STUDY DESIGNIn vitro wear simulation.OBJECTIVETo determine the type and amount of wear produced by experimental metal-on-metal artificial discs for the lumbar spine. To minimize the amount of wear by changing the carbon content, clearance, and presence of a keel and notch.SUMMARY OF BACKGROUND DATAIn contrast to the extensive number of hip joint replacement simulator studies examining the effects of individual design variables on wear, existing artificial lumbar disc wear publications have measured wear using only the final version of each product. That is, the effects of individual variables such as material, diameter, or clearance on wear of artificial discs are not known, even though the importance of such variables has been established in artificial hip wear studies.METHODSExperimental metal-on-metal artificial discs for the lumbar spine were tested in a 3-station, biaxial spine wear simulator designed and constructed by the investigators. Two versions of the implants were manufactured with differences in carbon content, clearance, and the presence of a keel. Additionally, implants were tested with or without a surgical notch.RESULTSThe wear rates of the experimental metal-on-metal lumbar discs in the current study ranged from 6.2 to 15.8 mm3/million cycles. However, changing the carbon content of the ball from low to high, decreasing the initial clearance, and eliminating the anteroposterior keel reduced the wear rate from 12.4 to 7.6 mm3/million cycles. Furthermore, removing the surgical notch reduced the wear rate from 7.6 to 6.2 mm3/million cycles. The surface damage was generally consistent with low lubrication and varying degrees of abrasive and fatigue wear, with impingement of nonbearing surfaces observed at 1.5 million cycles for the longer-term test.CONCLUSIONAlthough the implants tested in the current study were experimental, the results suggest that metal-on-metal lumbar discs have the potential to produce wear of this magnitude and mechanism in vivo. Therefore, careful consideration of individual design variables, including those considered in the current study, is necessary to avoid production of excessive wear in artificial lumbar discs.