Methods: Fifty smears from oral brush biopsies (35 from patients with and 15 without oral cancer) were examined. After Feulgen restaining, DNA measurements were performed using a DNA-image-cytometer. The following measurements were blinded to the first. Six test slides were additionally prepared from a rat as rat liver imprints and were analyzed according to standardized protocols of the ESACP. For the oral smears the occurrence of stemline- and single cell-aneuploidy, number and modal values of stemlines and CVs (coefficient of variation) of reference cells were compared between both measurements. Reproducibility was assessed by Paired Samples t Test and Kappa Statistics.
Results: All non-neoplastic smears revealed DNA-euploidy and all from oral cancers DNA-aneuploidy. In the group of oral cancers, DNA-single cell- and stemline- aneuploidy was confirmed in all cases by both measurements. The modal values of aneuploid stemlines which contained most of the tumor cells varied up to ±0.15c of the DNA content. The overall diagnostic reproducibility was 100 %, k=1, and the 95% CI=1. The CVs of rat hepatocytes (mean=1,54±0,03) and lymphocytes (mean=0,75±0,30) were found far below the standard set by the ESACP of 5%.
Conclusion: Nuclear DNA measurements on standardized rat liver imprints and normal oral mucosa cells allowed a high precision, far below the thresholds set by the ESACP. The interobserver reproducibility of the diagnostically relevant findings of DNA-euploidy versus aneuploidy in oral smears was 100 %. Differences in modal values of aneuploid stemlines were attributed to the well-known heterogeneity of chromosomal constitution of tumor cells. We conclude that diagnostic DNA-ICM is a highly reproducible method in the non-invasive early detection of oral cancer in brush biopsies.