P1037: DOES EARLY INTERVENTION IN MYELOFIBROSIS IMPACT OUTCOMES? A POOLED ANALYSIS OF THE COMFORT I AND II STUDIES

Abstract

Background: Myelofibrosis (MF) is characterized by cytopenias, splenomegaly, burdensome symptoms, and poor overall survival (OS). Few studies have investigated if earlier intervention with targeted MF therapies affects response and OS. In a pooled analysis of the COMFORT I and II trials of ruxolitinib (RUX), patients (pts) who received RUX at randomization or after crossover from placebo (PBO) or best available therapy (BAT) had improved OS. Aims: To assess the association of MF disease duration before treatment with disease outcomes using pooled COMFORT data Methods: COMFORT I (NCT00952289) and COMFORT II (NCT00934544) were phase 3 trials of RUX vs PBO or BAT, respectively, in pts with intermediate-2 or high-risk MF who provided written informed consent. In this post hoc analysis, data from RUX-treated pts in both studies were combined (RUX group), and data from the PBO/BAT arms were pooled (control group). Pt subgroups were defined based on disease duration before treatment initiation (≤12 mo or >12 mo from diagnosis). Assessments included frequency of thrombocytopenia (platelets [PLT] <100 Gi/L or PLT transfusion) and anemia (hemoglobin <100 g/L or red blood cell transfusion) events, spleen volume response (SVR; spleen volume reduction ≥35% from baseline [SVR35]), symptom response (MF-Symptom Assessment Form total symptom score [TSS] reduction ≥50% from baseline [TSS50]; available in COMFORT I only), and OS. OS was assessed using the Kaplan-Meier method; pts randomized to PBO/BAT were included in the PBO/BAT group regardless of crossover. A multivariable analysis (MVA) using a logistic regression model was used to examine factors associated with SVR. Results: There were 525 pts in the analysis (RUX: ≤12 mo, n=84; >12 mo, n=216; PBO/BAT: ≤12 mo, n=66; >12 mo, n=159). Median age across groups ranged from 65 to 70 y. Baseline clinical characteristics were generally similar across subgroups, but pts with shorter vs longer disease duration were slightly younger with higher blood counts. Among pts who received RUX, fewer thrombocytopenia events were observed among those treated earlier (≤12 mo vs >12 mo), with differences seen as early as Wk 4–8 on treatment (18% vs 33%) and sustained over time; a similar trend was observed for anemia events (Wk 4–8, 59% vs 72%). The proportion of pts with SVR35 was greater among those who initiated RUX earlier (≤12 mo vs >12 mo) at Wk 24 (48% vs 33%; P=0.0610) and 48 (44% vs 27%; P=0.0149; Table). A numerically greater proportion of pts who initiated RUX at ≤12 mo vs >12 mo achieved TSS50 at Wk 24 (56% vs 40%; P=0.0829; Table). OS at Wk 240 was improved among pts who initiated RUX at ≤12 mo vs >12 mo (63% [95% CI, 51%–73%] vs 57% [95% CI, 49%–64%]; P=0.0430; Table). Comparatively, OS was longer with RUX vs PBO/BAT regardless of disease duration. A sensitivity analysis using a 24-mo cutoff was also conducted but yielded weaker associations between disease duration and SVR, TSS, and OS. In the MVA, a significantly greater binary SVR was seen among pts with shorter (≤12 mo) vs longer (>12 mo) MF disease duration (odds ratio, 2.1; P=0.022). Image:Summary/Conclusion: These findings suggest that earlier RUX initiation in MF may improve clinical outcomes, including cytopenias, SVR, symptom burden, and OS. While “watch and wait” remains a common treatment approach for newly diagnosed patients, these data suggest that pts with MF may benefit from earlier intervention. Additional studies to further evaluate the impact of early intervention are warranted.