1

Introduction

Dental composites are the most frequently used direct restorative materials and have become the first choice of a majority of practitioners world-wide for the restoration of posterior teeth . The primary reason for this ascent from its introduction to dentistry approximately 50 years ago is mostly related to esthetics. The importance of the ability to replace lost or damaged tooth structure in a convenient and cost effective manner, and with an excellent esthetic outcome, cannot be overstated. In addition, the ability to use adhesive dentistry to provide restoration resistance and retention form allows for a minimal intervention approach, providing another significant advantage by conserving tooth structure . However, the longevity of composite restorations, as well as the durability of the composite material itself as a tooth replacement, is often questioned. Many believe that the most serious issue with dental composites is the fact that the polymerization reaction is accompanied by a volumetric shrinkage that generates stress within the material and leads to compromised adhesion to the tooth and a poor seal of the restoration.

Clinical studies vary widely in terms of the success rate for composite restorations, and proponents and detractors alike often use essentially the same data to support their opinions. Recent efforts by the National Institute for Dental and Craniofacial Research of the National Institutes of Health in the U.S. have targeted the development of new dental composite restoratives with enhanced service life, specifically requesting that new materials double the longevity of current materials. To justify these initiatives, NIDCR documents point to an average lifetime of 6–7 years for dental composite restorations . While it is important to note that reviews show that many clinical studies report much greater longevity for these materials , there is significant evidence for this relatively short lifespan for composites . Further, even with evidence that composites may have similar service life as dental amalgam, composite failure due to caries is typically higher than for amalgam . The longevity of both composite and amalgam is reduced in patients with high caries risk status , but the effect is more significant for composites . An extensive review of clinical studies in which dental composite and amalgam have been directly compared shows that while one retrospective, single-practice study reported improved longevity for composite vs. amalgam, the preponderance of clinical evidence demonstrates the overall enhanced longevity of amalgam restorations ( Table 1 ). This conclusion is supported by a recent Cochrane review in which the odds ratio for failure of composite over amalgam was nearly 2:1, with the increased risk for composite being due to secondary caries; admittedly the evidence was considered weak due to the limited number of acceptable studies . Thus, the general consensus is that dental composite restorations do not last as long as the profession desires, or perhaps even consider acceptable. The latter statement accounts for the fact that some dentists still hesitate to embrace this material for routine direct restoration of posterior teeth in their practice.

The question that then becomes of critical importance is why do dental composite restorations not demonstrate greater longevity? To address this question, it is important to clarify the reasons for replacement and failure of these materials. There is a distinction between these two terms, and this has been clarified for dentistry many years ago. It is likely that many restorations that may still be serviceable are replaced, and for a variety of reasons, perhaps even because it is difficult to determine their true quality . For example, decay around restorations is often difficult to confirm without removal of the existing restoration to visualize the actual state of the tooth. Stained margins, gaps at margins, fractured margins, and other obvious deficiencies, possibly with or without the presence of symptoms, may leave the dentist with a dilemma about the need for immediate treatment to prevent greater problems at a later date . In any case, these conditions are most likely related to a deterioration of the restoration with time. But the existence of a deficiency at the time of placement cannot be ruled out either. What is known is that the primary reason for replacement of dental composite restorations, even in recent studies, is caries associated with the restoration . Whether this is a recurrence of the original caries or a new caries lesion formed specifically due to the presence of the restoration (i.e. secondary caries) may be a matter of discussion and debate. Composites also fail due to chipping or fracture of the material, fracture of the tooth, discoloration, and excessive wear. All point in full or in part to deficiencies in the material or the placement technique.

While an initial deficiency is most likely a problem of technique , the fact remains that in many cases it is the characteristics of the material that influence the manifestation of this deficiency. This problem pertains to the “technique sensitivity” of dental composites, a phenomenon that has been related to their handling characteristics, shrinkage during polymerization, limited depth of cure, and requirement to be bonded to the tooth with a separate adhesive while incorporating meticulous isolation. Another potential problem is that amalgam, a material with a long history of use and familiarity, and admittedly a more forgiving placement procedure, preceded composite as the most common direct placement restorative material. This, and initial hesitance on the part of dental schools to embrace the materials and provide adequate training, prolonged the time it has taken for many clinicians to become comfortable with composites as a direct posterior material . The two main problems focused on by the profession are the polymerization shrinkage occurring during curing and the need for bonding. The two are directly related in that high initial bond strengths would likely not be required to produce a sealed interface if the materials did not shrink during curing. This article will focus on this issue of polymerization shrinkage and its accompanying stress, in terms of its potential relevance to the clinical situation.

2

Polymerization shrinkage stress – the phenomenon

In recent years, dental manufacturers have tried to address the deficiencies in dental composite restoratives through the development of enhanced dental bonding agents and composites with reduced shrinkage, or reduced shrinkage stress. The ideal dental composite would undergo zero, or at least low, shrinkage during setting. Zero shrinkage would ensure that the material remained physically adjacent to the tooth surface if originally placed there, with no subsequent dimensional change if the material itself did not absorb water with time. However, the typical dimethacrylate monomers used in commercial composites do take up water , and therefore, it may be beneficial overall for there to be some contraction of the composite, at least those based on current monomers, that will be subsequently compensated by a delayed expansion during service. In any case, the production of composites with low shrinkage, often stated as being less than 1.0% by volume, has been a goal of manufacturers for many years.

But the true concern regarding curing shrinkage, which is inevitable due to the nature of the vinyl polymerization involving reductions in intermolecular dimensions and free volume, is the internal stress created within the material . This stress is a product of the constraint of the free shrinkage of the polymer, and is dependent upon a number of factors, including the size and nature of the monomers, the acquisition of stiffness of the material during polymerization, the rate of the reaction, and the external constraints imposed by the bonding to the tooth. The many factors involved in the evolution of these stresses have been the subject of many comprehensive reviews and will not be readdressed here . The fact is that this reaction produces internal stress that cannot adequately be relaxed, either by changes in the molecular structure within the composite or by deformation at its free surfaces. Thus, these stresses are transferred to the bonded interfaces with the tooth structure creating delamination or tooth fracture whenever and wherever the localized stress exceeds the adhesion strength or the strength of the adjacent residual tooth structure . Furthermore, these stresses may increase with time, causing delayed damage to cavity margins .

It may be stated that it took many years of work by dental researchers highlighting the potential significance of polymerization shrinkage stress to influence manufacturers to develop composites with lower polymerization stress. Consider that what may have been the first article highlighting this phenomenon of curing stress was published by Bowen in 1967 . But it was not until the series of studies conducted by Davidson and Feilzer in Amsterdam during the mid-1980s that refocused attention on this issue. These studies set the stage for what may be described as a deluge of investigations into this phenomenon that followed for more than two decades. However, possibly the first composites marketed as having low shrinkage (Aelite LS, Bisco; Filtek LS, 3M ESPE; InTen-S, Ivoclar Vivadent), were not introduced to the market until approximately the mid-2000s, despite the fact that significant research was being conducted on the development of alternative monomer systems throughout the 1990s .

What is considered to be high shrinkage stress for dental composite? The answer is likely any stress that approaches or exceeds the local adhesive force or residual tooth structure strength, resulting in gap formation. However, residual stresses that remain within the restoration apply continual force on the adhesive interface with the tooth, and may further degrade that interface during function leading to debonding or damage at a later date. The end result of all of these stresses is localized or generalized delamination, either at the interface or at some point distant to the interface if the stress results in tooth fracture ( Fig. 1 ), which has been implicated as a primary factor leading to the formation of caries around dental composite restorations. The question is, “Does the evidence, either in vitro or in vivo, support this hypothesis?”

Schematic showing crack formation within enamel and delamination at margin of dental composite restoration and enamel.

2

Polymerization shrinkage stress – the phenomenon

In recent years, dental manufacturers have tried to address the deficiencies in dental composite restoratives through the development of enhanced dental bonding agents and composites with reduced shrinkage, or reduced shrinkage stress. The ideal dental composite would undergo zero, or at least low, shrinkage during setting. Zero shrinkage would ensure that the material remained physically adjacent to the tooth surface if originally placed there, with no subsequent dimensional change if the material itself did not absorb water with time. However, the typical dimethacrylate monomers used in commercial composites do take up water , and therefore, it may be beneficial overall for there to be some contraction of the composite, at least those based on current monomers, that will be subsequently compensated by a delayed expansion during service. In any case, the production of composites with low shrinkage, often stated as being less than 1.0% by volume, has been a goal of manufacturers for many years.

But the true concern regarding curing shrinkage, which is inevitable due to the nature of the vinyl polymerization involving reductions in intermolecular dimensions and free volume, is the internal stress created within the material . This stress is a product of the constraint of the free shrinkage of the polymer, and is dependent upon a number of factors, including the size and nature of the monomers, the acquisition of stiffness of the material during polymerization, the rate of the reaction, and the external constraints imposed by the bonding to the tooth. The many factors involved in the evolution of these stresses have been the subject of many comprehensive reviews and will not be readdressed here . The fact is that this reaction produces internal stress that cannot adequately be relaxed, either by changes in the molecular structure within the composite or by deformation at its free surfaces. Thus, these stresses are transferred to the bonded interfaces with the tooth structure creating delamination or tooth fracture whenever and wherever the localized stress exceeds the adhesion strength or the strength of the adjacent residual tooth structure . Furthermore, these stresses may increase with time, causing delayed damage to cavity margins .

It may be stated that it took many years of work by dental researchers highlighting the potential significance of polymerization shrinkage stress to influence manufacturers to develop composites with lower polymerization stress. Consider that what may have been the first article highlighting this phenomenon of curing stress was published by Bowen in 1967 . But it was not until the series of studies conducted by Davidson and Feilzer in Amsterdam during the mid-1980s that refocused attention on this issue. These studies set the stage for what may be described as a deluge of investigations into this phenomenon that followed for more than two decades. However, possibly the first composites marketed as having low shrinkage (Aelite LS, Bisco; Filtek LS, 3M ESPE; InTen-S, Ivoclar Vivadent), were not introduced to the market until approximately the mid-2000s, despite the fact that significant research was being conducted on the development of alternative monomer systems throughout the 1990s .

What is considered to be high shrinkage stress for dental composite? The answer is likely any stress that approaches or exceeds the local adhesive force or residual tooth structure strength, resulting in gap formation. However, residual stresses that remain within the restoration apply continual force on the adhesive interface with the tooth, and may further degrade that interface during function leading to debonding or damage at a later date. The end result of all of these stresses is localized or generalized delamination, either at the interface or at some point distant to the interface if the stress results in tooth fracture ( Fig. 1 ), which has been implicated as a primary factor leading to the formation of caries around dental composite restorations. The question is, “Does the evidence, either in vitro or in vivo, support this hypothesis?”

Schematic showing crack formation within enamel and delamination at margin of dental composite restoration and enamel.

3

Polymerization stress – in vitro evidence

Numerous laboratory studies have been conducted to evaluate the potential negative effects of contraction stress by examining marginal leakage, gap formation, cuspal deflection, bond strength, tooth cracking, and mechanical properties of the restorative. Typically these studies have incorporated situations in which stress is expected to be a significant factor, i.e. constraint of free resin shrinking by a confining 3-D cavity preparation, rapid curing protocols, and bulk placement of “non-bulk-fill” materials. While there are inconsistencies in the outcomes of different studies, there is significant evidence that higher contraction stresses lead to a greater incidence of problems. Following is a review of some of these studies and the extent to which they support the case for the deleterious effects of stress.

Many studies were conducted throughout the past 30 years or more to establish that stress is produced during the polymerization reaction of dental composite, and other dental materials. A quick search of PubMed with the terms “dental composite polymerization shrinkage stress” returns 362 articles spanning 1977 to the present. These studies show that the magnitude of the generated stress depends on a number of factors, such as monomer composition and extent of cure , filler amount and stiffness , external and internal constraint of polymer deformation , curing method , etc. Others have attempted to explain the relative contribution of the different factors when the materials are placed within cavity preparations differing in compliance . Certain studies have been conducted to probe specific aspects that relate to the effect, rather than the simple presence or causes, of these stresses.

The most obvious concern with respect to the polymerization shrinkage stresses is the detrimental effect on marginal integrity and seal of the composite restoration to the tooth. Numerous studies show that even when attempting to bond the material to a cavity preparation, the marginal seal is typically compromised, resulting in marginal gaps, stains, or leakage . This deficiency remains true even with current “improved” adhesive systems . In fact most in vitro leakage studies have shown that dental composites do not provide a perfect seal to either enamel or dentin, independent of bonding or placement method . This leakage has been described as occurring on the nano-, micro-, and macro-scale, but the most important observation may be the fact that it simply happens at all, and the true clinical consequence of it is not entirely clear. In any case, there is evidence that materials that demonstrate reduced contraction stress produce better marginal seal and less leakage .

Experiments have been conducted to show a direct correlation between enhanced leakage and higher contraction stress for different brands of composites . These studies provided a physically observable outcome for the stress that had been proven to exist in dental composites, and made the concept of stress generation more “clinically relevant” to the practitioner and researcher alike. In most studies involving the placement of composite in preparations with margins in both enamel and dentin, sealing is typically better to acid-etched enamel , suggesting that the presence of a higher bond strength at this margin produces a more perfect seal. It is also true that when stress builds up within the curing material, it will tend to be relieved at the more vulnerable dentin interface and remain sealed with the enamel interface. Thus it would be surprising to find a situation where a restoration margin demonstrated total debonding or leakage as a result of contraction stress, and rather, such an occurrence would most likely be due to inadequate adhesion ( Fig. 2 ).

Gap formation at the pulpal floor of a dental composite restoration as a result of polymerization contraction in the absence of adequate adhesion.

It is possible that the interface may remain bonded, due to adequate adhesion forces, but continue to exist in a state of stress from the polymerization contraction. This would likely only be evident in situations where the curing composite is under significant constraint imposed by the adhesion to the walls of a three-dimensional cavity, as described many years ago and characterized by the configuration factor (C-factor) . It has been shown that the C-factor alone cannot be used to predict contraction stresses for a given material, and other factors, such as the volume of the material also may have a significant effect . But studies in which composite has been placed within preparations and then either pushed out or sectioned to test adhesion have shown reduced bond strength under conditions of greater constraint, suggesting that the higher stresses negatively affected the bonds . This outcome is consistent with the evidence already noted for unsealed margins around composite restorations.

Studies have also been conducted to measure the stress produced during composite contraction by actually measuring the propagation of cracks in tooth structure and dental ceramics. A popular, though somewhat controversial, method for measuring the fracture toughness of a ceramic is through microindentation. Cracks propagating from the tips of the indent can be used to relate the resistance of the material to such propagation, i.e. a measure of its fracture toughness. Likewise, cracks propagating from existing cracks may be used to demonstrate that stresses have been produced, as well as their magnitude. This method has been used to show that stress generated by a curing composite in a 3-dimensional cavity are greatest closest to the margin and result in propagation of existing cracks . This outcome suggests that brittle enamel at margins may be subjected to cracking under the influence of composite contraction stress. Studies using this method have shown more extensive cracking for direct placement composites vs. cemented inlays and for larger restorations, i.e. larger volumes of composite , and less extensive cracking with some lower shrinkage composite formulations . This cracking in the tooth structure has also been related to the formation of “white lines” around just cured dental composite restorations in the clinical setting, which are especially apparent when the tooth surface is dried due to the highlighting of the discontinuity within the enamel or at the interface caused by the crack .

Further evidence for the effect of stress has been generated by studying the mechanical properties of composites cured under conditions of high constraint. Flexure strength and modulus has been shown to be reduced under these conditions in specimens that were cut from a construct that provided strong adhesion to the walls to ensure high contraction stress conditions . It is not known whether the reduced strength and stiffness of the material resulting from the residual stresses remaining in the composite after curing produces clinically relevant outcomes. But as stated earlier, it is likely that the residual stresses would be reduced as water sorption occurs, providing an enhanced environment for polymer molecular motions to relieve the stresses, thus negating concerns over time. Perhaps the negative effects on bond strength are of greater concern.

Some investigators have studied the direct effect of polymerization contraction on cuspal flexure of mesial-occlusal-distal cavity preparations, showing greater flexure when composites are very well cured using bulk-cure vs. incremental placement techniques , and reduced flexure when certain flexible liners are placed . A criticism of some of these studies is the amount of illumination time used to ensure complete cure of the composite, which would be considered to be excessive clinically, and leading to unrealistically high extents of cure that produce greater contraction stress and more deleterious effects. While this creates some uncertainty about the clinical relevance of the outcomes, the effect of the contraction stress, independent of how it was achieved, remains. Other studies have more recently shown that some commercial composites specifically designed to have reduced polymerization contraction and/or stress produce less deleterious outcomes on marginal adaptation or cuspal deflection .

Some evidence against the effect of contraction stress has been generated when examining models of polymerization of composites in preparations using the finite element model or other mathematical models. Certain studies do not show enhanced negative effects, such as cuspal deflection, when placing composites with equivalent curing characteristics in bulk, where stresses would be expected to be greater than when placed incrementally . These results are explained by the incremental deformation of the cavity walls with the curing of each successive composite increment, thereby reducing the total volume of composite required to fill the cavity. The result is higher residual stress than if the composite was placed and cured in bulk.

Taking all of this work into consideration, there is significant in vitro evidence that polymerization contraction stress in resin dental composite materials produces negative outcomes. There are sufficient controlled studies to demonstrate this directly. Though these studies provide further evidence for a link between contraction stress and clinically relevant outcomes, they do not confirm it.