New Results from the UK-NCI Pediatric CT scans study

Amy BERRINGTON, NCI, United States
SALOTTI J. 1 , MCHUGH K. 2 , LITTLE M. 5 , HARBRON R. 1 , LEE C. 5 , NEIGE J. 5 , TRANH V. 5 , MOROZ B. 5 , PARKER L. 3 , RAJARAMAN P. 5 , STILLER C. 4 , STEWART D. 5 , CRAFT A. 1 , PEARCE M. 1

1 Institute of Health & Society, University of Newcastle, Newcastle, UK
2 Radiology, Great Ormond Street Hospital, London, UK
3 Faculty of Medicine, Dalhousie University, Halifax, Canada
4 New College, University of Oxford, Oxford, UK
5 Radiation Epidemiology Branch, NCI, Bethesda, USA

Purpose: CT scans save lives and avoid unnecessary surgery.  However, they may also be associated with an increased risk of cancer, particularly in children who are more sensitive to radiation exposure.  We published the first evidence of the potential risk of cancer from pediatric CT scans in our UK cohort.  Here we present results from analyses using new data on underlying conditions, dose uncertainty and the first results for lymphomas.
Methods: Retrospective record linkage cohort study of approximately 180,000 patients who underwent CT scans between before 2002 aged <22 years in the UK.  We collected and reviewed clinical information to assess potential bias from underlying conditions and a set of 1000 CT films to assess the impact of dose uncertainty.  We analysed the risk of leukemia/MDS (n=74), brain tumors (n=135) and lymphomas (n=119) in relation to estimated organ dose from CT scans using Poisson regression.
Results: Underlying conditions that pre-dispose to cancer did not bias the risk estimates for leukemia or brain tumors, but did introduce some bias for lymphomas.  Previous cancers that had not been reported to the cancer registry were a source of bias, particularly for brain tumors.  However, significant dose-response relationships remained even after exclusion of these previous cancers for leukemia/MDS and brain tumors.  The CT films provided new information on trends in doses showing a decrease starting in the mid-1990s and age-adjustments becoming more common.  Assessment of the impact of dose uncertainty and lymphomas will be completed early 2016.
Conclusions: Our new findings, based on additional clinical data, still support a relationship between pediatric CT scans and subsequent leukemia/MDS and brain tumors.  These results have clinical and public health importance as they re-inforce the need to avoid unnecessary CT scans in children, and also provide direct evidence of cancer risks in the low-dose range (<100mGy).