How anatomic should anatomic total shoulder arthroplasty be? Evaluation of humeral head reconstruction with the best-fit circle.

ABSTRACT

BACKGROUND: Humeral implant designs for anatomic total shoulder arthroplasty (aTSA) focus on anatomic reconstruction of the articular segment. Likewise, the pathoanatomy of advanced glenohumeral osteoarthritis often results in humeral head deformity. We hypothesized the anatomic reconstruction of the humeral head in aTSA risks overstuffing the glenohumeral joint. METHODS: Ninety-seven cases (52 females) of primary glenohumeral osteoarthritis in patients treated with aTSA were evaluated. Preoperative computed tomography scans were used to classify glenoid morphology according to the Walch classification. Coronal plane images in the plane of the humerus were used to determine the anatomic best-fit circle as described by Youderian et al. Humeral head thinning was determined as the distance from the center of rotation of the best-fit circle to the nearest point along the humeral articular surface. aTSA was modeled with a predicted anatomic humeral head and a simulated 4-mm polyethylene glenoid component. The change in the position of the native humerus was determined. Wilcoxon Rank Sum tests were used to evaluate differences in humeral head thinning and humeral lateralization between monoconcave and biconcave glenoid morphologies. Spearman's rank correlation coefficients were used to assess the relationship between humeral head thinning with preoperative active forward elevation and external rotation. RESULTS: The mean radius of the best-fit circle was 25.0 ± 2.1 mm. There was a mean thinning of 2.4 ± 2.0 mm (range -1.7 to 8.3). The mean percent thinning of the humeral head was 9.4% ± 7.7%. The mean humeral lateralization was 6.4 ± 2.0 mm. Humeral head thinning was not significantly associated with active forward elevation (r = -0.15, P = .14) or active external rotation (r = -0.12, P = .25). There were no significant differences in the percentage of humeral head thinning (P = .324) or humeral lateralization (P = .350) between concentric and eccentric glenoid wear patterns. CONCLUSIONS: Utilization of the best-fit circle as a guide in aTSA may risk excessive lateralization of the humerus and overstuffing the glenohumeral joint. This may have implications for subscapularis repair and healing, as well as glenoid implant and rotator cuff longevity. These findings call into question whether recreation of normal glenohumeral anatomy in aTSA is appropriate for all patients. Humeral head reconstruction in aTSA should account for glenohumeral joint volume and soft tissue contracture.