Reproducibility of DNA measurements in imprints of thyroid adenoma. Variation sources.

DNA measurement represents a type of quantitative analysis which allows us to gain prognostic information on malignant tumors and to study the natural history of the epithelial neoplasia. However, there are sources of variation in evaluating the DNA content. These include variation due to field selection, variation between observers (interobserver variation), and variation between laboratories (interlaboratory variation). The influence of various variation sources was studied in 4 experiments. When DNA measurements were made from the same microscopic fields, the results did not differ remarkably. However, observer training proved to be important. Intra- and interobserver variation was lower among experienced morphometrists than among inexpert observers. Different laboratories and image analyzers may give different results when the same case is measured. To overcome at least part of the potential variation sources, undergraduates and postgraduates at the Departments of Pathology of Ancona and Kuopio Universities are specially trained in the use and application of morphometry. Special sampling rules should be applied and observers encouraged to follow the rules as uniformly as possible.