Should Dukes' B patients receive adjuvant therapy? A statistical perspective

Abstract

The benefit of adjuvant therapy, such as 5-fluorouracil (5-FU) combined with leucovorin, is a matter of debate for patients with Dukes' B colon cancer. Several approaches have been taken to address this issue. Initially, studies were conducted to assess treatment benefits in both Dukes' B and Dukes' C patients. These studies identified an overall benefit of adjuvant treatment and enrolled enough Dukes' C patients to determine a treatment benefit for adjuvant 5-FU/leucovorin in this subpopulation. However, the individual studies were insufficiently powered to detect a treatment benefit in Dukes' B patients. An analysis of four separate studies (National Surgical Adjuvant Breast and Bowel project) compared the benefit of adjuvant treatment in Dukes' B patients with that in Dukes' C patients and showed similar relative reductions in mortality and disease-free survival in Dukes' B and in Dukes' C patients. The Liver Infusion Meta-Analysis Group also reported similar relative benefits from a portal vein infusion of 5-FU-based chemotherapy in Dukes' B and Dukes' C patients. The International Multicenter Pooled Analysis of Colon Cancer Trials B2 study, which combined data from patients with Dukes' B colon cancer in five separate trials, failed to show a statistically significant benefit of adjuvant 5-FU/leucovorin compared with surgery alone. We review the advantages and limitations of different approaches to detect treatment benefits in patients with Dukes' B colon cancer, and we argue that there is a need for a meta-analysis of all adjuvant trials to reliably address this question.