Seminars in Fetal & Neonatal Medicine 20 (2015) 130e137
Contents lists available at ScienceDirect
Seminars in Fetal & Neonatal Medicine journal homepage: www.elsevier.com/locate/siny
Review
Non-central nervous system fetal magnetic resonance imaging Elspeth Whitby a, *, Peter Wright b a b
Academic Unit of Reproductive and Developmental Medicine, Sheffield, UK Medical Imaging and Medical Physics, Radiology, Royal Hallamshire Hospital, Sheffield, UK
s u m m a r y Keywords: Fetal MRI EXIT Placenta Fetus
Fetal magnetic resonance imaging (MRI) is currently offered in a limited number of centers but is predominantly used for suspected fetal central nervous system abnormalities. This article concentrates on the role of the different imaging sequences and their value to clinical practice. It also discusses the future of fetal MRI. Crown Copyright © 2015 Published by Elsevier Ltd. All rights reserved.
1. Introduction Magnetic resonance imaging (MRI) of the fetal central nervous system (CNS) was first suggested in 1983 [1]. Its success was limited by fetal motion. In France, curare was injected into the umbilical cord to effectively paralyze the fetus for the duration of the scan [2]. In the USA, high doses of maternal sedation were preferred [3,4]. In the UK, fetal MRI did not really start until the late 1990s when ultrafast sequences were developed, effectively “freezing” fetal motion without any intervention [5]. A flurry of research followed from a small number of centers [6e9]. Initially, all imaging was T2weighted (T2W) and this has remained the workhorse sequence of fetal MRI. The research was influenced by local laws pertaining to termination of pregnancy. In countries where terminations are only allowed in early pregnancy the imaging was done earlier in gestation e for example, Italy, where groups have published normative data prior to 24 weeks' gestational age [10]. Others, such as France, have provided more data from 24 weeks to birth and are gradually adding growth curves for earlier gestational ages [11]. There are numerous reviews [12e14] and research papers [15e17] on the value of fetal MRI for CNS abnormalities. A recent systematic review is an ideal starting point [18]. Movement into clinical practice has been gradual but by the early 2000s there was a smattering of centers in the UK offering MRI evaluation of the fetal CNS, often provided by neuroradiologists or paediatric neuroradiologists, and imaging of other areas of
* Corresponding author. Address: Academic Unit of Reproductive and Developmental Medicine, 4th floor, Jessop Wing, Tree Root Walk, Sheffield S10 2SF, UK. Tel.: þ44 (0)114 2261081. E-mail address: e.whitby@sheffield.ac.uk (E. Whitby). http://dx.doi.org/10.1016/j.siny.2015.04.001 1744-165X/Crown Copyright © 2015 Published by Elsevier Ltd. All rights reserved.
the fetal body were essentially neglected. A few centers developed their services because of an interest by obstetric radiologists, and these centers currently image a wider spectrum of pathology. With respect to fetal CNS imaging, research centers are looking beyond diagnoses based on gross structure and now examine CNS development using spectroscopy and tractography techniques. These are technically more challenging and often time-consuming. The success rate is low but there is the potential to diagnose the more complex abnormalities in greater detail. This may give parents more information upon which to base clinical decisions. 1.1. Current UK service and best practice In the UK currently there are enough centers to provide adequate service for the population if patients are willing to travel, and if clinicians are aware of the available services and know whom to contact. We have not yet developed a robust network and referral pattern. Also needing further refinement are the imaging parameters and sequences best used for different pathologies. Fetal MRI for all body-regions should be an adjunct to ultrasound and will remain a targeted examination, following an abnormal ultrasound examination, not a screening tool. However, all the information on the image should be used, so that, even if not formally examined, the placenta, uterine cavity, and other areas of the fetus that appear in the images should be given some consideration. Centers with easy access to good quality fetal MRI will use it more frequently and often supplant a repeat ultrasound examination. Fetal MRI should be used as part of the multi-disciplinary team (MDT) approach to patient management. Most centers agree that the workhorse sequence is a T2weighted (T2W) steady-state fast spin echo (SSFSE), in three orthogonal planes of the body area of concern. Most centers include
E. Whitby, P. Wright / Seminars in Fetal & Neonatal Medicine 20 (2015) 130e137
a T1W image in at least one orientation. Good quality images can usually be obtained within 20e30 min. Other useful sequences include diffusion-weighted imaging (DWI) and balanced gradient echo. A three-dimensional T2W, balanced gradient echo is also useful to allow reformatting; however, the spatial resolution remains low, and currently these are not good enough to replace single plane T2W SSFSE images, especially at 1.5 Tesla (T). The choice of coil depends on local availability. Most centers use a surface array body coil placed over the area to be imaged; cardiac coils can also be used but require careful positioning and may need re-positioning during the scan, especially in later gestation, and this will increase the examination time. In practice, we have observed that the fetus settles with repeated similar noises, so it is advisable to do all T2W sequences first, then moving on to T1W sequences and finishing with the DWI, as this is the loudest and longest sequence. The resting state of the mother is also important as this affects the quality of the images. A relaxed mother will result in a relaxed, quiet fetus whereas a stressed mother causes increased fetal movements. Explaining the procedure and making the mother comfortable both increase the image quality. The woman should be placed feet-first into the scanner to avoid claustrophobia. If possible, she should be imaged in the most comfortable position for her; the semi-decubitus position will avoid hypotension from the uterus compressing the inferior vena cava and preventing venous return. Certain scanners are designed such that a semi-decubitus position is difficult because of the nature of the surface body coil, and if the woman is able to lie supine, she can be imaged in this position. Pads and cushions should be used to support the mother's body and to encourage her to relax. Women who return for a follow-up scan are often more relaxed than for the first scan, and this improves the imaging quality. Some American groups provide juice and cookies 15 min prior to the examination, whereas some French groups advocate nothing by mouth for a few hours prior to the examination (E. Whitby, personal communication). Each set of images is based on the previously acquired set. This is a non-conventional approach to imaging for the MR radiographers, who are used to planning from several image sets to obtain a good anatomical imaging plane. As the fetus is in continual motion, speedy positioning from the last sequence is the most reliable way to obtain good images. Most radiographers can adapt to this technique, and, with experience, the speed at which they can select the sequences and position for them increases, enhancing the quality of the images.
2. Sequences for fetal MRI
131
Fig. 1. Exomphalus, 20 weeks' gestation. T2-weighted image, sagittal section. All the liver *, most of the stomach (arrow) and some of the bowel are outside the fetal abdomen.
performed at 34e36 weeks' gestation to avoid the need for postnatal imaging prior to surgery. The T2W sequence is useful for all types of lung pathology. Fetal MRI can securely differentiate between congenital cystic pulmonary lesions (Fig. 2a) and a CDH with just the bowel in the thorax (Fig. 2c and d). It is also extremely useful for imaging any cystic lung lesion to define the cyst and its origin, e.g. neuroenteric, bronchogenic (Fig. 2b), or duplication cyst. Infradiaphragmatic pulmonary sequestrations can be assessed; they are the same signal intensity as the lung, usually located adjacent to the diaphragm and are often wedge-shaped. They are easier to define as separate from the adjacent organs with fetal MRI than with ultrasound. There are still limitations with the T2W sequence even for structural information e the heart is seen as a dark signal area, and there is little detail in fetal bones. Pathologic diagnosis is often enhanced by additional sequences.
2.1. T2-weighted imaging
2.2. T1-weighted imaging
The T2W image provides the most anatomical detail in all areas of the fetus. On this image, the brain, lungs, kidneys, bladder, stomach, small and large bowel, liver, and gall bladder are clearly seen. Other structures are sometimes seen but are not always evident, and these include the fetal adrenal glands and pancreas. The detail seen in the fetal brain is aided by the inherent contrast difference between the parenchyma (dark) and the cerebrospinal fluid (CSF) (bright). Structural abnormalities are detected on this sequence and the location of organs with associated defects can also be determined such as the contents of an exomphalos (Fig. 1), the contents of the hemithorax in congenital diaphragmatic hernia (CDH), displacement by tumors, and so forth. The anatomical detail is often useful to plan postnatal surgery and to discuss this with the parents prior to delivery. In a few centers, a fetal MRI is
The detail in the T1W image is limited, partly because of low spatial resolution and partly because of little inherent T1 contrast in the tissues secondary to their large water content, leading some people to underestimate its clinical value. The normal thyroid tissue is evident as two small tiny areas of brightness in the fetal neck. The normal liver is bright on T1, and this is valuable to locate small areas of liver in CDH (Fig. 2d). Meconium is also bright on T1W imaging, raising possibilities for fetal MRI to be used to diagnose bowel pathology. However, be aware that fetal fat is not bright on T1 until mid-third trimester and this is often a source of error, especially in diagnosing lipomas associated with corpus callosal abnormalities and spinal defects. Hemorrhage is bright on T1W imaging. It is useful to acquire a T1W image in at least one orientation for every fetal MRI scan
132
E. Whitby, P. Wright / Seminars in Fetal & Neonatal Medicine 20 (2015) 130e137
Fig. 2. Lung pathologies, 20 weeks' gestational age. (a) Congenital cystic pulmonary malformation, numerous small cysts and high signal lung on T2-weighted (T2W) images due to increased fluid content of the lung tissue. (b) Simple lung cyst, high signal area on T2W images. (c) Congenital diaphragmatic hernia. T2W images: the liver is seen in the hemithorax. (d) T1W images: the liver is high signal and clearly seen in the hemithorax as is the meconium-filled bowel.
performed, so that the reporting clinician understands the normal appearances and can readily detect abnormalities. Imaged in the coronal orientation the thyroid gland will have the typical butterfly pattern (Fig. 3a). Similarly a fetal goiter will be uniformly bright on T1 and have the typical appearance but be much larger than normal (Fig. 3b). It is important to assess for the presence of a patent airway. Other neck pathology looks very different. The typical fetal teratoma is of mixed signal characteristic on T2W imaging and is usually low to intermediate mixed signal on T1 unless it contains hemorrhage (Fig. 3c, d). An unusual mimic of the fetal goiter on T2 is a rhabdosarcoma, but this is low signal on T1 (Fig. 3e). Again, the T2W sequence is essential for establishing the state of the airway. Cystic lesions in the neck will have a high T2 signal and low T1 signal, matching the signal characteristics of amniotic fluid unless the cyst contains fluid with increased protein or hemorrhage, where the T2 (Fig. 3f) and T1 intensities will be intermediate.
The liver is high signal on T1W imaging, which helps to locate it in CDH (Fig. 2d) [19,20]. The echogenicity of the liver is similar to that of the lung with ultrasound, making location of the liver difficult. It is generally accepted that, in CDH, if the liver is in the thorax, the outlook is poor. It has become evident from fetal MR that in general this is the case, but in cases where only a small amount of the left lobe of the liver is in the thorax, the outlook depends more on the lung volume. Imaging of the bowel is helped by visualization of meconium as high signal on T1W images (Fig. 4a) in fetal bowel, initially in the rectum from the second trimester and then progressively more proximally with increasing gestational age. The length of the meconium signal relates to gestational age but normative values have yet to be established [21]. In addition, the width of the meconium at each level may help detect pathologic states (Fig. 4b). For example, with a sacrococcygeal teratoma the tumor can
E. Whitby, P. Wright / Seminars in Fetal & Neonatal Medicine 20 (2015) 130e137
133
Fig. 3. Neck pathologies. (a) Normal thyroid gland [21 weeks' gestational age (GA)] is seen as a small area of high signal on T1-weighted (T1W) coronal images (arrows). (b) Fetal goiter (enlarged thyroid) (21 weeks' GA). Normal signal intensity for thyroid tissue and homogeneous but much larger than normal (*). (i) T1W coronal images. (ii) T2W images (note the patent, fluid-filled airway, arrow). (c) Teratoma (34 weeks' GA): predominantly cystic, mixed heterogeneity and irregular shape. T2W sagittal images. (d) Teratoma (20 weeks' GA): predominantly solid, mixed heterogeneity and irregular shape. T2W sagittal images. (e) Teratoma: same patient as above but at 30 weeks' gestation; severe hydrops has developed. (f) Rhabdosarcoma of the neck. Homogeneous signal intensity but different distribution and shape to a fetal goiter. (i) T2W sagittal images. These are low signal on T1W images (ii). (g) Fluid-filled complex lymphangiohaemangioma (20 weeks' GA). Note the high-signal fluid areas (arrow) and the mid-signal haemorrhagic and vascular areas (*).
displace or compress the rectum, and this is detectable by changes in the meconium pattern. Ileal atresia is seen as a dilated fluid-filled loop of proximal bowel on MRI, and absence of meconium on fetal MR images may indicate a microcolon. Sometimes this may be a temporary problem with a dilated loop on fetal MRI (Fig. 4b) and postnatal imaging (Fig. 4c) that resolves without intervention. The lack of meconium alerts the clinician to potential problems at birth that might otherwise be unsuspected [22,23].
2.3. Diffusion-weighted imaging (DWI) DWI has been used for imaging the brain. The diffusion pattern is gestational age dependent, and to ensure accurate interpretation, it is advisable always to obtain a DWI image so that the reporting clinician is aware of the normal appearance. The pattern is usually symmetrical. It is useful to detect small areas of hemorrhage, infarction, and damage from infection [24,25].
134
E. Whitby, P. Wright / Seminars in Fetal & Neonatal Medicine 20 (2015) 130e137
Fig. 4. T1-weighted images of fetal bowel. (a) Normal sagittal orientation. Note the high signal from meconium in the rectum (arrow). (b) T1 Weighted Coronal MRI scan demonstrating a small loop of dilated bowel and no meconium in the distal bowel at the level of the rectum (arrow). (c) Abdominal x ray post delivery of the baby in Fig. 4b. Note the lack of air in the distal bowel. This resolved spontaneously in the first 24 hours after birth.
E. Whitby, P. Wright / Seminars in Fetal & Neonatal Medicine 20 (2015) 130e137
135
Fig. 5. Normal diffusion-weighted imaging of the fetal kidneys (20 weeks' gestational age). Note the high signal from the parenchyma.
DWI is also extremely useful to locate the kidneys. Normally functioning parenchyma has a bright DWI signal and conforms to the renal shape (Fig. 5). This is helpful to locate ectopic kidneys, small kidneys, and to confirm bilateral renal agenesis [26,27]. 2.4. Balanced gradient echo The balanced gradient echo sequence is useful to detect the edges of organs and lesions. The detail of the organ is lost in this sequence but the edges are enhanced. This is useful for cysts in the brain, abdomen, and lungs. It helps delineate skin and membrane and is useful in spina bifida, exomphalos, etc., as it helps provide the surgeon with additional information prior to delivery. It is currently an underused sequence that has the potential to help image other areas and pathology in the fetus, especially heart (Fig. 6) and bones (Fig. 7).
Fig. 6. Balanced gradient echo sequences showing the fetal heart (arrow) e an area yet to be researched in any detail.
and should be examined thoroughly. Understanding normal placental changes with gestation will allow the reporting clinician to detect an abnormality. Recently, several reports of placental abnormalities defined by MRI have appeared [33,34]. In addition,
2.5. Susceptibility-weighted imaging This has potential to provide detail on fetal bones, especially in spina bifida, but requires a longer time and is thus more affected by fetal movement [28]. 2.6. Spectroscopy Spectroscopy is not yet fully researched and remains of restricted use in clinical practice. There are a few reports suggesting the potential use of the technique but normal values are still to be established [29e32]. This technique may be of value in differentiating fetal stress at any point during the pregnancy, and abnormalities on spectroscopy may indicate potential problems after birth but it is an area that requires further exploration. The value extends beyond the CNS, as lactate can be detected in the amniotic fluid and in the brain of the sick fetus. 3. Further applications 3.1. The placenta It should not be forgotten that the placenta is an important organ throughout gestation. It can also be seen on numerous images
Fig. 7. Balanced gradient echo sequences showing the fetal bone (arrow) e an area of potential clinical benefit but requiring further research.
136
E. Whitby, P. Wright / Seminars in Fetal & Neonatal Medicine 20 (2015) 130e137
an increasing number of pregnancies are complicated by placental adhesive disorder, and clinicians are requesting imaging for the diagnosis and to aid surgical planning [35,36]. 3.2. Other anatomical areas Other anatomical areas are still under-imaged with fetal MRI and additional research is needed. The fetus with abnormal bones is often referred for MRI to assess other associated abnormalities. New sequences have shown potential, especially susceptibilityweighted imaging, which gives a “black bone” appearance [28]. There are a few reports describing the use of MRI to image the fetal heart [37e39]. 3.3. Planning surgery Any pathology that compresses or obstructs the fetal airway increases the risk for hypoxia at birth. To safely deliver an affected fetus, the oxygen supply needs to be maintained. Such pregnancies may be delivered using the EXIT (ex-utero intrapartum treatment) procedure, where the head and neck are delivered and operated on while maintaining utero-placental circulation. These are technically challenging deliveries, and the more detailed knowledge of the anatomy, the better the outcome ought to be. Fetal MRI can provide detail beyond ultrasound and plays a central role in these cases [40,41]. Imaging plays a central role in deciding on management and the clinician responsible for the imaging should be part of the MDT [41].
It will be essential that the reporting clinicians know the outcomes of the cases they see in order to maintain standards and increase their own level of experience. This requires a close working collaboration with neonatologists, surgeons, obstetricians, genetic counsellors, pediatric radiologists and pediatric pathologists.
5. The future of fetal MRI In the last decade, fetal MRI has gone from a research tool to an established clinical practice. It has advanced our understanding of pathology and the natural history of some diseases. For example, fetuses with cystic pulmonary malformations have a favourable prognosis unless there is associated ascites. Some reports suggest that the ascites may resolve, improving the outlook [43]. Networking will increase the speed at which we increase our understanding and improve patient imaging and patient care. Timing of imaging is key. Most centers like to image as late in gestation as possible, but the most useful time is immediately following identification of the problem on ultrasound. Earlier gestation fetal MRI will depend on local practice and on what degree of certainty or what pathology is required. The UK national guidelines (Medicine and Healthcare Products Regulatory Agency, November 2014) advise imaging of the fetus at any age whilst remaining in normal mode at both 1.5 and 3 T. Early second trimester fetal MRI may be beneficial. For example, the parents of a fetus with CDH might consider termination if the prognosis is poor. The main problem with imaging in early pregnancy is that there is little experience, so care needs to be taken in interpretation of images.
4. Current practice In the UK, centers are limiting services for various reasons, including the limited experience of available radiologists. Some offer a selective service or have limited experience at early gestational ages and only scan in the third trimester. Knowledge of the embryology and normal imaging appearances at each gestational age is essential for accurate diagnosis [42]. Imaging in early gestation is technically more challenging but can often answer a clinical question and guide management, even if detailed imaging is not possible. Experienced centers offer a larger range of services, including imaging all areas of the fetus. However, this has resulted in a large variation in the services offered throughout the UK. Methods to improve the service that are currently being established include: (a) Establishing a large network within the UK to support all active radiologists and fetal medicine experts. This would allow the major tertiary referral centres with large fetalematernal units to establish an imaging service supported by the more established groups. The network would not only provide clinical support but also act as a forum to discuss difficult cases, improve image quality, and even provide cross-coverage. With the current means of electronic communication and transfer of images, this would be easily achievable and improve the current imaging service. (b) Establishing a central group to develop imaging guidelines to ensure good image quality but to be flexible enough for centers still to be able to develop their own imaging sequences for the benefit of all. This has the added value that image quality will be maintained, and improvements in image quality or development of additional sequences that are of clinical value can be widely disseminated to benefit all patients at little cost to the NHS.
Practice points Fetal MRI is developing rapidly but areas other than the CNS have not been fully evaluated. Fetal MRI is a useful adjunct to sonography in some cases. Fetal MRI of the heart and bones is still investigational.
Research directions The role of fetal MRI for chest pathology. The role of fetal MRI for abdominal pathology. The role of MRI in placental disorders.
Conflict of interest statement None declared. Funding sources None. References [1] Smith FW, Adam AH, Phillips WD. NMR imaging in pregnancy. Lancet 1983 Jan 1;1(8314e5):61e2. PubMed PMID: 6129387. [2] Girard N, Raybaud C, Dercole C, Boubli L, Chau C, Cahen S, et al. In vivo MRI of the fetal brain. Neuroradiology 1993;35(6):431e6. PubMed PMID: 8377914. [3] Levine D, Barnes PD, Madsen JR, Li W, Edelman RR. Fetal central nervous system anomalies: MR imaging augments sonographic diagnosis. Radiology 1997 Sep;204(3):635e42. PubMed PMID: 9280237.
E. Whitby, P. Wright / Seminars in Fetal & Neonatal Medicine 20 (2015) 130e137 [4] Levine D, Hatabu H, Gaa J, Atkinson MW, Edelman RR. Fetal anatomy revealed with fast MR sequences. AJR Am J Roentgenol 1996 Oct;167(4):905e8. PubMed PMID: 8819380. [5] Levine D, Barnes PD, Sher S, Semelka RC, Li W, McArdle CR, et al. Fetal fast MR imaging: reproducibility, technical quality, and conspicuity of anatomy. Radiology 1998 Feb;206(2):549e54. PubMed PMID: 9457211. [6] Levine D. MR imaging of fetal central nervous system abnormalities. Brain Cogn 2002 Dec;50(3):432e48. PubMed PMID: 12480488. [7] Hubbard AM, Harty MP, States LJ. A new tool for prenatal diagnosis: ultrafast fetal MRI. Semin Perinatol 1999 Dec;23(6):437e47. PubMed PMID: 10630540. [8] Hubbard AM. Ultrafast fetal MRI and prenatal diagnosis. Semin Pediatr Surg 2003 Aug;12(3):143e53. PubMed PMID: 12961108. [9] Whitby E, Paley MN, Davies N, Sprigg A, Griffiths PD. Ultrafast magnetic resonance imaging of central nervous system abnormalities in utero in the second and third trimester of pregnancy: comparison with ultrasound. BJOG 2001 May;108(5):519e26. PubMed PMID: 11368139. [10] Parazzini C, Righini A, Rustico M, Consonni D, Triulzi F. Prenatal magnetic resonance imaging: brain normal linear biometric values below 24 gestational weeks. Neuroradiology 2008 Oct;50(10):877e83. PubMed PMID: 18563404. [11] Tilea B, Alberti C, Adamsbaum C, Armoogum P, Oury JF, Cabrol D, et al. Cerebral biometry in fetal magnetic resonance imaging: new reference data. Ultrasound Obstet Gynecol 2009 Feb;33(2):173e81. PubMed PMID: 19172662. [12] Coakley FV, Glenn OA, Qayyum A, Barkovich AJ, Goldstein R, Filly RA. Fetal MRI: a developing technique for the developing patient. AJR Am J Roentgenol 2004 Jan;182(1):243e52. PubMed PMID: 14684546. [13] Glenn OA, Coakley FV. MRI of the fetal central nervous system and body. Clin Perinatol 2009 Jun;36(2):273e300. viii. PubMed PMID: 19559321. [14] Grant JA. Fetal MRI. Pediatr Neurosurg 2000 Sep;33(3):167. PubMed PMID: 11096365. [15] Prayer D, Kasprian G, Krampl E, Ulm B, Witzani L, Prayer L, et al. MRI of normal fetal brain development. Eur J Radiol 2006 Feb;57(2):199e216. PubMed PMID: 16413984. [16] Prayer D, Brugger PC, Kasprian G, Witzani L, Helmer H, Dietrich W, et al. MRI of fetal acquired brain lesions. Eur J Radiol 2006 Feb;57(2):233e49. PubMed PMID: 16413156. [17] Simon EM. MRI of the fetal spine. Pediatr Radiol 2004 Sep;34(9):712e9. PubMed PMID: 15316691. [18] Whitby E. personal communication. 2014. [19] Mayer S, Klaritsch P, Petersen S, Done E, Sandaite I, Till H, et al. The correlation between lung volume and liver herniation measurements by fetal MRI in isolated congenital diaphragmatic hernia: a systematic review and metaanalysis of observational studies. Prenatal Diagn 2011 Nov;31(11):1086e96. PubMed PMID: 21915885. [20] Terui K, Omoto A, Osada H, Hishiki T, Saito T, Sato Y, et al. Prediction of postnatal outcomes in congenital diaphragmatic hernia using MRI signal intensity of the fetal lung. J Perinatol 2011 Apr;31(4):269e73. PubMed PMID: 21052047. [21] Farhataziz N, Engels JE, Ramus RM, Zaretsky M, Twickler DM. Fetal MRI of urine and meconium by gestational age for the diagnosis of genitourinary and gastrointestinal abnormalities. AJR Am J Roentgenol 2005 Jun;184(6):1891e7. PubMed PMID: 15908548. [22] Zizka J, Elias P, Hodik K, Tintera J, Juttnerova V, Belobradek Z, et al. Liver, meconium, haemorrhage: the value of T1-weighted images in fetal MRI. Pediatr Radiol 2006 Aug;36(8):792e801. PubMed PMID: 16799788. [23] Manganaro L, Saldari M, Bernardo S, Vinci V, Aliberti C, Sollazzo P, et al. Role of magnetic resonance imaging in the prenatal diagnosis of gastrointestinal fetal anomalies. La Radiologia Medica 2014 Oct 28. PubMed PMID: 25348138. [24] Kasprian G, Del Rio M, Prayer D. Fetal diffusion imaging: pearls and solutions. Top Magn Reson Imaging 2010 Dec;21(6):387e94. PubMed PMID: 22158132. [25] Prayer D, Prayer L. Diffusion-weighted magnetic resonance imaging of cerebral white matter development. Eur J Radiol 2003 Mar;45(3):235e43. PubMed PMID: 12595108.
137
[26] Manganaro L, Francioso A, Savelli S, Tomei A, Fierro F, Di Maurizio M, et al. Fetal MRI with diffusion-weighted imaging (DWI) and apparent diffusion coefficient (ADC) assessment in the evaluation of renal development: preliminary experience in normal kidneys. La Radiologia Medica 2009 Apr;114(3):403e13. PubMed PMID: 19381763. [27] Savelli S, Di Maurizio M, Perrone A, Tesei J, Francioso A, Angeletti M, et al. MRI with diffusion-weighted imaging (DWI) and apparent diffusion coefficient (ADC) assessment in the evaluation of normal and abnormal fetal kidneys: preliminary experience. Prenatal Diagn 2007 Dec;27(12):1104e11. PubMed PMID: 17849498. [28] Robinson AJ, Blaser S, Vladimirov A, Drossman D, Chitayat D, Ryan G. Fetal “Black Bone” MRI: Utility in assessment of the fetal spine. Br J Radiol 2014 Dec 11:20140496. PubMed PMID: 25496509. [29] Berger-Kulemann V, Brugger PC, Pugash D, Krssak M, Weber M, Wielandner A, et al. MR spectroscopy of the fetal brain: is it possible without sedation? AJNR Am J Neuroradiol 2013 Feb;34(2):424e31. PubMed PMID: 22821916. [30] Girard N, Gouny SC, Viola A, Le Fur Y, Viout P, Chaumoitre K, et al. Assessment of normal fetal brain maturation in utero by proton magnetic resonance spectroscopy. Magn Reson Med 2006 Oct;56(4):768e75. PubMed PMID: 16964617. [31] Heerschap A, Kok RD, van den Berg PP. Antenatal proton MR spectroscopy of the human brain in vivo. Child's Nerv Syst 2003 Aug;19(7e8):418e21. PubMed PMID: 12811484. [32] Fenton BW, Lin CS, Macedonia C, Schellinger D, Ascher S. The fetus at term: in utero volume-selected proton MR spectroscopy with a breath-hold techniqueea feasibility study. Radiology 2001 May;219(2):563e6. PubMed PMID: 11323489. [33] Dekan S, Linduska N, Kasprian G, Prayer D. MRI of the placenta e a short review. Wien Med Wochenschr 2012 May;162(9e10):225e8. PubMed PMID: 22717878. [34] Masselli G, Gualdi G. MR imaging of the placenta: what a radiologist should know. Abdom Imaging 2012 Jul 15. PubMed PMID: 22797659. [35] Bour L, Place V, Bendavid S, Fargeaudou Y, Portal JJ, Ricbourg A, et al. Suspected invasive placenta: evaluation with magnetic resonance imaging. European Radiology 2014 Dec;24(12):3150e60. PubMed PMID: 25070274. [36] Elhawary TM, Dabees NL, Youssef MA. Diagnostic value of ultrasonography and magnetic resonance imaging in pregnant women at risk for placenta accreta. J Matern Fetal Neonatal Med 2013 Sep;26(14):1443e9. PubMed PMID: 23489020. [37] Votino C, Jani J, Damry N, Dessy H, Kang X, Cos T, et al. Magnetic resonance imaging in the normal fetal heart and in congenital heart disease. Ultrasound Obstet Gynecol 2012 Mar;39(3):322e9. PubMed PMID: 21837757. [38] Yamamura J, Kopp I, Frisch M, Fischer R, Valett K, Hecher K, et al. Cardiac MRI of the fetal heart using a novel triggering method: initial results in an animal model. J Magn Reson Imaging 2012 May;35(5):1071e6. PubMed PMID: 22246623. [39] Manganaro L, Vinci V, Bernardo S, Sollazzo P, Sergi ME, Saldari M, et al. Magnetic resonance imaging of fetal heart: anatomical and pathological findings. J Matern Fetal Neonatal Med 2014 Aug;27(12):1213e9. PubMed PMID: 24102352. [40] MacArthur CJ. Prenatal diagnosis of fetal cervicofacial anomalies. Curr Opin Otolaryngol Head Neck Surg 2012 Dec;20(6):482e90. PubMed PMID: 22913933. [41] Dighe MK, Peterson SE, Dubinsky TJ, Perkins J, Cheng E. EXIT procedure: technique and indications with prenatal imaging parameters for assessment of airway patency. Radiographics 2011 MareApr;31(2):511e26. PubMed PMID: 21415194. [42] Glenn OA. Normal development of the fetal brain by MRI. Semin Perinatol 2009 Aug;33(4):208e19. PubMed PMID: 19631082. [43] Kunisaki SM, Barnewolt CE, Estroff JA, Ward VL, Nemes LP, Fauza DO, et al. Large fetal congenital cystic adenomatoid malformations: growth trends and patient survival. J Pediatr Surg 2007 Feb;42(2):404e10. PubMed PMID: 17270558.
Contents lists available at ScienceDirect
Seminars in Fetal & Neonatal Medicine journal homepage: www.elsevier.com/locate/siny
Review
Non-central nervous system fetal magnetic resonance imaging Elspeth Whitby a, *, Peter Wright b a b
Academic Unit of Reproductive and Developmental Medicine, Sheffield, UK Medical Imaging and Medical Physics, Radiology, Royal Hallamshire Hospital, Sheffield, UK
s u m m a r y Keywords: Fetal MRI EXIT Placenta Fetus
Fetal magnetic resonance imaging (MRI) is currently offered in a limited number of centers but is predominantly used for suspected fetal central nervous system abnormalities. This article concentrates on the role of the different imaging sequences and their value to clinical practice. It also discusses the future of fetal MRI. Crown Copyright © 2015 Published by Elsevier Ltd. All rights reserved.
1. Introduction Magnetic resonance imaging (MRI) of the fetal central nervous system (CNS) was first suggested in 1983 [1]. Its success was limited by fetal motion. In France, curare was injected into the umbilical cord to effectively paralyze the fetus for the duration of the scan [2]. In the USA, high doses of maternal sedation were preferred [3,4]. In the UK, fetal MRI did not really start until the late 1990s when ultrafast sequences were developed, effectively “freezing” fetal motion without any intervention [5]. A flurry of research followed from a small number of centers [6e9]. Initially, all imaging was T2weighted (T2W) and this has remained the workhorse sequence of fetal MRI. The research was influenced by local laws pertaining to termination of pregnancy. In countries where terminations are only allowed in early pregnancy the imaging was done earlier in gestation e for example, Italy, where groups have published normative data prior to 24 weeks' gestational age [10]. Others, such as France, have provided more data from 24 weeks to birth and are gradually adding growth curves for earlier gestational ages [11]. There are numerous reviews [12e14] and research papers [15e17] on the value of fetal MRI for CNS abnormalities. A recent systematic review is an ideal starting point [18]. Movement into clinical practice has been gradual but by the early 2000s there was a smattering of centers in the UK offering MRI evaluation of the fetal CNS, often provided by neuroradiologists or paediatric neuroradiologists, and imaging of other areas of
* Corresponding author. Address: Academic Unit of Reproductive and Developmental Medicine, 4th floor, Jessop Wing, Tree Root Walk, Sheffield S10 2SF, UK. Tel.: þ44 (0)114 2261081. E-mail address: e.whitby@sheffield.ac.uk (E. Whitby). http://dx.doi.org/10.1016/j.siny.2015.04.001 1744-165X/Crown Copyright © 2015 Published by Elsevier Ltd. All rights reserved.
the fetal body were essentially neglected. A few centers developed their services because of an interest by obstetric radiologists, and these centers currently image a wider spectrum of pathology. With respect to fetal CNS imaging, research centers are looking beyond diagnoses based on gross structure and now examine CNS development using spectroscopy and tractography techniques. These are technically more challenging and often time-consuming. The success rate is low but there is the potential to diagnose the more complex abnormalities in greater detail. This may give parents more information upon which to base clinical decisions. 1.1. Current UK service and best practice In the UK currently there are enough centers to provide adequate service for the population if patients are willing to travel, and if clinicians are aware of the available services and know whom to contact. We have not yet developed a robust network and referral pattern. Also needing further refinement are the imaging parameters and sequences best used for different pathologies. Fetal MRI for all body-regions should be an adjunct to ultrasound and will remain a targeted examination, following an abnormal ultrasound examination, not a screening tool. However, all the information on the image should be used, so that, even if not formally examined, the placenta, uterine cavity, and other areas of the fetus that appear in the images should be given some consideration. Centers with easy access to good quality fetal MRI will use it more frequently and often supplant a repeat ultrasound examination. Fetal MRI should be used as part of the multi-disciplinary team (MDT) approach to patient management. Most centers agree that the workhorse sequence is a T2weighted (T2W) steady-state fast spin echo (SSFSE), in three orthogonal planes of the body area of concern. Most centers include
E. Whitby, P. Wright / Seminars in Fetal & Neonatal Medicine 20 (2015) 130e137
a T1W image in at least one orientation. Good quality images can usually be obtained within 20e30 min. Other useful sequences include diffusion-weighted imaging (DWI) and balanced gradient echo. A three-dimensional T2W, balanced gradient echo is also useful to allow reformatting; however, the spatial resolution remains low, and currently these are not good enough to replace single plane T2W SSFSE images, especially at 1.5 Tesla (T). The choice of coil depends on local availability. Most centers use a surface array body coil placed over the area to be imaged; cardiac coils can also be used but require careful positioning and may need re-positioning during the scan, especially in later gestation, and this will increase the examination time. In practice, we have observed that the fetus settles with repeated similar noises, so it is advisable to do all T2W sequences first, then moving on to T1W sequences and finishing with the DWI, as this is the loudest and longest sequence. The resting state of the mother is also important as this affects the quality of the images. A relaxed mother will result in a relaxed, quiet fetus whereas a stressed mother causes increased fetal movements. Explaining the procedure and making the mother comfortable both increase the image quality. The woman should be placed feet-first into the scanner to avoid claustrophobia. If possible, she should be imaged in the most comfortable position for her; the semi-decubitus position will avoid hypotension from the uterus compressing the inferior vena cava and preventing venous return. Certain scanners are designed such that a semi-decubitus position is difficult because of the nature of the surface body coil, and if the woman is able to lie supine, she can be imaged in this position. Pads and cushions should be used to support the mother's body and to encourage her to relax. Women who return for a follow-up scan are often more relaxed than for the first scan, and this improves the imaging quality. Some American groups provide juice and cookies 15 min prior to the examination, whereas some French groups advocate nothing by mouth for a few hours prior to the examination (E. Whitby, personal communication). Each set of images is based on the previously acquired set. This is a non-conventional approach to imaging for the MR radiographers, who are used to planning from several image sets to obtain a good anatomical imaging plane. As the fetus is in continual motion, speedy positioning from the last sequence is the most reliable way to obtain good images. Most radiographers can adapt to this technique, and, with experience, the speed at which they can select the sequences and position for them increases, enhancing the quality of the images.
2. Sequences for fetal MRI
131
Fig. 1. Exomphalus, 20 weeks' gestation. T2-weighted image, sagittal section. All the liver *, most of the stomach (arrow) and some of the bowel are outside the fetal abdomen.
performed at 34e36 weeks' gestation to avoid the need for postnatal imaging prior to surgery. The T2W sequence is useful for all types of lung pathology. Fetal MRI can securely differentiate between congenital cystic pulmonary lesions (Fig. 2a) and a CDH with just the bowel in the thorax (Fig. 2c and d). It is also extremely useful for imaging any cystic lung lesion to define the cyst and its origin, e.g. neuroenteric, bronchogenic (Fig. 2b), or duplication cyst. Infradiaphragmatic pulmonary sequestrations can be assessed; they are the same signal intensity as the lung, usually located adjacent to the diaphragm and are often wedge-shaped. They are easier to define as separate from the adjacent organs with fetal MRI than with ultrasound. There are still limitations with the T2W sequence even for structural information e the heart is seen as a dark signal area, and there is little detail in fetal bones. Pathologic diagnosis is often enhanced by additional sequences.
2.1. T2-weighted imaging
2.2. T1-weighted imaging
The T2W image provides the most anatomical detail in all areas of the fetus. On this image, the brain, lungs, kidneys, bladder, stomach, small and large bowel, liver, and gall bladder are clearly seen. Other structures are sometimes seen but are not always evident, and these include the fetal adrenal glands and pancreas. The detail seen in the fetal brain is aided by the inherent contrast difference between the parenchyma (dark) and the cerebrospinal fluid (CSF) (bright). Structural abnormalities are detected on this sequence and the location of organs with associated defects can also be determined such as the contents of an exomphalos (Fig. 1), the contents of the hemithorax in congenital diaphragmatic hernia (CDH), displacement by tumors, and so forth. The anatomical detail is often useful to plan postnatal surgery and to discuss this with the parents prior to delivery. In a few centers, a fetal MRI is
The detail in the T1W image is limited, partly because of low spatial resolution and partly because of little inherent T1 contrast in the tissues secondary to their large water content, leading some people to underestimate its clinical value. The normal thyroid tissue is evident as two small tiny areas of brightness in the fetal neck. The normal liver is bright on T1, and this is valuable to locate small areas of liver in CDH (Fig. 2d). Meconium is also bright on T1W imaging, raising possibilities for fetal MRI to be used to diagnose bowel pathology. However, be aware that fetal fat is not bright on T1 until mid-third trimester and this is often a source of error, especially in diagnosing lipomas associated with corpus callosal abnormalities and spinal defects. Hemorrhage is bright on T1W imaging. It is useful to acquire a T1W image in at least one orientation for every fetal MRI scan
132
E. Whitby, P. Wright / Seminars in Fetal & Neonatal Medicine 20 (2015) 130e137
Fig. 2. Lung pathologies, 20 weeks' gestational age. (a) Congenital cystic pulmonary malformation, numerous small cysts and high signal lung on T2-weighted (T2W) images due to increased fluid content of the lung tissue. (b) Simple lung cyst, high signal area on T2W images. (c) Congenital diaphragmatic hernia. T2W images: the liver is seen in the hemithorax. (d) T1W images: the liver is high signal and clearly seen in the hemithorax as is the meconium-filled bowel.
performed, so that the reporting clinician understands the normal appearances and can readily detect abnormalities. Imaged in the coronal orientation the thyroid gland will have the typical butterfly pattern (Fig. 3a). Similarly a fetal goiter will be uniformly bright on T1 and have the typical appearance but be much larger than normal (Fig. 3b). It is important to assess for the presence of a patent airway. Other neck pathology looks very different. The typical fetal teratoma is of mixed signal characteristic on T2W imaging and is usually low to intermediate mixed signal on T1 unless it contains hemorrhage (Fig. 3c, d). An unusual mimic of the fetal goiter on T2 is a rhabdosarcoma, but this is low signal on T1 (Fig. 3e). Again, the T2W sequence is essential for establishing the state of the airway. Cystic lesions in the neck will have a high T2 signal and low T1 signal, matching the signal characteristics of amniotic fluid unless the cyst contains fluid with increased protein or hemorrhage, where the T2 (Fig. 3f) and T1 intensities will be intermediate.
The liver is high signal on T1W imaging, which helps to locate it in CDH (Fig. 2d) [19,20]. The echogenicity of the liver is similar to that of the lung with ultrasound, making location of the liver difficult. It is generally accepted that, in CDH, if the liver is in the thorax, the outlook is poor. It has become evident from fetal MR that in general this is the case, but in cases where only a small amount of the left lobe of the liver is in the thorax, the outlook depends more on the lung volume. Imaging of the bowel is helped by visualization of meconium as high signal on T1W images (Fig. 4a) in fetal bowel, initially in the rectum from the second trimester and then progressively more proximally with increasing gestational age. The length of the meconium signal relates to gestational age but normative values have yet to be established [21]. In addition, the width of the meconium at each level may help detect pathologic states (Fig. 4b). For example, with a sacrococcygeal teratoma the tumor can
E. Whitby, P. Wright / Seminars in Fetal & Neonatal Medicine 20 (2015) 130e137
133
Fig. 3. Neck pathologies. (a) Normal thyroid gland [21 weeks' gestational age (GA)] is seen as a small area of high signal on T1-weighted (T1W) coronal images (arrows). (b) Fetal goiter (enlarged thyroid) (21 weeks' GA). Normal signal intensity for thyroid tissue and homogeneous but much larger than normal (*). (i) T1W coronal images. (ii) T2W images (note the patent, fluid-filled airway, arrow). (c) Teratoma (34 weeks' GA): predominantly cystic, mixed heterogeneity and irregular shape. T2W sagittal images. (d) Teratoma (20 weeks' GA): predominantly solid, mixed heterogeneity and irregular shape. T2W sagittal images. (e) Teratoma: same patient as above but at 30 weeks' gestation; severe hydrops has developed. (f) Rhabdosarcoma of the neck. Homogeneous signal intensity but different distribution and shape to a fetal goiter. (i) T2W sagittal images. These are low signal on T1W images (ii). (g) Fluid-filled complex lymphangiohaemangioma (20 weeks' GA). Note the high-signal fluid areas (arrow) and the mid-signal haemorrhagic and vascular areas (*).
displace or compress the rectum, and this is detectable by changes in the meconium pattern. Ileal atresia is seen as a dilated fluid-filled loop of proximal bowel on MRI, and absence of meconium on fetal MR images may indicate a microcolon. Sometimes this may be a temporary problem with a dilated loop on fetal MRI (Fig. 4b) and postnatal imaging (Fig. 4c) that resolves without intervention. The lack of meconium alerts the clinician to potential problems at birth that might otherwise be unsuspected [22,23].
2.3. Diffusion-weighted imaging (DWI) DWI has been used for imaging the brain. The diffusion pattern is gestational age dependent, and to ensure accurate interpretation, it is advisable always to obtain a DWI image so that the reporting clinician is aware of the normal appearance. The pattern is usually symmetrical. It is useful to detect small areas of hemorrhage, infarction, and damage from infection [24,25].
134
E. Whitby, P. Wright / Seminars in Fetal & Neonatal Medicine 20 (2015) 130e137
Fig. 4. T1-weighted images of fetal bowel. (a) Normal sagittal orientation. Note the high signal from meconium in the rectum (arrow). (b) T1 Weighted Coronal MRI scan demonstrating a small loop of dilated bowel and no meconium in the distal bowel at the level of the rectum (arrow). (c) Abdominal x ray post delivery of the baby in Fig. 4b. Note the lack of air in the distal bowel. This resolved spontaneously in the first 24 hours after birth.
E. Whitby, P. Wright / Seminars in Fetal & Neonatal Medicine 20 (2015) 130e137
135
Fig. 5. Normal diffusion-weighted imaging of the fetal kidneys (20 weeks' gestational age). Note the high signal from the parenchyma.
DWI is also extremely useful to locate the kidneys. Normally functioning parenchyma has a bright DWI signal and conforms to the renal shape (Fig. 5). This is helpful to locate ectopic kidneys, small kidneys, and to confirm bilateral renal agenesis [26,27]. 2.4. Balanced gradient echo The balanced gradient echo sequence is useful to detect the edges of organs and lesions. The detail of the organ is lost in this sequence but the edges are enhanced. This is useful for cysts in the brain, abdomen, and lungs. It helps delineate skin and membrane and is useful in spina bifida, exomphalos, etc., as it helps provide the surgeon with additional information prior to delivery. It is currently an underused sequence that has the potential to help image other areas and pathology in the fetus, especially heart (Fig. 6) and bones (Fig. 7).
Fig. 6. Balanced gradient echo sequences showing the fetal heart (arrow) e an area yet to be researched in any detail.
and should be examined thoroughly. Understanding normal placental changes with gestation will allow the reporting clinician to detect an abnormality. Recently, several reports of placental abnormalities defined by MRI have appeared [33,34]. In addition,
2.5. Susceptibility-weighted imaging This has potential to provide detail on fetal bones, especially in spina bifida, but requires a longer time and is thus more affected by fetal movement [28]. 2.6. Spectroscopy Spectroscopy is not yet fully researched and remains of restricted use in clinical practice. There are a few reports suggesting the potential use of the technique but normal values are still to be established [29e32]. This technique may be of value in differentiating fetal stress at any point during the pregnancy, and abnormalities on spectroscopy may indicate potential problems after birth but it is an area that requires further exploration. The value extends beyond the CNS, as lactate can be detected in the amniotic fluid and in the brain of the sick fetus. 3. Further applications 3.1. The placenta It should not be forgotten that the placenta is an important organ throughout gestation. It can also be seen on numerous images
Fig. 7. Balanced gradient echo sequences showing the fetal bone (arrow) e an area of potential clinical benefit but requiring further research.
136
E. Whitby, P. Wright / Seminars in Fetal & Neonatal Medicine 20 (2015) 130e137
an increasing number of pregnancies are complicated by placental adhesive disorder, and clinicians are requesting imaging for the diagnosis and to aid surgical planning [35,36]. 3.2. Other anatomical areas Other anatomical areas are still under-imaged with fetal MRI and additional research is needed. The fetus with abnormal bones is often referred for MRI to assess other associated abnormalities. New sequences have shown potential, especially susceptibilityweighted imaging, which gives a “black bone” appearance [28]. There are a few reports describing the use of MRI to image the fetal heart [37e39]. 3.3. Planning surgery Any pathology that compresses or obstructs the fetal airway increases the risk for hypoxia at birth. To safely deliver an affected fetus, the oxygen supply needs to be maintained. Such pregnancies may be delivered using the EXIT (ex-utero intrapartum treatment) procedure, where the head and neck are delivered and operated on while maintaining utero-placental circulation. These are technically challenging deliveries, and the more detailed knowledge of the anatomy, the better the outcome ought to be. Fetal MRI can provide detail beyond ultrasound and plays a central role in these cases [40,41]. Imaging plays a central role in deciding on management and the clinician responsible for the imaging should be part of the MDT [41].
It will be essential that the reporting clinicians know the outcomes of the cases they see in order to maintain standards and increase their own level of experience. This requires a close working collaboration with neonatologists, surgeons, obstetricians, genetic counsellors, pediatric radiologists and pediatric pathologists.
5. The future of fetal MRI In the last decade, fetal MRI has gone from a research tool to an established clinical practice. It has advanced our understanding of pathology and the natural history of some diseases. For example, fetuses with cystic pulmonary malformations have a favourable prognosis unless there is associated ascites. Some reports suggest that the ascites may resolve, improving the outlook [43]. Networking will increase the speed at which we increase our understanding and improve patient imaging and patient care. Timing of imaging is key. Most centers like to image as late in gestation as possible, but the most useful time is immediately following identification of the problem on ultrasound. Earlier gestation fetal MRI will depend on local practice and on what degree of certainty or what pathology is required. The UK national guidelines (Medicine and Healthcare Products Regulatory Agency, November 2014) advise imaging of the fetus at any age whilst remaining in normal mode at both 1.5 and 3 T. Early second trimester fetal MRI may be beneficial. For example, the parents of a fetus with CDH might consider termination if the prognosis is poor. The main problem with imaging in early pregnancy is that there is little experience, so care needs to be taken in interpretation of images.
4. Current practice In the UK, centers are limiting services for various reasons, including the limited experience of available radiologists. Some offer a selective service or have limited experience at early gestational ages and only scan in the third trimester. Knowledge of the embryology and normal imaging appearances at each gestational age is essential for accurate diagnosis [42]. Imaging in early gestation is technically more challenging but can often answer a clinical question and guide management, even if detailed imaging is not possible. Experienced centers offer a larger range of services, including imaging all areas of the fetus. However, this has resulted in a large variation in the services offered throughout the UK. Methods to improve the service that are currently being established include: (a) Establishing a large network within the UK to support all active radiologists and fetal medicine experts. This would allow the major tertiary referral centres with large fetalematernal units to establish an imaging service supported by the more established groups. The network would not only provide clinical support but also act as a forum to discuss difficult cases, improve image quality, and even provide cross-coverage. With the current means of electronic communication and transfer of images, this would be easily achievable and improve the current imaging service. (b) Establishing a central group to develop imaging guidelines to ensure good image quality but to be flexible enough for centers still to be able to develop their own imaging sequences for the benefit of all. This has the added value that image quality will be maintained, and improvements in image quality or development of additional sequences that are of clinical value can be widely disseminated to benefit all patients at little cost to the NHS.
Practice points Fetal MRI is developing rapidly but areas other than the CNS have not been fully evaluated. Fetal MRI is a useful adjunct to sonography in some cases. Fetal MRI of the heart and bones is still investigational.
Research directions The role of fetal MRI for chest pathology. The role of fetal MRI for abdominal pathology. The role of MRI in placental disorders.
Conflict of interest statement None declared. Funding sources None. References [1] Smith FW, Adam AH, Phillips WD. NMR imaging in pregnancy. Lancet 1983 Jan 1;1(8314e5):61e2. PubMed PMID: 6129387. [2] Girard N, Raybaud C, Dercole C, Boubli L, Chau C, Cahen S, et al. In vivo MRI of the fetal brain. Neuroradiology 1993;35(6):431e6. PubMed PMID: 8377914. [3] Levine D, Barnes PD, Madsen JR, Li W, Edelman RR. Fetal central nervous system anomalies: MR imaging augments sonographic diagnosis. Radiology 1997 Sep;204(3):635e42. PubMed PMID: 9280237.
E. Whitby, P. Wright / Seminars in Fetal & Neonatal Medicine 20 (2015) 130e137 [4] Levine D, Hatabu H, Gaa J, Atkinson MW, Edelman RR. Fetal anatomy revealed with fast MR sequences. AJR Am J Roentgenol 1996 Oct;167(4):905e8. PubMed PMID: 8819380. [5] Levine D, Barnes PD, Sher S, Semelka RC, Li W, McArdle CR, et al. Fetal fast MR imaging: reproducibility, technical quality, and conspicuity of anatomy. Radiology 1998 Feb;206(2):549e54. PubMed PMID: 9457211. [6] Levine D. MR imaging of fetal central nervous system abnormalities. Brain Cogn 2002 Dec;50(3):432e48. PubMed PMID: 12480488. [7] Hubbard AM, Harty MP, States LJ. A new tool for prenatal diagnosis: ultrafast fetal MRI. Semin Perinatol 1999 Dec;23(6):437e47. PubMed PMID: 10630540. [8] Hubbard AM. Ultrafast fetal MRI and prenatal diagnosis. Semin Pediatr Surg 2003 Aug;12(3):143e53. PubMed PMID: 12961108. [9] Whitby E, Paley MN, Davies N, Sprigg A, Griffiths PD. Ultrafast magnetic resonance imaging of central nervous system abnormalities in utero in the second and third trimester of pregnancy: comparison with ultrasound. BJOG 2001 May;108(5):519e26. PubMed PMID: 11368139. [10] Parazzini C, Righini A, Rustico M, Consonni D, Triulzi F. Prenatal magnetic resonance imaging: brain normal linear biometric values below 24 gestational weeks. Neuroradiology 2008 Oct;50(10):877e83. PubMed PMID: 18563404. [11] Tilea B, Alberti C, Adamsbaum C, Armoogum P, Oury JF, Cabrol D, et al. Cerebral biometry in fetal magnetic resonance imaging: new reference data. Ultrasound Obstet Gynecol 2009 Feb;33(2):173e81. PubMed PMID: 19172662. [12] Coakley FV, Glenn OA, Qayyum A, Barkovich AJ, Goldstein R, Filly RA. Fetal MRI: a developing technique for the developing patient. AJR Am J Roentgenol 2004 Jan;182(1):243e52. PubMed PMID: 14684546. [13] Glenn OA, Coakley FV. MRI of the fetal central nervous system and body. Clin Perinatol 2009 Jun;36(2):273e300. viii. PubMed PMID: 19559321. [14] Grant JA. Fetal MRI. Pediatr Neurosurg 2000 Sep;33(3):167. PubMed PMID: 11096365. [15] Prayer D, Kasprian G, Krampl E, Ulm B, Witzani L, Prayer L, et al. MRI of normal fetal brain development. Eur J Radiol 2006 Feb;57(2):199e216. PubMed PMID: 16413984. [16] Prayer D, Brugger PC, Kasprian G, Witzani L, Helmer H, Dietrich W, et al. MRI of fetal acquired brain lesions. Eur J Radiol 2006 Feb;57(2):233e49. PubMed PMID: 16413156. [17] Simon EM. MRI of the fetal spine. Pediatr Radiol 2004 Sep;34(9):712e9. PubMed PMID: 15316691. [18] Whitby E. personal communication. 2014. [19] Mayer S, Klaritsch P, Petersen S, Done E, Sandaite I, Till H, et al. The correlation between lung volume and liver herniation measurements by fetal MRI in isolated congenital diaphragmatic hernia: a systematic review and metaanalysis of observational studies. Prenatal Diagn 2011 Nov;31(11):1086e96. PubMed PMID: 21915885. [20] Terui K, Omoto A, Osada H, Hishiki T, Saito T, Sato Y, et al. Prediction of postnatal outcomes in congenital diaphragmatic hernia using MRI signal intensity of the fetal lung. J Perinatol 2011 Apr;31(4):269e73. PubMed PMID: 21052047. [21] Farhataziz N, Engels JE, Ramus RM, Zaretsky M, Twickler DM. Fetal MRI of urine and meconium by gestational age for the diagnosis of genitourinary and gastrointestinal abnormalities. AJR Am J Roentgenol 2005 Jun;184(6):1891e7. PubMed PMID: 15908548. [22] Zizka J, Elias P, Hodik K, Tintera J, Juttnerova V, Belobradek Z, et al. Liver, meconium, haemorrhage: the value of T1-weighted images in fetal MRI. Pediatr Radiol 2006 Aug;36(8):792e801. PubMed PMID: 16799788. [23] Manganaro L, Saldari M, Bernardo S, Vinci V, Aliberti C, Sollazzo P, et al. Role of magnetic resonance imaging in the prenatal diagnosis of gastrointestinal fetal anomalies. La Radiologia Medica 2014 Oct 28. PubMed PMID: 25348138. [24] Kasprian G, Del Rio M, Prayer D. Fetal diffusion imaging: pearls and solutions. Top Magn Reson Imaging 2010 Dec;21(6):387e94. PubMed PMID: 22158132. [25] Prayer D, Prayer L. Diffusion-weighted magnetic resonance imaging of cerebral white matter development. Eur J Radiol 2003 Mar;45(3):235e43. PubMed PMID: 12595108.
137
[26] Manganaro L, Francioso A, Savelli S, Tomei A, Fierro F, Di Maurizio M, et al. Fetal MRI with diffusion-weighted imaging (DWI) and apparent diffusion coefficient (ADC) assessment in the evaluation of renal development: preliminary experience in normal kidneys. La Radiologia Medica 2009 Apr;114(3):403e13. PubMed PMID: 19381763. [27] Savelli S, Di Maurizio M, Perrone A, Tesei J, Francioso A, Angeletti M, et al. MRI with diffusion-weighted imaging (DWI) and apparent diffusion coefficient (ADC) assessment in the evaluation of normal and abnormal fetal kidneys: preliminary experience. Prenatal Diagn 2007 Dec;27(12):1104e11. PubMed PMID: 17849498. [28] Robinson AJ, Blaser S, Vladimirov A, Drossman D, Chitayat D, Ryan G. Fetal “Black Bone” MRI: Utility in assessment of the fetal spine. Br J Radiol 2014 Dec 11:20140496. PubMed PMID: 25496509. [29] Berger-Kulemann V, Brugger PC, Pugash D, Krssak M, Weber M, Wielandner A, et al. MR spectroscopy of the fetal brain: is it possible without sedation? AJNR Am J Neuroradiol 2013 Feb;34(2):424e31. PubMed PMID: 22821916. [30] Girard N, Gouny SC, Viola A, Le Fur Y, Viout P, Chaumoitre K, et al. Assessment of normal fetal brain maturation in utero by proton magnetic resonance spectroscopy. Magn Reson Med 2006 Oct;56(4):768e75. PubMed PMID: 16964617. [31] Heerschap A, Kok RD, van den Berg PP. Antenatal proton MR spectroscopy of the human brain in vivo. Child's Nerv Syst 2003 Aug;19(7e8):418e21. PubMed PMID: 12811484. [32] Fenton BW, Lin CS, Macedonia C, Schellinger D, Ascher S. The fetus at term: in utero volume-selected proton MR spectroscopy with a breath-hold techniqueea feasibility study. Radiology 2001 May;219(2):563e6. PubMed PMID: 11323489. [33] Dekan S, Linduska N, Kasprian G, Prayer D. MRI of the placenta e a short review. Wien Med Wochenschr 2012 May;162(9e10):225e8. PubMed PMID: 22717878. [34] Masselli G, Gualdi G. MR imaging of the placenta: what a radiologist should know. Abdom Imaging 2012 Jul 15. PubMed PMID: 22797659. [35] Bour L, Place V, Bendavid S, Fargeaudou Y, Portal JJ, Ricbourg A, et al. Suspected invasive placenta: evaluation with magnetic resonance imaging. European Radiology 2014 Dec;24(12):3150e60. PubMed PMID: 25070274. [36] Elhawary TM, Dabees NL, Youssef MA. Diagnostic value of ultrasonography and magnetic resonance imaging in pregnant women at risk for placenta accreta. J Matern Fetal Neonatal Med 2013 Sep;26(14):1443e9. PubMed PMID: 23489020. [37] Votino C, Jani J, Damry N, Dessy H, Kang X, Cos T, et al. Magnetic resonance imaging in the normal fetal heart and in congenital heart disease. Ultrasound Obstet Gynecol 2012 Mar;39(3):322e9. PubMed PMID: 21837757. [38] Yamamura J, Kopp I, Frisch M, Fischer R, Valett K, Hecher K, et al. Cardiac MRI of the fetal heart using a novel triggering method: initial results in an animal model. J Magn Reson Imaging 2012 May;35(5):1071e6. PubMed PMID: 22246623. [39] Manganaro L, Vinci V, Bernardo S, Sollazzo P, Sergi ME, Saldari M, et al. Magnetic resonance imaging of fetal heart: anatomical and pathological findings. J Matern Fetal Neonatal Med 2014 Aug;27(12):1213e9. PubMed PMID: 24102352. [40] MacArthur CJ. Prenatal diagnosis of fetal cervicofacial anomalies. Curr Opin Otolaryngol Head Neck Surg 2012 Dec;20(6):482e90. PubMed PMID: 22913933. [41] Dighe MK, Peterson SE, Dubinsky TJ, Perkins J, Cheng E. EXIT procedure: technique and indications with prenatal imaging parameters for assessment of airway patency. Radiographics 2011 MareApr;31(2):511e26. PubMed PMID: 21415194. [42] Glenn OA. Normal development of the fetal brain by MRI. Semin Perinatol 2009 Aug;33(4):208e19. PubMed PMID: 19631082. [43] Kunisaki SM, Barnewolt CE, Estroff JA, Ward VL, Nemes LP, Fauza DO, et al. Large fetal congenital cystic adenomatoid malformations: growth trends and patient survival. J Pediatr Surg 2007 Feb;42(2):404e10. PubMed PMID: 17270558.
