Enrollment in Children's Oncology Group's clinical trials: Population-based linkage with the National Childhood Cancer Registry

Background Improvements in outcomes among children and adolescents diagnosed with cancer are attributable to many factors, including clinical trials such as those administered through the Children's Oncology Group (COG) as well as population-based resources such as the National Childhood Cancer Registry (NCCR). The objective of this study was to link COG trial data with the NCCR to evaluate overall enrollment patterns. Methods Data were received from the NCCR and COG that were linked using an array of variables, then compared to evaluate enrollment patterns in COG studies from 2007 to 2018. Multivariable logistic regression was used to identify characteristics associated with not being enrolled in a COG study. Results Among 134 696 NCCR patients with cancer, 51 062 matched with COG study enrollees. There were several differences in demographic and clinical characteristics between individuals enrolled and not enrolled in COG studies. Enrollment was higher among children aged from birth to 4 years compared with adolescents aged 15-19 years (53.7% vs 20.1%). Differences by race and ethnicity were also observed; for example, individuals who identified as non-Hispanic White were more likely to be enrolled than were individuals who identified as non-Hispanic Asian or Pacific Islander (38.8% vs 32.9%). In a multivariable logistic regression model, several characteristics were strongly associated with not being enrolled in a COG study, including age at diagnosis, year of diagnosis, race and ethnicity, and cancer type. Conclusion Our results suggest that several groups are underrepresented in COG clinical trials. This information can help guide the prioritization of population groups for engagement in future studies.