Posterior spinal fusion is performed for a variety of lumbar spine conditions for relief of low back pain. Success relies on an effective fusion. Autograft is associated with donor site comorbidity and limited supply. Allograft has the potential for infection and has limited osteoinductive activity. Bone morphogenic proteins (BMPs) have been promoted for use in posterior spinal fusion despite considerable cost and limited evidence to their efficacy.

The aim of this study was to compare the clinical response, donor site morbidity and radiologic rates of fusion in patients undergoing posterior spinal fusion looking at the choice of bone graft or substitute. A retrospective review of 141 patients undergoing instrumented posterior lumbar spinal fusion by a single surgeon for degenerative disc disease, degenerative spondylolithesis or lytic spondylolithesis between 2000 and 2005 was undertaken. Patients were contacted and assessed for donor site morbidity and scored with the Oswestry Disability Index (ODI). Radiographs were taken and assessed by an independent blinded radiologist using the Ferguson score. Simple analysis was performed of these results to compare bone grafting techniques.

One hundred and forty-one patients were available for review. Fusions were performed for lytic spondylolithesis in 12.4%, degenerative spondylolithesis in 46% and for degenerative disc disease in 41.6% of patients. BMP-2 was used in 19.6%, allograft in 59.8% and iliac crest bone graft in 20.5% of patients. The BMP-2 and non BMP-2 groups were equally spread between the diagnosis and levels of surgery. The overall Ferguson score radiographic fusion rates for these patients was A in 67.9%, B in 17.9% and C in 11.9%. The BMP-2 group patients scored 76.9% (A group) and 23.1% (B group). The non-BMP-2 group scored 57.1% (A group), 23.8% (B group) and 19.1% (C group). The Oswestry Disability Index for patients with BMP-2 improved from 49.7% to 19%, whereas with no BMP-2 improved from 50.0% to 20.9%. Donor site morbidity was not identified as a problem in patients who had an autograft procedure.

Over the course of several years a single spinal surgeon’s posterior lumbar spinal fusion practice has evolved as a variety of bone grafting techniques have been trialled in an effort to increase the rate of bony fusion. There was no obvious difference in Oswestry Disability Index score but there was a modest difference in the Ferguson radiologic fusion score for the BMP-2 group. Morbidity in the autograft group was not a problem. These results have confirmed the efficacy of both allograft and autograft in fusion.