It is estimated that in 1993 there will be > 155,000 new cases of adenocarcinoma of the colon and rectum in the United States (1). Surgical resectability offers the best chance of cure, and the results of surgical and pathologic staging are important factors in deciding whether the patient should be offered adjuvant chemotherapy in the case of colon cancer or combined chemotherapy and radiation therapy for patients with rectal cancer. Because there is significant variation in the rate of recurrence and overall survival within TNM stage categories, there is a need for additional, independent factors to better predict outcome in patients that have had a potentially curative surgical resection.To assess the available data on the possible clinical utility of DNA flow cytometry, a Consensus Conference was recently held. The aims of this work, which reports the conclusions from the ColodRectum Organ System Group that participated in the conference, are: (1) to analyze the current published literature regarding the clinical utility of flow cytometric measurements of DNA ploidy and cell proliferative activity (PA, based on %S or %S + G2M) as prognostic factors for recurrence and survival, and (2) to suggest guidelines and standards for future investigation of DNA ploidy and PA so that these measurements can be appropriately applied to specific colorectal cancer patient care situations.The majority of studies in the literature (2-23; Table 1) suggest that DNA ploidy has significant prognostic value. There are no reports demonstrating DNA aneuploidy as a favorable feature, but several reports (e.g., 3,7,14,18,19,20,21,24) have concluded that DNA ploidy provides no additional significant prognostic information, particularly when considered following multivariate analysis that includes traditional prognostic factors. With exceptions (e.g., 7,17,25), most studies suggest a trend toward an increasing frequency of DNA aneuploidy in higher Dukes stages and for tumors located in the left colon and rectum. Although there are exceptions, most of the published studies suggest that DNA ploidy and PA are unrelated to numerous clinical and pathologic observations in colorectal cancers (26). These include evaluation of age, tumor grade, infiltrative growth pattern, host inflammatory and desmoplastic response, presence of vascular invasion, carcinoembryonic antigen, or mutant p53 status. Fewer studies have examined the prognostic significance of PA. As summarized in Table 2, these series with one exception (7) have found PA to be a significant prognostic factor in colorectal cancer. These conclusions were reached despite the use of four different mathematical methods for S-phase modeling, some with debris subtraction and some without. This emphasizes the need for careful methodological description and, wherever possible, standardization of methods as described subsequently.Contradictory study results have been obtained following the use of a wide variety of technical and analytical methods, many of which may compromise the accurate det...