The Effect of Sandblasting on Bond Strength of Soft Liners to Denture Base Resins: A Systematic Review and Meta-Analysis of In Vitro Studies

Objectives This study aimed to evaluate the effect of sandblasting on the bond strength of denture base resin to soft liners. Materials and Methods This report follows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement. PubMed, Embase, Cochrane, Scopus, and OpenGrey databases were searched for in vitro studies that compared sandblasting with no treatment in terms of the tensile, shear, and peel bond strength of resilient lining materials (acrylic-based or silicone-based) to polymethyl methacrylate denture base resin. Based on the outcome, the analysis was carried out in three groups of tensile, shear, and peel bond strength. Subgroup analysis was done for the effect of size of particles on sandblasting, blasting pressure, and type of soft liner whenever possible. Heterogeneity was evaluated among the studies, and meta-analysis was performed with random effect models (p < .05). Results After screening, 16 articles met the inclusion criteria for meta-analyses. No treatment showed significantly higher tensile (p < 0.001) or peel (p=0.04) bond strength, although shear bond strength of sandblasted resin was significantly better (p=0.008). Results of subgroup analyses of particle size favored the control group in 50 µ Al2O3 particle size (p < 0.001). In analyses of blasting pressure, the control group had significantly better tensile bond strength than specimens with blasting pressure ≤1 bar (p < 0.001) while specimens with blasting pressure beyond 1 bar showed significantly more tensile strength than control group (p=0.03). In silicon-based liners, groups without any surface treatment had significantly higher tensile bond strength (p < 0.001). Conclusion According to the in vitro studies, sandblasting would not lead to significant increase in bond strength of soft liner to the denture base resin.


Introduction
Prolonged use of dentures is common among elderly patients. It could cause denture soreness and serve bone resorption [1]. Resilient lining materials are used to distribute the pressure equally and prevent localization of force by a cushion effect under the denture bases [2][3][4][5][6][7][8][9][10][11]. Relining materials offer dentists a quick, convenient, and short time solution for patient problems. Indications of resilient lining materials are seen in patients with exostosis due to uneven bone resorption, tender soft tissues, bony undercuts, immediate dentures, treatment dentures after implantation or healing period, presence of parafunctional habits, xerostomia, ill fitted dentures, wearing facial prostheses, and demand for better rhythm of chewing strokes. ey also compensate for the volumetric shrinkage of acrylic resin [6,[12][13][14][15]. ese materials can be provisional or permanent, and auto-or heat-cure-polymerized [16][17][18]. Five types of soft liners exist according to their chemical structures, namely, plasticized acrylic resins (chemical or heat-polymerized), vinyl resins, polyurethane, polyphosphazene, and silicone rubbers (heat-cured or room-temperature-vulcanized) [8]. All types of resilient liner materials have some drawbacks such as insufficient color stability, losing resiliency over time, poor abrasion resistance, presence of surface defects and porosity, water uptake, microbial gathering, bond failures to denture base resin, unsatisfied taste over the time, mephitis, difficulty in cleaning, and premature hardening due to plasticizers solubilization [12,[19][20][21][22][23][24].
Two-layer dentures can only be successful when there is strong adhesion between different layers of materials [25] Tensile bond strength with a minimum of 0.44 MPa (4.5 kg/ cm 2 ) between acrylic resin and liner is needed to be acceptable for clinical usage [26][27][28]. To overcome the weak bond strength between denture base resin (DBR) and liners, sandblasting with alumina, laser application, chemical cauterization or primers, acrylic drills, or mesh textured glass fibers have been used by researchers [2,6,9,11,[29][30][31][32][33].
e investigators tried to roughen DBR surface with airborne particles before adding the liners to improve the bond strength [26,31,[34][35][36][37]. Controversy exists regarding efficiency of sandblasting in improving bond strength. While some investigations have shown improved bond strength, [30,32] others have reported that mechanical surface treatment of DBR decreases the adhesion bond strength [21,25,29,31]. Meanwhile, existing reviews evaluated sandblasting without considering the role of sandblasting parameters in the final outcome. e aim of this systematic review was to analyze the effect of sandblasting on bond strength of resilient lining materials applied to DBR considering size of particles in sandblasting, blasting pressure, and type of soft liner.

Materials and Methods
is systematic review was reported according to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) statement [38,39]. e study design focuses on answering the following question, according to PICO strategy: "what is the effect of sandblasting on bond strength of resilient liners to polymethyl methacrylate (PMMA)?" In this process, the population was denture base resins that were bonded to soft liners; the intervention was sandblasting the substrate; the comparison was made with groups without any surface treatment; and the outcomes were tensile bond strength, shear bond strength, or peel bond strength between acrylic denture base and soft liner. e secondary aim of the study was to answer two further questions: "What is the effect of Al 2 O 3 particle size and blasting pressure on the bond strength between denture base resin and resilient liner materials?" "How could the type of soft liners affect the bond of sandblasted groups and control groups?" e review question, aims of the study, suitability criteria, search strategy, and data analysis were specified in the beginning with clarity and were included in the study content. A systematic literature search was done in the PubMed, Embase, Cochrane, Scopus, and OpenGrey (https://www.opengrey.eu) databases until January 2020 (Table 1). Furthermore, the reference and citations' lists of the selected articles were reviewed for selecting potential inclusions.
Eligible studies were experimental, in vitro, and laboratory studies which evaluated the bond strength of resilient lining materials (acrylic-based or silicone-based) to polymethyl methacrylate (PMMA) denture base resin and compared no treatment (control) with sandblasting surface pretreatment (experimental) in the same study. In addition, the study should report the mean and standard deviation (SD) of tensile, shear, or peel bond strength. Studies that evaluated other materials for denture base except PPMA, critiques, case reports, systematic reviews, and expert opinion papers were excluded. 38 studies that did not provide essential data after contacting the authors via e-mail were also excluded. Moreover, included studies had to be published in English.
Title, abstract, and full text selection were carried out by two authors (F. H. and M. A.) independently. Finally selected full text studies, based on inclusion and exclusion criteria, were those with proper control group having no surface treatment and with experimental group in which no further treatment, such as applying adhesive, was performed after sandblasting. Disagreements on selection process were resolved by a third investigator (M. H.), and finally consensus was reached through discussion. Two investigators extracted study content and data independently using a standard form prepared in software (Office Excel 2013 software, Microsoft Corporation, Redmond, WA, USA). e following data were extracted: sample size, name of acrylic, liner material, particle size of sandblasting, pressure of sandblasting, time of sandblasting, distance from sandblasting tip to specimen, storage condition of specimen before testing, thermocycling, mean and SD of experimental and control group, and failure mode. Any disagreements between investigators were resolved by discussion. In studies where enough information was not provided, the authors were contacted via e-mail.
Two authors (F. H. and M. H.) evaluated the methodological quality of each enrolled study independently bases on reported tools of previous systematic reviews of in vitro studies. [40,41] erefore, the following parameters were checked for risk of bias evaluation: specimen randomization, single operator protocol implementation, blinding of the testing machine operator, presence of a control group, standardization of sample preparation, failure mode evaluation, use of materials according to the manufacturer's recommendation, description of sample size calculation, and discarded unacceptable samples. If the article reported the parameter, it received "yes" for that parameter. If information is not provided or the article does not follow the parameters, it received "not mentioned" or "no," respectively. Articles with one to three reported items were considered as high risk of bias, four to five as medium risk of bias, and six to nine as low risk of bias.
For meta-analysis, the outcomes were categorized into three groups of tensile, shear, and peel bond strength. Sandblasted and control groups were analyzed in each category both globally and by subgroups. e effect of size of particles on sandblasting, blasting pressure, and type of soft liner was analyzed in subgroups in categories with sufficient data. Studies with several independent experimental and control groups were assumed as independent comparisons in meta-analysis. For studies with multiple correlated comparisons (control group in common), groups were combined with specific formula for mean and SD to create a single pairwise comparison in order to overcome a unit-ofanalysis error.
Meta-analysis was based on inverse-variance method. As MPa was accepted unit for reporting bond strength values, values of different units were converted to MPa. Bond strength was the continuous outcome evaluated for mean difference (MD) and the corresponding confidence interval. A p value ≤0.05 was considered statistically significant in Z test. Heterogeneity among studies was calculated using I 2 and chi 2 tests. All analyses were done using random effect model in Review Manager software (version 5.1, Cochrane Collaboration, Copenhagen, Denmark).

Results
e process of screening the articles is summarized in Figure 1 according to PRISMA statement. 106 articles were identified from databases after reading the titles, of these 53 were eligible for full text evaluation. Finally, 37 studies were excluded for the reasons presented in Table 2, and 16 articles were enrolled for meta-analysis. Study characteristics and descriptive evaluation of studies are presented in Table 3.
e effect of type of soft liner was investigated with strata of silicon-based liner and acrylic resin-based liner. As study groups were not sufficient in shear and peel bond strength categories, this subgroup analysis was only conducted for tensile bond strength. e results showed that groups without any surface treatment had significantly higher tensile bond strength when silicon-based liner was used (p < 0.001). Meanwhile, the two studies that used acrylic resin-based liners showed no significant difference between control and sandblasting groups ( Figure 9).
Results of quality assessment showed that six studies had medium risk of bias and ten studies had low risk of bias (Table 4). e most not reported items were "single operator protocol implementation" and "blinding of the testing machine operator."

Discussion
Different methods have been introduced to improve bond of denture base resins to soft liners. e influence of these methods has been evaluated in two systematic reviews. [84,85] e enhancement mechanisms can be divided into three general categories: first, increasing the available surface area for bonding by increasing surface roughness; second, improving the chemical behavior of substrate to improve wettability; and finally establishing hydrogen bond between acrylic group of PMMA and adhesive primers. Treating the surface by laser, sandblasting, and chemical solvent influences the bond strength through increasing surface roughness.
e surface of material that is candidate for bonding can be sandblasted by spraying a stream of Al 2 O 3 particles under high pressure. [86] Global results from two systematic reviews showed that airborne particle abrasion decreases the bond strength between denture base resin and soft liners. [84,85]       International Journal of Dentistry sandblasting. [29,30,73,75,76] Several different parameters and strategies are used for sandblasting, and this could obscure getting the real impact of this procedure on the bond strength. Factors that could affect the bond strength values between the liner materials and denture base resin are the type of lining materials, particle size of sands, blasting pressure and time, test methods, thermocycling, speed of head of testing machine, and thickness of lining material. is review and meta-analysis tried to consider variables in sandblasting including particle size, blasting pressure, and type of liner to identify the effect of this pretreatment in improving bond strength. [11,15,25,31,33,36,78,84,85,[87][88][89][90][91][92].
Quality of resilient lining materials is evaluated by their tensile properties. e bond strength between denture base resin and resilient lining materials is usually assessed by tensile test due to reliable results and also easiness of performance. [35,85,93] Results of our meta-analysis showed that in general sandblasting could not improve tensile bond strength significantly. Increasing the bond strength after sandblasting is expected as it provides more bonding surface and creates mechanical locks at bond site, also removing contaminants. [29] It results in irregularities, valleys, depressions, many small pits, and scratches in acrylic resin treated surface. [94,95] SEM investigation also shows that sandblasted surfaces are rougher and have no debris. [30]  Test for subgroup differences: Chi 2 = 58.99, df = 1 (P < 0.00001), I 2 = 98.3% Heterogeneity: Tau 2 = 3.05; Chi 2 = 74.09, df = 3 (P < 0.00001); I 2 = 96% Test for overall effect: Z = 2.04 (P = 0.04) Figure 7: Forest plot for subgroup analysis of particle size for peel bond strength.  Soft lining material could flow into the irregularities of the acrylic resin that resulted in significant effect on adhesive values. [31] However, the size of the irregularities may not be adequate to allow the resilient lining material to penetrate into them without leading to a significant increase in tensile bond strength. [30] As flowing into resin irregularities by the liners is dependent on their viscosity, the liquidity of the elastic materials in a clarified contact angle and surface energy define the penetration. [26,31] e penetration coefficient (PC) for liquids into a cavity is given by PC � c cos θ/2η, where c is the surface tension, θ is the contact angle, and η is the viscosity. is can state the lower tensile strengths of sandblasted specimens subjected to the reviewed studies. On the other hand, creation of microcracks, vacancies, and voids during packing the resilient lining material on resin surface may trap air bubbles, compensate for the effect of irregularities for increasing the contact surface, and result in reduced bond strength. [35] e other explanation for strength reduction is the stress induced at the junction of PMMA and soft liner, or stress concentration because of discontinuities on the surface. [46] Another hypothesis for reduced bond strength is separated resin or Al 2 O 3 particles, which remain in the irregularities of the treated surface and will decrease the bond strength. [32] e rate of Al 2 O 3 adhesion may be varied in the used denture materials.
e results of meta-analysis of two studies showed that sandblasting do not increase peel bond strength; though shear bond strength increased significantly. Al-Athel et al. designed an investigation of the effect of test methods on bond strengths of the liners. [88] ey demonstrated that roughening the surface increased shear bond strength while tensile bond strength decreased. [88] Such finding could be explained by the fact that, in roughened surface, more force is needed to move two surfaces along each other as friction is increased. [88] It should be noticed that the distance between the two surfaces and where the force is applied are the most important factors that could affect shear test values. [13] As debonding begins at the edge of the lining materials, the most similar test to intraoral situations for bonded two-layer dentures is peel test. [7,60,96] is test directly measures the debonding force, and the site of applying force is closely similar to the real situation in the mouth. [7,31,97] However, it is not possible to catch it at the liner acrylic resin interface directly because the possibility of soft liners tearing is high in peel test. [96] erefore, thickness of liner seems to be critical as cohesive failure is higher. [89,98,99] Moreover, surface energy is different in roughened surface and smoothed one. [100] Pretreatment of denture surface affects its geometry that results in alteration of surface energy. [7,31] e amount of force recommended in peeling test is related to surface energy of the used materials, so it should be mentioned accurately.
In this study, the reviewed studies were categorized into three groups with strata of small particle size (50 µ Al 2 O 3 ), medium particle size (60 µ Al 2 O 3 < particle size <125 µ Al 2 O 3 ), and large particle size (particle size ≥250 µ Al 2 O 3 ). Studies with small particle size of Al 2 O 3 (less than 50 µm) showed adverse effect on bond strength of the acrylic resin denture base to resilient material. e result of blasting with particle size in the range of 60 µm to 125 µm showed increasing bond strength; though the difference was not   significant. By increasing the particle size to 250 µm, the results again favored not sandblasted groups. e rationale for these findings is that the size of roughening by sandblasting with 50 µm Al 2 O 3 particles may not be sufficient to allow liner material penetration. [31,62,76] As the penetration coefficient of the liners is inversely related to their viscosity, liners with higher viscosity have less penetration into PMMA surface pores. [35] On the other hand, sandblasting with large size particles (250 µm) also reduces the bond strength due to stress concentration of large size particles. Akin et al. suggested sandblasting with particle size of 120 µm in comparison to 50, 60, and 250 µm for maximum bonding. [11]. e second subgroup analysis was performed to investigate the effect of blasting pressure. Pair comparison groups were categorized into two strata based on blasting pressure (0.2 bar ≤ blasting pressure ≤1 bar; 1 bar < blasting pressure ≤ 4 bar). Meta-analysis showed less tensile bond strength for blasted specimens when the blasting pressure was ≤ 1 bar (p < 0.001). By increasing the blasting pressure to more than 1 bar, sandblasting became significantly more effective than control group (p � 0.03). is finding can be explained by more irregularities caused by high pressure of sand steam.
Surface treatment should be selected according to the type of the resilient lining material to achieve acceptable bond strength. [85] Among the included studies, nine evaluated silicone-based soft liners, three used acrylic-based resilient liners, and four evaluated both types of liners. e results showed that groups without any surface treatment had significantly higher tensile bond strength when siliconbased liner was used (p < 0.001). Meanwhile, the two studies that used acrylic resin-based liners showed contrary results. Khakbaz et al. showed improved bond strength of acrylic soft liner after sandblasting while Kulkarani et al. indicated higher strength in not blasted group. Overall, there is still controversy about the superiority of silicon and acrylic soft liners. Several articles claimed that the similarity of acrylic resin-based liners to denture bases caused higher bond strength values in comparison with silicon-based lining materials. [75,78,83] As methyl methacrylate and ethyl methacrylate are monomers that are basically similar, they can mix through polymerization procedure resulting in a copolymer. Silicone liners do not have any chemical bonding to acrylic denture bases because of their structural differences. [31,78] On the other hand, some studies demonstrated that heat-polymerized silicone-based resilient lining materials had better bond strength than soft liners that contained plasticizer. ese heat-polymerized liners had the greatest bond strengths to acrylic resin denture bases, and the autopolymerized silicone liners had insufficient bonding to acrylic base. [25,[87][88][89][90][91][92] e most important justification in these articles for superior bond strength of silicon liner in comparison to acrylic-based liner was related to minimal water absorption of silicon-based soft liners. [101]. e high level of heterogeneity in analyses indicates great variation of methodology as well as various influencing factors in the main outcome. ese factors include type of liner, size of particle, pressure of blasting, speed of head of testing machine, time of blasting, distance from blasting tip to the specimen, storage condition before testing, and thermocycling. e first three items are discussed in this study with quantitative analyses, and the other five items are presented descriptively. A straight correlation between the tensile strength values and the speed of head of testing machine is reported. [88] It has been shown that the amount of tensile strength between acrylic base and resilient lining material increased significantly up to 40 mm/min speed of machine head, and after that it had reverse effect. [88] Out of our included studies, nine tested the specimens with universal testing machine at a cross head speed of 5 mm/min, and two used cross head speed of 10 mm/min. [9,11,33,[75][76][77][78][79][80] e time of blasting in included studies varied between 10 and 60 seconds, which could be an influencing factor. It has been shown that sandblasting at different distances and angles contributes differences in surface roughness when it is applied to zirconia or titanium materials [102]. However, no study identified the effect of this parameter on the roughness of acrylic resin. ermocycling also affects the values of bond strength. When resilient liner is immersed in water, it will absorb water and saliva, and the plasticizer and solvent agent will leach out of the liner. e balance between these two mechanisms determines the dimensional stability of the material and bond strength. [16] Two studies evaluated the effect of thermocycling and reported that tensile bond strengths were significantly lower than those in the same sets before thermocycling. [76,77] ermocycling could also change the mode of failure to adhesive failure. Nakhaei et al. reported mixed failure for group without thermocycling and adhesive failure for specimen thermocycled between 5 and 55°C for 5,000 cycles. [76].
Taking all of these factors into account, it can be concluded that these factors could affect the final outcome, and more in vitro studies with uniform parameters of testing are encouraged to limit the conflicting factors. e authors could not find any clinical studies that compared the effect of sandblasting on the longevity of bond between denture base and liners, and one of the limitation of this study is that the results are based on in vitro studies. Further clinical studies are needed to indicate the long-term effect of sandblasting as a pretreatment surface preparation.

Conclusion
Within the limitations of this study, these points can be emphasized: (1) Sandblasting decreases the tensile and peel bond strength of resilient lining materials to denture base resins. However, it improves the shear bond strength.
(2) In 50 µ Al 2 O 3 particle size, the amount of bond strength of control group is higher than that of experimental group. However, as the particle size goes beyond 50 µ, no significant difference exists between the two groups. In particle size subgroup analysis of shear bond strength, sandblasting with 50 µ Al 2 O 3 resulted in significantly higher shear bond strength. Groups which were sandblasted with 250 µ Al 2 O 3 had no significant difference with no treatment specimen. (3) Meta-analysis showed higher tensile bond strength for control group when the blasting pressure was ≤1 bar. By increasing the blasting pressure beyond 1 bar, sandblasting became significantly more effective than control group. (4) Groups without any surface treatment had significantly higher tensile bond strength when siliconbased liner was used, while the two studies that used acrylic resin-based liners showed no significant difference between control and sandblasting groups.
Data Availability e data that support the findings of this study are available from the corresponding author upon reasonable request.

Conflicts of Interest
e authors declare that there are no conflicts of interest regarding the publication of this paper.