Subependymal giant cell astrocytoma (SEGA): Is it an astrocytoma? Morphological, immunohistochemical and ultrastructural study.

Subependymal giant-cell astrocytoma (SEGA) is a rare intra-ventricular low-grade tumor which frequently occurs as a manifestation of tuberous sclerosis complex. The histogenesis of SEGA is controversial and its astrocytic nature has been doubted. First studies suggested the astrocytic nature of SEGA while several recent reports demonstrate its glio-neuronal nature. In spite of this, in the recently revised WHO classification of the CNS tumors, SEGA has been still included in the group of astrocytomas. We studied nine tuberous sclerosis complex-associated SEGAs. Patients were 1-18 years old. Eight patients (89%) had a solitary lesion located in the lateral ventricle close to of the head of the caudate nucleus, the remaining patient (11%) had two tumors, one located close to the head of the left caudate nucleus and the other in the central part of the right lateral ventricle. Histologically, tumors were composed of three types of cells: spindle, gemistocytic and ganglion-like. Four tumors (44%) had a prominent vascularization and three (33%) showed an angiocentric pattern. Calcifications were observed in six cases (66%). By immunohistochemistry, the majority of the tumors were GFAP- (9; 100%), neurofilament- (8, 89%), neuron-specific enolase- (9, 100%), and synaptophysin- (8; 89%) positive. Ultrastructural studies were performed on four cases. In all four there were glial cell processes filled with intermediate filaments. In one case dense core putative neurosecretory granules were appreciable. Our results emphasize the glio-neuronal nature of SEGA. We suggest moving it into the group of mixed glio-neuronal tumors under the denomination of subependymal giant cell tumor.

Liquid-based endometrial cytology in the management of sonographically thickened endometrium.

Liquid-based cytology represents an opportunity to re-evaluate endometrial cytology. We evaluated the accuracy of liquid-based endometrial cytology as compared to biopsy in 670 women scheduled for histeroscopy because of thickened endometrium (>4 mm), as evaluated by transvaginal sonography. Endometrial biopsy detected pathology in 41 (6%) of cases (21 of which were adenocarcinomas). Cytologic study found pathology in 62 (9%) cases (19 of which were adenocarcinomas). Two hundred ninety-one biopsies (43%) and 28 (4%) cytologies were inadequate. The sensitivity and the specificity were estimated, respectively, at 95% and 98%; the positive and negative predictive values were estimated, respectively, at 83% and 99%. Cytology provided sufficient material more often than biopsy (P < 0.01). We consider endometrial cytology an efficacious diagnostic opportunity. It could be usefully applied in association with transvaginal sonography. The combination of these procedures might reduce more invasive and expensive diagnostic procedures.

NF2 gene expression in sporadic meningiomas: relation to grades or histotypes real time-pCR study.

One of the most common regions involved in the meningiomas tumorigenesis is chromosome 22q where the NF2 gene resides. The deficiency or loss of the NF2 gene product, merlin/schwannomin, plays a role in tumor development and metastatization. Conflicting results have been reported on the prognostic value of merlin in meningiomas. Several studies have indicated NF2 gene inactivation as an early tumorigenic event unrelated to the histological grade or clinical behavior. On the contrary, the NF2 gene alteration rate differs between the different histotypes. A pathogenesis independent from the NF2 gene has been suggested in meningothelial meningiomas. In the present work, we studied the NF2 gene expression through real time-PCR (RT-PCR) in 30 meningiomas. The average of the NF2 gene expression of all meningiomas was considered as reference value. The average of expression of WHO grade I and II meningiomas was higher than the average of all meningiomas, whereas that of WHO grade III meningiomas was lower. When we compared the NF2 gene expression in the different meningioma grades we did not note a significant difference (P = 0.698) despite the tendency to decrease from grade I to grade III. The average expression of meningothelial meningiomas was higher than the reference value, and that of non-meningothelial meningiomas was lower. The difference in NF2 gene expression between meningothelial and non-meningothelial meningiomas was statistically significant (P = 0.013). Our data supports the finding that alterations in NF2 gene alteration are histotype related but not grade related.

Genetic and expression profiles of cerebellar liponeurocytomas.

Cerebellar liponeurocytoma, a rare, newly identified CNS neoplasm of adults, is characterized by advanced neuronal/neurocytic and focal lipomatous differentiation, low proliferative potential and a favorable clinical prognosis. Despite the different age distribution and benign biological behavior, the cerebellar liponeurocytoma shares several features with the cerebellar medulloblastoma, which may include an origin from the periventricular matrix of the fourth ventricle or the external granular layer of the cerebellum. To establish the genetic profile of cerebellar liponeurocytomas, we have formed an international consortium and collected tumor samples from 20 patients. DNA sequencing revealed TP53 missense mutations in 4 (20%) of 20 cerebellar liponeurocytomas, a frequency higher than in medulloblastomas. There was no case with PTCH, APC, or beta-catenin mutations, each of which may be present in subsets of medulloblastomas. Isochromosome 17q, a genetic hallmark of classic medulloblastomas, was not observed in any of the cases investigated by FISH analysis. cDNA array analyses were carried out on 4 cerebellar liponeurocytomas, 4 central neurocytomas, and 4 classic medulloblastomas. Cluster analysis of the cDNA expression data of 1176 genes grouped cerebellar liponeurocytomas close to central neurocytomas, but distinct from medulloblastomas. These results suggest cerebellar liponeurocytoma as a distinct tumor entity that is genetically different from medulloblastoma. Furthermore, the cDNA expression array data suggest a relationship to central neurocytomas, but the presence of TP53 mutations, which are absent in central neurocytomas, suggests that their genetic pathways are different.