Cerebellar Hypoplasia

Cerebellar hypoplasia (CH) is a purely descriptive term related to a reduction of the cerebellar size and volume^1,2. Hypoplasia may be global or focal (unilateral, with or without vermian involvement) and is associated with a highly heterogeneous group of diseases, developmental and acquired. When the hypoplastic process affects only the cerebellar vermis it should be called vermian hypoplasia and is not included in the scope of this chapter.

213000 Cerebellar hypoplasia
 

0001321 Cerebellar hypoplasia
0007360 Cerebellar hypoplasia / atrophy
0100307 Cerebellar hemisphere hypoplasia
 

The prenatal prevalence of CH is not known, with few single center reports of pediatric patients only. The incidence varies according to the etiology (prenatal infection, exposure to teratogens, chromosomal aberrations, metabolic disorders and other monogenic syndromes and complex brain malformations)³.

Etiologies can be divided into primary (malformative, genetic or metabolic) and secondary lesions (disruptive, post-hemorrhagic, acquired (such as prenatal infections, teratogen exposure)³. Since the introduction of fetal MRI and recent advances in molecular biology, the number of conditions associated with CH (isolated or as part of a complex syndrome) is rising constantly. CH of monogenic origin can have either autosomal recessive, dominant or X-linked inheritance. Regarding chromosomal anomalies, CH is the classic cerebellar malformation associated to trisomy 18 (and not Dandy Walker Malformation, as previously thought; in some cases, the vermian component is severely affected, mimicking DWM).

The cerebellar involvement is often heterogeneous: the whole cerebellum may be hypoplastic in equal or variable scale; the hypoplasia may selectively involve either the vermis alone, only the hemispheres (uni or bilateral) sparing the vermis, or one hemisphere (unilateral cerebellar hypoplasia) including or not the vermis ³. Disruptive lesions often do not follow a constant pattern and may compromise differently the cerebellar structures in individual patients. Postnatally, specific imaging patterns may provide a clue for the underlying genetic cause 4, but during intrauterine life these patterns are not always clear, especially in the second trimester. If the hypoplasia affects only the vermis, it is called vermian hypoplasia, a condition outside the scope of this chapter. Of note, hypoplasia/dysplasia may affect also the entire vermis or only part of it and any part may be involved, not only the inferior vermis as previously thought.  In terms of prognosis, it is important to differentiate between CH and cerebellar atrophy (loss of cerebellar parenchyma with secondary enlargement of the infratentorial spaces).⁵ Nevertheless, in utero this differential diagnosis is quite challenging given that the fetal brain is changing and evolving constantly.

Recurrence risk depends on the specific etiology and, therefore, each case requires tailored counseling. In autosomal recessive entities the recurrence rate is 25%, such as in Smith–Lemli–Opitz syndrome, molybdenum cofactor deficiency, isolated sulfite oxidase deficiency. Rhombencephalosynapsis is mostly sporadic and no specific mutation has yet been identified. Trisomies 13, 17, 18, 21, partial trisomy 12q, monosomy 21q, monosomy 1p36, ring chromosome 6, microdeletion 13q12.3-q14.11, deletion 1q44, deletion 22q11.2 are mostly de novo. Cerebellar hypoplasia as part of lissencephalic syndromes are mostly related to de novo mutations (DCX, LIS1 and tubulin genes). In some cases, autosomal recessive inheritance is described, such in RELN and VLDLR mutations. The α-dystroglycanopathies have frequently a recessive pattern of inheritance such as in POMT, FCMD and LARGE mutations³. 

The finding of a small transverse cerebellar diameter (TCD) is the first sonographic sign to raise suspicion of CH, regardless the etiology. Attalah et al published the only study addressing specifically the clinical significance of a small-for-gestational-age TCD as an independent factor for abnormal outcome. The authors propose a threshold of TCD below the 5th centile / below -2 SD for growth restricted fetuses for the identification of fetuses at high risk for chromosomal anomalies, major structural anomalies, and adverse neurodevelopmental outcome, independently of the presence or absence of additional findings6. Another study (in abstract form only) assessed 27 patients with TCD below the 10th centile and mean referral gestational age of 25.¹ W. Associated structural anomalies were found in 85% of cases and included CNS anomalies (70%) and non-CNS anomalies. In 26% of cases a genetic cause was detected (either chromosomal, or genic) Nine families opted for continuation of pregnancy and normal postnatal development was observed only in 2 children (22% of live births).   

It is important to notice that in many cases, and especially in CH of neurodegenerative causes, the TCD will be normal or near-normal in mid-gestation and the decline in cerebellar growth will be noticed only in the 3rd trimester.  This fact justifies the TCD measurement in third trimester examinations, especially in high risk cases and in countries in which late termination of pregnancy is allowed in the presence of severe neurological disease. 

The presentation in axial views will vary depending on the way the cerebellum is affected. The classic presentation will be a reduced TCD with a normal or near normal cerebellar morphology. The CM may or may not be dilated (>10mm) in this cases. Of note, if the CM is enlarged in the presence of a small TCD the diagnosis involves CH and not megacisterna magna or arachnoid cyst. The morphology of the hemispheres, vermis, brainstem structures,7 4th ventricle, along with the supratentorium, should be strictly analyzed (either by dedicated neurosonography and/or MRI) and also followed up into the 3rd trimester in order to obtain additional information for the diagnosis. If the cerebellum is round-shaped the diagnosis of complete rhombencephalosynapsis shall be suspected (see VISUOG chapter on RES). In some cases an abnormal 4th ventricular  shape is observed, with inversion of its proportions, being longer in the antero-posterior axis than large latero laterally.⁸ If an associated pontine hypoplasia is observed, the diagnosis should be amplified to pontocerebellar hypoplasia and not CH. If an open 4th ventricle to the cisterna magna is present (after 18 weeks), indicating a possible vermian anomaly, the first step should be to exclude Blake’s pouch cyst, and Dandy Walker malformation. Cerebellar hypoplasia with significant vermian component is a feature of different syndromes, among them Walker-Warbug Syndrome and trisomy 18. Of note, most ciliopathies have severe vermian hypoplasia without compromise of the hemispheres that may be even enlarged (as in 20% of the Joubert syndrome cases).
The clues for the definite etiological diagnosis will depend on the presence of additional findings as described below in the section on differential diagnosis. Prenatally, the relatively low resolution of US and MRI (compared with postnatal imaging), the timeline changes of the cerebellum throughout gestation and the imaging overlap between entities often pose a challenge to the accurate diagnosis using imaging only. The highest diagnostic yield for CH in the genetic evaluation will come from whole exome sequencing (and not from karyotyping or microarray). 

Postnatal cerebellar hypoplasia-related entities were classified into five big groups adapted from Poretti et al3: a. Unilateral Cerebellar Hypoplasia (UCH); b. CH with mainly vermian involvement (not discussed in this chapter); c. Global CH with involvement of both vermis and hemispheres; d. Pontocerebellar hypoplasia (not discussed in this chapter); e. CH associated with malformations of cortical development.

Unilateral Cerebellar Hypoplasia (UCH) is secondary to brain disruption (hypoxic-ischemic-hemorrhagic) ⁹. It is mostly acquired and frequently occurs after an event of cerebellar hemorrhage with further atrophy of the affected side vermis. It has a wide spectrum of presentation, varying from the complete absence of one cerebellar hemisphere to different degrees of asymmetry between sides, abnormal folia, cerebellar clefts and variable vermian involvement. It is frequently diagnosed only late in the second / third trimester, after the acute hemorrhagic phase has passed and the affected structures have  undergone complete or partial aplasia^10,11. In one series, the mean gestational age at diagnosis was 26 weeks⁸. Conditions that predispose to brain disruption may be present such as mutations in COL4A1 gene¹². There are also reports of UCH secondary to congenital infections, such as CMV13 and Zika virus. UCH is present in about 75% of PHACE(S) syndrome patients with posterior fossa involvement, due to persistent embryonic carotid-basilar vascular connections leading to cerebellar unilateral disruption There are two reports of UCH in patients with osteogenesis imperfecta and mutations in WNT1. When the vermis is compromised in the hemorrhage, different degrees of vermian dysgenesis may be observed and these cases are related to poorer neurodevelopmental outcomes ³. Isolated UCH cases with vermian sparing mostly have completely normal postnatal outcomes.

Global CH with involvement of both vermis and hemispheres includes the largest number of entities.
a)    Cerebellar agenesis may be either developmental (PTF1A mutations) or disruptive

b)    Congenital infections, particularly cytomegalovirus and Zika virus. Cerebellar involvement is more severe with fetal infection earlier in pregnancy when compared to seroconversion in the third trimester; The pattern of cerebellar involvement in these cases includes global cerebellar hypoplasia with reduced foliation, secondary to a decrease in the proliferation of granule neurons in the external granular layer, abnormal cerebellar cortical migration, and altered morphology of the Purkinje cells^14,15.

c)    Teratogens exposure such as anticonvulsant drugs (e.g., phenytoin and valproic acid), retinoic acid, alcohol and cocaine;

d)    Rhomboencephalosynapsis (RES) - described in details elsewhere in VISUOG Encyclopedia;

e)    Chromosomal aberrations, classically featuring in trisomies 13 and 18, but also in trisomy 21, partial trisomy 12q and monosomy 21q, trisomy 17 mosaicism, monosomy 1p36, ring chromosome 6, X;8 translocation and monosomy X;

f)    Metabolic disorders as in Adenylosuccinase deficiency (a disease of purine metabolism), Zellweger syndrome, nonketotic hyperglycinemia, mitochondrial disorders, Molybdenum cofactor deficiency, isolated sulfite oxidase deficiency and mucopolysaccharidoses. Many of these disorders have an autosomal recessive inheritance;

g)    Other genetic syndromes as in Smith–Lemli–Opitz syndrome, Ritscher‐Schinzel (craniocerebello‐cardiac) syndrome, Hoyeraal‐Hreidarsson syndrome, CHARGE, Endosteal sclerosis, Delleman syndrome (Oculocerebrocutaneous), More rarely may be part of Neurofobromatosis type 1, Aicardi-Goutierrez Syndrome (Pseudo-TORCH), oculodentodigital syndrome, Cohen syndrome, Cri du chat, Pallister–Killian, Galloway–Mowat, and Sengers syndromes³.

CH associated with malformations of cortical development: Including lissencephaly (LIS1, DCX, ARX), polymicrogyria, dysgyria, periventricular nodular heterotopia (FLNA mutations), primary microcephaly typically in association with RELN, VLDLR, and TUBA1A mutations and cerebellar dysplasias. Cerebellar dysplasias, which are characterized by abnormal morphology of the folia and fissures along with microscopic features of abnormal cerebellar white matter and gray-white matter junction. The dysplasia may be global, as in the Alpha-dystroglycanopathies, presenting with Cobblestone Malformation (previous Lissencephaly type 2) and congenital muscle dystrophy. This pattern may result from mutations in 15 genes responsible for the O‐ and rarely N‐glycosylation of a‐dystroglycan. Different phenotypes have been described: Fukuyama congenital muscular dystrophy, muscle‐ eye‐brain disease, and Walker–Warburg syndrome.in tubulinopathies or focal as in Joubert syndrome and related disorders. These disorders are also described in more details elsewhere in VISUOG Encyclopedia.

The clinical phenotype associated with CH is wide and depends also on associated brain malformations or additional unrelated symptoms. The main presentation is muscular hypotonia and global developmental delay. Children develop cerebellar signs only later on. Neurological findings include truncal ataxia (49–93%), hypotonia (47–49%), ocular movement disorders (40–46%), dysarthria (38%), intention tremor (9–35%), microcephaly (20%) and seizures (28–56%). Intellectual disability is present (>60%) and is severe in 35% of them. Speech and language disorders range from mild impairment to total absence of language development, behavioral abnormalities are common and so are autistic features (5–20%)³.    
In UCH, the surface loss of cerebellar hemisphere does not always correlate with poor prognosis. UCH with normal vermis and no associated findings can be associated with normal outcome. In one series, 24 out of 26 were live born and 16 had normal development in infancy even with cerebellar surface loss prenatally was >50% in the majority of cases. The infants with severe developmental delay had severely restricted growth, PHACE syndrome, congenital CMV infection, and some even had normal vermis^13. 

1.     Boltshauser E. Cerebellum?small brain but large confusion: A review of selected cerebellar malformations and disruptions. Am J Med Genet. 2004;126A(4):376-385. doi:10.1002/ajmg.a.20662
2.     Schmahmann JD, Boltshauser E. Cerebellar Disorders in Children. London: Mac Keith Press; 2012. http://search.ebscohost.com/login.aspx?direct=true&db=nlebk&AN=503758&site=ehost-live.
3.     Poretti A, Boltshauser E, Doherty D. Cerebellar hypoplasia: Differential diagnosis and diagnostic approach. Am J Med Genet Part C Semin Med Genet. 2014;166(2):211-226. doi:10.1002/ajmg.c.31398
4.     Graham JM, Spencer AH, Grinberg I, et al. Molecular and neuroimaging findings in pontocerebellar hypoplasia type 2 (PCH2): Is prenatal diagnosis possible? Am J Med Genet Part A. 2010;152(9):2268-2276. doi:10.1002/ajmg.a.33579
5.     Poretti A, Wolf NI, Boltshauser E. Differential diagnosis of cerebellar atrophy in childhood. Eur J Paediatr Neurol. 2008;12(3):155-167. doi:10.1016/j.ejpn.2007.07.010
6.     Atallah A, Guibaud L, Gaucherand P, Massardier J, Desportes V, Massoud M. Fetal and perinatal outcome associated with small cerebellar diameter based on second or third-trimester ultrasonography. Prenat Diagn. 2019. doi:10.1002/pd.5465
7.     Leibovitz Z, Haratz KK, Malinger G, Shapiro I, Pressman C. Fetal posterior fossa dimensions: normal and anomalous development assessed in mid-sagittal cranial plane by three-dimensional multiplanar sonography: Fetal posterior fossa biometry. Ultrasound Obstet Gynecol. 2014;43(2):147-153. doi:10.1002/uog.12508
8.     Haratz KK, Shulevitz SL, Leibovitz Z, et al. Fourth ventricle index: sonographic marker for severe fetal vermian dysgenesis/agenesis. Ultrasound Obstet Gynecol. 2019;53(3):390-395. doi:10.1002/uog.19034
9.     Boltshauser E, Steinlin M, Martin E, Deonna T. Unilateral cerebellar aplasia. Neuropediatrics. 1996;27(1):50-53. doi:10.1055/s-2007-973748
10.     Malinger G, Zahalka N, Kidron D, Ben-Sira L, Lev D, Lerman-Sagie T. Fatal outcome following foetal cerebellar haemorrhage associated with placental thrombosis. Eur J Paediatr Neurol. 2006;10(2):93-96. doi:10.1016/j.ejpn.2006.02.002
11.     Malinger G, Lev D, Lerman‐Sagie T. The fetal cerebellum. Pitfalls in diagnosis and management. Chitty L, Pilu G, eds. Prenat Diagn. 2009;29(4):372-380. doi:10.1002/pd.2196
12.     Vermeulen RJ, Peeters-Scholte C, Van Vugt JJM, et al. Fetal origin of brain damage in 2 infants with a COL4A1 mutation: fetal and neonatal MRI. Neuropediatrics. 2011;42(1):1-3. doi:10.1055/s-0031-1275343
13.     Massoud M, Cagneaux M, Garel C, et al. Prenatal unilateral cerebellar hypoplasia in a series of 26 cases: Significance and implications for prenatal diagnosis. Ultrasound Obstet Gynecol. 2014;44(4):447-454. doi:10.1002/uog.13217
14.     Haratz KK, Lerman-Sagie T. Prenatal diagnosis of brainstem anomalies. Eur J Paediatr Neurol. 2018;22(6). doi:10.1016/j.ejpn.2018.06.011
15.     Cekinovic D, Lisnic VJ, Jonjic S. Rodent models of congenital cytomegalovirus infection. Methods Mol Biol. 2014;1119:289-310. doi:10.1007/978-1-62703-788-4_16

Please note that the maximum upload size is 5MB, and larger images and video clips can be sent to [email protected].