R. J. G. De Moor & G. M. G. Hommez
Department of Operative Dentistry and Endodontology, Dental School, University Hospital, University of Ghent, Ghent, Belgium.

Introduction.
The provision of a well-compacted and tightly adapted root filling is one of the goals of root canal treatment. However, contemporary obturation techniques and filling materials do not seal completely the root canal system up to the level of the cemento-enamel junction. Moreover, it is accepted that both apical and coronal leakage can occur following apparently successful root canal treatment (Hovland & Dumsha 1985, Saunders & Saunders 1994a).
Several factors appear to influence the extent of both apical and coronal leakage. The influence of the smear layer on leakage has been evaluated and its removal has been advocated in order to reduce leakage (Kennedy et al . 1986, Taylor et al . 1997). Furthermore, various root canal filling techniques based on heated or preheated gutta-percha have been introduced in order to enhance complete filling of the root canal. These include warm vertical condensation (Schilder 1967), warm lateral condensation (Endotec 1986), thermatic compaction (McSpadden 1980), hybrid condensation, i.e. a combination of cold lateral condensation and thermomechanical compaction (Tagger et al . 1984), thermoplasticized gutta-percha as a coating on a flexible carrier ( Johnson 1978), and injection moulded thermoplasticized guttapercha (Yee et al . 1977). With all these techniques an endodontic sealer is recommended.
Evaluation of the long-term sealing ability of root canal sealers as a function of time and at various time intervals within the same experimental group has been limited (Gutmann et al . 1993, Saunders & Saunders 1994b, Wu et al . 1995, Kontakiotis et al . 1997). From these studies it is clear that there is a tendency for sealers to leak more as a function of time. When the performance of different sealers was compared (AH26, Ketac-Endo, Sealapex, Tubliseal, Roth, PCS-EWT), it has often been shown that AH26 performed better than the other root canal sealers (Wu et al . 1995, Kontakiotis et al . 1997).
However, there have been no studies on the long-term sealing ability of AH26 used in conjunction with various gutta-percha techniques. The aim of the present study was therefore to evaluate both coronal and apical leakage of root canals obturated by various techniques in combination with AH26. A second objective was to verify the findings of a previous study (De Moor & De Boever 2000) investigating the apical sealing ability of AH26.

Materials and methods.
Seven hundred and ninety extracted human straight single-rooted teeth with mature apices were used in this study. Both carious (limited occlusal and/or interproximal lesions without pulp exposure) and non-carious teeth were included. All teeth were stored in 10% formalin until the sample was completed. In order to exclude teeth with multiple root canals, all teeth were radiographed from two angles before root canal treatment. Organic debris was removed by submerging the teeth in 2.5% sodium hypochlorite for 8 h. Subsequently, they were washed with tap water for 1 h, and stored in saline until used.

Sample preparation.
The crowns were removed 2 mm above the cementoenamel junction with a high-speed fissure bur and water spray. After gross removal of pulp tissues, a size 10 Flexofile (Dentsply Maillefer, Ballaigues, Switzerland) was introduced into the canal until it could be seen in the major apical foramen. The working length was determined by subtracting 1 mm from this length. The root canals were prepared by means of a crown-down/stepback technique by one operator (De Moor & Martens 1999, De Moor & De Boever 2000). The coronal half of the root canals was preflared with Gates Glidden drills (Dentsply Maillefer) in a larger to smaller sequence (numbers 4-3-2) and the canals were copiously irrigated with 2.5% sodium hypochlorite solution with a 27- gauge endodontic needle (Monoject, Sherwood Medical, St Louis, MO, USA). Smear layer was removed using File- Eze (Ultradent Products Inc., South Jordan, UT, USA) during canal preparation. The apical half of the canal was then prepared with the step-back technique. The master file varied between size 35 and 45. The canals were dried with paper points and the patency of the apical foramen was confirmed with a size 10 Flexofile. The roots were randomly divided into 10 experimental groups of 75 roots each, 20 positive and 20 negative control roots. Following the drying with paper points, the canals were obturated by one of the following techniques. The access cavities in the groups selected for testing apical leakage were filled with Ketac-Fil (Espe, Seefeld, Germany).

Group 1: cold lateral condensation of gutta-percha.
A standard size gutta-percha cone (Dentsply Maillefer) that matched the master apical file was fitted to the working length with tug back. AH26 sealer was mixed according to the manufacturer’s instructions and placed in the canal with a rotary paste filler to working length. The master cone was coated with AH26 and gently seated at the working length. Lateral condensation was then carried out using size 20 and 25 accessory gutta-percha cones with endodontic finger spreaders (Dentsply Maillefer) placed in the first instance to within 1 mm of the working length. The gutta-percha cones coated with sealer were laterally condensed until they could not be introduced more than 3 mm into the root canal.
Following obturation, the gutta-percha was removed from the coronal cavity up to the level of the cementoenamel junction with a warm instrument (PK Thomas Waxing Instrument, N PKT-2, Hu Friedy, Leimen, Germany) and vertically condensed with Machtou pluggers (Dentsply Maillefer).

Group 2: warm vertical condensation of gutta-percha.
A standard size gutta-percha cone that matched the master apical file was fitted to the working length with tug back. AH26 sealer was mixed following the manufacturer’s instructions and placed in the canal with a rotary paste filler to working length. The master cone was coated with AH26 and gently seated at the working length. The coronal portion of the master cone was removed with red-hot heat carriers (PK Thomas Waxing Instrument, N PKT-1 and PKT-2, Hu Friedy, Leimen, Germany). The gutta-percha was then condensed in an apical direction with cold Machtou pluggers. This procedure was repeated until the apical 5 mm of the root canal remained filled with gutta-percha after condensation. The latter was verified by subtracting 5 mm from the working length and marking this length by means of a rubber stop on the Machtou plugger of appropriate size. Subsequently, segments of gutta-percha, 3–4 mm long, were attached to a slightly heated plugger and condensed apically into the root canal. To soften the gutta-percha, additional heat was transferred to the gutta-percha mass with heat carriers. In between the addition of the gutta-percha and heat transfer, cold prefitted pluggers were used for condensation of the gutta-percha to a homogeneous mass. This was continued until the entire root canal was packed.

Group 3: hybrid condensation of gutta-percha.
The basis for the hybrid condensation technique was the same as for the cold lateral gutta-percha condensation, except that cold lateral condensation stopped when the apical 5 mm were condensed. Once this was achieved, a gutta-percha condenser (Dentsply Maillefer) of appropriate size was chosen. The size of the condensers varied between 25 and 40. The diameter of the condenser allowed insertion into the canal to at least a depth of 8 mm from the apical preparation. The condenser was then rotated without apical pressure (9000 r.p.m.). After 1 s the gutta-percha was plasticized and the condenser was carried apically up to 3 mm from the apical stop. The instrument was kept for 1 s at the desired level and gradually withdrawn, whilst rotating at full speed. Withdrawal speed was determined by the speed of the instrument itself. This obturation technique was a modification of the combined lateral condensation and thermatic compaction as described by Tagger et al . (1984). At the end of condensation, the gutta-percha was removed from the coronal cavity up to the level of the cementoenamel junction with a warm instrument and vertically condensed with Machtou pluggers.

Group 4: Thermafil obturation.
The correct size of the plastic core Thermafil Obturator (Dentsply Maillefer) was selected using the verification kit. The obturators were then placed in the Therma-Prep oven (Dentsply Maillefer) according to the manufacturer’s instructions. AH26 was sparingly introduced into the coronal third of each canal using a rotary paste filler, after which the plasticized Thermafil device was inserted to the apical stop. The shank of each carrier was cut at the canal orifice using an inverted cone bur in a high-speed handpiece and the gutta-percha was compacted vertically with a plugger.

Group 5: Soft-Core obturation.
The correct size of the plastic core Soft-Core (Soft-Core System Aps, Copenhagen, Denmark) obturator was selected using the Size Verifier. The obturators were then placed in the Soft-Core Oven (Soft-Core System). When the oven indicated that the obturator was ready, it was removed from one of the slots in the top of the oven. AH26 was sparingly introduced into the coronal third of each canal using the rotary paste filler, after which the plasticized Soft-Core device was inserted into the apical stop. The handle and insertion pin were removed by a twisting motion. Excess plastic core material was removed with a small inverted cone bur and any extra gutta-percha removed. The gutta-percha was then compacted vertically with a Machtou plugger.

Controls.
In addition to the specimens in the experimental groups, the root canals of a further 40 teeth were prepared. The canals were not obturated in 20 of the teeth, but the coronal opening, as well as the apex, was covered with cyanoacrylate cement before all of the surfaces of the root were coated with two layers of nail varnish. These were the negative controls. In the remaining 20 teeth, the root canals were not obturated and sealed neither apically nor coronally prior to leakage testing. These were the positive controls.

Staining, longitudinal splitting and dye measurement.
The samples were kept at 37 C in Vacutainers for 1 day, 1 week, and 4, 6 and 12 months in 80% relative humidity. Before storing the teeth, radiographs were taken from the buccal and the mesial side of every tooth. At the end of this period, the roots were covered with nail varnish. Clear acrylic nail varnish was applied to the entire tooth surface for the negative controls and all surfaces except the apical 2–3 mm for the teeth in the apical leakage testing group and the coronal opening in the coronal leakage testing group. The first coat was allowed to dry and a second coat was applied.
All teeth were immersed in a bath of India ink (Pelikan, Hannover, Germany) and stored at 37 C for 90 h, after which they were thoroughly washed with running water. The nail varnish was removed with a razor blade and the teeth soaked in acetone for 1 h to remove residues of nail varnish. The teeth were then air dried. Longitudinal shallow grooves were made on the buccal and lingual surface with a rotating diamond disk of small diameter under continuous water cooling, and the teeth carefully fractured and sectioned with a sharp chisel. Care was taken to include the apical foramen in the fracture line.
The degree of microleakage was determined by measuring

1. the linear extent of India ink penetration from the apical end of the preparation (apical leakage testing) and
2. the surface of the coronal gutta-percha at the level of the amelocemental junction to the position of the maximum dye penetration apically (coronal leakage testing).

To determine the most coronal (apical leakage testing) and apical (coronal leakage testing) point of linear leakage, two measurements were made. First, after splitting the roots, dye penetration was measured on the basis of dye visible on the gutta-percha filling material or root canal wall. Secondly, after the gutta-percha filling material was removed with an endodontic explorer, dye penetration was measured on the canal walls. The measurements of dye penetration were compared, and the larger one selected as the definitive point of linear dye leakage.
To eliminate bias, apical and coronal leakage were measured independently by two evaluators who were unaware of the obturation techniques used.
All measurements were obtained by means of a stereomicroscope (Stemi SR, Zeiss, Oberkochen, Germany) at x6 magnification with calibrated scale ocular.

The negative control teeth showed no dye penetration. The positive teeth showed dye penetration along the entire length of each root canal.
Agreement amongst examiners was evaluated using Cohen’s kappa (kappa = 0.92). No significant differences amongst the observers were scored, so that the calculation of the average leakage values of the two observers for each root was justified.
Since the data indicated a non-normal distribution, leakage was assigned using the following categories; Apical leakage: 0 = no leakage detected; 1 = up to 0.5 mm; 2 = 0.5–1 mm; 3 = 1–2 mm; 4 = >2 mm leakage (Table 1). Coronal leakage: 0 = no leakage detected; 1 = up to 1 mm; 2 = 1–2 mm; 3 = 2–4 mm; 4 = >4 mm leakage (Table 2). Statistical analysis was carried out using the Kruskal–Wallis test for-non-parametric data to determine whether there were significant differences between the groups. Pairs of groups were compared using the Mann–Whitney U -test.

Table 1. Apical leakage scores for the five obturation techniques.


Leakage scores: 0, no leakage detected; 1, up to 0.5 mm; 2, 0.5-1 mm; 3, 1-2 mm; 4, >2 mm leakage.


Table 2. Coronal leakage scores for the five obturation techniques.


Leakage scores: 0, no leakage detected; 1, up to 1 mm; 2, 1-2 mm; 3, 2-4 mm; 4, >4 mm leakage.

Linear apical leakage of the experimental roots.
There were statistically significant differences between groups after 7 days ( P < 0.05) but not after 1 day or after 4, 6 and 12 months ( P > 0.05). Pairs of groups were compared using the Mann–Whitney U -test and statistically significant differences ( P < 0.05) were found between Soft-Core obturators and vertical condensation after 1 week and after 4, 6 and 12 months; between Soft- Core obturators and lateral condensation after 1 day and 1 week; and between Soft-Core obturators and hybrid condensation at all time periods ( P < 0.05). From these differences in leakage scores the hybrid condensation was superior at all maturation times. Table 3 contains the data on mean apical leakage plus standard deviations, and the minimum and maximum extent of apical leakage by technique and time. Root fillings with Soft- Core obturators and AH26 leaked more than when AH26 was combined with the other four obturation techniques.

Linear coronal leakage of the experimental roots.
The results for coronal leakage are provided in Table 4. There were significant differences in coronal leakage between the various obturation techniques after 1 and 7 days, and after 4 and 12 months ( P < 0.05). There were, however, no statistically significant differences in leakage between groups after 6 months ( P > 0.05). According to the Mann–Whitney test no statistically significant differences ( P > 0.05) were found between Soft-Core obturators and Thermafil obturators at any evaluation time. Statistically significant differences were found between Soft-Core obturators and lateral condensation after 1 day, 1 week and 4 months ( P < 0.001) and 6 and 12 months ( P < 0.05); between Soft-Core obturators and vertical condensation after 1 day, 1 week, 4 and 12 months ( P < 0.001) and 6 months ( P < 0.05); between Soft-Core obturators and hybrid condensation after 1 day, 1 week and 4 months ( P < 0.001) and 6 and 12 months ( P < 0.05). There were statistically significant differences between Thermafil and lateral condensation after 1 week ( P < 0.05); between Thermafil and vertical condensation after 1 day, 1 week and 4 months ( P < 0.001); between Thermafil and hybrid condensation after 1 day, 1 week and 4 months ( P < 0.001).

Discussion.
In order to evaluate the sealing ability of root fillings, several in vitro methods have been designed. It is important to appreciate that not only is the apical seal of the root canal of importance, but the coronal seal is of equal importance for the success of treatment (Saunders & Saunders 1994a). The most common method used to assess leakage remains the measurement of dye penetration (Wu & Wesselink 1993). The results of dye penetration studies, however, are confusing and often result in variable conclusions (Wu & Wesselink 1993, Dalat & Spångberg 1994). This lack of agreement has been discussed by Wu & Wesselink (1993), who questioned the validity of leakage studies and recommended that more research should be devoted to leakage study methodology. The nature and the amount of leakage observed with this technique cannot be extrapolated to an in vivo situation. On the other hand, laboratory testing remains the only valuable preclinical screening test that can predict or indicate clinical performance (De Moor & Martens 1999).
Longitudinal sectioning of roots and the linear measurement of dye penetration were used in the present study for the measurement of leakage. On the one hand, it is clear that this experimental model cannot mimic the in vivo situation. On the other hand, it enables the observation of dye penetration and whether or not that is associated with porosities in the gutta-percha, the presence of empty spaces, stripping of the gutta-percha from the solid core system or changes in the structure of thermoplasticized gutta-percha. Splitting the root longitudinally combined with dye penetration also enables the demonstration of the pattern of dye penetration.
One of the aims of the present investigation was to verify the findings of a previous study (De Moor & De Boever 2000). In the present study and in our previous fourmonth study (De Moor & De Boever 2000) obturation with Soft-Core obturators resulted in greater leakage scores with higher standard deviations when compared with the other filling techniques. This may indicate that Soft-Core obturators are ineffective in the apical obturation of well-instrumented straight root canals under ideal conditions. In addition, it was also seen that apical leakage increased with time up to 4 months in all five obturation techniques. These findings were in general agreement with those of others (Gurney et al . 1971), in which leakage increased during the five-month evaluation period. Other studies on long-term apical sealing ability as a function of time and of the filling techniques investigated in the present study do not appear to have been published.

Table 3. Group means, standard deviations, and minimum and maximum extent of apical dye penetration in mm.


 
Table 4. Group means, standard deviations, and minimum and maximum extent of coronal dye penetration in mm.

The results regarding coronal leakage revealed no statistically significant differences between the three non-carrier systems and between the two carrier systems. This lack of a significant effect was not surprising, given that a similar procedure was used to seal the orifices of the specimens.
In our previous study (De Moor & De Boever 2000) it was found that the preheated gutta-percha of the Soft- Cores appeared to be porous when viewed under the microscope. Despite previously reported data in favour of the use of AH26 in combination with the gutta-percha coated system Thermafil (Dalat & Spångberg 1994), the presence of porosities in the gutta-percha of the Soft-Core obturators might explain the greater amounts of apical leakage found in the present and in our previous studies (De Moor & Martens 1999, De Moor & De Boever 2000). These porosities did not appear to be of importance when scoring the amount of coronal dye leakage.

References.

Dalat DM, Sp?ngberg LSW (1994) Comparison of apical leakage   in root canals obturated with various gutta-percha techniques using a dye vacuum   tracing method. Journal of Endodontics 20 , 315-9.
  De Moor RJG, De Boever JG (2000) The sealing ability of an epoxy resin root   canal sealer used with five gutta-percha obturation techniques. Endodontics   and Dental Traumatology 16 , 291-7.
  De Moor RJG, Martens LC (1999) Apical microleakage after lateral condensation,   hybrid gutta-percha condensation and Soft-Core obturation: an in vitro evaluation.   Endodontics and Dental Traumatology 15 , 239-43.
Endotec Thermal Endodontic Condenser System (1986) The Warm Lateral Condensation   Technique Clinical Manual . Dentsply Int. Inc. Milford DE, USA: The LD Caulk   Division.
  Gurney BF, Best EJ, Gervasio G (1971) Physical measurements on gutta-percha.   Oral Surgery, Oral Pathology and Oral Medicine 32 , 260-70.
  Gutmann JL, Saunders WP, Saunders EM, Nguyen L (1993) An assessment of the   plastic Thermafil obturation technique. Part 2. Material adaptation and sealability.   International Endodontic Journal 26 , 179-83.
  Hovland EJ, Dumsha TC (1985) Leakage evaluation in vitro of the root canal   sealer cement Sealapex. International Endodontic Journal 18 , 179-82.
  Johnson WB (1978) A new gutta-percha technique. Journal of Endodontics 4 , 184-8.  
  Kennedy WA, Walker WA, Gough RW (1986) Smear layer removal effects on apical   leakage. Journal of Endodontics 12 , 21-7.
  Kontakiotis EG, Wu M-K, Wesselink PR (1997) Effect of sealer thickness on   long-term sealing ability: a 2 year follow-up study. International Endodontic   Journal 30 , 307-12.
  Saunders WP, Saunders EM (1994a) Coronal leakage as a cause of failure in   root canal therapy: a review. Endodontics and Dental Traumatology 10 , 15-8.  
  Saunders WP, Saunders EM (1994b) Influence of the smear layer on the coronal   leakage of Thermafil and laterally condensed gutta-percha root fillings with   glass ionomer sealer. Journal of Endodontics 20 , 155-8.
  Schilder H (1967) Filling root canals in three dimensions. Dental Clinics   of North America 11 , 723-44.
  Tagger M, Tamse D, Katz A, Korzen BH (1984) Evaluation of the apical seal   produced by a hybrid root canal filling method, combining lateral condensation   and thermatic compaction. Journal of Endodontics 10 , 299-303.
  Taylor JK, Jeansonne BG, Lemon RR (1997) Coronal leakage: effects of smear   layer, obturation technique, and sealer. Journal of Endodontics 23 , 508-12.  
  Wu M-K, Wesselink PR, Boersma J (1995) A 1-year follow-up study on leakage   of four root canal sealers at different thicknesses. International Endodontic   Journal 28 , 185-9.
  Wu MK, Wesselink PR (1993) Endodontic leakage studies reconsidered. Part   I: Methodology, application and relevance. International Endodontic Journal   26 , 37-43.
  Yee FS, Marlin J, Krakow AA, Grin P (1977) Three-dimensional obturation of   the root canal injection molded, thermoplasticized dental gutta-percha.   Journal of Endodontics 3 , 168-74.