Dental adhesive microtensile bond strength following a biofilm-based in vitro aging model

Abstract Laboratory tests are routinely used to test bonding properties of dental adhesives. Various aging methods that simulate the oral environment are used to complement these tests for assessment of adhesive bond durability. However, most of these methods challenge hydrolytic and mechanical stability of the adhesive- enamel/dentin interface, and not the biostability of dental adhesives. Objective To compare resin-dentin microtensile bond strength (μTBS) after a 15-day Streptococcus mutans (SM) or Streptococcus sobrinus (SS) bacterial exposure to the 6-month water storage (WS) ISO 11405 type 3 test. Methodology A total of 31 molars were flattened and their exposed dentin was restored with Optibond-FL adhesive system and Z-100 dental composite. Each restored molar was sectioned and trimmed into four dumbbell-shaped specimens, and randomly distributed based on the following aging conditions: A) 6 months of WS (n=31), B) 5.5 months of WS + 15 days of a SM-biofilm challenge (n=31), C) 15 days of a SM-biofilm challenge (n=31) and D) 15 days of a SS-biofilm challenge (n=31). μTBS were determined and the failure modes were classified using light microscopy. Results Statistical analyses showed that each type of aging condition affected μTBS (p<0.0001). For Group A (49.7±15.5MPa), the mean μTBS was significantly greater than in Groups B (19.3±6.3MPa), C (19.9±5.9MPa) and D (23.6±7.9MPa). For Group D, the mean μTBS was also significantly greater than for Groups B and C, but no difference was observed between Groups B and C. Conclusion A Streptococcus mutans- or Streptococcus sobrinus-based biofilm challenge for 15 days resulted in a significantly lower μTBS than did the ISO 11405 recommended 6 months of water storage. This type of biofilm-based aging model seems to be a practical method for testing biostability of resin-dentin bonding.


Introduction
Although dental composite restorations demonstrate favorable immediate bonding properties, failures occur over clinical service due to secondary caries, marginal defects or staining, chipping, and fractures. 1 The degradation of the resin-dentin interface is considered the weak link and it is cited as one of the main reasons for failure of composite restorations. 2 These facts, as well as newer dental adhesives claiming improved bonding effectiveness with simplified application techniques, emphasize the relevance of testing the resin-dentin interface integrity before dental bonding agents are marketed.
Clinical trials provide an accurate and effective determination of the long-term bond effectiveness for new materials, and they constitute the highest standard for testing them. 3,4 However, due to time and  3,4,7 However, these techniques present disadvantages to simulated aging, which ultimately limits their use. Water storage is time-consuming, requiring a minimum of 6 months. 6,8 Thermocycling methods lack a consensus regarding the ideal number of cycles or cycling protocol. 9 Furthermore, these methods challenge mechanical and hydrolytic stability only, leaving enzymatic stability untested. Therefore, it would be desirable to adopt a more relevant method for simulated aging in a laboratory that can simultaneously speed the aging process for resin-dentin interfaces and help evaluate an adhesive long-term bonding properties.
In the oral cavity, bacteria can form biofilms on both soft and hard tissue, which includes a variety of restorative materials, such as ceramics, resin composites, and amalgams. 10 Plaque bacteria and resin-based dental materials interact dynamically. because each extraction was performed for purely clinical reasons. Furthermore, the teeth could not be connected to the patient from which they were extracted.

Specimen Preparation
Thirty-one extracted teeth were cleaned and stored in 0.5% Chloramine-T trihydrate bactericidal reagent until they were mounted in dental stone using a customized mold. The teeth were then trimmed to create a flat coronal dentin using a water-cooled diamond wheel (Whip Mix, Louisville, KY, USA) and the Computer Numeric Controlled (CNC) machine (University of Iowa, Iowa City, IA, USA). The exposed dentin surface was etched with 37% phosphoric acid A radiometer (Demetron/Kerr, Danbury, CT, USA) was used to evaluate the stability of energy output throughout the study. Each bonded assembly was then segmented perpendicular to the resin-dentin interface into four sticks using an Isomet 1000 sectioning machine (Buehler, Lake Bluff, IL, USA). Each of the 2 mm x 2 mm resin-dentin stick was further trimmed using the CNC machine into a dumbbell with crosssectional area of 0.5 mm 2 , gauge length of 1 mm, and radius of curvature or 'neck' of 0.6 mm. Dumbbells were sterilized by storing them in 0.5% Chloramine-T disinfectant reagent (0.5% of chloramine-T trihydrate with autoclaved water) for 24 hours followed by rinsing five times with autoclaved water before being exposed types of storage media by Wald chi-square test. 8,22,23 The significance for all tests was 0.05, and the SAS for Windows version 9.4 (SAS Institute Inc., Cary, NC, USA) was used to perform data analysis.

Results
Mixed model ANOVA and Weibull regression model (Tables 1 and 2, respectively) showed that the method of simulated aging significantly (p<0.0001) affected the μTBS of dentin adhesive tested. Mean μTBS following 6 months of water storage was significantly higher than that following 5.  (Table 1).
Also, mean μTBS of specimens subjected to 15 days of S. sobrinus challenge was significantly greater than that following 5.5 months of water storage + 15 days of S. mutans challenge, or 15 days of S. mutans challenge, whereas no difference was found between the latter two groups ( Figure 2). Weibull distribution was similar for all groups, as represented by the shape parameter. The scale parameter represented by η (eta) (63.2% probability of failure) is also shown in Table 2. Regarding the failure mode, most of specimens (74.2%, 83.9%, 80.6%) exposed to bacterial challenge (Groups B, C and D, respectively), had apparent cohesive failures within the dentin substrate, very close to the adhesive interface ( Figure   3). In total, 55% of the specimens exposed to water challenge (Group A) had apparent cohesive failures within the dentin or dental composite substrate.  However, as we did not find a difference between both groups likely signifies that most of the degradation presented in Group B (5.5 months of WS + 15 days of SM) was due to SM leading to low μTBS.

Groups (N=31) Description of Aging Conditions Mean (SD) Microtensile Bond Strength (Mpa)*
In past studies, researchers have investigated how a resin-dentin interface degrades when challenged with a multi-species biofilm. 20,32,33 Although more clinically simulative, this kind of multi-species biofilm challenge is hard to control and standardize, since one of the bacterial species might dominate and outgrow the others. In this study, single species biofilms developed from S. mutans-and S. sobrinus were used. The strain of S. mutans selected for the study was UA159. This strain was chosen for it is well-studied and exhibits greater esterase activity on resin substrates, which are common in dental restorations, than other S. mutans strains. 18 The strain of S. sobrinus selected for the study was ATCC 33478. This strain was used for it displays typical properties of the species. For the first 24 hours, a BHI medium was supplemented with sucrose to establish a biofilm. Sucrose was not used afterwards to avoid S. mutans to bind to and build up too much on the specimens for the remainder of the biofilm challenge. If active gripping with glue was used to test for bond strength, a larger biomass would have created difficulties, but since we used a mechanically passive gripping device without glue, it did not affect our study.
The μTBS test was used instead of the traditional macro-shear test due to improved stress distribution at the true resin-dentin interface and to achieve accelerated degradation at short diffusional distance and relatively larger adhesive dentin margin exposure. 7 Despite being technically demanding, cylindrical dumbbell-shaped specimens were used instead of rectangular resin-dentin beams due to more uniform stress distribution at the dentin-resin interface under tensile load, thereby providing more reliable results. 7,34 For data analysis of the results, both ANOVA procedures and a Weibull regression model were used. The Weibull regression analysis is highly recommended, since it can account for variations in μTBS results and for clusters of samples that occur when four dumbbells obtained from a single tooth are used. 8,22,35 Most research studies avoid the test because of the large sample size it requires. 22 In this study, a sample size one-third larger would have been adequate if we were to use only ANOVA for data analysis. Future work may address some limitations of this study. Fractography and tracing the bacterial infiltration pathway using Scanning Electron Microscopy or Confocal Laser Scanning Microscopy would be useful for understanding biodegradation. Our study did not include an uninoculated BHI control. However, a significant difference was found between μTBS values of groups exposed to S. mutans and to S. sobrinus, therefore indicating probable degradation due to the bacteria used, and not the BHI media. Furthermore, a study measuring the quantity of Bis-GMA-derived degradative product bishydroxy-propoxy-phenylpropane (Bis-HPPP) did not find a significant difference in degradation of composite (Z-250)  It is also important to understand that different degradation mechanisms occur at distinct parts of the restorations and enamel or dentin tooth substrates.
Future studies could compare exposed restored teeth with and without enamel margins as well as individual test specimens; however, we chose to study the resin-dentin bond as this is a common restorative cavosurface margin and site of margin degradation to include recurrent caries. The resin-dentin interface is the weak link of adhesive dentistry due to the dentin substrate nature and the in vivo degradative mechanisms.
In summary, biofilm challenge for just 15 days produced significantly greater degradation and resulted in much lower μTBS values than did water storage for 6 months. This suggests that a biofilm challenge used to evaluate the hydrolytic and biostability of dental adhesives has a clear purpose while testing mechanical properties of bonding agents in the laboratory. However, the extent to which this model decreases the resin-dentin bond strength when compared to non-aged specimens cannot be determined in this study. The bacterial challenge can be helpful to assess how newer antibacterial resin monomers bond to dentin substrate. 37 The method of simulated aging in this study has been deliberately kept simple and it is easily to reproduce.
The biofilm-based model seems to be a promising in vitro method for simulated aging. However, this area needs further refinement and exploration into how well it complements or can replace other in vitro aging models dedicated to testing μTBS properties of dental adhesives.

Conclusion
Within the limitations of this in vitro study, a Streptococcus mutans-or Streptococcus sobrinusbased biofilm challenge for 15 days resulted in a significantly lower μTBS than the ISO 11405 recommended 6 months of water storage. This type of biofilm-based aging model seems to be a practical method for testing biostability of resin-dentin bonding.