2.1

Prospective clinical trial

Patients selected for this study met the following criteria: (1) absence of pain from the tooth to be restored; (2) possible application of rubber dam during placement of restoration; (3) no further restorations planned in other posterior teeth; (4) good level of oral hygiene; (5) absence of any active periodontal and pulpal pathology; (6) restorations required in two different quadrants (split mouth design); (7) age 18–65; (8) no pregnancy.

The study was approved by an Ethics Committee (University Clinic Erlangen, Germany). All patients were required to give written informed consents before starting the study and agreed to participate in a recall program. Thirty patients (23 female, 7 male, mean age 32.9 (24–59) years) with a minimum of two fillings to be replaced in different quadrants received at least two different restorations in a random decision according to recommendations of the CONSORT statement . Sample size calculation was carried out according to previous clinical studies , occluding teeth were not excluded.

Thirty-six Grandio fillings were bonded using an etch-and-rinse technique using Solobond M (Voco, Cuxhaven, Germany) and 32 Tetric Ceram restorations were bonded with Syntac (Ivoclar Vivadent, Schaan, Principality of Liechtenstein). All fillings (only Class II, 52 MO/OD, 16 MOD or more surfaces (i.e. buccal or lingual), no complete cusp replacements) were re-restorations made by one dentist in a private practice (31 upper bicuspids, 12 upper molars, 14 lower bicuspids, 11 lower molars). Reasons for replacement were caries ( n = 19), insufficient esthetics ( n = 2), and secondary caries beneath amalgam restorations ( n = 47). For all teeth receiving restorations, current X-rays (within six months of the procedure) were present. After evaluating the radiographs, 53 cavities (78%) were treated as “caries profunda”. Twenty-four cavities (35%) revealed no enamel at the floor of the proximal box, while 33 cavities (49%) exhibited a proximal enamel width of <0.5 mm.

All fillings were inserted in permanent vital teeth without pain symptoms. The cavities were cut using coarse diamond burs under profuse water cooling (80 μm diamond, Komet, Lemgo, Germany), and finished with a 25 μm finishing diamond. Inner angles of the cavities were rounded and the margins were not beveled. After cleaning and drying under rubber dam isolation (Coltene/Whaledent Inc., Altstätten, Switzerland), adhesive procedures were performed with Solobond M (2-step etch-and-rinse adhesive) and Syntac (4-step etch-and-rinse adhesive). The resin composite materials were applied into the cavity in layers of approximately 2 mm thickness and adapted to the cavity walls with a plugger. Each layer was light cured for 40 s (Elipar Trilight, 3M ESPE, Seefeld, Germany). The occlusal region was modeled as exactly as possible under intraoral conditions, avoiding visible overhangs. The light-emission window was placed as close as possible to the cavity margins. The intensity of the light was checked periodically with a radiometer (Demetron Research Corp., Danburg, CT, USA) and was found to be constantly above 650 mW/cm ² .

As soon as polymerization was completed, the surface of the restoration was controlled for defects and corrected when necessary. Visible overhangs were removed with a scaler and the rubber dam was removed. Contacts in centric and eccentric occlusion were controlled with foils (Roeko, Langenau, Germany) and adjusted with finishing diamonds (Komet Dental, Lemgo, Germany), shaped with flexible discs (3M Dental, St. Paul, USA), super-fine discs (3M Dental, St. Paul, USA) and polishing brushes (Hawe-Neos Dental, Bioggio, Switzerland).

At the initial recall (baseline, i.e. within 2 weeks), and after six months, one, two, four, six, eight, and 10 years, all restorations were assessed according to the modified United States Public Health Service (USPHS) criteria ( Table 1 ) by two independent investigators (dentists, both chairpersons) using loups with 3.5× magnification, mirrors, probes, bitewing radiographs, and intraoral photographs (Nikon D90, Sigma-Makro 105 mm) with and without occlusal contact points. Investigators were blinded, trained and calibrated through eight previous clinical studies and additional calibration sessions. Recall assessments were not performed by the clinician who initially placed the restorations.

Statistical appraisal was computed with SPSS for Windows XP 15.0 (SPSS Inc., Chicago, IL, USA). Statistical unit was one tooth, Kolmogorov Smirnov test revealed no normal distribution of values, so differences between restorative groups were evaluated using Mann–Whitney U -test, changes regarding different criteria over time were calculated with the Friedman test ( p = 0.05). Survival analysis was conducted using the Kaplan–Meier algorithm, additionally confidence intervals for both types of restorations were calculated.

2.2

SEM evaluation

In each recall, silicone impressions of restored teeth were taken (Dimension® Penta and Garant, 3M Oral Care, Germany, applied with syringes to get proximal parts recorded; and Minitray®, Hagen & Werken, Duisburg, Germany) and replica made using an epoxy resin (AlphaDie, Dürr-Dental, Rosbach, Germany). Inclusion criteria for SEM analysis were (1) no missed recall for continuous evaluation (max. 10% deviation), (2) impressions free of artifacts from every recall, (3) two restorations per patient (split-mouth), (4) both restorative systems evaluable in one patient. Inclusion criteria were fulfilled by 21 patients (i.e. 42 restorations, 21 of each material). Average observation time was 12.1 years. Localization of restored teeth were 19 upper premolars, 8 upper molars, 9 lower premolars, and 6 lower molars.

Replicas were mounted on specimen holders (Leit-C, Plano GMBH, Wetzlar, Germany) and then sputtered with gold at 20 mA for 75 s (Polaron SC 502 Sputter Coater, Quorum Technologies, Laughton East Sussex, UK). Replicas were assessed at 10×- and 200×-magnification at 12–16 kV under a SEM (Amray 1610 Turbo, KLA, Milpitas, CA, USA) using an overlap technique (Fiji-win32, http://fiji.sc/wiki/index.php/Downloads , and KHKs_iQuantiGap by Karl-Heinz Kunzelmann, LMU Munich, Germany). Semi-quantitative margin analysis was carried out using the criteria “perfect margin”, “overhang”, “positive step formation (restoration level higher than surrounding tooth)”, “negative step formation” (restoration level lower than surrounding tooth), “gap”, “fracture”, and “not judgeable” ( Fig. 1 ).

Evaluation criteria of margin analysis (perfect margin: red line; negative step formation: yellow line; overhang: purple line; positive step formation: pink line; gap formation: dark blue line; fracture of tooth or restoration: green line; not assessable: bright blue line). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Statistical appraisal was computed with SPSS for Windows 15.0 (SPSS Inc., Chicago, IL, USA). Statistical unit was one specimen, Kolmogorov Smirnov test revealed no normal distribution of values, so differences between groups were evaluated using Mann–Whitney U -test, changes over time regarding all SEM criteria were calculated using the Friedman test ( p = 0.05).