Dynamic navigation versus conventional freehand technique for injectable gutta-percha removal in endodontic retreatment: A comparative in vitro study

Abstract

Objectives

Dynamic navigation systems are increasingly used in endodontics to improve accuracy, particularly in complex cases. However, their effectiveness in routine procedures such as gutta-percha removal remains unexplored. This study compared the accuracy and efficiency of Dynamic Navigation System (DNS) versus the conventional freehand technique in gutta-percha retrieval.

Materials and methods

Single-rooted mandibular first premolar teeth obturated with injectable thermoplasticized gutta-percha were used in this study. Tooth samples were mounted on a 3-dimensional (3D) printed mandibular cast. Pre-operative and post-operative Cone Beam Computed Tomography were performed. The samples (n = 56) were randomly divided into two groups (I: DNS; II: conventional technique). All procedures for both groups were performed by a single clinician who had been trained in DNS. Accuracy and efficiency were assessed by evaluating coronal and apical deviation, angular deflection, procedural time, and the extent of gutta-percha removal.

Results

The data were analysed using IBM SPSS Statistics for Windows, Version 20 (IBM Corp., Armonk, N.Y., USA). The Kolmogorov-Smirnov test was used to assess the normality of the data, and it was determined that all parameters were normally distributed. The mean and standard deviation were calculated as the descriptive statistics. An Independent t-test was employed to analyze intergroup comparisons of all parameters. The statistical significance level was determined at p < 0.05. The conventional group showed less coronal and apical deviation compared to the DNS group (p = 0.000). The conventional group exhibited less angular deflection than DNS group (p = 0.005). The conventional group required less procedural time than the DNS group (p = 0.010).

Conclusion

The conventional free-hand technique significantly improved the operator's accuracy and efficiency compared with DNS for the retrieval of injectable gutta-percha.

Highlights

• •Dynamic navigation accuracy enhances with experience, while freehand retreatment remains viable for gutta-percha removal.

1. Introduction

Non-surgical root canal treatment is intended to prevent and treat apical periodontitis, either by maintaining or restoring natural teeth.¹ Notably, a meta-analysis revealed a global prevalence of 39% for apical periodontitis in root-filled teeth.² Non-surgical root canal retreatment is indicated when primary root canal treatment fails to resolve periapical disease and the quality of the previously performed procedure is unacceptable.³ The main goals of retreatment procedure are to completely remove any obstructing restorative materials and previous root filling material, eradicate the microbes, guarantee thorough disinfection, and achieve fluid-tight obturation within the root canal system.⁴ Therefore, effective retreatment requires the removal of the previously filled material without removing the tooth structure.^3,4

The removal of previously filled materials has been proposed using various instruments and techniques, including hand instruments, rotary or reciprocating systems, and sonic and ultrasonic irrigation systems.^5,6 It is well established that the procedure time and fatigue for the operator are reduced by nickel-titanium (NiTi) rotary systems, which are more effective than hand files.⁷ The ProTaper Universal retreatment system (Dentsply Maillefer, Ballaigues, Switzerland) is a common NiTi retreatment system used in continuous rotary motion.⁸

Guided endodontics is a technique used to manage obliterated root canals, post removal, autotransplantation, and periradicular surgery with fewer mishaps.9, 10, 11 Guided endodontics can be static or dynamic.¹² The dynamic navigation system (DNS) uses a preoperative cone-beam computed tomography (CBCT) scan to virtually plan and guide the bur during the procedure. The 3D-DNS operated similarly to a global positioning system for the handpiece and drill, providing real-time visual assistance for navigating surgical instruments during procedures and allowing plan modifications, if appropriate. DNS has the potential to enhance the accuracy and safety of less invasive procedures performed by dental surgeons while simultaneously reducing the likelihood of errors.^10,12

While dynamic navigation improves accuracy in complex endodontic cases, such as calcified canals, fiber post removal, and microsurgery, its role in routine retreatment, especially gutta-percha removal, remains unclear. Unlike access cavity preparation or apical surgery, gutta-percha retrieval depends on tactile control, canal centricity, and file flexibility rather than on predefined paths. Most studies have focused on hard tissue targeting accuracy rather than on removing thermoplasticized materials. However, evidence that real-time guidance surpasses freehand techniques for gutta-percha removal is lacking. This study aimed to compare the accuracy and efficiency of the DNS technique with those of conventional freehand approach for gutta-percha retrieval during retreatment procedures. The main objective of this study was to compare and evaluate dynamic navigation and conventional retreatment techniques with respect to canal deviation, procedural time, and efficacy of gutta percha removal.

2. Methods

2.1. Ethics and reporting guidelines

The Institutional Scientific and Ethical Committee Review Board of the SRM Institute of Science and Technology (SRMIST) University approved the study and the ethical clearance number SRMIEC-ST0323-422 was assigned. This study was conducted in accordance with the Declaration of Helsinki. This research adhered to the PRILE 2021 requirements for laboratory endodontic investigations; the PRILE flowchart is provided in Supplementary Fig. 1.¹³

2.2. Sample size calculation

Using G*Power software (3.1.9.2 version) and considering a priori power analysis, t tests – Means: Difference between two independent means (two groups), and by maintaining the following input parameters: α error as 0.05 at 95% CI, β error as 0.05. power of the test (1- β error) as 95%, and allocation ratio N1/N2 as 1. Two-tailed hypothesis and effect size (Cohen's d statistic) as 0.9982444 (determined from Connert T; 2021). The estimated total sample size for the current study with 95% power was 56 samples (n). Fifty-six human mandibular single-rooted first premolar teeth, freshly extracted for orthodontic treatment, were used in this study. The teeth were obtained from the Oral and Maxillofacial Surgery Department of SRM Kattankulathur Dental College and Hospital. Informed consent was obtained from the patients for the use of extracted teeth for research purposes. After sterilization in a steam autoclave, all teeth were kept hydrated in distilled water. Since no prior studies have specifically evaluated the effect sizes for dynamic navigation–assisted gutta-percha removal, a conservative effect size was chosen based on research on guided endodontic accuracy (Figure-1).

Fig. 1.

Shows the workflow of the study.

2.3. Root canal treatment procedure

The samples were standardized after radiographic pre-screening using periapical images obtained at multiple angulations to eliminate anatomical variations. They included only single-canal teeth with a root length of 22 mm. An access cavity was prepared using an endo access bur connected to an air turbine handpiece. Canal patency was checked using a #10 K-file. The working length was determined radiographically using a #15 K-file which was 1 mm short of the apex. Instrumentation was performed using ProTaper Gold rotary files in the sequence of SX, S1, S2, F1, and F2 (#25/0.08) using a crown-down technique. Irrigation was performed with 5.25% sodium hypochlorite between files, followed by 17% EDTA for 3 min and a final saline rinse. Obturation was performed using a zinc oxide-eugenol sealer and F2 (25/0.08) gutta-percha (Dentsply), followed by thermoplasticized backfill using Smart Gutta (Dentsply), with vertical compaction to minimize shrinkage.^14,15 The sealer was subsequently allowed to fully set by storing the tooth samples in a 100% humidity environment at 37°C for seven days. An endodontic resident performed all root canal treatment procedures (Groups I and II) under the direct supervision of endodontic faculty with more than 15 years of academic and clinical experience.

2.4. Pre-operative CBCT scans

A 3D printed resin mandibular cast was used to affix the tooth samples. The samples were stabilized using modelling wax to prevent movement during the CBCT acquisition. The mandibular cast was used to mount two to four tooth samples. Sixteen models were used. A voxel resolution of 0.120 mm^3 was achieved in a single-arch preoperative CBCT scan (CS 9300; Carestream, Atlanta, GA). The teeth were randomly divided into two groups by simple randomization using a coin-tossing method without stratification: Group I: DNS (n = 28) and Group II: conventional freehand (n = 28). A single operator, who was trained and calibrated in DNS, executed all procedures for both groups. A single operator performed all procedures to reduce inter-operator variability, however this might have caused operator-preference bias.

2.5. Group I: Dynamic navigation system (DNS)

The first step involved the acquistition of Digital Imaging and Communications in Medicine (DICOM) data of natural premolars mounted on a 3D printed cast and importing it into the DNS planning software (Navident; ClaroNav, Toronto, Canada). The canal access pathway for injectable gutta-percha retrieval was planned by determining the drilling entry point, angle, trajectory, and depth along the canal axis using 3D visualization on the software interface. The second phase comprised "trace registration" of the tooth surfaces, which ensured that the tooth and canals were accurately positioned throughout the procedure. The third phase involved calibrating the endomotor and retreatment files. In a two-step procedure, the endomotor tracked by the system was calibrated. The axis was calibrated first, and the tip of the retreatment file was calibrated subsequently. This enabled the system to continuously monitor the direction and position of the retreatment files and present the details to the operator in real time on a computer screen.

All procedures were performed by a single operator with 12 years of clinical experience to ensure standardization and to minimize inter-operator variability. The operator underwent training and calibration sessions using 30 trial attempts before the study to gain proficiency in the DNS.

Gutta-percha removal was performed using DNS, X-smart endomotor (Dentsply Sirona, Ballaigues, Switzerland), and ProTaper Universal Rotary Retreatment Files (Dentsply Sirona, Ballaigues, Switzerland). New retreatment files were used for each sample. Calibration of the system and the drilling trajectory were planned. After the consistency between the actual position of the ProTaper retreatment rotary files and the virtual location in the DNS software was verified, gutta-percha was drilled under the comprehensive supervision of DNS (Navident) with an X-smart Endomotor (Dentsply) at 500 rpm using ProTaper retreatment files. After attaining straight-line access to the canal, 2.5 ml of gutta-percha solvent was drawn into a syringe. The canals were flushed with a few drops of gutta-percha solvent between the files. The gutta-percha was retrieved with D1, D2, and D3 in the coronal, middle, and apical thirds of the canal, respectively. Gutta-percha removal was completed when the D3 file reached the end of the designated drilling trajectory. The DNS software displayed a virtual image during the procedure, allowing the operator to modify the tooth location, angle, and depth, as well as the position of the rotary files, in real time as illustrated in Fig. 2, Fig. 3.

Fig. 2.

Clinical working view of DNS-assisted gutta-percha retrieval in the laboratory model.

Fig. 3.

Real-time navigation screen display showing the planned and active drilling path during gutta-percha removal.

2.6. Group II: Conventional technique

The procedure-planning methods of the conventional group were nearly identical to those of the DNS group. Drilling was stopped when the D3 retreatment rotary files reached the full length of the root canal, as confirmed radiographically, without the need for DNS guidance.

2.7. Post-operative CBCT scans

All samples underwent a postoperative CBCT scan and were subsequently entered into DNS Evalunav software. The amount of gutta-percha removal, procedure time, and canal deviation in both groups were assessed by superimposing preoperative and postoperative CBCT images.

2.8. Outcome measures

The amount of gutta-percha removal was assessed by comparing preoperative and postoperative CBCT images for both groups. Visual evaluation of the root filling material removal was performed by a single blinded operator on a desktop computer with a 21-inch monitor under ambient lighting. The specimens were coded by an independent investigator who evaluated and assessed them in a randomized order without group identifiers. The Evalunav superimposition software (Navident Technologies) was used to assess the accuracy of preoperative and postoperative CBCT images. The accuracy of the DNS and traditional methods was assessed by comparing the differences in coronal and apical positions and angle changes from the planned drilling path in preoperative CBCT scans with those in postoperative CBCT scans. Postoperative drilling trajectory deviations from the preplanned trajectory were quantified in the X, Y, and Z dimensions. A blinded, experienced clinician who had been involved in the planning and execution of the gutta-percha retrieval procedure conducted the evaluations. Efficiency was evaluated by measuring the procedure duration in minutes using a stopwatch. The DNS technique was recorded from the start of drilling to the end of the planned path, whereas the traditional method was recorded from the beginning of retrieval until the D3 files reached the working length. The completeness of gutta-percha removal was qualitatively assessed using preoperative and postoperative CBCT images, as volumetric segmentation of thermoplasticized gutta-percha was beyond the scope of this study.

2.9. Statistical analysis

The data were analysed using IBM SPSS Statistics for Windows, Version 20 (IBM Corp., Armonk, N.Y., USA). The Kolmogorov-Smirnov test was used to assess the normality of the data, and it was determined that all parameters were normally distributed. The mean and standard deviation were calculated as the descriptive statistics. An Independent t-test was employed to analyze intergroup comparisons of all parameters. The statistical significance level was determined at p < 0.05.

3. Results

The mean ± standard deviation values for the coronal and apical deviations, angular deflection, and procedural time are presented in Table 1. Quantitative parameters derived from CBCT imaging, including coronal deviation, apical deviation, and angular deflection, were analysed using appropriate parametric tests following confirmation of normal data distribution by the Kolmogorov–Smirnov test. The descriptive statistics, including the means and standard deviations for coronal deviation, apical deviation, angular deflection, and procedural time, are summarized in Table 1. The comparison of canal deviation between the two techniques revealed statistically significant differences; the dynamic navigation system (DNS) group showed a greater mean coronal deviation (1.04 ± 0.22 mm) than the conventional freehand group (0.39 ± 0.16 mm) (p < 0.001), with a similar pattern in apical deviation (0.77 ± 0.25 mm vs. 0.53 ± 0.21 mm, p < 0.001). Angular deflection analysis also indicated a significant difference, as the DNS group exhibited a higher mean angular deviation (2.45 ± 1.33°) than the conventional technique (1.65 ± 0.54°, p = 0.005), reflecting greater deviation during gutta-percha retrieval. Procedural efficiency, measured by total removal time, favored the conventional method, which had a shorter average duration (7.04 ± 1.21 min) than the DNS group (8.24 ± 2.03 min; p = 0.010). CBCT evaluation was performed through visual assessment by a single blinded evaluator using pre-operative and post-operative CBCT images to ensure a comprehensive three-dimensional analysis. As the qualitative assessment did not involve a predefined scoring system, the observations were recorded descriptively, particularly for parameters such as the completeness of gutta percha removal. Qualitative CBCT evaluations demonstrated successful gutta-percha removal in both groups. However, removal was more consistent and complete in the conventional group, with residual obturation material observed more frequently in the DNS samples, particularly in the apical third. Additionally, within the DNS group, a trend toward reduced angular and linear deviations across successive samples suggested a learning curve, which could have influenced the accuracy as operator familiarity with the system increased. The conventional group exhibited significantly lower coronal and apical deviations and angular deflection than the DNS group (P < 0.05). The conventional group had a shorter average time for gutta-percha removal (7.04 ± 1.21 min) than the DNS group (8.24 ± 2.03 min) (P < 0.05). Fig. 4(A–F) illustrates the extent of gutta-percha removal, demonstrating that the conventional technique accomplished complete gutta-percha removal compared to DNS. Successive samples showed decreased angular deflection and deviation in the dynamic navigation group, indicating an operator's learning curve.

Table 1.

Intergroup comparison of coronal and apical deviation, angular deflection, and procedural time (mean ± standard deviation).

Parameter	Groups	N	Mean ± SD	Std. Error Mean	Mean difference	Std. Error difference	95% Confidence Interval of the Difference		df	Test valueb (p-value)
							Lower	Upper
Coronal Deviation (mm)	I	28	1.04 ± 0.22	0.042	0.648	0.052	0.543	0.753	54	12.349 (0.000)a
	II	28	0.39 ± 0.16	0.030
Apical Deviation (mm)	I	28	0.77 ± 0.25	0.047	0.240	0.062	0.115	0.366	54	3.852 (0.000)a
	II	28	0.53 ± 0.21	0.040
Angular deflection (°)	I	28	2.45 ± 1.33	0.252	0.798	0.272	0.251	1.345	54	2.928 (0.005)a
	II	28	1.65 ± 0.54	0.102
Time Taken (mins)	I	28	8.24 ± 2.03	0.384	1.202	0.447	0.305	2.100	54	2.686 (0.010)a
	II	28	7.04 ± 1.21	0.229

^aStatistically Significant (p ≤ 0.05); Group I - 3D Dynamic Navigation, Group II - Conventional Technique.

^bIndependent t-test value.

Fig. 4.

(A-F): Pre-operative and Post-operative CBCT scans of the DNS and conventional groups for gutta-percha removal.

Legends: Fig. 4(A–F): Pre-operative and Post-operative CBCT scans of the DNS group and the conventional group for gutta-percha removal. [A: Pre-operative CBCT – DNS, B: Post-operative CBCT – DNS (Sample 1), C: Post-operative CBCT – DNS (Sample 2), D: Pre-operative CBCT – Conventional technique, E: Post-operative CBCT – Conventional technique (Sample 1); F: Post-operative CBCT – Conventional technique (Sample 2)].

4. Discussion

This study aimed to evaluate the effectiveness of two techniques (conventional freehand and DNS) for the retrieval of injectable gutta-percha from an extracted tooth. The efficacy of gutta-percha removal using ProTaper instruments and H-files, with or without chloroform, was evaluated in a 2013 study by Khalilak et al.¹⁶ The results suggest that ProTaper NiTi instruments are more effective for gutta-percha retrieval than H-files in canals with minimal or no curvature.¹⁶ Consequently, ProTaper Universal retreatment files were used to remove the gutta-percha in both groups. Although dynamic navigation improves accuracy in real time, its clinical benefits are procedure-specific. In retreatment involving gutta-percha removal, success depends on canal anatomy, file flexibility, and tactile feedback, not trajectory targeting. Therefore, navigation-assisted drilling may not enhance the performance of gutta-percha retrieval, especially in straight canals.

The mandibular first premolar was chosen for this investigation because of its high failure rate and treatment difficulty. The extreme variations in root canal morphology in these teeth may help explain the treatment failure.¹⁷ This study used thermoplasticized gutta-percha for three-dimensional obturation because of the complex canal morphology of the mandibular first premolar.

The accuracy and precision of the DNS and conventional techniques were assessed by comparing the drilling trajectory from the pre-operative CBCT with that from the post-operative CBCT using superimposition software. In this study, the coronal and apical deviations and angular deflection were 1.04 mm, 0.77 mm, and 2.45° for DNS, and 0.39 mm, 0.53 mm, and 1.65° for the conventional approach, indicating that DNS exhibited higher coronal and apical deviations and angular deflection than the conventional method. This may be attributed to the flexibility of NiTi rotary files, which may have caused a deviation from the intended route during the accuracy assessment with the DNS. However, the angular deflection for retrieving gutta-percha in the first and final samples of the DNS group was 4.34° and 0.71°, respectively. This indicates that proficiency in DNS utilization improves with experience. The improvement in angular deviation across DNS samples showed a learning curve, indicating that operator experience is crucial for effective dynamic navigation–assisted retreatment.

A study conducted by Jain et al.¹⁸ assessed the precision of the DNS for intraosseous anesthesia, revealing an angular deflection of 1.36°, demonstrating the precision achievable with real-time guided systems. The authors attributed this accuracy to the enhanced control over angulation and depth provided by continuous visual feedback and pre-operative planning. Prior research by Dianet et al.¹⁹ examined the precision of the DNS in identifying calcified canals and reported an angular deflection of 1.69°. The greater angular deviation in the conventional technique is attributed to the absence of real-time guidance and increased operator dependency. Dianat et al.²⁰ recently assessed the DNS for precision in root-end resection, demonstrating an angular deviation of 2.54° with the DNS technique and 12.38° with the traditional approach. The research conducted by Tang et al.²¹ demonstrated that deviations in the angles, volumes, lengths, and depths of root-end resections were markedly reduced in the SN and DNS groups compared with the FH group. In contrast, the present study demonstrated greater deviation in the DNS group, which may be attributed to the technique sensitivity of the system, including the influence of the operator learning curve, calibration accuracy, and the demands of the real-time hand-eye coordination. Optimal performance requires refined motor control, hand-eye coordination, and continued practice to achieve proficiency, as outcomes are highly dependent on operator proficiency and precise calibration, as well as differences in study design and procedural complexity. Similarly, a study by Karim et al.²² indicated that although dynamically guided trephining with an extractor may recover separated instruments, it is less effective than 3D microscope guidance in terms of treatment duration, procedural mishaps, and volume changes. The findings of the present study align with those of Karim et al.,²² as earlier investigations on DNS indicate that the guided approach is preferable. The data from this study suggest that although DNS is a guided technique, its efficacy also depends on the operator's proficiency.

Efficiency was assessed by measuring the duration of the procedure in minutes. In a study by Aldahmash et al.,²³ the time required for osteotomy and root-end resection, as well as the total procedure time, was considerably reduced with DNS compared with the freehand technique. The DNS and freehand methods had total procedural durations of 13.3 ± 4.5 min and 23.71 ± 8.16 min, respectively.²³ In another study, Janabi et al.¹⁴ observed that the mean duration of gutta-percha removal in DNS (4.03 ± 0.43 min) was shorter than that with the freehand technique (8.30 ± 4.65 min). The present study found that the average duration for gutta-percha removal in the conventional group (7.04 min) was shorter than that of the DNS group (8.24 min). The efficacy ascertained in the present investigation is in direct contrast to the findings of other studies.

The findings of this study suggest that DNS is less accurate than the conventional freehand technique for removing injectable gutta-percha, consistent with the study by Karim et al.²² Therefore, the null hypothesis is rejected. Nevertheless, the DNS accuracy increased significantly from the first to the 28th sample, which can be attributed to the inherent learning curve of the DNS. Preoperative CBCT is essential for accurate planning before this procedure, as its precision is contingent on the later. Additionally, the operator must consistently verify the CBCT image for any distortion caused by patient movement, and the dentist provides dynamic guidance, unlike in a static system. It is uncommon to observe the patient rather than the monitor when preparing access cavities, which requires experience. Another disadvantage of DNS is their increased procurement costs. Furthermore, as both DNS and conventional procedures were performed by a single operator, blinding during the intervention phase was not feasible, which may have introduced a performance bias despite minimizing inter-operator variability. No former calibration or intra-observer analysis was performed, which represents another limitation of this study.⁹

The present study has limitations as an in vitro investigation, lacking clinical variables such as patient movement, canal curvature, and intraoral constraints. Only straight, single-rooted mandibular premolars were evaluated, limiting the applicability to more complex anatomies. This may influence the procedural accuracy and efficiency, as complex canal anatomies, including curvature and anatomical variations, can significantly affect instrument navigation and gutta percha removal in clinical scenarios. Gutta-percha removal was qualitatively assessed using CBCT imaging. However, this approach is inherently limited by subjectivity and the absence of quantitative validation, as advanced volumetric segmentation techniques may yield more precise, objective, and reproducible three-dimensional assessments. A further limitation of the present study is that the qualitative CBCT evaluation was not based on a validated scoring system, and no assessment of inter-observer or intra-observer reliability was performed. Another limitation of the present study is the lack of assessment of remaining dentin thickness following retreatment procedures. Future studies incorporating micro-CT or advanced CBCT-based volumetric analysis, standardized scoring system with reliability assessment and remaining dentin thickness evaluation are recommended to ensure a more comprehensive assessment of clinical performance and outcomes. All procedures were performed by a single operator, which may have introduced operator-dependent bias, despite reducing inter-operator variability. Although the operator was trained and calibrated prior to the study, a learning curve effect was observed with the DNS, with improved accuracy noted in the later samples compared to the earlier ones. This suggests that operator experience and proficiency with the system may influence outcomes and should be considered when interpreting results.

The strength of this study includes the use of DNS technology, which enables real-time visual feedback, CBCT-based three-dimensional evaluation for accurate assessment, comparative study design, and standardized methodology that enhances the reliability and reproducibility of the findings.

5. Conclusion

In this in vitro study, the conventional freehand technique showed lower canal deviation and shorter procedural time than dynamic navigation for gutta-percha removal. The findings suggest that dynamic navigation may not confer an advantage for gutta-percha removal in routine straight canals and may be influenced by operator experience and procedural indications. Further research on curved canals, volumetric assessments, and multiple operators is required. Overall, the study indicates that the conventional method is more effective and accurate than DNS, considering the limitations of the study.

Human and animal statements

There was no Human subjects and animal subjects were used for the study on “Dynamic Navigation Versus Conventional Freehand Technique for Injectable Gutta-Percha Removal in Endodontic Retreatment: A Comparative In Vitro Study”. All the authors agree to the same.

Declaration of competing interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article: