Background Psoriatic arthritis (PsA) can cause important structural damages which can lead to disability. TNF antagonists have shown their clinical efficacy in PsA, but only a few data are available regarding their structural effect.

Objectives To determine whether TNF antagonists have an effect on radiographic progression in patients with PsA by performing a systematic review and meta-analysis based on data from randomized controlled trials (RCTs) versus placebo. To check if the combination with methotrexate (MTX) is superior to monotherapy.

Methods A systematic review of literature was performed until March 2011. Bibliographic references were selected from Embase and Medline databases, and from the two last EULAR and the ACR annual meetings. Radiographs progression were scored by Sharp method modified for PsA (mTSS). Primary endpoint was the proportion of patients with no radiographic progression at week 24 (defined by a mTSS variation score ≤0.5). Secondary end point included the proportion of non progressors at week 48 and mean variation of mTSS at week 24. The Mantel-Haenszel method was used to provide a common odds-ratio (OR) estimate and 95% confidence interval (CI) in TNF antagonists (+/– DMARDs) versus placebo (+/– DMARDs) treated patients for infliximab, etanercept, adalimumab and golimumab RCTs. Statistical heterogeneity was assessed on the Q test (χ²), using a significance level of 0.05. OR and 95% CI were shown on forest plots.

Results Search found out 206 articles and 3 abstracts. Retrieved data allowed meta-analysis on 4 articles and 1 abstract for the proportion of non progressors at week 24. 1110 patients were included, 484/584 (82,9%) were considered as non-progressors at week 24 in the TNF antagonists group versus 362/526 (68,8%) in the placebo group (OR=2.68 [1.99; 3.60] p<0.0001) without heterogeneity (I²=3%;p=0.39) (figure 1). Based on 3 studies, same results were found at week 48 in favor of TNF antagonists group (OR=2.42 [1.57;3.71] I²=0%;p=0.91). Among 533 patients receiving TNF antagonists versus 454 receiving placebo in 3 studies (4 comparisons), the mean change of the mTSS at week 24 was lower in the TNF antagonist group versus placebo (Mean difference= -0,69 [-1.12; -0.27], with heterogeneity (I²=76%;p=0.006). Only two RCTs provided datas on the combination with MTX: the mean or median change of the mTSS was similar in adalimumab or infliximab subgroups, irrespective of the MTX use.

• Download figure
• Open in new tab
• Download powerpoint

Conclusions This meta-analysis of RCTs showed that all TNF antagonists lead to a better control of structural damages of the PsA than a placebo after 24 and 48 weeks of treatment. The respective role of DMARDS was not determined due to a lack of data.

Disclosure of Interest None Declared