To the Editor:

Patients with rheumatoid arthritis (RA) have a 1.5-times excess risk of cardiovascular (CV) disease compared to the general population, attributed to chronic inflammation.1,2 In the general population, detectable levels of high-sensitivity cardiac troponin (hs-cTn) are associated with higher risk of major adverse CV events (MACE) and all-cause mortality.3 A study in RA reported an association between higher levels of hs-cTn I with occult coronary atherosclerotic plaque burden and risk of incident CV events, with an average follow-up of 60 months.4,5 The objective of this study was to test the longer-term association between clinically established thresholds for detectable hs-cTN T (hs-cTnT) with MACE and all-cause mortality in RA.

We studied 331 patients in a longitudinal observational RA cohort study, the Brigham and Women’s Hospital Rheumatoid Arthritis Sequential Study (BRASS). In BRASS, subjects are assessed every 6 months for RA disease activity, and blood samples are collected yearly; high-sensitivity C-reactive protein (hsCRP) is also measured annually. As this study was focused on identifying subclinical risk in patients not on standard primary prevention CV therapies, we selected the earliest available blood sample for each subject and excluded those without samples or who were on a statin or lipid-lowering therapy in the year prior or up to the date of blood sample collection, as defined previously.6 The timepoint at which the earliest blood sample was collected was considered the baseline (T0) for subjects in this study. Hs-cTnT, a marker of subclinical myocardial injury, and routine lipids were measured in all subjects at T0. These data, along with additional data extracted from medical record review, were used to estimate the 10-year atherosclerotic CV disease (ASCVD) risk at T0.7 Outcomes for the primary outcome MACE (acute coronary syndrome, stroke, and CV death) and secondary outcome all-cause mortality were obtained using medical record review from linked electronic health record (EHR) data (2 independent reviewers, MJ and cardiologist BW). Patients were followed from baseline until MACE, death, last EHR encounter, or 10 years from T0, whichever occurred first. The associations between detectable hs-cTnT with MACE and all-cause mortality were determined using Cox proportional hazards; models were further adjusted by baseline ASCVD risk score and hsCRP.

The mean age of the cohort was 57.9 years; 83.1% were female, 65.9% (193/293) were anti-CCP positive, and median RA disease duration was 12.0 years (Supplementary Table S1, available with the online version of this article). RA treatments at baseline included 55.9% methotrexate, 42.3% tumor necrosis factor inhibitor (TNFi), and 26% oral glucocorticoids. The prevalence of CV risk factors was overall low, with 22.7% hypertension, 4.8% diabetes mellitus, 9.1% hyperlipidemia, and 7.9% active smoking. The median hsCRP at baseline was 4.35 (IQR 1.09-18.29) and the median Disease Activity Score in 28 joints based on CRP (DAS28-CRP) was 3.32 (moderate disease activity). The median calculated 10-year ASCVD risk was 3.87% (IQR 1.18-9.87). A total of 117 (35.3%) had detectable hs-TnT, with a median level of 8.98 mg/dL (7.49-12.96). A total of 16 MACE (4.8%) events occurred in 10 years with 50 all-cause deaths (15.1%). Detectable hs-cTnT was associated with future MACE (hazard ratio [HR] 7.13, 95% CI 2.29-22.10; Figure), which remained significant after adjusting for ASCVD risk and log hsCRP (HR 4.29, 95% CI 1.31-14.10, P = 0.02), as well as baseline ASCVD risk and DAS28-CRP (HR 5.79, 95% CI 1.49-22.40; Table). Detectable hs-cTnT was further associated with all-cause mortality (HR 7.2, 95% CI 4.85-10.90) including after adjusting for baseline ASCVD risk and log hsCRP (HR 4.18, 95% CI 2.60-6.72), and baseline ASCVD risk and DAS28-CRP (HR 4.74, 95% CI 2.20-10.20; Supplementary Table S2). The ASCVD risk score alone was significantly associated with MACE (HR 1.06, 95% CI 1.02-1.10).

Association of detectable hs-cTnT with future MACE. Kaplan-Meier curve demonstrating the association of detectable and undetectable hs-cTnT with the primary composite outcome of MACE (acute coronary syndrome, ischemic stroke, CV mortality) among individuals with RA without known atherosclerotic CV disease. CV: cardiovascular; hs-cTnT: high-sensitivity cardiac troponin T; hsTNT: high-sensitivity troponin T; MACE: major adverse cardiovascular event; RA: rheumatoid arthritis.

Associations between baseline hs-cTnT and MACE at 10 years.

In this RA cohort with overall low estimated ASCVD risk, detectable hs-cTnT was associated with future MACE independent of traditional CV risk factors, levels of systemic inflammation (hsCRP), and RA disease activity (DAS28-CRP). Hs-cTnT was further associated with all-cause mortality. Although this is one of the largest studies of patients with RA with hs-cTnT and MACE outcomes, the relatively low event rate observed may be explained in part by patients who receive primary care or present with MACE outside of the hospital system. The low event rate may also reflect contemporary data suggesting a decrease in CV events among patients with RA,8 a trend observed in the general population9 secondary to preventive measures and public health initiatives. However, the increased CV mortality observed in RA compared to the general population remains.10 More recently, there has been an increase in the prevalence of cardiometabolic risk factors, with steep rises projected moving forward—a trend predicted to similarly occur in RA.11 These findings are also consistent with prior work in RA identifying an association between higher hsTnI levels with incident CV events.5 Our current study builds on the prior findings by testing the association between now-established clinical cut-offs for detectable hs-cTnT with incident MACE and all-cause mortality, providing information on the clinical relevance of a detectable hs-cTnT in patients with RA not identified for primary prevention. Additionally, our study provides a longer follow-up period. The sample size did not allow analysis of the effect of specific RA treatments on MACE. From a clinical standpoint, the identification of patients with RA who have a higher CV risk, prompting initiation of CV prevention medications (ie, statins), remains a challenge. Future studies are needed to examine the association of hs-cTnT and existing atherosclerotic plaque among asymptomatic patients with RA to determine the utility of screening specific populations using hs-cTnT in the clinic. Overall, these findings suggest that hs-cTnT may be a useful marker to improve CV risk assessment among patients with RA with overall low estimated ASCVD risk.

KPL received funding from the Gordon and Llura Gund Foundation, the National Institutes of Health (NIH R01 HL127118, P30 AR072577), the Harold and DuVal Bowen Fund, and the BeBrave Fund. BNW is funded in part by the NIH National Heart, Lung, and Blood Institute (NHLBI K23 HL159276-01) and the American Heart Association (AHA 21CDA851511). GCM is funded in part by a Rheumatology Research Foundation Scientist Development Award.

MEW reports grants from BMS, Eli Lilly, Aqtual, AbbVie, and Janssen; personal consulting fees from AbbVie, Aclaris, Amgen, Aqtual, Bayer, BMS, CorEvitas, Eli Lilly, EQRx, Genosco, GSK, Gilead, Horizon, J&J, Kyvrena, Pfizer, Rani, Revolo, Sanofi, Scipher, SciRhom, Set Point, and Tremeau, outside the submitted work. MDC reports institutional research grants from Gilead and Amgen; and personal consulting fees from MedTrace, outside the submitted work. BNW reports consulting (scientific advisory board roles) for Kinsika, NovoNordisk, BMS, and Horizon, outside of the submitted work. KPL received a 1-time consulting fee from UCB.