Page 394 - 2014 Printable Abstract Book
P. 394
are underway. A challenge in the stomach cancer studies is that the size, shape and location of the
stomach are unknown and variable, which leads to considerable uncertainty in the estimated stomach
dose. Nevertheless it is clear from these studies that high-dose, fractionated radiotherapy substantially
increases risk of upper GI cancers.
SND. CT SCANS AND CANCER RISKS: NEW RESULTS AND METHODOLOICAL ISSUES
The number of CT scans performed annually in the U.S. has risen dramatically in the last thirty
years from 1 million to 90 million scans per year; with less rapid increases seen in most developed
countries around the world. Despite the great clinical utility of CT, concerns have been raised about the
potential cancer risks from the associated ionizing radiation, particularly for children, because of their
higher radiosensitivity and higher doses. Results from the first two historical pediatric cohorts were
recently published, which both suggested elevated cancer risks with increasing numbers of CT scans/dose
even in the low-dose range of about 50mGy. As well as informing clinical practice these, and other ongoing
pediatric CT cohorts, have the potential to inform low-dose radiation protection standards more broadly.
In this multi-disciplinary session we plan to focus on presentation of new results and discussion of the
methodological issues, which may bias the study results either in an upward or downward direction. This
will include presentation of results about the impact of underlying conditions such as Down Syndrome
and other data on indication for the CT, state-of-the-art CT dosimetry and analyses of the impact of
uncertainties in dose estimates. The session will conclude with a panel discussion of experts including
clinicians, dosimetrists, statisticians and epidemiologists.
(SND01) Cancer risks from low-dose radiation - results from the Australian cohort exposed to
diagnostic CT scans and nuclear medicine procedures. John D. Mathews, MD PhD; Zoe Brady, BSc PhD;
Anna Forsythe, BSc; and Darren Wraith, PhD, The University of Melbourne, Melbourne, Australia
De-identified Medicare records for 11 million young Australians (including records of diagnostic
CT scans and nuclear medicine procedures) were linked to national cancer registrations. Cancer risk was
compared between persons exposed to diagnostic radiation and those who were unexposed. As
previously reported (1), the excess relative risk (ERR) of cancer was found to increase by an average of
16% (95% confidence interval 13-19%) for every diagnostic CT procedure before the age of 20 years.
Excess risk was greatest after exposures at younger ages. For persons developing a brain cancer after a
head scan at age <5 yr, the probability of causation was approximately 60%. Our estimates of ERR per unit
dose for brain cancer and leukemia were consistent with (2) and (3). For solid cancers, our estimate,
0.027(0.017-0.037) per mSv, (see Table 9 in (1)), was significantly greater than estimates based on (3).
Here we report that cancer risks are also increased after exposures to ionizing radiation from diagnostic
nuclear medicine procedures. We also use a simple theoretical model to explain why the ERR/dose
declines with increasing attained age and with increasing age at radiation exposure. Our earlier findings
have been criticized because of the possibility of reverse causation bias (eg from symptoms of a pre-
cancerous condition prompting a CT), and because some excess cancers appeared as soon as 2-5 years
after radiation exposure. However, as radiation-induced leukemia is well recognized within 5 years of
stomach are unknown and variable, which leads to considerable uncertainty in the estimated stomach
dose. Nevertheless it is clear from these studies that high-dose, fractionated radiotherapy substantially
increases risk of upper GI cancers.
SND. CT SCANS AND CANCER RISKS: NEW RESULTS AND METHODOLOICAL ISSUES
The number of CT scans performed annually in the U.S. has risen dramatically in the last thirty
years from 1 million to 90 million scans per year; with less rapid increases seen in most developed
countries around the world. Despite the great clinical utility of CT, concerns have been raised about the
potential cancer risks from the associated ionizing radiation, particularly for children, because of their
higher radiosensitivity and higher doses. Results from the first two historical pediatric cohorts were
recently published, which both suggested elevated cancer risks with increasing numbers of CT scans/dose
even in the low-dose range of about 50mGy. As well as informing clinical practice these, and other ongoing
pediatric CT cohorts, have the potential to inform low-dose radiation protection standards more broadly.
In this multi-disciplinary session we plan to focus on presentation of new results and discussion of the
methodological issues, which may bias the study results either in an upward or downward direction. This
will include presentation of results about the impact of underlying conditions such as Down Syndrome
and other data on indication for the CT, state-of-the-art CT dosimetry and analyses of the impact of
uncertainties in dose estimates. The session will conclude with a panel discussion of experts including
clinicians, dosimetrists, statisticians and epidemiologists.
(SND01) Cancer risks from low-dose radiation - results from the Australian cohort exposed to
diagnostic CT scans and nuclear medicine procedures. John D. Mathews, MD PhD; Zoe Brady, BSc PhD;
Anna Forsythe, BSc; and Darren Wraith, PhD, The University of Melbourne, Melbourne, Australia
De-identified Medicare records for 11 million young Australians (including records of diagnostic
CT scans and nuclear medicine procedures) were linked to national cancer registrations. Cancer risk was
compared between persons exposed to diagnostic radiation and those who were unexposed. As
previously reported (1), the excess relative risk (ERR) of cancer was found to increase by an average of
16% (95% confidence interval 13-19%) for every diagnostic CT procedure before the age of 20 years.
Excess risk was greatest after exposures at younger ages. For persons developing a brain cancer after a
head scan at age <5 yr, the probability of causation was approximately 60%. Our estimates of ERR per unit
dose for brain cancer and leukemia were consistent with (2) and (3). For solid cancers, our estimate,
0.027(0.017-0.037) per mSv, (see Table 9 in (1)), was significantly greater than estimates based on (3).
Here we report that cancer risks are also increased after exposures to ionizing radiation from diagnostic
nuclear medicine procedures. We also use a simple theoretical model to explain why the ERR/dose
declines with increasing attained age and with increasing age at radiation exposure. Our earlier findings
have been criticized because of the possibility of reverse causation bias (eg from symptoms of a pre-
cancerous condition prompting a CT), and because some excess cancers appeared as soon as 2-5 years
after radiation exposure. However, as radiation-induced leukemia is well recognized within 5 years of
