Determination of humeral head size in anatomic shoulder replacement for glenohumeral osteoarthritis

Abstract

Background: We hypothesized that a sphere mapped to specific preserved nonarticular landmarks of the proximal humerus can accurately predict native humeral head radius of curvature (ROC) and head height (HH) in the osteoarthritic, deformed humeral head.

Methods: Three consistent nonarticular landmarks were defined with a 3-dimensional sphere (and 2-dimensional circle in midcoronal plane) placed along the articular surface in 31 normal cadaveric humeri. Side-to-side differences in ROC and HH were determined in 22 pairs of normal shoulders. Using the nonarticular landmarks and sphere method, 3 independent blinded observers performed 2 sets of measurements in 22 pairs of shoulders with unilateral glenohumeral osteoarthritis. The predicted native ROC and HH in the pathologic shoulder were compared with the normal side control.

Results: The mean side-to-side difference in normal shoulders was 0.2 mm (ROC) and 0.6 mm (HH). In the unilateral osteoarthritis cases, the intraobserver mean differences for the normal side were 0.3 mm (ROC) and 0.9 mm (HH). The pathologic side ROC and HH, defined by the sphere, exhibited intraobserver differences of 0.5 mm (ROC) and 1.0 mm (HH). The mean side-to-side differences between the normal and pathologic sides were 0.5 mm (ROC) with concordance correlation coefficient of 0.95 and 1.3 mm (HH) with concordance correlation coefficient of 0.66.

Conclusion: A sphere mapped to preserved nonarticular bone landmarks can be used for accurate preoperative measurement of premorbid humeral head size and therefore the selection of an anatomically sized prosthetic head. This is applicable postoperatively, as is a circle method for 2-dimensional assessment of anatomic humeral reconstruction in the coronal plane.

Keywords: 3D CT scan; Anatomic shoulder replacement; glenohumeral arthritis; humeral head size; humerus templating; preoperative planning; shoulder arthroplasty; sphere model.