Martyn Snow, Martyn Snow, Lee Middleton, Lee Middleton, Samir Mehta, Samir Mehta, Andrew Roberts, Andrew Roberts, Richardson Jb, Jan Herman Kuiper, Jan Herman Kuiper, Martyn Snow, Martyn Snow, Lee Middleton, Lee Middleton, Samir Mehta., Samir Mehta., Andrew Roberts, Andrew Roberts, Richard Grey, James Hutchinson, Jan Herman Kuiper, Jan Herman Kuiper, Steve White, Simon Roberts, David Griffiths, Aslam Mohammed, Kirti Moholkar, Tanweer Ashraf, Marcus Green, James Hutchinson, Tony Bhullar, Sanjeev Chitnis Andrew Shaw, L. van Niekerk, Anthony Hui, Jon Olav Drogset, Gunnar Knutsen, Mike McNicholas, Mark Bowditch, David Johnson, Turner Pg, Sanjiv Chugh, Neil Hunt, Salman Ali, Simon Palmer, Andrew Perry, Alastair Davidson, Peter Hill, Sunny Deo, Venkat Satish, M Radford, Ron Lang- staff, David Houlihan-Burne, Dominic Spicer, Padman Phaltankar, Ahmed Hegab, David Marsh, Steve Cannon, Tim Briggs, Rob Pollock, Richard Carrington, John Skinner, G. Carter Bentley, Andrew Price, Peter Schranz, Vipul Mandalia, Shaun O’Brien
OBJECTIVE: To determine at 10 years whether ACI was superior to alternative forms of cartilage management in patients after a failed previous treatment for chondral or osteochondral defects in the knee. METHODS: 390 patients were randomly assigned to receive either ACI or alternative management. Patients aged 18-55 years with symptomatic cartilage defects who had failed one previous therapeutic surgical procedure were included. Randomization was stratified by age, defect size, defect location, chosen intended alternative treatment and baseline Lysholm knee score. The primary outcome was a patient-completed Lysholm knee score. The main secondary outcome was time to treatment failure. Analysis was performed on all available data and the intention-to-treat principle using a mixed linear model. RESULTS: in both groups. At 10 years, 70% completed all patient Lysholm scores, however 94% provided 1 response and contributed to the analysis. The mean Lysholm score in those randomised to ACI was significantly higher (mean difference 7.3 points; 95% CI 2.5-12.1). Sensitivity analysis on the impact of further operations following treatment failure suggests a treatment effect that still favours ACI but was somewhat reduced (mean difference 3.8 points, 95%CI: -1.8-9.5). Ten-year treatment failure rates were comparable (29% and 25%, HR 1.04, 95%CI 0.69-1.6). In the ACI arm, mean scores were not affected by previous marrow stimulation, whereas in the alternative arm, mean scores were significantly worse. INTERPRETATION: ACI results in superior knee function at 10 years follow-up compared to alternate surgical treatments. The results of ACI were not impacted by prior marrow stimulation. REGISTRATION: International Standard Randomised Controlled Trial Number: 48911177.