Vertebral bone marrow diffusivity in normal adults with varying bone densities at 3T diffusion-weighted imaging

Abstract

Background

There has been controversy surrounding the relationship between diffusivity and bone mineral density (BMD) in vertebral bone marrow. Moreover, sex-related differences of vertebral bone marrow diffusivity in relation to varying bone densities have not yet been evaluated.

Purpose

To prospectively investigate the role of diffusion-weighted imaging (DWI) in assessing vertebral marrow changes in normal adults with varying bone densities.

Material and Methods

A total of 124 normal adult volunteers were enrolled in this study. Sagittal magnetic resonance (MR) DWI of the lumbar spine was performed. The ADC values of vertebral bone marrow were measured. Volumetric BMD measurement was performed by quantitative computed tomography (QCT) using Mindways QCT analysis software. All participants were divided into three groups according to BMD (normal, osteopenia, osteoporosis). The differences of the apparent diffusion coefficient (ADC) values of the three groups was compared, and partial correlation analysis was used to evaluate the correlation between ADC values and BMD.

Results

ADC values decreased as BMD decreased in female participants. When compared with the normal bone density group, ADC values were signiﬁcantly decreased in the osteoporotic group and in the osteopenic group of female participants. ADC values of female participants were signiﬁcantly higher than of male participants in the normal bone density group (P < 0.001). ADC values correlated positively with BMD values (r = 0.307, P = 0.016) for female participants.

Conclusion

The diffusivity in vertebral bone marrow with varying bone densities differed by sex. ADC values correlated positively with BMD in women. DWI can quantitively evaluate osteoporosis in women.

Keywords

Osteoporosis bone marrow magnetic resonance imaging (MRI)diffusion- weighted imaging (DWI)apparent diffusion coefficient (ADC)bone mineral density (BMD)

Introduction

Osteoporosis is characterized by low bone mass and microarchitectural deterioration of bone tissue, leading to increased bone fragility and a consequent increase in fracture risk. Currently, osteoporosis is diagnosed primarily by determination of bone mineral density (BMD). Dual-energy X-ray absorptiometry (DXA) and quantitative computed tomography (QCT) are the most commonly used methods of BMD evaluation (1). However, bone strength has been shown to depend not only on BMD but also on trabecular bone quality (2). There is an increased amount of fat in the vertebral marrow in patients with osteoporosis (3,4). Diffusion-weighted imaging (DWI) exploits the random translational mobility of water molecules. Changes in bone marrow composition can lead to changes in the extracellular space, leading to changes in bone marrow diffusivity (3,5). Few reports have described the relationship between diffusivity and BMD in vertebral bone marrow with varying bone densities, and controversial findings have been reported (3 –10). However, these studies measure BMD by DXA, in which spinal degeneration and abdominal aortic calcification may be associated with the overestimation of BMD and the underestimation of osteoporosis. Moreover, sex-related differences of the vertebral bone marrow diffusivity in relation to varying bone densities have not yet been evaluated.

The purpose of this study was to prospectively investigate the potential of DWI in assessing bone marrow changes and the differences related to sex in normal adult volunteers with varying bone densities.

Material and Methods

Participants

Our ethics committee approved this prospective study, and all participants provided written informed consent. The study group was composed of 124 normal adult volunteers (age range, 20–90 years; mean age, 53.7 ± 13.6 years). There were 72 women (age range, 21–90 years; mean age, 55.1 ± 15.8 years) and 52 men (range age, 20–82 years; mean age, 52.2 ± 17.4 years), with no significant demographic differences related to sex (P = 0.9). Participants were not recruited if they had: (i) clinical or imaging evidence of metabolic bone disease or metastases or lumbar compression fracture; (ii) a history of lumbar spinal surgery or chemotherapy and/or irradiation; or (iii) a contraindication to magnetic resonance (MR) examination.

MR technique

All MR examinations were performed using a spinal phased array coil on a 3.0 Tesla superconducting MR System (Verio, Siemens Healthineers, Forchheim, Germany). The maximum gradient amplitude and maximum slew rate were 45 mT/m and 200 mT/m/s, respectively.

Routine MR imaging (MRI) of the lumbar spine was performed using sagittal T1-weighted (T1W) (TR/TE 624/11 ms) and T2-weighted (T2W) (TR/TE 3500/88 ms) sequences and transversal T2W (TR/TE 3800/105 ms) sequences. The following imaging parameters were used: slice thickness, 4 mm; intersection gap, 0.4; field of view (FOV), 300 × 100 mm; matrix, 384 × 288; and number of excitation (NEX), 2. All MR images were used to identify pre-existing abnormalities of the lumbar vertebrae.

Sagittal lumbar vertebral DWI was performed using spin-echo, single-shot, echo-planar imaging (EPI) sequences (TR/TE, 5500/100 ms; section thickness, 5 mm; intersection gap, 0; FOV, 280 × 200 mm; NEX, 2; matrix, 164 × 132). Two DW images were simultaneously obtained with diffusion sensitivity parameters (b values) of 0 and 600 s/mm² in the orthogonal planes (x, y, z).

QCT measurement of BMD

All participants underwent a spiral computed tomography (CT) scan of the spine from the level of the second to the fourth lumbar vertebral body (L2–L4) (16 slices High Speed, GE Healthcare, Milwaukee, WI, USA). Scan parameters were 120 kVp, 100 mAs, 3 mm slice thickness, and 40 cm FOV. Participants were positioned supine on the CT table with Mindways CT calibration phantom placed under the participants to cover levels L2–L4. Images were transferred to the QCT PRO PC (Mindways Software Inc., Austin, TX, USA). A region of interest (ROI) within the trabecular bone of each of three vertebral bodies was placed semi-automatically for the BMD measurement, to avoid cortical bone and the posterior veins. The average trabecular BMD of L2–L4 was calculated. The error of this technique is less than 1.5 % (11). Average BMD is expressed in milligrams per cubic centimeter of calcium hydroxyapatite. For the BMD of spinal trabecular bone, 80 mg/cm³ ≤ BMD ≤ 120 mg/cm³ for osteopenia and BMD < 80 mg/cm³ for osteoporosis. These were suggested by the International Society for Clinical Densitometry (ISCD) in 2007 (12) and by the American College of Radiology in 2008 (13). All participants were divided into three groups (normal, osteopenia, osteoporosis). The difference of ADC of three BMD groups and a partial correlation analysis were used to evaluate the correlation between ADC values and BMD.

Quantitative image analysis

ADC values were calculated from the automatically generated ADC maps of the DW images. A circular ROI was drawn manually by one investigator (FH), and all ROIs were identical in size (1.0 cm² area). The ROI was set on the center of the second to forth lumbar vertebral body to weaken the effect of vertebral end-plate fatty changes or edema secondary to disc degeneration. The mean ADC value was obtained from the three vertebral bodies, which was measured three times in each vertebral body (Fig. 1).

Fig. 1.

DWI images and ADC maps obtained at the second to the fourth lumbar vertebra when the b value was 0, 600; ADC values were measured in ADC maps. From: Jie H, Hao F, Na LX. Vertebral bone marrow diffusivity in healthy adults at 3T diffusion-weighted imaging. Acta Radiol 2016;57:1238–1243. With permission.

Statistical analysis

The mean ADC values of each group of BMD was calculated (mean plus or minus standard deviation). One-way analysis of variance (ANOVA) was performed to compare the difference of ADC values among the different BMD groups (normal, osteopenia, and osteoporosis) for each sex. A t-test was performed to compare the difference of ADC values between the sexes in the same BMD group. A partial correlation analysis was performed to analyze the correlation between ADC values and BMD, and a P value < 0.05 was considered to indicate statistical significance. All statistical analyses were performed with SPSS software (version 17.0; SPSS Inc., Chicago, IL, USA).

Results

All participants underwent both QCT and conventional MRI with DWI. The mean ADC values of the lumbar vertebral bodies for sex and BMD are summarized in Tables 1 and 2.

Table 1.

Comparison of ADC values of different BMD groups in normal adult male participants.

Group	n	Age (years)	ADC (× 10^–3 mm²/s)	BMD (mg/cm³)
Normal	22	45.5 ± 17.0	0.310 ± 0.087	153.8 ± 25.0
Osteopenia	16	60.1 ± 11.5	0.291 ± 0.065	105.4 ± 10.7
Osteoporosis	14	68.9 ± 7.0	0.301 ± 0.031	66.1 ± 9.9

No significant differences among BMD group in men, P > 0.05

Table 2.

Comparison of ADC values of different BMD groups in normal adult female participants.

Age	n	Age (years)	ADC (× 10^–3 mm²/s)	BMD (mg/cm³)
Normal	26	39.6 ± 12.6	0.450 ± 0.130	164.2 ± 26.3
Osteopenia	21	58.5 ± 9.9	0.303 ± 0.068*	99.0 ± 11.9
Osteoporosis	25	65.0 ± 9.1	0.287 ± 0.048*^†	59.0 ± 19.8

Compared with normal group, P < 0.001.

†

Compared with osteopenia group, P > 0.05.

There was no definite trend among male participants. The ADC values decreased with decreasing BMD among female participants. There were significant differences in the ADC values among the three bone density groups in women (normal versus osteopenia and normal versus osteoporosis, P < 0.001 and P < 0.001, respectively) (Fig. 2), whereas there were no significant differences in men (P = 0.503) (Fig. 3). The ADC values of the normal bone density group were significantly higher than of the osteopenic and osteoporotic groups for women. There were no significant differences in the ADC values between the osteopenic and osteoporotic groups in women.

Fig. 2.

Graph showing the mean ADC values of L2–L4 for women with varying BMD groups. There were significant differences among the three bone density groups. The ADC values of the normal bone density group were significantly higher than of the osteopenic and osteoporotic groups. There were no significant differences in the ADC values among the osteopenic and osteoporotic groups. The horizontal bar shows the mean value for each group.

Fig. 3.

Graph showing the mean ADC values of L2–L4 for men in varying BMD groups. There were no significant differences among the three bone density groups. The horizontal bar shows the mean value for each group.

There were also significant sex differences for the ADC values in the normal bone density group (P < 0.001), although there were no significant sex differences among the ADC values in the osteopenic group and osteoporotic groups (P = 0.625 and 0.731, respectively). The ADC values of the normal bone density group in female participants were significantly higher than those of the male participants (t = 12.71, P < 0.001) (Fig. 4).

Fig. 4.

Graph showing the mean ADC values for males and females in varying BMD groups. There were significant sex differences between the normal BMD groups (P < 0.001), although no significant sex differences were observed in the osteopenic and osteoporotic groups (P = 0.625 and 0.731, respectively).

In addition, the ADC values were significantly weakly positively correlated with BMD for all participants (r = 0.221, P = 0.025). The ADC values were not significantly correlated with BMD for male participants (r = −0.090, P = 0.576), whereas it was significantly and positively correlated with BMD in female participants (r = 0.307, P = 0.016) (Fig. 5).

Fig. 5.

Scatter diagram showing the positive correlation between ADC values of bone marrow in the lumbar vertebrae and BMD for normal adult female participants (r = 0.307, P = 0.016)

Discussion

Our study measured the BMD of vertebral bone marrow with QCT rather than DXA. The diagnostic criteria for DXA established by WHO in 1994 have long been used as the gold standard in the clinical diagnosis of osteoporosis. DXA determines BMD in two dimensions, including both trabecular and cortical bone, with the results expressed as areal density (grams per square centimeter). Spinal degeneration, abdominal aortic calcification, and other sclerotic lesions may lead to the overestimation of BMD. In contrast, QCT is truly a three-dimensional technique for quantifying BMD. It allows volumetric measurement of trabecular bone density without superimposition of cortical bone and other tissues. Even though osteoporotic bone loss occurs in cortical and trabecular bone concomitantly, the measurement of BMD in trabecular bone is more sensitive than in cortical bone because the turnover rate of trabecular bone is eight times than that of cortical bone. Thus, the measurement of BMD in trabecular can reflect the change of calcium hydroxyapatite in body early. A study by Li et al. (14) showed a significant difference in osteoporosis detection rates between DXA and QCT, providing clinical evidence that QCT has a greater diagnostic sensitivity than DXA.

The present study shows that ADC values tend to decrease with decreasing BMD in women. ADC values in women with osteopenia and osteoporosis were signiﬁcantly lower than in those with normal BMD. A statistically signiﬁcant positive correlation existed between ADC values and BMD. Previous research has shown that the vertebral bone marrow in decreased BMD consists of fatty bone marrow (15,16). Histologic studies have found similar results (17). The adipose tissue ﬁlls the space between the thinned and ruptured trabeculae and prevents extracellular water molecules from diffusing, resulting in decreased ADC values. In accordance with our investigation, some studies have examined the diffusivity of lumbar vertebral marrow with varying bone densities and found that the ADC values decreased with BMD and were positively correlated with BMD (5 –8,10). However, the major limitation of these studies was the lack of evaluation of sex differences. In contrast, a study by Grifﬁth et al. found that a decrease in bone density was associated with a corresponding increase in vertebral marrow fat content and a decrease in maximum enhancement and enhancement slope. But the study did not ﬁnd a signiﬁcant difference in the diffusion restrictions among the osteoporotic, osteopenic, and normal bone density groups, and no relationship between bone marrow ADC and bone density (3). Participants in the study were postmenopausal women aged more than 65 years (mean age, 73 years) and BMD was measured using DXA. Spinal degeneration and abdominal aortic calcification are most commonly seen in the elderly, and may reduce diagnostic sensitivity for osteoporosis. Koyama et al. reported a signiﬁcantly negative correlation between the ADC values of vertebrae and BMD (9). However, this study included a small number of participants (16 cases), and the BMD distribution was uneven (no participants in the osteopenic group). Our study used partial correlation analysis to exclude the inﬂuence of age on the results of reduced diffusivity of bone marrow.

Our study indicated that the ADC value of female participants (0.450 × 10^–3 mm²/s) was significantly higher than that of male participants (0.310 × 10^–3 mm²/s) in the normal bone density group. Some previous MR spectroscopy (MRS) studies have found that men have more vertebral bone marrow fat than women, and it was believed that increased fat marrow content in healthy men may be a physiologic phenomenon (16,18 –20). The lipid component of fatty bone marrow primarily consists of lipid-bound water protons, which conceivably have a more restricted mobility than unbound protons. The adipose cells fill the space of the trabecular bone, thus preventing extracellular water molecules from diffusing and results in decreased ADC values. Our findings are in agreement with these studies, although these earlier reports involved participants of a wide age range whose bone density was unknown.

Our study also demonstrated that the sex differences in diffusivity of the bone marrow were related to BMD. This sex difference only occurred in normal bone density group. There were no significant sex differences in the osteopenic or the osteoporotic groups. Our previous study showed that sex difference in diffusivity of the bone marrow occurs among normal participants aged less than 50 years without regard to bone density (21). In fact, osteoporosis and osteopenia are most commonly seen in women aged over 50 years. The ADC values were markedly reduced in osteoporosis and osteopenia in women. However, the ADC values were not markedly reduced in men; therefore, the difference between the sexes disappeared when the participants were osteoporotic and osteopenic. Chen et al. (22) found that age-related differences in vertebral marrow perfusion and fat deposition do exist between men and women. Women demonstrated a higher marrow perfusion rate than men aged less than 50 years. However, no difference was demonstrated in participants aged more than 50 years. This finding is approximately consistent with our results. However, that study only focused on the relationship between bone marrow perfusion change and aging and sex, without an analysis of bone mineral density. Sex hormones are important factors in higher diffusivity in premenopausal women. The higher diffusivity of the bone marrow in normal BMD group of women noted in our study may suggest a possible relationship between diffusivity and the female sex hormone and deserves further investigation. Thus far, it remains unknown whether sex differences influence vertebral bone marrow conditions.

Our study obtained mean ADC values of 0.450, 0.303, and 0.287 in normal bone density, osteopenia and osteoporosis group in vertebral bodies of women. Tang et al. (7) reported results of 0.47, 0.41, and 0.39 for normal bone density, osteopenia and osteoporosis in female vertebral bodies. Another study reported 0.465, 0.408, and 0.338 respectively (10). Our results agree with these values in the normal bone density group, but the values in the current study were lower in the osteopenic and osteoporotic groups. In fact, the accuracy of ADC values depends on many factors, such as the field strength, signal-to-noise ratio (SNR), spatial resolution, and the number and amplitude of the b values used in the data acquisition. There are many reasons for the difference, which include field strength, b value, uneven age distributions (participants were postmenopausal women in these studies). The difference criterion for bone density may be a main reason (QCT versus DXA).

Our study had several limitations. First, this study did not include participants aged less than 20 years. Second, in this study, the age distributions in the groups were slightly uneven, particularly for osteoporotic men. Third, we did not exclude all participants with conditions which can affect the bone marrow such as anemia. Fourth, the slice thickness (5 mm) used for DWI differed at T1W and T2W imaging (4 mm) in terms of the SNR of DWI.

In conclusion, our study demonstrated sex-related difference of the vertebral bone marrow conditions with varying bone densities. In particular, the ADC value was positively correlated with BMD in women, and a sex-related difference was observed in the normal BMD group. DWI techniques might have some value in estimating bone marrow changes in women associated with osteopenia and osteoporosis.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Lentle

. Osteoporosis and bone densitometry: does the emperor have clothes? CMAJ 1998; 159: 1261–1264.

Eriksson

Isberg

Lindgren

. Prediction of vertebral strength by dual photon absorptiometry and qualitative computed tomography. Calcif Tissue Int 1989; 44: 243–250.

Griffith

Yeung

Antonio

et al.

Vertebral marrow fat content and diffusion and perfusion indexes in women with varying bone density: MR evaluation. Radiology 2006; 241: 831–838.

Fanucci

Manenti

Masala

et al.

Multiparameter characterisation of vertebral osteoporosis with 3-T MR. Radiol Med 2007; 112: 208–223.

Yeung

Wong

Griffith

et al.

Bone marrow diffusion in osteoporosis: evaluation with quantitative MR diffusion imaging. J Magn Reson Imaging 2004; 19: 222–228.

Hatipoglu

Selvi

Ciliz

et al.

Quantitative and diffusion MR imaging as a new method to assess osteoporosis. Am J Neuroradiol 2007; 28: 1934–1937.

Tang

et al.

Evaluation of MR spectroscopy and diffusion-weighted MRI in detecting bone marrow changes in postmenopausal women with osteoporosis. Clin Radiol 2010; 65: 377–381.

Liu

Tang

et al.

Assessment of bone marrow changes in postmenopausal women with varying bone densities: magnetic resonance spectroscopy and diffusion magnetic resonance imaging. Chin Med J (Engl) 2010; 123: 1524–1527.

Koyama

Yoshihara

Kotera

et al.

The quantitative diagnostic capability of routine MR imaging and diffusion-weighted imaging in osteoporosis patients. Clin Imaging 2013; 37: 925–929.

10.

Agrawal

Agarwal

Chopra

et al.

Evaluation of MR spectroscopy and diffusion-weighted MRI in postmenopausal bone strength. Cureus 2015; 7: e327–e327.

11.

Blight

Bidaut

White

et al.

Helical multidetector row quantitative computed tomography (QCT) precision. Acad Radiol 2009; 16: 150–159.

12.

Engelke

Adams

Ambrecht

et al.

Clinical use of quantitative computed tomography and peripheral quantitative computed tomography in the management of osteoporosis in adults: the 2007 ISCD Official Positions. J Clin Densitom 2008; 11: 123–162.

13.

American College of Radiology. ACR Practice Guideline for the Performance of Quantitative Computed Tomography (QCT) Bone Densitometry (Resolution 33). Reston, VA: ACR, 2008. Available at: http://www.acr.org//media/ACR/Documents/PGTS/guidelines/QCT.pdf.

14.

et al.

Comparison of QCT and DXA: Osteoporosis detection rates in postmenopausal women. Int J Endocrinol 2013; 2013: 474–478.

15.

Schellinger

Lin

Lim

et al.

Bone marrow fat and bone mineral density on proton MR spectroscopy and dual-energy X-ray absorptiometry: their ratio as a new indicator of bone weakening. Am J Roentgenol 2004; 183: 1761–1765.

16.

Schellinger

Lin

Hatipoglu

et al.

Potential value of vertebral proton MR spectroscopy in determining bone weakness. Am J Neuroradiol 2001; 22: 1620–1627.

17.

Dunnill

Anderson

Whitehead

. Quantitative histological studies on age changes in bone. J Pathol Bacteriol 1967; 94: 275–291.

18.

Schellinger

Lin

Fertikh

et al.

Normal lumbar vertebrae: anatomic, age, and sex variance in subjects at proton MR spectroscopy–initial experience. Radiology 2000; 215: 910–916.

19.

Liney

Bernard

Manton

et al.

Age, gender, and skeletal variation in bone marrow composition: a preliminary study at 3.0 Tesla. J Magn Reson Imaging 2007; 26: 787–793.

20.

Kugel

Jung

Schulte

et al.

Age and sex specific differences in the 1H-spectrum of vertebral bone marrow. J Magn Reson Imaging 2001; 13: 263–268.

21.

Jie

Hao

. Vertebral bone marrow diffusivity in healthy adults at 3T diffusion-weighted imaging. Acta Radiol 2016; 10: 1238–1243.

22.

Chen

Shih

Chen

et al.

Vertebral bone marrow perfusion evaluated with dynamic contrast-enhanced MR imaging: significance of aging and sex. Radiology 2001; 220: 213–218.