A Protocol for the Development of a New Three‐Dimensional Classification of Dens Invaginatus Using Cone Beam Computed Tomography Imaging

Dens invaginatus (DI) is a dental developmental anomaly that arises when the enamel organ folds into the dental papilla during the morpho-differentiation phase of tooth development (Hülsmann 1997). Several factors may contribute to this disruption in enamel organ development, including genetic predisposition (Sarraf-Shirazi et al. 2010), physical trauma, infections, and irregular tooth bud growth (Hülsmann 1997). Overall prevalence rates of DI vary widely, ranging from 0.39% to 13.5% of teeth (Capar et al. 2015; Hegde et al. 2022). However, more recent data utilising three-dimensional (3D) imaging techniques suggest an overall global prevalence of approximately 7% across all teeth, with no significant difference between males and females (Dos Alves Santos et al. 2023). The maxillary lateral incisors are the most frequently affected teeth, exhibiting a prevalence of approximately 5%, with unilateral presentations being more common (Dos Alves Santos et al. 2023). The hallmark feature of DI is the presence of an enamel-lined invagination within the tooth that opens into the oral cavity. Clinically, the entrance to this invagination often appears as a pit, groove, or a deep foramen coecum, typically located on the palatal or occlusal surface of the crown (Siqueira Jr et al. 2022). The invagination is usually lined with enamel; however, histological, radiographic, and microscopic investigations of DI have reported inconsistent observations. Deficiencies or defects in the enamel lining of the invagination can expose the underlying intact dentine (Kramer 1953) with several studies reporting irregularities in the structure of the surrounding dentine leading to pulp communications with the oral cavity (Vincent-Townend 1974; Beynon 1982; Piatelli and Trisi 1993). The invagination can extend down to the apical third of the root and communicate with the periodontal ligament through a pseudo-foramen either apically or laterally. Thus, DI, along with its anomalous clinical characteristics, is often associated with caries, pulpal infection, and periradicular pathosis (Nosrat and Schneider 2015; Hülsmann 1997; Siqueira Jr et al. 2022) because the invagination, after eruption, is filled with saliva, bacteria, and debris (Rushton 1958). Given the complex relationship between the root canal and the invagination, variations in their spatial relationship occur. Using cross-sectional computed tomography, Zhu et al. (2017) defined several variations: (a) the invagination being in the centre and surrounded by the main root canal on all sides, (b) the invagination expands and displaces the main canal, resulting in a C-shaped canal system, (c) the invagination occupies the centre of the tooth and the main canal is displaced into a crescent shape on either side of the invagination, (d) the main canal and invagination are separate without any communication, with the invagination being present on the lateral side of the root, (e) the main canal lies in the centrally with no invagination visible in this section. Multiple categorizations of DI have been proposed based on the two-dimensional (2D) radiographic appearance of the invagination. Initially, four categories were proposed based on clinical and radiographic characteristics (Hallett 1953) with several further classifications being proposed in the subsequent decades (Oehlers 1957; Ulmansky and Hermel 1964; Schulze and Brand 1972; Vincent-Townend 1974). Currently, the clinical classification proposed by Oehlers (1957) is the most commonly used. Oehlers (1957) categorised DI into three basic types based on the vertical extent of the invagination as it appears on conventional 2D radiographs: ‘Type I—the invagination is confined to the tooth crown; Type II—the invagination extends apically to the root and ends in a “blind sac”; Type IIIa—the invagination goes beyond the cementoenamel junction and communicates laterally with the periodontal ligament through a pseudo-foramen; and Type IIIb—the invagination extends apically of the root and communicates with the periodontal ligament through the apical foramen’. Based on CBCT imaging, several modifications to Oehlers classification have been proposed in order to incorporate the wide range of anatomical variations that occur. For example, Type II has been subcategorised into three types based on the extent of the invagination along the root length (Kritika et al. 2022). Gul et al. (2020) proposed a modification to Type IV based on the cross-sectional anatomy of the tooth observed on CBCT images when the main canal is displaced laterally with both the invagination and the main canal having a communication with the periodontal ligament (PDL). In recent years, cone beam computed tomography (CBCT) imaging has become an invaluable tool in assessing the extension of an invagination, its communication between the surrounding tissues, and the presence of apical and periodontal lesions (Alkadi et al. 2021). A retrospective analysis of data from 300 patients revealed that the examination of two-dimensional images (OPG) only allowed the recognition of a low prevalence of DI (3%) in comparison to that of CBCT images (10.7%) (Capar et al. 2015). The deficiencies in the use of 2D radiographs were further emphasised in a retrospective study by Yalcin et al. (2022), who demonstrated that when interpreting panoramic radiographs, only 73.97% of teeth with DI were diagnosed when compared to CBCT. Thus, CBCT offers an advantage for the diagnosis, classification, and management of DI and is critical in treatment planning as well as determining whether modern endodontic techniques such as 3D guides for access cavity preparation would facilitate treatment (Abella Sans et al. 2025). It is obvious that the deficiencies of existing DI classifications based on 2D imaging can be overcome by a more comprehensive system based on the information gained from 3D CBCT images, which have the potential to offer more detailed and clinically relevant data. Although others have made use of CBCT data, the modifications to the classification they suggested did not involve a fundamental re-evaluation of the Oehlers system and did not include a comprehensive reassessment of the wide range of parameters involved. It is now time to move beyond 2D-based classifications. In summary, a robust, universally accepted 3D classification has the potential to transform both clinical practice and research. Therefore, the objective of this editorial is to outline the methodology to be followed in the development of a novel classification system for defining DI based on 3D imaging. The project directors (V.N., P.D.) recognised the necessity of a new 3D classification system for DI to benefit clinicians, researchers and ultimately patients. The steering committee established by the project directors will be responsible for the development of the initial draft classification and the refinement of the consensus-building process using the expertise of endodontic specialists, oral radiologists, general dental practitioners, and endodontic postgraduate students. The steering group will establish the classification criteria that are measurable, reproducible, consistent, and transparent, and that have an impact on the diagnosis, treatment, and prognosis of DI. The initial draft classification will be developed by the steering committee using a range of criteria, for example, the apical extent of the invagination, the cross-sectional anatomy of the invagination and its spatial relationship with the root canal system, communication with the periodontal ligament, communication with the pulp, the apical anatomy of the canal and root apex, and the condition of the tooth and surrounding tissues. The steering group will establish a panel of experts to solicit their opinions on the clarity, relevance, and utility of the draft classification using an online Delphi methodology. Delphi methodology is a structured and multi-round process that is employed to achieve a valid and reliable consensus that entails a series of questionnaires completed by a group of individuals (Hasson et al. 2000). The panel will consist of 17 academic or clinical endodontic specialists, 3 specialists in oral radiology, 3 endodontic postgraduate students, and 2 general practitioners. The experts will be chosen from across the world in order to provide a global perspective. Each Delphi panel member will provide their personal assessment of the clarity, relevance and specific text/wording of each category of the new classification system using a dichotomous scale (yes/no) and 9-point Likert scale (1—definitely excluded to 9—definitely included). Each category in the new classification will be included if it receives a score of 7–9 from at least 70% of members and a score of 1–3 from less than 30% of members. Categories will be excluded if they receive a score of 1–3 from over 70% of members and a score of 7–9 from less than 30% of members (Agha et al. 2017). The criteria and detailed text of each category will be revised in light of the anonymous feedback provided by the members and incorporated into each subsequent round of the Delphi exercise (Agha et al. 2017). The panel will also have the opportunity to suggest new categories for the classification. Using anonymized CBCT images of teeth with DI, the consistency and reproducibility of the new classification system will be evaluated and assessed independently by five endodontists with a minimum of 5 years of academic or clinical experience. The CBCT images utilised will be procured from existing images stored at the University of Sharjah, other universities, or from dental practices. No subject will undergo a CBCT scan for the purposes of the study. Thus, the CBCT images will have been prescribed for other purposes, such as the removal of impacted teeth, or diagnostic and treatment planning purposes. The identities of the participants whose CBCTs will be evaluated will not be disclosed, and the images will not contain any identifiable features. As a result of the retrospective nature of this project, it will not be possible to obtain informed consent from the individuals. Each assessor will be instructed to classify the type of DI according to the proposed classification using the CBCT images. The inter-rater reliability among the five assessors will then be assessed using Kappa statistics and other relevant statistical tests. If the inter-rater reliability between assessors is low, the steering committee will revise the classification, and the process will continue until the inter-rater reliability between assessors is adequate. Based on the comments received from the Delphi process and reliability test, the steering committee will amend and finalise the classification. The steering group will publish the new classification in the form of a manuscript in a suitable journal. The steering group will also present the classification at international and national congresses for the benefit of undergraduate and postgraduate students, academics, clinicians, and researchers. The development and validation of a new 3D classification for DI will address the shortcomings of traditional systems by incorporating CBCT-based data and clinically relevant parameters. It will ultimately result in enhanced patient and clinician-reported outcomes by providing a standardised and comprehensive approach to diagnosing and managing this complex anomaly. This initiative is indicative of a more comprehensive endeavour to establish a more precise and clinically meaningful framework for future research and practice in the management of this complex anomaly. The authors acknowledge that the accuracy of CBCT images used to define DI is influenced by factors such as field of view, resolution, and voxel size. All the authors made substantial contributions to the manuscript. All the authors have read and approved the final version of the manuscript. The authors declare no conflicts of interest. The authors have nothing to report.