The anatomy of neuroepithelial tumours

Kevin Akeret,Michael Weller,Niklaus Krayenbühl
DOI: https://doi.org/10.1093/brain/awad138
IF: 14.5
2023-08-01
Brain
Abstract:Many neurological conditions conceal specific anatomical patterns. Their study contributes to the understanding of disease biology and to tailored diagnostics and therapy. Neuroepithelial tumours exhibit distinct anatomical phenotypes and spatiotemporal dynamics that differ from those of other brain tumours. Brain metastases display a preference for the cortico-subcortical boundaries of watershed areas and have a predominantly spherical growth. Primary CNS lymphomas localize to the white matter and generally invade along fibre tracts. In neuroepithelial tumours, topographic probability mapping and unsupervised topological clustering have identified an inherent radial anatomy and adherence to ventriculopial configurations of specific hierarchical orders. Spatiotemporal probability and multivariate survival analyses have identified a temporal and prognostic sequence underlying the anatomical phenotypes of neuroepithelial tumours. Gradual neuroepithelial de-differentiation and declining prognosis follow (i) an expansion into higher order radial units; (ii) a subventricular spread; and (iii) the presence of mesenchymal patterns (expansion along white matter tracts, leptomeningeal or perivascular invasion, CSF spread). While different pathophysiological hypotheses have been proposed, the cellular and molecular mechanisms dictating this anatomical behaviour remain largely unknown. Here we adopt an ontogenetic approach towards the understanding of neuroepithelial tumour anatomy. Contemporary perception of histo- and morphogenetic processes during neurodevelopment permit us to conceptualize the architecture of the brain into hierarchically organized radial units. The anatomical phenotypes in neuroepithelial tumours and their temporal and prognostic sequences share remarkable similarities with the ontogenetic organization of the brain and the anatomical specifications that occur during neurodevelopment. This macroscopic coherence is reinforced by cellular and molecular observations that the initiation of various neuroepithelial tumours, their intratumoural hierarchy and tumour progression are associated with the aberrant reactivation of surprisingly normal ontogenetic programs. Generalizable topological phenotypes could provide the basis for an anatomical refinement of the current classification of neuroepithelial tumours. In addition, we have proposed a staging system for adult-type diffuse gliomas that is based on the prognostically critical steps along the sequence of anatomical tumour progression. Considering the parallels in anatomical behaviour between different neuroepithelial tumours, analogous staging systems may be implemented for other neuroepithelial tumour types and subtypes. Both the anatomical stage of a neuroepithelial tumour and the spatial configuration of its hosting radial unit harbour the potential to stratify treatment decisions at diagnosis and during follow-up. More data on specific neuroepithelial tumour types and subtypes are needed to increase the anatomical granularity in their classification and to determine the clinical impact of stage-adapted and anatomically tailored therapy and surveillance.
What problem does this paper attempt to address?