Total versus partial knee replacement in patients with medial compartment knee osteoarthritis: the TOPKAT RCT

• Published in: Health technology assessment (Winchester, England) England), 2020-04, Vol.24 (20), p.1-98
• Main Authors: Beard, David J, Davies, Loretta J, Cook, Jonathan A, MacLennan, Graeme, Price, Andrew, Kent, Seamus, Hudson, Jemma, Carr, Andrew, Leal, Jose, Campbell, Helen, Fitzpatrick, Ray, Arden, Nigel, Murray, David, Campbell, Marion K
• Format: Article
• Language: English
• Subjects: arthroplasty; decision-making; equipoise; expertise; knee replacement; osteoarthritis; partial knee replacement; total knee replacement
• ISSN: 1366-5278; 2046-4924
• DOI: 10.3310/hta24200

Late-stage medial compartment knee osteoarthritis can be treated using total knee replacement or partial (unicompartmental) knee replacement. There is high variation in treatment choice and insufficient evidence to guide selection. To assess the clinical effectiveness and cost-effectiveness of partial knee replacement compared with total knee replacement in patients with medial compartment knee osteoarthritis. The findings are intended to guide surgical decision-making for patients, surgeons and health-care providers. This was a randomised, multicentre, pragmatic comparative effectiveness trial that included an expertise component. The target sample size was 500 patients. A web-based randomisation system was used to allocate treatments. Twenty-seven NHS hospitals (68 surgeons). Patients with medial compartment knee osteoarthritis. The trial compared the overall management strategy of partial knee replacement treatment with total knee replacement treatment. No specified brand or subtype of implant was investigated. The Oxford Knee Score at 5 years was the primary end point. Secondary outcomes included activity scores, global health measures, transition items, patient satisfaction (Lund Score) and complications (including reoperation, revision and composite 'failure' - defined by minimal Oxford Knee Score improvement and/or reoperation). Cost-effectiveness was also assessed. A total of 528 patients were randomised (partial knee replacement,  = 264; total knee replacement,  = 264). The follow-up primary outcome response rate at 5 years was 88% and both operations had good outcomes. There was no significant difference between groups in mean Oxford Knee Score at 5 years (difference 1.04, 95% confidence interval -0.42 to 2.50). An area under the curve analysis of the Oxford Knee Score at 5 years showed benefit in favour of partial knee replacement over total knee replacement, but the difference was within the minimal clinically important difference [mean 36.6 (standard deviation 8.3) (  = 233), mean 35.1 (standard deviation 9.1) (  = 231), respectively]. Secondary outcome measures showed consistent patterns of benefit in the direction of partial knee replacement compared with total knee replacement although most differences were small and non-significant. Patient-reported improvement (transition) and reflection (would you have the operation again?) showed statistically significant superiority for partial knee replacement only, but both of these variables could b