Prognostic value of image and flow cytometric DNA ploidy assessments in invasive breast cancer

Conflicting results have been reported as to the prognostic value of DNA ploidy assessed by image cytometry (ICM) in breast cancer. The aim of this study was therefore to evaluate the prognostic value of ICM-DNA ploidy in a retrospective group of 118 invasive breast cancer patients with long term follow up, in comparison with flow cytometric (FCM) DNA ploidy and other established prognostic factors. Cell suspensions were prepared from paraffin blocks and partly used for FCM, partly to prepare cytospins for ICM after Feulgen staining. At least 1000 cells were fully automatically measured by ICM. All DNA histograms were interpreted with the MultiCycle program. Based on the ICM-DNA index, cases were classified in 5 subclasses as follows: DNA diploid (only one cell cycle present), DNA low aneuploid (1 <DNA-index £ 1.8), DNA tetraploid (1.8 <DNA-index £ 2.2), DNA high aneuploid (DNA-index > 2.2), and DNA multiploid (more than 2 cell cycles). Eleven cases (9%) provided poor ICM-DNA histograms with very broad peaks, which could not be interpreted. For the remaining 107 cases, the ICM-DNA index correlated well with the FCM-DNA index (R = 0.81). In 69 cases (65%), there was complete concordance between ICM and FCM determination of the DNA ploidy class. When grouping into DNA diploid versus DNA non-diploid, the concordance was 83%. Of the 38 cases that were DNA diploid by ICM, 8 (21%) were DNA non-diploid by FCM, and of the 40 cases DNA diploid by FCM, 10 (25%) were DNA non-diploid by ICM. In univariate survival analysis, ICM-DNA ploidy did not provide significance for either grouping system: the above mentioned 5 subclasses, DNA diploid/tetraploid vs DNA aneuploid, or DNA diploid vs DNA non-diploid. The same applied for FCM-DNA ploidy. In multivariate survival analysis, DNA ploidy measured by either ICM or FCM did not have additional prognostic value to the Multivariate Prognostic Index (a combination of mitotic index, lymph node status and tumour size). In conclusion, it is possible to reliably assess DNA ploidy by fully automated ICM measurement of a large number of cells (1000 or more) in a time comparable to FCM. Neither FCM nor ICM-DNA ploidy had prognostic value in this retrospective long term follow up study in invasive breast cancer.