Radiographic Analysis of Glenoid Size and Shape After Arthroscopic Coracoid Autograft Versus Distal Tibial Allograft in the Treatment of Anterior Shoulder Instability

Abstract

Background: The Latarjet procedure for autograft transposition of the coracoid to the anterior rim of the glenoid remains the most common procedure for reconstruction of the glenoid after shoulder instability. The anatomic glenoid reconstruction using distal tibial allograft has gained popularity and is suggested to better match the normal glenoid size and shape. However, concerns about decreased healing and increased resorption arise when an allograft bone is used. Purpose: To use radiological findings to evaluate the arthroscopic reconstruction of the glenoid with respect to the size, shape, healing, and resorption of coracoid autograft versus distal tibial allograft. Study Design: Cohort study; Level of evidence, 3. Methods: A retrospective review was performed of 48 consecutive patients who had an arthroscopic bony reconstruction of the glenoid (12 coracoid autograft, 36 distal tibial allograft), diagnosed anterior shoulder instability, and computed tomography (CT)–confirmed glenoid bone loss more than 20%. Coracoid autograft was performed only when tibial allograft was not accessible from a bone bank. Two fellowship-trained musculoskeletal radiologists reviewed pre- and postoperative CT scans at a minimum follow-up of 6 months for the following: graft position, glenoid concavity, cross-sectional area, width, version, total area, osseous union, and graft resorption. Clinical outcome was noted in terms of instability, subluxation, and dislocation at a minimum follow-up of 2 years. Simple logistic regression, 2-tailed independent-sample t tests, paired t tests, and Fisher exact tests were performed. Results: Graft union was seen in 9 of the 12 patients (75%) who had coracoid autograft and 34 of the 36 patients (94%) who had tibial allograft (odds ratio, 5.66; 95% CI, 0.81-39.20; P =.08). The odds ratio comparing allograft to coracoid for overall resorption was 7.00 (95% CI, 1.65-29.66; P =.008). Graft resorption ≥50% was seen in 3 (8%) of the patients who had tibial allograft and none of the patients who had coracoid autograft. Graft resorption less than 50% was seen in the majority of patients in both groups: 27 (73%) patients with tibial allograft and 5 (42%) patients with coracoid autograft. No statistically significant difference was found between the 2 procedures regarding anteroposterior diameter of graft (P =.81) or graft cross-sectional area (P =.93). However, a significant difference was observed in step formation between the 2 procedures (P