Significance of variation in DNA flow cytometric analyses from paraffin-embedded breast cancers. Evaluation of the grading efficiency of ploidy determination, DNA index, and S-phase fraction.

116 paraffin-embedded breast cancer samples were analyzed by flow cytometry. From each sample 3 consecutive 50 microns sections were cut for the study. The presence of neoplastic tissue was verified from light microscope sections cut before and after the adjacent sections. One laboratory started with one section from each block and was allowed extra sections when needed for analysis. At the end the laboratory obtained results from all 116 cases. The rest of the samples were studied by 2 other laboratories. Samples with results from 2 or 3 laboratories then allowed variability analysis and the estimation of the grading efficiency in a 2-grade system. Inconsistency in diploidy/aneuploidy distinction was present in 36 of 111 (32.4%) cases studied by two or three laboratories. This inconsistency was less obvious in samples graded as multiploid by at least one of the laboratories. The grading efficiency as analyzed from the results of 3 laboratories was 0.89, and of 2 laboratories, 0.84. The DNA index showed a slightly higher grading efficiency. At the cutoff point of 1.3, 91% of cases could be expected to be correctly classified into low ploidy and high ploidy groups (grading efficiency 0.91). The S-phase fraction had a mean grading efficiency of 0.89, a performance comparable to that of diploidy/aneuploidy distinction. In the light of the available data the flow cytometric analysis can add to the consistency of grading, especially when compared with subjective histological grading. However, the data do not suggest that flow cytometric analysis of paraffin sections as a grading method would be more consistent than quantitative histopathology from sections.