Total disc replacement arthroplasty using the AcroFlex lumbar disc: a non-human primate model

Using a non-human primate model, the current study was undertaken to investigate the efficacy of the AcroFlex lumbar disc as an intervertebral disc prosthesis, based on biomechanical, histopathologic and histomorphometric analyses. A total of 20 mature male baboons (Papio cynocephalus, mean weight 30 kg) were randomized into two equal groups based on post-operative time periods of 6 (n=10) and 12 months (n=10). Each animal underwent an anterior transperitoneal surgical approach to the lumbar spine, with intervertebral reconstructions performed at L3-L4 and L5-L6 using the following techniques: (1) tricortical iliac autograft and (2) AcroFlex lumbar disc. The two treatments were equally randomized between the non-contiguous operative lumbar levels. Post-mortem analysis included histopathologic assessment of the systemic reticuloendothelial tissues, multi-directional flexibility testing of the operative functional spinal units and quantitative histological analysis of trabecular bone coverage at the prosthesis endplates. Data were statistically compared using a one-way ANOVA with the Student-Newman-Keuls test. All animals survived the operative procedure and post-operative interval without significant intra- or peri-operative complications. Histopathologic analysis of the paraffin-embedded systemic reticuloendothelial tissues indicated no significant pathologic changes at the 6- or 12-month intervals. Plain film radiographic analysis showed no lucencies or loosening of any prosthetic vertebral endplate. Biomechanical testing of the 6-month autograft, reconstructions with AcroFlex lumbar disc and non-operative control (n=10) intact motion segments indicated no significant differences in peak range of motion (ROM) in axial compression. However, axial rotation produced significantly lower ROM for the autograft treatment compared to the intact and AcroFlex groups (P<0.05). The most significant differences in peak ROM were noted between all treatment groups under flexion/extension and lateral bending loading modalities (P<0.05). By 12 months, the intact condition indicated significantly more motion in all bending modes compared to the AcroFlex and autograft treatments, which were not statistically different from each other (P>0.05). Gross histopathologic analysis of the AcroFlex disc prosthesis demonstrated excellent ingrowth at the level of the implant-bone interface, without evidence of fibrous tissue or synovium. BioQuant histomorphometric analysis at the metal-bone interface (bone contact area/total endplate area) indicated the mean ingrowth was 54.59+/-13.24% at 6 months and 56.79+/-5.85% at 12 months. Radiographic analysis showed no lucencies or loosening of the AcroFlex vertebral endplate. Based on multi-directional flexibility testing, motion was preserved in axial rotation, but significantly diminished in the other bending modalities, particularly at the 12-month interval. This effect may be secondary to the limited surface area of device-vertebral endplate contact. Histomorphometric analysis of porous ingrowth coverage at the vertebral bone-metal interface was more favorable for total disc arthroplasty compared to historical reports of cementless femoral components. This project serves as the first comprehensive in vivo investigation into the AcroFlex disc prosthesis, and establishes an excellent research model in the evaluation of total disc replacement arthroplasty.