Selective bundle anterior cruciate ligament (ACL) reconstruction and/or remnant ACL preservation may be reasonable options for some patients. However, the frequency of isolated anteromedial (AM) or posterolateral (PL) bundle injuries in patients undergoing ACL reconstruction is unknown, and the value of MRI for prediction of this injury pattern is likewise unknown.Questions/purposes
We sought to determine (1) the proportion of knees with an intact AM or PL bundle in patients undergoing ACL reconstruction; (2) whether MRI predicted the bundle conditions seen at the time of surgery; and (3) whether the accuracy of the MRI prediction was affected by the timing of MRI after injury.Methods
During primary ACL reconstructions of 156 knees, conditions of AM and PL bundles were separately examined and classified into three categories: (1) completely torn; (2) attenuated; and (3) intact. Then, the bundles were assessed by blinded observers on MRI and classified into the corresponding three categories for 77 patients who had an MRI at our institution using a standard protocol. Diagnostic accuracy of MRI was computed, and the early MRI group (≤ 6 weeks from injury to MRI acquisition) was compared with the late MRI group (> 6 weeks).Results
Only 11 (7%) of the 156 knees we treated had an intact AM (one knee) or PL bundle (10 knees). Another 55 knees (35%) had a structurally continuous but attenuated AM or PL bundle. The overall diagnostic accuracy of MRI was 83%; accuracy was better for the AM bundle than the PL bundle (91% versus 78%; p = 0.026). MR prediction was less accurate in the early MRI group, particularly for PL bundle injury.Conclusions
An isolated bundle tear is uncommon in patients with ACL tears undergoing reconstruction. MRI can help surgeons predict bundle injury pattern with satisfactory precision, but caution should be used in predicting PL bundle injury using MRI with early acquisition time from injury.Level of Evidence
Level III, diagnostic study. See the Guidelines for Authors for a complete description of levels of evidence.