at the gingival
Alton M. Lacy, MS, PhD, DDS,a Craig Wada, DDS,b Weiming Larry Watanabe, BSd
Du, DDS,C and
University of California, School of Dentistry, San Francisco, Calif. This in vitro study investigated the effect of dental surface treatments on the seal of porcelain and resin laminate veneers to tooth structure. Nine groups of 20 teeth were prepared for labial veneers-six groups with the apical margin of the veneer in dentin, two groups with the veneer wholly bordered by enamel, and one group with the apical margin coincident with the apical margin of a previously placed glass ionomer restoration. The teeth were treated with a variety of dentin-bonding agents prior to restoration of eight groups with porcelain veneers and one group with Visio-Gem microfilled resin veneers. After placement, the veneers were stored in water for 4 days, then subjected to thermocycling before silver staining and sectioning. Results showed no leakage around the margins of the veneers wholly bordered by enamel. Complete leakage was found around the glass ionomer restorations, and nearly total leakage was found along the resin-dentin interface of the resin veneers. The porcelain veneers that extended onto dentin showed variable but limited leakage for four of five of the dentin-bonding treatments. (J IPEoSTHETDENT 1992;67:7-io.) T he gingival '. margin of a porcelain or resin laminate veneer may be placed within enamel parallel to the cementoenamel junction (CEJ) of a tooth, along or close to the CEJ or in the dentin and cementum of the root. The ability to bond restorations to etched enamel by using composite resin has been demonstrated in a large number of studies and is now routine in clinical dentistry. The relative inability to seal restorative resins to dentin is also well
known.1-8 Porcelain veneers whose margins are completely in enamel appear to seal well and provide excellent clinical service. Much less is known about the gingival seal of laminate veneers terminating in dentin. Previous in vitro research with porcelain and resin inlays3 indicates that when small thicknesses of bonding resin are used to secure prefabricated inlays in class II preparations, the contraction stresses are not suffmient to form a gap, and there is no microleakage. After thermocycling, however, restorations made from materials with high coefficients ofthermal expansion (microfilled resins, for example) do show mar-
ginal leakage as the initial seal is broken by differential thermal expansion, whereas ceramic inlays, which have a thermal expansion similar to natural teeth, seem to retain their seal.
This study investigated the effectiveness of certain dentin-bonding agents in inhibiting gingival microleakage at the margins of porcelain and a microfilled resin laminate veneer bonded to extracted teeth, to provide insight into specific treatments that will be clinically reliable.
Nine groups of 20 extracted anterior teeth of indeterminate age,which had been stored in a 0.02 % thymol solution, were randomly selected and prepared for anterior veneers. The preparations are depicted in Fig. 1. These groups were designated A through I. Groups A through F were prepared with the gingival margin 1 mm apical to the CEJ, exposing an average of 4.4 mm of dentin at the gingivofacial aspect of each prepared tooth. For group G, class V Ketac-Fil (ESPE-Premier, Norristown, Pa.) glass ionomer cement restorations (1 mm in in&al gingival width and approximately 1 mm deep) were placed in dentin at the CEJ of each tooth. Veneer preparations were then made similar to groups A through F with the gingival margin coincident with the apical margin of the cement restoration. Groups H and I were prepared with the gingival margins of the veneers 1 mm incisal to the CEJ, providing a border entirely in enamel.
Presented before the Pacific Coast Society of Prosthodontists, meeting, Napa, Calif. aAssociate Professor, Department of Restorative Dentistry. bGraduate student, Postgraduate Prosthodontics, Department of Restorative Dentistry. CVisiting Research Scientist, Department of Restorative Dentistry. dResearch Technician, Department of Restorative Dentistry. 10/1/296&S TEE
G-Cera porcelain (G-C International, Scottsdale, Ariz.) veneers were made on refractory dies for all groups except group F, which received Visio-Gem (ESPE-Premier) microfilled resin veneers. All porcelain veneers were etched with 9.5% hydrofluoric acid (Porcelain Etch, Ultradent Products Inc., Salt Lake City, Utah) for 4 minutes before application of a silane coupling agent (Scotchprime, 3-M St. Paul, Minn.). The Visio-Gem resin veneers were inter7
LACY ET AL
Table I. Tooth treatments Bet A B c D E F G I-3 I
Mirage Bond Tenure solution Gluma Scot&bond dual-cure ScotchbondMirage Bond Ketac Fi! + Scotchbond dual cure 40% Phosphoric acid Mirage Bond
Type of veneer
Porcelain onto dentin Porcelain onto dentin Porcelain onto dentin Porcelain onto dentin Porcelain onto dentin Visio-Gem onto dentin Porcelain onto glass ionomer Porcelain onto enamel Porcelain onto enamel
nally sandblasted with 50 pm aluminum bonding to the test teeth.
All veneers were bonded to the teeth with G-Cera light-
cured resin after treatment of each set of teeth according to the scheme illustrated in Table I. For set A, the exposed dentin and enamel were simultaneously treated with Mirage Bond (Mirage Porcelains Systems, Kansas City, Kan.) followed by Scotchbond dualcure resin (3-M) before application of the porcelain veneer with G-Cera resin. Set B was treated with Tenure (Den-Mat, Santa Maria, Calif.) solution and Visar-Seal (Den-Mat) resin according to manufacturer’s directions after acid etching of the visible exposed enamel with 40 % phosphoric acid for 20 seconds (Ultraetch, Ultradent Products Inc.) followed by rinsing and drying. Set C was treated with Gluma and Gluma resin (Columbus Dental/Miles Inc., St. Louis, MO.), according to the manufacturer’s directions, after similarly etching, rinsing, and drying of the enamel. Set D was treated with Scotchbond dual-cure dental adhesive after etching, rinsing, and drying of the exposed enamel. (The resin was blown to a thin film with a stream
of clean compressed air and was not light-cured before veneer placement.) Set E was treated with Scotchbond- resin (3-M) after etching of the exposed enamel, but varying the manufacturer’s directions by eliminating the precure of Scotchbond-2 resin before seating of the veneer. (Following the manufacturer’s directions to precure the resin in this case would prevent seating of the veneer). 8
of tooth preparations
for each set
for sets A through
Set F (the Visio-Gem veneers) was bonded after simultaneous treatment of the dentin and enamel with Mirage Bond followed by Scotchbond dual-cure resin (which was not precured before veneer placement). Set G (containing the glass ionomer cement restorations) was bonded after acid etching of the enamel and glass ionomer restorations for 10 seconds with 40 % phosphoric acid gel before application of Scotchbond dual-cure resin and veneer placement with the G-Cera composite resin. Set H (no exposed dentin) was etched 20 seconds with 40% phosphoric acid, rinsed and dried before bonding of the porcelain veneers with Scotchbond dual-cure resin and G-Cera composite resin. Set I (no exposed dentin) was treated with Mirage Bond according to manufacturer’s instructions before application of Scotchbond dual-cure resin, and the veneers were bonded with G-Cera composite resin. All veneered teeth were stored for 48 hours in water at 3’7” C, then thermocycled between 5’ and 55’ C for 4 hours with a 60-second dwell at each temperature. The teeth were then soaked 4 hours in 50 wt% aqueous solution of silver nitrate. The stained teeth were sectioned to observe the degree of microleakage followed by 2-hour immersion in photographic developer (Insta-Neg, Microcopy, Kennesaw, Ga.). The extent of silver nitrate penetration was assessed with a measuring microscope (Gaertner Scientific Corp., Chicago, Ill.) for all sets except group F. This group was excluded from quantitative comparative analysis because of the different geometric pattern of leakage around the cement restoration.
RESULTS The average extent of leakage seen in each set is shown in Table II. None of the specimens in groups I-I and I, where the margin of the veneer was totally in enamel, showed any evidence of leakage at the gingival (or other) margins. All veneers placed over glass ionomer restorations (group G) whose apical margins were coincident with the gingival margin of the veneer, showed extensive leakage around the glass ionomer restoration with a good seal at the ionomer/ resin interface (Fig. 2). Of the six remaining
groups, the extent of gingival leak-
age is illustrated in Fig. 3. Extensive leakage was found at the dentinal interface of all of the Visio-Gem veneers (group F), but comparatively little leakage around the porJANUARY
Fig. 2. Typical leakage patterns observed. Le jt: Complete seal with no leakage; micldle: total leakage along dentin-resin interface (most typical of Visio-Gem veneers); right: leakage around glass ionomer restorations (typical of all specimens). Table
Average leakage Leakage
A B c D E F G I-I H
0.39 i 0.47 0.39 2 0.30 0.51 -t 0.56 0.51 + 0.76 1.92 + 1.18 2.94 * 1.98 Totally around cement 0 0
Mirage Bond Tenure solution Gluma Scotchbonddual-cure ScotchbondMirage Bond Ketac Fil 40% Phosphoric acid Mirage Bond
celain veneers (group A), which were similarly bonded to the teeth. Analysis of variance of the data indicated no significant differences between sets A through D. Set E and F are significantly different from sets A through D and from each other.
ISCUSSION The absence of leakage in groups H and I, where the veneers were bounded only by enamel, confirms the tenacity and stability of the resin-etched enamel bond and clearly indicates the clinical superiority of this situation wherever it can be achieved. The phosphoric acid and the 2.5% nitric acid found in Mirage Bond appear equally effective in providing the resin-to-enamel seal. The extensive leakage seen around the Visio-Gem veneers (group F) is consistent with previous observations on Visio-Gem inlays with subgingival margins and is probably the result of breaking of the seal with the tooth, under the influence of differential thermal expansion and contraction as the restoration is subjected to thermocycling. THE
Type of veneer
Porcelain onto dentin Porcelain onto dentin Porcelain onto dentin Porcelain onto dentin Porcelain onto dentin Visio-Gem onto dentin Porcelain onto glassionomer Porcelain onto enamel Porcelain onto enamel
These results suggest that microfilled resin veneers are contraindicated where total enamel bonding cannot be achieved. With the exception of Scotchbond- the dentin-bonding agents used in this in vitro study seem equally effective in limiting microleakage at the gingival margin of veneers bonded to dentin. Although other studies have shown superior strength of resin-to-dentin bonds when the agents that remove the smear layer are used (in comparison to Scotchbond dual-cure, which does not remove smear iayer), the similar effectiveness of both types of agents used here suggests that bond strength and microleakage are not necessarily related. Further, the manufacturer suggests that if dentin bonding is to be achieved, the resin of Scotchbond- should be precured into a hard, relatively thick film before the addition of overlying composite. This requirement cannot be applied to a veneer system without compromising the fit of the veneer. The reason for the failure of Scotchbond- to effectively inhibit microleakage in this study is not known. We have observed that the acidic primer of Scotchbond9
2.5 z 2 02 b f;
F z z" L
NEW MIRAGE BOND
(VISI~;GEM) NEW MIRAGE BOND
SCOTCHBOND DUAL CURE
Fig. 3. Extent of stain penetration for groups A through F (all veneers extending onto dentin).
is caustic to mucosal tissue, further limiting the clinical use of this product. Because glass ionomer restorations require significant clinical effort and provide no resistance to microleakage, their placement prior to porcelain veneers seemsto offer no advantage except that slow dissolution of the restoration will locally fluoridate the interface, possibly resisting recurrent caries. Similar results indicating severe microleakage around glass ionomer restorations have been reported by other investigatorslo-l3 Finally, it should be noted that there is no verified correlation between in vivo behavior and in vitro results obtained from studies such as this. It is well known that materials that show poor results in the laboratory often can provide long clinical service. Standardized in vivo testing is difficult and often impossible. Simulation of oral conditions in the laboratory are equally difficult. However, laboratory studies that compare materials under similar conditions give us insight into probable clinical behavior and provide guidance for clinical trials.
1. Hansen E. Effect of three dentin adhesives on marginal adaptation of two light-cured composites. Stand J Dent Res 1986;94:82-6. 2. Gross J, Retief D, Bradley E. Microleakage of posterior composite restorations. Dent Mater 1985;1:7-10.
3. Fayyad M, Shortall AC. Microleakage of dentine-bonded posterior composite restorations. J Dent 1987;15:67-72. 4. Eakle W, Ito R. Effect of insertion technique on microleakage in MOD composite resins [Abstract]. J Dent Res 1987;66:293. 5. Blunck U, Roulet JF. In vitro marginal quality of dentin-bonded composites in class V cavities. Quint Int 1989;20:407-12. 6. Davidson CL, DeGee AJ, Feilzer A. The competition between composite dentin bond strengths and the polymerization contraction stress. J Dent Res 1984;63:1396-9. 7. Suzuki M, Gwinnett AJ, Jordan R. Relationship between composite resins and dentin treated with bonding agents. J Am Dent Assoc 1989; 118:75-7. 8. Davila JM, Gwinnett AJ, Robles JC. Marginal adaptation of composite resins and dentinal bonding agents. J Dent Child 1988;55:25-8. 9. Lacy A, Zhang K, Koh A, Wiltshire WA, Watanabe L. Marginal microleakage around class II resin and Dicer inlays [Abstract]. J Dent Res 1988;67:196. 10. Wenner KK, Fairhurst CW, Morris CF. Hawkins IK, Ringie RD. Microleakage at root restorations. J Am Dent Assoc 1988;117:825-5. 11. Shortall AC, Fayyad MA, Williams JD. Marginal seal of injection molded ceramic crowns cemented with three adhesive systems. J PROSTHET DENT 1989;61:24-7.
12. Shortall AC, Baglis RL, Grundy JR. Marginal seal of class V composite/glass ionomer sandwich restorations. Rest Dent 1988;4(4)Nov:80-87. 13. Thornton JB, Retief DH, Bradley EL. Marginal leakage of two glass ionomer cements: Ketac-Fil and Ketac-Silver. Am J Dent 1988;1:35-8. Reprint requests to; DR. ALTON M. LACY SCHOOL OF DENTISTRY, BOX 0758 U~'IVERSITY OF CALIFORNIA SAN FRANCISCO, CA 94143