Variation in enrollment of colorectal cancer patients in clinical trials

e15094 Background: Correct estimation of patient enrollment is an important success factor for planning clinical studies including studies for metastatic colorectal cancer (CRC). Methods: We reviewed all CRC studies published in the Journal of Clinical Oncology and Annals of Oncology between 01/2007 and 10/01/2008. 43 studies were found and the following data were collected from 39 studies: indication, phase, number of sites, number of patients enrolled, mean patient age, recruitment time, sponsor (Industry, NIH, Organization, University), region, line of treatment and type of drug (6 categories). 4 studies were omitted from analysis as recruitment data were largely missing. Our analysis is based on a literature review as information from unpublished trials is unavailable. This implies some limitations regarding the data interpretation. Results: An average enrollment of 0.92 Patient/Site/Month (range 0.10–7.38) was observed for these trials. The highest recruitment efficacy with a median of 4.11 Pt/S/M (range 1.81–7.38) was found in 5 single institution phase II trials. For multi-center phase II and III studies the median enrollment was 1.82 and 0.32 Pt/S/M respectively, with significant higher recruitment in phase II studies. The highest enrollment rate was observed for studies located in Europe or in USA (0.77 and 2.21 Pt/S/M respectively, p=0.03). No correlation was seen with the mechanism of action (targeted drug vs. chemotherapy), sponsor (NIH vs. Industry vs. IIT), line of treatment (first line vs. 2nd and subsequent line). For phase I recruitment analysis we retrieved 2 studies that investigated novel agents in solid tumor patients including advanced or metastatic colorectal cancer patients with a median recruitment of 0.46 Pt/S/M. For phase Ib and I/II recruitment analysis 5 studies were found with a median recruitment of 0.78 Pt/S/M. Conclusions: Single institution phase II clinical trials on novel agents with high potential to change future treatment standard demonstrate almost a tenfold higher than average recruitment rate for multi-center trials (0.45 Pt/S/M). Despite some limitations in the interpretation of results our analysis provides important information to support estimation of patient recruitment in future clinical trials for colorectal cancer. No significant financial relationships to disclose.