To assess the efficacy and safety of DOACs in MPN patients given their thrombogenic and hemorrhagic tendencies, we conducted a large-scale retrospective cohort study. Among MPN patients, 63% used DOACs for AF and 37% for VTE. Several reasons might explain why the AF cohort was larger than the VTE cohort. First, AF is relatively common in MPN patients; one study found that 13.5% of patients with ET or PV had atrial arrhythmias, 97% of which were AF. This prevalence increases to 31% in patients aged 80 or older [21]. Second, DOACs are prescribed to AF patients not only for secondary but also for primary prevention of stroke. Third, there may be stronger clinical guidelines or more substantial evidence supporting the use of DOACs in AF compared to VTE. Although DOACs are recommended for treating PE and DVT among VTE cases, there is no clear evidence supporting their use in unusual site VTEs [22, 23]. Unusual site thrombosis, accounting for about 10% of all VTE cases, occurs more frequently in MPN patients than in the general population [23–25]. In the current study, splanchnic vein thrombosis accounted for 16.8% of the VTE cohort.
With a median follow-up of 22.3 months, 12.2% of patients experienced thrombotic events on DOACs, with 7.9% being arterial and 5.4% being venous. The one-year cumulative incidence of thrombosis on DOACs was 11.1%. Due to differences in patient populations, study design, data sources, numerous biases, and variations in follow-up time, it is difficult to directly compare our data with previous studies. However, it appears that thrombosis events are somewhat higher during DOAC use in our study compared to prior research. Increasing age has consistently proven to be an independent predictor of future events in MPN [6, 26]. The current study included a median age of 74 years, which is older compared to previous studies that reported median ages of 65 to 71 years [17, 18]. Cardiovascular risk factors were significantly and independently associated with an increased rate of total thrombosis in ET patients [27]. Our study had a median CHA2DS2-VASc score of 4, indicating a high cardiovascular risk. Although we defined incident cases of thrombosis through several steps, the use of national insurance claims data likely included prevalent cases as well.
Within the AF cohort, 13.0% experienced thrombotic events, with 9.5% being arterial and 4.3% venous. According to the Mahé study, MPN patients with atrial arrhythmias have a very high incidence of thrombosis (HR = 2.38; 95% CI, 1.23–4.60, p = 0.01), predominantly arterial events [21]. There was no statistically significant difference in total thrombosis outcomes on DOACs between the AF and VTE cohorts. However, the VTE cohort experienced statistically significantly more VTE outcomes compared to the AF cohort.
Traditional risk factors for MPN-associated thrombosis include advanced age (> 60 years), sex, a history of thrombosis, leukocytosis, and the presence of the JAK2 V617F mutation [27]. The risk factors for thrombosis recurrence were age over 60 and a history of prior thrombosis, while the subtype of MPN does not predict recurrence [12, 13, 28]. How et al. reported that significant risk factors for thrombosis while using DOACs included a history of previous thrombosis, specific DOACs (dabigatran or edoxaban), and age < 65 [17]. In the current study, CHA2DS2-VASc scores ≥ 3, concomitant use of antiplatelet agents, and cytoreduction therapy were identified as statistically significant risk factors for thrombosis on DOACs. While MPN subtype was not linked to overall thrombosis events, ET had the highest incidence of arterial thrombosis events, followed by PV and then MF (p = 0.106).
Thrombosis recurrence risk can be reduced with antithrombotic therapy and/or cytoreduction. According to a systematic review of MPN patients with a history of VTE, a combination of anticoagulation and cytoreduction may offer the lowest recurrence risk [29]. However, most studies had a high risk of bias, and clinical and statistical heterogeneity led to inconsistent findings, warranting caution in interpretation. In this study, the use of concomitant antiplatelet agents and cytoreduction emerged as significant risk factors for thrombosis outcomes. This observation can be explained by treatment-risk bias, where patients with higher baseline risk factors are more likely to receive intensive treatment. Patients with AF and additional cardiovascular conditions such as coronary artery disease or peripheral artery disease may require simultaneous use of anticoagulants and antiplatelet agents to reduce thromboembolism risk [30]. For MPN patients with venous thrombosis who have a JAK2 mutation or cardiovascular risk factors, it is recommended to use both anticoagulation and aspirin together [31]. In the overall GIMEMA cohort, cytoreduction halved the risk of thrombosis, but its effect was limited to first arterial thrombosis [12]. In the ELN cohort, cytoreduction was not associated with a reduction in recurrent thrombosis [13].
MPN patients are prone to bleeding. A variety of factors, including differences in study design, definition of bleeding, selection bias, recording bias, and discrepancies in follow-up time, contribute to the inconsistent reporting of the incidence and risk factors of bleeding in MPN. Wille et al. reported a 5.7% incidence of major bleeding in MPN patients and found that the use of antiplatelet agents or anticoagulants was associated with bleeding outcomes, while age, gender, MPN subtype, and mutation status were not [29]. Kaifie et al. found that 8.2% of MPN patients experienced major bleeding, not linked to thrombocytosis, thrombocytopenia, or use of antiplatelet agents, VKAs, or DOACs, but rather to previous thromboembolic events, splenomegaly, and heparin administration [32]. The MPN-DOACs study reported a 5.9% incidence of major bleeding in MPN patients, identifying the MF subtype and the use of dabigatran as significant risk factors for bleeding on DOACs [18]. In this study, 4.6% of patients were hospitalized for major bleeding events while on DOAC therapy, including 1.9% with ICH and 2.7% with GI bleeding. Although the number of bleeding events was low, necessitating cautious interpretation, no statistically significant risk factors for bleeding outcomes with DOAC use were identified.
This study has several limitations. First, erroneous coding could result in misclassification bias, especially in the absence of laboratory validation results in HIRA databases. Second, due to the retrospective observational nature of the study, only associations can be drawn, and causal relationships cannot be inferred. Third, although an incident case was defined through several steps, prevalent cases may be included in this study. Fourth, a limitation of analyzing data using claims records is the inability to access information on laboratory findings related to thrombosis and bleeding risk factors, such as leukocytosis, thrombocytosis, thrombocytopenia, mutation status, and underlying inherited thrombophilia. Fifth, it was not possible to verify adherence and compliance. In a real-world setting, adherence to DOACs might be lower, potentially leading to suboptimal anticoagulation and a higher risk of thrombotic events. Lastly, the current study did not use a known bleeding score, such as HAS-BLED (Hypertension, Abnormal renal and liver function, Stroke, Bleeding, Labile international normalized ratio, Elderly, Drugs, or alcohol), which has been validated for VKAs but not for DOACs.[33] According to previous studies on MPN patients with AF, HAS-BLED scores were poor predictors of bleeding outcomes [34]. There is no proven bleeding score system for DOAC use yet, but a suitable system is expected to be established in future studies.
To date, the optimal anticoagulant approach remains challenging in MPN. Despite the limitations, such as the retrospective nature of the analysis, and the short follow-up period, our findings show that DOACs were both effective and safe. Future directions would include prospective longitudinal studies to assess the long-term efficacy of DOACs in anticoagulated MPN patients.