Prevalence and distribution of osteoporosis in degenerative lumbar scoliosis based on Hounsfield units measurement from CT examination

孙卓然,王龙杰,姜帅,李危石,邹达,王辉
DOI: https://doi.org/10.3760/cma.j.cn101202-20200805-00255
2021-01-01
Abstract:Objective:To demonstrate the prevalence and distribution of osteoporosis in degenerative lumbar scoliosis (DLS) patients based on Hounsfield units (HU) measurement and to provide a reference for the selective use of cement-augmented pedicle screws.Methods:Ninety-five DLS patients surgically treated in Peking University Third Hospital from January 2015 to September 2019 were retrospectively reviewed. Patients comprised 20 males and 75 females with a mean age of (63.3±6.5) years old. Cobb's angle and coronal balance distance (CBD) were measured by using the full-spine X-ray. Regions of interest were measured on coronal images of the vertebrae at three separate locations, namely, immediately posterior to the anterior vertebrae margin, in the middle of the vertebral body, and anterior to the posterior vertebrae margin. For each measurement, the largest possible elliptical region of interest was drawn. The cortical margins were excluded to prevent volume averaging. The average of the HU values from the three coronal slices was obtained to provide a mean HU value for each vertebral body. Then, HU measurements for the concavity and convexity of the vertebrae were obtained separately. For HU-based criterion, a cut-off of ≤110 HU was used for osteoporosis diagnosis. Patients were divided into two groups according to their diagnosis. General data and HU value of scoliosis related vertebrae were compared between the two groups.Results:The mean Cobb angle was 25.5°±11.1° degrees, and the mean CBD was (22.8±18.7) mm. Coronal imbalance type A was detected in 74 patients, type B was detected in 12 patients, and type C was detected in 9 patients. Fifty-eight patients (61.1%) presented with normal bone density, whereas 37 patients (38.9%) presented with osteoporosis. Female patients (34/75, 45.3%)were more likely to present with osteoporosis than males(3/20, 15.0%)(χ 2=6.110, P<0.05). No significant difference was found in age, body mass index, Cobb's angle, CBD, apex orientation of the major curve, and coronal imbalance types between patients with normal bone density and those with osteoporosis ( P>0.05 for all). The HU values were (122.2±45.0) Hu of L 1, (124.7±44.8) Hu of stable vertebrae, (120.9±42.9) Hu of neural vertebrae, (124.3±46.1) Hu of upper end vertebrae, (134.4±49.4) Hu of apex vertebrae, (153.3±63.8) Hu of lower end vertebrae, and (169.8±75.8) Hu of L 5. All scoliosis-related vertebrae within the major curve presented similar incidence of osteoporosis to L 1 vertebrae (all P values>0.05). L 5 vertebrae (17/95, 17.9%) presented lower incidence of osteoporosis when compared with the L 1 vertebrae(37/95, 38.9%) (χ 2=10.349, P<0.05). The incidence of osteoporosis within the convex side was significantly higher than that within the concave side at L 5, lower end, and apex vertebrae (all P values<0.01). No significant difference was found between convex and concave side HU values in the upper end, neural, and stable vertebrae (all P values>0.05). Conclusions:The prevalence of osteoporosis is 38.9% in DLS patients based on HU measurements. Females were more likely to present with osteoporosis. The vertebrae within compensatory curve presents lower incidence of osteoporosis. The convex side of apical, lower end, and L 5 vertebrae are the most common sites of osteoporosis and which needs to be strengthened by cement-augmented pedicle screws.
What problem does this paper attempt to address?