Correlation of root canal morphology of endodontically treated premolar and molar teeth with procedural errors using cone-beam computed tomography in an Iranian population

Introduction

Comprehensive knowledge about the root canal morphology has a fundamental role in the success of endodontic treatment [1]. Therefore, it is imperative to have adequate knowledge about the root canal configurations for a successful endodontic treatment. Such knowledge would enhance appropriate treatment planning and prevent technical errors in the course of treatment [2].
In various cases, some of the root canals may not be easily identified, and may be missed in root canal therapy. Inadequate knowledge about the root canal anatomy and not sufficiently negotiating the additional canals are the two main reasons for this problem [3]. Thorough negotiation as well as adequate instrumentation and obturation of additional canals can maximize treatment success, yielding excellent results. Correct access cavity design and adequate knowledge about the root canal ana­tomy can contribute to achieving the above-mentioned goal [4].
Although the outcome of endodontic treatment depends highly on the operator’s skills, procedural errors that may occur at any step of the treatment course reflect poor attention to canal configuration and faulty treatment plan [5-7].
Vertucci et al. [8] have shown complexities of the root canal by using clearing technique and hematoxylin staining.
Factors related to the root canal configuration complexities include racial differences, age, gender [9-11], and different methodologies, such as sectioning [12], staining of the root canal [13], and radiographic assessment [9]. Staining and sectioning are invasive procedures, which damage specimens. Conventional radio­graphy provides two-dimensional (2D) images, which do not reveal comprehensive information about the root canal morphology due to superimposition of anatomical structures. Panoramic radiography also has limitations, such as non-uniform magnification, distortion, and superimposition of anatomical structures [14]. Cone-beam computed tomography (CBCT) is superior to 2D imaging modalities due to the provision of high-quality, three-dimensional (3D) images without distortion and superimposition of structures. However, CBCT has a higher cost and radiation dose than 2D modalities. Compared with computed tomography, CBCT is the preferred choice due to excellent sub-millimeter resolution and lower patient radiation dose [15]. Different CBCT scanners have differences in resolution. Moreover, the type of detector can affect image quality and diagnostic results [16]. Even so, CBCT is not primarily used for initial assessment of the root canal morphology, and is only indicated when conventional imaging modalities cannot provide sufficient information for appropriate treatment planning [17]. According to Ok et al. [1], the root canal shape of premolar teeth of both jaws is complex and variable. Also, Nouroloyouni et al. [18] showed that the prevalence of root canal treatment errors was higher in molar teeth.

Objectives

The present study was conducted to assess the most common procedural errors, and evaluate the correlation of procedural errors with the root canal morphology of premolar and molar teeth according to the Vertucci’s classification using CBCT.

Material and methods

Number of samples

This retrospective cross-sectional study was conducted on eligible CBCT scans of 171 patients (65 males and 106 females, with a mean age of 49.0 ± 10.6 years) presenting at the Hamadan Dental School from 2019 to 2021 for dental diagnosis and treatment planning. Sample size included 131 endodontically treated premolar and 99 endodontically treated molar teeth (total of 375 roots). All CBCT images were obtained using the same CBCT scanner (NewTom 3G, Verona, Italy), and exposure parameters of 3.9-5.6 mA, 110 kVp, 0.2 mm voxel size, 6 × 6-inch field of view, and 3.6 seconds exposure time. The images were saved in NNT viewer software (Quantitative Radiology, Verona, Italy). An oral radiologist and a post-graduate student of oral radiology exa­mined the images.
Inclusion criteria were the presence of at least one endodontically treated molar and one endodontically treated premolar tooth on CBCT images, optimal quality of CBCT images for evaluation of root canal morphology of the teeth, and visualization of both jaws, posterior teeth, and periapical regions. Exclusion criteria were suspected teeth with a radiolucent line similar to fracture line, and teeth with endo-perio lesions.

Assessment of images

Prior to the study, the observers were calibrated by using a PowerPoint presentation of 30 CBCT images of endodontically treated teeth with different root canal configurations according to the Vertucci’s classification. The images were viewed with a 0.5 mm slice thickness, 30 mm width, and 1 mm interval on a 20-inch monitor (Samsung, Seoul, South Korea) with a 1024 × 1200-pixel resolution and 32-bit color depth. Axial and cross-sectional images were evaluated first, and coronal and sagittal images were reviewed if necessary. Data were extracted and recorded in a checklist. Eight Vertucci’s canal types were assessed, as showed in Figure 1.
To evaluate root filling length, the distance between the apical end of the root and the initiation point of the root filling was measured. Values > 2 mm indicated under-filling, values between 0 and 2 mm designated optimal length of root filling, and extrusion of root filling through the apex was recorded as over-filling [19]. In assessment of homogeneity of the root filling, the presence of voids and irregularities indicated a non- homogenous filling. Periapical lesion was recorded in case of a well-defined periapical radiolucency detected on 3D reconstructions, or the presence of a periodontal ligament space equal or wider than 0.5 mm [19-21]. After observation of the images, disagreements between the observers were resolved by discussion. To assess intra-observer agreement, 30% of the images were evaluated again by the same observer after a 2-week interval. Intra- and inter-observer agreements were estimated using kappa coefficient of agreement.
Frequency and percentage values were reported for qualitative variables, and mean and standard deviation values were reported for quantitative variables. Correlation between endodontic procedural errors and root canal morphology was analyzed using chi-square test with SPSS version 23, at 5% level of significance.

Results

The intra-observer agreement was found to be excellent (0.88). Maxillary premolars, mandibular premolars, maxillary molars, and mandibular molars accounted for 30%, 26.95%, 20.86%, and 22.17% of the teeth under study, respectively. The maxillary premolars had the highest share. Vertucci’s canal types in the maxillary and mandibular molar and premolar teeth are presented in Tables 1 and 2. Also, the distribution of different types of endodontic procedural errors according to the root canal morphology is reported in Table 3.
According to the chi-square test, over-filling (p < 0.05) and missed canals (p < 0.003) showed a significant correlation with the type of root canal. Moreover, the canal type had a significant correlation with the presence of periapical lesion (p < 0.009). Correlation of the presence of a periapical lesion with endodontic procedural errors and frequency distribution of different error types based on tooth types are reported in Tables 4 and 5, respectively. Moreover, procedural errors and canal configurations are demonstrated in Figures 2-6.

Discussion

Periapical radiography has been the most commonly used modality for the detection of endodontic procedural errors and periapical lesions. However, 3D-CBCT with no superimposition of adjacent structures has enabled more accurate evaluation of periapical region [22-24] and quality of root filling [20]. A recent study, in which CBCT was used for evaluation of root canal length reported that this modality had high accuracy for working length determination, especially in the posterior teeth as compared with conventional imaging modalities [25].
In the present study, all disto-buccal and palatal roots of maxillary molars showed the least complexity and were Vertucci’s type I, which was in agreement with findings of the previous studies [26-28]. Here, the mesio- buccal root was Vertucci’s type II in most cases, while Araujo et al. [29] reported that the mesio-buccal root was mostly Vertucci’s type IV in their study.
The root canal morphology, under-filling, and non- homogenous filling were the most frequent procedural errors. However, the most common procedural errors were under-filling and missed canals in a study by Araujo et al. [29]. An interesting finding was that the prevalence of technical errors increased with an increase in the complexity of root canal morphology, which was similar to Nascimento et al. findings. The Vertucci’s type I has one single canal that extends from the pulp chamber towards the apical foramen, and type IV has two separate canals extending from the pulp chamber towards the apex. The other root canal morphologies have branches in their paths, which make them more susceptible to endodontic procedural errors [30].
In the present study, under-filling was the most common error observed in canal types I and II, which was similar to findings of a previous study [29]. In type IV canals, missing a canal was the most frequent error, probably due to not detecting multiple apical branches on periapical radiographs, and mistaking the canal type with Vertucci’s type I [25]. Karabucak et al. [31] reported high prevalence of missed canals in maxillary molars (42.5-59.0%). A relatively similar result was obtained in our study (65.2%).
Teeth with missed canals are more susceptible to the development of periapical lesions compared with endodontically treated teeth with no missed canals. In the present study, the prevalence of periapical lesions associated with a missed canal was 52.1%, which was lower than the rate (87%) reported by Aysal et al. [32]. Araujo et al. [29] noted that in case of occurrence of endodontic technical errors, the risk of development of a radiolucent periapical lesion in type IV canals would be higher than that in type II canals, which was in agreement with the present findings. The reason is that in type II canals, there are two canals that end at one apical foramen, and in case of under-filling of one canal, apical seal may be achieved by appro­priate length root filling of other canal.
One limitation of cross-sectional studies is that they cannot reveal whether the detected periapical lesion is healing or developing [33]. Clinical information obtained from the patient may help in this respect. However, unavailability of such data was a limi­tation of the current study. Therefore, it is important to emphasize that the presence of a periapical lesion alone should not be considered a sign of endodontic treatment failure, and further information regarding clinical history of patients are required to cast a final judgment in this regard [34]. Furthermore, the pre­sence of a periapical radiolucency, when there is no detectable endodontic procedural error, may be attributed to the presence of accessory canals, which may not be visible on CBCT images, except for rare cases with large dimensions [35, 36]. CBCT has limitations in detection of such canals, especially in endodontically treated teeth [29]. Nevertheless, the efficacy of CBCT for pre-operative assessment and treatment planning of teeth suspected of having a complex morphology should be compared with periapical radiography.
According to recent recommendations, CBCT should be used for specific endodontic cases, and not routinely, to prevent unnecessary radiation exposure of patients [37]. In the literature, many studies have proposed CBCT as a reliable modality for the assessment of root canal system, endodontic procedural errors, and periapical lesions [2, 25, 38, 39]. Therefore, CBCT may be recommended when periapical radiographs suggest a complex root canal morphology to prevent possible procedural errors due to root canal complexity.

Conclusions

According to the present results, significant correlations were found between procedural errors, including over-filling and missing a canal with canal type. Moreover, a significant correlation was detected between the canal type and the presence of a periapical lesion, such that the highest frequency of periapical lesions was noted in canal types III and V.

Disclosures

1. Institutional review board statement: The study was approved by the Research Deputy of Hamadan University of Medical Sciences, Hamadan, Iran, with approval No.: 1400.821.
2. Assistance with the article: The authors would like to express their gratitude to the Research Deputy of Hamadan University of Medical Sciences.
3. Financial support and sponsorship: None.
4. Conflicts of interest: The authors declare no potential conflicts of interest concerning the research, authorship, and/or publication of this article.

References