If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Division of Radiation Oncology, National Cancer Centre Singapore, SingaporeDivision of Physics and Applied Physics, Nanyang Technological University, Singapore
Gy dose threshold for fetal dose to limit adverse effect of radiation.
•
Estimate fetal dose with Monte Carlo method or phantom measurement before treatment.
•
Shielding and dose reduction techniques to further reduce out-of-field dose.
•
Neutrons contribute most to fetal out-of-field dose in proton therapy.
•
Use of daily CBCT in IGRT for pregnant patient should be carefully considered.
Abstract
This paper aims to review on fetal dose in radiotherapy and extends and updates on a previous work1 to include proton therapy. Out-of-field doses, which are the doses received by regions outside of the treatment field, are unavoidable regardless of the treatment modalities used during radiotherapy. In the case of pregnant patients, fetal dose is a major concern as it has long been recognized that fetuses exposed to radiation have a higher probability of suffering from adverse effects such as anatomical malformations and even fetal death, especially when the threshold is exceeded. In spite of the low occurrence of cancer during pregnancy, the radiotherapy team should be equipped with the necessary knowledge to deal with fetal dose. This is crucial so as to ensure that the fetus is adequately protected while not compromising the patient treatment outcomes. In this review paper, various aspects of fetal dose will be discussed ranging from biological, clinical to the physics aspects. Other than fetal dose resulting from conventional photon therapy, this paper will also extend the discussion to modern treatment modalities and techniques, namely proton therapy and image-guided radiotherapy, all of which have seen a significant increase in use in current radiotherapy. This review is expected to provide readers with a comprehensive understanding of fetal dose in radiotherapy, and to be fully aware of the steps to be taken in providing radiotherapy for pregnant patients.
]. Some of the most common types of cancer among pregnant women include lymphoma, melanoma, leukemia, breast, cervical, and thyroid cancer, of which the majority would require radiotherapy as part of the treatment [
]. During the irradiation of the target, there would inevitably be dose received by the region outside of the target and the primary field edge, resulting in out-of-field (OOF) dose [
]. The three main sources of OOF dose from photon beams are treatment head leakage, scattered radiation from collimators or beam modifiers such as wedges and blocks, and scattered radiation within the patients. For higher-energy beam with energy >10 MV, photoneutron interactions also give rise to an additional OOF dose component through neutron production [
Despite being low, OOF dose to the fetus can have severe effects - lethality, anatomical malformations, mental retardation, growth retardation to name a few [
]. This must lead to a meticulous discussion and planning before radiotherapy is given, which is often a multidisciplinary effort involving the obstetricians, radiation oncologists, medical oncologists, medical physicists and other related professionals [
]. The final goal is to achieve a balance between the risk and benefits, and the patient needs to be fully informed of the risk if they choose to continue with the gestation.
For the purpose of patient management, it is recommended [
] that the fetal dose should be estimated by calculation or performing measurement in a phantom. Special shielding should be designed and constructed to ensure that the fetal dose falls within the acceptable threshold limit. Throughout the radiotherapy course and with fetal growth expected, in vivo measurement also needs to be carefully done to monitor the actual fetal dose.
There have been a number of comprehensive articles or reports addressing various aspects of concerns regarding fetal dose. Some of the examples include the ICRP Publication 84 [
], which all detailed the biological effects of fetal dose based on the absorbed dose level and the gestational age, and recommendations on treating pregnant patients. While the AAPM TG36 report focused mainly on photon radiotherapy, the ICRP Publication 84 touched on broader topics that include diagnostic imaging and nuclear medicine, in addition to radiotherapy.
] are either less physics-oriented or do not include proton therapy and image-guided radiotherapy (IGRT), both of which are increasingly being used in modern radiotherapy. Particle therapy, and in particular proton therapy, is gaining clinical use due to its conformal dose contribution with no exit dose. IGRT is viewed to be indispensable in the treatment with highly conformal dose distribution and hypofractionated regimen. The imaging dose from IGRT which could contribute to fetal dose has been hardly examined in current literatures. This review therefore aims to align and extend our knowledge about fetal dose with the most technologically modern and current radiotherapy practice to date.
In this review paper, fetal dose will first be discussed from the biological and clinical aspects. The main focus will then be on fetal dose delivered by photon therapy, proton or other types of particle therapy in general, and imaging for the purpose of IGRT. In particular, the factors affecting fetal dose level, the dose estimation methods, and the dose reduction techniques including shielding will be covered. For ease of reference, Section III and Section VI(A) will be stratified based on various cancer sites as the dose levels for each site are generally different and so are the corresponding shielding requirements. The outline of this review paper is depicted in Fig. 1.
Fig. 1Outline of topics covered in this review paper.
Radiation impact on fetus based on epidemiological evidences
Ionizing radiation’s main effects on the mammalian embryo and fetus include embryonic and fetal death, malformations, growth impairment, mental retardation, induction of malignancies, and hereditary defects. The absorbed dose, type of radiation, and gestational age at which exposure occurs are some of the factors affecting the frequency and severity of these effects. Here, the effects of radiation are discussed in general terms at different human postconception (PC) stage. Radiation effects can be classified as either deterministic or stochastic. Deterministic effect has a cause-and-effect relationship where the effect will not occur until a certain threshold is crossed. Upon crossing the threshold, the effect’s significance will increase linearly with dose. Stochastic effect occurs randomly and there is no threshold dose observed. In this section, the deterministic and stochastic effects of ionizing radiation on embryos and fetuses will be summarized [
National Council on Radiation Protection and Measurements. & National Council on Radiation Protection and Measurements. Scientific Committee 4-4 on the Risks of Ionizing Radiation to the Developing Embryo, F. Preconception and prenatal radiation exposure: health effects and protective guidance. (2013).
]. Table 1 provides an overview of the risk of effects with respect to the PC stage as well as the dose levels.
Table 1Risk of adverse effects at different PC stages and dose levels below or above the threshold.
Postconception (PC) days
Weeks
Dose (Gy)
Lethality
Malformation
Mental retardation
Growth retardation
Malignant disease
0 to 8 days PC (Preimplantation stage)
1
< 0.1
NCRPR-174: No increased risk of pregnancy loss at any stage of gestation
–
–
–
ICRP-84: 99.1 % probability that child will develop cancer
> 0.1
TG-36: Based on animal studies, 1 % − 2 % chance of early death after doses on the order of 0.1 Gy corresponding to a median lethal dose of about 1 Gy.
NCRPR-174: Insufficient evidence for preimplantation stages
–
–
–
–
8 to 56 days PC (Organogenesis stage)
2 to 8
< 0.1
–
ICRP-84: 97 % probability that child have no malformation
–
–
–
> 0.1
TG-36: Little risk of damage
NCRPR-174: The no-adverse-effect level for spontaneous abortion is estimated to be in the range of 0.25 to 0.5 Gy. The no-adverse-effect level for increased risk of embryonic death increases throughout gestation.
NCRPR-174: At all stages of organogenesis the risk of radiation-induced anatomical malformations with a dose of < 0.5 Gy is very low.
–
TG-36: High risk of damage
NCRPR-174: Better ability to recuperate from growth retardation than older fetuses
TG-36: Low risk of damage
56 to 105 days PC (Early fetal)
9 to 15
< 0.1
–
ICRP-84: 97 % probability that child have no malformation
–
–
–
> 0.1
–
–
–
TG-36: Significant risk of damage
NCRPR-174: No adverse effect between 0.25 Gy and 0.5 Gy.
TG-36: Low risk of damage
1
–
NCRPR-174: 40 % of fetuses were mentally retarded (A-bomb data)
–
105 to 175 days PC (Mid fetal)
16 to 25
> 0.1
–
–
–
TG-36: Low risk of damage
TG-36: Low risk of damage
1
–
–
NCRPR-174: 15 % of fetuses were mentally retarded (A-bomb data)
There is little to no evidence about the lethal effects of radiation on human during early stages of pregnancy. This is because of the high spontaneous abortion rate during that period, of which more than half arises from chromosomal abnormalities that occurred before conception during the development of the sperm or ova, or maternal disease states [
]. The lethality effect of radiation occurs as an “all or none” phenomenon where the pregnancy withstands the radiation exposure unharmed or is resorbed. The deductions of no-adverse-effect level of dose for lethal effects were drawn from in vitro and in vivo experiments conducted on animal cells and animals (rats and mice). During the first 14 days PC, the developing embryo is at the highest risk, where a dose of is sufficient to cause embryonic death in mice. There is no evidence that doses cause significant increase in embryo or fetal mortality. However, there is evidence that throughout pregnancy, the no-adverse-effect level for increased risk of embryonic death increases. This ranges from approximately to during organogenesis to in late gestation period [
RUSSELL LB, RUSSELL WL. Hazards to the embryo and fetus from ionizing radiation. in Proceedings of the International Conference on Peaceful Uses of Atomic Energy 175–178 (1956).
Congenital malformation occurs sporadically in all mammalian populations and its principal cause has been attributed to the incorrect interplay of genetic factors. It is widely accepted that the average incidence of malformation in live-born children throughout the world is 6 % [
]. Evidence has shown that high-dose radiation on rats during early organogenesis can produce anatomical congenital malformations of the brain, heart, craniofacial structures, viscera, and limbs [
RUSSELL LB, RUSSELL WL. Hazards to the embryo and fetus from ionizing radiation. in Proceedings of the International Conference on Peaceful Uses of Atomic Energy 175–178 (1956).
Utilization of developmental basic science principles in the evaluation of reproductive risks from pre- and postconception environmental radiation exposures.
Konermann G. Consequences of prenatal radiation exposure on perinatal and postnatal development. in Developmental Effects of Prenatal Irradiation 237–250 (1982).
]. Most data on the malformation rate as a function of dose and dose rate related to experimental animals as there are insufficient data dealing with acute high-dose radiation exposures during organogenesis in human pregnancies. Records from the Japanese atomic bomb survivors in 1945 did not provide adequate clinical information on infants that were exposed to the atomic bombs during organogenesis as many embryos may not have survived to term, or the offspring may have died during infancy before Radiation Effects Research Foundation (RERF) had developed accurate records in the late 1940s [
In summary, during the organogenesis period in early pregnancy, the risk of malformations is higher. The threshold for possible prenatal radiation effects for a fetus under 14 weeks PC, is approximately to . After 14 weeks PC, this threshold rises to at least to . At about 18 to 23 weeks PC, or late in the second trimester, the fetus is relatively resistant to the teratogenic effects of ionizing radiation [
Miller RW. Delayed radiation effects in atomic-bomb survivors. Major observations by the Atomic Bomb Casualty Commission are evaluated. Science (New York, N.Y.)166, 569–74 (1969).
]. Studies on the effects of acute irradiation in developing embryo or fetus were conducted on children who were exposed to radiation in utero from the atomic bombing of Hiroshima and Nagasaki [
]. It was found that there was no observation of mental retardation for exposure before the 8th week and after the 25th week PC. With a weighted uterine dose at , the incidence of mental retardation from the 8th week to the 15th week PC among the exposed pregnancies is 40 %. And from the 15th week to 25th week PC, mental retardation occurs in 15 % of the exposed pregnancies. Thus, the risk of mental retardation when exposed in utero is most prominent during 8th to 15th weeks PC. With a weighted uterine dose of less than , there is no observed effect on school performance at any stage of pregnancy. There is no evidence of an increased risk for microcephaly, mental retardation, lowered IQ, seizures, or impairment of neuromuscular performance in embryos and fetuses exposed to a dose of throughout pregnancy.
D. Growth Retardation
High doses to fetuses exposed to the atomic bomb detonation, as well as embryos or fetuses exposed to radiation therapy, have been associated to growth retardation [
Miller RW. Delayed radiation effects in atomic-bomb survivors. Major observations by the Atomic Bomb Casualty Commission are evaluated. Science (New York, N.Y.)166, 569–74 (1969).
]. However, there is insufficient evidence to definitively document human effects as a function of dose levels to the embryo or fetus. In a study by Rugh et al. [
], 18 groups of pregnant mice, each group on one particular day of pregnancy, were irradiated with to the whole body. The newborns were then examined at adulthood. The study found that the mice embryos irradiated in early pregnancy (0–4 days) did not suffer from growth reduction. Instead, growth reduction was observed in mouse embryos irradiated in mid-pregnancy (5–18 days). Data from the atomic bomb survivors from Otake and Schull studies are consistent with these findings [
]. However, the extent of human growth retardation following in utero irradiation cannot be determined until the irradiated population reaches maturity because the ability to recover from the growth retarding effects varies with the actual stage of exposure.
RERF investigators have spent more than two decades studying children and adults who were exposed in utero to radiation from Hiroshima and Nagasaki atomic bomb. Their publications [
] on the effects of in utero irradiation on growth primarily focused on small head size (microcephaly). Microcephaly is caused by radiation exposure, according to extensive analysis and evaluation in their publications [
], but these investigators can only speculate on whether radiation primarily affects skull growth or whether the growth of the skull is related to the effect of radiation on the size of the brain. With increasing weighted uterine dose, there is a statistically significant difference in the height of the 10- to 11-year-olds who were exposed in utero (P = 0.1). These publications contain no information on the linear growth no-adverse-effect level. In conclusion, the pre-implanted embryo and the embryo during early organogenesis have a greater ability to recover from growth retardation than older fetuses, who are likely to have a no-adverse-effect level for adult growth retardation with to dose exposure.
] investigated the association of hypertension, hypercholesterolemia, and cardiovascular diseases in adult populations exposed to radiation in utero and postnatally in children aged <10 years old among Hiroshima and Nagasaki atomic bomb survivors. Between 1978 and 2003, 506 adults were exposed to radiation in utero, and 1,053 children were followed with biennial clinical examinations. In the entire in utero exposed cohort or in trimester of-exposure subgroups, there were no statistically significant dose effects for any of these diseases. In the childhood exposure cohort, statistically significant dose effects for hypertension and cardiovascular disease were discovered. One reason for the negative results in the in utero radiation cohort is that 94 % of the in utero doses were less than (weighted uterine dose). Only 29 subjects in the in utero group received more than . Given the prevalence of hypertension, hypercholesterolemia, and cardiovascular diseases in adults, as well as the numerous known causes of these diseases in adults, it is unlikely that these data would have revealed a causal relationship between in utero irradiation and hypertension, hypercholesterolemia, and cardiovascular diseases.
] studied the relationships between in utero exposure and increment in thyroid disease among Hiroshima and Nagasaki atomic-bomb survivors. The average age of the 328 survivors who were irradiated in utero was 55.2 years. The subjects were examined between March 2000 and February 2003. The weighted uterine doses were between and , with a mean value of . A total of 229 people were given . 60 subjects were exposed in the first trimester, 73 in the second trimester, and 48 in the third trimester. The presence of thyroid nodules did not have a statistically significant dose–response relationship, according to the researchers (P = 0.22). Radiation exposure was not associated with the prevalence of cysts or autoimmune thyroid disease. Because of the small number of cases, it was not possible to assess the relationship between radiation exposure and malignant tumours or benign nodules.
G. Cancer (In utero exposure to occupational or environmental sources, and Japanese atomic-bomb survivors)
National Council on Radiation Protection and Measurements. & National Council on Radiation Protection and Measurements. Scientific Committee 4-4 on the Risks of Ionizing Radiation to the Developing Embryo, F. Preconception and prenatal radiation exposure: health effects and protective guidance. (2013).
], there were very few epidemiologic studies that found conclusive evidence of an increased risk of childhood leukemia, other childhood cancers, or adult cancers in offspring of nuclear or medical radiation workers’ mothers or fathers.
] have hardly revealed a link with natural background radiation exposure, but these epidemiologic studies lack the statistical power to detect small risks.
There was a statistically significant dose-related increase in the incidence of solid tumours in adulthood among Japanese atomic bomb survivors exposed to radiation in utero and those exposed in early childhood, though excess absolute rates (EARs) increased markedly with attained age among those exposed in early childhood but showed little change with time in the in utero cohort [
]. This disparity in EARs between the two cohorts suggests that the lifetime cancer risks associated with in utero exposure may be significantly lower than those associated with early childhood exposures. The leukemia mortality among atomic bomb survivors exposed in utero was based on only two cases in adults, both of whom were exposed to low doses, thus preventing attempts to compare the leukemia mortality risks between the in utero and early childhood cohorts [
Residents living near the radionuclide-contaminated Techa River who were exposed to radiation in utero and/or in early childhood before the age of 5 years experienced a non-statistically significant increase in solid cancers and a nearly statistically significant (P = 0.09) increase in leukemia deaths over a 49-year period [
]. There has been little evidence of an increase in childhood leukemia among the offspring who were in utero at the time of the accident, according to studies of nuclear reactor accidents [
]. Investigations of fallout from nuclear weapons testing or living near nuclear power plants yield little information about the risks of childhood leukemia or other cancers in the offspring of pregnant women [
]. Among the offspring of 1,494 mothers who lived in fallout-contaminated areas, a screening study to evaluate thyroid cancer and other thyroid diseases in offspring of women who were pregnant at the time of the Chernobyl nuclear reactor accident found seven thyroid neoplasms (a large, but not statistically significant excess) [
]. Radiotherapy plays a pivotal role in the management of the first three cancers. The impact of radiotherapy on the fetus will depend on the prescribed dose to the treatment site, treatment modality, distance to the fetus and the fetal or embryonic age. The risks to childhood development as a function of the fetal age have been reviewed in detail in the previous section, while the remaining three factors are closely related to the types of cancers. In general, from the perspective of the magnitude of fetal doses, cancer types can be divided into supradiaphragmatic or subdiaphragmatic. The latter will certainly result in a higher fetal dose.
A. Brain and head and neck tumours
Tumours in these sites are supradiaphragmatic and are located a distance away from the fundus to have an appreciable fetal dose in excess of [
Effect of tertiary multileaf collimator (MLC) on foetal dose during three-dimensional conformal radiation therapy (3DCRT) of a brain tumour during pregnancy.
]. Several measurement studies with anthropomorphic phantom have been conducted for brain tumours and re-affirmed this statement. For example, a prescribed dose to a grade 3 anaplastic astrocytoma with a 6 MV photon beam yields a resulting fetal dose of [
Effect of tertiary multileaf collimator (MLC) on foetal dose during three-dimensional conformal radiation therapy (3DCRT) of a brain tumour during pregnancy.
] reported fetal doses of and to the embryo’s head and legs respectively while treating a pregnant woman with grade 3 astrocytoma during third trimester using Cyberknife delivering to the tumour. Simulated phantom measurements showed that 3-d radiation therapy of brain tumours leads to fetal doses of less than at all gestational ages even without fetal shielding [
A similar magnitude of fetal doses is also reported for treatment of head and neck tumours in general when appropriate shielding is applied to the abdomen area. M. Podgorsak et al. [
] also reported a maximum dose of (with Cerrobend shielding) when treating head and neck squamous cell carcinoma with a dose of to the tongue. A more recent report80 on fetal doses when treating a pregnant woman with tonsil cancer using IMRT with a planned dose of to the primary tumour reported a maximum dose of (with lead apron).
Effect of tertiary multileaf collimator (MLC) on foetal dose during three-dimensional conformal radiation therapy (3DCRT) of a brain tumour during pregnancy.
] show that radiotherapy to brain and head and neck tumours typically does not result in a fetal dose exceeding which serves as the threshold for deterministic effect of radiation to the fetus.
B. Breast cancer
Most case studies have shown that in breast radiotherapy, it is plausible for pregnancy to be carried to full term. Early works by S. Ngu et al. [
] reported a dose of for a third trimester fetus and a dose of for a first trimester fetus respectively when proper shielding were implemented. It was also estimated that a prescribed dose to chest wall will only deliver about to the fetus [
]. However, studies have also shown that the fetus could receive doses exceeding the deterministic threshold of close to at an advanced gestational age as the fetus lies closer to the treatment field [
]. In general, the prognosis associated with breast cancer in a pregnant woman is not poorer than that in a non-pregnant woman, and patients are encouraged to continue their pregnancy to term [
]. In most cases involving surgery and chemotherapy, radiotherapy can take place after giving birth. An exception arises when chemotherapy is not indicated and it is impossible to wait until after delivery due to the time duration after surgery.
C. Hodgkin’s and non-Hodgkin’s Lymphomas
Radiotherapy can typically be delivered with minimal risk to the fetus if the lymphomas are presented in the supradiaphragmatic form. S. Woo et al. [
] has reported on the treatment of early-stage Nodular Sclerosing Hodgkin’s lymphoma in 16 pregnant women with treatment field to mediastinal, axillar and cervical lymph node levels who receive doses between . With lead shielding in the uterus area, the estimated fetal doses lie within with 6 MV photons. A case study was also reported by F. Peccatori et al. [
] whereby a was delivered to the right occipital region with 6 MeV electrons and a maximum dose of was estimated to be delivered to the third trimester fetus. Case studies dealing with mantle-field irradiation for Hodgkin’s disease during pregnancy resulted in fetal doses at the top of the fetus of [
Conceptus dose from involved-field radiotherapy for Hodgkin’s lymphoma on a linear accelerator equipped with MLCsDie Strahlendosis im Fetus von Involved-Field-Radiotherapie wegen Hodgkin-Lymphom an einem Linearbeschleuniger mit Multileafkollimatoren (MLC).
] measured the conceptus dose from radiotherapy for Hodgkin’s lymphoma with 6 MV photons. They found that cervical node irradiation results in fetal doses below during the entire pregnancy. The corresponding dose due to radiotherapy in the regions of axilla, mediastinum and neck-mediastinum may exceed the threshold value of depending upon the stage of gestation. Decisions about the use of fetal shielding and/or pregnancy termination should be taken. These days, there have been lesser indications of radiotherapy for treatment of Hodgkin’s and non-Hodgkin’s lymphomas. If indicated, the volumes and doses are also decreasing, which would presumably induce a smaller impact on fetal dose.
D. Gynaecological tumours
Cervical cancers are the most diagnosed gynaecological cancers during pregnancy with incidence ranging from 1 in 10,000 to 4 in 100,000 pregnancies [
]. Pelvic irradiation is the standard treatment for locally advanced cervical cancer and similar to the subdiaphragmatic form of lymphomas, is contraindicated during pregnancy due to the high risk of abortion and fetal damage [
] to preserve the fetus and not compromise the prognosis.
Factors affecting fetal dose in radiotherapy
A. Megavoltage X-rays
OOF doses can be contributed from photon leakage through the treatment head of the machine, radiation scattered from the collimators and beam modifiers, and radiation scattered within the patient from treatment beams [
], as illustrated in Fig. 2. In regions that are close to the treatment field, dominant source of OOF scatter originates from patient scatter. For distances >20 cm, head leakage becomes the dominant source of OOF dose [
Studies have shown that leakage through treatment head of machine varies for different manufacturers. This is mainly due to the orientation of the waveguide, shielding design of the machine and nature of the beam collimation system. For example, Lonski et al. [
] found that within the treatment delivery head, the horizontal waveguide produced the most amount of leakage. It has also been observed that Varian 600C (Varian Medical System, Palo Alto, California), a single-energy mode linear accelerator (Linac), had the least leakage at 0.15 % while Siemens Primus (Siemens Medical Systems, Concord, California) produced the most leakage at five times that of Varian 600C.
Apart from that, head leakage and collimator scatter have been studied by Fraass and van de Geijn [
] and it was found that these two factors together gave rise to OOF dose with the same order of magnitude as patient scatter. Leakage from different machines was also seen to vary up to a factor of 2, which agrees with the findings of Lonski et al. [
] revealed that collimator scatter contributed 20 to 40 % of the OOF radiation dose depending on machine, field size and distance from the field. At distances beyond 60 cm from the central axis, leakage from treatment head was found to be the main contributor of OOF doses. Kourinou et al. [
] provided data about the fetal dose components for radiotherapy of nasopharyngeal cancer, breast carcinoma, Hodgkin’s disease and lung cancer. They reported that the contribution of head leakage and collimator scatter to the total fetal dose varies by 52–80 % by the treatment site in early pregnancy.
Furthermore, addition of physical wedge would increase the OOF dose near the beam by a factor of 2 to 4 [
]. In the case of a universal wedge which is located inside the gantry head, the OOF doses within 30 cm from the field borders are not elevated compared to open portals [
]. However, with the use of enhanced dynamic wedge (EDW), wedging effect can be produced without the addition of a physical wedge. This is clearly manifested in the study done by Gopalakrishnan et al. [
], where a decrease in scattered dose to structures outside of treatment field was achieved while retaining the same conformity index which is used for evaluating plan quality. Hence, the use of EDW would be a better choice than a physical wedge as it results in lower OOF doses.
], OOF doses were measured while varying beam energies from 4 MV to 25 MV, field size from 5 × 5 cm to 25 × 25 cm, and depth from 2 cm to 15 cm. In general, the most important factor affecting OOF doses is found to be the distance from the field edge, which exhibited a nearly exponential relation. Changes in depth showed a minimal effect on OOF doses, while comparison across different energies with the same field size and depth revealed that OOF doses were qualitatively similar and of the same order of magnitude. In other studies, changes in OOF dose due to distance from field edge were compared with scattering from within the patient. It was found that scattering contributes more significantly to OOF doses when measured closer to the beam edge [
For higher energy photon beams (>10 MV), photonuclear reaction occurs when the beams hit the target, flattening filter and collimation system found within the Linac head [
]. Neutrons are produced as the high energy photon beams interact with the materials with high atomic number (Z values) in the Linac head. In the study done by Howell et al. [
], photon beam energy was varied for different Linac to understand the production of neutron. It was found that total neutron fluence increased with photon energy, with Varian Linac having the highest neutron production. This is due to the higher primary electron energy used in the machine, producing a higher peak energy of photons that are responsible for neutron productions [
Nuclear reactions would also take place when photon beams interact with treatment room wall or patient body. However, they can be considered insignificant when compared with neutron production in the head of the Linac [
Neutron contamination from medical electron accelerators: recommendations of the National Council on Radiation Protection and Measurements. (The Council, 1984).
] asserted that the low-Z elements constituting human tissues caused the neutron production to be negligible. In addition, neutron productions are also dependent on the number of MUs delivered which can vary between treatment plans [
Despite the numerous factors mentioned, the difference in design of Linacs and the energy of the beams play the most critical role in the photoneutron production when high-energy photons are used [
], different Linacs from various manufacturers were measured for their neutron source strength values (Q values). Upper jaws in Linacs from Elekta were replaced with multi-leaf collimator (MLC), those from Siemens had their lower jaw replaced by MLC, whereas Varian Linac added an MLC while retaining both their jaws. Measurements using gold foil activation in Bonner spheres were done with MLCs and jaws closed for all Linacs. Varian recorded a Q value of (1.02–2.46) × 1012 while Elekta and Siemens had a Q value of (2.46–5.86) × 1011. The results obtained from different Linacs, which showed an increasing neutron dose with increasing treatment energy, agree with each other.
B. Proton therapy
The use of proton therapy for treatment can potentially achieve a higher therapeutic ratio due to the presence of a Bragg peak, allowing the sparing of critical organs and tissues [
]. However, the use of proton therapy beams produce neutrons, secondary protons, light charged ions, recoil heavy ions, and photons during treatment, resulting in secondary dose equivalent [
], the main contribution to the overall OOF dose is due to neutrons in treatment with protons, which originate from nuclear interactions within components along the beam line and within patients themselves. The contribution of the secondary photons is limited to 10 % of that associated with neutrons. Two commonly used proton beam delivery methods are the passive scattering and spot scanning techniques. Passive scattering technique uses scatterers, beam-flattening devices, collimators/apertures and compensators at the beam delivery nozzle to obtain a homogenous and flat dose in the target. On the other hand, spot scanning technique uses a pencil beam to scan across the target volume with magnetic dipoles, without the need for other beam components at the beam delivery nozzle [
] found that the scatter dose equivalent per proton treatment was between 33 and 80 mSv/Gy. The 200 MeV proton beam line using passive scattering technique with two scattering foils and three collimators produced too much scatter to the OOF dose that additional shielding was required before treatment could begin. This scattering was also found to decrease with increasing distance from the final collimator closest to the patient.
], the neutron dose resulting from spot scanning and passive scattering technique irradiation were evaluated and it was found that the former had a dose reduction of at least 10 times compared to the latter. The maximum scattered neutron dose in healthy tissue from a 177 MeV proton beam was measured to be 0.39 mSv/Gy for a small target to 4.09 mSv/Gy for a big target. These values are similar in magnitude to that from an abdominal CT scan that has a dose of 4 mSv/Gy [
], neutron dose equivalent per proton treatment dose was measured at different lateral distance from the centre of the proton beam with the neutron dosimeter placed at fetus representative position. Using a Rando phantom (Phantom Laboratory Inc., Salem, NY) with three different points of measurements, it was found that a 200 MeV beam energy produced 0.26 mSv/Gy of neutron dose equivalent at a distance of 18.9 cm. This reduced to 0.10 mSv/Gy at a distance of 54.5 cm from the centre of the proton beam.
Regardless of proton delivery technique, dose due to internal and external neutron decreases with distance due to attenuation through elastic collision. In passive scattering beam made up of multiple collimation systems, neutron dose is generally higher as field collimators are made up of materials with high Z-value and are generally closer to the patient.
C. Imaging dose from image-guided radiotherapy
The use of IGRT has become almost essential in modern radiotherapy due to the increased use of highly conformal treatment techniques (e.g. Volumetric Modulated Arc Therapy or VMAT) as it offers the ability to provide precise dose delivery to the target to improve local control and allow sparing of dose to healthy tissues [
]. Using on-board kilovoltage cone beam computed tomography (CBCT) systems on the Linac, daily image registration can be done to better align the patient. However, the imaging dose from CBCT cannot be neglected, as Perks et al. [
], where thermoluminescent dosimeters (TLDs) were placed at different intervals along the superior-posterior axis on the surface and the central plane within the phantom. A dose of and were obtained for a distance of 7 cm and 12 cm from imaging field edge, respectively.
] has found that daily CBCT could deliver a considerable amount of dose to critical organ. For instance, an abdomen scan can cause the small intestine and rectum to receive an additional dose of to in 35 fractions. Also in the same study, the use of low-dose mode for CBCT was claimed to be beneficial to the patient, based on their finding that the effective dose from low-dose CBCT was approximately-one fifth of those from standard mode CBCT.
It is therefore important to weight the advantages and disadvantages of using IGRT on pregnant patients. CBCT may be considered when the edge of the imaging field is at a distance away from the fetus. As for treatment that require CBCT around the fetus, the low-dose mode CBCT should be recommended despite the decrease in image quality.
Fetal dose estimation
Estimation of fetal dose plays a very crucial role in the planning of radiotherapy for pregnant patients. This is because the occurrence and severity of adverse effects on fetus depend strongly on the amount of fetal dose received, as discussed earlier in Section II. If the fetal dose exceeds the threshold of , countermeasures such as shielding and modifications of treatment techniques ought to be taken, which will be elaborated in Section VI. In this section, various methods of dose estimation will be covered, following closely the chronological development.
A. X-rays and proton therapy
a. Measurements using Physical Phantom
Anthropomorphic phantom has been very widely used along with TLDs or ionisation chambers to estimate fetal dose since decades ago. For instance, the Rando phantom (Alderson Research Labs, Stanford, CA) was used for fetal dose estimation during breast cancer [
], the phantom is made up of numbered slices of 2.5 cm thick each. Different slices were replaced by Lucite slices to simulate the fetus position at different gestational age. To measure the dose absorbed by various parts of the fetus, the TLDs are inserted into the prefabricated holes at different depths of the slice, corresponding to different points of interest. The types of TLDs typically used are lithium fluoride (LiF) and calcium fluoride (CaF2) TLDs. For measurement with ionisation chambers, the slices could be replaced with PMMA slabs [
] to accommodate large-volume ionisation chambers. In a recent study, optically stimulated luminescence detectors (OSLDs) were used to measure fetal dose arising from an innovative technique called Virtual Tangential-fields Arc Therapy (ViTAT) [
As the measurement using physical phantom could be onerous and provides less flexibility in terms of optimizing irradiation conditions and varying patient features [
], computational phantoms have come into play. Computational human phantoms are models that represent organs or tissues of the body, or the whole body mathematically [
Stylized phantom is the earliest type of computational phantom being used. It is made up of geometrical shapes depicting different structures, which can be described by simple surface equations [
]. Despite its simplicity, it was found to be lacking in its ability to reflect realistic anatomical geometries, hence giving rise to the development of voxelized phantom. Voxelized phantom, as the name itself suggests, consists of a large number of voxels, each corresponding to a particular tissue or organ type. It is worth pointing out that this class of phantom was developed based on segmented high-resolution medical images, thus it was only made possible following the advancement of imaging modalities such as computed tomography (CT) and magnetic resonance imaging (MRI) [
That being said, voxelized phantom turned out to be less accommodating to variability of patient parameters including but not limited to body size or organ position [
], leading to the design of a BREP phantom, which is sometimes also known as a hybrid phantom because it combines the desirable features of stylized and voxelized phantoms [
] have constructed three hybrid fetal phantoms for gestational age of 20, 31 and 35 weeks based on magnetic resonance images. These phantoms are patient-specific, realistic and allow for accurate fetal dose estimation. Besides, Paulbeck et al. [
] has also come up with a new hybrid phantom series called the J45 series, which was modified from the existing ones based on the Japanese atomic survivor populations in 1945. The pregnant phantom series has been used alongside with Monte Carlo in a study to evaluate the fetal dose due to 3D conformal radiotherapy for breast cancer [
In addition to these three main categories, the Department of Nuclear Energy, Federal University of Pernambuco, Brazil has developed a type of phantom known as the mesh phantom by modifying the vertices, edges and faces of polygonal meshes [
Commercial treatment planning system (TPS) is the most common dose calculation platform in clinical use currently. Nevertheless, it is known and acknowledged by the vendor that TPS is only accurate in the treatment field itself and faces limitation in calculating OOF doses accurately [
Accuracy and sources of error of out-of field dose calculations by a commercial treatment planning system for intensity-modulated radiation therapy treatments.
] have been found to underestimate the OOF doses when compared to either phantom measurement or Monte Carlo simulation. In a study done by Ogretici et al. [
] that investigated two different algorithms (anisotropic analytical algorithm and Acuros XB algorithm) of Eclipse v13.0.33 (Varian Medical Systems, Palo Alto, CA, USA), the TPS algorithms failed to accurately account for OOF doses for area further than 13 cm away from the field edges. Consequently, the fetal dose was given as while phantom measurement estimated a mean dose of . This finding questions the applicability of TPS to fetal dose estimation.
Before the advent of Monte Carlo simulation, a few calculation methods have been introduced to quantify OOF doses. One example is the Clarkson’s algorithm which is based on the scatter maximum ratios and measured primary beam dose profiles [
]. It comprises of two major parts: the construction of patient phantom and the dose calculation. Both of these methods have shown reasonably good agreements with phantom measurements using TLD.
Currently, a computational phantom, which has been discussed in the previous section, is often coupled with Monte Carlo radiation transport code for dose estimation purpose. One of the most frequently used codes is the Monte Carlo N-Particle transport code (MCNP) which was developed by Los Alamos National Laboratory and comes in different versions due to constant modifications [
Lecture Notes in PhysicsMonte-Carlo Methods and Applications in Neutronics, Photonics and Statistical Physics. Springer-Verlag,
Berlin/Heidelberg1985: 33-55
]. It has been used to simulate a 6 MV photon beam for a number of studies to evaluate the shielding effect during the radiotherapy for Hodgkin’s, breast cancer [
A feasibility study to calculate unshielded fetal doses to pregnant patients in 6-MV photon treatments using Monte Carlo methods and anatomically realistic phantoms.
] made use of another Monte Carlo simulation method called the Particle and Heavy Ion Transport code System (PHITS) to evaluate the fetal dose resulting from 3D conformal radiotherapy for a breast cancer patient. Overall, the development of accurate dose calculation algorithms and computational phantom has given the medical physicist an alternate method to decide on the best treatment modality to achieve minimal dose to the fetus.
B. Image-Guided radiotherapy
IGRT ensures that minimum dose is delivered to the organs at risk (OARs) by improving the patient set-up accuracy. Despite this advantage, it could result in extra imaging dose to the patient which is not only limited to the imaging area but also to OARs nearby. To the best of our knowledge, fetal dose due to daily imaging has not been investigated specifically, but there are a number of studies which looked into the imaging dose both inside and outside the field area for non-pregnant patients.
Similar to the dose estimation methods discussed previously, the imaging dose from IGRT has been studied using physical phantoms as well as computational phantoms coupled with Monte Carlo simulation. Yet, it is worth noting that water phantom, which has rarely been used for evaluation of OOF doses from radiotherapy, appeared more frequently in the literature for imaging dose estimation. Using two cylindrical water phantoms with 30 cm and 16 cm diameter (to simulate an average body and an average head, respectively) and a Farmer chamber, Islam et al. [
] investigated the imaging dose arising from gantry-mounted kV-CBCT systems. The former performed measurements on Elekta’s (Elekta Oncology Systems, Norcross, GA) X-ray Volumetric Imaging (XVI) system while the latter studied an additional system of Varian (Varian Medical Systems, Palo Alto, California) On-Board Imager (OBI).
While the Alderson Rando phantom has been a rather popular tool for simulating a human body, various other anthropomorphic phantoms have been constructed for the purpose of estimating imaging dose during IGRT. One example is the phantom known as RPHAT [
], which is sliced in the coronal plane and thus allows the simulation of different thicknesses of patients’ tissue. This was reported in a study by Perks et al. [
], on the other hand, measured the MVCT imaging dose during helical tomotherapy using an anthropomorphic male ATOM phantom (CIRS Inc, Norfolk, VA).
Apart from the more conventional measurement technique, more advanced methods such as Monte Carlo simulation has also been used to estimate the dose to radiosensitive organs resulting from full- and half-fan kV-CBCT scans on a Varian OBI system [
]. The simulation was done using the Vanderbilt-Monte-Carlo-Beam-Calibration (VMCBC; Vanderbilt University, Nashville, TN) algorithm, where the X-rays were generated with the BEAMnrc code and the dose was calculated with the DOSXYZnrc code after creating the CT-based patient phantoms from the CBCT images.
Judging from the diversity of imaging dose estimation methods that have been introduced to date, it is clear that imaging dose is not negligible. This is more so in the case of pregnant patients, as the fetuses’ well-being is highly dependent on the amount of OOF dose received, hence the need for a systematic study of fetal dose following repeated imaging for IGRT.
Fetal dose shielding and dose reduction technique
In radiation therapy for the pregnant patients, it is impossible to eliminate all radiation to the fetus. Therefore, there is a need to understand all the sources of secondary radiation that will contribute to fetal dose in order to provide adequate protection for the fetus. As mentioned earlier in Section IV(A), the principal sources of OOF dose include photon leakage through the Linac gantry head, external radiation scatter from collimators and beam modifiers, and internal radiation scatter within the patient’s body [
]. Thus, careful planning is essential during radiation treatment for the pregnant patient by using appropriate shielding techniques or other dose reduction techniques (e.g. for IMRT and VMAT).
] classified the design of the shields into three types, namely Bridge over patient, Table over treatment couch and Mobile Shields, as shown in Fig. 3. The design of shields is required to allow treatment for Anterior-Posterior Posterior Anterior (AP-PA) and lateral radiation fields situated above the diaphragm and lower extremities. Irradiation using oblique fields might be limited by the design of the shield, where weight is the primary concern. Consequently, safety is paramount in the selection and design of the shields to eliminate the possibility of injury to the patient or radiation therapists. As recommended by the AAPM TG-36, a block of lead is sufficient to reduce the photon radiation dose to the fetus to less than which poses little risk of radiation damage to the fetus. The following section reviews the type of shielding (e.g. design, size, material) that had been used for certain tumour sites as well as for pre-treatment CT imaging. The dose reduction reported is subject to the point of measurement (i.e. distance from irradiation field), type of measurement (e.g. Monte Carlo simulation, experimental measurement on anthropomorphic/plastic water phantoms), type of detector (e.g. ion chamber, film), treatment modalities (e.g. IMRT, 3DCRT) and gestational age. In general, these data show that fetal dose may be effectively reduced to below for treatment of tumours located superiorly to the fetus by using a thick lead1.It is also clear that the increase in gestational age and field size causes the fetal dose to increase. Monte Carlo simulation studies have also shown that fetal dose increases with increasing stage of gestation and this is related to the decrease in distance between the irradiation field edge and the fetus [
A feasibility study to calculate unshielded fetal doses to pregnant patients in 6-MV photon treatments using Monte Carlo methods and anatomically realistic phantoms.
]. In addition, factors such as the design of the shielding (i.e. size, weight, material), the position of the shielding during irradiation, and the inclusive weight of both the patient and shielding may not exceed the weight limit of the couch. These are the main factors which the clinicians and medical physicists should consider when planning a treatment for pregnant women.
Fig. 3Types of Shielding Devices (A) Bridge over patient (B) Table over treatment couch (C) Mobile Shield.
There are many dosimetric studies on shielding devices using Monte Carlo simulation and measurement using anthropomorphic phantoms. In head and neck tumours, Moeckli et al. [
] achieved a dose reduction of 70 %, using a lead shield of positioned over the abdomen and lateral region of the patient (Bridge over patient), with a lead sheet from the bridge coming down to the patient’s skin. Subsequent work by Josipovic et al. [
] demonstrated that a thick Cerrobend () could reduce the fetal dose by a factor of 2.8–4.1, leading to less than . This shield extended caudally from the field border was placed on the block holder of the linear accelerator enabling fetal protection during IMRT. Han et al. [
] found that common head and neck and brain treatment plans of less than will result in less than to the fetus. A lead or Cerrobend (Bridge over patient) was effective enough in reducing the fetal dose by a factor of 3 and hence deemed the recommendation by AAPM TG-36 of 5 cm lead to achieve fetal dose to less than conservative. Costa et al. [
] demonstrated that by using a Cerrobend (Bridge over patient) for a beam photon treatment plan, a 91 % reduction of fetal dose to as low as was possible. Horowitz et al. [
] was able to reduce the fetal dose to below by using only lead shielding (similar to Table over couch). Additionally, a custom made U-shaped lead shield with a detachable posterior shield was designed and experimented by Owrangi et al. [
] placed a 1-cm-thick lead shield having the shape of a tunnel over phantom’s abdomen. This shield together with a movable lead wall reduced fetal dose up to 51 % during cerebral radiotherapy. They varied the planning techniques, field sizes and distance from field edge and a comprehensive peripheral dose dataset was presented with and without shielding to aid clinicians in deciding treatment techniques. Based on these studies, it can be seen that treatment for brain tumours generally contributes little dose to the fetus and with a lead thickness of , it should be sufficient to reduce the fetal dose significantly. Similarly it is possible for head and neck tumours.
ii. Breast Cancer
For breast cancer patients, most works have shown that an appropriate implementation of shielding was able to reduce fetal dose to below . A study done by Atarod et al. [
] revealed that a thick Cerrobend (Bridge over patient)was sufficient to reduce the dose by from to when treating breast tumours with Co-60 beam. Likewise, Filipov et al. [
] managed to reduce the fetal dose to a level within the tolerable limits as specified in ICRP 84 and AAPM TG-36, by using a thick lead brick () on the lateral side of the abdomen region and thick lead slabs () above and below the abdomen region. Ogretici et al. [
] found that a lead apron of frontal thickness and rear thickness could reduce the fetal dose by from to , thus suggesting it to be a potential alternative for any centres that lack special shielding equipment. Additionally, Mazonakis et al. [
] ( and ) had used a similar shielding region where lead is placed on the anterior and the lateral region of the abdomen. Both studies had managed to reduce the fetal dose significantly. Instead of using external beam therapy, Candela-Juan et al. [
] performed a study on breast irradiation using 192Ir Brachytherapy, and demonstrated that with proper shielding ( between the breast and the table, two lead bricks of thick on the lateral sides of the breast and lead above the breast), it was possible to reduce the peripheral dose from at from caudal edge of breast to less than at . Ultimately, a thick lead is shown to be able to reduce the fetal dose significantly, to below the recommended fetal dose value ().
iii. Hodgkin’s and non-Hodgkin’s Lymphomas
Mantle field irradiation is usually used for Hodgkin’s and non-Hodgkin’s lymphomas treatment. Nuyttens et al. [
] estimated the fetal dose before treatment using an anthropomorphic phantom. Two different shielding designs were used in this study for two different patients. One shielding consisted of 6 Cerrobend blocks of on the lateral side of the patient with 8 lead aprons of over the patient. On the other hand, a Bridge over patient design was employed for the second patient. Both shielding designs made use of similar materials and were found to reduce the fetal dose by . The same designs were adopted for the actual treatment and the findings suggested that malformations or growth and mental retardation were not detected after mid-fetus period when fetal doses were below . In addition, Mazonakis et al. [
] presented results of embryo dose using anthropomorphic phantoms at the first trimester of gestation. They suggested that a lead Bridge over patient setup could reduce embryo dose to below and shielding would be required if the irradiation field is in the neck-mediastinum and mantle region. Shielding is instead optional if the irradiation field is in the neck or axilla region but it is recommended whenever it is available, so as to minimize unforeseen delayed radiation effects. Monte Carlo simulations revealed that the change in fetal dose calculated with a 2-cm-thick lead shield over phantom’s abdomen and that computed with a 5-cm-thick shield was 23.4 % [
]. However, using a lead thickness does not show further significant fetal dose reduction (less than ). These works again indicate that a thick lead is sufficient to bring the fetal dose down to the recommended level of . The decreasing indications, volumes and doses of radiotherapy for treatment of Hodgkin’s and non-Hodgkin’s lymphomas nowadays is expected to entail an even lower fetal dose.
iv. Pre-Treatment CT imaging
In pre-treatment CT imaging, fetal dose reduction using simple shielding such as lead apron or bismuth were studied. Chatterson et al. [
] suggested that adjustments to the scanning protocols, such as reducing the voltage and limiting the scan regions, are more effective than using shields. In the third trimester, reducing the voltage from to reduced the fetal dose to without shielding. Similar study had been done by Moore et al. [
] where they recommended to use despite recording lowest fetal dose when was used. A bismuth shield was used in this study. Subsequently, a literature review by Ghaznavi [
] discussed various approaches to reduce fetal dose due to CT scan, such as tube current modulation, scan parameters, external shielding, internal shielding and reconstruction algorithms. The review advised to minimize fetal dose with appropriate approaches while preserving diagnostic image quality. Although low kVp was shown to reduce fetal dose, redundant reduction of kVp and mAs is not acceptable due to the increase in image noise. From these studies, despite the low fetal dose measured, shielding is always recommended whenever possible.
B. Dose reduction technique
In the review presented above, one of the trends across the case studies was that they were based on energy. This is due to the generation of neutrons when energies higher than were used, hence high-energy beams were not advisable for clinical treatment. This is especially a concern in proton therapy where the major contribution to fetal dose comes from neutrons [
] suggested to include borated polyethylene as a shielding material to reduce peripheral neutron doses. Similarly, Co-60 beams are not recommended for treatment due to the increase radiation leakage from the unit, leading to higher OOF doses [
], 3DCRT planning technique could be a better option in reducing fetal dose as compared to IMRT and this is due to the generally higher monitor units (MUs) used which resulted in an increase in fetal dose when using intensity-modulated techniques. Modern techniques such as IMRT and VMAT are generally not recommended, according to the authors, due to a lack of clinical data. In spite of this, Josipovic et al. [
] had however suggested a shielding design which is a Cerrobend block attached to the gantry head to reduce collimator scatter contribution to fetal dose during IMRT. Collimator scatter contributes the highest dose to the OOF dose and a significant reduction of OOF/peripheral dose was obtained at 40 cm away from isocentre of a pregnant patient with rhinopharynx cancer treated with IMRT [
]. Special care should be given by the radiation oncologists in the selection of the minimum margins in the definition of the target volumes during treatment planning process.
Studies have also shown that the use of wedge filters or Cerrobend blocks may increase the peripheral doses due to extra scattered radiation by beam modifiers [
Fetal dose management is an effort that requires broad knowledge ranging from the biological effects of radiation on the fetus, shielding design, measurement and the physical set-up to be implemented for each individual clinical case. Meanwhile the fetus will grow throughout the radiotherapy treatment and constant monitoring is important. This review paper has approached the topic from various perspectives to provide a comprehensive understanding of the possible issues and solutions that might be useful while deciding to treat or implement radiotherapy for pregnant patients. The constant advancement of radiotherapy definitely will require periodic revision of our current knowledge and approaches. However our main priority remains the same, if radiotherapy is indicated and unavoidable for pregnant patients – to maximise treatment for our pregnant patients and to provide the best protection to the fetus.
Funding
Hong Qi Tan is supported by the Duke-NUS Oncology Academic Program Goh Foundation Proton Research Programme (08/FY2021/EX(SL)/92-A146)
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
References
Mazonakis M.
Damilakis J.
Estimation and reduction of the radiation dose to the fetus from external-beam radiotherapy.
National Council on Radiation Protection and Measurements. & National Council on Radiation Protection and Measurements. Scientific Committee 4-4 on the Risks of Ionizing Radiation to the Developing Embryo, F. Preconception and prenatal radiation exposure: health effects and protective guidance. (2013).
RUSSELL LB, RUSSELL WL. Hazards to the embryo and fetus from ionizing radiation. in Proceedings of the International Conference on Peaceful Uses of Atomic Energy 175–178 (1956).
Utilization of developmental basic science principles in the evaluation of reproductive risks from pre- and postconception environmental radiation exposures.
Konermann G. Consequences of prenatal radiation exposure on perinatal and postnatal development. in Developmental Effects of Prenatal Irradiation 237–250 (1982).
Miller RW. Delayed radiation effects in atomic-bomb survivors. Major observations by the Atomic Bomb Casualty Commission are evaluated. Science (New York, N.Y.)166, 569–74 (1969).
02709-0&id=gr1.jpg){kind=link}
02709-0&id=gr2.jpg){kind=link}
02709-0&id=gr3.jpg){kind=link}