Reliability of a CT reconstruction for preoperative surgical planning in the arthroscopic Latarjet procedure.

PURPOSE: The arthroscopic Latarjet procedure has provided reliable results in the treatment of anterior shoulder instability. However, this procedure remains technically challenging and is related to several complications. The morphology of the coracoid and the glenoid are inconsistent. Inadequate coracoid and glenoid preparing may lead to mismatching between their surfaces. Inadequate screws lengthening and orientation are a major concern. Too long screws can lead to suprascapular nerve injuries or hardware irritation, whereas too short screws can lead to nonunions, fibrous unions or migration of the bone block. The purpose of the study was to investigate the application of virtual surgical planning and digital technology in preoperative assessment and planning of the Latarjet procedure.

METHODS: Twelve patients planned for an arthroscopic Latarjet had a CT scan evaluation with multi-two-dimensional reconstruction performed before surgery. Interobserver and intraobserver reliability were evaluated. The shape of the anterior rim of the glenoid and the undersurface of the coracoid were classified. Coracoid height was measured, respectively, at 5 mm (C1) and 10 mm (C2) from the tip of the coracoid process, corresponding to the drilling zone. Measurements of the glenoid width were then taken in the axial view at 25 % (G1) and 50 % (G2) of the glenoid height with various α angles (5°, 10°, 15°, 20°, 25°, 30°) 7 mm from the anterior glenoid rim. Shapes of the undersurface of the coracoid and the anterior rim of the glenoid were noted during the surgical procedure. Post-operative measurements included the α angle.

RESULTS: Concerning coracoid height measurements, there was an almost perfect to substantial intra- and inter-reliability, with values ranging from ICC = 0.75-0.97. For the shape of the coracoid, concordances were, respectively, perfect (ICC = 1) and almost perfect (0.87 [0.33; 1]) for the intra- and interobserver reliabilities. Concerning the glenoid, concordance was always almost perfect for 50 % height. Concordance was almost perfect for 25 % height 15° and 30° for inter- and intraobserver, for intraobserver at 0° and 25°. All the other values were still showing moderate concordance. Shape of the coracoid analysis reproducibility was perfect for both intra- and interobserver ICC = 1. There was a total agreement (ICC = 1) between the preoperative evaluation of the shape of the glenoid and the coracoid and the intraoperative assessment.

CONCLUSION: The ideal and accurate preoperative planning of screwing of the coracoid graft in the arthroscopic Latarjet can be achieved in the real surgery assisted by the virtual planning. The clinical importance of this study lies in the observation that this new preoperative planning could offer a simple, effective and reproducible tool for surgeons helping them to prepare in the best possible way a technically challenging procedure usually associated with a high rate of complications.