To achieve re-osseointegration on implant surfaces exposed to peri-implant infections, treatment should re-establish biocompatibility. The aim of this study was to test whether air powder abrasive treatment (APA) using osteoconductive powders can, in addition to cleaning, increase the biocompatibility of the contaminated implant surface. Ninety-six in vitro Ca-precipitated, organic film layer–coated sandblasted and acid-etched titanium discs were treated by APA using erythritol, hydroxylapatite (HA), and biocalcium phosphate (BioCaP) powders (n = 16 per group). Six treatment modalities were created (HA or erythritol cleaning with/without BioCaP coating). MC3T3-E1cells were seeded on discs, and cell attachment, viability, proliferation, and differentiation were evaluated. Pristine discs were used as control (control 1). Contaminated and nontreated discs were used as control (control 2). The cells were stretched and attached in all test groups. The cell viability and proliferation (DNA amount) in all test groups were significantly higher than in the pristine and contaminated disc groups. There was no significant difference between the test groups. The differentiation (alkaline phosphatase activity) of the cells on treated discs was significantly higher than on the contaminated discs but lower than in the pristine group. The cell viability in control 2 was significantly lower than the control 1. The APA with osteoconductive powder on contaminated titanium surfaces promoted the cell viability, proliferation, and differentiation of the MC3T3-E1 cells. The biocompatibility of the surface was higher than that of the contaminated discs. The tested aspects of cell response, with the exception of differentiation, reached to the level of the pristine surface. The in vitro results showed that APA with osteoconductive powders could be a promising method for implant surface treatment.
Peri-implantitis is an infectious condition of the tissues around osseointegrated implants that is characterized by loss of supporting bone and clinical signs of inflammation (bleeding and/or suppuration on probing).1
The main cause of the peri-implantitis is often seen as the infection of the peri-implant tissues and the microbial colonization of dental implants.2 As a result of the microbial contamination of the implant surface, the chemical composition of the titanium oxide layer on the surface changes. This alters the surface energy and has negative effects on the adhesion capacity of osseous cells and bone tissue to titanium.3 Because bone conduction is not only dependent on conditions for bone repair (biological factors such as growth factors, blood supply), but also on the biomaterial used and response toward it, the materials that are too toxic to allow osteoconduction will not be osseointegrated either.4 In the case of peri-implantitis, the infection exposed implant has a poor osteoinduction capacity because of the altered surface energy. As a result of this phenomenon, the re-osseointegration is not achieved either. Therefore, a successful treatment of peri-implantitis must focus on not only cleaning the implant and the infection in the tissue but also the re-establishment of the biocompatibility and osteoconductive properties of the titanium surface. Only in this case can the condition of the exposed implant surface be restored to such a level bone to implant contact can be re-established by re-osseointegration.
The surface treatment methods that are currently used aim to improve the impaired biocompatibility of the surface by only removing the biofilm and cleaning the implant surface. However, several studies focusing on the capacity of different treatment methods in re-establishing the biocompatibility of contaminated titanium surfaces5–9 show that the biocompatibility of the surface depends on the instrumentation used to remove the biofilm and not only on the cleaning efficiency. In most of the studies, the cell response toward the treated surface is not as good as toward the sterile nontreated surface5,6 ; however, certain treatments show higher cell attachment and cell viability on treated surfaces compared to others.10 A review paper has concluded that the applied instrumentation method may have a selective influence on the attachment of different cells.11 The methods that have limited cleaning efficacy and cause a deposition of debris from the instrument on the surface fail to restore biocompatibility. The surfaces that have higher biocompatibility are the ones that are cleaned without surface damage, leaving residues from the instrumentation on the implant surface, and have lower percentages of carbon concentration on the surface after cleaning.11,12 Being one of the mechanical methods, although it is not perfect, the air powder abrasive treatment is reported as the best-performing treatment with regard to this aspect. It has been proven to be very effective on removing the biofilm from the titanium surface in several in vitro and in vivo studies.12–14 Apart from the cleaning, it has no damaging effect on the surface topography and is successful in reducing the carbon concentration of the chemical content of the surface oxide layer.15 However, these results are also dependent on the powder type used. The powder not only impacts the cleaning effect but also modifies the chemical content of the surface by leaving particles attached on the surface after the treatment.16
Based on this fact, we think that using osteoconductive powders like calcium phosphate (CaP) might improve the surface modification aspect of air powder abrasive treatment. CaP is a biocompatible and osteoconductive material that is widely applied for different applications, such as bone substitute material. Its composition is similar to bone mineral. When used as bone substitute material, CaP is able to achieve an early and functional bone apposition on dental implants.17–19
MC3T3-E1 cells are preosteoblastic cells that have been used to evaluate the osteoblast-like cell behavior toward modified titanium surfaces.20–23 The cell attachment, morphology, viability, proliferation, and osteogenic differentiation are indicators of the hard tissue compatibility of the tested surface.
The hypothesis of the present study was that the use of osteoconductive powders with air powder abrasive treatment in peri-implantitis surface treatment would improve the reduced biocompatibility of the contaminated titanium implant surface.
The aim of this study was to test whether the air powder abrasive treatment using biocompatible powders can re-establish the osteoconductivity of the contaminated titanium surface. This was done by testing for cell attachment, cell proliferation, and cell viability of MC3T3-E1 cells toward contaminated and treated surfaces of titanium discs.
Materials and Methods
Titanium discs of 10 mm in diameter (Ningbo Cibei Medical Treatment Appliance Co, Zhejiang, China) were used to prepare the specimens. The surfaces of the discs were sandblasted and acid etched (SLA), showing the same surface properties as SLA surface of oral implants.
Ca-precipitated organic film layer formation on the discs
To create a strongly attaching film layer with organic and nonorganic elements, the following model was used. Sterile titanium discs (SLA surface, 10 mm in diameter) were incubated in α-minimum essential medium (α-MEM; Gibco, Grand Island, NY) supplemented with 20% fetal bovine serum (FBS; Gibco) at 37°C with 5% carbon dioxide and 95% humidity for 24 hours to create a pellicle on the titanium surface and facilitate the biofilm formation. Subsequently, every disc was immersed in 1 mL fresh unstimulated saliva from a healthy donor and incubated for another 24 hours. The following day, fresh α-MEM supplemented with 20% FBS was added on each disc and incubated for 72 hours. The medium was refreshed, and the discs were incubated again for 72 hours. The discs were transferred into new well plates containing 2 mL saturated (0.02 M) Ca(OH)2 solution and incubated for 72 hours to create Ca precipitation on the discs.
Air powder abrasive device
All discs were treated by air powder abrasive treatment. An EMS Air flow device (AIR-FLOW master and AIR-FLOW Perio, EMS, Nyon, Switzerland) was used with a standard air-flow nozzle. The EMS AirFlow Chamber and Perio Plus Chamber were used for the cleaning and coating step, respectively. The air pressure for all steps was 240 000 Pa (2.4 Bar). The water flow was 42 mL/min for the cleaning, 2 mL/min for the first coating, and 15 mL/min for the second coating.
The pressure inside the chamber, the consumed powder amount, and the water flow were measured during the treatment using a manometer (DPI 802 P GE Druck, Manchester, UK) and a balance (Mettler Toledo PR 8002, Columbus, Ohio).
The cleaning step was performed using either pure Erythritol (AIRFLOW PLUS powder, EMS) or a mixture of erythritol and hydroxylapatite (HA; EMS). Erythritol powder is a commercially available powder with a mean particle size of approximately 14 μm. HA powder has microparticles with a mean of 5 μm that are made up of nanoparticles. The HA-erythritol mixture is prepared with the ratio of 4% HA and 96% erythritol powder according to their weight.
The coating step was performed using a biomimetic CaP (BioCaP)-erythritol mixture. BioCaP powder has a particle size between 15 and 75 μm. This powder is produced under physiologic conditions and developed at ACTA (Academisch Centrum Tandheelkunde, Amsterdam, The Netherlands).24 The 2 powders were mixed in the ratio of 4% BioCaP and 96% erythritol according to their weight.
Treatment and test groups
The treatment consisted of 2 subsequent steps. The first step was aimed at cleaning and the second step at modification of the surface by BioCaP coating.
The test groups were as follows: group 1, HA cleaning + BioCaP coating 1; group 2, HA cleaning + BioCaP coating 2; group 3, erythritol cleaning + BioCaP coating 1; group 4, erythritol cleaning + BioCaP coating 2; group 5, HA cleaning + no coating; group 6, erythritol cleaning + no coating; control 1, no biofilm coating, no treatment; control 2, biofilm coating but no treatment; control 3, plastic well plate (Table 1).
The treatment was applied with a fixed nozzle angle of 60° at a distance of 4 mm to the disc. The disc was moved circularly parallel to the ground to let the air spray uniformly and reach the whole surface. The cleaning and coating steps continued each for 30 seconds.
For the treatment evaluation, we used 2 samples per test group for cell attachment, 6 samples per test group and 4 samples per control for cell viability, and 4 samples per test group and 2 samples per control for DNA/alkaline phosphatase (ALP) tests.
After the treatment, the discs were autoclaved and placed into 48-well plates. MC3T3 osteoblast-like cells were used for the in vitro evaluation. Cells in complete α-MEM supplemented with 10% FBS, 100 U/mL penicillin, 100 μg/mL streptomycin sulfate, and 0.5% fungizone were incubated at 37°C with 5% carbondioxide and 95% humidity.
Cell attachment and SEM observation
A total of 1 × 105 cells were seeded on the titanium discs in 1 mL osteogenic medium (complete α-MEM supplemented with 10% FBS, 100 U/mL penicillin, 100 μg/mL streptomycin sulfate, 0.5% fungizone, 1% β-GP, and 1% vitamin C) and incubated at 37°C with 5% carbondioxide and 95% humidity for 2 hours. Then, the cells were fixed, and samples were prepared for scanning electron microscope (SEM) observation. High magnitude SEM photos were taken (Zeiss Sigma Field Emission, Gemini, Cambridge, UK).
After 4 and 6 days of incubation, the Alamar blue (AB) fluorescent assay (Life Technologies, Grand Island, NY) was used to measure cell viability.25 AB measures the reductive activity inside living cells. To measure the viability of disc-adherent cells, 10% AB reagent was added inside the medium on each well. Cells were incubated with reagent for 4 hours in culture conditions, and then media samples were collected in an opaque 96-well plate for fluorescent detection (excitation = 530 nm, emission = 590 nm) using a BioTek Synergy microplate reader (BioTek Instruments, Winooski, Vt). A viability assay was conducted using 3 samples per group and 2 culture replicates per sample.
Total DNA content and ALP activity: Cell proliferation
Total DNA content and ALP activity were measured 6 days after seeding the 1 × 105 cells in osteogenic medium on the titanium discs. After 6 days of incubation, the medium was vacuumed, and the discs were washed with phosphate-buffered saline (PBS) buffer (Gibco) and placed in new well plates; 500 μL lysis buffer was added to each well. DNA was extracted by 3 cycles of freezing and thawing. Total DNA content was quantified using the Cyquant Cell Proliferation Assay (Molecular Probes, Eugene, Ore) according to the manufacturer's protocol. ALP activity was measured according to the method described by Lowry.26
The analysis of the different test groups was carried out by using a multifactor analysis of variance (ANOVA) statistical design (Excel program designed by J. Lemaître) with 2 cleaning powders, 3 coating methods, and 3 repetitions (the different samples per test group are analyzed explicitly in a group named “Repetition”). As the factorial ANOVA assumes that the dependent variable approximates a multivariate normal distribution, the assumption was checked graphically using a Q-Q-Plot. As the factorial ANOVA assumes homoscedasticity of error variances, which means that the error variances of all data points of the dependent variable are equal or homogenous throughout the sample, this was tested with the Levene's test with a significance level of 0.05.
Only the factors with P < .05 value were considered significant for the statistical model.
Considering that the control groups had less repetitions and that the differences between groups are not high, the comparative study of all the treated data with regard to the controls is handled with a Welch's t-test. The significance level was P < .05.
The SEM analysis of the titanium discs showed the morphology of the cells attached on the titanium surface. The cell morphology displayed clear differences between the sterile discs (control 1) and treated discs (test groups). Approximately half of the cells on the sterile discs (control 1) were round and not well attached on the surface; the other half were stretched and attached. On the treated discs, most of the cells were stretched. These cells were near each other and attempted to communicate with the help of focal adhesion sites. The stress fibers were a sign that they were preparing to migrate. The untreated discs (control 2) showed mostly flat and attached cells. However, these cells did not show many focal adhesion cites and stress fibers like the ones seen in the test groups. All the test groups showed similar morphology; however, test groups 1 and 2 showed higher numbers of stretched cells. The groups without the coating step showed the least number of stretched cells on the SEM photographs (Figure 1).
There were no signs of fungal or bacterial contamination of the well chambers during the experiment. The cell viability was assessed on days 4 and 6 by an AB assay. For all experiments, cells directly seeded on the well plate floor were used as a positive control.
On day 4, the mean of all treatment groups showed significantly higher cell viability than both sterile (C1) and untreated discs (C2) (Figure 2). The cell viability counts of the treatment groups that had HA cleaning (T1, T2, T5) were approximately 4 times higher than the sterile control, whereas the groups that had erythrithol cleaning (T3, T4) were approximately 5 times higher than the sterile control. Erythrithol cleaning without coating (T6) showed the most pronounced difference compared with the sterile control—that is, 6 times higher (Figure 2).
For the statistical analysis of the different test groups, the cleaning step was the major factor that caused a statistically significant difference for cell viability on day 4: the groups with erythrithol cleaning showed higher cell viability compared with the groups with HA cleaning. The coating step only showed minimal influence on day 4 results (Figure 2).
On day 6, like day 4, the mean cell viability of all treatment groups was significantly higher than both sterile (C1) and untreated discs (C2). The cell viability counts of the treated discs were approximately 2.5 to 3 times higher than the sterile control (Figure 2). However, on day 6, the cleaning step did not have a significant influence on cell viability compared with day 4. This time, the BioCaP coating step had influence and caused a significant difference compared with the groups without the BioCaP coating (Figure 2).
Contrary to our expectations, the untreated discs (C2) showed higher cell viability than the sterile discs after both 4 and 6 days; this could possibly mean that the nonremoved but autoclaved dead bacteria layer increased the cell viability of the discs.
DNA amount: Cell proliferation
Cell proliferation was checked with a CyQuant assay on day 6, and there was no significant difference between the test groups. However, the mean of all treatment groups showed significantly higher DNA amount than all control groups (Figure 3). Similar to the cell viability, the untreated discs (C2) showed 2 times higher cell proliferation than the sterile discs.
ALP activity: Cell differentiation
The ALP activity was checked on day 6 and normalized to DNA results. There were no significant differences among the treatment groups. The mean ALP activity of all treatment groups had significantly higher results than on the untreated discs; however, it was not as high as on the sterile discs (Figure 3). The ALP activity on the untreated discs (C2) was significantly lower than the sterile discs (C1) unlike the DNA amount, meaning that, although the dead bacteria layer increased the proliferation, it inhibits the cell differentiation, as shown by the ALP activity.
Re-osseointegration is a process that takes place based on osteoinduction and osteoconduction. Because bone conduction is dependent on the biomaterial used and the response toward it, the implant surface biocompatibility influences this process crucially.4 Therefore, implant surface treatment has a very important role to achieve re-osseointegration.
This study aimed to test the biocompatibility and restoring capacity of a new implant surface treatment approach. Air powder abrasive treatment was applied using biocompatible and osteoconductive powders in 2 steps. The first step aimed to remove the biofilm, and the second step aimed to create a CaP coating. To determine the effect of the cleaning step alone and cleaning and coating steps together, 6 test groups were created. In addition, different water settings were tested because the water flow influences the amount of the powder particles that attach on the titanium surface. Therefore, we applied the coating step either in high or low water flow and tested the biocompatibility of the resulting surface. Biofilm-covered untreated discs were used as negative controls, and pristine and untreated discs without biofilm coating were used as positive controls.
The cell response tests were carried out using MC3T3-E1 cells, which is a nontransformed cell line established from newborn mouse calvaria. These cells exhibit an osteoblastic phenotype as evidenced by the expression of ALP activity,27 the synthesis of extracellular matrix (ECM) components such as osteocalcin and type 1 collagen28 and their ability to mineralize the ECM.
The first parameters assessed were the cell attachment and morphology. The cells were attached on the surface in all groups; however, the discs in test groups 1 and 2 showed higher numbers of stretched cells. These 2 groups had both HA and BioCaP in cleaning and coating steps; therefore, these discs had the highest Ca content. Additionally, the coating step created a rough CaP layer, and the cells attached on this layer with the focal adhesion sites and communicated with each other. The groups without the coating step showed the least number of stretched cells on SEM photos. Therefore, we speculate that the CaP coating step had a positive effect on the cell attachment. Another interesting finding was the morphology of the cells in the control 2 group. Although contamination of the surface impairs the biocompatibility of the titanium surface via the bacteria exotoxins and increases the percentage of carbon, surprisingly in our study, there were many flat cells that were well attached to this contaminated surface. We consider that the reason of this unexpected finding could be that the protein content of the sterilized but unremoved biofilm layer acted as an extracellular matrix, and the cells attached better on this layer. The only group that showed some amount of round and not well-attached cells was the pristine, noncontaminated control group.
Second, the cell viability after 4 and 6 days was measured. According to the results, the treatment, independent of the settings and cleaning and coating combinations, enhances the cell viability compared with the untreated discs and even sterile discs. It was expected that the treatment would cause a higher biocompatibility than the untreated discs simply because the biofilm was mechanically cleaned. However, the treatment increased the biocompatibility more than the sterile discs. This was not a common finding in previous studies. John et al29 performed a similar study with air powder abrasive treatment based on glycine and tricalcium phosphate on titanium surfaces. They showed significantly lower viabilities of all treated titanium test groups than the control groups with native specimens without biofilm formation. They concluded that there was no indication for re-establishment of the biocompatibility after biofilm formation and removal via the tested devices.29 Other studies testing several treatment modalities such as air powder abrasive treatment with sodium bicarbonate and amino acid glycine,6 ultrasonic scalers,8 or titanium brush5 showed results in accordance with the study by John et al29 All treatment groups showed less viability than the sterile, noncontaminated control group. Kreisler et al30 performed the only study that reported the cell growth on treated surfaces not significantly different from that on sterile specimens. They tested the biocompatibility of the contaminated, laser treated and air powder abrasive–treated titanium surfaces. The cell growth on treated surfaces was significantly higher than contaminated and nontreated specimens and not significantly different from the sterile specimens.30 When the same treatment methods were applied on noncontaminated titanium surfaces, without any biofilm formation beforehand, the surface viability results were different. Toma et al12 treated pristine, sterile titanium discs with an air powder abrasive device using aminoacide-glycine or titanium brush or applied implantoplasty and tested the cell viability following the treatments. The air powder abrasive group showed higher cell viability than all treatment methods and the pristine surface. The fact that the viability on the treated discs was higher than the untreated discs shows that the treatment itself does not deteriorate the biocompatibility of the surface. The air powder abrasive treatment improved the biocompatibility of the surfaces both on contaminated and sterile surface application. However, when applied on the contaminated discs, it does not reach to the level of the pristine titanium. This could be due to the strong cytotoxic effect of the virulence factors of the bacteria, such as lipopolysaccharides and proteinases on the host cells.31 Therefore, most of the treatment modalities could not eliminate the adverse effect of the bacteria cell components on the cell growth even if they clean the surface. However, in this study, APA treatment not only compensated but also improved the viability of the contaminated discs (exceeded the sterile control). This may be due to the effect of the HA and BioCaP powder particles attached to the titanium surface. These osteoconductive particles may have compensated the adverse effects of the previous contamination.
The viability levels increased from day 4 to day 6 in all groups except group 6. However, this increase in the test groups was slower compared to the control. This means that the effect of the treatment appeared around 4 days and this effect decreased from day 4 to day 6. The cell viability on day 4 was mostly influenced by the type of cleaning and not the coating step. However, the coating step had an effect on day 6. There was a more pronounced increase with the BioCaP-treated discs (the groups that have a coating step) for cell viability on day 6. However, the effect was not seen on day 4. This can be interpreted by the late-appearing effect of the BioCaP on the cell viability. The HA cleaning group without a coating step showed a minimum increase in cell viability from day 4 to day 6. The cell viability of the only group that had neither HA nor BioCaP even decreased. Therefore, this shows that any type of CaP—either HA or BioCaP—has an influence on later cell viability.
Another interesting finding was that biofilm coated untreated discs had higher cell viability counts than the pristine non–biofilm-coated discs. Although the biofilm was not mechanically cleaned from these discs, all the discs were autoclaved prior to the cell cultures. Therefore, the dead bacteria could be seen in the SEM pictures as a layer covering the titanium surface (Figure 4). As we have already mentioned for cell attachment, we speculate that the proteins of the bacteria acted as an ECM. The cells could easily attach on this layer, as shown on the SEM photographs, and they could also proliferate, as shown with the viability assay counts.
In the present study, the cell proliferation and differentiation levels of the cells were also checked by the DNA and ALP levels. The DNA results were in line with the viability results, whereas the ALP results were different. The treatment improved the differentiation of the cells (shown by ALP levels) on treated discs compared to the untreated discs (except treatment 3); however, it does not reach the differentiation level of the pristine, non-biofilm-coated discs. Because the biofilm leads to changes in composition of the superficial titanium oxide layer, this can alter the surface energy and impair the cell attachment.3,32 Although the treatment compensated the adverse effects of the contamination to a certain level, the cells attaching to this treated surface did not differentiate at the level of the cells on the pristine surface. The negative control group results also show the impairing effect of the biofilm on the differentiation level of the cells. As we mentioned before, the cells surprisingly attached and proliferated well on this contaminated surface, but they did not differentiate as well as on the sterile or treated discs, as shown by the low ALP levels. We speculate that although the dead bacteria layer increased the cell viability and proliferation, it inhibited the differentiation compared to the pristine surface. In brief, the differentiation levels of the cells were negatively influenced by contamination.
Our treatment increased the impaired differentiation level of the cells but could not totally restore the surface and reach the level of the pristine surface.
The only exception to the above-mentioned phenomenon was treatment 3. This group showed very high cell viability but very low ALP results—lower than even untreated contaminated discs. The fact that the cells were very viable, but they do not actively go into the differentiation phase may be due to the influence of intense CaP coating on the surface. However, on the other treatment modules, the coating step had a beneficial effect on late cell viability results and, although not significant, a positive effect on cell differentiation.
While drawing conclusions out of this study, it needs to be kept in mind that in vitro surface biocompatibility is one of the several factors influencing the re-osseointegration process. Therefore, this study should be followed by in vivo clinical studies.
In conclusion, under the limitations of this study, our treatment was shown to improve the biocompatibility of the surface by advancing the viability, the proliferation, and differentiation of the cells. The treatment was not cytotoxic toward cells. The adverse effects of contamination on differentiation was compensated, whereas the adverse effect of contamination on viability and proliferation was not only compensated but improved more than sterile discs.
Marcel Donnet is a scientist working in the dental research group of Electro Medical Systems Company. The EMS air flow device is used in this study. We thank Dr. R.A.M. Exterkate from the Department of Preventive Dentistry at ACTA for help in the microbiology laboratory and R.I. Oke for support with the production of the figures.
The authors state that there are no conflicts of interest.