Review

Porcelain veneers: a review of the literature

M. Peumans*, B. Van Meerbeek, P. Lambrechts, G. Vanherle

Department of Operative Dentistry and Dental Materials, School of Dentistry, Oral Pathology and Maxillo-Facial Surgery, Catholic University of Leuven,Kapucijnenvoer 7, B-3000 Leuven, Belgium

Received 15 April 1999; received in revised form 25 June 1999; accepted 10 September 1999

Abstract

Objectives:Porcelain veneers are steadily increasing in popularity among today’s dental practitioners for conservative restoration ofunaesthetic anterior teeth. As with any new procedure, in vitro and in vivo investigations are required to assess the ultimate clinical efficacyof these restorations. The current literature was therefore reviewed in search for the most important parameters determining the long-termsuccess of porcelain veneers.

Data sources:Laboratory studies focusing on parameters in prediction of the clinical efficacy of porcelain veneers such as the toothpreparation for porcelain veneers, the selection and type of the adhesive system, the quality of marginal adaptation, the resistance againstmicroleakage, the periodontal response, and the aesthetic characteristics of the restorations have been reviewed. The clinical relevance ofthese parameters was then determined by reviewing the results of short and medium to long-term in vivo studies involving porcelain veneersperformed during the last 10 years.

Conclusions:The adhesive porcelain veneer complex has been proven to be a very strong complex in vitro and in vivo. An optimal bondedrestoration was achieved especially if the preparation was located completely in enamel, if correct adhesive treatment procedures werecarried out and if a suitable luting composite was selected. The maintenance of aesthetics of porcelain veneers in the medium to long termwas excellent, patient satisfaction was high and porcelain veneers had no adverse effects on gingival health inpatients with an optimal oralhygiene. Major shortcomings of the porcelain veneer system were described as a relatively large marginal discrepancy, and an insufficientwear resistance of the luting composite. Although these shortcomings had no direct impact on the clinical success of porcelain veneers in themedium term, their influence on the overall clinical performance in the long term is still unknown and therefore needs further study.q 2000Elsevier Science Ltd. All rights reserved.

Keywords: Porcelain veneers; Adhesion; Clinical effectiveness; Laboratory performance; Literature review

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1641.1. The adhesion complex: tooth/luting composite/porcelain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164

1.1.1. Tooth surface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1641.1.2. Porcelain veneer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1651.1.3. Luting composite. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1661.1.4. The adhesion complex: tooth/luting composite/porcelain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

1.2. Marginal adaptation of the porcelain veneer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1701.3. Microleakage at the tooth/luting composite/porcelain interface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1721.4. Periodontal response. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1731.5. Aesthetic characteristics of porcelain veneers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174

2. Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174

Journal of Dentistry 28 (2000) 163–177

Journalof

Dentistry

0300-5712/00/$ - see front matterq 2000 Elsevier Science Ltd. All rights reserved.PII: S0300-5712(99)00066-4

www.elsevier.com/locate/jdent

* Corresponding author. Tel.:132-16-332432; fax:132-16-332440.

1. Introduction

The patients’ demand for treatment of unaestheticanterior teeth is steadily growing. Accordingly, severaltreatment options have been proposed to restore theaesthetic appearance of the dentition. For many years,the most predictable and durable aesthetic correction ofanterior teeth has been achieved by the preparation offull crowns. However, this approach is undoubtedly mostinvasive with substantial removal of large amounts ofsound tooth substance and possible adverse effects onadjacent pulp and periodontal tissues. The great progressin bonding capability to both enamel and dentine madewith the introduction of multi-step total-etch adhesivesystems [1–3], along with the development of high-performance and more universally applicable small-particle hybrid resin composites has led to more conser-vative restorative adhesive techniques to deal withunaesthetic tooth appearance. Resin composite veneerscan be used to mask tooth discolorations and/or tocorrect unaesthetic tooth forms and/or positions.However, such restorations still suffer from a limitedlongevity, because resin composites remain susceptibleto discoloration, wear and marginal fractures, reducingthereby the aesthetic result in the long term [4,5]. Insearch for more durable aesthetics, porcelain veneershave been introduced during the last decade. Glazedporcelain veneers were proposed to be durable anteriorrestorations with superior aesthetics. The idea of porce-lain veneers is not a new one. In 1938, Dr CharlesPincus [6] described a technique in which porcelainveneers were retained by a denture adhesive during cine-matic filming. The fragile restorations had to beremoved after filming because no adhesive systemexisted at that time to permanently attach them. Simon-sen and Calamia [7] as well as Horn [8] reactivated theinterest in porcelain veneers by introducing special acid-etching procedures that substantially improved the long-term porcelain veneer retention. They demonstrated thatthe bond strength of an hydrofluoric acid-etched andsilanated veneer to the luting resin composite is routi-nely greater than the bond strength of the same lutingresin to the etched enamel surface [9]. From the momentporcelain veneers could be adhesively luted, the clinicaland laboratory techniques have continued to be refined.

As with any new procedure, in vitro studies are requiredto analyse the different aspects of this new system, whichare important for clinical functioning. And finally, in vivostudies are needed to assess the ultimate clinical efficacy ofthese restorations. Therefore, laboratory studies focusing onthe most important parameters in prediction of the clinicalefficacy of porcelain veneers have been reviewed. Theclinical relevance of these parameters was then determinedby reviewing the results of short and medium to long-term invivo studies involving porcelain veneers performed duringthe last 10 years.

1.1. The adhesion complex: tooth/luting composite/porcelain

The porcelain veneer technique includes the bonding of athin porcelain laminate to the tooth surface using adhesivetechniques and a luting composite in order to change thecolour, form and/or position of anterior teeth. The success ofthe porcelain veneer is greatly determined by the strengthand durability of the formed bond between the three differ-ent components of the bonded veneer complex, as there arethe tooth surface, the luting composite and the porcelainveneer.

1.1.1. Tooth surfaceConcepts regarding the preparation of teeth for porcelain

veneers have changed over the past few years. Althoughearly concepts suggested minimal or no tooth preparation[8,10–13], current beliefs support removal of varyingamounts of tooth structure [10,14–19].

Enamel reduction is required to improve the bondstrength of the resin composite to the tooth surface [20–22]. Doing so, the aprismatic top surface of mature unpre-pared enamel, which is known to offer only a minor reten-tion capacity, is removed. In addition, care must be taken tomaintain the preparation completely in enamel to realise anoptimal bond with the porcelain veneer [23]. Although theresults of the newest generation dentin adhesive systems arevery promising, the bond strength of porcelain bonded toenamel is still superior when compared with the bondstrength of porcelain bonded to dentin [24,25].

The vast majority of teeth receiving porcelain laminateveneers should have some enamel removal, usually approxi-mately 0.5 mm, which allows for the minimal thickness ofporcelain. Christensen [26] states that 0.75 mm is the opti-mum amount of enamel that should be removed. Accordingto Ferrari et al. [27], however, the extent and thickness ofenamel in the gingival area of anterior teeth does not permita reduction of 0.5 mm without encroaching upon thedentine. In addition, Natress et al. [28] found that in caseof freehand preparation, the proximal and cervical enamelwas reduced more than 0.5 mm in the vast majority of caseswith exposure of dentine in most teeth.

If dentine is exposed, protection is recommended for theperiod between preparation and cementation in order toprevent post-operative sensitivity and bacterial invasion[29,30]. The temporary materials (resin composite or acrylicresin) currently in use only partially seal the surface [31,32].More effectively, the exposed dentine can be protected bymeans of a primer, which is a hydrophilic reactive monomerin an organic solvent [33,34]. The use of these primers ordesensitisers after preparation seems not to deteriorate adhe-sion to dentine when the exposed dentine surface is adequatelyre-treated at the final appointment prior to the actual cementa-tion [35]. Paul and Scha¨rer [36] even proposed the applicationof the dentin bonding agent immediately after completion oftooth preparation. This new dentin bonding agent application

M. Peumans et al. / Journal of Dentistry 28 (2000) 163–177164

technique may prevent the development of bacterial leakageand dentin sensitivity during the temporary phase, and thetechnique is associated with improved bond strength in vitro.If temporary resin veneers must be placed because of aestheticand/or phonetic reasons, it is indicated to use an eugenolfree temporary cement in order to maintain the original bondstrength [37]. Alternatively temporary veneers may beconstructed in composite, held in place by a small area ofetched enamel [14,17].

Regarding the incisal preparation, three basic types ofpreparation have been described namely, the window orintra-enamel preparation, the overlapped incisal edgepreparation and the feathered incisal preparation. Severalauthors favoured the overlapped incisal preparation[16,17,26]. With this type of incisal preparation, thedental technician has more control on the aesthetic char-acteristics of the incisal part of the porcelain veneer. Inaddition, this preparation will make the restoration moreresistant to incisal fractures. Highton et al. [38]confirmed this latter statement in vitro using a two-dimensional photo-elastic stress analysis. This type ofpreparation distributed the occlusal load over a widersurface area and, consequently, reduced stress concen-tration within the veneer. On the contrary, an in vitrostudy by Hui et al. [39] and Gilde et al. [40] demon-strated that an overlap porcelain veneer design willtransmit maximum stress on the veneer and increasesthe risk of cohesive fracture. A window design preparedentirely into enamel withstood axial stress most favour-ably in this investigation. They concluded that wherestrength is an important requisite, the most conservativetype of veneer, namely the window preparation, was thedesign of choice. However, in the clinical study ofMeijering et al. [41] no relation was seen betweensurvival and incisal preparation design (no incisal over-lap versus incisal overlap) for both indirect resin

composite and porcelain veneers after 2.5 years of clin-ical functioning. Further in vivo studies have to pointout if a similar result would be noticed in the longterm.

1.1.2. Porcelain veneerVeneers are mainly fabricated from conventional low

fusing feldspathic porcelain. Two methods for fabricationof these porcelain veneers have been described: the plati-num foil technique [8,11,14,42] and the refractory die tech-nique [14]. At present, the refractory die technique ispreferred to the platinum foil technique in most laboratories[43].

By etching the inner side of the porcelain veneer withhydrofluoric acid and subsequently silanizing the etchedsurface, the bond strength of a luting composite to theetched porcelain surface has been measured to be higherthan the bond strength of a luting composite to etchedenamel and even exceeding the cohesive strength of theporcelain itself [9,22,44–47]. Etching the inner side of theporcelain veneer with hydrofluoric acid creates a retentiveetch pattern. SEM of the etched porcelain surface showed anamorphous micro-structure with numerous porosities[44,47–50] (Fig. 1). These micro-porosities increase thesurface area for bonding and lead to a micro-mechanicalinterlocking of the resin composite. Several factors likethe etching time, concentration of the etching liquid, fabri-cation method of the porcelain restoration [7,44], and typeof porcelain [51,52] determine the micro-morphology of theetch pattern and consequently the bond strength of the resincomposite to the etched porcelain.

In addition to micro-porosities, micro-cracks wereobserved that grow when the etching time increases [49].These cracks can act as sources of crack initiation andslightly, although not significantly, decrease the flexuralstrength of the etched porcelain. Weakening of the

M. Peumans et al. / Journal of Dentistry 28 (2000) 163–177 165

Fig. 1. Field-emission SEM photomicrograph of a GC Cosmotech porcelain surface etched with 9.5% hydrofluoric acid (Porcelain Etch gel, Ultradent) for 60 sand ultrasonically cleaned. The etched porcelain surface showed an amorphous micro-structure with numerous micro-porosities. (Bar� 10mm;Magnification� 10 000× ).

porcelain by etching was also noted in other in vitro studies[53,54].

Ultrasonic cleaning of etched porcelain in 95% alcohol,acetone or distilled water is indicated to remove all residualacid and dissolved debris from the surface. Inadequaterinsing after etching the porcelain surface may leave re-mineralised salts, which can be recognised as a white resi-due or deposit [55]. Some authors [50,56] studied the etchpatterns of hydrofluoric acid on feldspathic porcelain withSEM and concluded that the best surface, in terms of pene-trability, was obtained by immersion of the etched porcelainin an ultrasonic bath. Aida et al. [57], however, observed nosignificant differences in surface morphology and bondstrength between etched (5% GC Hydrofluoric Acid, GCCorp., Tokyo, Japan) feldspathic porcelain with and withoutultrasonic cleaning.

Silanization of etched porcelain with a bi-functionalcoupling agent provides a chemical link between the lutingresin composite and porcelain. A silane group at one endchemically bonds to the hydrolysed silicon dioxide at theceramic surface, and a methacrylate group at the other endcopolymerises with the adhesive resin. Single-componentsystems contain silane in alcohol or acetone and requireprior acidification of the ceramic surface with hydrofluoricacid to activate the chemical reaction. With two-componentsilane solutions, the silane is mixed with an aqueous acidsolution to hydrolyse the silane, so that it can react directlywith the ceramic surface. If not used within several hours,silane will polymerise to an unreactive polysiloxane [58].Several authors reported differences in bond strength depen-dent on the silane treatment used [45,46,51,56]. In addition,heating of the silane-coated porcelain to 1008C resulted in abond strength twice as high than if no heating was used [52].

The bond strength of resin composite to a pre-treated

ceramic restoration has been described to be negativelyinfluenced by external factors like water sorption [52], ther-mocycling [22,59], and fatigue [60]. Contamination of thepre-treated surface with die stone [61], latex gloves [62],saliva [46,59], silicone-based fit-checker paste [63,64], andtry-in paste [65] will also lower the bond strength. Severalcleaning methods were proposed to restore the original bondstrength. In case of contamination with saliva, re-etching theinner side of the porcelain with 37% phosphoric acidrestored the bond strength [46,66]. Acetone cleaning, afterremoval of the try-in paste, produced a marked reduc-tion in bond strength [61,65]. This cleaned surface hadto be silanated again to restore the original bondstrength [65]. A decreased bond strength due to contam-ination with fit-checker paste was restored by re-etchingand silanizing the porcelain surface [64], whereas Shethet al. [63] reported that the original bond strength couldnot be restored due to chemical contamination of theporcelain surface.

1.1.3. Luting compositeFor cementation of porcelain veneers a light-curing luting

composite is preferred [55]. A major advantage of light-curing is that it allows for a longer working time comparedwith dual cure or chemically curing materials. This makes iteasier for the dentist to remove excess composite prior tocuring, and greatly shorten the finishing time required forthese restorations. In addition, their colour stability is super-ior compared with the dual-cured or chemical-curedsystems. Nevertheless, it is important that there is enoughlight transmittance throughout the porcelain veneer to poly-merise the light-curing luting composite. The porcelainveneer absorbs between 40–50% of the emitted light. Thethickness of the porcelain veneer is the primary factor

M. Peumans et al. / Journal of Dentistry 28 (2000) 163–177166

Fig. 2. The interface tooth/luting composite/porcelain of a bonded to tooth porcelain veneer shows the interdigitation of the luting composite intothe etched-and-silanized porcelain surface and the etched enamel surface. (Bar� 25mm; Magnification� 625× ) Center: Magnification of the luting composite (OpalLuting Composite, 3M). At the luting composite/porcelain interface, the luting composite is diluted with unfilled resin (u) (Scotchbond Multi-Purpose, 3M).The undiluted luting composite contains densely packed rounded filler particles of different sizes. (Bar� 1 mm; Magnification� 5000× ).

M.

Pe

um

an

se

ta

l./

Jou

rna

lof

De

ntistry

28

(20

00

)1

63

–1

77

167Table 1Descriptive statistics of clinical trials involving porcelain veneers (evaluation criteria*: color (1), surface texture (2), wear (3), marginal adaptation (4), marginal discoloration (5), caries (6), fracture (7), retention(8), post-operative sensitivity (9), gingiva response (10), patient satisfaction (11))

Author Number of veneers Number of patients Porcelain/adhesive-system Type of preparation Observation-period Evaluation criteria*

Clyde and Gilmoure [42] 200 Not specified Chameleon(Terec)/Duo-cure(Terec)

Feathered incisal edgeIncisal bevelPalatal overlap

1–30 months 1,7,8,9,10

Calamia [89] 115 17 Chameleon (Terec)/Comspan1 Ultrabond(Den-Mat)

No preparationSlight incisal overlap

2–3 y 4,5,6,7,8,10(USPHS criteria)

Jordan et al. [90] 80 12 Not specified/dual cure(not specified)

Conventional(no incisal overlap)

4 y 1,2,3,4,6,8,9,10(modified Ryge criteria)

Rucker et al. [84] 44 16 Vitadur-N(Vita)/Heliolink 1 Dual cement(Vivadent)

Incisal bevel 2 y 1,4,7,8,10,11

Feathered incisal edgeChristensen and Christensen[91]

163 45 Cerinate (Den-Mat)/Ultrabond (Den-Mat)

Feathered incisal egde 3 y 1,4,5,6,7,10,11

Nordbq et al. [85] 135 41 Ceramco (Ceramco Inc.)/Porcelite LC (Kerr)

Conventional(no incisal overlap)

3 y 3,4,5,7,8

Jager et al. [86] 80 25 Mirage/MirageFLC 1 Mirage Bond (FAMirage)

Palatal overlap 1–7 y 4(SEM),5,6,7,10,11

(10 veneers.4 y)Strassler and Nathanson [83] 291 60 Cerinate (Den-Mat)/

Ultrabond (Den-Mat)No preparationConventional(no incisal overlap)

18–42 months 1,4,5,7,8(modified USPHS criteria)

Strassler and Weiner [92] 115 21 Cerinate (Den-Mat)/Ultrabond (Den-Mat)

No preparationConventional(no incisal overlap)

7–10 y 1,4,5,6,7,8(modified USPHS criteria)

Walls [93] 54 12 Fiber reinforced porcelain(not specified)/Heliolink(Vivadent)1 Gluma (Bayer)

Special preparation forworn teeth

5 y 5,7,8,10

Meijering [41] 56 Not specified Flexo-ceram (ElephantCeramics)/not specified

Conventional(no incisal overlap)

2.5 y 1,4,5,6,7,8,9,10,11(modified USPHS criteria)

Palatal overlapPeumans et al. [87] 87 25 GC Cosmotech Porcelain/GC

Cosmotech Bonding Set(GC)1 Scotchbond 2 (3M)

Palatal overlap 5–6 y 1,2,4,5,6,7,8,10,11

Kihn et al. [88] 59 12 Ceramco Colorlogic/Ceramco Colorlogic BondingSystem

Conventional (noincisal overlap)Palatal overlap

48 months 1,4,5,6,9,10

M.

Pe

um

an

se

ta

l./

Jou

rna

lof

De

ntistry

28

(20

00

)1

63

–1

77

168

Table 2Results of clinical trials involving porcelain veneers

Author Retention1 Fracture rate (%) Excellent margins (%) Microleakage (%) Caries (%) Periodontal health Maintenance of aesthetics (%) Patient satisfaction

Clyde and Gilmoure [42] 1 – – – Excellent (not specified) – –Calamia [89] 3 93 17 1 Acceptable 100 –Jordan et al. [90] 3 83 – 0 Excellent 100 –Rucker et al. [84] 0 100 – – Excellent (not specified) 100 100% excellentChristensen and Christensen[91]

13 65 8 1 Slight gingival irritation 100 87% excellent

Nordbq et al. [85] 5 Not specified Negligible – – 100 –Jager et al. [86] 1 55 1 1 Excellent – 84% excellentStrassler and Nathanson [83] 1.7 96 0 0 – 100 –Strassler and Weiner [92] 7 88 14 1 – 100 –Walls [93] 14 Not specified 28 – Excellent – –Meijering et al. [41] Not specified 98 22 0 Excellent 100 93% excellentPeumans et al. [87] 1 14 25 2 Slight gingival irritation 100 80% excellentKihn et al. [88] 0 85 2 0–1 No gingival response 100 /

determining the light transmittance available for polymer-isation [67–73]. The colour and the opacity of the porcelainwould have less influence on the amount of absorbed light[69,72,73]. Linden et al. [72] reported that the opacity ofporcelain became more important for facings with a thick-ness of 0.7 mm or more. Consequently, the presence of aporcelain veneer increases the setting time of the resincomposite used beneath the veneer [69–72]. O’Keefe et al.[73] argued that it was necessary to double the recom-mended exposure time.

In case of porcelain with a thickness of more than 0.7 mm[72], light-cured resin composites do not reach their maxi-mum hardness. A dual-cured luting composite, whichcontains the initiation systems for both chemically andlight-cured composites, is advisable in these situations.With these latter luting agents a stronger bond can beobtained with the porcelain [68]. Also higher values of hard-ness were reported for the dual-cure resin cements than forthe light-cured luting composites because of their higherdegree of polymerisation [72,74].

Regarding the classification of luting composites, Albers[55] ranked these luting composites following the classifi-cation of resin composites for restorative purposes proposedby Lutz and Phillips [75]. According to Inokoshi et al. [76],it is difficult to classify the luting composites followingthese criteria because their high content of small particlesrequires that many of them would be classified as hybridresins or heavily filled resins. A classification of 14 dual-cure luting composites was made according to their maxi-mum filler size. This maximum filler size varied extremelyfrom less than 1 to 250mm. The predominant filler size forall products was much smaller than the maximum filler size.The filler weight varied from 36 to 77%. No correlation wasobserved between filler loading, maximum filler size andconsistency of these dual-cured luting composites, whereasa strong correlation was found between the consistency andthe film thickness of the luting agents. This relation wassupported by the temperature dependence of the film thick-ness reported for these dual-cure luting composites [77].The great diversity in the currently available productsmakes clear specification for luting composites urgentlyneeded.

1.1.4. The adhesion complex: tooth/luting composite/porcelain

The adhesion complex porcelain/luting composite/toothwas studied by Stacey [22]. He reported that a very strongcomplex was obtained in vitro by luting the porcelainveneer. The strength of the combined porcelain/lutingcomposite/enamel bond (63 MPa) was significantly higherthan the separate composite/etched enamel (31 MPa) andluting composite/etched-and-silanized porcelain (33 MPa)bond strengths. This assumes that a close apposition of theenamel and the porcelain surface synergistically influencesthe bond strength of the luting composite/enamel and lutingcomposite/porcelain bonds. In addition, Andreasen et al.

[78] and Stokes and Hood [79] noted that extracted incisorsrestored with porcelain veneers were recovered to theiroriginal strength. A laboratory investigation analysingultra-morphologically the adhesive interface of porcelainveneers bonded to tooth structure confirmed the high reten-tion of bonded porcelain veneers [50]: FE-SEM photo-graphs showed a strong micro-mechanical interlocking ofthe luting composite in the micro-retentive pits of the acid-etched tooth surface at one side and in the etch pits of theacid-etched porcelain surface at the other side (Fig. 2).

Magne and Douglas [80] also demonstrated that porcelainveneers restore the mechanical behaviour and microstruc-ture of the intact tooth in vitro even when they are bonded toan extensive dentin surface using an optimised applicationmode of dentine adhesives. Nevertheless, significant cyclictemperature changes can induce the development of flaws inporcelain veneers. A sufficient and even thickness of cera-mic combined with a minimal thickness of luting compositewill provide the restoration with a favourable configurationwith regard to crack propensity, namely, a ceramic andluting composite thickness ratio above 3 [81]. This ratioalso appears to have a relevant influence on the stress distri-bution in porcelain laminates. Restorations that are too thin,combined with poor internal fit, resulted in higher stresses atboth the surface and interface of the restoration [82].

Ultimately, the clinical relevance of these in vitro resultsmust be determined in vivo. A review of clinical trials invol-ving porcelain veneers, that were conducted during the last10 years, is summarised in Table 1 with their respectivedescriptive statistics (Table 1) and results (Table 2). Thesein vivo results confirmed the strong bond between a porce-lain veneer and the underlying tooth tissue as most short andmedium-term clinical studies reported a very low failurerate (0–5%) due to loss of bonding and fracture [42,83–88]. Somewhat higher failure rates were noted by Christen-sen and Christensen [91], 13% after three years and byStrassler and Weiner [92], 7% after 7–10 years. Walls[93] also observed a high number of fractures and/or lossof porcelain veneers (14%) after five years of clinical func-tioning. Unfavourable occlusion and articulation seemed tohave been a determining factor for the high failure rate ofporcelain veneers as the latter restorations were placedfor aesthetic and functional reconstruction of fracturedand worn anterior teeth in patients with a history ofbruxism. The large exposed dentine surfaces to bebonded in these fractured and worn teeth (using athird generation dentin adhesive) may also have contrib-uted to the high failure rate.

Some authors [94,95] reported higher failure rates in vivowhen porcelain veneers were partly bonded to underlyingcomposite restorations. This bonding is based on delayedresin-to-resin bonding, which may decrease the bondstrength of the porcelain veneer–tooth complex.

As a conclusion, in vitro and in vivo studies indicated thatporcelain veneers are strong and durable restorations in themedium to long term when enough intact tooth tissue is left

M. Peumans et al. / Journal of Dentistry 28 (2000) 163–177 169

to bond the porcelain veneer and when occlusion and articu-lation are not pathological.

1.2. Marginal adaptation of the porcelain veneer

For any cemented restoration the weak link is at therestoration-cement–tooth interface. In porcelain veneersthe composite luting agent is the weak link in the system.When used as a luting agent, the bulk of composite resin isgreatly reduced. However, there is still polymerisationvolumetric shrinkage in the amount of 2.6–5.7% [96],which may create a marginal opening or loss of themarginal seal. The thermal expansion coefficient (TEC) isalso different from the tooth tissues and the porcelain.Finally, composite resins may wear and the wear will begreater in case of larger gap widths compared to smallerones [97]. In addition, in vitro studies have demonstrateda dissolution of the resin matrix of composite resin in oralfluids [98–100]. Therefore it is desirable to minimise thecomposite component and maximise the porcelain compo-nent by having as close an adaptation of the porcelainveneer as possible.

Sorensen et al. [101] compared in vitro the marginal fit ofporcelain veneers fabricated by the two proceduresdescribed above: the platinum foil technique and the refrac-tory die technique. They reported that the mean verticalmarginal discrepancy (for all positions combined) for plati-num foil veneers (187mm) was significantly less than thatfor veneers made with the refractory die technique(242mm). The same observation was done by Sim andIbbetson [102] (60 versus 290mm) and Wall et al. [103](74 versus 132mm), although they reported smallermarginal gap widths. This finding seems however incongru-ous, considering that the platinum foil occupies 25mm. Limand Ironside [104] reported that divesting with aluminiumoxide abrasive may account for inadvertent abrasion of the

delicate inner porcelain surface and causes larger marginaldiscrepancies (114mm with sandblasting versus 97mmwithout sandblasting).

With the refractory die technique, the importance ofmatching the thermal expansion coefficient between therefractory die and the ceramic must be emphasised inorder to fabricate restorations with a significantly improvedmarginal accuracy [105]. Intermixing of individual ceramicand refractory products seems to be commonplace in manylaboratories with some combinations having differentthermal expansion coefficients [43].

For both fabrication techniques, the marginal openings atthe gingivo-proximal corners were two to four times largerthan at the mid-labial position [101,102,106]. This is prob-ably the result of the shrinkage of porcelain towards theregion of greatest bulk (the centre) and the geometry ofthe margins. Clinically this poorer fit at the gingivo-proximal corners of the veneers would be furthercompounded by the difficulty in access for finishing of theluted veneers in these regions.

When luting porcelain veneers it is important that themarginal discrepancies are completely filled with the lutingcomposite, in order to polish the cement layer to a smoothmargin. Harasani et al. [106] observed after finishing of theveneers still a considerable amount of excess luting agent atthe veneer margins. Also Coyne and Wilson [107] reportedthat only a small proportion of the margins of each porcelainveneer was found to have an ideal marginal adaptationmicroscopically. Hannig et al. [108] noted that especiallythe cervical region seems to be a problematic area to achieveperfect marginal adaptation.

Tay et al. [109] advised to remove the excess of non-polymerised composite cement with a brush moistenedwith bonding resin. This will reduce the dragging outtendency of the resin out of the marginal gap and ensure asmoother margin that is polishable.

M. Peumans et al. / Journal of Dentistry 28 (2000) 163–177170

Fig. 3. SEM photomicrograph of the cervical margin of a 5-year old porcelain veneer (P) showing a small marginal defect and a border of roughened porcelain(arrows). (G, gingiva; C, luting composite) (Bar� 250mm; Magnification� 52× ).

Finishing of the veneer margins corrects the inherentmarginal defects but results in removal of the glaze fromthe porcelain [87] (Fig. 3). This will cause increased plaqueretention and gingival reaction on the one hand and wear ofthe antagonistic elements on the other hand, unless theporcelain can be polished to a smooth surface.

Some authors showed that polishing procedures canproduce a polished surface, which is equal to a glazedporcelain surface [110–112]. While these polishing instru-ments perform satisfactorily on flat accessible surfaces athigh speeds, none of them are well suited for finishingcrucial gingival or interproximal regions of a bondedveneer. Haywood et al. [113] evaluated finishing andpolishing in these crucial areas in vitro. According to

these authors, a finish equal or superior in smoothness toglazed porcelain was achieved through the use of a series offinishing grit diamonds (Micron Finishing System) followedby a 30-fluted carbide bur and diamond polishing paste.Other finishing combinations produced surface textures,which were not as smooth as glazed porcelain. Polishingunder waterspray produced also a smoother surface for agiven diamond sequence than did dry polishing [114].However, the effect of all these finishing procedures atthe cervical margins of the veneer with their difficultaccessibility for polishing instruments, must be evaluatedin vivo.

Regarding the marginal adaptation of porcelain veneersafter several years of clinical functioning, most in vivo

M. Peumans et al. / Journal of Dentistry 28 (2000) 163–177 171

Fig. 4. Field-emission SEM photomicrograph of the luting composite/etched enamel interface of a bonded to tooth porcelain veneer. At the most cervicalenamel area, the luting composite is bonded to aprismatic enamel. Atypically shaped prisms are present. (C, Luting composite; E, Enamel, u, unfilled resin)(Bar� 2mm; Magnification� 5000× ).

Fig. 5. Field-emission SEM photomicrograph of the luting composite/etched enamel interface of a bonded to tooth porcelain veneer at the area of prismaticenamel. The enamel prismata are longitudinally sectioned. The perifery of the enamel prisms is preferentially dissolved forming macro-tags with a depth ofapproximately 5mm (large arrows). Micro-tags are present between the hydroxyapatite crystals (small arrows). The macro-tags and micro-tags contribute to anenormous increase in surface area for bonding. (C, Luting composite; E, Enamel) (Bar� 2mm; Magnification� 5000× ).

studies reported a relatively high number of restorationswith an excellent marginal adaptation (65–98%)[41,83,84,88–92], while only few clinical studies reportedthat small marginal defects were frequently observed alongthe entire outline of the porcelain veneer after five years offunctioning [86,87] (Table 2). SEM examination showedthat these small marginal defects were due to the wearingout of the composite luting agent and loss of bonding[86,87]. A similar phenomenon was reported as submargina-tion in several in vivo studies of porcelain inlays [115–117].

Further research should be directed towards improve-ment of marginal adaptation of the porcelain veneers.Attention must be given to the laboratory procedure toreduce the marginal gap width and to the finishingprocedure to diminish the excess of resin composite at themargins. Finally more wear-resistant luting compositesshould be developed.

1.3. Microleakage at the tooth/luting composite/porcelaininterface

The polymerisation shrinkage of the luting composite andthe difference in thermal expansion coefficient between theluting composite and both the tooth and the porcelain veneercauses stress at the tooth/luting composite/porcelain inter-face. According to the theories by Feilzer et al. [118], thereis only a limited potential for relaxation of polymerisationstress due to flow because the luting composite is bonded atall sides in a narrow “gap” with only a limited area ofunbonded surface. Consequently, Feilzer et al. [119] foundthat when the thickness of the resin composite is thinneddown, as in the case of a luting agent, the wall-to-wallpolymerisation shrinkage might be three times the normallinear contraction of bulk resin composite. Residual poly-merisation stress could then be expected to be even greater

M. Peumans et al. / Journal of Dentistry 28 (2000) 163–177172

Fig. 6. Field-emission SEM photomicrograph of the luting composite/etched dentine interface at the cervical third of a bonded to tooth porcelain veneer. Thedentine tubules (arrows) are longitudinally sectioned. The dark area represents the hybrid layer and the layer of unfilled adhesive resin (H). No lossof bondingis observed in this area using a modern total etch dentine adhesive system (Scotchbond Multi-Purpose, 3M). (C, Luting composite; D, dentine) (Bar� 10mm;Magnification� 1000× ).

Fig. 7. The oblique sectioned dentine tubules at the luting composite/etched dentine interface are filled with resin tags (arrow). No loss of bonding is observedin this area using a modern total etch dentine adhesive system (Scotchbond Multi-Purpose, 3M). (C, Luting composite; D, dentine; H, hybrid layer and layer ofunfilled adhesive resin) (Bar� 5mm; Magnification� 2000× ).

than that reported for resin composite restorations. Due tothis contraction stress there exists a competition between theadhesive forces of the two bonded interfaces: the porcelain/luting composite interface and the luting composite/toothinterface. The interface with the lowest adhesive forceswill fail namely the luting composite/tooth interface. Micro-leakage will occur at this interface and may then lead tostaining, post-operative sensitivity and recurrent caries.

Microleakage at the luting composite/porcelain interfacewas negligible [101,120,121]. At the interface lutingcomposite/tooth the microleakage was minimal when thepreparation margins were located completely in enamel[101,120–123]. Only at the cervical margin, microleakagewas more pronounced. This would be due to the presence ofaprismatic enamel in the cervical region [50,124] (Figs. 4and 5). Porcelain veneer preparations generally end in thisregion of aprismatic enamel, which therefore may be largelyresponsible for the poor marginal sealing reported at thecervical margins of porcelain veneers in vitro [101,120–123]. When the cervical preparation margin was located indentine, significantly greater microleakage was recorded atthe luting composite/tooth interface [125]. The use of a thirdgeneration dentine bonding agent can reduce the occurrenceof microleakage according to several authors [120,125] incontrast with the results obtained by Sim et al. [123],who reported that none of the tested (third generation)dentine bonding agents significantly reduced microleak-age at the dentine margins. Dietschi et al. [126]reported an obvious reduction in marginal leakage ofceramic inlays when bonded to dentin using twomodern dentine adhesives, although the integrity ofthe adhesive interface did not yet appear optimal indentine bonded ceramic restorations because bondingfailures occurred mainly between the hybrid layer andthe overlaying resin. In another in vitro study debondingwas never noticed along the luting composite/dentininterface when a modern multi-step total-etch dentinadhesive system was used [50] (Figs. 6 and 7). Magneand Douglas [80] noticed in vitro that the method ofdentin adhesive application during placement of theporcelain veneer had an influence on the quality ofthe luting composite dentin/interface. Their scanningelectron microscope observations demonstrated that thetraditional application of a multi-step total etch dentinadhesive (dentin adhesive applied when proceeding toluting the veneer) was associated with bonding failuresbetween the hybrid layer and the overlaying resin,whereas unbroken and continuous interfaces wereobtained with a new method using the same dentinadhesive (dentin adhesive applied to dentin and curedbefore taking the impression of the veneer). However,none of the modern adhesive systems appears yet to beable to guarantee hermetically sealed restorations withmargins free of discoloration for a long time [25].

Another effective method to reduce microleakage is bypost-finishing sealing the margins of the veneer with a

bonding resin [125]. Further research is necessary to evalu-ate the retention of the sealer.

In addition to the location of the preparation margins, thetype of luting composite determines the occurrence ofmicroleakage as the thermal expansion coefficient and theamount of polymerisation shrinkage vary among the type ofresin composite. A high filler loading reduces the thermalexpansion coefficient and polymerisation shrinkage.Because of these reasons, a luting composite with an opti-mal filler loading is preferred to lute the porcelain veneers[101,121,122,123]. However, the viscosity of such cementsis high and hence the positioning of the veneer during theluting procedure may be delicate. The luting procedure withthese highly filled luting composites can be simplified byusing an ultrasonic insertion technique [127]. Clinicalconfirmation of this method is however required.

Light-cured and dual-cured luting composites show asimilar leakage pattern at the luting composite/tooth inter-face according to Zaimoglu et al. [122], whereas Jankowskiet al. [128] observed less microleakage when using dual-cured luting composites.

In clinical studies, microleakage was more frequentlyobserved when dentine was exposed during preparationfor porcelain veneers [89], even when a third generationdentine adhesive system was used [87,93]. In other clinicalstudies of porcelain veneers with complete intra-enamelpreparations, microleakage was reported less frequently[83,85,88,91,92]. Finally, microleakage was rarely asso-ciated with the presence of caries in vivo [41,83,86–92].

As a conclusion, microleakage can be minimised bylocating the preparation margins of the veneer in enameland by selecting a highly filled luting composite.

1.4. Periodontal response

Porcelain is considered as the most aesthetic and biocom-patible material in dentistry with the ability to imitate soundenamel. Several studies showed that porcelain retainsplaque less than other restorative materials or enamel[129–132], that the plaque is removed more rapidly fromporcelain surfaces and/or that the bacterial plaque vitality onthese surfaces was smaller [133]. The low surface roughnessof the glazed porcelain may to a large part be responsible forthis phenomenon [134]. Based on these observations onewould expect no or even a positive reaction of the marginalgingival tissues towards porcelain veneers.

Several short-term [89,135] and medium-term[88,93,136] clinical studies of porcelain veneers confirmedthis expectation. They reported no increase of even adecrease in plaque accumulation on these restorations.Kourkata et al. [135] even described a significant lowerbacterial plaque vitality immediately after placement ofthe porcelain veneers. Peumans, however, observed a slightincrease in plaque retention at the cervical margins of 5-yearold porcelain veneers [137]. This slight increase wasexplained by the increased surface roughness at the cervical

M. Peumans et al. / Journal of Dentistry 28 (2000) 163–177 173

border of the restoration caused by removal of the glaze ofthe porcelain during finishing with microfine finishingdiamonds (Fig. 3). In addition to this roughened cervicalborder, the presence of small cervical marginal defectscan lead to an increased plaque retention [87].

Concerning the gingival response, most clinical studiesobserved no change in gingival health at the restored teeth[41,42,84,86,88–90,93,136] (Table 2). Only a few clinicalstudies reported a slight gingival inflammation at therestored teeth [91], especially in patients with a moderateor bad oral hygiene [137].

According to Pippin et al. [136], the location of the cervi-cal extension of the restoration in location to the gingivalmargin also plays an important role in the reaction of thegingival tissues. They reported that the gingival reactionincreased as the extension was located closer to or belowthe gingival margin, however, the gingival reaction forporcelain veneers was at the same location lower than forthe metal ceramic restorations.

In conclusion, no or a minimal periodontal response canbe expected at teeth restored with porcelain veneers. Anoptimal oral hygiene of the patient and smoothly finishedmargins are important factors to maintain an optimal period-ontal health around the restored teeth.

1.5. Aesthetic characteristics of porcelain veneers

There is a general agreement among the practitioners thatporcelain veneers will continue to play a vital role in elec-tive dental aesthetics. This places high demands on predict-ability, especially with colour matching and maskingmethods. The final shade of the veneers depends not onlyon the colour, opacity and thickness of the porcelain but alsoon the colour of the underlying tooth and the colour andthickness of the luting composite [138–143].

Colour matching of one discoloured tooth with a porce-lain veneer to the surrounding natural teeth must be consid-ered as most difficult. It is impossible to mask a strongdiscoloration by a thin layer of porcelain (0.3–0.7 mm)without making the restoration opaque and lifeless. Conse-quently, the restored tooth will never have the same trans-lucency as the surrounding natural teeth. [14,15,42,55,87].

Regarding the aesthetics (colour stability and surfacesmoothness) of porcelain veneers after several years of clin-ical functioning, all clinical studies confirmed the mainte-nance of aesthetics of porcelain veneers in the short termand in the medium to long term (Table 2). In addition,patient acceptance of porcelain veneers in these clinicalstudies was high. The percentage of patients that werecompletely satisfied with the porcelain veneers variedfrom 80 to 100% (Table 2). Some studies even reportedan increase in patient satisfaction after several years[91,144]. This increase was explained by the habituationof the patients to the aesthetic improvement of their denti-tion with porcelain veneers.

2. Conclusion

The adhesive porcelain veneer complex appeared to be avery strong complex in vitro and in vivo. An optimal bondwas obtained if the preparation was located completely inenamel, if correct surface treatment procedures were carriedout and if a suitable composite luting agent was selected.However, from an aesthetic and periodontal point of view acomplete intra-enamel preparation cannot always berealised. The quality of the restoration was inferior ifdentine was exposed to a large extent, as the current dentinbonding agents are not yet able to prevent microleakage atthe dentin margins in the long term.

The periodontal response to porcelain veneers variedfrom clinically acceptable to excellent. Regarding theaesthetic properties of the porcelain veneers, these restora-tions maintained their aesthetic characteristics in themedium to long term and patient satisfaction was high.

The major shortcoming of porcelain veneers was the rela-tively wide marginal discrepancy. At these marginal open-ings the luting composite was exposed to the oralenvironment and the wear resistance of the composite lutingagents was not yet optimal.

Nevertheless, these shortcomings had no direct impact onthe success of porcelain veneers in the medium term,however, their influence on the overall clinical performancein the long term is still unknown.

References

[1] Nakabayashi N, Kojima K, Masuhara E. The promotion of adhesionby infiltration of monomers into tooth substrates. Journal of Biome-dical Materials Research 1982;16:265–73.

[2] Van Meerbeek B, Vanherle G, Lambrechts P, et al. Dentin- andenamel-bonding agents. Current Opinion in Dentistry 1992;2:117–27.

[3] Pashley DH, Ciucchi B, Sano H, et al. Permeability of dentin toadhesive agents. Quintessence International 1993;24:618–31.

[4] Peumans M, Van Meerbeek B, Lambrechts P, et al. The five-yearclinical performance of direct composite additions to correct toothform and position. Part I: aesthetic qualities. Clinical Oral Investiga-tions 1997;1:12–18.

[5] Peumans M, Van Meerbeek B, Lambrechts P, et al. The five-yearclinical performance of direct composite additions to correct toothform and position. Part II: marginal qualities. Clinical Oral Investi-gations 1997;1:19–26.

[6] Pincus CR. Building mouth personality. Journal of South CaliforniaDental Association 1938;14:125–9.

[7] Simonsen RJ, Calamia JR. Tensile bond strength of etched porce-lain. Journal of Dental Research 1983;62:297 Abstract 1154.

[8] Horn RH. Porcelain laminate veneers bonded to etched enamel.Dental Clinics of North America 1983;27:671–84.

[9] Calamia JR, Simonsen RJ. Effect of coupling agents on bondstrength of etched porcelain. Journal of Dental Research1984;63:179 Abstract 79.

[10] Calamia JR. Etched porcelain veneers: the current state of the art.Quintessence International 1985;1:5–12.

[11] Plant CG, Thomas GD. Porcelain facings: a simple clinical andlaboratory method. British Dental Journal 1987;163:231–4.

M. Peumans et al. / Journal of Dentistry 28 (2000) 163–177174

[12] McLaughlin G, Morrison JE. Porcelain fused to tooth: the state of theart. Restorative Dentistry 1988;4:90–4.

[13] Reid JS, Murray MC, Power SM. Porcelain veneers—a four yearfollow-up. Restorative Dentistry 1988;5:42–55.

[14] Garber DA, Goldstein RE, Feinman RA. Porcelain laminate veneers,Lombourg: Quintessence Publishing Co, 1987.

[15] McComb D. Porcelain veneer technique: a promising new methodfor restoring strength and esthetics to damaged or discolored teeth.Ontario Dentist 1988;65:25–32.

[16] Weinberg LA. Tooth preparation for porcelain laminates. New YorkState Dental Journal 1989;5:25–8.

[17] Nixon RL. Porcelain veneers: an esthetic therapeutic alternative. In:Rufenacht CR, editor. Fundamentals of esthetics, Lombourg: Quin-tessence Publishing Co, 1990. p. 329–68.

[18] Garber DA. Porcelain laminate veneers: to prepare or not toprepare?. Compendium of Continuing Education in Dentistry1991;XII:178–82.

[19] Friedman MJ. Augmenting restorative dentistry with porcelainveneers. Journal of the American Dental Association 1991;122:29–34.

[20] Schneider PM, Messer LB, Douglas WH. The effect of enamelsurface reduction in vitro on the bonding of composite resin topermanent human enamel. Journal of Dental Research1981;60:895–900.

[21] Black JB. Esthetic restoration of tetracycline-stained teeth. Journalof the American Dental Association 1982;104:846–51.

[22] Stacey GD. A shear stress analysis of the bonding of porcelainveneers to enamel. The Journal of Prosthetic Dentistry1993;70:395–402.

[23] Troedson M, De´rand T. Shear stresses in the adhesive layer underporcelain veneers. A finite element method study. Acta Odontolo-gica Scandinavica 1998;56:257–62.

[24] Van Meerbeek B, Peumans M, Gladys S, et al. Three-year clinicaleffectiveness of four total-etch dentinal adhesive systems in cervicallesions. Quintessence International 1996;27:775–84.

[25] Van Meerbeek B, Perdigao J, Lambrechts P, et al. The clinicalperformance of adhesives. Journal of Dentistry 1998;26:1–20.

[26] Christensen GJ. Have porcelain veneers arrived? Journal of theAmerican Dental Association 1991;122:81.

[27] Ferrari M, Patroni S, Balleri P. Measurement of enamel thickness inrelation to reduction for etched laminate veneers. The InternationalJournal of Periodontics and Restorative Dentistry 1992;23:407–13.

[28] Nattress BR, Youngson CC, Patterson JW, et al. An in vitro assess-ment of tooth preparation for porcelain veneer restorations. Journalof Dentistry 1995;23:165–70.

[29] Olgart L, Brannstrom M, Johnson G. Invasion of bacteria into dent-inal tubules; experiments in vivo and in vitro. Acta OdontologicaScandinavica 1974;32:61–70.

[30] Brannstrom M. Etiology of dentin hypersensitivity. Proceedings ofthe Finnish Dental Society 1992;88(Suppl. 1):7–13.

[31] Ellege DA, Schorr BI. A provisional restoration technique for lami-nate veneer preparation. The Journal of Prosthetic Dentistry1989;62:139–42.

[32] Rada RE, Jankowski BJ. Provisional laminate veneer provisionali-zation using visible light-curing acrylic resin. Quintessence Interna-tional 1991;22:291–3.

[33] Nikaido T, Burrow MF, Tagami J, et al. Effect of pulpal pressure onadhesion of resin composite to dentin: bovine serum versus saline.Quintessence International 1995;26:221–6.

[34] Cagidiaco MC, Ferrari M, Garberoglio R, et al. Dentin contamina-tion protection after mechanical preparation for veneering. Ameri-can Journal of Dentistry 1996;9:57–60.

[35] Cobb DS, Reinhardt JW, Vargas MA. Effect of HEMA-containingdentin desensitizers on shear bond strength of a resin cement. Amer-ican Journal of Dentistry 1997;10:62–5.

[36] Paul SJ, Scha¨rer P. The dual bonding technique: a modified methodto improve adhesive luting procedures. International Journal of Peri-odontics & Restorative Dentistry 1997;17:537–45.

[37] Kelsey WP, Latta AM, Blankenau RJ. Effect of provisional restora-tions on dentin bond strength of resin cements. American Journal ofDentistry 1998;11:67–70.

[38] Highton R, Caputo AA, Matyas J. A photoelastic study of stresses onporcelain laminate preparations. The Journal of Prosthetic Dentistry1987;58:157–61.

[39] Hui KKK, Williams B, Davis EH, et al. A comparative assessmenton the strengths of porcelain veneers for incisor teeth dependent ontheir design characteristics. British Dental Journal 1991;171:51–5.

[40] Gilde H, Lenz P, Furst U. Untersuchungen zur belastbarkeit vonKeramikfacetten. Deutsche Zahna¨rtzliche Zeitschrift 1989;44:869–71.

[41] Meijering AC, Creughers NHJ, Roeters FJM, et al. Survival of threetypes of veneer restorations in a clinical trial: a 2.5-year interimevaluation. Journal of Dentistry 1998;26:563–8.

[42] Clyde JS, Gilmoure A. Porcelain veneers: a preliminary review.British Dental Journal 1988;164:9–14.

[43] Wildgoose DG, Winstanley RB, van Noort R. The laboratoryconstruction and teaching of ceramic veneers: a survey. Journal ofDentistry 1997;25:119–23.

[44] Stangel I, Nathanson D, Hsu CS. Shear strength of the compositebond to etched porcelain. Journal of Dental Research 1987;66:1460–5.

[45] Lacy AM, Laluz J, Watanabe LG, et al. Effect of porcelain surfacetreatment on the bond to composite. The Journal of ProstheticDentistry 1988;60:288–91.

[46] Nicholls JI. Tensile bond of resin cements to porcelain veneers. TheJournal of Prosthetic Dentistry 1988;60:443–7.

[47] Lu R, Harcourt JK, Tyas MJ, et al. An investigation of the compositeresin/porcelain interface. Australian Dental Journal 1992;37:12–19.

[48] Schaffer H, Dumfahrt H, Gausch K. O¨ berflachenstruktur undsubstanzverlust beim a¨tzen keramischer materialien. SchweizerischeMonatsschrift Zahnmedicin 1989;90:530–43.

[49] Yen TW, Blackmann RB, Baez RJ. Effect of acid etching on theflexural strength of a feldspathic porcelain and a castable glass cera-mic. The Journal of Prosthetic Dentistry 1993;70:224–33.

[50] Peumans M, Van Meerbeek B, Yoshida Y, et al. Porcelain veneersbonded to tooth structure: an ultra-morphological FE-SEM exami-nation of the adhesive interface. Dental Materials 1999.

[51] Calamia JR, Vaidyanathan J, Vaidyanathan TK, et al. Shear bondstrength of etched porcelains. Journal of Dental Research1985;64:828 Abstract 1096.

[52] Roulet JF, So¨derholm KJM, Longmate J. Effects of treatment andstorage conditions on ceramic/composite bond strength. Journal ofDental Research 1995;74:381–7.

[53] Hussain MA, Bradford EW, Charlton G. Effect of etching on thestrength of aluminious porcelain jacket crowns. British Dental Jour-nal 1979;147:89–90.

[54] Jones DW. The strength and strengthening mechanisms of dentalceramics. In: McLean JW, editor. Dental ceramics, Proceedings ofthe First International Symposium on Ceramics, Chicago: Quintes-sence Publishing Co, 1983. p. 83–136.

[55] Albers HF. Bonded tooth colored restoratives: indirect bondedrestorations, Santa Rosa: Alto Books, 1989. p. 1–42 Supplement.

[56] Jones GE, Boksman L, McConnell RL. Effect of etching techniqueon the clinical performance of porcelain veneers. Quintessence ofDental Technology 1986;10:635–7.

[57] Aida M, Hayakawa T, Mizukawa K. Adhesion of composite toporcelain with various surface conditions. The Journal of ProstheticDentistry 1995;73:464–70.

[58] Suh BI. All Bond—Fourth generation dentin bonding system. Jour-nal of Esthetic Dentistry 1991;3:139–46.

[59] Muller G. Atzen und silaniseren dentaler keramiken. DeutscheZahnartzliche Zeitschrift 1988;43:438–41.

[60] Williamson RT, Mitchell RJ, Breeding LC. The effect of fatigue onthe shear bond strength of resin bonded porcelain. Journal ofProsthodontics 1993;2:115–9.

M. Peumans et al. / Journal of Dentistry 28 (2000) 163–177 175

[61] Swift B, Walls AWG, McCabe JF. Porcelain veneers: the effect ofcontaminants and cleaning regimens on the bond strength of porce-lain to composite. British Dental Journal 1995;179:203–8.

[62] Holtan JR, Lua MJ, Belvedere P, et al. Evaluating the effect of glovecoating on the shear bond strength of porcelain laminate veneers.Journal of the American Dental Association 1995;126:611–6.

[63] Sheth J, Jensen M, Tolliver D. Effect of surface treatment on etchedporcelain bond strength to enamel. Dental Materials 1988;4:328–37.

[64] Tjan AHL, Dunn JR, Grant BE. Effect of fitting paste on bondstrength of composite/porcelain. Journal of Dental Research1989;68:270 Abstract 714.

[65] Della Bona A, Northeast SE. Shear bond strength of resin bondedceramic after different try-in procedures. Journal of Dentistry1994;22:103–7.

[66] Aboush YE. Removing saliva contamination from porcelain veneersbefore bonding. The Journal of Prosthetic Dentistry 1998;80:649–53.

[67] Brodbelt RHW, O’Brien WJ, Fan PL. Translucency of dental porce-lains. Journal of Dentistry 1980;59:70–75.

[68] Nathanson D, Hassan F. Effect of etched porcelain thickness onresin–porcelain bond strength. Journal of Dental Research1987;66:245 Special Issue, Abstract 1107.

[69] Strang R, McCrosson J, Muirhead GM, et al. The setting of visiblelight-cured resins beneath etched porcelain veneers. British DentalJournal 1987;163:149–51.

[70] Chan KC, Boyer DB. Curing light-activated composite cementthrough porcelain. Journal of Dental Research 1989;68:476–80.

[71] Blackman R, Barghi N, Duke E. Influence of ceramic thickness onthe polymerization of light-cured resin cement. The Journal of Pros-thetic Dentistry 1990;63:295–300.

[72] Linden JJ, Swift EJ, Boyer DB, et al. Photo-activation of resincements through porcelain veneers. Journal of Dental Research1991;70:154–7.

[73] O’Keefe KL, Pease PL, Herrin HK. Variables affecting the spectraltransmittance of porcelain through porcelain veneer samples. TheJournal of Prosthetic Dentistry 1991;66:434–8.

[74] Cardash HS, Baharav H, Pilo R, et al. The effect of porcelain coloron the hardness of luting composite resin cement. The Journal ofProsthetic Dentistry 1993;69:620–3.

[75] Lutz F, Phillips RW. A classification and evaluation of compositeresin systems. The Journal of Prosthetic Dentistry 1983;50:480–8.

[76] Inokoshi S, Willems G, Van Meerbeek B, et al. Dual-cure lutingcomposites. Part I: filler particle distribution. Journal of Oral Reha-bilitation 1993;20:133–46.

[77] Van Meerbeek B, Inokoshi S, Davidson CL, et al. Dual cure lutingcomposites—Part II: Clinically related properties. Journal of OralRehabilitation 1994;21:57–66.

[78] Andreasen FM, Flu¨gge E, Daugaard-Jensen J, et al. Treatment ofcrown fractured incisors with laminate veneer restorations: anexperimental study. Endodontics and Dental Traumatology1992;8:30–5.

[79] Stokes AN, Hood JAA. Impact fracture characteristics of intactcrown human central incisors. Journal of Oral Rehabilitation1993;20:89–95.

[80] Magne P, Douglas WH. Porcelain veneers: dentin bondingoptimization and biomimetic recovery of the crown. The Interna-tional Journal of Prosthodontics 1999;12:111–21.

[81] Magne P, Kwon KR, Belser UC, et al. Crack propensity of porcelainlaminate veneers: a simulated operatory evaluation. Journal of Pros-thetic Dentistry 1999;81:327–34.

[82] Magne P, Versluis A, Douglas WH. Effect of luting compositeshrinkage and thermal loads on the stress distribution in porcelainlaminate veneers. Journal of Prosthetic Dentistry 1999;81:335–44.

[83] Strassler HE, Nathanson D. Clinical evaluation of etched porcelainveneers over a period of 18 to 42 months. Journal of EstheticDentistry 1989;1:21–8.

[84] Rucker ML, Richter W, MacEntee M, et al. Porcelain and resin

veneers clinically evaluated: 2-year results. Journal of the AmericanDental Association 1990;121:594–6.

[85] Nordbq H, Rygh-Thoresen N, Henaug T. Clinical performance ofporcelain laminate veneers without incisal overlapping: 3-yearresults. Journal of Dentistry 1994;22:342–5.

[86] Jager K, Stern M, Wirz J. Laminates-reif fur die Pra¨xis? Quintessenz1995;46:1221–30.

[87] Peumans M, Van Meerbeek B, Lambrechts P, et al. Five-year clin-ical performance of porcelain veneers. Quintessence International1998;29:211–21.

[88] Kihn PW, Barnes DM. The clinical evaluation of porcelain veneers:a 48-month clinical evaluation. Journal of the American DentalAssociation 1998;129:747–52.

[89] Calamia JR. Clinical evaluation of etched porcelain veneers. Amer-ican Journal of Dentistry 1989;2:9–15.

[90] Jordan RE, Suzuki M, Senda A. Clinical evaluation of porcelainlaminate veneers: a four-year recall report. Journal of EstheticDentistry 1989;1:126–37.

[91] Christensen GJ, Christensen RP. Clinical observations of porcelainveneers: a three year report. Journal of Esthetic Dentistry1991;3:174–9.

[92] Strassler HE, Weiner S. Seven to ten year clinical evaluation ofetched porcelain veneers. Journal of Dental Research 1995;74:176Abstract 1316.

[93] Walls AWG. The use of adhesively retained all-porcelain veneersduring the management of fractured and worn anterior teeth. Part II:clinical results after 5-years follow-up. British Dental Journal1995;178:337–9.

[94] Dunne SM, Millar J. A longitudinal study of the clinical perfor-mance of porcelain veneers. British Dental Journal 1993;175:317–21.

[95] Shaini FJ, Shortall ACC, Marquis PM. Clinical performance ofporcelain laminate veneers. A retrospective evaluation over a periodof 6.5 years. Journal Oral Rehabilitation 1997;24:553–9.

[96] Bausch JR, De Lange K, Davidson CL, et al. Clinical significance ofpolymerization shrinkage of composite resins. The Journal of Pros-thetic Dentistry 1982;48:59–67.

[97] Shinkai K, Suzuki S, Leinfelder KL, et al. Effect of gap dimensionon wear resistance of luting agents. American Journal of Dentistry1995;8:149–51.

[98] Mc Kinney JE, Wu W. Chemical softening and wear of dentalcomposites. Journal of Dental Research 1985;64:1326–31.

[99] Roulet JF, Walti C. Influence of oral fluids on composite resin andglass–ionomer cement. The Journal of Prosthetic Dentistry1984;52:182–9.

[100] Vrijhoef MMA, Hendriks FHJ, Letzel H. Loss of substance of dentalcomposite restorations. Dental Materials 1985;1:101–5.

[101] Sorensen JA, Strutz JM, Avera SP, et al. Marginal fidelity andmicroleakage of porcelain veneers made by two techniques. TheJournal of Prosthetic Dentistry 1992;67:16–22.

[102] Sim C, Ibbetson RJ. Comparison of fit of porcelain veneers fabri-cated using different techniques. International Journal of Prostho-dontics 1993;6:36–42.

[103] Wall JG, Reisbick MH, Espelata KG. Cement luting thicknessbeneath porcelain veneers made on platinum foil. The Journal ofProsthetic Dentistry 1992;68:448–50.

[104] Lim CC, Ironside JG. Grit blasting and the marginal accuracy of twoceramic veneer systems—a pilot study. The Journal of ProstheticDentistry 1997;77:359–64.

[105] McIntyre MF, Bochiechio RA, Johnson R. Marginal gap width witha new refractory porcelain system. The Journal of Prosthetic Dentis-try 1993;69:564–7.

[106] Harasani MH, Isidor F, Kaaber S. Marginal fit of porcelain andindirect composite laminate veneers under in vitro conditions. Scan-dinavian Journal of Dental Research 1991;99:262–8.

[107] Coyne B, Wilson NHF. The marginal adaptation of porcelain lami-nate veneers. Journal of Dental Research 1987;66:885 Abstract 452.

M. Peumans et al. / Journal of Dentistry 28 (2000) 163–177176

[108] Hannig M, Jepsen S, Jasper V, et al. Der Randschlub glaskera-mischer Veneers mit zervikaler Schmelz-oder Dentinebegrenzung.Deutsche Zahna¨rtzliche Zeitschrift 1995;50:227–9.

[109] Tay WM, Lynch E, Auger D. Effects of some finishing techniques oncervical margins of porcelain laminates. Quintessence International1987;18:599–602.

[110] Goldstein GR, Barnhard BR, Penugonda B, Profilometer SEM.Profilometer, SEM, and visual assessment of porcelain polishingmethods. The Journal of Prosthetic Dentistry 1991;65:627–34.

[111] Patterson CJW, McLundie AC, Stirrups DR, et al. Efficacy of porce-lain refinishing systems in restoring surface finish after grinding withfine and extra-fine diamond burs. The Journal of Prosthetic Dentistry1992;68:402–6.

[112] Ward MT, Tate WH, Powers JM. Surface roughness of opalescentporcelains after polishing. Operative Dentistry 1995;20:106–10.

[113] Haywood VB, Heymann HO, Kusy RP, et al. Polishing porcelainveneers: an SEM and specular reflectance analysis. Dental Materials1988;4:116–21.

[114] Haywood VB, Heymann HO, Scurria MS. Effect of water, speed,and experimental instrumentation on finishing and polishing porce-lain intra-orally. Dental Materials 1989;5:185–8.

[115] Roulet JF. Margin quality: criteria and techniques for assessment. In:Anusavice Kenneth J, editor. Quality evaluation of dental restora-tions: criteria for placement and replacement, Lombourg: Quintes-sence Publishing Co, 1989. p. 223–9.

[116] Suzuki S, Leinfelder K, Shinkai K. Wear resistance of resin cements.American Journal of Dentistry 1995;8:83–7.

[117] Gladys S, Van Meerbeek B, Inokoshi S, et al. Clinical and semi-quantitative marginal analysis of four tooth-coloured inlay systemsat 3 years. Journal of Dentistry 1996;23:329–38.

[118] Feilzer AJ, De Gee AJ, Davidson CL. Setting stress in compositeresin in relation to configuration of the restoration. Journal of DentalResearch 1987;66:1636–9.

[119] Feilzer AJ, De Gee AJ, Davidson CL. Increased wall-to-wall curingcontraction in thin bonded resin layers. Journal of Dental Research1988;68:48–50.

[120] Tjan AHL, Dunn JR, Sanderson IR. Microleakage patterns of porce-lain and castable ceramic laminate veneers. The Journal of ProstheticDentistry 1989;61:276–82.

[121] Lacy AM, Wada C, Du W, et al. In vitro microleakage at the gingivalmargin of porcelain and resin veneers. The Journal of ProstheticDentistry 1992;67:7–10.

[122] Zaimoglu A, Karaaglac¸lioglu L, Uctacli. Influence of porcelainmaterial and composite luting resin on microleakage of porcelainlaminate veneers. Journal of Oral Rehabilitation 1992;19:319–27.

[123] Sim C, Neo J, Kiam Chua EK, et al. The effect of dentin bondingagents on the microleakage of porcelain veneers. Dental Materials1994;10:278–81 Abstract 215.

[124] Gaspercic D. Micromorphometric analysis of cervical enamel struc-ture of human upper third molars. Archives of Oral Biology1995;40:453–7.

[125] Zaimoglu A, Karaagac¸lioglu L. Microleakage in porcelain laminateveneers. Journal of Dentistry 1991;19:369–72.

[126] Dietschi D, Magne P, Holz J. Bonded to tooth ceramic restorations:in vitro evaluation of the efficiency and failure mode of two modern

adhesives. Schweizerische Monatsschrift Zahnmedicin 1995;105:229–305.

[127] Done B. Techniques for placement of porcelain veneers. Journal ofthe American Dental Association 1987;114:155–6.

[128] Jankowski JL, Olin PS, Holtan JR. Microleakage of porcelainveneers bonded with dual cured luting agents and fatigued. Journalof Dental Research 1993;72:130.

[129] Newcomb GM. The relationship between the location of subgingivalcrown margins and gingival inflammation. Journal of Periodontol-ogy 1974;45:151–4.

[130] Janenko C, Smales RJ. Anterior crowns and gingival health. Austra-lian Dental Journal 1979;24:225–30.

[131] Chan C, Weber H. Plaque retention on restored teeth with full-cera-mic crowns: a comparative study. The Journal of Prosthetic Dentis-try 1986;56:666–71.

[132] Olsson J, van der Heijde Y, Holmberg K. Plaque formation in vivoand bacterial attachment in vitro on permanently hydrophobic andhydrophilic surfaces. Caries Research 1992;26:428–33.

[133] Hahn R, Weiger R, Netuschil L, et al. Microbial accumulation andvitality on different restorative materials. Dental Materials1993;9:312–6.

[134] Quirynen M, Bollen CML. The influence of surface roughness andsurface-free energy on supra- and subgingival plaque formation inman. A review of the literature. Journal of Clinical Periodontology1995;22:1–14.

[135] Kourkata S, Walsh TF, Davis LG. The effect of porcelain laminateveneers on gingival health and bacterial plaque characteristics. Jour-nal of Clinical Periodontology 1994;21:638–40.

[136] Pippin DJ, Mixson JM, Sodan-Els P. Clinical evaluation of restoredmaxillary incisors: veneers vs. PFM crowns. Journal of the Amer-ican Dental Association 1995;126:1523–9.

[137] Peumans M. The clinical performance of veneer restorations andtheir influence on the periodontium. Thesis. Katholieke UniversiteitLeuven. 1997.

[138] Davis BK, Aquilino SA, Lund PS, et al. Subjective evaluation of theeffect of porcelain opacity on the resultant color of porcelainveneers. International Journal of Prosthodontics 1990;3:567–72.

[139] Davis BK, Scott JO, Johnston WM. Effect of porcelain shades onfinal shade of porcelain veneers. Journal of Dental Research1991;70:475 Abstract 1671.

[140] Davis BK, Aquilino SA, Lund PS, et al. Colorimetric evaluation ofthe effect of porcelain opacity on the resultant color of porcelainveneers. International Journal of Prosthodontics 1992;5:130–6.

[141] Mormann WH, Link C, Lutz F. Color changes in veneer ceramicscaused by bonding composite resins. Acta Med Dent Helv1996;1:97–102.

[142] Davis BK, Papcum LJ, Johnston WM. Effect of cement shade onfinal shade of porcelain veneers. Journal of Dental Research1991;70:297 Abstract 250.

[143] Davis BK, Johnston WM, Saba RF. Kubelka-Munk reflectancetheory applied to porcelain veneer systems using a colorimeter.International Journal of Prosthodontics 1994;7:227–33.

[144] Meijering AC, Roeters JM, Mulder J, et al. Patients’ satisfactionwith different types of veneer restorations. Journal of Dentistry1997;25:493–7.

M. Peumans et al. / Journal of Dentistry 28 (2000) 163–177 177