Three-dimensional measures of posterior bone loss and retroversion in Walch B2 glenoids predict the need for an augmented anatomic glenoid component.

Hypothesis: The purpose of this study is to evaluate whether the amount of measured posterior bone loss on 2- and 3-dimensional (2D and 3D) imaging of Walch B2 glenoids can reliably predict the plan for an augmented anatomic glenoid component.

Methods: Patients with Walch B2 glenoids and preoperative computed tomography (CT) scans were retrospectively identified. 2D axial CT scans were reviewed and posterior bone loss was measured by 3 independent reviewers. Images were then formatted into BluePrint (Wright Medical) preoperative planning software. The same 3 reviewers again measured posterior bone loss on 3D imaging. Additionally, all cases were planned with BluePrint software. An augment was used when the following criteria were unable to be satisfied with standard implants: <10° retroversion, <10° superior inclination, ≥90% backside contact, <2 mm medial reaming, and ≤1 peg perforation.

Results: Forty-two patients were included in the final analysis with a mean age of 63.1 ± 6.3 years. As measured by BluePrint, the mean retroversion was 23° ± 7° (range = 9°-40°), the mean superior inclination was 5° ± 6° (range = -9° to 22°), and the mean posterior subluxation was 80% ± 17% (range = 41%-95%). The mean 2D bone loss measurements (3.5 ± 1.6 mm) were significantly lower than the mean 3D bone loss (4.0 ± 1.8 mm) measurements (P = .03). There was substantial agreement between reviewers on both 2D and 3D measurements with an interclass correlation of 0.815 (95% confidence interval [CI] 0.714-0.889, P < .001) and an interclass correlation of 0.802 (95% CI 0.683-0.884, P < .001), respectively. Augments were used in 73.8%, 63.4%, and 63.4% of cases by reviewers 1, 2, and 3, respectively, with moderate agreement with a Fleiss kappa of 0.592 (95% CI 0.416-0.769, P < .001). Augment size was moderately, positively correlated with the amount of bone loss on 3D imaging but not with 2D imaging. After multivariate logistic regression, both 3D bone loss and retroversion were found to be predictive for a plan to use an augment.

Conclusion: Planning for a posterior augment in Walch B2 glenoids is better predicted with 3D imaging than with 2D imaging, as 2D imaging may underestimate posterior bone loss. Additionally, use of a larger augment size is moderately correlated with posterior bone loss on 3D imaging but not 2D imaging. Standard 2D imaging may be limited in cases of posterior bone loss, and 3D imaging may be beneficial for preoperative planning in Walch B2 glenoids.