Benefits and disadvantages of tooth-coloured alternatives to amalgam

OBJECTIVES: To give the practising dentist scientifically based data to assist him/her in the responsible decision-making process necessary to weigh the options available to the patient if she/he prefers not to have an amalgam placed. DATA SOURCES: Based on the literature and on the research work, which was done in the author's department, the indications and limitations of the known alternatives of amalgam were formulated. DESCRIPTION OF ALTERNATIVES TO AMALGAM: With the exception of cast gold restorations, all alternatives require the strict use of adhesive techniques. When compared with similar amalgam restorations, placing composite restorations (if they are indicated) takes approximately 2.5 times longer because complex incremental techniques are needed. Despite all the efforts, direct composite restorations placed in large cavities still show unacceptable amounts of marginal openings. Tooth-coloured inlays are a better alternative for large restorations. These restorations must be inserted with adhesive techniques. With composite inlays it is difficult to achieve a composite-composite bond. Ceramic inlays may be micromechanically bonded to the luting composite. They all show clinically a good marginal behaviour and the use of ultrasonic energy may further simplify the application technique of aesthetic inlays. STUDY SELECTION: Papers describing the different techniques were used as a base for the corresponding chapter. To assess and compare the longevity of the different restoration types, literature data were used. We limited ourselves to papers reporting at least 5-year clinical data. Longitudinal, clinically controlled studies were preferred. However, to be more complete, retrospective, cross sectional studies were also included. LONGEVITY OF POSTERIOR RESTORATIONS: Amalgam shows excellent longevity data with studies up to 20 years. The average annual failure rate is 0.3-6.9%. Posterior composites are in the same range (0.5-6.6%), however, the study times are much shorter (max. 10 years). For tooth-coloured inlays much less data are available. Longevity is reported up to 6 years with annual failure rates of 0.6-5%. CONCLUSIONS: All aesthetic alternatives to amalgam require more complex procedures and more time. If cost benefit considerations are a concern, amalgam is still the most convenient restorative material for posterior teeth.     

Resin-modified glass ionomers for luting posterior ceramic restorations

OBJECTIVES: Until recently, esthetic inlay restorations in posterior teeth have been limited to cavities surrounded by enamel. Dentin adhesive systems in combination with luting composites and light-cured resin-modified glass ionomer cements offer a possibility for bonding ceramic inlays to cavities when the cervical margin is in dentin. This study was designed to compare in vitro marginal integrity of ceramic inlays bonded to dentin to restorations placed in cavities with margins located entirely in the enamel. METHODS: In the present in vitro study, the sealing abilities of a dentin bonding agent/luting composite combination (Syntac/Dual Cement, Vivadent) and resin-modified glass ionomers (Photac Fil, Photac Bond, ESPE; Dyract, De Trey Dentsply; Fuji II LC, GC Dental Industrial Corp.; and Vitremer, 3M Dental Products) used as luting agents in cavities extending beyond the cemento-enamel junction, were compared to the sealing abilities of a conventional luting composite (Vita Cerec Duo Cement, Vita) in cavities within sound enamel. SEM analysis and dye penetration were performed to evaluate marginal integrity at the cervical cavity margins. RESULTS: The dentin bonding agent/luting composite combination (Syntac/Dual Cement) rendered a marginal seal within the dentin similar to the quality obtained with the conventional luting procedures within sound enamel. When three out of the five resin-modified glass ionomers were used as luting agents (Dyract, Fuji II LC and Vitremer), the results were comparable to those reported for the dentin bonding agents and the conventional method. SIGNIFICANCE: Light-cured resin-modified glass ionomer cements may be considered as an alternative to dentin bonding agents when the cavity margins of ceramic inlay restorations are within the dentin. However, further studies, e.g., wear resistance, must be performed.     

In vitro milling accuracy and fit of 'natural inlays'

OBJECTIVE: This study aimed to investigate the milling behaviour of natural inlays (NI) dental restorations constructed from sound extracted teeth. This was done by comparing the milling accuracy and fit of NI to those of industrial porcelain inlays (PI), milled in the same way. METHODS: A calibration pro-inlay was used to mill three NI and three PI. These were cemented in six acrylic cavities reproduced from the calibration mould, using composite luting cement. No etching and bonding were done. After storage for 24 h in water at room temperature the specimens were sliced buccolingually in a standardised way. For each specimen, two sections were photographed under a light-microscope, resulting in enlarged pictures of the sections. An acetate matrix with the measurement points was placed over each picture in a standardised way. The interfaces between inlays and moulds were measured at 13 fixed points per section using computerised image analysis software. The mean vertical- and floor-interfaces were calculated for each picture, and the overall means were found for each group. Confidence intervals were used for comparison of the differences. The profiles of the milled materials were examined using scanning electron microscopy. RESULTS: There were no differences between NI and PI in the mean interfaces (NI, 102 +/- 8 micrometers; PI, 107 +/- 8 micrometers). Electron microscopy revealed no apparent differences in the profiles of the milled surfaces. CONCLUSION: These findings indicate that the milling accuracy and the fit of natural inlays and milled porcelain inlays are comparable.     

New England eye care survey

Amalgambond Plus with a high-performance additive was evaluated for its ability to bond a resin composite or an amalgam alloy to deep dentin in primary teeth with nonretentive cavity preparations. The clinical performance of amalgam and resin composite mesio-occlusodistal restorations bonded with Amalgambond Plus was evaluated at 15 months and 2 and 3 years. There was no statistically significant difference in the retention, marginal adaptation, secondary caries, and post-operative sensitivity over the times of the evaluation or between amalgam and resin composite restorations. After 3 years, most of the teeth were extracted because it was their exfoliation time, and 29 restorations limited within buccal and lingual cusps were evaluated for marginal leakage. There were no significant differences in microleakage between amalgam and resin composite restorations lined with Amalgambond Plus. Amalgambond Plus has the potential for serving successfully as a cavity liner with either amalgam or resin composite restorations.     

Value of the search for anticardiolipin antibodies during the course of systemic lupus erythematosus

This study investigates the shear bond strengths (SBS) of different products used for bonding amalgam to dentine and the marginal seal provided by these materials. The SBS test was carried out by bonding cylinders of Dispersalloy amalgam to human dentine with an intermediate layer of Amalgambond Plus, All-Bond 2, Imperva Bond/Dual or Scotch Bond Multi-Purpose was first placed. Thereafter the samples were tested to failure in the shear mode. For the microleakage test, standardized class V cavities were prepared in human molars with one cavo margin abutting enamel and another dentine/cementum. Amalgam was condensed into the cavities, pretreated with one of the products mentioned above. In addition two further groups were prepared. One receiving treatment with Polyvar varnish and the other remained untreated as the control. Microleakage was assessed with a reflecting light microscope using I.S.O. criteria and the fracture sites and marginal gaps were examined in a scanning electron microscope. Results indicate that All-Bond 2 and Amalgambond Plus induced similar SBS and Scotchbond Multi-Purpose the weakest. At both enamel and dentine/cementum junctions resin bonding agents reduced microleakage. However, this reduction was greater at the enamel interface than at the other. Varnish allowed the greatest amount of marginal leakage and leakage was similar to the untreated control. Amalgam bonding agents are more effective at preventing marginal leakage at the enamel margin than at the dentine/cementum margin. Cavity varnish is not effective in preventing microleakage around amalgam restorations. There is a correlation between shear bond strength and marginal leakage. Materials giving the highest shear bond strength also exhibited the least marginal leakage. All-Bond 2, Amalgambond Plus and Imperva Bond/Dual are recommended to improve the seal around amalgam restorations.     

Bonding of amalgam to composite: tensile strength and morphology study

OBJECTIVES: The present study was based on the premise that it may be possible to produce useful dental restorations by bonding freshly triturated amalgam to a cured composite restoration (Group 1 specimens), or by bonding uncured composite to hardened amalgam (Group 2 specimens). METHODS: To determine the validity of this premise, a phosphonate adhesive resin cement was used to produce simulated, layered dental restorations for each test group. RESULTS: The mean tensile bond strength of 24 hour-old Group 1 specimens (6.74 MPa +/- 1.63 MPa) was almost twice that of 24 hour-old Group 2 specimens. Cohesive failure of the amalgam-substrate layer was a prominent feature of the fracture pattern of Group 1 specimens. On the other hand, rupture of all Group 2 specimens occurred mainly along the adhesive-amalgam interface. Findings from SEM examination of the layers of amalgam, adhesive cement, and resin composite of intact Group 1 specimens suggested that inclusions of adhesive resin cement may be the cause of the persistent cohesive failure of the amalgam layer. SIGNIFICANCE: It may be possible to improve the strength of bonded couples made from freshly triturated amalgam and cured resin composite by minimizing the thickness of the adhesive layer.     

Comparison of marginal fit and microleakage of ceramic and composite inlays: an in vitro study

The purpose of this in vitro study was to evaluate the marginal fit and microleakage of four types of ceramic or resin composite inlays. Sixty extracted human third molar teeth were randomly assigned into four groups. MOD cavities without bevels were prepared. The mesiogingival margin was cut in enamel, whereas the distogingival margin was placed below the amelocemental junction. The preparations and inlays were made according to the manufacturers' instructions. After acid-etching and treatment with a dentine bonding agent the inlays were cemented with dual curing CEREC resin composite luting material. After placement and polishing 10 teeth of each group were thermocycled 2500 times between 14 degrees C and 62 degrees C. Each tooth was bisectioned and from each part, one occlusal and three proximal sections of 75-100 microns thickness were cut and ground (EXAKT Cutting and Grinding system). The marginal discrepancy and the thickness of the luting cement were measured microscopically at each section and an average for each tooth was calculated. Considerable variation of marginal fit was seen, within inlays, and among different types of inlays. In general, the Vita Dur N inlays showed the best fit, e.g. the thickness of the luting cement for Vita Dur N was 114 microns occlusally and 119 microns mesiogingivally compared to 199 microns occlusally for CEREC and 219 microns mesiogingivally for Estilux. At enamel margins, a tendency of less microleakage was seen for ceramic inlays compared to composite inlays.     

The epidemiology of Neisseria gonorrhoeae isolates in Greenland 1979-1990: the emergence, spread and disappearance of non-PPNG strains carrying the conjugative 38.9 kb plasmid

The aim of this study was to compare the clinical performance of an amalgam, a glass polyalkenoate (ionomer) cerment material and a resin-based composite material used in small Class II cavities in permanent teeth. All restorations were inserted under rubber dam. They were examined yearly for 3 years. One clinician continued the study up to 5 years. The clinical examination focused on two criteria: clinically acceptable and failure. In addition, impressions were taken of the prepared cavities immediately before restoration and at each clinical examination using an elastomeric material. The study comprised 274 Class II restorations (88 amalgams, 95 cerments and 91 resin composites) placed in 142 adolescent patients. One hundred and sixty-seven restorations were in molar and 107 in premolar teeth. Patient dropout after 5 years resulted in the loss of 161 restorations, evenly distributed for restorative material and type of tooth involved. Four amalgam restorations, 22 glass ionomer cerment and nine resin composite restorations failed. The glass ionomer cerment and amalgam restorations failed primarily due to bulk fractures, while the resin composite restorations failed due to secondary caries and bulk fractures.     

Contraction gaps in Class II restorations with self-cured and light-cured resin composites

PURPOSE: To evaluate in vitro the cervical gap formation in composite restorations in which an increment of autopolymerizing composite was interposed between the light-cured filling material and the dentin bonding system. Two different self-cured materials and two different adhesives were used and compared. MATERIALS AND METHODS: In 40 Class II cavity preparations with the gingival margins in dentin or cementum, an increment of either of two different autopolymerizing composites, Palfique or Bisfil 2B, was interposed between the light-cured filling material (Palfique Estelite) and either of two different dentin bonding systems, All-Bond and Superbond D (20 restorations each). The specimens were immersed in toluidine blue dye. A section was taken from each specimen and examined under an optical microscope. The extent of dye penetration was measured. The specimens were then stored in water for 1-4 months and replicas of the sections were examined by SEM. RESULTS: Good marginal adaptation was observed for All-Bond/Bisfil 2B in all the restorations. All-Bond/Palfique showed good adaptation in 7 out of 10 cases, the failures occurring between bonding agent and composite. Superbond D/Palfique and Superbond D/Bisfil 2B showed marginal gap formation in 8 out of 10 and in all the restorations respectively. In the latter, four gaps occurred between bonding agent and dentin. As a consequence, a very significant difference (P<0.001) was found between All-Bond and Superbond treatments; no significant difference was found between Bisfil 2B and Palfique in each group treated with either All-Bond or Superbond. After 1-4 months storage in water, hygroscopic expansion of the restorative materials reduced the total number of gaps and the efficacy of the treatments was no more significantly different.     

The use of a new strain of Vibrio cholerae O139 as a producer of an enteric chemical vaccine

Amalgam remains unchallenged as a posterior restorative material. But its inability to bond to the teeth leads to some amount of microleakage at the restoration-tooth interface with associated problems such as post operative sensitivity, pulpal complications etc. Also a broken amalgam restoration requires replacement which will further weaken the tooth structure. Recently, 4-META has been introduced which can graft amalgam and composite to enamel, dentin and old amalgam restorations. In this study, the bonding and marginal sealing abilities of 4-META was assessed both at the tooth-amalgam interface and old amalgam fresh amalgam interface.     

Histological changes induced by CO2 laser microprobe specially designed for root canal sterilization: in vivo study

Sixty-six class-II CAD/CAM-manufactured ceramic inlays (Cerec) were placed in 27 patients. Each patient received at least one inlay luted with a dual-cured resin composite and one inlay luted with a chemically cured resin composite. The inlays were examined 5 years after luting using the California Dental Association (CDA) criteria. Eighty-nine percent of the 66 inlays were rated 'satisfactory'. During the follow-up period replacement was required for 3 inlays because of inlay fractures (4.5%) and 1 inlay because of fracture of the tooth substance (1.5%). All those inlays were luted with the dual-cured resin composite luting agent. Of the remaining 62 inlays the CDA rating 'excellent' was given to 84% for color, 97% for surface, and 81% for anatomic form. 'Excellent' margin integrity was seen in 52% of the dual-cured resin composite luted inlays and in 61% of the chemically cured resin composite luted inlays. No statistically significant (P> 0.05) difference was observed between the two luting agents.     

Electrochemical study of insulating properties of dental amalgam bonding polymers

The standard techniques used for amalgam restorations often result in a lack of adhesion to mineralized dental tissues. The bonding of amalgam with polymer has been suggested to improve its adaptation to dental tissues. Moreover the polymer involved in the bonding should inhibit the corrosion and the diffusion of metallic ions. The aim of this study was to evaluate in vitro the capacity of bonded amalgam to prevent ionic diffusion and migration. In this respect, an original method employing electrochemical techniques was used to determine the leakage current of bonded amalgam restorations. The electrochemical behaviour of conventional and bonded amalgam restorations was compared using a potentiostat driven by a computerized system (Voltamaster, Radiometer Analytical) with software for specific applications such as chronoamperometry or cyclic voltammetry. Samples of recently extracted teeth of young patients were first examined, and then the results were checked by other experimental assays using protected and unprotected copper sticks. The measurements obtained with chronoamperometry (E = +300 mV/SCE) in Ringer's solution at 37 degrees C showed that after polarization for 30 h the oxidation current decreased threefold for bonded samples (10 microA cm(-2)) as compared with the unprotected samples (35 microA cm(-2)). These results, as well as those obtained with the copper wires, demonstrated that even with two layers of adhesive the bonded joint is permeable to ions probably as a result of the hydrophylic properties of HEMA, a component of the adhesive. However, using five layers of adhesive reduced the ionic current by a factor as large as 10(6).     

Hematological changes in infectious diseases. Hematological consequences of bacterial, protozoal and spirochetal infections (2)

Thirty-six Class V amalgam restorations were placed in 29 patients. Prior to insertion of the amalgam in 12 cavities the walls were treated with Amalgam Liner and in another 12 cavities with Copalite. The remaining 12 cavities were not treated before amalgam insertion. The margin quality was determined quantitatively with the SEM using replicas obtained immediately after the polishing of the restorations and 6 months later. At baseline the cavities treated with Copalite showed a margin quality identical to the untreated cavities. Cavities treated with Amalgam Liner showed an inferior margin quality. After 6 months the margin quality of the three groups showed no statistical differences. With neither of the liners tested was it possible to improve the margin quality of amalgam restorations.     

Selection of restorative materials, reasons for replacement, and longevity of restorations in Florida

The purpose of the present study was to obtain information relating to the types of restorative materials used, the main reason for replacement of restorations, and the age of failed restorations. Dentists in general dental practice in Florida were invited to record details from their own work pertaining to restorations. The diagnostic criteria were described and coded. The clinicians were not calibrated in the use of the criteria but they could call in for further explanation if needed. The 27 clinicians involved placed 2,035 restorations of which 53% were replacements of failed restorations. The increased use of resin based restorative material was clearly evident including posterior composites. The clinical diagnosis secondary caries was the most common reason for replacement of amalgam (56%) and composite (59%) restorations. Only discoloration showed a statistically significant difference in the reason for replacement of the two types of materials. The median age of the replaced amalgam restoration was 15 years and that of composite restoration was 8 years.     

Middle digestive cancers in Texas Hispanics: a New World syndrome?

The aim of the study was to evaluate the esthetic performance of direct composite additions in correcting tooth form and position at 5 years. Composite additions were directly placed using the acid-etch technique and enamel bonding on 87 intact maxillary anterior teeth in 23 young patients (12-19 years old). The restorations were made by one operator using an ultrafine midway-filled densified restorative composite. Color slides were made at baseline and 5 years. At the 5-year recall, esthetic performance was assessed clinically by two evaluators at chair-side in subterms of color match, translucency/opacity, surface roughness, and anatomical form. Five additional examiners scored esthetics on the 5-year slides. Of the restorations, 89% were still esthetically satisfactory after 5 years of clinical service. The remaining restorations needed replacement, mainly because of severe loss of anatomical form, to a lesser degree because of severe color mismatch. Central incisors and small unilateral restorations generally showed the best results concerning color match (68% and 74%, respectively) and surface smoothness (84% and 100%, respectively). The slide scores on color match and translucency/opacity were generally similar or somewhat better than the direct clinical scores; however, the difference were not statistically significant (P > 0.05). As far as surface roughness is concerned, the results were significantly better (P < 0.05) when recorded indirectly than by direct clinical evaluation. In conclusion, direct composite additions are a valuable and effective procedure for esthetic and conservative treatment of malformed and misaligned anterior teeth. Loss of anatomical form due to wear points to a shortcoming of the composite material used, with which a durable esthetic result cannot always be guaranteed in the long term.     

Effect of topical timolol on tissue circulation in optic nerve head

OBJECTIVE: The aim of the present study was to determine whether removal of all amalgam restorations might significantly affect mercury levels in plasma and urine and whether the use of rubber dams might reduce patient exposure to mercury during amalgam removal. METHODS: All amalgam restorations were removed from 18 subjects during a single treatment session in which a rubber dam was used and from 10 subjects when a rubber dam was not used. All amalgam restorations were removed by the same dentist using high-speed cutting, water coolant, and high-volume evacuation. The levels of mercury in plasma and urine were analyzed both before and during the subsequent twelve months after amalgam removal. In order to determine whether removal of all amalgam restorations might cause an exposure large enough to significantly increase the mercury levels in two indicator media for mercury exposure, i.e., plasma and urine, and to determine if the removal might cause a significant decrease in the mercury levels found over time, the one-tailed, paired Students' t-test was used. For each individual, the pre-removal levels were compared with both the levels found in plasma on d 1 and in urine on d 10, and also with the levels found 1 y after removal. Furthermore, in order to examine whether the use of rubber dams had any effect on the mercury levels found after removal, the changes in the mercury levels found were compared between the groups using the Wilcoxon-Mann-Whitney rank sum test. RESULTS: After removal of all amalgam restorations, only the non-rubber dam group showed significant increases in the mercury levels found in plasma (p = 0.012) and urine (p = 0.037). However, one year later, the mercury levels in plasma and urine had sunk significantly below the pre-removal levels for both groups. When the changes in the mercury levels found were compared between the groups, the non-rubber dam group showed a significantly higher increase of mercury in plasma than the rubber dam group the day after removal (p = 0.0010). Compared to the pre-removal mercury levels in plasma and urine, the levels found 1 y after removal of all amalgam restorations were on average 52 +/- 23% (range 4-89%) lower in plasma and 76 +/- 21% (range 20-94%) lower in urine. SIGNIFICANCE: The study showed that dental amalgam had a statistically significant impact on the mercury levels found in plasma and urine in the patients tested, and that the use of a rubber dam during removal of all amalgam restorations significantly reduced the peak of mercury in plasma following removal.     

Microleakage of dentin bonding systems used with spherical and admixed amalgams

PURPOSE: To evaluate the effectiveness of the dentin bonding systems (DBS) Tenure, Syntac, Amalgambond and All-Bond 2 using a resin-lined amalgam technique in preventing short-term microleakage in Class V cavity preparations restored with a spherical and an admixed alloy. MATERIALS AND METHODS: Class V cavity preparations were made on the mesial and distal surfaces of non-carious human mandibular molars with the occlusal margins in enamel and gingival margins in cementum. Prepared teeth were randomly assigned to 12 treatment groups (n = 10) consisting of each DBS, no liner and Copalite for both amalgam alloys. Samples were thermocycled, stained and sectioned to evaluate microleakage. RESULTS: The DBS-lined restorations had significantly less microleakage at both occlusal and gingival margins than the Copalite and unlined restorations. Among DBS, Syntac had significantly more microleakage with both alloys than either All-Bond 2 or Amalgambond. These results do not totally correlate to a previous study which used the same experimental design but included viscous liners in addition to the same DBS, indicating a potential difference between resin-lined versus resin-bonded amalgam techniques in preventing microleakage.     

In vitro evaluation of marginal and internal adaptation of class II resin composite restorations after thermal and occlusal stressing

The aim of this in vitro trial was to evaluate the external and internal adaptation of class II composite restorations to tooth structure by means of replica scanning electron microscope (SEM) observation. Standard MOD preparations were cut in human extracted teeth with margins located above and below the cementoenamel junction. Cavities were restored with either a direct multilayered technique, inlays (using conventional or dual DBA application) or prepolymerized composite inserts, using similar restorative material (Syntac, Variolink and Tetric). Evaluations were performed after mechanical and thermocycling. For the marginal adaptation in enamel, the bevelled margins of the direct group resulted in higher percentages of "continuity" (92.5% mesially and 94.6% distally), while the "marginal tooth fracture" was the most commonly found defect in groups with a butt preparation (up to 29.4%, with inserts). For the marginal adaptation in dentin, "continuity" percentages varied from 59% (inlay) to 87.9% (insert). As regards the internal adaptation, results for the whole dentin interface varied from 43.1% (inlay) to 63.9% (inlay with dual bonding) of "continuity". No regional difference in internal adaptation was found between the different preparation areas, except between gingival dentin and gingival enamel (44.9% vs. 80.7% of continuity, all group pooled data). Debonding occurred only at the dentin-restoration interface and consistently took place at the top of the hybrid layer. In the present experimental conditions, the inlay or insert techniques, which make use of the Dual bonding concept, proved to have the best potential to maintain the integrity of dentin-restoration interface.     

Artificial caries formation around fluoride-releasing restorations in roots

Secondary caries is one of the most important factors leading to replacement of dental restorations. This investigation assessed the capacity of fluoride-releasing restorative materials to resist caries in vitro when used in roots. Class 5 cavities were prepared in the buccal and lingual surfaces of 30 extracted premolars. The six materials used were: glass-ionomer cement (Fuji), glass-ionomer cement with silver particles added (Ketac-silver), fluoride-containing composite resin (Tetric), composite resin (Silux plus), fluoride-containing amalgam (Fluor-Alloy) and high-copper amalgam (Dispersalloy). After 5 weeks in an acid gel for caries-like lesion formation, the teeth were sectioned longitudinally and examined with polarized light. The results showed that repair with glass-ionomer materials of a carious lesion may be of great importance in the prevention of secondary caries around the restorations in roots.     

Influence of finishing time on marginal sealing ability of new generation composite bonding systems

Although variation in finishing techniques has been shown to affect microleakage, little research has been published on the influence of finishing time on the marginal sealing ability of new generation composite bonding systems. The objective of the present study was to evaluate the influence of finishing time on the enamel and dentine marginal sealing ability of four new generation composite systems. Two class V preparations, which were solely in enamel or dentine, were made on the buccal surfaces of 96 freshly extracted molar teeth. The teeth were randomly divided into four groups of 24 and restored with composite resin (Silux Plus) utilizing the following bonding systems: Scotchbond Multi-purpose, Fuji Bond LC, Prime & Bond 2.0 and Bisco One-step. Half of the restorations in each group were finished immediately after light polymerization and stored for 1 week. For the remaining restorations, finishing was delayed for 1 week. The storage medium was isotonic saline at 37 degrees C throughout the experiment. All restorations were then thermocycled, subjected to dye penetration testing, sectioned and scored. The results suggest that the finishing of composite restorations, bonded with the bonding systems evaluated, should be carried out immediately after light polymerization. Delayed finishing does not improve but instead can be detrimental to the marginal seal of the restorations. The effects of delayed finishing are, however, both bonding system and tissue specific.