This study characterized sequential molecular and cellular events in the porcine mandibular distraction osteogenesis (DO) wound. Nineteen Yucatan minipigs were divided into three treatment groups: Group A, unilateral mandibular distraction with 0 day latency, 1 mm/day rate for 12 days, 24 days fixation ( n = 16); Group B, acute lengthening 12 mm ( n = 2); Group C, sham control ( n = 1). Group A was further divided by death date: mid-DO ( n = 5), end-DO ( n = 4), mid-fixation ( n = 5) and end-fixation ( n = 2). Groups B and C were killed on postoperative day 36, corresponding to end-fixation. Specimens were stained for proliferating cell nuclear antigen (PCNA) and bone morphogenetic protein-4 (BMP4). Cellular proliferation (PCNA) was assessed quantitatively and BMP4 staining was assessed on a semi-quantitative scale. Progenitor cell proliferation was greatest during mid-DO and decreased from end-DO through end-fixation. Proliferation in the acute lengthening group was elevated relative to sham control and comparable to end-DO. BMP4 staining intensity (localized to the periosteal cambium layer) was greatest during mid- and end-DO, decreased at mid-fixation and was undetectable at end-fixation. Progenitor cell proliferation and BMP4 expression are greatest during mid-DO and decrease progressively thereafter. At the time of death of the acute lengthening group, only increased cell proliferation was demonstrated.
Distraction osteogenesis (DO) is clinically effective for the treatment of craniofacial defects requiring bone lengthening. DO produces gradual skeletal expansion and eliminates the need for bone and soft tissue grafting procedures. Greater lengthening can be achieved by DO compared to standard techniques of acute jaw advancement with bone grafts. The quality and quantity of bone formation in response to distraction is improved compared to that achieved by acute jaw lengthening of equal magnitude. Disadvantages of DO include the need for patient cooperation during active distraction, device failure and a significantly increased duration of treatment compared to standard osteotomies. A better understanding of the molecular events of healing may allow development of DO protocols to improve the rate and quality of bone formation and to reduce lengthy treatment times.
It has been demonstrated that DO prolongs osteoinductive growth factor expression for the duration of active distraction in comparison to the transient increases observed during fracture repair. Bone morphogenetic protein-4 (BMP4), a member of the transforming growth factor beta (TGFβ) superfamily, initiates osseous healing by promoting the migration of osteoprogenitor cells and stimulating their proliferation and osteogenic differentiation. Transitory increases in BMP4 expression and cellular proliferation are observed in response to bone fracture. BMP4 levels rapidly return to baseline once bone ossification begins and decreasing cellular proliferation accompanies callus maturation. In contrast, DO prolongs the early phase of bone repair by maintaining elevated levels of BMP4 for the duration of active distraction as well as stimulating increased cellular proliferation. There are a limited number of large animal studies in which the authors investigate molecular and cellular events during DO. Most reported small animal experiments have compared events of DO healing to osteotomies without actual bone advancement. To the authors’ knowledge there have been no reports comparing DO to acute lengthening of equal magnitude in a large animal model using molecular methods.
The purpose of this study was to document molecular and cellular events during healing of the porcine mandibular distraction wound by documenting serial changes in cellular proliferation and BMP4 expression. The authors hypothesize that during active distraction: a peak in cellular proliferation will occur; and BMP4 expression will be localized and markedly enhanced in the periosteal cambium layer.
Materials and methods
Nineteen female Yucatan minipigs in the mixed dentition stage (age 6 months) were used in this study. They were housed in the animal facility 1 week prior to operation to become acclimatized to diet and housing conditions. Animals were randomly divided preoperatively into three treatment groups: Group A, unilateral distraction ( n = 16); Group B, acute lengthening ( n = 2); Group C, sham control ( n = 1). Experimental use of the animals was approved by the Subcommittee on Research and Animal Care at the Massachusetts General Hospital (SRAC #2002N000270) and complied with the standards set forth by the Association for Assessment and Accreditation of Laboratory Animal Care.
A reproducible, standardized mandibular DO model in the Yucatan minipig has been reported by the authors’ group. In summary, the right mandibular body and ramus are exposed through a submandibular incision and sub-periosteal dissection. A continuous vertical corticotomy extending from the junction of the mandibular body and ramus to the inferior portion of the mandibular angle is made. A modified, semiburied, unidirectional, mandibular distraction device (Synthes CMF, Westchester, PA, USA) is placed perpendicular to the corticotomy and fixed with four 2.0 mm bicortical screws. Radiopaque marker screws are placed on each side of the corticotomy inferior to the distractor footplates for radiographic measurement of bone lengthening. After fixation of the distractor, the osteotomy is completed using osteotomes. The wound is closed in layers and postoperative analgesics immediately administered. Elizabethan collars are placed around the animals’ necks to minimize displacement of the distractors and to allow device activation out of their visual fields. Animals are fed a pureed diet for the duration of the study.
Animals from Group A ( n = 16) underwent unilateral distraction osteogenesis with 0 day latency, 1 mm/day rate (0.5 mm twice daily rhythm) for 12 days (gap size 12 mm). The period of rigid fixation was 24 days (twice the duration of active distraction). Animals undergoing distraction were further divided into four groups which were determined by date of death: mid-DO ( n = 5), end-DO ( n = 4), mid-fixation ( n = 5) and end-fixation ( n = 2). Group B animals ( n = 2) underwent immediate 12 mm acute lengthening following placement of distraction device and Group C ( n = 1) sham control underwent corticotomy and distractor placement without activation. Groups B and C were killed and mandibles harvested at the end-fixation time point (postoperative day 36).
Clinical and radiographic evaluation
Serial clinical examination for signs of pain and infection and to document mandibular position and occlusion were performed. Lateral and submental plain radiographs were used to measure mandibular lengthening as the change in distance between marker screws after the completion of distraction.
Right hemi-mandibles were harvested, fixed and decalcified in Immunocal solution (10% formic acid in 10% formalin solution) for 10 days using a previously described protocol. Specimens were then placed in 10% phosphate buffered formalin for 48 h and stabilized in 30% sucrose solution for 24 h.
Experimental specimens were divided in half in the coronal plane through the midpoint of the regenerate zone creating two (anterior and posterior) segments and embedded in paraffin. Tissue sections were prepared at 6 μm intervals from the middle of the regenerate zone from both the anterior and posterior paraffin blocks and mounted on Fisher–Frost slides.
Prepared slides were placed in a 60 °C oven overnight and deparaffinized in xylenes. Rehydration was achieved with a series of ethanol dilutions (100%, 95%, 70%, 50%) followed by a 10 min rinse in 1× phosphate buffered saline (PBS). A conventional microwave with a rotating tray was used to accomplish manufacture-suggested heat induced epitope retrieval. Slides were submersed in 1× PBS and brought to the boil for 10 min followed by a 20 min cooling period. Endogenous peroxidase activity and non-specific binding sites were blocked using reagents included in DAKO LSABII kit (LSABII-HRP, K0675).
Primary antibodies against proliferating cell nuclear antigen (PCNA) and BMP4 were used in this study. Sections were incubated in anti-PCNA antibody (mouse monoclonal, 1:100, Dako M0879) or anti-BMP4 antibody (rabbit polyclonal, 1:75, Abcam 39973) at room temperature for 2.5 h in a humidity box. Primary antibodies were diluted with commercially available diluent (Dako Antibody Diluent S0809). Detection of target antigen binding was accomplished with biotinylated secondary antibodies, streptavidin conjugated HRP and visualized with AEC substrate chromogen (DAKO K3464). All slides were counterstained with Mayer’s Haematoxylin-Lillie’s modification (Dako S3309).
A Nikon Eclipse 80i microscope equipped with a Q Imaging microdigital colour camera (Canada, Burnaby, BC) was used to take all digital photomicrographs.
For PCNA, three images (400×) were taken at random from the superior, middle and inferior thirds on slides from each half of the regenerate (nine images per slide, 18 images per animal used in this study). The inferior alveolar nerve (IAN) separated the superior and middle thirds. The area inferior to the IAN was bisected to separate the middle from the inferior third. Two independent investigators analysed each image using ImageJ software (National Institute of Heath, Bethesda, MD, USA) and determined the percentage positive PCNA as the ratio of PCNA immunopositive nuclei to total nuclei in each image. Antigen specificity was confirmed using porcine skin samples as positive controls to confirm nuclear staining of proliferating cells in the basal layer.
For BMP4, semi-quantitative analysis of extracellular protein expression has been validated in several studies. Two blinded investigators used a semi-quantitative approach to assess BMP4 staining intensity by location. The periosteal and sub-periosteal regions were systematically scanned by each investigator who assigned the regions a staining intensity score based on a 4-point scale (0–3), where: 0 is no staining, 1 is minimal staining intensity of <50% of the surface area, 2 is moderate staining intensity of 50–100% of the surface area and 3 is uniform dense staining of 100% of the surface area.
Data analysis was performed using a standard statistical software package (SPSS version 17.0, SPSS Inc., Chicago, IL, USA). One-way analysis of variance (ANOVA) with Tukey post hoc comparison was performed on PCNA data. For all analyses, a P -value <0.05 was considered statistically significant. Inter-examiner reliability analysis, using the Kappa statistic, was performed on the BMP4 data.
Clinical and radiographic evaluation
All animals ( n = 19) survived the surgical procedure and observation period without device failure. At the time of death, an occlusal crossbite with deviation towards the non-operated side was present in all animals with unilateral mandibular advancement (Groups A and B). Two animals from Group A (one mid-DO and one mid-fixation) developed postoperative infections and were removed from the study, leaving these groups with four animals each for analysis. Desired mandibular advancement, measured by marker screw separation on lateral and submental radiographs, was achieved.
Progenitor cell proliferation, as determined by the percentage of positive PCNA cells, was greatest during mid-DO and decreased sequentially through end-fixation ( Figs 1 and 2 ). There was a significant decrease in proliferation between mid- and end-DO time periods ( P = 0.005) and between mid- and end-DO groups and the fixation groups. Within the fixation groups there was no statistically significant difference in proliferation. Proliferation in the acute lengthening group was not significantly different from that in the end-DO group but was greater than in the other end-treatment groups (end-fixation, sham control). Proliferation in the end-fixation group was nearly equal to sham control. No staining was observed on the negative control specimens.
Inter-examiner reliability tests showed high correlation in the evaluation of BMP4 staining intensity (Kappa correlation coefficient = 0.721). During active distraction (mid- and end-DO) extracelluar BMP4 expression was localized to the inner cambium layer of the periosteum and extended near contiguously around the specimens (two moderate to three maximal intensity). A distinct delineation was observed between the cambial and adjacent outer fibrous layer which had no observable staining ( Fig. 3 ). During distraction, limited intracellular staining was observed in osteoblasts rimming bony islands. BMP4 intensity decreased dramatically following active distraction. At mid- and end-fixation, ‘minimal’ (1) or ‘no staining’ (0) was observed in all evaluated sites. Staining in the acute and sham groups was comparable to the end-fixation group. No staining was observed on the negative controls.