Objectives: Anterior cruciate ligament (ACL) repair has regained interest in recent years as it has been demonstrated to provide favorable outcomes in patients with proximal ACL tears or avulsions. Magnetic resonance imaging (MRI) is the primary method uses to pre-operatively evaluate the ACL injury and help decide whether to perform an ACL repair or reconstruction. However, intraoperative findings often do not correspond with the MRI findings. This discrepancy may cause the surgeon to deviate from their initial surgical plan. Furthermore, some surgeons may disregard the possibility of performing an ACL repair solely due to the MRI report. Thus, potential candidates for ACL repair may be ignored due to the MRI report. This study aims to evaluate the reliability of MRI in predicting the degree of ACL tears planning for surgical treatment. Methods: We reviewed patients that underwent ACL repairs by a single surgeon at our institution between June 2015 to August 2018. All patients were initially seen for an ACL injury and were consulted for both ACL reconstruction and ACL repair. The final treatment decision was made following diagnostic arthroscopy. An ACL repair was performed only if the patient had a proximal tear, an intact but vertical ACL graft from a previous ACLR, or a tibial spine avulsion. Any patients that were missing MRI reports or operative reports were excluded. Pre-operative MRI of sagittal views of the ACL were then reviewed by a single radiologist and rated on integrity of the ligament’s midsubstance, and its tibial and femoral attachments. Intraoperative reports were then compared with the MRI reports. Results: We identified 40 consecutive ACL repairs of which 25 were included. Fifteen patients were excluded due to missing either operative or MRI reports. Of the included patients, 16 were males (64%) and the mean ± SD of age was 33.5±9.5 years (range 14-50 years). Two patients also had previous ACL reconstructions that failed and were subsequently revised with an ACL repair. MRIs demonstrated 15 full thickness tears with 9 torn at midsubstance, 2 torn at the tibial attachment, and 4 torn at the femoral attachment. Nine were reported as partial tears, with 4 torn at midsubstance, 2 torn at tibial attachment, and 3 torn at femoral attachment. One MRI reported no tear present. Intraoperatively, the surgeon identified 11 full-thickness tears, 11 partial tears, 2 tibial spine avulsions, and 2 incidences where a previous ACL reconstruction had failed due to a vertical graft. All of the full-thickness and partial tears were at the femoral attachment. The overall MRI sensitivity for the presence of an ACL injury was 96% with 2 false positives (8%), both of which revisions for failed ACLR. The overall accuracy was 32%, with specificity for tear-degree and tear-location 44% and 36%, respectively. Conclusion: Our results suggest that MRI may not be completely reliable in assessing the degree and location of an ACL tear. Furthermore, our results demonstrate that some patients who had repairable ACLs had MRI reports that suggested otherwise. Thus, our findings suggest that the surgeon should not solely rely on MRI imaging for ruling out possible ACL repair. This may also suggest that diagnostic tools, such as in-office diagnostic arthroscopy, be used in conjunction with MRI to more accurately assess the nature of an ACL tear. Future prospective studies with a larger sample size are necessary to confirm our findings that MRI may be less reliable for assessing ACL repair candidates.
Evidence too weak to guide surgical treatment decisions for anterior cruciate ligament injury: a systematic review of the risk of new meniscal tears after anterior cruciate ligament injury
