- Email: [email protected]
P maps of bone architecture in 3D synchrotron radiation microtomographic images
ARTICLE IN PRESS
Applied Radiation and Isotopes 62 (2005) 569–575 www.elsevier.com/locate/apradiso
High resolution Ca/P maps of bone architecture in 3D synchrotron radiation microtomographic images M. Tzaphlidoua,, R. Spellerb, G. Royleb, J. Griffithsb, A. Olivoc, S. Panic, R. Longoc a Laboratory of Medical Physics, Medical School, University of Ioannina, PO Box 1186, 45110 Ioannina, Greece Department of Medical Physics and Bioengineering, University College London, 11-20 Capper Street, London WC1E 6JA, UK c Department of Physics, University of Trieste, Italy
b
Received 26 January 2004; received in revised form 4 October 2004; accepted 12 October 2004
Abstract The Ca/P ratio was measured in cortical bone samples from the femoral neck and tibia of different animal species, using synchrotron radiation microtomography. Use of a monoenergetic X-ray beam, as provided by the synchrotron facility, generates accurate 3D maps of the linear attenuation coefficient within the sample and hence gives the ability to map different chemical components. Also, by comparing normal and abnormal bones, i.e. osteoporotic (induced by inflammation), changes in the Ca/P ratio brought about by bone diseases can be detected. MicroCT data sets were collected at 20 and 28 keV for each bone sample and two calibration phantoms. From the 3D data sets, multiple 2D slices were reconstructed with a slice thickness of 30 mm. Regions of interest were defined around suitable sites and were converted to Ca/P ratios using the data collected from the test phantoms. A significant difference ðpo0:001Þ between osteoporotics and age-matched normals at both energies was detected. Differences between different bone sites from the same animal are not significant ðp40:5Þ while those between the same bone sites from different animals are highly significant ðpo0:001Þ: Differences between estimates made at 20 and 28 keV are not significant ðp40:5Þ: An important aspect is the ability to map the spatial distribution of the Ca/P ratio. r 2004 Elsevier Ltd. All rights reserved. Keywords: Synchrotron radiation microtomography; Ca/P ratio; Cortical bone
1. Introduction Reliable measurement of the dynamic and static strength of the skeleton is a requirement for the rational approach to the clinical problem of skeletal diseases. It is generally accepted that a measurement of bone mineral content is an appropriate estimate of bone quality and Corresponding author. Tel.: +30 2651 097595; fax: +30 2651 097854. E-mail address: [email protected] (M. Tzaphlidou).
most current clinical techniques make an estimate of bone mineral density without considering the role played by the different chemical components that are present. The aim of this work was to measure the Ca/P ratio in different kinds of bones and to seek information on whether this ratio changes with bone abnormalities. This is because changes in the amounts of Ca and P in biological apatites do not go hand-in-hand; therefore, a decrease in bone density may be due to a decrease in either Ca or P, or to dissimilar decreases in both. Consequently, the determination of the Ca/P ratio may
0969-8043/$ - see front matter r 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.apradiso.2004.10.003
ARTICLE IN PRESS 570
M. Tzaphlidou et al. / Applied Radiation and Isotopes 62 (2005) 569–575
provide a sensitive measure of bone mineral changes and may add to our understanding of the changes occurring in bone diseases. A new clinical method has been put forward for assessing the skeletal Ca/P ratio in vivo, using photon absorptiometry (Fountos et al., 1997, 1998, 1999). The new system uses two photon energies of 39 and 89 KeV obtained by placing cerium and samarium filters in the X-ray beam. It is optimized for measuring the Ca/P ratio at a fixed site in the distal third of the right radius. The method assumes that bone is a three-component system: Ca, PO4 and water. This may be a useful measurement to assesss bone disorders and age changes, particularly if it can be accomplished noninvasively. In order to evaluate the value of the new method as a diagnostic indicator for bone disorders, the precision and accuracy of the technique must be sufficient to satisfy the requirement for being able to distinguish differences between study populations. To determine the accuracy of in vivo Ca/P ratio measurements, we need to compare them with in vitro measurements, by this way the range of Ca/P values in different parts of the same bone or in different bones can be calculated most accurately. Synchrotron sources provide new, more accurate possibilities to assess bone mineral content from intact bone samples (Peyrin et al., 2000; Nuzzo et al., 2002a; Postnov et al., 2003). Use of a monoenergetic, high photon flux X-ray beam, as provided by a synchrotron facility, allows high resolution and high signal-to-noise ratio imaging (Salome et al., 1999). In the study of hard tissue such as bone, it is particularly useful because of its ability to accurately measure the linear attenuation coefficient (Davis and Wong, 1996). Thus, in a CT procedure, 3D maps of the linear attenuation coefficient within the sample, unaffected by beam hardening can be generated. When produced with high spatial resolution these data sets provide the ability to map different chemical components (Kinney et al., 2000; Nuzzo et al., 2002a) throughout the volume of the bone sample. Computed microtomography has been applied successfully to trabecular and cortical bones from humans (Salome et al., 1999; Peyrin et al., 2000; Nuzzo et al., 2002a,b) and rats (Stenstrom et al., 1998, 2000; Barbier et al., 1999; Laib et al., 2000). It has also been used for mice, where the thinner bone structures require very high spatial resolution (Martin-Badosa et al., 2003). It is known that the mechanical strength of bone depends first of all, on the condition of the cortical bone (Stein and Granik, 1976; Stenstrom et al., 2000). Hence, in the present study, rabbit cortical bone from the femoral neck and tibia regions was analyzed with two objectives in mind:
Firstly,
to estimate the bulk Ca/P ratio for intact bones as a basis for comparison with in vivo studies.
Secondly,
to study regions most affected by bone disease and to identify the significance of the distribution of the Ca/P ratio.
2. Materials and methods 2.1. Bone samples Animals were housed in groups of two and were allowed to freely take solid diet and tap water. The breeding room was light controlled (light period from 7.00 am to 7.00 pm) and room temperature (2071 1C). Cortical right tibia samples from four female normal subjects were analysed. Two of them were from 1-yearold lambs and two were from 3-year-old sheep. In addition, cortical bone from the right tibia and femoral neck from two female normal 7-month-old rabbits, and cortical bone from the right tibia from two osteoporotic (induced by inflammation) sex- and age-matched rabbits was measured. These samples were chosen as a model for evaluating the sensitivity of the Ca/P ratio estimates as they covered different animals, different sites and different bone conditions. Inflammation-mediated osteoporosis was induced by subcutaneous injections of magnesium silicate (talkum) in the flanks, as previously described (Kounadi et al., 1998). Animals were injected once a day for 20 consecutive days and sacrificed 1 day after the last injection. This method causes osteoporosis associated with chronic inflammation to be provoked (Minne et al., 1984; Bauss et al., 1985). As previously reported, osteoporosis induced by this way has significant effects on the skeletal Ca/P ratio (Fountos et al., 1998) and on the bone collagen structure (Kounadi et al., 1998). In all cases the samples were held in thin walled plastic tubes of 10 mm diameter. A section approximately 4 mm long was imaged for each sample. Throughout the animal experiments, which were performed at Ioannina University, care was taken to minimize pain or discomfort. All studies were approved by the Ioannina University Institutional Animal Care and Use Committee. 2.2. Equipment and techniques All experiments were performed at the ELETTRA Synchrotron Light Laboratory in Trieste, Italy. MicroCT data sets were collected using the CT set-up on the SYRMEP (Synchrotron Radiation for Medical Physics) beamline. This set-up consists of a high precision rotary table mounted upon high-accuracy translators and cradles. This allows precise alignment of the rotation axis of the sample with the detector pixels. The detector is an X-ray sensitive photonic
ARTICLE IN PRESS M. Tzaphlidou et al. / Applied Radiation and Isotopes 62 (2005) 569–575
science CCD camera and was operated in a 2 2 binning mode, providing 28 mm square pixels. A full data set consisted of 360 views over 1801 and for each sample, data were collected at 20 and 28 keV. These two energies were chosen as the best compromise that provided the greatest energy separation but that also took into account the competing processes of photon attenuation in the sample and photon intensity at the source. Two calibration phantoms containing Ca(HPO4) 2H2O and Ca(H2PO4)2(H2O) with densities of 2.31 and 2.22 g/ml, respectively, as bone substitutes, were also imaged. Previous experiments had shown that the calibration materials demonstrated a well-understood behaviour ð Z3effective Þ over the restricted range of Ca/P to be measured in this study and hence only two calibration phantoms were used. These phantoms were chosen for calibration as they were also used for the in vivo measurements and thus, they should be considered to be the ideal phantoms for the in vitro measurements. 2.3. Analysis Using a filtered backprojection routine, multiple 2D slices (see example in Fig. 1) were reconstructed from the 3D data sets with a slice thickness of 30 mm. All of the software was written in IDL (Interactive Data Language, Research Systems Inc, Boulder, CO 80301) and a full data set typically contained 140 slices. Analysis of these images was carried out in two ways:
For
each slice, regions of interest were identified around suitable sites. A total of four, uniformly
571
distributed sites were used per slice and all selected sites were away from obvious boundaries and inclusions. Mean CT values within the ROI were converted to Ca/P ratios using the data collected from the calibration phantoms. Bulk estimates for the Ca/P ratio were then computed as the mean and standard deviation of the Ca/P ratio values from the individual slices. For selected bone sites, the spatial distribution of the Ca/P ratio was studied by forming contour plots using the appropriate slices.
The reliability of the difference between different bone sites, as well as between different animal species was evaluated by the Student’s t-test.
3. Results Table 1 summarises the results from the analysis of the bulk Ca/P ratios. Results show a significant difference ðpo0:001Þ between osteoporotics and agematched normals at both energies. Differences between different bone sites from the same animal are not significant ðp40:5Þ while those between the same bone sites from different animals are highly significant ðpo0:001Þ: Table 2 shows the groups of samples that were statistically compared and the significance of the differences (p values). Differences between estimates made at 20 and 28 keV are small without any statistical significance ðp40:5Þ: However, an important aspect that is unique to the measurements made at a synchrotron source such as ELETTRA is the ability to map the spatial distribution of the Ca/P ratio. This has been investigated by using contour plots for selected slices, where the local gradients exhibited in the contour plots are a measure of the changing Ca/P ratio and hence can be used to study the changing distribution. Fig. 2 shows examples
Table 1 Bulk values of the CA/P ratio measured at 20 and 28 keV
Fig. 1. MicroCT image of a normal rabbit tibia cortical sample contained within a thin-walled plastic tube. The image is taken at 20 keV using 360 projections over 1801. The image has been reconstructed using filtered backprojection.
Bone type
Mean Ca/P (20 keV)
Mean Ca/P (28 keV)
Femoral rabbit (1) Femoral rabbit (2) Tibia rabbit (1) Tibia rabbit (2) Tibia osteoporotic rabbit (1) Tibia osteoporotic rabbit (2) Tibia lamb (1) Tibia lamb (2) Tibia sheep (1) Tibia sheep (2)
1.5570.18 1.2270.19 1.8870.06 1.7970.05 1.4570.04 1.6670.06 1.0170.04 1.1970.04 1.2870.04 1.5570.05
1.5470.37 1.1470.26 1.8470.31 1.9770.05 1.6770.18 1.7570.13 0.9670.05 1.0570.06 1.1770.07 1.6570.04
ARTICLE IN PRESS 572
M. Tzaphlidou et al. / Applied Radiation and Isotopes 62 (2005) 569–575
Table 2 Groups of samples that were statistically compared and the significance of the differences (p values) Groups compared
20 keV p values
28 keV p values
Tibia rabbit (1,2) Tibia osteoporotic rabbit (1,2)
o 0.001
o 0.001
Femoral rabbit (1) Tibia rabbit (1)
40.5
40.5
Femoral rabbit (2) Tibia rabbit (2)
40.5
40.5
Femoral rabbit (1) Femoral rabbit (2)
o0.001
o0.001
Tibia rabbit (1) Tibia rabbit (2)
o0.001
o0.001
Tibia sheep (1) Tibia sheep (2)
o0.001
o0.001
p values
Femoral rabbit (1) 20 keV Femoral rabbit (1) 28 keV
40.5
Femoral rabbit (2) 20 keV Femoral rabbit (2) 28 keV
40.5
Tibia rabbit (1) 20 keV Tibia rabbit (1) 28 keV
40.5
Tibia lamb (1) 20 keV Tibia lamb (1) 28 keV
40.5
Tibia lamb (2) 20 keV Tibia lamb (2) 28 keV
40.5
bone and to use spatial measurements to look at the distribution of Ca and P throughout the sample. 4.1. Ca/P ratio and bone condition A new method has been developed at the University of Ioannina for the estimation of the in vivo Ca/P ratio in the human radius. This approach uses a modified dual energy X-ray absorptiometry system (Fountos et al., 1999). Using this instrument, a significant difference in the mean Ca/P ratio between postmenopausal osteoporotic patients and premenopausal controls has been observed (Fountos et al., 1999) and from studies in rabbits, a significant decrease in the tibia Ca/P ratio in animals with inflammation-mediated osteoporosis has also been found (Fountos et al., 1998). The same authors studying the relationships between the skeletal Ca/P ratio and the structural changes in the bone in response to osteoporosis, suggest that alterations in the Ca/P ratio may be indicative of underlying changes in the organic matrix of bone, mainly in collagen. Other studies (Kounadi et al., 1998) demonstrate that in osteoporotic rabbits (induced by inflammation) severe structural abnormalities in the rabbit tibia collagen occur. These abnormalities consist of a significant alteration in mean fibril diameter as well as marked disorganization in the packing of the fibrils. Our results at ELETTRA confirm that a statistically significant decrease is observed in rabbit tibia Ca/P ratio for osteoporotic samples suggesting that there is a relationship between bone loss and a lowered Ca/P ratio (see Table 1). 4.2. Ca/P ratio and bone type
of the Ca/P contour plots for different bone types, sites and bone conditions taken from the 28 keV images. Contours have been plotted at Ca/P ratios of 60%, 80%, 85%, 90% and 97% of the maximum values in the image.
4. Discussion The experiments described in this paper, using the microtomography system on the SYRMEP beamline at the ELETTRA synchrotron light laboratory, were designed to confirm and extend previous results on
The variations with different animals and different sites detected in this work indicate the importance of selecting suitable animal models when applying results to human studies. Other workers performing measurements on intact bones by neutron activation analysis have shown that the Ca/P ratio depends upon the kind of bone studied (Tzaphlidou and Zaichick, 2002, 2003; Zaichick and Tzaphlidou, 2002, 2003). The different values obtained in the different intact bone sites are due to the influence of the organic matrix, i.e. fat, lipids and marrow (particularly that of red marrow) which depend on bone kind (Bolotin and Sievanen, 2001; Fountos et al., 1997; Tzaphlidou and Zaichick, 2002).
Fig. 2. Contour plots, derived from the microCT images (600 600 pixels), of the Ca/P ratio for normal and osteoporotic rabbit tibia samples. Contours at 60%, 80%, 85%, 90% and 97% of maximum Ca/P within the sample have been plotted in each sample. It can be seen that the contour density is increased in osteoporotic samples and can also demonstrate regions of non-uniformity. The pixel size is approximately 12 mm.
ARTICLE IN PRESS M. Tzaphlidou et al. / Applied Radiation and Isotopes 62 (2005) 569–575
573
ARTICLE IN PRESS 574
M. Tzaphlidou et al. / Applied Radiation and Isotopes 62 (2005) 569–575
4.3. Ca/P spatial distribution and bone condition The spatial distribution of the Ca/P ratio can be seen in Fig. 2. It appears that normal bone shows less variation in the distribution of Ca/P as demonstrated by the near absence of contours throughout the bone volume. However, the osteoporotic examples show a higher density of Ca/P contours in general, but also display regions of significantly higher Ca/P gradient changes. These changes are not fully understood but could be due to local exchanges of Ca and P. A further interesting observation is the variation in the contour density within a single sample. It can be seen in Fig. 2 that a sample which is judged to be ‘osteoporotic’ from the bulk estimate of the Ca/P ratio may show regions of ‘osteoporotic’ behaviour and regions of ‘normal’ behaviour. Further work is required to confirm and understand these observations and their full interpretation is still ongoing. In conclusion, synchrotron radiation microtomography is potentially an outstanding in vitro technique to map different bone chemical components and to detect changes brought about by bone diseases. As has been reported (Tzaphlidou and Zaichick, 2002, 2003; Zaichick and Tzaphlidou, 2002), the Ca/P ratio may provide greater reliability for diagnosis of bone disorders and hence microtomography may become a valuable technique to be used during bone therapeutic and diagnostic trials.
Acknowledgements This work was undertaken at the Elettra synchrotron radiation source and was funded by the Human Potential Programme—Transnational Access to Major Research Infrastructures of the European Community.
References Barbier, A., Martel, C., de Vernejoul, M.C., Tirode, F., Nys, M., Mocaer, G., Morieux, C., Murakami, H., Lacheretz, F., 1999. The visualization and evaluation of bone architecture in the rat using three-dimensional X-ray microcomputed tomography. J. Bone Mineral. Metab. 17, 37–44. Bauss, F., Minne, H.W., Sterz, H., Weng, U., Wesch, H., Ziegler, R., 1985. Comparative bone analysis via inflammation-mediated osteopenia (IMO) in the rat. Calcif. Tissue Int. 37, 539–546. Bolotin, H.H., Sievanen, H., 2001. Inaccuracies inherent in dual-energy X-ray absorptiometry in vivo bone mineral density can seriously mislead diagnostic/prognostic interpretations of patient-specific bone fragility. J. Bone Mineral Res. 16, 799–805. Davis, G.R., Wong, F.S., 1996. X-ray microtomography of bones and teeth. Physiol. Meas. 17, 121–146.
Fountos, G., Yasumura, S., Glaros, D., 1997. The skeletal calcium/phosphorus ratio: a new in vivo method of determination. Med. Phys. 24, 1303–1310. Fountos, G., Kounadi, E., Tzaphlidou, M., Yasumura, S., Glaros, D., 1998. The effect of inflammation-mediated osteoporosis (IMO) on the skeletal Ca/P ratio and on the structure of rabbit bone and skin collagen. Appl. Radiat. Isot. 49, 657–679. Fountos, G., Tzaphlidou, M., Kounadi, E., Glaros, D., 1999. In vivo measurement of radius calcium/phosphorus ratio by X-ray absorptiometry. Appl. Radiat. Isot. 51, 273–278. Kinney, J.H., Haupt, D.L., Balooch, M., Ladd, A.J., Lane, N.E., 2000. Three-dimensional morphometry of the L6 vertebra in the ovariectomized rat model of osteoporosis: biomechanical implications. J. Bone Miner. Res. 15, 1981–1991. Kounadi, E., Fountos, G., Tzaphlidou, M., 1998. The influence of inflammation-mediated osteopenia (IMO) on the structure of rabbit bone and skin collagen fibrils. Connect. Tissue Res. 37, 69–76. Laib, A., Barou, O., Vico, L., Lafage-Proust, M.H., Alexandre, C., Ruegsegger, P., 2000. 3D micro-computed tomography of trabecular and cortical bone architecture with application to a rat model of immobilisation osteoporosis. Med. Biol. Eng. Comput. 38, 326–332. Martin-Badosa, E., Elmoutaouakkil, A., Nuzzo, S., Amblard, D., Vico, L., Peyrin, F., 2003. A method for the automatic characterization of bone architecture in 3D mice microtomographic images. Comp. Med. Imag. Graphics 27, 447–458. Minne, H.W., Pfeilschifter, J., Scharla, S., Mutschelknauss, S., Schwarz, A., Krempien, B., Ziegler, R., 1984. Inflammationmediated osteopenia in the rat: a new animal model for pathological loss of bone mass. Endocrinology 115, 50–54. Nuzzo, S., Peyrin, F., Cloetens, P., Baruchel, J., Boivin, G., 2002a. Quantification of the degree of mineralization of bone in three dimensions using synchrotron radiation microtomography. Med. Phys. 29, 2672–2681. Nuzzo, S., Lafage-Proust, M.H., Martin-Badosa, E., Boivin, G., Thomas, T., Alexandre, C., Peyrin, F., 2002b. Synchrotron radiation microtomography allows the analysis threedimensional microarchitecture and degree of mineralization of human iliac crest biopsy specimens: effects etidronate treatment. J. Bone Mineral Res. 17, 1372–1382. Peyrin, F., Salome, M., Nuzzo, S., Cloetens, P., Laval-Jeantet, A.M., Baruchel, J., 2000. Perspectives in three-dimensional analysis of bone samples using synchrotron radiation microtomography. Cell Mol. Biol. 46, 1089–1102. Postnov, A.A., Vinogradov, A.V., Van Dyck, D., Saveliev, S.V., De Clerck, N.M., 2003. Quantitative analysis of bone mineral content by X-ray microtomography. Physiol. Meas. 24, 165–178. Salome, M., Peyrin, F., Cloetens, P., Odet, C., Laval-Jeantet, A.M., Baruchel, J., Spanne, P., 1999. A synchrotron radiation microtomography system for the analysis of trabecular bone samples. Med. Phys. 26, 2194–2204. Stenstrom, M., Olander, B., Carlsson, C.A., Carlsson, G.A., Lehto-Axtelius, D., Hakanson, R., 1998. The use of computed microtomography to monitor morphological changes in small animals. Appl. Radiat. Isot. 49, 565–570.
ARTICLE IN PRESS M. Tzaphlidou et al. / Applied Radiation and Isotopes 62 (2005) 569–575 Stenstrom, M., Olander, B., Lehto-Axtelius, D., Madsen, J.E., Nordsletten, L., Carlsson, G.A., 2000. Bone mineral density and bone structure parameters as predictors of bone strength: an analysis using computerized microtomography and gastrectomy-induced osteopenia in the rat. J. Biomech. 33, 289–297. Stein, I.D., Granik, G., 1976. Rib structure and bending strength: an autopsy study. Calcif. Tissue Res. 20, 61–73. Tzaphlidou, M., Zaichick, V., 2002. Neutron activation analysis of calcium/phosphorus ratio in rib bone of healthy humans. Appl. Radiat. Isot. 57, 779–783.
575
Tzaphlidou, M., Zaichick, V., 2003. Calcium, phosphorus, calcium-phosphorus ratio in rib bone of healthy humans. Biol. Trace Element Res. 93, 63–74. Zaichick, V., Tzaphlidou, M., 2002. Determination of calcium, phosphorus, and the calcium/phosphorus ratio in cortical bone from the human femoral neck by neutron activation analysis. Appl. Radiat. Isot. 56, 781–786. Zaichick, V., Tzaphlidou, M., 2003. Calcium and phosphorus concentrations and the calcium/phosphorus ratio in trabecular bone from the femoral neck of healthy humans as determined by neutron activation analysis. Appl. Radiat. Isot. 58, 623–627.
Applied Radiation and Isotopes 62 (2005) 569–575 www.elsevier.com/locate/apradiso
High resolution Ca/P maps of bone architecture in 3D synchrotron radiation microtomographic images M. Tzaphlidoua,, R. Spellerb, G. Royleb, J. Griffithsb, A. Olivoc, S. Panic, R. Longoc a Laboratory of Medical Physics, Medical School, University of Ioannina, PO Box 1186, 45110 Ioannina, Greece Department of Medical Physics and Bioengineering, University College London, 11-20 Capper Street, London WC1E 6JA, UK c Department of Physics, University of Trieste, Italy
b
Received 26 January 2004; received in revised form 4 October 2004; accepted 12 October 2004
Abstract The Ca/P ratio was measured in cortical bone samples from the femoral neck and tibia of different animal species, using synchrotron radiation microtomography. Use of a monoenergetic X-ray beam, as provided by the synchrotron facility, generates accurate 3D maps of the linear attenuation coefficient within the sample and hence gives the ability to map different chemical components. Also, by comparing normal and abnormal bones, i.e. osteoporotic (induced by inflammation), changes in the Ca/P ratio brought about by bone diseases can be detected. MicroCT data sets were collected at 20 and 28 keV for each bone sample and two calibration phantoms. From the 3D data sets, multiple 2D slices were reconstructed with a slice thickness of 30 mm. Regions of interest were defined around suitable sites and were converted to Ca/P ratios using the data collected from the test phantoms. A significant difference ðpo0:001Þ between osteoporotics and age-matched normals at both energies was detected. Differences between different bone sites from the same animal are not significant ðp40:5Þ while those between the same bone sites from different animals are highly significant ðpo0:001Þ: Differences between estimates made at 20 and 28 keV are not significant ðp40:5Þ: An important aspect is the ability to map the spatial distribution of the Ca/P ratio. r 2004 Elsevier Ltd. All rights reserved. Keywords: Synchrotron radiation microtomography; Ca/P ratio; Cortical bone
1. Introduction Reliable measurement of the dynamic and static strength of the skeleton is a requirement for the rational approach to the clinical problem of skeletal diseases. It is generally accepted that a measurement of bone mineral content is an appropriate estimate of bone quality and Corresponding author. Tel.: +30 2651 097595; fax: +30 2651 097854. E-mail address: [email protected] (M. Tzaphlidou).
most current clinical techniques make an estimate of bone mineral density without considering the role played by the different chemical components that are present. The aim of this work was to measure the Ca/P ratio in different kinds of bones and to seek information on whether this ratio changes with bone abnormalities. This is because changes in the amounts of Ca and P in biological apatites do not go hand-in-hand; therefore, a decrease in bone density may be due to a decrease in either Ca or P, or to dissimilar decreases in both. Consequently, the determination of the Ca/P ratio may
0969-8043/$ - see front matter r 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.apradiso.2004.10.003
ARTICLE IN PRESS 570
M. Tzaphlidou et al. / Applied Radiation and Isotopes 62 (2005) 569–575
provide a sensitive measure of bone mineral changes and may add to our understanding of the changes occurring in bone diseases. A new clinical method has been put forward for assessing the skeletal Ca/P ratio in vivo, using photon absorptiometry (Fountos et al., 1997, 1998, 1999). The new system uses two photon energies of 39 and 89 KeV obtained by placing cerium and samarium filters in the X-ray beam. It is optimized for measuring the Ca/P ratio at a fixed site in the distal third of the right radius. The method assumes that bone is a three-component system: Ca, PO4 and water. This may be a useful measurement to assesss bone disorders and age changes, particularly if it can be accomplished noninvasively. In order to evaluate the value of the new method as a diagnostic indicator for bone disorders, the precision and accuracy of the technique must be sufficient to satisfy the requirement for being able to distinguish differences between study populations. To determine the accuracy of in vivo Ca/P ratio measurements, we need to compare them with in vitro measurements, by this way the range of Ca/P values in different parts of the same bone or in different bones can be calculated most accurately. Synchrotron sources provide new, more accurate possibilities to assess bone mineral content from intact bone samples (Peyrin et al., 2000; Nuzzo et al., 2002a; Postnov et al., 2003). Use of a monoenergetic, high photon flux X-ray beam, as provided by a synchrotron facility, allows high resolution and high signal-to-noise ratio imaging (Salome et al., 1999). In the study of hard tissue such as bone, it is particularly useful because of its ability to accurately measure the linear attenuation coefficient (Davis and Wong, 1996). Thus, in a CT procedure, 3D maps of the linear attenuation coefficient within the sample, unaffected by beam hardening can be generated. When produced with high spatial resolution these data sets provide the ability to map different chemical components (Kinney et al., 2000; Nuzzo et al., 2002a) throughout the volume of the bone sample. Computed microtomography has been applied successfully to trabecular and cortical bones from humans (Salome et al., 1999; Peyrin et al., 2000; Nuzzo et al., 2002a,b) and rats (Stenstrom et al., 1998, 2000; Barbier et al., 1999; Laib et al., 2000). It has also been used for mice, where the thinner bone structures require very high spatial resolution (Martin-Badosa et al., 2003). It is known that the mechanical strength of bone depends first of all, on the condition of the cortical bone (Stein and Granik, 1976; Stenstrom et al., 2000). Hence, in the present study, rabbit cortical bone from the femoral neck and tibia regions was analyzed with two objectives in mind:
Firstly,
to estimate the bulk Ca/P ratio for intact bones as a basis for comparison with in vivo studies.
Secondly,
to study regions most affected by bone disease and to identify the significance of the distribution of the Ca/P ratio.
2. Materials and methods 2.1. Bone samples Animals were housed in groups of two and were allowed to freely take solid diet and tap water. The breeding room was light controlled (light period from 7.00 am to 7.00 pm) and room temperature (2071 1C). Cortical right tibia samples from four female normal subjects were analysed. Two of them were from 1-yearold lambs and two were from 3-year-old sheep. In addition, cortical bone from the right tibia and femoral neck from two female normal 7-month-old rabbits, and cortical bone from the right tibia from two osteoporotic (induced by inflammation) sex- and age-matched rabbits was measured. These samples were chosen as a model for evaluating the sensitivity of the Ca/P ratio estimates as they covered different animals, different sites and different bone conditions. Inflammation-mediated osteoporosis was induced by subcutaneous injections of magnesium silicate (talkum) in the flanks, as previously described (Kounadi et al., 1998). Animals were injected once a day for 20 consecutive days and sacrificed 1 day after the last injection. This method causes osteoporosis associated with chronic inflammation to be provoked (Minne et al., 1984; Bauss et al., 1985). As previously reported, osteoporosis induced by this way has significant effects on the skeletal Ca/P ratio (Fountos et al., 1998) and on the bone collagen structure (Kounadi et al., 1998). In all cases the samples were held in thin walled plastic tubes of 10 mm diameter. A section approximately 4 mm long was imaged for each sample. Throughout the animal experiments, which were performed at Ioannina University, care was taken to minimize pain or discomfort. All studies were approved by the Ioannina University Institutional Animal Care and Use Committee. 2.2. Equipment and techniques All experiments were performed at the ELETTRA Synchrotron Light Laboratory in Trieste, Italy. MicroCT data sets were collected using the CT set-up on the SYRMEP (Synchrotron Radiation for Medical Physics) beamline. This set-up consists of a high precision rotary table mounted upon high-accuracy translators and cradles. This allows precise alignment of the rotation axis of the sample with the detector pixels. The detector is an X-ray sensitive photonic
ARTICLE IN PRESS M. Tzaphlidou et al. / Applied Radiation and Isotopes 62 (2005) 569–575
science CCD camera and was operated in a 2 2 binning mode, providing 28 mm square pixels. A full data set consisted of 360 views over 1801 and for each sample, data were collected at 20 and 28 keV. These two energies were chosen as the best compromise that provided the greatest energy separation but that also took into account the competing processes of photon attenuation in the sample and photon intensity at the source. Two calibration phantoms containing Ca(HPO4) 2H2O and Ca(H2PO4)2(H2O) with densities of 2.31 and 2.22 g/ml, respectively, as bone substitutes, were also imaged. Previous experiments had shown that the calibration materials demonstrated a well-understood behaviour ð Z3effective Þ over the restricted range of Ca/P to be measured in this study and hence only two calibration phantoms were used. These phantoms were chosen for calibration as they were also used for the in vivo measurements and thus, they should be considered to be the ideal phantoms for the in vitro measurements. 2.3. Analysis Using a filtered backprojection routine, multiple 2D slices (see example in Fig. 1) were reconstructed from the 3D data sets with a slice thickness of 30 mm. All of the software was written in IDL (Interactive Data Language, Research Systems Inc, Boulder, CO 80301) and a full data set typically contained 140 slices. Analysis of these images was carried out in two ways:
For
each slice, regions of interest were identified around suitable sites. A total of four, uniformly
571
distributed sites were used per slice and all selected sites were away from obvious boundaries and inclusions. Mean CT values within the ROI were converted to Ca/P ratios using the data collected from the calibration phantoms. Bulk estimates for the Ca/P ratio were then computed as the mean and standard deviation of the Ca/P ratio values from the individual slices. For selected bone sites, the spatial distribution of the Ca/P ratio was studied by forming contour plots using the appropriate slices.
The reliability of the difference between different bone sites, as well as between different animal species was evaluated by the Student’s t-test.
3. Results Table 1 summarises the results from the analysis of the bulk Ca/P ratios. Results show a significant difference ðpo0:001Þ between osteoporotics and agematched normals at both energies. Differences between different bone sites from the same animal are not significant ðp40:5Þ while those between the same bone sites from different animals are highly significant ðpo0:001Þ: Table 2 shows the groups of samples that were statistically compared and the significance of the differences (p values). Differences between estimates made at 20 and 28 keV are small without any statistical significance ðp40:5Þ: However, an important aspect that is unique to the measurements made at a synchrotron source such as ELETTRA is the ability to map the spatial distribution of the Ca/P ratio. This has been investigated by using contour plots for selected slices, where the local gradients exhibited in the contour plots are a measure of the changing Ca/P ratio and hence can be used to study the changing distribution. Fig. 2 shows examples
Table 1 Bulk values of the CA/P ratio measured at 20 and 28 keV
Fig. 1. MicroCT image of a normal rabbit tibia cortical sample contained within a thin-walled plastic tube. The image is taken at 20 keV using 360 projections over 1801. The image has been reconstructed using filtered backprojection.
Bone type
Mean Ca/P (20 keV)
Mean Ca/P (28 keV)
Femoral rabbit (1) Femoral rabbit (2) Tibia rabbit (1) Tibia rabbit (2) Tibia osteoporotic rabbit (1) Tibia osteoporotic rabbit (2) Tibia lamb (1) Tibia lamb (2) Tibia sheep (1) Tibia sheep (2)
1.5570.18 1.2270.19 1.8870.06 1.7970.05 1.4570.04 1.6670.06 1.0170.04 1.1970.04 1.2870.04 1.5570.05
1.5470.37 1.1470.26 1.8470.31 1.9770.05 1.6770.18 1.7570.13 0.9670.05 1.0570.06 1.1770.07 1.6570.04
ARTICLE IN PRESS 572
M. Tzaphlidou et al. / Applied Radiation and Isotopes 62 (2005) 569–575
Table 2 Groups of samples that were statistically compared and the significance of the differences (p values) Groups compared
20 keV p values
28 keV p values
Tibia rabbit (1,2) Tibia osteoporotic rabbit (1,2)
o 0.001
o 0.001
Femoral rabbit (1) Tibia rabbit (1)
40.5
40.5
Femoral rabbit (2) Tibia rabbit (2)
40.5
40.5
Femoral rabbit (1) Femoral rabbit (2)
o0.001
o0.001
Tibia rabbit (1) Tibia rabbit (2)
o0.001
o0.001
Tibia sheep (1) Tibia sheep (2)
o0.001
o0.001
p values
Femoral rabbit (1) 20 keV Femoral rabbit (1) 28 keV
40.5
Femoral rabbit (2) 20 keV Femoral rabbit (2) 28 keV
40.5
Tibia rabbit (1) 20 keV Tibia rabbit (1) 28 keV
40.5
Tibia lamb (1) 20 keV Tibia lamb (1) 28 keV
40.5
Tibia lamb (2) 20 keV Tibia lamb (2) 28 keV
40.5
bone and to use spatial measurements to look at the distribution of Ca and P throughout the sample. 4.1. Ca/P ratio and bone condition A new method has been developed at the University of Ioannina for the estimation of the in vivo Ca/P ratio in the human radius. This approach uses a modified dual energy X-ray absorptiometry system (Fountos et al., 1999). Using this instrument, a significant difference in the mean Ca/P ratio between postmenopausal osteoporotic patients and premenopausal controls has been observed (Fountos et al., 1999) and from studies in rabbits, a significant decrease in the tibia Ca/P ratio in animals with inflammation-mediated osteoporosis has also been found (Fountos et al., 1998). The same authors studying the relationships between the skeletal Ca/P ratio and the structural changes in the bone in response to osteoporosis, suggest that alterations in the Ca/P ratio may be indicative of underlying changes in the organic matrix of bone, mainly in collagen. Other studies (Kounadi et al., 1998) demonstrate that in osteoporotic rabbits (induced by inflammation) severe structural abnormalities in the rabbit tibia collagen occur. These abnormalities consist of a significant alteration in mean fibril diameter as well as marked disorganization in the packing of the fibrils. Our results at ELETTRA confirm that a statistically significant decrease is observed in rabbit tibia Ca/P ratio for osteoporotic samples suggesting that there is a relationship between bone loss and a lowered Ca/P ratio (see Table 1). 4.2. Ca/P ratio and bone type
of the Ca/P contour plots for different bone types, sites and bone conditions taken from the 28 keV images. Contours have been plotted at Ca/P ratios of 60%, 80%, 85%, 90% and 97% of the maximum values in the image.
4. Discussion The experiments described in this paper, using the microtomography system on the SYRMEP beamline at the ELETTRA synchrotron light laboratory, were designed to confirm and extend previous results on
The variations with different animals and different sites detected in this work indicate the importance of selecting suitable animal models when applying results to human studies. Other workers performing measurements on intact bones by neutron activation analysis have shown that the Ca/P ratio depends upon the kind of bone studied (Tzaphlidou and Zaichick, 2002, 2003; Zaichick and Tzaphlidou, 2002, 2003). The different values obtained in the different intact bone sites are due to the influence of the organic matrix, i.e. fat, lipids and marrow (particularly that of red marrow) which depend on bone kind (Bolotin and Sievanen, 2001; Fountos et al., 1997; Tzaphlidou and Zaichick, 2002).
Fig. 2. Contour plots, derived from the microCT images (600 600 pixels), of the Ca/P ratio for normal and osteoporotic rabbit tibia samples. Contours at 60%, 80%, 85%, 90% and 97% of maximum Ca/P within the sample have been plotted in each sample. It can be seen that the contour density is increased in osteoporotic samples and can also demonstrate regions of non-uniformity. The pixel size is approximately 12 mm.
ARTICLE IN PRESS M. Tzaphlidou et al. / Applied Radiation and Isotopes 62 (2005) 569–575
573
ARTICLE IN PRESS 574
M. Tzaphlidou et al. / Applied Radiation and Isotopes 62 (2005) 569–575
4.3. Ca/P spatial distribution and bone condition The spatial distribution of the Ca/P ratio can be seen in Fig. 2. It appears that normal bone shows less variation in the distribution of Ca/P as demonstrated by the near absence of contours throughout the bone volume. However, the osteoporotic examples show a higher density of Ca/P contours in general, but also display regions of significantly higher Ca/P gradient changes. These changes are not fully understood but could be due to local exchanges of Ca and P. A further interesting observation is the variation in the contour density within a single sample. It can be seen in Fig. 2 that a sample which is judged to be ‘osteoporotic’ from the bulk estimate of the Ca/P ratio may show regions of ‘osteoporotic’ behaviour and regions of ‘normal’ behaviour. Further work is required to confirm and understand these observations and their full interpretation is still ongoing. In conclusion, synchrotron radiation microtomography is potentially an outstanding in vitro technique to map different bone chemical components and to detect changes brought about by bone diseases. As has been reported (Tzaphlidou and Zaichick, 2002, 2003; Zaichick and Tzaphlidou, 2002), the Ca/P ratio may provide greater reliability for diagnosis of bone disorders and hence microtomography may become a valuable technique to be used during bone therapeutic and diagnostic trials.
Acknowledgements This work was undertaken at the Elettra synchrotron radiation source and was funded by the Human Potential Programme—Transnational Access to Major Research Infrastructures of the European Community.
References Barbier, A., Martel, C., de Vernejoul, M.C., Tirode, F., Nys, M., Mocaer, G., Morieux, C., Murakami, H., Lacheretz, F., 1999. The visualization and evaluation of bone architecture in the rat using three-dimensional X-ray microcomputed tomography. J. Bone Mineral. Metab. 17, 37–44. Bauss, F., Minne, H.W., Sterz, H., Weng, U., Wesch, H., Ziegler, R., 1985. Comparative bone analysis via inflammation-mediated osteopenia (IMO) in the rat. Calcif. Tissue Int. 37, 539–546. Bolotin, H.H., Sievanen, H., 2001. Inaccuracies inherent in dual-energy X-ray absorptiometry in vivo bone mineral density can seriously mislead diagnostic/prognostic interpretations of patient-specific bone fragility. J. Bone Mineral Res. 16, 799–805. Davis, G.R., Wong, F.S., 1996. X-ray microtomography of bones and teeth. Physiol. Meas. 17, 121–146.
Fountos, G., Yasumura, S., Glaros, D., 1997. The skeletal calcium/phosphorus ratio: a new in vivo method of determination. Med. Phys. 24, 1303–1310. Fountos, G., Kounadi, E., Tzaphlidou, M., Yasumura, S., Glaros, D., 1998. The effect of inflammation-mediated osteoporosis (IMO) on the skeletal Ca/P ratio and on the structure of rabbit bone and skin collagen. Appl. Radiat. Isot. 49, 657–679. Fountos, G., Tzaphlidou, M., Kounadi, E., Glaros, D., 1999. In vivo measurement of radius calcium/phosphorus ratio by X-ray absorptiometry. Appl. Radiat. Isot. 51, 273–278. Kinney, J.H., Haupt, D.L., Balooch, M., Ladd, A.J., Lane, N.E., 2000. Three-dimensional morphometry of the L6 vertebra in the ovariectomized rat model of osteoporosis: biomechanical implications. J. Bone Miner. Res. 15, 1981–1991. Kounadi, E., Fountos, G., Tzaphlidou, M., 1998. The influence of inflammation-mediated osteopenia (IMO) on the structure of rabbit bone and skin collagen fibrils. Connect. Tissue Res. 37, 69–76. Laib, A., Barou, O., Vico, L., Lafage-Proust, M.H., Alexandre, C., Ruegsegger, P., 2000. 3D micro-computed tomography of trabecular and cortical bone architecture with application to a rat model of immobilisation osteoporosis. Med. Biol. Eng. Comput. 38, 326–332. Martin-Badosa, E., Elmoutaouakkil, A., Nuzzo, S., Amblard, D., Vico, L., Peyrin, F., 2003. A method for the automatic characterization of bone architecture in 3D mice microtomographic images. Comp. Med. Imag. Graphics 27, 447–458. Minne, H.W., Pfeilschifter, J., Scharla, S., Mutschelknauss, S., Schwarz, A., Krempien, B., Ziegler, R., 1984. Inflammationmediated osteopenia in the rat: a new animal model for pathological loss of bone mass. Endocrinology 115, 50–54. Nuzzo, S., Peyrin, F., Cloetens, P., Baruchel, J., Boivin, G., 2002a. Quantification of the degree of mineralization of bone in three dimensions using synchrotron radiation microtomography. Med. Phys. 29, 2672–2681. Nuzzo, S., Lafage-Proust, M.H., Martin-Badosa, E., Boivin, G., Thomas, T., Alexandre, C., Peyrin, F., 2002b. Synchrotron radiation microtomography allows the analysis threedimensional microarchitecture and degree of mineralization of human iliac crest biopsy specimens: effects etidronate treatment. J. Bone Mineral Res. 17, 1372–1382. Peyrin, F., Salome, M., Nuzzo, S., Cloetens, P., Laval-Jeantet, A.M., Baruchel, J., 2000. Perspectives in three-dimensional analysis of bone samples using synchrotron radiation microtomography. Cell Mol. Biol. 46, 1089–1102. Postnov, A.A., Vinogradov, A.V., Van Dyck, D., Saveliev, S.V., De Clerck, N.M., 2003. Quantitative analysis of bone mineral content by X-ray microtomography. Physiol. Meas. 24, 165–178. Salome, M., Peyrin, F., Cloetens, P., Odet, C., Laval-Jeantet, A.M., Baruchel, J., Spanne, P., 1999. A synchrotron radiation microtomography system for the analysis of trabecular bone samples. Med. Phys. 26, 2194–2204. Stenstrom, M., Olander, B., Carlsson, C.A., Carlsson, G.A., Lehto-Axtelius, D., Hakanson, R., 1998. The use of computed microtomography to monitor morphological changes in small animals. Appl. Radiat. Isot. 49, 565–570.
ARTICLE IN PRESS M. Tzaphlidou et al. / Applied Radiation and Isotopes 62 (2005) 569–575 Stenstrom, M., Olander, B., Lehto-Axtelius, D., Madsen, J.E., Nordsletten, L., Carlsson, G.A., 2000. Bone mineral density and bone structure parameters as predictors of bone strength: an analysis using computerized microtomography and gastrectomy-induced osteopenia in the rat. J. Biomech. 33, 289–297. Stein, I.D., Granik, G., 1976. Rib structure and bending strength: an autopsy study. Calcif. Tissue Res. 20, 61–73. Tzaphlidou, M., Zaichick, V., 2002. Neutron activation analysis of calcium/phosphorus ratio in rib bone of healthy humans. Appl. Radiat. Isot. 57, 779–783.
575
Tzaphlidou, M., Zaichick, V., 2003. Calcium, phosphorus, calcium-phosphorus ratio in rib bone of healthy humans. Biol. Trace Element Res. 93, 63–74. Zaichick, V., Tzaphlidou, M., 2002. Determination of calcium, phosphorus, and the calcium/phosphorus ratio in cortical bone from the human femoral neck by neutron activation analysis. Appl. Radiat. Isot. 56, 781–786. Zaichick, V., Tzaphlidou, M., 2003. Calcium and phosphorus concentrations and the calcium/phosphorus ratio in trabecular bone from the femoral neck of healthy humans as determined by neutron activation analysis. Appl. Radiat. Isot. 58, 623–627.