53156257 Relationship Between Bond Strength Tests and Clinical Outcomes

d e n t a l m a t e r i a l s 2 6 ( 2 0 1 0 ) e100–e121

avai lab le at www.sc iencedi rec t .com

journa l homepage: www. int l .e lsev ierhea l th .com/ journa ls /dema

Review

Relationship between bond-strength tests and clinicaloutcomes

B. Van Meerbeek ∗, M. Peumans, A. Poitevin, A. Mine, A. Van Ende, A. Neves,J. De MunckLeuven BIOMAT Research Cluster, Department of Conservative Dentistry, School of Dentistry, Oral Pathology and Maxillo-Facial Surgery,Catholic University of Leuven, Kapucijnenvoer 7, B-3000 Leuven, Belgium

a r t i c l e i n f o

Article history:

Received 19 November 2009

Accepted 19 November 2009

Keywords:

a b s t r a c t

One often alleges that laboratory bond-strength testing cannot predict clinical effectiveness

of adhesives. Major argument to sustain this claim is the wide variation in bond-strength

values recorded for one specific adhesive among different research institutes worldwide.

The main reason for these inconsistent bond-strength measurements is supposedly the

current lack of a standard bond-strength testing protocol. This paper (and presentation)

aimed to report on an extensive literature review with regard to the different laboratory

bond-strength test methods and their data provided, along with a second extensive litera-
Review
Bond strength

Clinical effectiveness

Clinical trial

ture review on clinical effectiveness data of adhesives in terms of retention rates of adhesive

Class-V restorations. Combining both systematic reviews, we have subsequently searched

for a potential relationship between bond-strength data and clinical outcomes.

emy

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e101

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e101

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e101

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e103

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e104

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e105. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e105

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e105. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e107

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e107

Retention

Adhesive

© 2009 Acad

Contents

1. Dental adhesive technology ANNO 2009. . . . . . . . . . . . . . . . .1.1. Etch&rinse. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1.2. Self-etch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1.3. Self-adhesive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2. Measuring bond strength. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2.1. Macro-shear . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2.2. Macro-tensile/push-out . . . . . . . . . . . . . . . . . . . . . . . . . . . .2.3. Micro-tensile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2.4. Micro-shear . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2.5. Fatigue. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.6. Bond durability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2.7. KULeuven systematic review on bond strength . . . . . . . .2.8. Fracture mechanics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
∗ Corresponding author. Tel.: +32 16 33 75 87; fax: +32 16 33 27 52.E-mail address: [email protected] (B. Van Meerb

0109-5641/$ – see front matter © 2009 Academy of Dental Materials. Pudoi:10.1016/j.dental.2009.11.148

of Dental Materials. Published by Elsevier Ltd. All rights reserved.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e107. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e108. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e109

eek).blished by Elsevier Ltd. All rights reserved.

mailto:[email protected]

dx.doi.org/10.1016/j.dental.2009.11.148

d e n t a l m a t e r i a l s 2 6 ( 2 0 1 0 ) e100–e121 e101

3. Assessment of sealing ability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e1093.1. Micro-leakage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e1093.2. Marginal adaptation/gap formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e1093.3. Bacterial leakage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e1103.4. Permeability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e1103.5. Nano-leakage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e1103.6. 3D-leakage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e110

4. Clinical performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e1104.1. Class-V clinical data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e1104.2. KULeuven systematic review on Class-V clinical data, Part I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e1134.3. KULeuven systematic review on Class-V clinical data, Part II . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e114

5. Relationship between laboratory and clinical bonding effectiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e1145.1. Relationship I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e1155.2. Relationship II. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e1155.3. Relationship III . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e1155.4. Relationship IV. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e115

6. Conclusions and closing remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e116. . . . .

. . . . .

1

Tsdicstrwbasbdcciasest[

ea

1

IpiiHao2

Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. Dental adhesive technology ANNO 2009

he fast progress in dental adhesive technology has exten-ively influenced modern restorative dentistry. Althoughecayed/fractured teeth can be reconstructed minimal-

nvasively and nearly invisibly using adhesive technology, thelinical longevity of composite restorations is today still toohort [1,2]. Despite the enormous advances made in adhesiveechnology during the last 50 years, the bonded interface itselfemains the Achilles heel of an adhesive filling [3,4]. Mainlyater sorption is thought to destabilize the adhesive–toothond, though the actual interfacial degradation mechanismsre far from understood. In this context, several aspectshould be considered with regard to the strength and dura-ility of the bond to the two dental hard tissues, enamel andentin. These include the heterogeneity of tooth structure andomposition, the features of the dental surface exposed afteravity preparation, and the characteristics of the adhesivetself, such as its strategy of interaction with both substratesnd its basic physicochemical properties. Furthermore, allorts of chemical and mechanical challenges that are inher-nt to the oral environment should be taken into account,uch as there are moisture, masticatory stresses, changes inemperature and pH, and dietary and chewing related habits5].

Modern adhesive approaches can be divided into (1) antch&rinse, (2) a self-etch (or etch&dry), and (3) nowadays alsoself-adhesive approach [6].

.1. Etch&rinse

n brief, the multi-step etch&rinse approach involves a phos-horic acid-etch step that at enamel produces deep etch-pits

n the hydroxyapatite (HAp)-rich substrate, and at dentin dem-neralizes up to a depth of a few micrometers to expose a
Ap-deprived collagen mesh. The next step involves eitherseparate priming step followed by the application/curing
f a combined primer/adhesive resin following a simplified-step procedure, or a separate primer and adhesive resin

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e116. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e116

step following a 3-step procedure. The final objective is tomicro-mechanically interlock upon diffusion and in situ poly-merization of monomers into the enamel etch-pits (Fig. 1a),the opened dentin tubules and the exposed collagen network,the latter forming the well-documented hybrid layer (Fig. 2a).Without doubt, the micro-mechanical interlocking of tinyresin tags within the acid-etched enamel surface is still todaythe best achievable bond to enamel [6–8]. It not only effec-tively seals the restoration margins on the long term, but alsoprotects the more vulnerable bond to dentin against degrada-tion [9]. On the contrary, etching dentin is a rather aggressiveprocedure as it dissolves and removes (through rinsing) thenatural protection of collagen (Fig. 2a), thereby producing aresin–collagen complex that is vulnerable to degradation uponwater sorption, possibly enhanced by the documented enzy-matic degradation process [10–12]. As the most intimate andstable intermolecular interaction possible, primary chemicalinteraction between resin and the mainly organic substanceremaining at acid-etched dentin would definitely contribute tothe bond durability, but is however lacking [13,14]. This defi-cient chemical interaction should most likely be regarded asthe major shortcoming of today’s etch&rinse approach. Never-theless, traditional 3-step etch&rinse adhesives are still todayregarded as ‘gold-standard’.

1.2. Self-etch

The self-etch approach can be further subdivided into a‘strong’ (pH < 1), an ‘intermediately strong’ (pH ≈ 1.5), a ‘mild’(pH ≈ 2), and an ‘ultra-mild’ (pH ≥ 2.5) self-etch approachdepending on the self-etching or demineralization intensity[15]. Self-etching only dissolves the smear layer, but does notremove the dissolved calcium phosphates, as there is no rinsephase. The more intense the self-etching, the more calciumphosphates are dissolved and embedded within the interfacial
transition zone [16]. Such resin-encapsulated calcium phos-phates within the exposed collagen mesh are however rathersoluble (Fig. 3) and may explain the lower laboratory and clin-ical bonding performance of strong self-etch adhesives, in

e102 d e n t a l m a t e r i a l s 2 6 ( 2 0 1 0 ) e100–e121

Fig. 1 – TEM photomicrographs illustrating the adhesive–enamel interface for an etch&rinse (a) versus a ‘mild’ self-etchadhesive (b). Note the partially dissolved HAp-crystals and much deeper interaction at phosphoric-acid-etched enamel(following the etch&rinse approach in (a)) versus the visibly hardly ‘touched’ HAp-structure and superficial but tightinteraction at self-etched enamel (following the mild self-etch approach in (b)). The white porosities in between theHAp-crystals (b) should be regarded as an artifact due to the diamond knife-sectioning procedure through the brittle enamelsubstrate.

Fig. 2 – TEM photomicrographs illustrating the adhesive–dentin interface for an etch&rinse (a) versus a ‘mild’ self-etchadhesive (b), taken after heavy-metal staining and at the same magnification. While the etch&rinse approach (using 30–40%phosphoric acid) exposes collagen deeply (a), all collagen remained protected by HAp following the mild self-etch approach (b).

d e n t a l m a t e r i a l s 2 6 ( 2

Fig. 3 – TEM photomicrograph illustrating theadhesive–dentin interface produced by a ‘strong’ self-etchadhesive. Besides deep hybridization, typical of such anadhesive are the intensely stained interfibrillar spaces inbetween the exposed collagen fibrils of the hybrid layer,m

pg[m[tcttputc[ia

Fb

ost likely representing dissolved calcium phosphates.

articular to dentin [17]. At enamel, they however perform ineneral much better due to this more aggressive self-etching18]. The less intense the self-etching, the more bur-smear

ay interfere with the eventual bonding performance (Fig. 4)19,20]. In particular ‘mild’ (pH ≈ 2) self-etch adhesives appearo deal reasonably well with bur-smear, producing a submi-ron hybrid layer with substantial HAp-crystals still protectinghe collagen fibrils (Fig. 2b). Functional monomers, in par-icular like 10-MDP (10-methacryloyloxydecyl dihydrogenhosphate), have been proven to interact with this resid-al HAp through primary ionic binding [21,22]. The resultantwofold micro-mechanical and chemical bonding mechanism
losely resembles that of glass-ionomers (Fig. 5; see below)6,21,23–27]. However, chemical bonding potential on its owns insufficient; the formed ionic bonds should also be stable inn aqueous environment. Chemical bonding promoted by 10-
ig. 4 – TEM photomicrographs showing the resin–smear compleonded to enamel in (a) and to dentin in (b).

0 1 0 ) e100–e121 e103

MDP appeared not only more effective, but also more stable inwater than that provided by other functional monomers like4-MET (4-methacryloyloxyethyl trimellitic acid) and phenyl-P (2-methacryloyloxyethyl phenyl phosphoric acid), in thisorder [22]. The dissolution rate of the respective calcium saltsof these three monomers, as measured by AAS (or atomicabsorption spectroscopy), was inversely related to their chem-ical bonding potential revealed by XPS: the more intense thechemical bonding potential, the less the resultant calcium saltcould be dissolved. This finding was further explained in the‘AD-concept’ or the adhesion–decalcification concept that dic-tates if molecules will either adhere to or decalcify mineralizedtissues [28,29].

Two-step self-etch adhesives involve the application ofa separate, more hydrophobic adhesive resin after thehydrophilic self-etch primer. This makes the interface morehydrophobic and thus better seals it to the direct benefit ofbond durability. Finally, the most simple- and fast-to-use 1-step (self-etch) adhesives generally come with some sacrificein bonding performance. This lower bonding efficiency hasbeen thoroughly documented in laboratory research, and mustbe ascribed to, among others, to a lower polymerization con-version and thus inferior mechanical properties, enhancedwater sorption through osmosis from the host dentin, poten-tial phase-separation effects when the adhesive solution islow in or free of HEMA (2-hydroxyethyl methacrylate), filler de-bonding within the adhesive resin through hydrolysis of thesilane coupling, potential smear interference for ultra-mildself-etch adhesives, and reduced shelf life in particular withregard to one-component formulations [30–35].

1.3. Self-adhesive

Glass-ionomers and resin-modified glass-ionomers are ‘self-adhesive’ through submicron hybridization, combined withwell-proven primary ionic interaction of polyalkenoic acid
with calcium within HAp (Fig. 5). Polyalkenoic acid possessesabundant functional carboxylic groups that ‘grab’ HAp simul-taneously at different and remote sites. Other ‘self-adhesive’materials are the so-called self-adhesive luting composites
x at the interface of an ‘ultra-mild’ self-etch adhesive

e104 d e n t a l m a t e r i a l s 2 6

Fig. 5 – TEM photomicrograph illustrating the twofold‘self-adhesive’ mechanism of (resin-modified)glass-ionomers, consisting of micro-mechanicalinterlocking through submicron hybridization, along withprimary electrostatic binding of the carboxyl groups of

used [39]. The bond strength can be measured statically using

polyalkenoic acid as the principal functional polymer withthe glass-ionomer formulation to HAp.

that have been introduced some years ago [36–38]. They areoften mistakenly termed as ‘self-etching’, while they inter-act only very superficially with dentin without clear signs ofdemineralization (Fig. 6). Finally, it is in the line of expecta-tions that such self-adhesive luting composites will soon lead
to the development of self-adhesive flowable and later full-restorative composites.
While adhesive–enamel/dentin interfacial characteriza-tion using scanning and even more reliably using transmission

Fig. 6 – TEM photomicrographs illustrating the interface of a selfsmear-free) dentin in (a) and to bur-cut dentin in (b).

( 2 0 1 0 ) e100–e121

electron microscopy (potentially supplemented with chemicalinterfacial analysis) definitely reveals a deeper insight in theunderlying mechanisms of adhesion, the actual bonding effec-tiveness of today’s adhesive approaches should be measuredusing a mechanical bond-strength test.

2. Measuring bond strength

By definition, the ideal bond-strength test should be in thefirst place easy (meaning low technique-sensitivity) and rel-atively fast. In general, advantages of ‘laboratory testing’ are,among others, (1) the quickness to gather data on a specificparameter/property, (2) the relative ease of the test method-ology commonly used, (3) the possibility (and necessity) tomeasure one specific parameter, while keeping all othervariables constant, (4) to be able to directly compare the per-formance of a new and/or experimental material/techniquewith that of the current ‘gold-standard’, (5) to be able to testsimultaneously many (of course within certain limits) exper-imental groups within one study set-up, and (6) to be ableto mostly use relatively unsophisticated and inexpensive testprotocols/instruments. The final objective of a laboratory testshould obviously be to gather data in prediction of the eventualclinical outcome.

In order to measure the bonding effectiveness of adhesivesto enamel and dentin, diverse methodologies can today be

a MACRO- or MICRO-test set-up, basically depending upon thesize of the bond area. The MACRO-bond strength, with a bondarea larger than 3 mm2, can be measured in ‘shear’, ‘tensile’,or using a ‘push-out’ protocol.

-adhesive luting composite to either fractured (and thus

6 ( 2

2

DtsstumitmDfcdbttctrAavsMrcqrtWscitpfm

t‘teotatta

rleBowao

d e n t a l m a t e r i a l s 2

.1. Macro-shear

efinitely most commonly used is the shear bond-strengthechnique [39]; it was found to have been used in 26% ofcientific papers reporting on bond strength. As no furtherpecimen processing is required after the bonding procedure,he shear bond-strength test undoubtedly thanks its high pop-larity in companies and other research institutes as it is theost easy and fastest method. In an attempt to standard-

ze the test protocol, specific jigs have been prepared, suchhe Ultradent jig (Ultradent, Salt Lake City, UT, USA) and the

ore recent SDI rig (SDI, Bayswater, Victoria, Australia) (Fig. 7).espite such standardization attempts, a meta-analysis of

actors involved in bond-strength testing revealed the signifi-ant influence of various parameters, like those related to theentin substrate (i.e. nature of teeth), to the composite andonding area (i.e. composite stiffness), to the storage condi-ions of the bond assemblies (i.e. thermo-cycling), and to theest design (i.e. crosshead speed) [40]. In particular, a strongorrelation was found between the mean bond strength andhe failure mode: the higher the bond strength, the higher theate of cohesive failure. Also the type of composite is crucial.

stiffer composite will result in different stress distributionst the interface and lead to a higher apparent bond-strengthalue. Our extensive literature review conducted on the occa-ion of this presentation at the Portland Academy of Dentalaterials (ADM) meeting (see below) revealed a positive cor-

elation between bond strength and the E-modulus of theomposite used [41,42]. Somewhat surprisingly, adhesives areuite often tested in bond-strength studies along with theespective composite of the same company in the sense thathe composite was fine-tuned to the adhesive (or vice versa).

hile this may be true to a certain extent, doing so, conclu-ions can only be drawn at the level of the adhesive/compositeombination, and certainly not at the level of the adhesivetself. Another source of variability are the different configura-ions employed to apply the shear force, including wire loops,oints and knife edges [43]. In particular, the use of a wire loopor shear bond-strength tests appeared to concentrate stress

ore near the interface rather than a knife edge.The most frequently ignored guideline in the test pro-

ocol following the ISO/TS 11405 specification (2003) is thata limitation of the bonding area is important’ [44]. Manyimes, the adhesive is applied to the entirely exposednamel/dentin substrate, by which the bonding area is notnly substantially larger, but also the shear load is appliedo the adhesive–composite interface rather than to thedhesive–tooth interface (Fig. 7). Besides such major flaws inhe test protocol, there is a good consensus in current litera-ure that the shear stress is not uniformly distributed acrossnd not necessarily focused at the true interface [43,45–48].

Nevertheless, the shear bond-strength test probablyemains a very popular test to screen new adhesive formu-ations on their bonding effectiveness. The most noteworthyxample of shear bond-strength testing is the ‘Battle ofonds’ series by Degrange et al. [49,50]. A gigantic dataset
f more than 16,000 ‘immediate’ bond strengths to dentinas collected, and revealed upon pooling of the data per
dhesive class a statistically significant higher bond strengthf etch&rinse adhesives than of self-etch adhesives. More-

0 1 0 ) e100–e121 e105

over, 3-step etch&rinse adhesives outperformed the 2-stepetch&rinse adhesives, and they on their turn the 2-step self-etch adhesives that presented with a pooled bond strengththat was not statistically significant from that recorded forthe most simple-to-use 1-step adhesives. The significantlyhighest bond strength was measured for OptiBond FL (Kerr,Orange, CA, USA) that somewhat surprisingly performed onlynot significantly better than the relatively new 1-step adhesiveEasy Bond (3M ESPE, Seefeld, Germany). Very striking is alsothe operator effect, widely varying for each product. Indeed,very pertinent information gathered in this mega-test is thecoefficient of variance that gives some idea on the technique-sensitivity of individual products. Products that presentedwith the highest bond strength, were not necessarily the leasttechnique-sensitive. Care should nevertheless be taken not toover-interpret the data, this in the first place because of theshear bond strength set-up (see above) and in particular alsobecause of the solely ‘immediate’ bond-strength data recordedwithout any kind of bond aging involved (see below).

2.2. Macro-tensile/push-out

Less popular is a macro-tensile bond-strength approach thatnevertheless can be used to measure for instance the bondstrength of cements to hard materials such as ceramics andmetal alloys [51,52]. A push-out approach has also beenemployed, in particular to dynamically test the fatigue resis-tance of adhesive–dentin bonds [53–56]. It has however neverbeen adopted as a universal bond-strength test method, mostlikely because of the more laborious specimen preparationinvolved as well as the more time-consuming methodology.This method appeared however very useful to test the reten-tion of posts luted in root canals [55,56].

2.3. Micro-tensile

Common MICRO-bond strength is typically measured intensile, as micro-tensile bond-strength testing (�TBS) wasdeveloped in 1994 by Sano et al. [57]. The bond area testedis much smaller compared to that of the ‘MACRO’ tests, beingabout 1 mm2 or less. After the bonding procedure, some fur-ther specimen processing or the actual preparation of themicro-specimens is required, rendering the test more labo-rious and technique-sensitive. Nevertheless, a long list ofadvantages is typically ascribed to �TBS when compared tomacro-bond-strength testing, of which the most importantare the better economic use of teeth (with multiple micro-specimens originating from one tooth), the better control ofregional differences (e.g. peripheral versus central dentin),the better stress distribution at the true interface (avoid-ing cohesive failure in tooth substrate or composite), etc.[57,58]. This renders the �TBS test more versatile, as multiplespecimens obtained from a single tooth enable more inven-tive study set-ups and better controlled substrate variables.However, several micro-specimen preparation protocols arebeing used worldwide, one being more technique-sensitive
than the other. Today, so-called ‘trimmed’ and ‘non-trimmed’micro-specimens are prepared, both having advantages anddisadvantages [59,60]. Non-trimmed micro-specimens aredefinitely most easy and fast to prepare. Trimming the micro-


Fig. 7 – Figure showing the Ultradent jig (top) and the SDI rig (bottom), both specifically designed and fabricated for shearbond-strength testing. An inappropriate, though commonly used specimen-preparation technique involves the applicationof the adhesive to the entire tooth surface, on top of which a composite cylinder is bonded (top). As shown by the fracturedspecimen (top right), the shear stress caused the specimen to de-bond at a much larger area than the area to which the

ables
composite cylinder was bonded. The SDI rig much better enconfined area (bottom).
specimens at the interface to so-called hourglass-shapedspecimens better concentrates stress at the interface, butinvolves a more invasive specimen procedure. When thistrimming is not carefully performed, interfacial defects mayeasily be introduced and early set off crack-propagation dur-ing the tensile loading of the micro-specimen, eventuallycausing the interface to fail prematurely at a lower (bond)strength [61]. Originally, interfaces were trimmed by free handusing a dental handpiece [57,62]. Despite being very labori-ous, such micro-specimen trimming largely depends on theskills of the operator, thereby introducing a learning curveand non-negligible factor of technique-sensitivity. Therefore,semi-automatic trimming of micro-specimens using a so-called MicroSpecimen Former (University of Iowa, Iowa City,IA, USA) is highly advisable to trim rectangular specimens ina well-controlled and standardized way into specimens with acircular cross-section. Also other factors such as specimen–jigattachment, specimen-loading speed and specimen align-ment, do influence the final outcome, and therefore shouldbe standardized within the test set-up [58,63,64].

One major issue of dispute in current literature is therequired number of individual teeth from which many micro-specimens can be prepared to be statistically sound [65]. Anelegant way to handle this problem is to use every tooth as its
own control. This means that the tooth surface for bonding isdivided into 2 or, for practical reasons, a maximum of 4 sectors,each receiving a different treatment. Probably, the best way isto know which micro-specimen comes from which tooth, and
both the adhesive and the composite be applied to a

then use linear mixed models (ANOVA with random effects)to analyze the data. As also recommended by the ISO Tech-nical specification on testing adhesion to tooth structure (ISONo. 11405, 2003) [44], another way would be to apply survivalanalysis like the Weibull model or Cox proportional hazardusing the force that is required for bond failure [66]. Althoughnot generally agreed upon, a minimum of 5 teeth seems rea-sonable in order to keep the advantage of more economic useof teeth using a �TBS approach.

A second issue involves the hard-to-neglect amount ofpre-testing failures often recorded with micro-tensile bond-strength testing [65]. Several approaches have been applied todeal with the pre-testing failures: (a) exclude all pre-testingfailures from further (statistical) analysis, which obviouslyoverestimates the actual bond strength; (b) assign a bond-strength value of for instance 0 MPa to each pre-testing failure.This actually penalizes the product too severely, as there wasa certain bond strength; and (c) a modification of the for-mer approach by assigning a pre-determined value to eachpre-testing failure, as for example the lowest �TBS measuredwithin the respective group [66–68]. Apart from the obviouseffect on the mean �TBS value, these data transformationsmay also affect the subsequent statistical analysis. Especiallyin case of many pre-testing failures, methods (b) and (c) will
result in a data distribution that is inevitably skewed. As aresult, more powerful parametric statistical analyses cannotbe applied anymore and one has to rely on more basic non-parametrical methods such as the Kruskall–Wallis test. Just

6 ( 2

ibrplnspsmatd

cbaoe

2

AfwopatbTpttcbsptv[

2

BmistHittpfc(fa(


gnoring pre-testing failures, following method (a), should note done (perhaps except when the micro-specimen failedemote from the interface in one of the adherents), as a highroportion of pre-testing failures is typically associated with

ow �TBSs measured for the other micro-specimens that didot de-bond prior to testing [65]. Moreover, adhesives that con-istently present with relatively high �TBSs, do not suffer fromre-testing failures. Eventually, micro-specimen processinghould be as a-traumatic as possible and the use of specialeasures to avoid pre-testing failures, such as the use of

lginate or gypsum to fill up the space between the slabs,hereby better supporting the slabs during the second 90-egree-turned cut, is highly advisable.

Although hard to scientifically prove, a micro-tensile proto-ol appears to be able to discriminate adhesives better on theironding performance than a traditional shear bond-strengthpproach (see above), most likely the reason why up to 60%f current scientific papers reporting on bond strengths havemployed the �TBS approach [41,42].

.4. Micro-shear

lso in order to generate as much specimens as possiblerom a single tooth, a micro-shear bond-strength test (�SBS)as introduced in 2002 [69,70]. This test combines the easef manipulation with the ability to test several specimenser tooth. The very fine composite build-up (cylinder) withtypical diameter of 0.7 mm, in combination with a rela-

ive thick adhesive layer, may however result in considerableending and variable and non-uniform loading conditions.his non-uniform stress distribution is probably even moreronounced as compared to macro-shear bond testing. Fur-hermore, it is impossible to confine the adhesive to the areaested, as required by ISO-standard No. 11405 (2003) [44]. Basi-ally due to these major shortcomings, this �SBS test has noteen adopted very well, since only 7% of recent bond-strengthtudies have used this protocol [41,42]. In a recent study com-aring both MICRO-bond methodologies, it was shown thathe micro-shear values were about 1/3 of the micro-tensilealues, while no difference in failure analysis was observed71].

.5. Fatigue

esides static bond-strength tests, theoretically clinicallyore relevant is to test adhesive interfaces dynamically, as

n the clinical situation tooth-composite bonds are seldomubjected to acute tensile/shear stresses. It is rather exposedo cyclic sub-critical loadings produced during chewing [3].owever, since such fatigue tests are even more labor-

ntensive and time-consuming than static bond-strengthests, a steadily growing, but still only low number of fatigueests have been tried out throughout recent years on theirotential to predict clinical effectiveness. In literature, 6 dif-erent fatigue tests have been reported on, as there are,hronologically, (1) a macro-push-out fatigue test [53,54,72],
2) a macro-shear fatigue test [53,72–77], (3) a micro-rotaryatigue test [78–80], (4) a micro-shear fatigue test [81], (5)
micro-4-point-bend fatigue test [82,83] and most recent6) a micro-tensile fatigue test [84,85]. Despite the alleged

0 1 0 ) e100–e121 e107

need for more fatigue testing of adhesives and even thoughseveral typical fatigue phenomena can be observed (Fig. 8)[72,78], little new information on bonding effectiveness isprovided than that revealed by the easier and faster staticbond-strength tests. For example, micro-rotary [79] as well asmicro-tensile fatigue testing [84,85] revealed a similar superiorbonding effectiveness of the 3-step ‘gold-standard’ etch&rinseadhesive OptiBond FL (Kerr) over the 2-step ‘gold-standard’self-etch adhesive Clearfil SE Bond (Kuraray, Tokyo, Japan), thaton its turn bonds significantly better than the 1-step adhesiveG-Bond (GC, Tokyo, Japan). In addition, these fatigue tests havelargely been applied to dentin with bonding to enamel beingmuch more difficult to assess in fatigue.

2.6. Bond durability

While early or 24-h bond strength is mostly measured, thereis a definite need to test bonding effectiveness of adhesivesunder more clinically relevant circumstances or upon agingof the specimens. Indeed, many currently available dentaladhesives have presented with a relatively high short-termbond strength, while not always equally favorable clinicalresults have been obtained. Therefore, more laboratory effortshould definitely go to durability testing of adhesion, ratherthan measuring the ‘immediate’ bond strength. In up to35% of the bond-strength studies, some kind of ‘aging’ fac-tor is added to the study design [41,42]. Water storage andthermo-cycling are the most popular artificial aging meth-ods, but also other methodologies as mechanical loading anddegradation by enzymes and various chemical substanceshave been employed in literature [3]. Actually, the probablymost widely used aging technique is simple thermo-cycling,although not always sufficiently long. A short thermo-cyclingregimen of 500 cycles, as recommended by the ISO TR 11450standard (2003) [44], is of little use, while sometimes onlyvery long thermo-cycling up to 100,000 cycles can discrimi-nate differences in bond durability of different adhesives [86].A well-validated method to assess bond durability is how-ever the storage of micro-specimens in water [87]. Doing so,it was reported that within about 3 months, all classes ofadhesives exhibited mechanical and morphological evidenceof degradation that resembles in vivo aging [88]. Today, thecurrent challenge in adhesive dentistry is indeed to make theadhesive–tooth interface more resistant against aging, therebyrendering the restorative treatment more predictable in termsof clinical performance on the long term. The longevity ofbonded restorations is to a large extent related to the degrada-tion of the adhesive interface, which may occur in a relativelyshort term, depending on the way the adhesive has beenmanipulated, on the actual adhesive approach and on theadhesive composition. Hydrolysis of interface components,such as dentinal collagen and resin, due to water sorption,potentially enhanced by enzymatic degradation, and sub-sequent elution of the break-down products are the majorfactors lately identified as involved in this bond degradationprocess [3,4,10,26,30,89–91]. In fact, none of the current adhe-
sives or techniques is today able to produce an interface thatis absolutely resistant to degradation, but fortunately manyresearch efforts are currently devoted to improving bond dura-bility.


Fig. 8 – Figure illustrating the micro-rotary fatigue approach developed by De Munck et al. [78–80], thereby showing someture
typical fatigue phenomena such as ‘beach marks’ at the frac
2.7. KULeuven systematic review on bond strength

At Leuven, we have recently thoroughly reviewed the literatureusing the search term ‘dentin bond strength’ in PubMed with alimitation to PubMed-registered papers published in the last 5years [41,42]. This search was done on 18/01/2009 and revealed1019 studies, of which at the time of this paper 744 papers wereevaluated on the following inclusion criteria: (a) measuringbond strength to dentin; (b) involving at least two commercialadhesives; (c) only in combination with light-cure composites(no glass-ionomers, nor resin cements for instance); (d) andreporting at least the type of bond-strength test, the meanbond strength in MPa, the standard deviation and the totalnumber of specimens tested. So far, 214 studies were includedin the dataset, leading to 1828 bond-strength data for spe-cific adhesive/composite combinations. Most popular testsconducted were the micro-tensile bond-strength tests (1012individual test groups were inserted in the database), followedby macro-shear (287), micro-shear (221), and macro-tensile(113) bond-strength tests. Only 26 macro-push-out and 12micro-push-out bond-strength tests were reported on. Clearly,the most recent 1-step adhesives were tested most frequently(531), followed closely by both the 2-step etch&rinse (510)and 2-step self-etch adhesives (450), while somewhat surpris-ingly the traditional 3-step etch&rinse adhesives have beentested least frequently (173). This finding confirms that still toooften in many of today’s bond-strength studies so-called ‘gold-
standard’ controls, by preference a well-documented andconsistently good-performing 3-step etch&rinse or 2-step self-etch adhesive, are lacking in the study set-ups. Perhaps shouldjournals be more strict in publishing bond-strength studies
plane.

solely when a proper ‘gold-standard’ control is included. Alsofollowing the systematic review [41,42], the bond strengthof (resin-modified) glass-ionomers is scarcely tested, whichcould be due to the well-known fact that they tend to failcohesively within the material itself, while the actual bondstrength to tooth tissue can hardly be determined. Amongthe individual adhesives tested, the 2-step self-etch adhe-sive Clearfil SE Bond (Kuraray) was tested most (250 times),followed closely by the 2-step etch&rinse adhesive Scotch-bond 1/Single Bond (3M ESPE) that was tested 211 times.All other adhesives were substantially less frequently tested,among which the 1-step adhesive Adper Prompt L-Pop (3MESPE) was tested 76 times, the 1-step adhesive Clearfil S3Bond (Kuraray) 73 times, the 1-step adhesive One-up BondF (Tokuyama, Tokyo, Japan) 62 times, the 2-step etch&rinseadhesive One-Step (Bisco, Schaumburg, IL, USA) 56 times, the2-step etch&rinse adhesive Prime&Bond NT (Dentsply, Kon-stanz, Germany) 52 times, the 3-step etch&rinse adhesiveOptiBond FL (Kerr) 50 times, etc. From this dataset, alreadysome preliminary conclusions can be drawn. At first, thereis an excessive number of bond-strength data published inliterature and the variability between the studies is huge.Especially the difference between the different test set-ups isremarkable; a simple general linear model revealed that theweighted mean of all micro-tensile bond-strength values isabout twice as high as the weighted mean of the macro-sheartest (16 versus 30 MPa). As the fine details of the bond-strength
test in the first place determine the height of the absolutebond strength, it therefore does make not much sense to statethat a certain absolute value in MPa is needed for a product tobe clinically effective. Likewise, striving for a standard bond-


Fig. 9 – Upon the systematic review recently conducted onbond strength at Leuven [41,42], the weightedbond-strength means were found to clearly differ peradhesive class. The weighted mean and large confidenceinterval for the (2-step) glass-ionomers is less trustworthy,st

staitms1(pa2aiow

2

A[tutpcIaattNnm[oma

ince only few products could be included at the momenthis paper was written.

trength test, enabling bond strength to be measured as arue material property (thus independent from the test set-up,nd reproducible in different research institutes worldwide),s most likely never achievable. Following the present sys-ematic review [41,42], significant differences in the ‘pooled’

ean bond strength can be observed, as the weighted bond-trength means of individual adhesives ranges from about2 MPa (for Absolute, Dentsply-Sankin, Tokyo, Japan) to 49 MPafor OptiBond FL, Kerr) and the weighted bond-strength meanser adhesive class range from 31 MPa for 3-step etch&rinsedhesives, to 29 MPa for 2-step self-etch adhesives, 26 MPa for-step etch&rinse adhesives, and 20 MPa for 1-step self-etchdhesives (Fig. 9). The weighted mean and large confidencenterval for the 2-step glass-ionomers is less trustworthy sincenly few products could be included at the moment this paperas written.

.8. Fracture mechanics

s concluded in the ADM review paper from Scherrer et al.92], the high scatter in dentin bond-strength data found inhe literature, regardless which current bond-strength test issed, should be attributed primarily to non-uniform stress dis-ributions due to a number of geometrical, loading, materialroperty and specimen-preparation variables, as this was alsoonfirmed by Finite Element Analysis [45,47,48,59,60,93,94].t has been proposed that an interfacial fracture mechanicspproach that studies the failure of an interface by the initi-tion and growth of a single, large, dominant crack, such ashe fracture toughness test, would be more appropriate foresting the relatively brittle adhesive–tooth interface [92,95].evertheless, this fracture mechanics approach has so farot received much support despite several fracture toughnesseasurements of adhesives have previously been reported on

96–103]. Probably, the big hurdle to take is the more labori-us and time-consuming specimen preparation, by which thisethodology no longer meets with the basic requirements of

n easy and fast, and ‘first’ product-screening test. In addition,

0 1 0 ) e100–e121 e109

one should not underestimate the perhaps higher technique-sensitivity involved in fracture mechanics, by which also usingthis methodology large scattering of the data among thediverse research centers worldwide can hardly be avoided.

3. Assessment of sealing ability

As the longevity of an adhesive composite restoration ismainly affected by leakage of oral fluids along the interfacebetween restorative material and tooth substrate [104,105],probably more clinically relevant than bond-strength stud-ies is to evaluate the capacity of an adhesive to maintainthe tooth-restoration transition sealed. It is especially thoughtto predict better the clinical performance of adhesives withregard to the occurrence of postoperative sensitivity and/orsecondary caries [105].

3.1. Micro-leakage

Leakage on the micrometer scale (micro-leakage) is oftenassessed in vitro microscopically on cross-sections using awide variety of tracers or dyes that are able to infiltratethe composite–tooth interface, such as silver nitrate, methy-lene blue, basic fuchsin, and erythrosine [104], and evensome radioactive markers [105]. Although such micro-leakageassessment is in principle even easier than bond-strengthmeasurements (and therefore done very frequently), thereliability of conventional micro-leakage protocols remainscontroversial [104,106,107]. Like for bond-strength studies,a systematic review of dye penetration studies for restora-tive materials concluded that a comparison of study resultsis impossible due to the great variability of methodol-ogy parameters employed worldwide [104]. In addition, thecross-section-based procedure is destructive and part of thespecimen is lost through slicing. More importantly, the pro-cedure allows only a limited 2D-view on the distribution ofthe marker fluid, by which it is for instance not possible todetermine the point of deepest leakage with certainty. Newerso-called micro-permeability tests make use of fluorescentdyes like rhodamine, that are ‘loaded’ to the pulp chamberin order to investigate at higher resolution using confocalmicroscopy the sealing ability of adhesives at the interfaceitself [108–110].

3.2. Marginal adaptation/gap formation

Another technique to assess marginal sealing ability involvesa semi-quantitative marginal analysis of replicas of restoredteeth using SEM, combined with custom-made softwareto measure the length of the margin part that exhibits a(gap-free) ‘continuous margin’ versus any kind of marginaldefect following a pre-determined set of margin qualities[106,111,112]. Besides being laborious and time-consuming(most likely being the main reasons why this methodologyis solely consistently used in a few research centers), another
major drawback of this technique is that it only enables toassess marginal sealing in 2D at the outer surface. The advan-tage of using replica’s however is that the same margins canbe assessed multiple times after having exposed the spec-

s 2 6
e110 d e n t a l m a t e r i a l
imens to some kind of aging procedure [112]. A simplifiedmarginal sealing screening test evaluates the gap formationaround standardized cylindrical cavities, a test introduceda long time ago [106,113]. Based on this simple approach,an automated system has recently been developed that candetect marginal gaps automatically using an optical sensor[114]. Using this automated methodology, 21 adhesives werefor instance screened on their ability to produce gap-freemargins [115]. This study revealed that the etch&rinse adhe-sives demonstrated the best marginal adaptation, followedby the 2-step self-etch adhesives. The 1-step self-etch adhe-sives yielded the poorest marginal adaptation and the highestscatter in test results between materials and within the samematerial.

3.3. Bacterial leakage

A bacterial micro-leakage model has mainly been applied inendodontic research [116–119].

3.4. Permeability

A full-quantitative test to measure sealing ability has beendeveloped on basis of the early dentin permeability studiesby Reeder and Pashley et al. [120–123], but is perhaps tootechnique-sensitive to become a standard test to screen adhe-sives on their sealing ability [124,56].

3.5. Nano-leakage

Just like ‘classical’ micro-leakage assessment, also qual-itative imaging and certainly quantification of so-callednano-leakage (as per definition being ‘the diffusion of smallions or molecules within the hybrid layer in the absenceof gap formation’ [125,126]), using SEM, TEM or confocalmicroscopy [127–130] should also be interpreted with care[131]. The high regional variability makes it very difficult toobtain representative information with regard to the resis-tance of adhesives against nano-leakage. Moreover, at least5 independent studies agree on the absence of any correla-tion between nano-leakage and bond strength of adhesives[132–135].

3.6. 3D-leakage

For leakage, either micro- or nano-leakage, to be a valuablecriterion for evaluating adhesive performance, a fully quan-titative methodology that measures leakage in three ratherthan two dimensions is most likely required [131]. In searchfor a full-quantitative 3D leakage test, microfocus-CT with apotential resolution of the order of 1 �m (depending on thehardware capabilities and X-ray source characteristics) hasrecently been tried out [132–140].

4. Clinical performance

Despite the importance of laboratory studies attempting topredict clinical performance of biomaterials, clinical trialsremain the ultimate way to collect scientific evidence on the

( 2 0 1 0 ) e100–e121

clinical effectiveness of a restorative treatment [2,6,141,142].The popularity of in vitro studies in the field of adhesive den-tistry may in part be ascribed to the rapid evolution of dentaladhesive technology and the resultant high turnover of adhe-sive systems, which often tempts manufacturers to release asuccessor product on the market even before its precursor hasbeen clinically evaluated, at least on the long term. By carryingout in vivo studies, all possible aging factors play at the sametime, thereby disclosing whether an adhesive is truly reliablefor routine clinical practice.

4.1. Class-V clinical data

Clinical effectiveness of adhesives should best be determinedusing Class-V clinical trials, because (1) such lesions do notprovide any macro-mechanical retention, so that ineffectivebonding will result in early restoration loss, (2) Class-V restora-tion margins are located in enamel as well as in dentin,(3) lesions are commonly located on vestibular surfaces ofanterior teeth and premolars, thus providing good accessfor the restorative procedure as well as evaluation (visuallyusing an explorer and magnifying glasses, and even by SEMif epoxy replicas are made), (4) preparation and restorationof Class-V lesions is minimal and relatively easy, therebysomewhat reducing practitioner variability, (5) lesions are rel-atively wide spread and prevail on multiple teeth, facilitatingpatient selection and enabling split-mouth study designs, and(6) Class-V lesions have a relatively small C-factor, by whichthe mechanical properties of the composite used are moreor less unimportant and the actual adhesive performancedetermines the eventual clinical outcome to a great extent[2,6,142–145]. Retention, marginal integrity and clinical micro-leakage are usually the key parameters recorded to judgeupon clinical effectiveness of adhesives. Definitely the mostobjective criterion is retention, by which Class-V clinical datarecorded worldwide can be mutually compared [2].

At KULeuven (Belgium), there is a long track record in con-ducting Class-V clinical trials [2,6,141,142,145–158], as part ofour overall translational research line within BIOMAT on thedevelopment of new adhesive dental restorative materials.This research line involves basic research investigating molec-ular interactions of functional monomers with inorganic toothtissue components, applied research up to mechanical testingof adhesive interfaces, and eventually clinical research intothe ultimate clinical performance of new adhesive materi-als/procedures in the mouth of patients (KULeuven UniversityHospital). From the start in the mid-1980s, almost the samestandard study design has been employed. The data collec-tion was done by well-trained and calibrated examiners in anattempt to maintain the study quality equally high, basicallyenabling the Class-V data that have been collected throughoutthe past 25 years be compared mutually. The major outcomestrived for was always the basic clinical effectiveness of exper-imental and/or commercial adhesives, while sufficient carewas taken to include an appropriate control(s) within eachindividual study. The major modifications of the initial clinical
trial protocol to the current one entail a gradually increas-ing number of restorations/patients (a), as well as a shift of a‘split-mouth’ design to a ‘pair-tooth’ study design (b). We cur-rently place at least 50 Class-V restorations per experimental

d e n t a l m a t e r i a l s 2 6 ( 2 0 1 0 ) e100–e121 e111

Fig. 10 – Graph representing the retention rate of Class-Vrestorations up to 13 years of clinical service for three3-step etch&rinse adhesives [156,161]. This study isconducted at the Catholic University of Leuven in Belgium(KULeuven). Be aware that the retention rate (in the Y-axis)s

gct(tpthnrptl‘ituttomcoarwtsVcTattp

(

Fig. 11 – Graph representing the retention rate of Class-Vrestorations up to 8 years of clinical service for a 2-stepself-etch adhesive with and without selective phosphoricacid etching of enamel [157,165]. This study is conducted at

require urgent restoration repair and certainly not restorationreplacement. As flexure at the tooth cervix has been advancedas one of the three principal factors causing non-carious cer-vical lesions [162,163], it was reasonable to hypothesize that

Fig. 12 – Graph representing the retention rate of Class-Vrestorations up to 3 years of clinical service for two 1-stepself-etch adhesives as compared to that of a 3-stepetch&rinse control adhesive [158,166,167]. This study isconducted at three university institutes, namely the

tarts at 75%.

roup in at least 50 patients, by which we comply with theurrent recommendations for conducting randomized con-rolled clinical trials (RCT) of dental restorative materials [159].a) Although in the latter recommendations it is stated thathe clinical investigator performs only two restorations peratient (one restoration per experimental group), we believehat it is of primary importance to include a sufficientlyigh number of restorations in the clinical trial. We thereforeowadays employ a ‘multi-restoration approach’, by which weestore ALL Class-V lesions that need to be restored within oneatient, equally dividing ad random the lesions per experimen-al group. It is of course correct that treating more than oneesion per condition within one mouth introduces a potentialpatient factor’ (dependent data), but this does not necessar-ly mean that the power of the study cannot be increased byhis multi-restoration approach. Provided that a suitable (butnfortunately more complex) statistical analysis is applied,he supplementary restorations may still add new informa-ion. The eventual increase in power of the study will dependn the strength of correlation between the repeated measure-ents (within one patient). Only in the worst case, when the

linical outcome is completely determined by patient factors,ne can return to a one-restoration-per-condition-per-patientnalysis by ad random (and multiple times) selecting pairs ofestorations within each mouth. Most often, the patient factorill have some influence, but will never completely determine

he outcome so that treating more lesions per patient with theame material (as patients often present with multiple Class-

lesions) is an easy and valid option to prevent conductinglinical trials with low sample size and thus power [160]. (b)he second ‘pair-tooth’ design indicates that restorations willlways be placed in pairs of two restorations (each experimen-al group) in nearly equal teeth (first and second premolar at
he same side, left and corresponding right incisor, canine orremolar, respectively).
At the moment, we run three Class-V studies at LeuvenFigs. 10–12). (KUL-1) A long-term study evaluates the clinical

the Catholic University of Leuven in Belgium (KULeuven).Be aware that the retention rate (in the Y-axis) starts at 50%.

effectiveness of two 3-step etch&rinse adhesives in combi-nation with a hybrid, stiffer composite versus a micro-filled,more flexible composite (Fig. 10) [156,161]. After 13 years ofclinical service, 94% of the restorations in the control group(OptiBond FL in combination with Prodigy, Kerr) were still infunction. After this long clinical service, obviously marginaldefects and discolorations were observed at a steadily grow-ing incidence, but most are only minor defects that do not

University of Sao Paulo in Brazil (USP), the SuleymanDemirel University at Turkey (SDU), and the CatholicUniversity of Leuven in Belgium (KUL). Be aware that theretention rate (in the Y-axis) starts at 50%.

s 2 6
similar stress is imposed to restored Class-V lesions, therebypromoting marginal degradation or even dislodgement of therestoration [150,164]. This study however failed to prove thatthe composite stiffness affects the clinical longevity of cervicalcomposite restorations. (KUL-2) In a second long-term study,the clinical effectiveness of a ‘mild’ 2-step self-etch adhesive(Clearfil SE Bond) was evaluated (Fig. 11) [157,165]. At 8 years,the clinical effectiveness appeared excellent, with selectiveacid-etching of the enamel margins only having some minoreffect on secondary clinical parameters (higher incidence ofsmall marginal defects/discolorations at enamel that were ofclinical negligible relevance, since they did not require anyrestorative intervention at the 8-year recall). (KUL-3) The mostrecent clinical Class-V study is multi-center based and evalu-ates the clinical performance of two 1-step self-etch adhesives(G-Bond, GC; Clearfil S3 Bond, Kuraray) in comparison tothe ‘gold-standard’ 3-step etch&rinse adhesive (OptiBond FL,Kerr) (Fig. 12) [158,166,167]. Each of the studies did exam-ine two adhesives following the same multi-restoration andpaired-tooth design explained above. Up to 2 years of clinicalservice, no major differences in clinical performance betweenthe 1-step adhesives and the 3-step etch&rinse control wereobserved. Obviously, longer-term clinical evaluation is neededin respect to bond durability. In a recent Class-V clinical trial byBurrow and Tyas [168], a 3-year retention rate of, respectively,100% and 97% was recorded for the 1-step adhesives G-Bond(GC, in combination with the composite Gradia, GC) and forClearfil S3 Bond (Kuraray, in combination with the compositeClearfil ST, Kuraray).

At UNC (USA), a wide spectrum of adhesives, among
which 3-step etch&rinse adhesives (OptiBond Dual cure, Kerr;ProBond, Dentsply; Scotchbond MP, 3M-ESPE; Gluma SolidBond, Hereaus-Kulzer, Hanua, Germany), 2-step etch&rinseadhesives (Excite, Ivoclar-Vivadent, Schaan, Liechtenstein;
Fig. 13 – Graph representing the retention rate of Class-V restoracommercial adhesives (Single Bond (3M ESPE) was tested twice).of North Carolina at Chapel Hill (UNC) [168–175]. Be aware that th

( 2 0 1 0 ) e100–e121

OptiBond Solo, Kerr; Prime&Bond NT, Dentsply; Single Bond,3M ESPE), 1-step (self-etch) adhesives (Adper Prompt LP-2, 3MESPE; iBond, Hereaus-Kulzer; One-up Bond, Tokuyama; Vari-glass, Dentsply), and (resin-modified) glass-ionomers (Fuji IILC, GC; Variglass, Dentsply), have been evaluated in non-carious Class-V lesions in a consistent way (Fig. 13) [169–176].For the 3-step etch&rinse adhesive, OptiBond Dual Cure (Kerr),12-year data are available, reporting a very high retention rateof 93% [175]. While almost all other retention curves showa clear downward trend, that of OptiBond Dual Cure (Kerr)remains high and rather stable on a long-term perspective.Especially the difference with its 2-step etch&rinse successor(OptiBond Solo, Kerr) is striking, as for this adhesive after 8years the retention rate dropped to 69% [176]. Likewise, the8-year retention rate of the acetone-based 2-step etch&rinseadhesive, Prime&Bond 2.1 (Dentsply), dropped to even 59%[176]. Also the water-based 3-step and 2-step etch&rinse adhe-sives Scotchbond MP and Single bond (both from 3M ESPE)performed rather equally with retention rates of, respectively,78% and 75% at 3 years [172]. They both however cannotcompete with the ethanol-based 3-step etch&rinse adhesiveOptiBond Dual Cure (Kerr), being the precursor of the still cur-rently commercially available successor OptiBond FL (Kerr).Regarding the clinical performance of 1-step adhesives, it isworth to mention the steep drop in retention rate of AdperPrompt LP-2 (3M ESPE) to 81% at 1.5 years of clinical service[172], while for the more recently marketed ‘mild’ 1-step self-etch adhesives, iBond (Hereaus-Kulzer), so far no losses wererecorded (100% retention rate at 1.5 years) [174], and for One-up Bond (Tokuyama) a less favorable retention rate of 90% was
recorded [173].
Two other long-term Class-V studies worth mentioningwere performed at Umeå University (Sweden) by Van Dijkenet al. (Fig. 14) [177,178]. At Umeå, the retention rates of Class-V

tions up to 12 years of clinical service for 13 differentThese studies are (have been) conducted at the Universitye retention rate (in the Y-axis) starts at 50%.


Fig. 14 – Graph representing the retention rate of Class-Vrestorations up to 13 years of clinical service for 13 differentcommercial adhesives. These studies have been conductedat Umeå University (Sweden) [176,177]. Some recent clinicalretention rates [178] are depicted by letters in the graph. Forthe 3-step etch&rinse adhesive cmf (‘a’: Saremco; incombination with the composite els, Saremco), a 1.5-yearretention rate of 96% was recorded versus a 1.5-yearretention rate of 91.8% for the 2-step etch&rinse adhesiveXP-Bond (‘b’: Dentsply; also in combination with thecomposite els, Saremco). For the 2-step self-etch adhesiveClearfil Protect Bond (Kuraray), a 1.5-year retention rate of98.2% was recorded when used in combination with boththe composite Tetric Ceram (‘c’: Ivoclar-Vivadent) and els(‘d’: Saremco). A 3-year retention rate of, respectively,96.7%, 96.5% and 93.7% have been recorded at UmeåUniversity for G-Bond (GC), AdheSE (Ivoclar-Vivadent) andC

rtcmwoaCeArLsgew(saamtg

ionomers. Their low annual failure rates can be ascribed totheir unique self-adhesiveness, based on the twofold micro-mechanical and chemical bonding mechanism (see above).Despite their excellent clinical performance in terms of reten-

learfil SE Bond (Kuraray).

estorations placed using 13 different adhesives were sys-ematically recalled up to 13 years of clinical service. Fororrect interpretation of the Umeå data, it is important toention that all restorations were placed in dentin lesionsithout any intentional enamel involvement. In this respect,nly the clinical bonding performance to dentin was evalu-ted, in contrary to most other (also the above-mentioned)lass-V studies, in which the adhesive was applied to bothnamel and dentin in a so-called mixed enamel/dentin lesion.fter 13 years, only 5 adhesives presented with a retention

ate higher than 50%: the 3-step etch&rinse adhesive Clearfiliner Bond (Kuraray) with 74%; the 3-step etch&rinse adhe-ive OptiBond Dual Cure (Kerr) with 59%; the resin-modifiedlass-ionomer Vitremer (3M ESPE) with 64%; the 2-step self-tch adhesive ART Bond (Coltène, Altstätten, Switzerland)ith 59% and the 3-step etch&rinse adhesive Syntac Classic

Ivoclar-Vivadent) with 64%. The four worst-performing adhe-ives (retention rates below 30% at 13 years) were all obsoletedhesives that fortunately are no longer commercially avail-
ble. The clinical bonding performance of the more recentlyarketed adhesives is clearly better, corresponding to the bet-
er laboratory data reported on in literature and thus with thereat progress dental adhesive technology has undergone in

0 1 0 ) e100–e121 e113

the last decade. As at Umeå recalls have systematically beendone on a year basis, the data clearly show the continuousdegradation in bonding effectiveness of all adhesives tested,though at a significantly varying speed for each individualadhesive. Again, if enamel would have been bonded to as well,the retention data would obviously have been much higher.This once again confirms that bonding to dentin remains verychallenging, in the laboratory as well as clinically. Some morerecent Class-V retention rates of studies conducted at UmeåUniversity are mentioned in Fig. 14 [179]. They again confirmthe improved clinical effectiveness of recent-generation adhe-sives, even when applied in less application steps.

Considering only the really long-term clinical data, notmuch information on long-term clinical effectiveness of adhe-sives is today available, especially when abstracting the Umeådata because of the dentin-only study design. The most favor-able long-term (retention) data have been recorded for the3-step etch&rinse adhesives Optibond FL (Kerr) and its precur-sor Optibond Dual-Cure (Kerr). At Leuven, Optibond FL (Kerr)presented with a 94% retention rate at 13 years [161], whileeven a 97% retention rate was recorded at Chicago [180]. AtChapel Hill, an almost equally high retention rate of 93% wasrecorded for OptiBond Dual Cure (Kerr) at 12 years [175]. Theonly other favorable retention rate (97%) was recorded for the2-step self-etch adhesive Clearfil SE Bond (Kuraray) at 8 years[157,165].

4.2. KULeuven systematic review on Class-V clinicaldata, Part I

In an attempt to get a better insight in the current over-all clinical performance of adhesives, we refer to a relativelyrecently conducted systematic review, in which the annualfailure rates were determined per adhesive class (Fig. 15)[2]. According to that review, the best clinical performancewith regard to retention has so far been achieved by glass-

Fig. 15 – Graph representing the mean annual failure ratesper adhesive class, determined according to a systematicreview of Class-V clinical trials of adhesives during theperiod 1998–2004 [2].

s 2 6 ( 2 0 1 0 ) e100–e121

Fig. 16 – Graph representing the mean annual failure ratesper adhesive class, determined according to a systematicreview of Class-V clinical trials of adhesives during theperiod 2004–2009 [182], in continuation of the earlier con-ducted review during the period 1998–2004 [2]. The overall


tion, glass-ionomers commonly present with lower aestheticfeatures when compared to resin-based restorative materi-als. The poorer mechanical properties of glass-ionomers alsoexplain the lower scores achieved in bond-strength tests whencompared to those of resin-based adhesives, by which a glass-ionomer typically fails cohesively rather than it de-bondsfrom the tooth surface (see above). Besides glass-ionomers,in particular 3-step etch&rinse adhesives have exhibited areasonably good clinical effectiveness (Fig. 15) [2]. The clin-ical durability of 3-step etch&rinse adhesives confirms theirgenerally superior laboratory results, in which they are con-sidered as ‘gold-standard’ and often employed as controlto compare the performance of new-generation adhesiveswith. According to the same standard, 2-step self-etch adhe-sives tend to approach the clinical performance of 3-stepetch&rinse adhesives in terms of low annual failure rates(Fig. 15) [2]. Their ability to provide a shallow but uniformhybrid layer, along with their capability to chemically bondto the dentin substrate seems to play an important role toresist long-term hydrolytic degradation (see above). Com-monly, the clinical performance of such self-etch adhesivesdoes not vary substantially from one study to another, whichis indicative of their rather low technique-sensitivity. Further-more, another clinically important benefit is, as mentionedbefore, that these self-etch adhesives are repeatedly associ-ated with very low levels of postoperative sensitivity [181].In general, 2-step etch&rinse adhesives have performed lessfavorably than the conventional 3-step version (Fig. 15) [2].Laboratory studies have corroborated these results, ascribingtheir poorer performance to their higher hydrophilicity andreduced hybridization potential (see above). It is noteworthythat irrespective of the number of application steps, acetone-based etch&rinse adhesives have generally performed lesssatisfactorily than their water/ethanol-based alternatives. Theabove-mentioned high technique-sensitivity of acetone-basedadhesives must be the reason for their compromised long-term clinical data. According to this review [2], a so farrather inefficient clinical performance has been noted forthe newest generation of 1-step adhesives. Widely varyingretention scores have been recorded, indicating their hightechnique-sensitivity despite their user-friendliness (Fig. 15).Such lower bonding performance must be ascribed to themany concerns advanced earlier.

4.3. KULeuven systematic review on Class-V clinicaldata, Part II

On the occasion of this manuscript and the ADM 2009 meet-ing at Portland, the systematic review of the literature thatappeared in the period 1998 until 2004 [2], was continued inthe same way for the literature that appeared in the period2004–2009 and has provided new data with regard to annualfailure rates per adhesive class [182,183] (Fig. 16). Besides thecalculation of the annual failure rates per adhesive class forthe new period 2004–2009, also both datasets were combined,revealing annual failure rates for the 10-year period 1998–2009.
Still the lowest annual failure rates have been recorded for theglass-ionomers, irrespective of the time period during whichthe literature was reviewed. Equally favorably performed theso-called ‘mild’ and ‘intermediately strong’ 2-step self-etch
mean annual failure rates per adhesive class for the total10-year review period (1998–2009) are mentioned as well.

adhesives. This is not that surprising taking into accountthat these self-etch adhesives make use of a bonding mech-anism that most closely approaches that of glass-ionomers(see above). Including all 2-step self-etch adhesives (also theso-called ‘strong’ self-etch adhesives) substantially increasesthe annual failure rates. This can also be explained as these‘strong’ self-etch adhesives significantly performed worse inlaboratory, especially regarding their bonding efficiency todentin. As mentioned above, the aggressive self-etch proce-dure of these ‘strong’ self-etch adhesives relatively deeplydissolves HAp (exposing collagen). As these adhesive do notinvolve a rinse step, the dissolved calcium phosphates areembedded within the hybrid layer. These calcium phosphatesare however not very hydrolytically stable, thereby seriouslyjeopardizing the bond longevity. Somewhat higher annual fail-ure rates were recorded for the 3-step etch&rinse adhesives,but they decreased in the more recent literature review, alsohaving resulted in a decreased annual failure rate for thewhole 10-year review period (1998–2009). The annual failurerate of the 2-step etch&rinse adhesives remained nearly sta-ble, while that of the most simple-to-use 1-step adhesivessubstantially improved in the most recent literature, even toan annual failure rate in line with that recorded for the 3-step etch&rinse adhesives in the same period. For the whole10-year review period, the annual failure rate of the 3-stepetch&rinse adhesives is still significantly lower. Nevertheless,it is noteworthy that the most recent generation of 1-stepadhesives have substantially improved and evolve definitelyin the good direction (Fig. 16).

5. Relationship between laboratory and
clinical bonding effectiveness
The ultimate question is if there is a relationship betweenthe bonding effectiveness measured in the laboratory with

6 ( 2

twAqtb

5

AtPs

5

Ibafidcticbte

FdfsrCcp


he clinical effectiveness determined in patients. In otherords, can we predict clinical effectiveness in the laboratory?lthough it is hard to give a straightforward answer to thisuestion, there are some trends that certainly point to cer-ain associations between laboratory and clinical data on theonding effectiveness of adhesives.

.1. Relationship I

good correlation coefficient of 0.81 was found betweenhe annual failure rates reported in the systematic review ofeumans et al. [2] and the ‘Battle-of-the-Bonds’ shear bond-trength data from Degrange et al. [49,50].

.2. Relationship II

n a study by Heintze et al. [184], searching for a correlationetween semi-quantitative margin analysis data (see above),s performed in two research centers, and a somewhat arti-cial clinical index, consisting of retention loss, marginaliscoloration, and marginal integrity (weighted), only a weakorrelation was found. The in vitro measured ‘marginal adap-ation’ appeared to only have a mediocre value in terms ofts ability to predict the clinical performance of adhesives (in
ervical cavities). Unfortunately, the analysis was hamperedy the great variability in clinical results, but also by the facthat no correlation was found between the in vitro methodsmployed at both research institutes involved.
ig. 17 – Correlation analysis between bond-strength data [41,42ata were extracted and a weighted mean was calculated, consid

or those data that were recorded after the specimens were subjetorage, thermo-cycling, etc.) (bond strength – durability). For theetention rate at 2 years) as well as after 5 years of clinical servicorrelation analysis revealed that the bond strength seems not torrelation). The bond-strength data gathered after in vitro agingerformance on the longer term (5 years) to some extent (r = 0.58

0 1 0 ) e100–e121 e115

5.3. Relationship III

In a more recent study by Heintze et al. [185], a correlationanalysis was performed between micro-tensile bond-strengthdata and the same clinical index. Again, no significant correla-tion could be found between the micro-tensile bond-strengthdata and the clinical index, nor with the clinical retentionrate and marginal integrity. The only significant correlationfound existed between micro-tensile bond strength, espe-cially measured after 6-month water storage, and the marginaldiscoloration of Class-V restorations. The authors thereforesuggested that early marginal staining could be predictive forfuture retention loss of non-carious cervical restorations. Oneshould however keep in mind that the clinical evaluation ofmarginal staining is rather subjective, and may lead to sub-stantially more variance in data when considered worldwide.

5.4. Relationship IV

On the occasion of this manuscript and the ADM 2009meeting at Portland, we have searched for a potential rela-tionship of the laboratory bond-strength data obtained inthe recently performed systematic review [41,42], with theclinical retention rates collected in the systematic reviewon the clinical effectiveness of adhesives in Class-V lesions
(Fig. 17) [2,182,183]. For this analysis, the complete database(bond strength – all) was used, but also a database filteredfor bond-strength data measured after the specimens hadbeen subjected to some kind of in vitro aging procedure (bond
] and clinical Class-V retention rates [2,182]. Bond-strengthering all data in the database (bond strength – all), or onlycted to some kind of durability factor (as for example waterclinical data, retention rates after 2 years (Class-V

e (Class-V retention rate at 5 years) were gathered.o predict the 2-year clinical outcome (no significantappear more clinically relevant and do predict the clinical11, p = 0.0475).

s 2 6

r


strength – durability). Regarding the clinical data, the reten-tion rate at 2 years was first chosen (Class-V retention rateat 2 years), as this was the point in time when most reten-tion rates were available. In a second phase, also the retentionrate at 5 years was used (Class-V retention rate at 5 years),which is the longest recall time that still revealed sufficientdata. Using these criteria, the preliminary correlation anal-ysis did not reveal any significant correlation between bothparameters for the 2-year clinical data and the 5-year clini-cal data, the latter correlated to the complete bond-strengthdataset (Fig. 17). A significant, quite reasonable correlation wasnevertheless found between the aged bond-strength data andthe 5-year clinical data. The fact that no one-to-one relation-ship between the in vitro and in vivo 2-year data exists, shouldbe attributed to the fact that many adhesives then still showa very high retention rate (about half of the studies revealedretention rates higher than 95% at 2 years). Another plausiblereason is that regarding the aged bond-strength data no cor-rection was for instance performed for the kind of aging. So, aregime of 500 thermo-cycles was for example pooled with 1-year water-storage data, although the latter aging procedureshould have resulted for sure in more bond degradation thanthe former. On the other hand, as could be expected fromthe longer-term 5-year clinical data, a trend for associationbetween in vitro bond strength and the clinical data becomesapparent for the complete dataset, and even significant whenall non-aged bond-strength data were excluded.

6. Conclusions and closing remarks

Adhesive technology has undergone great progress in the lastdecade. In light of the major drawbacks attributed to all-in-one adhesives, conventional 3-step etch&rinse adhesives and(mild) 2-step self-etch adhesives are still the benchmarks fordental adhesion in routine clinical practice. When bondingto enamel, an etch&rinse approach is definitely preferred,indicating that simple micro-mechanical interaction appearssufficient to achieve a durable bond to enamel. When bondingto dentin, a mild self-etch approach is superior, as it involves(like with glass-ionomers) additional ionic bonding with resid-ual HAp. This additional primary chemical bonding definitelycontributes to bond durability. Altogether, when bonding toboth enamel and dentin, selective etching of enamel followedby the application of the 2-step self-etch adhesive to bothenamel and dentin currently appears the best choice to effec-tively and durably bond to tooth tissue.

Along with the fast evolution of multi-step adhesivestoward 1-step adhesives, luting composites to adhesively luteindirect ceramic restorations underwent a similar evolutiontoward simple-to-use and less technique-sensitive 1-step lut-ing agents. Hence, these self-adhesive luting composites donot need any kind of pre-treatment of the tooth substrate,thereby bringing the development of self-adhesive compositescloser to reality.

Design and synthesis of functional monomers with strong
and stable chemical affinity to HAp is definitely a worthwhilepathway toward new and improved adhesive technology.Actually, many studies on dental adhesive technology arenowadays empirical, basically testing the bond strength of dif-
( 2 0 1 0 ) e100–e121

ferent cocktails of adhesive solutions to enamel and dentin inthe laboratory. The adhesive formulation that scores best com-monly makes it rapidly to the market, after which the superiorlaboratory performance of the product is hopefully confirmedin independent randomized controlled clinical trials. Muchknowledge on the underlying mechanisms of adhesion toenamel and dentin has been gathered by numerously imag-ing adhesive–tooth interfaces with all sorts of microscopes.On the contrary, the actual molecular interactions at the inter-face have hardly been explored. Also many basic questionsstill remain unanswered, like for instance why certain func-tional monomers have better bonding potential than others,and why mild self-etch adhesives with additional chemicalbonding do not perform better at enamel. Thus, fundamen-tal for future design and development of dental adhesivesis the further unravelling of the interfacial molecularinteractions.

Finally, the ultimate objective of this paper (and presen-tation) was to search for a potential relationship betweenlaboratory bond-strength data and clinical outcomes. After anextensive discussion of the different laboratory test method-ologies to measure bond strength (and sealing ability), andthe presentation of the most recent clinical effectivenessdata (with regard to retention rates of adhesive Class-Vrestorations), some clear indications for correlation of labo-ratory bond strength with clinical retention rates of Class-Vrestorations have been advanced, in particular of ‘aged’bond-strength data with medium-term retention rates. Con-sequently, besides measuring the ‘immediate’ bond strengthof adhesives to enamel and dentin, measuring the ‘aged’ bondstrength should be encouraged in order to predict the clinicaleffectiveness of adhesives.

Acknowledgments

The authors would like to thank Michael Burrow et al. (Univer-sity of Melbourne), Ed Swift et al. (University of North-Carolinaat Chapel Hill), and Jan Van Dijken et al. (Umeå University) fordisclosing their most recent Class-V clinical data, and Sieg-ward Heintze et al. (Ivoclar-Vivadent) for sharing their mostrecent laboratory and clinical correlation data.

e f e r e n c e s

[1] Manhart J, Chen H, Hamm G, Hickel R. Buonocore memoriallecture. Review of the clinical survival of direct and indirectrestorations in posterior teeth of the permanent dentition.Oper Dent 2004;29:481–508.

[2] Peumans M, Kanumilli P, De Munck J, Van Landuyt K,Lambrechts P, Van Meerbeek B. Clinical effectiveness ofcontemporary adhesives: a systematic review of currentclinical trials. Dent Mater 2005;21:864–81.

[3] De Munck J, Van Landuyt K, Peumans M, Poitevin A,Lambrechts P, Braem M, et al. A critical review of the
durability of adhesion to tooth tissue: methods and results.J Dent Res 2005;84:118–32.
[4] Breschi L, Mazzoni A, Ruggeri A, Cadenaro M, Di Lenarda R,Dorigo ES. Dental adhesion review: aging and stability ofthe bonded interface. Dent Mater 2008;24:90–101.

6 ( 2
[5] Mjör IA, Shen C, Eliasson ST, Richter S. Placement andreplacement of restorations in general dental practice inIceland. Oper Dent 2002;27:117–23.

[6] Van Meerbeek B, De Munck J, Yoshida Y, Inoue S, Vargas M,Vijay P, et al. Buonocore memorial lecture: adhesion toenamel and dentin: current status and future challenges.Oper Dent 2003;28:215–35.

[7] Buonocore MG. A simple method of increasing theadhesion of acrylic filling materials to enamel surfaces. JDent Res 1955;34:849–53.

[8] Gwinnett AJ, Matsui A. A study of enamel adhesives. Thephysical relationship between enamel and adhesive. ArchOral Biol 1967;12:1615–20.

[9] De Munck J, Van Meerbeek B, Yoshida Y, Inoue S,Vargas M, Suzuki K, et al. Four-year water degradation oftotal-etch adhesives bonded to dentin. J Dent Res2003;82:136–40.

[10] Pashley DH, Tay FR, Yiu C, Hashimoto M, Breschi L,Carvalho RM, et al. Collagen degradation by host-derivedenzymes during aging. J Dent Res 2004;83:216–21.

[11] Hebling J, Pashley DH, Tjäderhane L, Tay FR. Chlorhexidinearrests subclinical degradation of dentin hybrid layers invivo. J Dent Res 2005;84:741–6 [Erratum in: J Dent Res 200685:384].

[12] De Munck J, Van den Steen PE, Mine A, Van Landuyt KL,Poitevin A, Opdenakker G, et al. Inhibition of enzymaticdegradation of adhesive–dentin interfaces. J Dent Res2009;88:1101–6.

[13] Xu J, Stangel I, Butler IS, Gilson DF. An FT-Ramanspectroscopic investigation of dentin and collagen surfacesmodified by 2-hydroxyethylmethacrylate. J Dent Res1997;76:596–601.

[14] Nishiyama N, Suzuki K, Komatsu K, Yasuda S, Nemoto K. A13C NMR study on the adsorption characteristics of HEMAto dentinal collagen. J Dent Res 2002;81:469–71.

[15] De Munck J. An in vitro and in vivo study on the durabilityof biomaterial–tooth bonds. PhD dissertation, KULeuven;2004.

[16] Inoue S, Van Meerbeek B, Vargas M, Yoshida Y,Lambrechts P, Vanherle G. Adhesion mechanism ofself-etch adhesives. In: Tagami J, Toledano M, Prati C,editors. Advanced adhesive dentistry. Granadainternational symposium 1999. Cirimido (Como): GraficheErredue; 2000. p. 131–48.

[17] Brackett WW, Covey DA, St Germain Jr HA. One-yearclinical performance of a self-etching adhesive in class Vresin composites cured by two methods. Oper Dent2002;27:218–22.

[18] Perdigão J, Lopes MM, Gomes G. In vitro bondingperformance of self-etch adhesives. II. Ultramorphologicalevaluation. Oper Dent 2008;33:534–49.

[19] Ermis RB, De Munck J, Cardoso MV, Coutinho E, VanLanduyt KL, Poitevin A, et al. Bond strength of self-etchadhesives to dentin prepared with three different diamondburs. Dent Mater 2008;24:978–85.

[20] Cardoso MV, Coutinho E, Ermis RB, Poitevin A, Van LanduytK, De Munck J, et al. Influence of dentin cavity surfacefinishing on micro-tensile bond strength of adhesives. DentMater 2008;24:492–501.

[21] Van Meerbeek B, Yoshida Y, Van Landuyt K, Perdigão J, DeMunck J, Lambrechts P, et al. In: Summitt JB, Robbins JW,Hilton TJ, Schwartz RS, editors. Fundamentals of operativedentistry. A contemporary approach. 3rd ed. Chicago:Quintessence Publishing; 2006. p. 183–260.

[22] Yoshida Y, Nagakane K, Fukuda R, Nakayama Y, Okazaki M,Shintani H, et al. Comparative study on adhesiveperformance of functional monomers. J Dent Res2004;83:454–8.

0 1 0 ) e100–e121 e117

[23] Yoshida Y, Van Meerbeek B, Nakayama Y, Snauwaert J,Hellemans L, Lambrechts P, et al. Evidence of chemicalbonding at biomaterial–hard tissue interfaces. J Dent Res2000;79:709–14.

[24] Fukuda R, Yoshida Y, Nakayama Y, Okazaki M, Inoue S,Sano H, et al. Bonding efficacy of polyalkenoic acids tohydroxyapatite, enamel and dentin. Biomaterials2003;24:1861–7.

[25] Yiu CK, Tay FR, King NM, Pashley DH, Sidhu SK, Neo JC, etal. Interaction of glass-ionomer cements with moistdentin. J Dent Res 2004;83:283–9.

[26] Tay FR, Sidhu SK, Watson TF, Pashley DH. Water-dependentinterfacial transition zone in resin-modified glass-ionomercement/dentin interfaces. J Dent Res 2004;83:644–9.

[27] Coutinho E, Van Landuyt K, De Munck J, Poitevin A, YoshidaY, Inoue S, et al. Development of a self-etch adhesive forresin-modified glass ionomers. J Dent Res 2006;85:349–53.

[28] Yoshida Y, Van Meerbeek B, Nakayama Y, Yoshioka M,Snauwaert J, Abe Y, et al. Adhesion to and decalcification ofhydroxyapatite by carboxylic acids. J Dent Res2001;80:1565–9.

[29] Yoshioka M, Yoshida Y, Inoue S, Lambrechts P, Vanherle G,Nomura Y, et al. Adhesion/decalcification mechanisms ofacid interactions with human hard tissues. J Biomed MaterRes 2002;59:56–62.

[30] Tay FR, Pashley DH. Have dentin adhesives become toohydrophilic? J Can Dent Assoc 2003;69:726–31.

[31] Tay FR, Frankenberger R, Krejci I, Bouillaguet S, Pashley DH,Carvalho RM, et al. Single-bottle adhesives behave aspermeable membranes after polymerization. I. In vivoevidence. J Dent 2004;32:611–21.

[32] Yiu CK, Hiraishi N, Chersoni S, Breschi L, Ferrari M, Prati C,et al. Single-bottle adhesives behave as permeablemembranes after polymerisation. II. Differentialpermeability reduction with an oxalate desensitiser. J Dent2006;34:106–16.

[33] Van Landuyt KL. Optimization of the chemical compositionof dental adhesives. Towards a simplified and durable,universal enamel/dentin adhesive. PhD dissertation,KULeuven; 2008.

[34] Van Landuyt K, Mine A, De Munck J, Jaecques S, PeumansM, Lambrechts P, et al. Are one-step adhesives easier to useand better performing? Multi-factorial assessment ofcontemporary one-step self-etch adhesives. J Adhes Dent2008;11:175–90.

[35] Van Landuyt KL, De Munck J, Mine A, Peumans M,Lambrechts P, Van Meerbeek B. Predominant failureunderneath hybrid-layer with mild self-etch adhesivesafter aging. In: CED-IADR. 2009 [Abstract #31].

[36] Hikita K, Van Meerbeek B, De Munck J, Ikeda T, Van LanduytK, Maida T, et al. Bonding effectiveness of adhesive lutingagents to enamel and dentin. Dent Mater 2007;23:71–80.

[37] Radovic I, Monticelli F, Goracci C, Vulicevic ZR, Ferrari M.Self-adhesive resin cements: a literature review. J AdhesDent 2008;10:251–8.

[38] Monticelli F, Osorio R, Mazzitelli C, Ferrari M, Toledano M.Limited decalcification/diffusion of self-adhesive cementsinto dentin. J Dent Res 2008;87:974–9.

[39] Burke FJT, Hussain A, Nolan L, Fleming GJP. Methods used indentine bonding tests: an analysis of 102 investigations onbond strength. Eur J Prosthodont Rest Dent 2008;16:158–65.

[40] Leloup G, D’Hoore W, Bouter D, Degrange M, Vreven J.Meta-analytical review of factors involved in dentinadherence. J Dent Res 2001;80:1605–14.

[41] De Munck J, Peumans M, Poitevin A, Mine A, Van Ende A,Van Meerbeek B. Literature review on factors affecting thebond strength to enamel/dentin. Unpublishedobservations.

s 2 6
[42] De Munck J, Mine A, Poitevin A, Van Ende A, Van MeerbeekG. Testing bond strength. A review of the literature. In:Academy of Dental Materials 2009 Abstract. 2009.

[43] DeHoff PH, Anusavice KJ, Wang Z. Three-dimensional finiteelement analysis of the shear bond test. Dent Mater1995;11:126–31.

[44] ISO. Dental materials – testing of adhesion to toothstructure. Technical specification no. 11405; 2003.

[45] Van Noort R, Noroozi S, Howard IC, Cardew G. A critique ofbond strength measurements. J Dent 1989;17:61–7.

[46] Pashley DH, Sano H, Ciucchi B, Yoshiyama M, Carvalho RM.Adhesion testing of dentin bonding agents: a review. DentMater 1995;11:117–25.

[47] Versluis A, Tantbirojn D, Douglas WH. Why do shear bondtests pull out dentin? J Dent Res 1997;76:1298–307.

[48] Sudsangiam S, van Noort R. Do dentin bond strength testsserve a useful purpose? J Adhes Dent 1999;1:57–67.

[49] Degrange M, Hitmi L, Bouter D, Gonthier S, Basset F, BIijaouiJ. Efficiency of new enamel–dentin bonding systems:assessment by general practitioners. In: Wilson NHF, RouletJF, Fuzzi M, editors. Advances in operative dentistry.Quintessence Publishing; 2001. p. 173–84.

[50] Degrange M, Lapostolle B. L’expérience des batailles desadhésifs. L’Information Dentaire 2007;89:112–8.

[51] Abreu A, Loza MA, Elias A, Mukhopadhyay S, Looney S,Rueggeberg FA. Tensile bond strength of an adhesive resincement to different alloys having various surfacetreatments. J Prosthet Dent 2009;101:107–18.

[52] Kern M, Barloi A, Yang B. Surface conditioning influenceszirconia ceramic bonding. J Dent Res 2009;88:817–22.

[53] Drummond JL, Sakaguchi RL, Racean DC, Wozny J,Steinberg AD. Testing mode and surface treatment effectson dentin bonding. J Biomed Mater Res 1996;32:533–41.

[54] Frankenberger R, Krämer N, Petschelt A. Fatigue behaviourof different dentin adhesives. Clin Oral Investig 1999;3:11–7.

[55] Goracci C, Tavares AU, Fabianelli A, Monticelli F, Raffaelli O,Cardoso PC, et al. The adhesion between fiber posts androot canal walls: comparison between microtensile andpush-out bond strength measurements. Eur J Oral Sci2004;112:353–61.

[56] Zicari F, Couthino E, De Munck J, Poitevin A, Scotti R, NaertI, et al. Bonding effectiveness and sealing ability offiber-post bonding. Dent Mater 2008;24:967–77.

[57] Sano H, Shono T, Sonoda H, Takatsu T, Ciucchi B, CarvalhoR, et al. Relationship between surface area for adhesionand tensile bond strength – evaluation of a micro-tensilebond test. Dent Mater 1994;10:236–40.

[58] Pashley DH, Carvalho RM, Sano H, Nakajima M, YoshiyamaM, Shono Y, et al. The microtensile bond test: a review. JAdhes Dent 1999;1:299–309.

[59] Neves Ade A, Coutinho E, Cardoso MV, Jaecques S,Lambrechts P, Sloten JV, et al. Influence of notch geometryand interface on stress concentration and distribution inmicro-tensile bond strength specimens. J Dent2008;36:808–15.

[60] Neves Ade A, Coutinho E, Poitevin A, Van der Sloten J, VanMeerbeek B, Van Oosterwyck H. Influence of jointcomponent mechanical properties and adhesive layerthickness on stress distribution in micro-tensile bondstrength specimens. Dent Mater 2009;25:4–12.

[61] Sadek FT, Monticelli F, Muench A, Ferrari M, Cardoso PE. Anovel method to obtain microtensile specimensminimizing cut flaws. J Biomed Mater Res B Appl Biomater2006;78:7–14.

[62] Inoue S, Vargas MA, Abe Y, Yoshida Y, Lambrechts P,Vanherle G, et al. Microtensile bond strength of elevencontemporary adhesives to dentin. J Adhes Dent2001;3:237–45.

( 2 0 1 0 ) e100–e121

[63] Poitevin A, De Munck J, Van Landuyt K, Coutinho E,Peumans M, Lambrechts P, et al. Influence of threespecimen fixation modes on the micro-tensile bondstrength of adhesives to dentin. Dent Mater J 2007;26:694–9.

[64] Poitevin A, De Munck J, Van Landuyt K, Coutinho E,Peumans M, Lambrechts P, et al. Critical analysis of theinfluence of different parameters on the microtensilebond strength of adhesives to dentin. J Adhes Dent2008;10:7–16.

[65] Roulet J-F, Van Meerbeek B. Editorial: statistics: a nuisance,a tool, or a must? J Adhes Dent 2007;9:287–8.

[66] Eckert GJ, Platt JA. A statistical evaluation of microtensilebond strength methodology for dental adhesives. DentMater 2007;23:385–91.

[67] Sarr M, Mine A, De Munck J, Cardoso MV, Kane AW, VrevenJ, et al. Immediate bonding effectiveness of contemporarycomposite cements to dentin. Clin Oral Investig 2009;25[Epub ahead of print].

[68] Mine A, De Munck J, Cardoso MV, Van Landuyt KL, PoitevinA, Kuboki T, et al. Bonding effectiveness of twocontemporary self-etch adhesives to enamel and dentin. JDent 2009;37:872–83.

[69] Shimada Y, Yamaguchi S, Tagami J. Micro-shear bondstrength of dual-cured resin cement to glass ceramics.Dent Mater 2002;18:380–8.

[70] Shimada Y, Senawongse P, Harnirattisai C, Burrow MF,Nakaoki Y, Tagami J. Bond strength of two adhesivesystems to primary and permanent enamel. Oper Dent2002;27:403–9.

[71] Yildirim S, Tosun G, Koyutürk AE, Sener Y, Sengün A, OzerF, et al. Microtensile and microshear bond strength of anantibacterial self-etching system to primary tooth dentin.Eur J Dent 2008;2:11–7.

[72] Ruse ND, Shew R, Feduik D. In vitro fatigue testing of adental bonding system on enamel. J Biomed Mater Res1995;29:411–5.

[73] Frankenberger R, Strobel WO, Krämer N, Lohbauer U,Winterscheidt J, Winterscheidt B, et al. Evaluation of thefatigue behavior of the resin–dentin bond with the use ofdifferent methods. J Biomed Mater Res B Appl Biomater2003;67:712–21.

[74] Erickson RL, De Gee AJ, Feilzer AJ. Fatigue testing of enamelbonds with self-etch and total-etch adhesive systems. DentMater 2006;22:981–7.

[75] Erickson RL, De Gee AJ, Feilzer AJ. Effect of pre-etchingenamel on fatigue of self-etch adhesive bonds. Dent Mater2008;24:117–23.

[76] Erickson RL, Barkmeier WW, Kimmes NS. Fatigue of enamelbonds with self-etch adhesives. Dent Mater 2009;25:716–20.

[77] Barkmeier WW, Erickson RL, Latta MA. Fatigue limits ofenamel bonds with moist and dry techniques. Dent Mater2009;(August) [Epub ahead of print].

[78] De Munck J, Braem M, Wevers M, Yoshida Y, Inoue S, SuzukiK, et al. Micro-rotary fatigue of tooth–biomaterialinterfaces. Biomaterials 2005;26:1145–53.

[79] De Munck J, Van Landuyt KL, Coutinho E, Poitevin A,Peumans M, Lambrechts P, et al. Fatigue resistance ofdentin/composite interfaces with an additionalintermediate elastic layer. Eur J Oral Sci 2005;113:77–82.

[80] De Munck J, Arita A, Shirai K, Van Landuyt KL, Coutinho E,Poitevin A, et al. Microrotary fatigue resistance of aHEMA-free all-in-one adhesive bonded to dentin. J AdhesDent 2007;9:373–9.

[81] Braem M. Microshear fatigue testing of tooth/adhesive
interfaces. J Adhes Dent 2007;9:249–53.
[82] Staninec M, Kim P, Marshall GW, Ritchie RO, Marshall SJ.Fatigue of dentin–composite interfaces with four-pointbend. Dent Mater 2008;24:799–803.

6 ( 2
[83] Staninec M, Nguyen H, Kim P, Marshall GW, Rithchie RO,Marshall SJ. Four-point bending evaluation ofdentin–composite interfaces with various stresses. MedOral Patol Oral Cir Bucal 2008;1(13):E81–4.

[84] Poitevin A, De Munck J, Vivan Cardoso M, Mine A, PeumansM, Van Meerbeek B, et al. Dynamic versus static bondstrength testing: critical analysis and clinical relevance.Unpublished observations.

[85] Poitevin A, De Munck J, Vivan Cardoso M, Mine A, PeumansM, Van Meerbeek B. Comparison of dynamic versus staticbond strength testing. In: Academy of Dental Materials2009 Abstract. 2009.

[86] Inoue S, Koshiro K, Yoshida Y, De Munck J, Nagakane K,Suzuki K, et al. Hydrolytic stability of self-etch adhesivesbonded to dentin. J Dent Res 2005;84:1160–4.

[87] De Munck J, Van Landuyt K, Coutinho E, Poitevin A,Peumans M, Lambrechts P, et al. Micro-tensile bondstrength of adhesives bonded to Class-I cavity-bottomdentin after thermo-cycling. Dent Mater 2005;21:999–1007.

[88] Shono Y, Terashita M, Shimada J, Kozono Y, Carvalho RM,Russell CM, et al. Durability of resin–dentin bonds. J AdhesDent 1999;1:211–8.

[89] Tay FR, Pashley DH. Water treeing – a potential mechanismfor degradation of dentin adhesives. Am J Dent2003;16:6–12.

[90] De Munck J, Shirai K, Yoshida Y, Inoue S, Van Landuyt K,Lambrechts P, et al. Effect of water storage on the bondingeffectiveness of 6 adhesives to Class I cavity dentin. OperDent 2006;31:456–65.

[91] De Munck J, Van den Steen PE, Mine A, Van Landuyt KL,Poitevin A, Opdenakker G, et al. Inhibition of enzymaticdegradation of adhesive–dentin interfaces. J Dent Res2009;88:1101–6.

[92] Scherrer SS, Cesar PF, Swain MV. Direct comparison of thebond strength results of the different test methods: acritical literature review. In: Academy of Dental Materials2009 Review paper. 2009.

[93] Betamar N, Cardew G, Van Noort R. The effect of variationsin hourglass specimen design on microtensile bondstrength to dentin. J Adhes Dent 2007;9:427–36.

[94] Betamar N, Cardew G, Van Noort R. Influence of specimendesigns on the microtensile bond strength to dentin. JAdhes Dent 2007;9:159–68.

[95] Söderholm K-J. Letter to the Editors. J Adhes Dent2008;10:163–4.

[96] Tam LE, Pilliar RM. Fracture toughness ofdentin/resin–composite adhesive interfaces. J Dent Res1993;72:953–9.

[97] Lin CP, Douglas WH. Failure mechanisms at the humandentin–resin interface: a fracture mechanics approach. JBiomech 1994;27:1037–47.

[98] Toparli M, Aksoy T. Fracture toughness determination ofcomposite resin and dentin/composite resin adhesiveinterfaces by laboratory testing and finite element models.Dent Mater 1998;14:287–93.

[99] Armstrong SR, Boyer DB, Keller JC, Park JB. Effect of hybridlayer on fracture toughness of adhesively bondeddentin–resin composite joint. Dent Mater 1998;14:91–8.

[100] Tantbirojn D, Cheng YS, Versluis A, Hodges JS, Douglas WH.Nominal shear or fracture mechanics in the assessment ofcomposite–dentin adhesion? J Dent Res 2000;79:41–8.

[101] Tam LE, Khoshand S, Pilliar RM. Fracture resistance ofdentin–composite interfaces using different adhesive resinlayers. J Dent 2001;29:217–25.

[102] Far C, Ruse ND. Effect of bleaching on fracture toughness ofcomposite–dentin bonds. J Adhes Dent 2003;5:175–82.

[103] Setien VJ, Armstrong SR, Wefel JS. Interfacial fracturetoughness between resin-modified glass ionomer and

0 1 0 ) e100–e121 e119

dentin using three different surface treatments. DentMater 2005;21:498–504.

[104] Raskin A, D’Hoore W, Gonthier S, Degrange M, Déjou J.Reliability of in vitro microleakage tests: a literature review.J Adhes Dent 2001;3:295–308.

[105] de Almeida JB, Platt JA, Oshida Y, Moore BK, Cochran MA,Eckert GJ. Three different methods to evaluatemicroleakage of packable composites in Class IIrestorations. Oper Dent 2003;28:453–60.

[106] Heintze SD. Systematic reviews. I. The correlation betweenlaboratory tests on marginal quality and bond strength. II.The correlation between marginal quality and clinicaloutcome. J Adhes Dent 2007;9:77–106.

[107] Amarante de Camargo DA, Sinhoreti MAC, Correr-SobrinhoL, de Sousa Neto MD, Consani S. Influence of themethodology and evaluation criteria on determiningmarginal leakage in dentin–restorative interfaces. Clin OralInvestig 2006;10:317–23.

[108] Griffiths BM, Watson TF, Sherriff M. The influence ofdentine bonding systems and their handlingcharacteristics on the morphology and micropermeabilityof the dentine adhesive interface. J Dent 1999;27:63–71.

[109] Sauro S, Mannocci F, Toledano M, Osorio R, Pashley DH,Watson TF. EDTA or H3PO4/NaOCl dentine treatments mayincrease hybrid layers’ resistance to degradation: amicrotensile bond strength andconfocal-micropermeability study. J Dent 2009;37:279–88.

[110] D’Alpino PH, Pereira JC, Svizero NR, Rueggeberg FA, PashleyDH. Use of fluorescent compounds in assessing bondedresin-based restorations: a literature review. J Dent2006;34:623–34.

[111] Roulet JF, Reich T, Blunck U, Noack M. Quantitative marginanalysis in the scanning electron microscope. ScanningMicrosc 1989;3:147–58 [discussion 158–9].

[112] Blunck U, Zaslansky P. Effectiveness of all-in-one adhesivesystems tested by thermocycling following short andlong-term water storage. J Adhes Dent 2007;9(Suppl.2):231–40.

[113] Hansen EK. Effect of Gluma in acid-etched dentin cavities.Scand J Dent Res 1987;95:181–4.

[114] Heintze SD, Forjanic M, Jakob G. Automatic gap detection atrestoration margins with an optical sensor in vitro. J AdhesDent 2005;7:95–105.

[115] Heintze SD, Forjanic M, Roulet FJ. Automated marginanalysis of contemporary adhesive systems in vitro:evaluation of discriminatory variables. J Adhes Dent2007;9:359–69.

[116] Deveaux E, Hildelbert P, Neut C, Boniface B, Romond C.Bacterial microleakage of Cavit, IRM, and TERM. Oral SurgOral Med Oral Pathol 1992;74:634–43.

[117] Torabinejad M, Rastegar AF, Kettering JD, Pitt Ford TR.Bacterial leakage of mineral trioxide aggregate as aroot-end filling material. J Endod 1995;21:109–12.

[118] Bae KS, Baumgartner JC, Nakata TT. Development of ananaerobic bacterial leakage model. J Endod 1998;24:233–5.

[119] Matharu S, Spratt DA, Pratten J, Ng YL, Mordan N, WilsonM, et al. A new in vitro model for the study of microbialmicroleakage around dental restorations: a preliminaryqualitative evaluation. Int Endod J 2001;34:547–53.

[120] Reeder Jr OW, Walton RE, Livingston MJ, Pashley DH. Dentinpermeability: determinants of hydraulic conductance. JDent Res 1978;57:187–93.

[121] Pashley DH, Thompson SM, Stewart FP. Dentinpermeability: effects of temperature on hydraulic
conductance. J Dent Res 1983;62:956–9.
[122] Pashley DH, Andringa HJ, Derkson GD, Derkson ME,Kalathoor SR. Regional variability in the permeability ofhuman dentine. Arch Oral Biol 1987;32:519–23.

s 2 6
[123] Ciucchi B, Bouillaguet S, Holz J, Pashley D. Dentinal fluiddynamics in human teeth, in vivo. J Endod 1995;21:191–4.

[124] Bouillaguet S, Duroux B, Ciucchi B, Sano H. Ability ofadhesive systems to seal dentin surfaces: an in vitro study.J Adhes Dent 2000;2:201–8.

[125] Sano H, Takatsu T, Ciucchi B, Horner JA, Matthews WG,Pashley DH. Nanoleakage: leakage within the hybrid layer.Oper Dent 1995;20:18–25.

[126] Sano H, Yoshiyama M, Ebisu S, Burrow MF, Takatsu T,Ciucchi B, et al. Comparative SEM and TEM observations ofnanoleakage within the hybrid layer. Oper Dent1995;20:160–7.

[127] Dörfer CE, Staehle HJ, Wurst MW, Duschner H, Pioch T. Thenanoleakage phenomenon: influence of different dentinbonding agents, thermocycling and etching time. Eur J OralSci 2000;108:346–51.

[128] Li H, Burrow MF, Tyas MJ. Nanoleakage patterns of fourdentin bonding systems. Dent Mater 2000;16:48–56.

[129] Tay FR, Pashley DH, Yoshiyama M. Two modes ofnanoleakage expression in single-step adhesives. J DentRes 2002;81:472–6.

[130] Yuan Y, Shimada Y, Ichinose S, Tagami J. Qualitativeanalysis of adhesive interface nanoleakage usingFE-SEM/EDS. Dent Mater 2007;23:561–9.

[131] Van Meerbeek B. The “myth” of nanoleakage. J Adhes Dent2007;9:491–2.

[132] Okuda M, Pereira PN, Nakajima M, Tagami J. Relationshipbetween nanoleakage and long-term durability of dentinbonds. Oper Dent 2001;26:482–90.

[133] Guzmán-Armstrong S, Armstrong SR, Qian F. Relationshipbetween nanoleakage and microtensile bond strength atthe resin–dentin interface. Oper Dent 2003;28:60–6.

[134] Hashimoto M, De Munck J, Ito S, Sano H, Kaga M, Oguchi H,et al. In vitro effect of nanoleakage expression onresin–dentin bond strengths analyzed by microtensilebond test, SEM/EDX and TEM. Biomaterials 2004;25:5565–74.

[135] Ding PG, Wolff D, Pioch T, Staehle HJ, Dannewitz B.Relationship between microtensile bond strength andnanoleakage at the composite–dentin interface. Dent Mater2009;25:135–41.

[136] De Santis R, Mollica F, Prisco D, Rengo S, Ambrosio L,Nicolais L. A 3D analysis of mechanically stresseddentin–adhesive–composite interfaces using X-raymicro-CT. Biomaterials 2005;26:257–70.

[137] Mollica F, De Santis R, Ambrosio L, Nicolais L, Prisco D,Rengo S. Mechanical and leakage behaviour of thedentin–adhesive interface. J Mater Sci Mater Med2004;15:485–92.

[138] Eden E, Topaloglu-Ak A, Cuijpers V, Frencken JE. Micro-CTfor measuring marginal leakage of Class II resin compositerestorations in primary molars prepared in vivo. Am J Dent2008;21:393–7 [Erratum in: Am J Dent 2009 22:64].

[139] Jaecques S, Vivan Cardoso M, Zicari F, Coutinho E, Naert I,Van Meerbeek B. Microleakage in composite dentalrestorations: 3D imaging by micro-CT. 8th worldbiomaterials congress, 28 May–1 June 2008, Amsterdam,NL, paper 2289.00, poster P-Sat-O-752, CD-ROM publishedby Marathon International ©WBC2008; 2008.

[140] Jaecques SVN, Cardoso MV, Coutinho E, De Munck J, VanMeerbeek B. Sensitivity of micro-CT for microleakagedetection around dental composite restorations. In:PEF-IADR London. 2008 [Abstract 0289].

[141] Van Meerbeek B, Peumans M, Verschueren M, Gladys S,
Braem M, Lambrechts P, et al. Clinical status of ten dentinadhesive systems. J Dent Res 1994;73:1690–702.
[142] Van Meerbeek B, Perdigão J, Lambrechts P, Vanherle G. Theclinical performance of adhesives. J Dent 1998;26:1–20.

( 2 0 1 0 ) e100–e121

[143] Browning WD, Brackett WW, Gilpatrick RO. Two-yearclinical comparison of a microfilled and a hybridresin-based composite in non-carious Class V lesions. OperDent 2000;25:46–50.

[144] Tyas MJ, Burrow MF. Three-year clinical evaluation ofOne-Step in non-carious cervical lesions. Am J Dent2002;15:309–11.

[145] Peumans M, DeMunck J, VanLanduyt K, Lambrechts P, VanMeerbeek B. Five-year clinical effectiveness of a two-stepself-etching adhesive. J Adhes Dent 2007;9:7–10.

[146] Vanherle G, Verschueren M, Lambrechts P, Braem M.Clinical investigation of dental adhesive systems. Part I. Anin vivo study. J Prosthet Dent 1986;55:157–63.

[147] Lambrechts P, Braem M, Vanherle G. Buonocore memoriallecture. Evaluation of clinical performance for posteriorcomposite resins and dentin adhesives. Oper Dent1987;12:53–78.

[148] Vanherle G, Lambrechts P, Braem M. An evaluation ofdifferent adhesive restorations in cervical lesions. JProsthet Dent 1991;65:341–7.

[149] Van Meerbeek B, Braem M, Lambrechts P, Vanherle G.Two-year clinical evaluation of two dentine–adhesivesystems in cervical lesions. J Dent 1993;21:195–202.

[150] Van Meerbeek B, Braem M, Lambrechts P, Vanherle G.Evaluation of two dentin adhesives in cervical lesions. JProsthet Dent 1993;70:308–14 [Comment in: J Prosthet Dent1994 72:672–3].

[151] Van Meerbeek B, Peumans M, Gladys S, Braem M,Lambrechts P, Vanherle G. Three-year clinical effectivenessof four total-etch dentinal adhesive systems in cervicallesions. Quintessence Int 1996;27:775–84.

[152] Peumans M, Van Meerbeek B, Lambrechts P, Vanherle G.Two-year clinical effectiveness of a resin-modifiedglass-ionomer adhesive. Am J Dent 2003;16:363–8.

[153] Van Meerbeek B, Kanumilli PV, De Munck J, Van Landuyt K,Lambrechts P, Peumans M. A randomized, controlled trialevaluating the three-year clinical effectiveness of two etch& rinse adhesives in cervical lesions. Oper Dent2004;29:376–85.

[154] Van Meerbeek B, Kanumilli P, De Munck J, Van Landuyt K,Lambrechts P, Peumans M. A randomized controlled studyevaluating the effectiveness of a two-step self-etchadhesive with and without selective phosphoric-acidetching of enamel. Dent Mater 2005;21:375–83.

[155] Peumans M, Munck J, Van Landuyt K, Lambrechts P, VanMeerbeek B. Three-year clinical effectiveness of a two-stepself-etch adhesive in cervical lesions. Eur J Oral Sci2005;113:512–8.

[156] Peumans M, De Munck J, Van Landuyt KL, Kanumilli P,Yoshida Y, Inoue S, et al. Restoring cervical lesions withflexible composites. Dent Mater 2007;23:749–54.

[157] Peumans M, De Munck J, Van Landuyt K, Lambrechts P, VanMeerbeek B. Five-year clinical effectiveness of a two-stepself-etching adhesive. J Adhes Dent 2007;9:7–10.

[158] Van Landuyt KL, Peumans M, Fieuws S, De Munck J, CardosoMV, Ermis RB, et al. A randomized controlled clinical trial ofa HEMA-free all-in-one adhesive in non-carious cervicallesions at 1 year. J Dent 2008;36:847–55.

[159] Hickel R, Roulet JF, Bayne S, Heintze SD, Mjör IA, Peters M,et al. Recommendations for conducting controlled clinicalstudies of dental restorative materials. Science CommitteeProject 2/98 – FDI World Dental Federation study design(Part I) and criteria for evaluation (Part II) of direct andindirect restorations including onlays and partial crowns. J
Adhes Dent 2007;9(Suppl. 1):121–47 [Erratum in: J AdhesDent 2007 9:546].
[160] Van Landuyt K, Fieuws S, Van Meerbeek B. Guesteditorial. Sample size considerations for

6 ( 2

[185] Heintze SD, Chaiyasri T, Armstrong SR, Rousson V.


restoration-longevity randomized controlled trials. J AdhesDent 2008;10:247.

[161] Peumans M, De Munck J, Van Landuyt KL, Lambrechts P,Van Meerbeek B. Thirteen-year clinical effectiveness of twothree-step etch&rinse adhesives in cervical lesions.Unpublished data, KULeuven.

[162] Grippo O. Abfractions: a new classification of hard tissuelesions of teeth. J Esthet Dent 1991;3:14–9.

[163] Braem M, Lambrechts P, Vanherle G. Stress-inducedcervical lesions. J Prosthet Dent 1992;67:718–22.

[164] Heymann HO, Sturdevant JR, Bayne SC, Wilder AD, SluderTB, Brunson DW. Examining tooth flexure effects. J AmDent Ass 1992;122:41–7.

[165] Peumans M, De Munck J, Van Landuyt KL, Lambrechts P,Van Meerbeek B. Eight-year clinical effectiveness of a mildtwo-step self-etch adhesive in cervical lesions with andwithout selective phosphoric-acid etching of enamel.Unpublished data, KULeuven.

[166] Ermis RB, Van Landuyt KL, Cardoso MV, De Munck J, VanMeerbeek B, Peumans M. Clinical effectiveness of aone-step self-etch adhesive in non-carious cervical lesionsat two years. Unpublished data, Suleyman DemirelUniversity and KULeuven.

[167] Cardoso MV, Moretto SG, Van Landuyt K, Peumans M,Carvalho RCR, Russo EMA, et al. Randomized clinical trialevaluating the clinical effectiveness of two one-stepself-etch adhesives in cervical Class-V non-carious lesions.Unpublished data, University of Sao Paulo and KULeuven.

[168] Burrow NF, Tyas MJ. Retention rates of the one-stepadhesives G-Bond and Clearfil S3 Bond in Class-V lesions.Unpublished data, University of Melbourne.

[169] Sturdevant JR, Bayne SC, Heymann HO, Wilder AD,Roberson TM, Sluder TB. 2-Year clinical evaluation ofresin-modified glass ionomer filling material. Trans AcadDent Mater 1994;7(100):21.

[170] Wilder AD, Bayne SC, Heymann HO, Sturdevant JR,Roberson TM. Two-year clinical evaluation of wettingeffects on glass ionomer bonding. J Dent Res 1995;74(A):34[Abstract 180].

[171] Sturdevant JR, Bayne SC, Heymann HO, Wilder AD,Roberson TM. 3-Year clinical evaluation of acetone-basedVLC dentin bonding system. J Dent Res 1999;78(B):390[Abstract 2279].

[172] Wilder AD, Perdigão J, Heymann HO, Swift EJ, Roberson TM,Bayne SC. Six-month clinical study of an all-in-oneadhesive. J Dent Res 2001;80:65 [Abstract 234].

[173] Pereira PNR, Swift EJ, Heymann HO, Ritter AV, SturdevantJR, Wilder AD. Clinical evaluation of a self-etching adhesive

0 1 0 ) e100–e121 e121

and a flowable composite. J Dent Res 2006;85(Spec Iss A)[Abstract 352].

[174] Ritter AV, Heymann HO, Swift EJ, Sturdevant JR, Wilder AD.Clinical evaluation of an all-in-one adhesive in non-cariouscervical lesions with different degrees of dentin sclerosis.Oper Dent 2008;33:370–8.

[175] Wilder AD, Swift EJ, Heymann HO, Ritter AV, Sturdevant JR,Bayne SC. 12-Year clinical evaluation of a three-step dentinadhesive in non-carious cervical lesions. J Am Dent Assoc2009;140:526–35.

[176] Ritter AV, Swift EJ, Heymann HO, Sturdevant JR, Wilder AD.An eight-year clinical evaluation of filled and unfilledone-bottle dental adhesives. J Am Dent Assoc2009;140:28–37.

[177] Van Dijken JW, Sunnegårdh-Grönberg K, Lindberg A.Clinical long-term retention of etch-and-rinse andself-etch adhesive systems in non-carious cervical lesions.A 13 years evaluation. Dent Mater 2007;23:1101–7.

[178] Van Dijken JW, Pallesen U. Long-term dentin retention ofetch-and-rinse and self-etch adhesives and aresin-modified glass ionomer cement in non-cariouscervical lesions. Dent Mater 2008;24:915–22.

[179] Van Dijken JW. Durability of adhesiev restorative systemswithout HEMA and TEGDMA. In: CED-IADR at Munich. 2009[Abstract 363].

[180] Boghosian AA, Drummond JL, Lautenschlager E. Clinicalevaluation of a dentin adhesive system: 13 year results. In:IADR New Orleans. 2007 [Abstract 0228].

[181] Perdigão J. New developments in dental adhesion. DentClin North Am 2007;51:333–57.

[182] Peumans M, De Munck J, Lambrechts P, Van Meerbeek B.Systematic review of literature reporting on the clinicaleffectiveness of contemporary adhesives up to 2009.Unpublished data, KULeuven.

[183] Peumans M, De Munck J, Mine A, Van Meerbeek B.Adhesives for the restoration of cervical non-cariousClass-V lesions. Cochrane Database Syst Rev; inpreparation.

[184] Heintze SD, Blunck U, Göhring TN, Rousson V. Marginaladaptation in vitro and clinical outcome of Class Vrestorations. Dent Mater 2009;25:605–20.

Correlation of micro-tensile bond strength data and clinicaloutcome of Class V restorations. Unpublished data,Ivoclar-Vidavent, Chulalongkorn University Bangkok, TheUniversity of Iowa, University of Lausanne.

Documents

53156257 Relationship Between Bond Strength Tests and Clinical Outcomes