- Email: [email protected]
Cranio-caudal asymmetries in trabecular architecture reflect vertebral fracture patterns
Bone 95 (2017) 102–107
Contents lists available at ScienceDirect
Bone journal homepage: www.elsevier.com/locate/bone
Full Length Article
Cranio-caudal asymmetries in trabecular architecture reflect vertebral fracture patterns Ge Yang a, Michele C. Battié b, Steven K. Boyd c, Tapio Videman b, Yue Wang a,⁎ a b c
Spine lab, Department of Orthopedic Surgery, The First Affiliated Hospital of Zhejiang University, Hangzhou, China Faculty of Rehabilitation Medicine, University of Alberta, Edmonton, Canada Cumming School of Medicine, University of Calgary, Calgary, Canada
a r t i c l e
i n f o
Article history: Received 26 August 2016 Revised 15 November 2016 Accepted 18 November 2016 Available online 19 November 2016 Keywords: Lumbar spine Vertebral fractures Trabecular microstructure μCT Disc degeneration
a b s t r a c t Clinically, vertebral fractures often occur in the upper lumbar spine and involve the superior endplate of a vertebra (which is immediately caudal to a disc). Knowledge that the cranial endplate of a disc is thicker and has greater bone mineral density (BMD) than the corresponding caudal endplate helps to explain this phenomenon. In this study, we investigated structural differences in vertebral trabeculae on either side of a lumbar disc to provide further insight into vertebral fracture risk. As the focus is trabecular difference within a spinal motion segment, we define cranial and caudal vertebral trabeculae relative to the disc. Ninety-two spinal motion segments from 46 cadaveric lumbar spines (males, mean age 50 years, range 21–63 years) were studied. Disc narrowing on radiography and spread of barium sulfate (BaSO4) on discography were measured to indicate disc degeneration. Micro-computed tomography (μCT) images were obtained at a resolution of 82 μm for each vertebra and processed to include only vertebral trabeculae. Using image processing, the vertebral trabeculae were divided into superior and inferior halves, and then into central and peripheral regions which were approximately opposite to the disc pulposus and annulus, and further into anterior and posterior sub-regions. Microarchitecture measurements for each vertebral region were obtained to determine the differences between the cranial and caudal trabeculae (relative to disc) and their associations with age and disc degeneration within each spinal motion segment. Data from the upper (L1/2–L3/4) and lower (L4/5) lumbar segments were analyzed separately. In the upper lumbar region, the trabeculae cranial to a disc on average had 5.3% greater BMD and trabecular bone volume, 3.6% greater trabecular number, 9.7% greater connectivity density, and 3.7% less trabecular separation than the corresponding caudal trabeculae (P b 0.05 for all). Similar trends were observed in peripheral, anterior and posterior regions, but not in central region. No structural difference was observed in the trabeculae of L4/5 segment. Structural asymmetries of vertebral trabeculae were not associated with age, disc degeneration, or disc narrowing. Vertebral trabecular parameters cranial to the disc were greater than caudally in the upper but not in the lower lumbar region. Findings further explain why vertebral fractures are more common in the upper lumbar region and more frequently involve the endplate caudal to a disc. © 2016 Elsevier Inc. All rights reserved.
1. Introduction Traumatic and osteoporotic vertebral fractures are among the most common skeletal injuries in adults. Interestingly, thoracolumbar vertebral fractures involve the superior portion of the vertebral body, including the endplate and trabeculae, more often than the inferior portion [1,2]. This clinical phenomenon may be partially explained by previous findings that the superior endplate of a vertebra is thinner and the trabeculae are less dense in the superior portion of the vertebra than inferiorly [3,4]. While the reason for the structural differences within a vertebral body remains unexplained, the superior and inferior portions ⁎ Corresponding author at: 79# Qingchun Road, Hangzhou 310003, China. E-mail address: [email protected] (Y. Wang).
http://dx.doi.org/10.1016/j.bone.2016.11.018 8756-3282/© 2016 Elsevier Inc. All rights reserved.
of a vertebra originate from different sclerotomes that fuse together in a developmental process called “resegmentation” [5]. Yet, the trabecular structure on either side of the disc within a spinal motion segment also seems to be considerably different, although they are derived from one sclerotome. A spinal motion segment, including two adjacent vertebrae and the intervening intervertebral disc, is often studied as a working unit in spine mechanics. During resegmentation, portions of two adjacent sclerotomes, a cranial one with densely packed sclerotome cells and a caudal one with loosely packed cells, fuse into a vertebral precursor. In other words, within a spinal motion segment, the half of the vertebral body cranial to a disc has less cell density than the half of the vertebral body caudal to the disc when the embryonic spine is formed (Fig. 1) [6,7]. Given the fact that the sclerotome cells are the major source of chordacentrum
G. Yang et al. / Bone 95 (2017) 102–107
103
Fig. 1. Sclerotomes resegmentation. A: In a sclerotome, cells are loosely packed in the cranial portion but densely packed in the caudal portion. B: In sclerotomes resegmentation, notochord expanded to form intervertebral discs, which separate adjacent vertebrae. In an embryonic spinal segment, therefore, vertebra immediately caudal to a disc have densely packed cells and that cranial to a disc have loosely packed cells. (Adapted from Dar G. et al. [5]).
mineralization [8] and the early period of ontogeny has a substantial influence on bone quality and structure [9], theoretically, vertebral trabeculae caudal to a disc could be expected to have better bone quality than the corresponding cranial trabeculae. On the other hand, a healthy intervertebral disc is able to distribute mechanical loads upward and downward evenly, and the endplates and trabeculae cranial and caudal to a disc overall share equivalent amounts of loads. Age related disc degeneration, however, may alter stress distribution within a spinal motion segment [10,11]. Our previous studies revealed that the endplate cranial to a disc is more concave, thicker, and of higher bone mineral density (BMD) than the caudal endplate [12,13]. In addition, we observed that a greater degree of disc degeneration was associated with greater BMD in the vertebra caudally adjacent to disc, but not in the cranial vertebra [14]. Taking evidence together, we hypothesized that within a spinal motion segment there is also an asymmetry of vertebral trabecular structure on either side of the disc. In this study, we further sought to determine if the asymmetry of trabecular structure is associated with aging or disc degeneration, or merely a design feature of the lumbar spine. As the focus of this paper is trabecular difference within a spinal motion segment, we define cranial and caudal vertebral trabeculae relative to the disc, if not specified. 2. Materials and methods 2.1. Sample We used a lumbar spine archive comprised of 149 Caucasian male cadavers, with a mean age of 50 years at the time of death and a range from 21 to 64 years. The archive included men under the age of 65 years, employed up to the time of death, who passed away with a short illness history [15]. After a routine autopsy examination, X-ray images and discography of the lumbar spine were performed. Then, spinal ligaments, muscles and intervertebral discs were removed and lumbar vertebrae were washed, processed, dried and preserved at room temperature. As some archived vertebrae data were lost, we selected and obtained micro-computed tomography (μCT) scans of 150 lumbar vertebrae from 48 of the cadavers, which are with adjacent disc degeneration data, to study the interactions between vertebra and disc [14]. Using the same sample, the purpose of the present study was to investigate differences in vertebral trabeculae within a spinal motion
segment. Thus, complete motion segments with data of both vertebrae and the intervening disc were required. The research was approved by the Health Research Ethics Board at the University of Alberta. 2.2. Measurements of disc degeneration Disc degeneration was measured using discography and disc space narrowing. Discography was performed by injecting 2–5 ml of barium sulfate (BaSO4) anteriorly into the center of an intervertebral disc. According to the spread of BaSO4 on discogram, a 4-grade ordinal scale (none, slight, moderate or severe) was used to rate the degree of disc degeneration, as reported previously [16]. Disc space narrowing was assessed from lateral X-ray radiographs using a three-point scale as none, slight to moderate, or severe. The reliability of the disc degeneration measurements was good (intra-rater kappa = 0.81 for discographic degeneration and kappa = 0.71 for disc space narrowing measurements) [17]. 2.3. Measurement of vertebral trabeculae microstructure Each vertebra was scanned using a μCT system (XtremeCT, Scanco Medical, Brüttisellen, Switzerland) with standard scanning parameters: 60 kVp, 1000 μA, 200 ms integration time, and 750 projections. The dried vertebrae were scanned with a nominal isotropic resolution of 82 μm (field of view 125 mm, 1536 × 1536 pixels) in air. Regions of interest were identified for the vertebral body in each microradiograph and contoured using a semi-automated contouring method [18]. In so doing, only the trabeculae of the vertebral body were included in data analysis and other components of the vertebra, such as the cortex of the vertebral body and posterior elements of the vertebra, were excluded. Using a technique we previously reported [17], each vertebral body was first divided into superior and inferior halves. Then, each half vertebra was further divided into central and peripheral regions, which were approximately corresponding to the areas connecting to nucleus pulposus and annulus fibrosus. Further, the peripheral region was segmented into anterior and posterior sub-regions (Fig. 2). Structural and densitometric analyses were performed (Image Processing Language, v5.15, Scanco Medical AG, Brüttisellen, Switzerland) to obtain trabecular bone volume fraction (BV/TV; %), trabecular number (Tb.N; 1/mm), BMD (mg/cm3), trabecular thickness (Tb.Th; mm),
104
G. Yang et al. / Bone 95 (2017) 102–107
Fig. 2. Schematic diagram illustrating the division of vertebral trabeculae for a spinal motion segment. The trabeculae of a vertebral body were separated as superior and inferior two halves, which were referred as cranial or caudal trabeculae adjacent to the disc. Each half was divided into central and peripheral sections, corresponding to the regions connecting to nucleus pulpous and annular fibrosus, respectively. The peripheral section of the vertebral trabeculae was further segmented into anterior and posterior sub-regions.
trabecular separation (Tb.Sp; mm) and connectivity density (Conn.D; 1/ mm3) parameters for each defined section.
with greater discographic disc degeneration (OR = 1.66 for each 10 years, P b 0.01), greater disc narrowing (OR = 1.32 for each 10 years, P b 0.01), and less vertebral BMD (coef. = −0.99, P b 0.01).
2.4. Statistical analysis As spinal biomechanics and prevalence of disc degeneration are considerably different between the upper lumbar region (L1/2, L2/3 and L3/ 4 discs) and lower lumbar region (L4/5 and L5/S1 discs), data for each lumbar region were analyzed separately. Paired t-tests were used to compare the differences of trabecular microstructures between cranial and caudal trabeculae within a spinal motion segment. Taking L4/L5 segment as an example, we matched trabecular measurements of the inferior half of the L4 vertebra with those of the superior half of L5 vertebra. The differences in regional architectural parameters were presented as the percentage of difference between cranial and caudal vertebrae. Multivariable regression was used to examine the associations between trabecular differences and age, body mass index (BMI; kg/cm2), disc degeneration and disc space narrowing. Data analysis was performed using Stata (Release 13.0, StataCorp, Texas, USA).
3.1. Differences of trabecular architecture within a spinal motion segment In the upper lumbar region, overall, the half vertebra cranial to a disc had greater BMD, greater trabecular bone volume fraction, more connections, a greater number of trabeculae and less trabecular separation than the corresponding vertebrae caudal to the disc (Table 2). Further, greater trabecular BMD, BV/TV, Tb.N, and Conn.D measurements were observed in the peripheral section, but not in the central section (Table 2). When the peripheral trabeculae were further divided into
Table 2 Differences in vertebral trabeculae for spinal motion segments from the upper lumbar region (N = 46). Vertebral trabeculae
3. Results
Measures
Two cadaveric lumbar spines were removed from data analysis due to missing data. As a result, 92 spinal motion segments (13 L1/L2, 15 L2/ L3, 18 L3/L4, 46 L4/L5 segments) from 46 cadaveric lumbar spines (mean age 50.0 ± 0.72 years, range 21–63 years) were included in the current study. There were 46 spinal motion segments from the upper lumbar region and 46 spinal motion segments from the lower lumbar region. Degree of disc degeneration was statistically significantly greater in the lower lumbar region than the upper lumbar region (χ2 test, P b 0.05 for both measurements) (Table 1). Greater age was associated Table 1 Disc degeneration measurements in relation to lumbar regions. Upper lumbar region
Lower lumbar region
Discographic disc degeneration⁎ None 10 (21.7%) Slight 20 (43.5%) Moderate 7 (15.2%) Severe 9 (19.6%) Disc space narrowing⁎⁎
5 (10.9%) 9 (19.6%) 13 (28.3%) 19 (41.2%)
None Slight to moderate Severe
20 (43.5%) 20 (43.5%) 4 (8.7%)
36 (78.3%) 6 (13.1%) 2 (4.3%)
The upper lumbar region includes L1/2, L2/3 and L3/4 discs; the lower lumbar region includes L4/5 and L5/S1 discs. χ2 tests were used. ⁎ P b 0.05. ⁎⁎ P b 0.01.
Half vertebrae
BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th Central section BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th Peripheral BMD section BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th Anterior BMD sub-region BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th Posterior BMD sub-region BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th
Cranial to disc
Caudal to disc
Difference
149.75 ± 17.99 0.12 ± 0.01 1.41 ± 0.29 1.01 ± 0.11 0.88 ± 0.11 0.12 ± 0.01 138.69 ± 23.80 0.12 ± 0.02 1.42 ± 0.45 1.03 ± 0.13 0.88 ± 0.14 0.11 ± 0.01 152.64 ± 17.02 0.13 ± 0.01 1.39 ± 0.27 1.01 ± 0.10 0.88 ± 0.11 0.13 ± 0.01 141.35 ± 21.24 0.12 ± 0.02 1.21 ± 0.29 0.98 ± 0.11 0.91 ± 0.15 0.12 ± 0.01 159.22 ± 18.71 0.13 ± 0.02 1.55 ± 0.35 1.06 ± 0.12 0.83 ± 0.12 0.13 ± 0.01
143.07 ± 21.51 0.12 ± 0.02 1.29 ± 0.28 0.98 ± 0.10 0.91 ± 0.10 0.12 ± 0.01 138.15 ± 23.46 0.12 ± 0.02 1.36 ± 0.40 1.01 ± 0.12 0.89 ± 0.13 0.11 ± 0.02 144.36 ± 21.58 0.12 ± 0.02 1.26 ± 0.25 0.97 ± 0.09 0.91 ± 0.10 0.12 ± 0.01 139.61 ± 26.98 0.12 ± 0.02 1.06 ± 0.28 0.93 ± 0.10 0.96 ± 0.13 0.13 ± 0.02 145.42 ± 23.07 0.12 ± 0.02 1.35 ± 0.28 1.00 ± 0.10 0.89 ± 0.10 0.12 ± 0.01
6.68 ± 13.42⁎⁎ 0.01 ± 0.01⁎⁎ 0.12 ± 0.16⁎⁎ 0.03 ± 0.06⁎⁎ −0.03 ± 0.06⁎⁎ 0.002 ± 0.001 0.54 ± 16.37 0.00 0.57 ± 0.04 0.02 ± 0.01 −0.02 ± 0.09 −0.002 ± 0.002 8.28 ± 13.90⁎⁎ 0.01 ± 0.01⁎⁎ 0.13 ± 0.16⁎⁎ 0.04 ± 0.06⁎⁎ −0.03 ± 0.06⁎⁎ 0.003 ± 0.01 1.74 ± 22.68 0 ± 0.02 0.15 ± 0.20⁎⁎ 0.05 ± 0.07⁎⁎ −0.05 ± 0.09⁎⁎ −0.01 ± 0.02 13.81 ± 8.68⁎⁎ 0.01 ± 0.01⁎⁎ 0.19 ± 0.24⁎⁎ 0.05 ± 0.09⁎⁎ −0.05 ± 0.09⁎⁎ 0.01 ± 0.01⁎⁎
Paired t-tests were used to examine the difference in the cranial and caudal trabeculae of defined sections. ⁎⁎ P b 0.01.
G. Yang et al. / Bone 95 (2017) 102–107
anterior and posterior sub-regions, greater trabecular parameters were observed in the cranial vertebrae than the caudal remained. For the lower lumbar region, however, similar findings were not observed and there were few statistically significant differences in the microstructural measurements between the cranial and caudal vertebrae (Table 3).
Table 4 Associations of cranio-caudal differences in trabecular parameters with age, BMI and discographic disc degeneration: results for upper lumbar vertebrae (N = 46). Discographic disc degeneration
Half vertebrae
3.2. Associations of cranio-caudal trabecular differences with age and disc degeneration Using regression models, the structural differences between cranial and caudal trabeculae within a spinal motion segment in the upper lumbar spine were not explained by age, BMI, or disc degeneration, as indicated from either discography (Table 4) or disc narrowing (Table 5). Similarly, the few statistically significant differences at L4/5 were not explained by age, BMI or disc degeneration (data not reported).
Peripheral section
Anterior sub-region
4. Discussion Using a large sample of lumbar spinal motion segments, microstructural differences of vertebral trabeculae cranial and caudal to the disc were investigated. In the upper lumbar region (L1/L4 spinal motion segments), vertebral trabeculae cranial to the disc had greater values of BMD, BV/TV, Tb.N, Conn.D and less Tb.Sp measurements than caudally. Moreover, this structural asymmetry lies in the peripheral region of the vertebra which connects to the annulus fibrosus of the disc, but not in the central portion of the vertebra to which the nucleus pulposus attaches. Similar findings, however, were not present in the L4/L5 spinal motion segment. In addition, the asymmetry of vertebral trabecular microstructure in the upper lumbar region was not related to age or disc degeneration, suggesting it may be an inherent spinal characteristic.
Table 3 Differences in vertebral trabeculae for spinal motion segments from the lower lumbar region (N = 46).
105
Posterior sub-region
Difference in
Age
BMI
No
Slight
Moderate
Severe
BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th
0.18 0.00 0.00 0.00 0.00 0.0 0.26 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.48 0.00 0.01 0.00 0.00 0.00
−0.03 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 −0.23 0.00 −0.01 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
1.49 0.00 0.02 0.02 −0.02 0.00 0.99 0.00 0.02 0.02 −0.02 0.00 6.15 0.01 −0.03 0.00 0.00 0.01 1.17 0.00 −0.09 −0.02 0.02 0.00
−3.63 0.00 0.06 0.02 0.00 0.00 −5.36 0.00 0.08 0.01 0.00 −0.01 −23.16⁎ 0.02 −0.12 −0.04 0.05 −0.01 −8.53 −0.01 0.06 −0.01 0.04 0.00
−5.56 0.00 −0.10 −0.03 0.03 0.00 −5.88 0.00 −0.08 −0.03 0.03 0.00 −5.42 0.00 −0.10 −0.04 0.05 0.00 −16.32 −0.01 −0.19 −0.08 0.08 0.00
Values are regression coefficients (coef.) in multivariable regression models. Differences in structural measurements are the dependent variables and no disc degeneration was the reference. ⁎ P b 0.05.
Together with previous findings that in the upper region endplates are thinner and vertebrae have less BMD than in the lower region, and the vertebral endplate cranial to a disc is thicker and has greater BMD than the caudal endplate [12,14], the findings help to explain why vertebral fractures are more common in the upper lumbar region and more frequently involve the endplate caudal to a disc.
Vertebral trabeculae
Half vertebrae
Central section
Peripheral section
Anterior sub-region
Posterior sub-region
Measures
Cranial to disc
Caudal to disc
Difference
BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th
148.80 ± 27.52 0.12 ± 0.02 1.28 ± 0.42 0.97 ± 0.13 0.93 ± 0.15 0.13 ± 0.02 135.57 ± 31.16 0.11 ± 0.03 1.31 ± 0.61 0.99 ± 0.17 0.93 ± 0.2 0.11 ± 0.02 151.86 ± 27.10 0.13 ± 0.02 1.26 ± 0.39 0.97 ± 0.13 0.92 ± 0.15 0.13 ± 0.02 146.24 ± 29.93 0.12 ± 0.03 1.14 ± 0.44 0.94 ± 0.02 0.96 ± 0.19 0.13 ± 0.02 155.89 ± 28.25 0.13 ± 0.02 1.32 ± 0.41 0.99 ± 0.14 0.90 ± 0.16 0.13 ± 0.02
153.39 ± 33.35 0.13 ± 0.03 1.34 ± 0.50 0.99 ± 0.19 0.97 ± 0.65 0.13 ± 0.03 136.46 ± 40.11 0.11 ± 0.03 1.37 ± 0.61 0.99 ± 0.21 1.07 ± 1.11 0.12 ± 0.20 158.47 ± 33.56 0.13 ± 0.03 1.31 ± 0.47 1.00 ± 0.18 0.94 ± 0.50 0.14 ± 0.03 154.63 ± 41.34 0.13 ± 0.03 1.12 ± 0.45 0.95 ± 0.18 1.00 ± 0.09 0.14 ± 0.03 160.07 ± 36.55 0.13 ± 0.03 1.43 ± 0.52 1.04 ± 0.18 0.89 ± 0.43 0.13 ± 0.02
−4.60 ± 24.41 −0.004 ± 0.02 −0.05 ± 0.24 −0.03 ± 0.10 −0.05 ± 0.58 −0.003 ± 0.03 −0.89 ± 29.36 −0.001 ± 0.02 −0.06 ± 0.31 −0.04 ± 0.10 −0.13 ± 1.0 −0.001 ± 0.02 −6.62 ± 26.05 −0.01 ± 0.02 −0.05 ± 0.24 −0.03 ± 0.10⁎ −0.02 ± 0.44 −0.004 ± 0.03 −8.39 ± 34.24 −0.007 ± 0.03 0.02 ± 0.25 −0.01 ± 0.11 −0.04 ± 0.53 −0.01 ± 0.003 −4.17 ± 28.81 −0.03 ± 0.02 −0.11 ± 0.35⁎ −0.06 ± 0.11⁎⁎ 0.01 ± 0.34 0.003 ± 0.02
Paired t-tests were used to examine the differences in the cranial and caudal vertebrae. ⁎ P b 0.05. ⁎⁎ P b 0.01.
Table 5 Associations of cranio-caudal differences in trabecular parameters with age, BMI and disc narrowing: results for upper lumbar vertebrae (N = 46). Disc narrowing Differences in Age Half vertebrae
Peripheral section
Anterior sub-region
Posterior sub-region
BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th
−0.08 0 0 0 0 0 −0.02 0 0.01 0 0 0 −0.16 0 0 0 0 0 0.09 0 0.01 0 0 0
BMI
No Slight to moderate Severe
0.02 0 −0.02 0 0 0 0.14 0 0 0 0 0 0.13 0 −0.01 0 0 0 0.12 0 −0.01 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
−1.09 0 −0.1 −0.02 0.01 0 −2.33 0 −0.08 −0.02 0.02 0 −0.78 0 −0.06 −0.02 0 −0.01 −15.46 −0.01 −0.12 −0.04 0.05 −0.01
14.63 0.01 0.08 0.05 −0.07 0.01 11.68 0.01 0.07 0.05 −0.06 0 3.93 0 0.01 0.02 −0.04 −0.02 11.41 0.01 0.1 0.05 −0.06 0
Values are regression coefficients (coef.) in multivariable regression models. Differences in structural measurements are the dependent variables and no disc narrowing was the reference.
106
G. Yang et al. / Bone 95 (2017) 102–107
It seems that trabeculae cranial and caudal to a disc should be different in structure, as cell densities in the precursors of centrums are different. The number of cells in the sclerotomes are important in vertebral development [5,8] and sclerotome cells are major source of chordacentrum mineralization [19]. As such, better trabecular quality was expected in the trabeculae caudal to a disc rather than that cranial as there are more sclerotome cells in an embryonic vertebra (Fig. 1). Better trabecular structure, however, was observed in the trabeculae cranial to a disc rather than caudally, and in the peripheral region of a vertebra, but not in the central region. It seems that some postnatal exposures, such as characteristic mechanical loading patterns in the upper lumbar spine, play a role in the structural differences of vertebral trabeculae within a spinal motion segment. Findings that greater microstructural measurements are present in the vertebra cranial rather than caudal to the disc and that such structural differences exist in the upper lumbar spine, but not in the lower lumbar spine, may have clinical implications. Previous research revealed that structural differences in a vertebral body may predispose the lumbar vertebra to fracture [4,20]. In the current study, asymmetric vertebral trabecular structure at the two sides of a disc are observed in the upper lumbar segments, but not in the lower lumbar region, help to explain why fractures more commonly involve L1–3 vertebrae rather than L4–5 vertebrae [2]. Further, relatively inferior structure in the trabeculae caudal to a disc partially explains why lumbar spine fractures commonly occur in the endplate caudal to a disc [1,2]. Similarly, the findings may also relate to the distribution pattern of Schmorl's nodes, which also typically involve the upper lumbar spine [21] and the vertebra caudally adjacent to a disc [5]. We postulate that better microstructure in the trabeculae cranial to a disc than that caudal in the upper but not the lower lumbar spine may be related to vertebra anatomy and sagittal alignment of the lumbar spine. For a lumbar vertebra, the superior portion of trabeculae, which are caudally adjacent to a disc, connect to posterior elements through a pair of pedicles. In the upper lumbar spine there is little lumbar lordosis [22], and the compressive loads the vertebral body experiences are partially transferred to the neural arch, which in return stress-shields the anterior portion of the vertebra [23]. For the L4/L5 segment, however, there is significant lumbar lordosis [22] and kinematic behaviors are different from the upper lumbar motion segments (e.g. contributes to more rotation, greater velocities than the upper lumbar spine) [24,25]. Increased lordosis and a different kinematic pattern in the L4/5 segment may offset the stress-shielding effect from the neural arches to certain degree, resulting in similar trabecular structures at two sides of a disc. Although degenerated or narrowed discs can change or alter mechanical distribution patterns [10,26], the vertebral trabecular asymmetry observed was independent of age and disc degeneration. Given the complicated stress distributions in the lumbar spine, which are also influenced by posture and endplate fractures [27,28], our theory cannot fully explain the structural asymmetries observed. To some degrees, it appears that the structural asymmetry of vertebral trabeculae is a natural feature of spinal motion segments in the upper lumbar spine. Stress distribution patterns inside an intervertebral disc are different in the nucleus pulposus and the annulus fibrosus [10]. In the current study, we identified that structural differences of vertebral trabeculae are mainly in the peripheral region where the annulus fibrosus attaches, but not in the central trabeculae adjacent to the nucleus pulposus. This likely because the nucleus pulposus is a fluid-like tissue which is able to distribute compressive stress evenly. Age and disc degeneration related structural changes within the annulus, however, may lead to a transfer of load from the nucleus to the posterior annulus [10], resulting in varied trabeculae structure there. This is the first study of the microstructural variation of vertebral trabeculae within a spinal motion segment. A large sample of lumbar vertebrae was studied using μCT, which is a reliable and accurate approach to measure structural parameters for vertebral trabeculae. Moreover, we used discography and disc space narrowing to measure the degree
of disc degeneration. In addition, measurements of trabeculae and disc from a spinal motion segment were matched in data analyses, allowing the study of a site-specific association. With respect to limitations, as only motion segments from middle-aged men were studied, it is possible that trabeculae structures are different in women. In addition, μCT data were obtained from dried specimens, which may differ from those obtained from fresh samples or in vivo. 5. Conclusions In summary, trabecular bone was found to have superior parameters immediately cranial to upper lumbar discs compared to bone that was caudal to them, no such asymmetry was found at L4/5 disc. Such asymmetry of vertebral trabeculae may be a natural feature of upper lumbar spinal motion segments, which is not related to age, BMI or disc degeneration. The observed structural asymmetry helps to explain why fractures and Schmorl's nodes most commonly occur in the upper lumbar region and in the trabeculae caudal to a lumbar disc. Conflict of interests All the authors declare that they have no conflict of interest. Acknowledgments This study was partially supported by National Natural Science Foundation of China (NSFC, 81371995). References [1] L.Y. Dai, X.Y. Wang, C.G. Wang, L.S. Jiang, H.Z. Xu, Bone mineral density of the thoracolumbar spine in relation to burst fractures: a quantitative computed tomography study, Eur. Spine J. 15 (2006) 1817–1822. [2] F. Magerl, M. Aebi, S.D. Gertzbein, J. Harms, S. Nazarian, A comprehensive classification of thoracic and lumbar injuries, Eur. Spine J. 3 (1994) 184–201. [3] P.A. Hulme, S.K. Boyd, S.J. Ferguson, Regional variation in vertebral bone morphology and its contribution to vertebral fracture strength, Bone 41 (2007) 946–957. [4] F.D. Zhao, P. Pollintine, B.D. Hole, M.A. Adams, P. Dolan, Vertebral fractures usually affect the cranial endplate because it is thinner and supported by less-dense trabecular bone, Bone 44 (2009) 372–379. [5] G. Dar, Y. Masharawi, S. Peleg, N. Steinberg, H. May, B. Medlej, et al., Schmorl's nodes distribution in the human spine and its possible etiology, Eur. Spine J. 19 (2010) 670–675. [6] B. Brand-Saberi, B. Christ, Evolution and development of distinct cell lineages derived from Somites, Curr. Top. Dev. Biol. 48 (1999) 1–42. [7] K.M. Kaplan, J.M. Spivak, J.A. Bendo, Embryology of the spine and associated congenital abnormalities, Spine 5 (2005) 564–576. [8] W. Bernd, B. Anita, H. Ann, R. Joerg, W.P. Eckhard, W. Christoph, Conditional ablation of osteoblasts in medaka, Dev. Biol. 364 (2012) 128–137. [9] C. Cooper, M.K. Javaid, P. Taylor, K. Walker-Bone, E. Dennison, N. Arden, The fetal origins of osteoporotic fracture, Calcif. Tissue Int. 70 (2002) 391–394. [10] M.A. Adams, D.S. Mcnally, P. Dolan, ‘Stress’ distributions inside intervertebral discs. The effects of age and degeneration, J. Bone Joint Surg. Br. 78 (1996) 81–87. [11] P. Pollintine, P. Dolan, J.H. Tobias, M.A. Adams, Intervertebral disc degeneration can lead to “stress-shielding” of the anterior vertebral body: a cause of osteoporotic vertebral fracture? Spine 29 (2004) 774–782. [12] Y. Wang, M.C. Battie, S.K. Boyd, T. Videman, The osseous endplates in lumbar vertebrae: thickness, bone mineral density and their associations with age and disk degeneration, Bone 48 (2011) 804–809. [13] Y. Wang, M.C. Battie, T. Videman, A morphological study of lumbar vertebral endplates: radiographic, visual and digital measurements, Eur. Spine J. 21 (2012) 2316–2323. [14] Y. Wang, S.K. Boyd, M.C. Battie, Y. Yasui, T. Videman, Is greater lumbar vertebral BMD associated with more disk degeneration? A study using microCT and discography, J. Bone Miner. Res. 26 (2011) 2785–2791. [15] T. Videman, M. Nurminen, J.D. Troup, Lumbar spinal pathology in cadaveric material in relation to history of back pain, occupation, and physical loading, Spine 15 (1990) 728–740. [16] Y. Wang, T. Videman, M.C. Battie, Morphometrics and lesions of vertebral end plates are associated with lumbar disc degeneration: evidence from cadaveric spines, J. Bone Joint Surg. Am. 95 (e26) (2013) 1–7. [17] Y. Wang, J.S. Owoc, S.K. Boyd, T. Videman, M.C. Battie, Regional variations in trabecular architecture of the lumbar vertebra: associations with age, disc degeneration and disc space narrowing, Bone 56 (2013) 249–254. [18] H.R. Buie, G.M. Campbell, R.J. Klinck, J.A. Macneil, S.K. Boyd, Automatic segmentation of cortical and trabecular compartments based on a dual threshold technique for in vivo micro-CT bone analysis, Bone 41 (2007) 505–515.
G. Yang et al. / Bone 95 (2017) 102–107 [19] W. Bernd, H. Seintsch, R. Amstad, G. Burri, V. Weber, Organization model for postoperative pain management in a basic-care hospital, Anaesthesist 53 (2004) 531–542. [20] M.A. Adams, P. Dolan, Biomechanics of vertebral compression fractures and clinical application, Arch. Orthop. Trauma Surg. 131 (2011) 1703–1710. [21] W. Yue, T. Videman, M.C. Battié, Modic changes: prevalence, distribution patterns, and association with age in white men, Spine 12 (2012) 411–416. [22] E.S. Lee, C.W. Ko, S.W. Suh, S. Kumar, I.K. Kang, J.H. Yang, The effect of age on sagittal plane profile of the lumbar spine according to standing, supine, and various sitting positions, J. Orthop. Surg. Res. 9 (2014) 1–10. [23] M.A. Adams, P. Pollintine, J.H. Tobias, G.K. Wakley, P. Dolan, Intervertebral disc degeneration can predispose to anterior vertebral fractures in the thoracolumbar spine, J. Bone Miner. Res. 21 (2006) 1409–1416.
107
[24] S. Parkinson, A. Campbell, W. Dankaerts, A. Burnett, P. O'Sullivan, Upper and lower lumbar segments move differently during sit-to-stand, Man. Ther. 18 (2013) 390–394. [25] R.S. Alqhtani, M.D. Jones, P.S. Theobald, J.M. Williams, Investigating the contribution of the upper and lower lumbar spine, relative to hip motion, in everyday tasks, Man. Ther. 21 (2016) 268–273. [26] R.C. Mulholland, The myth of lumbar instability: the importance of abnormal loading as a cause of low back pain, Eur. Spine J. 17 (2008) 619–625. [27] P. Pollintine, A.S. Przybyla, P. Dolan, M.A. Adams, Neural arch load-bearing in old and degenerated spines, J. Biomech. 37 (2004) 197–204. [28] P.R. Landham, H.L.A. Baker-Rand, S.J. Gilbert, P. Pollintine, D.J. Annesley-Williams, M.A. Adams, et al., Is kyphoplasty better than vertebroplasty at restoring form and function after severe vertebral wedge fractures? Spine J. 15 (2015) 721–732.
Contents lists available at ScienceDirect
Bone journal homepage: www.elsevier.com/locate/bone
Full Length Article
Cranio-caudal asymmetries in trabecular architecture reflect vertebral fracture patterns Ge Yang a, Michele C. Battié b, Steven K. Boyd c, Tapio Videman b, Yue Wang a,⁎ a b c
Spine lab, Department of Orthopedic Surgery, The First Affiliated Hospital of Zhejiang University, Hangzhou, China Faculty of Rehabilitation Medicine, University of Alberta, Edmonton, Canada Cumming School of Medicine, University of Calgary, Calgary, Canada
a r t i c l e
i n f o
Article history: Received 26 August 2016 Revised 15 November 2016 Accepted 18 November 2016 Available online 19 November 2016 Keywords: Lumbar spine Vertebral fractures Trabecular microstructure μCT Disc degeneration
a b s t r a c t Clinically, vertebral fractures often occur in the upper lumbar spine and involve the superior endplate of a vertebra (which is immediately caudal to a disc). Knowledge that the cranial endplate of a disc is thicker and has greater bone mineral density (BMD) than the corresponding caudal endplate helps to explain this phenomenon. In this study, we investigated structural differences in vertebral trabeculae on either side of a lumbar disc to provide further insight into vertebral fracture risk. As the focus is trabecular difference within a spinal motion segment, we define cranial and caudal vertebral trabeculae relative to the disc. Ninety-two spinal motion segments from 46 cadaveric lumbar spines (males, mean age 50 years, range 21–63 years) were studied. Disc narrowing on radiography and spread of barium sulfate (BaSO4) on discography were measured to indicate disc degeneration. Micro-computed tomography (μCT) images were obtained at a resolution of 82 μm for each vertebra and processed to include only vertebral trabeculae. Using image processing, the vertebral trabeculae were divided into superior and inferior halves, and then into central and peripheral regions which were approximately opposite to the disc pulposus and annulus, and further into anterior and posterior sub-regions. Microarchitecture measurements for each vertebral region were obtained to determine the differences between the cranial and caudal trabeculae (relative to disc) and their associations with age and disc degeneration within each spinal motion segment. Data from the upper (L1/2–L3/4) and lower (L4/5) lumbar segments were analyzed separately. In the upper lumbar region, the trabeculae cranial to a disc on average had 5.3% greater BMD and trabecular bone volume, 3.6% greater trabecular number, 9.7% greater connectivity density, and 3.7% less trabecular separation than the corresponding caudal trabeculae (P b 0.05 for all). Similar trends were observed in peripheral, anterior and posterior regions, but not in central region. No structural difference was observed in the trabeculae of L4/5 segment. Structural asymmetries of vertebral trabeculae were not associated with age, disc degeneration, or disc narrowing. Vertebral trabecular parameters cranial to the disc were greater than caudally in the upper but not in the lower lumbar region. Findings further explain why vertebral fractures are more common in the upper lumbar region and more frequently involve the endplate caudal to a disc. © 2016 Elsevier Inc. All rights reserved.
1. Introduction Traumatic and osteoporotic vertebral fractures are among the most common skeletal injuries in adults. Interestingly, thoracolumbar vertebral fractures involve the superior portion of the vertebral body, including the endplate and trabeculae, more often than the inferior portion [1,2]. This clinical phenomenon may be partially explained by previous findings that the superior endplate of a vertebra is thinner and the trabeculae are less dense in the superior portion of the vertebra than inferiorly [3,4]. While the reason for the structural differences within a vertebral body remains unexplained, the superior and inferior portions ⁎ Corresponding author at: 79# Qingchun Road, Hangzhou 310003, China. E-mail address: [email protected] (Y. Wang).
http://dx.doi.org/10.1016/j.bone.2016.11.018 8756-3282/© 2016 Elsevier Inc. All rights reserved.
of a vertebra originate from different sclerotomes that fuse together in a developmental process called “resegmentation” [5]. Yet, the trabecular structure on either side of the disc within a spinal motion segment also seems to be considerably different, although they are derived from one sclerotome. A spinal motion segment, including two adjacent vertebrae and the intervening intervertebral disc, is often studied as a working unit in spine mechanics. During resegmentation, portions of two adjacent sclerotomes, a cranial one with densely packed sclerotome cells and a caudal one with loosely packed cells, fuse into a vertebral precursor. In other words, within a spinal motion segment, the half of the vertebral body cranial to a disc has less cell density than the half of the vertebral body caudal to the disc when the embryonic spine is formed (Fig. 1) [6,7]. Given the fact that the sclerotome cells are the major source of chordacentrum
G. Yang et al. / Bone 95 (2017) 102–107
103
Fig. 1. Sclerotomes resegmentation. A: In a sclerotome, cells are loosely packed in the cranial portion but densely packed in the caudal portion. B: In sclerotomes resegmentation, notochord expanded to form intervertebral discs, which separate adjacent vertebrae. In an embryonic spinal segment, therefore, vertebra immediately caudal to a disc have densely packed cells and that cranial to a disc have loosely packed cells. (Adapted from Dar G. et al. [5]).
mineralization [8] and the early period of ontogeny has a substantial influence on bone quality and structure [9], theoretically, vertebral trabeculae caudal to a disc could be expected to have better bone quality than the corresponding cranial trabeculae. On the other hand, a healthy intervertebral disc is able to distribute mechanical loads upward and downward evenly, and the endplates and trabeculae cranial and caudal to a disc overall share equivalent amounts of loads. Age related disc degeneration, however, may alter stress distribution within a spinal motion segment [10,11]. Our previous studies revealed that the endplate cranial to a disc is more concave, thicker, and of higher bone mineral density (BMD) than the caudal endplate [12,13]. In addition, we observed that a greater degree of disc degeneration was associated with greater BMD in the vertebra caudally adjacent to disc, but not in the cranial vertebra [14]. Taking evidence together, we hypothesized that within a spinal motion segment there is also an asymmetry of vertebral trabecular structure on either side of the disc. In this study, we further sought to determine if the asymmetry of trabecular structure is associated with aging or disc degeneration, or merely a design feature of the lumbar spine. As the focus of this paper is trabecular difference within a spinal motion segment, we define cranial and caudal vertebral trabeculae relative to the disc, if not specified. 2. Materials and methods 2.1. Sample We used a lumbar spine archive comprised of 149 Caucasian male cadavers, with a mean age of 50 years at the time of death and a range from 21 to 64 years. The archive included men under the age of 65 years, employed up to the time of death, who passed away with a short illness history [15]. After a routine autopsy examination, X-ray images and discography of the lumbar spine were performed. Then, spinal ligaments, muscles and intervertebral discs were removed and lumbar vertebrae were washed, processed, dried and preserved at room temperature. As some archived vertebrae data were lost, we selected and obtained micro-computed tomography (μCT) scans of 150 lumbar vertebrae from 48 of the cadavers, which are with adjacent disc degeneration data, to study the interactions between vertebra and disc [14]. Using the same sample, the purpose of the present study was to investigate differences in vertebral trabeculae within a spinal motion
segment. Thus, complete motion segments with data of both vertebrae and the intervening disc were required. The research was approved by the Health Research Ethics Board at the University of Alberta. 2.2. Measurements of disc degeneration Disc degeneration was measured using discography and disc space narrowing. Discography was performed by injecting 2–5 ml of barium sulfate (BaSO4) anteriorly into the center of an intervertebral disc. According to the spread of BaSO4 on discogram, a 4-grade ordinal scale (none, slight, moderate or severe) was used to rate the degree of disc degeneration, as reported previously [16]. Disc space narrowing was assessed from lateral X-ray radiographs using a three-point scale as none, slight to moderate, or severe. The reliability of the disc degeneration measurements was good (intra-rater kappa = 0.81 for discographic degeneration and kappa = 0.71 for disc space narrowing measurements) [17]. 2.3. Measurement of vertebral trabeculae microstructure Each vertebra was scanned using a μCT system (XtremeCT, Scanco Medical, Brüttisellen, Switzerland) with standard scanning parameters: 60 kVp, 1000 μA, 200 ms integration time, and 750 projections. The dried vertebrae were scanned with a nominal isotropic resolution of 82 μm (field of view 125 mm, 1536 × 1536 pixels) in air. Regions of interest were identified for the vertebral body in each microradiograph and contoured using a semi-automated contouring method [18]. In so doing, only the trabeculae of the vertebral body were included in data analysis and other components of the vertebra, such as the cortex of the vertebral body and posterior elements of the vertebra, were excluded. Using a technique we previously reported [17], each vertebral body was first divided into superior and inferior halves. Then, each half vertebra was further divided into central and peripheral regions, which were approximately corresponding to the areas connecting to nucleus pulposus and annulus fibrosus. Further, the peripheral region was segmented into anterior and posterior sub-regions (Fig. 2). Structural and densitometric analyses were performed (Image Processing Language, v5.15, Scanco Medical AG, Brüttisellen, Switzerland) to obtain trabecular bone volume fraction (BV/TV; %), trabecular number (Tb.N; 1/mm), BMD (mg/cm3), trabecular thickness (Tb.Th; mm),
104
G. Yang et al. / Bone 95 (2017) 102–107
Fig. 2. Schematic diagram illustrating the division of vertebral trabeculae for a spinal motion segment. The trabeculae of a vertebral body were separated as superior and inferior two halves, which were referred as cranial or caudal trabeculae adjacent to the disc. Each half was divided into central and peripheral sections, corresponding to the regions connecting to nucleus pulpous and annular fibrosus, respectively. The peripheral section of the vertebral trabeculae was further segmented into anterior and posterior sub-regions.
trabecular separation (Tb.Sp; mm) and connectivity density (Conn.D; 1/ mm3) parameters for each defined section.
with greater discographic disc degeneration (OR = 1.66 for each 10 years, P b 0.01), greater disc narrowing (OR = 1.32 for each 10 years, P b 0.01), and less vertebral BMD (coef. = −0.99, P b 0.01).
2.4. Statistical analysis As spinal biomechanics and prevalence of disc degeneration are considerably different between the upper lumbar region (L1/2, L2/3 and L3/ 4 discs) and lower lumbar region (L4/5 and L5/S1 discs), data for each lumbar region were analyzed separately. Paired t-tests were used to compare the differences of trabecular microstructures between cranial and caudal trabeculae within a spinal motion segment. Taking L4/L5 segment as an example, we matched trabecular measurements of the inferior half of the L4 vertebra with those of the superior half of L5 vertebra. The differences in regional architectural parameters were presented as the percentage of difference between cranial and caudal vertebrae. Multivariable regression was used to examine the associations between trabecular differences and age, body mass index (BMI; kg/cm2), disc degeneration and disc space narrowing. Data analysis was performed using Stata (Release 13.0, StataCorp, Texas, USA).
3.1. Differences of trabecular architecture within a spinal motion segment In the upper lumbar region, overall, the half vertebra cranial to a disc had greater BMD, greater trabecular bone volume fraction, more connections, a greater number of trabeculae and less trabecular separation than the corresponding vertebrae caudal to the disc (Table 2). Further, greater trabecular BMD, BV/TV, Tb.N, and Conn.D measurements were observed in the peripheral section, but not in the central section (Table 2). When the peripheral trabeculae were further divided into
Table 2 Differences in vertebral trabeculae for spinal motion segments from the upper lumbar region (N = 46). Vertebral trabeculae
3. Results
Measures
Two cadaveric lumbar spines were removed from data analysis due to missing data. As a result, 92 spinal motion segments (13 L1/L2, 15 L2/ L3, 18 L3/L4, 46 L4/L5 segments) from 46 cadaveric lumbar spines (mean age 50.0 ± 0.72 years, range 21–63 years) were included in the current study. There were 46 spinal motion segments from the upper lumbar region and 46 spinal motion segments from the lower lumbar region. Degree of disc degeneration was statistically significantly greater in the lower lumbar region than the upper lumbar region (χ2 test, P b 0.05 for both measurements) (Table 1). Greater age was associated Table 1 Disc degeneration measurements in relation to lumbar regions. Upper lumbar region
Lower lumbar region
Discographic disc degeneration⁎ None 10 (21.7%) Slight 20 (43.5%) Moderate 7 (15.2%) Severe 9 (19.6%) Disc space narrowing⁎⁎
5 (10.9%) 9 (19.6%) 13 (28.3%) 19 (41.2%)
None Slight to moderate Severe
20 (43.5%) 20 (43.5%) 4 (8.7%)
36 (78.3%) 6 (13.1%) 2 (4.3%)
The upper lumbar region includes L1/2, L2/3 and L3/4 discs; the lower lumbar region includes L4/5 and L5/S1 discs. χ2 tests were used. ⁎ P b 0.05. ⁎⁎ P b 0.01.
Half vertebrae
BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th Central section BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th Peripheral BMD section BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th Anterior BMD sub-region BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th Posterior BMD sub-region BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th
Cranial to disc
Caudal to disc
Difference
149.75 ± 17.99 0.12 ± 0.01 1.41 ± 0.29 1.01 ± 0.11 0.88 ± 0.11 0.12 ± 0.01 138.69 ± 23.80 0.12 ± 0.02 1.42 ± 0.45 1.03 ± 0.13 0.88 ± 0.14 0.11 ± 0.01 152.64 ± 17.02 0.13 ± 0.01 1.39 ± 0.27 1.01 ± 0.10 0.88 ± 0.11 0.13 ± 0.01 141.35 ± 21.24 0.12 ± 0.02 1.21 ± 0.29 0.98 ± 0.11 0.91 ± 0.15 0.12 ± 0.01 159.22 ± 18.71 0.13 ± 0.02 1.55 ± 0.35 1.06 ± 0.12 0.83 ± 0.12 0.13 ± 0.01
143.07 ± 21.51 0.12 ± 0.02 1.29 ± 0.28 0.98 ± 0.10 0.91 ± 0.10 0.12 ± 0.01 138.15 ± 23.46 0.12 ± 0.02 1.36 ± 0.40 1.01 ± 0.12 0.89 ± 0.13 0.11 ± 0.02 144.36 ± 21.58 0.12 ± 0.02 1.26 ± 0.25 0.97 ± 0.09 0.91 ± 0.10 0.12 ± 0.01 139.61 ± 26.98 0.12 ± 0.02 1.06 ± 0.28 0.93 ± 0.10 0.96 ± 0.13 0.13 ± 0.02 145.42 ± 23.07 0.12 ± 0.02 1.35 ± 0.28 1.00 ± 0.10 0.89 ± 0.10 0.12 ± 0.01
6.68 ± 13.42⁎⁎ 0.01 ± 0.01⁎⁎ 0.12 ± 0.16⁎⁎ 0.03 ± 0.06⁎⁎ −0.03 ± 0.06⁎⁎ 0.002 ± 0.001 0.54 ± 16.37 0.00 0.57 ± 0.04 0.02 ± 0.01 −0.02 ± 0.09 −0.002 ± 0.002 8.28 ± 13.90⁎⁎ 0.01 ± 0.01⁎⁎ 0.13 ± 0.16⁎⁎ 0.04 ± 0.06⁎⁎ −0.03 ± 0.06⁎⁎ 0.003 ± 0.01 1.74 ± 22.68 0 ± 0.02 0.15 ± 0.20⁎⁎ 0.05 ± 0.07⁎⁎ −0.05 ± 0.09⁎⁎ −0.01 ± 0.02 13.81 ± 8.68⁎⁎ 0.01 ± 0.01⁎⁎ 0.19 ± 0.24⁎⁎ 0.05 ± 0.09⁎⁎ −0.05 ± 0.09⁎⁎ 0.01 ± 0.01⁎⁎
Paired t-tests were used to examine the difference in the cranial and caudal trabeculae of defined sections. ⁎⁎ P b 0.01.
G. Yang et al. / Bone 95 (2017) 102–107
anterior and posterior sub-regions, greater trabecular parameters were observed in the cranial vertebrae than the caudal remained. For the lower lumbar region, however, similar findings were not observed and there were few statistically significant differences in the microstructural measurements between the cranial and caudal vertebrae (Table 3).
Table 4 Associations of cranio-caudal differences in trabecular parameters with age, BMI and discographic disc degeneration: results for upper lumbar vertebrae (N = 46). Discographic disc degeneration
Half vertebrae
3.2. Associations of cranio-caudal trabecular differences with age and disc degeneration Using regression models, the structural differences between cranial and caudal trabeculae within a spinal motion segment in the upper lumbar spine were not explained by age, BMI, or disc degeneration, as indicated from either discography (Table 4) or disc narrowing (Table 5). Similarly, the few statistically significant differences at L4/5 were not explained by age, BMI or disc degeneration (data not reported).
Peripheral section
Anterior sub-region
4. Discussion Using a large sample of lumbar spinal motion segments, microstructural differences of vertebral trabeculae cranial and caudal to the disc were investigated. In the upper lumbar region (L1/L4 spinal motion segments), vertebral trabeculae cranial to the disc had greater values of BMD, BV/TV, Tb.N, Conn.D and less Tb.Sp measurements than caudally. Moreover, this structural asymmetry lies in the peripheral region of the vertebra which connects to the annulus fibrosus of the disc, but not in the central portion of the vertebra to which the nucleus pulposus attaches. Similar findings, however, were not present in the L4/L5 spinal motion segment. In addition, the asymmetry of vertebral trabecular microstructure in the upper lumbar region was not related to age or disc degeneration, suggesting it may be an inherent spinal characteristic.
Table 3 Differences in vertebral trabeculae for spinal motion segments from the lower lumbar region (N = 46).
105
Posterior sub-region
Difference in
Age
BMI
No
Slight
Moderate
Severe
BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th
0.18 0.00 0.00 0.00 0.00 0.0 0.26 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.48 0.00 0.01 0.00 0.00 0.00
−0.03 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 −0.23 0.00 −0.01 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
1.49 0.00 0.02 0.02 −0.02 0.00 0.99 0.00 0.02 0.02 −0.02 0.00 6.15 0.01 −0.03 0.00 0.00 0.01 1.17 0.00 −0.09 −0.02 0.02 0.00
−3.63 0.00 0.06 0.02 0.00 0.00 −5.36 0.00 0.08 0.01 0.00 −0.01 −23.16⁎ 0.02 −0.12 −0.04 0.05 −0.01 −8.53 −0.01 0.06 −0.01 0.04 0.00
−5.56 0.00 −0.10 −0.03 0.03 0.00 −5.88 0.00 −0.08 −0.03 0.03 0.00 −5.42 0.00 −0.10 −0.04 0.05 0.00 −16.32 −0.01 −0.19 −0.08 0.08 0.00
Values are regression coefficients (coef.) in multivariable regression models. Differences in structural measurements are the dependent variables and no disc degeneration was the reference. ⁎ P b 0.05.
Together with previous findings that in the upper region endplates are thinner and vertebrae have less BMD than in the lower region, and the vertebral endplate cranial to a disc is thicker and has greater BMD than the caudal endplate [12,14], the findings help to explain why vertebral fractures are more common in the upper lumbar region and more frequently involve the endplate caudal to a disc.
Vertebral trabeculae
Half vertebrae
Central section
Peripheral section
Anterior sub-region
Posterior sub-region
Measures
Cranial to disc
Caudal to disc
Difference
BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th
148.80 ± 27.52 0.12 ± 0.02 1.28 ± 0.42 0.97 ± 0.13 0.93 ± 0.15 0.13 ± 0.02 135.57 ± 31.16 0.11 ± 0.03 1.31 ± 0.61 0.99 ± 0.17 0.93 ± 0.2 0.11 ± 0.02 151.86 ± 27.10 0.13 ± 0.02 1.26 ± 0.39 0.97 ± 0.13 0.92 ± 0.15 0.13 ± 0.02 146.24 ± 29.93 0.12 ± 0.03 1.14 ± 0.44 0.94 ± 0.02 0.96 ± 0.19 0.13 ± 0.02 155.89 ± 28.25 0.13 ± 0.02 1.32 ± 0.41 0.99 ± 0.14 0.90 ± 0.16 0.13 ± 0.02
153.39 ± 33.35 0.13 ± 0.03 1.34 ± 0.50 0.99 ± 0.19 0.97 ± 0.65 0.13 ± 0.03 136.46 ± 40.11 0.11 ± 0.03 1.37 ± 0.61 0.99 ± 0.21 1.07 ± 1.11 0.12 ± 0.20 158.47 ± 33.56 0.13 ± 0.03 1.31 ± 0.47 1.00 ± 0.18 0.94 ± 0.50 0.14 ± 0.03 154.63 ± 41.34 0.13 ± 0.03 1.12 ± 0.45 0.95 ± 0.18 1.00 ± 0.09 0.14 ± 0.03 160.07 ± 36.55 0.13 ± 0.03 1.43 ± 0.52 1.04 ± 0.18 0.89 ± 0.43 0.13 ± 0.02
−4.60 ± 24.41 −0.004 ± 0.02 −0.05 ± 0.24 −0.03 ± 0.10 −0.05 ± 0.58 −0.003 ± 0.03 −0.89 ± 29.36 −0.001 ± 0.02 −0.06 ± 0.31 −0.04 ± 0.10 −0.13 ± 1.0 −0.001 ± 0.02 −6.62 ± 26.05 −0.01 ± 0.02 −0.05 ± 0.24 −0.03 ± 0.10⁎ −0.02 ± 0.44 −0.004 ± 0.03 −8.39 ± 34.24 −0.007 ± 0.03 0.02 ± 0.25 −0.01 ± 0.11 −0.04 ± 0.53 −0.01 ± 0.003 −4.17 ± 28.81 −0.03 ± 0.02 −0.11 ± 0.35⁎ −0.06 ± 0.11⁎⁎ 0.01 ± 0.34 0.003 ± 0.02
Paired t-tests were used to examine the differences in the cranial and caudal vertebrae. ⁎ P b 0.05. ⁎⁎ P b 0.01.
Table 5 Associations of cranio-caudal differences in trabecular parameters with age, BMI and disc narrowing: results for upper lumbar vertebrae (N = 46). Disc narrowing Differences in Age Half vertebrae
Peripheral section
Anterior sub-region
Posterior sub-region
BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th BMD BV/TV (%) Conn.D Tb.N Tb.Sp Tb.Th
−0.08 0 0 0 0 0 −0.02 0 0.01 0 0 0 −0.16 0 0 0 0 0 0.09 0 0.01 0 0 0
BMI
No Slight to moderate Severe
0.02 0 −0.02 0 0 0 0.14 0 0 0 0 0 0.13 0 −0.01 0 0 0 0.12 0 −0.01 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
−1.09 0 −0.1 −0.02 0.01 0 −2.33 0 −0.08 −0.02 0.02 0 −0.78 0 −0.06 −0.02 0 −0.01 −15.46 −0.01 −0.12 −0.04 0.05 −0.01
14.63 0.01 0.08 0.05 −0.07 0.01 11.68 0.01 0.07 0.05 −0.06 0 3.93 0 0.01 0.02 −0.04 −0.02 11.41 0.01 0.1 0.05 −0.06 0
Values are regression coefficients (coef.) in multivariable regression models. Differences in structural measurements are the dependent variables and no disc narrowing was the reference.
106
G. Yang et al. / Bone 95 (2017) 102–107
It seems that trabeculae cranial and caudal to a disc should be different in structure, as cell densities in the precursors of centrums are different. The number of cells in the sclerotomes are important in vertebral development [5,8] and sclerotome cells are major source of chordacentrum mineralization [19]. As such, better trabecular quality was expected in the trabeculae caudal to a disc rather than that cranial as there are more sclerotome cells in an embryonic vertebra (Fig. 1). Better trabecular structure, however, was observed in the trabeculae cranial to a disc rather than caudally, and in the peripheral region of a vertebra, but not in the central region. It seems that some postnatal exposures, such as characteristic mechanical loading patterns in the upper lumbar spine, play a role in the structural differences of vertebral trabeculae within a spinal motion segment. Findings that greater microstructural measurements are present in the vertebra cranial rather than caudal to the disc and that such structural differences exist in the upper lumbar spine, but not in the lower lumbar spine, may have clinical implications. Previous research revealed that structural differences in a vertebral body may predispose the lumbar vertebra to fracture [4,20]. In the current study, asymmetric vertebral trabecular structure at the two sides of a disc are observed in the upper lumbar segments, but not in the lower lumbar region, help to explain why fractures more commonly involve L1–3 vertebrae rather than L4–5 vertebrae [2]. Further, relatively inferior structure in the trabeculae caudal to a disc partially explains why lumbar spine fractures commonly occur in the endplate caudal to a disc [1,2]. Similarly, the findings may also relate to the distribution pattern of Schmorl's nodes, which also typically involve the upper lumbar spine [21] and the vertebra caudally adjacent to a disc [5]. We postulate that better microstructure in the trabeculae cranial to a disc than that caudal in the upper but not the lower lumbar spine may be related to vertebra anatomy and sagittal alignment of the lumbar spine. For a lumbar vertebra, the superior portion of trabeculae, which are caudally adjacent to a disc, connect to posterior elements through a pair of pedicles. In the upper lumbar spine there is little lumbar lordosis [22], and the compressive loads the vertebral body experiences are partially transferred to the neural arch, which in return stress-shields the anterior portion of the vertebra [23]. For the L4/L5 segment, however, there is significant lumbar lordosis [22] and kinematic behaviors are different from the upper lumbar motion segments (e.g. contributes to more rotation, greater velocities than the upper lumbar spine) [24,25]. Increased lordosis and a different kinematic pattern in the L4/5 segment may offset the stress-shielding effect from the neural arches to certain degree, resulting in similar trabecular structures at two sides of a disc. Although degenerated or narrowed discs can change or alter mechanical distribution patterns [10,26], the vertebral trabecular asymmetry observed was independent of age and disc degeneration. Given the complicated stress distributions in the lumbar spine, which are also influenced by posture and endplate fractures [27,28], our theory cannot fully explain the structural asymmetries observed. To some degrees, it appears that the structural asymmetry of vertebral trabeculae is a natural feature of spinal motion segments in the upper lumbar spine. Stress distribution patterns inside an intervertebral disc are different in the nucleus pulposus and the annulus fibrosus [10]. In the current study, we identified that structural differences of vertebral trabeculae are mainly in the peripheral region where the annulus fibrosus attaches, but not in the central trabeculae adjacent to the nucleus pulposus. This likely because the nucleus pulposus is a fluid-like tissue which is able to distribute compressive stress evenly. Age and disc degeneration related structural changes within the annulus, however, may lead to a transfer of load from the nucleus to the posterior annulus [10], resulting in varied trabeculae structure there. This is the first study of the microstructural variation of vertebral trabeculae within a spinal motion segment. A large sample of lumbar vertebrae was studied using μCT, which is a reliable and accurate approach to measure structural parameters for vertebral trabeculae. Moreover, we used discography and disc space narrowing to measure the degree
of disc degeneration. In addition, measurements of trabeculae and disc from a spinal motion segment were matched in data analyses, allowing the study of a site-specific association. With respect to limitations, as only motion segments from middle-aged men were studied, it is possible that trabeculae structures are different in women. In addition, μCT data were obtained from dried specimens, which may differ from those obtained from fresh samples or in vivo. 5. Conclusions In summary, trabecular bone was found to have superior parameters immediately cranial to upper lumbar discs compared to bone that was caudal to them, no such asymmetry was found at L4/5 disc. Such asymmetry of vertebral trabeculae may be a natural feature of upper lumbar spinal motion segments, which is not related to age, BMI or disc degeneration. The observed structural asymmetry helps to explain why fractures and Schmorl's nodes most commonly occur in the upper lumbar region and in the trabeculae caudal to a lumbar disc. Conflict of interests All the authors declare that they have no conflict of interest. Acknowledgments This study was partially supported by National Natural Science Foundation of China (NSFC, 81371995). References [1] L.Y. Dai, X.Y. Wang, C.G. Wang, L.S. Jiang, H.Z. Xu, Bone mineral density of the thoracolumbar spine in relation to burst fractures: a quantitative computed tomography study, Eur. Spine J. 15 (2006) 1817–1822. [2] F. Magerl, M. Aebi, S.D. Gertzbein, J. Harms, S. Nazarian, A comprehensive classification of thoracic and lumbar injuries, Eur. Spine J. 3 (1994) 184–201. [3] P.A. Hulme, S.K. Boyd, S.J. Ferguson, Regional variation in vertebral bone morphology and its contribution to vertebral fracture strength, Bone 41 (2007) 946–957. [4] F.D. Zhao, P. Pollintine, B.D. Hole, M.A. Adams, P. Dolan, Vertebral fractures usually affect the cranial endplate because it is thinner and supported by less-dense trabecular bone, Bone 44 (2009) 372–379. [5] G. Dar, Y. Masharawi, S. Peleg, N. Steinberg, H. May, B. Medlej, et al., Schmorl's nodes distribution in the human spine and its possible etiology, Eur. Spine J. 19 (2010) 670–675. [6] B. Brand-Saberi, B. Christ, Evolution and development of distinct cell lineages derived from Somites, Curr. Top. Dev. Biol. 48 (1999) 1–42. [7] K.M. Kaplan, J.M. Spivak, J.A. Bendo, Embryology of the spine and associated congenital abnormalities, Spine 5 (2005) 564–576. [8] W. Bernd, B. Anita, H. Ann, R. Joerg, W.P. Eckhard, W. Christoph, Conditional ablation of osteoblasts in medaka, Dev. Biol. 364 (2012) 128–137. [9] C. Cooper, M.K. Javaid, P. Taylor, K. Walker-Bone, E. Dennison, N. Arden, The fetal origins of osteoporotic fracture, Calcif. Tissue Int. 70 (2002) 391–394. [10] M.A. Adams, D.S. Mcnally, P. Dolan, ‘Stress’ distributions inside intervertebral discs. The effects of age and degeneration, J. Bone Joint Surg. Br. 78 (1996) 81–87. [11] P. Pollintine, P. Dolan, J.H. Tobias, M.A. Adams, Intervertebral disc degeneration can lead to “stress-shielding” of the anterior vertebral body: a cause of osteoporotic vertebral fracture? Spine 29 (2004) 774–782. [12] Y. Wang, M.C. Battie, S.K. Boyd, T. Videman, The osseous endplates in lumbar vertebrae: thickness, bone mineral density and their associations with age and disk degeneration, Bone 48 (2011) 804–809. [13] Y. Wang, M.C. Battie, T. Videman, A morphological study of lumbar vertebral endplates: radiographic, visual and digital measurements, Eur. Spine J. 21 (2012) 2316–2323. [14] Y. Wang, S.K. Boyd, M.C. Battie, Y. Yasui, T. Videman, Is greater lumbar vertebral BMD associated with more disk degeneration? A study using microCT and discography, J. Bone Miner. Res. 26 (2011) 2785–2791. [15] T. Videman, M. Nurminen, J.D. Troup, Lumbar spinal pathology in cadaveric material in relation to history of back pain, occupation, and physical loading, Spine 15 (1990) 728–740. [16] Y. Wang, T. Videman, M.C. Battie, Morphometrics and lesions of vertebral end plates are associated with lumbar disc degeneration: evidence from cadaveric spines, J. Bone Joint Surg. Am. 95 (e26) (2013) 1–7. [17] Y. Wang, J.S. Owoc, S.K. Boyd, T. Videman, M.C. Battie, Regional variations in trabecular architecture of the lumbar vertebra: associations with age, disc degeneration and disc space narrowing, Bone 56 (2013) 249–254. [18] H.R. Buie, G.M. Campbell, R.J. Klinck, J.A. Macneil, S.K. Boyd, Automatic segmentation of cortical and trabecular compartments based on a dual threshold technique for in vivo micro-CT bone analysis, Bone 41 (2007) 505–515.
G. Yang et al. / Bone 95 (2017) 102–107 [19] W. Bernd, H. Seintsch, R. Amstad, G. Burri, V. Weber, Organization model for postoperative pain management in a basic-care hospital, Anaesthesist 53 (2004) 531–542. [20] M.A. Adams, P. Dolan, Biomechanics of vertebral compression fractures and clinical application, Arch. Orthop. Trauma Surg. 131 (2011) 1703–1710. [21] W. Yue, T. Videman, M.C. Battié, Modic changes: prevalence, distribution patterns, and association with age in white men, Spine 12 (2012) 411–416. [22] E.S. Lee, C.W. Ko, S.W. Suh, S. Kumar, I.K. Kang, J.H. Yang, The effect of age on sagittal plane profile of the lumbar spine according to standing, supine, and various sitting positions, J. Orthop. Surg. Res. 9 (2014) 1–10. [23] M.A. Adams, P. Pollintine, J.H. Tobias, G.K. Wakley, P. Dolan, Intervertebral disc degeneration can predispose to anterior vertebral fractures in the thoracolumbar spine, J. Bone Miner. Res. 21 (2006) 1409–1416.
107
[24] S. Parkinson, A. Campbell, W. Dankaerts, A. Burnett, P. O'Sullivan, Upper and lower lumbar segments move differently during sit-to-stand, Man. Ther. 18 (2013) 390–394. [25] R.S. Alqhtani, M.D. Jones, P.S. Theobald, J.M. Williams, Investigating the contribution of the upper and lower lumbar spine, relative to hip motion, in everyday tasks, Man. Ther. 21 (2016) 268–273. [26] R.C. Mulholland, The myth of lumbar instability: the importance of abnormal loading as a cause of low back pain, Eur. Spine J. 17 (2008) 619–625. [27] P. Pollintine, A.S. Przybyla, P. Dolan, M.A. Adams, Neural arch load-bearing in old and degenerated spines, J. Biomech. 37 (2004) 197–204. [28] P.R. Landham, H.L.A. Baker-Rand, S.J. Gilbert, P. Pollintine, D.J. Annesley-Williams, M.A. Adams, et al., Is kyphoplasty better than vertebroplasty at restoring form and function after severe vertebral wedge fractures? Spine J. 15 (2015) 721–732.