A case of metastatic brain tumor causing multifocal cerebral embolism.

The patient was a 72-year-old woman who had previously undergone treatment for femoral chondrosarcoma (histologically rated as myxofibrosarcoma). She suddenly developed left homonymous hemianopsia and was diagnosed with cerebral embolism. Because she had atrial fibrillation, we treated her for cardiogenic cerebral embolism. About 3 months later, however, she developed left hemiplegia, and head magnetic resonance imaging revealed multiple tumorous lesions affecting the previously detected infracted area and several new areas. We assumed that a tumor embolus had caused cerebral embolism, which resulted in growth of the tumor from the embolus and formation of a metastatic brain tumor. The metastatic foci formed from the tumor embolus were visualized by diagnostic imaging, and histological examination of the resected tumor confirmed that the brain tumor had occluded the brain vessel (tumorigenic cerebral embolism). No such case has been reported to date, and this case seems to be important.

Morphological and flow cytometric analysis of cell infiltration in glioblastoma: a comparison of autopsy brain and neuroimaging.

Even when we successfully perform a total extirpation of glioblastoma macroscopically, we often encounter tumor recurrence. We examined seven autopsy brains, focusing on tumor cell infiltration in the peripheral zone of a tumor, and compared our findings with the MR images. There has so far been no report regarding mapping of tumor cell infiltration and DNA histogram by flow cytometry, comparing the neuroimaging findings with the autopsy brain findings. The autopsy brain was cut in 10-mm-thick slices, in parallel with the OM line. Tissue samples were obtained from several parts in the peripheral zone (the outer area adjacent to the tumor edge as defined by postcontrast MRI) and then were examined by H&E, GFAP, and VEGF staining. We defined three infiltrating patterns based on number of infiltrated cells as follows: A zone, 100%-60% of the cells infiltrated tumor cells compared with tumor cell density of the tumor mass; B zone, 60%-20%; C zone, 20%-0%. In the autopsy brain, the tumor was easily identified macroscopically. We found that (1) the tumor cells infiltrated the peritumoral area; and (2) tumor cell infiltration was detected over an area measuring from 6 to 14 mm from the tumor border in the A zone. When performing surgery on glioblastoma, a macroscopic total extirpation of the tumor as defined by the contrast-enhanced area in MRI is therefore considered to be insufficient for successfully reducing tumor recurrence.

Interleukin-5 and interleukin-10 are produced in central nervous system tumor cysts.

Interleukin-5 and interleukin-10, as important mediators of vascular permeability, contribute to the development of various pathologic effusions. However, little is known regarding the involvement of these two cytokines in the formation of cysts associated with central nervous system (CNS) tumors. Twenty-eight patients with various cystic CNS tumors were investigated for expression of interleukin-5 and interleukin-10 in cyst fluid and their matched cytokine receptors in tumor tissue. Interleukin-5 and interleukin-10 were detected in cyst fluid, and interleukin-5 concentration was significantly correlated with interleukin-10 concentration (r=0.508, p=0.006). Moreover, both receptors were also detectable in the tumor tissue specimens and high levels of expression were also found in perivascular cells. Therefore, the local production of interleukin-5 and interleukin-10 might be implicated in some types of cyst formation.

An immunohistochemical and electron microscopic study of atypical teratoid/rhabdoid tumor.

We report two infant cases with atypical teratoid/rhabdoid tumor (AT/RT) located in the cerebellar vermis and spinal cord. MRI showed the tumors were isointense on T1-weighted images and mixed intensity of isointense and slight high intensity on T2-weighted images. Postcontrast MRI demonstrated clear margin of tumor and heterogeneous strong enhancement. It was difficult to differentiate the tumor from medulloblastoma by hematoxylin and eosin staining. However, immunohistochemical staining showed that these tumor cells react positively for cytokeratin, smooth muscle actin (SMA), and epithelial membrane antigen (EMA) and helped us with the differentiation. Electron microscopic study has confirmed the presence of mesenchymal components, such as filaments and desmosome junctions in the rhabdoid cells, but no neuronal components. The tumors rapidly increased in size, showing high MIB-1 index, and the prognosis was gave.