Comparison of Fracture Resistance Based on Ferrule Presence vs Material Type in Maxillary Central Incisor Endocrown Restorations

Introduction

Maxillary incisors are particularly prone to traumatic injuries due to their anterior position in the arch. Endodontic treatment may become necessary, depending on the depth and extent of the fracture [1]. In such cases, indirect restorations supported by the root canal system, such as post-and-core or endocrowns, are commonly preferred to restore function and aesthetics [2].

Selecting an appropriate restorative technique is crucial for ensuring the long-term survival and functional efficiency of these teeth [3]. Post-and-core restorations have traditionally been considered the standard for restoring damaged anterior teeth [4]. However, these restorations require additional tooth preparation, which can weaken the remaining tooth structure and increase the risk of root fractures [3,5].

Advancements in adhesive techniques and the widespread use of CAD/CAM technology have shifted the clinical approach toward minimally invasive dentistry. As a result, endocrowns, a type of indirect adhesive restoration, are increasingly favored as a more conservative, practical, and cost-effective option for restoring severely compromised endodontically-treated teeth. Endocrowns provide macro-retention from the pulp chamber and micro-retention through adhesive bonding without the need for post-space preparation [6,7]. Preserving enamel enhances the marginal seal and bonding strength. Although mainly used for posterior teeth with extensive tissue loss, their use in anterior teeth remains debated [8–11]. Clinical success relies on preparation design, material choice, and the presence of a ferrule [12–14]. The ferrule effect improves fracture resistance by evenly distributing occlusal forces via a circumferential dentin collar [15–21].

Various ceramic materials have been introduced for CAD/CAM-fabricated endocrowns [22,23]. Lithium disilicate-reinforced glass ceramics are among the most preferred materials for fabricating endocrowns within the category of all-ceramic restorations due to their superior esthetic properties, high bond strength compared to other restorative materials, reduced plaque accumulation, and better aging resistance relative to composite resins [22–24]. However, a notable limitation of lithium disilicate ceramics is their relatively low mechanical strength. In contrast, zirconia offers excellent mechanical properties but is aesthetically inferior due to its inherent opacity. ZLS ceramics were developed to combine the favorable characteristics of glass and zirconia. These materials incorporate zirconia particles to arrest crack propagation within the ceramic matrix, enhancing mechanical durability. Fine lithium silicate crystals also contribute to improved translucency and aesthetics [22,25,26].

Thermocycling is commonly used in vitro to simulate the aging of restorative materials. This process mimics the temperature changes in the oral environment, offering insights into the long-term adaptation and fracture resistance of restorations [27,28]. The rapid temperature shifts during thermocycling cause stress at the adhesive interface due to differences in thermal expansion between the tooth structure and restorative materials, which can lead to marginal microleakage, crack formation, and a gradual decline in fracture resistance [27,28,29]. Considering thermal aging, fracture resistance is a key factor for the long-term success of restorations. Occlusal forces vary throughout the dental arch, from 400–800 N in mandibular molars to about 150 N in the anterior region [30–32]. While average masticatory forces generally range from 20 to 120 N, they can reach as high as 800 N in patients with bruxism [33]. Therefore, restorative materials need to withstand these varying functional loads. Their mechanical performance is typically assessed by static tests, which measure resistance to steady forces, and dynamic fatigue tests, which simulate chewing cycles to evaluate long-term clinical performance [34–37].

This study investigated the effect of ferrule design and restorative material on fracture resistance, where both variables are simultaneously at play. By investigating the influence of ferrule presence and ceramic type, this study aims to provide clinical insights into the optimal design and material selection for anterior endocrown restorations, guiding evidence-based restorative decision-making.

The null hypotheses were:

Material and Methods

SAMPLE COLLECTION/PREPARATION AND ROOT CANAL TREATMENT PROCEDURE:

This in vitro study was conducted in the Department of Prosthodontics, Faculty of Dentistry, Dicle University, and the Dicle University Faculty of Dentistry Research Laboratory, using teeth extracted for periodontal reasons only. The extracted teeth were examined under 10× magnification to ensure the absence of caries, cracks, fractures, or previous restorations. Only specimens of similar dimensions were included. The average tooth length was 22±1 mm, while the mesiodistal width was 9±1 mm. Periodontal tissues and any surface residues were mechanically removed using a scaler. Following extraction, the teeth were stored in a formalin solution for 24 hours. Subsequently, they were kept in distilled water at 37°C for a period not exceeding 6 months, as per ISO 11405 guidelines, consistent with other studies in the literature, until use [38–42].

The crowns were sectioned 3 mm above the cementoenamel junction (CEJ) using a low-speed diamond saw (Isomet, Buehler Ltd., USA) under continuous water irrigation to standardize the procedure [39–42]. Root canal treatment was then started. The working length was set 1 mm short of the apical foramen, and canals were shaped using a NiTi rotary system (VDW SILVER, Munich, Germany) at a 30# (0.06 taper). Irrigation was performed sequentially with 17% EDTA, distilled water, and 2.5% NaOCl, followed by a final rinse to prevent chemical precipitation. Canals were then dried with sterile paper points (Pearl Endo, Vietnam) and obturated with a calcium hydroxide-based sealer (Sealapex, Kerr, USA) and 30# gutta-percha cones (Pearl Endo, Vietnam). Excess gutta-percha was removed 1 mm apical to the canal orifice and sealed with temporary filling material (Vladmiva Dentin-Paste, Ukraine). Specimens were stored in a 100% humid environment for 1 week to allow complete sealer setting.

STUDY GROUPS DIVISION:

After numbering the 40 teeth, they were randomly assigned into 2 groups using https://www.random.org/. To assess the effect of ferrule absence on fracture resistance, 20 specimens were prepared without a ferrule (Ferrule (−) group) using a black-banded wheel diamond bur (DIMEI Piranha, BR837-021C, China) to create a butt-joint finish at the CEJ, followed by refinement with a red diamond bur (DIMEI Piranha, BR001-021C, China) freehand.

For the Ferrule (+) group (n=20), a 2 mm circumferential ferrule was prepared on teeth with a 3 mm coronal structure. A 1 mm-wide shoulder finish line was established 1 mm above the CEJ using a green-banded cylindrical diamond bur, with final refinement performed using a red-banded bur (DIMEI Piranha, BR001-021C, China). Burs were replaced after every 5 preparations in both groups to maintain precision.

A central retention cavity was prepared to a depth of 3 mm to provide a proper insertion path and rounded internal angle, but due to a 2 mm ferrule, the total cavity depth was 5 mm in the Ferrule (+) group (Figure 2). The canal orifices were sealed using a self-etch adhesive system (Clearfil SE Bond, Kuraray Co., Ltd., Osaka, Japan) and a flowable composite resin (Clearfil Majesty Flow, Kuraray Co., Ltd., Osaka, Japan).

All specimens were embedded in cold acrylic resin using cylindrical molds made from 30 mm-long PVC tubes (25 mm diameter). This ensured the roots were positioned perpendicular to the ground plane, with 1 mm of the root surface exposed below the CEJ. The tooth numbers were written on the molds using a permanent marker.

The specimens were randomly divided by using https://www.random.org/ into 4 subgroups based on ferrule presence and the material (n=10):

ENDOCROWN FABRICATION AND CEMENTATION PROTOCOL:

Digital impressions of all specimens were taken using CEREC Omnicam (Dentsply Sirona, Germany). Endocrowns were designed using inLab CAD software (Dentsply Sirona, Germany) and milled from CAD/CAM ceramic blocks. Following the manufacturer’s recommendations, each restoration was polished and crystallized at 850°C for 10 minutes [9,12,43].

The internal surfaces of the endocrowns were etched with 5% hydrofluoric acid for 20 seconds, silanized, and then coated with a universal adhesive. Cementation was performed using a dual-cure adhesive resin cement (G-CEM LinkForce, GC Corporation, Japan) under standardized loading (5 N) for 60 seconds. Specimens were then stored in distilled water at 37°C for 24 hours [10,11,44–46].

THERMOCYCLING AND FRACTURE RESISTANCE TESTING:

All endocrown-restored specimens were stored in distilled water at 37°C for 24 hours to simulate intraoral aging. Thermal cycling was then performed using an automated thermocycler (SD Mechatronics THERMOCYCLER) for 5000 cycles, spanning a temperature range of 5°C to 55°C. Each specimen was immersed for 25 seconds at each temperature, with a 10-second transition time between baths [47,48]. After thermal cycling, fracture resistance was tested using a universal testing machine (Lloyd LRX, Lloyd Instruments, UK). A 1-mm-diameter steel loading tip was applied to the palatal surface, 3 mm below the incisal edge, at a 45° angle (Figure 3). Specimens were stabilized to prevent movement during testing. The load was applied at a crosshead speed of 1 mm/min until fracture occurred, and the failure load was recorded in Newtons (N) [12,40,49–51].

The fracture types of all broken specimens were examined using a stereomicroscope (10×) under a light source. Fractures were classified into 4 groups [52]:

STATISTICAL ANALYSIS:

In numerous studies examining fracture resistance, the number of specimens per group varies. When the sample size is insufficient, it becomes challenging to perform a reliable statistical analysis [53]. Frater et al [54] used 10 specimens, Sharath et al [55] used 15, while Turker et al [56] and Silva-Sousa et al [57] each used 10 specimens per experimental group. Based on these studies, the total sample size for the present study was determined to be 40, with 10 specimens (n=10) in each group.

Data analysis was performed using the licensed version of SPSS 21.0 (IBM, Inc., Chicago, IL, USA). Normality testing was conducted using the Shapiro-Wilk and/or Kolmogorov-Smirnov tests, with P values less than 0.05 indicating a non-normal distribution and those greater than 0.05 indicating a normal distribution. For non-normally distributed data, Mann-Whitney U and Kruskal-Wallis H tests were applied. Post hoc multiple comparison tests were used to identify significant differences in Kruskal-Wallis H test results. Pearson’s chi-squared test and a Monte Carlo simulation were used for nominal variables due to the low expected cell counts. A two-way analysis of variance (two-way ANOVA) was conducted to evaluate the effects of ferrule presence (present/absent) and restorative material type (lithium disilicate and ZLS) on fracture resistance. The interaction effect between ferrule and material type was also assessed. A significance level of 0.05 was applied, with P values less than 0.05 indicating statistical significance.

Results

As shown in Table 1, there was no statistically significant difference in fracture resistance values between the material groups, regardless of the presence of a ferrule (P>0.05).

As indicated in Table 2, the fracture resistance value of the Ferrule (−) group was significantly lower than that of the Ferrule (+) group (P<0.05).

In Table 3, the fracture resistance value of the Ferrule (−) e max group was significantly lower compared to the Ferrule (+) e max and Ferrule (+) Celtra Duo groups (P<0.05). The presence of a ferrule had no statistically significant effect on the fracture resistance of Celtra Duo endocrowns (P>0.05).

All tested materials in the study, whether they have a ferrule or not, are suitable for endocrown materials in anterior teeth (Figure 4).

In the e max material group, a statistically significant difference in fracture resistance was observed between the ferrule groups (P<0.05). Specifically, the fracture resistance of the Ferrule (−) group was significantly lower than that of the Ferrule (+) group. In the Celtra Duo material group, although no statistically significant difference was found between the ferrule groups (P>0.05), specimens with a ferrule exhibited higher mean fracture resistance values (Table 4). Ferrule presence significantly increased fracture resistance in both material groups (P<0.05). While both materials performed similarly with a ferrule, ZLS demonstrated superior resistance in ferrule-free conditions. A significant interaction effect between ferrule presence and material type was observed (P=0.00013), indicating that the impact of material choice becomes more critical in the absence of a ferrule.

Table 5 shows that no Type 1 fractures were observed in the study. Although the results were not statistically significant (P>0.05), 50% of the Ferrule (+) e max group, 70% of the Ferrule (+) Celtra Duo group, 90% of the Ferrule (−) e max group, and 100% of the Ferrule (−) Celtra Duo group exhibited Type 4 fractures. This outcome highlights the potential importance of ferrule presence in minimizing the occurrence of non-restorable fractures.

Discussion

Maxillary central incisors are particularly susceptible to trauma due to their anatomical position and exposure to angular forces, which often necessitate indirect restorative approaches such as crowns or endocrowns [58].

According to Fehrenbach et al, rehabilitating endodontically-treated anterior teeth with extensive coronal loss remains a clinical challenge, as there is no definitive consensus in the literature regarding the ideal restorative material or technique. Therefore, investigating novel approaches that simplify the procedure and enhance clinical outcomes is of great importance [2].

For a successful restoration, several factors must be considered, including the remaining tooth structure, ferrule height, tooth morphology, position in the arch, and the direction of functional forces. Aesthetic outcomes, durability, reparability, and cost-effectiveness are critical parameters influencing the restorative decision [59].

Maxillary central incisors are susceptible to lateral forces, parafunctional habits such as bruxism, and high-stress concentrations, all of which significantly affect the long-term success of restorations [60]. Although numerous studies have investigated fracture resistance, variations in tooth type, age, preparation techniques, restorative materials, and testing protocols make direct comparisons challenging [61].

Endocrown restorations used in the anterior teeth have certain disadvantages compared to those in the posterior teeth. Mainly, the limited ceramic thickness at the margins increases the risk of marginal failure and ceramic chipping, which can lead to both aesthetic and functional problems [62]. Additionally, because of their shape, anterior teeth experience higher lateral and oblique forces, raising the risk of fracture and affecting long-term durability [2]. A comparative study also found that anterior endocrown restorations showed more microleakage after thermomechanical cycling than post-core restorations [63]; this is a significant disadvantage affecting both biological health and aesthetic stability. Therefore, anterior endocrowns have limitations that require careful consideration, especially regarding marginal integrity, stress distribution, microleakage, and, indirectly, aesthetics.

On the other hand, although traditional restorations, still recognized as the standard method, often involve placement of crowns on post-and-core systems with a ferrule effect, such approaches can also be prone to failure due to loss of post retention, fracture, deformation, or root perforation [4]. More conservative approaches have gained prominence with advancements in adhesive dentistry, leading to a re-evaluation of the necessity of traditional post-and-core techniques [9]. In this context, endocrowns have emerged as a valuable alternative, offering maximum preservation of tooth structure, high aesthetic performance, and reduced dependence on macro-retentive features [5]. While endocrowns are predominantly used for posterior teeth, their application in anterior teeth is also increasing [64].

While the influence of the ferrule effect on fracture resistance has been widely examined [12–14], a comprehensive literature review revealed that few in vitro studies have compared different restorative materials and the presence of ferrule, specifically in endocrowns applied to maxillary central incisors [2,57,65,66]

Finite element analyses by Aversa et al demonstrated that endocrowns reduce high-stress concentrations and improve the biomechanical behavior of restored teeth [67]. Additionally, since they are fabricated from a single block, this helps minimize interfacial stress [67–69]. In an in vivo study by Bindl and Mörmann, only 1 failure was reported out of 19 endocrowns followed over a 28-month period [6]. That study emphasized a minimally invasive approach, focusing on CAD/CAM-fabricated endocrown designs while excluding post-and-core systems. Based on this rationale, the present study aimed to evaluate the fracture resistance of endocrown restorations fabricated by CAD/CAM applied to endodontically-treated human maxillary central incisors, concerning material type and the presence of a ferrule.

Despite their advantages, a standardized preparation protocol for endocrowns is lacking, highlighting the need for further research to optimize their clinical application [12,57,70]. Data on optimal cavity depth remains particularly limited in anterior endocrown restorations. In a finite element analysis, Li et al found no significant differences in stress distribution between cavity depths of 3 mm, 4 mm, and 5 mm in maxillary central incisors, identifying 3 mm as the best depth [71]. Kanat Ertürk et al compared endocrowns with 3 mm and 6 mm canal extensions fabricated from different CAD/CAM blocks, reporting that lithium disilicate and zirconia-based materials could be safely used at both depths. However, resin nano-ceramic and feldspathic ceramic materials were determined to lack sufficient durability for anterior applications [72]. Moreover, studies on molars have indicated that increasing cavity depth leads to greater marginal and internal misfit [9,73]. Accordingly, in the present study, a canal extension depth of 3 mm was chosen, with an internal incisal-cervical taper of 6° and a supragingival margin established 1 mm above the CEJ.

Numerous studies have evaluated the effectiveness of various ferrule heights and configurations. Most of these studies indicate that a 2 mm circumferential ferrule significantly improves fracture resistance for endodontically-treated teeth [15–21]. In an in vitro study by Tan et al [15], 50 maxillary central incisors were divided into 3 groups: a 2 mm circumferential ferrule, a 2 mm partial ferrule (2 mm buccal-lingual, 0.5 mm proximal), and a no-ferrule control. Their results revealed that teeth with a complete 2 mm circumferential ferrule exhibited significantly higher fracture resistance than the other groups. These and some of the recent studies’ findings are consistent with the outcomes of our study, which led to the rejection of our first null hypothesis. Similarly, Pantaleón et al [16] assessed fracture resistance using a universal testing machine across groups with different ferrule designs and heights, including a 2 mm circumferential ferrule, partial ferrules of 2, 3, 4, and 6 mm (with 1 missing interproximal wall), and a ferrule-free group. The group with a 2 mm partial ferrule missing 1 cavity wall had the lowest fracture resistance. In comparison, the highest was recorded in the group with a 2 mm circumferential ferrule.

The clinical relevance of the ferrule effect has also been emphasized in long-term studies. A 17-year follow-up clinical study evaluated 304 teeth restored with various metal post-and-core systems and reported that a 2 mm circumferential ferrule was critical for long-term survival [20]. In another study, Li et al [71] performed a finite element analysis on maxillary central incisors with oblique fractures restored using various techniques, including fiber posts, cast posts, and endocrowns with cavity depths of 3–5 mm. The monolithic endocrowns exhibited superior stress distribution to conventional restorations involving multiple interfaces. Unlike our study, however, no significant difference was found between the 1 mm and 2 mm ferrule groups, possibly due to the oblique nature of the fracture lines and the absence of a true circumferential ferrule.

An in vitro study by Elsaid et al involved 40 mandibular molars, half prepared with a 1 mm ferrule and the other half without any ferrule. Endocrowns were fabricated from lithium disilicate (IPS e max CAD, Ivoclar Vivadent) blocks. Following thermal aging, fracture resistance was tested by applying axial forces using a universal testing machine. The ferrule group exhibited significantly higher fracture resistance (3528.69 N±569.83) compared to the non-ferrule group (2393.91 N±759.71) [74]. Although the fracture resistance values in our study were lower, the trend aligns with their findings. The lower values in our study may be attributed to several factors: the reduced bonding surface area in anterior teeth compared to molars, the application of oblique rather than axial forces, and the increased lever arm effect due to a greater crown-to-root ratio in anterior teeth. Additionally, our findings agree with several other studies in the literature [75–78].

In an in vitro study by Einhorn et al, 36 mandibular third molars were prepared with 3 different ferrule configurations: no ferrule, 1 mm ferrule, and 2 mm ferrule. Endocrowns were fabricated from lithium disilicate (IPS e max CAD, Ivoclar Vivadent) for all groups. The characteristics of each preparation were confirmed, and the available bonding surface area was measured using a digital recording microscope. The authors reported that the bonding surface area increased by approximately 36% in the 1 mm ferrule group and by more than 47% in the 2 mm ferrule group compared to the standard endocrown preparation [12]. In our study, the mean fracture resistance values of endocrowns with a ferrule were higher than those of the non-ferrule (butt-joint) groups. Although our study was conducted on anterior teeth, the findings are consistent, unlike Einhorn et al’s study on molars. The higher average fracture resistance observed in ferrule-containing specimens can be attributed to the increased adhesive surface area, which likely enhanced bonding efficiency.

Endocrowns can be fabricated using a variety of materials, including lithium disilicate, zirconia-reinforced lithium silicate (ZLS), monolithic zirconia, leucite-based ceramics, resin composites, and polymer-infiltrated ceramics. While these materials present specific advantages, they also exhibit certain limitations that warrant further development [22–24,79]. In a study by Elsaka and Elnaghy, ZLS materials demonstrated superior flexural strength, fracture resistance, elastic modulus, and hardness compared to lithium disilicate ceramics [26]. Although the ZLS group exhibited higher mean fracture resistance values in our study, the difference was not statistically significant. This outcome may be attributed to the limited sample size. Furthermore, the absence of dynamic fatigue testing, such as cyclic loading to simulate chewing forces, may have limited the ability to reveal clinically relevant mechanical performance differences between the materials.

In an in vivo study conducted by Bakke et al, the average maximum occlusal force applied to anterior teeth was reported to be 222 N [80]. In our study, the mean fracture resistance values were 534.10 N for the e max group, 588.34 N for the Celtra Duo group with a ferrule, 365.38 N for the e max group without a ferrule, and 425.87 N for the Celtra Duo group without a ferrule. All measured values exceeded the maximum physiological forces typically exerted on anterior teeth. Therefore, it can be concluded that the restorations used in this study have sufficient fracture resistance for the anterior region, particularly in patients without parafunctional habits such as bruxism. However, we believe long-term clinical follow-up studies are needed to validate these results.

In vitro studies indicate that applying force at a 45° angle tends to produce more catastrophic failures. Oblique loading increases fracture risk even under lower forces, as the resulting stresses are not evenly distributed along the long axis of the tooth and tend to concentrate in the cervical region. This stress concentration increases the likelihood of non-repairable fractures [81,82]. When evaluating the fracture patterns in our study, type 4 fractures were the most frequently observed failure mode in both material groups, regardless of ferrule presence, while type 1 fractures were not observed. Despite the high fracture resistance values recorded across all groups, most failures were non-repairable. This can be attributed to the oblique loading protocol and the increased crown height of the maxillary central incisors compared to the posterior teeth, which results in a longer lever arm. Therefore, although not statistically significant, we suggest that a ferrule plays a more critical role than the restorative material in preventing catastrophic failures and promoting more favorable, repairable fracture patterns.

This in vitro study has several limitations that should be acknowledged. Although thermal aging was simulated through 5000 thermocycles between 5 °C and 55 °C, the temperature changes of an oral cavity under 6 months of clinical service, to approximate intraoral temperature fluctuations, like some studies [47,48], no mechanical aging procedures – such as cyclic fatigue loading or chewing simulation – were performed. Consequently, the potential influence of functional stresses and fatigue-induced degradation on the long-term behavior of anterior endocrowns could not be evaluated. Using a single static load until failure provides insight into initial fracture resistance but fails to reflect time-dependent crack propagation or cumulative stress patterns that occur clinically. Furthermore, only 1 type of adhesive resin cement was used throughout the study. While this approach ensured standardization, it may not fully reflect the clinical variability associated with different adhesive strategies and cementation protocols. Additionally, the relatively small sample size may limit the statistical power and generalizability of the findings. Future research should incorporate mechanical loading regimens such as cyclic loading to simulate mastication, various adhesive systems, and larger sample sizes to thoroughly evaluate the clinical performance of different materials and preparation designs.

Conclusions

A 2 mm ferrule significantly improved the fracture resistance of anterior endocrowns, regardless of the ceramic material. While both lithium disilicate and ZLS performed similarly with a ferrule, ZLS showed better resistance and fewer catastrophic failures without it. Preserving ferrule structure is recommended, and ZLS may be preferred when it is not achievable. Future in vivo studies and long-term clinical trials are needed to validate these findings further and assess the long-term performance.