Patients and Methods: One thousand sixteen patients with B2 colon cancer entered onto five separate trials were randomized to FU + LV or observation, A pooled analysis for event-free (EFS) and overall survival (OS) using a stratified lag-rank and Cox model was performed.

Results: The median follow-vp duration was 5.75 years. Patients receiving FU + LV did not experience a significant increase in EFS or OS. The hazards ratio at 5 years was 0.83 (90% confidence interval, 0.72 to 1.07) for EFS and 0.86 (90% confidence interval, 0.68 to 1.07) for OS. The 5-year EFS was 73% for controls and 76% for FU + LV.The 5-year OS war 80% for controls and 82% for FU + LV. Increasing age and pearly differentiated tumors were significant indicators of poor prognosis (P <.02).

Conclusion: This data set does not support the routine use of + LV in all patients with B2 colon cancer. ranger follow-up may identify a small benefit. At present, studies in B2 colon cancer designed with a no-treatment control arm should be considered appropriate. J Clin Oncol 17:1356-1363. (C) 1999 by American Society of Clinical Oncology.