The importance of heterogeneity and of multiple site sampling in the prospective determination of deoxyribonucleic acid flow cytometry.
Multiple fresh specimens from 59 nonsmall cell carcinomas of the lung, 38 carcinomas of the gastric tract and 55 carcinomas of the colon and rectum were analyzed by deoxyribonucleic acid (DNA) flow cytometry (FC) after radical resection to evaluate tumor ploidy as an independent prognostic factor. The minimum follow-up period was five years (range of five to ten years). Aneuploidy was observed in 98.0 percent of carcinomas of the lung, in 70.9 percent of carcinomas of the colon and rectum and in 63.1 percent of carcinomas of the gastric tract. FC DNA heterogeneity, in terms of different number of DNA stem lines or different DNA indices between core and periphery, or both, was found in 50.0 percent of carcinomas of the lung, 47.0 percent of carcinomas of the colon and rectum and in 34.5 percent of carcinomas of the gastric tract. A diploid pattern was more frequently observed in less advanced stages of the gastrointestinal tract. By univariate analysis (Kaplan-Meier), patients with carcinoma of the lung with hypodiploid or hypertetraploid peaks, or both, and aneuploid gastric tumors had poorer prognosis. These differences were only marginally significant. Cox analysis demonstrated that the single most important prognostic variable for predicting the overall survival rate was the stage of disease. Tumor DNA content can be considered a marker of advanced stages, particularly in tumors of the gastrointestinal tract, but there is no evidence that it is an independent prognostic variable able to predict long term survival in patients who have been radically resected.