Correlation of transversal microradiography and microhardness on in situ-induced demineralization in irradiated and nonirradiated human dental enamel.

The objective was to evaluate whether a correlation exists between microhardness (MH) and transversal microradiographical (TMR) data for in situ-induced caries lesions in irradiated and nonirradiated human enamel. Enamel specimens were prepared from the lingual and buccal surfaces of 20 freshly extracted, caries-free third molars. The surfaces were polished (4000 grit). Either the lingual or the buccal specimen of each tooth was irradiated with 60 Gy; the other sample was not irradiated. Two irradiated and two nonirradiated specimens were inserted into both buccal aspects of each of five intraoral mandibular appliances. The appliances were worn by five persons for 6 weeks day and night. One side was brushed daily with a fluoride-free toothpaste; on the other side, dental plaque was allowed to grow. Individual oral hygiene was performed without any fluorides. During meals, the appliance was stored in 10% sucrose solution. Then the enamel specimens were cut perpendicular to their oral surface. The cut surface was polished (4000 grit) and the Knoop hardness number (KHN) was measured across the lesions, at 25, 50 and 75 microm from the oral surface. After MH testing the samples were polished again, thereby reducing the surface by about 10 microm. Subsequently, the slabs were ground to a thickness of 90 microm, and studied by means of TMR. Mineral loss was calculated with dedicated software at the locations corresponding to the MH testing. A total of 120 paired data was submitted to linear-regression analysis. Neither MH nor TMR showed significant differences between irradiated and nonirradiated enamel lesions. A linear relation was found between square root of KHN and the mineral volume percent. In naturally induced caries lesions, MH values can be converted to mineral volume percent using the equation [21.19 + 3.66 x square root of KHN]. This equation fits the data with a reliable correlation coefficient (r2=0.915).