Journal of Orthopaedic Research 823%?46 Raven Press, Ltd., New York 0 1990 Orthopaedic Research Society
Internal Remodeling of Periosteal New Bone During Fracture Healing Hannu T. Aro, Burkhard W. Wippermann, Stephen F. Hodgson, and Edmund Y. S. Chao Biomechanics Laboratory, Department of Orthopedics, and *Department of Internal Medicine, Mayo CliniclMayo Foundation, Rochester, Minnesota, U.S.A.
Summary: A closed fracture model of the rat tibia was employed to study internal remodeling of periosteal new bone during fracture repair. Static histomorphometric parameters of osteoid surface (or perimeter) and eroded surface (resorption surface) were used as indicators of appositional bone formation and resorption of bone trabeculae, respectively. Intracortical remodeling at the fracture site was evaluated using quantitative tetracycline histology and microradiography. The extents of osteoid and eroded bone surfaces did not differ significantly in the periosteal woven new bone in the early phases of fracture healing. Later on, the periosteal new bone had significantly more osteoid surface than eroded surface (p < 0.001). The number of osteoclasts also decreased significantly over time during fracture healing (p = 0.028). Cortical bone showed a continuous increase of porosity (p < 0.01) between 1 and 6 weeks after fracture. These results suggest that there is a time-related change in the balance of periosteal bone formation and resorption during the progress of fracture repair. We hypothesize that this change was related to the restoration of bony continuity. Further studies are, however, needed to indicate the histomorphometric features of periosteal new bone in fracture nonunions. Key Words: Fracture-Callus-Fixation-Osteoblast-Osteoclast.
allows the analysis of periosteal, intracortical, and endosteal new bone formation and porosity (3,19,29). These techniques determine the morphometric progress of fracture union and remodeling, but they do not show the cell activities of bone formation and resorption. Histologic quantification of osteoblastic and osteoclastic activities is a standardized diagnostic method in metabolic bone diseases (18). The use of this method has been uncommon in fracture healing studies (26). In this technique, the osteoblastic activity is estimated based mainly on the two-dimensional or three-dimensional measurement of osteoid production by osteoblasts (osteoid surface or perimeter, osteoid volume). The osteoclastic activity is estimated by the relative presence of scalloped resorbing lacunae (eroded surface).
Optimal mechanical conditions for fracture healing are still unknown. This is partially due to methodological difficulties in measuring cell activities of new bone formation and bone resorption in healing bones under different mechanical conditions. If reliable and accurate quantitative methods can be developed for the estimation of cellular activities during fracture repair, then the direct cellular responses to mechanical factors may be defined in well-planned experimental models. Quantitative tetracycline histology, combined with high-resolution microradiography (13,14,27), ~~
Received August 2, 1988; accepted May 16, 1989. Address correspondence and reprint requests to Dr. E. Y. S. Chao at Biomechanics Laboratory, Department of Orthopedics, Mayo ClinicMayo Foundation, Rochester, MN 55905, U.S.A.
238
REMODELING OF PERIOSTEAL NEW BONE
239
Our objective in the current study was to determine the histomorphometric characteristics of periosteal new bone during the sequential stages of fracture repair. The time-related changes in the histomorphometric properties of the periosteal new bone were correlated with the progress of intracortical remodeling. MATERIALS AND METHODS Animal Model Twenty-five 2-month-old male Sprague-Dawley rats (mean weight, 226 g; SD, 7g) were used. Closed standardized fracture of the tibia was produced bilaterally in each animal. The tibia was prenailed with three loose-fitting stainless steel intramedullary pins (diameter, 0.4 mm) and then was fractured manually (2). The fixation was able to maintain axial alignment of the fracture but without interference with bone fragment relative motion at the fracture site. Such intramedullary fixation allowed formation of external fracture callus (l), and immediate weight bearing was allowed after surgery. The animals resumed normal walking within a few days after fracture. An intramuscular injection of oxytetracycline 25 mg/kg was administered at the time of surgery and once a week thereafter. The time between administration of the tetracycline label and sacrifice was 7 days. The animals were sacrificed at a random sequence: five animals at 1 week and four animals each at 2 , 3 , 4 , 5 , and 6 weeks after surgery. Two-plane radiographs obtained immediately after surgery and before sacrifice were used to determine the type of fracture produced, the state of union, and the position of the intramedullary rods (Fig. 1). After sacrifice, the hind legs were disarticulated at the knee joints, and the intramedullary rods were extracted. The first six rats, one animal at each time interval, were used to standardize the experimental methods and the testing condition. The remaining 19 animals were used for the analysis. Preparation of Fracture Callus Specimens After careful dissection of soft tissues, the fracture region of interest (3-mm-thick segment of the midportion of the fracture) was precisely sectioned by a diamond saw. The thickness of the specimens was measured using a micrometer caliber. This fracture segment contained the critical zone of frac-
FIG. 1. Rat tibia1 fractures fixed by intrarnedullary rods united within 4 weeks, as demonstrated radiographically.
ture union process where the advancing front of the anchoring external callus ultimately unite (2). The mineral content of the sectioned specimens was determined by means of micro-bone-densitometry , which is a high-resolution single-photon absorptiometry (12). Micrsbone-densitometry is a custommade, noncommercial device constructed for the analysis of small bone specimens. The device uses a collimated narrow beam (1 m in diameter) of gamma rays from a low-energy isotope source (1251,27.5 keV, 70 mCi) for the precise rectilinear scanning of the specimen. Scanning is performed at 1-s intervals and slow speed (6 mdmin) with small 0.1-mm increments. The count data of the attenuated beam are acquired every 1/2 s. The data processing is performed using a bone edge detection algorithm. The device provides high-resolution computer screen display of the mineral distribution. The specimen was subjected to mechanical testing as reported elsewhere (4). The fracture callus specimens were fixed in 70% alcohol, dehydrated in increasing concentrations of alcohol, defatted in xylene, and embedded in methyl methacrylate to which glycol methacrylate, polyethylene glycol, and dibutyl phthalate (Polysciences, Inc., Warrington, PA, U.S.A.) were added to improve cutting properties and decrease shrinkage. Undecalcified 5-pm transverse sections were cut with a microtome (Reichert-Jung 1140, Reichert Scientific Instruments, Buffalo, NY, U.S.A.) and were stained with a modified Goldner’s trichrome stain (9). The sections were examined under light microscopy. The tissue composition of the
J Orthop Res, Vol. 8, No. 2, 1990
H . T . ARO ET AL.
240
transverse fracture sections was measured using a digitizing tissue-planimetry , as previously described (2). Each histologic section was projected on a digitizing table using magnification of 25 times, and a diagram of the external callus was sketched. The cross-sectional area of the whole periosteal callus and the area of periosteal new bone (including porosity) were measured from the sketches using a modified SLICE program (17). Previous studies of a rabbit fracture model have shown that the restoration of fracture strength and stiffness appear to be related to the amount of new bone connecting the fragments measured from cross sections (5). In the present study, the structural strength of the healing fractures was not determined. The mineral content as determined by microbone-densitometry was expressed as the total amount of bone mineral (BMC). This value represented the mineral content of the whole 3-mm-thick specimen, including the original cortical bone. The normalized density of bone minerals was calculated per unit volume of each specimen and the mineral density of external fracture callus, 6, could be expressed as S=
(BMC) - (BMC)o A x t
where (BMC) = bone mineral content in mg, (BMC), = bone mineral content of 1-week specimen in milligrams, A = area of the specimen scanned in square millimeters, and t = thickness of the specimen in millimeters. This study did not include any sham-operated group. The normalization was performed against the mean mineral content of the 1-week specimens (16.7 k 2.3 mg, n = 4) to give an approximate estimation of the contribution to the measurement of the cortical bone mass. This normalization did not take into account any time-related changes or specimen-related differences in the mineral content of the cortical bone and endosteal new bone. Histologically, the one-week specimens showed new of the cross-sectional callus area. bone in 4% Thus, our estimation of fracture callus mineral density was considered acceptable. Histomorphometric Evaluation Histomorphometric analysis was performed using a semiautomated skeletal histomorphometric technique (Bioquant, R 8z M Biometrics, Inc.,
J Orthop Res, Val. 8, N o . 2 , 1990
Nashville, TN, U.S.A.). The measurements were two-dimensional and were performed on four standardized quarters of the external fracture callus, on the anterior, posterior, medial, and lateral aspect of the tibia. The following static histomorphometric parameters were analyzed. 1. Osteoid surface = (trabecular new bone (or perimeter) surface length covered by osteoid)/(total trabecular new bone surface length) x 100 (%) (trabecular new bone 2. Eroded surface = surface length made up by scalloped resorbing osteoclast-positive or osteoclast-negative lacunae)/(total trabecular new bone surface length) x 100 (%) 3 . Osteoclast number = ( n u m b e r of o s t e o clasts)/(100 mm of trabecular bone surface length) (dl00 mm) 4. Quiescent surface = 100 - (osteoid surface (a derived index) + eroded surface) (%) In these measurements, trabecular new bone referred to periosteal (wovedcancellous) new bone. The results for the four quarters of each specimen were averaged because of the small dimensions involved. Evaluation of Intracortical Remodeling
A 200-pm transverse section was cut with a diamond saw (Buehler Isomet, Lake Bluff, IL, U.S.A.) from the methyl metacrylate block of each undecalcified fracture specimen next to the thin histomorphometric section. The section was ground to 120 p,m and was prepared for contact microradiography performed on a high-resolution contact film (Kodak spectroscopic plates, Eastman Kodak, Rochester, NY, U.S.A.). The unstained sections were viewed under ultraviolet fluorescence microscopy. The intracortical areas of new bone formation, as indicated by the tetracycline labeling, and bone porosity were measured according to the point counting method of Harris and Weinberg (11). New bone formation within osteocyte lacunae was not included in the analyses of new bone formation. The area of osteocyte lacunae was not included in cortical bone porosity calculation.
REMODELING OF PERIOSTEAL NEW BONE
Statistical Analysis For statistical purposes, the experimental unit was one animal. Selection of the fracture for the statistical analysis in each animal was randomized. One fracture healed without substantial external callus, and the contralateral fracture of this animal was used for the analysis. The time-related changes in the different parameters of new bone and cortical remodeling were analyzed using analysis of variance (SAS Software, SAS Institute, Inc., Cary, NC, U.S.A.). If a significant change was observed, Tukey's studentized range test at a confidence level of 95% (p < 0.05) was used to indicate the healing period when the significant change occurred. RESULTS Progress of Fracture Healing Transverse or slightly oblique fractures were obtained at the distal tibiofibular junction. Radiographic union by external callus was observed within 4 weeks (Fig. 1). The total mineral content of the fracture specimens reached the maximum value also by 4 weeks (Table 1). At this stage of healing, nonosseous tissue (fibrous tissue, mineralized and nonmineralized cartilage) constituted a total of 22% of the cross-sectional callus area. The crosssectional callus area decreased significantly (p < 0.05) between 4 and 6 weeks (Table l), indicating the start of external remodeling. The estimated mineral density showed a 2.1-fold increase (p < 0.05) between 4 and 6 weeks (Table 1).
241
woven new bone at 2, 3, and 4 weeks after fracture (Table 2). Neither of these parameters nor the total amount of inactive bone surface showed statistically significant changes with the healing time. There was, however, a tendency to a decreasing extent of eroded surface after 4 weeks. The extent of osteoid surface tended to increase slightly over time. Consequently, the periosteal new bone had significantly more osteoid than eroded surface at 5 and 6 weeks (p < 0.001, two-way analysis of variance). The number of osteoclasts on the periosteal new bone (Figs. 2A and B) was highest at 2 weeks and thereafter significantly decreased with the healing time (p = 0.028) (Table 2). Between 4 and 6 weeks, the periosteal woven new bone maturated into lamellar cancellous bone, which formed an osseous shell around the healing bone (Fig. 3A). The outer surface of the shell showed bone resorption lacunae (Fig. 3B), indicating the initiation of external remodeling processes. The spatial distribution of osteoblastic activity, as indicated by the spatial location of the osteoid, was nonuniform at this stage of healing. Osteoid formation was most prominent in the inner surface of the shell (Fig. 3C).
Histomorphometry The extents of osteoid surface and eroded bone surface did not differ significantly in the periosteal
Intracortical Remodeling Intracortical porosity showed a 3.3-fold increase (p < 0.01) during the first 6 weeks of fracture healing (Table 3). The resorption canals were formed in both the axial and transverse directions of the bone (Fig. 4A). Osteoclasts were occasionally seen to form cutting cones, characteristic of the formation of intracortical resorption canals. The resorption canals were generally lined by a thin layer of new bone (Fig. 4B). Intracortical new bone did not increase significantly with the fracture healing time.
TABLE 1. Cross-sectional callus area, exterior callus new bone area, and bone mineral content of fracture specimens (mean 5 SD, n = 3) Healing time (wks) Cross-sectional callus area (mm*) Periosteal new bone area (mm') (%o)c Bone mineral content (mg) Bone mineral density (mg/mm3 x 10')
2 27.2 i 5.0 4.5 i 2.2 17.4 2 10.8 27.4 i 2.5 12.3 i 1.5
3 35.4 i 6.1 9.7 i 7.7 25.4 2 18.0 39.0 i 12.3 18.6 i 7.7
4 29.0 i 3.6 22.4 i 2.5 78.4 2 16.8 45.9 i 7.4 31.5 2 7.8
5 20.6 2 4.3 20.5 i 4.4 99.6 i 6.7 44.9 i 10.0 43.0 2 5.6
6 12.2 f 3.9b 12.2 i 3.9 100.0 2 0.0 40.6 i 4.3 66.8 i 16.4d
p p
P" 0.0012 = 0.0037
=
NS p =0.0002
" Statistically significant in time-related change (NS
= not significant). The 6-week value significantly lower than the 4-week value (Tukey's studentized range test, p < 0.05). Expressed as a percentage of the cross-sectional area. The 6-week value significantly higher than the 4-week value (Tukey's studentized range test, p < 0.05).
J Orthop Res, Voi. 8, N O . 2 , 1990
242
H . T . ARO ETAL. TABLE 2. Changes in histomorphometric properties of the periosteal new bone in the rat tibia during fracture healing (mean 2 SD, n = 3) Healing time (wks)
Osteoid surface (%) Eroded surface (%) Quiescent surface (%) Osteoclast number
2 19.4 2 17.0 12.5 2 4.7 68.1 i 16.0 123.5 i 60.5
3 20.8 2 10.4 14.3 i 2.1 65.0 2 9.0 64.8 2 48.2
4 28.6 i 16.5 10.5 2 6.8 60.9 i 8.5 28.9 i 44.0
5
34.6 9.1 56.4 9.8
i 4.1** i 3.8 i 2.5 i 12.0
6
Pa
37.2 i 6.0b 4.0 6.8 I 56.0 t 2.0 1.5 2 8.0
NS NS NS p = 0.028
Significance in time-related changes (one-way analysis of variance); NS, not significant. Amount of osteoid surface significantly higher compared with that of eroded surface (Tukey’s test and studentized range test, p < 0.05). a
DISCUSSION Quantitative analysis of bone formation and bone resorption is essential to the evaluation of control mechanisms of the bone repair processes. The current experiment showed that standard histomorphometric methods may give useful information about the cellular activities of internal remodeling processes at the fracture site. For the convenience of discussing the results, bone resorption and formation activities occurring within the callus shell are described as internal remodeling. On the other hand, cellular activities related to the change of cross-sectional areas of callus are defined as external bone remodeling. The extents of osteoid and eroded bone surfaces
were similar in the woven new bone of the external callus in the early phases of fracture healing. This finding supports the concept (24) that the internal remodeling of periosteal new bone starts before fracture union. Previous studies of the same rat fracture model showed that the expression of type I collagen genes, as indicated by quantitative measurement of mRNA levels for type I procollagen chains per unit volume of the callus, reached maximum by 2 weeks and remained stable between 2 and 4 weeks (16). In-situ hybridization revealed a strong labeling with pro l(1) collagen cDNA probe in osteoblasts lining trabeculae of the woven new bone (21). The current histomorphometric evaluation also showed a continuous existance of osteoid surface in the perios-
FIG. 2. A: The remodeling of periosteal new bone included appositional new bone formation on the surface of woven new bone (NB), as indicated by the arrow, the presence of osteoid layer (0s)lined by cuboidal osteoblasts (OBL). A newly entrapped osteocyte (OST) is visible within the bone trabeculae. B: Multinucleated osteoclasts (OCL) were commonly observed in the immediate vicinity of osteoblasts (OBL) lining a osteoid layer (0s).Note the eroded lacunae of the bone surface (ES) next to the osteoclast. (Two weeks after fracture, histologic sections stained with Goldner’s stain, original magnification ~250.)
J Orthop Res, Vol. 8, No. 2 , 1990
REMODELING OF PERIOSTEAL NEW BONE
243
C
FIG. 3. A: Periosteal new bone formed a thick outer shell around the fracture site of the tibia during the remodeling phase of healing. OCB, original cortical bone; HRL, Howship’s resorption lacunae; NB, new bone; BFS, bone formation surface; f, fibula (microradiograph, 6 weeks after fracture, original magnification x25). B:The outer surface of the bony callus was rough in certain regions because of the formation of Howship’s resorption lacunae (HRL). These lacunae occasionally contained osteoclasts in histologic sections. Note that the corresponding inner surface of the shell contained bone formation surface (BFS) that appeared smooth; NB, new bone; OCB, original cortical bone (microradiograph, 5 weeks after fracture, original magnification x50). C: The mature cancellous lamellar bone of the callus shell showed thick osteoid (0s) layer in its inner surface indicating the progress of the closure of the space between the shell and the original cortical bone; NB, new bone; OBL, osteoblasts (histologic section stained with Goldner’s stain, 5 weeks after fracture, original magnification ~ 5 0 ) .
J Orthop Res, Vol. 8,No. 2,1990
H . T . ARO ET AL.
244
TABLE 3 . Intracortical remodeling of the rat tibia during fracture healing (mean -t SO)" Healing time (wks) 1
2
3
4
5
Pb
6 ~~
Original cortical bone (unlabeled) (%) Intracortical new bone (%) Intracortical porosity (%) a
87.8 i 2.1 8.8 t 1.4 3.4 i 0.8
82.0 i 5.5 14.3 i 4.7 2.7 2 1.2
81.8 i 4.9 11.7 i 3.0 6.5 i 1.9
77.8 2 4.0 14.2 i 2.0 8.0 i 2.5
78.2 i 6.0 13.0 i 1.9 8.9 2 4.1
74.4 2 5.5 14.1 2 0.9 11.5 i 4.6
p
< 0.025 NS
p < 0.01
n = 3 except for the 1-week group (n = 4). Significance in time-related change (one-way analysis of variance); NS, not significant.
teal new bone. These results showed that the internal remodeling of periosteal new bone involves a considerable steady collagenous bone matrix production by osteoblasts. After the early phases of fracture healing, the periosteal bone had more osteoid than eroded surface. Theoretically, this change should result in increased density of the periosteal new bone because
of the continuous new bone formation in the presence of decreased bone resorption. Absorptiometry techniques are accurate in the measurements of bone minerals (15,28). In the current study, the periosteal new bone showed a significant increase in the density of bone minerals during fracture healing. This result confirmed that the observed change in the balance between appositional formation of
FIG. 4. lntracortical remodeling of the rat tibia during fracture healing. A: Longitudinal and transverse resorption canals (arrows) developed in cortical bone (3 weeks after fracture, microradiograph original magnification x50). B: The examination of the same specimen under ultraviolet microscopy (new bone labeled by tetracycline white) showed that these resorption canals were partly covered by a thin layer of new bone (arrows). PNB, periosteal new bone; ENB, endosteal new bon (original magnification x50).
J Orthop Res, Vol. 8, N o . 2 , 1990
REMODELING OF PERIOSTEAL NEW BONE
lamellar bone on the surface of woven bone and resorption of woven bone trabeculae had a positive net effect on bone density. It should be emphasized that the present experiment included only an approximate estimation of callus mineral density because both the effects of intracortical bone resorption and endosteal new bone formation could not be determined precisely. Cellular activities of the external remodeling processes of the periosteal new bone were not quantitated in this study. However, the decrease in crosssectional area of the callus and the qualitative evidence of bone resorption showed that external remodeling started at 4 weeks. This external remodeling resulting in an obvious decrease of callus volume, providing the explanation of why the absolute mineral mass of the periosteal new bone did not show a continuous increase during fracture healing. It seems that external remodeling is a response to the increasing density and maturation of the periosteal new bone after fracture union. Osteoclasts are of monocytic-macrophagic origin and invade the fracture site of the rat tibia at an early stage of healing (10). The close presence of multinucleated osteoclasts to osteoblasts lining the osteoid layer suggests an interaction between these two cell types during the early stages of fracture healing. Osteoclasts have been shown to exhibit expression of mRNAs for a transforming growth factor-p (TGF-P) and cellular protooncogene c-fos during endochondral ossification processes of human growth plates (21,22). Both TGF-p and protooncogene c-fos have been implicated to have important roles in the metabolism of chondrocytes and osteoblasts (8,20,25). All these data and the current study suggest that osteoclasts could have a function in the control of endochondral fracture healing processes. Rat cortical bone does not possess a haversian system. However, the rat appears to develop a marked remodeling process of the cortical bone during fracture healing. This process seems to take part in the revascularization process of the fracture ends. In the current model, the radial growth of penetrating blood vessels from the medullary cavity to the external callus may have contributed to the observed intracortical porosity. Intracortical bone porosity increase may also have been a sequential change in response to load by passing through periosteal new bone formation. It is unlikely that the fracture fixation method used (loose-fit or flexible
245
intramedullary wires) could cause bone stressprotection, thereby leading to osteopenia. The amount of intracortical new bone did not increase significantly during the experimental period, indicating a lag period between intracortical resorption and new bone formation. This finding is compatible with the theory of Charnley (6). He proposed that the bridging mechanism of periosteal callus is to pass the necrotic fracture ends before the slow start of cortical healing. Ultimately, the cortical reconstruction by secondary osteons seems to be important to regain the ultimate bone union strength (3,7,23). Factors that control internal and external bone remodeling processes of the external fracture callus are not completely known. The remodeling seems to be, at least after fracture union, related to the mechanical factors as dictated by the functional demands according to Wolff's law. Fracture union constitutes, histologically, the establishment of a bony bridge between the fracture fragments and, biomechanically , the restoration of the hard-tissue characteristics of the bone. The current experiment showed that the progress of fracture healing is associated with a change in the balance of periosteal new bone formation and resorption. This change coincided with the time of callus ossification and the start of external remodeling, suggesting a relationship with the restoration of bony continuity. However, further studies are needed to exclude a similar time-related change of cell activities in fractures that fail to unite. These results may help further research to optimize the biomechanical properties of fracture fixation devices. Acknowledgment: The authors thank Jan Palmer for her collaboration, and Glenda Evans, Michael Bateman, and James Bronk for their technical assistance.
REFERENCES Aro H, Eerola E, Aho AJ: Osteolysis after rigid fixation. The possible role of periosteal neural mechanoreceptors in bone remodeling. Clin Orthop 166:292-300, 1982 Aro H, Eerola E, Aho AJ: Determination of callus quantity in 4-week-old fractures of the rat tibia. J Orthop Res 3: 101108, 1985 Aro HT, Kelly PJ, Lewallen DG, Chao EYS: Comparison of the effects of dynamization of constant rigid fixation on rate and quality of bone osteotomy union in external fixation. Orthop Trans 12:378, 1988 Aro HT, Wippermann BW, Hodgson SF, Wahner HW, Lewallen DG, Chao EYS: Noninvasive monitoring of fracture callus mineralization using high-resolution single-photon absorptiometry. Orthop Trans 12:489, 1988
J Orthop Res. Vol. 8, No. 2, 1990
246
H . T. ARO ETAL.
5. Black J, Perdigon P, Brown N, Pollack SR: Stiffness and strength of fracture callus. Relative rates of mechanical maturation as evaluated by a uniaxial tensile test. Clin Orthop 182:278-287, 1984 6. Chamley J: The Closed Treatment of Common Fractures, 3rd ed. Edinburgh, London, New York, Churchill Livingstone, 1974, p 1-42 7. Claes L, Bum C, Gerngross H, Mutschler W: Bone healing stimulated by plasma factor XIII. Osteotomy experiments in sheep. Acta Orthop Scand 5657-62, 1985 8 . Ellisgsworth LR, Brennan JE, Fok K, Rosen DM, Bentz H, Piez KA, Seydin SM: Antibodies to the N-terminal portion of cartilage-inducing factor A and transforming growth factor p. J Biol Chem 261:12362-12367, 1986 9. Goldner J: A modification of the Masson trichrome technique for routine laboratory purposes. Am J Pathol 14:237243, 1938 10. Gothlin G, Ericsson JLE: The osteoclast. Review of ultrastructure, origin, and structure-function relationship. Clin Orthop 120:201-231, 1976 11. Harris WH, Weinberg EH: Microscopic method of measuring increases in cortical bone volume and mass. Calcif Tissue Res 8:19&196, 1972 12. Hodgson SF, Wahner HW, Dunn WL, Bateman MD: Micro bone densitometry of human transiliac bone biopsies: a new approach to the direct measurement of skeletal mass in man. Bone Mineral Measurements by Photon Absorptiometry: Methodological Problems, ed by J Dequeker, P Geusens, HW Wahner, Leuven, Belgium, Leuven University Press, 1988, p 461 13. Jowsey J, Kelly PJ, Riggs BL, Bianco AJ Jr, Scholz DA, Gershon-Cohen J: Quantitative microradiographic studies of normal and osteoporotic bone. J Bone Joint Surg [Am] 47:785-806, 1965 14. Kelly PJ, Peterson LFA, Janes JM: A method of using sections of bone prepared for microradiography for subsequent histologic study. Proc Staff Meet Mayo Clin 34:274-283, 1959 15. Mazess RB: The noninvasive measurement of skeletal mass. In: Bone and Mineral Research, by WA Peck, Amsterdam, Oxford, Princeton, Excerpta Medica, 1983, pp 223-279 16. Multimaki P, Aro H, Vuorio E: Differential expression of fibrillar collagen genes during callus formation. Biochem Biophys Res Commun 142:536-541, 1987 17. Nagurka ML, Hayes WC: An interactive graphics package
J
Orthop Res, Vol. 8, NO. 2, 1990
for calculating cross-sectional properties of complex shapes. J Biomech 135944, 1980 18. Parftt AM, Drezner MK, Glorieux FH, Kanis JA, Malluche H, Meunier PJ, Ott SM, Recker RR: Bone histomorphometry: standardization of nomenclature, symbols, and units. J Bone Mineral Res 2595410, 1987 19. Rand JA, An KN, Chao EYS, Kelly PJ: A comparison of the effect of open intramedullary nailing and compression plate fixation on fracture-site blood flow and fracture union. J Bone Joint Surg [Am] 63:427442, 1981 20. Ruther U , Garber C, Komitowski D, Muller R, Wagner EF: Deregulated c-fos expression interferes with normal bone development in transgenic mice. Nature 325:412416, 1987 21. Sandberg M, Aro H, Multimaki P, Aho H, Vuorio, E: In situ localization of collagen production by chondrocytes and osteoblasts in fracture callus. J Bone Joint Surg 71:69-77, 1989 22. Sandberg M, Vuorio T, Hirvonen H, Alitalo K, Vuorio E: Enhanced expression of TGF-a and c-fos mRNAs in the growth plates of developing human long bones. Development 102:461-470, 1988 23. Schenk RK: Histophysiology of bone remodeling and bone repair. In: Perspectives on Biomaterials, ed by OCC Lin, EYS Chao, Amsterdam, Elsevier Science Publishers B.V., 1986, pp 75-99 24. Sevitt S : Bone Repair and Fracture Healing in Man. Edinburgh, London, Melbourne and New York, Churchill Livingstone, 1981 25. Seyedin SM, Thomas TC, Thompson AY, Rosen DM, Piez KA: Purification and characterization of two cartilageinducing factors from bovine demineralized bone. Proc Natl Acad Sci USA 82:2267-2271, 1985 26. Shih M-S, Norrdin RW: Effect of prostaglandin El on the periosteal regional acceleratory phenomenon in fractured ribs: histomorphometry study in beagles. A m J Vet Res 48:828-830, 1987 27. Vanderhoeft PJ, Kelly PJ, Peterson LFA: Determination of growth rates in canine bone by means of tetracycline-labeled patterns. Lab Invest 11:714-726, 1962 28. Wahner HW, Riggs BL: Methods and application of bone densitometry in clinical diagnosis. CRC Crit Rev Clin Lab Sci 24:3:217-233, 1986 29. Wu J-J, Shyr HS, Chao EYS, Kelly PJ: Comparison of osteotomy healing under external fixation devices with different stiffness characteristics. J Bone Joint Surg [ A m ] 66:1258-1264, 1984
Internal Remodeling of Periosteal New Bone During Fracture Healing Hannu T. Aro, Burkhard W. Wippermann, Stephen F. Hodgson, and Edmund Y. S. Chao Biomechanics Laboratory, Department of Orthopedics, and *Department of Internal Medicine, Mayo CliniclMayo Foundation, Rochester, Minnesota, U.S.A.
Summary: A closed fracture model of the rat tibia was employed to study internal remodeling of periosteal new bone during fracture repair. Static histomorphometric parameters of osteoid surface (or perimeter) and eroded surface (resorption surface) were used as indicators of appositional bone formation and resorption of bone trabeculae, respectively. Intracortical remodeling at the fracture site was evaluated using quantitative tetracycline histology and microradiography. The extents of osteoid and eroded bone surfaces did not differ significantly in the periosteal woven new bone in the early phases of fracture healing. Later on, the periosteal new bone had significantly more osteoid surface than eroded surface (p < 0.001). The number of osteoclasts also decreased significantly over time during fracture healing (p = 0.028). Cortical bone showed a continuous increase of porosity (p < 0.01) between 1 and 6 weeks after fracture. These results suggest that there is a time-related change in the balance of periosteal bone formation and resorption during the progress of fracture repair. We hypothesize that this change was related to the restoration of bony continuity. Further studies are, however, needed to indicate the histomorphometric features of periosteal new bone in fracture nonunions. Key Words: Fracture-Callus-Fixation-Osteoblast-Osteoclast.
allows the analysis of periosteal, intracortical, and endosteal new bone formation and porosity (3,19,29). These techniques determine the morphometric progress of fracture union and remodeling, but they do not show the cell activities of bone formation and resorption. Histologic quantification of osteoblastic and osteoclastic activities is a standardized diagnostic method in metabolic bone diseases (18). The use of this method has been uncommon in fracture healing studies (26). In this technique, the osteoblastic activity is estimated based mainly on the two-dimensional or three-dimensional measurement of osteoid production by osteoblasts (osteoid surface or perimeter, osteoid volume). The osteoclastic activity is estimated by the relative presence of scalloped resorbing lacunae (eroded surface).
Optimal mechanical conditions for fracture healing are still unknown. This is partially due to methodological difficulties in measuring cell activities of new bone formation and bone resorption in healing bones under different mechanical conditions. If reliable and accurate quantitative methods can be developed for the estimation of cellular activities during fracture repair, then the direct cellular responses to mechanical factors may be defined in well-planned experimental models. Quantitative tetracycline histology, combined with high-resolution microradiography (13,14,27), ~~
Received August 2, 1988; accepted May 16, 1989. Address correspondence and reprint requests to Dr. E. Y. S. Chao at Biomechanics Laboratory, Department of Orthopedics, Mayo ClinicMayo Foundation, Rochester, MN 55905, U.S.A.
238
REMODELING OF PERIOSTEAL NEW BONE
239
Our objective in the current study was to determine the histomorphometric characteristics of periosteal new bone during the sequential stages of fracture repair. The time-related changes in the histomorphometric properties of the periosteal new bone were correlated with the progress of intracortical remodeling. MATERIALS AND METHODS Animal Model Twenty-five 2-month-old male Sprague-Dawley rats (mean weight, 226 g; SD, 7g) were used. Closed standardized fracture of the tibia was produced bilaterally in each animal. The tibia was prenailed with three loose-fitting stainless steel intramedullary pins (diameter, 0.4 mm) and then was fractured manually (2). The fixation was able to maintain axial alignment of the fracture but without interference with bone fragment relative motion at the fracture site. Such intramedullary fixation allowed formation of external fracture callus (l), and immediate weight bearing was allowed after surgery. The animals resumed normal walking within a few days after fracture. An intramuscular injection of oxytetracycline 25 mg/kg was administered at the time of surgery and once a week thereafter. The time between administration of the tetracycline label and sacrifice was 7 days. The animals were sacrificed at a random sequence: five animals at 1 week and four animals each at 2 , 3 , 4 , 5 , and 6 weeks after surgery. Two-plane radiographs obtained immediately after surgery and before sacrifice were used to determine the type of fracture produced, the state of union, and the position of the intramedullary rods (Fig. 1). After sacrifice, the hind legs were disarticulated at the knee joints, and the intramedullary rods were extracted. The first six rats, one animal at each time interval, were used to standardize the experimental methods and the testing condition. The remaining 19 animals were used for the analysis. Preparation of Fracture Callus Specimens After careful dissection of soft tissues, the fracture region of interest (3-mm-thick segment of the midportion of the fracture) was precisely sectioned by a diamond saw. The thickness of the specimens was measured using a micrometer caliber. This fracture segment contained the critical zone of frac-
FIG. 1. Rat tibia1 fractures fixed by intrarnedullary rods united within 4 weeks, as demonstrated radiographically.
ture union process where the advancing front of the anchoring external callus ultimately unite (2). The mineral content of the sectioned specimens was determined by means of micro-bone-densitometry , which is a high-resolution single-photon absorptiometry (12). Micrsbone-densitometry is a custommade, noncommercial device constructed for the analysis of small bone specimens. The device uses a collimated narrow beam (1 m in diameter) of gamma rays from a low-energy isotope source (1251,27.5 keV, 70 mCi) for the precise rectilinear scanning of the specimen. Scanning is performed at 1-s intervals and slow speed (6 mdmin) with small 0.1-mm increments. The count data of the attenuated beam are acquired every 1/2 s. The data processing is performed using a bone edge detection algorithm. The device provides high-resolution computer screen display of the mineral distribution. The specimen was subjected to mechanical testing as reported elsewhere (4). The fracture callus specimens were fixed in 70% alcohol, dehydrated in increasing concentrations of alcohol, defatted in xylene, and embedded in methyl methacrylate to which glycol methacrylate, polyethylene glycol, and dibutyl phthalate (Polysciences, Inc., Warrington, PA, U.S.A.) were added to improve cutting properties and decrease shrinkage. Undecalcified 5-pm transverse sections were cut with a microtome (Reichert-Jung 1140, Reichert Scientific Instruments, Buffalo, NY, U.S.A.) and were stained with a modified Goldner’s trichrome stain (9). The sections were examined under light microscopy. The tissue composition of the
J Orthop Res, Vol. 8, No. 2, 1990
H . T . ARO ET AL.
240
transverse fracture sections was measured using a digitizing tissue-planimetry , as previously described (2). Each histologic section was projected on a digitizing table using magnification of 25 times, and a diagram of the external callus was sketched. The cross-sectional area of the whole periosteal callus and the area of periosteal new bone (including porosity) were measured from the sketches using a modified SLICE program (17). Previous studies of a rabbit fracture model have shown that the restoration of fracture strength and stiffness appear to be related to the amount of new bone connecting the fragments measured from cross sections (5). In the present study, the structural strength of the healing fractures was not determined. The mineral content as determined by microbone-densitometry was expressed as the total amount of bone mineral (BMC). This value represented the mineral content of the whole 3-mm-thick specimen, including the original cortical bone. The normalized density of bone minerals was calculated per unit volume of each specimen and the mineral density of external fracture callus, 6, could be expressed as S=
(BMC) - (BMC)o A x t
where (BMC) = bone mineral content in mg, (BMC), = bone mineral content of 1-week specimen in milligrams, A = area of the specimen scanned in square millimeters, and t = thickness of the specimen in millimeters. This study did not include any sham-operated group. The normalization was performed against the mean mineral content of the 1-week specimens (16.7 k 2.3 mg, n = 4) to give an approximate estimation of the contribution to the measurement of the cortical bone mass. This normalization did not take into account any time-related changes or specimen-related differences in the mineral content of the cortical bone and endosteal new bone. Histologically, the one-week specimens showed new of the cross-sectional callus area. bone in 4% Thus, our estimation of fracture callus mineral density was considered acceptable. Histomorphometric Evaluation Histomorphometric analysis was performed using a semiautomated skeletal histomorphometric technique (Bioquant, R 8z M Biometrics, Inc.,
J Orthop Res, Val. 8, N o . 2 , 1990
Nashville, TN, U.S.A.). The measurements were two-dimensional and were performed on four standardized quarters of the external fracture callus, on the anterior, posterior, medial, and lateral aspect of the tibia. The following static histomorphometric parameters were analyzed. 1. Osteoid surface = (trabecular new bone (or perimeter) surface length covered by osteoid)/(total trabecular new bone surface length) x 100 (%) (trabecular new bone 2. Eroded surface = surface length made up by scalloped resorbing osteoclast-positive or osteoclast-negative lacunae)/(total trabecular new bone surface length) x 100 (%) 3 . Osteoclast number = ( n u m b e r of o s t e o clasts)/(100 mm of trabecular bone surface length) (dl00 mm) 4. Quiescent surface = 100 - (osteoid surface (a derived index) + eroded surface) (%) In these measurements, trabecular new bone referred to periosteal (wovedcancellous) new bone. The results for the four quarters of each specimen were averaged because of the small dimensions involved. Evaluation of Intracortical Remodeling
A 200-pm transverse section was cut with a diamond saw (Buehler Isomet, Lake Bluff, IL, U.S.A.) from the methyl metacrylate block of each undecalcified fracture specimen next to the thin histomorphometric section. The section was ground to 120 p,m and was prepared for contact microradiography performed on a high-resolution contact film (Kodak spectroscopic plates, Eastman Kodak, Rochester, NY, U.S.A.). The unstained sections were viewed under ultraviolet fluorescence microscopy. The intracortical areas of new bone formation, as indicated by the tetracycline labeling, and bone porosity were measured according to the point counting method of Harris and Weinberg (11). New bone formation within osteocyte lacunae was not included in the analyses of new bone formation. The area of osteocyte lacunae was not included in cortical bone porosity calculation.
REMODELING OF PERIOSTEAL NEW BONE
Statistical Analysis For statistical purposes, the experimental unit was one animal. Selection of the fracture for the statistical analysis in each animal was randomized. One fracture healed without substantial external callus, and the contralateral fracture of this animal was used for the analysis. The time-related changes in the different parameters of new bone and cortical remodeling were analyzed using analysis of variance (SAS Software, SAS Institute, Inc., Cary, NC, U.S.A.). If a significant change was observed, Tukey's studentized range test at a confidence level of 95% (p < 0.05) was used to indicate the healing period when the significant change occurred. RESULTS Progress of Fracture Healing Transverse or slightly oblique fractures were obtained at the distal tibiofibular junction. Radiographic union by external callus was observed within 4 weeks (Fig. 1). The total mineral content of the fracture specimens reached the maximum value also by 4 weeks (Table 1). At this stage of healing, nonosseous tissue (fibrous tissue, mineralized and nonmineralized cartilage) constituted a total of 22% of the cross-sectional callus area. The crosssectional callus area decreased significantly (p < 0.05) between 4 and 6 weeks (Table l), indicating the start of external remodeling. The estimated mineral density showed a 2.1-fold increase (p < 0.05) between 4 and 6 weeks (Table 1).
241
woven new bone at 2, 3, and 4 weeks after fracture (Table 2). Neither of these parameters nor the total amount of inactive bone surface showed statistically significant changes with the healing time. There was, however, a tendency to a decreasing extent of eroded surface after 4 weeks. The extent of osteoid surface tended to increase slightly over time. Consequently, the periosteal new bone had significantly more osteoid than eroded surface at 5 and 6 weeks (p < 0.001, two-way analysis of variance). The number of osteoclasts on the periosteal new bone (Figs. 2A and B) was highest at 2 weeks and thereafter significantly decreased with the healing time (p = 0.028) (Table 2). Between 4 and 6 weeks, the periosteal woven new bone maturated into lamellar cancellous bone, which formed an osseous shell around the healing bone (Fig. 3A). The outer surface of the shell showed bone resorption lacunae (Fig. 3B), indicating the initiation of external remodeling processes. The spatial distribution of osteoblastic activity, as indicated by the spatial location of the osteoid, was nonuniform at this stage of healing. Osteoid formation was most prominent in the inner surface of the shell (Fig. 3C).
Histomorphometry The extents of osteoid surface and eroded bone surface did not differ significantly in the periosteal
Intracortical Remodeling Intracortical porosity showed a 3.3-fold increase (p < 0.01) during the first 6 weeks of fracture healing (Table 3). The resorption canals were formed in both the axial and transverse directions of the bone (Fig. 4A). Osteoclasts were occasionally seen to form cutting cones, characteristic of the formation of intracortical resorption canals. The resorption canals were generally lined by a thin layer of new bone (Fig. 4B). Intracortical new bone did not increase significantly with the fracture healing time.
TABLE 1. Cross-sectional callus area, exterior callus new bone area, and bone mineral content of fracture specimens (mean 5 SD, n = 3) Healing time (wks) Cross-sectional callus area (mm*) Periosteal new bone area (mm') (%o)c Bone mineral content (mg) Bone mineral density (mg/mm3 x 10')
2 27.2 i 5.0 4.5 i 2.2 17.4 2 10.8 27.4 i 2.5 12.3 i 1.5
3 35.4 i 6.1 9.7 i 7.7 25.4 2 18.0 39.0 i 12.3 18.6 i 7.7
4 29.0 i 3.6 22.4 i 2.5 78.4 2 16.8 45.9 i 7.4 31.5 2 7.8
5 20.6 2 4.3 20.5 i 4.4 99.6 i 6.7 44.9 i 10.0 43.0 2 5.6
6 12.2 f 3.9b 12.2 i 3.9 100.0 2 0.0 40.6 i 4.3 66.8 i 16.4d
p p
P" 0.0012 = 0.0037
=
NS p =0.0002
" Statistically significant in time-related change (NS
= not significant). The 6-week value significantly lower than the 4-week value (Tukey's studentized range test, p < 0.05). Expressed as a percentage of the cross-sectional area. The 6-week value significantly higher than the 4-week value (Tukey's studentized range test, p < 0.05).
J Orthop Res, Voi. 8, N O . 2 , 1990
242
H . T . ARO ETAL. TABLE 2. Changes in histomorphometric properties of the periosteal new bone in the rat tibia during fracture healing (mean 2 SD, n = 3) Healing time (wks)
Osteoid surface (%) Eroded surface (%) Quiescent surface (%) Osteoclast number
2 19.4 2 17.0 12.5 2 4.7 68.1 i 16.0 123.5 i 60.5
3 20.8 2 10.4 14.3 i 2.1 65.0 2 9.0 64.8 2 48.2
4 28.6 i 16.5 10.5 2 6.8 60.9 i 8.5 28.9 i 44.0
5
34.6 9.1 56.4 9.8
i 4.1** i 3.8 i 2.5 i 12.0
6
Pa
37.2 i 6.0b 4.0 6.8 I 56.0 t 2.0 1.5 2 8.0
NS NS NS p = 0.028
Significance in time-related changes (one-way analysis of variance); NS, not significant. Amount of osteoid surface significantly higher compared with that of eroded surface (Tukey’s test and studentized range test, p < 0.05). a
DISCUSSION Quantitative analysis of bone formation and bone resorption is essential to the evaluation of control mechanisms of the bone repair processes. The current experiment showed that standard histomorphometric methods may give useful information about the cellular activities of internal remodeling processes at the fracture site. For the convenience of discussing the results, bone resorption and formation activities occurring within the callus shell are described as internal remodeling. On the other hand, cellular activities related to the change of cross-sectional areas of callus are defined as external bone remodeling. The extents of osteoid and eroded bone surfaces
were similar in the woven new bone of the external callus in the early phases of fracture healing. This finding supports the concept (24) that the internal remodeling of periosteal new bone starts before fracture union. Previous studies of the same rat fracture model showed that the expression of type I collagen genes, as indicated by quantitative measurement of mRNA levels for type I procollagen chains per unit volume of the callus, reached maximum by 2 weeks and remained stable between 2 and 4 weeks (16). In-situ hybridization revealed a strong labeling with pro l(1) collagen cDNA probe in osteoblasts lining trabeculae of the woven new bone (21). The current histomorphometric evaluation also showed a continuous existance of osteoid surface in the perios-
FIG. 2. A: The remodeling of periosteal new bone included appositional new bone formation on the surface of woven new bone (NB), as indicated by the arrow, the presence of osteoid layer (0s)lined by cuboidal osteoblasts (OBL). A newly entrapped osteocyte (OST) is visible within the bone trabeculae. B: Multinucleated osteoclasts (OCL) were commonly observed in the immediate vicinity of osteoblasts (OBL) lining a osteoid layer (0s).Note the eroded lacunae of the bone surface (ES) next to the osteoclast. (Two weeks after fracture, histologic sections stained with Goldner’s stain, original magnification ~250.)
J Orthop Res, Vol. 8, No. 2 , 1990
REMODELING OF PERIOSTEAL NEW BONE
243
C
FIG. 3. A: Periosteal new bone formed a thick outer shell around the fracture site of the tibia during the remodeling phase of healing. OCB, original cortical bone; HRL, Howship’s resorption lacunae; NB, new bone; BFS, bone formation surface; f, fibula (microradiograph, 6 weeks after fracture, original magnification x25). B:The outer surface of the bony callus was rough in certain regions because of the formation of Howship’s resorption lacunae (HRL). These lacunae occasionally contained osteoclasts in histologic sections. Note that the corresponding inner surface of the shell contained bone formation surface (BFS) that appeared smooth; NB, new bone; OCB, original cortical bone (microradiograph, 5 weeks after fracture, original magnification x50). C: The mature cancellous lamellar bone of the callus shell showed thick osteoid (0s) layer in its inner surface indicating the progress of the closure of the space between the shell and the original cortical bone; NB, new bone; OBL, osteoblasts (histologic section stained with Goldner’s stain, 5 weeks after fracture, original magnification ~ 5 0 ) .
J Orthop Res, Vol. 8,No. 2,1990
H . T . ARO ET AL.
244
TABLE 3 . Intracortical remodeling of the rat tibia during fracture healing (mean -t SO)" Healing time (wks) 1
2
3
4
5
Pb
6 ~~
Original cortical bone (unlabeled) (%) Intracortical new bone (%) Intracortical porosity (%) a
87.8 i 2.1 8.8 t 1.4 3.4 i 0.8
82.0 i 5.5 14.3 i 4.7 2.7 2 1.2
81.8 i 4.9 11.7 i 3.0 6.5 i 1.9
77.8 2 4.0 14.2 i 2.0 8.0 i 2.5
78.2 i 6.0 13.0 i 1.9 8.9 2 4.1
74.4 2 5.5 14.1 2 0.9 11.5 i 4.6
p
< 0.025 NS
p < 0.01
n = 3 except for the 1-week group (n = 4). Significance in time-related change (one-way analysis of variance); NS, not significant.
teal new bone. These results showed that the internal remodeling of periosteal new bone involves a considerable steady collagenous bone matrix production by osteoblasts. After the early phases of fracture healing, the periosteal bone had more osteoid than eroded surface. Theoretically, this change should result in increased density of the periosteal new bone because
of the continuous new bone formation in the presence of decreased bone resorption. Absorptiometry techniques are accurate in the measurements of bone minerals (15,28). In the current study, the periosteal new bone showed a significant increase in the density of bone minerals during fracture healing. This result confirmed that the observed change in the balance between appositional formation of
FIG. 4. lntracortical remodeling of the rat tibia during fracture healing. A: Longitudinal and transverse resorption canals (arrows) developed in cortical bone (3 weeks after fracture, microradiograph original magnification x50). B: The examination of the same specimen under ultraviolet microscopy (new bone labeled by tetracycline white) showed that these resorption canals were partly covered by a thin layer of new bone (arrows). PNB, periosteal new bone; ENB, endosteal new bon (original magnification x50).
J Orthop Res, Vol. 8, N o . 2 , 1990
REMODELING OF PERIOSTEAL NEW BONE
lamellar bone on the surface of woven bone and resorption of woven bone trabeculae had a positive net effect on bone density. It should be emphasized that the present experiment included only an approximate estimation of callus mineral density because both the effects of intracortical bone resorption and endosteal new bone formation could not be determined precisely. Cellular activities of the external remodeling processes of the periosteal new bone were not quantitated in this study. However, the decrease in crosssectional area of the callus and the qualitative evidence of bone resorption showed that external remodeling started at 4 weeks. This external remodeling resulting in an obvious decrease of callus volume, providing the explanation of why the absolute mineral mass of the periosteal new bone did not show a continuous increase during fracture healing. It seems that external remodeling is a response to the increasing density and maturation of the periosteal new bone after fracture union. Osteoclasts are of monocytic-macrophagic origin and invade the fracture site of the rat tibia at an early stage of healing (10). The close presence of multinucleated osteoclasts to osteoblasts lining the osteoid layer suggests an interaction between these two cell types during the early stages of fracture healing. Osteoclasts have been shown to exhibit expression of mRNAs for a transforming growth factor-p (TGF-P) and cellular protooncogene c-fos during endochondral ossification processes of human growth plates (21,22). Both TGF-p and protooncogene c-fos have been implicated to have important roles in the metabolism of chondrocytes and osteoblasts (8,20,25). All these data and the current study suggest that osteoclasts could have a function in the control of endochondral fracture healing processes. Rat cortical bone does not possess a haversian system. However, the rat appears to develop a marked remodeling process of the cortical bone during fracture healing. This process seems to take part in the revascularization process of the fracture ends. In the current model, the radial growth of penetrating blood vessels from the medullary cavity to the external callus may have contributed to the observed intracortical porosity. Intracortical bone porosity increase may also have been a sequential change in response to load by passing through periosteal new bone formation. It is unlikely that the fracture fixation method used (loose-fit or flexible
245
intramedullary wires) could cause bone stressprotection, thereby leading to osteopenia. The amount of intracortical new bone did not increase significantly during the experimental period, indicating a lag period between intracortical resorption and new bone formation. This finding is compatible with the theory of Charnley (6). He proposed that the bridging mechanism of periosteal callus is to pass the necrotic fracture ends before the slow start of cortical healing. Ultimately, the cortical reconstruction by secondary osteons seems to be important to regain the ultimate bone union strength (3,7,23). Factors that control internal and external bone remodeling processes of the external fracture callus are not completely known. The remodeling seems to be, at least after fracture union, related to the mechanical factors as dictated by the functional demands according to Wolff's law. Fracture union constitutes, histologically, the establishment of a bony bridge between the fracture fragments and, biomechanically , the restoration of the hard-tissue characteristics of the bone. The current experiment showed that the progress of fracture healing is associated with a change in the balance of periosteal new bone formation and resorption. This change coincided with the time of callus ossification and the start of external remodeling, suggesting a relationship with the restoration of bony continuity. However, further studies are needed to exclude a similar time-related change of cell activities in fractures that fail to unite. These results may help further research to optimize the biomechanical properties of fracture fixation devices. Acknowledgment: The authors thank Jan Palmer for her collaboration, and Glenda Evans, Michael Bateman, and James Bronk for their technical assistance.
REFERENCES Aro H, Eerola E, Aho AJ: Osteolysis after rigid fixation. The possible role of periosteal neural mechanoreceptors in bone remodeling. Clin Orthop 166:292-300, 1982 Aro H, Eerola E, Aho AJ: Determination of callus quantity in 4-week-old fractures of the rat tibia. J Orthop Res 3: 101108, 1985 Aro HT, Kelly PJ, Lewallen DG, Chao EYS: Comparison of the effects of dynamization of constant rigid fixation on rate and quality of bone osteotomy union in external fixation. Orthop Trans 12:378, 1988 Aro HT, Wippermann BW, Hodgson SF, Wahner HW, Lewallen DG, Chao EYS: Noninvasive monitoring of fracture callus mineralization using high-resolution single-photon absorptiometry. Orthop Trans 12:489, 1988
J Orthop Res. Vol. 8, No. 2, 1990
246
H . T. ARO ETAL.
5. Black J, Perdigon P, Brown N, Pollack SR: Stiffness and strength of fracture callus. Relative rates of mechanical maturation as evaluated by a uniaxial tensile test. Clin Orthop 182:278-287, 1984 6. Chamley J: The Closed Treatment of Common Fractures, 3rd ed. Edinburgh, London, New York, Churchill Livingstone, 1974, p 1-42 7. Claes L, Bum C, Gerngross H, Mutschler W: Bone healing stimulated by plasma factor XIII. Osteotomy experiments in sheep. Acta Orthop Scand 5657-62, 1985 8 . Ellisgsworth LR, Brennan JE, Fok K, Rosen DM, Bentz H, Piez KA, Seydin SM: Antibodies to the N-terminal portion of cartilage-inducing factor A and transforming growth factor p. J Biol Chem 261:12362-12367, 1986 9. Goldner J: A modification of the Masson trichrome technique for routine laboratory purposes. Am J Pathol 14:237243, 1938 10. Gothlin G, Ericsson JLE: The osteoclast. Review of ultrastructure, origin, and structure-function relationship. Clin Orthop 120:201-231, 1976 11. Harris WH, Weinberg EH: Microscopic method of measuring increases in cortical bone volume and mass. Calcif Tissue Res 8:19&196, 1972 12. Hodgson SF, Wahner HW, Dunn WL, Bateman MD: Micro bone densitometry of human transiliac bone biopsies: a new approach to the direct measurement of skeletal mass in man. Bone Mineral Measurements by Photon Absorptiometry: Methodological Problems, ed by J Dequeker, P Geusens, HW Wahner, Leuven, Belgium, Leuven University Press, 1988, p 461 13. Jowsey J, Kelly PJ, Riggs BL, Bianco AJ Jr, Scholz DA, Gershon-Cohen J: Quantitative microradiographic studies of normal and osteoporotic bone. J Bone Joint Surg [Am] 47:785-806, 1965 14. Kelly PJ, Peterson LFA, Janes JM: A method of using sections of bone prepared for microradiography for subsequent histologic study. Proc Staff Meet Mayo Clin 34:274-283, 1959 15. Mazess RB: The noninvasive measurement of skeletal mass. In: Bone and Mineral Research, by WA Peck, Amsterdam, Oxford, Princeton, Excerpta Medica, 1983, pp 223-279 16. Multimaki P, Aro H, Vuorio E: Differential expression of fibrillar collagen genes during callus formation. Biochem Biophys Res Commun 142:536-541, 1987 17. Nagurka ML, Hayes WC: An interactive graphics package
J
Orthop Res, Vol. 8, NO. 2, 1990
for calculating cross-sectional properties of complex shapes. J Biomech 135944, 1980 18. Parftt AM, Drezner MK, Glorieux FH, Kanis JA, Malluche H, Meunier PJ, Ott SM, Recker RR: Bone histomorphometry: standardization of nomenclature, symbols, and units. J Bone Mineral Res 2595410, 1987 19. Rand JA, An KN, Chao EYS, Kelly PJ: A comparison of the effect of open intramedullary nailing and compression plate fixation on fracture-site blood flow and fracture union. J Bone Joint Surg [Am] 63:427442, 1981 20. Ruther U , Garber C, Komitowski D, Muller R, Wagner EF: Deregulated c-fos expression interferes with normal bone development in transgenic mice. Nature 325:412416, 1987 21. Sandberg M, Aro H, Multimaki P, Aho H, Vuorio, E: In situ localization of collagen production by chondrocytes and osteoblasts in fracture callus. J Bone Joint Surg 71:69-77, 1989 22. Sandberg M, Vuorio T, Hirvonen H, Alitalo K, Vuorio E: Enhanced expression of TGF-a and c-fos mRNAs in the growth plates of developing human long bones. Development 102:461-470, 1988 23. Schenk RK: Histophysiology of bone remodeling and bone repair. In: Perspectives on Biomaterials, ed by OCC Lin, EYS Chao, Amsterdam, Elsevier Science Publishers B.V., 1986, pp 75-99 24. Sevitt S : Bone Repair and Fracture Healing in Man. Edinburgh, London, Melbourne and New York, Churchill Livingstone, 1981 25. Seyedin SM, Thomas TC, Thompson AY, Rosen DM, Piez KA: Purification and characterization of two cartilageinducing factors from bovine demineralized bone. Proc Natl Acad Sci USA 82:2267-2271, 1985 26. Shih M-S, Norrdin RW: Effect of prostaglandin El on the periosteal regional acceleratory phenomenon in fractured ribs: histomorphometry study in beagles. A m J Vet Res 48:828-830, 1987 27. Vanderhoeft PJ, Kelly PJ, Peterson LFA: Determination of growth rates in canine bone by means of tetracycline-labeled patterns. Lab Invest 11:714-726, 1962 28. Wahner HW, Riggs BL: Methods and application of bone densitometry in clinical diagnosis. CRC Crit Rev Clin Lab Sci 24:3:217-233, 1986 29. Wu J-J, Shyr HS, Chao EYS, Kelly PJ: Comparison of osteotomy healing under external fixation devices with different stiffness characteristics. J Bone Joint Surg [ A m ] 66:1258-1264, 1984
