INTRODUCTION

The preservation of the remaining coronal structure has emerged as a crucial aspect for the clinical success of post-retained restorations and seems to be more important than the post choice.^1-8 Clinical trials have confirmed the strong scientific evidence that, irrespective of restorative technique of pulpless teeth, the preservation of at least one coronal wall and 2 mm of ferrule to post placement significantly reduces the clinical failure risk.^6,9,10

Several available post systems have been proposed for the rehabilitation of endodontically treated teeth (ETT). It is known that cast posts and cores are associated with high rates of irreversible fractures^11,12 because they transfer more stress to the root dentin compared with fiber posts.^13-16 Fiber posts have an elastic modulus similar to that of dentin and are usually associated with repairable failures^17,18 because they more homogeneously distribute stress along the root and thereby prevent root fracture.^13,19-21

Given that the use of fiber posts has increased, it is important to assess how the magnitude and direction of functional loads play a major role in the concentration of stress within teeth restored with posts.^22,23 Horizontal loads lead to a significantly higher concentration of stress within dentin than loads parallel to the long axis of the tooth.^24-28 Because the loads are applied at different levels along the dental arch, anterior teeth are most likely to be subjected to more horizontally directed loads due to their inclination in relation to posterior teeth.^23,29 For example, a force applied at an angle of 90° to the anterior teeth causes the appearance of tensile and compressive stresses,³⁰ which can cause damage to teeth restored with posts because these teeth have weaker supporting structures.

In addition, resolution of the load applied at 45° into force vectors using fractographical analysis facilitated the modeling of the specific features of stress (tensile, compression, and shear) found in an anterior ETT fractured after mechanical cycling.³¹ These data are corroborated by the results of another study that found that the elastic modulus of a post, with regard to concentration, magnitude, and direction of dentinal stress, was dependent on the direction of the applied load.¹⁴ When examining models of posts with a high modulus, it was found that horizontal loads led to more stress on the apical area of the root; such loads suggested a vertical root fracture. On the other hand, when low-modulus posts were modeled, forces at 45° and 90° caused more stress on the cervical area, with a direction that suggests debonding of the post.

The direction of the applied force and the fiber arrangement of fiber-reinforced polymeric materials (FRP) directly influence their mechanical properties.³² Glass fiber posts (classified as a FRP) generally present longitudinal fibers (parallel fibers, 0°) distributed inside the polymeric matrix, so they can support high tensile stress when the forces are applied along their central axis.³³ However, when oblique forces (45°) are applied (anterior teeth), shear stresses are induced inside the polymer, leading to fracture by lower loads than when only tensile stresses are generated.^32-34 Recently, in vitro studies showed that shear stresses can be as harmful as tensile and compression stresses when oblique forces are applied in ETT restored with fiber posts^31,35 or when only glass fiber posts were submitted to static^33,34 and fatigue loads.³³

Thus, our aim was to perform a fractographical analysis of an endodontically treated upper lateral incisor restored with a glass fiber post and a metal-ceramic crown that fractured after three years of clinical service and compare that with in vitro findings available in the literature. We sought to assess the forces exerted on the fractured tooth with both a finite element and a scanning electron microscopy (SEM) analysis as well as to validate the in vitro findings.

METHODS

Case Description

A 49-year-old man presented at the Division of Prosthodontics with a fracture in the upper left lateral incisor, incurred while eating. Previously, on September 28, 2010, the patient had been referred to the Division of Prosthodontics of the Faculty of Dentistry with a large coronal fracture in the tooth, which had already been endodontically treated. A detailed anamnesis was performed. The patient had good general and dental health but was missing the maxillary right second premolar, first molar and second molar, as well as the maxillary left first premolar, second premolar, and first molar. The initial focus of the treatment was on the rehabilitation of the fractured tooth, followed by replacement of the other lost teeth. After clinical and radiographic examinations, we proposed restoration of the lateral incisor with an intraradicular post and metal-ceramic crown as well as a removable partial denture (RPD) for the maxillary posterior segments. The fractured lateral incisor presented two proximal contacts, was a sound tooth antagonist with periodontal support, had a mobility grade of 0, and had a remaining root length of 16 mm.

Rehabilitation consisted of the cementation of a glass fiber post (White Post DC, FGM, Joinville, Brazil) using a self-adhesive cement (RelyX U100; 3M ESPE, St Paul, MN, USA). It was cemented at a 10-mm length, with a 6-mm coronal length and 1.6-mm coronal diameter.

On January 3, 2013, the patient sought assistance again due to a post-meal fracture in the previously rehabilitated lateral incisor. The patient brought the fractured restoration, and during examination, the following was observed: 1) the metal-ceramic crown, composite core, and fiber post were still luted together; (2) the fiber post had fractured into two pieces, approximately at cervical level, as described in the literature;¹⁴ 3) resin cement had adhered to the dentin, indicating an adhesive cement-post failure; and 4) according to the patient, an RPD had not been affixed in the posterior maxilla due to financial reasons.

We observed the root integrity and proposed cementing a new post and creating a new restoration. We explained to the patient the importance of an RPD to stabilize the occlusion, and the RPD was then manufactured to avoid overloading of the anterior segment of the maxilla. The parts of the fractured restoration were analyzed using a stereomicroscope (Discovery V20, Carl Zeiss, Göttingen, Germany; Figure 1) and an SEM (Jeol JSM 5400, Jeol Ltd, Tachikawa, Japan; Figure 2).

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

Finite Element Analysis

A two-dimensional model of a lateral incisor was created using the software CAD Rhinoceros (version 4.0SR8, McNeel North America, Seattle, WA, USA) to simulate the bone; the periodontal ligament (0.3 mm); the root (length: 15 mm; width: 7 mm); the gutta-percha; the resin cement thickness between the fiber post and root dentin (100 μm); the fiber post; a resin composite core (height: 7 mm; width: 5.5 mm); the cement thickness between the core and the metal-ceramic crown (100 μm); and a metal-ceramic crown (height: 9 mm; infrastructure thickness: 0.5 mm; ceramic width: 0.7 mm). A chamfer of 1.2 mm was designed at the vestibular and lingual portions. The fiber post was modeled with 10 mm inside the root canal and 5 mm at the coronal portion.

After modeling, the geometry was imported into an STP format to Ansys software (Ansys 13.0, Houston, TX, USA) for boundary conditions and numerical simulation. Tetrahedron elements were used, generating a total of 27,264 elements and 29,549 nodes. After the convergence test, the mean size of the elements was 0.15 mm, with the exception of the fiber post and the resin cement, which presented elements of 0.05 mm. The interfaces were considered bonded, and the base and lateral faces of the bone were considered fixed in the x, y, and z directions. A force of 70 N (an intermediate value used in the study of Wandscher and others)³¹ was applied at 45° to an area of 1 mm² situated 2 mm below the incisal edge of the crown. The fiber posts were considered orthotropic, whereas other materials were considered isotropic (Table 1).^36-43 All materials were considered homogeneous and linear elastic. The maximum principal stress, minimum principal stress, and shear stress were evaluated using the model.

Table 1 Materials, Elastic Modulus, Poison Values, and References Consulted to Obtain the Values

View Fullscreen

RESULTS

Analysis of the Prosthetic Fragment

The images derived from the stereomicroscope showed that the failure occurred in the fulcrum region, below the tooth cervical level (approximately 3 mm), in which one post piece was broken and the other was attached to the restorative core/crown (Figure 1a,b). It is possible to observe palatine staining (brown spots) on the marginal cement and on the fiber post (Figure 1b) due to marginal leakage. The SEM images showed the fractured parts, consequences of the tensile and compression stresses (Figure 2a,b) and shear stress (scallops: Figure 2c; hackle lines: Figure 2d).

Figure 3 shows the bending moments and the fulcrum lines. The periodontal ligament permitted tooth movement and the 45° load (F) promoted tooth bending, forming a fulcrum line at the bone crest level (fulcrum 1: red line). In addition, at the cervical level another fulcrum line was formed by a 45° load and the cervical surface of the root (fulcrum 2: green line). The bending moment (M) was measured by the applied force (F) and the distance between the load application point and the fulcrum line (d).⁴⁴ Because the moment is directly proportional to the distance, the higher the distance, the higher the bending moment. The distance between the loading point and fulcrum 1 is higher than the distance between the loading point and fulcrum 2, so moment 1 (M1) is higher than moment 2 (M2). Thus, the consequences of M1 on the post were higher, leading to a fracture at that point (Figure 1b).

View Figure

• Download as Powerpoint
• Download Figure

Figure 4 presents the stresses that acted on the restored tooth, the graphic of tension, and the formulas. By means of the parallelogram law,^44,45 the 45° force has been mathematically decomposed into a cartesian axis in force vectors to obtain the horizontal (F_x) and vertical (F_y) components of F (Figure 4a). F_x produces compressive loading (C) uniformly distributed in the cross-section as shown in the graphic of tension (Figure 4b). F_y produces transverse loading bending on the dental structure, generating normal tensile stresses (T) on the lingual surface and normal compression stresses (C) on the buccal surface as presented in the graphic of tension. These stresses tend to be zero or a minimum in the center of the dental element (N_L) and a maximum in the outer portions (Figure 4c). The sum of B and C results in the A graphic of tension, where it is possible to observe displacement of the neutral line to the lingual surface because there is more compressive stress acting in the structure. F_y also produces shear stress in parallel planes to the longitudinal axis of the structure due to the transverse loading. This stress is at a minimum at the outer portions and a maximum at the center, as noted in the graphic of tension (Figure 4d).^31,33,44,45

View Figure

• Download as Powerpoint
• Download Figure

Finite Element Analysis

Finite element analysis (Figure 5) presents numeric stress values of compression, tensile, and shear on the post after 45° load application. Compression and tensile stresses are considered normal tensions because they act in the long axis of the tooth. These stresses are at a maximum in the outer portions of the post. It is possible to observe that the higher value of compression stress is on the buccal region of the post (point C—Figure 5a) and the higher value of tensile stress is on the lingual region of the post (point A—Figure 5b). In relation to shear stress, the values were maximum in the center and minimum in the outer portions (points B and B′—Figure 5c).

View Figure

• Download as Powerpoint
• Download Figure

It is important to note that features found in the fractographic analysis coincide with biomechanical and finite element analysis (scallops and hackle lines: maximum shear zones; glass fibers exposed without matrix: maximum tensile zones; kneading: maximum compression zones).

DISCUSSION

Clinical studies^1,2,6,7 and literature reviews^3,4,5,8 have shown that the greater the remaining coronal tooth structure, the greater the survival of posts. In addition, long-term follow-up investigations assessing several tooth types have demonstrated that the survival of teeth with substantial tooth tissue is unaffected by the use of a post.^7,46-48 This means the presence of remaining coronal structure rather than the type of post is the most important clinical condition for success of ETT. In vitro studies have stated that teeth with at least 2 mm of remaining coronal structure provide higher fracture resistance,^9,49,50 and greater ferrule promotes a more homogeneous stress distribution in ETT and a lower probability of clinical failure.^50-52

Before discussing the biomechanical issues of the current forensic investigation, it is important to emphasize that the patient in this case report had no remaining coronal tooth structure, which might have caused (or elevated the risk of) the clinical failure. The factors discussed next should be considered secondary to remaining coronal tissue when analyzing ETT restored with posts.

Anterior teeth experience different load levels, principally oblique loads that lead to bending of the restored tooth.^14,23,29 Such loads result in extremely harmful stresses (tensile/compression^30,31,33-35and shear stresses^31,33,34,35) on an ETT restored with a post.

It is likely that the fracture of the lateral incisor can be attributed to an association of reasons. According to Figure 1a and b, the fracture occurred in the fulcrum zone 3 mm below the cervical level. As seen in Figure 3, two fulcrum lines formed on the post: one at the bone crest level and the other at the cervical level. The bending moment is the reaction induced in a structure when an external force, or moment, is applied to the element, causing bending.⁴⁴ The current fracture occurred at the farther point of load application, on M1 (where the bending effects are greatest).

Second, failure analysis made it possible to observe areas characterized by tensile, compression, and shear forces. Anterior teeth are positioned in the dental arch at an approximate angle of 45°. Any oblique load exerted under these teeth leads to bending of the dental structure and results in the onset of tensile, compression, and shear stresses.^31,33,35 A classic study described tensile and compressive stresses in a tooth undergoing bending.³⁰ More recently, other in vitro studies have analyzed these concepts in the failure analysis of both fiber posts and ETT restored with posts;^31,33,35 these observed that shear stress was present and may be as detrimental to restorative structures as tensile and compressive stresses.

During bending, tensile and compression stresses operate on the teeth, with the stresses at a maximum in the outer portions and a minimum in the center of the restoration assembly. The opposite occurs with shear stress; that is, the stress is at a maximum in the center and a minimum in the outer portions (Figure 4d).^{31,33-35,44,45} The effects of tensile and compression stresses are observed in Figure 2 under the assembly crown/core/post. Primarily, due to tensile stress,⁵³ an adhesive failure between tooth and post buildup (subcritical tensile failure of buildup/dentin interface⁵⁴) promoted a marginal leakage, first at the margin (brown staining spots), and then penetrating into the restoration (brown spots on fiber post, Figure 1b). This adhesive failure indicates the presence of a crack in the palatal region⁵⁵ as a consequence of the debonding of the core/crown due to the tensile stress that promoted bending of the intraradicular glass fiber post and the catastrophic failure of the crown.

Moreover, tensile stresses may be observed in the fiber post fracture. The lost part indicates the surface exposed to tensile stress (lingual portion). It is characterized by the presence of lost fibers without a matrix in both fractured parts (Figure 2a,b), as opposed to the surface exposed to compressive stress (Figures 1b and 2a,b). The compressive area is characterized by a “kneading” in which fiber bending occurs with compression of the matrix. These features were also found in an in vitro study that evaluated mechanical cycling and fracture load of weakened roots restored with posts.³¹ Finite element analyses showed the values of force in the areas of tensile and compression stresses inside the post (Figure 5). These values presented low magnitudes, when compared with the force applied on the palatal region, and could be explained by the dissipation of the stresses through the model.

Shear stress also negatively affected the restored ETT (Figure 4d), and it was concentrated strongly in both the center of the post and in the cement layer (considered a fragile area and subject to defects). In Figure 5c, it is possible to observe the virtual values of shear stress (higher in the center and lower in the outer portions) in the horizontal and vertical planes of fiber posts, which explains the failure behavior of this system.^{31,33,35,44,45} This explains why the failure in the central region of the post was classified as intralaminar mode II in-plane shear inside the post,^32-34 and it presents as hackles and scallops where they intersect adjoining areas of fiber-to-matrix separation (Figure 2c,d). Another explanation for the fiber post failure could be related to an initial adhesive failure between fiber post and root dentin, which could lead to concentrated stresses at the fiber post and the root dentin, as explained by Santos and others,⁵⁶ and could explain why the fiber post failed under a low magnitude of force. Unfortunately, it was not possible to model a nonbonding condition (nonlinear). Recent in vitro studies have found the same fracture mode by shear stress in fiber posts subjected to bending,^33,34 as well as fracture resistance after mechanical cycling of roots restored with fiber posts.³⁵ These in vitro results support the finding of the current forensic investigation.

Another important topic is the fiber post behavior when a load is applied to the restorative structure. A fiber post with longitudinal parallel fibers (0°) presents high resistance if force is applied to the longitudinal axis, but when oblique forces are applied, the post's response is different. The fiber arrangement inside the post directly affects its mechanical properties.^32-34 Fiber posts with parallel fibers (as in the current forensic investigation) have limited ability to support an oblique load (ie, 45° inclination) and may fracture with lower loads due to shear stress compared with tensile loads.³³ An alternative to this problem could be the development of fiber posts with different fiber arrangements (other fiber alignment angles) that are able to support loads in different directions. This finding supports the results presented by previous in vitro studies that used fiber posts with parallel fiber arrangements (0°).^33,34

It is difficult to predict exactly where the fracture initiated; however, it is likely that initially there was a failure by shear stress inside the post due to the limited capacity to support bending stresses, leading to an adhesive failure in the lingual portion of the core (indicated by brown spots on Figure 1b).

The criteria for an acceptable dental occlusion involve the presence of axial bilateral posterior contacts and either absence of contacts or smooth contacts in the anterior region.⁵⁷ In this clinical situation, the patient had lost posterior dental occlusion, which overloaded the restoration and intensified the consequences of the tension, compression, and shear loads in the restored lateral incisor.

Factors such as the quantity of the remaining dental structure, the position of the tooth in the arch, the absence of posterior support, the selection and adequate application of the restorative strategy, the type of antagonist, and the presence of RPDs are important issues to be evaluated. The negligence of these factors may result in a greater effect of the bending loads (tensile, compression, and shear) on ETT anterior teeth. For ETT restored with fiber posts, the remaining coronal tooth structure and the resulting ferrule are very important to increase the survival of the restoration.

In conclusion, it is important to emphasize that two factors were critical to the fracture of the restored lateral incisor: the significant loss of remaining coronal tooth structure and the lack of posterior occlusal support. If the coronal structure was larger and there were occlusal bilateral contacts, the effects of the stresses would be minimized and the biomechanical stability of the restorative assembly would probably be assured.

CONCLUSION

Given the limitations of this clinical report, some considerations can be drawn.

• 1.

  The fracture did not occur by one single factor but due to the association of several factors.

• 2.

  When restoring an ETT, the preservation of the remaining coronal tooth structure is a must.

• 3.

  The theories formulated in in vitro studies appear to be correct: Anterior ETT restored with posts suffer tensile, compressive, and shear stresses across the buccolingual depth at the cervical level. The failure described in the clinical situation of the current study is in accordance with other in vitro studies that described favorable or reparable failures in ETT restored with fiber posts.

• 4.

  The development of fiber posts with different fiber arrangements capable of withstanding both axial and oblique forces may be one possible solution to assist in anterior ETT rehabilitation.

• 5.

  In teeth restored with posts and cores, knowledge of the direction of the forces operating is strongly advisable, especially considering that these forces vary according to the position of the restored teeth in the dental arch (anterior or posterior teeth).