Journal of Oral Rehabilitation 1999 26; 364–371
Five year evaluation of class III composite resin
restorations in cavities pre-treated with an oxalic- or a
phosphoric acid conditioner
J. W. V. VAN DIJKEN*, A. L. OLOFSSON† & C. HOLM*
Health Service, Dental School, Umeå University, Umeå, Sweden
SUMMARY An oxalic acid solution has been pro-
posed as a conditioning agent for resin composite
restorations in two commercial adhesive systems.
The durability of 163 class III restorations, includ-
ing 12 class IV restorations, in cavities pre-treated
with an oxalic acid total etch technique or an
enamel etch with phosphoric acid was studied.
Each of 52 patients received at least one of each of
three experimental restorations. The restorat-
ions were evaluated yearly with slightly modified
Introduction
Resin composites have been used in anterior teeth
since the late 1960s. Despite their extensive use in class
III cavities only a few long term evaluations have been
published regarding their durability (van Dijken, 1986;
Smales & Gerke, 1992; van Noort & Davis, 1993; Qvist
& Ström, 1995). The early macrofiller resin composites
were gradually replaced by microfilled and hybrid
composites. The change from auto-cured resins to visi-
ble-light-cured materials resulted in improved colour
stability (van Dijken, 1986). Most of the recently intro-
duced resin composites are of hybrid type and contain
glass fillers with an average size around or below 1 mm.
Buonocore (1955) reported bonding of acrylic resins
to enamel etched with an 85% phosphoric acid. Since
the late 1970s phosphoric acid has been used routinely
in clinics to increase retention of the composite mate-
rial to enamel cavity margins. As a result of increased
enamel roughness and a higher surface energy an
adequate micromechanical bond was obtained between
364
*Department of Oral Biology and †Public Dental
United States Public Health Service (USPHS) crite-
ria. After 5 years 95% of the restorations were
evaluated as acceptable. Reasons for failure were
the fracture of four fillings, including three class
IV, secondary caries contiguous to two fillings and
a non-acceptable colour match for one restoration.
For eight class III restorations a fracture of the
incisal tooth structure was registered. No differ-
ences were seen between the three experimental
restorations.
resin and enamel. The use of newer enamel/dentin
adhesive systems, which contained more hydrophylic
monomers, made it possible also to obtain a bond to
the wet dentin (Nakabayashi, Kajima & Mashura,
1982). The mechanism of bonding to dentin following
acid etching is apparently similar to the enamel-etch
technique. A zone of porous dentin, created by condi-
tioning with an acid, is penetrated by hydrophylic
monomers and a so called hybrid zone is formed,
giving mechanical retention after polymerization. An-
other development in adhesive dentistry has been the
introduction of conditioning agents which etch enamel
and dentin simultaneously in the so-called total-etch
technique.
New bonding systems have been continuously devel-
oped during recent years, aiming to improve adhesion
of the resin composite to the tooth tissues and also to
simplify the clinical procedure. In an attempt to market
conditioning methods which were thought to be less
aggressive to use on dentin, mild conditioning agents
such as citric, maleic, nitric and oxalic acid in various
© 1999 Blackwell Science Ltd
 CAVITY PRE-TREATMENT WITH OXALIC ACID
concentrations have been proposed for use in new
dental adhesive systems, replacing the 35 – 40% phos-
phoric acid etching gel.
A bonding system which shows a relatively high
bond strength in the laboratory, and in vivo retention
rates required by the guidelines of the American Dental
Association for class V abrasion/erosion lesions, is the
original Gluma system (Eliades, Caputo & Vougiouk-
lakis, 1985; Hörstedt-Bindslev, Knudsen & Baelum,
1988; van Dijken, 1990). The system consists of four
steps. The first two steps condition enamel and dentin
with a 37% phosphoric acid and a neutralized 0·5 M
EDTA solution, respectively. Bowen and co-workers
(Bowen, Cobb & Rapson, 1982; Bowen et al., 1987)
devised in the early 1980s a dentin bonding system
using a dentin conditioner with an acidulated solution
of aluminium oxalate. Asmussen & Bowen (1987)
showed later that an increase in bond strength is ob-
tained by pretreating the dentin with certain amino
acids. Conditioning with an acidic solution of alu-
minium oxalate containing glycine, followed by
modified Gluma resin, yielded strong bonds to enamel
and dentin and simplifies the system (de Araujo &
Asmussen, 1989). This principle has been applied in
the formulation of the modified Gluma bonding sys-
tem, marketed as the two step Gluma 2000 bonding
system. The bonding system in vitro reduced signifi-
cantly microleakage at restoration margins in enamel
and dentin (Chan & Swift, 1994). The conditioner
contains oxalic acid, aluminium nitrate, glycine and
water for simultaneous conditioning of enamel and
dentin. Recently, another simplified Gluma adhesive
system, the Gluma 3-step bonding system became
available on the US market. The original Gluma 1
etchant and Gluma 2 cleanser are replaced with the
Gluma 2000 step 1 conditioner, while the original
Gluma 3 primer and Gluma 4 sealer are unaltered.
Although a milder etching pattern is obtained with the
oxalic solution, the bond strength of the modified
bonding system compares favourably with the original
Gluma 1 etchant (Strickland et al., 1992; Strydom et al.,
1995).
It was the purpose of this clinical investigation to
evaluate and compare intra-individually the clinical
effectiveness of the Gluma 2000 bonding system and
the Gluma 3-step bonding system, both with an oxalic
acid cavity conditioning and with a conventional phos-
phoric acid etch technique in class III cavities.
© 1999 Blackwell Science Ltd, Journal of Oral Rehabilitation 26; 364 – 371
365
Materials and methods
A total of 163 class III fillings, including 12 class IV
fillings, were placed in 52 patients, who were fully
informed of the study. All the patients, 20 men and 32
women, with a mean age of 56 years (range: 33–
76 years), received at least three class III fillings, one of
each of the experimental bonding groups. The fillings
were placed by random assignment in the teeth se-
lected. The majority of the cavities were replacements
of old composite fillings as a result of secondary caries
or for aesthetic reasons. The mostly large-sized cavities
were as close in size as possible within each patient.
The amount of enamel (%) surrounding the lesions
was at a maximum of 75% of the marginal circumfer-
ence. A 45° bevel was placed on all enamel margins.
After colour selection the teeth were isolated with a
rubber dam. In each patient each of the three following
treatments were performed:
(i) Gluma 2000 bonding system: Conditioning and
bonding of the cavity with the Gluma 2000 sys-
tem* (batch 2017J and 2022I), followed by inser-
tion of the hybrid-type resin composite Pekafill
PLT* (batch 1157G and 2068H) (group G2000).
The Gluma 2000 solution 1 contains oxalic acid,
aluminium nitrate, glycine and water for simulta-
neous conditioning of enamel and dentin. The
solution 2 primer was an aqueous solution of a
bifunctional monomer n-methacryloyloxyethyl-n-
methylformamide and ethanol.
(ii) Gluma 3-step bonding system: Conditioning of the
cavity with the Gluma 2000 solution 1 was fol-
lowed by application of Gluma 3 primer, the
Gluma 4 sealer* (batch 1156G and 2005I) and
Pekafill (group G 3-step). Gluma 3 primer contains
5% glutaraldehyde and 35% HEMA in water, and
the Gluma 4 sealer is an unfilled low viscosity
bonding resin containing bis-GMA and TEGDMA
monomers.
(iii) Control: Conventional acid etching, with a 37%
phosphoric acid gel (Gluma 1), of enamel parts of
the cavity for 30 s, thoroughly rinsing with water
and drying with compressed air was followed by
application of Gluma 4 sealer and Pekafill (group
C).
* Bayer, Dormhagen, Germany.
366 J . W . V . V A N D I J K E N et al.

Table 1. Criteria for direct clinical evaluation

Score

Category Acceptable Unacceptable Criteria

Anatomical form 0 The restoration is continuous with tooth anatomy

1 Slightly under- or over-contoured restoration; marginal ridges slightly under
contoured; contact slightly open (may be self-correcting); occlusal height
reduced locally
2 Restoration is under contoured, dentin or base exposed; contact is faulty, not
self-correcting; occlusal height reduced; occlusion affected
3 Restoration is missing or traumatic occlusion; restoration causes pain in tooth
or adjacent tissue
Marginal adaptation 0 Restoration is continuous with existing anatomic form, explorer does not catch
1 Explorer catches, no crevice is visible into which explorer will penetrate
2 Crevice at margin, enamel exposed
3 Obvious crevice at margin, dentin or base exposed
4 Restoration mobile, fractured or missing

Colour match 0 Very good colour match

1 Good colour match
2 Slight mismatch in colour, shade or translucency
3 Obvious mismatch, outside the normal range
4 Gross mismatch
Marginal discoloration 0 No discoloration evident
1 Slight staining, can be polished away
2 Obvious staining cannot be polished away
3 Gross staining
Surface roughness 0 Smooth surface
1 Slightly rough or pitted
2 Rough, cannot be refinished
3 Surface deeply pitted, irregular grooves
Caries 0 No evidence of caries contiguous with the margin of the restoration
1 Caries is evident contiguous with the margin of the restoration

The manufacturer’s directions were closely adhered
to regarding the placement of the bonding systems and
the restorative material. The solution 1 of the Gluma
2000 system was applied to the whole prepared area
including the bevel, using a dabbing action for 30 s.
The cavity was then rinsed with water for 10 s and air
dried. The solution 2 of Gluma 2000 was applied for
30 s using a soaked pellet, followed by gentle and
careful air-drying. The Gluma 3 primer was applied for
30 s using a soaked pellet and then carefully air dried.
The Gluma 4 sealer was carefully brushed over the
entire cavity surface and excess was removed by a
gentle blast of air. As recommended by the manufac-
turer, the sealer was not separately light cured. The
composite resin was dispensed from pre-loaded tubes
(PLT) with a Centrix† syringe. A three layer incremental
technique was used with the first layer receiving 60 s
and the other two layers receiving 40 s light activation
(Luxor light curing unit‡). The restorations were
finished by using multifluted carbide or diamond
finishing burs to remove gross excess, followed by
proximal finishing strips. The restorations were made
by three experienced dentists.
Evaluation
The restorations were evaluated directly after finishing
(baseline), after 6, 12, 24, 36, 48 and 60 months. The
following parameters were used: anatomical form,
marginal adaptation, colour stability, marginal discol-
oration, surface texture, caries, postoperative sensitivity
and vitality. Criteria used were a slight modification of
USPHS criteria as earlier described (Table 1)(van Di-
jken, 1986). All restorations were independently in-

‡
†
Hawe Neos, Gentilino, Switzerland. ICI, Macclesfield, U.K.

© 1999 Blackwell Science Ltd, Journal of Oral Rehabilitation 26; 364–371

 CAVITY PRE-TREATMENT WITH OXALIC ACID
vestigated by two evaluators. The vitality of the teeth
was tested at each recall during the first 3 years using
an electric pulp tester. The patients were surveyed
regarding the presence or absence of post-operative
sensitivity, 1 week after placement and at the recalls.
The data were recorded according to the criteria of
Borgmeijer et al. (1991).
A prediction of the caries risk, expressed as the
potential caries activity was made for each patient. The
individual caries risk was estimated from the net effect
of the patient’s oral hygiene, intake of fermentable
carbohydrates, salivary microbial counts, salivary flow
rate and buffering capacity. These factors were re-
garded as negative factors when certain values were
exceeded. Details have been reported previously (van
Dijken, 1986).
Intra-individual comparisons of the quality of the
three composite resin restorations were made. The
quality of the restorations was evaluated directly with
regard to marginal adaptation, anatomical form, mar-
ginal discoloration and presence of secondary caries, as
earlier described by van Dijken (1991). Then the three
restorations were rank ordered within each patient. In
this manner each patient served as a statistical unit.
The sums of the ranks of each experimental group
were then tested using Friedman’s two-way analysis of
variance test (Siegel, 1956). The null hypothesis was
rejected at the 1% level.
Results
All except 14 restorations were evaluated at the 5 year
recalls. Seven fillings were not evaluated because of
crown therapy, while the remaining seven were not
evaluated because the patients were unable to attend
the recalls. A cumulative rate of 149 restorations were
evaluated during the 5 year period and 143 of these
(96%) were assessed as acceptable. Seven restorations
were evaluated as non-acceptable. These included
three class IV and one class III, because of total or
partial fracture of the restorations after 1, 3, 4 and
5 years (three G2000, one C), two others for secondary
caries, both after 3 years (G 3-step, C), and one for
non-acceptable colour match at the 5 year recall. The
restoration with non-acceptable colour match was not
replaced as to the patient’s wishes. Two initial caries
lesions (G 3-step, C) observed at 2 and 3 years were
estimated as arrested at 5 years.
© 1999 Blackwell Science Ltd, Journal of Oral Rehabilitation 26; 364 – 371
367
Eight class III restorations showed fracture of the
incisal tooth structure: four G2000 teeth, one at 1 year,
two at 3 years and one at 5 years; three G 3-step teeth,
at 2, 3 and 5 years; and one C tooth at 3 years. In two
cases the restoration was replaced (G2000 and G 3-
step), and in three cases the fracture was repaired with
resin composite, while the original filling was still in
place. No treatment was given in the other three cases.
In total seven restorations (4%) were replaced during
the 5 year period. Frequencies of the scores (%) for the
evaluated variables at 5 years are given in Table 2. No
significant differences were seen between the three
experimental groups.
A small amount of calcium hydroxide base§ was
placed in 15 restorations, with little remaining dentin
(seven G2000, four G 3-step and four C). Post-opera-
tive sensitivity was reported for two restorations at
baseline for temperature changes during 1 week (one
G2000 and one G 3-step). Another C filling was sensi-
tive at baseline in occlusion, which disappeared after
correction of the occlusion. At the other recalls no
complaints were registered. Four teeth with low vital-
ity reactions at baseline (one G2000, one G 3-step and
two C) showed no signs of vitality at the 3 year recall.
The prediction of the potential caries activity of each
patient showed that 12 out of 47 (25%) of the patients
were considered as high caries-risk patients.
Discussion
Since the introduction of the enamel-etch technique
the dental profession has changed the concentration of
phosphoric acid and the etching time several times.
Buonocore (1955) reported about the effect of 85%
phosphoric acid. The concentration has decreased over
the years and in most adhesive systems used today it is
between 32% and 37%, whilst even 10% and 20%
gels are marketed. Clinical follow-ups have shown that
the original etching time of 120 s could be decreased to
15 s (van Dijken, 1986). During recent years several
other acids, such as citric, nitric, maleic and oxalic acid,
were introduced for use in total-etch techniques. One
of the reasons was the widespread opinion that phos-
phoric acid would be too aggressive for the dentin
tissue. However, the milder acids did not gain clinical
acceptance, as controversy existed about the etching
§
Dycal LD Co., Milford, DE, U.S.A.
368 J . W . V . V A N D I J K E N et al.

Table 2. Scores for the evaluated parameters at the 5 year phoric acid or conditioned with the oxalic solution,
recalls (%)
was similar, although the second conditioner was less
G2000 G 3-step C aggressive. They concluded that the retention to
enamel was more a function of increase in surface area
Score No. % No. % No. %
and increase in wettability of the etched enamel than
Anatomical 0 53 98·1 48 100 48 97·9 of depth of pattern. On the other hand Swift & Cloe
form (1993) found a significantly lower shear bond strength
1 0 0 0
after conditioning with the oxalic acid solution com-
2 1·9 0 2·1
pared to treatment with 10% or 35% phosphoric acid.
Marginal 0 53 62·3 48 56·3 48 70·8 The same research group showed that the Gluma 2000
adaptation* 1 30·2 39·6 27·1
2 3·8 0 0 system effectively reduced microleakage at both the
3 1·9 4·2 0
enamel and dentin margins, comparable to the All-
4 1·9 0 2·1 Bond 2¶ and Scotchbond MP** systems, which used
Colour match 0 45 42·2 45 53·3 44 65·9 phosphoric- and maleic acid as conditioners, respec-
1 53·3 42·2 27·3 tively (Chan & Swift, 1994). However, in two other
2 2·2 4·4 6·8
studies they reported that in comparison to other cur-
3 2·2 0 0 rent generation dentin adhesive systems the Gluma
4 0 0 0
2000 and the Gluma 3-step system had the lowest
Marginal 0 45 82·2 45 95·6 44 84·1
bond strengths and the greatest leakage (Triolo &
discoloration 1 17·8 4·4 13·6
2 0 0 2·3 Swift, 1992; Fortin et al., 1994). Bond strength and
3 0 0 0 microleakage figures are commonly used parameters
Surface 0 45 97·8 45 93·4 44 95·4
for the characterization or estimation of adhesive sys-
roughness 1 2·2 6·6 4·6 tems. Shear bond strength is not necessarily an accu-
2 0 0 0 rate predictor of the quality of the marginal adaptation
Caries 0 53 100 48 97·9 48 97·7 and no significant correlation was found between bond
1 0 2·1 2·3
strength and microleakage (Fortin et al., 1994). These
* Cumulative scores. data are screening figures and can only be seen as
relative indicators of the anticipated clinical perfor-
mance. Clinical studies are required to determine the
efficacy of the alternative acids and their clinical valid- long-term efficacy of new systems.
ity was not proven. Their etching effect on enamel was Over the years only a few long term evaluations of
less controllable by the dentist, and after some years anterior restorations have been published. Replace-
the manufacturers replaced many of these acids with ment rates of a range of chemically-cured resin com-
the traditional phosphoric acid (Finger & Uno, 1996). posites in class III, IV and V cavities varied between
Good clinical results were found for Gluma bonding 15% and 58% (medium 30%) after 6 years (van
system in class V and class III studies (Hörstedt-Bind- Dijken, 1986). The replacement frequencies for the
slev et al., 1988; van Dijken, 1990; Ferrari, Bertelli & same type of composite materials in class III cavities,
Finger, 1993). Subsequently the system was reformu- but performed in a multi-center study by general den-
lated and simplified to fewer clinical steps. In both tists ranged from 30% to 44% after 5 years (van Noort
Gluma 2000 and Gluma 3-step the first and second & Davis, 1993). Better 5 year survival rates were re-
step of the original Gluma system, 37% phosphoric ported by Bentley & Drake (1986) with a 17% failure
acid and EDTA was replaced by an oxalic acid/glycine rate for chemically cured composite restorations placed
solution. The etching pattern created by phosphoric by students, and by Smales & Gerke (1992) with an
acid was deeper and more pronounced than that re- 8% failure rate of restorations performed by a private
sulting from the oxalic acid solution. Inoue, Finger &
Mueller (1994) showed that the retentive strength be- ¶
Bisco, Inc., Schaumburg, IL, U.S.A.
tween the resin and ground enamel, etched with phos- ** 3M Co., St Paul, MN, U.S.A.

© 1999 Blackwell Science Ltd, Journal of Oral Rehabilitation 26; 364–371

 CAVITY PRE-TREATMENT WITH OXALIC ACID
practitioner. Enamel bonding resins were combined
with the enamel acid-etch technique in all these stud-
ies. In the study by van Noort and Davis it was not
indicated that etching was used. Most frequent reasons
for replacements in these studies were secondary caries
and body discoloration. In an attempt to prevent mar-
ginal gap formation Qvist & Ström (1995) re-etched
the margins of class III microfilled composite restora-
tions and applied bonding resin after finishing. This
resulted in failure rates of 8% after 6 years and 20%
after 11 years. In our study seven (5%) of the restora-
tions were registered as non-acceptable, while in seven
other teeth the tooth incisal of the restorations frac-
tured. The restorations in two of the tooth fractures
and six of the non-acceptable restorations were re-
placed (6%). Two of the other tooth fractures could be
repaired easily, while the other patients would not
consider replacement. The high incidence of tooth frac-
tures for class III restorations found in this study has
also be reported by others (van Dijken, 1986; van
Noort & Davis, 1993; Qvist & Ström, 1995). When we
consider the class III restorations in this study only 4%
failed. The class IV fillings show a high failure rate of
30%. All of the failures were in patients with high
parafunctional habits. The higher survival rate of class
III restorations has been shown in several earlier stud-
ies (van Dijken, 1986; Smales & Gerke, 1992; Millar,
Robinson & Inglis, 1997). Ferrari et al. (1993) studied
40 anterior restorations performed with a microfilled
resin composite bonded with the original Gluma sys-
tem. They reported a 15% failure rate after 5 years.
van Dijken (1996) evaluated 56 class III Pekafil com-
posite fillings bonded with the Gluma system during
3 years. Two incisal enamel fractures were registered,
which gives a failure rate of 3%. The high survival rate
for the class III restorations performed with the bond-
ing system Gluma 2000 is in strong contrast with the
findings in class V abrasion/erosion lesions. High losses
of retention were noted in three clinical evaluations
during the first year (van Dijken, 1993; Van Meerbeek
et al., 1994; Bayne et al., 1995). A higher failure rate
was found when bonded to sclerotic dentin (Bayne et
al., 1995). Concerning the controversial findings it can
be questioned how relevant the class V retention eval-
uation studies are for other cavity types.
Besides the high survival rates assessed, the most
surprising finding in the study was that no difference
© 1999 Blackwell Science Ltd, Journal of Oral Rehabilitation 26; 364 – 371
369
was found in survival rates for the three adhesive
techniques investigated. Combined with the hybrid
composite resin, they functioned very well during the
5 year follow up as a class III restorative. Despite the
fact that the cervical margins of all cavities were situ-
ated in dentin there was a low failure rate as in the
control group, where an enamel bonding and no
dentin priming was used. It cannot be excluded that
part of the dentin tissue in the control cavities was also
etched during the enamel etching and following water
rinsing step. This could possibly have resulted in better
sealing as a result of penetration of the enamel bonding
resin in the etched dentin, as shown in a recently
performed scanning electron microscope evaluation of
the interfacial adaptation of different adhesive tech-
niques in vivo (van Dijken, Hörstedt & Waern, 1998).
In the control group fillings in this study where only an
enamel bonding agent was used after total-etch of the
cavity with phosphoric acid, a 73% gap free adaptation
in the dentin was observed. This could be compared to
87% in the groups where an efficient last generation
dentin adhesive system was used.
The only difference found between the three adhe-
sive techniques studied was the number of teeth which
showed incisal fractures. All tooth fractures, except
one, were found in the two cavity groups conditioned
with the oxalic acid solution. One possible reason
could be the less pronounced enamel etch pattern
caused by the oxalic acid. On the other hand, no
differences were found in the quality of the enamel
marginal adaptation when we observed the three tech-
niques at baseline and after 1 year in vivo ageing using
the scanning electron microscope replica technique
(van Dijken & Hörstedt, 1998).
Only two secondary lesions were registered during
the evaluation. The incidence of secondary caries was
low despite the fact that 25% of the participating
patients were considered as caries risk patients. The
findings are in contrast with findings in earlier studies,
where secondary caries were the major reason for
replacement of resin composite fillings (van Dijken,
1986; Qvist, Qvist & Mjör, 1990). The ultimate proof of
efficacy of the bonding systems is the clinical trial. The
equal good results in all experimental groups in this
clinical evaluation raise doubts about whether a dentin
priming is necessary. The clinical picture after 5 years
did not confirm the hypothesis that the use of the
dentin bonding systems improved the durability of the
370 J . W . V . V A N D I J K E N et al.
class III restorations. The use of dentin bonding systems
in cavities of small volume can therefore be
questioned.
Acknowledgments
This study was supported in part by the Swedish Medi-
cal Research Council, the County Council of Västerbot-
ten, Bayer and the Swedish Dental Society.
References
ASMUSSEN, E. & BOWEN, R.L. (1987) Adhesion to dentin medi-
ated by Gluma: effect of pre-treatment with various amino
acids. Scandinavian Journal of Dental Research, 66, 1386.
DE ARAUJO, P.A. & ASMUSSEN, E. (1989) Aluminium oxalate/
glycine solutions as pre-treatment in the Gluma bonding sys-
tem. Scandinavian Journal of Dental Research, 97, 552.
BAYNE, S.C., HEYMANN, H.O., WILDER, A.D., STURDEVANT, J.R. &
ROBERTSON, T.M. (1995) 1-year clinical study of sclerotic vs
nonsclerotic dentin bonding. Journal of Dental Research, 74,
169 [Abstract no. 1221].
BENTLEY, C. & DRAKE, C.W. (1986) Longevity of restorations in
dental school clinic. Journal of Dental Education, 50, 594.
BORGMEIJER, P.J., KREULEN, C.M., AMERONGEN, W.E., AKER-
BOOM, H.B.M. & GRUYTHUYSEN, R.M.J. (1991) The prevalence
of postoperative sensitivity in teeth restored with class II
composite resin restorations. Journal of Dentistry for Children,
58, 378.
BOWEN, R.L., COBB, E.N. & RAPSON, J.E. (1982) Adhesive bond-
ing of various materials to hard tissue: improvement in bond
strength to dentin. Journal of Dental Research, 61, 1070.
BOWEN, R.L., ASMUSSEN, E., TUNG, M.S. & BLOSSER, R.L. (1987)
Dentin and enamel bonding agents. International Dental Jour-
nal, 37, 158.
BUONOCORE, M.G. (1955) A simple method of increasing the
adhesion of acrylic filling materials to enamel surfaces. Jour-
nal of Dental Research, 34, 849.
CHAN, K.C. & SWIFT, E.J. (1994) Marginal seal of new-genera-
tion dental bonding agents. Journal of Prosthetic Dentistry, 72,
420.
VAN DIJKEN, J.W.V. (1986) A clinical evaluation of anterior
conventional, microfiller and hybrid composite resin fillings.
A six year follow up study. Acta Odontologica Scandinavica, 44,
357.
VAN DIJKEN, J.W.V. (1990) A 3-year evaluation of Gluma and
Gluma/Scotchbond for restoration of cervical lesions. Scandi-
navian Journal of Dental Research, 98, 341.
VAN DIJKEN, J.W.V. (1991) A six year follow-up of three dental
alloy restorations with different copper contents. Swedish
Dental Journal, 15, 259.
VAN DIJKEN, J.W.V. (1993) Klinisch gedrag van dentine-adhe-
sieven. Tandartspraktijk, 14, 27.
VAN DIJKEN, J.W.V. (1996) 3-year evaluation of a compomer, a
© 1999 Blackwell Science Ltd, Journal of Oral Rehabilitation 26; 364–371
resin-modified glass ionomer and a resin composite in class
III cavities. American Journal of Dentistry, 9, 195.
VAN DIJKEN, J.W.V. & HÖRSTEDT, P. (1998) The effect of pre-
treatment with an oxalic acid solution on marginal adapta-
tion to enamel in vivo. Journal of Prosthetic Dentistry, 80, 75.
VAN DIJKEN, J.W.V., HÖRSTEDT, P. & WAERN, R (1998) Directed
polymerization shrinkage versus a horizontal incremental
filling technique. Interfacial adaptation in vivo in class II cav-
ities. American Journal of Dentistry, 11, 165.
ELIADES, G.C., CAPUTO, A.A. & VOUGIOUKLAKIS, G.J. (1985)
Composition, wetting properties and bond strength with
dentin of six new dentin adhesives. Dental Materials, 1, 170.
FERRARI, M., BERTELLI, E. & FINGER, W. (1993) A 5-year report
on a enamel-dentinal bonding agent and microfilled resin
system. Quintessence International, 24, 735.
FINGER, W.J. & UNO, S. (1996) Bond strength of Gluma CPS
using the moist dentin bonding technique. American Journal
of Dentistry, 9, 27.
FORTIN, D., SWIFT, E.J., DENEHY, G.E. & REINHARDT, J.W. (1994)
Bond strength and microleakage of current dentin adhesives.
Dental Materials, 10, 253.
HÖRSTEDT-BINDSLEV, P., KNUDSEN, J. & BAELUM, V. (1988)
Dentin adhesive materials for restoration of cervical erosions.
Two and three year clinical observations. American Journal of
Dentistry, 1, 195.
INOUE, M., FINGER, W.J. & MUELLER, M. (1994) Effect of filler
content of restorative resins on retentive strength to acid-
conditioned enamel. American Journal of Dentistry, 7, 161.
MILLAR, B.J., ROBINSON, P.B. & INGLIS, A.T. (1997) Clinical
evaluation of an anterior hybrid composite over 8 years.
British Dental Journal, 182, 26.
NAKABAYASHI, N., KAJIMA, K. & MASHURA, E. (1982) The pro-
motion of adhesion by infiltration of monomers into tooth
substrates. Journal of Biomedical Material Research, 16, 265.
VAN NOORT, R. & DAVIS, L.G. (1993) A prospective study of the
survival of chemically activated anterior resin composite
restorations in general dental practice: 5-year results. Journal
of Dentistry, 21, 209.
QVIST, V., QVIST, J. & MJÖR, I.A. (1990) Placement and
longevity of tooth coloured restorations in Denmark. Acta
Odontologica Scandinavica, 48, 305.
QVIST, V. & STRÖM, C. (1995) 11-year assessment of class-III
resin restorations completed with two restorative procedures.
Acta Odontologica Scandinavica, 51, 253.
SIEGEL, S. (1956) Non Parametric Statistics for the Behavioural Sci-
ences. Mc Graw-Hill, New York.
SMALES, R.J. & GERKE, D.J. (1992) Clinical evaluation of four
anterior composite resins over five years. Dental Materials, 8,
246.
STRICKLAND, S.S., RETIEF, D.H., RUSSELL, C.M. & DENYS, F.R.
(1992) Gluma shear bond strength to enamel and dentin
treated with pyruvic acid and glycine. American Journal of
Dentistry, 5, 97.
STRYDOM, C., RETIEF, D.H., RUSSELL, C.M. & DENYS, R.M. (1995)
Laboratory evaluation of the Gluma 3-step bonding system.
American Journal of Dentistry, 8, 93.
 CAVITY PRE-TREATMENT WITH OXALIC ACID 371

SWIFT, E.J. & CLOE, B.C. (1993) Shear bond strength of new
enamel etchants. American Journal of Dentistry, 6, 162.
TRIOLO, P.T. & SWIFT, E.J. (1992) Shear bond strengths of ten
dentin adhesive systems. Dental Materials, 8, 370.
VAN MEERBEEK, B., PEUMANS, M., VERSCHUEREN, M., GLADYS, S.,
BRAEM, M., LAMBRECHTS, P. & VANHERLE, G. (1994) Clinical
status of ten dentin adhesive systems. Journal of Dental Research,
73, 1690.
Correspondence: Dr Jan W. V. van Dijken, Institution of Oral
Biology, Dental School Umeå, Umeå University, S-901 87 Umeå,
Sweden. E-mail: Jan.van.Dijken@oralbio.umu.se

.
.

© 1999 Blackwell Science Ltd, Journal of Oral Rehabilitation 26; 364 – 371