Differences between clinical trial participants and patients in a population-based registry: the German Rectal Cancer Study vs. the Rostock Cancer Registry.

PURPOSE: There are few data on whether the samples of randomized phase III studies are representative for cancer patients in general populations. METHODS: We compared patient and disease characteristics of patients with stage II or III rectal cancer from the German Rectal Cancer Study (657 patients, 1995-2002) or the Rostock Cancer Registry (371 patients, 1997-2003). Differences between the Study and the Registry were analyzed for subgroups who received neoadjuvant chemoradiotherapy before resection or primary resection with or without postoperative chemoradiotherapy. RESULTS: Study and Registry patients differed in age (median, 61.7 vs. 65.0 years, P < 0.001) and proportion of women (31.3 percent vs. 38.4 percent, P < 0.004). Significant age and gender differences were seen in primary resection but not in neoadjuvant subgroups. In neoadjuvant and in primary resection subgroups, Study participants were more likely than Registry patients to have tumor location in the lower third of the rectum, a higher rate of R0 resection, a greater number of lymph nodes assessed, and fewer T4 tumors. In the primary resection subgroups, Study participants were more likely to have received postoperative chemoradiotherapy. Multivariate analyses showed no effect of population type (Study vs. Registry) on disease-free or overall survival in neoadjuvant subgroups, but increased risk for Registry patients in primary resection subgroups. CONCLUSIONS: Participants in clinical trials such as the German Rectal Cancer Study are not representative of all cancer patients of a general population. To enable wider extrapolation of results, future studies should include elderly and high-risk patients.