Abstract
PURPOSEUsing the technique of meta-analysis, we aim to illustrate the potential benefit, or lack of it, in adding chemotherapy to locoregional definitive treatment in a prospective randomized setting.PATIENTS AND METHODSMantel-Haenszel summary analyses were used to test 42 prospective and properly randomized trials for statistically significant differences in the proportion with side effects and in the proportion with response to treatment between the experimental treatment arm (including chemotherapy) and control arm (local definitive treatment only) of the study. Summarized estimates of relative risks of side effects and relative proportions of positive responses were obtained using the summarizing options in PROC FREQ in the SAS computer package. In 25 of 42 studies, sufficient survival information was available to estimate the effect of chemotherapy on the rate of dying per person per unit of time.RESULTSChemotherapy, when added to local definitive treatment, was found to increase toxicity. This incre...

This publication has 0 references indexed in Scilit: