Combined Tin-Containing Fluoride Solution and CO 2 Laser Treatment Reduces Enamel Erosion in vitro

Transcription

1 Original Paper Received: April 9, 2015 Accepted after revision: August 10, 2015 Published online: September 30, 2015 Combined Tin-Containing Fluoride Solution and CO 2 Laser Treatment Reduces Enamel Erosion in vitro Marcella Esteves-Oliveira a Nadine Witulski a Ralf-Dieter Hilgers b Christian Apel a, c Hendrik Meyer-Lueckel a Carlos de Paula Eduardo d Departments of a Operative Dentistry, Periodontology and Preventive Dentistry and b Medical Statistics, and c Department of Tissue Engineering and Textile Implants, Applied Medical Engineering, Helmholtz Institute for Biomedical Engineering, RWTH Aachen University, Aachen, Germany; d Department of Restorative Dentistry, School of Dentistry, University of São Paulo, São Paulo, Brazil Key Words Demineralization Enamel Erosion Fluoride Laser Prevention Tin Wear Abstract The aim of this in vitro study was to evaluate the effect of combined CO 2 laser and tin-containing fluoride treatment on the formation and progression of enamel erosive lesions. Ninety-six human enamel samples were obtained, stored in thymol solution and, after surface polishing, randomly divided into 6 different surface treatment groups (n = 16 in each group) as follows: no treatment, control (C); one CO 2 laser irradiation (L1); two CO 2 laser irradiations (L2); daily application of fluoride solution (F); combined daily fluoride solution + one CO 2 laser irradiation (L1F), and combined daily fluoride solution + two CO 2 laser irradiations (L2F). Laser irradiation was performed at 0.3 J/cm 2 (5 μs/226 Hz/10.6 μm) on day 1 (L1) and day 6 (L2). The fluoride solution contained AmF/NaF (500 ppm F), and SnCl 2 (800 ppm Sn) at ph 4.5. After surface treatment the samples were submitted to an erosive cycling over 10 days, including immersion in citric acid (2 min/0.05 M /ph = 2.3) 6 times daily and storage in remineralization solution ( 1 h) between erosive attacks. At the end of each cycling day, the enamel surface loss (micrometers) was measured using a 3D laser profilometer. Data were statistically analyzed by means of a 2-level mixed effects model and linear contrasts (α = 0.05). Group F ( 3.3 ± 2.0 μm) showed significantly lower enamel surface loss than groups C ( ± 4.1 μm), L1 ( 18.3 ± 4.4 μm) and L2 ( 16.3 ± 5.3 μm) but higher than L1F ( 1.0 ± 4.4 μm) and L2F (1.4 ± 3.2 μm, p < 0.05). Under the conditions of this in vitro study, the tin-containing fluoride solution caused 88% reduction of enamel surface loss, while its combination with CO 2 laser irradiation at 0.3 J/cm 2 hampered erosive loss almost completely S. Karger AG, Basel Polyvalent metal fluoride compounds like TiF 4 and SnF 2 have been shown to be highly effective in preventing not only enamel erosive surface loss but also chemical-mechanical erosive wear [Ganss et al., 2008; Hove et al., 2008; Stenhagen et al., 2013; Wiegand et al., 2014]. M. Esteves-Oliveira, N. Witulski and R.-D. Hilgers contributed equally to this work. karger@karger.com S. Karger AG, Basel /15/ $39.50/0 Priv.-Doz. Dr. Marcella Esteves-Oliveira Department of Operative Dentistry, Periodontology and Preventive Dentistry (ZPP) RWTH Aachen University, Pauwelsstrasse 30 DE Aachen (Germany) ukaachen.de

2 For the application as solutions most of the evidence currently available indicates that the greatest effectiveness is obtained using Sn-containing fluoride compounds [Huysmans et al., 2014]. In situ studies show a 67 and 94% reduction of enamel surface loss, respectively for AmF/NaF/SnCl (ph = 4.5) and for SnF 2 (ph = 2.5) solutions [Ganss et al., 2010; Stenhagen et al., 2013]. In contrast to NaF, which acts primarily by the formation of CaF 2 -like reservoirs [Rolla and Saxegaard, 1990], releasing fluoride ions when the ph drops to acidic values at the tooth-saliva interface, TiF 4 causes the formation of a glaze-like surface coating on enamel [Wiegand et al., 2009] and Sn-containing products cause the deposition of an amorphous Sn-rich layer over the surface, as well as the incorporation of Sn into enamel [Wei and Forbes, 1974; Schlueter et al., 2009b]. In contrast to SnF 2 solutions which cause a high reduction of enamel surface loss alone, with fluoride solutions containing stannous chloride both main active ions (fluoride and tin) seem to play an important role. In a recent study pure solutions of either component (AmF/NaF or SnCl 2 ) caused only moderate erosion reduction [Ganss et al., 2008]. For solutions containing 500 1,500 ppm fluoride as AmF or NaF, increasing the Sn content from 800 to 2,800 ppm significantly increased the antierosive effect. However, there are side effects in clinical use such as the dull sensation on the teeth. Thus, recently an effective combination between an antierosive effect and control of clinical side effects was found with solutions containing 800 ppm Sn (as SnCl 2 ) and 500 ppm F (as AmF/NaF) [Schlueter et al., 2009c]. As NaF formulations (ph >4) often only offer mild protection against dental erosion [Hove et al., 2008; Stenhagen et al., 2013; Souza et al., 2014] or require an intensive regime [Ganss et al., 2004], and considering that currently no long-term clinical trial on the effectiveness of antierosive compounds is available, the Sn-containing fluoride products appear to be the best option for symptomatic therapy of patients at high risk for erosion [Ganss et al., 2010; Huysmans et al., 2014]. In the last 30 years several investigations showing the thermal, chemical and acid solubility resistance effect of CO 2 laser irradiation on dental enamel have been published [Fox et al., 1992; Fried et al., 1996; Featherstone et al., 1998a]. Very good acid solubility resistance results were shown with pulses of 100 μs [Featherstone et al., 1998a], but it was also observed that pulses of this length combined with relative high energy densities (5 10 J/ cm 2 ) tended to cause surface microcracks [McCormack et al., 1995]. In further investigations the same research group demonstrated that by shortening the pulse duration, less energy is necessary to create less soluble enamel (by elimination of carbonate impurity of the hydroxyapatite) [Fried et al., 1999], while reducing the chances of surface damage. So, specifically using the 9.6-μm wavelength, it was demonstrated that, using a low-energy density (1 J/cm 2 ) and reducing the pulse duration from 20 to 8 μs twice, as much inhibition of enamel dissolution could be obtained [Gerard et al., 2005]. Moreover, even in a short-term clinical trial 87% inhibition of enamel demineralization has been shown [Rechmann et al., 2011]. Following these principles and considering that the 10.6-μm wavelength penetrates 10 times deeper in enamel than the 9.6 μm, we found a specific combination of laser parameters for this alternative wavelength that indeed also caused a very high reduction of enamel demineralization (81%), while causing neither damage to the surface or excessive pulp chamber temperature increase [Esteves-Oliveira et al., 2009]. This specific set of parameters have been recently studied in short-term erosion and abrasion models [Esteves-Oliveira et al., 2011a] also showing a high reduction of enamel dissolution and a rehardening of previously softened enamel [Esteves-Oliveira et al., 2011b]. Nevertheless, it is still not yet clear how long-lasting the effect of using these laser parameters on the increase in enamel resistance to dissolution is or how this technique would interact with the most successful strategies for preventing enamel erosion like the use of tin-containing fluoride products. Therefore, the primary aim of this in vitro study was to examine the effect of CO 2 laser treatment combined with a tin-containing fluoride solution on the formation and progression of enamel erosive tissue loss over 10 days in vitro. A secondary aim was to investigate whether a second laser irradiation after 5 days would enhance its protective effect. Finally, the null hypotheses formulated was that the mean reductions of enamel surface loss (measured by profilometry) caused by the following: (1) tincontaining fluoride treatment alone, (2) CO 2 laser irradiation alone applied once or twice or (3) a combined fluoride-laser treatment do not differ from each other after 10 days. Materials and Methods Sample Preparation Ethical approval was obtained from the Ethics in Research Committee of the Medical Faculty of RWTH Aachen University (EK093/09). Ninety-six enamel specimens were prepared from 566 Esteves-Oliveira et al.

3 sound human third molars collected in German dental offices (fluoride concentration in drinking water <0.2 mg/l). The teeth were stored in 0.1% thymol solution and after careful cleaning with scaler and rotating brushes the smooth surfaces were examined by means of a stereomicroscope (magnification 4) and only those presenting no cracks or structural defects were included in the study. The teeth were cut into halves with a diamond saw (Exakt, Norderstedt, Germany) and embedded in blocks with a light curing resin (Technovit 7200 VLC; Heraeus Kulzer, Wehrheim, Germany). The embedded samples were polished using P800, P1200, P2400, and P4000 silicon carbide grinding papers (Struers, Willich, Germany) on a rotating polishing machine (Exakt) and immersed for 30 s in a sonication bath. A standard removal of 400 ± 40 μm [Carvalho and Lussi, 2015] was controlled with a digital micrometer (Mitutoyo Deutschland, Neuss, Germany). In the center of the polished surface, a round tape of 2.5 mm in diameter was used to delineate the experimental area with an acid-resistant varnish (Essence, Sulzbach, Germany). Outside of this circle the surface was covered with a removable acrylic-coated cloth tape (4651 black; Tesa, Hamburg, Germany) in order to preserve a reference area for the profilometric analysis. This procedure was tested before the beginning of the study and proved not to cause any significant surface loss at the reference area. Subsequently, each of the hemisections were fixed on custom-made plastic slides (dimensions 8 2 cm) to enable the later attachment to a rack, allowing simultaneous insertion of all samples into falcon tubes (2 samples/ tube, approx. 4.6 ml solution/mm 2 of exposed enamel). Samples were stored in 100% humidity at all times. Sample Size A pilot study was conducted to decide on the number of line scans, positions and lesion sites within a reduced setting of only 2 groups (negative control and fluoride 1, n = 3). Surface loss measurements were completed with 3 line scans, 6 positions and at 3 lesion sites (left, middle, right) as previously described [Esteves-Oliveira et al., 2011a]. The variation calculated by fitting a variance component model, including the main effects only to the data, was about 0.08 between the 3 lines, 0.99 between the 6 positions, between the 3 sites, 3.35 between the 3 samples, and between the 10 days, with a measurement error of This suggested that measuring 1 single line on 1 sample is sufficient since the variation between the lines and positions was rather small and thus the number of replications could be kept small. However, the variation between sites suggests a greater number of replications here. To establish a difference between the mean erosion of the control group (32.27 ± 2.46 μm) and the fluoride group (30.90 ± 4.33 μm) at the end of our 10-day observation period, 105 samples per group would be necessary (2-group Satterthwaite t test of equal means, unequal variances, 5% significance level, 80% power). Because such a high sample size was related to the rather low effect caused by a nonfrequent fluoride application (only once), we decided to modify our study design by increasing the number of fluoride applications from once at the start of the experiments to a daily application of the fluoride solution, which is also closer to a clinical situation. From the clinical point of view, obviously a rather large effect (e.g. mean difference between two treatments divided by the standard deviation, SD = 1) should be established in an in vitro study to ensure translation to patients. This resulted in a sample size of 16 per group (2-group t test of equal means, variances, 5% significance level, 80% power). Groups Samples were randomly allocated to 6 groups (n = 16) receiving different surface treatments. Group C served as a negative control and received no pretreatment; group F served as a positive control and received daily application of fluoride solution; group L1 was laser irradiated once before the cycling; group L2 was laser irradiated twice (the first time before the cycling and the second time before the sixth cycling day); group L1F was fluoridated daily and laser irradiated once before the cycling, and group L2F was fluoridated daily and laser irradiated twice (before the cycling and before the sixth cycling day). For the combined laser-fluoride treatments (L1F and L2F) the first fluoride application was completed before laser irradiation. Fluoride Treatment Fluoride application was performed using an antierosive solution (ph = 4.5) containing 500 ppm of free fluoride as amine fluoride (125 ppm F) and sodium fluoride (375 ppm F) as well as 800 ppm of tin as stannous chloride (Elmex Erosionsschutz; Gaba, Lörrach, Germany). Specimens were immersed at the beginning of each cycling day in this solution for 30 s under agitation on an orbital shaker (Orbit; Labnet, Edison, N.J., USA). Laser Treatment For laser irradiation a CO 2 laser (Rofin SCx 30; Rofin-Sinar Laser, Hamburg, Germany) was used. Irradiation was performed at a wavelength of 10.6 μm, pulse duration of 5 μs, repetition rate of 226 Hz, energy density of 0.3 J/cm 2, and without water cooling, as described previously [Esteves-Oliveira et al., 2009]. In order to allow adequate determination of the energy densities, and considering that the laser beam had Gaussian distribution and radial symmetry, the beam diameter at 1/e 2 of the intensity level was determined using the knife-edge method. The emitted energy was controlled using an energy meter (Coherent Field Master GS + Detector LM45; Coherent, Santa Clara, Calif., USA). Irradiation was performed at a distance of 19.8 cm, and the beam diameter at the sample surface was 2.5 mm. The total number of overlapped pulses applied was 2,036 for the irradiation during 9 s. Erosive Cycling All specimens were demineralized 6 times daily for 2 min with 0.05 M citric acid (Merck) at ph 2.3, as described previously [Schlueter et al., 2009a]. Between demineralization periods there was an interval of at least 1 h in which the specimens were stored in a supersaturated mineral solution (4.08 m M H 3 PO 4, m M KCl, m M Na 2 CO 3, and 1.98 m M CaCl 2, ph 6.5; chemicals from Merck) [Gerrard and Winter, 1986]. By means of a special rack, the specimens of all groups could be moved simultaneously in and out of the falcon tubes each containing 45 ml of the de- and remineralization solutions. Each rack supported 8 falcon tubes, and a sample holder (plastic slide) for 2 samples was fixed in each tube. Before transfer to the next solution, the specimens were rinsed with tap water. During the whole ph cycling period the samples were stored at room temperature and under agitation on an orbital shaker at 50 rpm overnight; the specimens were kept in the remineralization solution and all solutions were renewed daily. Profilometric Measurement The enamel surface profiles were measured after the first treatment but before beginning the erosive cycling (first treatment), as well as daily after cycling, using a 3D laser profilometer (VK-X100, CO 2 Laser and Tin Fluoride Reduction of Erosion 567

4 20 objective; Keyence, Neu-Isenburg, Germany). Before the measurements, the tape was removed from the reference areas and replaced afterwards. After tape removal, the reference areas were carefully cleaned with cotton pellets and checked under the laser microscope to make sure that they were free of any adhesive remnants. In order to assure that the measurement of the enamel surface loss was at the same position as at the start of the experiment, two marks were engraved on the reference area of each sample using a fissure bur (Hager & Meisinger, Neuss, Germany) one at the left and one at the right side of the experimental area. For each specimen a rectangle 4, μm wide was scanned, beginning in the reference area at one side, then going through the experimental area and ending in the reference area on the other side. With the corresponding analysis software (VK Analysis Application, Keyence) the scanned surface was adjusted so that the reference areas of both sides were exactly at the same height. After that, a horizontal line scan connecting the two bur grooves and going through the center of the lesion was used to conduct the measurements. Along this line scan the average height difference between the reference areas and the experimental area was determined. Each day, the line was relocated at the same position to ensure the correct measurement of any height alteration in the experimental area. Repeated measurements (n = 5) of 1 randomly selected specimen gave an SD of 0.1 μm and the maximum height resolution of the optical profilometer was 5 nm. Scanning Electron Microscopy In order to verify the effects obtained by profilometry and observe the effect of cycling and of the different treatments on the enamel surface and subsurface, morphological investigations were performed. After the erosive cycling 3 samples of each group were randomly selected and sectioned with a band saw through the center of the lesion in order to allow a cross-sectional view. The samples were then serially dehydrated in alcohol, immersed in hexamethyldisilazane (HMDS) for further dehydration, covered with a thin gold layer and examined under an environmental scanning electron microscope (SEM; ESEM XL30 Field Emission Gun; Phillips, Eindhoven, the Netherlands). The images were obtained using a GSED (gaseous secondary electron detector) detector with the sample s chamber pressure at around 1 mbar, using an accelerating voltage of 10 kv. Statistical Analysis The differences in enamel surface loss were described by means and SD for each variable (laser, fluoride and day). The groups L1F and L2F, as well as L1 and L2, were pooled in the analysis up to day 5. For statistical analysis a 2-level mixed effects model was fitted to the data, with samples as level 2 variable and day as level 1. The treatment variables laser (3 levels: no/1 /2 ) and fluoride (2 levels: no/yes) were covariates of level 2. We modelled the day variable (days 1 10) as random effect with an unstructured covariance structure for intercept and linear slope. All interaction terms up to order 3 were included in the model, as well as the experimental setting factors sample holder (2 levels) and rack (8 levels). Specific questions were analyzed by means of linear contrasts. To clarify, the results were given by degrees of freedom for nominator (NDF), denominator (DDF), F or t value, and p value. The significance level was set to 5%. SAS (version 9.2, TS2M3, 64 bit for Windows 7; SAS Institute Inc., Cary, N.C., USA) statistical software (PROC MIXED) was used for computations. Table 1. Mean rate of enamel surface loss/day and standard error for all groups over all experimental days Rate of enamel surface loss Enamel surface loss at day 10 mean (SE), μm/day Results 95% CI m ean ± SD, μm C 1.66 (0.05) 1.75, ± 4.1 F 0.35 (0.05) 0.44, ± L (0.04) 1.39, ± L (0.07) 1.39, ± L1F 0.38 (0.06) 0.49, ± L2F 0.22 (0.09) 0.40, ± % reduction relative to negative control Additionally, for day 10 the means and SD of enamel mineral loss with the respective % reduction in relation to negative control are described. SE = Standard error. Profilometric Measurements During the model fit, 13 of 1,056 observations showed a marked influence on the fitting marker restricted likelihood distance. Usually, in mixed model fitting, if the restricted likelihood distance shows outliers, fitting of the model is questionable and there is the recommendation in the literature to sequentially delete the observations by which the largest outliers occurs. Thus the samples which had values of restricted likelihood distance almost 2 times higher than the next following value were deleted. Excluding these samples also results in fairly good normal distribution of the residuals and thus shows a reasonable model fit. Modelling the correlation matrix for the time effect leads by an autoregressive model to a further improvement of the model fit. There were significant differences in mean enamel loss between the day laser fluoride group interactions (NDF 2, DDF 850, F = 10.07, p < ), as well as for day fluoride groups (NDF 1, DDF 850, F = 1,033.1, p < ) and for day laser groups (NDF 2, DDF 850, F = 17.08, p < ). Moreover, a significant main effect of day (NDF 1, DDF 850, F = 1,841.38, p < ) and laser use (NDF 2, DDF 20, F = 15.98, p < ) was shown. The highest enamel surface loss at all days ( fig. 1 ) and the highest average rate of surface loss per day were observed in the negative control group ( table 1 ). This group presented statistically significant lower means than all 568 Esteves-Oliveira et al.

5 15 10 C L1 L2 L1F L2F F Color version available online 5 E E 0 Enamel surface loss (μm) B C D First treatment Day 1 Day 2 Day 3 Day 4 Day 5 Day 6 Day 7 Day 8 Day 9 Day 10 A Fig. 1. Mean enamel surface height after first treatment and mean enamel surface loss after every cycling day in the 6 different surface treatment groups: C, F, L1, L2, L1F, and L2F. Different letters indicate statistically significant differences between the groups at the end of the experiments (day 10). other groups (all p < ). Daily fluoride application ( 3.3 ± 2.0 μm) led to statistically significant lower surface loss per day compared to both groups with laser irradiation alone (L1 = 18.3 ± 4.4 μm, L2 = 16.3 ± 5.3 μm, both p < ). Until day 3, though, erosion reduction caused by one laser irradiation ( 1.2 ± 3.9 μm) and three applications of fluoride solution ( 1.1 ± 3.2 μm) were in the same range ( fig. 1 ). Nevertheless, the highest erosion reduction over the whole experiment was observed by the groups combining fluoride application and laser irradiation (L1F and L2F), which was even significantly higher than that of the positive control (both p < ). In percentage terms, the tin-containing fluoride solution caused 88% reduction of mean enamel surface loss, while its combination with CO 2 laser irradiation hampered erosion almost completely (104 and 106% reduction; table 1 ). In this case of combined treatment (L1F and L2F), laser irradiating the enamel surface once or twice did not cause any statistically significant difference in daily enamel loss (p = ). Immediately after the first treatments the average enamel surface height in group C was 1.19 ± 0.7 μm, which was thus lower than the other groups (L1: 4.45 ± 3.4 μm, L2: 5.67 ± 4.75 μm, L1F: 3.85 ± 3.19 μm, and L2F: 4.81 ± 4.99 μm). SEM Analysis The cross-sectional morphological analyses revealed lesion formation with clear borders for the groups C and F ( fig. 2 ). In the laser groups lesion borders were not so clear, especially in groups L1F und L2F, where the whole lesions were difficult to identify. Enamel surface loss in these groups was often only recognizable at the borders and almost invisible in the center of the treated area (fig. 2 ). CO 2 Laser and Tin Fluoride Reduction of Erosion 569

6 Fig. 2. Cross-section SEM images of the center of the lesions after 10 days of erosive cycling in the 6 different surface treatment groups: C, F, L1, L2, L1F, and L2F. 570 Esteves-Oliveira et al.

7 Discussion A wide variety of models have been described to simulate erosion in vitro [Wiegand and Attin, 2011]. In the present study we chose one simulating frequent acid attack that might happen in persons at high risk for erosion [Schlueter et al., 2009a], who may consume acidic drinks, fruits and food 4 or more times per day, while being exposed to other risk factors [Lussi and Hellwig, 2014]. In addition, it was also intended to use a study design that allowed the evaluation of the treatment effect over time, since we have previously shown the preventive effect of CO 2 laser irradiation against erosion, but in a rather short-term model simulating early erosion lesions [Esteves-Oliveira et al., 2011b]. Thus it was still not clear how long such an effect would last. However, one disadvantage of the design used here is that it did not simulate the protective effect of the pellicle, which acts as a diffusion barrier for the acids [Hannig et al., 2004]. In this way we probably had a higher rate of enamel surface loss than in a clinical situation, but having fewer variables influencing the results it was possible to isolate the effect of the laser irradiation on mineral dissolution. Furthermore, the differences between the groups observed here were also probably not coming from differences between the substrate conditions, as care was taken to have all specimens polished similarly and to expose enamel at approximately the same depth, since this can also influence erosion susceptibility [Carvalho and Lussi, 2015]. To the best of our knowledge, this is the first time that the CO 2 laser irradiation (λ = 10.6 μm) at 0.3 J/cm 2 and 5 μs has been combined with the application of a tin-containing fluoride solution. The erosion reduction caused by the fluoride solution alone (SnCl/AmF/NaF) compared to no treatment was 88%, which is very similar to the values currently published in the literature. When tincontaining fluoride solutions with the same ph and concentration as used here were tested, erosion reduction varied from 65 to 67% [Schlueter et al., 2009a; Ganss et al., 2010]. However, when either higher concentrations or lower ph, which can be more related to side effects, were used, higher reductions of around 80 94% have been reported [Schlueter et al., 2009a; Stenhagen et al., 2013]. This represents a very high reduction, but in the present study the combination of this solution with the CO 2 laser irradiation prevented erosion almost completely. On average, it resulted in a reduction of % compared to the negative control, thus meaning that, for several samples, there was not only little or no erosion lesion formation but most probably a build-up of material or formation of a coating layer, seen as an increase in surface height [Stenhagen et al., 2013]. In fact, the rates of erosion reduction caused by the daily use of tin-containing fluoride solution alone are already very high (65 94%), even under severe erosive conditions [Hove et al., 2008; Schlueter et al., 2009c; Ganss et al., 2010, 2011]. This being the case, one must ask if it is really meaningful or necessary to go from 88 to 100% reduction. Home use of the fluoride solution is very easy, while combination with the laser treatment requires a professional application and has higher costs. From a preventive perspective it is possible to imagine that this will certainly not make sense for all patients but may do so for the minority of the population at high risk for erosion. Such a treatment could be of special interest, for example, for individuals who for any reason are not able (or will need a long time) to control the intrinsic or extrinsic causes of erosion such as patients who have bulimia, anorexia or are resistant to behavioral changes and have a low salivary flow rate. The protective effect of Sn-containing solutions has been related to the following two mechanisms: (1) the incorporation of tin into the enamel mineral and (2) the formation of an approximately 500-nm-thick tin-containing coating layer over the enamel surface [Schlueter et al., 2009b]. The latter is thought to prevent or delay the contact of the acids with the enamel mineral [Huysmans et al., 2011]. In the present study on the first experimental day (day 1), increased surface height, possibly indicating some deposition of material over the enamel surface, could indeed be observed in all groups where the tin fluoride solution or laser irradiation was applied. However, over time this layer seemed to disappear in the fluoride group but continued to be detected until the end of the experiments for the groups with combined laser and fluoride treatment. It is likely that the formation of this layer over the surface may have been responsible for the very high erosion resistance observed in these groups. As this was an unexpected effect in this study, the mineral composition of these deposits was not analyzed and its exact constitution is unknown. However, considering that the application of the tin-containing fluoride solution alone normally leads to the deposition of a tin-containing metal layer, it can be speculated that the laser irradiation may have caused a better fixation of this coating through a melting process. During CO 2 laser irradiation at 10.6 μm the temperature increase in the first 1 5 μm of enamel surface can be as high as 900 1,000 C for some fractions of a second [Zuerlein et al., 1999]. Considering that tin melts at temperatures between 170 and 232 C [Jo et al., CO 2 Laser and Tin Fluoride Reduction of Erosion 571

8 2011], it is highly probable that tin melting has occurred here. However, this must be clarified in future studies. For the increase in surface height observed in the groups solely laser irradiated, it is difficult to find an explanation. Since no extra material was applied to the surface it may possibly only indicate a nontransient increase in enamel volume. Though such an effect is imaginable and there is a report on the increase of hydroxyapatite crystal size in dentin after CO 2 laser irradiation [Kantola, 1973], as far as we know such a volumetric expansion has never been reported before and will be further investigated in our future studies. A build-up or positive values of surface height after fluoride treatment has been already observed in other studies and has been related to the formation of a coating layer over enamel [Hove et al., 2008; Stenhagen et al., 2013]. Although this is a real possibility, if it was the case here it has yet to be confirmed. The use of scanning electron microscopy of cross-sectioned samples is definitely necessary to better characterize and measure the thickness of this layer formed on the enamel surface and such an investigation is already planned for a better understanding of the present results. Our profilometric measurements only allow a rough estimation of the thickness of these materials by subtracting the average value of the control group. According to this, the thicknesses were around 2.67 and 3.63 μm, respectively, for LF1 and LF2. Regarding the homogeneity of this deposited layer it is not possible to drawn any conclusion, as the measurements were always carried out in the same position. Another issue that still has to be clarified is how brushing resistant this coating layer is. The deposited layer has indeed promoted a very high protective effect against erosion alone, but it is not clear how it would behave under combined abrasive and erosive wear. It has been recently demonstrated that toothbrushing clearly removes and alters precipitates formed after the application of TiF 4, AmF and SnF 2 solutions [Wiegand et al., 2014] and decreases the protective efficacy of tin-containing toothpastes [Ganss et al., 2012]. It may also be that after laser treatment these deposits become more mechanically resistant and stable under toothbrushing conditions. This issue must also be clarified in future studies. In a clinical situation tooth surfaces are covered by a salivary pellicle, which was not simulated here [Hannig et al., 2004]. Therefore, it is not known if it could prevent or reduce the effect of the fluoride solution and the laser irradiation. Regarding the tin fluoride, one study showed that natural pellicle formed in situ did not result in any negative influence on its protective effect [Hove et al., 2008]. Also, when the pellicle was simulated in an in vitro study significantly less erosive wear for Sn-containing toothpastes was observed [Hara et al., 2013]. Regarding the CO 2 laser irradiation, it is also probable that it would not negatively influence the effects, since pellicles have a high protein content (proteins, peptides, lipids, and other macromolecules) and these components show a very poor absorption at 10.6 μm, being almost transparent [Fried et al., 1999; Zuerlein et al., 1999]. In the present study we chose to work with stannous chloride as the source of Sn instead of SnF 2. High enamel erosion reductions of about 91 94% for solutions containing SnF 2 have been shown in some studies [Hove et al., 2008; Stenhagen et al., 2013]. However, these SnF 2 solutions also present a much lower ph and as it is not yet clear if they could cause side effects clinically, we preferred to use an Sn-containing fluoride solution which is already readily available. The set of laser parameters tested in the present study were obtained after systematic investigation of several parameters, including laser pulses of short duration (5 μs), low fluence (0.3 J/ cm 2 ) and a high number of overlapped pulses [Esteves-Oliveira et al., 2009]. Besides not causing either surface damage or excessive pulpal temperature increase, as observed in a 3D finite element model, these parameters have also repeatedly been shown to increase enamel resistance to demineralization both in a caries model [Esteves-Oliveira et al., 2009] as well as in shortterm erosion and abrasion models [Esteves-Oliveira et al., 2011a, b]. So it appears that these positive results are not happening by chance but are rather related to the short pulse duration and low fluence used. This is also potentially the reason why the other few studies evaluating other CO 2 lasers for erosion prevention have obtained either no significant increase in enamel resistance to erosive demineralization or lead to surface damage [Steiner-Oliveira et al., 2010; Wiegand et al., 2010] all of them used higher fluence or longer pulse duration than those used here. Thus, at least the effects on the enamel surface with the same laser system appear to be reproducible and predictable (same institute, other models). The only obstacle for the reproducibility of the outcomes found here is the fact that only some CO 2 laser equipment, readily available, emits laser irradiation in the same conditions as our system does. So, even if one would consider this protocol as established for the prevention of dental erosion, there are currently some problems for additional clinical and laboratorial tests: (1) no clinical lasers are available that can reproduce the irradiation conditions proposed here and 572 Esteves-Oliveira et al.

9 (2) most of the dental faculties in the world do not have such a laser system, as it was designed for industrial applications. However, it is imaginable that in the future the size and the costs of such a laser system will decrease due to technological evolution and mass production demanded by other industries (automobiles, aircrafts and civil engineering). The mechanism through which laser irradiation increases enamel resistance to acid dissolution has been related to chemical changes in enamel. Enamel mineral is composed of an impure form of hydroxyapatite, containing carbonate, sodium and magnesium. The presence of these impurities, especially of carbonate ions, in the crystal structure increases the mineral solubility. By contrast, when these impurities are eliminated or reduced, enamel becomes less acid soluble [Shellis et al., 2014]. This has been suggested as the main mechanism of action of the CO 2 laser for reducing enamel acid dissolution, since irradiation conditions causing a very high increase of enamel resistance to acid dissolution also promote a very high reduction (98%) of carbonate from the 1 μm of the outer enamel surface [Featherstone et al., 1998a, b; Zuerlein et al., 1999]. There is evidence that the ideal surface temperature increase for eliminating carbonate from enamel ( C) can be obtained with both the 9.6 and the 10.6 μm CO 2 wavelengths [Fried et al., 1996]. In the present study the CO 2 laser irradiation alone significantly reduced erosive enamel loss compared to no treatment. However, as expected, this effect is high during the first few days (day 3: L1 = 85%, L2 = 85%) but decreases over time (day 10: L1 = 33%, L2 = 44%). This indicates that the laser-modified enamel layer, which is supposed to be at least as thick as the penetration depth of the μm wavelength in enamel, namely 12 μm [Zuerlein et al., 1999], at some point is dissolved and nonmodified deeper enamel layers are exposed. It may be concluded that under the conditions of this in vitro study a tin-containing fluoride solution causes a high reduction (88%) of enamel erosive surface loss, while its combination with CO 2 laser irradiation at 0.3 J/cm 2 (10.6 μm, 5 μs, 226 Hz) hampers erosive loss almost completely. This laser irradiation condition has been previously reported as not causing either damage to the surface or an excessive pulp temperature increase in previous investigations. However, further in vivo tests (animal models) are necessary to ensure total absence of irreversible pulp reaction before this irradiation protocol can be used clinically. Acknowledgments This work was supported by the START Program of the Medical Faculty of the RWTH Aachen University (grant No. AZ43/09). The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. Author Contributions M.E.-O., R.-D.H. and C.P.E. conceived and designed the experiments. N.W. and M.E.-O. performed the experiments. M.E.- O., R.-D.H., N.W., and H.M.-L. analyzed the data. M.E.-O., R.- D.H., N.W., C.A., H.M.-L., and C.P.E. wrote the paper. Disclosure Statement We herewith declare that this study comprises original results and that there are no relationships that might lead to a conflict of interest. References Carvalho TS, Lussi A: Susceptibility of enamel to initial erosion in relation to tooth type, tooth surface and enamel depth. Caries Res 2015; 49: Esteves-Oliveira M, Pasaporti C, Heussen N, Eduardo CP, Lampert F, Apel C: Prevention of toothbrushing abrasion of acid-softened enamel by CO 2 laser irradiation. J Dent 2011a;39: Esteves-Oliveira M, Pasaporti C, Heussen N, Eduardo CP, Lampert F, Apel C: Rehardening of acid-softened enamel and prevention of enamel softening through CO 2 laser irradiation. J Dent 2011b;39: Esteves-Oliveira M, Zezell DM, Meister J, Franzen R, Stanzel S, Lampert F, Eduardo CP, Apel C: Co 2 laser (10.6 μm) parameters for caries prevention in dental enamel. Caries Res 2009; 43: Featherstone JD, Barrett-Vespone NA, Fried D, Kantorowitz Z, Seka W: CO 2 laser inhibition of artificial caries-like lesion progression in dental enamel. J Dent Res 1998a;77: Featherstone JD, Fried D, Duhn CW: Surface dissolution kinetics of dental hard tissue irradiated over a fluence range of 1 8 j/cm 2 ; in Wigdor HA, Featherstone JD, White JM, Katzir A (eds): Lasers in Dentistry IV. Proceedings of the SPIE Meeting. San Jose, Bellingham, 1998b, vol 3248, pp Fox JL, Yu D, Otsuka M, Higuchi WI, Wong J, Powell G: Combined effects of laser irradiation and chemical inhibitors on the dissolution of dental enamel. Caries Res 1992; 26: Fried D, Murray MW, Featherstone JDB, Akrivou M, Dickenson KM, Duhn C, Ojeda OP: Dental hard tissue modification and removal using sealed tea lasers operating at Fλ = 9.6 and 10.6 μm; in: Lasers in Dentistry V. Proceedings of the SPIE Meeting. San Jose, Bellingham, 1999, vol 3593, pp Fried D, Seka W, Glena RE, Featherstone JDB: Thermal response of hard dental tissues to 9- through 11-μm CO 2 -laser irradiation. Opt Eng 1996; 35: CO 2 Laser and Tin Fluoride Reduction of Erosion 573

10 Ganss C, Klimek J, Brune V, Schurmann A: Effects of two fluoridation measures on erosion progression in human enamel and dentine in situ. Caries Res 2004; 38: Ganss C, Lussi A, Grunau O, Klimek J, Schlueter N: Conventional and anti-erosion fluoride toothpastes: effect on enamel erosion and erosion-abrasion. Caries Res 2011; 45: Ganss C, Neutard L, von Hinckeldey J, Klimek J, Schlueter N: Efficacy of a tin/fluoride rinse: a randomized in situ trial on erosion. J Dent Res 2010; 89: Ganss C, Schlueter N, Hardt M, Schattenberg P, Klimek J: Effect of fluoride compounds on enamel erosion in vitro: a comparison of amine, sodium and stannous fluoride. Caries Res 2008; 42: 2 7. Ganss C, von Hinckeldey J, Tolle A, Schulze K, Klimek J, Schlueter N: Efficacy of the stannous ion and a biopolymer in toothpastes on enamel erosion/abrasion. J Dent 2012; 40: Gerard DE, Fried D, Featherstone JD, Nancollas GH: Influence of laser irradiation on the constant composition kinetics of enamel dissolution. Caries Res 2005; 39: Gerrard WA, Winter PJ: Evaluation of toothpastes by their ability to assist rehardening of enamel in vitro. Caries Res 1986; 20: Hannig M, Fiebiger M, Guntzer M, Dobert A, Zimehl R, Nekrashevych Y: Protective effect of the in situ formed short-term salivary pellicle. Arch Oral Biol 2004; 49: Hara AT, Lippert F, Zero DT: Interplay between experimental dental pellicles and stannouscontaining toothpaste on dental erosionabrasion. Caries Res 2013; 47: Hove LH, Holme B, Young A, Tveit AB: The protective effect of TiF4, SnF2 and NaF against erosion-like lesions in situ. Caries Res 2008; 42: Huysmans MC, Jager DH, Ruben JL, Unk DE, Klijn CP, Vieira AM: Reduction of erosive wear in situ by stannous fluoride-containing toothpaste. Caries Res 2011; 45: Huysmans MC, Young A, Ganss C: The role of fluoride in erosion therapy. Monogr Oral Sci 2014; 25: Jo YH, Jung I, Choi CS, Kim I, Lee HM: Synthesis and characterization of low temperature Sn nanoparticles for the fabrication of highly conductive ink. Nanotechnology 2011; 22: Kantola S: Laser-induced effects on tooth structure. VII. X-ray diffraction study of dentine exposed to a CO 2 laser. Acta Odontol Scand 1973; 31: Lussi A, Hellwig E: Risk assessment and causal preventive measures. Monogr Oral Sci 2014; 25: McCormack SM, Fried D, Featherstone JDB, Glena RE, Seka W: Scanning electron microscope observations of CO 2 laser effects on dental enamel. J Dent Res 1995; 74: Rechmann P, Fried D, Le CQ, Nelson G, Rapozo- Hilo M, Rechmann BM, Featherstone JD: Caries inhibition in vital teeth using 9.6-μm CO 2 -laser irradiation. J Biomed Opt 2011; 16: Rolla G, Saxegaard E: Critical evaluation of the composition and use of topical fluorides, with emphasis on the role of calcium fluoride in caries inhibition. J Dent Res 1990; 69: ; discussion Schlueter N, Duran A, Klimek J, Ganss C: Investigation of the effect of various fluoride compounds and preparations thereof on erosive tissue loss in enamel in vitro. Caries Res 2009a;43: Schlueter N, Hardt M, Lussi A, Engelmann F, Klimek J, Ganss C: Tin-containing fluoride solutions as anti-erosive agents in enamel: an in vitro tin-uptake, tissue-loss, and scanning electron micrograph study. Eur J Oral Sci 2009b;117: Schlueter N, Klimek J, Ganss C: In vitro efficacy of experimental tin- and fluoride-containing mouth rinses as anti-erosive agents in enamel. J Dent 2009c;37: Shellis RP, Featherstone JD, Lussi A: Understanding the chemistry of dental erosion. Monogr Oral Sci 2014; 25: Souza BM, Lima LL, Comar LP, Buzalaf MA, Magalhaes AC: Effect of experimental mouthrinses containing the combination of NaF and TiF4 on enamel erosive wear in vitro. Arch Oral Biol 2014; 59: Steiner-Oliveira C, Nobre-dos-Santos M, Zero DT, Eckert G, Hara AT: Effect of a pulsed CO 2 laser and fluoride on the prevention of enamel and dentine erosion. Arch Oral Biol 2010; 55: Stenhagen KR, Hove LH, Holme B, Tveit AB: The effect of daily fluoride mouth rinsing on enamel erosive/abrasive wear in situ. Caries Res 2013; 47: 2 8. Wei SH, Forbes WC: Electron microprobe investigations of stannous fluoride reactions with enamel surfaces. J Dent Res 1974; 53: Wiegand A, Attin T: Design of erosion/abrasion studies insights and rational concepts. Caries Res 2011; 45(suppl 1): Wiegand A, Magalhaes AC, Navarro RS, Schmidlin PR, Rios D, Buzalaf MA, Attin T: Effect of titanium tetrafluoride and amine fluoride treatment combined with carbon dioxide laser irradiation on enamel and dentin erosion. Photomed Laser Surg 2010; 28: Wiegand A, Schneider S, Sener B, Roos M, Attin T: Stability against brushing abrasion and the erosion-protective effect of different fluoride compounds. Caries Res 2014; 48: Wiegand A, Waldheim E, Sener B, Magalhaes AC, Attin T: Comparison of the effects of TiF4 and NaF solutions at ph 1.2 and 3.5 on enamel erosion in vitro. Caries Res 2009; 43: Zuerlein MJ, Fried D, Featherstone JDB: Modeling the modification depth of carbon dioxide laser-treated dental enamel. Lasers Surg Med 1999; 25: Esteves-Oliveira et al.