DIAGNOSIS - Primary diffuse leptomeningeal primitive neuroectodermal tumor (PDL PNET)
One of the first mentions of primary diffuse leptomeningeal (PDL) tumors, in which no systemic or central nervous system intraparenchymatous lesion was found reaches back to the twenties of the last century with the description of a case of "sarcomatosis" of the meninges, and a case of invasion of the leptomeninges by "undifferentiated embryonic cells", these broad terms based on morphological grounds only (2). Subsequently, as neuropathological accuracy for the identification of neoplastic cells nature improved, more accurate terms were introduced like PDL gliomatosis, PDL ependymoblastoma and oligodendroglioma, leptomeningeal melanoma, primary leptomeningeal melanomatosis, or true PDL sarcomatosis.
The last WHO classification of tumor of the central nervous system (CNS) clearly defines CNS/PNET as an heterogeneous group of embryonal tumors composed of poorly differentiated neuroepithelial cells which have the capacity for divergent differentiation along neuronal, astrocytic, and ependymal lines, and which arise in the cerebral hemispheres, brain stem, or spinal cord (9).
The morphological and immunohistochemical profile of this case was clearly consistent with a CNS/ supra tentorial PNETs (6, 9), the lack of immunostaining for CD99 and the absence of EWS on cytogenetical analysis definitely ruling out a Ewing sarcoma/ "peripheral" PNET (3), and the absence of imbalance on MYCC gene partially helping to discard a medulloblastoma (8). Other CNS/PNETs tumors, like the medulloepithelioma or the ependymoblastoma, as other embryonal tumors, such as the atypical teratoid/rhabdoid tumour, could also be ruled out on histopathological grounds.
However, literature concerning PDL PNETs is scarce and frequently discloses a considerable difficulty in properly labelling the described tumors, though the differentiation from PDL medulloblastomas is not always clear-cut. Indeed, medulloblastomas presenting with leptomeningeal dissemination without a true mass lesion, as shown in one case by MRI (10), and in another by CT scan (10), have been reported, although these cases differ from the PDL PNETs because the deposits were confined to the posterior fossa. In a report of 7 cases concerning a meningeal presentation of PNETs, but without detailed neuropathological data (7), no diagnostic uniformity was reached: three PNETs, three small-cells undifferentiated tumors, and one medulloblastoma being mentioned. Three other cases were described as PDL PNETs, one retrieved among a clinic pediatric series of 36 supratentorial PNETs (4), another diagnosed by CSF cytology only (1), and the last with post-mortem confirmation (6).
It is intriguing that in our patient the vermis presented, even if superficially, a high contrast enhancement on MRI. Given the small specimen received for neuropathological diagnosis, we wonder if it could represent a primary deposit of a diffuse leptomeningeal process, a very initial phase of the development and growth of a medulloblastoma which first presented as a leptomeningeal dissemination (even if the histopathology does not favor it), or even if a true parenchymatous mass could present later on. Indeed, cases of PDL PNETs which escaped early detection by MRI, including one with a voluminous, supratentorial, extra-axial placed, most probably a secondary deposit, were reported (5). However, our patient was alive eight months after the diagnosis, was submitted to radio and chemotherapy and no intraparenchymal mass has been detected so far.
Given the possible different radio and chemotherapy protocols to be used, an attempt should be done to precisely identify the nature of these so-called PDL PNETs, by having access to both extended neuropathological and cytogenetics studies.
Contributed by Henda Foreid, Cândida Barroso, Herculano Carvalho, Carlos Morgado, Lúcia Roque, José Pimentel