Fifty-two single-rooted human teeth with a single root canal, extracted for periodontal and orthodontic reasons, were used in this study. Superficial debris was removed from the roots with an ultrasonic scaler and the teeth were stored in physiologic saline until ready for use. The crowns were removed at the cementoenamel junction using a water-cooled tapered diamond bur. The working length was determined by passing a 15 K-type file into the canal until the tip was just visible at the apical foramen and then subtracting 0.5mm from this length.

The canals were cleaned and shaped with .04 and .06 taper Profile (Dentsply-Maillefer, Ballaigues, Swiss) using a crowndown pressureless technique. The apical portion of canal was enlarged to minimum ISO size #30 and maximum ISO size #55 file, depending on the size of original canal size. The apical patency was checked by passing a size #15 file through the apex.

Copious irrigation with 3.5% sodium hypochlorite was used throughout the cleaning and shaping of the canals using a syringe and 27 gauge endodontic needle. The canals were then flushed with 10ml of deionized water and dried with paper points. The prepared roots were randomly divided into four experimental groups of 12 teeth each. Four teeth were served as negative and positive controls and were prepared in the same manner as the experimental groups and received no filling (Group 5 and 6). Two teeth were used as negative controls (Group 6) and these teeth were completely coated with two coats of nail varnish.

The teeth in Group 1 were filled with gutta-perch and AH26 after All Bond 2 dental adhesive (Bisco, Itasca, IL, USA) application by the Continuous Wave of Condensation Technique. The canal walls were etched with 37% orthophosphric acid for 15 seconds, rinsed with 10ml-deionized water and dried with paper points. AH26 was then prepared according to the manufacture's directions. All Bond 2 primer was applied inside the canals. All Bond 2 Dentin Enamel Bonding Resin and Pre-Bond Resin was mixed and applied into the canal on a soaked paper point. The master gutta-percha point was coated with AH26 sealer and seated in the canal to the full working length. The root canal down-packed with System B (Analytic Endodontics, Orange, CA, USA) and backfilled with Obtura II (Obtura Spartan, Fenton, MO, USA) by the continuous wave of condensation technique. The appropriately tapered System-B Heat Plugger is preheated 200℃.

The teeth in Group 2 were filled with gutta-percha and AH26 after Prime & Bond NT Dual Cure Dental Adhesive System (Dentsply, Konstanz, Germany) application. The canal dentin was etched with Caulk 34% Tooth Conditioner Gel for 15 seconds. This was immediately rinsed for 10 seconds with 10ml deionized water and dried with paper points. Prime & Bond NT Dual Cure Dental Adhesive System was mixed according to the manufacturer's instructions and applied in the canal. Root canal filling procedure was then performed as described for the teeth in Group 1.

The smear layer of the teeth in Group 3 was removed by slowly injecting 10ml of 17% EDTA solution into the canal over a 5-minute-period. This was then followed by flushing the canal with 10ml of 3.5% NaOCl. No adhesive system was used in this group. Except this, the canal filling procedure was performed as described for the teeth in Group 1 and 2.

The root canals of Group 4 were filled with gutta-percha and AH26. There was no treatment of EDTA. The smear layer was still remaining. The root canals were filled by same procedure of Group 1 and 2.

The coronal ends of gutta-percha were cut off until a uniform filling length of 7mm was remained for all roots. After obturation, all roots were stored at 100% humidity and 37℃ incubator for 3 days to allow full setting of the sealer and adhesives.

A dual chamber anaerobic bacterial leakage model was assembled using a 5ml irrigation syringe and tooth as the upper chamber and a 20 ml scintillation vial (Samwoo Scientific Co., Seoul, Korea) as the lower chamber. The tooth was attached with cyanoacrylate cement and silicon glue to the tip of the syringe and the joint was sealed with two coats of nail varnish. The syringe was secured via a hole drilled through the cap of a 20 ml scintillation vial.

The apparatus was sterilized with ethylene oxide gas. The vials were placed in the anaerobic chamber (Coy Laboratory Products Inc., Ann Arbor, MI, USA) for 48 h to eliminate any oxygen in the system. Sterile brain heart infusion broth (Difco, Sparks, MD, USA) with yeast extract, hemin, menadion, and bromcresol purple (Sigma, St. Louis, USA) (bpBHI) was placed in each vial to a level of 4-5mm above the root end. Bromcresol purple is a chromatic indicator8). Using a sterile micropipette, 100 µl of bpBHI broth with Fusobacterium nucleatum (VPI 10197) was carefully placed into the upper chamber syringe reservoirs along with 3ml of sterile broth. 3ml of fresh sterile broth was added to each vial every week. The experimental samples were incubated in the anaerobic chamber at 37℃ and observed everyday for turbidity and color change of broth. Table 1 explains the leakage score obtained by criteria. Data was statistically analyzed using the Friedman Reapeated Measures Analysis.

Two roots of each group were used in scanning electron microscope (SEM) observation. One root was split-fractured along the axis of teeth and the other one was cross-sectioned at apical, middle and coronal levels of the root.

Vertical grooves were then cut along the buccal and lingual side of the roots with a slow-speed diamond saw (Isomet, Buhler Ltd., Lake Bluff, NY, USA), after which the roots were split in half occluso-apically with a triangular chisel and mallet. Horizontal sections were made approximately every 2mm with a low-speed diamond saw. Each two sections were made in the apical, middle and coronal thirds of each root.

The samples were first rinsed with gentle stirring in sterile saline solution to remove the non-attached bacteria and subsequently fixed in a 2.5% glutaraldehyde solution for 30min. After rinsing with 0.1M phosphate buffer, they were dehydrated by immersion in increasingly concentrated alcohol solutions (30%, 50%, 70%, and 100%) for 30min in each bath. Samples were subsequently critical point-dried to preserve the bacterial structure. Sputter coating the samples with carbon-gold completed the preparation of samples.

They were examined for the presence of the hybrid layer, the penetrated resin tags into the dentinal tubules, presence of void or gaps in the dentin-resin interface or resin-gutta percha interface, and Fusobacterium nucleatum that had attached to the canal surface using a scanning electron microscope (JSM-840S, JEOL, Tokyo, Japan).

The positive controls showed a color change and turbidity in the bpBHI within 1 to 2 days. In contrast, the negative controls did not show any color changes for experimental duration of 60 days. Table 2 shows the leakage scores of experimental groups. The samples of group 4 (AH26 alone without treatment of EDTA) leaks significantly more than the other three groups (p<0.0005). No statistical significant differences between group 1 and 2, between 2 and 3, and between 1 and 3.

The scanning electron micrographs revealed that removal of the smear layer allowed the adhesive agents and AH26 to enter dentinal tubules. A hybrid-like layer was observed in all the specimens of group 1 and 2 (Fig. 1 and Fig. 2). The dentinal tubules were also open and AH26 could penetrate into the dentinal tubules in Group 3 (Fig. 3). AH26 in group 4 less penetrated into the dentinal tubules than the other groups and no penetration was observed in some area (Fig. 4).