Abstract
Introduction
Due to its excellent mechanical and physico-chemical properties PEEK has been used for decades as a structural implant material in various non-articulating medical devices for the spine. It is self-evident that PEEK has been proposed as a possible bearing material for total disk replacement. This study aimed to assess the wear performance of an all PEEK ball-and-socket cervical TDR concept in an invitro wear simulation at different test frequencies.
Materials and Methods
Samples The articulation geometry of the test samples was based on a ball and socket design with an articulating radius of 7.5 mm. The articulation radii were rs = 7.64 mm (concave parts) and rb = 7.43 mm (convex parts). The radial clearance was 210 µm and the surface roughness of the wear samples Ra was approximately 0.16µm.
Test set-up All PEEK samples were gamma irradiated and presoaked in demineralized water at ambient temperature for at least 2 weeks before testing. Wear tests were performed using an EndoLab servo-hydraulic wear simulator. The test set-up followed the protocol for cervical load profiles as defined in the standard ISO-18192-1. All tests were performed at 37°C in newborn calf serum with a protein content of 30 g/L.
Tests were performed at 2 Hz for 10 million cycles and at 1 Hz for 3 million cycles with two samples per test frequency. The wear was determined gravimetrically after 0.5 million cycles, after 1 million cycles, and subsequently after each interval of 1 million cycles. The wear for the samples with same test frequency was pooled and the mean wear was calculated. To document the appearance of the bearing surfaces micrographs were taken at each measuring point.
Results
The wear data for 2 Hz test frequency showed in an initial phase between 0 and approximately 3 million cycles, a high wear rate of 5.09 mg/million cycles and in a second phase starting at 4 million cycles up to 10 million cycles, a markedly reduced wear rate of 1.44 mg/million cycles. At 1 Hz, a run in period with low wear was observed over the first million cycles, followed by a sudden wear rate increase. The wear rate at 1 Hz reached a value of 8.52 mg/million cycles, which was 67% larger compared to the initial wear rate at 2 Hz. Independent of the test frequency the images of the bearing surfaces indicated a dramatic wear damage with oval-shaped wear traces and severe scratching (see image/graph).
Conclusion
In literature Grupp et al,1 reported wear rates of 1.75 mg/million cycles for a similar set-up tested at 1 Hz. A second set of PEEK-on-PEEK wear simulator data for cervical loading but 2 Hz test frequency has been published by Brown et al2 who reported very low wear rates in the range of 0.4 mg/million cycles. The observed deterioration of the PEEK bearing surfaces (see image/graph) is in line with wear test results described by Grupp et al1 and Kraft et al3 but contradicts the self-matting behavior of a PEEK bearing as postulated by Brown et al2.
The high wear rates for PEEK found in this study correlate with a severe deterioration of the bearing surfaces. These findings indicate that under loads and motion pattern relevant to cervical TDR, the use of PEEK as a bearing material might be subject to severe long-term damage of the bearing surfaces. Despite the excellent properties of PEEK as an implant material its use for articulating surfaces for cervical TDR must be critically reviewed.
Yes
Grupp TM, Meisel HJ, Cotton JA, et al. Alternative bearing materials for intervertebral disk arthroplasty. Biomaterials 2010;31(3):523–531
Brown T, Bao Q, Hallab N. Biotribology assessment of NUNEC, a PEEK on PEEK cervical disk replacement according to ISO and ASTM recommendations. 9th Annual Meeting of the International Society for the Advancement of Spine Surgery 2009, abstract no 94
Kraft M, Koch D, Bushelow M; Evaluation of PEEK on PEEK as a new articulation concept for cervical total disk replacement - an in-vitro wear simulation study. European Spine Journal 2011;20:2050, abstract no P97