Zone specific bone density evaluation of the acromion may predict postoperative acromion stress fracture in patients undergoing a reverse total shoulder arthroplasty.

Abstract

Background: The goal of this study was to utilize preoperative computed-tomography(CT) scans to identify differences in the Hounsfield units(HU) of the acromion in patients who did and did not develop a postoperative acromial and scapular-spine fracture(ASF) after primary reverse total shoulder arthroplasty (rTSA).

Methods: A retrospective analysis was performed at a single institution. All patients undergoing a rTSA with either a 135° neck/shaft angle(NSA) humeral inlay design combined with a lateralized center-of-rotation(COR) glenosphere or a 145° NSA onlay combined with a medialized COR glenosphere design between 2011-2021 with a minimum follow-up of 24-months were included. Demographic characteristics and clinical outcome metric scores were recorded. Preoperative CT scans were analyzed to obtain acromion trabecular bone density measurements in HU in each zone of the scapula based on the Levy classification. Radiographic parameters were evaluated to determine their association with ASF.

Results: In total 263-patients were included, 140-patients with a 135° NSA humeral-inlay design;123-patients with a 145° NSA humeral-onlay design. There were no significant differences in baseline demographics between cohorts. The rate of ASF was 6.4%(9/140) for the 135° NSA-inlay-design versus 2.4%(3/123) in the 145° NSA-onlay design. In the non-fracture cohort there was a linear increase in bone density from zone-1(173.9HU)→zone-3(396.5HU)(lateral→medial). In the fracture cohort there was a decrease in bone density from zone-1(282.6HU)→zone-3(154.5HU). Measuring preoperative bone density in all Levy specific fracture-zones resulted in an AUC of 0.96 correlating to excellent predictive value. A threshold cutoff of 99.9 resulted in a sensitivity of 91.6% and specificity of 75.3%. A HU of 99.9 in any of the three-zones resulted in OR 5.1(p<0.0001) for sustaining an ASF postoperatively. A threshold of<50HU was associated with an 8-times higher-likelihood of developing a fracture in that specific zone. Greater than 5° of superior tilt in combination with ≥24mm of distalization was associated with an OR 6.4(p=0.0004) of sustaining an ASF.

Conclusion: The current study demonstrates an accurate method of measuring HU at each of the described Levy fracture zones with excellent predictability of patients who are at risk of an ASF following rTSA. Additionally, we found that a HU threshold of <50 HU at any of the three Levy zones was associated with a nearly 8 times higher likelihood of developing a fracture in that specific zone. Lastly, we found that >5° of superior tilt in combination with ≥24mm of distalization was associated with 6.4 times higher likelihood of sustaining an ASF agnostic to prosthesis design.