Total ankle prosthesis: preliminary experimental results on wear rates.

INTRODUCTION There is a renewed interest in total ankle replacement, but still little is known on wear rate. The aim of this study was to assess this in a new prosthesis design, in-vitro by a knee wear simulator and in-vivo in retrieval specimens by spectroscopy. METHODS Four intact specimens of the BOX Ankle (Finsbury Orthopaedics Ltd, Leatherhead, UK), ’small’ size, were analysed in-vitro. The talar and tibial components (CoCrMo, ASTM F-75) were coupled with menisci inserts (low calcium medical grade UHMWPE). Wear test was performed using a four-station knee joint simulator (Shore Western, USA), one only in soak for control. Kinematics was set in accordance with standards ISO 14243/1,2,3 and a computer model of stance in walking. Two million cycles were performed with deionised water as lubricant. Wear of the menisci components was assessed in terms of weight loss, measured using a microbalance (Sartorius AG, Germany, uncertainty ±0.03 mg). An additional retrieval specimen was obtained at revision surgery after two years, as for unresolved pain and limited mobility in a Charcot-Marie-Tooth foot. Micro-Raman spectra were obtained using a Jasco NRS2000C instrument with a microscope at 20× of magnification, and recorded on both the faces of the soak control, in-vitro specimen #2 and the retrieval. Crystallinity was calculated by PLS analysis. Dimensional measures were performed using a Coordinate Measuring Machine (CMM- Zeiss Prismo 7 VAST) to obtain working surface maps to quantify the volume variation with a reference volume (prosthesis CAD template). RESULTS After 2 million cycles the weight loss in the 3 specimens was 32.68 ± 10.32 mg, 14.78 ± 4.85 mg, and 62.28 ± 17.70 mg. Wear did not induce considerable morphology changes neither in-vitro nor in-vivo. CONCLUSION The knee wear simulator was able to reproduce accurately load-motion cycles at the replaced ankle during walking. The Raman results suggest that knee wear simulator is suitable for reproducing the morphology behaviour observed in-vivo. This fact was also confirmed by the results of the CMM analysis (Fig. 1). The volume variation did not show considerable differences among the specimens, irrespectively to wear simulated in-vitro or measured in-vivo.