PO:04:050 | Cluster-Based Endotyping of Psoriatic Arthritis in a Monocentric Italian Cohort: A Real-World study

Eneida Cela¹, Arianna D'Antonio¹, Luigi Fiannacca¹, Alberto Bergamini¹, Maxime Breban², Felicie Costantino², Maria Sole Chimenti¹ | ¹Rheumatology, Allergology and Clinical Immunology, Department of System Medicine, University of Rome Tor Vergata Rome, Italy; ²Rheumatology Department, AP-HP, Ambroise-Paré Hospital Paris, France

Background. Psoriatic arthritis (PsA) is a chronic inflammatory arthritis with a broad spectrum of skin and joint manifestations. Identifying cluster-based endotypes could offer a more personalized treatment strategy and improve long-term outcomes. This study aimed to apply cluster analysis to a real-world cohort of PsA patients initiating biologic disease modifying anti-rheumatic drugs (bDMARDs), in order to define clinically meaningful endotypes.

Methods. Factorial analysis of mixed data followed by hierarchical clustering on principal components was performed in 151 patients with active PsA initiating a biologic treatment for the first time. Clinical, demographic, and disease activity variables were used for clustering.

Results. Among the 151 patients included, the optimal number of clusters was determined to be two with each cluster having a comparable size (Cluster 1: n = 76 (50.3%); Cluster 2: n = 75 (49.7%) (Figure 1). There was a similar distribution in terms of age and disease duration in both clusters, whereas Cluster 2 contained a higher proportion of females (p=0.007). (Table 1) Cluster’s stability was moderate to good with a mean ARI of 0.61. The most distinctive features between the clusters were related to joint and skin involvement. Cluster 1 was characterized by a higher frequency of PsO by history (61% vs. 21%; p<0.001) and at inclusion (40,3% vs. 4%; p<0.001). In contrast, patients in this cluster exhibited milder articular symptoms than in Cluster 2, including lower TJC and SJC, lower DAPSA, and better functional status (HAQ-DI) (all p < 0.001). Moreover, patients in Cluster 2 exhibited with a higher frequency of both peripheral and axial manifestations, including inflammatory back pain (51% vs. 30%, p = 0.011) by history. Notably, 96% of patients in Cluster 2 had no active PsO at the time of assessment. Extra-articular manifestations such as uveitis and inflammatory bowel disease did not differ between both groups. No notable differences were found between clusters in terms of other comorbidities, including fibromyalgia and metabolic syndrome. Additionally, we observed no significant difference for peripheral and axial joint radiographic damage and the use of NSAIDs, corticosteroids and csDMARDs was similarly distributed across both clusters.

Conclusions. We identified two distinct endotypes among PsA patients referred to initiate bDMARDs: a PsO-dominant endotype and an articular-dominant one with more severe joint disease. This two-cluster model could be used to stratify patients for personalised treatment decision and prediction of long-term outcome.