Diagnostic imaging in pregnancy: Making informed decisions

Short abstract

The use of diagnostic imaging involving ionising radiation may be necessary in pregnancy and requires an assessment of the most appropriate and safest imaging modality which will provide the necessary information balanced with the potential risks to the mother and fetus. In most cases, this will involve a potential fetal radiation dose well below 50 mGy. At these doses, there is no risk of lethality, genetic damage/epigenetic change, teratogenicity, growth impairment or sterility. Older epidemiological data indicating a potential increased cancer risk have been contradicted by newer data and better understanding of the biology of low dose radiation. The linear no-threshold rule has been challenged by many and more realistic estimates of oncogenicity risk along with the potential risks of contrast agents are summarised in this review. Imaging in the pregnant population is increasing in both the number of examinations performed and the number of patients being imaged, with the greatest increase being computed tomography scans. Counselling and obtaining informed consent for imaging that involves radiation requires the clinician to communicate with the woman and her family a realistic estimate of the potential radiation dose to herself and her fetus, to describe and quantitate the risks of this estimated dose, to outline the benefits of the imaging procedure and to respond to any questions or concerns. As almost all diagnostic imaging involves doses below the 50 mGy threshold, clinically indicated investigations should not be withheld during pregnancy. All allied staff must also be well informed to ensure the patient receives a consistent message about the risks and benefits of the proposed test.

Introduction

The use of diagnostic imaging in pregnancy requires an assessment of the most appropriate and safest imaging modality which will provide maximal information balanced with the potential risks to the mother and fetus. Imaging techniques that do not involve electromagnetic radiation, i.e. ultrasound and magnetic resonance imaging (MRI) are considered safe in pregnancy. Various guidelines have been published expressing confidence in the use of MRI throughout pregnancy, including the first trimester, with no evidence of adverse outcomes on pregnancy or the fetus/neonate.^1–6

X-rays, computerised tomography (CT) with and without iodinated contrast, nuclear scanning and positron emission tomography scanning all involve potential exposure of both mother and fetus to ionising radiation. Such imaging may occur inadvertently or by intention during pregnancy, necessitating appropriate counselling. In addition, CT, MRI or even ultrasound may require administration of a contrast agent to improve image definition or other additional information.

The major radiation source for most women during pregnancy is environmental, arising from a number of sources including cosmic, gamma and radon. The magnitude of the exposure varies (see Table 1) with a global average estimated radiation exposure over the nine months of pregnancy being 2.3 mSv.^7,8 For a fetus, the dose received is much smaller than that of the mother, as a result of attenuation through the maternal tissue.⁹

Table 1.

Sources of environmental radiation.⁷

Source	Exposure (mSv)
Airport scannerAir travel:	0.005 per exposure0.005 per short flight*0.03 per long flight*
Aircrew/frequent flyers	1–5 per year
Background radiation
• UK or Australia	1.8 per year
• USA	6.2 per year
• Kerala, India	30 per year
• Some areas of Brazil and Sudan	40 per year
• Ramsar in Iran	Up to 260 per year
Uranium miners/ nuclear industry workers	Additional to background: 1.5–2.5 per year

The unit mSv refer to exposure whilst the absorbed dose is measured in mGy.*At higher latitudes, multiply by a factor of 2--3.

When a pregnancy is declared, by a female working in the presence of radiation, the embryo or fetus is afforded the same level of protection as the general public, i.e. 1 mSv per year. The governance of the radiation industry requires that all operators adhere to an ‘as low as (is) reasonably achievable’ (ALARA) principle, which means making every reasonable effort to maintain exposures to ionising radiation as far below the dose limits as practical.

Radiation dose and pregnancy

The calculation of radiation dose, both maternal whole body and organ specific as well as the estimated fetal absorbed dose, is calculated based on established dosimetry tables or specific calculations made by a medical physicist on a case-by-case basis. Published dosimetry tables do vary but for counselling purposes, diagnostic imaging procedures can be divided into negligible (<0.1 mGy), low-moderate (0.1–10 mGy) and higher dose (10–50 mGy).⁴ All radiation doses are potentially additive and where multiple procedures are performed, cumulative dosage should be calculated by a qualified medical physicist.

For most standard radiography, the fetal dose is very low (Table 2) although adjustments are often made to procedures to lower the fetal absorbed dose even further.⁹ This should only be done if image quality is not compromised. In addition to technical aspects of the radiation procedure, e.g. type of procedure, the site and radiation dose administered, maternal factors will influence the fetal absorbed dose. During X-rays and CT scanning, the ‘thickness’ of the mother, which alters as pregnancy progresses, will influence the penetration of the dose and hence the fetal absorbed dose. The use of appropriate shielding will significantly reduce the fetal absorbed dose as well as the dose to adjacent maternal tissues.

Table 2.

Estimated fetal dose from common imaging procedures.

Fetal dose range (mGy)	Examples of procedures
0.001–1.0	X-rays: beyond 10 cm of uterus including head, chest, breast, teeth, extremitiesX-rays: abdomen, pelvis, hippulmonary angiogramCT: head or neck, pelvimetry, chest, computed tomography pulmonary angiogram (CTPA)Nuclear scan: Lung ventilation, liver/spleen, ventilation/perfusion (V/Q)
1.1–10	X-rays: Barium enema, intravenous pyelography, lumbar spineCT: lumbar spine, abdomen, coronary artery angiographyNuclear scan: renal, white cell, bone, cardiac.
11–50	CT: pelvis, pelvis and abdomen, pelvis and abdomen and chestNuclear scan: PET/CT whole body scan, myocardial perfusion

CT: computed tomography; PET: positron emission tomography.

Note: Cumulative dose estimates are required if more than one imaging procedure is performed during pregnancy. If possible, for dose estimates >10 mGy, individual patient dose calculations should be made by a medical physicist.^5,9,10

In the case of nuclear medicine studies, fetal absorbed dose will represent the cumulative effect of external irradiation from the maternal tissues as well as placental transfer and fetal uptake of radiopharmaceuticals. For radioisotopes that are excreted in urine, radiation from bladder contents or the placenta will have a greater impact in early pregnancy compared with later pregnancy. In general, the whole body fetal absorbed dose from nuclear scanning studies tends to decrease throughout pregnancy.¹¹ By using smaller administered doses and longer imaging times, the fetal absorbed dose may be further reduced, but there is concern that this may compromise the value of the study. Although a reduction of the injected dose can be compensated for by increasing the scanning time, if the acquisition time is longer patients are more likely to move or ask to interrupt the scan to void the bladder (not a rare occurrence in pregnant women).¹² Strategies to aid excretion of radiopharmaceuticals, e.g. ensuring prompt and complete bladder emptying will reduce the fetal absorbed dose without compromising image quality and should be encouraged.

Nuclear scanning has been enhanced by the addition of hybrid single photon emission computed tomography (SPECT) which allows simultaneous acquisition of combined multi-modality imaging, with seamless fusion of three-dimensional volume datasets. The application of SPECT provides enhanced information with simultaneous imaging of function and anatomical localisation. The fetal radiation dose from such procedures will be greater than from any single modality but the additional information may be critical.

Risks of ionising radiation

A number of national and international bodies have issued guidelines or policy documents which indicate a negligible risk to the fetus for radiation doses below 50 mGy, well below the dose from any single diagnostic imaging event although cumulative studies could lead to such a dose.^4–6 The effects of radiation exposure in pregnancy depend on the time of exposure as well as the fetal absorbed dose. Until the placenta implants, the cells of the conceptus are hypoxic and therefore less radiosensitive. In the very early embryo, the effect of radiation is more likely to be failure to implant or undetectable death of the embryo, i.e. an all or none phenomenon.¹³

Radiation effects may be classified as either deterministic, also called tissue reactions, or stochastic. Cell killing, leading to fetal death, gross malformation, developmental abnormalities or growth retardation is a deterministic event, i.e. there is a threshold dose below which no effect is seen and the higher the dose the greater the effect. In comparison, malignancy and hereditary abnormalities are stochastic effects, i.e. the absorbed dose influences the probability but not the severity of the effect. Both may be relevant when imaging in pregnancy although the evidence for stochastic effects is scant and inconsistent.¹⁴

Experimental assessment of very high dose radiation has identified seven specific areas of potential concern to the pregnant woman and her fetus.

• Lethality

• Genetic damage/epigenetic change

• Teratogenicity

• Growth impairment

• Sterility

• Oncogenicity

With regard to potential fetal dosage from individual diagnostic radiation procedures, there is no evidence for any increase in lethality (miscarriage or stillbirth), teratogenicity, genetic damage/epigenetic changes, growth impairment, mental retardation or sterility (Table 3).

Table 3.

Threshold dose effects of radiation at different stages of pregnancy.^15–17

Gestation (weeks)	Effect	Estimated threshold (mGy)
2–4	Miscarriage or no effect	>50–100
4–10	Congenital anomalies: skeleton, eyes, genitals Intrauterine growth restriction	200
10–17	Microcephaly Intellectual impairment	>200
>18	Intellectual impairment	>250

Oncogenicity

It has generally been assumed that the relationships observed between radiation dose and adverse effects at high levels of radiation exposure also apply (proportionally) to low levels: the linear no-threshold (LNT) hypothesis. This concept remains a source of controversy amongst experts with the International Commission on Radiological Protection and the World Health Organisation supporting the theory whilst the World Nuclear Association and the United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) argue that there is compelling biologic and epidemiologic data to refute the LNT modelling of radiation risk. The LNT model does not account for proven biological adaptive responses that change the shape of the LNT dose–response line in the low dose range (Figure 1) including antioxidant production, apoptosis, immune system-mediated effects and repair of DNA double-strand breaks. All biologic and experimental data suggest the existence of practical thresholds for carcinogenesis. This concept means that below the dose threshold, the carcinogenic risk, if it exists, is so small that it is without clinical importance. Cuttler¹⁸ has used the leukaemia incidence data in Hiroshima survivors to compare the estimated (based on LNT) versus actual incidence (gathered by UNSCEAR) to illustrate the discrepancies (Figure 1).

Figure 1.

Actual leukaemia incidence (UNSCEAR data) in the Hiroshima survivors for 1950–57. Graphic extrapolation based on estimated dose (±standard error) compared with LNT model. The dotted line represents a re-estimate of the doses because many of these individuals had features of radiation complications indicating a potential underestimation of their dose based on distance alone. Source: Reproduced from Cuttler¹⁸ under CC-BY-NC license.

A large number of epidemiological studies have been performed to assess the possible effects of prenatal radiation on the incidence of malignant disease. All are flawed by problems such as retrospectivity, small study size, inadequate or inappropriate case and control selection, and variability in the determination of radiation exposure and measurement of outcome parameters. The studies in which prenatal radiation exposure have been associated with an increased incidence of malignancy have found a relative risk ratio of all cancer of 1.5–2.4.^19–21 However, a number of other epidemiological studies have failed to establish any statistically significant association between prenatal exposure to radiation and childhood malignancy.^22–24 At least part of the increased risk associated with irradiation could be accounted for by the fact that mothers with a higher incidence of illness during pregnancy (a susceptibility which might be associated with an increased risk of tumour in their offspring) had a greater incidence of exposure to diagnostic radiation. The National Radiological Protection Board has adopted an estimated additional risk (EAR) co-efficient for cancer incidence under 15 years of age following low-dose irradiation in utero of 0.006% per mGy compared with a risk of 0.0018% per mGy for a dose received just after birth. Estimates for the extra risk of childhood cancer from a fetal radiation dose of 1000 mGy range from 0.022 (Oxford Survey Childhood Cancer) through to 0.060.^25,26 Given a baseline risk of childhood cancer of between 1.0 and 2.5 per 1000, Table 4 gives estimates of risk comparing a low risk versus the high risk modelling based on available data.²⁷

Table 4.

Estimates of cancer risk from fetal radiation in pregnancy – comparing a low risk versus high risk model.²⁷

Fetal dose (mGy)	Low risk model	High risk model
10	1 in 4545	1 in 1667
20	1 in 2272	1 in 834
30	1 in 1515	1 in 556
40	1 in 1136	1 in 417
50	1 in 909	1 in 334

Perhaps, a simpler statistic is that quoted by Tirada et al.⁹ of an EAR of cancer following a 50 mGy in utero exposure of 1.1–3.0 events per 1000 exposures. The data regarding the risk at different gestations and on various systems and organs are imprecise with wide confidence intervals (CIs) owing to the low doses and the limited sample sizes in available studies.¹⁷

Radioiodine and its effects on the fetal thyroid

Nuclear scanning of the thyroid is generally avoided in pregnancy; however, inadvertent scans do occur in early pregnancy and the following data are helpful when counselling such women.²⁸ Table 5 describes the estimated fetal radiation dose from thyroid nuclear scans based on current dosimetry assessment.¹¹

Table 5.

Estimated fetal radiation dose from thyroid scanning at different stages of pregnancy.¹¹

Radiopharmaceutical	Fetal dose (mGy) at three months	Fetal dose (mGy) at six months	Fetal dose (mGy) at nine months
99mTc	5.52	1.17	0.66
131I	0.31a	0.7a	0.51a

^aIn the case of ¹³¹I, the fetal thyroid dose is significantly larger than this tabulated fetal dose.

The special case of radioiodine administered for treatment of thyrotoxicosis or thyroid cancer requires specific mention. ¹³¹I given for therapeutic rather than diagnostic purposes crosses the placenta readily and the fetal thyroid begins to accumulate iodine from about the end of the first trimester. In early pregnancy, the major risk is from external gamma radiation from the maternal bladder whilst after 12 weeks, fetal and/or placental uptake increases and the fetal thyroid becomes functional leading to accumulation of tracer. Hence, the cumulative fetal thyroid dose in later pregnancy is much greater than the dose to the rest of the fetus.^10,29 If therapeutic doses of radio-iodine are inadvertently given to a pregnant woman with thyrotoxicosis, there is a significant risk of fetal thyroid damage after 12 weeks’ gestation. Inadvertent use of radioiodine prior to 12 weeks is unlikely to damage the fetal thyroid.³⁰ If pregnancy is confirmed shortly after dose administration, maternal hydration and frequent voiding should be encouraged and potassium iodide may be given as a thyroid-blocking agent. If therapeutic ¹³¹I is administered to a woman of childbearing age, a minimum period of six months before conceiving is recommended by some experts.²⁸ However, this is based on an intention to achieve stable thyroid function post thyroid ablation, rather than any risk of radiation or iodine damage to the fetal thyroid. From a radiation safety point of view, conception should be delayed until the potential fetal dose from residual radioiodine is less than 1 mGy. A number of observational studies have demonstrated no impact on fertility, pregnancy or neonatal outcome after radioiodine therapy,^31–33 and termination should not be recommended even if the period to conception was less than six months as the risk of harm is low.

Maternal risks of radiation in pregnancy and postpartum

The main area of concern both during pregnancy and lactation has been the effects of radiation on breast tissue. This is particularly relevant with regard to investigation of suspected pulmonary embolism where there is a great deal of debate about CTPA versus V/Q scanning.^34–40 The fetal radiation dose varies with gestation but is below 1 mGy with both CTPA and V/Q, but the maternal breast dose with CTPA is 10–70 mGy, significantly greater than with V/Q (<1.5). In a recent paper, Isidoro et al.⁴¹ published estimations of radiation dose to both the fetus and maternal breast at varying gestations (Table 6). The use of bismuth breast shields or modulation of the CT technique can reduce this dose but may impair image quality. The lung dose from either study is similar at 7–14 mGy.

Table 6.

Estimated absorbed dose (mGy) for maternal breast and fetus from CT and nuclear imaging for investigation of suspected pulmonary embolism.⁴¹

Organ		CTPA non-contrast	CTPA contrast enhanced	Perfusion (Q) SPECT scan only	V/Q SPECT
Estimated maternal breast dose		12	22	0.25	0.50
Estimated fetal absorbed dose by gestation
Early		0.16	0.21	0.14	0.18
Three months		0.20	0.28	0.20	0.25
Six months		0.43	0.73	0.25	0.31
Nine months		0.42	0.57	0.20	0.25

CTPA: computed tomography pulmonary angiogram; SPECT: single photon emission computed tomography.

Technical factors caused by pregnancy, such as hyperdynamic circulation and hemodilution, can affect contrast opacification in CTPA. Poor lung expansion and breath-holding secondary to the gravid uterus can also affect the CTPA. Other imaging options include CXR with perfusion scanning (Q) only or magnetic resonance angiography (MRA). The latter has lower diagnostic yield but no radiation.

Previous assessments have estimated the lifetime risk of breast cancer from a dose of 20 mGy to the breast is approximately 1/1200 at age 20, 1/2000 at age 30 and 1/3500 at age 40.^40,42 From these estimates, the additional breast cancer risk from CTPA in a 30-year-old woman is estimated as 1/2000 compared with 1/40,000 for a V/Q scan. This figure is not very high, but studies have suggested that this rate is seven times higher in the pregnant woman and hence should be considered in the diagnostic algorithm.^43,44 In breastfeeding women, only tiny amounts of iodinated or gadolinium-based contrast medium given to a lactating mother are excreted into breast milk, and only a minute proportion enters the baby’s gut. Therefore, after CTPA or MRA it is not necessary to discard breast milk.

After a V/Q scan, only very tiny amounts of radioactivity can be detected in breast milk.⁴⁵ Interruption of breastfeeding for 12 hours is recommended but milk pumped during the waiting period can usually be stored and used later, with the storage duration dependent on the radioactive isotope’s half-life.^46,47

In addition to the fetal radiation dose, a number of other factors may determine the choice between CTPA and V/Q scanning including sensitivity, specificity, rate of inconclusive studies, the availability of each investigation, the hemodynamic stability of the woman, the presence or absence of a likely alternative diagnosis and the presence or absence of deep venous thrombosis. If multiple investigations are performed, a cumulative fetal and maternal breast radiation dose should be calculated.

Contrast agents

A number of X-ray and CT procedures require the administration of iodine-containing contrast agents. There are no reports of teratogenesis from iodinated contrast agents but the amount of inorganic iodine available to interfere with thyroid metabolism is about 0.1% of the dose administered. Non-ionic contrast agents have been shown to cross the placenta and inhibit Type II and III deiodinases which can reduce intracellular triiodothyronine, the levels of which directly affects transcription of many genes that are important in fetal development.⁴⁸ In addition, depending on the dose of iodine, there is a theoretical risk of fetal thyroid blockade although to date this has not been reported.⁴⁹

The use of gadolinium, although not teratogenic, has generated concern as gadolinium has been shown to cross the placental barrier and will remain in the amniotic fluid indefinitely. There are concerns that the gadolinium ion may dissociate from its chelate molecule leading to possible adverse effects such as neurotoxicity.⁵⁰ A recent large retrospective study of first trimester exposure with gadolinium (n = 397) identified a slight increase in stillbirths or neonatal death with an adjusted relative risk of 3.70 (95% CI 1.55–8.85) although this may be related to the condition necessitating the MRI rather than the procedure.³ There was a significant increase in a composite outcome of rheumatological/inflammatory/infiltrative skin conditions (adjusted hazard ratio for all MRI exposure in first trimester 1.36 95% CI 1.09–1.69) but there was no increase in connective tissue or skin disease resembling nephrogenic systemic fibrosis, a syndrome associated with gadolinium exposure outside of pregnancy. These outcomes were not increased in MRI without gadolinium but adverse events were all rare. Expert bodies have concluded that the effects of MRI contrast agents remain unknown and may be harmful. There is no requirement to cease breastfeeding following the administration of gadolinium to a lactating woman as it has low lipid solubility with less than 0.04% measurable in breast milk.⁴

Contrast-enhanced ultrasound is a technique combining traditional ultrasound with contrast agents such as small air bubbles or gas-filled microbubbles to enhance echogenicity in order to more accurately image blood flow, e.g. for assessment of septal defects in the heart. Although the European Federation of Societies for Ultrasound in Medicine and Biology have recommended against its use because of limited experience, small studies in human pregnancy have demonstrated no harm.^51,52 After a stroke or transient ischaemic event, the presence or absence of a right–left cardiac shunt may influence management and should not be withheld because of pregnancy.

Counselling and informed consent

Counselling and obtaining informed consent for imaging that involves radiation requires the clinician to communicate with the woman and her family a realistic estimate of the potential radiation dose to herself and her fetus, to describe and quantitate the risks of this estimated dose, to outline the benefits of the imaging procedure and to respond to any questions or concerns. All allied staff must also be well informed to ensure the patient receives a consistent message about the risks and benefits of the proposed test(s).

For practical purposes, no specific counselling is required for women undergoing diagnostic imaging with a predicted fetal absorbed dose of less than 10 mGy. This includes all X-ray and CT scanning not involving the abdomen and most nuclear scans. For direct exposures or nuclear scanning with a potential exposure greater than 10 mGy, the women should be counselled on a risk/benefit basis. Whilst there is an ongoing debate in the literature regarding the appropriateness of applying a LNT model for counselling, all current guidelines use such a model in their recommendations. If there is any specific risk, it is limited to childhood malignancy, but, as described above, for each 10 mGy exposure, theoretical projections based on a LNT assumption indicate a likely maximum risk of one additional case of childhood cancer (not death) per 1700 exposures of 10 mGy.¹⁷ This must be balanced against the benefit of the imaging or treatment in terms of management of the maternal condition.

Conclusion

Audit data from the United States have shown that imaging in the pregnant population is increasing in both the number of examinations performed and the number of patients being imaged, with the greatest increase being CT scans.⁵³ When the mother’s condition necessitates diagnostic radiation it is necessary to balance the risks of the procedure with the benefits to be gained. As almost all diagnostic imaging involves doses below the 50 mGy threshold, clinically indicated investigations should not be withheld because of concerns regarding fetal radiation exposure. In all cases it is essential that the radiologist, radiographer or nuclear physician be informed that the woman is pregnant so that they may make appropriate provisions. If direct pelvic or abdominal radiation is considered necessary during pregnancy, the woman should be counselled regarding the benefits and risks of the procedure stressing the very low incidence of complications and the importance of the information to be derived. The ‘as low as reasonably possible’ principle should apply to both patients and occupational exposures to minimise radiation exposure at all times.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Ethical approval

Not applicable

Informed consent

Not applicable

Guarantor

SL

Contributorship

SL is the sole author.