Nerve injury and recovery after lateral lumbar interbody fusion with and without bone morphogenetic protein-2 augmentation: a cohort-controlled study.

Background Context: Despite common use of intraoperative electrophysiologic neuromonitoring, injuries to the lumbar plexus during lateral lumbar interbody fusion (LLIF) have been reported. Emerging data suggest that recombinant human bone morphogenetic protein-2 (rhBMP-2) use during an anterior or transforaminal lumbar interbody fusion may be associated with an increased risk of neurological deficit. Clinical data on the sequelae of rhBMP-2 implantation in close proximity to the lumbosacral plexus during LLIF remains to be understood.

Purpose: The purpose of this study was to compare the incidence of neurologic deficits and pain in patients undergoing LLIF with and without rhBMP-2.

Study Design/setting: Retrospective outcome analysis in controlled cohorts undergoing the lateral exposure technique for LLIF with and without rhBMP-2.

Methods: The electronic medical records of patients undergoing LLIF with and without supplemental posterior fusion for degenerative spinal conditions were retrospectively reviewed over a 6-year period. Patients with previous lumbar spine surgery or follow-up of less than 6 months were excluded. Patients were divided into 2 groups, Group 1 (rhBMP-2 use; n=72) and Group 2 (autograft/allograft use; n=72), and were matched according to the age at the time of surgery, gender, weight, body mass index, side of approach, total number of treated spinal segments, use of supplemental posterior fusion, and length of follow-up.

Results: Immediately after surgery, a sensory deficit was recorded in 33 patients in Group 1 and 35 patients in Group 2 (odds ratio [OR] 0.895; 90% confidence interval [CI] 0.516-1.550; p=.739). At last follow-up, a persistent sensory deficit was identified in 29 patients whose LLIF procedure was supplemented by rhBMP-2 and 20 patients in whom autograft/allograft was used (OR 1.754; 90% CI 0.976-3.151; p=.115). A motor deficit was recorded in 37 patients immediately after the rhBMP-2 procedure and 28 patients treated with autograft/allograft (OR 1.661; 90% CI 0.953-2.895; p=.133). A persistent motor deficit was recorded in 35 and 17 patients in Groups 1 and 2, respectively, at last follow-up (OR 3.060; 90% CI 1.681-5.571; p=.002). During the first postoperative examination, 37 patients in Group 1 and 25 patients in Group 2 complained of anterior thigh or groin pain (OR 1.987; 90% CI 1.133-3.488; p=.045). At last follow-up, there was a significantly higher number of patients in Group 1 who complained of persistent anterior thigh or groin pain than Group 2 (8 vs. 0 patients) (OR 16.470; 90% CI 1.477-183.700; p=.006).

Conclusions: Our results provide evidence of an increased rate of postoperative neurologic deficit and anterior thigh/groin pain after LLIF using rhBMP-2, when compared with matched controls without rhBMP-2 exposure. This study suggests a potential direct deleterious effect of rhBMP-2 on the lumbosacral plexus.