A study into the laboratory techniques for investigating by chanel lam

UNIVERSITY OF SIENA SCHOOL OF DENTAL MEDICINE PhD PROGRAM: “DENTAL MATERIALS AND THEIR CLINICAL APPLICATIONS”

A STUDY INTO THE LABORATORY TECHNIQUES FOR INVESTIGATING THE RESISTANCE TO FRACTURE AND THE CLINICAL PERFORMANCES OF ENDODONTICALLY-TREATED TEETH RESTORED WITH FIBER POSTS AND DIFFERENT RESTORATIVE MATERIALS CONFIGURATIONS. MECHANICAL TESTS AND FINITE ELEMENT ANALYSES.

Candidate: Roberto Sorrentino Promoter: Prof. Marco Ferrari Copromoters: Prof. Franklin R. Tay, Prof. Fernando Zarone

CONTENTS

CHAPTER I: Assessing the validity of mechanical tests to investigate the resistance to fracture of endodontically-treated teeth................................................................pag.5 Mechanical resistance to fracture of endodontically-treated teeth: review of the literature....................................................................................................pag.6 Introduction..........................................................................................................pag.6

The role of posts in the restoration of endodontically treated teeth...............................pag.7

Corornal restorations.....................................................................................................pag.8

Restorative materials.....................................................................................................pag.9

Experimental studies: review of the literature..............................................................pag.11

Mechanical tests..........................................................................................................pag.13 Load type...........................................................................................................pag.14 Load application area........................................................................................pag.15 Load frequency..................................................................................................pag.17 Load jig characteristics......................................................................................pag.18 Load speed........................................................................................................pag.19 Load intensity....................................................................................................pag.20 Angle of load application...................................................................................pag.21 Simulation of supporting tissues........................................................................pag.23 2

Evidence based selection of dental materials..............................................................pag.24

The experimental environment....................................................................................pag.25

Mechanical properties of dental materials...................................................................pag.26

Step by step procedure to accomplish laboratory and clinical evaluations of dental materials and restorative systems.............................................pag.28

Protocol to perform mechanical static tests.................................................................pag.30

CHAPTER II: Investigating the mechanical resistance to fracture of endodonticallytreated premolars simulating different restorative systems.................................pag.51 Effect of post and core restorations on the resistance to fracture of endodontically treated maxillary premolars in different restorative systems........................................................................................................................pag.52

Effect of post retained composite restorations of mesial-occlusal-distal cavities on the fracture resistance of endodontically treated teeth..............................pag.75

CHAPTER III: Investigating the mechanical resistance to fracture of endodonticallytreated molars simulating different restorative systems.....................................pag.102 Effect of fiber post and resin core restorations on fracture resistance and pattern of endodontically treated mandibular molars..........................................pag.103 3

CHAPTER IV: Assessing the validity of laboratory mechanical tests to investigate the resistance to fracture of endodontically-treated teeth by means of virtually simulated clinical situations...................................................................................pag.116 Evaluation of the biomechanical behaviour of maxillary central incisors restored by means of endocrowns compared to a natural tooth: a linear static 3D Finite Elements Analysis.............................................................................pag.117

Three dimensional finite element analysis of strain and stress distributions in endodontically treated maxillary central incisors restored with different post, core and crown material....................................................................................pag.144

Three dimensional finite element analysis of stress and strain distributions in endodontically treated maxillary central incisors restored with post and core direct and indirect restorations..........................................................................pag. 170

CHAPTER I

Assessing the validity of mechanical tests to investigate the resistance to fracture of endodontically-treated teeth.

Mechanical Resistance to Fracture of Endodontically-Treated Teeth: Review of the Literature.

Introduction Endodontically-treated teeth are affected by a higher risk of biomechanical failure than vital teeth (Hansen et al., 1990; Khers et al., 1990; Testori et al., 1993; Tamse et al., 1999; Llena-Puy et al., 2001; Fennis et al., 2002). Most dental fractures are the result of the loss of tooth structure due to carious lesions and/or cavity preparation (Sedgley & Messer, 1992). In particular, the access preparation for endodontic treatment causes the loss of the roof of the pulp chamber: this may account for the relatively high fracture incidence documented in maxillary premolars (Ross, 1980; Salis et al., 1987).

The role of posts in the restoration of endodontically treated teeth Posts are necessary to build up and retain coronal restorations but they do not reinforce dental roots (Caputo & Standlee, 1976). Moreover, some authors asserted that posts may interfere with the mechanical resistance of treated teeth, leading to an increased risk of damage to residual tooth structure (Sornkul & Stannard, 1992; Akkayan & Gulmez, 2002). To date, there is still no agreement regarding which material or technique can be considered ideal for the restoration of endodontically-treated teeth (Hudis & Goldstein, 1986; Creugers et al., 1993; Ortega et al., 2004). Long-term follow-up studies on the post-and-core technique reported highly variable survival rates, demonstrating that root fractures do occur in clinical practice (Sorensen & Martinoff, 1984; Torbjรถrner et al., 1995; Yamada et al., 2004). In order to avoid unrestorable root fractures, posts with biomechanical characteristics similar to those of dentin have been advocated (Akkayan & Gulmez, 2002).

Corornal restorations The interfaces of materials with different moduli of elasticity represent the weak point of a restorative system, as the toughness/stiffness mismatch influences the stress distribution (Assif & Gorfil, 1994; Ausiello et al., 1997). Thus, the strength of endodontically-treated teeth is affected by the material as well as the design of the postand-core (Golberg & Burstone, 1992; Assif & Gorfil, 1994; Libman & Nicholls, 1995; Akkayan & Caniklioglu, 1998). Several techniques and materials have been suggested to increase the fracture resistance of restorative systems but none of them has demonstrated the ability to reduce the incidence of fractures in clinical practice (Hudis & Goldstein, 1986; Creugers et al., 1993; Ortega et al., 2004). Numerous studies have been performed to evaluate the mechanical resistance of endodontically-treated teeth and in particular maxillary premolars (Sornkul & Stannard, 1992; Libman & Nicholls, 1995; Akkayan & Caniklioglu, 1998; Mannocci et al., 1999; Akkayan & Gulmez, 2002; Ortega et al., 2004; Yamada et al., 2004), as a high incidence of fracture for this group of teeth has been reported (Ross, 1980; Salis et al., 1987). Most studies focused on the materials and techniques used to increase the strength of the tooth-restoration complex (Hudis & Goldstein, 1986; Creugers et al., 1993; Ortega et al., 2004). The choice of appropriate restorations should be guided by both physical properties and esthetics (Ferrari et al. 2000a; Creugers et al., 2005a; Creugers et al., 2005b). A restoration has several purposes: to repair the cavity, to strengthen the tooth and to provide an effective seal between the root canal system and the oral environment (Belli et al. 2005). Endodontic treatment should not be considered complete until the coronal restoration has been placed (Wagnild & Mueller 2002). Coronal leakage may lead to bacterial contamination; on the contrary, the longevity of an endodontic treatment is significantly increased by a correct coronal restoration (Tronstad et al. 2000). 8

Restorative materials For building up the coronal restoration after luting a fiber post, all types of composite materials have been proposed (Ferrari et al. 2000a, Ferrari & Scotti 2002, Grandini et al. 2005a, Monticelli et al. 2004, Ferrari et al. 2000b). Microhybrid and flowable resin materials in the self-curing or light-activated formulation are characterized by different strength, stiffness and elasticity: such properties could affect the longevity of the restoration (Ferrari et al. 2000a, Asmussen et al. 1999). Such resin composites can be used together in the so called â&#x20AC;&#x153;sandwich techniqueâ&#x20AC;?, in order to join the mechanical properties of both materials (Croll 2004). Up to date, there is still no agreement regarding the best composite for direct coronal build up of endodontically treated teeth (Ferrari et al. 2000a, Monticelli 2004). Adhesive interfaces of bonded restorations transmit and distribute occlusal forces to the remaining tooth structures homogeneously, potentially strengthening the restored tooth and increasing its resistance to fracture (Hernandez et al. 1994). On the other hand, the interfaces of materials with different moduli of elasticity represent the weak point of a restorative system, as the toughness/stiffness mismatch influences the stress distribution (Assif & Gorfil 1994). The C-factor, defined as the ratio of bonded to unbonded surface areas of cavities, is highly unfavorable in root canals, where it can range from 20 to 200 (Morris et al. 2001). A rigid bond between composite materials and remaining tooth structures could cause stress concentration at the adhesive interfaces (Kemp-Scholte & Davidson 1990). In order to reduce the risk of debonding due to polymerization shrinkage, low viscosity flowable composites with low elastic modulus can be placed between the bonding agent and the composite restoration (Van Meerbeek et al. 1992). Such an intermediate resin layer acts as a stress adsorber (Alhadainy & Abdalla 1996), releasing contraction stresses and improving marginal integrity of the restoration (Belli et al. 2005). Versluis et al. (1996) reported that filling a cavity with a bulk of composite generates less volumetric shrinkage than using the incremental technique within identical cavity shapes. 9

When using composite materials, the resistance of a tooth to cuspal and/or vertical fracture may be affected by several factors, such as the mechanical properties of the restorative system used (Grandini et al. 2005a), the presence of fiber posts (Grandini et al. 2005a) and the cavity shape and dimensions (Hannig et al. 2005). Different studies proved that mesial-occlusal-distal (MOD) cavity preparations reduce the resistance to fracture of root filled teeth, especially maxillary premolars (Belli et al. 2005).

Experimental studies: review of the literature Over the past 10 years, several in vitro (Akkayan & Gulmez 2002, Mannocci et al. 1999, Asmussen et al. 1999, Drummond et al. 1999, Akkayan & Caniklioglu 1998, Akkayan 2004; Fokkinga et al., 2004) and in vivo studies (Ferrari et al. 2000a, Ferrari & Scotti 2002, Ferrari et al. 2000b; Creugers et al., 2005a; Creugers et al., 2005b) were performed to assess the mechanical performances of fiber posts (Asmussen et al. 1999, Grandini et al. 2002, Grandini et al. 2005b). Several studies demonstrated convincing clinical results for the adhesion of fiber posts to root canal substrates (Nakabayashi & Pashley 1998, Ferrari et al. 2000c, Ferrari et al. 2001a, Goracci et al. 2004), the reliability of luting procedures (Ferrari et al. 2001b) and coronal build up (Gateau et al. 1999, Freedman 2001). However, studies on the post-and-core technique reported highly variable survival rates, demonstrating that unfavourable root fractures may occur (Yamada et al. 2004). Teeth restored with bonded CAD/CAM ceramic inlays fractured with a significantly high number of severe fractures as well (Hannig et al. 2005). Recently, several papers supported the use of a direct restoration without placing any post for restoring endodontically treated teeth (Stockton et al. 1998, Baratieri et al. 2000, Assif et al. 2003). Adhesive restorations have been claimed to enhance cuspal stiffness in teeth restored without crown coverage (St.Georges et al. 2003). Moreover, adhesive techniques preserve the maximum amount of sound tooth structure (Mannocci et al. 2002). As to fracture strength and fracture patterns, Krejci et al. (2003) showed no significant differences between teeth restored with and without posts. Moreover, some studies pointed out that mechanical resistance to fracture of endodontically treated teeth could be affected by the presence of posts and the risk of residual tooth structure damage could increase (Akkayan & Gulmez 2002, Akkayan 2004, Strub et al. 2001). In endodontically treated teeth, occlusal loads could be transferred intraradicularly by postand-core restorations, increasing the occurrence of vertical root fractures (Trope et al. 11

1986). On the contrary, other authors noticed that fiber posts reduced the risk of root fractures (Mannocci et al. 1999). Actual consensus in restorative dentistry indicates that decementation or failure of posts is preferable than fracture of residual tooth structure (Akkayan 2004, Ferrari et al. 2000b). The less residual tooth structure the more important the physical properties of post-and-core systems (Christensen 1996).

Mechanical tests Clinicians often experience the clinical fracture of endodontically treated teeth (Assif & Gorfil 1994, Ferrari et al. 2000a, Akkayan & Gulmez 2002). In particular, several studies have been performed to assess the mechanical resistance to fracture of pulpless teeth (Ferrari et al. 2000a, Akkayan & Gulmez 2002, Trope et al. 1996, Salis et al. 1987, Ferrari & Scotti 2002, Mannocci et al 1999, Asmussen et al. 1999, Drummond et al. 1999, Akkayan & Caniklioglu 1998, Akkayan 2004, Ferrari et al. 2000b, Grandini et al. 2002, Grandini et al. 2005a, Yamada et al. 2004). Most studies have tried to identify the best technique and combination of materials to be used to increase the strength of the toothrestoration complex (Ferrari et al. 2000a, Monticelli et al. 2004, Ferrari et al. 2000b, Hannig et al. 2005). Grandini et al. (2005a) demonstrated that restoration of endodontically treated teeth with fiber posts and direct resin composites is a treatment option, that in the short term conserves remaining tooth structure and results in good patient compliance. Reviewing the literature about the loading conditions used to investigate the mechanical resistance to fracture of endodontically treated teeth, some common features can be recognized (Heydecke et al., 2001; Heydecke et al., 2002; Akkayan & Gulmez, 2002; Newman et al., 2003; Yamada et al., 2004; Cheung, 2005; Fokkinga et al., 2005; Hannig et al., 2005; Hayashi et al., 2005; Nagasiri & Chitmongkolsuk, 2005; Schwartz & Fransman, 2005): load type, load application area, load frequency, load jig characteristics, load speed, load intensity, angle of load application, simulation of supporting tissues.

Load type Most experimental studies have been accomplished applying static loads (Akkayan & Gulmez, 2002; Newman et al., 2003; Fokkinga et al., 2005; Hannig et al., 2005; Hayashi et al., 2005; Nagasiri & Chitmongkolsuk, 2005), whereas only few investigations have adopted dinamic loading (Heydecke et al., 2001; Heydecke et al., 2002; Krejci et al., 2003). Such a choice is taken considering the object of the study as well as some practical factors, just like time and costs. The use of static forces permits to simplify the study realization (Akkayan & Gulmez, 2002; Newman et al., 2003; Fokkinga et al., 2005; Hannig et al., 2005; Hayashi et al., 2005; Nagasiri & Chitmongkolsuk, 2005) and requires universal testing machines easier to use and less expensive than fatigue and thermomechanical cycling test workstations (Heydecke et al., 2001; Heydecke et al., 2002; Kreyci et al., 2003; Fokkinga et al., 2005) (fig.1). Particularly, static tests allow to investigate the mechanical properties of a material to evaluate toughness, stiffness or static strength to different kind of loads (Sorrentino et al., 2006a; Sorrentino et al., 2006b; Salameh et al., 2006) whereas dynamic assessments are required to analyze the mechanical performances of materials or restorative systems during function over time (Heydecke et al., 2001; Heydecke et al., 2002). To better simulate the conditions of oral environment, dynamic assessments should be performed maintaining samples dipped in artificial saliva at 37Â°C during tests (Krejci et al., 2003).

Load application area Area of load application has been widely described as one of the paramount factors to achieve reliable laboratory results (Heydecke et al., 2001; Heydecke et al., 2002; Akkayan & Gulmez, 2002; Newman et al., 2003; Fokkinga et al., 2005; Hannig et al., 2005; Hayashi et al., 2005; Nagasiri & Chitmongkolsuk, 2005). Nevertheless, the choice of such an area sistematically differs according to the gnatologic statements of the school the researchers who performed the investigation belong to (Heydecke et al., 2001; Heydecke et al., 2002; Akkayan & Gulmez, 2002; Newman et al., 2003; Fokkinga et al., 2005; Hannig et al., 2005; Hayashi et al., 2005; Nagasiri & Chitmongkolsuk, 2005). Furthermore, load application area changes in relation to tooth anatomy and type of tooth (Heydecke et al., 2001; Heydecke et al., 2002; Akkayan & Gulmez, 2002; Newman et al., 2003; Fokkinga et al., 2005; Hannig et al., 2005; Hayashi et al., 2005; Nagasiri & Chitmongkolsuk, 2005). When anterior teeth are evaluated (Heydecke et al., 2001; Heydecke et al., 2002), specifically maxillary incisors, the load has been always applied on the palatal aspect in an area ranging from 2 to 3 mm from the incisal margin. This was due to the gnatologic role of incisors, designed to bear non axial loads rather than axial forces. The area of load application coincides with the area of incisor guidance during stomatognatic functions. When the fracture resistance of endodontically treated posterior teeth is investigated (Akkayan & Gulmez, 2002; Fokkinga et al., 2005; Hannig et al., 2005; Hayashi et al., 2005; Nagasiri & Chitmongkolsuk, 2005), mainly maxillary premolars, the load has been applied in regions varying from the center of occlusal surface to the supporting cusps (fig.2); in just one case, different regions were used in the same study (Hayashi et al., 2005). Such a choice was based on different gnatologic considerations trying to accomplish the best load simulation of occlusion during function. Some authors (Hannig et al., 2005) pointed out the importance of applying loads in unaltered areas of teeth if 15

possible, to achieve reliable mechanical data avoiding the influence of restorative materials or, even worse, of tooth/restoration interface. The position of the loading site may influence the failure mode, particularly in relation to the position of the post (Fokkinga et al., 2005). The choice of “favourable” or “unfavourable” load application areas and/or angulations would lead to “expected” fracture lines and/or failure patterns (Fokkinga et al., 2005) (fig.3 and fig.4). Furthermore, some authors (Hannig et al., 2005) have proposed to prepare small cavities in unaltered areas of the specimens to place the loading jigs with cyanoacrylate adhesive, even though during oral functions cusp slopes have the possibility of sliding one to each other. Such arbitrary choices could affect the reliability of laboratory results.

Load frequency Such a parameter has been considered only in studies using dynamic loading conditions, just like fatigue tests and thermomechanical cycling analyses (Heydecke et al., 2001; Heydecke et al., 2002). Specimens have been loaded at a frequency of 1.3 Hz, based on a review of the literature on the masticatory function investigating speed of mandibular movements, rate of chewing and forces developed during stomatognatic functions (Bates et al., 1975).

Load jig characteristics Specimens have been loaded with different kinds of jigs, whose width had been chosen according to tooth anatomy, occlusal morphology and/or restoration type, in order to reproduce the mean width of antagonist teeth. The jigs described in the literature were different in shape and material (Heydecke et al., 2001; Heydecke et al., 2002; Akkayan & Gulmez, 2002; Newman et al., 2003; Yamada et al., 2004; Cheung, 2005; Fokkinga et al., 2005; Hannig et al., 2005; Hayashi et al., 2005; Nagasiri & Chitmongkolsuk, 2005; Schwartz & Fransman, 2005). Usually rounded tips have been chosen to homogeneously apply loads (Heydecke et al., 2001; Heydecke et al., 2002; Akkayan & Gulmez, 2002; Newman et al., 2003; Yamada et al., 2004; Cheung, 2005; Fokkinga et al., 2005; Hannig et al., 2005; Hayashi et al., 2005; Nagasiri & Chitmongkolsuk, 2005; Schwartz & Fransman, 2005); on the contrary, sharp tips are well known to develop stress concentration areas. Most authors have chosen ball tips whose diameters ranged from 2.5 to 6 mm based on specimen shape and anatomy (Heydecke et al., 2001; Heydecke et al., 2002; Hannig et al., 2005). Only a few investigators adopted cylindric jigs (Newman et al., 2003). Furthermore, different materials have been used to perform load application, just like ceramics (Heydecke et al., 2001; Heydecke et al., 2002) and steel (Akkayan & Gulmez, 2002; Hannig et al., 2005).

Load speed Speed of load application should simulate oral functions just like chewing. When mechancal investigations are performed, crosshead speed is usually set according to the specifications of the loading machines used to test the specimens. High speed would cause not homogenoeus stress arising in both tooth tissues and restorative materials, whereas low speed would not be representative of the oral functions. On such basis, specimens have been loaded at a speed ranging from 0.5 to 2 mm/min (Heydecke et al., 2001; Heydecke et al., 2002; Newman et al., 2003; Fokkinga et al., 2005; Hannig et al., 2005; Hayashi et al., 2005; Nagasiri & Chitmongkolsuk, 2005).

Load intensity As to static investigations, specimens have always been loaded from 0 Newtons till fracture occurred, in order to record data about the maximum fracture resistance (Newman et al., 2003; Fokkinga et al., 2005; Hannig et al., 2005; Hayashi et al., 2005). On the contrary, in dynamic tests an arbitrary load has been usually chosen based on material characteristics or restoration design, according to literature reviews on forces developed during stomatognatic functions (De Boever et al., 1978; Kampe et al., 1987; Heydecke et al., 2001; Heydecke et al., 2002). In most fatigue analyses an arbitrary load of 30 Newtons has been applied to assess the mechanical resistance to fracture of endodontically treated teeth (De Boever et al., 1978; Kampe et al., 1987; Heydecke et al., 2001; Heydecke et al., 2002). Static tests should be performed first to obtain reliable data about limit and/or mean values of static strength of materials (Goracci et al., 2004; Sorrentino et al., 2006a; Sorrentino et al., 2006b). The load applied by fatigue test machines during dynamic tests should be set on the basis of such results. Thermocycling could be considered to simulate aging of materials (Heydecke et al., 2001; Heydecke et al., 2002).

Angle of load application According to gnatology concepts, teeth are subjected intraorally to different loading condition due to their specific function . As a consequence, posterior teeth have to withstand occlusal and masticatory forces (Akkayan & Gulmez, 2002; Newman et al., 2003; Yamada et al., 2004; Cheung, 2005; Fokkinga et al., 2005; Hannig et al., 2005; Hayashi et al., 2005; Nagasiri & Chitmongkolsuk, 2005; Schwartz & Fransman, 2005) whereas anterior teeth are responsible for tearing and functional guidances (Heydecke et al., 2001; Heydecke et al., 2002). It has been well documented that fracture resistance of teeth depends on the angle of applied load and axial forces are less detrimental than oblique forces (Loney et al., 1995); in particular, shear stresses directly applied to th cusps may lead to mechanical failure of adhesive restorative systems (Uyehara et al., 1999). As a consequence, direction of loads is a crucial factor to the longevity of root filled teeth (Krejci et al., 2003). Moreover, natural variations in tooth morphology may affect the long-term success of restorations (Ortega et al., 2004). Reviewing the literature, experimental load angulation remains a controversial topic (Heydecke et al., 2001; Heydecke et al., 2002; Akkayan & Gulmez, 2002; Newman et al., 2003; Fokkinga et al., 2005; Hannig et al., 2005; Hayashi et al., 2005; Nagasiri & Chitmongkolsuk, 2005). Different forces have been applied to assess mechanical resistance to compression, shear and tension respectively. However, static analyses of resistance to fracture have usually been performed with compressive tests till fractures occurred (Akkayan & Gulmez, 2002; Newman et al., 2003; Fokkinga et al., 2005; Hannig et al., 2005; Hayashi et al., 2005; Nagasiri & Chitmongkolsuk, 2005). When anterior teeth are investigated (Heydecke et al., 2001; Heydecke et al., 2002), specifically maxillary incisors, loads at 130Â° to the longitudinal axis of teeth have

been widely described due to both the average value of interincisal angle and incisor guidance such teeth are devoted to. Conversely, in case of in vitro assessments on posterior teeth (Akkayan & Gulmez, 2002; Newman et al., 2003; Fokkinga et al., 2005; Hannig et al., 2005; Hayashi et al., 2005; Nagasiri & Chitmongkolsuk, 2005), forces with an angulation ranging between 30째 and 45째 to the longitudinal axis of teeth have been proposed (fig.5 and fig.6). Many authors applied loads perpendicular to cusp slopes, noticing mainly fractures of tooth tissues than of restorative materials (Fokkinga et al., 2005; Hannig et al., 2005; Hayashi et al., 2005). On such basis, it is worth discussing the choice of load angulation which, in most cases, appears arbitrary. Most static mechanical fracture tests reported in the literature are characterized by experimental loads on the supporting cusps at 130째-150째 to the longitudinal axis of posterior teeth (Fokkinga et al., 2005; Hannig et al., 2005; Hayashi et al., 2005). Such a condition generates compressive loads perpendicular to the cuspal side (fig.7). On the contrary, during function, the occlusion generates non-axial forces resolved into their vectors along the cuspal side following the parallelogram of forces (fig.8). Such a phenomenon was emphasized by different recent in vitro investigations in which it was decided to load the specimens in a direction parallel to teeth longitudinal axes in order to distribute the stresses more evenly between the residual dental tissues and the restorative material, simulating a physiological occlusion (Sorrentino et al., 2006a; Sorrentino et al., 2006b; Salameh et al., 2006).

Simulation of supporting tissues It has been proved that the use of rigid materials, such as acrylic resin, to embed extracted teeth to be used as specimens in mechanical tests may lead to unreliable load values and may affect the mode of failure of samples (Sirimai et al., 1999; Newman et al., 2003). The fracture strength values of restored teeth without artificial ligament were higher than those with simulated ligaments, as the acrylic resin produced a ferrule effect preventing root fractures (Hayashi et al., 2005). Consequently, different root coatings for the specimens have been proposed to simulate the tooth surrounding anatomcal structures (Mendoza et al., 1997; Sirimai et al., 1999; Newman et al., 2003). In almost all the reviewed in vitro studies, specimens have been embedded in acrylic resin blocks and a silicon index has been used to simulate cortical bone and periodontal ligament respectively (Heydecke et al., 2001; Heydecke et al., 2002; Newman et al., 2003; Hayashi et al., 2005). Self curing silicone has been usually used to simulate the periodontal membrane (Heydecke et al., 2001; Heydecke et al., 2002; Hayashi et al., 2005) (fig.9).

Evidence based selection of dental materials The biomechanical behaviour of biologic structures as well as restorative systems is influenced by several factors interacting one with each other (Zarone et al., 2005; Zarone et al., 2006a). In the oral environment several variables contribute to determine the long term success of restorations. Some of them are dependent on the individual, just like occlusion, load intensity and direction, temperature, moisture, wear, presence of sound tooth structure and quality of supporting tissues, whereas other factors are not controllable, such as structural integrity, microleakage, fatigue and time. Furthermore, teeth and restorative materials are characterized by intrinsic physical characteristics which are responsible for their mechanical performances during functions over time (Van Noort, 2002). Evidence based selection of dental materials requires a close working relationship between dental materials scientists and dental clinicians. Dental materials laboratory investigations are necessary to understand the properties of dental materials, to appreciate correctly the complex relationship between materials properties, design and environment and to develop step by step procedures for clinical validation of experimental models. Furthermore, laboratory and clinical research evidences are necessary to prove the efficacy and safety of rehabilitative treatments (Van Noort, 2002).

The experimental environment Laboratory investigations should be designed as simulations of the clinical conditions to assess physical, mechanical and optical properties of a material. Once experimental protocol have been validated, laboratory tests should be agreed universally as a predictor of clinical performance (Van Noort, 2002). Nevertheless, clinical situations can be only partially simulated by in vitro tests (Fokkinga et al., 2005). Clinical oral functions cause dynamic loading of teeth, in which applied forces, load frequency and direction vary greatly. Laboratory test designs seldom tend to simulate the worst case scenario (Fokkinga et al., 2005) but arbitrary choices may result in large variations of data (Hayashi et al., 2005). Consequently, also due to the large number of secondary variables involved (i.e. tooth type and condition, restorative procedures and materials), it is also impossible to to compare experimental data extrapolated by different in vitro studies (Fokkinga et al., 2005).

Mechanical properties of dental materials When the mechanical properties of a material are discussed, it is not proper to consider the “resistance” or the “strength” but specifications are required. Several kinds of forces can act on a material: tensile, compressive, flexural and/or shear loads; consequently, according to experimental conditions, laboratory tests will investigate tensile, compressive, flexural and shear strength (Van Noort, 2002). For example, as it adhesively link different interfaces, bond strength can be considered as the force the adhesive system oppose to a tensile load. Moreover, the variable “time” could deeply influence the performances of a material; it could be deduced that the same material could obtain different results as it is investigated statically or dynamically. When an experimental test is designed, the first goal should be the cleraly definition of the study variables. On such basis, the proper mechanical analysis can be chosen correctly. Dental tissues and restorative materials are characterized by specific mechanical properties, just like static strength (i.e. tensile strength, compressive strength, flexural strength) and dynamic strength (i.e. fatigue strength). Furthermore, other parameters are paramount to understand the biomechanical behaviour of restorative systems and evaluate their clinical performances, such as stiffness, toughness, resilience and fatigue. Such properties can be defined as follows (Van Noort, 2002): •

Stiffness is the ability of a structure to maintain its shape when acted upon by a load;

•

Toughness is the amount of energy a material can absorb up to the point of fracture;

•

Resilience is the amount of energy a material can absorb without undergoing any plastic deformation;

â&#x20AC;˘

Fatigue is a process of slow crack initiation and propagation under cyclic loads; fatigue failure occurs at lower loads than the tensile strength of a material with no outward signs of problems when subjected to repeated stress cycles.

When a restorative system is loaded, it is able to adsorb the applied forces generating peculiar stress and strain distributions. The evaluation of such patterns, for example using the Finite Element Analysis, could be a reliable predictive parameter to forecast areas under risk of possible mechanical failures (Zarone et al., 2005; Zarone et al., 2006a) (fig.10). Stress can be defined as the force per unit cross-sectional area, that is acting on a material, whereas strain is the fractional change in the dimensions caused by the force (Van Noort, 2002). If correctly analysed, data extrapolated from stress-strain curves are paramount to evaluate the clinical performances of restorative materials (Zarone et al., 2005; Zarone et al., 2006a).

Step by step procedure to accomplish laboratory and clinical evaluations of dental materials and restorative systems A correct evaluation of the mechanical and clinical performances of dental materials and restorative systems should pass through different steps, from the simulation of clinical conditions to the intraoral confirmation of the effectiveness of restorations (Zarone et al., 2005; Zarone et al., 2006a) (fig.11). Computerized mathematical models allow to create bioengineering virtual simulations of the oral environment, just like Finite Element Analysis (FEA). Tooth morphologies can be acquired by optical or laser scanning procedures; then, mechanical properties of dental tissues and restorative materials can be simulated using threedimensional solid elements (Zarone et al., 2005; Zarone et al., 2006a) (fig.12). Nowadays, it is possible to mechanically characterize not only enamel and dentin but also bone tissues, periodontal ligament and adhesive interfaces (Zarone et al., 2005; Zarone et al., 2006a) (fig.13). Different loading conditions can be evaluated with FEA in order to evaluate stress and strain distributions within teeth and restorations and hypothesize predictive areas under risk of clinical failure (Zarone et al., 2005; Zarone et al., 2006a). Data extrapolated from virtual simulations have to be confirmed by in vitro results. Consequently, laboratory investigations have to be used as validation of the simulated mathematical models (Zarone et al., 2005; Zarone et al., 2006a). Mechanical tests are necessary to analyze the characteristics of dental tissues and restorative materials alone or of their interaction when assembled in restorative systems (Sorrentino et al., 2006a; Sorrentino et al., 2006b; Salameh et al., 2006). Static tests should be performed to assess the mechanical properties of materials under different loading conditions (Sorrentino et al., 2006a; Sorrentino et al., 2006b; Salameh et al., 2006). Differently, dynamic investigations are useful to simulate the mechanical performances of materials or restorative systems 28

during function (Heydecke et al., 2001; Heydecke et al., 2002). Restored teeth should be tested for static fracture as well as for fatigue resistance, since they have to withstand both types of load in the oral environment (Hatashi t al., 2005). Mechanical static tests (i.e. fracture test, microtensile bond strength test, push out test, pull out test) should be used to identify the limit and/or mean values of compressive, tensile and shear strength of materials (Goracci et al., 2004; Sorrentino et al., 2006a; Sorrentino et al., 2006b). Then, such data should be used to set the load applied by fatigue test machines during dynamic tests. Furthermore, when dynamic investigations are performed, specimens can be first subjected to thermocycling to simulate aging of materials, reproducing clinical functions over time (Heydecke et al., 2001; Heydecke et al., 2002). Once FEA, static and dynamic results have been correlated, in vivo invstigations are necessary to validate laboratory data. Clinical trials should represent the final step to assess the safety and effectiveness of any rehabilitative potocol (Ferrari et al., 2000a; Ferrari et al., 2000b; Zarone et al., 2005; Zarone et al., 2006a).

Protocol to perform mechanical static tests The present paragraph aims at describing the ideal conditions to investigate the mechanical resistance to fracture of natural or restored teeth. The same procedures fit both to sound and endodontically treated teeth (Sorrentino et al., 2006a; Sorrentino et al., 2006b; Salameh et al., 2006). Inclusion and exclusion criteria should be clearly stated before sampling, for example selecting tooth type, absence of caries or previous restorations, crown tissues up to a certain amount and absence of fracture lines. Furthermore, the choice of similar crown and root sizes is paramount to achieve clinically relevant results (Hayashi et al., 2005; Sorrentino et al., 2006a; Sorrentino et al., 2006b; Salameh et al., 2006). Differently from in vitro studies on dental materials, it is not possible to replicate restorations since biologic substrates just like teeth are different one from each other. In order to reproduce experimental conditions as tidily as possible, a statistical analysis of mesial-distal and buccal-palatal widths as well as of root length should be performed to notice the homogeneity of the samples (Hayashi et al., 2005; Sorrentino et al., 2006a; Sorrentino et al., 2006b; Salameh et al., 2006). It has been widely described how testing human teeth results in large standard deviations (Krejci et al., 2003; Fokkinga et al., 2005; Hayashi et al., 2005). It would be desirable to use sound teeth as control group if absolute values of mechanical resistance are to be assessed (Sorrentino et al., 2006a). Conversely, if a specific restorative system has to be investigated, the control group should be chosen according to study variables, for example using a previously investigated restorative material or design (Sorrentino et al., 2006b). As to study groups, random distributions of specimens has to be accomplished (Sorrentino et al., 2006a; Sorrentino et al., 2006b; Salameh et al., 2006). In order to make up for anatomic variability, experimental groups should be made up of 10 specimens each at least (Sorrentino et al., 2006a; Sorrentino et 30

al., 2006b; Salameh et al., 2006). Teeth should be used within 3 months following extraction (Hayashi et al., 2005). Sterile specimens should be stored in water at room temperature until further processing, using a stove if necessary (Sorrentino et al., 2006a; Sorrentino et al., 2006b; Salameh et al., 2006). The oral environment is characterized by approximately 90% humidity (Musanje & Darvell, 2003); consequently, a moisture-controlled machine should be used to store the specimens. If such a machine is not available, teeth should not be dipped in water but enveloped in wet sheet of paper. Crown, cavity and/or endodontic preparation should be performed strictly following clinical protocols, manufacturersâ&#x20AC;&#x2122; instructions and/or standardized criteria, stated according to the needs of the investigation and extrapolating data from literature (Sorrentino et al., 2006a; Sorrentino et al., 2006b; Salameh et al., 2006). Tooth preparations should be performed using diamond rotary burs under constant irrigation and cooling in order not to overheat the samples (Sorrentino et al., 2006a; Sorrentino et al., 2006b; Salameh et al., 2006). If necessary, parallelometer drilling devices can be used to maintain preparations parallel or perpendicular to reference axes, such as tooth longitudinal axis (Hayashi et al., 2005). Thickness preparation, tooth residual structure as well as dental materials increments should be checked with digital callipers (Ferrari et al., 1992; Sorrentino et al., 2006a; Sorrentino et al., 2006b; Zarone et al., 2006b). If these are not available or their use is not possible because of cluttered spaces, silicon indexes or hand and/or rotary instruments whose dimensions are well known (i.e. periodontal probes, burs) can be used (Ferrari et al., 1992). Anatomic references can be used to make such measurements, just like the cemento-enamel junction (CEJ) (Sorrentino et al., 2006a; Sorrentino et al., 2006b). After preparation, specimens should be stored in disinfectant solutions (i.e. chloramine 1%, thymol 0.02%) for 24 hours at least to eliminate eventual impurities (Strawn et al., 1996; Francescut et al., 2006). 31

Once completed the sample preparation, mechanical tests should be performed according to the needs of the investigation. Static or dynamic loads have to be applied on the basis of the specific mechanical property that is the object of the study (Sorrentino et al., 2006a; Sorrentino et al., 2006b; Salameh et al., 2006). As previously reported, the present paragraph aims at describing an in vitro investigation on the compressive strength of restored maxillary premolars. Before performing mechanical tests, specimens have to be embedded in selfpolymerizing acrylic resin; stainless steel cylinder can be used as mould for resin (Sorrentino et al., 2006a; Sorrentino et al., 2006b; Salameh et al., 2006). If present, the CEJ should be situated approximately 1.5 mm above the level acrylic resin to simulate bone crest whereas the buccal side of the root should be located about 1 mm from the outer surface of the acrylic cylinder to simulate buccal bone crest (Fokkinga et al., 2005; Sorrentino et al., 2006a; Sorrentino et al., 2006b; Salameh et al., 2006). During such procedures, samples have to be continuously irrigated with water to avoid overheating due to resin polymerization (Sorrentino et al., 2006a; Sorrentino et al., 2006b). Then, specimens have to be stored at 90% humidity as previously described (Musanje & Darvell, 2003) for 48 hours at least to allow resin to complete polymerization (Fokkinga et al., 2005). To simulate the periodontal ligament, roots have to be covered with a 200 mÎź thick layer of polyvinylsiloxane impression material, as its width and modulus of elasticity are similar to those of natural periodontal membranes (Coolidge, 1937; Mendoza et al., 1997; Sirimai et al., 1999; Newman et al., 2003; Hayashi et al., 2005). A universal testing machine has to be used to perform static mechanical tests. Such machine should be linked to a personal computer provided with a dedicated software to record fracture values and load-deformation curves (Sorrentino et al., 2006a; Sorrentino et al., 2006b; Salameh et al., 2006). Samples should be fixed in the metal holder of the 32

testing machine with the longitudinal axis parallel to load direction, so that the teeth themselves distribute the applied force along cuspal slopes (Sorrentino et al., 2006a; Sorrentino et al., 2006b; Salameh et al., 2006). Any angulation of the axis of the specimen would be arbitrary, influencing the failure mode (Heydecke et al., 2001; Heydecke et al., 2002; Newman et al., 2003; Fokkinga et al., 2005; Hannig et al., 2005; Hayashi et al., 2005). Stainless steel rods with rounded tips have to be used to load specimens (Hannig et al., 2005; Sorrentino et al., 2006a; Sorrentino et al., 2006b; Salameh et al., 2006). The width of such tips has to be chosen considering the occlusal morphology of teeth. For example, when maxillary premolars are investigated, a stainless steel rod with tip diameter of 1 mm can be considered optimal as it reproduces the mean width of antagonist teeth present in the simulated clinical situation (Sorrentino et al., 2006a; Sorrentino et al., 2006b; Salameh et al., 2006). A crosshead speed of 1 mm/min would be desirable Sorrentino et al., 2006a; Sorrentino et al., 2006b); anyway, it could range according to values reported in literature (0.5-2 mm/min) (Heydecke et al., 2001; Heydecke et al., 2002; Newman et al., 2003; Fokkinga et al., 2005; Hannig et al., 2005; Hayashi et al., 2005). If static mechanical tests have to be performed, specimens have to be subjecetd to a constant load from 0 Newtons (N) until fracture occurs (Newman et al., 2003; Fokkinga et al., 2005; Hannig et al., 2005; Hayashi et al., 2005; Sorrentino et al., 2006a; Sorrentino et al., 2006b; Salameh et al., 2006). Conversely, if fatigue tests are required, the load intensity should be chosen on the basis of evidence based data about the mean value of compressive and/or tensile strength of both the restorative materials and the type of tooth (Bates et al., 1975; Heydecke et al., 2001; Heydecke et al., 2002). As previously described for the shape of loading tips, the loading area should be chosen according to the occlusal morphology of the specimen (Sorrentino et al., 2006a; 33

Sorrentino et al., 2006b; Salameh et al., 2006). On the basis of gnatologic concepts, teeth should be loaded on the slope of the supporting cusp, simulating a physiological occlusion; it is well known that during function antagonist cusps do not reach the central fissure of the occlusal surface of teeth (Sorrentino et al., 2006a; Sorrentino et al., 2006b; Salameh et al., 2006). If mechanical resistance to fracture is investigated, specimens should be loaded until fracture occurs (Newman et al., 2003; Fokkinga et al., 2005; Hannig et al., 2005; Hayashi et al., 2005; Sorrentino et al., 2006a; Sorrentino et al., 2006b; Salameh et al., 2006). Two independent observers should evaluate macroscopically the characteristics of fractures by visual inspection and disagreements should be resolved by discussion or with the arbitrate of a third calibrated observer (Hayashi et al., 2005). Fracture lines can be highlighted with ink perfusion and a stereomicroscope can be used to evaluate the fracture pattern Sorrentino et al., 2006a; Sorrentino et al., 2006b; Salameh et al., 2006. If fracture lines move towards the part of root embedded in acrylic resin, digital radiographs can be used to detect internal crack propagation (Hayashi et al., 2005). Several classifications for mode of failure have been proposed (Heydecke et al., 2001; Heydecke et al., 2002; Akkayan & Gulmez, 2002; Fokkinga et al., 2005; Hayashi et al., 2005); the following one considers the mechanical cause of fracture (i.e. adhesive vs cohesive) as well as its clinical implications (i.e. restorable vs unrestorable – fig.14) (Sorrentino et al., 2006a; Sorrentino et al., 2006b): •

Adhesive fracture: the fracture of an adhesive bond/interface between tooth tissues and restorative materials;

•

Cohesive fracture: the fracture inside the bulk of tooth tissues or restorative materials with no exposure of any adhesive layer;

•

Restorable fracture: a fracture interesting only the coronal part of a tooth;

•

Unrestorable fracture: a fracture interesting the root of a tooth. 34

References 1. Akkayan B. An in vitro study evaluating the effect of ferrule length on fracture resistance of endodontically treated teeth restored with fiber-reinforced and zirconia dowel systems. J Prosthet Dent 2004; 92: 155-162. 2. Akkayan B, Caniklioglu B. Resistance to fracture of crowned teeth restored with different post systems. Eur J Prosthod Restor Dent 1998; 6: 13-18. 3. Akkayan B, Gulmez T. Resistance to fracture of endodontically treated teeth restored with different post systems. J Prosthet Dent 2002; 87: 431-437. 4. Alhadainy HA, Abdalla AI. 2-year clinical evaluation of dentin bonding systems. Am J Dent 1996; 9: 77-79. 5. Asmussen E, Peutzfeldt A, Heitmann T. Stiffness, elastic limit and strength of newer types of endodontic posts. J Dent 1999; 27: 275-278. 6. Assif D, Gorfil C. Biomechanical considerations in restoring endodontically treated teeth. J Prosthet Dent 1994; 71: 565-567. 7. Assif D, Nissan J, Gafni Y, Gordon M. Assessment of the resistance to fracture of endodontically treated molars restored with amalgam. J Prosthet Dent 2003; 89: 462-465. 8. Ausiello P, Gee AJ, Rengo S, Davidson CL. Fracture resistance of endodonticallytreated premolars adhesively restored. Am J Dent 1997; 10: 237-241. 9. Baratieri LN, De Andrada MA, Arcari GM, Ritter AV. Influence of post placement in the fracture resistance of endodontically treated incisors veneered with direct composite. J Prosthet Dent 2000; 84: 180-184. 10. Belli S, Erdemir A, Ozcopur M, Eskitascioglu G. The effect of fibre insertion on fracture resistance of root filled molar teeth with MOD preparations restored with composite. Int Endod J 2005; 38: 73-80.

11. Caputo AA, Standlee JP. Pins and posts – why, when and how. Dent Clin North Am 1976; 20: 299-311. 12. Cheung W. A review of the management of endodontically treated teeth: post, core and the final restoration. J Am Dent Assoc 2005; 136: 611-619. 13. Christensen GJ. When to use fillers, build-ups or posts and cores. J Am Dent Assoc 1996; 127: 1397-1398. 14. Creugers NH, Mentink AG, Kayser AF. An analysis of durability data on post and core restorations. J Dent 1993; 21: 281-284. 15. Creugers NHJ, Mentink AGB, Fokkinga WA, Kreulen CM. 5-year follow-up of a prospective clinical study on various types of core restorations. Int J Prosthodont 2005a; 18: 34-39. 16. Creugers NHJ, Kreulen CM, Fokkinga WA, Mentink AGB. A 5-year prospective clinical study on core restorations without covering crowns. Int J Prosthodont 2005b; 18: 40-41. 17. Croll TP. The “sandwich” technique. J Esthet Restor Dent 2004; 16: 210-212. 18. Drummond JL, Toepke TR, King TJ. Thermal and cyclic loading of endodontic posts. Eur J Oral Sci1999; 107: 220-224. 19. Fennis WMM, Kuijs RH, Kreulen CM, Roeters FJ, Creugers NH, Burgersdijk RC. A survey of cusp fractures in a population of general dental practices. Int J Prosthodont 2002; 15: 559-563. 20. Ferrari M, Mannocci F, Vichi A, Cagidiaco MC, Mjor IA. Bonding to root canal: structural characteristics of the substrate. Am J Dent 2000c; 13: 255-260. 21. Ferrari M, Scotti R. Fiber posts: characteristics and clinical applications, Milan, Italy: Masson, 2002. 22. Ferrari M, Vichi A, Garcia-Godoy F. Clinical evaluation of fiber-reinforced epoxy resin posts and cast post and cores. Am J Dent 2000b; 13: 15B-18B. 36

23. Ferrari M, Vichi A, Grandini S. Efficacy of different adhesive techniques on bonding to root canal walls: a SEM investigation. Dent Mater 2001a; 17: 422-429. 24. Ferrari M, Vichi A, Grandini S, Goracci C. Efficacy of a self-curing adhesive-resin cement system on luting glass-fiber posts into root canals: a SEM investigation. Int J Prosthodont 2001b; 14: 543-549. 25. Ferrari M, Vichi A, Mannocci F, Mason PN. Retrospective study of the clinical performance of fiber posts. Am J Dent 2000a; 13: 9B-13B. 26. Fokkinga WA, Kreulen CM, Vallittu PK, Creugers NHJ. A structured analysis of in vitro failure loads and failure modes of fiber, metal and ceramic post-and-core systems. Int J Prosthodont 2004; 17: 476-482. 27. Fokkinga WA, Le Bell AM, Kreulen CM, Lassila LVJ, Vallittu PK, Creugers NHJ. Ex vivo fracture resistance of direct resin composite complete crowns with and without posts on maxillary premolars. Int Endod J 2005; 38: 230-237. 28. Freedman GA. Esthetic post-and-core treatment. Dent Clin North Am 2001; 45: 103-116. 29. Gateau P, Sabek M, Dailey B. Fatigue testing and microscopic evaluation of post and core restorations under artificial crowns. J Prosthet Dent 1999; 82: 341-347. 30. Goldberg AJ, Burstone CJ. The use of continuous fibre reinforcement in dentistry. Dent Mater 1992; 8: 197-202. 31. Goracci C, Tavares AU, Fabianelli A, Monticelli F, Raffaelli O, Cardoso PC, Tay FR, Ferrari M. The adhesion between fiber posts and root canal walls: comparison between microtensile and push-out bond strength measurements. Eur J Oral Sci 2004; 112: 353-361. 32. Grandini S, Balleri P, Ferrari M. Scanning electron microscopic investigation of the surface of fiber posts after cutting. J Endod 2002; 28: 610-612.

33. Grandini S, Goracci C, Tay FR, Grandini R, Ferrari M. Clinical evaluation of the use of fiber posts and direct resin restorations for endodontically treated teeth. Int J Prosthodont 2005a; 18: 399-404. 34. Grandini S, Goracci C, Monticelli F, Tay FR, Ferrari M. Fatigue resistance and structural characteristics of fiber posts: three-point bending test and SEM evaluation. Dent Mater 2005b; 21: 75-82. 35. Hannig C, Westphal C, Becker K, Attin T. Fracture resistance of endodontically treated maxillary premolars restored with CAD/CAM ceramic inlays. J Prosthet Dent 2005; 94: 342-349. 36. Hansen EK, Asmussen E, Christiansen NC. In vivo fractures of endodontically treated posterior teeth restored with amalgam. Endod Dent Traumatol 1990; 10: 4955. 37. Hayashi M, Takahashi Y, Imazato S, Ebisu S. Fracture resistance of pulpless teeth restored with post-cores and crowns. Dent Mater 2005, in press. 38. Hernandez R, Bader S, Boston D, Trope M. Resistance to fracture of endodontically treated premolars restored with new generation dentine bonding systems. Int Endod J 194; 27: 281-284. 39. Heydecke G, Butz F, Strub JR. Fracture strength and survival rate of endodontically treated maxillary incisors with approximal cavities after restoration with different post and core systems: an in vitro study. J Dent 2001; 29: 427-433. 40. Heydecke G, Butz F, Hussein A, Strub JR. Fracture strength after dynamic loading of endodontically treated teeth restored with different post and core systems. J Prosthet Dent 2002; 87: 438-445. 41. Hudis SI, Goldstein GR. Restoration of endodontically treated teeth: A review of the literature. J Prosthet Dent 1986; 55: 33-38.

42. Kemp-Scholte CM, Davidson CL. Marginal integrity related to bond strength and strain capacity of composite resin restorative systems. J Prosthet Dent 1990; 64: 658-664. 43. Khers SC, Carpenter CW, Vetter JD, Staley RN. Anatomy of cusps of posterior teeth and their fracture potential. J Prosthet Dent 1990; 64: 139-147. 44. Krejci I, Duc O, Dietschi D, de Campos E. Marginal adaptation, retention and fracture resistance of adhesive composite restorations on devital teeth with and without posts. Oper Dent 203; 28: 127-135. 45. Libman WJ, Nicholls JI. Load fatigue of teeth restored with cast posts and cores and complete crowns. Int J Prosthodont 1995; 8: 155-161. 46. Llena-Puy MC, Forner-Navarro L, Barbero-Navarro I. Vertical root fracture in endodontically treated teeth: A review of 25 cases. Oral Sur Oral Med Oral Pathol Oral Radiol Endod 2001; 92: 553-555. 47. Loney RW, Moulding MB, Ritsco RG. The effect of load angulation on fracture resistance of teeth restored with cast post and cores and crowns. Int J Prosthodont 195; 8: 247-251. 48. Mannocci F, Ferrari M, Watson TF. Intermittent loading of teeth restored using quartz fibre, carbon-quartz fibre and zirconium dioxide ceramic root canal posts. J Adhes Dent 1999; 1: 153-158. 49. Mannocci F, Bertelli E, Sherriff M, Watson TF, Ford TR. Three-year clinical comparison of survival of endodontically treated teeth restored with either full cast coverage or with direct composite restoration. J Prosthet Dent 2002; 88: 297-301. 50. Monticelli F, Goracci C, Ferrari M. Micromorphology of the fiber post-resin core unit: a scanning electron microscopy evaluation. Dent Mater 2004; 20: 176-183.

51. Morris MD, Lee KW, Agee KA, Bouillaguet S, Pashley DH. Effects of sodium hypochlorite and RC-prep on bond strengths of resin cement to endodontic surfaces. J Endod 2001; 27: 753-757. 52. Musanje L, Darvell BW. Aspects of water sorption from the air, water and artificial saliva in resin composite restorative materials. Dent Mater 2003; 19: 414-422. 53. Nagasiri R, Chitmongkolsuk S. Long-term survival of endodontically treated molars without crown coverage: a retrospective cohort study. J Prosthet Dent 2005; 93: 164-170. 54. Nakabayashi N, Pashley DH. Hybridization of dental hard tissue. Berlin, Germany: Quintessence, 1998. 55. Newman MP, Yaman P, Dennison J, Rafter M, Billy E. Fracture resistance of endodontically treated teeth restored with composite posts. J Prosthet Dent 2003; 89: 360-367. 56. Ngoh EC, Pashley DH, Loushine RJ, Weller RN, Kimbrough F. Effects of eugenol on resin bond strengths to root canal dentin. J Endod 2001; 27: 411-414. 57. Ortega VL, Pegoraro LF, Conti PCR, Do Valle AL, Bonfante G. Evaluation of fracture resistance of endodontically treated maxillary premolars, restored with ceromer or heat-pressed ceramic inlays and fixed with dual-resin cements. J Oral Rehabil 2004; 31: 393-397. 58. Ross IF. Fracture susceptibility of endodontically treated teeth. J Endod 1980; 6: 560-565. 59. Salameh Z, Sorrentino R, Papacchini F, Ounsi HF, Tashkandi E, Goracci C, Ferrari M. Fracture resistance and failure patterns of endodontically treated mandibular molars restored using resin composite with or without translucent glass fiber posts. J Endod 2006; 32: 752-755.

60. Salis SG, Hood JAA, Stokes ANS, Kirk EEJ. Pattern s of indirect fracture in intact and restored human premolar teeth. Endod Dent Traumatol 1987; 3: 10-14. 61. Schwartz RS, Fransman R. Adhesive dentistry and endodontics: materials, clinical strategies, and procedures for restoration of access cavities: a review. J Endod 2005; 31: 151-165. 62. Sedgley CM, Messer HH. Are endodontically treated teeth more brittle? J Endod 1992; 18: 332-335. 63. Sorensen JA, Martinoff JT. Intracoronal reinforcement and coronal coverage: A study of endodontically treated teeth. J Prosthet Dent 1984; 51: 780-784. 64. Sornkul E, Stannard JG. Strength of roots before and after endodontic treatment and restoration. J Endod 1992; 18: 440-443. 65. Sorrentino R, Monticelli F, Goracci C, Zarone F, Tay FR, Garcia-Godoy F, Ferrari M. Effect of post retained composite restorations on the fracture resistance of endodontically-treated teeth related to the amount of coronal residual structure. Am J Dent 2006a; in press. 66. Sorrentino R, Salameh Z, Zarone F, Tay FR, Ferrari M. Effect of post retained composite restorations of mesial-occlusal-distal cavities on the fracture resistance of endodontically-treated teeth. J Adhes Dent 2006b; in press. 67. St.Georges AJ, Sturdevant JR, Swift EJ Jr, Thompson JY. Fracture resistance of prepared teeth restored with bonded inlay restorations. J Prosthet Dent 2003; 89: 551-557. 68. Stockton L, Lavelle CL, Suzuki M. Are posts mandatory for the restoration of endodontically treated teeth? Endod Dent Traumatol 1998; 14: 59-63. 69. Strub JR, Pontius O, Koutayas S. Survival rate and fracture strength of incisors restored with different post and core systems after exposure in the artificial mouth. J Oral Rehabil 2001; 28: 120-124. 41

70. Tamse A, Fuss Z, Lustig J, Kaplavi J. An evaluation of endodontically treated vertically fractured teeth. J Endod 1999; 7: 506-508. 71. Testori T, Badino M, Castagnola M. Vertical root fractures in endodontically treated teeth: A clinical survey of 36 cases. J Endod 1993; 19: 87-91. 72. TorbjĂśrner A, Karlsson S, Ă&#x2013;dman PA. Survival rate and failure characteristics for two post designs. J Prosthet Dent 1995; 73: 439-444. 73. Tronstad L, Asbjornsen K, Doving L, Pedersen I, Eriksen HM. Influence of coronal restorations on the periapical health of endodontically treated teeth. Endod Dent Traumatol 2000; 16: 218-221. 74. Trope M, Langer I, Maltz D, Tronstad L. Resistance to fracture of restored endodontically treated premolars. Endod Dent Traumatol 1986; 2: 35-38. 75. Uyehara MY, Davis RD, Overton JD. Cuspal reinforcement in endodontically treated molars. Oper Dent 1999; 24: 364-370. 76. Van Meerbeek B, Lambrechts P, Inokoshi S, Braem M, Vanherle G. Factors affecting adhesion to mineralized tissues. Oper Dent 1992; Suppl 5: 111-124. 77. Van Noort R. Introduction to dental materials, 2nd Edition. Mosby Ltd, 2002. 78. Versluis A, Douglas W, Cross M, Sakaguchi RL. Does an incremental filling technique reduce polymerization shrinkage stresses? J Dent Res 1996; 75: 871878. 79. Wagnild GW, Mueller KI. Restoration of the endodontically treated tooth. In: Cohen S, Burns RC, eds. Pathways of the Pulp, 8th edn. St Louis, MO, USA: Mosby Inc., 2002; 765-795. 80. Yamada Y, Tsubota Y, Fukushima S. Effect of restoration method on fracture resistance of endodontically treated maxillary premolars. Int J Prosthodont 2004; 17: 94-98.

81. Zarone F, Apicella D, Sorrentino R, Ferro V, Aversa R, Apicella A. Influence of tooth preparation design on the stress distribution in maxillary central incisors restored by means of alumina porcelain veneers: a 3D-finite element analysis. Dent Mater 2005; 21: 1178-1188. 82. Zarone F, Sorrentino R, Apicella D, Valentino B, Ferrari M, Aversa R, Apicella A. Evaluation of the biomechanical behavior of maxillary central incisors restored by means of endocrowns compared to a natural tooth: A 3D static linear finite elements analysis. Dent Mater 2006a Jan 10; [Epub ahead of print]. 83. Zarone F, Epifania E, Leone G, Sorrentino R, Ferrari M. Dynamometric assessment of the mechanical resistance of porcelain veneers related to tooth preparation: a comparison between two techniques. J Prosthet Dent 2006b; 95: 354-363.

Fig. 1 â&#x20AC;&#x201C; Endurance fatigue test machine (Si-Plan Electronics Research LTD).

Fig. 2 â&#x20AC;&#x201C; Load applied on the palatal cusp of a maxillary first premolar, 2 mm from the tip of the cusp towards the central fossa, in order to simulate an occlusal load. 44

Fig. 3 – Expected restorable fracture (in red) under a 45° angulated load to the longitudinal axis of the tooth.

Fig. 4 – Expected unrestorable fracture (in red) under a 30° angulated load to the longitudinal axis of the tooth. 45

Fig. 5 – 45° angulated load to the longitudinal axis of the tooth.

Fig. 6 – 30° angulated load to the longitudinal axis of the tooth.

Fig. 7 â&#x20AC;&#x201C; Arbitrary compressive load perpendicular to the cuspal slope (yellow arrow: main force; blue arrow: resultant force perpendicular to the cuspal slope; red dashed line: translated longitudinal axis of the tooth).

Fig. 8 â&#x20AC;&#x201C; Vectorial resolution of forces along cuspal slope after the application of a load parallel to the longitudinal axis of the tooth (yellow arrow: main force; blue arrow: main force component perpendicular to the cuspal slope; red arrow: main force component parallel to the cupal slope). 47

Fig. 9 – Silicone simulated periodontal ligament around a tooth embedded in acrylic resin block.

Fig. 10 – Three dimensional finite element stress (MPa) strain (ε) distributions after load application.

Fig. 11 â&#x20AC;&#x201C; Step by step experimental procedure.

Fig. 12 â&#x20AC;&#x201C; Three dimensional finite element model of a maxillary central incisor.

Fig. 13 â&#x20AC;&#x201C; Three dimensional finite element model of a tooth and surrounding supporting tissues.

Fig. 14 â&#x20AC;&#x201C; Restorable (left side) and unrestorable (right side) fractures after static mechanical compressive test. 50

CHAPTER II

Investigating the mechanical resistance to fracture of endodontically-treated premolars simulating different restorative systems.

Effect of Post Retained Composite Restorations on the Fracture Resistance of Endodontically-Treated Teeth Related to the Amount of Coronal Residual Structure.

Sorrentino R., Monticelli F., Goracci C., Zarone F., Tay F.R., Garcia-Godoy F., Ferrari M. â&#x20AC;&#x201C; American Journal of Dentistry, 2006 in press.

ABSTRACT Purpose. To compare the fracture resistance and failure patterns of endodontically-treated teeth with a progressively reduced number of residual walls restored using resin composites, with or without translucent glass fiber posts. Methods. Ninety extracted human single rooted maxillary premolars were used. After endodontic treatment, the following groups were created: Group 1 (control group): endodontically-treated single rooted maxillary premolars with four residual walls; Group 2: three residual walls; Group 3: two residual walls; Group 4: one residual wall, and Group 5: no residual wall. Groups 2-5 were each divided into two subgroups: subgroups “a” were restored with resin composites, while subgroups “b” were restored with translucent glass fiber posts and resin composites. Static fracture resistance tests were performed. One-way ANOVA, Tukey post hoc test and t-test (P=0.05) were conducted. Results. The mean failure loads (N) were 502.4±152.5 (Gr.1), 416.4±122.2 (Group 2a), 423.0±103.3 (Gr.2b) 422.1±138.9 (Gr.3a) 513.2±121.7 (Gr.3b), 488.7±153.7 (Gr.4a) 573.4±169.2 (Gr.4b), 856.7±112.2 (Gr.5a) and 649.5±163.5 (Gr.5b), respectively. The samples restored with fiber posts exhibited predominantly restorable fractures. The number of residual cavity walls influenced the mechanical resistance of endodonticallytreated teeth. Clinical significance. The results of the present study would allow clinicians to make an informed choice from among available materials to restore endodontically treated teeth, in order to decide in which clinical situations it would be advisable to use fiber posts.

Introduction Endodontically-treated teeth are affected by a higher risk of biomechanical failure than vital teeth

1-6

. Most dental fractures are the result of the loss of tooth structure due to

carious lesions and/or cavity preparation 7. In particular, the access preparation for endodontic treatment causes the loss of the roof of the pulp chamber: this may account for the relatively high fracture incidence documented in maxillary premolars 8,9. Posts are necessary to build up and retain coronal restorations but they do not reinforce dental roots

. Moreover, some authors asserted that posts may interfere with

the mechanical resistance of treated teeth, leading to an increased risk of damage to residual tooth structure

11,12

. To date, there is still no agreement regarding which material

or technique can be considered ideal for the restoration of endodontically-treated teeth 15

13-

. Long-term follow-up studies on the post-and-core technique reported highly variable

survival rates, demonstrating that root fractures do occur in clinical practice

16-18

. In order

to avoid unrestorable root fractures, posts with biomechanical characteristics similar to those of dentin have been advocated 12. The interfaces of materials with different moduli of elasticity represent the weak point of a restorative system, as the toughness/stiffness mismatch influences the stress distribution

19,20

. Thus, the strength of endodontically-treated teeth is affected by the

material as well as the design of the post-and-core

20-22

. Several techniques and materials

have been suggested to increase the fracture resistance of restorative systems but none of them has demonstrated the ability to reduce the incidence of fractures in clinical practice 13-15

. The first posts to be introduced in clinical practice were gold post and cores. Later

on, prefabricated metal posts were proposed, the physical properties of which were very different from those of the core material and dentin. These posts generated high stresses, often leading to unrestorable root fractures

. In order to avoid such problems, metal-free

posts with mechanical characteristics similar to those of dental tisues were developed Subsequently, fiber-reinforced post systems were introduced

23,24

. They were initially

strengthened by incorporating glass fibers in an epoxy resin matrix

. More recently,

translucent quartz fiber posts have been introduced, in order to achieve optimal aesthetic properties. These systems allow the diffusion of the polymerizing light through the post during cementation, reaching the apical third of the root canal 24. An in vitro study25 pointed out that zirconia posts showed a higher fracture resistance than both titanium and fibrereinforced systems. Conversely, fiber-reinforced posts greatly reduced the risk of root fractures 24. The present study compared the fracture resistance and failure patterns of endodontically-treated single rooted maxillary premolars with a progressively reduced number of residual cavity walls, that were restored using resin composites with or without translucent glass fiber posts. The null hypothesis tested was that there was no association between the fracture resistance of endodontically-treated single rooted maxillary premolars restored by means of resin composite with or without glass fiber -reinforced posts, and the number of residual cavity walls remaining in teeth that require these restorations. Materials and Methods Ninety human single rooted maxillary premolars, extracted for orthodontic or periodontal reasons, were selected for the study. Teeth with caries and previous restorations were excluded. Dental plaque, calculus and periodontal tissues were removed. The teeth were stored in 0.9% saline solution at 37Â°C until they were used for the mechanical tests. Canal morphology was verified from standardized apical radiographs (70 kV and 0.08 seconds) both in the mesial-distal and bucco-lingual directions. The pulp chamber of each tooth was opened following a standardized procedure. Canal length was determined visually by passing a size #15 K-file (Dentsply Maillefer, Ballaigues, Switzerland) into the root canal until its tip was visible from the apical foramen. The 55

working length was established 1 mm short of the apex. The root canals were instrumented using stainless steel K-files (Dentsply Maillefer, Ballaigues, Switzerland) and rotary Ni-Ti instruments (Flex-Master, VDW, Munich, Germany). All canals were prepared to ISO size 30, 0.04 taper. The instrumentation was performed strictly according to the manufacturerâ&#x20AC;&#x2122;s instructions. Each set of Flex-Master was discarded after preparing 8 canals. The root canals were irrigated in between instrumentation with 5.25% sodium hypochlorite at 37Â°C and 10% ethylene diamine tetra-acetic acid solution alternately. All the teeth were obturated by means of the warm vertical condensation technique, using gutta-percha and an endodontic sealer (Pulp Canal Sealer, Kerr, Orange, CA, USA). To minimize the influence of size and shape variations of the root canals on the results, the teeth were classified according to their mesial-distal and bucco-lingual dimensions and randomly distributed into 5 groups: Group 1 was made up of 10 teeth (control group) while Groups 2 to 5 were made up of 20 teeth each. The specimens were prepared as follows (Fig. 1): -

Group 1 (control group): the endodontic access to the pulp chamber was filled for at least 2 mm by means of a flowable resin composite material (X-Flow, Dentsply Caulk, York, PA, USA); a micromatrix resin composite material (Esthet-X, Dentsply Caulk) was used to build-up the occlusal surface. All the coronal walls were intact;

Group 2: the distal wall of each tooth was removed, using the limits of the marginal crest as anatomical references. The cavity was extended towards the access preparation, so as to create a divergent disto-occlusal standardized adhesive preparation. The cervical margin was placed 1 mm apical to the cemento-enamel junction (CEJ).

Group 3: both the distal and the mesial walls of each tooth were removed, so as to create a mesial-occlusal-distal (MOD) cavity. The same preparation criteria used in Group 2 were adopted. 56

Group 4: the distal, mesial and buccal walls of each tooth were removed. The same preparation criteria used in previous groups were adopted,

Group 5: the whole crown of each tooth was removed 1 mm coronally to the most occlusal point of the CEJ. Groups 2, 3, 4 and 5 were divided into two subgroups, consisting of 10 specimens

each. In subgroups b (2b-5b), according to the morphology of root canals, gutta-percha was removed with a Largo drill No. 1 or No.2 (Dentsply Maillefer, Ballaigues, Switzerland) using a handpiece at 800-1220 rpm. At least 4 mm of root filling were left to preserve the apical seal. According to the morphology of root canals, translucent glass fiber posts #1 or #2 (DT Light Post, RTD, St. Egeve, France) were used for retention in the post cases created. Each post was tried into the root canal and cut to adequate length with a diamond bur, in order to cover its occlusal end with at least 2 mm of resin composite. The post surface was silanized with Monobond-S (Ivoclar-Vivadent, Schaan, Liechtenstein) for 60 s. A universal self-priming dental adhesive (Prime&Bond NT Dual-Cure, Dentsply Caulk) was used to bond both the fibre posts into the root canals and the composite restorations on the endodontically treated teeth. The following bonding protocol was adopted, strictly according to the manufacturerâ&#x20AC;&#x2122;s instructions. The root canal surfaces were thoroughly washed with air/water spray and then gently blown with an air syringe, not to desiccate the dentin surface. The total etch technique was performed, applying 36% phosphoric acid (DeTrey Conditioner 36, Dentsply Caulk) to the canal walls surfaces for 15 seconds. The conditioned areas were rinsed thoroughly with water for at least 15 seconds. Water was removed from the rinsed canal with a soft blow of air and paper points. A moist surface was left, to avoid desiccating the dentine. Once the surfaces were treated, they were carefully kept uncontaminated. An equal number of drops of Prime&Bond NT DualCure and Self Cure Activator (Dentsply Caulk) were mixed in a clean mixing well for 2 57

seconds. The activated self-priming dental adhesive was dispensed onto a disposable micro-brush and immediately rubbed on all root canal surfaces for at least 20 seconds. The solvent was removed by blowing gently with air from a dental syringe for at least 5 seconds. Paper points were introduced into the root canal to remove excess bonding agent. The mixed adhesive/activator was applied on the post, which was then dried with air from a dental syringe for at least 5 seconds. The posts were then luted with a dualcured resin cement (Calibra, Dentsply Caulk), following the manufacturerâ&#x20AC;&#x2122;s instructions. The luting agent was applied both to the post and post-space. The posts were seated into the root canals and the cement excess was cleaned up with paper points. The resin cement and the adhesive material were simultaneously light-cured through the coronal portion of the cemented post for 40 seconds using a halogen light-curing unit (Optilux 401, Demetron, Kerr, Danbury, USA) at 750 mW/cm2. In all specimens, the access preparations were filled for at least 2 mm with a flowable light-cured resin composite (X-Flow); cavity fillings and crown build-ups were performed using a micromatrix resin composite (Esthet-X). The composite restorations were bonded using a universal self-priming dental adhesive (Prime&Bond NT Dual-Cure, Dentsply Caulk) according to the bonding protocol previously described. The enamel was etched for at least 15 seconds. Transparent matrices (Hawe Striproll, KerrHawe, Bioggio, Switzerland) were used in order to hold the restorative material. A standardized incremental composite build-up technique was employed, by light-curing the resin composites at 2 mm increments. Each increment was light-cured for 40 seconds on each aspect of the specimens, ensuring uniform exposure of all cavity surfaces according to the morphology of the restoration. A simplified anatomical build-up technique was used to restore all the crowns. Each tooth was embedded in a block of self-curing acrylic resin (Jet Kit, Lang Dental MFG. CO., Wheeling, IL, USA) surrounded by aluminum cylinders with the long 58

axis perpendicular to the base of the block, leaving at least half of the root exposed so as to evaluate morphologically the eventual root fractures. An addition-cure polyvinylsiloxane impression material (Flexitime, Heraeus Kulzer, Hanau, Germany) was applied in the root region to a thickness of 200 to 400 Îźm to simulate the periodontal ligament. In order to dissipate the heat generated from the polymerization reaction of the resin, the crowns were continuously moistened with water spray. The specimens were stored in distilled water at room temperature until they were used for static fracture resistance measurement; the storage time ranged from a minimum of 24 hours to a maximum of 36 hours. A universal loading machine (Triaxal Tester T400 Digital, Controls srl, Cernusco s/N., Italy) was used for this purpose. Each specimen was inserted into the holding device with an inclination of 90Â° in relation to the horizontal plane (Fig. 2). A controlled load was applied by means of a stainless steel rod with a tip diameter of 1 mm in a direction parallel to the longitudinal axis of the tooth. Pressure from the rod tip was applied on the resin composite at the palatal cusp, 2 mm from the tip of the cusp towards the central fossa, in order to simulate an occlusal load. The load was applied at a crosshead speed of 1 mm/minute. All samples were loaded until fracture and the maximum breaking loads were recorded in Newtons (N) by a computer (Digimax Plus, Controls srl, Cernusco s/N., Italy) connected to the loading machine. After mechanical failure, all the samples were perfused with Indian ink, in order to highlight the fracture lines. The mode of failure was recorded and classified as restorable or unrestorable. The failure mode was macroscopically evaluated at sight and the restorations were classified as unrestorable with the appearance of root fractures. In case of micro-cracks, the fracture pattern was evaluated using a 10x stereomicroscope (Zeiss OpMi1, Zeiss, Oberkochen, Germany). The data were statistically analyzed with the software SPSS 12.0 (SPSS Inc., Chicago, IL, USA). The Kolmogorov-Smirnov test was used to verify the normality of the 59

data distribution. The data recorded from the control group (group 1) and the groups without fiber posts (2a, 3a, 4a and 5a) were analyzed with one-way analysis of variance (ANOVA) followed by Tukey post hoc test for comparisons to reveal the influence of the number of residual walls on the fracture resistance. Differences in the fracture loads between the groups without (subgroups a) and with fiber posts (subgroups b) within group 2 to group 5 were analyzed with t-test. For all the statistical tests, the level of significance was set at p = 0.05.

Results The highest mean fracture resistance was recorded for subgroup 5a at 856.7±112.2 N, followed by subgroup 5b at 649.5±163.5 N, subgroup 4b at 573.4±169.2 N, subgroup 3b at 513.2±121.7 N, Group 1 at 502.4±152.5 N, subgroup 4a at 488.7±153.7 N, subgroup 2b at 423.0±103.3 N, subgroup 3a at 422.1±138.9 N and subgroup 2a at 416.4±122.2 N. The one-way ANOVA and the Tukey post hoc test revealed that the group with no residual walls (group 5) reached significantly higher fracture resistance than all the other groups (P< 0.001); on the contrary, the mean fracture loads of Groups 1 to 4 were statistically comparable (P> 0.05) (Table 1). The t-test between the groups without (subgroups a) and with fiber posts (subgroups b) within Group 2 to Group 5 showed that subgroup 5a reached significantly higher fracture resistance than subgroup 5b (P =0.004); on the contrary, difference was not significant between subgroup 2a and 2b (P = 0.898), 3a and 3b (P = 0.136) and 4a and 4b (p = 0.257) (Table 2). For the mode of failure, the single rooted maxillary premolars restored with fiber posts mostly exhibited restorable fractures while teeth restored without fiber posts mostly exhibited unrestorable failures (Fig. 3). The results are summarized in Table 3. Most of the unrestorable fractures were evident in teeth restored without fiber posts. For example, in subgroup 5a, seven premolars exhibited catastrophic fractures, while in 60

subgroup 5b, eight teeth were characterized by restorable partial crown fractures. The presence of fiber posts was statistically significant for enhancing the fracture resistance of the specimens. No unrestorable fractures were noticed in subgroup 4b. Catastophic failures were characterized by an oblique fracture line on the palatal surface of the teeth, extending from the coronal to the middle third of the root. Complete crown fracture and serious root damage occurred only in one specimen in subgroup 4a and in one specimen in subgroup 5a. In all groups, coronal failures of both the residual dental tissues and the composite restorations occurred at the level of the palatal cusp. Cohesive fractures of the composite material were noticed in all groups. Adhesive failure of the whole composite bulk was evident only in few specimens in subgroups 2a and 2b. Only one specimen in subgroup 4b displayed the exposure of a fiber post after the fracture of the crown.

Discussion Numerous studies have been performed to evaluate the mechanical resistance of endodontically-treated teeth and in particular maxillary premolars incidence of fracture for this group of teeth has been reported

8,9

11,12,14,18,21,22,24

, as a high

. Most studies focused on

the materials and techniques used to increase the strength of the tooth-restoration complex

13-15

. The present study was performed to evaluate the influence of both the post-

and-core restorative system and the number of residual dental walls on the fracture resistance of these teeth. As fracture resistance was improved with the use of a fiber post and with a decreasing number of residual cavity walls, the null hypothesis tested was rejected. As to the preparation of the post space, several studies recommended a post length equal to Âž of root canal length

or at least equal to the length of the crown

. Moreover,

according to traditional teachings, a minimum of 3 to 5 mm of gutta-percha should remain in the apical portion of the root to maintain an adequate seal 26,28. However, a recent study 61

demonstrated that 3 mm of gutta-percha provide an unreliable apical seal, therefore, 4 to 5 mm are recommended

. Such observations were considered for the specimen

preparations of the present study. There is concern that eugenol-containing root canal sealers may interfere with the polymerization of resin cements weakening the adhesion between root dentin, luting agents and fiber posts

30-33

. Several studies investigated the effects of residual eugenol

from eugenol containing temporary cements on the bond strength of resin luting agents and post retention

34,35

. Some studies found eugenol to have no effect on shear bond

strengths of resin adhesives

33,36

. On the contrary, other studies proved that eugenol-

containing sealers reduced post retention 34. However, it was reported that such a problem can be avoided by mechanical or chemical decontamination of the dentin from residual eugenol. Post space preparation can remove the contaminated dentin layer 37% phosphoric acid etching of the canal walls

37-40

. Moreover,

and the use of alcohol and/or EDTA

are effective in removing residual eugenol and restoring the retentive strength of resins to dentin

. In the present study, after preparing mechanically the post space with calibrated

drills, EDTA was used to decontaminate root canal dentin from residual eugenol. Most static mechanical fracture tests reported in the literature are characterized by a loading of the premolars on the palatal cusp at 130Â°-150Â° to the longitudinal axis. Such a condition generates a compressive load perpendicular to the cuspal side

12,18

. On the

contrary, during function, the occlusion generates non-axial forces resolved into their vectors along the cuspal side following the parallelogram of forces. As a consequence, it was decided to load the specimens in a direction parallel to the longitudinal axis of the tooth, in order to distribute the stresses more evenly between the residual dental tissues and the restorative material, simulating a physiological occlusion. During function, a patient with no parafunctional habits generates a maximum biting load of about 350-500 N in the region of premolars 41. These values were confirmed by the 62

mean failure load recorded in Group 1 (control group). Subgroup 3b showed the closest mechanical resistance to group 1 while mean fracture loads lower than that of the control group were recorded in subgroups 2a, 2b, 3a and 4a. In general, the use of fiber posts permitted higher failure loads to be obtained than those recorded in teeth restored by resin composites only. Several studies pointed out that failure of endodontically-treated teeth is exasperated by the loss of tooth structure

. As far as adhesive restorations are

concerned, the results of this study indicated that irrespective of the adjunctive use of fiber posts, lower mean failure loads were predominantly associated with the presence of more residual cavity walls. Such a phenomenon may be explained by the higher cavity configuration factors that are associated with cavities with more residual walls

. An

adhesive interface represents the transition area between materials with different moduli of elasticity. In an adhesive restoration, each bondable dental surface creates an adhesive interface where stress accumulates during the polymerization of composite materials 19. As a consequence, an increase in the number of cavity walls for bonding will simultaneously reduce the capacity for relief of shrinkage stresses via flow of the unset resinous materials. Moreover, an adhesive interface represents a weak link which can fail under loading. In subgroups 2a, 2b, 3a and 3b, the progressive increase in adhesive interfaces between the tooth structure, the composite restoration and the fiber post probably enhanced the propagation of microcracks generated in the load application area, leading to the mechanical failure of the palatal cusps and to the cohesive fractures observed in both the dentin and the composite restoration. As the number of cavity walls in subgroups 4a, 4b, 5a and 5b were progressively reduced, high values of fracture resistance were observed. Since the cavity configuration factors were reduced, there was an increasing chance of stress relief during the initial polymerization phase and a more homogeneous distribution of stresses from the load 63

application area during the loading phase. Only one specimen in subgroup 4a was affected by an unrestorable fracture while no catastrophic failures were observed in subgroup 5b. This highlighted the optimal polymerization and mechanical behaviour of the composite restorations under these less taxing conditions, enabling a more homogeneous load distribution to be subsequently achieved. The presence of a fiber post introduces an additional adhesive interface in the restorative system, which can participate in the propagation of microcracks, leading to tooth fracture. Such a phenomenon may explain the lower mean load fracture recorded in subgroup 5b, when compared with subgroup 5a. Nevertheless, the mode of fracture in the subgroup 5b specimens is conducive to repair, whereas irreparable root fractures were observed in the subgroup 5a specimens. From a functional point of view, the restored premolars of both subgroups 5a and 5b can sustain occlusal loads. However, the more favorable fracture patterns observed in the fiber post-retained restorations in subgroup 5b represent a paramount factor that is highly desirable for the long-term success of these restorations. With the use of ink perfusion, microcracks could be seen propagating from the load application area towards the inner portion of the composite bulk and following the adhesive interfaces, leading to the mechanical failure of the palatal cusp. Analysis of these failure patterns revealed fracture lines that were almost parallel to the direction of the applied load. The crack propagation was caused mainly by the loading direction rather than the potentially weak sites (e.g. voids or contact contamination along composite increments) within the restorative system itself, confirming the optimal mechanical characteristics of both the adhesive system and the composite materials. Within the limitations of this study, it may be concluded that the mechanical resistance of endodontically-treated teeth is enhanced by the use of a bonded fiber postcomposite core restoration. Resistance to fracture is strongly associated with the number 64

of remaining cavity walls in the tooth to be restored. In teeth that exhibited the same number of cavity walls, higher fracture loads were observed in teeth that were restored with fiber posts. Moreover, fractures that occur in teeth restored with fiber posts are repairable, while those restored with resin composites only demonstrate frequent catastrophic root fractures that necessitate the removal of these damaged teeth. Further clinical trials should be conducted to validate the results of this in vitro study.

References 1. Llena-Puy MC, Forner-Navarro L, Barbero-Navarro I. Vertical root fracture in endodontically treated teeth: A review of 25 cases. Oral Sur Oral Med Oral Pathol Oral Radiol Endod 2001; 92: 553-555. 2. Testori T, Badino M, Castagnola M. Vertical root fractures in endodontically treated teeth: A clinical survey of 36 cases. J Endod 1993; 19: 87-91. 3. Tamse A, Fuss Z, Lustig J, Kaplavi J. An evaluation of endodontically treated vertically fractured teeth. J Endod 1999; 7: 506-508. 4. Fennis WMM, Kuijs RH, Kreulen CM, Roeters FJ, Creugers NH, Burgersdijk RC. A survey of cusp fractures in a population of general dental practices. Int J Prosthodont 2002; 15: 559-563. 5. Khers SC, Carpenter CW, Vetter JD, Staley RN. Anatomy of cusps of posterior teeth and their fracture potential. J Prosthet Dent 1990; 64: 139-147. 6. Hansen EK, Asmussen E, Christiansen NC. In vivo fractures of endodontically treated posterior teeth restored with amalgam. Endod Dent Traumatol 1990; 10: 4955. 7. Sedgley CM, Messer HH. Are endodontically treated teeth more brittle? J Endod 1992; 18: 332-335. 8. Ross IF. Fracture susceptibility of endodontically treated teeth. J Endod 1980; 6: 560-565. 9. Salis SG, Hood JAA, Stokes ANS, Kirk EEJ. Pattern s of indirect fracture in intact and restored human premolar teeth. Endod Dent Traumatol 1987; 3: 10-14. 10. Caputo AA, Standlee JP. Pins and posts â&#x20AC;&#x201C; why, when and how. Dent Clin North Am 1976; 20: 299-311. 11. Sornkul E, Stannard JG. Strength of roots before and after endodontic treatment and restoration. J Endod 1992; 18: 440-443. 66

12. Akkayan B, Gulmez T. Resistance to fracture of endodontically treated teeth restored with different post systems. J Prosthet Dent 2002; 87: 431-437. 13. Hudis SI, Goldstein GR. Restoration of endodontically treated teeth: A review of the literature. J Prosthet Dent 1986; 55: 33-38. 14. Ortega VL, Pegoraro LF, Conti PCR, Do Valle AL, Bonfante G. Evaluation of fracture resistance of endodontically treated maxillary premolars, restored with ceromer or heat-pressed ceramic inlays and fixed with dual-resin cements. J Oral Rehabil 2004; 31: 393-397. 15. Creugers NH, Mentink AG, Kayser AF. An analysis of durability data on post and core restorations. J Dent 1993; 21: 281-284. 16. Sorensen JA, Martinoff JT. Intracoronal reinforcement and coronal coverage: A study of endodontically treated teeth. J Prosthet Dent 1984; 51: 780-784. 17. TorbjĂśrner A, Karlsson S, Ă&#x2013;dman PA. Survival rate and failure characteristics for two post designs. J Prosthet Dent 1995; 73: 439-444. 18. Yamada Y, Tsubota Y, Fukushima S. Effect of restoration method on fracture resistance of endodontically treated maxillary premolars. Int J Prosthodont 2004; 17: 94-98. 19. Ausiello P, Gee AJ, Rengo S, Davidson CL. Fracture resistance of endodonticallytreated premolars adhesively restored. Am J Dent 1997; 10: 237-241. 20. Assif D, Gorfil C. Biomechanical considerations in restoring endodontically treated teeth. J Prosthet Dent 1994; 71: 565-567. 21. Akkayan B, Canikioglu B. Resistance to fracture of crowned teeth restored with different post systems. Eur J Prosthod Restor Dent 1998; 6: 13-18. 22. Libman WJ, Nicholls JI. Load fatigue of teeth restored with cast posts and cores and complete crowns. Int J Prosthodont 1995; 8: 155-161.

23. Goldberg AJ, Burstone CJ. The use of continuous fibre reinforcement in dentistry. Dent Mater 1992; 8: 197-202. 24. Mannocci F, Ferrari M, Watson TF. Intermittent loading of teeth restored using quartz fibre, carbon-quartz fibre and zirconium dioxide ceramic root canal posts. J Adhes Dent 1999; 1: 153-158. 25. Asmussen E, Peutzfeldt A, Heitmann T. Stiffness, elastic limit and strength of newer types of endodontic posts. J Dent 1999; 27: 275-278. 26. Goodacre CJ, Spolnik KJ. The prosthodontic management of endodontically treated teeth: a literature review. Part III. Tooth preparation considerations. J Prosthodont 1995; 4: 122-128. 27. Sorensen JA, Martinoff JT. Clinically significant factors in dowel design. J Prosthet Dent 1984; 52: 28-35. 28. Goodacre CJ, Spolnik KJ. The prosthodontic management of endodontically treated teeth: a literature review. Part II. Maintaining the apical seal. J Prosthodont 1995; 4: 51-53. 29. Abramovitz L, Lev R, Fuss Z, Metzger Z. The unpredictability of seal after post space preparation: a fluid transport study. J Endod 2001; 27: 292-295. 30. Hansen EK, Asmussen E. Influence of temporary filling materials on effect of dentin-bonding agents. Scand J Dent Res 1987; 95: 516-520. 31. Craig RG. Restorative Dental Materials 8th ed. St. Louis: CV Mosby Co., 1989; 272. 32. Ngoh EC, Pashley DH, Loushine RJ, Weller RN, Kimbrough F. Effects of eugenol on resin bond strengths to root canal dentin. J Endod 2001; 27: 411-414. 33. Schwartz RS, Robbins JW. Post placement and restoration of endodontically treated teeth: a literature review. J Endod 2004; 30: 289-301.

34. Tjan A, Nemetz H. Effect of eugenol-containing endodontic sealer on retention of prefabricated posts luted with an adhesive composite resin cement. Quintessence Int 1992; 22: 839-844. 35. Schwartz

RS,

Murchison

DF,

Walker

WA.

Effects

eugenol

and

noneugenolendodontic sealer cements on post retention. J Endod 1998; 24: 564567. 36. Peutzfeldt A, Asmussen E. Influence of eugenol-containing temporary cement on efficacy of dentin-bonding systems. Eur J Oral Sci 1999; 107: 65-69. 37. Mayhew JT, Windchy AM, Goldsmith LJ, Gettleman L. Effect of root canal sealers and irrigation agents on retention of preformed posts luted with a resin cement. J Endod 2000; 26: 341-344. 38. Boone KJ, Murchison DF, Schindler WG, Walker WA. Post retention: the effect of sequence of post-space preparation, cementation time, and different sealers. J Endod 2001; 27: 768-771. 39. Mannocci F, Ferrari M, Watson TF. Microleakage of endodontically treated teeth restored with fibre posts and composite cores after cyclic loading: a confocal microscopic study. J Prosthet Dent 2001; 85: 284-291. 40. Wolanek GA, Loushine RJ, Weller RN, Kimbrough WF, Volkmann KR. In vitro bacterial penetration of endodontically treated teeth coronally sealed with a dentin bonding agent. J Endod 2001; 27: 354-357. 41. Korioth TW, Versluis A. Modeling the mechanical behavior of the jaws and their related structures by finite element (FE) analysis. Crit Rev Oral Biol Med 1997; 8: 90-104.

ONE-WAY ANOVA

Walls (R)

Walls (C)

1 wall b 0 walls

1 wall b

2 walls b

Mean difference (R-C)

Standard error

p value Lower bound

Upper bound

367.91

61.22

<0.001

193.95

541.87

434.51

61.22

<0.001

260.55

608.47

3 walls 4 walls b

440.27 354.22

61.22 61.22

<0.001 <0.001

266.31 180.26

614.23 528.18

0 walls a 2 walls b 3 walls b 4 walls b

-367.91 66.60 72.36 -13.68

61.22 61.22 61.22 61.22

<0.001 0.812 0.761 0.999

-541.87 -107.36 -101.60 -187.64

-193.95 240.56 246.32 160.27

0 walls a 1 wall b 3 walls b

-434.51 -66.60 5.76

61.22 61.22 61.22

<0.001 0.812 1.000

-608.47 -240.56 -168.20

-260.55 107.36 179.72

4 walls b

-80.29

61.22

0.686

-254.25

93.67

0 walls 1 wall b 2 walls b

-440.27 -72.36 -5.76

61.22 61.22 61.22

<0.001 0.761 1.000

-614.23 -246.32 -179.72

-266.31 101.60 168.20

4 walls b

-86.05

61.22

0.627

-260.01

87.91

-354.22

61.22

<0.001

-528.18

-180.26

2 walls

3 walls b

95% Confidence Interval

0 walls

a b

1 wall 13.68 61.22 0.999 -160.27 187.64 b 2 walls 80.29 61.22 0.686 -93.67 254.25 b 3 walls 86.05 61.22 0.627 -87.91 260.01 Table 1 â&#x20AC;&#x201C; One-way ANOVA in the interaction between the number of residual dental walls. Values are mean differences of fracture loads Âą standard deviations; standard errors are also calculated. Subgroups with the same superscripts are not statistically significant (p>0.05). [R = reference group; C = comparison group] 4 walls b

t-TEST

Group

Mean fracture load (N)

2a § (3 walls, no post)

416.38

Standard deviation

Standard error mean

122.17

38.63

2b (3 walls, post)

422.98

103.31

32.67

3a † (2 walls, no post)

422.14

138.92

43.93

3b † (2 walls, post)

513.17

121.67

38.47

4a ‡ (1 wall, no post)

488.74

153.75

48.62

‡

4b (1 wall, post)

573.43

169.19

53.50

5a ↕ (0 walls, no post)

856.65

112.21

35.48

∫

5b (0 walls, post)

649.47

163.54

p value

-0.13

0.898

-1.56

0.136

-1.17

0.257

3.30

0.004

51.72

Table 2 – Results of the t-tests in the interaction between the number of residual dental walls and the restorative system. Values are means ± standard deviations; standard error means are also calculated. Mean fracture loads are expressed in Newtons (N). Subgroups with the same superscripts are not statistically significant (p>0.05). [df = degrees of freedom]

MODE OF FAILURE Restorable fractures

Unrestorable fractures

Group 1 (control group)

Subgroup 2a

Subgroup 2b

Subgroup 3a

Subgroup 3b

Subgroup 4a

Subgroup 4b

Subgroup 5a

Subgroup 5b

Table 3 â&#x20AC;&#x201C; Mode of failure of the specimens.

Fig. 1 â&#x20AC;&#x201C; Pulp chamber opening and cavity preparation designs.

Fig. 2 â&#x20AC;&#x201C; Fracture test loading direction.

Fig. 3 â&#x20AC;&#x201C; Restorable (left side) and unrestorable fracture (right side).

Effect of Post Retained Composite Restoration of Mesial-Occlusal-Distal Cavities on the Fracture Resistance of Endodontically Treated Teeth.

Sorrentino R., Salameh Z., Zarone F., Tay F.R., Ferrari M. â&#x20AC;&#x201C; Journal of Adhesive Dentistry, 2006 in press.

ABSTRACT Purpose. The present study aimed at comparing the fracture resistance and failure patterns of endodontically treated premolars with MOD cavities restored using different material combinations. The null hypothesis postulated that there was no association between the fracture resistance of endodontically treated premolars and the resin composite materials or the post-and-core system used to build up the restorations. Materials and Methods. 80 single rooted maxillary premolars were used. After endodontic treatment and preparation of MOD cavities, 8 groups were created made up of 10 samples each using the following material combinations: group 1 (control), flowable and microhybrid resin control composites; group 2, flowable A; group 3, flowable B; group 4, microhybrid resin A; group 5, microhybrid resin B; group 6, flowable B + microhybrid resin B; group 7, flowable A + microhybrid resin A + post A; group 8, flowable B + microhybrid resin B + post B. Mechanical static fracture tests were performed loading the specimens till fracture. Results. The mean failure loads (N) were 502.4 [control], 470.5 [group 7], 445.5 [group 8], 440.8 [group 6], 405.1 [group 5], 363.9 [group 4], 316.9 [group 2] and 301.9 [group 3] respectively. Statistically significant differences were found between group 1 vs 2, 1 vs 3 and 3 vs 7 (p<.05). Conclusion. The fracture resistance of endodontic treated premolars with MOD cavities was enhanced by the use of the sandwich technique. The samples restored with posts showed predominantly restorable fractures, while teeth restored without posts mostly displayed unrestorable failures.

Introduction Endodontically treated teeth usually present inadequate remaining coronal structure due to cavity preparation (Assif & Gorfil 1994, Ferrari et al. 2000a, Akkayan & Gulmez 2002). The type of definitive restoration to restore endodontically treated maxillary teeth may be influenced by the amount of hard tissues remaining after tooth preparation (Seow et al. 2005). Their survival is endangered by substantial loss of dentin and they are considered more susceptible to fracture than healthy teeth (Assif & Gorfil 1994). In particular, relatively high fracture incidence was documented in maxillary premolars (Trope et al. 1986, Salis et al. 1987). The choice of appropriate restorations should be guided by both physical properties and esthetics (Ferrari et al. 2000a). A restoration has several purposes: to repair the cavity, to strengthen the tooth and to provide an effective seal between the root canal system and the oral environment (Belli et al. 2005). Endodontic treatment should not be considered complete until the coronal restoration has been placed (Wagnild & Mueller 2002). Coronal leakage may lead to bacterial contamination; on the contrary, the longevity of an endodontic treatment is significantly increased by a correct coronal restoration (Tronstad et al. 2000). Posts were originally designed to build up and retain coronal restorations in case of severely damaged teeth (Ferrari et al. 2000a, Ferrari & Scotti 2002). The growing patientsâ&#x20AC;&#x2122; demand for esthetics led to the development of metal-free post-and-core systems (Ferrari et al. 2000a, Akkayan & Gulmez 2002, Mannocci et al. 1999) or, alternatively, of build up procedures for restoring endodontically treated teeth using fiber posts in combination with direct esthetic restorations (Grandini et al. 2005a). Prefabricated posts currently available on the market are characterized by different physical properties, chemical composition and handling (Monticelli et al. 2004). Fiber-, glass- and quartz-reinforced posts were developed to obtain restorative systems with elastic moduli comparable to that of dentin (Asmussen et al. 1999, Drummond et al. 1999). Later on, translucent fiber posts allowed clinicians to 77

use light-curing cements and achieve optimal esthetics (Mannocci et al. 1999). Due to the inclusion of chemical initiators, dual-cure resin cements are able to polymerize even where light transmission is incomplete, just like the apical third of root canals (Foxton et al. 2005). Over the past 10 years, several in vitro (Akkayan & Gulmez 2002, Mannocci et al. 1999, Asmussen et al. 1999, Drummond et al. 1999, Akkayan & Caniklioglu 1998, Akkayan 2004) and in vivo studies (Ferrari et al. 2000a, Ferrari & Scotti 2002, Ferrari et al. 2000b) were performed to assess the mechanical performances of fiber posts (Asmussen et al. 1999, Grandini et al. 2002, Grandini et al. 2005b). Several studies demonstrated convincing clinical results for the adhesion of fiber posts to root canal substrates (Nakabayashi & Pashley 1998, Ferrari et al. 2000c, Ferrari et al. 2001a, Goracci et al. 2004), the reliability of luting procedures (Ferrari et al. 2001b) and coronal build up (Gateau et al. 1999, Freedman 2001). However, studies on the post-and-core technique reported highly variable survival rates, demonstrating that unfavourable root fractures may occur (Yamada et al. 2004). Teeth restored with bonded CAD/CAM ceramic inlays fractured with a significantly high number of severe fractures as well (Hannig et al. 2005). Recently, several papers supported the use of a direct restoration without placing any post for restoring endodontically treated teeth (Stockton et al. 1998, Baratieri et al. 2000, Assif et al. 2003). Adhesive restorations have been claimed to enhance cuspal stiffness in teeth restored without crown coverage (St.Georges et al. 2003). Moreover, adhesive techniques preserve the maximum amount of sound tooth structure (Mannocci et al. 2002). As to fracture strength and fracture patterns, Krejci et al. (2003) showed no significant differences between teeth restored with and without posts. Moreover, some studies pointed out that mechanical resistance to fracture of endodontically treated teeth could be affected by the presence of posts and the risk of residual tooth structure damage could increase (Akkayan & Gulmez 2002, Akkayan 2004, Strub et al. 2001). In endodontically treated teeth, occlusal loads could be transferred intraradicularly by post78

and-core restorations, increasing the occurrence of vertical root fractures (Trope et al. 1986). On the contrary, other authors noticed that fiber posts reduced the risk of root fractures (Mannocci et al. 1999). Actual consensus in restorative dentistry indicates that decementation or failure of posts is preferable than fracture of residual tooth structure (Akkayan 2004, Ferrari et al. 2000b). The less residual tooth structure the more important the physical properties of post-and-core systems (Christensen 1996). For building up the coronal restoration after luting a fiber post, all types of composite materials have been proposed (Ferrari et al. 2000a, Ferrari & Scotti 2002, Grandini et al. 2005a, Monticelli et al. 2004, Ferrari et al. 2000b). Microhybrid and flowable resin materials in the self-curing or light-activated formulation are characterized by different strength, stiffness and elasticity: such properties could affect the longevity of the restoration (Ferrari et al. 2000a, Asmussen et al. 1999). Such resin composites can be used together in the so called â&#x20AC;&#x153;sandwich techniqueâ&#x20AC;?, in order to join the mechanical properties of both materials (Croll 2004). Up to date, there is still no agreement regarding the best composite for direct coronal build up of endodontically treated teeth (Ferrari et al. 2000a, Monticelli 2004). Adhesive interfaces of bonded restorations transmit and distribute occlusal forces to the remaining tooth structures homogeneously, potentially strengthening the restored tooth and increasing its resistance to fracture (Hernandez et al. 1994). On the other hand, the interfaces of materials with different moduli of elasticity represent the weak point of a restorative system, as the toughness/stiffness mismatch influences the stress distribution (Assif & Gorfil 1994). The C-factor, defined as the ratio of bonded to unbonded surface areas of cavities, is highly unfavorable in root canals, where it can range from 20 to 200 (Morris et al. 2001). A rigid bond between composite materials and remaining tooth structures could cause stress concentration at the adhesive interfaces (Kemp-Scholte & Davidson 1990). In order to reduce the risk of debonding due to polymerization shrinkage, low viscosity flowable composites with low elastic modulus can be placed between the 79

bonding agent and the composite restoration (Van Meerbeek et al. 1992). Such an intermediate resin layer acts as a stress adsorber (Alhadainy & Abdalla 1996), releasing contraction stresses and improving marginal integrity of the restoration (Belli et al. 2005). Versluis et al. (1996) reported that filling a cavity with a bulk of composite generates less volumetric shrinkage than using the incremental technique within identical cavity shapes. When using composite materials, the resistance of a tooth to cuspal and/or vertical fracture may be affected by several factors, such as the mechanical properties of the restorative system used (Grandini et al. 2005a), the presence of fiber posts (Grandini et al. 2005a) and the cavity shape and dimensions (Hannig et al. 2005). Different studies proved that mesial-occlusal-distal (MOD) cavity preparations reduce the resistance to fracture of root filled teeth, especially maxillary premolars (Belli et al. 2005). A recent in vitro study showed that the structural integrity of the bonding interface between the core material and the post can play an important role on the longevity of the restoration of endodontically treated teeth (Monticelli et al. 2004). In order to reinforce the post-and-core system, the coronal end of the post should be surrounded by at least 1 mm in length of the core material (Monticelli et al. 2004). It has been well documented that fracture resistance of teeth depends on the angle of applied load and axial forces are less detrimental than oblique forces (Loney et al. 1995); in particular, shear stresses directly applied to th cusps may lead to mechanical failure of adhesive restorative systems (Uyehara et al. 1999). As a consequence, direction of loads is a crucial factor to the longevity of root filled teeth (Krejci et al. 2003). Moreover, natural variations in tooth morphology may affect the long-term success of restorations (Ortega et al. 2004). The present study aimed at comparing the fracture resistance and failure patterns of endodontically treated single rooted maxillary premolars with MOD cavities restored using

microhybrid and flowable resin composites in different combinations with or without translucent glass fiber posts. The null hypothesis postulated that there is no association between the fracture resistance of endodontically treated single rooted maxillary premolars and the resin composite materials or the post-and-core system used to build up the restorations.

Materials and methods Eighty human single rooted maxillary premolars, extracted for orthodontic or periodontal reasons, were selected for the study. Teeth with caries and previous restorations were excluded. Only sound teeth with an average length of 23.5±1 mm, bucco-lingual coronal dimensions of 9±1 mm and mesial-distal coronal dimensions of 7±1 mm (Kraus et al. 1969) were included in the study. Dental plaque, calculus and external debris were removed with an ultrasonic scaler (Cavitron SPS Ultrasonic Scaler, Dentsply Caulk, York, PA, USA). In order to simulate the oral environment, the teeth were stored in an incubator at 37°C in 90% humidity until the execution of the mechanical tests. Canal morphology was verified from standardized apical radiographs (70 kV and 0.08 s) both in the mesial-distal and bucco-lingual directions. Because of anatomic variability, the teeth were prepared free-hand. Both access and MOD cavities were prepared using high-speed diamond rotary cutting instruments (314 846 KR 016, Komet, Lemgo, Germany) under constant water cooling. The MOD cavities were prepared down to the root canal orifices according to standardized dimenions, so that the thickness of residual walls measured 2 mm occlusally and 2.5 mm at the cementum-enamel junction on the buccal aspect and 1.5 mm occlusally and 1.5 mm at the cementum-enamel junction on the lingual aspect. Standardized depth was verified with a scaled periodontal probe (23/UNC 15, Hu Friedy, Chicago, IL, USA). All internal cavity surfaces and margins were finished with diamond

rotary finishing instruments (314 8846 KR 016, Komet) under constant water cooling with the use of a stereomicroscope at 15x magnification. Canal length was determined visually by passing a size #15 K-file (Dentsply Maillefer, Ballaigues, Switzerland) into the root canal until its tip was visible from the apical foramen. The working length was established 1 mm short of the apex. The root canals were instrumented using stainless steel K-files (Dentsply Maillefer) and rotary Ni-Ti instruments (Flex-Master, VDW, Munich, Germany). All canals were prepared to ISO size 30, 0.04 taper. The instrumentation was performed strictly according to the manufacturerâ&#x20AC;&#x2122;s instructions. Each set of Flex-Master was discarded after preparing 8 canals. The root canals were irrigated in between instrumentation with 5.25% sodium hypochlorite at 37Â°C and 10% ethylene diamine tetra-acetic acid solution alternately. All the teeth were obturated by means of the warm vertical condensation technique, using gutta-percha and an endodontic sealer (Pulp Canal Sealer, Kerr, Orange, CA, USA). To minimize the influence of size and shape variations of the root canals on the results, the teeth were classified according to their mesial-distal and bucco-lingual dimensions and randomly distributed into 8 groups. Two-way ANOVA was used to assess coronal bucco-lingual, coronal mesial-distal and root dimensions and no significant differences were noted among the various groups (p>0.05). Each group was made up of 10 teeth. The specimens were prepared as follows (fig. 1) (Table 1): -

Group 1 (control): both the distal and the mesial walls of each tooth were removed, so as to create a mesial-occlusal-distal (MOD) cavity. The cervical margin was placed 1 mm apical to the cemento-enamel junction (CEJ). A tolerance of 0.3 mm was used to include preparations in the tests. the endodontic access to the pulp chamber was filled for at least 2 mm by means of a flowable resin composite material (X-Flow, Dentsply Caulk); a microhybrid resin composite material (EsthetX, Dentsply Caulk) was used to build-up the occlusal surface. 82

Group 2: a mesial-occlusal-distal cavity was prepared as previously described; the crown build-up was performed using a flowable composite material (flowable A: Aeliteflo, Bisco Inc., Schaumburg, IL, USA);

Group 3: a mesial-occlusal-distal cavity was prepared as previously described; the crown build-up was performed using a flowable composite material (flowable B: Tetric-flow, Ivoclar-Vivadent, Schaan, Liechtenstein);

Group 4: a mesial-occlusal-distal cavity was prepared as previously described; the crown build-up was performed using a microhybrid resin composite material (microhybrid A: Light-Core, Bisco Inc.);

Group 5: a mesial-occlusal-distal cavity was prepared as previously described; the crown build-up was performed using a microhybrid resin composite material (microhybrid B: Tetric-ceram, Ivoclar-Vivadent);

Group 6: a mesial-occlusal-distal cavity was prepared as previously described; the endodontic access and the floor of the cavity was filled with a flowable material while the crown build-up was performed using a microhybrid resin composite material (flowable B + microhybrid B);

Group 7: a mesial-occlusal-distal cavity was prepared as previously described; translucent fiber posts (post A: DT Light Post, RTD, St. Egeve, France) were seated in the root canal; the endodontic access and the floor of the cavity was filled with a flowable material while the crown build-up was performed using a microhybrid resin composite material (flowable A + microhybrid A);

Group 8: a mesial-occlusal-distal cavity was prepared as previously described; translucent fiber posts (post B: FRC Postec, Ivoclar-Vivadent) were seated in the root canal; the endodontic access and the floor of the cavity was filled with a flowable material while the crown build-up was performed using a microhybrid resin composite material (flowable B + microhybrid B). 83

In groups 7 and 8, according to the morphology of root canals, gutta-percha was removed with a Largo drill No. 1 or No.2 (Dentsply Maillefer) using a handpiece at 8001220 rpm. At least 4 mm of root filling were left to preserve the apical seal. Dowel spaces were prepared with calibrated diamond rotary cutting instruments dedicated to each post system following the manufacturerâ&#x20AC;&#x2122;s instructions. EDTA was used to decontaminate root canal dentin from residual eugenol. According to the morphology of root canals, translucent glass fiber posts size #1 or #2 were used for retention in the post space created. Each post was tried into the root canal and cut to adequate length with a diamond rotary cutting instrument, in order to cover its occlusal end with at least 2 mm of resin composite. The post surface was silanized with Monobond-S (Ivoclar-Vivadent) for 60 s. All the materials used in root canals were applied using microbrushes. A universal self-priming dental adhesive (Prime&Bond NT Dual-Cure, Dentsply Caulk) was used to bond both the fiber posts into the root canals and the composite restorations on the endodontically treated teeth. The following bonding protocol was adopted, strictly according to the manufacturerâ&#x20AC;&#x2122;s instructions. The root canal surfaces were washed thoroughly with air/water spray and then gently blowed with an air syringe, not to desiccate the dentine structure. The total etch technique was performed, applying 36% phosphoric acid (DeTrey Conditioner 36, Dentsply Caulk) to the canal walls surfaces for 15 s. The conditioned areas were rinsed thoroughly with water for at least 15 s. Water was removed from the rinsed canal with a soft blow of air and paper points. A moist surface was left, to avoid desiccating the dentine. Once the surfaces were treated, they were carefully kept uncontaminated. An equal number of drops of Prime&Bond NT Dual-Cure and Self Cure Activator (Dentsply Caulk) were mixed in a clean mixing well for 2 s. The activated selfpriming dental adhesive was dispensed onto a disposable micro-brush and immediately rubbed on all root canal surfaces for at least 20 s. The solvent was removed by blowing 84

gently with air from a dental syringe for at least 5 s. Paper points were introducted into the root canal to remove bonding excess. The mixed adhesive/activator was applied on the post, which was then dried with air from a dental syringe for at least 5 s. The posts were then luted with a dual-cured resin cement (Calibra, Dentsply Caulk in group 7 and Variolink II, Ivoclar in group 8), following the manufacturerâ&#x20AC;&#x2122;s instructions. The luting agents were mixed for 10 s and applied to the post with a microbrush and to the post-space with a periodontal probe. The posts were seated into the root canals and stabilized. The cement excess was cleaned up with paper points. The resin cement and the adhesive material were simultaneously light-cured for 40 s using a halogen light-curing unit (Optilux 501, Demetron, Kerr, Danbury, USA) at 800 mW/cm2 with the tip of the light unit directly in contact with the post. In order to ensure appropriate light intensity, the emitted light was measured before each exposure with the digital radiometer of the light unit. The composite restorations were bonded using a universal self-priming dental adhesive (Prime&Bond NT Dual-Cure, Dentsply Caulk) according to the bonding protocol previously described. The enamel was etched for at least 15 s. Transparent matrices (Hawe Striproll, KerrHawe, Bioggio, Switzerland) were used in order to hold the restorative materials. A standardized incremental composite build-up technique was employed, by light-curing the resin composites at 2 mm increments. Each increment was light-cured for 40 s on each aspect of the specimens, ensuring uniform exposure of all cavity surfaces according to the morphology of the restoration. A simplified anatomical build-up technique was used to restore all the crowns. Each tooth was embedded in a block of self-curing acrylic resin (Jet Kit, Lang Dental MFG. CO., Wheeling, IL, USA) surrounded by steel cylinders with the long axis perpendicular to the base of the block, leaving at least half of the root exposed so as to evaluate morphologically the eventual root fractures. In order to dissipate the heat generated from the polymerization reaction of the resin, the specimens were continuously 85

moistened with water spray. A thin layer of glycerin was applied with a microbrush on dental roots, in order to remove the specimens before resin completed the polymerization. After the first signs of polymerization, the teeth were carefully removed manually from the rein blocks. An addition-cure polyvinylsiloxane impression material (Flexitime, Heraeus Kulzer, Hanau, Germany) was injected into the acrylic resin molds and the specimens were inserted again into the steel cylinders filled with resin, in order to simulate periodontal ligaments through standardized silicone layers with a thickness of 200 to 400 Îźm. As previously described, in order to simulate the oral environment, the specimens were stored in an incubator at 37Â°C in 90% humidity until the execution of static fracture resistance measurement; the storage time ranged from a minimum of 24 h to a maximum of 36 h. A universal loading machine (Triaxal Tester T400 Digital, Controls srl, Cernusco s/N., Italy) was used for this purpose. Each specimen was inserted into the holding device with an inclination of 90Â° in relation to the horizontal plane (fig. 2). A controlled load was applied by means of a stainless steel rod with a tip diameter of 1 mm in a direction parallel to the longitudinal axis of the tooth. Pressure from the rod tip was applied at the level of the palatal cusp, 2 mm from the tip of the cusp towards the central fossa, in order to simulate an occlusal load. The load was applied at a crosshead speed of 1 mm/min. All samples were loaded until fracture and the maximum breaking loads were recorded in Newtons (N) by a computer (Digimax Plus, Controls srl, Cernusco s/N., Italy) connected to the loading machine. After mechanical failure, all the samples were immersed in Indian ink, in order to highlight the fracture lines. The mode of failure was recorded and classified as restorable or unrestorable. The failure mode was macroscopically evaluated at sight and the restorations were classified as unrestorable with the appearance of root fractures. In case of micro-cracks, the fracture pattern was evaluated using a 10x stereomicroscope (Zeiss OpMi1, Zeiss, Oberkochen, Germany). 86

The data were statistically analyzed with the software SPSS 12.0 (SPSS Inc., Chicago, IL, USA). The Kolmogorov-Smirnov test was used to verify the normality of the data distribution. The data recorded were analyzed with one-way analysis of variance (ANOVA) followed by Tukey post hoc test for comparisons to reveal the influence of the different restorative system on the fracture resistance. For all the statistical tests, the level of significance was set at p = 0.05.

Results The highest mean fracture resistance was recorded for group 1 (control group) at 502.43±152.47 N, followed by group 7 at 470.55±118.98 N, group 8 at 445.55±77.48 N, group 6 at 440.82±102.68 N, group 5 at 405.07±92.65 N, group 4 at 363.95±155.72 N, group 2 at 316.97±131.45 N and group 3 at 301.99±88.95 N (fig. 3). The one-way ANOVA and the Tukey post hoc test revealed that the groups restored only with flowable composites (group 2 and 3) reached significantly lower fracture resistance values than group 1 (control) and group 7 did (p<0.05). As a consequence, the null hypothesis was rejecetd. On the contrary, the mean fracture loads of group 1 (control) and groups 4 to 8 were statistically comparable (p>0.05) (fig. 3). As to the mode of failure, no statistically significant differences (p>0.05) were noticed between the various material combinations used to restore the single rooted maxillary premolars (fig. 4). Nevertheless, the teeth restored with fiber posts of group 8 mainly showed restorable partial crown fractures. Catastrophic fractures occurred more frequently in group 3. In group 7 the same number of specimens experienced restorable and unrestorable fractures (Table 2). Catastophic failures were characterized by an oblique fracture line on the palatal surface of the teeth, extending from the coronal to the middle third of the root. In all the

groups, coronal failures of both the residual dental tissues and the composite restorations occurred at the level of the palatal cusp, where the arbitrary load was applied. In all the restored groups (group 2 to 8), both cohesive and adhesive fractures were observed. In particular, 31 cohesive (44.3%) and 39 adhesive failures (55.7%) were noticed. Cohesive fractures inside the composite bulk mainly occurred in group 2 whereas adhesive failures of the interfaces were mainly evident in group 3. In group 7 and 8, no specimens displayed the exposure of a fiber post after the fracture of the crown.

Discussion Clinicians often experience the clinical fracture of endodontically treated teeth (Assif & Gorfil 1994, Ferrari et al. 2000a, Akkayan & Gulmez 2002). In particular, several studies have been performed to assess the mechanical resistance to fracture of maxillary premolars (Ferrari et al. 2000a, Akkayan & Gulmez 2002, Trope et al. 1996, Salis et al. 1987, Ferrari & Scotti 2002, Mannocci et al 1999, Asmussen et al. 1999, Drummond et al. 1999, Akkayan & Caniklioglu 1998, Akkayan 2004, Ferrari et al. 2000b, Grandini et al. 2002, Grandini et al. 2005a, Yamada et al. 2004). Most studies have tried to identify the best technique and combination of materials to be used to increase the strength of the tooth-restoration complex (Ferrari et al. 2000a, Monticelli et al. 2004, Ferrari et al. 2000b, Hannig et al. 2005). Grandini et al. (2005a) demonstrated that restoration of endodontically treated teeth with fiber posts and direct resin composites is a treatment option, that in the short term conserves remaining tooth structure and results in good patient compliance. The present study was performed to evaluate the influence of both microhybrid and flowable resin composites in different combinations with or without translucent glass fiber posts on the fracture resistance and failure patterns of endodontically treated single rooted maxillary premolars with MOD cavities. As fracture resistance decreased restoring teeth only with flowable resin composites, the null hypothesis tested was rejected. 88

The post space was prepared according to the most common criteria reported in the literature, inserting the posts for Âž of root canal length (Goodacre & Spolnik 1995) and leaving at least 4 mm of gutta-percha at level of the apex to provide a reliable apical seal (Abramovitz et al. 2001). There is concern that eugenol-containing root canal sealers may weaken the adhesion between root dentin, luting agents and fiber posts (Ngoh et al. 2001) interfering with the polymerization of resin cements (Ngoh et al. 2001, Schwartz & Robbins 2004). Several studies were performed to investigate the effects of eugenol containing temporary cements on the bond strength of resin cements and post retention (Tjan & Nemetz 1992, Schwartz et al. 1998). Some studies proved that eugenol has no effect on the shear bond strength of resin adhesives (Schwartz & Robbins 2004). On the contrary, other studies found eugenol-containing sealers to reduce post retention (Tjan & Nemetz 1992). However, it has been documented that root canal dentin contamination can be avoided both mechanically and/or chemically: post space preparation can remove residual eugenol from the contaminated dentin layer (Schwartz & Robbins 2004) and 37% phosphoric acid etching of the canal walls and the use of alcohol and/or EDTA are effective in decontaminating the root canal dentin, restoring the retentive strength of resins to dentin (Schwartz & Robbins 2004). In the present study, after preparing mechanically the post space with calibrate drills, EDTA was used to decontaminate root canal dentin from residual eugenol. In most static mechanical fracture tests reported in the literature, the premolars were loaded on the palatal cusp at 130Â°-150Â° to the longitudinal axis of the tooth, generating a compressive load perpendicular to the cuspal side (Akkayan & Gulmez 2002, Yamada et al. 2004). On the contrary, during function, the occlusion generates non-axial forces resolved into their vectors along the cuspal side following the parallelogram of forces (Sorrentino et al. 2006). As a consequence, in the present study the specimens 89

were loaded in a direction parallel to the longitudinal axis of the tooth, in order to distribute the stresses more evenly between the residual dental tissues and the restorative material, simulating a physiological occlusion. Groups 7 and 8 showed the closest mechanical resistance to group 1 (control) and mean fracture loads lower than that of the control group were recorded in all the other experimental groups. In the region of premolars, a patient with no parafunctional habits generates a maximum biting load of about 350-500 N during function (Korioth & Versluis 1997). These values were confirmed by the mean failure load recorded in group 1 (control). A recent study (Sorrentino et al. 2006) demonstrated that MOD cavities represent the restorative condition in which the mechanical resistance is closest to that of caries-free single rooted maxillary premolars. In the present study, the specimens which reached a fracture load of at least 450 N showed unrestorable fractures; over this value the restorative system transfers the load intraradicularly leading to catastrophic root fractures. As a consequence, the use of adhesive restorations without prosthetic crowns may represent a risk for the long-term success in parafunctional patients. Only one in vivo study (Mannocci et al. 2002) proved that the short term clinical success rates of endodontically treated premolars restored with fiber posts and direct composite restorations were equivalent to those of teeth restored with full coverage metalceramic crowns. The mean fracture loads recorded showed that flowable composites alone were not sufficient to sustain occlusal loads; the same conclusion can be suggested for the MOD cavities restored only with microhybrid resin composites. Better results were obtained with the so called sandwich technique (Croll 2004): microhybrid resins provide stiffness that is advisable to achieve stable support of the restorations (Monticelli et al. 2004), whereas flowable composites are more elastic, easier to handle and provide better integration with the fiber post surface (Monticelli et al. 2004). Even though the presence of fiber posts does not strengthen restored teeth, their use usually leads to restorable fractures. From a 90

functional point of view, the more favourable fracture patterns observed in the fiber postretained restorations represent a paramount factor that is highly desirable for the long-term success of these restorations. The ink perfusion allowed to observe the propagation of microcracks from the load application area towards the inner portion of the composite bulk and the tooth-restoration adhesive interfaces, causing cohesive and adhesive fractures respectively. An adhesive interface represents the transition area between materials with different moduli of elasticity; as a consequence, the adhesive interfaces are considered the weak link of adhesive restorations which can fail under load, leading to the mechanical failure of the palatal cusps. Such kind of fracture pattern was due to the morphology of the MOD cavities, leaving limited amounts of residual tooth struture at level of the cervical margin of the specimens. The failure pattern analysis showed fracture lines that were almost parallel to the direction of the applied load. The crack propagation was caused mainly by the loading direction rather than the potentially weak sites (e.g. voids or contact contamination along composite increments) within the restorative system itself, confirming the optimal mechanical characteristics of the adhesive systems and the composite materials tested. All the cohesive fractures affected the composite restorations. This highlighted the optimal polymerization and mechanical behaviour of both the flowable and microhybrid composite materials tested, enabling a homogeneous load distribution and acting as stress adsorbers. The bonding agent failed where the shear stress prevailed over the compressive load. Within the limitations of this study, the following conclusions can be drawn: -

the mechanical resistance of endodontically treated maxillary premolars with MOD cavities was enhanced by the combined use of flowable and microhybrid resin composites in the so called sandwich technique;

resistance to fracture was associated with the presence of fiber posts; 91

in case of failure, the use of fiber posts contributed to notice restorable fractures, whereas the use of resin composites alone may cause catastrophic root fractures, leading to the extraction of the damaged teeth.

Clinically, direct restoration of endodontically treated premolars with fiber posts and resin composites without crown coverage might be considered a valid alternative to the less conservative and more expensive traditional prosthetic crowns in the short term period. Nevertheless, the use of adhesive techniques with no crown coverage represents a risk for the long term success of restorations in patients expressing high biting forces and in parafunctional patients. Further clinical trials should be conducted to validate the results of this in vitro study.

References 1. Abramovitz L, Lev R, Fuss Z, Metzger Z. The unpredictability of seal after post space preparation: a fluid transport study. Journal of Endodontics 2001;27:292-295. 2. Akkayan B. An in vitro study evaluating the effect of ferrule length on fracture resistance of endodontically treated teeth restored with fiber-reinforced and zirconia dowel systems. Journal of Prosthetic Dentistry 2004;92:155-162. 3. Akkayan B, Caniklioglu B. Resistance to fracture of crowned teeth restored with different post systems. European Journal of Prosthodontics and Restorative Dentistry 1998;6:13-18. 4. Akkayan B, Gulmez T. Resistance to fracture of endodontically treated teeth restored with different post systems. Journal of Prosthetic Dentistry 2002;87:431437. 5. Alhadainy HA, Abdalla AI. 2-year clinical evaluation of dentin bonding systems. American Journal of Dentistry 1996;9:77-79. 6. Asmussen E, Peutzfeldt A, Heitmann T. Stiffness, elastic limit, and strength of newer types of endodontic posts. Journal of Dentistry 1999;27:275-278. 7. Assif D, Gorfil C. Biomechanical considerations in restoring endodontically treated teeth. Journal of Prosthetic Dentistry 1994;71:565-567. 8. Assif D, Nissan J, Gafni Y, Gordon M. Assessment of the resistance to fracture of endodontically treated molars restored with amalgam. Journal of Prosthetic Dentistry 2003;89:462-465. 9. Baratieri LN, De Andrada MA, Arcari GM, Ritter AV. Influence of post placement in the fracture resistance of endodontically treated incisors veneered with direct composite. Journal of Prosthetic Dentistry 2000;84:180-184. 10. Belli S, Erdemir A, Ozcopur M, Eskitascioglu G. The effect of fibre insertion on fracture resistance of root filled molar teeth with MOD preparations restored with composite. International Endodontic Journal 2005;38:73-80. 11. Christensen GJ. When to use fillers, build-ups or posts and cores. Journal of the American Dental Association 1996;127:1397-1398. 12. Croll TP. The â&#x20AC;&#x153;sandwichâ&#x20AC;? technique. Journal of Esthetic and Restorative Dentistry 2004;16:210-212. 93

13. Drummond JL, Toepke TR, King TJ. Thermal and cyclic loading of endodontic posts. European Journal of Oral Sciences 1999;107:220-224. 14. Ferrari M, Mannocci F, Vichi A, Cagidiaco MC, Mjor IA. Bonding to root canal: structural characteristics of the substrate. American Journal of Dentistry 2000c;13:255-260. 15. Ferrari M, Scotti R. Fiber posts: characteristics and clinical applications. Milan, Italy: Masson 2002. 16. Ferrari M, Vichi A, Garcia-Godoy F. Clinical evaluation of fiber-reinforced epoxy resin posts and cast post and cores. American Journal of Dentistry 2000b;13:15B18B. 17. Ferrari M, Vichi A, Grandini S. Efficacy of different adhesive techniques on bonding to root canal walls: an SEM investigation. Dental Materials 2001a;17:422-429. 18. Ferrari M, Vichi A, Grandini S, Goracci C. Efficacy of a self-curing adhesive-resin cement system on luting glass-fiber posts into root canals: an SEM investigation. International Journal of Prosthodontics 2001b;14:543-549. 19. Ferrari M, Vichi A, Mannocci F, Mason PN. Retrospective study of the clinical performance of fiber posts. American Journal of Dentistry 2000a;13:9B-13B. 20. Foxton RM, Nakajima M, Tagami J, Miura H. Adhesion to root canal dentine using one and two-step adhesives with dual-cure composite core materials. J of Oral Rehabilitation 2005;32:97-104. 21. Freedman GA. Esthetic post-and-core treatment. Dental Clinics of North America 2001;45:103-116. 22. Gateau P, Sabek M, Dailey B. Fatigue testing and microscopic evaluation of post and core restorations under artificial crowns. Journal of Prosthetic Dentistry 1999;82:341-347. 23. Goodacre CJ, Spolnik KJ. The prosthodontic management of endodontically treated teeth: a literature review. Part III. Tooth preparation considerations. Journal of Prosthodontics 1995;4:122-128. 24. Goracci C, Tavares AU, Fabianelli A et al. The adhesion between fiber posts and root canal walls: comparison between microtensile and push-out bond strength measurements. European Journal of Oral Sciences 2004;112:353-361. 94

25. Grandini S, Balleri P, Ferrari M. Scanning electron microscopic investigation of the surface of fiber posts after cutting. J of Endodontics 2002;28:610-612. 26. Grandini S, Goracci C, Tay FR, Grandini R, Ferrari M. Clinical evaluation of the use of fiber posts and direct resin restorations for endodontically treated teeth. International Journal of Prosthodontics 2005a;18:399-404. 27. Grandini S, Goracci C, Monticelli F, Tay FR, Ferrari M. Fatigue resistance and structural characteristics of fiber posts: three-point bending test and SEM evaluation. Dental Materials 2005b;21:75-82. 28. Hannig C, Westphal C, Becker K, Attin T. Fracture resistance of endodontically treated maxillary premolars restored with CAD/CAM ceramic inlays. Journal of Prosthetic Dentistry 2005;94:342-349. 29. Hernandez R, Bader S, Boston D, Trope M. Resistance to fracture of endodontically treated premolars restored with new generation dentine bonding systems. International Endodontic Journal 1994;27:281-284. 30. Kemp-Scholte CM, Davidson CL. Marginal integrity related to bond strength and strain capacity of composite resin restorative systems. Journal of Prosthetic Dentistry 1990;64:658-664. 31. Korioth TW, Versluis A. Modeling the mechanical behavior of the jaws and their related structures by finite element (FE) analysis. Critical Reviews in Oral Biology and Medicine 1997;8:90-104. 32. Kraus BS, Jordan RE, Abrams L. A study of the masticatory system â&#x20AC;&#x201C; Dental anatomy and occlusion. Baltimore, USA: Williams & Wilkins 1969. 33. Krejci I, Duc O, Dietschi D, de Campos E. Marginal adaptation, retention and fracture resistance of adhesive composite restorations on devital teeth with and without posts. Operative Dentistry 2003;28:127-135. 34. Loney RW, Moulding MB, Ritsco RG. The effect of load angulation on fracture resistance of teeth restored with cast post and cores and crowns. International Journal of Prosthodontics 1995;8:247-251. 35. Mannocci F, Bertelli E, Sherriff M, Watson TF, Ford TR. Three-year clinical comparison of survival of endodontically treated teeth restored with either full cast

coverage or with direct composite restoration. Journal of Prosthetic Dentistry 2002;88:297-301. 36. Mannocci F, Ferrari M, Watson TF. Intermittent loading of teeth restored using quartz fiber, carbon-quartz fiber, and zirconium dioxide ceramic root canal posts. Journal of Adhesive Dentistry 1999;1:153-158. 37. Monticelli F, Goracci C, Ferrari M. Micromorphology of the fiber post-resin core unit: a scanning electron microscopy evaluation. Dental Materials 2004;20:176-183. 38. Morris MD, Lee KW, Agee KA, Bouillaguet S, Pashley DH. Effects of sodium hypochlorite and RC-prep on bond strengths of resin cement to endodontic surfaces. Journal of Endodontics 2001;27:753-757. 39. Nakabayashi N, Pashley DH. Hybridization of dental hard tissue. Berlin, Germany: Quintessence 1998. 40. Ngoh EC, Pashley DH, Loushine RJ, Weller RN, Kimbrough F. Effects of eugenol on resin bond strengths to root canal dentin. Journal of Endodontics 2001;27:411414. 41. Ortega VL, Pegoraro LF, Conti PC, do Valle AL, Bonfante G. Evaluation of fracture resistance of endodontically treated maxillary premolars, restored with ceromer or heat-pressed ceramic inlays and fixed with dual-resin cements. Journal of Oral Rehabilitation 2004;31:393-397. 42. Salis SG, Hood JA, Stokes AN, Kirk EE. Patterns of indirect fracture in intact and restored human premolar teeth. Endodontics & Dental Traumatology 1987;3:10-14. 43. Schwartz

RS,

Murchison

DF,

Walker

WA.

Effects

eugenol

and

noneugenolendodontic sealer cements on post retention. Journal of Endodontics 1998;24:564-567. 44. Schwartz RS, Robbins JW. Post placement and restoration of endodontically treated teeth: a literature review. Journal of Endodontics 2004;30:289-301. 45. Seow LL, Toh CG, Wilson NH. Remaining tooth structure associated with various preparation designs for the endodontically treated maxillary second premolar. European Journal of Prosthodontics and Restorative Dentistry 2005;13:57-64.

46. St.Georges AJ, Sturdevant JR, Swift EJ Jr, Thompson JY. Fracture resistance of prepared teeth restored with bonded inlay restorations. Journal of Prosthetic Dentistry 2003;89:551-557. 47. Stockton L, Lavelle CL, Suzuki M. Are posts mandatory for the restoration of endodontically treated teeth? Endodontics & Dental Traumatology 1998;14:59-63. 48. Strub JR, Pontius O, Koutayas S. Survival rate and fracture strength of incisors restored with different post and core systems after exposure in the artificial mouth. J of Oral Rehabilitation 2001;28:120-124. 49. Tjan A, Nemetz H. Effect of eugenol-containing endodontic sealer on retention of prefabricated posts luted with an adhesive composite resin cement. Quintessence International 1992;22:839-844. 50. Tronstad L, Asbjornsen K, Doving L, Pedersen I, Eriksen HM. Influence of coronal restorations on the periapical health of endodontically treated teeth. Endodontics & Dental Traumatology 2000;16:218-221. 51. Trope M, Langer I, Maltz D, Tronstad L. Resistance to fracture of restored endodontically treated premolars. Endodontics & Dental Traumatology 1986;2:3538. 52. Uyehara MY, Davis RD, Overton JD. Cuspal reinforcement in endodontically treated molars. Operative Dentistry 1999;24:364-370. 53. Van Meerbeek B, Lambrechts P, Inokoshi S, Braem M, Vanherle G. Factors affecting adhesion to mineralized tissues. Operative Dentistry 1992;Suppl 5:111124. 54. Versluis A, Douglas W, Cross M, Sakaguchi RL. Does an incremental filling technique reduce polymerization shrinkage stresses? Journal of Dental Research 1996;75:871-878. 55. Wagnild GW, Mueller KI. Restoration of the endodontically treated tooth. In: Cohen S, Burns RC, eds. Pathways of the Pulp, 8th ed. St Louis, MO, USA: Mosby Inc., 2002; 765-795. 56. Yamada Y, Tsubota Y, Fukushima S. Effect of restoration method on fracture resistance of endodontically treated maxillary premolars. International Journal of Prosthodontics 2004;17:94-98. 97

MATERIALS AND MANUFACTURERS

Flowable resin composite

Microhybrid resin composite

Fiber post

Group 1

X-Flow (Dentsply Caulk)

Esthet-X (Dentsply Caulk)

-----

Group 2

Aeliteflo (Bisco, Inc.)

-----

Group 3

Tetric-flow (Ivoclar-Vivadent)

-----

Group 4

-----

Light-Core (Bisco, Inc.)

-----

Group 5

-----

Tetric-ceram (Ivoclar-Vivadent)

-----

Group 6

Tetric-flow (Ivoclar-Vivadent)

Tetric-ceram (Ivoclar-Vivadent)

-----

Group 7

Aeliteflo (Bisco, Inc.)

Light-Core (Bisco, Inc.)

DT Light Post (RTD)

Group 8

Tetric-flow (Ivoclar-Vivadent)

Tetric-ceram (Ivoclar-Vivadent)

FRC Postec (Ivoclar-Vivadent)

Table 1 â&#x20AC;&#x201C; Products used and manufacturers.

FRACTURE PATTERNS

GROUP

AD + RE

AD + UN

CO + RE

CO + UN

-----

Table 2 â&#x20AC;&#x201C; Fracture pattern distribution (AD: adhesive; CO: cohesive; RE: restorable; UN: unrestorable).

Fig. 1 â&#x20AC;&#x201C; Preparations of the specimens: pulp chamber opening and cavity preparation designs.

Fig. 2 â&#x20AC;&#x201C; Static mechanical test to fracture: loading direction.

100

Fig. 3 â&#x20AC;&#x201C; Fracture load-group histogram: fracture values are reported in Newtons (N). The same symbols indicate a statistically significant difference between groups.

Fig. 4 â&#x20AC;&#x201C; Mode of failure: restorable (A) and unrestorable fracture (B).

101

CHAPTER III

Investigating the mechanical resistance to fracture of endodontically-treated molars simulating different restorative systems.

102

Fracture Resistance and Failure Patterns of Endodontically Treated Mandibular Molars Restored Using Resin Composite With or Without Translucent Glass Fiber Posts.

Salameh Z., Sorrentino R., Papacchini F., Ounsi H.F., Tashkandi E., Goracci C., Ferrari M. â&#x20AC;&#x201C; Journal of Endodontics, 2006; 32(8): 752-755.

103

ABSTRACT The elastic modulus of the restorative material is important in restoring endodontically treated teeth. This study aimed to compare the fracture resistance and failure patterns of 90 mandibular molars restored using resin composites with or without fiber posts, with respect to the number of residual cavity walls. Five restoration types were performed corresponding to different wall defects (groups 1-5). Groups were divided in two subgroups corresponding to the use or absence of fiber posts. Teeth were loaded and resistance of specimens was measured as the axial compressive load to cause fracture and macroscopic fracture patterns were observed. One way ANOVA revealed a significant difference in fracture resistance (p _ 0.001). Tukey post hoc test also revealed significant differences between groups as samples restored with fiber posts exhibited mostly restorable fractures. It was concluded that the resistance of endodontically treated mandibular molars restored with composite resins is mainly affected by the number of residual walls. Using fiber-reinforced posts optimized fracture patterns.

104

Introduction Endodontically treated teeth (ETT) are considered to have a higher risk of fracture because of their inherently poor structural integrity as a result of pre-existing caries and/or tooth preparation (1, 2). Loss of the roof of the pulp chamber and/or the marginal ridges is further factors that are likely to influence the fracture resistance of such teeth (3). The fracture potential of ETT have been studied, yet to date, no definite causal relationship between fracture and the type of restoration has been established, and controversies remain regarding which material or technique would be ideal for their rehabilitation. Although posts are necessary to retain coronal build-up materials, they do not reinforce roots and may even weaken them through loss of radicular dentin necessitated by post-space preparation (4). Furthermore, and particularly regarding prefabricated posts, the interfaces between materials of different moduli of elasticity represent areas of weakness as local discrepancies influence stress-strain distribution (4). In this respect, posts with similar biomechanical properties to dentin, viz. carbon fiber-reinforced posts, were developed (5â&#x20AC;&#x201C;7). These were followed by translucent glass or quartz fiber-reinforced posts with better esthetic properties. A major advantage of such posts is the possibility of using composite restorative materials to rebuild missing coronal structure, which offers better interfacial integrity through the use of materials of similar elastic moduli (8, 9). Furthermore, since chemical bonding is not possible with the matrices used in post fabrication, chemical bonding can be achieved between post fibers and core material by applying a silane agent (10). Etching of the post surface is also possible using 10% hydrogen peroxide, thus increasing the surface area and improving the micromechanical retention at the post-core interface (11). The aim of the present study was to compare the fracture resistance and failure patterns of endodontically treated mandibular molars restored using resin composites with or without translucent glass fiber posts, and with respect to the number of residual cavity walls. The null hypothesis tested was that there is 105

no association between the fracture resistance (and patterns) of endodontically treated mandibular molars restored by means of resin composite with or without glass fiberreinforced posts, and the number of residual cavity walls remaining coronally.

Materials and Methods Ninety human mandibular first and second molars, extracted for periodontal reasons, were selected. Teeth with caries and/or previous restorations were excluded. Dental plaque, calculus and periodontal tissues were removed. The teeth were stored in 0.9% saline solution at 37Â°C. Canal morphology was verified from standardized apical radiographs (70 kV and 0.08 s) both in the mesio-distal and bucco-lingual directions. The pulp chamber of each tooth was opened and working length was determined visually by placing a size #10 K-file (Dentsply-Maillefer, Ballaigues, Switzerland) at the apical foramen. Root canals were instrumented using stainless steel K-files #10, 15, 20 (Dentsply-Maillefer) followed by rotary Ni-Ti instruments (ProTaper, Dentsply- Maillefer) according to the manufacturerâ&#x20AC;&#x2122;s instructions. All canals were prepared to the F2 size and instruments discarded after use in four root canals or if instrument deformation was visible. Root canals were irrigated between instrumentation with 2ml 5.25% sodium hypochlorite. All teeth were obturated using the warm vertical condensation technique, using calibrated gutta-percha points (F2, Dentsply- Maillefer) and an endodontic sealer (AH26, DentsplyMaillefer). To account for the influence of root canal morphological variations on the results, teeth were classified according to their mesio-distal and bucco-lingual dimensions and proportionately distributed among the experimental groups so as to have similar representation of morphologies within them. Experimental group 1 comprised 10 teeth (control group) while groups 2 to 5 comprised 20 teeth each. The specimens were prepared as follows (Fig. 1a): 106

- Group 1 (control group): the pulp chamber was filled with a flowableresin composite material (X-Flow, Dentsply-Caulk, York, PA) and a micromatrix resin composite material (Ceram X, Dentsply-Caulk); all coronal walls were left intact; - Group 2: the distal wall of each tooth was removed, using the limits of the marginal crest as an anatomic reference. The cavity was extended towards the access preparation to create a divergent disto-occlusal standardized adhesive preparation. The cervical margin was placed 1 mm apical to the CEJ; - Group 3: both the distal and mesial walls of each tooth were removedto create a mesio-occluso-distal (MOD) cavity; the same preparation criteria used in group 2 was adopted; - Group 4: the distal, mesial and buccal walls of each tooth were removed; the same preparation criteria used in previous groups were adopted; - Group 5: the whole crown of each tooth was removed 1 mm coronally to the most occlusal point of the CEJ. Groups 2 through 5 were divided into two subgroups, designated a and b (n_10 each). Subgroups 2a through 5a were restored with an approximately 2 mm thick layer of flowable resin composite material (X-Flow, Dentsply-Caulk), followed by several layers of micromatrix resin composite material (Ceram X, Dentsply-Caulk). In subgroups 2b through 5b, the coronal build-up was preceded by placement of a translucent glass fiber post (DT Light Post, RTD, St. Egreve, France). Each post was tried into the root canal and cut to adequate length with a diamond bur so as to cover its occlusal end with at least 2 mm of composite resin. The post surface was silanized with Calibra Silane (Dentsply-Caulk) for 60 s. The canal walls were etched with 36% phosphoric acid for 15 s, and then rinsed and dried with paper points. Prime & Bond NT Dual-Cure (Dentsply-Caulk) was used as an adhesive and light-cured for 20 s using a halogen light-curing unit (Astralis 10, IvoclarVivadent, Schaan, Liechtenstein) at 750 mW/cm2, before luting the posts with a dual-cure 107

resin cement (Calibra, Dentsply-Caulk), according to the manufacturerâ&#x20AC;&#x2122;s instructions. Light curing was performed through the post for 40 s. Using transparent matrices (Hawe Striproll, Kerr-Hawe, Bioggio, Switzerland), a standardized incremental composite build-up technique was used, consisting of light-curing for 40 seconds of each 2 mm of resin composite increment. A simplified anatomic build-up technique was used to restore the occlusal surface of the crowns. Each tooth was embedded in a block of self-curing acrylic resin (Orthoresin, Lang Dental MFG., Co., Wheeling, IL) using a silicone mold, leaving 2 to 3 mm of the root exposed so as to morphologically evaluate the eventual root fractures. A 0.5-mm layer of polyvinylsiloxane impression material (Flexitime, Heraeus Kulzer, Hanau, Germany) was applied in the root region to simulate the periodontal ligament before embedding the tooth. Specimens were stored for less than 1 week in distilled water at room temperature before testing. A universal loading machine (Triaxal Tester T400 Digital, Controls srl, Cernusco s/N., Italy) was used for evaluating static fracture resistance. Each specimen was inserted vertically into the holding device and a stainless steel rod having a 3 mm tip diameter was used to apply the controlled load in a direction parallel to the longitudinal axis of the tooth. The point of load application was 2 mm from the tip of the buccal cusp towards the central fossa, to simulate occlusal load. Crosshead speed was 1 mm/minute, and all samples were loaded until fracture while maximum breaking loads were recorded in Newtons (N) by a computer (Digimax Plus, Controls srl) connected to the loading machine. Fracture resistance of the test specimens was specifically measured as the axial compressive load to cause fracture and determined by noting an evident load drop with the mechanical testing machine. Macroscopic fracture patterns were observed after ink perfusion to highlight fracture lines, photographs were taken using a digital camera, and the mode of failure was classified as restorable or unrestorable (fractures were classified as unrestorable if root fractures occurred). Data were statistically analyzed with SPSS 12.0 108

(SPSS, Inc., Chicago, IL). The Kolmogorov-Smirnov test was used to verify the normality of the data distribution. The one-way ANOVA was then used, followed by Tukey post hoc test for multiple comparisons; p was set to 0.05 for all statistical tests.

Results According to the Kolmogorov-Smirnov test, the data had a normal distribution that allowed for further statistical analyses. One way ANOVA revealed that the difference in fracture resistance of the specimens was statistically significant (p _ 0.001). Mean fracture resistances, standard deviations, and Tukey post-hoc test results are given in Table 1. Results of modes of failure (restorable, Fig. 1b; unrestorable, Fig. 1c) are given in Table 2.

Discussion This study was designed to assess the fracture resistance and pattern of failure of mandibular molars restored using microhybrid resin composite with or without translucent fiber-reinforced posts. The study also attempted to take into consideration the degree of destruction of the crown before restoration. Given the finding that the number of residual cavity walls and the resistance to fracture was related, as was the fracture pattern and the presence or absence of fiber-reinforced posts, the null hypothesis was rejected. The use of resin-based cement in this experimental design was intended to circumvent the potentially detrimental influence that eugenol- containing root canal sealers have on the adhesion between root dentin, luting agents and fiber posts. However, the potential negative influence of sodium hypochlorite on bond strength was not taken into account. It has elsewhere been shown that the presence of a periodontal analogue is of importance in fracture testing, resulting in significant modifications in modes of fracture (10). The present study took this into consideration by adding a layer of silicone simulating the periodontium. The choice of load direction (parallel to the long axis of the tooth) was also designed to 109

simulate physiological function and to obtain a degree of nonaxial loading through existing occlusal contact variations. For this reason, it was not deemed necessary to have the loading tip contact simultaneously the two inner cuspal sides as this would have generated a wedge effect that might have skewed the results (11). The forces placed on the dentition during normal masticatory function are generally small compared to the maximal biting force. Anderson (12, 13) was the first to measure loads on mandibular molars using strain gauges and found that the maximum whole tooth load varied between 7.2 and 14.9 kg (70.6 and 146 N) when eating meat, biscuit or carrots. De Boever et al. (14) reported forces of between 2.4 and 7.2 kg (23.5 and 70.6 N) using transmitters in removable pontics, and concluded that functional chewing forces are variable from session to session and change with the consistency and viscosity of the food. More recently, maximum biting force on the first molar was reported as approximately 859 N (15), and elsewhere as (878 N) (16). The mean fracture load recorded in this study for the control group was 1198 N that is higher than both the maximum chewing and biting loads reported. Groups 2 and 3 also displayed fracture resistances that were greater than the maximum loads. Groups 4 and 5, however, showed fracture resistance values in the range of maximal biting loads but greater than physiological masticatory forces. Thus, it may be suggested from these data that crown coverage is not necessary in molars restored with composite resins. However, this study did not take into account the effect of aging of dental bonds (17), longterm behavior of such restorations (18), or the influence of parafunctional habits (19). No significant differences between sub-groups, representing the influence of a post for a given coronal restoration, were noted. Significant differences did emerge, however, among the groups, whereby the more the residual walls, the higher the resistance, with the exception of group 1 that exhibited a significantly lower fracture resistance than group 2a. Data regarding compressive and flexural properties of the posts and core materials used were not available, and thus the observed behaviors of specimens under loading 110

conditions could not be interpreted with regard to the relative differences in mechanical properties of post material, composite resins and tooth structure (20). However, it could be suggested that the influence of cavity design, as reflected by the behavior of multi-walled restorations (groups 1 and 2) is important. However, corroborating evidence for this is needed through further investigations. The ink perfusion produced an interesting finding. Contrary to the findings with metallic posts, it would appear that the use of fiber-reinforced posts has a positive effect on the fracture pattern, resulting in most fractures being restorable. Recently, it has been found that post geometry can significantly affect post retention (21, 22), and there is every reason to suppose that a variation of the geometry of posts used in this study could have produced a different outcome. Within the limitations of this study, it can be concluded that the resistance to fracture of endodontically treated mandibular molars restored with composite resins is mainly affected by the number of residual coronal walls. More walls are clearly beneficial, although it also seems that one wall could be sacrificed to compensate for the C-factor. Fracture resistance is not affected by the presence or absence of fiber reinforced posts. While coronal coverage may remain the recognized standard of care for posterior ETT that are also subjected to parafunctionalforces, the findings suggest that many such teeth that are notsubjected to heavy occlusal forces, may be adequately restored with bonded resin composites. In such cases, the use of posts seems to optimize fracture patterns and so facilitate re-treatment. Further research is still necessary to investigate the longevity of such restorations especially in clinical conditions, and the possible influence of parafunction.

111

References 1. Tidmarsh BG. Restoration of endodontically treated posterior teeth. J Endod 1976;2:374 –375. 2. Eakle WS, Maxwell EH, Braly BV. Fracture of posterior teeth in adults. J Am Dent Assoc 1986;112:215–21 8. 3. Reeh ES, Douglas WH, Messer HH. Stiffness of endodontically treated teeth related to restoration technique. J Dent Res 1989;68:1540–1544. 4. Assif D, Gorfil C. Biomechanical considerations in restoring endodontically treated teeth. J Prosthet Dent 1994;71:565–567. 5. Ferrari M, Vichi A, Mannocci F, Mason PN. Retrospective study of the clinical performance of fiber posts. Am J Dent 2000;13:9B–13B. 6. Fredriksson A, Astback J, Pamenius M, Arvidson K. A retrospective study of 236 patients with teeth restored by carbon fiber reinforced epoxy resin posts. J Prosthet Dent 1998;80:151–157. 7. Sidoli GE, King PA, Setchell DJ. An in vitro evaluation of carbon fiber based post and core system. J Prosthet Dent 1997;78:5–9. 8. Trope M, Langer I, Maltz D, Tornstad L. Resistance to fracture of endodontically treated premolars. Endo Dent Traumatol 1986;2:35– 38. 9. Steele A, Johnson BR. In vitro fracture strength of endodontically treated premolars. J Endod 1999;25:6–8. 10. Soares CJ, Pizi ECG, Fonseca RB, Martins LRM. Influence of root embedment material and periodontal ligament simulation on fracture resistance tests. Braz Oral Res 2005;19:11–16. 11. Korioth TWP, Versluis A. Modeling the mechanical behavior of the jaws and their related structures by finite element analysis. Crit Rev Oral Biol Med 1997;8:90–104.

112

12. Anderson DJ. Measurements of stress in mastication 1. J Dent Res 1956;135:664 – 670. 13. Anderson DJ. Measurements of stress in mastication 2. J Dent Res 1956;135:671– 673. 14. De Boever JA, McCall WD, Holden S, Ash MM. Functional occlusal forces: an investigation by telemetry. J Prosthet Dent 1978;40:326 –333. 15. Cosme DC, Baldisserotto SM, Canabarro Sde A, Shinkai RS. Bruxism and voluntary maximal bite force in young dentate adults. Int J Prosthodont 2005;18:328–332. 16. Ahlberg JP, Kovero OA, Hurmerinta KA, Zepa I, Nissinen MJ, Kononen MH. Maximal bite force and its association with signs and symptoms of TMD, occlusion, and body mass index in a cohort of young adults. Cranio 2003;21:248–252. 17. Ferrari M, Mason PN, Goracci C, Pashley DH, Tay FR. Collagen degradation in endodontically treated teeth after clinical function. J Dent Res 2004;83:414–419. 18. Armstrong SR, Vargas MA, Chung I, et al. Resin-dentin interfacial ultrastructure and microtensile dentin bond strength after five-year water storage. Oper Dent 2004;9:705– 712. 19. Fernandes AS, Dessai GS. Factors affecting the fracture resistance of post-core reconstructed teeth: a review. Int J Prosthodont 2001;14:355– 363. 20. Galhano GA, Valandro LF, de Melo RM, Scotti R, Bottino MA. Evaluation of the flexural strength of carbon fiber-, quartz fiber-, and glass fiber-based posts. J Endod 2005;31:209 –211. 21. Goracci C, Fabianelli A, Sadek FT, Papacchini F, Tay FR, Ferrari M. The contribution of friction to the dislocation resistance of bonded fiber posts. J Endod 2005;31:608–612. 22. Pirani C, Chersoni S, Foschi F, et al. Does hybridization of intraradicular dentin really improve fiber post retention in endodontically treated teeth? J Endod 2005;31:891–894.

113

Walls left (group)

No post (subgroups a)

Post (subgroups b)

Mean

0 walls (5)

833.4c

179.7

677.2c

21.3

1 wall (4)

653.1c

36.5

706.8c

130.8

2 walls (3)

1195.2a

332.9

878.2ac

135.6

3 walls (2)

1606.8b

410.5

1512.2ab

267.0

4 walls (1)

1197.9a

363.5

-----

Table 1. Means and Standard Deviations of Fracture Resistance (in Newtons). Different letters indicate statistically significant differences. For the number of residual walls, the Tukey post hoc test showed that between groups 4a, 4b, 5a, and 5b, differences in the fracture resistance of the teeth were not statistically significant (p_0.001), but the differences were statistically significant with group 1, 2a, 2b, 3a, with group 3a significantly different from group 2a, 2b, and group 1 significantly different from 2a.

Walls left (group)

No post (subgroups a)

Post (subgroups b)

Unrestorable

Restorable

Unrestorable

Restorable

0 walls (5)

40%

60%

100%

1 wall (4)

40%

60%

100%

2 walls (3)

60%

40%

30%

70%

3 walls (2)

70%

30%

50%

4 walls (1)

60%

40%

-----

Table 2. Modes of failures Samples restored with fiber posts mostly exhibited restorable fractures while teeth restored without fiber posts mostly exhibited unrestorable fractures (Fig. 1b). This was particularly true for groups with one or no walls left (groups 4 and 5) were subgroups b showed 100% restorable fractures (Fig. 1c).

114

Fig.1 â&#x20AC;&#x201C; (a) Schematic drawing representing the loss of dentinal walls in each group. (b) Sample from group 3a exhibiting a nonrestorable vertical fracture extending in the mesial root. (c) Sample from group 5b exhibiting a restorable fracture. The acrylic resin was removed with a carbide bur to show that radicular fracture did not occur. It appears clearly that the failure occurred rather in the restorative material before root fracturing stress levels could be obtained.

115

CHAPTER IV

Assessing the validity of laboratory mechanical tests to investigate the resistance to fracture of endodontically-treated teeth by means of virtually simulated clinical situations.

116

Evaluation of the biomechanical behaviour of maxillary central incisors restored by means of endocrowns compared to a natural tooth: a linear static 3D Finite Elements Analysis.

Zarone F., Sorrentino R., Apicella D., Valentino B., Ferrari M., Apicella A. â&#x20AC;&#x201C; Dental Materials, 2006 Jan 10; Epub ahead of print.

117

ABSTRACT Objective: The present study aimed at evaluating different restoring configurations of a crownless maxillary central incisor, in order to compare the biomechanical behaviour of the restored tooth with that of a sound tooth. Materials and methods: A 3D FE model of a maxillary central incisor is presented. An arbitrary static force of 10 N was applied with an angulation of 125Â° to the tooth longitudinal axis at level of the palatal surface of the crown. Different material configurations were tested: composite, syntered alumina, feldspathic ceramic endocrowns and glass post resorations with syntered alumina and feldspathic ceramic crown. Results: High modulus materials used for the restoration strongly alter the natural biomechanical behaviour of the tooth. Critical areas of high stress concentration are the restoration-cement-dentin interface both in the root canal and on the buccal and lingual aspects of the tooth-restoration interface. Materials with mechanical properties underposable to that of dentin or enamel improve the biomechanical behaviour of the restored tooth reducing the areas of high stress concentration. Significance: The use of endocrown restorations present the advantage of reducing the interfaces of the restorative system. The choice of the restorative materials should be carefully evaluated. Materials with mechanical properties similar to those of sound teeth improve the reliability of the restoartive system.

118

Introduction Restoration of endodontically treated teeth is a common problem in restorative dentistry, related to the fractures occurring in such teeth [1][2]. Endodontically-treated teeth are affected by a higher risk of biomechanical failure than vital teeth [3][4]. The access preparation for endodontic treatment causes the loss of the roof of the pulp chamber: this may account for the relatively high fracture incidence documented in pulpless teeth [5]. Posts have been often described not to reinforce endodontically treated teeth [6]. Moreover, some authors noticed that posts may interfere with the mechanical resistance of teeth, increasing the risk of damage to residual tooth structure [7]. Their role of maintaining the core material is particulary relevant for posterior teeth, where masticatory loads are essentially compressive [8]. However, when loaded transversely, as in the case of incisors, the flexural behaviour of posts should be carefully considered [9]. The magnitude and the angle of incisal load greatly influence the long-term success of restorative systems involving central incisors. In particular, the geometry of the restoration is of less importance than the masticatory loading conditions [10]. In the complex systems of post restorations, the stress distribution within the structure is multiaxial, not uniform and depending on the magnitude and direction of the applied external loads. Type of external loading, geometry of structure and residual stresses are some of the causes of multiaxiality stress distribution [6]. The placement of an endodontic post creates an unnatural restored structure since it fills the root canal space with a material that has a defined stiffness unlike the pulp. Hence it is not possible to recreate the original stress distribution of the tooth. Steel posts are the most dangerous for the root, potentially leading to its fracture. Even working on the cement layer stress adsorbing effect by using less rigid cements, it is not possible to improve the stress arising in the system because of the high rigidity of the steel post [7]. In a carbon post restoration, the elastic modulus of the cement layer strongly influences the stress 119

adsorbing effect of the system. The use of glass fiber posts results in the best stress distribution, since the cement layer rigidity is less relevant compared to steel and carbon post configurations [11]. To summarize, two parameters strongly influences the mechanical behaviour of endodontically treated teeth restored with posts: the characteristics of the interfaces and the rigidity of the materials. Some of the fractures affecting post restorations could be related to concentration of forces [12]; fatigue loading must be considered an additional cause of root fracture. Failures related to post restored teeth have been related to fatigue more than to maximal loading [13]. Different critical restoration regions and fracture patterns have been described to occur in fatigue testing on titanium and composite posts and amalgam cores [14]. On the contrary, static strength tests (i.e. maximum load at failure) performed on the same systems led to similar mechanical behaviour [14]. Several authors reported that specimens loaded for more than 105 cycles showed the formation of gaps at level of the core-tooth interface. As to composite posts, such gaps were induced by the deformation of the dentin interface, where load transfer from the post to dentin occurs. On the contrary, as to the more rigid amalgam and titanium systems, gaps were due to the deformation of the core at level of the facial aspect moving away from the tooth [7]. In order to assess all the eventual variables, the type of materials used for core build-up and crown restoration have to be carefully evaluated [15]. For many years, all-ceramic anterior crown restorations have been used as an alternative to metal-ceramic crowns [16]. Microcracks and fractures could affect traditional feldspathic porcelain, therefore reinforced ceramic core materials have been developed to support veneering ceramics. Alumina core crowns have shown high values of mechanical resistance to fracture [17]. Innovative CAD-CAM technologies have introduced new systems for dental restorations. In particular, CAD-CAM endocrown assembles the endocanalar-post, the core and the crown in one component. The CAD-CAM endocrowns can be built from both a single block 120

of feldspathic ceramics or a syntered-alumina block including a radicular portion and a ceramic coated core. Feldspathic endocrowns show a lower flexural strength [18] than syntered-alumina endocrowns [19]. The ongoing research for biocompatible materials with physico-mechanical properties similar to those of natural tooth tissues has introduced a new generation of composite blocks for CAD-CAM processing. To date, there is still no agreement in the literature about which material or technique can optimally restore endodontically-treated teeth [20]. Moreover, in the literature there are no data about the effect of endocrown restorations on the biomechanical behaviour of restored teeth. The aim of the present study was to compare the stress distribution patterns of a sound tooth with those of teeth restored with the following different material configurations: alumina endo-crowns, feldsphatic endocrowns, composite endocrowns and traditional crowns built up on glass post and core units. As to endodontically treated maxillary central incisors restored with post and core systems, three different null hypotheses were tested in the present study: 1) there is no association between the strain and stress distribution patterns and crown materials; 2) there is no association between the endocrown material and the biomechanical behaviour of restored teeth; 3) there is no association between the use of endocrowns with biomechanical properties similar to those of tooth tissues and the possibility to reproduce a physiological strain and stress distribution.

Materials and Methods A linear static structural analysis has been performed to calculate the stress distribution in different restoring configurations. An arbitrary load of 10 N was applied at 60Â° angle with 121

tooth longitudinal axis at the palatal surface of the crown to simulate tearing function (Figure 1). The static load of 10N has been defined in order to compare the systems in the field of low deformation where a linear-static analysis can be performed with reliable results. In order to identify areas of highest stress concentration where possible fatigue failure are more expected to occur, the choice of the pertinent stress representation criterion was based on the evaluation of failure predictive potential of the analysis performed. Von Mises (equivalent stresses) energetic criterion has been then chosen. Under fatigue loading, in fact, the calculated stresses should relate to fracture probability and, therefore, to the assumption that different stress states having the same effect are equivalent when determining the system failure at critical stress values (failure criterion). In such cases, accurate predictions have been observed [21][22][23].

Solid and FE models preparation The solid model was generated using literature data [24] for dentin and enamel internal volumes and morphologies, while the external shape of the maxillary central incisor was obtained by laser based 3D digitising (Cyberware) of a plaster cast (Thanaka manufacturer, Japan). The scanned profiles were assembled in a three dimensional wire frame structure using a 3D CAD (Autocad 12, Autodesk, Inc.) and exported into a 3D parametric solid modeller (Pro-Engineering 16.0 Parametric Technologies, USA). The tooth volumes were generated by fitting of the horizontal and vertical profiles. The geometries and volumes for the cement layer and post were also generated at this stage. The FEM model was obtained by importing the solid model into ANSYS rel. 9.0 FEM software (Ansys, Inc. Houston) using IGES format. The volumes were redefined in the new environment and meshed with 8 nodes brick with 3 degree of freedom per node, finally resulting in a model with 13272 elements and 15152 nodes (fig. 1). Accuracy of the model 122

has been checked by convergence tests. Particular attention has been devoted to the refinement of the mesh resulting from the convergence tests at the cement layer interfaces. Different material properties were coupled with the elements and geometries according to the volume material defined in figure 1 (enamel, dentin, restored crown, core, cement and post). The following seven models with different materials configuration were considered: Model 1 (MOD1): sound tooth, mechanical properties of dentine and enamel were assigned according to literature data [25][26][27] (Table 1) Model 2 (MOD2): endodontically treated maxillary central incisor restored with glass fiber post cemented with a high modulus dual-cure cement (Panavia, Kuraray, Japan). The same properties of the dual-cure cement were asigned to the core material. A feldspathic crown was considered. Model 3 (MOD3): endodontically treated maxillary central incisor restored with glass fiber post cemented with a high modulus dual-cure cement (Panavia). The same properties of the dual-cure cement were assigned to the core material. An alumina crown was considered Model 4 (MOD4): endodontically treated maxillary central incisor restored with a composite CAD-CAM machined endocrown. The canalar and core portions were considered as a one-component machined from a prepolymerised composite block (Paradigm MZ100, 3MESPE, Germany). A dual-cure cement layer (Panavia) was simulated as bonding system. A feldspathic crown was considered Model 5 (MOD5): endodontically treated maxillary central incisor restored with a composite CAD-CAM machined endocrown. The canalar and core portions were considered as a one-component machined from a prepolymerised composite block (Paradigm MZ100). A dual-cure cement layer (Panavia) was simulated as bonding system. A syntered-alumina crown was considered. 123

Model 6 (MOD6): endodontically treated maxillary central incisor restored with a feldspathic CAD-CAM machined endocrown. A dual-cure cement layer (Panavia) was simulated as bonding system. Model 7 (MOD7): endodontically treated maxillary central incisor restored with a syntered alumina CAD-CAM machined endocrown. A hybrid glass ionomer cement layer (RelyX Unicem, 3M ESPE, Germany) was simulated as bonding system. The mechaincal properties of all the simulated materials are reported in Table 1. Fig. 1 shows the FE models of the tested configurations. The glass fiber posts were considered made up of long fibers (glass fibers) embedded into a polymeric matrix. This composite material was considered orthotropic, so that it showed different mechanical properties along the fiber direction (x direction) and along the other two normal directions (y and z direction). The elastic properties of the isotropic materials are reported in Table 1. The glass fiber post mechanical characteristics are reported in Table 2. Ex, Ey, Ez represent the elastic moduli along the three dimensional directions while νxy, νxz, νyz and Gxy, Gxz, Gyz are the Poisson’s ratios and the shear moduli in the orthogonal planes (xy, xz and yz) respectively. Due to the comparative aim of the structural evaluations, the given arbitrary commercially available post geometry has been used: - 6% conicity; - tip diameter 1.0 mm; - 10 mm insertion depth (about 2/3 of the root length). All the nodes on the external surface of the root were constrained in all directions. The following assumptions have been made: - complete bonding between post and cement was considered; - dentine was assumed as an elastic isotropic material according to Darendelier [28] and Versluis 124

[29]; - rigid constrains have been considered at level of the root.

Results Figures 2 to 8 show the stress (A) and strain (B) distributions in the analyzed models. Model 1: (Figure 2A) Maximum equivalent stress occurs at level of the cervical region both on the facial and palatal aspect of the tooth. At level of the cemento-enamel junction (CEJ), more on the root surface than on the coronal structure. Stresses arising at level of the CEJ either on the palatal and facial aspect progressively decrease toward the inner part of the tooth. A low stress concentration region is evident at the incisal margin. The coronal enamel is subjected to a limited stress concentration, making the dentin below to be interested by a lower amount of stress. (Figure 2B) The root dentine deforms mainly in the areas closer to the CEJ and strain values decrease toward the inner part of the tooth. Lower strain values are recorded in both the incisal and middle third of the crown. Minimum strain values are evident at level of the apex. Model 2: (Figure 3A) Stress values comparable to those recorded in Model 1 are evident at level of the CEJ on both the facial and palatal aspect of the tooth. Higher stress concentration (9 MPa) is recorded on the facial aspect of the post-cement-dentine interface. Stress arising is evident in the whole glass-post (from 2 to 9 MPa). Stress in the cement core ranges from 1 to 2 MPa. (Figure 3B) The tooth deforms mainly in the cervical region; higher strain values are recorded in root dentin while lower values are evident at level of the cervical region of the crown. A strain value of 0.4e-003Îľ is recorded in root dentin below the CEJ. As to the glass post, higher deformations are evident on the facial aspect, where a strain value of

125

0.1e-003Îľ is recorded. In the coronal area, higher strain values are evident in the core material than in the glass post. Model 3: (Figure 4A) Stress occurs on the facial side of the post-cement-dentine interface (9 MPa), in the glass post structure stress arising is limited to the radicular portion of the post. High stress values are recorded in the crown-core-dentine interface (16 MPa - out of scale in the models) either on the facial and palatal side of the tooth. Stress in the cement core arise near the dentine-core interface (2 MPa), low stress values are recorded in the coronal portion of the glass post. (Figure 4B) The tooth deforms mainly in the cervical region of the root while no deformation are recorded in the crown. A strain value of 0.5e-003Îľ is recorded in the root dentin below the CEJ. Deformations occurring in the glass post are comparable to those described in Model 2. Model 4: (Figure 5A) Stress values along the radicular part of the endocrown-cementdentine interface range from 1 to 2 MPa, stress progressively increases form the longitudinal axis of the root toward the facial and palatal surface. Stress are homogeneous distributed in the composite core and ranges from 4 to 7 MPa in the crown-core-dentine interface. (Figure 5B) The tooth deforms mainly in the cervical region; higher strain values are recorded in root dentin while lower values are evident at level of the cervical region of the crown. A strain value of 0.5e-003Îľ is recorded in root dentin below the CEJ. In the root part of the composite endocrown, strain values are consistent with those obtained in root dentin. As to the coronal part of the core composite material, the strain progressively decrease from the cervical third to the incisal margin. The deformations recorded in the feldspathic ceramics are consistent with those obtained in the core material. Model 5: (Figure 6A) Stress values along the radicular part of the endocrown-cementdentine interface range from 1 to 2 MPa, stress profressively increases form the 126

longitudinal axis of the root toward the facial and palatal surface. On the countrary stress in the whole composite core structure reach a maximum value of 1 MPa, while stress in the crown-core-dentine interface ranges from 15 to 18 MPa (grey- out of scale in the model). (Figure 6B) The tooth deforms mainly in the cervical region of the root while no deformation are recorded in the crown. A strain value of 0.5e-003ε is recorded in the root dentin below the CEJ. In the root part of the composite endocrown, strain values are consistent with those obtained in root dentin. Model 6: (Figure 7A) Stress along the radicular part of the endocrown-cement-dentine interface range from 2 to 5 MPa. Stress in the CORONAL PART OF THE ENDOCROWNcement-dentine interface on the facial and palatal side range from 5 to 7 MPa while in the centaral areas of the CORONAL PART OF THE ENDOCROWN-cement-dentine interface range from 1 to 2 MPa. (Figure 7B) The tooth deforms mainly in the cervical region; higher strain values are recorded in root dentin while lower values are evident at level of the cervical region of the crown. A strain value of 0.444-003ε is recorded in root dentin below the CEJ. As to the coronal part of the endocrown, the maximum deformations are recorded at level of the CEJ. In the root part of the feldspathic endocrown, higher deformations are evident on the facial aspect, where a strain value of 0.111e-003ε is recorded. Model 7: (Figure 8A) Maximum stress values are recorded along the facial side of the RPE-cement-dentine interface, ranging from 6 to 18 MPa (grey- out of scale in the model). Stress in the coronal part of the endocrown-cement-dentine interface either on the facial and palatal side ranges from 6 to 10 MPa while in the central areas of the coronal part of the endocrown-cement-dentine interface range from 1 to 2 MPa.

127

(Figure 8B) The deformations are reorded mainly in the root dentin. In the cervical region, the dentin undergoes a value of 0.444e-003Îľ. No significant strain values are recorded in the overall coronal part of the alumina restoration.

Discussion Maxillary central incisors occlusaly protect posterior teeth through the protrusive movement, making posterior teeth disclosing during mouth opening. Moreover, they have the function to tear foods and stresses arising during tearing are of paramount importance for the long-term success of restorations. An endocrown is a full or compact crown that extends a post into the pulp chamber and/or pulp canals as one unit and not a separate post [30]. Stress concentrates where a not homogeneous material distribution is present just like the interface regions. The interfaces of materials with different moduli of elasticity represent the weak point of a restorative system, as the toughness/stiffness mismatch influences the stress distribution [31][32]. In many applications, a system is subjected to the repeated application of a stress below the yield strenght of the material. Even though the stress is below the yield strenght, the material may fail after a large number of applications of the stress. Failure occurs by a process involving nucleation of a crack, slow propagation of the crack, catastrophic failure of the system. Cracks nucleate at locations of highest stress and lowest local strenght. Systems investigated either in static and fatigue loading conditions show underposable highest stress concentration areas. In both cases systems showed similar failure aspects [33][34][35]. A static linear analysis can be succesfully applied to extrapolate reliable information about the relative susceptibility of systems to fatigue loading conditions. Fatigue loading condition, the most common in the oral environment, may emphasize 128

stress arising in critical areas leading the system to failure [13]. On this basis our assumption is that systems showing homogeneous stress disribution in a static analysis (better stress distributing capability) would shows a lower fatigue sensitivity in clinical applications. The present study was designed to compare the stress distribution of the different systems in order to identify areas of high stress concentration where possible fatigue failures are more expected to occur. As a consequnce, the attention was mainly focused on the overall stress distribution arising in a maxillary central incisor where the crown hard tissues were lost. Six different restorative configurations were tested in order to evaluate which one better mimrix the biomechanical behaviour of a sound tooth. As to the assumption of perfect bonding, usually interface imperfactions are randomly distributed and influence stress distribution in localized areas. As a consequence, imperfect bonding was not considered in order to compare the analyzed systems. The periodontal ligament has an high non-linear behavior [36]. In the present analysis it has been implicitly assumed that both displacements and strains developed in the structure are small. In practical terms this means that geometry of the elements remains basically unchanged during the loading process and that first-order, infinitesimall and linear strain approximation can be used. Accordingly the periodontal ligament was not simulated and rigid constrains was considered. In MOD1 (sound tooth) stresses were almost uniformely distributed in the tooth strucure. Nevertheless, the stress concentrated at the cervical region, where the cemento-enamel junction creates a physiological discontinuity of the tissues mechanical properties. In MOD2 and MOD3, the concentration of stress in the glass post-cement-dentin interfaces could be explained considering the higher elastic modulus (EM) of the glass post compared with the lower EM of dentin. Glass post increase the rigidity of the root and reduce its deformability (Fig.3b), the lower dentine deformation determines a decrease in physiological stress concentration described in MOD1 (Fig.2). On the other hand the 129

difference in resistance to deformation between glass post and root dentine produces a stress arising in the glass post-cement-dentin interfaces (Fig.3a). In MOD3 the high rigidity alumina crown determines a lack of deformation in the coronal part of the models and acts as strain shielder on the core material. The low deformability of this system determines a stress arising in the allumina crown and a lack of stress in the core material. The high rigidity of the coronal part increases strain and stress in the crown-core-dentin interface on the palatal and facial aspects of the restored tooth, while in MOD2 the higher flexibility of the pheldspar crown permits the coronal part to has higher deformation values and to dissipate stress in the core material. The higher deformability of the coronal part in MOD2 reduces stress arising in the crown-core-dentin interface on the palatal and facial aspects of the restored tooth. In MOD4 and MOD5 the low modulus (16GPa)

composite

endocrown permits to the canalar part to deform uniformelly with the root dentine (Fig.5b and 6b) thus not significant stress arising are recorded in the canalar portion-cementdentine interface. Strain values recorded in the root part of MOD4 and MOD5 are comparable with strain values recorded in MOD1 (Fig.2b) in the same area. MOD4 and MOD5 shows differents behavior in the coronal part; pheldspar crown in MOD4 undergoes an higher deformation than the allumina one in MOD5. In MOD4 the deformation is transferred unformelly from the crown to the core material, as happens in the sound tooth simulated in MOD1. In MOD5 the high modulus allumina crown has low deformation values on the other hand its low deformability increase stress values in its structure. The allumina crown produces a lack of strain and conseguentelly a lack of stress in the core material; as a conseguence the coronal part acts as one rigid component transferring a great amount of energy to the cervical region of the root where strain and stress values are increased. Stress values obtained in root cervical region in MOD5 are comparable with that obtained in MOD3, where the same allumina crown is simulated, while stress values in root cervical region in MOD4 are comparable with those obtained in MOD1 and MOD2. 130

The more flexible pheldspar crown deforms similary as the enamel structure, when the core structure is reinforced by the coronal estension of a glass-post, its overall rigidity is increased and deformations are recorded in the core areas interposed between the post and the crown MOD2 (Fig.3b). When a material with a resistance to deformation similar to that of dentine is used for the core structure, the core deforms under the crown MOD4 (Fig.5b). The simultaneous deformation of core and crown materials permits to contain strain and stress arising in the crown-core-dentin interface MOD2 (Fig.3a,b) MOD4 (Fig.5a,b). Either in MOD2 and MOD4 stress values at level of the crown-core-dentin interface are underposable to the stress values recorded in MOD1 at level of the dentinenamel junction. According to the results of the present study, the first null hypothesis was rejected since the crown material influenced the stress patterns of the restored tooth. In MOD6, the more flexible feldspathic endocrown deforms uniformelly either in the coronal and canalar portions, deformations occurring in this system improve the stress dissipating capability of the restoration; stress noticed in MOD6 are slightly higher than those recorded in the composite restored systems MOD4 (Fig.5a), MOD5 (Fig.6a). Deformations occurring in MOD7 (Fig.8b) are significantly lower than those recorded in the other investigated systems. Stress recorded in MOD7 (Fig.8a) in the canalar portioncement-dentin interface and in the core-crown-dentin interface are significantly higher than those noticed in the other investigated systems. In MOD7 the high EM alumina endocrown behaves as a rigid block producing high stress concentration either in the canalar portioncement-dentin interface and in the core-crown-dentin interface. Although using a low modulus glass-ionomer cement in order to improve its stress adsorbing function, it was not possible to reduce the stress arising in the system because of the high rigidity of the alumina endocrown. Some studies reported that post restored teeth with highly rigid materials showed the formation of gaps at level of both the core-tooth and post-tooth interfaces; such gaps were induced by the deformation of the dentine interface, where load 131

transfer from the restoration materials to the dentin occurs [16]. It could be hypothesized that an alumina endocrown may deforms significantly the dentin interface as reported in highly rigid post and core resorations. According to the results of the present study, the second null hypothesis was rejected, since the endocrown material influenced the biomechanical behaviour of the restored tooth. MOD 4 and MOD5 show the best biomechanical behaviour if compared to the stress and strain distribution in the sound tooth (MOD1). The use of a single block of a material with an EM similar to that of dentin to restore both the canalar and core portions of an endocrown enables the system to undergoes similar deformation of a sound tooth and to dissipate the stress along the overall structure of the restored tooth. Furthermore the endocrown configuration significantly reduces the number of interfaces of the system. According to the results of the present study, the third null hypothesis was rejected, since the biomechanical properties of the restorative system significantly influence the deformation and stress absorbing capability of the restored system.

Conclusions High stiffness materials just like alumina significantly withstand deformations, generating high stress concentrations in the interfaces. As a consequence, such materials negatively modify the biomechanical behaviour of the restorative system. On the contrary, low stiffness materials as composite resins accompany the natural flexural movements of the tooth, reducing stress arising in the interfaces. Within the limits of the present linear static analysis, it can be concluded that restorative materials with mechanical properties as much as possible similar to that of natural tooth hard tissues enable the restored system to mimic the mechanical behaviour of a natural tooth. From a clinical point of view, since they are producted from a single block, endocrown restorations present the advantage of reducing the interface of the restorative system. 132

Moreover, composite resins seem to be the most reliable materials to build-up endocrown restoration, as they generates low amounts of stress concentration.

133

References [1]

Yeh CJ. Fatigue root fracture: a spontaneous root fracture in non-endodontically

treated teeth. Br Dent J 1997; 182: 261-6. [2] Ferrari M, Vichi A, Mocci F, Mason PN. Retrospective study of the clinical performance of posts. Am J Dent 2000; 13: 9B-14B. [3] Llena-Puy MC, Forner-Navarro L, Barbero-Navarro I. Vertical root fracture in endodontically treated teeth: A review of 25 cases. Oral Surg, Oral Med, Oral Pathol, Oral RadiolEndod 2001; 92: 553-5. [4] Fennis WMM, Kuijs RH, Kreulen CM, Roeters FJ, Creugers NH, Burgersdijk RC. A survey of cusp fractures in a population of general dental practices. Int J Prosthodont 2002; 15: 559-63. [5] Salis SG, Hood JAA, Stokes ANS, Kirk EEJ. Patterns of indirect fracture in intact and restored human premolar teeth. Endod Dent Traumatol 1987; 3: 10-4. [6] Glazer B. Restoration of endodontically treated teeth with carbon fibre posts: a pospective study. J Can Dent Assoc 2001; 67: 70-1. [7] Akkayan B, Gulmez T. Resistance to fracture of endodontically treated teeth restored with different post systems. J Prosthet Dent 2002; 87: 431-7. [8] Guzy GE and Nicholls JI. In vitro comparison of intact endodontically treated teeth with and without endo-post reinforcement. J Prosthet Dent, 1979; 42: 39-44. [9] Heydecke G, Butz F, Strub JR. Fracture strength and survival rate of endodontically treated maxillary incisors with approximal cavities restoration with different post and core systems: an in vitro study. J Dent 2001; 29: 427-33. [10] Troedson M, Derand T. Effect of margin design, cement polymerization, and angle of loading on stress in porcelain veneers. J Prosthet Dent 1999; 82: 518-24. [11] Lanza A, Aversa R, Rengo S, Apicella D, Apicella A. 3D FEA of cemented steel, glass and carbon posts in a maxillary incisor. Dent Mater 2005; in press. 134

[12] Peters MCRB, Poort HW, Farah JW, Craig RG. Stress analysis of tooth restored with post and core. J Dent Res 1983; 62:760-3. [13] Schatz D, Alfter G, Goz G. Fracture resistance of human incisors and premolars: morphological and patho-anatomical factors. Dent Traumatol 2001; 17: 167-73. [14] Huysmans MC, van der Varst PG, Schafer R, Peters MC, Plasschaert AJ, Soltesz U. Fatigue behavior of direct post-and-core-restored premolars. J Dent Res, 1992; 71: 114550. [15] Zarone F, Apicella D, Sorrentino R, Apicella A. Influence of tooth preparation design on the stress distribution in maxillary central incisors restored by means of alumina porcelain veneers: a 3D Finite Elements Analysis. Dent Mater 2005; in press. [16] Castellani D, Bacetti T, Giovannoni A, Bernardini DU. Resistance to fracture of metalceramic and all-ceramic crowns. Int J Prosthodont 1994; 7: 149-54. [17] Chai J, Takahashi Y, Sulaiman F, Chong K and Lautenschlager EP. Probability of fracture of all-ceramic crowns. Int J Prosthodont, 2000; 13: 420-4. [18] Tinschert J, Zwez D, Marx R and Anusavice KJ. Structural reliability of alumina-, feldspar-, leucite-, mica- and zirconia-based ceramics. J Dent, 2000; 28: 529-35. [19] Magne P and Belser U. Esthetic improvements and in vitro testing of In-Ceram Alumina and Spinell ceramic. Int J Prosthodont, 1997; 10: 459-66. [20] Ortega VL, Pegoraro LF, Conti PCR, Do Valle AL, Bonfante G. Evaluation of fracture resistance of endodontically treated maxillary premolars, restored with ceromer or heatpressed ceramic inlays and fixed with dual-resin cements. Journal of Oral Rehabilitation 2004; 31: 393-97. [21] Peters MC, Poort HW. Biomechanical stress analysis of the amalgam-tooth interface. J Dent Res 1982; 62: 358-62. [22] Williams KR, Edmundsen JT. A finite element stress analysis of an endodontically restored tooth. Eng Med; 13(4):167-73. 135

[23] Pao YC, Reinhardt KA, Krejci RF. Root stress with tapered-end post design in periodontally compromised teeth. J Prosthet Dent 1987; 57:281-286. [24] Wheeler RH. Dental anatomy, physiology and occlusion. In Philadelphia WB Saunders USA 1974; 151. [25] De Santis R, Prisco D, Apicella A, Ambrosio L, Rengo S, Nicolais L. Carbon post adhesion to resin luting cement in the restoration of endodontically treated teeth. J Mater Sci Mater Med 2000; 11:201–6. [26] Ferrari M, Vichi A, Garcia-Godoy F. Clinical evaluation of reinforced epoxy resin posts and cast post and cores. Am J Dent 2000; 13:15B–118. [27] Asmussen E, Peutzfeldt A, Heitmann T. Stifness, elastic limit and strength of newer types of endodontic posts. J Dent 1999; 27:275–8. [28] Darendeliler SY, Alacam T, YamanY. Analysis of stress distribution in a maxillary central incisor subjected to various post and core applications. J Endodont 1998; 24:10711 [29] Versluis A, Douglas WH, Cross M, Sakaguchi RL. Does an incremental filling technique reduce polimerization shrinkage stresses? J Dent Res 1996; 3:871-8. [30] Otto T. Computer-aided direct all-ceramic crowns: preliminary 1-year results of a prospective clinical study. Int J Periodontics Restorative Dent 2004; 24(5): 446-55. [31] Ausiello P, Gee AJ, Rengo S, Davidson CL. Fracture resistance of endodonticallytreated premolars adhesively restored. Am J Dent 1997; 10: 237-41. [32] Assif D, Gorfil C. Biomechanical considerations in restoring endodontically treated teeth. J Prosthet Dent 1994; 71: 565-7. [33] Reis PNB, Ferreira JAM, Costa JDM, Richardson MOW. Fatigue damage in a glass fibre reinforced polypropylene composite. In: Found MS. Experimental techniques and design in composite materials 4. A.A. Balkema Publishers, Lisse, 2002.

136

[34] Found MS, Quaresimin M. Fatigue damage of carbon fibre reinforced laminates under two-stage loading. In: Found MS. Experimental techniques and design in composite materials 4. A.A. Balkema Publishers, Lisse, 2002. [35] De Iorio A, Ianniello D, Iannuzzi R, Penta F, Apicella A, Di Palma L. Strenght criteria for composite material structures. Experimental techniques and design in composite materials 4. A.A. Balkema Publishers, Lisse, 2002. [36] Cattaneo PM, Dalstra M, Melsen B. The finite element method: a tool to study orthodontic tooth movement. J Dent Res 2005; 84: 428-33.

137

Material/component

Elastic modulus (GPa)

Poissonâ&#x20AC;&#x2122;s ratio

Dentin

18.6

0.32

Enamel

84.1

0.33

Feldspathic ceramics

0.3

Syntered alumina

418

0.22

Composite (Paradigm MZ100)

0.3

High modulus cement (Panavia)

18.6

0.28

Low modulus cement (RelyX Unicem)

5.1

0.27

Table 1 â&#x20AC;&#x201C; Mechanical characteristics of the materials.

138

Property

Glass post

Ex, Gpa

Ey, Gpa

9.5

Ez, Gpa

9.5

νxy

0.27

νxz

0.34

νyz

0.27

Gxy

3.10

Gxz

3.50

Gyz

3.10

Table 2 – Mechanical properties of the glass fiber post.

139

Figure 1 â&#x20AC;&#x201C; 3D FE models and loading conditions.

Figure 2 â&#x20AC;&#x201C; Stress (A) and strain (B) distributions in Model 1.

140

Figure 3 â&#x20AC;&#x201C; Stress (A) and strain (B) distributions in Model 2.

Figure 4 â&#x20AC;&#x201C; Stress (A) and strain (B) distributions in Model 3.

141

Figure 5 â&#x20AC;&#x201C; Stress (A) and strain (B) distributions in Model 4.

Figure 6 â&#x20AC;&#x201C; Stress (A) and strain (B) distributions in Model 5.

142

Figure 7 â&#x20AC;&#x201C; Stress (A) and strain (B) distributions in Model 6.

Figure 8 â&#x20AC;&#x201C; Stress (A) and strain (B) distributions in Model 7.

143

Three Dimensional Finite Element Analysis of Strain and Stress Distributions in Endodontically-Treated Maxillary Central Incisors Restored with Different Post, Core and Crown Materials.

Sorrentino R., Apicella D., Ferro V., Zarone F., Ferrari M., Apicella A. â&#x20AC;&#x201C; Dental Materials, 2006; in press.

144

145

Introduction Restoration of endodontically treated teeth is a common problem in restorative dentistry, related to the fractures occurring in such teeth [1][2]. Posts are used to maintain the core material particulary in posterior teeth, where masticatory loads are essentially compressive [3]. However, when loaded transversely, as in the case of incisors, the flexural behaviour of posts should be carefully considered [4]. The magnitude and the angle of incisal load greatly influence the long-term success of restorative systems involving central incisors [5]. In the complex systems of post restorations, the stress distribution within the structure is multiaxial, not uniform and depending on the magnitude and direction of the applied external loads. Type of external loading, geometry of structure and residual stresses are some of the causes of multiaxiality stress distribution [6]. The placement of an endodontic post creates an unnatural restored structure since it fills the root canal space with a material that has a defined stiffness unlike the pulp. Hence it is not possible to recreate the original stress distribution of the tooth [7]. The use of glass fiber posts results in the best stress distribution, since the cement layer rigidity is less relevant compared to steel and carbon post configurations [8]. To summarize, two parameters strongly influences the mechanical behaviour of endodontically treated teeth restored with posts: the characteristics of the interfaces and the rigidity of the materials. Some of the fractures affecting post restorations could be related to concentration of forces [9]; fatigue loading must be considered an additional cause of root fracture. In order to assess all the eventual variables, the type of materials used for core build-up and crown restoration have to be carefully evaluated [10]. To date, there is still no agreement in the literature about which material or technique can optimally restore endodontically-treated teeth [11].

146

Proper understanding of the physical variables affecting the mechanical behaviour of biomedical devices during function is paramount to validate the clinical effectiveness and predict the long-term success of prostheses and restorative systems [12][10]. The simultaneous interaction of the many variables affecting a restorative system can be studied by means of a simulation in a computerized model. The Finite Elemet Analysis (FEA) consists in dividing a geometric model into a finite number of elements, each with specific physical properties. The variables of interest are approximated with some mathematical functions. Stress and strain distributions in response to different loading conditions can be simulated with the aid of computers provided with dedicated softwares. As a consequence, detailed evaluations of the mechanical behaviour of biologic and/or restorative systems are achievable, even in non-homogeneous bodies [10][13][14][15]. If correctly validated, the FEA could be useful in optimizing restorative design criteria and material choice and in predicting fracture potential under given circumstances [10][12]. Searching the literature, the majority of papers based on FE modelling of strain and stress distribution in post and crown restored teeth is based on two dimensional (2D) models [16][17][18] and only few on three dimensional (3D) ones [8][19][20]. Although reliable when considering axial symmetric systems, two dimensional meshing procedures do not allow to correctly assess the spatial distribution of stresses and strains affecting a restorative system. In order to overcome such a problem, three dimensional FEA was introduced to obtain more realistic models [8][10][12]. Nevertheless, this approach could present coarse meshes that would not allow a proper representation of both preparation details or materials used in thin layers, just like luting agents. The use of 3D-FEA with solid brick elements could be useful in achieving ordinate mapped meshes of complex biologic structures or restorative systems [10][12]. The present study compared the stress and strain distribution patterns of a sound maxillary central incisor with those of endodontically treated teeth restored with different 147

post, core and crown direct and indirect materials within the restorative materials and at the adhesive interfaces as well. The aim of the present comparative analysis was to estimate which combination of restorative materials resulted in the most homogeneous stress and strain distributions, so that to obtain the most similar configuration to that of a sound tooth. Two different null hypotheses were tested: - there is no association between the mechanical properties of core and crown materials on stress and strain concentration areas along the dentin/cement/post interfaces; - there is no association between the mechanical properties of core and crown materials on stress and strain levels along the dentin/cement/post interfaces. As to clinical significance, the results of the present study would allow clinicians to make an informed choice from among available materials to restore endodontically treated teeth.

Materials and methods The solid model was generated using literature data [24] for dentin and enamel internal volumes and morphologies, while the external shape of the maxillary central incisor was obtained by laser based 3D digitising (Cyberware) of a plaster cast (Thanaka manufacturer, Japan). The scanned profiles were assembled in a three dimensional wire frame structure using a 3D CAD (Autocad 12, Autodesk, Inc.) and exported into a 3D parametric solid modeller (Pro-Engineering 16.0 Parametric Technologies, USA). The tooth volumes were generated by fitting of the horizontal and vertical profiles. The geometries and volumes for the cement layer and post were also generated at this stage. The FEM model was obtained by importing the solid model into ANSYS rel. 9.0 FEM software (Ansys, Inc. Houston) using IGES format. The volumes were redefined in the new environment and meshed with 8 nodes brick with 3 degree of freedom per node, finally resulting in a 3D FE model with 13272 elements and 15152 nodes (fig.1). All the nodes on 148

the external surface of the root were constrained in all directions. Accuracy of the model was checked by convergence tests. Particular attention was devoted to the refinement of the mesh resulting from the convergence tests at the cement layer interfaces. Different material properties were coupled with the elements and geometries according to the volume material defined in figure 1 (enamel, dentin, restored crown, core, cement and post). Eighteen experimental models with different materials configuration were simulated as reported in Table 1. The mechaincal characteristics of all the simulated materials are reported in Table 2; the mechanical properties of dentine and enamel were assigned according to literature data [25][26][27]. The posts were considered made up of long glass fibers embedded into a polymeric matrix. This composite material was considered orthotropic, so that it showed different mechanical properties along the fiber direction (x direction) and along the other two normal directions (y and z direction). The elastic properties of the isotropic materials are reported in Table 1. The mechanical characteristics of the glass fiber post are reported in Table 2. Ex, Ey, Ez represent the elastic moduli along the three dimensional directions while νxy, νxz, νyz and Gxy, Gxz, Gyz are the Poisson’s ratios and the shear moduli in the orthogonal planes (xy, xz and yz) respectively. Due to the comparative aim of the structural evaluations, the following arbitrary commercially available post geometry was used: - 6% conicity; - 1.0 mm tip diameter; - 10 mm insertion depth (about 2/3 of the root length). A linear static structural analysis was performed to calculate the strain and stress distributions in different restoring configurations. An arbitrary load of 10 N was applied at 60° angle with tooth longitudinal axis on the palatal surface of the crown to simulate 149

tearing function (fig. 2). The static load of 10 N was defined to compare the systems in the field of low deformation where a linear static analysis can be performed with reliable results. Results of a static linear analysis are lineary proportional with the applied load In order to identify areas of strain and stress concentration where possible fatigue failures are more expected to occur, the choice of the pertinent stress representation criterion was based on the evaluation of failure predictive potential of the analysis performed. Von Mises (equivalent stresses) energetic criterion was then chosen. Under fatigue loading, in fact, the calculated stresses should relate to fracture probability and, therefore, to the assumption that different stress states having the same effect are equivalent when determining the system failure at critical stress values (failure criterion). In such cases, accurate predictions were observed [21][22][23]. The following assumptions were made: - dentin was assumed as an elastic and isotropic material [28] [29]; - complete bonding between crown and cement and between post and cement was considered; - rigid constraints were considered at level of the root. The assumption of perfect bonding resulted in homogeneous stress distributions reducing problems regarding possible stress concentrations caused by the presence of interface imperfections or debonded areas. The periodontal ligament was not simulated and rigid constraints were considered in all the experimental models: such an aspect can be avoided due to the comparative purpose of the present study. Nevertheless, the presence of a viscoelastic constraint would result in a different angle of application of the load and in a more significant effect on flexural stress build-up.

150

Results Tables 3 and 4 show the maximum values and concentration areas of strain and stress recorded in the present study respectively. Moreover, the peak values of the different anatomical and restorative components of each model were reported: crown, abutment, post/cement, root, combined crown interface (CI) and root interface (RI) were considered.

Strain analysis In all the models the values of strain recorded at the middle third of the buccal aspect of the root surface were out of scale. On the contrary, the minimum values were noticed at level of both the apical portion of the post and the root apex. In Mod.1 (control) the maximum deformations were evidenced at level of the cementoenamel junction (CEJ) on both the buccal and palatal aspects of root cement and dentin. Strains progressively decreased from the outer to the inner part of the root and from the CEJ towards the incisal margin of the crown as well (fig.3). In Mod.2 the maximum strains were evident at level of the root side of the CEJ next to the prosthetic margin and progressively decreased from the outer to the inner part of the root . The crown showed minimum strains at the cervical region, whereas the middle and incisal thirds were almost free of deformations (fig.3). In Mod.3, Mod.4, Mod.5 and Mod.6 strain distributions quite similar to those described for Mod.2 were evident but the prosthetic crown appeared even less defromed at the cervical third. The glass fiber post was interested by minimal strains on the buccal aspect (fig.3). In Mod.7 the maximum strains were evident at level of the root side of the CEJ next to the prosthetic margin and progressively decreased from the outer to the inner part of the root and from the cervical third towards the incisal margin of the crown as well (fig.4). In Mod.8, Mod.9, Mod.10 and Mod.11 the maximum strains were evident at level of the root side of the CEJ next to the prosthetic margin and progressively decreased from the 151

outer to the inner part of the root and from the cervical third towards the incisal margin of the crown as well. Deformations concentrated on the palatal aspect of either the prosthetic crown and the abutment. Moreover, strains were noticed in the cement-post interface at the CI and in the palatal side of the coronal part of the cement layer. The glass fiber post was interested by minimal strains on the buccal aspect (fig.4). In Mod.12 the maximum strains were evident on both the buccal and palatal aspects of the cervical margin and on the palatal concavity of the prosthetic crown. Deformations progressively decreased from the outer to the inner part of the root and from the cervical third towards the incisal margin of the crown as well. The glass fiber post was interested by limited strains particularly in its coronal part (fig.4). In Mod.13, Mod.14, Mod.15 and Mod.16 strain distributions quite similar to those described for Mod.11 were evident but higher deformations were noticed in both the prosthetic crown and the abutment, particularly at the palatal concavity; such a phenomenon was less evident in Mod.16. A strain peak interested the CI and the glass fiber post was interested by limited strains on the buccal aspect (fig.5). In Mod.17 strain distributions quite similar to those described for Mod.13-15 were evident but even higher deformations interested both the cervical third of the prosthetic crown and the abutment, particularly at the palatal concavity (fig.5). In Mod.18 deformations quite similar to those noticed in Mod.13-15 were recorded (fig.5).

progressively decreased from the outer to the inner part of the root and from the CEJ towards the incisal margin of the crown as well (fig.6). In Mod.2 stress distributions quite similar to those described for Mod.1 were noticed. Stress concentrated at level of the root side of the CEJ buccally whereas it mainly interested the prosthetic margin palatally. The coronal residual dentin appeared less stressed than the control natural tooth (fig.6). In Mod.3, Mod.4, Mod.5 and Mod.6 stress distributions quite similar to those described for Mod.2 were evident. The abutment was interested by minimal stress concentrations. Tha glass fiber post resulted particularly stressed at the buccal aspect of the RI (fig.6). In Mod.7 the maximum stress was evident at level of the root side of the CEJ next to the prosthetic margin and progressively decreased from the outer to the inner part of the root and from the cervical third towards the incisal margin of the crown as well (fig.7). In Mod.8, Mod.9, Mod.10 and Mod.11 lower stresses than in Mod. 3-6 were recorded in the cervical area of the restorative systems. On the contrary, higher stresses interested the buccal side of the glass fiber post particularly at the CI (fig.7). In Mod.12 stresses concentrated at the root side of the prosthetic margin particularly on the buccal aspect. Stresses progressively decreased from the cervical third to the incisal margin of the prosthetic crown. The abutment was mainly stressed in the cervical region. The glass fiber post showed a stress concentration at the CI (fig.7). In Mod.13, Mod.14, Mod.15 and Mod.16 stresses concentrated at the root side of the prosthetic margin particularly on the buccal aspect. Stresses progressively decreased from the cervical third to the incisal margin of the prosthetic crown. Stress concentrations were noticed along the cement-post interface particluarly at the CI either buccally and palatally. Stress on the palatal side of the Ci disappeared in Mod.16 where higher stress concentrations were evident in the cervical region of the abutment (fig.8).

153

In Mod.17 and Mod.18 stress distributions quite similar to those described for Mod.13-15 were recorded. Lower stresses were noticed in the cervical area of the abutment whereas slightly higher stresses concentrated at level of the root side of the prosthetic margin (fig.8).

Discussion According to the results of the present study, the mechanical properties of the crown and core material influenced both the position of concentration areas and the level of stress and strain along the dentin/cement/post interfaces; consequently, both the null hypotheses tested were rejecetd. Stress concentrates where not homogeneous material distributions are present just like interfaces. Interfaces of materials with different moduli of elasticity represent the weak link of restorative systems, as the toughness/stiffness mismatch influences the stress distribution [31][32]. In the oral enviroment, restorative systems are subjected to fatigue stress, namely the repeated application of lower loads than the yield strenght of the restorative materials. Nevertheless, such a cycling load application might produce microcracks causing the failure of the restorative system [13]. Such micro-cracks nucleate at locations of highest stress and lowest local strenght. Systems investigated either in static and fatigue loading conditions show underposable highest stress concentration areas and similar failure patterns [33][34][35]. Static linear analyses can be successfully applied to extrapolate reliable information about the relative susceptibility of systems to fatigue loading conditions. On this basis, the assumption that systems showing homogeneous stress disributions in a static analysis would show a lower fatigue sensitivity in clinical applications can be state.

154

The present study compared the stress distributions of different restorative systems to identify areas of high stress concentration, where eventual fatigue failures are more expected to occur. When the loading of a sound tooth (Mod.1 â&#x20AC;&#x201C; control) was simulated, the stress was almost uniformely distributed in tooth strucure. Nevertheless, the stress concentrated at the cervical region, where the cemento-enamel junction creates a physiological discontinuity of the tissue mechanical properties. Due to its increased rigidity (E=69 GPa), the feldspathic crown luted on a natural tooth abutment (Mod.2) transmitted lower deformations to coronal dentin concentrating the strain on root dentin next to the CEJ. Similarly, when the feldspathic crown was luted on composite cores (Mod.3, Mod.4, Mod.5 and Mod.6), the ceramic produced a stress and strain shielding effect on the core materials: as a consequence, deformations concentrated on root dentin next to the prosthetic crown interface. One one side, root dentin absorbed part of the strain but, on the other side, it conveyed limited deformations to the glass post. Stress arose at the RI due to the difference in mechanical resistance between the post and the dentin. Stress and strain values recorded in Mod.3, Mod.4, Mod.5 and Mod.6 were limited because of the good deformation absorbing capability of the dentin. Moreover, when a relatively rigid material just like feldspathic ceramic was used to build a prosthetic crown, the mechanical properties of the more elastic composite materials did not significantly influence the stress and strain concentrations within the core. On the contrary, when a prosthetic crown made up of a high modulus composite material (e.g. Gradia Forte, E=13.7 GPa) was simulated and luted on composite abutments (Mod.7, Mod.8, Mod.9, Mod.10 and Mod.11), the lower resistance to deformation of the composite crown produced higher deformation values in the core structure than those recorded simulating the presence of a feldspathic crown (Mod.3, Mod.4, Mod.5 and Mod.6). The deformations occurring in the core structure were transferred to the coronal part of the 155

glass post. As a consequence, the discontinuity in the mechanical properties of the core materials, the cement, the post and the dentin caused a stress concentration at the CI. Differently than the restorative systems provided with a feldspathic crown, in case of a high modulus composite crown the mechanical characteristics of the core materials significantly influenced the stress arising at the CI. Cores with lower elastic modulus transferred higher deformations to the glass post, increasing the material properties discontinuity and producing higher stress values at the CI. Conversely, higher elastic modulus cores reduced the strain of the glass post and decreased the mechanical properties mismatch, leading to lower stress arising at the CI. When a low modulus composite (e.g. Gradia, E=6.4 GPa) crown was simulated on a natural tooth abutment (Mod.12), the stress and strain distributions in tooth tissues remained basically unchanged. Due to the low resistance to deformation of the composite material, most strains arose at the cervical third of the crown. When a low modulus composite crown was luted on composites cores (Mod.13, Mod.14, Mod.15 and Mod.16), the deformation transferring phenomenon previously described for Mod.8, Mod.9, Mod.10 and Mod.11 was significantly emphasized. Perhaps, the values of stress concentrating at the CI were significantly higher than those recorded in both the feldspathic crown and high modulus composite crown systems. In Mod.16 the higher modulus of the core material significantly reduced the stress arising at the CI when compared to Mod.13, Mod.14 and Mod.15. Mod.17 and Mod.18 simulated direct restorations of glass post-reinforced teeth by means of a low modulus direct composite (e.g. Gradia Direct Anterior) and a high modulus direct composite (e.g. Gradia Direct Posterior) respectively. Due to the lower rigidity, Mod.17 configuration transferred higher deformations to the coronal part of the glass post and increased stress arising at the CI. On the contrary, the higher resistance to deformation of

156

the high modulus direct composite material reduced strains of the glass post, leading the restorative system to undergo lower stress and strain values at the CI.

Conclusions In FEA, assumptions related to material properties of simulated structures are not usually absolute representations of the structure. In reality the structures modeled are much more dynamic. Moreover, the physical characteristics of tissues vary from site to site and from individual to individual. The dimensions of natural restored teeth may deviate from those selected for the study and the mechanical constants involved may vary as well. In spite of these limitations, the present study may be considered to provide relative values that are meaningful representations of qualitative clinical trends. Within the limitations of the present theoretical study, the following conclusions can be drawn: - the mechanical properties of the crown and core material influenced the position of concentration areas of stress and strain along the dentin/cement/post interface; - the mechanical properties of the crown and core material influenced the level of stress and strain along the dentin/cement/post interface; - the cervical region of the restored tooth was subjected to the highest strain and stress concentrations; - the higher the rigidity of the crown and core materials the more apical the stress and strain concentrations along the adhesive interfaces. Further in vitro and in vivo analyses should be performed to confirm the clinical validity of the data provided by the 3D simulations of the present study.

157

References [1] Yeh CJ. Fatigue root fracture: a spontaneous root fracture in non-endodontically treated teeth. Br Dent J 1997; 182: 261-6. [2] Ferrari M, Vichi A, Mannocci F, Mason PN. Retrospective study of the clinical performance of posts. Am J Dent 2000; 13: 9B-14B. [3] Guzy GE and Nicholls JI. In vitro comparison of intact endodontically treated teeth with and without endo-post reinforcement. J Prosthet Dent, 1979; 42: 39-44. [4] Heydecke G, Butz F, Strub JR. Fracture strength and survival rate of endodontically treated maxillary incisors with approximal cavities restoration with different post and core systems: an in vitro study. J Dent 2001; 29: 427-33. [5] Troedson M, Derand T. Effect of margin design, cement polymerization, and angle of loading on stress in porcelain veneers. J Prosthet Dent 1999; 82: 518-24. [6] Huysmans MC, Van der Varst PG. Finite element analysis of quasistatic and fatigue of post and cores. J Dent 1993; 21: 57-64. [7] Akkayan B, Gulmez T. Resistance to fracture of endodontically treated teeth restored with different post systems. J Prosthet Dent 2002; 87: 431-7. [8] Lanza A, Aversa R, Rengo S, Apicella D, Apicella A. 3D FEA of cemented steel, glass and carbon posts in a maxillary incisor. Dent Mater 2005; 21: 709-715. [9] Peters MCRB, Poort HW, Farah JW, Craig RG. Stress analysis of tooth restored with post and core. J Dent Res 1983; 62:760-3. [10] Zarone F, Apicella D, Sorrentino R, Ferro V, Aversa R and Apicella A. Influence of tooth preparation design on the stress distribution in maxillary central incisors restored by means of alumina porcelain veneers: a 3D Finite Elements Analysis. Dent Mater 2005; 21: 1178-1188. [11] Ortega VL, Pegoraro LF, Conti PCR, Do Valle AL, Bonfante G. Evaluation of fracture resistance of endodontically treated maxillary premolars, restored with ceromer or heat158

pressed ceramic inlays and fixed with dual-resin cements. Journal of Oral Rehabilitation 2004; 31: 393-97. [12] Zarone F, Sorrentino R, Apicella D, Valentino B, Ferrari M and Apicella A. Evaluation of the biomechanical behaviour of maxillary central incisors restored by means of endocrowns compared to a natural tooth: a linear static 3D Finite Elements Analysis. Dental Materials, 2006; Jan 10 [Epub ahead of print]. [13] Ausiello P, Apicella A, Davidson CL and Rengo S. 3D-Finite Elements Analysis of cusp movements in a human upper premolar, restored with adesive resin-based composites. J Biomech 2001; 34: 1269-1277. [14] Dalstra M, Huiskes R and van Erning L. Development and validation of a threedimensional finite element model of the pelvic bone. J Biomech Eng 1995; 117: 272-278. [15] Zarone F, Apicella A, Nicolais L, Aversa R and Sorrentino R. Mandibular flexure and stress build-up in mandibular full-arch fixed prostheses supported by osseointegrated implants. Clin Oral Impl Res 2003; 1: 319-328. [16] Yang HS, Lang LA, Molina A and Felton DA. The effects of dowel design and load direction on dowel-and-core restorations. J Prosthet Dent 2001; 85: 558-567. [17] Toparli M. Stress analysis in a post-restored tooth utilizing the finite element method. J Oral Rehabil 2003; 30: 470-476. [18] De Castro Albuquerque R, De Abreu Polleto LT, Fontana RHBTS and Cimini Jr CA. Stress analysis of an upper central incisor restored with different posts. J Oral Rehabil 2003; 30: 936-943. [19] Pierrisnard L, Bohin F, Renault P and Barquins M. Corono-radicular reconstruction of pulpless teeth: a mechanical study using finite element analysis. J Prosthet Dent 2002; 88: 442-448. [20] Asmussen E, Peutzfeldt A and Sahafi A. Finite element analysis of stresses in endodontically treated, dowel-restored teeth. J Prosthet Dent 2005; 94: 321-329. 159

[21] Schatz D, Alfter G and Goz G. Fracture resistance of human incisors and premolars: morphological and patho-anatomical factors. Dent Traumatol 2001; 17: 167-173. [22] Williams KR and Edmundsen JT. A finite element stress analysis of an endodontically restored tooth. Eng Med 13:167-173. [23] Pao YC, Reinhardt KA and Krejci RF. Root stress with tapered-end post design in periodontally compromised teeth. J Prosthet Dent 1987; 57: 281-286. [24] Wheeler RH. Dental anatomy, physiology and occlusion. In Philadelphia WB Saunders USA 1974; 151. [25] De Santis R, Prisco D, Apicella A, Ambrosio L, Rengo S and Nicolais L. Carbon post adhesion to resin luting cement in the restoration of endodontically treated teeth. J Mater Sci Mater Med 2000; 11: 201-206. [26] Ferrari M, Vichi A and Garcia-Godoy F. Clinical evaluation of reinforced epoxy resin posts and cast post and cores. Am J Dent 2000; 13: 15B-18B. [27] Asmussen E, Peutzfeldt A and Heitmann T. Stifness, elastic limit and strength of newer types of endodontic posts. J Dent 1999; 27: 275-278. [28] Darendeliler SY, Alacam T and Yaman Y. Analysis of stress distribution in a maxillary central incisor subjected to various post and core applications. J Endodont 1998; 24: 107111. [29] Versluis A, Douglas WH, Cross M and Sakaguchi RL. Does an incremental filling technique reduce polimerization shrinkage stresses? J Dent Res 1996; 3: 871-878. [30] Otto T. Computer-aided direct all-ceramic crowns: preliminary 1-year results of a prospective clinical study. Int J Periodontics Restorative Dent 2004; 24: 446-455. [31] Ausiello P, Gee AJ, Rengo S and Davidson CL. Fracture resistance of endodonticallytreated premolars adhesively restored. Am J Dent 1997; 10: 237-241. [32] Assif D and Gorfil C. Biomechanical considerations in restoring endodontically treated teeth. J Prosthet Dent 1994; 71: 565-567. 160

[33] Reis PNB, Ferreira JAM, Costa JDM and Richardson MOW. Fatigue damage in a glass fibre reinforced polypropylene composite. In: Found MS. Experimental techniques and design in composite materials 4. A.A. Balkema Publishers, Lisse, 2002. [34] Found MS and Quaresimin M. Fatigue damage of carbon fibre reinforced laminates under two-stage loading. In: Found MS. Experimental techniques and design in composite materials 4. A.A. Balkema Publishers, Lisse, 2002. [35] De Iorio A, Ianniello D, Iannuzzi R, Penta F, Apicella A and Di Palma L. Strenght criteria for composite material structures. Experimental techniques and design in composite materials 4. A.A. Balkema Publishers, Lisse, 2002.

161

MODEL

CROWN

ABUTMENT

1 (control)

POST/CEMENT

Natural tooth

Feldsp.

Natural tooth

Feldsp.

GDA

Glass fibers + Unifil Core

Feldsp.

GDA + GFL

Glass fibers + Unifil Core

Feldsp.

GFL

Glass fibers + Unifil Core

Feldsp.

Unifil Core

Glass fibers + Unifil Core

GFR

GDA

Glass fibers + Unifil Core

GFR

GDA + GFL

Glass fibers + Unifil Core

GFR

GFL

Glass fibers + Unifil Core

GFR

Unifil Core

Glass fibers + Unifil Core

Gradia

GDA

Glass fibers + Unifil Core

Gradia

GDA + GFL

Glass fibers + Unifil Core

Gradia

GFL

Glass fibers + Unifil Core

Gradia

Unifil Core

Glass fibers + Unifil Core

Natural tooth

GDA

Glass fibers + Unifil Core

GDP

Glass fibers + Unifil Core

Table 1 â&#x20AC;&#x201C; Experimental models (Feldsp.: feldspathic ceramic; GDA: Gradia Direct Anterior; GFL: Gradia Flow; GFR: Gradia Forte; GDP: Gradia Direct Posterior).

162

Material/component

Elastic modulus (GPa)

Poissonâ&#x20AC;&#x2122;s ratio

Dentin

18.6

0.32

Enamel

84.1

0.33

Feldspathic ceramics

0.3

Syntered alumina

418

0.22

Composite (Paradigm MZ100)

0.3

High modulus cement (Panavia)

18.6

0.28

Low modulus cement (RelyX Unicem)

5.1

0.27

Table 2 â&#x20AC;&#x201C; Mechanical characteristics of the materials.

163

MOD.

MAXIMUM

CROWN

ABUTMENT

POST/CEMENT

440 (root cervical area)

55-440

-----

500 (root cervical area)

55-167

55-110

55-500

-----

500 (root cervical area)

55-167

110-330

55-110

55-500

55-330

55-167

500 (root cervical area)

55-167

110-330

55-167

55-500

55-330

55-167

500 (root cervical area)

55-167

55-500

55-167

500 (root cervical area)

55-167

55-500

55-167

440 (root cervical area)

55-330

55-220

-----

500 (abutment/post interface)

55-440

167-500

55-167

55-440

110-389

55-167

500 (abutment/post interface)

55-330

167-500

55-167

55-440

110-440

55-167

440 (root cervical area)

55-440

167-389

55-167

55-440

110-330

55-167

440 (root cervical area)

55-440

110-278

55-167

55-440

110-222

55-167

500 (crown cervical area)

55-500

110-440

-----

500 (crown cervical area)

55-500

167-500

55-167

55-440

167-500

55-330

500 (crown cervical area)

55-500

167-500

55-167

55-440

167-500

55-330

500 (crown cervical area)

55-500

167-440

55-167

55-440

110-389

55-167

500 (crown cervical area)

55-500

110-330

55-167

55-440

110-167

55-167

500 (crown cervical area)

55-500

167-500

55-220

55-440

167-500

55-220

500 (crown cervical area)

55-500

167-500

55-167

55-440

110-440

55-220

55-330

ROOT

55-440

Table 3 – Strain values expressed in με (CI: combined crown interface; RI: root interface).

164

MOD.

MAXIMUM

CROWN

ABUTMENT

7.8 (root cervical area)

1.1-7.8

-----

8.9 (root cervical area)

1.1-8.9

1.1-2.2

1.1-8.9

-----

8.9 (root cervical area+RI)

1.1-8.9

1.1-2.2

1.1-7.8

1.1-8.9

1.1-5.5

1.1-7.8

10 (root cervical area+RI)

1.1-10

1.1-2.2

1.1-7.8

1.1-8.9

1.1-5.5

1.1-7.8

10 (root cervical area+RI)

1.1-10

1.1-2.2

1.1-7.8

1.1-8.9

1.1-5.5

1.1-7.8

10 (root cervical area+RI)

1.1-10

1.1-2.2

1.1-7.8

1.1-8.9

2.2-5.5

1.1-7.8

7.8 (root cervical area)

1.1-5.5

1.1-4.4

-----

10 (CI)

1.1-6.7

1.1-2.2

1.1-10

1.1-7.8

2.2-10

1.1-7.8

10 (CI)

1.1-6.7

1.1-2.2

1.1-10

1.1-7.8

2.2-10

1.1-7.8

8.9 (CI)

1.1-6.7

1.1-2.2

1.1-8.9

1.1-7.8

2.2-8.9

1.1-7.8

8.9 (root cervical area+CI+RI)

1.1-6.7

1.1-3.3

1.1-7.8

2.2-7.8

1.1-7.8

7.8 (root cervical area)

1.1-3.3

1.1-6.7

-----

10 (CI)

1.1-3.3

1.1-10

1.1-7.8

2.2-10

1.1-10

10 (CI)

1.1-3.3

1.1-10

1.1-7.8

2.2-10

1.1-10

10 (CI)

1.1-3.3

1.1-10

1.1-7.8

2.2-10

1.1-10

10 (CI)

1.1-3.3

1.1-6.7

1.1-10

1.1-7.8

3.3-10

1.1-10

10 (CI)

1.1-3.3

1.1-10

1.1-7.8

2.2-10

1.1-10

10 (CI)

1.1-3.3

1.1-10

1.1-7.8

2.2-10

1.1-10

1.1-6.7

POST/CEMENT

ROOT

1.1-7.8

Table 4 â&#x20AC;&#x201C; Stress values expressed in MPa (CI: combined crown interface; RI: root interface).

165

Fig. 1 â&#x20AC;&#x201C; Three dimensional finite element model of the restorative systems.

Fig. 2 â&#x20AC;&#x201C; Arbitrary load application.

166

Fig. 3 – Strain distributions in models 1 to 6 (strains expressed in με).

Fig. 4 – Strain distributions in models 7 to 12 (strains expressed in με). 167

Fig. 5 – Strain distributions in models 13 to 18 (strains expressed in με).

Fig. 6 – Stress distributions in models 1 to 6 (stresses expressed in MPa).

168

Fig. 7 â&#x20AC;&#x201C; Stress distributions in models 7 to 12 (stresses expressed in MPa).

Fig. 8 â&#x20AC;&#x201C; Stress distributions in models 13 to 18 (stresses expressed in MPa).

169

Three dimensional finite element analysis of stress and strain distributions in endodontically treated maxillary central incisors restored with post and core direct and indirect restorations.

Sorrentino R., Apicella D., Ferro V., Zarone F., Ferrari M., Apicella A. â&#x20AC;&#x201C; Journal of Adhesive Dentistry, 2006; in press.

170

ABSTRACT Purpose: The present study aimed at estimating which combination of restorative materials resulted in the most homogeneous stress and strain distributions. Materials and Methods: Eight experimental finite element models with different materials configuration were simulated; both indirect and direct restorations were considered. An arbitrary load of 10 N was applied at 60Â° angle with tooth longitudinal axis on the palatal surface of the crown to simulate tearing function. Results: In all the models the values of both strain and stress recorded at the middle third of the buccal aspect of the root surface were out of scale. On the contrary, the minimum values were noticed at level of both the apical portion of the post and the root apex. Conclusion: The mechanical properties of the crown and core material influenced both the position of of concentration areas and the level of stress and strain along the dentin/cement/post interface.

171

Introduction Fractures frequently affect endodontically treated teeth [1][2]. The materials used to build up the crown is often retained by means of posts: such devices are effective in withstanding compressive loads, just like in case of posterior teeth restorations [3], whereas their flexural behaviour should be carefully considered when posts are loaded transversely, as in case of incisors [4]. In post and core restorations, the stress distribution within the structure is multiaxial, not uniform and depending on the magnitude and direction of the applied external loads [5]. Type of external loading, geometry of structure and residual stresses are some of the causes of multiaxial stress distributions [6]. The presence of a post characterized by a stiffness different from that of the pulp originates an unnatural restored structure: consequently, it is not possible to recreate the stress distribution of natural teeth [7]. Glass fiber posts offer the best stress distribution, since the rigidity of the cement layer is less relevant compared to steel and carbon post systems [8]. Briefly, two parameters strongly influence the mechanical behaviour of endodontically treated teeth restored with posts: the characteristics of the interfaces and the rigidity of the materials. Some of the fractures affecting post restorations could be related to concentration of forces [9]; fatigue loading must be considered an additional cause of root fracture. In order to assess all the eventual variables, the type of materials used for core build-up and crown restoration have to be carefully evaluated [10]. To date, there is still no agreement in the literature about which material or technique can optimally restore endodontically-treated teeth [11]. Proper understanding of the physical variables affecting the mechanical behaviour of biomaterials is paramount to validate the clinical effectiveness and predict the long-term success of restorative systems [12][10].

172

Computerized models could be useful in simulating the simultaneous interaction of the many variables affecting a restorative system. The Finite Elemet Analysis (FEA) consists in dividing a geometric model into a finite number of elements, each with specific physical properties. The variables of interest are approximated with some mathematical functions. Stress and strain distributions in response to different loading conditions can be simulated with the aid of computers provided with dedicated softwares. As a consequence, detailed evaluations of the mechanical behaviour of biologic and/or restorative systems are achievable, even in non-homogeneous bodies [10][13][14][15]. If correctly validated, the FEA could be useful in optimizing restorative design criteria and material choice and in predicting fracture potential under given circumstances [10][12]. Most studies based on FEA of strain and stress distribution in post and crown restored teeth is based on two dimensional (2D) models [16][17][18] and only few on three dimensional (3D) ones [8][19][20]. Although reliable when considering axial symmetric systems, two dimensional meshing procedures do not allow to correctly assess the spatial distribution of stresses and strains affecting a restorative system. In order to overcome such a problem, three dimensional FEA was introduced to obtain more realistic models [8][10][12]. Nevertheless, this approach could present coarse meshes that would not allow a proper representation of both preparation details or materials used in thin layers, just like luting agents. The use of 3D-FEA with solid brick elements could be useful in achieving ordinate mapped meshes of complex biologic structures or restorative systems [10][12]. The present study compared the stress and strain distribution patterns of a sound maxillary central incisor with those of endodontically treated teeth restored with different post and core direct and indirect materials within the restorative materials and at the adhesive interfaces as well.

173

The aim of the present comparative analysis was to estimate which combination of restorative materials resulted in the most homogeneous stress and strain distributions, so that to obtain the most similar configuration to that of a sound tooth. Two different null hypotheses were tested: - there is no association between the mechanical properties of core and crown materials on stress and strain concentration areas along the dentin/cement/post interfaces; - there is no association between the mechanical properties of core and crown materials on stress and strain levels along the dentin/cement/post interfaces. As to clinical significance, the results of the present study would allow clinicians to make an informed choice from among available materials to restore endodontically treated teeth.

the mesh resulting from the convergence tests at the cement layer interfaces. Different material properties were coupled with the elements and geometries according to the volume material defined in figure 1 (enamel, dentin, restored crown, core, cement and post). Eight experimental models with different materials configuration were simulated as reported in Table 1; both indirect (Mod.2 - Mod.4) and direct restorations (Mod.5 - Mod.8) were considered. The mechanical properties of dentine and enamel were assigned according to literature data [25][26][27]. The elastic properties of the isotropic materials are reported in Table 2. Two different kinds of posts were considered: steel and glass fiber posts. The latter were considered made up of long glass fibers embedded into a polymeric matrix. Such a composite material was considered orthotropic, showing different mechanical properties along the fiber direction (x direction) and along the other two normal directions (y and z direction). Due to the comparative aim of the structural evaluations, the following arbitrary commercially available post geometry was used: - 6% conicity; - 1.0 mm tip diameter; - 10 mm insertion depth (about 2/3 of the root length). A linear static structural analysis was performed to calculate the strain and stress distributions in different restoring configurations. An arbitrary load of 10 N was applied at 60Â° angle with tooth longitudinal axis on the palatal surface of the crown to simulate tearing function (fig. 2). The static load of 10 N was defined to compare the systems in the field of low deformation where a linear static analysis can be performed with reliable results. Results of a static linear analysis are linearly proportional with the applied load The choice of the pertinent stress representation criterion was based on the evaluation of failure predictive potential of the analysis performed, in order to identify areas of strain and 175

stress concentration where possible fatigue failures are more expected to occur. Von Mises (equivalent stresses) energetic criterion was then chosen. Under fatigue loading, in fact, the calculated stresses should relate to fracture probability and, therefore, to the assumption that different stress states having the same effect are equivalent when determining the system failure at critical stress values (failure criterion). In such cases, accurate predictions were observed [21][22][23]. The following assumptions were made: - dentin was assumed as an elastic and isotropic material [28] [29]; - complete bonding between crown and cement and between post and cement was considered; - rigid constraints were considered at level of the root. The assumption of perfect bonding resulted in homogeneous stress distributions reducing problems regarding possible stress concentrations caused by the presence of interface imperfections or debonded areas. The periodontal ligament was not simulated and rigid constraints were considered in all the experimental models: such an aspect can be avoided due to the comparative purpose of the present study. Nevertheless, the presence of a viscoelastic constraint would result in a different angle of application of the load and in a more significant effect on flexural stress build-up.

Results Table 3 and table 4 show the maximum values and concentration areas of strain and stress recorded in the present study respectively. Furthermore, the peak values of the different anatomical and restorative components of each model were reported: crown, abutment, post/cement, root, combined crown interface (CI) and root interface (RI) were considered. 176

Strain analysis In all the models the values of strain recorded at the middle third of the buccal aspect of the root surface were out of scale. On the contrary, the minimum values were noticed at level of the apical portion of the root. In Mod.1 (control) (fig. 2) the maximum deformations were evidenced at level of the cemento-enamel junction (CEJ) on both the buccal and palatal aspects of root cement and dentin. Strains progressively decreased from the outer to the inner part of the root and from the CEJ towards the incisal margin of the crown as well. In Mod.2, Mod.3 and Mod.4 (figg. 3-5) the maximum strains were evident at level of the root side of the CEJ next to the prosthetic margin and progressively decreased from the outer to the inner part of the root. The crown was almost free from deformations. Very low strain values were noticed in both the abutment and the post. As to the latter, deformations were more evident on the vestibular side. Relatively high deformations occurred next to the apical portion of the post. As to the cement layer, the maximum strain values were recorded at the CI on the palatal side of the adhesive interface, particularly at level of the coronal third of the root; such a phenomenon was more evident in Mod.3. In Mod.5 and Mod.6 (fig. 6 and 7) the maximum strains were evident at level of the root side of the CEJ next to the prosthetic margin and progressively decreased from the outer to the inner part of the root and from the cervical third to the incisal margin of the crown. A strain pattern comparable to that of the crown was noticed in the abutment as well. Very low strain values were noticed in the post, even though relatively high deformations occurred next to its apical portion. Furthermore, deformations were more evident on the vestibular side of the post. As to the cement layer, the maximum strain values were recorded at the CI; such a phenomenon was more evident in Mod.6. In Mod.7 and Mod.8 (figg. 8 and 9) strain distributions quite similar to those described for Mod.5 and Mod.6 were evident but the deformations recorded were higher. Very high 177

strain values were noticed in the cervical area of the crown and at the palatal concavity. A peak of strain was evident in the palatal region next to the most coronal portion of the post. As to the cement layer, higher deformations occurred in Mod.8.

Stress analysis In all the models the values of stress recorded at the middle third of the buccal aspect of the root surface were out of scale. On the contrary, the minimum values were noticed at level of both the apical portion of the post and the root apex. In Mod.1 (control) (fig. 2) the maximum stresses were evidenced at level of the cementoenamel junction (CEJ) on both the buccal and palatal aspects of root cement and dentin. Stress progressively decreased from the outer to the inner part of the root and from the CEJ towards the incisal margin of the crown as well. In Mod.2, Mod.3 and Mod.4 (figg. 3-5) stress concentrated at level of both the prosthetic margin and the root side of the CEJ. Stress progressively decreased from the outer to the inner portion of the root as well as from the cervical third to the incisal margin of the crown. The abutment was interested by very low stress arising. High stress values were noticed at the RI on the buccal side of the post particularly in Mod.3 and Mod.4; furthermore, in Mod.3 and Mod.4, such a stress concentration extended towards the palatal side of the post as well at the CI. In Mod.5 and Mod.6 (figg. 6 and 7) stress concentrated at level of the root side of the CEJ next to the prosthetic margin and progressively decreased from the outer to the inner portion of the root. Similarly, both the abutment and the crown were inrested by decreasing stress values from the cervical third to the incisal margin. High stress values were noticed in the buccal portion of the post in Mod.5, particularly at level of the CI. Even higher values were recorded in Mod.6, where stress also extended towards the palatal side of the post. Moderate stresses were evident at the apical tip of the post. 178

In Mod.7 and Mod.8 (figg. 8 and 9) stress distributions quite similar to those described for Mod.5 and Mod.6 were evident. Furthermore, stress values recorded in Mod.7 and Mod.8 were higher and located more coronally, particularly at level of the CI. Relatively high stress concentrations were also noticed at level of the coronal tip of the post.

Discussion According to the results of the present study, the mechanical properties of the crown and core material influenced both the position of concentration areas and the level of stress and strain along the dentin/cement/post interfaces; consequently, both the null hypotheses tested were rejecetd. Interfaces showed stress concentrations due to the different moduli of elasticity and the toughness/stiffness mismatch of not homogeneous materials [31][32]. In the oral enviroment, the mechanical performances of restorative systems may be affected by the cycling application of lower loads than the yield strenght of the restorative materials. Fatigue stress can result in microcracks that nucleate at locations of highest stress and lowest local strenght causing the failure of the restorative system [13]. Systems investigated either in static and fatigue loading conditions show underposable highest stress concentration areas and similar failure patterns [33][34][35]. Static linear analyses can be successfully applied to extrapolate reliable information about the relative susceptibility of systems to fatigue loading conditions. On this basis, the assumption that systems showing homogeneous stress disributions in a static analysis would show a lower fatigue sensitivity in clinical applications can be state. The present study compared the stress distributions of different restorative systems to identify areas of high stress concentration, where eventual fatigue failures are more expected to occur.

179

In the natural tooth (Mod.1 - control) the stress was almost uniformely distributed, even though the stress concentrated at the cervical region, where the CEJ creates a physiological discontinuity of the tissue mechanical properties. In indirect restorations (Mod.2 - Mod.4), the zirconia crown transmitted lower deformations to coronal dentin due to its increased rigidity and concentrated the strain on root dentin next to the CEJ. Despite the good deformation adsorbing capability of the dentin, limited deformations were conveyed to the glass post. Stress arose at the RI due to the difference in mechanical resistance between the post and the dentin. The mechanical properties of the more elastic composite materials did not significantly influence the stress and strain concentrations within the core. On the contrary, when a direct prosthetic crown made up of a high modulus composite material was simulated (Mod.5 and Mod.6) , the lower resistance to deformation of the composite resulted in higher deformation values within the core. The deformations occurring in the core structure were transferred to the coronal part of the post causing a stress concentration at the CI due to the discontinuity in the mechanical properties of the restorative materials. When a low modulus composite crown was simulated (Mod.7 and Mod.8) , the stress and strain distributions in tooth tissues remained basically unchanged. Due to the low resistance to deformation of the composite material, most strains arose at the cervical third of the crown. The deformation transferring phenomenon previously described for Mod.5 and Mod.6 was significantly emphasized. In all the tested material configurations, the use of a steel post significantly increased the stress concentration at the RI and the tensional stress of the interface reached the CI as well (Mod.3, Mod.4, Mod.6 and Mod.8). Such a phenomenon was due to both the decreased deformability of the root and the discontinuity in the mechanical properties of the restorative materials at the interfaces. The stiffness of the steel post limited the 180

deformations but caused evident stress oncentrations within the material. The use of a low modulus cement with a lower resistance to deformation did not affect the stress concentrations at the RI and CI. Conversely, the use of a glass-fiber post resulted in more homogeneous stress distributions due to the increased deformability of the root (Mod.2, Mod.5 and Mod.7). In the presence of a steel post, the lower the modulus of elasticity of the restorative material the higher the stress concentrations.

Conclusions In FEA, assumptions related to material properties of simulated structures are not usually absolute representations of the structure, since in reality they are much more dynamic. Furthermore, the physical characteristics of tissues vary inter- and intraindividually. The dimensions of natural teeth may deviate from those selected for this study and the mechanical constants involved may vary as well. In spite of these limitations, the present study may be considered to provide relative values that are meaningful representations of qualitative clinical trends. Within the limitations of the present theoretical study, the following conclusions can be drawn: - the mechanical properties of the crown and core material influenced both the position of of concentration areas and the level of stress and strain along the dentin/cement/post interface; - the cervical region of the restored tooth was subjected to the highest strain and stress concentrations; - the higher the rigidity of the crown the higher the stress in the cervical area; - the lower the rigidity of the crown the higher the stress transmitted to the post; - steel posts caused very high stress concentrations within the material.

181

Further in vitro and in vivo analyses should be performed to confirm the clinical validity of the data provided by the 3D simulations of the present study.

182

186

MODEL

CROWN

ABUTMENT

1 (control)

POST

CEMENT

Natural tooth

LAVA zirconia

Unicem

Glass-fibers

Unicem

LAVA zirconia

Rely-X ARC

Steel

Rely-X ARC

LAVA zirconia

Unicem

Steel

Unicem

Z 250

Glass-fibers

Unicem

Z 250

Steel

Rely-X ARC

Filtek flow

Glass-fibers

Unicem

Filtek flow

Steel

Rely-X ARC

Table I â&#x20AC;&#x201C; Experimental models.

187

Material/component

Elastic modulus (GPa)

Dentin

18.6

Enamel

84.1

Zirconia (Lava)

210

Microhybrid composite (Z-250)

Flowable composite (Filtek-flow)

High modulus cement (Unicem)

9.5

Low modulus cement (Rely-X ARC)

5.7

Table 2 â&#x20AC;&#x201C; Mechanical characteristics of the materials.

188

MOD.

MAXIMUM

CROWN

ABUTMENT

POST

7.8 (root cervical area)

-----

10 (prosthetic margin)

0 - 10

0 - 1.1

0 - 7.8

0 - 3.3

0 - 10

0 - 4.4

3.3 - 7.8

10 (prosthetic margin, post)

0 - 10

0 - 1.1

0 - 10

0 - 3.3

0 - 10

0 - 8.9

2.2 - 10

10 (prosthetic margin, post)

0 - 10

0 - 1.1

0 - 10

0 - 3.3

0 - 10

0 - 8.9

2.2 - 10

8.9 (post)

0 - 5.5

0 - 3.3

0 - 8.9

0 - 3.3

0 - 10

2.2 - 7.8

3.3 - 7.8

10 (post)

0 - 5.5

0 - 3.3

2.2 - 10

0 - 3.3

0 - 10

2.2 - 10

3.3 - 10

10 (post)

0 - 3.3

0 - 10

0 - 3.3

0 - 10

2.2 - 10

3.3 - 7.8

10 (post)

0 - 3.3

2.2 - 10

0 - 3.3

0 - 10

2.2 - 10

3.3 - 10

0 - 3.3

CEMENT

ROOT

0 - 10

Table 3 â&#x20AC;&#x201C; Stress values.

189

MOD.

MAXIMUM

CROWN

ABUTMENT

POST

440 (root cervical area)

-----

500 (root cervical area)

0 - 111

111 - 333

0 - 500

0 - 167

111 - 167

500 (root cervical area)

0 - 111

111 - 333

0 - 500

0 - 167

111 - 167

500 (root cervical area)

0 - 111

111 - 333

0 - 500

0 - 167

111 - 167

440 (root cervical area)

0 - 389

0 - 333

0 - 167

111 - 333

0 - 440

0 - 278

111 - 167

440 (root cervical area)

0 - 389

0 - 333

0 - 167

111 - 389

0 - 440

0 - 278

111 - 167

500 (prosthetic margin, palatal concavity)

0 - 500

167 - 500

0 - 167

111 - 389

0 - 500

167 - 389

111 - 167

500 (prosthetic margin, palatal concavity)

0 – 500

167 - 500

0 - 167

111 - 500

0 - 500

167 - 333

111 - 167

0 - 389

CEMENT

ROOT

0 - 440

Table 4 – Strain values.

190

Fig. 1 â&#x20AC;&#x201C; Arbitrary load application.

Fig. 2 â&#x20AC;&#x201C; Strain and stress distributions in model 1.

191

Fig. 3 â&#x20AC;&#x201C; Strain and stress distributions in model 2.

Fig. 4 â&#x20AC;&#x201C; Strain and stress distributions in model 3.

192

Fig. 5 â&#x20AC;&#x201C; Strain and stress distributions in model 4.

Fig. 6 â&#x20AC;&#x201C; Strain and stress distributions in model 5.

193

Fig. 7 â&#x20AC;&#x201C; Strain and stress distributions in model 6.

Fig. 8 â&#x20AC;&#x201C; Strain and stress distributions in model 7.

194

Fig. 9 â&#x20AC;&#x201C; Strain and stress distributions in model 8.

195