Laboratory methods to simulate the mechanical degradation of resin composite restorations

Abstract

Objectives:

This study reviewed the literature to identify in vitro approaches that have been used to simulate the mechanical degradation and fatigue of resin composite restorations.

Methods:

A search for articles was carried out in 4 databases and included studies in which composite restorations were bonded to teeth and subject to cyclic loading. Articles were assessed for eligibility, and the following items were the extracted from the included studies: authors, country, year, materials tested, simulation device and details including load magnitude and frequency, number of cycles, type of antagonist, test medium, and temperature. Data were analyzed descriptively.

Results:

The 49 studies included showed a high level of heterogeneity in methods, devices, and test parameters. Nineteen different simulation devices were used, applying loads varying between 30 and 2900 N, and frequencies varying between 0.4 and 12 Hz. The load and frequency used most often were ~ 50 N (63.3%) and 1.5–1.7 Hz (32.7%). The number of cycles varied between 10 K and 2.4 M, 1.2 M was the most prevalent (40.8%). The majority of studies combined cyclic loading with at least one additional aging method: static liquid storage, thermo-mechanical cycling applied simultaneously, and thermal cycling as a discrete aging step were the three most frequent methods. The overall evidence indicated reporting problems, and suggested a lack of clinical validation of the research methods used.

Significance:

Validation studies, underlying clinical supporting data, and better reporting practices are needed for further improving research on the topic. Specific suggestions for future studies are provided.

1. Introduction

Polymer-based restorative materials are subject to aging and deterioration in the oral environment, a process that may result in irreversible changes in their original properties. These changes may involve the chemical structure, physical properties, or esthetic appearance of restorative materials, and thus affect their clinical performance [1-3]. The complex, dynamic oral environment poses a multitude of biological, chemical, physical, and mechanical challenges [1,4]. During masticatory cycles, for instance, a wide range of loads may generate stresses on the dentition and restorative materials [5,6]. In sound tooth structures, occlusal and other mechanical loads are transmitted through the dental hard tissues mainly as compressive forces. In restored teeth, occlusal loads may also develop tensile and shear stresses within the structure or along the tooth-restoration interface [1,7], due to a stark difference in elastic properties between restorative materials and tooth structures [8] coupled with the complex geometry of restorations [9].

Because of the dynamic interactions among mechanical and other challenges in the mouth, fundamental material properties of dental resin composites, e.g. flexural strength, elastic modulus, and fracture toughness may not be precise predictors of their clinical performance [10,11]. Fracture failures often occur due to damage accumulating from cyclic loading rather than a single-cycle overloading event; thus, it is suggested that the fatigue properties of resin composite restorations may help in predicting their clinical performance [12]. Fatigue is a mechanical degradation process [13] in which the load-bearing capacity of a material or restored tooth is reduced by repetitive or cyclic stresses, with accumulation and growth of damage or subcritical defects at subcritical loads [10,11]. In order to predict the behavior of resin composites in vivo, in vitro experimental setups and methods should simulate the oral environment and emulate the degradation processes that occur in the mouth.

Laboratory methods that intend to reproduce clinical conditions should involve clinically-relevant specimen geometries, preparation of restorations, and loading conditions [10]. Several chewing simulators and different methods to promote fatigue and mechanical degradation of resin-based composites are available, but to date no overview or mapping of these methods have been carried out. This type of information could be valuable for future development, evaluation, and validation of in vitro methods that aim to simulate intraoral conditions. The aim of this literature review was to identify in vitro approaches that have been used to simulate the mechanical degradation and fatigue of resin composite restorations bonded to tooth structures. This study also aimed to summarize the methods used in the publications and identify current problems and future challenges related to the in vitro simulation of the mechanical fatigue of resin composite restorations.

2. Methods

2.1. Eligibility criteria

The inclusion criteria consisted of in vitro studies in which resin composite restorations were subject to cyclic loading, i.e., laboratory research that endeavored to generate and evaluate the mechanical degradation of composite restorations. Studies that used mechanical devices to simulate chewing or other intraoral mechanical challenges were included, whereas the absence of a mechanical cyclic loading component was a factor for exclusion. Studies evaluating the fracture resistance, number of cycles until restoration failure, restoration retention, and fatigue resistance were included, whereas studies that used cyclic loading but restricted the evaluation to wear and surface effects, marginal integrity of restorations, and gap formation were excluded. In addition, the studies had to have tested resin composite specimens that were adhesively bonded to human tooth structures to be considered restorations, which means that studies using non-bonded specimens (e.g., disks, bars, cylinders) and tooth analogs were excluded. Studies reporting load to fracture tests were included, provided that the mechanical degradation had been simulated before the test was performed. Studies testing only fiber-reinforced composites were excluded to concentrate the investigation on conventional composites, which represent the current standard of care for restorative adhesive approaches. Articles that combined a mechanical cyclic component with non-mechanical aging procedures (e.g., thermal cycling, static water storage) also were included and the additional procedures were registered. Studies on finite element analysis alone were excluded. Only articles published in English were considered, with no restrictions on the date of publication.

2.2. Information sources and search

Four electronic databases were independently searched by two reviewers (VPL, JBM): PubMed/MEDLINE, Embase, Scopus, and Web of Science. A search strategy was developed using MeSH terms, which were provided by PubMed (Table 1) and adapted to the other databases. The last search was performed in July, 2020. The reviewers also manually searched the reference lists of the articles included for additional relevant studies.

Table 1.

Search strategy used in PubMed and adapted to the other databases.

	Search terms
#5	Search #1 AND #2 AND #3 AND #4
#4	Aging[tiab] OR Ageing[tiab] OR Chew*[tiab] OR Cycl*[tiab] OR load*[tiab] OR Masticat*[tiab] OR Simulat*[tiab]
#3	"Dental Stress Analysis"[Mesh] OR "Dental Restoration Failure"[Mesh] OR Compressive Strength[tiab] OR Degradation[tiab] OR failure behavio*[tiab] OR Fatigue[tiab] OR Flexural Strength[tiab] OR Fracture*[tiab] OR Load to failure[tiab] OR Mechanical Stress*[tiab] OR Restoration failure*[tiab] OR Stress analys*[tiab]
#2	"Dental Restoration, Permanent"[Mesh] OR restoration*[tiab] OR filling*[tiab] OR inlay*[tiab] OR onlay*[tiab]
#1	"Composite Dental Resin" [Supplementary Concept] OR "Composite Resins"[Mesh] OR composite resin*[tiab] OR composite[tiab]

2.3. Selection of sources of evidence

Search results were exported to Endnote X7 (Thompson Reuters, Philadelphia, PA, USA). Duplicates were removed and the results were imported into an online management platform for systematic reviews (Rayyan by Qatar Computing Research Institute, Doha, Qatar). Titles and abstracts were screened by the two independent reviewers, who determined whether or not the studies should be considered for full-text reading. Any disagreement regarding the eligibility criteria was resolved through discussion and consensus, or a third reviewer was consulted (RRM).

2.4. Data charting, data items, and analysis

A data extraction template was developed using a spread-sheet software (Excel, Microsoft, Redmond, Washington, USA), which was pilot tested by two reviewers (VPL, RRM) to reach a consensus on which data to collect and how. Then, one reviewer (VPL) extracted the following data items: authors; country of the corresponding author; year of publication; study objective; materials tested; type of simulation of oral conditions (e.g., mechanical cyclic loading, thermo-mechanical cyclic loading, step-stress analysis); simulation device; simulation details including load magnitude and frequency, number of cycles, type of antagonist material, test medium and temperature; and main findings. Extractions were revised by a second reviewer (RRM). Data were analyzed descriptively.

3. Results

The search resulted in the retrieval of 6767 records (Fig. 1). After removal of duplicates, 2960 unique articles were screened, and 2824 were excluded based on the eligibility criteria. A total of 136 full-text articles was assessed for eligibility, of which 87 were excluded for the reasons presented in Fig. 1. The main reason for exclusion was that the specimens were not bonded resin composite restorations. In total, 49 studies were included for charting and discussion.

Fig. 1.

Flowchart of selection of studies in this scoping review.

3.1. Characteristics of studies

Table 2 presents characteristics of all 49 studies, including variables related to tests and specimens. The studies were published between 1994 and 2020, with one team contributing to 14.3% of the articles included [14-20]. The following were the most frequent countries in which the corresponding authors resided: Brazil (24.5%), Germany (24.5%), Switzerland (10.2%), and the USA (10.2%). In total, 19 different simulation devices from varied manufacturers were used, although 7 studies did not provide information on which simulation device was tested and 5 other studies only reported generic information on the device, with no identification of brand or manufacturer. The mechanical challenge was applied to composite restorations with loads varying between 30 N and 2900 N, and frequencies varying between 0.4 Hz and 12 Hz. The load and frequency used most often were ~ 50 N (63.3%) and 1.5–1.7 Hz (32.7%), respectively. All studies reported the loading forces that were applied to specimens, but 8.2% of articles did not report the frequency of mechanical cycling. The number of cycles varied between 10 K and 2.4 M, with 1.2 M as the most common number (40.8%). Spheres and balls were the most frequent antagonist shapes (51%), the most frequent antagonist materials were metal, steel, and stainless steel (30.6%). Distilled water (47%) was the most common type of storage medium used during the test; the test temperatures varied between 4 °C and 60 °C. The majority of studies tested indirect or CAD-CAM resin composite materials (59.2%) and 93.9% of the studies focused on the posterior dentition, usually testing molars (59.2%) and often mesial-occlusal-distal tooth preparations (34.7%).

Table 2.

Summarizing chart for the studies included (n = 49).

First author, year	Characteristics of the test								Characteristics of the specimens
	Aging method	Simulation device (manufacturer)	Load, N	Frequency, Hz	Number of cycles	Antagonist	Test medium	Temperatureb	Resin composite	Format	Tooth
Brunner and Özcan, 2020 [68]	TMCL + load to fracture	Custom-made chewing simulator - CoCoM (University of Zurich)	50	1.67	1.2 × 106	Steel sphere	Distilled water	5–55 °C	Direct	Fixed dental prosthesis	Premolars and molars
Frankenberger et al., 2020 [69]	Liquid storage + TMCL	Chewing simulator CS4 + THE-1100 (SD Mechatronik, Germany)	50	0.5	1 × 105; 5 × 105	Steatite	Water	5–55 °C	Direct	MOD restorations	Molars
Grubbs et al., 2020 [70]	Liquid storage + mechanical cyclic loading + load to fracture	Wear instrument - Modified University of Alabama wear simulator (USA)	65	1.2	1 × 105	Steel sphere	Water	37 °C	Direct	MOD restorations	Molars
Prechtel et al., 2020 [71]	TMCL	Chewing simulator - CS-4.8 (SD Mechatronik, Germany)	50	1.2	1.2 × 106	Stainless steel sphere	Distilled water	5–55 °C	Direct	Class I restorations	Molars
Rosentritt et al., 2020 [72]	Liquid storage + TMCL + load to fracture	Chewing simulator - eGo Kältesysteme (Regensburg, Germany)	50	1.6	1.2 × 106	Steatite ball	Distilled water	5–55 °C	CAD/CAM	Crowns	Molars
Shafiei et al., 2020 [73]	Liquid storage + mechanical cyclic loading + thermal cycling + load to fracture	Chewing Simulator CS4 (SD Mechatronik, Germany)	50	0.5	1 × 105	Stainless steel	Water	1500 cycles; 5–55 °C	Direct and Indirect	MOD restorations	Premolars
Zhang et al., 2020 [74]	Liquid storage + thermal cycling + mechanical cyclic loading + load to fracture	N.I.	50	1.3	1.2 × 106	Stainless steel sphere	Water	6400 cycles; 5–55 °C	Indirect	Occlusal veneers	Premolars
Bijelic-Donova et al., 2020 [75]	Mechanical cyclic loading + load to fracture	Chewing simulator - CS-4.2 (SD Mechatronik, Germany)	85	1.5	1.2 × 105	Metal sphere	Distilled water	23 °C	Direct	MOD restorations	Molars
de Kuijper et al., 2020 [76]	TMCL + load to fracture	Chewing simulator - CS-4.8 (SD Mechatronik, Germany)	50	1.7	1.2 × 106	Ceramic sphere	N.I.	8000 cycles; 5–55 °C	Direct	Composite built up	Molars
Heck et al., 2020 [77]	Liquid storage + mechanical cyclic loading	Computer-controlled chewing simulator (MUC 2; Willytec, Germany)	50	1	1 × 106	Ceramic sphere	Double distilled water	N.I.	CAD/CAM	Occlusal veneers	Molars
Pivetta Rippe et al., 2020 [78]	Liquid storage + mechanical cyclic loading (1)+ liquid storage + mechanical cyclic loading (2)+ load to fracture	Mechanical device ER-11000 (Erios, Brazil)(1); Electrodynamic testing machine (ElectroPuls E3000, Instron, USA)(2)	80(1); 400(2)	2(1); 10(2)	1 × 106 (1); 1.5 × 106(2)	Stainless steel cylinder	Distilled water	37 °C	CAD/CAM	MOD inlays	Premolars
Rosentritt et al., 2019 [79]	Liquid storage + TMCL + load to fracture	Chewing simulator - eGo (Germany)	50	1.6	2.4 × 106	Steatite ball	Distilled water	6000 cycles; 5–55 °C	CAD/CAM	Crowns	Molars
Schwendimann and Özcan, 2019 [27]	TMCL + load to fracture	Custom-made chewing simulator - CoCoM (University of Zurich, Switzerland)	50	1.67	1.2 × 106	Steel sphere	Distilled water	5–55 °C	Direct	Veneers	Incisors
Silva et al., 2018 [80]	Humidity storage + Mechanical cyclic loading + load to fracture	Chewing simulator - Biocycle (Biopdi, Brazil)	50	2	1.2 × 106	Stainless steel sphere	Distilled water	37 °C	Direct	Class II restorations	Molars
Soares et al., 2018 [14]	Liquid storage + mechanical cyclic loading + cyclic load to fracture	Closed-loop servohydraulics - Acumen 3 (MTS Systems, USA)	200; 100–2900	5	1.85 × 105; 8.7 × 103	Resin composite sphere	Distilled water	N.I.	Indirect	MOD inlays	Molars
Montagner et al., 2017 [81]	Liquid storage + mechanical cyclic loading	Mechanical device - Rub&Roll (Radboud University, The Netherlands)	30	0.4	7.5 × 105	Metal rod in a silicone tube	Water	Room temperature	Direct	Composite built up	Molars
Rosentritt et al., 2017 [82]	Liquid storage + TMCL + load to fracture	Chewing simulator – eGo (Kältesysteme, Germany)	50	1.6	1.2 × 106	Steatite ball	Distilled water	2 × 3000 cycles; 5–55 °C	CAD/CAM	Crowns	Molars
Sawalt et al., 2017 [83]	Mechanical cyclic loading + liquid storagea + thermal cycling	Chewing simulator - Willytec Kausimulator 3.1.3 (Willitec, Germany)	49	N.I.	1.2 × 106	Ceramic sphere	Distilled water	10,000 cycles; 5–55 °C	Direct	MOD restorations	Premolars
van den Breemer et al., 2017 [84]	TMCL + load to fracture	Chewing simulator - CS-4.8 (SD Mechatronik, Germany)	50	1.7	1.2 × 106	Ceramic	Distilled water	8000 cycles; 5–55 °C	CAD/CAM	MOD inlays	Molars
Montagner et al., 2016 [85]	Liquid storage + mechanical cyclic loading	Mechanical device - Rub&Roll (Radboud University, The Netherlands)	30	0.4	2.5 × 105; 5 × 105; 7.5 × 105	Metal rod in a silicone tube	Distilled water	N.A.	Direct	Composite built up	Molars
Goldberg et al., 2016 [86]	Liquid storage + mechanical cyclic loading + load to fracture	Closed-loop servohydraulic - MiniBionix II (MTS Systems, USA)	200–1400	5	5 × 103; 3 × 104	Steel cylinder	Distilled water	N.A.	CAD/CAM	Crowns	Molars
Barreto et al., 2015 [87]	Thermal cycling + mechanical cyclic loading + load to fracture	Mechanical loading device - LRP 3000 (Erios, Brazil)	50	4	3 × 105	N.I.	Distilled water	6000 cycles; 5, 37 and 55 °C	Direct	MOD restorations	Premolars
Ilgenstein et al., 2015 [88]	TMCL + load to fracture	Computer controlled masticator - CoCoM 2 (PPK, Zwitzerland)	49	1.7	1.2 × 106	Molar	N.I.	3000 cycles; 5–55 °C	CAD/CAM	Onlays	Molars
Magne et al., 2015 [16]	Liquid storage + mechanical cyclic loading	Closed-loop servohydraulics - MiniBionix II (MTS Systems, USA)	200–1400	10	1.85 × 105	Resin composite sphere	Distilled water	N.I.	CAD/CAM	Crowns	Molars
Monaco et al., 2015 [89]	TMCL + load to fracture	N.I.	49	1.7	1.2 × 106	Premolar	N.I.	3000 cycles; 5–55 °C	Indirect	Onlays	Premolars
Rosatto et al., 2015 [90]	Mechanical cyclic loading + load to fracture	Chewing simulator - Biocycle (Biopdi, Brazil)	0–50	2	1.2 × 106	Stainless steel sphere	Water	37 °C	Direct	MOD restorations	Molars
Mehl et al., 2014 [91]	Liquid storage + mehanical cyclic loading + thermal cycling + load to fracture	Chewing simulator - Munich chewing simulator (Willytec, Germany)	50	N.I.	5 × 104	N.I.	N.I.	1660 cycles; 5–55 °C	Indirect	MOD inlays	Premolars
Schwindling et al., 2014 [92]	Thermal cycling + mechanical cyclic loading + load to fracture	Chewing simulator - Willytec CS4 (SD Mechatronik, Germany)	64	N.I.	1.2 × 106	N.I.	Water	10,000 cycles; 6.5–60 °C	Direct	Composite built up	Molars
Rosentritt et al., 2014 [21]	TMCL + load to fracture	N.I.	50	N.I.	1.2 × 106	Steatite sphere	N.I.	5–55 °C	Direct	Veneers	Incisors
Taha et al., 2014 [93]	Humidity storage + mechanical cyclic loading	Servo-hydraulic materials testing machine - MTS model 801 (MTS Systems, USA)	50	3–5	7.5 × 104	Steel cylinder	N.I.	N.I.	Direct	MOD restorations	Premolars
Zamboni et al., 2014 [94]	Liquid storage + mechanical cyclic loading + load to fracture	N.I.	50	1	1 × 105	N.I.	Water	37 °C	Indirect	MOD inlays	Premolars
Batalha-Silva et al., 2013 [17]	Liquid storage + mechanical cyclic loading	Closed-loop servohydraulics - MiniBionix II (MTS Systems, USA)	200–1400	5	1.85 × 105	Resin composite sphere	Distilled water	N.I.	Indirect	MOD inlays	Molars
Salaverry et al., 2013 [95]	Liquid storage + mechanical cyclic loading + load to fracture	Mechanical device ER-11000 (Erios, Brazil)	200	1	5 × 105	N.I.	Distilled water	N.I.	Indirect	MOD inlays	Premolars
Aggarwal et al., 2012 [96]	Mechanical cyclic loading + load to fracture	N.I.	60	5	1.5 × 105	N.I.	N.I.	N.I.	Indirect	Crowns	Premolars
Magne et al., 2012 [18]	Liquid storage + mechanical cyclic loading	Closed-loop servohydraulics - MiniBionix II (MTS Systems, USA)	50–1050	5	1.8 × 105	Resin composite sphere	Distilled water	N.I.	Indirect	MOD inlays	Molars
Batalocco et al., 2012 [97]	Liquid storage + thermal cycling + mechanical cyclic loading + load to fracture	Custom-made chewing simulator (N.I.)	49	4.16	2 × 106	Flat stainless-steel piston	Saline solution	10,000 cycles; 5–55 °C	Direct	Veneers	Incisors
Kassem et al., 2012 [98]	Liquid storage + mechanical cyclic loading	Instron testing machine (Instron, USA)	60–600	12	1 × 106	Steel sphere	Distilled water	Room temperature	CAD/CAM	Crowns	Molars
Krastl et al., 2011 [99]	Liquid storage + TMCL + load to fracture	Computer controlled masticator - CoCoM 2 (PPK, Zwitzerland)	49	1.7	1.2 × 106	Human cusps	N.I.	3000 cycles; 5–50.5 °C	Indirect	Crowns	Premolars
Laegreid et al., 2011 [100]	Thermal cycling + mechanical cyclic loading + load to fracture	N.I.	300	1	1 × 104	N.I.	Water-based tracer material	5000 cycles; 5–55 °C	Direct	Class II restorations	Molars
Schlichting et al., 2011 [19]	Liquid storage + mechanical cyclic loading + load to fracture	Closed-loop servohydraulic -MiniBionix II (MTS Systems, USA)	200–1400	5	5 × 103; 3 × 104	Resin composite sphere	Distilled water	N.A.	CAD/CAM	Veneers	Molars
Heintze and Cavalleri, 2010 [101]	Liquid storage + thermal cycling + mechanical cyclic loading + TMCL	Chewing simulator with pneumatic force actuators (N.I.)	100	1	6.4 × 105; 1.2 × 106	Lithium disilicate conical stylus	N.I.	5–55 °C	Direct	Class V restorations	Premolars
Magne et al., 2010 [20]	Liquid storage + mechanical cyclic loading	Closed-loop servohydraulic - MiniBionix II (MTS Systems, USA)	200–1400	5	5 × 103; 3 × 104	Resin composite sphere	Distilled water	N.A.	CAD/CAM	Occlusal veneers	Molars
Attia et al., 2006 [102]	Liquid storage + TMCL + load to fracture	Chewing simulator - Willytec Kausimulator 3.1.3 (Willitec, Germany)	49	1.2	6 × 105	Steatite ball	N.I.	3500 cycles; 4–58 °C	CAD/CAM	Crowns	Premolars
Heintze and Cavalleri, 2006 [23]	Liquid storage + TMCL	Chewing simulator (N.I.)	50	1.6	1.2 × 106	Lithium disilicate conical stylus	N.I.	100,000 cycles; 5–55 °C	Direct	Class V restorations	Premolars
Stappert et al., 2006 [22]	TMCL + Load to fracture	Chewing simulator - Willytec CS4 (SD Mechatronik, Germany)	49	1.6	1.2 × 106	Steatite ball	N.I.	5300 cycles; 5–55 °C	Indirect	MOD inlays	Molars
Kuijs et al., 2006 [66]	Mechanical cyclic loading + load to fracture	Servo-hydraulic materials testing machine (MTS Inc., USA)	200–1000	5	1 × 104; 5 × 104	Cylindrical stainless-steel bar	Water	37 °C	Direct and Indirect	Class II restorations	Premolars
Lehmann et al., 2004 [103]	Thermal cycling + mechanical cyclic loading + load to fracture	N.I.	50	1.66	1 × 104	Spherical indenter	N.I.	10,000 cycles; 5–55 °C	Indirect	Crowns	Molars
Behr et al., 2003 [104]	Liquid storage + TMCL + load to fracture	Thermo-mechanical cycling device (N.I.)	50	1.66	1.2 × 106	N.I.	N.I.	6000 cycles; 5–55 °°C	Indirect	Crowns	Molars
Krejci et al., 1994 [105]	Liquid storage + TMCL + load to fracture	Chewing simulator (N.I.)	49	1.7	1.2 × 106	N.I.	N.I.	3000 cycles; 5–55 °C	Indirect	Crowns	Premolars

TMCL: Thermo-mechanical cyclic loading = both aging methods applied simultaneously. N.I.: Not informed; N.A.: Not applicable; MOD: mesial-occlusal-distal; MO: mesial-occlusal; DO: distal-occlusal.

^aLiquid storage every 2 × 10⁶ cycles.

^bNumber of thermal cycles alone.

As shown in Fig. 2, a large variability in test methods was reported across the studies, resulting in 16 different combinations of laboratory methods used to age the specimens. The majority of studies used an initial step of static liquid storage before subjecting the restorations to the mechanical challenge (59.2%). Thermo-mechanical cycling, i.e., both thermal and mechanical cycles applied simultaneously was reported in 38.8% of the studies, whereas 20.4% used thermal cycling as a discrete aging step. A load to fracture test was applied in 69.4% of the articles. In addition, the choice of whether or not to simulate the periodontal ligament, and how this simulation was accomplished varied in the studies included.

Fig. 2.

Combinations of methods used to age resin composite restorations in the studies included. In total, 16 different combinations of methods were reported across 49 articles. Total number of studies using each combination is presented.

3.2. Synthesis of results

This review identified studies that attempted to simulate oral conditions to produce mechanical degradation and fatigue of resin composite restorations. Significant heterogeneity across methods, devices, substrates, and test conditions was observed. Of the 49 studies included, consistent testing methods and protocols were found mainly in studies from a same research group; heterogeneity was sometimes present even when studies from a same group were compared. In addition to issues regarding the inconsistency in methods, reporting problems also were observed in this sample of articles. Relevant items of information that could be important for reproducibility were sometimes missing, including basic aspects such as the load and frequency of mechanical cycling. A positive finding was that the studies endeavored to simulate other chemical and/or physical factors that could occur in the mouth in spite of their main outcome being to produce mechanical degradation of the restorations. In this sample, only one study mentioned that achieving clinical relevance was one of their objectives [23]. The majority of the studies included also aimed to correlate the parameters adopted in vitro with those of the in vivo scenario, as observed in their discussion chapters. The overall evidence indicated reporting problems, and suggested a lack of clinical validation of contemporary research methods used to simulate the mechanical degradation and fatigue of resin composite restorations, which may hinder extrapolation of findings and could limit comparisons among composite restorations tested under different laboratory conditions. The main research methods used across the studies are discussed as follows.

4. Discussion

4.1. Mechanical cyclic loading

This aging method refers to the application of repeated and continuous mechanical forces with the overall aim of simulating the challenges imposed by chewing, bruxing, and/or clenching habits on teeth and restorative materials in vivo. The method aims to generate fluctuating stresses within the restorative material and adjacent tooth, leading to fatigue degradation of the restored tooth structure, which may eventually fail, as a result of an accumulation of minor and major damages including subcritical crack generation, crack propagation, surface irregularities, wear, loss of anatomical shape, marginal breakdown, and fractures. Cyclic loading can be exerted by different antagonists (actuators, plungers, styli, or other structures) with different shapes and sizes, which are generally brought into contact with the restoration (specimen) for load application, i.e., usually without generation of impact. Previous studies have indicated that the antagonist material, its shape, size, and number of contact points with the restoration are factors that might influence the mechanical challenge and damage promoted by cyclic loading [24-26].

Depending on the configuration of the specimens and antagonists used, cyclic loading could generate different types of stresses on the restored structures, including compressive, tensile, and shear stresses. The purpose of the aging method is not to cause immediate fractures, but to show that these could happen over time, as the restorative material and surrounding tooth become more brittle. The cracks propagated may eventually coalesce and favor the occurrence of chipping or catastrophic fractures. The specimen is usually in a state of rest but, in some cases, the specimen or the antagonist can be moved back-and-forth to promote sliding. In the present sample, the variability in load application methods included a tripod contact (mesiobuccal, distobuccal, and lingual cusps) with the antagonist [14,16,17,19,20], load applied from the lingual surface at angulations varying between 30 and 105 degrees to the antagonist [15,18,27], and combination of loads applied centrally in the occlusal fossa then eccentrically on the lingual cusps [23], among others. This remarkable variability could hinder any metasearch analysis, for instance. The loading regimens also varied among methods and studies, including the forces and number of mechanical cycles. In several studies, the load magnitude varied during cycling to simulate conditions closer to masticatory function, which oscillates during jaw movements [28,29]. The present study also showed that 1.2 M was the number of mechanical cycles most often used, but there was no consensus in the literature on what this number of cycles may represent in terms of how many months or years they simulated in the clinical service time. It seems that there is room for research on that aspect since it could benefit the methods available for laboratory simulation of mouth conditions and improve homogeneity among studies.

The microstructure of polymeric restoratives plays a major role in their response to strain [30-32], including porosity, defects, and contaminants left within the material. This is particularly relevant for direct restorative procedures. As the operator prepares the restorations by using different resin composite layers, air bubbles and other imperfections left in between these layers could compromise the fatigue resistance of restorations by acting as areas of stress concentration or magnification [33,34]. Appropriate curing is another crucial factor for composites as it affects their degree of C═C conversion and development of mechanical properties [35-38]. This factor means that the photoactivation or other curing procedures are important steps in the methodology and should be reported in full detail. Since the objective is to mimic the clinical situations, perhaps the adhesive procedures, preparation of restorations, and photoactivation procedures also should be planned to simulate these clinical conditions. The bonding between the resin composite and adjacent tooth structure could affect the restorative material ability not to concentrate all stresses but partially to dissipate them to the surrounding structure [1,32,39]. Researchers are urged to clearly report how the tooth surfaces were prepared and bonding procedures were carried out, including details on how direct and indirect restorations were prepared and the surfaces were treated. Finishing and polishing procedures also should be reported thoroughly for their potential to generate rougher or smoother resin composite surfaces [40,41] and affect the surface damage and wear caused by the antagonist.

4.2. Static liquid storage before mechanical testing

Static liquid storage is a method to reproduce the humidity of the intraoral environment and is widely used in dental materials research as an aging method. In the included studies, the storage media used were distilled water, water, and saline, at both room and 37 °C temperatures. In the first few days after a resin composite restoration is prepared, unreacted, leachable components such as free monomers and polymerization promoters can be eluted into the storage medium [42-44]. Static liquid storage could also help to induce hydrolytic degradation of the polymer networks and influence their later response to cyclic loading. In the present review, however, the storage times usually ranged from 24 h to 7 days before testing. Absorption and solubility could be precursors to permanent deterioration processes, but perhaps the storage times should be longer for plasticization, oxidation, and hydrolytic effects to take place within the bulk of the restoratives [45].

An alternative to accelerate the polymeric degradation process during storage could be the use of enzymes present in saliva [46,47]. The choice of liquid media also could be a factor to be considered. Reports have indicated that composites stored in artificial saliva eluted more filler particles than those stored in distilled water [48,49]. The pH also may vary among the different storage media, but pH values were generally not reported in the studies included. Albeit the oral environment is kept near neutrality, resin composites are exposed to multiple compounds from foods and beverages, mainly acidic, and saliva also undergoes variations in pH [50]. A meta-analytical study on the release of components from resin-based materials [51] reported that no significant correlation could be observed between the pH of the medium and the release of filler components from composites. In cases of longer storage periods, degradation of the adhesive bonds between the restoration and tooth structure could also occur and affect their mechanical performance. For a better understanding of the influence of the static liquid storage on the response of composite restorations to mechanical challenges, researchers are encouraged to clearly report the following details: type and volume of medium used, immersion time, temperature, pH, and whether the medium was renewed or not during the storage period.

4.3. Thermal cycling during or before mechanical testing

In several cases, mechanical cyclic loading was applied associated with a simultaneous thermal cycling or the thermal cycling was a discrete aging step used before loading. Thermal cycling is a method that endeavors to challenge and induce damage to the restorations by generating expansion and contraction stresses arising from the alternate immersion in liquid media at low and high temperatures. The constant temperature changes could affect the mechanical performance of restorations by straining both the composite material and restored structures, as well as their bonded interface. Within the composite structure, the different components (i.e., polymer matrix and filler particles) have different coefficients of thermal expansion and therefore may expand and contract dissimilarly during thermal changes, straining the bonds between the particles and polymer, as well as those between the restorative material and tooth structure. Perhaps the simultaneous combination of mechanical and thermal cycling (thermo-mechanical cycling) could induce faster mechanical degradation and fatigue of resin composite restorations, but this evidence was not available from the studies included in this review.

Previous authors have questioned the absence of concrete evidence that failures in clinical practice may occur because of thermal stresses [52], and have also addressed the uncertainty about whether failure could occur due to flow in one or other of the layers in the bonded structure. The same authors suggested that immersion during thermal cycling could facilitate the breakdown by hydrolysis of the adhesive bond between the composite and tooth tissue. This reaction could be assisted by stress and a potential mechanism for fatigue failure, but time at temperature would be a relevant condition to be further analyzed. In Section 4.1, we mentioned that the number of mechanical cycles had no standardized protocols across the included studies, which also seemed to be the case for thermal cycling. A review article on thermal cycling in dental materials research commented that the choice of cycling regimens, including temperature, number of cycles, and dwell times is widely dissimilar in the literature and seemed to be selected by convenience [53]. In a subsequent study, the same authors [54] concluded that thermal cycling was able to affect the flexural properties of resin composites but at least 30 K cycles were needed depending on the material tested.

4.4. Load to fracture test and different loading stages

Several studies used load to fracture tests in which load was applied until fracture occurred with a general aim of providing a quantitative measure of the resistance or strength of the aged restored structure, at least in the specimens that survived the mechanical cycling. The method, also known as crunch-the-crown test, applies a uniaxial compressive force at a constant strain rate to the specimen until catastrophic fracture occurs. However, in the mouth, restorations are subject to a wide range of low intensity loads and tend to degrade or fail due to accumulated damage from cyclic loading rather than a single-cycle loading event [12,13]. For this reason, load to fracture tests are not considered good predictors of clinical failure of resin composite restorations [12,34]. Fractures are reported as one of the main reasons for failure in clinical studies of resin composites [55-59], but the load magnitude until fracture occurs in this type of test can be extreme and the concentration of stress so high that the failure types may not resemble those observed in the clinical setup due to the high contact stresses. A report stated that most contact stresses in the test may exceed between 25 and 125 times the contact stresses at tooth wear facets in clinical contexts [60].

A load to fracture test may, however, be useful in some situations, for example when later applying an accelerated step-stress fatigue test [61,62]. In this type of test, incremental stress levels are applied, and load and number of cycles to failure are registered. Before starting the actual test, Shembish et al. [62] determined the step-stress profiles (mild, moderate, and aggressive) based on the mean load obtained by a load to fracture test. The mild profile typically begins from a stress level around 30% of the mean and the aggressive from 60% value, for instance [63], but the validity of these percentages still have to be determined. In step-stress tests, the number of cycles decreases as the load increases with the aim of distributing failures across different loads [63]. The aggressive profiles in step-stress tests aim to simulate extreme clinical situations such as grinding, clenching, or intrinsic masticatory accidents [64,65]. Several articles included in this study employed different loading stages in a same experiment [15-17,19,20,66], usually starting with a so-called preconditioning phase, i.e. a lower force applied initially followed by subsequent stages of increasing forces, a method sometimes called staircase loading [18] or accelerated test [14].

4.5. Limitations and future challenges

Fatigue properties could be useful predictors of the clinical performance of resin composite restorations, but attention must be paid to the clinical relevance of the laboratory models used. In Table 3, we summarize factors that should be considered in future laboratory studies on the mechanical degradation of resin composite restorations, including aspects to improve reporting practices and the reproducibility of experiments, in addition to opportunities for research on the topic. Clinical reports have indicated a difference in the failure behavior between anterior and posterior restorations [59,67]. However, the majority of the studies included herein evaluated teeth in the posterior region, which might be related to the configuration of the simulation devices and the higher loads observed in the posterior dentition, which increases the risk for restoration failures. This observation highlights the need for more in vitro models for evaluating anterior teeth. The great advantage of laboratory studies when compared with clinical evaluations is that they can be standardized, validated, and reproduced without variables related to patients. However, since validation may need homogeneity and convergence of methods, the high heterogeneity and reporting problems observed in this sample of articles seem to be challenges to be addressed in future research on the topic. The low correspondence across studies may hinder metasearch analysis; the reporting problems raised issues about reproducibility. In addition, applicable clinical data needs to be available as a reference for translational research approaches.

Table 3.

Factors to be considered in future studies and opportunities for research on the simulation of the mechanical degradation of resin composite restorations.

Items for improved reporting practices and reproducibility
Restorative procedures:
Detailed direct, semi-direct, or indirect techniques used to prepare the restorations Complete information on adhesive procedures Specifics for all surface treatments and photoactivation methods Details on procedures for finishing and polishing of restorations
Mechanical loading method:
Load (force), frequency, and method of load application Antagonist material, size, and shape Contacts (number, area) and inclination between the antagonist and restoration Test medium, pH, and temperature
Other associated aging approaches:
Storage time before and/or during mehanical loading Type, volume, pH, and temperature of storage medium Frequency of renewal of the storage medium Number of other cycles (e.g. thermal cycles) and dwell times
Opportunities for future research on the topic
Mechanical degradation in anterior teeth restored with resin composites Comparability across different loading devices Correspondence between the number of mechanical cycles and clinical service time Clinical validation of in vitro methods for simulating the mechanical degradation of restorations Translational approaches including laboratory and clinical data

5. Conclusions

The literature contains several in vitro publications on different methods used to simulate the mechanical degradation and fatigue of resin composite restorations, with a high level of heterogeneity in the methods, devices, specimens, and test parameters adopted. The large variability across the studies raises questions about between-laboratory validity of methods and the overall comparability of findings among studies, and may be an issue for translating the laboratory results to the clinics. Validation studies, underlying clinical supporting data, and better reporting practices are needed.