Page 415 - 2014 Printable Abstract Book
P. 415
of gestation were explored using linear regression and Cox proportional hazards models. The analyses
are based on the 1,167 mother-child pairs with complete data, whose mean fetal I-131 dose is .05 Gy
(SD=0.13). Preliminary analyses found no association of fetal 131 exposure with birth weight, chest
circumference or preterm delivery but did show dose-dependent relationships with head circumference
(HC) and gestational length. In Cox proportional hazards models adjusted for parity, gender and trimester
at exposure, higher dose was strongly associated with late-term/post-term delivery (P<.001), although it
is unclear how radiation may have affected triggers for term delivery. Regression analyses adjusted for
the same variables confirmed, somewhat unexpectedly, the relation-ship of thyroid 131 to head
circumference, with a decrement on the order of 0.2 cm per unit increase in log fetal dose. The reduced
head size among in utero survivors of the atomic bombings was associated with exposure to gamma
radiation, which affects the brain, not only the thyroid. The biologic mechanism for reduced HC in a
cohort exposed to radioactive iodine is less certain but the finding was robust.
(PS7-65) Computed pediatric tomography exposure and radio-induced cancers: the first results from a
1
1
1
national cohort study in France. Marie Odile Bernier ; Neige Journy ; Jean Luc Rehel ; Hubert Ducou Le
1
1
2
3
Pointe ; Jean François Chateil ; Sylvaine Caer-Lorho ; and Dominique Laurier , IRSN, Fontenay aux Roses,
3
1
2
France ; Trousseau Hospital, Paris, France ; and Bordeaux Hospital, Bordeaux, France
Context: The increasing use of computed tomography (CT) scans for the pediatric population
raises the question of the possible impact of such ionising radiation (IR) exposure on the occurrence of
radio-induced cancers. Recent epidemiological studies in the UK and Australia have suggested an
increased risk of cancer among children receiving CT scans. In France, a nationwide study has been
launched to assess cancer risks associated with the use of CT scans in pediatrics. This study is part of the
Epi-CT collaborative European project. Material and methods: In 21 French University hospitals, children
less than 10 years old, subjected to at least one CT scan between 2000 and 2011, were included.
Cumulative organ doses were estimated according to the protocols retrieved from the radiology
departments, using specifically designed simulation software. Clinical information recorded during
hospitalization was used to determine whether the children had medical conditions likely to increase their
risk of cancer. Cancer incidence and mortality data were retrieved through national registries. Results: At
the end of 2011, 67 000 children were included, 30% of whom were exposed to a first CT scan before the
age of 1 year. Examinations of the head represent 69% of the CT scans. Considering various periods of
exclusion, no significant increase of excess relative risk (ERR) of radio-induced cancer was observed.
Nevertheless, the observed ERR of brain tumor, lymphoma and leukemia decreased after an adjustment
on clinical conditions predisposing to cancer. Conclusion: These first results quantify the indication bias,
which could be suspected in a study focusing on people non representative of the general population. It
emphasizes how some genetic defects and other conditions associated with a high risk of cancer could
affect the assessment of the risk attributed to CT scans.
are based on the 1,167 mother-child pairs with complete data, whose mean fetal I-131 dose is .05 Gy
(SD=0.13). Preliminary analyses found no association of fetal 131 exposure with birth weight, chest
circumference or preterm delivery but did show dose-dependent relationships with head circumference
(HC) and gestational length. In Cox proportional hazards models adjusted for parity, gender and trimester
at exposure, higher dose was strongly associated with late-term/post-term delivery (P<.001), although it
is unclear how radiation may have affected triggers for term delivery. Regression analyses adjusted for
the same variables confirmed, somewhat unexpectedly, the relation-ship of thyroid 131 to head
circumference, with a decrement on the order of 0.2 cm per unit increase in log fetal dose. The reduced
head size among in utero survivors of the atomic bombings was associated with exposure to gamma
radiation, which affects the brain, not only the thyroid. The biologic mechanism for reduced HC in a
cohort exposed to radioactive iodine is less certain but the finding was robust.
(PS7-65) Computed pediatric tomography exposure and radio-induced cancers: the first results from a
1
1
1
national cohort study in France. Marie Odile Bernier ; Neige Journy ; Jean Luc Rehel ; Hubert Ducou Le
1
1
2
3
Pointe ; Jean François Chateil ; Sylvaine Caer-Lorho ; and Dominique Laurier , IRSN, Fontenay aux Roses,
3
1
2
France ; Trousseau Hospital, Paris, France ; and Bordeaux Hospital, Bordeaux, France
Context: The increasing use of computed tomography (CT) scans for the pediatric population
raises the question of the possible impact of such ionising radiation (IR) exposure on the occurrence of
radio-induced cancers. Recent epidemiological studies in the UK and Australia have suggested an
increased risk of cancer among children receiving CT scans. In France, a nationwide study has been
launched to assess cancer risks associated with the use of CT scans in pediatrics. This study is part of the
Epi-CT collaborative European project. Material and methods: In 21 French University hospitals, children
less than 10 years old, subjected to at least one CT scan between 2000 and 2011, were included.
Cumulative organ doses were estimated according to the protocols retrieved from the radiology
departments, using specifically designed simulation software. Clinical information recorded during
hospitalization was used to determine whether the children had medical conditions likely to increase their
risk of cancer. Cancer incidence and mortality data were retrieved through national registries. Results: At
the end of 2011, 67 000 children were included, 30% of whom were exposed to a first CT scan before the
age of 1 year. Examinations of the head represent 69% of the CT scans. Considering various periods of
exclusion, no significant increase of excess relative risk (ERR) of radio-induced cancer was observed.
Nevertheless, the observed ERR of brain tumor, lymphoma and leukemia decreased after an adjustment
on clinical conditions predisposing to cancer. Conclusion: These first results quantify the indication bias,
which could be suspected in a study focusing on people non representative of the general population. It
emphasizes how some genetic defects and other conditions associated with a high risk of cancer could
affect the assessment of the risk attributed to CT scans.
