Bilateral Cavitations of Ganglionic Eminence: A Fetal MR Imaging Sign of Halted Brain Development

Discussion

From our preliminary observation, it appears that GE cavitations are an aspect of more complex conditions associated with severe cerebral structural derangement. The awareness of this rare developmental abnormality involving the GE region may have implications in better understanding of the complex malformations caused by defective cellular proliferation and migration, such as lissencephalies. The fact that the lesions were bilateral and symmetric with an inverted “regular” open C shape, an epithelium-like lining, regular margins, and no apparent signs of hemorrhage (including the 2 cases with T1-weighted images) suggests a malformative rather than necrotic-clastic origin. Familial recurrence in 2 cases and the presence of frontal band heterotopias in 1 case further support such a hypothesis. Although the data are not extensive enough to conclude a statistical significance, the available ADC values (cases 1 and 4), albeit showing a slight trend toward mild reduction (0.85 SD, 0.09 m²/s), remained higher than ADC values reported in fetal brain acute ischemia cases (ie, 0.4 m²/s in Righini et al⁸). Moreover, in the case illustrated in On-line Fig 2, GE histology showed a higher cell body/interstitial space ratio (a possible sign of structural developmental anomaly) but not signs of acute cytotoxic edema (such as neuropil swelling). Nevertheless, because we do not have detailed histologic data for all of our cases, we cannot completely rule out the necrotic nature of the observed GE region cavitations. Moreover, the existence of a recent report,⁹ which showed basal ganglia vasculopathy in a lissencephaly case, further warns of the possibility that ischemia may play a role in cavitation onset.

According to Barkovich et al,¹⁰ lissencephalies are primarily caused by abnormal neuronal migration. In such conditions, abnormal development of the GE, together with deficiency of other fetal transient structures such as the germinal matrix, may play a key role. We may speculate that cavitations associated with GE might derive from abnormal proliferation of neural precursors in the region and subsequently affect migration of inhibitory interneurons from GE toward their final cortical destination. However, detailed cytologic-immunohistochemical assessment could not be performed in our cases because of faulty tissue preservation. However, the cystic anomaly of the GE region might contribute to causing lissencephaly even by simply physically affecting cell migration, even in the absence of proliferation defect, because, in at least 2 of our cases, GE was larger than in healthy controls of the same GA.

The presence in our 5 cases of opercularization defects, agenesis-severe hypoplasia of CC, abnormal cortical plate findings at histology, and heterotopic bands, further supports the hypothesis that GE cavitation anomalies are part of complex malformations, which involve cell migration and possibly a cell differentiation phase.

To the best of our knowledge, GE cavitations have not been described in animal models of lissencephaly; neither have they been reported in the existing extensive postnatal neuroimaging literature regarding human lissencephalies. This could be explained by their size decrease as the whole brain progressively grows. In postnatal MR studies, cavitation remnants might have been occasionally interpreted in the past just as enhanced perivascular spaces in the basal ganglia region; cavitations may relatively decrease in size with respect to the progressive growth of the whole brain.

A major limitation of our report is the lack of substantial molecular genetics characterization of our cases. Banded karyotyping, which turned out to be normal, was available in 3 cases (On-line Table 1); however, karyotyping is usually of limited diagnostic power in lissencephaly. Molecular study of the X-linked gene ARX, which was not possible, might have been indicated in cases 2 and 3, although the severe phenotype observed in the sister and parental consanguinity makes it unlikely to be an X-linked inheritance. Converging experimental, neuropathologic, and imaging evidence suggests that mutations in this gene cause an abnormal developmental process primarily affecting GE development. Impaired proliferation of neuronal elements in the GE and subsequent abnormal proliferation of GABAergic interneurons has been demonstrated in the ARX knock out model.¹¹ Microcystic changes in the basal ganglia are observed in the brains of children with X-linked lissencephaly with abnormal genitalia, both at neuropathology¹² and by use of MR imaging scan.¹³ Multiple small areas of abnormal signal intensity, possibly representing microcystic changes, are observed in the basal ganglia of boys with ARX mutations and profound cognitive impairment but no lissencephaly.¹⁴

Albeit limited by the rarity of these findings and by the difficulty of histology studies on fetal human brain before the third trimester of GA, additional cases investigated by cytologic labeling of GE-originating neuroblasts and cortical interneurons are needed to further elucidate the meaning of the anomaly we have reported and to definitively rule out the possibility of a lesional etiology.