Risk of Hemorrhage in Patients with Polycythemia Vera Exposed to Aspirin in Combination with Anticoagulants: Results of a Prospective, Multicenter, Observational Cohort Study (REVEAL)

Background

Polycythemia vera (PV) is associated with an increased risk of thrombosis and major hemorrhage. Aspirin (ASA) is recommended for primary thromboprophylaxis in PV, but is commonly discontinued at the time of initiation of an anticoagulant (AC) due to concern for an increased risk of hemorrhage. High rates of recurrent venous thromboembolism have been reported with ACs alone in patients (pts) with myeloproliferative neoplasms, but limited data exist on hemorrhagic outcomes for ACs combined with ASA in pts diagnosed with PV. To assess the safety of dual antiplatelet and AC therapy in pts with PV, we assessed the rates of hemorrhage among pts enrolled in a large, prospective, observational cohort study.

Methods

The REVEAL study is a prospective, non-interventional, observational study of pts with PV conducted at 204 sites in the United States. Retrospective data up to 6 months before enrollment and prospective data following enrollment were entered into an electronic data capture form. Retrospective and prospective data pertaining to bleeding events and use of ACs and/or ASA were extracted for this analysis.

Post-enrollment hemorrhagic events were graded according to CTCAE by treating physicians. Pt subgroups (AC only, ASA only, AC + ASA, or neither) were defined by treatments received prior to enrollment. Exposure-adjusted hemorrhagic event rates and the probability of hemorrhage in the post-enrollment period were estimated by subgroups. The association between risk of hemorrhage and the use of AC in combination with ASA after enrollment was assessed with a Cox proportional hazards model, with AC and ASA use as a time-dependent covariate, adjusting for age, sex, disease duration, and history of bleeding prior to enrollment. Time to first post-enrollment hemorrhagic event was modeled with time censored at death, date of discontinuation, or last visit; analyses were conducted for all hemorrhagic events and severe (grade 3 or 4) events.

Results

A total of 2510 pts were enrolled in REVEAL and included in this analysis. At the time of enrollment, the median age was 67 (range 22 to 95) years, 54.2% were men, 56.0% were diagnosed with hypertension, and 19.9% reported a history of thrombotic events. The ASA only subgroup consisted of 1490 (59.4%) pts, AC only 183 (7.3%) pts, and AC + ASA 108 (4.3%) pts. The most frequently prescribed ACs were warfarin, rivaroxaban, and apixaban.

With a median (range) follow-up of 2.3 (0 to 3.6) years, 110 (4.4%) pts developed a hemorrhagic event. The most common sites were gastrointestinal (n=51, 46.4%), cutaneous (n=27, 24.5%), and genitourinary (n=12, 10.9%). Forty-six pts (1.8%) developed ≥1 severe hemorrhage, most commonly gastrointestinal (23/46, 50%) and central nervous system (7/46, 15.2%). Of the 46 pts with severe events, 38 (82.6%) required hospitalization and 7 (15.2%) were fatal (2 gastrointestinal, 2 subdural hemorrhages, 1 subdural hematoma, 1 cerebral hemorrhage, and 1 hemorrhagic stroke). At the end of the study period, 8.4% of pts (n=211) were receiving warfarin, 3.9% (n=99) rivaroxaban, and 2.6% (n=65) apixaban.

The exposure-adjusted rate (per 100 pt-years [95% CI]) for any hemorrhage in the overall cohort was 2.0 (1.64, 2.39); in the ASA only subgroup and in the AC + ASA subgroup, these rates were 1.5 (1.09, 1.93) and 7.9 (4.17, 11.56), respectively. The overall 3-year cumulative incidence was 22.1% (12.42%, 33.52%) for any hemorrhagic event (Figure) and 3.0% (0.80%, 7.81%) for severe hemorrhagic events. Overall, pts receiving AC in combination with ASA in the post-enrollment period were over 4-fold as likely to have a hemorrhage (HR [95% CI] = 4.22 [2.57, 6.94]; P <0.0001) compared with those who received either one of the treatments alone, or neither treatment. Notably, the risk of severe hemorrhage was significantly higher for pts receiving AC and ASA combination in the post-enrollment period (HR [95% CI] = 3.06 [1.35, 6.95]; P=0.0074).

Conclusions

In a large prospective cohort of pts with PV, the combination of ACs with ASA was associated with an increased risk of bleeding compared with ASA alone. Based on the observed 3-fold increased risk of severe hemorrhage associated with the combination of ACs and ASA, caution is required when prescribing antiplatelet agents with ACs for the treatment of thrombotic disorders in pts with PV.

View largeDownload slide

View largeDownload slide

Close modal