Purpose: To compare two different enamel pretreatments and their effect on the efficiency of penetration of a one-component adhesive into natural carious lesions.

Methods: Eight extracted human molars and premolars with non-cavitated interproximal lesions were selected. ICDAS code 1-2 was assessed by visual, microscopic, X-ray and Diagnocam record analysis. Samples were cut vertically across the demineralization to obtain two symmetrical lesions, (n=16). After isolating the cut surfaces with nail varnish, paired lesion halves' surfaces were pretreated with two different techniques: Group 1: surfaces were firstly abraded with fine diamond-coated metallic strips (Steelcarbo) and then etched with 37% H₃PO₄ acid (Omni-etch, 120 seconds); Group 2: lesion surfaces were etched with 15% HCl acid (Icon-etch, 120 seconds). All teeth were stained with rhodamine isothiocyanate (RITC) solution (12 hours) and subsequently stored in dry chamber (3 hours). All samples were penetrated with a one-component adhesive (Scotchbond Universal) for 180 seconds and coated with a thin layer of flowable composite (Tetric Flow). After light curing, un- encapsulated dye was bleached by immersion in 30% hydogen peroxide for 12 hours at 37°C. Remaining lesion pores were stained with sodium fluorescein solution. Thin cuts of the teeth were observed with confocal microscopy and computer image analysis was performed (ImageJ). The percentage of penetration (area of resin penetration/area of total demineralization x100) was calculated.

Results: Pretreatment with fine aluminum oxide-coated metallic strip followed by 37% H₃PO₄ acid showed a larger infiltration area (51.7% ± 12.2) in almost all samples compared to pretreatment with 15% HCl acid alone (22.1% ± 13.2). Statistical analysis using t-test showed a significant difference between the two groups (P = 0.011).