Induction chemotherapy in locally recurrent rectal cancer 29 CHAPTER 2 disease. Response to this treatment could also be used as a selection criterion for further procedures. Good responders not developing metastases may be better candidates for this extensive surgery, whereas those who exhibit local progression or even developmetastases during chemotherapy could be spared unnecessarymorbidity and mortality, and undergo the best palliative treatment in the meantime. One of the major drawbacks of this study is that it was not designed as a prospective study to achieve a complete response. Negative selection bias on the basis of more or less unfavourable conditions may have influenced the results. The accidental finding that most of these patients could undergo an R0 resection and unexpectedly showed a high pCR rate is hypothesis-generating, and requires further validation in future studies. To show an increase in the pCR rate following ICT from5 to 15 percent (two-tailed α = 0.05 and power of 80 percent), the number of patients in each arm would need to be 140. A difference of 15 percent in the R0 resection rate, which is the strongest predictor of oncological outcome, would also require 140 patients per arm. To demonstrate a 10 percent increase in theDFS rate at 3 years, more than 700 patients would be required in each study arm, which is an unrealistic number for a study with such a heterogeneous group as patients with locally recurrent rectal cancer. Alternative study designs, such as a matched case–control study, would require a relatively small cohort of patients to undergo the intervention.