This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright
Transcript
This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies areencouraged to visit:
d e n t a l m a t e r i a l s 2 7 ( 2 0 1 1 ) e1–e9
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
Round robin test: Wear of nine dental restorative materialsin six different wear simulators – Supplement to the roundrobin test of 2005
S.D. Heintzea,∗, W.W. Barkmeierb, M.A. Lattab, V. Roussonc
a R&D, Ivoclar Vivadent AG, Bendererstrasse 2, FL-9494 Schaan, Liechtensteinb Department of General Dentistry, Creighton University School of Dentistry, Omaha, NE, USAc Biostatistics Unit, Institute for Social and Preventive Medicine, University of Lausanne, Switzerland
a r t i c l e i n f o
Article history:
Received 2 June 2009
Received in revised form 7 June 2010
Accepted 6 September 2010
Keywords:
Wear
Round robin test
In vitro
Composite
Ceramic
Amalgam
Wear simulator
a b s t r a c t
Objective. The purpose of the present study was to submit the same materials that were
tested in the round robin wear test of 2002/2003 to the Alabama wear method.
Methods. Nine restorative materials, seven composites (belleGlass, Chromasit, Estenia,
Heliomolar, SureFil, Targis, Tetric Ceram) an amalgam (Amalcap) and a ceramic (IPS Empress)
have been submitted to the Alabama wear method for localized and generalized wear. The
test centre did not know which brand they were testing. Both volumetric and vertical loss
had been determined with an optical sensor. After completion of the wear test, the raw data
were sent to IVOCLAR for further analysis. The statistical analysis of the data included loga-
rithmic transformation of the data, the calculation of relative ranks of each material within
each test centre, measures of agreement between methods, the discrimination power and
coefficient of variation of each method as well as measures of the consistency and global
performance for each material.
Results. Relative ranks of the materials varied tremendously between the test centres. When
all materials were taken into account and the test methods compared with each other, only
ACTA agreed reasonably well with two other methods, i.e. OHSU and ZURICH. On the other
hand, MUNICH did not agree with the other methods at all. The ZURICH method showed the
lowest discrimination power, ACTA, IVOCLAR and ALABAMA localized the highest. Material-
wise, the best global performance was achieved by the leucite reinforced ceramic material
Empress, which was clearly ahead of belleGlass, SureFil and Estenia. In contrast, Heliomolar,
Tetric Ceram and especially Chromasit demonstrated a poor global performance. The best
consistency was achieved by SureFil, Tetric Ceram and Chromasit, whereas the consistency
of Amalcap and Heliomolar was poor. When comparing the laboratory data with clinical
data, a significant agreement was found for the IVOCLAR and ALABAMA generalized wear
method.
Significance. As the different wear simulator settings measure different wear mechanisms,
it seems reasonable to combine at least two different wear settings to assess the wear
e2 d e n t a l m a t e r i a l s 2 7 ( 2 0 1 1 ) e1–e9
1. Introduction
In 2005 the results of a round robin test on the wear of den-tal materials were published [1]. Ten dental materials hadbeen tested with five laboratory wear methods. The resultswere quite different. When allocating relative ranks to thematerials there was only little agreement between the fivewear methods. One explanation was that the wear methodsfollow different wear generating concepts which result in adifferent ranking of the materials. The test centres used dif-ferent wear simulators, different forces, different antagonistmaterials, different number of cycles, with or without ther-mocycling, etc. Some used abrasive mediums and differentmethods to evaluate the material loss. ZURICH additionallyincluded 5 h of simulated toothbrushing between the phasesas well as storage of the samples in ethanol. Furthermore,some methods showed a low discriminatory power whichcan be explained by the device that is used in conjunctionwith the method. Detailed qualification and validation pro-tocols that show that the wear device is qualified and thewear method is validated are not available [2]; this holdstrue for all devices that are included in the round robintest.
The laboratory Alabama wear method that has – accord-ing to a review on wear – the highest citation frequency inthe dental literature was not included in the round robintest at that time [2]. The Alabama wear method is alsoincluded in the ISO Technical Specification on the wear bytwo/three body contact [3]; three of the other five methods(ACTA, OHSU, Zurich) are also included in the ISO TechnicalSpecification.
The Alabama wear method was developed by Leinfelderand Suzuki and is therefore also called Leinfelder-Suzuki wearmethod. The method was first published in 1989 [4] and isa modification of a device that has been originally designedby Roulet [5]. Several major modifications were made overthe years. In the first publication in 1989, a polyethylenetape was used as intermediate substance, driven by a tapeadvancing system [4]. The tape was replaced by a slurry ofPMMA beads ten years later [6]. The original force was 55 N,which was increased to 75 N ten years later. In the first pub-lication, a stainless steel stylus with 2 mm radius hit thespecimen without rotation. In the new method an additional30◦ clockwise rotation was integrated as soon as the sty-lus hits the surface of the specimen which was then called“localized wear”. Additionally, with a new specimen a flat sty-lus made of polyacetal is brought into contact with the flatspecimen; the wear produced by this approach is called “gen-eralized wear”. The materials specimens are incorporated intoextracted molars that are trimmed flat. The stylus for gener-alized wear is made of polyacetal, the one for localized wearis stainless steel. The original publication states that eachspring is calibrated with a 200 kg load cell in conjunction witha universal testing machine prior to testing, but no data havebeen reported with regard to the deviations, the scattering ofresults and the time intervals of force measurements or thereplacement frequency of the spring. Most of the early pub-lished data come from the same authors (Leinfelder and/orSuzuki [4,6,16–17]).
Later, modifications to the original methods reported byLeinfelder and Suzuki were introduced to increase the repro-ducibility and reliability of the overall method. In an effort tomove away from placing test specimens in extracted humanteeth, a standardized cavity in a stainless steel custom fixturewas used for positioning the tests materials in the wear sim-ulator. To eliminate variations due to the wear of the acetalgeneralized wear stylus an identically shaped flattened stain-less steel was introduced [7]. For localized wear, a customantagonist fixture was used that could accommodate a stain-less steel ball bearing with a radius of 2.387 mm was usedin place of the original hardened steel cone-shaped stylus.The original hardened steel localized antagonist tip surfacedegraded with use altering the surface finish of the stylus.Using ball bearings facilitates a cost-effective way to provide anew, standardized antagonist for each test specimen and eachexperiment trial [8,9].
The aim of the present study was to submit the same dentalmaterials that were included in the round robin test publishedin 2005 to the Alabama wear method and to compare theresults to the other five methods with the same statisticalmethodology. Therefore, this publication is a supplement tothe 2005 publication; Section 2 as well as Section 5 are abbre-viated and the reader is asked to consult the first publication.
2. Materials and methods
The selected materials were the same as in the first phaseof the round robin test except for Targis 130 ◦C. As the oventhat cured Targis specimens at a temperature of 130 ◦C was nolonger available, this material had to be excluded. However,Targis specimens cured at 95 ◦C were included. Except for twomaterials (belleGlass and Targis) batches other than the onesused in the first phase of the round robin test (2002–2003) hadto be used since the batch was not longer available or hadalready expired. Table 1 lists the materials with their batchnumbers of the first and second phase of the round robintest.
The materials were produced in the same way as in the firstphase. They were produced at Ivoclar by one operator, codedwith numbers and sent to the test centre so that the centrewas not aware which material they were testing.
2.1. ALABAMA wear method
The ALABAMA wear method was carried out at the Center forOral Health Research at Creighton University (Omaha, USA)using three Leinfedler-Suzuki wear simulators that were thor-oughly calibrated before testing the specimens. The materialswere submitted to both generalized wear testing using a flatsurface stainless steel stylus and localized wear using a stain-less steel ball bearing. Specimens for both wear models wereplaced in a water bath with slurry of PMMA beads (averageparticle size 44 �m). Each generalized and localized specimenwas surface scanned with the Proscan 2000 non-contact opti-cal profilometer (Scantron Industrial Products, Ltd, Taunton,England) using a S38/3 sensor with a depth of field of 3000 �m.Proscan and ProForm software were used for quantificationof material changes between the “before” and “after” surfaceprofiles.
d e n t a l m a t e r i a l s 2 7 ( 2 0 1 1 ) e1–e9 e3
Table 1 – List of materials, manufacturer, type of material, processing and batch-number.
Material Manufacturer Type of material Processing Batch2003
Batch2008
Amalcap Plus Ivoclar Vivadent Amalgam Mixing C25527 J14154Empress TC1 Ivoclar Vivadent Leucite reinforced ceramic Pressing C35146 G17472belleGlass enamel KerrLab Fine particle hybrid composite Polymerization by heat and light 911422 911422Chromasit S4 Ivoclar Vivadent Microfilled composite Polymerization by heat and pressure C15082 K15062Estenia Enamel E2 Kuraray Hybrid composite Polymerization by heat and light 202CA 00020B/E2Heliomolar 210 B Ivoclar Vivadent Microfilled composite Polymerization by light 29157 K30118SureFil Shade Dentsply Fine particle hybrid composite Polymerization by light 990615 051013Targis Incisal S1 Ivoclar Vivadent Fine particle hybrid composite Polymerization by heat and light C05051 C05051Tetric Ceram 210 Ivoclar Vivadent Fine particle hybrid composite Polymerization by light C16761 G08104
2.2. Ivoclar Vivadent-Method (IVOCLAR)
After processing and before testing, the specimens (n = 8) werekept dry at a temperature of 37 ◦C for 24 h. The samples weremounted in a chewing simulator that is commercially avail-able by Willytec (SD Mechatronik, Germany). Antagonists aremade by pressed IPS Empress ceramic (Ivoclar Vivadent) whichwas glazed two times at a temperature of 870 ◦C. The diameterof the antagonist is 2.36 mm at a height of 600 �m measuredfrom the tip of the antagonist. A weight of 5 kg had been puton each vertical bar. The sliding movement has been fixedat 0.7 mm. The frequency of the antagonist movement was1.6 Hz. A total of 120,000 cycles of unidirectional antagonistmovements have been carried out. Thermocycling with a fre-quency of 320/120,000 cycles has been included in the wearsetting with a temperature difference between 5 ◦C and 55 ◦C.After completing the wear generating procedure, replicas weremade with white super hard plaster (Fuji Superhard rock, GCCorporation, Japan). The plaster replicas have been analyzedby means of a commercially available laserscanner device(Laserscan 3D, Willytec, Germany) and the appropriate Match3D software [10]. The volumetric (IVVOL) as well as the maxi-mal vertical loss (IVVERT) (1% percentile) have been calculatedby the software.
2.3. Zurich-Method (ZURICH)
After processing and before testing the samples (n = 8) werekept in water at a temperature of 36.5 ◦C for 2 weeks. Thismethod has been described elsewhere in detail [11]. With aload of 49 N and a frequency of 1.7 Hz the palatal cusps cutout of similar upper molars pushed against the surface ofthe specimens (n = 6) that were mounted on a rubber socketat 45◦ angle allowing the antagonist to glide over the surfaceof the test specimen. The test specimens were kept in waterwith exchanged temperature according to a thermocyclingprotocol (3000× 5 ◦C/55 ◦C). After 120,000, 240,000, 640,000, and1,200,000 cycles of loading the samples were submitted to atoothbrushing device with a slurry of toothpaste for 30 min,30 min, 100 min, and 140 min resp. [12]. Additionally, at theend of the first phase (120,000 cycles) the samples were putinto a solution of 75% ethanol for 20 h to simulate the chemi-cal degradation. After each thermomechanical sequence, themaximal vertical loss of both the specimens (occlusal contactarea OCA) and the antagonists as well as the vertical loss inthe contact free area (CFA) were calculated by using a comput-
erized 3D-scanner [13]. The scanner is driven by step motorswhich scan the object in 1 �m steps in the z-direction and in100 �m steps in the xy-direction [5]. In each test sequence sixdifferent materials have been submitted to wear which wererandomly allocated to the six test chambers.
2.4. OHSU-Method (OHSU)
This method has been described elsewhere in detail [14]. Inprinciple, enamel cusps were forced into contact with thespecimens through a layer of food like slurry (mixture of poppyseeds and PMMA beads). The enamel cusps were drilled out ofhuman upper molars of similar shape giving them a spher-ical shape of a diameter of 10 mm. The enamel stylus werepolished with 600 grit and 1000 grit silicon carbide slurry andpolished with 5 �m aluminium oxide paste and then ultrasoni-cally cleaned for 1 min. Before mounting the specimens (n = 10)in the wear simulator, they were kept in water for 24 h at 37 ◦C.First the cusp was forced with a load of 50 N on the surface,sliding across a linear path of 8 mm producing abrasive wear.At the end of each path, a static load of 80 N was applied toproduce localized attrition wear. For a whole test sequence100,000 cycles at 1 Hz with unidirectional movements wererun. The mean vertical loss of the abrasion and attrition wearfacets were measured with a profilometry device at 10 definedtracks. The values of tracks 4–6 correspond to the abrasionwear (OHSUABR) and the tracks 8–9 to the attrition wear facet(OHSUATT).
2.5. Munich-Method (MUNICH)
For this method, a prototype of the wear simulator used forthe Ivoclar Vivadent method has been used but with the con-figuration according to a pin-on-block-design allowing thetest specimens (n = 8) gliding under permanent contact to thespherical antagonist (Degusit aluminium oxide, 5 mm diam-eter) at a linear distance of 8 mm and a vertical load of50 N. After processing and before testing the samples (n = 8)were kept in physiological sodium chloride solution at roomtemperature for 7 days. During the chewing simulation thesamples were rinsed with distilled water at 37 ◦C. At 10,000,30,000, and 50,000 double cycles (bidirectional forth- and back-movement), replicas have been made with white plaster (FujiSuperhard rock, GC Corporation, Japan). The volumetric lossof the wear facet has been determined on the plaster modelswith the Laserscan 3D device described above (Willytec, Ger-
e4 d e n t a l m a t e r i a l s 2 7 ( 2 0 1 1 ) e1–e9
many). In each test sequence eight different materials havebeen submitted to wear which were randomly allocated to theeight test chambers.
2.6. ACTA-Method (ACTA)
Two metal wheels rotate in different directions with about 15%difference in the circumferential speed while having near con-tact [15]. The test specimens (n = 24 and n = 28 resp.) are placedon the circumference of the wheel while the other wheelserves as antagonist. The force with which the two wheelscome together was adjusted to 15 N. The wheels were placed ina slurry of white millet seeds in a buffer solution. After 50,000,100,000 and 200,000 cycles the maximal vertical loss of the testspecimens was measured with a profilometry device.
3. Statistical methods
For the sake of clarity, only the wear data after completing allcycles are presented.
As each variable has a different scale it is impossible tocompare the variables with each other. Thus, the logarithmictransformation of the data was applied to stabilize the vari-ance for each variable. In Fig. 1 the transformed raw data forall methods and variables are presented.
3.1. ANOVA and power of discrimination
ANOVA was applied for each of the eleven variables usingthe log-transformed data. The power of discrimination foreach variable could be measured by the R2-value provided byANOVA. This value represents the percentage of the total vari-ance in the data, caused by the variation between materials.
3.2. Agreement between methods
The sample means of the log-transformed data were usedfor ranking the 9 materials with respect to each variable bycreating “relative ranks”. For each variable, the relative rankwas assigned 1 to the material with the lowest mean log-value, denoted by m, and the relative rank 9 to the materialwith the highest mean log-value, denoted by M. Then, relativeranks were defined in order to respect the relative differencesbetween materials. Thus, the relative rank of a material withmean log-value x was set to 1 + 8(x − m)/(M − m).
Table 2 – Discrimination power (%) of the differentvariables.
For the agreement between two variables the Mean of theAbsolute Deviations of the Ranks (MADR) has been calculated(for the formula see [1]).
3.3. Comparison of materials
Ranks had also been used to compare the materials with eachother with respect to “global performance” and “consistencyof performance” (across the methods). In order to give thesame weight to each method, only the variable IVVERT (forIVOCLAR), the variable OHSUABR (for OHSU) and the variableALABAMA LOCVERT (for ALABAMA) were considered for thisprocedure. The global performance Gi of a material can bemeasured by the median of its relative ranks with respect tothe six methods. The consistency of performance Ci was cal-culated by the mean of absolute deviations with respect toglobal performance.
3.4. Reliability of methods
The reliability of the different wear measurement methods isrelated to its variabilities which were assessed by calculatingthe coefficient of variation.
4. Results
4.1. ANOVA and power of discrimination
ANOVA led to clear-cut significant results (p < 0.0001) for eachvariable, meaning that at least some of the materials differedfrom the others.
Except for the Zurich method (53.5%) and the ALABAMAgeneralized wear method (80%) all other methods achieveda discrimination power of about 90% or more (Table 2). TheALABAMA localized wear had a higher discriminatory powerthan ALABAMA generalized wear.
4.2. Agreement between methods
The relative ranks of the materials in relation to each vari-able are given in Table 3 and graphically illustrated in Fig. 2.One can see at first glance that the MUNICH method wasvery different from all the other methods. This was con-firmed by the MADR values provided in Table 4. By contrast,the values IVVERT and IVVOL, the variables OHSUABR andOHSUAT, ALABAMA LOCVOL and ALABAMA GENVOL as well as
d e n t a l m a t e r i a l s 2 7 ( 2 0 1 1 ) e1–e9 e5
Fig. 1 – Boxplot graphs of log-transformed wear raw data for each test method. 1 = Empress, 2 = belleGlass, 3 = SureFil,4 = Estenia, 6 = Amalcap, 7 = Targis, 8 = Heliomolar, 9 = Tetric Ceram, 10 = Chromasit. Note: Material no. 5 (Targis cured at130 ◦C) has been excluded for the analysis of the present study.
ALABAMA LOCVERT and ALABAMA GENVERT strongly agreedwith each other (MADR = 0.6 for IVOCLAR and OHSU, and 0.1for ALABAMA). It is also noteworthy that ACTA was the methodwhich agreed the most with the other variables, in particularwith ZURICH and OHSUABR as well as with all variables relatedto the ALABAMA method. The ALABAMA localized wear vari-ables correlated well with the ZURICH method.
4.3. Comparison of materials
The best global performance was achieved by the ceramicmaterial Empress (G = 1.3) which was clearly ahead of belle-Glass (G = 4.7), SureFil (G = 5.2) and Estenia (G = 5.2) (Table 3);belleGlass and Estenia are composite materials for indirectuse. In contrast to these materials, Heliomolar (G = 7.6), Tetric
Author's personal copy
e6 d e n t a l m a t e r i a l s 2 7 ( 2 0 1 1 ) e1–e9
Table 3 – Relative ranks of materials with respect to each variable global performance Gi and consistency Ci for eachmaterial (1: IVOCLAR VOL, 2: IVOLCLAR VERT, 3: ZURICH, 4: OHSUABR, 5: OHSUATT, 6: MUNICH, 7: ACTA, 8: ALABAMALOCVOL, 9: ALABAMA LOCVERT, 10: ALABAMA GENVOL; 11: ALABAMA GENVERT).
Fig. 2 – Agreement between laboratory methods measured by relative ranks 1–9 (y-axis).
Ceram (G = 7.9) and especially Chromasit showed a poor globalperformance (G = 9.0). The best consistency was achieved bySureFil, Tetric Ceram and Chromasit (C = 1.3 each), whereasAmalcap and Heliomolar demonstrated a poor consistency(C = 2.4 and C = 2.5). The consistency of all the other materialswas between 1.7 and 1.9.
4.4. Reliability of method
When the coefficients of variation are compared with eachother (Table 5), it can be noted that the variables IVVERTand ACTA were those which varied the least, followedby ALBAMA LOCVERT. OHSUABR, OHSUATT, MUNICH and
d e n t a l m a t e r i a l s 2 7 ( 2 0 1 1 ) e1–e9 e7
Table 5 – Coefficient of variation (multiplied by 100) for each material and each variable, as well as mean and median of 9coefficients of variation for each variable (1: IVOCLAR VOL, 2: IVOLCLAR VERT, 3: ZURICH, 4: OHSUABR, 5: OHSUATT, 6:MUNICH, 7: ACTA, 8: ALABAMA LOCVOL, 9: ALABAMA LOCVERT, 10: ALABAMA GENVOL; 11: ALABAMA GENVERT).
above all ZURICH had the highest coefficient of variation(Table 5).
5. Discussion
When comparing the Alabama wear results of specific mate-rials of the present round robin study with studies that usedthe Alabama method and the same materials, similar resultswere obtained. In a study that evaluated composite materialsfor indirect use the ranking of the materials Estenia, belle-Glass and Targis was the same compared to the present studywith regard to localized wear [16]; for generalized wear bel-leGlass and Estenia were in the same rank whereas in thepresent study the wear was significantly lower for Esteniathan for belleGlass. The difference as well as the differencesin the absolute values may be explained by the fact that inthe cited study the materials were submitted to 100,000 cyclescompared to 400,000 cycles in the present study. Differencesmay have also been due to the better standardization of thespecimens, the antagonists, and the greater degree of accuracypossible with the optical scanning system. The high localizedwear compared to the generalized wear for the amalgam isline with another study that evaluated the amalgam mate-rials Tytin and Dispersalloy which were quite similar [17]: inthe present study the difference between localized and gen-eralized wear was about 11-fold for Amalcap whereas it wasabout 12- to 13-fold in the cited study.
When the relative ranks of the materials were calculated,the results varied tremendously between the test centres.When taking all materials into account and comparing thetest methods, only ACTA agreed reasonably well with threeother methods, ZURICH, OHSU abrasion and ALABAMA local-ized wear, achieving mean values of absolute deviations of theranks below 2 (Table 4). On the other hand, MUNICH did notagree with all the other methods at all which was already thecase in the first phase of the round robin test. It should benoted that the range of stress generated by the respective wearsystems varies greatly.
In-vitro wear evaluations done in the very early stagesof composite resin development showed that the wear rateof amalgam was virtually constant across a range of stresswhereas, resin composite wear was significantly influenced by
the stress generated in wear simulation [18]. This is attributedto the ductile mode of surface failure exhibited by amal-gam compared to the more complex crack initiated mode offailure that is often observed with resin composites [19,20].With respect to the comparison of the wear of ceramic andresin composite, the ALABAMA model is consistent with thesuperior performance of the ceramic evidenced by previousinvestigations [1].
For the ALABAMA system the fact that the absolute valuesof generalized and localized wear are so different is clearly areflection of the reproduction of different wear factors for eachmodel [20]. For the generalized model, abrasion and erosionare likely the dominant forces, but their relative contributionis not known. For localized wear, it is likely that abrasion andfatigue or attrition are predominant [20]. The same abrasivemedia was used for both localized and generalized wear. Thecontribution to wear of the media is far more important forgeneralized wear, as the flattened wear antagonist in the gen-eralized method does not come into direct contact with thetest material.
As for the test method both vertical and volumetric losshad been evaluated for ALBAMA localized wear and ALBAMAgeneralized wear. Both variables within the same test methodwere strongly associated for IVOCLAR and ALABAMA gener-alized wear (mean value of absolute deviations of the ranksequal to 0.6 for Ivoclar and 0.1 for Alabama generalized wear),but less for ALABAMA localized wear (1.0, Table 4). Consid-ering the single materials and relative ranks, a difference ofmore than 1.5 units between the variables ALABAMA LOCVOLand ALABAMA LOCVERT was found for the materials Estenia,belleGlass and SureFil.
For a detailed discussion on the possible reasons for thewear of the different materials in relation to the wear sim-ulation method the authors refer to the first publication ofthe round robin test on wear [1]. In that publication three dif-ferent wear types could be differentiated. The Alabama wearmethod belongs to the second type (Type 2) to which also ACTAand OHSU abrasion belongs; all these methods use an abra-sive medium. The three-body-wear test is mainly an abrasiveprocess with low pressure loads. The wear rate of this testis principally influenced by the hardness and fracture tough-ness of the material. The wear simulators OHSU, Alabamaand ACTA are quite different from each other. Nevertheless, it
Author's personal copy
e8 d e n t a l m a t e r i a l s 2 7 ( 2 0 1 1 ) e1–e9
seems that the wear processes are similar and only Amalcapand Tetric Ceram represent exceptions in the correlation.
To date, all wear simulation methods lack the scientific evi-dence to prove that the in vitro simulation corresponds tothe in vivo situation. This has been stated already in 1989by the Council on Dental Materials [21] and not many firminsights have been gathered since then. The ZURICH methodclaims that 1,200,000 cycles in the simulator correspond tofive years in vivo. However, this assumption has not beensystematically verified in longitudinal clinical studies with dif-ferent materials and is only based on the extrapolation of alimited number of amalgam and indirect composite restora-tions [22,23]. For the Alabama wear simulator a comparativeanalysis with clinical data revealed that 100,000 cycles in thesimulator correspond to about 3.6 months in vivo [8]. Again,only a small number of fillings in vivo were involved in thiscomparison, although they were followed up for four yearswith yearly measurements and an additional measurementafter six months of clinical service.
For the ACTA and OHSU method reasonable correlationcoefficients between in vitro and in vivo data had also beenpublished [14,15]. These correlations, however, were in partestablished using semi-quantitative methods for assessingthe wear in vivo on pooled data from clinical trials. Meth-ods which use scale models, like the Vivadent Scale or MLScale, with defined steps at the restoration margin assumewear at the margin to be predictable in terms of general wearand occlusal contact wear [24]. However, they systematicallyunderestimate wear [25], as marginal breakdown, and under-/overfilling are possible confounders. Furthermore, wear at themargins is not correlated with wear of the occlusal contactarea. More sophisticated methods employ special microscopicdevices [26], mechanical computer-aided scanners [5,27,28], orlaser equipment [10].
Clinical studies that quantitatively measure wear onreplicas are complicated and time-consuming and requiresophisticated laboratory equipment. Therefore, only few stud-ies have been carried out so far. From the few data on clinicalwear one may learn that the results differ from study tostudy, which may be due to different wear analyzing methodsand sample selection. Furthermore, the results scatter widely,mainly due to patient- but also dentist-related factors [29].The coefficient of variation (CV) is found to be between 30%and 50% [2]. Most often a very limited number of fillings havebeen analyzed (<20 per material). A large data pool of 28 directcomposite materials, one amalgam, six indirect compositeresins, enamel and five ceramic materials, was published bythe Clinical Research Associates Newsletter [30–32]. The obser-vation period was one to four years and about 30 restorationshad been placed per material. However, no statistical analy-sis based on the variability of the results has been performed.Nevertheless, the results of the CRA data showed that enamel,amalgam, and ceramic materials exhibit low wear rates aftertwo years of service. For the materials that were includedin the present round robin test, CRA data showed low clini-cal wear for Heliomolar as well as for belleGlass and SureFil.Higher wear rates were given for Targis which was, however,evaluated in molar crowns. If the wear rates of the roundrobin test were correlated with the clinical wear rates of CRAusing relative ranks, Alabama generalized wear showed the
best correlation followed by the OHSU method (abrasion) andthe Ivoclar method. The results of the other methods showeda poor or even a negative correlation.
More longitudinal clinical wear data on different materi-als incorporated in a sufficient number of subjects are neededto really validate the in vitro data with regard to the wearmethod’s ability to predict clinical wear. Probably, the results oftwo different wear methods have to be combined to accuratelyassess the clinical wear resistance of dental materials.
6. Conclusions
- IVOCLAR (vertical loss) and ACTA were the best methodswith respect to the coefficient of variation. The discrimina-tion power of the ZURICH method was clearly inferior to allother methods.
- The variables related to the same method largely agreedwith each other for the IVOCLAR (vertical/volumetric) andOHSU method (abrasion/attrition) but not for the ALABAMAmethod (localized/generalized).
- The MUNICH method strongly disagreed with the other fivemethods.
- As the different wear simulator settings measure differentwear mechanisms, it seems reasonable to combine at leasttwo different wear settings to assess the wear resistance ofa new material.
- A systematic and comprehensive comparative analysis withclinical wear data shall be carried out to assess the predic-tive value of each wear method.
Acknowledgements
The authors would like to thank Till Göhring (University ofZurich), Karl-Heinz Kunzelmann (University of Munich), Mar-tin Rosentritt (University of Regensburg), and John Sorensen(University of Portland) for conducting the wear tests in therespective wear simulator settings and sending the raw datato IVOCLAR for further analysis. Furthermore, we would liketo thank Mrs. G. Zellweger (R&D Ivoclar Vivadent) for the fab-rication of the specimens.
r e f e r e n c e s
[1] Heintze SD, Zappini G, Rousson V. Wear of ten dentalrestorative materials in five wear simulators – results of around robin test. Dent Mater 2005;21:304–17.
[2] Heintze SD. How to qualify and validate wear simulationdevices and methods. Dent Mater 2006;22:712–34.
[3] ISO. Dental materials – guidance on testing of wear. Part 2.Wear by two-and/or three body contact. TechnicalSpecification 2001, No. 14569-2.
[4] Leinfelder KF, Beaudreau RW, Mazer RB. An in vitro devicefor predicting clinical wear. Quintessence Int 1989;20:755–61.
[5] Roulet J-F. Degradation of dental polymers. Basel: Karger;1987.
[6] Leinfelder KF, Suzuki S. In vitro wear device for determiningposterior composite wear. J Am Dent Assoc1999;130:1347–53.
[7] Barkmeier WW, Latta MA, Erickson RL, Wilwerding TM. Wearsimulation of resin composites and the relationship toclinical wear. Oper Dent 2008;33:177–82.
Author's personal copy
d e n t a l m a t e r i a l s 2 7 ( 2 0 1 1 ) e1–e9 e9
[8] Barkmeier WW, Latta MA, Erickson RL, Lambrechts P.Comparison of laboratory and clinical wear rates of resincomposites. Quintessence Int 2004;35:269–74.
[9] Kimmes N, Barkmeier WW, Latta MA. Localized wear ofresin composites using three antagonists. J Dent Res2010;89(Spec issue A) [Abstract 459].
[10] Mehl A, Gloger W, Kunzelmann KH, Hickel R. A new optical3-D device for the detection of wear. J Dent Res1997;76:1799–807.
[11] Krejci I, Reich T, Lutz F, Albertoni M. In-vitro-Testverfahrenzur Evaluation Dentaler Restaurationssysteme 1. Computer-gesteuerter Kausimulator. Schweiz Monatsschr Zahnmed1990;100:953–60.
[12] Krejci I, Albertoni M, Lutz F. In-vitro-Testverfahren zurEvaluation dentaler Restaurationssysteme 2. Zahnbürsten-/Zahnpastaabrasion und chemische degradation. SchweizMonatsschr Zahnmed 1990;100:1164–8.
[13] Krejci I, Reich T, Bucher W, Lutz F. Eine neue Methode für diedreidimensionale Verschleissmessung. Schweiz MonatsschrZahnmed 1994;104:160–9.
[14] Condon JR, Ferracane JL. Evaluation of composite wear witha new multi-mode oral wear simulator. Dent Mater1996;12:218–26.
[15] De Gee AJ, Pallav P. Occlusal wear simulation with the ACTAwear machine. J Dent 1994;22:S21–7.
[16] Suzuki S, Nagai E, Taira Y, Minesaki Y. In vitro wear ofindirect composite restoratives. J Prosthet Dent2002;88:431–6.
[17] Suzuki S. Does the wear resistance of packable compositeequal that of dental amalgam? J Esthet Restor Dent2004;16:355–65.
[18] Bailey WF, Rice SL. Comparative sliding-wear behavior of adental amalgam and a composite restorative as a functionof contact stress. J Dent Res 1981;60:731–2.
[19] Roberts JC, Powers JM, Craig RG. Wear of dental amalgam. JBiomed Mater Res 1977;11:513–23.
[20] Condon JR, Ferracane JL. Factors effecting dental compositewear in vitro. J Biomed Mater Res 1997;38:303–13.
[21] Council on Dental Materials, Instruments, and Equipment.Obstacles to the development of a standard for posteriorcomposite resins. J Am Dent Assoc 1989; 118:649–51.
[22] Lutz F, Krejci I. Mesio-occlusodistal amalgam restorations:quantitative in vivo data up to 4 years. A data base for thedevelopment of amalgam substitutes. Quintessence Int1994;25:185–90.
[23] Krejci I, Lutz F. In-vitro-Testverfahren zur Evaluationdentaler Restaurationssysteme-Korrelation mitIn-vivo-Resultaten. Schweiz Monatsschr Zahnmed1990;100:1445–9.
[24] Bryant RW. Comparison of three standards for quantifyingocclusal loss of composite restorations. Dent Mater1990;6:60–2.
[26] Christensen R, Bangerter V. Apparatus for automated,non-contact measurements of surface changes. J Dent Res1990;69:126 [Abstract 140].
[27] DeLong R, Pintado M, Douglas WH. Measurement of changein surface contour by computer graphics. Dent Mater1985;1:27–30.
[28] Krejci I, Reich T, Buchner W, Lutz F. Eine neue Methode zurdreidimensionalen Verschleissmessung. Schweiz MedZahnmed 1994;104:147–57.
[29] Söderholm KJ, Lambrechts P, Sarrett D, Abe Y, Yang MC,Labella R, et al. Clinical wear performance of eightexperimental dental composites over three yearsdetermined by two measuring methods. Eur J Oral Sci2001;109:273–81.
[30] CRA. Restorative resins: current generation Class 2 resinstatus report #3–2-year clinical performance. CRA Newsl1996;20:1–3.
[31] CRA. Posterior full crowns 2001. Part 2. Resin crowns, 4 yearclinical status. CRA Newsl 2001;25:1–3.
[32] CRA. Posterior full crowns 2001. Part 3. 2-year clinicalpeformance of CAD-CAM copings & full crowns. CRA Newsl2001;25:1–3.
