Longevity of direct and indirect resin composite restorations in permanent posterior teeth: A systematic review and meta-analysis

Accepted Manuscript Title: Longevity of direct and indirect resin composite restorations in permanent posterior teeth: A systematic review and meta-analysis Author: Ana Maria Antonelli da Veiga Amanda Carneiro Cunha Daniele Masterson Tavares Pereira Ferreira Tatiana Kelly da Silva Fidalgo Thomaz Kauark Chianca K´atia Rodrigues Reis Lucianne Cople Maia PII: DOI: Reference:

S0300-5712(16)30160-9 http://dx.doi.org/doi:10.1016/j.jdent.2016.08.003 JJOD 2655

To appear in:

Journal of Dentistry

Received date: Revised date: Accepted date:

9-2-2016 13-5-2016 10-8-2016

Please cite this article as: {http://dx.doi.org/ This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Title of paper: Longevity of direct and indirect resin composite restorations in permanent posterior teeth: a systematic review and meta-analysis

Short title: Longevity of direct and indirect resin restorations: systematic review

Name of authors: Ana Maria Antonelli da Veigaa, Amanda Carneiro Cunhab, Daniele Masterson Tavares Pereira Ferreirac, Tatiana Kelly da Silva Fidalgob, Thomaz Kauark Chiancab, Kátia Rodrigues Reisa, Lucianne Cople Maiab

Job titles and adresses: a

Department of Prosthodontics and Dental Materials, Federal University of Rio

de Janeiro, Av. Carlos Chagas Filho, 373, Prédio do CCS, Bloco K, Ilha da Cidade Universitária, Rio de Janeiro, Brazil. b

Department of Pediatric Dentistry and Orthodontics, Federal University of Rio

de Janeiro, Av. Carlos Chagas Filho, 373, Prédio do CCS, Bloco K, Ilha da Cidade Universitária, Rio de Janeiro, Brazil. c

Center of Health Science, Federal University of Rio de Janeiro, Av. Carlos

Chagas Filho, 373, Prédio do CCS, Bloco I, Ilha da Cidade Universitária, Rio de Janeiro, Brazil

Corresponding Authors: Dr Lucianne Cople Maia Disciplina de Odontopediatria da FO-UFRJ, caixa postal 68066 – CEP: 21941971- Cidade Universitária – Rio de Janeiro – RJ – BRAZIL. Tel +55 21 39382098 Endereço eletrônico: [email protected] (Cople Maia, L)

Keywords: resin composite restoration, inlay, direct composite, longevity, systematic review, meta-analysis.

ABSTRACT

Longevity of direct and indirect resin composite restorations in permanent posterior teeth: a systematic review and meta-analysis

Objectives: The aim of this systematic review and meta-analysis was to assess the differences in clinical performance in direct and indirect resin composite restorations in permanent posterior teeth. Sources: PubMed, the Cochrane Library, Web of Science, Scopus, LILACS, BBO, ClinicalTrials.gov and SiGLE were searched without restrictions. Study selection: We included randomized clinical trials (RCTs) that compared the clinical performance of direct and indirect resin composite restorations in Class I and Class II cavities in permanent teeth, with at least two years of follow-up. The risk of bias tool suggested by Cochrane Collaboration was used for quality assessment. Data: After duplicate removal, 912 studies were identified. Twenty fulfilled the inclusion criteria after the abstract screening. Two articles were added after a hand search of the reference list of included studies. After examination, nine RCTs were included in the qualitative analysis and five were considered to have a ‘low’ risk of bias. The overall risk difference in longevity between direct and indirect resin composite restorations in permanent posterior teeth (p > 0.05) at five-year follow-up was 1.494 [0.893–2.500], and regardless of the type of tooth restored, that of molar and premolars was 0.716 [0.177–2.888] at three-year follow-up. Conclusions: Based on the findings, there was no difference in longevity of direct and indirect resin composite restorations regardless of the type of material and the restored tooth. Clinical significance: Contemporary dentistry is based on minimally invasive restorations. Any indication of a less conservative technique must have

unquestionable advantages. In vitro and in vivo studies reveal contradictory evidence of the clinical performance of direct and indirect resin composite restorations in posterior teeth. Thus this study clarified this doubt. Keywords: resin composite restoration, inlay, direct composite, longevity, systematic review, meta-analysis.

1. Introduction

Direct and indirect resin composite restorations are widely used in contemporary dentistry to restore posterior teeth.1-3 Traditionally, the choice between the use of direct and indirect techniques for resin composites in posterior teeth is based on the size of the cavity to be restored. Small and medium cavities are usually restored with direct composite resin restorations. On the other hand, in large cavities, where the width of the isthmus exceeds two-thirds of the distance between facial and lingual cusp tips, indirect restorations become indicated.3-6 However, because of the evidence that direct resin composite restorations have properties suitable for use in posterior teeth,711

do not require invasive preparation and12 are made in only one session at low

cost, many dentists are also using them in large cavities, making the clinical decision challenging.14 Bis-GMA-based

resin

composites

could

have

considerable

polymerization linear shrinkage around 0.36–0.88 %15 and volumetric shrinkage of about 1.5 to 3.4 %.16 This phenomenon is a consequence of the reduction of intermolecular distances by creation of single covalent bonds between resin monomers during the formation of the polymer network. These volumetric changes may lead to the formation of local interfacial gaps and consequent microleakage.17,18 However, it is still not clear whether these gaps could exceed the clinically relevant width of around 60µm at the outer margin of the restoration.17 The magnitude of this contraction depends on factors like resin matrix formulation, amount of filler used in the resin composite, degree of conversion19 and incremental filling technique.20 The stress generated by this polymerization shrinkage in direct resin composites is much higher than in

indirect ones (13 times).21 For indirect resin composite restorations, postcure using light, heat, pressure or atmosphere of nitrogen5 and the thin layer of adhesive cement help to relax the stress of the contraction of polymerization.22 The only shrinkage which may influence interfacial adaptation occurs in the cement layer.19 Thus, it is expected theoretically that the mechanical and physical properties23-25 and consequently the clinical performance of indirect resin composite restorations should be improved.21, 25-27 However, when both techniques are compared, in vitro21, 26-28 and in vivo studies2,

29

seem not to be unanimous regarding the best results of one

technique over the other, including when it is used to restore large cavities in posterior teeth.1, 2, 4, 30 Thus, the aim of the present systematic review (SR) and meta-analysis was to answer the following question focus: is there a difference in the clinical longevity of restorations performed of direct and indirect resin composite restorations in permanent posterior teeth?

2.

Materials and methods

2.1 Protocol and registration

The

study

protocol

was

registered

on

the

PROSPERO

database

(http://www.crd.york.ac.uk) under number CDR42015020210 on 05/06/2015 and we followed as closely as possible the preferred reporting items of systematic review and meta-analysis protocols (PRISMA-P) 2015.31

2.2 Eligibility criteria

The controlled vocabulary (MeSH terms) and free terms were used to define the search strategy (Table 1) based on the elements of PICOS questions (S – study design)32 as follows: 1. Population (P): permanent posterior teeth of humans, with Class I or Class II cavities, with or without cusp involvement, from decay, exchange of pre-existing dental material or dental erosion. 2. Intervention (I): indirect resin composite restorations.

3. Comparison (C): direct resin composite restorations. 4. Outcome (O): longevity of the direct and indirect resin composite restorations. 5. Study design (S): randomized clinical trial (RCTs) with at least two years of follow-up. Only RCTs that compared the longevity of direct and indirect resin composite restorations in Class I or Class II cavities with or without cusp involvement and with at least two years of follow-up were selected and included. There were no restrictions regarding setting, language or year of publication. The failure rate of direct and indirect resin composite restorations in posterior teeth was the primary outcome. Secondly, the failure rate in a subgroup that compared direct resin composite (DRC) with indirect resin composite (IRC) and direct composite resins (DRC) and direct inlay/onlay restorations (DIO) was evaluated, along with whether the type of teeth, bruxism and risk of caries have some effect of clinical longevity of the resin composite restorations.

2.3 Exclusion criteria

Non-controlled clinical trials, case reports, series of cases, reviews, abstracts, in vitro studies, observational studies, discussions, interviews, editorial and expert opinion were excluded. Additionally, studies that do not relate to the issue because they were conducted in relation to another type of cavity, cavities extending to the root surface, restorations of anterior teeth, deciduous teeth, not vital tooth and tooth without antagonist were also excluded.

2.4 Information sources and search

The literature search strategy was developed using a combination of MeSH terms with free terms most frequently cited in published literature related to direct and indirect resin composite restorations in posterior teeth. The search process was performed by two independent researchers (A.M.A.V and A.C.C) under the guidance of an expert librarian (D.T.P.F). The search strategies were adapted according to the requirement of the base researched and are described

individually (Table 1). The terms were searched in the fields Title and Abstract without application of any filter or limit regards to the idiom. For manuscripts in idioms other than English (such as Spanish, French and Portuguese), there was at least one researcher in the team to translate them. The last update was performed on August 18, 2015 and the following electronic databases were searched: MEDLINE via PubMed, Cochrane Library, Web of Science, Scopus, Latin American and Caribbean Health Sciences (LILACS) and the Brazilian Library of Dentistry (BBO). To locate unpublished and ongoing clinical trials related to the review question we follow the website ClinicalTrials.gov (www.clinicaltrial.gov), and the grey literature was explored using the database System for Information on Grey Literature in Europe (SIGLE) (Table 1).

2.5 Study collection and data collection process All references were collected in web software (www.myendnoteweb.com) during the selection process. Articles were selected by title and abstract and when appearing in more than one database they were considered only once. When there was insufficient information in the title and abstract, full-text articles were obtained to make a clear decision. Two of the reviewers (A.M.A.V and A.C.C) classified those who met the inclusion criteria. If multiple reports of the same study were identified but with, for example, a different follow-up, only the report with the longer follow-up was considered to avoid overlapping data. However, in the case of doubts regarding methodology, the article that gave rise to the study series was accessed or its authors were consulted. We hand-searched the reference lists of the selected articles in order to find any studies that had not been identified by the electronic search strategy. Selected articles about which the two authors did not agree were included or excluded in a consensus meeting with a third author (L.C.M), following the predefined eligibility criteria. Each study was identified by ID, combining the first author and year of publication. A pilot test was conducted using a sample of study reports to verify that the eligibility criteria were consistent with the research question. The following

details

were

extracted

by

customized

forms

based

on

recommendations by Cochrane Handbook 5.0.2 (www.handbook.cochrane.org):



Details of the sources including first researcher, year of publication and reviewer’s identification.



Details of eligibility including study design, population, intervention, control and follow-up.



Details of participants including number of restorations per group, age and gender.



Details of intervention and control including type of intervention, material and technique used and time spent until the outcome.



Details of outcomes including success and failure rates, and number of and reasons for dropouts.

In addition, information was extracted about source of funding and number of attempts to contact the authors, if necessary (up to three attempts in over a period of two months through the electronic means available). When the corresponding authors did not answer the researchers, the same number of attempts were made to contact other co-authors. 2.6 Risk of bias in individual studies Qualitative analysis of the studies was carried out using the Collaboration’s tool for assessing risk of bias in RCTs. Due to the methodological characteristics of the studies, only four domains were considered key domains for the assessment of the risk of bias: sequence generation, allocation concealment, incomplete outcome data and selective outcome reporting. Blinding of participants, personnel and outcome assessors was not considered key due to the specific characteristics of the studies that compare direct and indirect restorations. The risk of bias for each entry recording was judged as ‘no’ to indicate high risk of bias, ‘yes’ to indicate low risk bias and ‘unclear’ to indicate either lack of information or uncertainty over the potential risk of bias. When a study was judged as ‘unclear’ in any of fields, contact with the authors by electronic message was made in order to obtain more information and to enable the judgment of low or high risk of bias. During the extraction of the data, in the case of disagreements between reviewers, these were resolved through discussion with an experienced researcher (L.C.M).

2.7 Summary measures and synthesis of the results A meta-analysis was performed using the Comprehensive Meta-Analysis software (version 3.2, Biostat, Englewood, USA) to assess differences in the clinical longevity of direct and indirect resin composite restorations. The studies showed that failure rates for each group and the total number of teeth were included in the meta-analysis. Since the studies had the same follow-up, failure rates were obtained and pooled in the meta-analysis according to the follow-up. The overall failure rate was only computed for studies that had a follow-up of five years. In addition to the general rate, two subgroups were created for comparing analysis: (i) direct resin composite (DRC) against indirect resin composite (IRC) and (ii) DRC against direct inlay/onlay (DIO) based on five years of follow-up. A final analysis was performed comparing the clinical performance of the DRC against the IRC in molars and premolars at three-year follow-up. A fixed effects model was employed. Heterogeneity was assessed using the Inconsistency Index (I2) and the relative risk was also calculated (p < 0.05). The I2 describes the percentages of total variation across studies that are due to heterogeneity rather than chance.33 When necessary, sensitivity analysis and subgroup analysis were used. 3. Results 3.1 Characteristics of included articles After the database screening and removal of duplicates, 912 studies were identified (Fig. 1). After title and abstract screening, 20 studies were selected by eligibility. Two more studies were added following a manual search of the references of these 20 studies. Among them, 13 were excluded due to the following reasons: (1) no access to article,-34-36 (2) overlapping data,37-41 (3) no relation to the topic, 41, 42 (4) no RCTs,43-45 (5) no comparison direct with indirect resin composite restorations.

46

There was only one paper in an idiom we did

not have anyone to translate (Chinese). However, the paper’s abstract was in English, with enough information for us to identify that it did not fit the eligibility criteria to be included in the systematic review.

The characteristics of the nine studies selected for the qualitative data analysis are listed in Table 2. All studies were developed at universities with follow-ups ranging from two to 11 years. The minimum number of enrolled patients was 28 and the maximum 157. The ages were quite heterogeneous. Four studies used a paired-tooth design, in one study the subjects received only one type of restoration (intervention or control) and in another one patients received five restorations. However, in three studies it was unclear how the study was conducted. The United States Public Health Service (USPHS) and the United States Public Health Service – Modified (USPHS-M) indexes were employed for almost all of the studies. Only one used a qualitative analysis itself. Six studies compared the performance of DRC restorations with IRC and three compared the performance of DRC with DIO. All cavities were Class I or Class II, and in six studies cavities were medium-to-large-size Class II restorations, and may have had three or more surfaces involved, including one or more cusps. With regard to tooth type (molars and premolars), in studies in which there was a comparison of DRC with IRC restorations, only two studies used both teeth. When the intervention was DIO restorations, only one of them described the number of teeth used for each group and the type of restoration they received. Despite the large variety of composite brands used, we had difficulty describing them. Resin composites were divided into high filler load (>60% vol.) and filler load (<60% vol.)6,

10

In studies that compared DRC

restorations with IRC, resin composites with high filler load were employed for IRC and filler load <60% vol. for DRC. In three studies in which the intervention was DIO, both groups, control and intervention, used resin composite with filler load <60% vol. With regard to patient-related variables such as caries risk and bruxism, only three studies reported such analysis.1,

13, 52

The most common general

failures reported were (i) DRC: fracture of restoration, anatomical form, tooth fracture and marginal adaptation; (ii) IRC: marginal discolouration and marginal adaptation, fractures and debonding of restoration; and (iii) DIO: secondary caries.

3.2 Assessment of the risk of bias An assessment of the risk of bias of the studies selected is presented in Fig. 2. According to the predetermined key domains, three studies were considered ‘high’ risk of bias,4,

5, 51

five were classified as ‘low’1,

13, 48, 52, 54

and one as

‘unclear’3 risk of bias. All studies with ‘low’ risk of bias were considered for the meta-analysis, and after susceptibility testing was conducted and there were no significantly different results with or without the inclusion of the ‘unclear’ risk of bias, this was also included in the meta-analysis. In summary, six studies met the best requirement features for quantitative analysis. 3.3 Synthesis of the results: meta-analysis Following the guidance of Cochrane Handbook 5.0.2, after conducting a general meta-analysis, there was no statistically significant difference for the outcomes studied, regardless of the inclusion of ‘unclear’ risk of bias (p> 0.05). Fig. 3 is the meta-analysis for the overall failure rate of all studies with five-year follow up.1,

3, 13, 47, 48

These studies showed low heterogeneity (I2 = 4.09%) and a

relative risk of 1.494 [0.893 – 2.500] for indirect restorations. However, the pooled meta-analysis showed no statistically significant difference in clinical longevity for direct and indirect resin composite restorations (p = 0.126). Fig. 4A shows the subgroup analysis comparing (i) DRC against IRC at five-year follow-up. The relative risk was 1.278 [0.663–2.465] in relation to IRC, with no statistically significant difference between the two groups (p = 0.464) and low heterogeneity (I2 = 35.44%). Fig. 4B shows the subgroup analysis comparing (ii) DRC with DIO with five-year follow-up. There was no statistical difference between the groups, with a relative risk of 1.915 [0.837–4.385] (p = 0.124) in relation to DIO and without heterogeneity between groups (I2 = 0.00%). A pooled meta-analysis is presented in Fig. 5, which compares molars and premolars restored with DRC and IRC with three-year follow-up. The heterogeneity was low for both molars (I2 = 25.03%) and premolars (I2 = 0.00%). The overall relative risk was 0.716 [0.177–2.888], without statistical difference (p = 0.638).

4. Discussion Systematic review and meta-analysis studies are complementary to RCTs, since they help summarize the current knowledge available in health with the common goal of seeking solid scientific evidence for use in clinical practice.46,47 To reduce the individual risk of bias of RCTs included in an SR it is necessary for them to be evaluated for their methodological quality. The current recommendation of Cochrane Handbook 5.0.2 (www.handbook.cochrane.org) for assessing the risk of bias in clinical trials is the use of a domain-based tool, i.e. a critical assessment is made separately for different aspects of the risk of bias.38 However, the assessment of the overall risk of bias involves the consideration of the relative importance of the different areas according to the peculiarities of each SR, thereby helping the authors decide on which domains they consider to be key.36 This tool recommends that only studies with ‘low’ risk of bias in all key domains must be considered in meta-analysis, because they synthesize the best evidence for clinical practice. Although including only RCTs is an important inclusion criterion, this decision usually results in a limited number of studies and, consequently, of restorations that can be included in a study, posing the possibility of underpowering the study’s outcomes. Despite the fact that, over the time, this study showed that there were no statistical differences in clinical longevity between direct and indirect resin composite restorations, it is important to highlight that all nine studies included in this metaanalysis performed sample size calculation. Therefore, they had sufficient power to detect significant differences between groups if those differences were present. In addition, the effect sizes of all individual studies were analyzed in a pooled meta-analysis, showing no statistically significant difference in clinical longevity for direct and indirect resin composite restorations. For this reason, the authors believe that the number of studies (n = 9) and restorations (n = 899) included did not underpower the study’s results and conclusions. The randomization is important in clinical studies to allow an unpredictable result and prevent selection bias, which can lead to overestimation of the effect of intervention according to Cochrane Handbook 5.0.2. Unfortunately, a recurring problem encountered in the RCTs included in this SR was the lack of clarity in describing the sequence generation. Another difficulty was in relation to the

allocation concealment. Authors rarely described the method used to conceal the allocation sequence in sufficient detail to determine whether intervention allocations could have been foreseen in advance of, or during, enrolment. To clarify these two key domains, it was necessary to contact, through e-mail, the majority of the authors,1,

4, 5, 13, 48, 51

showing that this information is rarely

described in the methodology of the RCTs. Although blinding of participants, operators and outcome assessors is very important to avoid bias of performance and detection, there are intrinsic technical differences between direct, direct inlay/only and indirect resin composite restorations that make it impossible to blind operators and patients (Fennis et al.48). Assessor blinding is also a major challenge in such studies. Only three studies included in our systematic review (SR) indicated they blinded the assessors. They asserted the assessors did not have previous knowledge about the type of restoration or the materials used in the restorations they were about to assess. Even though not explicitly mentioned in their papers, it is fair to assume the other six studies included in our SR also did not provide assessors with up-front information about the type of the restorations they were about to assess. However, not knowing in advance which type of restorations they were going to assess does not prevent assessors from knowing immediately, by looking at the restorations, which type they are. In the case of indirect restorations (made through a laboratory), only professionals with limited training or experience are unable to distinct them from a direct resin composite restoration when they evaluate one; there are clear differences in clinical appearance,

for

instance,

in

terms

of

morphology,

polishing,

and

presence/absence of line of cementation. The differences between direct resin composite and direct inlay/onlay restorations, however, are more subtle, especially when they have been in the mouth for some time; for inlay/onlay the cement line is not the same as for ceramic restorations and their anatomic form is rather similar to direct restorations. Given assessor blinding is a source of bias almost impossible to be completely eliminated in any study of this kind, we have not considered it as a key domain (point) of bias in our selection criteria. The report of incomplete outcome data is important because it describes the number of participants lost in each intervention group compared with the

total number of randomized participants. If caution isn’t taken with regard to the report of incomplete outcome data, there may be the risk of creating attrition/exclusion bias due to the disproportion of participants in one group compared to another. Fortunately, in our SR only one study did not report how many participants were lost.51 Loss of participants can be due to patients moving out of the area or a loss of interest in returning for recall appointments. This can be justified because clinical trials take a lot of time to be concluded and specifically in dental trials the benefits are more for researchers than for patients.47 Selective outcome reporting was a key domain because through this domain we can examine whether the reports were free of suggestion of selective outcome reporting and the outcomes were according to the preexisting protocol. In this SR the majority of studies were free of such bias (Fig. 2). With regard to the restorative technique employed in the intervention groups of RCTs included in our SR, indirect composite resin restorations were named in different ways as ‘inlays’13 or ‘indirect resins’3, 4, 5, 48, 52 when they were made in a dental lab and ‘direct inlay/onlay’1, 47 or only as ‘indirect restorations’51 when they were made using conventional direct resin composite placed into the tooth cavity previously isolated and posteriorly placed for secondary curing. To avoid confounding variables, the authors subdivided indirect resin composite restorations into two subgroups called ‘Indirect resin composite’ and ‘Direct inlay/onlay’ to analyse these subgroups individually and when pooled in a single group in the meta-analysis. As there was no statistically significant difference in clinical performance between them (Figs. 3, 4A and 4B) and there is no unanimity among the classifications, we suggest considering terms such as ‘Indirect resin composite restorations’ or just ‘Resin composite inlay/onlay’ and the technique could be better described in the material and method section. Moreover, the resin composites employed in indirect techniques of these RCTs have not always been indicated exclusively for dental laboratories;13 all received secondary curing and were cemented with an adhesive cement agent. Unification of this terminology could make access to the literature available easier for researchers, clinicians and students.

In In vitro studies,23-25 secondary curing is seen as one of the advantages of indirect resin composite restorations; however, it does not seem clinically applicable for improving the longevity of these materials. This can be justified because clinically patient-related variables such as bruxism and caries risk8, 10, 17, 19, , 55

can significantly impact on the long-term success of restorations.

Unfortunately, in our SR it was not possible to perform meta-analysis for these factors because it was considered in three of the RCTs1, 13, 40 and in only one of them52 were the failures associated with individual patient risk. In addition to the factors described above, other factors such as material and tooth-related variables have also been identified as key factors for the clinical success of resin composite restorations.10, 11, 17, 53, 54 With regard to resin composite materials, due to the variety among them, we could not perform a meta-analysis but did conduct a descriptive analysis in which it was shown that both high filler load (>60% vol) and filler load (<60% vol.) resin composites were used regardless of technique. One limitation of this systematic review is that some of the materials used in the studies that fulfilled the eligibility criteria are not on the market anymore. However, independently of the materials and techniques employed, it did not influence the longevity of restorations, once they were considered clinically acceptable, and the number of failures was considered low. On the other hand, considering the improvements on the more recent restorative materials, it is suggested that new longitudinal studies comparing direct and indirect resin composite restorations using materials currently in the market should be implemented. The tooth-related variable, only two trials with 'low' risk of bias employed both types of teeth, molars and premolars, and a meta-analysis was possible only with three-year follow-up. Overall, there was no statistically significant difference (p = 0.638) between them. Although this meta-analysis only included two trials with a short follow-up, these results are in accordance with Pallesen and Van Dijken,56 who also found no difference in longevity between molars and premolars restored with DRC over a 30-year follow-up.

Failures

were

analysed descriptively and do not appear in the meta-analysis due to variable factors reported in each of the three groups of restorations. For DRC restorations, fracture of the restoration and changes in anatomical form were

most frequently reported, followed by failures in marginal adaptation of restorations. In longitudinal studies that evaluated the performance of DRC,8, 10, 11, 56

these results were also found to be the most prevalent failures followed by

dental caries. However, in our SR, no study has reported decay as the most prevalent factor of failure in DRC restorations, and only two trials1,

47

reported

failures in DIO due to dental caries. For IRC, marginal adaptation was the most frequently reported factor of failures. Both in vivo6,

57

and in vitro

26

studies,

crevices at the margins and marginal discolouration were prevalent factors of failures. Making clinical decisions about indirect resin composite restorations is not always easy, especially when the professional is faced with issues related to increased wear of the remaining tooth structure, time and cost without having the clear advantage of longevity compared to direct resin composite restorations. One example is the research by Laegreid et. al.,14 in which, in a questionnaire answered by 270 dentists from Bergen, Norway, when asked about which type of restorative material they preferred to employ in largely destroyed posterior teeth, with losses of up to two cusps, 65.1% answered that they rarely or never employed indirect resin composite restoration while 83.5% often used direct resin composite for such clinical situations. Perhaps the predilection for direct composite resin restorations is due to very favourable results regarding the clinical performance of these restorations in posterior teeth shown by clinical trials.7, 11, 56, 58 These findings are consistent with those found in our meta-analysis, which showed no statistically significant difference in the clinical performance of the DRC when compared to IRC at five-year follow-up (p = 0.126). 5. Conclusion Based on the results of this systematic review and meta-analysis, there is evidence of no difference in terms of clinical longevity between direct and indirect resin composite restorations. This conclusion remains valid even when the type of restored tooth is taken into account. Therefore, it seems more reasonable to suggest that direct restorations should be given preference to

indirect restorations in many situations, since the former require less effort and cost. 6. Conflict of interest None. Acknowledgments This study was conducted as part of the master degree of Ana Maria Antonelli da Veiga under the supervision of Professor Lucianne Cople Maia. The authors of this study would like to thank the following authors who kindly provided information not available in their full texts: Jan Van Djiken, Ulla Pallesen, David Bartlett and Willem Fennis.

REFERENCES 1. Van Dijken JW. Direct resin composite inlays/onlays: an 11 year follow-up. Journal of Dentistry 2000;28:299-306. 2. Spreafico RC, Krejci N, Dietschi D. Clinical performance and marginal adaptation of class II direct and semidirect composite restorations over 3.5 years in vivo. Journal of Dentistry 2005;33:499-507. 3. Cetin AR, Unlu N, Cobanoglu N. A Five-Year Clinical Evaluation of Direct Nanofilled and Indirect Composite Resin Restorations in Posterior Teeth. Operative Dentistry 2013;38:E31-E41. 4. Manhart J, Neuerer P, Scheibenbogen-Fuchsbrunner A, Hickel R. Three-year clinical evaluation of direct and indirect composite restorations in posterior teeth. Journal of Prosthetic Dentistry [Internet] 2000;84:[289-96 pp.]. Available from: http://onlinelibrary.wiley.com/o/cochrane/clcentral/articles/054/CN00326054/frame.html. 5. Ozakar-Ilday N, Zorba YO, Yildiz M, Erdem V, Seven N, Demirbuga S. Three-year clinical performance of two indirect composite inlays compared to direct composite restorations. Medicina oral, patología oral y cirugía bucal 2013;18:e521-8. 6. Huth KC, Chen HY, Mehl A, Hickel R, Manhart J. Clinical study of indirect composite resin inlays in posterior stress-bearing cavities placed by dental students: results after 4 years. Journal of Dentistry 2011;39:478-88. 7. Manhart J, Chen HY, Hickel R. Three-year results of a randomized controlled clinical trial of the posterior composite QuiXfil in class I and II cavities. Clinical Oral Investigations 2009;13:301-7. 8. Da Rosa Rodolpho PA, Donassollo TA, Cenci MS, Loguercio AD, Moraes RR, Bronkhorst EM, et al. 22-Year clinical evaluation of the performance of two posterior composites with different filler characteristics. Dental Materials 2011;27:955-63. 9. Demarco FF, Correa MB, Cenci MS, Moraes RR, Opdam NJ. Longevity of posterior composite restorations: not only a matter of materials. Dental Materials 2012;28:87-101. 10. Opdam NJ, van de Sande FH, Bronkhorst E, Cenci MS, Bottenberg P, Pallesen U, et al. Longevity of posterior composite restorations: a systematic review and meta-analysis. Journal of Dental Research. 2014;93:943-9. 11. Frankenberger R, Reinelt C, Kraemer N. Nanohybrid vs. fine hybrid composite in extended class II cavities: 8-year results. Clinical Oral Investigations 2014;18:125-37. 12. Magne P. Conservative restoration of compromised posterior teeth with direct composites: a 7-year report. Practical Periodontics and Aesthetic Dentistry 2000;12:747-9. 13. Pallesen U, Qvist V. Composite resin fillings and inlays. An 11-year evaluation. Clinical Oral Investigations 2003;7:71-9. 14. Laegreid T, Gjerdet NR, Johansson A, Johansson AK. Clinical decision making on extensive molar restorations. Operative Dentistry 2014;39:E231-40. 15. Kweon HJ, Ferracane J, Kang K, Dhont J, Lee IB. Spatio-temporal analysis of shrinkage vectors during photo-polymerization of composite. Dental Materials 2013;29:1236-43.

16. Boaro LC, Goncalves F, Guimaraes TC, Ferracane JL, Pfeifer CS, Braga RR. Sorption, solubility, shrinkage and mechanical properties of "low-shrinkage" commercial resin composites. Dental Materials 2013;29:398-404. 17. Nedeljkovic I, Teughels W, De Munck J, Van Meerbeek B, Van Landuyt KL. Is secondary caries with composites a material-based problem? Dental Materials 2015;31(11):e247-77. 18. Schneider LF, Cavalcante LM, Silikas N. Shrinkage Stresses Generated during Resin-Composite Applications: A Review. Journal of Dental Biomechanics 2010;2010. 19. van Dijken JW, Lindberg A. A 15-year randomized controlled study of a reduced shrinkage stress resin composite. Dental Materials 2015;31(9):1150-8. 20. Van Ende A, Mine A, De Munck J, Poitevin A, Van Meerbeek B. Bonding of low-shrinking composites in high C-factor cavities. Journal of Dentistry 2012;40(4):295-303. 21. Dejak B, Młotkowski A. A comparison of stresses in molar teeth restored with inlays and direct restorations, including polymerization shrinkage of composite resin and tooth loading during mastication. Dental Materials 2015;31:e77-e87. 22. Bagis YH, Rueggeberg FA. Effect of post-cure temperature and heat duration on monomer conversion of photo-activated dental resin composite. Dental Materials 1997;13:228-32. 23. Asmussen E, Peutzfeldt A. Mechanical properties of heat treated restorative resins for use in the inlay/onlay technique. Scandinavian Journal of Dental Research 1990;98(6):564-7. 24. Khan AM, Satou N, Shintani H, Taira M, Wakasa K, Yamaki M. Effects of post-curing by heat on the mechanical properties of visible-light cured inlay composites. Journal of Oral Rehabilitation 1993;20:605-14. 25. Coelho-de-Souza FH, Camacho GB, Demarco FF, Powers JM. Fracture resistance and gap formation of MOD restorations: Influence of restorative technique, bevel preparation and water storage. Operative Dentistry 2008;33:37-43. 26. Aggarwal V, Logani A, Jain V, Shah N. Effect of cyclic loading on marginal adaptation and bond strength in direct vs indirect Class II MO composite restorations. Operative Dentistry 2008;33:587-92. 27. Duquia RCS, Osinaga PWR, Demarco FF, Habekost LV, Conceicao EN. Cervical microleakage in MOD restorations: In vitro comparison of indirect and direct composite. Operative Dentistry 2006;31:682-7. 28. de Paula AB, Duque C, Correr-Sobrinho L, Puppin-Rontani RM. Effect of restorative technique and thermal/mechanical treatment on marginal adaptation and compressive strength of esthetic restorations. Operative Dentistry 2008;33:434-40. 29. Kuijs RH, Fennis WMM, Kreulen CM, Roeters FJM, Verdonschot N, Creugers NHJ. A comparison of fatigue resistance of three materials for cuspreplacing adhesive restorations. Journal of Dentistry 2006;34:19-25. 30. Thordrup M, Isidor F, Horsted-Bindslev P. A 5-year clinical study of indirect and direct resin composite and ceramic inlays. Quintessence International 2001;32:199-205. 31. Moher D, Shamseer L, Clarke M, Ghersi D, Liberati A, Petticrew M, et al. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Systematic Review 2015;4:1.

32. Maia LC, Antonio AG. Systematic reviews in dental research. A guideline. The Journal of Clinical Pediatric Dentistry 2012;37:117-24. 33. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. British Medical Journal 2003;327(7414):557-60. 34.Conti GF, Ragni G, Foglio Bonda PL, Centini S, Micheletti P, Sapelli PL. Direct inlays and onlays in composite material. Dental Cadmos 1987;55:83-8. Lowe E. Restoring form, function, and aesthetics in Class II direct composite restorations. Dentistry Today 1998;17:72-4. 36. Trushkowsky R. Direct and indirect composites for the restoration of decimated dentition in elderly patients. Dentistry Today 2000;19:92-9. 37. Cetin AR, Unlu N. Clinical wear rate of direct and indirect posterior composite resin restorations. International Journal of Periodontics Restorative Dentistry 2012;32:e87-94. 38. Scheibenbogen-Fuchsbrunner A, Manhart J, Kremers L, Kunzelmann KH, Hickel R. Two-year clinical evaluation of direct and indirect composite restorations in posterior teeth. Journal of Prosthetic Dentistry 1999;82:391-7. 39. Wassell RW, Walls AW, McCabe JF. Direct composite inlays versus conventional composite restorations: three-year clinical results. British Dental Journal 1995;179:343-9. 40. Pallesen U, Qvist V. Clinical evaluation of resin fillings and inlays: 8-year report. Journal of Dental Research 1998;77:914(Abstract nº 2254). 41. Freilich MA, Goldberg AJ, Gilpatrick RO, Simonsen RJ. Direct and indirect evaluation of posterior composite restorations at three years. Dental Materials 1992;8:60-4. 42. Kreulen CM, van Amerongen WE, Gruythuysen RJ, Akerboom HB, Borgmeijer PJ. Evaluation of treatment times for Class II resin composite inlays. Journal of Dentistry for Children 1992;59:333-7. 43. Spreafico RC, Krejci N, Dietschi D. Clinical performance and marginal adaptation of class II direct and semidirect composite restorations over 3.5 years in vivo. Journal of Dentistry 2005;33:499-507. 44. Zhang LL, Liu TJ, Fang MX, Li W. Clinical evaluation of composite inlays in defective molars. Zhonghua Kou Qiang Yi Xue Za Zhi 2008;43:44-7. 45. Harb-Kadiri J, Khairallah W. [Restoration of cusped teeth using composite resins: direct or indirect method?]. Le-Chirurgien-dentiste de France 1986;56:43-7. 46. Thordrup M, Isidor F, Horsted-Bindslev P. A 5-year clinical study of indirect and direct resin composite and ceramic inlays. Quintessence International 2001;32:199-205. 47. Wassell RW, Walls AW, McCabe JF. Direct composite inlays versus conventional composite restorations: 5-year follow-up. Journal of Dentistry 2000;28:375-82. 48. Fennis WM, Kuijs RH, Roeters FJ, Creugers NH, Kreulen CM. Randomized Control Trial of Composite Cuspal Restorations: Five-year Results. Journal of Dental Research 2014;93:36-41. 49. Berwanger O, Suzumura EA, Buehler AM, Oliveira JB. How to Critically Assess Systematic Reviews and Meta-Analyses? Revista Brasileira de Terapia Intensiva 2007;19:475-480.

50. Carvalho APV, SilvaI V, GrandeIl AJ. Avaliação do risco de viés de ensaios clínicos randomizados pela ferramenta da colaboração Cochrane. Diagnóstico e Tratamento 2013;18:38-44. 51. Wendt Jr SL, Ziemiecki TL, Leinfelder KF. Proximal wear rates by tooth position of resin composite restorations. Journal of Dentistry 1996;24:33-9. 52. Bartlett D, Sundaram G. An up to 3-year randomized clinical study comparing indirect and direct resin composites used to restore worn posterior teeth. International Journal Prosthodontic 2006;19:613-7. 53. Signore A, Benedicenti S, Covani U, Ravera G. A 4- to 6-year retrospective clinical study of cracked teeth restored with bonded indirect resin composite onlays. International Journal Prosthodontic 2007;20:609-16. 54. Van Dijken JW. Durability of resin composite restorations in high C-factor cavities: a 12-year follow-up. Journal of Dentistry 2010;38:469-74. 55. Van de Sande FH, Opdam NJ, Rodolpho PA, Correa MB, Demarco FF, Cenci MS. Patient risk factors' influence on survival of posterior composites. Journal of Dental Research 2013;92:78s-83s. 56. Pallesen U, van Dijken JW. A randomized controlled 30 years follow up of three conventional resin composites in Class II restorations. Dental Materials 2015;31:1232-44. 57. Dukic W, Dukic OL, Milardovic S, Delija B. Clinical evaluation of indirect composite restorations at baseline and 36 months after placement. Operative Dentistry 2010;35:156-64. 58. Opdam NJ, Bronkhorst EM, Loomans BA, Huysmans MC. 12-year survival of composite vs. amalgam restorations. Journal of Dental Research 2010;89:1063-7.

FIGURE CAPTIONS Fig. 1: Flow diagram of literature search. Fig. 2: Summary of the risk of bias assessment according to the Cochrane Collaboration tool. The underlined authors provided extra information by e-mail to enable assessment of the risk of bias. Fig. 3: Forest plot of the risk of failures in permanent posterior restorations performed with direct vs. indirect resin composite restorations with five-year follow-up (p = 0.126). Fig. 4A: Forest plot of the risk of failures in permanent posterior restorations performed with subgroup analysis comparing DRC vs. IRC at five-year followup, showing no differences between the two groups (p = 0.464). Fig. 4B: Forest plot of the risk of failures in permanent posterior restorations performed with subgroup analysis comparing DRC vs. DIO at five-year followup, showing no differences between the two groups (p = 0.124). Fig. 5: Forest plot of the risk of failures in permanent posterior restorations performed comparing DRC vs. IRC in molars and premolars with three-year follow-up. There was no statistical difference in a global analysis between the groups (p = 0.638).

Table 1 - Eletronic databases and research strategies (18/ AUG/ 2015) PUBMED #1Molar [MH] OR Molar [TIAB] OR Molars [TIAB] OR Bicuspid [MH] OR Bicuspid [TIAB] OR Premolar* [TIAB] OR Dentition, Permanent [MH] OR Permanent Dentition [TIAB] OR Tooth erosion [MH] OR Tooth erosion [TIAB] OR Erosive tooth wear [TIAB] OR Posterior teeth [TIAB] OR Posterior tooth [TIAB] OR Dental caries [MH] OR Dental caries [TIAB] OR Dental decay [TIAB] OR Class I [TIAB] OR Class II [TIAB] #2 Indirect composite resin* [TIAB] OR Indirect resin* [TIAB] OR Indirect posterior composite [TIAB] OR Indirect restoration* [TIAB] OR Indirect resin composite [TIAB] OR Indirect composite [TIAB] OR Indirect [TIAB] OR Inlays [MH] OR Inlays [TIAB] OR Inlay [TIAB] OR Overlay* [TIAB] OR Onlay* [TIAB] #3 (Composite resins [MH] OR Composite resins [TIAB] OR Composite resin [TIAB] OR Composite restorative materials [TIAB] OR Resin composit* [TIAB] OR Resin composite restoration* [TIAB] OR Posterior composite restoration*[TIAB] OR Resin-based composite*[TIAB] OR Tooth-colored restorat* [TIAB] OR Dental composite* [TIAB] OR Composite restoration* [TIAB]) Direct composite resin* [TIAB] OR Direct resin composite restoration* [TIAB] OR Direct composite restorations [TIAB] OR Direct posterior composite* [TIAB] OR Direct restoration* [TIAB] OR Direct resin composite [TIAB] OR Direct composite [TIAB] OR Direct [TIAB]) #1 AND #2 AND #3 SCOPUS #1 TITLE-ABS-KEY (“Molar” OR “Molars” OR “Bicuspid” OR “Premolar” OR “Premolars” OR “Permanent Dentition” OR “Tooth erosion” OR “Erosive tooth wear” OR “Posterior teeth” OR “Posterior tooth” OR “Dental caries” OR “Dental decay” OR “Class I” OR “Class II”) #2 TITLE-ABS-KEY (“Indirect composite resin” OR “Indirect composite resins” OR “Indirect resin” OR “Indirect resins” OR “Indirect posterior composite” OR “Indirect restoration” OR “Indirect restorations” OR “Indirect resin composite” OR “Indirect composite” OR “Indirect” OR “Inlay” OR “Inlays” OR “Overlay” OR “Overlays” OR “Onlay” OR “Onlays”) #3 TITLE-ABS-KEY (“Composite resins” OR “Composite resin” OR “Composite restorative materials” OR “Resin composite” OR “Resin composites” OR “Resin composite restoration” OR “Resin composite restorations” OR “Posterior composite restoration” OR “Posterior composite restorations” OR “Resin-based composite” OR “Resin-based composites” OR “Tooth-colored restoration” OR “Tooth-colored restorations” OR “Dental composite” OR “Dental composites” OR “Composite restoration” OR “Composite restorations”) AND (“Direct composite resin” OR “Direct composite resins” OR “Direct resin composite restoration” OR “Direct resin composite restorations” OR “Direct composite restorations” OR “Direct posterior composite” OR “Direct posterior composites” OR “Direct restoration” OR “Direct restorations” OR “Direct resin composite” OR “Direct composite” OR “Direct”) #1 AND #2 AND #3 WEB OF SCIENCE #1 Topic : (“Molar” OR “Molars” OR “Bicuspid” OR “Premolar” OR “Premolars” OR “Permanent Dentition” OR “Tooth erosion” OR “Erosive tooth wear” OR “Posterior teeth” OR “Posterior tooth” OR “Dental caries” OR “Dental decay” OR “Class I” OR “Class II”) #2 Topic: (“Indirect composite resin” OR “Indirect composite resins” OR “Indirect resin” OR “Indirect resins” OR “Indirect posterior composite” OR “Indirect restoration” OR “Indirect restorations” OR “Indirect resin composite” OR “Indirect composite” OR “Indirect” OR “Inlay” OR “Inlays” OR “Overlay” OR “Overlays” OR “Onlay” OR “Onlays”) #3 Topic: (“Composite resins” OR “Composite resin” OR “Composite restorative materials” OR “Resin composite” OR “Resin composites” OR “Resin composite restoration” OR “Resin composite restorations” OR “Posterior composite restoration” OR “Posterior composite restorations” OR “Resin-based composite” OR “Resin-based composites” OR “Tooth-colored restoration” OR “Toothcolored restorations” OR “Dental composite” OR “Dental composites” OR “Composite restoration” OR “Composite restorations”) AND (“Direct composite resin” OR “Direct composite resins” OR “Direct resin composite restoration” OR “Direct resin composite restorations” OR “Direct composite restorations” OR “Direct posterior composite” OR

“Direct posterior composites” OR “Direct restoration” OR “Direct restorations” OR “Direct resin composite” OR “Direct composite” OR “Direct”) COCHRANE LIBRARY ID Search Hits #1 MeSH descriptor: [Molar] explode all trees #2 molar #3 molars #4 #1 or #2 or #3 #5 MeSH descriptor: [Bicuspid] explode all trees #6 bicuspid #7 premolar* #8 #5 or #6 or #7 #9 MeSH descriptor: [Dentition, Permanent] explode all trees #10 dentition, permanent #11 permanent dentition #12 #9 or #10 or #11 #13 MeSH descriptor: [Tooth Erosion] explode all trees #14 tooth erosion #15 erosive tooth wear #16 #13 or #14 or #15 #17 posterior tooth or posterior teeth #18 MeSH descriptor: [Dental Caries] explode all trees #19 dental caries #20 dental decay #21 #18 or #19 or #20 #22 "class I" or "class II" #23 #4 or #8 or #12 or #16 or #17 or #21 or #22 #24 Indirect composite resin* or Indirect resin* or Indirect restoration* or Indirect resin composite or Indirect composite or Indirect #25 Indirect posterior composite #26 #24 or #25 #27 MeSH descriptor: [Inlays] explode all trees #28 inlays #29 inlay #30 overlay or overlays or onlay* #31 #27 or #28 or #29 or #30 #32 #26 or #31 #33 MeSH descriptor: [Composite Resins] explode all trees #34 composite resins #35 composite resin or Resin composit* or Resin composite restoration* or Resinbased composite* or Tooth-colored restorat* or Dental composite* or Composite restoration* #36 Composite restorative materials #37 #33 or #34 or #35 or #36 #38 Direct composite resin* or Direct resin composite restoration* or Direct composite restorations or Direct posterior composite* or Direct composite or Direct #39 #37 and #38 #40 #23 and #32 and #39 LILACS and BBO #1 TW:(MH:molar OR diente molar OR dente molar OR molars OR dientes molares OR dentes molares OR MH:bicuspid OR diente premolar OR dente pré-molar OR dentición permanente OR dentição permanente OR MH: dentition, permanente OR MH:tooth erosion OR erosión de los dientes OR erosão dentária OR MH:dental caries OR cáries dental OR cárie dentária OR caries decay OR posterior tooth OR diente posterior OR dente posterior OR posterior teeth OR dientes posteriores OR dentes posteriores OR class I OR clase I OR classe I OR class II OR clase II OR classe II) #2 TW:(MH:Composite resins OR Resinas compuestas OR Resinas compostas OR Composite resin OR Composite resins OR Composite restorative materials OR Resin composit$ OR Resina composta$ OR Resin composite restoration$ OR Posterior composite restoration$ OR Resin-based composite$ OR Tooth-colored restorat$ OR

Dental composite$ OR Compuesto dental$ OR Compósito dental$ OR Composite restoration$ OR Restauración de compusto$ OR Restauração de compósito$ OR Direct composite resin$ OR Direct resin composite restoration$ OR Direct composite restorations OR Direct restoration$ OR Restaración directa$ OR Restauração direta$ OR Direct composite OR Compuesto directo OR Compósito direto OR Direct OR Directa OR Direta) #1 AND #2 CLINICAL TRIALS.GOV Posterior teeth and Inlay and Direct composite resin SIGLE (Molar OR Molars OR Bicuspid OR Premolar OR Premolars OR “Class I” OR “Class II”) AND (“Indirect composite resin*” OR “Indirect resin*” OR “Indirect restoration*” OR Indirect OR Inlay OR Inlays OR Overlay OR Overlays OR Onlay OR Onlays) AND (“Composite resin*” OR “Resin composite*” OR “Resin composite restoration*” OR “Tooth-colored restoration*” OR “Dental composite*” OR “Composite restoration*”)

Table 2: Study and patients characteristics for qualitative analysis. Study/ Setting

Bartlett &Sundara m, 2006

Folowup (years)

3

Study design

Splitmouth

Criteria

USPHS

o

N of pati ents (% men )

Mean age/ range

29 (n.r.)

TW: 43 (25-62)

University

Cetin et al. 2013

o

N restorat ions (DRC /IRC or DIO*)

Tooth

Cavity

16 (16)

PM (DRC /IRC or DIO) 11 (11)

M (DRC /IRC or DIO) 5 (7)

Control: 39 (28-65)

13 (13)

6 (6)

7 (7)

Control: Extensive caries lesions at least 1 cusp missing

_

67 (41)

5

Splitmouth

USPHSM

22 (32)

23(20-28)

67 (41)

5

19 (1/patient) 138 (2/ patient)

Clinical examina tion

77 (80)

54,9 (3581)

92 (84)

University

Fennis et al. 2014 University

92 (84)

_

TW: multiple worn

o

N restorat ions/ dropout

Manufacturing procedure

Results

Conclusions

29 (2)

IRC: A developmental, light/heat-cured, microfilled resin composite material DRC: Heliomolar HB, Ivoclar Vivadent

TW: Lost: IRC 5 PM and DRC 3PM + 1M; Fractured: IRC 3PM + 1M and DRC 2PM + 1M Control: Lost: IRC 2 PM; Fractured: IRC 1 M

TW: high fracture rate for direct and indirect resin composite Control: both performed satisfactorily

Class I and II cavities (small to medium)

54 (0)

IRC: Estenia [E]; Tescera [TATL] DRC: Filtek Supreme XT [FS XT]; Tetric EvoCeram [TEC]; AELITE Aesthetic [AA]

Alfa score: Gingival adaptation and Retention: 100%; Color match: 100% FS XT, TEC, E; 95% AA, TATL; Marginal Integraty: 100% FS XT, 95% TEC, 90%TATL, 84% E, 82% AA; Surface texture: 95% E, TEC, TATL, 82% FS, AA; Marginal dislocoloration: 95% TEC, 87%FS XT, 73%E, 70%TALT, 64% AA.

Annual failure rates: DRC: 1.6% IRC: 2.5%. All acceptable.

Class II (Fracture of palatal or buccal cusp of upper PM)

157(18)

IRC: Estenia, Kuraray DRC: AP-X, Kuraray.

Failure rate: DRC: Fracture remaining cusp (37.5%) and cohesive failure restoration (25%). RCI: Dislodged restoration

Overall survival: DRC: 91.2% e IRC:83.2%. No significant difference

Manhart et al. 2000

3

n.r.

USPHSM

n.r

n.r.

43 (45)

17 (20)

3

n.r.

USPHSM

28 (21)

32

20 (40)

20 (40)

11

Splitmouth

USPHSM

8 (20)

35(19-64)

54 (81)

n.r

11

n.r

USPHSM

24 (16)

48(27-70)

34 (100)

20 (84)

13 (10)

RCI: Tetric [T], Vivadent; Blend-a-lux, Procter & Gamble [BL] e Pertachibrid Unifil, ESPE [PHU] RCD: Tetric [T], Vivadent; Blend-a-lux, Procter & Gamble [BL] e Pertachibrid Unifil, ESPE [PHU]

Alfa score: Surface texture: IRC 67% and DRC 33%; Marginal integrity: IRC 57% and DRC 40%; Integrity of the restoration: DRC 97% and IRC 77%, RCD (p=.023). .

IRC: Large class II DRC: Small to médium Class II cavities

49 (0)

IRC: Brilliant DI, Coltene [DI]; Tescera ATL, Bisco [T]. DRC: Valux Plus, 3M ESPE.

Annual overall survival: 93% [T], 86% [DI] and 67% RCD.

n.r

Large class II cavities

27(1)

IRC: Brilliant Dentin, Coltene [BD]; Estilux Posteirior, Kulzer [EP]; SR-Isosit [ISO]. DRC: BD e EP.

Failure rate: DRC 16% and IRC 17% (p>0.05). M: DRC 14% and IRC 20%; PM: DRC 5% e IRC 8%. Most common failures: DRC: Color match and minor fractures; IRC: wear of luting composite, marginal discoloration and match color.

No significant difference in the long-term survival Additional oven curing had only a minor influence on the fracture resistance and did not improve the wear resistance in IRC.

14 (16)

Large class II

96 (33) restorati ons

IOD: Briliant DI, Coltene. RCD: Fulfil, DeTray, Dentsply.

Failure rate: DIO 17.7% and DRC 27.3%. Main reasons for fail: fracture DIO 8.3% and DRC 12.1%, occlusal wear in contact áreas DIO 4.2% and DRC 6.1% abd secondary caries: DIO

The mechanical properties of the material apparently were not improved by secondary

_

University

Van Dijken, 2000 University

Overall survival: IRC: 93% and DRC: 87%. No significant difference between the techniques.

88 (28) rest.

University

Pallesen & Qvist, 2003

between the techniques.

Large Classe I, II cavities

University

OzakarIlday et al. 2013

(26.7%) and dilodged & cohesive restoration (20%).

IRC [T] Best result.

Wassell et al. 2000

5

Splitmouth

USPHS

19 (54)

29,6 (+10 years)

50 (20)

n.r

n.r

Class II

100 (35) restorati ons

University

Wendt et al.1996 University

IOD: Briliant, Coltene. RCD: Briliant, Coltene.

2

1/ patient

USPHS

n.r

n.r

50 (20)

n.r

n.r

Class II

n.r

DIO e DRC: P-30, P-50, (3M), Heliomolar (Vivadent),Clearfil (Kuraray) e Brilliant DI(Coltene).

4.2% and DRC 9.1%. Higher mechanical failures in M than PM.

cure.

Failure rate: DIO 17.4% and DRC 7.5%. Main reasons for fail: pressure sensitivity: 4 DIO, 1 DRC; periapical abscess 2 DIO, 2 DRC, fracture restoration 2 DIO, 1 DRC; fracture tooth 3 DIO, 1 DRC and secondary caries 1 DIO.

Direct inlays show no advantage over the direct placement restorations and have a trend to a higher failure rate.

DIO P-50 < proximal wear.

There were no differences in the amount of proximal wear with regard to tooth position.

DRC: Direct resin composite, DIO: Direct Inlay/Onlay and IRC: Indirect resin composite; USPHS: United States Public Health Service; USPHS-M: United States Public Health Service modified; PM: Premolars; M: Molars; RC: resin composite; TW: Tooth wear.

Identification

Fig 1

PubMed (n=825)

Scopus (n=247)

Web of Science (n=132)

Cochrane Library (n=89)

Lilacs e BBO (n=4)

Clinical Trials.gov (n=1)

SIGLE (n=1)

Screening

Records identified through database searching (n=1300)

Records after duplicates removed (n=912)

Records excluded after title and abstract screen (n=892)

Elegibility

Full-text articles assessed by database searching (n=20) Hand searching (n=2)

Full-text articles assessed for elegibility (n=22)

Included

Articles excluded, with reasons (n=13) No access of article (n=3)34-36 Overlapping data (n=4)37-40 No related to the topic (n=2)41,42 No RCTs (n=3)43-45 No compared direct with indirect composite resin restorations (n=1)46 Studies included in qualitative synthesis (n=9)

Studies included in quantitative synthesis (meta-analysis) (n=6)

Figure

Figure

Figure

Figure

Figure