Complications with recombinant human bone morphogenetic protein-2 in posterolateral spine fusion associated with a dural tear

Abstract

Background context Potential complications related to ectopic bone formation, as seen with transforaminal lumbar interbody fusion, have always been a concern with the use of bone morphogenetic proteins (BMPs). Although less clearly anticipated, complications related to the proinflammatory effects of recombinant human bone morphogenetic protein-2 (rhBMP-2), such as swelling and edema in the cervical spine, have been observed as well. Until recently, risks related to intradural exposure to BMP have not been widely considered. However, a recent animal study reports “in the presence of a SCI and/or dural tear, rhBMP-2 diffuses intrathecally and activates a signaling cascade in all major CNS cell types, which may increase glial scarring and impact neurologic recovery.” Although this study was conducted at the spinal cord level, the observation generates obvious concerns for the much more common scenario of a dural tear associated with lumbar decompression and fusion. Purpose The purpose of this study was to look for any indication of neurologic injury or impaired neurologic recovery in patients treated with rhBMP-2 for lumbar fusion complicated by dural tear. Study design/setting Propensity score matched case-control study. Patient sample From consecutive series of 1,037 patients who underwent decompression and posterolateral lumbar spine fusion using rhBMP-2/absorbable collagen sponge between 2003 and 2006, intraoperative dural tear was reported in 58 cases (5.59%). Outcome measures Preoperative and 2-year postoperative Oswestry Disability Index, Short Form-36 (SF-36), leg pain, and back pain scores. Methods Fifty-eight cases in which decompression and posterolateral spinal fusion were complicated by dural tear, where propensity score matched to a group without dural tear, based on age, smoking status, number of surgical levels and preoperative Oswestry Disability Index, SF-36 Physical Composite Summary score, SF-36 Mental Composite Summary (MCS) score, and back and leg pain scores. The patients with a dural tear were then compared with the matched cohort with regard to baseline and 2-year patient-based outcome measures. Particular attention was given to indices of leg pain that might reflect an influence of rhBMP-2 on neurologic function or impaired neurologic recovery. Results No patient in the group with a dural tear and three patients in the group without a dural tear complained of new onset radiculopathy postoperatively, with one requiring oral steroids. The radiculopathy resolved within 6 months postoperatively in all three patients. Statistically significant improvement was observed in all health-related quality of life (HRQOL) measures, except SF-36 MCS, at both 1 and 2 years postoperatively in both groups. There were no significant differences in any HRQOL parameter between the groups with or without a dural tear at either 1 or 2 years postoperatively. In particular, the leg pain improvement, 2.2 points in the group with a dural tear and 2.4 points in the group without a dural tear, was statistically equivalent. Conclusions The data suggests fairly convincingly that the presence of a repairable dural tear is not necessarily an impediment to the use of rhBMP-2 in posterolateral fusion. Further studies are needed to address the less common clinical scenario of BMP use in conjunction with spinal cord injury, as studied in the animal model that prompted this investigation. Finally, avoidance of BMP use may still be prudent in the setting of an unrepairable dural tear.