Methodology.
Forty-eight single-rooted human teeth, with fully formed apices and straight root canals were used. The root canals were accessed and instrumented using a stepback technique with H-files. They were randomly assigned to four groups and subsequently filled with a combination of lateral and vertical condensation of gutta-percha and one of the following sealers: Roth 811, AH26, Endion and Roekoseal. The root fillings were removed 1 year later, using either H-files in combination with GG drills or the ProFile Ni-Ti system. Teeth were then grooved longitudinally and split. The amount of gutta-percha and sealer remaining on the root canal walls was traced and scored visually with the aid of a stereomicroscope. The scores were analysed and statistically compared with the Kruskal-Wallis test between the ProFile and H-file groups, as well as among the four sealer subgroups. Two samples from each group were studied under the scanning electron microscope to enhance inspection of canal walls and remaining material.

Results.
Sealer remnants were observed with both techniques mainly in the middle and apical third of the root canal. The ProFile system and the H-files were associated with similar amounts of remaining filling material (P > 0.05). In the cervical third of the root canal all sealer remnants were removed with both techniques. In the middle and apical third AH26 was associated with a statistically significant greater quantity of remnants on the root canal walls with both removal techniques (P < 0.05). Endion, Roth 811 and Roekoseal were associated with approximately the same amount of filling material in the middle third of the root canal (P > 0.05), whereas in the apical third Endion was associated with significantly more remnants of filling material than the other two sealers with either ProFile or H-files (P < 0.05).

Introduction.
Nonsurgical root canal retreatment is the treatment of choice when root-filled teeth are associated with disease. An important step in retreatment is the removal of existing filling material to regain access to the entire canal, expose remnants of necrotic tissue and microorganisms and facilitate their removal.
Several techniques, employing various instruments have been proposed for removing root filling materials (Wilcox 1989, Friedman et al. 1992, Moshonov et al. 1994, Hu?lsmann & Stotz 1999, Farge et al. 1998, Betti & Bramante 2001, Barrieshi-Nusair 2002, Viduc?ic? et al. 2003). Heat and a variety of solvents such as, chloroform, eucalyptol, orange oil, xylol and halothane, have also been used to soften gutta-percha and facilitate its removal. However, it has been reported that none of the retreatment methods produces completely clean root canal walls (Wilcox et al. 1987, Imura et al. 2000, Ferreira et al. 2001, Barrieshi- Nusair 2002).
The difficulty in gutta-percha removal is directly related to canal preparation and filling techniques, the type of sealer used as well as the time elapsed since the original treatment (Lambrianidis 2001). The majority of the laboratory studies examined teeth with root fillings performed a short time before retreatment, without investigating all representative types of sealers (Friedman et al. 1992, Imura et al. 2000, Barrieshi- Nusair 2002).
The aim of this study was to compare the efficacy of ProFile rotary nickel-titanium (Ni-Ti) endodontic instruments (Dentsply Maillefer, Baillagues, Switzerland) and Hedstroem-files (H-files; Antaeos, Vereinigte Dentalwerke GmbH & Co., Munich, Germany), used in combination with Gates-Glidden drills (GG; Antaeos) during the removal of root fillings comprising of guttapercha and one of the four representative sealers.

Materials and methods.
Forty-eight single-rooted human teeth were used. The teeth were stored for 2 days immediately after extraction at room temperature in 3% sodium hypochlorite to dissolve organic debris. Subsequently, they were cleaned with an ultrasonic scaler and washed with distilled water to remove calculus and soft-tissue debris and then immersed in 10% formalin solution until use. Criteria for tooth selection were: existence of a single root canal, no visible root caries, fractures or cracks on examination with a ·4 magnifying glass, no signs of internal or external resorption or calcification and a fully formed apex. Only roots with less than 5_ of curvature according to Schneider (1971) were included. Preoperative mesiodistal and buccolingual radiographs of each root were taken to confirm the existence of a single straight canal. Only root canals in which the first file that fitted at the apex was a size 15 were included.
Instrumentation was performed by a single operator. Access cavities were prepared and a size 10 H-file was introduced into the canal until it was visible at the apical foramen. The working length was determined by subtracting 1 mm from this measurement. This same file was used during the root canal preparation to maintain patency of the canal. Root canal preparation was performed using H-files with a stepback technique (Ingle et al. 2002). Instrumentation was standardized with a size 30 H-file reaching full working length, a size 55 file 5 mm coronally and a final coronal flaring with GG sizes 3 and 2. A 15% EDTA gel was used as a chelating agent. The canals were irrigated between successive instruments with 5 mL of 2.5% NaOCl delivered with 27-gauge needle tips placed passively into the canal, as far as 3 mm from the apical foramen without binding.
Samples were wrapped in aluminium foil, embedded in acrylic blocks and randomly divided into four groups A, B, C and D, comprising 12 teeth each. The sealers were prepared according to the manufacturer's instructions. The root canal walls were dried with paper points and then coated with sealer using a lentulo to 1-2 mm from the apical constriction. The root filling was performed with a combination of lateral and vertical condensation techniques using gutta-percha and either Roth 811 (Roth International, Chicago, IL, USA), AH26 (Dentsply Maillefer), Endion (Voco Dental Products, Cuxhaven, Germany) or RS Roekoseal Single Dose (Roeko Dental Products, Langenau, Germany) for groups A, B, C and D respectively.
The quality of the root filling was evaluated with mesiodistal and buccolingual radiographs and was considered satisfactory when no voids could be detected. The access cavity was sealed temporarily (Cavit-G, Espe, Seefeld, Germany) and teeth were stored in a humidity chamber (100% humidity and 37 _C) for 1 year. The teeth in all four groups were then randomly divided in two subgroups comprising six teeth each.
In subgroups A1, B1, C1 and D1, the filling was removed using H-files in combination with GG drills, whereas in subgroups A2, B2, C2 and D2, the ProFile system was used for gutta-percha removal. In subgroups A1, B1, C1 and D1, GG size 3 and subsequently 2 were used to remove the coronal guttapercha and create a reservoir for solvent. Increments of chloroform (0.1 mL) were placed in the canal to soften the gutta-percha; only up to two applications were used. The gutta-percha was removed using H-files, with a circumferential filling motion and copious irrigation with 2.5% NaOCl. Each stainless steel file was used five times and then discarded. The last instrument introduced to the full working length was a size 35 H-file, one size larger than the master apical file used in the original instrumentation.
In subgroups A2, B2, C2 and D2, gutta-percha was removed using the ProFile system in a crowndown technique. Orifice Shapers No. 4 (size 50, taper 0.07) and No. 3 (size 40, taper 0.06) were used to remove the coronal gutta-percha and create a reservoir for solvent. Increments of chloroform (0.1 mL) were placed in the canal to soften gutta-percha; only up to two applications were used. ProFile instruments sizes 30 and 25, with a 0.06 taper were introduced up to two-third within the root canal followed by sizes 30 and 25, with a 0.04 taper, each progressing to the full working length; 15% EDTA gel was used as a chelating agent and copious irrigation with 2.5% NaOCl was performed during the procedure. Ni-Ti instruments were used with a speed-reduction handpiece (KaVo Dental Gmbh & Co., Biberach, Germany) and a torque control motor (ATR Tecnika, Pistoia, Italy). The instruments were used with a continuous axial motion, with a range of 2-3 mm, exerting slight apical pressure. Each Ni-Ti instrument was used five times, and then discarded. The last instrument introduced to the full working length was a size 30, with a 0.04 taper.
Criteria for the assessment of removal of the root filling material were the detection of smooth canal walls and absence of gutta-percha or sealer on the last instrument to be used. Magnifying loupes ·4.5 (Carl Zeiss, Oberkochen, Germany) were used to enhance vision.
The teeth were grooved buccolingually with a diamond disc and sectioned longitudinally. Each half was divided into coronal, middle and apical thirds. The scoring was carried out visually by three independent pre-calibrated examiners under the stereomicroscope at ·4 magnification for each half in the coronal, middle and apical third separately. In case of disagreement, all three examiners repeated the scoring together. Evaluation scales used were: score 0, no gutta-percha and no sealer; score 1, debris of sealer; score 2, debris of sealer and gutta-percha; score 3, severe debris of sealer and gutta-percha. Mean values were calculated from the scores gathered for all 12 halves of each group (A1, B1, C1, D1, A2, B2, C2 and D2) in the cervical, middle and apical thirds and in the total root canal surface (Table 2). Statistical analysis of the scores was performed with the Kruskal-Wallis test.