Abstract
Objectives
Inhibition of hybrid layer degradation, for example via inhibition of matrix-metallo-proteinases (MMP) could reduce risk of retention loss and failure of adhesively placed restorations. This systematic review investigated such inhibitory pretreatment qualitatively and via meta – and trial-sequential-analysis.
Data sources
We included randomized clinical trials comparing degradation inhibitory cavity pretreatment versus no, placebo or alternative treatments prior adhesive placement of resin-based restorations. Trials reporting retention loss or failure (graded bravo-delta in USPHS or similar criteria) were included. Trial selection, data extraction, and risk of bias assessment were conducted independently by two reviewers. Fixed- or random-effects intention-to-treat, per-protocol, and scenario meta -analyses were performed, and trial-sequential-analysis used to control for risk of random errors. Electronic databases (PubMed, Embase, Cochrane CENTRAL) were systematically screened, and hand searches and cross-referencing performed.
Study selection
The ten included trials involved 208 patients (695 cavities) and used chlorhexidine (seven trials), ethanol-wet-bonding (two trials), and quaternary ammonium compounds for degradation inhibition. All but one trial had high risk of bias. Follow-up ranged from 6 to 36 months. Risk of retention loss was not significantly decreased after pretreatment (per-protocol OR [95% CI] 1.37 [0.68/2.77], intention-to-treat: 1.25 [0.76/2.04]). This was found for risk of restoration failure as well (per-protocol: 0.86 [0.56/1.34], intention-to-treat: 1.22 [0.83/1.80]). Scenario analyses found great uncertainty introduced by attrition. According to trial sequential analysis, no firm evidence was reached.
Conclusions
There is insufficient evidence to recommend or refute degradation inhibitory cavity pretreatment prior adhesively placing resin-based restorations. This may change if teeth are followed-up for longer.
Clinical significance
Dentists can perform cavity pretreatments for inhibition of hybrid layer degradation, but a beneficial effect is not supported by sufficient evidence. The impact of further effects (e.g. disinfection, pulp-irritation) remains unclear.
1
Introduction
Although adhesively placed resin-based restorations bear acceptable survival rates, dentists still spend a significant amount of time replacing clinically failed restorations. The most common reasons for failure are reported as fracture and secondary caries, whereas fracture is assumed to occur early and secondary caries predominantly later within the function period of the restoration . Different factors such like operator-, patient-, materials- and tooth-related variables might contribute to restoration failure. Among these variables caries risk of the individual and cavity size and/or geometry seem to significantly impact restoration failure . Another potential factor that could contribute to clinical failures is degradation of the adhesive-dentine interface. Some laboratory studies suggest that enzymatic degradation of the hybrid layer could weaken the adhesive-dentin bond . However, it is unclear if weakening of the adhesive dentin-bond by hybrid layer degradation can promote restoration failure in clinical situations.
Most current adhesive systems use a hydrophilic component either for rewetting and thus re-expanding the collapsed collagen network (as part of etch-and-rinse adhesives) or for ionizing acidic components (usually in self-etch adhesives). The resulting hybrid layer oftentimes contains water components and incompletely resin-covered collagen fibrils, which are prone for collagenolytic and gelatinolytic cleavage . This enzymatic degradation involves dentin matrix-metallo proteinases (MMP) , which are activated by the acidic resins applied during dentin etching . Long-term interfacial degradation is thought to reduce adhesive strengths, allowing gap formation , and thus leading to leakage and secondary caries lesion formation as main long-term failure of resin-based restorations .
Hybrid layer degradation inhibitory pretreatment of the cavity prior to adhesive placement of these restorations could thus reduce risk of failure and retention loss. A variety of pretreatment strategies have been suggested. They either involve MMP-inhibition by denaturation via strong acids, chelating of the Zn-component of MMPs via chlorhexidine, tetracycline, EDTA, and quaternary ammonium salts like 12-methacryloyloxydodecyl pyridinium bromide (MDPB), or prevention of collagen cleavage by protecting fibers. For the latter, ethanol-wet bonding is suggested, which allows to replace water within the collagen network with ethanol, creating greater inter-fibrillar space to allow full infiltration of the network with hydrophobic resins. This is thought to allow a more complete resin-protection of collagen fibers . Both MMP-inhibitory pretreatments and ethanol-wet bonding have been found to increase bond strengths in vitro and ex vivo . However, the clinical relevance of such pretreatment remains unclear.
Our investigation was aimed at comparing the risk of retention loss and failure of adhesively placed resin-based restorations after degradation inhibitory cavity pretreatment versus no, placebo or alternative pretreatments. To do so, a systematic review and meta analysis was performed on available randomized trials. As traditional meta -analyses do not sufficiently account for random errors, i.e., spuriously significant results (type I errors) or spuriously insignificant results (type II errors) , trial sequential analysis (TSA) was also conducted. TSA uses calculation of a required information size (RIS), and handles meta -analyses not reaching this RIS with more stringent statistical thresholds . These so-called monitoring boundaries are calculated for superiority/inferiority, but also futility, thus indicating when to stop conducting more randomized trials . Based on this review, recommendations for clinical practice and research should be deducted.
2
Methods
2.1
Eligibility criteria
Randomized clinical trials investigating humans receiving adhesively placed resin-based restorations, with minimum two treatment groups comparing degradation inhibitory cavity pretreatment with no such treatment, placebo treatment, or alternative pretreatments, were included. Treatment groups should only differ with regards to pretreatment, other provided treatments (moisture control, carious tissue removal, bonding strategy, restoration) should be identical. Not performing any such pretreatment was regarded as control group, i.e., standard of care. Pretreatment was seen as experimental group. We did not specify what kind of pretreatment was used, and used unspecific search terms accordingly (see below).
Trials needed to have recorded restoration retention loss, which we treated as primary outcome, and/or failure, which was the secondary outcome. Failure was defined as the number of restorations without alpha-rating in the United States Public Health Services (USPHS) criteria or similar rating systems. Note that discoloration of restorations was not used to assess failure, as this was not assumed to be affected by pretreatment. No restrictions with regard to language or publication date were applied.
2.2
Search, data extraction, risk of bias
Electronic databases (Embase, Medline, Cochrane Central) were screened using a defined search protocol, which was adapted for each database. Studies were cross-referenced via bibliographies of identified full-texts. Blinding for authors or journals was not performed. Title and abstract of identified studies were screened independently by two calibrated reviewers (FS and GG) for eligibility, and inclusion decided by both reviewers as well. Consensus was obtained by discussion.
Duplicate data extraction was performed independently by two reviewers (FS, GG) using a piloted spreadsheet. The following items were extracted: study name, year of publication, setting, source of funding, sample (size, age, characteristics, number of teeth and lesions), moisture control, pre-treatment and control, bonding strategy, restoration placed, follow-up, attrition, outcome measure, findings. If several time points were recorded and reported, the latest reported time point was used for analysis. Risk of bias was assessed by two independent reviewers (FS, GG) as outlined by the Cochrane Collaboration .
2.3
Meta-analysis
Meta- and trial-sequential-analyses were performed for both outcomes, retention loss and failure. No quality threshold with regards to risk of bias was used to decide inclusion into quantitative analyses. Continuity correction of +1 was used for trials with zero events. Heterogeneity was assessed using both Cochran’s Q and I 2 -statistics . Depending on heterogeneity (I 2 < 35% or above), fixed or random-effects meta -analysis was performed. Odds ratios and 95% confidence intervals (95% CI) were calculated. Four analyses were performed: (1) per-protocol analysis (PP), i.e. assessment of participants based on the intervention they received and their availability for follow-up. PP accounts for possible bias introduced by attrition and protocol deviations . (2) intention-to-treat analysis (ITT), i.e. assessment of participants as randomized regardless of whether they received the intervention or were available for follow-up. For ITT, it was assumed that all missing participants experienced an event. and scenario analyses following ITT principle. Here, attrition was handled differently in experimental and control group. In the best-case analysis, it was assumed drop-outs in the control, but not the experimental group to be associated with events. In the worst-case scenario, this was reversed. Scenario analyses explore the uncertainty stemming from missing data via most extreme imputations .
Publication bias was assessed graphically via simple test of asymmetry of funnel plots and odds ratio adjusted (ORa) for systematically missing values using trim-and-fill . Meta-analysis was performed with Comprehensive Meta Analysis (2.2.064, Biostat, Englewood, NJ, USA).
Given the low number of trials and their uniform design, no further subgroup or meta -regression analyses were performed. Data were not adjusted to account for possible clustering of teeth in studies using split-mouth design etc., as it was assumed that effects of such clustering would be limited. Quality of evidence for each outcome effect estimate was graded according to the GRADE working group of evidence using GRADE Profiler 3.6 .
2.4
Trial sequential analysis (TSA)
Conventional meta -analysis uses Z-values to compare two interventions, with Z = 0.0 indicating no difference between intervention groups. If Z exceeds ±1.96, a difference is traditionally assumed to be statistically significant (p ≤ 0.05, two-sided test). As for repeated updates of meta -analyses, a new Z-value is calculated for each update. In TSA, this series of Z-values is plotted against the accumulated number of patients, events, or information . This ‘cumulative Z-curve’ is then assessed regarding its relation to the conventional significance boundaries (Z = ±1.96), the required information size (RIS), and the trial sequential monitoring boundaries (TSMB) for benefit, harm, or futility. The RIS was calculated based on type I error risk of α=0.05, a type II error risk of β=0.20 (equivalent to a power of 0.80) and the control event proportion. Relative risk reduction (RRR) was based on an a priori defined worthwhile interventional effect of 20%. It should be noted that smaller intervention effects might well be relevant . This, however, would increase the RIS even further . The required RIS was further adjusted for the diversity in the meta -analysis (diversity-adjusted RIS: DARIS) . The Lan-DeMets version of the O’Brien–Fleming function was used for calculating the TSMBs. Results of cumulative Z-value crossing the conventional boundary of significance (Z = ±1.96) but not the TSMBs for benefit or harm were defined as spuriously significant. Firm evidence was assumed to be reached when the Z-curve crossed the TSMB for benefit or harm before the DARIS was reached. Firm evidence of futility was confirmed by the Z-curve crossing the TSBM for futility. TSA 0.9 (Copenhagen Trial Unit, Copenhagen, Denmark) was used .
2
Methods
2.1
Eligibility criteria
Randomized clinical trials investigating humans receiving adhesively placed resin-based restorations, with minimum two treatment groups comparing degradation inhibitory cavity pretreatment with no such treatment, placebo treatment, or alternative pretreatments, were included. Treatment groups should only differ with regards to pretreatment, other provided treatments (moisture control, carious tissue removal, bonding strategy, restoration) should be identical. Not performing any such pretreatment was regarded as control group, i.e., standard of care. Pretreatment was seen as experimental group. We did not specify what kind of pretreatment was used, and used unspecific search terms accordingly (see below).
Trials needed to have recorded restoration retention loss, which we treated as primary outcome, and/or failure, which was the secondary outcome. Failure was defined as the number of restorations without alpha-rating in the United States Public Health Services (USPHS) criteria or similar rating systems. Note that discoloration of restorations was not used to assess failure, as this was not assumed to be affected by pretreatment. No restrictions with regard to language or publication date were applied.
2.2
Search, data extraction, risk of bias
Electronic databases (Embase, Medline, Cochrane Central) were screened using a defined search protocol, which was adapted for each database. Studies were cross-referenced via bibliographies of identified full-texts. Blinding for authors or journals was not performed. Title and abstract of identified studies were screened independently by two calibrated reviewers (FS and GG) for eligibility, and inclusion decided by both reviewers as well. Consensus was obtained by discussion.
Duplicate data extraction was performed independently by two reviewers (FS, GG) using a piloted spreadsheet. The following items were extracted: study name, year of publication, setting, source of funding, sample (size, age, characteristics, number of teeth and lesions), moisture control, pre-treatment and control, bonding strategy, restoration placed, follow-up, attrition, outcome measure, findings. If several time points were recorded and reported, the latest reported time point was used for analysis. Risk of bias was assessed by two independent reviewers (FS, GG) as outlined by the Cochrane Collaboration .
2.3
Meta-analysis
Meta- and trial-sequential-analyses were performed for both outcomes, retention loss and failure. No quality threshold with regards to risk of bias was used to decide inclusion into quantitative analyses. Continuity correction of +1 was used for trials with zero events. Heterogeneity was assessed using both Cochran’s Q and I 2 -statistics . Depending on heterogeneity (I 2 < 35% or above), fixed or random-effects meta -analysis was performed. Odds ratios and 95% confidence intervals (95% CI) were calculated. Four analyses were performed: (1) per-protocol analysis (PP), i.e. assessment of participants based on the intervention they received and their availability for follow-up. PP accounts for possible bias introduced by attrition and protocol deviations . (2) intention-to-treat analysis (ITT), i.e. assessment of participants as randomized regardless of whether they received the intervention or were available for follow-up. For ITT, it was assumed that all missing participants experienced an event. and scenario analyses following ITT principle. Here, attrition was handled differently in experimental and control group. In the best-case analysis, it was assumed drop-outs in the control, but not the experimental group to be associated with events. In the worst-case scenario, this was reversed. Scenario analyses explore the uncertainty stemming from missing data via most extreme imputations .
Publication bias was assessed graphically via simple test of asymmetry of funnel plots and odds ratio adjusted (ORa) for systematically missing values using trim-and-fill . Meta-analysis was performed with Comprehensive Meta Analysis (2.2.064, Biostat, Englewood, NJ, USA).
Given the low number of trials and their uniform design, no further subgroup or meta -regression analyses were performed. Data were not adjusted to account for possible clustering of teeth in studies using split-mouth design etc., as it was assumed that effects of such clustering would be limited. Quality of evidence for each outcome effect estimate was graded according to the GRADE working group of evidence using GRADE Profiler 3.6 .
2.4
Trial sequential analysis (TSA)
Conventional meta -analysis uses Z-values to compare two interventions, with Z = 0.0 indicating no difference between intervention groups. If Z exceeds ±1.96, a difference is traditionally assumed to be statistically significant (p ≤ 0.05, two-sided test). As for repeated updates of meta -analyses, a new Z-value is calculated for each update. In TSA, this series of Z-values is plotted against the accumulated number of patients, events, or information . This ‘cumulative Z-curve’ is then assessed regarding its relation to the conventional significance boundaries (Z = ±1.96), the required information size (RIS), and the trial sequential monitoring boundaries (TSMB) for benefit, harm, or futility. The RIS was calculated based on type I error risk of α=0.05, a type II error risk of β=0.20 (equivalent to a power of 0.80) and the control event proportion. Relative risk reduction (RRR) was based on an a priori defined worthwhile interventional effect of 20%. It should be noted that smaller intervention effects might well be relevant . This, however, would increase the RIS even further . The required RIS was further adjusted for the diversity in the meta -analysis (diversity-adjusted RIS: DARIS) . The Lan-DeMets version of the O’Brien–Fleming function was used for calculating the TSMBs. Results of cumulative Z-value crossing the conventional boundary of significance (Z = ±1.96) but not the TSMBs for benefit or harm were defined as spuriously significant. Firm evidence was assumed to be reached when the Z-curve crossed the TSMB for benefit or harm before the DARIS was reached. Firm evidence of futility was confirmed by the Z-curve crossing the TSBM for futility. TSA 0.9 (Copenhagen Trial Unit, Copenhagen, Denmark) was used .
3
Results
3.1
Results of the searches
From 1006 identified studies, 18 were evaluated full-text, and 10 randomized trials included ( Fig. 1 ). Excluded studies and reasons for exclusion can be found in the appendix (Table S1).
Overall, 208 patients with 695 cavities were treated ( Table 1 ). Six trials treated permanent teeth in adults, two primary teeth in children, and one premolars scheduled for extraction in children. One trial did not state the age of the patients, and treated third molars scheduled for extraction. Seven trials used chlorhexidine as cavity pretreatment, two performed ethanol wet-bonding, and one used quaternary ammonium compounds (MDPB). Five trials used rubberdam for moisture isolation. All but one trial had used the etch-and-rinse technique; two used a self-etch adhesive in an additional group. Follow-up ranged from 6 to 36 months, with drop-outs of 0% up to 73%. Six trials had used USPHS criteria for assessment of failure, one FDI criteria, and three had only determined survival of restorations. All but one trial had high risk of bias ( Table 2 ).
| Study | Type, country | Sample | Teeth, cavity class | Treatments | Follow-up (mos), outcome measure | Attrition | Retained | With α-rating |
|---|---|---|---|---|---|---|---|---|
| De Araujo 2013 | Split-mouth, Brazil | 17 adults (23–54) with low caries risk and at least three non-carious cervical lesions | 91 teeth, class V | 3 groups (n = 31): ER with Adper Scotchbond MB (ADP), SE with Adper Easy One (AEO, 3 M Espe), Ethanol-wet-bonding (EWB, sequence of 50%, 100% ethanol each for 20s, followed by experimental hydrophobic primer (ADP primer diluted in 100% ethanol to 10 m%) for 5s, then ADP adhesive for 10s; relative moisture control, restoration with Filtek Z350 XT (3 M) | 12, USPHS | 3% (1 patient, 3 teeth) |
ADP: 31/31, AEO: 27/30, EWB: 27/29 | ADP: 23/31, AEO: 10/30, EWB: 25/29 |
| Dutra-Correa 2013 | Split-mouth, Brazil | 37 adults (27–79) with min. 20 teeth, 2–4 of them needing a class V restoration | 120 vital teeth, mostly premolars and molars, class V | 4 groups (n = 30): ER with Xeno Bond Plus (XBP, Dentsply Caulk, Milford, USA), SE with Xeno V (XEN, Dentsply Detrey, Konstanz, Germany), both groups treated with 2% CHX (GFM, Joinville, Brazil) prior placing adhesive; relative moisture control, restored with Esthet X (Dentsply Caulk) | 18, USPHS | 23% (13 patients, 28 teeth) |
XPB: 24/24 XBP/CHX: 20/22 XEN: 21/23 XEN/CHX: 22/23 | XPB: 21/24 XBP/CHX: 20/22 XEN: 20/23 XEN/CHX: 21/23 |
| Sartori 2013 (44) | Split-mouth, Brazil | 20 adults (33–64) with min. 20 teeth, 2–4 of them with non-carious cervical lesions | 70 vital teeth, class V |
2 groups (n = 35): ER with Adper Single Bond 2 (AS2, 3 M Espe, St. Paul, USA), ER with ADP + 2% CHX (Dermus, Florianopolis, Brazil), relative moisture control, restored with Filtek Supreme XT (3 M Espe) | 36, USPHS | 29% (6 patients, 20 teeth) |
AS2: 22/25, ADP/CHX: 21/25 | AS2: 18/25, ADP/CHX: 17/25 |
| Brackett 2007 | Split-mouth, Mexico | 6 children (median 15 years) with pairs of premolars scheduled for extraction for orthodontic reasons | 12 premolars, class I | 2 groups (n = 6): ER with Adper Single Bond Plus (ASP, 3 M Espe), ER with ASP + 2% CHX (Cavity Cleanser, BISCO, Schaumburg, IL, USA) prior placing adhesive; rubberdam isolation, restored with Epic TMTP, Parkell, Edgewood, NY, USA) and Filtek Supreme (3 M Epe) | 6, survival |
0% (0,0) | ASP: 6/6 ASP/CHX: 6/6 | not reported |
| Carrilho 2007 | Split-mouth, Brazil | 12 patients (age not reported), non-carious third molars scheduled for extraction | 18 non-carious third molars, class I | 2 groups (n = 9): ER with Single Bond (SB, 3 M Espe), ER with SB + 2% CHX (PRODERMA, Piracicaba, SP, Brazil) prior placing adhesive; rubberdam isolation; restored with Clearfil Protect Liner (Kuraray Medical Inc., Osaka, Japan) and Z250 (3 M Espe) | 14, survival | 0% (0,0) | SB 9/9 SB/CHX 9/9 | not reported |
| Helbling 2005 | Split-mouth, Brasil | 11 children (8–12) with carious primary molars | 22 primary molars, class I | 2 groups (n = 22); ER with Single Bond (SB, 3 M Espe), ER with SB + 2% CHX (Cavity Cleanser, Bisco), rubberdam isolation; restored with EPIC-TMPT (Parkell Inc., Farmingdale, NY, USA) and Z250 (3 M Espe) | 6, USPHS | 73% (8 patients, 16 teeth) |
SB 3/3 SB/CHX 3/3 | SB 3/3 SB/CHX 3/3 |
| Montagner 2015 | Split- mouth, Brazil | 42 adults (21–76) with min 20 teeth, 2 of them with non-carious cervical lesions | 169 incisors, canines or premolars, class V | 2 groups; ER with Adper Single Bond 2 (AS2, 3 M Espe, St. Paul, USA) (n = 81), ER with AS2 + 2% CHX (Manipulation Solution; Uso Indicado Pharmacy Pelotas, RS, Brazil) (n = 88), relative moisture control, restored with Filtek Z 350 (3 M Espe) | 6, FDI | 0% [0, 0] | AS2 78/81 AS2/CHX 85/88 | AS2 26/81 AS2/CHX 39/88 |
| Türkün 2005 | Split- mouth, Turkey | 39 adults (26–59) with non-carious cervical lesions and normal periodontal health | 163 teeth in normal occlusion relationship to opposing tooth, class V | 2 groups; Self etch adhesive Xeno III (X3, Dentsply, Konstanz, Germany) (n = 78); Self etch Clearfil Protect Bond (CPB, Kuraray) (n = 85), relative moisture control, restoration with Esthet X (Dentsply) | 12, USPHS | 0% (0,0) | X3 75/78 CPB 85/85 | X3 76/78 CPB 84/85 |
| Mortazavi 2012 | Split-mouth Iran | 12 adults (30–60) with non-carious cervical lesions who were generally healthy and had good oral hygiene | 36 vital teeth, class V | 3 groups (n = 36): ER Optibond FL (OFL, Kerr, USA); OFL + Ethanol wet Bonding, SE with Clearfil SE Bond (CSE, Kuraray, Japan), rubberdam isolation, restored with Grandio (Voco, Cuxhaven, Germany) | 12, USPHS | 0% (0,0) | OFL 12/12 OFL/Eth 11/12 CSE 12/12 | OFL 12/12 OFL/Eth 11/12 CSE 11/12 |
| Ricci 2010 | Split mouth Brazil | 13 children (8–11) with at least two contralateral carious primary molars, collection of teeth after exfoliation | 8 primary molars with occlusal caries lesions limited to the outer half of dentin, class I | 2 groups (n = 8); ER adhesive Prime & Bond NT (PBNT, Dentsply Caulk, Milford, USA); ER with PBNT + 2% CHX (FGM Produtus Odontologicos, Joinville, SC, Brazil); rubber dam isolation; restored with Z250 (3 M Espe) | 10−12 month (n = 4), 18−20 month (n = 4), survival | not extractable | PBNT (10–12) 2/2 PBNT + CHX (10–12) 2/2 PBNT (18–20) 2/2 PBNT + CHX (18–20) 2/2 | not reported |
Stay updated, free dental videos. Join our Telegram channel
VIDEdental - Online dental courses
