Several studies suggest that maternal radiosensitivity increases during pregnancy, and that this effect is mediated, at least in part, by the elevated levels of steroid and/or nonsteroid maternal hormones during pregnancy [
The Adult Health Study (AHS) population at the Radiation Effects Research Foundation (RERF) includes approximately 20,000 participants exposed to radiation during the bombings of Hiroshima and Nagasaki, Japan in 1945 that have been followed with biennial medical exams. In addition to significant effects on cancer induction, there are a number of long-term dose-response effects detectable in A-bomb survivors [
Our analysis focused on previously evaluated general population dose-response lateeffects of radiation in A-bomb survivors on serum cholesterol (increased) [
We herein assess whether pregnancy at the time of the bombings increased a woman’s radiosensitivity by comparing the long-term effects doseresponses in pregnant women with that of non-pregnant women in the seven parameters noted above.
For purposes of analysis, the seven parameters were broadly categorized as “indirect” or “direct” effects. The delineation was loosely based on whether or not the endpoint measured depended on a more “indirect” effect of radiation, that is, requiring a cellular intermediary or sustained cytokine interaction to manifest the end effect (cholesterol, WBC, ESR, HGB) or was the result of a more “direct” effect on a cell-type (stable chromosome aberration, GPA locus mutation, and naïve CD4 T-cell count). Those categories are not strictly defined, but are helpful in interpreting our results.
Subject data was obtained from mothers of the 3,631
Due to the disparity in data availability between the “indirect” and “direct” effects, we conducted two separate cross-sectional analyses for this report. The first analysis was performed on data collected during AHS Cycle 2 (1960–1962), which had the maximum participation (250) of women that were pregnant ATB. Laboratory data from this cycle included the “indirect” effect markers (serum cholesterol, WBC, ESR, and HGB). The second analysis involved data from women that participated in AHS Cycle 22 (2000–2002) when “direct” effect test data (chromosome aberration, GPA locus mutation, and naïve CD4 T-cell count) became available. Due to attrition, data from a maximum of only 85 pregnant women was available for analysis in Cycle 22.
Separate and distinct populations for each of the cross-sectional analyses were also drawn from the AHS database for age and city-matched women that were not pregnant ATB. Since the number of pregnant women was limited, we expanded the age range for non-pregnant women to include as large an amount of women as possible. Pregnant women and non-pregnant women with unknown dose were excluded from either analysis. The final ratio of non-pregnant to pregnant women was approximately 10 : 1 for all study parameters. Demographic information for the entire study cohort is presented in Table
Distribution of pregnant and non-pregnant women in A-bomb survivors by cycle, city, and age at the bombs.
Cycle | City | Pregnant | Age ATB | |||||
---|---|---|---|---|---|---|---|---|
<20 | <25 | <30 | <35 | 35+ | Total | |||
2 | Hiro | No | 326 | 480 | 368 | 389 | 1264 | 2827 |
Yes | 2 | 50 | 35 | 33 | 38 | 158 | ||
Naga | No | 167 | 186 | 127 | 85 | 211 | 776 | |
Yes | 4 | 10 | 17 | 5 | 9 | 45 | ||
22 | Hiro | No | 217 | 319 | 145 | 99 | 45 | 825 |
Yes | 3 | 32 | 18 | 7 | 1 | 61 | ||
Naga | No | 139 | 129 | 64 | 31 | 13 | 376 | |
Yes | 3 | 6 | 10 | 2 | 0 | 21 |
H: Hiroshima; N: Nagasaki.
Number of pregnant and non-pregnant women in A-bomb survivors by measurements, trimester and city.
Items | Pregnant | Not pregnant | Two sided | |
---|---|---|---|---|
Cholesterol | ||||
Number | 58 | 819 | ||
Median age ATB (range) | 27 (17, 43) | 31 (17, 47) | 0.090 | |
First trimester (%) | 22 (37.9) | |||
Second trimester (%) | 26 (44.8) | |||
Third trimester (%) | 10 (17.2) | |||
Hiroshima (%) | 36 (62.1) | 577 (70.5) | 0.231 | |
Nagasaki (%) | 22 (37.9) | 242 (29.5) | ||
Median dose (mGy) ((range) | 172.0 (0, 2862) | 115.5 (0, 3380) | 0.170 | |
WBC | ||||
Number | 166 | 2368 | ||
Median age ATB (range) | 28 (17, 44) | 30 (17, 47) | 0.019 | |
First trimester (%) | 58 (34.9) | |||
Second trimester (%) | 67 (40.4) | |||
Third trimester (%) | 41 (24.7) | |||
Current Smoker (%) | 24 (14.5) | 403 (17.0) | >0.5 | |
Former (%) | 3 (1.8) | 56 (2.4) | ||
Never (%) | 139 (83.7) | 1909 (80.6) | ||
History of Inflammatory of disease (%) | 40 (24.1) | 643 (27.2) | 0.443 | |
Cancer (%) | 1 (0.6) | 12 (0.5) | 0.693 | |
Hiroshima (%) | 129 (77.7) | 1862 (78.6) | 0.856 | |
Nagasaki (%) | 37 (22.3) | 506 (21.4) | ||
Median dose (mGy) (range) | 97.2 (0, 2862) | 134.6 (33.325) | 0.223 | |
ESR | ||||
Number | 101 | 1466 | ||
Median age ATB, range | 28 (19, 43) | 32 (19, 47) | 0.001 | |
First trimester (%) | 35 (34.7) | |||
Second trimester (%) | 41 (40.6) | |||
Third trimester (%) | 25 (24.8) | |||
Current Smoker (%) | 14 (13.9) | 249 (17.0) | 0.206 | |
Former (%) | 0 (0.0) | 33 (2.3) | ||
Never (%) | 87 (86.1) | 1184 (0.808) | ||
History of Inflammatory of disease (%) | 17 (0.168) | 340 (23.2) | 0.177 | |
Cancer (%) | 1 (1.0) | 11 (0.8) | 0.747 | |
Hiroshima (%) | 101 (1.0) | 1466 (1.0) | < 0.001 | |
Nagasaki (%) | 0 (0.0) | 0 (0.0) | ||
Median dose (mGy) (range) | 76.3 (0, 2391) | 157.4 (0, 3278) | 0.033 |
HGB | ||||
---|---|---|---|---|
Number | 166 | 2368 | ||
Median age ATB, range | 28 (17, 44) | 30 (17, 47) | 0.019 | |
First trimester (%) | 58 (34.9) | |||
Second trimester (%) | 67 (40.4) | |||
Third trimester (%) | 41 (24.7) | |||
Current Smoker (%) | 24 (14.5) | 403 (17.0) | 0.607 | |
Former (%) | 3 (1.8) | 56 (2.4) | ||
Never (%) | 139 (83.7) | 1909 (80.6) | ||
History of Inflammatory of disease (%) | 40 (24.1) | 643 (27.2) | 0.443 | |
Cancer (%) | 1 (0.6) | 12 (0.5) | 0.693 | |
Hiroshima (%) | 129 (77.7) | 1862 (78.6) | 0.856 | |
Nagasaki (%) | 37 (22.3) | 506 (21.4) | ||
Median dose (mGy) (range) | 97.2 (0, 2862) | 134.6 (0, 3325) | 0.223 | |
Stable chromosome aberration frequency | ||||
Number | 27 | 313 | ||
Median age ATB (range) | 23 (20,34) | 21 (17, 33) | 0.003 | |
First trimester (%) | 10 (37.0) | |||
Second trimester (%) | 11 (40.7) | |||
Third trimester (%) | 6 (22.2) | |||
Hiroshima (%) | 18 (66.7) | 213 (68.1) | 0.947 | |
Nagasaki (%) | 9 (33.3) | 100 (32.0) | ||
Median Dose (mGy) (range) | 121.7 (0, 1518) | 287.6 (0, 3252) | 0.328 | |
GPA locus mutation | ||||
Number | 26 | 327 | ||
Median age ATB (range) | 24 (18, 32) | 21 (17, 38) | 0.005 | |
First trimester (%) | 7 (26.9) | |||
Second trimester (%) | 11 (42.3) | |||
Third trimester (%) | 8 (30.8) | |||
Hiroshima (%) | 23 (88.5) | 214 (65.4) | 0.029 | |
Nagasaki (%) | 3 (11.5) | 113 (34.6) | ||
Median Dose (mGy), range | 4.3 (0, 1621.4) | 80.1 (0, 2977) | 0.322 | |
Naïve CD4 T cell count | ||||
Number | 75 | 887 | ||
Median age ATB (range) | 24 (17, 38) | 21 (17, 38) | <0.001 | |
First trimester (%) | 26 (34.7) | |||
Second trimester (%) | 33 (44.0) | |||
Third trimester (%) | 16 (21.3) | |||
Hiroshima (%) | 57 (76.0) | 606 (68.3) | 0.211 | |
Nagasaki (%) | 18 (24.0) | 281 (31.7) | ||
Median Dose (mGy), range | 66.9 (0, 2027) | 81.9 (0, 3311) | 0.554 |
Nonfasting serum cholesterol levels were measure by the Kendall-Abell method [
Anticoagulated blood samples were collected for measurement. WBC counts were obtained manually by the Melangeur method [
Anti-coagulated blood samples were collected for measurement. ESR was measured by the Wintrobe method [
Hemoglobin level was measured by a manual procedure with quality control procedures in place to maintain reproducibility and consistency in laboratory results [
Chromosome spreads from peripheral blood lymphocytes were prepared and Giemsa stained by conventional procedures [
A single-beam cell sorter, FACStar (Becton Dickinson Immunocytometry Systems, San Jose, CA), was used to sort four types of variant erythrocytes lacking the expression of one GPA allele that were distinguished from normal MN heterozygous cells. Those four variants are M
Analytical flow cytometry was conducted in a FACScan machine (BD Biosciences, San Jose, CA). Expression of CD45RA and CD4 molecules was analyzed with FITC-labeled and PE-labeled antibodies, respectively. In every measurement, approximately 20,000 cells were analyzed. The percentage of CD45RA
Separate linear regression analyses were performed based on the indirect and direct outcomes selected. The outcomes for WBC, cholesterol and GPA were logtransformed with base 10 to approximate the transformed outcome as normal distribution.
For three of the indirect outcomes (HGB, WBC count, and ESR), our model accounted for the influence of city (Hiroshima or Nagasaki), smoking history (current, former, never), age ATB (age at exposure), history of inflammatory disease/inflammatory process (yes or no), cancer history (yes or no), bone marrow radiation dose (DS02) in Gy (continuous variable), and pregnancy (yes or no ATB). Thus, the assumed simple model for the
For total cholesterol, and the direct outcomes (naïve CD4 T-cell count and GPA mutation rate), the independent variables are city, adjusted age ATB, radiation dose and pregnancy. Thus the assumed model for the
Because of the level of dispersion in the chromosome aberration frequency results, a binomial model with linear probability was used that had city, age ATB, smoking, radiation dose, and pregnancy indicator as independent variables and overdispersion was accounted for. The model for chromosome aberration probability is
A tabular summary of our findings is presented in Tables 3.
No overall dose-response effect was found in serum cholesterol, WBC count, or ESR. Neither was there a statistically significant difference in the slopes of the regression lines for pregnant versus non-pregnant women. There was a suggestion of a statistically significant overall dose-response decrease (
Chromosome aberration frequency demographic information is presented in Table
Scatter plot of chromosome aberration frequency in pregnant versus non-pregnant women with linear trend lines for both groups. Open circles reflect control data from non-pregnant women, while solid triangles reflect data from women pregnant ATB.
Scatter plot of chromosome aberration frequency in pregnant versus non-pregnant women (first two trimesters separated from third trimesters) with linear trend lines for each group. Open diamonds reflect control data from non-pregnant women, solid triangles reflect women pregnant in the first two trimesters ATB, and solid squares reflect women pregnant in the third trimester ATB.
A statistically significant overall dose response increase in GPA locus mutation rate of 26.25 Gy−1 (
Several authors have found that the number of chromosome aberrations in
There are several preclinical studies that suggest that a biochemical environment that emulates pregnancy may be associated with an increased rate of radiosensitivity, albeit early radiosensitivity as measured primarily by chromosome aberrations in PBL shortly after irradiation. This literature is summarized below.
Sharma and Das. demonstrated that there was a statistically significant increase in spontaneous sister chromatid exchanges (SCE) in human PBL from females in the third trimester of pregnancy compared to non-pregnant females (10.7 versus 6.5,
Roberts et al. demonstrated a statistically significant difference between females and male subjects in the total number of chromosome aberrations (
Building on previous work showing similar results in a pregnant mouse model [
Ricoul et al. subsequently demonstrated that the
Kanda and Hayata investigated the effect of incubation of human PBL with various levels of estradiol during and after irradiation to 3 Gy at 1.4 Gy/min [
Baeyens et al. found higher levels of micronuclei in an
Similar to the speculation by Kanda et al. noted above, Vares et al. investigated the role of progesterone in modulating the effects of radiation on several human breast cancer cell lines [
In summary, several researchers have found that
In conducting this retrospective analysis, we expected that the radiation dose-response in pregnant women, as indicated by the slope of the individual regression line, for our selected parameters would be steeper than that found in the group of non-pregnant female A-bomb survivors.
It should be noted that there is a potential for significant selection bias at several levels in our analysis. First, as noted in the methods section, only a small fraction of the women that were pregnant at the time of bombings participated in the AHS, and not all of these women had appropriate data available for analysis. In addition, the selection of women to participate in the original chromosome aberration and GPA locus mutation rates published by previous authors was predicated on a higher radiation exposure dose so extrapolation of the dose rate effect to the lower doses experienced by the women in the present study may be problematic.
Wong et al. reported a significant dose-dependent increase in serum cholesterol level in a study involving more than 9,800 A-bomb survivors with a mean dose of 1.09 Sv [
These results are based on data collected during Cycle 2 of the AHS, 13–15 years after the bombings, and may not reflect the true long-term incidence of late effects in these parameters. The published dose-response effects to date have involved the study of responses over much longer time intervals. However, Cycle 2 contained the largest number of women that were pregnant at the time of the bombings, so using a later cycle for analysis would further limit the statistical power of any findings.
In a study involving over 3,000 A-bomb survivors, Kodama et al. demonstrated the existence of stable chromosome aberrations (at least one translocation or inversion per 100 lymphocytes per person) occurred at a rate related to the radiation exposure dose with an alpha/beta ratio of approximately 1.7 Sv (95% CI 0.9–4) [
For our study, an overall positive dose-response chromosome aberration effect was found that is in accordance with that found by Kodama et al. Although their results are not statistically different, both pregnant and non-pregnant women show a significant positive dose-response effect. Graphically in Figure
As shown in Figure
Regarding GPA locus mutation, Kyoizumi et al. studied the mutation rate among more than 1,000 A-bomb survivors with a mean age of 64 and found a doubling dose of 1.2 Sv (95% CI 0.95–1.56) with a minimum dose to detect increased mutation of 0.24 Sv (95% CI 0.041–0.51) [
An overall positive dose-response effect with respect to GPA locus mutation was found that is in accordance with that found by Kodama et al.; the overall dose-response rate was 26 mutations Gy−1. However, there was no significant difference in the regression lines for pregnant versus non-pregnant women.
Regarding naïve CD4 T-cell counts and building on previous work that demonstrated a decreased count in 159 A-bomb survivors exposed to more than 1.5 Gy of radiation [
An overall negative dose-response effect was found that is in accordance with that found by Kusonoki et al.; the overall dose-response rate was a decrease of 1.42 Gy−1. However, there was no significant difference in the regression lines for pregnant versus non-pregnant women.
Numeric result data for all parameters is presented in Table
Summary of expected and actual findings in pregnant A-bomb survivors—direct effects.
Parameter | Expected change with pregnancy | Regression Coefficients | Standard Error | ||
---|---|---|---|---|---|
Log-Cholesterol (143–220 mg/dl) | (Log-mg/dl/Gy) | ||||
Overall | +0.006 | 0.005 | .259 | ||
−0.015 | 0.020 | .435 | |||
Not pregnant | +0.007 | 0.005 | .166 | ||
Difference | .264 | ||||
Log-WBC count (33–91103 cells/cm | (Log-cells/cm3/Gy) | ||||
Overall | −0.001 | 0.004 | .726 | ||
−0.009 | 0.019 | .633 | |||
Not pregnant | +0.002 | 0.004 | .646 | ||
Difference | .573 | ||||
ESR (0–40 mm/hr) | (mm/hr/Gy) | ||||
Overall | +0.485 | 0.427 | .255 | ||
−2.884 | 2.189 | .188 | |||
Not pregnant | +0.617 | 0.435 | .156 | ||
Difference | .117 | ||||
HGB (11–15 g/dl) | (g/dl/Gy) | ||||
Overall | −0.067 | 0.039 | .088 | ||
+0.139 | 0.187 | .456 | |||
Not pregnant | −0.077 | 0.040 | .057 | ||
Difference | .258 |
Parameter | Expected change with pregnancy | Regression Coefficients | Standard Error | ||
---|---|---|---|---|---|
Chromosome aberration | (freq/cell/Gy) | ||||
Overall | +0.0708 | 0.00439 | <.001 | ||
+0.0864 | 0.0194 | <.001 | |||
Not pregnant | +0.0699 | 0.00450 | <.001 | ||
Difference | .403 | ||||
Log-GPA locus mutation | (Log-mut/Gy) | ||||
Overall | +0.281 | 0.029 | <.001 | ||
+0.414 | 0.141 | .004 | |||
Not pregnant | +0.275 | 0.029 | <.001 | ||
Difference | .337 | ||||
Naïve CD-4 T-cell count | (cells/Gy) | ||||
Overall | −1.425 | 0.419 | .001 | ||
−1.551 | 1.939 | .424 | |||
Not pregnant | −1.419 | 0.429 | .001 | ||
Difference | .947 |
In general, diagnostic imaging and radiation therapy are proscribed during pregnancy, more for the well-documented potential for harm to the developing fetus than for the protection of the expectant mother. However, there can be inadvertent radiation exposure for diagnostic imaging during an occult pregnancy. Also, diagnostic images are occasionally obtained in women during the third trimester since by that time the fetus is generally thought to be less susceptible to catastrophic radiation damage (intrauterine death, microcephaly, mental retardation) and the low-rate of long-term cancer induction risk to the fetus may be outweighed by the need for diagnostic imaging or treatment in the mother [
What we learn from studies on radiosensitivity and hormones may have implications for understanding the role of the hormonal milieu on radiation therapy-induced tumor and normal tissue effects. Radiation therapy is commonly used in women who may benefit from antiestrogens and aromatase inhibitors as part of their treatment for breast cancer. It is important to understand if these agents, which affect the production or interaction of estrogen with its cellular receptor, markedly alter the sensitivity of the tumor and/or normal tissues to the effects of radiation therapy. Small studies have demonstrated in vitro cell-cycle redistribution effects and alterations to tumor cell survival curves with concurrent anti-estrogen exposure [
Clinically, several retrospective studies have demonstrated no increased acute or late adverse effects or impact on local control with concurrent versus sequential radiation therapy and hormonal therapy for early stage breast cancer [
To date, there are no published
It is also possible that other metabolic conditions other than the level of circulating hormones may also affect the relative radiosensitivity of women who are pregnant at the time of exposure. For example, as pregnancy is a hypermetabolic state and is associated with increased oxidative stress [
There is no statistically significant evidence of changes in various biomarkers to suggest an increased radiosensitivity in women that were pregnant ATB in Hiroshima and Nagasaki as compared to non-pregnant control subjects. We compared the late effects of changes in serum cholesterol, WBC count, ESR, HGB, stable chromosome aberration, GPA locus mutations, or naïve CD4 T-cell counts.
The increased slope of the trend line in the third trimester with respect to stable chromosome aberrations is suggestive of an increased radiosensitivity, but there are not enough pregnant women in the third trimester in the AHS database to demonstrate statistical significance.
A separate protocol is underway to determine if cancer incidence (particularly breast, skin and thyroid), cancer mortality, or all-cause mortality is increased in women that were pregnant ATB. This study will have the advantage of using the much larger Life Span Study (LSS) database, containing information on approximately 120,000 survivors, and may have sufficient power to identify even small differences.
The relative efficacy and toxicity associated with the concurrent use of anti-estrogens and aromatase inhibitors should be evaluated in a randomized trial setting with long-term followup to determine if these agents change the therapeutic index by altering the production and/or cellular interaction of estrogen.
The authors report no conflict of interests. The authors alone are responsible for the content and writing of the paper.
The views expressed in this paper are those of the author(s) and do not necessarily reflect the official policy or position of the Department of the Navy, Department of Defense, or the United States Government. The author is a military service member. This work was prepared as part of his official duties. Title 17 U.S.C. 105 provides that “Copyright protection under this title is not available for any work of the United States Government.” Title 17 U.S.C. 101 defines a United States Government work as a work prepared by a military service member or employee of the United States Government as part of that person’s official duties.
The Radiation Effects Research Foundation (RERF), Hiroshima and Nagasaki, Japan is a private, nonprofit foundation funded by the Japanese Ministry of Health, Labour and Welfare (MHLW) and the U.S. Department of Energy (DOE), the latter in part through the National Academy of Sciences. This publication was supported by RERF Research Protocol RP 2-75.