BackgroundCholesterol loading capacity (CLC) describes the ability of serum to deliver cholesterol to cells. It is linked to foam cell formation, a pivotal step in atherosclerotic plaque development. Rheumatoid arthritis (RA) serum promoted foam cell formation significantly more than control serum. Likewise, RA patients display greater plaque burden and higher-risk features than non-RA controls. bDMARDs and statins lower cardiovascular risk by reducing new coronary plaque formation, promoting regression, altering the composition and stabilizing prevalent atherosclerotic lesions.ObjectivesTo evaluate the associations between CLC, coronary plaque burden and cardiovascular event risk in patients with RA. We further explored the conditioning effects of RA treatments on these relationships.Methods140 patients underwent coronary CT angiography for atherosclerosis evaluation and were prospectively followed for cardiovascular events over 6.0±2.4 years. Coronary artery calcium score (CAC), number of segments with plaque (segment involvement score [SIS]) and plaque composition were assessed. CLC was the macrophage cholesterol content, measured by fluorometric assay, after a 24-hour incubation with whole serum. Robust linear regression examined the effects of CLC and the interaction between CLC and bDMARD use on SIS and CAC. Negative binomial regression evaluated CLC and CLC × bDMARD interaction effects on number of high-risk (low-attenuation) plaques. With data discretized into 1-month intervals, weighted pooled logistic regression models with robust variance estimation evaluated CLC and time-varying bDMARD use as predictors of event risk, and the effect of CLC × time-varying bDMARD use on risk. Stabilized inverse probability of treatment and censoring weights were estimated as a function of ASCVD risk, SIS, RA duration, and baseline and time-varying CRP and statin use.ResultsMean (SD) CLC was 12.67 (2.83) μg/mg protein. In analyses adjusting for ASCVD score, HDL, prednisone and statin use, CLC (per 1-SD unit) was not related to SIS (β -0.05 [95%CI -1.19,0.09]), number of high-risk plaques (rate ratio [RR] 1.20 [95%CI 0.80-1.80]) or ln-transformed CAC (β 0.017 [95%CI -0.133,0.147]). However, in analyses stratified by baseline bDMARD use, CLC (per 1-SD unit) was positively related to number of high-risk plaques (RR 2.14 [95%CI 1.04-4.40]) and ln-transformed CAC (β 0.21 [95%CI 0.01-0.41]) among bDMARD-naïve individuals (Figure 1). In addition, CLC inversely associated with SIS (per SD increment; β -0.16 [95%CI -0.32, -0.01]) only in bDMARD-treated patients. Baseline statin use did not significantly modify the effect of CLC on coronary plaque (not shown). CLC associated with cardiovascular event risk (per SD increment; adjusted odds ratio 2.02 [95%CI 1.27-3.50], p=0.011) covarying for ASCVD score and time-varying bDMARD use. The CLC × time-varying bDMARD use interaction also predicted event risk (p =0.010); current bDMARD use associated with lower event risk at higher (1 SD above the mean) CLC levels (p=0.037) but not average or lower (1 SD below the mean) CLC levels (p=0.064 and 0.756, respectively).ConclusionCLC associated with greater CAC score and high-risk plaque burden in bDMARD-naïve RA patients and lower total plaque burden in bDMARD-treated patients at baseline. CLC also predicted long-term cardiovascular risk and its effect was mitigated by bDMARD use.Disclosure of InterestsGeorge Karpouzas Speakers bureau: Sanofi-Genzyme-Regeneron, Janssen, Bristol-Meyer-Squibb, Consultant of: Sanofi-Genzyme-Regeneron, Janssen, Bristol-Meyer-Squibb, Grant/research support from: Pfizer, Bianca Papotti: None declared, Sarah Ormseth: None declared, Marcella Palumbo: None declared, Elizabeth Hernandez: None declared, Cinzia Marchi: None declared, Francesca Zimetti: None declared, Matthew Budoff Consultant of: Pfizer, Nicoletta Ronda: None declared
Read full abstract