Micro-CT evaluation of residual material in canals filled with Activ GP or GuttaFlow following removal with NiTi instruments

Transcription

1 doi: /j x Micro-CT evaluation of residual material in canals filled with Activ GP or GuttaFlow following removal with NiTi instruments M. J. Roggendorf 1, M. Legner 2, J. Ebert 3, E. Fillery 2, R. Frankenberger 1 & S. Friedman 4 1 Department of Operative Dentistry and Endodontology, University Medical Center, Giessen and Marburg, Germany; 2 Oral Microbiology, Faculty of Dentistry, University of Toronto, Toronto, ON, Canada; 3 Dental Clinic 1 Operative Dentistry and Periodontology, University Hospital Erlangen, Erlangen, Germany; and 4 Department of Endodontics, Faculty of Dentistry, University of Toronto, Toronto, ON, Canada Abstract Roggendorf MJ, Legner M, Ebert J, Fillery E, Frankenberger R, Friedman S. Micro-CT evaluation of residual material in canals filled with Activ GP or GuttaFlow following removal with NiTi instruments. International Endodontic Journal, 43, , Aim To assess the efficacy of removing Activ GP or GuttaFlow from canals using NiTi instruments. Methodology Root canals in 55 extracted premolars were prepared to apical size 40, 0.04 taper. The teeth were imaged with micro-ct, and 30 teeth selected that had consistent apical size and taper of the shaped canals. They were randomly assigned to root filling with either the glass-ionomer-based ActivGP system (n = 15) or the polyvinylsiloxane-based GuttaFlow system (n = 15). After 2 weeks, canals were retreated stepwise with size EndoSequence 0.04 taper instruments. Micro-CT scans (8 lm) were taken after use of each instrument to detect root filling residue in the coronal, middle and apical segment, and the retreatment time recorded. Residue, expressed as percentage of canal surface area, was compared between groups with t-tests, and within groups with repeated measures anova and Bonferroni-adjusted pairwise comparisons. Retreatment time was analysed with one-way anova. Results The percentage of sealer residue-coated canal surface was consistently highest (P < 0.001) in the apical third of canals, and it did not differ significantly between the two root filling groups. Stepwise enlargement from size 40 to 50 significantly decreased the amount of sealer residue in both groups (P < 0.001). Retreatment time did not differ significantly between groups. Conclusions Both root fillings with ActivGP and GuttaFlow were removed with nickel-titanium rotary instruments. Enlargement of canals up to two sizes beyond the pre-retreatment size was necessary to minimize the amount of sealer remaining. Keywords: endodontic retreatment, Root filling. Received 14 June 2009; accepted 9 November 2009 Introduction Epidemiologic studies have reported that over 40% of root filled teeth in the population present with periradicular radiolucency consistent with endodontic Correspondence: Dr med. dent. Matthias Johannes Roggendorf, Department of Operative Dentistry and Endodontology, Medical Center for Oral and Maxillofacial Sciences, University Hospital Marburg, Georg-Voigt-Str. 3, D Marburg, Germany (Tel.: ; fax: ; dr.roggendorf@mac.com). infection (Georgopoulou et al. 2005, Sunay et al. 2007, Chen et al. 2007). In the majority of the affected teeth, the treatment of choice is orthograde retreatment of the root canal system (Friedman 2002). The immediate aim of retreatment is to disinfect the root canal space to create conditions conducive to periradicular healing; therefore, the retreatment procedure requires re-access of the root canal system and the complete removal of the existing root filling (Stabholz & Friedman 1988). For this reason, root filling materials should be retreatable (Grossman 1970). 200 International Endodontic Journal, 43, , 2010 ª 2010 International Endodontic Journal

2 Roggendorf et al. Retreatment efficacy of Activ GP and GuttaFlow Conventional root fillings consist of a core material, primarily compacted gutta-percha, combined with a root canal sealer (Grossman 1970). Commonly used sealers include zinc oxide eugenol-based cements, glass-ionomer cements, resins and silicones. To remove the root filling materials, hand, rotary and ultrasonic instruments can be used (Friedman et al. 1992, 1993, Moshonov et al. 1994, Wolcott et al. 1999, Bramante & Betti 2000, Imura et al. 2000, Sae-Lim et al. 2000, Ferreira et al. 2001, Ezzie et al. 2006, Schirrmeister et al a d, Zmener et al. 2006, Barletta et al. 2007, 2008, Hassanloo et al. 2007, Hammad et al. 2008, Iizuka et al. 2008, Somma et al. 2008, Tasdemir et al. 2008a,b) with or without solvents or heat to facilitate the procedure (Tamse et al. 1986, Wilcox et al. 1987, Wolcott et al. 1999). Regardless of the retreatment technique and the materials removed, studies have invariably reported the presence of root filling residue on the canal walls (Wilcox et al. 1987, Friedman et al. 1992, 1993, Moshonov et al. 1994, Masiero & Barletta 2005, Hassanloo et al. 2007). The amount of the residue can be reduced when canal enlargement during retreatment exceeds the initial preparation size (Hassanloo et al. 2007). The efficacy of removing the root filling materials during retreatment has been assessed mostly by methodologies that destroy the specimens. These methodologies include horizontal (Wolcott et al. 1999) or vertical splitting of the examined teeth (Wilcox et al. 1987, Friedman et al. 1992, 1993, Moshonov et al. 1994, Barrieshi-Nusair 2002, Zmener et al. 2006, Hassanloo et al. 2007, Giuliani et al. 2008, So et al. 2008), or clearing to render the teeth transparent (Schirrmeister et al a d, Gu et al. 2008). In fewer studies, the examined specimens were imaged with either conventional radiographs (Ferreira et al. 2001, Masiero & Barletta 2005, Gergi & Sabbagh 2007, Zanettini et al. 2008) or micro-ct (Barletta et al. 2007, 2008, Hammad et al. 2008). The latter, in particular, can provide detailed imaging of the root canals (Peters et al. 2001, 2003, Hübscher et al. 2003, Paqué et al. 2005) and their content (Jung et al. 2005, Barletta et al. 2007, 2008). The non-destructive imaging methods are best suited for the assessment of stepwise enlargement of the canal space during retreatment (Hassanloo et al. 2007). Recently, two new root filling systems have been advocated for use without compaction. The Activ GP Precision Obturation System (Brasseler USA, Savannah, GA, USA) comprises a hand mixed glass-ionomer cement sealer and gutta-percha that contains silanated glass-ionomer cement and is coated with glass-ionomer cement particles at a thickness of 2 lm. The GuttaFlow system (Coltène Whaledent, Langenau, Germany) comprises a polyvinylsiloxane-based sealer that contains small gutta-percha fillers and nano-silver particles. It is syringed into the root canal or applied with a lentulo spiral and combined with a single gutta-percha cone. The manufacturers of both systems advocate the single-cone technique to be used mostly in round canals, to reduce the amount of sealer in the fillingdentine interface, but the quick setting reaction of both systems precludes the use of compaction techniques. Consequently, root fillings performed with these new systems may contain thicker layers of sealer than those performed with conventional compaction techniques. Therefore, the purpose of this study was to assess the efficacy of stepwise retreatment of Activ GP and GuttaFlow root canal fillings, using non-destructive micro-ct imaging to measure the amount of root filling residue. Material and methods Sample size Previous studies (Friedman et al. 1992, 1993, Moshonov et al. 1994, Hassanloo et al. 2007) have reported differences of 25% in the amount of residue when retreating canals filled with different materials. Assuming a difference of 25% between the materials assessed in this study, the sample size was calculated as 15 teeth per group for an analysis with 80% power and 5% significance level. Root specimens Fifty-five extracted human pre-molars with straight bulky roots had all adhering soft tissue removed and were stored in 0.5% chloramine T solution (Merck Eurolab, Darmstadt, Germany) until used. The pulp chambers and canals were accessed conventionally and the tooth crowns sectioned 10 mm from the root-end to standardize the length of the specimens. A size 15 reamer (Dentsply Maillefer, Ballaigues, Switzerland) was inserted through the canal to 1 mm beyond the apical foramen to establish apical patency, and the working length established 1 mm shorter than the length of the extruded instrument. Canals were flared coronally with no. 3 Gates-Glidden drills (Dentsply Maillefer) and shaped in a crown-down sequence with K taper instruments (Sybron Endo Inc., Newport ª 2010 International Endodontic Journal International Endodontic Journal, 43, ,

3 Retreatment efficacy of Activ GP and GuttaFlow Roggendorf et al. Beach, CA, USA) to size 40 at the working length. Apical patency was maintained throughout with size 15 reamers. Canals were irrigated intermittently with 2 ml of 2.5% sodium hypochlorite, and finally rinsed with 5 ml of 17% EDTA (Smear Clear; Sybron Endo, Glendora, CA, USA), 5 ml of 2.5% sodium hypochlorite and 5 ml of sterile saline. The tooth specimens were stored under humid conditions (37 C, 95% humidity) until scanned. The tooth specimens were embedded in silicone blocks (Provil Novo; Heraeus Kulzer, Inc., South Bend, IN, USA) and scanned with a cone-beam micro-ct unit (General Electric Locus SP, Model MS 8X-90/CAB; General Electric Medical Systems, London, ON, Canada) to identify specimens with round canal cross-sections, such as are recommended for root filling with the single-cone technique with both systems. Scans were evaluated for regularity of canal shape and presence of an apical constriction. Twenty-four specimens where canals had fins, oval cross-sections or no apical stop were discarded (Fig. 1a), leaving 30 specimens with regular, round canals for evaluation (Fig. 1b). One additional tooth was instrumented for replacement as described later. Root filling Specimens were randomly assigned to two root filling groups (n = 15/group). In canals filled with the Activ GP system (Activ GP Cement, Liquid LOT , Powder LOT B), a size 40, 0.04 taper Activ GP gutta-percha cone (LOT No ) was fitted in each canal with tug back at working length. It was then used to apply the Activ GP sealer to the canal and finally seated to length. In canals filled with the GuttaFlow system (LOT ), a size 40, 0.04 taper gutta-percha cone (LOT ; Hygenic SpectraPoint, Coltène Whaledent, Cuyahogen Falls, OH, USA) was pre-fitted in the canal with tug back. The GuttaFlow capsule was mixed for 30 sec in a triturator (Ultramat 2; Southern Dental Industries, Austin, TX, USA) and used to directly apply the sealer into the canal at working length. The pre-fitted master cone was thinly coated with GuttaFlow and seated to length. All specimens had excess gutta-percha removed with a hot excavator and canals coronally sealed with Cavit (Premier Dental, Plymouth Meeting, PA, USA). They were subsequently stored for 2 weeks (37 C, 95% humidity) until retreatment. All specimens were prepared and filled by the same operator (M.R.) to reduce inter-operator variation. Specimens were Figure 1 Pre-filling micro-ct images of tooth specimens after root canal enlargement to size 40 taper (a) The tooth was deemed unsuitable because of fins and irregular canal shape and was excluded from the study. (b) The tooth was deemed suitable because of round, regular canal shape and was included in the study. radiographically exposed from the buccal and lingual aspects to verify the apical extent and homogeneity of the root filling. One specimen from the GuttaFlow group was discarded because of voids, and replaced by a new one. Retreatment A no. 3 Gates-Glidden drill (Dentsply Maillefer) was used to remove the coronal 2 mm of the root canal filling in all specimens. Subsequently, canals were retreated in three steps. In the first step, EndoSequence 0.04 taper instruments (Brasseler USA) were used in a crown-down sequence. The instruments were operated in an electric motor handpiece (Kerr K3 ETCM; Sybron Endo) at 500 rpm using moderate apical pressure, until working length was reached with a size 40 instrument. Canals were irrigated with 2.5% sodium hypochlorite between successive instruments, and finally rinsed with 5 ml of 17% EDTA (Smear Clear; Sybron Endo), 5 ml of 2.5% sodium hypochlorite and (a) (b) 202 International Endodontic Journal, 43, , 2010 ª 2010 International Endodontic Journal

4 Roggendorf et al. Retreatment efficacy of Activ GP and GuttaFlow 5 ml of sodium chloride to wash out debris and root filling residue and to avoid precipitation of irrigants on the canal surface during moist storage of the teeth until the specimens were scanned again. Canals were dried with paper points, and the specimens scanned with micro-ct (see below). In the second step, canals were further retreated to working length with a size 45, 0.04 taper EndoSequence instrument, rinsed and dried as described previously, and scanned again with micro-ct. In the third step, canals were retreated to size 50, 0.04 taper and the specimens scanned. Instruments used for retreatment were replaced after three uses. For all retreated specimens, the time to reach working length with the size 40, 0.04 taperinstrument was recorded to the nearest second with a stopwatch. The times needed for the subsequent steps of retreatment, and the total time including instrumentation, irrigation and drying, were also recorded. and each retreatment step (enlargement to size 40, 45 and 50) was recorded. Analysis Statistical analysis was computed with SPSS 16.0 for Windows (SPSS Co., Chicago, IL, USA). Differences in PRCS between the two root filling groups were analysed with series of t-tests, for each canal segment at every retreatment step. The differences within each group amongst the coronal, middle and apical segments were analysed with repeated measures anova and Bonferroni-adjusted pairwise comparisons. The differences between successive retreatment steps at each canal segment within each group were analysed using Friedman tests and Wilcoxon tests. The difference in retreatment time between the two groups was analysed using one-way anova. The level of significance was set at 5% for all statistical tests. Micro-CT imaging The specimens were scanned using a Locus SP micro- CT unit (General Electric). They were mounted in silicone putty (Provil Novo; Heraeus Kulzer, Inc.) that allowed precise placement in the scanning unit without interference with the scanning process. To view the specimens canal shapes before root filling, a short scan setting was used (80 kv, 80 la, 720 slices, 200, 16 lm resolution). For the post-retreatment scans, the full scan setting was used (80 kv, 80 la, 900 slices, 200, 8 lm resolution). A total of 145 scans were performed (55 preliminary scans and 30 post-retreatment scans for each of the three retreatment steps), captured digitally and stored onto a Unix server (General Electric). Results None of the EndoSequence instruments fractured or showed signs of deformation during the retreatment procedure. Test imaging of a few root filled specimens allowed easy discrimination of dentine, gutta-percha and sealer (Fig. 2). Similarly after retreatment, root filling residue was easily discriminated from the root dentine as well as pulpal and dentinal debris. Root filling residue was detected in both groups and was Outcome assessment Scans were segmented, and residual root filling materials visualized using Amira 4.1 software (Mercury Systems, Berlin, Germany). Computed raw images were converted into DICOM files. The root canals were separated into three segments (apical, middle, coronal) of 3 mm each. The amount of root filling residue on the canal walls was quantified with OsiriX software (The OsiriX Foundation, Geneva, Switzerland). Areas covered with residue were expressed as percentage of the canal wall surface area. The mean percentage of residue-coated surface (PRCS) for each canal segment Figure 2 Micro-CT image of a tooth specimen with Activ GP root filling. The root dentine (green), gutta-percha (orange) and sealer (yellow) are clearly detectable. Note the relatively uniform thickness of the sealer all along the gutta-percha cone. ª 2010 International Endodontic Journal International Endodontic Journal, 43, ,

5 Retreatment efficacy of Activ GP and GuttaFlow Roggendorf et al. Table 1 Percentage of residue-coated canal wall surface (PRCS) in root canals filled with Activ GP or GuttaFlow, after stepwise retreatment with apical enlargement to sizes 40/0.04, 45/0.04 and 50/0.04 Mean PRCS per retreatment step (SD) Group Canal segment Step 1 (size 40) Step 2 (size 45) Step 3 (size 50) Activ GP Apical 3.37 (3.78) 0.57 (0.73) 0.16 (0.34) Middle 1.43 (2.40) 0.69 (1.07) 0.24 (0.79) Coronal 0.29 (0.62) 0.16 (0.61) 0.05 (0.19) Total 1.34 (1.38) 0.43 (0.57) 0.14 (0.34) GuttaFlow Apical 2.46 (3.80) 0.62 (0.84) 0.24 (0.60) Middle 0.94 (2.53) 0.57 (1.97) 0.24 (0.90) Coronal 1.14 (1.92) 0.17 (0.26) 0.05 (0.20) Total 1.36 (1.41) 0.41 (0.77) 0.16 (0.35) invariably identified as sealer without any residual gutta-percha being detected. Table 1 presents the mean PRCS values for both root filling groups in the three canal segments at each retreatment step. The differences between the two groups were not statistically significant (t-test) for either canal segment at each retreatment step. After the first retreatment step with size 40, 0.04 taper instruments, the differences amongst the apical, middle and coronal canal segments were significant (repeated measures anova, P < 0.001) in each group, with the apical segments having the highest mean PRCS (Fig. 3a). A consistent reduction in the mean PRCS was observed from one retreatment step to the next in both groups at each of the canal segments (Fig. 3b). The results for the three retreatment steps were significantly different (Friedman test, P < 0.001). The difference between steps 1 and 3 was significant for all regions and the whole root canal surface (Wilcoxon tests, P < 0.001), whilst the differences between step 2 and the other two steps were partially not significant. The mean time to complete each of the three retreatment steps is summarized in Table 2. The time required for the first step of retreatment was significantly (one-way anova, P = 0.015) shorter in canals filled with GuttaFlow than in canals filled with Activ GP. The time required to complete the two subsequent steps of retreatment was shorter and did not differ significantly between both root filling groups. The total time required for all three steps of retreatment also did not differ significantly between the two groups. Discussion Several models have been to assess the amount of residual root filling materials after retreatment. Destructive models (Friedman et al. 1992, 1993, Moshonov et al. 1994, Schirrmeister et al a d) limit the assessment of residue to one end-point of the retreatment process. Non-destructive radiographic imaging (Masiero & Barletta 2005, Gergi & Sabbagh 2007) allows the assessment of residue at different stages of retreatment; however, it yields qualitative rather than quantitative outcomes (Hassanloo et al. 2007). Hassanloo et al. (2007) recently introduced a disassembly-reassembly model whereby roots embedded in acrylic blocks were split vertically, reassembled to carry out the retreatment procedures, and disassembled to measure the amount of residue by direct digital imaging. This model allowed quantitative, stepwise assessment of residue after progressive enlargement of the canal (Hassanloo et al. 2007), but it presented various technical challenges. Recently, Barletta et al. (2007) used micro-ct imaging to assess the efficacy of retreatment. This methodology allows non-destructive, stepwise assessment by repeated scans of the same root specimen. By coupling the finely detailed three-dimensional imaging provided by micro-ct with appropriate software, areas of interest within the canal lumen and root structure such as the volume of removed dentine during cleaning and shaping (Peters et al. 2001, 2003, Hübscher et al. 2003, Paqué et al. 2005), voids within root fillings (Jung et al. 2005, Hammad et al. 2009) or root filling residue after retreatment (Barletta et al. 2007, 2008), can be quantified with precision and viewed in three dimensions. Indeed, in the present study, the micro-ct images allowed the dentine, sealer and gutta-percha to be clearly distinguished from each other (Fig. 2), and they readily displayed residual sealer on the canal wall after retreatment (Fig. 3a,b). The sealer-coated areas were expressed as percentage of the total canal wall surface at each root section, similar to several previous reports (Wilcox et al. 1987, Betti & Bramante 2001, Masiero & Barletta 2005, Hassanloo et al. 2007, Tasdemir et al. 2008b). Clinicians may better appreciate 204 International Endodontic Journal, 43, , 2010 ª 2010 International Endodontic Journal

6 Roggendorf et al. Retreatment efficacy of Activ GP and GuttaFlow (a) Figure 3 Micro-CT images of a tooth specimen filled with Activ GP, after the first and last steps of retreatment. (a) After retreatment with size 40 taper 0.04 instrument, residue of sealer is predominantly present in the apical segment of the root canal. Less sealer residue is present in the middle and coronal segments. (b) After retreatment with size 50 taper 0.04 instrument, considerably less residue of sealer is present in the apical segment of the canal, whilst no sealer residue can be seen in the middle and coronal segments (Residue within the apical foramen was excluded from analysis). the extent of residual sealer expressed in this manner than reported as absolute surface area. Notwithstanding the clear benefits of the micro-ct model, in the present study, it proved to be extremely time consuming. Pilot testing revealed less detail in scans of 16 lm resolution compared with 8 lm resolution. The 55 pre-filling scans were performed at 16 lm (Barletta et al. 2007, 2008, Hammad et al. 2008, 2009) whilst the 90 post-retreatment scans were performed at 8 lm, requiring much more time for each scan. The total scanner-time of nearly 300 h resulted in (b) a drastically increased cost when compared with previous models, casting into doubt the feasibility of future use of this state-of-the-art methodology for similar studies. In contrast to most previous studies where root fillings were compacted laterally (Friedman et al. 1992, Moshonov et al. 1994, Imura et al. 2000, Barrieshi- Nusair 2002, de Oliveira et al. 2006, Zmener et al. 2006, Hassanloo et al. 2007) or vertically (Ferreira et al. 2001, Barletta et al. 2007, 2008), canals in the present study were filled with sealer and a single cone each. Recognizing the better suitability of the singlecone technique for round canals, only roots with round canal cross-sections after shaping were selected. Care was taken to verify the homogeneity and proper apical extent of the root fillings. The root fillings were allowed to set for 2 weeks, to exclude the possibility of incompletely set sealer at the time of retreatment. Removal of the root fillings was performed with engine-driven instruments, shown to be effective in previous studies (Sae-Lim et al. 2000, Hülsmann & Bluhm 2004, Masiero & Barletta 2005, De Carvalho Maciel & Zaccaro Scelza 2006, Schirrmeister et al. 2006c,d, Hassanloo et al. 2007, Saad et al. 2007). Retreatment with engine-driven instruments has become increasingly popular, following a similar trend in the use of these instruments for cleaning and shaping. Although specific instruments have been developed for retreatment, the design of the common engine-driven nickel-titanium instruments is well suited for drawing root filling debris coronally (Sae-Lim et al. 2000, Masiero & Barletta 2005, De Carvalho Maciel & Zaccaro Scelza 2006, de Oliveira et al. 2006, Schirrmeister et al. 2006c,d, Hassanloo et al. 2007). In the coronal 2 mm of the canals, Gates-Glidden drills were used to facilitate the retreatment procedure (Friedman et al. 1990, Hassanloo et al. 2007). The Gates-Glidden drills could be taken as far as 5 6 mm into the root canals leaving a minimal amount of root filling to remove (Schirrmeister et al. 2006b). However, because such procedures can result in excessive removal of root dentine and the subsequent weakening of the root, it was considered unfavourable. Also, in curved canals Gates-Glidden drills can only be taken 2 3 mm apically without the risk of canal transportation (Schirrmeister et al. 2006b). The remaining root filling was removed with EndoSequence instruments, primarily to conform to the retreatment guidelines for the Activ GP root filling system. Endo- Sequence instruments were previously shown to effectively remove gutta-percha without the use of solvent or heat (Schirrmeister et al. 2006d). Use of solvent, therefore, was not deemed necessary in this study. ª 2010 International Endodontic Journal International Endodontic Journal, 43, ,

7 Retreatment efficacy of Activ GP and GuttaFlow Roggendorf et al. Table 2 Time required for completion of retreatment steps with apical enlargement to sizes 40/0.04, 45/0.04 and 50/0.04 in canals filled with Activ GP or GuttaFlow. The measured time included instrumentation until reaching the working length, instrument changes, irrigation and drying of the root canal Mean retreatment time (s) (SD) Group Step 1 (size 40) Step 2 (size 45) Step 3 (size 50) Total time Activ GP Sealer (45.3) (16.0) (15.2) (54.3) GuttaFlow (30.0) (28.8) (10.4) (34.2) P-value As in the previous studies (Wilcox et al. 1987, Friedman et al. 1992, 1993, Moshonov et al. 1994, Imura et al. 2000, Baratto Filho et al. 2002, Masiero & Barletta 2005, Kosti et al. 2006, Barletta et al. 2007, Hassanloo et al. 2007, Saad et al. 2007, Hammad et al. 2008), neither retreated material in the present study could be removed completely. Both Activ GP and GuttaFlow could be removed to a similar extent without significant differences in the amount of sealer residue. Interestingly, the <4% of sealer-covered canal wall surface in the present study was considerably smaller than the 6 36% reported in the previous studies (Wilcox et al. 1987, Betti & Bramante 2001, Masiero & Barletta 2005, Hassanloo et al. 2007, Tasdemir et al. 2008b). Although a direct comparison cannot be drawn with the qualitative reports on retreatment of the glass-ionomer cement sealer Ketac- Endo (Friedman et al. 1992, 1993, Moshonov et al. 1994) or Endion (Kosti et al. 2006), it appeared that the amount of residue in those studies also exceeded that observed in the present study. A previous study on the removal of GuttaFlow (Tasdemir et al. 2008b) found more sealer to be present after retreatment. Several reasons can be suggested for the better retreatment efficacy in this study than in the previous ones. The first reason may be the amenability of the retreated materials to removal. GuttaFlow is a siliconebased material that may adapt closely but does not adhere to the canal wall dentine. It also remains soft after setting. In contrast, the glass-ionomer cementbased Activ GP sealer is expected to bond to the canal walls (Saleh et al. 2002) and to fully solidify after setting (Fisher et al. 2007). Nevertheless, the bond strength of the Activ GP sealer to dentine is low (Saleh et al. 2002). Also, as any glass-ionomer cement, set Activ GP sealer may be brittle, a property that may facilitate its removal with sharp instruments. Thus, the second reason for the effective retreatment may be the cutting nature of the EndoSequence instruments used. When operated at 500 rpm, the instruments effectively withdrew the core gutta-percha cone so that virtually no gutta-percha residue was found in any of the canals. The third reason for effective removal may be the taper of the retreating instruments, matching that of the filled canal shapes. Thus, removal of the sealer may have been enhanced relative to what would be achieved if the retreating instruments had a smaller taper. The fourth reason for the effective retreatment could be the presence of a relatively uniform layer of sealer in the retreated canals (Fig. 2). Using only canals with a round cross-section, recommended for filling with a single cone and sealer, most likely helped to reduce the thickness of the sealer. In addition, guttapercha cones that matched the taper of the prepared canals were used, further reducing the gaps occupied by sealer. Also, the use of the single-cone technique may have resulted in less lateral displacement of the sealer into possible canal irregularities, such as occurs when root fillings are condensed (Blum et al. 1998). After the first step of retreatment with instrument size 40, 0.04 taper, the amount of sealer residue was the largest in the apical canal level, in agreement with the previous reports (Friedman et al. 1992, 1993, Moshonov et al. 1994, Masiero & Barletta 2005, Hassanloo et al. 2007). Further canal enlargement by one size beyond the filled canal size (to size 45, 0.04 taper) already resulted in significant reduction in the amount of sealer residue. Additional enlargement by two sizes (to size 50, 0.04 taper) significantly reduced the amount of residual sealer (Fig. 3a,b), with the greatest improvements, 10- to 20-fold, observed in the apical canal level, and more moderate 4- to 6-fold reductions observed in the middle and coronal canal levels (excluding the coronal level in the GuttaFlow group, where an excessive amount of residual sealer was noted in the coronal level after the first step of retreatment). These findings corroborated those reported by Hassanloo et al. (2007), but the effect of further enlargement was far greater in the present study, possibly because of the reasons suggested earlier. Taken together, these results emphasized the critical importance during retreatment of canal enlargement by at least two sizes beyond the filled canal size, even 206 International Endodontic Journal, 43, , 2010 ª 2010 International Endodontic Journal