Controlled Clinical Trial on the Outcome of Glass Fiber

Composite Cores Versus Wrought Posts and Cast Cores for

the Restoration of Endodontically Treated Teeth:
A 5-Year Follow-up Study
Ellen Cloet, DDS1/Elke Debels, DDS1/Ignace Naert, DDS, PhD2

Purpose: The aim of this study was to compare the 5-year outcomes of glass fiber
composite with cast posts and cores for the restoration of endodontically treated teeth.
Materials and Methods: A total of 143 patients in need of 203 full ceramic restorations
on endodontically treated teeth were included. After primary stratification based on the need
for post or no post, teeth were randomly allotted to test group 1 (prefabricated glass fiber
posts), 2 (custom-made glass fiber posts), or 3 (composite cores without posts). The control
group was treated with gold alloy-based wrought posts and cast cores. Success (original
present) and survival (present after intervention) probability lifetime curves, corrected for
clustering, were drawn over the entire data set. Results: The mean follow-up time was
5.8 years (range: 0.5 to 7.2 years). At 5 years, the success and survival probabilities were
85.2% and 91.5%, respectively. Lifetime curves did not show any significant differences
between the test and control groups for success (P = .85) or survival (P = .57). Moreover,
no significant differences for success or survival could be found among the four groups (the
three test groups and the control group). Conclusion: After 5 years of follow-up, cast gold
and composite post-and-core systems on teeth with ceramic full restorations provided with
a ferrule performed equally well. Int J Prosthodont 2017;30:71–79. doi: 10.11607/ijp.4861

T he restoration of endodontically involved teeth re-

mains a challenge. Lack of tooth tissue may result
in insufficient retention and resistance for the crown
that further weakens the tooth.5–7 Several post ma-
terials, designs, and techniques are available.8 The
cast gold alloy-based post and core has been used
restoration. Eventually, a root canal post is needed to for decades. In spite of its long clinical application,
increase the anchorage of the core buildup material some important disadvantages remain, such as the in-
for retention.1–3 Anterior teeth, and to a lesser extent vasiveness of the technique, especially in anatomical
premolars, are more prone to nonaxial loading com- small roots; the extraoral fabrication procedure lead-
pared with molars during (para)function. A root canal ing to higher costs; and the dark color of the core that
post may then be beneficial for the former.4 For mo- may compromise esthetically demanding cases.
lars, a post-retained core may become obsolete be- In vitro studies have reported increased root frac-
cause sufficient thick dentin walls are left. Besides, tures rates using cast/threaded posts.3,9–11 Composite
larger undercuts in the pulp chamber offer extra re- fiber and ceramic posts were introduced in the early
tention for the core buildup material. 1990s as an alternative to cast alloy-based posts and
The extra retention offered by a post should be cores.12–14 Because fiber posts have elastic moduli
weighed against the sacrifice of healthy tooth tissue similar to that of dentin, the risk of vertical root frac-
ture is said to be reduced.2,3,11,15 Quartz or glass fi-
ber posts (white or translucent) are used in situations
where esthetic demands are high.16 Increased post re-
tention and root fracture resistance have been report-
1Assistant Professor, Unit of Prosthetic Dentistry, Department of ed in in vitro studies using adhesively cemented root
Oral Health Sciences, KU Leuven, Leuven, Belgium. canal posts instead of conventional.17–20 The resultant
2Full Professor, Unit of Prosthetic Dentistry, Department of
homogenous entity in the post allows a more uni-
Oral Health Sciences, KU Leuven, Leuven, Belgium.
form stress distribution as observed in in vitro stud-
Corresponding author: Dr Ignace Naert, Unit of Prosthetic Dentistry, ies, which may better protect the weakened tooth in
KU Leuven, Kapucijnenvoer 7 Box 7001, B-3000 Leuven, Belgium.
patients.12,13,21 A large number of in vitro studies have
Fax: +32 (0) 16 33 23 09. Email: ignace.naert@med.kuleuven.be
compared different restorative techniques for end-
©2017 by Quintessence Publishing Co Inc. odontically treated teeth. However, different materials

Volume 30, Number 1, 2017 71

© 2017 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Glass Fiber Composite Cores Versus Wrought Posts and Cast Cores

and methods have been used, often leading to contra- Stratification and Randomization Procedures
dictory results.22 Therefore, clinical use of cast, metal,
fiber, or no posts should not be encouraged based on The clinical protocol advised by Bolla et al features
those in vitro studies alone. well-defined inclusion criteria, such as delineation
The failure criteria reported in several studies have of the number of residual coronal walls, for a clearer
not always been clearly defined. Different failure assessment of the influence of the remaining tooth
rates have been reported for postendodontic resto- structure on treatment outcomes.34 Therefore, before
rations.23–27 Among the baseline factors influencing randomization took place, two strata were followed:
the clinical outcome of restored pulpless teeth, tooth insufficient or sufficient tissue left to build up the tooth
type and position in the dental arch in relation to oc- core with or without a root canal post.34 Teeth with at
clusal forces and the presence or absence of proxi- least two dentin walls ≥ 2 mm thick and with wide pulp
mal contacts have been identified as relevant.23,28–30 chambers were considered as having sufficient re-
Moreover, preservation of coronal dental tissue has maining tooth tissue to bond the composite core with-
emerged as the critical factor for well-functioning out the use of a root canal post. This procedure clearly
postendodontic restorations.25,31–33 In a Cochrane reflects the tooth distribution in the study population.
systematic review on post-retained restorations, the Within the stratum ‘insufficient remaining tooth tis-
authors advised that clinical protocols should fea- sue left,’ teeth were randomized, with attention to allo-
ture well-defined inclusion criteria, including delin- cation concealment, to either the control group (gold
eation of the number of residual coronal walls, for a alloy-based wrought post with cast cores [Parapost,
clearer assessment of the influence of the remaining Coltène-Whaledent, and Medior 3 Cendres+Métaux])
tooth structure on treatment outcomes.34 Observation or test group 1 or 2. Within both test groups, another
times longer than 3 years are another prerequisite.34 stratification took place depending on the perimeter
For this purpose, the present 5-year prospective clini- of the root canal entrance—small or wide—again re-
cal trial was designed. The null hypothesis was that flecting the influence of remaining tooth structure on
the direct composite fiber post and core techniques treatment outcomes.34
had an equal success/survival probability compared The root canal was considered small when the pe-
with the indirect gold alloy-based wrought posts and rimeter of the 1.4-mm-diameter prefabricated glass
cast cores for teeth covered with full ceramic restora- fiber post used in this study made ≥ 180° contact with
tions. This article addresses the 5-year data following the shaped root canal lumen. A prefabricated glass
the same group of patients as reported in an interim fiber post was used in this case (test group 1; Parapost
report.24 FibreLux, Coltène-Whaledent). It was considered wide
when the 1.4-mm-diameter post made < 180° contact.
Materials and Methods This was the case in ovoid canals and in root canals
in which the anatomical shape was lost due to previ-
Between January 2006 and June 2008, all patients at ous overinstrumentation. A custom-made glass fiber
the Unit of Prosthetic Dentistry, KU Leuven, Belgium, post was used in this case (test group 2; everStick,
in need of ceramic single full restorations on endodon- StickTech). When doubt arose, the principal investiga-
tically treated teeth were screened for participation in tor was asked for his judgment.
the study. Patient and tooth exclusion criteria were set Within the stratum of teeth with sufficient tooth tis-
before intake. At patient level, the exclusion criteria sue remaining, teeth were randomized to either test
were medical conditions that hampered a follow-up group 3 (composite core without post; Clearfill AP-X
for at least 5 years, known allergies to products that and Clearfill SE, Kuraray) or to the control group (gold
were planned to be used in the study, untreated peri- alloy-based wrought post with cast core; Parapost and
odontitis (pocket probing pocket depth ≤ 3 mm) or Medior 3). In test group 3, the core buildup started
high decay sensitivity, and inability or unwillingness to from the pulp chamber and was used only in premo-
give informed consent for participation. At tooth level, lars and molars, because of their anatomical dimen-
the exclusion criteria were serving as abutments to sions, to bond the composite directly without the use
retain a removable or fixed partial dental prosthesis, of a post. Figure 1 shows the stratification/randomiza-
lack of antagonists, a vertical root crack diagnosed tion procedure for the 144 patients in need of a single
under 3× magnification, recurrent endodontic com- restoration on 205 endodontically treated teeth.
plications, and crown fractures or caries extending The clinical coordinator (E.D.) deemed it mandatory
below the bone level crest where crown-lengthening to specify teeth as having insufficient or sufficient
procedures to create a ferrule were either refused or tooth tissue remaining. To calibrate the clinical inves-
contraindicated. The latter three criteria eventually re- tigator and the principal investigator (I.N.), a randomly
sulted in tooth extraction. selected series of extracted teeth (n = 20) was used

72 The International Journal of Prosthodontics

© 2017 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
 Cloet et al

Patients (n = 144)

Teeth (n = 205)

Insufficient tooth tissue left Sufficient tooth tissue left

88.5% (n = 181) 11.5% (n = 24)

Randomization Randomization

Stratification

Control Test 1 Test 2 Control Test 3

49.7% 35.9% 14.4% 45.8% 54.2%
(n = 90)a (n = 65) (n = 26) (n = 11) (n = 13)

Direct
Prefabricated Prefabricated
Prefabricated Custom- composite
wrought post wrought post
post made post buildup
and cast core and cast core
without post

Fig 1   Flow chart of stratification/randomization of the original 144 patients in need of a single restoration on 205 endodontically treated teeth.

to train them to rate the remaining tooth tissue. To test To further standardize the procedure, all post-core
the inter-rater reliability, Cohen’s kappa (κ = 0.71) systems were luted with the same dual-curing adhe-
was calculated. From those teeth, slides were made sive cement Panavia F 2.0/ED Primer II (Kuraray). In
to be used as a reference in case of doubt during the all test groups, the core was built up with highly filled
intake period. When doubt arose, the principal inves- posterior composite Clearfill AP-X and bonded with
tigator was asked for his judgment and consultation the self-etch adhesive system Clearfill SE. All teeth
was organized until agreement was reached. were restored with all-ceramic Al2O3 single crowns
(Procera, Nobel Biocare) and luted with Panavia F 2.0/
Procedures ED Primer II.
Core buildup and full ceramic restorations were
The control group, consisting of a wrought post with made by 29 operators, all of whom were graduated
cast core, differed from the traditionally described dentists employed at the Unit of Prosthetic Dentistry.
fully cast post and core only in that the cast post was A training session in which the protocol was theo-
replaced by a wrought post that was cast onto the retically explained in detail and preclinical hands-on
metal core. This strategy has been followed in the training were given at the start of the study and re-
department for the last 29 years to avoid weakening peated on a biannual basis for calibration purposes.
of the transition between the fully cast post and core All clinical procedures for post placement, composite
during casting procedures, especially in small-diame- buildup, and restoration cementation were performed
ter root canals. Indeed, air turbulence during casting strictly following the manufacturers’ instructions.
procedures at the post-core transition may result in
a porous cast and eventually in early post fracture.35 Variables
Except for the apical third, where the metal post is
mostly congruent with the root canal lumen after root Independent variables included age, sex, tooth, rea-
canal preparation, the mid- and coronal canal luminae son for tooth restoration, remaining tooth tissue, size
are still waxed around the wrought cylindrical post of the root canal lumen, restoration type, and time of
before casting. This results in a thin cement layer and post and/or core insertion. Dependent variables were
mechanically reliable post-core transition without ad- antagonistic status, and occlusion and articulation
ditional root weakening. patterns.

Volume 30, Number 1, 2017 73

© 2017 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Glass Fiber Composite Cores Versus Wrought Posts and Cast Cores

Chipping and fracture of the ceramic

100
restorations were considered compli-
90 cations since they were not intrinsically
80 related to the restoration of the root ca-
nal–treated tooth. All failures occurring
Patients in follow-up (%)

70
between consecutive follow-ups were
60 recorded as well. The study protocol
50 was approved by the Ethical Committee
for Clinical Trials of KU Leuven, 2006.
40
30 Statistical Analysis
20
Besides descriptive statistics, Kaplan-
10
Meier estimates were used to calcu-
0 late success and survival of teeth. The
0 1 2 3 4 5 6 7 8 9
95% pointwise confidence intervals
Time (y)
(CIs) were corrected for clustering of
Fig 2   Kaplan-Meier curve indicating the follow-up time of the restorations and censoring multiple teeth within one patient by re-
the 25 failed restorations (absolute failure) at their moment of failure (symbols). Two failures placing Greenwood’s variance formula
are not shown (arrows). with the formula of Ying and Wei.37
The success rate was compared be-
tween groups using a Cox regression
with the robust sandwich estimate of
Lin and Wei38 for the covariance ma-
Baseline and Follow-up trix. Follow-up times were censored
when tooth loss without failure oc-
Baseline corresponds to the time of post and/or core placement, since curred (competing event). This was
all teeth were loaded with a provisional restoration until the final ce- true for one tooth that was extracted
ramic restoration was ready. Follow-up of patients was scheduled after for orthodontic reasons. Analyses
1, 3, and 5 years (2-year window) from the start of the study. All res- were performed using SAS version 9.2
torations were examined clinically and radiographically. Antagonistic (SAS Institute) and the R package.
status, marginal integrity (decay), periodontal status, and occlusion
and articulation patterns were evaluated. Baseline and follow-up ex- Results
aminations were carried out by one blinded operator.
A distinction was made between absolute and relative failures. Root Patients
fractures or irreparable fractures of the post/core restoration (eg,
fractures where it was impossible to remove the metal post without Originally, 144 patients with 205 res-
further weakening the root) leading to tooth extraction were consid- torations were included at baseline.
ered absolute failures. Loss of post retention, in which recementation One patient with two teeth restored
of the post and core could save the tooth, and reparable fractures never got her final crown restorations
of the core without further weakening of the tooth were considered and was retrieved for the final analysis.
relative failures. Eventually, 143 patients (53% women)
Success was defined as the absence of absolute and relative fail- with a mean age of 47 ± 8.7 years
ures, while survival was defined as the absence of absolute failures. (range: 18–80 years) in need of 203 full
Although endodontic and periodontal failures leading to tooth extrac- single restorations on endodontically
tion are not directly related to the type of root restoration, they were treated teeth were followed for at least
not considered in isolation. Depending on whether or not the actual 5 years (mean: 5.8 years; range: 0.5–7.2
post and core was in situ, they were considered in the survival and years). Patients were censored at their
success lifetime analyses. last follow-up. A total of 9 patients with
When the fiber post (test groups 1 and 2) or the composite core 10 restorations (5 control/5 test) failed
(test group 3) fractured, the failed core buildup was replaced by a cast to attend the 5-year recall; 3 refused
gold alloy-based post and core but kept in its original group accord- further participation ;and the other 6
ing to the intention-to-treat principle,36 as requested by the ethical could not be traced. Restoration recall
committee of KU Leuven. Depending on the failure mode, the same or rates at 1, 3, and 5 years were; 97.0%,
a newly made all-ceramic crown was placed. 92.0%, and 87.2%, respectively (Fig 2).

74 The International Journal of Prosthodontics

© 2017 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
 Cloet et al

100 100

75 3 y: 96.4% (CI: 93.8%–99.0%) 75 3 y: 93.3% (CI: 89.9%–96.6%)

5 y: 91.5% (CI: 87.04%–95.5%) 5 y: 85.2% (CI: 80.0%–90.4%)
Survival (%)

Survival (%)
7 y: 84.4% (CI: 78.4%–90.4%) 7 y: 74.3% (CI: 67.3%–81.3%)
50 50

25 25

0 0
0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7
Time in situ (y) Time in situ (y)

Teeth (n) 203 200 195 189 180 179 174 164 129 91 45 Teeth (n) 203 200 195 187 177 174 167 157 122 86 41
Patients (n) 143 140 136 131 128 128 125 120 95 68 34 Patients (n) 143 140 136 131 127 125 121 115 89 65 32

a b
Fig 3   Lifetimes curves, corrected for clustering, for successful (a) and surviving (b) root-treated teeth sampled over test and control groups.
Number of patients and teeth for each time point are indicated. Survival/success probabilities are calculated from baseline onward.

Root Restorations fracture into the root canal (n = 2), caries (n = 5),
endodontic failure (n = 4), and periodontal failure
Originally, 88.5% of the teeth were rated as having (n = 2).
insufficient remaining tooth tissue; the remainder A total of 21 relative failures was observed, of which
(11.5%) were considered sufficient (Fig 1). Of the res- 10 belonged to the control and 11 to the test groups.
torations, 49.3% (n = 101) were cast gold alloy-based Of these, 7, 3, and 1 belonged to test groups 1, 2, and
post and cores, 32.0% (n = 65) were prefabricated 3, respectively. The reasons for relative failures were
glass fiber posts with composite cores, 12.8% (n = loss of retention of the post (n = 9), endodontic failure
26) were custom-made glass fiber posts with com- (n = 4), and post fracture requiring post replacement
posite cores, and 5.9% (n = 13) were composite cores (n = 8).
without posts. The majority of the patients (71%) had Lifetime curves, corrected for clustering, were
one restoration only, while 18%, 8%, 2%, and 1% of drawn for successful and surviving teeth sampled
the patients had two, three, four, and five restorations, over test and control groups. At year 5, the success
respectively. and survival probabilities, respectively, were 85.2%
and 91.5%. At year 7, the success and survival prob-
Failures abilities, respectively, were 74.3% and 84.4% (Fig 3).
Lifetime curves, corrected for clustering, could not
At the 1-year recall, neither absolute, relative, nor show significant differences between the four groups
endodontic failures were observed. However, four all- separately (test groups 1 to 3 and control group) for
ceramic restorations showed complications such as success or survival (Fig 4). The 5-year success prob-
chipping of the layering ceramic (n = 3) or fracture abilities were 87.8% (CI: 74.7–100.0) for test group 2,
(n = 1). The former were polished, while the latter was 86.9% (CI: 80.0–93.8) for the control group, 81.6% (CI:
replaced by a new restoration and the tooth was kept 71.8–91.4) for test group 1, and 83.3% (CI: 62.2–100.0)
in function. for test group 3, while the survival probabilities were
At the 5-year recall visit a total of 25 absolute fail- 92.1% (CI: 81.7–100.0) for test group 2, 91.4% (CI:
ures was observed. A Kaplan-Meier curve, censoring 84.2–98.6) for test group 1, 91.7% (CI: 76.0–100.0) for
the failed restorations at their moment of failure, is test group 3, and 91.2% (CI: 85.3–97.0) for the control
shown in Fig 2. Of the failures, 14 belonged to the group.
control group and 11 to the test group (6 to group The failure distribution between anterior and poste-
1, 2 to group 2, and 3 to group 3). The reasons for rior teeth for absolute and relative failures is shown in
the absolute failures were root fracture (n = 12), post Table 1. For absolute failures, 14% and 12% occurred in

Volume 30, Number 1, 2017 75

© 2017 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Glass Fiber Composite Cores Versus Wrought Posts and Cast Cores

100 100

75 75
Survival (%)

Survival (%)
50 50
Control Control
Test 1 Test 1
Test 2 Test 2
25 Test 3 25 Test 3

0 0
0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7
Time in situ (y) Time in situ (y)

Control 100 98 96 95 89 88 85 81 68 50 24 Control 100 98 96 93 87 85 82 78 66 49 23

Test 1 65 64 62 59 57 57 55 51 40 26 14 Test 1 65 64 62 60 56 55 51 48 37 23 12
Test 2 26 26 26 24 23 23 23 21 15 10 5 Test 2 26 26 26 23 23 23 23 21 14 9 4
Test 3 12 12 11 11 11 11 11 11 6 5 2 Test 3 12 12 11 11 11 11 11 10 5 5 2

a b
Fig 4   Lifetimes curves, corrected for clustering, for successful (a) and surviving (b) root-treated teeth for test groups 1, 2, and 3 and the control
group. Number of patients and teeth for each time point are indicated. Survival/success probabilities are calculated from baseline onward.

Table 1   F ailure Distribution Between Anterior and Posterior Teeth for Absolute and Relative Failures
Control Test 1 Test 2 Test 3
Failure Anterior Posterior Anterior Posterior Anterior Posterior Posterior
type (n = 40) (n = 60) (n = 16) (n = 49) (n = 12) (n = 14) (n = 12)
Absolute 6 8 1 5 2 0 3
Relative 6 4 3 4 3 0 1

anterior versus posterior teeth, respectively (P > .05), calculation. Based on existing data in the literature at
while for the relative failures the rates were 17% and the start of the study, a power calculation resulted in
7%, respectively (P < .05). a sample size of 800 teeth to reach a power of 80%
More wear facets were seen at the last recall com- (two-sided α = .05). This would have lengthened the
pared with baseline (test: 13.8%, control: 11.2%). study intake period by at least 3 times and therefore
During excursion movements, no real pattern of ar- was not realistic. When the literature between 1990
ticulation (frontal guidance, canine guidance, group and 2015 was searched for prospective studies deal-
guidance) could be observed either from baseline to ing with the same topic, 20 articles were isolated.
year 5 or between test and control groups. However, only 4 of these had numbers exceeding the
present study (Monticelli et al39 [n = 225]; Fokkinga
Discussion et al40 [n = 307]; Salvi et al41 [n = 325]; and Ferrari et
al42 [n = 360]), but all were far less than 800 teeth. A
The present clinical trial determined the survival prob- multicenter approach could have partially solved the
ability of endodontically treated teeth restored with recruitment of patients, but the present authors were
either gold alloy-based wrought post and cast cores not successful in this respect. This limitation should
or one of three alternative composite fiber post/core be taken into account when interpreting the results.
restorations. Following a proper prosthodontic treat- On the other hand, the follow-up is at least 5 years,
ment plan, a total of 203 restorations placed in 143 the study is prospective, and the teeth were stratified
patients were followed up. This article deals with the to the amount of tissue left before randomization took
≥ 5-year recall data and uses the same data set as the place. This eliminated an important confounding fac-
interim report by Zicari et al.24 tor. However, it also resulted in test group 3 having
A limitation of this monocenter study is that the only 13 teeth with few events. Low numbers in groups
sample size was determined by the predefined in- 2 and 3 inherently reflect the distribution of endodon-
clusion period and not by an a priori formal power tically treated teeth in a study population based on

76 The International Journal of Prosthodontics

© 2017 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
 Cloet et al

tooth tissue left after tooth preparation. For this rea- according to the expected dentin height left after
son, comparisons were made only between test and tooth preparation. The 17-year survival rates at resto-
controls and not among the different test groups. ration level ranged from 71% to 80%, and at tooth level
The null hypothesis, that the success/survival prob- from 83% to 92%. The type of core restoration showed
ability for the direct composite fiber post and core no influence on the survival probability either at tooth
techniques was equal to that of the indirect gold alloy- or at restoration level. The authors could not correlate
based wrought posts and cast cores for teeth covered the occurrence of irreparable failures and the need for
with full ceramic restorations, was accepted. tooth extraction with the placement of metallic posts.
The overall survival and success probabilities at This outcome fully corroborates the outcome of the
5 years were 91.5% and 85.2%, respectively. This present study.
outcome is in line with those reported in previous Ferrari et al42 investigated the influence of the
prospective and retrospective studies.23,26,27,42–45 amount of remaining coronal dentin in endodontically
However, the success rates are difficult to compare treated premolars in a 6-year RCT. Regardless of the
because of differences in study designs and criteria. restorative procedure, the preservation of at least one
A retrospective study by Ferrari et al in 2000 found coronal wall significantly reduced failure risk, even
survival rates of 84% and 95% (P < .001) for resto- when a ferrule was absent. This finding was not in-
rations with cast posts and cores and carbon fiber vestigated in the present study but is not in line with
posts, respectively, after 4 years of service.46 The others on the ferrule effect.5,7,31,32 A possible explana-
authors also observed that more unfavorable failures tion for this discrepancy identified by the authors as a
occurred in teeth restored with cast posts and cores, shortcoming is that the teeth were assigned based on
of which 9% were irreparable root fractures. In the the amount of coronal dentin left after root canal treat-
present study, 14 of 25 absolute failures occurred in ment but before abutment preparation. All teeth in-
the control group (teeth restored with cast posts and volved in our study were selected only when it became
cores). This difference in outcome is probably due evident that after root/core preparation, a full restora-
to the study design, in which teeth were originally tion was the only option. Overall, the present results
stratified according to the amount of tooth tissue left. corroborate well in terms of survival probabilities.
Within this constraint, it seems that the mode of root The most recent systematic review could not find dif-
restoration does not play a role. ferences in outcome for metal versus fiber posts, which
In a long-term retrospective study in which the corroborates the present results.47 However, limitations
clinical outcome of fiber posts (n = 985) was evalu- were noted, such as the inclusion of retrospective co-
ated, survival rates between 89% and 93% were hort studies and lack of information about the amount
observed after 7 to 11 years.26 Teeth were restored of coronal tissue remaining, the presence of a ferrule
with either full-crown restorations or direct compos- effect, the location of the tooth (anterior versus poste-
ite restorations. A total of 79 failures were recorded: rior), or occlusion features (antagonists). All these fac-
39 due to endodontic reasons, 1 root fracture, 1 fiber tors were well controlled in the present study.
post fracture, 17 crown dislodgements, and 21 due to Although absolute failures did not differ between
post debonding. Although no reference was made to anterior and posterior teeth in this study, twice as
the amount of residual coronal structure remaining, many relative failures occurred in anterior teeth. The
the authors suggested that the mechanical failures most common type of failure was dislodgement of
were primarily related to the lack of coronal tooth the post, which is in line with previous clinical find-
structure. ings.23,26,44,45,48–50 Of these post dislodgements, 30.9%
Prospective studies are rather scarce, and random- occurred in anterior teeth versus 18.02% in posterior
ized controlled trials (RCTs) even more so. Differences teeth. This is in line with previous clinical studies in
in study design, inclusion criteria, and observation which root posts in incisors and canines failed ap-
periods, make those that are available obsolete for proximately three times as often compared with res-
comparisons. torations in premolars or molars.28,49,50
Fokkinga et al40 collected the survival data for 307 Besides tooth type and position in the dental arch,
restorations in a RCT with follow-up of up to 17 years. the type of the final crown restoration (single crowns
Restorations were made by 18 operators and were cast versus fixed partial dental prostheses) and the ab-
posts and cores, prefabricated metal post with resin sence of approximal contacts have been mentioned
composite cores, and post-free all-composite cover- as significant predictors for failure.30,49 In this study,
age restorations. The final restoration was either a full only single teeth restored with all-ceramic crowns
or partial overlay restoration, depending on the tooth and with antagonists were selected, which is con-
tissue left after endodontic treatment. Before treat- sidered less favorable than when fixed partial dental
ment allocation, the recipient tooth was categorized prostheses are used.

Volume 30, Number 1, 2017 77

© 2017 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Glass Fiber Composite Cores Versus Wrought Posts and Cast Cores

Due to the study design, the number of teeth under 11. Sirimai S, Riis DN, Morgano SM. An in vitro study of the frac-
investigation in test groups 2 and especially 3 were ture resistance and the incidence of vertical root fracture
of pulpless teeth restored with six post-and-core systems.
limited, and uncertainty remains for the latter group.
J Prosthet Dent 1999;81:262–269.
Nevertheless, the results corroborate well with those 12. Duret B, Reynaud M, Duret F. New concept of coronoradicular
of other studies. The cast post and core has been used reconstruction: The Composipost (1) [in French]. Chir Dent Fr
for many decades, and this study, which started in 1990;540:131–141.
2007, used this as a viable control. This restoration type 13. Duret B, Reynaud M, Duret F. A new concept of corono-
radicular reconstruction: The Composipost (2) [in French]. Chir
is still frequently used all over the world, but hard data
Dent Fr 1990;60:69–77.
remain scarce. Although the results did not reveal sig- 14. Duret B, Reynaud M, Duret F. Intérêt des matériaux à structure
nificant differences between test and control groups, unidirectionnelle dans les reconstructions corono-radiculaires.
there are surely practical benefits, such as decreased J Biomat Dent 1992;7:45–47.
cost and time, that support using direct techniques. 15. Deutsch AS, Musikant BL, Cavallari J, Lepley JB. Prefabricated
dowels: A literature review. J Prosthet Dent 1983;49:498–503.
However, this is outside the scope of this study.
16. Vichi AM, Ferrari M, Davidson CL. Influence of ceramic and
cement thickness on the masking of various types of opaque
Conclusions posts. J Prosthet Dent 2000;83:412–417.
17. Cohen BI, Pagnillo MK, Newman I, Musikant BL, Deutsch AS.
After 5 years of follow-up, both cast gold and com- Retention of three endodontic posts cemented with five dental
cements. J Prosthet Dent 1998;79:520–525.
posite post-and-core systems on teeth with ceramic
18. Rosenstiel SF, Land MF, Crispin BJ. Dental luting agents: A re-
full restorations, provided with a ferrule, performed view of the current literature. J Prosthet Dent 1998;80:280–301.
equally well. No clinically relevant differences were 19. Hagge MS, Wong RD, Lindemuth JS. Retention strengths of
seen between test and control groups. five luting cements on prefabricated dowels after root ca-
nal obturation with a zinc oxide/eugenol sealer: 1. Dowel
space preparation/cementation at one week after obturation.
Acknowledgments
J Prosthodont 2002;11:168–175.
20. O’Keefe KL, Miller BH, Powers JM. In vitro tensile bond strength
The authors reported no conflicts of interest related to this study. of adhesive cements to new post materials. Int J Prosthodont
2000;13:47–51.
References 21. Pegoretti A, Fambri L, Zappini G, Bianchetti M. Finite element
analysis of a glass fiber reinforced composite endodontic post.
Biomaterials 2002;23:2667–2682.
1. Freedman GA. Esthetic post-and-core treatment. Dent Clin 22. Heydecke G, Peters MC. The restoration of endodontically
North Am 2001;45:103–116. treated, single-rooted teeth with cast or direct posts and cores:
2. Sorensen JA, Martinoff JT. Intracoronal reinforcement and A systematic review. J Prosthet Dent 2002;87:380–386.
coronal coverage: A study of endodontically treated teeth. 23. Naumann M, Blankenstein F, Dietrich T. Survival of glass fibre
J Prosthet Dent 1984;51:780–784. reinforced composite post restorations after 2 years-an obser-
3. Trope M, Maltz DO, Tronstad L. Resistance to fracture of restored vational clinical study. J Dent 2005;33:305–312.
endodontically treated teeth. Endod Dent Traumatol 1985; 24. Zicari F, Van Meerbeek B, Debels E, Lesaffre E, Naert I. An
1:108–111. up to 3-year controlled clinical trial comparing the outcome of
4. Torbjörner A, Fransson B. Biomechanical aspects of prosthetic glass fiber posts and composite cores with gold alloy-based
treatment of structurally compromised teeth. Int J Prosthodont posts and cores for the restoration of endodontically treated
2004;17:135–141. teeth. Int J Prosthodont 2011;24:363–372.
5. Creugers NH, Mentink AG, Fokkinga WA, Kreulen CM. 5-year 25. Mancebo JC, Jiménez-Castellanos E, Cañadas D. Effect of tooth
follow-up of a prospective clinical study on various types of type and ferrule on the survival of pulpless teeth restored with
core restorations. Int J Prosthodont 2005;18:34–39. fiber posts: A 3-year clinical study. Am J Dent 2010;23:351–356.
6. Yang HS, Lang LA, Molina A, Felton DA. The effects of dowel 26. Ferrari M, Cagidiaco MC, Goracci C, et al. Long-term retro-
design and load direction on dowel-and-core restorations. spective study of the clinical performance of fiber posts. Am J
J Prosthet Dent 2001;85:558–567. Dent 2007;20:287–291.
7. Sorensen JA, Engelman MJ. Ferrule design and fracture re- 27. Ferrari M, Cagidiaco MC, Grandini S, De Sanctis M, Goracci C.
sistance of endodontically treated teeth. J Prosthet Dent 1990; Post placement affects survival of endodontically treated pre-
63:529–536. molars. J Dent Res 2007;86:729–734.
8. Asmussen EA, Peutzfeldt, Heitmann T. Stiffness, elastic limit, 28. Naumann M, Blankenstein F, Kiessling S, Dietrich T. Risk fac-
and strength of newer types of endodontic posts. J Dent 1999; tors for failure of glass fiber-reinforced composite post resto-
27:275–278. rations: A prospective observational clinical study. Eur J Oral
9. King PA, Setchell DJ. An in vitro evaluation of a prototype CFRC Sci 2005;113:519–524.
prefabricated post developed for the restoration of pulpless 29. Aquilino SA, Caplan DJ. Relationship between crown place-
teeth. J Oral Rehabil 1990;17:599–609. ment and the survival of endodontically treated teeth.
10. Maccari PC, Cosme DC, Oshima HM, Burnett LH Jr, Shinkai RS. J Prosthet Dent 2002;87:256–263.
Fracture strength of endodontically treated teeth with flared 30. Caplan DJ, Kolker J, Rivera EM, Walton RE. Relationship be-
root canals and restored with different post systems. J Esthet tween number of proximal contacts and survival of root canal
Restor Dent 2007;19:30–36. treated teeth. Int Endod J 2002;35:193–199.

78 The International Journal of Prosthodontics

© 2017 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
 Cloet et al

31. Juloski J, Radovic I, Goracci C, Vulicevic ZR, Ferrari M. Ferrule 43. Fredriksson M, Astbäck J, Pamenius M, Arvidson K. A ret-
effect: A literature review. J Endod 2012;38:11–19. rospective study of 236 patients with teeth restored by car-
32. Stankiewicz NR, Wilson PR. The ferrule effect: A literature re- bon fiber-reinforced epoxy resin posts. J Prosthet Dent 1998;
view. Int Endod J 2002;35:575–581. 80:151–157.
33. Zhi-Yue L, Yu-Xing Z. Effects of post-core design and ferrule 44. Ferrari M, Vichi A, Mannocci F, Mason PN. Retrospective study
on fracture resistance of endodontically treated maxillary cen- of the clinical performance of fiber posts. Am J Dent 2000;
tral incisors. J Prosthet Dent 2003;89:368–373. 13:9B–13B.
34. Bolla M, Muller-Bolla M, Borg C, Lupi-Pegurier L, Laplanche O, 45. Malferrari S, Monaco C, Scotti R. Clinical evaluation of teeth
Leforestier E. Root canal posts for the restoration of root filled restored with quartz fiber-reinforced epoxy resin posts. Int J
teeth. Cochrane Database Syst Rev 2007;1:CD004623. Prosthodont 2003;16:39–44.
35. Bergman M, Holmlund L, Wictorin L. Noble metal alloy wires in 46. Ferrari M, Vichi A, Garcia-Godoy F. Clinical evaluation of fiber-
cast posts. Mechanical properties and microstructure. Odontol reinforced epoxy resin posts and cast post and cores. Am J
Revy 1974;25:273–287. Dent 2000;13:15B–18B.
36. Pocock SJ. Clinical Trials: A Practical Approach. Chichester: 47. Figueirido FE, Martins-Filho PR, Fari ES. Do metal post–re-
John Wiley & Sons, 1984. tained restorations result in more root fractures than fiber
37. Ying Z, Wei LJ. The Kaplan-Meier estimate for dependent fail- post–retained restorations? A systematic review and meta-
ure time observations. J Multivar Anal 1994;50:17–29. analysis. J Endod 2015;41:309–316.
38. Lin DY, Wei LJ. The robust inference for the Cox Proportional 48. King PA, Setchell DJ, Rees JS. Clinical evaluation of a car-
Hazards Model. J Am Stat Assoc 1989;84:1074–1077. bon fibre reinforced carbon endodontic post. J Oral Rehabil
39. Monticelli F, Grandini S, Goracci C, Ferrari M. Clinical behav- 2003;30:785–789.
ior of translucent-fiber posts: A 2-year prospective study. Int J 49. Naumann M, Sterzenbach G, Alexandra F, Dietrich T.
Prosthodont 2003;16:593–596. Randomized controlled clinical pilot trial of titanium vs glass fi-
40. Fokkinga WA, Kreulen CM, Bronkhorst EM, Creugers NH. Up ber prefabricated posts: Preliminary results after up to 3 years.
to 17-year controlled clinical study on post-and-cores and cov- Int J Prosthodont 2007;20:499–503.
ering crowns. J Dent 2007;35:778–786. 50. Torbjorner A, Fransson B. A literature review on the prosthetic
41. Salvi GE, Segrist Guldener BE, Amstad T, Joss A, Lang NP. treatment of structurally compromised teeth. Int J Prosthodont
Clinical evaluation of root filled teeth restored with or with- 2004;17:369–376.
out post-and-core systems in a specialist practice setting. Int
Endod J 2007;40:209–215.
42. Ferrari M, Vichi A, Fadda GM, et al. A randomized controlled
trial of endodontically treated and restored premolars. J Dent
Res 2012;91:72S–78S.

Volume 30, Number 1, 2017 79

© 2017 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.