Goldberg, Richard M.

ABSTRACT: Prospective clinical trials are the gold standard for evidence-based methodology used to support changes in the practice of medicine. Clinical researchers, regulatory agencies, payers, and the public embrace the conduct of phase I, II, and III clinical trials as integral to improving patient care. The National Cancer Institute (NCI) funds a number of cooperative oncology groups to conduct such clinical trials in the United States. In order to protect enrolling patients, the NCI requires expedited reporting to allow rapid identification of severe side effects on NCI-sponsored clinical trials. However, chemotherapy drugs frequently cause predictable side effects, the rapid reporting of which would potentially overwhelm the system. This article describes the development and documents the performance of a real-time toxicity reporting system implemented by the North Central Cancer Treatment Group. The goal of this system is to supplement the currently required NCI adverse event monitoring procedures and to permit study teams to identify the need to modify ongoing clinical trials. The system has proven its value in the monitoring of phase II and III trials, including trial N9741, a three-arm, phase III, advanced colorectal cancer chemotherapy study exploring combinations of irinotecan, oxaliplatin, and fluorouracil. We believe the methods described present opportunities for improving patient safety in clinical research.

PURPOSE: To evaluate the value of thymidylate synthase (TS), Ki-67, and p53 as prognostic markers in patients with Dukes’ B2 and C colon carcinoma. METHODS: We conducted a retrospective analysis to evaluate the prognostic value of TS, Ki-67, and p53 in 465 patients with Dukes’ B2 (220 patients) or Dukes’ C (245 patients) colon carcinoma. Patients represent a nonrandom subset obtained from five randomized phase III trials and were treated with either surgery alone (151 patients) or surgery plus fluorouracil-based chemotherapy (314 patients). All three markers were assayed using immunohistochemical techniques. RESULTS: With a minimum follow-up of 5 years, our retrospective analysis failed to demonstrate a consistent and significant association between TS, Ki-67, or p53 and either disease-free survival or overall survival. Exploratory analyses did not reveal a convincing explanation for these results that are in conflict with the published literature. Notable interactions were observed. In particular, high Ki-67 levels were associated with increased (decreased) survival in patients with low (high) TS intensity. Patients whose tumors stained positively for p53 seemed to benefit substantially from the use of adjuvant chemotherapy compared with those who were not treated (P = .05). CONCLUSION: This retrospective investigation failed to demonstrate a significant association between TS, Ki-67, or p53 staining and clinical outcome.