Quantitative CT for Preoperative Assessment of Lumbar Degenerative Spondylolisthesis: The Unique Impact of L4 Bone Mineral Density on Single-Level Disease

Background: Quantitative computed tomography (qCT) efficiently measures 3-dimensional vertebral bone mineral density (BMD), but its utility in measuring BMD at various vertebral levels in patients with lumbar degenerative spondylolisthesis (LDS) is unclear. Purpose: We sought to determine whether qCT could differentiate BMD at different levels of LDS, particularly at L4-L5, the most common single level for LDS. In addition, we sought to describe patterns of BMD for single-level and multiple-level LDS. Methods: We conducted a study of patients undergoing surgery for LDS who were part of a larger longitudinal study comparing preoperative and intraoperative images. Preoperative patients were grouped as single-level or multiple-level LDS, and qCT BMD was obtained for L1-S1 vertebrae. Mean BMD was compared with literature reports; in multivariable analyses, BMD of each vertebra was assessed according to the level of LDS, controlling for covariates and for BMD of other vertebrae. Results: Of 250 patients (mean age: 67 years, 64% women), 22 had LDS at L3-L4 only, 170 at L4-L5 only, 13 at L5-S1 only, and 45 at multiple levels. Compared with other disorders reported in the literature, BMD in our sample similarly decreased from L1 to L3 then increased from L4 to S1, but mean BMD per vertebra in our sample was lower. Nearly half of our sample met criteria for osteopenia. In multivariable analysis controlling for BMD at other vertebrae, lower L4 BMD was associated with LDS at L4-L5, greater pelvic incidence minus lumbar lordosis, and not having diabetes. In contrast, in similar multivariable analysis, greater L4 BMD was associated with LDS at L3-L4. Bone mineral density of L3 and L5 was not associated with LDS levels. Conclusion: In our sample of preoperative patients with LDS, we observed lower BMD for LDS than for other lumbar disorders. L4 BMD varied according to the level of LDS after controlling for covariates and BMD of other vertebrae. Given that BMD can be obtained from routine imaging, our findings suggest that qCT data may be useful in the comprehensive assessment of and strategy for LDS surgery. More research is needed to elucidate the cause-effect relationships among spinopelvic alignment, LDS, and BMD.