Abstract
Objective
The purpose of this study was to compare the use of different variables to measure the clinical wear of two denture tooth materials in two analysis centers.
Methods
Twelve edentulous patients were provided with full dentures. Two different denture tooth materials (experimental material and control) were placed randomly in accordance with the split-mouth design. For wear measurements, impressions were made after an adjustment phase of 1–2 weeks and after 6, 12, 18, and 24 months. The occlusal wear of the posterior denture teeth of 11 subjects was assessed in two study centers by use of plaster replicas and 3D laser-scanning methods. In both centers sequential scans of the occlusal surfaces were digitized and superimposed. Wear was described by use of four different variables. Statistical analysis was performed after log-transformation of the wear data by use of the Pearson and Lin correlation and by use of a mixed linear model.
Results
Mean occlusal vertical wear of the denture teeth after 24 months was between 120 μm and 212 μm, depending on wear variable and material. For three of the four variables, wear of the experimental material was statistically significantly less than that of the control. Comparison of the two study centers, however, revealed correlation of the wear variables was only moderate whereas strong correlation was observed among the different wear variables evaluated by each center.
Significance
Moderate correlation was observed for clinical wear measurements by optical 3D laser scanning in two different study centers. For the two denture tooth materials, wear measurements limited to the attrition zones led to the same qualitative assessment.
1
Introduction
Knowledge of the wear resistance of dental materials and of the likely effects of the materials on opposing natural teeth are important aspects of restorative dentistry. Although numerous wear-testing machines have been developed for preclinical study of the wear of dental materials , no method of wear simulation has yet been accepted internationally, and results from most of the methods used do not include clinical wear data . Only recently have some research groups shown that in vitro wear tests can be clinically representative . The major problem has been that all laboratory wear machines simulate only one or two of the wear mechanisms simultaneously present in the mouth. They are unable to replicate the oral environment with all its biological variations .
Clinical data on the wear behavior of dental materials are, therefore, particularly valuable, even if clinical wear measurements are methodologically difficult and require much effort. Clinical wear studies are essential, first for better assessment of the wear behavior of new dental materials and, second, to monitor the wear of natural teeth. Tooth wear with different etiology has been recognized as a major problem with increasing prevalence .
For measurement of clinical wear, either clinical categorization systems or indirect methods with study casts have been used. Some popular methods, for example the United States Public Health Service (USPHS) system , enable only subjective categorization of wear. With another method, the Leinfelder system , subjective, but at least semiquantitative, measurements of wear are possible. Most wear-evaluation systems have the disadvantage of assessing wear at restoration margins only and do not enable quantitative measurement of wear of whole restorations or tooth surfaces . They are, therefore, unsuitable for measurement of the wear of denture teeth or full crowns. These methods also systematically underestimate true wear .
Advances in measurement techniques have led to the use of 3D laser-scanning devices which enable non-contact surface profilometry and wear measurements by superimposition of baseline and follow-up scans (occlusal matching) . Nowadays, such objective quantitative procedures, rather than subjective evaluation scales, are recommended, and are adequate for studies of clinical wear. Non-contact 3D wear measurement with a laser scanner is currently regarded as the most accurate and effective technique for clinical wear analysis .
Several clinical studies have used 3D laser-scanning devices and occlusal matching to quantify the occlusal wear of different biomaterials . Although these studies used the same principle of quantitative wear measurement, they differ in minor, but probably important, methodological details. They also used different variables to measure wear (volume wear versus vertical height loss) and different occlusal areas (attrition zones versus complete occlusal surfaces) to report clinical wear data. It is not yet clear to what extent wear data from different centers, obtained by use of 3D laser scanning, are comparable, or the extent to which their accuracy and precision is affected by the scanning hardware, matching software, and operator.
The purpose of this study was to compare the use of different variables for measurement of clinical wear by use of optical 3D devices in two different centers. Wear data for two different denture tooth materials from patients with complete dentures were collected by one center over a period of 24 months by use of a split-mouth design. One tooth material was a double-cross-linked polymer (DCL = control material); the other was an improved version of the first material. In a laboratory study tooth brushing resulted in less wear of the experimental material than of the DCL material .
Casts of the dentures were evaluated for wear in two different centers. Both centers performed non-contact 3D wear measurements with the same methodological approach but with different laser-scanner devices, different matching routines, and different variables for reporting wear data. We intended (1) to investigate the extent to which the different wear variables correlated with each other and (2) to compare statistically the performance of the two materials.
2
Materials and methods
2.1
Subjects
This study was part of a multi-center clinical trial (seven centers) involving the same denture tooth materials. Wear results from all the test centers have been published elsewhere . The participants in this clinical trial were initially ten edentulous patients with indication for new full dentures. Two further patients were subsequently recruited because of dropouts early in the study. Exclusion criteria were subjects with allergy to the ingredients of the denture base or the denture tooth material, subjects who wear their existing dentures for fewer than 6 h per day, subjects for whom compliance could not be expected, and subjects who received their first set of full dentures less than 12 months ago. The mean age of the participants was 74.6 years (SD 10), seven were female. All participants were required to sign a consent form. The study protocol was approved by the local review board of Heidelberg University Hospital (ethical approval no. 375/2006).
2.2
Clinical procedures
All participants were treated in the Department of Prosthodontics of Heidelberg University Hospital, Germany, and provided with complete dentures in the maxilla and mandible. For one subject the lower denture was retained by use of four implants. The dentures were made in accordance with the usual routines of complete denture treatment. For occlusal adjustment, centric occlusion and the principle of bilateral balanced dynamic occlusion were used. All dentures were fabricated from the denture base material ProBase High Impact (Ivoclar Vivadent, Schaan, Liechtenstein) in accordance with the manufacturer’s instructions for use (standard polymerization).
Two different denture tooth materials were used to manufacture the dentures. One material (control group) was a double-cross-linked polymer (DCL; Ivoclar Vivadent). The other denture tooth material (study group) was an improved version of the first material containing 20% UDMA/PMMA fillers (experimental material; Ivoclar Vivadent) ( Table 1 ). In the latter material the polymer and the matrix were homogeneously cross-linked; this was achieved by subjecting both the pre-cross-linked polymer and the matrix to a secondary cross-linking process. The occlusal anatomy was identical for both materials and for all subjects. The posterior teeth were produced in SR Ortholingual molds and the anterior teeth in SR Vivodent molds, at Ivoclar Vivadent. Both tooth materials were available in Vita classical color shades A1, A2, A3, and A3.5. After determination of size and color, the two different tooth materials were coded with numbers and sent to the Department of Prosthodontics of Heidelberg University Hospital. None of the dental technicians, clinical operators, or patients knew which material was used for which quadrant.
| Enamel and dentin | Cervical | ||
|---|---|---|---|
| Experimental material (wt%) | DCL (control) (wt%) | Both materials (wt%) | |
| Polymethyl methacrylate | 72 | 33–35 | 94–97 |
| Dimethacrylate | 8 | 5–7 | 1–4 |
| Cross-linked polymethyl methacrylate | 0 | 59 | 0 |
| UDMA/PMMA fillers | 20 | 0 | 0 |
| Pigments | <0.5 | <0.5 | <0.5 |
| Initiators and stabilizers | <0.5 | <0.5 | <0.5 |
The two denture tooth materials were placed in accordance with the split-mouth design. The study teeth and the control teeth were randomly assigned to the left and right sides of the dentures. Randomization was performed by Ivoclar Vivadent, by use of the minimization technique and including the variables gender, age, and smoker/non-smoker.
To evaluate wear, extra-oral impressions of the dentures were prepared at baseline (i.e. after an occlusal adjustment period of 1–2 weeks), and after 6 months (24 ± 2 weeks after baseline), 12 months (48 ± 2 weeks after baseline), 18 months (72 ± 4 weeks after baseline), and 24 months (96 ± 4 weeks after baseline). Two impressions (one each of the upper and lower dentures) were obtained with a polyvinylsiloxane impression material (Flexitime Putty/Flexitime Correct Flow; Heraeus Kulzer, Germany) by use of the dual-viscosity technique. Metal, nonperforated stock trays with retentive rims (Ergolock XL; Omnident, Rodgau, Germany) were used to make the impressions; these were then used to perform wear measurements in center 1 (Department of Prosthodontics, Heidelberg University Hospital, Germany). Furthermore, four impressions (one in each quadrant) were made with another polyvinylsiloxane impression material (Virtual putty/Virtual light, Ivoclar Vivadent) by use of the same technique and plastic trays for each quadrant (Systemp. Impression Trays, Ivoclar Vivadent). These impressions were sent to center 2 (Ivoclar Vivadent, Schaan, Liechtenstein) for evaluation of wear.
Before taking impressions the dentures were inspected for stains, calculus, and foreign debris, and, if necessary, cleaned in an ultrasonic cleaner. If dental calculus present in the fissures was not removed by the ultrasonic cleaner, the calculus was carefully removed with an ultrasonic handpiece. Red occlusion foil (Hanel 12 μm; Coltène Whaledent, Langenau, Germany) was placed to mark the occlusal contact points in centric occlusion. The clinical status, particularly the marked occlusal contact points of the lower and upper dentures, was photographed. These photographs were used to determine the occlusal contact areas for the wear measurements and to determine reference areas for occlusal matching.
2.3
Wear analysis in center 1 (Heidelberg, Germany) and center 2 (Ivoclar Vivadent Schaan, Liechtenstein)
In both centers impressions were poured with the same plaster, type IV dental stone (GC Fujirock Hard Rock; GC, Tokyo, Japan). Wear was determined indirectly by use of the casts with commercially available laser-scanning devices. Center 1 used the Laserscan 3D (Willytec, Gräfelfing, Germany). This scanner is not commercially available any more since 2006. Center 2 used the etkon es1 (formerly Willytec GmbH, now Straumann CADCAM, Gräfelfing, Germany). The difference between the scanners is that the Laserscan 3D scans the object from one vertical perspective only whereas the etkon es1 scanner scans the object from nine different sides and combines the images in one image with the result that steep angles and undercuts are scanned. Comparative analysis of the devices revealed only slight differences between wear measurements if the same specimens were scanned and analyzed (less than 2%; unpublished data).
Before scanning, the casts were evaluated for voids and/or pearls. Plaster pearls were carefully removed with a sharp scalpel blade, paying attention to remove the pearls only. In center 2 a stereo microscope (6× magnification) was used to do this.
Quantification of the wear of all casts was performed by one operator in each center who was trained in use of the method. First, the occlusal surfaces of the casts were digitized with the laser-scanning devices. Each denture quadrant was scanned separately. In both centers the data sets obtained in this way were then checked for surface changes (wear) by use of surface-analysis software (Match 3D, Version 1.6; Willytec, Gräfelfing, Germany). The extent of wear was calculated by superimposing the baseline and follow-up scans (occlusal matching) by means of an automated superimposition process, as described by Mehl et al. .
Before the baseline and follow-up scans were superimposed, at least three reference areas or reference points that were not abraded by wear, according to the intraoral pictures, were selected from at least two adjacent teeth (e.g. 1st molar and 2nd premolar). The reference areas were mostly located on the buccal and/or lingual side of the teeth. The matching process was performed with a minimum of 800 points and 20,000 iterations in center 1, whereas a minimum of 1200 points and 8000 iterations were used in center 2. The matching process was repeated several times until the standard deviation of the last iterations was less than 20 μm. In analysis center 2, standard deviations of up to 30 μm were allowed in specific cases if other criteria were met (see below). In analysis center 1 a threshold value of −30 μm was defined to prevent surface changes as a result of wear or artifacts in the casts (voids, blebs) from impairing the superimposition process; this was not done in analysis center 2. This routine means that all areas of the follow-up cast which differed from the baseline cast by more than 30 μm in the negative direction were not included in the matching process. Details of the mathematical algorithms used for this occlusal matching have been published elsewhere .
In analysis center 2, the standard deviation was only one criterion for a satisfactory match. Other criteria were the differential picture and the distribution of the z -values. The differential picture results from the superimposition and matching process for the baseline and follow-up models and includes negative—which means material loss—and positive values. A symmetrical bell-shaped distribution of the z -values is an indicator of adequate matching. An additional criterion for adequate matching is the distribution of the z -values in the differential picture in areas for which little wear is observed: the z -values should be in the range ±15 μm in these areas. If the standard deviation was above the acceptance level or the histogram revealed an uneven distribution of z -values the matching process was regarded as unsuccessful and the teeth were not included in the wear analysis.
For wear analysis in center 1 the occlusal surfaces of the first and second premolars and the first molars were evaluated. The maximum vertical loss of the entire occlusal surface, calculated by use of the surface-analysis software, is referred to below as HD max. Separate analysis was conducted for attrition zones (occlusal contact areas) identified by means of clinical pictures with marked occlusal stops. For these attrition zones maximum vertical loss was also calculated; this is referred to below as HD oca max.
Wear analysis in center 2 was performed as follows. First, areas of attrition were identified from the clinical pictures. The wear zones on the differential pictures usually corresponded to the occlusal stops on the clinical picture. These areas were digitally cut, and for each area the vertical loss was calculated by use of the software. For both premolars, two wear measurements were taken (buccal and lingual). For both molars, four wear measurements were taken (mesiobuccal, distobuccal, mesiolingual, and distolingual). For each tooth, the mean and maximum of these two or four measurements were regarded as summary measures, and are referred to below as Ivo mean and Ivo max. In both analysis centers the 99%-quantile was used to eliminate outliers.
2.4
Statistical analysis
To analyze the correlation among the four wear variables (HD max, HD oca max, Ivo mean, and Ivo max) a log-transformation (natural logarithms) was performed so all wear measurements better approached a normal distribution. Both the classical Pearson correlation and the Lin correlation were calculated. The former is a measure of (linear) association, penalizing only random differences between the wear variables. The latter is a measure of agreement (also interpretable as a type of intraclass correlation), penalizing both random and systematic differences (and hence the bias) between the variables. Thus, the latter is, by definition, smaller than or equal to the former.
For statistical assessment of the combined effects of material and time, we considered, for the different outcomes (Log HD max, Log HD oca max, Log Ivo mean, and Log Ivo max), a linear mixed model including a random subject effect to account for the correlations within measurements of the same subject. Other factors and covariates included in the model were the time of the measurement (as a categorical variable with four possible values: 6, 12, 18, or 24), the material (DCL or experimental), the type of tooth (4, 5, 6, or 7), the jaw (mandible or maxilla), and the quadrant (left or right).
2
Materials and methods
2.1
Subjects
This study was part of a multi-center clinical trial (seven centers) involving the same denture tooth materials. Wear results from all the test centers have been published elsewhere . The participants in this clinical trial were initially ten edentulous patients with indication for new full dentures. Two further patients were subsequently recruited because of dropouts early in the study. Exclusion criteria were subjects with allergy to the ingredients of the denture base or the denture tooth material, subjects who wear their existing dentures for fewer than 6 h per day, subjects for whom compliance could not be expected, and subjects who received their first set of full dentures less than 12 months ago. The mean age of the participants was 74.6 years (SD 10), seven were female. All participants were required to sign a consent form. The study protocol was approved by the local review board of Heidelberg University Hospital (ethical approval no. 375/2006).
2.2
Clinical procedures
All participants were treated in the Department of Prosthodontics of Heidelberg University Hospital, Germany, and provided with complete dentures in the maxilla and mandible. For one subject the lower denture was retained by use of four implants. The dentures were made in accordance with the usual routines of complete denture treatment. For occlusal adjustment, centric occlusion and the principle of bilateral balanced dynamic occlusion were used. All dentures were fabricated from the denture base material ProBase High Impact (Ivoclar Vivadent, Schaan, Liechtenstein) in accordance with the manufacturer’s instructions for use (standard polymerization).
Two different denture tooth materials were used to manufacture the dentures. One material (control group) was a double-cross-linked polymer (DCL; Ivoclar Vivadent). The other denture tooth material (study group) was an improved version of the first material containing 20% UDMA/PMMA fillers (experimental material; Ivoclar Vivadent) ( Table 1 ). In the latter material the polymer and the matrix were homogeneously cross-linked; this was achieved by subjecting both the pre-cross-linked polymer and the matrix to a secondary cross-linking process. The occlusal anatomy was identical for both materials and for all subjects. The posterior teeth were produced in SR Ortholingual molds and the anterior teeth in SR Vivodent molds, at Ivoclar Vivadent. Both tooth materials were available in Vita classical color shades A1, A2, A3, and A3.5. After determination of size and color, the two different tooth materials were coded with numbers and sent to the Department of Prosthodontics of Heidelberg University Hospital. None of the dental technicians, clinical operators, or patients knew which material was used for which quadrant.
