ABSTRACT
BACKGROUND: The design of humeral implants has evolved from the initial Grammont design, notably regarding the neck-shaft angle (NSA) and through the use of cementless prostheses. Stress shielding was frequently reported with cementless implants, the 2 main risk factors being humeral stem misalignment and high filling ratios. Our hypothesis was that standard length cementless stems with an NSA of 140° would facilitate good alignment with moderate filling ratios, thereby limiting stress shielding and ensuring good clinical outcomes. METHODS: A single-center retrospective study was conducted of patients who underwent reverse total shoulder arthroplasty between January 2015 and August 2017, with at least 2 years' follow-up. Clinical evaluations included range of motion measurements and Constant and subjective shoulder values scores. Frontal alignment was assessed in terms of the angle (α) between axes of the stem and of the humerus. Filling ratios were measured in the metaphysis of the humerus, in the diaphysis, and at the distal end of the stem and considered excessive above 0.7. Stress shielding was evidenced radiographically by the observation of medial cortical narrowing, medial metaphysis thinning, lateral metaphysis thinning or under-the-baseplate osteolysis. RESULTS: Eighty-two shoulders were included and 70 had radiographic follow-up data available. The mean patient age was 78 years and 63/81 patients (78%) were female. The mean follow-up time was 39 ± 7 months. The mean α angle was 1.4° ± 0.9° and was less than 5° in all cases. The mean metaphyseal, diaphyseal, and distal filling ratios were 0.61 ± 0.06, 0.70 ± 0.08, and 0.64 ± 0.09, respectively. The mean Constant score improved from 28 ± 11 preoperatively to 64 ± 14 at last follow-up. The mean subjective shoulder values score at last follow-up was 81 ± 12. Seventy-nine percent of patients (55/70) had at least 1 form of stress-shielding related, which were not associated with clinical outcomes, apart from lateral metaphysis thinning, which was associated with lower active anterior elevation (mean, 106° ± 30° vs. 126° ± 28°; P = .01) and lower Constant scores (mean, 56 ± 17 vs. 65 ± 14; P = .06). CONCLUSION: The use of cementless reverse shoulder prostheses with a NSA of 140° was associated with good clinical outcomes at 2 years' follow-up. The prosthesis stem was correctly aligned with the humeral axis and the filling ratios were <0.7 in all cases. Stress-shielding was common but, apart from lateral metaphysis thinning, this had no impact on clinical outcomes.
