The authors state that they have no conflicts of interest.
Severity of Vertebral Fractures Is Associated With Alterations of Cortical Architecture in Postmenopausal Women†
Article first published online: 8 DEC 2008
Copyright © 2009 ASBMR
Journal of Bone and Mineral Research
Volume 24, Issue 4, pages 737–743, April 2009
How to Cite
Sornay-Rendu, E., Cabrera-Bravo, J.-L., Boutroy, S., Munoz, F. and Delmas, P. D. (2009), Severity of Vertebral Fractures Is Associated With Alterations of Cortical Architecture in Postmenopausal Women. J Bone Miner Res, 24: 737–743. doi: 10.1359/jbmr.081223
- Issue published online: 4 DEC 2009
- Article first published online: 8 DEC 2008
- Manuscript Accepted: 5 DEC 2008
- Manuscript Revised: 6 NOV 2008
- Manuscript Received: 25 SEP 2008
- bone microarchitecture;
- postmenopausal women;
- vertebral fractures;
Patients with vertebral fractures (VFx) have trabecular architectural disruption on iliac biopsies. Because cortical bone is an important determinant of bone strength, we assessed cortical and trabecular microarchitecture at peripheral sites in patients with VFx of varying number (N) and severity (S). Bone architecture and volumetric density (vBMD) were assessed at the distal radius and tibia with HR-pQCT (XTreme CT; Scanco Medical, Bassersdorf, Switzerland) in 100 women with VFx (age, 74 ± 9 yr) of different S (GI, n = 23; GII, n = 35; GIII, n = 42) and in 362 women (age, 69 ± 7 yr) without peripheral or VFx (G0) from the OFELY study. Spine areal BMD (aBMD) was assessed by DXA. Among all women, at the radius and after adjustment for age and aBMD, there were significant trends in lower vBMD, cortical thickness (Cort.Th), trabecular number (Tb.N) and thickness (Tb.Th), higher trabecular separation (Tb.Sp), and distribution of separation (Tb.Sp.SD) with greater VFx S and N. Among women with VFx, lower Cort.Th and cortical vBMD (D.Cort) were associated with severe (GIII) and multiple (n > 2) VFx (p < 0.05). The age-adjusted OR for each SD decrease of Cort.Th was 2.04 (95% CI, 1.02–4.00) after adjustment for aBMD. At the tibia, there were trends for lower vBMD, Tb.N, Tb.Th, and higher Tb.Sp and Tb.Sp.SD with greater VFx S and N (p < 0.001). Among women with VFx, lower Cort.Th and D.Cort were associated with severe and multiple (n > 3) VFx (p < 0.01). In postmenopausal women, VFx are associated with low vBMD and architectural decay of trabecular and cortical bone at the radius and tibia, independently of spine aBMD. Severe and multiple VFx are associated with even more alterations of cortical bone.
The risk of subsequent vertebral fracture (VFx) in women with preexisting VFx is approximately four times that of women without a fracture history, increasing with the number of prior vertebral fractures.(1,2) Furthermore, the occurrence of a new VFx increases the risk of subsequent VFx within the following year.(3) Women with more severe degrees of vertebral deformities measured as a relative decrease in vertebral height have an increased risk of subsequent vertebral and nonvertebral fractures.(4) VFx are also associated with an increased mortality, particularly for clinical VFx.(5,6)
Microstructural deterioration of bone tissue has been reported in the presence of VFx, from ex vivo studies on iliac crest bone biopsies, showing lower trabecular bone volume and number and higher trabecular separation in women with vertebral fractures.(7–9) Alterations of cortical bone have also been reported in women with VFx.(9–13)
Alterations in bone microarchitecture according to the severity of VFX have been little studied. Trabecular architecture alterations associated in patients with greater VTx severity have been reported from iliac biopsies in a study of 190 postmenopausal women with osteoporosis, such as decreasing bone volume, trabecular number, and connectivity density, and increasing trabecular separation.(14) The cortical architecture was not evaluated. To our knowledge, no study has analyzed the association of the number of VFx with alterations of bone architecture.
The aim of this study was to assess the association of both cortical and trabecular architecture impairment at the distal radius and tibia with VFx severity and number in postmenopausal women. The objective was to identify if VFx severity and number were associated with alterations of bone microarchitecture independently of BMD assessed by DXA.
MATERIALS AND METHODS
One hundred women (mean age, 74 ± 9 yr) with VFx were analyzed. Thirty-one women with VFx belonged to the OFELY study and 69 women were admitted for hospitalization (n = 37) or consultation (n = 32) in the rheumatology unit. Three hundred sixty-two women (mean age, 69 ± 7 yr) without peripheral or VFx from the OFELY study who came to their 14th visit at the same time as women with VFx were also analyzed. Thus, all the women were measured between January 2005 and March 2007. OFELY (os des femmes de Lyon) is a prospective study of the determinants of bone loss in 1039 volunteer women, 31–89 yr of age, recruited between February 1992 and December 1993 and randomly selected from the affiliates of a large health insurance company (Mutuelle Générale de l' Education Nationale) from the Rhône district with an annual follow-up. The OFELY cohort has been described elsewhere.(15,16)
VFx were identified on lateral X-ray films of the thoracic and lumbar spine. According to the semiquantitative method of Genant, each vertebrae from T4 to L4 was graded as normal (0) or with mild (I), moderate (II), or severe (III) deformity. A mild fracture was defined as a 20–25% reduction in anterior, middle, or posterior vertebral height. Moderate and severe fracture were defined as 25–40% and >40% reduction in any height, respectively.(17) The most severe deformity was considered for each woman, and all women were classified into subgroups: G0, n = 362; GI, n = 23; GII, n = 35; GIII, n = 42.
In the OFELY study, the clinical VFx (radiographically confirmed) were collected in the annual questionnaire, whereas VFx that did not reach clinical attention were assessed on the X-ray films performed every 4 yr. The latest radiographs were performed at the 13th visit (i.e., 1 yr before bone microarchitecture evaluation).
We excluded vertebral fractures that occurred because of major trauma. Peripheral fragility fractures were annually registered during the 13-yr follow-up, all confirmed by radiographs (all sites were included, except head, toes, and fingers). Only low-trauma fractures (i.e., those occurring after falls from standing height or less) were taken into account.
Bone microarchitectural measurement
The nondominant forearm (or the nonfractured forearm in the case of prior fracture) and distal tibia were scanned using a high-resolution 3D pQCT device (XTreme CT; Scanco Medical, Bassersdorf, Switzerland). This system uses a 2D detector array in combination with a 0.08-mm point-focus X-ray tube, enabling the simultaneous acquisition of a stack of parallel CT slices with a nominal isotropic resolution (voxel size) of 82 μm. The details of the acquisition and analysis has been described previously.(18) At each skeletal site, 110 CT slices were obtained, thus delivering a 3D representation of ∼9 mm in the axial direction. The entire volume of interest was automatically separated into a cortical and trabecular region. Thus, we obtained volumetric total (D.tot), trabecular (D.trab)and cortical (D.cort) bone densities (mg HA/cm3). Mean cortical thickness (Cort.Th) was defined as the mean cortical volume divided by the outer bone surface. Trabecular bone volume fraction (BV/TV, %) was derived from trabecular density assuming fully mineralized bone to have a mineral density of 1.2 g HA/cm3. Among the trabecular structural parameters, trabecular number (Tb.N, 1/mm) was measured, whereas trabecular thickness (Tb.Th, μm) and separation (Tb.Sp, μm) were derived from BV/TV and Tb.N using standard methods from histomorphometry [i.e., Tb.Th = (BV/TV)/Tb.N and Tb.Sp = (1 − BV/TV)/Tb.N].(7) Distance transformation techniques also enable the calculation of the distribution of trabecular separation (Tb.Sp.SD, μm) reflecting the heterogeneity of the trabecular network.(7) The in vivo precision error of density (total, trabecular, and cortical) measurements, expressed as CV, ranged from 0.7% to 1.5%. The reproducibility of structural parameters was slightly lower, with CVs ranging from 0.9% to 4.4%.(18)
Areal BMD (aBMD) was measured at the same visit by DXA, at the anteroposterior lumbar spine and the total hip (QDR 4500, software version V8.26a; Hologic, Waltham, MA, USA). The in vivo precision error of DXA, expressed as the CV, was 0.9% for the spine and 1% for total hip. A control phantom was scanned every day. The in vitro long-term precision of DXA over the duration of the study was 0.41%.
A trend test was conducted for each continuous variable according to the number and the severity of the VFx, before and after adjusting for age and spine aBMD.
The association between measurement values and fracture status was calculated by logistic regression and expressed as ORs (with 95% CIs) per 1 SD change from the reference group. All statistical analyses were performed using the Statistical Analysis Software (SAS V8; SAS Institute, Cary, NC, USA).
Characteristics of women according to fracture severity are presented in Table 1. Significant trends were observed with increasing severity of VFx for increasing age, years since menopause, number of VFx and previous peripheral fractures, and decreasing height and T-score at the lumbar spine. Among women with VFx, both the existence and the number of previous peripheral fractures did not differ according to the severity and the number of VFx.
Fifty-six percent of the women with VFx were currently receiving an osteoporosis-related drugs (bisphosphonate, n = 47; selective estrogen receptor modulator [SERM], n = 3; estrogens, n = 3; teriparatide, n = 3; tibolone, n = 1) compared with 22% of controls (bisphosphonate, n = 33; estrogens, n = 33; SERM, n = 11; tibolone, n = 4; p < 0.0001). None of the women with VFx were treated with bone loss inducing therapy compared with 2% of controls (anti-aromatases, n = 7; corticosteroids, n = 2; p = 0.2).
At the distal radius, after adjusting for age and spine aBMD assessed by DXA, there were significant trends in decreasing volumetric BMDs, cortical thickness (Cort.Th) and trabecular number (Tb.N) and thickness (Tb.Th) and increasing trabecular separation (Tb.Sp) and distribution of separation (Tb.Sp.SD) with greater VFx severity (p < 0.0001 to p < 0.04; Table 2). Adjusting for the presence of osteoporosis-related drugs did not affect the results except for Tb.Th (p = 0.13). Total volumetric BMD (D.Tot) was 16%, 21%, and 26% lower (p < 0.0001) and Cort.Th was 9%, 19%, and 30% lower (p = 0.02) in women with GI, GII, and GIII, respectively, compared with women with G0.
Among women with VFx, D.Cort and Cort.Th were significantly lower in GIII compared with GI (p < 0.05). When the analysis was restricted to the 81 women who did not sustain a VFx from the L1 to L4 levels (that could have overestimated the aBMD values), trend tests were still significant after adjusting for age and spine aBMD.
At the tibia, after adjusting for age and spine aBMD, there were significant trends in decreasing volumetric total and trabecular BMDs, Tb.N, Tb.Th and increasing Tb.Sp and Tb.Sp.SD with greater VFx severity. Women with GI, GII, and GIII had significantly lower BMDs and more microarchitectural alterations than women with G0. In contrast, no significant differences were found between GI and GIII (Table 3). Adjusting for the presence of osteoporosis-related drugs did not affect the results.
Thirty-nine women had one VFx, 19 had two VFx, and 42 had more than two VFx. The number of VFx was associated with greater alterations of volumetric total, cortical, and trabecular BMDs, cortical thickness, and trabecular microarchitecture (Tb.N, Tb.Th, Tb.Sp, Tb.Sp.SD) assessed at the radius (adjusted p for age and spine aBMD <0.0001 to <0.001). When the microarchitecture was assessed at the tibia, the number of VFx was associated with greater alterations of all those parameters (p < 0.0001) except for Cort.Th (p = 0.06; data not shown).
By combining number and severity of VFx, we found that women with severe and multiple VFx (GIII and N > 2, n = 30) had lower cortical thickness and density assessed at the radius compared with women with mild or moderate and one or two VFx. The age-adjusted OR for each SD decrease of Cort.Th was 2.13 (95% CI, 1.10–4.17) and 2.04 (95% CI, 1.02–4.00) before and after adjusting for spine aBMD, respectively. When the microarchitecture was assessed at the tibia, the risk did not increase in women with GIII and N > 2. Nevertheless, with more drastic conditions such as severe VFx and more than three vertebral fractures (GIII and N > 3, n = 21), lower cortical thickness and density assessed at both sites were associated with severe and multiple vertebral fracture (Fig. 1).
After adjusting for total hip aBMD (instead of spine aBMD), women with severe and multiple VFx still had lower cortical thickness and density assessed at the radius compared with women with mild or moderate and less than two or three VFx. The age-adjusted OR for each SD decrease of Cort.Th was 2.08 (95% CI, 1.13–3.83) for women with GIII and N > 2 and 2.26 (95% CI, 1.20–4.47) for women with GIII and N > 3. When the microarchitecture was assessed at the tibia, the risk did not increase in women with GIII and N > 2 or N > 3, after adjusting for total hip BMD.
We report that, in postmenopausal women, VFx are associated with low volumetric BMD and architectural alterations of trabecular and cortical bone assessed at the radius and tibia. This association is independent of decreased areal BMD. Moreover, severe and multiple VFx are associated with more alterations of cortical architecture.
Areal BMD measurement by DXA is widely used to estimate the risk of osteoporotic fractures. However, many fractures occur in women with BMD above the WHO threshold of osteoporosis (i.e., at a T-score > −2.5).(19–21) The relevance of trabecular microstructure to bone strength has been well documented,(22) and the evaluation of both microarchitecture and BMD may improve estimation of the risk of fracture. In a previous case-control study from the OFELY cohort, we showed that fragility fractures were associated with low volumetric BMD and architectural alterations of trabecular and cortical bone assessed in vivo with the same high-resolution method. The architectural alterations of bone were partially independent of aBMD assessed by DXA.(23)
In this study focusing on VFx, we confirmed this, showing lower trabecular bone volume (BV/TV), trabecular number, and higher trabecular separation in women with VFx. These results are in agreement with ex vivo studies on iliac crest bone biopsies.(7,8,24) We also found lower trabecular thickness that has been reported in some histological studies,(7,25) whereas others found higher trabecular thickness in women with VFx.(8,24) A study using multidetector row CT (MDRCT) at the third lumbar vertebra with a higher spatial resolution than conventional CT, performed in 82 postmenopausal women, showed that both volumetric BMD and microstructure parameters were associated with a higher relative risk for prevalent vertebral fracture than did aBMD obtained by DXA.(26) In another study using high-resolution MRI at the radius, all structural parameters, except trabecular thickness, differed between women without fracture and those with VFx.(27) All those studies cited above compared patients with and without VFx.
Only one study has analyzed alterations of microarchitecture according to the severity of VFx. In that study, trabecular parameters were assessed by 3D μCT on iliac biopsies from 190 osteoporotic postmenopausal women. After adjusting for age, height, and lumbar spine BMD, there were significant trends in decreasing BV/TV, trabecular number (but not trabecular thickness), and connectivity density, and increasing trabecular separation with greater vertebral fracture severity, graded from 0 to 3. Among women with VFx, BV/TV was lower, whereas the other trabecular parameters were not more altered in women with severe VFx than in women with mild VFx.(14) We did not find more alterations of trabecular parameters in women with severe VFx than in women with mild VFx. The low number of women with severe VFx in the first study (n = 12) could partially explain these discrepancies. Moreover, in our study, women with mild VFx had lower BV/TV and trabecular number and higher trabecular separation and distribution of separation than women without VFx. Our findings suggest that trabecular architecture is widely altered in presence of VFx even with mild VFx. This is in agreement with results from an analysis of the placebo groups of two large trials, showing that mild vertebral fractures (and even only one) increase significantly the risk of incident fracture in osteoporotic women.(28)
Cortical bone is another important structural determinant of bone strength, but the load sharing between the cortical shell and trabecular bone in the human vertebra remains unclear. In agreement with histomorphometric studies(10,11) and with our previous study, we found lower cortical thickness values in women with vertebral fractures. Moreover, women with more severe VFx have thinner cortices at peripheral sites, contrasting with no more decrease of aBMD or trabecular alterations than women with mild VFX. This is in agreement with a study performed with pQCT at the radius showing a lower cortical area in women with VFx compared with postmenopausal controls.(29) That suggests a predominant role of cortical architecture on the preservation of the vertebral shape that should be confirmed by further studies evaluating cortical bone at the spine, because bone structure is different in long bones and vertebral bodies. Indeed, the cortical shell of the vertebral bodies is thinner than in long bones and seems to be a more thickened trabecular bone of different structure than the true cortical bone (i.e., with Haversian canals and lamellar structure) of the forearm.(30,31) Moreover, the relationship between the cortical thickness at the radius and at the spine is poorly documented because measurements with QCT at the spine are often misleading. Inadequate spatial resolution (voxel size between 0.6 and 0.8 mm) relative to cortical thickness (less than one half of a millimeter)(10,30) leads to overestimate measurements with QCT because of volume averaging artefacts. In a study of 16 first lumbar vertebrae from human cadaveric spines comparing direct and QCT-based measurements of the vertebral shell, Silva et al.(30) showed an overestimation of the thickness by a factor of at least two with CT scans.
Nevertheless, recent studies also reported an important role of cortical bone-assessed at the spine level-in VFx. Melton et al.(13) showed that the apparent vertebral cortical thickness, assessed with lumbar QCT, was smaller in women with VFx, whereas no significant differences were observed among the trabecular parameters assessed by HRpQCT at the radius. Using high-resolution CT-based finite element models on vertebral bodies obtained from female human cadavers, the authors reported that the percent bone mass attributable to the cortical shell-not including the endplates-was large, ranging from 21% to 39% across vertebrae, even though the average shell thickness was only on the order of 0.38 mm.(32)
The greater alterations of bone microarchitecture in women with severe and multiple VFx could be the result of greater bone loss producing structural changes or previous abnormal bone structure at the peak bone mass. Other studies are necessary to test these hypothesis. Nevertheless, our findings are in agreement with a study conducted from the placebo group of the MORE study showing a greater risk of subsequent fracture in women with severe VFx fractures at baseline compared with women with mild prevalent fractures, even after controlling for differences in baseline characteristics of women. In that study, baseline BMD and the number and severity of prevalent vertebral fractures were all predictors of the risk of subsequent vertebral fractures but fracture severity grade was a stronger predictor of future nonvertebral fracture risk than either baseline BMD or the presence or number of baseline vertebral fractures. The authors suggested that the severity of existing vertebral fractures may better reflect impaired bone architecture leading to a decline in bone strength than does BMD or the number of prevalent fractures.(2)
We did not observed lower cross-sectional area assessed at the radius and tibia in women with VFx. In contrast, some studies indicated that compressive strength of vertebrae is determined not only by their BMD but also by their dimensions such as cross-sectional area.(33–35)
Our study has limitations. The OFELY study is a population-based cohort study, and all fragility fractures were prospectively assessed during a long follow-up. The repetition of spine radiographs every 4 yr allowed an optimal ascertainment of vertebral fractures. Nevertheless, 69 women with VFx did not belong to the OFELY study; they were used to increase the number of women with different grades of VFx. Another limitation is that bone microarchitecture was assessed at peripheral sites and not at the spine. The case-control design of our study does not permit establishment of a causal relationship between alterations of microarchitecture and VFx. Finally, a certain number of the women with VFx were currently on an osteoporosis-related drug, especially bisphosphonates, that could have modified bone microarchitecture. Nevertheless, alterations of bone architecture are probably worse in women taking those treatments and taking account their presence did not affect the results.
In conclusion, our findings suggest that, in postmenopausal women, VFx are associated with low volumetric BMD and architectural alterations of trabecular and cortical bone that can be assessed noninvasively by HRpQCT. Severe and multiple VFx are associated with even more alterations of cortical bone. Prospective studies should be undertaken to determine whether HRpQCT can predict the risk of severe and multiple vertebral fractures.
The authors thank A Bourgeaud, S Ailloud, and W Wirane for excellent technical assistance.
- 62003 Prevalent vertebral deformities predict increased mortality and increased fracture rate in both men and women: A 10-year population-based study of 598 individuals from the Swedish cohort in the European Vertebral Osteoporosis Study. Osteoporos Int 14: 61–68., , , ,
- 111997 The thickness of human vertebral cortical bone and its changes in aging and osteoporosis: A histomorphometric analysis of the complete spinal column from thirty-seven autopsy specimens. J Bone Miner Res 12: 89–95., , , ,
- 232007 Alterations of cortical and trabecular architecture are associated with fractures in postmenopausal women, independently of decreased bone mineral density measured by DXA. The OFELY study. J Bone Miner Res 22: 425–433., , ,