BackgroundThe differential contribution of IL-6 and IL-17 pathways to the pathogenesis of psoriatic arthritis (PsA) is not fully understood. Upadacitinib (UPA), an oral JAK inhibitor, was more effective than placebo (PBO) in improving key clinical manifestations of PsA in two global phase 3 trials, SELECT-PsA 1 (non-biological DMARD-IR, nbDMARD-IR) and SELECT-PsA 2 (biological DMARD-IR, bDMARD-IR).1,2 Targeted proteomic analysis suggested that UPA modulates multiple biological pathways in innate and adaptive immune systems via direct and indirect inhibition of key regulators, including IL-6 and IL-17 pathways, with a possible shift from Th1 predominance in nbDMARD-IR PsA to a more Th17 bias in bDMARD-IR PsA.3ObjectivesWe assessed the relationship between IL-6 and IL-17 pathway modulation and different clinical outcomes after UPA treatment in nbDMARD-IR and bDMARD-IR PsA patients.MethodsA subset of patients was randomly selected from SELECT-PsA 1 (n=74 of UPA 15 mg QD, n=74 of PBO) and PsA 2 studies (n=90 of UPA 15 mg QD, n=81 of PBO). Serum levels of IL-6, IL-17A, IL-17F, and beta-defensin 2 (BD2) proteins were measured at baseline, week 2, and week 12 by validated immunoassays. The quantitative cytokine measurements were transformed as log10, and PASI score was transformed as log10 (x+1) prior to analysis. A Repeated Measure Mixed Linear Model was used to compare UPA versus PBO treatment effects in overall selected patients and between responders and non-responders defined by PASDAS score ≤ 3.6 (Minimal Disease Activity, MDA)4 and PASI75 at week 12, respectively. The relationships between cytokines and clinical outcomes (PASI and DAS28-CRP) were assessed by Pearson’s correlation at baseline and after treatment.ResultsIn nbDMARD-IR PsA patients, baseline IL-17A, IL-17F, and BD2 levels correlated with each other and with PASI, while IL-6 appeared independent from the IL-17 pathway and correlated with DAS28-CRP. At week 12, UPA treatment significantly decreased IL-6 and BD2. The decrease of IL-6 was more pronounced in PASDAS MDA responders and correlated with DAS28-CRP improvement, but the decrease of BD2 was significant in PASI75 responders and correlated with PASI improvement. In contrast, IL-17A and IL-17F were not significantly changed after UPA treatment, neither correlated with clinical outcomes at week 12.In bDMARD-IR PsA patients, baseline IL-17A level was significantly elevated compared to nbDMARD-IR patients but weakly correlated with other cytokines and show no correlation with PASI. At week 12, the reduction of IL-6 after UPA treatment was not different between responders and non-responders (PASDAS MDA or PASI75) and did not correlate with DAS28-CRP improvement, while the reduction of BD2 remained significant in PASI75 responders and correlated with PASI improvement. Further, UPA treatment significantly reduced IL-17A in PASDAS MDA responders and IL-17F in PASI75 responders compared to non-responders, respectively. The reduction of IL-17F correlated with PASI improvement at week 12.ConclusionIL-6 and IL-17 pathway inhibition after UPA treatment showed different profiles in relationship with clinical outcomes in nbDMARD-IR versus bDMARD-IR PsA patients. IL-6 decrease was more pronounced in nbDMARD-IR PsA patients and associated with joint manifestation improvement, while IL-17A and IL-17F decreases were only observed in bDMARD-IR PsA patients and associated with psoriasis improvement. BD2, a biomarker of Th17-associated skin pathology, significantly decreased after UPA treatment in both nbDMARD-IR and bDMARD-IR PsA studies, which likely contributed to UPA effects on psoriasis improvement in a broad range of PsA patients.