Abstract Studies have reported that dental resin-basedmaterials release substances which have biological lia-bilities. However, some current methods for detectingthese substances may not be adequate to detect biologi-cally relevant concentrations. In the current study, wehypothesized that resin-based materials exhibit cytotoxiceffects and alter cellular function in vitro when high-pressure liquid chromatography (HPLC-UV detection)cannot detect any release of substances. We further hy-pothesized that this release continues even after agingthe samples in artificial saliva. Five types of compositeor compomer materials (Z-100, Tetric Ceram, Dyract AP,Solitaire, and Clearfil AP-X) and one organically modi-fied ceramic material (Definite) were tested after agingin artificial saliva for 0, 7, or 14 days. Cytotoxicity wasassessed using direct contact with fibroblasts and mea-surement of succinic dehydrogenase activity after 48 h ofexposure post aging. Release of substances from the ma-terials was assessed using HPLC with UV detection. Al-tered cellular function was estimated by measuring pro-liferation of MCF-7 cells with sulforhodamine staining.HPLC showed that whereas initial release of substanceswas higher without aging, this release dropped signifi-cantly after 7 or 14 days of aging, and was equivalent tothe Teflon controls after 14 days for four of the materials(Tetric Ceram, Definite, Solitaire, and Clearfil AP-X).Without aging in saliva, all materials had cytotoxicities>50% of the Teflon negative controls. After 14 days ofaging, all materials except the Definite continued toshow severe cytotoxicity. Only the Definite could betested for its ability to alter cellular function because ofthe continuing toxicity of the other materials. This modi-fied ceramic material caused a significant proliferativeeffect on the MCF-7 cells indicating that sufficient sub-stances were released to alter cellular function. We con-

cluded that all of these commercially available resin-based dental materials continue to release sufficientcomponents to cause lethal effects or alter cellular func-tion in vitro even after 2 weeks of aging in artificial saliva.

Key words Cell culture · Resin composite · Compomer ·Biocompatibility · High-pressure liquid chromatography

Introduction

Several studies have reported that dental resin-based re-storative materials release substances which have biolog-ical liabilities [1–3]. These substances have a wide rangeof toxic potencies [4] and their effective toxicity is re-duced but often not eliminated by the presence of dentin[5–7]. Ferracane and Condon [1] have suggested that themajority of the relevant release of substances from com-posites occurs within 24 h, and that component releaseafter 24 h is negligible. Thus, the biological liability ofresin materials has often been assumed to be limited tothe first 24 h. Investigations of the potential toxicity ofthese materials have therefore focused mainly on the ear-ly intervals [2,3]. Despite this focus, other research indi-cates that a substantial portion of resin-based materialsremains unconverted to polymer and therefore may beavailable for longer-term release [8].

Clinically, the pulpal or gingival effects of resin-based restorative materials have been suspected [9–11]but never proven. The role of microleakage of bacteria orbacterial products has often been presumed to be themain cause of any pulpal sensitivity [12,13]. Neverthe-less, there are several reasons to suspect that resins mayalter pulpal physiology. First, resin-based materials areplaced on cut and often etched dentin which is perme-able, or more recently, directly on exposed pulpal tissue[14]. Second, studies have shown the ability of these res-ins to diffuse through dentin [15–17]. Finally, other stud-ies have shown that dental resin monomers may havesublethal effects at low concentrations [2,3,18–21] or

J.C. Wataha (✉) · F.A. Rueggeberg · C.A. Lapp · J.B. LewisP.E. Lockwood · J.W. Ergle · Donald J. MettenburgSchool of Dentistry, Medical College of Georgia, Augusta, GA 30912–1260, USAE-mail: watahaj@mail.mcg.eduTel.: ++ 706 721 3354, Fax: ++ 706 721 8349

Clin Oral Invest (1999) 3:144–149 © Springer-Verlag 1999

O R I G I N A L A RT I C L E

J.C. Wataha · F.A. Rueggeberg · C.A. LappJ.B. Lewis · P.E. Lockwood · J.W. ErgleD.J. Mettenburg

In vitro cytotoxicity of resin-containing restorative materialsafter aging in artificial saliva

Received: 19 April 1999 / Accepted: 15 July 1999

Downloaded from http://www.elearnica.ir

145

may synergize to increase their effective toxicity [22].Identification of the connection between biological ef-fects and release of components from resin-based materi-als has been hampered by the relative insensitivity ofmethods used to detect resin release. Currently, mosthigh-pressure liquid chromatography methods with ultra-violet detection (HPLC-UV) have detection limits ofabout 1 µg/ml [2,3,23]. The biological effects of resinsmay occur well below these limits.

In the current study, we hypothesized that the biologi-cal liabilities of resin-based dental restorative materialsextend beyond 24 h, are below the level of detection ofcurrent HPLC-UV methods, and may alter cellular func-tion. To test these hypotheses, we (1) compared the cyto-toxicity of several resin-based materials before and afteraging them in artificial saliva for up to 14 days, (2) relat-ed the cytotoxicity to mass released from the materials,and (3) investigated the ability of less toxic materials toalter cellular growth rates.

Materials and methods

Materials and sample fabrication

Composites and compomers (Table 1) were selected frommaterials currently used in dental practice. In addition, anew material, classified as an organically modified ce-ramic (Definite; Degussa AG, Hanau, Germany), wastested. All specimens were treated aseptically to preventthe need for disinfection before biological testing. Speci-mens were discs 1 cm in diameter and 1 mm thick with asurface area of 1.88 cm2. A stainless-steel mold was usedto fabricate the samples. The resin material was carefullypacked into the mold to avoid voids. Mylar; (DuPont,Wilmington, Delaware, USA) was added to both sides ofthe material, and the resin was sandwiched between glassslides with clamps. The surface was therefore protectedsomewhat from oxygen inhibition. The specimens werecured with a composite curing light (Optilux 500, Demet-ron-Kerr, Danbury, Conn. USA) with an intensity of atleast 500 mW/cm2 for one minute from each side of theclamped apparatus. The intensity of the light was testedthrough the glass and Mylar with a radiometer (Demet-ron-Kerr). After curing, any gross flash of material wasremoved with a sterile scalpel blade, and specimens werestored in sterile 24-well tissue culture plates in the dark atroom temperature (25°C) for less than 7 days before test-ing. Trimming of the samples was kept to a minimum tolimit alteration of the surface of the cured specimens.

Aging of the samples

Specimens were divided into three groups. Specimens inthe first group were tested without aging in artificial sali-va. The second and third groups used specimens aged for7 or 14 days, respectively, in 1 ml/specimen of artificialsaliva containing 4.1 mM KH2PO4, 4.0 mM Na2HPO4,24.8 mM KHCO3, 16.5 mM NaCl, 0.25 mM MgCl2,4.1 mM citric acid, and 2.5 mM CaCl2 [24]. The pH ofthe artificial saliva was adjusted to 6.7, then sterilized byfiltration before use. All specimens were made simulta-neously and were tested in parallel in the same experi-ment. The non-aged specimens were stored for 14 dayswhile the parallel samples were aged in artificial saliva.

Cytotoxicity

After aging in the artificial saliva, specimens (n=6) weretested for cytotoxicity by placing them in cultures ofBalb/c 3T3 fibroblasts for 48 h (American Type CultureCollection, Rockville, Md. USA). The Balb cells wereused based on previous use [25] and their conformationwith specifications of the International Standards Orga-nization [26]. Cells were plated at 12,500 cells/cm2 in24-well format. One milliliter of cell-culture medium[DMEM, 3% NuSerum, 200 mM glutamine, 125 units/mlpenicillin, 125 µg/ml streptomycin, all from Gibco(Gibco BRL, Grand Island, N.Y. USA)] was plated perwell. Cells were routinely maintained in this same medi-um. Thus, the surface area to volume ratio was1.88 cm2/ml, which was midrange in the ISO require-ments (0.5–6.0 cm2/ml). Immediately after plating thecells, the resin specimens were added into the centers ofthe wells and secured to prevent any movement. Con-trols were Teflon (Tf) discs of equivalent size to adjustfor the physical presence of the sample and blank wellswith cells but no samples to ensure cell behavior wasnormal. Direct contact tests were chosen rather than test-ing extracts from the material to minimize the need tochange medium on the cells and to allow for direct inter-action of any leachable substances from the materialswith the cells.

After 48 h of exposure, the viability of the cells in thewells was assessed by measuring succinic dehydroge-nase (SDH) activity using the MTT method. This meth-od has been shown to be one reproducible way of quick-ly assessing the cytotoxic effects of biomaterials in vitro[25,27]. SDH activity was expressed as a percentage ofthe Tf negative controls and groups were compared us-

Table 1 Materials, manufac-turers, and lot numbers Product Manufacturer Lot no.

Definite Degussa AG (Hanau, Germany) DEA 227T1; 209a

Dyract AP Dentsply Detrey (Baar, Germany) 9706000733Z100 3 M Dental Products (St. Paul, MN USA) 19971002Solitaire Heraeus-Kulzer (Wehrheim, Germany) 00–12–31Clearfil AP-X Kuraray Dental (Japan) 0392A6Tetric Ceram Vivadent Ets. (Schaan, Lichtenstein) 916079

a Two different lot numberswere tested

ing ANOVA and Tukey multiple comparison intervals (αpreset to 0.05).

High-pressure liquid chromatography

After aging in artificial saliva, specimens for use withhigh-pressure liquid chromatography (HPLC; n=2) wereadded to fresh artificial saliva (1 ml/specimen) for 48 h,after which the saliva was collected and analyzed for therelease of components by HPLC (pump: Beckman Mod-el 110 A, Fullerton, Calif. USA; injector: Altex Technol-ogies Model 210, Santa Clara, Calif. USA; precolumnand column: Supelguard LC-18 and Supelcosil LC18 5micron 15×4.6 mm, Supelco, Belefonte, Pa. USA; UVdetector: Perkin Elmer Model LC75, Norwalk, Conn.USA). Thus, the assessment for released componentswas done at the same time as the cytotoxicity evaluation.HPLC methods used isocratic conditions with an aceto-nitrile-water (70/30% v/v) as the mobile phase and thedetector set to 208 nm. Isocratic conditions were used toobtain maximum resolution of peaks. The areas under allvisible peaks were integrated electronically to approxi-mate the total mass of the released substances. Mass re-lease was compared to Tf negative controls using AN-OVA and Tukey intervals at the 14 day elution time. AN-OVA and Tukey intervals were also used to detect signif-icant changes in mass release with aging time for eachmaterial.

Cellular growth tests

Only materials and conditions that caused less than 50%suppression of SDH activity were tested because suble-thal effects were moot in materials with higher cytotox-icities. Specimens (n=6) were aged for either 7 or 14days in artificial saliva, then placed into a cell-culturemedium containing DMEM (phenol-red free) and 10%human serum stripped of estrogens with dextran-coatedcharcoal [28] for an additional 7 days. The medium ex-tracts (not the artificial saliva extracts) were then placedonto MCF-7 cells (BUS strain received from C. Sonnen-schein, Tufts University, Boston Mass.) for 7 days.MCF-7 cells were used because these cells are known toproliferate in an estrogenic environment [20,28] andcould therefore be used to assess altered cellular func-tion caused by substances released from the material.Cells were established at 5000 cells/cm2 in 24-well for-mat for 24 h. The plating medium was then carefully re-moved and the wells were washed with Hanks bufferedsalt solution. Medium extracts from the resin specimenswere then added. After incubation for 7 days, cell num-ber was estimated using sulforhodamine B staining in96-well format. The optical density of cell-free blanksof the staining solution were subtracted from all values[29]. The results were expressed as the optical densityof the sulforhodamine stain, and compared statisticallyusing ANOVA and Tukey multiple comparison inter-vals.

146

Because of the system complexity and potential forartifacts with the cell-growth experiments, numerouscontrols were used. Positive controls were 10–9 mol/l17β-estradiol (E2), a naturally occurring estrogen knownto cause growth of the MCF-7 cells, and 10–6 mol/l bis-phenol A (BPA), a resin component known to have weakestrogenic effects [20]. Negative controls were (1) cellculture medium that was conditioned in the 24-well dish-es for 7 days to control for release of substances from thedishes, (2) medium from artificial saliva-conditionedwells for 7 or 14 days, and (3) 0.5% (v/v) ethanol in me-dium to control for the concentration of ethanol in the E2or BPA solutions. Teflon was not used as a negative con-trol because we were primarily interested in any effectsfrom the composite materials. In the proliferation test,the true negative control was the cell culture dish itself,since substances from the dishes may have leached intothe medium to alter the proliferation of the MCF-7 cells.In this sense, the Tf samples would have confounded thecontrol of the system.

Results

The cytotoxicity of all materials was greater than 50%(cell dehydrogenase activity was less than 50% of Tfcontrols) when no aging in artificial saliva was used(Fig. 1). After aging for 7 days in saliva, the Definite im-proved significantly from 48% cell activity to 80%. Al-though several other materials showed some increase incell activity (Z-100, Clearfil), none of these increaseswas statistically significant. After aging for 14 days inartificial saliva, Z-100, Tetric, and Clearfil all improvedsignificantly. However, all remained severely cytotoxic

Fig. 1 Cytotoxicity of samples after no aging (0-day), aging for 7days in artificial saliva (7-day), or aging for 14 days in artificialsaliva (14-day). Cytotoxicity was measured in terms of cell activi-ty as a percentage of the Teflon negative controls. Thus, a low cellactivity indicates a high cytotoxicity. Error bars indicate ±1 stan-dard deviation (n=6). Within each material, identical lowercaseletters indicate no statistical differences among the aging times(ANOVA, Tukey, α=0.05). Among materials, identical uppercaseletters on the 14-day values indicate no statistical differences be-tween the various materials and the Teflon negative controls

147

with cell activities <25% of the Teflon negative controls.Dyract and Solitaire showed no improvement. The Defi-nite improved to 90% cell activity, although this increasewas not significantly different from aging for 7 days. At14 days, all materials except the Definite still caused sta-tistically less cell activity than the Teflon.

For the HPLC studies, the release of mass from thenon-aged specimens varied among materials (Fig. 2).Z-100, Tetric, and Dyract showed the most mass release,Solitaire’s release was intermediate, while the Definiteand Clearfil had the least release (at or below the detec-tion limit of the technique). After aging for 7 days, allmaterials showed a significant drop in the mass released.In some cases, this drop was 70–80% of the non-agedvalues (Z-100, Tetric). After aging for 14 days, the massrelease continued to drop for all materials, although thedecreases weren’t always statistically different fromthose for 7-day aging. After aging for 14 days, the Defi-nite, Tetric, Solitaire, and Clearfil were all statisticallyequivalent to the Teflon controls in terms of mass releaseestimated by this method. In Fig. 2, the reduction inmass loss from the Tf samples was statistically signifi-cant but probably not meaningful because values lessthan 250,000 pixel units were below the detection limitsof this HPLC system. Similarly, the mass release for theDefinite and the Clearfil show statistical changes, butthese changes may not be meaningful.

The Definite material was the only material selectedfor the cell-growth test because all other materials re-mained severely cytotoxic even after aging for 14 days(Fig. 1). The negative controls (medium, ethanol in me-dium, artificial saliva) showed an expected amount ofcellular proliferation over the 7 days of the test (Fig. 3).The bisphenol A and E2 positive controls showed the ex-pected significant increases in cellular proliferation rela-tive to the negative controls. The extracts from the Defi-

nite caused cellular proliferation statistically greater thanthat of the negative controls, but statistically less thanthat of the positive controls. The effect was more pro-nounced for the extracts in the first week, but statistical-ly the first and second-week extracts were equivalent.

Discussion

The current study implies that the cytotoxicity of severalpopular resin-based restorative materials continues evenafter aging for 14 days in artificial saliva (Fig. 1) and oc-curs even when high-pressure liquid chromatographic(HPLC-UV) analysis shows no significant mass releaseversus Tf (Fig. 2). Aging the resin specimens in artificialsaliva significantly reduced the mass released, whichagrees with the work of Ferracane and Condon [1]. How-ever, the current studies contrast with these authors’work because enough components continued to be re-leased after 24–48 h to cause significant and often severecytotoxic effects in vitro. At first glance, the reduction inmass release from the materials would seem to be incon-sistent with their continued cytotoxicity. However, thispremise assumes that the analytical methods have detec-tion limits that are low enough to detect all of the massthat may be released. Furthermore, low levels of resinsmay have cellular effects over longer periods of time.Given the results of the current study, some HPLC meth-ods may not be sensitive enough to detect biologicallyrelevant release from these materials.

Fig. 2 Mass released from the samples after no aging (0-day), ag-ing for 7 days in artificial saliva (7-day), or aging for 14 days inartificial saliva (14-day). Mass release was estimated by the totalpeak area (in pixel units) seen during HPLC analysis. Peak areawas determined electronically. Error bars indicate ±1 standard de-viation (n=2). Within each material, identical lowercase letters in-dicate no statistical differences among the aging times (ANOVA,Tukey, α=0.05). Among materials, identical uppercase letters onthe 14-day values indicate no statistical differences between thevarious materials and the Teflon negative controls

Fig. 3 Results of the cellular proliferation test for the Definitematerial. Cellular proliferation for the various conditions was as-sessed by growth of MCF-7 human breast cancer cells and ex-pressed as the optical density of sulforhodamine B dye staining af-ter 7 days of exposure to the various extracts. The negative con-trols were cell-culture medium (Med), 0.5% ethanol in medium(EtOH), medium from wells that had artificial saliva in the wellsfor 1 week (Sal-1wk), and artificial saliva in dishes for 2 weeks(Sal-2wk). The positive controls were bisphenol A in ethanol(0.5%)-medium and 17β-estradiol in ethanol (0.5%)-medium (E2).Extracts from the Definite samples are shown for 1 (Def-1wk) or 2weeks (Def-2wk). Error bars indicate ±1 standard deviation (n=6).Identical uppercase letters indicate no statistical differences be-tween groups (ANOVA, Tukey, α=0.05)

The current study used unpolished specimens. Al-though polishing may have reduced cytotoxic liability byremoving uncured surface components, we chose to useunpolished specimens for several reasons. First, whenresin materials are placed against dentin in a cavity prep-aration, it is the unpolished material with which oral tis-sues must interact. Second, the unpolished condition rep-resents the ‘worst case’ which is desirable for an in vitroscreening test of this nature. Third, subgingival contactmay be with unpolished surfaces because the restorationcannot be accessed or because open margins may occurat these locations. Finally, the use of unpolished speci-mens has been reported in other studies of this type[1–3].

The detection limit for the HPLC analysis in the cur-rent study was estimated to be near 1 µg/ml for sub-stances such as Bis-GMA and TEGDMA which are like-ly leachables from the tested materials. The current studyhas assumed that peak area in HPLC at 208 nm is a goodestimator of massed released. Technically, this assump-tion may not be accurate if the extinction coefficients ofthe various released components vary widely, or if com-ponents are released which do not absorb well at208 nm. Thus, even after aging for 14 days, the compos-ites with no identifiable peak area may have been releas-ing components not detected by this method. An alterna-tive which is more inclusive is using simple masschange, which has been used by previous authors [1].However, mass change has many confounding problemsincluding water sorption, loss of volatile componentsduring drying, relative insensitivity, and concurrent ad-sorption of mass onto the specimens. Thus, the methodused in the current study appeared reasonable, and wasuseful for comparing mass loss from the same materialsafter different aging intervals. More sensitive methods ofHPLC detect such as mass spectroscopy may improvethe current results.

It should be emphasized that mass loss may not corre-late well with biological effects because of the varyingbiological potencies of released components [2–4]. Wemade no effort to identify the components released fromthe materials in the current study. Since identification ofthe exact components released from these types of mate-rials can be difficult and time consuming, the methodused in the current study gives a quick approximation ofamount of leached mass. The limitations for this methodin predicting exact biological performance are severeand are apparent by comparing Figs. 2 and 3. Overall, bi-ological tests would appear to be more sensitive indica-tors of release from composite materials.

Only the Definite was tested for the effect on cellularproliferation because this material was the only materialwith low cytotoxicity after aging for 14 days. Any suble-thal effects for the other materials would have been ob-scured by their overt toxicity (Fig. 1). The current studyshowed that even when overt cytotoxicity is low, cellfunctions may be altered. Thus, the Definite materialcaused significant cellular proliferation of the MCF-7cells even when the material was slightly cytotoxic (Fig.

3). Other work has established these types of sublethaleffects for metal ions [30] and cytokine secretion frommonocytes [31–33]. The other resin materials in the cur-rent study can only be screened for these types of effectsif their toxicity can be decreased. Future studies that agethese materials for even longer periods might allow theseassessments. The trend toward increasing cellular activi-ty in Fig. 1 for Z-100, Tetric, and Clearfil even after ag-ing for 14 days indicates that longer aging times mightbe informative in this regard. An alternative method ofassessing these materials would have been to serially di-lute extracts from the materials. This is one avenue forfurther investigation.

A question arises about the interpretation of the cellu-lar proliferation test used in the current study. Do the re-sults in Fig. 3 indicate an estrogenic effect for Definite?Cellular proliferation with breast cancer cell lines hasbeen used as an endpoint to screen for estrogenic effectsof resin materials [28]. However, it seems prudent to re-serve judgment about the estrogenicity of Definite sincecellular proliferation may be caused by mechanisms thanestrogenic responses. For example, silver ions have beenshown to cause increased cellular proliferation in culture[30], yet silver ions are not considered xenoestrogens.Thus, strictly equating the estrogenicity of materials withcellular proliferation in one cell type may be inappropri-ate. Additional experiments should be done to test if theproliferation is caused specifically through estrogenicmechanisms. The results of the current study only showthat sublethal effects of released components can bedemonstrated and should be taken into considerationwhen evaluating the biological liabilities of these materi-als.

Conclusions

In summary, we have shown that several commonly usedresin-based restorative materials continue to release bio-logically relevant amounts of mass into artificial salivain vitro even after aging for 2 weeks. This result con-trasts with those of some authors who have indicated thatthe extent of mass release from some of these materialsis negligible after 24 h. We have shown further that someHPLC methods may not be adequate to monitor the bio-logical liabilities of materials, and that biological testsare currently more sensitive and must be considered as aviable means to assess loss of components from resin-based materials.

Acknowledgments The authors thank Degussa Corporation andthe Medical College of Georgia Biocompatibility Program fortheir support of this work.

148

19. Jontell M, Hanks CT, Bratel J, Bergenholtz G (1995) Effectsof unpolymerized resin components on the function of acces-sory cells derived from the rate incisor pulp. J Dent Res 74:1162–1167

20. Schafer TE, Lapp CA, Hanes CM, Lewis JB, Wataha JC,Schuster GS (1999) Estrogenicity of bisphenol A and bisphe-nol A dimethacrylate in vitro. J Biomed Mater Res 45:192–197.

21. Bouillaguet S, Wataha JC, Virgillito M, Gonzalez L, RakichDR, Meyer J-M (1999) Effect of sub-lethal concentrations ofHEMA (2-hydroxyethyl methacrylate) on THP-1 humanmonocyte-macrophages, in vitro. Dent Mater (accepted)

22. Ratansathien S, Wataha JC, Hanks CT, Dennison JB (1995)Cytotoxic interactive effects of dentin bonding components onmouse fibroblasts. J Dent Res 74:1602–1606

23. Bagis YH, Rueggeberg FA (1997) Effect of post-cure tempera-ture and heat duration on monomer conversion of photo-acti-vated dental resin composite. Dent Mater 13:228–232

24. Arvidson K Johansson EG (1985) Galvanic currents betweendental alloys in vitro. Scand J Dent Res 93:467–473

25. Wataha JC, Craig RG, Hanks CT (1992) Precision of and newmethods for testing in vitro alloy cytotoxicity. Dent Mater8:65–71

26. International Standards Organization (1997) Biological evalu-ation of medical devices: Part 5, Tests for cytotoxicity: in vitromethods. 10993–10995, Geneva, Switzerland

27. Pearse AG (1972) Histochemistry: theoretical and applied.Williams and Wilkins, Baltimore, pp 881–920

28. Olea N, Pulgar R, Pérez P, Olea-Serrano F, Rivas A, Novillo-Fertrell A, Pedraza V, Soto AM, Sonnenschein C (1996) Estro-genicity of resin-based composites and sealants used in den-tistry. Environ Health Perspectives 104:298–305

29. Skehan P, Storeng R, Scudiero D, Monds, A, McMahon J, Vis-tica D, Warren JT, Bokesch H, Kenney S, Boyd MR (1990)New colorimetric cytotoxicity assay for anticancer-drugscreening. J Natl Cancer Inst 82:1107–1112

30. Wataha JC, Hanks CT, Craig RG (1991) The in vitro effects ofmetal cations on eukaryotic cell metabolism. J Biomed MaterRes 25:1133–1149

31. Rakich DR, Wataha JC, Lefebvre CA, Weller RN (1999) Ef-fect of dentin bonding agents on the secretion of inflammatorymediators from macrophages. J Endodont 25:114–117

32. Wataha JC, Ratanasathien S, Hanks CT, Sun ZL (1996) In vit-ro IL-1beta and TNF-alpha release from THP-1 monocytes inresponse to metal ions. Dent Mater 12:322–327

33. Schedle A, Rausch-Fan XH, Samorapoompichit P, Franz A,Leutmezer F, Spittler A, Baghestanian M, Lucas T, Valent P,Slavicek R, Boltz-Nitulescu G (1998) Effects of dental amal-gam and heavy metal cations on cytokine production by pe-ripheral blood mononuclear cells in vitro. J Biomed Mater Res42:76–84

149

References

1. Ferracane JL, Condon JR (1990) Rate of elution of leachablecomponents from composite. Dent Mater 6:282–287

2. Geurtsen W, Spahl W, Leyhausen G (1999) Variability of cyto-toxicity and leaching of substances from four light-curing pitand fissure sealants. J Biomed Mater Res 44:73–77

3. Geurtsen W, Spahl W, Leyhausen G (1998) Residual mono-mer/additive release and variability in cytotoxicity of light-curing glass-ionomer cements and compomers. J Dent Res77:2012–2019

4. Hanks CT, Strawn SE, Wataha JC, Craig RG (1991) Cytotoxiceffects of resin components on cultured mammalian fibro-blasts. J Dent Res 70:1450–1455

5. Hanks CT, Wataha JC, Parsell RR, Strawn SE (1992) Delinea-tion of cytotoxic concentrations of two dentin bonding agentsin vitro. J Endodont 18:589–596

6. Bouillaguet S, Wataha JC, Hanks CT, Ciucchi B, Holz J(1996) In vitro cytotoxicity and dentin permeability of HEMA.J Endodont 22:244–248

7. Bouillaguet S, Virgillito M, Wataha J, Ciucchi B, Holz J (1998)The influence of dentine permeability on cytotoxicity of fourdentine bonding systems, in vitro. J Oral Rehab 25:45–51

8. Loza-Herrero MA, Rueggeberg FA, Caughman WR, SchusterGS, Lefebvre CA, Gardner FM (1998) Effect of heating delayon conversion and strength of a post-cured resin composite. JDent Res 77:425–431

9. Schmalz G (1998) The biocompatibility of non-amalgam den-tal filling materials. Eur J Oral Sci 106:696–706

10. Stanley HR (1995) Dental iatrogenesis, part 2. Dentistry To-day, February: 76–81

11. Bergenholtz G (1998) In vivo pulp responses to bonding ofdental restorations. Trans Acad Dent Mater 123–147

12. Brännström M, Nyborg H (1972) Pulpal reaction to compositeresin restorations. J Prosthet Dent 72:181–189

13. Cox CF, Suzuki S, Suzuki SH (1995) Biocompatibility of den-tal adhesives. Can Dent J 23:13–41

14. Tsuneda Y, Hayakawa T, Yamamoto H, Ikemi T, Nemoto K(1995) A histopathological study of direct pulp capping withadhesive resins. Operative Dent 20:223–229

15. Pashley DH (1996) Dynamics of the pulpo-dentin complex.Crit Rev Oral Biol Med 7:104–133

16. Hanks CT, Wataha JC, Parsell RR, Strawn SE, Fat JC (1994)Permeability of biological and synthetic molecules throughdentine. J Oral Rehab 21:475–487

17. Gerzina TM, Hume WR (1995) Effect of hydrostatic pressureon the diffusion of monomers through dentin in vitro. J DentRes 71:369–373

18. Jewett A, Hume WR (1998) HEMA and TEGDMA trigger hu-man lymphocyte cytokine secretion in vitro. J Dent Res77:240 (abstract 1074)