BACKGROUND:
Extensor mechanism disruption is a devastating complication after TKA. Extensor mechanism reconstruction can be performed using an allograft or synthetic mesh. Recent studies have disagreed about the durability of these reconstructions. Evaluating a larger series of reconstructions that use current popular techniques and principles provides information to guide patient and surgeon expectations while bringing to light the potential fate of patients who experience an infection postoperatively.

QUESTIONS/PURPOSES:
(1) What was the 5-year survival free of repeat revision surgery or persistent extensor lag of more than 30° after allograft or synthetic mesh reconstruction of extensor mechanism disruptions after TKA, and did this differ between techniques? (2) What was the 5-year survival free of infection after allograft or mesh reconstruction, and what proportion of those patients experienced a secondary severe complication (arthrodesis or amputation)?

METHODS:
Between April 2008 and December 2020, 123 patients underwent extensor mechanism reconstruction after TKA at one center. Of those, 37% (45) were lost to follow-up before 2 years, had not been seen in the past 5 years, did not reach a study endpoint (repeat revision or extensor lag > 30°) before that time, or underwent primary repair and thus could not be analyzed, leaving 63% (78) for analysis here. During the study period, we considered extensor mechanism surgery when acute or chronic disruption of the patellar or quadriceps tendon was present or there was a fracture of the patella resulting in extensor lag. The decision to use either mesh or an allograft largely depended on surgeon experience and familiarity with either technique. We collected patient demographics, operative details (location of disruption and presence of a hinged prosthesis), subsequent periprosthetic joint infection (PJI), and postoperative SF-12 and Knee Injury and Osteoarthritis Outcome Score, Joint Replacement, scores. We used Kaplan-Meier survivorship analysis, with endpoints of extensor lag more than 30° or revision of the reconstruction and PJI. The mean follow-up duration was 5.3 ± 3.2 years.

RESULTS:
Survivorship free from revision or recurrent extensor lag was 54% (95% CI 44% to 66%) at 5 years. There was no difference between those undergoing reconstruction with allografts and those with mesh in terms of survivorship free from those endpoints at that timepoint (52% [95% CI 40% to 68%] versus 57% [95 CI% 41 to 78%]; p = 0.99). Survivorship free from PJI was 81% (95% CI 73% to 90%) at 2 years. There was no difference between those with allografts and those with mesh in terms of survivorship free from PJI at that timepoint (79% [95% CI 69% to 92%] versus 83% [95% CI 71% to 98%]; p = 0.75). Of the 17 patients who experienced PJI, four ultimately underwent arthrodesis and three more underwent transfemoral amputation.

CONCLUSION:
Allograft and synthetic mesh reconstructions commonly left patients with persistent limb dysfunction, while several patients had PJI and some underwent salvage procedures. Although improvements have been made with regard to treating extensor mechanism disruptions, patients and surgeons should be well aware of the potential adverse outcomes when determining management and use these findings to conduct risk-benefit analyses. There is a need for future studies to identify protocols that improve the durability of reconstructions and to determine whether interventions such as prolonged prophylactic antibiotic therapy have a role in minimizing the risk of PJI.

LEVEL OF EVIDENCE:
Level III, therapeutic study.