Dandy-Walker Phenotype with Brainstem Involvement: 2 Distinct Subgroups with Different Prognosis

DISCUSSION

Neurologic development is variable in patients with DW,¹ with the clinical course depending not only on the degree of cerebellar hypoplasia but also on the extent and severity of associated CNS malformations. In this study, patients with DW with associated brainstem involvement were evaluated to understand the potential imaging variability of malformations in this structure, their potential associations, and its clinical impact on the DW population. Although all patients with DW and brainstem involvement shared characteristic posterior fossa cystic malformation with the required features for the current diagnostic criteria,¹⁵ there were 2 different imaging groups of patients with brainstem involvement: 1) those with the mild form: considered a less severe phenotype in which changes were related only to the size and proportion of the brainstem structures (anterior-posterior embryologic defect);²⁴ and 2) those with the severe form in which an additional dysgenetic appearance of the brainstem was noted, with a peculiar posterior tegmentum fold, bump, and cleft. In line with our concept of the critical role of the brainstem in neurologic functions, these patients with extensive morphologic abnormalities in the brainstem were also more severely affected clinically, including a significantly increased frequency of bulbar and other autonomic nervous system dysfunctions, seizures, and higher mortality rates, compared with the mild form. From the genetic perspective, we noted that genetic abnormalities and associated syndromes in our DW cohort with brainstem malformation (30%) are almost 2 times higher than expected in the overall DW population (16%).³ This observed difference suggests that patients with DW with brainstem malformations may be better candidates for an extensive genetic work-up.

The mild brainstem phenotype was the most frequent form in our cohort (78%), characterized mainly by the presence of pontine hypoplasia and, more rarely, hypoplasia or hyperplasia of other segments of the brainstem, including the medulla and midbrain. Apart from the brainstem dysmorphology resulting from a presumed abnormal embryologic anteroposterior defect, the measurements and appearance of the posterior fossa, including posterior fossa perimeter, fastigial angle blunting, and choroid plexus/taenia tela choroidea complex morphology, were not significantly different from those in the overall cohort already described in 2022.¹⁵ There were also no differences between the additional abnormalities of the brain or clinical features commonly noted among patients with DW.^25,26

Patients with severe brainstem malformation, on the other hand, had additional findings that clearly distinguish them from the mild brainstem DW cohort. In particular, we identified the presence of a posterior tegmental dysplasia as a hallmark of this group. The patients with tegmental brainstem dysplasia characterized by folding, bumps, and clefts in the tegmentum were highly associated with having more severe/extreme vermian hypoplasia, massive hydrocephalus, and larger posterior fossa perimeters compared with the mild brainstem subgroup. Additionally, dysgenesis and agenesis of the corpus callosum and supratentorial interhemispheric cysts were significantly associated with this phenotype. Most interesting, during the retrospective imaging review and analysis, we noted that some of the patients with brainstem tegmentum dysplasia shared similar features already described in tegmental and bulbar cap dysplasias, including focal posterior protuberance of the brainstem with a small and flat pons, features related to an underlying abnormal axonal guidance pathophysiology.^22,23 These phenotypic similarities shed light on the likely embryologic developmental intersection between both disorders.

There is substantial evidence that disruption of the rhombic lip underlies the cerebellar DW.^27,28 The rhombic lip is a progenitor zone located along the dorsal edge of the hindbrain in the rhombomere, one of the anterior hindbrains adjacent to the roof of the fourth ventricle. It gives rise to multiple cell lines that are relevant to the phenotypes we have described in this report. Progenitors from rhombomere 1 rhombic lip develop into cerebellar granule cell progenitors, which migrate over the cerebellar anlage and then undergo massive proliferation, driving most cerebellar growth.²⁸ Earlier in development, these progenitors also give rise to the glutamatergic pontine nuclei neurons, which earlier migrate over the developing cerebellar anlage and continue ventrally to form the pontine nuclei at the brainstem.^18⇓-20 Failure of generation and/or migration of these early-born brainstem neurons in mice results in phenotypes similar to those seen in our patient population.²⁰ Additionally, rhombic lip progenitors from rhombomere 1 and other more posterior hindbrain regions also give rise to choroid plexus epithelial cells, which play a role in the expansion of the fourth ventricle plexus from its lateral edges. Mouse models have shown that very early misspecification of cerebellar rhombic lip progenitors toward the choroid plexus fate, at the expense of cerebellar and brainstem fates, contributes to the expansion of the fourth ventricle roof and both cerebellar and brainstem hypoplasia.^14,29,30

Therefore, the simultaneous association of brainstem and cerebellar malformation and even fourth ventricle expansion is anticipated in patients with DW. The severity of these abnormalities is likely related to the timing and extent of the insult during early embryologic stages. Early insults during the critical period of rhombic lip development can lead to more severe and extensive abnormalities, whereas insults occurring later in development may result in milder phenotypes. Thus, the precise timing and extent of the disruption to the rhombic lip progenitors determine the severity of the disorder, emphasizing the importance of early embryologic events in shaping the neurologic outcomes in individuals with DW.

The variability of neuroimaging findings from mild to severe brainstem groups associated with varied severity of clinical presentations and outcomes as well as different rates of underlying genetic causes reinforce the importance of detailed imaging descriptions and neuroimaging stratification to improve the understanding of these conditions and to better support neurologists and geneticists in the clinical care and genetic counseling of these patients and their families. In patients with DW, the brainstem should be carefully scrutinized for differences in size, shape, and/or proportions. Direct comparison with aged-matched normal brain MR imaging examinations may be useful in questionable cases.

Although our study yielded notable findings, it is important to acknowledge its limitations. First, the retrospective study design and cross-sectional analysis focused primarily on postnatal studies. This approach overlooked the consideration of neurodevelopmental outcomes and prenatal imaging findings, which could have provided valuable insights. As a result, there is the possibility of selection and misclassification bias within our groups, which hinders the establishment of a stronger causal relationship. To enhance the reliability of our conclusions, future research should address these limitations by incorporating prospective designs, comprehensive assessments of prenatal factors, and animal model studies. Another limitation was the small number of severe cases in which DW and posterior dysplasia of the brainstem were noted. However, given that these cases were collected from numerous tertiary referral centers with expertise in pediatric neuroimaging, this small number may reflect the true rarity of severe brainstem dysplasia in postnatal life.