PRENATAL DIAGNOSISPrenat Diagn 2006; 26: 707–710.Published online 9 June 2006 in Wiley InterScience(www.interscience.wiley.com) DOI: 10.1002/pd.1485

SHORT COMMUNICATION

Outcome of fetal cerebral posterior fossa anomalies

Audrey Long*, Paul Moran and Stephen Robson1Fetal Medicine Unit, Royal Victoria Infirmary, Newcastle Upon Tyne, UK.

Introduction Limited data exist on the outcome of Dandy–Walker malformation (DWM), Dandy–Walkervariant (DWV) and mega-cisterna magna (MCM). We report the first population-based study of posterior fossaanomalies from the northern region of England.

Methods Cases were identified from the Northern Congenital Abnormality Survey (NorCAS) and regionalFetal Medicine Unit databases for the period 1986–2004 for DWM/V and 1995–2005 for MCM (defined as acisterna magna ≥10 mm). Outcome data was obtained from pediatric records and/or general practitioner/healthvisitor questionnaires for all survivors.

Results A prenatal diagnosis of a posterior fossa abnormality was made in 91 cases, with a further 12 casesof DWM/V diagnosed postnatally, giving incidences of DWM/V and MCM of 1/11574 and 1/8268 birthsrespectively. In five cases where DWM/V was suspected prenatally, the diagnosis was not confirmed. Of the47 with DWM/V, 41 (87%) had additional anomalies. There were three survivors, all with neurodevelopmentaldisability. Of the 39 cases of MCM, 24 (62%) had additional anomalies. There were 30 survivors; one childdied at 3 months and the outcome was normal in 25 children including 12/13 (92%) with isolated MCM.

Summary Posterior fossa anomalies are relatively common. The outcome is very poor in DWM/V owing tothe high rate of associated anomalies. The outcome appears better with MCM, especially if this is an isolatedfinding. Copyright 2006 John Wiley & Sons, Ltd.KEY WORDS: Dandy–Walker malformation; Dandy–Walker variant; mega-cisterna magna; outcome; prenataldiagnosis

INTRODUCTION

Examination of the posterior fossa is part of routineultrasound screening for fetal anomaly. Posterior fossaanomalies can be divided into Dandy–Walker malfor-mation (DWM), comprising cystic dilation of the fourthventricle, enlarged posterior fossa, complete or par-tial agenesis of cerebellar vermis, and elevated tento-rium, Dandy–Walker variant (DWV), comprising vari-able hypoplasia of the cerebellar vermis with or withoutenlargement of the cisterna magna and mega-cisternamagna (MCM) with normal cerebellar vermis and fourthventricle (Ecker et al., 2000; Pilu, 2000; Pilu et al.,2000; Bernard et al., 2001). However, recent evidencefrom MRI imaging suggests DWM, DWV and MCMrepresent a continuum and should be referred to as theDandy–Walker complex (Barkovich et al., 1989). Pre-vious studies of the prenatal detection and outcome ofposterior fossa anomalies have all come from referralcenters. The aim of this population-based study is todescribe the outcome of posterior fossa anomalies in thenorthern region of England.

*Correspondence to: Audrey Long, Fetal Medicine Unit, RoyalVictoria Infirmary, Newcastle Upon Tyne, UK.E-mail: audlong@doctors.org.uk

METHODS

The Northern Congenital Abnormality Survey (Nor-CAS) collects data on all major congenital abnormalitiesin fetuses (including spontaneous abortions and termi-nations of pregnancy), stillbirths and liveborn infants ofmothers resident in the former Northern Region. Thepopulation of approximately 3 million is stable, withlittle migration (Atkins and Hey, 1991; Rankin, et al.,1999; Sparey et al., 2000). The register is voluntary buthas a high notification rate and has collected data since1984 (1992).

Guidelines for screening for cerebral anomalies atthe 18–20 week scan were standardized in 1995 toinclude routine measurement of the cisterna magna andnotification to NorCAS if ≥10 mm. All cases of DWMand DWV notified to NorCAS between 1986 and 2004,and MCM between 1995 and 2004 were identified andthe outcomes reviewed. The Fetal Medicine Unit inNewcastle provides a regional referral service; the unitdatabase was also searched for the same period, andany additional cases identified. Outcome informationwas obtained from pediatric records where availableand a questionnaire sent to general practitioners andhealth visitors. Formal neurodevelopmental assessmentwas not performed. The outcome was designated normalif the child had normal routine developmental checks.Barkovich’s classification of posterior fossa anomalieswas used (Barkovich et al., 1989).

Copyright 2006 John Wiley & Sons, Ltd. Received: 3 January 2006Revised: 14 March 2006

Accepted: 4 April 2006Published online: 9 June 2006

708 A. LONG ET AL.

RESULTS

A prenatal diagnosis of a posterior fossa anomaly wasmade in 91 cases, and a further 12 cases were diagnosedpostnatally. In five cases where DWM/V was suspectedprenatally, the diagnosis was not confirmed (false-positive rate 5%); one was reclassified as normal at asubsequent scan 14 weeks later and the remaining fourhad ventriculomegaly associated with other intracranialpathology at postmortem (cytomegalovirus infection [1],agenesis of corpus callosum [1] and Arnold Chiarimalformation [2]). In one case diagnosed postnatallyon ultrasound imaging, the diagnosis was revised to anarachnoid cyst after CT imaging.

There were 47 cases of confirmed DWM/V diag-nosed prenatally (incidence 1/11574 births); of these15 were DWV and 32 were DWM. The mean (SD)gestation at diagnosis was 20.0 (3.3) weeks and themale : female ratio was 1.5 : 1. Outcomes are shown inTable 1. Associated structural anomalies were presentin 39/47 (83%) cases; of these 86% were detectedby prenatal ultrasound. Karyotyping was performed in32 cases with 13 anomalies (3 unbalanced transloca-tions, 2 deletions, 3 trisomies [18,18,21], 1 triploidyand four sex chromosome anomalies [47,XXY, 47,XXX,49,XXXXY, 45X0/46XX]) of which 11 had additionalstructural anomalies. Thirty-eight (81%) families optedfor termination of pregnancy (TOP). Five infants wereliveborn and three survived (median [range] follow-up96 [10–156] months), all with neurodevelopmental dis-advantage. However, in one case (with a left hemiple-gia), this was secondary to cardiac surgery for Tetral-ogy of Fallot. Postmortem examination was available in24/44 (54.5%) terminations and deaths, and confirmedDWM/V in 22 (91.6%), with ‘suboptimal visualization’of the posterior fossa in 2.

There were 39 cases of MCM diagnosed prenatally(incidence 1/8268 births). The mean (SD) gestation at

diagnosis was 28.9 (5.2) and the mean gestation at deliv-ery was 36.8 (5.2) weeks. The median [range] sizeof the cisterna magna was 10 [10–20] mm and themale : female ratio was 3.75 : 1. Outcomes are shown inTable 1. Associated structural anomalies were presentin 24/39 (62%) cases, of which 96% were detected byprenatal ultrasound. Seven cases underwent karyotypingwith two anomalies; trisomy 21 in a case with isolatedMCM and mosaic trisomy 7q monosomy t9p associ-ated with ventriculomegaly. Three (8%) families optedfor TOP. Thirty-five infants were liveborn and 30 sur-vived, of which 25 were normal at follow-up (median[range] 36 [12–96] months). Postmortem examinationwas available in one neonatal death (age 1 day), whichshowed a left frontal lobe hemorrhage due to alloim-mune thrombocytopenia.

DISCUSSION

We report the first population-based study of posteriorfossa anomalies diagnosed in the north of England overan 18-year period. Although formal neurodevelopmentalassessment was not undertaken, we obtained follow-upinformation on all cases. Although there is evidence thatDWM/V and MCM represent a continuum (Barkovichet al., 1989), we report the outcome data separatelybecause they were so different; no fetus with DWM/Vhad a normal outcome in contrast to 92% of fetuses withisolated MCM.

The incidences of DWM/V and MCM in our serieswere 1/11574 and 1/8268 births respectively. There isno data on the prenatal incidence of posterior fossaanomalies. DWM is usually quoted as having an inci-dence of 1/25 000–35 000 based on the series of (Hirschet al., 1984). However, this figure was indirectly derivedfrom a postnatal population with hydrocephalus. It has

Table 1—Outcome of posterior fossa anomalies

OutcomeNeurodevelopmental

outcome (LB)N seen in

Category N FMU TOP FD NND Survivors Normal Abnormal

DWM/V Prenatal diagnosis Isolated 6 5 4 2 0 0 n/a n/aAssociated abnormalitiesa 41a 31 34 2 2 3 0 3b

Postnatal diagnosis Isolated 3 0 0 0 0 3 2 1Associated abnormalities 8 2 4 0 2 2 0 2

MCM Prenatal diagnosis Isolated 13 12 0 0 0 13 12 1d

Associated abnormalitiesc 26c 24 3 1 5 17e 13 3f

N, number; FMU, fetal medicine unit; TOP, termination of pregnancy; FD, fetal death; NND, neonatal death; LB, live birth; DWM/V,Dandy–Walker malformation/variant; MCM, mega-cisterna magna.a Fifteen cases involved one organ system (cerebral [n = 7—all ventriculomegaly], cardiac [n = 2], renal [n = 1], facial [n = 1] and skeletal[n = 1]), while 26 cases had multiple anomalies (including 20 with cerebral anomalies [12 with ventriculomegaly and five with agenesis corpuscallosum]).b Age 13—left hemiplegia from right-sided CVA (following repair of cardiac defect), normal intellectual development. Age 7—mild globaldevelopmental delay, epilepsy, behavioral problems; Age 4—severe expressive speech delay.c Twenty-two cases involved one organ system (cerebral [n = 12—ventriculomegaly in 11, agenesis of corpus callosum in 1], renal [n = 8],cardiac [n = 1], liver [n = 1], while two cases had multiple anomalies (both included cerebral anomalies [1 with ventriculomegaly]).d Age 4—delayed motor development (poor feeding and delayed walking [29 months]).e One death at 3 months.f Age ten months—VP shunt, HC 98th centile, delayed motor development; Age 3—delayed speech/language and deafness; Age 8—delayedcognitive and speech/language development (diagnosis congenital cerebral ataxia).

Copyright 2006 John Wiley & Sons, Ltd. Prenat Diagn 2006; 26: 707–710.DOI: 10.1002/pd

OUTCOME OF FETAL CEREBRAL POSTERIOR FOSSA ANOMALIES 709

been previously reported that there is a slight femalepreponderance (Hirsch et al., 1984; Klein et al., 2003)for DWM, whereas we found a male preponderance of1.5 : 1 for DWM/V and 3.75 : 1 for MCM.

Excluding terminations, the mortality of DWM/V inour series was 66%, slightly higher than the figuresof 35–55% previously reported (Russ et al., 1989;Estroff et al., 1992). This partly reflects the high rateof associated anomalies (87%), the majority of whichwere detected prenatally. Normal neurodevelopmentaloutcome in prenatal series has ranged from 20–50%(Estroff et al., 1992; Aletebi and Fung 1999; Eckeret al., 2000; Kolble et al., 2000) with higher rates of60–75% in postnatal series (Hirsch et al., 1984; Goldenet al., 1987). We are unable to draw meaningful con-clusions about neurodevelopmental outcomes in theDWM/V group, as only three survived, all with handi-caps.

Associated structural anomalies were less commonin the MCM group. The majority involved the brainwith ventriculomegaly being present in 12/22 (54.5%)of cases. Although termination rates were low (7.6%),there was still a high fetal/neonatal mortality rate inthe MCM group (16.7%), mostly related to associ-ated anomalies. MCM is reported to be associatedwith aneuploidy, in particular, trisomy 18 (Steigeret al., 1995; Ek et al., 1998). In our series, therewere only two known karyotype abnormalities, bothof which had additional anomalies and neither weretrisomy 18. Neurodevelopmental outcome was normalin 25/29 survivors, and this was more likely wherethe MCM was isolated (12/13). Thus, it would appearthat isolated MCM carries a favorable prognosis. Thisis supported by one small series of 15 cases diag-nosed in the third trimester, which resulted in a nor-mal pregnancy and neonatal outcome (Haimovici et al.,1997).

There was a high concordance between postmortemand prenatal ultrasound diagnosis of DWM/V (91.6%)in our series, which compares well with previous studieswhere the concordance has been as low as 43% (Carrollet al., 2000). In our series, no diagnosis was madeprior to 18 weeks when development of the cerebellarvermis may not have been complete (Bromley et al.,1994). The diagnosis of DWM/V was not confirmedin five cases but in two cases visualization of theposterior fossa was recorded as ‘suboptimal’. Four ofthe five cases had other significant intracranial pathology(ventriculomegaly and agenesis of corpus callosum) andthe fifth had a karyotype abnormality.

This series predates the introduction of fetal MRI,which may be a useful adjunct to prenatal diagno-sis, especially in the third trimester (Yuh et al., 1994;Adamsbaum et al., 2005). Use of half-Fourier acquisi-tion single-shot spin echo (HASTE) sequences allowsoptimal visualization of the posterior fossa (Stazzoneet al., 2000). MRI may be of prognostic value; abnor-mal lobulation of the cerebellar vermis has been associ-ated with abnormal development (Boddaert et al., 2003;Klein et al., 2003).

In summary, posterior fossa anomalies are rela-tively common. The majority of DWM/V had asso-ciated anomalies and termination rates were veryhigh. Although associated anomalies are also com-mon with MCM (66.7%), the outcome appears bet-ter, being apparently normal in 92% of survivorswith isolated MCM and in 81% with associatedanomalies.

REFERENCES

1992. Fetal abnormality: an audit of its recognition and management.Northern Regional Survey Steering Group. Arch Dis Child 67(7Spec No): 770–774.

Adamsbaum C, Moutard ML, Andre C, et al. 2005. MRI of the fetalposterior fossa. Pediatr Radiol 35(2): 124–140.

Aletebi FA, Fung KF. 1999. Neurodevelopmental outcome afterantenatal diagnosis of posterior fossa abnormalities. J UltrasoundMed 18(10): 683–689.

Atkins AFJ, Hey EN. 1991. The northern regional fetal abnormalitysurvey. In Antenatal Diagnosis of Fetal Abnormalities, Donnai D,Drife JO (eds). Springer Verlag: London; 13–30.

Barkovich AJ, Kjos BO, Norman D, et al. 1989. Revised classifica-tion of posterior fossa cysts and cystlike malformations based on theresults of multiplanar MR imaging. AJR Am J Roentgenol 153(6):1289–1300.

Bernard JP, Moscoso G, Renier D, et al. 2001. Cystic malformationsof the posterior fossa. Prenat Diagn 21(12): 1064–1069.

Boddaert N, Klein O, Ferguson N, et al. 2003. Intellectual prognosisof the Dandy–Walker malformation in children: the importance ofvermian lobulation. Neuroradiology 45(5): 320–324.

Bromley B, Nadel AS, Pauker S, et al. 1994. Closure of the cerebellarvermis: evaluation with second trimester US. Radiology 193(3):761–763.

Carroll SG, Porter H, Abdel-Faltah S, et al. 2000. Correlation ofprenatal ultrasound diagnosis and pathologic findings in fetal brainabnormalities. Ultrasound Obstet Gynecol 16(2): 149–153.

Ecker JL, Shipp TD, Bromley B, et al. 2000. The sonographicdiagnosis of Dandy–Walker and Dandy–Walker variant: associatedfindings and outcomes. Prenat Diagn 20(4): 328–332.

Ek S, Anadakumar C, Wang YC, et al. 1998. Enlargement of cisternamagna as an indicator of chromosomal abnormalities in a low-riskAsian population. J Perinat Med 26(4): 325–327.

Estroff JA, Scott MR, Benacerraf BR, et al. 1992. Dandy–Walkervariant: prenatal sonographic features and clinical outcome.Radiology 185(3): 755–758.

Golden JA, Rorke LB, Brue DA, et al. 1987. Dandy–Walkersyndrome and associated anomalies. Pediatr Neurosci 13(1):38–44.

Haimovici JA, Doubilet PM, Benson LB, et al. 1997. Clinicalsignificance of isolated enlargement of the cisterna magna(>10 mm) on prenatal sonography. J Ultrasound Med 16(11):731–734, quiz 735–6.

Hirsch JF, Pierre-Kahn A, Renier D, et al. 1984. The Dandy–Walkermalformation. A review of 40 cases. J Neurosurg 61(3): 515–522.

Klein O, Pierre-Kahn A, Boddaert N, et al. 2003. Dandy–Walkermalformation: prenatal diagnosis and prognosis. Childs Nerv Syst.19: 484–489.

Kolble N, Wisser J, Kurmanavicius J, et al. 2000. Dandy–walkermalformation: prenatal diagnosis and outcome. Prenat Diagn 20(4):318–327.

Pilu G. 2000. The Dandy–Walker complex and fetal sonography.Ultrasound Obstet Gynecol 16: 115–117.

Pilu G, Perolo A, Falco P, et al. 2000. Ultrasound of the fetal centralnervous system. Curr Opin Obstet Gynecol 12(2): 93–103.

Rankin J, Dillon E, Wright C, et al. 1999. Congenital anteriorabdominal wall defects in the north of England, 1986–1996:occurrence and outcome. Prenat Diagn 19(7): 662–668.

Russ PD, Pretorius DH, Johnson M, et al. 1989. Dandy–Walkersyndrome: a review of fifteen cases evaluated by prenatalsonography. Am J Obstet Gynecol 161(2): 401–406.

Copyright 2006 John Wiley & Sons, Ltd. Prenat Diagn 2006; 26: 707–710.DOI: 10.1002/pd

710 A. LONG ET AL.

Sparey C, Jawaheer G, Barrett AM, et al. 2000. Esophageal atresiain the Northern Region Congenital Anomaly Survey, 1985–1997:prenatal diagnosis and outcome. Am J Obstet Gynecol 182(2):427–431.

Stazzone MM, Hubbard AM, Bilaniuk LT, et al. 2000. Ultrafast MRimaging of the normal posterior fossa in fetuses. AJR Am JRoentgenol 175(3): 835–839.

Steiger RM, Porto M, Lagrew DC, et al. 1995. Biometry of the fetalcisterna magna: estimates of the ability to detect trisomy 18.Ultrasound Obstet Gynecol 5(6): 384–390.

Yuh WT, Nguyen HD, Fisher DJ, et al. 1994. MR of fetal centralnervous system abnormalities. AJNR Am J Neuroradiol 15(3):459–464.

Copyright 2006 John Wiley & Sons, Ltd. Prenat Diagn 2006; 26: 707–710.DOI: 10.1002/pd