Material and methods

One hundred premolars were collected from various oral surgery practices in southeast Michigan. Upon collection, the teeth were washed in tap water and stored at 4°C in 0.1% thymol solution. Only carefully selected premolars with a healthy facial enamel surface were included.

The premolars were randomly assigned to 5 groups of 20 each. In the first group, the brackets were bonded with Aegis-Ortho resin cement. In the remaining 4 groups, the brackets were bonded with Transbond XT (3M Unitek, Monrovia, Calif). Of the 4 groups receiving the Transbond XT bonding, 1 served as the control group and 3 as the experimental groups: Vanish fluoride varnish (5% sodium fluoride white varnish [3M, St Paul, Minn]), MI Paste, and Pro-seal.

The roots of the premolars were sectioned and discarded, and the remaining crowns were embedded in acrylic.

The bracket was placed 2 mm gingivally to the buccal cusp tip and in the mesiodistal center of the clinical crown.

To standardize and limit the enamel area exposed to the etching and bonding procedures, the enamel surface was protected by adhesive tape during all adhesive procedures. By using a hole puncher, a window was cut from the adhesive tape, leaving an enamel area the size of the orthodontic bracket base exposed. A metal orthodontic bracket was bonded to each tooth by using a light-cured bonding resin, depending on the group assignment. Universal premolar brackets, Victory Series (3M Unitek) with a 0.022-in slot and 0° of tip and torque, were used in this study.

Each tooth was cleaned with nonfluoridated pumice with a rubber prophy cup on a slow-speed hand piece. Remaining residue was washed away with tap water, and each tooth was thoroughly dried for 15 seconds. For all groups, the enamel window in the tape was etched for 30 seconds with 35% phosphoric acid gel (Bosworth), rinsed with water for the same amount of time, and dried with compressed air.

For the Aegis-Ortho group, Aqua bond (Bosworth) was used as the primer. For the remaining groups having Transbond XT adhesive cement, Transbond XT primer (3M Unitek) was selected as the bonding agent. Immediately after the adhesive was applied, each bracket was placed onto its tooth surface, and excess adhesive was removed. The resin was light-cured with an Ortholux LED curing light (3M Unitek) for 10 seconds. The adhesive tape was removed from each tooth. To avoid interference of general tissue demineralization of the crown, the entire tooth was covered by varnish except for the area of interest. All teeth were painted with a thin coat of acid-resistant nail varnish (Revlon, New York, NY) on all surfaces except the 1.0 to 1.5 mm area gingival to the bracket, leaving a small rectangular box exposed for demineralization ( Fig 1 ). For quantitative light-induced fluorescence (QLF) measurements, this exposed enamel was compared with the nail varnish-covered area of unexposed enamel just gingival to the box.

Bracketed premolar with nail varnish painted around the exposed enamel surface. The dotted rectangle represents the area of interest. Asterisks ( ^∗ ) indicate the control area for QLF measurements, covered with nail varnish.

Teeth in the varnish group were dried, and the exposed enamel box was painted with Vanish. The varnish was allowed to dry for 5 minutes. Teeth in the CPP-ACP group were also dried, and the exposed enamel in the box was painted with a pea-size amount of MI Paste with a cotton swab. It was left undisturbed for 3 minutes. The MI Paste on each tooth was removed with a cotton roll and allowed to sit for another 2 minutes. Finally, the remaining residue was washed away with tap water. Following the manufacturer’s instructions, we applied the MI Paste daily during the experiment. The teeth in the resin sealer group were acid-etched in the exposed enamel box, and a layer of Pro-seal resin was applied. The Pro-seal was cured for 20 seconds with the curing light.

The groups were cycled in a demineralization solution for 8 hours per day. The demineralization solution was based on lactic acid and Carbopol C907 (Lubrizol, Wickliffe, OH) (pH 5.0), 50% saturated with hydroxyapatite. The solution was refreshed daily during the experimental period of 15 days.

After 8 hours of exposure to the caries-like challenge, all teeth in each group were rinsed with water and placed in artificial saliva for 30 minutes. The teeth were cleaned in no particular order with a Sonicare FlexCare power toothbrush (Philips Electronics North America Corp, Andover, MA). The brush bristles were rinsed with water, and a pea-size amount of Crest Cavity Protection toothpaste (1,100 ppm sodium fluoride, Procter & Gamble, Cincinnati, Ohio) was applied to the brush. The brush was lightly applied to the exposed enamel box for 2 seconds per tooth. After the 5 groups were separately brushed and rinsed with tap water, they were placed once again in the artificial saliva solution for approximately 15 hours until the cycle was repeated. With the MI Paste group only, the teeth were removed from the artificial saliva solution 10 minutes before the next demineralization period started to receive an MI Paste application.

On day 15, all teeth were removed from the saliva solution, rinsed under tap water, and stored in 100% humidity. The remaining fluoride varnish residue on the teeth from the Vanish group was removed with a plastic scaler.

To assess demineralization, QLF and confocal laser scanning microscopy (CLSM) were used. Both procedures were carried out at the Oral Health Research Institute, Indiana University, Indianapolis, Ind. The use of QLF as a quantitative measurement tool has the advantage of being noninvasive, and it can be applied clinically. CLSM, on the other hand, using sections of teeth, allows for a more accurate, quantitative analysis of surface and subsurface areas, providing enamel lesion depth data in larger tissue sections. These methods are complementary.

QLF images were taken of all teeth. To calculate fluorescence loss, a computer-generated rectangle (approximately 1.5 × 4.0 mm ² ) captured the enamel gingival to the bracket region. The fluorescence values of the pixels in the rectangle exposed to the treatments and cycling were compared with the fluorescence values of a control area of a similar size: a relevant enamel area that was previously covered by nail varnish during the cycling. The value obtained was the average change in fluorescence (fluorescence loss:ΔF) of the pixels under the threshold.

For CLSM, a casting resin (Meyers Plastics, Indianapolis, Ind) was applied on top of each specimen and allowed to polymerize overnight. An approximately 1-mm thick section was obtained from each specimen by using a Silverstone-Taylor hard-tissue microtome (Scientific Fabrications Laboratory, Lafayette, Colo). Each section was stained overnight with a 0.1-mmol/L solution of Rhodamine B (Aldrich Chemical, Milwaukee, Wis). The stained surfaces of each specimen’s section were allowed to air dry before being analyzed with a confocal microscope (LSM510-META, Carl Zeiss Microimaging, Thornwood, NY) to determine the extent of the lesions. The images were examined with Metamorph software (version 4.1.6, Universal Images, West Chester, Pa). After being brought into focus (by using a 10-times Plan Neofluar objective, N.A. 0.3, Carl Zeiss Microimaging), the specimens were illuminated with an argon laser with a 543-nm excitation wavelength. Confocal slits were set at 82 μm with a 580 to 633 nm filter, and the argon laser intensity was set at 20%. Areas were scanned planoparallel to the transversal cut surface of the specimen and perpendicular to the natural surface of the tooth. The value obtained was the lesion depth in the area of exposed enamel.