gms | German Medical Science

Objective: Anterior median corpectomy has become the most widely used procedure for the treatment of multilevel cervical stenosis. Implantation of autologous bone graft, mostly taken from the iliac crest, combined with an anterior plate-screw osteosynthesis (PSO) was the “golden standard” and resulted in fusion rates between 70 and 100%. In recent times a numerous amount of non-biological implants have been developed to avoid donor site morbidity. This study compared autologous iliac grafts with bone-filled PEEK cage implants with regard to clinical and radiological results.

Method: The clinical and radiological data of 46 patients with degenerative multilevel cervical stenosis who underwent one- or two-level anterior median corpectomy between 2004 and 2012 were analyzed. From 2004 to 2007 patients received vertebral replacement with an autologous iliac graft (group 1, n=19), from 2007 to 2012 with a PEEK cage implant (group 2, n=27). Both groups were compared using the visual analogue scale and EMS-Score. In X-rays Cobb-angle and regional cervical lordosis angle, rate of fusion and loss of height were analyzed. Periprocedural complications and revision surgery were compared.

Results: Mean follow-up was 20 ± 14 months. In both groups, pain decreased and myelopathic signs improved equally until the end of the follow-up period. In X-rays the Cobb-angle decreased by 1.4° from early postoperative to late follow-up in patients with iliac grafts and by 0.7° in patients with PEEK xenografts. Loss of height was 3.6 ± 4.1% in patients with iliac grafts vs. 5.2 ± 3.7% in patients with PEEK implants. Mean ratio between height of the cage and height of the fused vertebrae decreased in group 1 and increased in group 2. The fusion rate was 100% in patients with iliac grafts and 89% in patients with PEEK xenografts. Rate of implant failure with need for revision surgery was markedly higher in patients with PEEK cage (0% vs. 11%). In group 1, donor site morbidity occurred in one patient.

Conclusions: Preoperative pain and myelopathic symptoms improve after decompression irrespective of the material used for vertebral replacement. In our series, the use of a PEEK xenograft showed a higher risk of non-fusion and graft failure. Radiological data suggest that an intrusion of the cage into the adjacent vertebrae might contribute to a higher rate of implant and screw/plate dislocation. A prospective, randomized study is needed to gather more evidence for or against the use of PEEK cage implants.