Flow cytometry of prostate cancer: relationship of DNA content to survival

Between March 1970 and December 1978 there were 366 patients with prostatic cancer treated by 125I seed implants and pelvic lymph node dissection. All had a minimum of 5 years follow-up. One hundred thirty-three patients had metastatic prostatic cancer in lymph nodes (Stage D1) at the time of lymph node dissection and seed implantation. Ninety-one of the 133 patients were judged to have sufficient metastatic prostatic cancer in their nodal tissue (greater than 50% replacement with tumor) to justify flow cytometric cellular DNA measurements on the involved paraffin-embedded nodal tissue. Nine patients were excluded due to uninterpretable DNA histograms leaving 82 patients for analysis. Forty-nine patients had aneuploid and 33 had diploid tumors. There was no statistical bias between the aneuploid and diploid groups due to age (P = 0.970, chi 2 test), time between diagnosis and implantation (P = 0.217, chi 2 test), number of positive nodes (P = 0.669, two-sample t test of means), or tumor grade (P = 0.332, chi 2 test). Median survival time of the aneuploid and diploid groups was 5.0 and 8.8 years, respectively (P = 0.0109, log rank test). Cox regression analysis confirmed the effect of aneuploidy versus diploidy on survival by controlling for other potentially confounding variables (age, time from diagnosis to implantation, number of positive nodes, and grade). Grade as a predictor of survival did not approach statistical significance in this series of relatively small size (P = 0.116). Thirty-eight of the 82 patients had moderately differentiated neoplasms. Nineteen of these were aneuploid and 19 diploid. The median survival was 5.8 and 9.1 years, respectively, for these grade-matched aneuploid and diploid groups (P = 0.039, log rank test). We conclude that flow cytometric DNA measurements on archived paraffin-embedded tumor in nodal metastases appear to be a strong predictor of survival for Stage DI prostatic cancer.