Does Er,Cr:YSGG reduce the microleakage of restorations when used for cavity preparation? A systematic review and meta-analysis

The focused question according to the Participants, Interventions, Control and Outcomes (PICO) principle was: ‘During cavity preparation, does Er,Cr:YSGG laser result in less microleakage of the restoration in comparison to traditional burs?’ and ‘Does the application of prior acid etching increase the adhesive potential of self-etching and etch-and-rinse adhesives in laser-prepared cavities?’

A literature search was conducted in Pubmed, EBSCO, Embase, and the Cochrane Controlled Register of Trials (CENTRAL) up to July 2019. Appropriate search algorithms were developed for each database, using the following Boolean phrases: (laser) AND (microleakage OR leakages, dental OR dental leakage OR leakage, dental) AND (cavity preparation OR preparation, dental cavity OR preparations, dental cavity OR cavity preparation, dental OR dental cavity preparations). The search had no restriction in the publication language.

Two blinded, independent investigators screened the titles and abstracts from the electronic searches to find potentially eligible studies. The full texts of all the seemingly eligible studies were obtained and further evaluated in detail to make sure they really met all the selective criteria. To avoid missing eligible studies, the reference lists of all the selected full-text studies were also screened. The agreement between the reviewers was calculated in the selection procedure by Cohen’s kappa statistics [14], which yielded a value of 0.93. Any discrepancies were resolved by discussion until a consensus was reached.

The investigators selected the included studies based on the following criteria:

1. Randomized controlled trials (RCTs) or quasi-RCTs.

2. Comparison of the microleakage of cavities prepared by Er,Cr:YSGG laser versus cavities prepared by traditional burs.

3. Comparison of the microleakage of cavities prepared by laser with additional acid etching with that by laser alone.

4. Use of human teeth.

5. Drop-out rate: < 20%.

Case reports, review papers, letters to the editor, monographs, conference abstracts, and animal studies were excluded.

Two reviewers independently extracted and managed data on the characteristics of the included studies as follows: year of publication, the country of origin, study design, number of teeth, restorative materials and adhesive systems used, cavity type, cavity size, tooth type, groups, number of drop-outs, and microleakage test. The following parameters of Er,Cr:YSGG laser were also recorded: mode, manufacturer, tip, wavelength, pulse frequency, pulse duration, mean power, and method of application. Any difference was resolved by discussion. In cases in which research data were incomplete or missing, the authors were contacted to ask for further information.

Statistical analysis was performed using RevMan 5.3 software provided by the Cochrane Collaboration. Risk ratio (RR) was used, along with 95% confidence intervals (CIs), for dichotomous data, while the mean difference (MD) was used with 95% CIs for continuous data. In addition, the z-test was used. I² test on the level of α = 0.10 was used to evaluate statistical heterogeneity. When there was significant statistical heterogeneity (I² > 50%), a random-effect model was used to analyze the data. If I² was ≤50%, a fixed-effect model was used. The statistical significance for the hypothesis testing was set at α < 0.05 (2-tailed z-tests). When the data could not be pooled in the meta-analysis, they were summarized qualitatively.

Two investigators independently assessed the quality of the target studies using Cochrane Collaboration’s tool for assessing the risk of bias. The assessment for each article included seven domains, and each domain was divided into three categories: low risk of bias, unclear risk of bias, and high risk of bias. The study was judged as low risk of bias if all the domains were deemed low risk, as moderate risk if one or more were considered as unknown risk, as high risk if all were deemed high risk. The agreement between the reviewers was assessed based on Cohen’s kappa statistics, assuming κ = 0.6 to be a favorable score. Any disagreement was resolved by discussion, and a third investigator was consulted if arbitration was required.

Totally, 357 articles were retrieved initially from the databases, which decreased to 170 after removing the duplicates (Fig. 1). After screening the titles and abstracts, the full texts of 21 articles were obtained for more detailed data. Finally, 13 studies were included, 11 of which were selected for meta-analysis [4‐9, 12, 15‐18].

Table 1 presents the characteristics of the studies included. The 13 studies were published from 2001 to 2018; five studies were conducted in Iran, two in Turkey, two in Brazil, and the remaining were conducted in Germany, Spain, Brazil, and India, respectively. Among them, five studies selected primary teeth as the experimental models, while the others selected premolars, third molars, or other posterior teeth. Class V cavities accounted for 10 out of 13 studies. Table 2 presents the parameters of the lasers used.

As shown in Fig. 2, all the included studies were considered to have a moderate risk of bias (Fig. 2). Among them, only one study described the randomization methods clearly [17], and five studies referred to the blinding of outcome assessment [4‐6, 15, 16]. All the studies had unclear information about ‘allocation concealment’ and ‘blinding of wpersonnel.’

Except for two studies that did not report specific microleakage scores or indexes [19], the remaining 11 studies were selected for meta-analysis (Fig. 3). Subgroup analysis was conducted based on different measuring positions (enamel margin, dentin margin, and the whole marginal line). Six studies compared the microleakage of cavities prepared by burs with that prepared by Er,Cr:YSGG lasers on enamel and dentin margins [4, 6, 8, 9, 12, 17], while the remaining five measured microleakage on all the marginal lines [5, 7, 15, 16, 18]. The results revealed significant heterogeneity among the studies (χ² = 79.41, I² = 80%, P < 0.00001). Meta-analysis with a random model indicated that the incidence of microleakage was a little higher in traditional bur groups both on the dentin and the whole marginal line, while it did not show any significant difference (RR = 1.03, 95% CI range: 0.85–1.25, P = 0.74), (dentin margin: RR = 1.26, 95% CI range: 0.67–2.38, P = 0.47; whole marginal line: RR = 1.27, 95% CI range: 0.44–3.67, P = 0.66). However, in the enamel margin subgroup, the results revealed an insignificant increase in microleakage in the laser group (RR = 0.87, 95% CI range: 0.60–1.27, P = 0.47). The study by Rossi et al., which was excluded from the meta-analysis, also reported no statistically significant difference between the bur and Er,Cr:YSGG laser groups on the whole marginal line [19]. However, Geraldo-Martins et al. reported that the Er,Cr:YSGG group had a higher microleakage index compared to the traditional bur group [20].

Four out of 13 studies measured the effect of acid etching on the adhesive potential of self-etch and etch-and-rinse adhesives after laser preparation [7, 9, 12, 16]. Two studies used acid etching with self-etching adhesives and evaluated the microleakage value on enamel and dentin margins, respectively [9, 12], while the remaining two applied etch-and-rinse adhesives and measured the value on the whole marginal lines [7, 16]. It was reported that prior acid etching improved the adhesive potential of self-etching adhesives and decreased microleakage after laser preparations significantly (χ² = 1.28, I² = 0%, P = 0.73, RR = 2.69, 95% CI range: 1.74–4.15, P < 0.00001). The significant difference was detected both in the enamel and dentin margin subgroups (enamel margin: RR = 3.0, 95% CI range: 1.54–5.83, P = 0.001; dentin margin: RR = 2.44, 95% CI range: 1.38–4.34, P = 0.002) (Fig. 4). However, the difference was not significant for the etch-and-rinse adhesives (χ² = 5.20, I² = 81%, P = 0.02, RR = 1.18, 95% CI range: 0.63–2.22, P = 0.60) (Fig. 5).

Figure 2 presents the quality of the included studies. All the studies were considered to have a moderate risk of bias. Unclear information about ‘random sequence generation,’ ‘blinding of personnel’ and ‘allocation concealment’ were the main risk factors. Although all the studies were conducted randomly, only one study mentioned the randomization method adopted in the experiment. Computer programs, random tables, or other randomization methods are necessary to balance the assignment to experimental and control groups. In addition, the importance of allocation concealment and blinding techniques must be emphasized. Improper blinding and allocation concealment might result in the overestimation of the effect of experimental treatments, causing bias. The quality of the included studies was not favorable, possibly decreasing the reliability of conclusions drawn in the present study.

With advances in dental materials, composite resin restorations have replaced amalgam restorations gradually. However, polymerization shrinkage, the major disadvantage of composite materials, can give rise to a marginal gap at the tooth–restoration interface, affecting the long-term success of restorations [10]. The phenomenon during which bacteria, liquids, molecules, or ions pass through the marginal gaps is known as microleakage and is often regarded as the most important factor resulting in secondary caries and pulpal infection [8, 21]. From this viewpoint, many suggestions have been made to reduce microleakage, such as the use of low-shrinkage resins and adequate preparation of the tooth [22]. Currently, the most common method for cavity preparation is the use of burs. However, the generated heat and pressure may cause pain and pulpal damage during cavity preparation if there is no adequate refrigeration, always posing a challenge for doctors and patients. For years, numerous new techniques have been developed as alternatives to traditional burs, of which the application of erbium family lasers has won broader and wider acceptance. Owing to its specific mechanism, Er,Cr:YSGG laser can cut enamel and dentin effectively, with less vibration, sharp noise, and pain for patients during cavity preparation. Also, adverse thermal effects on the pulp and surrounding tissues can be prevented effectively with the use of a water mist spray [23]. However, it should be noted that the parameters of lasers, such as repetition rate, air-water ratio, and mean power, are very critical as overheating can cause not only pulp damage but also undesirable morphological changes, such as cracks, carbonizations, etc., resulting in pain and irreversible damage. Some investigations have indicated that the Er,Cr:YSGG preparation improved the bonding process of adhesives and reduced microleakage [8, 24]. However, there is continuing controversy over the effect of Er,Cr:YSGG preparation on microleakage. Some studies found no difference between cavities prepared by bur and Er,Cr:YSGG laser [5, 15, 25], with some even reporting that Er,Cr:YSGG laser resulted in higher microleakage scores [24]. These reports have not been evaluated systematically to date.

Generally, microleakage tests can be conducted in vitro and in vivo; however, in vitro studies are more common owing to their precise and easy procedural steps [4]. For microleakage tests, several techniques have been widely used, including dye penetration, scanning electron microscopy, chemical tracers, air pressure, and neutron activation [5], with dye penetration being the most common technique for its ease of implementation, low cost, and safety [26]. Therefore, it is not surprising that all the studies included in our review were in vitro studies and used dye penetration for evaluation of microleakage.

As shown in our meta-analysis, there was no significant difference between Er,Cr:YSGG and bur preparations in terms of microleakage. Similarly, the same conclusion has been reached in a study by Rossi et al. With high absorbance in water and hydroxyapatite, the Er,Cr:YSGG laser beams can heat the water content of dental hard tissues, causing a micro-explosion of water particles [6]. The process of laser ablation produces an irregular, rough, and moist dental surface with exposed dentinal tubules, intact enamel rods, and no smear layer [5]. Considering these microscopic changes, some studies reported that Er,Cr:YSGG preparations might be more suitable for adhesion of restorative materials [8, 27]. However, some studies have indicated that the enamel melting, minimal cracking, and acid-resistant surfaces created by Er,Cr:YSGG irradiation during preparation might have adverse effects on the bonding process of adhesives, especially with ultra-mild self-etch adhesives [28‐30].

Additionally, the ablation of dentin can fuse the collagen fibrils and reduce interfibrillar spaces, thus limiting the penetration of adhesives and resins [31]. Such unfavorable marginal sealing in Er,Cr:YSGG preparation has been observed by Geraldo-Martins et al. They measured the total length of the tooth–restoration interface and the interface infiltrated by dye to calculate the infiltration index [20]. The results indicated that Er,Cr:YSGG group had a higher microleakage index than the traditional bur group. The inconsistent conclusion drawn by Geraldo-Martins et al. might be explained by the fact that they used carious teeth as a study model and showed that laser could not remove the carious tissue completely, possibly affecting the marginal sealing of restorations. Under clinical conditions, bonding to dentin is more challenging because of its higher water and organic matter content. However, as shown in the present study, no significant difference was observed between the cavities prepared by lasers and those by burs in terms of microleakage rate, either on enamel or dentin margins.

It should be noted that there was significant heterogeneity among the included studies, and the conclusions reached in the present study should be interpreted cautiously. This heterogeneity can be explained by the variability in tooth type, cavity type, restoration materials, adhesives, and the irradiation parameters. For example, some studies used conventional methacrylate-based microhybrid composite resins for restorations, while some used low-shrinkage composite resin systems, which might have affected the results, accounting for the inconsistent conclusions [4, 9].

Dental adhesives can be classified into two categories based on the way they react with the smear layer [32]. Etch-and-rinse adhesives need a prior acid etching procedure, while self-etch adhesives contain acidic functional monomers that can remove the smear layer selectively. In recent years, there has been a growing debate on the use of acid etching after laser preparation. Previous studies observed that prior acid etching improved the bond strength of self-etch and etch-and-rinse adhesives following laser preparation [33, 34]. However, some authors reported that acid etching did not affect the adhesive procedure in laser-prepared cavities, even weakening the bonding efficacy of self-etch adhesives [35, 36]. Based on these studies, acid etching might decrease the hydroxyapatite content of dental tissue, thereby weakening the chemical bonding of adhesives [12, 37]. In the present study, self-etch adhesives, in combination with prior acid etching, exhibited less microleakage than those without acid etching on enamel and dentin margins; however, the results were insignificant for etch-and-rinse adhesives. The following explanations might account for this result. According to many studies, a lack of water in dental tissue after laser preparation might affect the etching capacity of self-etching adhesives, thus limiting its penetration through the laser-modified layer. While additional acid etching can remove this layer, increase surface wettability, and benefit the hybrid layer formation, all of which can help reduce microleakage in laser-prepared cavities [12]. As reported by Obeidi et al., the etching time should be prolonged to 40 s, but not to 60 s, to improve the bond strength of self-etch adhesives in Er,Cr:YSGG-prepared teeth [32].

This systematic review and meta-analysis had some limitations. Firstly, the effect of different restorative materials on microleakage was not fully considered. With inconsistent shrinkage rates, different resin-based systems might interfere with the results of further microleakage evaluation and cause bias.

Secondly, selecting an effective adhesive is also crucial for restorations. The applied adhesive systems varied widely among studies, including different self-etch and etch-and-rinse adhesives. Some studies showed that self-etch and etch-and-rinse adhesives had similar microleakage values in Er,Cr:YSGG laser preparations. However, the impact of different adhesives on microleakage should be further elucidated. Moreover, the differences in tooth characteristics, cavity type, and irradiation parameters might also affect microleakage. For example, although some studies showed that the morphology of primary teeth prepared by Er:YAG laser was similar to that of permanent teeth, the primary teeth often had a higher degree of mineralization, which might result in different reactions to laser beams.

Geraldo-Martins et al. indicated that it is essential to use carious teeth rather than healthy teeth as study models for the in vitro test of the microleakage. However, only one out of the 13 included studies used carious teeth to prepare cavities. Further studies are needed to verify the authenticity of healthy teeth as the research model in microleakage tests. The laser parameters, such as spot size, emission model, pulse duration, cooling rate, mean power, energy density, and focus, are vital for the application of lasers [23]. In addition, the chemical composition of the dental tissue influenced its efficacy. For example, the effect of laser on enamel was different from those observed in dentin. However, the laser parameters and the height of cavities varied markedly among the included studies, and some studies even lacked a detailed description of the information.

Considering the limitations mentioned above, the following suggestions are proposed for future research. Firstly, more RCTs should be conducted strictly according to the Cochrane’s criteria for the risk of bias. Secondly, researchers should report more data on the laser parameters applied in the study, and it is useful to explore the best parameters of Er,Cr:YSGG laser for cavity preparation. Thirdly, further investigations should focus on the long-term effects of laser preparation on restorations. Lastly, it is necessary to demonstrate whether it is possible to use sound teeth as study models to prepare cavities and test the microleakage of restorations.