Red Blood Cell Transfusion Independence Status Is an Independent Predictor of Survival: A Post Hoc Time-Dependent Analysis of the Phase 3 Simplify-1, Simplify-2, and Momentum Trials
Vikas Gupta,Claire N Harrison,Boris Gorsh,Bharat Patel,Zhaohui Wang,Molly Purser,Catherine Ellis,Bryan Strouse,Dwaipayan Patnaik,Jun Kawashima,Ruben A. Mesa
DOI: https://doi.org/10.1182/blood-2023-178772
IF: 20.3
2023-11-28
Blood
Abstract:Introduction: The management of anemia in myelofibrosis (MF), a chronic progressive myeloproliferative neoplasm, remains an ongoing medical need as it is present in approximately 30% of patients (pts) at diagnosis and >50% after 1 year, contributing to negative impacts such as fatigue, medical complications, and poor quality of life. Furthermore, anemia and resulting transfusion need are independent predictors of poor prognosis. Momelotinib (MMB) is a Janus kinase (JAK) 1/JAK2/activin A receptor type 1 (ACVR1) inhibitor that has demonstrated spleen, symptom, and anemia benefits, including increases in hemoglobin (Hb) and reduction or elimination of red blood cell (RBC) transfusion requirements in pts with MF. Data from the phase 3 SIMPLIFY-1 (S1), SIMPLIFY-2 (S2), and MOMENTUM studies have shown an association between transfusion independence at wk 24 and prolonged overall survival (OS) with MMB (Mesa R, et al. Leukemia. 2022; Verstovsek S, et al. ASH 2022. Poster 3028). Because these analyses did not adjust for additional prognostic factors or effect modifiers nor potential longitudinal changes in transfusion status over time, the impact of transfusion status on OS has not been fully characterized. Here, we describe an analysis investigating the prognostic influence of RBC transfusion status over time and other covariates on OS from the S1 (NCT01969838), S2 (NCT02101268), and MOMENTUM (NCT04173494) studies irrespective of the treatment received. Methods: These treatment-agnostic OS analyses incorporated data from 3 multicenter, international, randomized trials investigating MMB vs an active comparator. Analyses were performed in the safety populations of S1 (N=430; MMB vs ruxolitinib [RUX], JAK inhibitor [JAKi] naive), S2 (N=156; MMB vs best available therapy [88.5% RUX], JAKi experienced), and MOMENTUM (N=195; MMB vs danazol, JAKi experienced). In the current analysis, which accounts for evaluations of transfusion status every 4 wk, time-dependent covariates (RBC transfusion status, categorized as either transfusion independent [TI; absence of RBC transfusion and no Hb level <8 g/dL in the prior 12 wk] or non-TI [not satisfying the criteria for TI]) and fixed baseline covariates (age, sex, MF subtype, Eastern Cooperative Oncology Group [ECOG] status, JAK2 mutational profile, Total Symptom Score, International Prognostic Scoring System [IPSS] score for S1, Dynamic IPSS [DIPSS] score for S2 and MOMENTUM, platelet count, baseline spleen volume as assessed by central reading of magnetic resonance imaging/computed tomography scan) were evaluated to determine their prognostic impact. An extended Cox model was used to generate an overall hazard ratio (HR) using all available follow-up time for each study, and independent variables were retained if a backward variable selection method determined that they were significantly associated with OS. Results: Significant covariates associated with survival across the 3 trials are presented in the table, with RBC transfusion status as a consistent prognostic variable. In S1, age, platelet count, and baseline spleen volume were also significantly associated with OS, and in S2, significant covariates included age, baseline spleen volume, and DIPSS. In MOMENTUM, only RBC transfusion status was significantly associated with survival. When accounting for differences in prognostic factors and effect modifiers as well as changes in transfusion status over time, results from MOMENTUM demonstrate a strong and significant relationship between transfusion status and OS (HR, 5.18; 95% CI, 1.86-14.47; P=.0017), suggesting that non-TI pts have a higher risk of all-cause mortality of more than 5 times that of TI pts. Similar results with respect to the survival benefits of TI status were observed in both S1 (HR, 3.32; 95% CI, 2.31-4.78; P<.0001) and S2 (HR, 1.87; 95% CI, 1.07-3.29; P=.0287). Conclusions: These data demonstrate that when accounting for differences in known prognostic factors and effect modifiers as well as changes in RBC transfusion status over time, a consistent and statistically significant relationship was observed between transfusion status and OS across all 3 phase 3 MMB trials. Consistent with inclusion of transfusion status in DIPSS-plus risk assessment, these data validate TI status as a clinically relevant and meaningful endpoint that is prognostic for survival in JAKi-naive and JAKi-experienced disease settings.
hematology