Reliability of outer lesion secondary caries as a predictor of wall lesions

To assess the reliability of in vitro developed secondary caries outer lesions as an indicator of presence and extent of wall lesions.

Class I occlusal cavities were cut in 132 extracted premolars: 12 teeth were not restored, 120 restored with one of 20 different restoration combinations of silver amalgam, base and varnish. Specimens were aged in 1% NaCl for up to 1 year where after 80 restored and 12 unrestored teeth were subjected to an in vitro bacterial challenge for 36 days, the other 40 specimens were similarly incubated in an acidified broth (pH = 4.0). Sections were prepared for polarized light microscopy and outer and wall lesion areas, outer lesion width, wall lesion length and enamel thickness measured. Data were subjected to Fisher's exact test and Pearson Correlation.

There was a significant association (P < 0.05) between outer and wall lesions in all cariously challenged specimen groups. Sensitivity was high 95-96% but specificity was 26% for acid broth, 68% for restored bacterial and 100% for unrestored specimens. Restoration variables impacted widely on specificity. Correlation values were too low to show clinically relevant associations between wall and outer lesion size. Frequency distribution showed that wall lesions extending beyond the dentin-enamel junction were generally associated with potentially remineralizable outer lesions.