Cementless shoulder arthroplasty Aramis: does humeral stress shielding depend on stem size?

Abstract

Background

Cementless total shoulder arthroplasty (TSA) preserves humeral bone stock and thus facilitates revision surgery. Less experienced surgeons intuitively tend to use large filling ratios to ensure good fixation and alignment of the humeral implant. The aim of this study was to radiographically quantify stress shielding at two years' follow-up in 2 groups of patients with the same cementless stem implanted 5 years apart. The main hypothesis was that lower filling ratios would not be associated with greater stem misalignment. The secondary hypothesis was that lower filling ratios would be associated with reduced humeral stress shielding.

Methods

This single-center retrospective study involved 2 groups of patients who underwent anatomic or reverse total shoulder arthroplasty (rTSA) for primary or secondary osteoarthritis or massive rotator cuff tear, with the same Aramis humeral implants, either in 2013-14 (group 1, N = 59) or in 2019 (group 2, N = 47). Coronal stem alignment and filling ratios were measured on immediate postoperative and last follow-up radiographs, at a mean follow-up of two years. Stress shielding was evaluated in terms of medial cortical narrowing, medial metaphysis thinning, and lateral metaphysis thinning.

Results

The proportion of patients with the 2 narrowest stem sizes (8.5 or 10) was 51% (30/59) in group 1 and 96% (45/47) in group 2. Filling ratios in group 2 were significantly lower than in group 1 (P < .001), but stem alignment angles were clinically equivalent and actually lower in group 2 than in group 1 (mean angle, 2.2° vs. 2.7°, P = .03). Anatomic TSA patients had mild medial cortical narrowing and metaphysis thinning and none or minimal lateral metaphysis thinning, with no difference between the 2 groups. rTSA patients had mild medial cortical narrowing and metaphysis thinning, with no significant difference between the 2 groups, but lateral metaphysis thinning was less frequent and less severe among group 2 patients (0.69 ± 0.32 vs. 0.87 ± 0.17; P = .01).

Conclusion

In these 2 groups of cementless TSA patients, treated by the same surgeon and with the same implant 5 years apart, more recently treated patients had significantly narrower but similarly well-aligned stems, supporting the study hypothesis. These reduced filling ratios were associated with less frequent and less severe lateral stress shielding in patients with rTSA. The prevalence of medial stress shielding for both TSA and of lateral stress shielding for rTSA suggests different mechanisms are involved.