Biological significance of long-term bisphenol A release in the saliva of patients wearing orthodontic appliances: A systematic review and meta-analysis

Background Orthodontic appliances contain Bisphenol A and are controversial due to its potential risks for human health. Thus, the aim of the present research was to identify the presence of Bisphenol A in the saliva of patients wearing orthodontic appliances. Material and Methods A systematic search of the literature was conducted in four electronic databases (PubMed, Embase, Scopus, and Web of Science) and a manual search of grey literature. Research was done up to March 2023, without language restrictions. Based on inclusion/exclusion criteria, data were extracted by two independent reviewers. Results A total of 2293 potentially eligible articles were identified, of which 8 were finally included. The studies included a total of 238 patients and showed a moderate quality in the PEDro scale. All the devices studied released Bisphenol A into the saliva, with the polycarbonate brackets being the ones that released it for a longer time. The most significant increase occurred in the first 30 minutes after bonding with composites, reaching 697 µg/g. with polycarbonate brackets. Conclusions Although a statistically significant increase of Bisphenol A levels in the saliva of orthodontic patients were found, this increase does not exceed the maximum allowable daily intake. Thus, the use of these materials can be considered safe for human health. Key words:BPA, Bisphenol-A, cytotoxicity, Orthodontic materials, Composite resins.


Introduction
In orthodontics, bracket placement through adhesive compounds is a daily chore that requires efficient and trustworthy firm adhesion.The material sciences of resin-based composites have suffered a strong and continual advent of improvement with multiple generations of adhesive systems introduced in the market along the past years (1,2).The adhesive composites in orthodontics have similar chemical composition than those of dental restorative composite materials, consisting of inorganic fillers, a solvent base and an organic matrix that most often includes glycidyl methacrylate (Bis-GMA), which breaks down into a biochemical monomer called bisphenol A (BPA) (3).Temperature and pH changes, mechanical wear, and bacterial or salivary enzymatic action accelerate this process.Apart from resin composites, this monomer is also used in the production of a wide variety of orthodontic materials such as aesthetic brackets, elastomeric ligatures, thermoplastic aligners, chains and acrylic retainers (4,5).Despite the successful use of these materials, there is still some uncertainty about the safety and biocompatibility of some potentially harmful monomers and BPA is the most controversial one.Despite numerous discussions and a lack of consensus about its safety, BPA was classified by the European Chemicals Agency as a 'substance of very high concern' since it was identified as an endocrine disruptor with risks for human health (toxic for human reproduction) and environment, as determined in the Regulation Statement (6).When adhesive composites or some orthodontic materials degrade in the mouth, BPA is released into the saliva.From saliva, passes into the rest of the body fluids, having been found in urine, adult and fetal blood, amniotic fluid, the placenta or even in breast milk (7).Its molecular structure mimics the structure of natural estrogens, producing disruptions in the endocrine system and having teratogenic effect, even at low doses.It has also been reported to accelerate the onset of puberty, to cause feminization in men, to have carcinogenic effects (breast and prostate) (8) or serve as a risk factor in human fertility.In addition, BPA could have an impact on the psychosocial health of children as it has been associated with higher levels of anxiety, depression and social stress.Apart from that, many cases of allergic contact dermatitis have also been reported among dental personnel (7).
Although sufficient medical research has proved the harmful effects of materials that release BPA, the doses usually released in the dental field do not exceed 50 mg/ kg/day, which is the maximum allowable daily intake established by the United States Environmental Protection Agency (EPA) to consider it toxic or harmful (9).Since orthodontic materials are expected to have a ser-vice life of several years in the mouth, it is necessary to investigate the long-term release of those potentially harmful ingredients Also, the debonding process may lead to the release of BPA in the saliva (5).Studies on the release of BPA from materials used in orthodontics have solely focused on identifying, in vitro, whether these materials have estrogenic potential or not (9).There are few studies and no reviews focused on studying the release of BPA in vivo, in patients wearing orthodontic appliances.Therefore, the main objective of this work was to analyze the presence of BPA in the saliva of patients wearing orthodontic or retention appliances and to assess if the levels are within the limits reported by the international protection agencies.

Material and Methods
-Study protocol and registration This systematic review was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement (10).and was previously registered in PROSPERO (CRD42022380640).The focused PICO research question was: Do orthodontic materials (I) release BPA in the saliva (O) of patients wearing orthodontics or retention appliances (P)? -Eligibility criteria The following eligibility criteria were applied: 1) Population: clinical trials, conducted on patients wearing orthodontic appliances.2) Interventions: presence of orthodontic or retention appliances.4) Outcome measure: assessment of the levels of BPA in saliva in relation to the type of material and time of exposure.The exclusion criteria were in vitro studies, patients who did not wear orthodontic or retention appliances and studies about resin o sealant restorations, case reports, review and systematic review articles, retrospective studies, editorials, opinions, surveys, guidelines, conferences, commentary articles and in vivo animal studies.
-Information sources and search strategy An electronic search was conducted through Pubmed, Scopus, Embase and Web of Science.In addition, a search of the grey literature was performed through Opengrey as well as a hand search of the bibliography.The search was conducted up to March 2023, with no constraints in terms of publication year or language.The search strategy is shown in Table 1.Two researchers (RG-B and AV-C), working independently, systematically assessed the titles and abstracts of all identified articles.If disagreement occurred, a third author (BT) was consulted.If the abstract did not contain enough information to include or exclude a particular article, the authors read the full article before making a final decision.Once potential studies were identified for inclusion, both authors retrieved and reviewed the full texts.e914

Risk of bias in individual studies
The quality of clinical trials has been assessed using the Physiotherapy Evidence Database (PEDro) scale (11,12).Any disagreement between the two initial investigators was resolved by consensus and, when in doubt, a third investigator was consulted (Table 2).-Quantitative synthesis (meta-analysis) The studies were combined with a random effects model using the maximum likelihood method.The effect size used was the standardized mean difference and was interpreted using the Cohen scale (d= 0.2-0.5 small effect size; d=0.5-0.8 medium; and d>0.8 significant).The forest plot shows the analysis by subgroups according to the time point.Heterogeneity was assessed with the Q test, p value<0.1 and by I2 (25-50% mild;50-75% moderate;>75% important).
The presence of publication bias was analyzed using the Trim and Fill study imputation method to obtain symmetry in the funnel plot and with the Egger regression intercept at p<0.1.

-Selection of studies
The electronic search identified 2293 preliminary references of which 597 were from Pubmed, 586 from Embase, 528 from Scopus and 582 from Web of Science.
Duplicate articles were manually discarded resulting in 709 entries.From these, 676 were excluded by reading the title and abstract.After reading the resulting 33 articles, 25 were discarded because they did not meet the inclusion criteria.The reasons for rejection of the excluded articles were recorded (Supplement 1) (http://www.medicinaoral.com/medoralfree01/aop/jced_61735_s01.pdf).Finally, 8 were included for qualitative synthesis and 2 for the meta-analysis.The PRISMA 2020 flowchart (Fig. 1) provides an overview of the study selection process.

Study characteristics
The eight studies included in this systematic review were clinical trials (Table 3).(15), that studied photopolymerizable composite and also started from detectable levels of BPA in saliva, 0.56 ng/g-155 ng/l respectively.In addition, Kang et al. (16) found that the hybrid composite released more BPA (7.2676 ng/ml vs. 2.3211 ng/ ml) with a statistically significant difference between the groups.The study by Kloukus et al. (15), found statistically significant differences in the increase in BPA after bonding and in the decrease in BPA after the water rinse, but not with the water/ethanol rinse.The studies by Moreira et al. (9) and Kang et al. (16) agreed that the greatest release of BPA occurs 30 minutes after bonding, but Kang et al. (16) did not find detectable levels per month and Moreira et al. (9), yes.Raghavanan et al. (8), found that essix retainers released more BPA than self-cured Hawley retainers followed by heat-cured Hawley retainers.On the other hand, the highest levels of BPA released by the essix retainer were detected after a week and with the Hawley this happened after their placement.Watanabe et al. (13) found a progressive increase of BPA levels in saliva after bonding polycarbonate brackets, even 40 months later.Furthermore, the release of this monomer was higher in vivo than in vitro.
-Risk of Bias The articles have obtained moderate quality on the PE-Dro scale (Table 2).This scale is made up of 11 items, with 11 being the highest score.However, item 9 has not been assessed, due to the absence of a control group because the studies compare the initial values of BPA and the values after inserting the orthodontic device in the same patient.Thus, the maximum possible score is 10.Two studies obtained 6/10 (14,15), two studies 5/10 (3,8) and three 4/10 (9,13,16).The study by Seifi et al. (5) obtained the highest score because it is the only one that meets item number 7 that corresponds to the masking of the evaluators.
-Quantitative synthesis (meta-analysis) Two studies (9,14) with four time points have been combined using a random effects model.The estimated effect size was the standardized mean difference (d) between the intervention group and the control group at each of the four time points using a subgroup meta-analysis.Figure 2 shows how a release of BPA has been estimated with an effect size at time point 1 (30 minutes after the brackets were bonded) of 3.03 (CI 1.66-4.40).In this way, and according to the Cohen scale, we found an important effect size, being statistically significant (p value < 0.001).The effect size decreases over time, ceasing to be statistically significant.Thus, at time point 2 it is 2.42 (IC -1.37-6.21),at time point 3 it is 1.81 (IC -0.13-3.76)and finally at time point 4 it is 1.51 (IC -0.68 -3.71).The meta-analysis has presented significant heterogeneity with a Q test p value <0.001 and I2 = 91.9%.
-Publication Bias Through the trim and fill analysis (Fig. 3), we can observe the imputation of 3 studies to the left of the funnel plot to form a symmetrical image, resulting in a new estimate that differs significantly from the initial estimate.In addition, with Egger's regression analysis, we obtained an intercept value of 15.9 (CI 13.4-18.4),statistically significant with a p value <0.001.For all these reasons, we can affirm that the meta-analysis is exposed to a possible publication bias.

Discussion
Even though BPA is present in most of the composites used for bonding brackets and retainers, the literature on e918 the release of BPA in saliva is still limited.Although this monomer has been found to be an estrogenic disruptor that can have serious repercussions at high doses causing potentially health problems, is still unclear whether its presence in dental composites can have a harmful effect.There is published literature related to the release of BPA from orthodontics materials, such as the articles of Iliadi et al. (4) and Halimi et al. (7), but these reviews included both in vivo and in vitro studies.In this systematic review it was decided to include only in vivo studies, since the study by Watanabe et al. (13), found that BPA was released 10 times more in vivo than in vitro.So, the inclusion of in vitro studies could lead to underestimating the levels of BPA.The systematic review of Iliadi et al. (4), focused on analyzing the cytotoxic effects and the release of BPA only from thermoplastic aligners and retainers, unlike the present review that analyzed all the materials used in orthodontics.The only article included in both reviews was the one of Raghavan et al. (8).On the contrary, the systematic review of Halimi et al. (7) included 21 articles, of which 10 were in vitro studies and 11 studied the release of BPA from orthodontic materials.Only the studies of Kloukus et al. (15) and Kang et al. (16), were in vivo studies and both have also been included in the present review.The studies have a medium/low quality and none of them meets items 3,5,6, which correspond to blinding, since both the professionals and the participants knew the group to which each patient belonged.Only the study by Seifi et al. (5), who achieved the highest score, meets item 7, which corresponds to the masking of the evaluators.
Secondly, the study with the longest follow-up was the one of Watanabe et al. (13).These authors estimated that the release of BPA from polycarbonate brackets over 34 months was 374 μg/g.This value corresponded to 8.2 μg per bracket (bracket weight, 22 mg).If 10 brackets were applied to a 40 kg body weight patient for 34 months, then the total BPA intake would be 82 μg or a daily intake of 2.0 ng/kg/day.This value is 100 times lower than the maximum non-estrogenic effect level of BPA of 200 ng/kg/day, whose value was calculated in mice.That is, even if 10 times more BPA were released in vivo than in vitro, that amount would still have little or no estrogenic effect.BPA was found in restorative materials such as composites and sealants, so a patient with many composite fillings or fissure sealants would be expected to have higher levels of BPA in saliva, especially if these fillings are recent.Some studies (3,8,9,15,16), started from detectable levels of BPA in saliva.Despite this, only three studies (5,14,15) included patients without composite restorations but even between authors there were discrepancies.Manoj et al. (14) found no BPA in the saliva before braces were placed, Kloukus et al. (15) found initial levels of 155 ng/l for the ethanol + water rinse group and 130 ng/l for the tap water rinse group and for Seifi et al.Kloukus et al. (15) showed that the oxygen inhibited layer composed of unreacted monomer species could dissolve in saliva or rinsing media and released intraorally.According to these authors, the removal of the oxygen inhibited layer immediately after bonding should be considered mandatory.Pumice prophylaxis of air-exposed cured surfaces has been shown to reduce BPA release in sealants and resin composites used as orthodontic retainers.
According to the study by Görükmez et al. (3), the repositioning of incorrectly placed brackets can make patients more exposed to additional residual monomers.Therefore, indirect cementation with a self-curing composite would be preferable, because, according to their study, they release less BPA than light-curing composites.In contrast Manoj et al. (14) did not find any difference.Such difference could be explained by the different storage temperatures of the samples used, since, as mentioned above, BPA is temperature sensitive.Also, the study by Seifi et al. (5) demonstrated that removal of residual adhesive by the tungsten carbide bur instead of the ultrasonic scaler can released less BPA in the saliva and is a faster method (1.008 µg./mL vs. 2.83 µg./mL).This must be considered to minimize the intake of BPA by our patients.The study by Görükmez et al. (3), concluded that the use of a chromatographic technique in tandem with mass spectrometry provided more precise and sensitive results.While gas or liquid chromatography devices alone can measure down to 1 ppm (one millionth), chromatography combined with mass spectrometry can measure down to 1 ppd (one billionth).Despite this, the studies by Raghavan et al. (8) and Watanabe et al. (13) did not use this combination of techniques.Manoj et al. (14), used high-performance liquid chromatography to avoid bias, since, although it is a relatively slower technique, high-performance liquid chromatography can be performed at room temperature, which allowed the analysis of heat-sensitive compounds like BPA, since it can be affected by changes in oral temperature over time and gas chromatography is performed at elevated temperatures.This leads us to understand that the studies that used the liquid chromatography in tandem with mass spectrometry would obtain more reliable results than those that used gas chromatography.Regarding the limitations of this systematic review, we found few in vivo studies on the subject and diversity in methodology.The comparisons between studies were difficult to perform since each study used a different unit, technique, and storage temperature.In sum, this study provides relevant information regar-ding the release of BPA of different materials: is maximum after cementing metal brackets or retainers and decreases with time; thermoplastic retainers release higher levels of BPA and for a longer time than the hawley type retainers.Removal of remaining adhesive with a tungsten carbide bur or rinsing with water after bonding can help to reduce BPA to baseline levels, although in the case of the aesthetic polycarbonate brackets the release of BPA remains active until they are removed.The amount of BPA released by these dental materials has not been clinically related to harmful effects on human health.

( 5 )
the basal values were not reflected.This difference may be due to the selection of the sample, the different storage temperatures and the different techniques used.In Manoj et al. (14) and Seifi et al. (5) studies, patients who smoked or chewed tobacco, with kidney or liver diseases and with acrylic work or resin restorations were excluded, whereas in the Kloukus et al. (15) study, the exclusion criteria included only the presence of resin or e919 sealant restorations.Aspects such as the consumption of hot drinks in plastic cups and environmental exposure to BPA should also be considered, as Manoj et al. (14) reported.

Table 2 :
Quality of clinical trials on the PEDro scale.

Table 3 :
Main characteristics of the studies.