Zipeng Ye1, Junjie Xu1, Jiebo Chen1, Yi Qiao1, Chenliang Wu1, Guoming Xie2, Shikui Dong3, Jinzhong Zhao4. 1. Department of Sports Medicine, Shanghai Jiao Tong University Affiliated Sixth People's Hospital, 600 Yishan Road, Shanghai, 200233, China. 2. Department of Sports Medicine, Shanghai Jiao Tong University Affiliated Sixth People's Hospital, 600 Yishan Road, Shanghai, 200233, China. xieguoming2006@163.com. 3. Department of Sports Medicine, Shanghai Jiao Tong University Affiliated Sixth People's Hospital, 600 Yishan Road, Shanghai, 200233, China. ntmcdsk@126.com. 4. Department of Sports Medicine, Shanghai Jiao Tong University Affiliated Sixth People's Hospital, 600 Yishan Road, Shanghai, 200233, China. jzzhao@sjtu.edu.cn.
Abstract
PURPOSE: To identify the radiological predictive risk factors for anterior cruciate ligament reconstruction (ACLR) failure, compare the diagnostic accuracies of different parameters of conventional radiographs and magnetic resonance imaging (MRI), and determine the cutoff values for patients at higher risk. METHODS: Twenty-eight patients who were diagnosed as ACLR failure via MRI or arthroscopic examination were included in the study group. They were matched to 56 patients who underwent primary ACLR with the same surgical technique and without graft failure at the minimum 24-month follow-up by age, sex, and body mass index. On true lateral whole-leg radiographs, the posterior tibial slope (PTS) referenced to the tibial mechanical axis (PTS-mechanical), PTS referenced to the tibial proximal anatomical axis (PTS-anatomical), and anterior tibial translation (ATT) were measured. On the sagittal slices of MRI, the medial tibial slope (MTS), medial tibial plateau (MTP) subluxation (MTPsublx), lateral tibial slope (LTS), and lateral tibial plateau (LTP) subluxation (LTPsublx) were obtained. Receiver operator characteristic (ROC) curves were constructed to compare the diagnostic performance and determine the cutoff values of different radiological parameters. RESULTS: The study group demonstrated higher values of PTS-mechanical (10.7° ± 2.9° vs 8.7° ± 1.9°, p = 0.003), PTS-anatomical (13.2° ± 2.8° vs 10.5° ± 2.5°, p < 0.001), ATT (10.7 ± 3.3 mm vs 8.9 ± 2.2 mm, p = 0.014), LTS (9.4° ± 2.1° vs 5.5° ± 2.5°, p < 0.001), and LTPsublx (8.2 ± 2.8 mm vs 6.8 ± 1.9 mm, p = 0.009) as compared with the control group. The area under the ROC curve of LTS was significantly larger than that of PTS-mechanical (p = 0.006) and PTS-anatomical (p = 0.020). Based on the maximum Youden indexes, the cutoff values of PTS-mechanical, PTS-anatomical, and LTS were 10.1° (sensitivity, 64.3%; specificity, 78.6%), 12.0° (sensitivity, 71.4%; specificity, 71.4%), and 7.7° (sensitivity, 85.7%; specificity, 80.4%), respectively. CONCLUSION: Due to the morphological asymmetry of the MTP and LTP, steep LTS measured on MRI is the best radiological predictor of ACLR failure. Detailed measurement of the LTS on MRI is recommended to evaluate the risk of ACLR failure prior to the surgery. LEVEL OF EVIDENCE: III.
PURPOSE: To identify the radiological predictive risk factors for anterior cruciate ligament reconstruction (ACLR) failure, compare the diagnostic accuracies of different parameters of conventional radiographs and magnetic resonance imaging (MRI), and determine the cutoff values for patients at higher risk. METHODS: Twenty-eight patients who were diagnosed as ACLR failure via MRI or arthroscopic examination were included in the study group. They were matched to 56 patients who underwent primary ACLR with the same surgical technique and without graft failure at the minimum 24-month follow-up by age, sex, and body mass index. On true lateral whole-leg radiographs, the posterior tibial slope (PTS) referenced to the tibial mechanical axis (PTS-mechanical), PTS referenced to the tibial proximal anatomical axis (PTS-anatomical), and anterior tibial translation (ATT) were measured. On the sagittal slices of MRI, the medial tibial slope (MTS), medial tibial plateau (MTP) subluxation (MTPsublx), lateral tibial slope (LTS), and lateral tibial plateau (LTP) subluxation (LTPsublx) were obtained. Receiver operator characteristic (ROC) curves were constructed to compare the diagnostic performance and determine the cutoff values of different radiological parameters. RESULTS: The study group demonstrated higher values of PTS-mechanical (10.7° ± 2.9° vs 8.7° ± 1.9°, p = 0.003), PTS-anatomical (13.2° ± 2.8° vs 10.5° ± 2.5°, p < 0.001), ATT (10.7 ± 3.3 mm vs 8.9 ± 2.2 mm, p = 0.014), LTS (9.4° ± 2.1° vs 5.5° ± 2.5°, p < 0.001), and LTPsublx (8.2 ± 2.8 mm vs 6.8 ± 1.9 mm, p = 0.009) as compared with the control group. The area under the ROC curve of LTS was significantly larger than that of PTS-mechanical (p = 0.006) and PTS-anatomical (p = 0.020). Based on the maximum Youden indexes, the cutoff values of PTS-mechanical, PTS-anatomical, and LTS were 10.1° (sensitivity, 64.3%; specificity, 78.6%), 12.0° (sensitivity, 71.4%; specificity, 71.4%), and 7.7° (sensitivity, 85.7%; specificity, 80.4%), respectively. CONCLUSION: Due to the morphological asymmetry of the MTP and LTP, steep LTS measured on MRI is the best radiological predictor of ACLR failure. Detailed measurement of the LTS on MRI is recommended to evaluate the risk of ACLR failure prior to the surgery. LEVEL OF EVIDENCE: III.