Background Psoriasis (Pso) is one of the most common chronic inflammatory skin diseases in Europe. Psoriatic arthritis (PsA) is closely associated to Pso. Up to 30% of the Pso patients will develop PsA during skin disease course. Defined and validated approaches for early detection are still missing. Beside biomarkers from blood or imaging, clinical characteristics of the patients may be of value to detect PsA patients in the transition state early.

Objectives To perform an AI-based cluster analysis in a cohort of Pso patients at-risk for development of PsA to assess clinical characteristics as markers for early PsA.

Methods Clinical data sets from the recently published XCITING study [1] were used to perform an AI based analysis using the attributes “tenderness of joints (68 TJC)” “tenderness at entheses (LEI)” to detect clinical profiles for early detection of inflammatory musculoskeletal (MSK) processes defined as positive findings in clinical examination (SJC, PsA diagnosis), MSUS or fluorescence optical imaging. Main criteria for inclusion into XCITING as at-risk study were Pso without PsA diagnosis, nail involvement and/or MSK complaints within the last 6 months. For clustering the patients, the AI-based analysis contained three steps: (1) Reducing random noise by performing Non-Negative Matrix Factorization, (2) embedding into 2D using t-SNE and (3) clustering using DBSCAN.

Results Characteristics of the XCITING cohort were described previously [1]. After performance of the cluster analysis, seven different cluster types were identified (cluster 0-6) according to the clinical data sets of the cohort by use of the attributes and tested for their significance to predict the presence or absence of MSK inflammation in Pso (swollen joints, positive MSUS or positive FOI). Three “tenderness clusters” out of 7 were identified with significant correlation: while as expected cluster 2 (no major findings in LEI and TJC) is associated with no inflammation, “feet- type” involvement (cluster 4) and dominance at PIP and DIP joints (cluster 6) (Figure 1, Table 1) are associated with MSK inflammation at the hands.

Conclusion Markers for early detection of Pso patients who will develop PsA are missing. Within this analysis we show, that by use of clinical data sets only, risk profiles developed from finding of tenderness at different anatomical regions might be helpful for detection of inflammatory MSK processes. Interestingly, the feet tenderness can also predict MSK inflammation at the hands. A combination of both, clinical data sets and liquid/imaging biomarkers may be identified on base of this observation to increase the potential to detect Pso patients with high-risk profiles for PsA early.

Reference [1]Koehm, M., et al., Fluorescence-optical imaging as a promising easy-to-use imaging biomarker to increase early psoriatic arthritis detection in patients with psoriasis: a cross-sectional cohort study with follow-up. RMD Open, 2022. 8(2).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Cluster analysis of patients with psoriasis only but increased risk for PsA according to the attributes “tenderness of joints (68 TJC)” and “tenderness at entheses (Leeds enthesitis index)”

Acknowledgements This project is part of the working program within the HIPPOCRATES consortium. HIPPOCRATES has received funding from the Innovative Medicines Initiative 2 Joint Undertaking (JU) under grant agreement No. 101007757. The JU receives support from the European Union’s Horizon 2020 research and innovation programme and EFPIA.

Disclosure of Interests None Declared.