Clinical Features Associated with Long-Term Ruxolitinib Exposure in Myelofibrosis and Predictive Role of Neutrophils

Introduction: Ruxolitinib (RUX), has been the first approved JAK1/2 inhibitor in the USA for patients with advanced stage myelofibrosis (MF). Median duration of RUX therapy in the initial approval studies has been < 3 years, and about 30% of patients completed 5 years of treatment.

Objective: We decided to characterize patients on long-term RUX therapy, e.g. ≥5 years. By comparing their disease features at therapy initiation to patients who received RUX for 6 months to ≤ 3 years, we evaluated factors associated with ≥5 years of therapy and overall survival (OS).

Methods: We performed retrospective chart review of 276 consecutive patients with MF who presented to our institution with newly diagnosed disease, and during their follow-up received RUX for either 6 months to ≤ 3 years (RUX 6m-3y; n= 203) or for ≥5 years (RUX ≥5y; n=73). Both, primary MF (PMF) and post-essential thrombocythemia MF and post-polycythemia vera MF (PPV/PET-MF) patients were included. Responses to RUX (according to IWG-MRT criteria) included all defined responses expressed as one dichotomic variable. Fisher exact test and χ2 were used to compare clinical characteristics between groups at the time of RUX initiation. Overall survival from the time of RUX initiation was estimated by Kaplan-Meier method with log-rank test. Cox proportional hazard model was used for multivariate analysis of variables associated with RUX ≥5y and OS.

Results: Median age of all patients was 65 years (range, 20-94), 62% were males. Major clinical differences between 203 patients with RUX 6m-3y and 73 patients with RUX ≥5y were that RUX ≥5y group had more females and more patients with PPV/PET-MF, and these patients had higher hemoglobin, platelets and neutrophils percentage, but lower white cells, monocytes and blasts (Table 1). No differences were noticed in age, symptoms, splenomegaly, distribution of abnormal karyotype or driver mutations (JAK2, MPL, CALR). Only 33% of patients had next generation sequencing for high-risk molecular mutations (ASXL1, IDH1/2, EZH2, U2AF1/SRSF2) at RUX initiation with non-significantly higher proportion in RUX 6m-3y (12% vs 35%, p = 0.06).

There were more responders on RUX ≥5y (58, 79%) than on RUX 6m-3y (88, 43%), p < 0.001. At least 80% of patients from each group started on RUX at ≥15 mg twice daily. Multivariate analysis showed that achievement of response, having PET/PPV-MF, higher white cells, hemoglobin and neutrophil percentage, and lower peripheral blasts were associated with RUX ≥5y (Table 2A). Response achievement, longer RUX therapy, higher hemoglobin, neutrophils percentage, platelets, and lower age were associated with OS in the entire cohort (Table 2).

Within the group of RUX ≥5y, the only predictor of OS was age (HR 1.09, 95% CI 1.16-1.42, p < 0.01). In those with RUX 6m-3y, factors associated with OS were age (HR 1.05, 1.03-1.08, p < 0.001), platelets ≥100 x 10 ⁹ /L (HR 0.53, 95% CI 0.47-0.90, p = 0.02), neutrophils percentage (HR 0.98, 95% CI 0.97-0.99, p < 0.001) and achievement of response (HR 0.67, 95% CI 0.47-0.86, p = 0.03). Estimated OS from RUX start was 128 months (95% CI, 91-165) and 38 months (95% CI, 29-39) for RUX ≥5y and RUX 6m-3y, respectively (HR 3.95, 95% CI 2.59-6.01, p < 0.001). 8-year survival from RUX initiation was 76% and 18% for RUX ≥5y and RUX 6m-3y, respectively.

Subsequent analysis performed separately for patients with PMF and PPV/PET-MF revealed that more differences between RUX ≥5y and RUX 6m-3y were noticed in PPV/PET-MF patients, whilst those with PMF only differed in blasts and neutrophils percentage (Table 1). The only clinical variables that were associated with RUX ≥5y and OS were age for PMF and PPV/PET-MF, neutrophil percentage for PMF and platelets for PPV/PET-MF, respectively

Conclusion: Our results demonstrate that patients with "less cytopenic" MF and those with PET/PPV-MF can more often remain on RUX ≥ 5 years. Our findings highlight the role of neutrophils percentage as valuable predictive factor of long-term RUX benefit as well as longer survival from RUX initiation. We have also noticed that age at RUX start remained the only factor impacting survival of patients who were on RUX ≥ 5 years, whereas more clinical variables retained significance with shorter RUX exposure.

Figure 1

View largeDownload slide

Figure 1

View largeDownload slide

Close modal