• Amy Berrington de Gonzalez, Jane A Salotti, Kieran McHugh, Mark P Little, Richard W Harbron, Choonsik Lee, Estelle Ntowe, Melissa Z Braganza, Louise Parker, Preetha Rajaraman, Charles Stiller, Douglas R Stewart, Alan W Craft, and Mark S Pearce.
• Radiation Epidemiology Unit, Division of Cancer Epidemiology and Genetics, NCI, Bethesda, MD, USA.
• Br. J. Cancer. 2016 Feb 16; 114 (4): 388-94.

BackgroundWe previously reported evidence of a dose-response relationship between ionising-radiation exposure from paediatric computed tomography (CT) scans and the risk of leukaemia and brain tumours in a large UK cohort. Underlying unreported conditions could have introduced bias into these findings.MethodsWe collected and reviewed additional clinical information from radiology information systems (RIS) databases, underlying cause of death and pathology reports. We conducted sensitivity analyses excluding participants with cancer-predisposing conditions or previous unreported cancers and compared the dose-response analyses with our original results.ResultsWe obtained information from the RIS and death certificates for about 40% of the cohort (n∼180 000) and found cancer-predisposing conditions in 4 out of 74 leukaemia/myelodysplastic syndrome (MDS) cases and 13 out of 135 brain tumour cases. As these conditions were unrelated to CT exposure, exclusion of these participants did not alter the dose-response relationships. We found evidence of previous unreported cancers in 2 leukaemia/MDS cases, 7 brain tumour cases and 232 in non-cases. These previous cancers were related to increased number of CTs. Exclusion of these cancers reduced the excess relative risk per mGy by 15% from 0.036 to 0.033 for leukaemia/MDS (P-trend=0.02) and by 30% from 0.023 to 0.016 (P-trend<0.0001) for brain tumours. When we included pathology reports we had additional clinical information for 90% of the cases. Additional exclusions from these reports further reduced the risk estimates, but this sensitivity analysis may have underestimated risks as reports were only available for cases.ConclusionsAlthough there was evidence of some bias in our original risk estimates, re-analysis of the cohort with additional clinical data still showed an increased cancer risk after low-dose radiation exposure from CT scans in young patients.

26 keywords...

hide…