The Role of Direct Oral Anticoagulants in Managing Myeloproliferative Neoplasms Patients

Article information

J Korean Cancer Assoc. 2024;.crt.2024.738

Publication date (electronic) : 2024 September 20

doi : https://doi.org/10.4143/crt.2024.738

Ji Yun Lee, Ju-Hyun Lee, Woochan Park, Jeongmin Seo, Minsu Kang, Eun Hee Jung, Sang-A Kim, Koung Jin Suh, Ji-Won Kim, Se Hyun Kim, Jeong-Ok Lee, Jin Won Kim, Yu Jung Kim, Keun-Wook Lee, Jee Hyun Kim, Soo-Mee Bang

Department of Internal Medicine, Seoul National University Bundang Hospital, Seoul National University College of Medicine, Seongnam, Korea

Correspondence: Soo-Mee Bang, Department of Internal Medicine, Seoul National University Bundang Hospital, Seoul National University College of Medicine, Gumi-ro 173 Beon-gil, Bundang-gu, Seongnam 13620, Korea Tel: 82-31-787-7039 E-mail: smbang7@snu.ac.kr

Received 2024 August 5; Accepted 2024 September 19.

Abstract

Introduction

Philadelphia chromosome-negative myeloproliferative neoplasms (MPNs), which include polycythemia vera (PV), essential thrombocythemia (ET), and primary myelofibrosis (MF), are chronic hematopoietic disorders defined by an excessive production of mature myeloid blood cells [1]. Venous and arterial thrombosis is a primary cause of morbidity and mortality due to a marked prothrombotic condition [2,3]. Compared to the general population, MPN patients have a threefold and tenfold higher rate of arterial and venous thrombosis, respectively [4]. At diagnosis, over 20.0% of MPN patients had experienced thrombosis, with arterial and venous thromboses accounting for 16.2% and 6.2% of cases, respectively [5]. During follow-up, the predicted cumulative rates of arterial thrombosis were 5.5%, 1% to 3%, and 1.75% per patient-year for PV, ET, and MF, respectively [6-8]. The estimated cumulative rates of venous thrombosis during follow-up were 1.05%, 0.6%, and 0.76% per patient-year for PV, ET, and MF, respectively [8-10]. MPNs are associated with an increased risk of bleeding [11]. Among newly diagnosed MPN patients, the pooled prevalence of hemorrhagic sequelae was 6.2%, with rates of 6.9%, 7.3%, and 8.9% in PV, ET, and MF, respectively [5]. Gastrointestinal, mucosal, and cutaneous bleeding were all common [5].

The choice of the course and duration of antithrombotic therapy in MPN patients is particularly challenging due to limited data. Major guidelines favor the use of direct oral anticoagulants (DOACs) as first-line therapies to prevent stroke or systemic embolism associated with atrial fibrillation (AF) [12,13]. DOACs are also the first-choice treatment for acute venous thromboembolism (VTE) in the general population and in cancer patients without gastrointestinal (GI) tumor [14]. In the general population, DOACs have been shown to be at least as effective as vitamin K antagonist (VKA), but with a lower rate of major bleeding complications. Given their propensity for thrombotic and hemorrhagic complications, MPN patients might benefit from reduced bleeding risks when using DOACs compared to VKAs. There are no randomized controlled trials comparing VKAs and DOACs in MPN patients; only a handful of retrospective studies exist, leaving the safety and efficacy of DOACs in this population uncertain. How et al. [15] reported that the 1-year cumulative incidence of thrombosis and bleeding on DOACs in MPN patients was 5.5% and 12.3%, respectively. According to the MPN-DOACs study, over a median follow-up period of 1.7 years, there was a 7.2% recurrence of thrombosis and a 5.9% incidence of major hemorrhagic complications on DOACs [16]. DOACs and VKAs appear to have a substantially similar risk-benefit profile in MPN, but this interpretation is based on very limited data. Therefore, this study aimed to conduct a large cohort study using a national healthcare database to examine the incidence and risk factors for thrombotic and bleeding complications in MPN patients with AF and VTE who are treated with DOACs in real-world clinical settings.

Materials and Methods

Results

1. Patient characteristics

The present study included 567 patients with MPN who were newly administered DOACs for AF or VTE during an 11-year period (S1 Fig.). Out of the 368 MPN patients included in the final analysis, 231 (62.8%) were treated with DOACs for AF, and 137 (37.2%) for VTE. Clinical baseline characteristics of the study cohort are shown in Table 1. Patients treated with DOACs for AF had a statistically significant higher median age, exceeding that of patients treated for VTE by 2 years (75 years vs. 73 years, p=0.048). Additionally, the prevalence of elderly patients aged 75 and older was significantly greater in the AF cohort compared to the VTE cohort (51.9% vs. 42.3%, p=0.007). Regarding gender distribution, the proportion of males was significantly higher in the AF group than in the VTE group (58.9% vs. 43.8%, p=0.005).

Table 1.

Clinical features of the cohort, by DOAC indication

Among the MPN subtypes, ET was the most prevalent (48.6%), followed by PV (32.3%), and MF (19.0%). There was a trend in MPN subtype distribution between the AF and VTE groups, especially with a notable difference in the proportion of MF (15.6% vs. 24.8%, p=0.041). Overall, the CHA2DS2-VASc score was statistically significantly higher in the AF group compared to the VTE group. Additionally, the prevalence of hypertension and congestive heart failure was significantly greater in the AF group than in the VTE group (p < 0.001 for both conditions). Among the DOACs, rivaroxaban was the most used (37.2%), followed by apixaban (30.7%), edoxaban (24.5%), and dabigatran (7.6%). More than half of the patients (51.1%) were on antiplatelet therapy, with aspirin being the most common (44.6%), followed by clopidogrel (23.1%). Furthermore, 79.3% of the patients received cytoreductive therapy, with hydroxyurea being the most frequently used (64.9%), followed by anagrelide (27.4%) and ruxolitinib (12.2%). Hydroxyurea was the most used agent in both groups, whereas anagrelide was more frequently used in the AF group compared to the VTE group (31.2% vs. 21.2%, p=0.038). Conversely, ruxolitinib was more commonly used in the VTE group than in the AF group (17.5% vs. 9.1%, p=0.017). In MPN patients treated with DOACs for VTE, the most prevalent subtype was DVT (37.2%), followed by PE alone (34.3%), unusual site VTE (16.8%), and DVT combined with PE (11.7%) (S3 Table).

2. Incidence of thrombosis

The median follow-up for all patients was 22.3 months (IQR, 7.1 to 44.1). In our cohort, 45 patients (12.2%) experienced thrombotic events on DOACs, including 29 (7.9%) with arterial thrombosis and 20 (5.4%) with venous thrombosis. The 1-year cumulative incidence of thrombosis on DOACs was 11.1% (95% CI, 7.7 to 15.2) (Fig. 1A).

Fig. 1.

Cumulative incidence of thrombosis on direct oral anticoagulant in all myeloproliferative neoplasm patients (A), by indication (B), and by myeloproliferative neoplasm subtype (C). AF, atrial fibrillation; ET, essential thrombo cythemia; MF, primary myelofibrosis; PV, polycythemia vera; VTE, venous thromboembolism.

The median follow-up for the AF cohort was 24.7 months (IQR, 7.8 to 46.3). In this cohort, 30 patients (13.0%) experienced thrombotic events, including 22 (9.5%) with arterial thrombosis and 10 (4.3%) with venous thrombosis. The 1-year cumulative incidence of thrombosis on DOACs was 10.8% (95% CI, 3.8 to 15.8). The median follow-up for the VTE cohort was 17.3 months (IQR, 6.2 to 39.4). In the VTE cohort, 15 patients (10.9%) had thrombotic events, including seven (5.1%) with arterial thrombosis and 10 (7.3%) with venous thrombosis. The 1-year cumulative incidence of recurrent thrombosis on DOACs was 11.7% (95% CI, 6.0 to 19.6). No significant difference in thrombosis incidence was observed between the AF and VTE cohorts (p=0.591) (Fig. 1B). Venous thrombosis outcomes occurred significantly more often in the VTE cohort than in the AF cohort, as shown in the S4 Fig. Although there was no association between MPN subtypes and overall thrombosis events, arterial thrombosis events were most common in the ET group, followed by the PV and MPN groups (Fig. 1C, S5 Fig.)

3. Incidence of bleeding

Seventeen patients (4.6%) were hospitalized for major bleeding events while on DOACs, including seven (1.9%) with ICH and 10 (2.7%) with hospitalization for GI bleeding. The 1-year cumulative incidence of hospitalized for major bleeding on DOACs was 3.7% (95% CI, 2.0 to 6.4) (Fig. 2A). There was no significant difference in bleeding incidence between the AF and VTE cohorts (p=0.539) (Fig. 2B). Bleeding outcomes were not related to MPN subtypes (Fig. 2C).

Fig. 2.

Cumulative incidence of bleeding on direct oral anticoagulant in all myeloproliferative neoplasm patients (A), by indication (B), and by myeloproliferative neoplasm subtype (C). AF, atrial fibrillation; ET, essential thrombo cythemia; MF, primary myelofibrosis; PV, polycythemia vera; VTE, venous thromboembolism.

In the AF cohort, 13 patients (5.6%) were hospitalized for major bleeding events, including five (2.2%) with ICH and eight (3.5%) for GI bleeding. The 1-year cumulative incidence of hospitalization for major bleeding on DOACs was 4.3% (95% CI, 2.0 to 8.0). In the VTE cohort, four patients (2.9%) were hospitalized for major bleeding events, including two (1.5%) with ICH and two (1.5%) for GI bleeding. The 1-year cumulative incidence of hospitalization for major bleeding events on DOACs was 2.5% (95% CI, 0.7 to 6.5).

4. Risk factors for thrombosis and bleeding

In thrombosis events, CHA2DS2-VASc scores ≥ 3 (hazard ratio [HR], 3.48; 95% CI, 1.23 to 9.85; p=0.019), the use of concomitant antiplatelet agent (HR, 2.57; 95% CI, 1.35 to 4.88; p=0.004), and cytoreduction (HR, 2.20; 95% CI, 1.08 to 4.49; p=0.030) were statistically significant risk factors for thrombosis (Table 2). Both univariate and multivariate analyses revealed no significant risk factors associated with hospitalization for major bleeding (S6 Table).

Table 2.

Univariate analysis and multivariate analysis for thrombosis

Discussion

This study represents the first investigation into thrombosis and bleeding in MPN patients using DOACs for AF or VTE in an East Asian population. Considering the ethnic differences in thrombosis and bleeding risks, which are influenced by genetic, environmental, and lifestyle factors, our research holds significant importance.

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 [19]. 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 [20,21]. Unusual site thrombosis, accounting for about 10% of all VTE cases, occurs more frequently in MPN patients than in the general population [21-23]. In the current study, splanchnic vein thrombosis accounted for 16.8% of the VTE cohort.

During the median follow-up duration of 22.3 months, 12.2% and 4.6% of patients encountered thrombotic and major bleeding events, respectively. Due to differences in patient populations, study designs, data sources, various biases, and follow-up durations, it is challenging to directly compare our data with previous DOAC studies. While our study showed a somewhat higher rate of thrombotic events compared to the MPN-DOACs study [16], which reported a 7.2% recurrence of thrombosis and a 5.9% incidence of major hemorrhagic complications on DOACs, this apparent difference can be explained by several factors. Increasing age has consistently proven to be an independent predictor of future thrombotic events in MPN [6,24]. Cardiovascular risk factors were significantly and independently associated with an increased rate of total thrombosis in ET patients [25]. The patient cohort included in our study is characterized by an elderly population with a median age of 74 years and a high cardiovascular risk profile, as indicated by a median CHA2DS2-VASc score of 4. These characteristics suggest that our study population was older and had a more significant burden of cardiovascular risk factors compared to those in previous DOAC studies.

The GIMEMA (Gruppo Italiano Malattie Ematologiche dell’Adulto) study reported that among 494 patients with PV or ET treated for arterial or venous thrombosis, 33.6% experienced recurrent thrombosis, while 5.4% suffered major bleeding, with 18.2% receiving long-term VKA therapy [26]. Similarly, the European Leukemia Network (ELN) found that among 206 MPN patients with typical site VTE, approximately 70% of whom were on long-term VKA therapy, 21.8% experienced recurrent thrombosis, and 5.8% experienced major bleeding [27]. To date, no clinical studies have directly compared the efficacy and safety of DOACs versus VKAs in MPN patients. Although comparing retrospective studies has its limitations, the consistent findings of no significant increase in the risk of recurrent thrombosis or bleeding events support the potential efficacy of DOACs in this patient population.

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 [25]. 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 [26-28]. How et al. [15] reported that significant risk factors for thrombosis while using DOACs included a history of previous thrombosis, specific DOACs (dabigatran or edoxaban), and age < 65. 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. The observation that concomitant antiplatelet therapy and cytoreduction are significant risk factors for thrombotic events in patients using DOACs may be attributed to several underlying biases and mechanisms. Firstly, treatment-risk bias is a critical factor; patients with a higher baseline thrombotic risk are more likely to receive intensive treatment regimens, including both antiplatelet agents and cytoreductive therapy. For patients with AF who have a CHA2DS2-VASc score of 2 or greater, along with additional cardiovascular conditions, the simultaneous use of anticoagulants and antiplatelet agents may be required to reduce the risk of thromboembolism. This is particularly relevant for those who have undergone percutaneous coronary intervention for acute coronary syndrome or vascular intervention for peripheral artery disease [30]. For MPN patients with venous thrombosis who have a JAK2 mutation or cardiovascular risk factors, the use of both anticoagulation and aspirin together is recommended [31]. While this dual therapy approach may be necessary for some, it inherently increases the complexity of managing thrombotic risk, potentially leading to an increased incidence of thrombosis despite anticoagulation.

In the context of MPNs, cytoreduction is typically employed to lower elevated blood cell counts and mitigate thrombotic risk. However, the efficacy of cytoreduction in preventing recurrent thrombosis remains inconsistent across studies. In the GIMEMA cohort, cytoreduction halved the risk of thrombosis, but its effect was limited to first arterial thrombosis [26]. In contrast, in the ELN cohort, cytoreduction was not associated with a reduction in recurrent thrombosis [27]. This inconsistency may reflect variations in patient populations, the degree of cytoreduction achieved, and the underlying disease biology. Given the nature of observational studies, unmeasured confounding factors, such as disease characteristics, including JAK2 mutation status, erythrocytosis, and thrombocytosis, may have influenced the results. Additionally, patients receiving concomitant antiplatelet therapy or cytoreductive treatment might have different DOAC dose adjustments or levels of treatment adherence, which could impact thrombosis risk. Further prospective studies are needed to clarify the precise impact of these treatments on thrombosis risk in this patient population.

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. [32] 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. Kaifie et al. [33] 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. 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 [16]. 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 [34]. According to previous studies on MPN patients with AF, HAS-BLED scores were poor predictors of bleeding outcomes [35]. 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.

Electronic Supplementary Material

Article information Continued