Abstract

6514 Background: Myelodepleted myelofibrosis (MF), characterized bycytopenias, lower JAK2V617F allele burden (JAK2%) and shorter benefit from JAKi ruxolitinib (RUX), has worse survival (OS) compared to myeloproliferative MF. Lower JAK2% and inferior OS is more typical for primary MF (PMF) vs MF from essential thrombocythemia or polycythemia vera (PPV/PET-MF). We sought to investigate the impact of JAK2% (</≥ 50%), cytopenias and the use of RUX in outcome of PMF/PPV-PET-MF patients from our center. Methods: 601 medical charts of JAK2 mutated patients with MF (known JAK2%) were retrospectively reviewed. We divided patients based on the absence (-) or presence (+) of cytopenias (hemoglobin < 10 g/dL or platelets < 100 x10^9/L) and leukocytosis (WBC ≥ 25 x10^9/L) into: Gr1 = (-)/(-) [absence of both]; Gr2 = (-)/(+) [proliferative]; Gr3 = (+)/(-) [cytopenic]; Gr4 = (+)/(+) [cytopenic and proliferative]) and evaluated OS per JAK2 </≥ 50% and PMF vs PPV/PET-MF. We assessed the tolerance of RUX ≥3 years. We used descriptive statistics, Kaplan-Meier curve with log-rank test and regression analysis for demographics, estimation of OS and its comparison. OS was censored at the time of stem cell transplantation. Results: Median age of the entire cohort was 75 years (64% males). Patients with JAK2 <50%, more likely PMF, had more anemia and thrombocytopenia, higher blasts, less leukocytosis, and smaller splenomegaly. JAK2 </≥ 50% did not discriminate OS of the entire group, PMF or PPV/PET-MF (median OS of ~ 45 months for each; only PPV/PET-MF with 64 months). OS of Gr1-4 according to JAK2 </≥ 50% and PMF/ PPV/PET-MF and combined OS of patients with comparable outcome (A-D) are shown in the table. The final groups with distinct outcomes included: A+B) non-cytopenic PET/PPV-MF and PMF irrespectively of proliferation and JAK2% - with one exception of non-cytopenic, proliferative PMF with JAK2 ≥50% - had median OS of 70 months (range, 54-86); D) cytopenic and proliferative PMF with any JAK2% and PPV/PET-MF with JAK2 < 50% of median OS 19 months (range, 13-25), and C) the rest of the cytopenic patients and non-cytopenic, proliferative PMF with JAK2 ≥50% with a median OS of 36 months (range, 32-42). Among the 3 final groups, similar proportion of patients were exposed to RUX during their follow-up: 36% (A+B), 31% (B) and 28% (C), respectively (p =0.13). The most patients in (A+B) group were able to tolerate RUX for ≥3 years (p =0.036). All patients had improved OS with RUX (data to be presented). Conclusions: Cytopenias and/or proliferation, rather than JAK2 </≥50%, define the outcome of PMF and PPV/PET-MF patients. [Table: see text]