Induced Experimental Peri-Implantitis and Periodontitis: What Are the Differences in the Inflammatory Response?

In clinical practice, the problems of periodontitis (POIS) and of peri-implantitis (PIIS) are frequently encountered and are of a major concern to dentists. Treatment endeavors for these inflammatory diseases aim at eliminating the inflammation and infection (if present), improving the local hygiene conditions, and re-establishing tooth integration or implant reosseointegration.^1–5  However, therapeutic measures such as mechanical debridement, adjunctive antiseptic therapy, adjunctive antibiotic therapy, and hygiene instruction,^6,7  are often impaired by lack of sufficient patient compliance, smoking, and so on, and thus are of limited effectiveness. Efforts to find novel and/or alternative therapies are continuously being made using animal models that allow the mimicking of the human pathology.^3,8  Adequate experimental animal models are thus needed for disease simulation under laboratory conditions. Currently no useful in vitro test systems for the simulation of POIS and PIIS are available, even though from the perspective of the 3R philosophy,⁹  this would be highly welcome. Some encouraging attempts have been made, however.^10,11  The animal models developed thus far to induce experimental POIS and PIIS mostly use the ligature-model^12–14 ; however, these models are controversial with respect to their usefulness and impact. Some authors see this model rather as a foreign body study^15,16  than a useful inflammation model; moreover, a number of factors such as the nutrition (type of food used), the local hygiene, the bacterial flora, the animal species used, and so on,^12,17,18  have different influences that render the models quite diverse in the resulting pathology. However, although this diversity is observed in model-dependent outcomes, it surprisingly is generally found that similar cell populations of inflammation are involved in these models but to different degrees.^12,17,18  Using human biopsy materials, it was found that PIIS and POIS indeed do not show significant differences in cellular histopathology.^10,19–22 

Respecting the possible pathogenic factors for PIIS and POIS, many authors assume that PIIS develops after a successful implantation of an endosseous implant from an imbalance between bacterial challenges and the host response.²³  Results from clinical and experimental studies have revealed that the tissue's response to plaque formation at both the tooth site and the site of dental implants are indeed similar,²⁴  in particular, in humans.¹⁹  A large number of studies have shown that persistent inflammatory activities result in the destruction of soft and mineralized periodontal tissues.^25,26 

One particular aspect of the inflammatory response that hitherto was not specifically addressed relates to the dendritic cells that capture and present antigens to B and T cells of acquired immunity. Activated CD4 T-helper cells produce subsets of cytokines that define phenotypically distinguished immune responses: Th1 and Th2.²⁷  Each phenotype is characterized by unique signaling pathways and expression of specific transcription factors, notably T-β for Th1 and GATA3 for Th2. A number of previous studies focused on the aspects of cytokine activities associated with T cells in periodontal diseases, such as interferon (IFN)γ,^28–30  interleukin (IL)-4,^29,31–33  IL-17,^30,34,35  and IL-10,^32,36  secreted by Th1, Th2, Th17, and T-regulatory cells (Tregs), respectively, as well as peri-implant cervical fluid³⁷ ; the involvement of T cells in experimental animal models has not been studied directly in PIIS and POIS, as research focused mainly on human POIS (biopsy materials^19,38–41), and the results of these studies were found to be quite variable,⁴²  some showing basically no differences.¹⁹ 

In this comparative study, we hypothesize that the immunohistologic response in the induced PIIS in dogs is different from induced POIS, with disease-specific pathophysiologic events operating. In an experimental dog study, we induce POIS and PIIS using the well-established ligature method^12,15,43  and compared the composition of the inflammatory cellular populations in the inflamed periodontal tissues; moreover, we specifically investigated the T-cell reaction pattern of cells infiltrating the gingival tissues, with a particular view to Th1 and Th2 cellular activities, and analyzed secondary effects such as the associated bone resorption degrees in the experimental dog model for POIS and PIIS.

Six 1-year-old male beagle dogs were used in this study. Permission for the animal experiments was obtained from XXXXX. The rules and requirements for a healthy, well-protected and humane animal life were assured for the dogs from the beginning to the end of the study.

For the extraction of the first and second premolars on both sides of the mandible, dogs were put under general anesthesia (3% pentobarbital was injected intraperitoneally), and standard tooth extraction techniques were used.

During the experimental course (Figure 1) with the need of repeated general anesthesia (for tooth extraction, implant placement, and ligature placement), we lost 2 dogs during general anesthesia because of cardiac arrest.

After a 3-month healing time, 1 implant was placed on each side (ie, left and right side of the mandible). Then, again 3 months after implant placement, cotton ligatures (surgical cotton suture materials were obtained from Cibei Company) were placed in a submarginal position around the neck of the implants (obtained from Jinhuan Medical Company); these titanium implants had surfaces of the SLA type (sand-blasted, acid etched) to induce PIIS (Figure 1). The same method was used at the same time point on the first molar to induce POIS on each side (ie, left and right side). Probing depths were obtained by standard clinical measuring approaches. PIIS was defined to have been achieved if probing depth measurements were found to be larger than approximately 2.1 mm, according to Lang et al¹⁸  and Fickl.⁴⁴  Because this study is a comparative study between POIS and PIIS within the same individuals, no untreated control groups were established. Ligatures were examined twice a week and retied if the ligature had loosened. No oral hygiene measures were applied after ligation. All animals received standard food regimes and were euthanized according to the locally approved university protocol (pentobarbital overdose by intravenous injections) 3 months after ligature placements (Figure 1).

One control radiograph was taken just before implant placement at each implantation site (Figure 1); thereafter, radiographs were obtained at the time of ligature placement and at euthanasia (Figure 1) to assess the depth and width of bone resorption. The radiographs were analyzed using a microscope (Olympus SZH 10 Research Stereo, Olympus). To quantify the degree of PIIS-induced bone resorption effects, the distances between the top of the implant and the bottom of the bone defect were assessed as an indicator of the depth of bone resorption, and the distance between the lateral surface of the implant and the lateral bone wall of the defect was measured to quantify the bone resorption in width (Figure 2). In the POIS teeth, the distance between the cement-enamel junction (CEJ) and the bottom of the bone defect was measured as the depth of the bone resorption, and the distance between the lateral surface of the teeth (in the mesio-distal plane) and the lateral bone walls of the defect (also in the mesio-distal plane) was measured to obtain the width (Figure 2) of bone resorption. The lateral width measurements were performed using a series of lines equally spaced and placed perpendicular to the implant surface (or the tooth root surface) area (Figure 2), with a random starting point at the implant (or tooth root) proximal level (oral cavity side) such that approximately 3 to 4 measurements within the gaps were obtained toward the distal end of the implant (or the tooth root apex, respectively).

After gentle surgical removal of marginal gingiva under general anesthesia (before overdosing with pentobarbital for euthanasia) at the end of the experiment (Figure 1), this gingival material was processed for flow cytometry analyses (to be compared with the buccal gingiva removed before tooth extraction at the beginning of the experiment; Figure 1). The remaining materials (bone with the whole teeth and bone with the whole implants) were then dissected free with surgical saws and chemically fixed in unbuffered 10% (v/v) formaldehyde, containing 1% (v/v) CaCl₂ (to prevent decalcification), for 2 days at room temperature, and then washed with tap water for 24 hours. Thereafter, samples were dehydrated in 70% ethanol for 10 days, in 80% ethanol for another 10 days, and then in 95% ethanol for 10 days; samples were finally dehydrated in 100% ethanol for 3 rounds of 10-day periods. The samples were subsequently placed into pure xylene twice for 10 days each, and then in MMA + 15% dibutyl phthalate for 15 days, MMA + 15% dibutyl phthalate + 0.5% Perkadox for 15 days at 4 °C, and MMA + 20% dibutyl phthalate + 1% Perkadox at 12°C until they were polymerized.

Using a Leco diamond saw (Model VC50), the tissue blocks were cut into 5 to 7 slices, each one of about 600-μm thickness and 1 mm apart, according to a systematic random sampling protocol,^45,46  glued to plastic holders, and polished down to a thickness of about 80 μm before surface staining with McNeals tetrachrome and acid fuchsin⁴⁷  and examining with a light microscope (Nikon K15018, Nikon).

Low- and high-magnification images were taken of the POIS and PIIS areas and printed on photographic papers for comparative visual examination by 2 independent observers: 1 pathologist and 1 medical scientist. Cell types specifically analyzed and estimated in numerical area density were, according to previous publications,^22,48  the following: fibroblasts, polymorphonuclear neutrophils, lymphocytes, plasma cells, macrophages, mast cells, and foreign body giant cells.¹⁶ 

CD4⁺ T-cell lines were obtained from gingival tissues measuring 10 × 10 mm that were removed by excision buccally (negative control tissue) before surgery and lingually after experimental PIIS and POIS²⁵  had been established. The tissues were cut into small pieces and placed in 1 mL phosphate-buffered saline (PBS) + 2% fetal calf serum (FCS), 2 mg/mL collagenase type II, and 1 mg/mL DNAse Type I. Thereafter, the tissues were further minced with a no. 15 sterile surgical blade and then transferred from the plate to a 15-mL conical test tube. The tissue/cells remaining on the plate were washed with 1 mL PBS + 2% FCS and transferred to the same test tube, which was incubated in a shaker incubator for 20 minutes at 37°C and 200 rpm. We then added 20 μL 0.5 M ethylendiamintetraacetate (EDTA) and incubated the material for another 20 minutes at 37°C and 200 rpm. The test tube was filled to 12 mL with PBS + 2% FCS and peripheral blood (control) and centrifuged at 4°C at 400g for 12 minutes. The supernatant was removed, and cells were resuspended with 2 mL PBS + 2% FCS. The sample was then filtered with a 70-μm cell strainer, centrifuged at 4°C and 320g for 5 minutes, and then the supernatant was removed and cells were resuspended with 500 μL RPMI 1640 liquid and 10% thermal inactivation fetal bovine serum (FBS) in the tube. Cell density was adjusted to 1 × 10⁶/mL, and the test volume was 500 μL.

For the preparation of CD4⁺ T-cell lines from fresh peripheral blood (for control purposes, ie, from an uninflamed body compartment other than dental tissue), blood samples were collected in a heparin anticoagulant tube and centrifuged at 250g at 18°C to 22°C for 10 minutes. The upper plasma layer was then discarded, and a 1:2 dilution was made with 1% PBS. We carefully transferred the sample into a tube with a lymphocyte separation medium (TBDscience), measuring the same volume as the sample. Two layers were observed and centrifuged for 20 minutes at 400g. The supernatant was removed and aspirated with the lymphocyte layer, and 5 mL PBS was added. This sample was centrifuged at 250g for 10 minutes, and the supernatant was then discarded and the sample washed again. The cells were then resuspended with RPMI 1640 + 10% FBS. Finally, the cell density was adjusted to 1 × 10⁶/mL, and the test volume thus obtained was about 500 μL.

Cells were incubated in a cell stimulation cocktail (Catalog Number 00-4975, eBioscience) plus 1 μL protein transport inhibitors (Catalog Number 00-4975, eBioscience), maintained in 5% CO₂ at 37°C for 6 hours, and then 2 mL PBS was added, and cells were centrifuged for 6 minutes at 1500 rpm/min. Cells were then resuspended in 100 μL PBS, containing 2% FBS, and were labeled with 5 μL/CD4-APC (Catalog Number 17-5040, eBioscence) and 5 μL/CD25-FITC (Catalog Number 11-0250, eBioscence) during an incubation period of 30 minutes at 4°C. After this, resuspended cell samples were washed with PBS and 2% FBS and centrifuged for 5 minutes at 4°C (the upper plasma then being discarded). Thereafter, 500 μL permeabilization buffer (perm/wash buffer, Catalog Number 00-5523, eBioscence) was added for an incubation time of 40 minutes at 4°C and a 1-time washing step with buffer. The solution was then centrifuged for 7 minutes, and the supernatant was removed. Finally, samples were washed once with 1 mL perm/wash buffer, centrifuged, and analyzed by flow cytometry (Cytomics TM FC500, Beckman).

For the quantitative estimation of possible significant differences in bone resorption effects between the 2 experimental groups, standard paired t test analyses were applied. The level of significance was defined as P < .05. An independent professional statistician checked both the methods used and the computations. Because of the small number used in this exploratory study, the data are to be considered rough preliminary indications.

The cotton-ligature technique to induce POIS is illustrated by macroscopic images in Figure 3a and b: Figure 3a was taken at the time point of ligature placement, and Figure 3b was taken 3 months later when the inflammatory response had developed. The same ligature technique was applied at fully osseointegrated dental implants. Figure 3c illustrates by a CBCT picture an example of such an implant that successfully osseointegrated, the bony tissue typically being in tight contact with the implant surface structures. Three months after ligature placement (Figure 1), the CBCT pictures reveal (see Figure 3d as an example) the tight bone-implant contact to be destroyed, and that bone resorption had taken place, in particular around the mesio-distal aspects of the implants. However, at the tip of the implants (ie, at their distal ends), bony resorption activities were detected (Figure 3d). Very similar phenomena were observed around teeth (Figure 3d) where POIS was induced; bony resorption phenomena also were present at the lateral surfaces of the tooth root, although to a somewhat lesser degree, and at the tips of the roots. The illustrations chosen are typical and representative of the structural changes observed 3 months after ligature placement around implants and/or teeth. The morphometric measurements of the areas with bone loss along the lateral (ie, in the mesio-distal plane) and at the distal ends revealed (Figure 4) that the depth of the bone resorption extent (Figure 4a) was the same in both PIIS and POIS (ie, no significant differences were encountered). However, the variation of the resorption width varied considerably, being much higher in POIS, is illustrated nicely by the much higher coefficient of error (CE = 8.8%) than in PIIS (CE = 5.3%). Almost the reverse effects were found in the data relating to the mesio-distal width resorption: here, the variation of the results was much higher in the PIIS cases (CE = 6.3%) compared with the POIS cases (CE = 1.6%). In addition, in this case, we encountered a significantly higher bone resorption activity (paired t test; P < .05; Figure 4b) that was associated with the inflammatory process; whereas the resorption gap showed a width of 2.1 mm (mean value) in case of POIS, the mean width of the resorption gap in case of ligature-induced PIIS was 4 mm (Figure 4b).

The low-magnification light micrograph of Figure 5a illustrates a representative case of PIIS 3 months after ligature placement. Figure 5b and c show at higher magnification osteoclast-mediated bone resorption activities at multiple sites along the partially destroyed previously osseointegrated interface. The richness of inflammatory cell infiltration of the peri-implant tissue (predominantly lymphocytes) was clearly identifiable, and the loose and swollen characteristics of the intercellular spaces are obvious. Figure 5d, taken at low magnification, illustrates a representative case of POIS, and in Figure 5e and f, the richness in inflammatory cell infiltrates was also readily recognizable. However, here the tissue was much richer in collagen fiber bundles and appeared less loose and much less swollen (than in PIIS). The periodontal ligament with the typical richness in collagen fibers appeared somewhat loosened and infiltrated by inflammatory cells but still exhibited an ordered functional arrangement. In both cases (ie, in PIIS and POIS), the richness of inflammatory cell infiltration was obvious and appeared to be of the same degree, also with respect to the richness in blood vessels and their numerical area density. Figure 6a illustrates in the peri-implant PIIS connective tissue that the blood vessel density is mostly of a regular degree; in some areas, a somewhat higher degree of blood vessel density was encountered. Figure 6b shows the bony tissue area of PIIS in which the blood vessel area density appears significantly increased. In Figure 6a, the extensive infiltration by lymphocytes is readily recognizable and is dominating the intercellular tissue spaces. Plasma cells, mast cells, and macrophages were found to be only sparsely present, and foreign body giant cells could not be found; the same picture of inflammatory cell type populations was found in both PIIS and POIS. The high degree of bone resorption activities, affected by large numbers of osteoclasts, is illustrated in Figure 5b and c (and at higher magnification in Figure 6b). A very striking finding was the absence of osteogenic activities and of osteoblasts in this phase of the postoperative inflammatory course.

The percentages of Th1 and Th2 cells in the peripheral blood (nondental tissue control) were investigated by flow cytometry before and after the inflammation model was established around the teeth and the implants. Th1 and Th2 cells in peripheral blood within the CD4 subsets of cell populations showed no significant difference before and after the inflammatory model was established (Figure 7). Also, the relative changes of the Th1 and Th2 cell proportions between PIIS and POIS (Figure 8) did not show any significant differences and thus basically remained unchanged. A particular finding was, however, that the variance of the Th2 changes (in percent) was quite small in comparison to the others (Figure 8). Similar findings of different variances were obtained relating to the relative changes of these cellular subpopulations within the inflamed gingival tissues in PIIS and POIS compared with the normal (buccal) gingival tissues (Figure 9; ie, the variances being significantly larger in the cell populations of the diseased tissues compared with those of the normal control: buccal gingiva, pre-experimental time point).

In this exploratory investigation, we hypothesized that the inflammatory responses of experimentally induced PIIS and POIS in the dog model are different, in particular with respect to Th1 and Th2 cell involvement. In addition, we expected that secondary tissue changes such as bone resorption activities, the degree of local inflammation, and inflammation-induced neovascularization would also be different. These hypotheses were brought forward for experimentally induced PIIS and POIS, although several studies had indicated that the local responses to bacterial infection in clinical PIIS are, immunologically and biochemically, fairly similar in periodontal diseases, in particular in the human tissues.^19,49–51  A number of immune response studies of experimental PIIS describe the role of cytokines^42,52–55  but also point out the additional inflammatory role of the ligature itself, which may contribute to the model pathology.^10,16  However, the question if T-cell populations in experimental animal models for PIIS and POIS are actively involved has not been addressed, and if possible, whether T-cell involvement is different or the same in PIIS and POIS. We indeed found no significant differences in the patterns of Th1 and Th2 cell population changes in PIIS and POIS in this experimental dog model. Also, the presence of inflammatory cell infiltrates did not seem to vary significantly in amount and in composition of inflammatory cell population composition between the 2 different experimental groups in the dog model (PIIS vs POIS). Given the different tissue compartments present in PIIS and POIS (ie, peri-implant bone tissue vs periodontal collagenous tissue), this is a surprising finding. An observation that the degree of neovascularization in the inflammatory tissue in POIS appeared significantly higher than in PIIS may be related to the physiologically higher degree of blood vessel numerical densities in these connective tissues compared with bone tissue. However, because of the absence of other differences in cellular parameters, it appears to be of limited functional impact in the inflammatory processes. In particular, the secondary tissue changes (bone resorption, gap formation, etc) and effects of inflammation are quite similar in PIIS and POIS, except for the lateral width in the mesio-distal plane of bony resorption activities, which was found to be significantly higher in PIIS than in POIS. This difference may be associated with the somewhat protective effect of the periodontium respecting activation of osteolytic activities in the immediate tooth surrounding and the ligamentous type of structural collagenous anchoring (periodontal anchoring) in the adjacent bone tissue and thus may have a barrier membrane–like effect on the expansion of the inflammatory and resorptive processes into the surrounding bony spaces, as was previously concluded in similar studies.^50,56  However, we wish to emphasize again that, because of the small sample size used in this study, the statistical values presented are to be interpreted as preliminary indicators only, for which reason the study is defined as exploratory in nature.

On the other hand, there are the physiologically present osteoclast cell populations in the bony environment of PIIS that can be more readily and rapidly activated, and additional osteoclast cell populations that can be recruited. The differences in these bone tissue resorption patterns between PIIS and POIS do not seem to be related to the inflammatory cell populations involved because these were found to be practically identical in composition and numerical density in both groups. An interesting observation is the fact that at the time point chosen for this investigation, 3 months after surgical ligature placing, the signs of inflammatory chronification,^19,57  indicated by the typical lymphocyte cell populations in the inflamed areas, are quite similar in both POIS and PIIS. Whenever inflammatory responses occur in the body, T cells become involved and infiltrate these locations, in particular, when signs of long-lasting inflammations are operative, and thus, chronification of the processes is developing.⁵⁸  It is not surprising, however, that foreign body giant cells were not observed to be present (although the application of the ligature represents the introduction of a foreign body). These cells and macrophages are predominantly present in acute phases of inflammations relating to tissue reactions to foreign bodies⁵⁹  and not in chronic inflammatory processes.¹⁹ 

It is of particular interest that within the investigated time period, no signs of osteoblastic or osteogenic activities were encountered. This indicates that active inflammatory responses are still ongoing at that time point, and it also points to the chronification of the process. These data imply also the presence of ongoing inflammation-induced inhibiting activities of osteogenesis by the inflammation itself and by the associated release of high levels of antiosteogenic inhibitory cytokines (ie, proinflammatory cytokines).^46,60  Based on these findings, it may be speculated that it could make sense that a disease-modulatory therapy respecting the inflammatory PIIS activity (ie, a host-response modulatory drug therapy) may be useful and considered as a therapeutic option for such pathologic conditions. The therapeutic immunologic modulation of the host response is nowadays considered a novel promising treatment approach. To date, a host modulator drug therapy is, for example, a subantimicrobial dose of doxycycline, which inhibits host-derived matrix metalloproteinases that are responsible for soft and mineralized tissue degradation, possibly resulting in a reduced progression of periodontitis.⁶¹  Given that the POIS and PIIS host responses appear similar in nature and histopathology, such a drug-based treatment approach could thus also be considered for use in PIIS treatment concepts. Given the various limitations of the study (small number, preliminary data set, experimental model limitations, absence of direct relationship to the human pathology, animal model limitation; see Table 1 for summary), such suggestions are of limited value but nonetheless of interest given the analogous soft tissue and inflammatory pathology encountered in both groups and also in the literature and in the human pathology.

As we pointed out above, abundant osteoclast-induced bone resorption activities are present. Several studies with osteoprotegerin (OPG) have shown that activated T cells are able to promote osteoclast differentiation through production and expression of OPG.^62,63  These direct and indirect modes of T-cell involvement in periodontal bone resorption appear to be dependent on the extent of Th1-type T-cell recruitment in inflamed tissue.⁶⁴  However, in this study, there was no significant difference found in Th1 and Th2 after inflammatory gingivitis was established for PIIS and POIS.

There are indeed a number of different ways to induce POIS and PIIS as alternatives to the cotton-ligature technique: Aggregatibacter actinomycetemcomitans infection, an oral lavage model, a lipopolysaccharide injection model, and/or a calvarial model.¹⁷  Placement of ligatures around the teeth has been used in many different types of animals, ranging from rats to nonhuman primates, which then can lead to the accumulation of dental plaques and microulceration of the sulcular epithelium. This in turn facilitates the invasion of periodontal pathogens into the connective tissue.¹⁷  The experimental dog model used in this study clearly has its limitations. Although the cellular inflammatory infiltrates were found to be similar, as seems to be the case also in human tissues,¹⁹  the concept of introducing a ligature in this model as a foreign body requires this material to be considered a contributing factor to the histopathologic processes that in the clinical situation is not present. Moreover, the absence of a hygiene control in this study and the bacterial plaque formation situation may represent factors that contribute to a more drastic pathology than encountered in the clinical reality. Another aspect of model limitation may be the fact that the PIIS and POIS pathologies were generated in adjacent teeth in this model. One cannot exclude any proximity effects of 1 site to the other. However, clinical experience tells us that single tooth POIS does indeed occur in patients without affecting the tissues of the neighbor tooth,⁶⁵  implying that a spatial restriction of the pathology (and also a treatment success) to a single tooth indeed is possible. This may relate in particular to local diffusible factors such as cytokines and chemokines but less so to the cellular infiltration processes. Given these aspects and the fact of basic differences in the oral biology, tissue composition, and species-specific reactivity between and among animals and the human patient, data interpretation thus cannot be applied directly to the human situation; however, the model still may have its usefulness in testing novel therapeutic concepts.

In this exploratory study, we established 2 types of animal models: PIIS and POIS. We found that the bone resorption effects in PIIS are more extensive than those of POIS, which is in agreement with the findings of a previous study²⁴ ; however, in this study, we compared the depth and width of the bone resorption. The depths of the bone resorption showed no significant difference between the 2 groups, whereas the widths of the bone resorption did show significant differences, very likely because the periodontitis lesion is somewhat walled off by an intact supracrestal connective tissue fiber compartment.⁶⁶  Thus, PIIS lesion progression occurred on a facilitated basis, probably because of the absence of a healthy connective tissue fiber compartment, walling off the lesion from the alveolar bone. The basic inflammatory processes were found to be quite similar in POIS and PIIS in this dog model. The data are, however, to be considered preliminary in nature because of the small sample sizes investigated. They also indicate that further model refinements of in vivo testing approaches are needed to be developed to better simulate the human pathologic conditions for relevant developments of new therapeutic strategies. The basic dilemma of incongruency between human and animal pathologies will, however, remain, and will continue to require careful data interpretation.

This preliminary study did not confirm the hypothesis that experimental POIS and PIIS in a dog model are different. Although the data interpretation is of limited scope because of a small number chosen in this study and because of the inherent experimental model limitations, the indications are that the inflammatory and histopathologic responses are essentially the same. Differences are encountered only in the pattern and extent of bone resorption activities, and these corelate to clinically known differences to a limited extent. The POIS and PIIS method of experimental induction (ligature methods) needs to be refined/modified for improved simulation of the human pathology in an experimental in vivo model for a more focused approach when experimental explorations of novel and more specific therapeutic principles are envisaged.

Abbreviations

CBCT:

cone-beam computed tomography

CEJ:

cement-enamel junction

PIIS:

peri-implantitis

POIS:

periodontitis

This study was supported by the Foundation of the Department of Osteoporosis, Inselspital Bern University Hospital. The funding source had no influence on the writing and the decision of submission of this article. The authors thank Long-Feng Lu, Chinese Academy of Sciences, Wuhan, China, for statistical advice and for checking all the statistical procedures used for correctness.

The authors declare no conflicts of interest.