The objective of the present study was to evaluate the outcomes of autogenous bone graft (AB) and bioglass (BG) associated or not with leukocyte-poor platelet-rich plasma (LP-PRP) in the rabbit maxillary sinus (MS) by histomorphometric and radiographic analysis. Twenty rabbits divided into 2 groups (G1, G2) were submitted to sinus lift surgery. In G1, 10 MS were grafted with AB and 10 MS were grafted with BG. In G2, 10 MS were grafted with AB + LP-PRP and 10 MS were grafted with BG + LP-PRP. After 90 days, the animals were killed and specimens were obtained, x-rayed, and submitted to histomorphometric, radiographic bone density (RD) and fractal dimension analysis. Radiographic bone density mean values (SD), expressed as aluminum equivalent in mm, of AB, BG, AB + LP-PRP, and BG + LP-PRP groups were 1.79 (0.31), 2.04 (0.39), 1.61 (0.28), and 1.53 (0.30), respectively. Significant differences (P < 0.05) were observed between BG and AB, and BG + PRP and BG. Fractal dimension mean values were 1.48 (0.04), 1.35 (0.08), 1.44 (0.04), and 1.44 (0.06), respectively. Significant differences were observed between BG and AB, and AB + LP-PRP and BG. Mean values for the percentage of bone inside MS were 63.30 (8.60), 52.65 (10.41), 55.25 (7.01), and 51.07 (10.25), respectively. No differences were found. No correlations were observed among percentage of bone, RD and FD. Histological analysis showed that MS treated with AB presented mature and new bone formation. The other groups showed minor bone formation. Within the limitations of this study, the results indicated that at a 90-day time end point, AB yielded better results than AB + LP-PRP, BG, and BG + LP-PRP and should be considered the primary material for MS augmentation.
Since the discovery of the biological phenomenon of osseointegration by Per-Ingvar Branemark in the 1950s,1 the use of osseointegrated implants has become a frequent dental procedure with a high rate of success.
However, alveolar process atrophy due to tooth extraction and sinus pneumatization in the posterior edentulous maxilla represents a challenge for the placement of endosseous implants. The resolution of this anatomic restriction requires a therapeutic approach called sinus lift procedure,2,3 which interposes a bone graft between the membrane and the sinus floor, increasing bone height and enabling the placement of optimal length implants.4–6
Although autogenous bone is considered the primary choice for maxillofacial bone reconstruction,5,6 some biomaterials such as bioglass have been used4,7,8 with a high degree of predictability in different surgical sites. This synthetic ceramic material is formed by silica, calcium, phosphorus, and sodium ions and is considered a biocompatible, atoxic, resorbable,8 and radiopaque material,9 with osteoconductive, hemostatic,10 and antimicrobial properties.11 Its main characteristic is to provide a strong chemical bonding to the bone.12
Platelet-rich plasma (PRP), an autogenous platelet concentrate, which is commonly used in combination with several biomaterials, is known to release a significant amount of growth factors at the surgical site, increasing vascularity and bone maturation.13,14 A new protocol to obtain leukocyte-poor PRP (LP-PRP) in rabbits shown on Table 1 and used in this study, which avoids leukocyte contamination and early platelet activation, has been previously described in the literature by the authors.15
There are no prospective controlled studies evaluating the quality of newly formed bone inside maxillary sinuses grafted with bioglass alone or in combination with LP-PRP by means of objective parameters. According to the European Association for Osseointegration,16 radiographic assessment of the grafted region is the most important method for postoperative control of sinus lift surgery. Radiographic bone density and fractal analysis of trabecular bone, in particular, allow an objective evaluation of the graft integration process, and determine whether the surgery was successful or not.17,18
Therefore, this study was undertaken to compare the outcomes of bioglass or autogenous bone associated with LP-PRP for maxillary sinus augmentation in rabbits by using radiographic analysis (density and fractal dimension) along with histomorphometric analysis, in order to validate both radiographic methods as indicators of bone neoformation.
Materials and Methods
Study design and surgical procedure
This study was approved by the Ethics Committee on Animal Experimentation/FOAR-UNESP. Twenty adult male rabbits (New Zealand) aged 28 weeks and with a mean weight of 4 ± 0.33 kg were submitted to sinus lift surgery.
Twenty animals were divided into 2 groups. In group 1 (G1), 10 maxillary sinuses (MS) were randomly grafted with particulate autogenous bone and 10 MS were grafted with bioglass (PerioGlas, NovaBone Products LLC, Alachua, Fla). In group 2 (G2) 10 MS were randomly grafted with a 1∶1 mixture of particulate autogenous bone and autogenous LP-PRP, and 10 MS were grafted with a 1∶1 mixture of bioglass (PerioGlas) and autogenous LP-PRP. The volume of either autogenous bone or bioactive glass and autogenous LP-PRP used for sinus floor augmentation was the amount required for completely filling sinus cavities. No sinus membrane perforations were observed.
In order to prepare LP-PRP, 8 mL of venous blood was collected from the marginal ear vein of each rabbit in a Vacutainer tube containing 1 mL of adenine citrate dextrose anticoagulant to prevent clot formation. This procedure was performed before general anesthesia; otherwise, resulting vasoconstriction would not allow the collection. After that, venous blood was centrifuged at 300g for 10 minutes, and the plasma was separated from the red and white blood cells and was set apart in a sterile plastic tube. The plasma was centrifuged again at 2860g for 5 minutes. The lower third of the preparation was considered to be LP-PRP. Blood cell counts were performed for each sample of LP-PRP in order to confirm the obtention of a platelet concentrate with a low leukocyte value. Finally, 10% CaCl2 was added to LP-PRP in a ratio of 10 PRP to 2 CaCl2 vol/vol for gelling PRP. The protocol15 is summarized in Table 1.
General anesthesia was induced after blood collection by intramuscular injections of 10% ketamine (veterinarian use; Ketamina Agener, Agener União, Brazil) at 0.35 mL/kg body weight and xylazine chlorhydrate (veterinarian use; Calmium, Agener União Saúde Animal) at 0.25 mL/kg body weight. Local anesthesia was accomplished with 2% mepivacaine with epinephrine 1∶100 000 (Mepiadre, DFL, Rio de Janeiro, RJ, Brazil).
Both lateral portions of the maxilla, as well as the right iliac bone region were shaved, and the surgical field was scrubbed with iodine solution—PVPI (Aster, Sorocaba, SP, Brazil). The surgical procedure was performed according to Watanabe et al (1999)19 (ie, skin incisions were made a few millimeters above the inferior border of the incisive bone and the maxilla on both sides). The subcutaneous tissue, the masseter muscle and the maxillary periosteum were incised and elevated dorsally. A trap door in the lateral antral wall of the maxilla was made with a diamond bur under profuse saline irrigation. The sinus membrane (Figure 1a and b) was elevated with a Lucas curette.5 A corticocancellous bone block was harvested from the right iliac crest of the animal (Figure 1c), particulated and placed in the maxillary sinus, associated or not with LP-PRP. Figure 1d shows maxillary sinus grafted with bioactive glass. The skin flap was sutured to promote healing.
All animals received intramuscular injections of 0.10 mg/kg body weight of antibiotics (veterinary use, small size; Pentabiótico, Fort Dodge, Saúde animal Ltda, Brazil) for 3 days. For pain relief, paracetamol (Tylenol Bebê, JANSEN-CILAG Farmacêutica, São José dos Campos, SP, Brazil) at 10 mg/kg body weight was administered. The animals were fed with solid chow and water ad libitum, and received veterinary care whenever needed throughout the experiment. No potential complications such as wound infection or swelling were encountered.
Ninety days after the experiment, a lethal intramuscular dose of 40% chloral hydrate was administrated and each animal's maxilla was dissected. Skulls were cut along the median line and both maxillary sinuses were set apart in order to avoid the overlapping of other structures with maxillary sinus radiographic images. After image acquisition, specimens were immersed in a 10% formaldehyde solution, pH 7.4, for 48 hours for fixation.
Radiographic acquisition and examination
Radiographs were taken using a GE 1000 X-ray unit and occlusal film (Kodak Insight F) with the X ray at a right angle to the maxillary sinus with the aid of a custom-made positioning device. Exposure ranges (70 kVp, 10 mA, 15 pulses) and focus-film distance (53 cm) were previously established in a pilot study. In all radiographic exposures, an aluminum 10-step wedge was placed next to the specimen (Figure 2a). Following automatic processing (Dent-X 9000), the films were digitized using a flatbed scanner equipped with a transparency adapter (AGFA, Snapscan 1236S).
The acquired digitized images were filed as tagged image file format (TIFF), at maximum resolution (2400 dpi) in the computer hard disk. For analysis, a region of interest (RI) and a region of control (RC) were selected on the digitized radiographic images (Figure 2a). RI corresponded to the maxillary sinus grafted area (Figure 2b), while RC to a nongrafted maxillary sinus located superiorly to the grafted sinus (Figure 2c). In order to standardize measurements, RI size was defined as the largest possible area drawn within the smallest maxillary sinus included in the sample (312 × 192 pixels). Eleven other areas (100 × 100 pixels) were also selected on each image. Ten of these areas corresponded to each step of the aluminum step wedge image, and one corresponded to the most radiolucent region (background) located in the vicinity of the thinnest aluminum step image, where the film was directly exposed to the X-ray beam.
Using ImageJ software (version 1.36b, NIH software, http://rsb.info.nih.gov/ij/index.html) on each image, radiographic density was measured in the region of each step of the aluminum step wedge and in the RI and RC. To eliminate film exposure and processing related influences, density values were plotted against the corresponding thickness of the aluminum step wedge in millimeters, using Microsoft Office Excel 2003. The obtained values provided the equivalent aluminum mm to the density measured in the specimen image.
Prior to fractal dimension determination, each image was processed with ImageJ software, as follows: (1) the original image (Figure 2a) was converted into binary (ie, a black and white image; Figure 2b); (2) the binary image was skeletonized and the bone tissue area was reduced to a single pixel line (Figure 2c); and (3) the skeletonized images were inverted (Figure 2d; ie, the bone area made white while the medullar region was made black). This last image was saved in bmp (bitmap) format for fractal dimension analysis. Fractal dimension was calculated by the box-counting method20 by means of Benoit 1.3 software (Trusoft International Inc, St Petersburg, Fla) as follows: If S = [log (number of windows) × log (window size)] then FD = (1−S).
Histologic processing and histomorphometric analysis
Histologic samples were prepared according to standard laboratory routine procedures. Specimens were fixed in 10% neutral buffered formalin for 1 month and were decalcified with Morse solution (1∶1 of sodium citrate 20% and formic acid 50%). The specimens were embedded in paraffin, and histologic serial sections were prepared. Microtomy of the paraffin blocks containing maxillary sinuses was performed using an automatic microtome (Jung Supercut 2065, LEICA Instruments GmbH, Heidelberg, Germany), and serial sections were obtained. Three 5-µm sections from each maxillary sinus were assessed. Two of them were stained with hematoxylin-eosin and one with Masson's trichrome. The first section was obtained from the center of the maxillary sinus, and the other two from its edges.
The mean percentage of bone filling in all 3 sections, measured by histometry, was used for data analysis. The assessed area comprised the entire bone extension formed by the vestibular cortex, alveolar ridge, palatal cortex, and the most prominent point of the sinus membrane in relation to the ridge located above the graft. Histomorphometric analysis was performed to quantify bone formation after 90 days. These measurements were taken under a light microscope at up to ×40 magnification (Olympus BX51). The images were then selected and transferred to a microcomputer (Pentium 4 Intel) through a photographic camera (Olympus CAMEDIA C5060, wide zoom) connected to the light microscope.
Using a specific image analysis software (ImageTool, UTHSCSA, version 3.0), the total area was determined and quantified as 100%. Strictly bone areas were delimited excluding other structures such as lacunae, cells, blood vessels, neural bundles, and biomaterial particles. Thus, the amount of bone present in that predetermined area was established. Values were expressed in percent to facilitate statistical comparison between groups.
Descriptive histologic analysis assessed the type and the quality of newly formed bone tissue, the characteristics of the cells present, and the presence or absence of bioglass (BG) or autogenous bone (AB) graft using a light microscope at ×4, ×10, and ×40 magnifications (Olympus BX51).
Student's paired t test was used to compare treatment outcomes as well as RI and RC in the same animal (AB vs BG and AB + LP-PRP vs BG + LP-PRP). Student's t test for independent samples was used for the other comparisons (AB vs AB + LP-PRP, BG vs BG + LP-PRP, AB vs BG + LP-PRP, and BG vs AB + LP-PRP). Bonferroni correction was used to keep the global significance level at 5%, considering a total of 6 comparisons. A multiple linear regression model was used to assess the relationship between density/fractal dimension and histometry.21
The mean values (standard deviation) expressed in terms of percentage for bone filling were 63.30% (8.60%) in maxillary sinuses grafted with autogenous bone, and 52.65% (10.41%) in those grafted with bioglass. Maxillary sinuses grafted with a combination of autogenous bone + LP-PRP and bioglass + LP-PRP presented bone filling percentages of 55.25% (7.01%) and 51.07% (10.25%), respectively. These differences were not statistically significant.
Radiographic bone density and fractal dimension
Table 2 shows the mean values of RI and RC radiographic density in the 4 groups. Mean density values observed in the RI of MS grafted with AB, BG, AB + LP-PRP, and BG + LP-PRP were significantly higher than in their RC. Significantly higher density values were observed in the RI of MS grafted with BG when compared with those grafted with AB or BG + LP-PRP.
Table 2 also shows the mean values of RI and RC fractal dimensions in the 4 groups. Mean values observed in the RI of AB, AB + LP-PRP, and BG + LP-PRP groups were not significantly different when compared to their RC. However, in the BG group, mean RI fractal dimension value was significantly smaller than in its RC.
Mean values of fractal dimension in the RI of the BG group were significantly smaller than values in the AB and the AB + LP-PRP groups. Further significant differences were not found in other comparisons.
Correlation between radiographic density/fractal dimension and histometry
Multiple regression analysis showed no correlation between radiographic bone density and histometry for AB (r = 0.23), BG (r = 0.12), AB + LP-PRP (r = 0.23), and BG + LP-PRP (r = 0.12), or between fractal dimension and histometry for AB (r = 0.21), BG (r = 0.08), AB + LP-PRP (r = 0.21), and BG + LP-PRP (r = 0.08).
Descriptive histologic analysis
Autogenous Bone Group
Newly formed bone was observed next to the bone plate corresponding to the alveolar process and the palatal wall (Figure 3a). This bone was at different maturation stages exhibiting trabecular arrangement with nodular disposition involving predominantly adipose medullary tissue, sometimes characterized by a highly vascularized fibroadipose structure. The newly formed bone tissue was either concentrically disposed forming Havers channels, a characteristic of mature bone, or highly cellularized involving randomly arranged osteocytes with osteoblasts surrounding the tissue in formation, regularly disposed amid osteoclasts (Figure 3b). This organization suggests the continuing formation and remodeling of the original tissue, which characterizes fully active bone tissue. No inflammatory cells were observed in this group.
Newly formed bone trabeculae were observed at the sinus membrane side of the maxillary sinus bone, exhibiting large medullary spaces filled with adipose tissue. Closer to the maxillary sinus membrane, vacuole formations containing an internal reddish structure suggestive of bioglass particle remains were seen in all cases (Figure 3c). Most of these structures were covered by a thin layer of newly formed bone tissue at different maturation stages. No inflammatory cells were observed in the region (Figure 3d). In one case, biomaterial fragments were encapsulated by fiber tissue without surrounding bone formation.
Autogenous Bone + LP-PRP Group
Maxillary sinuses grafted with a combination of autogenous bone and LP-PRP presented, from the sinus membrane side of the maxillary sinus bone to the center of the maxillary sinus, newly formed bone exhibiting large medullar spaces and irregular and thin bone trabeculae (Figure 4a). Sometimes these trabeculae presented mature bone with Haversian systems and were envolted by a connective tissue. However, no active osteoblasts and osteoid tissue were observed on the surface of the preexisting or newly formed bone, indicating that the bone formation process was finished (Figure 4b). No inflammatory cells were observed.
Bioglass + LP-PRP Group
Light microscopy showed a sparse neoformed trabecular bone in the vicinity of the preexisting bone wall with an irregular arrangement and medullar spaces predominantly filled with adipose tissue (Figure 4c). In the central areas of the maxillary sinuses and next to the sinus membrane, bioglass particles were integrated within newly formed bone, although restricted bone neoformation was observed. The great majority of the biomaterial particles were in fibrous tissue. No inflammatory cells or osteoblasts were encountered in the implanted material suggesting no signs of new bone formation or bone remodeling (Figure 4d).
Results of the present study, assessed by histometric analysis, showed that maxillary sinuses grafted with AB + LP-PRP presented bone filling of approximately 55%; sinuses grafted with BG + LP-PRP had approximately 51% of their area filled with bone, and sinuses grafted only with BG presented approximately 53% of bone filling. On the other hand, the percentage of bone formation in maxillary sinuses grafted only with AB was higher (63%).
Even though these differences among types of grafts were not statistically significant, descriptive histologic analysis showed that the results observed in MS treated with AB alone were superior to those with BG, AB + LP-PRP, and BG + LP-PRP, where minor bone formation was encountered. These findings can be explained by the fact that histometric analysis only included bone tissue quantification, not taking into account other parameters such as the percentage of blood vessels, biomaterial particles, medullary bone, and, most of all, lacunae. In other words, the percentage of bone observed in the group grafted with AB (63%) consisted of newly formed bone exhibiting characteristics of mature bone tissue under active neoformation and remodeling. The remaining 37% was mostly composed of medullary spaces, blood vessels, nervous fibers and bone cells (ie, structures commonly found in mature bone tissue).
On the other hand, in MS grafted with BG, the percentage of new bone formation was 53%, most of it found close to the sinus membrane side of the maxillary sinus bone, where blood supply was higher and osteogenic cells were present. The remaining 47% mostly included large medullary spaces filled with adipose tissue, few vasculonervous bundles and biomaterial particles still being resorbed, without any signs of new bone formation. The reason why there were no active osteoblasts and osteoid tissue on the surface of the preexisting or newly formed bone in the BG, AB + LP-PRP, and BG + LP-PRP groups remain unclear. Studies are needed to explain the factors involved.
Therefore, superior bone formation was definitively found with the use of autogenous bone alone. Furthermore, it is important to consider the presence of outliers in the AB group. If discrepant results were excluded from the original sample (63% ± 8.60%), a higher bone filling mean with a smaller standard deviation (67% ± 3.74%) would be found.
In summary, comparing the results of maxillary sinuses grafted with AB + LP-PRP and BG + LP-PRP and the groups treated only with AB or BG, we may conclude that the introduction of an autogenous component was not helpful in improving the outcomes. Actually, LP-PRP decreased not only the quantity of bone but also the quality of bone tissue formed. This must be due to the fact that less quantities of autogenous bone graft were placed inside maxillary sinuses because of the physical presence of LP-PRP gel. It is important to consider that the protocol used to obtain LP-PRP, described in a previous study,15 presented a reduced amount of leukocytes and a high concentration of intact platelets with a mean platelet enrichment percentage of 327% in relation to whole blood, a level considered adequate for a platelet concentrate.22,23
Some important aspects of the present study should be highlighted. First, to our knowledge, there are no other studies that have tested bioglass alone or in combination with LP-PRP in maxillary sinus grafting. The majority have investigated bioglass associated with autogenous bone, thus not demonstrating the actual osteogenic properties of bioglass. The second aspect refers to the advantages of testing materials in animals instead of humans. In fact, biopsies in humans can only be obtained from sites where implants will be placed and not from the total sinus area (ie, the assessment of surgical outcomes is primarily based on radiographic analysis, not on histometric analysis, as performed in the present study).
By analyzing radiographic density, it was possible to distinguish grafted maxillary sinuses from an empty area, regardless of graft type. As expected, density was higher in cavities filled with a given material than in a similar, but empty, cavity. Radiographic density was higher in maxillary sinuses grafted with BG (2.04 ± 0.39), because the residual material gives this higher density, as reported in the literature,24 followed by those grafted with AB (1.79 ± 0.31), AB + LP-PRP (1.61 ± 0.28), and BG + LP-PRP (1.53 ± 0.30). Significant differences were observed only between BG and AB, and between BG and BG + LP-PRP.
Fractal dimension analysis showed that in maxillary sinuses grafted with AB, AB + LP-PRP, and BG + LP-PRP, mean RI values (1.48 ± 0.04, 1.44 ± 0.04, and 1.44 ± 0.06, respectively) were very close to those observed in their RCs (1.47 ± 0.05, 1.46 ± 0.06, and 1.48 ± 0.05, respectively), suggesting a structural similarity of texture. On the other hand, the structural pattern of the area grafted with BG (1.35 ± 0.08) was not so close to the one grafted with AB or to its RC mean value (1.49 ± 0.04), because BG modifies the distribution of trabecular bone in relation to the AB group, even after 90 days. This finding must be seen with caution in clinical practice because no study was done yet investigating the behavior of dental implant surfaces in contact with bone tissue containing particles of bioglass.
Although no correlation was observed between histometry and density or between histometry and fractal dimension, it is noteworthy that both methods allowed the detection of important differences between treatments not only in the comparison of grafted areas with empty areas, but also in the assessment of graft type, as compared to clinical conditions evidenced by descriptive histologic analysis.
In a current study that is already being developed by the authors, other structures not quantified in this experiment, such as mineralized and nonmineralized tissue, as well as remaining graft particles, are under assessment. Moreover, the bone-to-graft contact areas are also being studied.
In conclusion, the radiographic and descriptive histologic analyses were both effective in assessing bone neoformation in the maxillary sinuses of rabbits, providing complementary information about surgical outcomes. Based on these methods, it was possible to infer that at a 90-day time end-point, maxillary sinuses treated with autogenous bone presented the best results, while bioglass produced less satisfactory results showing lower osteoconductive potential.
Finally, the extrapolation of these data to humans should be done with care. A prospective, randomized, controlled study to investigate clinical and radiographic aspects of bone grafts inside maxillary sinuses, as well as the behavior of the grafts tested herein when in contact with the implant surface, is necessary.