A comparative analysis of fusion rates and donor-site morbidity for autogeneic rib and iliac crest bone grafts in posterior cervical fusions.

OBJECT: Autogeneic bone graft is often incorporated into posterior cervical stabilization constructs as a fusion substrate. Iliac crest is used frequently, although donor-site morbidity can be substantial. Rib is used rarely, despite its accessibility, expandability, unique curvature, and high bone morphogenetic protein content. The authors present a comparative analysis of autogeneic rib and iliac crest bone grafts, with emphasis on fusion rate and donor-site morbidity.

METHODS: A review was conducted of records and radiographs from 600 patients who underwent cervical spinal fusion procedures in which autogeneic bone grafts were used. Three hundred patients underwent rib harvest and posterior cervical fusion. The remaining 300 patients underwent iliac crest harvest (248 for an anterior cervical fusion and 52 for posterior fusion). The analysis of fusion focused on the latter subgroup; donor-site morbidity was determined by evaluating the entire group. Fusion criteria included bony trabeculae traversing the donor-recipient interface and long-term stability on flexion-extension radiographs. Graft morbidity was defined as any untoward event attributable to the graft harvest. Statistical comparisons were facilitated by using Fisher's exact test.

CONCLUSIONS: Demographic data obtained in both groups were comparable. Rib constructs were placed in the following regions: occipitocervical (196 patients), atlantoaxial (35 patients), and subaxial cervical spine (69 patients). Iliac crest grafts were placed in the occipitocervical (28 patients), atlantoaxial (10 patients), and subaxial cervical (14 patients) regions. Fusion occurred in 296 (98.8%) of 300 rib graft and 49 (94.2%) of 52 iliac crest graft constructs (p = 0.056). Graft morbidity was greater with iliac crest than with rib (p < 0.00001). Donor-site morbidity for the rib graft was 3.7% and included pneumonia (eight patients), persistent atelectasis (two patients), and superficial wound dehiscence (one patient). Pneumothorax, intercostal neuralgia, and chronic chest wall pain were not encountered. Iliac crest morbidity occurred in 25.3% of the patients and consisted of chronic donor-site pain (52 patients), wound dehiscence (eight patients), pneumonia (seven patients), meralgia paresthetica (four patients), hematoma requiring evacuation (three patients), and iliac spine fracture (two patients). Even when chronic pain was not considered, morbidity encountered in obtaining iliac crest still exceeded that encountered with rib harvest (p = 0.035). The fusion rate and donor-site morbidity for rib autograft compare favorably with those for iliac crest when used in posterior cervical constructs. To the authors' knowledge, this represents the largest series to date in which the safety and efficacy of using autogeneic bone graft materials in spinal surgery are critically analyzed.