Lung cancer trial patients show disparities in optional biospecimen study participation

BACKGROUND: Biospecimen analyses may provide important insights into patient selection and pharmacodynamic effects. To provide generalizable findings, such studies must enroll adequate and diverse populations. METHODS: We analyzed patient and institutional characteristics according to agreement to participate in optional, embedded biospecimen studies among patients enrolled in the E1505 and E5508 phase 3 lung cancer therapeutic trials. Differences were compared using Wilcoxon rank sum test, Pearson's Chi-squared test, and logistic regression. RESULTS: Overall, 3,017 patients were enrolled in the two trials. Mean age was 63 years, 49% were female, and 83% were non-Hispanic white. Among these individuals, 2,692 (89%) agreed to participate in at least one biospecimen study, and 2,577 (85%) agreed to studies requiring future biospecimen collection. In multivariable logistic regression, compared to non-Hispanic white patients, other patients were less likely to agree to participate: OR 0.59 (95% CI, 0.45-0.79; P < 0.001) for any biospecimen study; OR 0.62 (95% CI, 0.48-0.80; P < 0.001) for studies requiring future biospecimen collection. Women and patients treated outside main academic institutions (e.g., affiliates, community sites) were also less likely to participate. CONCLUSIONS: Among patients enrolled in lung cancer clinical trials, women, racial and ethnic minorities, and patients treated outside major academic centers are less likely to participate in optional biospecimen studies. Because some of these populations may already be under-represented in trial populations, this pattern may exacerbate disparities in translational and clinical research. TRIAL REGISTRATION: NCT00324805, NCT01107626.