Characteristics of Local Delivery Agents for Treating Peri-Implantitis on Dental Implant Surfaces: A Preclinical Study

The success rate of dental implantation procedures is improving/¹ Dental implants are now considered one of the most effective treatment options for tooth loss, and the increasing popularity of implantation^2–4  could lead to an increased occurrence of peri-implant disease in the future.⁵  Peri-implantitis is an infectious disease impairing peri-implant tissue that causes mucosal inflammation and alveolar bone loss, and it is reportedly present in approximately 10% of implants and 20% of implant patients 5–10 years after the procedure.^6,7 

Even though various methods for treating peri-implantitis have been developed and applied, clear and common treatment standards remain to be established.⁶  Peri-implantitis treatments can be divided into nonsurgical and surgical methods. There is evidence that the nonsurgical treatment of submucosal mechanical debridement combined with an adjunctive local antibiotic therapy, submucosal polishing of the implant fixture with glycine powder, or irradiating the implant fixture with an Er:YAG laser can mitigate inflammation of peri-implant tissue,⁸  while surgical treatments include periodontal resective and regenerative surgery of peri-implantitis lesions.⁹ 

Most peri-implantitis treatment options are based on treating periodontitis, due to the similarity between the etiologies of these 2 diseases.⁹  The use of locally administered antibiotics and antiseptics is becoming a routine procedure in periodontitis treatments. The localized delivery of agents such as chlorhexidine, tetracycline, minocycline (7-dimethylamino-6-dimethyl-6-deoxytetracycline),¹⁰  and metronidazole makes it possible to maintain them at a constant level at the infection site.¹¹  Similarly, the local application of tetracycline for peri-implantitis was found to improve the peri-implant probing depth and bleeding tendency as well as to reduce total anaerobic bacterial counts at the peri-implantitis lesion.¹²  Also, chlorhexidine and minocycline are clinically effective in improving the bleeding score of pathologic peri-implant tissue.¹³  Several kinds of dental ointments containing antimicrobial substances, such as minocycline and doxycycline, are currently used clinically, including Periocline (Sunstar, Osaka, Japan), Minocline (DongKook Pharmaceutical, Seoul, Korea), MinoCure (NIBEC, Seoul, Korea), Arestin (OraPharma, Horsham, Pa), and Atridox (Tolmar, Fort Collins, Colo).

Several recent studies have found the local delivery of antibiotics, such as minocycline, to be an effective adjunctive therapy with mechanical debridement.^5,13,14  However, these studies have been limited by not focusing on the influence of local agents on the reosseointegration of periodontally compromised implant fixtures, instead being focused on decreasing bleeding on probing and periodontal pocket depth. Other studies have suggested that the wettability of the implant surface is one of the key factors for successful osseointegration.^15–18  According to Sartoretto et al,¹⁹  hydrophilicity of the implant surface accelerates osseointegration and positively affects the area of the bone-to-implant interface. Therefore, maintaining the hydrophilicity of the implant surface is an important prerequisite for osseointegration, which is the ultimate purpose of peri-implantitis treatment.^17,20 

Clinical irrigation is frequently applied to peri-implantitis lesions, but there is a possibility of local delivery agents (LDAs) without drug efficacy remaining on the implant surface after irrigating with normal saline. It has been demonstrated that some of these chemicals lose their efficacies by around 7 days after administration.^21,22  It is probable that ineffective LDA residues not only disturb reosseointegration on the implant surface when there is a history of peri-implantitis but also potentially act as a reservoir of microorganisms.

The purpose of this study was to quantify the LDAs that remain on implant surfaces after water cleaning using scanning electron microscopy (SEM) (Vega II LSU, Tescan Orsay Holding, Brno, Czech Republic) and energy-dispersive X-ray spectroscopy (EDS) (Iridium 500i, IXRF Systems, Austin, Tex). Because LDA residues have no effect on drug efficacies, a preclinical study model was designed to check for unwanted residues on implant surfaces.

Internal-connection implant fixtures with a sand-blasted, large-grit, acid-etched surface (Implantium, Dentium, Seoul, Korea) were used for this study. They were received in their original sterile packaging and kept sealed until the start of the investigation. Four kinds of widely used LDAs were prepared as follows: chlorhexidine gel (Hanall Biopharma, Seoul, Korea), tetracycline (Chong Kun Dang Pharmaceutical Corporation, Seoul, Korea), Minocline, and Periocline. Tetracycline solution was made by dissolving tetracycline powder in normal saline. Normal saline (0.9% NaCl solution; CJ HealthCare Corporation, Seoul, Korea) was used as a control.

The implants were divided into 5 groups based on surface treatment methods. In order to ensure that each implant was completely covered with the LDA, it was dipped and then rolled in the LDA for 5 minutes. Group 1 was a control group in which normal saline was applied to the implant surface. In group 2, chlorhexidine gel was applied to the implant surface, while tetracycline solution, Minocline, and Periocline were used in groups 3, 4, and 5, respectively (Figure 1). Each group was further divided into 5 subgroups based on the cleaning methods and durations. For subgroups A, B, and C, sufficient LDA was applied to the implant surface and then rinsed with running distilled water (JW Pharmaceutical, Seoul, Korea) for 10 seconds, 5 minutes, and 15 minutes, respectively (Figure 2a). For subgroups D and E, the implant surfaces with applied LDA were washed with the water spray of a dental-unit chair for 10 seconds and 5 minutes, respectively (Figure 2b). All 25 implants were dried at room temperature for 5 minutes and then packed and sealed.

After applying the previously described processes, all groups were investigated to identify the characteristics of the implant surfaces by SEM and to analyze the ingredients of the LDA residues by EDS. The middle-third portion of each implant surface was evaluated.

We used SEM to analyze the morphology of implant surfaces to reveal the pattern of LDA residue attachment. All implants were mounted on a sample mount with sterile forceps and then rinsed with ethyl alcohol to remove foreign particles and dust. The implants were electrically connected to the sample holder to prevent the electron beam from charging the sample and distorting the image. Use of SEM involves irradiating an area of interest with an electron beam, with the reflected electrons then being detected. Therefore, the implant surfaces were first coated with a conductive material (gold) using a sputter deposition system in the vacuum environment inside the chamber before capturing images for the SEM analysis.²³ 

In addition, EDS was applied to evaluate the chemical composition of LDA residues on the implant surfaces. This involves using an X-ray analysis system to detect the energy spectrum of X rays originating from a specimen material and analyze which elements are producing the X rays. A random area of each coated implant surface was selected for the examination. The device was set to a vacuum level of 10^–5 torr, an accelerating voltage of 20 kV, a takeoff angle of 35°, a beam current of 1 nA, a resolution of 130 eV, and a working distance of 20 mm to capture the X rays emitted from the implant surfaces. The EDS data were analyzed both qualitatively and quantitatively to identify the chemical composition of the implant surfaces.²³ 

The total areas of LDA residues on the implant surfaces were measured using an automated image analysis system (Image-Pro Plus, Media Cybernetics, Silver Spring, Md). All measurements were taken by 2 different dentists at least 3 to 5 times in the middle third of implant surface.

These raw data were reviewed by the independent statistician and then analyzed by the following procedure.

Statistical analysis was performed using SPSS software (version 20.0, IBM Corporation, Armonk, NY). Data were not normally distributed. The Kruskal-Wallis H test and Mann-Whitney U test were used to compare the areas of each of the LDA residues under the same cleaning condition and to compare the maximum of 2 cleaning methods in the same LDA groups, respectively. Also, the Kruskal-Wallis H test and Mann-Whitney U test were used to compare the areas of each of the LDA residues according to the application time increase of each cleaning methods. The cutoff for statistical significance was set at P < .05. In addition, post hoc analysis was performed by Mann-Whitney U test using Bonferronic correction method. The significance level (α) was equal to 0.05 divided by the number of groups, ₅C₂.

The SEM images for group 1 (the control group) demonstrated pure implant surfaces under all of the cleaning conditions (ie, subgroups A–E). The chlorhexidine gel in groups 2-A, 2-B, 2-C, and 2-D had been completely rinsed away, whereas some tetracycline residue was evident in groups 3-A, 3-B, 3-C, and 3-D. In contrast, the implant surfaces in groups 4-A and 5-A mostly remained coated with the LDA. More of the Periocline had been removed in group 5-B than group 5-A. Similarly, more Minocline was rinsed off in group 4-B than group 4-A, although large amounts remained. Minimal LDA residue was found in group 5-C, while group 4-C remained coated with Minocline (Figure 3). The LDA remained completely coated on the implant surface in groups 4-D and 5-D. No LDA residues were observed on the implant surfaces after washing in groups 2-E, 3-E, and 5-E, while large amounts of Minocline residues still remained in group 4-E (Figure 4).

In order to examine the amount of LDA residue on the implant surfaces more effectively, the area ratio of each implant surface covered by the LDA was calculated and then compared between implants. In all subgroups, regardless of the cleaning method, the amount of the LDA residue covering the implant surfaces decreased as the cleaning duration increased. In group 1, no LDA remained on the implant surface even after rinsing or washing for only 10 seconds.

In this study, as a result of the various cleaning (ie, rinsing and washing) durationd, all test groups showed statistically significant results in the area changes of LDA residues, compared with the control group. Generally, in the oral cavity natural flow of gingival crevicular fluid (GCF) and saliva, or clinical irrigation and mouth gargling, applies on LDA residues with various durations and pressures; thus this study properly reflects the actual environment (Table 1).

In group 2, the chlorhexidine gel was almost completely removed from the implant surface after rinsing for the longer duration under running distilled water. In groups 3 and 5, the amount of LDA residue on the implant surfaces in subgroup C was less than half that in subgroup A. However, in group 4, more than 90% of the implant surface was still covered with Minocline even after rinsing with running distilled water for 15 minutes (Figure 5). The area changes of each of the LDA residues in accordance with rinsing time increased in all LDA groups and showed significant results (Table 2).

Using the water spray of a dental-unit chair to wash the implant surfaces in subgroups D and E involved applying a higher cleaning pressure. Except in group 1, in which it disappeared immediately upon cleaning, the amount of LDA residue tended to significantly decrease as the cleaning pressure increased regardless of the application time (Table 3). In groups 1 and 2, the LDAs were removed after washing for only 10 seconds. Therefore, no significant changes were shown based on longer washing time (Table 2). In group 3, when the washing duration was increased from 10 seconds to 5 minutes, LDAs were no longer present on the implant surface. In group 5, 99% of the LDA residues disappeared when the washing duration was increased to 5 minutes. On the other hand, Minocline covered about 90% of the implant surface in groups 4-D and 4-E, despite the application of a higher cleaning pressure compared with the rinsing method (Figure 6). Therefore, LDA residues significantly decreased in groups 3, 4, and 5, as washing time increases (Table 2).

As per the independent statistician's review, the statistical analysis methods were well set up and applied for this study. Therefore, this study conclusion has also been statistically appropriately drawn. In addition, the aforementioned statistical analysis provided reliable results because they were analyzed by the independent dentist and statistician.

The LDA residues were analyzed in more detailed by examining the implants in subgroup C (rinsing in running distilled water for 15 minutes) and subgroup E (washing with the water spray of a dental-unit chair for 5 minutes). After applying the rinsing or washing process, a part of the surface of each implant where the LDA remained was randomly selected, and then the components at that site were analyzed.

Titanium (Ti) was detected at concentrations of 94.370%, 92.324%, and 95.459% in groups 1-C, 2-C, and 3-C, respectively. In addition, inorganic substances contained in each LDA residue and in the distilled water used for cleaning was detected at levels of 0–6.713%. Since the main component of an implant surface is Ti, the detection of more than 90% Ti can be attributed to most of the LDA being removed from the implant surfaces and the implant surfaces being exposed. On the other hand, Ti was detected at much lower levels in groups 4-C and 5-C, of 0% and 0.892%, respectively. This means that the LDA residues covered more than 99% of the implant surfaces, indicating that Minocline and Periocline had not been completely rinsed away in running distilled water. This is supported by the detected levels of carbon (C) and oxygen (O) in minocycline hydrochloride (whose molecular formula is C₂₃H₂₇N₃O₇ · HCl), which is a major component of both LDAs. The detected levels of C and O were 98.813% and 89.949%, respectively. The detected levels of other inorganic substances in groups 4-C and 5-C were similar to those in groups 1-C, 2-C, and 3-C (Figure 7).

In groups 1-E, 2-E, and 3-E, the detected levels of Ti were 90.938%, 90.701%, and 89.551%, respectively, indicating that the LDAs were completely removed from approximately 90% of the implant surface. Other inorganic substances were detected at small amounts, from 0 to 8.897%. In group 5-E, the detected level of Ti was 90.232%, which is much higher than that in group 5-C; this indicates that Periocline is easily washed off when using a higher cleaning pressure. Also, other inorganic substances were detected at levels similar to those in the control group and mostly consisted of C and nitrogen (N), totaling 9.158%. However, in group 4-E, Ti was detected at 0.570%, while both C and O were detected at 98.788%, similar to that in group 4-C. It can therefore be assumed that more than 99% of the Minocline remained on the implant surface without being washed off, even when a higher cleaning pressure was applied (Figure 8).

In all of the groups except group 4, the amount of LDA residues removed from the implant surfaces appeared to increase with the applied cleaning pressure. However, unlike other LDAs, applying water at a higher cleaning pressure did not result in the removal of Minocline from the implant surface—it can be assumed that the water insolubility of Minocline is the main factor underlying these results.

Chemical treatment, one of the nonsurgical methods of peri-implantitis treatments, involves the direct application of an appropriate LDA to the implant surfaces for disinfection and decontamination.²⁴  Also, LDAs are commonly used for periodontitis treatments, with citric acid, hydrogen peroxide, chlorhexidine, and saline mainly being used as LDAs in clinical practice; these LDAs have shown similar effects in experimental studies.²⁴ 

The present study focused on the 5 LDAs that are commonly used in clinical practice. The duration of their efficacy at the applied site is limited, and those LDAs that are no longer effective constitute useless residues. These LDA residues should be removed by the flow of GCF and saliva as well as regular clinical irrigation and mouth gargling.

The results of this study demonstrate that the removal patterns of the 5 tested LDAs differed with the cleaning method. In subgroups C and E of groups 1, 2, and 3, normal saline, chlorhexidine gel and tetracycline solution were fully removed from the implant surfaces under continuous cleaning. Therefore, the continuous flow of GCF and saliva in the periodontal tissue will be sufficient to naturally remove these LDAs. Groups 5-C and 5-E showed that cleaning with a sufficient pressure is effective for removing Periocline, indicating that clinical irrigation or mouth gargling is essential. On the other hand, the removal pattern of Minocline did not differ between groups 4-C and 4-E.

The following three reasons may be suggested to explain the observations made when attempting to remove Minocline: (1) short cleaning duration, (2) low cleaning pressure, and (3) the water insolubility of Minocline. This study found that the cleaning duration and pressure are not key factors influencing the removal of Minocline from the implant surface, and so it can be deduced that its water insolubility is the main reason for Minocline remaining on the implant surface. In other words, components other than the main component (minocycline hydrochloride) play an important role in determining the physical characteristics of Minocline and Periocline.

A study of the effects of normal saline²⁵  found that the lipopolysaccharide level was significantly lower compared with untreated implant surfaces when cotton pellets were immersed in 0.9% normal saline for 1 minute and the infected implant surfaces were wiped off. However, since normal saline is rarely used alone, it is difficult to evaluate its value as a therapeutic agent; instead, it is mainly used as an additional agent in other antitoxin treatments.

Chlorhexidine, an antimicrobial agent, is a cationic bisbiguanide with broad antibacterial effects, low toxicity, strong adsorptivity on the skin and mucous membranes, and continuous bacteriostatic action,²⁶  Its mechanism of action involves changing the bacterial adhesion to teeth and disrupting the bacterial cell walls so as to induce cell death, and its major effect is the aggregation of large cytoplasmic molecules due to increased cell-wall permeability.²⁷  In vitro studies have shown that chlorhexidine acts this way on Ti, plasma-sprayed, and hydroxyapatite-coated implants.²⁸  Tetracycline^27,29  is bacteriostatic and exhibits broad-spectrum activity, inhibiting both Gram-positive and Gram-negative organisms. This hinders bacterial protein synthesis and human tissue collagenase activity, and its concentration is 2–10 times higher in the gingival sulcus than in plasma, thus making it more effective in periodontitis treatment. Also, it prevents periodontal tissue destruction via slow release after adhering to the tooth surface.

Minocycline is a second-generation and semisynthetic tetracycline analog that is stable in vivo and is characterized by a high absorption rate when applied orally.^29–32  Also, minocycline has been a useful agent for treating local bacterial infections in the form of minocycline hydrochloride.³³  Minocline and Periocline are LDAs containing minocycline hydrochloride, and they are clinically used for both periodontitis and peri-implantitis treatments.

A 0.5-g syringe of Periocline contains 10 mg of minocycline hydrochloride. It also contains a hydrophilic biodegradable substrate comprising hydroxyethyl-cellulose, aminoalkyl-methacrylate, triacetin, and glycerine.^34–36  The minocycline hydrochloride is released slowly while its substrate is degraded,²⁹  and the remaining Periocline (ie, the substrate components without minocycline hydrochloride) is absorbed physiologically. The concentration of minocycline hydrochloride in a periodontal pocket was measured at 1300 μg/mL at 1 hour after administering 0.05 mL of Periocline, and it subsequently decreased to 90 μg/mL after 7 hours.³⁷  Periocline not only exerts a significant effect on periodontitis lesions but also is hydrophilic and biocompatible, and so its risk of side effects is low due to its residues being removed from the implant surface after a certain time period.³⁵ 

Like Periocline, a 0.5-g syringe of Minocline also contains 10 mg of minocycline hydrochloride. An effective concentration is reportedly maintained even 100 hours after applying Minocline to Porphyromonas gingivalis,³⁸  and it remains in the periodontal pocket for up to 7 days.³⁹  Unlike Periocline, Minocline contains the substrates comprising alginate and chitosan.³⁹  Alginates are refined from seaweed and become viscous and insoluble when mixed with water.⁴⁰  Chitosan is formed by deacetylation of chitin followed by conversion into the amino group; it is nontoxic in vivo and inexpensive. However, it is insoluble in most organic solvents and can be dissolved only in dilute acidic solutions with pH values below 6.0.^41,42  Chitosan in polymeric form is not easily absorbed into the human body, and oral absorption is more difficult.⁴³  Therefore, chitosan does not disappear physiologically or naturally even after its drug efficacy has finished, and it will remain unnecessarily on the implant surface.

The presence of the aforementioned components will physically block the migration and attachment of cytokines and cells associated with bone formation and healing, thereby interfering with the reosseointegration of a dental implant. In addition, this can facilitate the colonization of microorganisms such as bacteria and the formation of reservoirs, which will eventually lead to an environment in which peri-implantitis can recur. Microorganisms can colonize more rapidly on hydrophobic and nonpolar surfaces,⁴⁴  and the plaque growth over 9 days was found to be about 10 times greater on a hydrophobic surface than on a hydrophilic surface.⁴⁵ 

Minocline was found to be stickier than Periocline after being applied to the implant surface. In addition, even after a long-duration or high-pressure cleaning method, 90% of the Minocline remained on the implant surface. In other words, Minocline residues—without any drug efficacy—coated on the implant surface can form a hydrophobic and nonbiodegradable environment. Ultimately this acts as a reservoir of microorganisms, such as hydrophobic bacteria, and disturbs re-osseointegration of the implant, which is the ultimate goal of peri-implantitis treatment.

While Periocline contains minocycline hydrochloride (as does Minocycline), it is also hydrophilic and biodegradable. Therefore, when Periocline was cleaned for a longer duration or using a higher pressure, the cleaning results were similar to those for the other LDAs used in this study. In other words, Periocline residues—without drug efficacy—can be removed from the implant surface through the continuous release of GCF and saliva as well as regular clinical irrigation and mouth gargling.

Various surgical and nonsurgical peri-implantitis treatments have been suggested. The effective use of LDAs is very important because such nonsurgical treatments are applicable to a wide range of patients and can be used to complement surgical treatment and thereby produce effective results.

This study investigated the physical properties of LDAs, which are essential in clinical practice, and how they actually work on the implant surface. This study was limited by the experiments not being performed under normal oral conditions, but the findings are still meaningful for understanding the characteristics of LDAs applied to implant surfaces in a controlled situation. A good understanding of the characteristics of each LDA will be helpful to dentists in their clinical practices. In addition, the results of this study and those of future clinical studies will make it possible to compare and analyze the characteristics and effects of various LDAs.

Different LDAs based on minocycline hydrochloride exhibit different properties when they are applied to the surface of a dental implant, and these properties depend on the constituent substrates. The results of this study indicate that greater attention needs to be paid to cleaning procedures and careful follow-up planning when using Minocline in clinical applications compared with applications using Periocline.

Abbreviations

EDS

energy-dispersive X-ray spectroscopy

GCF

gingival crevicular fluid

LDA

local delivery agents

SEM

scanning electron microscopy

This study was supported by a grant awarded in 2015 by Ilsan Hospital, National Health Insurance Service (Grant no. 2016-19). The raw data were reviewed by the celebrated scholar and statistician Kwang-Hak Bae, chief director, Oral Health Science Research Center, Apple Tree Dental Hospital. The authors would like to show gratitude to him.

The authors declare that they have no conflict of interest.