Background IL-23 is considered to play an important role in the development of enthesitis. Ustekinumab (UST), a combined inhibitor of IL-12/IL-23 shows efficacy in psoriatic arthritis (PsA), which is driven by enthesial and synovial disease, while it has no therapeutic role in diseases driven by synovitis alone, such as rheumatoid arthritis. We therefore speculated that inhibition of IL-23 is particularly effective in enthesitis-driven PsA patients.

Objectives To compare the efficacy of UST with tumor necrosis factor inhibitor (TNFi) treatment in clearing enthesitis in PsA patients.

Methods ECLIPSA is a prospective randomized-controlled open study. Patients with PsA with active enthesitis were randomized 1:1, receiving either standard doses of UST (arm 1) or TNFi (arm 2). At baseline the following parameters were assessed: age, gender, BMI, disease duration, previous DMARDs, use of corticosteroids, use of NSAIDs, swollen and tender joint count, VAS-pain, VAS-global, NAPSI, PASI, MASES, SPARCC, LDI, BASDAI, BASFI, HAQ-DI, SF-36, FACIT-F, ESR and CRP. Primary endpoint was a SPARCC of 0 after 6 months. Patients were seen every 3 months and followed for a total of 6 months. In order to investigate the effects of study treatment over time we used 2x3 mixed design ANOVA models for both physician's and patient's reported outcomes. Furthermore, exploratory logistic regression was used to predict a SPARCC of 0 at month 6 from baseline SPARCC, PASI, NAPSI, FACIT-F and BASDAI while additionally accounting for age, gender, PsA duration and study treatment.

Results 51 patients (UST=25; TNFi=26) were screened and 47 patients (UST=23; TNFi=24) were enrolled with 4 patients not presenting signs of active enthesitis at baseline. Mean ± SD age was 59.11±12.16 years and mean ± SD disease duration was 6.4±7.79 years. Mean± SD SPARCC at baseline was 4.87±2.69 in the UST group and 3.88±2.52 in the TNFi group. Other baseline characteristics were similar between both groups with exception of HAQ-DI, BASFI and SF-36 mental scale. In regards to the effect of study treatment (TNFi vs. UST) and time, the corresponding ANOVAs suggested an important interaction of both factors for measures of enthesitis (MASES and SPARCC), patient-reported disease activity (BASDAI and BASFI), physical well-being (SF-36 physical component summary scale), and PASI all p≤0.044 with superiority of UST. However, TNFi was superior to UST with respect to improvement of fatigue (FACIT-F), p<0.001. After 6 months, 17 out of 24 UST patients (70.8%) and 10 out of 26 TNFi patients (38.4%) reached the primary endpoint (SPARCC=0). Logistic regression predicting enthesitis-free state of disease was significantly related to study treatment only, with patients receiving UST being more likely to show no signs of enthesitis at month 6 (OR=0.037; p=0.005).

Conclusions These results show that UST is superior to TNFi in resolving the enthesitis component of disease in PsA patients with active enthesial disease. Based on these data more stratified treatment approaches can be designed in PsA patients, where enthesitis-driven patients are targeted by IL-23/IL-17 pathway inhibitors, while more systemic disease manifestations of PsA, such as patients with polyarticular disease or those with high-level fatigue are targeted by TNFi.

Disclosure of Interest None declared