Evaluation of the clinical outcomes and factors affecting the prognosis of endodontic microsurgery using cone-beam computed tomography: a retrospective cohort study

Abstract

Background

This retrospective cohort study aimed to assess clinical outcomes and identify prognostic factors of endodontic microsurgery through cone-beam computed tomography (CBCT) analysis.

Methods

Seventy-one teeth diagnosed with apical periodontitis and indicated for endodontic microsurgery were included. Two endodontic specialists performed the surgery following a standardized protocol. Clinical and radiographical reevaluations were conducted at least one year postoperatively. Two examiners independently measured the periapical radiolucency volumes before and after surgery from CBCT scans. Radiographic healing was classified as complete, limited, uncertain, or unsatisfactory. Multivariate logistic regression was used for prognostic analysis.

Results

Sixty-three teeth (88.7%) were available for follow-up; Sixty-two teeth (Sixty-one asymptomatic) were included in the final prognostic analysis. One tooth was excluded because it was extracted after the metal post-and-core fell off. The volume of periapical radiolucency measured at the final follow-up (≥ 1 year) showed a mean reduction of 89.4% compared to preoperative volumes (95% CI: 78.9% to 99.99%; P < .001). Complete or limited healing was observed in 53 teeth (85.5%), while 9 teeth (14.5%) showed uncertain or unsatisfactory healing. The angle of root resection was identified as a significant prognostic factor (P < .05).

Conclusion

This retrospective study provided information about the outcomes and prognostic factors of endodontic microsurgery, so as to detect the disease prognosis trend as early as possible, pre-plan the next treatment strategy, and effectively avoid prognostic factors that affect clinical outcomes.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12903-026-08017-y.

Introduction

Endodontic microsurgery is a surgical procedure that aims to remove persistent periapical lesions from endodontically treated teeth and regenerate periapical tissues [1]. Owing to the advances in endodontic microsurgery such as surgical operating microscope, the use of ultrasonic devices and micro-instruments, cone beam computed tomography (CBCT) imaging, superior root-end filling materials, and minimally invasive surgical philosophy, practitioners have more options for managing complex cases in which root canal treatment fails and other clinical strategies (e.g., endodontic retreatment, intentional replantation, tooth hemisection, and root amputation) is not feasible [2]. It is important to note that such failure often arises from several factors, including anatomic and pathologic challenges such as canal calcification or obstruction, as well as procedural complications including instrument separation, ledge formation, perforation, or apical transportation. Collectively, these issues prevent complete debridement and result in a compromised apical seal [3]. Given these challenges, endodontic microsurgery has a success rate of approximately 90% and is considered reliable for managing complex cases [4–6]. A more predictable outcome is expected.

With the widespread adoption of cone-beam computed tomography (CBCT), its role in enhancing surgical planning has been associated with improved outcomes in endodontic microsurgery, as investigated in studies analyzing factors affecting success [1]. Research has shown that CBCT is superior to periapical radiographs for detecting apical periodontitis [7–9]. A reduction in the size of the radiolucent area is indicative of successful healing of both periapical lesions and bone tissue. However, periapical radiography cannot reflect the true size of lesions surrounding the apex [8]. CBCT has been increasingly used for diagnosis, as researchers have reported a high degree of consistency between CBCT and histopathological diagnoses [7]. For routine follow-up, the benefits of CBCT far outweigh its associated risks of radiation exposure, unless the status of healing is difficult to ascertain [10]. By using both CBCT data and the software’s volume rendering function, the volume of periapical lesions before and after surgery can be measured and compared, allowing quantitative assessment of changes in periapical lesions [11, 12]. However, during the last 2 decades, literature on the volumetric changes in periapical lesions following endodontic microsurgery have been scarce [10, 11, 13–17].

The aim of this retrospective cohort study was to use 3D volumetric measurements of periapical radiolucency to assess the clinical outcomes and factors affecting the prognosis of endodontic microsurgery at least 1 year after the procedure.

Methods

Patient inclusion

The research protocol received approval from the ethics board of China-Japan Friendship Hospital, Beijing, China (2024-KY-257).

We screened patients who had undergone endodontic microsurgery at the Oral Medical Center of China-Japan Friendship Hospital between June 2023 and April 2024. Inclusion criteria were: availability of preoperative CBCT scan, a confirmed diagnosis of apical periodontitis, and an indication for endodontic microsurgery (Table 1) [3]. We excluded pregnant women, patients with poorly controlled systemic diseases on the basis of previous medical records with American Society of Anesthesiologists class III to V [18], and teeth with identified root fractures prior to surgery. For all included cases, data on whether the tooth had undergone nonsurgical endodontic retreatment prior to microsurgery was ascertained through review of medical and radiographic records. A total of 71 teeth from eligible patients were enrolled, all of whom provided informed consent.

Table 1.

The indications of endodontic microsurgery [3]

The indications of endodontic microsurgery
1. When root canal treatment fails and retreatment is not possible or fails again.
2. When there are anatomic variations at the root apex involving complex structures.
3. When there is a periapical lesion associated with a root canal that is obstructed by calcifications and cannot be negotiated.
4. When there is a periapical lesion associated with a root canal exhibiting severe curvature that cannot be adequately instrumented.
5. When complications of root canal treatment cannot be corrected through non-surgical retreatment: (1) A non-retrievable separated instrument associated with a periapical lesion; (2) The formation of an impassable ledge in the canal preventing apical debridement; (3) Apical transportation (zipping) or perforation resulting from over-instrumentation; (4) Persistent symptoms or inflammation due to significant overfilling of the root canal.
6. The presence of a periapical lesion in a tooth with an existing post-and-core restoration where post removal is impossible or would risk root fracture.
7. Failure of a periapical cyst to heal following nonsurgical root canal treatment.

Treatment protocol

Endodontic specialists performed all endodontic microsurgeries using an operating microscope (Leica M525 F20; Leica, Heerbrugg, Schweiz). Local anesthesia was achieved with 4% articaine containing 1:100,000 epinephrine (Primacaine; Acteon Pharma, Bordeaux, France). Following the raising of a full-thickness mucoperiosteal flap, osteotomy was performed by using fissure burs (Lindemann H161 burs; Brasseler, USA; Savannah, GA) to expose the root apex. A 3 mm segment of the apex of the root was resected perpendicular to the long axis with a high-speed diamond bur (Dentsply 700 010; Dentsply, Ballaigues, Switzerland) under sterile water. The resected surface was stained with methylene blue dye and inspected under ×20 magnification to identify any fractures or anatomical complexities. An ultrasonic tip (KiS; Obtura Spartan, Algonquin, IL) powered by a piezoelectric unit (Spartan MTS; Obtura Spartan) was used to prepare a 3-mm-deep root end cavity, which was then dried and filled with iRoot BP (Dentsply, Tulsa Dental Specialties, Tulsa, OK). An intraoperative X-ray was performed to confirm proper material placement within the root canal. The blood clot formed naturally from the osseous bleeding following debridement. To stabilize this endogenous clot and support the healing architecture, a resorbable collagen sponge was placed as a biocompatible scaffold over the retrograde obturation, specifically to protect the clot and guide early wound healing. The flap was repositioned and sutured with 4–0 nonabsorbable sutures (Mersilk W586, Ethicon, Johnson & Johnson Medical Ltd., Shanghai, China), which were removed after 5–7 days. Patients received postoperative instructions for a soft diet and oral rinsing with 0.1% chlorhexidine gluconate (hexamidine) for 1 week. Antibiotics were generally not prescribed unless the patient’s medical history warranted them, while an analgesic (ibuprofen, 400 mg) was allowed when necessary [2, 19–22].

Review

Patients were followed up via telephone interview or email more than 1 year after surgery. Patients were encouraged to comply with the recommended follow-up plan. Those who had moved to new locations were reached out to once more. For patients who couldn’t come back to the hospital for follow-up, telephone interviews were conducted for follow-up purposes. During these interviews, the function and symptoms of the treated teeth were noted down. Regarding patients whose teeth had already been extracted, the time of extraction and the reason were documented.

The clinical examination involved evaluations of subjective discomfort, swelling, the formation of sinus tracts, tenderness upon palpation or percussion, tooth mobility, periodontal pocket depth, and the quality of coronal restoration. The healing status on radiographic images was determined according to the modified PENN 3D [23] criteria and categorized into 4 groups: complete, limited, uncertain, or unsatisfactory healing (Table 2). For teeth with multiple roots that underwent endodontic microsurgery, the root showing the poorest healing was utilized to classify the radiographic healing. Complete healing and limited healing with no clinical signs or symptoms were considered successful. Moreover, uncertain healing and unsatisfactory healing were labeled failures according to Su et al.‘s study [2]. Cases where teeth were extracted due to persistent clinical symptoms or other reasons were also considered failures.

Table 2.

Modified Penn 3D criteria (radiographic) [23]

modified PENN 3D Criteria (radiographic)
(1) Complete Healing
(A) Re-formation of periodontal space of normal width and lamina dura over the entireresected and un-resected root surfaces.
(B) Slight increase in width of apical periodontal space over the resected root surface, butless than twice the width of non-involved parts of the root.
(C) Small defect in the lamina dura surrounding the root-end filling.
(D) Complete bone repair with discernible lamina dura; bone bordering the apical areadoes not have the same density as surrounding non-involved bone.
(E) Complete bone repair. Hard tissue covering the resected root-end surface completely.No apical periodontal space can be discerned.
(2) Limited Healing
Complete healing can be observed in immediate vicinity of the resected root surface, butthe site demonstrates one of the following conditions:
(A) The continuity of the cortical plate is interrupted by an area of lower density.
(B) A low density area remains asymmetrically located around the apex or has an angularconnection with the periodontal space.
(C) Bone has not fully formed in the area of the former access osteotomy.
(D) In areas with pre-existing periodontal disease or physiologic fenestrations un-resectedroot surfaces do not demonstrate bone coverage and/or periodontal reattachment.
(3) Uncertain Healing
The volume of the low density area appears decreased and demonstrates one of thefollowing conditions:
(A) The thickness is larger than twice the width of the periodontal space.
(B) The location is symmetrically around the apex as a funnel-shaped extension of theperiodontal space.
(4) Unsatisfactory Healing
The volume of the low density area appears enlarged or unchanged.

For the small operative field (6 × 6 cm), CBCT scans (both preoperative and postoperative follow-up) were obtained by a NewTom VGi (NewTom, Verona, Italy) at a high resolution level (slice thickness = 0.25 mm and intervals = 0.125 mm). The operating parameters were configured as 110 kVp, 13.04 mA, and an exposure time of 5.4 s. The CBCT images were then reconstructed using the matching NNT software, version 4.00.1 (NNT, Verona, Italy).

Two endodontists (chosen from the endodontic specialists) received training to independently assess the CBCT images. They were blinded to both the CBCT images and the follow-up time. In a dark room, the preoperative and postoperative CBCT images were projected onto a large screen and presented in a random order. A periapical lesion was identified when a disruption of the lamina dura was noticed, and the radiolucency around the radiographic apex was at least two times the width of the periodontal ligament space on at least two planes of the CBCT images. In case of any disagreement, the matter was discussed until a consensus was achieved. The volume of radiolucency in the CBCT scans was then measured separately by two examiners with the help of 3D Slicer (version 5.6.1; National Alliance for Medical Image Computing, USA). The measurement results were saved in the digital imaging and communication in medicine format. Initially, the area of periapical bone lesions was approximately segmented into rectangular forms. On the layers where the bone lesions were most distinctly observable across different axes, the threshold range for the lesion area was established. By applying this threshold range, the edges of the bone lesions were automatically outlined layer by layer. Subsequently, each layer was inspected to verify the accuracy of the identified lesion edges. Ultimately, the volume of each lesion was computed (Fig. 1). The measurement was carried out twice, with a 1-month interval between the two measurements, and the average of the first measurement was adopted. The volumes of radiolucency in the CBCT images before treatment and at the postoperative follow-up were contrasted, and the percentage of changes was computed [2, 12, 19–22].

Fig. 1.

Periapical lesion volume calculation using 3D Slicer. A Manual segmentation of a periapical defect in sagittal, axial and coronal views. B Selection of periapical defects using a grayscale value range selection tool. Then semiautomatic defect volume recognition and reconstruction of 3D image on a lesion. Finally calculation of the lesion volume

Clinical factors assessed [2, 19–21, 24–26].

Preoperative factors

Sex

Male or female.

Age

≤ 45 years or > 45 years.

Tooth type

Anterior, premolar, or molar.

Arch type

Maxillary or mandibular.

Quantification of lesion size (CBCT-PAI)

The periapical lesion size was measured using the CBCT periapical index (CBCT-PAI) [25, 26]. The largest diameter of the lesion, in conjunction with cortical bone expansion or destruction, served as the reference value. Apical periodontitis was categorized based on the maximum diameter of the radiolucent area observed in a slice from one of the CBCT planes, using the following criteria: 0: Intact periapical bone structures; 1 = 0.5–1 mm (widened periodontal ligament space); 2 = 1–2 mm; 3 = 2–4 mm; 4 = 4–8 mm; 5 = > 8 mm. Variables that could be added to the scores were also included: E - cortical expansion and D - cortical destruction. Teeth with more than one root with an apical periodontitis were assigned the root score with the highest CBCT-PAI (Fig. 2).

Fig. 2.

Examples of CBCT-PAI scores (0 to 5) in separate rows. Each row displays the same periapical lesion across the three orthogonal planes: sagittal (left), coronal (middle), and axial (right) slices, with scoring primarily based on the lesion's maximum diameter

Preoperative lesion volume

≤ 65 mm³ or > 65 mm³ (based on the volume setting of Zhang et al.‘s study [19]).

Quality of orthograde root filling

The quality of orthograde root filling was ascertained on preoperative CBCT scans in coronal and sagittal sections by considering both the length and density of the root filling in accordance with the classifications in the studies of Zhang et al. [19] and Liang et al. [27]. Satisfactory root filling quality was characterized as a flush length and satisfactory density; otherwise, the root filling was regarded as unsatisfactory.

Abutment

Whether the tooth was an abutment.

Intraoperative factors

Angle of root resection

The angle of root resection was measured between the resection plane and the reference line perpendicular to the long axis of the root. This value was subtracted from 90° to calculate the resection angle relative to a reference line perpendicular to the long axis of the root (Fig. 3). An adequate resection angle was defined as an angle less than or equal to 20°, whereas an inadequate angle was defined as an angle greater than 20° [28].

Fig. 3.

Schematic illustration of the resection angle in the bucco-lingual plane (angle 1 = angle measured with software tool; angle 2 = calculated angle, i.e. 90° minus angle 1)

Root-end filling quality

Adequate root-end filling should completely fill the root-end cavity without voids and should be extended to 3 mm or more in length.

Postoperative factors

Coronal restoration

The quality of coronal restoration was evaluated through clinical examination. Satisfactory restoration was defined as evidence of no discrepancy, discoloration, or recurrent caries at the restoration margin and no gingival recession [29].

Follow-up period

12–16 months or 17–20 months.

Statistical analysis

Statistical analyses were carried out with IBM SPSS Statistics for Apple Version 29.0 (IBM Corp., Armonk, NY). To assess the consistency among and within reviewers, the Cohen kappa coefficient and the intraclass correlation coefficient (ICC) were employed. A Wilcoxon signed rank test was utilized to compare the lesion volume before and after surgery. Regarding the analysis of predictors, the dependent variable was the dichotomous clinical outcome (categorized as success or failure). The chi-square test or Fisher’s exact test was applied to examine the bivariate correlation between treatment outcomes and all variables. Multivariate logistic regression analysis was conducted to determine the factors influencing the prognosis of endodontic microsurgery and to evaluate the impact of each factor on the surgical prognosis. The significance level was established at α = 0.05.

Results

Among the 71 teeth involved, the clinical data of 63 teeth were reviewed at least 1 year after treatment (Fig. 4). The average follow-up duration was 15 months, with a range from 12 to 20 months. The percentage of teeth under follow-up was 88.7% (63 out of 71). One tooth was extracted because the metal post-and-core had fallen off. The remaining 62 teeth (40 belonging to women and 22 to men, aged 23–72 years old) were clinically and radiographically examined. Eight teeth were lost during the follow-up process. Specifically, 2 patients with 2 teeth became pregnant, and the other 6 patients with 6 teeth either moved to other places or declined to cooperate with the follow-up protocol. For these 8 patients who were lost to follow-up, a telephone interview was carried out, and it was found that all of their teeth were functioning properly and showed no symptoms.

Fig. 4.

Flow chart illustrating the inclusion and exclusion of teeth

When determining whether periapical radiolucency was present or absent, the kappa scores for intraexaminer agreement were 0.937 and 0.962, and the kappa score for interexaminer agreement was 0.954. Regarding the CBCT volumetric measurements, the intraclass correlation coefficients (ICCs) for intraexaminer agreement were 0.976 and 0.992, while the ICC for interexaminer agreement was 0.968.

Based on the preoperative and postoperative CBCT data (with at least 1-year follow-up after surgery), the periapical radiolucency volume and the percentage change in volume of 62 teeth are presented in Table 3. Among the 62 teeth under follow-up, all but one (tooth no. 36) showed no clinical signs or symptoms. Tooth no. 36 developed a periapical fistula. Based on radiographic criteria of PENN 3D, the treatment outcome was successful in 53 teeth (85.5%), comprising 37 teeth (69.8%) with complete healing and 16 teeth (30.2%) with limited healing (Fig. 5). The remaining 9 teeth (14.5%) showed uncertain or unsatisfactory healing (Fig. 6). When comparing the volume before surgery with that at the postoperative follow-up, which was carried out at least 1 year after the operation, a significant difference was found (P<.001). The volume of periapical radiolucency measured at the final follow-up (≥ 1 year) showed a mean reduction of 89.4% compared to preoperative volumes (95% CI: 78.9% to 99.99%; P < .001). The distribution of the percentage of volumetric reduction is illustrated in Fig. 7.

Table 3.

The volume and percentage of change for periapical lesions based on preoperative and at least 1-year postoperative cone-beam computed tomographic data

No. of teeth	Preoperative lesion volume (mm3)	Postoperative lesion volume (mm3)	Percentage of lesion volume reduction (%)*	No. of teeth	Preoperative lesion volume (mm3)	Postoperative lesion volume (mm3)	Percentage of lesion volume reduction (%)*
1	427.62	0	100	32	71.49	177.18	-148
2	406.82	12.94	97	33	68.53	0	100
3	389.08	0	100	34	65.83	0	100
4	345.93	10.38	97	35	63.37	0	100
5	237.07	0	100	36	55.52	6.89	88
6	204.21	0	100	37	53.67	0	100
7	195.72	0	100	38	50.96	0	100
8	190.21	0	100	39	47.72	0	100
9	184.82	25.34	86	40	46.74	0	100
10	180.65	0	100	41	44.22	0	100
11	172.84	0	100	42	42.35	0	100
12	170.01	0	100	43	40.34	0	100
13	165.04	0	100	44	39.41	0	100
14	161.06	85.73	47	45	36.87	0	100
15	157.93	0	100	46	34.36	0	100
16	152.28	0	100	47	32.57	0	100
17	147.46	0	100	48	30.92	0	100
18	137.11	20.79	85	49	27.84	0	100
19	132.57	0	100	50	25.34	0	100
20	128.74	184.39	-43	51	23.38	0	100
21	122.28	0	100	52	22.46	0	100
22	116.49	0	100	53	20.47	0	100
23	111.05	0	100	54	18.21	0	100
24	102.48	0	100	55	17.37	0	100
25	102.36	0	100	56	15.53	0	100
26	92.17	151.04	-64	57	13.39	0	100
27	87.26	0	100	58	10.07	0	100
28	86.44	0	100	59	9.38	0	100
29	84.13	0	100	60	7.72	0	100
30	80.72	0	100	61	4.34	0	100
31	73.28	0	100	62	4.07	0	100

*Percentage of lesion volume reduction (%) = (preoperative lesion volume - postoperative lesion volume)/preoperative lesion volume

Fig. 5.

Reconstructed images in multiplanes and 3 dimensions based on CBCT scans with successful clinical outcomes. The preoperative (A–D and I–L) and the follow-up (E–H and M–P) CBCT scans. Case 1: Tooth 25 showed complete healing 1 year after a properly performed surgery. The preoperative periapical lesion with a volume of 44.22 mm3 had completely disappeared. Case 2: Tooth 16 was performed endodontic microsurgery. CBCT imaging at 1.5 years postoperatively showed that the volume of periapical lesion (122.28 mm3) had resolved, but the continuity of the cortical plate was interrupted by an area of lower density, indicating limited healing. The teeth in both cases were absence of clinical symptoms

Fig. 6.

Radiographic healing on CBCT images of teeth with poor clinical outcomes. The preoperative (A–D and I–L) and the follow-up (E–H and M–P) CBCT scans. Case 3: A 63-year-old man with a large periapical lesion on a mandibular central incisor underwent endodontic microsurgery. The postoperative CBCT images revealed the volume of periapical lesions significantly reduced from 161.06 mm3 to 85.73 mm3 at 2 years after treatment, with the labial lamina completely absent, indicating uncertain healing. Case 4: Tooth 26 was treated with endodontic microsurgery. However, the volumetric measurement revealed an enlargement in the volume of radiolucency from 71.49 mm3 at the preoperative assessment to 177.18 mm3 at the 1 year of follow-up evaluation, so the outcome of the tooth was determined as unsatisfactory healing

Fig. 7.

The scatterplot showing the percentage change in the volume of periapical lesions at least 1 year after endodontic microsurgery. Points above the x-axis indicate a decrease in the volume of the radiolucency, while points below the axis represent an increase in volume. The dashed line represents the average percentage change in volume, which is 89.4%

Table 4 presents a summary of the influence of clinical factors on dichotomous outcomes. Through multivariate logistic regression analysis, it was found that the angle of root resection (P<.05; odds ratio = 14.86; CI: 1.19–185.92) was a significant determinant of surgical outcome (Table 5). The average calculated angle of the resection plane for all roots was 11.7° ± 8.6°. The Nagelkerke R² value of this multivariable model was 0.64. The failure risk was 14.86 times greater for teeth with inadequate angles of root resection than for those with adequate angels. Although the P value for the clinical factor “quality of orthograde root filling” was less than 0.05, the odds ratio was extremely low (odds ratio = 0.03). Therefore, in this experimental context, it was not regarded as clinically significant.

Table 4.

A summary of bivariate analysis for the effects of clinical factors on radiographic outcome at the at least 1 year of follow-up

Factors		Radiographic outcome		Adjusted
	No. of teeth	Successful (%)	Failure (%)	P values*
Sex				0.326
Male	22	17 (77.3)	5 (22.7)
Female	40	36 (90.0)	4 (10.0)
Age				0.278
≤ 45 years	30	28 (93.3)	2 (6.7)
> 45 years	32	25 (78.1)	7 (21.9)
Tooth type				0.158
Anterior	21	19 (90.5)	2 (9.5)
Premolar	27	25 (92.6)	2 (7.4)
Molar	14	9 (64.3)	5 (35.7)
Arch type				0.446
Maxillary	48	42 (87.5)	6 (12.5)
Mandibular	14	11 (78.6)	3 (21.4)
CBCT-PAI				0.278
0	N/A	N/A	N/A
1 (0.5–1 mm)	2	2 (100.0)	0 (0.0)
2 (1–2 mm)	12	12 (100.0)	0 (0.0)
3 (2–4 mm)	32	27 (84.4)	5 (15.6)
4 (4–8 mm)	12	10 (83.3)	2 (16.7)
5 (> 8 mm)	4	2 (50.0)	2 (50.0)
Preoperative lesion volume				0.159
≤ 65mm3	28	27 (96.4)	1 (3.6)
> 65mm3	34	26 (76.5)	8 (23.5)
Quality of orthograde root filling				0.004
Satisfactory	7	2 (28.6)	5 (71.4)
Unsatisfactory	55	51 (92.7)	4 (7.3)
Abutment				0.326
Yes	2	1 (50.0)	1 (50.0)
No	60	52 (86.7)	8 (13.3)
Angle of root resection				0.043
Adequate	59	52 (88.1)	7 (11.9)
Inadequate	3	1 (33.3)	2 (66.7)
Root-end filling quality				0.278
Adequate	61	53 (86.9)	8 (13.1)
Inadequate	1	0 (0.0)	1 (100.0)
Coronal restoration				0.475
Satisfactory	58	50 (86.2)	8 (13.8)
Unsatisfactory	4	3 (75.0)	1 (25.0)
Follow-up period				0.326
12–16 months	39	35 (89.7)	4 (10.3)
17–20 months	23	18 (78.3)	5 (21.7)
Total	62	53	9

N/A Not available

*P values were adjusted for multiple comparisons using the Benjamini-Hochberg false discovery rate (FDR) method

Table 5.

Multivariate logistic regression analysis of the association between prognostic factors and endodontic microsurgery outcome

Factors	No. of teeth	Radiographic outcome
		Successful (%)	Failure (%)	OR value	OR 95% CI	P Values	R 2
Sex							0.64
Male	22	17 (77.3)	5 (22.7)
Female	40	36 (90.0)	4 (10.0)	0.38	0.09,1.59	0.184
Age
≤ 45 years	30	28 (93.3)	2 (6.7)
> 45 years	32	25 (78.1)	7 (21.9)	3.92	0.74,20.65	0.107
Tooth type
Anterior	21	19 (90.5)	2 (9.5)
Premolar	27	25 (92.6)	2 (7.4)	0.76	0.10,5.90	0.793
Molar	14	9 (64.3)	5 (35.7)	5.28	0.85,32.62	0.073
Arch type
Maxillary	48	42 (87.5)	6 (12.5)
Mandibular	14	11 (78.6)	3 (21.4)	1.91	0.41,8.88	0.409
CBCT-PAI
0	N/A	N/A	N/A
1 (0.5–1 mm)	2	2 (100.0)	0 (0.0)
2 (1–2 mm)	12	12 (100.0)	0 (0.0)	1.00	0.00,Inf	1.000
3 (2–4 mm)	32	27 (84.4)	5 (15.6)	21416443.46	0.00,Inf	0.997
4 (4–8 mm)	12	10 (83.3)	2 (16.7)	23129758.94	0.00,Inf	0.997
5 (> 8 mm)	4	2 (50.0)	2 (50.0)	115648794.70	0.00,Inf	0.997
Preoperative lesion volume
≤ 65mm3	28	27 (96.4)	1 (3.6)
> 65mm3	34	26 (76.5)	8 (23.5)	8.31	0.97,71.14	0.053
Quality of orthograde root filling
Satisfactory	7	2 (28.6)	5 (71.4)
Unsatisfactory	55	51 (92.7)	4 (7.3)	0.03	0.00,0.22	0.001
Abutment
Yes	2	1 (50.0)	1 (50.0)
No	60	52 (86.7)	8 (13.3)	0.15	0.01,2.71	0.201
Angle of root resection
Adequate	59	52 (88.1)	7 (11.9)
Inadequate	3	1 (33.3)	2 (66.7)	14.86	1.19,185.92	0.036
Root-end filling quality
Adequate	61	53 (86.9)	8 (13.1)
Inadequate	1	0 (0.0)	1(100.0)	103690265.20	0.00,Inf	0.994
Coronal restoration
Satisfactory	58	50 (86.2)	8 (13.8)
Unsatisfactory	4	3 (75.0)	1 (25.0)	2.08	0.19,22.58	0.546
Follow-up period
12–16 months	39	35 (89.7)	4 (10.3)
17–20 months	23	18 (78.3)	5 (21.7)	2.43	0.58,10.18	0.224
Total	62	53	9

N/A Not available

Discussion

A case in which endodontic treatment failed is challenging to manage [30]. Information on treatment outcomes is essential for the decision-making process. These decisive factors, such as the necessity of endodontic microsurgery, can be weighed against alternative methods of treatment (e.g., nonsurgical root canal retreatment, hemisection, and tooth extraction) [31]. The aim of this retrospective cohort study was to evaluate the outcomes of endodontic microsurgery at least 1 year after the procedure and the potential factors affecting its prognosis to guide clinical treatment, prognosis assessments, and decision-making in the future.

In this study, endodontic microsurgery was performed by professionally trained endodontic specialists. Strict aseptic conditions were maintained during the procedures, and tools and equipment such as operating microscopes, ultrasound instruments, CBCT imaging and bioceramic materials were utilized to ensure smooth execution of the surgery. Patients were followed up for at least 1 year after surgery. Halse et al. [32] and Jesslen et al. [33] reported that the 5-year prognosis can be predicted from the 1-year assessment with accuracies of 91% and 95%, respectively. Additionally, Song et al. [34] reported no significant difference in clinical outcomes between one-year and four-year or longer follow-up after endodontic microsurgery. Therefore, in our study, data collected at least 1 year after endodontic microsurgery may be able to predict the long-term outcomes of the procedure.

At least 1 year after surgery, 62 out of 71 teeth underwent clinical and radiographic examinations. A comparison of the preoperative and postoperative CBCT scans revealed that 85.5% of the teeth were free of radiolucency, indicating a reduction in intracanal infection and the restoration of healthy periapical conditions. In Zhang et al.‘s [20] research, 130 teeth were examined 12–48 months after surgery, all of which were asymptomatic, and the clinical success rate was 87.8%. In another study of 332 teeth, 198 followed up for 1–4 years (middle-term follow-up) had a healing rate of 86.9%, whereas 134 followed up for 5–9 years (long-term follow-up) had a healing rate of 67.2% [35]. In a systematic review of studies with a longer follow-up period, the pooled success rate of endodontic microsurgery ranged from 78% to 91% for patients followed up for 2–13 years [36]. In our study, the success rate of endodontic microsurgery was 85.5%, which was generally consistent with the success rates reported in the above literature.

In this study, the healing of periapical lesions after endodontic microsurgery was assessed through preoperative and postoperative CBCT scans. The correlation between CBCT imaging findings and the histological features of periapical lesions has been confirmed in animal experiments [7]. Schloss et al. [10] reported that CBCT scans could more accurately evaluate the status of healing after endodontic microsurgery than periapical (PA) radiographs could. PA radiographs are limited in that three-dimensional anatomical structures are condensed into two-dimensional images, and the surrounding anatomical structures are subjected to interference and geometric distortion. Additionally, compared with PA radiographs, CBCT assessments have better repeatability and reproducibility in evaluating the status of postoperative healing [10, 28]. They can provide three-dimensional imaging data of the examination site, offering more comprehensive information for diagnosing periapical lesions [37]. Therefore, in this study, we used CBCT instead of PA radiographs to evaluate the prognosis of the surgery.

In the current study, considering the risks of radiation exposure and its benefits after approval by the ethics committee, CBCT scans were not performed immediately after the completion of surgery. Instead, they were taken during the follow-up examination more than one year after the surgery. Small field of view CBCT imaging was applied, and the exposure settings were in compliance with the ALARA (as low as reasonably achievable) principle while ensuring the quality of the radiographs. Additionally, thyroid collar protection was provided for all patients. With respect to volume, the preoperative and postoperative volumes of periapical lesions have been examined in only a few studies, as has their correlation with the volume corresponding with modified PENN 3D criteria for evaluating the prognosis of surgery. In addition, in order to gain a clearer understanding of the lesion in three dimensions, the CBCT-PAI also employed in the study, which helped to reduce false-negative rates, minimize interobserver variability, and enhance the reliability of epidemiological studies on apical periodontitis prevalence and severity based on high-resolution CBCT images [25]. Recently, published clinical studies have shown that the modified PENN 3D standard has high interobserver agreement and internal consistency in the evaluation of 3D results [2, 38]. Therefore, by integrating the volumetric insights from the PENN 3D criteria with the structured scoring of the CBCT-PAI, the present study enabled buccolingual and mesiodistal evaluation of periapical lesions, which can obtain a more accurate classification of treatment results and a general understanding of the prognosis trend.

In this study, only the angle of root resection was identified as a predictor of surgical success, related similar studies on prognostic factors have not reached this conclusion [1, 2, 17, 19–21, 28, 39]. While the quantity of inadequate cases was too small, resulting in a deviation in the estimation (CI: 1.19–185.92), which suggested that the sample size in our study was insufficient, and larger, prospective studies were warranted for validation. Compared with the success rate of 88.1% in adequate cases, the rate in inadequate cases was 33.3%, which was significantly lower than that in adequate cases. Theoretically, clinicians should bevel the root end resection as perpendicularly to the long axis of the root as possible to benefit from optimized elimination of apical ramifications combined with limited removal of the root length, reduce the surface area of the resection plane and the number of exposed dentinal tubules, and minimize the risk of bacterial leakage from the periphery after retrofilling [28]. In a retrospective study of prognostic variables, 116 teeth treated with apical surgery had a “minimal” bevel of 0–10°, whereas 55 teeth had a “pronounced” bevel > 10°. The rate of failure was 10.3% for teeth with a minimal bevel but 29.1% for teeth with a pronounced bevel [39]. In addition, Monteiro et al. [40] reported that apicectomy at 90° promoted greater homogeneity in the stress distribution on the fiber post, cement layer and root dentin, which suggested a lower probability of failure. In addition, an inadequate angle of apical resection might lead to root end leakage through numerous exposed dentinal tubules [41]. Hence, an adequate angle of root resection is crucial for the success of endodontic microsurgery. When performing apicoectomy, it should be as perpendicular as possible to the long axis of the root.

During the follow-up process, one tooth was extracted due to detachment of the post-and-core, but the radiographic examination before extraction revealed that the periapical lesion had healed, indicating that the endodontic microsurgery was successful. However, for this discrepancy between radiographic and clinical success, we considered it to be a failure as well, because it was stipulated in the methods that cases in which teeth were extracted because of other reasons were regarded as failures. In addition, the follow-up examination included both clinical and radiographical, the affected tooth without the post-and-core crown restoration could not perform its normal function, even if the periapical lesion had healed. The purpose of endodontic microsurgery was to cure the lesion and enable the tooth to function. Only when both of the above conditions were met can it be considered a success. This served as a reminder that, in addition to evaluating the extent of the periapical lesion and the difficulty of surgery before operation, assessing the quality of the coronal restoration is equally important. Since endodontic microsurgery involves removing a portion of the root end, it alters the crown-to-root ratio, resulting in decreased fracture resistance of the tooth. Therefore, if the quality of the restoration is poor, little tooth tissue remains, the risk of restoration detachment is high, and the possibility of repairability is low, extraction may be the preferred treatment option.

There were several limitations to this study. First, since our hospital is a general rather than a specialized institution, the number of patients in the Oral Medical Center is naturally lower than that in specialized hospitals, leading to a relatively smaller sample size. Additionally, the follow-up period was relatively short. Third, regarding the choice of root resection technique, the use of a high-speed bur in this study may introduce variability in the resection angles, representing a potential source of methodological bias. Finally, the type of periodontal defect was not included in the analysis of prognostic factors, which might be considered critical and could potentially confound healing outcomes. Therefore, the generalizability of these results should be interpreted with caution given the limited population and setting. Nonetheless, these findings offer valuable insights. Further studies incorporating larger sample sizes, longer follow-up periods, the use of ultrasonic device and a more comprehensive range of prognostic factors are necessary.

Conclusion

Considering the limitations of this study, the clinical outcomes and factors influencing the prognosis of surgery were determined and measured through CBCT scans. The success rate of endodontic microsurgery treatment was 85.5%, with a follow-up period of at least 1 year. An adequate angle of root resection had a positive effect on the success rate. With the information above, it helped to detect the disease prognosis trend as early as possible, pre-plan the next treatment strategy, and effectively avoid prognostic factors that affect clinical outcomes, so as to guide clinical treatment, prognosis assessment, and decision-making.

Abbreviations

CBCT

Cone beam computed tomography

CBCT-PAI

CBCT periapical index

ICC

Intraclass correlation coefficient

PA

Periapical

ALARA

As low as reasonably achievable

Authors’ contributions

Z.J contributed to the study conception and design. W.JS and W.J analyzed the data and wrote the main manuscript text. Clinical procedures and data collection were performed by G.LJ and Q.LY. H.P, L.SS and X.J prepared figures and tables. W.JS and W.J supervised data analysis and thoroughly revised the manuscript. All authors read and approved the final manuscript.

Funding

The study was funded by the National High Level Hospital Clinical Research Funding (grant no. 2024-NHLHCRF-PYII-25 and 2023-NHLHCRF-PY-03).

Data availability

All data generated or analysed during this study are included in this published article and its supplementary information files.

Declarations

Ethical approval and consent to participate

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and national research committee and with the 1964 Helsinki declaration. This study was approved by the local ethical committee at China-Japan Friendship Hospital, Beijing, China (Approval Number: 2024-KY-257). Informed written consent to participate was obtained from all patients.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.