The evolvement of cervical and lumbar disc replacement designs as alternatives to spinal fusion has resulted in a substantial number of ongoing United States Food and Drug Administration (FDA)-sponsored clinical trials. While these trials seek to establish "safety and effectiveness," they are of limited in vivo duration and benefit from long-term, benchtop laboratory comparison. To ensure implant durability, mechanical and biological evaluation of these devices, particularly with regard to long-term fatigue behavior, is necessary.
Both ASTM International (in ASTM Standard F2423 - 05)1 and the International Organization for Standardization (in ISO 18192-1:2008)2 have provided guidance in the evaluation of the performance of artificial spinal discs. These guides propose the biochemical environment, motions, and loading that are appropriate to simulate long-term use of prostheses employed in total disc arthroplasty. The evaluated parameters include wear measured by gravimetric weight loss as well as changes in the articular surface shape and roughness to the extent that these may influence function.
This study describes the wear characteristics of an articulating cervical disc replacement during approximately eighty years of simulated loading in an electromechanical, multiaxial spinal disc simulator (U.S. Patent No. 7,493,828 B2)3 with comparison to clinical retrievals. These results are useful in demonstrating the safety and effectiveness of this device and also present a preclinical evaluation methodology for future disc-replacement designs.
The challenge with any simulation design is the accuracy to which the design replicates physiologic human motions and loads such that the information derived is predictive of anticipated in vivo performance. Compliance with current FDA, ASTM, and ISO guidelines suggests that spinal device simulators have multiple simultaneous rotations, joint compression, and a hydrated, temperature-controlled environment. The daily activities of the patient, including walking, running, and stair-climbing, are translated to realistic simulations of spinal flexion-extension, lateral bending, and axial rotation under load.
Each specimen was aligned in an anatomical position for an erect patient and loaded statically to an average force (and standard deviation) of 100 N ± 2 N of joint compression through the center of rotation of the prosthesis via a compression spring. Under displacement control, the prosthesis was articulated simultaneously through 6.0° ± 0.5° of fully reversing axial rotation (r = −1) and 10.0° ± 0.5° of fully reversing lateral bending (r = −1) about the center of rotation of the prosthesis. All motions were sinusoidal and in phase. The vector of the compressive force tracked the bending motion as a follower load. Test frequencies were 1 Hz for the first 2 million cycles and 1.35 Hz from 2 million to 10 million cycles (Fig. 2).
Five specimens were tested in a series of identical, custom, single-station motion simulators for 10 million motion cycles (Fig. 3). All tests were conducted in 37° ± 3°C filtered and sterilized bovine calf serum lubricant diluted to a total protein concentration of 20 grams per liter with phosphate buffered saline solution, following ASTM Standard F1714 - 964. The lubricant contained 0.2% sodium azide and ethylenediaminetetraacetic acid (EDTA) at a concentration of 20 mM. Distilled water was added as needed to correct for lubricant evaporation.
Wear was determined by the gravimetric weight-loss method described in ASTM Standard F1714. Specimens were evaluated before testing, every 500,000 cycles for 1 million cycles, and every 1 million cycles thereafter. Individual components were weighed on a Sartorius ME235S analytical balance (Sartorius AG, Goettingen, Germany) with a resolution of 0.01 mg. The cumulative weight loss and wear rate were reported for each specimen at each interval after correcting for fluid absorption in the load soak controls (n = 2). At each of the test intervals, the lubricant was exchanged to enable wear particle analysis.
The Simulation
All five devices completed 10 million cycles of simulated activity without functional failure and with minimal debris generation. The ultra-high molecular weight polyethylene convex (caudal) articular surface of the specimens presented with considerable areas of polishing while the cobalt-chromium-molybdenum concave (cephalic) articular surface presented with a few long and isolated scratches that were 3 to 5 mm in length (Fig. 4). The mean volumetric wear rates were 9.5 mm3 per million cycles and 0.2 mm3 per million cycles for the ultra-high molecular weight polyethylene convex (caudal) and cobalt-chromium-molybdenum concave (cephalic) components, respectively (Figs. 5 and 6). Furthermore, surface evaluation of the ultra-high molecular weight polyethylene convex (caudal) components demonstrated that the articulations remained spherical, with only a slight flattening detected in most specimens.
The Retrieved Clinical Specimens
The majority of the ultra-high molecular weight polyethylene convex (caudal) components demonstrated wear, defined as areas in which an absence of machine marks was evident (Fig. 7). Severe gouging marks associated with explantation were observed on several components and were confined to the anterior region. Fracture did not occur in any of the specimens, and third-body particulate was found embedded in five of the twenty-five retrievals. Separation of the ultra-high molecular weight polyethylene from the metal backing in one of the twenty-five specimens and damage to the locking mechanism in one of the twenty-five specimens were noted; however, whether these occurred in vivo or during removal cannot be clearly determined.
The PCM cervical disc replacement demonstrated a lower volumetric wear rate per million cycles than that reported for contemporary hip and knee arthroplasty designs in the simulator study. The components exhibited functional durability that would suggest success in long-term in vivo use. Similar wear patterns, particular to the removal of machine marks on the ultra-high molecular weight polyethylene convex (caudal) component, were noted in the majority of retrievals. Several damage modes were observed on the retrieved components that were not captured in the simulator study. Future studies are needed to correlate individual surgical factors, with the goal being to identify the cause of these wear changes.
These laboratory simulator evaluations assist us in understanding how material wear characteristics affect the anticipated performance of cervical and lumbar spinal disc replacements. These evaluations should be required in all future designs (both articulating and elastomeric) to ensure their functional and mechanical integrity.