How are patient-reported pain outcomes associated with biomarker and structural pathology subtypes in knee osteoarthritis? An explorative evaluation in the IMI-APPROACH cohort

Objective: To explore associations between patient-reported pain outcomes and knee osteoarthritis (OA) subtypes based on systemic biochemical markers and joint structural pathology as defined by MRI. Methods: Data were obtained from 297 knee OA patients from the IMI-APPROACH study. Pain outcomes were assessed using the KOOS, WOMAC, ICOAP, NRS, PainDETECT, and a pain diary. Biochemical markers in serum and urine were used to classify patients into systemic biomarker subtypes (low tissue turnover, structural damage, and systemic inflammation) via k-means clustering. Structural pathology subtypes were determined using MRI into an inflammatory, meniscus/cartilage damage, and subchondral bone pathology subtype. Associations between pain measures and subtypes were analyzed using multivariable regression models adjusted for age, sex, and BMI. Results: The systemic inflammation biomarker subtype was significantly associated with higher KOOS pain, WOMAC weight-bearing pain, NRS knee pain, and PainDETECT scores (all p ​≤ ​0.042 and β ​≥ ​0.12). The low tissue turnover subtype negatively associated with lower KOOS, WOMAC, and ICOAP constant pain (all p ​≤ ​0.22 and β ​≤ ​−0.13), and the structural damage subtype with lower PainDETECT scores (more nociceptive-like pain; p ​= ​0.046 and β ​= ​−0.12). Among MRI subtypes, meniscus/cartilage damage was significantly associated with lower PainDETECT scores (p ​= ​0.005 and β ​= ​−0.16). No significant associations were found for the subchondral bone subtype or pain diary outcomes. Conclusion: For commonly used pain questionnaires, pain severity seems linked with inflammatory activity more than structural damage. Structural damage is primarily associated with nociceptive-like pain according to PainDETECT, which might be valuable for patient selection to clinical trials and observational studies.