Incidence of radiographic humeral bone remodeling after reverse total shoulder arthroplasty and its impact on clinical outcome.

BACKGROUND

Reverse total shoulder arthroplasty (rTSA) has proven to be a reliable surgical treatment option for a wide variety of shoulder pathologies. Bony remodeling around the humeral component occurs frequently but its impact on clinical outcome remains unclear.

METHODS

We retrospectively reviewed 102 patients who underwent primary rTSA between 2008 and 2020. At final follow-up (FU), patients underwent standardized radiographs and clinical examination, including patient-reported outcome measures and the adjusted Constant Murley score (aCMS). Serial radiographs (pre-operative, 6 weeks and ≥2 years post-operative) were evaluated for: bone remodeling (cortical thickness, pedestal formation, spot welds, reactive lines, osteolysis, radiolucent lines) and implant stem parameters (subsidence, size, alignment and filling ratio). Stress shielding was quantitatively defined as a decrease in combined cortical thickness measured at four sites (2 medial [M1 and M2], 2 lateral [L1 and L2]). Subgroup analysis was conducted to determine correlations between the radiographic findings and clinical outcome. The relationship between stress shielding severity and functional outcome was investigated by using percent-based thresholds to explore potential cutoff values.

RESULTS

In 102 patients (mean age 70 years, 71% female, mean FU 65 months), the mean aCMS was 80. Stress shielding occurred in 74%, with combined cortical thickness decreasing from 3.0 to 2.2 mm at final FU; proximal lateral thinning (L1) was most pronounced. A ≥10% cortical thinning threshold at L1 is associated with a lower aCMS (77 vs. 86, P = .051). Stem-related parameters, including filling ratio, did not correlate with cortical thinning. Radiographic bone remodeling such as spot welds (36%), reactive lines (49%), and osteolysis (25%) were frequently observed but have no correlation with clinical outcome.

DISCUSSION: This study assessed the long-term radiographic evaluations of humeral bone remodeling after rTSA. The quantitative assessment of cortical thickness showed that the proximal lateral humerus is most affected by stress shielding. A >10% cortical thinning at this site showed a tendency toward inferior clinical outcome, contrasting with prior reports suggesting no functional impact. Other radiographic bone remodeling was common but did not impair function. Our findings support the standardize assessment of cortical resorption with special attention to the region of the proximal lateral humerus during patient FU.