Association between adolescent and young adult (AYA) age, acuity of illness at diagnosis, and clinical trial enrollment among patients with hematologic malignancies at a pediatric center

Abstract Background: Adolescent and young adult (AYA) patients with hematologic malignancies have inferior disease-free and overall survival compared to their younger peers. Race/ethnicity, socioeconomic status, and insurance status have all been linked to lower overall survival in AYA patients. We have previously demonstrated the association between high acuity of illness at cancer diagnosis and early mortality in pediatric acute myeloid leukemia (AML) and acute lymphoblastic leukemia (ALL). Clinical trials may offer the best medical and social support for AYA patients; thus, understanding barriers to clinical trial enrollment, which is dependent on patient, provider, and health system factors, may address the survival disparities seen in AYAs. We evaluated potential drivers of clinical trial enrollment in AYA patients with hematologic malignancies treated at a pediatric center. Methods: We performed a retrospective study of individuals, ages 0-21 years, diagnosed and treated at Children's Healthcare of Atlanta (CHOA) for hematologic malignancies between 2010-2018. The primary study outcome was enrollment in a frontline therapeutic clinical trial (yes vs. no). Local institutional review board records were used to determine trial availability at CHOA, and individual charts were reviewed to determine if a patient met the eligibility criteria (disease and organ sufficiency) for an open trial at time of diagnosis. We evaluated AYA age (≥ 15 years old) and high acuity of illness at diagnosis as the primary exposure variables. Acuity of illness was defined by any use of intensive care unit resources from 30 days before through 72 hours after cancer diagnosis based on pre-defined criteria via manual abstraction from the electronic medical records. Logistic regression models estimated the associations of age and acuity of illness at diagnosis with clinical trial enrollment (odds ratios (ORs) and associated 95% confidence intervals (CI)). Covariates examined include sex, race/ethnicity, primary language (English vs. non-English), health insurance type (public vs. private), and blood cancer subtype (leukemia vs. lymphoma). An adjusted model was determined by a stepwise selection approach (R v.4.4.1). Results: Between 2010-2018, 1030 patients were diagnosed and treated for a hematologic malignancy at CHOA. An open frontline clinical trial was available at CHOA for 621/1030 patients (60.3%). Among these, 99 (15.9%) patients were AYA age, 172 (27.7%) patients identified as non-Hispanic Black, and 160 (25.8%) patients were classified as having high acuity of illness at diagnosis. However, 19 (3.1%) patients did not meet trial eligibility criteria. Among the 602 patients offered an open clinical trial and determined eligible, 437 patients (72.6%) enrolled on the trial. Unadjusted models showed that AYA age (OR: 0.34 [95% CI: 0.22-0.53]), public insurance (OR: 0.70 [95% CI: 0.50-0.99]), non-Hispanic Black race/ethnicity (OR: 0.46 [95% CI: 0.46-0.69]), and high acuity of illness (OR: 0.61 [95% CI: 0.42-0.90]) were negatively associated with clinical trial enrollment, while leukemia diagnosis (OR: 6.29 [95% CI: 3.69-11.0]) was positively associated with clinical trial enrollment. In the adjusted model, AYA age (aOR: 0.45 [95% CI: 0.28-0.73]), non-Hispanic Black race/ethnicity (aOR: 0.51 [95% CI: 0.32-0.80]), and leukemia diagnosis (aOR: 5.36 [95% CI: 3.05-9.64]) remained significant. Conclusions: Clinical trial enrollment must be evaluated at an institutional and individual patient level. In a large academic pediatric cancer center with robust clinical trial commitment, we found that AYA age and non-Hispanic Black race/ethnicity have independent associations on non-enrollment in a clinical trial. High acuity of illness at diagnosis was nonsignificant after adjustment. Ongoing analyses will characterize the reasons for clinical trial refusal/ineligibility, and a mediation analysis will determine the extent to which high acuity of illness at diagnosis and area-based measures of social vulnerability account for the observed trial enrollment disparities by AYA age and race/ethnicity. Future studies will expand this analysis to include AYA patients treated for hematologic malignancies at an affiliated adult cancer center and expand the cohort through 2023.