Necessity of bevels for box only Class II composite restorations

Niek J. M. Opdam, DMD, PhD,a Joost J. M. Roeters, DMD, PhD,b Ruud Kuijs, DMD,c and
Rob C. W. Burgersdijk, DMD, PhDd
University of Nijmegen, Nijmegen, The Netherlands
Statement of problem. The tooth preparation of a bevel is recommended to improve marginal quality of
a composite restoration. However, in small Class II restorations, it is unclear if a bevel also contributed to a
better marginal fit.
Purpose. This study investigated the influence of tooth preparation design on microleakage of minimal
posterior Class II composite restorations.
Material and methods. Box-shaped Class II tooth preparations for posterior composite restorations in
maxillary premolars were restored with a total etch technique. The tooth preparations were beveled or non-
beveled and the box prepared at a right angle cervically or additionally excavated. The facial and lingual box
margins were also either beveled or unbeveled. The teeth were thermocycled and immersed in a dye solu-
tion. After sectioning specimens, dye penetration at the facial and palatal margins was recorded.
Results. A bevel-reduced microleakage both at the cervical and ascending walls. Enamel cracks were
observed along certain unbeveled margins as recorded in this study. The additional excavation did not con-
tribute to reduction of microleakage.
Conclusions. Tooth preparation of a bevel is recommended for an optimal marginal seal in small box-type
Class II composite restorations. (J Prosthet Dent 1998;80:274-9.)

CLINICAL IMPLICATIONS
It is recommended to place a bevel at the margins of small box-type Class II composite
restorations to control marginal leakage.

W hen a tooth is restored for the first time, an

adhesive restoration can preserve sound tooth tissue
of restorative materials has great clinical significance.
Sensitivity, bacterial infiltration, and secondary caries
compared with conventional silver amalgam.1 This is may develop from microleakage. An adhesive restora-
particularly true when a tooth is initially restored. tion technique can reduce microleakage particularly
Therefore small carious lesions for the proximal sur- when advanced adhesive systems are selected.8,9 The
faces of premolars and molars should conserve sound dentist has the opportunity to modify the tooth prepa-
tooth tissue. ration design of the carious surface with minimal
Tunnel restorations or box-type (saucer-shaped) removal of sound tooth tissue, especially with previous-
restorations are viable alternatives. However, the tunnel ly unrestored teeth. During removal of dental caries of
restoration has been difficult to control because of a a Class II defect, the dentist has the opportunity to
limited access to the cavity and poor clinical perfor- either extend the cervical outline of the tooth prepara-
mance.2,3 Although clinical studies have been limited, tion to the carious-free dentinoenamel junction or
current investigations have demonstrated a suitable restrict the tooth preparation to sound enamel. The
clinical performance for box-type restorations.4,5 dental caries is later removed at the dentinoenamel
Considerable effort has been made to reduce the junction with a round bur.
microleakage of restorations. The inability of restora- Tooth preparation of a bevel is recommended to
tive materials to seal the interface with tooth structure improve the bonding surface and reduce microleakage.
encouraged microleakage of intraoral irritants. 6 However, tooth preparation of a bevel is arduous in
Bergenholtz et al.7 stated positively that microleakage small cavity preparations when the approximal margins
are in contact with the adjacent tooth. In extensive loss
This investigation was financially supported by the Dutch Dental of tooth structure, tooth preparation of a bevel at a cer-
Association (NMT)-Nieuwegein, The Netherlands.
aAssistant Professor, Department of Cariology and Endodontology.
vical margin may result in removal of remaining enam-
bAssistant Professor, Department of Cariology and Endodontology. el. Another effect of beveling is an enlargement of the
cAssistant Professor, Department of Cariology and Endodontology. tooth preparation, unless unsupported enamel is pre-
dProfessor, Department of Cariology and Endodontology. served.

274 THE JOURNAL OF PROSTHETIC DENTISTRY VOLUME 80 NUMBER 3

OPDAM ET AL THE JOURNAL OF PROSTHETIC DENTISTRY

Fig. 1. Unbeveled tooth preparation (group 1). Fig. 2. Beveled tooth preparation (group 2).

The purpose of this study was to investigate the tooth long axis. The cervical margin ended 2 mm supe-
influence of a bevel on microleakage of minimal Class rior to the enamel-cement junction.
II composite restorations. Teeth were divided into 4 groups of cavity designs,
according to tooth preparations:
MATERIAL AND METHODS
Group 1: Tooth preparation remained as described,
Box-shaped Class II tooth preparations were com- which resulted in a butt-joint outline at the cervical
pleted for 20 extracted sound maxillary premolars, margin and a concave-shaped margin at facial and lin-
which were stored in a 1% chloramine solution after gual walls of the box because of the shape of the dia-
extraction. Each tooth received 2 box tooth prepara- mond stone (Fig. 1).
tions at the mesial and distal surfaces, which were per- Group 2: Tooth preparation was beveled at the
formed with a diamond stone (Meisinger 836012). facial, lingual, and cervical margins. The 45-degree
The width of the box was 3 mm buccolingual and bevel, 1 mm in width, was made with a fine diamond
the axial depth was 1.5 mm at the cervical margin. The stone (Horico FG 539 F 007) (Fig. 2).
axial wall was prepared in a parallel direction with the Group 3: Same procedures as group 1, except at the

SEPTEMBER 1998 275

THE JOURNAL OF PROSTHETIC DENTISTRY OPDAM ET AL

Fig. 3. Unbeveled tooth preparation combined with excava- Fig. 4. Beveled tooth preparation combined with excavation
tion (group 3). (group 4).

cervical wall of the tooth preparation, excavation was with water spray for 20 seconds, then gently dried with
simulated with a round bur (Meisinger HM 1012) a stream of air. Clearfil Photo Bond material (Kuraray,
(Fig. 3). Osaka, Japan) was mixed and applied to the cavity,
Group 4: Same procedures as group 3; however, the thinned with a gentle stream of air, and cured during
tooth preparation was beveled at the facial, lingual, and 10 seconds of light polymerization (Translux, Kulzer,
cervical margins according to the procedure in group 2 Wehrheim, Germany) at an intensity of 600 mW/cm2.
(Fig. 4). Composite (Clearfil Ray Posterior, Kuraray, Osaka,
The teeth were mounted in an artificial jaw and were Japan) was injected in the cavity with a Centrix tip
in contact with adjacent teeth to simulate clinical con- (Hawe Neos, Bioggio, Switzerland), in a composite
ditions. A Lucifix transparent matrix with light-reflect- delivery gun (Caulk - DeTrey Dentsply, Konstanz,
ing Luciwedges material (Hawe Neos, Bioggio, Germany).
Switzerland) was placed around each tooth. Tooth Composite was applied in 2 layers. The first layer was
preparations were etched with 35% phosphoric acid inserted on the floor of the cavity and carefully shaped
(DMG, Hamburg, Germany) for 15 seconds, rinsed against the palatal wall of the cavity. This layer was

276 VOLUME 80 NUMBER 3

OPDAM ET AL THE JOURNAL OF PROSTHETIC DENTISTRY

Table I. Results for dye penetration

0 1 2 3 4

Cervical dye penetration

Group 2–bevel 10
Group 4–bevel with 5 4 1 *
excavation
Group 3–no bevel with 2 3 2 1 2
excavation
Group 1–no bevel 1 1 6 2
Approximal dye penetration
Group 1/3–no bevel 6 12 2
Group 2/4–bevel 19 1
*Results within same level of significance are marked.

cured for 30 seconds through the light-reflecting

wedge and palatal cusp. Subsequently, a second layer
was injected in the cavity to complete the restoration.
This layer was cured for 30 seconds from the buccal
side through the cusp. Finally, the restoration was
cured for an additional 30 seconds from the occlusal
surface (Fig. 5).
The restorations were removed from the artificial
jaw and finished occlusally with a fine grit diamond
stone (830 C 016, Meisinger, Düsseldorf, Germany)
and approximally with Soflex disks (3M, St. Paul,
Minn.). After 24 hours storage in tap water, the teeth
were thermocycled (500 cycles at 5° to 55°C) and coat-
ed with 2 layers of nail varnish with the surfaces of Fig. 5. Restorative procedure. A, Application and polymer-
restorations remaining free. The teeth were then ization of first layer. B, Application and polymerization of
immersed in a dye solution (basic fuchsin) for 24 hours. second layer.
After embedding the teeth in an acrylic resin (Fas-
tacryl), the teeth were sectioned with a 0.5-mm thick
diamond saw (Bronwill, Rochester, N.Y.).
RESULTS
Teeth were sectioned horizontally so the cervical
half of the restorations remained intact. The section Table I presents the microleakage values for various
was inspected from the occlusal surface with a micro- tooth preparation designs. The Kruskal-Wallis ANOVA
scope. Dye penetration at the facial and palatal margins revealed a significant influence of tooth preparation
was scored on a 5-point ordinal rating scale: 0, no leak- design on cervical microleakage. The differences in
age; 1, leakage restricted to enamel; 2, leakage in microleakage between groups were statistically signifi-
dentin not reaching axial cavity wall; 3, leakage to axial cant (Mann-Whitney U-test; P<.05) (Table I).
cavity wall; 4, leakage beyond axial cavity wall into the A bevel at the facial and palatal margins resulted in
pulp. Each tooth was subsequently sectioned mesiodis- 95% of restorations without microleakage. The absence
tally along the fissure system, and cervical dye penetra- of a bevel resulted in 30% restorations without leakage
tion was also scored according to the identical ordinal and these differences were statistically significant
rating scale. (Mann-Whitney U-test, P<.05). Microleakage at the
Dye penetration at the palatal and facial margins ascending wall extended into the dentin in 10% of the
was statistically compared by using a Mann-Whitney restorations.
U-test at P<.05. The data for cervical dye penetration For microleakage at the cervical margin, the cervical
were statistically compared by using a Kruskal-Wallis bevel used in groups 2 and 4 resulted in 75% of the
analysis of variance (ANOVA) at P<.05. Statistically restorations without microleakage at the cervical mar-
significant differences in microleakage between sepa- gins. Group 2 without microleakage produced the best
rate groups were computed with the Mann-Whitney results. However, an unbeveled cervical outline result-
U-test at P<.05, corrected to .01 for multiple group ed in only 15% of restorations without leakage and 70%
testing. of the restorations exhibited microleakage in dentin.

SEPTEMBER 1998 277

THE JOURNAL OF PROSTHETIC DENTISTRY OPDAM ET AL

The differences of the groups with additional excava- the cavity wall.17 In addition, the narrow-shaped area
tion (3 and 4) with comparable groups without excava- of the bevel may have hindered the flow of composite
tion (1 and 2) were not statistically significant. The pat- to the cervical outline. However, a suitable adaptation
tern of microleakage recorded for groups with excava- to the cavity wall was discovered in this study because
tion was commonly dye penetration through a the selected composite had a medium consistency and
horizontal fracture between enamel prisms adjacent to was injected in the cavity.18,19
the cervical margin of the restoration. Enamel cracks This study indicated that a bevel is recommended
inducing microleakage were also observed in restora- when a tooth preparation is made for a minimal Class
tions of group 1. II composite restoration. However, in a 5-year clinical
study, different responses were not recorded for Class
DISCUSSION
II composite restorations with or without a bevel.20 An
Research has focused on alternatives to dental silver 11-year clinical study of beveled and nonbeveled Class
amalgam in the treatment of dental caries. Composites III restorations reported no difference in survival
are available that have recorded satisfactory results in rate.21 Therefore a clinical study that compares beveled
clinical studies. A 3-year clinical study with the same and nonbeveled small Class II box-restorations is nec-
composite selected for this study demonstrated excel- essary to establish the clinical relevance of this study.
lent responses of the restorations without signs of post-
CONCLUSIONS
operative sensitivity.10
Insertion of a composite restoration has become less The following conclusions were drawn from this
complicated because of the development of total etch study.
adhesives that can be applied to dentin and enamel. 1. Tooth preparation of a bevel-reduced microleak-
The adhesive in our study has demonstrated adequate age at both the cervical and ascending walls of a small
performance in other in vitro and in vivo studies.8-10 Class II box-type composite resin restoration.
Therefore, in our study, excellent results were possibly 2. Tooth preparation of a bevel is recommended for
restricted to small Class II cavities. The beveled restora- both cavity designs in this study.
tions recorded virtually no leakage. The experimental
tooth preparation designs in this study were used to REFERENCES
investigate the influence of either a simplification of 1. Walls AW, Murray JJ, McCabe JF. The management of occlusal caries in
tooth preparation design (no bevel) or a design that permanent molars. A clinical trial comparing a minimal composite
preserved cervical enamel. restoration with an occlusal amalgam restoration. Br Dent J 1988;164:
288-92.
Tooth preparation of a bevel appeared to reduce the 2. Do Amaral Zenkner JE, Baratieri LN, Monteiro S, Caldeira De Andrada
microleakage both at the cervical and ascending walls. MA, Vieira LCC. Clinical and radiographic evaluation of cermet tunnel
This was in agreement with other studies that also restorations on primary molars. Quintessence Int 1993;24:783-91.
3. Strand GV, Nordbø H, Tveit AB, Espelid I, Wikstrand K, Eide GE. A 3-year
observed the phenomenon of enamel cracks along an clinical study of tunnel restorations. Eur J Oral Sci 1996;104:384-9.
unbeveled margin11,12; these cracks were also reported 4. Nordbø H, Leirskar J, Von der Fehr FR. Saucer-shaped cavity preparation
by Crim and Chapman.13 The additional excavation for composite resin restorations in Class II carious lesions: three-year
results. J Prosthet Dent 1993;69:155-9.
did not contribute to a reduction of microleakage. In 5. Kreulen CM. A study on assessing quality aspects of class II restorations.
similar research, a retentive groove at the cervical wall [Thesis.] Amsterdam: ACTA; 1992. p. 75-81.
was discovered beneficial to reduce microleakage.14 6. Brännström M. Infection beneath composite resin restorations: Can it be
avoided? Oper Dent 1987;12:158-63.
This experimental tooth preparation with a cavosurface 7. Bergenholtz G, Cox CF, Loesche WJ, Syed SA. Bacterial leakage around
design resembling the cavity preparation of group 4 dental restorations: its effect on the dental pulp. J Oral Pathol 1982;11:
reduced microleakage.15 However, in one study, a 439-50.
8. Abdalla AI, Davidson CL. Comparison of the marginal integrity of in vivo
chemical-curing composite was selected, with a differ- and in vitro Class II composite restorations. J Dent 1993;21:158-62.
ent shrinkage pattern.15 Another factor may be that, in 9. Abdalla AI, Davidson CL. Comparison of the marginal axial wall integrity
this study, more unsupported enamel remained, as frac- of in-vivo and in-vitro made adhesive class V restorations. J Oral Rehabil
1993;20:257-69.
tures in the enamel close to the margin of the restora- 10. Wendt SL, Leinfelder KF. Clinical evaluation of Clearfil photo posterior: 3-
tion were observed. These fractures were also verified year results. Am J Dent 1992;5:121-5.
by Oilo and Jørgensen.16 11. Lösche GM, Neuerburg CM, Roulet JF. Die adhäsive Versorgung konserv-
ativer Klasse II Kavitäten. Deutsch Zahn Zeitschr 1993;48:26-30.
Marginal leakage in the ascending walls in this study 12. Hilton TJ, Schwartz RS, Ferracane JL. Microleakage of four Class II resin
was restricted to enamel in 95% of the cases, but a bevel composite insertion techniques at intraoral temperature. Quintessence Int
did affect diminished microleakage. Evidently, 1997;28:135-44.
13. Crim GA, Chapman KW. Reducing microleakage in Class II restorations:
microleakage was more pronounced along the cervical an in-vitro study. Quintessence Int 1994;25:781-5.
margin. The results may also have been influenced by 14. Shahani DR, Menezes JM. The effect of retention grooves on posterior
the consistency of the composite selected and the mode composite resin restorations: an in vitro microleakage study. Oper Dent
1992;17:156-64.
of application. Excessive manipulation of the compos- 15. Porte A, Lutz F, Lund MR, Swartz ML, Cochran MA. Cavity designs for
ite may have also resulted in a reduced adaptation to composite resins. Oper Dent 1984;9:50-6.

278 VOLUME 80 NUMBER 3

OPDAM ET AL THE JOURNAL OF PROSTHETIC DENTISTRY

16. Oilo G, Jørgensen KD. Effect of beveling on the occurrence of fracture in Reprint requests to:
the enamel surrounding composite resin fillings. J Oral Rehabil 1977;4: DR. N. J. M. OPDAM
305-8. DEPARTMENT OF CARIOLOGY AND ENDODONTOLOGY
17. Hansen EK, Asmussen E. Efficacy of dentin-bonding agents in relation to UNIVERSITY OF NIJMEGEN
application technique. Acta Odontol Scand 1989;47:117-20. PO BOX 9101
18. Opdam N, Roeters JJ, Peters TC, Burgersdijk RC, Teunis M. Cavity wall NL 6500 NIJMEGEN
adaptation and voids in adhesive Class I resin restorations. Dent Mater THE NETHERLANDS
1996;12:230-5.
19. Opdam N, Roeters JJ, Peters TC, Burgersdijk RC, Kuyijs RH. Consistency Copyright © 1998 by The Editorial Council of The Journal of Prosthetic
of resin composites for posterior use. Dent Mater 1996;12:350-4. Dentistry.
20. Wilson NH. Performance of occlusion in butt-joint and bevel-edged 0022-3913/98/$.500 + 0. 10/1/91495
preparations: five-year results. 4. Dent Mater 1991;7:92-8.
21. Qvist V, Strøm C. Eleven-year assessment of Class-III resin restorations
completed with two restorative procedures. Acta Odontol Scand
1993;51:253-62.

Effect of heating delay on conversion and strength of a

Noteworthy Abstracts post-cured resin composite
of the Loza-Herrero MA, Rueggeberg FA, Caughman WF, Schuster
Current Literature GS, Lefebvre CA, Gardner FM. J Dent Res 1998;77:426-31.

Purpose. This investigation examined the effect of delayed and post-cure heating (after initial
light-polymerization) on the strength and monomer conversion of 1 commercially available com-
posite.
Material and Methods. A total of 160 disk-shaped specimens (10 × 1 mm thick) were fabricat-
ed from a commercially available composite material (Herculite XRV, Shade A3, Kerr/Sybron).
All specimens were initially light-polymerized for 60 seconds on each side in an aluminum mold
using a controlled light source. Two types of specimens were fabricated. One type was used as a
“light-polymerized only” control, and the other was subjected to post-cure heat treatment for
7 minutes at 100°C. All specimens were coded and stored at room temperature in a light tight
container until needed. The following time intervals between initial light-polymerization and
post-cure treatment of the specimens were 5 and 30 minutes, and 6, 24, 48, 72, 96, and
120 hours. Ten test specimens and 10 control specimens were used for each time interval. After
the designated time delay, the unheated and post-cured composite specimens were subjected to
testing to determine biaxial flexural strength. Ten specimens per test group were then analyzed
for monomer conversion using infrared spectroscopy. Analyses of variance (α = 0.05), Fischer’s
Protected LSD tests, Dunnett’s t tests, and Student’s t tests were used to analyze the data.
Results. Both delay time and type of treatment revealed a significant effect upon the biaxial flex-
ure strength of the specimens, as did the interaction of these parameters. Control specimens
seemed to gain strength with increasing time. This was confirmed by linear regression analysis.
No such correlation existed between strength and delay time for post-cure heated specimens. The
only time intervals that demonstrated greater strength values for the post-cured specimens com-
pared with the unheated control group specimens were the 5- and 30-minute delays. Both delay
time and type of treatment significantly affected the observed cure values (monomer conversion)
and the interaction of these variables. Cure values of unheated specimens were observed to
increase with time after initial light-polymerization. Monomer conversion of heated specimens
also declined with delay in heat application.
Conclusions. Post-cure heating did not significantly affect the biaxial flexural strength of the
composite tested, and time delay of heat application was not a significant factor in strength. Post-
cure heating of the composite significantly affected the increase in monomer conversion, and the
time after which heat was applied following initial light-polymerization was influential. Early
application of heat, after initial light-polymerization, resulted in higher monomer conversion val-
ues. For the greatest amount of polymerization, post-cure heating should be applied within
6 hours after light-polymerization. 17 References.—DL DIXON

SEPTEMBER 1998 279