[DNA flow cytometry is a useful prognostic guide in the treatment of breast cancer.]

It is widely agreed that the presence or absence of axillary lymph-node involvement (N) is the most reliable predictor of relapse or survival in breast cancer, together with tumor size (T) and the presence or absence of distant metastasis (M). These prognostic factors are the cornerstones of the TNM staging system. The aim of the present study was to ascertain, in all patients diagnosed with invasive primary breast cancer in Iceland during the years 1981-84 (n=347), whether flow cytometric DNA analysis of ploidy status and fraction of cells in the S-phase contribute prognostic information, addi nottional to that obtained with TNM staging variables. Paraffin fixed tumor material was available from 340 patients (98%) and DNA ploidy and S-phase fraction was assessed with flow cytometry. DNA ploidy could be analysed in 98% of tumor samples (n=334), of which 114 (34%) were diploid and 220 (66%) non-diploid. S-phase fraction could be analysed in 97% of the tumor samples (n=329), the median S-phase value was 7.0%, and was higher in non-diploid than diploid tumors (p<0.0001, 9.3% vs. 2.7%). Median duration of patient follow-up was 7.5 years. The disease-free survival at that point of time was 15% higher in patients with diploid tumors than non-diploid ones (p=0.004, 69% vs 54%). Similar survival comparison in relation to S-phase fraction was 30% when the median S-phase value was used as cut-off point (p<0.0001, S-phase<7.0% being 74% vs. S-phase ;7.0% being 44%). Multivariate analyses with regard to breast cancer survival and disease-free survival, which included both ploidy status and S-phase categories adjusting for age, tumor size and lymph node involvement, showed the S-phase value categories to be independent prognostic variables (p<0.0001). Patients with high S-phase tumors had a three-fold higher risk of recurrence than patients with low S-phase tumors. Ploidy status was not an independent prognostic variable, if however the S-phase categories were excluded from analysis, ploidy status was on the borderline of being an independent variable (p=0.09). In node-negative patients the S-phase fraction was the only useful variable in determining prognosis. We conclude that the S-phase value is a useful prognostic guide for the clinician and will be used for this purpose in the treatment of breast cancer in Iceland.