PDF
Citation
Print
INTRODUCTION
Rheumatoid arthritis (RA) is characterized as a persistent, systemic inflammatory disorder mediated by the immune system; it predominantly impacts the joints and frequently presents with extra-articular manifestations [1,2]. Patients with RA have a 38% increased risk of acute myocardial infarction (AMI) compared with the general population, which represents the leading cause of death in RA [3,4]. Systemic inflammation can accelerate the atherosclerotic process and is a key mechanism in the pathogenesis of cardiovascular disease in patients with RA [5,6]. Numerous studies have shown that high RA activity increases the risk of major adverse cardiovascular events (MACEs), suggesting that effective management of RA activity could potentially reduce this risk [7,8].
Glucocorticoids and nonsteroidal anti-inflammatory drugs (NSAIDs) have been linked to an increased risk of cardiovascular complications; however, the cardiovascular effects of anti-rheumatic medications—including conventional synthetic disease-modifying anti-rheumatic drugs (csDMARDs) and biological DMARDs (bDMARDs)—are currently under debate [9]. The complexity in evaluating the cardiovascular impact of these drugs arises from their diverse mechanisms and the combined use of various medications in RA therapy; these drugs affect cardiovascular health either directly (through endothelial damage) or indirectly (by influencing insulin resistance, lipid metabolism, blood pressure, and thrombotic pathways) [9]. A comprehensive understanding of both inflammation and atherosclerotic risk factors is therefore essential for assessing the cardiovascular implications of anti-rheumatic agents in RA treatment.
This study therefore aimed to examine the association between the use of different anti-rheumatic agents and the risk of AMI in patients with RA using a nationally representative cohort of medical claims data in Korea.
MATERIALS AND METHODS
Data source
This study is a retrospective, nationwide, population-based analysis using data from the Korean Health Insurance Review and Assessment Service (HIRA) claims database. The HIRA database comprehensively encompasses the health information of approximately 50 million individuals under South Korea’s National Health Insurance (NHI) program [10]. This database provides details on patient demographics (such as age at 5-year intervals and sex), diagnostic codes, medical and surgical procedures, and medication prescriptions [11].
This study was conducted following the ethical guidelines of the Declaration of Helsinki. Due to the retrospective nature of the study, the Asan Medical Center Institutional Review Board (IRB) waived the need for informed consent (IRB no. 2021-1365).
Study population
The diagnosis of RA was based on the International Classification of Disease 10th revision (ICD-10, adapted for the Korean healthcare framework) and the Korean Rare Intractable Disease (RID) registration code for RA. In the Korean RID system, RA is confirmed based on the 2010 American College of Rheumatology/European League Against Rheumatology classification criteria for RA, along with a positive test for rheumatoid factor or anti-citrullinated peptide antibodies [12].
The RA cohort comprised patients who were newly diagnosed with seropositive RA (ICD-10 code M05; RID code V223) between January 2011 and December 2020. We defined “newly diagnosed RA” as patients assigned the RA diagnostic code for the first time and with no record of anti-rheumatic drug prescriptions from January 2010 until the date of RA code assignment. Based on this operational definition of RA, we included patients who had at least 2 times outpatient clinic visits, received prescriptions for anti-rheumatic agents, and were registered in the Korean RID system. Patients with a history of AMI or stroke prior to the diagnosis of RA were excluded, and the risk observation window was set from a minimum of 12 months to a maximum of 120 months. The study focused on individuals aged ≥20. To reduce potential confounding factors in assessing RA treatment impact, we excluded conditions like seronegative RA, other rheumatic diseases, interstitial lung disease, cancer, human immunodeficiency virus infection, solid organ transplantation, and end-stage renal disease requiring dialysis (Supplementary Table 1). We excluded individuals with an observation period of <90 days.
Cases and controls
In this study, patients with AMI were defined as patients with RA who were hospitalized with a new diagnosis of AMI, identified by ICD-10 code I21 and AMI-related procedures (Supplementary Table 2). For the control group, we used a greedy algorithm to match AMI patients at a 1:4 ratio with controls from the RA cohort. The matching criteria were age, sex, and the RA diagnosis year. The index date for cases was the first AMI diagnosis date; for controls, the index date was aligned with their matched case.
Assessment of exposure to anti-rheumatic agents
We determined exposure to anti-rheumatic agents through prescription records, including csDMARDs (such as hydroxychloroquine, methotrexate, leflunomide, sulfasalazine, and tacrolimus), bDMARDs (such as anti-tumor necrosis factor-alpha [TNF-α] agents [etanercept, infliximab, adalimumab, and golimumab], abatacept, and tocilizumab), Janus kinase (JAK) inhibitors (such as tofacitinib, baricitinib, upadacitinib), and glucocorticoids.
To evaluate drug exposure, the period from the diagnosis of RA to the AMI event was examined, with retrospective data collection extending back to 2010 for individuals diagnosed in 2011. Categories were formed based on the duration of exposure to each anti-rheumatic agent and the exposure risk window. Initially, the exposure risk window was divided into current exposure, defined as a prescription within 3 months of the AMI event, and past exposure, where the prescription had ceased >3 months before the AMI event. Further, exposure was classified into two durations based on a 12-month period: short-term (<12 months) and long-term (≥12 months). Finally, a subgroup analysis was conducted, organizing patients into five distinct groups: non-exposure, past short-term exposure, past long-term exposure, current short-term exposure, and current long-term exposure.
Confounding variables
In our analysis of potential AMI determinants, we analyzed data covering the period 1 year before the index date. This data encompassed various parameters, including age at the time of RA diagnosis, sex, and a range of comorbidities (including hypertension, dyslipidemia, diabetes mellitus, carotid stenosis, asthma, chronic obstructive pulmonary disease, chronic kidney disease, congestive heart failure, and atrial fibrillation). We recorded medication history; specifically, the use of low-dose aspirin, NSAIDs, statins, and antihypertensive drugs. For an accurate evaluation of these comorbidities, the Charlson Comorbidity Index (CCI) was used [13].
Statistical analysis
We analyzed participant characteristics, including age groups, comorbidities, and medications. Categorical variables were represented as counts and percentages, and continuous variables as means with standard deviations. The standardized mean difference (SMD) was employed to quantify the balance of baseline characteristics between AMI patients and matched controls. An SMD <0.1 signified a reasonable balance of confounders between groups. To examine the specific impact of each anti-rheumatic agent on AMI risk, we used conditional logistic regression models. These models compared the risk of new-onset AMI for each anti-rheumatic agent against non-exposure. Adjustments were made for comorbidities, CCI, NSAIDs, low-dose aspirin, statins, antihypertensive agent use, and other anti-rheumatic agents. The outcomes were represented as adjusted odds ratios (aORs) with 95% confidence intervals (CIs). Statistical significance was set at a p-value <0.05. All statistical evaluations were conducted using SAS Enterprise Guide software, version 7.1 (SAS Institute, Cary, NC, USA).
RESULTS
Study population and baseline characteristics
A total of 35,133 patients newly diagnosed with RA between 2011 and 2020 were identified using Korean HIRA claims data. Among these individuals, 484 (1.4%) were diagnosed with AMI. These patients were matched to 1,924 control patients without AMI at a 1:4 ratio based on age, sex, and year of diagnosis (Figure 1).
Table 1 displays the baseline demographics of the participants; in both cohorts, the predominant age range was 60~69 years, with females comprising 65.6% of patients. Those who suffered an AMI had a higher prevalence of conditions such as hypertension, dyslipidemia, diabetes mellitus, and chronic kidney disease and were more likely to have elevated CCI scores than those in the non-AMI group.
Table 2 outlines the relationship between the current use of specific anti-rheumatic agents and the incidence of AMI; non-exposure or past exposure to these drugs served as a reference. Patients currently treated with glucocorticoids (aOR: 2.301, 95% CI: 1.741~3.041) and leflunomide (aOR: 1.525, 95% CI: 1.196~1.944) exhibited an elevated AMI risk compared with those with non-exposure or past exposure.
Table 3 describes the association between long-term exposure to each anti-rheumatic agent and AMI, with non-exposure and short-term exposure to each anti-rheumatic agent as the references. Notably, groups with long-term exposure to glucocorticoids (aOR: 1.792, 95% CI: 1.378~2.330) and leflunomide (aOR: 1.740, 95% CI: 1.327~2.207) demonstrated higher risks of AMI than those in the non-exposure and short-term exposure groups.
Supplementary Table 3 describes the results of the subgroup analysis, delineating the AMI risk in relation to the duration and exposure level of various anti-rheumatic agents. All groups with glucocorticoid exposure had higher risks of AMI than those in the non-exposure group ([past short-term exposure] aOR: 7.065, 95% CI: 2.341~21.318; [past long-term exposure] aOR: 9.115, 95% CI: 2.964~28.035; [current short-term exposure] aOR: 4.694, 95% CI: 1.315~16.751; [current long-term exposure] aOR: 12.704, 95% CI: 4.146~38.926). Current long-term (aOR: 1.736, 95% CI: 1.312~2.298) and past long-term (aOR: 1.566, 95% CI: 1.049~2.338) exposure to leflunomide, and current long-term exposure to tacrolimus (aOR: 1.781, 95% CI: 1.143~2.773) were significantly associated with an increased risk of AMI compared with those in the non-exposure group. Conversely, past short-term (aOR: 0.450, 95% CI: 0.281~0.722) and past long-term (aOR: 0.611, 95% CI: 0.423~0.884) exposure to hydroxychloroquine were significantly associated with a decreased risk of AMI compared with that in the non-exposure group.
DISCUSSION
In this study, we observed that 1.4% of patients newly diagnosed with RA in Korea developed AMI. A similar study analyzing claims data in Taiwan reported that AMI occurred in 1.1% of patients with RA, demonstrating concordance with our findings [14]. This population-based nested case-control study revealed a markedly increased risk of AMI associated with current and long-term exposure to glucocorticoids and leflunomide.
Glucocorticoids are known for their potent anti-inflammatory and immunosuppressive properties, which can be beneficial for controlling the symptoms of RA. However, they have negative effects on glucose metabolism, blood pressure, and cholesterol metabolism, which affect the occurrence of atherosclerosis [9]. According to the results of a meta-analysis, patients taking glucocorticoids demonstrated a higher risk of MACEs (rate ratio: 1.27, 95% CI: 1.15~1.40) [15]. Additionally, glucocorticoid use increases the risk of cardiovascular disease in patients with RA in a dose- and duration-dependent manner [16,17]. Similar trends were observed in our study, where all categories of glucocorticoid use—including past short-term, past long-term, current short-term, and current long-term—were associated with an elevated risk of AMI in patients with RA compared with patients with glucocorticoid non-exposure. The highest aOR for AMI, 12.704 (95% CI: 4.146~38.926), was observed in patients undergoing current long-term glucocorticoid therapy. This underscores the dual role of glucocorticoids, where their potent anti-inflammatory properties help control RA disease activity, while prolonged use may be linked to an elevated risk of AMI. These findings emphasize the complexity of RA management, requiring careful consideration of treatment strategies that balance effective disease control with the potential for increased cardiovascular risk. Specifically, these results are consistent with existing literature that links chronic glucocorticoid use to heightened cardiovascular risk in patients with RA [16,17]. Furthermore, this highlights the necessity for further research to delineate the direct cardiovascular effects of glucocorticoids from those associated with uncontrolled RA disease activity, which often requires glucocorticoid therapy.
The current study demonstrated that current exposure and long-term exposure to leflunomide was associated with an increased risk of AMI in patients with RA. Leflunomide is an isoxazole derivative that exhibits anti-inflammatory effects by inhibiting pyrimidine synthesis [18]. Leflunomide has been linked to an increased risk of hypertension, making it unsuitable as a first-line anti-rheumatic agent in patients with hypertension or cardiovascular diseases [9]. Contrary to the findings of our study, another study reported a significant reduction in AMI risk among patients with RA treated with leflunomide (adjusted rate ratio: 0.28) [19]. This discrepancy highlights the need for further investigation to reconcile these opposing outcomes and understand the full cardiovascular impact of leflunomide in patients with RA.
Tacrolimus, a calcineurin inhibitor, is an anti-rheumatic agent that can be used in methotrexate-resistant patients with RA, according to the Asia Pacific League of Associations for Rheumatology [20]. Tacrolimus reportedly increases the risk of abnormal electrocardiogram rhythm and diabetes mellitus in patients with RA [21,22]. Additionally, a multicenter prospective investigation reported that tacrolimus can have adverse cardiovascular effects (hypertrophic cardiomyopathy and sinus tachycardia) in patients who have undergone kidney transplantation [23]. In the current study, current long-term exposure to tacrolimus was associated with an increased risk of AMI (aOR: 1.781, 95% CI: 1.143~2.773). Since tacrolimus is currently approved and used for RA treatment in only a few countries—including Korea, Japan, and Canada—there is limited evidence of this association [24]. Therefore, further research on the potential link between tacrolimus and AMI is needed.
Hydroxychloroquine has been shown to mitigate atherosclerosis risk by influencing factors such as lipid and glucose levels, coagulation, and endothelial functions in individuals with RA and systemic lupus erythematosus [25]. Despite these benefits, potential risks—including heart failure and arrhythmias when using hydroxychloroquine—have been highlighted [26]. Recent analyses of Medicare data indicate that in patients with a prior history of heart failure, the incidence of MACEs is greater with hydroxychloroquine than with methotrexate [27]. Contrarily, another retrospective observational cohort study of patients with RA treated with hydroxychloroquine revealed a reduced occurrence of AMI, with a hazard ratio (HR) of 0.9 (95% CI: 0.85~0.96) [25]. Consistent with these findings, our study demonstrates a significant association between past hydroxychloroquine use and a lower risk of AMI.
In our study, anti-TNF-α agents did not significantly reduce the risk of AMI. Previous reports have suggested that anti-TNF-α agents are associated with a lower risk of cardiovascular events (HR: 0.71, 95% CI: 0.52~0.97) compared with csDMARDs [28]. Anti-TNF-α agents exert a cardioprotective effect by enhancing cholesterol transport, improving glucose metabolism, downregulating adhesion molecules, and counteracting inflammatory effects on coagulation [9]. Recent findings from the Oral Rheumatoid Arthritis Trial Surveillance indicated that tofacitinib was associated with an increased risk of MACEs compared with anti-TNF-α agents in patients with RA aged ≥50 years with cardiovascular risk [29]. In our study, the use of JAK inhibitors did not influence the risk of AMI. However, this observation is limited by the small number of participants (only three) who used JAK inhibitors before their AMI event, restricting our ability to draw comprehensive conclusions about the link between JAK inhibitors and cardiovascular risk. Consequently, further research involving a larger cohort of patients treated with JAK inhibitors is warranted to elucidate this relationship.
The current study had certain limitations. First, we could not analyze social risk factors—including smoking and alcohol consumption—associated with AMI due to the limited clinical information available in the claims database. Moreover, the absence of data on RA disease activity, which is a key determinant in treatment decisions and cardiovascular risk, may have overestimated the effects of DMARDs on AMI risk. Furthermore, the absence of detailed information regarding medication adherence precluded us from conducting an in-depth analysis. Second, the occurrence of an AMI was identified based on registered diagnostic and procedure codes, which may not fully capture the severity of the AMI. In addition, the study lacked data on the interval between RA diagnosis and AMI occurrence, potentially omitting the impact of disease duration on cardiovascular risk. Third, channeling bias may have been a relevant factor, as patients with severe RA may have current and long-term use of glucocorticoid therapy and thus have had a higher baseline risk of developing AMI. Additionally, this study did not analyze AMI risk based on glucocorticoid dosage; therefore, further research on the impact of high-dose glucocorticoids on the risk of AMI is warranted. Fourth, subgroup analysis of abatacept, tocilizumab, and JAK inhibitors was unavailable owing to the relatively small number of patients with AMI. Fifth, this study predominantly included patients aged 60 and above with multiple cardiovascular risk factors; therefore, generalizability to younger patients or those with fewer risk factors may be limited. Nevertheless, the major strength of this study lies in its real-world design, which utilizes a nationwide claims database involving a large population to assess the effects of various anti-rheumatic agents on the risk of AMI.
CONCLUSION
This study demonstrated that current long-term exposure to glucocorticoids, leflunomide, and tacrolimus was associated with an increased risk of incident AMI. The effect of anti-rheumatic agents on the development of incident AMI was more pronounced with a longer treatment duration (>1 year) and current exposure. Clinicians should therefore be aware of the increased risk of AMI in patients with RA and carefully consider the cardiovascular implications when selecting anti-rheumatic agents.
SUPPLEMENTARY DATA
Supplementary data can be found with this article online at https://doi.org/10.4078/jrd.2024.0104
XML Download