Validation of noninvasive quantification of bone marrow fat volume with microCT in aging rats

https://doi.org/10.1016/j.exger.2011.01.001Get rights and content

Abstract

Marrow fat infiltration is one of the hallmarks of age-related bone loss. This fat infiltration has been quantified by invasive and noninvasive methods. However, the validity of the noninvasive methods has not been correlated with a gold standard. In this study we aim to validate the usefulness of marrow fat quantification by correlating microCT (μCT) images with histology analysis. Fat volume (FV) and bone volume (BV) of distal femora of young (4 months) and old (27 months) Louvain/c (LOU) rats (n = 22) were quantified by histology and compared with μCT images analyzed by an image analysis software (SliceOMatic).

We found that for SliceOMatic/μCT the intra-rater reliability for duplicate measurements was 0.94 (p < 0.001) and the inter-rater reliability for FV/BV ratio in young and old rats was 98% and 99% respectively. Both methods showed a significant increase (~ 2 fold) in the FV/BV ratio in the old rats as compared with their young counterparts (p < 0.001). A significantly higher correlation (r2 = 0.85) in the old rats was found between our noninvasive method and histology. Furthermore, our noninvasive method showed good agreement with histology.

In conclusion, noninvasive quantification of FV/BV ratio using an image analysis software is as reliable as histology for identifying age related marrow fat changes with high inter and intra-rater reliability. These findings provide a new noninvasive method for quantifying marrow fat, which is useful and can be tested not only in animals but also in human studies.

Research Highlights

► Marrow fat infiltration is one of the hallmarks of senile osteoporosis. ► Current methods of marrow fat quantification lack comparison with a gold standard. ► Image analysis of marrow fat has a good agreement with the gold standard. ► Image analysis is a promissory method to quantify fat infiltration in aging bone.

Introduction

Osteoporosis and its most devastating sequelae, fractures, is a rising global health, economic and social burden. Thus the early detection and treatment of individuals at risk of fractures is a priority before the fragility fracture cascade sets in. Bone Mineral Density (BMD) and Dual Energy X-ray Absorptiometry (DXA) have long been the recommended fracture surrogate and non-invasive tool respectively that estimate fracture risk. However, there is evidence that BMD alone, as defined by the World Health Organization (WHO), does not reliably predict fractures (Marshall et al., 1996), does not identify the majority who are at moderate risk (Pasco et al., 2006), and is limited for monitoring the effect of drug therapy (Delmas and Seeman, 2004). This has led to the development of clinical tools for predicting fracture risks such as the WHO Fracture Risk Assessment Tool (FRAX) (Kanis et al., 2005) and the Garvan fracture risk calculator (Nguyen et al., 2007). However, the validity of a purely clinical tool to predict fractures is still controversial (Leslie and Hans, 2009) and its accuracy may be limited by differences between cohorts (Sandhu et al., 2010).

In recent years it is increasingly recognized that trabecular microarchitecture confers bone its strength (Seeman and Delmas, 2006) and hence may explain the discrepancy between BMD and fracture risk (Delmas and Seeman, 2004). Several non-invasive methods [e.g. magnetic resonance imaging (MRI), and computed tomography (CT) scan] have been used to assess the microarchitecture of the different components of bone (Brandi, 2009). However despite their usefulness, there are other components of bone microarchitecture that have not been fully assessed. One of them is the presence of increasing levels of marrow fat (Burkhardt et al., 1987). In contrast to menopausal bone loss, age-related bone loss is not only associated with high levels of bone resorption, but also with increased adipogenesis (Rozman et al., 1989) and reduced osteoblastogenesis (Zhou et al., 2008), which affects bone mass. Biopsy studies have shown significant increase in marrow fat with age (Tanaka and Inoue, 1976), as well as an inverse relationship between fat volume (FV) and bone volume that was independent of sex and correlated with the changes seen in people with osteoporosis (Justesen et al., 2001).

Currently there are few non-invasive methods that have quantified marrow fat in humans. Among them, magnetic resonance imaging (MRI) has been the main modality showing increased marrow fat in older subjects (Schellinger et al., 2001) and in osteoporotic individuals (Yeung et al., 2005). However the correlations between MRI and BMD by DXA have been inconsistent and were limited to the vertebrae (Griffith et al., 2005, Schellinger et al., 2004, Shen et al., 2007). Furthermore, although studies using MRI suggest that the degree of marrow adiposity may be a better predictor of bone loss and thus could be used as a surrogate for fracture risk (Schellinger et al., 2004), the lack of histology correlation limits the validity of this conclusion. Due to these reasons and other technical limitations (e.g. metallic prosthesis), a reliable alternative method to quantify fat infiltration is still required. With the emerging use of high resolution imaging methods to assess bone microarchitecture, such as peripheral quantitative CT (pQCT) in humans and microCT (μCT) in bone samples, the development of new image analysis methods to quantify marrow fat could offer an alternative to MRI. In this study, we aimed to validate a new method of non-invasive quantification of bone marrow fat by correlating μCT image analyses with the gold standard (histology) in aging rats. We expect that this new method will facilitate marrow fat quantification in animal and human studies looking at the mechanisms of age-related bone loss and senile osteoporosis.

Section snippets

Animals

Twenty-two young mature (4-month-old, n = 12) and old (27-month-old, n = 10) male Louvain/c/rqrv (LOU) rats were studied. Male rats were selected due to their significantly higher levels of marrow fat infiltration as compared with old female LOU rats (Duque et al., 2009). Rats were obtained from the Aging LOU Rat Colony Infrastructure of the Quebec Network for Research on Aging (RQRV; www.rqrv.com). The rats were killed by rapid decapitation in block design fashion. Their bones were rapidly

Invasive and non-invasive identification of marrow fat in young and old LOU rats

Fig. 1 illustrates both bone micro architecture by histology analysis in young (A) and old (F) bone as compared with BV and FV labeling using SliceOMatic analysis of μCT sections obtained from the same femur in young (B, C–E) and old (G, H–J) male LOU rats. In agreement with fat identification using histological analysis (A and F), μCT labeling shows higher levels of FV in old rats (H–J) as compared to young rats (C–E).

Intra- and inter-rater reliability

For SliceOMatic/μCT the intra-rater reliability for duplicate measurements

Discussion

In this study we successfully correlated image and histology analysis of the marrow fat/bone relationship in the aging skeleton. We validated a non-invasive method of FV quantification with μCT images using well-established imaging software. Consistent with previous invasive studies (Justesen et al., 2001, Verma et al., 2002) and with our own previous report in this same model (Duque et al., 2009), our noninvasive quantification indicates that FV increases and BV decreases with age in this

Conclusion

In conclusion, non-invasive quantification of FV/BV ratio using image analysis software is a useful and reliable tool to quantify one of the hallmarks of age-related bone loss and senile osteoporosis. The fact that this method is comparable to histology for identifying age related changes in marrow adiposity assures the use of this noninvasive method for aging bone research in the near future.

Acknowledgements

This study was supported by a project grant from the Australian National Health and Medical Research Council (NHMRC, Grant # 632767) and by an operating grant from the Nepean Medical Research Foundation. A/Prof. Duque and Dr Li hold Fellowships from the University of Sydney Medical Research Foundation. Dr Demontiero holds a scholarship from The Rebecca Cooper Medical Research Foundation. The authors are deeply grateful to the members of the Quebec Network for Research in Aging for providing us

References (40)

  • A.J. Burghardt et al.

    A longitudinal HR-pQCT study of alendronate treatment in post-menopausal women with low bone density: relations between density, cortical and trabecular micro-architecture, biomechanics, and bone turnover

    J. Bone Miner. Res.

    (2010)
  • E. Demerath et al.

    Validity of a new automated software program for visceral adipose tissue estimation

    Int. J. Obes.

    (2007)
  • J. Griffith et al.

    Vertebral bone mineral density, marrow perfusion, and fat content in healthy men and men with osteoporosis: dynamic contrast-enhanced MR imaging and MR spectroscopy

    Radiology

    (2005)
  • T. Hirano et al.

    Bone marrow plays a role in bone metabolism: histomorphometry of iliac bone in postmenopausal women

    Calcif. Tissue Int.

    (1992)
  • I. Janssen et al.

    Estimation of skeletal muscle mass by bioelectrical impedance analysis

    J. Appl. Physiol.

    (2000)
  • J. Justesen et al.

    Adipocyte tissue volume in bone marrow is increased with aging and in patients with osteoporosis

    Biogerontology

    (2001)
  • J. Kanis et al.

    Assessment of fracture risk

    Osteoporos. Int.

    (2005)
  • J.L. Kuhn et al.

    Evaluation of a microcomputed tomography system to study trabecular bone structure

    J. Orthop. Res.

    (1990)
  • W. Leslie et al.

    Commentary on lumpers and splitters: the FRAX debate continues

    J. Bone Miner. Res.

    (2009)
  • X.S. Liu et al.

    Bone density, geometry, microstructure and stiffness: relationships between peripheral and central skeletal sites assessed by DXA, HR-pQCT, and cQCT in premenopausal women

    J. Bone Miner. Res.

    (2010)
  • Cited by (25)

    • Tongue muscle mass is associated with total grey matter and hippocampal volumes in Dementia with Lewy Bodies

      2022, Archives of Gerontology and Geriatrics
      Citation Excerpt :

      Slice-O-Matic measures, on the other hand, have been verified against gold standards in the past, exhibiting excellent inter- and intra-rater reliability, as well as good reliability and validity. ( Demontiero et al., 2011) The sample size was small due to imaging availability and quality selection, and longitudinal analyses were not performed because images were only available at baseline. Tongue muscle volume at diagnosis was associated with total grey and hippocampal volume in DLB, indicating that measuring the tongue muscle can offer a simple, non-invasive and cost-effective measurement of muscle volume since most patients with suspected or confirmed dementia will have MRI scans.

    • Simultaneous three-dimensional visualization of mineralized and soft skeletal tissues by a novel microCT contrast agent with polyoxometalate structure

      2018, Biomaterials
      Citation Excerpt :

      Aging induced a significant increase in the bone marrow adipocyte volume fraction (8.9-fold; Fig. 5D), adipose density (5.3-fold; Fig. 5F) and average adipocyte diameter (1.6-fold; Fig. 5E). This effect corresponds well with the literature, reporting an aging-associated increase in the bone marrow adiposity [72,73]. Different from most published data, we were now also able, because of the high spatial and contrast resolution of our images, to quantify the morphological parameters of individual adipocytes, such as the true adipocyte diameter and its distribution (measured in 3D).

    • Use of osmium tetroxide staining with microcomputerized tomography to visualize and quantify bone marrow adipose tissue in vivo

      2014, Methods in Enzymology
      Citation Excerpt :

      If adequate analysis is not performed, traditional sectioning methods can easily miss changes in MAT volume and/or distribution. In species ranging in size from rat to humans indirect imaging techniques, including computed tomography and magnetic resonance (MR) spectroscopy, have been applied with success (Bredella et al., 2009; Demontiero, Li, Thembani, & Duque, 2011; Regis-Arnaud et al., 2011). Although MR has been tried in isolated mouse femurs, quantitation of fat verses water is very imprecise (C.J. Rosen, unpublished).

    • Lamin A/C deficiency is associated with fat infiltration of muscle and bone

      2011, Mechanisms of Ageing and Development
      Citation Excerpt :

      Muscle and inter-myofiber fat volumes were quantified as a relative percentage of their summation (total muscle volume). As per our previous report (Demontiero et al., 2011), there was a significant correlation (r2 = 0.92) between volumes obtained by histology vs. those quantified using our image analysis software. Our histology and microCT analysis showed that muscle volume was significantly reduced from 86.6% in WT mice to 68.4% in heterozygous (Lmna+/−) mice to 43.8% in Lmna−/− mice (p < 0.001) (Fig. 1).

    View all citing articles on Scopus
    View full text