Fetal cerebral ventriculomegaly: outcome in 176 cases




To evaluate the outcome of fetuses affected by different degrees of ventriculomegaly.


We studied 176 fetuses with ventriculomegaly and evaluated the pregnancy outcome and the neurodevelopmental outcome at age ≥ 24 months. The population was divided into three groups according to ventricular width: A (mild ventriculomegaly, 10 to 12 mm); B (moderate, 12.1 to 14.9 mm) and C (severe, ≥ 15 mm).


Ventriculomegaly was more often an isolated finding in Group A (44/75; 58.7%) than in Group B (10/41; 24.4%) and Group C (24/60; 40%). When the ventriculomegaly was an isolated finding, 97.7% of fetuses with mild, 80% with moderate and 33.3% of those with severe dilatation were alive at ≥ 24 months. The neurodevelopmental outcome was normal in 93% of Group A, 75% of Group B and 62.5% of Group C.


Our results suggest that the definition of borderline ventriculomegaly should be limited to ventricular width below 12 mm. Cases with measurements above this value are more often associated with malformations and have a normal neurodevelopmental outcome less frequently. Copyright © 2005 ISUOG. Published by John Wiley & Sons, Ltd.


A standard examination of the fetus includes prenatal sonographic evaluation of the cerebral ventricles, which can be detected as early as 15 weeks' gestational age, and their measurement is part of the assessment of the central nervous system (CNS).

Severe ventriculomegaly (width of the atria of the lateral ventricles ≥ 15 mm) is often associated with an unfavorable outcome1. However, the significance of lesser degrees of dilatation, also called borderline dilatations (between 10 and 15 mm), is still unclear: the dilatation may be associated with structural or chromosomal anomalies, or be a harmless finding with no pathological significance2–6. Therefore, the detection of dilatation raises several problems in obstetric management and counseling of prospective parents.

The aim of the present study was to assess the outcome of fetuses with different degrees of prenatally diagnosed ventricular dilatation.


We reviewed 204 consecutive cases of ventriculomegaly referred to our ultrasound unit in a 10-year period (June 1990 to June 2000). Twin pregnancies (n = 14) and cases with incomplete follow-up (n = 14) were excluded, leaving 176 cases for analysis.

Indications for scanning were: suspected ventriculomegaly and/or other CNS malformations in 123 cases (70%), suspected non-CNS malformations in 30 cases (17%) and no risk factors (routine scanning) in 23 cases (13%). The scans were performed transabdominally with different ultrasound equipment: Aloka SSD-680, Aloka SSD-1700 (Aloka Company Ltd, Tokyo, Japan), ATL 3000 and ATL 5000 (Advanced Technology Laboratories, Bothwell, WA, USA). Ventriculomegaly was defined as the diameter of one or both lateral ventricles ≥ 10 mm. The ventricular atria were measured on an axial plane at the level of the thalami by positioning electronic calipers on the internal margins of the ventricular wall perpendicular to the long axis of the ventricles7.

In each case, a thorough sonographic evaluation of fetal anatomy was performed, including fetal echocardiography. Information about the fetal karyotype was available in 152 cases. Twenty-four mothers refused the cytogenetic analysis (17 had a normal child and seven had the pregnancy terminated because of multiple malformations of the fetus). Screening for infections (TORCH) was performed in only 30 cases. Data on pregnancy outcome were available in all cases: spontaneous miscarriage, termination of pregnancy (TOP), type of delivery, perinatal, neonatal and infant morbidity and mortality. In cases of miscarriage, TOP, neonatal or infant death the postmortem examination was available. A detailed clinical examination and instrumental examination, where appropriate, were obtained in liveborn infants. All the surviving neonates were at least 24 months old at the time of the study (range, 2–12 years).

Data on their neurodevelopmental condition were obtained in structured interviews with the parents. The questionnaire included the evaluation of: locomotor activities and co-ordination of movements; hearing and visual capacity (also activity of the oculomotor muscles); development and quality of speech and socialization skills; learning performance; evolution of diagnosed anomalies or new pathologies and their therapies (often chronic) (Figure 1).

Figure 1.

Questionnaire on the neurodevelopmental condition of the surviving neonates.

We defined severe neurodevelopmental outcomes as: cerebral palsy (with or without the need for orthopedic help, urinary incontinence, daily enemas, etc.), epilepsy, bradyacusia with prothesis, mono/bilateral blindness and mental retardation. Mild anomalies were: moderate motor skill problems (with or without tone and reflex anomalies), squint, nystagmus, mild speech difficulty, moderate learning problems and ventricular-peritoneal shunt with normal motor development8.

The population was divided into three groups according to the degree of ventricular dilatation: mild, from 10 to 12 mm (Group A), moderate, from 12.1 to 14.9 mm (Group B), and severe, ≥ 15 mm (Group C). There were 75 cases (43%) in Group A, 41 (23%) in Group B and 60 (34%) in Group C. Outcomes in the three groups were compared by means of χ2 test and Fisher's exact test.


The median gestational age at diagnosis of ventriculomegaly was 24 (range, 15–36) weeks in Group A, 29 (range, 16–39) weeks in Group B and 25 (range, 15–39) weeks in Group C. The number of cases diagnosed before 24 weeks was significantly higher (P = 0.003) in Groups A (46/75; 61%) and C (34/60; 57%) than in Group B (12/41; 29%). Different outcomes of the total population are summarized in Figure 2.

Figure 2.

Flow chart showing the different outcomes of the total population. *Associated with structural and/or chromosomal malformations (see Table 1). TOP, termination of pregnancy.

Ventriculomegaly was more often an isolated finding in Group A (44/75; 58.7%) than in Group B (10/41; 24.4%) (P < 0.001) and Group C (24/60; 40%) (P = 0.047). Structural and/or chromosomal anomalies were found in 31 fetuses (41.3%) of Group A, 31 (75.6%) of Group B and 36 (60%) of Group C (Table 1). All the malformations were diagnosed in utero, although not always at the first examination. In Group A, there were three cases of chromosomal anomalies not associated with structural malformations; the maternal age in these three cases was 25, 30 and 42 years. The ventriculomegaly was unilateral in five cases: four in Group A, one of which had associated malformations (spina bifida), and one in Group B (associated with a nasopharyngeal tumor).

Table 1. Structural and chromosomal anomalies in Groups A, B and C
GroupNo. of casesStructural malformationsChromosomal anomalies
  1. CNS, central nervous system; VSD, ventricular septal defect

A 3 47,XY,+21
 5Omphalocele, ascites, heart malformations47,XY,+18
Pyelectasis (Potter 4)46,XY/47,XY,+8 (mosaicism)
Clinodactyly, dysmorphia47,XX,+21
Common atrioventricular canal47,XX,+21
Brachycephaly, single umbilical artery, polydactyly, VSD, Dandy–Walker syndrome47,XY,+13
 5Spina bifida/myelomeningocele 
 4Agenesis of the corpus callosum 
 3Dandy–Walker syndrome 
 1Cerebral hemorrhage 
10Non-CNS malformations: multiple malformations (7), congenital heart disease (1), others (2) 
B 1Closed choanas, hand deformations, renal cysts and pyelectasis, heart anomalies47,XY,+21
10Spina bifida/myelomeningocele 
 6Agenesis of the corpus callosum 
 2Dandy–Walker syndrome 
 3Other CNS malformations: nasopharyngeal tumor, cephalocele, pachygyria 
 9Non-CNS malformations: genetic syndromes (5), multiple malformations (3), congenital heart disease (1) 
C 2Dysmorphia, sacral meningocele47,XXX
Dysmorphia, clubfoot, claw hands (Klinefelter)47,XXY
10Spina bifida/myelomeningocele 
11Agenesis of the corpus callosum 
 1Dandy–Walker syndrome 
 1Cerebral hemorrhage 
10Non-CNS malformations: multiple malformations (8), diaphragmatic hernia (1), esophageal atresia (1) 

Seven fetuses (23% of the 30 that were tested) presented with an underlying infection: five due to cytomegalovirus (two in Group A, one in Group B and two in Group C) and two due to Toxoplasma gondii (both in Group C). Major sonographic anomalies (hydrops, ascites, cerebral calcifications) were present in all these cases. The male/female ratio was close to 1 in Groups B and C, while males outnumbered females in Group A (46 vs. 29).

TOP was chosen in 65 cases (37% of the total population; 64% of the malformations diagnosed before 24 weeks). Almost half of them (31) were in Group C, 22 in Group A and 12 in Group B. In 14 cases, ventriculomegaly was the only finding and 13 of them were in the severe form. Data on TOP, spontaneous miscarriage, intrauterine death, neonatal death and infant death in each group are reported separately for cases of isolated and non-isolated ventriculomegaly (Table 2).

Table 2. Outcome of 176 fetuses with ventriculomegaly
 Group A (75)Group B (41)Group C (60)Total (176)
Isolated (44)Associated (31)Isolated (10)Associated (31)Isolated (24)Associated (36)
Termination of pregnancy 022 111131865
Spontaneous miscarriage 1 2 0 1 0 1 5
Intrauterine death 0 0 0 0 0 0 0
Neonatal death 0 2 1 2 2 2 9
Infant death 0 0 0 2 1 1 4
Alive at ≥ 24 months of age43 5 815 81493

Since the outcome in cases with structural and/or chromosomal anomalies depends on the type of the associated malformations, we further considered outcome data only for isolated cases. When ventriculomegaly was an isolated finding, 97.7% of fetuses with mild, 80% of those with moderate and 33% of those with severe dilatation were alive at ≥ 24 months. The neurodevelopmental outcome at the age of ≥ 24 months was normal in 93%, 75% and 62.5% of the surviving neonates, respectively (Table 3). There were two cases of mild handicap in Group A (language problems requiring a speech therapist; nystagmus), two cases in Group B (mild intellective impairment requiring a delay in starting primary school and the need for a support teacher; ventriculoperitoneal shunt with normal neurodevelopmental outcome and mild strabismus) and one in Group C (hydrocephalus with ventriculoperitoneal shunt and mild neurodevelopmental delay at 1 year of age).

Table 3. Neurodevelopmental outcome of survivors with isolated ventriculomegaly
Outcome n (%)ABC
  • *

    Three cases of unilateral ventriculomegaly.

  • Group A vs. C, P = 0.042.

  • A case of retinoblastoma.

Normal40* (93)6 (75)5 (62.5)
Mild handicap2 (4.6)2 (25)1 (12.5)
Severe handicap1 (2.3)0 (0)2 (25)

Among the 106 liveborn fetuses, the ventricular dilatation improved and eventually disappeared during intrauterine life in 50 cases, remained unchanged in 37 cases and increased in 19 cases. A normal outcome was found significantly more often when ventriculomegaly improved (92%) than when it remained unchanged (35%) (P < 0.001) or worsened (21%) (P < 0.001) (Table 4).

Table 4. Outcome of the liveborn fetuses according to the evolution of ventriculomegaly in utero
Evolution of ventriculomegalyNormalMild handicapSevere handicapDeadTotal (106)
Improved in:
 Group A3721 40
 Group B51  6
 Group C4  4
  Total n (%)46 (92)3 (6)1 (2) 50
Stable in:
 Group A7 1210
 Group B217414
 Group C413513
  Total n (%)13 (35)2 (5)11 (30)11 (30)37
Worsened in:
 Group A  0
 Group B2141 8
 Group C217111
  Total n (%)4 (21)2 (10.5)11 (58)2 (10.5)19


Abnormal dilatation of the cerebral lateral ventricles was one of the first malformations to be recognized prenatally in the late seventies. The diagnosis of hydrocephaly was made when the ventricle/hemisphere ratio was above the upper limits of reference ranges. Subsequent studies characterized the in utero growth of different parts of the ventricular system and reference values of ventricular width were established throughout pregnancy7, 9–11. The accepted standard today is the measurement of atrial width at the level of the choroid plexus.

Atrial width above 15 mm from 14–16 weeks onward is usually associated with a poor prognosis1. Our study confirms previous data: 60% of the cases with ventricular width ≥ 15 mm had associated structural malformations of the CNS and/or other organs. Pregnancy was terminated in most cases diagnosed before 24 weeks (88%) and there was one case of spontaneous miscarriage. Of the 60 fetuses with severe ventriculomegaly, only 10 (16%) were alive and normally developed at the age of at least 2 years. Even when the analysis was limited to cases of isolated ventriculomegaly, the rate of normal neurodevelopmental outcome was 62.5%. Graham et al.12 found major neurological morbidity at pediatric follow-up in 33% (3/9) of survivors with isolated severe ventriculomegaly. It is noteworthy that we diagnosed only a relatively small percentage of cases (57%) before 24 weeks of gestational age; since it is easy to diagnose severe ventriculomegaly, this means that almost half of severe ventricular dilatations developed later in prenatal life.

The published reports on borderline ventriculomegaly, defined as ventricular width between 10 and 15 mm, are relatively few and often not comparable, especially with regard to neurodevelopmental outcome because of the variable length of follow-up and different methodological approaches4–6, 11–16.

We reviewed 116 such cases which makes ours the largest study. We separately analyzed cases of mild (10 to 12 mm) and moderate (12.1 to 14.9 mm) ventriculomegaly and found that the latter ones were significantly more often associated with structural anomalies (75%) than mild cases (41%). The large variation in the frequency of associated anomalies reported in the literature on borderline ventriculomegaly2, 3, 5, 6, 14, 15, ranging from 10%6 to 76%2, might be due to variation in the proportion of cases with mild and moderate ventriculomegaly; unfortunately, in most studies data were not separately analyzed for cases with atrial width < 12 mm or > 12 mm. Only Vergani et al.5 found that an atrial width < 12 mm was associated with other anomalies in 6% of the cases, compared with 56% when the atrial width was ≥ 12 mm.

Once structural malformations are ruled out, there is still a risk of chromosomal anomalies. The frequency of chromosomal anomalies reported in the literature in these cases ranges from 0 to 14%3, 5, 6, 12, 14, 15, 17, 18; separate analysis for mild and moderate ventriculomegaly is never carried out. Although our sample is too small to draw any final conclusion, our data suggest that, at variance with structural anomalies, the degree of ventriculomegaly is not predictive of the risk of aneuploidies. None of the cases with ventricular width > 12 mm had isolated chromosomal anomalies; there were three isolated aneuploidies in cases of mild ventriculomegaly.

Screening for infections (usually TORCH) is recommended when ventriculomegaly is diagnosed5, 6, 13, 19. However, there are few data to argue in favor or against such an approach. Although it suggests that TORCH infections are frequent (23%), our study is not conclusive because the number of screened cases was small (n = 30).

We collected follow-up data when the infants were at least 24 months old by means of interviews with the parents. Although this way of obtaining information about the condition of infants produces a very high rate of follow-ups, the results must be viewed with caution because parents are sometimes not objective. A strength of our study is that the infants were 24 months or older at the time of the interviews. It is well-known that this age must be reached in order to exclude moderate to severe neuromotor disabilities, and parents usually report this type of damage. A longer follow-up, to school age, and a neurological evaluation are needed to exclude mild motor or behavioral abnormalities which are often not reported by the parents.

Another strength of our study is that we analyzed the data separately according to the degree of ventriculomegaly: mild (10 to 12 mm) and moderate (12.1 to 14.9 mm). Neither group had severe neurodevelopmental outcomes, if we exclude the case of retinoblastoma, which is most likely a chance association. However, the overall outcome was better when the ventriculomegaly was mild: it was normal in 93% of the cases compared with 75% in moderate cases. Vergani et al.5 reported similar results in isolated cases: a significantly lower rate of developmental delay when atrial width was < 12 mm (3%) than when it was 12–15 mm (23%). Pilu et al.6, in a review of the literature on 141 cases, reported a 3.8% rate of abnormal neurodevelopmental outcome when the atrial width was < 12 mm, compared to 14% when it was 12–15 mm6. Recently, Signorelli et al.16 published results in 60 cases of mild isolated ventriculomegaly ⩽ 12 mm, reporting normal neurodevelopmental outcome in 100% of the cases. They suggested considering this atrial width a variant of the norm, once structural and chromosomal anomalies have been excluded. The fact that only in Group A there was a prevalence of male fetuses (61%) would support the observation that male fetuses have a slightly greater atrial width compared to females6.

The evolution of ventriculomegaly in utero seems to be related to the outcome, with better prognosis when ventriculomegaly improves or disappears, independently of the severity at first presentation (mild, moderate or severe).

In conclusion, our results suggest that the definition of mild ventriculomegaly should be limited to a ventricular width below 12 mm. Cases with measurements above this value, but below 15 mm, are more often associated with structural malformations. Moreover, they have a normal neurodevelopmental outcome less frequently, although numbers are too small to reach statistical significance. We suggest that they should be considered separately (and defined as moderate). When ventriculomegaly is found, the diagnostic work-up must include targeted ultrasound examinations and fetal echocardiography since the rate of associated malformations is high and CNS or extraneural anomalies may be present. It must be underscored that in some cases associated structural malformations may be diagnosed later in pregnancy. Although the rate of chromosomal anomalies in isolated ventriculomegaly is fairly low, it still justifies offering fetal karyotyping.


We thank Marilisa Biolcati for the statistical analysis and Dr Peter Christie for reviewing the English language.