Abstract
Introduction
Reverse total shoulder arthroplasty (RTSA) is a semi-constrained joint replacement with an articulating cobalt-chromium glenosphere and ultra-high molecular weight polyethylene (PE). Because of its limited load bearing, surgeons and implant manufacturers have not elicited the use of highly cross-linked PE in the shoulder, and to date have not considered excessive PE wear in the reverse shoulder a primary concern. As the number of shoulder procedures is expected to grow exponentially in the next decade, however, it is important to evaluate how new designs and bearing materials interact and to have an understanding of what is normal in well-functioning joint replacements. Currently, no in vivo investigation into RTSA PE wear has been conducted, with limited retrieval and simulation studies. In vitro and in silico studies demonstrate a large range in expected wear rates, from 14.3 mm3/million cycles (MC) to 126 mm3/MC, with no obvious relationship between wear rate and polyethylene diameter. The purpose of this study is to evaluate, for the first time, both volumetric and linear wear rates in reverse shoulder patients, with a minimum six-year follow-up using stereo radiographic techniques.
Methods
To date, seven patients with a self-reported well-functioning Aequalis Reversed II (Wright Medical Group, Edina, MN, USA) RTSA implant system have been imaged (mean years from surgery = 7.0, range = 6.2 to 9). Using stereo radiographs, patients were imaged at the extents of their range of motion in internal and external rotation, lateral abduction, forward flexion, and with their arm at the side. Multiple arm positions were used to account for the multiple wear vectors associated with activities of daily living and the shoulder's six degrees of motion. Using proprietary software, the position and orientation of the polyethylene and glenosphere components were identified and their transformation matrices recorded. These transformation matrices were then applied to the CAD models of each component, respectively, and the apparent intersection of the glenosphere into the PE recorded. Using previously validated in-house software, volumetric and maximum linear wear depth measurements were obtained. Linear regression was used to identify wear rates.
Results
The volumetric and linear wear rates for the 36 mm PE liners (n = 5) were 39 mm3/y (r2= 0.86, range = 24 to 42 mm3/y) and 0.09 mm/y (r2 = 0.96, range = 0.08 to 0.11 mm/y), respectively. Only two patients with 42 mm PE liners were evaluated. For these, volumetric and linear wear rates were 110 mm3/y (r2 = 0.81, range = 83 to 145 mm3/y) and 0.17 mm/y (r2 = 0.99, range = 1.12 to 1.15 mm/y), respectively.
Conclusion
For the first time, PE wear was evaluated in the reverse shoulder in vivo. More patients are required for conclusive statements, but preliminary results suggest first order volumetric and linear wear rates within those predicted by simulation studies. It is interesting to note the increased wear with larger PE size, likely due to the increased contact area between congruent faces and the potential for increased sliding distance during arm motion.
