Fracture load and mode of failure of ceramic veneers with different

preparations
Jacopo Castelnuovo, DDS, MSD, a Anthony H. L. [jan, Dr Dent, DDS, PhD, b Keith Phillips, D M D ,
MSD, c Jack I. Nicholls, PhD, d and John C. Kois, D M D , MSD e
University of Washington, School of Dentistry, Seattle, Wash.

S t a t e m e n t o f p r o b l e m . Fracture is a clinical failure modality for ceramic veneers. Whether design of

tooth preparation can affect the strength of ceramic veneers remains controversial.
P u r p o s e . This in vitro study evaluated fracture load and mode of failure of ceramic veneers, with 4 tooth
preparation designs, that were bonded on extracted human maxillary central incisors. Identical parameters
were also measured on unrestored intact teeth for comparison.
M a t e r i a l a n d m e t h o d s . Fifty maxillary central incisors were randomly divided into 5 equal groups. Each
group was assigned a diftkrent tooth preparation design: (1) no incisal reduction, (2) 2 mm incisal reduc-
tion without palatal chamfer (butt joint), (3) 1 mm incisal reduction and 1 mm height palatal chamfer, (4)
4 mm incisal reduction and 1 mm height palatal chamfer, and (5) unrestored (control). Forty teeth were
prepared to accommodate ceramic veneers of equal thicl~ness and incisocervical length. Stone dies were fab-
ricated and veneers made from IPS Empress ceramic. Ceramic veneers were bonded and all teeth mounted
in phenolic rings with epoxy resin. Fracture loads were recorded with a mechanical testing machine.
R e s u l t s . Mean f?acture loads (SD) in kgfwere as follows: group 1, 23.7 (6.11); group 2, 27.4 (9.63);
group 3, 16.4 (3.44); group 4, 19.2 (6.18); and group 5, 31.0 (10.38). Modes of failure were also analyzed
for both ceramic veneers and teeth. One-way ANOVA with multiple comparisons revealed 3 significant sub-
sets: groups 1-2-5, groups 4-1, and groups 3-4 (P<.05). Groups 1 and 2 had no ceramic veneer fractures;
group 3 had 3 ceramic veneer fractures, and group 4 had 6 ceramic veneer fractures.
C o n c l u s i o n . Groups 1 and 2 recorded the greatest fracture loads that were comparable to an unrestored
control. (J Prosthet Dent 2000;83:171-80.)

A study conducted in partial fulfillment of the requirements for the

degree of Master of Science in Dentistry, and supported in part Ceramic veneers, w h i c h have b e c o m e a p o p u l a r ,
by the C.N.R. Grant AI97.00112.04. w e l l - a c c e p t e d , a n d p e r c e p t i v e d e n t a l p r o c e d u r e , were
Presented before the International Association for Dental Research, i n t r o d u c e d t o d e n t i s t r y d u r i n g t h e late 1 9 2 0 s a n d
Nice, France, June 1998; the American Academy of Esthetic Den-
tistry, Philadelphia, Pa., August 1998; The American College of 1930s.1, 2 T h e clinical survival rate o f ceramic veneers
Prosthodontists, San Diego, Calif., September 1998; and Italian b o n d e d t o e n a m e l has b e e n predictable. 3 O b s e r v a t i o n
Academy of Prosthetic Dentistry (AIOP) Research Forum; recipi- p e r i o d s o f ceramic veneers r e p o r t e d in t h e l i t e r a t u r e
ent "Best Oral Presentation" Award, Bologna, Italy, November range f r o m 18 m o n t h s u p t o 15 years. 3-5 O n e s t u d y
1998. r e p o r t e d t h a t 93% o f 3 5 0 0 ceramic veneers placed over
aGraduate Prosthodontics, Department of Restorative Dentistry.
bFormer Professor and Director of Biomateria[s Research, Depart- a 15-year p e r i o d were r a t e d as successful, whereas 7%
ment of Restorative Dentistry; Emeritus Professor and Consultant were c o n s i d e r e d failures. 3
in giomaterials Research, School of Dentistry, koma Linda Uni- Ceramic veneers arc indicated for teeth with m o d e r a t e
versity, Loma Linda, Calif. d i s c o l o r a t i o n 6 8 caused by tetracycline, fluoride, age, 9
cActing Assistant Professor and Director, Graduate Prosthodontics, and amelogenesis impcrtkcta. 1 Thus, ceramic veneers arc
Department of Restorative Dentistry.
dprofessor, Department of Restorative Dentistry. c h o s e n to p r o v i d e excellent esthetics. 10-12 C e r a m i c
eAffiliate Clinical Associate Professor, Graduate Prosthodontics, veneers can also be selected for the restoration o f trau-
Department of Restorative Dentistry. m a t i z e d , f r a c t u r e d , and w o r n d e n t i t i o n (Fig. 1). 7 9

FEBRUARY2000 THE JOURNAL OF PROSTHETIC DENTISTRY 171

THE JOURNAL OF PROSTHETIC DENTISTRY CASTELNUOVO ET AL

Fig. 2. Five test groups. Horizontal line represents load

applied consistently 2.5 mm from incisal edge. Unsupport-
ed ceramic (mm) is indicated for each group.

ceramic veneers during the clinical observation pcriod

up to 15 years. 3 Different tooth preparation designs for
ceramic veneers arc described in the literature. Clyde
and Gilmour, 6 and Hui ct al ll dcscribcd the thathered
incisal edge tooth preparation, the incisal 0.5 to 1.0 m m
bevel preparation, the intraenamel or "window" tooth
preparation in which 1 m m of incisal edge is preserved,
and the overlapped incisal edge tooth preparation. Wein
berg I4 suggested a 1-mm incisal reduction with round-
Fig. 1. A, Fractured maxillary central incisors after comple- ed line angles for improved translucency of a veneer.
tion of orthodontic treatment. B, Bonded ceramic veneers
Sheets and Taniguchi 1° described a tooth preparation
restore both function and esthetics.
with a chamfer for adequate porcelain thickness and with
a r o u n d e d incisal edge and lingual hea W chamfer.
According to Calamia, 1 a tooth preparation that incor-
Abnormal tooth anatomy or malposition can also be cor- porated an incisal overlap was preferable because the
rected with bonded ceramic laminates.LS,9 relative porcelain veneer was stronger and provided a
Ceramic veneers are contraindicated for edge-to- positive seat during cementation. designes
edge and cross bite occlusal relationships because o f Clinical cohesive ceramic fractures have occurred
excessive stress during functionA ° It has also been mainly at the incisal edge of the veneer because o f
r e p o r t e d that ceramic veneers should be avoided in greater stress. 3 It was believed that a palatal chamfer was
patients with heavy occlusion, extreme facial version, necessary to strengthen ceramic veneers.l°, is Unfortu-
p o o r oral hygiene, and severe dentinal demineraliza- nately, most of the data regarding the clinical behavior of
tion, or t o o t h fluoridation, n Delamination, inadver- different tooth preparation designs originated from
tent pulpal damage, periodontal irritation, and unnat- anecdotal reports. It remains controversial whether dif-
ural appearance are potential complications o f ceramic ferent tooth preparation designs can affect fracture
veneers. 12 Ceramic veneers should theoretically be strength of ceramic veneers or whether one configura-
exposed to minimal occlusal loads. 9 Toh et a113 report- tion o f tooth preparation is superior to another.
ed that ceramic laminates are indicated exclusively to The purpose of this in vitro study was to measure the
restore esthetics and n o t function. However, Fried- fracture load o f 4 designs of t o o t h preparations tbr
man 12 reported that ceramic veneers not only provide ceramic veneers. The configuration o f the incisal edge
suitable esthetics, but also reliable functional strength. and the amount of unsupported incisal ceramic were the
Therefore, they can be used to provide anterior guid- independent variables. Unrestored intact teeth were the
ance by restoring appropriate incisal length. control, and the mode of failure was also recorded.
Failures associated with ceramic veneers are related to
MATERIAL AND METHODS
fracture, microleakage, and debonding. 3 Fracture alone
accounted for 67% o f the total failures recorded for The materials selected tbr the fabrication of ceramic

172 VOLUME 83 NUMBER 2

CASTELNUOVO ET AL THE JOURNAL OF PROSTHETIC DENTISTRY

Table I. Materials used for the fabrication of ceramic veneers

Material Manufacturer Batch number

Reprosil type II-medium viscosity L.D. Caulk, Milford, De[. 950112

Triad VLC Dentsply, York, Pa. --
Die Keen Heraeus Kulzer Inc, Irvine, Calif. --
Cement Spacer Belle de St Claire, Kerr, Romulus, Mich. 013302
Die Lube J.M. Ney Co, Degussa, S Plainfield, N.J. --
Proart Wax Williams-lvoclar-Vivadent, Amherst, N.Y. 226017-631

Table II. Materials used for the fabrication of ceramic veneers

Material Manufacturer Batch number

Ultra-Etch 35% phosphoric acid UItradent Products Inc., S Jordan, Utah --

IPS Empress System Ivoclar, Schaan, Liechtenstein --
IPS Ceramic Etching Gel Ivoc[ar, Schaan, Liechtenstein 701405
IPS Empress Cem-Kit Ivoc[ar, Schaan, Liechtenstein
Variolink Base (white) 560794
Variolink Base (yellow) 604959
Catalyst 614243
Liquid Strip (glycerine gel) 560259

veneers are presented in Tables I and II. Fifty extracted USA) were selected to refine the preparation. All tooth
intact human maxillary central incisors with 10-mm preparations were completed entirely in enamel, without
anatomic crown length, h o m o g e n e o u s mesiodistal sharp line angles. Cervical finish lines were developed at
width, and faciopalatal thickness were selected. Each the cementoenamel junction (CEJ) so the final ceramic
t o o t h was free o f dental caries or restorations. The veneers were 10 mm in length, to reproduce dimensions
teeth were cleaned and stored in distilled water at room of the anatomic crown of the tooth.
temperature from the day o f extraction until testing.
Impression making and master die fabrication
Teeth were randomly assigned to 5 groups o f 10
specimens. Each o f 4 groups was assigned a different Impressions for the 40 prepared teeth were made
tooth preparation for ceramic veneers (Fig. 2): with a medium viscosity polyvinyl siloxane impression
1. no incisal reduction with feathered incisal edge; material (Reprosil, LD Caulk, Milford, Del.). Custom
2. 2 mm incisal reduction without palatal chamfer acrylic trays (Triad VLC; Dentsply, York, Pa.) were
(butt joint); used, and each tray allowed an impression o f 3 pre-
3. 1 mm incisal reduction with 1 m m height palatal pared teeth. Impressions were cast in vacuum mixed die
chamfer; stone (Die Keen, Heraeus Kulzer Inc USA, Irvine,
4 . 4 mm incisal reduction with 1 mm height palatal Calif.), according to the manufacturer with respect to
chamfer; and water-powder ratio and mixing time. Stone dies were
5. unrestored intact teeth as control. recovered from impressions, and 2 coats o f die spacer
(Cement Spacer; Belle De St Claire, Kerr, Romulus,
Tooth preparation
Mich.) were painted 0.5 m m short o f the finish lines o f
The facial surface of the teeth in groups 1 through 4 the preparations. Two coats o f die lubricant (Die Lube;
was prepared to accommodate veneers of equal thickness. Degussa, South Plainfield, N.J.) were then applied to
Facial reduction was 0.3 mm at the cervical third and each die (Fig. 3).
0.5 mm at the middle and incisal thirds. 16 Incisal and
Ceramic veneer fabrication
palatal reduction for each group was performed as
described (Fig. 2). Tooth preparations were extended to All 40 veneers in groups i through 4 were waxed to
include interproximal contacts with rotary instruments a uniform thickness o f 0.6 mm with beige wax (Proart
(Brasseler USA, Savannah, Ga.) and a water coolant. Self- Wax; Williams, Ivoclar-Vivadent, Amherst, N.Y.) and
limiting depth-cutting disks o f 0.3 mm (834-31-016, sprued. Ceramic veneers were then pressed after invest-
Brasseler USA) and 0.5 mm (834-31-021 Brasseler USA) ment. All procedures were p e r f o r m e d with IPS
were used to define the depth cuts, then 1.2 mm cham- Empress materials and protocol (Ivoclar, Schaan,
fer diamond burs (KS1;35005-31-52-012, Brasseler Liechtenstein). After divestment, the veneer fit was ver-

FEBRUARY 2000 173

THE JOURNAL OF PROSTHETIC DENTISTRY CASTELNUOVO ET AL

Fig. 3. Stone dies o1 different tooth preparation designs.

Fig. 4. Testing with Instron machine.

ified with green aerosol (Occlude; Pascal, Bellevue,

Wash.) sprayed over a stone die. High spots on the
ceramic veneers w e r e removed with a diamond medium Cement was a combination o f 25% Variolink ycllow
grit r o u n d bur ( 8 0 1 - 1 1 - 0 0 9 , Brasseler USA). All base, 25% Variolink white base, and 80% catalyst. This
ceramic veneers were then reduced to 0.3 mm at the cement was hand mixed and applied to both prepared
cervical third and 0.5 mm at the incisal two thirds with teeth and ceramic veneers. The ceramic veneers were
green stones (661-120, Brasseler USA). seated on the prepared teeth with light finger pressure
The IPS Empress layering technique was selected tbr and excess ceraent was removed with an explorer
all treated ceramic veneers. Ceramic was "cutback" before an oxygen barrier was applied to the margins
before both a wash firing and application o f enamel (Liquid Strip; Ivoclar). Photopolymerization was per-
ceramic and relative firing cycles were performed. The formed for 40 seconds for facial and palatal margins of
amount o f cutback was 0.2 m m facially at the incisal each ceramic veneer. Margins were then finished with
two thirds, and 0.5 mm incisally for ceramic veneers in finishing diamond burs (379F-31-023; 132F-31-008,
groups 2, 3, and 4. Enamel ceramic was then applied. Brasseler USA).
For group 1, the cutback was perfbrmed only facially
Specimen testing
because incisal reduction was not programmed for that
tooth preparation design. Ceramic veneers were then The 50 maxillary incisors were mounted individually
finished with diamond burs ( 8 6 3 - 1 1 - 0 1 6 , Brasseler in 1 × 1-in. phenolic rings with epoxy resin (Buehler
USA), and their dimensions standardized again after Ltd, Lake Bluff; Ill.) with the long axis parallel to center
measurement with an electronic Boley gauge caliper line of the ring. Each tooth was suspended in the mid-
(Mitutoyo Corp, Kawasald, Japan) for the height and a dle o f the ring by means of a no. 1 paper clip (ACCO
dial caliper (Mitutoyo) fbr thickness. Final dimensions USA Inc, Wheeling, Ill.) that engaged the tooth at the
for all ceramic veneers were 0.3 mm and 0.5 mm thick- CEJ and rested on the edges of the ring. When the axis
nesses at cervical third and incisal two thirds respec- o f the tooth was positioned correctly, as judged by 2
tively, and 10 mm in length. The veneers were then investigators, epoxy resin was poured in the ring. All
glazed in a ceramic oven ( C o m m o d o r e II-VPF; specimens were embedded up to 2 mm below the CEJ
Jelenko, Armonk, N.Y.). to simulate a natural biologic width.17a 8
The fracture loads (kgf) were determined using a
Bonding ceramic veneers
universal testing machine (Instron Corp, Canton,
The 40 prepared teeth were acid etched tbr 15 sec- Mass.). Load was applied at a 90-degree angle to the
onds with 35% phosphoric acid gel (Ultra-Etch; Ultra- lingual surface o f the test tooth. This orientation was
dent Products, South Jordan, Utah) and thoroughly standardized by securing the phenolic rings in a
rinsed with water for 30 seconds. Ceramic veneers were m o u n t i n g jig (Fig. 4). The load was consistently
treated with fluoridric acid IPS ceramic etching gel applied at 2.5 mm fi:om the incisal edge. A customized
(Ivoclar) for 1 minute, in accordance with the manu- plunger with round tip and a fixed lateral stainless steel
facturer's instructions, and rinsed with water for 30 sec- arm that engaged the incisal edge was used for this pur-
onds. IPS Empress Cem Kit (Ivoclar) was used to bond pose (Fig. 5). The plunger was attached to the Instron
the ceramic veneers to prepared teeth according to the load cell, and crosshead speed was 0.5 m m / m i n .
manufacturer's recommendations. Modes o f failure were macroscopically assessed and

174 VOLUME 83 NUMBER 2

CASTELNUOVO ET A L THE JOURNAL OF PROSTHETIC DENTISTRY

3 o ;~! ................
z z
fi ....
2s~i ,,
Fracture 20 71
Load !.
(Kgf) ::¢
Io / !
, {

5 >~ {

0 ~ m ..........
GP | GP2 GP3 GP4 GP5

Preparation Design

Fig. 6. Mean (SD) fracture load of different designs for tooth

preparations.

Table III. M e a n (SD) m a x i m u m v e n e e r s f r a c t u r e l o a d (kgf)

Preparation design X SD

5 31.0 [ 10.38
2 27.4 9.63
1 23.7 [ 6.11
4 19.2 I 6.18
3 1 6.4 3.44

Instron crosshead speed: 0.5 mm/min.

Means connected by vertical lines are not significantly different.

Fig. 5. Custom-made plunger.

Table IV. Frequency of veneers' failure modality (n=10 per
group)
Failure modality
analyzcd after spccimcn tcsting for both vcnccrs Bond failure
(1 = intact; 2 = f~acture; 3 = bond failure + fracture) Preparation design Intact Fracture + fracture Total
and tecth (1 = intact; 2 = coronal fracture; 3 = cervical
1 10 0 0 10
fracture; 4 = root fracture). 2 10 0 0 10
Statistical analysis 3 7 3 0 10
4 4 5 1 10
Fracture load data were log-transformed 19 to con- 5 (control) . . . .
firm homogeneity o f variances, as verified by the Fmax Chi square = 15.1935; df = 6; prob = 0.0188; 0~=.05.
test. A 1-way analysis o f variance (ANOVA) with
Duncan's multiple comparisons test was p e r f o r m e d
for this transformed data. The failure mode frequen-
Fracture l o a d
cies were analyzed with the chi-square test. A level o f
significance o f ~=.05 was satisfied during the entire Group 2 (2 m m o f unsupported incisal ceramic and
statistical analyses. butt joint) yielded a greater f?acture load than groups 3
and 4 (veneer designs with 1.0 m m and 4.0 m m o f
RESULTS
unsupported incisal ceramic both with palatal chamfer,
Table III and Figure 6 present the fracture load respectively). This was not significantly different from
means and standard deviation for the 4 designs o f group 1 (feathered incisal edge veneer design).
ceramic veneers and the unrestored control. Tables IV G r o u p 1 (feathered incisal edge) and group 4
and V present the failure modality frequency for (4.0 m m o f unsupported incisal ceramic and palatal
veneers and fbr teeth, respcctively. One-way ANOVA chamfer) recorded f?acture load values that were not
disclosed significant differences between mean fracture significantly different. Groups 2 ceramic veneers
loads (Table VI; P=.0011). The chi-square test results (2.0 m m o f u n s u p p o r t e d incisal ceramic and b u t t
were significant for both ceramic veneer (P=.0188) and joint) and 1 (feathered incisal edge) were comparable
tooth mode of failure (P=.0196 ) (Tables IV and V). in fracture load to the control (group 5).

FEBRUARY 2000 175

THE JOURNAL OF PROSTHETIC DENTISTRY CASTELNUOVO ET AL

Table V. Frequency of teeth failure modality (n=l O) Table VI. Summary of analysis of variance

Failure mode Source of Sum of Mean of

variation df squares squares f P
Preparation Coronal Cervical Root
design Intact fracture fracture fracture Total Between groups 4 0.457764 0.114441 5.50504 .0011
1 0 4 2 4 10
Within groups 45 0.935478 0.0207884
2 0 5 1 4 10
Total 49 1.39324
3 0 5 3 2 10 c~=.05.
4 2 7 0 1 10
5 0 2 0 8 10
ing. In our study, ceramic veneers were reduced to 0.3
Chi square = 24.1152; df = 12; prob = 0.0196; 0~-.05.
m m and 0.5 m m at the cervical and incisal third sur-
faces, respectively. This procedure allowed tbr conserv-
ative tooth preparation and a more predictable bond
M o d e o f failure
that was restricted to an enamel surface.
Ceramic veneers were consistently intact (100%) Silanc coupling agents have c o m m o n l y rendered
after testing for groups 1 and 2 t o o t h preparation etched ceramics more wettablc by composite cement to
designs. Forty percent of intact ceramic veneers were improve the b o n d between the 2 materials. 3,21-2s In
recovered for group 4 (4.0 m m u n s u p p o r t e d incisal our study, etched ceramic veneers were bonded direct-
ceramic and palatal chamfer), which r e c o r d e d the ly to etched enamel without a silane coupling agent
greatest n u m b e r o f fractures a m o n g the groups because the manufacturer did not specifically recom-
(P=.0047). Group 3 ceramic veneers recorded 30% o f m e n d its application using Variolink resin cement. Dur-
t?actures, but difference in frequency of mode o f failure ing testing, only 1 partial d e b o n d associated with a
was not significantly different from groups 1 and 2 ceramic veneer fracture was recorded (Table IV), which
(P=.1546). N o significant difference in ceramic veneer represented only 1% o f total ceramic veneer failures.
failure modality was recorded between groups 3 and 4 Simulation of a periodontal ligament was unneces-
( P = . 6 7 7 5 ) . Unrestored teeth in group 5 exhibited a sary in this study because the progressive load applied
significantly greater n u m b e r o f root fractures compared to the coronal portion o f the embedded tooth would
with teeth in the other groups (P=.0196). not have been mitigated by interposition o f a softer
medium between the root of" the tooth and surround-
DISCUSSION
ing epoxy resin. An interposed soft m e d i u m would
Tooth preparations for bonded restorations, includ- have been meaningful during an impact fracture test
ing ceramic veneers, should bc restricted to enam- when a blow was delivered to a specimen. Because of
el 10,12,16 because extensive exposure o f dentin can the natural anatomic variations observed in h u m a n
reduce bond strengths 7 and encourage microleakage. 2° maxillary incisors, it was not plausible to use a single
Many have suggested a minimal thickness for t o o t h mounting jig to position teeth in phenolic rings with
preparations o f 0.5 m m tbr ceramic v e n e e r s . 1,2,6,7,14 the long axis parallel to the center line. Therefore, per-
Ferrari et al 7 reported that the enamel of 114 sectioned sonal judgment o f 2 researchers was used when posi-
anterior teeth was 0.3 to 0.5 m m at the gingival third, tioning each specimen.
0.6 to 1.0 m m at the middle third, and 1.0 to 2.1 m m Other studies that compared fracture strengths of dif-
at the incisal third surfaces, and observed that enamel ferent designs for ceramic veneer tested the specimens
at the gingival third might be totally removed to pre- by loading the veneer-tooth system directly at the incisal
vent an overcontourcd ceramic restoration. In our in edge and in a direction parallel to the long axis o f the
vitro study, 40 t o o t h preparations a c c o m m o d a t e d t o o t h . l l, 24 The orthognathic interincisal angle being
veneers with a 0 . 3 - m m gingival third thickness and a 135 degrees, 25 stresses that affect maxillary ceramic
0 . 5 - m m middle third thickness. 16 All ceramic veneers veneers during mastication and protrusive mandibular
in our study were bonded to etched enamel. excursions are not usually directed parallel to the long
Wax patterns for the 40 ceramic veneers were fabri- axis of a tooth. Elevated loads generated during para-
cated at a thickness o f 0.6 m m to allow for accurate function are not applied in this direction either. These
pressing of ceramic ingots with the IPS Empress systcm parafunctional loads are derived from occlusal strengths
(Ivoelar). This m a n u f a c t u r e r ' s r e c o m m e n d a t i o n that can be 6 times greater than those recorded for non-
ensured a precisc marginal fit for the restorations. parafunctional patients, 26 and are deleterious for most
There was no concern about thickness o f the wax pat- dental restorations. Such loads are usually directed
tern in relation to strength o f the pressed ceramic toward the palatal surface of maxillary incisors and are
veneer. Therefore, it was acceptable to reduce the not parallel to the long axis of a tooth.
thickness o f pressed ceramic veneers to that o f conven- In comparison, forces applied on incisal edges dur-
tional feldspathic laminates betbre finishing and bond- ing function were markedly reduced in magnitude.

176 VOLUME 83 NUMBER 2

CASTELNUOVO ET AL THE JOURNAL OF PROSTHETIC DENTISTRY

Fig. 8. Examples of failed spe(:imens from group 3.

strongest, II and the recorded ftacturc loads wcrc toni

parablc to al~othcr invcstigation. -~7 Tccth rcstorcd with
that saint venccr dcsign and tested with facial impact
also disclosed a mean fiacturc energy comparable to
unrestorcd maxillary incisors. 2s Conversely, the
amount of" unsupportcd ccralnic may have been ulti-
matcly the weakness fi)r thc inordinate n u m b e r of"
vcnccr fractures recorded ff)r group 4 (60%). The rcla
tivcly high fi'acturc load values in this group may have
Fig. 7. Orthognathic interincisal angle. Red a r r o w indicates bcen the result of the improved physical properties of"
horizontal load applied by manclibular incisors. the lcucitc-reinfbrccd ceramic. 2°
The palatal chalnfer prepared fbr 2 of the 3 ccramic
veneer designs with incisal overlapping (groups 3 and
Both functional and parafunctional loads applied on 4:) appeared to bc thc weak feature even for ccramic
palatal surfaces move the ceramic veneers facially. 22 veneer designs with unsupported incisal ceramic less
Ceramic is more susceptible to failure when exposed to than 2 m m (group 3). Fractured specimens in group 3
tensile loads. 3 For these reasons, in our study only the displayed cracks that extended from the palatal chamfer
horizontal c o m p o n e n t o f load applied by mandibular to the fhcial surface o f the ceramic veneer (Fig. 8). In
incisors on the palatal surface of maxillary incisors was addition, fractures in ceramic veneers o f g r o u p 4
considered when positioning specimens for testing with consistently involved the palatal chamfer. One study
an Instron machine (Fig. 7). Thus, veneers were loaded discovered no diff}~rence in fracture load for mandibu-
at a 90-degree angle to the long axis o f the tooth. This lar incisor ceramic veneers with incisal ceramic length
angle also prevented the Instron crosshead from sliding ranging from 0.0 to 2.0 ram. 3° In our study, when the
along the palatal surface o f natural teeth. height o f u n s u p p o r t e d ceramic was increased f r o m
Clinical studies have reported that ceramic veneers 1 m m (group 3) to 4 m m (group 4), no significant dif-
bonded to mandibular incisors exhibited a lower frac- ferences in m o d e o f failure and fracture load values
ture rate because of the less destructive nature of com- were evident between groups. Thus, it is possible that
pressive loads applied on veneer incisal edges. 3 Most the unsupported ceramic was not a critical factor. Con-
clinical fractures have occurred on ceramic veneers versely, the only c o m m o n feature between groups 3
b o n d e d to maxillary incisors, 3 so our study was and 4 was the palatal chamfer.
designed to reproduce a similar clinical condition. Ceramic veneers that remained intact after testing
The ceramic veneer design with 4.0 m m o f unsup- were able to resist the stresses transmitted from the fail-
ported incisal ceramic and palatal chamfer (group 4) ing tooth structure to restorations; and for those spec-
was included in this study as a restorative option for a imens, it was concluded that tooth structure failed first.
fractured tooth. Fracture loads for this ceramic veneer The opposite scenario was recorded for the 2 teeth in
design were comparable to veneer design with feath- group 4 that remained intact after testing: the veneers
ered incisal edge without incisal unsupported ceramic fractured first, protecting the tooth structure. For spec-
(group 1). This design has been reported as one o f the imens in groups 3 and 4, for wflich both veneers and

FEBRUARY 2000 177

THE J O U R N A L OF PROSTHETIC DENTISTRY C A S T E L N U O V O ET AL

designs because an increase in fracture load was

obtained, and the incidence o f fractured ceramic
veneers was reduced to zero for veneers with 2 m m o f
unsupported incisal ceramic and butt-joint configura-
tion. Clinically, ceramic veneers with 1 or 2 m m o f
ineisal reduction and palatal chamfer were successful
and widely performed. Ceramic veneers from group 3
perfbrmed poorly in this study, but this does not imply
that this preparation design does not meet the stan-
dards for clinical longevity and predictability.
Conversely, another study reported that, when the
ratio between ceramic and luting composite thickness
increased, a more favorable ceramic veneer configura-
tion was possible because it prevented ceramic cracks
after thermal cyclic loading. 31 It was also reported that
the optimal ceramic thickness should have been at least
3 times greater than the luting composite. In that same
study, indirect SEM observation o f the failed veneers
revealed that thermocycling-induccd ceramic cracks
were also present on the palatal chamfer surface, in
proximity o f the margin. Clinically, that was the loca-
tion where ceramic can be thin in relation to the thick-
ncss of the luting composite. Finite element analysis
evaluation confirmed the importance of having a suffi-
cient ceramic bulk and minimal luting composite thick-
ness to reduce thermal and polymerization shrinkage
stresses applied to the ceramic. 32 It is critical tbr the
dentist to understand that tooth preparation can affect
the longevity of ceramic veneers. The substitution of a
Fig. 9. A, Butt-joint incisal configuration allows for different
palatal chamtkr with a butt-joint incisal design, which
paths of insertion ranging from facial-palatal to incisocervi-
provides a favorable ceramic/luting composite ratio at
cal. B, Palatal chamfer incisal configuration allows only
incisocervical path of insertion. the palatal surface, can reduce the risk o f postinsertion
palatal cracks caused by shrinkage o f the composite
cement during polymerization and by natural thermal
changes in the oral environment. The most substantial
teeth fractured, it was probable that tooth structure finding of our study was not the poor performance o f
failed first and the restorations were unable to resist the veneers from group 3, but that other tooth preparation
transmittcd stress. designs, such as that o f group 2, showed superior
All ceramic veneers were loaded lingually 2.5 mm results with several clinical advantages.
from the incisal edge, to reproduce an average vertical The butt-joint incisal configuration still permitted
overlap. The assumption was made that patients who preservation o f a peripheral enamel layer around all
brux would start parafunctioning fYom that position margins. The presence o f enamel is critical tbr eliminat-
when the mandible m o v e d protrusively. Functional ing microleakage at the palatal tooth-restoration inter-
mandibular movements also would originate from that face, 15,20 and effectively counteracting shear stresses.
position. The load for specimens in groups 2 and 3, The orientation o f enamel rods at the palatal surface of
with their tooth preparation design, was falling 0.5 mm central incisors approaches the 90-degree angle with
apical to the palatal margin o f the veneers, entirely on the long axis o f the tooth. It is necessary to remove
the t o o t h structure. This may have been caused by both prismatic and interprismatic mineral crystals to
loading directly over the palatal finish line, which produce a more effective enamel etching. 33 Thus, it was
caused a higher fracture ratc for the ceramic veneers in advisable to expose the cross sections of the palatal
group 3 with thin palatal ceramic. The full thickness enamel rods versus their sides during development of a
ceramic butt joint featured in group 2 could be the finish line to enhance the etchant contact with the
rationale for recording 100% intact veneers in that interprismatic enamel. Tooth preparation with a palatal
group. finish line at an angle with t o o t h surface larger than
Our study suggests the elimination o f a palatal 90 degrees, and without cxccssivc reduction o f the
chamfer in overlapping incisal edge ceramic veneer thickness of the palatal ceramic at the tooth-restoration

178 VOLUME 83 NUMBER 2

CASTELNUOVO ET AL THE JOURNAL OF PROSTHETIC DENTISTRY

margin, can be better achieved with a butt-joint incisal parable to more conservative designs in relation to the
design. For teeth with a thin facial-palatal dimension, amount of unsupported incisal ceramic.
the palatal chamfer can expose dentin, which would 4. Ceramic veneers with incisal butt joint offered
result in the same amount of peripheral cnamel created several clinical advantages such as tooth preparation,
by a butt joint. However, with the palatal chamfer the ceramic veneer fabrication, manipulation, and insertion.
finish line would not be at an angle with, but rather We thank Mr Mauro Pasinelli, Graphic Designer, for the com-
parallel to, enamel rods. puter generated illustrations.
Because of the relative low flexural strength of felds-
pathic ceramics, a4 the strength necessary for all-ceram- REFERENCES
ic restorations to survive intraorally and resist incisal
1. Calamia JR. The etched porcelain veneer technique. NY State Dent l
shear loads is provided mainly by a bond to underlying 1988,54:48-50.
tooth structure, and not by the resistance-retention 2. Calamia JR. Clinical evaluation of etched porcelain veneers. Am J Dent
form o f tooth preparation. Ceramic fails at a critical 1989;2:9-15.
3. Friedman MJ. A 15-year review of porcelain veneer failure--a clinician's
strain of 0.1%, 3s so if the bond to tooth structure fails, observations. Compend Contin Educ Dent 1998;19:625-36.
the ceramic fractures because it is brittle. Therefore, the 4. Strassler HE, Nathanson D. Clinical evaluation of etched porcelain
presence of a palatal chamfer would not provide appre- veneers over a period of 18 to 42 months. J Esther Dent 1989;1:21-8.
5. Christensen GJ, Christensen RE Clinical observation of porcelain veneers:
ciable resistance-retention to tooth structure *br the a three-year report. J Esthet Dent 1991 ;3:174-9.
ceramic veneer compared with the resistance-retention 6. Clyde JS, Gi[mour A. Porcelain veneers: a preliminary review. Br Dent J
provided by a sole bond to enamel. Treatment with 1988;164:9-14.
7. Ferrari M, Patroni S, Ba[[eri P. Measurement of enamel thickness in rela-
bonded all-ceramic restorations is dependent on adhe- tion to reduction for etched laminate veneers. Int J Periodontics Restora-
sive dentistry and not macromechanical retention from tive Dent 1992;23:407-13.
geometric tbrm. 8. Tjan AH, Dunn JR, Sanderson IR. Microleakage patterns of porcelain and
castable ceramic laminate veneers. J Prosthet Dent 1989;61:276-82.
The ceramic veneer design with 2 mm of unsup- 9. Christensen GJ. Veneering of teeth. State of the art. Dent Clin North Am
ported incisal ceramic and butt joint compared with 1985;29:373-91.
feathered incisal edge allowed tbr characterization of 10. Sheets CG, Taniguchi 1. Advantages and limitations in the use of porcelain
veneer restorations. J Prosthet Dent 1990;64:406-11.
incisal thirds o f restorations. Compared with t o o t h
11. Hui KK, Williams B, Davis EH, Holt RD. A comparative assessment of the
preparation with palatal chamfer, the butt-joint prepa- strengths of porcelain veneers for incisor teeth dependent on their design
ration was simpler, less time-consuming, and the palatal characteristics. Br Dent J 1991 ;171:51-5.
12. Friedman M. Multiple potential of etched porcelain laminate veneers. J
finish line was precisely reproduced on the master cast.
Am Dent Assoc ] 987;115 (Spec. Iss.):83E-87E.
This translated into easier ceramic veneer fabrication 13. Toh GC, Setcos JC, Weinstein AR. Indirect dental laminate veneers--an
for the dental technician because it was pertbrmed on a overview. J Dent 1987;15:117-24.
fiat surface that provided a more adequate support for 14. Weinberg LA. Tooth preparation for porcelain laminates. NY State Dent J
1989;55:25-8.
ceramic veneers. The path of insertion of these restora- 15. Garber DA, Goldstein RE, Feinman RA. Porcelain laminate veneers.
tions followed different directions ranging from buccal- Chicago: Quintessence Publishing Co; 1988. p. 44.
palatal to incisal-cervical. The risk of fracture for thin 16. Rufenacht CR. Fundamentals of esthetics. 1st ed. Chicago: Quintessence
Publishing Co; 1990. p. 329%8.
unsupported palatal ledges of ceramic was also con- 17. Gargiulo AW, Wentz FM, Orban B. Dimensions and relations of the den-
trolled with this design. Clinically, this is relevant, togingiva[ junction in humans. J Periodontol 1961 ;32:261-7.
especially with simultaneous cementation of multiple 18. Kois JC. Altering gingival levels: the restorative connection. Part h biolog-
ic variables. J Esthet Dent 1994;6:3-9.
units (Fig. 9). After the veneer was bonded to the 19. Howell DC. Fundamental statistics for the behavioral sciences. 3rd ed.
tooth, the butt joint represented a favorable feature in Belmont: Duxbury Press; 1995. p. 310.
the prevention o f palatal ceramic cracks that occur 20. Castelnuovo J, Tjan AH, Liu R Microleakage of multi-step and simplified-
step bonding systems. Am J Dent 1996;9:245-8.
because o f intraoral thermal stresses. 21. Calamia JR, Simonsen RJ. Effect of coupling agents on bond strength of
etched porcelain. J Dent Res 1984;63:197.
CONCLUSIONS 22. Nicholls JI. Tensile bond of resin cements to porcelain veneers. J Prosthet
Dent 1988;60:443-7.
Within the limits of this study, the following conclu- 23. Calamia JR. Etched porcelain veneers: the current state of the art. Quin-
sions were drawn: tessence [nt 1985;16:5-12.
1. Ceramic veneers with 2.0 mm o f unsupported 24. Highton R, Caputo AA, Matyas J. A photoelastic study of stresses on
porcelain laminate preparations. J Prosthet Dent 1987;58:157-61.
incisal ceramic and butt joint and ceramic veneers with
25. Reitz PV, Aoki H, Yoshioka M, Uehara J, Kubota Y. A cephalometric study
feathered incisal edge were the strongest and remained of tooth position as related to facial structure in profiles of human beings:
intact. a comparison of Japanese (Oriental) and American (Caucasian) adults. J
Prosthet Dent 1973;29:157-66.
2. A palatal chamfer did not provide increased
26. Gibbs CH, Mahan PE, Mauderli A, Lundeen HC, Walsh El(. Limits of
strength for ceramic veneers. human bite strength. J Prosthet Dent 1986;56:226 9.
3. On the basis of this in vitro investigation, fi:ac- 27. Andreasen FM, Flugge E, Daugaard-Jensen J, Munksgaard EC. Treatment
tured teeth with up to 4.0 mm of missing tooth struc- of crown fractured incisors with laminate veneer restorations. An experi-
mental study. Endod Dent Traumato11992;8:30-5.
ture can be restored with leucite-reinforced ceramic 28. Stokes AN, Hood JA. Impact fracture characteristics of intact and crowned
veneers. The fracture load values recorded were com- human central incisors. J Oral Rehabil 1993;20:89-95.

FEBRUARY 2000 179

THE JOURNAL OF PROSTHETIC DENTISTRY CASTELNUOVO ET AL

29. Dong JK, Luthy H, Wohlwend A, Scharer P. Heat-pressed ceramics: tech- 35. McLaren EA. All-ceramic alternatives to conventional metal-ceramic
nology and strength. Int J Prosthodont 1992;5:9-16. restorations. Compend Contin Educ Dent 1998;19:307-10.
30. Wall JG, Reisbick MH, Johnston WM. Incisal-edge strength of porcelain
laminate veneers restoring mandibular incisors. Int J Prosthodont 1992;5: Reprint requests to:
441-6. DR JAcoPo CASTELNUOVO
31. Magne P, Kwon KR, Belser UC, Hodges JS, Douglas WH. Crack propen- VIA ARCHIMEDEN.185
sity of porcelain laminate veneers: A simulated operatory evaluation. J 00197 ROME
Prosthet Dent 1999;81:327-34. ITALY
32. Magne P, Verslius A, Douglas WH. Effect of luting composite shrinkage
and thermal loads on the stress distribution in porcelain laminate veneers. Copyright © 2000 by The Editorial Council of The Journal of Prosthetic
] Prosthet Dent 1999;81:335-44. Dentistry.
33. Nakabayashi N, Pash[ey DH. Hybridization of dental hard tissues. 1st ed. 0022-3913/2000/$12.00 + 0. 10/1/104369
Tokyo: Quintessence Publishing Co Ltd; 1998. p. 37-9.
34. Seghi RR, Daher 1-, Caputo AA. Relative flexural strength of dental restora-
tive ceramics. Dent Mater 1990;6:181-4.

Tooth-implant and implant supported fixed partial den-

tures: A 10-year report
Gunne J, Qstrand P, Lindhe T, Borg K, Olsson M. I n t J
Prosthodon t 1999;12:216-21.

Purpose. Although the use of implants for the restoration of partially edentulous patients is
increasing, the possibilities of placing implants in the posterior mandible are limited. The purpose
of this longitudinal study was to evaluate, after a 10-year follow-up, the use of short implants to
support fixed partial dentures (FPDs) in the posterior mandible and to compare implant sup-
ported FPDs with tooth-implant supported FPDs.
Material and methods. Twenty-three patients with maxillary complete dentures and a Kennedy
Class I MOD 1 mandibular residual dentition were included in this longitudinal study. Each
patient was treated on one side of their mouth with an implant-supported FPD and the con-
tralateral side with a FPD supported by 1 implant and 1 tooth to allow for intraindividual com-
parison during the study period. Brfinemark implants (Nobel Biocare) were used in this study. At
the 10-year follow-up, nical and radiologic evaluations were performed that assessed: FPD stabil-
ity, implant stability, supporting tooth mobility, bleeding on probing, tightness of gold screws,
sensory function of the mental region, and marginal bone level. Statistical evaluation of the data
was made.
Results. After 10 years of function, there was no clinical difference regarding failure rates or
changes in marginal bone levels between FPDs supported by 2 implants and FPDs supported by
1 implant connected to 1 natural tooth. Shorter implants were not fbund to be less favorable than
long implants.
Gonclusion. The authors concluded that the connection between 1 implant and a natural tooth
to support a FPD in the posterior mandible is a cost-effective, predictable, and reliable treatment
and may be superior to an FPD supported by implants alone in select clinical situations. 24 Ref-
erences. - - R P Rennet

180 VOLUME 83 NUMBER 2