A Prospective Cohort Analysis of Adjacent Vertebral Body Bone Mineral Density in Lumbar Surgery Patients With or Without Instrumented Posterolateral Fusion

Abstract

A prospective, cohort study of 11 patients who underwent either a posterior lumbar spinal fusion with instrumentation (n = 7) or a lumbar laminotomy and discectomy alone (n = 4) and who were evaluated by dual energy radiograph absorptiometry (DEXA) to determine bone mineral density (BMD) at the adjacent three cephalad vertebral levels. To determine long-term BMD changes that occur at the adjacent three levels above an instrumented posterolateral lumbar fusion or an isolated laminotomy and lumbar discectomy. No long-term prospective study has evaluated the affects of instrumented lumbar fusions on bone remodeling at adjacent vertebral levels. Several studies in animals and humans have reported a decrease in BMD at the adjacent level during the first 6 months after spinal fusion with a return to baseline at 1-year follow-up in up to 60% of patients. DEXA was performed initially at a mean postoperative follow-up of 4 years (range, 2.3-5.5 years) and again at a mean of 10.8 years (range, 9.1-2.4 years). Eleven patients were divided into two groups: laminotomy and discectomy (n = 4) and instrumented posterior spinal fusion (n = 7). All patients underwent surgical procedures at the L4-L5 or L5-S1 levels with DEXA analysis being performed on the adjacent three cephalad levels. The discectomy group (mean age, 57.8 years) underwent lumbar hemilaminotomy without fusion whereas the other group (mean age, 60 years) underwent pedicle-screw instrumentation and posterolateral lumbar fusion. Peripheral sites, including the femoral neck, were included in the DEXA analysis to normalize for individual differences in bone mineral metabolism. At the mean 10.8-year follow-up, the fusion group was noted to have at the adjacent level, two levels cephalad, and three levels cephalad normalized BMDs of 1.47, 1.39, and 1.27, respectively. A 14.8%, 10.8%, and 9.5% increase respectively in normalized BMD was observed when compared with the mean 4-year fusion values (P < 0.05). This increase was also noted on comparative T-score, Z-score, and absolute BMD values (P < 0.05). The discectomy group when evaluated revealed no statistically significant change from the mean 4 to 10.8-year follow-up (BMD, normalized BMD, T-score, Z-score). No statistically significant difference was noted in hip BMD at the mean 4-year and 10.8-year follow-up (1.05 versus 1.03), suggesting that the effects were local. The local BMD adjacent to an instrumented lumbar fusion is increased at a mean of 10.8-years after surgery. There is a gradual decrease in BMD changes with increasing distance from the fusion level. Alterations in fusion site biomechanics and modulus mismatch between the host bone and the spinal instrumentation most likely result in chronic, localized bone remodeling with an increased BMD that decreases the greater the distance from the fusion mass.