CT imaging findings in symptomatic patients with and without revision surgery after reverse shoulder arthroplasty

Sophia S. Goller, Anna L. Falkowski, Rainer J. Egli, Georg C. Feuerriegel, Samy Bouaicha, Reto SutterSkeletal 2025 · DOI

TL;DR

This study investigated CT imaging findings in patients experiencing shoulder pain after reverse shoulder arthroplasty (RSA) to identify predictors of revision surgery (RS). The key takeaway is that prosthesis dislocation was the only CT finding significantly associated with the need for RS, suggesting it's a critical indicator of impending revision.

What Problem Does This Paper Address?

Reverse shoulder arthroplasty (RSA) is a common procedure for treating severe shoulder arthritis and complex fractures. While generally successful, a significant number of patients experience complications requiring revision surgery (RS). Identifying which imaging findings on CT scans are most indicative of this need for revision is crucial for guiding clinical decisions and potentially intervening earlier. Currently, the clinical relevance of various CT findings in symptomatic patients after RSA, especially in distinguishing those who will require revision from those who won't, remains unclear.

Methods

This was a retrospective, single-center study. Two radiologists independently reviewed CT scans of 99 symptomatic patients who had undergone RSA. They assessed a comprehensive list of potential CT findings, including:

  • Material fracture and loosening: of the peg, baseplate, screws, and humeral stem.
  • Screw positioning: including medial protrusion of screw tips.
  • Prosthesis dislocation.
  • Glenoid notching: classified using the Nerot-Sirveaux system.
  • Fractures: specifically acromial stress fractures, coracoid fractures, other scapula fractures, and periprosthetic humeral fractures.
  • Deltoid muscle quality: assessed for volume atrophy and fatty infiltration using standardized classifications.

Patients were divided into two groups: Group 1 (underwent revision surgery, n=52) and Group 2 (did not undergo revision surgery, n=47). Statistical analysis, including chi-square tests and Mann-Whitney U tests, was used to compare findings between groups. Odds ratios were calculated to quantify associations, and a multivariate regression model was employed to account for confounding factors like age, sex, time since surgery, infection history, and symptom severity.

Datasets

Not discussed in the paper. The study utilized CT imaging data from 99 patients who underwent clinically indicated CT scans for symptom clarification after RSA between January 2015 and January 2020.

Key Insights and Ideas

  • Dislocation is King: The most striking finding is that prosthesis dislocation was the only CT imaging feature that significantly predicted the need for revision surgery. This highlights its critical importance as a sign of implant failure.
  • Subtle Trends: While not statistically significant, there was a notable trend for increased loosening of the peg, baseplate, and superior screw, as well as periprosthetic humeral fractures, in patients who eventually required revision. This suggests these issues, while not definitive predictors on their own, are more prevalent in the revision group.
  • Glenoid Notching is Ubiquitous: Glenoid notching was a very common finding in both groups, and interestingly, it was more prevalent in patients who did not undergo revision surgery. This challenges the idea that notching alone is a primary driver for revision.
  • Deltoid Muscle Quality Matters (Potentially): The study observed a trend towards higher volume atrophy in the lateral and posterior deltoid muscle in the revision group. This, along with a trend for increased fatty infiltration, suggests that muscle quality might play a role in RSA outcomes, though further research is needed.
  • Symptoms Trump Imaging (Sometimes): The multivariate analysis indicated that moderate to severe symptoms and a history of infection were significant predictors of revision, underscoring the importance of clinical presentation alongside imaging findings.

Key Results

  • Prosthesis Dislocation: Significantly more frequent in the revision group (19.2%) compared to the non-revision group (2.1%) (P=0.007).
  • Loosening Trends:
    • Peg loosening: 30% in revision vs. 16% in non-revision (P=0.49).
  • Baseplate loosening: 15% in revision vs. 6% in non-revision (P=0.16).
  • Superior screw loosening: 18% in revision vs. 7% in non-revision (P=0.10).
  • Periprosthetic Humeral Fractures: 29% in revision vs. 13% in non-revision (P=0.09).
  • Glenoid Notching: Present in 78.9% of the revision group and 93.6% of the non-revision group (P=0.035). Severe grade 4 notching was more frequent in the revision group (31% vs. 17%), but not statistically significant (P=0.11).
  • Deltoid Muscle Atrophy: Lateral third atrophy was more frequent in the revision group (17.3% vs. 4.3%, P=0.039) and posterior third atrophy (9.6% vs. 0%, P=0.029), though confidence intervals for odds ratios crossed 1.

Limitations and Future Work

  • Retrospective Design: The study's retrospective nature means clinical data was extracted from charts, which can introduce biases.
  • Selection Bias: Only symptomatic patients who underwent CT were included, limiting generalizability to asymptomatic patients or those not requiring CT.
  • Confounding Factors: While a confounder analysis was performed, it's acknowledged that surgical decisions are complex and influenced by many individual factors not fully captured in a retrospective study.
  • Small Sample Size for Specific Findings: Some specific findings (e.g., severe glenoid notching, certain fracture types) had small numbers, limiting statistical power.

Future work could focus on prospective studies to validate these findings and explore the role of deltoid muscle quality in more detail. The authors emphasize the importance of collaboration between radiologists and surgeons to integrate imaging findings with clinical presentation for optimal patient management.

Summary updated

Abstract

In this cohort of symptomatic patients after RSA, prosthesis dislocation was the only CT imaging finding associated with RS. Besides, there was a trend with higher numbers of loosening of the peg, baseplate, and superior screw, as well as periprosthetic humeral fractures in patients with RS, though not reaching significance.
BibTeX
@inproceedings{goller2025ct,
  author = {Sophia S. Goller and Anna L. Falkowski and Rainer J. Egli and Georg C. Feuerriegel and Samy Bouaicha and Reto Sutter},
  title = {{CT imaging findings in symptomatic patients with and without revision surgery after reverse shoulder arthroplasty}},
  booktitle = {Skeletal},
  year = {2025},
  doi = {10.1007/s00256-025-04867-9},
}
Figure