Results of Anterior Cruciate Ligament Reconstruction With Patellar Tendon Autografts: Objective Factors Associated With the Development of Osteoarthritis at 20 to 33 Years After Surgery

Abstract

Background: Few studies exist that report objective radiographic and physical examination results at >20 years after anterior cruciate ligament (ACL) reconstruction. Hypothesis: The risk of osteoarthritis (OA) at >20 years after surgery would be statistically significantly higher if an increased severity of factors was present. Study Design: Cohort study; Level of evidence, 2. Methods: Between 1982 and 1994, 1428 knees met the inclusion criteria of being primary ACL surgery, having no existing OA or other ligamentous laxity, and having no known graft tear. Prospective data analyzed included demographics; meniscus and articular cartilage status; ratings of bilateral standing posteroanterior weightbearing, lateral, and Merchant view radiographs; KT-1000 arthrometer measurements; and range of motion measurements. Radiographs were graded for joint space narrowing, sclerosis, and osteophytes according to International Knee Documentation Committee (IKDC) criteria. Multivariate and univariate logistic models were used to determine the effect of potential predictors on the odds of having radiographic evidence of OA. IKDC and Cincinnati Knee Rating System (CKRS) subjective surveys were conducted. Results: A minimum 20-year objective follow-up was obtained for 423 knees at a mean of 22.5 ± 2.1 years postoperatively. If a patient lacked normal extension or flexion at discharge, the odds of lacking normal extension or flexion at follow-up were 19.7 and 7.97, respectively ( P < .001). Radiographic ratings were normal for 35.2%, nearly normal for 36.2%, abnormal for 20.1%, and severely abnormal for 8.5%. Multivariate analysis showed that the predictive factors for the presence of OA in the long-term were medial meniscectomy, older age at surgery, and less than normal knee extension at discharge. The odds ratios for statistically significant factors for the presence of any OA at follow-up were 2.02 for knee extension loss at discharge, 2.98 for medial meniscectomy, 1.65 for lateral meniscectomy, 1.06 for age at follow-up, 1.62 for chronic injury type, and 2.17 for articular cartilage damage. The IKDC and CKRS subjective scores were progressively and statistically significantly lower as the level of OA was greater. Conclusion: The prevalence rate of developing moderate to severe OA at >20 years after ACL reconstruction was 28.6%. Significant factors predictive of OA in the long-term were older age at surgery, medial meniscectomy, and knee extension loss.