Preoperative metaphyseal cancellous bone density is associated with intraoperative conversion to stemmed total shoulder arthroplasty

Abstract

Methods to determine whether a stemless humeral component is appropriate for anatomic total shoulder arthroplasty are varied and often subjective. Objective preoperative data regarding metaphyseal bone quality may help guide surgical decision-making. This study sought to evaluate preoperative proximal humeral bone quality and determine whether it is predictive of intraoperative conversion to a stemmed humeral component. Consecutive patients who underwent primary anatomic total shoulder arthroplasty from a single-surgeon practice were enrolled. All patients received a preoperative computed tomography (CT) scan for surgical templating purposes. The exclusion criteria were lack of a preoperative CT scan, preoperative plan for a stemmed component, and intraoperative conversion to a stem for a reason other than bone quality (ie, fracture). Preoperative CT scans were analyzed with an automated templating software. Cortical index and thickness were calculated, and bone density of the proximal diaphysis, cancellous metaphysis, and cortical metaphysis was obtained by averaging Hounsfield units (HU) across anatomically defined regions using a previously validated technique. The decision to convert to a stemmed humeral component was made intraoperatively based on a lack of stability of the trial stemless component. Bone quality measurements were compared between stemless and stemmed groups. An exact logistic regression was used incorporating gender and age. A total of 79 patients who underwent primary anatomic total shoulder arthroplasty were included in this study. Of these patients, 6 underwent intraoperative conversion to a stemmed humeral component (7.6%). There was no significant difference between cohorts in terms of cortical index and bone density within the proximal diaphysis and cortical metaphysis. On univariate analysis, cortical thickness, metaphyseal cancellous bone density, and gender were significantly different between groups. Patients receiving a stem had significantly lower metaphyseal cancellous bone density than those receiving stemless components (5.5 ± 11.2 HU vs. 47.6 ± 29.4 HU, P<.001). All patients converted to stems were female individuals (P= .01) and had metaphyseal cancellous bone density less than 20 HU (P<.001). Metaphyseal cancellous bone density can be calculated on preoperative CT scans and is associated with intraoperative conversion to a stemmed humeral component in anatomic shoulder arthroplasty. A threshold of 20 HU can be used to predict which patients are more likely to require stemmed components.