Red Blood Cell Distribution Width May Predict Drug-Induced Anemia and Prognosis in Patients Affected by Primary/Secondary Myelofibrosis Treated with Ruxolitinib

Abstract

Introduction: Myelofibrosis (MF) is often characterized by a multifactorial anemia determined, in part, by bone marrow (BM) fibrosis, extramedullary erythropoiesis and splenomegaly. Ruxolitinib (RUX) is the first-in-class janus kinase 2 (JAK2) inhibitor approved for treatment of MF, proved to reduce spleen volume and decrease symptom burden. The red cell distribution width (RDW) is the measure of erythrocyte volume variability (anisocytosis). RDW has been recognized as a marker of clinical and subclinical systemic inflammation, and its elevation has also been associated with poor outcome in a wide spectrum of benign disorders and in different types of neoplasms.

Methods: We retrospectively evaluated RDW in a single-center series of 200 consecutive patients with primary and secondary MF at RUX treatment initiation and examined any possible correlation with adverse MF features or drug-related anemia and any prognostic impact.

Results: We suggested 20.5% as the optimal cutoff point in RDW values at start of RUX to dichotomize patients in receiver operating characteristic (ROC) analysis for spleen response and for survival. Higher RDW values at RUX start were associated with clinical and laboratory features of an aggressive MF phenotype. Lower spleen response (p < 0.001) and greater odds of drug-related anemia at 3 (p = 0.006) and 6 months (p < 0.001) were also seen in patients with higher RDW. Both increased RDW (considered as a continuous variable) and RDW ≥ 20.5% were associated with shorter overall survival (OS) from RUX initiation in univariate and multivariate analysis: HR 1.25 (95% confidence interval [CI], 1.12-1.40) (p < 0.001) and HR 3.01 (95% CI 1.81-4.99) (p < 0.001), respectively. RDW ≥ 20.5% at RUX start seems to possibly improve patients' sub-stratification along with anemia and conventional prognostic scoring systems.

Conclusions: RDW at RUX start might represent a good indirect measure of MF features and might have prognostic significance for RUX-treated patients affected by MF, aiding in the rapid detection of patients with poor prognosis.

Keywords: Drug-induced anemia; Myelofibrosis; Overall survival (OS); Red blood cell distribution width (RDW); Ruxolitinib (RUX); Spleen response.

Fig. 1

Patient disposition flowchart. Patient flowchart reporting the total number of patients included in the database and the number of patients excluded from the analysis with the corresponding motivations, according to selection criteria. Pts patients, RUX ruxolitinib, Allo-HSCT allogenic hematopoietic stem cell transplant, OS overall survival

Fig. 2

Kaplan–Meier curve for overall survival probability from ruxolitinib initiation using a baseline RDW cutoff of 20.5%; patients with baseline RDW ≥ 20.5% presented shorter OS than patients with RDW < 20.5% [34.6 months (95% CI 23.1–46.2) vs. 84.2 months (95% CI 75.8–92.6)] [HR 3.35 (95% CI 2.11–5.32)] (p < 0.001). CI confidence interval, RUX ruxolitinib, OS overall survival, HR hazard ratio, RDW red cell distribution width

Fig. 3

Kaplan–Meier curves for overall survival probability from ruxolitinib initiation using a baseline RDW cutoff of 20.5% and a baseline HB cutoff of 10 g/dL. A Patients with Hb ≥ 10 g/dL. B Patients with Hb ≥ 10 g/dL. C Combination of all cases with baseline RDW cutoff of 20.5% and baseline HB cutoff of 10 g/dL. RUX ruxolitinib, OS overall survival, CI confidence interval, HR hazard ratio, Hb hemoglobin, RDW red cell distribution width

Fig. 4

Comparison of Kaplan–Meier curves for overall survival probability from ruxolitinib initiation using a baseline RDW cutoff of 20.5% and DIPSS or RR6 risk scores. A Int-2 DIPSS risk patients. B Comparison of sub-stratification based on RDW in Int-2 DIPSS risk patients with other DIPSS risk categories. C RR6 high-risk patients. D Comparison of sub-stratification based on RDW in RR6 high-risk patients with other RR6 risk categories. RUX ruxolitinib, OS overall survival, CI confidence interval, HR hazard ratio, DIPSS Dynamic International Prognostic Scoring System, RR6 response to ruxolitinib after 6 months, Int intermediate, RDW red cell distribution width

Fig. 5

A Association between hemoglobin levels at RUX and baseline RDW, scatter plot. B Box-and-whisker plot showing that patients developing drug-related anemia at 6 months after ruxolitinib initiation presented a higher median baseline RDW. RUX ruxolitinib, OS overall survival, Hb hemoglobin, RDW red cell distribution width. [Drug-related anemia definition can be found in the outcome measures section of the text]