INTRODUCTION — Diagnostic imaging is sometimes necessary during pregnancy and in nursing patients. The safety of diagnostic ultrasound during pregnancy and in nursing patients is well established. However, other types of imaging evaluation may also be required. Although the safety of radiation exposure during pregnancy is a common concern, a missed or delayed diagnosis can pose a greater risk to patients and their pregnancies than any hazard associated with ionizing radiation. In many cases, the perception of fetal risk is higher than the actual risk.
This topic will review issues related to the safety of diagnostic imaging other than ultrasound in pregnant and nursing patients. Issues related to the safety of diagnostic ultrasound in pregnancy are discussed separately. (See "Overview of ultrasound examination in obstetrics and gynecology", section on 'Safety'.)
Many of the diagnostic imaging studies discussed in this topic are also performed on newborns. The effects of radiation in pediatric populations are also discussed separately. (See "Radiation-related risks of imaging" and "Clinical manifestations, evaluation, and diagnosis of acute radiation exposure", section on 'Pediatric considerations'.)
NOMENCLATURE — Any discussion of the effects of radiation requires knowledge of radiation nomenclature and dosimetry.
The amount of energy from ionizing radiation deposited in any absorbing material (eg, tissue) is the radiation-absorbed dose or "rad." An absorbed dose of 1 rad means that 1 gram of material absorbed 100 ergs of energy from exposure to radiation. The gray (Gy) is the radiation absorption dose measured in international (SI) units.
●1 rad = 0.01 Gy
The equivalent dose reflects the biologic effect of radiation exposure on human tissue. The unit for measuring the equivalent dose is the roentgen-equivalent man (rem). The sievert (Sv) is the radiation equivalent dose measured in SI units. Depending on the type of radiation (beta, gamma, alpha, or neutron), the absorbed dose may be the same as or lower than the equivalent dose.
●1 rem = 0.01 Sv
Some organs are more sensitive to radiation than others, and this difference is reflected by the effective dose. The effective dose is calculated by multiplying the equivalent dose to an organ by the tissue weighting factor for that organ. The unit for measuring the equivalent dose to an organ is also the rem or Sv.
Additional details of radiation dosimetry are provided separately. (See "Clinical manifestations, evaluation, and diagnosis of acute radiation exposure", section on 'Measures of radiation'.)
BACKGROUND RADIATION — In the United States, the average person is exposed to a radiation dose equivalent of approximately 3.1 mSv (310 mrem) whole-body exposure per year from natural sources [1]. The United States Nuclear Regulatory Commission recommends that occupational radiation exposure (ie, exposure related to employment) of pregnant patients not exceed 5 mSv (500 mrem) to the fetus during the entire pregnancy [2]. The dose equivalent to the fetus is the sum of the deep-dose equivalent to the pregnant patient and the dose equivalent to the embryo/fetus resulting from radionuclides in the embryo/fetus and radionuclides in the pregnant patient. External exposures are monitored using individual monitoring devices. When indicated, internal exposures are estimated by measuring the radiation emitted from the body or by measuring the radioactive materials contained in biological samples. (See "Clinical manifestations, evaluation, and diagnosis of acute radiation exposure", section on 'Biologic effects of radiation'.)
CHOICE OF IMAGING STUDY — Ultimately, the choice of the optimal imaging study for a pregnant or nursing patient is based on the information needed in the specific clinical setting, but with consideration of the potential fetal risks from exposure to radiation or contrast agents, or the risks to the infant from exposure to isotopes or iodinated contrast agents transferred into breast milk. These risks are discussed below. The choice of imaging study or studies is best made jointly by the clinical (medical, surgical, obstetric) providers and the radiologist, who can sometimes modify the technique to minimize fetal/infant risk without significantly compromising the information needed for maternal diagnostic evaluation and management.
TIMING
●Patients of reproductive age – At the time of the radiologic examination, all patients of childbearing potential should be asked if they could be pregnant [3]. If any doubt exists, the results of a pregnancy test should be obtained before proceeding.
The ideal time to schedule nonurgent imaging studies in patients of reproductive age is during the first 10 days of the menstrual cycle, if possible, as this will aid in avoiding imaging during an unrecognized pregnancy. Preconceptional ovarian exposure to diagnostic levels of ionizing radiation has no measurable effect on future fertility or pregnancy outcomes.
●Pregnant patients – Imaging studies should be performed in pregnant patients without regard to gestational age when the information is expected to provide information that is needed for care. The potential risks of the relevant radiologic imaging techniques and the potential risks of avoiding imaging (eg, nondiagnosis or inaccurate diagnosis, worsening disease) need to be considered on a case-by-case basis.
Multiple national and international organizations have written guidelines on imaging the pregnant patient. A comprehensive resource including 17 organizations and their 33 reports was published in 2011 [4]. In 2017, the American College of Obstetricians and Gynecologists (ACOG) published a committee opinion regarding "Guidelines for diagnostic imaging during pregnancy and lactation" [5]. The information in this topic is derived, in part, from these reports and can help in decision making.
●Nursing patients – Imaging studies are performed in lactating patients when medically indicated. Ionizing radiation to the breast does not affect the breastfed infant.
Need for iodinated contrast agents does not affect timing of the study. (See 'Risks to breastfed infants' below.)
In nuclear medicine studies, breastfeeding should be suspended for the period of time that radioactivity is present in milk, which depends upon the half-life of the specific agent. (See 'Risks to breastfed infants' below.)
Screening mammography is not routinely performed in patients at average risk of breast cancer under age 40, is a shared decision at age 40 to 49, and is recommended in patients age 50 and above. Ideally, age-appropriate screening is performed after nursing has been completed. The hormonal changes in pregnant and lactating patients may cause proliferation of ducts and lobules that result in increased density and nodularity of the breast parenchyma on mammography. These changes make it difficult to identify small nodules, asymmetries, and architectural distortion and thereby decrease the sensitivity of the examination, although not all patients undergo these changes in breast density. (See "Screening for breast cancer: Strategies and recommendations", section on 'Pregnancy and lactation'.)
MATERNAL RISKS FROM IONIZING RADIATION — Maternal radiation-related risks are generally similar to those in nonpregnant patients. These issues are reviewed in detail separately. (See "Radiation-related risks of imaging" and "Clinical manifestations, evaluation, and diagnosis of acute radiation exposure".)
However, pregnant patients experience proliferative changes in their breast tissue during pregnancy. Since proliferating breast tissue is more sensitive to radiation, it has been hypothesized that exposing the chest to ionizing radiation during pregnancy or lactation may lead to a small increase in the lifetime risk of breast cancer, but this has not been proven [6].
FETAL RISKS FROM IONIZING RADIATION
General principles
●There are no high-quality studies in pregnant humans from which to derive data on risks of ionizing radiation on the fetus; most of our information is based upon case reports and extrapolation of data from investigations of survivors of the atomic bomb in Japan and the Chernobyl accident [7-13]. Based on these data, the potential deleterious consequences of ionizing radiation can be divided into four main categories [14,15]:
•Pregnancy loss (miscarriage, stillbirth)
•Malformation
•Disturbances of growth or development
•Mutagenic and carcinogenic effects
The occurrence of each outcome depends upon the gestational age at the time of radiation exposure, the dose of radiation absorbed by the fetus, and the performance of fetal cellular repair mechanisms. Cellular damage caused by low levels of radiation exposure is usually repaired by a number of physiologic processes. By contrast, high-level exposure can interrupt important events in cell development and maturation, which may cause permanent injury or death.
●Deterministic effects only happen above a threshold dose and the severity of the effect increases as the dose increases above the threshold. A large number of affected cells results in more significant clinical problems. If injury to these cells occurs during a critical stage of organogenesis (primarily but not exclusively two to eight weeks after conception (figure 1)), impairment, agenesis, or deformity of the developing organ can occur. For example, microcephaly develops if a large number of differentiating central nervous system cells are injured. Pregnancy loss and growth restriction are other deterministic effects (table 1).
Stochastic effects can occur at any radiation dose and result in changes to the cell genome and altered differentiation and function of the affected cells. The probability, but not the severity, of the effect increases with increasing radiation dose. Childhood cancer is the primary stochastic effect of fetal radiation exposure. As an example, the increased risk of thyroid cancer as a result of in utero exposure to radiation after the Chernobyl accident is a stochastic effect [7]. (See 'Carcinogenesis' below.)
●The developing human is most sensitive to the lethal effects of ionizing radiation during the first 14 days after conception. During this period, the radiation-exposed pregnancy either survives undamaged or is resorbed (termed the "all or none" phenomenon) [16]. Radiation-induced teratogenesis, growth restriction, or carcinogenesis are not observed during this stage of development [17], presumably because of the pluripotent nature of each cell of the very early embryo. For human exposure, a conservative estimate of the threshold for death at this stage is more than 0.1 Gy rads (100 mGy, 10 rads) [18]. An embryonic dose of 1 Gy (1000 mGy, 100 rads) will likely kill 50 percent of embryos.
●During the period of organogenesis (approximately 2 to 8 weeks after conception or 4 to 10 weeks after the last menstrual period), the embryo may be damaged as a result of radiation-induced cell death, disturbances in cell migration and proliferation, or mitotic delay [19]. Lethality is rare. The major sequelae of radiation damage at this stage are fetal growth restriction and congenital malformations, particularly of the central nervous system (eg, microcephaly, gross eye abnormalities), which may be associated with intellectual disability. Microcephaly is the most frequently cited manifestation of radiation injury in utero [20]. In the absence of any of these findings, the presence of other types of malformations in humans should not be attributed to radiation exposure [17]. Carcinogenesis and an increase in the frequency of naturally occurring genetic mutations are other potential effects.
Fetal dose
Effects of dose less than 0.05 Gy — Diagnostic imaging studies typically expose the fetus to less than 0.05 Gy (50 mGy, 5 rads), and there is no evidence of an increased risk of fetal anomalies, intellectual disability, growth restriction, or pregnancy loss from ionizing radiation at this dose level [17,21]. The margin of safety is augmented by the fact that most human exposures from diagnostic imaging will be fractionated over a period of time; this type of exposure is less harmful than a single large acute exposure [17]. (See 'Fetal dose' above.)
Effects of dose more than 0.05 Gy — The threshold at which an increased risk of deterministic effects are observed in radiation-exposed embryos/fetuses has not been definitively determined. The evidence suggests the risk begins to increase at doses above 0.1 Gy (100 mGy, 10 rads), and particularly above 0.15 to 0.2 Gy (150 to 200 mGy, 15 to 20 rads).
Whether there is a risk between 0.05 and 0.1 Gy (50 to 100 mGy, 5 to 10 rads) is unclear [22]. In one retrospective study including 97 pregnant patients who underwent abdominal or lumbar radio-diagnostic imaging during the first trimester and in whom follow-up data were available, of the five patients exposed to 0.05 to 0.09 Gy (50 to 92 millisievert [mSv]), one congenital anomaly was observed; no congenital anomalies were observed in patients receiving <0.05 Gy, but numbers were too small to determine statistical significance [23].
It is important to note that even those diagnostic imaging studies associated with high fetal radiation exposure (eg, abdominal or pelvic computed tomography [CT], barium enema, cystourethrogram) almost never expose the fetus to this level of radiation. (See 'Fetal dose' above.)
Dose from specific imaging studies — Examples of the estimated fetal exposures for some common imaging studies involving ionizing radiation are listed in the table (table 2) [19,24,25]. It is important to recognize that radiography and CT imaging of regions far from the uterus minimally impact the fetus since the abdomen is shielded and it is only subjected to scattered radiation.
Although several fetal radiation dose tables are available, dosimetry calculations vary widely, which can be confusing for clinicians and patients. Therefore, when counseling a pregnant patient about the radiation risks associated with a diagnostic imaging study, the estimated dose for the specific fetus should generally be calculated by a radiologist or radiology physicist familiar with dosimetry. Factors to be considered include the number and type of projections, exposure time, distance between the target and the fetus, radiograph output, shielding, and use of digital acquisition systems designed to limit dose.
Techniques for minimizing fetal exposure — Preimaging consultation with the radiologist can help to ensure that the information necessary for maternal diagnosis and management is obtained at the lowest possible fetal radiation dose. Various techniques can be used to help safeguard the fetus.
●Nonabdominopelvic plain radiography – The patient should wear a lead apron to minimize fetal exposure from radiation scatter whenever nonabdominopelvic sites are being imaged. A fast film/screen combination or digital radiography can also be used to reduce total radiation exposure. Nevertheless, diagnostic radiographs of the head, neck, chest, and limbs (which do not include the fetus in the imaging field) produce almost no scatter to the fetus; thus, any radiation received would not result in a measurably increased risk of any adverse outcome.
●Abdominopelvic plain radiography – The following techniques can be used to minimize fetal radiation exposure during studies in which the fetus is directly in the field of view:
•A posterior-anterior exposure lowers the fetal radiation dose by 0.02 to 0.04 mGy (0.00002 to 0.00004 Gy, 2 to 4 mrad) compared with the traditional anterior-posterior exposure because the uterus is located in an anterior pelvic position.
•Shutters can be employed to collimate the radiation beam and reduce scatter.
•Avoiding magnification near the uterus and use of grids decrease the fetal dose of radiation.
•Minimize repeat examinations.
●Fluoroscopy and angiography – During fluoroscopic and angiographic imaging studies, modifying the exposure time, number of images obtained, beam size, and imaging area can reduce the amount of radiation exposure.
●Nuclear medicine – Maternal hydration and frequent voiding reduces fetal exposure to radionuclides excreted in the urine and accumulating in the maternal bladder.
●Computed tomography – The fetal radiation dose from a CT scan is affected by several variables, including the number, location, and thickness of slices. When CT imaging is performed in pregnancy, using a narrow collimation and wide pitch (ie, the patient moves through the scanner at a faster rate) results in a slightly reduced image quality but provides a large reduction in radiation exposure. Scanning protocols should also be modified. As an example, if performing a CT scan with contrast, the number of acquisitions can be reduced by eliminating the precontrast series. (See "Principles of computed tomography of the chest".)
Potential consequences
Malformations — For the developing fetus under 16 weeks of gestation, the threshold for possible prenatal radiation effects is approximately 0.1 to 0.2 Gy (100 to 200 mGy, 10 to 20 rads) [18]. After 16 weeks of gestation, the consensus of most researchers is that this threshold is much higher, at least 0.5 to 0.7 Gy (500 to 700 mGy, 50 to 70 rads). After approximately 20 to 25 weeks of gestation, the fetus is relatively resistant to teratogenic effects of ionizing radiation [26].
Developmental delay — Studies in survivors of the atomic bombing of Hiroshima demonstrated that the risk of intellectual disability and microcephaly was highest for radiation exposures at 8 to 15 weeks after conception [8]. The abnormalities were attributed to alterations in neuronal development. No cases of severe intellectual disability were identified in the children of atomic bomb survivors who were exposed prior to 8 weeks or after 25 weeks following conception. The risk appeared to be a linear function of dose, with a threshold of 0.12 Gy (120 mGy, 12 rads) at 8 to 15 weeks, and 0.21 Gy (210 mGy, 21 rads) at 16 to 25 weeks [9-12].
At 8 to 15 weeks, the average intelligence quotient (IQ) loss was approximately 25 to 31 points per Gy (per 1000 mGy or 100 rads) above 0.1 Gy (100 mGy, 10 rads), and the risk for severe intellectual disability was approximately 40 percent per Gy (per 1000 mGy or 100 rads) above 0.1 Gy (100 mGy, 10 rads). By comparison, at 16 to 25 weeks, the average IQ loss was approximately 13 to 21 points per Gy (per 1000 mGy or 100 rads) at doses above 0.7 Gy (700 mGy, 70 rads), and the risk of severe intellectual disability was approximately 9 percent per Gy above 0.7 Gy (700 mGy, 70 rads).
Growth restriction — Atomic bomb survivor data showed a permanent restriction of physical growth with increasing radiation dose, particularly above 1 Gy [18]. This was most pronounced when the exposure occurred in the first trimester. A 3 to 4 percent reduction in height at age 18 occurred when the dose was greater than 1 Gy (1000 mGy, 100 rads).
Carcinogenesis — Animal data suggest that carcinogenic effects are most pronounced during late fetal development [21]. Low levels (eg, 0.01 to 0.02 Gy [10 to 20 mGy; 1 to 2 rad]) of in utero radiation exposure may increase the risk of childhood cancer, particularly leukemia, by a factor of 1.5 to 2 over the baseline incidence of approximately 1 in 3000 [5,17,27]. Similarly, newborn radiation exposure of 0.01 Gy (10 mGy, 1 rad) increases the lifetime risk of developing a childhood malignancy, particularly leukemia, from the background rate of approximately 0.2 to 0.3 percent to approximately 0.3 to 0.7 percent [28]. However, the carcinogenic potential of low-level radiation is controversial since nonirradiated siblings of these children also have a higher incidence of leukemia. Furthermore, children exposed in utero at the time of the bombings of Hiroshima and Nagasaki have not developed a significantly increased rate of cancer [29]. An estimate of the risk of childhood leukemia after in utero exposure to radiation in various populations is shown in the table (table 3).
Solid cancer incidence rates have been examined among survivors of the atomic bombings of Hiroshima and Nagasaki who were in utero (n = 2452) or younger than 6 years (n = 15,388) at the time of the bombings [30]. Both the in utero and early childhood groups exhibited statistically significant dose-related increases in incidence rates of solid cancers, but the lifetime risks following in utero exposure were much lower than for early childhood exposure. At age 50, the estimated excess absolute rate per 10,000 person-years per Gy was 6.8 (95% CI <0-49) for those exposed in utero and 56 (95% CI 36-79) for those exposed as young children. There was no increase in oncogenic risk for exposures less than 0.2 Gy (200 mGy, 20 rads).
Genetic mutations — Radiation may increase the frequency of naturally occurring genetic mutations; it does not induce mutations unique to this source. Small increases in the rate of genetic mutation are difficult to detect because the background rate of spontaneous mutation is already high (approximately 10 percent), recessive mutations take several generations to become apparent, and autosomal dominant mutations are rare [10]. There is no way to distinguish radiation-induced genetic mutations from similar conditions arising from other environmental exposures. Studies attempting to estimate the incidence of radiation mutagenesis have been based largely upon animal and plant experiments. Very few human data are available, apart from observations in the offspring of atomic bomb survivors. An increased risk of genetic disorders induced by ionizing radiation has not been demonstrated in any human population at any radiation dose [2,10,31]. However, a perturbation of the normal sex ratio of live births, with an increase in the male-to-female ratio, has been reported and thought to be secondary to dysfunction of the paternal X chromosome [32].
FETAL AND INFANT RISKS FROM IODINATED CONTRAST MATERIALS
Fetal risks — Iodinated contrast materials may be used in pregnancy when indicated. They do not appear to be teratogenic or carcinogenic. However, iodinated contrast materials cross the placenta and can produce transient depressive effects on the developing fetal thyroid gland. Although the fetal thyroid begins to trap iodine in the first trimester and produces T4 and T3 by midgestation, clinical sequelae from brief exposures to iodinated contrast material in the second and third trimester have not been reported [33,34].
Risks to breastfed infants — Since most iodinated intravenous contrast agents are highly protein bound and rapidly cleared from the maternal circulation (half-life <60 minutes), they are present only at very low levels in breast milk [35-37]. Moreover, contrast agents have low oral bioavailability, so the infant absorbs a minimal amount of iodine. In a study of 10 newborns who received intravenous contrast media for urographic studies, thyroid function tests were checked at the time of the study and 10 and 30 days later [38]. No abnormalities were found, suggesting that even therapeutic doses of contrast media administered directly to infants do not affect infant thyroid function.
In their statement on administration of contrast medium to breastfeeding patients, the American College of Radiology (ACR) estimated that less than 0.01 percent of the maternal dose of iodinated contrast is absorbed by the breastfeeding infant. The ACR concluded that it is safe for patients to breastfeed after receiving contrast media, but they should be informed of the theoretical risks of direct toxicity or allergic reaction [39].
Patients who are concerned about theoretical adverse effects may pump to remove breast milk before administration of the contrast agent and then express and discard milk for 24 hours after the imaging study, which is also the position taken by most manufacturers in their package inserts.
FETAL AND INFANT RISKS FROM NUCLEAR MEDICINE
Fetal risks
Overview — Nuclear medicine studies (eg, pulmonary ventilation-perfusion, thyroid, bone, and renal scans) use a radioisotope bound to a chemical agent. The effect of these substances on the fetus depends upon maternal uptake and excretion, placental permeability, fetal distribution and tissue affinity, as well as the half-life, dose, and type of radiation emitted. Substances that can localize in specific fetal organs and tissues, and thus may be of concern, include iodine-131 (I131) or iodine-123 (I123) in the thyroid, iron-59 in the liver, gallium-67 in the spleen, and strontium-90 and yttrium-90 in the skeleton. Fetal exposure also results from proximity to radionuclides excreted into the maternal bladder; maternal hydration and frequent voiding can reduce this type of exposure.
Pregnant patients may have contact with individuals who have received radioactive materials as part of a diagnostic study; the minimal residual radioactivity does not result in a measurably increased risk to the fetus. Radiation exposure from close contact is higher after some types of therapeutic radiation (eg, radioiodine therapy of thyroid cancer, brachytherapy implants for prostate cancer) [40,41]. A period of restricted contact may be prudent, depending upon the type of therapy and dose administered.
Selected imaging studies
Thyroid scan — By the 10th to 12th week of gestation, radioiodine isotopes are readily absorbed by the fetal thyroid. Although there are no reports of adverse fetal effects from diagnostic doses of radioactive iodine, it should not be administered to pregnant patients because induction of thyroid cancer in the offspring is a concern [19]. If a diagnostic scan of the thyroid is required, the preferred agent is Technetium-99m or I123 (avoid I131) [5].
Positron emission tomography — There is minimal information regarding positron emission tomography (PET) in pregnancy. This technique involves injection of a radioisotope, fluorodeoxyglucose F 18. Animal reproduction studies have not been conducted with fluorodeoxyglucose F 18 Injection, and it is not known whether fluorodeoxyglucose F 18 Injection can cause fetal harm when administered to a pregnant patient or can affect reproduction capacity.
Because of the lack of safety data in human pregnancy, magnetic resonance imaging (MRI) or computed tomography (CT) are generally preferred to PET as they usually provide similar information, but the decision needs to be made on a patient-specific basis.
Risks to breastfed infants — Breastfeeding should be suspended for the period of time that radioactivity is present in milk; this will depend upon the half-life of the specific agent. The agent with the shortest half-life should be used. Before receiving the agent, the patient should express milk and store it in a freezer for later use. After the study, the patient should continue to express milk to maintain production and prevent engorgement, but discard the milk until the radioactive compound has been cleared. Radiology departments can also screen milk samples for residual radioactivity before the patient resumes breastfeeding [42].
The Committee on Drugs of the American Academy of Pediatrics advises interruption of breastfeeding for a minimum of three weeks after the patient has received I131, I125, N22, and Ga67 [43]. Because the lactating breast has a greater I131 affinity than the nonlactating breast and radioactivity persists after imaging, patients should stop breastfeeding at least four weeks before whole-body scans with I131 and should not resume breastfeeding thereafter to reduce the radiation dose, and potential cancer risk, to maternal breast tissue.
FETAL RISKS FROM MAGNETIC RESONANCE IMAGING
Overview — Magnetic resonance imaging (MRI) uses electromagnetic radio waves, rather than ionizing radiation, to generate detailed images. At the cellular level, possible direct biologic effects of MRI consist of 1) induction of local electric fields and currents from the static and time-varying magnetic fields, and 2) radiofrequency radiation resulting in heating of tissue. Other potential maternal dangers include trauma from projection of metal objects into the magnetic field (eg, small metal fragments can be projected into the eyes), interference with the operation of electronic devices (eg, cardiac pacemakers) or position of metallic implants, burns from heating of conductive materials in implants, and acoustic damage from high-intensity noise.
Despite these concerns, there are no reported adverse maternal or fetal effects from MRI during pregnancy [44-47]. In the largest study (MRI: n = 1737 deliveries; no MRI: n = 1,418,451 deliveries), first-trimester MRI was not associated with significantly increased risks for stillbirth, neonatal death, congenital anomaly, neoplasm, or vision or hearing loss in children followed up to age four years when adjustments were made for differences between exposure groups [47]. Magnetic field strengths were not reported but were likely at or below 1.5 Tesla since most cases can be done adequately at this level. A retrospective study of 81 neonates exposed in utero to 3 Tesla during MRI for maternal or fetal indications reported no significant difference in mean birth weight or prevalence of hearing impairment by 12 months of age between the exposed and control neonates [48]. We believe it is probably safe to image at 3 Tesla for neurologic imaging of the maternal head/neck but would only do this if there will be a clear advantage in diagnostic quality (ie, at the recommendation of a neuroradiologist). However, almost all safety studies have been performed predominantly at or below 1.5 Tesla magnetic field strengths. There may be an increased risk of tissue heating at higher field strengths. For example, in an animal study at 3 Tesla, heating effects were shown in amniotic fluid and fetal tissue [49].
In some cases, MRI is the preferred diagnostic modality because it provides better images than ultrasonography while avoiding the ionizing radiation of computed tomography (CT). As an example, first-trimester MRI is a reasonable option in a pregnant patient with suspected appendicitis in whom the appendix cannot be visualized by ultrasound examination. An expert panel on MRI safety opined that MRI should be performed at any stage of pregnancy when the information requested from the MRI study cannot be acquired by other nonionizing studies, the data are needed to potentially affect the care of the patient or fetus during the pregnancy, and it is not prudent to wait until the patient is no longer pregnant to obtain these data [50].
MRI is sometimes used during pregnancy specifically to image the fetus or placenta.
Use of gadolinium — Gadolinium, the contrast agent most commonly used for MRI because of its magnetic properties, crosses the placenta and is excreted by the fetus into the amniotic fluid. It is then swallowed; thus, it can be reabsorbed into the fetal circulation. Given the potentially long half-life in the fetus and few data from human pregnancy, it is not recommended for use in the pregnant patient unless the potential benefit justifies the potential risk to the fetus (ie, "if it significantly improves diagnostic performance and is expected to improve fetal or maternal outcome" [5]) [51,52].
Gadolinium increases the signal from tissues that have increased blood flow, particularly in the setting of inflammation or neoplasm. In the mother, it is used for MRI evaluation of the brain and spinal cord, suspected inflammatory joint disease, inflammatory bowel disease, and inflammatory and neoplastic conditions of solid organs. It may also be useful for evaluation of inflammatory and neoplastic conditions of bone, muscle, and connective tissue.
Animal studies have shown very low levels of gadolinium in fetal tissues (predominately bone) after in-utero exposure [52]. Human data from gadolinium exposed pregnancies are limited but concerning [47,53-57]. The largest human study comparing gadolinium MRI during pregnancy (n = 397) with no MRI (n = 1,418,451) reported gadolinium exposure at any time during pregnancy was associated with an increased risk for rheumatologic, inflammatory, or infiltrative skin conditions (adjusted hazard ratio 1.36, 95% CI 1.09-1.69) and for stillbirths and neonatal deaths (adjusted relative risk 3.70, 95% CI 1.55-8.85, 7 versus 9844 events) but not for nephrogenic systemic fibrosis or congenital anomalies [47].
Nephrogenic systemic fibrosis (NSF) is a rare disorder that usually occurs in adults and could be misdiagnosed as a connective tissue or skin disease in young children. It has been hypothesized that persistence of gadolinium in the fetus may increase the risk of NSF in children exposed in utero. The adjusted HR for NSF in this study was 1.00, but the confidence interval was wide (95% CI 0.33-3.02) given the small number of cases (≤5) and the challenge of making the diagnosis in a young child. For this reason, and because of the increased risk for the broader outcome of any rheumatologic, inflammatory, or infiltrative skin condition, an increased risk of NSF could not be conclusively excluded, particularly with first-trimester exposure. (See "Nephrogenic systemic fibrosis/nephrogenic fibrosing dermopathy in advanced kidney disease".)
This study involved a previous generation of gadolinium-based contrast agents. The newer generation is more stable and may have a lower potential for fetal toxicity, but no data are available. Therefore, gadolinium should generally be avoided in the pregnant patient unless its use significantly improves diagnostic performance and is likely to improve patient outcome.
Risks to breastfed infants — Gadolinium-based contrast agents are present at very low levels in human milk and not absorbed well by the infant gut; no adverse effects have been reported in infants exposed through lactation [58-60].
In their statement on administration of contrast medium to breastfeeding mothers, the American College of Radiology (ACR) estimates that less than 0.0004 percent of gadolinium-based contrast is absorbed by the breastfeeding infant. The ACR concluded that it is safe for patients to breastfeed after receiving contrast media, but mothers should be informed of the theoretical risks of direct toxicity or allergic reaction [61,62]. The American College of Obstetricians and Gynecologists (ACOG) also concluded that breastfeeding should not be interrupted after gadolinium administration [5]. Patients who are concerned about theoretical adverse effects may pump to remove breast milk before administration of the contrast agent and then express and discard milk for 24 hours after the imaging study, which is also the position taken by most manufacturers in their package inserts.
By comparison, the Contrast Media Safety Committee of the European Society of Urogenital Radiology concluded that breastfeeding should be avoided for 24 hours after injection of gadolinium-based contrast medium if agents with a high risk for NSF are used (eg, gadodiamide, gadopentetate dimeglumine, gadoversetamide) [63]. Patients who receive contrast agents with intermediate risk for NSF (gadofosveset, gadobenate dimeglumine, gadoxetate disodium) may wish to consider the procedure described above (pumping before the imaging study, and expressing and discarding breast milk for 24 hours after the study).
FETAL RISKS FROM ULTRASOUND — No biologic effects have been documented from diagnostic ultrasound in the pregnant patient, despite intensive use over several decades. The potential for deleterious consequences from heat and cavitation exists since ultrasound uses sound waves that interact with biological tissues. B-mode and M-mode imaging operate at acoustic outputs that do not produce harmful temperature rises. However, Doppler ultrasound does have this potential; therefore, guidelines for Doppler use in pregnancy have been formulated to minimize exposure time and acoustic output. The safety of ultrasound in pregnancy is discussed in detail separately. (See "Overview of ultrasound examination in obstetrics and gynecology", section on 'Safety'.)
APPROACHES TO COMMON DISEASE-SPECIFIC IMAGING IN PREGNANT PATIENTS
●Renal stones – (See "Kidney stones in adults: Kidney stones during pregnancy".)
●Pulmonary embolism – (See "Diagnosis of pulmonary embolism in pregnancy".).
●Dental radiography – The radiation dose to the fetus from maternal dental radiography is minute, 0.0001 mGy (0.01 mrads) for an average study, and is not considered harmful. Although one population based case-control study found an association between antepartum dental radiography of >0.4 mGy (40 mrads) to the maternal thyroid and low birth weight (less than 2500 g) [64], this association is not consistent with findings from multiple other studies and is not biologically plausible [15]. Further investigation is needed before any change is made to the recommendations for dental imaging in pregnant patients.
●Evaluation of a breast mass – (See "Gestational breast cancer: Epidemiology and diagnosis", section on 'Diagnosis and staging' and "Breast imaging for cancer screening: Mammography and ultrasonography", section on 'Abnormalities on mammography'.)
●Deep vein thrombosis – (See "Deep vein thrombosis in pregnancy: Epidemiology, pathogenesis, and diagnosis", section on 'Imaging'.)
●Appendicitis – (See "Acute appendicitis in pregnancy", section on 'Imaging'.)
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Imaging in pregnancy".)
SUMMARY AND RECOMMENDATIONS
●When performing any imaging study in a patient of childbearing age, always consider whether the patient may be pregnant before performing the study. (See 'Timing' above.)
●Ideally, we suggest scheduling nonurgent imaging studies during the first 10 days (follicular phase) of the menstrual cycle. All patients of childbearing potential should be asked if they could be pregnant at the time of a radiologic examination. If any doubt exists, a pregnancy test should be obtained prior to the diagnostic imaging studies. The perceived risk of radiation exposure is much greater than the actual risk, but a full explanation of these risks to the patient and the patient's family is best given prior to, rather than after, the exposure. (See 'Introduction' above.)
●During pregnancy, ultrasound examination and magnetic resonance imaging (MRI) are generally preferred to imaging modalities that involve ionizing radiation. (See 'Fetal risks from ultrasound' above and 'Fetal risks from magnetic resonance imaging' above.)
However, concern about the possible effects of ionizing radiation should not prevent medically indicated diagnostic imaging studies using the best available modality for the clinical situation. When imaging studies requiring ionizing radiation are necessary, various techniques can be employed to minimize the radiation dose. (See 'Fetal risks from ionizing radiation' above.)
●The choice of imaging study or studies is best made jointly by the clinical (medical, surgical, obstetric) providers and the radiologist, who can sometimes modify the technique to minimize fetal/infant risk without significantly compromising the information needed for maternal diagnostic evaluation and management. (See 'Choice of imaging study' above.)
●Radiation risks should be discussed with the pregnant patient, including an explanation of the background population risk for miscarriage, congenital anomalies, genetic disease, and growth restriction (approximately 20, 4, 10, and 10 percent, respectively), as well as the risk of developmental disorders. (See 'Fetal risks from ionizing radiation' above.)
●At doses less than 0.05 Gy, there is no evidence of an increased risk of fetal anomalies, intellectual disability, growth restriction, or pregnancy loss from ionizing radiation. There may be a small increased risk of childhood cancer, 1 in 1500 to 2000 versus the 1 in 3000 background rate. (See 'Fetal risks from ionizing radiation' above.)
●During the first 14 days after fertilization, intact survival or death are the most likely outcomes of radiation exposure above 0.05 Gy (5 rads). A conservative estimate of the threshold for intrauterine death is more than 0.1 Gy (10 rads). (See 'Fetal risks from ionizing radiation' above.)
●After the first 14 days, radiation exposure over 0.5 Gy may be associated with an increased risk of congenital malformations, growth restriction, and intellectual disability. (See 'Fetal risks from ionizing radiation' above.)
●Gadolinium should generally be avoided in the pregnant patient, unless its use significantly improves diagnostic performance and is likely to improve patient outcome. Gadolinium-based contrast agents are present at very low levels in human milk and not absorbed well by the infant gut; no adverse effects have been reported in infants exposed through lactation. (See 'Use of gadolinium' above.)
●Although there are no reports of adverse fetal effects from diagnostic doses of radioactive iodine, it should not be administered to pregnant patients because induction of thyroid cancer in the offspring is a concern. If a diagnostic scan of the thyroid is required, the preferred agents are Technetium-99m or iodine-123 (but not iodine-131). (See 'Thyroid scan' above.)
●MRI can be performed at any stage of pregnancy when the information requested from the study cannot be acquired by nonionizing imaging studies, and the data are needed to care for the patient or fetus during the pregnancy. (See 'Fetal risks from magnetic resonance imaging' above.)
●For patients undergoing nuclear medicine scans with radioisotopes, breastfeeding should be suspended for the period of time that radioactivity is present in milk; this will depend upon the half-life of the specific agent (see 'Risks to breastfed infants' above). It is safe for patients to breastfeed after receiving iodinated contrast media. (See 'Risks to breastfed infants' above.)
