In vivo and in vitro models of medulloblastomas and other primitive neuroectodermal brain tumors of childhood.

Recent advances in understanding the basic biology of the neoplastic cells that populate childhood primitive neuroectodermal tumors (PNET) of the central nervous system (CNS) underline several unique properties of these common pediatric brain neoplasms. For example, studies of posterior fossa cerebellar medulloblastomas (MB), a prototypical group of brain tumors that comprise the largest class of PNET, suggest that the molecular phenotype of subpopulations of neoplastic cells in MB partially recapitulates stages in the acquisition of the neuronal phenotype by normal developing human CNS progenitor cells. However, as reviewed here, it appears that the neoplastic cells in MB exhibit one or more molecular defects in the sequence of normal maturational events that enable CNS progenitor cells to exit the cell cycle, become committed to the neuronal lineage, and undergo terminal differentiation into fully mature, permanently postmitotic CNS neurons. Indeed, since PNET emerge almost exclusively in early childhood, the induction of PNET may result from genetic lesions that arise in developing CNS progenitor cells thereby preventing these neural precursors from executing normal programs of lineage commitment and differentiation in the CNS. Clarification of how lineage commitment and maturation in PNET comprised of neuron-like tumor cells deviate from normal CNS development may clarify how oncogenes and tumor suppressor genes exert their effects in a cell type specific manner at different stages in the normal maturation of CNS cells. Recently, a number of potentially effective in vitro and in vivo model systems of PNET have been developed. Since these model systems could facilitate efforts to elucidate mechanisms of neoplastic transformation and tumor progression in the CNS, we review the potential utility of several recently described in vitro (e.g., MB cell lines) and in vivo (e.g., transgenic mice) experimental systems as models of authentic childhood CNS neoplasms.