Correlation between ICDAS and histology: Differences between stereomicroscopy and microradiography with contrast solution as histological techniques

Abstract

Detection of occlusal caries with visual examination using ICDAS correlates strongly with histology under stereomicroscopy (SM), but dentin aspects under SM are ambiguous regarding mineral content. Thus, our aim was to test two null hypotheses: SM and microradiography result in similar correlations between ICDAS and histology; SM and microradiography result in similar positive (PPV) and negative predictive values (NPV) of ICDAS cut-off 1-2 (scores 0-2 as sound) with histological threshold D3 (demineralization in the inner third of dentin). Occlusal surfaces of extracted permanent teeth (n = 115) were scored using ICDAS. Undemineralized ground sections were histologically scored using both SM without contrast solution and microradiography after immersion in Thoulet's solution 1.47 for 24 h (MRC). Correlation between ICDAS and histology differed from SM (0.782) to MRC (0.511) (p = 0.0002), with a large effect size "q" of 0.49 (95% CI: 0.638/0.338). For ICDAS cut-off 1-2 and D3, PPV from MRC (0.56) was higher than that from SM (0.28) (p< 0.00001; effect size h = 0.81), and NPV from MRC (0.72) was lower than that from SM (1,00) (p < 0.00001; effect size h = 1.58). In conclusion, SM overestimated the correlation between ICDAS and lesion depth, and underestimated the number of occlusal surfaces with ICDAS cut-off 1-2 and deep dentin demineralization.

Fig 1. Histology of ICDAS score 4.

Internal histological features of an occlusal surface with ICDAS score 4 visualized under SM (A), microradiography without contrast (B), and MRC (C), with similar lesion depth in all techniques. The contrast solution infiltrated demineralized dentin following the paths of the dentinal tubules. Both sites pointed by the black and white arrows have the same aspect (dentin sclerosis) under both microradiography without contrast (B) and MRC (C), but under SM those sites have different aspects: the site pointed by the black arrow has white opaque aspect, and the site pointed by the white arrow has a translucent aspect (A).

Fig 2. Disagreements between SM and MRC.

Histological lesion depths under SM and MRC images of grounds sections of occlusal surfaces with different ICDAS scores (arrow indicates the location of ICDAS on occlusal surface): ICDAS score 0 with histological score 0 under SM (A) and histological score 4 under MRC (B); ICDAS score 1 with histological score 2 under SM (C) and histological score 4 under MRC (D); ICDAS score 2 with histological score 2 under SM (E) and histological score 4 under MRC (F); and ICDAS score 3 with histological score 2 under SM (G) and histological score 4 under MRC (H).

Fig 3. Correlation between ICDAS and histology (SM and MRC).

Plots of ICDAS scores against the histological scores measured by SM (A) and MRC (B), along with corresponding correlation coefficients, 95% confidence intervals, statistical significance, and power. Most data points are overlapped; this is why there are a few filled circles.