Do preoperative factors and implant design features influence humeral stem extraction efforts?

• Published in: Journal of shoulder and elbow surgery 2022-07, Vol.31 (7), p.1515-1523
• Main Authors: Gauci, Marc-Olivier, Diaz, Miguel A., Christmas, Kaitlyn N., Simon, Peter, Frankle, Mark A.
• Format: Article
• Language: English
• Subjects: humeral stem; revision shoulder surgery; Stem extraction; TSA
• ISSN: 1058-2746; 1532-6500
• DOI: 10.1016/j.jse.2021.12.029

Variations in humeral component designs in hemiarthroplasty and anatomic total shoulder arthroplasty cases can impact the degree of difficulty during a revision surgery that necessitates the removal of the humeral stem. However, no metric exists to define stem extraction effort nor to identify associated factors that contribute to extraction difficulty. The purpose of this study is to describe a method to quantify stem extraction difficulty and to define features that will impact the effort during stem removal. This was a retrospective review of 58 patients undergoing revision of hemiarthroplasty or anatomic total shoulder arthroplasty requiring stem extraction. Each included patient had existing preoperative radiographic examination, an intraoperative video of the stem removal process, and explants available for analysis by 3 surgeons. The following factors were assessed for the impact on extraction difficulty: (1) preoperative features such as cement use, fill of proximal humerus, and stem design features; (2) intraoperative data on extraction time and bone removal; and (3) postoperative findings related to extraction artifacts (EAs). A scoring system was established to distinguish easy (Easy group) and difficult (Difficult group) stem removal cases and further used to identify the features that may affect intraoperative difficulty of stem removal. The Difficult group accounted for 26% (15/58) of the study population with an 18-minute average stem extraction time, average EA count of 69, and 35 mm of bone removed. The Easy group accounted for 74% (43/58) of patients, with a 4-minute average extraction time, average EA count of 23, and 10 mm of bone removed. Logistic regression model was able to correctly classify 82% of the cases, explaining 26.7% of the variance in humeral stem removal with cement and proximal coating variables. The likelihood of cemented stem removal being difficult is 5 times greater compared to an uncemented stem, and having proximal coating doubles the likelihood of a difficult stem removal compared to cases with no coating. Quantifying stem extraction difficulty is possible with intraoperative video as well as explant analysis. Preoperative features of the fixation type and specific features of stem design such as proximal coating will impact difficulty of stem extraction.