Ethical approval (#10228) was obtained from the Ethics Committee of Shiraz Dental School. Eighty mature human mandibular first and second molars with mesial root curvature within the range of 20°-40° were selected from an existing pool of extracted human teeth in the Biomaterial Research Center, Shiraz, Iran. Root curvatures were measured after pre-operative radiographs according to the methodology of Schneider . Teeth were excluded if they had heavily calcified canals, canals with apical foramina larger than a size 15 hand file (i.e. less than a size 20 file) as assessed externally, or with preexisting fractures or cracks when examined under a light microscope at ×20 magnification (Dino-Lite Pro2 AD413TL; AnMo Electronics Corp, New Taipei City, Taiwan). The teeth were stored in 0.1% chloramine T solution at 4°C throughout the study.
After access cavity preparation, a size 10 K-file (MANI, INC. Utsanomiya, Japan) was placed into the canal until it was visible at the apical foramen and 1mm was subtracted to establish working length (WL). A glide path was then prepared to a size 15 K-file. Each canal was irrigated with 2 mL of a freshly prepared 1% NaOCl solution between each instrument and dried with paper points. The teeth were then randomly divided into the following four groups (n=20):
Group 1- Non-instrumented root canals (Control group)
Root canals were only irrigated as far into the canal as feasible; there was no glide path negotiation, instrumentation, or obturation.
Group 2- Instrumentation with stainless steel hand files (SS).
Canal preparation was performed using a step-back technique with the apical portion enlarged up to a size 25 K-file (MANI, INC. Utsanimiya, Japan). Progressively larger K-files were used to step-back in 1 mm increments to 5 mm short of WL. Gates Glidden (GG) drills (MANI, INC. Utsanimiya, Japan) were used to enlarge the middle and coronal portions of the root canals as follows: GG #2 to 5 mm short of the WL, GG #3 to 7 mm short, and GG #4 was used just into the orifices.
Group 3- Instrumentation with ProTaper Next (PTN)
Canals were prepared using PTN rotary NiTi instruments (Dentsply Maillefer, Ballaigues, Switzerland) using an adjustable torque- and speed-controlled endodontic motor (Endo-Mate DT; NSK Nakanishi, Inc, Kanuma, Japan) according to the manufacturer’s instructions. In brief, X1 (0.17/0.04 ) and X2 (0.25/0.06) instruments were sequentially used to WL in a crown-down manner.
Group 4- Instrumentation using Mtwo (M2)
Canals were instrumented using a modified Mtwo (VDW GmbH, Munich, Germany) protocol to a standardised 35/0.04 instrument, after initial crown-down preparation with the 25/.07 (5mm short of WL) and 30/.05 (2mm short of WL) instruments. The same electric motor was used as for the PTN, and torque settings were selected for each instrument according to the manufacturer’s instructions. In all groups, each instrument was used only once.
After SS preparation, root canals were dried and obturated with minimal force using thermoplastic compaction (BeeFill® 2in1, VDW GmbH, Munich, Germany). After the preparation was completed for the NiTi groups, the roots were filled with their respective gutta-percha (GP) systems (GP size matched with the master apical file) using a single-cone technique and 2Seal easymiX® (VDW GmbH, Munich, Germany) as a canal sealer.
Postoperative radiographs were taken in bucco-lingual and mesio-distal directions to confirm the adequacy of the root canal obturation. The obturated teeth were examined again using the microscope to exclude any teeth with cracks that may have been created during the treatment.
Mounting the roots and measuring fracture resistance
The mesial roots were sectioned at orifice level using diamond discs under water-cooling leaving a length of approximately 12 mm. To calculate true fracture resistance (FR), i.e. Force/Surface area (Anusavice et al. 2013), root length and volume of the roots were considered according to the equation: SA=3(V1-V2)/h where SA=surface area, h=root length, V1=volume of the root and V2=volume of the canal.
The root volume was measured by taking a silicone impression of the outer surface of the root up to the CEJ. This was then weighed on an electronic balance (GR-3000, A & D CL Toshiba, Tokyo, Japan) to an accuracy of 0.1 mg and filled with water. The amount of water required to fill the silicone mould was also weighed and subtracted from the volume of the mould to determine the total volume of the root (V1). Because the shape of the root canal preparation was essentially a truncated cone (Fig. 1), the volume of the root canal space (V2) was based on the canal taper and apical size for each instrumentation method and calculated using the following equation: V2 = πh(r12 + r1r2 + r22) / 3, where h is the height (i.e. length of root), r2 is half the apical preparation size, and r1 is the radius at that plane of the root based on the dimensions of the final instrumentation system.
All the roots were mounted vertically in autopolymerizing acrylic in PVC tubes with diameter 20 mm and length 20 mm. The roots were covered with a 1 mm layer of light body polyvinyl siloxane (Affinis, Coltene AG/Whaledent Inc, Altstatten, Switzerland) to simulate a periodontal ligament and positioned in the center of the acrylic resin with 3 mm of the coronal root portion exposed. The mounted specimen was placed in a jig at a 45-degree angle and aligned vertically in the universal testing machine (Zwick/ Roell Z020; Zwick GmbH & Co, Germany). A vertical loading force to fracture was applied using a cone-shaped metal rod (0.7 mm diameter blunt tip) mounted on the machine directly over the canal opening of each root at a rate of 1.0 mm per min. The force, measured in Newton (N), was recorded and the maximum load was designated as the fracture load (FL).
The collected data were analysed using the SPSS package (version 22, SPSS Inc., Chicago, IL, USA). The normality assumption was assessed using the Kolomogorov- Smirnov test. The assumption of homogeneity of variances for data in the FR group was not supported; therefore a natural logarithm was applied to make the variation of data homogenous. A one-way ANOVA with post-hoc Duncan’s test was used for each value group (FL and FR) to compare the three instrumentation methods and the control group, with P < 0.05.