Background
Peri-implantitis is a biofilm-mediated progressive inflammatory disease in the tissues surrounding the dental implant, which ultimately may lead to its loss [
1]. To date there is no consensus on a treatment protocol for peri-implant diseases. Therapies researched have mostly been modifications of periodontitis treatment modalities [
2].
In 1990, Lozada and co-workers presented a case report where peri-implantitis was treated by open flap debridement in addition to recontouring the exposed implant surface with high-speed diamond and aluminum oxide burs [
3], a procedure now referred to as implantoplasty. The advantage of a smooth versus a rough surface is facilitated oral hygiene and a reduction in bacterial colony forming units [
4].
A previous clinical trial compared open flap debridement with or without implantoplasty. No change in radiographic bone loss from baseline to the 3-year follow-up was observed in the test group, whereas a mean marginal bone loss of 1,44 mm was found in the control group [
5]. Another clinical trial has been performed as well, albeit not comparing groups with and without implantoplasty. Schwarz and co-workers did implantoplasty on 38 patients as part of a treatment with and without the use of Er-YAG laser [
6]. These limited studies have presented promising clinical outcomes following implantoplasty.
Rimondini et al. investigated in vitro differences in the topographical surface roughness parameters between different implantoplasty bur protocols. All bur sequences tested rendered surfaces that were smoother than the plasma-spray-coated control implant, whereas no significant differences were found between the different bur sequences and the machined control [
7]. More recent in vitro studies have focused on the optimum bur sequence [
8‐
11], the heat generation following titanium polishing [
11‐
13], and biocompatibility [
14,
15]. Fracture resistance has also been the focus of in vitro studies [
16], but according to a recent systematic review no fractures following implantoplasty have been reported in the literature [
17].
A possible advantage of implantoplasty in addition to facilitated oral hygiene is a potential improvement of the soft tissue adaptation to the dental implant. An ideal implant surface should impede bacteria and biofilm growth and adhesion while at the same time allow rapid connective tissue attachment. A significant impact of the surface topography on connective tissue attachment has been demonstrated [
18]. Previous studies have shown that human gingival fibroblasts (HGF) spread more readily on smooth as compared to rough surfaces, and that the connective tissue adhesion is affected by surface properties [
19‐
23]. More knowledge of the mechanisms involved in the re-establishment of a soft tissue seal, of which fibroblasts play a key role, subsequent to implantoplasty treatment, is warranted. Therefore it is of interest to explore how the surface modifications made by clinicians during implantoplasty may affect HGFs.
The aim of the present in vitro study was to characterize the substrate topographies following experimental implantoplasty and to examine the fibroblast growth, attachment, morphology and cytokine secretion following culture on the various titanium substrates. The null hypothesis was that surface modifications by experimental implantoplasty have no effect on fibroblast growth, attachment, morphology or cytokine secretion.
Discussion
The initial growth of fibroblasts demonstrate a weak to moderate negative correlation to the surface roughness (Sa) following a selection of experimental implantoplasty strategies. All CB sequences rendered smoother surfaces than the DB sequences, and the additional use of BG resulted in smoother surfaces than Ark when combined with both CB and the DB sequences. The different implantoplasty bur sequences affected the elemental composition of the titanium surfaces, but when comparing the secretion of IL-6, VEGF, MCP1, MCP3 and IP-10 and fibronectin levels, the rough control (SLA) in general demonstrated higher levels whereas only small differences were observed between the implantoplasty test groups.
That fibroblasts respond differently according to substrate roughness is already known [
17,
18,
27]. However, this has only been shown on surfaces prepared in laboratories with delicate equipment and procedures impossible to replicate intraorally in patients. In case of peri-implantitis, surface alterations of rough implants may be desirable to facilitate hygiene measures, but potentially also to improve the soft tissue adaptation. This study is the first to demonstrate that chairside treatment with the use of only a few bur sequences has the capacity to influence subsequent in vitro fibroblast growth and adhesion. This indicates that implantoplasty treatment outcomes may affect soft tissue healing, adaptation and homeostasis, and not only the ease of microbial disruption in oral hygiene.
Experimental implantoplasty procedures including BG rendered the lowest S
a values, which is in agreement with Ramel and co-workers. Although they analyzed cylindrical dental implants with a two-dimensional stylus profilometer, the order of surface roughness for BG, Ark and DB as measured by R
a is in accordance with the present study [
9]. Bollen et al. suggested that bacterial colonization is not affected as long as the substrate roughness is below Ra 0.2 μm [
28]. In the present study only the POL control group had a S
a value below this threshold, which is in agreement with previous studies [
9,
10,
13]. To the best of the authors’ knowledge, only Costa-Berengeuer and co-workers have reported S
a values of less than 0.2 μm by the use of mere chairside bur-sequences [
16]. Potential explanations for these conflicting findings may be that Costa-Berenguer and co-workers used a high-speed handpiece and changed burs for every implant.
The DB-treated and SLA coins demonstrated very different fibroblast growth. Despite clear discrepancies in both the profilometer analysis and the SEM images, the Sa, Sz and Sq values were similar for coins in the DB and SLA group. This questions the validity of the use of these roughness parameters alone to determine surface roughness and clinical applicability of implantoplasty. One may hypothesize whether other surface roughness parameters or combinations of parameters would be more suitable for use in this context. In the present study the parameter Spk seemed to better differentiate SLA and DB. Spk represents the mean height of peaks above the core surface, and a large Spk value indicates a surface of high peaks providing a small initial contact area, which may be an explanation for the poor HGF growth in the SLA group.
The numerous dark spots that covered the surface of the CB-treated coins were not visible to the same extent in the other groups. One may hypothesize these dark-spots are debris following the CB sequence. The highest percentage of Si was observed on the surface of, and in the debris from, coins treated with BG. This demonstrates that BG burs leave behind more Si than the other burs, which is not surprising as BG are silicone burs. However, it also suggests not all silicon is lost as debris but some may be found on the implant surface.
Higher numbers of fibroblasts were found in groups with lower surface roughness (S
a value) for both diamond and carbide sequences. However, the POL control group with the lowest S
a value did not have significantly more fibroblasts as compared to any of the test groups at any time point. This may indicate that fibroblast growth and adhesion was not only affected by the surface S
a value in the present study. The enhanced growth on the smoothest surfaces observed in the present study is in agreement with the findings from Könönen et al. who compared fibroblast proliferation on three different titanium substrates. They also found that the fibroblasts cultured on the roughest surface were round and flat and had aberrant morphology after 3 days. Other previous studies have also reported higher viability and proliferation on smoother titanium surfaces [
20,
29,
30].
Studies have suggested smooth or finely grooved titanium substrates may be optimal for soft tissue adaptation due to its support of integrin-receptor clustering into focal and ECM contacts [
26]. One of the main functions of focal adhesion proteins is to promote cell-attachment to the extracellular matrix [
31]. These proteins are also important for cell-motility, normal cell function and interaction with the environment [
32,
33]. Fibronectin is a major structural glycoprotein that contributes attachment and spreading of fibroblasts [
34]. The distribution of fibronectin was investigated in the present study, but no overall trend was observed according to the different implantoplasty surface treatments at day 3 or 6. At day 30 however, the fibronectin level was higher in the SLA group. However, one must keep in mind that at day 30 very few cells were present in the SLA group whereas the fibronectin remained, which at this time point greatly affected the results presented relative to the SLA group.
Vinculin is a cytoskeletal protein involved in formation of focal adhesion [
35], and for this reason we aimed to quantitatively and qualitatively analyse it. Previous studies have indicated conflicting results with respect to HGFs’ expression of vinculin [
20,
36,
37].
Since the effect of the experimental implantoplasty treatment on the fibroblast growth was limited beyond day 6 in the present study, the analyses of cytokine secretion to the cell medium was performed at the two earliest time points only. Also the DB and CB sequences were left out of the Luminex analysis since these rendered the roughest surfaces of the experimental implantoplasty treatments, and would therefore not be considered in a clinical setting. The different cell densities were used to facilitate the chosen analyses. The cell seeding density of ~ 70 cells/coin were used for growth analysis, whereas the density of ~ 350 cells/coin were used for growth analysis and Luminex analysis. Wells seeded with the lower cell density were used for cell growth analysis in order to avoid early confluence due to rapid cell growth and to characterize the morphology of single isolated cells. The higher cell density was used in wells included in the Luminex assay to increase the concentration of cytokines secreted to the cell medium. One may speculate to what extent the various experimental implantoplasty substrates contribute different biological responses. For example, the concentration of IP-10 was higher in the cell medium from HGFs cultured on CB + Ark as compared to CB + BG at day 3. Whether such findings have any clinical relevance must be addressed in in vivo studies and clinical research. In the Luminex assay, a limited set of factors known to be expressed and secreted by fibroblasts that have potential stimulatory and/or inhibitory effects on surrounding cells and soft tissue in vivo, and/or potential implications in bone metabolism, was chosen.
Fibroblast adhesion and growth is only one of few events taking place following implantoplasty-treatment. The epithelial and in vivo soft tissue adaptation has not been addressed in this study but play an important role. Implantoplasty is above all performed to counter microbial challenges, and its impact on preventing bacterial recolonization and facilitating removal of bacterial colonization is considered pivotal to healing and homeostasis of peri-implant health following implantoplasty treatment in response to peri-implantitis challenges. So far in vitro studies on implantoplasty have focused on surface roughness [
8‐
10], heat generation [
11‐
13], and fracture resistance [
14,
16,
17]. This study provides some new insights on the soft tissue component following experimental implantoplasty. The establishment of a healthy soft tissue adaptation to the implant surface may be an important part of implantoplasty. Obtaining the smoothest surface possible may hence not be the ultimate goal of implantoplasty, if the soft tissue adaptation can be improved without deterioration of bacterial considerations. A number of studies on implantoplasty have been published in previous years, but it remains as a controversial therapy. There is limited scientific evidence to support an effect on the course of peri-implant diseases. Furthermore, the procedure leads to the release of titanium debris in vast quantities to the peri-implant tissues, which may have adverse biological effects [
38]. For this reason, only the supracrestal parts of the implant exposed following bone loss due to peri-implantitis, or as a result of mucosal recessions should be carefully considered for implantoplasty therapy.
This study has noteworthy limitations. Collecting the coins following culture required turning the 96-well plates upside down. Consequently, few coin cell layers were partly damaged. Only intact areas of the coins were used for confocal pictures and analysis. Other limitations include the use of titanium coins, which clearly differ from the cylindrical implants used in patients, and lack of standardization of parameters such as pressure and alignment during the experimental implantoplasty procedure. Although efforts were made to collect the experimental implantoplasty debris, particles may have been lost as aerosols during drilling. The fibronectin imaging was not possible to perform with standardized laser strength in all cases, which may have influenced the subsequent arbitrary quantification. Three of the cytokines were below the detection limit in the immunoassay analysis. No further attempts were made to adapt the cell medium to reach the detection limit. Furthermore, RT-PCR would have been useful in this study in order to verify the cytokine findings in this study also at mRNA level. Attempts were made to measure the total surface area after the experimental implantoplasty but this required the use of a mathematical model and assumptions we could not make. The different cellular behavior observed in this study may indeed also be explained by surface texture parameters not assessed in this study or non-topographical factors such as the altered chemistry of the surface following experimental implantoplasty as demonstrated in the present study. Another factor which may have influenced the results is corrosion from the titanium coins. This study was not designed to identify corrosion and therefore we cannot rule out any titanium corrosion from the coins and potential impact on cells during the 30 days experiment. This needs to be addressed in future research.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.