| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Jefferson Medical College and Alfred duPont Hospital for Children
Wilmington, DE 19899
New Mexico Federal Regional Medical Center University of New Mexico
Albuquerque, NM 87108
We were surprised by the article published by El-Khoury et al. [1] in the August 2003 issue of the AJR. The most important requirements for a physician or counselor who advises reproductive-aged women about the risks of radiation before or during pregnancy are a scholarly and complete awareness of all the literature dealing with the risks of intrauterine radiation exposure, compassion and sensitivity to the emotionality of the counseling situation, and avoiding the unnecessary promotion of anxiety.
The ultimate message that should go to pregnant women is that diagnostic radiation doses (including radiation with CT) are associated with an extremely small risk of potential cancer, but no evidence has been found of any measurable increase in mental retardation, congenital malformations, or microcephaly.
Unfortunately, the article by El-Khoury et al. [1] falls short. The consent form to be presented to the patient does not convey the current scientific understanding of in utero radiation effects and does not even accurately reflect what is contained in the text of the article. In particular, the consent form for fetal doses in the range of 0.01–0.05 Gy (1–5 rad) published by the authors indicates that risk increases for congenital malformations, mental retardation, and miscarriage (spontaneous abortion) from exposures of 0.01–0.05 Gy and greater than 0.05 Gy of low linear energy transfer (LET) ionizing radiation. Even in the penultimate sentence of the first paragraph of their own discussion, the authors are inconsistent by stating that "Somatic effects such as body size and mental retardation appear to have dose thresholds in the range of 50–100 mSv (5–10 rem)."
Actually, congenital malformations, mental retardation, microcephaly, and miscarriage are threshold phenomena for which risk does not increase until the fetal dose reaches 0.01–0.20 Gy or more. In 1977, when Handbook 54 [2] was published by the National Council on Radiation Protection and Measurements (NCRP), the 0.05-Gy value was selected for these reasons: below that, no increased risks appeared for deterministic effects; a 0.05-Gy absorbed fetal dose was well below the dose that would produce the deterministic effects just mentioned; and 99% of radiology exposures during pregnancy resulted in fetal absorbed doses less than 0.05 Gy.
For some reason, the authors of the "policy for a new era" attach more significance to the 0.05-Gy value than is warranted. The 1997 NCRP annual meeting dealt with the reproductive effects of all types of radiation, and all the presentations were published in the April 1999 issue of Teratology. Four articles dealt with the reproductive effects of in utero ionizing radiation [3–6]. El-Khoury et al. [1] did not mention the current recommendations of the International Commission on Radiological Protection [7], which deals with counseling pregnant women at risk for radiation exposure.
An article by Brent [3] summarizes the stochastic and deterministic risks of ionizing radiation. The results are based on thousands of exposures and hundreds of experiments in animals as well as human epidemiology studies. The thresholds for congenital malformations, growth retardation, pregnancy loss, and neurobehavioral effects are all greater than 0.20 Gy [3]. Growth retardation and embryonic death exhibit even higher thresholds, which vary with the stage of embryonic development. For instance, the lethal dose for the fetus in the latter stages of pregnancy is similar to that of the mother.
Mental retardation has been the most controversial deterministic effect, primarily because of the article by Otake and Schull in 1984 [8]. The authors found no threshold for mental retardation, but the risk for mental retardation doubled with exposures of 0.02 Gy. Their conclusion was criticized by their colleagues. To their credit, Schull and Otake [4] reanalyzed their data and performed further research. In 1999, they concluded that the threshold for mental retardation was approximately 0.18 Gy for 8–15 weeks of gestation, and although they do not provide an estimate, it appears that the threshold for mental retardation during the latter part of pregnancy is approximately 0.5 Gy, because the lower end of the 95% confidence interval was stated to be 0.28 Gy. Miller [6] analyzed the risk of mental retardation in one of the articles published in 1999. Miller and Schull spent a good part of their professional career involved with this issue. Miller presented risk levels in a simple table indicating that the threshold is greater than 0.5 Gy, which is much higher than exposures received during diagnostic radiology procedures.
On the basis of biologic plausibility and the results of animal studies, it would appear that data support the viewpoint that mental retardation is a deterministic effect with a threshold above 0.2 Sv.
Basic science information [3] indicates why 1 rad (0.01 Gy) would not double the incidence of mental retardation. Teratogenesis, which involves the perturbation of many cells, is a threshold phenomenon. It does not result from injury to one cell, as happens with mutagenesis and carcinogenesis. Studies of in utero exposure to ionizing radiation indicate approximately 30 IQ points may occur per gray during the most sensitive period of human brain development, indicating that severe mental retardation would not occur even if no threshold existed because a linear relationship to exposure would predict a 0.3 IQ point loss at 0.01 Gy. At 0.01 Gy, no observable pathologic effects in the developing brain could account for severe central nervous system effects. Neurobehavioral evaluations of animals exposed in utero found a threshold for behavioral effects at the same dose (0.2 Gy) as for other teratologic effects. Pathologic examinations of irradiated brains exposed to 0.02 Gy reveal no pathologic consequences that could account for severe mental retardation. That would mean that the pattern of effects produced by ionizing radiation that accounts for mental retardation when the fetus is exposed to absorbed doses of 0.5–2 Sv does not occur at low exposures. Furthermore, subsequent studies by Schull and Otake [4] quantified the risk of reduced intellect in patients after in utero ionizing radiation exposures. They estimated that intellect was reduced by approximately 30 IQ points per sievert. Even if a linear relationship existed between the dose and IQ reduction, one could predict that 0.01 Gy would not account for a doubling of the incidence of mental retardation because a linear extrapolation of Schull and Otake's data would represent a maximum reduction of only three tenths of an IQ point at 0.01 Gy. Behavioral studies in animals were unable to identify neurobehavioral effects at doses less than 0.02 Gy [9–12]. Although one has to be careful in extrapolating animal data to humans, the lack of neurobehavioral effects from in utero irradiation supports the other findings that indicate that mental retardation is a deterministic effect.
The risks for congenital malformations, fetal growth retardation, and miscarriage are not increased from absorbed doses below 0.20 Gy, and in fact, the thresholds for growth retardation and miscarriage are much higher [3, 13].
One other important factor is the matter of fractionation and protraction of the radiation. When diagnostic radiologic examinations involve procedures that take hours or are a combination of procedures that involve more than 1 day, the effective dose to the embryo is reduced [14].
The risk of cancer from in utero radiation exposure is a controversial subject. Exposure to ionizing radiation represents an oncogenic risk, but the magnitude of the risk is unclear. The risks cannot be estimated accurately because of conflicting data and an inability to make precise risk estimates.
Boice and Miller [5] summarize the dilemma as follows:
Since the reports in 1956 and 1958 that in utero radiation was associated with an increased risk of leukemia and solid cancers during childhood, the issue has been debated. Many epidemiologic studies have been performed. Evidence for a causal association derives almost entirely from case-control studies, whereas practically all cohort studies find no association, most notably in the atomic bomb survivors who were irradiated in utero. Although it is likely that in utero radiation represents a leukemogenic risk to the fetus, the magnitude of the risk remains uncertain. The causal nature of the risk of cancers other than leukemia is less convincing, and the similar relative risk (RR = 1.5) for virtually all forms of childhood cancer suggests an underlying bias. Few studies have addressed the potential risk of adult cancer after intrauterine exposure. Radiotherapy given to newborns, however, has been linked to cancer of the thyroid and breast later in life.
The last comment made by Boice and Miller [5] is of interest because animal studies reported by Brent [15, 16], Rugh et al. [17], and Reincke et al. [18] indicate that the juvenile mouse is more sensitive to the oncogenic effects of radiation than the embryo. None of these authors were enthusiastic to apply their results to the human, but human epidemiology is far from consistent in estimating the oncogenic risks of intrauterine radiation. The inconsistencies are one of the most prominent aspects of this area of research. A publication from the famous Collaborative Perinatal Project of the United States National Institutes of Health by Shiono et al. [19] emphasized that the offspring of mothers who were exposed to diagnostic radiation during pregnancy did not have an increased rate of malignancy, but an increased rate of cancer appeared in offspring of women who received radiation before pregnancy. These results are the opposite of what was found in the Radiation Effects Research Foundation studies in Japan.
When MacMahon [20] wrote his interpretation of the oncogenic risk of fetal radiation in 1985, many were surprised by his pessimism. He said, "The meaning of this association [statistical] remains unclear." In his summary paragraph he said, "It seems likely that the question of the association between fetal irradiation and childhood cancer will fade into medical history unresolved and remain the source of more confusion than enlightenment."
Twenty years later, it appears that he was partly correct. We are still confused, but the issue has not faded. Investigators are still attempting to determine the risk. Wakeford and Little [21] updated their excess absolute risk coefficient estimate in 2003 to be 0.08 Gy and said:
However, the statistical dosimetry, modeling and other uncertainties associated with these risk estimates are appreciable, and there is reason to believe that these coefficients could be systematic overestimates. This implies that doses to the foetus in utero of the order of 10 mSv discernibly increase the risk of childhood cancer. However, uncertainties in risk estimates are such that it is difficult to conclude reliably from these epidemiologic data what the level of risk from these low doses might be, beyond the inference that the risk is not zero or has been grossly understated.
How do you interpret the cancer risk for your patients who have received an absorbed conceptus or fetal dose of 100 mGy or less? It is quite simple. If you take the worst case scenario, you can convince yourself and the patient that the cancer risk is insignificant compared with the risk of spontaneous incidence of cancer and reproductive effects. Furthermore, it appears that the embryo is no more sensitive to the oncogenic effects of radiation than the young child.
Let us deal with these issues directly.
If sonography or MRI is available, it should be used for a pregnant patient when a radiologic procedure is needed.
An attempt should be made to keep the dose to the conceptus at 0.05 Gy or less if a diagnostic radiologic procedure is needed for a pregnant patient. However, the dose to the conceptus can be allowed to exceed 0.05 Gy if the procedure is vital for the health care of the mother and fetus. A radiologic procedure should not be performed on a woman if it is an elective follow-up examination or a preemployment screening examination, unless you are certain that she is not pregnant.
A pregnancy test should be performed if a patient is not certain whether she is pregnant. The pregnancy test may not be positive for 8–10 days after conception, which would be 23–25 days after the first day of the last menstrual period.
If the radiologic procedure must be performed because it is the best procedure and no nonionizing procedure would be equally satisfactory, then explain to the woman that the procedure will not increase the risk for birth defects, miscarriage, growth retardation, prematurity, or mental retardation. However, every healthy mother begins a pregnancy with a 3% risk for birth defects, a 15% risk for miscarriage, a 4% risk of prematurity, a 4% risk of growth retardation, and a 1% risk of mental retardation or neurologic developmental problems. Explain to the patient that physicians and scientists are not yet in a position to alter these background risks.
The risk of cancer from radiation is anxiety-provoking to a future mother who is aware of the issue. The physician or counselor should explain that the estimates of this risk are controversial. Every 10,000 infants that are born each year have a lifelong fatal cancer risk of approximately 18% (1,800/10,000). If the family is concerned about radiation-induced cancer, they should be reassured by the fact that approximately one extra case of cancer in this population of 10,000 would appear, if all 10,000 received 0.01 Gy. A family who has been counseled appropriately should be made aware of their background risks of 1,500 miscarriages, 300 birth defects, 1,100 early and late onset genetic disease, 400 premature babies, and 400 growth-retarded babies. These figures dwarf the small risk of cancer [3, 5].
In summary, we recommend that El-Khoury et al. [1] revise their "new era" permission forms and submit them to the NCRP, the Health Physics Society, and the radiologic societies for their comments. Plaintiff attorneys would be pleased to bring suit against radiologists who use the consent form published by El-Khoury et al. because the misinformation that it contains provides a basis for nonmeritorious negligence litigation.
Perhaps it would be simpler to require a consent form only when doses of 0.1 Gy (10 rad) more than fetal absorbed doses will be given. After all, below this level only small risks of cancer years later are present, except for large background risks. One can place this in context by realizing that current radiology practice often does not use a consent form for IV contrast material, which carries a risk (albeit small) of prompt death.
Our other criticism of the form as published is that it has attempted to provide generic counseling as if the embryo was equally sensitive throughout pregnancy. Informed physicians and health physicists recognize that thresholds for various radiation embryologic effects vary at different stages of pregnancy. The threshold figures that we have presented apply during the most sensitive periods. Even high exposures of radiation during the first 14 days of human embryonic development are unlikely to result in an increase in surviving abnormal embryos (the "all or none" period) [3, 21]. Counseling women who may be pregnant about radiation risks is a process that requires considerable knowledge of embryology and radiation embryology and cannot be summarized in a few sentences on a consent form [22, 23].
Finally, appropriate counseling should include not only the potential radiation risk but also the risk of harm that could result if the patient decides not to undergo the radiologic examination.
References
University of Iowa Hospitals and Clinics Iowa City, IA
52242-1077
Boston Medical Center Boston, MA 02118
University of Iowa Hospitals and Clinics Iowa City, IA
52242
University of Iowa Iowa City, IA 52242
We much appreciate the thorough comments and insights of Drs. Brent and Mettler on our work [1]. The intent of our policy was to take a consistent approach to situations in which the fetus is irradiated that included the patient and had an informed consent component. Another important intent was to satisfy those who interpret the literature to indicate a "benign" risk and those who have more concern from a biologic or medicolegal viewpoint. With this in mind, one has to set an action limit in which it becomes appropriate to use the consent form.
The 10-mSv dose that we chose is well below a level that would cause threshold-related effects, as we point out in the text and as Brent and Mettler assert in their letter, but several reasons justify its selection. Most radiographic procedures deliver well below 10 mSv to the fetus so that the need for this consent form is relatively infrequent. At our institution, we have used this consent form less than 12 times in the 24 months since the policy has been implemented. This dose is twice that allowed for the fetus of a radiation worker specified by the United States Nuclear Regulatory Commission. The 5-mSv limit is public information, and fetal doses in excess of that amount are likely to cause concern to pregnant patients.
The main issue that Brent and Mettler address is the information that is included in the 10- to 50-mSv consent form. It is written in the context of the fetus receiving 50 mSv and states that this is associated with a minimally increased risk of the listed radiation detriments to the fetus. The policy requires that the radiologists discuss the procedure with the patient so that the magnitude of the radiation risks using even the most conservative interpretation of the literature can be put in the proper perspective. We agree that in particular the inclusion of miscarriage on the 10- to 50-mSv consent form is inappropriate and should be omitted. We also suggest that others who wish to implement the policy described in our article carefully consider how they express the risks associated with radiation in their consent form. However, it would be difficult to ignore the childhood cancer risk because this has been the subject of numerous publications heightened by the concern over pediatric CT [3–5]. A main goal of our article was to provide a basis for thorough, sensitive counseling and discussion. The review provided by Brent and Mettler assists in this process.
References
1208
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  S. J. Patel, D. L. Reede, D. S. Katz, R. Subramaniam, and J. K. Amorosa Imaging the Pregnant Patient for Nonobstetric Conditions: Algorithms and Radiation Dose Considerations RadioGraphics, November 1, 2007; 27(6): 1705 - 1722. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. D. Lyerly, L. M. Mitchell, E. M. Armstrong, L. H. Harris, R. Kukla, M. Kuppermann, and M. O. Little Risks, Values, and Decision Making Surrounding Pregnancy Obstet. Gynecol., April 1, 2007; 109(4): 979 - 984. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. K. Yousefzadeh, M. B. Ward, and C. Reft Internal Barium Shielding to Minimize Fetal Irradiation in Spiral Chest CT: A Phantom Simulation Experiment. Radiology, June 1, 2006; 239(3): 751 - 758. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
