AUTHORS
Yan Xu, MD, MSc, Rob Sellar, FRCPath, MBBChir2, Ranjeeta Mallick, PhD3, Marc Rodger, MD, MSc4
AFFILIATIONS
1 Ottawa Hospital Research Institute, Ottawa, Ontario, Canada
2 University College London, London, United Kingdom
3 Ottawa Hospital Research Institute, Ottawa, Ontario, Canada
4 McGill University, Montréal, Quebec, Canada
ABSTRACT
Background: Clonal hematopoiesis (CH) involving the JAK2 gene has been identified as a novel predictor of incident venous thromboembolism (VTE). However, whether JAK2 CH impacts long-term outcomes following a diagnosis of unprovoked VTE remains unclear. We sought to evaluate the impact of JAK2 CH on risks of recurrent VTE and anticoagulant-associated major bleeding after a first episode of unprovoked VTE.
Methods: REVERSE-2 study was an international, multi-centre prospective management study that included unselected patients from 44 secondary and tertiary care centres. Patients were included if they had a first episode of unprovoked, objectively identified VTE treated with 5–12 months of anticoagulation. Baseline demographic characteristics, clinical data and DNA samples were collected at enrolment. Females with 0 or 1 risk factor for VTE recurrence (hyperpigmentation, erythema and swelling; VIDAS D-dimer ≥250 ng/mL; age ≥ 65 years; body mass index [BMI] ≥ 30 kg/m2) stopped anticoagulation, whereas other individuals were suggested to continue anticoagulation. The study exposure was JAK2V617F CH identified using digital droplet PCR ≥ 0.2% variant allele fraction (VAF). The primary outcome was recurrent VTE during follow-up, and secondary outcome was major bleeding. Crude hazard ratios with Firth correction were calculated with JAK2 wildtype as reference, then adjusted by age, sex, and anticoagulant type (direct oral anticoagulants vs. vitamin K antagonists).
Results: Of 709 REVERSE-2 participants who consented to DNA storage, 12 (2.4%) had a JAK2V617F CH mutation identified (median VAF 2.2%, range 0.2% – 62.4%), while five had a mutation detected at VAF <0.2%. Median follow-up was 1 year with no difference by JAK2V617F CH status. Mean age was higher among JAK2V617F CH carriers compared to non-carriers (61.1 vs. 54.7 years, P=0.035), with no difference with BMI or site of index VTE. VIDAS D-dimer was similarly distributed between the two groups (median 216.8 ng/mL vs. 218.4 ng/mL, P = 0.676). Among JAK2 CH and non-CH carriers, 25% were stratified as low risk for VTE recurrence and stopped anticoagulation in both groups, while the remainder continued anticoagulant treatment.
Among 531 individuals who continued anticoagulation, recurrent VTE occurred in one of nine JAK2V617F CH carriers, corresponding to a cumulative incidence of 11.2% (95% CI 0.3% to 62.6%) over the 1-year follow-up. On the other hand, 8 of 522 (1.2%, 95% CI 0.4% to 2.6%) of non-CH carriers had recurrence while on anticoagulation during the same period. Compared to non-CH carriers, individuals with JAK2 CH had a 9-fold higher risk of recurrent VTE during extended anticoagulant therapy (hazard ratio 9.0, 95% CI 1.3 – 64.8, P=0.029). There was no difference by JAK2V617F CH status in risks of major bleeding during anticoagulation (P = 0.488), recurrent VTE following anticoagulant discontinuation (P = 0.411), and all-cause mortality (P = 0.09).
Conclusion: After a first unprovoked VTE, JAK2 CH was associated with higher age but no other established predictor of recurrent VTE. Nonetheless, JAK2 CH conferred an increased risk of recurrent VTE during extended anticoagulation. The impact of JAK2V617F CH on optimal anticoagulant duration, dosing and role of adjunctive therapy among patients with unprovoked VTE warrant further exploration.
DISCLOSURES
Yan Xu, MD, MSc: No relevant conflicts of interest to declare.