No abstract

In knee osteoarthritis (OA), pain sensitization has been linked to a more severe symptomatology, but the prognostic implications of pain sensitivity in people undergoing conservative treatment such as physiotherapy are not established. This study aimed to prospectively investigate the association between features of pain sensitization and clinical outcome (nonresponse) after guideline-based physiotherapy in people with knee OA. Participants (n = 156) with moderate/severe knee OA were recruited from secondary care. All participants completed self-administered questionnaires and underwent quantitative sensory testing at baseline, thereby establishing subjective and objective measures of pain sensitization. Participants (n = 134) were later classified after a physiotherapy intervention, using treatment responder criteria (responder/nonresponder). Quantitative sensory testing data were reduced to a core set of latent variables using principal component analysis. A hierarchical logistic regression model was constructed to investigate whether features related to pain sensitization predicted nonresponse after controlling for other known predictors of poor outcome in knee OA. Higher temporal summation (odds ratio 2.00, 95% confidence interval 1.23-3.27) and lower pressure pain thresholds (odds ratio 0.48, 95% confidence interval 0.29-0.81) emerged as robust predictors of nonresponse after physiotherapy, along with a higher comorbidity score. The model demonstrated high sensitivity (87.8%) but modest specificity (52.3%). The independent relationship between pain sensitization and nonresponse may indicate an underlying explanatory association between neuroplastic changes in nociceptive processing and the maintenance of ongoing pain and disability in knee OA pain. These preliminary results suggest that interventions targeting pain sensitization may warrant future investigation in this population.

This systematic review found insufficient evidence to support an independent predictive relationship between QST measures of nervous system sensitization and treatment outcome. Self-report measures demonstrated better predictive ability. Further high-quality prognostic research is warranted.

Objectives: Pain sensitization in knee osteoarthritis (OA) is associated with greater symptom severity and poorer clinical outcomes. Measures that identify pain sensitization and are accessible to use in clinical practice have been suggested to enable more targeted treatments. This merits further investigation. This study examines the relationship between quantitative sensory testing (QST) and clinical measures of pain sensitization in people with knee OA. Methods: A secondary analysis of data from 134 participants with knee OA was performed. Clinical measures included: manual tender point count (MTPC), the Central Sensitization Inventory (CSI) to capture centrally mediated comorbidities, number of painful sites on a body chart, and neuropathic pain-like symptoms assessed using the modified PainDetect Questionnaire. Relationships between clinical measures and QST measures of pressure pain thresholds (PPTs), temporal summation, and conditioned pain modulation were investigated using correlation and multivariable regression analyses. Results: Fair to moderate correlations, ranging from −0.331 to −0.577 (P<0.05), were identified between MTPC, the CSI, number of painful sites, and PPTs. Fair correlations, ranging from 0.28 to 0.30 (P<0.01), were identified between MTPC, the CSI, number of painful sites, and conditioned pain modulation. Correlations between the clinical and self-reported measures and temporal summation were weak and inconsistent (0.09 to 0.25). In adjusted regression models, MTPC was the only clinical measure consistently associated with QST and accounted for 11% to 12% of the variance in PPTs. Discussion: MTPC demonstrated the strongest associations with QST measures and may be the most promising proxy measure to detect pain sensitization clinically.