Insoluble NaF in Duraphat ® May Prolong Fluoride Reactivity of Varnish Retained on Dental Surfaces

varnish should be maintained on enamel surfaces without suffering mechanical disturbance by the patient. Considering the importance of calcium fluoride (CaF2)-like reservoirs on the anticaries effect of professional fluoride application, an in vitro study was designed to test the reactivity of Duraphat® varnish with enamel forming these reservoirs as a function of time. Since most fluoride in Duraphat® varnish is insoluble to react and form products on enamel, the relative contribution of the varnish soluble and insoluble fluoride fractions to the reactivity was also evaluated. For this, whole-varnish, containing soluble and insoluble fluoride (total fluoride concentration of 23699±384 μg F/g), or centrifuged varnish, containing only soluble fluoride (fluoride concentration of 258±97 μg F/g), were applied in a standardized manner on enamel slabs (n=8/varnish group/time), which were immersed in continuously renewed artificial saliva for up to 36 h. CaF2-like reservoirs formed on enamel by varnish application were extracted using 1 M KOH and fluoride concentration was measured with ion specific electrode. The results were expressed as μg F/cm2 of enamel area. Whole varnish formed significantly higher fluoride concentration on enamel than centrifuged varnish, reaching maximum concentration at 24 h (22.0±4.5 μg F/cm2). Centrifuged varnish reached maximum concentration at 6 h (3.20±0.81 μg F/ cm2). In conclusion, a longer varnish retention time than the usually recommended could improve the anticaries effect of Duraphat® varnish, allowing that NaF particles, initially insoluble in the varnish matrix, prolong the reactivity with enamel. Insoluble NaF in Duraphat® May Prolong Fluoride Reactivity of Varnish Retained on Dental Surfaces


Introduction
Among commercial formulations for professional fluoride application, the anticaries effect of fluoride gel and varnish (Duraphat ® ) is based on evidence (1,2).Their mechanism of action is attributed to the reactivity of the fluoride present in these formulations with enamel/ dentine, forming loosely-bound, calcium fluoride (CaF 2 )like reservoirs (3), whose effect on enamel demineralization is concentration-dependent (4).However, while in gels all fluoride is available for an immediate (1-4 min) reaction with enamel/dentine because the NaF salt used is dissolved in water (5,6), in the varnish the colophony hydrophobic matrix is dissolved in ethanol, in which only a small percentage of NaF is soluble.
Based on studies suggesting that the reaction of Duraphat ® with enamel is time-dependent (7,8), a prolonged retention of NaF varnishes has been clinically recommended to allow for a more thorough fluoride reaction with enamel.However, there is no consensus among the authors about the time recommended to maintain the varnish, ranging from 4 h (9) to the next morning after application (10)(11)(12).This may have arisen from the limitations of the studies to establish a precise time for varnish maintenance on dental surfaces.Thus, Retief et al. (7) showed that fluoride uptake by enamel was increased when the contact with Duraphat ® varnish was raised from 1 to 24 h, but only total fluoride was measured, and not the loosely-bound fluoride reservoirs.On the other hand, Bruun and Givskov (8) could not observe differences in the reactivity of Duraphat ® varnish with enamel after 6 or 18 h, but a simulation of the intraoral salivary clearance was not performed.Therefore, a detailed kinetics study on the reactivity of fluoride varnish with enamel is still lacking.Also, in both cited studies the relative importance of the fraction of fluoride that is soluble in varnish for an immediate reaction with dental hard tissues forming CaF 2 -like products versus the insoluble NaF particles present in the varnish matrix that could prolong the reaction when dissolved by saliva was not tested.
Thus, it was hypothesized that the reaction of the varnish with enamel would be dependent, on the shortterm, on its soluble fluoride concentration, and on the long-term, on dissolution of insoluble NaF in the matrix.Therefore, it was used whole varnish and the supernatant of the centrifuged varnish, in which all fluoride is soluble, to isolate the effect of total and soluble fluoride on the kinetics of CaF 2 -like formation on enamel.

Experimental Design
This was an in vitro, randomized, blind study, in which 112 bovine enamel slabs (3x3 mm) prepared from sound bovine incisors and with their surfaces polished flat, were randomly allocated (n=8 in each varnish group and time), into the study factors: 1. composition of the varnish, i.e., whole varnish (containing soluble and insoluble fluoride (Duraphat ® , A. Natherman & Cie, Cologne, Germany.Lot N°. 011001)) and centrifuged varnish (the supernatant of centrifuged Duraphat ® (5 min, 11.000 g), containing only soluble fluoride) (Fig. 1); and 2. reaction time, from 5 min to 36 h.Additionally, 16 untreated slabs were used for baseline loosely-bound (KOH-soluble) fluoride concentration determination.The dimension of each slab surface was measured with a digital caliper (± 0.01 mm).Only the outer enamel surface was exposed to the treatments, and the other surfaces were isolated with wax.
The reactivity test was performed by applying whole or centrifuged varnish (5.0±0.5 mg) on the surface of each enamel slab, which was immersed in constantly renewing artificial saliva (13) for 5 min, 30 min, 2 h, 6 h, 12 h, 24 h or 36 h.In each test, 4 slabs of the same varnish group and time were immersed together in 40 mL of artificial saliva (10 mL/slab).Saliva was continuously renewed (0.5 mL/min/slab), simulating salivary flow rate, using two peristaltic pumps, which kept saliva volume within a narrow oscillation range (±1 mL) during the immersion periods.Each test was repeated twice for each of the experimental times, totalizing n=8 slabs/time/group.At the end of each test period, the system was disassembled, the slabs were collected, the varnish removed with acetone-imbibed cotton swab and each slab thoroughly rinsed with acetone for 30 s to ensure complete removal of the varnish but not of the CaF 2 -like reservoirs formed (8,14), followed by a 30 s rinse with purified water.CaF 2 -like concentration formed on enamel was determined.

Determination of CaF 2 -like Concentration on Enamel
Slabs were individually immersed in 0.4 mL of 1 M KOH (4.44 mL/cm²) for 24 h under agitation at room temperature (15).Fluoride in the alkali extract was determined using a specific ion electrode (Orion model 96-09, Orion Research, Cambridge, MA, USA) coupled to an ion analyzer (Orion EA-740) after buffering with TISAB II containing 1 M HCl, and expressed as µg F/cm 2 of enamel area.

Determination of Fluoride Concentration in Varnish
Fluoride concentration in whole and centrifuged Duraphat® varnish (Fig. 1) was determined after a standardized extraction using acetone.Varnish was weighed (10±0.1 mg) in a centrifuge microtube to which 100 µL of acetone was added.After vigorous homogenization in vortex for 15 s to ensure complete varnish dissolution, the extracts of whole or centrifuged varnishes were transferred to 100 or 10-mL volumetric flasks, respectively, and the microtube was washed 3x with 1 mL of water and once again with 100 µL of acetone for total varnish removal.The extract was diluted to 100 or 10 mL using purified water, for whole or centrifuged varnish, and fluoride concentration was measured in this solution using a specific ion electrode after buffering with TISAB II.

Determination of Fluoride Concentration In Artificial Saliva
Aliquots (n=2 to 11) of artificial saliva exiting the flow system were collected at different time-points for determination of fluoride concentration using the fluoride electrode, after buffering with TISAB III.

Statistical Analysis
CaF 2 -like fluoride data were analyzed by two-way ANOVA, considering as factors type of varnish (two levels) and application time (seven levels).The assumption of normality of errors and homogeneity of variances were satisfied by transforming the data to the potency of 0.3 for analysis.The SAS system (SAS Institute Inc., version 9.2, Cary, NC, USA) was used in the analyses, with significance level set at 5%.
Baseline CaF 2 -like concentration in the untreated enamel blocks was 0.11±0.02µg F/cm 2 .Exposure to varnishes significantly increased CaF 2 -like concentration on enamel (Fig. 2).This increase was significantly higher when the whole varnish was applied when compared to the centrifuged varnish (p<0.05).Moreover, although the reactivity of the whole varnish with enamel was stabilized only after 24 h (Fig. 3), for the centrifuged varnish the maximum CaF 2 -like concentration was reached after 6 h.
Fluoride concentration in the continuously flowing artificial saliva was below the detection limit of the fluoride analysis (0.03 µg F/mL) for the collections made after 6 h for the whole varnish and for all times for the centrifuged.In the whole varnish group, the maximum fluoride concentration in artificial saliva (Fig. 3) was observed at 30 min (0.07±0.01 µg F/mL).

Discussion
Since most fluoride in Duraphat ® varnish is insoluble and the CaF 2 -like fluoride reservoirs chemically formed on enamel from soluble fluoride reactivity are considered responsible for the anticaries mechanism of action of professional fluoride application (4), it is in principle challenging to explain how this product is effective to control caries (1).
The current study presents an explanation to the reaction with enamel surface of the soluble and insoluble fluoride fractions in the varnish, and also extends the findings of Retief et al. (7) and Bruun and Givskov (8), which support the importance of retaining the varnish applied on dental surfaces for longer time.Furthermore, our results provide a clue on how the anticaries potential of different varnishes formulation currently found in the market should be evaluated (16).
The findings (Fig. 2) showed that Duraphat ® reaction with enamel is time-dependent and that the insoluble fluoride in the formulation plays a significant role in the process.The greater fluoride concentration formed on enamel by the whole varnish than by the centrifuged one may be attributed to the dissolution of insoluble NaF particles present in the varnish matrix ( 8) and the increase

Fluoride varnish reaction is time-dependent
of free fluoride concentration at the varnish-enamel surface interface, but not to the fluoride released to saliva.Indeed, the concentration of fluoride found in saliva was very low (Fig. 3) to explain the increase of CaF 2 -like fluoride formed on enamel overtime (17).Therefore, fluoride release to saliva from fluoride varnishes may not be considered the best predictor of their anticaries potential because it is short-lived and the anticaries effect of these products is maintained for longer periods, giving support to their use 2 to 4 times/year (1).Also, fluoride varnish should not be considered a slow-releasing material able to protect all dentition against caries because its site-specific action is based on the formation of CaF 2 -like reservoirs on the dental surface where it was applied, with limited reactivity with the surrounding dentition (18).These loosely-bound fluoride reservoirs would be released to a biofilm newly formed on the treated surface (4,12).Likewise, microscopic remnants of the varnish retained on retentive surfaces for long periods of time might function as slow-release fluoride reservoirs, but with limited local effect.
Any model has limitations to adequately simulate the in vivo conditions (19).The use of bovine enamel as a surrogate of human enamel can be considered acceptable given that similar reactivity with both of them has been observed after the use of professional fluoride products (4,6).Also, in the present study, although artificial saliva was used as the immersion media to simulate the kinetics of varnish imbibition and insoluble NaF dissolution, the specimens were subjected to continuous flow simulating what happens in the oral cavity.Furthermore, the relationship between CaF 2 -like concentration and time that varnish is retained on dental surfaces should be tested regarding its effect on enamel-dentine demineralization.Finally, the kinetics of CaF 2 -like formation as a function of varnish retention time should also be tested in situ or in vivo giving support to a later removal (20).
In conclusion, the findings of the present study show that insoluble fluoride in whole-varnish is an important fluoride source that would be dissolved during intraoral exposure, increasing fluoride reactivity with enamel overtime.Considering the anticaries importance of looselybound fluoride reservoirs formed by professional fluoride application, the results suggest that the recommendation that patients avoid removing the varnish applied on dental surfaces could actually improve the anticaries effect of Duraphat ® .

Figure 1 .
Figure 1.Whole and centrifuged Duraphat® varnish used in the experiment.Note the precipitation of NaF particles by centrifugation, while in the cloudy whole varnish, they are dispersed in the matrix.

Figure 3 .
Figure 3. Fluoride concentration (avg±se; n=2 to 11) in artificial saliva according to the time (h) that the varnishes were kept on enamel surfaces.

Figure 2 .
Figure 2. CaF 2 -like (KOH-soluble fluoride) concentration (µg F/cm 2 ) formed on enamel by the whole or centrifuged varnish according to the time under continuous artificial saliva flow (avg±sd;n=8).Capital letters represent the statistical differences between the varnishes conditions, and lower case among the application times for each condition (p<0.05).