Materials and methods.
Sixty single-rooted teeth, including maxillary incisors, maxillary and mandibular canines and single-rooted maxillary and mandibular premolar teeth were collected from a number of sources and stored in a solution of water and 5% thymol at room temperature. They were examined under 2x magnification to ensure that they were free of caries, restorations or cracks, which might affect their resistance to experimental loading. Mesiodistal and buccolingual dimensions were recorded at the cervical margin with the aid of a digital caliper (Anglia, STMicroelectronics, Edinburgh, UK) to ensure that each experimental group would contain teeth of a similar size distribution. Forty teeth were selected from the initial 60 teeth and allocated to four groups (Fig.1).
- Group A: The control group. Ten posts and cores were cemented in dowel channels with no retained buccal dentine.
- Group B: Ten posts and cores were cemented in dowel channels with 3 mm of retained buccal dentine.
- Group C: Ten posts and cores were cemented in dowel channels with 4 mm of retained buccal dentine.
- Group D: Ten posts and cores were cemented in dowel channels with 5 mm of retained buccal dentine.
Each group contained 10 teeth with a similar distribution of mesiodistal and buccolingual dimensions. The measurements are given in Tables 1and 2.
Figure 1. The completed specimens.
Table 1. Mesiodistal dimensions (mm) of sample teeth.
Table 2. Buccolingual dimensions (mm) of sample teeth.
Teeth in group A were sectioned15 mm from the apex, 2 mm coronal to the amelo-cemental junction, at a level corresponding to the clinical gingival margin (Turner 1982, Patel & Gutteridge 1996). In groups B, C and D, the teeth were sectioned18, 19 and 20 mm, respectively, from the apex. Post channels were prepared using the Parapost system (Whaledent Int., NY, USA) using the black1.5-mmdrill under constant irrigation with saline, to create a post space of 9 mm in group A, with the root face at right angles to the channel and a potential post length of12,13 and14 mm in groups B, C and D, respectively. As differing tooth types were used in each group, it was felt that retaining 3, 4 and 5 mm of coronal dentine in the experimental groups would be most likely to demonstrate significant effects on the strength of the tooth, and on this occasion, smaller 1- and 2-mm preparations were not included.
A shoulder was prepared around the most coronal retaining dentine in groups B, C and D. To facilitate this, a jig was waxed up on a smooth, black, plastic Parapost to produce a flat central button, covering an area of root face dentine, 1.5 mm wide, buccally and lingually. Mesially and distally, the width of the dentine core covered was reduced to1mmto accommodate the oval anatomical outline of the root face. These dimensions were selected to produce a clinically realistic shoulder width of approximately1mm. Once cast, the jig was positioned in the prepared posthole and the shoulder prepared around it, ensuring that the amount of retaining coronal dentine was standardized. The shoulder was prepared with a depth limited diamond bur and finished with a tapered, plain cut tungsten carbide bur, sectioned to leave 3, 4 or 5 mm of cutting length. An antirotational notch was cut on the lingual aspect of teeth in all the groups.
Each specimen was embedded indie stone (Begostone, BEGO Bremer Goldschl gereiaeilh, Bremen, Germany), 2 mm from the shoulder in groups B, C and D and 2 mm from the sectioned tooth surface in group A, contained within a specially made stainless steel tube. The same tube was used to hold specimen under loading. A surveyor (Degussa, AG, Geschaftsbereich Dental, Frankfurt I, Germany)was used to align the post channel with the outside of the steel tube by ensuring that the analysing rod positioned within the post hole was parallel to the outside of the tube.
Post and cores were then waxed up for each tooth using black, laboratory burn-out posts and a standard mould for waxing of the core section. This was a silicone index of an initial wax-up of a tooth in group A. After casting with nickel-chrome alloy (Heraeus, Heraenium NA, Heraeus Kulzer GmbH, Hanau, Germany) and finishing, the post and cores were cemented using zinc phosphate cement (Elite cement 100, GC corporation, Tokyo, Japan), spun down the channels with a spiral paste filler (Dentsply Maillefer Instruments, SA, Ballaigues, Switzerland) prior to seating the castings firmly with digital pressure for 3 min. Specimens were stored in 100% humidity at room temperature and tested 1week later (Tjan & Whang 1985). As the study was designed to test the strength of a tooth restored with a partial post and core where the height of retained coronal dentine was the only variable, specimen preparation was simplified by omitting the construction of a crown over the post and core (Volwiler et al.1989, Patel & Gutteridge1996). In the test groups B, C and D, cementation of a crown would have introduced a ferrule effect as compared to the control group A and this would have made interpretation of the results more difficult.
The stainless steel tube containing the specimen was mounted in a specially designed retaining arm of an Instron universal testing machine (Instron1195, Instron Limited, Buckinghamshire, UK). The load was applied at1308 from the horizontal with a steel rod, as previous studies have shown this to be the most clinically comparable angle of loading in anterior teeth (Eshelman & Sayegh 1983, Volwiler et al. 1989, Patel & Gutteridge 1996). The point of load application was standardized by the inclusion of a small mark in the silicone mould used to construct the core. The cross-head speed was 10 mm min_1, and the specimens were loaded to failure with a chart speed of 100 m min. Results a represented in the form of force at failure (N) from the force defection curve. Subsequently, a representative specimen from each group was selected and photographed to record the pattern of coronal fracture.
Comparison of results from the four test groups was carried out using one-way analysis of variation (anova) and Fisherâ€™s PLSD (Protected Least Significant Difference) test to demonstrate differences between pairs of groups.