RESUMO
OBJECTIVES: To assess leakage of class-I restorations using a gas-enhanced permeation test (GEPT) as compared with conventional SEM or dye analysis. MATERIALS AND METHODS: Pressure differences over time and penetrating water volumes were measured simultaneously in a two-chamber system (GEPT) before and after class I cavity preparation in 30 molars. Ten teeth were restored with a composite restoration without bonding (A1), a composite restoration with bonding (A2) or a ceramic indirect restoration (B). Five intact teeth served as controls (C). Another GEPT measurement was performed and impressions were taken. Teeth were subjected to thermodynamic loading (1 200 000 cycles) and final GEPT measurements and impressions were made. SEM evaluation of the marginal continuity was performed and teeth were subjected to a Fuchsin dye penetration test. Spearman's rank test was used to compare results from different tests. RESULTS: The GEPT and SEM values did not correlate before loading (0.359, p = 0.051), but significantly correlated afterwards (0.662, p < 0.0001). The correlations between the Fuchsin dye penetration test and GEPT and SEM surface marginal analysis were significant (0.777 and 0.534, p-values < 0.0001 and 0.002, respectively). CONCLUSIONS: SEM marginal analysis was mainly limited in reflecting the surface restoration integrity. GEPT evaluation may, therefore, serve as a tool to non-destructively assess restoration sub-surface integrity over time. CLINICAL RELEVANCE: The current study provided proof that restoration margin quality does not necessarily reflect its leakage behaviour.