Background/Purpose

Currently, there are a number of effective therapies for psoriatic arthritis (PsA). The objective of this systematic review was to assess the efficacy (PSARC, ACR 20/50/70 and PASI75 response) and adverse event profile (AEs) (any AEs, serious AEs, and infection rate) of using TNF inhibitors or new therapies including Ustekinumab and Apremilast to treat active PsA.

Methods

This review included all published/unpublished RCTs comparing biologics or Apremilast to placebo in PsA patients. Multiple mechanisms of action (MOA) were included: TNF inhibitor (Adalimumab, Certolizumab, Etanercept, Infliximab and Golimumab), IL12/23 inhibitor(Ustekinumab) and PDE4 inhibitor(Apremilast). The databases MEDLINE(n=245), EMBASE(n=1551) and CENTRAL(n=223) were searched. Conference abstracts from the ACR and American Academy of Dermatology (2009-13) were hand searched. Two reviewers independently screened titles/abstracts/full text for eligibility. Data extraction and risk of bias assessment were performed in duplicate using Cochrane’s Risk of Bias Assessment.

Results

18 studies (5433 participants) were included. Kappa for full text and risk of bias was 0.8545 and 0.9217 respectively. When the MOAs were pooled, the risk ratio (RR) for achieving PSARC, ACR20,50,70, and PASI75 at 12-16 and 24 weeks(wk) were all statistically significant. The likelihood of achieving the primary outcome ACR20 (12-16wk) had RR=2.74(95%CI=2.27-3.32) with significant heterogeneity (I²= 65%) which was explained by a priori subgroup analysis (by MOA). When comparing the ACR20 (12-16wk) response by MOA, the RR was 3.99 (95%CI: 2.91-5.46), 2.03 (95%CI: 1.70-2.42) and 2.08 (95%CI: 1.66-2.59) respectively for the TNF, PDE4 and IL12/23 inhibitors. RR of achieving ACR20 response (24wk) by MOA is shown in figure1. Low heterogeneity (I²<40%) was seen for the outcomes ACR50,70 at 12-16/24wk respectively, suggesting there may be comparable efficacy of each MOA class for these outcomes. There was no significant increase in risk for any AEs or severe AEs, except a mild increase in risk of infections at 24wk (RR=1.15, 95%CI= 1.00-1.32).

Conclusion

TNF inhibitors, Ustekinumab & Apremilast all appear to produce significant benefits on joint and skin disease for individuals with PsA.  The safety profile, aside from a slight increased risk of infections, appears acceptable. When comparing the three MOAs, the TNF inhibitors appeared most effective in achieving an ACR20 response at 12-16wk. However, by 24wk there was no significant difference between TNF inhibition and Apremilast while a significant difference remained between TNF inhibition and Ustekinumab. Whilst head to head trials are needed to definitively draw such conclusions, incorporating additional RCTs in future analyses will provide greater power to characterize this relationship.

Fig1: RR for achieving ACR20 response (24wk) vs placebo by MOA