Do fluoroquinolone agents produce therapeutic benefits or harmful effects in patients with periodontitis? A systematic review and meta-analysis

Abstract

The adjunctive use of fluoroquinolone (FQ) agents in patients with periodontitis produces contradictory results. There has been no meta-analysis performed based on the evaluations of FQ use that would enable making appropriate clinical decisions. Our study aimed to evaluate, via a systematic review and meta-analysis conducted in accordance with the PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) guidelines, the clinical benefits, antimicrobial effects and safety profiles of the FQ agents administered to periodontitis patients under a conventional treatment regime. Relevant databases were searched for studies published up to May 2020, with the quality and risk of bias evaluations performed on the selected studies, and meta-analyses, funnel plots and heterogeneity tests carried out based on the obtained data. Any finding of p-value less than 0.05 was considered statistically significant. Quality and the risk of bias ranged from high to low. With acceptable heterogeneity and no reporting bias, the meta-analyses showed that local or systemic FQ use produced the following results: a reduced probing depth change (ΔPD) (p < 0.00001 at ≤3 months); reduced bleeding on probing (BOP) (p  < 0.00001 at 3–6 months); reduced subgingival detection of Aggregatibacter actinomycetemcomitans for up to 12 months (p-values from <0.00001 to 0.001); and an insignificant number of adverse events (p ≥ 0.05) in patients subjected to a conventional therapy as compared to those subjected to an antibiotic-free therapy. Our study found evidence to show that FQ administration provides clinical benefits and ensures antibacterial effects in periodontitis patients subjected to a conventional therapy regime.

Keywords: periodontitis, meta-analysis, fluoroquinolone, adjunctive therapy

Introduction

It is well-known that dental plaque can initiate gingival inflammation (gingivitis), which, in many cases, can progress, leading to the destruction of the underlying connective tissue and the alveolar bone (periodontitis), resulting in the loss of the affected tooth.¹ Periodontitis is one of the most predominant oral diseases; it affects the majority of the global population.² Variability in the individual host’s response to local factors plays a crucial role in this condition, and the loss of balance between the host’s response and the oral microbiome leads to development and progression of the disease.¹ Aggregatibacter actinomycetemcomitans (A. actinomycetemcomitans) and members of the red complex (e.g., Porphyromonas gingivalis – P. gingivalis, Tannerella forsythia – T. forsythia and Treponema denticola – T. denticola) are the most significant species associated with periodontitis.³ During the course of this disease, the foregoing bacteria penetrate the tissues and the periodontium, triggering the host’s immune response to the invading bacteria.⁴ Periodontitis also has been associated with the pre­sence of various systemic disorders, such as coronary heart disease, diabetes, cerebrovascular disease, and pancreatic cancer.^{5, 6}

The conventional treatment of periodontitis involves an oral hygiene program administered at home, and professional management (e.g., scaling) conducted via surgical or nonsurgical access to the affected sites. This kind of treatment is effective for the majority of patients and is followed by a recall program, which provides periodontal care every 3–4 months and helps to prevent the progression of the disease to a chronic state. However, the use of adjunctive treatment is necessary in the case of patients for whom the recall program has been unsuccessful. Adjunctive therapies comprise the local or systemic administration of anti­microbials,⁷ of which fluoroquinolone (FQ) agents are one of the most important antibiotics prescribed by oral healthcare professionals.^{8, 9} Fluoroquinolone agents, a family of broad-spectrum antibacterial agents acting against a wide range of aerobic gram-positive and gram-negative organisms,¹⁰ act by binding to an intracellular target in the cytosol of bacteria, where they inhibit the activity of DNA gyrase, with a high selectivity for prokaryotic enzymes.^{11, 12}

While numerous studies have evaluated the clinical and microbiological efficacy as well as the safety profiles of FQ agents in patients with periodontitis,^{3, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26} the results are often contradictory. A precise clinical evaluation requires the use of statistical methods, such as meta-analysis, to combine the results obtained through independent studies. Meta-analysis provides a more exact estimate of the health effects of treatment than those derived from individual studies.²⁷ To date, there have been no review and meta-analysis carried out to evaluate the efficacy and safety profiles of the FQ agents used as adjuncts to the conventional treatment of periodontitis patients. Therefore, the aim of our study was to evaluate, using the systematic review and meta-analysis methodology, the clinical, antimicrobial and harmful effects of the FQ agents admini­stered in combination with a conventional treatment for periodontitis.

Methods

Search strategy

The present study was conducted in accordance with the PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) guidelines.^{28, 29} In addition, the protocol for the systematic review portion of the present study was registered in the International Prospective Register of Systematic Reviews (PROSPERO) of the United Kingdom’s National Institute for Health Research (NIHR).

The PubMed, MEDLINE, Cochrane, LILACS, and Imbiomed databases were searched from their earliest records to May 31, 2020 to identify interventional studies that employed FQ agents for the adjunctive treatment of periodontitis. The following search terms were used for the PubMed, MEDLINE, Cochrane, and LILACS databases: “chronic periodontitis” AND “fluoroquinolone”; “chronic periodontitis” AND “levofloxacin”; “chronic periodontitis” AND “moxifloxacin”; “chronic periodontitis” AND “ciprofloxacin”; “chronic periodontitis” AND “ofloxacin”; “chronic periodontitis” AND “sitafloxacin”; “chronic periodontitis” AND “sparfloxacin”; “aggressive periodontitis” AND “fluoroquinolone”; “aggressive periodontitis” AND “levofloxacin”; “aggressive periodontitis” AND “moxifloxacin”; “aggressive periodontitis” AND “ciprofloxacin”; “aggressive periodontitis” AND “ofloxacin”; “aggressive periodontitis” AND “sitafloxacin”; “aggressive periodontitis” AND “sparfloxacin”; “periodontitis” AND “fluoroquinolone”; “periodontitis” AND “levofloxacin”; “periodontitis” AND “moxifloxacin”; “periodontitis” AND “ciprofloxacin”; “periodontitis” AND “ofloxacin”; “periodontitis” AND “sitafloxacin”; and “periodontitis” AND “sparfloxacin”. In this manner, 21 combinations of 2 different terms were used for each database. Due to the nature of the platform, the following single search terms were used for the Imbiomed database: levofloxacin (LVX); moxifloxacin (MOX); ciprofloxacin (CPX); ofloxacin (OFX); sitafloxacin (STX); sparfloxacin (SPX); fluoroquinolone; chronic periodontitis; aggressive periodontitis; and periodontitis.

Eligibility criteria

Research papers were selected based on the following inclusion criteria: a randomized or non-randomized controlled clinical trial that employed a parallel or split-mouth design to systemically treat healthy patients diagnosed with chronic, adult or aggressive periodontitis; at least 1 test group received an FQ agent; the report had been published in English or Spanish. The exclusions criteria were as follows: an FQ agent was not administered as an adjunctive therapy; an antibiotic-free group was not included in the study; the study participants had been prescribed an anti-inflammatory medication; the patients had taken an antibiotic in the previous 3 months; all the data used in the study had been taken from another study by the same author. This last criterion did not include duplicates, as those studies were not the exact copies of each other.

Data extraction

First, 2 researchers independently screened the titles and abstracts of the articles found, and then reviewed full-text papers. Once the initial evaluation was complete, a third researcher reviewed the work. Any discrepancy between the evaluations was resolved by consensus with the input of another experienced researcher. The following characteristics were extracted from each study: first author; year of publication; the participants’ age; study design; confirmed diagnosis of the disease; number of patients; intervention cha­racteristics (active principle, concentration, dose inter­val, and route of administration); periodontal para­meters; adverse effects; and the number of patients in whom periodontitis-related microorganisms had been identified via subgingival detection. The periodontal parameters used were probing depth (PD), clinical attachment level (CAL), PD change (ΔPD), gain of CAL (ΔCAL), percentage of sites with bleeding on probing (%BOP), plaque index (PI), and gingival index (GI). When a study featured more than 1 group treated with the same FQ agent, but at a different concentration, the information was harvested solely from the test group which received the dose with the most beneficial effects. The extracted information was grouped into the following periods of time: ≤3 months post-intervention; >3 and ≤6 months post-intervention; and >6 and ≤12 months post-intervention. When a study involved 2 or more examinations in an established timeframe, the information was collected solely from the examination with the longest timeframe.

Assessment of quality and the risk of bias

The quality of the selected studies was assessed using the Oxford Quality Scale,³⁰ as described previous­ly.^{31, 32, 33} Clinical trials with scores ≥3 were classified as high-quality studies, while those with scores <3 were considered low-quality studies. The internal vali­dity of the selected studies was evaluated using the Cochrane Collaboration’s risk of bias (RoB) tool.³⁴ This tool uses the following criteria to assess the risk of bias: random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting, and other bias. According to the procedure described by Higgins et al., each criterion was categorized as a low, unclear or high risk of bias,³⁵ with a risk of bias graph used to show the proportion of the selected studies falling into each category. Each quali­ty and risk of bias assessment was reviewed by an additional researcher, with any discrepancies resolved as described above.

Data analysis

The meta-analyses were performed based on the data obtained from studies of periodontitis patients treated with an adjunctive control or FQ intervention, administered either locally or systemically. For conti­nuous data, each meta-analysis was carried out with the use of the inverse variance (IV) method. The mean differences (or standardized mean differences) and their 95% confidence intervals (CIs) were analyzed with the fixed effects model, using the Review Manager 5.3 software.³⁴ For dichotomous data, the meta-analysis was performed with the Mantel–Haenszel (MH) method. The odds ratios (ORs) (or risk differences) and their 95% CIs were analyzed with the fixed effects model, using the software mentioned above, and statistic Z, statistic I² and p-values were used to evaluate the overall effect, heterogeneity and probability, respectively. Statistic I² ranging from 0 to 40%, from 40 to 70%, or from 70 to 100% was considered as absent, acceptable or considerable heterogeneity, respectively, with proba­bility of less than 0.05 accepted as significant. With a funnel plot used to detect reporting bias in each conducted meta-analysis conducted, the presence of any reporting bias produced an asymmetrical funnel.³⁶

Results

Characteristics and evaluation
of the studies

Thirty-one studies were identified as complying with the selection criteria of using an adjunctive FQ agent to treat patients with periodontitis (Figure 1). Of these stu­dies, nineteen were excluded: 1 did not employ an FQ insert as an adjunctive therapy,²⁵ 14 did not include a comparator group without an antibiotic,^{9, 21, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48} 3 studies used the data extracted from a study by the same author that had already been considered in the present study,^{13, 15, 49} and 1 other study used ibuprofen together with MOX.⁵⁰ Thecharacteristics and outcomes of the studies selected for analysis are summarized in Table 1 and Table 2. A dose of 0.4% MOX was selected for our review, as Flemmig et al. claimed that this dose produced better results than other doses.¹⁹ The studies conducted by Ardila et al.,¹⁴ Pradeep et al.,³ Pradeep et al.,¹⁷ Flemmig et al.,¹⁹ Khan et al.,¹⁶ Nakajima et al.,¹⁸ Guentsch et al.,²⁰ Kimura et al.,²⁶ Tezel et al.,⁵¹ Kleinfelder et al.,²³ Parthasarathy et al.,²² and Nagaraju et al.²⁴ received quality scores of 5, 5, 5, 5, 3, 3, 3, 3, 2, 2, 1, and 1, respectively.

Our quantitative analyses did not include the studies of Parthasarathy et al.²² and Kimura et al.,²⁶ as the former used incomplete data (means without standard deviations) and the latter solely reported undetectable subgingival levels of A. actinomycetemcomitans in their patients during the entire course of the therapeutic intervention. Furthermore, some of the information from the studies included in the present review could not be extracted, as it was presented solely in the graphic form (e.g., the PD, CAL, %BOP, and GI values presented by Khan et al.¹⁶). Since 2 studies by the same author,^{3, 17} which were included in the present study, were conducted in the same place and period of time, the data extraction procedure was performed in such a way as to avoid redundancy. Figure 2 shows the risk of bias for the 10 studies subjected to the quantitative analysis presented below.

Periodontal, bacteriological
and safety examinations

At ≤3 months post-intervention, the 95% CI, Z- values and p-values obtained in the meta-analyses showed a reduction in ΔPD and the number of patients with A. actinomycetemcomitans due to the adjunctive use of FQ, a finding that contrasted with the control data (Figure 3). On the contrary, the PI values as well as the subgingival levels of P. gingivalis and T. forsythia were not significantly modified by FQ use (Figure 3). Each of the above-mentioned analyses exhibited acceptable heterogeneity and reporting bias (Figure 3 and Figure 4). While the meta-analyses con­ducted for ΔCAL, CAL, PD, %BOP, and GI showed signifi­cant changes in their Z-values, significant heterogeneity and reporting bias were observed (Table 3).

At >3 and ≤6 months post-intervention, the 95% CI, Z-values and p-values obtained in the meta-analyses showed a reduction in the %BOP and the number of patients with A. actinomycetemcomitans due to the adjunctive use of FQ, a finding that contrasted with the control group (Table 3). On the other hand, the PI values and the number of subjects with P. gingivalis and T. forsythia were not modified by the FQ agents (Table 3). All the foregoing analyses found the absence of heterogeneity and reporting bias (Table 3). While the meta-analyses of ΔCAL, CAL, ΔPD, and PD showed statistical changes in their Z-values, these results presented with considerable heterogeneity and reporting bias (Table 3).

At >6 and ≤12 months post-intervention, the 95% CI, Z-values and p-values obtained in the meta-analyses showed a reduction in the number of patients with detectable A. actinomycetemcomitans due to the use of FQ agents as adjuncts to a conventional therapy, a finding that contrasted with the control group (Table 3). During this period, the control group exhibited reduced %BOP, in contrast with the test group, while the meta-analyses conducted for CAL and PD did not show beneficial changes due to the adjunctive use of FQ. Acceptable heterogeneity and reporting bias were found in these analyses (Table 3).

For the safety evaluation, the overall analysis (95% CI, Z-values and p-values) did not show a significant presence of drug-related adverse events due to the use of FQ in patients with periodontitis, in contrast to the control data. The absence of heterogeneity and reporting bias was observed for this analysis (Figure 5).

Discussion

To our knowledge, this is the first systematic review and meta-analysis that has evaluated the beneficial and harmful effects of the use of FQ agents as adjuncts to a conventional therapy for patients with periodontitis. The present study included patients diagnosed with chronic, adult or aggres­sive periodontitis (Table 1). It should be noted that the term ‘adult periodontitis’ was replaced with ‘chronic periodonti­tis’ in 1999 to avoid a diagnostic dilemma for clinicians.⁵² Previously, authors either classified periodontal disease in accordance with the American Academy of Periodontology’s 1999 classification system^{3, 14, 17, 19, 51} or reported the dia­gnosis of the disease without any description of the clas­sification system used.^{16, 18, 20, 22, 23, 24, 26} While we confirmed the classification of periodontal disease used in each study se­lected for the present paper,^{1, 52} we cannot support the dia­gnosis of the disease in 3 studies due to the limited information presented on the classification used.^{16, 22, 24} The review presented here analyzed aggressive and chronic periodontitis together, given that the updated guidance on the ma­nagement of periodontitis proposes that they be combined as a single entity, despite having different phenotypes.¹ The meta-analysis included only 1 study conducted on patients with aggressive periodontitis.¹⁴

All of the studies included in the meta-analyses exa­mined either locally or systemically administered CPX, LVX, MOX, SPX, STX, or OFX in combination with conventional treatment for periodontitis (Table 1). Our literature search found 2 previous reviews and meta-analyses in this area, including more than 25 studies.^{53, 54} Both these works compared the efficacy of systemic antibiotics combined with scaling and root planing (SRP) with the use of SRP alone. Each meta-analysis included only 1 study that had tested the use of MOX,²⁰ while the remainder of the analyzed studies used doxycycline, amoxicillin/metronidazole, metronidazole, azithromycin, clarithromycin, tetracyclines, amoxicillin plus clavulanic acid, ornidazole, or spiramycin in their test groups.^{53, 54} The combined use of CPX with metronidazole was mentioned in a previous extensive review that evaluated the effects of adjunctive antimicrobial therapies on A. actinomycetemcomitans-associated periodontitis.⁵⁵ The remaining reviews and meta-analyses evaluating the effects of antibiotics in patients with periodontitis did not include studies that used FQ in their test groups.^{56, 57, 58, 59, 60} Our study focused on FQ agents, since other types of antibiotics have already been extensively analyzed, as mentioned above. Moreover, the subgroup or independent analysis should be performed separately for each class of antibiotics in order to correctly evaluate the data from the application of meta-analysis methodologies.⁶¹

As P. gingivalis, T. forsythia and other bacteria compose the subgingival plaque biofilm, their microbial communi­ties can negatively affect host immunity in the oral ca­vity.⁶² Porphyromonas gingivalis is known to be the key pathogen underlying the pathogenesis of chronic periodontitis,⁶³ while A. actinomycetemcomitans plays a crucial role in the etiology of aggressive periodontitis, and is also associated with the etiology of chronic periodontitis.^{17, 64} Furthermore, very few clinical cases of aggressive periodontitis have been associated with the presence of P. gingivalis.⁶⁵ The present study found that systemic FQ use as an adjunct to a conventional therapy reduced the number of patients with positive A. actinomycetemcomitans detection in the subgingival plaque for up to 12 months post-intervention (Table 2, Table 3 and Figure 3). This microbiological benefit of the use of FQ agents results from their greater distribution in the gingival crevicular fluid as compared to other antibiotics, their strong antibacterial activity against A. actinomycetemcomitans, and their robust inhibitory effect on the early and mature phases of biofilm formation of A. actinomycetemcomitans.^{3, 9, 14, 66} The present study did not find evidence of a reduction in the number of patients with subgingival detection of P. gingivalis and T. forsythia for up to 6 months after a systemic FQ intervention (Table 3 and Figure 3). A previous study conducted on patients with chronic periodontitis showed that P. gingivalis isolates possessed low susceptibility to CPX as compared to other antibio­tics, such as doxycycline, amoxicillin/clavulanic acid and azithromycin.⁶⁷ However, a study conducted on patients with generalized aggressive periodontitis showed that P. gingivalis and T. forsythia isolates were greatly susceptible to MOX.⁶⁸ The above-mentioned in vitro susceptibility tests do not accurately reflect the clinical efficacy of antibiotics, and the strains can present different resistance profiles in different geographical areas, depending on the local use and abuse of antimicrobials.⁶⁸ While resistance against MOX in the isolates of periodontal bacteria from patients is yet to be reported, eventually, MOX may be found to be ineffective due to its structural similarity to CPX.¹¹ The present study quantitatively analyzed the information pertaining to A. actinomycetemcomitans, P. gingivalis and T. forsythia, as these periodontopathogens are strongly associated with the clinical para­meters of periodontal disease. Moreover, these bacteria are predictors for the treatment outcome¹⁴ and sufficient data was extracted to enable the corresponding analyses.

It is well-known that PD is a useful overall indicator of the state of periodontal pockets, while %BOP reflects the current status of periodontal inflammation and CAL reflects the accumulative periodontal damage.⁶⁹ The present study showed that the levels of ΔPD and %BOP improved at ≤3 months and at >3 and ≤6 months after local/systemic or systemic FQ use, respectively (Table 3 and Figure 3). However, the %BOP, CAL and PD results did not show beneficial changes for patients after 6 months of a systemic FQ intervention (Table 3). The improvement observed in the periodontal parameters was a consequence of 2 mechanisms exerted by FQ agents. Firstly, the antibacterial activity of FQ reduced the inflammatory response of the host due to a reduction in the level of periodontopathogens observed in pockets.⁴ Secondly, the direct modulation by FQ of the toll-like receptor 4-myeloid differentiation protein-2/nuclear factor-κB signaling reduced the synthesis and release of pro-inflammatory cytokines in immune cells.^{8, 70} It should be noted that Table 3 shows that a control therapy had a more beneficial effect on %BOP at >6 months than an FQ intervention. This estimate was produced based on the data extracted from a study by Kleinfelder et al., whose baseline examination presented a higher level of %BOP in the test group than in the control group.²³ Therefore, a control intervention was more effective, despite a similar level of change in the baseline %BOP for both groups.²³

The PI values were recorded in 7 out of the 12 selected studies (Table 2), as dental plaque plays an important role in the development and progression of gingival inflam­mation.⁷¹ The present study found that the adjunctive administration of FQ agents did not change supragingival plaque levels for up to 6 months (Table 3 and Figure 3). This was observed despite the antibacterial effect of FQ agents against bacteria that colonize the supragingival plaque biofilm, such as Streptococcus mitis (S. mitis), Streptococcus oralis (S. oralis) and Streptococcus mutans (S. mutans), their inhibitory action against the biofilm formation of S. mutans,^{62, 72, 73} and their extensive penetration into saliva after their systemic administration.⁷⁴ It should be noted that adequate oral hygiene routine, the use of mouth rinse and supragingival teeth cleaning (STC), which can contribute to disrupting supragingival plaque, were used in both the test and control groups in some of the studies analyzed in the present paper.^{71, 75, 76, 77}

The present study also found that the adjunctive use of FQ agents exhibited an acceptable safety profile in patients with periodontitis, as these antibiotics did not produce a significant change in the number of adverse events as compared to the control group (Figure 5). Dizziness, diarrhea and light-headedness were reported in patients receiving a systemic FQ intervention in 3 out of the selected studies,^{3, 17, 18} while gastrointestinal system disorders and resistance mechanism disorders were observed in both the local FQ and placebo therapy groups in 1 study analyzed here.¹⁹ On the other hand, 3 studies reported the absence of adverse events in patients subjected to both local or systemic drug delivery (Table 2).^{14, 16, 20} As the widespread use of antibiotics should not be promoted in dental practice,⁶⁰ dental care professionals must balance the clinical benefits against the adverse effects of FQ use, and consider the potential development of both microbial resistance to antibiotics and gut dysbiosis in patients. This latter condition can produce dysbiosis-related systemic diseases, such as increased susceptibility to infectious diseases, altered immune homeostasis, allergic diseases, and metabolic syndrome.^{78, 79} As resistance to FQ agents is mediated by a reduction in the number of porins and reduced accumulation of the drug in the bacteria,⁸⁰ the induction of resistant bacterial mutants is rapidly promoted by exposure to low FQ concentrations. The spontaneous mutation rate also contributes to drug resistance via the strain- and quinolone-dependent mechanism. The mechanism of resistance to FQ agents comprises a single amino acid mutation, which leads to a Ser83→Phe substitution in DNA gyrase.¹¹ In addition, the administration of FQ agents exposes the subject to a high risk of developing gut dysbiosis.⁷⁸ Since the protection of the gut microbiome during antibiotic therapies should be a priority for the dental care professional, the use of probiotics, such as Saccharomyces boulardii or Lactobacillus rhamnosus GG, can help to prevent gut dysbiosis. These probiotics compete with bacterial pathogens for attachment sites on intestinal cells, and then may exert their biological effects, including the modulation of the content of the gut micro­biota and the immune response.⁷⁸ Due to the beneficial health effects of probiotics, treatment with probiotics should be started as soon as possible after the commencement of FQ treatment.

We found 66.7% and 33.3% of the studies selected for the present study to be classified as high- and low-quality re­search, respectively, and acceptable levels of heterogeneity and reporting bias were found. All of the analyzed studies were categorized with a low (65.7%), unclear (28.6%) or high (5.7%) risk of bias (Table 3, Figure 2, Figure 3, and Figure 4). All said, the analyses found evidence that FQ can be used safely to improve clinical and microbiological parameters in periodontitis patients receiving a conventional therapy. The outcomes reported in the present study are similar to the beneficial health effects produced by the systemic or local application of other antibiotics (amoxicillin, clavulanic acid, metronidazole, azithromycin, clarithromycin, doxycycline, tetracycline, or chlorhexidine) in diabetic patients with periodontitis, smokers with chronic periodontitis or subjects with aggressive periodontitis.^{57, 58, 59, 60} Additionally, our results are in agreement with a recent consensus report on the use of adjunctive anti­biotics with SRP in patients diagnosed with periodontitis.⁵⁶ Given the previous absence of a review and meta-analysis of data taken from studies that used FQ combined with a conventional therapy, the report recommended the sensible and restricted administration of a combination of amoxicillin and metronidazole, metronidazole alone or azithromycin in patients with periodontitis.⁵⁶

The remainder of our analyses mostly showed that for up to 6 months post-intervention, FQ agents exerted more beneficial effects on the periodontal parameters as compared to controls (Table 3). However, as the meta-analyses conducted in the present study showed high levels of heterogeneity and reporting bias, these results should not be taken into consideration.⁸¹ Our study shows that there is an urgent need for further high-quality studies with a low risk of bias to add to the currently available evidence in this clinical context. The main limitation of our study was the small number of studies included in some of the meta-analyses undertaken, which may impact the interpretation of the results. Nevertheless, the use of the available data for appropriate statistical analyses is essential to support the growing demand for effective dental, medical and public health decisions.³²

Conclusions

Despite its limitations, the present study showed that the local or systemic administration of FQ agents produces positive effects on some clinical and microbiological parameters of periodontitis in patients receiving a conventional therapy. These positive effects, which include an improvement in ΔPD at ≤3 months, reduced %BOP at >3 and ≤6 months and a low prevalence of patients with A. actinomycetemcomitans for up to 12 months, are not observed following an antibiotic-free conventional therapy. All of the beneficial effects of FQ agents were observed with an insignificant presence of mild adverse events. The adjunctive use of FQ should be considered only when the expected benefits are greater than the potential adverse clinical consequences, including bacterial resistance, dysbiosis and dysbiosis-related diseases.

Ethical approval

This study was not performed on human participants or animals.