The outcomes of nonsurgical root canal treatment and retreatment assessed by CBCT: a systematic review and meta-analysis

Abstract

Cone-beam computed tomography (CBCT) has been demonstrated to identify apical radiolucency with higher accuracy than two-dimensional radiography. The outcome of root canal treatment varies depending on the imaging modality employed. This systematic review aimed to: (1) estimate the success rate of nonsurgical root canal treatment and retreatment when assessed by CBCT and (2) investigate the influence of some factors suspected to be associated with treatment outcomes. An electronic search was performed in the following databases: MEDLINE, Embase, Web of Science, the Cochrane Library, and gray literature. Article selection and data extraction were independently conducted by 2 reviewers. The terms ‘strict’ (complete resolution of periapical lesion) or ‘loose’ (reduction in size of existing periapical lesion) were used to describe the outcome criteria. The Grading of Recommendations, Assessment, Development and Evaluation (GRADE) approach was used to evaluate the quality of the included studies. Meta-analysis and meta-regression established pooled outcome rates, 95% confidence intervals (CIs), and significant clinical prognostic factors (P < .05). Nineteen articles were included. The overall quality of evidence was moderate. The estimated weighted pooled overall success rates for NSRCT and NSReRCT assessed by CBCT were 41.03% (95% CI: 28.68%–53.95%; I² = 94.76%) under strict criteria and 85.01% (95% CI: 80.85%–88.75%; I² = 68.36%) under loose criteria. This review’s findings offer valuable insights to guide the design of future studies assessing root canal treatment and retreatment outcomes through CBCT. However, they should be interpreted with caution due to the retrospective and heterogeneous nature of the data. Review registration: This protocol was registered in the international prospective register of systematic reviews: the PROSPERO database (CRD42024591017). Clinical trial number: Not applicable.

Supplementary Information

The online version contains supplementary material available at 10.1007/s44445-025-00021-2.

Background

The primary goal of root canal treatment is to eliminate microbial infection within the root canal system, reducing it to a level compatible with tissue healing. This is achieved by thoroughly cleaning and shaping, then placing a durable root filling to prevent reinfection. This process aims to either prevent the onset of apical periodontitis (AP) or resolve existing AP, ultimately preserving the tooth’s natural function (Gilheany et al. 1994).

The success of root canal treatment is evaluated based on how effectively these objectives are met. Treatment is deemed successful if the tooth remains free of clinical symptoms at follow-up intervals of 1 to 4 years and radiographs show normal periapical tissues. Conversely, the presence or persistence of apical radiolucency indicates treatment failure, which may necessitate further clinical intervention (ESE 2006).

Traditionally, periapical radiography has been used to evaluate the outcome of root canal treatment. However, periapical radiographs are inherently limited in diagnostic accuracy due to their 2D nature, geometric distortions, and anatomical noise (Patel and Horner 2009). In vivo and ex vivo studies showed that lesions located within the cancellous bone might be misdiagnosed by periapical radiography. Although lesions of a specific size can be identified in areas with a thin cortical bone layer, the same-sized lesions may remain undetectable in regions with thicker cortical bone (Estrela et al. 2008).

Cone beam computed tomography (CBCT) has been widely used in medicine and dentistry since the 1980s (Brenner and Hall 2007; Tachibana and Matsumoto 1990). In many situations where periapical radiolucency is not visible on radiographs, CBCT can identify periapical lesions. CBCT has been demonstrated to identify apical radiolucency with higher accuracy than two-dimensional radiography. When histopathology was utilized as a reference standard, CBCT outperformed periapical radiography in the diagnosis of apical radiolucency in human cadavers. It is also demonstrated that the results of root canal treatment vary depending on the imaging modality employed; periapical radiographs demonstrate 79% favourable outcomes, whereas CBCT only shows 35% (Vandenberghe et al. 2008).

Findings from outcome studies empower clinicians to provide patients with accurate information regarding the prognosis of different treatment options. Several published systematic reviews have assessed the outcome of root canal treatment. Nevertheless, findings from periapical radiographs have been used in the previously published reviews (Sabeti et al. 2024; Ng et al. 2007; Ng et al. 2008; Burns et al. 2022). This review aimed to: (1) estimate the success rate of nonsurgical root canal treatment and retreatment when assessed by CBCT and (2) investigate the influence of some factors suspected to be associated with treatment outcomes.

Methods

A detailed protocol for this systematic review and meta-analysis was defined, following the guidelines of the Preferred Reporting Items for Systematic reviews and Meta-Analyses Protocols (PRISMA-P) statement (Page et al. 2021). This protocol was registered in the international prospective register of systematic reviews: the PROSPERO database (CRD42024591017). The PRISMA checklist is included in Supplementary Table S1.

Focused question

The research question was developed as follows:

1. Participants: Permanent teeth which underwent NSRCT or NSReRCT.

2. Intervention: NSRCT or NSReRCT assessed by CBCT.

3. Comparison: None.

4. Outcome: Periapical healing assessed via CBCT.

Literature search

A comprehensive literature search was undertaken until 31 December 2023 with no start date limit. The following four electronic databases were searched: EMBASE, Web of Science, PubMed (including MEDLINE) and the Cochrane Central Register of Controlled Trials (CENTRAL). The search strategy for the four databases was constructed with a librarian. A manual search of Google Scholar was also conducted to detect relevant unpublished manuscripts, research reports, doctoral dissertations and other grey literature, with complementary searches through Open Grey (http://www.opengrey.eu), Networked Digital Library of Theses and Dissertations (http://www.ndltd.org), and Open Access Theses and Dissertations (https://oatd.org). There were no language restrictions applied. The following keywords and phrases were used in the search strategy: root canal treatment, root canal therapy, endodontic treatment, retreatment, recurrence, previously treated, Cone-Beam Computed Tomography, CBCT, 3D imaging, treatment outcome, radiographic evaluation, success rate, prognosis, healing, evaluation, assessment, failure, and survival. These keywords were adjusted and tailored to each database to ensure comprehensive coverage of relevant studies (Supplementary Table S2).

Inclusion criteria

Studies were selected based on the following inclusion criteria:

1. Study design: Prospective or retrospective cohort studies and randomized controlled trials.

2. A stratified analysis of NSRCT or NSReRCT outcome is available if other treatment types have been included.

3. Sample size provided.

4. Radiographic success was determined using CBCT by either ‘strict’ or ‘loose’ criteria:

   • (i)
     ‘Strict’ radiographic criteria of success: the absence of apical radiolucency at follow-up examination.
   • (ii)
     ‘Loose’ radiographic criteria of success: reduction in the size of apical radiolucency at follow-up examination.

5. Overall success rate given or could be calculated from the raw data provided.

6. Outcomes were assessed after at least 6 months of follow-up.

7. Additional criteria: the study was available in English or translated by the author, the clinical protocol was clearly described, the data was not duplicated from a previously included study, and the study met appropriate quality standards.

Studies on primary dentition, irrelevant designs, animal/microbiological studies, ex vivo/in vitro studies, case reports, and review articles were excluded.

Study selection and extraction

The selection of studies was carried out in two distinct phases. All the results from the database searches were exported to Endnote X9 (Clarivate, London, UK). Duplicate results were immediately removed. All titles and abstracts of retrieved manuscripts were assessed independently and duplicated by two reviewers (LA and RA). Publications that did not meet the eligibility criteria listed above were excluded, and full texts of initially selected articles were obtained for further in-depth evaluation. In the second phase of study selection, the selected articles were independently assessed by the two reviewers (LA and RA), and data extraction was performed (Supplementary Table S3). Disagreements between the two reviewers were resolved through discussion, and when consensus could not be reached, a third reviewer was consulted.

Risk of bias assessment

Two reviewers (LA and RA) rated the risk of bias. Randomized controlled trials were assessed using the Cochrane Risk of Bias tool (RoB 2), and cohort studies were examined using the Cochrane Risk of Bias tool for non-randomized trials (ROBINS-I) (Table 1). A classification of “low risk” was assigned if all domains were judged as having a low bias risk, “moderate risk” if any domain had minor concerns, and “high risk” if any domain had major concerns. Disagreements were resolved by discussion.

Table 1.

Risk of bias assessment

Quality of evidence

The quality of evidence was assessed using the five domains of Grading of Recommendations, Assessment, Development and Evaluation framework, as presented in Supplementary Table S4. This method categorized evidence quality as high, moderate, or low.

Quantitative analyses

Statistical analysis utilized Stata version 14 (StataCorp., College Station, TX, USA). Pooled success rates with 95% confidence intervals (CIs) were calculated. Heterogeneity, measured by the I² index, was assessed using the Cochran Q test, with I² > 50% indicating high heterogeneity. Publication bias was examined using funnel plots and Egger’s test. In the presence of publication bias, outcome rates were adjusted using the trim-and-fill method. To investigate potential sources of statistical heterogeneity among study characteristics, meta-regression models were employed. P < 0.05 was considered statistically significant.

Results

Search results

The final list of articles generated included 1279 studies. After title and abstract screening, 32 articles were obtained for full-text review. After a full-text review, 19 articles remained for inclusion in this systematic review (Fig. 1). The excluded studies at the full-text stage are listed in Supplementary Table S5. The publication date ranged from 2011 to 2022. A total of 1365 teeth were included in the meta-analysis.

Fig. 1.

The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flowchart for the study selection

Quality of evidence

The quality of evidence of the 19 included studies was assessed using the GRADE approach (Supplementary Table S4). Two studies, both randomized controlled trials, were considered to provide ‘high’ quality evidence — one investigating the effect of enhanced infection control protocols (Zahran et al. 2021) and the other examining ultrasonic activation of irrigants (Liang et al. 2013). Ten studies were rated as providing ‘moderate’ quality evidence (Verma et al. 2020; Borden et al. 2013; Knight et al. 2020; Fernandez et al. 2013, 2017; Restrepo-Restrepo et al. 2019; Patel et al. 2012; Zhang et al. 2021; Al-Nuaimi et al. 2017; Davies et al. 2016), while seven were classified as ‘low’ quality (Castro et al. 2016; Gudac et al. 2022; Zavattini et al. 2020; Saidi et al. 2015; Liang et al. 2011; Curtis et al. 2018; Metska et al. 2013). Factors contributed to downgrading included small sample sizes (n = 2) (Verma et al. 2020; Castro et al. 2016), limited number or lack of calibration of radiographic observers (n = 2) (Castro et al. 2016; Saidi et al. 2015), low re-examination rates (< 50%) (n = 1) (Liang et al. 2011), and incomplete reporting of treatment procedures (e.g. rubber dam, magnification, treatment techniques, or number of visits) (n = 7) (Verma et al. 2020; Fernandez et al. 2013, 2017; Restrepo-Restrepo et al. 2019; Castro et al. 2016; Zavattini et al. 2020; Saidi et al. 2015). Additional concerns involved insufficient reporting on subject loss (n = 1) (Saidi et al. 2015), unclear analysis of extracted or retreated teeth (n = 1) (Fernandez et al. 2017), and combining outcomes for primary and secondary root canal treatments into a single group (Borden et al. 2013; Gudac et al. 2022).

Methodological characteristics of the included studies

The characteristics of the included studies are presented in Table 2 and summarized below:

Table 2.

Characteristics of the included studies

Author year	Study design	Country	Operator	Sample size	Recall rate (%)	Outcome assessment	CBCT outcome assessment method	Radiographic assessment criteria	Follow up	≥ 2 Radiographic observers	Calibration	Reliability test	Statistical analysis
Gudac 2022	Prospective	Lithuania	Endodontist	176 teeth 403 roots	N/A	C + L	COPI index (diameter of radiolucency)	S	2 years	__ No	✓	✓	χ2
Zahran 2021	RCT	United Kingdom	Endodontist & postgrad	115 teeth	80	C + L	6-point scale	L	1 year	✓	✓	✓	Fisher’s exact Mann Whitney Log. Reg
Zhang 2021	Prospective	China	Endodontist	58 teeth	64	C + L	Lesion volume	S + L	4 years	✓	✓	✓	χ2 Wilcoxon Fisher’s exact Log. Reg
Verma 2020	RCT	India	Missing	57 teeth	89	C + L	CBCT PAI index	S + L	1 year	N/A	N/A	✓	χ2
Knight 2020	Prospective	United Kingdom	Postgrad	65 teeth	74	C + L	6-point scale	S + L	1 year	✓	✓	✓	χ2 McNemar’s Fisher’s exact Log. Reg
Zavattini 2020	NRCT	United Kingdom	Endodontist & postgrad	104 teeth	69	C + L	Visual assessment	L	1 year	✓	✓	N/A	Fisher’s Exact
Restrepo 2019	Ambispective	Columbia	Postgrad	125 teeth	83	C + L	CBCT PAI index	S + L	2 years	✓	✓	✓	χ2 Fisher exact Log. Reg
Fernandez 2017	Ambispective	Colombia	Postgrad	208 roots 132 Teeth	100	C + L	CBCT PAI index	S	10 years	✓	✓	✓	the Kaplan–Meier method the Pearson or Fisher exact test, and the Cox proportional hazards model multivariate survival analyses with stepwise extended Cox proportional hazards regression Hazard ratios (HRs)
Fernandez 2013	Ambispective	Colombia	Postgrad	208 roots 132 Teeth	58	C + L	CBCT PAI index	S	5 years	✓	✓	✓	χ2 odds ratio
Del castro 2016	RCT	Brazil	Endodontist	26 teeth	93	C + L	Lesion volume	S + L	1 year	__ No	✓	✓	χ2 Mann Whitney T-test Cramer’s V
Saidi 2015	Ambispective	Beirut	Postgrad	156 teeth	52	C + L	Visual assessment	S	5 years	__ No	N/A	N/A	Fisher Exact McNemar
Van der bordon 2013	Ambispective:	Amsterdam	Missing	71 roots 50 teeth	75	C + L	Lesion volume	S + L	10–37 months	__ No	N/A	✓	χ2 McNemar
Liang 2013	RCT	China	Missing	84 teeth	82	C + L	Lesion volume	S + L	10–19 months	✓	N/A	✓	χ2 T test McNemar Log. Reg
Patel 2012	Prospective	United Kingdom	Missing	218 roots 123 teeth	82	C + L	6-point scale	S + L	1 year	✓	✓	✓	χ2 McNemar’s
Liang 2011	Ambispective	China	Endodontist	143 roots 115 teeth	36	C + L	6-point scale	S	2 years	✓	N/A	✓	χ2 Log. Reg
Curtis 2018	Ambispective	USA	Postgrad	68 teeth	N/A	C + L	Lesion volume	S + L	22 months	✓	✓	N/A	N/A
Alnuaimi 2017	Prospective	United Kingdom	Postgrad	137 teeth	88	C + L	6-point scale	S + L	1 year	✓	✓	✓	χ2 McNemar’s Fisher’s exact Log. Reg
Davies 2015	Prospective	United Kingdom	Endodontist & postgrad	98 teeth 215 roots	86	C + L	6-point scale	S + L	11–18 months	✓	✓	✓	χ2 McNemar
Metska 2013	Prospective	Amsterdam	Postgrad	35 teeth	78	C + L	Lesion volume	S + L	1 year	✓	✓	✓	Wilcoxon

N/A, not available; RCT, randomized controlled trial, USA, United States of America; C, clinical assessment; R, radiographic assessment; S, Strict criteria for radiographic assessment; L, loose criteria for radiographic assessment; Log. Reg.: logistic regression; χ2: Chi-square

Among the 19 included studies, four were randomized controlled trials (Zahran et al. 2021; Liang et al. 2013; Verma et al. 2020; Castro et al. 2016), one was a non-randomized controlled trial (Zavattini et al. 2020), seven were ambispective cohort studies (Borden et al. 2013; Fernandez et al. 2013, 2017; Restrepo-Restrepo et al. 2019; Saidi et al. 2015; Liang et al. 2011; Curtis et al. 2018), and seven were prospective cohort studies (Knight et al. 2020; Patel et al. 2012; Zhang et al. 2021; Al-Nuaimi et al. 2017; Davies et al. 2016; Gudac et al. 2022; Metska et al. 2013).

Sample sizes varied from 26 (Castro et al. 2016) to 176 teeth (Gudac et al. 2022). In 12 studies, teeth were the unit of analysis (Zahran et al. 2021; Liang et al. 2013; Verma et al. 2020; Knight et al. 2020; Restrepo-Restrepo et al. 2019; Zhang et al. 2021; Al-Nuaimi et al. 2017; Castro et al. 2016; Zavattini et al. 2020; Saidi et al. 2015; Curtis et al. 2018; Metska et al. 2013), whereas in 4 studies, it was roots (Borden et al. 2013; Fernandez et al. 2013; Gudac et al. 2022; Liang et al. 2011). The other three studies used both teeth and roots as units of analysis (Fernandez et al. 2017; Patel et al. 2012; Davies et al. 2016).

Seventeen out of nineteen studies reported recall rates, either at the tooth or patient level, ranging from 36 to 100% and having a median of 80% (Zahran et al. 2021; Liang et al. 2013, 2011; Verma et al. 2020; Borden et al. 2013; Knight et al. 2020; Fernandez et al. 2013, 2017; Restrepo-Restrepo et al. 2019; Patel et al. 2012; Zhang et al. 2021; Al-Nuaimi et al. 2017; Davies et al. 2016; Castro et al. 2016; Zavattini et al. 2020; Saidi et al. 2015; Metska et al. 2013).

All studies assessed treatment outcomes through both clinical and radiographic evaluations. Most studies (n = 15) used preoperative and postoperative CBCT imaging for radiographic evaluation (Zahran et al. 2021; Liang et al. 2013; Verma et al. 2020; Borden et al. 2013; Knight et al. 2020; Restrepo-Restrepo et al. 2019; Patel et al. 2012; Zhang et al. 2021; Al-Nuaimi et al. 2017; Davies et al. 2016; Castro et al. 2016; Gudac et al. 2022; Zavattini et al. 2020; Curtis et al. 2018; Metska et al. 2013), while four studies relied solely on postoperative CBCT (Fernandez et al. 2013, 2017; Saidi et al. 2015; Liang et al. 2011). For radiographic assessments, five studies exclusively applied strict criteria, requiring complete resolution of periapical radiolucency (Fernandez et al. 2013, 2017; Gudac et al. 2022; Saidi et al. 2015; Liang et al. 2011), whereas two studies used only loose criteria, focusing on the reduction in the size of the radiolucency (Zahran et al. 2021; Zavattini et al. 2020). In the remaining 12 studies, outcomes were calculated using both strict and loose criteria (Liang et al. 2013; Verma et al. 2020; Borden et al. 2013; Knight et al. 2020; Restrepo-Restrepo et al. 2019; Patel et al. 2012; Zhang et al. 2021; Al-Nuaimi et al. 2017; Davies et al. 2016; Castro et al. 2016; Curtis et al. 2018; Metska et al. 2013).

In most studies (n = 14), at least two observers assessed radiographic outcomes (Zahran et al. 2021; Liang et al. 2013, 2011; Knight et al. 2020; Fernandez et al. 2013, 2017; Restrepo-Restrepo et al. 2019; Patel et al. 2012; Zhang et al. 2021; Al-Nuaimi et al. 2017; Davies et al. 2016; Zavattini et al. 2020; Curtis et al. 2018; Metska et al. 2013). Four studies used only one observer (Borden et al. 2013; Castro et al. 2016; Gudac et al. 2022; Saidi et al. 2015), while one study did not specify the number of observers (Verma et al. 2020). Radiographic observers were reported to be calibrated, and/or inter-observer reliability calculations were performed in 17 studies (Zahran et al. 2021; Liang et al. 2013, 2011; Borden et al. 2013; Knight et al. 2020; Fernandez et al. 2013, 2017; Restrepo-Restrepo et al. 2019; Patel et al. 2012; Zhang et al. 2021; Al-Nuaimi et al. 2017; Davies et al. 2016; Castro et al. 2016; Gudac et al. 2022; Zavattini et al. 2020; Curtis et al. 2018; Metska et al. 2013).

The included studies utilized various methods to assess periapical status through CBCT. Gudac et al. (2022) employed the Complex Periapical Index (COPI) to categorize the size of periapical radiolucency based on its diameter (Gudac et al. 2022). Six studies focused on measuring the volume of the radiolucency, with three of them using a 20% cutoff point for volume changes (Liang et al. 2013; Borden et al. 2013; Zhang et al. 2021). In comparison, the other three studies did not specify a cutoff (Castro et al. 2016; Curtis et al. 2018; Metska et al. 2013). Five studies used visual assessments with a 6-point scale (Zahran et al. 2021; Knight et al. 2020; Patel et al. 2012; Al-Nuaimi et al. 2017; Davies et al. 2016), and three additional studies used visual assessment methods without employing a specific scale (Zavattini et al. 2020; Saidi et al. 2015; Liang et al. 2011). The CBCT PAI index was used in four studies, but the interpretation of the CBCT PAI scores varied. For instance, Verma et al. (2020) considered scores 0, 1, and 2 indicative of normal periapical tissues (Verma et al. 2020), Restrepo-Restrepo et al. (2019) classified only scores 0 and 1 as normal (Restrepo-Restrepo et al. 2019), Fernandez et al. (2013, 2017) interpreted only score 0 as normal periapical tissues (Fernandez et al. 2013, 2017).

The included studies used the chi-square test, Fisher’s exact test, Mann–Whitney U test, and t-test to assess prognostic factors. Eight studies also included multivariate analyses (Zahran et al. 2021; Liang et al. 2013, 2011; Knight et al. 2020; Fernandez et al. 2013, 2017; Restrepo-Restrepo et al. 2019; Zhang et al. 2021; Al-Nuaimi et al. 2017).

Overall success rates

The success rates reported from the included studies ranged from 63 to 100% based on loose criteria and from 0% to 81.3% based on strict criteria. Meta-analyses estimated the weighted pooled overall success rates of NSRCT and NSReRCT assessed using CBCT (from 13 studies (Liang et al. 2013; Verma et al. 2020; Knight et al. 2020; Fernandez et al. 2017; Restrepo-Restrepo et al. 2019; Patel et al. 2012; Zhang et al. 2021; Al-Nuaimi et al. 2017; Davies et al. 2016; Castro et al. 2016; Saidi et al. 2015; Curtis et al. 2018; Metska et al. 2013)) to be 41.03% (95% CI: 28.68%–53.95%; I² = 94.76%) under strict criteria and 85.01% (95% CI: 80.85%–88.75%; I² = 68.36%) under loose criteria (from 13 studies (Zahran et al. 2021; Liang et al. 2013; Verma et al. 2020; Knight et al. 2020; Restrepo-Restrepo et al. 2019; Patel et al. 2012; Zhang et al. 2021; Al-Nuaimi et al. 2017; Davies et al. 2016; Castro et al. 2016; Zavattini et al. 2020; Curtis et al. 2018; Metska et al. 2013)) (Fig. 2).

Fig. 2.

Meta-analysis of studies reporting on the outcome of nonsurgical root canal treatment/retreatment using CBCT under ‘strict’ and ‘loose’ criteria

Effects of clinical characteristics on success rates

Table 3 reports the weighted pooled success rates by study characteristics and outcome criteria.

Table 3.

Pooled weighted outcomes by clinical factors

Subgroup		Success
		Strict				Loose
		Number of studies	Number of teeth	Pooled rate (%) (95% CI)	I2 index	Number of studies	Number of teeth	Pooled rate (%) (95% CI)	I2 index
Treatment procedure	NS-RCT	8	750	42 (27–58)	94.96%	8	699	88 (83–92)	69.29%
	NS-ReRCT	5	396	39 (19–62)	95.08%	5	396	80 (73–86)	57.72%
Preoperative apical radiolucecny	Present	10	720	35.19 (22.00–49.60)	93.39%	12	843	84.28 (78.68–89.20)	76.14%
	Absent	5	223	88.47 (73.37- 98.33)	80.48%	6	148	92.84 (84.45–98.58)	43.67%
Extent of root filling	Adequate	2	126	64.30 (55.57–72.59)	0%	2	164	81.48 (75.08–87.15)	0%
	Short	2	15	32.30 (8.30–61.09)	0%	1	5	60.00 (23.07–88.24)	0%
	Long	2	42	62.41 (46.41–76.97)	0%	1	1	100.00 (20.65–100.00)	0%
Number of visits	Single	5	360	42.53 (17.22–70.03)	96.16%	5	439	86.19 (79.26–91.98)	68.67%
	Multiple	8	556	43.58 (30.04–57.61)	90.36%	6	439	84.25 (79.34–88.64)	36.90%
Tooth type	Anterior	4	139	82.22 (60.51–96.84)	87.46%	2	43	94.21 (84.13–99.83)	0%
	Posterior	5	436	52.32 (40.37–64.13)	83.31%	4	397	82.69 (77.11–87.65)	46.44%
Quality of coronal restoration	Adequate	5	430	52.83 (30.37–74.72)	95.37%	4	368	86.97 (77.94–94.00)	78.73%
	Inadequate	3	34	13.28 (0.94–32.26)	0%	2	28	78.71 (60.78–92.75)	0%
Irrigation	Activated	6	417	48.58 (30.90–66.43)	92.42%	8	636	83.27 (76.28–89.29)	78.21%
	Non-activated	6	466	24.14 (7.42–46.14)	95.66%	5	334	88.98 (81.29–94.97)	68.53%
Operator	General dentist	1	84	19.05 (12.08–28.72)	0%	1	84	91.67 (83.78–95.90)	0%
	Postgrad	7	690	42.89 (27.75–58.72)	94.25%	5	430	82.75 (76.01–88.62)	63.83%
	Specialist	2	83	46.81 (36.06–57.70)	0%	1	58	86.21 (75.07–92.84)	0%
	Specialist & Postgrad	1	98	61.22 (51.33–70.27)	0%	4	343	84.60 (73.98–92.93)	81.55%
Magnification	Yes	9	765	48.40 (35.15–61.75)	92.69%	9	803	81.27 (77.65–84.65)	36.21%
	Not mentioned	6	381	34.06 (12.35–59.62)	95.30%	3	167	94.82 (88.78–98.84)	0%
Rubber dam	Yes	11	979	43.89 (31.20–56.99)	94.06%	11	965	83.94 (79.78–87.73)	63.22%
	Not mentioned	2	29	0.00 (0.00–2.73)	0%	2	130	88.12 (81.76–93.37)	0%
Follow up duration	1 year	8	625	30.15 (15.55–47.08)	94.57%	10	844	84.75 (79.36–89.49)	74.52%
	2 years	2	193	43.74 (36.77–50.84)	0%	2	193	86.71 (81.47–91.22)	0%
	≥ 4 years	3	328	69.07 (63.66–74.23)	0%	1	58	86.21 (75.07–92.84)	0%

Qualification of operator

Root canal treatment procedures were performed by different operators: endodontists (n = 4) (Zhang et al. 2021; Castro et al. 2016; Gudac et al. 2022; Liang et al. 2011), postgraduate endodontic residents (n = 8) (Knight et al. 2020; Fernandez et al. 2013, 2017; Restrepo-Restrepo et al. 2019; Al-Nuaimi et al. 2017; Saidi et al. 2015; Curtis et al. 2018; Metska et al. 2013), and a mixed group of endodontists and postgraduate students (n = 3) (Zahran et al. 2021; Davies et al. 2016; Zavattini et al. 2020). Four studies did not report operator qualifications (Liang et al. 2013; Verma et al. 2020; Borden et al. 2013; Patel et al. 2012).

Tooth type

Most studies included a combination of tooth types (n = 7) (Knight et al. 2020; Fernandez et al. 2013, 2017; Restrepo-Restrepo et al. 2019; Patel et al. 2012; Zhang et al. 2021; Liang et al. 2011); however, two studies focused exclusively on single-rooted teeth (Liang et al. 2013; Verma et al. 2020), one on posteriors (Al-Nuaimi et al. 2017) and another on molars (Zahran et al. 2021), and eight studies did not report the tooth type (Borden et al. 2013; Davies et al. 2016; Castro et al. 2016; Gudac et al. 2022; Zavattini et al. 2020; Saidi et al. 2015; Curtis et al. 2018; Metska et al. 2013). The pooled success rate for treatments on anterior teeth was [strict: 82.22% (95% CI: 60.51–96.84); loose: 94.21% (95% CI: 84.13–99.83)] which is higher than treatments on posterior teeth [strict: 52.32% (95% CI: 40.37–64.13); loose: 82.69% (95% CI: 77.11–87.65)].

Treatment procedure

Twelve studies focused on NSRCT (Zahran et al. 2021; Liang et al. 2013, 2011; Verma et al. 2020; Knight et al. 2020; Fernandez et al. 2013, 2017; Restrepo-Restrepo et al. 2019; Patel et al. 2012; Castro et al. 2016; Zavattini et al. 2020; Saidi et al. 2015), five studies on NSReRCT (Zhang et al. 2021; Al-Nuaimi et al. 2017; Davies et al. 2016; Curtis et al. 2018; Metska et al. 2013), and two studies included both NSRCT and NSReRCT without providing stratified outcome data based on the treatment procedure (Borden et al. 2013; Gudac et al. 2022). The weighted pooled success rate for NSRCT assessed using CBCT, based on data from 8 studies, under strict and loose criteria was 42% (95% CI: 27%–58%) and 88% (95% CI: 83%–92%), respectively. While the weighted pooled success rates for NSReRCT assessed using CBCT, based on data from 5 studies (Zhang et al. 2021; Al-Nuaimi et al. 2017; Davies et al. 2016; Curtis et al. 2018; Metska et al. 2013), was 39% (95% CI: 19%–62%) using strict criteria and 80% (95% CI: 73%–86%) using loose criteria.

Rubber dam & magnification

Only four out of 19 studies did not mention use of rubber dam isolation (Castro et al. 2016; Zavattini et al. 2020; Saidi et al. 2015; Liang et al. 2011). Similarly, using a dental operating microscope was common, with 5 out of 19 studies not reporting its use for magnification (Liang et al. 2013, 2011; Verma et al. 2020; Castro et al. 2016; Saidi et al. 2015).

Instrumentation & obturation

Thirteen studies used rotary instrumentation (Zahran et al. 2021; Liang et al. 2013, 2011; Verma et al. 2020; Borden et al. 2013; Knight et al. 2020; Patel et al. 2012; Zhang et al. 2021; Al-Nuaimi et al. 2017; Davies et al. 2016; Castro et al. 2016; Gudac et al. 2022; Zavattini et al. 2020), while three studies employed both hand and rotary instrumentation, providing outcome data stratified by the method of instrumentation (Fernandez et al. 2013, 2017; Restrepo-Restrepo et al. 2019). Three studies did not report the instrumentation method (Saidi et al. 2015; Curtis et al. 2018; Metska et al. 2013).

Eight studies utilized warm vertical compaction as the obturation technique (Zahran et al. 2021; Liang et al. 2013; Knight et al. 2020; Patel et al. 2012; Zhang et al. 2021; Al-Nuaimi et al. 2017; Davies et al. 2016; Gudac et al. 2022), while three studies employed cold lateral compaction (Verma et al. 2020; Castro et al. 2016; Liang et al. 2011). Additionally, five studies used both methods (Borden et al. 2013; Fernandez et al. 2013, 2017; Restrepo-Restrepo et al. 2019; Metska et al. 2013), with three providing stratified outcome data (Fernandez et al. 2013, 2017; Restrepo-Restrepo et al. 2019). Two studies did not report the obturation technique (Saidi et al. 2015; Curtis et al. 2018). A non-randomized clinical trial compared the outcome of NSRCT between teeth obturated with a single cone and bioceramic sealer versus warm vertical compaction with epoxy resin sealer, finding no significant difference in the success rates between the two groups (Zavattini et al. 2020).

Irrigation

Sodium hypochlorite was the primary irrigant solution used in all the included studies. Nine studies employed ultrasonic-activated irrigation (Zahran et al. 2021; Borden et al. 2013; Knight et al. 2020; Patel et al. 2012; Zhang et al. 2021; Davies et al. 2016; Gudac et al. 2022; Zavattini et al. 2020; Metska et al. 2013). Two randomized controlled trials compared the outcome of NSRCT carried out with and without activated irrigation (Liang et al. 2013; Verma et al. 2020). Verma et al. (2020) found that ultrasonic and laser irrigation significantly enhanced the outcome of NSRCT more than non-activated irrigation (Verma et al. 2020). In contrast, Liang et al. (2013) did not find any significant difference (Liang et al. 2013). The pooled success rate for treatments carried out with activated irrigation was higher than treatments with non-activated irrigation [activated irrigation: strict 48.58% (95% CI: 30.90–66.43), loose 83.27% (95% CI: 76.28–89.29); non-activated irrigation: strict 24.14% (95% CI: 7.42–46.14), loose 88.98 (95% CI: 81.29–94.97)].

Coronal restoration

The majority of studies reported placing a permanent restoration after the treatment (n = 14) (Zahran et al. 2021; Liang et al. 2013, 2011; Verma et al. 2020; Borden et al. 2013; Knight et al. 2020; Patel et al. 2012; Zhang et al. 2021; Al-Nuaimi et al. 2017; Davies et al. 2016; Castro et al. 2016; Gudac et al. 2022; Zavattini et al. 2020; Metska et al. 2013). One study used a combination of permanent and temporary restorations and provided stratified outcome data (Restrepo-Restrepo et al. 2019). Four studies did not report the restoration type placed upon completion of endodontic treatment (Fernandez et al. 2013, 2017; Saidi et al. 2015; Curtis et al. 2018). Only five studies did not make note of the quality of coronal restoration at follow-up (Verma et al. 2020; Borden et al. 2013; Saidi et al. 2015; Curtis et al. 2018; Metska et al. 2013). The pooled success rate for root-filled teeth with adequate restorations was higher than those with inadequate restorations at follow-up [Adequate restoration: strict 52.83% (95% CI: 30.37–74.72), loose 86.97% (95% CI: 77.94–94.00); inadequate restoration: strict 13.28% (95% CI: 0.94–32.26), loose 78.71% (95% CI: 60.78–92.75)].

Number of visits

In four studies, treatment was completed on a single visit (Zahran et al. 2021; Liang et al. 2013; Borden et al. 2013; Patel et al. 2012), while in six studies, it was completed over multiple visits (Knight et al. 2020; Zhang et al. 2021; Al-Nuaimi et al. 2017; Davies et al. 2016; Zavattini et al. 2020; Curtis et al. 2018). Five studies used both single and multiple visits (Fernandez et al. 2013, 2017; Restrepo-Restrepo et al. 2019; Gudac et al. 2022; Liang et al. 2011), with three of them providing stratified outcome data (Fernandez et al. 2013, 2017; Restrepo-Restrepo et al. 2019). Three studies did not report the number of visits (Verma et al. 2020; Saidi et al. 2015; Metska et al. 2013). A randomized controlled trial compared the outcomes of single versus two-visit treatments, finding no significant difference in the success rates between the two groups (Castro et al. 2016). The pooled success rates for treatments completed in single and multiple visits were comparable [single visit: strict 42.53% (95% CI: 17.22–70.03), loose 86.19% (95% CI: 79.26–91.98); multiple visits: strict 43.58% (95% CI: 30.04–57.61), loose 84.25 (95% CI: 79.34–88.64)].

Periapical status

Three studies included teeth without preoperative apical radiolucency (Fernandez et al. 2013, 2017; Liang et al. 2011), while seven included only those with preoperative apical radiolucency (Liang et al. 2013; Verma et al. 2020; Borden et al. 2013; Restrepo-Restrepo et al. 2019; Zhang et al. 2021; Castro et al. 2016; Metska et al. 2013). Additionally, other seven studies included both groups (Zahran et al. 2021; Knight et al. 2020; Patel et al. 2012; Al-Nuaimi et al. 2017; Davies et al. 2016; Gudac et al. 2022; Curtis et al. 2018), and six of them provided outcome data stratified by the presence of preoperative periapical lesions (Zahran et al. 2021; Knight et al. 2020; Patel et al. 2012; Al-Nuaimi et al. 2017; Davies et al. 2016; Curtis et al. 2018). In comparison, two studies did not report the status of periapical tissues of included teeth (Zavattini et al. 2020; Saidi et al. 2015). The pooled success rate was higher in cases without preoperative periapical lesions [5 studies; strict: 88.47% (95% CI: 73.37–98.33), three studies; loose: 92.84% (95% CI: 84.45–98.58)] compared to those with preoperative lesions [10 studies; strict: 35.19% (95% CI: 22.00–49.60), seven studies; loose: 84.28% (95% CI: 78.68–89.20)].

Apical extent of root filling

Seven studies provided stratified outcome data based on the apical extent of the root filling (Zahran et al. 2021; Knight et al. 2020; Fernandez et al. 2013, 2017; Restrepo-Restrepo et al. 2019; Zhang et al. 2021; Liang et al. 2011). The length of the root filling was categorized into three groups: short (< 2 mm short of the radiographic apex), adequate (0–2 mm of the radiographic apex), and long (extrusion). The pooled success rates were comparable for adequate obturation [strict: 64.30% (95% CI: 55.57–72.59); loose: 81.48% (95% CI: 75.08–87.15)] and long obturation [strict: 62.41% (95% CI: 46.41–76.97); loose: 100% (95% CI: 20.65–100.00)], with reduced outcomes observed for short obturation [strict: 32.30% (95% CI: 8.30–61.09); loose: 60.00% (95% CI: 23.07–88.24)].

Duration of follow-up after treatment completion

The follow-up duration after treatment completion ranged from 1 to 12 years. Ten studies had a follow-up of 1 year (Zahran et al. 2021; Liang et al. 2013; Verma et al. 2020; Knight et al. 2020; Patel et al. 2012; Al-Nuaimi et al. 2017; Davies et al. 2016; Castro et al. 2016; Zavattini et al. 2020; Metska et al. 2013), five studies had a follow-up of 2 years (Borden et al. 2013; Restrepo-Restrepo et al. 2019; Gudac et al. 2022; Liang et al. 2011; Curtis et al. 2018), one study had a 4-year follow-up (Zhang et al. 2021), two studies had a follow-up of 5 years (Fernandez et al. 2013; Saidi et al. 2015), and one study had a follow-up of 10 years (Fernandez et al. 2017). The pooled success rates of root-filled teeth, when assessed under loose criteria, were comparable across different follow-up periods [1 year: 84.75% (95% CI: 79.36–89.49), 2 years: 86.71% (95% CI: 81.47–91.22), ≥ 4 years: 86.21% (95% CI: 75.07–92.84)]. However, under strict criteria, the highest pooled success rate was observed at the 4-year follow-up, followed by the 2-year and 1-year follow-ups [1 year: 30.15% (95% CI: 15.55–47.08), 2 years: 43.74% (95% CI: 36.77–50.84), ≥ 4 years: 69.07% (95% CI: 63.66–74.23)].

Sources of heterogeneity

To further investigate the variability observed in treatment outcomes, exploratory meta-regression was performed on ten prespecified study-level covariates under both strict and loose radiographic success criteria (Table 4). These covariates reflect factors commonly considered influential in endodontic treatment outcomes including presence of preoperative apical radiolucency, treatment procedure, tooth type, operator, number of visits, irrigant activation, use of magnification, extent of root filling, quality of coronal restoration, and follow-up time. Among these, 4 covariates showed statistically significant contributions to explaining heterogeneity under the strict criteria: presence of preoperative apical radiolucency (QM = 12.1177, p = 0.001, R² = 51.98%), quality of coronal restoration (QM = 9.0342, p = 0.003, R² = 85.37%), tooth type (QM = 5.8642, p = 0.016, R² = 41.47%), and follow-up duration (QM = 6.3140, p = 0.043, R² = 30.09%). Under the loose criteria, only one covariate – the use of magnification –demonstrated a statistically significant contribution (QM = 9.8241, p = 0.002, R² = 69.99%). Interestingly, the coefficient for magnification use (− 1.1148) was negative, suggesting that studies reporting the use of magnification had lower odds of successful outcomes under loose criteria compared to those not reporting magnification use.

Table 4.

Exploratory Meta-Regression Results for Covariates Associated with Treatment Success under Strict and Loose Criteria

Covariate	β Coefficient (95% CI)	Odd Ratio (95% CI)	P Value	QM (df)	Tau2	R2 (%)
Presence of PA
Loose
Absence	0			0.9041 (1)	0.2309	0.00
Present	−0.3829 (−1.1722, 0.4064)	0.6819 (0.3097, 1.5014)	0.342
Strict
Absence	0			12.1177 (1)	0.8830	51.98
Present	− 2.1945 (− 3.4301, − 0.9589)	0.1114 (0.0324, 0.3833)	0.001
Extend of Root Filling
Loose
Adequate	0			2.2391 (3)	0.2749	0.00
Long	− 0.4813 (− 3.9505, 2.9879)	0.6180 (0.0192, 19.8440)	0.786
Short	− 1.1744 (− 3.4088, 1.0599)	0.3090 (0.0331, 2.8863)	0.303
Long & Short	− 1.1744 (− 3.0161, 0.6673)	0.3090 (0.0490, 1.9489)	0.211
Strict
Adequate	0			2.5805 (2)	0.6227	0.00
Long	−0.6560 (−2.4426, 1.1306)	0.5189 (0.0869, 3.0976)	0.472
Short	−1.6670 (−3.7017, 0.3678)	0.1888 (0.0247, 1.4446)	0.108
Number of Visits
Loose
Single	0			0.0417 (1)	0.0516	0.00
Multiple	−0.0493 (−0.5220, 0.4235)	0.9519 (0.5933, 1.5273)	0.838
Strict
Single	0			1.8000 (1)	0.6910	0.00
Multiple	0.7018 (−0.3234, 1.7270)	2.0173 (0.7237, 5.6236)	0.180
Tooth Type
Loose
Anterior	0			1.4413 (1)	0.0542	19.22
Posterior	−0.7944 (−2.0912, 0.5025)	0.4519 (0.1235, 1.6529)	0.230
Strict
Anterior	0			5.8642 (1)	0.4590	41.47
Posterior	−1.3154 (−2.3801, −0.2508)	0.2684 (0.0925, 0.7782)	0.016
Quality of Restoration
Loose
Inadequate	0			0.3078 (1)	0.0338	0.00
Adequate	0.2849 (−0.7215, 1.2913)	1.3296 (0.4860, 3.6375)	0.579
Strict
Inadequate	0			9.0342 (1)	0.2734	85.37
Adequate	1.9177 (0.6672, 3.1682)	6.8053 (1.9487, 23.7654)	0.003
Follow Up Time
Loose
1 Year	0			0.4483 (2)	0.1752	0.00
2 Years	0.2301 (−0.5595, 1.0196)	1.2587 (0.5715, 2.7722)	0.568
> 4 Years	0.2469 (−0.9121, 1.4058)	1.2800 (0.4017, 4.0791)	0.676
Strict
1 Year	0			6.3140 (2)	0.7773	30.09
2 Years	0.2884 (−1.1457, 1.7225)	0.3180 (5.5985, 0.6935)	0.694
> 4 Years	1.5741 (0.3394, 2.8088)	4.8262 (1.4040, 16.5897)	0.013
Irrigation
Loose
Not Activated	0		0.123	2.3815 (1)	0.0679	28.32
Activated	−0.4010 (−0.9104, 0.1083)	0.6696 (0.4024, 1.1144)
Strict
Not Activated	0
Activated	0.9890 (−0.4083, 2.3862)	2.6885 (0.6648, 10.8725)	0.165	1.9245 (1)	1.3182	6.86
Operator
Loose
Postgraduate	0
General Dentist	0.8532 (−0.2565, 1.9630)	2.3472 (0.7737, 7.1208)	0.132	3.4027 (3)	0.1218	14.09
Specialist	0.2879 (−0.8029, 1.3787)	1.3337 (0.4480, 3.9700)	0.605
Specialist & Postgraduate	−0.1706 (−0.7933, 0.4522)	0.8432 (0.4523, 1.5718)	0.591
Strict
Postgraduate	0
General Dentist	−1.1000 (−4.0766, 1.8767)	0.3329 (0.0170, 6.5320)	0.469	0.9493 (3)	1.9379	0.00
Specialist	−0.2382 (−2.7452, 2.2689)	0.7881 (0.0642, 9.6688)	0.852
Specialist & Postgraduate	0.8037 (−2.1508, 3.7582)	2.2338 (0.1164, 42.8725)	0.594
Magnification
Loose
No	0		0.002	9.8241 (1)	0.0401	69.99
Yes	−1.1148 (−1.8119, −0.4177)	0.3280 (0.1633, 0.6586)
Strict
No	0		0.445	0.5823 (1)	1.1859	0.00
Yes	0.4571 (−0.7169, 1.6311)	1.5795 (0.4882, 5.1095)
Treatment Procedure
Loose
RCT	0		0.095	2.7822 (1)	0.1284	8.49
ReRCT	−0.4589 (−0.9982, 0.0803)	0.6319 (0.3685, 1.0837)
Strict
RCT	0		0.729	0.1196 (1)	1.2455	0.00
ReRCT	−0.2323 (−1.5488, 1.0842)	0.7927 (0.2125, 2.9570)

Publication bias

Publication bias was assessed with Egger’s test and the trim-and-fill method (Fig. 3 and Supplementary Table S6). A potential bias was noted for strict periapical healing only. Nevertheless, after applying the trim-and-fill adjustment, pooled rates remained similar.

Fig. 3.

Funnel plot to assess publication bias in outcomes under (A) strict criteria; (B) loose criteria

Discussion

This systematic review of 19 studies evaluating CBCT outcomes for nonsurgical root canal treatment (NSRCT) and retreatment (NSReRCT) found a moderate overall quality of evidence. Based on loose criteria, NSRCT showed an overall success rate of 88%, and NSReRCT showed an overall success rate of 80%. As this is the first systematic review to evaluate root canal treatment outcomes based on CBCT, previous comparisons are limited.

Conventional radiographic images offer a two-dimensional view of three-dimensional anatomical structures. In contrast, CBCT provides a three-dimensional assessment and has been shown to be more sensitive in detecting apical lesions. This increased sensitivity enables CBCT to identify apical periodontitis even when conventional radiographic assessments indicate complete healing (strict criteria) or reduction in lesion size (loose criteria). Despite earlier recommendations to increase reporting of outcomes using CBCT (Patel and Horner 2009; Wu et al. 2009), there is still a limited number of studies employing this method to evaluate the outcome of NSRCT and NSReRCT. Only 19 studies were eligible for inclusion in this review.

This meta-analysis estimated the weighted pooled success rate of NSRCT and NSReRCT assessed using CBCT as 42% and 39% under strict criteria and 88% and 80% under loose criteria, respectively. These findings are notably lower than those reported in previous systematic reviews using periapical radiographs. For instance, Burn et al. (2022) reported success rates for NSRCT of 92.6% (loose) and 82.0% (strict) (Burns et al. 2022), while Ng et al. (2007) found similar pooled rates of 85.2% (loose) and 74.7% (strict) (Ng et al. 2008). In the case of NSReRCT, Sabti et al. (2024) reported success rates of 87.5% (loose) and 78.8% (strict) (Sabeti et al. 2024), and Ng et al. (2007) found rates of 77.2% (loose) and 76.7% (strict) (Ng et al. 2008). The substantially lower pooled success rate under strict criteria in this CBCT-based review (42% and 39%) may be attributed to the higher sensitivity of CBCT in detecting periapical lesions, which might remain undetected on 2D radiographs. Therefore, comparisons across imaging modalities should be interpreted with caution, as they reflect not only treatment outcomes but also diagnostic sensitivity.

The meta-regression findings underscore the prognostic value of preoperative apical status, tooth type, quality of coronal restoration, and follow-up duration. These variables significantly contributed to the heterogeneity observed under strict criteria and are consistent with previous research emphasizing their role in treatment outcomes (Burns et al. 2022; Ng et al. 2011; Gillen et al. 2011).

Teeth with preoperative apical radiolucency demonstrated a lower success rate than those without existing apical radiolucency at the time of treatment. This finding aligns with previous systematic reviews (Ng et al. 2008; Burns et al. 2022). It could be attributed to the presence of bacterial biofilms within the root canal system and periapical tissues, which are more resistant to disinfection, as well as the complexity of the host immune response in resolving chronic inflammation. Notably, the majority of teeth included in this review (85%) presented with preoperative apical radiolucency, which may have influenced the overall results.

Anterior teeth demonstrated better treatment outcomes compared to posterior teeth, which is consistent with previous reviews (Ng et al. 2011; Kojima et al. 2004). One possible explanation for this trend is that anterior teeth generally have simpler root canal anatomy and more accessibility, making them easier to clean, shape, and obturate. Additionally, posterior teeth are subject to higher occlusal forces, which can increase the risk of treatment failure over time.

Teeth with adequate coronal restorations had significantly better outcomes than those with inadequate restorations. Adequate restorations help maintain the coronal seal and protect against reinfection. These findings support prior research (Ng et al. 2011; Gillen et al. 2011), highlighting the importance of high-quality restorations in the long-term success of root canal therapy.

Follow-up duration emerged as a critical factor in interpreting outcomes. The healing of apical periodontitis is a gradual process that involves immune responses and tissue remodeling, often requiring extended periods for complete resolution. Short-term observations may show initial signs of healing, but a follow-up duration of at least 3 to 4 years is recommended for a more accurate assessment of treatment outcomes, as complete lesion resolution can sometimes take over a decade (Duncan et al. 2023; Fristad et al. 2004). While pooled success rates under loose criteria were stable across different follow-up periods [1 year: 84.75% (95% CI: 79.36–89.49), 2 years: 86.71% (95% CI: 81.47–91.22), ≥ 4 years: 86.21% (95% CI: 75.07–92.84)], success rates under strict criteria improved with time [1 year: 30.15% (95% CI: 15.55–47.08), 2 years: 43.74% (95% CI: 36.77–50.84), ≥ 4 years: 69.07% (95% CI: 63.66–74.23)]. This pattern suggests that longer follow-up durations allow more time for complete healing and that short-term assessments might underestimate the actual long-term success of root canal treatments. Thus, using loose criteria can be valuable for capturing gradual improvements over time.

Interestingly, the use of magnification was associated with lower success under loose criteria. This counterintuitive finding may reflect selection bias, where magnification is used more frequently in complex cases with poorer prognoses, or residual confounding by factors such as operator expertise or disease severity. It may also point to reporting inconsistencies across studies, where the presence and type of magnification (e.g., dental loupes vs. operating microscopes) were variably defined or not explicitly stated.

In contrast, the remaining covariates —including operator level, irrigant activation, extent of root filling, primary vs. secondary treatment, and number of visits—did not significantly explain heterogeneity in either criterion, likely due to inconsistent reporting or limited variation across included studies. Future research should aim to standardize the reporting of these variables and include them more consistently in study designs to better evaluate their potential impact on treatment outcomes and inter-study heterogeneity.

This review also highlights a key limitation observed in prior systematic reviews: the inconsistent handling of extracted or retreated teeth. In this review, outcomes were analyzed as reported by the original studies. Two studies classified extracted or retreated teeth as failures (Al-Nuaimi et al. 2017; Zavattini et al. 2020), while two others excluded them from analysis (Zhang et al. 2021; Davies et al. 2016). One study did not report their inclusion (Fernandez et al. 2017). The remaining 14 studies reported no such events (Zahran et al. 2021; Liang et al. 2013, 2011; Verma et al. 2020; Borden et al. 2013; Knight et al. 2020; Fernandez et al. 2013; Restrepo-Restrepo et al. 2019; Patel et al. 2012; Castro et al. 2016; Gudac et al. 2022; Saidi et al. 2015; Curtis et al. 2018; Metska et al. 2013). This highlights the need for standardized reporting of such cases to improve comparability.

While this review showed that CBCT provides valuable diagnostic information, its routine use for outcome assessment is not recommended. Instead, CBCT should be utilized selectively in cases where conventional radiographs fail to provide clear information or when persistent symptoms or complex anatomical features necessitate more detailed assessment. The decision to use CBCT should be guided by clinical judgment, patient presentation, and adherence to established radiographic guidelines (Bhatt et al. 2021; Patel et al. 2019).

Limitations

This systematic review has several limitations:

1. Some of the included studies utilized only postoperative CBCT imaging, which limited outcomes assessment to strict criteria alone. The absence of preoperative CBCT in these studies reduces the ability to identify partial healing or improvements that fall short of complete healing, which could otherwise be captured using loose criteria.

2. The heterogeneity among the included studies may limit the comparability and generalizability of the results. Variations in study design, sample size, tooth type, treatment protocols, and follow-up durations complicate direct comparisons. Notably, the wide variation in sample sizes—ranging from small-scale studies to those with larger cohorts—may have impacted the precision of the pooled success rates, as studies with smaller sample sizes are more susceptible to variability and potential bias. These differences reduce the ability to draw definitive conclusions about the overall effectiveness of root canal treatment and retreatment based on CBCT outcomes and highlight the need for standardized protocols in future research.

3. The included studies employed a range of approaches for assessing periapical tissues in CBCT, including visual assessments, lesion size or volume measurements, and the CBCT Periapical Index (CBCT PAI). This lack of standardization complicates the comparison of outcomes and reduces the reliability of pooled analyses, as the results may be influenced by the specific assessment methods employed by each study. Future research in this area would benefit from adopting standardized criteria for CBCT-based periapical healing assessments to enable more consistent and comparable outcome reporting.

4. The short follow-up periods reported in many of the included studies is a notable limitation of this review, with most ranging between 1 to 2 years. Given that the healing of apical periodontitis is a slow and progressive process that may take several years—or even over a decade—for complete resolution, short-term follow-up may not fully capture the long-term success of root canal treatment. This is particularly relevant when applying strict radiographic criteria, which may underestimate healing in the early stages. Therefore, the lack of long-term data limits the ability to assess the true outcomes of treatment comprehensively and highlights the need for studies with extended follow-up durations.

Despite its limitations, this review has notable strengths. It is the first systematic review to evaluate the outcomes of root canal treatment/retreatment using CBCT. A rigorous search process was employed, applying strict criteria to identify relevant and methodologically robust studies. Additionally, several clinical factors that may influence the outcome of NSRCT and NSReRCT were investigated. These findings help guide case selection and highlight the need for future research, particularly prospective studies with long-term follow-up using CBCT.

Conclusion

The estimated weighted pooled overall success rates for NSRCT and NSReRCT assessed by CBCT were 41.03% (95% CI: 28.68%–53.95%; I² = 94.76%) under strict criteria and 85.01% (95% CI: 80.85%–88.75%; I² = 68.36%) under loose criteria. However, the quality of evidence regarding treatment factors influencing treatment outcomes was suboptimal, with substantial variation observed among the studies. Greater standardization in endodontic outcome research is needed, particularly in the consistent use of both strict and loose criteria and the adoption of standardized units of analysis (e.g., tooth vs. root). Additionally, establishing a consistent protocol for post-treatment restorations is crucial to ensure reliable outcome assessments.

In conclusion, while this review’s findings offer valuable insights, they should be interpreted with caution due to the retrospective and heterogeneous nature of the data. The review does not provide definitive conclusions but offers essential indications of the factors that may influence treatment success, which can help guide the design of future studies assessing root canal treatment and retreatment outcomes through CBCT.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

The study was conducted by a single author (conceptualization, methodology, data curation, data extraction, statistical analysis, writing, project administration).

Data collection was performed by two independent reviewers.

Funding

The author declares that no funding was used to support this publication.

Data availability

The data that supports the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

The study did not require ethical approval.

Consent for publication

Not applicable.

Competing interests

The author declares that there are no conflicts of interest in connection with this article.