Airborne-particle Abrasion and Dentin Bonding: Systematic Review and Meta-analysis
The literature reviewed suggests that airborne particle abrasion has no negative effects on the bond strength of resin-based materials to dentin and that a positive influence on dentin bond strength was only achieved in specific air-abrasion conditions. In this systematic review the authors investigated how airborne-particle abrasion (APA) using aluminum oxide affects the bond strength of resin-based materials to dentin. The search was performed in three databases. In vitro studies (Type of study) comparing the bond strength of resin-based materials (Outcome) to air-abraded (Intervention) compared with non–air-abraded (Comparison) human dentin (Population) were included (the PICOT elements are given parenthetically). From 5437 unique articles, 65 were read in full, 33 were included in the qualitative synthesis, and 32 were included in the meta-analysis. Methodologic quality and risk of bias were assessed. Comparisons were performed between air-abraded and control dentin groups by adopting a random-effects model (α=0.05). Additional analyses were carried out for the different parameters used in APA: type of surface treatment in the control group, particle size, air pressure, and APA duration. The bond strength to air-abraded dentin was favored only when the control surface was treated with a hand excavator. For particle size, APA was favored when the particle size was >30 μm and the controls were no treatment or hand excavator or when the particle size was ≤30 μm and the control was bur. In addition, the results favored air-abraded groups only when the pressure was > 5 bar and bur was used in the control group. No significant differences were observed for duration of APA. No comparison on bond strength considering the presence of aging conditions was possible in the included studies due to the low number of studies that aged the specimens. In conclusion, APA had no negative effects on the bond strength of resin-based materials to dentin and was able to improve the dentin bond strength only when the particle size was > 30 μm and air pressure was > 5 bar. PROSPERO registration protocol: CRD42018096128Clinical Relevance
SUMMARY
INTRODUCTION
Airborne-particle abrasion (APA) is a procedure used for several applications in dentistry, with the first report dating back to the 1940s.1 Different air-abrasion devices have been introduced to the market for applications including cavity preparation,1–3 prophylaxis and removal of surface stains,1 selective caries removal,4 tribochemical coating,5 and surface polishing or roughening.6 APA involves propelling a well-defined, sharply focused stream of particles expelled from a small nozzle under high pressure against a surface. The fluid used is usually compressed air, and many particle types, such as sodium bicarbonate, glycine, and aluminum oxide, can be used depending on the intended goal of APA.7,8 The particle size, pressure, and duration of APA may also vary depending on the clinical application9–12 and affect the result of the abrasion process.
Depending on the abrasive particle, the kinetic energy of the accelerated hard particles may result in rapid substance removal on impact.13 Whereas sodium bicarbonate is usually used for polishing procedures, APA with aluminum oxide is commonly used to prepare surfaces to enhance micromechanical retention of restorative materials, such as glass ceramics,14,15 oxide ceramics,16,17 and resin composites.18,19 The objective is usually to increase the area for micromechanical interlocking of adhesive materials.20–22 Despite a limited number of clinical studies, clinical applications of APA of dental substrates using aluminum oxide particles have been reported as a cleaning method, a pretreatment technique before adhesive luting of indirect restorations, and surface treatment before resin composite restorations.23–25
Despite the potential benefits for bonding restorative materials, APA has also been shown to produce surface flaws and microcracks that can compromise the strength of ceramic restorations.26,27 Thus, evaluation of the effects of APA on human dentin is warranted. In several in vitro studies, investigators have examined the effect of APA on dentin10,21,22,28–32 and have usually focused on applying air-abrasion to improve the bond strength of adhesive materials.10,17,24,33–38 Large variability exists among the size of the abrasive particles used, as well as the time duration and pressure used in APA. Pooled in vitro data could help determine whether APA has a positive effect on dentin and ascertain whether the technique can be applied clinically to dentin surfaces without major concerns. The aim of this systematic review of in vitro studies, therefore, was to investigate how APA using aluminum oxide particles affects the bond strength of resin-based materials to human dentin. The null hypothesis was that APA does not have a negative effect on dentin bond strength.
METHODS AND MATERIALS
This systematic review adheres to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.39 The review protocol was registered with the international database for systematic reviews – PROSPERO (protocol CRD42018096128). The PICOT elements were as follows: Population, human dentin; Intervention, air abrasion considering a procedure in which dentin surfaces were subjected to abrasive blasting by a stream of aluminum oxide particles propelled under high pressure with compressed air as the fluid; Comparison, non–air-abraded dentin; Outcome, bond strength of resin-based materials to dentin; and Type of study, in vitro tests.
Systematic Literature Search
The literature search aimed to identify all studies that evaluated the effect of APA using aluminum oxide particles on dentin. The search was systematically performed by two independent reviewers (VPL and VSC) using three online international scientific databases: The National Library of Medicine (MED-LINE/PubMed), ISI Web of Science, and Scopus. The search strategy used in PubMed is shown in Table 1. The strategy was adapted to the other databases accordingly. The final search was performed in October 2018. After the articles were searched, all were imported into Endnote X7 software (Thompson Reuters, Philadelphia, PA, USA) to remove duplicates.
Titles and abstracts were read to verify the inclusion criteria: in vitro studies that reported comparison between air-abraded and non–air-abraded dentin bond strengths. When the study did not clearly define the control group, the non–air-abraded group was considered the control. The following terms were considered in the inclusion criteria: “air abrasion,” “airborne particle abrasion,” “air polishing,” or “sandblasting.” Aluminum oxide particles were the only abrasive eligible for this review. Only studies that evaluated the bond strength of resin-based materials to sound dentin from human teeth exposed to APA were included. Studies that evaluated bovine dentin and abrasive particles other than aluminum oxide were excluded. Only articles published in English were considered, with no restrictions on year of publication. Any disagreement regarding the eligibility of the included studies was resolved through discussion and consensus, or a third reviewer (RRM) was consulted. Only studies that fulfilled all eligibility criteria were included. Whenever information relevant to eligibility was unavailable in the abstract or the abstract itself was unavailable, the article was selected for full-text reading. The reviewers manually searched the reference lists of the included articles for additional relevant studies.
Data Recorded From the Selected Studies
For each included study, the following data and information were recorded using a standard form in spreadsheet format (Excel for Mac version 16.31, Microsoft Corporation, Redmond, WA, USA): control group and its surface treatment, particle size, air-abrasion distance, angle with the surface during air-abrasion, air-abrasion pressure, air-abrasion duration, cleaning method or surface treatment after APA, type of bond strength test, bond strength mean values in MPa, standard deviations, and number of specimens tested.
Data Analysis
Pooled effect estimates were obtained by comparing the standardized mean difference between the air-abraded and control groups within each study with estimated 95% confidence intervals. The standardized mean difference was used to minimize differences in bond strength values measured by different methods, for example, shear or tensile tests. The analyses were performed by adopting a random-effects model using Review Manager version 5.1 (The Nordic Cochrane Centre, The Cochrane Collaboration, Copenhagen, Denmark). As the studies adopted different surface treatments in the control groups, a meta-analysis was carried out considering the same intervention-control subgroups, that is, the control groups were separated according to their surface treatments: no treatment, bur, SiC abrasive paper, hand excavator, or acid etching. Additional analyses were carried considering the different air-abrasion parameters adopted, that is, the abrasive particle size (≤30 μm or >30 μm), air pressure (≤5 bar or >5 bar), APA duration (≤10 seconds or ≥15 seconds) and presence of aging conditions (yes/no). All meta-analyses considered the same combinations of intervention-control groups. Statistical heterogeneity of the treatment effect among studies was appraised using the Cochran Q test, in which values >50% were considered to suggest substantial heterogeneity.40 Multiple groups from the same study were analyzed according to the Cochrane guidelines formula for combining groups.40
Quality Assessment and Risk of Bias
The methodologic quality and risk of bias of the included studies was assessed according to Cochrane guidelines40 and criteria adapted from previous studies41,42 as follows: selection bias (random sequence generation), sample-size calculation, presence of a clearly defined control group, and performance and detection bias (blinding of operator/examiner). Each criterion was judged to have high, low, or unclear risk of bias, which was also used for quality assessment. The assessment of risk of bias was performed using Review Manager version 5.1.
RESULTS
The search resulted in the retrieval of 7340 articles, as shown in the study flowchart presented in Figure 1. After removing duplicates, 5437 unique publications were screened, of which 5372 were excluded because they did not meet the eligibility criteria. A total of 65 articles were assessed in full, including two found in the manual search. From these 65 publications, 32 were excluded for reasons detailed in Figure 1. The list of articles excluded after the eligibility screening is provided as supplementary material in the Appendix. A total of 33 studies were included in the qualitative synthesis, and 32 were included in the meta-analysis. One study that evaluated bond strength43 was not included in the quantitative analysis because the standard deviation value was missing; nevertheless, this study reported similar bond strengths between air-abraded and control dentin.
Citation: Operative Dentistry 46, 1; 10.2341/19-216-L
Varied APA parameters were used across the studies: 70% used aluminum oxide particles with size >30 μm, the air pressure most commonly used was up to 5 bar (51.5%), and air-abrasion duration was usually up to 10 seconds (54.5%). All studies included in this review and the details of the air-abrasive procedures are provided in the Appendix. Only data that were within the scope of this study are reported. Regarding quality assessment (Figure 2), most included studies presented low risk of bias relative to random sequence generation, and the majority of studies presented a clearly defined control group. Sample-size calculation was reported in two studies, and blinding of operator/examiner was not reported in any of the included studies. The risk of bias for each item judged in each included study is shown in Figure 3.
Citation: Operative Dentistry 46, 1; 10.2341/19-216-L
Citation: Operative Dentistry 46, 1; 10.2341/19-216-L
The meta-analysis on dentin bond strength values considering the different combinations of intervention-control comparisons is presented in Figure 4. A significant difference was found between the groups favoring dentin subjected to APA compared with non–air-abraded dentin only when the control surface was treated with a hand excavator (p=0.02, I2=67%). When the other control dentin surface treatments were considered (no treatment, bur, SiC paper, or acid etching), no significant differences between air-abraded and non–air-abraded dentin were detected.
Citation: Operative Dentistry 46, 1; 10.2341/19-216-L
Considering the APA parameter particle size (Figure 5), the results favored APA when the particle size was > 30 μm and the controls were no treatment (p=0.02, I2=64%) and hand excavator (p<0.00001, I2=0%). APA was also favored when the particle size was ≤30 μm and the control was bur (p=0.0004, I2=0%). For the other control surfaces no significant differences between the experimental and control groups were observed. When the parameter air-abrasion pressure was considered in the meta-analysis (Figure 6), the results favored air-abraded groups only when the air pressure was >5 bar and bur was used to treat the control surfaces (p=0.01, I2=0%), with no other significant differences. Two studies44,45 did not report air pressure, thus were not considered in the subgroup analysis. Regarding the parameter APA duration (Figure 7), no significant differences were observed in bond strength between control and air-abraded dentin. Three studies did not report the duration of air abrasion.46–48 and, in two studies, air-abrasion duration was not standardized since it depended on the removal of cement from the surface20,49 or tooth preparation.8,50
Citation: Operative Dentistry 46, 1; 10.2341/19-216-L
Citation: Operative Dentistry 46, 1; 10.2341/19-216-L
Citation: Operative Dentistry 46, 1; 10.2341/19-216-L
It was not possible to perform a meta-analysis considering the presence of aging conditions in the studies as only four articles reported dentin bond strengths for immediate and aged groups.45,49,51,52 Each study had a different type of control, hindering comparisons within the same combinations of intervention-control groups.
DISCUSSION
The null hypothesis tested was accepted since application of APA with aluminum oxide particles had no detrimental effect on the bond strength of resin-based materials to dentin in any of the meta-analyses. Previous studies have reported possible negative effects of APA on dentin characteristics. These studies showed that APA may result in more irregular11,12,26,53–55 or rougher dentin surfaces compared with non–air-abraded dentin.56–59 However, findings of the present investigation suggest that those irregular surface aspects may not negatively interfere with the bonding of resin-based materials to dentin. In fact, increasing dentin surface roughness and producing a more irregular surface texture are the goals of APA in many clinical cases. Propelling aluminum oxide particles to dentin may result in substance removal from the surface because of the kinetic energy of the accelerated particles and differences in hardness between the abrasive and the dentin tissue. Aluminum oxide particles have a Vickers hardness of approximately 1200 kg/mm2, whereas the Vickers hardness of dentin is approximately 57–60 kg/mm2.60 The rougher dentin surfaces may improve micromechanical interlocking between adhesive agents and restorative materials or improve the wettability of dentin surfaces. This may explain the findings showing that APA was able to improve the dentin bond strength, although only in a few cases and only in the short term.
The improved bond strength to air-abraded dentin was dependent on particle size and pressure of the air stream used.20,47,61 Aluminum oxide particles >30 μm in size generally yielded better bond strength, although the same effect was observed for particles ≤30 μm when the control was bur. Application of any particle size could potentially increase surface roughness and, thus, the interaction of adhesive agents with dentin. The differences observed for the distinct particle sizes could be explained by their distinct ability in generating morphologic changes for micromechanical keying on dentin surfaces. In addition, air pressures >5 bar, which are in the range of air pressures produced by dental turbines, also led to improved dentin bond strength in a few cases, whereas lower pressures did not yield the same result. In contrast, APA duration was not particularly important for the bond strength to dentin. Therefore, it appears reasonable to suggest that in cases of use of APA in dentin as surface pretreatment seeking for improved bonding,23–25 aluminum oxide particle sizes >30 μm and air pressure >5 bar should be preferred, although further analyses in this regard are warranted. In case the dentist intends to apply APA to the dentin for other clinical purposes, such as surface cleaning, any particle size or air pressure could be used. It should be highlighted, however, that the actual ability of APA in cleaning the dentin was not investigated here.
Variability in methods used among studies for air-abrading the dentin surfaces was observed, including particle sizes, distances, angulations, cleaning methods after abrasion, and air pressures. Analyses according to the parameters employed were used to aid in minimizing those heterogeneities. The surface treatments applied to the dentin specimens before the bond strength tests were not homogeneous either. Therefore, to minimize the clinical heterogeneity regarding surface conditions, the extracted data were separated according to same intervention-control subgroups, that is, the dentin surface treatments: no treatment, bur, SiC abrasive paper, hand excavator, or acid etching. Comparisons were restricted to same intervention-control conditions. The in vitro literature is known for having problems regarding good reporting practices, especially because no guidelines are available for reporting the results of the numerous types of in vitro tests used in dentistry. In addition, the extracted data were limited to sound dentin to reduce structural and morphologic variability regarding the dentin substrate; thus, the conclusions should not be extrapolated to caries-affected or sclerotic dentin.
Most studies included in this review tested only immediate bond strengths to dentin; that is, there was no evaluation of the effects of water degradation or other aging method on the dentin bonds. In vitro studies are urged to always include a storage group when testing adhesive bond strengths. Previous studies showed that differences reported between treatments in the short term were not observed when the bonded specimens were stored in water for some time before testing.45,49 This finding is of particular importance for APA enthusiasts: the positive effects of the treatment may not persist in the long term. Therefore, it seems that the decision to apply APA to dentin should be made by the dentists, taking into account their own clinical experience; the literature cannot give a definitive answer on whether APA may effectively generate dentin bonds that last longer than non–air-abraded dentin. However, no negative effects for APA applied to dentin were observed either; thus, the treatment seems to be safe with respect to bonding to dentin.
CONCLUSION
Within the limitations of this study, the following conclusions can be drawn:
APA with aluminum oxide particles had no negative effects on the bond strength of resin-based materials to dentin.
In a few subgroup analyses, air abrasion was able to improve the immediate bond strength to dentin when the particle size was >30 μm and air pressure was >5 bar.
APA duration had no significant effect on immediate dentin bond strengths.
Flowchart of the systematic review.
Risk of bias graph: proportion of studies with low, unclear, or high risk of bias for each item.
Risk of bias for each item judged as low, unclear or high in each included study.
Meta-analysis on dentin bond strength values considering the different combinations of intervention-control comparisons. Statistically significant difference was observed when the control surface was treated with a hand excavator (p=0.02).
Meta-analysis for particle size (≤30 μm or >30 μm) with same intervention-control groups. Statistically significant differences were observed when the particle size was >30 μm and the controls were no treatment (p=0.02) or hand excavator (p<0.00001). Considering particle size ≤30 μm, a statistically significant difference was observed when the control was bur (p=0.0004).
Meta-analysis for air pressure (≤5 bar or >5 bar) with same intervention-control groups. Statistically significant difference was observed when the air pressure was >5 bar and bur was used to treat the control surfaces (p=0.01).
Meta-analysis for airborne-particle abrasion duration (≤10 seconds or ≥15 seconds) with same intervention-control groups. No significant differences were observed in bond strength between control and air-abraded dentin.
Contributor Notes
Verônica P. Lima, DDS, MS, PhD candidate, Graduate Program in Dentistry, Federal University of Pelotas, Pelotas, Brazil
Karla D.A. Soares, DDS, Graduate Program in Dentistry, Federal University of Sergipe, Aracaju, Brazil
Vanderson S. Caldeira, DDS, School of Dentistry, Federal University of Pelotas, Pelotas, Brazil
André L. Faria-e-Silva, DDS, MS, PhD, professor, Graduate Program in Dentistry, Federal University of Sergipe, Aracaju, Brazil
Bas A.C. Loomans. DDS, MS, PhD, assistant professor, Department of Dentistry, Radboud University Medical Center, Nijmegen, The Netherlands
*Rafael R. Moraes, DDS, MS, PhD, professor, Graduate Program in Dentistry, Federal University of Pelotas, Pelotas, Brazil