Dental fracture can be a consequence of tooth weakening, abrasion, attrition, erosion, abfraction, caries lesions, restorative procedures, and trauma [1234]. Weakened teeth are characterized by a decrease in their volume and the loss of healthy dentin due to caries lesions and tooth preparation [15]. The most common causes of dental fractures are high-impact loads during chewing or premature occlusal contact [56]. It has been reported that dental fractures occur more frequently in teeth with large restorations than in teeth with small restorations or sound teeth [5].

Extensive restorations are often associated with endodontic treatment due to pulp vitality impairment. Endodontically-treated teeth are at a higher risk of biomechanical failure than vital teeth [7]. The factors posing the highest risk of fracture include loss of dental tissue [8], preparation of an endodontic access cavity to disinfect the root canal [6], and/or changes in the physical properties of dentin due to dehydration [8]. In cases of extensive tooth destruction, intraradicular retainers are required to provide sufficient retention for the restorative material since the dental remnant is insufficient [910]. However, the use of intraradicular posts in order to strengthen dental structure has already been reported to be ineffective [1112].

For posterior teeth with a mesio-occlusal-distal (MOD) cavity, the most suitable treatment consists of a direct restoration using composite resin due to its high bond strength to the dental substrate [3131415]. In cases where dental pulp has been compromised and endodontic treatment has been performed, the best protocol prior to performing the restoration, which may increase the resistance of the dental structure, is still not well-defined [1617]. It has been generally reported in the literature that glass ionomer cement should be used to support the composite resin restoration [1819]. Other authors have reported the use of fiberglass posts and filling the pulp chamber with resin cement [11]. Finally, some manufacturers offer pre-framed metal posts with this indication. Therefore, this study evaluated whether there were any advantages in using a prefabricated post in terms of load-bearing capacity, cusp deflection, and stress distribution of endodontically-treated human upper premolars with MOD restoration. The first null hypothesis was that different techniques would not influence the cusp deflection and load-bearing capacity of human upper premolars with MOD restoration, while the second null hypothesis was that the different restorative techniques would not influence the stress concentration in the cervical area of the teeth.

One-way ANOVA revealed that the different treatments significantly influenced the mean values of cusp deflection (p = 0.027) and load-bearing capacity (p = 0.001). The restorations using a metal post (MP group) showed the highest cusp deflection and the lowest mean value for load-bearing capacity. The GIC and CR groups showed lower mean values for cusp deflection and higher mean values for load-bearing capacity. The descriptive statistical analysis and results of the Tukey test are shown in Table 2. According to the finite element analysis, all modalities presented a similar profile for cusp linear deformation (Figure 3). However, more fringes were present in the MP group, followed by the FGP, GIC, and CR groups. In analyzing the tooth surfaces (Figure 4), it was noted that the MP group presented more areas with red color on the colorimetric stress map, meaning that it experienced the highest stress concentration. Figure 5 shows that both metallic and fiber posts concentrated more stress (metal > FGP) in the cervical area, increasing the stress concentration below the restoration compared to the GIC and CR groups. There was no significant difference among the groups in the stress distribution in the composite resin restoration (Figure 6). All groups showed the highest stress peak on the palatal surface (5.45–5.89 MPa), in comparison with the buccal face (2.51–3.23 MPa). The results are summarized in Table 2.

This study aimed to evaluate the influence of prefabricated posts on cusp deflection, load-bearing capacity, and stress distribution in endodontically-treated upper human premolars with a MOD restoration. The microstrain results showed statistically significant differences among the experimental groups, leading the first null hypothesis to be rejected. The stress distribution results showed different values among the evaluated treatments, causing the second null hypothesis to be rejected.

Restoration of teeth with composite resin without endodontic treatment has already been reported to show promising mechanical behavior [2728]. These previous studies have suggested that composite resin restoration decreases cusp deflection. This can be explained through the adhesive characteristics of this type of restoration, which partially restore the rigidity of the dental element through adequate bond strength at the tooth/restoration interface [29].

Finite element analysis showed that the mechanical behavior was very similar across all groups in an ideal situation in which the adhesive interface is perfectly bonded between the tooth and restoration (Figure 6). Thus, the possibility of dental fracture due to stress concentration in the dental structure seems to be more influenced by the restorative treatment after endodontic access than by marginal infiltration along the restoration. Figure 4 shows that the palatal cusp was the most susceptible to fracture in response to a load being exerted on the restorative material. This finding is in agreement with previous studies that evaluated the biomechanical behavior of healthy [22] and restored teeth [23] using finite element analysis. In addition, it is important to note that intraradicular posts are often indicated for palatal canals [11], which further contributes to this cusp being stressed during masticatory load. This effect is even greater for metal posts than fiberglass posts (Figures 3 and 4).

Glass ionomer cement is used as a base material for restorations because its mechanical properties are similar to those of dentin [1930]. This advantage is reflected in our results, in which the teeth that had their pulp chamber filled with glass ionomer cement and were directly restored with composite resin presented similar cusp deflection values to those restored with composite resin without endodontic treatment (Table 2). The good performance of glass ionomer cement as a substitute for lost dentin during access opening for endodontic treatment obtained in this study is in agreement with previous studies [2431] which indicated that glass ionomer cement is a suitable material to be used under a composite resin restoration.

The use of a metallic post inside the root canal, in addition to removing more dental tissue [11], means that a non-adhesive restorative material is in contact with the composite resin and the remaining structures; this increases the possibility of cusp movement, and thereby elevates the risk of tooth fracture. The results showed a significance difference between the group restored with a metal post and the group restored with a fiberglass post (Table 2). In addition, as shown in Figure 5, a higher stress concentration was present below the restoration in both groups with an intraradicular post than in the GIC and CR groups, with larger zones of red color for the MP group. This fact corroborates previous reports that fiber posts are preferable to prefabricated metal posts [32]. A correlation between the cusp deflection and fracture resistance values could also be observed, which has also been reported in the literature [3334].

The group that presented the highest cusp deflection means herein also showed the lowest load-bearing capacity. Despite the results of the present study, it is worth noting that it has limitations because it is a laboratory study. The effects of cyclic fatigue, oblique load in the restoration [35], thermal aging [36], different composite materials [37] and sliding occlusal loads can modify the mechanical behavior of restored teeth, thereby weakening the adhesive strength and altering the final results. Moreover, pH variation and defects in the restorative material, such as bubbles within the restoration, were not reproduced in the numerical simulation, as ideal conditions of a homogeneous (without bubbles or defects in the cement or composite) and isotropic material were assumed. Another important limitation of this study is that the finite element simulation was performed with general materials, using mechanical properties reported in the literature [192425]. However, the values used in this study cannot represent the mechanical properties of the current in vitro materials during the strain gauge analysis. This simplification was present for all groups; for that reason, the mechanical behavior may not present precisely the same values for each method, but the same profile is likely to be observed. Finally, even though the results are valid, they should be extrapolated with care and explored in future longitudinal clinical studies to define the best clinical approach.

There is no mechanical advantage in using intraradicular posts when an endodontically-treated premolar requires MOD restoration. Filling the pulp chamber with glass ionomer cement and restoring the tooth with only composite resin showed the most promising results in terms of cusp deflection, failure load, and stress distribution.