Journal of the National Cancer Institute Advance Access published online on March 11, 2008
JNCI Journal of the National Cancer Institute, doi:10.1093/jnci/djn045
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© The Author 2008. Published by Oxford University Press.
ARTICLES |
Solid Cancer Incidence in Atomic Bomb Survivors Exposed In Utero or as Young Children
Affiliations of authors: Hirosoft International Corporation, Eureka, CA (DLP); Departments of Statistics (HC, SF) and Epidemiology (NN, FK), Radiation Effects Research Foundation (KK, RES), Hiroshima, Japan; Department of Epidemiology, Radiation Effects Research Foundation, Nagasaki, Japan (AS, MS); Radiation Epidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Bethesda, MD (KM)
Correspondence to: Dale L. Preston, PhD, Hirosoft International Corporation, 1335 H St, Eureka, CA 95501-2331 (e-mail: preston{at}hirosoft.net).
 |
ABSTRACT |
|---|
 Top Abstract Context and CaveatsSubjects and MethodsResultsDiscussionFundingReferencesNotes |
|---|
Background: In utero exposure to radiation is known to increase risks of childhood cancers, and childhood exposure is associated with increased risks of adult-onset cancers. However, little is known about whether in utero exposure to radiation increases risks of adult-onset cancers.
Methods: Solid cancer incidence rates were examined among survivors of the atomic bombings of Hiroshima and Nagasaki who were in utero (n = 2452) or younger than 6 years (n = 15388) at the time of the bombings. Poisson regression was used to estimate and compare the levels and temporal patterns of the radiation-associated excess risks of first primary solid cancers among these survivors at ages 12–55. All statistical tests were two-sided.
Results: There were 94 eligible cancers in the in utero group and 649 in the early childhood group. The excess relative risk (ERR) increased with dose for both in utero (age 50, ERR = 1.0 per Sv, 95% confidence interval [CI] = 0.2 to 2.3 per Sv) and early childhood (age 50, ERR = 1.7 per Sv, 95% CI = 1.1 to 2.5 Sv) exposures. The ERR declined (P = .046) with increasing attained age in the combined cohort. Excess absolute rates (EARs) increased markedly with attained age among those exposed in early childhood but exhibited little change in the in utero group. At age 50, the estimated EARs per 10000 person-years per Sv were 6.8 (95% CI = <0 to 49) for those exposed in utero and 56 (95% CI = 36 to 79) for those exposed as young children.
Conclusions: Both the in utero and early childhood groups exhibited statistically significant dose-related increases in incidence rates of solid cancers. The apparent difference in EARs between the two groups suggests that lifetime risks following in utero exposure may be considerably lower than for early childhood exposure, but further follow-up is needed.
Prior knowledge Exposure to ionizing radiation in utero and in childhood is associated with increased risks of cancers in childhood and in adulthood, respectively. Study design Excess risks of solid cancers at ages 12–55 among survivors of the atomic bombings of Hiroshima and Nagasaki who were in utero and young children at the time of the bombings were determined. Contributions Excess relative risks of solid cancers increased with radiation dose for both groups of survivors; they declined with increasing attained age in the combined cohort. Excess absolute rates increased with attained age among those who were exposed in childhood but remained steady among those exposed in utero. Implications The difference in excess absolute rates between the two groups of survivors suggests that lifetime risks after exposure may be lower for those exposed in utero than those exposed in childhood, but additional follow-up is necessary. Limitations Due to the limited population size available for analysis, data regarding temporal patterns and risks of site-specific cancers were not available.
|
The Radiation Effects Research Foundation (RERF) tracks the mortality and cancer incidence among survivors of the 1945 atomic bombings of Hiroshima and Nagasaki. Observations of those exposed in utero have been analyzed and periodically reported since 1970. A dose-related increase in cancer mortality before age 15 (ie, childhood cancer mortality) could not be demonstrated in this group due to the small numbers of cancers (1–4). However, as the cohort has aged and cancers have accumulated, so has evidence of a dose-related increase in cancer mortality (5–7).
People who were in utero or young children at the time of the bombings are now attaining ages at which background cancer rates begin to rise sharply. A previous analysis considered solid cancer and leukemia mortality over the age range 15–46 years in these groups (7). Because the in utero cohort is small and follow-up time was limited, the data included only eight deaths from solid cancers and two from leukemia among those exposed to at least 0.01 Sv. However, it was possible to show a statistically significant excess relative risk (ERR) of solid cancers (ERR = 2.4 per Sv, 95% confidence interval [CI] = 0.3 to 6.7 per Sv) (7). The magnitude of this excess did not differ from that of those exposed during the first 6 years of life (ERR = 1.4 per Sv, 95% CI = 0.4 to 3.1). The number of leukemia deaths was too small for a dose–response analysis.
In this report, we consider solid cancer incidence in the age range of 12–55 years for the period 1958–1999 among a cohort of atomic bomb survivors who were either in utero or in the first 6 years of life at the time of the bombings. We pay particular attention to differences in the temporal pattern of the radiation-associated excess risk of solid cancers following exposure in utero or during early childhood. Analyses of the risk of leukemia and other malignant neoplasms of the lymphohematopoietic system will be presented in a separate paper because the case ascertainment methods and follow-up period differ from those used for solid cancers.
|
Subjects and Methods |
|---|
|
TopAbstractContext and CaveatsSubjects and MethodsResultsDiscussionFundingReferencesNotes |
|---|
Study Cohorts
This study was reviewed and approved by the human subjects study review committees of the RERF and the Hiroshima and Nagasaki Tumor registries.
The study population consisted of a cohort of 3268 people who were in utero at the time of the bombings (August 6, 1945, in Hiroshima; August 9, 1945, in Nagasaki) and an early childhood cohort including the 15899 members of the RERF Life Span Study (LSS) who were younger than 6 years at the time of bombings. Individuals in both groups were alive and had no documented history of cancer before January 1, 1958, when tumor registries were established in each city. Follow-up for analyses of mortality in the LSS cohort began on October 1, 1950. Between this date and the end of 1957, there was only one cancer death among those exposed in utero and none among those exposed in early childhood. Individual radiation doses were determined using the Dosimetry System 2002 (DS02) (8–10). The gamma dose was combined with the neutron dose, which was weighted (ie, multiplied by 10) to reflect the greater biologic effect of neutron radiation. Because DS02 does not provide fetal dose estimates, the mother's uterine dose was used as a surrogate for fetal dose in persons who were exposed in utero (11–13). The DS02-weighted colon dose was used for persons who were exposed as children. DS02 estimates could not be computed for 738 persons (227 exposed in utero, 511 exposed in early childhood) who were exposed within 3 km of the hypocenter and for whom the effects of shielding by buildings or terrain could not be adequately characterized. These people were excluded from the analyses.
The in utero cohort also included 589 people born to women whose exposure status is unknown. Because it is believed that these women were not exposed to radiation from the bombs, it has been customary to treat their children as having received zero dose. However, rather than make that assumption, we excluded the children of these women from the current analyses. Interestingly, age- and sex-adjusted solid cancer incidence rates in this group appear to be lower (RR = 0.35, 95% CI = 0.15 to 0.67) than those for the cohort members who received little or no radiation dose, suggesting that they differed from others with regard to some factors affecting their baseline cancer rates.
Follow-up began on January 1, 1958, when the tumor registries started, except, as described in (7), for 468 (14%) of the in utero cohort members who were identified after 1958, largely through a supplement to the 1960 Japanese national census. Follow-up for these 468 cohort members began on October 1, 1960. Follow-up ended on the date of the first primary cancer diagnosis, the date of death from any cause, the date of loss to follow-up, the date of reaching age 55, or December 31, 1999, whichever occurred first. A total of 35 cohort members (12 in utero) were lost to follow-up due to migration from Japan. The age 55 cutoff was used to ensure compatibility because all in utero cohort members were younger than this at the end of follow-up on December 31, 1999.
After exclusions, 2452 survivors who were in utero and 15388 who were young children at the time of the bombings were included in the study (Table 1). People whose mothers normally resided in the city and met the other cohort eligibility criteria but were "not in city" at the time of the bombing were included in the study population because they contribute to the estimation of background rates and, hence, to the precision of the estimated excess rate per Sv of radiation exposure. The not-in-city group for the LSS was identified on the basis of special censuses conducted in Hiroshima and Nagasaki cities in 1950, 1951, and 1953. As noted elsewhere (14), in view of the way in which the group was selected, it seems that members of the not-in-city group were more likely to have been residents of areas near the hypocenters than more distal residents.
|
Incident cancers were ascertained by linkage to the Hiroshima and Nagasaki tumor registries, which provide relatively complete population-based case ascertainment for residents of Hiroshima and Nagasaki and the surrounding areas. Complete mortality follow-up data for both the in utero and early childhood cohorts are available from the mandatory national family registry system (koseki). Tumor registry case ascertainment and data quality were discussed in (15), and mortality follow-up procedures have been described in many reports [including (16,17)]. In view of the incomplete ascertainment among nonresidents, it would be ideal to limit analyses of cancer incidence to periods when cohort members were residents of the Hiroshima and Nagasaki tumor registry catchment areas. However, such detailed individual residence history information is not available. Therefore, as in analyses of cancer incidence in the LSS (18,19), migration-adjusted person-years at risk were estimated using city-specific, calendar year–specific, age-specific, and sex-specific residence probability estimates that were derived from a subset of individuals who were contacted biennially for RERF's clinical follow-up study (18,20). Additional information on residence probabilities is included online (Supplementary Figure 1, available online). It should also be noted that participation in the RERF clinical examination program has little impact on cancer ascertainment because this examination is not the primary source of medical care for cohort members and is not, with the exception of some short-term programs concerned with thyroid disease (21) and skin neoplasms (22), a cancer screening exam. Few cancers were initially diagnosed in the RERF clinical study.
Statistical Analysis
Analyses included first primary solid cancers [International Classification of Disease for Oncology (ICD-O) version 3 (23) topography codes C00–C76 and C80 and behavior code 3] and first primary tumors of the brain, central nervous system, and meninges (ICD-O topography codes C70–C72), including benign tumors or tumors of uncertain behavior (ICD-O behavior codes 0 and 1, respectively). In situ tumors (behavior code 2) were not included. Like most major registries (24), the Hiroshima and Nagasaki registries routinely collect and report information on the incidence of tumors of the brain and central nervous system with benign or uncertain behavior. As in other analyses of cancer incidence among the atomic bomb survivors (25), all tumors of the brain and central nervous system were included in these analyses. Approximately 90% of the cancers were histologically confirmed. Cancers among cohort members who lived outside Hiroshima or Nagasaki prefecture at the time of diagnosis were not included in the analyses.
Analyses were based on simple parametric ERR and excess absolute rate (EAR) models fit to a detailed stratification of cancers and person-years using Poisson regression methods (26,27). The person-year table was stratified on city, sex, in utero vs childhood exposure, age at exposure (trimester for in utero and 0–2 years and 3–5 years for children), attained age (2-year categories from age 12 to 53, with an additional category for age 54), maternal distance from the hypocenter or exposure status (<1500, 1500–2999, 3000–10000 m, and not in city), and 13 adjusted DS02 dose categories (with cut points at weighted doses of 0.005, 0.02, 0.05, 0.1, 0.2, 0.5, 0.75, 1, 2, 2.5, and 3 Sv). The dose-error–adjustment method (28,29) was used, assuming 35% random error in individual dose estimates to allow for the impact of uncertainty in individual dose estimates on risk estimates. The resulting table had nonzero person-years in 7361 of the 84480 potential cells. In addition to person-years and the number of solid cancers, each cell also contained information on the numbers of several specific types of cancer and person-year–weighted means of attained age, age at exposure, year, distance, and dose.
ERR analyses were based on models using the form
 |
for which 
0 is a parametric model for the baseline rates that depends, as described below, on attained age (a) and sex (s) and 
(d)
(z) describes the shape of the dose response ((d)) and radiation effect modification ((z)). The dose response was generally found to be linear in dose, with a slope that may differ for those exposed in utero and those in early childhood. Effect modification was described using a log-linear function of factors of interest, such as sex, attained age, or age at exposure. The primary effect modifiers considered in these analyses were log attained age and sex.
The EAR or excess rate models used the form
 |
in which the second term describes the excess rate. The dose response and effect modification terms were the same as those considered for the ERR models.
The logarithms of the sex-specific baseline rates were described as quadratic functions of log attained age, which implies that baseline rates are proportional to a power of age that varies with logarithm of age. This model can be written as
 | [1] |
City, being in utero, and location at the time of the bombs (proximal defined as being within 3 km of the hypocenter, distal defined as being 3–10 km from the hypocenter, and not in city defined as being more than 10 km from the hypocenter) were considered as potential modifiers of the baseline rates. The 10 km cutoff has been used previously to define the not-in-city (unexposed) group (16). The 3 km cut point, which has been used in other reports on the LSS (8,17,25), was chosen because the estimated maximum possible dose at this distance is low, that is, less than annual natural background radiation levels.
Age effects on the ERR and EAR were modeled as log-linear in log age and hence can be described as proportional to age to a (constant) power. In models that include age effects on the excess risk, the dose–response slope is the sex-averaged risk for a 50 year old who received 1 Sv. Although age 50 is near the upper end of the age range used in these analyses, most of the cancers in these cohorts were diagnosed among participants between the ages of 45 and 55.
The shape of the dose–response curve was examined using methods described in several recent reports on cancer incidence and mortality in the LSS (8,17,30). These methods make use of both linear quadratic and nonparametric (dose category–specific) descriptions of the dose response. The extent of nonlinearity in the linear quadratic model was described in terms of the curvature, which was defined as the ratio of the quadratic coefficient to the linear coefficient. The curvature is zero in a linear model; it is negative if the dose response is concave downward and positive if it is upward, becoming infinite for a pure quadratic dose response. Because radiation protection is concerned with curvature at low doses, we focused on the 0 to 2 Sv dose range to reduce the effects that high-dose exposures might have (due, eg, to cell killing or dose error) on inference about the nature of the dose response at lower doses.
In the linear quadratic models, we allowed the coefficient of the linear term in dose to differ for in utero and childhood exposures but constrained the curvature to be the same for the two groups. The nonparametric descriptions of the dose response assumed that in utero and childhood exposure risks were proportional, with the same constant of proportionality over all dose categories. These rather strong assumptions were necessary because of the small size of the in utero–exposed cohort.
Hypothesis tests and confidence intervals were based on likelihood ratio tests applied to the profile likelihood (31). Ninety-five percent CIs were used for specific model parameters. All statistical tests were two-sided, and P values less than .05 were considered statistically significant.
|
Results |
|---|
|
TopAbstractContext and CaveatsSubjects and MethodsResultsDiscussionFundingReferencesNotes |
|---|
Descriptive Statistics and Crude Rates
In the full cohort, 1216 solid cancers were recorded during the follow-up period (January 1, 1958, to December 31, 1999), including 901 first primary cancers that were diagnosed before age 55. A total of 34 of these first primary tumors occurred among people whose exposure status or DS02 dose was unknown, and 124 occurred among people who did not reside in the catchment area at the time of diagnosis (Table 2). Dose–response analyses were based on the remaining 743 cancers.
|
The 743 eligible first primary solid cancers included 336 cancers among men and 407 among women. Cancers of the digestive system were the most common, accounting for 70% of male and 30% of female cancers, and nearly half of the cancers were stomach cancers. Cancers of the breast and reproductive organs accounted for 48% of the cancers among women. Thyroid cancers accounted for 3% of male and 11% of female cancers. Only eight of the solid cancers were diagnosed during adolescence (ie, between ages 14 and 19), of which seven were among the early childhood exposure group (including cancers of the stomach, bone, soft tissue, skin, and thyroid and two central nervous system tumors) and one in the in utero group (a Wilms tumor diagnosed at age 14). In large part, the types of cancers in these cohorts seem consistent with what one would expect in an unexposed young adult Japanese population. Additional information on the distribution of types of cancer by sex is available as supplementary material (Supplementary Table 1, available online).
Background Rate Models
Because the members of these study cohorts were born within a few years of each other and all were exposed at the same time, there is little likelihood of birth cohort effects on the baseline rates. Thus, the primary factors considered in modeling baseline rates were attained age and sex. However, we also looked for evidence of differences in the baseline rate level with exposure cohort (in utero, childhood), city, and location at the time of the bombs (proximal, distal, not in city, or unknown exposure status). These analyses were carried out with allowance for separate dose effects for in utero and childhood exposure.
Baseline rates and the nature of their variation with age differed by sex. For both men and women, the log age-specific rates were well described by a linear quadratic function in log age. The quadratic term in log age was statistically significant for men (P = .008) but not for women (P = .10). No difference in the Hiroshima and Nagasaki baseline rates was observed (P = .13, Nagasaki to Hiroshima rate ratio = 1.0, 95% CI = 0.85 to 1.2), nor was there any indication of differences between the baseline rates for the in utero and childhood exposure groups (P > .5).
Age-specific baseline rates of solid cancer incidence were estimated for men and women after allowance for a linear radiation dose response (Figure 1). The pattern was typical of many populations in that women had higher rates of solid cancers than men before age 50, and rates for both men and women began to increase dramatically after age 40. The increase in rates between ages 40 and 55 was roughly proportional to age to the fourth power for men and to age to the third power for women.
|
There was statistically significant heterogeneity in the baseline rates for the proximal, distal, and unexposed groups (P < .001). Baseline rates for the distal exposure group were about 50% greater than those for the proximal exposure group (RR = 1.46, 95% CI = 1.20 to 1.77), whereas rates for the not-in-city group were virtually the same as those for the proximal exposure group (RR = 0.99, 95% CI = 0.80 to 1.2). The difference between rates in the distal and proximal group exposure groups was in the same direction as, but considerably larger than, the difference noted in (14,30) for the full LSS cohort. As in most LSS analyses, we included the distal survivors without any special adjustments. Adjusting for possible proximal–distal differences in baseline rates increased risk estimates by about 25% but had little impact on the estimates of temporal patterns that are described below.
Dose Response and Effect Modification
We examined the dose distribution of solid cancers by cohort (Table 3) and calculated crude rates and crude relative risks for three dose categories stratified by sex and attained age (Table 4). Although the number of cancers was not large, especially for the in utero group, the results suggested that risks were elevated among those exposed to doses in excess of 0.2 Sv and that radiation-associated risks for the in utero cohort may have a somewhat different temporal pattern than those for the childhood exposure cohort. These patterns will be explored more formally below.
|
|
Excess Relative Risk Models
In a model with the same time-constant ERR for in utero and childhood exposures, the estimated ERR per Sv (ERR1Sv) was 1.9 (95% CI = 1.4 to 2.6; P < .001). Allowing the dose response for in utero and childhood exposures to differ, the ERR1Sv estimates were 1.3 (95% CI = 0.2 to 2.8) for in utero exposure and 2.0 (95% CI = 1.4 to 2.8) for childhood exposure. The difference between these ERR estimates was not statistically significant (P = .3). Allowing for different ERRs, the estimated numbers of radiation-associated cancers were nine in the in utero group and 87 in the early childhood group.
Using the effect modification model described in equation 1 to describe variation in the ERR with attained age, the ERR decreased with increasing age (P = .046). This decrease was proportional to age to the –1.3 power (95% CI = –2.4 to –0.06). As indicated in the upper portion of Table 5, allowing for this temporal trend, the ERR1Sv estimates at age 50 for in utero and early childhood exposure were 1.0 (95% CI = 0.20 to 2.3) and 1.7 (95% CI = 1.1 to 2.5), respectively. Radiation effect parameter estimates were also determined from a more general model that included a sex effect and allowed different attained age effects for in utero and childhood exposure (Table 5). In this model, the ERR decreased in proportion to age to the power –2.8 for those exposed in utero and to the power –1.1 for early childhood exposure (Figure 2). The difference in the decrease between the two groups was not statistically significant (P = .3). Using this model, the ERR1Sv estimates at age 50 were 0.42 (95% CI = <0.00 to 2.0) and 1.7 (95% CI = 1.1 to 2.5) for in utero and childhood exposures, respectively. There was a weak suggestion of a sex difference in the ERRs (P = .13).
|
|
Excess Absolute Rate Models EAR models with effects for attained age and sex described the data as well as the ERR models discussed above. The EAR for childhood exposure (at age 50, EAR = 56 cancers per 10000 person-years per Sv, 95% CI = 36 to 79) increased statistically significantly with increasing attained age (P < .001), with the increase estimated to be proportional to age cubed (Table 5 and Figure 2, B). However, there was no evidence of a statistically significant change in the EAR with attained age (P > .5) among those exposed in utero (at age 50, EAR = 6.8 cancers per 10000 person-years per Sv, 95% CI = <0 to 49). Because of the small number of radiation-associated cancers in the in utero group, this difference in temporal risk patterns, although striking, was not statistically significant (P = .14).
A statistically significant difference in the EAR estimates of men and women was observed (Table 5). Excess rates for women were about twice those for men.
Shape of Dose–Response Curve For doses in the 0 to 2 Sv range, there was a suggestion of upward curvature in the dose–response curve (P = .09), with a curvature estimate of 1.0 (95% CI = –0.07 to 212). Assuming the same curvature for in utero and early childhood exposures, the low-dose slope in the linear quadratic model for in utero exposure was about 50% of that for the linear model, but this ratio was quite uncertain (95% CI = 0.06% to 300%). A nonparametric dose–response function that was computed by smoothing dose category–specific ERR estimates was similar to the simple linear dose–response function (Figure 3).
|
Variation in Risk by Trimester or Age at Exposure
No variation in the ERR by trimester of exposure was observed for those exposed in utero (P > .5), and the point estimates (at age 50 in a model that allows for effect modification by attained age) were virtually identical: 1.1 (95% CI = <0 to 3.4) for the first trimester, 0.9 (95% CI = <0 to 2.8) for the second trimester, and 1.0 (95% CI = –0.06 to 3.7) for the third trimester. In addition, no variation in risks with age at exposure was observed for those with early childhood exposure (P > .5). The ERR1Sv estimates (at age 50) were 1.8 (95% CI = 1.1 to 2.8) for those exposed before age 3 and 1.5 (95% CI = 0.8 to 2.5) for those exposed at ages 3–5.
|
Discussion |
|---|
|
TopAbstractContext and CaveatsSubjects and MethodsResultsDiscussionFundingReferencesNotes |
|---|
This study provides direct evidence that radiation exposure is associated with increased risks of adult-onset solid cancers in atomic bomb survivors exposed in utero or in early childhood. For those exposed in early childhood, the ERRs may decrease with time. The absolute risks among those exposed in utero are therefore likely to be considerably lower than simple projections based on studies of childhood cancers in other in utero–exposed populations [which have been estimated to be approximately 6% per Sv by age 15 (19)] and may be lower than absolute risks among those exposed early in life. However, additional follow-up of this cohort is necessary before definitive conclusions can be made about the nature of the risks for those exposed in utero.
This study is one of the only cohort studies of in utero exposure with long-term, continuous active follow-up. This study also provides a unique opportunity to compare effects of in utero and early childhood exposures. However, the power of the study to characterize temporal patterns is limited by the small number of cohort members who received appreciable radiation exposures (eg, >100 mSv), especially among those exposed in utero, and by the fact that the oldest surviving in utero exposed cohort members were only 55 years of age at the end of follow-up. Because of these limitations, site-specific analyses are not yet feasible. However, the types of cancers seen to date (ie, primarily stomach, lung, and breast cancer) appear to be typical of what is seen in Japanese populations (24). Furthermore, because comprehensive data on solid cancer incidence are unavailable for the period from 1945 to 1957, this study cannot provide information on the effect of radiation on the incidence of childhood cancers.
Cancer incidence in the early childhood cohort with 1 year less follow-up than in this study was considered in the recently published analyses of cancer incidence in the full LSS cohort of atomic bomb survivors (25) (which includes the early childhood cohort considered here). In those analyses, simple parametric models were used to describe variation in the excess risks with attained age and age at exposure and the early childhood exposure risk estimates are similar to those obtained directly from the analyses of the early childhood performed in this study.
This study of atomic bomb survivors is one of the few human studies that have specifically examined adult-onset cancers following in utero exposure. Earlier analyses of solid cancer mortality in this cohort (7) provided some indication of elevated rates among those exposed in utero but no evidence of differences in excess rates for in utero and early childhood exposures. Although follow-up for the current analyses began more than 7 years after the start of follow-up for the mortality analyses, the number of cancers used in the current analyses (n = 94) is considerably greater than the number of deaths considered in the mortality analyses (n = 57). This increase is due to the inclusion of follow-up at older ages and because less fatal types of cancer, such as breast and thyroid cancer, account for a relatively high proportion of cancers seen in young adults. We are aware of only one other relevant study in a different population, in which cancer mortality to age 49 was examined among 3097 residents near the Techa River who were exposed to radiation in utero and/or postnatally before the age of 5 (32). In that study, prenatal total body doses ranged from 0 to 0.2 Gy and postnatal doses ranged from 0 to 0.46 Gy and a non–statistically significant excess of solid cancers (30 observed, 25.4 expected) was found. The combined prenatal and postnatal bone marrow dose, which averaged 0.3 Gy and ranged up to 2.0 Gy, was nearly statistically significantly associated with leukemia incidence (P = .09).
Little or no apparent dose response was found for chromosome aberrations among in utero atomic bomb survivors (33), and mouse experimental data (34) suggest that chromosome aberrations do not persist after in utero exposure. The lack of a chromosome aberration dose response among the in utero exposed group may be related to the differences in excess risks for the in utero and early childhood exposure groups. Excess mammary tumors have been seen in rats (35) and excess liver tumors have been observed in mice (36) after in utero irradiation, primarily after doses of greater than 2 Gy. Fetal exposure of beagles to either 0.16 or 0.8 Gy led to increases in lymphoma incidence and in total lifetime fatal malignancies (37). However, other studies of mice and dogs (38–43) suggest that cancer risks associated with in utero exposure may be lower than those associated with postnatal exposures. Notably, Upton et al. (43) found no excess leukemia or cancer risk in RF mice after in utero exposure to 3 Gy, Di Majo et al. (36) found no excess cancer in BC3F1 mice after in utero exposure to 300 mGy, and Ellender et al. (44) reported no excess of intestinal tumors in Apc/Min+ mice after acute in utero exposure to 2 Gy x-rays. However, each of these studies showed increased risks following comparable doses administered postnatally.
Thousands of pregnant women are exposed to radiation each year, either occupationally or as patients, and in utero exposure is still a public health concern (45,46). Several reviews (19,47,48) have summarized the numerous studies on fetal x-ray exposures and childhood cancer with general support for an association between fetal exposure and childhood leukemia. However, there is less consensus regarding fetal radiation exposure and solid cancer risk, ranging from doubts about whether such an effect even exists (47) to being generally positive but with caveats (48) and to a conclusion that the total childhood cancer risk is large (19)—an absolute risk on the order of 6% per Gy. Much less is known about the long-term health consequences of in utero radiation exposure.
The present data suggested that increases in risks of adult-onset cancer among those exposed to radiation in utero may be smaller than for those exposed in early childhood. Moreover, we found a statistically significant decrease in the ERR for adult-onset solid cancer with increasing attained age for in utero as well as for early childhood exposures to radiation, and this decrease may be more marked for those exposed in utero than as children. The difference in temporal patterns for in utero and early childhood exposures was most striking when the radiation effects were described in terms of the EAR, with the estimated EAR for in utero exposure being virtually constant over the age range considered here and that for postnatal exposure increasing markedly with age. This apparent difference suggests that lifetime risks following in utero exposure may be considerably lower than for early childhood exposures. Further follow-up is needed to determine whether this is the case. Whether or not differences in the level and temporal pattern of excess risks for in utero and early childhood exposures to radiation prove to be statistically significant in future analyses, the finding of a decrease in the ERR with increasing age for both in utero and early childhood exposures in the atomic bomb survivor data indicates that lifetime risks of cancer in those exposed in utero are likely to be considerably less than projections based on relative risks derived from studies of childhood cancer incidence (19).
Atomic bomb survivors who were exposed to radiation in utero are just reaching ages at which baseline cancer rates increase markedly. Thus, further follow-up of this cohort is needed to provide new information on risks of adult-onset cancers following in utero radiation exposure.
|
Funding |
|---|
|
TopAbstractContext and CaveatsSubjects and MethodsResultsDiscussionFundingReferencesNotes |
|---|
RERF is a private nonprofit foundation funded equally by the Japanese Ministry of Health, Labour, and Welfare and the United States Department of Energy through the National Academy of Sciences; US National Cancer Institute (NCI; N01-CP31012); Division of Cancer Epidemiology and Genetics, NCI Intramural Research Program.
|
NOTES |
|---|
|
TopAbstractContext and CaveatsSubjects and MethodsResultsDiscussionFundingReferencesNotes |
|---|
This publication is based on research performed at the RERF, Hiroshima and Nagasaki, Japan. The analyses make use of data obtained from the Hiroshima and Nagasaki Tumor registries that are managed by RERF on behalf of the Hiroshima City and Hiroshima and Nagasaki Prefectures. We are especially grateful to the RERF staff involved in the ascertainment of vital status and cause of death and the tumor registry staff who trace and abstract records of cancer diagnoses.
The sponsors had no role in the analysis, the interpretation of the results, the preparation of the manuscript, or the decision to submit the manuscript for publication.
|
REFERENCES |
|---|
|
TopAbstractContext and CaveatsSubjects and MethodsResultsDiscussionFundingReferencesNotes |
|---|
1. Kato H, Keehn RJ. Mortality in live born children who were in-utero at the time of the atomic bombs (ABCC TR 13-66). (1966) Hiroshima, Japan: Atomic Bomb Casualty Commission.
2. Jablon S, Kato H. Childhood cancer in relation to prenatal exposure to atomic-bomb radiation. Lancet (1970) 2(681):1000–1003.[Web of Science][Medline]
3. Kato H. Mortality in children exposed to the A-bombs while in utero, 1945–1969. Am J Epidemiol (1971) 93(6):435–442.
4. Ishimaru T, Ichimaru M, Mikami M. Leukemia incidence among individuals exposed in utero, children of atomic bomb survivors, and their controls; Hiroshima and Nagasaki (RERF TR 11-81). (1981) Hiroshima, Japan: Radiation Effects Research Foundation.
5. Yoshimoto Y, Kato H, Schull WJ. Risk of cancer among children exposed in utero to A-bomb radiations, 1950–84. Lancet (1988) 2(8612):665–669.[Web of Science][Medline]
6. Yoshimoto Y, Delongchamp R, Mabuchi K. In-utero exposed atomic bomb survivors: cancer risk update. Lancet (1994) 344(8918):345–346.[CrossRef][Web of Science][Medline]
7. Delongchamp RR, Mabuchi K, Yoshimoto Y, Preston DL. Cancer mortality among atomic bomb survivors exposed in utero or as young children, October 1950–May 1992. Radiat Res (1997) 147(3):385–395.[CrossRef][Web of Science][Medline]
8. Preston DL, Pierce DA, Shimizu Y, et al. Effect of recent changes in atomic bomb survivor dosimetry on cancer mortality risk estimates. Radiat Res (2004) 162(4):377–389.[CrossRef][Web of Science][Medline]
9. Young RW, Kerr GD, eds. Reassessment of the Atomic-Bomb Radiation Dosimetry for Hiroshima and Nagasaki: Dosimetry System 2002. (2005) Hiroshima, Japan: Radiation Effects Research Foundation.
10. Cullings HM, Fujita S, Funamoto S, Grant EJ, Kerr GD, Preston DL. Dose estimation for atomic bomb survivor studies: its evolution and present status. Radiat Res (2006) 166(1):219–254.[CrossRef][Web of Science][Medline]
11. Fujita S. Versions of DS86. RERF Update (1989) 1(2):2–3.
12. Kerr GD. Organ dose estimates for the Japanese atomic-bomb survivors. Health Phys (1979) 37(4):487–508.[Web of Science][Medline]
13. Hashizume T, Maruyama T, Nishizawa K, Nishimura A. Dose estimation of human fetus exposed in utero to radiation from atomic bombs in Hiroshima and Nagasaki. J Radiat Res (1973) 14(4):346–362.[CrossRef][Medline]
14. Cologne JB, Preston DL. Longevity of atomic-bomb survivors. Lancet (2000) 356(9226):303–307.[CrossRef][Web of Science][Medline]
15. Mabuchi K, Soda M, Ron E, et al. Cancer incidence in atomic bomb survivors. Part I: use of the tumor registries in Hiroshima and Nagasaki for incidence studies. Radiat Res (1994) 137(2 suppl):S1–S16.[CrossRef][Web of Science][Medline]
16. Beebe G, Usagawa M. The Major ABCC Samples. (1968) Hiroshima, Japan: Radiation Effects Research Foundation (RERF TR 12-68).
17. Preston DL, Shimizu Y, Pierce DA, Suyama A, Mabuchi K. Studies of mortality of atomic bomb survivors. Report 13: solid cancer and noncancer disease mortality: 1950–1997. Radiat Res (2003) 160(4):381–407.[CrossRef][Web of Science][Medline]
18. Thompson DE, Mabuchi K, Ron E, et al. Cancer incidence in atomic bomb survivors. Part II: solid tumors, 1958–1987. Radiat Res (1994) 137(2 suppl):S17–S67.[Web of Science][Medline]
19. Doll R, Wakeford R. Risk of childhood cancer from fetal irradiation. Br J Radiol (1997) 70(830):130–139.[Abstract]
20. Sposto R, Preston DL. Correcting for catchment area nonresidency in studies based on tumor registry data. (1992) Hiroshima, Japan: Radiation Effects Research Foundation. (RERF CR 1–92).
21. Imaizumi M, Usa T, Tominaga T, et al. Long-term prognosis of thyroid nodule cases compared with nodule-free controls in atomic bomb survivors. J Clin Endocrinol Metab (2005) 90(9):5009–5014.
22. Yamada M, Kodama K, Fujita S, et al. Prevalence of skin neoplasms among the atomic bomb survivors. Radiat Res (1996) 146(2):223–226.[CrossRef][Web of Science][Medline]
23. Fritz AG, Percy C, Jack A, Sobin LH, Parkin MD. International classification of diseases for oncology (ICD-O-3). (2000) 3rd ed. Geneva, Switzerland: World Health Organization.
24. Parkin DM, Whelan SL, Ferlay J, Teppo L, Thomas DB, eds. Cancer Incidence in Five Continents, Vol. VIII. IARC Scientific Publications No. 155 (2002) Lyon, France: IARC.
25. Preston DL, Ron E, Tokuoka S, et al. Solid cancer incidence in atomic bomb survivors: 1958–1998. Radiat Res (2007) 168(1):1–64.[CrossRef][Web of Science][Medline]
26. Breslow N, Day N. Statistical Methods in Cancer Research Vol 2. The Design and Analysis of Cohort Studies (1988) Lyon, France: International Agency for Research on Cancer.
27. Preston DL, Lubin JH, Pierce DA, McConney ME. Epicure Users Guide (1993) Seattle, WA: Hirosoft International Corporation.
28. Pierce DA, Stram DO, Vaeth M. Allowing for random errors in radiation dose estimates for the atomic bomb survivor data. Radiat Res (1990) 123(3):275–284.[CrossRef][Web of Science][Medline]
29. Pierce DA, Preston DL, Stram DO, Vaeth M. Allowing for dose-estimation errors for the A-bomb survivor data. J Radiat Res (1991) 32(suppl):108–121.[CrossRef][Medline]
30. Pierce DA, Preston DL. Radiation-related cancer risks at low doses among atomic bomb survivors. Radiat Res (2000) 154(2):178–186.[Web of Science][Medline]
31. Cox DR, Hinckley DV. Theoretical Statistics (1974) New York: Chapman Hall.
32. Ostroumova E, Akleyev A, Hall P. Infant mortality among offspring of individuals living in the radioactively contaminated Techa River area, southern Urals. Acta Medica Nagasakiensia (2005) 50(1):23–28.
33. Ohtaki K, Kodama Y, Nakano M, et al. Human fetuses do not register chromosome damage inflicted by radiation exposure in lymphoid precursor cells except for a small but significant effect at low doses. Radiat Res (2004) 161(4):373–379.[CrossRef][Web of Science][Medline]
34. Nakano M, Kodama Y, Ohtaki K, et al. Chromosome aberrations do not persist in the lymphocytes or bone marrow cells of mice irradiated in utero or soon after birth. Radiat Res (2007) 167(6):693–702.[CrossRef][Web of Science][Medline]
35. Inano H, Suzuki K, Onoda M, Yamanouchi H. Susceptibility of fetal, virgin, pregnant and lactating rats for the induction of mammary tumors by gamma rays (1996) 145(6):708–713.
36. Di Majo V, Coppola M, Rebessi S, Covelli V. Age-related susceptibility of mouse liver to induction of tumors by neutrons. Radiat Res (1990) 123(2):227–234.
37. Benjamin SA, Lee AC, Angleton GM, Saunders WJ, Keefe TJ, Mallinckrodt CH. Mortality in beagles irradiated during prenatal and postnatal development. II. Contribution of benign and malignant neoplasia. Radiat Res (1998) 150(3):330–348.[CrossRef][Web of Science][Medline]
38. Di Majo V, Rebessi S, Pazzaglia S, Saran A, Covelli V. Carcinogenesis in laboratory mice after low doses of ionizing radiation. Radiat Res (2003) 159(1):102–108.[CrossRef][Web of Science][Medline]
39. Schmahl W. Synergistic induction of tumours in NMRI mice by combined foetal X-irradiation with low doses and ethylnitrosourea administered to juvenile offspring. Carcinogenesis (1988) 9(8):1493–1498.
40. Seed TM, Kaspar LV, Tolle DV, Fritz TE. Chronic radiation leukemogenesis: postnatal hematopathologic effects resulting from in-utero irradiation. Leuk Res (1987) 11(2):171–179.[CrossRef][Web of Science][Medline]
41. Sasaki K, Kasuga T. Life-shortening and carcinogenesis in mice irradiated at the prenatal period with gamma rays. In: Life-Span Radiation Effects Studies in Animals: What Can They Tell Us? (CONF-830951)—Thompson RC, Mahaffey JA, eds. (1986) Washington, DC: Office of Scientific Information and Technology, Department of Energy.
42. Schmahl W. Long-term effects after prenatal irradiation. In: Effects of Prenatal Irradiation With Special Emphasis on Late Effects—Streffer C, Patrick G, eds. (1984) Luxembourg: Commission of the European Communities.
43. Upton AC, Odell TT Jr, Sniffen EP. Influence of age at time of irradiation on induction of leukemia and ovarian tumors in RF mice. Proc Soc Exp Biol Med (1960) 104(5):769–772.[CrossRef][Medline]
44. Ellender M, Harrison JD, Kozlowski R, Szluinska M, Bouffler SD, Cox R. In utero and neonatal sensitivity of ApcMin/+ mice to radiation-induced intestinal neoplasia. Int J Radiat Biol (2006) 82(3):141–151.[CrossRef][Web of Science][Medline]
45. International Commission on Radiological Protection. ICRP publication 84: pregnancy and medical radiation. Ann ICRP (2000) 30(1):1–43.[Medline]
46. Cousins C. Radiation and pregnancy. Advances in Radiation Protection in Medicine (2007) Bethesda, MD: National Council on Radiation Protection and Measurements. www.ncrponline.org/Annual_Mtgs/2007_Ann_Meet_Prog.pdf. Accessed February 19, 2008.
47. Boice JD Jr, Miller RW. Childhood and adult cancer after intrauterine exposure to ionizing radiation. Teratology (1999) 59(4):227–233.[CrossRef][Web of Science][Medline]
48. International Commission on Radiological Protection. ICRP publication 90: Biological effects after prenatal irradiation (embryo and fetus). Ann ICRP (2003) 31(1–2):1–200.
Manuscript received September 13, 2007; revised December 12, 2007; accepted February 4, 2008.
Correspondence about this Article
Related Article in JNCI
CiteULike 
Connotea 
Del.icio.us What's this?
J Natl Cancer Inst 2008 100: 1482-1483.
J Natl Cancer Inst 2008 100: 375.
This article has been cited by other articles:
|
|
 |
|
  D. Milkovic, V. Garaj-Vrhovac, M. Ranogajec-Komor, S. Miljanic, G. Gajski, Z. Knezevic, and N. Beck Primary DNA Damage Assessed With the Comet Assay and Comparison to the Absorbed Dose of Diagnostic X-rays in Children International Journal of Toxicology, September 1, 2009; 28(5): 405 - 416. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Iwanaga, M. Tagawa, K. Tsukasaki, T. Matsuo, K.-i. Yokota, Y. Miyazaki, T. Fukushima, T. Hata, Y. Imaizumi, D. Imanishi, et al. Relationship between monoclonal gammopathy of undetermined significance and radiation exposure in Nagasaki atomic bomb survivors Blood, February 19, 2009; 113(8): 1639 - 1650. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. B. Richardson and S. Wing Re: Solid Cancer Incidence in Atomic Bomb Survivors Exposed In Utero or as Young Children J Natl Cancer Inst, October 15, 2008; 100(20): 1482 - 1483. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||