Examining the potential direct cardiovascular benefit of tumor‐necrosis factor inhibitor in rheumatoid arthritis: Natural and controlled direct effect analyses

Abstract

Background/Purpose: Tumor necrosis factor inhibitors (TNFi) may have a direct benefit on cardiovascular (CV) disease beyond reducing rheumatoid arthritis (RA) disease activity measured by the Clinical Disease Activity Index (CDAI). Methods:We compared TNFi initiators and methotrexate (MTX) monotherapy initiators from the CorEvitas RA registry. Two approaches to the "direct effect" of TNFi beyond CDAI were used. In the natural direct effect (NDE) analysis, the potential CV benefit of TNFi was partitioned into NDE and the natural indirect effect (NIE) mediated by CDAI during the first 6 months. We also estimated the controlled direct effects (CDE), corresponding to the direct benefit of TNFi when CDAI trajectories were hypothetically equalized between the initiators of TNFi and MTX monotherapy at a constant value. Estimates were given on the hazard ratio scale. Results:We identified 5764 initiators of TNFi and 3588 initiators of MTX monotherapy. TNFi initiators were younger (58 vs. 64 years) with a shorter disease duration.Our total effect estimates (TNFi vs. MTX [reference]) were protective in direction (0.76-0.91). The NDE estimate was 0.76 [95% confidence interval (CI) 0.59, 0.98], whereas the NIE estimate was 1.00 [95%CI 1.00, 1.00]. In the CDE analyses accounting for longitudinal CDAI, the CDE estimates was 1.27 [95%CI 0.60, 2.69]. Conclusions:We could not convincingly demonstrate a direct benefit of TNFi outside its impact on CDAI. At present, the emphasis should be on the stringent control of RA disease activity, a known important CV risk factor, regardless of medication choice.

Reducing cardiovascular risk in immune‐mediated inflammatory diseases: Tumour necrosis factor inhibitors compared to conventional therapies—A systematic review and meta‐analysis

“…The systematic search resulted in a total of 8724 hits. After the title, abstract and full‐text selection, 54 articles matched inclusion criteria 35–88 and two additional articles 89,90 were found to be eligible during citation search. Two articles had corrections (errata) 91,92 that are referenced in the results in addition to the citations of the original articles 39,48…”

“…Altogether 56 studies were included, and 29 articles were involved in the quantitative analyses, 35,36,38,43,44,46,48–53,56,59,60,62,64,67,68,70,71,73,75,76,82,83,86,88,89,92 whereas findings of 27 studies were evaluated qualitatively in the systemic review 37,39–42,45,47,54,55,57,58,61,63,65,66,69,72,74,77–81,84,85,87,90,91 . The selection and screening process is shown in Figure 1.…”

“…Two articles had corrections (errata) 91,92 that are referenced in the results in addition to the citations of the original articles. 39,48 Altogether 56 studies were included, and 29 articles were involved in the quantitative analyses, 35,36,38,43,44,46,[48][49][50][51][52][53]56,59,60,62,64,67,68,70,71,73,75,76,82,83,86,88,89,92 whereas findings of 27 studies were evaluated qualitatively in the systemic review. 37,[39][40][41][42]45,47,54,55,57,58,61,63,65,66,69,…”

“…Seven studies including 84,320 patients examined the risk of MACE, using multivariate HRs. 49,50,67,75,76,82,88 One of the analysed articles included Pso and PsA patients, 82 five studies examined patients with RA, 49,50,67,75,88 and one study included Pso, PsA and RA patients as well. 76 The overall effect size showed that the TNFi-treated group had 26% lower risk compared to the CSNB group (HR 0.74, 95% CI 0.58-0.95), reaching statistical significance (p = 0.025).…”

“…Fourteen studies were analysed for risk of MACE comparing the IRR of 79,909 TNFi-treated to 112,768 CSNB-treated patients. 36,43,44,[48][49][50]56,67,70,71,76,83,88,89,92 Seven additional studies were assessed compared to results with multivariate HRs. 36,43,48,56,70,83,89,92 Four of the seven studies included Pso or PsA patients, 36,48,70,83,92 whereas two articles investigated patients with RA.…”