Abstract
Introduction
Atrial fibrillation (AF) poses a substantial worldwide health concern, significantly increasing the risk of stroke and morbidity. Direct oral anticoagulants (DOACs) such as apixaban are recommended over vitamin K antagonists for the management of AF. However, the impact of thyroid abnormalities on DOACs, specifically apixaban in AF patients remains underexplored. Given the limited data, this study aims to evaluate the effectiveness and safety of apixaban in AF patients with uncontrolled hypothyroidism.
Methods
This study was a retrospective cohort analysis that categorized patients into two sub-cohorts according to their hypothyroidism status at the time they began apixaban treatment: a control group (without hypothyroidism) and an active group (with uncontrolled hypothyroidism). The primary outcome assessed was the rate of thrombosis events following the initiation of apixaban, while bleeding, stroke, and venous thromboembolism (VTE) events were considered as secondary outcomes. Logistic regression analysis was performed, with a p-value of less than .05 deemed statistically significant.
Results
Among 292 patients included, 51 had uncontrolled hypothyroidism, and 241 were in the control group. Both groups had a median age of 70 years, with predominantly female patients. Any thrombosis events were higher in the uncontrolled hypothyroidism at crude analysis (17.6% vs 8.4%; p-value = .04); as well as higher odds at regression analysis [aOR: 2.40, 95%CI 0.99-5.83; p-value = .05]. In addition, stroke and major bleeding events were significantly higher in the uncontrolled hypothyroidism group (aOR: 4.26, 95%CI 1.51-12.00; p-value = .006 and aOR: 6.21, 95%CI 1.73-22.19; p-value = .005, respectively). The rate of VTE events and minor bleeding did not differ significantly between the two groups.
Conclusions
The use of apixaban in patients with AF and uncontrolled hypothyroidism was linked to higher rates of thrombosis and major bleeding compared to those without known hypothyroidism. These findings highlight the need for further research through larger prospective studies in this often-overlooked population.
Introduction
Atrial fibrillation (AF) is one of the most common supraventricular arrhythmias resulting from irregular heartbeats. Approximately 59.6 million patients suffer from AF worldwide. 1 AF is associated with a 4 to 5-fold increase in the risk of stroke, with the high mortality and morbidity associated with ischemic stroke,1–3‐3 AF imposes a 1.5-fold increase in the cost of care, and the risk of death from AF-related stroke is doubled. 3
In 2010, the global prevalence of AF was 596.2 and 373.1 per 100,000 in male and female individuals, respectively. 4 Prevalence and incidence rates increased significantly with increasing age. 4 Other comorbidities that are commonly associated with AF include obstructive sleep apnea, chronic kidney disease (CKD), pulmonary disease, and hypo- and hyperthyroidism. 3
The pharmacological management of AF includes rate and rhythm control, and cardioembolic stroke prevention with anticoagulants. 5 The clinical guidelines favor direct oral anticoagulants (DOACs) over vitamin K antagonists, as they are associated with lower major bleeding and comparable or even better effectiveness in the prevention of cardioembolic stroke. 5 The ARISTOTLE trial revealed that apixaban significantly reduced the risk of stroke, systemic embolism, major bleeding, and death in patients with AF. 6
The relationship between AF and hypothyroidism can be multifaceted: first, overtreatment of hypothyroidism can cause hyperthyroidism, which increases the risk of AF. 7 Second, hypothyroidism is associated with some cardiovascular diseases, such as hypertension and diastolic dysfunction, that usually coexist with AF.8,9 Additionally, hypothyroidism is associated with other risk factors for atherosclerotic cardiovascular disease that can further increase the risk of thrombosis and stroke, including increased serum cholesterol, C-reactive protein, and homocysteine.9,10
Interestingly, thyroid function influences the hemostatic system, where low levels of thyroid hormone shift the hemostatic system to a hypocoagulable and hyperfibrinolytic state. Hyperthyroidism is known as a risk factor for the development of AF. 11 It is associated with a shortening in the repolarization phase of the intracellular potential in the atrium, which could lead to increased heart rate and, ultimately, result in an irregular rhythm. 12 On the other hand, the risk of cardiovascular events is reported more in hypothyroidism patients, although the direct association with AF remains conflicting. 13
The lack of clinical evidence in this population highlights concerns regarding the use of apixaban in patients with thyroid abnormalities. Therefore, this study aims to evaluate the effectiveness and safety of standard dosing apixaban in patients with uncontrolled hypothyroidism.
Materials and Methods
Study Design
This retrospective cohort study was conducted at King Abdulaziz Medical City in Riyadh, spanning from January 01, 2016, to December 31, 2019. The study enrolled patients with confirmed AF who were prescribed apixaban for the prevention of cardioembolic stroke. Patients were categorized into two sub-cohorts based on their thyroid function status at the time of apixaban initiation and within three months thereafter: those without hypothyroidism (control group) and those with uncontrolled hypothyroidism (active group). Uncontrolled hypothyroidism is defined as known to have hypothyroidism with an abnormality of the thyroid panel (ie high TSH > 4.5 mIU/L, low T4 < 9 pmol/L, and/or low T3 < 2.9 pmol/L) within three months of apixaban initiation. All eligible patients were followed for at least one year. The study, conducted under the auspices of the King Abdullah International Medical Research Center (KAIMRC) in October 2023 (Ref. # NRC23R/638/10). Due to the retrospective observational design, informed consent was waived, with all methods adhering to relevant guidelines and regulations.
Study Setting
This study was conducted at King Abdulaziz Medical City, a tertiary-care academic referral hospital located in Riyadh, Saudi Arabia, with a bed capacity of 1923. King Abdulaziz Medical City provides a wide range of healthcare services, from primary healthcare to highly specialized tertiary care, providing to the diverse health needs of its target population.
Study Participants
All adult patients (age ≥18 years) with confirmed AF who received apixaban during the study period were assessed for eligibility. Patients were excluded if they were known to have liver cirrhosis (Child-Pugh score of C), a mechanical valve, hyperthyroidism, antiphospholipid antibody syndrome (APLS), or a poor adherence history, as documented by their medical records. Other exclusion criteria include patients with stable hypothyroidism (defined as patients who have an established diagnosis of hypothyroidism on levothyroxine and thyroid-stimulating hormone; TSH, thyroxine; T4, and triiodothyronine; T3 within the normal value), incomplete data/laboratory results, or those with a follow-up period of less than three months.
Data Collection
We collected patients’ demographic data, comorbidities, history of bleeding, laboratory tests including coagulation profile (ie INR, activated partial thromboplastin time (aPTT), D-dimer, and fibrinogen), liver and renal function tests, complete blood count (ie hemoglobin and platelet levels), and thyroid panel (TSH, T4, and T3). Additional information on apixaban dose and timing of initiation, concurrent use of antiplatelet medications, and gastrointestinal (GI) prophylaxis were also documented. Furthermore, we collected data on various clinical events, such as stroke events (either ischemic or hemorrhagic), venous thromboembolism (VTE), any thrombosis events, major and clinically relevant non-major bleeding (CRNM) as defined by the International Society of Thrombosis and Hemostasis (ISTH). Lastly, risk scores for stroke and bleeding, specifically the CHA2DS2-VASc and HAS-BLED scores. The data for each patient was collected and extracted from the electronic hospital's record system, Best Care 2.0A, and compiled into the Research Electronic Data Capture (REDCap) software.
Outcomes
This study aimed to compare the effectiveness and safety of apixaban in patients with and without uncontrolled hypothyroidism. The primary outcome was to evaluate the thrombotic events after apixaban initiation. The secondary outcomes were stroke, VTE and major and minor bleeding events.
Outcomes Definitions
Stroke is verified through medical documentation or brain MRI and computed tomography (CT) scans.
Thrombotic events were identified using the International Classification of Diseases, Tenth Revision, Clinical Modification (ICD10-CM) code (ie, stroke, pulmonary embolism, deep vein thrombosis, splenic infarctions, and basilic vein thrombosis), chart review documentation, and/or radiology findings.
Major bleeding was defined according to the ISTH as clinically overt bleeding associated with a fall in hemoglobin by ≥20 g/L, transfusion of ≥2 Units of packed red blood cells (PRBCs) or whole blood, retroperitoneal or intracranial bleeding, or fatal bleeding. 14
CRNM bleeding was defined according to the ISTH definition as any sign or symptom of bleeding that does not fit the criteria for the ISTH definition of major bleeding but does meet at least one of the following criteria: requiring medical intervention by a healthcare professional, bleeding leading to hospitalization or increased level of care. 15
Statistical Analysis
The baseline characteristics and outcome variables were reported as medians with interquartile ranges (IQR) for continuous data and as numbers, and percentages for categorical data. A comparison of continuous variables between the two groups was conducted using the Mann–Whitney U test, while categorical or binary data were compared using the Chi-square test.
To examine the correlation between uncontrolled hypothyroidism (considered a risk group) and the study outcomes, multivariable logistic regression analysis was utilized, and odds ratios (OR) with 95% confidence intervals (CIs) were reported. The following independent variables were included in the model based on their potential association with the outcomes and demonstrated clinical and statistical significance during the initial analysis: Comorbidities including e.g., chronic kidney disease (CKD), and heart failure (HF), CHA2DS2-VASc, and HAS BLED scores.
To evaluate the appropriateness of the logistic regression model, the Hosmer-Lemeshow goodness-of-fit test was employed, and a non-significant p-value (eg, p ≥ .05) indicates that the model provides a good fit. No imputation was made for missing data, as the cohort of patients in our study was not derived from random selection. A p-value < .05 was considered to be statistically significant. STATA software version 17 BE was used for all statistical analyses.
Sample Size Calculation
Group sample sizes of 54 patients in each arm would achieve 80% power to detect a meaningful difference of 25% between the two groups. A two-sided Z test with pooled variance was used, with a significance level of 5%.16–18‐18
Results
Demographic and Clinical Characteristics
A total of 1433 patients were screened, of whom 292 met the inclusion criteria. Of the included patients, 51 had uncontrolled hypothyroidism, and 241 were assigned to the control group (Figure 1). Most included patients were female (54.9%) with a median age of 70 years in both groups. Among patients with uncontrolled hypothyroidism, 91% were on levothyroxine. Apixaban total maintenance daily doses (MDD) were either 5 mg or 10 mg, with the majority receiving the 10 mg dose among the control (74.4%) and the active (66.7%) groups, with a p-value of .26. Comorbidities between the two groups were balanced except for HF and CKD, which were more prevalent in the uncontrolled hypothyroidism group than those in the control group (62.7% vs 47.1%; p-value = .042), (39.2% vs 22.3%; p-value = .011), respectively, as shown in Table 1.
Eligibility Criteria Flowchart.
Baseline Characteristics.
The median CHA2DS2-VASc score was statistically higher in patients with uncontrolled hypothyroidism than in the control group [5 versus 4; p-value = .029]. The median HAS-BLED score tends to be higher but was not significantly different [3 versus 2; p-value = .069] as shown in Table 1. Other laboratory parameters were similar between the two groups except for TSH, which was higher in the uncontrolled hypothyroidism group (6.59 vs 1.96; p-value <.001). In addition, the baseline platelet counts and blood urea nitrogen (BUN) were higher in patients with uncontrolled hypothyroidism compared to the controlled group [269 versus 242; p-value = .063 and 7.5 versus 6.2; p-value = .038, respectively], as shown in Table 2.
Baseline Laboratory.
^Mann–Whitney U test is used to calculate the p-value.
Thromboembolic Outcomes
In the crude analysis, any thrombosis events were higher in the uncontrolled hypothyroidism group; it was statistically significant (9[17.6%] vs 20[8.4%]: p-value = .04). After adjusting for the potential confounders, the adjusted OR (aOR) was 2.40 [95%CI 0.99-5.83; p-value = .05], revealing a higher odd of any thrombosis. In addition to that, patients with uncontrolled hypothyroidism had higher events of stroke in the crude analysis compared to those in the control group (15.7% vs 4.2%; p-value =.001). The results were consistent with regression analysis [aOR; 4.26, 95%CI 1.51-12.00; p-value = .006]. In contrast, VTE events were not statistically significant between the two groups [aOR 0.96, 95%CI 0.20-4.73; p-value = .96] as shown in Table 3.
Thrombosis and Bleeding Outcomes.
^^Chi-square test is used to calculate the P-value.
Bleeding Outcomes
In the crude analysis, the major bleeding events were higher in patients with uncontrolled hypothyroidism with statistically significant differences [11.8% versus 2.1%; p-value = .001] as well as in regression analysis [aOR 6.21, 95%CI 1.73-22.19; p-value = .005]. Regarding the type of major bleeding, intracranial bleeding was higher in the uncontrolled hypothyroidism group [9.8% versus 0.4%, p-value <.003]. On the other hand, CRNM bleeding was not statistically different in the uncontrolled hypothyroidism group in the crude [21.6% versus 13.8%; p-value = .14] and regression analyses [aOR 1.66, 95%CI 0.76-3.62; p-value = .20] as shown in Table 3.
Discussion
To the best of our knowledge, this is the first real-world study to examine the effectiveness and safety outcomes associated with apixaban use in patients with AF and concomitant uncontrolled hypothyroidism. We found that patients with uncontrolled hypothyroidism had significantly higher odds of developing new thrombotic events, including stroke, compared to patients without hypothyroidism, even after adjusting for relevant confounders such as apixaban dose, comorbidities, platelet count, and CHA2DS2-VASc score. Additionally, major bleeding events were significantly more frequent in the uncontrolled hypothyroidism group, indicating a potential dual risk of thrombosis and bleeding. These findings suggest that uncontrolled hypothyroidism may alter the risk-benefit profile of apixaban in AF patients. Prior studies in this area were limited to post-hoc analyses of randomized trials that did not account for real-time thyroid status or differentiate between controlled and uncontrolled thyroid function.6,19
Limited data exist on apixaban use in uncontrolled hypothyroidism, making it difficult to assess the trade-off between thrombotic and bleeding risk in this population. Thus, it requires relative attention when the literature review is interpreted. In contrast to our findings, the data from the ARISTOTLE trial showed no difference in thrombotic events between hypothyroidism and euthyroidism with apixaban use after covariate adjustment; however, the study includes both controlled and uncontrolled cases of hypothyroidism. 6 It has been suggested that the effect of oral anticoagulants could be reduced with high TSH levels (>10mU/L, indicative of hypothyroidism). 20 Bucerius et al observed low INR values of <2.0 in 76% of hypothyroid cancer patients on warfarin, but these results may be skewed by known increased thrombotic risk associated with malignancy. 20 Moreover, Vasilopoulou et al found an increased risk of AF-related hospitalization (including stroke and bleeding) in patients with hypothyroidism (aHR: 1.77; p = .019). However, when the analysis was limited to stroke, there was no significant difference (aHR: 1.64; p = .227). 19
In terms of bleeding safety profile, our study identified a concerningly higher bleeding risk in uncontrolled hypothyroid patients with AF receiving apixaban, which contradicts the ARISTOTLE trial. 6 Several factors might explain the increased bleeding risk. Firstly, uncontrolled hypothyroidism itself is an independent predictor of a hypercoagulable state, resulting in a higher risk of bleeding. 11 Secondly, hypothyroidism is known to decrease levels of clotting factors like VIII, IX, XI, and von Willebrand factor (VWF), leading to prolonged bleeding times and abnormal clotting tests, prothrombin time (PT) and aPTT.21,22 This is supported by studies demonstrating hemostatic abnormalities in severe hypothyroidism. 23 Additionally, although not pharmacodynamically assessed in our study, hypothyroidism might potentiate the anticoagulant effect of apixaban, further raising bleeding risk. Lastly, our study population with uncontrolled hypothyroidism had a higher proportion of CKD and dialysis dependence in baseline characteristics when compared to the control group, which are independent risk factors for bleeding and might have confounded the results, especially with apixaban use.
The 2023 American College of Cardiology/American Heart Association AF guidelines provide special recommendations for patients with a comorbid thyroid disorder. 5 However, the recommendations addressed were specific to hyperthyroidism, in which anticoagulation for patients with hyperthyroidism and AF who were at risk for stroke was recommended until the normalization of thyroid functions is achieved and sinus rhythm is maintained. The association between hyperthyroidism and AF is well-established compared to hypothyroidism, which has a weaker association. 8
Although this study reports important findings, it should be interpreted while considering the study's limitations. First, the retrospective, single-center design introduced some bias, and therefore a cause-effect relationship cannot be established. Moreover, since there were no planned follow-up visits for the study, it is possible that some outcomes were not captured in patients who sought medical care in other hospitals. Additional multicenter, prospective studies are needed to confirm the results. Second, there were some baseline differences between the two study groups in terms of comorbidities, dialysis, and CHA2DS2VASc score. However, a robust analysis, including multivariable regression analysis was performed to minimize the effect of confounders. Third, despite a clear definition of hypothyroidism used at baseline based on laboratory results within three months of apixaban initiation, follow-up thyroid function tests were not evaluated. Hence, it is possible that some patients achieved euthyroid or even hyperthyroid states with their treatment, which could impact the outcomes. Future prospective studies are needed to evaluate the thyroid and AF-specific outcomes throughout the study period to minimize the risk of bias. Future studies should also consider adding patients with controlled hypothyroidism as a third comparator group to better understand the relationship between hypothyroidism and the coagulopathy risks. Lastly, the study was limited to patients who received apixaban for AF. Therefore, the generalizability of the findings to other anticoagulants or other indications may be limited due to potential variations in clotting and bleeding risks associated.
Conclusion
Our study showed that the use of apixaban in patients with AF and uncontrolled hypothyroidism was linked to higher rates of thrombosis and CRNM bleeding compared to those without known hypothyroidism. These findings highlight the need for further research through larger prospective studies in this often-overlooked population.
Footnotes
Acknowledgments
The authors would like to thank the Deanship of Scientific Research at Shaqra University for supporting this work. In addition, we express our sincere appreciation to all researchers affiliated with the Saudi Critical Care Pharmacy Research (SCAPE) platform and the invaluable support provided by the Saudi Society for Multidisciplinary Research Development and Education. Their assistance was instrumental in completing this project.
Ethics Approval and Consent to Participate
The study was approved in October 2023 by King Abdullah International Medical Research Center Institutional Review Board, Riyadh, Saudi Arabia (Ref.# NRC23R/638/10). Participants’ confidentiality was strictly observed throughout the study by using anonymous unique serial numbers for each subject and restricting data only to the investigators. Informed consent was not required due to the research's method as per the policy of the governmental and local research center.
Competing Interests
No author has a conflict of interest in this study.
Author Contributions
All authors critically revised the manuscript, agreed to be fully accountable for ensuring the integrity and accuracy of the work, and read and approved the final manuscript.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Availability of Data and Material
The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.