DNA heterogeneity determined by flow cytometry in prostatic adenocarcinoma--necessitating multiple site analysis.

Although DNA ploidy analysis of prostate cancer is generally associated with grade, stage, clinical outcome, and responsiveness to androgen therapy, one possible reason cited for contrary reports may be tumor heterogeneity. A preliminary report using flow cytometric analysis of punch biopsies demonstrated DNA heterogeneity in five of nine patients. We evaluated 75 patients by cutting whole mounts of formalin fixed prostatectomy tissue every 0.6 cm. All malignant areas and a selected normal area were circumscribed, excised, remounted, and 1-3 50 mu thick sections removed. The nuclei were extracted by a Hedley technique and the DNA stained with propidium iodide. Each whole mount had an average of 1 distinct malignant area (range of 1-6 areas per whole mount block). Nuclei were analyzed on a Becton Dickinson (San Jose, CA) FACScan flow cytometer equipped with RFIT DNA software program. After excluding histograms with CVs > 8.0% and/or "suspicious" diploid histograms having a right "shoulder," 75 or 87 patients still had > or = 2 malignant sites available for analysis (average 4, range 2-9 malignant sites/patient). The 322 histograms had an average CV of 4.4%. Thirty of 75 patients (40%) showed DNA heterogeneity in multiple samples taken from the same prostate. There were 37 prostates with only diploid (D), 1 with only tetraploid (T), 7 with only aneuploid (A), 20 with D plus A, 7 with D plus T, 2 with D plus T plus A, and 1 with a D plus suspected hypodiploid DNA content. Exclusion of the tetraploid and "near diploid aneuploid" cases still resulted in 16% (12/75) of the patients having a diploid versus aneuploid DNA content heterogeneity. Because 40% of the prostates contained a different ploidy depending on which area was sampled, this report suggests multiple sites of malignancy must be analyzed to more accurately assess the ploidy status of prostatic adenocarcinoma.