Failures and Reoperations After Matrix-Assisted Cartilage Repair of the Knee: A Systematic Review

Purpose: To quantify the reported failures and reoperations for the emerging technique of matrix-assisted cartilage repair at short-term and midterm follow-up. Methods: We conducted a systematic review of 3 databases from March 2004 to February 2014 using keywords important for articular cartilage repair. Two authors reviewed the articles, the study exclusion criteria were applied, and articles were determined to be relevant (or not) to the research question. All studies with a minimum of 2 years' clinical follow-up were reviewed for all reported reoperations. The reasons for reoperations were recorded. Results: We reviewed 66 articles from the 301 articles identified in the original systematic search. There were 60 articles on matrix-assisted cartilage transplantation and 6 articles on matrix-induced chondrogenesis. The matrix-assisted cartilage transplantation studies reported on a total of 1,380 patients at 2 to 5 years' follow-up. Among these, there were 72 reoperations (5%) including 46 treatment failures (3%). These numbers increased to an 11% reoperation rate and 9% treatment failure rate at minimum 5-year follow-up of 961 patients. The most common procedures performed other than revision cartilage surgery or arthroplasty were manipulation under anesthesia for arthrofibrosis (0.7%) and debridement for graft hypertrophy (1.2%). The matrix-induced chondrogenesis studies reported on 163 patients. Among these, there were 15 reoperations (9%) that included 4 treatment failures (2%), 9 manipulations under anesthesia (6%), and 2 debridements for graft hypertrophy (1%). Conclusions: Treatment failure rates for matrix-assisted cartilage repair increase from short-term to midterm follow-up, with 11% of patients having undergone further surgery at a minimum of 5 years' follow-up. These data can be used to counsel patients on the potential need for further operative intervention after this emerging cartilage repair technique.