OP is a significant public health problem that is characterized by a systemic impairment of bone mass and microarchitecture. This condition is most commonly reported among postmenopausal women. Bone loss in the OVX rat, which shares striking similarities with postmenopausal bone loss in aged women, is considered the “gold standard” animal model in postmenopausal OP studies [
6]. Radiological assessment is mostly used in the detection of OP to prevent fragile fractures. However, BMD alone cannot predict fracture risk reliably or earlier [
7]. Therefore, other aspects which may reflect the bone microenvironment have been investigated in the last two decades.
Recently, a few studies have proposed the hypothesis that reduced bone perfusion is closely related to compromised BMD [
1,
4,
8‐
10]. Our longitudinal observation showed a tendency of a decrease in the semi-quantitative perfusion parameter E
max accompanied with a decrease in BMD in the OVX group from week 6 on, which is consistent with other reports [
1,
9]. Semi-quantitative DCE-MRI analysis consists of a group of parameters such as E
max and enhancement slope that require calculation based on TIC, which allows for noninvasive evaluation of marrow perfusion in OP [
1,
4,
11,
12]. Nonetheless, semi-quantitative DCE-MRI is more influenced by individual hemodynamic fluctuations and MRI protocols, and its hemodynamic parameters lacks a clear interpretation related to the underlying physiology [
13,
14].
In the past decade, quantitative DCE-MRI parameters have been used to successfully reflect histological changes in vasculature. It is because of developments in pharmacokinetic models have enabled them to resolve those problems for semi-quantitative DCE-MRI [
12,
15]. Therefore, a direct relationship between quantitative DCE-MRI parameters and bone marrow perfusion can be established. K
trans represents the volume transfer constant from the plasma space to the interstitial space. This physiological parameter is fully determined by plasma flow and the permeability-surface area product. K
ep, known as the interstitium-to-plasma rate constant, reflects the reverse transport of gadolinium back into the vascular space. V
e measures the interstitial volume, which is defined as the extravascular-extracellular volume fraction [
15].
To the best of our knowledge, few longitudinal animal-based studies have compared semi-quantitative with quantitative DCE-MRI to determine which better reflects the perfusion condition in OP. In our study, acute estrogen deficiency caused a significant decrease in K
trans of the bone marrow from week 3 after OVX. It was identified by TEM observation, which showed vascular endothelial dysfunction appearing as tighter gaps between VECs with swollen mitochondria. Gulhan et al. [
16] reported that postmenopausal women with OP may have an association with higher endothelin-1levels than those without OP. Animal studies indicated that the tendency of increased endothelin-1 serum levels in the OVX rats was one of the most likely causes of enhanced vasoconstriction and decreased permeability [
2]. These findings strongly support the hypothesis that vascular endothelial dysfunction after OVX induced low bone marrow perfusion at an early stage [
1]. It is notable that no significant change on MVD calculation was found between the OVX and control group until week 12. This phenomenon implies that the decrease in K
trans of the OVX group in the late stage may be attributed to the decrease of MVD in the bone marrow. Our study also found that the V
e values of the OVX group decreased significantly only at week 9 compared with those of the control group. It is obvious that the change of V
e values was later than that of the K
trans values in the OVX group. The increased and enlarged fat cells in the bone marrow of OVX rat occupied the trabecular space, which induced a reduction in extravascular-extracellular space [
1,
17,
18]. This indicates that microvascular dysfunction may occur earlier than the accumulation of fat cells. K
ep, as the ratio of K
trans to V
e, was not significantly different throughout the observation period, which may be due to the reduction of K
trans and V
e in synchrony. Ma et al. [
19] found that the two pharmacokinetic parameters (K
trans and V
e) showed a significant decrease in OP patients compared to normal subjects. In a few longitudinal animal-based studies, K
trans was reported to be a promising parameter to monitor acute ischemia in osteoporotic bone [
2], while K
ep may be a sensitive index to reflect long-term chronic ischemia in OP due to vessel rarefication and maturation in the bone marrow [
3]. Our results are consistent with these reports, which indicated that reduced bone marrow perfusion of OP could be directly reflected by quantitative DCE-MRI. However, no difference in E
max was found between the two groups until week 6 after the operation and the correlation coefficient of K
trans was the highest between MVD and DCE-MRI parameters. Together, these findings suggest that K
trans changes earlier and is more precise than E
max, although both of them have strong positive correlations with BMD and could reflect the reduced bone marrow perfusion in OP. In our study, DCE-MRI can be well performed on the vertebral body of the rat, which illustrates that this technique will be a feasible and reliable modality to quantify bone marrow perfusion in humans. Dual-energy X-ray absorptiometry (DXA) is a well-standardized, inexpensive, and convenient technique for BMD measurement, which has a low radiation dose. However, it is a two-dimensional measurement and is sensitive to degenerative diseases, which may alter the BMD value [
20]. Compared with DXA, DCE-MRI is a relatively high-cost technique. However, lack of radiation and acquisition of compromised bone perfusion parameters in earlier stage of OP makes it more attractive in the clinical practice and more sensitive in monitoring the therapeutic effect. A comprehensive evaluation including BMD and bone metabolism should be performed for an OP patient regardless of whether it is for a precise diagnosis or the assessment of efficacy.
Our study has several limitations. First, the number of rats at each time point was relatively small, but it conforms to the statistical regulation. Experimental errors may affect the results because of individual differences. Second, only the L5 vertebral body was studied by DCE-MRI, micro-CT examination and immunochemical analysis. TEM observation were performed on L4 vertebral bodies of randomly selected rats: one in the control group and two in the OVX group at each time point. Differences in bone marrow perfusion and bone marrow microenvironment may exist at different points along the lumbar spine with varying BMD values [
21,
22], The mismatch between imaging and TEM examination may affect the results. Third, only one semi-quantitative DCE-MRI parameter was selected to compare with the quantitative DCE-MRI parameters. However, E
max is one of the most common semi-quantitative indexes in the measurement of perfusion because of its relative stability and repeatability.