Immuno-histopathologic evaluation of mineralized plasmatic matrix in the management of horizontal ridge defects in a canine model (a split-mouth comparative study)

A split-mouth comparative experimental study design was conducted in accordance with the highest standards of ARRIVE guidelines under the approval of the Research Ethics Committee of the Faculty of Dentistry, Alexandria University for the conduct of research on experimental animals by the Faculty of Dentistry, Alexandria University (IRB No. 00010556-IORG 0008839). We used this study design to standardize all factors that could affect the regenerative procedure [26, 27].

Sample size was estimated assuming alpha error = 5% and study power = 80%, based on previous studies on the effect of BCP alloplast, MPM, and PRF on bone formation [28‐30]. To detect an effect size of 1.32 (difference in bone formation between experimental and control groups), using two-tailed test, the required sample size was calculated using G*Power 3.1.9.4 sample size calculator [31] to be 11 defects per group, increased to be 12 defects to make up for the loss to follow-up. Therefore, a sample of six dogs was the enough required sample for this study, as the aim of the study was to demonstrate intervention efficacy in a single group utilizing a split-mouth study design (4 defects per dog) (number of groups = 2) (total number of defects = 24 defects). To act as a negative control, another group of three dogs (12 defects; 4 defects in each dog) was added in which the defects were created and left empty. Accordingly, the total sample size was nine dogs (12 defects per group) (number of groups = 3) (total sample size = 36 defects).

For this study, we selected nine healthy, adult, male mongrel dogs (Canis familaris), about 18–24 months of age and weighing approximately 9–14 kg. The dogs were supplied and housed by the City of Scientific Research and Technological Applications (SRTA-City). We chose a canine model instead of a rodent animal since skeletally larger animals are more suitable in studies targeting dental tissue healing [32].

The mandibular third (P3) and fourth (P4) premolars of the nine dogs were assigned for this study. The study included three groups, each included 12 surgically created horizontal, well-contained critical-sized ridge defects. In group I (MPM experimental group), 12 defects on the left side of 6 dogs (2 defects in each dog) were created and managed with GBR using MPM composite and covered with an MPM compact layer followed by the application of collagen membrane (COLLAGUIDE™ by BIOLAND, Korea). In group II (PRF control group), 12 defects on the right side of the same 6 dogs (2 defects in each dog) were created and managed with GBR using PRF cuts mixed with BCP alloplast (Genesis BCP™ by Dio-implant, Korea) and covered with PRF compact membrane followed by the application of collagen membrane. Regarding the six dogs of the experimental and control groups, two dogs (number of defects = 8; 4 for each group) were randomly allocated using computer-generated list of random numbers and euthanized as follows: two dogs at 4-weeks, another two after 8-weeks, and the remaining two after 12-weeks. In group III (negative control group), 12 defects were created and left empty in three dogs (4 defects in each dog, 2 defects on the right side and 2 on the left) that were euthanized at 12 weeks. That was done to cover the whole study period showing that such a defect will not regenerate spontaneously without adjunctive measures, allowing for an unbiased strategy for analysis of the obtained results [33, 34]. This was done in contrary to the study conducted by Cakir et al. [35] to confirm the critical-sized defect dimensions.

The MPM was prepared according to the stated protocol published by our group [35]. Briefly, we draw 20 ml venous blood from the animal into anticoagulant-free plastic blood collection tubes (VACUETTE® Z No Additive by Greiner Bio-One GmbH, Austria; 10 ml each) for centrifugation (15 min at 2500 rpm). This led to the separation of blood into red blood cells in the bottom and an upper layer containing clear yellow plasma rich in leukocytes, platelets, mesenchymal stem cells, and fibrinogen. Then, we collected the upper layer and mixed it with BCP to form a homogenous mixture as shown in Fig. 1a–c.

PRF was prepared according to Chokroun’s technique [37]. Briefly, we draw 20 ml of venous blood from the cephalic vein of each dog and immediately divided into two anticoagulant-free glass blood collection tubes (SUMBOW® Plain Glass Vacuum Tube, China; 10 ml each) and centrifuged for 10 min at 3000 rpm. The fibrin clot formed in the middle of the tube after centrifugation, was gently removed and cut into small pieces and mixed with the BCP to fill the two created defects. The other fibrin clot was used to obtain the PRF membrane. (Fig. 1d–f).

The animals were anesthetized via an intravenous injection of sodium thiopental (13 mg/kg). We performed sulcular incisions and reflected the buccal mucoperiosteal flaps at the regions of the mandibular third (P₃) and fourth (P₄) premolars. Then, we extracted premolar teeth. Following Cologne Classification of Alveolar Ridge Defects (CCARD); the defects dimensions after extraction were surgically extended mesiodistally, buccolingually and apicocoronally to meet the criteria of horizontal medium sized defect 4–8 mm, inside the ridge contour (H.2.i) [38]. Accordingly, two horizontal well-contained critical-sized ridge defects were surgically created per jaw quadrant, each about 7-mm mesiodistally × 8-mm apicocoronally × 5-mm buccolingually [33]. (Fig. 2a,b) Acute-type surgical defects were used because these defects are well characterized, standardized wound models that are reproducible, allowing for an unbiased and appropriate strategy for analysis of the obtained results [34, 39].

For the experimental group (left side), we treated the defects with MPM composite covered with MPM compact layer. On the right side, the control group, we fill the defects with the mixture of the BCP with serum exudates and PRF cuts covered with PRF membrane. Then, we covered all the defects with collagen membrane and advanced the flaps to their original positions and sutured using simple interrupted 2–0 silk sutures (Ethicon silk suture, Johnson & Johnson, USA; Fig. 2c–h).

On the first postoperative day, animals were injected with ampicillin (Eipico Co., 10th of Ramadan City, Egypt; 1 g) intramuscularly and then received ampicillin administered in their food for 7 days. They were also injected intravenously with a nonsteroidal anti-inflammatory drug (Meloxicam DELTA PHARMA Factory-Industrial Zone B4,10th of Ramadan City, Egypt) on the first day. Dogs were placed on a soft diet throughout the postoperative period to reduce the possibility of local trauma to the site of operation. Sutures were removed after 10 days. If signs of surgical failure were noticed as wound dehiscence or infection, the defects were retreated and excluded from the study. Dogs were euthanized via an intravenous overdose injection of concentrated thiopental sodium.

After the animals were euthanized, jaw segments with the operated teeth were dissected and fixed in 10% buffered neutral formalin for 4 days. Segments were then decalcified in multiple baths of 10% trichloroacetic acid. After decalcification, the specimens were immersed in paraffin blocks to obtain semi serial (5-µm) thick histologic sections in the buccolingual direction through the entire mesiodistal plane of the premolars using a rotatory microtome. Finally, tissue sections were stained with hematoxylin and eosin (H&E) for general evaluation of tissue healing and cell activity and to show the sequential changes in regenerative features during osteogenesis between MPM and PRF [39]. To accomplish this, we examined H&E stained the sections at three follow-up intervals: 4, 8, and 12 weeks postoperatively. This was done to show the quality of newly formed bone and the distinction between immature woven bone and mature lamellar bone. In addition, Gomori trichrome stain was used to examine the new bone formation and collagen organization [40].

Histologic sections were analyzed quantitatively using ImageJ 1.46 r software (all obtained a magnification × 100) [41]. The mean percentage (%) of the newly formed bone surface area in the defect parameter was measured for the three groups at the three different observation periods. Five images were used for each section, and two different pathologists blindly evaluated all measurements, and the means were recorded.

Tissue sections were stained by Masson’s trichrome, and the stained sections were quantitatively and blindly analyzed using ImageJ 1.46 r software (all obtained at a magnification × 100) [41]. Five different microscopic fields were used for each specimen to quantify the collagen percentage.

Tissue sections were stained using osteopontin (OPN; Thermo Scientific, Cat. No. RB-9097-R7, USA) at a dilution ratio of 1:50 of rabbit polyclonal anti-OPN antibody. This was done to investigate early events in extracellular matrix mineralization. Immunohistochemical staining was performed using the labeled streptavidin–biotin complex method [42]. The immunoreactivity of the tissue sections was examined blindly by two different pathologists in five randomly selected microscopic fields at 400 × magnification to determine the intensity as the mean area percentage and were analyzed quantitatively using ImageJ 1.46 r software. Positive cells were counted in five different microscopic fields that demonstrated more intense staining.

Data analysis was performed using IBM SPSS software package version 22.0 (IBM Corp, Armonk, NY) [43]. Normality was checked for all variables using descriptive statistics, plots (histogram and boxplots), and Shapiro Wilk normality test. All variables showed normal distribution, so means and standard deviation (SD) were calculated, and parametric tests were used. Paired t test was used for comparisons between MPM and PRF groups, repeated-measures analysis of variance for comparisons between more than two periods or stages and post hoc tests (least significant difference) for pairwise comparisons.

The animals tolerated the surgical procedures well and recovered normally. Furthermore, all dogs survived the entire study period. There were no postoperative adverse reactions, such as allergies or infection. Moreover, neither exposure of the collagen membrane nor tissue dehiscence was observed.

We examined the tissue sections to show the sequential changes in regenerative features during osteogenesis between MPM and PRF. Four weeks postoperatively, tissue sections reveal the difference between dogs treated with MPM versus PRF. The defects treated with MPM appeared to be replaced by thick, dense bone trabeculae covered by periosteum and thick collagen membrane on the surface. In the PRF group the new bone appears as irregular, immature, unorganized woven bone starting to take the arrangement of primary osteons with multiple dilated blood vessels. (Fig. 3a–d).

Bone maturation continued during the 8-weeks follow-up, with a mature dense bone trabecula forming merged homogenous masses that filled almost all defects in the MPM group blending with the preexisting bone. However, in the PRF group, alternating bright and dark staining bone trabeculae were seen. The stain variation reflected the influence of the time factor on the amount and quality of the more mineralized forming bone, replacing areas of fibrous tissue. (Fig. 3e–i).

Throughout the entire defect, bone formation was evident and was made of well-organized, dense, compact bone masses formed of numerous equal-sized osteons in the MPM group 12-weeks postoperatively. On the other hand, the osteons in the PRF group were highly cellular, immature, and unequal in size, with a haphazard cellular organization around the large blood vessels (Fig. 3j–n).

In the current study, the negative control defect was left for 12-weeks to demonstrate whether any bone formation would spontaneously fill the critical defect space. The defect was infiltrated by large marrow spaces, with some granulation tissue in between with a limited amount of bone formation underlying the thick layer of fibrous tissue (Fig. 3o).

The MPM group demonstrated a statistically greater mean percentage of bone surface area at 4, 8, and 12 weeks postoperatively (85.24 ± 7.16, 95.28 ± 3.06, and 97.07 ± 3.47, respectively) than the PRF group (58.52 ± 12.08, 72.44 ± 19.83, and 80.54 ± 13.28, respectively; P < 0.001, 0.004, and 0.002, respectively; Fig. 4).

Upon quantifying the percentage of collagen, we found that defects treated with MPM induced regenerative bone formation after 4-weeks, with an average 17.11% ± 6.23% of collagen fibers compared with 30.17% ± 3.99% in those treated with PRF (P = 0.04). At the 8-weeks time point, the MPM and control groups showed a decrease in bone matrix collagen fibers average percentage, 10.86% ± 4.34% and 14.04% ± 1.96%, respectively, with no statistical significance (P = 0.39). The same results were noted during the 12-weeks in which a decrease in collagen percentage continued to occur, reaching 5.70 ± 1.34 and 10.17 ± 3.52 in the MPM and PRF groups, respectively (P = 0.25). Moreover, the decrease in collagen percentage in the MPM group that occurred during bone maturation across the three-time intervals showed no statistical significance (P = 0.11). Meanwhile, there was a significant difference in PRF collagen percentage between the 4- and 8-weeks time point (30.2% ± 4% and 14% ± 2%, respectively; P = 0.045; Fig. 5a–g).

After assessing the bone surface area, we analyzed the immunoreaction of OPN as a bone-specific protein found in the extracellular matrix. We observed high immune positivity to OPN in the MPM group as compared with the PRF-treated group (MPM 70.32 ± 7.52, 52.24 ± 2.41, 25.75 ± 1.94, respectively; PRF 62.25 ± 6.38, 43.97 ± 5.42, 20.96 ± 4.32, respectively) throughout the three-time point intervals. However, a statistically significant difference between MPM and PRF was noted only at the 8-weeks interval (P = 0.03). At the 12-weeks interval, a linear deposition of OPN immunoreaction was observed across the cement lines at the periphery of the Haversian system in the MPM group. On the other hand, a less homogeneously dispersed OPN immune positivity throughout the osteon was noted in the PRF group. Moreover, when we compared the reaction during the 12-weeks to the negative control group, we noted a decrease in the OPN reaction that was not statistically significant (P = 0.18; Fig. 6a–i).