Prospective Randomized Clinical Trial on the Survival of Lithium Disilicate Posterior Partial Crowns Bonded Using Immediate or Delayed Dentin Sealing: Short-term Results on Tooth Sensitivity and Patient Satisfaction

Because of the advances in adhesive technologies and ceramic materials, the dental profession has positively embraced ceramic indirect restorations, as they require minimally invasive preparations. Today, biomechanically or esthetically compromised teeth can be restored with partial indirect restorations at a minimum biological price, saving sound tooth tissues. 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 (5.5% to 27.2%).¹  The amount of tooth structure reduction could also decrease the degree of postoperative pain.¹  According to histological studies, where a distance of 0.5 mm from the pulp could cause pulpal reaction in 60% of the cases, a similar reaction was observed in only 5% of cases in which the dentin thickness was greater than 1 mm.²  Moreover, not only the preparation depth but also the provisional phase could affect pulp reaction, namely, within one or two weeks, microorganisms could reach the pulp when the dentin is not adequately sealed with a provisional restoration. Therefore, bacterial infection in the dentin tubules was postulated to be one of the main causes of pulpal damage and pain.³  To reduce the risk of postpreparation sensitivity and irritation to pulp tissue, occluding or sealing of exposed dentin tubules may be necessary. A combination product (GLUMA Desensitizer, Heraeus Kulzer GmbH, Wehreim, Germany) with glutaraldehyde and hydroxyethyl methacrylate content resulted in a significant reduction in postpreparation sensitivity.⁴  Another cause of postoperative pain is postcementation hypersensitivity. This symptom is characterized by a short, sharp pain when thermal or chemical stimuli are introduced to vital abutment teeth after the cementation of a crown or fixed dental prosthesis.⁵  This type of hypersensitivity is usually self-healing but is uncomfortable and can persist for up to 24 months.⁶  The underlying cause of dentinal hypersensitivity is in part explained by the hydrodynamic theory in that a stimulus causes fluid to flow in the dentinal tubules, and this transmits a variety of physical stimuli, which eventually stimulates electrical nerve activity, resulting in a typical, sharp, transient pain.⁷ 

The clinical success of ceramic restorations relies substantially on the quality of their adhesion to dental tissues, where adhesion to dentin in particular remains a clinical challenge to date. Improvements have been made over the years in terms of material properties, and more effective methods have been developed to condition intaglio surfaces of ceramic substrates. In an attempt to improve the adhesion of indirect restorations to dentin and thereby diminish postoperative sensitivity, the immediate dentin sealing (IDS) technique has been suggested as an alternative to conventional adhesive luting, also referred to as delayed dentin sealing (DDS).^8-14  The major difference between the IDS and DDS technique lies in the fact that with the IDS, a thin layer of adhesive resin is applied immediately after tooth preparation and prior to impression making, whereas in the DDS, the adhesive resin layer is applied just before luting the restoration. With the IDS technique, improved patient comfort during the provisional phase, less need of anesthesia during luting of the restoration, and reduced postoperative sensitivity were reported.^15,16  However, IDS is an additional step in the workflow of indirect restorations, and the stability of IDS after cleansing procedures is controversial.¹⁵  A recent systematic review study emphasized the lack of randomized controlled trials and affirmed the beneficial effect of IDS to minimize tooth sensitivity.¹⁵ 

Therefore, the objectives of this study were to evaluate tooth sensitivity (objective and subjective) and patient satisfaction after bonding partial ceramic restorations with IDS or DDS techniques on the vital molar teeth, in a prospective, randomized within-subject clinical trial. The tested null hypothesis was that there would be no significant difference in tooth sensitivity or patient satisfaction with partial indirect restorations bonded with either IDS or DDS during the first year of clinical service.

Between December 2013 and May 2016, a total of 30 patients (13 women, 17 men; mean age: 54 years) with an indication for two indirect partial ceramic restorations for first or second vital molar teeth were recruited. Included teeth would typically be mechanically compromised because of recurrent caries, fractured cusps, old restorations with marginal leakage, and/or large restorations that would be difficult to restore with composite. Further inclusion criteria were the following: physically and psychologically able to tolerate conventional restorative procedures, good oral hygiene, presence of an intact buccal wall of the tooth, normal response on cold test, possible rubber dam application, presence of the antagonistic tooth, and willingness of the patient to return for follow-up examinations. Radiographs were taken preoperatively to verify pulpal/periapical and coronal tooth status.

The two teeth randomly received either IDS (test group, n=30) or DDS (control group, n=30) through the randomization software (www.randomizer.org). Hence, the study can be characterized as a randomized controlled, single-blind clinical trial with within-subject comparison. A consort flow chart showing the enrollment, intervention allocation, follow-up, and data analysis is presented in Figure 1. The study was approved by the Medical Ethics Committee of the University Medical Center Groningen, the Netherlands (ABR number: NL 45130), and all the patients were provided with informed consent. Distribution of restorations and extension of the restorations are presented in Table 1.

The brands, types, manufacturers, chemical compositions, and batch numbers of the main materials used in this study are listed in Table 2. After isolating the teeth with a rubber dam (Hygenic Dental Dam, Coltène/Whaledent Inc, OH, USA), the existing restorations were removed. This step was a major determinant with respect to the preparation design, for which as much sound tooth structure as possible was maintained within the recommendations of the manufacturer for the ceramic material to be used for onlays and inlays (IPS-e.max press, Ivoclar Vivadent, Schaan, Liechtenstein). The outline configuration was a butt shoulder, prepared with a diamond bur and ultrasonic handpiece (SONICflex prep ceram, KaVo GmbH, Biberach/Riss, Germany). All internal angles were smoothed to reduce stress concentration. A minimal thickness of 1 mm for the restorative material was allowed in all areas. The cusps were covered (1.5 mm) if the remaining tooth structure wall was less than 2-mm thick from its occlusal aspect or when the outline of the restoration would be in an area with static or dynamic antagonist contacts. Slight divergence with an angle of 100° to 120° between the proximal cavity walls and the prospective proximal inlay surfaces was provided. The dental technician blocked out any incidental undercuts in the teeth that were allocated to the control group (DDS); the remaining cases were compensated for by the IDS. Contacts at the marginal ridge were avoided.

The teeth of the test group received IDS (Clearfil SE Primer and Adhesive, Clearfil Majesty Flow, Kuraray) immediately after preparation (Table 3). Electrosurgery was performed in cases in which retraction of the gingiva was required for proper impression making. Impressions were made using a silicone impression material (Heavy and Ultra Light Body Aquasil, Dentsply, Milford, DE, USA) using an individually designed impression tray. Temporary restorations were then made chairside using a chemically polymerized resin composite material (Protemp, 3M ESPE, Neuss, Germany) and cemented using polycarboxylate cement (Durelon, 3M ESPE, St Paul, MN, USA). Any temporary restorations that were debonded before the second appointment were noted.

One dental technician fabricated all lithium disilicate restorations (IPS-e.max press, Ivoclar Vivadent). Two weeks after preparation, the temporary cement was removed from the teeth with an ultrasonic tip and a scaler. The IDS was checked for stability after cleansing of the preparation using magnification, but no detrimental effects were observed in any of the cases. The sequence of the different tooth-conditioning and restoration procedures, before cementation, is presented in Tables 3 and 4. The adhesive procedure differed between the test and control group, as outlined in these tables. All partial restorations were luted using a heated (55°C) dual-polymerized luting composite (Variolink Ultra, Ivoclar Vivadent). Restorations were placed initially under slight pressure, where the excess material was removed immediately from the margins with a probe, a scaler, and dental floss (Oral-B, Rotterdam, the Netherlands). After increasing the pressure, the final excess cement was manipulated against the tooth to prevent marginal gaps. The restorations were photo-polymerized (>1,000 mW/cm², Bluephase Style, Ivoclar Vivadent) for 40 seconds from three sides, and this was repeated after the application of glycerin gel. Occlusion and articulation were checked carefully using 40-μm carbon paper. The margins of the restorations were finished using a scaler and an ultrasonic device (EVA handpiece 7LP in combination with a 61 LG, KaVo GmbH) and polished using ceramic polishers (CeraGloss blue and yellow, Edenta, Argau, Switzerland). An intraoral radiograph was then made to check for excess cement in the cervico-approximal region. A typical case is presented in Figure 2.

The test and control teeth were initially dried with gauze. Subsequently, objective tooth sensitivity was measured using a coolant spray (Kälte spray, Orbis Dental, Münster, Germany) on a cotton ball (4 mm) that was immediately applied to the buccal wall of the corresponding tooth in each group, and the time until the patient responded was recorded. Regarding subjective tooth sensitivity, the patients were asked to indicate which one of the two teeth was the most sensitive or whether they were equally sensitive upon cold testing, when drinking hot/cold beverages, or while chewing. The patients were also asked to score the degree of experienced discomfort after the placement of the restoration using a visual analog scale (VAS), with scores ranging from none to very painful (Figure 3). Both evaluations were performed preoperatively and repeated after one week and then three and 12 months postoperatively.

To evaluate patient satisfaction, the patients were asked how satisfied they were with the restorations in general, with regard to the color and shape of the restorations, and the ability to chew (VAS, ranging from very satisfied to not satisfied) per tooth (Figure 4). These questions were posed to the patient after one week and then three and 12 months postoperatively.

Patients were instructed to call if they experienced any kind of failure (debonding or fracture) or serious discomfort or other concerns.

Statistical analysis was performed using SPSS 22.0 software for Windows (SPSS Inc, Chicago, IL, USA). Objective tooth sensitivity was compared at all time points using the McNemar test. To compare the frequency of subjective tooth sensitivity across treatments (IDS vs DDS), a chi-square test was applied to the measurements at each time point. The VAS items regarding tooth sensitivity and patient satisfaction were first tested for normality. As none of the items were distributed normally, all sensitivity and satisfaction items of both treatments were compared using Wilcoxon signed rank tests. Temporary crown loosening was compared across treatments using McNemar test. p values less than 0.01 were considered to be statistically significant in all tests.

After 1-year follow-up, all restorations (N=60) were in situ and showed a normal reaction to cold testing. None of the patients called to report failure, discomfort, or other concerns before the 1-year follow-up. No further technical or biological complications were noted. No tooth sensitivity change was noticed with partial preparations.

Tooth sensitivity reports of the patients did not show a significant difference between the preoperative phase and at all other time points (p>0.01; Table 5). There was also no significant difference between IDS and DDS (p>0.01) for all items in the questionnaire.

Patient satisfaction according to the VAS scores was not significantly different between the IDS and DDS groups for all items in the questionnaire at all time points (p>0.01; Table 5).

The frequency of debonding in the temporary restorations in the IDS group was higher (N=7; inlay: 1, two-cusp: 1, three-cusp: 2, four-cusp replacement: 3) than the control group (DDS; N=3, two-cusp: 1, three-cusp: 1, four-cusp: 1), but the difference in debonding rate per patient was not significantly different in both groups, χ²(1, n=60)=1.125, p=0.289.

This study was undertaken to investigate tooth sensitivity (objective and subjective) and patient satisfaction after bonding partial ceramic restorations using the IDS or DDS techniques on vital molar teeth. The key finding of this study indicated no significant difference in tooth sensitivity or patient satisfaction with partial restorations bonded with either IDS or DDS during the first year of follow-up. Thus, the null hypothesis was accepted.

The amount of tooth structure reduction is considered to be an important factor affecting postoperative tooth sensitivity.¹  One study that compared the effects of two different cements on tooth sensitivity of full circumferential crowns concluded that postcementation tooth sensitivity was a frequent complaint.¹⁷  This is in contrast to our findings, in which none of the patients complained about postcementation sensitivity, probably because of the minimal invasive preparation design. The use of IDS was postulated to reduce postoperative tooth sensitivity, improve comfort during the provisional restoration period, and necessitate anesthesia less frequently during cementation.^15,16  In this study, however, the added value of IDS on perceived tooth sensitivity was not demonstrated using a minimally invasive preparation approach. This is also in contrast with the findings of Hu and Zhu,¹⁶  who found that the IDS technique significantly reduced postcementation tooth sensitivity in abutment teeth restored with fixed dental prostheses. In that study, full circumferential crown preparations were performed. More postoperative sensitivity may be expected in full circumferential preparations as more dentin tubules are opened on the surface and these are intimately connected to the pulp. Also, Hu and Zhu¹⁶  used a bonding system with a separate etching step. Self-etching systems have a tendency to minimize postoperative sensitivity, compared with total-etch or milder etching systems, because of their higher qualitative and quantitative capacity of penetration.¹⁸  When the risk of postoperative tooth sensitivity is expected to be high, such as in deep preparations or with the use of total-etch systems, the addition of IDS may make a difference in reducing postoperative sensitivity.

In this study, normal response was noted upon provocation with a cold test after drying the teeth with gauze followed by a question about the difference in sensitivity between the teeth as an outcome variable for perceived sensitivity. Cold tests have been proved to be more reliable than heat tests in determining pulp vitality, but their accuracy is certainly not absolute.¹⁹ 

All test and control group evaluations on satisfaction were comparable. In fact, when people within the control group were already very satisfied, then it is difficult to demonstrate any relevant increase in the test group. Also, satisfaction was strongly related to communication, approach, and involvement and not always directly related with the clinical outcome. Moreover, not all satisfaction items were relevant when comparing differences between the test and control group.

A meta-analysis related to the topic indicated restoration failures in the first year predominantly due to endodontic complications.²⁰  The results of this study after 1 year were promising since no endodontic complications occurred within the first year. Teeth from the test group had a tendency to debond more often in the temporary crown phase than teeth in the control group. This did not influence the difference in patient-perceived tooth sensitivity between the two teeth, but it is certainly not desirable in a clinical setting. The temporary cement in this study was polycarboxylate cement, which reacts chemically with the calcium ions in the tooth structure. This may explain why the cement attached better to the tooth in the control group. In the test group, however, the dentin was covered with adhesive from the IDS layer. In such a situation, the cement cannot adhere adequately to the tooth and may yield to debonding. Our null hypotheses that there would be no significant difference in tooth sensitivity and patient satisfaction with partial indirect restorations bonded using IDS or DDS during the first year of clinical service were not rejected.

From this clinical study, the following can be concluded:

1. No tooth sensitivity change could be detected with a minimally invasive preparation design.

2. Perceived tooth sensitivity after partial preparation of vital molar teeth and their subsequent restoration with adhesively luted glass ceramic was not affected with the implementation of IDS or DDS at all follow-up moments for all items in the questionnaire.

3. Patient satisfaction with regard to the color and shape of the restorations and the ability to chew were not significantly different between the IDS and DDS groups at all time points.

The authors acknowledge Alette van Elk (TTL Oosterwijk Dental Laboratory Oosterwijk/Elysee, Groningen, the Netherlands) for her assistance in fabricating the ceramic inlays. The authors extend their gratitude to Ivoclar Vivadent (Schaan, Liechtenstein) and Kuraray (Osaka, Japan) for their generous provision of some of the materials used in this study.

This study was conducted in accordance with all the provisions of the local human subjects oversight committee guidelines and policies of the Medical Ethics Committee of the University Medical Center Groningen. The approval code for this study is: ABR number: NL 45130.

The authors' institutions supported this study. The authors declare that they have no conflicts of interest and that they did not have any commercial interest in any of the materials used in this study.