This in vitro experimental study was conducted on 54 single-rooted maxillary canine teeth extracted for reasons not related to this study, such as poor periodontal prognosis. The study protocol was approved by the ethics committee of the university (IR.SBMU.DRC.REC.1401.060).

The sample size for this study was calculated to be 18 in each group assuming α = 0.05, β = 0.2, a study power of 80%, mean values of 1.2 and 1.9, and standard deviation values of 0.6 and 0.8 for global apical deviation based on data reported in a previous study [27]. In the SSG group, canine teeth were mounted unilaterally (n = 18), while in the DSG group, two canine teeth were mounted bilaterally in each cast (n = 36). A total of 36 casts and three-dimensionally printed templates were evaluated.

The inclusion criteria were a mean length of 25–27 mm, completely formed roots, having a single canal, small/no caries, and no/minimal restoration.

For the purpose of standardization, teeth were decoronated by a disc such that the root length was standardized at 20 mm.

After access cavity preparation, a #15 K-file (Mani, Tochigi-Ken, Japan) was introduced into the canal to ensure patency and to determine the working length. The root canals were then instrumented with ProTaper Gold (Dentsply Sirona, Ballaigues, Switzerland) using the single-length technique. All teeth were instrumented to F3 final size and irrigated with 2 mL 5.25% sodium hypochlorite between the files using a 30-gauge side-vented needle with a safety tip. The total volume of irrigation with 5.25% sodium hypochlorite at the end of the instrumentation was 10 mL for each canal. After completion of the root canal preparation, 5 mL 17% EDTA was applied to the canals for 1 min, after which the specimens were subsequently rinsed with saline. After the root canals were dried at paper points, the canals were obturated via the single cone technique using the corresponding ProTaper gutta-percha points and AH26 sealer (Dentsply Sirona). The specimens were wrapped in sterile gauze and stored at 100% humidity (to remain hydrated) for 1 week to allow the sealer to set.

The post space was prepared with a #4 Peeso reamer (Mani, Tochigi – Ken, Japan) 1.3 mm in diameter and 11 mm in length from the tooth surface. A #2 cylindrical glass fiber post (Nordin Glassix, Chailly, Switzerland) with a round end and a 1.2 mm diameter was inserted into the canal, and a fissure bur was used to shorten the fiber post 2 mm below the tooth margin for placement of the final restoration. The fiber post was then cemented in the canal by using self-adhesive resin cement (DentKist SuperCem, Gunpo-si, South Korea). Finally, a dual-cure composite (Tokuyama, Tokyo, Japan) with a 2 mm thickness was applied over the fiber post.

The teeth were mounted in 36 stone casts. In the SSG group, canine teeth were mounted unilaterally (n = 18; 9 on each side, R or L) to mitigate any potential left-hand or right-hand bias, while in the DSG group, two canine teeth were mounted bilaterally in each cast. To fix the teeth in the casts, a putty-wash impression was made from the maxillary dental arch of a resin model, and the extracted canine teeth were mounted at the site of canine teeth in the impression by using dental wax. The impressions were then poured with type 3 dental stone. The teeth in the impression were embedded in dental wax from their cusp tip to their cementoenamel junction (CEJ). Thus, the coronal parts above the CEJ remained exposed, while the areas below the CEJ were embedded and fixed in dental stone (Fig. 1).

(Right) SSG group stone cast. (Left) DSG group stone cast

All casts were subjected to CBCT with a NewTom VGi CBCT scanner (NewTom, Verona, Italy) with the following parameters: 110 kVp, 11.21 mA current, 60 × 60 mm field of view, and 0.15 mm voxel size. The scans were saved in DICOM format. One cast from the SSG group and one from the DSG group were visualized via CBCT, and the images were separated from each other by a cotton roll. After CBCT, all casts were fully scanned by an intraoral scanner (CS 3600 scanner; Carestream, Atlanta, USA), and the files were saved in STL format.

Both the CBCT and intraoral optical scan files were transferred to surgical planning software (Blue Sky Plan 4; Blue Sky Bio, LLC; Grayslake, IL) for automatic superimposition. An expert operator also supervised the process to ensure accurate superimposition of the two files. After superimposition of the files, the guide was designed according to the location of the fiber post such that the drill would have no contact with the root dentin and would only drill the fiber post along its path. The drilling path was terminated when the end of the fiber post was reached (11 mm from the tooth surface), and gutta-percha root filling was initiated (Fig. 2). With respect to tooth coverage by the guide, in the SSG group, two anterior and two posterior teeth were included in the guide design (a 5-unit guide). In the DSG, the guide was passed through the midline, which included both canines bilaterally and two teeth posterior to the canine teeth on each side (a 10-unit guide) (Fig. 3). The location of the metal sleeve in the resin template was also determined. The stainless-steel metal sleeve (Straumann T-sleeve, Basel, Switzerland) used in this study had a 2.5 mm external diameter, 1.35 mm internal diameter, and 7.5 mm length [28]. Additionally, to stop the drill at a length of 20 mm, a 1.5 mm space was considered between the metal sleeve and the tooth surface.

(Up) Guide design for the SSG group. (Down) Guide design for the DSG group. (a) Coronal, (b) axial, (c) sagittal, (d) three-dimensional, and (e) panoramic views

Dental coverage, location of the metal sleeve, verification windows of the guide, and location of the drill entry in the DSG (a) and SSG (b) groups

The design was then printed with a 3D printer (Anycubic, Shenzhen, China) using photopolymerized biocompatible polymer resin (PowerResins, Singapore) with a 3 mm thickness. The metal sleeve was placed mechanically (frictionally) in the created hole due to the presence of threads in the sleeve. The templates were tested in their respective casts to ensure their correct seating, retention, and stability. Additionally, verification windows were considered in resin in template design to ensure complete seating of the guide. The casts were placed in a phantom head and fixed by an incorporated magnet. The correctness of the guide on the cast was evaluated again by the verification window (Fig. 4).

(a, b, c) Drilling process; SSG and DSG along with the drill

A Straumann drill (Straumann Guided Surgery, Basel, Switzerland) 20 mm in length and 1.3 mm in diameter was used in combination with an implant motor (NSK, Nakanishi, Japan) operating at 1200 rpm with 25 N/cm torque for drilling of the fiber post to reach the gutta-percha [28]. Drilling of each tooth was performed in 3 steps. After each step, the guide was removed, and the drill debris, sleeve, and tooth were rinsed with saline. The entire drill length (20 mm) was entered into the metal sleeve such that 9 mm of the drill was involved with the sleeve and the distance between the sleeve and tooth surface, and the apical 11 mm of the drill was in the canal for fiber post removal. After completion of drilling, 17% EDTA and saline were used for the final rinse of debris. The entire drilling process was performed by a senior postgraduate student in endodontics (Fig. 4).

After completion of drilling, all casts in both groups underwent CBCT with the same parameters as explained for the preoperative CBCT. The postoperative CBCT scans were subsequently transferred to Blue Sky software (Blue Sky Plan 4; Blue Sky Bio, LLC, Grayslake, IL). The drill paths on the postoperative CBCT scans were superimposed on the virtually designed drill path, and the maximum coronal deviation (MCD) at the marginal entry point of the tooth in millimeters (mm), the maximum apical deviation (MAD) at 10 mm apical to the tooth margin in millimeters (mm), and the maximum angular deflection (MAnD) from the designed path in degrees were calculated and recorded (Figs. 5, 6 and 7).

A schematic view of the maximum coronal deviation (A), maximum apical deviation (B), and maximum angular deflection (Ɵ) is shown

(a) Superimposition of postoperative CBCT scans on the virtually designed drill path in the SSG group (the axial view (red circle) indicates the virtually designed path, and the blue circle indicates the actual drill path). (b) Calculation of the maximum coronal deviation (at 0 mm) and maximum apical deviation at 10 mm apical to the tooth margin in the SSG group (the blue rectangle indicates the virtually designed path, and the red rectangle indicates the actual drill path). (c) Superimposition of postoperative CBCT scan on the virtually designed drill path in the SSG group (3D view - green cylinder indicates the virtually designed path, and blue cylinder indicates the actual drill path)

(a) Superimposition of postoperative CBCT scans on the virtually designed drill path in the DSG group (the axial view (red circle) indicates the virtually designed path, and the blue circle indicates the actual drill path). (b, c) Calculations of the maximum coronal deviation (at 0 mm) and maximum apical deviation at 10 mm apical to the tooth margin in the DSG for the right and left canines, respectively (the blue rectangle indicates the virtually designed path, and the red rectangle indicates the actual drill path). (d) Superimposition of postoperative CBCT scans on the virtually designed drill path in the DSG group (3D view: the green cylinder indicates the virtually designed path, and the blue cylinder indicates the actual drill path)

The Shapiro‒Wilk test confirmed the normality of the data distribution, while Levene’s test confirmed the homogeneity of variance (P > 0.05). Thus, an independent t test was applied to compare the means of the dependent variables between the two groups (SSG and DSG). All the statistical analyses were carried out using SPSS version 25 (SPSS, Inc., IL, USA) at the 0.05 level of significance.