Nonspherical humeral arthroplasty increases internal rotation: a biomechanical comparison of the native humeral head to nonspherical and spherical humeral implants

Abstract

BackgroundNonspherical humeral head implants more closely resemble native humeral anatomy than spherical components and may better replicate native shoulder range of motion (ROM) and kinematics. The purpose of this study was to compare shoulder ROM and kinematics of a commercially available nonspherical humeral head implant with the native humeral head and a height matched, custom manufactured spherical implant. MethodsSix fresh frozen cadaveric shoulder specimens were used with a custom shoulder testing system. The native shoulder was tested in multiple positions under anatomic muscle loading. Each specimen was tested for ROM and glenohumeral joint kinematics by measuring the humeral head apex and humeral head center (HHC) translation per degree of rotation using a MicroScribe digitizer. Measurements were then repeated after implantation of a spherical and, subsequently, a nonspherical humeral head prothesis. ResultsThe nonspherical implant had significantly more internal rotation (IR) compared to the spherical implant at 0° abduction (10.6 ± 6.2° more IR, P = .004, 95% confidence interval [CI]: −13.3, 34.5), 30° abduction (5.7 ± 2.8°, P = .009, 95% CI: −12.6, 24.0) and 60° abduction (6.8 ± 2.7°, P = .002, 95% CI: −8.3, 22.1) in the scapular plane, and 60° abduction (6.9 ± 2.0°, P = .031, 95% CI: −12, 25.6) in the coronal plane. The nonspherical implant had more IR than the native head at 60° of abduction in the scapular plane (7.0 ± 2.2° P = .002, 95% CI: −10.3, 24.3). The spherical head had less IR than the native head at 0° abduction (7.2 ± 4.8°, P = .031, 95% CI: 32.5, 18.1). There were no differences in humeral head apex translation per degree of rotation noted between the spherical implant or nonspherical implant and the native shoulder. The nonspherical head had less HHC translation than the native shoulder at 30° abduction in the forward flexion plane (P = .007); otherwise, there were no statistically significant differences in HHC translation between the native shoulder, the spherical head, and the nonspherical head. There was no significant difference observed between the average difference in anterior-posterior and superior-inferior radius of curvature of the nonspherical implants (2.0 ± 0.7 mm) and the native humeral heads (1.9 ± 1.3 mm) [P = .926]. ConclusionThe results of this biomechanical study suggest that the commercially available nonspherical humeral head has improved IR when compared to a custom, height controlled spherical implant and offers ROM and kinematics similar to the native humeral head in a cadaveric model.