Correlation between ICDAS-Visual and ICDAS-Histology and comparison between Hemisections of Proximal Sound Surfaces and Initial-caries Lesions

Objectives: To establish the correlation between ICDAS-visual and ICDAS-histology assessments and compare the histological lesion depth between longitudinal-paired hemisections in proximal sound surfaces and initial-caries lesions.
Methods: Given Ethical-Board approval and signed-consent forms, for another study an expert in the ICDAS system selected 60 extracted premolars with 49 ICDAS-2 caries lesions (distinct-visual-change-in-enamel) and 11 ICDAS-0 (sound-surfaces). Teeth were longitudinally hemisected at the center of the lesion/surface, using a diamond saw (Isomet-1000, Buehler®), giving a total of 120 hemisections. Right/left stereomicroscopic-hemisection images were captured (Olympus-SZ2-ILST®). Using the Motic-Images-Plus-2.0 software® 3 examiners, by agreement, marked with ruled lines: the center-lesion external point, the related enamel-dentin junction (EDJ), and the maximum lesion depth (μm). The histological lesion depth was assessed (ICDAS-histological scores: 0 No-enamel-demineralization, 1 Enamel-demineralization-limited-to-outer-enamel-half, 2 Demineralization-between-enamel-50%-dentine-outer-1/3, 3 Demineralization-involving-middle-dentine-1/3). The software calculated in each hemisection the lesion depth (μm). Statistical analysis was conducted to calculate correlation between diagnosis methods (Spearman correlation coefficient) and differences of lesion depth between hemisections (U-Mann Whitney test).
Results: Sound surfaces did not show histological demineralization. Most lesions were ICDAS-histology code 1 (46.5%) and code 2 (46.5%), and 7% were code 3. Correlation between visual and histological assessments was substantial (r_s = 0.72). Carious lesions had an average depth of 604.98 ± 365.14 μm for left- and 664.35 ± 719.78 μm for right-side images, without significant difference between hemisections (p>0.05).
Conclusions: The correlation between ICDAS-visual and ICDAS-histology in this sample was substancial. Laboratory conditions, like over-drying of extracted teeth over time or loss of dental structure during the cutting process could have conduced to non-correlation cases, giving importance to care taken for inclusion/exclusion criteria of in vitro studies. On the other hand, either hemisection can be used for these studies.