How surface coatings on titanium implants affect keratinized tissue: A systematic review

Abstract Apart from osseointegration, the stability and long‐term survival of percutaneous titanium implants is also strongly dependent on a qualitative soft‐tissue integration in the transcutaneous region. A firm connective tissue seal is needed to minimize soft‐tissue dehiscence and epithelial downgrowth. It is well‐known that the implant surface plays a key role in controlling the biological response of the surrounding keratinized tissue and several coating systems have been suggested to enhance the soft‐tissue cell interactions. Although some promising results have been obtained in vitro, their clinical significance can be debated. Therefore, the purpose of this systematic review is to gain more insight into the effect of such coatings on the interface formed with keratinized soft‐tissue in vivo. A comprehensive search was undertaken in March 2021. Relevant electronic databases were consulted to identify appropriate studies using a set of search strings. In total, 12 out of 4971 publications were included in this review. The reported coating systems were assigned to several subgroups according to their characteristics: metallic, ceramic and composite. Notwithstanding the differences in study characteristics (animal model, implantation period, reported outcomes), it was noticed that several coatings improve the soft‐tissue integration as compared to pristine titanium. Porous titanium coatings having only limited pore sizes (<250 μm) do not support dermal fibroblast tissue attachment. Yet, larger pores (>700 μm) allow extensive vascularized soft‐tissue infiltration, thereby supporting cell attachment. Nanostructured ceramic coatings are found to reduce the inflammatory response in favor of the formation of cell adhesive structures, that is, hemidesmosomes. Biomolecule coatings seem of particular interest to stimulate the soft‐tissue behavior provided that a durable fixation to the implant surface can be ensured. In this respect, fibroblast growth factor‐2 entrapped in a biomimetic apatite coating instigates a close to natural soft‐tissue attachment with epidermal collagen fibers attaching almost perpendicular to the implant surface. However, several studies had limitations with respect to coating characterization and detailed soft‐tissue analysis, small sample size and short implantation periods. To date, robust and long‐term in vivo studies are still lacking. Further investigation is required before a clear consensus on the optimal coating system allowing enhancing the soft‐tissue seal around percutaneous titanium implants can be reached.

Surface modification of Ti to fine-tune the surface physicochemical properties has been suggested to augment soft tissue integration. 5 However, conventional surface modification techniques such as bead blasting, etching or anodization, alter the original surface of the substrate. Coatings do not have this effect. Rather, coating enables the complete coverage of the pristine metal surface with a biologically active material that encourages the host cell interaction, without modifying the original surface. 14 Several coatings, mainly materials mimicking the components of living tissue, have been investigated for their potential to activate epithelial and/or fibroblast functions, such as inorganic CaP based coatings or biological coatings of extracellular matrix (ECM) components or growth factors. 5 Yet, many of these studies only involve in vitro research, which sometimes varying outcomes. Moreover, the clinical significance of in vitro results is controversial because methodologies often do not consider the complexity of the in vivo situation.
With this systematic review focused on in vivo evaluation, we aim to gain more insight into the effect of coatings on the Ti implantkeratinized tissue interface characteristics with the purpose of identifying those coatings that significantly improve the peri-implant seal in vivo and therefore are most promising for further clinical investigation.

| Eligibility criteria
Studies eligible for this review were: original research papers, case reports, (non-) randomized control trials and prospective and retrospective studies/case series, systematic reviews and meta-analyses.
Technical notes, editorials, letters to the editor, opinions or commentaries, which did not present original data were withheld. Only studies regarding the effect of coated Ti implants on the keratinized tissue seal were included. Studies solely researching the interface between Ti and bone were excluded. If studies reported results concerning the effect on soft tissue and osseointegration, only the keratinized results were accounted for. Only in vivo research was considered, this included animal studies as well as studies involving human subjects.
In vitro research on soft tissue healing and fibrosis was not taken into account, as these studies often lack consensus. Furthermore, research methods applied for in vivo and in vitro studies differ too much, thereby hindering a reliable comparison of the results. No restrictions with respect to the publication date were imposed. Only the English, German, French and Dutch literature was checked.

| Information sources and search strategy
The systematic literature search was performed using the following electronic databases: PubMed Central (www.ncbi.nlm.nih. gov/pubmed), Cochrane Library (www.cochranelibrary.com), Embase  Table 1 and

| Study records
The selection process (screening, eligibility, and data extraction) was carried out by two independent researchers (C. V. D. B. and B. Z.).
Articles were included through title and abstract screening. If eligible, the full text was analyzed and assessed for inclusion. All articles eligible for the systematic review were stored electronically in a full-text version.

| Risk of bias in individual studies
An assessment of internal validity, performance, selection, and other types of bias for individual human and animal studies was performed using the OHAT Risk of Bias Rating Tool for Human and Animal Studies. The analysis was done at study level and was carried out by two independent reviewers (C. V. D. B. and B. Z.).

| Study selection
The search yielded a total of 4971 articles, a detailed overview of the search results per database is given in Table 1. After initial screening of title and abstract, 54 records were found. The full-text articles were further assessed for eligibility and a total of 12 studies could be included in this systematic review. The overall quality of the studies under review was assessed using the OHAT Risk of Bias Rating Tool for Human and Animal Studies. Results deviated but were acceptable as given in Table 2.

| Study material
A detailed evaluation and data extraction was performed for the 12 selected studies, the major characteristics are given in Table 3 Note: "++": definitely low risk of bias; "+": probably low risk of bias; "-": probably high risk of bias; "--": definitely high risk of bias. Note: If no information was given concerning the characteristics a "-" was placed.
Abbreviations: Ag-HA, silver substituted hydroxyapatite; BAHI, bone anchored hearing implant; C, carbon; cp, commercially pure; cp, commercially pure; DLC, diamond like carbon; EMD, enamel matrix derivative; FGF-2, fibroblast growth factor 2; Fn, fibronectin; HA, hydroxyapatite; PGA, poly(L-glutamic) acid; PLL, poly(L-lysine); pTi, porous titanium; rhPDGF, recombinant human platelet derived growth factor; Si-HA, silicon substituted hydroxyapatite; Ti, titanium.  Ti flanges. 16 It was found that the pTi coating reduced epithelial downgrowth, but the epithelial attachment was similar for both flange materials. Yet, an increased dermal attachment could be observed for pTi flanges and the median percentage soft tissue fill and median density of fibroblast nuclei within the inner pores of the implant was significantly increased for pTi coated as compared to drilled flanges.

| Subgroup 2 -Ceramic coatings
Ti and its alloys meet many of the biomechanical requirements for load-bearing implants. Moreover, the stable oxide layer that forms at the surface minimizes metal ion release into the biological environment, which largely explains its biocompatibility. However, the mate-

| Inorganic-inorganic coatings
One coating type involving multiple inorganic materials that was   Nevertheless, these covered a wide range of coating strategies, either addressing the surface topography (pTi) or chemistry (ceramics) or introducing truly biologically active organic components (biomolecules) at the implant surface, as well as combined approaches.

| Organic-organic coatings
Porous structured Ti surfaces are considered for soft-tissue integration as these offer an enlarged specific surface area available for cell attachment and tissue ingrowth. 8 Overall, the here reviewed pTi coatings showed a good soft-tissue reaction, although care should be taken to avoid Ti particle release from the coatings, a complication commonly associated with plasma sprayed Ti coatings and which was found to trigger inflammation. 4,30 The Altering the surface chemistry of an implant may be considered more effective in controlling cellular behavior than surface topography and is therefore another valuable approach to fine-tune soft-tissue integration. 35 16 Similarly, rhPDGF coatings on TiUnite implants only showed a beneficial effect on the short-term, as soluble growth factors are prone to rapid enzymatic degradation. 22 For a more durable attachment and activity of the biomolecules, physical entrapment in an inorganic surface coating can be considered. 24 However, the structural ECM protein collagen type I physically entrapped in an anodically grown oxide coating on cp Ti did not improve the soft-tissue integration. 25 The authors hypothesized that the collagen degraded prematurely. Promising results, on the other hand, were obtained for FGF-2 which was physically entrapped in biomimetically deposited calcium phosphate coatings. 26  properties. In fact, few papers went into detail about their surface characteristics. A reference was mostly made to earlier publications performed by the same research group, however this data was mostly limited to coating thickness and average roughness. We have recently made recommendations for a comprehensive surface characterization to correlate with the soft-tissue response. 37 Functional coatings tend to be fragile as compared to the high insertion forces applied during implantation and can lose activity upon sterilization or storage.
Another important restriction is that the methodologies for evaluating the soft-tissue/implant interface varied significantly between studies, both in approach, profundity and histomorphometric analyses.
As indicated above, dermal attachment is a prerequisite to prevent epithelial downgrowth. Unfortunately, not all studies compared the dermal and epidermal connection to coated and pristine implants individually.
A comparison of the effect of the different coatings on the soft tissue interface was therefore difficult. Moreover, only a few studies incorporated detailed high-resolution imaging allowing to analyze the orientation of dermal collagen fibers in contact with the implant surface, even though it is believed that a perpendicular insertion of fibers confirms the establishment of a firm bioseal. Another limitation that needs to be addressed is that most studies only report on a small sample size with a variable time of exposure. This could be due to various reasons, amongst others ethical considerations regarding use of animals in in vivo experimentation. A small sample size may reduce the statistical power of a study. In very small studies, there exists a possibility of (selection) bias. Therefore, the study quality was assessed by the OHAT risk of bias framework to evaluate risk of bias on study level in human and non-human animal studies. All studies were acceptable according to this assessment. No studies were excluded solely based on sample size, as this can lead to loss of important data.