Use of fetal magnetic resonance imaging in patients electing

termination of pregnancy by dilation and evacuation

Geeta Sharma, MD,a Linda Heier, MD,b Robin B. Kalish, MD,a Robert Troiano, MD,c and

Stephen T. Chasen, MDa

New York, NY

OBJECTIVE: The purpose of this study was to determine whether magnetic resonance imaging of the fetal

brain before dilation and evacuation enhances diagnosis when equivocal ultrasound findings and disrupted

autopsy specimens exist.

STUDY DESIGN: Patients with equivocal fetal brain abnormalities on ultrasound examination who were

considering termination of pregnancy were evaluated retrospectively. Abdominal and pelvic magnetic

resonance imaging was performed for further evaluation, and orthogonal fetal brain images were obtained. A

multidisciplinary team reviewed all cases and discussed the findings, possible causes, and recurrence risks

with each patient.

RESULTS: Seven patients with fetal brain anomalies underwent magnetic resonance imaging before dilation

and evacuation. Magnetic resonance imaging diagnoses included intracranial hemorrhages, semilobar

holoprosencephaly, intracranial teratoma, multiple cerebral infarcts, and unilateral cerebellar hypoplasia. In all

cases, magnetic resonance imaging provided valuable information and helped distinguish possible genetic

syndromes from likely sporadic disorders of brain development.

CONCLUSION: Magnetic resonance imaging can provide insight into diagnosis, cause, and recurrence risks

for patients who choose dilation and evacuation because of fetal brain abnormalities. (Am J Obstet Gynecol

2003;189:990-3.)

Key words: Ultrasound examination, magnetic resonance imaging, central nervous system, fetalanomaly

Antenatal ultrasound examination is effective in the

detection of most fetal central nervous system (CNS)

abnormalities.1,2 When ultrasound findings are equivo-

cal, magnetic resonance imaging (MRI) may enhance

the visualization of intracranial structures and lead to

a specific diagnosis. Abnormal fetal CNS findings on MRI

correlate well with pathologic examination3,4 and can

provide clinically relevant information regarding progno-

sis and recurrence risks.

FetalMRI was first described in the 1980s, but its use was

limited by slow capture time that required fetal sedation

with drugs that were administered directly by cord-

ocentesis or to the mother.5 Improved imaging tec-

hniques that were developed in the early 1990s allowed

imaging collection, with acquisition times of 400 msec

From the Departments of Obstetrics and Gynecology, Division ofMaternal-Fetal Medicine,a Neuroradiology,b and Radiology,c WeillMedical College of Cornell University.Presented at the Twenty-third Annual Meeting of the Society for MaternalFetal Medicine, San Francisco, California, February 3-8, 2003.Reprint requests: Geeta Sharma, MD, Department of Obstetrics andGynecology, Division of Maternal-Fetal Medicine, Weill Medical Collegeof Cornell University, 525 E 68th St, Room J-130, New York, NY 10021.E-mail: geeta_sharma@yahoo.com� 2003, Mosby, Inc. All rights reserved.0002-9378/2003 $30.00 + 0doi:10.1067/S0002-9378(03)00712-9

990

that minimized fetal movement artifact. No deleterious

fetal effects have been reported.6

Patients may choose the termination of pregnancy for

fetuses with severe congenital malformations.7,8 Induc-

tion with prostaglandins permits the delivery of an intact

fetus, but completion of the procedure may require up to

3 days.9 In addition, side effects of fever, diarrhea, and

pain are common, although induction protocols that

involve misoprostol and oxytocin (Pitocin) may be better

tolerated.10,11

Alternatively, abortion that is performed between 14 to

24 weeks of gestation can be accomplished by dilation and

evacuation (D&E). This procedure can be performed

safely in experienced hands as an ambulatory procedure.

Some authors have described lower complication rates for

D&E than for other techniques of second trimester

abortion.9,12 For these reasons, patients may choose to

undergo D&E for pregnancy termination.

Although D&E may be preferable from the maternal

standpoint, this procedure often precludes adequate

neuropathologic examination. The purpose of this study

was to determine the contribution of fetal MRI to patient

counseling in cases with severe sonographic fetal brain

abnormalities that were nondiagnostic in patients who

underwent D&E.

Volume 189, Number 4Am J Obstet Gynecol

Sharma et al 991

Table. Comparison of equivocal/nondiagnostic ultrasound findings and MRI results with proposed diagnoses

CaseGestational age (wk)

at ultrasound Ultrasound findingsGestational age(wk) at MRI MRI findings Diagnosis

1 18.4 Ultrasound 1: prominentforehead, small nasalprominence, cavumseptum pellicidum notseen,abnormalventricularconfiguration, smallcerebellum and limbs

18.9 Marked hydrocephalus with in-traventricular hemorrhageseen with blood fluid levels independent portions of ven-tricles; bilateral parasagittalcortical defects, small brain-stem and cerebellum; possi-ble hypoplasia or absence ofcerebellar vermis

Encephaloclastic defectwith acute intracra-nial hemorrhage

18.7 Ultrasound 2: severeventriculomegaly

2 20 Prominent third and lateralventricles, cystic mass inposterior fossa

20.4 Enlarged lateral ventricles,large blood fluid levels seenwithin bilateral occipitalhorns, right frontal horn, andwithin cystic retrovermianmass in the cerebellum

Acute intracranialhemorrhage

3 23.6 Midline intracranial masswith cystic component;mild ventriculomegaly,cavum septum pelluci-dum, thalami and thirdventricle not seen

23.8 Suprasellar mass with centralnecrosis, within third ventri-cle causing obstructive hy-drocephalus; no corpuscollosum visualized

Midline intracranialmass, likely neoplas-tic: teratoma, PNETor glioblastoma

4 17.4 Thin cerebral cortex, dis-ordered and enlargedlateral ventricles, cavumseptum pellicidum notvisualized, and oligohy-dramnios.

17.6 Fused thalami and frontallobes, large supratentorialdorsal cyst, monoventricle

Semilobar holoprosen-cephaly, triploidy

5 19.7 Ventriculomegaly/asym-metry; poorly visualizedmidline structures

19.9 Asymmetric enlargement oflateral ventricles, focal andasymmetric areas of absentcortex; cortical lesions appearto have resulted from vascularinsult; small fourth ventricle

Multiple vascular insultsand aqueductalstenosis

6 18.6 outsideultrasound

Ultrasound 1: Dandy-Walker malformationsuspected, mild hydro-cephalus, corpus collo-sum not visualized

21.6 Corpus collosum and cavumseptum pellicidum visualized,colpocephaly seen; absentmiddle and inferior cerebel-lar vermis and absent midand inferior left cerebellarhemisphere; ex vacuo en-largement of fourth ventricleand cisterna magna; posteriorfossa not enlarged

Primary disorder ofneuronal migration,unilateral cerebellaragenesis

22.6 Ultrasound 2: Dandy-Walker variant suspected

7 21.4 Microcephaly, asymmetriclateral ventricles—colpo-cephaly, cavum septumpellucidum seen, lemonsign

21.9 Right-sided colopocephaly; nosplenium of corpus collosumseen, cavum septum pellici-dum seen, interhemisphericdorsal subarachnoid cystcompressing left occipitalhorn and lobe; small cere-bellar hemispheres

Interhemispheric sub-arachnoid cyst andpartial agenesis ofcorpus collosum

Material and methods

Patients who were included in this study underwent

second-trimester ultrasound examination in our unit

from February 2001 to July 2002. When a severe fetal

CNS anomaly was suspected by ultrasound examination

but a specific diagnosis could not be identified, MRI was

offered for further evaluation. Cytogenetic testing was

also recommended. Patients who were included in this

study were those patients who opted to undergoD&E after

undergoing extensive counseling about the ultrasound

findings. MRI imaging permitted further discussion

regarding possible causes and recurrence risks.

Ultrasound images were obtained with a 6 MHz trans-

abdominal transducer with multihertz and harmonic

capability (Sequoia system 512; Acuson, Mountain View,

Calif). MRI imaging was performed on a 1.5 T General

October 2003Am J Obstet Gynecol

992 Sharma et al

Electric (Milwaukee, Wis) torso coil with multiple echo-

planar T2-weighted sequences. Single-shot fast spin echo

imaging (5-mm thick/0 spacing; field of view, 32; matrix,

256 3 256 with one acquisition) was performed through

the fetal brain in the axial, coronal, and sagittal planes.

Less than 1 week elapsed between our ultrasound and

MRI examinations. Maternal-fetal medicine specialists

reviewed all ultrasound images. MRI images were inter-

preted by a pediatric neuroradiologist and were reviewed

with maternal-fetal medicine specialists. Geneticists were

involved in the multidisciplinary approach, and all pa-

tients were encouraged to undergo genetic counseling.

Results

There were 7 patients who considered termination of

pregnancy because of fetal CNS anomalies that were

suggested by antenatal ultrasound examination. All 7

Fig 1. Case 1: Ultrasound examination and MRI images. A,Acute severe ventriculomegaly that was detected by ultrasoundexamination at 18.7 weeks that was not present at 18.4 weeks. B,MRI visualized a recent hemorrhagic event.

patients chose D&E as the mode for termination of

pregnancy. MRI results in all 7 cases confirmed the

presence of a CNS anomaly, further elucidated possible

diagnoses, and provided information for patient counsel-

ing that was relevant to recurrence risks in subsequent

pregnancies. All pregnancies were conceived spontane-

ously by healthy mothers with no known comorbidities,

consanguinity, or significant genetic histories.

Major ultrasound and MRI findings for the 7 cases are

described in the Table. Figs 1 and 2 show the ultrasound

andMRI images for cases 1 and 3. In all but 1 case (case 6),

the fetal brain was disrupted during evacuation, pre-

cluding neuropathologic examination. In case 6, autopsy

confirmed the MRI findings.

Fig 2. Case 3: Ultrasound examination and MRI images. A, Acircumscribed intracranial mass that was detected by ultrasoundexamination at 23.6 weeks was delineated by (B) MRI as probablyneoplastic.

Volume 189, Number 4Am J Obstet Gynecol

Sharma et al 993

Comment

MRI can be an effective tool to enhance or to confirm

diagnosis when fetal autopsy is not possible. Guo et al3

compared MRI findings to pathologic reports from

neonatal autopsy and induced abortions. Their results

demonstrated a significant correlation between antenatal

MRI and autopsy findings.

Our experience suggests that the use of MRI in patients

who undergo evacuation because of fetal CNS anomalies

can provide valuable information when sonographic

findings are inconclusive. The 2 cases of fetal intracranial

hemorrhage led to testing for alloimmune thrombo-

cytopenia, which commonly recurs and can be treated

aggressively in future pregnancies.13 Although, the testing

revealed no positive results, the small size of the current

series prohibits any conclusions. The preoperative di-

agnosis of fetal holoprosencephaly in 1 case underscored

the importance of cytogenetic testing, the results of which

excluded a syndrome with autosomal recessive inheri-

tance. The diagnoses of intracranial tumor, irregular

cortical clefts, unilateral cerebellar hypoplasia, and in-

terhemispheric subarachnoid cyst suggested sporadic

occurrence, which are unlikely to recur in the future.

Because these MRI diagnoses were made with a high

degree of confidence, we felt that D&E did not prevent

clinically relevant information from being obtained

because of fetal disruption.

In summary, fetal MRI may be indicated in those cases

in which the additional information may offer further

assurance to patients who must come to terms with an

anomalous fetus and the decision of pregnancy termina-

tion. Even in those cases in which the parents are resolved

to pregnancy termination because of sonographic evi-

dence of a severe brain anomaly, MRI may provide

diagnostic findings and reduce the importance of autopsy

in the achievement of a diagnosis.

REFERENCES

1. Levine D, Barnes PD, Madsen JR, Abbott J, Mehta T, Edelman RR.Central nervous system abnormalities assessed with prenatal mag-netic resonance imaging. Obstet Gynecol 1999;94:1011-9.

2. Simon EM, Goldstein RB, Coakley FV, Filly RA, Broderick KC, MusciTJ, et al. Fast MR imaging of fetal CNS anomalies in utero. Am JNeuroradiol 2000;21:1688-98.

3. Guo WY, Chang CY, Ho DMT, Wong TT, Sheu MH, Cheng HC, et al.A comparative MR and pathological study on fetal CNS disorders.Child Nerv Syst 2001;17:512-8.

4. Wagenvoort AM, Bekker MN, Go AT, Vandenbussche FP, VanBuchem MA, Valk J, et al. Ultrafast scan magnetic resonance inprenatal diagnosis. Fetal Diagn Ther 2000;15:364-72.

5. Hubbard AM, States LJ. Fetal magnetic resonance imaging. TopMagn Reson Imaging 2001;12:93-103.

6. Baker PN, Johnson IR, Harvey PR, Gowland PA, Mansfield P. A three-year follow up of children imaged in utero with echo-planarmagnetic resonance. Am J Obstet Gynecol 1994;170:32-3.

7. Saari-Kemppainen A, Karjalainen O, Ylostalo P, Heinonen OP.Ultrasound screening and perinatal mortality: controlled trial ofsystematic one-stage screening in pregnancy: the Helsinki Ultra-sound Trial. Lancet 1990;336:387-91.

8. Julian-Reynier C, Philip N, Scheiner C, Aurran Y, Chabal F, Maron A,et al. Impact of prenatal diagnosis by ultrasound on the prevalence ofcongenital anomalies at birth in southern France. J EpidemiolCommunity Health 1994;48:290-6.

9. Autry AM, Hayes EC, Jacobson GE, Kirby RS. A comparison ofmedical induction and dilation and evacuation for second-trimesterabortion. Am J Obstet Gynecol 2002;187:393-7.

10. Owen J, Hauth JC. Vaginal misoprostol vs concentrated oxytocin pluslow-dose prostaglandin E2 for second trimester pregnancy termina-tion. J Matern Fetal Med 1999;8:48-50.

11. Bebbington MW, Kent N, Lim K, Gagnon A, Delisle MF, Tessier F,et al. A randomized controlled trial comparing two protocols for theuse of misoprostol in midtrimester pregnancy termination. Am JObstet Gynecol 2002;187:853-7.

12. Kalish RB, Chasen ST, Rosenzweig LB, Rashbaum WK, Chervenak FA.Impact of midtrimester dilation and evacuation on subsequentpregnancy outcome. Am J Obstet Gynecol 2002;187:882-5.

13. Bussel JB, Zabusky MR, Berkowitz RL, McFarland JG. Fetalalloimmune thrombocytopenia. N Engl J Med 1997;337:22-6.