Acromion and scapular spine fractures following reverse total shoulder arthroplasty: operative versus nonoperative management

Background

The reported incidence of acromial and scapular spine fractures (ASFs) following reverse total shoulder arthroplasty (rTSA) is approximately 5%. There is no consensus on whether to treat ASFs operatively or nonoperatively. This study reviewed patients diagnosed with ASFs following rTSA to determine which patients are successfully managed nonoperatively versus those who undergo surgery. We hypothesized that nonoperative and operative treatment results in equivalent outcomes.

Methods

A database of prospectively enrolled patients was reviewed to identify patients who underwent rTSA and were diagnosed with a postoperative ASF from 2015 to 2022 with a minimum 2-year follow-up (n = 19). Nonoperative treatment was initially attempted on all patients. Patients were separated into operative (n = 7) and nonoperative (n = 12) groups based on treatment. Demographic information and clinical outcomes were compared between patients. ASF patients were then matched 1:1 to a group of nonfracture rTSA patients on age (±12 months), sex, body mass index (±2.0), follow-up duration (±6 months), and whether the dominant arm was affected, yielding an ASF group (n = 19) and a nonfracture control group (n = 19). Demographic information and clinical outcomes at last follow-up were compared between groups. Clinical outcomes included American Shoulder and Elbow Surgeons (ASES) score, Subjective Shoulder Value, Visual Analog Scale (VAS) pain, and range of motion (ROM).

Results

Patients with Levy type I fractures were more likely to be treated operatively (71.4% vs. 8.3%; P = .013). Patients with Levy type I fractures had higher VAS pain scores than those with type II or III fractures (4.7 vs. 1.8 vs. 1.0; P = .014). Patients with fractures lateral to the glenoid articular surface had lower ASES scores (57.7 vs. 71.8; P = .05) and higher VAS pain scores (3.7 vs. 1.2; P = .009) than patients with nonlateral ASFs. Regardless of operative or nonoperative treatment, ASF patients had lower ASES scores (64.3 vs. 88.2; P < .001), lower Subjective Shoulder Values (69.0 vs. 85.8; P < .001), and higher VAS pain scores (2.5 vs. 0.8; P = .007) compared to nonfracture controls. ASF patients had inferior ROM, including less flexion/extension, (131 vs. 156; P < .001), less external rotation (46 vs. 58; P = .024), and lower internal rotation scores (4.7 vs. 5.9; P < .001) than nonfracture controls.

Conclusion

The development of ASFs, regardless of treatment modality, was associated with inferior outcomes and ROM compared to nonfracture rTSA patients. Type I fractures were more likely to fail nonoperative treatment than medial fractures. Lateral fractures had inferior outcomes compared to medial fractures. Fall prevention and adequate lifting restrictions may significantly lower the incidence of ASF.