Bone Growth

μ-CT of the femur/Tibia (trabecular/cortical) and lumbar vertebrae-V (trabecular) was performed using SkyScan 1076 scanner (SkyScan). Scanning was done at 70 kV, 100 mA using a 0.5-mm aluminum filter and exposure set to 590 ms. In total, 1800 projections were collected at a resolution of 6.93 μm/pixel. For purpose of carrying out reconstruction process NRecon software was used. Bone length was determined from the rendered 3D images in CTAn software. The distance from trochanter to edge of femoral condyles defined total femoral length. Tibial length calculated as distance from medial condyle to medial malleolus. Manual segmentation of 2D slice of sagittal images was done to isolate growth plates from the surrounding bone tissue in micro-CT images followed by reconstruction to render 3D images. It was from these images that growth plate height was measured using data viewer software. For trabecular region analysis, ROI was drawn at a total of 100 slices in secondary spongiosa at a distance of 1.5 mm from distal border of growth plates excluding the parts of cortical bone and primary spongiosa. In vivo measurement of LV5 trabecular was done by encompassing 50 continuous slices which start from the beginning of trabecular bone within spinal body (Li et al., 2016 ; Dempster et al., 2013 (link)). The CTAn software was used to analyze cortical bone by taking into consideration 350 consecutive image slides discarded from growth plate to leave out all trabecular regions, out of these 200 continuous images were selected as such. Various bone histomorphometric trabecular parameters (3D) and cortical parameters (2D) were analyzed by already established protocols (Dempster et al., 2013 (link); Srivastava et al., 2013 ). For determination of BMD of LV5, femur and tibia μ-CT scans were utilized as determined from VOI made for trabecular and cortical region. For BMD calibration, 2-mm-diameter hydroxyapatite phantom rods with known BMD (0.25 and 0.75 g/cm³) were used. For each analysis, the estimated BMD was determined based on linear correlation between the μCT attenuation coefficient and BMD (Srivastava et al., 2013 ).

Four healthy male canines weighing ~25 kg with a clinically healthy periodontium were used for the present study. The animal selection and management and the surgical procedures were approved by the Ethics Committee on Animal Experimentation (IR.UMSHA.REC.1397.639).
An intramuscular injection of a mixture of Ketamine 10% (Ketamine alfasan, Woerden, Holland, The Netherlands; 10 mg/kg) and Xylazine 2% (5 mg/kg) was used for deep anesthesia. The duration of anesthesia with this method was 1 to 1.5 h. To continue anesthesia, intubation was performed using a combination of oxygen with 1.5% halothane gas (Halothane BP, Nicholas Piramal India Limited, Mumbai, India). Routine dental infiltration anesthesia and lidocaine infiltration (Persocaine-E, Lidocaine HCL 2% + Epinephrine 1/80,000, Daroupakhsh pharmaceutical. Mfg. Co. Tehran, Iran) were used at the surgical sites to control pain and bleeding. Four canines were employed for this study, with a total number of 32 defects. Prior to surgery, oral prophylaxis with 0.2% chlorhexidine was performed. An intra-crevicular incision was made on the buccal aspect of the treated sextants. Following the elevation of the buccal mucoperiosteal flap, four square-shaped dehiscence defects were prepared just below the cemento-enamel junction (CEJ) with dimensions of 5 mm × 5 mm (width × length) on the root surface of the canine, the first and second premolars (distal roots), and the mesial root of the first molars in each side of the mandible (four similar defects in each side of the jaw). The bone defects were prepared using rotating burs under sterile saline irrigation. Root planing was performed using Gracey curettes and chisels, and the cementum and periodontal ligaments were completely removed over the exposed root in the defect area (Figure 3). The CEJ and the most apical portion of the denuded dentin surface were used as histopathological markers to determine the reference points for evaluating bone and periodontal regeneration in the area.
Each defect was washed with normal saline serum and the 32 defects were randomly assigned to four groups: (1) GTR using Botiss Jason^® membrane Botiss Biomaterials GmbH, Zossen Germany (Jason), (2) GTR using Smartbrane membrane, Regedent, Zurich, Switzerland (Smartbrane), (3) The novel 3D-printed membrane, and (4) no membrane (Control) (Figure 3). Each membrane was carefully adapted to the defect to cover 2 mm beyond the defect edges. The placement of the 3D-printed membrane was conducted in a way to have the large pore size (400–500 µm) at the side of the bone tissue and the small pore size (50 to 150 µm) at the side of the soft tissue. This placement was used because the GTR membrane serves to create a physical barrier between the regenerating bone and the surrounding tissue, which helps to prevent the soft tissue from infiltrating the area where the bone is being regenerated and requires a smaller pore size at the side of the soft tissue. Moreover, 400–500 µm is appropriate for the growth of bone tissue.
The flap was repositioned and sutured tightly using 3-0 ePTFE (Osteogenics Biomedical, Inc., Lubbock, TX, USA) at the cemento-enamel junction. The flaps completely covered the membranes in a tension-free closure. The commercial membranes used in this study were porcine pericardium-derived collagen membranes.

After 9 weeks, the reconstruction surgery was performed as previously described [3 (link)]. The bioreactor on the fifth rib was utilized for the reconstruction, assuming bone growth could be observed and the vascular supply could be located. Briefly, two incisions were created over ribs 4 and 8, the bioreactors on ribs 3, 5, 7, and 9 were located by blunt dissection. The bioreactor on rib 5 was isolated with its adjacent intercostal artery and vein. All other bioreactors were harvested without vasculature and fixed in 10% neutral buffered formalin prior to analysis of bone formation characteristics.
An incision was created at the inferior border of the mandible. The lateral plate, bicortical screws, and space maintainer were removed. In the event of an infection around the space maintainer, the defect area was debrided and irrigated with copious amounts of normal saline. If a mucosal dehiscence was present, the mucosal opening was sutured closed. The vascular anastomoses were performed to branches of the facial artery and vein using a surgical microscope (Leica Microsystems). As all sheep had developed a robust callus medially that provided sufficient mechanical support for the mandible, smaller monocortical self-tapping screws and lateral plate (as in the partial segmental defect [3 (link), 10 (link)]) were utilized to hold the transferred bioreactor tissue in place.