Fetal Brain MRI: Novel Classification and Contribution to Sonography

Abstract

Purpose: 1) To evaluate and classify the indications for fetal brain MRI in a tertiary referral center. 2) To assess the contribution of fetal brain MRI to fetal neurosonography.

Materials and Methods: A retrospective study in a tertiary medical center during a two-year period (2011 – 2012) included pregnant women who underwent fetal brain MRI. MRI was implemented at 32 weeks of gestation unless a severe abnormality possibly requiring earlier medical intervention was suspected.

Results: 633 patients were included, 40 (6.3 %) underwent repeated examinations with a total of 733 fetal MRI scans. Patients were classified to three main indication cohorts: Suspected primary brain anomaly (52.9 %), non-CNS disorders (32.5 %) and obstetrical complications (14.6 %). These cohorts were further divided into 16 separate groups with lateral ventricle abnormalities being the most common (23.7 %), followed by exposure to TORCH (17.5 %) and cerebral cortex abnormalities (13 %). 149 (19.3 %) fetal MRI scans demonstrated additional findings. Repeated examinations were commonly implemented in complicated monochorionic-biamniotic (MCBA) twin pregnancies (34.6 %) and in cases of supra-tentorial cysts (19 %). The average gestational age for MRI scan in the MCBA group was 26 ± 5 weeks in comparison to ≥ 31st weeks in all other groups (p < 0.001).

Conclusion: The current study describes a detailed picture of fetal brain MRI indications. Most patients were referred because of CNS anomalies. The impressive diversity of 16 separate entities emphasizes the expanding use of fetal brain MRI. Complicated MCBA pregnancies, which may have dramatic events, constitute a unique challenge due to early and repetitive MRI examinations and may serve as a role model for the contribution of fetal MRI during antenatal evaluation. The contribution of MRI to prenatal evaluation in various indications is discussed.

Zusammenfassung

Ziel: 1) Bewertung und Klassifizierung der Indikationen für eine MRT des fetalen Gehirns in einem Tertiärzentrum. 2) Die Bewertung der MRT des fetalen Gehirns unterstützend zur Neurosonografie.

Material und Methoden: Eine retrospektive Studie in einem medizinischen Tertiärzentrum schloss Schwangere über einen Zeitraum von 2 Jahren (2011 – 2012) ein, bei denen eine fetale MRT des Gehirns erfolgte. Die MRT wurde in der 32. Schwangerschaftswoche durchgeführt, wenn nicht aufgrund des Verdachts auf schwere Anomalien eine frühere medizinische Intervention nötig war.

Ergebnisse: Eingeschlossen wurden 633 Patienten mit insgesamt 733 fetalen MRT-Untersuchungen, davon 40 (6,3 %), bei denen wiederholt Untersuchungen durchgeführt wurden. Die Patienten wurden nach der Hauptindikation in drei Kohorten eingeteilt: Verdacht auf primäre Hirnanomalien (52,9 %), Erkrankungen ohne ZNS-Beteiligung (32,5 %) und geburtshilfliche Komplikationen (14,6 %). Diese Kohorten wurden in 16 verschiedene Untergruppen eingeteilt, wobei die häufigsten Auffälligkeiten Anomalien der Seitenventrikel (23,7 %), Exposition zu TORCH-Erregern (17,5 %) und Anomalien der Hirnrinde (13 %) waren. Bei den fetalen MRT-Untersuchungen wurden in 149 Fällen (19,3 %) zusätzliche Befunde erhoben. Folgeuntersuchungen wurden im Allgemeinen in schwierigen monochoriotisch-diamniotischen (MCDA) Zwillingsschwangerschaften (34,6 %) und in Fällen mit supratentoriellen Zysten (19 %) durchgeführt. Das durchschnittliche Schwangerschaftsalter bei MRT-Untersuchung in der MCDA-Gruppe betrug 26 ± 5 Wochen im Vergleich zu ≥ 31. SSW in allen anderen Gruppen (p < 0,001).

Schlussfolgerung: Die vorliegende Studie beschreibt ein detailliertes Bild der Indikationen für eine MRT des fetalen Gehirns. Die meisten Patienten wurden aufgrund von lokalisierten ZNS-Auffälligkeiten untersucht. Die beeindruckende Vielfalt von 16 unterschiedlichen Untergruppen unterstreicht den zunehmenden Einsatz der MRT des fetalen Gehirns. Schwierige MCDA-Schwangerschaften, die dramatische Verläufe haben können, sind eine ganz besondere Herausforderung aufgrund der frühen und wiederholten MRT-Untersuchungen. Sie können als Vorbild dienen, was den Beitrag der fetalen MRT bei der vorgeburtlichen Analyse anbelangt. Die Leistung der MRT für die pränatale Bewertung bei verschiedenen Indikationen wird diskutiert.

Key words

brain - CNS - MR imaging

Introduction

Ultrasound (US) is the method of choice for fetal imaging regarding routine dating, screening, evaluation of growth and anomalies as well as monitoring of placental function, assessment of the risk for aneuploidy and guidance of invasive procedures [1]. Magnetic resonance imaging (MRI) is a useful complementary method for evaluation and decision making in the work-up of antenatal pathologies identified during obstetrical US screening [2] [3].

Since the first report of MRI in pregnancy [4], its use has rapidly expanded due to many diverse factors in medical prenatal practice [5]. First, MRI has several advantages in comparison to US including improved resolution and direct visualization of both sides of the fetal brain, thus overcoming some of the difficulties encountered in obstetrical US such as decreased amniotic fluid volume, fetal positioning and acoustic shadowing from the ossifying calvarium [6]. Second, as a consequence of its capabilities, fetal MRI can detect additional abnormalities which were not diagnosed by US and may lead to altered patient consultation and management [7] [8] [9]. For example, Morris et al. have demonstrated various additional findings demonstrated by MRI in 50 % of isolated ventriculomegaly such as germinal matrix bleeds, focal hemimegalencephaly, agenesis of the corpus callosum, aqueduct stenosis and others [10]. Third, MRI permits imaging in more than one plane in fetuses with difficult US-based diagnosis without fetal exposure to ionizing radiation [11]. The surge in the development of fast MRI techniques and further progressions (such as the rapid pulse sequences, parallel imaging with an increasing number of channels and advances in coil design) have improved imaging capabilities in pregnant women and fetuses [12] and contributed to the impressive evolution of prenatal MRI. Therefore, the increasing list of indications for fetal MRI performance [13] is not surprising.

The expanding use of MRI has previously been described. However, existing reports of fetal brain MRI indications lack uniformity. While some authors have reviewed the general conceptions leading to MRI performance such as adding certainty to US diagnosis, looking for additional abnormalities and research [1], others have described specific indications including ventriculomegaly, arachnoid cysts, Dandy Walker malformations, holoprosencephaly etc. [14]. Data regarding the incidence of each indication is limited. The aims of the current research were to classify the indications for fetal brain MRI and to evaluate the additional findings demonstrated by MRI per indication in order to supply a comprehensive, detailed and updated picture regarding that evolving field.

Materials and Methods

Population

In this retrospective study we evaluated the charts of all pregnant women who underwent fetal brain MRI conducted in a single tertiary center between 1/1/2011 and 31/12/2012. All patients were referred for a fetal brain MRI due to a specific suspected anomaly demonstrated by sonographic anatomical survey (mostly during 19 – 24 weeks of gestation) or in cases of obstetrical complications which may lead to brain damage. MRI was conducted after performing fetal neurosonogram by a specialized gynecologist in the field of fetal imaging according to International Society of Ultrasound in Obstetrics and Gynecology (ISUOG) guidelines [15]. Additionally, three-dimensional (3 D) neurosonography was performed according to the neurosonography specialist’s (E. K.) clinical judgment. Prior to MRI, clinical history regarding personal and family medical history was taken including obstetrical history and confirmation of gestational age determined by first trimester US. All patients gave their informed consent prior to MRI performance.

MRI examination

The preferred timing for fetal brain MRI in our institute is the 32^nd week of gestation due to the ability to assess brain maturation (sulcation and gyration) and parenchyma [6] [16] as well as confirmation of US diagnosis and evaluation of additional abnormal findings. However, MRI was conducted earlier in cases of suspected severe abnormality which may require medical intervention (such as termination of pregnancy, laser therapy in MCBA pregnancy, etc.) or parental anxiety [16] since termination of pregnancy during the second and third trimesters is performed in Israel in cases of suspected severe fetal disabilities including brain pathologies. Alternatively, in cases in which medical intervention was not planned before the 32^nd week, there was a gap between the first US pathological findings and MRI performance in order to achieve a comprehensive brain evaluation which is limited before the 32^nd week. Repetitive examination was performed in cases of optional radiological changes during pregnancy (such as after intrauterine procedures in MCBA pregnancies).

Scans were obtained using a 1.5 T MR system (GE Healthcare). Single-shot fast spin-echo T2-weighted sequences were used with the FOV determined by the size of the fetal head with a range of 24 – 30 cm. Other parameters were a matrix of 320 × 224, a TE of 90 ms, and a TR of 1298 ms. The fast-spoiled gradient recalled T1 sequence was performed only in the axial plane with a larger FOV of 40 cm, a TR of 160 ms, and a TE of 2.3 ms. Then a DWI sequence was performed with a 40-cm FOV, b-values of 0 and 1000 or 700 ms. The ADC calculation map was added [17].

Fetal MRI was performed and evaluated by an MR-ultrasonography collaboration team which included an obstetrician specialist in fetal US (EK) and a neuro-radiologist MRI expert (CH). Both are familiar with the field of fetal imaging and participated in all MR examinations and interpretations. In addition to brain examination, a routine evaluation of the head profile and face (ears, eyes, palate, and tongue) was conducted in order to exclude midline pathologies such as cleft lip/palate.

Indications classification

Fetal brain MRI indications were classified into three main cohorts: located CNS anomalies, non-CNS disorders and obstetrical complications. The intention of this classification was to clarify and organize the tremendously wide range of indications and to construct a comprehensive and reliable list of fetal MRI indications. Each cohort was comprised of several specific indications as described below.

CNS anomalies included the following suspected abnormalities:

1. Lateral ventricle examination was indicated in cases of ≥ 10 mm dilatation [18] with or without asymmetry of ≥ 2 mm between ventricles..

2. Cortical abnormality examination included suspected macrocephaly or microcephaly of more or less than 2 standard deviations, measured by US [19] [20]. Additional referrals in this category included suspected abnormal sulcation maturation which presented as either abnormal CSF volume, sulcation delay according to gestational age [21] or aberrant operculation of the Sylvian fissure [22].

3. Abnormal US findings of the midline structures included absence or abnormality of the corpus callosum [16] or cavum septi pellucidi [3].

4. Pathological findings in the posterior fossa included cerebellar or brain stem abnormalities, large cisterna magna or suspected Dandy-Walker malformations [6] [16].

5. Supratentorial cysts³ included midline (ex. cavum vergae) and non-midline cysts such as choroid plexus cyst and periventricular pseudocysts.

6. Craniosynostosis was suspected in cases of abnormal skull shape and/or early suture closure. Although MRI is not considered the method of choice for suture visualization, it enables the exclusion of pathological pressure on the brain tissue and assessment of the subarachnoid space in order to detect associated brain abnormalities [23].

7. NTD referrals [11] [24] included suspected encephalocele or myelomeningocele.

8. Brain tumors.

Non-CNS disorders included the following entities:

1. Infectious etiology constituted cases of maternal exposure to TORCH. Amniocentesis (AC) was recommended after 21 completed weeks of gestation and at least 7 weeks from first positive serological results [25] in order to diagnose fetal disease prior to MRI implementation. CMV isolation was performed by culture of fibroblasts, shell vial technique and polymerase chain reaction amplification of CMV DNA in amniotic fluid [26]. Patients who refused to undergo AC were referred to MRI in the absence of fetal exposure diagnosis [6].

2. Body anomalies which may be associated with brain malformations including unexplained bilateral club foot (after exclusion of oligohydramnios or breech presentation as possible etiologies) [27], heart anomalies except isolated ASD or VSD [28] or anomalies involving more than 2 different systems (urinary, musculoskeletal, heart, gastrointestinal etc.). Additional indications in that category were unexplained severe polyhydramnion with an amniotic fluid index (AFI) of over 300 mm with normal karyotype and absence of relevant abnormalities [29].

3. Suspected genetic disease [30] included previous pregnancy [16] or family member with neurological malformation or deficit, numeric or structural chromosomal abnormality demonstrated by karyotype or comparative genomic hybridization (CGH) and abnormal first/second trimester screening tests for Down syndrome.

4. Head and face anomalies included cleft lip/palate [14] or other suspected head and face anomalies (cataract, deformed ears etc.).

Obstetrical complications included the following indications:

1. Severe MCBA pregnancy (monochorionicity was based on first trimester ultrasound evaluation) complications including: twin to twin transfusion syndrome (TTTS); selective intrauterine growth restriction (IUGR); intrauterine fetal demise (IUFD) of one fetus; intrauterine procedures (laser ablation or selective termination by bipolar or radiofrequency) [17] [30] [31]. In cases of repetitive examinations due to different indication (ex. two MRI scans due to TTTS performed before and after laser therapy) each scan was calculated separately.

2. MRI was implemented in severe IUGR (less than 5^th percentile) in two scenarios. First, ischemic brain damage was evaluated in cases of early severe IUGR accompanied by abnormal Doppler parameters[3] [32]. Second, MRI was performed in order to exclude brain anomaly after exclusion of placental or infectious anomaly or anomaly which may result from IUGR[33].

3. Abnormal antepartum fetal evaluation leading to MRI in order to exclude brain etiology or ischemic damage included: fetal hypotonia; abnormal Doppler as a reflection of placental insufficiency; brain evaluation after intrauterine blood transfusion due to severe fetal anemia to exclude ischemic brain damage; loss of variability in non-stress test (NST) remote from term without explanation of the reduced variability[6].

4. Suspected reduced uterine perfusion during pregnancy due to bleeding ectopic pregnancy, maternal surgery or hypovolemic shock aimed at excluding ischemic brain damage[6].

Outcome measurements and statistical analysis

After all indications were evaluated and classified, the number of fetal scans was summarized and the following parameters were evaluated for each indication: incidence, twin pregnancies ratio, MRI imaging timing, rate of repeated MRI scan and the rate of additional findings. Additional findings were defined in the CNS anomalies as demonstration of pathological findings in addition to the located abnormality which was the indication for examination, while the definition of additional findings in the non-CNS disorders and obstetrical complications cohorts was demonstration of any abnormal finding.

Each indication group was compared to the average of all other indications. The t-test and chi square were used as appropriate. A p-value < 0.05 was considered significant.

Ethical approval

The research was approved by the Hospital Research Ethics Board.

Results

633 pregnant patients with a total of 773 fetuses were referred for fetal brain MRI during a two-year period (2011 – 2012). The average gestational age for MRI performance was 32 ± 4 weeks with 50 twin pregnancies (9.3 %) after exclusion of the MCBA group. 40 patients (6.3 %) had repeated examinations. 149 (19.3 %) of all fetal MRI scans demonstrated additional findings.

Division into three main classification cohorts revealed that detection of a CNS anomaly was the main referral indication for fetal MRI including 335 (52.9 %) patients. There were 206 (32.5 %) women in the non-CNS disorders cohort while 92 (14.6 %) were referred due to obstetrical complications ([Fig. 1]). Each cohort was further evaluated and subdivided in order to describe a detailed and precise description ([Table 1]).

Fig. 1 Fetal brain MRI: Indication distribution.

Table 1 Fetal MRI indications classification, incidences, timing, need for repetitive examinations and additional findings.

main classification	indication	patients (%)	gestation week[1]	patients with repeated examinations (%)	fetuses	additional findings[2]
CNS anomalies	lat. ventricles	150 (23.7)	32.1 ± 2.6	7 (4.7)	176	27 (15.3)
	cerebral cortex	82 (13)	34.7 ± 2.6	1 (1.2)	90	16 (19.5)
	midline structures	37 (5.9)	31.1 ± 4.3	1 (2.7)	41[3]	10 (24.4)
	post. Fossae	36 (5.7)	31.3 ± 4.4	0 (0)	38	11 (30.1)
	supratentorial cyst	19 (3)	32.7 ± 3.6	4 (19)	25	4 (16)
	craniosynostosis	5 (0.8)	34.6 ± 1.5	0 (0)	5	1 (20)
	NTD	4 (0.6)	34 ± 4.8	0 (0)	5	0 (0)
	tumor	2 (0.3)	35 ± 4.2	0	2	0 (0)
non-CNS disorders	infectious	111 (17.5)	33.4 ± 2.2	6 (5.4)	124	27 (21.6)
	body anomaly	42 (6.6)	33.3 ± 3.2	0 (0)	44	9 (20.5)
	genetic	41 (6.5)	31.8 ± 2.3	2 (4.9)	463	9 (20)
	head or face anomaly	12 (1.9)	30.5 ± 3.3	0 (0)	13	0 (0)
obstetrical complications	MCBA	52 (8.2)	26.2 ± 4.5	18 (34.6)	113	26 (23)
	IUGR	30 (4.8)	32.8 ± 2.5	1 (3.3)	403	8 (20)
	fetal evaluation	6 (0.9)	32.2 ± 1.7	0 (0)	7	1 (14.3)
	reduced uterine perfusion	4 (0.6)	31.3 ± 1.9	0	4	0 (0)
total		633 (100)	31.8 ± 3.8	40 (6.3)	773	149 (19.3)

¹ Mean ± standard deviation.

² Additional findings were defined in the CNS anomalies as demonstration of pathological findings in addition to the located abnormality which was the indication for examination, while the definition of additional findings in the non-CNS disorders and obstetrical complications cohorts was demonstration of any abnormal finding.

³ One patient with twin pregnancy underwent two examinations.

CNS anomalies:

1. The most common among the detected anomalies group as well as the whole research cohort was the lateral ventricles group, in which 109 patients were referred to MRI due to asymmetry with or without dilatation of one ventricle to ≥ 10 mm and 41 women were examined because of bilateral dilatation. Additional pathologic findings were demonstrated among 15.3 % fetuses ([Fig. 2]).

2. The abnormal cortical development group included suspected microcephaly in 55 patients, macrocephaly in 21 cases and sulcation delay in 6. MRI revealed additional pathologies in 19.5 % of cases.

3. Midline structure malformations included 24 patients with suspected abnormal corpus callosum and 13 with unsatisfying demonstration of the cavum septi pellucidi. The abnormal fetal examination rate in that group was 24.4 %.

4. Posterior fossa malformations included 25 cases of suspected megacysterna magna, 7 abnormally demonstrated cerebellums, 3 cases of pathological vermis and one case of suspected Blake’s pouch. 30.1 % of fetuses were diagnosed with additional brain insults ([Fig. 3]).

5. The 19 women referred for evaluation due to supratentorial cysts were divided into 15 non-midline and 4 midline cysts. 4 examinations demonstrated additional pathologies.

6. All 5 craniosynostosis cases were singleton pregnancies among which one examination was defined as abnormal.

7. The NTD group consisted of 3 cases of myelomeningocele and one of encephalocele. No additional brain insults were demonstrated.

8. There were 2 cases of suspected brain tumors without other brain pathologies.

Fig. 2 a Axial ultrasonographic view of the fetal brain demonstrating asymmetric lateral ventricles with irregular border, diagnosed at 32 weeks of gestation. b Axial T2 MR images demonstrating porencephalic cyst with damage to the brain parenchyma. c Sagittal T2 MR images demonstrating porencephalic cyst with damage to the brain parenchyma.

Fig. 3 a Dandy-Walker malformation at 27 weeks of gestation. Mid-sagittal ultrasonographic view of the fetal brain demonstrating agenesis of the vermis, enlarged posterior fossa with elevated tentorium, thick and short corpus callosum. b Dandy-Walker malformation at 27 weeks of gestation. T2 MR mid-sagittal plane image confirming the diagnosis and also emphasis of the abnormal development of the cortex. c Dandy-Walker malformation at 27 weeks of gestation. Post-mortem T2 MR scan.

Out of 382 fetuses which were included in the CNS anomalies group, 69 (18 %) had abnormal MRI examinations due to additional pathological findings. No significant differences were found between indications groups regarding abnormal examinations rates.

Non-CNS disorders:

1. The infectious entity included women who were exposed to TORCH: 108 to CMV and one to varicella, one to rubella and one to toxoplasma. Abnormal examination was diagnosed in 21.6 % of fetuses.

2. The group of body anomalies included 42 patients and was divided into 13 referrals due to bilateral club feet, 11 heart anomalies and 9 cases of multiple anomalies which involved more than 2 systems and 9 examinations because of polyhydramnion. 20.5 % of fetuses had abnormal examinations.

3. Suspected genetic pathologies (41 patients) included mostly patients with a prior abnormal fetus or family member with a neurological syndrome (28 women), numeric or structural chromosomal abnormality (10 patients) and an aberrant screening test for Down syndrome (3 women). Brain pathologies were demonstrated in 9 (20 %) cases.

4. Among the fetuses identified as having head or face anomalies, 7 patients had fetuses with cleft lip and/or palate, 3 with abnormal head findings and 2 with fetal face abnormalities. Fetal brain MRI was normal in all cases.

Out of 227 fetuses which were included in the non-CNS anomalies group, 45 (19.8 %) had abnormal MRI examinations due to pathological findings. No significant differences were found between indications groups regarding abnormal examination rates.

Obstetrical complications:

1. The MCBA group consisted of 52 patients. Of these, 18 women (34.6 %) underwent repeated examinations (3 patients had 5, 4 and 3 MR scans each compared to 2 examinations among the other 15) resulting in a total of 75 MR scans. 43 examinations were performed after intrauterine procedure (laser therapy or selective reduction) and 14 after IUFD of one fetus in order to evaluate the surviving fetus. Other scans were conducted due to selective IUGR or TTTS which did not necessitate intervention. Pathological findings were demonstrated in 26 (23 %) fetuses ([Fig. 4]).

2. IUGR patients were divided into 22 singleton pregnancies and 8 bichorionic-biamniotic (BCBA) twin pregnancies with IUGR of one fetus. 8 fetuses (20 %) had an abnormal examination.

3. Abnormal obstetrical fetal evaluation included 3 cases of reduced variability on NST and a single case of hypotonia, abnormal Doppler evaluation and brain evaluation after intrauterine blood transfusion due to severe fetal anemia. One examination (14.3 %) was defined as abnormal.

4. Four women had an earlier episode of hypotension during pregnancy and therefore were suspected for reduced uterine perfusion. These cases included abdominal surgeries in 3 women and one patient who had seizure. No brain insults were demonstrated in that group.

Fig. 4 a Monochorionic-biamniotic twin at 24 weeks of gestation after spontaneous demise of the co-twin two weeks earlier. Axial ultrasonographic view of the fetal brain demonstrating enlarged lateral ventricles with abnormal appearance of the brain parenchyma. b Monochorionic-biamniotic twin at 24 weeks of gestation after spontaneous demise of the co-twin two weeks earlier. Axial T2 MR image demonstrating massive bilateral cortical necrosis probably due to ischemic damage. c Monochorionic-biamniotic twin at 24 weeks of gestation after spontaneous demise of the co-twin two weeks earlier. T2 MR mid-sagittal plane image of the surviving fetus (upper) with the dead co-twin (lower).

Out of 164 fetuses which were included in the obstetrical complications group, 35 (21.3 %) had abnormal MRI examinations due to pathological findings. No significant differences were found between indications groups regarding abnormal examination rates.

Further analysis revealed significantly higher rates of repetitive examinations in the complicated MCBA pregnancy group (34.6 %) and the supratentorial cysts group (19 %) compared to all other indications (p < 0.001). Moreover, the average gestational age for the first MRI scan in the MCBA group was 26 ± 5 weeks in comparison to ≥ 31 weeks in all other groups (p < 0.001). There were no significant differences regarding twin pregnancy rate between indications groups except the obvious higher rate among the MCBA cohort.

Discussion

The emerging use of fetal MRI for antenatal evaluation is evolving. Deciding which indication necessitates referral for fetal MRI is still a major challenge for obstetricians and specialized obstetricians in the field of fetal imaging and for radiologists. The expanding referrals for fetal MRI require intensive knowledge in the fields of anatomical and developmental embryology. The current research included 633 patients and 773 fetal MR scans which were conducted over a 2-year period in a single tertiary center. To the best of our knowledge, this is the largest study for the assessment and classification of fetal brain MRI indications. The described results provide a comprehensive and detailed picture of the growing referrals for that examination.

Previous reports have described the indications for fetal MRI and demonstrated the CNS evaluation as the most common indication for fetal MRI [24] [34]. The current study has divided the large number of fetal brain MRI scans according to their indications. Classification of fetal MRI indications presents a great challenge because of the wide range of possible abnormal findings. The indications were divided in our classification into three cohorts: Detected CNS findings, suspected non-CNS disorders and obstetrical complications. This classification has several advantages. First, each cohort describes specific clinical and radiological settings with a typical MRI role. MRI was implemented in the first cohort in order to confirm US diagnosis and identify or rule out additional pathologies. The second cohort consists of patients suspected to have a non-CNS disorder which would not necessarily include a brain anomaly, yet MRI was performed to dismiss such injuries. MRI was performed in the third cohort to rule out brain pathology which may cause the obstetrical complications or to exclude an ischemic event resulting from that complication. Second, this classification simplifies the huge diversity of brain abnormalities and delineates the principles of antenatal MRI evaluation. Third, the current division of indications to three main cohorts may be suitable for additional rare indications (not included in the current study) such as vascular brain anomalies which may be added to the indications cohort of detected CNS findings. Our results emphasize detected CNS anomalies as the major refer for fetal brain MRI (52.9 %). However, one cannot ignore the important contribution of non CNS disorders and obstetrical considerations to MRI implementation.

The further division of each cohort into 16 separate entities emphasizes the impressive increase in indications for fetal brain MRI. Our results point out that lateral ventricle pathologies, exposure to TORCH and assessment of the cortex are the most common referrals (together constituting more than half of all evaluations). Rare indications include suspected brain tumors and reduced uterine perfusion due to previous maternal hypotension during pregnancy. The incidence of lateral ventricle abnormality evaluations in our study of 23.7 % coincides with previous reports. Levine described fetal brain MR scans of 214 patients; 25.7 % of them were referred due to ventriculomegaly[14] compared to 55.5 % of 86 cases published by Trompoukis et al. [34]. However, the incidences of TORCH exposure (17.5 %), cerebral cortex evaluation (13 %) and NTD (only 0.6 %) have not been previously published.

The contribution of MRI during complicated pregnancies is complex and depends on various clinical and imaging considerations. Comparison between all pathological MRI as well as US findings in each indication was beyond the scope of the current study, which included an extremely large cohort. Therefore, we had to define general considerations in order to emphasize the additional findings demonstrated by MRI. Additional findings in the detected CNS anomalies cohort included cases in which brain insults were demonstrated in addition to the primary pathology which was the indication for examination, while in the non-CNS anomalies and the obstetrical complications cohorts, any brain insult resulted in the classification of additional findings. Although such a general approach does not enable a detailed discussion regarding the various US versus MRI pathologies in each indication, it gives a reliable overview of the MRI contribution to antenatal evaluation.

19.3 % of the fetal brain MRI examinations in the current study demonstrated additional findings. Although the contribution of fetal brain MRI has been previously investigated in some indications [6] [35] [36], it has never been reported in others. For example, additional findings were found in 40 – 50 % of cases of ventriculomegaly and 63 % of cases referred due to suspected an abnormal corpus callosum [37] [38]. However, in the current research we found additional MRI findings in only 15.3 % and 24.4 % of ventriculomegaly and midline structures, respectively. The gap between these early observations and the current results can be explained by the great advances in prenatal US and especially sonographic CNS evaluation as demonstrated by the neurosonogram guidelines published on 2007 [15]. On the other hand, our results regarding the demonstration of additional findings in complicated MCBA pregnancies (23 %) correlate with a previous report which found MRI pathologic findings in 9/34 (26.5 %) cases in that population [17]. In addition, the current study evaluates the contribution of MRI in less investigated indications such as non-CNS anomalies. We assume that information may contribute to the debate regarding the contribution of MRI to US [39] [40].

The current study emphasizes that a complicated MCBA pregnancy is a unique entity with early gestational age during the MR scan (26 ± 5 weeks) and the need for repetitive examinations later in pregnancy (34.6 %) due to dramatic changes during these pregnancies which may require early intervention and repetitive evaluations. Therefore, complicated MCBA pregnancies may serve as a role model for the contribution of fetal MRI during antenatal evaluation. It should be noted that the supratentorial cyst group was demonstrated with a high rate of repetitive MR scans as well (19 %). An optional explanation is the tendency of these pathologies to continue developing during late stages of pregnancy and to possibly be accompanied by additional pathologies [41] [42]. Therefore, later evaluation is indicated.

The introduction of fetal brain MRI has been received with enthusiasm [39]. However, some doubts regarding possible biases in the studies comparing US and MRI have been speculated such as overenthusiastic approach towards MRI as well as methodological drawbacks[43]. Some authors have differentiated between specific abnormalities in which MRI is superior to US (gross space occupying lesions, posterior fossa and the posterior corpus callosum) as opposed to other anomalies in which MRI has no advantage over neurosonography (ventriculomegaly, holoprosencephaly, open NTD and the anterior corpus callosum) especially when performed by a dedicated neurosonographer [39] [44]. Additionally, as any other method, MRI has its own limitations such as fetal motion and maternal claustrophobia and discomfort during the scan which may result in movement artifacts and reduced study quality [6]. These conflicting attitudes towards fetal MRI emphasize the importance of large studies such as the current study which enable a comprehensive evaluation of the massive use of that method as a part of prenatal assessment.

Although the comparisons between US and MRI have been intensively investigated, data regarding the relationship between MRI and 3 D US is scarce. Neurosonogram performed in fetal neurology clinics at tertiary referral centers [39] as well as the ISUOG guidelines do not include routine 3 D evaluations[15]. However, evaluations of the spine [15], face [45] and fetal brain vasculature [46] should include 3 D assessment. 3 D neurosonography was performed in the current study as previously described [47]. Unfortunately, data regarding each 3 D assessment was not available in that retrospective study which included a very large sample size. Therefore, the added value of 3 D neurosonography was not examined and should be investigated in future studies.

As opposed to the complementary workup of prenatal MRI and US, in practice these modalities are usually conducted by different specialists: MRI is performed by radiologists who usually do not have a background in fetal medicine and do not perform fetal US, while US is often performed by OB/GYN experts who do not have special skills in neuro-anatomy and neuroimaging [3]. Glenn has emphasized the importance of continued multi-disciplinary collaborative efforts [7]. The current study sheds light on the enormous diversity of the findings during MR scans and therefore the strengths of the concept of the MR-ultrasonography collaboration team [3].

In conclusion, the current study included a very large number of fetal MRI examinations and gives a comprehensive evaluation and classification of the indications regarding the surge in this field. The stepwise classification may contribute to better understanding among physicians and caregivers regarding the indications for fetal brain MRI. Although the common indications have previously been published and evaluated, rare indications, which implement a unique antenatal challenge, should not be ignored. In spite of the fact that MRI results were not in the scope of the current study, MRI influence on the clinical outcome in these rare situations should be evaluated.

• References

• 1 Weston MJ. Magnetic resonance imaging in fetal medicine: a pictorial review of current and developing indications. Postgraduate medical journal 2010; 86: 42-51 ; quiz 50
• 2 Ghobrial PM, Levy RA, O'Connor SC. The fetal magnetic resonance imaging experience in a large community medical center. Journal of clinical imaging science 2011; 1: 29
• 3 Guibaud L. Contribution of fetal cerebral MRI for diagnosis of structural anomalies. Prenatal diagnosis 2009; 29: 420-433
• 4 Smith FW, Adam AH, Phillips WD. NMR imaging in pregnancy. Lancet 1983; 1: 61-62
• 5 Garel C. Fetal MRI: what is the future?. Ultrasound in obstetrics & gynecology: the official journal of the International Society of Ultrasound in Obstetrics and Gynecology 2008; 31: 123-128
• 6 Vazquez E, Mayolas N, Delgado I et al. Fetal neuroimaging: US and MRI. Pediatric radiology 2009; 39 (Suppl. 03) 422-435
• 7 Glenn OA. MR imaging of the fetal brain. Pediatric radiology 2010; 40: 68-81
• 8 Levine D, Barnes PD, Madsen JR et al. Central nervous system abnormalities assessed with prenatal magnetic resonance imaging. Obstetrics and gynecology 1999; 94: 1011-1019
• 9 Levine D, Barnes PD, Madsen JR et al. Fetal central nervous system anomalies: MR imaging augments sonographic diagnosis. Radiology 1997; 204: 635-642
• 10 Morris JE, Rickard S, Paley MN et al. The value of in-utero magnetic resonance imaging in ultrasound diagnosed foetal isolated cerebral ventriculomegaly. Clinical radiology 2007; 62: 140-144
• 11 Levine D, Barnes PD, Edelman RR. Obstetric MR imaging. Radiology 1999; 211: 609-617
• 12 Levine D, Hatabu H, Gaa J et al. Fetal anatomy revealed with fast MR sequences. Am J Roentgenol American journal of roentgenology 1996; 167: 905-908
• 13 Chapman T, Matesan M, Weinberger E et al. Digital atlas of fetal brain MRI. Pediatric radiology 2010; 40: 153-162
• 14 Levine D. Obstetric MRI. Journal of magnetic resonance imaging: JMRI 2006; 24: 1-15
• 15 International Society of Ultrasound in O, Gynecology Education C. Sonographic examination of the fetal central nervous system: guidelines for performing the 'basic examination' and the 'fetal neurosonogram'. Ultrasound in obstetrics & gynecology: the official journal of the International Society of Ultrasound in Obstetrics and Gynecology 2007; 29: 109-116
• 16 Salomon LJ, Garel C. Magnetic resonance imaging examination of the fetal brain. Ultrasound in obstetrics & gynecology: the official journal of the International Society of Ultrasound in Obstetrics and Gynecology 2007; 30: 1019-1032
• 17 Hoffmann C, Weisz B, Yinon Y et al. Diffusion MRI findings in monochorionic twin pregnancies after intrauterine fetal death. AJNR American journal of neuroradiology 2013; 34: 212-216
• 18 Cardoza JD, Goldstein RB, Filly RA. Exclusion of fetal ventriculomegaly with a single measurement: the width of the lateral ventricular atrium. Radiology 1988; 169: 711-714
• 19 Biran-Gol Y, Malinger G, Cohen H et al. Developmental outcome of isolated fetal macrocephaly. Ultrasound in obstetrics & gynecology: the official journal of the International Society of Ultrasound in Obstetrics and Gynecology 2010; 36: 147-153
• 20 Stoler-Poria S, Lev D, Schweiger A et al. Developmental outcome of isolated fetal microcephaly. Ultrasound in obstetrics & gynecology: the official journal of the International Society of Ultrasound in Obstetrics and Gynecology 2010; 36: 154-158
• 21 Cohen-Sacher B, Lerman-Sagie T, Lev D et al. Sonographic developmental milestones of the fetal cerebral cortex: a longitudinal study. Ultrasound in obstetrics & gynecology: the official journal of the International Society of Ultrasound in Obstetrics and Gynecology 2006; 27: 494-502
• 22 Lerman-Sagie T, Malinger G. Focus on the fetal Sylvian fissure. Ultrasound in obstetrics & gynecology: the official journal of the International Society of Ultrasound in Obstetrics and Gynecology 2008; 32: 3-4
• 23 Fjortoft MI, Sevely A, Boetto S et al. Prenatal diagnosis of craniosynostosis: value of MR imaging. Neuroradiology 2007; 49: 515-521
• 24 Santos XM, Papanna R, Johnson A et al. The use of combined ultrasound and magnetic resonance imaging in the detection of fetal anomalies. Prenatal diagnosis 2010; 30: 402-407
• 25 Lipitz S, Hoffmann C, Feldman B et al. Value of prenatal ultrasound and magnetic resonance imaging in assessment of congenital primary cytomegalovirus infection. Ultrasound in obstetrics & gynecology: the official journal of the International Society of Ultrasound in Obstetrics and Gynecology 2010; 36: 709-717
• 26 Shulman LM, Rudich C, Sayar Y et al. Detection of CMV-DNA in cells from peritoneal fluid of IPD/CAPD patients by polymerase chain reaction. Advances in peritoneal dialysis Conference on Peritoneal Dialysis 1992; 8: 258-264
• 27 Kammoun F, Tanguy A, Boesplug-Tanguy O et al. Club feet with congenital perisylvian polymicrogyria possibly due to bifocal ischemic damage of the neuraxis in utero. American journal of medical genetics Part A 2004; 126A: 191-196
• 28 Mlczoch E, Brugger P, Ulm B et al. Structural congenital brain disease in congenital heart disease: results from a fetal MRI program. European journal of paediatric neurology: EJPN: official journal of the European Paediatric Neurology Society 2013; 17: 153-160
• 29 Damato N, Filly RA, Goldstein RB et al. Frequency of fetal anomalies in sonographically detected polyhydramnios. Journal of ultrasound in medicine: official journal of the American Institute of Ultrasound in Medicine 1993; 12: 11-15
• 30 Triulzi F, Manganaro L, Volpe P. Fetal magnetic resonance imaging: indications, study protocols and safety. La Radiologia medica 2011; 116: 337-350
• 31 Kline-Fath BM, Calvo-Garcia MA, O'Hara SM et al. Twin-twin transfusion syndrome: cerebral ischemia is not the only fetal MR imaging finding. Pediatric radiology 2007; 37: 47-56
• 32 Gressens P, Luton D. Fetal MRI: obstetrical and neurological perspectives. Pediatric radiology 2004; 34: 682-684
• 33 Egana-Ugrinovic G, Sanz-Cortes M, Figueras F et al. Differences in Cortical Development assessed by fetal MRI in late-onset intrauterine growth restriction. American journal of obstetrics and gynecology 2013;
• 34 Trompoukis P, Papantoniou N, Chlapoutaki C et al. Fetal MRI: is it really helpful?. The journal of maternal-fetal & neonatal medicine: the official journal of the European Association of Perinatal Medicine, the Federation of Asia and Oceania Perinatal Societies, the International Society of Perinatal Obstet 2012; 25: 2363-2368
• 35 Adamsbaum C, Moutard ML, Andre C et al. MRI of the fetal posterior fossa. Pediatric radiology 2005; 35: 124-140
• 36 Epelman M, Daneman A, Blaser SI et al. Differential diagnosis of intracranial cystic lesions at head US: correlation with CT and MR imaging. Radiographics: a review publication of the Radiological Society of North America, Inc 2006; 26: 173-196
• 37 Glenn OA, Barkovich J. Magnetic resonance imaging of the fetal brain and spine: an increasingly important tool in prenatal diagnosis: part 2. AJNR American journal of neuroradiology 2006; 27: 1807-1814
• 38 Glenn OA, Barkovich AJ. Magnetic resonance imaging of the fetal brain and spine: an increasingly important tool in prenatal diagnosis, part 1. AJNR American journal of neuroradiology 2006; 27: 1604-1611
• 39 Malinger G, Ben-Sira L, Lev D et al. Fetal brain imaging: a comparison between magnetic resonance imaging and dedicated neurosonography. Ultrasound in obstetrics & gynecology: the official journal of the International Society of Ultrasound in Obstetrics and Gynecology 2004; 23: 333-340
• 40 Pistorius LR, Hellmann PM, Visser GH et al. Fetal neuroimaging: ultrasound, MRI, or both?. Obstetrical & gynecological survey 2008; 63: 733-745
• 41 Bats AS, Molho M, Senat MV et al. Subependymal pseudocysts in the fetal brain: prenatal diagnosis of two cases and review of the literature. Ultrasound in obstetrics & gynecology: the official journal of the International Society of Ultrasound in Obstetrics and Gynecology 2002; 20: 502-505
• 42 Sahinoglu Z, Uludogan M, Delikara MN. Prenatal sonographic diagnosis of dilated cavum vergae. Journal of clinical ultrasound: JCU 2002; 30: 378-383
• 43 Malinger G, Lev D, Lerman-Sagie T. Is fetal magnetic resonance imaging superior to neurosonography for detection of brain anomalies?. Ultrasound in obstetrics & gynecology: the official journal of the International Society of Ultrasound in Obstetrics and Gynecology 2002; 20: 317-321
• 44 Garel C, Moutard ML. Main congenital cerebral anomalies: how prenatal imaging AIDS counseling. Fetal diagnosis and therapy 2014; 35: 229-239
• 45 Andresen C, Matias A, Merz E. Fetal face: the whole picture. Ultraschall in der Medizin 2012; 33: 431-440
• 46 Pashaj S, Merz E. Prenatal demonstration of normal variants of the pericallosal artery by 3D ultrasound. Ultraschall in der Medizin 2014; 35: 129-136
• 47 Paladini D, Quarantelli M, Sglavo G et al. The role of MRI in the clinical management of foetuses with central nervous system abnormalities in a tertiary referral center. Ultrasound in obstetrics & gynecology: the official journal of the International Society of Ultrasound in Obstetrics and Gynecology 2013;