Confounding of the association between radiation exposure from CT scans and risk of leukemia and brain tumors by cancer susceptibility syndromes
A partir d'une revue systématique de la littérature, cette étude évalue l'influence de syndromes de prédisposition au cancer (syndrome de Down, syndrome de Noonan, etc) sur l'association entre une exposition à des rayonnements induits par des scanners durant l'enfance et le risque de leucémie ou de tumeurs cérébrales
Background: Recent studies linking radiation exposure from pediatric computed tomography (CT) to increased risks of leukemia and brain tumors lacked data to control for cancer susceptibility syndromes (CSS). These syndromes might confound the association since they are associated with an increased cancer risk and may increase the likelihood of CT scans performed in children. Methods: We identify CSS predisposing to leukemia and brain tumors through a systematic literature search and summarize prevalence and risk estimates. Since there is virtually no empirical evidence in published literature on patterns of CT use for most types of CSS, we estimate confounding bias of relative risks (RR) for categories of radiation exposure based on expert opinion about the current and previous patterns of CT scans among CSS patients. Results: We estimate that radiation-related RRs for leukemia are not meaningfully confounded by Down syndrome, Noonan syndrome or other CSS. In contrast, RRs for brain tumors may be overestimated due to confounding by tuberous sclerosis complex (TSC) while von Hippel-Lindau syndrome, neurofibromatosis type 1 or other CSS do not meaningfully confound. Empirical data on the use of CT scans among CSS patients is urgently needed. Conclusions: Our assessment indicates that associations with leukemia reported in previous studies are unlikely to be substantially confounded by unmeasured CSS, whereas brain tumor risks might have been overestimated due to confounding by tuberous sclerosis complex. Impact: Future studies should identify TSC patients in order to avoid overestimation of brain tumor risks due to radiation exposure from CT scans.