Prospective clinical evaluation of 765 partial glass-ceramic posterior restorations luted using photo-polymerized resin composite in conjunction with immediate dentin sealing

Partial indirect restorations can be indicated when it is difficult to restore form and function using direct composite restorations. Monolithic reinforced glass-ceramic restorations have gained popularity in posterior teeth as these restorations are less prone to fracture compared to feldspathic ceramic [1]. Lithium disilicate restorations have an increased fracture toughness from the crystallites as these induce the fracture to bow, deflect and branch [2]. Besides their strength, adhesive luting is possible and this increases strength and regain aesthetics at a minimum biological price, removing defect structures only [3‐5].

A recent systematic literature review revealed a cumulative survival rate for single glass-ceramics and feldspathic porcelain restorations of 92–95% after 5 years, and 91% after 10 years [6]. Due to high load in the posterior region, these restorations were more prone to fracture than crowns in the anterior region [7, 8]. In short-term evaluations, bulk fracture, chipping of the ceramic and adhesive problems have been reported as the main reason of failure [9]. Besides restoration failures, secondary caries was observed as a biological complication (1%) [6]. Comparing acid-etched e.max lithium disilicate monolithic and bilayered complete coverage restorations resulted in an estimated cumulative survival rate of 96.5% after 10.4 years for monolithic and 100% after 7.9 years for bilayered restorations. This difference was statistically significant [10]. Long-term data with a high number of these partial all-ceramic restorations is unfortunately limited, which precludes firm statements regarding their effectiveness [11].

Adhesive bonding to dentin has been considered as the weakest link in clinical durability and fracture resistance of ceramic restorations [12]. The so-called immediate dentin sealing concept (IDS) has been studied extensively in in vitro studies and obtained significantly increased bond strengths [13‐21]. An vitro study demonstrated that the application of an IDS layer with ceramic inlays significantly increased fracture strengths [22].

The use of a conventional photo-polymerized resin composite as a luting agent is debated [23, 24]. Manufacturers and studies claimed that the use of composite as a luting agent with thick or extensive restorations could lead to a decrease in degree of conversion [25‐29]. However, luting with composite has some practical advantages, i.e., increased working time, improved biomechanical properties, wear resistance and ease of removal the excess [25‐27, 30]. In addition, recent articles have proved their ability to have a higher bond strength in in vitro [26, 31] and in in vivo research [27, 30, 32].

The aim of this study is to evaluate the clinical survival and success rate of partial glass-ceramic posterior restorations luted with conventional photo-polymerized resin composite in conjunction with the use of IDS. Further evaluation will involve the location of the restoration (premolar/molar and mandibula/maxilla), pre-restorative endodontic status (vital/devitalised) and the extension of the indirect ceramic restoration (number of sides and cusps involved).

In total, 765 partial restorations in 158 patients (60 men, 98 women) could be included: of which 697 restorations on vital teeth and 68 on devitalised teeth. In total, 360 restorations were placed in the mandibula and 405 in the maxillary posterior teeth, 282 on premolars and 483 on molars. The mean observation time was 53.3 months (range 3–113 months).

In three cases, the extension of the indirect ceramic restoration could not be determined from the photograph made at baseline, nor from the patients’ record. The number of sides (buccal/lingual/palatinal/mesial/distal) that the restorations replaced were as follows: 2 sides: 12, 3 sides: 198, 4 sides: 262 and 5 sides: 290. The number of cusps that were replaced by the restoration were as follows: 0 cusp n = 409, 1 cusp n = 179, 2 cusp n = 110, 3 cusps n = 19 and 4 cusps n = 45.

The overall estimated cumulative survival rate according to Kaplan-Meier after 5 years of function and longer is 99.6% (SE 0.3%, 3 events) and the success rate are 98.6% (SE 0.5%, 9 events) after 5 years and 96% (SE 1.7%) after 7.5 years and longer (12 events, Figs. 1 and 2). Three cases were absolute failures and thus extracted, 2 being a fracture of a tooth after 10 months (devitalised) and a tooth after 33 months (vital) with restorations without cusp coverage. A non-salvageable apical re-infection occurred in another tooth after 18 months of function.

Over time, repairable and salvageable failures occurred in 9 teeth, being endodontic treatment needed (n = 7), secondary caries (n = 1) and debonding of the restoration (n = 1). Location (premolar/ molar and mandibular/maxilla), extension of the indirect ceramic restoration (number of sides and cusps involved) and pre-restorative endodontic treatment (vital/devitalised) did not significantly affect the cumulative success rate (log rank test, p > 0.05).

The USPHS data across the different observation periods are presented in Table 5 by means of frequency distributions. The condition of the vast majority of the restorations was excellent and remained unaffected after 5 years of function or longer, as also represented by the USPHS sum score. A Kruskal-Wallis test was performed to explore the USPHS sum scores in time, i.e. baseline to > 5 years. There is a statistically significant difference between the USPHS sum scores in time (p < 0.001). The results of the post-hoc tests after Bonferroni correction show a significant difference between the USPHS sum score at baseline and the scores for the subsequent time intervals (1, 2, 3, 4 and > 5 years). Hence, some deterioration of the general quality of the restoration is seen after the first year of clinical service, but it remained stable thereafter (Table 5).

In this prospective study, the clinical performance of 765 partial glass-ceramic posterior restorations luted by means of a conventional photo-polymerized resin composite in conjunction with Immediate Dentin Sealing were evaluated. Restorations presented excellent estimated cumulative survival and success rates after 5 years, being 99.6% and 98.6%, respectively. A representative case is presented in Fig. 3. No clinical study has been performed on partial ceramic restorations using Immediate Dentin Sealing and a conventional photo-polymerized resin composite.

The results of this study are better than other studies on all-ceramic partial restorations not using IDS and composite with short (2 years; 100% [32]) to medium-term (5–8 years; 94.8–97.4% [5], 7 years; 97%–100% [36],4 years; 93%–100% [37‐39]) observation periods. The evidence for such restorations with longer follow-up time and high numbers of patients is limited [22].

One review reported an estimated survival rate of 95% after 5 years and 93% after 10 years [6], and compared to the survival of conventional posterior full crowns (metal–ceramic (94.7%) and lithium disilicate reinforced glass-ceramic (96.6%) after 5 years) [40], the partial restorations in this study show a higher survival rate. One of the possibilities for the excellent outcome of this study is the use of a conventional photo-polymerized resin composite as the luting agent [41]. A dual-polymerized composite resin to lute ceramic inlays was used in the vast majority of other clinical studies [5, 6, 37, 38, 42, 43]. Previous studies which investigated the use of photo-polymerized composite resins for luting purposes produced conflicting results [28, 29]. Presumed reduced wear of luting composites could not be confirmed when using the higher filled luting material [28]. However, more recent studies on this subject are promising [25‐27, 31]. The higher filler content and lower initiator concentration compared to dual-polymerized resin cements may be beneficial in terms of mechanical strength and the wear properties at the exposed margins [29]. Even thick restorations are not contra-indicated with photo-polymerized luting composites in combination with IDS [25] but the use of a high power photo-polymerized (> 2000 mW/cm²) unit and extended polymerization time are considered of critical importance [26]. A decrease in marginal adaptation of ceramic restorations over time is reported in the current literature, as marginal deterioration can be attributed to degradation and wear of the composite [12, 44]. An important factor for the clinical long-term performance of partial coverage restorations is marginal degradation on the resin cement and deterioration of the all-ceramic materials during clinical function [45]. Based on the results of this study, the USPHS criteria according marginal adaptation showed very good results over time even with restorations up to 5 year or more because of the favourable properties of the conventional photo-polymerized resin composite.

Two absolute failures occurred due to fracture in the root of the teeth. One of these catastrophic fractures was in a vital teeth and one in a devitalised teeth. Both of these fractures occurred in teeth without cusp coverage. Several authors suggest cusp coverage to restore weak posterior teeth [46, 47]. However, the amount of 0 and 1 cusp-replaced restorations were high in this study. Almost no cusp coverage was performed due to removal of sound tissue as little as possible. The results of this study showed that the extension of the restoration and the pre-restorative endodontic status did not challenge the survival. Probably due to the adhesive quality with the use of IDS and conventional photo-polymerized resin composite, cusp coverage is no longer required. But further research is necessary to confirm this statement.

Fractures of the ceramic material or minor chipping in general were not observed, which may be contributed to meticulous polishing of the ceramic material when small occlusal corrections were deemed necessary after luting. This may have prevented micro cracks that could lead to catastrophic failures in due time [30]. An additional risk for crack formation is polymerization shrinkage of the luting composite, which creates stress concentrations at the adhesive interface and at the ceramic subsurface [48]. When using IDS with indirect bonded restorations, the delayed placement of the restorations and postponed occlusal loading facilitate the dentin bond to develop without stress [49]. The use of IDS may have led to less fractures and chippings in this study. However, the use of a lithium disilicate is also known to have higher mechanical properties that produces less fractures compared to other ceramics like leucite reinforced glass-ceramic and feldspathic glass-ceramics [1].

Chippings are reported to occur mainly at the marginal area of a restoration, involving small or severe material loss and leaving an irregular oblique fracture plane [50]. Because the condition of the restorations is difficult to assess (fracture lines and small chippings), impression taking for replicas with SEM recordings could provide additional information.

Most frequently observed failures were teeth needing endodontic treatment (1%, n = 7). The condition of the restorations involved with this complication was still good, with no secondary caries or defects of the restoration. From other studies, it has been reported that failures followed by endodontic complications are seen in 3% of the cases with ceramic and resin inlays, onlays and overlays [6] and in 15.6% of the cases with metal-ceramic crowns [51]. The low incidence in the presence study may be the result of the minimal invasive preparation design and the use of IDS. While a circumferential full-crown preparation is associated with the sacrifice of 67.5 to 75.6% of the original tooth structure, partial preparation is associated with substantially less sacrifice of healthy tooth tissue (5.5 to 27.2%) [52]. More invasive, circumferential crown preparations and the use of air turbines are correlated with an increase in pulpal complications [52, 53]. Exposed vital dentin immediately after tooth preparation is susceptible to insult from bacterial infiltration and micro-leakage during the provisionalization phase [13]. Bacterial and fluid penetration through the exposed dentinal tubules can result in colonization of micro-organisms, post-operative sensitivity and the potential for subsequent irritation of the pulp [13]. The use of Immediate Dentin Sealing is postulated to avoid these possible sequelae [13, 54], thereby playing an important role in keeping tooth vitality and possibly preventing teeth from hypersensitivity.

One experienced operator performed all operative procedures which increases internal, but decreases external, validity of the observations and stresses the importance of meticulous description of the operative procedures used. The difference between operators (for example seating pressure, operating time and experience) on clinical outcome is well recognized [26, 55]. However, in this study, a conventional photo-polymerized resin composite is used as a luting agent and this technique may be less technique-sensitive and therefore less prone to application errors. It is also easier to remove composite excess when a various direct composite resin rather than a low viscous material is used [26]. The use of high-quality intra-oral photographs mainly had a practical reason, so the clinical performance could be judged and critically appraised by one and the same observer, not being the treating physician, outside office times. This procedure may even be preferred over clinical evaluation when judging dental restorations [56].

The present investigation is a prospective clinical study; several clinical variables (restoration size and intra-oral distribution) could act as cofounders. Considering these limitations, further in vivo investigations would be necessary to validate the clinical performance and efficacy of lithium disilicate partial restorations, given different cementation procedures confirming the effectiveness of the approach used in this study.

Under the given circumstances and conditions as presented in this prospective study, partial glass-ceramic posterior restorations luted with a conventional photo-polymerized resin composite in conjunction with the use of Immediate Dentin Sealing have an excellent medium-term prognosis.