INTRODUCTION

During the development of subsurface caries lesions, mineral is dissolved

out of the enamel, resulting in increased porosities that appear clinically

as the so-called 'white-spot' lesions (Ten Cate et al., 2003). Today, these

lesions are commonly treated by enhancing remineralization, e.g., by

improving the individual's oral hygiene or fluoridation. However, in non-

compliant individuals with cavitated proximal lesions and greater lesion

extension, this strategy has considerable limitations.

A promising alternative therapy for the arrest of caries lesions might be

the infiltration of subsurface lesions with low-viscous light-curing resins.

Since porosities of enamel caries act as diffusion pathways for acids and

dissolved minerals, infiltration of these lesions with resins might occlude the

pathways, thus leading to an arrest of caries progression.

Several studies have demonstrated that artificial caries lesions can be

infiltrated by commercially available adhesives and fissure sealants (Davila

et al., 1975; Robinson et al., 2001; Schmidlin et al., 2004; Meyer-Lueckel etal., 2006). Moreover, it has been shown that infiltrated artificial lesions do

not progress in a cariogenic environment (Mueller et al., 2006; Paris et al.,2006). Thus far, only two in vitro studies have addressed the infiltration of

natural lesions. However, these early reports were mainly descriptive

(Davila et al., 1975), or used materials which were not clinically applicable

due to their unsanitary nature (Robinson et al., 1976). Since there are

substantial structural differences between both lesion types, it is not

applicable to transfer findings from artificial to natural lesions.

The surface layer of enamel caries lesions has a lower pore volume

compared with that of the lesion body underneath (Bergman and Lind, 1966;

Silverstone, 1973). Since the infiltration of enamel caries with light-curing

resins is mainly driven by capillary forces, the pore diameter and volume

influence the penetration speed (Paris et al., 2007). Therefore, the surface

layer forms a barrier, which might hamper the infiltration of the lesion body.

From this follows that removing or perforating the surface layer might be

essential for a successful infiltration of the lesion body. In artificial lesions,

brief etching with 37% phosphoric acid enhanced resin penetration (Gray

and Shellis, 2002). With thicker and more mineralized surface layers in

natural lesions (Bergman and Lind, 1966), it was assumed that this etching

procedure would not be effective in eroding the surface layer (Meyer-

Lueckel et al., 2007). The latter study confirmed that etching with 15%

hydrochloric acid gel leads to a more effective erosion of the surface layer

compared with 37% phosphoric acid gel, but did not focus on the

subsequent infiltration of resins into the lesions.

Therefore, the aim of the present study was to evaluate the penetration

of a commercial adhesive into natural proximal caries lesions, without pre-

treatment and with prior conditioning by two different etching gels in vitro.

The working hypotheses were:

(Hypothesis 1) The surface layer of natural un-cavitated caries is a

diffusion barrier, which hampers the penetration of resin. Therefore, no

ABSTRACTInfiltration of non-cavitated caries lesions with

light-curing resins could lead to an arrest of lesion

progression. The aim of this study was to evaluate

the penetration of a conventional adhesive into

natural enamel caries after pre-treatment with two

different etching gels in vitro. Extracted human

molars and premolars showing proximal white-

spot lesions were cut across the lesions

perpendicular to the surface. Corresponding lesion

halves were etched for 120 sec with either 37%

phosphoric acid gel (H3PO

4) or 15% hydrochloric

acid gel (HCl), and subsequently infiltrated with

an adhesive. Specimens were observed by

confocal micro scopy. Mean penetration depths

(SD) in the HCl group [58 (37) �m] were

significantly increased compared with those of the

H3PO

4group [18 (11) �m] (p < 0.001; Wilcoxon).

It can be concluded that etching with 15%

hydrochloric acid gel is more suitable than 37%

phosphoric acid gel as a pre-treatment for caries

lesions intended to be infiltrated.

KEY WORDS: caries, resin infiltration, etching,

acid gel.

Received July 23, 2006; Last revision February 9, 2007;

Accepted March 6, 2007

A supplemental appendix to this article is published

electronically only at http://www.dentalresearch.org.

Resin Infiltration of Natural Caries Lesions

S. Paris, H. Meyer-Lueckel*, and A.M. Kielbassa

Dept. of Operative Dentistry and Periodontology,University School of Dental Medicine, Campus BenjaminFranklin, Charité-Universitätsmedizin Berlin,Assmannshauserstr. 4-6, 14197 Berlin, Germany;*corresponding author, hendrik.meyer-lueckel@charite.de

J Dent Res 86(7):662-666, 2007

RESEARCH REPORTSBiomaterials & Bioengineering

662 at International Association for Dental Research on March 23, 2015 For personal use only. No other uses without permission.jdr.sagepub.comDownloaded from

International and American Associations for Dental Research

J Dent Res 86(7) 2007 Infiltration of Caries Lesions 663

resin penetration occurs without prior

etching; and (Hypothesis 2) pre-

treatment for 120 sec with 15%

hydrochloric acid gel leads to a more

pronounced resin penetration

compared with 37% phosphoric acid

gel.

MATERIALS & METHODSExtracted human molars and premolars

showing proximal white-spot lesions

were used in this study. The study

protocol conformed to the principles

outlined in the Central German Ethics

Committee's statement (2003) focusing

on the use of human body material in

medical research. After being carefully

cleaned of soft tissues, teeth were

stored in 20% ethanol solution until

used. Teeth were examined by 20

stereo microscopy (Stemi SV 11; Carl

Zeiss, Oberkochen, Germany), and

cavitated as well as damaged lesions

were excluded.

For radiographic examination,

teeth were positioned in a silicone base

with the buccal aspects facing a

radiographic tube (Heliodent MD;

Siemens, Bensheim, Germany). To

simulate cheek scatter, we placed a 15-

mm wall of clear Perspex between the

tube and the teeth. Standardized

radiographs (0.12 sec, 60 kV, 7.5 mA)

were taken of each tooth (Ektaspeed;

Kodak, Stuttgart, Germany) and

developed in an automatic processor

(XR 24-II; Dürr Dental, Bietigheim-

Bissingen, Germany). The radiographic

lesion depths were independently

assessed by two examiners and scored

(Marthaler and Germann, 1970): no

translucency (R0), translucency

confined to the outer half on enamel

(R1), translucency confined to the inner

half of enamel (R2), translucency

confined to the outer half of dentin

(R3), or translucency confined to the

inner half of dentin (R4). In case of

disagreement in an assessment of

radiographic lesion depth, a consensus

rank was reached.

The roots of the teeth were

removed, and the crowns were cut

across the caries lesions perpendicular

to the surface (Band Saw; Exakt

Apparatebau, Norderstedt, Germany),

providing two halves of each lesion

(Figs. 1a, 1b). Subsequently, the cut surfaces were examined

(stereo microscope, 20 ; Stemi SV 11) and classified with respect

to the histological lesion extension, according to the radiological

grading (Marthaler and Germann, 1970): C1, extension into the

outer half of enamel; C2, extension into the inner half of enamel;

or C3, extension into the outer half of dentin. Lesions extending

into the inner half of dentin (C4) were excluded. Corresponding

lesion halves showing the same grading level (C1-C3) in

Figure 1. Representative images of a lesion treated with the adhesive. (A) Clinical aspect of the mesialsurface of a human molar showing a white-spot lesion (dotted line). The lesion was cut in two halvesalong the dashed line. (B) Aspect of the cut surfaces of the same enamel lesion. (C-E) Confocalmicroscopic images of resin-infiltrated lesions (E, sound enamel; SL, surface layer; LB, lesion body; R,penetrated resin; S, lesion surface). (C) Deep resin penetration can be observed after etching withHCl. (D) The surface layer of this H3PO4-etched caries lesion was not eroded completely. Therefore,only superficial resin penetration occurred, as indicated by a fine rim of red fluorescence at the toothsurface. (E) Magnified image of an HCl-etched lesion (40x objective). The outermost 50-100 �m ofprism cores are filled with resin. In non-infiltrated parts of the lesion body, the highly porous prismcenters show green fluorescence.

at International Association for Dental Research on March 23, 2015 For personal use only. No other uses without permission.jdr.sagepub.comDownloaded from

International and American Associations for Dental Research

664 Paris et al. J Dent Res 86(7) 2007

histological caries extension were assigned to the treatment (TRT)

group (n = 10 each). In case the corresponding lesion halves

differed in lesion extension, the deeper lesion was used as the

control (CTR; n = 10), and the remaining half was disposed of.

Subsequently, the cut surfaces were covered with nail varnish.

In the treatment (TRT) group, corresponding lesion halves were

etched either with 37% phosphoric acid gel (H3PO

4; total etch;

IvoclarVivadent, Schaan, Liechtenstein), or with an experimental

15% hydrochloric acid gel (HCl). The HCl gel contained

hydrochloric acid 15%, glycerol 19%, highly dispersed silicon

dioxide 8%, and methylene blue 0.01% in aqueous solution. After

120 sec, the gels were rinsed thoroughly with water spray for 30

sec. In the control (CTR) group, no acid etching was performed.

Lesions were immersed in pure ethanol for 30 sec and

subsequently dried for 60 sec with oil-free compressed air.

A dental adhesive (Excite; IvoclarVivadent, Schaan,

Liechtenstein) labeled with 0.1% tetramethylrhodamine

isothiocyanate (TRITC; Sigma Aldrich, Steinheim, Germany) was

applied to the lesion surfaces. The resin was allowed to penetrate

the lesions for 5 min. Subsequently, excess material was removed

by means of cotton pellets, and the resin was light-cured for 30 sec

(Translux CL; Heraeus Kulzer, Hanau, Germany) at 400 mW/cm2.

The nail varnish was carefully removed, and specimen halves were

fixed on object holders parallel to the cut surface and polished

(Exakt Mikroschleifsystem, Abrasive Paper 2400, 4000; Exakt

Apparatebau, Norderstedt, Germany). To stain remaining pores,

we immersed the specimens in 50% ethanol solution containing

100 �M/L sodium fluorescein (Sigma Aldrich) for 3 hrs.

Specimens were observed by confocal laser scanning

microscopy (CLSM Leica TCS NT; Leica, Heidelberg, Germany)

in dual-fluorescence mode and with a 10x objective. The excitation

light had two wavelength maxima, at 488 and 568 nm. The emitted

light was split by a 580-nm reflection short-pass filter and passed

through a 525/50-nm band-pass filter for FITC and a 590-nm long-

pass filter for RITC detection. Images with a lateral dimension of

1000 x 1000 �m2 and a resolution of 1024 x 1024 pixels were

recorded and analyzed by AxioVision LE software (Zeiss,

Oberkochen, Germany). Penetration depths and thicknesses of the

(residual) surface layer for the lesion halves were measured at up

to 10 defined points (depending on the lesion size; indicated by a

100-�m grit), and mean values were calculated. Additionally to

CLSM analysis, acid-etched as well as infiltrated lesion surfaces

were observed by scanning electron microscopy (APPENDIX).

Statistical analysis was performed with SPSS software (SPSS

for Windows 11.5.1; SPSS, Chicago, IL, USA). Data were checked

for normal distribution by the Kolmogorov-Smirnov test. To

analyze differences in penetration depth between lesion halves/acid

gels, we used the Wilcoxon test for paired samples. For

comparison between unpaired groups, we performed Mann-

Whitney and Kruskal-Wallis tests. Penetration depths were

analyzed with regard to possible differences between various

histological lesion extensions (C1-C3) and radiological grades

(R0-R3). The level of significance was set at 5%.

RESULTSIn the CLSM images, the penetrated resin showed a red

fluorescence, whereas remaining pores within the lesion, as

well as dentin, appeared green (Figs. 1c-1e). Solid material,

such as sound enamel or the surface layer, was displayed black.

Penetration depths varied considerably. For lesion halves

etched with HCl gel, the mean penetration depth (standard

deviation) [58 (37) �m] was significantly higher compared

with that of those lesions treated with H3PO

4gel [18 (11) �m]

(p < 0.001; Wilcoxon) (Fig. 2). Without acid-etching, no resin

penetration was found [0 (1) �m]. Within treatment groups, no

significant differences for penetration depths could be observed

between various lesion extensions (C1-C3) (p > 0.05; Kruskal-

Wallis).

For radiological grading of lesion extensions, good inter-

observer agreement could be found (� = 0.804). Similar to

histological lesion extension (C1-C3), no significant

differences in penetration depth could be observed among

different radiological grades (R0-R3) (Table).

For those lesions where the surface layer was completely

removed (CTR, n = 0; H3PO

4, n = 2; HCl, n = 8), significantly

higher (p < 0.01; Mann-Whitney) penetration depths [65 (35)

�m] could be found compared with those lesions where

residues of the surface layer remained after etching [33 (31)

�m]. Surface layer thickness was significantly reduced after

HCl etching [20 (18) �m], compared with that in the lesions

etched with phosphoric acid [37 (25) �m] and with the non-

etched CTR group [42 (23) �m] (p > 0.05; Mann-Whitney).

DISCUSSIONIn previous studies where confocal microscopy was used, resin

penetration was visualized by labeling of the resin with

Figure 2. Mean penetration depths of resin for various pre-treatmentsand histological lesion extensions (box and whisker plots with quartilesand medians; n = 10 per group). Statistically significant differencesbetween groups are indicated with asterisks (*p < 0.05; **p < 0.01;***p < 0.001; Mann-Whitney). Abbreviations: C1, caries extension intothe outer half of enamel; C2, caries extension into the inner half ofenamel; C3, caries extension into the outer half of dentin.

at International Association for Dental Research on March 23, 2015 For personal use only. No other uses without permission.jdr.sagepub.comDownloaded from

International and American Associations for Dental Research

J Dent Res 86(7) 2007 Infiltration of Caries Lesions 665

fluorescent dyes (Schmidlin et al., 2004). In another

approach, the remaining (not infiltrated) pores were

marked with a dye-labeled polymer (Meyer-

Lueckel et al., 2006; Paris et al., 2006), or by

imbibition in a fluorescent solution (González-

Cabezas et al., 1998). In the present study, these

two methods were modified and combined. The

penetration of the resin was visualized by the red

fluorescence of TRITC. The remaining lesion pores

were indicated by the green fluorescence of sodium

fluorescein. Hard tissues with small pore volume

including the surface layer or sound enamel were

neither infiltrated by the resin nor stained by the

green solution and appeared black. Thus, the dual-

fluorescence technique used in the present study

allowed for the simultaneous observation of the

porous lesion structure, the penetrated resin, as well

as structures with small pore volume.

Caries infiltration might be a promising approach for the

treatment of uncavitated caries lesions. In contrast to fissure

sealing, where the diffusion barrier is placed on top of the

(lesion) surface, the infiltration technique aims to create the

diffusion barrier inside the lesion, replacing lost mineral with

resin. Therefore, the infiltration treatment should be

differentiated from sealing techniques, where a resin layer

(Goepferd and Olberding, 1989; García-Godoy et al., 1997;

Tantbirojn et al., 2000) was established on the caries lesions. A

recent clinical trial, where proximal enamel lesions were

superficially sealed with an adhesive, found a significantly

reduced but still relatively high (43.5%) lesion progression over

an 18-month observation period (Martignon et al., 2006).

Another clinical investigation did not find any significant

differences in lesion progression between the sealed and the

control group (Gomez et al., 2005). The latter authors

speculated that treatment failures might be due to incomplete

sealing or sealant disintegration over time. Moreover,

laboratory studies confirmed the inferior resistance of unfilled

resins to mechanical and chemical stress (Schmidlin et al.,2002, 2006). Therefore, it is questionable whether superficial

smooth-surface sealing with unfilled resins is, as yet, generally

applicable in daily practice.

Compared with the latter concepts, the infiltration

treatment might bear several advantages. With the infiltration

technique, excessive resin is removed from the tooth surface

before light-curing, whereby clinical application is greatly

simplified. With application strips coated on one side,

proximal lesions can be infiltrated without special protection

for the adjacent tooth, and after only minimal tooth

separation, e.g. , by means of a wooden wedge or an

orthodontic rubber band (unpublished results). Moreover,

with this treatment, no sealant margins are produced on the

tooth surface that could enhance plaque accumulation and

cause periodontal inflammation. Furthermore, infiltration of

the porous lesion structures might strengthen the lesion

mechanically and prevent cavitation.

To infiltrate a caries lesion, the penetrating resin needs

access to the porous spaces of the lesion body. It was assumed

that the penetration could be hampered by the highly

mineralized surface layer, where the pore volume is

considerably lower. In fact, in the present investigation, no

penetration was found without prior acid-etching. Moreover, it

could be demonstrated that penetration was significantly

increased in those parts of the lesion where the surface layer

was completely removed after acid-etching. Therefore,

hypothesis 1 could be corroborated.

Recently, 15% hydrochloric acid gel proved to erode the

surface layer more effectively than 37% phosphoric acid gel

(Meyer-Lueckel et al., 2007). In the present investigation, the

effects of etching with these acid gels on resin penetration were

compared. Phosphoric acid (37%) is frequently used in

restorative dentistry for adhesive purposes. Hydrochloric acid

has been previously used for enamel microabrasion

(McCloskey, 1984; Mathewson et al., 1987). Although short-

term contact of this strong acid with mucosa has been shown to

be harmless (Croll et al., 1990), safety precautions, such as a

rubber dam, should be used in clinical practice.

Since significantly higher penetration depths could be

found after etching with hydrochloric acid, hypothesis 2 could

be confirmed as well. However, the surface layer could not be

eroded completely in 67% of lesions in the HCl group. Thus,

longer application times should be considered to achieve

complete surface layer erosion. However, resin penetration was

not influenced by macroscopic or radiological lesion extension.

The penetration depths observed for Excite in natural

lesions in the present study were lower compared with those

observed in artificial lesions (104 �m/30 sec) in a previous

investigation (Meyer-Lueckel et al., 2006), although a ten-fold

longer penetration time was chosen in the present study. It

might be argued that the incomplete surface layer erosion in

natural lesions, even for those etched with HCl, could be

responsible for this contradiction. However, in specimens

where the surface layer could be totally removed, mean

penetration depths were lower as well. In contrast to artificial

lesions, the pores of natural caries might be contaminated with

organic materials, such as proteins and carbohydrates, that

might hamper resin penetration as well. This underlines that

findings from artificial lesions cannot necessarily be

extrapolated to natural lesions.

It can be concluded that the surface layer of a non-cavitated

natural caries lesion is a barrier that significantly hampers the

penetration of a light-curing resin. Therefore, no substantial

resin penetration could be observed without prior acid-

conditioning. Etching for 120 sec with hydrochloric acid gel led

to deeper resin penetration than etching with phosphoric acid

gel, although surface layers could not be removed in all cases.

Table. Mean Penetration Depths [�m (standard deviations)] for the VariousRadiological Caries Extensions

Group\Radiolucency R0* R1 R2 R3

CTR 0 ( 0)n = 8 0 ( 0) n = 10 1 ( 1) n = 8 0 ( 0) n = 4

TRTH3PO4 25 (15) n = 6 17 (12) n = 10 16 ( 7) n = 8 16 (10) n = 6HCl 47 (27) n = 6 65 (41) n = 10 52 (27) n = 8 67 (52) n = 6

* Abbreviations: R0, no radiographic translucency; R1, translucency confined tothe outer half of enamel; R2, translucency confined to the inner half of enamel;R3, translucency confined to the outer half of dentin; CTR, unetched controlgroup; TRT, treatment group; H3PO4, phosphoric acid etching; HCl, hydrochloricacid etching.

at International Association for Dental Research on March 23, 2015 For personal use only. No other uses without permission.jdr.sagepub.comDownloaded from

International and American Associations for Dental Research

666 Paris et al. J Dent Res 86(7) 2007

ACKNOWLEDGMENTSThis study was supported by the Deutsche Forschungs -

gemeinschaft (DFG; PA 1508/1-1). The authors are indebted to

Mrs. Anja Bartels and Mrs. Julia Heinrich (Dept. of Operative

Dentistry and Periodontology, CBF, Charité) for their excellent

contributions to the experiments, to Dr. Herbert Renz (Dept. of

Experimental Dentistry, CBF, Charité) for his assistance with

the SEM, and to Prof. Dr. Harald Stein (Institute for Pathology,

CBF, Charité) for providing the CLSM.

The Charité-Universitätsmedizin Berlin holds US

(US10/432,271) and European (EP06021966.4) patent

applications for an infiltration technique for dental caries lesions

in which the authors of this study are appointed as inventors.

REFERENCESBergman G, Lind PO (1966). A quantitative microradiographic study

of incipient enamel caries. J Dent Res 45:1477-1484.

Central German Ethics Committee (2003). Statement: [The use of

human body materials for the purposes of medical research].

20.02.2003 http://www.zentrale-ethikkommission.de/10/

31KoerpermatZ.html [site in German].

Croll TP, Killian CM, Miller AS (1990). Effect of enamel

microabrasion compound on human gingiva: report of a case.

Quintessence Int 21:959-963.

Davila JM, Buonocore MG, Greeley CB, Provenza DV (1975).

Adhesive penetration in human artificial and natural white spots. JDent Res 54:999-1008.

García-Godoy F, Summitt JB, Donly KJ (1997). Caries progression of

white spot lesions sealed with an unfilled resin. J Clin PediatrDent 21:141-143.

Goepferd SJ, Olberding P (1989). The effect of sealing white spot

lesions on lesion progression in vitro. Pediatr Dent 11:14-16.

Gomez SS, Basili CP, Emilson CG (2005). A 2-year clinical evaluation

of sealed noncavitated approximal posterior carious lesions in

adolescents. Clin Oral Investig 9:239-243.

González-Cabezas C, Fontana M, Dunipace AJ, Li Y, Fischer GM,

Proskin HM, et al. (1998). Measurement of enamel remineral -

ization using microradiography and confocal microscopy. A

correlational study. Caries Res 32:385-392.

Gray GB, Shellis P (2002). Infiltration of resin into white spot caries-

like lesions of enamel: an in vitro study. Eur J Prosthodont RestorDent 10:27-32.

Marthaler TM, Germann M (1970). Radiographic and visual

appearance of small smooth surface caries lesions studied on

extracted teeth. Caries Res 4:224-242.

Martignon S, Ekstrand KR, Ellwood R (2006). Efficacy of sealing

proximal early active lesions: an 18-month clinical study

evaluated by conventional and subtraction radiography. CariesRes 40:382-388.

Mathewson RJ, Morrison JT, Carpenter R (1987). Modification of

stained enamel surfaces: use of hydrochloric acid and pumice

mixture. J OK Dent Assoc 77:22-25.

McCloskey RJ (1984). A technique for removal of fluorosis stains. JAm Dent Assoc 109:63-64.

Meyer-Lueckel H, Paris S, Mueller J, Colfen H, Kielbassa AM (2006).

Influence of the application time on the penetration of different

dental adhesives and a fissure sealant into artificial subsurface

lesions in bovine enamel. Dent Mater 22:22-28.

Meyer-Lueckel H, Paris S, Kielbassa AM (2007). Surface layer

erosion of natural caries lesions with phosphoric and hydrochloric

acid gels in preparation for resin infiltration. Caries Res 41:223-

230.

Mueller J, Meyer-Lueckel H, Paris S, Hopfenmuller W, Kielbassa AM

(2006). Inhibition of lesion progression by the penetration of

resins in vitro: influence of the application procedure. Oper Dent31:338-345.

Paris S, Meyer-Lueckel H, Mueller J, Hummel M, Kielbassa AM

(2006). Progression of sealed initial bovine enamel lesions under

demineralizing conditions in vitro. Caries Res 40:124-129.

Paris S, Meyer-Lueckel H, Colfen H, Kielbassa AM (2007).

Penetration coefficients of commercially available and

experimental composites intended to infiltrate enamel carious

lesions. Dent Mater 23:742-748.

Robinson C, Hallsworth AS, Weatherell JA, Kunzel W (1976). Arrest

and control of carious lesions: a study based on preliminary

experiments with resorcinol-formaldehyde resin. J Dent Res55:812-818.

Robinson C, Brookes SJ, Kirkham J, Wood SR, Shore RC (2001). In

vitro studies of the penetration of adhesive resins into artificial

caries-like lesions. Caries Res 35:136-141.

Schmidlin PR, Gohring TN, Sener B, Lutz F (2002). Resistance of an

enamel-bonding agent to saliva and acid exposure in vitro assessed

by liquid scintillation. Dent Mater 18:343-350.

Schmidlin PR, Zehnder M, Pasqualetti T, Imfeld T, Besek MJ (2004).

Penetration of a bonding agent into de- and remineralized enamel

in vitro. J Adhes Dent 6:111-115.

Schmidlin PR, Kluck I, Zimmermann J, Roulet JF, Seemann R (2006).

Caries-preventive potential of an adhesive patch after

thermomechanical loading—a microbial-based in vitro study. JAdhes Dent 8:7-12.

Silverstone LM (1973). Structure of carious enamel, including the

early lesion. Oral Sci Rev 3:100-160.

Tantbirojn D, Rozzi SM, Mitra SB, Kedrowski BL, Douglas WH

(2000). In vitro inhibition of enamel demineralization by a

polymerizable amphiphilic film. Eur J Oral Sci 108:564-568.

Ten Cate JM, Larsen MJ, Pearce EIF, Fejerskov O (2003). Chemical

interactions between the tooth and oral fluids. In: Dental caries.

Fejerskov O, Kidd EAM, editors. Oxford: Blackwell Munksgaard,

pp. 49-69.

at International Association for Dental Research on March 23, 2015 For personal use only. No other uses without permission.jdr.sagepub.comDownloaded from

International and American Associations for Dental Research