Spatial mapping of humeral head bone density

Abstract

Background: Short-stem humeral replacements achieve fixation by anchoring to the metaphyseal trabecular bone. Fixing the implant in high-density bone can provide strong fixation and reduce the risk of loosening. However, there is a lack of data mapping the bone density distribution in the proximal humerus. The aim of the study was to investigate the bone density in proximal humerus.

Methods: Eight computed tomography scans of healthy cadaveric humeri were used to map bone density distribution in the humeral head. The proximal humeral head was divided into 12 slices parallel to the humeral anatomic neck. Each slice was then divided into 4 concentric circles. The slices below the anatomic neck, where short-stem implants have their fixation features, were further divided into radial sectors. The average bone density for each of these regions was calculated, and regions of interest were compared using a repeated-measures analysis of variance with significance set at P < .05.

Results: Average apparent bone density was found to decrease from proximal to distal regions, with the majority of higher bone density proximal to the anatomic neck of the humerus (P < .05). Below the anatomic neck, bone density increases from central to peripheral regions, where cortical bone eventually occupies the space (P < .05). In distal slices below the anatomic neck, a higher bone density distribution in the medial calcar region was also observed.

Conclusion: This study indicates that it is advantageous with respect to implant fixation to preserve some bone above the anatomic neck and epiphyseal plate and to use the denser bone at the periphery.

Keywords: Shoulder arthroplasty; humeral bone density; humeral component; implant design; implant fixation; implant loosening; short-stem devices.

Figure 1

Diagram of the most common short-stem humeral components. The hemispherical head is assembled to a variety of stem designs shown in the figure using a taper fit mechanism. The stem is press fitted into the cancellous bone beneath the anatomic neck cut.

Figure 2

Variation of bone density from the proximal to distal region across slices parallel to the anatomic neck. The range and the orientation of the slices are shown, with the anatomic neck at slice 6.

Figure 3

Variation of bone density from central to peripheral zones for each slice parallel to the anatomic neck. The range and the orientation of the slices are shown, with the anatomic neck at slice 6.

Figure 4

Graph of mean bone density stratified by concentric zone (1-4) and radial sectors (A-F). Asterisks and brackets represent significant comparisons (P < .05) between radial sectors for a given concentric zone. Also, note that data are for the average of the 6 slices distal to the anatomic neck.

Figure 5

Graph of mean bone density stratified by proximal to distal (proximal, middle, distal) position and concentric zone (1-4). Asterisks and brackets represent significant comparisons (P < .05) between concentric zones for a given proximal to distal region. Also, note that slice data were grouped and averaged into distal, middle, and proximal regions to make comparisons more clinically meaningful as described in the Statistics section.