Clinical Efficacy of Chlorhexidine as an Adjunct to Mechanical Therapy of Peri-Implant Disease: A Systematic Review and Meta-Analysis

Dental implants have been widely used as rehabilitation therapy for fully or partially edentulous ridges. However, the inflammatory condition of tissue around implants often leads to peri-implant diseases, including peri-implant mucositis and peri-implantitis.¹ 

Peri-implant mucositis is a plaque-induced inflammatory process infiltrating into soft tissue around the implant and may progress to peri-implantitis, characterized by inflammation plus loss of supporting bone.²  If not reversed, peri-implant soft and hard tissue breakdown results in a significant implant failure rate.³  Based on the 11th European Workshop on Periodontology, the mean prevalence of peri-implant mucositis and peri-implantitis reached 43%, and 22%, respectively.⁴ 

Various approaches have been suggested for the treatment of peri-implant diseases, but no consensus on clinical protocols has been achieved. Mechanical debridement is traditionally proposed for peri-implant lesions.⁵  In addition, some adjunctive therapies, such as antiseptics and local or systemic antibiotics, are also recommended. Chlorhexidine (CHX), a bisbiguanide class cationic agent, is one of the choices for implant surface decontamination. By preventing plaque accumulation, it is considered beneficial for inhibition of the inflammatory process.^6,7  A clinical study showed significant improvement in bleeding on probing (BOP) when CHX-containing gel was used in combination with mechanical therapy for peri-implant mucositis.⁸  However, other studies^9–11  concluded that the most important factor in treatment was mechanical biofilm control and that additional oral antiseptic could be redundant for peri-implant diseases. Thus, the adjunctive use of CHX in peri-implant disease remains controversial and is often based on clinical experience.

The objective of this systematic review and meta-analysis was to help guide therapy selection for treatment of peri-implant diseases by determining the efficacy of CHX as an adjunctive therapy to mechanical debridement vs mechanical debridement alone.

This review was conducted following the PRISMA guidelines,¹²  and the study protocol was registered with PROSPERO (CRD42019120775).

PubMed, EMBASE, CENTRAL (Cochrane library), Web of Science, and ClinicalTrials.gov were searched without any limitations from inception to December 23, 2018.

We used Boolean operators to combine MeSH (Medical Subject Heading) terms, keywords, and other free terms, such as “chlorhexidine” and “peri-implantitis” to identify relevant publications. The detailed search strategy is reported in the supplemental material. No language or date restrictions were imposed. Furthermore, a manual search to review the reference lists of related publications and review articles was conducted by the first author of this report.

The eligibility criteria were based on the PICOS principle¹³—(1) patients: adults diagnosed with peri-implantitis and/or peri-implant mucositis; (2) intervention: mechanical debridement with adjunctive chlorhexidine; (3) comparison: mechanical debridement alone or in combination with placebo; (4) outcomes: results evaluating inflammatory levels; (5) studies: randomized controlled trials (RCTs) were included, regardless of publication status, publication date, or risk of bias.

Studies were excluded if they did not meet the standards above, and additional exclusion criteria included the following: (1) patients received other treatment, such as surgery, antibiotics, and other chemical agents; and (2) cross-sectional studies, case-control studies, cohort studies, case reports, literature reviews, in vitro studies, and animal studies.

Two coauthors independently scanned the study titles and abstracts for eligibility, potentially eligible papers were retrieved for full-text evaluation, and the primary reasons for exclusion of any reports were recorded. Disagreements were resolved by discussion with a third investigator.

We extracted data on interventions and outcomes from the studies ultimately included in the analysis. BOP reduction was regarded as the primary outcome. Secondary outcomes were probing depth (PD) reduction and clinical attachment level (CAL) gain. We also recorded study and patient characteristics including age, number of participants, number of implants, and follow-up months. Authors were contacted to clarify ambiguities and to request complete data from primary reports. If mean scores and standard deviations (SDs) were only reported in graphs and we were unable to obtain them from the original authors, we used the GetData Graph Digitizer software (version 2.26, USA) to extract the data. Because the SD of difference (SD [diff]) between checkpoints and baseline were not reported in some studies, we used the SD (baseline) and SD (checkpoints) to calculate the SD (diff), and the correlation coefficient was considered 0.5.¹⁴ 

Two independent reviewers assessed the methodological quality of the selected studies, based on the Cochrane collaboration tool for assessing risk of bias and including the following domains: (1) random sequence generation, (2) allocation concealment, (3) blinding of participants and personnel, (4) blinding of outcome assessment, (5) incomplete outcome data, (6) selective reporting, and (7) other sources of bias (eg, similarity of groups at baseline).¹⁵  Differences were resolved after discussion and, if necessary, consultation with a third auditor.

The GRADE (Grades of Recommendation, Assessment, Development, and Evaluation) approach was used to assess the quality of evidence for all outcomes in the included studies. This technique appraised the confidence in the estimates of effect by considering within-study risk of bias, directness of the evidence, heterogeneity of the data, precision of effect estimates, and risk of publication bias.

After grouping studies according to follow-up period (checkpoints of 1, 3–4, and 6–8 months), we used the Revman 5.3 software (Cochrane Collaboration) to quantitatively combined data. Because of the continuous outcomes, mean difference (MD) and 95% confidence intervals (CIs) were selected to evaluate differences in adjunct CHX effects. The statistical significance was defined as a 2-sided P < .05. Statistical heterogeneity was measured with the I²  statistic. An I²  statistic greater than 50% was considered to show moderate to high heterogeneity.¹⁶  Considering anticipated heterogeneity, the MD and 95% CI were calculated on the basis of the random effects model.

We performed prespecified subgroup analyses based on the hypothesis that patients with peri-implantitis would derive different degrees of curative effect vs peri-implant mucositis. We also performed subgroup analyses according to the administered CHX method to diminish potential heterogeneity. Publication bias was not tested, because the number of included studies was insufficient (<10).

Our search of the 5 databases resulted in 277 references after excluding duplicates; and after screening the titles and abstracts, we also excluded 238 records that did not meet the inclusion criteria. Next, we examined full-text copies of the remaining 39 studies, and selected 6 studies^8,10,17–20  that met the inclusion criteria developed for this review. Because we were unable to obtain the required data from 1 of the selected studies, 5 RCTs^10,17–20  in total were included in the final quantitative analysis (Figure 1).

There were 207 participants with 327 implants in the 6 included RCTs^8,10,17–20  (details in Table 1). Only 1 trial¹⁸  was multicenter. The publication period extended from 2002 to 2017. One study¹⁸  included patients with peri-implantitis (with marginal bone loss > 2 mm), whereas 5 studies^{8,10,17,19,20}  investigated peri-implant mucositis (without marginal bone loss). All the studies compared mechanical debridement combined with CHX vs mechanical debridement alone. Two^8,17  studies incorporated dental gel containing CHX, 3^10,19,20 included a CHX solution rinse, and 1¹⁸ used biodegradable CHX chips. The checkpoints ranged from 1 to 8 months.

Five trials^{8,10,17,19,20}  were judged “unclear risk of detection bias” because blinding of the outcome assessors was not described and thus may have influenced the outcomes. Randomization and/or allocation were not described in detail in 2 studies^10,19 ; therefore, we rated them “unclear risk of selection bias.” One study¹⁹  was rated “high risk of attrition bias” because of a high overall proportion of attrition. We rated another study⁸ “high risk of reporting bias,” mainly because of the absence of reported measures of variance and means for each group.

In short, the risk of bias was considered high in 2 RCTs^8,19 , low in 1 RCT,¹⁸  and moderate in 3 RCTs^10,17,20  (Figure 2). The quality of evidence in the included studies resulted in a majority of moderate-quality outcomes and 2 low-quality outcomes (Table 2).

All included studies demonstrated that adjunctive CHX was not significantly effective in BOP improvement at follow-up compared with control groups at follow-up, excepting 1 study⁸  with incomplete data.

Meta-analyses were performed for quantitative synthesis. The BOP test and control group summary measures were not significant (P > .05) after 1 month (MD, 0.10; 95% CI, −0.06 to 0.25; I² = 57%), 3–4 months (MD, 0.06; 95% CI, −0.03 to 0.15; I² = 26%), and 6–8 months (MD, 0.06; 95% CI, −0.03 to 0.14; I² = 0%) of follow-up (Figure 3). Secondary outcomes confirmed that the test groups also did not show significant improvement for PD or CAL compared with control groups (Figures 4 and 5).

The subgroup analysis based on dosage form is shown in Table 3. Two studies including 155 participants comparing CHX rinse to placebo were combined in a meta-analysis that presented a significant effect in favor of CHX for BOP reduction after 3 months (MD, 0.09; 95% CI, 0.01–0.18; I² = 0%). No significant differences were found in other subgroups comparing different types of CHX methods.

We performed sensitivity analyses for those outcomes with low heterogeneity (<50%) by modifying the random effects model to a fixed-effects model; however, no changes in outcome were found.

The present systematic review aims to confirm whether adjunctive use of CHX improves peri-implant clinical parameters. Results showed that CHX therapy had no significant effect on BOP reduction, PD reduction, or CAL gain; therefore, mechanical debridement may be sufficient for peri-implant treatment. All studies included in the meta-analyses supported this conclusion.

BOP reduction was considered the primary outcome of this review. The inflammatory manifestations in gingivae around implants are remarkable clinical features of both peri-implant mucositis and peri-implantitis.²¹  Whether or not bleeding occurs on provocation may be used as an objective approach to measure inflammation level. However, because the probing approach may vary in some aspects such as inserted depth, applied force, instrument angulation, and calibration, credibility of the combined results of BOP is limited.²²  Therefore, more diverse indices to evaluate the degree of inflammation, such as gingival index, are recommended in future studies.

Although CHX is regarded as the standard drug for plaque control,²³  our systematic review demonstrated that additional CHX did not significantly reduce BOP in peri-implant diseases. Test and control groups in all the analyses demonstrated significant diminution of inflammation, emphasizing the importance of mechanical biofilm removal and indicating that CHX had limited effects on clinical parameters. According to several in vitro studies, CHX also failed to show effective implant surface decontamination.²⁴  However, other reviews reported that adjunctive CHX could enhance the mechanical treatment of gingivitis and periodontitis. A reason for this discrepancy may be that all studies included in this review reported high-quality oral hygiene instruction (OHI) and mechanical decontamination, and only a limited number of patients in general may exhibit good compliance with OHI. Proper management provided to both test and control groups may control plaque and biofilm effectively. Additionally, because of the lack of substantivity to implants, CHX may not provide a benefit equivalent to its use on natural teeth,²⁵  explaining the contradictory results.^14,26 

One study⁸  included in this review (but not incorporated in the meta-analysis) did show that adjunct brush-on gel containing CHX had a statistically significant effect on BOP reduction. This could be related to the administration of CHX for up to 12 weeks. However, long-term use of CHX is not recommended because of various adverse effects.²⁷  Notably, the incomplete data reported in that particular study compromised the evidence quality and therefore did not alter our conclusions regarding CHX.

No significant differences were found between the test and control groups in PD reduction and CAL gain, which are equally important parameters for peri-implantitis. De Waal et al²⁸  investigated the role of additional CHX + cetylpyridinium chloride in the surgical treatment of peri-implantitis and found no improvements in clinical results. However, nonsurgical treatment for peri-implantitis, such as mechanical therapy used alone, appears to be insufficient.²⁹  A previous cohort study³⁰  indicated that a water jet incorporating CHX could improve the clinical efficacy of nonsurgical treatments. However, our meta-analyses suggested that adjunct CHX provided no significant improvement in the peri-implantitis. That result may be attributed to the only study¹⁸  included in our analysis involving peri-implantitis, which incorporated a hydrolyzed gelatin matrix as a placebo, and another report suggested that the gelatin matrix might exhibit an antibacterial effect after degradation.³¹  Therefore, future clinical studies focusing on the adjunct use of CHX in peri-implantitis are warranted.

Heterogeneity was moderate in 2 comparisons of our meta-analyses, including BOP reduction within 1 month and CAL gain in 6–8 months. Considering the limited number of included papers in these analyses, it was not placed into critical status. Nonetheless, some of the observed heterogeneity could be explained by the variation in CHX application methods. All the meta-analyses were included because of the evident consistency of the effect, despite the heterogeneity.

In this review, the included studies were heterogeneous in some aspects because of a general lack of well-designed studies in this area. To minimize the risk of bias and improve the validity of the comparisons, subgroup analyses were performed according to diagnostic criteria, follow-up time, and dosage form, and bias assessment was strictly conducted. One subgroup analysis for BOP reduction incorporating a CHX solution rinse was statistically significant, suggesting that a CHX solution rinse may be more effective than a dental gel. Because the number of studies was restricted, we were unable to perform subgroup analyses based on CHX concentration and application methods. Several reviews have demonstrated the efficacy of CHX on plaque and gingivitis, regardless of concentration; therefore, we speculate that this potential bias had little effect on our study outcomes.^26,32,33  Additionally, the application methods (toothbrush, sulcus inserting, and rinsing) of CHX were based on the dosage form. The heterogeneity can be resolved mostly by previous subgroup analysis. However, CHX solution was also applied by irrigation subgingivally in some studies, but it was used only once after mechanical debridement, and rinsing was used as a main method for 2 weeks. Therefore, the small difference in application methods would not alter our conclusion.

Our systematic review and meta-analysis summarized the role of adjunctive CHX to mechanical therapy in peri-implant disease. Based on our results, CHX use is not necessary for peri-implant diseases when high-quality mechanical decontamination is performed, and patients show good adherence to clinical directions. Moreover, the possibility that self-medication with CHX may mask serious underlying periodontal diseases should be taken into consideration.³⁴  If required, CHX solution (opposed to other formulations) may be preferable for clinical use.

There are some limitations to our review. The small number of studies made testing for publication bias difficult. Potential biases may have occurred because we estimated SD (diff) for some studies and extracted data from graphs in 1 published report. Besides, CHX application methods were still somewhat heterogenous and may bias the results. Consequently, further clinical trials including rigorous design and larger numbers of subject are required to address these issues.

Within the limitations of our review, the results indicated that mechanical debridement with additional use of CHX did not enhance the clinical results compared with mechanical debridement alone. Additional high-quality clinical studies with more subjects should be performed in future to confirm these results.

Abbreviations

BOP:

bleeding on probing

CAL:

clinical attachment level

CHX:

chlorhexidine

CI:

confidence interval

GRADE:

grades of recommendation, assessment, development, and evaluation

MD:

mean difference

MeSH:

medical subject heading

OHI:

oral hygiene instruction

PD:

probing depth

PICOS:

participants, intervention, control, outcome, studies

RCT:

randomized controlled trials

SD:

standard deviation

This work was supported by the National Natural Science Foundation of China (Grant 81870779), the International Cooperation Project of Chengdu Municipal Science and Technology Bureau (Grant 2015-GH02-00035-HZ), and the International Scientific Cooperation and Exchanges Project of Sichuan Province (Grant 2017HH0078).

The authors declare no conflicts of interest.