S Dubé1, F Heyen, M Jenicek. 1. Department of Surgery, University of Montreal, Quebec, Canada.
Abstract
UNLABELLED: Despite the National Institutes of Health consensus regarding use of adjuvant chemotherapy in colorectal carcinoma, many general surgeons question the efficacy of this approach when considering costs involved for both the individual patient and society at large. PURPOSE: This study was designed to determine the real impact of adjuvant chemotherapy on five-year survival rates of patients. METHOD: A qualitative and quantitative meta-analysis of results from 39 randomized clinical trials published from 1959 to 1993 is described. RESULTS: Design quality of clinical trials had a mean score of 48.6 percent (+/-6.2 standard deviation). A small benefit of therapy in terms of overall survival was noted, with a mortality odds ratio (OR) of 0.91 (confidence interval (CI) 95 percent, 0.83-0.99). For the group of colon carcinomas, the OR was 0.81 (CI 95 percent, 0.69-0.94) with an OR of 0.64 (CI 95 percent, 0.48-0.85) for the group of rectal carcinomas. The effect size was 0.09 for the colon group and 0.20 for the rectal group. For those patients who receive chemotherapy, this effect size implies that we can expect an increase of 5 percent in the survival rate in the group with colon carcinoma and a 9 percent increase in the survival rate in the group with rectal carcinoma. CONCLUSION: Given the high incidence of colorectal carcinoma, the small benefit observed for those patients receiving chemotherapy is far from negligible. However, indications for adjuvant chemotherapy warrant further discussion.
UNLABELLED: Despite the National Institutes of Health consensus regarding use of adjuvant chemotherapy in colorectal carcinoma, many general surgeons question the efficacy of this approach when considering costs involved for both the individual patient and society at large. PURPOSE: This study was designed to determine the real impact of adjuvant chemotherapy on five-year survival rates of patients. METHOD: A qualitative and quantitative meta-analysis of results from 39 randomized clinical trials published from 1959 to 1993 is described. RESULTS: Design quality of clinical trials had a mean score of 48.6 percent (+/-6.2 standard deviation). A small benefit of therapy in terms of overall survival was noted, with a mortality odds ratio (OR) of 0.91 (confidence interval (CI) 95 percent, 0.83-0.99). For the group of colon carcinomas, the OR was 0.81 (CI 95 percent, 0.69-0.94) with an OR of 0.64 (CI 95 percent, 0.48-0.85) for the group of rectal carcinomas. The effect size was 0.09 for the colon group and 0.20 for the rectal group. For those patients who receive chemotherapy, this effect size implies that we can expect an increase of 5 percent in the survival rate in the group with colon carcinoma and a 9 percent increase in the survival rate in the group with rectal carcinoma. CONCLUSION: Given the high incidence of colorectal carcinoma, the small benefit observed for those patients receiving chemotherapy is far from negligible. However, indications for adjuvant chemotherapy warrant further discussion.