INTRODUCTION

Fracture is a frequent cause for extraction of an endodontically treated tooth. Two studies evaluating the reasons for the extraction of endodontically treated teeth demonstrated unrestorable cusp or vertical root fractures in 11.2 % to 28.5 % of cases.^1,2 Hence, it is of clinical importance that the restoration of an endodontically treated tooth results in a tooth-restoration complex that is resistant to fracture.³ The durability of the post-endodontic restoration largely depends on the amount of remaining tooth tissue.^4–6 Data from a retrospective study showed that endodontically treated molars with a large amount of remaining tooth structure had a better 5-year cumulative survival rate than molars with two remaining walls or less (78% vs 18%).⁵

Post-endodontic restoration can be achieved by an indirect or a direct operative procedure. A conventional indirect procedure of restoring severely mechanically compromised teeth is by creating a ferrule of at least 2 mm in height and cementation of a full contour crown in cases where there is ample remaining vertical height to provide retention and resistance to the indirect restoration.⁷ If the former is not the case, post-and-core buildup is commonly provided prior to fabrication of the crown.⁸ Endocrowns could be a viable adhesive alternative to these post-and-core buildups in structurally compromised molar teeth.^9,10 An endocrown is a monolithic restoration that extends into the pulp chamber. This type of restoration functions particularly well in molar teeth, since the pulp chamber of such teeth provides a large area for adhesion of the restoration. When fabricated from a glass ceramic or indirect composite, adhesive cementation results in the preservation of tooth structure and enamel because retention does not merely rely on macro-geometry. However, when enamel is lost, adhesion to dentin remains a clinical challenge.¹¹ For ceramic inlays and onlays with a cervical outline in dentin there is a 78% higher risk of failure compared with restorations with an outline in enamel.¹² Immediate dentin sealing (IDS) improves adhesion to dentin.¹³

In molars with no or only a single coronal wall remaining, the clinician and patient have to consider treatment alternatives in relation to costs and durability. Indirect restorations are relatively costly and frequently overstretch the patients' financial budget, especially after recent pre-restorative endodontic treatment. A possible alternative could be a directly made full composite buildup. Fabricating a full composite buildup, however, is a clinical challenge, since it is rather difficult to adequately restore the original morphology of the tooth. Polymerization shrinkage and the degree of conversion also affect the final result and the survival of the restoration.¹⁴ Incorporating glass fibers into the bulk of composite may result in less polymerization shrinkage, higher fracture resistance, and deflection of fractures to a more restorable failure.¹⁵

In a systematic review of the Cochrane collaboration,¹⁶ it was concluded that further clinical trials are needed to compare direct and indirect restorations in the rehabilitation of endodontically treated teeth. For a better understanding of the mechanical behavior, various types of restorations may be proposed, as well, which forms the rationale for the present study. The fracture strength of severely compromised endodontically treated molars restored by various direct and indirect means, using composite or glass ceramic materials, either with or without a post, were compared under the null hypotheses that there would be no difference in mean fracture strength and mode of failure among the various groups.

METHODS AND MATERIALS

A total of 105 sound extracted third molars, similar in size (mesiodistal length 9–11 mm; root length 10–13 mm as measured from the cementoenamel junction [CEJ] to the apex), were included (n=15 per group). Exclusion criteria were: presence of caries or cracks, abnormal morphology, roots <10 mm, presence of restorations, or root canal treatment. All molars were embedded 1 mm below the CEJ in polyvinylchloride tubes (height: 10 mm; diameter: 12 mm) using autopolymerizing acrylic resin (Autoplast, Condular, Wager, Switzerland). Teeth were stored for a maximum of four months in distilled water before preparation. All materials used are listed in Table 1.

Table 1 Brands, Types, Chemical Compositions, Manufacturers, and Batch Numbers of the Materials Used for the Experiments

View Fullscreen

Randomization and Specimen Preparation

After endodontic treatment all teeth were randomly assigned via sealed envelopes to a control group (control; no further preparation) or one of six treatment groups (n=15; see Figure 1). A three-dimensional scan was obtained for all teeth in the BioGeneric Copy mode of an intraoral scanning device (Cerec Omnicam, Sirona Dental Systems GmbH, Bensheim, Germany). A putty impression (Provil Novo fast set, Heraeus Kulzer GmbH, Hanau, Germany) of the intact molars was made. The crowns of teeth in all treatment groups were removed using a coarse diamond wheel bur (5909 FG, Komet Dental, Lemgo, Germany) 1 mm above the CEJ (preserving a ring of enamel around the outline) and immediately restored.

View Figure

• Download as Powerpoint
• Download Figure

RESULTS

All specimens survived the thermomechanical aging process. Wear facets were present on both the ceramic and composite restorations. An overview of the fracture strength results is presented in Table 2. Fracture strength was significantly affected by the type of restoration (F[6, 98]=5.89, p=0.000, ω=0.47). Post hoc analysis revealed that there was a significant difference between group LDS (3217±1052 N) on the one hand and control (1890±774 N; p=0.001, d=1.44), GFRC (1823±911 N; p=0.000, d=1.42), C (2192±752 N; p=0.018, d=1.12), and CP (1830±590 N; p=0.000, d=1.63) on the other hand. No significant differences were found between groups LDS, P-LDS (2697±665 N), and EC (2425±993 N).

Table 2 Fracture Strength Results (Newtons)^a

View Fullscreen

Table 3 shows the modes of failure per group. There was a significant association between the mode of failure and the type of restoration (p=0.005; two sided). The odds for a repairable fracture when GFRC was used were 4, 5.4, 8, and 28 times higher than for the full contour crowns, control group, EC, and other composite groups, respectively. Figures 2A through C and 3A through C show scanning electron microscope images of representative specimens. In group GRFC, a deflection of the fracture from the central part is visible (Figure 2A). Figure 3D shows that the adhesive failure occurred between the interface of the dentin and IDS layer, whereas the IDS layer, composite, and ceramic are still bonded.

Table 3 Failure Modes^a

View Fullscreen

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

DISCUSSION

The objectives of this study were to investigate the influence of the type of restoration on the fracture strength and mode of failure of severely compromised endodontically treated molars. According to the results, the hypothesis that the type of restoration had no influence on the fracture strength has to be rejected. Specimens from the LDS group obtained a significantly higher mean fracture strength than the control and direct composite groups. To the authors' knowledge, no fracture strength studies are currently available that compare the direct composite and indirect restoration of endodontically treated molars. Even after thermomechanical aging, all types of restorations were able to withstand the mean masticatory force in humans, ranging from 600 N to 900 N.^17–19 This is in line with other studies.^20–22

Only a few studies investigated the fracture strength of endodontically treated molars restored with cusp-replacing direct composite restorations; however, the specimens were not subjected to any form of aging.^20,23 Salameh and others reported fracture strengths of 677±21 N and 833±180 N for direct composite restorations with and without a glass fiber post, respectively.²³ In another study, the fracture strength of extensive direct composite restorations without posts was 1421±320 N.²⁰ The fracture strengths found in the present study are comparable or higher, although the specimens were thermomechanically aged to simulate five years of clinical service.²⁴ A possible explanation could be the use of a highly filled microhybrid composite resin (70% volume, 86% weight filler load).²⁵ Only one study described the fracture strength of endodontically treated molars restored with fiber-reinforced composite (2251±586 N).¹⁵

The mean fracture strength of endocrowns in this study is comparable to that of endocrowns as reported by Gresnigt and others under axial loading (2428±566 N).²¹ Carvalho and others reported values of 3181 N and 3265 N for full contour lithium disilicate crowns with a 4-mm composite buildup and endocrowns.²² This higher value compared with those in the present study may be explained by the use of an antagonist resin sphere, which deforms under compressive load. No significant difference was found among the three ceramic groups. This is in line with another in vitro study, where no significant difference in mean fracture strength was found for the full coverage and endocrown lithium disilicate restorations on endodontically treated molars (1076±132 and 989±109 N respectively).²⁶

Both for the direct composite and lithium disilicate group, the use of a post did not contribute to a higher fracture strength. This is in accordance with several other studies.^23,27–29 Lithium disilicate endocrowns performed significantly better than a composite post and core buildup and full contour crown (675±159 N vs 470±130 N).²⁷ In endodontically treated molars without ferrule, a post had no added benefit to the failure load of ceramic crowns covering a composite buildup.³⁰ This was also the case for large composite restorations with or without the use of a glass fiber post.^23,28

As for the hypothesis that the type of restoration had no effect on the failure mode, the results show that group GFRC resulted in significantly more repairable failures; thus, the hypothesis can be rejected. A reason for this could be the deflection of the fracture during crack propagation by the glass fibers because of their different orientation (see Figure 2B).¹⁵ Furthermore, glass fibers seem to absorb a lot of stress, as is apparent from the velocity hackles in Figure 2C.³¹ In contrast, the representative specimens of groups C, P-LDS, and EC showed fractures within the pulpal chamber, resulting in nonrepairable failures (Figure 3A through C). Strain in the direct composite transfers stress to the root dentin, which is confirmed by strain measurements and finite element analysis.³² Even when compared with the control, which only had a small class I filling, the odds for repairable failure of group GRFC were 5.4 times higher. The glass fiber reinforced composite therefore might be an interesting method to increase the chance of tooth survival after fracture. More research about the possibilities of this material is needed. The use of a post did not result in more favorable fractures for CP (only one repairable failure), as opposed to Salameh and others, who found 100% restorable fractures when a glass fiber post was present.²³ This can be explained because the specimens in this study failed under a much higher load. As for the full contour crowns, the use of a post did not result in more repairable failures (five repairable failures for LDS and P-LDS). This is confirmed by another study.²⁹

As is apparent from Figure 3D,E, when the adhesive interface is challenged by tensile forces due to fracture, the dentin-adhesive bond still remains the weakest link.

Several factors influence fracture resistance, such as direction of the force, crosshead speed and tooth embedment.^20,33 Some authors advise simulating the periodontal ligament.³³ Due to the risk of changes in the periodontal ligament material as a result of the extensive aging, no simulation of the periodontal ligament was incorporated.

In this study, the load was directed along the axis of the tooth, which resembles the direction of intraoral forces. However, a continuous load is not typically encountered in the mouth. During function, restorations are subjected to more fatigue loading. This was simulated by means of a chewing simulator. It is estimated that during normal function, the maximum whole tooth load varies between 70.6 N and 146 N.^34,35 In the present study, a load of 50 N was used during aging, which may have been too low. In contrast, Rosentritt and others stated that 1.2×10⁶ times 50N may be sufficient to estimate the survival rate in vivo.³⁶ In future studies, cyclic loading might be a better simulation of the clinical environment.

From a clinical point of view, it is important to consider the consequences of creating a ferrule on severely compromised teeth: enamel is lost and surgical crown lengthening is often necessary. Moreover, adhesive cementation in these cases might not be possible any more, thereby reducing the fracture strength of the glass ceramic.³⁷ The results of this study contribute to the evidence of adhesive rehabilitation of such biomechanically compromised endodontically treated molar teeth, whether by means of direct composite or glass ceramic endocrowns, although more clinical research is needed.

CONCLUSIONS

Within the limitations of this in vitro study, the following could be concluded:

• Under axial loading, direct composite restorations showed similar fracture strength values as lithium disilicate crowns with a composite post-and-core buildup and endocrowns;

• Endocrowns performed similar in terms of fracture strength compared with full contour lithium disilicate crowns;

• The use of a glass fiber post did not influence fracture strength;

• The use of a glass fiber reinforced composite core was associated with more repairable failures.