2123: Hierarchical Ranking of Biologic Disease-Modifying Antirheumatic Drugs and Targeted Systemic Therapies for Psoriatic Arthritis: A Network Meta-analysis

Jesus Diaz¹, Guillermo Quiceno² and Adela Castro³, ¹Hamilton Medical Center, Chattanooga, TN, ²UT Southwestern Medical Center, Dallas, TX, ³Hamilton Physician Group-Specialty Care, Dalton, GA

Background/Purpose: There are an increasing number of biological and targeted DMARDs in PsA. However, challenges around treatment selection remain, given the few head-to-head studies directly comparing their clinical efficacy (1)(2). In such situations, network meta-analysis (NMA) can be useful. This study aims to compare the efficacy of biologics and targeted synthetic therapies for patients with active PsA at the end of the induction period (12–24 weeks) and provide a ranking of these treatments based on their performance in skin and joints.

Methods: A systematic literature review was conducted until April 2022. Only studies reporting American College of Rheumatology 20% (ACR 20) response and Psoriasis Area Severity Index 90% (PASI 90) response were included. Bayesian random-effects NMA was performed to combine the direct and indirect evidence from randomized controlled trials (RCTs) of bDMARDs and tsDMARDs reporting joint outcomes by using ACR 20 and skin outcomes using PASI 90. NMA was conducted by Stata 17 MP software using odds ratio (OR) with 95% credible interval (CrI) to assess the clinical effectiveness. The surface Under Cumulative Ranking curve (SUCRA) was used to analyze the relative efficacy ranking of different treatments.

Results: We included 30 RCTs that enrolled 13,137 patients with PsA. There were 630 comparisons of 36 interventions, out of which 69 were direct comparisons. Compared with placebo, all interventions achieved ACR20 response rate, except for secukinumab 75mg and brepocitinib 10mg. Regarding PASI90, all the interventions showed an improvement compared to placebo except for brepocitinib 10mg and brepocitinib 30mg (table 1).

The SUCRA indicated that IV golimumab 2mg/kg, golimumab 100mg, IV infliximab 5mg/kg, golimumab 50mg, and bimekizumab 160mg had the highest probability of achieving ACR20. Regarding PASI90, the SUCRA indicated that IV infliximab 5mg/kg, golimumab 50mg, bimekizumab 160mg, golimumab 100mg, and bimekizumab 160mg LD had the highest probability of achieving the best outcome (table 1).

The ranking probability based on the SUCRA indicated that the therapies with the highest probability of achieving both ACR20 and PASI90 response rates were IV infliximab and golimumab, followed by bimekizumab 160mg and bimekizumab 160mg LD (Figure 1). Therapies were arbitrarily classified into subgroups according to their performance above or below the mean rank (Figure 2).

Conclusion: In this NMA, infliximab and golimumab demonstrated favorable efficacy achieving both ACR 20 and PASI 90 in patients with PsA. The selection of therapies for patients with PsA could be guided depending on the prevailing disease manifestation.

Ref.
1. Mease PJ, et al. Rheumatology (Oxford).2021 May 14;60(5):2109-2121
2. McInnes IB, et al. RMD Open.2022 Mar;8(1):e002074
Table 1. Odds ratio, 95% credible intervals and ranking of bDMARDs and tsDMARDS vs placebo.



Figure 1. Hierarchy of bDMARDs and tsDMARDS by mean rank of ACR20 and PASI90. Best possible outcome achieved by therapies on the right top corner and worst outcome achieved by therapies on lower left corner.
Disclosures: J. Diaz, None; G. Quiceno, None; A. Castro, None.