Methodological considerations on push-out tests in Endodontics

We would like to take this opportunity to comment on the Letter to the Editor that covered various issues relating to the use of push-out tests in endodontic research (Moinzadeh et al. 2014). First of all, we must stress that we are in agreement with the conclusion at the end of the Letter, that is: ‘Laboratory studies should be critically assessed from a methodological point of view, and efforts should be made to improve the current ex vivo models’, as we also believe that attempts are necessary to improve push-out experimental methods in Endodontics. However, we wish to address the main methodological aspects raised by the authors to improve push-out tests, in order to avoid misconceptions that may become mainstream in future research in this area.

1. Sliding friction, rather than true bond strength, largely contributes to dislocation resistance

This statement has been generally accepted in the literature and we agree with the authors over this point. However, from a materials science perspective, a clear misconception is to think the push-out tests in Endodontics as a reliable output for measuring true adhesion of filling materials to root canal walls, or even as ‘a potential surrogate marker for endodontic outcomes’. Adhesion is well known to be a complex process and is not likely to occur effectively within the root canal space. Consequently, it would be more important (i) to understand how different filling materials are prone to resist dislocation forces that might affect the materials inside the root canal and (ii) to be able to reliably rank the quality of filling materials and techniques. In short, push-out studies should be regarded simply as ordinary ‘bench’ tests that should be non-labour-intensive. A close-to-ideal laboratory test shall be used as a preliminary screening, where financial, ethical and practical issues cannot limit their application. Thus, these tests cannot be related to true ‘endodontic outcomes’ or have ‘clinical significance’.

       2. The geometry of the root canal in the thin-slice should be diverging in the forward direction of the applied load in order to reduce the contribution of frictional sliding to the dislocation resistance

In fact, this is not a new proposal and it has been used already in several studies. As pointed out by Moinzadeh and colleagues, the Poisson effect may increase the retention of materials due to their cross- sectional deformation induced by push-out forces. Thus, when using parallel-sided cavities, the Poisson effect may increase actual sliding friction and interfere with the final results. However, it should be stressed that good standardization of the cavity is certainly of greater importance than the cavity configuration itself. This standardization is difficult to create when using real root canals prepared with currently available instruments/burs for root canal preparation, as discussed below. In fact, our concern over the creation of a reliable anatomical baseline (well-standardized anatomical conditions) is indeed more relevant as it will enhance the internal validity of comparative assessments. It is also important to point out that the real impact of the Poisson effect on the results of push-out tests when using 1-mm-thin-sliced specimens remains unknown for testing root filling materials with two interfaces (core material and sealer).

       3. When testing different materials, differences in elastic modulus should be avoided or at least reported

From a theoretical standpoint, this is a reasonable measure that will improve the overall quality of research, but again it is not new (Chen et al. 2013). It is unknown (and maybe unlikely) that the elastic modulus of a sealer used in such a small volume/ thickness, when combined with a core material, would significantly affect the outcome of the push-out test; however, the ideal should be experimentally evaluate the actual role of the elastic modulus of the root filling materials in the final push-out results.

       4. Push-out specimens should be sliced after the application and setting of the tested material, in order for results to demonstrate a realistic influence of the C-factor.

Currently, the negative influence of the C-factor on the dislocation resistance of root filling materials is almost always restricted to methacrylate resin-based sealers. When considering the gold-standard class of root canal sealers, such as the epoxy resin-based as AH Plus (De Trey Dentsply, Konstanz, Germany), the C-factor has only a negligible effect on the overall test outcome (Kim et al. 2010). So, it is also reasonable to assume that physico-chemical properties other than the elastic modulus, such as dimensional change and the degree of polymerization, are also involved in the performance of the sealer within the root canal space when submitted to the push-out test. It is important to note, however, that root fillings consist of a core material plus a sealer; thus, there are at least two interfaces that can be ‘pushed’ out from the canal space, and this could create a systematic source of error. Hence, studies that focused only on the properties of the sealer, have filled the canal space only with the sealer to achieve better control on the mode of failure. In this way, bias related to the classification of the failure mode is avoided, as all failures in this model will have an undoubtedly adhesive nature, which also truly reflects the bond strength between the sealer and dentine (Neelakantan et al. 2011).

       5. The preparation and use of artificial cavities in dentine regions not corresponding to the prepared root canal region

From a material science standpoint, the creation of artificial dentine cavities has several advantages that the authors did not take into consideration. At first glance and as stated by Moinzadeh and colleagues, reproduction of the clinical situation is considered as a major advantage when using the intracanal dentine walls of extracted human teeth in push-out tests. However, the considerable variation in root canal morphology makes anatomical standardization challenging and thus creating balanced experimental and control groups a difficult task (Hülsmann et al. 2005). Considering that canals have dissimilar cross-sectional shapes at different levels of the root throughout the same tooth, sample selection based only on the use of single-rooted teeth results in poor standardization and means the experimental model becomes more divorced from a sound condition for experimental comparability. Therefore, one of the main limitations of push-out tests is the difficulty of creating a reliable baseline as the intricate anatomy and the variable morphology of the substrate (dentine) are confounding factors (De-Deus 2012). It also worth-mentioning that the presence of calcospherites-containing zones along the canal wall may increase sealer retention in the uninstrumented areas of the root canal, which could also have a random influence over push-out assays (Huffman et al. 2009). In addition, push-out studies using natural teeth have other potential confounders such as tooth age, storage time, the amount and distribution of sclerotic dentine, dentine microhardness and modulus of elasticity, which must account for some of the differences in outcomes between studies. In summary, the use of randomized small-sized experimental groups of root filled teeth is unable to overcome the important effects of the intrinsic biological–chemical–physical variances of root dentine and canal shape. When extracted human teeth are used for this type of assessment, the potential for variations is strong and can explain the large standard deviations reported in relation to the mean values found in some studies. Thus, differences in sliding friction or elastic modulus of the material that are considered relevant by the authors are somewhat less relevant if the baseline substrate is not well standardized. On the other hand, the use of well-standardized geometrical test specimens (artificial canal spaces) is an attempt to overcome the individual anatomic differences of natural canal spaces, which generally make push-out test outcomes impossible to compare. The employment of artificial canal spaces allows the creation of balanced experimental groups and also similar cleaning and shaping intracanal conditions, which is almost impossible to obtain when randomized small-sized experimental groups of root filled teeth are used.

       6. Tested surfaces should be prepared according to procedures corresponding to endodontic protocols

This recommendation is usually based on the philosophical rationale prevalent in the endodontic scientific field, which claims that laboratory studies should follow the real-life clinical treatment conditions in order to be considered valid. However, being ‘usual’ does not necessary mean being ‘correct’ from a scientific perspective. In reality, translating the results from laboratory studies to the clinical setting is not always simple or even possible. Reliable scientific understanding of a clinical treatment, material or technique has to start somewhere. Hence, a well-known evidence-based approach prerogative is that the scientific ‘starting point’ should be always laboratory studies as they are able to create a basic understanding that is rapid and safe, has minimal cost and overcomes many ethical concerns. Moreover, laboratory studies can be designed so as to control confounding variables and thus reliably isolating the variable of interest. The consequence is that the results from methodological sound laboratory studies can be reliable, comparable and reproducible. Laboratory studies have the advantage of using experimental designs that are able to maximize their internal validity, and this should be the focus of the improvements suggested for the push-out experimental models in Endodontics. Thus, it is not necessary to prepare test surfaces according to clinical procedures, as the main purpose is limited to create standardized samples of the dentine substrate, which is difficult to be created using procedures corresponding to real-life endodontic protocols. In other words, it is not possible to achieve ‘a correct shaping’ and ‘conditioning of the cavity walls’ using standard clinical endodontic instruments, which usually results in unpredictable final results due to the variable anatomical configuration of root canals.

Following the  same  rationale,  push-out results obtained using conventional thin slices of natural teeth cannot also be ‘extrapolated to what occurs inside an entire canal’. It is important to bear in mind that an ideal push-out model can only rank the filling materials/techniques, rather than being guidelines for clinical decision-making. Thus, standard rules and principles for conventional ‘bench tests’ should drive the research design of push-out experimental models for endodontic purposes in order to create reliable and comparable conditions and achieve reproducible results. Hence, it is important to highlight that the conclusion ‘push-out test is a valuable test’ is not supported by the rationale of the letter submitted by Moinzadeh and co-workers or any experimental research, thus far. Rather, it may be a prudent and wiser choice to bear in mind that push-out tests may contribute to the understanding of the properties of specific filling materials and their relationship to root dentine, but nothing more than that.

Authors: G. De-Deus, E. Souza, M. Versiani

References:

1. Chen WP, Chen YY, Huang SH, Lin CP (2013) Limitations of push-out test in bond strength measurement. Journal of Endodontics 39, 283–7.
2. De-Deus G (2012) Research that matters – root canal filling and leakage studies. International Endodontic Journal 45, 1063–4.
3. Huffman BP, Mai S, Pinna L, Weller RN, Primus CM, Gutmann JL (2009) Dislocation resistance of ProRoot Endo Sealer, a calcium silicate-based root canal sealer, from radicular dentine. International Endodontic Journal 42, 34–46.
4. Hülsmann M, Peters OA, Dummer PMH (2005) Mechanical preparation of root canals: shaping goals, techniques and means. Endodontic Topics 10, 30–76.
5. Kim YK, Grandini S, Ames JM et al. (2010) Critical review on methacrylate resin-based root canal sealers. Journal of Endodontics 36, 383–99.
6. Moinzadeh AT, Jongsma L, Wesselink PR (2014) Considerations about the use of the “push-out” test in Endodontic research. International Endodontic Journal doi: 10.1111/ iej.12416. [Epub ahead of print].
7. Neelakantan P, Subbarao C, Subbarao CV, De-Deus G, Zehnder M (2011) The impact of root dentine conditioning on sealing ability and push-out bond strength of an epoxy resin root canal sealer. International Endodontic Journal 44, 491–8.