Accepted Manuscript

Longevity of Self-Etch Dentin Bonding Adhesives compared to Etch-and-Rinse Dentin

Bonding Adhesives: A Systematic Review

Nader Masarwa, BDS, JDB., Senior Specialist- Conservative Dentistry, Ahmed

Mohamed, PhD, DDS., Iyad Abou-Rabii, DDS, PGCertPharma, PGDip OMFS
MResBiomedSc PhD FADFE, Senior research fellow, Rawan Abu Zaghlan, DDS
(Hons), MFDS RCSI, Specialist trainee, Liviu Steier, Dr.med.dent., Assoc. Clinical
Professor

PII: S1532-3382(16)30007-0
DOI: 10.1016/j.jebdp.2016.03.003
Reference: YMED 1114

To appear in: The Journal of Evidence-Based Dental Practice

Received Date: 28 February 2016

Accepted Date: 22 March 2016

Please cite this article as: Masarwa N, Mohamed A, Abou-Rabii I, Abu Zaghlan R, Steier L, Longevity
of Self-Etch Dentin Bonding Adhesives compared to Etch-and-Rinse Dentin Bonding Adhesives:
A Systematic Review, The Journal of Evidence-Based Dental Practice (2016), doi: 10.1016/
j.jebdp.2016.03.003.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT

Longevity of Self-Etch Dentin Bonding Adhesives compared to Etch-and-Rinse Dentin Bonding

Adhesives: A Systematic Review

Nader Masarwa, BDS, JDB.

Phone: (M) 00962797695726

Fax: 0096253241607

E-mail: nadermasarwa@gmail.com

Senior Specialist- Conservative Dentistry

Al-Hussein Medical City. The Royal Medical Services, Jordan Army.

Contribution to the paper: Hypothesis, searches, contributed significantly in writing all parts of the
manuscript.

Ahmed Mohamed, PhD, DDS.

University of Warwick, Coventry, UK.

Contribution to the paper: contributed substantially to the discussion.

Iyad Abou-Rabii,

DDS, PGCertPharma, PGDip OMFS MResBiomedSc PhD FADFE

Senior research fellow in Mantford University, Leicester, UK.

Contribution to the paper: Writing parts of Abstract and Introduction.

Rawan Abu Zaghlan, DDS (Hons), MFDS RCSI.

Specialist trainee- Department of Restorative Dentistry

Edinburgh Dental Institue.

Contribution to the paper: Database searches, contributed to writing Materials and Methods part.

Liviu Steier, Dr.med.dent.. Assoc. Clinical Professor.

University of Warwick, Coventry, UK.

Contribution to the paper: Idea, writing the abstract, revision and editing the manuscript.
 ACCEPTED MANUSCRIPT

ABSTRACT

Objectives: A systematic review and meta-analysis were performed to compare longevity of

Self-Etch Dentine Bonding Adhesives to Etch-and-Rinse Dentin Bonding Adhesives.

Material and methods: The following databases were searched for PubMed, MEDLINE, Web of
Science, CINAHL, the Cochrane Library complemented by a manual search of the Journal of
Adhesive Dentistry. The MESH keywords used were: “etch and rinse”, “total etch”, “self-etch”,
“dentine bonding agent”, “bond durability”, and “bond degradation”. Included were in-vitro
experimental studies performed on human dental tissues of sound tooth structure origin. The
examined Self-Etch Bonds were of two subtypes; Two Steps and One Step Self-Etch Bonds,
while Etch-and-Rinse Bonds were of two subtypes; Two Steps and Three Steps. The included
studies measured micro tensile bond strength (µTBs) to evaluate bond strength and possible
longevity of both types of dental adhesives at different times. The selected studies depended on
water storage as the aging technique. Statistical analysis was performed for outcome
measurements compared at 24 hours, 3 months, 6 months and 12 months of water storage.

Results: After 24 hours (p-value = 0.051), 3months (p-value = 0.756), 6 months (p-value =
0.785) and 12 months of water storage self-etch adhesives showed lower µTBs when compared
to the etch-and-rinse adhesives, but the comparisons were statistically insignificant.

Conclusion: In this study, longevity of Dentine Bonds was related to the measured µTBs.
Although Etch-and-Rinse bonds showed higher values at all times, the meta-analysis found no
difference in longevity of the two types of bonds at the examined aging times.

Key words: Dentine Bonding, Longevity, Self-Etch, Etch-and-Rinse.

 ACCEPTED MANUSCRIPT

Introduction

The evolution of adhesive materials and techniques has transformed the scope of dental
practice. Changing concepts of restorative dentistry from cementation and mechanical retention
into bonding is today’s method of direct and indirect restoration of teeth. The main goal of
bonding is to achieve an intimate contact between tooth structure and restorative materials.
While bonding to enamel has proven to be successful, bonding to dentine remains unreliable
(15). This is due to its higher water content, increased organic collagen concentration and its
structural variability. Dentine bonding is both micromechanical and chemical. Micromechanical
depends on the infiltration of bonding resin into the dentine, replacing the dissolved
hydroxyapetite crystals and surrounding the organic matrix content which is mainly collagen
fibrils. Here an inter-diffusion layer is formed and is called the hybrid layer (13). Chemical
bonding takes place with the effect of functional acidic monomers interaction with
hydroxyapetite crystals in Dentine. These monomers are composed of carboxylic, phosphonic
or phosphate groups. Phenyl-P, 10-methacryloyloxydecyl dihydrogenphosphate (10-MDP) is
one example of functional monomers which can interact with calcium ions in hydroxyapetite
crystals.(9)

The development of new dentin bonding materials, focused on minimizing the number of steps
involved in bonding procedure. A widely accepted classification of bonding systems depends
on whether a separate acid etching step is utilized or not. Accordingly, two main types are
available; self-etch (SE), and etch-and-rinse (ER) bonding systems (earlier known as total-etch
bonding system). Moreover, there is a sub-classification of these categories according to the
number of steps implied in each procedure depending on whether the prime and bond solutions
are used in one step from a mixed bottle or in two separate steps from two different bottles. So
for SE bonds there are one step (1SE) and two steps (2SE), while for ER bonds there are two
steps (2ER) and three steps (3ER). Omitting the etching step not only minimized time but also
prevented recontamination of the cavity by blood and saliva during etch washing and drying
which make SE a less technique sensitive procedure than ER bond (19).

Besides being a less technique sensitive, SE bonds are claimed to have other advantages, such
as reduced postoperative pain because residual smear plugs are left which expose less dentinal
tubules and causes less dentinal fluid flow than ER bonds (10).

Researchers found that durability of the bond plays a secondary role on the clinical survival of
restoration. Their laboratory durability studies found that resin-dentin bonds degrade at a much
faster rate than clinical restorations take to fail (16). Other authors concluded that the durability
of adhesive bond between resin and tooth structure is of significant importance for the longevity
of adhesive restorations (1). Even though most adhesives are capable of providing an adequate
immediate bond to enamel and dentine, the longevity of dentine bond remains an issue (6).

All dentine bonding materials undergo degradation mechanisms of which are not completely
understood. For Etch-and-rinse bonding systems, the major degradation is thought to be based
on both hydrolysis and enzymatic breakdown of the collagen fibrils and the polymerized resin
 ACCEPTED MANUSCRIPT

matrix in the hybrid layer (3).For Self-etch bonding systems, degradation was attributed to its
acidic content which increases the hydrophilicity of the adhesive layer and leads to water uptake
and plasticization (20).

Durability of dentine bonding can be affected by many factors such as: time, temperature
chemical and mechanical interactions. Several comparative studies have been published to
evaluate the changes of dentine bonding strength. These studies compared bond strength
values between different types of dentine bonding systems at different times and under various
changing factors. Most authors agree that measuring microtensile bond strength (µTBs) is of
fundamental importance to evaluate the bonding strength and consequently the possible
longevity of the resin-dentine bond (6). Furthermore, different artificial aging techniques were
used to resemble the changes in oral cavity. The most commonly used method is water storage,
where the bonded specimens are stored in 37°C water for a certain time. The results of these
studies varied also; while some studies found no differences in bonding strength between SE
and ER adhesives after 15 months (13), other studies found significant differences at as early
as six months (20).The differences were noticed under different aging times and conditions.

The purpose of this study was to review peer-reviewed literature that compare the bond
strength between SE dentine bonding systems and ER dentine bonding systems over time. The
question we need to answer here is “Do SE dentine bonding systems have more longevity
than ER dentine bonding systems?”
 ACCEPTED MANUSCRIPT

Materials and Methods

This paper will take the form of a literature review. A literature search was carried out and the
available literature was critically appraised to produce a well balanced picture by comparing and
contrasting different views on the issue.

There are two main categories of bonding systems in this study: self-etch bond and etch-and-
rinse bonds, which will be denoted as SE and ER respectively. Moreover, there is a sub-
classification of these categories according to the number of steps implied in each procedure:
One step (1SE), two steps (2SE, 2ER) and three steps (3ER).

Search strategies:

A literature search was performed on a personal computer using the following electronic
databases; Pubmed, MEDLINE (OVID), Web of Science, CINAHL and the Cochrane Library. In
addition, the Journal of Adhesive Dentistry was manually searched.
The literature review does lend itself to some extent to the PICO format:

Population: Teeth treated with dentine adhesives

Intervention: Self-etch adhesive systems
Comparison: Etch and rinse adhesive systems
Outcome: Microtensile bond strength
Different sets of keywords were used: “etch and rinse”, “total etch”, “self-etch”, “dentine bonding
agent”, “bond durability”, “bond degradation”.

Inclusion/exclusion criteria:

The following criteria were accepted in order to include studies with similar testing methods:

1. No restrictions to searches were made regarding language or date of publishing.

2. Only studies that compare between self-etch bonding systems and Etch-and-rinse
bonding systems were reviewed.
3. In-vitro experimental studies were selected for the purpose of this study.
4. Studies done on human dental tissues of sound tooth structure origin were included,
regardless the type of tooth or whether the included teeth were permanent or deciduous.
5. Studies on aged samples using water as storage media were selected.
6. Studies done on Enamel only were excluded.
7. Some papers provided insufficient raw data in the published articles, so attempts to
request the needed data were sent to corresponding authors by emails. The studies of
who didn’t respond were excluded for the purpose of insufficient data provided.

The full-text articles were independently read by two investigators. Nine articles have been
included for this systematic review (Table 1).
The included studies used extracted sound human teeth to test the strength of different
commercially used bonds. Different bonds were applied to cut tooth structure, and then
 ACCEPTED MANUSCRIPT

composite fillings were inserted. The resulting blocks were then sectioned into slices and stored
in water for different time periods. Testing the bond strength depended on measuring micro
tensile bond strengths (µTBs) of the stored blocks using certain measuring devices.

Statistical analysis:

Different commercial brands are grouped in 4 subtypes: 1SE, 2SE, 2ER, 3ER. Therefore, a first
aggregation of mean and standard deviation has been necessary in order to analyze the
outcome measure of the study.

First of all, the average mean and standard deviation of each subtype (1SE, 2SE, 2ER, 3ER)
was measured independently.

In order to measure the average µTBs of SE bonds, the mean of all the means of the
commercial brands belonging to 1SE and 2SE systems is calculated. The same method is done
for calculating the outcome measure of ER bonds.

The same strategy will be considered to calculate pooled standard deviations (SD)

The data were extracted from each study and overall SE and ER means (average µTBs
measured in MPs) and standard deviations were calculated and organized in a separate table
for the different periods of aging. (Tables 2).

The meta-analysis of this report is based on random-effects model, assuming that studies are
not always regarded as equal (different methods and/or different characteristics of the sample).

In other words, studies are assumed like a random selection of a hypothetical population of
studies. This population could be an infinite set of studies about the topic which could be
already done or being a project in the future.

Mean difference (MD) among groups of bond strength will be calculated for each study (Always
calculated as SE minus ER). Weighted mean difference (WMD) will be used to analyze the
global effect. It is assumed that all the individual studies have the same units of measurement
and conditions.

WMD estimations are accompanied of the confidence interval at 95% of the standard error and
of the p-value of nullity contrast (null hypothesis: adhesion factor doesn’t have an effect over
tensile bond strength).

The level of significance used in the analysis has been 5% (α=0.05)1.

1
Any p-value below 0.05 is indicative of a relation statistically significant. To the contrary, a p-value above or equal
to 0.05 indicates lack of relation.
 ACCEPTED MANUSCRIPT

Total sample size is 722 samples (teeth, laboratory models) at the basal time. There are 504
SE-type adhesions and 218 ER-type. With these samples sizes in a meta-database, t-test
reaches a power of 0.86 to detect as a significant effect size of 0.25 (small), with a confidence of
95%.

The software used to perform this meta-analysis has been R 3.0.2, `metafor’ package.
 ACCEPTED MANUSCRIPT

Results

Selected papers:

The search strategy of literature resulted in identification of 9 relevant published papers. The
aforementioned papers are in-vitro studies conducted on extracted human teeth. Unfortunately,
there are no studies of higher quality in this field as the search strategy has not detected any
human interventional or observational studies that are located high in the hierarchy of evidence.

Quality assessment:

The quality assessment of in-vitro studies often presents a challenge to researchers as there
are no published CASP tools for critical appraisal of in-vitro studies. CASP tools are considered
as the "gold standard" for critical appraisal of medical literature. Therefore, any alternative tools
for critical appraisal must be verified first for their validity and reliability.
In this review, the Timmer’s Analysis Tool is used to critically appraise the selected 9 papers.
The Timmer’s Analysis Tool had been verified for its validity and reliability and it therefore can
be used in systematically reviewing in-vitro studies (18). It consists of a checklist of 19 questions
designed to assess key aspects in in-vitro research. Each question has a value in points. The
more points a paper scores, the higher the quality of its research (18).

By applying the Timmer’s Analysis Tool on the selected 9 papers, a number of areas of
deficiencies have been detected. Table 3 shows a summary of quality assessment of the
selected papers.

Study design:

According to Timmer’s Analysis Tool, the type of study design plays an important role in
calculating the summary score which is the key number that estimates the quality of the
research.

In this review, all the included studies were basic science interventional studies (before and
after studies) with no controls used, except the studies by Van Landuyt et al 2010, De Munak et
al 2010 and Feitosa et al. 2012. Therefore, the later 3 papers score 4 points while the remaining
studies score 2 points. Subsequently, the maximum possible score for the later 3 studies will be
34, while the remaining studies can score maximum possible score of 32.

By applying the Timmer’s Analysis Tool on the selected papers, most of the papers scored
roughly similar summary scores (0.58 to 0.62) which could be understood that the studies were
 ACCEPTED MANUSCRIPT

of similar quality of evidence. It was also noticeable that the study by Martins et al 2009 scored
the highest summary score which would nominate it as the study with the highest quality of
research. The summary scores are shown in Table 4.

As to other elements of study design, all of the selected papers formulated research questions
properly and presented their questions, aims and objectives clearly. Their study designs were
deemed appropriate for the study questions and aims. However, only the studies by Van
Landuyt et al 2010, De Munak et al 2010 and Feitosa et al. 2012 used controls in their design
which would reduce any risk of bias in the later studies. It is not clear why the remaining studies
did not implement the introduction of controls in their designs. Theoretically, it was possible to
design their studies to include control groups.

Randomization:

All of the selected studies claimed randomised allocation of interventions into groups of
extracted teeth. However, none of the selected studies reported the randomisation protocol that
had been recruited. Randomisation protocols such as computer generated protocol or sealed
envelopes could have been used easily in all of the selected studies.

Blinding:

All of the selected studies covered their bonded cavities with composite fillings which would
introduce a form of single-blinding to the study design. However, none of the studies reported
who exactly prepared and filled the cavities and who exactly reviewed and assessed the
completely filled extracted teeth.

Sample size:

Despite its eminent importance to study designs, none of the selected studies reported any
power calculations or sample size calculations. Each study recruited a different number of
extracted teeth in their experiments, but none of the authors produced any justifications or
rationale for their sample sizes. As a result, it is difficult to formulate any relevance to the results
over the targeted populations. Table 5 shows the sample sizes in the selected studies.

Statistical tests:
 ACCEPTED MANUSCRIPT

All selected studies reported the statistical tests used in analyzing their data. The studies
utilized wide range of statistical test that are appropriate for their study designs. The list of
statistical tests used in the selected studies is shown in Table 6.
It was also observed that the authors ensured that all p-values were included with clear
indications to any statistically significant differences detected among the different groups in the
studies.

Statistical analysis:

Table 7 summarizes the results for all meta-analysis conducted.

Comparison between SE & ER bond strengths at initial time (24 hours water storage):
WMD value for µTBs is -6.59; SE group had lower µTBs than ER group, significant differences
were not reached but were close to it (p-value = 0.051).

Comparison between SE & ER at one month storage time:

Meta analysis was not possible since only one study compared between SE & ER at that time
interval.

Comparison between SE &ER at three months storage time:

WMD value for µTBs is -0.99; SE group had lower µTBs than ER group. There were no
significant differences (p=0.756); because that difference is too slight, eventually zero.

Comparison between SE & ER after six months storage time:

WMD value µTBs was -5.7; SE group had lower µTBs than ER group. There were no significant
differences reached (p=0.267).

Comparison between SE & ER after 12 months storage time:

WMD value for µTBs was -2.02; SE group had lower µTBs than ER group. There were no
significant differences were reached (p=0.785).

No meta-analysis was conducted at 15 months and 4 years because only one study provided
comparisons at those times.

Meta analysis was also conducted to compare between different subtypes of SE and ER
groups, where statistically significant differences were detected when comparison was made
between 1SE and ER bonds, and that was only at initial time (24 hours storage), (p=0.003).
Comparing between 2SE & ER bonds didn’t reveal any significant differences at any
examination time.
 ACCEPTED MANUSCRIPT

Other comparisons were made between SE & 2ER and between SE & 3ER, and all revealed no
significant differences between the measured mean µTBs at any time of the examination.
Significant differences were almost reached (p=0.087) when comparison was made between
SE & 3ER at six month storage time 3ER bonding system provides a mean strength 8.67 Mpa
higher than SE.
 ACCEPTED MANUSCRIPT

Discussion

The bonding of composite restorative materials to tooth structure has been an active area of
research over the last three decades. This can be considered as a logical consequence of early
concerns raised regarding the bonding strength of composite fillings and their longevity,
especially in comparison to the then commonly used amalgam fillings. Studies have found that
the bonding strength of composite restorative materials depends on a number of factors, but
primarily dependent on the bonding mode of their bonding systems. Researchers and
manufacturers alike have persistently attempted throughout the years to improve the micro-
mechanical bonding systems of composite fillings by producing successive generations of
bonding systems.

In literature, there have been two main types of bonding systems to date; Self-etch and Etch-
and-rinse bonding systems .Each bonding system has been found to be characterized by a
number of advantages and disadvantages that have direct subsequent clinical implications. The
aforementioned situation may increase the bonding options for clinicians, but it may also
conversely complicate their evidence-based decision making, especially that there has not been
any reliable comparative reviews, to date, that systematically collate and compare findings from
all relevant publications.

This systematic review aims to investigate the longevity of bonding systems by identifying any
possible significant differences in bonding strength over time in literature. Water storage is an
easy method for determining bond durability when compared to thermocycling or mechanical
loading (20). The tooth bond interface is directly subjected to hydrolysis and resin is subjected
to water uptake and subsequent plasticization which is a major factor in bond degradation.
The effect of direct water storage on dentin bond interface is significantly noticed when the
specimens were trimmed to expose dentin resin interface rather than trimming and exposing
tooth bond interface after end of water storage. According to Hashimoto et al, dentin-resin bond
strength of all tested adhesive systems declined at a greater rate when fully exposed to water
as trimmed specimens, the diffusional degradative effect of water is demonstrated. The
unproven clinical assumption is that a similar process occurs on exposed dentin cavity margins.
(2)
Water causes hydrolysis, plasticization and subsequent degradation of dentin-resin interface.
Saliva enzymes also play a role in degradation. Endogenous enzymes such as
Metalloproteinases attack the collagen fibrils in hybrid layer of bond adds to further degradation
and hydrolysis of dentin resin interface. Bacterial enzymes also add to overall degradation of
hybrid layer.
Many factors are thought and proved to play role in degradation of dentin-resin bond and water
is thought to be a major one.
The included studies involved testing blocks made of bonded composite restorations to sound
tooth structure. Although in clinical situations bonding is performed on carious or sclerosed tooth
 ACCEPTED MANUSCRIPT

structure, sound teeth are still widely used substrates in in-vitro studies performed for bond
testing. The reasons for this are not known for the authors of this review. One possible reason
could be due to technical difficulties in preparing bonded restorations on slices of carious teeth.

Dentin bond strength is also variable according to location in tooth whether it is cervical or
occlusal, and this is because of the variability of Dentin tissue itself. When occlusal superficial
Dentin is involved, less Dentinal tubules are present in square millimeter, with narrow diameter,
which make the bond primarily intertubular, while in deeper layers, wider and larger numbers of
dentinal tubules are encountered which make bonding primarily intratubular.

Coronal dentin was prepared mainly by trimming the occlusal enamel layer or creating a mid
coronal dentin block, or even classical class 1 cavity preparation (5). Deep dentin specimens
were also prepared in one study (9). As a result, we can’t apply the results of this study to
certain aspect of the tooth (for example cervical, occlusal, etc…) as tested blocks were
prepared from different surfaces and from different teeth levels.

In this review, micro-tensile strength testing was considered as the key indicator for the
strength of the material. Tensile strength is the maximum value of force a material can withstand
when pulled perpendicularly from the substrate surface. In literature, there are numerous
strength testing methods. However, micro-tensile strength testing provides more reliable results
as it provides less mixed and cohesive failures if compared to other methods (less than 20%).
Subsequently, this results in higher values for bond strength if compared to macro-tensile
strength testing (17,7).

Only 9 publications have been included in this review as they met the structured
inclusive/exclusive criteria. It was noticeable that the studies would be of low level of evidence
as they were all in-vitro studies. In fact, the studies were not only considered low in the
hierarchy of evidence, but they did also not score high summary scores according to the
Timmer’s Analysis Tool. Additionally, the studies lacked key statistical indicators resulting in
difficulties in formulating any reliable inferences or relevance to general population. For
example, all of the included studies did not report any sample or power calculations. As a result,
the findings and conclusions of these studies should be viewed with extreme caution.

In addition to the above deficiencies, it was noticed that there was only one study that included
all of the key time points in the experiments (i.e. De Munck, 2010). This would complicate any
systematic attempt to compare the findings of all of the included studies in this review. However,
despite of the heterogeneity among the included studies, the studies unanimously indicated that
the 1-step bonding system produced lower bonding strength if compared to 2-step or 3-step
 ACCEPTED MANUSCRIPT

bonding systems in all of the comparison time points (e.g. 24 hours, 3 months, 6 months and 12
months). However, these differences in bonding strength among the systems were not of any
statistical significance.

The in-vitro approach for medical or bio-medical research could be sometimes the only practical
approach for tests. However, in-vitro studies have been traditionally challenged by the
difficulties in accurately mimicking the biological, chemical and physical conditions in-situ. The
included studies used various incubation conditions; including the use of water, thymol, distilled
water, chloramine and simulated pulp pressure. However, it is not clear how these designs
would mimic the oral cavity and its properties. For instance, a number of studies claimed that
normal water was not characteristically similar to biological saliva. Indeed, the use of artificial
saliva have been seen in recent years as valid and reliable replica for biological saliva.
Historically, artificial saliva has been used in research since 1931 when it was used to test the
potential mercury poisoning from dental amalgam fillings (11). Leung and Darwell studied the
various formulations of artificial saliva that had been used over the years and reviewed the
effects of the various components on hydroxyapatite and assessed the stability of the
hydroxyapatite in these formulations. The study concluded that the formulation of a
standardized artificial saliva solution would improve their application in research (11).

In another example of poor quality of research in the included studies is the lack of controls in
some of the included studies. This does not only reduce the summary scores of the included
studies, but it does also introduce a substantial amount of bias to the analysis and interpretation
of the results and the findings of the experiments.

According to the results of this study, no significant difference was found between the bond
strength of SE adhesives when compared to the bond strength of ER adhesives although µTB s
values were higher for ER than SE at all testing times. All in all, the included studies showed
some evidence that 1SE bonding systems provide less bonding strength if compared to 2SE,
2ER and 3ER bonding systems. However, this evidence must be viewed with caution as the
included studies suffered from deficiencies in their design, experimentation, analysis and
interpretation which would, in turn, increase the amount of bias in the included studies.

CONCLUSION
 ACCEPTED MANUSCRIPT

1. In general, ER bond systems provide higher µTBs than SE ones.

2. At initial time (24 hours of storage) mean difference of µTBs is around 6.5 Mpa between
both categories. Strictly, significant differences were not reached but a very strong trend
is showed (p=0.051). At different time-points of examination, results become similar.

3. Differences are extreme when 1SE group is compared to ER group. µTBs with 1SE is
reduced at average in 11.34 Mpa respect to ER at the initial time. This conclusion is
statistically significant (p=0.003) and is the unique between all comparisons tested.
At 6-months it was detected a relevant trend in the same direction (p=0.061).

4. 2SE is the most homogeneous type to ER group. In fact, 2SE is the only system without
differences in µTBs any moment.

5. 2ER shows a higher µTBs than SE, especially at the initial time, but no significant
differences were obtained.

6. Similar conclusions are applied to the comparison between SE & 3ER, but the maximum
difference took place at the 6-month time-point.

7. In general, the heterogeneity of studies is remarkable. Between-studies variability seems

to be an important part of the total amount, including within-studies. But from other point
of view, most of studies are inside 95% confidence limits if heterogeneity is rejected. An
small number of studies are involved in each meta-analysis: Under these conditions
heterogeneity tests are sensible and lose statistical power.

8. No publication bias was detected. Studies with the largest standard errors not showed a
relevant asymmetry. Those with moderate values for standard errors range in a wide
interval of observed effects, but dispersion is not biased.

References

1. Abdalla, A. I. & Feilzer, A. J. Four-year water degradation of a total-etch and two self410
etching adhesives bonded to dentin. Journal of Dentistry 2008, 36 (8): 611-617.
 ACCEPTED MANUSCRIPT

2. Armstrong, S. R., Vargas, M. A., Fang, Q. & Laffoon, J. E. Microtensile bond strength of
a total-etch 3-step, total-etch 2-step, self-etch 2-step, and a self-etch 1-step dentin
bonding system through 15-month water storage. Journal of Adhesive Dentistry 2003, 5
(1): 47-56.

3. Breschi, L., Mazzoni, A., Ruggeri, A., Cadenaro, M., Di Lenarda, R. & De Stefano
Dorigo, E. Dental adhesion review: aging and stability of the bonded interface. Dental
Materials 2008, 24 (1): 90-101.

4. Carrilho, M.R.O., Carvalho, R.M., Tay, F.R., Yiu, C., Pashley, D.H. American journal of
dentistry 2005, 18:6; 315-319.

5. Chiaraputt, S., Roongrujmek, P., Sattabanasuk, V., Panish, N., Harniarattisal, C.,
Senawongse, P. Dental materials journal 2011, 30(6):814-826.

6. De Munck, J., Van Landuyt, K., Peumans, M., Poitevin, A., Lambrechts, P., Braem, M. &
Van Meerbeek, B. A critical review of the durability of adhesion to tooth tissue: methods
and results. Journal of Dental Research 2005, 84 (2): 118-132

7. Eliades G, Watts D, and Eliades T. Dental Hard Tissues and Bonding: Interfacial
Phenomena and Related Properties. Springer, 2005: 53-67.

8. Feitosa, V.P., Leme, A.A., Sauro, S., S obrinho, L.C., Watson, T.F., Sinhoreti, M.A., Correr,
A.B.. Journal of dentistry, 2012, 40(1134-1143).

9. Giannini, M., Makishi, P., Ayres, A.P.A., Vermelho, P.M., Fronza, B.M., Nikaido, T., Tagami,
J. Self-etch adhesive systems: A literature review. Brazilian dental journal 2015, 26 (1):3-10.

10. Hashimoto M, Ito S, Tay FR, Svizero NR, Sano H, Kaga M, Pashley DH. Fluid
movement across the resin-dentin interface during and after bonding. J Dent Res 2004; 83:843-
848.

11. Leung, V.W.-H, Darvell, B.W. Artificial salivas for in vitro studies of dental materials.
Journal of Dentistry 1997, 25 (6): 475-484.

12. Martins, G. C., Calixto, A. L., Gomes, O. M., Loguercio, A. D., D'Alpino, P. H. & Reis, A.
Effect of water storage on resin-dentin bond strengths formed by different bonding
approaches. Indian Journal of Dental Research 2009, 20 (4): 431-436.

13. Nakabayashi, N., Kojima, K. & Masuhara, E. The promotion of adhesion by the
infiltration of monomers into tooth substrates. Journal of Biomedical Materials Research,
1982 16 (3): 265-273.

14. Osorio, R., Pisani-Pronca, J., Erhardt, M.C.G., Osorio, E., Auguilera, F.S., Tay, F.R.,
Toledano, M. Journal of dentistry 2008 , 36:2; 163-169.

15. Pashley, D. H. Dynamics of the pulpo-dentin complex. Critical Reviews in Oral Biology &
Medicine 1996, 7 (2): 104-133.

16. Ricardo M. Carvalho. Durability of bonds and clinical success of adhesive restorations.
Dent Mater 2012, Jan;28 (1) . doi:10.1016/j.dental.2011.09.011.
 ACCEPTED MANUSCRIPT

Study title Authors Journal

17. Sakaguchi R and Powers J .Craig's restorative dental materials. 440

Elsevier/Mosby, 13th edition 2012

18. Timmer A, Sutherland L and Hilsden R. Development and evaluation of a quality score
for abstracts. BMC Medical Research Methodology 2003, 3:2

19. Toledano M, Osorio R, de Leonardi G, Rosales-Leal JI, Ceballos L,

Cabrerizo-Vilchez MA. Influence of self-etching primer on the resin adhesion
to enamel and dentin. Am J Dent 2001;14:205-210.

20. Van Landuyt, K. L., De Munck, J., Mine, A., Cardoso, M. V., Peumans, M. & Van
Meerbeek, B. Filler debonding & subhybrid-layer failures in self-etch adhesives. Journal
of Dental Research 2010, 89 (10): 1045-1050.

Table 1: the selected full-text articles included in the meta analysis:

 ACCEPTED MANUSCRIPT

Filler debonding & subhybrid –layer failure in Van Landuyt Journal of Dental
self-etch adhesives.(20) KL. et al Research.89(10):1045-50, 2010
Oct.

Microtensile bond strength of a total-etch 3- Armstrong Journal of Adhesive Dentistry.

step, total-etch 2-steps, and a self-etch 1- SR. et al 5(1):47-56,2003
step dentin bonding system through 15-
month water storage.(2)

Biodegradation of all-in-one self-etch Chiaraputt Dental materials journal.

adhesive systems at the resin-dentin Sirichan, et 30(6):814-826. 2011 NOV
interface.(5) al

Enzymatic degradation of adhesive-dentin De Munak European journal of oral

interfaces produced by mild self-etch Jan, et al sciences. 118(5):494-501. 2010
adhesives.(6) Oct.

Four-year water degradation of a total-etch Abdalla Ali, Journal of dentistery.36 :( 8);

and two self-etching adhesives bonded to et al 611-617. 2008 Aug
dentin.(1)

Resistance of ten contemporary adhesives to Osario Journal of dentistry. 36:2; 163-

resin-dentine bond degradation.(14) Raquel, et al 169. 2008 Feb

Effect of water storage on resin-dentin bond Martins, G C, Indian journal of dental

strengths formed by different bonding et al research. 20; 4; 431-436.2009.
approaches.(12) Oct-Dec.

Durability of resin-dentin bonds related to Carilho American journal of dentistry.

water and oil storage.(4) MRO, et al. 18:6; 315-319. Dec 2005.

Hydrolytic degradation of the resin-dentine Victor P. Journal of dentistry. 40(1134-

interface induced by the simulated pulpal Feitosa. 1143. 2012
pressure, direct and indirect water aging.(8)

Table No. 2: µTBs( measured in MPs), standard deviations (SD) for SE and ER at different aging times:
 At 24 hours At 3 months At 6 months
SE ER
Auth N M S N M S N M S N M S N M S N M S S
or ea D eaACCEPTED
D ea MANUSCRIPT
D ea D ea D ea D D
n n n n n n
Van 6 29. 8. 1 43. 8. 7 15. 6 1 37. 1
Laund 2 8 1 2 4 7 9 3 6 7 7.
uyt 2
KL. Et
al
Armst 2 43. 8. 4 43 1
rong, 2 9 3 2 1.
2002 6
Chiar 4 39. 1 1 49. 1 4 34. 1 1 37 1

PT
aputt, 5 4 1. 5 2 5. 5 9 5. 5 2
2011 7 1 6
De 2 31. 1 1 41. 1 2 34. 1 8 32. 1 3 21. 1 1 23. 2

RI
Munk 8 8 5. 6 1 2. 0 1 2. 7 4. 1 1 0. 4 4 5.
, 3 2 1 2 1 5
2010

SC
Abdal 4 39 3. 2 39 5.
la, 0 6 0 2
2008
Osori 2 23. 7. 6 39. 6. 2 19. 7. 6 39 9

U
o, 4 4 9 0 7 3 4 2 8 0
2008 0 0
AN
Marti 6 27. 7. 6 38. 1 6 26 8. 6 34. 8.
ns, 0 8 9 0 5 0. 0 7 0 5 7
2009 4
M

Carril 1 44. 7. 3 33. 1 1 31. 5. 2 21. 4.

do, 4 4 8 0 2 0. 4 9 5 7 1 5
2014 4
D

Feitos 1 43 3. 5 47. 5. 1 36. 4 5 35. 3.

a, 5 2 2 7 0 3 3 9
2012
TE
EP

At 12 months
SE ER
Author N Mean SD N Mean SD
Van
C

Launduyt
KL. Et al
AC

Armstrong
G, 2002
Chiaraputt,
2011
De Munk, 29 18 11.5 18 20.7 18
2010
Abdalla,
2008
Osorio, 150 16.1 7.5 60 34.4 9.9
2008
Martins,
 ACCEPTED MANUSCRIPT
2009
Carrilho, 20 20.7 6.2 35 11.3 3.9
2014

PT
RI
U SC
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT

Table No. 3: Quality assessment of the selected papers:

Van Landuyt et al 2010

Armstrong et al. 2003

Chiaraputt et al 2011

De Munak et al 2010

Feitosa et al. 2012

Carilho et al. 2005
Abdalla et al 2008

Martins et al 2009
Osario et al 2008

Question / objective sufficiently

described

Design evident and appropriate

to answer study question

Subject characteristics
sufficiently described

Subjects appropriate to the

study question
 ACCEPTED MANUSCRIPT

Controls used and appropriate

Method of subject selection

described and appropriate

Random allocation to treatment

groups described

Blinding of investigators to
intervention reported

Blinding of subjects to
intervention reported

Outcome measure well defined

and robust to measurement bias

Confounding accounted for

Sample size adequate

Post hoc power calculations for

statistically non-significant
results

Statistical analyses appropriate

Statistical tests stated

Exact p-values or confidence

intervals stated

Attrition of subjects and reason

for attrition recorded
 ACCEPTED MANUSCRIPT

Results reported in sufficient

detail

The results support the

conclusions

Table 4: Summary scores of selected studies

Study Points scored Summary

scores

Van Landuyt et al 2010 20 20/34 = 0.58

Armstrong et al. 2003 20 20/32 = 0.62

Chiaraputt et al 2011 20 20/32 = 0.62

De Munak et al 2010 22 22/34 = 0.65

Abdalla et al 2008 20 20/32 = 0.62

Osario et al 2008 20 20/32 = 0.62

Martins et al 2009 22 20/32 = 0.68

Carilho et al. 2005 20 20/32 = 0.62

Feitosa et al. 2012 22 22/34 = 0.65

 ACCEPTED MANUSCRIPT

Table 5: Selected studies and their sample sizes

Study Sample size

Van Landuyt et al 2010 153

Armstrong et al. 2003 44

Chiaraputt et al 2011 32

De Munak et al 2010 45

Abdalla et al 2008 54

Osario et al 2008 60

Martins et al 2009 24

Carilho et al. 2005 25

Feitosa et al. 2012 105

 ACCEPTED MANUSCRIPT

Table 6: Statistical tests used in selected studies

Study Statistical tests

Van Landuyt et al 2010 Kruskal-Wallis test

Armstrong et al. 2003 Log-rank & Wilcoxon tests

Chi square test

Fischer exact test

Chiaraputt et al 2011 Kolmogorov-Smirnov test

De Munak et al 2010 ANOVA test

Tukey–Kramer multiple
comparison tests.

Abdalla et al 2008 ANOVA test

Student-Newman-Keuls
 ACCEPTED MANUSCRIPT

multiple comparisons tests

Osario et al 2008 ANOVA test

Student-Newman-Keuls
multiple comparisons tests

Martins et al 2009 ANOVA test

Tukey’s post-hoc test

Carilho et al. 2005 ANOVA test

Tukey’s test

Feitosa et al. 2012 ANOVA test

Tukey’s test
 ACCEPTED MANUSCRIPT

Table 7: Meta analysis results comparing between SE& ER at related time lines:

24 h 3-months 6-months 12-months

N WMD p N WMD p N WMD p N WMD p

SE vs. ER 8 -6.59 0.051 2 -0.99 0.756 7 -5.70 0.267 4 -2.02 0.785

1SE vs. ER 8 -11.34 0.003** 2 -11.37 0.149 6 -12.11 0.061 4 -7.56 0.365

2SE vs. ER 7 -2.53 0.487 2 12.13 0.243 6 -2.05 0.667 3 6.50 0.194

SE vs. 2ER 7 -7.11 0.095 2 -0.99 0.756 6 -3.29 0.591 4 -2.02 0.785

SE vs. 3ER 2 -6.67 0.326 -- -- -- 3 -8.67 0.087 -- -- --

N: Number of studies
WMD: Weighted mean difference;
p: p-value for nullity test of WMD
*p<0,05; **p<0,01; ***p<0,001