Int. J. Oral Maxillofac. Surg. 2006; xxx: xxx–xxx doi:10.1016/j.ijom.2006.03.018, available online at http://www.sciencedirect.com

Research Paper Distraction Osteogenesis

Mineralization and mechanical properties of the canine mandible distraction wound following acute molding

C. Kunz1,2, N. Adolphs1,2, P. Bu¨scher1,2, B. Hammer1,2, B. Rahn1,2 1 The AO Research Institute Davos, Switzerland; 2The Clinic for Reconstructive Surgery, Department of Oral and Craniomaxillofacial Surgery, University Hospital Basel, Switzerland

C. Kunz, N. Adolphs, P. Bu¨scher, B. Hammer, B. Rahn: Mineralization and mechanical properties of the canine mandible distraction wound following acute molding. Int. J. Oral Maxillofac. Surg. 2006; xxx: xxx–xxx. # 2006 International Association of Oral and Maxillofacial Surgeons. Published by Elsevier Ltd. All rights reserved. Abstract. To investigate the influence of an acute single step callus manipulation immediately after distraction on mechanical properties and mineralization of the regenerate, custom made distraction devices were bilaterally placed in the mandibular angle of 15 beagle dogs, allowing to simultaneously compress and stretch the regenerate after completed linear distraction. The animals were divided in three groups (n = 5): Group 1 and 2 underwent manipulation of the regenerate, group 3 remained in a linear position. After 42 (group1) and 90 (group 2 and 3) days of consolidation the animals were sacrificed. The mechanical properties were assessed in an Instron1 testframe and bone density quantified by quantitative computed tomography and three- dimensionally assessed (Scion Image1 processing and analysis software). After 6 weeks of consolidation 25% of the specimens reached a stiffness which was 90% of the mean values of the unoperated reference hemi-mandibles. After a 13 week consolidation period, 62.5% were as stiff as the referenced specimens. Manipulated regenerates, allowed to heal under stable conditions for 13 weeks, had the same mechanical properties as specimens that underwent pure linear distraction. A temporary but not significant delay of osseous healing had to be postulated for the stretched zone after 6 weeks, indicating this area to be more critical than the compressed area.

Distraction osteogenesis represents a unique form of clinical tissue engineering for early correction of inborn malformation or reconstruction of bone defects after trauma or tumour resection. Lengthening of craniofacial bones is more complex 0901-5027/000001+06 $30.00/0

than in long bones because the bony deformity has to be corrected by shape and size, which makes multidimensional distraction necessary, especially for the correction of the hypoplastic mandible. In many publications11,12,16,23,24,26,27, as

Key words: distraction osteogenesis; callus manipulation; mechanical properties. Accepted for publication 6 March 2006

well as in the research of the current authors15, it has become evident that spatial vector control during multiplanar distraction is difficult and even multidimensional devices may not avoid deviations from planning4–6.

# 2006 International Association of Oral and Maxillofacial Surgeons. Published by Elsevier Ltd. All rights reserved. YIJOM-935; No of Pages 6

2

Kunz et al.

Distraction osteogenesis is a dynamic process interacting with changing soft tissue resistance and vector forces14. Thus, control of distraction vectors during a 3dimensional bone lengthening procedure for correction of severe mandibular deformities is a problem of major impact. Even if arced distraction is possible8, meticulous planning and multiplanar devices may not avoid deviations from preplanned results. Manipulations of the newly created regenerate, during or after the distraction procedure, may be necessary to correct the position of the mandible. It is the purpose of this study to investigate the mechanical characteristics and mineralization processes of a manipulated regenerate and to answer the question: does callus manipulation interfere with undisturbed osseous healing? Material and methods

Bilateral osteotomy of the mandibular angle was performed in the retromolar region of 15 fully grown female beagles (average age 13.1 months) with an average weight of 12.2 kg. To prevent infection, the molars M2 and M3, teeth with tiny roots, were extracted 6 weeks before. The complete osseous recovery was confirmed radiologically in a pilot series. Custommade distraction devices, facilitating defined angulation, were placed parallel to the sagittal plane on the mandibular angle (Fig. 1) and fixed with two 2-mm stainless steel Kirschner wires (Synthes1) on each side of the osteotomy. A custommade pointer mounted on the fulcrum of the device provided an accurate intraoperative positioning. Contact radiographs were taken before and after insertion of the device as well as before and after callus angulation using occlusal films (Kodak Ectaspeed Plus EO 41P1, Exposure 55 kV/25 ms, ATS Arco si 1001). After 5 days latency, distraction was initiated at a rate of 1  1 mm/day (0.5 mm on each side of the fulcrum). After regenerating 10 mm of bone, defined angulation of 208 was performed the day after distraction ended. Angulation was achieved simultaneously on both sides in an acute single-step procedure under general anaesthesia. The mechanical properties of the distracted and angulated hemi-mandibles were compared with a control group with linear distraction only. The following groups were assessed: group 1 (n = 5): linear distraction 10 mm, angulation 208, consolidation time 6 weeks;

Fig. 1. Protocol of distraction and angulation. (a) The custom-made distraction device is placed in the centre of the osteotomy; (b) after a 5-day latency period linear distraction at a rate of 1  1 mm/day follows until a regenerate of 10 mm of length is created; (c) immediately after the end of distraction the fresh regenerate is angulated. Harvesting of the hemi-mandibles was performed after 6 and 13 weeks of osseous consolidation.

group 2 (n = 5): linear distraction 10 mm, angulation 208, consolidation time 13 weeks; group 3 (n = 5): linear distraction10 mm, no angulation, consolidation time 13 weeks. Five intact, non-distracted beagle hemimandibles were chosen as a reference and underwent identical investigations. Reproducible contact radiographs of the hemi-mandibles using a Faxitron 804 unit (Faxitron Company, Illinois, USA) and an AGFA Structurix D4 Film (exposure 45 kV/5 min, Focus 53 cm) developed in an AGFA Structurix NDT M developing machine served for morphological assessment of the regenerate. The reproducibility was monitored by means of an aluminium step wedge which was included in each radiograph. The same contact radiographs of entire specimens were digitally imaged (Minolta RD 175, Macro lens 50 mm) using reproducible illumination. The radiological density of mineralized tissues within the distraction zone was then visualized using the Scion Image1 processing and analysis program (Scion Corporation, Maryland, USA, www.scioncorp.com), whereby x- and y-axes render the geometry of the projected hemi-mandible and the zaxis represents regional radiological density. Entire specimens underwent quantita-

tive computed tomography (Densiscan 10001 apparatus, Scanco Medical, Zu¨rich, Switzerland). The specimens were scanned in the manipulated area and the adjacent bone in 1 mm intervals in a sagittal plane. Data were recorded as a linear attenuation coefficient. For calibration of the attenuation coefficient the instrument uses an internal standard, which is monitored at intervals using an external phantom. For mechanical testing an Instron1 Compression/Tension test frame Type 4302 (Instron Corp, UK) with a load cell of 1 kN and compression test method was used. After dissection of the periosteum, the distal dentate segments of the harvested specimens were potted into a block of methylmethacrylate (Beracryl1) and fixed in the testing machine. Load was applied via a transosseous nail through the proximal segment. Loading was simulating an occlusal load and, after turning the specimens by 1808, simulating a basal load (Fig. 2). In order to avoid irreversible deformation and breaking, it was limited to the elastic range by manual interruption of the loading process as soon as the load/displacement curve deviated from its linear course. To measure stiffness, the slope of the force (N) versus displacement (mm) curve was determined in the linear portion of the curve, neglecting the initial non-linear portion produced by

Fig. 2. After embedding the hemi-mandibles, load is applied via a transosseous nail in a direction simulating an occlusal load (drawing) and at 1808 to this direction. The distance of the nail from the potting is standardized. The centre of rotation, not precisely definable, is assumed to be within the regenerate.

Mechanical properties after callus manipulation the adjustment mechanisms between machine and specimen. The stiffness values were represented as a box plot of each group (Fig. 4). After mechanical testing the specimens were transferred to histological processing. Statistical evaluation was performed with SAS Version 8. Comparisons between different treatment conditions were tested in a non-parametric way by the 2-sample Wilcoxon test and Kruskal– Wallis test, respectively. Comparison within mandibles was examined using Wilcoxon sign rank tests. Correlations were calculated by Spearman rank test. Each mandible side was treated as a statistical unit of its own, assuming that the effects of treatment were based on local conditions and not correlated within the 2 mandibular halves. Apart from the density outcomes, vascularization and mechanical stiffness did not, in fact, correlate significantly between respective mandibles. Results

No intraoperative complications occurred. One animal from group 2 died of a mechanical ileus caused by a foreign body, swallowed outside the barn. Minor postoperative complications were noticed in 2 dogs suffering from temporary superficial pin site infections which could be controlled by local treatment. By angulation a very symmetric degree of manipulation was achieved. The preplanned angle was realized in most cases (208  38 in 95%). Scion Image1 analysis of the intact specimens demonstrated that, in the case of undisturbed osseous healing, the density of the regenerate reached the same level as the normal adjacent bone in groups 2 and 3 after a 13-week consolidation time (Fig. 3). Apparently compressed crestal zones showed lower density values as compared to basal zones in group 3 but not in groups 1 and 2 (Wilcoxon sign rank test, P < 0.004). In group 1, after 6 weeks of consolidation, overall gap density was lower as compared to group 2 (P = 0.05) and group 3 (P = 0.09) (Table 1). Likewise, density within the compressed area as well as in the stretched area was lower in group 1 as compared to group 2 (P = 0.01 and 0.04, respectively) and group 3 (P = 0.34 and 0.03, respectively). If disturbance of ossi-

3

Fig. 3. Visualization of the density of mineralized tissues based on contact radiographs of the entire specimens using the Scion Image1 analysis program. x- and y-axes render geometry of the projected hemi-mandible, z-axis represents regional radiological density. (a) Regenerate with 208 angulation and a consolidation period of 13 weeks after uneventful osseous healing showing the same overall density as the adjacent bone. The central depression is caused by the decreased density in the region of the alveolar channel (arrow); (b) Regenerate after linear distraction without angulation after a 13-week consolidation period and uneventful healing revealing slightly less osseous density at the basal periphery; (c) density of the regenerate is slightly decreased when compared with the adjacent bone after a 6-week period of consolidation and uneventful osseous healing. When delayed ossification occurred, it was observed in the stretched basal zone (arrow).

fication was detected by histological and contact radiological analysis, it was visible within the stretched basal zone, indicating that this area is more critical than the compressed zone (Fig. 3c). Under occlusal and basal loading conditions, mean overall stiffness of group 1 (mean 156.2 N/mm occlusal load, 156.0 N/mm basal load) with a 6-week consolidation period was lower than in group 2 (mean 292.5 N/mm occlusal load, 264.2 N/mm with basal load, P = 0.07 and 0.04, respectively) and group 3 (mean 202.0 N/mm occlusal load, 188.7 N/mm basal load, P = 0.32 and 0.74, respectively) with 13 weeks of consolidation (Fig. 4). The variation was lower in group 1 compared with groups 2 and 3. Group 2 (angulated regenerate, 13 weeks consolidation) revealed a higher stiffness but this was not significant compared to group 3 (no angulation, 13 weeks consoli-

dation): P = 0.21 for occlusal zone and P = 0.14 for basal zone, respectively. Both 13-week consolidation groups showed a higher variation than group 1 (Fig. 4). The measurements demonstrated that manipulated regenerates, allowed to heal under stable conditions, have the same mechanical properties as specimens after a pure linear distraction. After 6 weeks of consolidation, 25% of the specimens reached a stiffness which was 90% of the values of the non-operated reference hemi-mandibles (223.5 N/mm occlusal load, 251.2 N/mm basal load). After a 13-week consolidation period, 62.5% of the manipulated hemi-mandibles and 55% of the linear lengthened specimens were as stiff as the referenced specimens. Low stiffness after 13 weeks of consolidation (groups 2 and 3) was always related to disturbed healing after pin instability at the bone/pin interface.

Table 1. Descriptive statistics of overall density of distraction gap as determined by quantitative computed tomography (Densiscan 10001) Group 1 (6 w, 208) 2 (13 w, 208) 3 (13 w, 08)

N Obs

Mean

Median

25th Pctl

75th Pctl

SD

Coefficient of variation

Minimum

Maximum

10 8 10

0.82 1.00 0.98

0.84 1.01 1.05

0.68 0.88 0.83

0.95 1.11 1.12

0.15 0.13 0.22

17.74 13.23 22.38

0.59 0.84 0.59

1.03 1.17 1.23

4

Kunz et al.

Fig. 4. Stiffness in crestal (=occlusal) loading (a) and basal loading (b). A clear increase of stiffness from a 6-week consolidation period after angulation (group 1) to a 13-week consolidation period after angulation (group 2) is observed (P = 0.07). The stiffness of angulated regenerates is higher than after a 13-week linear distraction (group 3), P = 0.21 and higher than the reference group.

According to these findings, angulation of the regenerate did not influence mineralization or mechanical properties permanently. Nevertheless, a temporary disturbance of osseous healing had to be postulated for the stretched basal zone after 6 weeks of consolidation. Discussion

Correction of severe craniofacial deformities, especially the hypoplastic mandible, by means of osteodistraction is more difficult than in long bones. Three-dimensional reconstruction and reshaping of the deformed skeleton is necessary and multiplanar spatial control of distraction vectors is a precondition for precise results. Changing soft tissue forces and interfering vector forces during distraction15 often result in major deviations from planning, which cannot be prevented even by multidimensional distraction devices. Thus, acute callus manipulation may be an important ‘lifeboat’ after loss of vector control, but also a part of a treatment plan

which includes angular changes as a part of the correction14. Mechanical characterization of the regenerated osseous tissue, the impact of strain application and tissue response during distraction osteogenesis have been the topic of many publications3,9,10,18–22,25,28– 30 . So far there has been no laboratory study investigating if a regenerate created by linear distraction osteogenesis can be manipulated in a single, acute procedure immediately after distraction, without decreasing its mechanical properties. In addition in this study it was possible to assess the effects of callus compression and stretching at the same distraction site. The fresh regenerate proved to be easily angulated without resistance to the molding procedure which was performed immediately after distraction was finished. Minor deviations from the preplanned angle of 208 could be due to the ongoing ossification of the regenerate and/or slight movements at the pin/bone interface17. Pin stability and rigidity were shown to be of the utmost importance for uneventful oss-

eous healing1. Device stability was assessed manually after sacrifice of the animals. Only 2 devices were unstable (6.6%). Pin stability was assessed for each pin after removal of the device. Instability of 1 single pin occurred in 6 specimens, but only in 1 case was the regenerate (group 1, 6 weeks recovery) clinically unstable. In 8 specimens at least 1 pin proximal and distal to the osteotomy was loose which in 3 regenerates (all in group 3, linear distraction, 13 weeks) resulted in instability (10%). In group 3 (linear distraction only, consolidation 13 weeks) most cases of pin instability occurred. The main reason may be problems with occlusion. After linear distraction without angulation, occlusal interferences with changed loading properties occurred, which were not observed when angulation was performed. Here, normal canine-to-canine relation resulted and a posterior open bite occurred with less occlusal interferences during the consolidation period. The system used for mechanical testing did not alter the morphology of the manipulated mandible, a precondition for further histological investigations. A similar cantilever-bending test described by PERROTT et al.28 proved to be useful for testing stiffness of an experimental mandibular distraction wound without destroying its structure. Low stiffness after 13 weeks of consolidation always occurred in combination with a considerable instability at the bone/pin interface, which was evaluated clinically as well as histologically, whereas uneventful osseous union was found under adequately stable conditions. After a 6-week consolidation time, density and stiffness were lower than in the groups with a neutral fixation of 13 weeks; this finding corresponds to a normal bone healing process2. KABAN et al.9 reported significantly lower than normal stiffness of a porcine distraction wound after 24 days of neutral fixation, in spite of clinical stability and high bone-fill scores by ultrasound and plain radiographic data. These findings indicate that not only morphological factors determine complete consolidation of the regenerate. In this study visible retardation of the bone healing process, indicated by lower bone density, occurred within the stretched basal zone only, not within the compressed area, indicating the more critical character of a stretched regenerate (Fig. 5). This retardation was completely compensated after 13 weeks of consolidation. Under stable conditions, 25% of the specimens of group 1 achieved the same

Mechanical properties after callus manipulation

Fig. 5. Contact radiographs of hemi-mandibles. Left: disturbed ossification in the basal, stretched zone of the angulated regenerate after 6 weeks of consolidation. Right: completed osseous healing after 13 weeks of consolidation.

level of stiffness (90%) after 6 weeks compared to the mean stiffness values of the control group. There was no obvious difference between angulated and only linearly distracted regenerates after a 13week consolidation. Different loading conditions (occlusal loading and loading 1808 to this direction) showed no obvious difference between all groups. Density measured by quantitative computed tomography (QCT) revealed a strong correlation to the stiffness of the regenerate, which confirmed the findings in literature7 (Fig. 6). These results demonstrate that a non-invasive QCT scan could predict the mechanical properties of a distracted mandibular regenerate, which may have major clinical impact on the timing of device removal. Density of the mineralized regenerate is not the only parameter with impact on the mechanical properties. The spatial pattern of bone formation bridging the distraction gap also seems to be of importance13. Osseous bridging of the fibrous interzone could be demonstrated by fluorochromes after

only 14–20 days in the compressed part of the manipulated regenerate, whereas bridging of the stretched zone usually occurred after 6 weeks. Wide bony bridging occurred after a consolidation period of 6 weeks, with the stretched part of the regenerate revealing a delay of 2–4 weeks when compared with the compressed zone. Further mineralization and remodelling of the adjacent bone continued after 13 weeks. The authors of the present study conclude that a regenerate created by linear distraction osteogenesis can be considerably manipulated immediately after finishing the lengthening process, without permanent impairment of the mechanical properties, whereas the newly formed bone during distraction is highly sensitive to unphysiological mechanical strain. Device and pin stability are preconditions for undisturbed osseous healing. Visible retardation of ossification within the stretched area compared to the compressed zone of the regenerate indicated the more critical character of a stretched

Fig. 6. Quantitative computed tomography (QCT Densiscan 1000) recorded data as a linear attenuation coefficient. Group 1 with a 6-week consolidation period after angulation showed less dense bone compared to specimens with 13 weeks of consolidation (groups 2 and 3) corresponding to a normal osseous healing at that early time. In the stretched regenerate (T = tension zone) mineralization was higher as compared to the compression zone (C) which was due to group 3, P = 0.004.

5

regenerate. Therefore tension forces acting on the completed regenerate should be minimized to prevent possible damage to the new bone. This can be achieved by gradually changing the vector of distraction during the lengthening process or by overdistraction prior to callus molding. Instead of being compressed on one side and distracted on the other side while it is angulated to reach the final position, the additional length gained by overdistraction would then allow that the regenerate is compressed over its full width to reach the same result. Acknowledgement. Supported by the Swiss National Science Foundation, Grant No. 3200-053983.98/1. References 1. Aronson J, Harrison BS, Boyd CM, Cannon DJ, Lubansky HJ. Mechanical induction of osteogenesis: the importance of pin rigidity. J Pediatr Orthop 1988: 8: 396–401. 2. Aronson J, Hogue WR, Flahiff CM, Gao GG, Shen XC, Skinner RA, Badger TM, Lumpkin CK. Development of tensile strength during distraction osteogenesis in a rat model. J Orthop Res 2001: 19: 64–69. 3. Carter DR, Beaupre GS, Giori NJ, Helms JA. Mechanobiology of skeletal regeneration. Clin Orthop Relat Res 1998: 355(Suppl.):41–55. 4. Cope JB, Samchukov ML, Cherkashin AM, Wolford LM, Franco P. Biomechanics of mandibular distractor orientation: an animal model analysis. J Oral Maxillofac Surg 1999: 57: 952–962. 5. Gateno J, Teichgraeber JF, Aguilar E. Computer planning for distraction osteogensis. Plast Reconstr Surg 2000: 105: 873–882. 6. Grayson BH, Santiago PE. Treatment planning and biomechanics of distraction osteogenesis from an orthodontic perspective. Semi Orthodont 1999: 5: 9–24. 7. Harp JH, Aronson J, Hollis M. Noninvasive determination of bone stiffness for distraction osteogenesis by computed tomography scans. Clin Orthop Relat Res 1994: 4: 42–48. 8. Jonsson B, Siemssen SJ. Arced segmental mandibular regeneration by distraction osteogenesis. Plast Reconstr Surg 1998: 101: 1925–1930. 9. Kaban LB, Thurmuller P, Troulis MJ, Glowacki J, Wahl D, Linke B, Rahn B, Perrott DH. Correlation of biomechanical stiffness with plain radiographic and ultrasound data in an experimental mandibular distraction wound. Int J Oral Maxillofac Surg 2003: 32: 296–304. 10. Kessler PA, Merten HA, Neukam FW, Wiltfang J. The effects of magnitude and frequency of distraction forces for tissue regeneration in distraction osteo-

6

11.

12.

13.

14.

15.

16.

17.

18.

Kunz et al. genesis of the mandible. Plast Reconstr Surg 2002: 109: 171–180. Klein C, Howaldt HP. Lengthening of the hypoplastic mandible by gradual distraction in childhood: a preliminary report. J Craniomaxillofac Surg 1995: 23: 68–74. Klein C, Howaldt HP. Correction of mandibular hypoplasia by means of bidirectional callus distraction. J Craniofac Surg 1996: 7: 258–266. Kunz C, Adolphs N, Buescher P, Hammer B, Rahn B. Possible problems of molding the regenerate in mandibular distraction osteogenesis – Experimental aspects in a canine model. J Craniomaxillofac Surg 2005: 33: 377–385. Kunz C, Brauchli L, Moehle T, Rahn B, Hammer B. Theoretical considerations for the surgical correction of the mandibular deformity in hemifacial microsomia patients using multifocal distraction osteogenesis. J Oral Maxillofac Surg 2003: 61: 364–368. Kunz C, Hammer B, Prein J. Manipulation of callus after linear distraction: a lifeboat or an alternative to multivectorial distraction osteogenesis of the mandible? Plast Reconstr Surg 2000: 105: 674–679. Losken HW, Patterson GT, Lazarou SA, Whitney T. Planning mandibular distraction: preliminary report. Cleft Palate Craniofac J 1995: 35: 71–76. Luchs JS, Stelnicki EJ, Rowe NM, Naijher NS, Grayson BH, McCarthy JG. Molding the Regenerate in mandibular distraction: part I: laboratory Study. J Craniofac Surg 2002: 13: 205–211. Meyer U, Meyer T, Schlegel W, Scholz H, Joos U. Tissue differentiation and cytokine synthesis during strainrelated bone formation in distraction osteogenesis. Br J Oral Maxillofac Surg 2001: 39: 22–29.

19. Meyer U, Meyer T, Vosshans J, Joos U. Decreased expression of osteocalcin and osteonectin in relation with high strains and decreased mineralization in mandibular distraction osteogenesis. J Craniomaxillofac Surg 1999: 27: 222– 227. 20. Meyer U, Meyer T, Wiesmann HP, Stratmann U, Kruse-Losler BM, Joos U. The effect of magnitude and frequency of interfragmentary strain on the tissue response to distraction osteogenesis. J Oral Maxillofac Surg 1999: 57: 1331– 1339. 21. Meyer U, Wiesmann HP, Kruse-Losler B, Handschel J, Stratmann U, Joos U. Strain-related bone remodeling in distraction osteogenesis. Plast Reconstr Surg 1999: 103: 800–807. 22. Meyer U, Wiesmann HP, Meyer T, Schulze-Osthoff D, Jasche J, Kruse-Losler B, Joos U. Microstructural investigations of strain-related collagen mineralization. Br J Oral Maxillofac Surgery 2001: 39: 381– 389. 23. Molina F, Ortiz Monasterio F. Mandibular elongation and remodelling by distraction: a farewell to major osteotomies. Plast Reconstr Surg 1995: 96: 825–840. 24. Moore MH, Guzman-Stein G, Proudman TW, Abbott AH, Netherway DJ, David DJ. Mandibular lengthening by distraction for airway obstruction in treacher Collins syndrome. J Craniofac Surg 1994: 5: 22–25. 25. Nomura S, Takano-Yamamoto T. Molecular events caused by mechanical stress in bone. Matrix Biology 2000: 19: 91–96. 26. Ortiz Monasterio F, Molina F, Andrade L, Rodriguez C, Arregui

27.

28.

29.

30.

JS. Simultaneous mandibular and maxillary distraction in hemifacial microsomia in adults: avoiding of occlusal disasters. Plast Reconstr Surg 1997: 100: 852–861. Pensler JM, Goldberg DP, Lindell B, Caroll NC. Skeletal distraction of the hypoplastic mandible. Ann Plast Surg 1995: 34: 130–136. Perrott DH, Rahn B, Wahl D, Linke B, Thurmuller P, Troulis MJ, Glowacki J, Kaban LB. Development of a mechanical testing system for a mandibular distraction wound. Int J Oral Maxillofac Surg 2003: 32: 523–527. Radomisli TE, Moore DC, Barrach HJ, Keeping HS, Ehrlich MG. Weight-bearing alters the expression of collagen types I and II, BMP2/4 and osteocalcin in the early stages of distraction osteogenesis. J Orthop Res 2001: 19: 1049–1056. Sato M, Ochi T, Nakase T, Hirota S, Kitamura Y, Nomura S, Yasui M. Mechanical tension-stress induces expressions of bone morphogenetic protein (BMP-)2 and BMP-4, but not BMP-6, BMP-7, and GDF-5 mRNA, during distraction osteogenesis. J Bone Min Res 1999: 14: 1084–1095.

Address: Christoph Kunz Clinic for Reconstructive Surgery Department of Oral and Craniomaxillofacial Surgery University Hospital Basel Spitalstrasse 21 CH - 4031 Basel Switzerland Tel: +41 61 2652525 Fax: +41 61 2657458. E-mail: [email protected]