INTERFERENCE SCREW FIXATION OF CERVICAL GRAFTS - A COMBINED IN-VITRO BIOMECHANICAL AND IN-VIVO ANIMAL STUDY

The dislodgement of an anterior bone graft in the cervical spine is a frequent complication of attempted fusion following discectomy or corp ectomy. It has been hypothesized that fixation augmented with interfer ence screws may increase the pullout strength of the construct and dec rease the rate of these complications. In vitro mechanical tests and i n vivo sheep studies were conducted to compare interference screw fixa tion methods for enhancing the fixation between the bone graft and the adjacent vertebra. Using human cadaver cervical spines, the anterior pullout strengths of cervical bone grafts were compared using fixation with and without the addition of interference screws for the in vitro mechanical testing. The mean pullout forces for a Smith-Robinson type bone graft alone was 58.1 N (SD 11.4 N); for the graft augmented with two 3.5 mm cancellous bone screws, 153.9 N ( 58.9 N); and for the gra ft with four 3.5 mm screws, 217.1 N (SD 69.9 N). The pullout strengths of the two- and four-3.5 mm screw constructs were significantly great er than the strength of the graft alone (P< 0.05). Similarly placed 2. 7 mm cortical screws of the same length provided increased pullout str ength (123.7 N 38.6 N and 142.5 N 38.2 N for two- and four-screw const ructs, respectively); however, in comparison to the graft alone, these differences were not statistically significant. For both screw types, the four-screw fixations were stronger than the two-screw fixations, although these differences were not statistically significant. For the in vivo portion of the study, a single-level anterior cervical discec tomy and fusion were performed on 20 sheep. The fixation was augmented in ten following surgery. At the end of 12 weeks, all nondisplaced ce rvical graft/vertebra constructs were harvested, radiographs were take n, and histology was performed. A very high rate of cervical graft dis lodgement was found in this sheep model of anterior cervical fusion. T he graft survival curve demonstrates that, for both groups, the high-r isk period of graft dislodgement is during the first week following su rgery. There was statistically-improved graft survival at 1, 2, 6, and 12 weeks for the grafts augmented with interference screws compared w ith that seen in the control group (P< 0.04). All nondisplaced grafts resulted in solid fusion by observed in direct visual examination, roe ntgenogram, and histology of the specimen. No adverse effect to graft incorporation was found for the interference screw group. This study d emonstrates that interference screw fixation augments the stability of the anterior bone graft construct and decreases the dislodgement foll owing anterior cervical grafting without compromising the graft incorp oration.