While endodontic treatment by itself increases the risk of root fracture , immature endodontically treated teeth are at an even higher risk of root fracture because of their thin dentinal walls [4, 5]. The situation becomes more complicated, if placement of an endodontic post is the treatment plan.
Traditionally, the long‐term use of calcium hydroxide has been the treatment of choice for induction of apexification in non-vital immature teeth . However, the introduction of MTA in 1993  provided a successful alternative with advantages including establishment of an instant apical barrier, setting ability in wet environment, promising sealing ability, provision of a shorter treatment period, improving patient compliance, ease of handling, and possibly increased fracture resistance of immature teeth [17,18,19,20,21]. CEM which was introduced in 2006 , has shown features comparable to those of MTA such as fracture resistance  and sealing ability . CEM has even shown some superior results such as higher antibacterial effect , significantly shorter setting time, easier handling, and no tooth discoloration .
Although recently developed posts such as zirconia posts, carbon posts and glass fiber posts have offered better esthetic results than cast metal posts [5, 26], the superiority of success rate is controversial. Some studies have reported reduced chair-side and laboratory time of the new post systems [10, 12, 26], while others reported the advantages of cast metal posts such as higher retention  or superiority of results in special circumstances such as when multiple teeth need post systems or in case of tooth mal-alignment, and in small teeth with minimal dental tissue .
Overall, cast metal posts are still one of the most used systems, especially for the posterior teeth [28, 29]. Unfortunately, caries of young permanent teeth is still highly prevalent among children and adolescents in many countries [30,31,32]; and excessive tooth destruction due to dental caries is still a major reason tooth loss in these age groups [33, 34]. In case of severe tooth destruction, cast metal posts are still regarded as the gold standard for restoration , but previous studies concerning fracture resistance of endodotically treated immature teeth only focused on fiber post restorations [4, 36, 37]. Fiber posts have elastic modulus similar to that of the dentin , hence the results cannot be generalized to other post systems including cast metal posts.
In the present study, the effect of type and thickness of apical plug materials (MTA vs CEM/ 3 mm vs 5 mm) on fracture resistance of endodontically treated immature teeth restored with cast metal posts was assessed. Several studies have shown that a full canal obturation or an apical plugging by bio-ceramics, increases the fracture resistance of either mature or simulated immature teeth compared to the roots which were instrumented but were not filled, or were filled only with gutta-percha and sealer [8, 39, 40]. Full canal obturation by bio-ceramics is not indicated when placement of an endodontic post is the treatment plan, because further removal of the material for post-space preparation might not be easy , and also the material is more expensive than gutta-percha and sealer. Therefore, in such cases, MTA or CEM are used as apical plugs and the rest of the root canal is filled with gutta-percha and sealer [42, 43].
The results of the current study showed that the fracture resistance of samples in all experimental groups were higher than the control group (gutta-percha and sealer). However, the superiority was not statistically significant except for teeth filled with a 5 mm MTA apical plug.
While this is the first study to compare the effect of MTA and CEM apical plugs on fracture resistance of teeth restored with cast metal posts, effect of the two materials on fracture resistance of teeth has been compared in other circumstances and controversial results were reported. Evren et al.  compared the fracture resistance of simulated immature human teeth using 4 mm apical plugs (MTA, CEM, and Biodentine), with fiber post and composite resin restoration. They reported no statistically significant difference of fracture resistance between the experimental groups, which is in accordance with the results from the current study. Evren et al. also reported that fracture resistance values for all experimental groups were significantly higher than the control group, while in the current study, the difference was only significant for MTA5 group. The difference between the results regarding the comparison of the experimental and control groups may be explained by methodological differences between studies (i.e., preparations of the control groups, thickness of the plugs, type of post systems, and restorative materials).
Sarraf et al.  compared the fracture resistance of immature bovine teeth completely filled with MTA, CEM, and Biodentine with no post-space preparation and placement. They reported that MTA and Biodentine showed superior results over CEM. These results seem to be inconsistent with the results of the current study which showed similar fracture resistance values for both CEM and MTA. However, these differences could also be explained by methodological differences between studies, particularly in using full canal obturation or apical plug. Sarraf et al. also reported that the fracture resistance was not different for CEM, gutta-percha and sealer, and control (dried cotton wool filling) groups. The difference between results regarding the comparison of the experimental and control groups in the two studies can also be explained by different preparations of control groups and the thickness of obturations.
Milani et al.  also compared the fracture resistance of simulated immature human incisors filled with MTA, CEM, and MTA plus composite resin with negative control (untreated teeth) and positive control (unfilled teeth) groups. The results of this study showed no significant differences among three experimental groups, which appears to be in consistent with the present study. Milani et al. also reported no significant difference among MTA plus Composite and CEM groups with positive and negative control groups. While MTA group had significantly higher strength values than positive control. Difference in results regarding the comparison of the experimental and control groups in the two studies is observed; which can be related to different preparations of the control groups and use of post systems.
The current study also showed no statistically significant difference of fracture resistance regarding the thickness of the apical plugs (3 mm or 5 mm). Although several studies have been performed to compare the effect of using different thicknesses of apical plugs on root- end sealing ability [45,46,47,48], the studies assessing the effect of thickness on the apical of mechanical properties are rare. Madani et al.  compared the fracture resistance of simulated immature teeth, filled with 3 and 5 mm apical plugs of MTA and CEM with a control group (5 mm gutta-percha). Teeth were restored with glass fiber post and composite resin. As consistent with the current study, Madani et al. reported no statistically significant difference of fracture resistance between the experimental groups. However, unlike the present study, no significant differences from the control group was found; which can be attributed to the fact that the studies used different post systems and restorations.
The effect of thickness on surface micro-hardness of MTA and CEM has also been evaluated in several studies. A study performed by Tabrizizadeh et al.  showed no statistical difference of surface micro-hardness between 4 and 8 mm MTA and CEM plugs. Login et al.  also reported no statistic difference of surface micro-hardness between 4 and 6 mm MTA plugs, while 10 mm plugs were significantly harder that 4 and 6 mm plugs. Although the results of both mentioned studies appear to be in consistent with the present study, testing variable mechanical properties (i.e., surface micro-hardness and fracture resistance), prevents accurate comparison.
Root-end sealing ability, in addition to mechanical resistance and hardness, should be noticed while comparing different apical plug materials and thicknesses. Adel et al.  who compared the root-end sealing ability of different thicknesses of MTA and CEM, reported a significantly higher sealing ability of 5 mm apical plugs compared to 3 mm apical plugs of both materials. Valois et al.  and Gosh et al. , also reported a higher root-end sealing ability of 4 mm MTA plugs compared to lower thicknesses. Thus, decisions for clinical use of apical plugs should not be made only on the basis of mechanical properties. And comprehensive in-vitro and in-vivo studies of various factors influencing the clinical outcome of endodontics and prosthodontic treatment of immature teeth is required.
Overall, the results of most studies on fracture resistance are not comparable to each other, because of the great methodological variations regarding sample type (e.g. bovine/human teeth, premolars/incisors), sample preparations (e.g. immaturity simulation, root canal preparation and obturation techniques, post-space preparation, post systems), obturations (e.g. full canal/ apical plug, material thickness), coronal restorations (e.g. composite resins, metal cores or crowns), testing machine (fatigue/static load, speed, angle) and several other factors. Hence, there is a need to standardize the methods, in order to perform fair comparisons and interpretations.
In the current study, no crowns were placed on the cores, as in some other studies [51, 52], in order to avoid the confounding effect of assemblage of several adhesively bonded parts. Although use of crowns could be more similar to the clinical situation, and the researchers could limit the confounding effect by considering the mode of failure. Evaluation of the mode of failure of the specimens could also bring more information about the mechanisms and reasons of failures. Therefore, not assessing the failure mode is noted as a limitation of the current study and it can be suggested to the future researchers to conduct the assessment. Findings of the current study is based on a relatively small sample size which can also be considered as a limitation. Other limitations of the study are that exact same post-space dimensions could not be achieved because of operator dependency of the preparations and slight anatomic differences. However, size of the apical opening could be measured which can be considered in future researches. And finally, the limitations of an in-vitro, static fracture resistance test are understood.