Journal of Oral Rehabilitation 1997 24; 20-25

The effect of dentine thickness on diffusion of resinmonomers in vitroA . H A M I D & W . R . H U M E Department of Restorative Dentistry, University of California, San Francisco, CA, U.S.A.

SUMMARY Forty extracted human third molar teethwere divided into four groups, each of 10 teeth, totest the hypothesis that dentine thickness variationinfluences diffusion of the monomers 2-hydroxy-ethylmethacrylate (HEMA) and triethylene glycoldimethacrylate (TEGDMA) from light-cured bondingresin-composite resin restorations to the pulp space.An occlusal cavity 6 mm in diameter was preparedin each tooth of four groups with remaining dentinethickness of 3-4-3-6, 2-4-2-6, 1-4-1-6 and 0-4-0-6 mm,respectively. A polypropylene chamber was attachedto the cemento-enamel junction of each tooth to

contain 1 mL of distilled water. Each cavity wastreated with Scotchbond Multipurpose (3M, U.S.A.)then restored with ZIOO (3M) and light activated for30 s. Water samples were retrieved over a time courseup to 30 days and analysed by high performanceliquid chromatography. Both HEMA and TEGDMAwere detected in the pulp samples for all teeth.Decreasing dentine thickness substantially increasedpulpward diffusion rate of both HEMA and TEGDMAduring the first day after placement, as well as thetotal release of both components from a bondingresin-composite combination in vitro.

Introduction

Although most placement of resin-based materials ontodentine is well tolerated by the pulp if bacteria areexcluded, there is strong evidence that bonding resin-composite resin restorations can cause pulpal damageindependently of bacterial microleakage when placedonto thin, acid-treated dentine in humans (Qvist, Staltze& Qvist, 1989) and monkeys (Horsted, 1987; Fujitani,Inokoshi & Hosada, 1992). The probable factorsresponsible for such damage are resin monomers whichdiffuse through dentine (Gerzina etal., 1991). Hankset al. (1988) showed in an in vitro model system thatthe presence of intervening dentine reduced thecytotoxicity of setting composite resin, greaterprotection against cytotoxicity being observed whencomposite resins were placed on 1-5 mm thicknessthan on 0-5 mm thickness dentine.

Dentine permeability is theoretically directly propor-tional to the number of exposed tubules and their

diameter, and is inversely proportional to dentine thick-ness (Pashley, 1985). Decreasing thickness can beexpected to increase tubule diameter and density. Thepresent study was carried out to test the hypothesisthat dentine thickness influences the rate and totalamount of chemicals diffusing through dentine to thepulp space. We did this by measuring diffusion overtime of monomers to the pulp space from a bondingresin-composite resin combination.

Materials and methods

Forty freshly extracted, non-carious human third molarteeth were collected from consenting patients at theUCSF School of Dentistry. The teeth were extractedfor therapeutic reasons unrelated to this study, weresterilized with gamma radiation as described by Whiteetal. (1994), then stored in distilled water at 4°C untiluse. The teeth were prepared as described by Hume(1985) and Gerzina & Hume (1994) (Fig. 1). In brief.

20 © 1997 Blackwell Science Ltd

D E N T I N E T H I C K N E S S O N M O N O M E R D I F E U S I O N 21

polypropylene/ chamber

1 ml distilledwater

Fig. 1. A sectional diagrammatic representation of a freshlyextracted htiman third molar tooth with an occlusal cavity, withremaining dentine thickness (RDT) and polypropylene chambercontaining 1 mL distilled water attached to the pulpal surface.

for each tooth the root system was removed from 2 mmbeyond the cemento-enamel junction and the root andpulp tissue discarded. A circular occlusal cavity wasprepared by hand with a tungsten carbide bur (#FG 56)at high speed with water spray. The diameter of eachocclusal cavity was as close to 6 mm as could bedetermined by multiple measurements during cavitypreparation. The remaining dentine thickness (RDT)between the pulpal wall of the cavity and the roof ofthe pulp chamber was measured at multiple pointsduring cavity preparation so that groups of 10 teetheach were prepared with RDT in the ranges 3-4-3-6 mm,2-4-2-6 mm, 1-4-1-6 mm and 0-4-0-6 mm. Each toothwas set into a polypropylene chamber and sealed atthe cemento-enamel junction with sticky wax. Onemillilitre of distilled water* was added to each chamber.Each cavity was treated with 10% maleic acid for 15 sand then washed with water and dried with a streamof air. Scotchbond Multipurpose (SMP) primer, whichcontains 2-hydroxyethylmethacrylate (HEMA), wasthen applied to the cavity walls and dried with astream of air without delay; immediately thereafterSMP adhesive (3M, U.S.A.), which contains both HEMAand 2,2-bis(p-2'hydroxy-3'methacryloxypropoxy)-phenylene propane (Bis-GMA) was applied then light-activated for 10 s using a visible light curing unit (Visilux2, 3M). The cavities were then restored with ZIOO (3M),which contains both triethylene glycol dimethacrylate(TEGDMA) and Bis-GMA to a depth of 2 mm and lightactivated for 30 s. The teeth and chambers were keptat 37°C. Chamber contents (eluates) were retrieved

*MilliQ RO6 Plus. Millipore corporation, Water chromatography

division, Milford. MA, U.S.A.

Table 1. HPLC conditions

ColumnMobile phase

Flow rate

Detector

Resolve 150 x 3-9 mm Silica CI8, 5 mmB 100% methanolIsocratlc 30% methanol for HEMAIsocratic 70% methanol for TEGDMA1-2 mL/min for HEMA1-6 mL/min for TEGDMAUV 215 nm

All components were from Waters Chromatography Division,Millipore Corporation, Milford, MA, U.S.A.

over a time course (14-4, 43-2, 144 and 432 min; 1, 3,10 and 30 days) and replaced with fresh distilled water.

Analyses of eluates were carried out by reversed-phase high performance liquid chromatography usinga 600E system controller, 717 auto sampler, cartridgepre-column, stainless steel silica C18 Resolve column, atunable UV/visible absorbance detector and Millenniumsoftware database (all components from Millipore Cor-poration, Waters Chromatography Division, Milford,MA, U.S.A.). The conditions for HPLC are summarizedin Table 1. Monomers were identified in the eluatesamples by comparing with the chromatograms ofauthentic standards of monomers HEMA and TEGDMA(both from Aldrich Chemical Co., Milwaukee, U.S.A.).The rate of release of each component was calculatedby dividing the amount in each eluate by the collectiontime. Cumulative release was calculated by addition ofthe amounts in each eluate. Release rate and cumulativerelease data were expressed as mean ± standard devi-ation of the mean. We used one-way repeated measuresANOVA to evaluate differences between dentine thicknesswith respect to rate of HEMA and TEGDMA release atsix timepoints (14-4 min, 43-2 min, 144 min, 432 min;1 day and 3 days), and the Tukey multiple comparisonprocedure at a significance level of alpha = 0-05 toevaluate pairwise differences at each time point. One-way ANOVA was also used to evaluate differencesbetween dentine thickness at day 3 for cumulativeHEMA and TEGDMA release, and Tukey multiple com-parison procedure was used at alpha = 0-05 to evaluatepairwise differences.

Results

Both HEMA and TEGDMA were detected in pulpchamber samples in all groups of teeth at times up to3 days. Neither monomer was detected in greater than

© 1997 Blackwell Science Ltd, Journal of Oral Rehabilitation 24; 20-25

22 A. HAMID & W. R. HUME

LUI

- 4

O 0-4 -0-6 mm RDT• 1-4 - 1-6 mm RDTA 2-4-2-6 mm RDT• 3-4 -3-6 mm RDT

-2-5 -1-5 -1 -0-5 0Time (log 10 (day))

Fig. 2. HEMA release rate for various dentine thickness groupsover time. Time log o represents 1 day.

0-6 vs. 1-4-1-6 mm. The cumulative release of HEMAat 3 days was significantly different (P < 0-05) betweenall groups with the exception of 2-4-2-6 vs. 3-4-3-6 mm.The overall HEMA cumulative release between thegroups were also significantly different with the excep-tion of 2-4-2-6 vs. 3-4-3-6 mm.

TEGDMA diffusion rates and cumulative amountsreleased are shown in Figs 4 and 5, respectively. Thehighest mean diffusion rate of TEGDMA was in thesecond sample period (14-4-43-2 min) for all teethand declined thereafter. The pairwise differences ofTEGDMA release rate between the groups are sum-marized in Table 3. The overall differences between thegroups were significant (P < 0-05). TEGDMA cumulat-ive release at 3 days was significantly different between0 4-0-6 mm thickness group vs. all other groups.

o<

O 0-4 - 0-6 mm RDT1-4- 1-6 mm RDT

A 2-4-2-6 mm RDTA 3-4-3-6 mm RDT

-0-2-2-5 -1-5 -1 -0-5 0

Time (log 10 (day))

Fig. 3. HEMA cumulative release for various dentine thicknessgroups over time.

trace amounts in 10 or 30 day samples. Bis-GMA wasnot detected in pulp chamber samples. Release ratesand cumulative release of HEMA and TEGDMA weremarkedly different depending on dentine thickness.Thinner dentine allowed more monomer diffusion thanthicker dentine.

HEMA diffusion rates and cumulative amounts areshown in Figs 2 and 3, respectively. The highest meandiffusion rate of HEMA was in the first sample period(0-14-4 min) for all teeth and declined thereafter. Thepairwise differences of HEMA release rate betweenthe groups are summarized in Table 2. The overalldifferences in HEMA release rate between the groupswere significant (P < 0-05), with the exception of 0-4-

Discussion

Visible light cure composite resins have been widelyaccepted as restorative materials due to their aestheticmerits and their capacity to bond to both enameland dentine. Resin-based bonding agents are used toincrease bond strength to dentine and to decreasemicroleakage. One shortcoming of composite resins isthat they are sometimes associated with adverse pulpalresponses (Johnson, Gordon & Bales, 1988; Borgmeijeretai, 1991). Given evidence on the permeability ofdentine (Pashley, 1985, 1990), it is very reasonable topropose that some adverse pulpal responses might bedue to diffusion of chemicals from the resin materialsthrough dentine to the pulp. As noted above, Qvist et al(1989) and Fujitani et al. (1992) demonstrated suchresponses in resin-restored teeth in the absence ofbacterial microleakage when dentine was thin.

Hume (1985) showed in an in vitro study thatchemicals from resin composite diffused through acid-treated dentine and killed test cells. TEGDMA wasshown to be a cytotoxic component of composite resinsmaterials in preliminary studies (Hood & Hume, 1990;Gerzina et al., 1991; Hume, Gezina & Rouse, 1993).TEGDMA is used as a component of many bonding andcomposite resins to reduce the viscosity of the resinsystem and to enhance manipulative properties (Ruyter& Sj0vik, 1981). HEMA is also cytotoxic (Hanks et ai,1992; Bruce, McDonald & Sydiskis, 1993), and isincluded in many bonding resins to enhance bondstrength to dentine (Nakabayashi & Takarada, 1992).The release of HEMA and TEGDMA from dental resins

© 1997 Blackwell Science Ltd, Journal of Oral Rehabilitation 24; 20-25

D E N T I N E T H I C K N E S S O N M O N O M E R D I F F U S I O N 23

Table 2. The significant differences of HEMA release rate at different timepoints between the groups. The notations above the shaded

cells correspond to the time to the left

RDT 0-4-0.6 mm 1-4-1-6 mm 2.4-2-6 mm 3-4—3-6 mm

14-4 min

144 min

1 day

0-4-0-6 mm1-4-1-6 mm2-4-2-6 mm3-4-3-6 mm

0-4-0-6 mm1-4-1-6 mm2-4-2-6 mm3-4-3-6 mm

0-4-0-6 mm1-4-1-6 mm2-4—2-6 mm3-4-3-6 mm

43-2 min

432 min

3 days

*Significant {P < 0.05); x = not significant.

Iaa

Q

- 2

i

i•

iiY \\ ^

O 0-4 - 0-6 mm RDT• 1-4 - 1-6 mm RDTA 2-4 - 2-6 mm RDTA 3-4 - 3-6 mm RDT

i

-2-5 -2 -1-5 -1 -0-5 0

Time (log 10 (day))0-5

Fig. 4. TEGDMA release rate for various dentine thickness groupsover time.

0-4 - 0-6 mm RDT1-4 - 1-6 mm RDT2-4 - 2-6 mm RDT3-4 - 3-6 mm RDT

-2-5 -2 -1-5 -1 -0-5 0 0-5Time (log 10 (day))

Fig. 5. TEGDMA cumulative release for various dentine thicknessgroups over time.

and from bonding resin-restorative resin combinationsand its movement through dentine in vitro was measuredby Gerzina &• Hume (1994, 1995). It was apparent fromthese studies that dentine very probably exerts itsprotective effect against the cytotoxicity of the twocompounds by reducing their peak concentration in theouter layers of dentine as a consequence of the timetaken for their diffusion across the dentine. The delayin passage muted the peak release rate of TEGDMA intothe pulp space to less than 1 % of the peak direct releaserate (Gerzina & Hume, 1994).

Several investigators have proposed that remainingdentine thickness is an important factor in the protec-

tion of the pulp (Lundy & Stanley, 1969; Stanley,Conti & Graham, 1975; Bergenholtz & Reit, 1980). Withall other factors being equal, for a given chemicalthe diffusion rate should be inversely proportionalto dentine thickness and directly proportional to thefraction of the cross-sectional area of dentine composedof dentinal tubules. In the present study the cavitysurface area (6 mm in diameter), temperature (37°C),depth of restoration (2 mm) and time of light activation(30 s) was the same for all groups of teeth. Our datashow that dentine thickness variation clearly affectedboth diffusion rates (in particular during the first dayafter placement) and cumulative release of the two

© 1997 Blackwell Science Ltd, Journal of Oral Rehabilitation 24; 20-25

24 A. HAMID & W. R. HUME

Table 3. The significant differences of TEGDMA release rate at different timepoints between the groups. The notations above the shadedcells correspond to the time to the left

RDT 0-4-0.6 mm 1-4—1-6 mm 2-4-2-6 mm 3-4-3-6 mm

14-4 min

144 min

1 day

0-4-0-6 mm1-4-1-6 mm2-4-2-6 mm3-4—3-6 mm

0-4-0-6 mm1-4-1-6 mm2-4—2-6 mm3-4-3-6 mm

0-4-0-6 mm1-4-1-6 mm2-4-2-6 mm3-4-3-6 mm

43-2 min

432 min

3 days

*Significant (P < 0.05); x = not significant.

trace molecules. Decreasing dentine thickness markedlyincreased pulpward diffusion rate and total diffusion ofboth monomers from bonding-resin restorations. It isvery likely therefore that substantial reduction ofdentine thickness will reduce the protective effect ofdentine (as was described above) for both HEMA andTEGDMA.

In the early sampling times (up to 1 day alter place-ment) the effects of decreased dentine thickness ondiffusion were more marked for HEMA than TEGDMA.For example, for the 0-14-4 min sampling period thethinnest dentine group (0-4-0-6 mm) had a HEMAdiffusion rate 48 times greater than the thickest group(3.4-3.6 mm), while the relative difference for thehighest TEGDMA diffusion rates between the same twothickness groups was only 2-2 times. For the thinnestgroup of teeth the maximum HEMA diffusion rate was317 times as great as the maximum TEGDMA rate,while for the thickest group it was 14 times as great.

The molecular weight of HEMA is 130-14 and thatof TEGDMA is 286-33. HEMA is also more water-soluble than TEGDMA. The lower, early TEGDMAdiffusion rate might be due both to its larger molecularsize and to a higher degree of polymeric conversion.It was apparent that TEGDMA, which was not presentin the bonding resin, was able to diffuse from thecomposite resin through the polymerized adhesiveresin, the hybrid layer and the resin tags. A similarphenomenon was observed in a separate study(Gerzina &• Hume, in press) where TEGDMA from a

urethane dimethacrylate-based resin diffused througha TEGDMA-Bis-GMA bonding resin. The relativelyhigh and early release of HEMA may indicate thatfree unpolymerized HEMA was present within thedentine tubules or the hybrid layer, even afterpolymerization of the adhesive resin. Residual waterpresent in the hybrid layer and the dentine tubulesmay result in poor polymeric conversion and thereforeincrease monomer release. The differences betweenTEGDMA and HEMA in the degree of influence ofdecreasing thickness on diffusion are most probablyrelated to changes in the properties of the cut dentinesurface as the tissue becomes thinner; larger tubulediameter and increasing overall wetness of the tissueare likely to increase the availability of bondingresin components more than components of thesubsequently placed restorative resin, for reasonsalready outlined. We did not detect Bis-GMA in ourpulp-space samples; this may be due to its relativelyhigh molecular weight, its low water solubility orto a high degree of polymeric conversion in ourexperimental system.

Our data support the hypothesis that remainingdentine thickness influences the concentration andamount of chemicals diffusing through dentine to thepulp space. Thin dentine appears likely to providemarkedly less protection against chemical toxins fromrestorative materials than does thick. It appears fromour data that the effect of dentine thickness variationon diffusion of resin components may be more

© 1997 Blackwell Science Ltd, Journal of Oral Rehabilitation 24; 20-25

D E N T I N E T H I C K N E S S ON M O N O M E R D I F F U S I O N 25

marked for smaller molecular weight substances inthe early hours after placement.

Acknowledgments

We would like to thank Dr K. Saeki for her helpin high performance liquid chromatography and P.Bacchetti for statistical analysis. This work was sup-ported by NIH grant ROl-DE 10331-OlAl.

References

BERGENHOLTZ, G. & REIT, C. (1980) Reactions of the dental pulp tomicrobial provocation of calcium hydroxide treated dentin.Scandinavian Journal of Dental Research, 88, 187-192.

BORGMEIJER, P.J.. KREULEN, CM. . VAN AMERONGEN. W . E . AKERBOOM.

H.B.M. & GRUYTHUYSHN. R.J .M. (1991) The prevalence ofpostoperative sensitivity in teeth restored with class twocomposite resin restorations. ASDC Journal of Dentistry for

Children, 58, 378-383.BRUCE, G.R., MCDONALD. N.J. & SYDISKIS, R.J. (1993) Cytotoxicity of

retrofill materials. Journal of Endodontics, 19, 288-292.FuJiTANi, M., iNOKOSHi. S. & Hos AD A, H. (1992) Effect of acid etching on

the dental pulp in adhesive composite restorations. International

Dental Journal, 42, 3-11.GERZINA, T.M. fr HUME. W.R. (1994) Effect of dentine on release

of TEGDMA from resin composite in vitro. Journal of Oral

Rehabilitation, 21, 4 6 3 ^ 6 8 .GERZtNA, T.M. & HUME, W.R. (1995) Effect of hydrostatic pressure

on the diffusion of monomers through dentin in vitro. Journal

of Dental Research, 74, 369-373.GERZINA. T.M. & HUME, W.R. (1996) The diffusion of monomers

from bonding resin-composite combinations through dentinein vitro. Journal of Dentistry, 24, 125-128.

GERZINA. T., PICKER. K., HOOD, A. & HUME, W. (1991) Toxicity and

quantitative analysis of TEGDMA and composite resin eluates.Journal of Dental Research, 70, 424.

HANKS, C.T., CRAIG, R.G., DIEHL, M.L. & PASHLEY, D.H. (1988)

Cytotoxicity of dental composites and other materials in a newin vitro device. Journal of Oral Pathology, 17, 396-403.

HANKS. C.T.. WATAHA. J.C.. PARSELL. R.R. & STRAWN. S.E. (1992)

Delineation of cytotoxic concentrations of two dentin bondingagents in vitro. Journal of Endodontics, 18, 589-596.

HOOD. A.M. fr HUME, W.R. (1990) Effect of sample thickness on thetoxicity of composite resins in vitro. Journal of Dental Research,

69, 943.HoRSTED, P.B. (1987) Monkey pulp reactions to caries treated with

Gluma Dentin Bond and restored with a microfilled composite.Scandinavian Journal of Dental Research, 95, 347-55.

HUME. W.R. (198 5) A new technique for screening chemical toxicityto the pulp from dental restorative materials and procedures.Journal of Dental Research, 64, 1322-1325.

HUME, W.R.. GERZINA. T.M. & ROUSE. S.R. (1993) TEGDMA

concentration and cytotoxicity in aqueous eluates of resincomposite. Journal of Dental Research, 71, 162.

JOHNSON, G.H.. GORDON. G.E. & BALES, D.J. (1988) Postoperative

sensitivity associated with posterior composite and amalgamrestorations. Operative Dentistry, 13, 66-73.

LuNDY, T. & STANLEY. H.R, (1969) Correlation of pulpalhistopathology and clinical symptoms in human teeth subjectedto experimental irritation. Oral Surgery, 27, 187-201.

NAKABAVASHL N. & TAKARADA. K. (1992) Effect of HEMA on bondingto dentin. Dental Materials, 8, 125-130.

PASHLEY. D.H. (1985) Dentin-predentin complex and itspermeability: Physiologic overview. Journal of Dental Research,

64, 613-620.PASHLEY, D.H. (1990) Clinical consideration of microleakage. Journal

of Endodontics, 16, 70-77.QvisT, V.. STALTZE. K. & QviST. J. (1989) Human pulp reactions to

resin restorations performed with different acid-etch restorativeprocedures. Acta Odontologica Scandinavica, 47, 253-267.

RuYTER. I.E. fr SJ0VIK. l.J. (1981) Composition of dental resin andcomposite materials. Acta Odontologica Scandinavica, 39, 133-146.

STANLEY. H.R.. CONTL A.J. & GRAHAM. C. (1975) Conservation of

human research teeth by controlling cavity depth. Oral Surgery,

39, 151-156.WHITE. J.M.. GOODIS. H.E., MARSHALL. S J . & MARSHALL, G.W. (1994)

Sterilization of teeth by gamma radiation. Journal of Dental

Research, 73, 1560-1567.

Correspondence: Dr W. R. Hume, Department of RestorativeDentistry, University of California, San Francisco, 707, ParnassusAve, San Francisco, CA 94143-0758, U.S.A.

© 1997 Blackwell Science Ltd, Journal of Oral Rehabilitation 24: 20-25