Background/Purpose: There are primarily six groups of biologic disease modifiers (BDM) for treatment of rheumatoid arthritis (RA): tumor necrosis factor inhibitors (infliximab (IFX), etanercept (ETN), adalimumab (ADM), certolizumab pegol (CZP) and golimumab (GLM), B-Cell depleting anti CD20 antibodies (rituximab), selective coestimulator blocker (abatacept (ABT), interleukin 6 inhibitor (tocilizumab (TCZ),  interleukin 1 receptor antagonist (anakinra) and Janus associated kinase inhibitor (tofacitinib). BMD are known to interfere in the immune system leading to concerns on their safety. In general, patient/physician preference is what mandates the choice of drug. This is partially due to lack of evidence comparing these drugs to one another (the majority of evidence is based on randomized controlled trials (RCT’s) that are compared to methotrexate (MTX). The objective of this research is to perform a network meta-analysis among these drugs, using the MTX group as a common comparator, to determine their efficacy and safety.

Methods: Studies were extracted from a computerized literature search of MEDLINE and EMBASE of all relevant RCT’s.  41 RCT’s, including 12,487 patients, were identified. There were three outcomes of interest for efficacy: ACR20, ACR50 and ACR70 response. There were several outcomes of interest for safety, primarily serious infections requiring intravenous antibiotics or hospitalization: pneumonia, cellulitis, skin abscess, sepsis, septic arthritis, urinary tract infection, TB and opportunistic infections, sudden cardiac death and death. For each outcome, a fixed-effects meta-analysis was employed to compare each drug to MTX. A mixed-treatment comparisons analysis was then used to compare each of these drugs to one another indirectly. Calculation of the probability that each treatment is best was implemented using the Bayesian Markov chain Monte Carlo method.

Results: In terms of ACR20 response, patients taking IFX (10mg/kg) had a highly statistically significant likelihood of achieving response compared to ADM (RR 0.33, 95% CI 0.11 – 1.00) and a similar benefit was seen in patients taking CZP compared to ADM (RR 0.21, 95% CI 0.06 – 0.69). In terms of ACR50 response, patients taking CZP had a statistically significant increased likelihood of achieving response compared to ABT (RR 0.32, 95% CI 0.11 – 1.04). In terms of ACR70 response, patients taking IFX (10 mg/kg) had a statistically significant higher likelihood of response compared to ABT (RR 0.31, 95% CI 0.09 – 0.90), ADM (RR 0.30, 95% CI 0.09 – 0.82), ETN (RR 0.22, 95% CI 0.06 – 0.70) and GLM (RR 0.29, 95% CI 0.07 – 0.98) and a similar benefit was seen in patients taking TCZ compared to ABT (RR 0.30, 95% CI 0.10 – 0.99), ADM (RR 0.29, 95% CI 0.09 – 0.87) and ETN (RR 0.21, 95% CI 0.06 – 0.80).  Patients taking IFX (3mg/kg) had a statistically significant decreased risk of death compared to ADM (RR 0.09, 95% CI 0.01 – 0.97. The degree of incoherence (measuring how closely the network fits together) was low for all outcomes.

Conclusion: This was the first attempt to include all BDM used in the treatment of RA in a network meta-analysis. Important results on safety and efficacy outcomes were discovered, future research is required to better elucidate these findings.

---

« Back to 2013 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/an-indirect-comparisons-analysis-between-biologic-disease-modifiers-in-the-treatment-of-rheumatoid-arthritis-to-evaluate-for-efficacy-and-safety/