Abstract
Emerging evidence suggest that quantitative sensory testing (QST) may predict the treatment response to pain-relieving therapies. This systematic review and meta-analysis focus on the predictive value of QST for pain management of knee osteoarthritis (OA). MEDLINE and EMBASE were systematically searched for all studies from year 2000 to 2023 on pretreatment QST and treatment of OA including surgical, pharmaceutical, and nonsurgical and nonpharmaceutical therapies. Preclinical studies and reviews were excluded. The systematic review followed the PRISMA guidelines and was pre-registered on the Open Science Framework website (link: https://osf.io/4FETK/, Identifier: DOI 10.17605/OSF.IO/4FETK). Meta-analysis were conducted to demonstrate the strength of the pre-treatment QST predictions on pain outcomes after OA treatments. Sixteen surgical (all on total knee arthroplasty [TKA], N = 1967), 5 pharmaceutical (4 on non-steroidal anti-inflammatory drugs [NSAIDs], N = 271), and 4 exercise-based therapy studies (N = 232) were identified. Pretreatment QST parameters predicted pain-relieving treatment outcomes in 81% of surgical, 100% of pharmaceutical, and 50% of exercise-based therapy studies. Meta-analyses found pretreatment QST profiles to predicted pain outcomes after TKA (random effects: 0.309, 95% confidence interval [CI]: 0.206-0.405, P < 0.001), NSAIDs (random effects: 0.323, 95% CI: 0.194-0.441, P < 0.001), and exercise-based therapies (random effects: 0.417, 95% CI: 0.138-0.635, P = 0.004). The overall risk of bias for the included studies was low to moderate. This systematic review and meta-analysis demonstrate weak-to-moderate associations between pretreatment QST and pain outcomes after standard OA pain treatments. Based on this work, it is hypothesized that a subset of specific pain sensitive patients with OA exist and that these patients do not respond adequately to standard OA pain treatments.