Search

The study of 18 oligodendrogliomas shows a number of characteristics which indicate a common hypothelial stem cell. The main cytologic characteristics are: a) high number of mitochondria, frequently atypical; b) polygonal crystalline structures in the cytoplasm in most of the tumors studied, and c) cytoplasmic microtubuli. None of these characteristics can be considered as pathognomic for the oligodendrogliomas. However, rows of folding cell processes represent a fundamental structural principle of all oligodendrogliomas and since they have not been found in any other brain tumor and may be regarded as an inherent and specific characteristic of the oligodendroglial group.

It is now clear that tissue culture has a role to play in the prediction of the degree of malignancy of human brain tumors. The fact that virtually all human brain tumors grow at least for some interval in culture allows application of tissue culture to the study of all tumors. This almost universal culturability of tumors is truly a singular virtue of intracranial neoplasms. No other human solid tumor group has been so amenable to in vitro growth or study. Up to the present time, despite our experience with over 1100 human brain tumor cultures, we have been extremely conservative in altering patient management on the basis of in vitro data. As is now apparent, the basis for direct input into the clinical milieu exists and it is necessary to work on a patient by patient basis to see how well the existing criteria can be applied to help guide management. Indeed the emphasis given to tissue culture by Rubinstein already confirms the current interest in the applicability of tissue culture data to neuropathological study of tumors (48). Clearly for certain tumors which tend to be benign, tissue culture can serve to alert the clinician to the perhaps unexpected malignant potential of the lesion. For the malignant tumors, the degree of malignancy, the probable biological behavior, the role of host defense factors and therapeutic agents can all be more quantitatively defined for the individual patient by the detailed study of the patient's cultured cells. With improvement in the surgical treatment of benign tumors and with better chemotherapeutic and radiotherapeutic measures for malignant neoplasms this more detailed and precise characterization of tumor behavior has become increasingly relevant to the optimization of the clinical management of the brain tumor patient.

A tumor with the clinical and light microscopic appearance of an oligodendroglioma that occurred in the lateral ventricles of a 25-year-old man is described. On further study this tumor proved to have the ultrastructural features typical of neuroendocrine tumors, and the presence of neuron-specific enolase was demonstrated by immunoperoxidase staining. This unusual presentation of a neuroendocrine tumor, which was entirely amitotic and free of atypia, raises important questions concerning both the true incidence of such cerebral differentiated neuroblastomas and their biologic behavior. The importance of electron microscopy and immunostaining techniques, which should be used more frequently to uncover additional cases of this tumor, is stressed.

A transmission electron microscope study was made of eight childhood brain tumors divided up into three zones, center, edge, infiltrating zone, and also of adjacent "normal-looking" brain. In seven of eight tumors, the numbers of Weibel-Palade bodies in endothelial cells were significantly increased in peripheral zones compared with central zones. A similar significant increase was observed after treatment of chick chorioallantoic membranes with tumor angiogenesis factor. It is suggested that large numbers of Weibel-Palade bodies may be a marker for proliferating endothelial cells in vivo.

Ultrastructural changes in the human epitumorous cerebral cortex were examined. A swelling of nerve cell perikarya, dendrites and axon terminals, and an occasional hyperplasia and disarray of microtubules and neurofilaments were observed. Accumulations of lysosomes, tubuloreticular structures and intracytoplasmic or intramitochondrial crystalloid inclusions were also found. Some myelinated axons were degenerated. Astrocytes and their processes were focally swollen. A mild swelling was found also in microgliocytes. Oligodendrocytes occasionally contained accumulations of dense bodies. Special attention was paid to dendritic spines. The spine surface morphology changed distinctly on swollen dendritic segments. The necks of most spines were short and wide, and numerous sessile forms were present in this location. The spine apparatus was often hypertrophied and disorganized. Smoothing out of spines on swollen dendrites is described and a possible functional significance of the observed changes in the symptomatology of brain tumors is hypothesized.

The fine structure of 26 isomorphic oligodendrogliomas excised from various areas of the Central Nervous System is described. These tumors are composed mainly of three types of cells: (a) those with large round or ovoid nuclei and scanty cytoplasm, (b) those with abundant cytoplasm, very rich in organelles and (c) cells very fibrillated with long cellular processes. The cytoplasm of the second cell type sometimes included crystalline bodies, numerous microtubules, myelin-like membranous structures and a large number of elongated mitochondria. The cells with the very fibrillated cytoplasm are eigher astrocytes intermixed with the neoplastic cells of the oligodendrocytic line or an astrocytic-like cellular evolution of the immature neoplastic cell of the oligodendrocytic line. Ultrastructural analysis provided evidence that isomorphic oligodendroglioma is not as homogeneous as was considered to be at the light microscopy level.

Oligodendrogliomas represent only 4% to 7% of intracranial gliomas, but accurate diagnosis of this neoplasm is important. They occur almost exclusively in the cerebral hemispheres and are most common in adults in middle life. Although classically associated with calcifications, these may be present in only 50% of cases. Special diagnostic techniques, including touch imprint preparations, electron microscopy, and immunohistochemistry can be useful in diagnosis. Grading systems correlating with prognosis have not been well established, but two recent careful clinical and pathologic studies have provided concrete and apparently valid criteria for grading. There is some evidence that radical surgical excision improves survival. Most patients receive postoperative radiation therapy. Prospective controlled therapeutic trials are lacking, however.

In 175 cases of oligodendroglioma (all grades, partly operated on several times), histological sections of the surgical material--in all, 226 preparations--were examined with regard to the presence and extent of lymphocytic infiltrates which could be assessed as manifestation of an immune reaction, and correlated with the clinical course. In case of infiltrates, the course was not more favourable than expected. Development of a tumour depends first of all on the inherent dynamics of the tumour itself. Prognosis can be favourably influenced by operation and possibly also by radiation.

Four mixed oligodendrogliomas and astrocytomas were investigated by electron microscopy and immunohistochemistry (GFAP, NSE and MBP). GFAP-positive oligodendroglioma cells and their transitional cells to GFAP-negative oligodendroglioma cells were present, suggesting successive morphological changes of astrocytic tumor cells. NSE-positive cells, suggestive of residual neurons, also exhibited round nuclei and perinuclear halos. On electron microscopy, oligodendroglioma cells that showed glial filaments, vascular end-feet and zonulae adherentes were occasionally present. The tumor cells with or without astrocytic characteristics showed common features of cytoplasmic organelles. These findings suggest that most oligodendroglioma cells in mixed gliomas are of an astrocytic nature and that characteristic microscopic features of oligodendroglioma are of a common cellular form that can be taken by various types of cells under certain circumstances.

Brain tumors were induced in 3-month-old rabbits of either sex by repeated intravenous injections of N-methyl-N-nitrosourea. Twelve brain tumors (6 pleomorphic gliomas, 5 grade 2--3 astrocytomas, 1 grade 2--3 oligodendroglioma) were established in culture and, with the exception of 2 neoplasms, were propagated in vitro as permanent cell lines. The glial nature of all cell lines was ascertained at several passage levels by testing the cells for the production of S-100 and GFA. It could be shown that most cells of all lines fluoresced positively for the S-100 protein, albeit differences in intensity of fluorescence were clearly noted between cells of the same culture and between different cultures. In general, astrocytoma cell lines had the strongest fluorescence. Pleomorphic glioma cells but especially astrocytoma cells reacted positively also for the GFA protein. Surprisingly enough, isolated cells of the oligodendroglioma line also showed evidence of GFA production. Exposure of cultures of rabbit glioma cells to db-cAMP for 8--10 hr resulted in inhibition of cell proliferation and stimulation of process formation. Furthermore, positive fluorescence for the S-100 and GFA proteins was more intense in cells treated with db-cAMP than in untreated cells. The latter observation may indicate that production and/or accumulation of glial proteins also was enhanced during the stationary phase of cell cultures.

The characteristics of brain tumors in vitro are analyzed in relation to their morphology, ultrastructure, immunocytochemistry and cell growth, showing that it is possible to identify the cell origin and to know several biological characteristics of the primitive tumor.

Tumors of the nervous system were induced in Sprague-Dawley and Long-Evans rats by weekly administrations of 6 mg/kg N-methyl-N-nitrosourea in the drinking water. Three of these tumors, a grade 2 mixed glioma, a grade 2 to 3 astrocytoma and a grade 1 to 2 oligodendroglioma, were established in culture and propagated in vitro. The mixed glioma strain (75SD-G-376) and the astrocytoma line (75SD-G-420) were repeatedly subcultured, cloned at passage 90 and 120 and designated as 75SD-G-376C and 75SD-G-420C clone, respectively. The growth rate of the oligodendroglioma cell strain (77LE-G-180) was very low and the cells died off after the 5th in vitro passage. The glial nature of all lines was ascertained by demonstrating the presence of the S-100 protein in the culture cells. 2 1/2 years after the establishment in vitro of the 75SD-G-376 and 75SD-G-420 primary cultures, mass cultures as well as clones derived from them are still producing S-100 and thus are clearly comparable to the primary cultures, at least in this respect. From a morphological standpoint based on light microscopy, cells of clonal lines with relatively few and short processes differ, however, from cells of primary cultures and their uncloned lines. Therefore, the cell morphology of these clones can be viewed upon as a form of adaptation to the in vitro conditions. It can be concluded that permanent cell lines with well-defined properties can be grown from experimental brain gliomas successfully established in culture and maintained in vitro.

The ultracytochemical acridine orange (AO) method has been employed to demonstrate DNA template activity within experimental brain tumors induced by transplacental administration ethylnitrosourea (ENU) in Sprague-Dawley rats. Ultrastructural examination revealed that acridine orange binds to DNA exclusively within the active extended euchromatin portion of the cell nucleus of brain tumor and of foci of atypical cell proliferation. No AO reaction products were visible in the non-neoplastic cell nuclei adjacent to the brain tumors. The percentages of AO positive cells in brain tumor cells in rats given an intraperitoneal injection of 3H-thymidine 1 hr before sacrifice were between 2.4% and 8.0%. The average number of AO chromatin reaction products per single section of a cell nucleus varied between 7 and 21. The ratios of areas of euchromatin to heterochromatin were found to be slightly larger in AO positive nuclei then in AO negative nuclei in each of the tumors. The present results suggested that rat brain tumors induced by transplacental administration ENU exhibit de-repression of DNA template normally repressed in the adult state. The usefulness of this method for study on the development of experimental brain tumors is briefly discussed.

Fifty-five cases of oligodendrogliomas and mixed oligoastrocytomas were evaluated using immunohistochemical (IH) study for glial fibrillary acidic protein (GFAP) and electron microscopic (EM) study. Most of the tumors in both of these groups showed many neoplastic oligodendroglial cells with GFAP-positive staining in their cytoplasm by IH study. By EM study too, many tumor cells showing features of oligodendroglial cells contained intermediate filaments. Our observations suggested the presence of a transitional form of cells in these tumors. The current study supports the contention that both oligodendrogliomas and oligoastrocytomas arise from a common progenitor cell capable of differentiation into both oligodendrocyte and astrocyte. The nature and degree of differentiation depends probably on gene expression and/or some microenvironmental factors.

Seven Oligodendrogliomas (2 with uniform cell type, 4 with cellular or tissue variability, and 1 with glioblastomatous changes) were examined ultrastructurally. The tumor cells were of two principal types with morphologic transitions between the two main types. The two principal cell types were identified as type 1 (undifferentiated) and type 2 (differentiated) on the basis of the number of anaplastic cells in an individual tumor and on the observations of Mori and Leblond (21) on non-neoplastic oligodendrocytes. Most of the tumor cells in all tumor exhibited similar histologic and ultrastructural characteristics including their arrangement and their tendency to form cytoplasmic processes which sometimes formed short stacks. These features were also recognizable in the glioblastomatous example and confirmed the presence of an oligodendroglial component. In addition to these characteristics, an increase in size and number of mitochondria, abundant intracytoplasmic structures, microtubules were regularly present in virtually all tumor cells. Cells rich in cytoplasmic filaments were present. These were identified as reactive astrocytes or as oligodendroglial tumor cells. Thus neither cytoplasmic filaments nor microtubules appear to be specific morphological markers for oligodendroglia or astrocytes; only the predominance of one of these structures permits cytogenetic identifications. The cytologic characteristics are not specific morphologic markers; however, recognition of their presence provides important diagnostic information.

Human fibroblasts and cells cultured from a glioblastoma multiforme and and oligodendroglioma were grown in standard tissue culture medium and in a medium in which D-valine was substituted for L-valine (DVM). Growth of all three types of cells was inhibited in DVM. Morphological changes of cells cultured in DVM also occurred: many fibroblasts became stellate-shaped and the oligodendroglioma cells had fewer processes. These findings indicate that DVM can not be used to decrease fibroblastic contamination in cultures of human gliomas without altering the morphological and growth characteristics of the neoplastic cells.

We have developed a model to study the neoplastic transformation of rat oligodendrocytes in culture. This procedure utilizes a technique previously developed by McCarthy and de Vellis which allows the preparation of 99% pure astrocyte and oligodendrocyte populations from 1- to 2-day-old rat cerebral cortices. Pregnant rats on the 19th day of gestation were given injections with either ethyl nitrosourea (10 micrograms/g body weight) in phosphate-buffered saline or phosphate-buffered saline, and oligodendrocyte cultures were prepared. Oligodendrocytes appear to be unstable in culture since transformation was observed with cells derived from either pups from pregnant rats either treated with nitrosourea or phosphate-buffered saline. Transformation required 78 to 108 days and 3 to 9 passages, at which time a marked increase in cellular proliferation was observed. The possibility that the transformed cells were derived from a nonoligodendroglial cell was excluded by the following evidence. Light and scanning electron micrographs of the transformed cells revealed cytological features essentially similar to those of primary oligodendroglial cultures. Furthermore, 2 biochemical oligodendroglial markers, the induction of lactate dehydrogenase by N6,O6-dibutyryl cyclic adenosine 3':5'-monophosphate and the presence of 2',3'-cyclic nucleotide 3' phosphohydrolase, were also retained. Conversely, another oligodendroglial marker, the hydrocortisone induction of glycerol phosphate dehydrogenase, was not found in any of the cell lines. These transformed cells grew as tumors when injected intracranially into 21-day-old rats. Histologically, these tumors did not appear as classical oligodendrogliomas, but their oligodendroglial origin was confirmed since the tumor tissue contained 2':3'-cyclic nucleotide 3'-phosphohydrolase activity, and the cells which grew from tumor explant cultures morphologically appeared similar to the parent cell line. The transformed cells were also characterized for in vitro properties which correlate with the expression of tumorigenicity. The transformed cells exhibited anchorage-independent growth and were agglutinated by concanavalin A treatment. Changes in fibrinolytic activity were not an exclusive property of transformed glial cells. This model should now allow us to study various mechanisms involved in the neoplastic transformation of oligodendrocytes.

Differentiated and anaplastic oligodendrogliomas revealed intracytoplasmic and nuclear presence of endogenous carbohydrate-binding proteins by application of labelled (neo)glycoproteins in histochemical analysis. The histochemical patterns showed differences between the differentiated and anaplastic forms of the same tumor type. Xylose-, lactose- and asialofetuin-specific carbohydrate-binding proteins could be detected in both types of tumors with the same staining intensity. However, maltose-specific carbohydrate-binding proteins were present only in the differentiated form. An inverse intensity of the histochemical reaction was observed with galactose-6-phosphate-, galactose-beta(1.3)-N-acetylglucosamine-N-acetylglucosamine- and mannose-(BSA-biotin) and fucose-(BSA-biotin) respectively, when the differentiated and anaplastic oligodendrogliomas were compared with each other. These differences document changes in the pattern of histochemically detectable carbohydrate-binding proteins, suggesting a role for endogenous carbohydrate-binding proteins in the tumor cell differentiation. These data indicate the potential usefulness of labelled (neo)glycoproteins as a new type of marker for histopathological diagnosis.

Ultrastructural concentric laminations have previously been thought to be specific to oligodendroglioma. However, these structures were also recognized in fibrillary astrocytomas, a mixed glioma and a glioblastoma. These laminations continued or closely related to attenuated processes or cytoplasm of astrocytic tumor cells. In addition, some lamellae contained glial filaments. It is considered that the concentric laminations are derived from attenuated astrocytic processes and have no relationship with myelin.