Bone, Vol. 15, No. 2, pp. M-202, 1994 Couvrieht 8 1994 Elsevier Science Ltd P&teain the USA. All rights reserved 8756-3282/94 $6.00 + .OO

Pergamon 8756-3282(93)EOOO7-8

DXA for Bone Density Measurement in Small Rats Weighing 150-250 Grams P. W. LU,’ J. N. BRIODY,2

R. HOWMAN-GILES,’

A. TRUBE3 and C. T. COWELL’

’ Robert Vines Growth Research Centre and 2 Department of Nuclear Medicine, Royal Alexandra Hospital for Children, Sydney, Australia 3 Department of Animal Science, The University of Sydney, Sydney, Australia Address for corresoondence and reorints: Sydney iO50, Australia.

Dr. Pei Wen Lu. Robert Vines Growth ResearchCentre, Royal Alexandra Hospital for Children,

Abstract

DXA employs the dual-energy projection technique and provides information on three body compartments: bone, lean tissue, and fat (Mazess 1990; Southard 1991). Compared with the abovementioned techniques on humans, DXA involves a lower dose of radiation and has more reliable performance (Mazess 1990). Chan et al. recently reported a high correlation between bone mineral content measured by DXA and ash weight in small chicken parts (Chan 1992), suggesting that it may be a useful technique for animal studies to assess total body bone mineral status. We have evaluated the use of DXA in the measurement of the total body and regional mineral content in live, small laboratory animals.

The present study evaluated the use of a small auhual total body software of dual energy x-ray absorptiometry (DXA) in the assessment of total body and regional bone mineral content (BMC) and bone mineral density (BMD) in small rats. Twenty-three rats, with weights ranging from 146 to 246 g, were included in the study. All were scanned using the same software version aud same scan procedure (speed and scanned area). Total body BMD, BMC, and body weight were measured by DXA in each rat. Femoral BMC and BMD were analyzed by ding regional analysis facilities. The repeatability (precision) of thii software version was assessed prior to the study and the coeffkients of variation (CV) were 2.9% for total body BMC, 0.8% for total body BMD, 1.2% for body weight, and 2.2% for mean femoral BMD. DXA measurements were compared with the measurements obtained by using established standards, namely weight and bone ash content. Total body ash content and femoral ash content were measured separately in all rats. There was a strong linear correlation between BMC and ash content in total body (4 = 0.98, p = 0.0001) and in femur (? = 0.94, P = 0.0001). There was also an excellent linear association between body weight measured by DXA and scale weight (? = 0.99, P = 0.0001). We conclude that thii software version is suitable for study on small animals and is a useful tool for assessment of regional as well as total body bone mineral status. Key Words: DXA-l3one

density-Rats-Ash

Materials and Methods DXA measurement

Measurements of total body bone mineral content (BMC), total body bone mineral density (BMD), total body weight of rats were obtained with a total body scanner using a specifically designed software version for small animals (DPX, Small Animal Software Version 1.0, Lunar Radiation Corp., Madison, WI). This scanner uses a constant potential X-ray source at 76 kVp and a K-edge filter (cerium) to achieve a congruent beam of stable dual-energy radiation. The effective energies are 38 and 70 keV. Throughout the study period, daily quality assurance tests were performed to ensure the effectiveness of the lights, beam, mechanics, and tissue value of the scanner. The collimation (size of x-ray beam at the source) used was 0.84 mm. The size of each sample point during the scan was 0.6 X 1.2 mm. With the DXA scanner, a series of transverse scans are made from the tip of nose to the midpoint of the tail of the rat at 1.2~mm intervals with a set scan area of 8 X 25 cm for each rat. Scans were performed in a transverse scan speed of 9.6 mm per second, giving a scan time of 30 min for the total body determination. No tissue equivalent material was used. All rats were scanned using the identical scan procedure (scan speed and scan area). BMD and BMC were also assessed on both femora using the regional analysis facility of DXA (Fig. 1). Boxes were drawn to outline the area of interest (femora). All scan files were analyzed (total body and regional) by the same technician (J.B.) to minimize the operation errors, particularly in

weight.

Introduction Methods that are available to evaluate whole body bone mineral content in vivo in laboratory rats include total body neutron activation analysis of calcium (Yasumura 1976) and radionuclide uptake studies (Lindgren 1976). However, the research application of these techniqu’es is limited because of several disadvantages including high radiation dose, high cost, and lack of regional information. In recent years dual energy x-ray absorptiometry (DXA) has been employed widely in human studies. 199

200

P. W. Lu et al.: DXA for bone density measurements

in small rats

Analysis Statistical analysis was performed using SAS (System for Elementary Statistical Analysis) software (Schlotzhauer & Lettell 1987). Results are given as means with 95% confidence intervals (CI). Associations between variables were assessed with linear regression and correlation.

Results

Fig. 1. Image obtained from DXA software version. Regions of interest (femora) are identified by boxes.

the regional analysis. Mean values of the two femora were used in the analysis. Animals Twenty-three female Sprague-Dawley rats at 6-12 weeks of age were used in the validation process. The body weight of rats ranged from 146.0 to 246.5 g. Rats were anesthetized with pentobarbitone (Nembutal40 mgikg, intraperitoneal injections) and placed on a piece of plastic sheet (0.2 mm thick, without measureable tissue or mineral equivalent materials) in a prone position for the scan. Rats remained immobilized during the 3WO min scan time. Only about half the tail of each rat was scanned (Fig. 1) because of the time limit. The study was approved by the Animal Ethics Committee, Royal Alexandra Hospital for Children.

The precision of DXA measurement was assessed by scanning a rat ten times with repositioning. The CV was 2.9% for total body BMC, 0.8% for total body BMD, and 1.2% for body weight. Precision of regional analysis was also assessed. The CV was 2.2% for mean femur BMD. The body weight ranged from 146.0 to 252.9 g (mean 193.1 g) . Significant correlations were found between live body weight and ash weight of total body and femur, total body BMC and BMD, and femur BMC and BMD (data not shown). There was a highly significant linear association between total body BMC measured by DXA and ash weight of the whole skeletons (Fig. 2). Details of equations are listed in Table I. The relationships between DXA measurements (BMC, weight) and established standards (ash content and scale weight) were further studied. In Figure 2, total body BMC measured by DXA was plotted against ash content, with the dotted line demonstrating unity (see Table I for equations). In the BMC measurement, the DXA method gave a lower reading as compared with ash content (mean 0.53 g, 95% CI: 0.41-0.65 g). As seen in Figure 2 the difference between BMC and ash content was size-dependent, and it decreased with increasing weight @ = 0.0003). The predictive model for BMC based on scale weight is: BMC = 4.89 - 0.05 X weight + 0.0002 X weight’(? = 0.94). In the body weight measurement, DXA method gave a higher reading than scale weight (Fig. 3; mean difference 6.8 g, 95% CI: 5.6-8.0 g). This discrepancy was not size-dependent @ = 0.11).

Ash weigh After DXA scan, rats were sacrificed using CO,. The whole skeleton of each rat as thoroughly dissected (excluding the distal half of the tail). The skeleton was then dried at 120°C for 24 h before cremation in a muffle furnace at 600°C for 24 h. During this process, all organic materials were burnt off and only bone minerals were left in the ash. Ash weight includes calcium, phosphate, magnesium, and other minerals and is commonly used as a synonym for bone mineral content. In this study, the skeletons of all rats were cremated in two parts; i.e., both femora and the remainder. Ash weight of the whole skeleton was obtained by summing the two parts. Mean femoral ash weight was calculated as ash weight of both femora divided by 2. Precision One dead rat was scanned ten times with repositioning between scans on the same day for assessment of precision. Precision of the software version on total body and regional measurements were both assessed. The coefficient of variation (CV) was expressed as the percentage of standard deviation (SD) of the mean.

0

2

4

Ash weight

6

8

(g)

Fig. 2. Association between total body BMC measured by DXA and total body ash content with regression line shown (solid line). Excellent linearity (ra = 0.98) is shown between these two measurements. However, in smaller rats, DXA gives a lower reading on total body BMC compared with ash content. This under estimation improves with increase in body size (close to identity-dotted line).

P. W. Lu et al.: DXA for bone density measurements

201

in small rats

Table I. Associationof BMC and weight measurementsbetween DXA and established methods--regression equations Equations

P

P

0.98

O.oool

0.99

0.0001

0.94

O.oool

0.3

Total body BMC = -1.32 + 1.19 X ash weight Total body weight = 35.2 + 0.90 x weight Mean femur BMC = -0.02 + 1.04 X femur ash weight

-

0.2

-

0.1

-

b

$” I’

/’ 4’ S(i’

$’ In the regional analysis, there was a highly significant association between the femur BMC and femur ash weight (Fig. 4; rz = 0.94, p = 0.0001).

oaf . 0.0

Discussion

0.1

Femur

This study has demonstrated that using a small animal software version of DXA, weight, BMC, and BMD can be readily measured in small animals. The biological associations such as the weight-related increase in BMD and BMC were also well demonstrated. A previous study (Ammann et al. 1992) demonstrated the usefulness of DXA for assessing changes in bone mass in different regions in live rats as small as 220 g. We have extended the low weight limit to 146 g such that this software may be useful to assess BMC and BMD before puberty in rats or in genetically small animals such as the growth hormone-deficient dwarf rats (Charlton .L988). Strong relationships were found between biological variables (live weight and ash weight) and variables measured by DXA including BMC, BMD, and weight. In this study, DXA software gave a consistently higher reading in body weight measurement compared with scale weight. It was also observed that this DXA software version gave a lower reading on bone mineral content compared with ash weight and this underestimation improves with increase in body size. DXA employs the dual energy pro-

250 200 150 100 50 L’ 50

100

150

Weight Fig. 3. Excellent correlation

200

-

250

scale

300

0.2

0.3

ash content

0.4

(g)

Fig. 4. Significant linear relationship between femur BMC measured by DXA and femur ash weight (? = 0.94).

jection technique which relies on the differential absorption (or attenuation) of photons at two energy levels within bone and soft tissue. Accordingly, body weight is divided into two compo-

nents: bone and soft tissue (Heymsfield 1989). The two energies of DXA were originally designed for human adults and the subsequent applications have been achieved by optimizing the software component. When used in humans, the influence of tissue thickness on results of BMC and BMD is accounted for in the algorithms so results are independent of thickness variation (Mazess et al. 1990). When used in small animals the weight independency may not hold. As rat weight decreases, the soft tissue thickness decreases and the bones get smaller and less dense. This causes difficulty in discriminating between low-density bones (seen in smaller animals) and soft tissue and thus results in underestimation of BMC in smaller rats (body weight < 200 g). Nevertheless, this software version provides excellent precision and high correlation to ash content. We believe that the influence of size on the accuracy of BMC measurement could be adjusted by using the predictive model (see results) when using animals lighter than 200 g. Regional analysis by DXA is useful in human studies to examine areas of potential pathology (spine, hip) and for determining BMC and BMD in areas that are predominantly made up of trabecular or cortical bone. With this software, we have found an excellent linear relationship between femur BMC and femur ash weight. We conclude from this study that this DXA software version is suitable for assessment of total body and regional bone mineral status in small, live laboratory animals. Its attributes of small radiation dose, good precision and accuracy makes research in small laboratory animals possible, though the size dependent variation in bone mineral content measurement needs to be considered.

350

(g)

between weight measured by DXA and scale weight (3 = 0.99). In general, DXA gives a higher reading compared with scale measurement (average 6.8 g higher). This overestimation is size-independent (dotted line illustrates identity).

The authors are grateful for technical assistance horn Mr. Joe Bisek and Mr. Robert Washenko of Lunar Radiation Corp.,

Acknowledgnenrs:

Madison, WI. This study was funded by the National Health & Medical Research Council (Australia) and Kabi-Pharmacia (Sweden).

202 References: Ammann, P.; Rizzoli, R.; Slosman, D.; Bonjour, 1. P. Sequential and precise in viva measurement of bone mineral density in rats using dual-energy x-ray absorption&y. J. Bone Miner. Res. 7(3):31 l-316; 1992. Ghan, G. M. Performance of dual energy X-ray absorptiometry in evaluation bone, lean body mass, and fat in pediatric subjects. J. Bone Miner. Res. 7:369-374; 1992. Charlton, H. M.; Clark, R. G.; Robinson, I. C. A. F.; Porter Goff, A. E.; Cox, B. S.; Bloch, B. A. Growth hormone-deficient dwarf!sm in the rat: a new mutation. J. Emiocrinoi. 1195-58; 1988. Heymstield, S. B.; Wang, J.; Heshka, S.; Kehayias, 1. I.; Pierson, R. N. Dualphoton absorptiometry: comparison of bone mineral and soft tissue mass measurements in viva with established methods. Am. J. Clin. Nurr. 49:128>1289; 1989. Lindgren, 1. U. Studies of calcium accretion rate of bone during immobilization in intact and thyroparathyroidectomized adult rats. C&if. Tissue Res. 22:41-47; 1976.

P. W. Lu et al.: DXA

for bone density measurements in small rats

Mazess, R. B.; Barden, H. S.; Bisek, J. P.; Hanson, J. Dual-energy X-ray absorptiometry for total body and regional hone mineral and soft tissue composition. Am. J. Clin. Nutr.51:110&1112; 1990. Mazess, R. B.; Howard, S. B.; Bisek, J. P.; Hanson, J. Dual energy X-ray absorptiometry for total-body and regional bone mineral and soft tissue composition. Am. J. Clin. NW. 51:1106-1112; 1990. Schlotzhauer, S. D.; Lettell, R. C. SAS systemfor elementary statistical analysis. Cay, NC: SAS Institute; 1987. Southard, R. N.; Morris, J. D.; Mahan, J. D.; Hayes, J. R.; Torch, M. A.; Sommer, A.; Zipf, W. B. Bone mass in healthy children: Measurement with quantitative DXA. Radiology 179~735738; 1991. Yasumura, S.; Elk, K. J.; Fairchild, E.; Brook, D.; Cohn, S. H. Effect of graded doses of cortisol on total body calcium in rats. Am. J. Physiol. 231:176&1763; 1976.

Dare Received: July 16, 1993 Dare Revised: October 14, 1993 Dare Accepred: October 15, 1993