Meta-analysis Comparing Allograft to Synthetic Reconstruction for Extensor Mechanism Disruption after Total Knee Arthroplasty.

Extensor mechanism disruption following total knee arthroplasty (TKA) is a devastating complication that causes high failure rates. There is controversy on what is the best way to do an extensor mechanism reconstruction. This study aims to compare both allograft and synthetic reconstructive techniques for success, reoperation, and infection rates and functional outcomes. The search on PubMed, MEDLINE, Embase, BIOSIS, and Cochrane databases was performed on March 15, 2019, using the following keyword groups: (1) "extensor mechanism" and "total knee arthroplasty," (2) "extensor mechanism" and "knee arthroplasty," (3) "extensor mechanism" and "revision total knee arthroplasty," and (4) "extensor mechanism" and "revision knee arthroplasty". Only studies on extensor mechanism disruption after TKA that included sufficient data to compare these two surgical techniques were included. Meta-analysis was performed with random effect model using the DerSimonian-Laird method. Thirty studies were included. The overall success rate of the reconstruction was 73.3% (95% confidence interval [CI]: 0. 651, 0.814). The success rate of allograft (72.8%, 95% CI: 0.626, 0.829) was not significantly different from synthetic material (78%, 95% CI: 0.707, 0.852, p = 0.416). There was no significant difference in revision rates between allograft (14.2%, 95% CI: 0.095, 0.189) and synthetic material (16%, 95% CI: 0.096, 0.223, p = 0.657). The overall relative risk of infection was 4.301 (95% CI: 1.885, 9.810). There was no significant difference in relative risk of infection between allograft (3.886, 95% CI: 1.269, 11.903) and synthetic material (4.851, 95% CI: 1.433, 16.419, p = 0.793). No statistically significant difference was found in mean postoperative Knee Society score (73.109 [95% CI: 67.296, 78.922] vs. 72.679 [95% CI: 69.184, 76.173], p = 0.901) between allograft and mesh reconstruction groups. There were no significant differences in overall failures, infections, functional outcomes, or revision reconstructions between allograft and synthetic material extensor mechanism reconstructions. Our results demonstrate the difficulty in treating this serious injury, independent of technique, as well as the significant risk for overall failure and infection.