- Email: [email protected]
A microradiographic investigation of cancellous bone healing after irradiation and hyperbaric oxygenation: a rabbit study
Int. J. Radiation Oncology Biol. Phys., Vol. 48, No. 2, pp. 555–563, 2000 Copyright © 2000 Elsevier Science Inc. Printed in the USA. All rights reserved 0360-3016/00/$–see front matter
PII S0360-3016(00)00638-6
BIOLOGY CONTRIBUTION
A MICRORADIOGRAPHIC INVESTIGATION OF CANCELLOUS BONE HEALING AFTER IRRADIATION AND HYPERBARIC OXYGENATION: A RABBIT STUDY ÅSE A. JOHNSSON, M.D.,*† MAGNUS JACOBSSON, M.D., PH.D.,† GO¨ STA GRANSTRO¨ M, M.D., D.D.S., PH.D.,‡ CARINA B. JOHANSSON, PH.D.,† KARL-GUSTAV STRID, M.SC.(ENG.), PH.D.,§ AND INGELA TURESSON, M.D., PH.D.㛳 Departments of *Radiology, †Biomaterials/Handicap Research, and ‡Otolaryngology, Head and Neck Surgery, §Biomaterials Research Group, Institute of Anatomy and Cell Biology, Sahlgrenska University Hospital, Go¨teborg University, Gothenburg Sweden; and 㛳 Department of Oncology, Academic Hospital, Uppsala, Sweden Purpose: To analyze the effect of irradiation on cancellous bone healing at different times after irradiation and to study if hyperbaric oxygen therapy (HBO) would affect the bone healing capacity, when delivered directly after irradiation. Methods and Materials: Rabbits were given a single dose of 15 Gy 60Co radiation to one hind leg, the other hind leg serving as control. A standardized defect through the femoral metaphysis of the rabbits was created by a trephine drill biopsy at different times after irradiation. New bone formation in the defect was evaluated by a new biopsy through the previous defect after a healing time of 8 weeks. The mineral contents of the biopsies were analyzed by microradiography and microdensitometry. Results: There was a large variation in the bone-forming capacity expressed as bone mineral content between the animals. No statistically significant differences could be detected regarding the effect of irradiation, HBO, or delayed surgery. Qualitative histology revealed more pronounced inflammation, fibrosis, and bone resorption in the irradiated bone. Conclusions: No definite conclusions can be drawn from the results of this study, however it might be hypothesized that cancellous bone recovers faster than cortical bone from radiation trauma. © 2000 Elsevier Science Inc. Cancellous bone healing, Hyperbaric oxygen treatment, Irradiation, Microdensitometry, Microradiography.
have been studied by Nilsson et al. (10) and Larsen et al. (11). Both authors stated beneficial effects of HBO on bone healing, and Larsen et al. (11) showed that the use of HBO improved implant integration in irradiated bone tissue. In a recent study, Chen et al. (12) reported that HBO might have different effects on irradiated and nonirradiated bone. The aim of the present study was to evaluate the boneforming capacity in predominantly cancellous bone, at different times after irradiation, and to study the effect of HBO on cancellous bone healing with and without preceding irradiation.
INTRODUCTION Radiotherapy in combination with surgery is the treatment generally used for malignant tumors in the craniofacial region. Several experimental investigations (1– 6) have studied the effects of irradiation on cortical bone healing. The damaging effect from radiotherapy on bone tissue is believed to be related to direct damage to the osteogenic cells (1, 3); however, vascular injury is also considered important (3). The ratio of cortical to cancellous bone varies in different parts of the craniofacial area. The modes and the speed of healing of cortical and cancellous bone differ, and cortical healing lags behind cancellous bone healing (7). However, the injury of irradiation to cells may be greater in the highly cellular cancellous bone (8), since cancellous bone usually has a higher rate of metabolic activity (9). The effects of hyperbaric oxygenation (HBO) on cortical bone healing
METHODS AND MATERIALS The following study design was approved by the Go¨teborg University Laboratory Animal Ethics Committee, Sweden. burg Medical Association, the Eivind and Elsa K:son Sylvan Foundation, the Assar Gabrielsson Foundation, the Go¨ranssons Foundation, the G. A. E. Nilsson Foundation, the A. A. Yhle´n Foundation, the King Gustaf V Jubilee Foundation, and the Swedish Medical Research Council (MFR). Accepted for publication 17 April 2000.
Reprint requests to: Dr. Åse A. Johnsson, Department of Biomaterials/Handicap Research, Institute for Surgical Sciences, Box 412, SE-405 30 Go¨teborg, Sweden. Acknowledgments—The following funds, here gratefully acknowledged, have supported the present study: the Hjalmar Svensson Foundation, the Greta and Einar Asker Foundation, the Gothen555
556
I. J. Radiation Oncology
●
Biology
●
Physics
Volume 48, Number 2, 2000
Table 1. Number of animals in the different groups Animals evaluated/ animals included Group A: Biopsies directly and 8 wk after irradiation 8/8 Group B: Biopsies directly and 8 wk after irradiation ⫹ HBO 6/10 Group C: Biopsies at 12 wk and 20 wk after irradiation 8/11 Group D: Biopsies at 52 wk and 60 wt after irradiation 8/12 Group B: One animal had to be sacrificed because of a diaphyseal fracture of the irradiated leg, two animals died after the last HBO treatment of unknown reasons, and one animal had to be sacrificed because of a spinal lesion secondary to seizures at the HBO treatment. Group C: Two animals had to be sacrificed because of fractures of the irradiated leg and one animal could not be evaluated due to preparation failure. Group D: One animal expired before the time of surgery, two animals had to be sacrificed because of fractures of the irradiated leg, and one animal had to be sacrificed because of a pulmonary infection.
Animals Adult (over 9 months old) New Zealand white rabbits were used in the study. The rabbits were group housed, the animals being females and neutered males. All animals received a single dose of 15 Gy 60Co radiation to one hind leg, the other hind leg serving as control. A cylindrical defect, 4.5 mm in diameter, through each femoral metaphysis was created immediately (group A and B), at 12 weeks (group C) and at 52 weeks (group D) after irradiation. Group B received HBO treatment during the first 4 postoperative weeks. After a healing time of 8 weeks a biopsy of the bone formed in the created defect was taken. Twelve animals were used to develop the surgical procedure. Forty-one animals received irradiation; 30 of them could be evaluated in the study. Five animals had to be sacrificed because of fractures of the irradiated leg. Five animals expired prior to evaluation and one animal could not be evaluated due to preparation failure. The number of animals in each group is presented in Table 1. Anesthetic procedure Intramuscular injections of fentanyl and fluanizon (Hypnorm, Janssen-Cilag, Buckinghamshire, England) at a dose of 0.5 ml/kg body weight and intraperitoneal injections of diazepam (Stesolid Novum, A/S Dumex Denmark, Pharmacia-Upjohn Stockholm, Sweden) at a dose of 1.5 mg/kg body weight were used for general anesthesia during irradiation and surgical procedures but not during HBO treatment. Local anesthesia with 1.0 ml of 5% lidocaine (Xylocaine, Astra, Mo¨lndal, Sweden) was administered to the femora prior to surgery. The shaved skin of the rabbits was washed with a mixture of iodine and 70% ethanol prior to surgery. In connection with surgery the animals received an antibiotic, benzyl-penicillin (PenoVet, Boeringer Ingelheim Agrovet, Hellerup, Denmark) at a dose of 20 mg/kg body weight. Postoperatively intramuscular injections of buprenorfin (Temgesic, Reckitt & Colman, Hull, England) were given in doses of 0.1 mg/kg body weight. At the time of sacrifice intravenous injections of a mixture of saline and barbiturates (Mebumal Vet., 60 mg/ml, Nord Vacc, Sweden) 1:4 were given.
Irradiation procedure The rabbits were irradiated to the distal femoral metaphysis and the proximal tibial metaphysis of the right hind leg, the left hind leg serving as control. During irradiation the femur and tibia were placed on a 5-cm-thick polystyrene phantom. Gamma irradiation (60Co) was used to minimize the difference in the absorbed dose in soft tissue and bone. Source-to-skin distance was 60 cm and field size was 7.5 ⫻ 7.5 cm. The dose rate was approximately 0.5 Gy per minute. A 5-mm bolus was applied to ensure full build-up, and 15 Gy were given as a single dose. Hyperbaric oxygen treatment On the third postoperative day (Monday) the animals in group B were placed in a pressure chamber of 75 liters volume (Gothenburg Diving Technique, Gothenburg, Sweden) and subjected to pure oxygen of 280 kPa for 2 hours. During this period, the first 10 minutes were used for successive compression up to 280 kPa; the pressure was kept constant for 90 minutes and followed by decompression for 20 minutes. The chamber temperature was kept at 23°C by a water cooling system. Produced carbon dioxide was eliminated by constant flow of oxygen with a flow rate of 1.5 L/min. The HBO treatment was performed once daily Monday to Friday and a total of 20 treatments was given. Guide and implant design In order to take standardized biopsies through the femoral metaphysis, a biopsy-guide-system was constructed by Wennbergs Finmek, Angered, Sweden. Illustrations of the guide and the trephine drill biopsy procedure are given in Figs. 1 a–1d. The surface to be placed in contact with the femur was modeled and slightly curved, making possible a close fit to the distal femoral diaphysis and metaphysis. There were four drill holes (canals) through the guide: one proximal, 4.5 mm in diameter, two oblique 4.55 mm in diameter, with a common opening on the surface facing toward the femoral metaphysis, and one distal, 1.6 mm in diameter. In order to guide the trephine drill through the femoral metaphysis, one of the two oblique canals was used for the
Cancellous bone healing after irradiation and hyperbaric oxygenation
Fig. 1. (a) The biopsy guide with its oblique canals (*), distal drill hole (∧), and proximal canal (#) used for the landmark fixture seen from above. (b) The modeled surface of the biopsy guide facing the medial aspect of the femur with the 1.6-mm drill (∧) and the 2.2-mm drill with connector (#) used to center the drill hole for the landmark fixture. The common opening of the two oblique canals (*) is placed between the drills. (c) The biopsy guide with the 1.6-mm drill (∧), the trephine drill, (*) and the landmark fixture (#) . (d) At the second biopsy the distal part of the femur was dissected free from soft tissue. The guide was matched to the previous defect by the landmark fixture and the 1.6-mm drill. During drilling the cortex of the lateral femoral metaphysis was inspected in order to ensure that the trephine drill did not deviate from the margins of the previous defect.
trephine drill, the other for saline cooling. The outer diameter of the trephine drill used was 4.5 mm and the inner diameter was 3.5 mm. A self-tapping, stainless steel implant (Wennbergs Finmek, Angered, Sweden) was used to stabilize the proximal part of the guide and to serve as a landmark for the second biopsy. The implant had a total length of 8.2 mm; the threaded part inserted in the femoral diaphysis was 5 mm in length, and the diameter was 2.4 mm. The top of the fixture was cylindrical, 4.5 mm in diameter, with an internal hexagon to fit a key used during insertion. Surgical procedure Under aseptic conditions a longitudinal incision was made through the skin and the muscle fascia of the distal medial side of the femur. Muscles were gently mobilized along the medial aspect of the femur. Parts of the medial patellar retinaculum were divided in order to displace the patellar tendon laterally. The periosteum was left intact except for the drilling sites, where it was incised and gently pushed aside. The guide described above was attached to the distal femoral metaphysis by a 1.6-mm drill. To stabilize the proximal part of the guide, a landmark fixture was inserted through the guide into the medial aspect of the femoral diaphysis. When the self-tapping implant had been inserted, its cylindrical top exactly fit the proximal canal and thus stabilized the guide. A 3.5-mm biopsy was taken from the
●
Å. A. JOHNSSON et al.
557
medial aspect of the femoral metaphysis through the medullary cavity and out through the lateral cortex of the femoral metaphysis. Ample cooling with saline was provided during drilling. After removal of the biopsy, the distal drill and the guide itself were removed, leaving a defect of 4.5 mm diameter through the distal femoral metaphysis and the 1.6-mm distal drillhole to heal. The landmark fixture was left in place. The 4.5-mm defect was irrigated and inspected in order to check that no visible bone was left in the defect. The patellar retinaculum and the fascia were closed with interrupted resorbable sutures and the skin with interrupted nonabsorbable sutures. Eight weeks after the primary operation, the animals were again anesthetized as described above and the distal part of the femur was dissected free from soft tissue. The guide was matched to the previous defect by the landmark fixture and the 1.6-mm drill. The created defect had healed, but it was still possible to recognize its margins. A new biopsy with the same diameter as the first one was taken through the guide and the newly formed bone in the previous defect (Fig. 1d). During drilling ample cooling with saline was provided and the cortex of the lateral femoral metaphysis was inspected in order to ensure that the trephine drill did not deviate from the margins of the previous defect. Finally the animals were sacrificed by intravenous injections of barbiturates. All biopsies were directly immersed in 4% neutral buffered formaldehyde and stored at 4°C. Microradiography The biopsies from the same animal, i.e., four specimens, were microradiographed simultaneously together with a density reference consisting of an aluminum step wedge on a Kodak high-resolution plate type 1A. The aluminum step wedge had 10 steps, each measuring 5 ⫻ 2.5 mm, and the height of each step was 0.3 mm, yielding a maximum wedge thickness of 3.0 mm. A Machlett roentgen tube, type OEG-50 with a copper target and a focal spot of 1 mm, generated a polyenergetic spectrum, which was prefiltered through 0.5 mm aluminum. The tube focus was fixed at 242 mm from the recording surface. Exposures were carried out at 27 kV and 20 mA for 40 minutes. The plates were processed immediately following exposure by development in fresh Kodak D-19 solution for 6 min at 20°C during standardized agitation and subsequent rinsing, fixing, washing and drying as proposed by the manufacturer. Strid and Ka¨lebo (13) have given a detailed description of the microradiographic procedure. Figure 2a demonstrates a survey microradiograph of a biopsy. Bone mass determination The microradiographs were digitized using a standard 35-mm slide scanner (Nikon CoolScan LS-10E). No contrast enhancement or other automatic image modification was performed during the scanning procedure. The microradiographs were scanned at a resolution of 106 pixels per mm. To cover the entire slide, each microradiograph was scanned in two portions, the step wedge being included in
558
I. J. Radiation Oncology
●
Biology
●
Physics
Volume 48, Number 2, 2000
Histology After microradiography, samples of newly formed bone from half of the animals in each group, and from two animals that were sacrificed prior to evaluation time, were decalcified in 16% formic acid and embedded in paraffin. Two, 3-m-thick, longitudinal sections were cut through the central part of each sample. The sections were stained with hematoxylin– eosin prior to light microscopic observations (Fig. 2b).
Fig. 2. (a) A survey picture of a microradiographed biopsy. (b) A survey picture of a decalcified biopsy stained with hematoxylin– eosin.
both scans in order to ensure that both scans showed the same density profile. The images were stored in TIFF format. The tagged image file format (TIFF) images were analyzed by a de facto standard image analysis program (NIH Image v. 1.62, National Institutes of Health). Quantitative analysis was performed using a custom program (AluDef 1.2) developed in-house using LabVIEW 5.0 (National Instruments) on an Apple Power Macintosh G3 computer. A detailed description of the procedure and assessment of the method is given by Strid et al. (14). The mineral content of the specimens was calculated in terms of aluminum equivalent mass.
Statistical methods Fisher’s test for pair comparisons was applied in assessing the effect of irradiation on bone formation. The effect of HBO and delayed surgery after irradiation was evaluated by Fisher’s permutation test (15) for irradiated as well as nonirradiated bone tissue. The numbers of animals in the different groups were based on results from previous studies (16 –18). RESULTS To express the bone-forming capacity (in terms of mineralization) in irradiated and nonirradiated bone tissue, a ratio between the mineral content of the newly formed bone in the biopsy, taken after a healing time of 8 weeks, and the mineral content of the biopsy taken to create the defect, was formed. A ratio of 1.0 would indicate that the same amount of mineralized bone tissue had formed after 8 weeks in the created defect as was present in the original biopsy. The results are presented in Table 2 and Fig. 3. There was a large variation in the bone-forming capacity between the animals. No statistically significant differences could be detected regarding the effect of irradiation, HBO, or delayed
Table 2. The bone-forming capacity expressed as a ratio between the mineral content of the newly formed bone in the biopsy, taken after a healing time of 8 weeks, and the mineral content of the biopsy taken to create the defect Group A Biopsy directly after irradiation
Group B Biopsy directly after irradiation HBO-treated animals
1 2 3 4 5 6 7 8
Non-irr
1.02 0.78 0.84 1.12 1.48 1.48 0.66 0.78 M 1.02 SD 0.32
Irr
Animal No.
Non-irr
Irr
0.62 0.66 0.79 0.96 1.61 0.96 0.96 1.26 0.98 0.33
1 2 3 4 5 6
0.83 0.78 1.45 0.91 1.72 1.33
0.88 0.88 0.99 0.87 0.9 1.6
Two animals expired 4 wk after surgery
1.17 0.39 0.69 1.05
1.02 0.29 0.62 0.7
Group D Biopsy 52 wk after irradiation Ratio of mineral content 8 wk biopsy/original biopsy
Ratio of mineral content 8 wk biopsy/original biopsy
Ratio of mineral content 8 wk biopsy/original biopsy
Ratio of mineral content 8 wk biopsy/original biopsy Animal No.
Group C Biopsy 12 wk after irradiation
Animal No. 1 2 3 4 5 6 7 8
Non-irr
Irr
Animal No.
Non-irr
Irr
0.43 0.73 1.23 0.89 0.85 0.98 0.75 1.07 0.87 0.24
0.77 0.61 0.87 0.78 0.94 0.46 1.08 0.35 0.73 0.25
1 2 3 4 5 6 7 8
0.57 1.16 0.73 0.74 0.69 0.8 0.67 0.69 0.76 0.18
0.55 1.13 0.83 1.02 0.6 0.91 0.64 0.76 0.8 0.21
0.82
0.82
One animal expired 6 wk after surgery
Cancellous bone healing after irradiation and hyperbaric oxygenation
●
Å. A. JOHNSSON et al.
559
Fig. 3. A diagram showing the results of the densitometric measurements as mean values of the ratios between the mineral content of the newly formed bone in the created defect and the mineral content of the biopsy taken to create the defect. There were no significant differences. Standard deviations (SD) are marked.
surgery on bone-forming capacity in terms of mineralization. However, there was a tendency toward reduced bone formation at 12 and 52 weeks after irradiation in both irradiated and nonirradiated bone tissue and a tendency toward improved bone formation after HBO in nonirradiated bone tissue. The samples that were evaluated histologically were examined in a blinded manner. The amount of fibrous tissue, inflammatory cells in the bone marrow, ongoing resorption, osteoclasts, osteoblasts, and blood vessels were categorized as abundant, moderately present, rare, or none for each sample judged. The results are summarized in Table 3. General qualitative observations on the decalcified hematoxylin– eosin-stained paraffin sections were also made in a blinded manner and revealed the following: Group A (directly after surgery, 8 weeks after irradiation). The irradiated test side revealed a higher amount of dense (fibrous) connective tissue in the bone marrow as compared to the nonirradiated control side, where a looser fibrous tissue dominated. The number of inflammatory cells was greater on the test side. Bone flakes were observed on both the test and control sections, seemingly more pronounced on the irradiated side. The bone flakes in the irradiated biopsies seemed to be acellular/pycnotic, whereas in the sections from nonirradiated bone, bone cells could be observed. A more active remodeling appeared to be present in the control biopsies. Large cores of woven/chondroid bone tissue were observed both in the test and in the control samples. However, the bone tissue structure itself appeared maturer in the latter. The mentioned cores had a morphologic appearance similar to chondroid tissue in the light
microscope (Figs. 4 and 5). However, with the aid of polarized light it was possible to detect cross-linked fibers in the tissue. Therefore we chose to call these cores “woven/ chondroid bone tissue.” Group B (directly after surgery, 8 weeks after irradiation and HBO treatment). The amount of fibrous tissue and inflammatory cells was greater on the irradiated side as compared to the nonirradiated biopsies. Moreover, the fat cells appeared more regular and the number of these cells was greater in the nonirradiated samples. There was a tendency of a higher degree of bone resorption on the irradiated test side. Often large cores of woven/chondroid bone tissue could be observed on the test and control side but seemingly larger cores were present on the former samples. Group C (8 weeks after surgery, 20 weeks after irradiation). An organized capsule-like fibrous tissue was more conspicuous on the irradiated side as compared to the control side. The number of fat cells seemed to be greater in the nonirradiated control biopsies, whereas inflammatory cells were more abundant in the test samples. There was a tendency of a more active bone remodeling in the irradiated biopsies compared to the control side. Group D (8 weeks after surgery, 60 weeks after irradiation). A loose connective tissue was observed in the nonirradiated biopsies as compared to a denser structure in the irradiated samples. The number of inflammatory cells was greater on the irradiated side and the amount of fat cells was less on this side in comparison to nonirradiated control biopsies. The bone tissue resorption was more pronounced and gave a more active impression on the irradiated side as compared to the nonirradiated samples.
560
I. J. Radiation Oncology
●
Biology
●
Physics
Volume 48, Number 2, 2000
Table 3. Histologic Evaluation* Group† A: Irr
A: N-irr
Fibrous tissue Abundant 1 0 Moderate 3 2 Rare 0 1 None 0 1 Inflammatory cells in the bone marrow Abundant 0 0 Moderate 4 0 Rare 0 2 None 0 2 Ongoing resorption Abundant 2 0 Moderate 1 3 Rare 1 1 None 0 0 Osteoclasts Abundant 2 0 Moderate 2 2 Rare 0 2 None 0 0 Osteoblasts Abundant 0 0 Moderate 3 4 Rare 1 0 None 0 0 Blood vessels Abundant 0 0 Moderate 4 4 Rare 0 0 None 0 0
B: Irr
B: N-irr
C: Irr
C: N-irr
D: Irr
D: N-irr
1 1 1 0
0 1 1 1
1 2 0 1
1 2 1 0
0 1 3 0
0 3 0 1
0 1 2 0
0 0 1 2
1 1 2 0
0 0 4 0
0 2 2 0
0 1 3 0
0 3 0 0
0 0 3 0
2 1 1 0
0 4 0 0
1 3 0 0
0 2 2 0
0 3 0 0
0 0 3 0
1 3 0 0
0 4 0 0
3 1 0 0
0 2 2 0
0 1 2 0
0 2 1 0
0 2 2 0
0 3 1 0
0 4 0 0
0 2 2 0
0 3 0 0
0 3 0 0
0 4 0 0
0 4 0 0
1 3 0 0
0 4 0 0
* The samples that were evaluated histologically were examined in a blinded manner. The parameters analyzed were: fibrous tissue, inflammatory cells in the bone marrow, ongoing resorption, osteoclasts, osteoblasts, and blood vessels. For each sample judged, the different parameters were categorized as abundant, moderate, rare, or none. (0 – 4) ⫽ number of samples in each category. Irr ⫽ irradiated, N-irr ⫽ nonirradiated. † Group A: Surgery directly after irradiation: 4 animals. Group B: Surgery directly after irradiation and HBO treatment: 3 animals. Group C: Surgery at 12 weeks after irradiation: 4 animals. Group D: Surgery at 52 after irradiation: 4 animals.
In summary, in all samples there were signs of both endochondral and intramembranous ossification. Cores of woven/chondroid bone tissue were present in many biopsies but more pronounced in irradiated bone. Generally there was more ongoing resorption and osteoclasts present in the irradiated side compared to the nonirradiated side. In the bone marrow there was a more pronounced inflammatory reaction in the irradiated side compared to the nonirradiated side in all groups. This finding was most conspicuous in the group that had surgery directly after irradiation. There was more fibrous tissue in the groups that had surgery directly after irradiation both with and without HBO. However, there were no discernable differences regarding the number of blood vessels. In one animal that expired after the last HBO treatment (4 weeks healing time) there was more fibrous tissue, osteoclasts, and bone resorption in the irradiated side compared to the nonirradiated side. One animal in group D (surgery 52 weeks after irradiation) that had to be sacrificed 6 weeks after surgery (because of a pulmonary
infection) showed no such differences. Examples of histologic findings are presented in Figs. 4 –7. DISCUSSION Shapiro (19) described the healing of cortical 2.4-mm diaphyseal defects in the rabbit tibia and femur. The initial source of repair tissue was the marrow. Vessels grew into the defect accompanied by mesenchymal cells. Bone formation was observed after 1 week, as woven bone at the endosteal surface at the periphery of the defects. After 2 weeks cores of woven bone were seen in the central parts of the defect. On the surface of these cores, osteoblasts formed osteoid in a lamellar manner. After 4 weeks only few cores of woven bone were observed. Few resorptive osteoclasts were observed during the healing. Haversian systems were beginning to transverse the peripheral parts of the newly formed repair bone after 6 – 8 weeks. Markel et al. (20) studied the reparative tissue and bone
Cancellous bone healing after irradiation and hyperbaric oxygenation
●
Å. A. JOHNSSON et al.
561
Fig. 4. Histologic specimen of newly formed bone from irradiated bone in an HBO-treated animal (group B) that expired 4 weeks after irradiation showing large cores of woven/chondroid bone formation. (Original magnification ⫻250, hematoxylin– eosin stain.)
Fig. 6. Histologic specimen of newly formed bone from irradiated bone (group A). Bone fragments and resorption was a pronounced phenomenon. (Original magnification ⫻250, hematoxylin– eosin stain.)
formed in 2-mm tibial defects in dogs. At 2 weeks, the tibial defect was filled mainly with undifferentiated connective tissue. From 2 weeks onward, the amount of bone progressively increased throughout the 12 weeks of the study. New bone was formed primarily by intramembranous ossification, with a small degree of endochondral ossification. The calcium content of the reparative tissue increased between 4 and 8 weeks and reached a plateau of 77% of that of normal cortical bone at 12 weeks. Yasui et al. (21) reported a third mechanism of ossification, via chondroid bone, a tissue intermediate between bone and cartilage formed directly by chondrocyte-like cells with transition from fibrous tissue to bone occurring gradually and consecutively without capillary invasion during distraction osteogenesis in rats. The cores of woven/chondroid bone found in the present study could have been formed in a manner suggested by Yasui (Figs. 4 and 5). The rabbits used in the present animal model have a bone turnover rate that is approximately three times faster than in the human, with integration of cortical implants occurring
within 6 weeks (22). Cancellous bone healing is faster than healing of cortical bone (7). Both the blood supply and the endosteal surface are proportionally greater in cancellous bone, which moreover usually has a higher rate of metabolic activity and remodeling than cortical bone (9). We chose a healing time of 8 weeks and a single dose of 15 Gy 60Co in the present study, to be able to correlate our results to previous studies measuring removal torque (i.e., implant stability) and histomorphometric parameters of implants placed in irradiated and nonirradiated bone tissue (16 –18). Both severe inhibition of cortical bone generation and partial recovery of cortical bone healing capacity with increasing time after irradiation have been shown after a single dose of 15 Gy 60Co (1). However, the radiation dose and fractionation may not be sufficient to truly reflect the radiation pathology in humans, as the tissue response has been reported to be species-specific (23). It is known that there is a large variation in cortical bone-forming capacity at different times in the same animal, as shown by Jacobsson
Fig. 5. Histologic specimen of newly formed bone from irradiated bone; 8 weeks after irradiation (group A), cores of woven/chondroid bone are now surrounded by lamellar bone. (Original magnification ⫻250, hematoxylin– eosin stain.)
Fig. 7. Histologic specimen of newly formed bone from the contralateral nonirradiated bone of the same animal seen in Fig. 6 (group A). Bone fragments and resorption was less pronounced compared to the irradiated side. (Original magnification ⫻250, hematoxylin– eosin.)
562
I. J. Radiation Oncology
●
Biology
●
Physics
et al. (24), using the bone harvest chamber in rabbits. In the present study there was a large variation in bone-forming capacity in terms of calcified tissue between the animals, and no significant differences could be detected regarding the effect of irradiation, HBO, or delayed surgery. There was a tendency to reduced bone formation at 12 and 52 weeks after irradiation in both irradiated and nonirradiated bone tissues, which might be due to an age-related effect. Shirota et al. (25) have shown that new bone formation around implants decreases with increasing age in an experimental study in rats. Several investigations (1– 6) have shown negative effects of irradiation on cortical bone healing. Emery et al. (26) evaluated the effects of irradiation on the healing of anterior vertebral grafts with use of a canine model. Preoperative irradiation did not have any deleterious effects in the situation with predominantly cancellous bone. However, the graft itself was not irradiated, a fact which might have influenced the results. The tendency to improved bone formation after HBO in nonirradiated bone tissue supports the findings of beneficial effects of HBO on bone healing stated by Nilsson et al. (10). However, Sawai et al. (27) found no significant effect of HBO on host bone when studying implants placed in autogenous bone grafts in rabbit mandible. Chen et al. (12) reported that HBO improved trabecular bone formation in irradiated bone around hydroxyapatite implants in rats, but the trabecular-bone volume in the nonirradiated side decreased compared to non-HBO-treated animals. The authors suggested that HBO might have different effects on irradiated and nonirradiated bone. In the present study no such differences could be detected. Five animals in the present study had to be prematurely sacrificed because of fractures of the irradiated leg, one in group B (HBO-treated), two in group C (surgery at 12 weeks after irradiation), and two in group D (surgery at 52 weeks after irradiation). This could be an indication of reduced mechanical properties of the bone that had been subjected to irradiation, since no fractures occurred in the nonirradiated bone. Sugimoto et al. (28) reported that the bending strength of irradiated rabbit bone was significantly decreased at 12 weeks after irradiation. Widmann et al. (29) found that prefracture irradiation significantly delayed the progressive increase in biomechanical parameters of fracture healing in rats up to 8 weeks.
Volume 48, Number 2, 2000
In the present study histologic examination revealed presence of bone fragments and bone resorption was a more pronounced phenomenon in irradiated bone. This could be an indication of delayed bone remodeling (persisting fragments from the first biopsy) or brittleness (fragmentation during the second biopsy) in the irradiated bone compared to the nonirradiated bone. Such bone fragments do contribute to the mean density (mineral content) of the biopsy and thus could have influenced the results. The trephine drill used for the biopsies inevitably produced some fragmentation of the bone tissue. A cancellous access port presented by Fox and Aufdemorte (30) has been reported to permit repeated sampling of cancellous bone tissue without fragmentation. A similar approach may reduce the error due to fragmentation. Fibrosis and inflammatory reactions in the bone marrow were more pronounced in the irradiated side. These findings are in line with the results presented by Rohrer et al. (31). The exposure parameters in the present study were chosen to avoid the contribution of soft tissue to the mean density of the biopsies. However, one cannot exclude that dense fibrous tissue may have given a small contribution to the mean density of a biopsy. A further development to avoid the contribution from soft tissue, would be to expose each sample twice at different exposures, subtract the images, and perform density analysis on the subtracted images. The findings in the present study could indicate that cancellous bone recovers faster than cortical bone from a radiation trauma. However, the large variation in boneforming capacity between the different animals, the limited number of animals, and qualitative differences between irradiated and nonirradiated bone (i.e., bone fragments and fibrosis) could also account for the lack of statistically significant differences regarding the densitometric results. In conclusion, when studying the bone-healing capacity, expressed as bone mineral content in predominantly cancellous bone, no statistically significant differences could be detected as to effects of irradiation, HBO, or delayed surgery. Qualitatively there was a more conspicuous inflammatory reaction in the marrow of the irradiated bone. Fibrosis and bone resorption were also more pronounced in the irradiated side.
REFERENCES 1. Jacobsson M, Jo¨nsson A, Albrektsson T, et al. Short- and long-term effects of irradiation on bone regeneration. Plast Reconstr Surg 1985;76:841– 848. 2. Maeda M, Bryant MH, Yamagata M,et al. Effects of irradiation on cortical bone and their time-related changes. J Bone Joint Surg 1988;70-A(3):392–399. 3. Takahashi S, Sugimoto M, Kotoura Y,et al. Long-term changes in the Haversian systems following high-dose irradiation. J Bone Joint Surg 1994;76-A:722–738. 4. Takahashi S, Sugimoto M, Kotoura Y, et al. The effects of
intraoperative radiotherapy on bone-healing ability in relation to different doses and postradiotherapy intervals. Int J Radiat Oncol Biol Phys 1994;30:1147–1152. 5. Arnold M, Kummermehr J, Trott K-R. Radiation-induced impairment of osseous healing: Quantitative studies using a standard drilling defect in rat femur. Radiat Res 1995;143:77– 84. 6. Arnold M, Stas P, Kummermehr J, et al. Radiation-induced impairment of bone healing in the rat femur: Effects of radiation dose, sequence and interval between surgery and irradiation. Radiother Oncol 1998;48:259 –265.
Cancellous bone healing after irradiation and hyperbaric oxygenation
7. Uhthoff H, Rahn B. Healing patterns of metaphyseal fractures. Clin Orthop Rel Res 1981;160:295–303. 8. Keller E. Placement of dental implants in the irradiated mandible: A protocol without adjunctive hyperbaric oxygen. J Oral Maxillofac Surg 1997;55:972–980. 9. Buckwalter J, Glimcher M, Cooper R, et al. Bone biology. Part I: Structure, blood supply, cells, matrix and mineralization. J Bone Joint Surg 1995;77-A:1256 –1275. 10. Nilsson P, Albrektsson T, Granstro¨m G,et al. The effect of hyperbaric oxygen treatment on bone regeneration: An experimental study using the bone harvest chamber in the rabbit. Int J Oral Maxillofac Implants 1988;3:43– 48. 11. Larsen P, Stronczek M, Beck F,et al. Osteointegration of implants in radiated bone with and without adjunctive hyperbaric oxygen. J Oral Maxillofac Surg 1993;51:280 –287. 12. Chen X, Matsui Y, Ohno K,et al. Histomorphometric evaluation of the effect of hyperbaric oxygen treatment on healing around hydroxyapatite implants in irradiated rat bone. Int J Oral Maxillofac Implants 1999;14:61– 68. 13. Strid KG, Ka¨lebo P. Bone mass determination from microradiographs by computer-assisted videodensitometry I. Methodology. Acta Radiol 1988;29:465– 472. 14. Strid KG, Kling-Petersen T, Johnsson ÅA. Bone mass determination from microradiographs using a slide scanner and modern software. In manuscript. 15. Bradley J. Fisher’s method of randomization. In: Distributionfree statistical tests. London: Prentice-Hall; 1968. p.68 – 86. 16. Hansson Å, Johnsson K, Jacobsson M, et al. Removal torques for titanium implants following irradiation. In: Laney W, Tolman D, editors. Tissue integration in oral, orthopedic and maxillofacial reconstruction. Chicago: Quintessence; 1992. p. 228 –233. 17. Johnsson K, Hansson Å, Granstro¨m G, et al. The effects of hyperbaric oxygenation on bone-titanium implant interface strength with and without preceding irradiation. Int J Oral Maxillofac Implants 1993;8:415– 419. 18. Johnsson ÅA, Sawaii T, Jacobsson M,et al. A histomorphometric study of bone reactions to titanium implants in irradiated bone and the effect of hyperbaric oxygen treatment. Int J Oral Maxillofac Implants 1999;14:699 –706. 19. Shapiro F. Cortical bone repair. J Bone Joint Surg 1988;70A:1067–1081.
●
Å. A. JOHNSSON et al.
563
20. Markel M, Wikenheiser M, Chao E. Formation of bone in tibial defects in a canine model. J Bone Joint Surg 1991;73A:914 –923. 21. Yasui N, Sato M, Ochi T, et al. Three modes of ossification during distraction osteogenesis in the rat. J Bone Joint Surg 1997;79-B:824 – 830. 22. Roberts W, Smith R, Zilberman Y,et al. Osseous adaptation to continuous loading of rigid endosseous implants. Am J Orthod 1984;86:95–111. 23. Marx R. Discussion: Histomorphometric examination of healing around hydroxylapatite implants in Co60-irradiated bone. J Oral Maxillofac Surg 1994;52:172–173. 24. Jacobsson M, Ka¨lebo P, Albrektsson T,et al. Provoked repetitive healing of mature bone tissue following irradiation: A quantitative investigation. Acta Radiol Oncol 1986;25:57– 62. 25. Shirota T, Ohni K, Suzuki K,et al. The effect of aging on the healing of hydroxylapatite implants. J Oral Maxillofac Surg 1993;51:51–56. 26. Emery S, Brazinski M, Koka A,et al. The biological and biomechanical effects of irradiation on anterior spinal bone grafts in a canine model. J Bone Joint Surg 1994;76-A:540 – 548. 27. Sawai T, Niimi A, Johansson C, et al. The effect of hyperbaric oxygen treatment on bone tissue reactions to c.p. titanium implants placed in free autogenous bone grafts. Clin Oral Impl Res 1998;9:384 –397. 28. Sugimoto M, Takahashi S, Toguchida J,et al. Changes in bone after high-dose irradiation. J Bone Joint Surg 1991;73-B:492– 497. 29. Widdman R, Pelker R, Friedlaender G,et al. Effects of prefracture irradiation on the biomechanical parameters of fracture healing. J Orthop Res 1993;11:422– 428. 30. Fox W, Aufdemorte T. Experience with osseous implants for bone and biomaterials research. J Long-Term Effects Med Implants 1993;3:1–27. 31. Rohrer M, Kim Y, Fayos J,et al. The effect of cobalt-60 irradiation on monkey mandibles. Oral Surg Oral Med Oral Path 1979;48:424 – 440.
PII S0360-3016(00)00638-6
BIOLOGY CONTRIBUTION
A MICRORADIOGRAPHIC INVESTIGATION OF CANCELLOUS BONE HEALING AFTER IRRADIATION AND HYPERBARIC OXYGENATION: A RABBIT STUDY ÅSE A. JOHNSSON, M.D.,*† MAGNUS JACOBSSON, M.D., PH.D.,† GO¨ STA GRANSTRO¨ M, M.D., D.D.S., PH.D.,‡ CARINA B. JOHANSSON, PH.D.,† KARL-GUSTAV STRID, M.SC.(ENG.), PH.D.,§ AND INGELA TURESSON, M.D., PH.D.㛳 Departments of *Radiology, †Biomaterials/Handicap Research, and ‡Otolaryngology, Head and Neck Surgery, §Biomaterials Research Group, Institute of Anatomy and Cell Biology, Sahlgrenska University Hospital, Go¨teborg University, Gothenburg Sweden; and 㛳 Department of Oncology, Academic Hospital, Uppsala, Sweden Purpose: To analyze the effect of irradiation on cancellous bone healing at different times after irradiation and to study if hyperbaric oxygen therapy (HBO) would affect the bone healing capacity, when delivered directly after irradiation. Methods and Materials: Rabbits were given a single dose of 15 Gy 60Co radiation to one hind leg, the other hind leg serving as control. A standardized defect through the femoral metaphysis of the rabbits was created by a trephine drill biopsy at different times after irradiation. New bone formation in the defect was evaluated by a new biopsy through the previous defect after a healing time of 8 weeks. The mineral contents of the biopsies were analyzed by microradiography and microdensitometry. Results: There was a large variation in the bone-forming capacity expressed as bone mineral content between the animals. No statistically significant differences could be detected regarding the effect of irradiation, HBO, or delayed surgery. Qualitative histology revealed more pronounced inflammation, fibrosis, and bone resorption in the irradiated bone. Conclusions: No definite conclusions can be drawn from the results of this study, however it might be hypothesized that cancellous bone recovers faster than cortical bone from radiation trauma. © 2000 Elsevier Science Inc. Cancellous bone healing, Hyperbaric oxygen treatment, Irradiation, Microdensitometry, Microradiography.
have been studied by Nilsson et al. (10) and Larsen et al. (11). Both authors stated beneficial effects of HBO on bone healing, and Larsen et al. (11) showed that the use of HBO improved implant integration in irradiated bone tissue. In a recent study, Chen et al. (12) reported that HBO might have different effects on irradiated and nonirradiated bone. The aim of the present study was to evaluate the boneforming capacity in predominantly cancellous bone, at different times after irradiation, and to study the effect of HBO on cancellous bone healing with and without preceding irradiation.
INTRODUCTION Radiotherapy in combination with surgery is the treatment generally used for malignant tumors in the craniofacial region. Several experimental investigations (1– 6) have studied the effects of irradiation on cortical bone healing. The damaging effect from radiotherapy on bone tissue is believed to be related to direct damage to the osteogenic cells (1, 3); however, vascular injury is also considered important (3). The ratio of cortical to cancellous bone varies in different parts of the craniofacial area. The modes and the speed of healing of cortical and cancellous bone differ, and cortical healing lags behind cancellous bone healing (7). However, the injury of irradiation to cells may be greater in the highly cellular cancellous bone (8), since cancellous bone usually has a higher rate of metabolic activity (9). The effects of hyperbaric oxygenation (HBO) on cortical bone healing
METHODS AND MATERIALS The following study design was approved by the Go¨teborg University Laboratory Animal Ethics Committee, Sweden. burg Medical Association, the Eivind and Elsa K:son Sylvan Foundation, the Assar Gabrielsson Foundation, the Go¨ranssons Foundation, the G. A. E. Nilsson Foundation, the A. A. Yhle´n Foundation, the King Gustaf V Jubilee Foundation, and the Swedish Medical Research Council (MFR). Accepted for publication 17 April 2000.
Reprint requests to: Dr. Åse A. Johnsson, Department of Biomaterials/Handicap Research, Institute for Surgical Sciences, Box 412, SE-405 30 Go¨teborg, Sweden. Acknowledgments—The following funds, here gratefully acknowledged, have supported the present study: the Hjalmar Svensson Foundation, the Greta and Einar Asker Foundation, the Gothen555
556
I. J. Radiation Oncology
●
Biology
●
Physics
Volume 48, Number 2, 2000
Table 1. Number of animals in the different groups Animals evaluated/ animals included Group A: Biopsies directly and 8 wk after irradiation 8/8 Group B: Biopsies directly and 8 wk after irradiation ⫹ HBO 6/10 Group C: Biopsies at 12 wk and 20 wk after irradiation 8/11 Group D: Biopsies at 52 wk and 60 wt after irradiation 8/12 Group B: One animal had to be sacrificed because of a diaphyseal fracture of the irradiated leg, two animals died after the last HBO treatment of unknown reasons, and one animal had to be sacrificed because of a spinal lesion secondary to seizures at the HBO treatment. Group C: Two animals had to be sacrificed because of fractures of the irradiated leg and one animal could not be evaluated due to preparation failure. Group D: One animal expired before the time of surgery, two animals had to be sacrificed because of fractures of the irradiated leg, and one animal had to be sacrificed because of a pulmonary infection.
Animals Adult (over 9 months old) New Zealand white rabbits were used in the study. The rabbits were group housed, the animals being females and neutered males. All animals received a single dose of 15 Gy 60Co radiation to one hind leg, the other hind leg serving as control. A cylindrical defect, 4.5 mm in diameter, through each femoral metaphysis was created immediately (group A and B), at 12 weeks (group C) and at 52 weeks (group D) after irradiation. Group B received HBO treatment during the first 4 postoperative weeks. After a healing time of 8 weeks a biopsy of the bone formed in the created defect was taken. Twelve animals were used to develop the surgical procedure. Forty-one animals received irradiation; 30 of them could be evaluated in the study. Five animals had to be sacrificed because of fractures of the irradiated leg. Five animals expired prior to evaluation and one animal could not be evaluated due to preparation failure. The number of animals in each group is presented in Table 1. Anesthetic procedure Intramuscular injections of fentanyl and fluanizon (Hypnorm, Janssen-Cilag, Buckinghamshire, England) at a dose of 0.5 ml/kg body weight and intraperitoneal injections of diazepam (Stesolid Novum, A/S Dumex Denmark, Pharmacia-Upjohn Stockholm, Sweden) at a dose of 1.5 mg/kg body weight were used for general anesthesia during irradiation and surgical procedures but not during HBO treatment. Local anesthesia with 1.0 ml of 5% lidocaine (Xylocaine, Astra, Mo¨lndal, Sweden) was administered to the femora prior to surgery. The shaved skin of the rabbits was washed with a mixture of iodine and 70% ethanol prior to surgery. In connection with surgery the animals received an antibiotic, benzyl-penicillin (PenoVet, Boeringer Ingelheim Agrovet, Hellerup, Denmark) at a dose of 20 mg/kg body weight. Postoperatively intramuscular injections of buprenorfin (Temgesic, Reckitt & Colman, Hull, England) were given in doses of 0.1 mg/kg body weight. At the time of sacrifice intravenous injections of a mixture of saline and barbiturates (Mebumal Vet., 60 mg/ml, Nord Vacc, Sweden) 1:4 were given.
Irradiation procedure The rabbits were irradiated to the distal femoral metaphysis and the proximal tibial metaphysis of the right hind leg, the left hind leg serving as control. During irradiation the femur and tibia were placed on a 5-cm-thick polystyrene phantom. Gamma irradiation (60Co) was used to minimize the difference in the absorbed dose in soft tissue and bone. Source-to-skin distance was 60 cm and field size was 7.5 ⫻ 7.5 cm. The dose rate was approximately 0.5 Gy per minute. A 5-mm bolus was applied to ensure full build-up, and 15 Gy were given as a single dose. Hyperbaric oxygen treatment On the third postoperative day (Monday) the animals in group B were placed in a pressure chamber of 75 liters volume (Gothenburg Diving Technique, Gothenburg, Sweden) and subjected to pure oxygen of 280 kPa for 2 hours. During this period, the first 10 minutes were used for successive compression up to 280 kPa; the pressure was kept constant for 90 minutes and followed by decompression for 20 minutes. The chamber temperature was kept at 23°C by a water cooling system. Produced carbon dioxide was eliminated by constant flow of oxygen with a flow rate of 1.5 L/min. The HBO treatment was performed once daily Monday to Friday and a total of 20 treatments was given. Guide and implant design In order to take standardized biopsies through the femoral metaphysis, a biopsy-guide-system was constructed by Wennbergs Finmek, Angered, Sweden. Illustrations of the guide and the trephine drill biopsy procedure are given in Figs. 1 a–1d. The surface to be placed in contact with the femur was modeled and slightly curved, making possible a close fit to the distal femoral diaphysis and metaphysis. There were four drill holes (canals) through the guide: one proximal, 4.5 mm in diameter, two oblique 4.55 mm in diameter, with a common opening on the surface facing toward the femoral metaphysis, and one distal, 1.6 mm in diameter. In order to guide the trephine drill through the femoral metaphysis, one of the two oblique canals was used for the
Cancellous bone healing after irradiation and hyperbaric oxygenation
Fig. 1. (a) The biopsy guide with its oblique canals (*), distal drill hole (∧), and proximal canal (#) used for the landmark fixture seen from above. (b) The modeled surface of the biopsy guide facing the medial aspect of the femur with the 1.6-mm drill (∧) and the 2.2-mm drill with connector (#) used to center the drill hole for the landmark fixture. The common opening of the two oblique canals (*) is placed between the drills. (c) The biopsy guide with the 1.6-mm drill (∧), the trephine drill, (*) and the landmark fixture (#) . (d) At the second biopsy the distal part of the femur was dissected free from soft tissue. The guide was matched to the previous defect by the landmark fixture and the 1.6-mm drill. During drilling the cortex of the lateral femoral metaphysis was inspected in order to ensure that the trephine drill did not deviate from the margins of the previous defect.
trephine drill, the other for saline cooling. The outer diameter of the trephine drill used was 4.5 mm and the inner diameter was 3.5 mm. A self-tapping, stainless steel implant (Wennbergs Finmek, Angered, Sweden) was used to stabilize the proximal part of the guide and to serve as a landmark for the second biopsy. The implant had a total length of 8.2 mm; the threaded part inserted in the femoral diaphysis was 5 mm in length, and the diameter was 2.4 mm. The top of the fixture was cylindrical, 4.5 mm in diameter, with an internal hexagon to fit a key used during insertion. Surgical procedure Under aseptic conditions a longitudinal incision was made through the skin and the muscle fascia of the distal medial side of the femur. Muscles were gently mobilized along the medial aspect of the femur. Parts of the medial patellar retinaculum were divided in order to displace the patellar tendon laterally. The periosteum was left intact except for the drilling sites, where it was incised and gently pushed aside. The guide described above was attached to the distal femoral metaphysis by a 1.6-mm drill. To stabilize the proximal part of the guide, a landmark fixture was inserted through the guide into the medial aspect of the femoral diaphysis. When the self-tapping implant had been inserted, its cylindrical top exactly fit the proximal canal and thus stabilized the guide. A 3.5-mm biopsy was taken from the
●
Å. A. JOHNSSON et al.
557
medial aspect of the femoral metaphysis through the medullary cavity and out through the lateral cortex of the femoral metaphysis. Ample cooling with saline was provided during drilling. After removal of the biopsy, the distal drill and the guide itself were removed, leaving a defect of 4.5 mm diameter through the distal femoral metaphysis and the 1.6-mm distal drillhole to heal. The landmark fixture was left in place. The 4.5-mm defect was irrigated and inspected in order to check that no visible bone was left in the defect. The patellar retinaculum and the fascia were closed with interrupted resorbable sutures and the skin with interrupted nonabsorbable sutures. Eight weeks after the primary operation, the animals were again anesthetized as described above and the distal part of the femur was dissected free from soft tissue. The guide was matched to the previous defect by the landmark fixture and the 1.6-mm drill. The created defect had healed, but it was still possible to recognize its margins. A new biopsy with the same diameter as the first one was taken through the guide and the newly formed bone in the previous defect (Fig. 1d). During drilling ample cooling with saline was provided and the cortex of the lateral femoral metaphysis was inspected in order to ensure that the trephine drill did not deviate from the margins of the previous defect. Finally the animals were sacrificed by intravenous injections of barbiturates. All biopsies were directly immersed in 4% neutral buffered formaldehyde and stored at 4°C. Microradiography The biopsies from the same animal, i.e., four specimens, were microradiographed simultaneously together with a density reference consisting of an aluminum step wedge on a Kodak high-resolution plate type 1A. The aluminum step wedge had 10 steps, each measuring 5 ⫻ 2.5 mm, and the height of each step was 0.3 mm, yielding a maximum wedge thickness of 3.0 mm. A Machlett roentgen tube, type OEG-50 with a copper target and a focal spot of 1 mm, generated a polyenergetic spectrum, which was prefiltered through 0.5 mm aluminum. The tube focus was fixed at 242 mm from the recording surface. Exposures were carried out at 27 kV and 20 mA for 40 minutes. The plates were processed immediately following exposure by development in fresh Kodak D-19 solution for 6 min at 20°C during standardized agitation and subsequent rinsing, fixing, washing and drying as proposed by the manufacturer. Strid and Ka¨lebo (13) have given a detailed description of the microradiographic procedure. Figure 2a demonstrates a survey microradiograph of a biopsy. Bone mass determination The microradiographs were digitized using a standard 35-mm slide scanner (Nikon CoolScan LS-10E). No contrast enhancement or other automatic image modification was performed during the scanning procedure. The microradiographs were scanned at a resolution of 106 pixels per mm. To cover the entire slide, each microradiograph was scanned in two portions, the step wedge being included in
558
I. J. Radiation Oncology
●
Biology
●
Physics
Volume 48, Number 2, 2000
Histology After microradiography, samples of newly formed bone from half of the animals in each group, and from two animals that were sacrificed prior to evaluation time, were decalcified in 16% formic acid and embedded in paraffin. Two, 3-m-thick, longitudinal sections were cut through the central part of each sample. The sections were stained with hematoxylin– eosin prior to light microscopic observations (Fig. 2b).
Fig. 2. (a) A survey picture of a microradiographed biopsy. (b) A survey picture of a decalcified biopsy stained with hematoxylin– eosin.
both scans in order to ensure that both scans showed the same density profile. The images were stored in TIFF format. The tagged image file format (TIFF) images were analyzed by a de facto standard image analysis program (NIH Image v. 1.62, National Institutes of Health). Quantitative analysis was performed using a custom program (AluDef 1.2) developed in-house using LabVIEW 5.0 (National Instruments) on an Apple Power Macintosh G3 computer. A detailed description of the procedure and assessment of the method is given by Strid et al. (14). The mineral content of the specimens was calculated in terms of aluminum equivalent mass.
Statistical methods Fisher’s test for pair comparisons was applied in assessing the effect of irradiation on bone formation. The effect of HBO and delayed surgery after irradiation was evaluated by Fisher’s permutation test (15) for irradiated as well as nonirradiated bone tissue. The numbers of animals in the different groups were based on results from previous studies (16 –18). RESULTS To express the bone-forming capacity (in terms of mineralization) in irradiated and nonirradiated bone tissue, a ratio between the mineral content of the newly formed bone in the biopsy, taken after a healing time of 8 weeks, and the mineral content of the biopsy taken to create the defect, was formed. A ratio of 1.0 would indicate that the same amount of mineralized bone tissue had formed after 8 weeks in the created defect as was present in the original biopsy. The results are presented in Table 2 and Fig. 3. There was a large variation in the bone-forming capacity between the animals. No statistically significant differences could be detected regarding the effect of irradiation, HBO, or delayed
Table 2. The bone-forming capacity expressed as a ratio between the mineral content of the newly formed bone in the biopsy, taken after a healing time of 8 weeks, and the mineral content of the biopsy taken to create the defect Group A Biopsy directly after irradiation
Group B Biopsy directly after irradiation HBO-treated animals
1 2 3 4 5 6 7 8
Non-irr
1.02 0.78 0.84 1.12 1.48 1.48 0.66 0.78 M 1.02 SD 0.32
Irr
Animal No.
Non-irr
Irr
0.62 0.66 0.79 0.96 1.61 0.96 0.96 1.26 0.98 0.33
1 2 3 4 5 6
0.83 0.78 1.45 0.91 1.72 1.33
0.88 0.88 0.99 0.87 0.9 1.6
Two animals expired 4 wk after surgery
1.17 0.39 0.69 1.05
1.02 0.29 0.62 0.7
Group D Biopsy 52 wk after irradiation Ratio of mineral content 8 wk biopsy/original biopsy
Ratio of mineral content 8 wk biopsy/original biopsy
Ratio of mineral content 8 wk biopsy/original biopsy
Ratio of mineral content 8 wk biopsy/original biopsy Animal No.
Group C Biopsy 12 wk after irradiation
Animal No. 1 2 3 4 5 6 7 8
Non-irr
Irr
Animal No.
Non-irr
Irr
0.43 0.73 1.23 0.89 0.85 0.98 0.75 1.07 0.87 0.24
0.77 0.61 0.87 0.78 0.94 0.46 1.08 0.35 0.73 0.25
1 2 3 4 5 6 7 8
0.57 1.16 0.73 0.74 0.69 0.8 0.67 0.69 0.76 0.18
0.55 1.13 0.83 1.02 0.6 0.91 0.64 0.76 0.8 0.21
0.82
0.82
One animal expired 6 wk after surgery
Cancellous bone healing after irradiation and hyperbaric oxygenation
●
Å. A. JOHNSSON et al.
559
Fig. 3. A diagram showing the results of the densitometric measurements as mean values of the ratios between the mineral content of the newly formed bone in the created defect and the mineral content of the biopsy taken to create the defect. There were no significant differences. Standard deviations (SD) are marked.
surgery on bone-forming capacity in terms of mineralization. However, there was a tendency toward reduced bone formation at 12 and 52 weeks after irradiation in both irradiated and nonirradiated bone tissue and a tendency toward improved bone formation after HBO in nonirradiated bone tissue. The samples that were evaluated histologically were examined in a blinded manner. The amount of fibrous tissue, inflammatory cells in the bone marrow, ongoing resorption, osteoclasts, osteoblasts, and blood vessels were categorized as abundant, moderately present, rare, or none for each sample judged. The results are summarized in Table 3. General qualitative observations on the decalcified hematoxylin– eosin-stained paraffin sections were also made in a blinded manner and revealed the following: Group A (directly after surgery, 8 weeks after irradiation). The irradiated test side revealed a higher amount of dense (fibrous) connective tissue in the bone marrow as compared to the nonirradiated control side, where a looser fibrous tissue dominated. The number of inflammatory cells was greater on the test side. Bone flakes were observed on both the test and control sections, seemingly more pronounced on the irradiated side. The bone flakes in the irradiated biopsies seemed to be acellular/pycnotic, whereas in the sections from nonirradiated bone, bone cells could be observed. A more active remodeling appeared to be present in the control biopsies. Large cores of woven/chondroid bone tissue were observed both in the test and in the control samples. However, the bone tissue structure itself appeared maturer in the latter. The mentioned cores had a morphologic appearance similar to chondroid tissue in the light
microscope (Figs. 4 and 5). However, with the aid of polarized light it was possible to detect cross-linked fibers in the tissue. Therefore we chose to call these cores “woven/ chondroid bone tissue.” Group B (directly after surgery, 8 weeks after irradiation and HBO treatment). The amount of fibrous tissue and inflammatory cells was greater on the irradiated side as compared to the nonirradiated biopsies. Moreover, the fat cells appeared more regular and the number of these cells was greater in the nonirradiated samples. There was a tendency of a higher degree of bone resorption on the irradiated test side. Often large cores of woven/chondroid bone tissue could be observed on the test and control side but seemingly larger cores were present on the former samples. Group C (8 weeks after surgery, 20 weeks after irradiation). An organized capsule-like fibrous tissue was more conspicuous on the irradiated side as compared to the control side. The number of fat cells seemed to be greater in the nonirradiated control biopsies, whereas inflammatory cells were more abundant in the test samples. There was a tendency of a more active bone remodeling in the irradiated biopsies compared to the control side. Group D (8 weeks after surgery, 60 weeks after irradiation). A loose connective tissue was observed in the nonirradiated biopsies as compared to a denser structure in the irradiated samples. The number of inflammatory cells was greater on the irradiated side and the amount of fat cells was less on this side in comparison to nonirradiated control biopsies. The bone tissue resorption was more pronounced and gave a more active impression on the irradiated side as compared to the nonirradiated samples.
560
I. J. Radiation Oncology
●
Biology
●
Physics
Volume 48, Number 2, 2000
Table 3. Histologic Evaluation* Group† A: Irr
A: N-irr
Fibrous tissue Abundant 1 0 Moderate 3 2 Rare 0 1 None 0 1 Inflammatory cells in the bone marrow Abundant 0 0 Moderate 4 0 Rare 0 2 None 0 2 Ongoing resorption Abundant 2 0 Moderate 1 3 Rare 1 1 None 0 0 Osteoclasts Abundant 2 0 Moderate 2 2 Rare 0 2 None 0 0 Osteoblasts Abundant 0 0 Moderate 3 4 Rare 1 0 None 0 0 Blood vessels Abundant 0 0 Moderate 4 4 Rare 0 0 None 0 0
B: Irr
B: N-irr
C: Irr
C: N-irr
D: Irr
D: N-irr
1 1 1 0
0 1 1 1
1 2 0 1
1 2 1 0
0 1 3 0
0 3 0 1
0 1 2 0
0 0 1 2
1 1 2 0
0 0 4 0
0 2 2 0
0 1 3 0
0 3 0 0
0 0 3 0
2 1 1 0
0 4 0 0
1 3 0 0
0 2 2 0
0 3 0 0
0 0 3 0
1 3 0 0
0 4 0 0
3 1 0 0
0 2 2 0
0 1 2 0
0 2 1 0
0 2 2 0
0 3 1 0
0 4 0 0
0 2 2 0
0 3 0 0
0 3 0 0
0 4 0 0
0 4 0 0
1 3 0 0
0 4 0 0
* The samples that were evaluated histologically were examined in a blinded manner. The parameters analyzed were: fibrous tissue, inflammatory cells in the bone marrow, ongoing resorption, osteoclasts, osteoblasts, and blood vessels. For each sample judged, the different parameters were categorized as abundant, moderate, rare, or none. (0 – 4) ⫽ number of samples in each category. Irr ⫽ irradiated, N-irr ⫽ nonirradiated. † Group A: Surgery directly after irradiation: 4 animals. Group B: Surgery directly after irradiation and HBO treatment: 3 animals. Group C: Surgery at 12 weeks after irradiation: 4 animals. Group D: Surgery at 52 after irradiation: 4 animals.
In summary, in all samples there were signs of both endochondral and intramembranous ossification. Cores of woven/chondroid bone tissue were present in many biopsies but more pronounced in irradiated bone. Generally there was more ongoing resorption and osteoclasts present in the irradiated side compared to the nonirradiated side. In the bone marrow there was a more pronounced inflammatory reaction in the irradiated side compared to the nonirradiated side in all groups. This finding was most conspicuous in the group that had surgery directly after irradiation. There was more fibrous tissue in the groups that had surgery directly after irradiation both with and without HBO. However, there were no discernable differences regarding the number of blood vessels. In one animal that expired after the last HBO treatment (4 weeks healing time) there was more fibrous tissue, osteoclasts, and bone resorption in the irradiated side compared to the nonirradiated side. One animal in group D (surgery 52 weeks after irradiation) that had to be sacrificed 6 weeks after surgery (because of a pulmonary
infection) showed no such differences. Examples of histologic findings are presented in Figs. 4 –7. DISCUSSION Shapiro (19) described the healing of cortical 2.4-mm diaphyseal defects in the rabbit tibia and femur. The initial source of repair tissue was the marrow. Vessels grew into the defect accompanied by mesenchymal cells. Bone formation was observed after 1 week, as woven bone at the endosteal surface at the periphery of the defects. After 2 weeks cores of woven bone were seen in the central parts of the defect. On the surface of these cores, osteoblasts formed osteoid in a lamellar manner. After 4 weeks only few cores of woven bone were observed. Few resorptive osteoclasts were observed during the healing. Haversian systems were beginning to transverse the peripheral parts of the newly formed repair bone after 6 – 8 weeks. Markel et al. (20) studied the reparative tissue and bone
Cancellous bone healing after irradiation and hyperbaric oxygenation
●
Å. A. JOHNSSON et al.
561
Fig. 4. Histologic specimen of newly formed bone from irradiated bone in an HBO-treated animal (group B) that expired 4 weeks after irradiation showing large cores of woven/chondroid bone formation. (Original magnification ⫻250, hematoxylin– eosin stain.)
Fig. 6. Histologic specimen of newly formed bone from irradiated bone (group A). Bone fragments and resorption was a pronounced phenomenon. (Original magnification ⫻250, hematoxylin– eosin stain.)
formed in 2-mm tibial defects in dogs. At 2 weeks, the tibial defect was filled mainly with undifferentiated connective tissue. From 2 weeks onward, the amount of bone progressively increased throughout the 12 weeks of the study. New bone was formed primarily by intramembranous ossification, with a small degree of endochondral ossification. The calcium content of the reparative tissue increased between 4 and 8 weeks and reached a plateau of 77% of that of normal cortical bone at 12 weeks. Yasui et al. (21) reported a third mechanism of ossification, via chondroid bone, a tissue intermediate between bone and cartilage formed directly by chondrocyte-like cells with transition from fibrous tissue to bone occurring gradually and consecutively without capillary invasion during distraction osteogenesis in rats. The cores of woven/chondroid bone found in the present study could have been formed in a manner suggested by Yasui (Figs. 4 and 5). The rabbits used in the present animal model have a bone turnover rate that is approximately three times faster than in the human, with integration of cortical implants occurring
within 6 weeks (22). Cancellous bone healing is faster than healing of cortical bone (7). Both the blood supply and the endosteal surface are proportionally greater in cancellous bone, which moreover usually has a higher rate of metabolic activity and remodeling than cortical bone (9). We chose a healing time of 8 weeks and a single dose of 15 Gy 60Co in the present study, to be able to correlate our results to previous studies measuring removal torque (i.e., implant stability) and histomorphometric parameters of implants placed in irradiated and nonirradiated bone tissue (16 –18). Both severe inhibition of cortical bone generation and partial recovery of cortical bone healing capacity with increasing time after irradiation have been shown after a single dose of 15 Gy 60Co (1). However, the radiation dose and fractionation may not be sufficient to truly reflect the radiation pathology in humans, as the tissue response has been reported to be species-specific (23). It is known that there is a large variation in cortical bone-forming capacity at different times in the same animal, as shown by Jacobsson
Fig. 5. Histologic specimen of newly formed bone from irradiated bone; 8 weeks after irradiation (group A), cores of woven/chondroid bone are now surrounded by lamellar bone. (Original magnification ⫻250, hematoxylin– eosin stain.)
Fig. 7. Histologic specimen of newly formed bone from the contralateral nonirradiated bone of the same animal seen in Fig. 6 (group A). Bone fragments and resorption was less pronounced compared to the irradiated side. (Original magnification ⫻250, hematoxylin– eosin.)
562
I. J. Radiation Oncology
●
Biology
●
Physics
et al. (24), using the bone harvest chamber in rabbits. In the present study there was a large variation in bone-forming capacity in terms of calcified tissue between the animals, and no significant differences could be detected regarding the effect of irradiation, HBO, or delayed surgery. There was a tendency to reduced bone formation at 12 and 52 weeks after irradiation in both irradiated and nonirradiated bone tissues, which might be due to an age-related effect. Shirota et al. (25) have shown that new bone formation around implants decreases with increasing age in an experimental study in rats. Several investigations (1– 6) have shown negative effects of irradiation on cortical bone healing. Emery et al. (26) evaluated the effects of irradiation on the healing of anterior vertebral grafts with use of a canine model. Preoperative irradiation did not have any deleterious effects in the situation with predominantly cancellous bone. However, the graft itself was not irradiated, a fact which might have influenced the results. The tendency to improved bone formation after HBO in nonirradiated bone tissue supports the findings of beneficial effects of HBO on bone healing stated by Nilsson et al. (10). However, Sawai et al. (27) found no significant effect of HBO on host bone when studying implants placed in autogenous bone grafts in rabbit mandible. Chen et al. (12) reported that HBO improved trabecular bone formation in irradiated bone around hydroxyapatite implants in rats, but the trabecular-bone volume in the nonirradiated side decreased compared to non-HBO-treated animals. The authors suggested that HBO might have different effects on irradiated and nonirradiated bone. In the present study no such differences could be detected. Five animals in the present study had to be prematurely sacrificed because of fractures of the irradiated leg, one in group B (HBO-treated), two in group C (surgery at 12 weeks after irradiation), and two in group D (surgery at 52 weeks after irradiation). This could be an indication of reduced mechanical properties of the bone that had been subjected to irradiation, since no fractures occurred in the nonirradiated bone. Sugimoto et al. (28) reported that the bending strength of irradiated rabbit bone was significantly decreased at 12 weeks after irradiation. Widmann et al. (29) found that prefracture irradiation significantly delayed the progressive increase in biomechanical parameters of fracture healing in rats up to 8 weeks.
Volume 48, Number 2, 2000
In the present study histologic examination revealed presence of bone fragments and bone resorption was a more pronounced phenomenon in irradiated bone. This could be an indication of delayed bone remodeling (persisting fragments from the first biopsy) or brittleness (fragmentation during the second biopsy) in the irradiated bone compared to the nonirradiated bone. Such bone fragments do contribute to the mean density (mineral content) of the biopsy and thus could have influenced the results. The trephine drill used for the biopsies inevitably produced some fragmentation of the bone tissue. A cancellous access port presented by Fox and Aufdemorte (30) has been reported to permit repeated sampling of cancellous bone tissue without fragmentation. A similar approach may reduce the error due to fragmentation. Fibrosis and inflammatory reactions in the bone marrow were more pronounced in the irradiated side. These findings are in line with the results presented by Rohrer et al. (31). The exposure parameters in the present study were chosen to avoid the contribution of soft tissue to the mean density of the biopsies. However, one cannot exclude that dense fibrous tissue may have given a small contribution to the mean density of a biopsy. A further development to avoid the contribution from soft tissue, would be to expose each sample twice at different exposures, subtract the images, and perform density analysis on the subtracted images. The findings in the present study could indicate that cancellous bone recovers faster than cortical bone from a radiation trauma. However, the large variation in boneforming capacity between the different animals, the limited number of animals, and qualitative differences between irradiated and nonirradiated bone (i.e., bone fragments and fibrosis) could also account for the lack of statistically significant differences regarding the densitometric results. In conclusion, when studying the bone-healing capacity, expressed as bone mineral content in predominantly cancellous bone, no statistically significant differences could be detected as to effects of irradiation, HBO, or delayed surgery. Qualitatively there was a more conspicuous inflammatory reaction in the marrow of the irradiated bone. Fibrosis and bone resorption were also more pronounced in the irradiated side.
REFERENCES 1. Jacobsson M, Jo¨nsson A, Albrektsson T, et al. Short- and long-term effects of irradiation on bone regeneration. Plast Reconstr Surg 1985;76:841– 848. 2. Maeda M, Bryant MH, Yamagata M,et al. Effects of irradiation on cortical bone and their time-related changes. J Bone Joint Surg 1988;70-A(3):392–399. 3. Takahashi S, Sugimoto M, Kotoura Y,et al. Long-term changes in the Haversian systems following high-dose irradiation. J Bone Joint Surg 1994;76-A:722–738. 4. Takahashi S, Sugimoto M, Kotoura Y, et al. The effects of
intraoperative radiotherapy on bone-healing ability in relation to different doses and postradiotherapy intervals. Int J Radiat Oncol Biol Phys 1994;30:1147–1152. 5. Arnold M, Kummermehr J, Trott K-R. Radiation-induced impairment of osseous healing: Quantitative studies using a standard drilling defect in rat femur. Radiat Res 1995;143:77– 84. 6. Arnold M, Stas P, Kummermehr J, et al. Radiation-induced impairment of bone healing in the rat femur: Effects of radiation dose, sequence and interval between surgery and irradiation. Radiother Oncol 1998;48:259 –265.
Cancellous bone healing after irradiation and hyperbaric oxygenation
7. Uhthoff H, Rahn B. Healing patterns of metaphyseal fractures. Clin Orthop Rel Res 1981;160:295–303. 8. Keller E. Placement of dental implants in the irradiated mandible: A protocol without adjunctive hyperbaric oxygen. J Oral Maxillofac Surg 1997;55:972–980. 9. Buckwalter J, Glimcher M, Cooper R, et al. Bone biology. Part I: Structure, blood supply, cells, matrix and mineralization. J Bone Joint Surg 1995;77-A:1256 –1275. 10. Nilsson P, Albrektsson T, Granstro¨m G,et al. The effect of hyperbaric oxygen treatment on bone regeneration: An experimental study using the bone harvest chamber in the rabbit. Int J Oral Maxillofac Implants 1988;3:43– 48. 11. Larsen P, Stronczek M, Beck F,et al. Osteointegration of implants in radiated bone with and without adjunctive hyperbaric oxygen. J Oral Maxillofac Surg 1993;51:280 –287. 12. Chen X, Matsui Y, Ohno K,et al. Histomorphometric evaluation of the effect of hyperbaric oxygen treatment on healing around hydroxyapatite implants in irradiated rat bone. Int J Oral Maxillofac Implants 1999;14:61– 68. 13. Strid KG, Ka¨lebo P. Bone mass determination from microradiographs by computer-assisted videodensitometry I. Methodology. Acta Radiol 1988;29:465– 472. 14. Strid KG, Kling-Petersen T, Johnsson ÅA. Bone mass determination from microradiographs using a slide scanner and modern software. In manuscript. 15. Bradley J. Fisher’s method of randomization. In: Distributionfree statistical tests. London: Prentice-Hall; 1968. p.68 – 86. 16. Hansson Å, Johnsson K, Jacobsson M, et al. Removal torques for titanium implants following irradiation. In: Laney W, Tolman D, editors. Tissue integration in oral, orthopedic and maxillofacial reconstruction. Chicago: Quintessence; 1992. p. 228 –233. 17. Johnsson K, Hansson Å, Granstro¨m G, et al. The effects of hyperbaric oxygenation on bone-titanium implant interface strength with and without preceding irradiation. Int J Oral Maxillofac Implants 1993;8:415– 419. 18. Johnsson ÅA, Sawaii T, Jacobsson M,et al. A histomorphometric study of bone reactions to titanium implants in irradiated bone and the effect of hyperbaric oxygen treatment. Int J Oral Maxillofac Implants 1999;14:699 –706. 19. Shapiro F. Cortical bone repair. J Bone Joint Surg 1988;70A:1067–1081.
●
Å. A. JOHNSSON et al.
563
20. Markel M, Wikenheiser M, Chao E. Formation of bone in tibial defects in a canine model. J Bone Joint Surg 1991;73A:914 –923. 21. Yasui N, Sato M, Ochi T, et al. Three modes of ossification during distraction osteogenesis in the rat. J Bone Joint Surg 1997;79-B:824 – 830. 22. Roberts W, Smith R, Zilberman Y,et al. Osseous adaptation to continuous loading of rigid endosseous implants. Am J Orthod 1984;86:95–111. 23. Marx R. Discussion: Histomorphometric examination of healing around hydroxylapatite implants in Co60-irradiated bone. J Oral Maxillofac Surg 1994;52:172–173. 24. Jacobsson M, Ka¨lebo P, Albrektsson T,et al. Provoked repetitive healing of mature bone tissue following irradiation: A quantitative investigation. Acta Radiol Oncol 1986;25:57– 62. 25. Shirota T, Ohni K, Suzuki K,et al. The effect of aging on the healing of hydroxylapatite implants. J Oral Maxillofac Surg 1993;51:51–56. 26. Emery S, Brazinski M, Koka A,et al. The biological and biomechanical effects of irradiation on anterior spinal bone grafts in a canine model. J Bone Joint Surg 1994;76-A:540 – 548. 27. Sawai T, Niimi A, Johansson C, et al. The effect of hyperbaric oxygen treatment on bone tissue reactions to c.p. titanium implants placed in free autogenous bone grafts. Clin Oral Impl Res 1998;9:384 –397. 28. Sugimoto M, Takahashi S, Toguchida J,et al. Changes in bone after high-dose irradiation. J Bone Joint Surg 1991;73-B:492– 497. 29. Widdman R, Pelker R, Friedlaender G,et al. Effects of prefracture irradiation on the biomechanical parameters of fracture healing. J Orthop Res 1993;11:422– 428. 30. Fox W, Aufdemorte T. Experience with osseous implants for bone and biomaterials research. J Long-Term Effects Med Implants 1993;3:1–27. 31. Rohrer M, Kim Y, Fayos J,et al. The effect of cobalt-60 irradiation on monkey mandibles. Oral Surg Oral Med Oral Path 1979;48:424 – 440.