PF671 RATE OF THROMBOSIS IN CONTEMPORARY PATIENTS WITH POLYCYTHEMIA VERA UNDER HYDROXYUREA REMAINS EXCEEDINGLY HIGH. RESULTS OF SYSTEMATIC REVIEW AND METANALYSIS

Background: Hydroxyurea (HU) is the recommended first line therapy in high risk polycythemia vera (PV) based on the results of PVSG protocol 08. In this prospective observational study, HU was more effective in reducing the rate of thrombotic events in 51 patients in comparison with historical controls treated with phlebotomy (PHL) alone. Since then, very few studies confirmed these results. Recently, a propensity score analysis of patients enrolled in the ECLAP trial documented superiority of HU in reducing thrombosis compared to PHL only. HU was compared to Interferon (IFN) in three recent randomized controlled trials in PV; unfortunately, primary end‐point of these trials was not the reduction of vascular complications but only hematological response, which is not deemed a valid surrogate of vascular events. Aims: We performed a meta‐analysis to determine the absolute risk of thrombosis, bleeding, acute myelogenous leukemia (AML) and myelofibrosis (MF) in contemporary patients (2008–2018) under HU treatment. Methods: We searched for relevant articles or abstracts in the following databases: Medline, EMBASE, clinicaltrials.gov, WHO International Clinical Trials Registry, LILACS. Sixteen out of 429 published studies reporting number of events using WHO diagnostic classification for PV, met the criteria of our study protocol (registered in PROSPERO; number CRD42018117814). Through a random effect logistic model, incidences, study heterogeneity and effects of confounders were estimated for each outcome at different follow‐ups. Results: Events were collected in 3,236 PV patients, during a mean follow‐up ranging from 0.3 to 12.4 years. Thrombosis: Overall major thrombosis incidence (n = 469) was about 3% per year, obtained by pooling event rates from each study. In meta‐regression analysis accounting for study‐specific confounders such as median age, antithrombotic therapy, cardio‐vascular risk factors and history of thrombosis, a slightly lower estimate of 2.8%.was found. This rate did not change over follow‐up time, as shown by a comparison between a logistic and a negative binomial model, and depended on age. Annual estimates of major thrombosis in patients with a median age of 60, 70 and 80 years were 1.6%, 3.6% and 6.8% respectively. This incidence is approximately 10‐fold higher than the one estimated in the general population. Bleeding: Based on 88 events over 1,485 patients, bleeding pooled incidence was 1% per year, independently of follow‐up duration and antithrombotic therapy, as shown by meta‐regression. Hematological transformations and mortality: Annual rate of AML (n = 63) was fairly constant over time and the cumulative 10‐year incidence was about 4% (0.4% patients/year). In contrast, rate of evolution into MF (n = 157), as predicted by meta‐regression, increased steeply after 5 years of follow‐up. In the 0–5/5–10 years of follow‐up the average annual rate of MF evolution was 1.0% and 5.1% respectively. Mortality (n = 522) followed a similar pattern as MF. Estimates were 2.4%, 12.6% and 56.2% at 1, 5 and 10 years respectively. Summary/Conclusion: The results of this meta‐analysis can be a valid reference for patient communication and counseling, and also constitute a help for sample size calculations in future comparative clinical trials adopting hard efficacy endpoints (thrombosis) in selected populations.