Glenoid "Pear Restoration" with the Traditional Latarjet versus Congruent Arc Modification: CT Assessment of Coracoid Morphology for Preoperative Planning

OBJECTIVES: The concept of the “inverted pear” glenoid was introduced by Lo and Burkhart as an indicator of significant attritional glenoid bone loss that predisposes patients to recurrent instability and higher failure of arthroscopic Bankart repair. Coracoid transfer (Latarjet) is often performed in this setting, and evaluation of coracoid dimensions may improve patient-specific preoperative planning. The purpose of this study was to utilize computed tomography (CT) to determine a reproducible method for coracoid measurement and develop a preoperative planning algorithm for glenoid pear restoration using the traditional Latarjet technique or congruent arc modification (CAM). We hypothesized that classic Latarjet technique would sufficiently restore at least 100% of the glenoid diameter, up to 30% of anterior bone loss. We also hypothesized that most female patients would have a coracoid thickness <10 mm, which may preclude the use of the CAM due to a risk of graft fracture with screw insertion. METHODS: Multiplanar reconstructed (MPR) shoulder CT scans were reviewed for patients aged 18-45. This age group was selected to capture an adult population that typically undergoes anterior shoulder stabilization surgery. Patients were excluded if CT scan demonstrated glenoid or humeral osteophyte formation, glenoid dysplasia, coracoid fracture, or tumor involving the glenoid or coracoid. All CT scans were obtained with the arm at the patient’s side, palm facing up, and the shoulder in external rotation. CT scans were performed by acquiring thin-slice data from the volumetric CT in the axial plane with a Discovery CT750 HD scanner (GE Healthcare) and the following scan parameters: 0.625-mm slice thickness, 0.313-mm slice spacing, reconstruction matrix of 512 × 512 pixels, reconstruction field of view of 40 to 45 cm, and pitch factor of 0.516. All measurements were performed using a PACS system (Sectra Workstation, Sectra AB, Linkoping, Sweden) by a board-certified musculoskeletal radiologist and two orthopaedic surgeons. The coracoid base was aligned in the coronal plane, and the long axis of the coracoid was aligned in the axial and sagittal planes (Figure 1). The coracoid length (proximal to distal) was measured on the modified axial view as the distance from the base to the tip along the long axis. The width (medial to lateral) and thickness (anterior to posterior) were measured on the modified coronal view. Coracoid dimensions were compared by sex and correlated with patient height and body mass index (BMI). Best-fit circle technique was used to quantify glenoid diameter and bone loss required for complete graft apposition of a 20- or 25-mm long coracoid graft to the glenoid neck. The proportion of glenoid diameter able to be restored using the classic Latarjet technique and the CAM was calculated and compared. The proportion of patients in which at least 20%, 30%, 35% or 40% of glenoid diameter could be restored using the classic Latarjet technique and the CAM was determined and compared. Based on these findings, a treatment algorithm was proposed considering glenoid bone loss, glenoid diameter and coracoid dimensions. RESULTS: A total of 120 CT scans were reviewed, and three were excluded (two with osteophytes and one with severe glenoid dysplasia). Measurements were performed and dimensions recorded on 117 CT scans (69.2% males). Coracoid dimensions varied considerably among patients (length: 17.5-31.8mm, width: 9.1-20.5mm, thickness: 6.1-15.7mm, Table 1). There was a strong positive correlation between patient height and coracoid length (r=0.629, p=0.000) and a moderate positive correlation between patient height and coracoid thickness (r=0.408, p=0.000). Mean coracoid and glenoid dimensions were larger in males than females (all p<0.0001). When comparing males and females, a higher proportion of males had a coracoid thickness >10mm (84.0% vs 27.8%, p<0.00001). While there was no difference in the proportion of males compared to females with coracoid length >20 mm (96.3% vs 94.4%, p=0.65), a higher proportion of males had coracoid length >25mm (48.2% vs 13.9%, p=0.0009). Compared to the classic Latarjet technique, the CAM was able to restore more native glenoid diameter in both males and females (Males: 54.1±6.6% vs 39.1±4.9%, p<0.0001; Females: 54.7±7.4% vs 36.2±4.3%, p<0.0001, Table 2). When comparing techniques, there was no difference in the proportion of patients in which at least 30% glenoid bone loss could be fully restored (Males: 98.8% vs 97.5%, p=1.0; Females: 100% vs 91.7%, p=0.24). However, the CAM was able to restore at least 35% and at least 40% glenoid bone loss in a higher proportion of males (98.8% vs 79% and 98.8% vs 40.7%, respectively, both p<0.0001) and females (100% vs 61.1% and 100% vs 16.7%, respectively, both p=0.00001). The amount of glenoid bone loss required for complete coracoid graft apposition of a 20mm or a 25mm graft was significantly different when comparing males and females, with less bone loss required for males (14.4±2.8% vs 20.7±5.5% and 26.8±7.5% vs 41.4±9.4%, respectively, both p<0.0001). Intra- and inter-rater reliability was excellent for all coracoid dimensions (ICC≥0.95±0.050 for each). CONCLUSIONS: We describe a reliable method of measuring coracoid dimensions for preoperative planning of glenoid pear restoration. The traditional Latarjet technique reliably restores the glenoid AP diameter with bone loss of up to 30%. The majority of females have coracoid thickness <10mm which may increase the risk of graft fracture when using CAM. The decision to utilize the classic Latarjet technique or CAM considers each individual’s glenoid and coracoid dimensions with a goal of achieving at least 100% restoration of native glenoid diameter (Figure 2). Alternative techniques may be considered if native glenoid pear restoration cannot be achieved with the Latarjet.