Abstract Background/Aims Osteoarthritis (OA) affects over 10 million people in the UK, leading to pain and reduced activity. Although no disease-modifying agents exist, intra-articular corticosteroid injections (IACI) can provide temporary pain relief. Pain sensitisation contributes to OA pain. Thrombospondin-4 (TSP-4) is an extracellular matrix protein associated with tissue repair, neurogenesis, pain and development of sensitisation. TSP-4 was evaluated as a potential biomarker predicting IACI efficacy. Methods In this real-world cohort study (IRAS ID 234659), participants with painful, symptomatic knee OA, RA and PSA (American College of Rheumatology (ACR) criteria) receiving usual NHS care and eligible for IACI were recruited. Baseline data included demographics (age, sex, ethnicity, body mass index, analgesia use) and patient-reported outcomes, including Visual Analogue Scale (VAS), painDETECT, and Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC). Participants with painDETECT scores ≥19 were classified as “sensitised” to pain. Blood samples were collected, and TSP-4 expression was quantified by ELH-TSP4 enzyme-linked immunosorbent assay (ELISA) kits (RayBiotech, Tububio, UK). Appropriate participants underwent ultrasound-guided IACI. At 3 months, VAS/WOMAC/painDETECT/TSP-4 expression was reassessed. Treatment response was defined as a ≥ 20% decrease in WOMAC pain score. Treatment-associated changes were assessed for significance using the Mann-Whitney test in R. Results Of 129 participants enrolled, 95% (122) received IACI, and 95 were analysed (67 OA, 22 RA, 6 PSA), with 34 lost to follow-up. Of these, 43% (n = 50) of the complete cohort and 42% (n = 32) of OA participants met the response criterion for decreased pain. Across all participants (three arthritis subtypes), responders demonstrated a significant decrease in TSP-4 expression at 3 months following IACI (p = 0. 006), while non-responders exhibited a significant increase (p = 0. 005). Within the responder group, only VAS pain score was significantly reduced (p = 0. 013), whereas for the OA responders, both VAS (p = 0. 0008) and WOMACP (p = 0. 04) were significantly reduced. There were no significant changes in painDETECT for any grouping, but across all participants there were significant decreases of VAS (p = 0. 0008) and WOMACP (p = 0. 04) with IACI. Conclusion We have shown, for the first time in human arthritis subtypes, a dynamic relationship between TSP4 and pain. The direction of change in TSP4 matches the treatment response, decreasing when pain levels are improved and increasing when they are not. In this cohort of predominantly non-sensitised participants (hence the non-significant change in painDETECT), higher levels of TSP4 were associated with responders, with a more significant pain reduction for OA patients. This study reinforces the relevance of assessing TSP-4 in mechanistic studies to better understand treatment selection and response. Further work in larger cohorts could be used to develop a predictive model, incorporating TSP-4 and other clinical/imaging biomarkers relating to pain that are subtype-specific. This could enable targeted IACI use, optimising patient outcomes, thereby reducing the economic and healthcare-related burden of arthritis. Disclosure A. Niakan: None. A. Paterson: None. K. Feather: None. A. Lambarth: None. M. Siebachmeyer: None. V. Ejindu: None. F. Howe: None. A. Rudnicka: None. A. Ezeonyeji: None. N. Sofat: None.
Niakan et al. (Wed,) studied this question.