Search for a command to run...
Background: Myelofibrosis (MF), particularly Intermediate‐2 (Int‐2) or High‐Risk disease, is a life‐threatening disease for which the Janus Kinase inhibitor (JAKi), ruxolitinib, is currently the only approved therapy. Currently, there is no approved or effective therapy for patients (pts) who lack or lose response to ruxolitinib, and these pts have a notably poor prognosis (median overall survival [OS] 12–14 mo) (Kuykendall et al Ann Hematol 2018; Newberry et al Blood 2017). Imetelstat, a 13‐mer oligonucleotide that specifically targets the RNA template of human telomerase, is a potent competitive inhibitor of telomerase enzymatic activity. In a Phase 2 study of imetelstat (2 doses) in JAKi‐relapsed/refractory (R/R) Int‐2 or High‐Risk MF (MYF2001; NCT02426086), spleen volume reduction and total symptom score improvements and median OS of 29.9 mo (95% CI 22.8, NE) were observed in the 9.4 mg/kg arm (Mascarenhas et al ASH 2018 #685). Aims: To further assess the potential OS benefit of imetelstat, pts treated with 9.4 mg/kg in MYF2001 were compared to a closely matched patient population from real‐world data (RWD), treated contemporaneously. Methods: MYF2001 pts had MF that was R/R to JAKi, defined as documented progressive disease (PD) during or after JAKi treatment. For the historical control, external RWD were collected from a single‐center study which included 114 pts who had discontinued ruxolitinib. A closely matched cohort was identified using the guidelines for inclusion and exclusion criteria as defined in MYF2001 protocol and consisted of pts with MF who had discontinued JAKi due to lack or loss of response and were subsequently treated with best available therapy (BAT) at the Moffitt Cancer Center. OS was measured from the time of JAKi discontinuation to death or censored at last follow‐up. To improve comparability, propensity score weighting approaches using average treatment effect for overlap population (ATO) and stabilized inverse probability treatment weighting (sIPTW) methods were used for critical baseline covariates (age, platelet count, time from diagnosis to JAKi discontinuation, JAKi duration, spleen size, sex, DIPSS score, ECOG, MF type, transfusion status, JAK2 mutation). Sensitivity analyses (to address early deaths post JAKi discontinuation observed in the RWD and the impact of subsequent transplant on OS) were explored to enhance the robustness of the results. Results: 57 pts (treated with imetelstat 9.4 mg/kg) from MYF2001 and 38 pts treated with BAT from RWD (median follow‐up, 23 and 43 mo, respectively) were analyzed with improved balanced baseline covariates after propensity score adjustment. Overlap weighting results (ATO) are summarized in the Figure, showing a 65% lower risk of death with imetelstat vs BAT (hazard ratio [HR], 0.35; 95% CI 0.18, 0.68; P = 0.0019). Sensitivity analysis results were consistent with the primary findings. With sIPTW, median OS was 30.69 mo (95% CI 25.2, not estimable) with imetelstat vs 12.04 mo (95% CI 9.5, 16.6) with BAT (HR, 0.33; 95% CI 0.18, 0.61; P = 0.0003), with 12‐mo OS rates of 85% and 52%, respectively and 24‐mo OS rates of 66% and 24%, respectively. Summary/Conclusion: These analyses showed that treatment with imetelstat was associated with a lower risk of death compared to BAT in closely matched pts with Int‐2 or High‐Risk MF after JAKi failure. Despite the limitations of such comparative analyses, this bridging of RWD with clinical trial data suggests favorable OS of imetelstat treatment in this very poor‐prognosis population and warrants further evaluation. image