Thank you all for your interest in this case.
First of all, I realized I mispelled mesencephalon as "mesocephalon" in the clinical history, I apologize.
As your can see in the images provided, this tumor was composed of variably sized cells with scant cytoplasm and pleomorphic round to oval nuclei, some with prominent nucleoli. We saw many mitotic figures, microvascular proliferation and focal necrosis. No perivascular rosettes, ependymal canals or rhabdoid differentiation was identified.
Given the location (thalamus), sex, age and MRI findings of a contrast enhancing, solid and partially cystic lesion we initially favored a germ cell neoplasm. Our differential diagnosis included all of the diagnoses kindly provided by you: Obviously a germ cell neoplasm, a supratentorial primitive neuroectodermal tumor (PNET), atypical teratoid/rhabdoid tumor, and high grade glial neoplasm (anaplastic astrocytoma, glioblastoma, anaplastic ependymoma, ependymoblastoma).
-GFAP was strongly positive in a large proportion of neoplastic cells (see Figure_9
-Scattered cells showed positivity with Synaptophysin (see Figure_10
) and Neurofilament protein.
-p53 showed strong nuclear immunoreactivity in about 50% of tumor cells (see Figure_11
-Ki-67 proliferation index was also in the 40-50% range.
-The following immunohistochemical markers were all negative in tumor cells: PLAP, beta-HCG, CD30 (Ki-1), alpha-fetoprotein, low-molecular-weight cytokeratin (CAM 5.2), and EMA.
The wording of the final diagnosis is mostly a matter of style (my personal opinion), and depending on your preference: Supratentorial PNET with advanced glial/astrocytic differentiation or a more straightforward WHO Grade IV astrocytoma/GBM are acceptable. Our preference was to call it a GBM because that diagnosis made it easy for her clinician to include her in an ongoing trial of experimental chemotherapy. p53 immunopositivity has been described in supratentorial PNETs (see references), and in my opinion cannot be used to distinguish between a GBM or an SPNET, but some neuropathologists may find it difficult to call this tumor a PNET with positive p53 staining. Recently a group at our institution has found important prognostic implications of p53 expression in malignant gliomas in children (5
Thanks again to Clóvis Klock and Guilherme Lopes, Romualdo Correia Lins Filho, Teresa Tuñón, Celso Rubens Vieira e Silva, Vidal Diez Sanchez, Jose Mari Arrinda and Javier Ortiz for their comments.