This retrospective study was approved by the Ethics Committee of Peking University School and Hospital of Stomatology (approval number: PKUSSIRB-201949122).
1. Primary teeth pulpectomy
1.1 Participants
The study participants were selected among healthy children below the age of 9 years, who received pulpectomy of the primary teeth in the Department of Pediatric Dentistry at Peking University Hospital of Stomatology. The inclusion criteria were as follows:
1) Teeth treated because of irreversible pulpitis or pulp necrosis, with or without periodontitis, for which radiographic examination showed no involvement of the permanent successor and minimal or no root resorption.
2) Teeth that had not undergone trauma.
3) Teeth that underwent regular postoperative examination with follow-up time ≥1.5 years.
4) Teeth for which complete records were available.
All pulpectomies were completed in compliance with the AAPD guidelines[4]. A six-month visit interval was recommended.
1.2 Data collection
The following information from an electronic medical record system (Beijing Jiahe Meikang Information Technology, Beijing, China) of Peking University Hospital of Stomatology was collected:
- Demographic characteristics.
- Information obtained from the first visit, including the date, rank of attending doctor (intern or expert), tooth position (anterior or posterior), treatment method (under general anesthesia (GA) or not [GA or non-GA]), periapical lesion (yes or no), clinical and radiographic manifestations, root filling material (iodoform zinc oxide paste[15] or Vitapex® [calcium hydroxide and iodoform paste, Neo-Dental, Tokyo, Japan]), and crown restoration material (Resin Filling, Preformed Metal Crown [PMC], or Glass Ionomer Cement [GIC]).
- Information obtained during follow-up, including following visit dates, chief complaints, and clinical and radiographic examination. If the tooth was already missing, the date of loss and the associated symptoms were recorded.
1.3 Clinical and radiological evaluation
The pulpectomy was considered clinically successful in the absence of pain, abnormal mobility, gingival pathology, and severe crown restoration defects necessitating root canal retreatment or extraction, in relation to the tooth. If not, it was labelled as clinical failure.
If the radiographic examination revealed a decrease in size or disappearance of the initial periapical lesions within 6 months[4], with no new appearances of periapical lesions and/or pathological root resorption, the treatment was classified as a radiological success; otherwise it was classified as a radiological failure.
The interventional outcome was defined as a success only if the treatment exhibited both clinical and radiological success. In addition, premature loss and delayed root resorption of primary teeth after treatment were classified as failure[4]. The primary teeth were evaluated for premature loss or delayed root resorption by comparison with the contralateral teeth and/or other adjacent teeth without pulp treatment. If there were no contralateral teeth and adjacent teeth without pulp treatment, it was estimated based on the development stage of the permanent successor.
All periapical films involved in this study were re-examined. Radiographic examinations were independently performed by two pediatric dentists. Cohen’s kappa statistic showed excellent reproducibility between the two investigators, with a measurement agreement of 0.85 and intra-examiner reliability (over two weeks) of 0.87.
2 Root canal morphology analysis
2.1 Data collection
Data of CBCTs, which were originally conducted for diagnosis, treatment, or regular examinations, were retrieved from the Department of Radiology at the Peking University Hospital of Stomatology from March 2015 to November 2019. Radiographic images of systemically healthy children aged 3–8 years, were selected. The maxillary first primary molars of these children meeting the following criteria were included:
- no abnormality in the crown and root morphology;
- no pulpal involvement and periapical lesions;
- had not undergone pulpotomy, pulpectomy or PMC restoration;
- the development of the roots was complete, without internal or external absorption;
- satisfactory image quality.
CBCT images were acquired using the 3D Accuitomo type F17 (Morita, Kyoto, Japan) at 80–90 kV and 5 mA. The voxel and slice thicknesses were both 0.125 mm. All CBCT exposures were performed by a licenced radiologist strictly in accordance with the manufacturer’s recommended protocol.
2.2 Measurements
The CBCT scan data were saved in DICOM format and transferred to Mimics 17.0 (Materialise Technologies, Leuven, Belgium). Several points and values were defined as below, to aid with the measurement of each root and canal (Fig. 1a).
Point A: the lowest point of the outer surface of the root furcation.
Point B: the anatomical root tip.
Point C: the point where the canal disappeared towards the root tip for the first time.
LR: length of the root measured from points A to B along the tooth axis.
LC: length of the canal measured from points A to C along the tooth axis.
Plane D: canal cross-section, 1/3*LR away from point A along the tooth axis.
If the root canal cavity reappeared after disappearing toward the root tip, it was recorded as a discontinuous canal, as shown in Fig. 1a. The LC and LR were measured.
The canal cross-section at plane D was reconstructed. First, an appropriate threshold for accurate segmentation of the canal cavity was selected for each canal. The threshold value was determined as follows. We first measured the grey value range of the lip in the CBCT at three different positions and calculated the difference between the minimum and maximum grey values for each position. Then, the mean of the three differences was calculated and recorded as “d”. The minimum grey value of the canal image was recorded as “c”, and d plus c was “p”, which was recorded as the reference threshold value of canal. Following this, the reference grey value “p” was slightly adjusted for improved accuracy according to the operator’s judgement.
After determination of a proper canal threshold, a three-dimensional (3D) model of the canal section at plane D was generated and measured, as shown in Fig. 1b. The detailed procedure is listed in Fig. 2. STL (stereo lithography) is a 3D graphics file format for rapid prototyping manufacturing.
The following outcome variables were calculated and recorded:
1) LC/LR: Ratio of the canal length and root length.
2) Ndis: The number of discontinuous root canals.
3) R.15: The ratio of the root canal wall area with an inner diameter lesser than 0.15 mm and the total root canal wall area at plane D.
4) Area of canal at plane D, measured using Image J (National Institutes of Health, Bethesda, USA)
Before starting the study, 69 canal cross-section samples not included in this study were used to assess the consistency of image segmentation with the index of mean intersection over union (MIoU). The MIoU was 0.81.
3 Statistical analysis
Statistical analyses were conducted using SPSS version 20.0 (SPSS Inc., Chicago, IL, USA). The survival rate of teeth after pulpectomy was determined using the Kaplan-Meier method. Univariate and multivariate analyses were performed using Cox regression analysis. Variables with P<0.1 in the univariate analysis were included in the multivariate analysis to be explored as possible risk factors. Factors previously reported to influence success were also included. The variables included in the Cox analysis were age, rank of attending doctor, treatment method, presence of periapical lesion, root canal filling material, and the crown restoration material. The Wald test (Backward: Wald) method was used, and the significance level of the selected and excluded variables was 0.05.
The normality of data from CBCT was tested using the Kolmogorov–Smirnov one-sample test. The Spearman correlation coefficients between R.15, LC/LR, area of canal at plane D, and age were calculated, and the linear regression coefficient between Ndis and age was calculated by binary logistic regression. The significance level was set at P<0.05.