G Model

JPOR 514 No. of Pages 7

journal of prosthodontic research xxx (2019) xxx–xxx

Contents lists available at ScienceDirect

Journal of Prosthodontic Research

journal homepage: www.elsevier.com/locate/jpor

Review

CAD/CAM or conventional ceramic materials restorations longevity: a

systematic review and meta-analysis
Stéfani Becker Rodriguesa , Patrícia Frankena , Roger Keller Celesteb ,
Vicente Castelo Branco Leitunea , Fabrício Mezzomo Collaresa,*
a
Department of Conservative Dentistry, Dental Materials Laboratory, School of Dentistry, Universidade Federal do Rio Grande do Sul (UFRGS), Porto Alegre,
Brazil
b
Department of Preventive and Social Dentistry, School of Dentistry, Universidade Federal do Rio Grande do Sul (UFRGS), Porto Alegre, Brazil

A R T I C L E I N F O A B S T R A C T

Article history: Purpose: This systematic review and meta-analysis aimed to evaluate the difference in longevity of tooth-
Received 9 May 2018 supported ceramic prostheses designed by conventional and computer-aided design/computer-aided
Received in revised form 23 October 2018 manufacturing (CAD/CAM) techniques.
Accepted 9 November 2018
Study selection: Two reviewers searched the Web of Science, PubMed, SCOPUS and LILACS databases
Available online xxx
between 1966 and October 2017. Clinical studies that compared the survival rate of CAD/CAM against
conventional restorations were included.
Keywords:
Results: Eleven randomized controlled trials and three prospective studies were included, n = 14. Three
Meta-analysis
Survival rate
types of tooth-supported restorations were searched in the included studies: single crown, multiple-unit
Prosthodontics and partial ceramic crown. The follow-up of patients in the studies ranged from 24 to 84 months. A total
Ceramics of 1209 restorations had been placed in 957 patients in the included trials, and failures were analyzed by
CAD/CAM type and material restoration. From a total of 72 restoration failures, the CAD/CAM system resulted in a
1.84 (IC95%: 1.28–2.63) higher risk than conventional manufacturing of ceramic restoration.
Nevertheless, when drop-outs were included as a failure risk, the CAD/CAM system resulted in a risk
of 1.32 (IC95%: 1.10–1.58). Multilevel analysis of tooth-supported ceramic restorations, considering drop-
outs as successes, resulted in rates of 1.48 and 2.62 failures per 100 restoration-years for the controls and
CAD/CAM groups, respectively. Considering drop-outs as failures, we found rates of 4.23 and 5.88 failures
per 100 restoration-years for the controls and CAD/CAM groups, respectively.
Conclusions: The meta-analysis results suggest that the longevity of a tooth-supported ceramic
prostheses made by CAD/CAM manufacturing is lower than that of crowns mad by the conventional
technique.
© 2018 Japan Prosthodontic Society. Published by Elsevier Ltd. All rights reserved.

1. Introduction German survey in which approximately 35,000 bridges, crowns,

onlays, and inlays were placed over a period of 3.5 years [11].
The evolution of computer-aided design/computer-aided Despite the fact that CAD/CAM technology has significantly
manufacturing (CAD/CAM) technology and use ceramics in improved over time and has been shown to be approximately 16%
dentistry allow industrial production following secondary milling more time efficient in ceramic prosthesis fabrication than the
[1–4]. CAD/CAM prosthetic treatment with high performance conventional method, ceramic adaptation, the differences in the
became an alternative to traditional techniques because of the ceramic material and restoration position should be considered
quickness of chairside fabrication of the final restoration [2,5–10]. [12,13]. CAD/CAM resulted in the worst results in terms of internal
Use of CAD/CAM technology has been increasing, as verified in a fit compared to conventionally manufactured lithium disilicate
[14]. Furthermore, the difference between the veneer and
framework could result in an inadequate veneer thickness, firing
and cooling rate errors and surface damage from CAD/CAM
ceramic restoration production [15].
* Corresponding author at: Dental Materials Laboratory, School of Dentistry,
Differences in ceramic material manufacturing techniques
Universidade Federal do Rio Grande do Sul, Rua Ramiro Barcelos, 2492 Rio Branco,
90035-003 Porto Alegre, RS, Brazil. bring into question which method presents higher survival rates
E-mail address: fabricio.collares@ufrgs.br (F.M. Collares). [16–18]. The estimated five-year survival of all-ceramic crowns

https://doi.org/10.1016/j.jpor.2018.11.006
1883-1958/ © 2018 Japan Prosthodontic Society. Published by Elsevier Ltd. All rights reserved.

Please cite this article in press as: S.B. Rodrigues, et al., CAD/CAM or conventional ceramic materials restorations longevity: a systematic review
and meta-analysis, J Prosthodont Res (2018), https://doi.org/10.1016/j.jpor.2018.11.006
G Model
JPOR 514 No. of Pages 7

2 S.B. Rodrigues et al. / journal of prosthodontic research xxx (2019) xxx–xxx

ranges between 90.7% and 96.6% (feldspathic/silica based ceramics

and leucite or lithium-disilicate reinforced glass ceramics),
whereas the clinical performance of CAD/CAM single tooth
restorations has been reported to be an estimated total survival
of 91.6% after five years [19,20]. Nevertheless, there is scant and
unpooled evidence regarding the survival of tooth-supported
ceramic restoration comparing the manufacturing techniques of all
conventional manufacturing processes (laboratory-produced, con-
ventional waxed-up/pressed ceramic) and CAD/CAM technology.
The purpose of this systematic review and meta-analysis was to
evaluate the longevity of conventional and CAD/CAM techniques
for tooth-supported ceramic prosthesis (single crown, multiple-
unit or partial ceramic crown) and to identify the complication
types associated with the main clinical outcomes.

2. Methods

This is a systematic review with an unpublished protocol. The

Preferred Reporting Items for Systematic Reviews and Meta-
Analyses (PRISMA) guidelines were used, and the checklist was
Fig. 1. Flowchart depicting the studies identified, included, and excluded with
followed in this study [21]. reasons.

2.1. Population, intervention, control, outcome and time (PICOT)

strategy were excluded. The manuscripts considered to be eligible were
selected for a full-text assessment. From this search, a collection of
P: Patients with an anterior and/or posterior tooth-supported studies to be evaluated by reviewers was created. The reviewers
single crown (SC) or multiple-unit (MU) or partial crowns (PCC). were not masked for included papers. The two aforementioned
I: Patients with at least one ceramic restoration made with the reviewers extracted data independently; discrepancies in data
CAD/CAM process. extraction were solved by consensus.
C: Patients with at least one ceramic restoration made with the From each study, the following data were obtained: study
conventional manufacturing processes. design, restoration type, follow-up time, mean patient age,
O: Longevity of CAD/CAM restorations. material framework and veneer, CAD/CAM system, setting,
T: Follow-up: at least two years. publication year, failure number and type, initial and final sample
size and number of patient drop-outs. Data not shown in published
2.2. Focused question articles were requested from the corresponding author.

“Longevity of tooth-supported ceramic restorations are more 2.5. Data analysis

influenced by conventional or CAD/CAM techniques?”
The risk of failure for each group of included studies was
2.3. Search methods and selection criteria calculated based on the number of baseline restorations (number
of initial restorations) and the number of failures at the end of the
In vivo studies that evaluated the survival of milling or CAD/ follow-up period. The relative risks (RR) from all studies were
CAM prostheses after a minimum of two years of follow-up time pooled in a fixed effect meta-analysis using the default Mantel and
were included. Failure of survival was defined as any need of a Haenszel method for binary variables. Bias was assessed using a
restoration replacement, such as loss of retention, loss of vitality, funnel plot, and heterogeneity was assessed using I2. The risk
secondary caries, tooth extraction and veneer or framework difference was also calculated.
fracture. The search strategy was performed from 1966 to October For the sensitivity analysis, we included drop-outs either as
10th, 2017, with no limits. The sources of published studies were failures or as successes [22]. Therefore, all analyses were
the Web of Science, PubMed, SCOPUS and LILACS databases. The duplicated to check for possible selection bias and robustness.
following MeSH terms and their combinations were used in the All analyses were carried out with Stata Software 13.1 (StataCorp
database searches: #1 ceramics, #2 ceramic dental, #3 clinical LP, College Station, TX, USA).
trial, #4 randomized controlled clinical trials, #5 clinical practice,
#6 cad cam, #7 dental milling, #8 (#1) OR (#2), #9 (#3) OR (#4) OR 2.6. Risk of bias assessment
(#5), #10 (#6) OR (#7), #11 (#8) AND (#9) AND (#10). In addition,
the references of all of the identified articles were manually Two reviewers (PF and SR) independently assessed the risk of
searched for further relevant studies. bias. Randomized controlled trials (RCT) were assessed using
Cochrane Collaboration’s tool guided by the Cochrane Handbook
2.4. Eligibility criteria and data collection for Systematic Reviews of Interventions (version 5.1.0) [23]. Six
specific domains were assessed: sequence generation, allocation
Two reviewers, PF and SR, independently assessed the titles and sequence concealment, blinding, incomplete outcome data, selec-
abstracts of all of the selected articles. The articles were first tive outcome reporting, and other sources of bias. Judgments of
reviewed by title and abstract and subsequently by a full text low, unclear or high were used to assess the overall risk of bias in
reading. This systematic review did not have a language limitation. each study. The risk of bias for each entry recording was judged as
Studies that did not evaluate the annual failure rate, survival rate or ‘no’ to indicate a high risk of bias, ‘yes’ to indicate a low risk of bias
conventional manufacturing (control group) against milling and ‘unclear’ to indicate either a lack of information or uncertainty
ceramic restorations of a tooth- supported SC or MU or PCCs over the potential risk of bias. The Newcastle–Ottawa Scale was

Please cite this article in press as: S.B. Rodrigues, et al., CAD/CAM or conventional ceramic materials restorations longevity: a systematic review
and meta-analysis, J Prosthodont Res (2018), https://doi.org/10.1016/j.jpor.2018.11.006
G Model
JPOR 514 No. of Pages 7

S.B. Rodrigues et al. / journal of prosthodontic research xxx (2019) xxx–xxx 3

Table 1. Study and patients characteristics of the reviewed studies.

Study Study Setting Mean age CAD/CAM Frame work and Frame work and R Mean N N Drop- Failure Setting
design (years) system veneer (C) veneer (T) follow upa (C/T) outs (C/T) (C/T) position
Monaco et al. [37] RCT U NR NR GC Zr SC 65.7 45/ 5/0 3/4 Posterior
45
Akin et al. [33] RCT U 29 Cerec D D SC 24 15/ 0/0 0/0 Anterior
15
Rinke et al. [36] RCT PP 49.6 Cercon MC Zr SC 60 50/ 18/9 2/3 Posterior
55
Passia et al. [31] RCT U 42 Kavo GC Zr SC 60 100/ 19/37 9/28 Posterior
everest 123
Vigolo and P PP 32 Procera/lava MC Zr SC 60 40/ 10/20 2/7 Posterior
Mutinelli [29] 40
Nicolaisen et al. RCT U/PP 51 BEGO MC Zr MU 36 17/ 0/0 3/2 Posterior
[35] 17
Christensen and RCT PP 50 Cercon/ MC Zr MU 24/36 96/ 9/12 5/15 Posterior
Ploeger [27] everest/lava 163
Pelaez et al. [30] RCT U Lava MC Zr MU 50 20/ 0/0 0/1 Posterior
20
Sailer et al. [26] RCT U 54.4 Cercon MC Zr MU 40 38/ 18/5 0/0 Posterior
38
Walter et al. [25] P U/PP 40 Procera GC MC MU 61 25/ 52/68 0/1 Posterior
22
Zenthöfer et al. RCT U 56 Lava MC Zr C- 36 10/ 10/9 2/4 Anterior
[34] MU 11
Federlin et al. [28] P U 37 Cerec GC F PCC’s 66 29/ 24/24 1/4 Posterior
29
Guess et al. [32] RCT U 45.5 Cerec D L PCC’s 84 40/ 52/52 0/1 Posterior
40
Molin et al. [24] RCT U 37 Cerec Gold, mirage, F PCC’s 60 60/ 0/0 7/2 Posterior
empress 20

R: restoration; NR: not reported; RCT: randomized controlled trial; P: prospective; U: university; PP: private practice; SC: single crown; MU: multiple unit; C-MU: cantilever
multiple unit; PCC’s: partial ceramic crowns; GC: gold crown; MC: metal-ceramic; D: dissilicate; Zr: zirconia; F: feldspar; C: control group; T: test group; L: leucite-based
glass-ceramics.
a
In months.

followed to assess the prospective studies. The methodological [26,29]. The types of failure according to the types of tooth-
quality was based on selection, comparability and outcome. supported restorations are shown in Table 2.

3. Results 3.2. Methodological quality

3.1. Study selection and description of studies The risk of bias in the studies included in the systematic review is
summarized in the supplementary material (Tables 3 and 4). For
Electronic database searches from all sources (PubMed, randomized controlled trials (RCTs), ten studies reported using
SCOPUS, LILACS and Web of Science) identified 1897 publications randomly assigned interventions; however, one study did not follow
(Fig. 1). After evaluating the titles and abstracts, 159 publications procedures to ensure a random sequence generation [36]. For the
were selected for full text reading. In the end, fourteen studies allocation concealment strategy, many studies had an unclear or high
(published between 1966 to October 10th, 2017) were included in risk of bias [24,26,30,32–37]. Six studies did not report blinding
this systematic review and seventeen comparisons groups were because it was not possible to differentiate them regarding the color
included in the meta-analysis (one publication presented one test of material (gold and ceramic/metal ceramic and ceramic). The
and three control groups) [24–36]. The included articles were variable blinding was considered to have a low risk in all studies since
published in English between 1999 and 2017; the characteristics of the outcomes are unlikely to have been influenced by a lack of
each study were presented in a descriptive table (Table 1). For all of blinding [38]. All studies correctly reported incomplete outcome
the studies included, 11 were randomized controlled trials and 3 data and stated the number and reasons for attrition and exclusion in
were prospective trials. Seven of these studies obtained financial the control and intervention groups. The studies used pre-specified
support or material donations from industry [25,26,30–32,35,36]. (primary and secondary) outcomes; all expected outcomes were
The majority of studies were conducted in universities (9 studies), included. No other sources of bias were identified. For a summary
some were conducted in private practices (3 studies) or jointly assessment of the outcomes within the studies, only three had a low
between a university and private practice (2 studies). Three types risk of bias, while five had an unclear risk of bias and three had a high
of tooth-supported restorations were searched in the included risk of bias. For prospective studies, the quality score for each study
studies, a single crown (5 studies), multiple-unit (6 studies was assessed; the major bias in these studies was a risk of bias in the
including in one that presented cantilever) and partial ceramic selection of samples because the participants were mostly from
crown (3 studies). A total of 1209 restorations were placed in 957 university and private dental offices. Furthermore, blinding was not
patients in the included trials. The follow-up of patients in the possible in these studies.
studies ranged from 24 to 84 months [32,33]. Considering articles
that presented results for the same sample, the most recent study 3.3. Results of analyses
was considered [28,31,32,36].
Trials that used United State Public Health Service (USPHS), 3.3.1. Fixed effects meta-analysis and multilevel regression
modified USPHS and California Dental Association (CDA) as The included studies had different follow-up periods; three
outcome measures, the categories Charlie/Delta and one studies had differences among their groups [24,27,29]. In addition,
(i.e., requiring repair or replacement) were counted as failures the number of drop-outs varied among studies and groups within

Please cite this article in press as: S.B. Rodrigues, et al., CAD/CAM or conventional ceramic materials restorations longevity: a systematic review
and meta-analysis, J Prosthodont Res (2018), https://doi.org/10.1016/j.jpor.2018.11.006
G Model
JPOR 514 No. of Pages 7

4 S.B. Rodrigues et al. / journal of prosthodontic research xxx (2019) xxx–xxx

Table 2. Failure type and number (control – test) according to the type of restoration.

Study Surface and Secondary Loss of Loss of Framework/ venner Extraction USPHS (C–D Modified ryge Restoration
color caries vitality retention fracture score) criteria type
Monaco et al. [37] 0–0 0–0 0–0 0–0 2–4 1–0 0–0 N/A SC
Akin et al. [33] N/A N/A N/A N/A N/A N/A N/A 0–0 SC
Rinke et al. [36] 0–0 1–0 1–1 0–1 0–1 0–0 N/A N/A SC
Passia et al.a [31] 9–28 N/A N/A SC
Vigolo and Mutinellia N/A N/A N/A N/A N/A N/A 1–7 N/A SC
[29]
Nicolaisen et al.d [35] 2–2 0–0 0–0 0–0 1–0 0–0 N/A N/A MU
Christensen and 0–0 1–2 0–0 0–0 4–13 0–0 N/A N/A MU
Ploeger [27]
Pelaez et al. [30] 0–0 0–0 0–0 0–0 0–0 0–1 N/A N/A MU
Sailer et al. [26] N/A N/A N/A N/A N/A N/A 0–0 N/A MU
Walter et al. [25] 0–0 0–0 0–0 0–0 0–1 0–0 N/A N/A MU
Zenthöfer et al.b [34] 0–0 0–0 1–2 0–0 1–2 0–0 0–0 N/A Cantilever
MU
Federlin et al.c [28] N/A N/A N/A N/A N/A N/A 1–4 N/A PCC’s
Guess et al.c [32] N/A N/A N/A N/A N/A N/A 0–1 N/A PCC’s
Molin et al.d [24] V 0–0 3–0 0–1 4–1 0–0 N/A N/A PCC’s

SC: single crown; MU: multiple unit; PCC’s: partial ceramic crowns.
a b c d
Not specified failure type-chipping down the coping and fracture of coping. None C or D score. Modified USPHS. CDA criteria.

Table 3. Cochrane summary assessment of risk of bias for randomized controlled trials.

Study Adequate sequence Allocation Blinding Incomplete outcome Selective outcome Other sources of Risk of bias
generation concealment data reporting bias
Monaco et al. [37] Yes Yes Yes Yes Yes Yes Low
Akin et al. [33] Yes Unclear Yes Yes Yes Yes Unclear
Rinke et al.a [36] No No Yes Yes Yes Yes High
Passia et al. [31] Yes Yes Yes Yes Yes Yes Low
Nicolaisen et al. [35] Yes Unclear Yes Yes Yes Yes Unclear
Christensen and Ploeger Yes Yes Yes Yes Yes Yes Low
[27]
Pelaez et al.b [30] Yes No Yes Yes Yes Yes High
Sailer et al.b [26] Yes No Yes Yes Yes Yes High
Zenthöfer et al. [34] Yes Unclear Yes Yes Yes Yes Unclear
Guess et al.c [32] Yes Unclear Yes Yes Yes Yes Unclear
Molin et al.c [24] Yes Unclear Yes Yes Yes Yes Unclear
a b c
Patient’s preference. List of randomization. Split-month design.

Table 4. Newcastle–Ottawa for quality rating of prospective studies.

Study Selection Comparability Outcome Total

of
star
Representativeness Selection of Ascertainment Demonstration that Comparability of Assessment Was follow-up Adequacy
of the exposed the non of exposure outcome of interest was cohorts on the basis of outcome long enough for of follow-
cohort exposed not present at start of of the design or outcomes to up of
cohort study analysis occur cohorts
Vigolo – 1 star 1 star 1 star – 1 star 1 star 1 star 6
and
Muttineli [29]
Federlin – 1 star 1 star 1 star 1 star 1 star 1 star 1 star 7
et al.a
[28]
Walter – 1 star 1 star 1 star – 1 star 1 star 1 star 6
et al.
[25]
a
Split-mouth study.

studies. As both issues violate the assumption of a fixed cohort, we Some studies presented more than one test group, and for this
incorporated them by calculating the Incidence Rate Ratios (IRR) reason, more groups were included in the meta-analysis (seven-
based on failures per 100 restoration-years. To calculate the time of teen) than in the systematic review (fourteen) [27,29]. The risk of
follow-up (restoration-year), we multiplied the number of failure among control groups was, on average, 8.5% (IC95%: 5.5–
restorations by the number of years of follow-up, with drop-outs 11.5), and among CAD/CAM was 14.4% (IC95%: 8.6–20.1); this meta-
counting as being followed for half of the period. For this analysis, analysis led to RR of 1.84 (IC95%: 1.28–2.63), Fig. 2. However, when
we pooled all groups in random intercept multilevel Poisson drop-outs were included as a failure, the risk of failure among the
regression. control groups was, on average, 22.2% (IC95%: 14.6–29.9); among

Please cite this article in press as: S.B. Rodrigues, et al., CAD/CAM or conventional ceramic materials restorations longevity: a systematic review
and meta-analysis, J Prosthodont Res (2018), https://doi.org/10.1016/j.jpor.2018.11.006
G Model
JPOR 514 No. of Pages 7

S.B. Rodrigues et al. / journal of prosthodontic research xxx (2019) xxx–xxx 5

Fig. 2. Forest plot of pooled studies with a relative risk of failure of 1.84 (1.28–2.63) Fig. 3. Forest plot of pooled studies with a relative of risk of failure of 1.32 (1.10–
in tooth-supported ceramic restorations not considering drop-outs. The 95% 1.58) in tooth-supported ceramic restorations considering drop-outs. The 95%
confidence intervals for survival rates are given in parenthesis. confidence intervals for survival rates are given in parenthesis.

CAD/CAM groups, the risk of failure was 28.2% (IC95%: 18.4–42.7). indicated that the CAD/CAM system resulted in more chances to
This meta-analysis led to RR = 1.32 (IC95%: 1.10–1.58), Fig. 3. fail compared to conventional manufacturing of a ceramic
Multilevel analysis of tooth-supported ceramic restorations restoration, considering losses of follow-up as failures or as
considering drop-outs as successes resulted in rates of 1.48 and successes.
2.62 failures per 100 restoration-years for the controls and CAD/ The RR was calculated as the ratio of the proportion of failures
CAM groups, respectively. This led to IRR = 1.76 (IC95%: 1.20–2.59) in CAD/CAM to that in conventional ceramic restorations according
times higher in the CAD/CAM groups. Considering drop-outs as to the meta-analysis, providing a consolidated value for failure
failures, we found rates of 4.23 and 5.88 failures per 100 compared with ceramics. A funnel plot was drawn to identify any
restoration-years for the control and CAD/CAM groups, respec- publication bias within the studies; none were observed (Figs. S3
tively, leading to IRR = 1.39 (IC95%: 1.11–1.74) times higher in the and S4). Of the two meta-analyses performed, one analyzed the RR
CAD/CAM groups. of failures that were reported by the authors and the other
analyzed the RR of failures including drop-outs as failures,
3.3.2. Assessment of heterogeneity simulating the worst possible outcome. The inclusion of drop-
The heterogeneity among the studies was quantified with the outs as failures decreases the difference among groups and
I2-statistic considered to be “high” if the statistical heterogeneity increases the sample size [22]. If most drop-outs were true failures,
levels were higher than 70% [39]. In this meta-analysis, the then conventional manufacturing would still have a significantly
heterogeneity in RR among the studies was I2 = 0% (p = 0.80) when lower risk of failure than CAD/CAM.
only considering restoration failures and was I2 = 37.0% (p = 0.06) Comparing implant-supported versus tooth-supported fixed
when drop-outs were considered as failures (Figs. 1 and 2). The dental prostheses restorations, the most frequent complication of a
cumulative meta-analysis suggests that that since 2013, the conventional tooth-supported prosthesis was caries and loss of
association with CAD/CAM restorations has been significant pulp vitality [16]. However, the most frequent reason for failure in
(Figs. S1 and S2). this study was veneer chipping fractures (Table 2), in corroboration
with previous studies [20,41]. Most likely, the tooth position, type
3.3.3. Assessment of publication bias and influent studies of material and process were responsible for these results since all
Publication bias was initially evaluated through an examination of the studies included in the meta-analysis used conventional
of funnel plot asymmetry [40]. Egger’s and Begg’s tests indicated impressions during the clinical assessment [4]. Most of the studies
no evidence of publication bias without drop-outs counted as that reported the failure of CAD/CAM restorations exceeded the
failures (p = 0.32 and p = 0.97, respectively) or with drop-outs control group except for the studies by Akin and Sailer (which
counted as failures (p = 0.45 and p = 0.71, respectively). Visual reported no failures in the control and test) and Molin (which
inspection of Begg’s funnel-plot (Figs. S3 and S4) also demonstrat- reported more failures in the control) [24,26,33]. From the selected
ed that no significant asymmetry was present for either analysis. studies, seven used zirconia as the framework; no study used a
However, analysis of influence showed that the study of Passia was monolithic zirconia crown. A recent systematic review showed
the most influential; removing it led the association to be not that monolithic crowns led no incidence of chipping compared
significant (p > 0.05), reducing the RR from 1.84 to 1.42 (IC95%: with bi-layered crowns [42]. Tensile or compressive residual stress
0.92–2.20) when drop-outs were considered as failures [31]. between the framework/veneer interface (zirconia/feldsphatic)
can happen as a result of the difference in the thermal coefficient of
4. Discussion materials [1]. Moreover, the lack of framework support, inadequate
experience with ceramics, veneer thickness, firing and cooling rate
This systematic review investigated the longevity of ceramic errors and surface damage from CAD/CAM production are the
restorations according to techniques (conventional or CAD/CAM) usual causes of failures in zirconia frameworks [15]. On one hand,
using a tooth-supported single crown, multiple-unit and partial the internal fit (between the tooth and prosthesis), type and
single crown. The fourteen studies included in the meta-analysis properties of ceramics and prosthetic treatment may increase the

Please cite this article in press as: S.B. Rodrigues, et al., CAD/CAM or conventional ceramic materials restorations longevity: a systematic review
and meta-analysis, J Prosthodont Res (2018), https://doi.org/10.1016/j.jpor.2018.11.006
G Model
JPOR 514 No. of Pages 7

6 S.B. Rodrigues et al. / journal of prosthodontic research xxx (2019) xxx–xxx

number of failures among CAD/CAM systems [14]. Although One included study had a limitation of no patient randomization
clinical data on monolithic zirconia restorations are still sparse, [25]. The selection of restoration type was based on the patient’s
these restorations are an option for eliminating the increased risk preference after information was presented on the two types of
of veneer ceramic failure. In contrast with this study, another restorations; thus the recruitment of patients in private practice
technical problem observed in zirconia-based single crowns was a was accelerated. Despite the fact that this selection mode could be
loss of retention [20]. On the other hand, conventional manufac- a possible risk, the numbers of the two types of restorations were
tured systems may decrease chipping failures as a result of the similar in both groups (metal-ceramic, n = 41 and zirconia, n = 50),
stronger bonding between the framework and veneer layer and no gender or age-related effects on the selected type of
(feldspathic veneer and metal, gold and disilicate framework). restoration was detected in this study. Additionally, the outcome
The tooth position also influences veneer chipping fractures; was determined by the examiner and not by patients, eliminating
most restorations of the included studies were located in the the bias of the selective outcome report [38]. Thus, this study was
posterior region [20,41]. Exceptions included two studies that considered to have a high risk of bias for adequate sequence
placed the prosthesis in the anterior region [33,34]. Crowns placed generation and allocation sequence concealment. Moreover, the
in anterior teeth have a longer survival time than those placed in authors of the included studies rarely described the method used
the posterior region as a result of the different direction of occlusal to conceal the allocation sequence, which made most key domains
forces [13]. The outcomes of anterior and posterior single crowns ‘unclear’.
were compared, and no statistically significant differences of
survival rates were found for metal-ceramic crowns, leucit or 5. Conclusion
lithium-disilicate reinforced glass ceramic crowns, and alumina or
zirconia-based crowns. Crowns made out of feldspathic or silica- The results of this meta-analysis suggest that the longevity of
based ceramics, however, exhibited significantly lower survival tooth-supported single crown, multiple unit or partial ceramic
rates in the posterior region than in the anterior region (87.8% vs. crowns made by CAD/CAM is lower than that of crowns made by
94.6%) [17]. According to the framework design, one study conventional techniques. The material type and process were the
considered the anatomical design of the framework and an most frequent reasons for CAD/CAM failures. However, studies that
adequate thickness of the veneer and found that these factors evaluate the difference between CAD/CAM generations and
cannot be considered critical factors for chipping in the zirconia software limitations should be performed to elucidate the reasons
group [30]. that CAD/CAM results in a higher risk of failure.
Two studies used a press lithium-disilicate glass ceramic
crown as the control group [32,33], and one used a CAD/CAM Acknowledgemens
lithium-disilicate glass ceramic crown as the experimental group
[33]. Both CAD/CAM and press crowns demonstrated a survival The authors gratefully acknowledge CAPES (Coordenação de
rate of 100% over the observation period. The lower failure rate of Aperfeiçoamento de Pessoal de Nível Superior) for the scholarship
the present study can be attributed to the improved flexural (S.B.R.). RKC and FMC hold PQ2 fellowship from CNPq.
strength of the pressed system (400 MPa) and the homogeneity of
the CAD/CAM material. The success rates of both materials, CAD/ Appendix A. Supplementary data
CAM and press lithium-disilicate glass crowns, were comparable
over time [16,32]. Supplementary data associated with this article can be found, in
Failures resulting from extensive fractures of the veneering the online version, at https://doi.org/10.1016/j.jpor.2018.11.006.
ceramic and loss of retention were frequently found technical
problems for all-ceramic and multiple unit zirconia crowns
References
compared with metal-ceramic crowns [20,41]. Consequently,
based on this study and previous reviews, zirconia-based restora- [1] Tholey MJ, Swain MV, Thiel N. Thermal gradients and residual stresses in
tions should not be considered as the primary treatment option. veneered Y-TZP frameworks. Dent Mater 2011;27:1102–10.
The biological and manufacturing differences between materials [2] Kollmuss M, Kist S, Goeke JE, Hickel R, Huth KC. Comparison of chairside and
laboratory CAD/CAM to conventional produced all-ceramic crowns regarding
may be considered to be one of the key decisive factors for the morphology, occlusion, and aesthetics. Clin Oral Invest 2016;20:791–7.
choice of ceramics as reconstructive material for rehabilitation. [3] Chochlidakis KM, Papaspyridakos P, Geminiani A, Chen CJ, Feng IJ, Ercoli C.
The included studies used different criteria to detect failure that Digital versus conventional impressions for fixed prosthodontics: a systematic
review and meta-analysis. J Prosthet Dent 2016;116:184–90.
may lead to certain measurement biases. Homogeneous data for
[4] Ahlholm P, Sipilä K, Vallittu P, Jakonen M, Kotiranta U. Digital versus
comparison in any systematic review is difficult and has been conventional impressions in fixed prosthodontics: a review. J Prosthodont
reported in a review on standardizing criteria for failure, success 2018;27:35–41.
and survival of ceramic prostheses [18]. The lack of detailed [5] Bindl A, Mörmann WH. Survival rate of mono-ceramic and ceramic-core CAD/
CAM-generated anterior crowns over 2-5 years. Eur J Oral Sci 2004;112:197–
information on technical complications, such as chipping fractures, 204.
refinishing, repairs, and bulk fractures, represents a major [6] Bindl L, Dresbach K, Lentze MJ. Incidence of acute respiratory distress
deficiency of previous reports on the performance and survival syndrome in German children and adolescents: a population-based study. Crit
Care Med 2005;33:209–312.
of ceramic restorations. Therefore, to compare and obtain the RR of [7] Strub JR, Rekow ED, Witkowski S. Computer-aided design and fabrication of
the studies, we took the reported failures by the authors into dental restorations: current systems and future possibilities. J Am Dent Assoc
account, which allowed a value to be placed on outcome and 2006;137:1289–96.
[8] Mainjot AK, Dupont NM, Oudkerk JC, Dewael TY, Sadoun MJ. From artisanal to
comparison. CAD-CAM blocks: state of the art of indirect composites. J Dent Res
The methodological quality of the included studies was verified, 2016;95:487–95.
and only those with a low risk of bias in all domains were [9] Alghazzawi TF. Advancements in CAD/CAM technology: options for practical
implementation. J Prosthodont Res 2016;60:72–84.
considered in the meta-analysis [23]. However, this decision [10] Sailer I, Benic GI, Fehmer V, Hämmerle CHF, Mühlemann S. Randomized
usually results in a limited number of studies and, consequently, of controlled within-subject evaluation of digital and conventional workflows for
ceramic restorations that can be included in the study, posing the the fabrication of lithium disilicate single crowns. Part II: CAD-CAM versus
conventional laboratory procedures. J Prosthet Dent 2017;118:43–8.
possibility of under-powering the study’s outcomes. The studies
[11] Belli R, Petschelt A, Hofner B, Hajtó J, Scherrer SS, Lohbauer U. Fracture rates
included in this systematic review had some design limitations, and lifetime estimations of CAD/CAM all-ceramic restorations. J Dent Res
and several studies involved small sample sizes (range, 10–163). 2016;95:67–73.

Please cite this article in press as: S.B. Rodrigues, et al., CAD/CAM or conventional ceramic materials restorations longevity: a systematic review
and meta-analysis, J Prosthodont Res (2018), https://doi.org/10.1016/j.jpor.2018.11.006
G Model
JPOR 514 No. of Pages 7

S.B. Rodrigues et al. / journal of prosthodontic research xxx (2019) xxx–xxx 7

[12] Joda T, Brägger U. Digital vs. conventional implant prosthetic workflows: a [28] Federlin M, Hiller KA, Schmalz G. Controlled, prospective clinical split-mouth
cost/time analysis. Clin Oral Implants Res 2015;26:1430–5. study of cast gold vs. ceramic partial crowns: 5.5 year results. Am J Dent
[13] Kassardjian V, Varma S, Andiappan M, Creugers NH, Bartlett D. A systematic 2010;23:161–7.
review and meta analysis of the longevity of anterior and posterior all-ceramic [29] Vigolo P, Mutinelli S. Evaluation of zirconium-oxide-based ceramic single-unit
crowns. J Dent 2016;55:1–6. posterior fixed dental prostheses (FDPs) generated with two CAD/CAM
[14] Papadiochou S, Pissiotis AL. Marginal adaptation and CAD-CAM technology: a systems compared to porcelain-fused-to-metal single-unit posterior FDPs: a
systematic review of restorative material and fabrication techniques. J 5-year clinical prospective study. J Prosthodont 2012;21:265–9.
Prosthet Dent 2018;119:545–51. [30] Pelaez J, Cogolludo PG, Serrano B, Serrano JF, Suarez MJ. A four-year
[15] Hallmann L, Ulmer P, Wille S, Kern M. Effect of differences in coefficient of prospective clinical evaluation of zirconia and metal-ceramic posterior fixed
thermal expansion of veneer and Y-TZP ceramics on interface phase dental prostheses. Int J Prosthodont 2012;25:451–8.
transformation. J Prosthet Dent 2014;112:591–9. [31] Passia N, Stampf S, Strub JR. Five-year results of a prospective randomised
[16] Pjetursson BE, Brägger U, Lang NP, Zwahlen M. Comparison of survival and controlled clinical trial of posterior computer-aided design-computer-aided
complication rates of tooth-supported fixed dental prostheses (FDPs) and manufacturing ZrSiO4-ceramic crowns. J Oral Rehabil 2013;40:609–17.
implant-supported FDPs and single crowns (SCs). Clin Oral Implants Res [32] Guess PC, Selz CF, Steinhart YN, Stampf S, Strub JR. Prospective clinical split-
2007;18:97–113. mouth study of pressed and CAD/CAM all-ceramic partial-coverage restora-
[17] Rekow ED, Silva NR, Coelho PG, Zhang Y, Guess P, Thompson VP. Performance tions: 7-year results. Int J Prosthodont 2013;26:21–5.
of dental ceramics: challenges for improvements. J Dent Res 2011;90:937–52. [33] Akin A, Toksavul S, Toman M. Clinical marginal and internal adaptation of
[18] Anusavice KJ. Standardizing failure, success, and survival decisions in clinical maxillary anterior single all-ceramic crowns and 2-year randomized
studies of ceramic and metal-ceramic fixed dental prostheses. Dent Mater controlled clinical trial. J Prosthodont 2015;24:345–50.
2012;28:102–11. [34] Zenthöfer A, Ohlmann B, Rammelsberg P, Bömicke W. Performance of zirconia
[19] Wittneben JG, Wright RF, Weber HP, Gallucci GO. A systematic review of the ceramic cantilever fixed dental prostheses: 3-year results from a prospective,
clinical performance of CAD/CAM single-tooth restorations. Int J Prosthodont randomized, controlled pilot study. J Prosthet Dent 2015;114:34–9.
2009;22:466–71. [35] Nicolaisen MH, Bahrami G, Schropp L, Isidor F. Functional and esthetic
[20] Sailer I, Makarov NA, Thoma DS, Zwahlen M, Pjetursson BE. All-ceramic or comparison of metal-ceramic and all-ceramic posterior three-unit fixed dental
metal-ceramic tooth-supported fixed dental prostheses (FDPs)? A systematic prostheses. Int J Prosthodont 2016;29:473–81.
review of the survival and complication rates. Part I: single crowns (SCs). Dent [36] Rinke S, Kramer K, Bürgers R, Roediger M. A practice-based clinical evaluation
Mater 2015;31:603–23. of the survival and success of metal-ceramic and zirconia molar crowns: 5-
[21] Moher D, Liberati A, Tetzlaff J, Altman DG, PRISMA Group. Preferred reporting year results. J Oral Rehabil 2016;43:136–44.
items for systematic reviews and meta-analyses: the PRISMA statement. J Clin [37] Monaco C, Llukacej A, Baldissara P, Arena A, Scotti R. Zirconia-based versus
Epidemiol 2009;62:1006–12. metal-based single crowns veneered with overpressing ceramic for restora-
[22] Wertz DC. Professional perspectives: a survey of Canadian providers. Health tion of posterior endodontically treated teeth: 5-year results of a randomized
Law J 1995;3:59–130. controlled clinical study. J Dent 2017;65:56–63.
[23] Higgins JPT, Green S. Cochrane Handbook for systematic reviews of [38] Schwendicke F, Göstemeyer G, Blunck U, Paris S, Hsu LY, Tu YK. Directly placed
interventions. Version 5.1.0 [updated March 2011]. The Cochrane Collabora- restorative materials: review and network meta-analysis. J Dent Res
tion; 2011. 2016;95:613–22.
[24] Molin MK, Karlsson SL. A randomized 5-year clinical evaluation of 3 ceramic [39] Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat
inlay systems. Int J Prosthodont 2000;13:194–200. Med 2002;21:1539–58.
[25] Walter M, Reppel PD, Böning K, Freesmeyer WB. Six-year follow-up of titanium [40] Light RJ, Pillemer DB. Summing up: the science of reviewing research.
and high-gold porcelain-fused-to-metal fixed partial dentures. J Oral Rehabil Cambridge: Harvard University Press; 1984.
1999;26:91–6. [41] Pjetursson BE, Sailer I, Makarov NA, Zwahlen M, Thoma DS. All-ceramic or
[26] Sailer I, Gottnerb J, Kanelb S, Hammerle CH. Randomized controlled clinical metal-ceramic tooth-supported fixed dental prostheses (FDPs)? A systematic
trial of zirconia-ceramic and metal-ceramic posterior fixed dental prostheses: review of the survival and complication rates. Part II: multiple-unit FDPs. Dent
a 3-year follow-up. Int J Prosthodont 2009;22:553–60. Mater 2015;31:624–39.
[27] Christensen RP, Ploeger BJ. A clinical comparison of zirconia, metal and [42] Elshiyab SH, Nawafleh N, George R. Survival and testing parameters of zirconia
alumina fixed-prosthesis frameworks veneered with layered or pressed based crowns under cyclic loading in an aqueous environment: a systematic
ceramic: a three-year report. J Am Dent Assoc 2010;141:1317–29. review. J Invest Clin Dent 2017;8:e12261.

Please cite this article in press as: S.B. Rodrigues, et al., CAD/CAM or conventional ceramic materials restorations longevity: a systematic review
and meta-analysis, J Prosthodont Res (2018), https://doi.org/10.1016/j.jpor.2018.11.006