The role of latency in mandibular osteodistraction

The role of latency in mandibular osteodistraction

Jom~al of Cranio-MaxiItojacial Surgery (1998) 26, 209 219 © 1998 European Association for Cranio-Maxillofacial Surgery The role of latency in mandibu...

8MB Sizes 1 Downloads 31 Views

Jom~al of Cranio-MaxiItojacial Surgery (1998) 26, 209 219 © 1998 European Association for Cranio-Maxillofacial Surgery

The role of latency in mandibular osteodistraction Kourosh Tavakoli, W. R. Walsh, Fiona Bonar, Richard Smart, Sue Wulf, Michael D. Poole

Department of Plastic and Reconstructive Surgery University of New South Wales and The St George Hospital, Sydney, Australia (Chief ProJessor M. D. Poole) S U M M A R Y . Even though osteodistraction has been well established in the extremities, the parameters used in craniofacial distraction have been essentially borrowed from orthopaedic experience. Latency is widely practised but its relevance has not been fully investigated. The purpose of this study was to establish the role of latency in mandibular distraction osteogenesis. Twenty-two growing Wethers sheep were allocated to four experimental groups. Six animals were allocated to each of Groups A, B and C and underwent bilateral mandibular corticotomies and attachment of an external lengthening device. Latent periods of 0, 4 and 7 days respectively were observed prior to beginning distraction. The distraction protocol consisted of a rate of 0.5 mm twice daily for 20 days, followed by a consolidation phase of 20 days after which the sheep were killed. Histology, bone densitometry and 3-point mechanical testing were performed on the harvested mandibles. Group D formed the control group (n=4). Histologically, the distracted bone exhibited bone formation primarily via intramembranous ossification with scattered islands of cartilage. The regenerated bone had mechanical properties significantly weaker than the undistracted control group (P<0.05), but between the experimental groups no statistically significant differences were demonstrable either in mechanical strength or DEXA density. These data indicate that a change in latency does not alter the properties of the regenerated bone in mandibular distraction osteogenesis and indeed no latent interval may be necessary at all in craniofacial distraction. This has implications for the duration of device fixation in distraction procedures.

INTRODUCTION

Wide variability both at the clinical and experimental level exists where delay periods of 0 to 14 days have been reported (Havlik and Bartlett, 1995; Chin and Toth, 1996; do Amaral et al., 1997). No group to date has specifically studied the issue of latency in the mandible. To this end, we embarked on this project to try and settle the debate over the significance of latency as it applies to mandibular osteodistraction.

Distraction osteogenesis, although initially developed for limb lengthening, is now being applied in the treatment of craniofacial deformities. Although the technique was initially described by Codivilla (1904), most basic science research into the biology of distraction osteogenesis was done by Ilizarov (1977 and 1989a, 1989b). Craniofacial surgery first embraced the concept of osteodistraction in 1973. Snyder et al. (1972) applied the technique of bone lengthening to the canine mandible. McCarthy et al. (1992) were able to apply distraction osteogenesis clinically to treat hemifacial microsomia. Since then, further work has demonstrated that distraction osteogenesis is an effective clinical tool with applications in the entire craniofacial skeleton (Rachmiel et al., 1993; Glat et al., 1994). The basic principles adopted in mandibular distraction osteogenesis have been essentially borrowed from the experience in orthopaedics (Ilizarov, 1989a and 1989b). Parameters such as latent period, ideal distraction rate and rhythm, and duration of the consolidation phase established by Ilizarov and others in tubular bones of the extremities have not been properly examined in the craniofacial skeleton (Chin and Toth, 1996). Latency represents a period of time when the distraction device is left intact after insertion on the patient's face prior to commencement of lengthening.

MATERIALS AND METHODS All experimentation was performed after authorization by the Animal Care and Ethics Committee of the University of New South Wales and complied with the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes (Australian Government Publishing Service, 1990). Twenty-two growing Wethers sheep were the subjects of the study. The animals were Species Pathogen Free and were all dipped and vaccinated two weeks prior to surgery. Eighteen animals were allocated to three experimental groups of 6 (A, B and C). The control group (D) comprised 4 sheep (Table 1). Insertion of the devices and corticotomy was performed under endotracheal halothane anaesthesia after initial induction using thiopentone. Stringent aseptic technique was observed at surgery and animals received intravenous cephalothin (Keflin), 30 mg/kg, 209

210 Journalof Cranio-MaxillofacialSurgery Table 1 - Summary of the experimentalprotocol

Group A Group B Group C Group D

Latent period

Distraction rate

Distraction rhythm

Consolidation period

Quantity

Days

mm

Per day

Days

6 6 6 4

0 4 7 N/A

0.5 0.5 0.5 N/A

BD BD BD N/A

20 20 20 N/A

Fig. 1 - Schematic drawing of the sheep skull demonstrating (A) the position of the distractor in relation to the interdentaI space and (B) mandibular lengthening post-distraction.

and intramuscular procaine penicillin (Benecillin), 15 mg/kg, at the time of induction of anaesthesia. The latter regime was repeated on a weekly basis until the animals were sacrificed. In Groups A, B and C, a sagittal midline submental incision was used to gain access to the interdental (ID) space on both sides of the mandible. The site of the proposed corticotomy was marked with Bonney's blue in the middle of the I D space, behind the mental foramen. Two Kirschner wires, (K-wires), 1.6 mm, alternating threaded and non-threaded were placed parallel through and through both sides of the mandible, two in front and two behind the corticotomy site. A distraction device was positioned loosely over the K-wires. Subperiosteal tunnelling gave access for the corticotomies using an air-driven drill and generous irrigation. The cortical osteotomy was next completed with narrow osteotomes taking care not to damage the neurovascular bundle or periosteal sheath. The distractor was secured and the pin sites dressed with Betadine ointment. Group D animals similarly received general anaesthesia and underwent soft tissue dissection around the I D space but neither K-wire insertion nor osteotomy was performed on the mandibles (Fig. la). The external fixation devices (Distraction Inc., Washington DC) used in this study were manufactured from aluminium and had a distraction capability such

that one 360 ° turn of the screws would amount to 0.5 m m lengthening. For Group A animals, distraction commenced immediately on the operating table. A delay period of 4 and 7 days was instituted for animals in Groups B and C respectively. Lengthening was performed at a rate of 1.0 mm/day in two increments of 0.5 ram, for 20 days. The placement of the distraction device was bilateral. This regimen was well tolerated by the animals without any requirement for analgesia. On completion of distraction, 20 days of consolidation followed with the device remaining in situ (Fig. 2). After the animals were humanely killed, the mandibles were split in the mid-sagittal and mid-coronal planes (at the level of the angle of jaw). The dissected mandibles were subsequently labelled as hemimandible A or B. Hemimandibles A and B were scanned using dualenergy X-ray absorptiometry (DEXA) in the lateral plane. Each hemimandible was positioned (lingual surface down) in a standard kidney dish filled with 500 ml normal saline solution. The D E X A scans were performed using a L U N A R E X P E R T system using an applied voltage of 134 kVp, a current 1 m A and a pixel size of 0.5 m m (the Left H a n d setting). The cust o m analysis setting was used to calculate the bone mass density (BMD) in the distraction sites.

The role of latencyin mandibular osteodistraction 211

Fig. 2 Sheepat completionof distractionperiod undergoinga routine devicecheck. For measurement, a 10 x 5.5 mm rectangular box was positioned overlying the distraction gap (View 1). A measurement of B M D in a standardized area in the centre of the distraction zone was thus possible. An identical rectangular box was measured immediately proximal to the distraction zone to be used as the normal (non-distracted) bone density for that animal (View 2). In the control group (D), bone density was measured in the ID gap of the mandible, corresponding to the site of the osteotomy and distraction in the experimental groups. The cortical bone was specifically not scanned in the controls. Hemimandibles B were mechanically tested using an MTS 858 servo-hydraulic materials testing machine (MTS Systems Corporation, Minneapolis, MN). The specimens, stored frozen in 0.9% NaC1 solution, were thawed on the day of testing to room temperature. Samples were tested in 3-point bending as previously reported by Elovic et al. (1994) at a rate of 2 mm/min. The hemimandible was placed lingual side down. The central loading point was aligned at the mid point. Load and displacement were collected and analysed using a personal computer. The data pertaining to peak load, stiffness and energy were obtained for all samples. The mode and site of failure was also recorded for each sample. Hemimandibles A were fixed in 10% buffered formalin and processed for light microscopy. The specimens were sectioned longitudinally in the axial plane with a slab of bone subsequently decalcified in 10% formic acid. Sections of the entire slab from anterior to posterior including the distraction zone were embedded in paraffin. Staining of 4 ~tm sections was done with haemotoxylin and eosin. Light microscopy examination, aided by polarization microscopy was then performed. The length of the distraction zone was measured as the intercortical distance. The type, extent and completeness of new bone formation and

other findings such as pin site infections were documented. Initially, the calculated means of all parameters per group were subjected to One-Way Analysis of Variance (ANOVA) with F as the test statistic. Subsequently, in those parameters where statistically significant results were found, multiple comparison test (Tukey's multiple range test) was applied in order to locate where the significant differences lay. Results from various parameters were then correlated using Pearson's Correlation Test in order to establish the reliability and consistency between various test results.

RESULTS

Clinical evaluation Mandibular lengthening led to prognathism (Fig. 1A, B and Fig. 3). Despite this, feeding was unaffected. Overall the protocol was well tolerated by the animals with a mean weight gain of 1.9 kg observed across the experimental groups. There was no mortality. A 33% pin track infection rate was observed across all the groups, This was treated satisfactorily with acetic acid (3%) dressings. At the time of sacrifice, gross examination of the lengthened mandible indicated bony union across the distraction zone, except in one case. Mandibular lengthening was confirmed radiographically with a generous amount of new bone seen across the distraction zones. Bony malunion was not observed (Fig. 4).

Histological findings After 3 weeks, the distraction zone was composed of a mixture of woven and maturing lamellar bony trabeculae lying in a loose fibrovascular stroma and

212 Joul7~alof Cranio-Maxillofacial Surgery

Fig. 3 - Facial CT scan (Lat.), 3-D formatted, demonstratingthe lengthened prognathicmandible (arrows).

arranged in a largely longitudinal manner within the gap (Fig. 5 A and B). Ongoing remodelling, manifested by osteoclastic and osteoblastic activity was evident. In 20% of the specimens, small foci of cartilaginous matrix could be seen but for the most part, ossification was intramembranous in nature (Fig. 5 C and D). Fibrous non-union with suppurative osteomyelitis was evident in three animals, one from each group (Fig. 5 E). In one case a large abscess occupied the distraction zone. There was no evidence of new bone formation and consequently the animal was excluded from further analysis. Other common findings were those of periosteal reaction and fibrocartilaginous metaplasia at the pin track sites adjacent to the distraction zone.

Fig. 4 Post-mortemX-ray (AP) of distracted mandibles showing the calcifiedlengthenedzone between the K-wires(arrows). specimens showed increased density values compared with the other groups. However, analysis of the distraction gap bone densities using one-way variance failed to reveal any statistical differences between the control (Group D) and experimental animals (Groups A, B and C). Furthermore no difference was found between the BMD values in the experimental groups (Fig. 7). The findings for the 'normal' bone of the premolar region, i.e. View 2, showed no significant differences either. Direct statistical comparison of B M D values in View 1 and View 2 of the same animal was not possible. In order to establish the relationship between Right vs Left, the one way variance test was applied which showed no statistically significant difference between distracted bone densities in the right and left hemimandibles of the same animal.

Bone densitometry The distracted zone was easily identified by D E X A scanning. The K-wire sites were also easily visible and less dense compared with the adjacent bone (Fig. 6). Bone formation was relatively symmetrical on the two sides of the mandible. Bone density in the distraction gaps was comparable to the interdental region of the non-operated animals. The mean recorded BMD values of the distraction zone for each group are given in Table 2. Group A

3-point mechanical testing The results of biomechanical analysis by 3-point bending are listed in Table 2. One way analysis of the results revealed a highly statistically significant difference in the mechanical results for peak load and stiffness between the experimental and control groups. Further resolution using the multiple variance analysis demonstrated that hemimandibles from control Group D were significantly stronger than distracted

Table 2 - Summaryof bone density and mechanical testing (mean) results Specimen

Group A Group B Group C Group D

DEXA

3-point Mechanical Testing

Bone density (View 1) g/cm2, (+/-SE)

Peak load N, (+/-SE)

Energy to failure J, (+/-SE)

Stiffness N/MM, (+/-SE)

0.5618 (+/-0.1078) 0.4513 (+/-0.0426) 0.4662 (+/-0.1501) 0.5077 (+/-0.0418)

385 (+/-144) 377 (+/-106) 345 (+/-119) 868 (+/-141)

740 (+/-484) 1062 (+/-436) 792 (+/-420) 1595 (+/-396)

154 (+/-80.2) 190 (+/-144) 121 (+/-77.0) 593 (+/-97.8)

The role of latency in mandibular osteodistraction 213

(A) showing cut end of cortex merging with longitudinal orientated new bone within distraction zone (H&E x 20). (B) Demonstrates combination of woven and lamellar bone in distraction site. Original cortical bone is entirely lamellar in nature (arrow) (Polarisation microscopyx 40). (C) D,istraction zone consists of ongoing ossification of fibrocartilaginousmatrix (H&E x 40) (D) Higher power view of (C) demonstrating ongoing osteoclastic and osteoblastic activity (H&E x 100). (E) Clear zone of fibrous non-union visible(H&E x 100). Fig. 5 -

hemimandibles as determined by peak load at failure and calculated stress levels during bending (P<0.05). However, further analysis of the experimental groups (A, B and C) failed to reveal any significance (Fig. 8 and Fig. 10) F-test was not found to be significant (P 0.13) for energy to failure (Fig. 9). In order to show any relationship a m o n g the 3point mechanical testing parameters, Pearson correlation coefficient was calculated for all possible pairs. Positive correlation was demonstrated (r +0.799) between peak load and stiffness. Furthermore, comparison of results of D E X A (View 1) and mechanical testing across each group revealed a positive correlation between that of peak

load and bone density (r +0.463) and stiffness and bone density (r +0.627).

DISCUSSION Although the experiments were a manipulation of regional anatomy, the prime concern was with the biology of bone deposition under different experimental conditions. There has been no attempt to study developmental growth, and alteration in morphology has been used as an endpoint only to quantify the degree of lengthening by distraction. We did not aim to establish the sequence of biological events

214

Journalof Cranio-Maxillofacial Surgery

Fig. 8 - 3-point mechanical testing: peak load in Newtons demonstrating significantly stronger bone in the control group D. P<0.05.

1062 1200 1000

....

~;ii~i!{~

80O

600

400

200

0

Fig. 6 D E X A scan of a lengthened mandible demonstrating bone density variations in View 1 (distracted bone) compared with View 2 (normal bone).

Fig. 9 - Three point mechanical testing: energy to failure in Joules. No significant changes across the four groups. P>0.13.

~9~ C3.5418 06

0.5077 0.4512

0,4622

O5

0.4

0.3 m 92

0.1

0

Fig. 7 - Bone mass density (mean) measurements across the experimental and control groups.

such as changes in the regional circulation that occur during latency. We intended to quantify the properties of the distracted bone at a single point in time. Thus the convention of analysing results at a fixed time after the conclusion of distraction was followed. It was found that a change in latency from 0 through to 7 days does not alter histological features, biomechanical properties and bone density mass of the regenerated bone analysed at 20 days post-distraction in the sheep mandibular model (Table 2).

Fig. 10 - 3-point mechanical testing: stiffness in Newtons/mm demonstrating significantly stronger bone in the control group D. P<0.05.

When using animal models for research, investigators must not only be aware of the limitations of these, but also recognize that the choice of animal might have a pronounced effect on the findings (Bardach and Kelly, 1987). For manipulations of craniofacial tissue, animals such as rabbits, dogs and sheep can be used, since humans, as mammals, share primitive bone and soft tissue responses with most other mammals (Siegel and Mooney, 1990).

The role of latency in mandibular osteodistraction

The anatomical, histological and surgical issues of using the sheep as a model for temporomandibular joint surgery have been previously described. Bosanquet and Goss (1987), concluded that the sheep skull exhibits similarities in shape, size and structure, providing a good comparative model for studying maxillofacial surgery in humans, both from the point of view of learning operative techniques and also studying the tissue characteristics. For research purposes, the use of sheep poses less ethical debate compared with dogs or non-human primates such as monkeys. The experience of other workers who have used sheep for cranio-maxillofacial osteodistraction has been similar to ours. They all concluded that the adult sheep is a good model for gradual bone expansion, being sufficiently large to allow craniofacial osteotomies to be performed. Sheep still in the growing phase provide a favourable model for comparison with the paediatric facial skeleton (Karaharju-Suvanto et al., 1990, 1992, 1996; Karaharju et al. 1993; Rachmiel et al., 1993, 1996). Although our experience with sheep was positive, some problems were encountered. The position of the distractor in the anterior portion of the mandible proved difficult with respect to maintaining adequate device hygiene and frame stability, due to the constant and rapid jaw movements, especially during feeding. However, these difficulties could be overcome. Many animal studies have followed the convention of adopting a delay prior to commencing distraction (Ilizarov 1989a, 1989 b; Aronson et al., 1988; Schumacher et al., 1994; Hamdy et al., 1995). It has been the generally accepted practice in orthopaedic surgery to observe a latent period in lower limb lengthening procedures (Kojimoto et al., 1988; Delloye et al., 1990). The recommended duration has been between 5 days (Ilizarov, 1989) and 10-15 days (DeBastiani et al., 1987). However, there has been

Table 3

215

wide variability in the latent period, in the literature on limb lengthening from a mere 36 h to 15 days (de Pablos et al., 1986; Aldergheri et al., 1989). White and Kenwright (1991) have recommended that the ideal timing should be determined by the surgeon based on the type of bone chosen for distraction, the site of osteotomy, the degree of accompanying soft tissue trauma and the age of the patient. For example, if the osteotomy performed has been extensive with very little medullary preservation in the older patient, then a longer delay period should be allowed. On the other hand, in the event of too much delay prior to starting distraction, there is the risk of premature union of the bone fragments and inability to achieve the desired length. In craniofacial surgery, the role of latency has never been resolved, Cousley and Calvert (1997), with wide variability existing in the literature both in experimental and clinical situations (Tables 3 and 4). Snyder et al. (1972), the first to apply distraction osteogenesis to the mandible in canines, used a delay period of 7 days prior to commencement of distraction. Karp et al. (1990), in their canine series investigating the histological changes during mandibular lengthening, allowed 10 days for the latency period. Conversely, Califano et al. (1994) demonstrated that after only a 12 h latent period, increased blood flow was observed in the distracted zone of the rabbit mandible. Since McCarthy's first report on the application of the distraction technique to craniofacial surgery, the clinical literature has been as diverse as the animal work. The lag period that he adopted at both clinical and experimental levels was 7 days (McCarthy et al., 1992). The latency period ranging from 4 to 7 days seems to be the current practice of many workers in the field of mandibular distraction (Perrot et al., 1993; Klein and Howalt, 1995; Rachmiel et al., 1995; Molina and Monasterio, 1995). However, a 14 day latency

Delay periods adopted for craniofacial osteo-distraction in animal experiments

Experimental Studies

Latent period

Mandible

Model

Age

Days

Snyder et al., 1972 Karp et al., 1990, 1992 Constantino et al., 1993 Karaharju-Suvanto et al., 1990, 1992, 1996 Block et al., 1993 Califano et al., 1994 Ganey et al., 1994 Guerrissi et al., I994 Fischer et al., 1997 Midface and cranium

Canine Canine Canine Sheep Canine Rabbit Canine Rabbit Canine

Unknown Growing Adult Growing Unknown Adult Adult Adult Growing

7 10 10 5 7 2 10 7 7

Rachmiel et al., 1993 Glat et al., 1994 Lalikos et al., t 995 Rachmiel et al., 1996

Sheep Canine Rabbit Sheep

Adult Growing Growing Adult

4 7 7 7

216 Journalof Cranio-MaxillofacialSurgery Table 4 Widevariabilityin latencyperiod reported in clinicalpractice Clinical studies

Conditions

McCarthy et al., 1992

Hemifacialmicrosomia(HFM) Nager's Syndrome Hypoglossia-hypodactylysynd HFM TMJ Ankylosis HFM Micrognathia HFM TMJ Ankylosis Moebius Syndrome Nager's Syndrome Rubenstein-TaybiSyndrome Cerebro Costo mandibular synd. HFM HFM HFM Treacher Collins Syndrome Piere Robin Syndrome TMJ Ankylosis Mandibular hypoplasia HFM HFM Goldenhar Syndrome Pfiefer's Syndrome Apert Syndrome Crouzon's disease

Perrot et al., 1993 Takato et al., 1993 Habalet al., 1994 Havlik and Bartlett, 1994 Klein and Howalt, 1995 Losken et al., 1995

Rachmielet al., 1995 Molina and Monasterio, 1995

Pensler et al., 1995 Diner et al., 1996 Chin and Toth., 1996 do Amaral et al., 1997

time has been reported by others (Takato et al., 1993; Havlik and Bartlett, 1994). More recently in a clinical trial from Brazil, do Amaral et al. (1997) successfully distracted the cranium in seven patients suffering from craniosynostosis with no latent period. The concept of latency was frequently observed by the early pioneers of distraction osteogenesis (Barr and Obey, 1933; Bosworth, 1938; Abbott and Saunders, 1939). Bosworth (1938), described latency as a period during which soft tissue inflammation could resolve. Latency was challenged by Kawamura et al. (1968), who questioned its importance in the presence of minimal soft tissue dissection. In more recent times, the term 'callostosis' has been coined in order to describe the process that occurs during latency, where increased bone proliferation has been observed (De Bastiani et al., 1987). It is believed to parallel the initial phase of fracture healing, during which the fibrinenclosed haematoma and inflammatory cells infiltrate the gap at the corticotomy site. The role of latency has been examined in a study of tibial osteodistraction in the rabbit model and greater extraosseous circulation and increased bone deposition was demonstrated by means of microangiography, histology and radiography in the lengthened tibia in the group with a 7 day delay period compared with no delay (White and Kenwright, 1990). The necessity for a delay period seems to be strongly linked to the state of intramedullary blood supply. Ilizarov, although a proponent of the general need for latency, did question its importance in the presence of intact endosteum and periosteum (Ilizarov

Latent period: Days 7 7 14 7 14 5 5

7 4

2 3 05 0

and Pereslytskikh, 1977). However, the preservation of endosteum has been shown not to be an absolute prerequisite to optimum bone regeneration (DeBastiani et al., 1987). The corticotomy technique utilized in our and other studies tends to preserve the intramedullary blood supply to a greater degree than was previously achieved with traditional osteotomy. In the latency study by White and Kenwright (1990), however, an osteotomy approach rather than medullary conserving corticotomy was adopted. This as acknowledged by the authors would have compromised the blood supply of the bone fragments and possibly influenced the healing pattern, and this raises questions regarding the emphasis put on the role of extraosseous circulation. Another interesting aspect of White and Kenwright's work was the relative instability of their lengthening frames which may have contributed towards their observations. The importance of an intact periosteal envelope over the osteotomy site has been previously documented (Kojimoto et al., 1988; Delloye et al., 1990). It has been proposed that as long as the periosteal sheath has not been severely damaged by factors such as power saw injury, it can contribute to bone regeneration (Tetsworth and Paley, 1995). In our study, the periosteum was preserved by creation of a sub-periosteal tunnel for the corticotomy to be performed, taking care not to damage the periosteum. In craniofacial surgery the issue of latency is further clouded by the introduction of a new set of variables. Generally, blood supply to the head and neck area is superior to that encountered in the extremities.

The role of latencyin mandibularosteodistraction 217 According to Aronson (1994), the regional vascular supply is, physiologically, perhaps the most vital factor in osteogenesis. Rapid healing seen in the craniofacial region compared with the lower limb has been generally attributed to a greater plexus of vessels and more abundant perforators. We believe that expeditious vascular regeneration is probably a crucial factor in our finding that there is no advantage from latency in mandibular osteodistraction. Furthermore, the intramembranous bones of the facial skeleton have different qualities to the endochondral bones (tubular) of the axial skeleton (Karaharju-Suvanto et al., 1990). Also, the axial forces acting in lower limb lengthening are not directly relevant to craniofacial distraction. Histological findings in the majority of specimens in this study revealed ossification had resulted in union across the distraction gap and the presence of osteoblastic and osteoclastic activity indicated ongoing bone deposition and remodelling (Fig. 5). Our findings are similar to those of Karaharju-Suvanto et al. (1990, 1992) from Finland, who are the only group to date who have reported osteodistraction on the sheep mandible. In the experience of that group, histological results at a similar stage post-distraction indicated that the new bone was laid down from the margins towards the centre of the gap, where the collagen bundles formed a frame for the new bone to form across the distraction gap. The direction of the bundles was longitudinal in the direction of distraction. Similar results were found in this study (Fig. 5). Despite the similarities there remained some differences in our study compared with the Finnish study. The sheep in our experiment were growing adults (12-14 months) compared with much younger animals (13.5-19 weeks) in the Finnish study. The position of the osteotomy in the ascending ramus also differed from the interdental gap chosen in the study by us (Karaharju-Suvanto et al., 1992). They allowed a latent period of 5 days and a distraction rate of 0.5-1.0 mm/day. The duration of consolidation phase at 35 days was slightly longer than the 20 days in the present study. The prevalence of cartilaginous areas in the distraction zone reported by Karaharju-Suvanto et al. (1990, 1992) is, as they concede, likely to be explained by their less stable unilateral fixation system and presence of infection, although the cartilaginous islands did eventually disappear in long-term follow-up. The process of ossification in our study was predominantly intramembranous despite the scattered presence of cartilage consistent with other similar studies (Karp et al., 1990; Komuro et al., 1994). In our study ongoing remodelling of these islands was present. In the tibial latency study by White and Kenwright (1990), fibrosis and scar tissue formation was evident distally along the distraction segment in the no delay group, but in our study the histological findings were similar to their 7 day delay group which demonstrated

woven bone and intervening cartilage. However, cartilaginous components were more prominent in tibial osteodistraction. In applying DEXA to craniofacial distraction, reproducible orientation is required to ensure consistent results. By systematically sampling a standard box it was hoped to reduce observer bias to a minimum (Fig. 6). No difference was recorded in the amount of bone deposited in the distraction gaps in the control and experimental groups (Table 2 and Fig. 7). Analysis of data in the latency groups also failed to reveal any statistically significant difference. It is noteworthy that in the control group the bony cortex was deliberately not scanned, which could explain the lack of difference when compared with the distracted bone. It may be that DEXA scanning was an insufficiently sensitive method of analysis to detect any difference, but others who have used it to demonstrate a positive effect of pulsed electrical stimulation applied via distraction pins use the DEXA method (Hamanishi et al., 1995). DEXA has also been used to measure the quantity and rate of formation of bone during leg lengthening (Eyres et al., 1993; Maffulli et al., 1997). The control group D hemimandibles were significantly stronger as determined by peak load at failure and stress levels during bending (Table 2, Fig. 9 and Fig. 10). This finding indicates that at 20 days post distraction, healing in the distraction gaps was not sufficiently advanced for the bone's biomechanical properties to replicate those of normal bone. This is also consistent with the histological observation of ongoing intense osteoblastic and osteoclastic activity within the distraction gap. As with DEXA analysis, no differences between the experimental groups were identified by 3-point bending; however, a positive correlation was demonstrated between the two investigative modalities adding weight to the consistency of the observed results. Energy to failure is a function of peak load, stiffness and elongation. The lack of significantly different outcome in energy values (Fig. 9) is perhaps due to the closeness of the other bending results (Fig. 8 and Fig. 10). Reports on the mechanical and physical properties of the mandible have been scarce (Elovic et al., 1994). There have been a few reports on the mechanical properties of distracted bone and callus in the limb lengthening literature (Schinckendantz et al., 1992; Walsh et al., 1994). Constantino et al. (1993), using bifocal distraction osteogenesis, estimated the strength of regenerated bone to be 77% of that of the normal bone. It was concluded that the strength of distracted bone could be accurately calculated using the biomechanical data. Bending experiments and torsion are frequently used to examine the properties of the fractured whole bone because they are convenient and expeditious (Black et al., 1984). However, bending and torsion involve complex internal stress fields and produce

218

Journal o f Cranio-Maxillofacial Surgery

results that are strongly dependent on testing skill and orientation of the test specimen (Walsh et al., 1994). Small variations in technique and surgical factors such as the degree of periosteal/endosteal preservation, differences in biomechanical environment in the lengthening sites between animals (Orbay et al., 1992) and general variability found amongst sheep of the same breed may in part contribute to the spectrum of bone properties observed. The process of osteodistraction can take up to 40 50 days to complete. The necessity for latency has been questioned (Karaharju-Suvanto et al., 1992) particularly by those dealing with paediatric maxillofacial surgery (Chin and Toth, 1996). The wide variation in latent periods in reports in the literature on craniofacial distraction osteogenesis demonstrates the lack of data in this area (Table 4). Our data suggests that latency alone does not have a significant influence on the physical and mechanical properties of mandibular distracted bone (Table 2) and that the traditional practice of observing a latent interval in craniofacial osteodistraction may not be necessary. In the clinical arena the surgeon, rather than arbitrarily designating a time frame before the start of lengthening should probably base his/her decision upon patient factors such as age and compliance. ACKNOWLEGEMENT The authors would like to acknowledge Dr Adele Awiss-Morsi & Mr John Rowlinson for their assistance with care of the animals; Dr Norman Cowen's distraction devices (Distraction Inc., Washington D.C.); Dr Thomas Martin for veterinary advice; Dr Borhan Uddin for assistance with statistical analysis; Ms Julia Janssen for technical aid with DEXA scanning; Mr Bob Haynes for illustrations; Mr Grant Taggart and Mr Mark Bromley for histotechnical expertise; and Ms Poppy Krimizis for administrative assistance.

REFERENCES Abbott, L. C., J B. de C. M. Saunders: Operative lengthening of tibia and fibula: a preliminary report on further development of principIes and technic. Ann. Surg. 110 (1939) 961-91. Aldegheri, R., L. Renzi-Brivio, S. Agostini: The callotasis method of limb lengthening. Clin. Orthop. 241 (1989) 137-45. Aronson, J, B. H. Harrison, C. L. Stewart, J H. Harp: The histology of distraction osteogenesis using different external fixators. Clin. Orthop. 241 (1988) 106-16. Aronson, J: Temporal and spatial increases in blood flow during distraction osteogenesis. Clin. Orthop. 301 (1994) 124-131. Bardach, J., K, Kelly: Role of animal modeIs in experimental studies of the craniofacial growth following cleft lip and palate repair. Cleft Palate J. 25 (1987) 103. Barr, J S., E R. Ober: Leg lengthening in adults. J. Bone. Joint Surg. 15 (1933) 674-678. Bosanquet, A., A. Goss: The sheep as a model for temporomandibular joint surgery. Int. J. Oral Maxillofac. Surg. 16 (1987) 600 603. Bosworth, D. M. : Skeletal distraction of the tibia. Surg. Gynecol. Obstet 66: (1938) 912 921. Califano, L., A. Cortesse, A. Zupi, G Tajana: Mandibular lengthening by external distraction: an experimental study in the rabbit. J. Oral Maxillofac. Surg. 52 (1994) 1179-1183.

Chin, M., B. A. Toth: Distraction osteogenesis in maxillofacial surgery using internal devices: a review of five cases. J. Oral. Maxillofac. Surg. 54 (1996) 45-53. Codivilla, A.: On the means of lengthening, in the lower limb the muscles and tissues which are shortened through deformity. Am. J. Orthop. Surg. 2 (1904) 353 369. Costantino, P. D., C. D. Friedman, M. L. Shindo, C. G Houston, G A. Sisson : Experimental mandibular regrowth by distraction osteogenesis. Arch. Otolaryngol. Head Neck Surg. 119 (i 993) 511-516. Cousley, R. R. J., ~L L. Calvert. Current concepts in the understanding and management of hemifacial microsomia. Brit. J. Plast. Surg. 50 (1997) 536-551. DeBastiani, G., R. Aldegheri, L. Renzi-Brivio, G Trivella. Limb lengthening by callus distraction (callotasis). J. Pediatr. Orthop. 7 (1987) 129 134. de Pablos, J, C. Villas, J Canadell: Bone lengthening by physeal distraction. An experiment study. Int. Orthop. 10 (1986) 163. Delloye, C, G. Delefortrie, L. Couteliet, A. Vincent: Bone regenerate formation in cortical bone during distraction lengthening. Clin Orthop 250 (1990) 34-42. do Amaral, C. M. R. et al.: Gradual bone distraction in craniosynostosis. Scand. J. Plast. Reconstr. Hand Surg. 31 (1997) 25-37. Elovic, R. P., J. A. Hipp, W. C Hayes: A method for measuring the structural properties of the rat mandible. Arch. Oral Biol. 39 (1994) 1029-1033. Eyres, IC S., M. J. Bell, J A. Kanis: New bone formation during leg lengthening. Br. J. Bone Joint Surg. 75 (1993) 96-106. Glat, P M., D. A. Saffenberg, N S. Karp, R. A. Holliday, G. Steiner, J G. McCarthy: Multidimensional distraction osteogenesis the canine zygoma. Plast. Reconstr. Surg. 94 (1994) 753-758. Hamdy, R. C., V~ Walsh, M, Olmedo, M. Wallach, M. G Ehrlich: Correlation between ultrasound imaging and mechanical and physical properties of lengthened bone: an experimental study in a canine model. J. Pediatr. Orthop. 15 (1996) 206-211. Haminishi, C., ~ Kawabata, 7~ Yoshii, S. Tanaka: Bone mineral density changes in distracted callus stimulated by pulsed direct electrical current. Clin. Orthop. 312 (1995) 247-251. Havlik, R. J., S. P. Bartlett: Mandibular distraction lengthening in the severely hypoplastic mandible: a problematic case with tongue aplasia. J. Craniofac. Surg. 5 (1994) 305 312. Ilizarov, G A., P. F Pereslytskikh: Regeneratsiia kostnatkani diafiza pri udlinenni posle zalryoti napravlennoi kosoi lil vintoobraznoi osteokazii. Vestn. Khir. 119 (1977) 89-93. Illizarov, G A: The tension-stress effect on the genesis and growth of tissues: Part 1. The influence of stability of fixation and soft-tissue preservation. Clin. Orthop. 238 (1989) 249-281. Ilizarov, G A: The tension-stress effect on the genesis and growth of tissues: Part 2. The influence of the rate and frequency of distraction. Clin. Orthop. 239 (1989) 263-285. Kamharju-Suvanto, ~, E. O. Karaharju, R. Ranta: Mandibular distraction; an experimental study in sheep. J. CranioMaxillofac. Surg. 18 (1990) 280 283. Karaharju-Suvanto, T., J Peltonen, A. Kahri, E. O. Karaharju: Distraction osteogenesis of the mandible. J. Oral Maxillofac. Surg. 2 (1992) 118 121. Karaharju, E. 0., K. Aalto, A. Kahri et al.: Distraction bone healing. Clin. Orthop. 297 (1993) 3843. Karaharju-Suvanto, Z, J Peltonen, O. Laitinen, A. Kahri: The effect of gradual distraction of the mandible on the sheep temporomandibular joint. Int. J. Oral Maxillofac. Surg. 25 (1996) 152-156. Karp, NS., C. H. M. Thorne, J G McCarthJ; 11. A. Sissons: Bone lengthening in the craniofacial skeleton. Ann. Plast. Surg. 24 (1990) 231-237. Kawamura, B., S. Honsono, T. Takahashi, T. Yano, Y. Kobayashi, JV Shibata, Y Shinoda: Limb lengthening by means of subcutaneous osteotomy. J. Bone Joint Surg. Am. 50 (1968) 851-77. Klein, C., H. Howaldt: Lengthening of the hypoplastic mandible by gradual distraction in childhood - a preliminary report. J. Cranio-Maxillofac. Surg. 23 (1995) 68-79. Kojimoto, H., N Yasui, T. Goto, S. Matsuda, Y Shimomura: Bone lengthening in rabbits by callus distraction: the role of periosteum and endosteum. Br. J. Bone Joint Surg. 70 (1988) 543-549.

The role of latency in mandibular osteodistraction Komuro, Y, T. Takato, I~ Harii, Y Yonemara: The histologic analysis of distraction osteogenesis of the mandible in rabbits. Plast. Reconstr. Surg. 94 (1994) 152-159. Maffulli, N, J. C. Y Cheng, A. Sher, T. t~ Lain: Dual-energy. X-ray absorptiometry predicts bone formation in lower limb callostosis lengthening. Ann. R. Coll. Surg. Engl. 79 (1997) 250-256. McCarthy, J. G, J. Schreiber, N Karp, C. H. Thorne, B. H. Grayson: Lengthening of the human mandible by gradual distraction. Plast. Reconstr. Surg. 89 (1992) 1-10. Molina, F , F O. Monasterio: Mandibular elongation and remodeling by distraction: a farewell to major osteotomies. Plast. Reconstr. Surg. 96 (1995) 825-840. National Health and Medical Research Council, Commonwealth Scientific and Industrial Research Organisation, Australian Agricultural Council: Australian code of practice for the care and use of animals for scientific purposes. Australian Government Publishing Service, Canberra. 1990, 1-75. Orbay, J L., V H. Frankel, J. E. Finkle, t~ Kummer: Canine leg lengthening by the Ilizarov technique. A biomechanical, radiological and morphological study. Clin. Orthop. 278 (1992) 265-273. Perrott, D. H., R. Berger, K. Vargervik, L. B. Kaban: Use of a skeletal distraction device to widen the mandible: a case report. J. Oral Maxillofac. Surg. 51 (1993) 435439. Rachmiel, A., Z. Porparic, L 7~ Jackson, T. Sugihara, L. Clayman, J S. Topf R. A. Forte: Midface advancement by gradual distraction. Br. J. Plast. Surg. 46 (1993) 201-207. Rachmiel, A., M. Levy, D. Laufer: Lengthening of the mandible by distraction osteogenesis: report of cases. J. Oral Maxillofac. Surg. 53 (1995) 838-846. Rachmiel, A., M. Levy., 12 Laufer, L. Clayman, L T Jackson: Multiple segmental gradual distraction of facial skeleton: an experimental study. Ann. Plast. Surg. 36 (1996) 52-59. Schickendantz, M. S., Y. T Watson, J. J Sferra, H. E. Kambic: A model for evaluating the strength of bones lengthened by distraction osteogenesis. Clin. Orthop. 275 (1992) 248-252.

219

Schumacher, R, J. Keller, I. Hvid: Distraction effects on muscle. Acta Orthop. Scand. 65 (1994) 647-650. Siegel, M. L, M. P Mooney: Appropriate animal models for craniofacial biology. Cleft Palate J. 27 (1990) 18-25. Snyder, C. C:, G A. Levine, H. M. Swanson, E. Z. Browne: Mandibular lengthening by gradual distraction; preliminary report. Plast. Reconstr. Surg. 51 (1972) 506-508. Takato, T., K. Harii, S. Hirabayashi, E Komuro, Y. Yonehara, 7~ Susami: Mandibular lengthening by gradual distraction: analysis using accurate skull replicas. Br. J. Plast. Surg. 46 (1993) 686-693. Tetsworth, K., D. Paley. Basic science of distraction histogenesis. Curr. Opin. Orthop. 6 (1995) 61 618. Walsh, W. R., R. C. Hamdy, M. G. Ehrlich: Biomechanical and physical properties of lengthened bone in a canine model. Clin. Orthop. 306 (1994) 230 238. White, S. H., J Kenwright: The timing of distraction of an osteotomy. Br. J. Bone. Joint. Surg. 72 (1990) 356-361. White, S. H,, J Kenwright: The importance of Delay in Distraction of Osteotomies. Orthopaedic Clinics of North America 22 (1991) 569-579.

Professor M. D. Poole

Department of Plastic and Reconstructive Surgery St George Hospital 30 Gray Street Kogarah Sydney, NSW 2217 Australia Tel: 0011-612-9350-2755; Fax: 0011-612-9350-2756; Email: R Krimizis @UNSW. edu.au Paper received 9 March 1998 Accepted 12 June 1998