There has been an escalating risk of brain cancer associated with increasing radiation dose to the brain from CT examinations, a cohort study of more than 650,000 children and young adults indicated.
Researchers observed a significant dose-response relationship such that for every 10,000 people who received a single head CT scan, approximately one patient developed brain cancer attributable to radiation exposure in the 5 to 15 years following the examination, reported Michael Hauptmann, PhD, biostatistician at Brandenburg Medical School in Neuruppin, Germany, and colleagues of the EPI-CT study.
“To put this finding into context, the number of pediatric head CT examinations per year during the past decade probably exceeds 1 million in the [European Union] and 5 million in the USA,” they reported in The Lancet Oncology.
“Only a small fraction of the attributable brain cancers might be preventable (i.e., those with unnecessarily high doses during CT examinations or the presumably larger group with clinically unjustified CT examinations),” Hauptmann and colleagues wrote. “Nevertheless, these figures emphasize the need to adhere to the basic radiological protection principles in medicine, namely justification (procedures are appropriate and comply with guidelines) and optimization (doses are as low as reasonably achievable).”
Indeed, CT use has already declined in recent years amid efforts to both justify and optimize these exams, according to Nobuyuki Hamada, PhD, radiation biologist of Japan’s Central Research Institute of Electric Power Industry in Tokyo, and Lydia Zablotska, MD, PhD, radiation epidemiologist of University of California San Francisco.
“Such continued efforts, along with extended epidemiological investigations, would be needed to minimize the risk of brain cancer after pediatric CT examination,” they wrote in an editorial accompanying the study.
In the EPI-CT study, Hauptmann and colleagues identified 978,174 individuals from nine European countries who had at least one CT before age 22 documented between 1977 and 2014.
For purposes of this analysis, the researchers included 658,752 people who were alive and cancer-free (including free from known brain tumors of benign or unknown behavior) 5 years after their first CT examination and followed them for an additional median 5.6 years.
Of the patients analyzed, 82% were younger than 30 years old at the end of follow-up, 56% were men, and 73% had received at least one head or neck CT.
During follow-up, 165 cases of brain cancer were reported, 73% of which were gliomas.
With a 5-year lag, mean cumulative brain dose had been 47.4 mGy among all individuals and 76.0 mGy among people with brain cancer.
Hauptmann and colleagues estimated that every 100 mGy of brain dose incurred a 127% increase in brain cancer risk (95% CI 51-269), 111% increase in glioma risk (95% CI 36-259), and 213% increase in brain cancer risk excluding gliomas (95% CI 25-136).
The average brain dose per head or neck CT in 2012-2014, 38 mGy, was associated with excess brain cancer risk compared with a brain dose of 0 mGy (RR 1.5, 95% CI 1.2-2.0).
The authors acknowledged several limitations to the observational study, including the possibility of reverse causation.
Although a 5-year exclusion period after the first recorded CT examination had been chosen to minimize the possibility of someone having pre-existing brain cancer, the authors acknowledged that participants had real, or suspected, medical problems for which a CT scan was indicated. Thus, they suggested, “these problems could, in theory, be early symptoms of subsequently diagnosed brain cancer.”
They also said there could be some uncertainly regarding dose estimates because they were obtained retrospectively from a number of data sources. “Detailed technical information for dose estimation was particularly scarce for years before 1990,” they noted.
Hamada and Zablotska added that longer follow-up would be necessary to understand age trends, as the average age at the end of the present study was only 22 years and it isn’t until an average 30 years of age that population rates of glioma start increasing.
The study was supported by the European Union’s Seventh Framework Programme and the Belgian Cancer Registry; La Ligue contre le Cancer, L’Institut National du Cancer, France; Ministry of Health, Labour and Welfare of Japan; German Federal Ministry of Education and Research; Worldwide Cancer Research; Dutch Cancer Society; Research Council of Norway; Consejo de Seguridad Nuclear, Generalitat de Catalunya, Spain; US National Cancer Institute; UK National Institute for Health Research; Public Health England.
Hauptmann, Hamada, and Zablotska had no disclosures.
A study co-author reported multiple relationships with industry.