EPID-22. MATERNAL USE OF TOBACCO, ALCOHOL, AND ILLICIT DRUGS DURING PREGNANCY AND ASSOCIATION WITH CHILDHOOD CANCER SUBTYPES

Abstract

Abstract OBJECTIVE To investigate the association between maternal prenatal substance use/abuse and diagnosis of childhood cancer subtypes, including childhood brain tumors. To minimize recall bias, we employed a case-only study design in which all respondents had a child diagnosed with cancer. STUDY DESIGN: We deployed a cross-sectional, case-only survey of parents of children diagnosed with cancer regarding prenatal smoking, drinking, and illicit drug use behaviors (scored on 1-7 Likert scale), demographic, gestational, and perinatal factors. Multivariable logistic regression assessed associations of gestational smoking, alcohol, and illicit drug use with childhood cancer subtypes, adjusting for child sex, race/ethnicity, birthweight, and household income. RESULTS 3145 families completed the survey, including 232 with astrocytoma (grades I-IV, including DIPG) and 210 with an embryonal CNS tumor (medulloblastoma, AT/RT, PNET). A minority reported ever smoking tobacco products during pregnancy (Likert score ≥ 2; 14%), ever consuming illicit drugs during pregnancy (Likert score ≥ 2; 4%), or drinking more than a moderate amount of alcohol during pregnancy (Likert score ≥ 5; 2%). Prenatal smoking was associated with lower odds of rhabdomyosarcoma (OR=0.48, P=0.023) and moderately elevated odds of non-Hodgkin lymphoma (OR = 1.52, P=0.098) and AML (OR = 1.52, P = 0.083). Moderate to heavy alcohol consumption was unassociated with cancer subtypes. Prenatal illicit drug use was associated with increased odds of retinoblastoma (OR = 5.03, 95% CI = 1.73-14.7, P = 0.003) and CNS embryonal tumors (OR = 2.56, 95% CI = 1.32-4.99, P = 0.006), with similar effects across subgroups of medulloblastoma, AT/RT, and supratentorial PNETs. CONCLUSIONS Associations between self-reported use of alcohol, tobacco, and illicit drugs during pregnancy and risk of childhood cancer is often biased by differential misreporting among parents of affected children. Consistent with prior reports, we observe elevated odds of hematologic malignancies in association with gestational tobacco smoke exposure and novel association between maternal illicit drug use during pregnancy and elevated odds of retinoblastoma and CNS embryonal tumors.