Dental Materials (2005) 21, 262–271
www.intl.elsevierhealth.com/journals/dema
Clinical performance of bonded leucite-reinforced
glass ceramic inlays and onlays after eight years
Norbert Krämer*, Roland Frankenberger
Department/Policlinic of Operative Dentistry and Periodontology, University of Erlangen-Nuremberg,
Glueckstrasse 11, Erlangen 91054, Germany
Received 21 August 2003; received in revised form 5 March 2004; accepted 25 March 2004
KEYWORDS
Glass ceramics; Resin
composites; Dentine
bonding; Clinical trial
Summary Objective: Ceramic inlays and onlays are a tooth colored alternative to
metallic restorations. Clinical long-term data are scarce though, especially about
inlays and onlays having proximal margins in dentin. The present prospective
controlled clinical study evaluated the clinical performance of IPS Empress inlays
and onlays with cuspal replacements and proximal margins below the cementoenamel
junction over eight years.
Methods: Ninety six ceramic restorations were placed in 34 patients by six dentists.
The restorations were bonded with an enamel/dentin bonding system (Syntac Classic)
and four different resin composite systems. The restorations were assessed after
placement by two calibrated investigators using modified USPHS codes and criteria at
the following time periods: baseline, 1,2,4,6 and 8 years.
Results: Eight of the 96 restorations investigated had to be replaced (failure rate 8%;
Kaplan – Meier): Six inlays suffered cohesive bulk fractures, two teeth required
endodontic treatment. After eight years of clinical service, significant deterioration
(Friedman 2-way ANOVA; P , 0:05) was found for marginal adaptation of the remaining
restorations. 98% of the surviving restorations exhibited marginal deficiencies,
independent of the luting composite.
Conclusions: IPS Empress inlays and onlays demonstrated to be successful even in
large defects. Neither the absence of enamel margins, nor cuspal replacement
significantly affected the quality of the restorations.
Q 2004 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved.
Introduction
During the last decade, many dental ceramics were
developed and clinically investigated [1 –5]. One of
these ceramics is the leucite-reinforced glass
ceramic IPS Empress having been marketed since
1990 [6 – 8].
Analyzing different clinical reports dealing
with ceramic inlays in the literature of the field,
*Corresponding author. Tel.: þ49-9131-1853-4202; fax: þ 499131-1853-3603.
E-mail address: kraemer@dent.uni-erlangen.de
bulk fractures are still the main problematic
scenario having been consequently observed with
all commercially available ceramic inlay systems
[9]. However, only a few controlled prospective
clinical studies present data about clinical
long-term performances of different systems
[10 – 13].
Although adhesively bonded to tooth hard tissues,
sintered ceramics suffered fractures in up to 20% of
clinically assessed cases [12]. Clinical failures with
sintered ceramic inlays and onlays were observed
throughout their clinical life. Extended Class-I
0109-5641/$ - see front matter Q 2004 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.dental.2004.03.009
Clinical performance of bonded leucite-reinforced glass ceramic inlays and onlays after eight years
restorations develop marginal fractures in the
majority of cases, whereas Class-II inlays fail
predominately due to bulk fractures [4,14]. However, certain clinical trials reported good clinical
perfomances [15,16]. Mirage II (Chameleon Dental
Products, Kansas City, USA) ceramic inlays being
characterized as glass – fiber-reinforced ceramic
system showed no failures after two years of clinical
service [17]. Dicor (Dentsply DeTrey, Konstanz,
Germany) glass ceramic inlays also demonstrated
to provide high success rates of 90% after six years
[18 – 22]. The leucite-reinforced ceramic system IPS
Empress (Ivoclar Vivadent, Schaan, Principality of
Liechtenstein) is similarly estimated [23].
When focussing on ceramic inlay restorations,
the majority of clinical trials are run on CAD/CAM
ceramic restorations [24 –27]. Reiss evaluated 1011
Cerec (Sirona Dental Systems, Bensheim, Germany)
inlays (ceramic: Vita Mark II, Vita Zahnfabrik, Bad
Säckingen, Germany) over a twelve year period
reporting a fracture rate of 8% [13]. The survival
analysis resulted in an 85% success rate after
12 years being representative of other Cerec
investigations [13,25,26].
Every clinical trial assessing ceramic inlays
reveals a certain deterioration of marginal quality
of these restorations [4,14,17,19,24]. This might be
caused by insufficient bonding to enamel or
degradation of the luting gap by degradation and
fatigue [17,19,28]. The vast majority of published
studies used a selective or total etch technique for
etching enamel,[4 – 7,9 – 11,13,14,16 – 22,24 – 27]
therefore an insufficient enamel bond should not
be the reason for the frequently observed marginal
deterioration. Focussing on the wear behavior of
luting composites, recent studies revealed no
significant influence of differently filled resin
composites for luting of ceramic inlays [28].
In this context, the aim of the present prospective clinical long-term trial was to evaluate the
performance of adhesively luted, IPS Empress inlays
and onlays with margins partially located below the
cementoenamel junction.
Materials and methods
Patients selected for this study met the following
criteria:
(1) Absence of pain from the tooth to be restored
(2) rubber dam application for placement of the
restoration
(3) proximal margins located below the cementoenamel junction (CEJ) in, if possible, 50% of
263
the teeth selected for restoration
(4) no further restorations planned in other posterior teeth
(5) high level of oral hygiene
(6) absence of any active periodontal and pulpal
desease.
Ninety six inlays (two surfaces: n ¼ 45; three
surfaces: n ¼ 27) and onlays ðn ¼ 24Þ were placed in
34 patients (11 male, 23 female; age 20 – 57
years, mean 33 years). Thirty percent of the
restorations were placed ðn ¼ 29Þ in maxillary
molars, 23% ðn ¼ 22Þ in maxillary premolars, 29%
ðn ¼ 28Þ in mandibular molars, and 18% ðn ¼ 17Þ in
mandibular premolars.
All patients were treated in the Department/
Policlinic of Operative Dentistry and Periodontology), University of Erlangen-Nuremberg (Germany),
by six different clinicians (assistant professors)
experienced in placing ceramic inlays and onlays.
All patients were required to give written informed
consent. The study was conducted according to EN
540 (Clinical investigation of medical devices for
human subjects, European Committee for Standardization). The patients agreed to a recall
programme of four years consisting of one appointment per year. The six and eight years recalls were
voluntary for the patients.
The preparations for the restorations were
performed slightly divergent without bevelling of
the margins using 80 mm diamond burs (Inlay PrepSet, Intensiv, Viganello-Lugano, Switzerland), and
finished with 25 mm finishing diamonds. The minimum depth of the cavity was 1.5 mm with rounded
occluso-axial angles. Dentin close to the pulp was
covered with a calcium hydroxide cement (Calxyl,
OCO-Praeparate GmBH, Dirmstein, Germany). A
glass ionomer cement (Ketac Bond, 3M Espe,
Seefeld, Germany) was used as the lining material.
Full-arch impressions were taken using a polyvinyl-siloxane material (Permagum High Viscosity,
3M Espe, Seefeld, Germany) and washed with a lowviscosity, syringeable material (Permagum Garant,
3M Espe, Seefeld, Germany) to record preparation
details.
One dental ceramist produced all the inlays and
onlays according to manufacturer’s instructions
within two weeks after impression taking.
The intraoral fit was evaluated under rubber
dam. Internal adjustments were performed using
finishing diamonds. Interproximal contacts were
assessed using waxed dental floss and special
contact gauges (YS Contact Gauge, YDM-Yamaura,
Tokyo, Japan). Prior to insertion, the thickness of
the inlays and onlays was recorded using a pair of
tactile compasses (Schnelltaster, Kroeplin,
264
N. Krämer, R. Frankenberger
Table 1 Composition of the investigated materials.
Adhesive
Treatment steps
Composition
Clinical procedure
Syntac
Etchant
37% phosphoric acid (enamel only)
Primer (Primer 1)
Maleic acid, triethylene glycol
dimethacrylate (TEGDMA), water/acetone
Polyethylene glycol dimethacrylate,
glutaraldehyde, water
Bisphenol glycidyl methacrylate (BisGMA),
triethylene glycol dimethacrylate, photoinitiator
Etch enamel margins separately
for 30 s.
Apply Primer for 20 s, dry.
Adhesive (Primer 2)
Heliobond (Bonding Agent)
Schluechtern, Germany) with an accuracy of
0.01 mm: The minimum thickness between deepest
fissure and fitting surface, minimal width in the
isthmus region for inlays, and the minimum thickness of the cuspal coverage in onlays were
measured.
The inlays were luted adhesively under rubber
dam and using the enamel etch technique. The
prepared teeth were thoroughly cleaned with
pumice slurry and etched with 37% phosphoric
acid gel. Afterwards, the dentin adhesive system
Syntac Classic (Ivoclar Vivadent) was applied
(Table 1). The internal surface of the restorations
was etched with 4.5% hydrofluoric acid (IPS Ceramic
etching gel, Ivoclar Vivadent, Schaan, Liechtenstein) for 60 s, rinsed, and then silanated with
Monobond S (Ivoclar Vivadent). After application of
the silane coupling agent, the solvent was evaporated with compressed air.
Adhesive insertion was performed with four
different luting composites: Dual Cement ðn ¼ 9Þ;
Variolink Low ðn ¼ 32Þ; Variolink Ultra ðn ¼ 6Þ; and
Tetric ðn ¼ 49Þ (all Ivoclar Vivadent, Schaan,
Liechtenstein, Table 2). The composite resins with
high viscosity (Variolink Ultra, Tetric) were used
according to the USI-technique (ultrasonic insertion
using EMS Piezon Master 400, Le Sentier,
Switzerland) utilizing the thixotropic properties of
the resin composite materials [29]. Polymerization
of the luting agents was performed by light-curing
for a total of 120 s from different positions
Apply Adhesive for 30 s, dry.
Apply bond, air thin.
Leave uncured and cure
together with the luting
composite.
(40 s from each direction). Prior to polymerization,
the luting composite was covered with glycerine gel
to prevent the formation of an oxygen inhibited
layer.
After light-curing and examining the luting areas
for defects, the rubber dam was removed. Centric
and eccentric occlusal contacts were adjusted using
diamond finishing burs (Intensiv, Viganello-Lugano,
Switzerland) prior to SofLex discs (3M, St. Paul, MN,
USA). Overhangs were removed and polished in the
same way, proximally with interdental diamond
strips (GC Dental Industrial Corp, Tokyo, Japan) and
interdental polishing strips (3M, St. Paul, MN, USA).
Final polishing was conducted using felt discs
(Dia-Finish E Filzscheiben, Renfert, Hilzingen,
Germany) with polishing gel (Brinell, Renfert,
Hilzingen, Germany).
Following placement of a restoration, the
restored tooth was covered with a fluoride solution
(Elmex Fluid, GABA, Lörrach, Germany) for 60 s to
treat potentially etched or ground adjacent enamel
areas [30].
At initial recall (six months recall was determined as baseline) and after one, two, four, six, and
eight years, all available restorations were assessed
according to modified United States Public Health
Service (USPHS) criteria (Tables 3 and 4) by two
independent investigators using mirrors, probes,
bitewing radiographs, and intraoral photographs
[7,31]. Recall assessments were not performed by
the clinician who had placed the restorations.
Table 2 Features of the restorations investigated.
Brand name
Type of composite
Mean particle size (mm)
Filler content (%wt)
Variolink Low
Variolink Ultra
Dual Cement
Fine particle hybrid
Fine particle hybrid
Inhomogeneously filled
microfill (pyrolytic silicone)
Fine particle hybrid
0.5–1
0.5–1
0.04
72
79
60
0.5–1
82
Tetric
Clinical performance of bonded leucite-reinforced glass ceramic inlays and onlays after eight years
Table 3 Modified USPHS criteria.
Surface roughness
Color match
Anatomic form
Anatomic form (margin)
Marginal integrity
Integrity tooth
Integrity inlay
Proximal contact
Changes in sensitivity
Complaints
Radiographic check
Subjective contentment
Within a previously calibrated pool of six investigators, one investigator was constantly assessing
the restorations throughout the whole study.
The statistical analysis was computed with SPSS
for Windows V11.0. The statistic unit was one
ceramic restoration, differences between the
groups were evaluated pair-wise with the Mann –
Whitney test (level of significance 0.05).
Results
The recall rate until the four years investigation was
100% and dropped to 60% due to the voluntary
character of the eight years recall (n ¼ 57 restorations). All patients were satisfied with their
restorations. 39 restorations could not be examined
after eight years due to failure ðn ¼ 8Þ or missed
recall investigation (n ¼ 31; drop out). Nine
patients were not available ðn ¼ 28Þ and one
patient lost her inlays due to prosthetic treatment
Table 4 Modified USPHS criteria.
Modified
criteria
Description
Analogous
USPHS criteria
‘Excellent’
Perfect
‘Alpha’
‘Good’
Slight deviations from ideal
performance, correction
possible without damage of
tooth or restoration
‘Sufficient’
Few defects, correction
‘Bravo’
impossible without damage
of tooth or restoration.
No negative effects expected
‘Insufficient’ Severe defects, prophylactic
removal for prevention of
severe failures
‘charlie’
‘Poor’
‘delta’
Immediate replacement
necessary
265
independent of the study. Over the whole observation period, the remaining 57 investigated restorations revealed no statistically significant
differences regarding proximal contact, sensitivity,
radiographic check (altogether only 12 X-rays after
eight years due to the voluntary character of this
recall) and subjective satisfaction (P . 0:05; Friedman 2-way ANOVA; Table 5).
Statistically significant differences over time
were observed for the following criteria (reasons/comments in parentheses): Surface roughness
(loss of gloss), color match (improving with time),
anatomic form (inlays were less worn under stress
than adjacent enamel), anatomic shape at margins
(step formations became rounded over time),
marginal integrity (especially at the eight years
recall: distinct deterioration with marginal fractures), tooth integrity (more cracks and abfractions
in enamel, but all within Bravo scores), inlay
integrity (continuous deterioration over time, predominantly chipping of the ceramic), and hypersensitivity (no complaints after eight years)
(P , 0:05; Friedman 2-way ANOVA).
Concerning the criteria marginal integrity, inlay
integrity, changes in sensitivity, and complaints, no
significant differences were found between the
luting composites Tetric and Variolink low
(P . 0:05; Mann – Whitney U-Test; Tables 6 and 7),
however, the analysis of inlay fractures between
Tetric and Variolink Low revealed P ¼ 0:051 in favor
of Variolink Low.
Statistically significant differences were
detected for the criteria tooth integrity (more
frequent enamel cracks with Tetric after the six
years recall) and radiographic check (more cases
of excess luting composite with Variolink Low
after four years; P , 0:05; Mann – Whitney U-Test).
In relation to tooth integrity, significant differences were detected especially in the Tetric
group between the one year, two years, four
years, six years, and eight years recall data
(P , 0:05; Friedman 2-way ANOVA). After eight
years, 82% of the restored teeth included small
enamel cracks representing an increase of 66%
from baseline. 16% of the restored teeth showed
enamel crack formation.
No statistically significant influence was attributed to the size of the inlay (two or more surfaces;
P . 0:05; Mann – Whitney test). The absence of
enamel in proximal boxes (15% with no enamel
and 54% of the restorations with less than 0.5 mm
residual enamel width in the proximal box) did not
have any influence on marginal performance or
secondary caries of the inlays and onlays.
The incidence of inlay defects judged with
‘Bravo’ or worse over time increased from 1% at
266
Table 5 Clinical results for IPS Empress according to modified USPHS-criteria on baseline, one, two, four, six and eight years.
Recall (SD)
Baseline ðn ¼ 89Þ (%)
1 year ðn ¼ 96Þ (%)
2 years ðn ¼ 95Þ (%)
4 years ðn ¼ 89Þ (%)
6 years ðn ¼ 67Þ (%)
8 years ðn ¼ 57Þ (%)
Alpha1
Alpha1
Alpha1
Alpha1
Alpha1
Alpha1
Alpha2
Bravo
Alpha2
0.5 years (0.14)
1.1 years (0.16)
92
84
66
43
40
96
84
85
94
87
–
41
59
77
15
34
93
90
87
94
90
–
Bravo
Alpha2
Bravo
2.1 years (0.17)
Alpha2
Bravo
4.0 years (0.14)
Alpha2
Bravo
Alpha2
6.0 years (0.33)
8.4 years (0.19)
84
92
91
18
4
61
35
94
100
100
100
81
96
90
49
2
67
18
95
97
100
92
Bravo
Criterion
Surface
Color matching
Anatomic form
Anatomic form (margin)
Marginal integrity
Integrity inlay
Integrity tooth
Proximal contact
Changes in sensitivity
Complaints
Radiographic assessment
a
b
c
8
16
34
57
60
3
16
15
6
13
–
1
–
59
39
23
85
60
5
9
13
6
6
–
2
5
2
1
4
–
41
66
74
11
17
92
78
84
100
98
88
59
33
25
88
65
3
21
10
1
1
1
18
4b
1
6
2
11
1
45
75
53
5
3
82
53
88
100
100
92
55
23
47
94
69
12
42
11
1
1
27a
5c
6
1
5
3
16
8
9
80
76
21
65
6
2
20
18
19
4
10
47
39
7
70
5
4
59
26
12
3
8
1 Inlay (2%) received ‘charlie‘ at the 4 years recall (gap formation).
1 Inlay (2%) received ‘delta‘ at the 2 years recall (chipping).
1 Inlay (2%) received ‘charlie‘ at the 4 years recall (chipping).
Table 6 Clinical results for IPS Empress luted with Variolink Low according to modified USPHS-criteria at baseline, one, two, four, six and eight years.
Criterion
a
1 year ðn ¼ 32Þ (%)
2 years ðn ¼ 32Þ (%)
4 years ðn ¼ 31Þ (%)
6 years ðn ¼ 22Þ (%)
8 years ðn ¼ 20Þ (%)
Alpha1
Alpha2
Bravo
Alpha1
Alpha2
Bravo
Alpha1
Alpha2
Bravo
Alpha1
Alpha2
Bravo
Alpha1
Alpha2
Bravo
Alpha1
24
93
97
100
90
–
76
7
3
63
3
9
6
6
–
6
16
97
91
100
100
84
63
3
9
22
26a
9
9
3
10
7
5
77
64
100
100
100
86
14
36
13
3
87
52
100
100
83
68
13
45
–
31
97
91
94
94
–
10
–
1 Inlay (2%) was rated ‘charlie‘ at the 4 years recall (gap formation).
–
3
65
35
90
100
86
Alpha2
Bravo
35
65
35
5
10
60
14
N. Krämer, R. Frankenberger
Marginal integrity
Integrity inlay
Integrity tooth
Changes in sensitivity
complaints
Radiographic assessment
Baseline ðn ¼ 29Þ (%)
54
25
18
43
11
71
4
64
11
100
100
100
28
25
67
25
83
2
4
6
15
8a
2
2
77
52
100
100
98
71
16
43
62
4
29
23
86
69
100
96
94
267
baseline to 7% after four years, and 26% after eight
years. Mainly chipping in the occlusal-proximal
contact areas was observed. Accurate comparisons
of clinical photographs revealed that these fractures mainly occurred in areas having been subjected to rotary occlusal adjustment (Figs. 1 and 2).
Until the eight year recall, seven IPS Empress
restorations had to be replaced. The survival rate
computed with the Kaplan– Meier algorithm was
92% after eight years (Fig. 3). Two inlays had to be
removed due to complaints (two because of
hypersensitivity, one was taken out of the protocol
after apex resection prior to the start of the study).
Bulk fractures of the ceramic material led to
replacement of six inlays. First catastrophic fractures were observed after 3 years, late failures
after 4.5 years. Although these time periods were
not part of the study, the patients had normal recall
appointments every six months where failures were
recorded and required replacements were carried
out. No catastrophic failure occurred after that
date.
The average ceramic dimensions measured prior
to insertion have been 1.4 mm below the deepest
fissure, 3.5 mm buccal-lingually at the isthmus,
and 2.0 mm below reconstructed cusps of onlays.
There was no statistically significant correlation
between dimensions of the inlay and fractures
observed ðP . 0:05Þ:
6
50
17
100
100
100
27
5b
5
b
1 Inlay (2%) received ‘delta‘ at the 2 years recall (chipping).
1 Inlay (2%) received ‘charlie‘ at the 4 years recall (chipping).
4
–
–
Marginal integrity
Integrity inlay
Integrity tooth
Changes in sensitivity
Complaints
Radiographic assessment
47
94
76
92
84
–
53
4
24
8
16
–
3
41
90
86
94
84
–
57
6
12
6
12
–
2
4
2
Discussion
a
Bravo
Alpha2
Alpha1
Bravo
Alpha2
Alpha1
Bravo
Alpha2
Bravo
Alpha2
Alpha1
Alpha1
Bravo
Bravo
Alpha2
Alpha1
Alpha2
Alpha1
6 years ðn ¼ 36Þ (%)
4 years ðn ¼ 44Þ (%)
2 years ðn ¼ 48Þ (%)
1 year ðn ¼ 49Þ (%)
Baseline ðn ¼ 49Þ (%)
Criterion
Table 7 Clinical results for IPS Empress luted with Tetric according to modified USPHS-criteria at baseline, one, two, four, six and eight years.
8 years ðn ¼ 31Þ (%)
Clinical performance of bonded leucite-reinforced glass ceramic inlays and onlays after eight years
The present study investigated the eight-year
performance of adhesively luted IPS Empress
ceramic inlays and onlays. It was the intention of
this study to include extended cavities, even with
proximal boxes extending below the CEJ.
The used modified USPHS criteria [31] proved to
be reliable for the tooth-colored restorations as
previously reported by Frankenberger et al. [32].
Especially in early years of ceramic inlay
studies, the split of the Alpha score into Alpha 1/2
proved to be an important tool over the whole
observation period.
For tooth-colored inlays, postoperative hypersensitivites have been initially reported to be
problematic due to possibly incompletely sealed
dentin or detachment between lining material and
dentin [33]. In the course of the present clinical
trial a certain amount of postoperative hypersensitivity (13% at baseline) was recorded, however, the
complaints were reduced rapidly in the majority of
cases. This is attributed to the use of a dentin
bonding agent in addition to a lining with glass
268
N. Krämer, R. Frankenberger
Figure 1 IPS Empress restorations in left upper premolars after eight years. Step formation and marginal ditching were
obvious (indicator, example for rating ‘Alpha2’). The inlays are still in situ.
ionomer cement. In the present study, the adhesive
system Syntac Classic was used with different luting
composites of the same manufacturer. The primarily intended focus was to compare a actual luting
composite (dual-curing, low viscosity, conventional
insertion technique: Variolink Low) with a restorative resin composite (light-curing, high viscosity,
ultrasonic insertion technique: Tetric). The
additionally used luting composites were developed
at the beginning of the study and were then added
to the protocol for fewer cases. In comparison with
previous studies, the additional use of the dentin
bonding led to limited hypersensitivity over the
observation period, independent of the composite
Figure 2 IPS Empress restorations in the left lower jaw after eight years. Small marginal fractures and chipping
fractures were detectable (indicators, judgement ‘Bravo’). In occlusal contact areas the gloss has diminished, a distinct
increase of surface roughness is already clinically visible.
Clinical performance of bonded leucite-reinforced glass ceramic inlays and onlays after eight years
Figure 3 Survival analysis using the Kaplan –Meier algorithm for IPS Empress with reasons for failure in every
case. Regarding the survival rate, no influence of inlay
size, localization, or luting composite was computed.
Therefore the data of different systems were pooled.
resin used for luting [34]. After one year, one
patient (having received four inlays) reported
occasional complaints (rated clinically ‘sufficient’).
During the whole observation period, two inlays had
to be replaced due to severe pulpal pain.
To date, numerous clinical studies have assessed
the luting space of tooth-colored inlays. The
majority of the restorations documented were
luted with resin composites of low viscositiy [24,
35,36]. The clinical consequences of luting composite wear are still not fully understood [37,38]. In
this context, the use of higher filled resin composites for luting was a hopeful approach [7,39,40].
The clinical data presented in this study cannot
confirm these expectations of enhanced abrasion
resistance with higher filled luting composites. For
the criterion ‘marginal integrity’ no clinically
noticeable difference between the luting agents
was found. The predominant ratings ‘good’ and
‘sufficient’ after six years were due to contact free
wear inside the luting areas characterized by
ditching having been detected throughout in all
groups. Although clinical evaluations alone might
not be sufficient tools for detecting differences of
the small luting space, previous studies using
mechnical profilometry exhibited similar results
[40]. Also from the clinical point of view, marginal
quality was subjected to considerable changes
beyond the six-years recall, expressed by an
increase of Bravo scores from 20% after six years
to 59% after eight years (Fig. 2). If this deterioration
progresses over the next years of clinical service,
it might be possible that further failures will occur
around or after the ten-years recall probably due to
269
decreased marginal quality. Especially, when luting
was performed with materials of low adhesion and
wear resistance e.g. glass ionomer cements, marginal fractures were frequently observed in previous studies [15,41]. Maybe this scenario is ten
years delayed and not completely avoided when
resin composites are used for luting.
Another critical point is still the questional
durability of adhesion to dentin. Warnings against
bonding to cervical dentin are still written consensus [42]. Nevertheless, in the present study 15%
of the inlays had proximal margins located
below the cementoenamel junction and 54% of
the restorations had less than 0.5 mm enamel
thickness in the proximal box with none of
the restorations radiographically revealing lower
clinical ratings or significant findings such as
secondary caries. This may be attributed to the
fact that an older three-step self-etching dentin
bonding agent (Syntac classic) was used. Recent
studies revealed that multi-step adhesives
providing a hydrophobic bonding agent as adhesion
promotor such as Heliobond as part of the
Syntac system achieve significantly better durability of resin-dentin bonding as reported by
DeMunck et al. for OptiBond FL over four years
and by Frankenberger et al. for Syntac over six years
water storage [43,44].
A further argument in favor of dentin bonding is
the estimation that the accurate use of this
additional bonding may prevent fractures of the
ceramic in bulk or along margins. This tends to
confirm the assumption that without any lining a
larger area of adhesion may enhance the stability of
the ceramic [4]. On the other hand, all cavities of
the present study received a lining with glass
ionomer cement. Therefore, from the present
results no additionally stabilizing effect of total
bonding of ceramic inlays can be assumed.
Significant differences were detected on the
enamel aspects of restored teeth, but these tooth
fractures did not have any significance on the
clinical survival behavior of the restorations.
About 16% of the enamel cracks (integrity tooth
‘good’) were already recognized at baseline. At
eight years, this observation increased to 82%.
However, no restoration had to be replaced due to
these cracks. Interestingly, there was a significant
difference with respect to enamel cracks when
different luting composites were used. When Tetric
as solely light-cured resin composite was used,
significantly more enamel cracks were observed
compared to the dual-cured Variolink Low after
six and more years. On the first approximation,
this may be an indicator that dual-curing luting
composites may be advantageous for adhesive
270
luting due to better stabilization caused by a more
reliable polymerization, independent of the distance of the light source during light polymerization. However, the facts clearly show that with the
dual-curing Variolink Low Syntac was used, providing a solely light-cured bonding agent (Heliobond).
So this assumption about better stabilization with
Variolink Low has to be rejected. Moreover, Tetric
seems to reveal a certain advantage regarding
handling, because on the X-rays less luting composite overhangs were found with Tetric. This seems
to be logical because the dentist definitively has
more time for excess removal prior to
photopolymerization.
The predominant failure scenario with ceramic
inlays is still the integrity of the inlay itself [7,
30]. ‘Half-moon’ fractures in the restorations
were detected as early as two years. These
fractures were observed exclusively in occlusally
loaded marginal ridges with pronounced overhangs in the direction of the approximating
tooth. Analyzing the clinical photographs resulted
in the finding that in each case of catastrophic
failure, occlusal adjustments were performed and
this trend continued throughout the study. Over
the eight years period, six inlays have failed due
to this reason. However, bruxism was considered
to be associated with the fractures in two cases.
Since the four-year report, one restoration had to
be replaced due to fracture [7]. It is interesting
whether the size and location of the individual
inlay influenced the clinical outcome. Neither the
number of restoration surfaces nor the size nor
the tooth type showed any significant influence
on clinical performance over the eight years
period.
No correlation was found between ceramic
thickness and fractures. The lowest cusp thickness
was recorded (0.3 mm) without having any clincial
consequences. There is considerable indication that
most of the fractures were attributed to fatigue
mechanisms [45,46,47]. Due to the difficult
intraoral situation, occlusal corrections may not
have been polished sufficiently and these microcracks may have been prone for later catastrophic
fractures. Therefore, the clinician should pay
attention to a careful polish of inlay areas having
been previously subjected to rotary occlusal corrections to prevent this particular problem. It is
interesting that all catastrophic fatigue fractures
occurred until the age of 4.5 years (Fig. 3).
Chippings that have been recorded at the eight
years recall (26% Bravo scores), seem to be
independent from rotary instrumentation during
occlusal adjustment directly after luting.
N. Krämer, R. Frankenberger
Conclusions
IPS Empress restorations revealed a 8% failure
rate with 98% of the remaining restorations
having marginal deficiencies after eight years.
The evaluated restorative system achieved satisfactory results for the restoration of larger
defects also in molar regions. Neither cusp
reconstruction nor preparation margins below
the cementoenamel junction were limiting factors for the good clinical success. Secondary
caries did not occur at all.
Acknowledgements
The authors thank Ivoclar Vivadent for supporting
this study.
References
[1] Banks RG. Conservative posterior ceramic restorations: a
literature review. J Prosthet Dent 1990;63:619—26.
[2] Jäger K, Wirz J, Schmidli F. Ceramic inlays as alternative to
amalgam? Schweiz Monatsschr Zahnmed 1990;100:
1345—52.
[3] Kelly JR, Nishimura I, Campbell SD. Ceramics in dentistry:
Historical roots and current persepectives. J Prosthet Dent
1996;75:18—32.
[4] Qualthrough AJE, Wilson NHF. A 3-year clinical evaluation
of a porcelain inlay system. J Dent 1996;24:317—23.
[5] Thonemann B, Federlin M, Schmalz G, Schams A. Clinical
evaluation of heat-pressed glass-ceramic inlays in vivo:
2-year results. Clin Oral Investig 1997;1:27—34.
[6] Fradeani M, Aquilano A, Bassein L. Longitudinal study of
pressed glass-ceramic inlays for four and a half years.
J Prosthet Dent 1997;78:346—53.
[7] Krämer N, Frankenberger R, Pelka M, Petschelt A. IPS
Empress inlays and onlays after four years—a clinical study.
J Dent 1999;27:325—31.
[8] Wohlwend A, Schärer P. Die Empress-Technik. Quintessenz
Zahntech 1990;16:966—78.
[9] Molin MK, Karlsson SL. A randomized 5-year clinical
evaluation of 3 ceramic inlay systems. Int J Prosthodont
2000;13:194—200.
[10] Fuzzi M, Rappelli G. Survival rate of ceramic inlays. J Dent
1998;26:623—6.
[11] Fuzzi M, Rappelli G. Ceramic inlays: clinical assessment and
survival rate. J Adhes Dent 1999;1:71—9.
[12] Hayashi M, Wilson NHF, Yeung CA, Worthington HV.
Systematic review of ceramic inlays. Clin Oral Investig
2003;7:8—19.
[13] Reiss B, Walther W. Clinical long-term results and 10 year
Kaplan-Meier analysis of Cerec restorations. Int J Comput
Dent 2000;3:9—23.
[14] Molin K, Karlsson. A 3-year clinical follow-up study of a
ceramic (Optec) inlay system. Acta Odontol Scand 1996;54:
145—9.
[15] Höglund ÄC, van Dijken JWV, Oloffson AL. Three-year
comparison of fired ceramic inlays cemented with compo-
Clinical performance of bonded leucite-reinforced glass ceramic inlays and onlays after eight years
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
site resin or glass ionomer cement. Acta Odont Scand 1994;
52:140—7.
Thordrup M, Isidor F, Hörstedt-Bindslev P. A one-year
clinical study of indirect and direct composite and ceramic
inlays. Scan J Dent Res 1994;102:186—92.
Friedl KH, Schmalz G, Hiller K-A, Saller A. In-vivo evaluation
of a feldspathic ceramic system: 2-year results. J Dent
1996;24:25—31.
Bessing C, Molin M. An in vivo study of glass ceramic (Dicor)
inlays. Acta Odontol Scand 1990;48:351—8.
Gladys S, Van Meerbeek B, Inokoshi S, Willems G, Braem M,
Lambrechts P, Vanherle G. Clinical and semiquantitative
marginal analysis of four tooth-colored inlay systems at
3 years. J Dent 1995;23:329—38.
Hayashi M, Tsuchitani Y, Miura M, Takeshige F, Ebisu S.
6-year clinical evaluation of fired ceramic inlays. Oper Dent
1998;23:318—26.
Roulet J-F. Longevity of glass ceramic inlays and
amalgam-results up to 6 years. Clin Oral Investig 1997;1:
40—6.
Stenberg R, Matsson L. Clinical evaluation of glass ceramic
inlays (Dicor). Acta Odontol Scand 1993;51:91—7.
El Mowafy O, Brochu J. Longevity and clinical performance
of IPS Empress ceramic restorations—a literature review.
J Can Dent Assoc 2002;68:233—7.
Heymann HO, Bayne SC, Sturdevant JR, Wilder AD,
Roberson TM. The clinical performance of CAD—CAMgenerated ceramic inlays. J Am Dent Assoc 1996;127:
1171—81.
Isenberg BP, Essig ME, Leinfelder KF. Three-year clinical
evaluation of CAD/CAM restorations. J Esthet Dent 1992;4:
173—6.
Mörmann WH, Krejci I. Computer-designed inlays after
5 years in situ: clinical performance and scanning
electron microscopic evaluation. Quintessence Int 1992;
23:109—15.
Sjögren G, Molin M, van Dijken JWV. A 5-year clinical
evaluation of ceramic inlays (Cerec) cemented with a dualcured or chemically cured resin composite luting agent.
Acta Odontol Scand 1998;56:263—7.
Krämer N, Frankenberger R. Leucite-reinforced glass
ceramic inlays after six years. Part II: wear of luting
composites. Oper Dent 2000;25:466—72.
Noack MJ, Roulet J-F, Bergmann P. A new method to lute
tooth coloured inlays with highly filled composite resins.
J Dent Res 1991;70—457. (Abstract 1528).
O’Reilly MM, Featherstone JD. Demineralization and
remineralization around orthodontic appliances: an in
vivo study. Am J Orthod Dentofacial Orthop 1987;92:
33—40.
271
[31] Cvar JF, Ryge G. Criteria for the clinical evaluation of
dental restoration materials. In: USPHS Publication No.
790, US Government Printing Office; 1971.
[32] Frankenberger R, Petschelt A, Krämer N. Leucite-reinforced
glass ceramic inlays and onlays after six years. Part I:
clinical behavior. Oper Dent 2000;25:459—65.
[33] Hickel R. The problem of hypersensitivities after insertion
of adhesive inlays (in German, Abstract in English). Dtsch
Zahnärztl Z 1990;45:740—2.
[34] Bodenheim G, Frankenberger R, Schoch M, Krämer N.
Clinical evaluation of a new ceramic luting system—six
months follow-up. J Dent Res 1999;78:308. (Abstract 369).
[35] Isidor F, Brondum K. A clinical evaluation of porcelain
inlays. J Prosthet Dent 1995;74:140—4.
[36] Jensen ME. A two-year clinical study of posterior etchedporcelain resin-bonded restorations. Am J Dent 1987;1:
27—32.
[37] Kawai K, Isenberg BP, Leinfelder KF. Effect of gap
dimension on composite resin cement wear. Quintessence
Int 1994;25:53—8.
[38] Krejci I, Lutz F, Reimer M. Wear of CAD/CAM ceramic inlays:
restorations, opposing cusps, and luting cements. Quintessence Int 1994;25:199—207.
[39] Krämer N, Pelka M, Petschelt A. Comparison of luting
composites using the high viscosity cementation technique.
J Dent Res 1995;74:537. (Abstract 1096).
[40] O’Neil SJ, Miracle RL, Leinfelder KF. Evaluating interfacial
gaps for esthetic inlays. J Am Dent Assoc 1993;124:48—54.
[41] Van Dijken JWV, Hörstedt P. Marginal breakdown of fired
ceramic inlays cemented with glass polyalkenoate (ionomer) cement or resin composite. J Dent 1994;22:265—72.
[42] Schmalz G, Federlin M, Geurtsen W. Are ceramic inlays and
veneers scientifically established? (in German, Abstract in
English). Dtsch Zahnärztl Z 1994;49:197—208.
[43] De Munck J, Van Meerbeek B, Yoshida Y, Inoue S, Vargas M,
Suzuki K, Lambrechts P, Vanherle G. Four-year water
degradation of total-etch adhesives bonded to dentin.
J Dent Res 2003;82:136—40.
[44] Frankenberger R, Strobel WO, Lohbauer U, Krämer N,
Petschelt A. The effect of six years water storage on resin
composite bonding to human dentin. J Biomed Mater Res
2004;69 B:25—32.
[45] Drummond JL, King TJ, Bapna MS, Koperski RD. Mechanical
property evaluation of pressable restorative ceramics. Dent
Mater 2000;16:226—33.
[46] Ohyama TM, Yoshinari M, Oda Y. Effects of cyclic loading on
the strength of all-ceramic materials. Int J Prosthodont
1999;12:28—37.
[47] Hickel R, Manhart J. Longevity of restorations in posterior
teeth and reasons for failure. J Adhes Dent 2001;3:45—64.