Ultrasound Obstet Gynecol 2006; 28: 899903
Published online 6 November 2006 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/uog.3865
Prenatal diagnosis of open and closed spina bifida
T. GHI*, G. PILU*, P. FALCO*, M. SEGATA*, A. CARLETTI*, G. COCCHI, D. SANTINI,
P. BONASONI, G. TANI and N. RIZZO*
Departments of *Obstetrics and Gynecology, Neonatology, Pathology and Pediatric Radiology, Policlinico S. Orsola-Malpighi and
University of Bologna, Bologna, Italy
K E Y W O R D S: alpha-fetoprotein; congenital anomalies; fetus; neural tube defects; prenatal diagnosis; spina bifida; ultrasound
ABSTRACT
Objective To identify criteria useful for differentiating
closed from open spina bifida antenatally.
Patients and methods A retrospective study of cases of
spina bifida diagnosed in a referral center between 1997
and 2004.
Results Of 66 cases of fetal spina bifida diagnosed at
a median gestational age of 21 (range, 1634) weeks,
detailed follow-up was available for 57. Of these, open
defects were found in 53 (93.0%) and closed defects in
four (7.0%). Closed spina bifida was associated in two
cases with a posterior cystic mass with thick walls and
a complex appearance, while in two cases the spinal
lesion could not be clearly differentiated from an open
defect, particularly at mid-gestation. Open spina bifida
was always associated with typical alterations of cranial
anatomy, including the so-called banana and lemon
signs, while in closed spina bifida the cranium was
unremarkable. When the data were available, levels of
amniotic fluid alpha-fetoprotein were always abnormally
elevated with open spina bifida and within normal limits
with closed forms.
Conclusion In this study 7% of cases of spina bifida
diagnosed in utero were closed. The differentiation
between open and closed forms is best shown by
the sonographic demonstration of abnormal or normal
cranial anatomy. Copyright 2006 ISUOG. Published
by John Wiley & Sons, Ltd.
INTRODUCTION
Spina bifida includes a continuum of anomalies that have
in common a defect of closure (dysraphism) of the neural
tube. Although many entities are found and different
Correspondence to: Dr G. Pilu, Clinica Ginecologica e Ostetrica, Policlinico S. Orsola-Malpighi, Via Massarenti 13, 40138 Bologna, Italy
(e-mail: pilu@aosp.bo.it)
Accepted: 6 January 2006
Copyright 2006 ISUOG. Published by John Wiley & Sons, Ltd.
terminologies are used, it is commonly accepted that two
main categories exist: open spina bifida (nervous tissue
and/or meninges exposed to the environment) and closed
spina bifida (skin-closed dysraphism)1 . Most studies on
the prenatal diagnosis of spina bifida have focused upon
open spina bifida, which is associated with an increased
concentration of alpha-fetoprotein (AFP) in amniotic fluid
and maternal serum and with typical cranial signs at
the sonographic examination2 5 . Intrauterine diagnosis of
closed spina bifida has also been reported6 10 . However,
in many of the available studies a precise distinction
between open and closed defects is not made11,12 . Yet
these entities, although embryologically related, are very
different from a pathophysiological as well as clinical
point of view1 . Closed spina bifida has a much more
favorable prognosis than the open forms. Indeed, many
affected individuals are asymptomatic13 15 .
We report here the experience of a referral center in
the prenatal diagnosis of spina bifida. The purpose of our
study was to identify criteria useful for the differential
diagnosis of open and closed defects.
PATIENTS AND METHODS
The archives of the ultrasound laboratory of our depart-
ment were retrospectively searched for all cases of isolated
fetal spina bifida that had been diagnosed antenatally
in the period 19972004. Prenatal sonographic findings
were noted, including the appearance of the spinal defect,
the presence and features of posterior masses associated
with the defect, and the cranial anatomy. Cranial signs of
spina bifida (the so-called lemon and banana signs) were
categorized as previously suggested3 . Ventriculomegaly
was diagnosed when the transverse diameter of the ven-
tricular atrium was 10 mm or greater16 . When the data
were available, levels of amniotic fluid AFP were also
ORIGINAL PAPER
900 Ghi et al.

noted. Concentrations higher than two standard devia-
tions above the mean for the reference values established
in our laboratory were considered abnormal. Postna-
tally, the spinal defects were classified as suggested by
Tortori-Donati et al.1 .
RESULTS
In the study period 66 fetuses with isolated spina bifida
were identified. The mean gestational age at diagnosis was
21 (range, 1634) weeks, and in 56 cases the diagnosis
was made before 24 weeks. The vast majority (n = 57) of
these fetuses were referred to our ultrasound laboratory
because of suspected anomaly on a routine scan. Other
indications included increased AFP in maternal serum
(n = 6) or in amniotic fluid (n = 3). Data on amniotic
fluid AFP were available in 15 cases.
Termination of pregnancy was elected in 59 cases, while
seven fetuses were delivered alive and survived. Detailed
follow-up was available in 57 cases, which represent the
study group. In 53 cases (93.0%) the defect was classified
postnatally as open. In four (7.0%) cases a closed defect
was found (three lipomeningoceles, one meningocele).
At mid-gestation, open defects were invariably associ-
ated with both the lemon sign and the banana sign, while
ventriculomegaly was present in 34/53 cases (64.2%).
Vertebral dysraphism with splaying of the lateral pro-
cesses was always demonstrated. In most cases (51/53)
this was associated with a thin walled cyst containing at
times internal septations, that was later identified as a
myelomeningocele (open spina bifida with a dorsal cyst)1 .
In the two remaining cases, no cystic structures above the
defect were seen, and these fetuses were diagnosed after
birth as having a myelocele (open spina bifida without a
dorsal cyst) (Figure 1)1 .
Of the four fetuses with closed spina bifida, three
were diagnosed before 24 weeks gestation and one at
25 weeks. They all had normal intracranial anatomy.
Spinal dysraphism was similar to the one encountered
with the open forms (Figure 2). However, it was always
confined to the sacral area and associated with cystic
structures that in two cases had thick walls and a
complex appearance with a mixture of anechoic areas
and echogenic material. These fetuses were diagnosed
after termination of pregnancy to have spina bifida with
a subcutaneous lipoma (Figure 3). The two remaining
fetuses had simple cystic structures with thin walls.
One of these (Figure 2) was found after birth to have
a meningocele; she received spinal surgery and is affected

Figure 1 Ultrasound images of open spina bifida at mid-gestation. (a) Cranial signs; (b) transverse section of the spine demonstrating
vertebral dysraphism with an associated cystic mass (arrow) that was later proven to be a myelomeningocele; (c) vertebral dysraphism
without posterior cyst (arrow) that was found to be a myelocele.

Figure 2 Ultrasound images of closed spina bifida with meningocele at 22 weeks gestation. (a) The intracranial anatomy is unremarkable;
(b, c) spinal dysraphism with an associated cystic mass (arrows) that was later proven to be a meningocele. The thickness of the cyst wall
does not appear overtly different from that of the myelomeningocele displayed in Figure 1.

Copyright 2006 ISUOG. Published by John Wiley & Sons, Ltd. Ultrasound Obstet Gynecol 2006; 28: 899903.
Prenatal diagnosis of open and closed spina bifida 901

Figure 3 (a, b) Ultrasound images of closed spina bifida with subcutaneous lipoma at 21 weeks gestation; this fetus had normal cranial
anatomy. The vertebral defect is associated with a posterior cyst (arrow) that has a very thick wall with an echogenic component, suggesting
a lipoma. (c) Spina bifida with intact skin and a lipoma are confirmed by pathological analysis after termination of pregnancy.

Figure 4 Closed spina bifida with subcutaneous lipoma (arrows); this fetus had normal cranial anatomy. The antenatal sonogram of the
spine at 28 weeks (a) is correlated with the postnatal appearance of the newborn (b) and with a magnetic resonance image (MRI) (c). Note
that while ultrasound demonstrates an anechoic lesion suggesting a meningocele, postnatal T2-weighted MRI demonstrates a hyperintense
lesion extending from the neural canal, indicative of a lipoma.

by urinary incontinence at the age of 7, without any other
neurologic morbidity. The other infant was found at birth
to have closed spina bifida with a lipoma and is doing
well at 1 month of age, seemingly free from neurologic
symptoms (Figure 4).
Amniotic fluid AFP level was available in 11 cases
with open spina bifida and was always increased when
compared with our reference values. Conversely, the four
fetuses with closed defects had normal levels.
Most of our patients were seen at mid-gestation and
elected termination of pregnancy. Only five fetuses with
open spina bifida were delivered at term. In the third
trimester, the banana sign persisted, while the lemon sign
disappeared and in general ventriculomegaly worsened.
In the two pregnancies with closed spina bifida that
continued, fetal intracranial anatomy was unremarkable
throughout gestation and after delivery.
DISCUSSION
Our study suggests that the most valuable sonographic
clue for differentiating closed from open spina bifida in
the fetus is the absence of cranial signs. With open spina
bifida there is leakage of cerebrospinal fluid within the
amniotic cavity. It has been suggested that the ensuing
hypotension of subarachnoid spaces triggers a cascade
of events that eventually results in the ArnoldChiari or
Chiari II malformation, a combination of small posterior
fossa, obliteration of the cisterna magna, prolapse
of cerebellum into the foramen magnum, obstructive

Copyright 2006 ISUOG. Published by John Wiley & Sons, Ltd. Ultrasound Obstet Gynecol 2006; 28: 899903.
902
hydrocephalus and abnormal calvarial development1 , that
in the fetus is characterized sonographically by the so-
called cranial signs1,3 5,11 . Open spina bifida and the
Chiari II malformation are parts of a malformative
sequence and are constantly associated. In closed spina
bifida the defect of the neural tube is sealed by skin, there
is no loss of cerebrospinal fluid and the cranial anatomy
is normal1 .
Previous reports noted the features of the cystic mass
associated with spinal dysraphism, which is usually
anechoic with a thin wall with open defects, while that
associated with the closed ones has a thick wall and/or
a complex appearance with echogenic components6 10 .
However, in our experience these findings were not clear-
cut. Similarly to previous reports, in half of our closed
defects the posterior cysts were completely anechoic
and the thickness of the wall was not overtly different
from that in myelomeningoceles, particularly at mid-
gestation. Our experience also confirms that in closed
spina bifida meningoceles and lipomas have a very
similar antenatal appearance and may be impossible to
distinguish7 . Lipomas typically appear sonographically as
echogenic masses. The reason why fetal spinal lipomas
are frequently anechoic (Figure 4) is unclear. It is worth
noting that intracranial lipomas are usually only detected
sonographically in late gestation17 .
The level of amniotic fluid AFP may assist the
distinction between open and closed spina bifida.
Increased concentrations of amniotic fluid AFP are
almost invariably found with open defects2,18 , while in
our experience closed defects always had values well
within the normal range. Determination of maternal
serum AFP levels would probably be valuable as
well, but our experience is limited. Amniotic fluid
acetylcholinesterase is more specific than AFP, and it
could also be useful in differentiating open from closed
defects. Unfortunately, this test was not performed in
our patients because it is not available in our laboratory
and therefore we cannot comment on that. Magnetic
resonance imaging has been used in previous studies, but
it has not been found to add significant information to
ultrasonography7,9 .
Exceptions to the general rule that normal intracranial
anatomy of the fetus and normal amniotic fluid AFP
predict closed lesions are expected. Open spina bifida
with minimal degrees of Chiari II malformation has
been described1 . This is, however, a rare finding, and
indeed we have never encountered it. False negatives of
cranial signs with spina bifida have occasionally been
reported, but in most of the reported cases the defects
were covered by skin19 . Although this may not be entirely
clear from the available literature, cranial signs do not
correlate with spina bifida in general, but only with the
Chiari II malformation that is typical of the open forms.
Conversely, leakage of cerebrospinal fluid may also occur
with closed defects as a consequence of a fistula opening
into the spinal canal, although usually this does not result
in the Chiari II malformation1 .
Copyright 2006 ISUOG. Published by John Wiley & Sons, Ltd.
Ghi et al.
Distinction between open and closed spina bifida has
prognostic implications. In the former, neurologic com-
promise is the consequence of two different mechanisms:
on one side, abnormal differentiation and development
of the neural cord resulting in variable degrees of motor
paralysis to the lower limbs and incontinence; on the other
side, hydrocephalus due to the Chiari II malformation. In
closed spina bifida there is usually a much lesser involve-
ment of the neural cord, and the Chiari II malformation
does not develop. In general, the outcome for infants with
closed spina bifida is good, although neurologic symptoms
of variable entity are frequently present. The interested
reader is referred to specific studies on this issue13 15 .
The format of our study does not allow us to comment
on the sensitivity of antenatal ultrasound in the prediction
of spina bifida. It is widely accepted that the diagnostic
accuracy of fetal ultrasound is excellent in pregnancies
at high risk of neural tube defects2,18,20 . Conversely,
in low-risk patients, the results vary, with detection
rates ranging between 40 and 80%20 . In general, better
results are obtained when ultrasonography is performed in
conjunction with AFP screening. In the Emilia Romagna
region, where our center is based, and where AFP
screening is not the standard of care, fetuses with spina
bifida are terminated in 80% of cases21 . Most of the
available studies, however, do not clearly separate open
from closed spina bifida, and the sensitivity of ultrasound
in the detection of the latter is unknown. Closed spina
bifida is indeed an elusive entity, whose real incidence
has not yet been clearly established. It is presumed
that it accounts for about 10% of cases22 , but this is
probably an underestimate. In a referral center for the
treatment of spina bifida, the closed forms were found
to be largely predominant1 . It is, however, difficult to
interpret these data, as selective abortion of fetuses with
open forms may have introduced a selection bias. In our
own series closed spina bifida was 7% of all cases, and
we speculate that most of these defects are not detected
prenatally. Identification of spinal dysraphism limited to
a few vertebral segments in the sacral area is difficult, and
particularly so when both intracranial anatomy and AFP
are normal.
In summary, a minority of cases of spina bifida
diagnosed in utero are closed, and they have a better
prognosis than open forms. The most valuable clue for
a specific recognition of these defects is the presence of
normal cranial anatomy of the fetus.
REFERENCES
1. Tortori-Donati P, Rossi A, Cama A. Spinal dysraphism: a
review of neuroradiological features with embryological cor-
relations and proposal for a new classification. Neuroradiology
2000; 42: 471491.
2. Filly RA, Callen PW, Goldstein RB. Alpha-fetoprotein screen-
ing programs: what every obstetric sonologist should know.
Radiology 1993; 188: 19.
3. Nicolaides KH, Campbell S, Gabbe SG, Guidetti R. Ultrasound
screening for spina bifida: cranial and cerebellar signs. Lancet
1986; 2: 7274.
Ultrasound Obstet Gynecol 2006; 28: 899903.
Prenatal diagnosis of open and closed spina bifida
4. Pilu G, Romero R, Reece EA, Goldstein I, Hobbins JC, Bovi-
celli L. Subnormal cerebellum in fetuses with spina bifida. Am J
Obstet Gynecol 1988; 158: 10521056.
5. Van den Hof MC, Nicolaides KH, Campbell J, Campbell S.
Evaluation of the lemon and banana signs in one hundred
and thirty fetuses with open spina bifida. Am J Obstet Gynecol
1990; 162: 322327.
6. Kim SY, McGahan JP, Boggan JE, McGrew W. Prenatal diag-
nosis of lipomyelomeningocele. J Ultrasound Med 2000; 19:
801805.
7. Pierre-Kahn A, Sonigo P. Lumbosacral lipomas: in utero diag-
nosis and prognosis. Childs Nerv Syst 2003; 19: 551554.
8. Meyer SH, Morris GF, Pretorius DH, James HE. Terminal
myelocystocele: important differential diagnosis in the prenatal
assessment of spina bifida. J Ultrasound Med 1998; 17:
193197.
9. Sattar TS, Bannister CM, Russell SA, Rimmer S. Pre-natal
diagnosis of occult spinal dysraphism by ultrasonography and
post-natal evaluation by MR scanning. Eur J Pediatr Surg 1998;
8 (Suppl 1): 3133.
10. Thorne A, Pierre-Kahn A, Sonigo P. Antenatal diagnosis of
spinal lipomas. Childs Nerv Syst 2001; 17: 697703.
11. Grandjean H, Larroque D, Levi S. The performance of routine
ultrasonographic screening of pregnancies in the Eurofetus
Study. Am J Obstet Gynecol 1999; 181: 446454.
12. Garne E, Loane M, Dolk H, De Vigan C, Scarano G, Tucker D,
Stoll C, Gener B, Pierini A, Nelen V, Rosch C, Gillerot Y, Fei-
joo M, Tincheva R, Queisser-Luft A, Addor MC, Mosquera C,
Gatt M, Barisic I. Prenatal diagnosis of severe structural con-
genital malformations in Europe. Ultrasound Obstet Gynecol
2005; 25: 611.
13. Kulkarni AV, Pierre-Kahn A, Zerah M. Conservative manage-
ment of asymptomatic spinal lipomas of the conus. Neuro-
surgery 2004; 54: 868873; discussion 873875.
Copyright 2006 ISUOG. Published by John Wiley & Sons, Ltd.
903
14. Pierre-Kahn A, Zerah M, Renier D, Cinalli G, Sainte-Rose C,
Lellouch-Tubiana A, Brunelle F, Le Merrer M, Giudicelli Y,
Pichon J, Kleinknecht B, Nataf F. Congenital lumbosacral
lipomas. Childs Nerv Syst 1997; 13: 298334; discussion 335.
15. Sattar MT, Bannister CM, Turnbull IW. Occult spinal dys-
raphism the common combination of lesions and the clinical
manifestations in 50 patients. Eur J Pediatr Surg 1996;
6 (Suppl 1): 1014.
16. Cardoza JD, Goldstein RB, Filly RA. Exclusion of fetal
ventriculomegaly with a single measurement: the width
of the lateral ventricular atrium. Radiology 1988; 169:
711714.
17. Pilu G, Sandri F, Perolo A, Pittalis MC, Grisolia G, Cocchi G,
Foschini MP, Salvioli GP, Bovicelli L. Sonography of fetal
agenesis of the corpus callosum: a survey of 35 cases. Ultrasound
Obstet Gynecol 1993; 3: 318329.
18. Watson WJ, Chescheir NC, Katz VL, Seeds JW. The role of
ultrasound in evaluation of patients with elevated maternal
serum alpha-fetoprotein: a review. Obstet Gynecol 1991; 78:
123128.
19. Benacerraf BR, Stryker J, Frigoletto FD Jr. Abnormal US
appearance of the cerebellum (banana sign): indirect sign of
spina bifida. Radiology 1989; 171: 151153.
20. Boyd PA, Wellesley DG, De Walle HE, Tenconi R, Garcia-
Minaur S, Zandwijken GR, Stoll C, Clementi M. Evaluation
of the prenatal diagnosis of neural tube defects by fetal
ultrasonographic examination in different centres across
Europe. J Med Screen 2000; 7: 169174.
21. International Clearinghouse for Birth Defects Monitoring
System. Annual report with data for 2000. The International
Centre for Birth Defects: Rome, 2002.
22. Main DM, Mennuti MT. Neural tube defects: issues in prenatal
diagnosis and counselling. Obstet Gynecol 1986; 67: 116.
Ultrasound Obstet Gynecol 2006; 28: 899903.