Objective: To determine the performance of different mathematical methods of correcting histograms obtained by densitometric DNA measurement on thin tissue sections for cutting artifacts in different types of human tissue.
Study design: The study group consisted of four different euploid-polyploidizing, nonmalignant human tissues (liver, breast cyst with apocrine metaplasia, adenoma of the thyroid, condyloma). The section thickness was accurately measured. The algorithms of Bins, McCready, Bacus and Haroske were applied to the histograms obtained from DNA measurements. The results were compared to those of the corresponding single-cell preparations.
Results: Correct ploidy equivalents were calculated using the algorithm of Bins in very thin sections (+/- 3 microns). This was also the case applying the formulas of McCready and Haroske to medium-thick sections (5.3-5.9 microns). The algorithm of Bacus did not lead to reliable results. Measurements on sections thicker than approximately 7 microns led to good results without mathematical correction.
Conclusion: Reliable results can be obtained only when knowing the exact section thickness. In very thin sections the method of Bins leads to reliable results; in medium-thick sections the methods of McCready and Haraske result in correct ploidy values. Sections thicker than 7 microns do not need histogram correction.