ABSTRACT
A question of current relevance in oncology is whether treatment with chemotherapy following the removal of a tumor is beneficial in patients with Dukes’ stage B2 colon cancer. The patients considered here have had a complete resection of their tumor, and at the time of resection, had no lymph-node involvement or distant metastases. Previous randomized controlled clinical trials conducted by major groups in Europe and North America have demonstrated a beneficial effect on overall survival from a
chemotherapeutic regimen containing two drugs, 5-fluorouracil (5-FU) and leucovorin (LV), in a broader group of patients that includes Dukes’ B2 patients. However, these trials were not designed or adequately powered to address whether the treatment is beneficial in the subgroup of B2 patients. These patients comprise less than 50% of the clinical-trial population, so a pooling of several datasets is necessary in order to provide sufficient power to detect a benefit. In this chapter, we will consider the pooling of individual-patient survival times from five randomized clinical trials that included both a no-treatment arm and at least one arm containing those treated with 5-FU+LV. Further details of the dataset are presented in Sec. IV.
