Page 1
347
Abstract: The purpose of this systematic review
was to quantitatively find out whether resin-modified
glass-ionomers (RM-GIC), in comparison to fluoride-
containing composite resin and composite resin without
fluoride, are associated with a more effective reduction
of demineralization in hard tooth tissues under caries
challenge. Five databases were systematically searched
on clinical trials up to 6 April 2009. Article inclusion
criteria: titles/abstracts relevant in answering the
review question, published in English, two-arm
(prospective) longitudinal trial; Exclusion criteria: not
all included subjects accounted for at the end of the trial;
subjects of both groups not followed up the same way;
no randomized, quasi-randomized controlled study
design for in situ and clinical trials; contains no
computable continuous data. Quality assessment of
the accepted in situ and clinical trials was performed.
Data were extracted in the form of datasets, containing
numbers of evaluated samples and mean result with
standard deviation for both groups. Fifteen articles
were selected for review. Two lacked computable data
and were excluded; nine laboratory trials, three
randomized in situ trials and one randomized control
trial were accepted. From these, 97 continuous datasets
were extracted. The evidence suggests that RM-GIC
is associated with a higher reduction of demineralization
in adjacent hard tooth tissue than composite resin
without fluoride. No difference was found when RM-
GIC was compared with fluoride-containing composite
resin. RM-GIC showed efficacy in reducing
demineralization. However, the internal validity of the
current evidence is limited and further high-quality
trials are needed. (J Oral Sci 52, 347-357, 2010)
Keywords: demineralization; resin-modified glass-
ionomer; composite resin; systematic
review.
Introduction
An important part of caries management is encouraging
hard tooth tissue remineralization (1). Ten Cate and van
Duinen have shown, in situ, a hyper-remineralization
effect in demineralized tooth tissues bordering glass-
ionomer cement (GIC) type restorations (2). The significant
remineralizing potential of GIC has been ascribed to the
release of fluoride ions, facilitated by a hydrophilic
environment (3). The remineralizing effect has been
explained clinically (4) on the basis of its fluoride release
into saliva, leading to an increase in the salivary fluoride
content from 0.04 to 0.30 ppm after one year (5). However,
the actual amount of fluoride in saliva required to have any
effect on the mineral content of teeth is still unclear (6).
Two recent systematic reviews with meta-analyses of
RCTs have confirmed the caries-preventive effect of GIC
on restoration margins (7) and on pits and fissures sealed
with GIC (8). These findings have been established for
conventional glass-ionomers (C-GIC) which set through
an acid-base reaction between fluoroaluminosilicate glass
powder and polyalkenoic acid liquid. However, C-GICs
remain sensitive to water uptake and are lost in the first
hours after setting, which led to the development of ‘resin-
modified’ GICs (RM-GIC). In the set material, approxi-
mately 10% of RM-GIC is resin, usually hydroxyethyl-
Journal of Oral Science, Vol. 52, No. 3, 347-357, 2010
Correspondence to Dr. Steffen Mickenautsch, Division of Public
Oral Health, University of the Witwatersrand, P. O. Box 2779,
Houghton 2041, South Africa
Tel: +27-11-717-2594
Fax: +27-11-717-2625
E-mail: neem@global.co.za
Demineralization of hard tooth tissue adjacent to
resin-modified glass-ionomers and composite resins:
a quantitative systematic review
Steffen Mickenautsch and Veerasamy Yengopal
Division of Public Oral Health, University of the Witwatersrand Johannesburg, Houghton, South Africa
(Received 27 January and accepted 12 April 2010)
Original
Page 2
348
methacrylate (HEMA) (9). Compared to other dental
materials, such as non-fluoride-containing composite
resins, laboratory research has shown a higher caries-
resistance in bovine enamel located considerably distant
from the margins of RM-GIC restorations (10). The in situ
trial by Cenci et al. showed lower demineralization in
both enamel and dentine around RM-GIC restorations
(11) and the RCT by Pascotto et al. reported RM-GIC to
be statistically more efficient in reducing enamel demin-
eralization around orthodontic brackets in clinics than
composite resin without fluoride (12).
One systematic review without quantitative synthesis has
been published regarding the secondary caries treatment
effect of GIC restorations (13). This review included C-
GIC and RM-GIC but did not distinguish differences
between these types of material. A more recent review by
Wiegand et al. included an overview covering the influence
of RM-GIC on the demineralization of enamel and dentin
(14). The results of this review indicated a reduction of
carious lesions adjacent to RM-GIC in laboratory trials.
However, no conclusive evidence was obtained from in situ
and clinical trials. Although the review by Wiegand et al.
included a systematic search strategy, it did not report on
quality aspects related to the internal validity of the included
trials and employed only a qualitative synthesis during the
assessment of the trial results (14).
To date, no systematic review using quantitative
synthesis, with or without meta-analysis, has been attempted
on this topic. Thus, the aim of this systematic review was
to quantitatively appraise the current evidence and to
answer the review question about whether RM-GIC, in
comparison to fluoride-containing composite resin and
composite resin without fluoride, is associated with a
higher reduction of demineralization in hard tooth tissues
under caries challenge.
Materials and Methods
Data collection
Five databases: Biomed Central, Cochrane Library,
Directory of Open Access Journals, PubMed and Science-
Direct were systematically searched for articles reporting
on clinical trials up to 6 April 2009. The strings of
MeSH/text search terms with boolean operators: i) “Tooth
Remineralization OR Tooth Demineralization AND Glass
Ionomer Cements AND Composite Resins” and ii) “Dental
Caries OR Dental Caries Susceptibility OR Root Caries
AND Glass Ionomer Cements AND Composite Resins”
were used to search the databases. Articles were selected
for review from the search results on the basis of their
compliance with the inclusion criteria:
1. Titles/abstracts relevant in answering the review
question;
2. Published in English;
3. Two-arm (prospective) longitudinal trial;
4. Focus on materials used for orthodontic and restorative
application.
It was expected that only a few RCTs would be found
relating to this topic. The investigation of the mineral
content of hard tooth tissue often requires evaluation of
extracted teeth under laboratory conditions. For this reason,
clinical trials in this field are challenged by ethical
considerations and randomized, double-blind short-term
in situ trials involving a small number of subjects appear
to be the study design of choice. Moreover, laboratory trials
may also provide additional valuable data on this topic.
However, laboratory trials present weak evidence only,
owing to the uncertainty of extrapolating their results to
physiological effects in humans (15). Thus, it was decided
to include laboratory, in situ and clinical trials in this
review but to assess their outcomes separately in accordance
with the evidence hierarchy (16). Where only a relevant
title without a listed abstract was available, a full copy of
the article was assessed for inclusion. References of the
included articles were checked, in order to identify further
trials suitable for inclusion.
Article review
Only articles that complied with the inclusion criteria
were reviewed further. Full copies of articles were reviewed
independently by two reviewers (VY and SM) in accord-
ance with the exclusion criteria (15):
1. Not all entered subjects accounted for at the end of
the trial;
2. Subjects of both groups not followed up the same way;
3. No randomized, quasi-randomized controlled study
design for in situ and clinical trials;
4. Contains no computable continuous data for extraction
(including the number of evaluated samples (n) and
the mean result of the measured outcome with standard
deviation (SD) for both material groups).
When several articles reporting on the same trial over
similar time periods were available, the article covering
the trial most comprehensively in accordance with the
exclusion criteria was accepted. Disagreements between
reviewers were resolved by discussion and consensus.
Quality of studies
The quality assessment of the accepted in situ and
clinical trials followed guidelines concerning the internal
validity of clinical studies (17) and was undertaken
Page 3
349
independently by two reviewers (VY and SM). Trials not
included in this review were used to pilot the process.
Subsequently, quality assessment rating scored by both
reviewers was derived through consensus. The following
criteria were used:
1) Generation of randomization sequence (allocation),
recorded as:
(A) Adequate – e.g., computer-generated random
numbers, table of random numbers,
(B) Unclear – not reported,
(C) Inadequate – e.g., case record number, date of birth,
date of administration, alternation;
2) Allocation concealment, recorded as:
(A) Adequate – e.g., central randomization, sequentially
numbered sealed opaque envelopes;
(B) Unclear – not reported;
(C) Inadequate – e.g., open allocation schedule, unsealed
or non-opaque envelopes;
3) Blind outcome assessment, recorded as:
(A) Adequate – Yes;
(B) Unclear – No information given as to whether
assessment was blinded;
(C) Inadequate – Reported in text that assessment was
not blinded;
(D) Not possible.
No quality assessment was done for accepted laboratory
trials.
Data extraction from accepted trials
Outcome measures related to the mineral content of
hard tooth tissue under caries challenge in contact with or
adjacent to either material were assessed. Two reviewers
(VY and SM) independently extracted data from the
accepted articles. Individual continuous datasets for the
control- and test-group were extracted from each article.
Where possible, missing data were calculated from
information presented in the text or tables. Authors of
articles were also contacted, in order to obtain missing
information. Data were extracted in the form of datasets,
each containing the number of evaluated samples (n) and
the mean result of the measured outcome with standard
deviation (SD) for both material groups. Disagreements
between reviewers during data extraction were resolved
through discussion and consensus.
Statistical analysis
A random effects model in RevMan Version 4.2 statistical
software by The Nordic Cochrane Centre, The Cochrane
Collaboration (Copenhagen; 2003) was used. Differences
in treatment groups were computed on the basis of mean
difference (MD) with 95% confidence intervals (CI). From
the accepted articles, extracted datasets were assessed for
their clinical and methodological heterogeneity, following
Cochrane guidelines (18). Datasets were considered
heterogeneous if they differed in type of study (laboratory,
in situ or clinical study type); whether the control material
(composite resin) contained fluoride or not; aspect and
definition of outcome measure; and type of hard tooth tissue.
In addition, datasets within each study type were considered
heterogeneous if they differed in the following aspects: i)
Laboratory study: initial exposure period; tissue distance
from material ii) in situ: saliva function; fluoride exposure
from other sources; tissue distance from material; follow-
up period iii) Clinical study: saliva function; fluoride
exposure from other sources; type of dentition; type of
cavity; follow-up period. The percentage of total variations
across datasets (I2), together with its associated P-value
(<0.10), was used in assessing statistical heterogeneity (19).
Only identified homogeneous datasets were considered
suitable for meta-analysis. All datasets were assigned a
Mantel-Haenszel weight directly proportionate to their
sample size.
Results
Systematic literature search and review
An initial search of PubMed, using both strings of
MeSH/Text words (i. and ii.), resulted in 403 and 490
articles, respectively. Of these, 15 articles (10-12,20-31)
complied with the inclusion criteria and were selected for
review. No further articles were identified for selection
during the subsequent search of the other four databases,
and during the reference check. From the 15 selected
articles, two were excluded because they lacked computable
data (20,21).
Thirteen articles; nine laboratory trials (10,22,23,26-31),
three randomized in situ trials (11,24,25) and one RCT were
accepted for further quality assessment and data extraction
(12).
Quality assessment and data extraction
For all in situ and clinical trials random allocation of
subjects, concealment of random allocation and evaluator
blinding were rated “B” (unclear), since no information
about these items was given in the text.
From the accepted laboratory, in situ and clinical trials,
51, 24 and 22, separate computable continuous datasets
with relevance to the review question were extracted,
respectively. The outcome measures of these datasets
related to the mineral content of hard tooth tissue were:
(A) Outcome measures that indicate the mineral loss after
Page 4
350
caries challenge:
a. Laboratory trials: Volume% mineral loss; Knoop
microhardness loss value; Reciprocal microhard-
ness value, as well as the difference in surface
microhardness before and after artificial caries
challenge; Lesion area and Lesion area + lesion
depth
b. In situ trials: Mineral loss; Lesion depth; Increase
of indention length
(B) Outcome measures that indicate the remaining
mineral content after caries challenge:
a. Laboratory trials: Mean density; Knoop micro-
hardness
b. Clinical trial: Knoop microhardness
The main characteristics of the extracted datasets are
described in Table 1-3. Large clinical and methodological
heterogeneity was observed between all datasets and
Table 1-2 Characteristics of datasets (DS) with potential influence on study outcome (laboratory trials) - contd.
Table 1 Characteristics of datasets (DS) with potential influence on study outcome (laboratory trials)
Page 5
351
therefore, no meta-analysis was attempted and statistical
heterogeneity was not further investigated. Instead, the mean
difference between the outcome effects of both material
groups was calculated with 95% confidence intervals
(MD; 95% CI) for each dataset. The results are presented
per study design in Figs. 1-3.
Comparison of RM-GIC versus fluoride-
containing composite resin
The results of the laboratory trials (Fig. 1) revealed no
Table 2 Characteristics of data sets (DS) with potential influence on study outcome (in situ trials)
Table 3 Characteristics of data sets (DS) with potential influence on study outcome (clinical trials)
End of preview.
