Background Pain in patients (pts) with psoriatic arthritis (PsA) has multifaceted origins; sustained improvement is difficult to achieve.1 Guselkumab (GUS), a fully human monoclonal antibody that selectively inhibits IL-23, is effective in treating multiple domains of PsA including joint, skin, and entheseal symptoms, and also elicits long-lasting improvements in pt-reported pain in the DISCOVER-1&2 trials of pts with active PsA.2

Objectives These post hoc analyses were conducted to identify determinants of changes in pt-reported pain in PsA pts using pooled data through 1 year of DISCOVER-1&2.

Methods Enrolled adult pts had active PsA despite standard therapies. DISCOVER-1 pts had ≥3 swollen and ≥3 tender joints and C-reactive protein (CRP) ≥0.3 mg/dL; DISCOVER-2 pts had ≥5 swollen and ≥5 tender joints and CRP ≥0.6 mg/dL. 31% of DISCOVER-1 pts received 1-2 prior tumor necrosis factor inhibitors; DISCOVER-2 pts were biologic-naïve. Pts were randomized 1:1:1 to GUS 100 mg every 4 weeks (wks) (Q4W); GUS 100 mg at W0, W4, then every 8 wks (Q8W); or placebo (PBO); PBO pts crossed over to GUS 100 mg Q4W at W24. Determinants with a statistically important effect (p<0.15) on pain (0-100 mm Visual Analogue Scale) in univariate Repeated Measures Generalized Linear Mixed Effects Models were included in a multivariate model employing backward stepwise selection (Pout=0.1) to identify independent determinants of pain improvement over 24 wks; the model was then tested separately in pts treated with PBO (through W24) and with GUS (through W24 and through W52).

Results GUS was associated with significantly greater improvement in pain compared to PBO as early as 2 wks post-treatment; there was a significant interaction between treatment group and time, with effect of GUS on pain continuously enhanced through W24. Higher baseline (BL) pain score, worse mental health (assessed with the Short-Form-36 Mental Component Summary [SF-36 MCS] score), and lower fatigue level and lower tender joint count [TJC] were also associated with significantly greater pain improvements at W24, while background use of NSAIDs was a negative predictor of pain improvement (Table 1). Treatment effect on pain was independent of PsA duration, gender, PsA subtype, prior TNFi exposure, BL skin disease, and BL swollen joint count (SJC). Continuous significant improvement from BL in pain with GUS extended through W52 even after adjustment for the identified determinants of pain improvement through W24 (Figure 1). At W52, predictors of change in pain remained significant with the exception of SF-36 MCS score (Table 1). Results did not exclude a small number of enrolled pts with fibromyalgia (FM: nGUS=8; nPBO=4). According to these exploratory findings, medical history of FM was associated with lower pain improvement through W24 (p=0.066); in the models run separately in pts with GUS and PBO, pts with FM treated with GUS had a mean (95% CI) pain improvement (-9.1 [-19.5, 1.2]) while pts treated with PBO had a mean worsening (0.7 [-12.5, 13.9]). Pain improvement through 52 wks was significant regardless of FM: pts with FM had a mean (95% CI) improvement of -14.7 (-25.9, -3.6) comparable to non-FM pts at W24, while pain improvement in pts with no FM was -22.2 (-24.0, -20.4).

Conclusion Early significant effects of GUS on pain were enhanced through 1 year. Significant predictors of change in pain were consistent at W24 and W52, with the exception of mental health measures. The impact of mental status on pt-reported pain and the potential for GUS to improve pain in pts with FM warrant further consideration.