Calcif Tissue Int (1992) 51:127-131
Calcified Tissue International 9 1992 Springer-Verlag New York Inc.
Laboratory Investigations Incorporation of Sodium Fluoride into Cortical Bone Does Not Impair the Mechanical Properties of the Appendicular Skeleton in Rats Thomas A. Einhorn, .1 G. K. Wakley, *.2 S. Linkhart, +3 E. B. Rush, .1 S. Maloney, *l E. Faierman, *l and D. J. Baylink +3 ~Department of Orthopaedics, Box 1188, Mount Sinai School of Medicine, New York, New York 10029-6574; ZDepartment of Orthopaedics, Medical Sciences Building, Mayo Clinic, Rochester, Minnesota; and 3Department of Medicine, Loma Linda University and Jerry L. Pettis Memorial Veterans Hospital, Loma Linda, California, USA Received August 14, 1991
Summary. Clinical studies on the use of sodium fluoride (NaF) in osteoporotic patients have demonstrated increased spinal bone mass without a reduction in vertebral fracture incidence, and a trend towards reduced appendicular bone mass with an increase in peripheral fracture incidence. As previous reports have suggested that NaF becomes incorporated into bone's crystal structure, possibly affecting bone strength, we sought to examine the relationship among bone fluoride content, bone mass, and skeletal fragility. Twentyone-day-old female Sprague-Dawley rats were treated with four different doses of NaF. The tibiae were subjected to histomorphometric and biochemical analyses, and the femora were tested in torsion for the properties of strength, stiffness, energy storage capacity, and angular deformation. The results showed that over 50% of the skeleton in these rats was turned over in the presence of NaF. The four different doses resulted in a linear increase in bone F concentration and suggested excellent absorption and incorporation of this drug. No changes in histomorphometric indices of bone formation or turnover were found. Despite the large fraction of bone formed during N a F treatment, and the linear increase in bone fluoride content in relation to dose, there were no changes observed in any of the mechanical properties. These results suggest that, even extensive incorporation of fluoride into bone, in the absence of an effect on bone mass or remodeling, does not significantly alter its capacity to withstand mechanical loads.
Key words: Sodium fluoride - Cortical bone - Mechanical properties - Histomorphometry - Bone mass.
Sodium fluoride (NaF) has been repeatedly shown to stimulate bone formation and increase cancellous bone mass in osteoporotic patients [1-4]. However, concern has arisen over the antifracture efficacy of N a F as well-controlled clinical studies have suggested that N a F therapy increases spinal bone mineral density but does not decrease the rate of
Offprint requests to: T. A. Einhorn
vertebral fractures [5, 6]. Moreover, in one study, bone density showed a nonstatistically significant trend to decrease in the appendicular (peripheral) skeleton which was associated with an increase in the incidence of fractures of the radius. [5]. Previous investigations had indicated that NaF treatment increased material bone density (calcium concentration per unit weight bone), microhardness, and crystal size in rats and humans [7-9]. Other reports suggested both positive and negative effects of this drug on the mechanical properties of bone [10, 11]. In order to determine the relationship among bone fluoride content, bone mass, and appendicular skeletal fragility, we developed a rat model in which a large portion of the skeleton is formed in the presence of NaF, and tested its biochemical, histomorphometric, and mechanical properties.
Materials and Methods Sixty 21-day-old female Sprague Dawley rats were divided into four groups. All animals were fed normal rat chow composed of 1.0% calcium, 0.67% phosphorus, and containing 25 parts per million (ppm) NaF (Purina Rat Chow, diet #5010, Purina Mills, Richmond, IN). They were offered distilled drinking water ad libitum. The four groups were distinguished by the concentrations of NaF contained in the drinking water. The concentrations used were 0 (control), 25, 50 and 75 ppm NaF resulting in estimated daily fluoride intakes (from the water) of 0, 0.5, 1.0, and 1.5 g/day, respectively. Animals were treated for 86 days in order that they turn-over a large portion of their skeletons during the course of this study. On the first day of the experiment, all animals were treated with an intraperitoneal (i.p.) injection of 10 mg tetracycline HC1 (Lederle Laboratories, Pearl River, NY). Sodium fluoride treatment began 2 days later. A second fluorochrome, demeclomycin 10 mg (Lederle), was injected i.p. on the 84th day. The animals were euthanized on the 86th day. At the time of euthanasia, both femora and tibiae were harvested. The left femora were frozen at -20~ For determining calcium (Ca) and F content, the femora were thawed, incubated in petroleum/ethyl ether (1:1), three changes per 48 hours each, and then dried at 100~ for 24 hours. A piece of this dry, defatted bone was weighed, and 6 N HC1 was added to each piece (20 ixl/mgbone). F concentration was determined using a fluoride-specific electrode (Orion Research, Inc., Boston, MA) and Ca content was measured using a Calcette Analyzer (Precision Systems, Inc., Natick, MA).
T. A. Einhorn et al.: NaF Does Not Impair Mechanical Properties of Cortical Bone
128
Table 1. Mechanical properties of right femora Stiffness (NM/~
Torsional strength (NM)
Energy absorption (NM x o)
Group
No.
Angular deformation (~
I (Control) II (25 ppm NaF) III (50 ppm NaF) IV (75 ppm NaF)
14
21.94 • 1.37
0.026 • 0.007
0.453 • 0.033
5.11 • 0.33
15
18.61 • 1.32
0.023 • 0.002
0.363 -+ 0.035
4.36 • 0.65
16
20.59 • 0.81
0.029 • 0.002
0.480 • 0.029
5.43 • 0.37
11
20.16 • 1.09
0.028 • 0.002
0.402 • 0.022
4.48 +- 0.39
Values mean --- SEM; NM = Newton-meter Note: None of these changes are significant
Table 2. Static tibial cancellous bone parameters (secondary spongiosum and proximal tibial metaphysis)
Group I (Control) II (25 ppm NaF) III (50 ppm NaF) IV
Bone volume (%)
Bone perimeter (mm -1)
Osteoclast no. (ram -2)
Osteoclast perimeter (ram -1 x 10 -3)
Osteoblast perimeter (mm -1 x 10 -3)
14.85 • 1.94 100.00 --- 13.03%
2.54 -+ 0.25 100.00 - 9.79%
0.39 - 0.08 100.00 --- 19.15%
9.00 -+ 1.79 100.00 --- 19.90%
33.81 -+ 6.89 100.00 -+ 20.39%
15.30 +- 1.19 102.87 • 7.99%
2.18 • 0.13 85.97 • 5.09%
0.61 -+ 0.10 153.74 - 25.05%
21.95 - 10.68 221.78 --- 107.96%
15.30 • 5.28 52.28 +- 20.11%
18.28 -+ 0.63 123.10 • 4.26%
2.56 • 0.10 100.87 • 3.82%
0.63 • 0.07 158.66 • 18.88%
12.00 • 201.64 •
16.74 • 1.32 112.71 --- 8.89%
2.53 • 0.13 99.48 • 5.07%
0.48 • 0.14 121.62 • 35.26%
1.85 18.65%
62.67 - 13.27 238.71 • 50.53%
9.00 --- 2.58 90.97 -+ 26.11%
45.83 • 13.29 174.59 • 50.63%
(75 ppm NaF) Mean + SEM BV = bone volume; TV = tissue volume (referent); T.Ar = tissue area (referent); Ob = osteoblast; Oc = osteoclast; Pm = perimeter Note: perimeter is a two-dimensional method of expressing surface
The right femora were subjected to mechanical testing and the left and right tibiae were fixed in formalin and prepared for histomorphometry.
Biomechanical Testing The femora were thawed, dissected free of their soft tissues, and the ends were potted in Cerrobend (Jackson-Walter, King of Prussia, PA). Changes in length and diaphyseal diameter were determined with a micrometer. The potted bones were mounted in the grips of a Servo-actuated (DC motor) torsion testing apparatus and loaded to failure. This apparatus uses a loading rate of 1 rps (360~ and is a motorized modification of the Burstein-Frankel system [12]. The data were digitized, displayed, and stored in an IBM PC/XT computer. Values were obtained for torsional strength, stiffness, angle of deformation, and energy storage capacity.
Histomorphometry The tibiae were fixed in neutral buffered 10% formal saline for a minimum of 72 hours and sectioned transversely in the middiaphysis at the tibio-fibular synostosis. The resulting cross-sections were ground on roughened glass to approximately 25 ~m thickness. Dynamic histomorphometric measurements were made on unstained sections using an Olympus BH-2 microscope (Tokyo, Japan) with UV objectives. A color video camera was mounted on the microscope and linked via a Microcomp digitizing bit-pad (SMI Corporation, GA) to a color monitor and Apple IIe microcomputer using Microcomp software (SMI).
Because rat bone differs from human bone in terms of its form and tissue distribution, the terminology approved by the Histomorphometry Nomenclature Committee of the American Society for Bone and Mineral Research [13] was modified and the following parameters and calculations were used: In diaphyseal cortical bone, medullary area was defined as the area delineated by the endocortical surface of the cross section; cross-sectional area was measured as the area of bone and marrow cavity within the periosteal surface of the specimen; cortical bone area was calculated as the difference between cross-sectional and medullary areas. The lengths of bone perimeter at the periosteal surface and endocortical surface were measured. Further, at the endocortical surface, the lengths of labeled perimeter and nonlabeled perimeter were also measured. The bone area between adjacent fluorochrome labels was measured at the periosteal and endocortical surfaces, respectively, and the group mean bone area between adjacent fluorochrome labels was calculated. This number was divided by the interlabel interval in days to obtain the average bone formation rates. The average formation rates were divided by the periosteal and endocortical label perimeters to obtain the mean periosteal and endocortical matrix apposition rates (MAR). Following removal of the cross sections, the proximal tibial metaphyses were decalcified in 5% formic acid in 10% neutral buffered formalin for 7 days. The bones were bissected in the midsagittal plane, and dehydrated in ascending grades of alcohol before infiltrating and embedding in glycol methacrylate (JB-4 plastic, Polysciences Inc., Warrington, PA). The resulting blocks were sectioned at 5 p,m thickness and the sections were stained for acid phosphatase by an azo-dye coupling technique and counterstained with methyl green thionine. Cancellous bone was examined in the secondary spongiosa of the proximal tibial metaphysis at a standard sampling site no less than 1
T. A. Einhorn et al.: NaF Does Not Impair Mechanical Properties of Cortical Bone
129
Table 3. Static tibial cortical bone parameters
Group I (Control) II (25 ppm NaF) III (50 ppm NaF) IV (75 ppm NaF)
Group I (Control) II (25 ppm NaF) III (50 ppm NaF) IV (75 ppm NaF)
Group I (Control) lI (25 ppm NaF) III (50 ppm NaF) IV (75 ppm NaF)
Crosssectional area (mm 2)
Medullary area (mm 2)
Cortical bone area (ramz)
Periosteal perimeter (mm)
Endocortical perimeter (mm)
3.23 -+ 0.064
0.69 -+ 0.01
2.54 -+- 0.06
6.53 -+ 0.07
3.01 -+ 0.03
3.40 -+ 0.08
0.76 P < 0.77 P
0.01 287.69 -+ 5.04 P > 0.001 369.12 -+ 7.56 P > 0.001
154.8 -+ 18.8 160.4 -+ 14.4 156.4 - 15.0
Values = mean -+ SEM
would indicate that if fluoride alters the relationship between bone strength and bone mass or density, it may do so only when bone formation or remodeling has been altered by fluoride administration. In the present study, none of the doses of NaF led to changes in bone histomorphometric measurements, and in particular, no increased bone formation was observed. One of the limitations in extrapolating these results to patients is the fact that F did not produce the increase in bone formation and changes in other bone histomorphometric measurements that would be expected based upon what is known of the effects of NaF in human bone. Although it is clear from the data that there was a direct linear relationship between the content of F in the drinking water and the incorporation of fluoride into bone, the ingestion of these amounts of N a F did not produce a change in bone mass. Two possible explanations for these findings are (1) NaF is less effective in stimulating new bone formation in very immature animals because their bones are already undergoing formation at maximal levels, and (2) the doses of F used were insufficient to stimulate bone formation in any animal. However, dose ranges of NaF between 25 and 75 ppm in drinking water, leading to an estimated daily NaF intake of between 0.5 and 1.0 g/rat, would seem to be sufficient to produce a formation-stimulating effect based on the results
of previously published works [7-11]. A recent report by Narita et al. [15] showed that doses of NaF similar to those used in the present experiment lead to a distribution of F in cortical bone whereby it is highest in the outer lamellae and lowest on the endosteal surface [15]. Moreover, these investigators found an excellent correlation between F intake and bone fluoride content. As the failure pattern of bone loaded in torsion is initiated in the outer cortex [16], these findings would suggest that the femora loaded in our experiments failed through sections of bone where the F concentration would have been significant. It should also be noted that only one method of mechanical testing, torsional loading, was used in this experiment. It is possible that other loading conditions such as compression, tension, or three- or four-point bending would yield different results. However, although vertebrae are most commonly loaded in compression, femoral and tibial fractures sustained in patients generally occur as a result of a combination of forces, the most significant of which is torsional loading [12, 16]. In order for a bone to be loaded in tension it would literally have to be pulled apart from either end. Moreover, for pure bending conditions to occur in nature, a long bone would have to be fixed at two points and have the direction of force occur at a perpendicular angle between the points of fixation. The value of torsional loading in mechanical testing is not only that it more closely approximates how fractures occur clinically, but it also fulfills other critical criteria essential to an experimental test. These include (1) the loading configuration subjects the bone to equally severe loading conditions at every section along its length so as to be able to identify weak sections; (2) the loading mode is not critically dependent upon bone geometry, in particular, bone length; and (3) the loading configuration allows for the tester to control for rates of application of loads, so that the time conditions produced in the test will be reproducible and preferably representative of those that occur during normal trauma [12]. Because the experimental conditions in this study did not produce a change in bone histology or bone mass, the results may not explain many of the current controversies surrounding the use of NaF in patients. However, the results clearly
T. A. Einhorn et al.: NaF Does Not Impair Mechanical Properties of Cortical Bone show that a marked increase in bone F content alone does not necessarily lead to an alteration in mechanical properties or an increase in the propensity to fracture under static load. This would suggest that the observed disparity between alterations in bone mass in patients treated with N a F and their lack of protection from spinal fractures may need to be explained on the basis of some effect of N a F on bone formation and remodeling as opposed to its mere incorporation into bone.
Acknowledgments. We are grateful to David Glass and Gita Bosch for technical assistance. This work was supported in part by NIH Grant 5 RO1 AR31062, the Orthopaedic Research and Education Foundation, and Veteran's Administration Merit Review Funds.
References 1. Riggs BL, Seeman E, Hodgson SF, Taves DR, O'Fallon WM (1982) Effect of the fluoride/calcium regimen on vertebral fracture occurrence in postmenopausal osteoporosis. N Engl J Med 306:446--450 2. Hansson T, Roos B (1978) Effect of combined therapy with sodium fluoride, calcium and vitamin D on the lumbar spine in osteoporosis. AJR 126:1294-1297 3. Jowsey J, Riggs BL, Kelly PJ, Hoffman DL (1972) Effect of combined therapy with sodium fluoride, vitamin D and calcium in osteoporosis. Am J Med 53:43-49 4. Charles P, Mosekilde L, Taagehoj Jensen F (1985) The effects of sodium fluoride, calcium phosphate and vitamin D2 for one to two years on calcium and phosphorous metabolism in postmenopausal women with spinal crush fracture osteoporosis. Bone 6:201-206
131
5. Riggs BL, Hodgson SF, O'Fallon WM, Chao EYS, Wahner HW, Muhs JM, Cedel SL, Melton LJ (1990) Effect of fluoride treatment on the fracture rate in postmenopausal women with osteoporosis. N Engl J Med 322:802-809 6. Kleerekoper M, Peterson E, Phillips E, Nelson D, Tilley B, Parfitt AM (1990) Continuous sodium fluoride therapy does not reduce vertebral fracture incidence in postmenopausal osteoporosis. J Bone Miner Res 4:5376 7. Schraer H, Posner AS, Schraer R, Zipkin I (1972) Effect of fluoride on bone minerals. Clin Orthop 86:260-286 8. Baud CA, Bang S (1973) Biophysical study of the bone tissue of fluoride-treated mice. J Dent Res 52:589 9. Franke J, Runge H, Grau P, Fengler F, Wanka C, Rempl H (1976) Physical properties of fluorosis bone. Acta Orthop Scand 47: 20-27 10. Carter DR, Beaupr6 GS (1990) Effects of fluoride treatment on bone strength. J Bone Miner Res 5:S177-S184 11. Riggins RS, Rucker RC, Chan MM, Zeman F, Beljan JR (1976) The effect of fluoride supplementation on the strength of osteopenic bone. Clin Orthop 114:352-357 12. Burstein AH, Frankel VH (1971) A standard test for laboratory animal bone. J Biomech 4:155-158 13. Parfitt AM, Drezner MK, Glorieux FH, Kanis JA, Malluche H, Meunier PJ, Ott SM, Recker RR (1987) Bone histomorphometry: standardization of nomenclature, symbols and units. J Bone Miner Res 2:595-610 14. Dunnet CW (1964) New tables for multiple comparisons with a control. Biometrics 20:482--491 15. Narita N, Kato K, Nakagaki H, Ohno N, Kameyama Y, Weatherell JA (1990) Distribution of fluoride content in rat's bone. Calcif Tissue Int 46:200-204 16. Hayes WC, Carter DR (1979) Biomechanics of bone. In: DJ Simmons, AS Kumin (eds) Skeletal research--an experimental approach. Academic Press, New York, pp 263-300
Calcified Tissue International 9 1992 Springer-Verlag New York Inc.
Laboratory Investigations Incorporation of Sodium Fluoride into Cortical Bone Does Not Impair the Mechanical Properties of the Appendicular Skeleton in Rats Thomas A. Einhorn, .1 G. K. Wakley, *.2 S. Linkhart, +3 E. B. Rush, .1 S. Maloney, *l E. Faierman, *l and D. J. Baylink +3 ~Department of Orthopaedics, Box 1188, Mount Sinai School of Medicine, New York, New York 10029-6574; ZDepartment of Orthopaedics, Medical Sciences Building, Mayo Clinic, Rochester, Minnesota; and 3Department of Medicine, Loma Linda University and Jerry L. Pettis Memorial Veterans Hospital, Loma Linda, California, USA Received August 14, 1991
Summary. Clinical studies on the use of sodium fluoride (NaF) in osteoporotic patients have demonstrated increased spinal bone mass without a reduction in vertebral fracture incidence, and a trend towards reduced appendicular bone mass with an increase in peripheral fracture incidence. As previous reports have suggested that NaF becomes incorporated into bone's crystal structure, possibly affecting bone strength, we sought to examine the relationship among bone fluoride content, bone mass, and skeletal fragility. Twentyone-day-old female Sprague-Dawley rats were treated with four different doses of NaF. The tibiae were subjected to histomorphometric and biochemical analyses, and the femora were tested in torsion for the properties of strength, stiffness, energy storage capacity, and angular deformation. The results showed that over 50% of the skeleton in these rats was turned over in the presence of NaF. The four different doses resulted in a linear increase in bone F concentration and suggested excellent absorption and incorporation of this drug. No changes in histomorphometric indices of bone formation or turnover were found. Despite the large fraction of bone formed during N a F treatment, and the linear increase in bone fluoride content in relation to dose, there were no changes observed in any of the mechanical properties. These results suggest that, even extensive incorporation of fluoride into bone, in the absence of an effect on bone mass or remodeling, does not significantly alter its capacity to withstand mechanical loads.
Key words: Sodium fluoride - Cortical bone - Mechanical properties - Histomorphometry - Bone mass.
Sodium fluoride (NaF) has been repeatedly shown to stimulate bone formation and increase cancellous bone mass in osteoporotic patients [1-4]. However, concern has arisen over the antifracture efficacy of N a F as well-controlled clinical studies have suggested that N a F therapy increases spinal bone mineral density but does not decrease the rate of
Offprint requests to: T. A. Einhorn
vertebral fractures [5, 6]. Moreover, in one study, bone density showed a nonstatistically significant trend to decrease in the appendicular (peripheral) skeleton which was associated with an increase in the incidence of fractures of the radius. [5]. Previous investigations had indicated that NaF treatment increased material bone density (calcium concentration per unit weight bone), microhardness, and crystal size in rats and humans [7-9]. Other reports suggested both positive and negative effects of this drug on the mechanical properties of bone [10, 11]. In order to determine the relationship among bone fluoride content, bone mass, and appendicular skeletal fragility, we developed a rat model in which a large portion of the skeleton is formed in the presence of NaF, and tested its biochemical, histomorphometric, and mechanical properties.
Materials and Methods Sixty 21-day-old female Sprague Dawley rats were divided into four groups. All animals were fed normal rat chow composed of 1.0% calcium, 0.67% phosphorus, and containing 25 parts per million (ppm) NaF (Purina Rat Chow, diet #5010, Purina Mills, Richmond, IN). They were offered distilled drinking water ad libitum. The four groups were distinguished by the concentrations of NaF contained in the drinking water. The concentrations used were 0 (control), 25, 50 and 75 ppm NaF resulting in estimated daily fluoride intakes (from the water) of 0, 0.5, 1.0, and 1.5 g/day, respectively. Animals were treated for 86 days in order that they turn-over a large portion of their skeletons during the course of this study. On the first day of the experiment, all animals were treated with an intraperitoneal (i.p.) injection of 10 mg tetracycline HC1 (Lederle Laboratories, Pearl River, NY). Sodium fluoride treatment began 2 days later. A second fluorochrome, demeclomycin 10 mg (Lederle), was injected i.p. on the 84th day. The animals were euthanized on the 86th day. At the time of euthanasia, both femora and tibiae were harvested. The left femora were frozen at -20~ For determining calcium (Ca) and F content, the femora were thawed, incubated in petroleum/ethyl ether (1:1), three changes per 48 hours each, and then dried at 100~ for 24 hours. A piece of this dry, defatted bone was weighed, and 6 N HC1 was added to each piece (20 ixl/mgbone). F concentration was determined using a fluoride-specific electrode (Orion Research, Inc., Boston, MA) and Ca content was measured using a Calcette Analyzer (Precision Systems, Inc., Natick, MA).
T. A. Einhorn et al.: NaF Does Not Impair Mechanical Properties of Cortical Bone
128
Table 1. Mechanical properties of right femora Stiffness (NM/~
Torsional strength (NM)
Energy absorption (NM x o)
Group
No.
Angular deformation (~
I (Control) II (25 ppm NaF) III (50 ppm NaF) IV (75 ppm NaF)
14
21.94 • 1.37
0.026 • 0.007
0.453 • 0.033
5.11 • 0.33
15
18.61 • 1.32
0.023 • 0.002
0.363 -+ 0.035
4.36 • 0.65
16
20.59 • 0.81
0.029 • 0.002
0.480 • 0.029
5.43 • 0.37
11
20.16 • 1.09
0.028 • 0.002
0.402 • 0.022
4.48 +- 0.39
Values mean --- SEM; NM = Newton-meter Note: None of these changes are significant
Table 2. Static tibial cancellous bone parameters (secondary spongiosum and proximal tibial metaphysis)
Group I (Control) II (25 ppm NaF) III (50 ppm NaF) IV
Bone volume (%)
Bone perimeter (mm -1)
Osteoclast no. (ram -2)
Osteoclast perimeter (ram -1 x 10 -3)
Osteoblast perimeter (mm -1 x 10 -3)
14.85 • 1.94 100.00 --- 13.03%
2.54 -+ 0.25 100.00 - 9.79%
0.39 - 0.08 100.00 --- 19.15%
9.00 -+ 1.79 100.00 --- 19.90%
33.81 -+ 6.89 100.00 -+ 20.39%
15.30 +- 1.19 102.87 • 7.99%
2.18 • 0.13 85.97 • 5.09%
0.61 -+ 0.10 153.74 - 25.05%
21.95 - 10.68 221.78 --- 107.96%
15.30 • 5.28 52.28 +- 20.11%
18.28 -+ 0.63 123.10 • 4.26%
2.56 • 0.10 100.87 • 3.82%
0.63 • 0.07 158.66 • 18.88%
12.00 • 201.64 •
16.74 • 1.32 112.71 --- 8.89%
2.53 • 0.13 99.48 • 5.07%
0.48 • 0.14 121.62 • 35.26%
1.85 18.65%
62.67 - 13.27 238.71 • 50.53%
9.00 --- 2.58 90.97 -+ 26.11%
45.83 • 13.29 174.59 • 50.63%
(75 ppm NaF) Mean + SEM BV = bone volume; TV = tissue volume (referent); T.Ar = tissue area (referent); Ob = osteoblast; Oc = osteoclast; Pm = perimeter Note: perimeter is a two-dimensional method of expressing surface
The right femora were subjected to mechanical testing and the left and right tibiae were fixed in formalin and prepared for histomorphometry.
Biomechanical Testing The femora were thawed, dissected free of their soft tissues, and the ends were potted in Cerrobend (Jackson-Walter, King of Prussia, PA). Changes in length and diaphyseal diameter were determined with a micrometer. The potted bones were mounted in the grips of a Servo-actuated (DC motor) torsion testing apparatus and loaded to failure. This apparatus uses a loading rate of 1 rps (360~ and is a motorized modification of the Burstein-Frankel system [12]. The data were digitized, displayed, and stored in an IBM PC/XT computer. Values were obtained for torsional strength, stiffness, angle of deformation, and energy storage capacity.
Histomorphometry The tibiae were fixed in neutral buffered 10% formal saline for a minimum of 72 hours and sectioned transversely in the middiaphysis at the tibio-fibular synostosis. The resulting cross-sections were ground on roughened glass to approximately 25 ~m thickness. Dynamic histomorphometric measurements were made on unstained sections using an Olympus BH-2 microscope (Tokyo, Japan) with UV objectives. A color video camera was mounted on the microscope and linked via a Microcomp digitizing bit-pad (SMI Corporation, GA) to a color monitor and Apple IIe microcomputer using Microcomp software (SMI).
Because rat bone differs from human bone in terms of its form and tissue distribution, the terminology approved by the Histomorphometry Nomenclature Committee of the American Society for Bone and Mineral Research [13] was modified and the following parameters and calculations were used: In diaphyseal cortical bone, medullary area was defined as the area delineated by the endocortical surface of the cross section; cross-sectional area was measured as the area of bone and marrow cavity within the periosteal surface of the specimen; cortical bone area was calculated as the difference between cross-sectional and medullary areas. The lengths of bone perimeter at the periosteal surface and endocortical surface were measured. Further, at the endocortical surface, the lengths of labeled perimeter and nonlabeled perimeter were also measured. The bone area between adjacent fluorochrome labels was measured at the periosteal and endocortical surfaces, respectively, and the group mean bone area between adjacent fluorochrome labels was calculated. This number was divided by the interlabel interval in days to obtain the average bone formation rates. The average formation rates were divided by the periosteal and endocortical label perimeters to obtain the mean periosteal and endocortical matrix apposition rates (MAR). Following removal of the cross sections, the proximal tibial metaphyses were decalcified in 5% formic acid in 10% neutral buffered formalin for 7 days. The bones were bissected in the midsagittal plane, and dehydrated in ascending grades of alcohol before infiltrating and embedding in glycol methacrylate (JB-4 plastic, Polysciences Inc., Warrington, PA). The resulting blocks were sectioned at 5 p,m thickness and the sections were stained for acid phosphatase by an azo-dye coupling technique and counterstained with methyl green thionine. Cancellous bone was examined in the secondary spongiosa of the proximal tibial metaphysis at a standard sampling site no less than 1
T. A. Einhorn et al.: NaF Does Not Impair Mechanical Properties of Cortical Bone
129
Table 3. Static tibial cortical bone parameters
Group I (Control) II (25 ppm NaF) III (50 ppm NaF) IV (75 ppm NaF)
Group I (Control) II (25 ppm NaF) III (50 ppm NaF) IV (75 ppm NaF)
Group I (Control) lI (25 ppm NaF) III (50 ppm NaF) IV (75 ppm NaF)
Crosssectional area (mm 2)
Medullary area (mm 2)
Cortical bone area (ramz)
Periosteal perimeter (mm)
Endocortical perimeter (mm)
3.23 -+ 0.064
0.69 -+ 0.01
2.54 -+- 0.06
6.53 -+ 0.07
3.01 -+ 0.03
3.40 -+ 0.08
0.76 P < 0.77 P
0.01 287.69 -+ 5.04 P > 0.001 369.12 -+ 7.56 P > 0.001
154.8 -+ 18.8 160.4 -+ 14.4 156.4 - 15.0
Values = mean -+ SEM
would indicate that if fluoride alters the relationship between bone strength and bone mass or density, it may do so only when bone formation or remodeling has been altered by fluoride administration. In the present study, none of the doses of NaF led to changes in bone histomorphometric measurements, and in particular, no increased bone formation was observed. One of the limitations in extrapolating these results to patients is the fact that F did not produce the increase in bone formation and changes in other bone histomorphometric measurements that would be expected based upon what is known of the effects of NaF in human bone. Although it is clear from the data that there was a direct linear relationship between the content of F in the drinking water and the incorporation of fluoride into bone, the ingestion of these amounts of N a F did not produce a change in bone mass. Two possible explanations for these findings are (1) NaF is less effective in stimulating new bone formation in very immature animals because their bones are already undergoing formation at maximal levels, and (2) the doses of F used were insufficient to stimulate bone formation in any animal. However, dose ranges of NaF between 25 and 75 ppm in drinking water, leading to an estimated daily NaF intake of between 0.5 and 1.0 g/rat, would seem to be sufficient to produce a formation-stimulating effect based on the results
of previously published works [7-11]. A recent report by Narita et al. [15] showed that doses of NaF similar to those used in the present experiment lead to a distribution of F in cortical bone whereby it is highest in the outer lamellae and lowest on the endosteal surface [15]. Moreover, these investigators found an excellent correlation between F intake and bone fluoride content. As the failure pattern of bone loaded in torsion is initiated in the outer cortex [16], these findings would suggest that the femora loaded in our experiments failed through sections of bone where the F concentration would have been significant. It should also be noted that only one method of mechanical testing, torsional loading, was used in this experiment. It is possible that other loading conditions such as compression, tension, or three- or four-point bending would yield different results. However, although vertebrae are most commonly loaded in compression, femoral and tibial fractures sustained in patients generally occur as a result of a combination of forces, the most significant of which is torsional loading [12, 16]. In order for a bone to be loaded in tension it would literally have to be pulled apart from either end. Moreover, for pure bending conditions to occur in nature, a long bone would have to be fixed at two points and have the direction of force occur at a perpendicular angle between the points of fixation. The value of torsional loading in mechanical testing is not only that it more closely approximates how fractures occur clinically, but it also fulfills other critical criteria essential to an experimental test. These include (1) the loading configuration subjects the bone to equally severe loading conditions at every section along its length so as to be able to identify weak sections; (2) the loading mode is not critically dependent upon bone geometry, in particular, bone length; and (3) the loading configuration allows for the tester to control for rates of application of loads, so that the time conditions produced in the test will be reproducible and preferably representative of those that occur during normal trauma [12]. Because the experimental conditions in this study did not produce a change in bone histology or bone mass, the results may not explain many of the current controversies surrounding the use of NaF in patients. However, the results clearly
T. A. Einhorn et al.: NaF Does Not Impair Mechanical Properties of Cortical Bone show that a marked increase in bone F content alone does not necessarily lead to an alteration in mechanical properties or an increase in the propensity to fracture under static load. This would suggest that the observed disparity between alterations in bone mass in patients treated with N a F and their lack of protection from spinal fractures may need to be explained on the basis of some effect of N a F on bone formation and remodeling as opposed to its mere incorporation into bone.
Acknowledgments. We are grateful to David Glass and Gita Bosch for technical assistance. This work was supported in part by NIH Grant 5 RO1 AR31062, the Orthopaedic Research and Education Foundation, and Veteran's Administration Merit Review Funds.
References 1. Riggs BL, Seeman E, Hodgson SF, Taves DR, O'Fallon WM (1982) Effect of the fluoride/calcium regimen on vertebral fracture occurrence in postmenopausal osteoporosis. N Engl J Med 306:446--450 2. Hansson T, Roos B (1978) Effect of combined therapy with sodium fluoride, calcium and vitamin D on the lumbar spine in osteoporosis. AJR 126:1294-1297 3. Jowsey J, Riggs BL, Kelly PJ, Hoffman DL (1972) Effect of combined therapy with sodium fluoride, vitamin D and calcium in osteoporosis. Am J Med 53:43-49 4. Charles P, Mosekilde L, Taagehoj Jensen F (1985) The effects of sodium fluoride, calcium phosphate and vitamin D2 for one to two years on calcium and phosphorous metabolism in postmenopausal women with spinal crush fracture osteoporosis. Bone 6:201-206
131
5. Riggs BL, Hodgson SF, O'Fallon WM, Chao EYS, Wahner HW, Muhs JM, Cedel SL, Melton LJ (1990) Effect of fluoride treatment on the fracture rate in postmenopausal women with osteoporosis. N Engl J Med 322:802-809 6. Kleerekoper M, Peterson E, Phillips E, Nelson D, Tilley B, Parfitt AM (1990) Continuous sodium fluoride therapy does not reduce vertebral fracture incidence in postmenopausal osteoporosis. J Bone Miner Res 4:5376 7. Schraer H, Posner AS, Schraer R, Zipkin I (1972) Effect of fluoride on bone minerals. Clin Orthop 86:260-286 8. Baud CA, Bang S (1973) Biophysical study of the bone tissue of fluoride-treated mice. J Dent Res 52:589 9. Franke J, Runge H, Grau P, Fengler F, Wanka C, Rempl H (1976) Physical properties of fluorosis bone. Acta Orthop Scand 47: 20-27 10. Carter DR, Beaupr6 GS (1990) Effects of fluoride treatment on bone strength. J Bone Miner Res 5:S177-S184 11. Riggins RS, Rucker RC, Chan MM, Zeman F, Beljan JR (1976) The effect of fluoride supplementation on the strength of osteopenic bone. Clin Orthop 114:352-357 12. Burstein AH, Frankel VH (1971) A standard test for laboratory animal bone. J Biomech 4:155-158 13. Parfitt AM, Drezner MK, Glorieux FH, Kanis JA, Malluche H, Meunier PJ, Ott SM, Recker RR (1987) Bone histomorphometry: standardization of nomenclature, symbols and units. J Bone Miner Res 2:595-610 14. Dunnet CW (1964) New tables for multiple comparisons with a control. Biometrics 20:482--491 15. Narita N, Kato K, Nakagaki H, Ohno N, Kameyama Y, Weatherell JA (1990) Distribution of fluoride content in rat's bone. Calcif Tissue Int 46:200-204 16. Hayes WC, Carter DR (1979) Biomechanics of bone. In: DJ Simmons, AS Kumin (eds) Skeletal research--an experimental approach. Academic Press, New York, pp 263-300
