Higher rheumatoid arthritis disease activity at study entry was associated with a greater risk of a first major adverse cardiovascular event, but the strength of that association differed according to patients’ combined rheumatoid factor and anticitrullinated protein antibody status.
Researchers analyzed data from 3,952 patients with prevalent rheumatoid arthritis and no established or suspected cardiovascular disease who were enrolled at 13 academic referral centers across 10 countries through the international ATACC-RA consortium. Enrollment began in 1985 and extended through 2012 or 2013, according to different sections of the report.
Patients were followed from enrollment until a first major adverse cardiovascular event, death, migration, or censoring. The primary outcome comprised nonfatal myocardial infarction, nonfatal stroke, or cardiovascular death.
Disease activity was measured at enrollment using the Disease Activity Score in 28 joints with C-reactive protein, or DAS28-CRP. In adjusted analyses, each 1-unit increase in baseline DAS28-CRP was associated with an 18% higher hazard of a first major adverse cardiovascular event. The models accounted for demographic characteristics, traditional cardiovascular risk factors, rheumatoid arthritis duration, and corticosteroid use.
During 22,981 patient-years of follow-up, 184 patients experienced a first major adverse cardiovascular event, including 102 myocardial infarctions, 68 strokes, and 14 cardiovascular deaths.
Rheumatoid factor (RF) and anticitrullinated protein antibody (ACPA) positivity did not reach conventional statistical significance as independent predictors when both markers were modeled alongside disease activity and cardiovascular risk factors. However, the estimates were imprecise, and the researchers cautioned that clinically meaningful independent associations could not be excluded.
The researchers identified an interaction among disease activity, RF status, and ACPA status. Higher DAS28-CRP scores were associated with cardiovascular events among patients who were negative for both antibodies and among those who were positive for both. Associations in the 2 single-positive groups did not reach statistical significance.
Those nonsignificant findings should be interpreted cautiously. The RF-negative, ACPA-positive group included 306 patients and experienced 14 cardiovascular events. The RF-positive, ACPA-negative group included 557 patients and experienced 28 events, resulting in wide and imprecise estimates.
Sensitivity analyses using separate RF- and ACPA-weighted Cox models produced a similar pattern. The results also remained consistent in a competing-risk analysis that treated 79 noncardiovascular deaths as competing events.
The findings suggested that combined serostatus may help explain why the association between current inflammatory activity and cardiovascular outcomes differs among patients with rheumatoid arthritis. The researchers advised against interpreting low disease activity as evidence of low cardiovascular risk or using it to justify de-escalation of established cardiovascular prevention. They also proposed that imaging-based risk reclassification might be useful in selected serologic groups, although those approaches were not evaluated in the study and require prospective confirmation.
The study had several limitations. Disease activity, cardiovascular risk factors, and medication use were assessed only at enrollment, preventing the researchers from accounting for changes during follow-up. Baseline DAS28-CRP data were available for 2,814 patients, and missing data were addressed using multiple imputation.
The researchers also lacked information on RF and ACPA titers, isotypes, and specificities. Cardiovascular events were not centrally adjudicated, surveillance may have varied among centers, and recruitment at academic centers with an interest in rheumatoid arthritis–related cardiovascular disease may have introduced referral bias. Residual confounding, survival bias, and left-truncation bias also could not be excluded.
Because the analysis was observational, it could not establish that disease activity caused the cardiovascular events or that lowering disease activity would confer different cardiovascular benefits according to serostatus.
Disclosures: The study was supported by an investigator-initiated grant from Pfizer to Karpouzas. Pfizer was not involved in the study design, procedures, data collection, analysis, interpretation, manuscript preparation, or submission. The researchers declared no competing interests.
Source: RMD Open