Background/Purpose: Treatment of rheumatoid arthritis (RA) with tocilizumab (TCZ), a humanized monoclonal antibody directed against the IL-6 receptor, is associated with clinical efficacy and marked reduction in systemic inflammatory markers associated with atherogenesis and atherothrombosis, but treatment-associated increases in low-density lipoprotein (LDL-C) have called into question the net cardiovascular risk-benefit ratio of TCZ in RA.

Methods: Seropositive RA patients with active disease (≥8 swollen and ≥8 tender joints and CRP >3 mg/L) and inadequate response to ≥1 nonbiologic DMARD were randomized 1:1 to receive open-label TCZ 8 mg/kg infused monthly or etanercept (ETA) 50 mg injected weekly. Enrollment was restricted to patients ≥50 years of age with ≥1 cardiovascular disease (CVD) risk factor, extra-articular RA manifestations, or history of a CVD event. The primary outcome was the comparison of incident major adverse CVD event (MACE), defined as CVD death, nonfatal myocardial infarction or nonfatal stroke, for TCZ vs ETA. Rates of other CVD events and non-CVD safety outcomes were explored as secondary outcomes. Lipid levels were tracked throughout the trial. CVD events were adjudicated by an independent committee. The study was powered to rule out ≥80% relative hazard of MACE for TCZ vs ETA.

Results: A total of 3080 RA patients were enrolled (TCZ n=1538; ETA n=1542), and 2957 (96%) completed the study with a full assessment of CVD events. Average follow-up time was 3.2 years. Early discontinuation of randomized treatment occurred in 23% of ETA- vs 26% of TCZ-treated patients. Baseline characteristics (mean age 61 years, 22% male, 29% current smokers, 71% with hypertension, 18% with diabetes, mean CRP 19.3 mg/L) were balanced between treatment groups. A total of 161 MACE qualified for inclusion into the primary analysis. In the intention-to-treat analysis, 83 MACE occurred over 4900 PYs in the TCZ arm vs 78 over 4891 PYs in the ETA arm (HR 1.05; 95% CI 0.77, 1.43). HRs for TCZ vs ETA for other secondary CVD outcomes are summarized in the Table. By week 4, total cholesterol, LDL-C, HDL-C, and triglycerides increased significantly in the TCZ arm compared with the ETA arm, with median LDL-C increasing 12% for TCZ vs 1% for ETA. Thereafter, average levels remained consistent to trial conclusion. The overall rate of adverse events, serious infections, and medically confirmed gastrointestinal perforations was numerically higher for TCZ vs ETA.

Conclusion: This comparative study of TCZ with ETA excluded a >43% relative increase in the hazard of MACE (HR 1.05; 95% CI 0.77, 1.43), with an estimated 5% increase in TCZ compared with ETA among RA patients with severe active disease and elevated baseline CVD risk. Average treatment-associated increases in LDL-C were higher for TCZ vs ETA. The CVD safety of TCZ relative to ETA should be interpreted in the context of its non-CVD safety and clinical efficacy.