2.1

Experimental design

The present study is registered at ClinicalTrials.gov (NCT01461239) and was a parallel group randomized controlled clinical trial. The study was approved by the Research and Ethics Committee (Protocol 122/2009) of the Federal University of Pelotas, and described according to the CONSORT recommendations and based on an assumption of equivalence of treatments.

Teeth were restored with a glass fiber post cemented with regular or self-adhesive resin cement, composite core and a direct or indirect restoration, according to the randomization process. Patients were recalled up to 60 months for clinical and radiographic examination. Survival curves were created and the type of failure was evaluated.

2.2

Sample size calculation

Sample size calculation was based on the fact that previously published papers with similar design showed no differences between direct and indirect restorations for endodontically treated teeth . In that sense, if there is truly no difference between the standard (crown) and experimental treatment (direct composite resin), and considering that the average tooth survival rate after 5 years would be of 96%, 30 teeth per group would be required to be 90% sure that the limits of a two-sided 90% confidence interval will exclude a difference between the standard and experimental group of more than 18% (considered to be a clinically significant threshold), based on the equivalence of the treatments, and taking into account a possible 20% patient loss.

2.3

Inclusion and exclusion criteria

Adult patients seeking treatment in the Federal University of Pelotas, in need of endodontic and restorative treatment in teeth with at least one entire coronal wall remaining after endodontic procedures, were selected. In addition, patients should have good oral and general health and bilateral occlusal posterior contacts. Patient exclusion criteria were: financial limitations; untreated temporomandibular joint disorder; or extensive removable partial/complete dentures in the opposing jaw. Tooth exclusion criteria were: tooth mobility; periodontally compromised condition; or a periapical lesion that did not resolve after endodontic treatment. All participants signed written informed consent before being accepted into the study.

2.4

Randomization procedures

All teeth were randomized and assigned to each group using a computer-generated list of random numbers. Each number was written on a white paper and placed into brown envelopes, by a researcher not involved in the study according to the treatment previously randomized. As a result, the clinician and the patient (double-blind study design) only knew which type of restoration was going to be the final restoration (direct or indirect) after cementing the post and making the resin composite restoration. Allocation only occurred after making the restoration, when the envelope was opened and if the paper had “crown” written, the crown preparation was performed. If the paper had “resin” written, no preparation was performed and the resin composite restoration was finished and polished. The randomization sequence was stratified by tooth type, anterior, premolar or molar. Due to the slow uptake of patients, blocks of 10 patients were randomized after one year of the clinical trial to minimize unbalancing.

2.5

Clinical procedures

All root canal treatments were performed under rubber dam isolation and all materials were used according to the manufacturers’ instructions. Trained undergraduate and graduate students performed all procedures, including restorative reconstruction.

A crown-down technique was performed (2.5% sodium hypochlorite as irrigant) using files in ascending sizes. The root canals were filled with gutta-percha points (Coltene/Whaledent, Langenau, Germany) and cement (Endo-fill, Dentsply/Maillefer, Petrópolis, Brazil) by lateral and vertical condensation. The gutta-percha was immediately partially removed with a heated spreader and a #2 Gates-Glidden drill, leaving 4 mm of apical seal. Where appropriate a waiting period with temporary restoration was observed, to evaluate peri-apical healing. The post space was prepared using a calibrated bur corresponding to the glass fiber post number (#0.5 or 1, White Post DC, FGM, Joinville, SC, Brazil). After checking the fit, the posts were cleaned with alcohol and pretreated with silane (ProSil, FGM) , and luted according to protocol for the assigned cement. For the regular resin cement (RelyX ARC, 3 M ESPE, St Paul, USA), the dentinal walls of the post space was acid-etched using 37% phosphoric acid (Condac, FGM) and an adhesive system was applied (Adper Single Bond or ScotchBond Multi Purpose—3 M ESPE) followed by insertion of the resin cement using Centrix syringe (DFL Indústria e Comércio S.A., Rio de Janeiro, Brazil). Digital pressure was applied for 5 min excess of cement was removed and light-cured for 40 s/surface. The same procedures were performed for self-adhesive resin cement (RelyX U100, 3 M ESPE), but without the adhesive. After post cementation, radiographs were taken to check the location of the post. All heads of the posts were 2 mm sub-occlusal.

Direct restorations were made using a microhybrid resin composite (ScotchBond Multi Purpose + Filtek Z250, 3 M ESPE) with an incremental technique. Each increment was light-cured for 40s. All restorations were immediately finished with fine and ultra-fine diamond finishing-burs (KG Sorensen, Barueri, SP, Brazil) under water spray, and polished with Sof-Lex discs (3 M ESPE) and 0.1 μm particle size diamond paste 1–7 days later.

For indirect restorations, the core restoration was made with Scotchbond MP and Filtek Z250 using an incremental technique (curing 40 s/layer) with at least 2 mm composite covering the post. For the metal-ceramic crown preparation diamond burs were used. All margins were at gingival level and finished with a chamfer outline. The preparation had a space of at least 1.5 mm in all vertical surfaces and 2 mm for the occlusal surface. Impressions were made (Impregum F, 3 M ESPE) and the same laboratory made all crowns using cobalt-chromium alloy and ceramic. A temporary acrylic resin crown was placed using calcium hydroxide cement (Dycal, Dentsply, Petrópolis, RJ, Brazil). All crowns were placed using the self-adhesive cement RelyX U100.

2.6

Evaluation parameters

The date of crown cementation or completed restorative treatment for each patient was considered as baseline. Patients were recalled after 6 months and 1 year, and then yearly up to 5 years for clinical and radiographic examination. A single trained and calibrated independent examiner (Kappa 0.92) carried out the evaluation following the FDI criteria including esthetic, functional and biological properties, with a threshold for failure between scores 3 (clinically sufficient/satisfactory) and 4 (clinically unacceptable) . In addition, the type of failure was also recorded. If the patient returned to the evaluation with a tooth with a debonded post and/or restoration, or even extracted, the time of failure was based on patient’s report.

2.7

Statistical analysis

Statistical analyses were performed using SPSS 22 for MAC (SPSS Inc, Chicago, IL). Descriptive analyses were used to describe the patients/teeth included in the study and the reasons for failure. The longevity of the restorations and the teeth was analyzed using Kaplan-Meier statistics and log-rank tests for differences between groups (p < 0.05). Type of tooth (anterior, premolar or molar) and position (mandible or maxilla) were assessed. Absolute and relative failures were considered in the analysis, i.e. success (clinically acceptable without repair) and survival (tooth present, including repair). Annual failure rates were also calculated.

2.1

Experimental design

The present study is registered at ClinicalTrials.gov (NCT01461239) and was a parallel group randomized controlled clinical trial. The study was approved by the Research and Ethics Committee (Protocol 122/2009) of the Federal University of Pelotas, and described according to the CONSORT recommendations and based on an assumption of equivalence of treatments.

Teeth were restored with a glass fiber post cemented with regular or self-adhesive resin cement, composite core and a direct or indirect restoration, according to the randomization process. Patients were recalled up to 60 months for clinical and radiographic examination. Survival curves were created and the type of failure was evaluated.

2.2

Sample size calculation

Sample size calculation was based on the fact that previously published papers with similar design showed no differences between direct and indirect restorations for endodontically treated teeth . In that sense, if there is truly no difference between the standard (crown) and experimental treatment (direct composite resin), and considering that the average tooth survival rate after 5 years would be of 96%, 30 teeth per group would be required to be 90% sure that the limits of a two-sided 90% confidence interval will exclude a difference between the standard and experimental group of more than 18% (considered to be a clinically significant threshold), based on the equivalence of the treatments, and taking into account a possible 20% patient loss.

2.3

Inclusion and exclusion criteria

Adult patients seeking treatment in the Federal University of Pelotas, in need of endodontic and restorative treatment in teeth with at least one entire coronal wall remaining after endodontic procedures, were selected. In addition, patients should have good oral and general health and bilateral occlusal posterior contacts. Patient exclusion criteria were: financial limitations; untreated temporomandibular joint disorder; or extensive removable partial/complete dentures in the opposing jaw. Tooth exclusion criteria were: tooth mobility; periodontally compromised condition; or a periapical lesion that did not resolve after endodontic treatment. All participants signed written informed consent before being accepted into the study.

2.4

Randomization procedures

All teeth were randomized and assigned to each group using a computer-generated list of random numbers. Each number was written on a white paper and placed into brown envelopes, by a researcher not involved in the study according to the treatment previously randomized. As a result, the clinician and the patient (double-blind study design) only knew which type of restoration was going to be the final restoration (direct or indirect) after cementing the post and making the resin composite restoration. Allocation only occurred after making the restoration, when the envelope was opened and if the paper had “crown” written, the crown preparation was performed. If the paper had “resin” written, no preparation was performed and the resin composite restoration was finished and polished. The randomization sequence was stratified by tooth type, anterior, premolar or molar. Due to the slow uptake of patients, blocks of 10 patients were randomized after one year of the clinical trial to minimize unbalancing.

2.5

Clinical procedures

All root canal treatments were performed under rubber dam isolation and all materials were used according to the manufacturers’ instructions. Trained undergraduate and graduate students performed all procedures, including restorative reconstruction.

A crown-down technique was performed (2.5% sodium hypochlorite as irrigant) using files in ascending sizes. The root canals were filled with gutta-percha points (Coltene/Whaledent, Langenau, Germany) and cement (Endo-fill, Dentsply/Maillefer, Petrópolis, Brazil) by lateral and vertical condensation. The gutta-percha was immediately partially removed with a heated spreader and a #2 Gates-Glidden drill, leaving 4 mm of apical seal. Where appropriate a waiting period with temporary restoration was observed, to evaluate peri-apical healing. The post space was prepared using a calibrated bur corresponding to the glass fiber post number (#0.5 or 1, White Post DC, FGM, Joinville, SC, Brazil). After checking the fit, the posts were cleaned with alcohol and pretreated with silane (ProSil, FGM) , and luted according to protocol for the assigned cement. For the regular resin cement (RelyX ARC, 3 M ESPE, St Paul, USA), the dentinal walls of the post space was acid-etched using 37% phosphoric acid (Condac, FGM) and an adhesive system was applied (Adper Single Bond or ScotchBond Multi Purpose—3 M ESPE) followed by insertion of the resin cement using Centrix syringe (DFL Indústria e Comércio S.A., Rio de Janeiro, Brazil). Digital pressure was applied for 5 min excess of cement was removed and light-cured for 40 s/surface. The same procedures were performed for self-adhesive resin cement (RelyX U100, 3 M ESPE), but without the adhesive. After post cementation, radiographs were taken to check the location of the post. All heads of the posts were 2 mm sub-occlusal.

Direct restorations were made using a microhybrid resin composite (ScotchBond Multi Purpose + Filtek Z250, 3 M ESPE) with an incremental technique. Each increment was light-cured for 40s. All restorations were immediately finished with fine and ultra-fine diamond finishing-burs (KG Sorensen, Barueri, SP, Brazil) under water spray, and polished with Sof-Lex discs (3 M ESPE) and 0.1 μm particle size diamond paste 1–7 days later.

For indirect restorations, the core restoration was made with Scotchbond MP and Filtek Z250 using an incremental technique (curing 40 s/layer) with at least 2 mm composite covering the post. For the metal-ceramic crown preparation diamond burs were used. All margins were at gingival level and finished with a chamfer outline. The preparation had a space of at least 1.5 mm in all vertical surfaces and 2 mm for the occlusal surface. Impressions were made (Impregum F, 3 M ESPE) and the same laboratory made all crowns using cobalt-chromium alloy and ceramic. A temporary acrylic resin crown was placed using calcium hydroxide cement (Dycal, Dentsply, Petrópolis, RJ, Brazil). All crowns were placed using the self-adhesive cement RelyX U100.

2.6

Evaluation parameters

The date of crown cementation or completed restorative treatment for each patient was considered as baseline. Patients were recalled after 6 months and 1 year, and then yearly up to 5 years for clinical and radiographic examination. A single trained and calibrated independent examiner (Kappa 0.92) carried out the evaluation following the FDI criteria including esthetic, functional and biological properties, with a threshold for failure between scores 3 (clinically sufficient/satisfactory) and 4 (clinically unacceptable) . In addition, the type of failure was also recorded. If the patient returned to the evaluation with a tooth with a debonded post and/or restoration, or even extracted, the time of failure was based on patient’s report.

2.7

Statistical analysis

Statistical analyses were performed using SPSS 22 for MAC (SPSS Inc, Chicago, IL). Descriptive analyses were used to describe the patients/teeth included in the study and the reasons for failure. The longevity of the restorations and the teeth was analyzed using Kaplan-Meier statistics and log-rank tests for differences between groups (p < 0.05). Type of tooth (anterior, premolar or molar) and position (mandible or maxilla) were assessed. Absolute and relative failures were considered in the analysis, i.e. success (clinically acceptable without repair) and survival (tooth present, including repair). Annual failure rates were also calculated.