Comparison of three-level anterior cervical discectomy and fusion using iliac crest bone graft and plate vs interbody cages with allograft and plate

Abstract

Anterior cervical plate fixation is believed to stabilize the operative motion segments, decreasing graft migration and preventing graft collapse that could lead to loss of cervical lordosis, malunion or nonunion and neurologic compromise. Although reports have noted high fusion rates in plated multilevel anterior cervical discectomy and fusion (ACDF) that range from 47% to 100% and are dependent on the number of levels fused, the efficacy of the combination of a titanium cage with allograft in comparison with autograft in such patients has not been investigated. We retrospectively analyzed 36 cases of three-level ACDF with anterior semi rigid plating. All cases were performed between August 2000 and June 2005. There were 19 males and 17 females with an average age of 51.6 (range from 35 to 69). Operated levels were C4-C7 in 30 patients, C3-C6 in 5 and C5-T1 in one patient. Nineteen patients (52%) had autologous iliac crest tricortical grafts, 17 (47%) had PEEK cages placed filled with allografts. The technique was the same in all cases: a standard left anteromedial approach to the cervical spine. The intervertebral discs were removed. All patients had burring of the end plates, 2 mm distraction and countersinking of the grafts by 2 mm from the anterior vertebral border. The autologous bone graft was tricortical and was harvested from the iliac crest using a low speed oscillating saw. The allograft used was fresh frozen, vacuum sealed cancellous chips and putty. An anterior cervical titanium plate was selected with variable angle locked screws. Clinical and radiographic data at 1 and 2 years postoperatively were obtained. Clinical outcome was assessed on 1 and 2 year follow-up and rated according to the Odom criteria. Fusion was achieved in 18 (94%) of the iliac crest group and 12 (71%) of the PEEK cage-allograft group. 1 patient in the iliac crest group and 5 in the cage-allograft group developed nonunions. Clinical outcome at 2 years was excellent in 5, good in 12 and fair in 2 patients of the iliac crest group. On the cage-allograft group it was excellent in 3, good in 13 and fair in 1. The 5 nonunions had good clinical outcome and 1 fair. We concluded that although autograft is the gold standard for 3 level ACDF, the use of allograft has the same functional status irrespective of the higher number of pseudarthrosis. The decision to use an autograft or a cage and an allograft for ACDF is therefore based mainly on personal preferences.