Arthroscopic Repair of Small and Medium-Sized Bony Bankart Lesions

Abstract

Background: There has been no study about treatment guidelines for arthroscopic repair according to the size of bony Bankart lesions of less than 25% of the glenoid width. Purpose: To evaluate the results of arthroscopic repair for bony Bankart lesions managed with different repair techniques based on their size. Study Design: Case series; Level of evidence, 4. Methods: Between March 2005 and February 2009, 44 of 52 consecutive patients with bony Bankart lesions with a size of less than 25% of the entire glenoid were managed with an arthroscopic approach. Of those patients, 34 (77%) were available for outcome analysis at a minimum 24 months’ follow-up (mean, 34 months; range, 24-60 months). The size of the fragment was measured by computed tomography (CT) and classified as small (<12.5% of the inferior glenoid width) and medium (12.5%-25%). Sixteen lesions were classified as small (small group), and 18 were classified as medium (medium group). For small lesions, capsulolabral repair using suture anchors without excision of the bony fragment was performed. For medium lesions, anatomic reduction and fixation using suture anchors was performed, and the adequacy of reduction was assessed by CT postoperatively. The visual analog scale (VAS) for pain score and modified Rowe score for bony Bankart repair were compared and the postoperative recurrence rate investigated. Results: One patient from the small group (6.3%) and 1 patient without anatomic reduction of the bony fragment in the medium group (5.6%) experienced traumatic redislocations. The mean VAS score improved from 1.7 preoperatively to 0.5 at final follow-up, and the mean Rowe score improved from 59 to 91 (both P < .001). The mean postoperative Rowe scores increased from 58 to 92 in the small group and from 60 to 91 in the medium group (both P < .001). Residual joint incongruity measuring ≤2 mm on both axial and coronal scans, which was considered an anatomic reduction, was present in 14 cases (77.8%) in the medium group. In the medium group, the mean postoperative Rowe scores increased from 60 to 95 in cases of anatomic reduction compared with an increase from 56 to 76 in cases of nonanatomic reduction. The Rowe score was statistically correlated with anatomic reduction of medium-sized bony fragments (P = .046). Conclusion: In small Bankart lesions, restoration of capsulolabral soft tissue tension alone may be enough, whereas in medium lesions, the osseous architecture of the glenoid should be reconstructed for more functional improvement and less pain.