The property of a root-extended Ti-6Al-4V endocrown utilizing 3D printing for a four-wall defect in endodontically treated molars: An in vitro study

Abstract

Background

Severely damaged molars pose significant challenges in restorative dentistry. Conventional methods, such as resin composite restorations, often fail to provide sufficient fracture resistance, marginal integrity, and sealing, leading to frequent failures and secondary caries. This study aims to compare the mechanical properties and marginal integrity of 3D-printed Ti-6Al-4V root-extended endocrowns with direct resin composite restorations in molars with 4-wall defects.

Methods

Forty-two extracted third molars were randomly assigned to a 3D-printed Ti-6Al-4V root-extended endocrown group (n = 21) or a nanohybrid resin composite group (n = 21). All specimens underwent thermocycling. Fracture resistance was evaluated in eight specimens per group using vertical loading. Microleakage was assessed by methylene blue staining, and marginal integrity was evaluated by SEM. Micro-CT was performed on endocrown specimens without visible failure to detect internal microcracks.

Results

The endocrown group showed a significantly higher fracture resistance than the resin group (4682 ± 981 N vs. 1334 ± 615 N, P < 0.05). No catastrophic failures occurred in the endocrown group, whereas the resin group exhibited multiple failure modes. Lower microleakage scores and fewer marginal defects were observed in the endocrown group (P < 0.05). Micro-CT revealed limited microcracks in two endocrown specimens, and marginal gaps remained within clinically acceptable limits.

Conclusions

The study innovatively designs a 3D-printed Ti-6Al-4V root-extended endocrown, combining the structural advantages of endocrowns and post-core restorations. In our study, the 3D-printed Ti-6Al-4V root-extended endocrowns exhibited markedly enhanced fracture resistance, diminished microleakage, improved marginal integrity in comparison to the direct resin restorations, hence endorsing their application for molars with severe defect.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12903-026-08012-3.

Background

The first permanent molar is the initial permanent molar to emerge in the oral cavity [1]. The early eruption and the deep occlusal pit/fissures of the first permanent molar, along with the lack of oral hygiene practices in children, result in a much greater caries risk compared to other permanent teeth [2, 3]. The first permanent molar is essential for preserving masticatory function, occlusal stability, and the integrity of the dental arch [4]. The severe defect and early loss of the first permanent molars not only result in the displacement of adjacent teeth, elongation of opposing teeth, and occlusal disturbances, but also precipitate long-term complications such as malocclusion and temporomandibular joint disorders, which severely impact the oromaxillofacial development and quality of life [1, 5].

The clinical methods for the first permanent molar post-root canal therapy include direct resin filling, stainless steel crowns (SSC), and post-core crown restoration. Direct resin filling is currently the predominant technique, optimizing the conservation of tooth tissue. However, it demonstrates insufficient mechanical strength and is prone to fracture under stress [6], with a heightened risk of secondary caries developing over prolonged use. The failure rate of direct resin restoration for extensive tooth defects is elevated, mostly due to fractures and subsequent caries [7, 8]. SSC provides protection for significant areas of deficient teeth; nonetheless, challenges such as periodontal irritation and insufficient marginal adaptation remain [9, 10]. The post-core crown exhibits superior retention and fracture resistance. However, post preparation may further weaken the tooth structure and significantly increase the risk of root fracture [11]. Specifically, the restricted crown height and occluso-gingival space in adolescents impede the application of traditional post-core crowns [12]. Post-core crowns are scarcely employed for adolescents’ permanent teeth. Thus, the rehabilitation of the first permanent molar with significant defect continues to be a therapeutic challenge in adolescents.

Recently endocrowns have garnered attention due to their limited tooth preparation and excellent fracture resistance [13]. The endocrown achieves retention via the integration of the restoration with the pulp chamber, while the retention is also enhanced by the dentin-restoration bonding, thus reducing the risk of root fracture to a certain extent [14, 15]. Nonetheless, certain unresolved issues persist in endocrowns. For instance, pulp chamber retention may be insufficient for stability in instances of severe defects, such as four-wall defects or when the remaining crown is excessively short, especially in the molar area subjected to substantial occlusal forces [15]. Moreover, the cement-retention mechanism of endocrowns necessitates precise marginal adaptation to prevent microleakage in both the endocrowns and root canals [16]. To address these challenges, this study proposed a root-extended restoration design (1 ~ 3 mm into the root canal), which minimizes excessive tooth structure removal and provides a more suitable solution for cases with severe coronal defects and limited remaining tooth structure.

Consequently, a root-extended endocrown was proposed in this study. To guarantee the thickness of root canals, the design must include a sufficient number of small post channels to slightly extend the restoration into the upper portion of the root canal, taking into account the retention capacity of the pulp cavity and root canal to improve overall stability. Furthermore, 3D printing technology provides superior flexibility and precision in manufacturing, enabling customization design. It proficiently produces restorations that correspond more accurately with particular anatomical characteristics [17]. A four-walled defect molar model was created to evaluate the property of this novel restoration. Following the completion of the root canal treatment and subsequent canal preparation, cone-beam computed tomography (CBCT) was utilized to obtain structural data of the tooth for restoration design. The Ti-6Al-4 V titanium alloy, recognized for its exceptional mechanical properties and advantageous biocompatibility, was selected for 3D printing to enable the creation of highly tailored and precise restoration [18]. This study aims to evaluate mechanical property and the marginal fitness of a root-extended endocrown utilizing 3D printing for a four-wall defect in vitro.

Methods

Experimental procedure

For research preparation, forty-two third molars were taken at West China Hospital, Sichuan University (Chengdu, Sichuan, China) between July 2024 and September 2024. Teeth with comparable dimensions and intact root morphology were included, while specimens with cracks or extensive caries were excluded. All samples were stored in 0.1% thymol solution prior to use [19]. The experimental workflow was illustrated in the Fig. 1. All samples were subjected to thermocycling aging. Following thermocycling, eight samples from the endocrown group and eight from the resin group underwent fracture-resistance testing, with sample size determined based on a priori power analysis for a two-independent-samples t-test as shown in the follow fomula [20] :

Fig. 1.

Experimental workflow of the study

According to the existing literature, eight samples [21] from each group were assessed for microleakage and five samples [22] from each group were examined using scanning electron microscopy (SEM). Furthermore, the samples that exhibited no failure in the fracture test were analyzed using micro-computed tomography (Micro-CT).

Tooth preparation

Root canal therapy was performed by an experienced dentist on the third molars. conducted on the third molars using hand and rotary nickel–titanium file system (Dentsply Sirona, USA) with a final apical preparation size of 30. The root canal sealer used was iRoot SP (Innovative BioCeramix Inc., Canada), which was prepared according to the manufacturer’s instructions and used in combination with master gutta-percha cones (Dentsply Sirona, USA). The sealer was delivered into the canal using the master apical file with gentle counterclockwise rotation. After obturation using a warm vertical compaction technique with a cordless gutta percha obturation system (B&L Biotech, USA), specimens were stored for 24 h to allow complete sealer setting [23]. Then the crown was sectioned 2 mm above CEJ as a four-wall defect. The control group was prepared for subseqent filling with resin (Fig. 2A). For the experimental group, canals deviating more than 60° from the long axis of the tooth were discarded, and 1 ~ 3 roots were selected for extension. Root extension were established along the canal’s trajectory gradually and concurrently. To create a consistent insertion trajectory, the orientations were adjusted to coincide with the tooth’s long axis, extended along the canal, and sufficiently expanded, with each extension prepared 1 mm deeper with each iteration. Preparation was suspended if the root extension deviated markedly from the longitudinal axis. The residual dentin thickness around the extension was maintained at ≥ 1 mm ensuring the fracture resistance [24], with a depth of 1 ~ 3 mm. Subsequently, all the margins of extension were modified to create rounded angles with the pulp chamber walls. A diamond bur was employed to eliminate undercuts in the pulp chamber, and the walls of the pulp chamber were modified to a taper of 3 to 5 degrees, ensuring a wall thickness above 2 mm and a flat chamber floor [25]. All the processes were conducted by a professional operator to guarantee uniformity.

Fig. 2.

Establishment of four wall defect model and further treatment. A Control group: The crown was excised 2 mm above the cementoenamel junction, then followed by root canal therapy and resin restoration. B Experimental group: the root extension was around 1 ~ 3 mm, with a minimum of 1 mm of residual dentin and a pulp chamber wall thickness exceeding 2 mm. The walls of the pulp chamber were modified to an angle of 3 ~ 5 degrees. (Figures were created with biorender.com.)

Design of the root-extended endocrown

The data of tooth structures (Table S1) in endocrown group was obtained by CBCT (Fig. 3A), valuating residual dentin thickness and depth of root extension. The root-extended endocrown was derived from the 3D model of CBCT data. To ensure an inserting path, we established a point P and a plane V. As illustrated in the Fig. 3B, point P was defined as the intersection point of the pulp chamber wall and the long axis of the tooth in the 3D model. A vertical plane (V) perpendicular to the direction of insertion was established at point P. The pulp chamber segment of the restoration above plane V was engineered in accordance with the pulp chamber expansion, but the root segment below plane V had an outside wall that was perpendicular to plane V and parallel to the long axis of the tooth for the insertion purposes. In the actual CAD process, the long axis of the tooth served as the initial path of insertion, which was subsequently refined for pulp chamber adaption (Fig. 3B and C), followed by root extension design to guarantee a fit of at least 50% with the prepared root extension (Fig. 3D). The restoration morphology adhered to the anatomy of the maxillary first molar, guaranteeing precise root fitting and endocrown adaption during insertion.

Fig. 3.

Design of root-extended endocrown. A CBCT was employed to evaluate whether the planned root canal extension depth (1 ~ 3 mm) and the remaining dentin thickness (≥ 1.5 mm) met the predefined inclusion criteria via buccal and occlusal sections respectively. And CBCT databases were further imported for three-dimensional reconstruction, providing accurate anatomical information for subsequent CAD procedures. B Design on CAD: The initial position (P) at which the angle between the pulp chamber wall and the tooth’s long axis is minimized, establishing the trajectory for the root-extended endocrown. C View of Insertion path: The path is calibrated according to the tooth’s long axis, ensuring accurate alignment of the restoration’s pulp chamber part. D The whole view of root-extended endocrown on four-wall defect molar. Cross-section of root extended design demonstrated the fit and unfit part with the prepared root extension

3D printing and restoration

Following the design phase, Ti-6Al-4 V (TianQi Additive) was selected for 3D printing via laser melting. Post-printing processing included ultrasonic cleaning to remove residual powder, high-temperature annealing to eliminate residual stress, and fine surface polishing to optimize fit and surface quality. Final dimensions and adaptability were checked before use. The tooth surfaces were meticulously cleaned prior to bonding. An ultrasonic scaler eliminated impurities, succeeded by air-drying to guarantee a clean, dry bonding contact. The restorations underwent ultrasonic cleaning for 5 min to eliminate printing remnants, followed by air drying. Self-adhesive resin cement (3 M ESPE RelyX™ U200 Automix) was uniformly applied to the interior of the restoration, positioned onto the defect model with a 0.5 kg load to verify stability and adaptation, resulting in the extrusion of surplus cement. Following the initial curing, surplus cement was eliminated using a microbrush, and the interface was further cured using an LED light (1200 mW/cm²) for 40 s [26]. Following bonding, marginal adaptation was assessed, residual cement was eliminated using an ultrasonic scaler, and the restoration was gently polished.

Resin restoration

In the resin group, each tooth was subjected to etching with 37% phosphoric acid for 15 s, followed by rinsing and air-drying until a frosty appearance was achieved following ultrasonic cleaning. A universal bonding agent (3 M ESPE Single Bond Universal) was utilized, air-dried to eliminate solvent, and polymerized with LED light for 20 s. A flowable resin (3 M™ Filtek™ Z350 XT) was initially applied to the pulp chamber floor as a base layer of ≤ 1 mm to facilitate adaptation and mitigate shrinkage stress, cured for 20 s, and subsequently filled in layers with nanocomposite resin (3 M™ Filtek™ Z350 XT) in increments of 1.5 ~ 2 mm, each cured for 20 s [27]. Following the anatomical shape, the resin surface was refined, and surplus material was eliminated to guarantee restoration precision. All operations were conducted by the same seasoned clinician to guarantee experimental uniformity and reproducibility [27].

Thermocycling

All groups underwent thermocycling to simulate the aging of restorations akin to a clinical scenario. The restorations underwent 5000 cycles between 5 °C and 55 °C, with a dwell duration of 30 s and a transfer period of 10 s in a thermocycling apparatus [28].

Fracture test

Samples were embedded in fast self-curing base resin (GC Unifast Trad) 2 mm below the CEJ and cured for 24 h. As shown in Fig. S1, samples were fixed to a universal testing machine (INSTRON 5565, INSTRON) and loaded vertically at the central fossa with a 5 mm stainless steel tip at 0.5 mm/min [29, 30]. Failure modes were classified into the following four types (Table S2): I - cohesive failure within restoration, II - restoration-tooth debonding, III - coronal tooth fracture, IV - root-involving fracture. Failures above the CEJ were considered repairable, while those extending below were deemed non-repairable [30].

Micro-CT examination

Samples exhibiting no visible fractures post-testing were analyzed using Micro-CT (Scanco uCT45, SCANCO Medical AG) at 80 kV, 110 µA, with an 18 μm pixel size, 18 μm slice thickness, and a 300 ms integration time. Three-dimensional reconstruction was conducted, and microcracks were assessed using multi-planar (sagittal, coronal, axial) examination utilizing CTAn software, incorporating local magnification and grayscale analysis. Microcracks were characterized as linear, low-density shadows measuring ≥ 50 μm in length, continuous across ≥ 2 consecutive slices [31].

Microleakage assessment

Restored teeth were submerged in 1% w/v methylene blue for 24 h in darkness, washed for 60 s, and sectioned mesio-distally through the center of the restoration under water cooling (DTQ-5, Wiyi) [32]. Microleakage was assessed using 40× stereomicroscopy (SZX2, Olympus) and scored as follows (Table S3): 0 - no dye penetration, 1 - Dye penetration into enamel, 2 - Dye penetration beyond enamel-dentino junction (EDJ), 3 - Dye penetration up to the pulp wall [33].

Scanning electron microscope assessment

SEM (JSM6400, JEOL) was employed to compare the visible cracks of thermocycled samples from different groups and analysis the marginal gaps of the endocrown group after thermocycling, at a magnification of 500×. Prior to SEM analysis, the specimens were sputter-coated with gold in a vacuum environment (K550X, Emitech Quorum Technologies) and mounted on a metallic stub with the tooth interface oriented perpendicular to the optical axis to ensure standardized projection angles [34]. Measurements were performed at the mid-buccal and mid-lingual regions, which were pre-marked before analysis. All assessments were conducted by a trained, blinded operator [35].

SEM images was analyzed using ImageJ software (NIH, Bethesda). Marginal integrity was evaluated by counting all visible cracks, and marginal gap widths were measured by randomly placing 30 measurement lines along the external buccal and lingual margins [36, 37]. In addition, to further evaluate the marginal adaptation of the endocrown group, the external marginal gaps (EMG) at the mid-buccal and mid-lingual surfaces were measured using thirty measurement lines drawn on each SEM image via ImageJ software [35]. After EMG measurement, the specimens were embedded in clear epoxy resin and sectioned longitudinally at the mid-buccal and mid-lingual sites under water cooling using a low-speed cutting machine (DTQ-5, Wiyi Metallographic Instrument Co.). The cut surfaces were then polished and examined under SEM to assess internal marginal gaps (IMG) using the same methods of EMG assessment.

Statistical analysis

Statistical analyses were conducted utilizing R Studio software (RStudio, Inc.). The Shapiro-Wilk test was employed to evaluate the normality of fracture strength data. For normally distributed data, intergroup comparisons were performed using the independent samples t-test. Microleakage scores, classified as ordinal non-parametric data, were assessed using the Mann-Whitney U test to determine group differences. The quantity and width of marginal cracks were compared between the endocrown and resin groups using Mann–Whitney U test. Additionally, differences in marginal gaps of endocrown group between the buccal and lingual surfaces were analyzed using the Wilcoxon signed-rank test.A P value of less than 0.05 was considered statistically significant. A P value of less than 0.05 was considered statistically significant.

Results

Restoration outcomes of endocrown group and resin group

In the 21 specimens within the 3D-printed root-extended endocrown group, 2 teeth were equipped with a single root extension, 11 teeth with two root extensions, and 8 teeth with three root extensions. As illustrated in Fig. 4A. The morphology of the restorations aligned with the anatomical characteristics of the first permanent molar. All restorations were effectively positioned, and the marginal adaption was adequate following cementation.

Fig. 4.

Representative images before and after restoration of different restorative strategies for severely damaged first permanent molars in vitro. A Restoration of severely damaged extracted first permanent molars using the 3D-printed endocrown approach. Pre- and post-restoration views demonstrate a root-extended endocrown with intracanal extension, anatomical reconstruction of the occlusal surface, and well-adapted marginal integrity after restoration. B Restoration of severely damaged extracted first permanent molars using the layered nanohybrid resin composite approach. Pre- and post-restoration views show that satisfactory marginal bonding was achieved after restoration. However, accurate reconstruction of the original occlusal morphology was more difficult to obtain

The restoration outcome of resin rgroup is shown in Fig. 4B. Severely defective tooth models were completely restored using adhesive resin composite, with the restoration surfaces carefully finished and polished to achieve a smooth contour. Although satisfactory marginal adaptation was obtained, the accurate reconstruction of the original occlusal morphology remained challenging due to the extensive loss of tooth structure.

The endocrown group exhibited better fracture resistance

In the fracture test, none of the endocrown samples exhibited visible fracture, permanent deformation, or detachment when subjected to the maximum compression load of the universal mechanical test (up to 5250.0 N), and all samples remained intact. Figure 5A illlustrated the peak compression load and peak compression stress throughout the fracture test. The maximum load varied from 2771.4 N to 5250.0 N, which corresponds to maximum stresses of 141.15 MPa to 288.71 MPa. As shown in Table 1, the mean maximum compression load of the endocrown group was 4682.1 ± 980.6 N, and the mean maximum compression stress was 228.13 ± 52.42 MPa. Conversely, all samples in the resin group succumbed to failure during loading, with the maximum compression load and maximum compression stress corresponding to the failure load and failure stress, respectively. The failure load varied from 657.8 N to 2362.6 N, which corresponds to failure stresses of 33.51 MPa to 120.33 MPa (Fig. 5A). The resin group’s average failure load was 1333.7 ± 615.0 N, and the average failure stress was 69.73 ± 28.16 MPa (Table 1). The endocrown group demonstrated a significantly greater capacity for compression loads and stresses than the resin group (P < 0.05).

Fig. 5.

Results of fracture test. A Comparison of maximum load and stress between different groups (P < 0.05). B Representative failure patterns observed after fracture testing. No failure or fracture was observed in the 3D-printed root-extended endocrown group, while the resin restoration group exhibited multiple failure modes, including Type I (cohesive failure within the restoration), Type II (restoration–tooth debonding), and Type IV (root-involving fracture). C The statistics of failure types in the endocrown group and the resin group (P < 0.05). D Microcracks (yellow arrow) detected in 3D-printed specimens with Micro-CT. The yellow arrow indicated the microcracks propagated from the apical region of the root extension through the inner canal wall to the external surface

Table 1.

Comparison of maximum load and stress between different restorative materials (Mean ± SD)

Groups	Maximum load (N)	P value	Maximum stress (MPa)	P value
Endocrown	4682.1 ± 980.7	< 0.001	228.13 ± 52.42	< 0.001
Resin	1333.7 ± 615.0		69.73 ± 28.16

The analysis of failure modes were shown in Fig. 5B and C. In the resin group, two samples exhibited Type I failure (cohesive fracture inside the restorative material), three had Type II failure (adhesive failure at the restoration-tooth interface), and three experienced Type IV failure (irreparable fracture extending into the root). Conversely, the endocrown group experienced no failures. The findings reveal that the root-extended endocrown exhibit enhanced uniformity in stress distribution and improved protection of tooth structure, highlighting significant advantages in fracture resistance and clinical reparability.

Micro-CT scanning revealed microcracks in two specimens. The two samples both withstood a maximum load of 5250.0 N during the fracture resistance test and revealed one microcrack each. As shown in Fig. 5D, the microcracks originated from the apical region of the root canal retention segment and extended outward through the inner canal wall toward the external cementum.

The marginal adaptation in the endocrown group was better than the control group

As shown in Fig. 6A, the endocrown group exhibited only slight microleakage. In contrast, Fig. 6B demonstrates pronounced dye penetration in the resin group, with evident leakage observed in all specimens. The distribution of microleakage scores is summarized in Fig. 6C. Specifically, in the endocrown group, 6 samples scored 0, 1 sample scored 1, and 1 sample scored 2, whereas in the resin group, 5 samples received a score of 3 and 3 samples scored 2. Quantitatively, the endocrown group showed significantly lower mean microleakage scores (0.625 ± 0.744) than the resin group (2.62 ± 0.517, P < 0.05). The results showed that endocrown group outperformed resin group in terms of leakage resistance and marginal sealing integrity.

Fig. 6.

Microleakage by MB staining and crack formations under SEM. A Microleakage assessment in the endocrown group. B Microleakage in the resin group (red arrow). C The microleakage score for the groups (P < 0.05). D The margin of endocrown specimens (red arrow). E The resin group specimens exhibited visible cracks (red arrow). F Comparison of quantities and widths of marginal cracks between the two groups (P < 0.05). G EMG analysis of the samples in endocrown group. H IMG analysis of the samples in endocrown group. I EMG and IMG of the endocrown croup were uniform in buccal and lingual sites (P > 0.05)

Following thermocycling, the marginal cracks in the surface of 5 samples per groups were evaluated by SEM. Among the endocrown group, only one specimen exhibited a single buccal crack (Fig. 6D), while the remaining had intact margins. In contrast, all the specimens in the resin group exhibited cracks (Fig. 6E), with a total of 10 cracks observed across the buccal and lingual surfaces. As shown in the quantitative analysis of cracks (Fig. 6F), The resin group exhibited a significantly higher number of cracks than the endocrown group (P < 0.05). Additionally, the crack widths revealed a significant difference between the two groups. The endocrown group showed minimal crack propagation, with a mean width of 1.05 ± 3.32 μm, whereas the resin group exhibited markedly wider cracks, averaging 14.06 ± 5.38 μm. The difference in crack widths between the two groups was statistically significant (P < 0.05). In addition, the marginal gaps described the marginal fit of crowns [35]. Consequently, the EMG and IMG were assessed in the endocrown group (Fig. 6G and H). Shown in Fig. 6I, the EMG measures 64.56 ± 17.88 μm on the buccal side and 67.76 ± 12.47 μm on the lingual side, resulting in an overall mean of 66.19 ± 14.66 μm. No significant difference was seen between the buccal and lingual surfaces (P > 0.05) (Fig. 6I). The IMG in the endocrown group demonstrated a mean of 47.66 ± 10.52 μm on the buccal aspect and 48.57 ± 10.03 μm on the lingual aspect, resulting in an overall mean of 48.19 ± 9.69 μm (Fig. 6I). No significant difference was detected as well (P > 0.05) (Fig. 6I).

Discussion

The rehabilitation of endodonticlly-treated permanent teeth in adolescents has consistently posed a challenge. This study combined the endocrown and post forms to create a root-extended endocrown using 3D printing for a four-wall defect in endodontically-treated molars.

The results indicated that the 3D-printed endocrowns had significantly enhanced fracture resistance in comparison to traditional resin restorations for the rehabilitation of first permanent molars with severe detect. Compression test demonstrated that the maximum load-bearing capacity of the endocrown group (4682.1 ± 980.7 N) was markedly superior to the resin group (1333.7 ± 615.0 N). Furthermore, the endocrown group maintained structural integrity, exhibiting no macroscopic failure or irreparable fracture, even when subjected to the universal testing machine’s maximum loading of 5250.0 N. In contrast, the resin group demonstrated cohesive failure within the restorative material, resulting in debonding at the adhesive interface and irreparable fractures with crack into the root. The findings corroborate the recognized advantage of indirect restorations, especially those featuring complete occlusal coverage and retentive properties, in enhancing the fracture resistance [38]. The improved effectiveness of the endocrown group can be attributed to two causes. Ti-6Al-4 V demonstrates a markedly high elastic modulus and yield strength, providing substantial support while reducing deformation and fracture. Secondly, the dual-retention design, incorporating both a pulp-chamber component and a root-canal extension, promotes enhanced and more equal stress distribution at the tooth-restoration interface [39].

Although no apparent failures were observed in the endocrowns group, Micro-CT analysis revealed microcracks in two samples under maximum loading of 5250.0 N. These microcracks were located in the apical region of the root extension segment and extended outward along the canal wall. This phenomenon can be attributed to the considerable difference in elastic modulus between the prosthetic metal material (~ 110 GPa) and dentin (~ 18.6 GPa), along with the restricted load-bearing area at the apex of the root extension segment. It resulted in stress concentration, which may induce microcrack formation under intense compressive loading. These microcracks may serve as initiation points for crack propagation and develop into irreversible fractures with prolonged masticatory pressure [40]. Limitation of occlusal forces and materials exhibiting an elastic modulus similar to dentin may be utilized to mitigate stress concentration and diminish microcrack formation [41]. Furthermore, to evaluate the fracture resistance under extreme conditions, mechanical tests in our study did not simulate the buffering effect of the periodontal ligament. However, the absence of periodontal supporting tissues can alter stress distribution, which may also induce microcrack formation under extreme stress conditions.

Moreover, the endocrown group exhibited significantly reduced total microleakage relative to the resin restoration group, indicating improved marginal integrity (P < 0.05). SEM analysis revealed that the total number of cracks in the endocrown group was significantly lower than that in the resin group (P < 0.05), and the observed cracks were markedly narrower in the endocrown group (P < 0.05). After thermocycling, both internal and external edges of the endocrown group remained unblemished and devoid of gaps, but significant marginal fractures were apparent at both borders of the resin restorations. This discrepancy may be attributed to the integrated monolithic structure and higher elastic modulus of the 3D-printed Ti-6Al-4 V endocrowns, which allow better distribution and absorption of thermal and mechanical stresses during loading [42, 43]. Thus, resin restorations are more susceptible to stress concentration and crack propagation, especially during cyclic fatigue. Marginal cracks strongly contribute to microleakage, increasing the likelihood of secondary caries and undermining long-term clinical durability [44]. Conversely, the 3D-printed endocrowns exhibited superior marginal integrity, highlighting their efficacy in minimizing microleakage and averting secondary caries.

According to clinical standards, the adhesives used in the two repair methods are different, which may also affect the micro-leakage results. However, the 3D printing group showed significantly stronger sealing properties. Therefore, we conducted additional tests on the edge gap of the 3D printing group. A marginal gap is the space between the finish line of the preparation and the cervical margin of the restoration, measured from the crown margin to the finish line [45, 46]. The results shown no statistically significant difference between EMG and IMG at the buccal and lingual sites, indicating a uniform marginal fit circumferentially. A marginal gap less than 120 μm is regarded clinically acceptable [45]. The endocrown group showed an average of 66.2 μm of EMG and 46.2 μm of IMG, both within the clinically acceptable range. Notably, the IMG of the endocrown group was even less than 50 μm, representing an ideal marginal adaptation to minimize plaque accumulation and microleakage. These findings highlight the superior marginal integrity and adaption of the 3D-printed Ti-6Al-4 V endocrowns, which also demonstrated reduced microleakage compared to resin restorations. The precise digital design and additive manufacturing contribute to the high precision and uniformity of the marginal seal, collectively supporting the clinical advantages of 3D-printed restorations [47].

In summary, the results of this study showed that the fracture resistance, microleakage prevention, and marginal adaptability of 3D-printed endocrowns were significantly better than those of resin composite restoration. Furthermore, regarding microcracks observed in endocrown group under ultimate load, our experiment lacked long-term validation, precluding a definitive understanding of their performance during long-term clinical application. This may be improved during the CAD/CAM process by changing the printing material and combining with the finite element analysis. In addition, the sample size of this study was small and limited to a single type of 3D printed crown, which was only compared with resin composite restoration. Therefore, future studies may consider increasing the sample size, explore more types of restorations and materials, and combine with clinical studies to comprehensively evaluate the clinical application performance of this 3D printed endocrown with endodontic retention.

Conclusions

This study designed a 3D-printed Ti-6Al-4 V endocrown combined with root canal retention in the restoration of severely damaged permanent first molars, focusing on its fracture resistance and marginal fitness. The results showed that the fracture resistance, microleakage prevention and of 3D-printed endocrowns were significantly better than resin composite restoration. The marginal gap of 3D-printed endocrowns was within or even less that the clinically acceptable range, showing ideal marginal adaptability.

Abbreviations

SSC

stainless steel crowns

CBCT

cone-beam computed tomography

CEJ

cementoenamel junction

SEM

scanning electron microscopy

Micro-CT

micro-computed tomography

EMG

external marginal gaps

IMG

internal marginal gaps

Authors’ contributions

All authors were involved in study interpretation, writing, and reviewing the manuscript, including final approval. Additionally, Yifei Chen was involved in data curation, original draft preparation, and investigation; Tingting Song contributed to conceptualization, methodology, and original draft preparation; Kexin Chen was involved in investigation, validation, and data curation; Yichen Xu contributed to investigation, visualization, and data curation; Lei Lei was involved in supervision, conceptualization, and reviewing and editing the manuscript; Ran Cheng contributed to supervision, methodology, and reviewing and editing the manuscript; Tao Hu was involved in software, validation, and visualization.

Funding

The research was supported by Clinical research project of West China Hospital of Stomatology, Sichuan University (LCYJ-MS-202301).

Data availability

The data from this study will not be publicly shared. Due to privacy, confidentiality, or other reasons, the related experimental data will not be made available to external researchers after publication. All data and materials will be managed and controlled by the research team. For inquiries regarding access to the raw data analyzed in this study, please contact Prof. Ran Cheng at chengran@scu.edu.cn.

Declarations

Ethics approval and consent to participate

All procedures were conducted in accordance with the Declaration of Helsinki (and its later amendments). The study protocol, mainly involving the collection of isolated teeth, was approved by the Medical Ethics Committee of West China Hospital of Stomatology, Sichuan University (Approval Number: WCHSIRB-D-2023-334). This study was conducted with the approval of the Medical Ethics Committee of West China Hospital of Stomatology, Sichuan University (Approval Number : WCHSIRB-D-2023-334). Written informed consent was obtained from all participants providing extracted teeth.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.