cladribine (32.6%), and hydroxyurea (7.9%). In 2L+ IPTW-weighted analyses, median (95% CI) OS was NR (30.2 months, NE) and 17.9 months (14.8, 36.5), and median DOT was 21.3 months (10.5, NE) and 5.4 months (3.5, 9.8) in avapritinib (n=67) and BAT (n=73) cohorts, respectively. In 2L+ IPTW-weighted Cox analyses, OS (HR [95% CI]: 0.34 [0.17, 0.75]; P=0.008) and DOT (HR [95% CI]: 0.35 [0.21, 0.58]; P<0.001) were significantly longer in avapritinib-treated vs BAT patients. Conclusions Comparisons of real-world data with clinical outcomes have intrinsic limitations, which we have ameliorated using IPTW. In this context, our findings support the use of avapritinib treatment for AdvSM as 1L treatment and for patients with prior treatment, as shown by the significantly better OS and DOT vs 1L midostaurin and 2L+ BAT, respectively.