RESUMO
AIM: To compare the amount of apically extruded debris and of remaining filling material during the removal of root canal filling material using three rotary NiTi retreatment instruments or Hedström files. METHODOLOGY: Ninety-six severely curved human molars of both jaws were selected. The root canals were prepared to size X2 (tip size 25, .06 taper) using the ProTaper Next system (Dentsply Sirona, Ballaigues, Switzerland), filled with gutta-percha and AH Plus sealer (Dentsply De Trey, Konstanz, Germany) and then randomly divided into four experimental groups (n = 24 each) with two subgroups of maxillary and mandibular teeth each. An experimental model was used as a phantom head to simulate the upper and lower jaws. The root filling materials were removed with one of the following files using a crown-down preparation technique: I. Hedström files (H-files) (VDW, Munich, Germany), II. R-Endo (Micro-Mega, Besançon, France), III. Reciproc (VDW) and IV. ProTaper Universal Retreatment system (PTU-R) (Dentsply Maillefer). Apically extruded material was collected in vials, which were weighed with a microbalance (10-5 g) before and after the retreatment. The area of residual filling material in the coronal, middle and apical root level was assessed using digital analysis. These data were analysed statistically using ANOVA and the Kruskal-Wallis test. RESULTS: Reciproc was associated with significantly less extruded debris than the H-files (P = 0.009). No significant differences were detected amongst the four retreatment techniques concerning residual filling material (P = 0.082). The amount of extruded debris and areas of remaining filling material were not correlated (P = 0.901). Location of teeth in the maxilla or mandible had no impact on the amount of extruded debris within each instrument group (P = 0.609). However, when teeth were evaluated in general irrespective of the instruments, significantly more debris was extruded in the mandibular location (P < 0.001). CONCLUSIONS: All retreatment systems were associated with apical extrusion of debris, but H-files extruded significantly more material than Reciproc. Remnants of filling material were observed in all samples with no significant differences between the four techniques.