Autogenous osseous grafts are often considered ideal for osseous regeneration. However, their use presents limitations, such as high morbidity to the donor site, limited availability, and relatively high and unpredictable resorption [
Alloplastic grafts are natural or synthetic materials that function as bone substitutes. Among the alloplastic grafts, few possess osteoconductive properties, whereas many of them act as space fillers. A bone substitute, when grafted in an osseous defect, should provide a proper environment for new bone formation and maintain the space where new bone could grow in [
Porous titanium granules were first used in Orthopedics [
PTG has been studied alone or in combination with xenograft [
The aim of the present study was the histological and histomorphometric comparison of the osseous regeneration and the graft integration in experimentally induced osseous defects in the rabbit femur treated with porous titanium granules or autogenous osseous graft.
Forty-five New Zealand white male rabbits (3.5–4 kg) were used. The animals were randomly assigned to three groups of 15 each: (1) treatment of the surgically induced femoral osseous defect with porous titanium granules (TIGRAN™-PTG) and resorbable collagen membrane (PTGM), (2) treatment of the surgically induced femoral osseous defect with autogenous osseous graft and resorbable collagen membrane (AGM), and (3) treatment of the surgically induced femoral osseous defect with resorbable collagen membrane alone (CM or control). They were acclimatized to the experimental conditions for one week prior to the study initiation.
The animals were housed one per cage in stainless steel wire net cages, fed a standard rodent diet with free access to water, and exposed to a 12 h light/dark cycle, at room temperature 18–22°C and relative humidity 55–65%. All animals were kept in their allocated cages for the entire study duration. The study was conducted in accordance with guidelines approved by the Council of the American Psychological Society (1980), the European Communities Council Directive of 24 November 1996 (86/609/EEC), and the Hellenic Presidential Decree 160/91. The study was performed meeting ARRIVE guidelines and approved by the University of Athens Ethics and Research Committee (Ref. 167/12.05.2011) and by the Veterinary Directorate of the Prefecture of Athens.
On day one of the study, the femoral osseous defect was surgically induced and treated with the allocated treatment for each animal. For premedication, 25 mg/kg ketamine hydrochloride (Ketaset, Ceva) and 5 mg/kg xylazine (Rompun, Bayer) were administered intramuscularly. Following premedication, 4 mg/kg carprofen (Rimadyl, Pfizer) and 5 mg/kg enrofloxacin (Baytril, Bayer) were administered subcutaneously for analgesia and antimicrobial prophylaxis, respectively. Additionally, thiopental sodium (Pentothal, Abbott) was administered intravenously, through a 21 G butterfly catheter placed in the marginal ear vein, in N/S 0.9% (10 ml/kg), for maintaining general anaesthesia. The animal was intubated with a 3.0 mm ID cuffed endotracheal tube (Mallinckrodt) and attached to a small animal ventilator (Harvard Apparatus model 683). The animal was monitored during surgery with pulse oxymetry (Kontron Instruments Pulse Oximeter 7840) and blood pressure monitoring (Dinamap-Criticon Vital Signs Monitor 1840). The right lateral femoral condyle was depilated, disinfected, and covered with a sterile drape. The incision was made, and the flap was elevated at the predetermined site of the right femur. Then, an osseous defect, 6 mm in diameter [
Immediately after euthanasia, the right femur was dissected and bone blocks containing the area of the osseous defect were removed. After fixation in buffered 4% formaldehyde solution [
Stained sections were photographed with the Olympus Dot Slide system 2.4 (Olympus, Tokyo, Japan), resulting in overview images with a resolution of 2.5726
In order to gather information about the quantity of bone regeneration within the defect, the percentage of newly formed bone within the complete region of interest (volume of newly formed bone per tissue volume; nBV/TV) was assessed. The amount of the bone substitute in the defect area (bone substitute volume per tissue volume; BSV/TV) was calculated to characterize the volume stability and packing density of the grafted materials. As new bone tissue can only be laid down in areas where there is no bone substitute material (PTG or AGM) present, the size of this available space between the particles has great influence on the amount of bone neoformation. Therefore, the percentage of newly formed bone in the available space between the bone substitute materials (newly formed bone per available volume; nBV/Av.V) was also calculated [
Positions of the regions of interest (ROIs) for the histomorphometric evaluation. A blue frame surrounds the cortical area and a green frame the medullary ROI. On the right side, classified newly formed bone is depicted in red and bone substitute material in yellow.
For variables nBV/TV, nBV/Av.V, Co.V/TV, BS.V/TV, and BBSC, descriptive statistics (mean, median, standard deviation (SD), interquartile range (IQR), minimum, and maximum) and boxplots were created. In order to achieve approximate normal distribution, the natural logarithm of nBV/TV, nBV/Av.V, Co.V/TV, and BS.V/TV was used, residuals were checked graphically. Models included area and treatment as independent variables and ID as the random variable. ANOVA was calculated to test for the influence of treatment. Post hoc Tukey tests were performed. As secondary hypotheses, the medullary and cortical areas were tested. Fisher–Pitman permutation tests were calculated for BBSC and PIR.
Nine animals (three in each group) died of anesthetic complications or postsurgical infection before they could finish the scheduled duration of six weeks and had for this reason to be excluded from the study. They were replaced by nine new animals which finished the planned course of experimentation. Therefore, 45 animals (15 in each group) in total completed the experiment and were finally studied. The survival rate was 83.33% for each animal group.
After six weeks, the old autochthonous and newly formed bones were readily distinguishable (Figures
Histological characterization of PTG osseointegration. In the center, the distribution of PTGs (gold-colored) within the drill hole is visible in an overview image. (a) Detail of the strong osseointegration in the cortical compartment. (b) Much weaker osseointegration in the medullary region. (c) In large areas, the PTGs showed no signs of osseointegration at all. (d) Remnants of the resorbable membrane were still detectable. (e) Newly formed bone tissue consisted mostly of woven bone (black arrowhead) compacted by parallel-fibred bone (white arrowhead). This primary bone had already been partly remodeled into secondary lamellar bone (asterisk). (f) Border of the drill hole (white arrowhead). (g) Newly formed bone tissue was laid down in the pores of the PTGs. (h) The spaces between the granules were mostly filled with fatty marrow (microphotograph of undecalcified thin‐ground section; Levai–Laczko stained; length of scale bar equals 200
All three treatment groups had healed well, granules were evenly distributed inside of the drill holes, and no strong displacements into the surrounding tissues were observed, but there were differences in the amount of regenerated bone tissue and also in its spatial distribution. In general, the strongest bone formation took place in the region of the aperture in the cortical bone (Figure
Comparison of defect healing and histological osseointegration in the three treatment groups (views of complete defects above and overview images of the lateral condyle below). The control group (resorbable membrane alone) showed very little new bone tissue in the marrow compartment of the defect. Bone formation was clearly stronger in the group treated with autogenous bone. Osseointegrated remnants of the graft (asterisks) were still present. In the defect filled with PTG, many titanium granules were detectable and well integrated into new bone tissue. The composite of granules and surrounding bone fills the defect to a larger extent than is the case in the other groups. Black dashed lines indicate the borders of the drill holes (microphotograph of horizontal undecalcified thin‐ground sections, Levai–Laczko stained).
In the deeper regions of the former drill hole, the differences were more pronounced. While in the CM group, these areas appeared to be almost free of bone tissue or were bridged by sparse cancellous trabeculae (Figure
There were no detectable differences in vascularization, tissue maturity, or the structure of the nonmineralized tissues in the test area, which predominantly consisted of fatty marrow (Figure
Descriptive statistics for nBV/TV, nBV/Av.V, BS.V/TV, Co.V/TV, BBSC, and PIR in the cortical and medullary areas are presented in Table
Descriptive statistics for nBV/TV, nBV/Av.V, Co.V/TV, BS.V/TV, BBSC, and PIR per treatment group and anatomical region.
Parameter | Area | Group | Mean | Median | SD | IQR | Min | Max |
---|---|---|---|---|---|---|---|---|
nBV/TV | Cortical | AGM | 27.3 | 25.4 | 6.9 | 7.4 | 14.1 | 39.3 |
CM | 17.7 | 19.6 | 10.8 | 14.0 | 0.6 | 38.5 | ||
PTGM | 23.8 | 21.0 | 12.0 | 11.0 | 4.8 | 54.8 | ||
Medullary | AGM | 12.2 | 13.4 | 6.3 | 6.9 | 0.0 | 25.3 | |
CM | 1.5 | 0.0 | 5.6 | 0.0 | 0.0 | 21.8 | ||
PTGM | 10.6 | 5.5 | 10.1 | 14.1 | 0.0 | 35.6 | ||
nBV/Av.V | Cortical | AGM | 29.6 | 29.5 | 9.3 | 8.9 | 14.3 | 52.0 |
CM | 17.7 | 19.6 | 10.8 | 14.0 | 0.6 | 38.5 | ||
PTGM | 39.4 | 40.1 | 15.6 | 13.1 | 8.5 | 66.0 | ||
Medullary | AGM | 13.7 | 14.1 | 7.4 | 9.7 | 0.0 | 27.3 | |
CM | 1.5 | 0.0 | 5.6 | 0.0 | 0.0 | 21.8 | ||
PTGM | 18.9 | 13.9 | 15.9 | 22.1 | 0.0 | 51.0 | ||
Co.V/TV | Cortical | AGM | 33.8 | 31.9 | 11.8 | 12.7 | 15.2 | 63.8 |
CM | 17.7 | 19.6 | 10.8 | 14.0 | 0.6 | 38.5 | ||
PTGM | 64.5 | 66.7 | 9.2 | 9.4 | 47.7 | 81.1 | ||
Medullary | AGM | 20.7 | 21.1 | 11.9 | 16.9 | 0.0 | 41.8 | |
CM | 1.5 | 0.0 | 5.6 | 0.0 | 0.0 | 21.8 | ||
PTGM | 56.1 | 54.5 | 11.4 | 13.5 | 31.4 | 78.0 | ||
BS.V/TV | Cortical | AGM | 6.5 | 4.8 | 6.5 | 7.8 | 0.5 | 24.4 |
CM | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | ||
PTGM | 40.7 | 43.4 | 10.4 | 11.2 | 16.9 | 53.9 | ||
Medullary | AGM | 8.5 | 7.5 | 7.7 | 6.3 | 0.0 | 28.2 | |
CM | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | ||
PTGM | 45.5 | 48.4 | 11.9 | 13.5 | 17.8 | 61.2 | ||
BBSC | Cortical | AGM | 84.3 | 85.4 | 6.8 | 9.6 | 70.6 | 92.3 |
PTGM | 26.9 | 26.3 | 11.8 | 13.8 | 5.9 | 47.0 | ||
Medullary | AGM | 77.8 | 78.5 | 11.2 | 9.7 | 51.3 | 100.0 | |
PTGM | 14.3 | 13.4 | 11.2 | 15.7 | 0.0 | 34.2 | ||
PIR | Cortical | AGM | 100.0 | 100.0 | 0.0 | 0.0 | 100.0 | 100.0 |
PTGM | 82.4 | 83.3 | 13.5 | 18.8 | 57.1 | 100.0 | ||
Medullary | AGM | 100.0 | 100.0 | 0.0 | 0.0 | 100.0 | 100.0 | |
PTGM | 74.6 | 83.7 | 18.2 | 23.5 | 30.4 | 93.3 |
Mean and median values, standard deviation (SD), interquartile range (IQR), minimum and maximum values.
Post hoc tests for the comparisons among the treatment groups.
Area | Group comparison | nBV/TV | nBV/Av.V | Co.V/TV | BS.V/TV | BBSC | PIR |
---|---|---|---|---|---|---|---|
|
|
|
|
|
| ||
Cortical area | AGM vs CM | 0.024 | 0.0108 | <0.001 | <0.001 | <0.001 | <0.001 |
PTGM vs CM | 0.197 | <0.001 | <0.001 | ||||
PTGM vs AGM | 0.644 | 0.6447 | 0.0083 | ||||
Medullary area | AGM vs CM | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 |
PTGM vs CM | <0.001 | <0.001 | <0.001 | ||||
PTGM vs AGM | 0.562 | 0.9905 | <0.001 |
Histomorphometric results for volumetric data: (a) percentage of newly formed bone tissue in the whole region of interest; (b) percentage of newly formed bone in the spaces available in between the bone substitute particles; (c) percentage of bone substitute material in the region of interest; (d) percentage of the composite consisting of newly formed bone plus bone substitute material in the region of interest.
Histomorphometric results concerning osseointegration: (a) percentage of the surface of the bone substitute particles that is in contact with newly formed bone in the region of interest; (b) percentage of the bone substitute particles that have at least one contact to the newly formed bone in the complete defect area.
For nBV/TV, there was a statistically significant difference in the comparisons of PTGM and AGM to CM in the medullary area and in the comparison between AGM and CM in the cortical area (Table
For nBV/Av.V, comparisons of PTGM and AGM to CM presented statistically significant differences, whereas the comparison between PTGM and AGM did not show statistical significance both in the cortical and medullary areas (Table
For Co.V/TV, comparisons between the groups presented statistically significant differences both in the cortical and medullary areas (Table
The possible influence of the treatment on nBV/TV, nBV/Av.V, and Co.V/TV was studied by one-way ANOVAs, for medullary and cortical areas separately. For cortical areas, the
Then, the possible different behavior of the variables nBV/TV, nBV/Av.V, and Co.V/TV in the medullary and cortical area was studied. For the hypothesis that the medullary and cortical areas did not differ in effect, the
The present study investigated the osseous regeneration and graft integration in well-established standardized defects in the rabbit femur [
In summary, osseous regeneration was comparably strong in the AGM and the PTGM group but significantly lower in the CM group. PTGM showed significantly higher graft volume stability but weaker osseointegration than AGM. Bone regeneration turned out to be highly associated with the stability of the grafting materials used.
In detail, PTGM exhibited a lower percentage of newly formed bone in the total region of interest than AGM. However, nBV in the available volume (i.e., the space that is not occupied by bone graft) was higher than in the AGM group. This paradoxical phenomenon can be explained by the fact that autogenous bone undergoes extensive resorption leading to the reduction of the graft volume as reported in various studies [
In other words, at the moment of grafting, tissue volume was the same for both groups since all drill holes had the same standardized dimensions. In the AGM group, new bone could continuously replace the resorbed autograft in an increasing available space. Within the constant available space between the titanium granules, this group performed as well as AGM. The osseous regeneration thus might be regarded as similarly effective in PTGM as compared to AGM.
No foreign body reaction was observed in either group. In addition to this good biocompatibility [
From a clinical point of view, the volume of the newly formed bone and the volume of the graft material together might be regarded as a biomechanically relevant entity. In oral implantology and orthopedic surgery, the osseointegrated bone graft serves as a physical foundation for the insertion of endosseous implants.
In this study, concerning the amount of “materials” that are biomechanically relevant (new bone plus bone substitute), PTGM was superior to AGM since these values were significantly higher for PTGM. In case of implant placement, the presence of a larger volume of “biomechanically relevant materials” should theoretically provide better support and stability for implants.
However, these positive findings in the PTGM group were somewhat diminished by the weaker osseointegration of the titanium granules. Bone-to-bone-substitute contact was lower for PTGM than AGM in both the cortical and medullary areas. In fact, a significant percentage (about 25%) of the PTG particles was not in contact with the newly formed bone at all while literally every piece of remaining, unresorbed autogenous bone showed the presence of some new bone on its surface. This proves that autogenous bone was better osseointegrated than PTG at six weeks. Such a lower bone-to-bone-substitute contact was found earlier for PTG as compared to xenograft at six months after maxillary sinus augmentation in rabbits [
Only when the bone substitute material is in direct contact with the network of newly formed bone, mechanical forces can be transferred from the implant to its osseous environment. A tighter connection between graft and newly formed bone therefore should ensure improved stability. In this aspect, PTG appears biomechanically inferior to autogenous graft, since a significant portion of the PTG particles can probably not contribute to provide clinically relevant mechanical stability.
Combining the findings on composite volume and on bone-to-bone-substitute contact, the present study showed a contradictory picture. PTG, as compared to autogenous graft, provided a significantly larger volume of materials for potential mechanical support at six weeks but actually achieved less contact with the newly formed bone and thus only undeterminable biomechanical potential. These considerations suggest that a larger amount of material available for osseointegration, which might be thought to offer a more stable environment for implant placement, does not necessarily result in better biomechanical stability. From the data at hand it cannot be deduced of which practical relevance these differences are. Only biomechanical testing of implants placed in areas augmented with the two materials studied here could answer this question.
It is legitimate to speculate that a closer contact and tighter connection between PTG and newly formed bone might be achieved with time. A possible delay in the histological osseointegration of bone substitute particles cannot be excluded with the use of PTG. Such a delayed healing compared to autogenous bone chips was documented when anorganic bovine bone was filled into extraction sockets [
Another noteworthy finding was the regional difference of osseous regeneration within the femoral defect: In all groups, less bone was formed in the medullary region than the cortical compartment of the defect. The higher regenerative potential of the cortical region might be attributed to the close local relationship to the periosteum [
In the present study, PTG particles were better integrated in the cortical region than the medullary region of the defect. This suggests the presence of an increased biomechanical stability in the cortical area, a fact that is of practical clinical importance since the critical primary stability of implants is strongly influenced by the amount of cortical bone present [
The lower regenerative potential of the medullary compartment allows conclusions about how bone substitute materials might perform in areas with impaired healing like the deeper regions of a sinus lift [
In conclusion, it can be said that in the present study, grafting the osseous defect with PTG was similarly effective in achieving osseous regeneration as with autogenous bone. The graft volume stability of the nonresorbable PTG was clearly far better than that of autogenous bone graft which was quickly resorbed. In total, more biomechanically relevant material, i.e., the combined volumes of newly formed bone and bone substitute material, was present in PTG-treated sites. These results suggest that PTG can equal autogenous bone grafting in facilitating bone formation and surpasses it as a space filler and in providing long-lasting graft stability. These properties might be advantageous for an application in defects where resorption is known to proceed fast such as in extraction sockets.
However, a large percentage of the PTG particles were not in contact with newly formed bone. It cannot be deduced from the results of this study if and to which degree this fact can reduce the mechanical stability of the augmented area. Additional biomechanical testing would be necessary to answer this important question.
The data used to support the findings of this study are available from the corresponding author upon request.
The authors declare that there are no conflicts of interest regarding the publication of this article.
The authors thank (1) Dr. Dimitra Tsarouchi and Dr. Alkistis Pantopoulou for their assistance in the pilot study; (2) Tigran Technologies AB SE-205 12 Malmö, Sweden, for providing the porous titanium granules free of charge (Natix®, PTG); and (3) Geistlich Pharma AG, Division Biomaterials, Bahnhofstrasse 40, 6110 Wolhusen, Switzerland, for providing the resorbable collagen membrane (Geistlich Bio-Gide®) free of charge.