Purpose: Interbody fusion cages have met with great success as an adjunct in the treatment of painful degenerative disc disease. One of the limitations is the need for the use of autogenous cancellous bone graft. In preclinical studies recombinant human bone morphogenetic protein-2 (rhBMP-2) delivered in a variety of carriers has been shown to be an effective substitute for autogenous bone, resulting in more rapid and reliable healing than that seen in control groups. The goal of this study was to report the early results of the first human trial attempting to use rhBMP-2 in interbody fusion cages.

Methods: This study was an FDA approved IDE multicenter pilot study. From 1/97 to 4/97, 14 patients were entered into a prospective, randomized trial. All patients had single level lumbar degenerative disc disease that was refractory to prolonged nonoperative care and were candidates for anterior interbody fusion of L4-5 or L5-S1. After consent, patients were randomized to either the control group (N-3) and received autogenous bone inside tapered titanium fusion cages (NOVUS LT, Sofamor Danek Memphis, TN) or to the investigational group (N = 11) and received rhBMP-2 (1.5 mg/ml)(Genetics Institute, Cambridge MA) delivered in a collagen sponge (Helistat, Integra Life Sciences, Plainsboro, NJ) inside the fusion cages. Depending on the size, the sponge in each cage was soaked with from 1.3 to 2.6 ml of the rhBMP2 solution. Patients were followed at regular intervals with plain x-ray, CT scan with reconstruction, and a full panel of blood tests. Radiographs were reviewed by an independent blinded radiologist with fusion defined as < 5 degrees of sagittal motion, absence of radiolucent lines, and presence of continuous bone through the cages. Clinical results were assessed using an outcomes questionnaire including the SF-36 general health status and Oswestry low back specific instruments.

Results: All 14 patients were available for 1-year follow-up. No cages displaced and no further surgeries were required. Mean hospital stay was 2.0 days for the rhBMP-2 patients compared to 3.3 days for the autograft controls. Of the 11 rhBMP2 patients, 10 of 11 were judged to be fused at 3 months. At 6 months and 1-year all 11 rhBMP-2 patients were noted to have a solid arthrodesis. Of the 3 control patients, 2 had solid arthodesis and one had an apparent nonunion at 1 year. On sagittal CT scan reconstruction new bone growth was seen throughout and anterior to the cages that were filled with rhBMP-2. No patients had bone formation outside of the desired area. The control patient with the nonunion had a halo surrounding the cage on the sagittal CT reconstruction. This patient had persistence of low back pain. Compared to preop, the Oswestry scores at 3 months were decreased in the rhBMP-2 group (39 to 30) compared to controls which were increased (35 to 43) and both mean scores were similar at 6 months (28 and 27). Conclusion: The preliminary results from this clinical trial with rhBMP-2 inside interbody fusion cages were excellent and support a larger pivotal trial. The arthrodesis was found to occur more rapidly and reliably than in the controls, although the sample size was limited. In addition to faster bone healing, a major advantage was the elimination of bone graft donor site morbidity and realization of decreased hospital stay. No evidence of excessive bone formation or systemic complications occurred. Moreover, this study provides one of the first demonstrations of consistent and unequivocal osteoinduction by a recombinant growth factor in humans.