Minipig Model of Maxillary Distraction Osteogenesis: Immunohistochemical and Histomorphometric Analysis of the Sequence of Osteogenesis

CRANIOMAXILLOFACIAL DEFORMITIES/COSMETIC SURGERY J Oral Maxillofac Surg 70:2629-2640, 2012

Minipig Model of Maxillary Distraction Osteogenesis: Immunohistochemical and Histomorphometric Analysis of the Sequence of Osteogenesis Maria E. Papadaki, DMD, MD, PhD,* Leonard B. Kaban, DMD, MD,† and Maria J. Troulis, DDS, MSc‡ Purpose: To document the sequence of bone formation in a minipig model of Le Fort I distraction

osteogenesis (DO) using immunohistochemistry and histomorphometry. Materials and Methods: Female Yucatan minipigs (N ⫽ 9) in the mixed-dentition stage underwent bilateral maxillary DO. The distraction protocol was 0 days of latency, with a distraction rate of 1 mm/d for 12 days and 24 days of fixation. Specimens were harvested and divided between the central incisors (18 hemi-maxillae) at the end of DO (n ⫽ 6), at mid-fixation (n ⫽ 6), and at the end of fixation (n ⫽ 6). Sections, including the advancement zone, were stained with hematoxylin-eosin, collagen II, CD34, proliferating cell nuclear antigen, and tartrate-resistant acid phosphatase. Light and fluorescence microscope images (original magnification ⫻200) were obtained, and percentage of surface area (PSA) of the advancement zone occupied by fibrous tissue, vessels, proliferating cells, osteoid, and bone was determined. An intact maxilla served as the control. Results: At the end of DO, in the advancement zone, the PSA (mean values) of proliferating cells was 33.16%; fibrous tissue, 52%; vessels, 4.35%; and new bone, 5.45%. At the end of fixation, the PSA of proliferating cells decreased to 10.53%, fibrous tissue to 2.3%, and vessels to 1.5% whereas the PSA of new bone increased to 44.9%. Conclusions: The results of this study indicate that the progression of osteogenesis in the maxillary DO wound begins with intense cellular proliferation and vascular fibrous tissue formation and progresses to mature, cancellous bone by the end of fixation. The PSA occupied by mature bone is significantly less than in the control maxilla at the end of fixation. This is consistent with the sequence in the mandibular DO wound. This is a US government work. There are no restrictions on its use. Published by Elsevier Inc on behalf of the American Association of Oral and Maxillofacial Surgeons. J Oral Maxillofac Surg 70:2629-2640, 2012

Received from the Department of Oral and Maxillofacial Surgery, Massachusetts General Hospital, Harvard School of Dental Medicine, Boston, MA. *Instructor. †Walter C. Guralnick Professor and Chairman. ‡Associate Professor and Director of Residency Training Program. This manuscript was presented at the 90th Annual Meeting of the American Association of Oral and Maxillofacial Surgeons, Seattle, WA, September 16-20, 2008, and won the Outstanding Poster Presentation Award. Funded by the AO Foundation (grant 03-K69), the Hanson Foun-

dation (Boston, MA), and the AO/Synthes/MGH Fellowship in Pediatric Oral and Maxillofacial Surgery. Address correspondence and reprint requests to Dr Papadaki: Department of Oral and Maxillofacial Surgery, Massachusetts General Hospital, Harvard School of Dental Medicine, Boston, MA 02114; e-mail: [email protected] This is a US government work. There are no restrictions on its use. Published by Elsevier Inc on behalf of the American Association of Oral and Maxillofacial Surgeons 0278-2391/12/7011-0$36.00/0 doi:10.1016/j.joms.2012.01.030

2629

2630 Distraction osteogenesis (DO) offers distinct advantages over standard osteotomies and acute lengthening for clinical bone expansion, including a less invasive operation, elimination of donor-site morbidity, the possibility of larger magnitudes of lengthening, soft tissue accommodation, and decreased risk to vital structures. Understanding the histogenesis of DO is of scientific and academic interest and is a prerequisite for the improvement of the technique and clinical outcomes. Once the cellular and tissue processes are understood, it may be possible to perform successful DO with a shorter treatment time and achieve increased quality and quantity of bone formation. We developed and described a Yucatan minipig model of maxillary DO at the Le Fort I level and showed that intramembranous bone formation was the predominant mechanism of healing in the advancement zone.1 Latency was not necessary for bone formation. The Yucatan minipig model has been used in a series of studies of mandibular DO biology in our laboratory (Skeletal Biology Research Center, Massachusetts General Hospital, Boston, MA) since 1997.2-7 The minipig model was chosen because the bone turnover rate in pigs is equal to that of humans, an important factor when one is studying the biology of a DO wound.8-12 Furthermore, the chewing pattern and transmission of chewing forces to the midface through the zygomatic buttresses and pyriform apertures are similar in minipigs and humans.10 Results in this large-animal model may, therefore, be more analogous to human DO than those in small-animal models.13 Finally, Yucatan minipigs are friendly and easy to handle in the laboratory. The purpose of this study was to analyze the detailed sequence of healing in a previously described minipig maxillary DO wound1 using immunohistochemistry and histomorphometry at 3 time points: end of distraction, mid-fixation, and end of fixation.

Materials and Methods ANIMALS

Nine female Yucatan minipigs, aged 6 months (range, 5.5-6.5 months), in the mixed-dentition stage, weighing 25 to 35 kg, were used in this set of experiments.1 The study protocol was approved by the Massachusetts General Hospital Subcommittee on Research Animal Care (SRAC, No. 2004N000294/1). Animals were housed for 2 to 7 days before the operation to acclimate to housing and diet. They were monitored daily for general appearance, activity level, weight, temperature, heart rate, and respiratory rate. Their diet included grain, yogurt, pudding, and apples. Preoperatively, the minipigs were fasted for 12 hours.

MINIPIG MODEL OF MAXILLARY DO SURGICAL PROCEDURE

The minipigs underwent bilateral maxillary DO under general anesthesia as previously described.1 A modified Le Fort I osteotomy was performed through a vestibular incision. The bone cut was performed with a reciprocating saw and extended from the piriform aperture to the zygomatic buttress in the usual fashion, ending with a vertical osteotomy between the 2 last molars bilaterally. Custom-made intraoral, unidirectional, semiburied Le Fort I distraction devices were fixed across the osteotomies bilaterally (Synthes CMF, West Chester, PA). The activating arms were externalized below the ears through small stab incisions. Fixation was achieved with eight 2-mm-diameter, 8-mm-long screws. Marker screws were placed above and below the osteotomy at the area of the piriform aperture bilaterally for measurement of the advancement. The osteotomy was completed with osteotomes, and the wounds were closed in 2 layers with 3-0 Monocryl suture (Ethicon, Somerville, NJ). The postoperative diet consisted only of mashed grains and yogurt to exclude chewing. DISTRACTION PROTOCOL

The distraction protocol was 0 days of latency, with a distraction rate of 1 mm/d for 12 days and 24 days of fixation. Animals were euthanized, and maxillae were harvested and divided in the midline between the central incisors, providing 2 maxillary specimens per pig and 6 per time point: end of DO (n ⫽ 6), mid-fixation (n ⫽ 6), and end of fixation (n ⫽ 6). Maxillary advancement was measured on ex vivo specimens on as the distance between the marker screws. One minipig aged 6 months did not undergo Le Fort I osteotomy or distraction and served as the control (n ⫽ 2 hemi-maxillary specimens). HISTOLOGY

Each hemi-maxilla was divided 2 areas of interest (piriform aperture and zygomatic buttress), resulting in 12 specimens at each time point (4 bony segments from each minipig ⫻ 3 minipigs at each time point). Each specimen included the advancement zone in the middle and native maxilla below and above the region of interest. All harvested experimental (n ⫽ 36) and control (n ⫽ 4) specimens were fixed, decalcified in 10% formic acid, divided 2 (for coronal and sagittal examination), and embedded in paraffin (Fig 1). All the specimens were sectioned at 5-␮m thickness, mounted on Fisherfrost slides (Fisher Scientific, Pittsburgh, Pennsylvania), and stained with hematoxylin-eosin (H&E). IMMUNOHISTOCHEMISTRY AND HISTOCHEMICAL STAINING

Immunohistochemical reactions for collagen II (cartilage), CD34 (cluster of designation) for endothelial cells,

2631

PAPADAKI, KABAN, AND TROULIS

FIGURE 1. A, Anterior half of left hemi-maxillary specimen after decalcification. It includes the advancement zone in the middle and native maxilla (M) above and below it. The specimen was cut to produce sagittal (B) and coronal (C) sections. B, The arrows point to the advancement zone. C, Coronal section of the left anterior hemi-maxilla that includes the middle and inferior turbinate (black arrows). The blue arrow points to the advancement zone. Papadaki, Kaban, and Troulis. Minipig Model of Maxillary DO. J Oral Maxillofac Surg 2012.

and proliferating cell nuclear antigen (PCNA) for proliferating cells were performed. Histochemical staining for tartrate-resistant acid phosphatase (TRAP) for osteoclasts was carried out. Four sections from each maxilla, right anterior (piriform aperture) and posterior (zygomatic buttress) and left anterior and posterior, were used for each immunohistochemical staining. Immunohistochemical reactions were performed with the indirect method of streptavidin-biotin. The la-

belled streptavidin biotin (LSAB2, Code K0672) kit (Dako, Carpinteria, California, USA) that includes the basic solutions necessary for this immunohistochemistry technique was used. After, paraffin sections were incubated at 60°C overnight, tissues were deparaffinized in xylene and alcohol solutions. The tissue outline was drawn with a PAP pen (Abcam, Cambridge, MA) to prevent applied solutions from diffusing. Antigen retrieval for PCNA and CD34 was performed by heating the tissues placed in phosphate-buffered solution (PBS) in a microwave for 5 minutes. Antigen retrieval for collagen II was performed by adding hyaluronidase 2% solution onto the tissues and maintaining it for 30 minutes. Tissue slides were then rinsed in PBS 3 times. Sections were incubated in peroxidase-blocking solution (3% hydrogen peroxide) for 10 minutes at room temperature (LSAB2 kit, Dako, Carpinteria, California, USA). Tissue slides were again rinsed in PBS 3 times. Sections were incubated with primary antibody (antibody added on section with pipette) at appropriate dilution: PCNA (M0879; DakoCytomation, Glostrup, Denmark), dilution 1:200 with incubation time of 3 hours; CD34 (M7165; DakoCytomation), dilution 1:50 with incubation time of 2 hours; and collagen II (7005; Chondrex, Redmond, WA), dilution 1:1,000 for 1 hour. Incubation for all antibodies was performed in a refrigerator at 3°C. Positive control markers included swine skin for PCNA and CD34 and swine auricular cartilage for collagen II. Tissue slides were rinsed in PBS 3 times. Sections were incubated in biotinylated secondary antibody (LSAB2, Dako, Carpinteria, California, USA) in PBS for 60 minutes at room temperature. Slides with tissues were rinsed in PBS 3 times. Sections were incubated in chromogen/substrate for 10 minutes (LSAB2). Slides with tissues were rinsed in PBS 3 times. Gill’s hematoxylin was added for 20 seconds to 3 minutes. Sections were rinsed in deionized water for 5 to 15 minutes and coverslips with mounting medium were added on the slides. 100 ␮L from each solution, was used for 1 slide. HISTOMORPHOMETRY

The advancement zone in each H&E and PCNAstained sagittal microsection, from either the piriform aperture or the zygomatic buttress, was divided into 5 regions from anterior to posterior: 〈, B, C, D, and E. Each of these regions was further divided into 6 zones from superior to inferior: 1, 2, 3, 4, 5, and 6 (Fig 2). Zones 3 and 4 correspond to the center of the advancement zone. Photomicrographs were taken at 200⫻ magnification with a microscope-integrated digital camera (Nikon Eclipse 80i Microscope/DS-Ri1; Nikon, Melville, NY). Images of zones 〈1 to A6, C1 to C6, and E1 to E6 were used for the histomorphometric (quantitative) analysis. Four H&E-stained sagittal microsections were used from each minipig: right and

2632

MINIPIG MODEL OF MAXILLARY DO

FIGURE 2. Sagittal schema of advancement zone (pink) and native maxilla (orange). The advancement zone was divided into zones A, B, C, D, and E from anterior to posterior. Each zone was further divided from superior to inferior into 6 areas. Microphotographs of areas 〈1 to A6, C1 to C6, and E1 to E6 under 200⫻ magnification were used in the histomorphometric analysis. Papadaki, Kaban, and Troulis. Minipig Model of Maxillary DO. J Oral Maxillofac Surg 2012.

left piriform aperture section and right and left zygomatic buttress section. We analyzed 72 images from each animal, 216 images from each group (end of DO, mid-fixation, and end of fixation), and 648 images in total.

Histomorphometric analysis was performed with ImageJ optical processing software (National Institutes of Health, Bethesda, MD) for digitized photomicrographs. The area of interest was marked and measured by use of the software (Figs 3-5). In each image

FIGURE 3. ImageJ software was used for histomorphometric analysis. The area of interest was outlined (green circles) by use of the “polygon” tool and measured in pixels. The “Results” table shows up after clicking “Analyze” and displays the actual surface of each outlined area. The percentage of the surface area to the TSA in each image was estimated. In this photograph, outlining and measurement of the surface area of PCNA-positive cells are shown. The arrows point to areas of osteoid. A, PCNA staining. B, H&E staining of two consecutive microsections (⫻200). 10-4: Pig code, R2: Right posterior segment, Sag: sagittal. Papadaki, Kaban, and Troulis. Minipig Model of Maxillary DO. J Oral Maxillofac Surg 2012.

2633

PAPADAKI, KABAN, AND TROULIS

FIGURE 4. Outlining and measurement of surface area (in pixels) of osteoid areas (1, 2, 3, and 4) at end of fixation (H&E stain, original magnification ⫻200). 11-5: Pig Code, R3: right posterior. Papadaki, Kaban, and Troulis. Minipig Model of Maxillary DO. J Oral Maxillofac Surg 2012.

the actual surface area and percentage of surface area (PSA) of proliferating cells (Fig 3), fibrous tissue, vessels, osteoblasts, osteoid (Fig 4), and new mature bone (Fig 5) were measured. Osteoblasts were outlined on the surface of the bone and in the osteoid. PCNA-positive proliferating cells were also outlined and included all types of cells in the advancement zone because the staining cannot distinguish among different types of cells (Fig 3). The PSA of each tissue type was calculated as the ratio of the surface area filled by each tissue type (measured in pixels) to the total surface area (TSA) of all tissues in the image (total image pixels). For comparison, these structures were measured at the corresponding region of an intact maxilla that was used as the control. The mean values of each tissue type’s PSA were calculated for each minipig and each study group (end of DO, mid-fixation, and end of fixation). Paired t tests were used for comparisons of 2 PSAs of the same tissue type among groups. Significance was set at P ⬍ .05.

Results The minipigs had the expected preoperative negative overjet (⫺12 mm); at the end of distraction, all

animals had developed an edge-to-edge occlusion (0 mm of overjet) (Fig 6). Maxillary advancement was 12 ⫾ 1 mm bilaterally and 2 to 3 mm in the vertical plane, as measured between the marker screws directly on the harvested specimens. END OF DO GROUP

Intense proliferating activity in the zone of advancement was observed during active distraction (Fig 7). At the end of DO, the zone of advancement consisted of 33.16% proliferating cells, 52% fibrous tissue, 8.12% osteoid, and 5.45% newly formed bone (mean values for the advancement zone as a whole are shown in Table 1). Almost all the cells (mesenchymal, endothelial, fibroblasts, osteoblasts) in this group were in the proliferating stage. Osteoblasts occupied 14.3% of the TSA, and vessels occupied 4.35% of the TSA (Table 2). Osteoblasts and proliferating cells were found to occupy 21% and 64.85% of the surface area, respectively, at the periphery of the advancement zone in proximity to the native bone. In the center of the advancement zone, the values were 7% and 12% of the TSA, respectively. Fibrous tissue was measured 20.25% in proximity to native bone and 77.44% in the center of the advancement zone. Newly formed bone

2634

MINIPIG MODEL OF MAXILLARY DO

FIGURE 5. An area of new calcified bone (yellow lines) with many osteocytes has been outlined and measured in pixels at the end of fixation Each photograph had been labeled with the pig code, the side (R vs L), 1 for anterior maxillary specimen or 2-3 for posterior specimen and the zone (A1 to A6, C1 to C6, E1 to E6). This photograph originated from pig 11-5 (end of fixation group), left maxilla, anterior specimen and A2 zone. (H&E stain, original magnification ⫻200). Papadaki, Kaban, and Troulis. Minipig Model of Maxillary DO. J Oral Maxillofac Surg 2012.

was measured 14% in proximity to native bone and 0% to 3% in the center of the advancement zone. Two zones were identified in the advancement zone: the zone of new trabeculae at the periphery and the hypercellular, fibrous zone in the center of the regenerate. MID-FIXATION GROUP

The zone of advancement consisted of 14.8% proliferating cells, 38.32% fibrous tissue, and 27.67% new bony trabeculae rimmed by osteoblasts (Table 1). Osteoblasts occupied 17.8% and vessels occupied 1.65% of the TSA (Table 2). END OF FIXATION GROUP

At the end of fixation, the regenerate consisted of 44.9% new bony trabeculae rimmed by osteoblasts and 4.3% fibrous tissue (Table 1). In all sections the gap had filled in with new bone except for a few areas in the posterior maxillary regions around the zygomatic buttress, where islands of fibrous tissue were encountered. Osteoclasts were more prominent in the end-of-fixation group (Fig 8). Cartilage was not found in the regenerate with collagen II immunohistochemistry staining in any group (Fig 9). At the end of fixation, the nasal septum

was attached to the maxillary crest with fibrous tissue (Fig 10). Cell proliferation was significantly higher in the end of DO group compared with mid-fixation and end of fixation groups. PSA occupied by new vessels decreased 3-fold at 15 days after active distraction was ceased and did not change by the end of fixation. Fibrous tissue decreased from 52% of the TSA at the end of DO to 38.32% at mid-fixation and 3.3% at the end of fixation (Table 1). Osteoblasts increased from 14.3% of the TSA at the end of DO to 17% at mid-fixation and decreased to 10% at the end of fixation (Table 2). New bone increased from 5.45% of the TSA at the end of DO to 44.9% at the end of fixation (Table 1). Osteoblasts, proliferating cells, and new trabeculae were found in greater concentrations in the peripheral zones of the regenerate at all time points. Vessels and fibrous tissue were present more predominantly in the center of the advancement zone during the active distraction and fixation period.

Discussion To our knowledge, this is the first description of the sequence of osteogenesis in a large-animal model of maxillary distraction at the Le Fort I level. No

PAPADAKI, KABAN, AND TROULIS

2635 In 2003 Stalmans et al18 performed 14 mm of DO at the Le Fort I level in 1 dog and performed follow-up of the animal radiographically for 1 year to document

FIGURE 6. Advancement of the maxilla by 12 mm was achieved with DO. A, Preoperative lateral photograph of minipig shows negative overjet of 12 mm. B, At the end of the fixation period, overjet was 0 mm. Papadaki, Kaban, and Troulis. Minipig Model of Maxillary DO. J Oral Maxillofac Surg 2012.

reports of histomorphometric analysis performed in midface DO were available before this study. We have shown the sequence from hematoma formation, fibrous proliferation, angiogenesis, bone formation, and remodeling in a minipig model. Biology of the midface distraction wound, especially at the Le Fort I level, has not been studied to the same extent as that of the mandible. A literature search of the electronic library PubMed yielded 199 animal studies on the biology of mandibular DO published between 2000 and 2011. During the same period, there were only 22 animal studies on midface DO.14-35 Two of these involved DO at the Le Fort I level of the midface,18,20 whereas the rest involved midface distraction at the Le Fort II and III levels. Three animal studies on the mechanism of osteogenesis, angiogenesis, and molecular biology of the midface at the Le Fort II level have been published.21-23

FIGURE 7. PCNA staining showed higher proliferating cell index in the end-of-DO group (A) compared with mid-fixation (B) and the end of fixation (C). The arrows point to PCNA-positive cells. All cells with a red or brown hue in the photographs are PCNA-positive cells. (original magnification ⫻200). Papadaki, Kaban, and Troulis. Minipig Model of Maxillary DO. J Oral Maxillofac Surg 2012.

2636

MINIPIG MODEL OF MAXILLARY DO

Table 1. HISTOMORPHOMETRIC ANALYSIS

PSA [Mean (SD)] (%)

End of DO Mid-fixation End of fixation Intact maxilla

Proliferating Cells (PCNA)

Osteoid

Cartilage

Fibrous Tissue

New Bone

33.16 (1.76) 14.8 (3.36) 10.53 (1.37) 0.56 (0.3)

8.12 (3.4) 24.40 (4.6) 41.27 (5.5) 10.93 (2.7)

0 0 0 0

52.25 (6.4) 38.32 (6.82) 3.30 (0.8) 2.82 (1.4)

5.45 (4.2) 27.67 (7.2) 44.9 (2.9) 24.07 (3.4)

Old Bone

0 0 0 60.19 (6.6)

NOTE. Osteoid was found among the new trabeculae and included osteoblasts (Fig 4). New bone refers to mature cancellous bone with multiple osteocytes indicative of recent bone formation. New bone was encountered in intact maxilla because bone continues to form rapidly in young subjects (6 months old). Old bone was mature, cancellous bone with thick trabeculae, no osteoblasts and fewer osteocytes compared with new bone. PCNA staining cannot distinguish among the various types of cells. In the PCNA PSA, all cell types are included (mesenchymal cells, osteoblasts, endothelial cells, preosteoblasts, fibroblasts). Papadaki, Kaban, and Troulis. Minipig Model of Maxillary DO. J Oral Maxillofac Surg 2012.

the position of the maxilla. No histologic or other microscopic studies were performed. In 2002 Weinzweig et al,20 in Philadelphia, studied a new, intraoral, internal device for midface distraction at the Le Fort I level in 4 adult rhesus Macaca mulatta monkeys. They achieved 20 mm of maxillary distraction at a rate of 1 mm/d. Histologic examination was limited to confirmation of bone bridging the gap at the end of the fixation period. No detailed microscopic analysis of the regenerate during DO was reported. Rachmiel et al36-38 performed a series of studies of DO at the Le Fort II level (in 1993, 1995, and 1998) using an external device. They gradually advanced the midface in 3 young sheep over a period of 21 days. Their protocol consisted of 5 days of latency, 21 days of distraction at a rate of 2 mm/d, and a fixation period of 6 weeks. The midface was advanced 36 mm in the nasofrontal area and 43 mm at the lateral portion of the maxilla. Regenerate at the nasofrontal area was examined with electron microscopy, H&E staining, Masson staining (for collagen fibers), and tartrateresistant acid phosphatase staining at the end of distraction, at the end of fixation, and at 1 year postoperatively. Concentrations of calcium and phosphorus were also calculated under the electronic microscope. The au-

Table 2. PSA OF VESSELS AND OSTEOBLASTS IN STUDY AND CONTROL GROUPS

PSA [Mean (SD)] (%)

End of DO Mid-fixation End of fixation Intact maxilla

Vessels

Osteoblasts

4.35 (3.2) 1.65 (1.01) 1.50 (0.3) 1.24 (0.5)

14.3 (7.04) 17.8 (3.8) 10.36 (4.2) 0.48 (0.4)

NOTE. Vessels were included in the fibrous tissue and osteoid presented in Table 1. Papadaki, Kaban, and Troulis. Minipig Model of Maxillary DO. J Oral Maxillofac Surg 2012.

thors reported fibrous tissue, new vessels, and mesenchymal cells in the regenerate at the end of distraction. New, fine trabeculae originating from the old/native bone at the osteotomy border, parallel to the distraction vector, were observed. More osteoclasts were found at the end of distraction compared with the end of fixation. The gap between bony segments had been bridged by osseous tissue, woven and lamellar type, many osteocytes, and few osteoclasts at the end of fixation. Finally, at 1 year postoperatively, the bone was found to be solid and lamellar. The same investigators performed a second series of DO experiments at the Le Fort II level in 7 sheep using the same distraction protocol except the rate was decreased to 1 mm/d (in 2001 and 2002).21,22 Sections from the regenerate were examined at the end of latency (fifth postoperative day) and on the fifth, 10th, 15th, and 20th days of the distraction period, as well as at the end of fixation (6 weeks). Microscopic examination was performed with H&E staining, Masson staining, and von Kossa staining for calcification foci and immunohistochemical staining for PCNA and Tie-2 for endothelial cells. No histomorphometric analysis was performed. The distraction achieved measured 20 mm. At the end of latency, the gap between the bony segments was filled by hematoma with newly formed capillaries, red cells, and very few spindle-shaped mesenchymal cells. During the distraction period, 3 zones were detected in the regenerate: the central or cellular zone with many mesenchymal-like cells and capillaries, the paracentral or fibrous zone, and the peripheral or calcification zone with new trabeculae originating from the old/ native bone. The investigators noted the presence of an “onion-like” configuration between the central and the paracentral zones that stained with Tie-2, indicative of new vessel formation. New trabeculae with osteoblast rimming were found at the end of distraction parallel to the distraction vector. The PCNA in-

PAPADAKI, KABAN, AND TROULIS

FIGURE 8. Tartrate-resistant acid phosphatase (TRAP) staining showed fewer osteoclasts at the end of DO (A) compared with mid-fixation (B) and the end of fixation (C) (original magnification ⫻200). Papadaki, Kaban, and Troulis. Minipig Model of Maxillary DO. J Oral Maxillofac Surg 2012.

dex was highest at the peripheral zone close to the osteotomy border. Our DO study differs from the experiments of Rachmiel et al21,22 because it was performed at the Le

2637 Fort I level, where the distal maxillary segment is advanced, in relation to the proximal segment, parallel to the osteotomy line. In contrast, the segments at the nasofrontal region in Le Fort II distraction separate from each other as in mandibular DO. Thus it is not surprising that findings would differ. In our study, 2 zones were identified in the advancement gap: a proliferating zone with proliferating cells adjacent to the native bone edge and a central zone predominantly with blood vessels and fibrous tissue. Furthermore, osteoclasts progressively increased from the end of DO to mid-fixation and the end of fixation, suggesting that remodeling begins in the zone adjacent to native bone at mid-fixation. Rachmiel et al found more osteoclasts at the end-of-DO period at the Le Fort II level. In our study there were more blood vessels at the end of DO in comparison with mid-fixation and the end of fixation, and there were more vessels in the central regions of the advancement zone in comparison with the peripheral zones closer to the native bone edges. Vascularity in a mandibular DO rat model has been studied by Glowacki et al.39 They showed that nicotine significantly and reproducibly inhibits osteogenesis and angiogenesis in the regenerate. Although primary and secondary cartilages contribute to the development of the midface,40 immunohistochemical staining for cartilage-specific collagen II was negative at all time points. Chondrocyte-like cells were found in 2 H&E-stained sections that immunohistochemical study failed to confirm. As distraction forces ceased during neutral fixation, the proliferation rate of mesenchymal cells, endothelial cells, and fibroblasts decreased, whereas osteoblasts increased from 14% at the end of DO to 17% at mid-fixation and then decreased to 12.5% at the end of fixation. This may be attributed to differentiation of mesenchymal cells to osteoblasts during distraction that continued to multiply during fixation, whereas mesenchymal cells ceased differentiating and multiplying during fixation. On the basis of these findings, a growth factor to promote differentiation of mesenchymal cells to osteoblasts could be added before the end of active distraction. Distraction results in intense proliferation of cells in the regenerate and rich neoangiogenesis in addition to directing all microstructures, that is, cells, collagen fibers, and vessels parallel to the distraction vector. Cell proliferation and angiogenesis at the Le Fort I level during the distraction period were impressive in this study. Termination of active distraction resulted in a decrease in both cell proliferation and angiogenesis during the fixation period. In 2010 Lawler et al7 studied the mandibular DO wound using the same animal model, distraction protocol, and methodology as in our study. The regenerate

2638

MINIPIG MODEL OF MAXILLARY DO

FIGURE 9. Coronal microsections from the same anterior maxillary region at the end of fixation stained with H&E (A) and collagen II (B). The purple area in A looked like cartilage. Immunohistochemical staining specifically for cartilage (collagen II) was negative (B). Periosteum can be seen on the right in both A and B. Papadaki, Kaban, and Troulis. Minipig Model of Maxillary DO. J Oral Maxillofac Surg 2012.

was examined after H&E or hematoxylin–alcian blue– Sirius red staining with histomorphometric analysis and determination of the PSA occupied by hematoma, fibrous tissue, cartilage, and bone. Analysis was performed at mid-DO, the end of DO, mid-fixation, and the end of fixation. Fibrous tissue decreased from mid-DO (PSA, 53.12% ⫾ 8.59%) to the end of fixation (PSA, 25.00% ⫾ 0.83%), similarly to the maxilla, from 52.25% ⫾ 6.4% to 3.3% ⫾ 0.8%. In the maxilla 41.27% (SD, 5.5%) of the advancement zone (regenerate) was occu-

FIGURE 10. Fibrous tissue was seen between the nasal septum and maxillary bone at the end of fixation. Papadaki, Kaban, and Troulis. Minipig Model of Maxillary DO. J Oral Maxillofac Surg 2012.

pied by osteoid at the end of fixation. Cartilage was present at the end of DO (PSA, 1.72% ⫾ 2.71%) and mid-fixation (PSA, 5.82% ⫾ 6.64%) in the mandible. No cartilage was found in the advancement zone in the maxilla. In the mandible, bone increased from mid-DO (PSA, 25.18% ⫾ 0.99%) to the end of fixation (PSA, 64.89% ⫾ 0.79%). Similarly, in the maxilla, bone increased from PSA of 5.45% (SD, 4.2%) at the end of DO to 44.9% (SD, 2.9%) at the end of fixation. It appears that bone formation occurs sooner in the mandible compared with the Le Fort I level of the maxilla. The bone formed in the maxilla at the end of fixation was calcified and included multiple osteocytes indicative of its recent formation. The latency period was 0 days in this model. Four to 10 days’ latency has been the standard for craniomaxillofacial DO.2 The rationale for the latency period is to permit soft callus formation that will be stretched in the next phase of DO. However, there are animal studies that have shown feasible DO with 0 days of latency in the mandible and long bones. In 1998 Tavakoli et al41 performed mandibular DO in sheep and found that biomechanical properties and bone density of the regenerated bone were equal in 3 groups with latency periods of 0, 4, and 7 days. Aronson and Shen42 in 1993 also found that bone regeneration and consolidation were best after 0 days of latency in metaphyseal and diaphyseal long bone lengthening. Similarly, experiments in our laboratory showed that mandibu-

PAPADAKI, KABAN, AND TROULIS

lar and maxillary DO could be successfully accomplished with no latency period.1-7 The anatomy and the rich blood supply of the craniomaxillofacial region may allow bone formation with 0 days of latency. This study indicated that osteogenesis in the maxillary DO wound at the Le Fort I level begins with intense cellular proliferation and vascular fibrous tissue formation and progresses to mature, cancellous bone by the end of fixation. Fibrous tissue in the advancement zone decreased significantly from the end of DO to the end of fixation when its PSA reached that of an intact maxilla. PSA occupied by new vessels decreased 3-fold after active distraction was ceased. Osteoblasts increased from 14.3% of the TSA at the end of DO to 17% at mid-fixation and decreased to 10% at the end of fixation. New bone increased from 5.45% of the TSA at the end of DO to 44.9% at the end of fixation. At the end of fixation, 86% of the advancement zone was occupied by osteoid and new bone. However, the PSA occupied by mature bone was significantly less than in the control maxilla. This is consistent with the sequence in the mandibular DO wound. Acknowledgments The authors thank Drs Haru Abukawa, Malikai Abulikamu, and Fardad Tayebety for their assistance with the procedures during the development of the animal model. They also thank Dr Angela Papageorgiou for her advice and technical guidance with PCNA staining and Dr Julie Glowacki for her comments and corrections of the manuscript.

References 1. Papadaki ME, Troulis MJ, Glowacki J, et al: A minipig model of maxillary distraction osteogenesis. J Oral Maxillofac Surg 68: 2783, 2010 2. Troulis MJ, Glowacki J, Perrott DH, et al: Effects of latency and rate on bone formation in a porcine mandibular distraction model. J Oral Maxillofac Surg 58:507, 2000 3. Kaban LB, Thurmüller P, Troulis MJ, et al: Correlation of biomechanical stiffness with plain radiographic and ultrasound data in an experimental mandibular distraction wound. Int J Oral Maxillofac Surg 32:296, 2003 4. Glowacki J, Shusterman EM, Troulis MJ, et al: Distraction osteogenesis of the porcine mandible: Histomorphometric evaluation of bone. Plast Reconstr Surg 113:566, 2004 5. Zimmermann CE, Thurmüller P, Troulis MJ, et al: Histology of the porcine mandibular distraction wound. Int J Oral Maxillofac Surg 34:411, 2005 6. Tayebaty FT, Williams WB, Baumann A, et al: Histologic and histomorphometric analysis of the porcine mandibular distraction wound. J Oral Maxillofac Surg 64:43, 2006 7. Lawler ME, Tayebaty FT, Williams WB, et al: Histomorphometric analysis of the porcine mandibular distraction wound. J Oral Maxillofac Surg 68:1543, 2010 8. Svendsen P: Anaesthesia and basic experimental surgery of minipigs. A practical exercise. Pharmacol Toxicol 80:23, 1997 (suppl 2) 9. Kuboki T, Shinoda IM, Orsini MC, et al: Viscoelastic properties of the pig temporomandibular joint articular soft tissues of the condyle and disc. J Dent Res 76:1760, 1997 10. Ciochon RL, Nisbett RA, Corruccini RS: Dietary consistency and craniofacial development related to masticatory function in minipigs. J Craniofac Genet Dev Biol 17:96, 1997

2639 11. de Vernejoul MC, Pointillart A, Golenzer CC, et al: Effects of iron overload on bone remodeling in pigs. Am J Pathol 116: 377, 1984 12. Gerard DA, Gotcher JE, Cooper EC, et al: Histomorphometric study of endosseous implants in minipigs: A pilot study. J Dent Res 70:460, 1991 13. Siegel MI, Mooney MP: Appropriate animal models for craniofacial biology. Cleft Palate J 27:18, 1990 14. Niu XG, Han XX: Evaluation of a new semiburied curvilinear distraction device in dogs. Br J Oral Maxillofac Surg 46:61, 2008 15. Liang L, Liu C, Bu R: Distraction osteogenesis for bony repair of cleft palate by using persistent elastic force: Experimental study in dogs. Cleft Palate Craniofac J 42:231, 2005 16. Gateno J, Seymour-Dempsey K, Teichgraeber JF, et al: Prototype testing for a new bioabsorbable Le Fort III distraction device: A pilot study. J Oral Maxillofac Surg 62:1517, 2004 17. Li M, Park SG, Kang DI, et al: Introduction of a novel internal spring-driven craniofacial bone distraction device. J Craniofac Surg 15:324, 2004 18. Stalmans K, Van Erum R, Verdonck A, et al: Cephalometric evaluation of maxillary advancement with an internal distractor in an adult boxer dog. Orthod Craniofac Res 6:104, 2003 19. Nadjmi V, Van Erum R, Schoenaers J, et al: Maxillary distraction using a trans-sinusal distractor: Technical note. Int J Oral Maxillofac Surg 32:553, 2003 20. Weinzweig J, Baker SB, Mackay GJ, et al: Immediate versus delayed midface distraction in a primate model using a new intraoral internal device. Plast Reconstr Surg 109:1600, 2002 21. Rachmiel A, Rozen N, Peled M, et al: Characterization of midface maxillary membranous bone formation during distraction osteogenesis. Plast Reconstr Surg 109:1611, 2002 22. Lewinson D, Maor G, Rozen N, et al: Expression of vascular antigens by bone cells during bone regeneration in a membranous bone distraction system. Histochem Cell Biol 116:381, 2001 23. Lewinson D, Rachmiel A, Rihani-Bisharat S, et al: Stimulation of Fos- and Jun-related genes during distraction osteogenesis. J Histochem Cytochem 51:1161, 2003 24. Sasaki A, Sugiyama H, Tanaka E, et al: Effects of sutural distraction osteogenesis applied to rat maxillary complex on craniofacial growth. J Oral Maxillofac Surg 60:667, 2002 25. Liu C, Song R, Song Y: Sutural expansion osteogenesis for management of the bony-tissue defect in cleft palate repair: Experimental studies in dogs. Plast Reconstr Surg 105:2012, 2000 26. Mao JJ, Wang X, Mooney MP, et al: Strain induced osteogenesis of the craniofacial suture upon controlled delivery of lowfrequency cyclic forces. Front Biosci 8:10, 2003 27. Wang E, Zhou S, Zhang J, et al: Three-dimensional sutural expansion osteogenesis to expand zygomatic bone: An experimental study. Br J Oral Maxillofac Surg 43:68, 2005 28. Cheung LK, Zhang Q: Radiologic characterization of new bone generated from distraction after maxillary bone transport. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 96:234, 2003 29. Cheung LK, Zhang Q: Healing of maxillary alveolus in transport distraction osteogenesis for partial maxillectomy. J Oral Maxillofac Surg 62:66, 2004 30. Cheung LK, Zheng LW: Effect of recombinant human bone morphogenetic protein-2 on mandibular distraction at different rates in an experimental model. J Craniofac Surg 17:100, 2006 31. Cheung LK, Zhang Q, Zhang ZG, et al: Reconstruction of maxillectomy defect by transport distraction osteogenesis. Int J Oral Maxillofac Surg 32:515, 2003 32. Henkel KO, Ma L, Lenz JH, et al: Closure of vertical alveolar bone defects with guided horizontal distraction osteogenesis: An experimental study in pigs and first clinical results. J Craniomaxillofac Surg 29:249, 2001 33. Boyne PJ, Herford AS: Distraction osteogenesis of the nasal and antral osseous floor to enhance alveolar height. J Oral Maxillofac Surg 62:123, 2004 34. Kawata T, Kohno S, Fujita T, et al: Transplantation of new autologous biomaterials into jaw cleft. J Int Med Res 29:287, 2001

2640 35. Kinoshita K, Hibi H, Yamada Y, et al: Promoted new bone formation in maxillary distraction osteogenesis using a tissueengineered osteogenic material. J Craniofac Surg 19:80, 2008 36. Rachmiel A, Potparic Z, Jackson IT, et al: Midface advancement by gradual distraction. Br J Plast Surg 46:201, 1993 37. Rachmiel A, Jackson IT, Potparic Z, et al: Midface advancement in sheep by gradual distraction: A 1-year follow-up study. J Oral Maxillofac Surg 53:525, 1995 38. Rachmiel A, Laufer D, Jackson IT, et al: Midface membranous bone lengthening: A one-year histological and morphological follow-up of distraction osteogenesis. Calcif Tissue Int 62:370, 1998

MINIPIG MODEL OF MAXILLARY DO 39. Glowacki J, Schulten AJM, Perrott D, et al: Nicotine impairs distraction osteogenesis in the rat mandible. Int J Oral Maxillofac Surg 37:156, 2008 40. Koski K: Cartilage in the face, in Bergsma D (ed): Morphogenesis and Malformation of Face and Brain. New York, NY, Alan R. Liss, 1975, p 231-254 41. Tavakoli K, Walsh WR, Bonar F, et al: The role of latency in mandibular osteodistraction. J Craniomaxillofac Surg 26:209, 1998 42. Aronson J, Shen X: Experimental healing of distraction osteogenesis comparing metaphyseal to diaphyseal sites. Clin Orthop Relat Res 25, 1993