Factors associated with humeral stem extraction difficulty in revision shoulder arthroplasty

Removal of the humeral stem during revision shoulder arthroplasty can be challenging and is associated with complications. The primary objective of this study was to identify risk factors for intraoperative humerus fractures or the need for a humerotomy. The secondary objective was to determine factors associated with the use of a standard or long revision stem.

In this retrospective monocentric study, a total of 127 consecutive shoulder arthroplasty revisions (53 men, 74 women, mean age 68 ± 11 years) were included. All included patients underwent a complete revision of the humeral component. Preoperative periprosthetic fractures were excluded. Assessed risk factors included patient-related factors, the etiology and characteristics of the primary prosthesis, time between the primary arthroplasty and revision, and the etiology of the revision. All risk factors were evaluated through a multivariate analysis with logarithmic regression.

A humerotomy was required in 7.9% (10/127), while an iatrogenic humerus fracture occurred in 24.4% (31/127) of cases, resulting in a total of 32.3% (41/127). Factors associated with difficulties during stem extraction included chronic infection (P = .001), the presence of a long stem (P = .04), and the use of a cemented stem (P = .01). Age and gender were not associated. During revision, a short, standard-length, and long stem were implanted in 7.1% (9/127), 55.1% (70/127), and 37.8% (48/127), respectively. Factors associated with the use of a standard or long revision stem included increased age (P = .006), lower preoperative external rotation (P = .028), a standard (P < .0001) or long (P = .002) primary stem, a cemented primary stem (P = .006), a longer interval between primary arthroplasty and revision (P = .025), and the occurrence of a fracture or the need for a humerotomy (P = .016), particularly in cases of diaphyseal localization (P = .009).

Difficulties and complications extracting the humeral stem were encountered in 32.3% of cases in our series. Chronic infection as the cause of revision and the use of a long or cemented stem were significant associated factors.