To evaluate the variability of bond strength test results of adhesive systems (AS) and to correlate the results with clinical parameters of clinical studies investigating cervical restorations.
Materials and methods
Regarding the clinical studies, the internal database which had previously been used for a meta-analysis on cervical restorations was updated with clinical studies published between 2008 and 2012 by searching the PubMed and SCOPUS databases. PubMed and the International Association for Dental Research abstracts online were searched for laboratory studies on microtensile, macrotensile and macroshear bond strength tests. The inclusion criteria were (1) dentin, (2) testing of at least four adhesive systems, (3) same diameter of composite and (4) 24 h of water storage prior to testing. The clinical outcome variables were retention loss, marginal discoloration, detectable margins, and a clinical index comprising the three parameters by weighing them. Linear mixed models which included a random study effect were calculated for both, the laboratory and the clinical studies. The variability was assessed by calculating a ratio of variances, dividing the variance among the estimated bonding effects obtained in the linear mixed models by the sum of all variance components estimated in these models.
Thirty-two laboratory studies fulfilled the inclusion criteria comprising 183 experiments. Of those, 86 used the microtensile test evaluating 22 adhesive systems (AS). Twenty-seven used the macrotensile test with 17 AS, and 70 used the macroshear test with 24 AS. For 28 AS the results from clinical studies were available. Microtensile and macrotensile (Spearman rho = 0.66, p = 0.007) were moderately correlated and also microtensile and macroshear (Spearman rho = 0.51, p = 0.03) but not macroshear and macrotensile (Spearman rho = 0.34, p = 0.22). The effect of the adhesive system was significant for microtensile and macroshear ( p < 0.001) but not for macrotensile. The effect of the adhesive system could explain 36% of the variability of the microtensile test, 27% of the macrotensile and 33% of the macroshear test. For the clinical trials, about 49% of the variability of retained restorations could be explained by the adhesive system. With respect to the correlation between bond strength tests and clinical parameters, only a moderate correlation between micro- and macrotensile test results and marginal discoloration was demonstrated. However, no correlation between these tests and a retention loss or marginal integrity was shown. The correlation improved when more studies were included compared to assessing only one study.
The high variability of bond strength test results highlights the need to establish individual acceptance levels for a given test institute. The weak correlation of bond-strength test results with clinical parameters leads to the conclusion that one should not rely solely on bond strength tests to predict the clinical performance of an adhesive system but one should conduct other laboratory tests like tests on the marginal adaptation of fillings in extracted teeth and the retention loss of restorations in non-retentive cavities after artificial aging.
In restorative dentistry, the largest area exposed after preparation is in most cases dentin. Therefore, bond strength on dentin is decisive for the restoration to be held in place. This is especially critical for those cavities/preparations with no or little mechanical retention like cervical restorations, crown stumps with reduced stump height and/or a high angle of convergence and overlay preparations. The sealing of the dentinal tubules is another important function of adhesive systems.
The effectiveness of an adhesive system to bond to dentin is commonly tested with a bond strength test. The first article on bond strength tests (tensile bond strength test) for dental materials was published in 1965 by Bowen . Since then, many more articles have been published. Today, 4960 articles are listed in PubMed when searching for “bond strength” and “dental”, 2695 articles for “bond strength” and “dentin” as well as 1545 articles with the search terms “bond strength” and “enamel” (search period 1955–2012, search month November 2012). A further 3716 abstracts were retrieved from International Association for Dental Research (IADR) abstracts online ( www.iadr.org , 2002–2012, search month November 2012, search term “bond strength” and “dentin”). These articles and abstracts advocate various test setups, such as the shear bond, microtensile, microshear, push-out and the fracture toughness test. There is only a small degree of standardization. Of the 4960 articles, only 12 relevant reviews critically evaluated the different bond strength tests in terms of their strengths and weaknesses . The remaining large number of studies on bond strength testing in dentistry investigated various modifications of test setups and substrates. Additionally, most of them focused on testing specific materials.
Only after 30 years of bond strength testing, efforts were made to relate the results of these tests to clinical findings and to look at the variability of test results. Even now, only few publications exist which correlate the in vitro bond strength data with the clinical outcome of the tested adhesive systems. The clinical model most often used to test the effectiveness of adhesive systems is the restoration of non-carious cervical defects. Such defects are especially suitable to test adhesive systems due to the following: (1) practically no macro-mechanical retention is present, (2) straightforward clinical placement of the restoration and evaluation of debonding, reducing operator and evaluator variability, and (3) high prevalence, which makes patient selection simple and allows for properly designed studies. Only three publications on the correlation between bond strength tests and the clinical performance of restorations placed with adhesive systems have been published so far. In one of these studies , the microtensile bond strength data of 15 adhesive/restorative systems placed by the same operator were correlated with the clinical studies of non-carious cervical Class V restorations. No correlation was found between the retention rate of cervical restorations after 3 years and the microtensile test results after 8 h or 6 months of water storage of specimens prior to testing. There was, however, a very moderate correlation between marginal staining and bond strength values after 6 months of water storage. A comprehensive database of microtensile bond strength data and an equally comprehensive database on the retention rates of clinical restorations placed in non-carious wedge-shaped defects at the same test institute (University of Leuven) found a moderate correlation between bond strength of artificially aged specimens and clinical retention. The correlation was higher for the 5-year data than for the 2-year data . However, most of product-related individual data – both clinical and laboratory – have not been published by that research group. Another attempt was made to correlate the bond strength data with the retention of cervical restorations. In 2010, Scherrer et al. published data of laboratory studies on six dentin adhesive systems, available in the literature, and four laboratory methods (macroshear, microshear, macrotensile and microtensile bond strength test). The review revealed a large variability for the same adhesive system evaluated with the same bond strength method, not only at different test institutes (inter-institute variability) but also at the same test institute (intra-institute variability). The variability was similar for each test method. Scherrer and colleagues pooled the data across the different studies in relation to the adhesive system and the bond strength test and calculated mean values and standard deviations. These data were correlated with estimated pooled 2-year retention rates of Class V restorations using the same adhesive systems and retrieved from the databank of the meta-analysis on cervical restorations . The results of the regression analysis for the pooled data demonstrated that only the macrotensile and microtensile tests but not the shear and the microshear tests correlated more accurately with the retention rate of cervical restorations. This finding allowed for two hypotheses: (1) tensile tests correlate with the retention loss of cervical restorations whereas the shear tests do not. (2) Pooled data across different institutes may correlate more accurately with retention loss of cervical fillings than individual data from one test institute. The explanation may be that these pooled data characterize the variability and efficacy of a certain adhesive system more appropriately.
These assumptions, however, are only based on six adhesive systems and the selection criteria of the laboratory studies were not very restrictive and included studies with one adhesive system only, with specimens submitted to thermocycling, different bonding areas, etc.
The aims of the present meta-analysis were threefold:
To investigate the variability of bond strength tests on dentin by retrieving literature data on the three most commonly used tests: microtensile, macrotensile and macroshear. Furthermore, the correlation between these three tests should be assessed.
To investigate the correlation between these tests and clinical parameters from studies on cervical restorations (retention, marginal discoloration, marginal integrity).
To investigate whether the correlation improves if pooled data from different studies are used instead of data from single studies.
The following null hypotheses were formulated:
There is no correlation between the three test methods.
There is no correlation between bond strength tests and the clinical outcome of cervical restorations.
The pooling of data does not improve the correlation.
Materials and methods
Selection of in vivo studies
For the clinical performance of the different adhesive systems, results from prospective studies on cervical restorations (Class V) were retrieved from literature. The internal database used for the meta-analysis on Class V restorations and published in Dental Materials in 2010 was updated with studies that were published between 2008 and 2012, using the same criteria as described in the publication, the same databases (PubMed and SCOPUS) and the same search words. The search period ranged from 2008 to 2012 and the search month was July 2012. However, for the in vitro/in vivo comparison only those studies with specific adhesive systems were used for which also in vitro data were available.
Selection of in vitro studies
For in vitro studies on bond strength, only studies that tested at least four different adhesive systems were included. The rationale was (1) that a reasonable ranking of adhesive systems in relation to the study should be possible, and (2) to simplify the literature search. However, there was no restriction with regard to the publication year. Other inclusion criteria were as follows:
Same diameter (macroshear, macrotensile) or diagonal (microtensile) of test specimen in relation to test method.
24 h of storage in water before conduction of bond strength test.
The databases PubMed (1955–2012) and IADR abstracts online (2002–2012) were searched using the following search terms: “bond strength” and “dentin”. The search month was July 2012. The search focused on the microtensile (Test 1), macrotensile (Test 2) and macroshear tests (Test 3).
The following data were retrieved from the studies:
Adhesive system (AS).
Mean bond strength value (MPa).
Number of specimens and/or teeth used per group.
Modeling of in vivo performance
The clinical performance was measured by means of the percentage of retention loss (R), the percentage of marginal discoloration (MD) and the percentage of detectable margins (MI). Although the percentage of secondary caries or caries at the restorative margins was also measured, it was not considered in this analysis as most experiments had 0% of marginal caries. In correspondence with the previous study by Heintze et al. , a clinical index defined as CI = (4R + 2MD + MI)/7 was calculated in order to summarize the clinical performance by weighing the three most commonly reported clinical parameters and giving higher importance to retention than to marginal discoloration and marginal integrity. In what follows, all these percentage values are expressed such that they are equal to 100% at baseline and decreasing afterwards. They were assessed after 0.5, 1, 1.5, 2, 3, 4, 5, 6, 7, 8 and 13 years (depending on the studies). Since measurement results after 13 years were only available for 3 experiments (from the same study), and since there was a gap of 5 years between 8 and 13 years, we restricted our attention in what follows to the first 8 years of follow-up.
Since a linear deterioration over time may imply a percentage below 0, which is by definition not possible, we are considering the following model:
Y 100 = exp ( − λ × T α × error )
In this model, the percentage Y also decreases over time, but not linearly so that the percentage value remains above 0 (as long as λ is positive). This model is equivalent to the following linear model:
log ( − log ( Y / 100 ) ) = β + α × log ( T ) + error
with β = log( λ ).
In this model, the parameter β summarizes the deterioration occurring in an experiment. The deterioration depends on the fixed characteristics of the experiment, i.e. the factors adhesive, preparation (no/yes/missing), beveling (no/yes/missing) and rubber dam (no/yes/missing). To account for the fact that partly the same patients were assessed in the different experiments of the same study, a random study effect was included in the model. Furthermore, to answer the correlations between the different measurements made in the same experiment, a random experiment effect was included. This results in a linear mixed effect model with the following variable value:
β = reference value + effect of adhesive system + preparation effect + beveling effect + rubber dam effect + study random effect + experiment random effect
The two random effects as well as the error term were assumed to be normally distributed. The reference value refers to the adhesive system No 2 (AdheSE) without preparation, without beveling and without rubber dam (in an average experiment from an average study). This reference value is thus a summary measure of the deterioration, i.e. of the in vitro performance of this adhesive. To get a summary measure of the in vitro performance of the other adhesives, the coefficients corresponding to the different adhesives estimated in our model may be added to this reference value. To fit a linear mixed effect model, we used the lme routine from the package nlme , implemented in the free statistical package R . Using this routine, it was possible to weigh a percentage Y according to the denominator used for its calculation, i.e. the number of subjects available at a given point in time. Thus, the percentages calculated from many subjects received a higher weight than the percentages calculated from few patients.
Modeling in vitro performance
Similar to the in vivo experiments, we summarized the in vitro performance of the different adhesives. To do so, we used the coefficients corresponding to the effect of the adhesive system in a linear mixed model for the average performances calculated in the different experiments, a fixed effect of the adhesive system and a random study effect. Thus, we modeled the measurements Y , provided by the test methods, using the linear model: Y = β + error where: β = reference value + effect of the adhesive system + study random effect. Since the measurements provided by the three test methods (in N/mm 2 ) do not share the same range (being much higher for Test 1 than for Test 2 and Test 3), this was done separately for each of the three test methods.
Correlation between clinical and in vitro performance
Using the coefficients corresponding to the effect of the adhesive system estimated in a linear mixed model as explained in the last two sections, it was possible to examine the correlation between the clinical and the in vitro performance of the different adhesives. Spearman correlations rho were calculated between the four measures of clinical performance (R, MD, MI, CI) and the three measures of in vitro performance.
Comparison of variability
It is not an obvious procedure to compare the variability of the estimates of clinical and in vitro performances as they are measured in different units. The in vitro performance is measured in Megapascal, whereas we summarized the clinical performance by means of a slope of deterioration for a percentage of clinical outcome parameters observed in a set of patients over time. To make such a comparison feasible, one possibility would be to calculate a ratio of variances, dividing the variance among the estimated effect of the adhesive systems obtained in our linear mixed models by the sum of all variance components estimated in these models.
For the in vitro performance the formula would be as follows:
percentage of variance due to adhesive = variance ( effect of adhesive system ) variance ( effect of adhesive system ) + variance ( study random effect ) + variance ( error ) .
For in vivo performance, considering experiments with 50 patients, the formula would be
percentage of variance due to adhesive = variance ( effect of adhesive system ) ( variance ( effect of adhesive system ) + variance ( study random effect ) + variance ( experiment random effect ) + variance ( error ) / 50 ) .