Hydra head activator-like immunoreactivity in human brain astrocytomas grade III-IV and the surrounding brain tissue.

We have developed a specific and sensitive radioimmunoassay for the hydra head activator neuropeptide (HHAP). The antibody recognizes the C-terminal portion of HHAP and shows no cross-reactivity with other neuropeptides. The assay allows measurement of HHAP-like material in tissue extracts with a minimum detectable concentration of 22 fmol/ml standard HHAP (2 fmol/tube). The concentration of immunoreactive (IR) HHAP in histopathologically characterized tissues was determined in 43 specimens from astrocytomas grade III-IV and surrounding brain tissue from 18 patients. Twenty-two control specimens of gray and white matter and five from hypothalamus were taken from 5 brains at autopsy. The concentration of IR-HHAP in the brain tumors, including the actively growing tumor front, is lower than in normal brain, and thus appears not to act as a growth factor. High performance liquid chromatography (HPLC) of extracts of hypothalamus and tumors revealed two major peaks of IR-HHAP; one eluted with the same elution volume as synthetic HHAP. HPLC of cerebral cortex extracts revealed only one major peak of IR-HHAP eluting close to the void volume, which may indicate a posttranslational processing variability of HHAP in different brain regions. By immunocytochemistry HHAP immunoreactivity was localized to the cytoplasma of neuroectodermal cells.

Tropomyosin isoform expression in normal and neoplastic astrocytes.

Changes of tropomyosin isoforms have previously been found accompanying morphologic alterations such as those associated with neoplastic transformations in mammalian cells. To determine whether an isoform change is associated with the malignancy of brain tumors, we employed both polyclonal antibodies specific to high and low molecular weight tropomyosin isoforms. We found, by using immunocytochemistry and immunoblotting, that an alteration of tropomyosin isoforms is associated with human astrocytomas. Differential staining patterns were seen for normal, reactive, and neoplastic astrocytes. Characterization of the antibodies revealed an increase in the higher molecular weight tropomyosin in more anaplastic astrocytomas than in those that were well differentiated. We also observed that the different isoforms have specific subcellular localizations. These findings suggest that neoplastic transformation is associated with alteration of tropomyosin isoforms in astrocytomas. We speculate that this change may be related to morphologic transformation in which microfilament functions, such as cell shape, interaction, and recognition are altered.

Sex steroids in human brain tumors and breast cancer.

The concentrations of three sex steroids, estradiol, progesterone and testosterone, were analyzed by radioimmunoassay after celite chromatography in brain tumor and breast cancer tissues. The concentrations in malignant gliomas and breast cancers showed interindividual variations, especially evident with regard to estradiol. High estradiol concentrations were recorded in two patients with malignant astrocytoma. The concentrations of 1.00 pg/mg and 3.32 pg/mg were 10 to 30 times as high as in normal female brain. In five of ten astrocytomas the estradiol concentration was higher than the lowest breast cancer value. The distribution of progesterone seemed more even, and the level was significantly lower in brain tumors and breast cancers as compared with female brain, perhaps indicating an increased metabolism. Testosterone levels were somewhat higher in brain tumors, as compared with breast cancers, but not different from values in brain tissue. There were no significant age or sex correlation or differences in the concentrations of steroids in the brain tumors. The results suggest that manipulation of sex steroid metabolism in malignant brain tumors can be of beneficial therapeutic value as has been shown for breast cancer and prostatic carcinoma.

The distribution of bromine content of dibromodulcitol in the central nervous system of patients with malignant gliomas.

The bromine content of human gliomas and white matter was determined by neutron activation analysis (NAA) following p.o. administration of a single dose of 400-500 mg/m2 dibromodulcitol (DBD). In another group of patients with brain gliomas, the bromine content was measured subsequent to application of a single dose of 334 mg/m2 of sodium bromide (equivalent dose regarding the bromine content of DBD). The bromine content of these two groups was compared to the values found in a third control group of untreated patients. The amount of bromine after DBD application was three to four times higher than in the untreated samples and the average accumulation ratio of 1.8 +/- 0.4 proved to be nearly identical both in tumour and white matter. The bromine values after NaBr treatment showed a different pattern of distribution. The accumulation was higher in the tumour tissue than in the normal white matter. These findings demonstrate that the pharmacokinetic properties of DBD- and NaBr-derived bromine are different, suggesting that the increase of bromine after DBD administration could be due to covalently bound bromine in DBD.

Sigma and opioid receptors in human brain tumors.

Human brain tumors (obtained as surgical specimens) and nude mouse-borne human neuroblastomas and gliomas were analyzed for sigma and opioid receptor content. Sigma binding was assessed using [3H]1,3-di-o-tolylguanidine (DTG), whereas opoid receptor subtypes were measured with tritiated forms of the following: mu, [D-ala2,mePhe4,gly-ol5]enkephalin (DAMGE); kappa, ethylketocyclazocine (EKC) or U69,593; delta, [D-pen2,D-pen5]enkephalin (DPDPE) or [D-ala2,D-leu5]enkephalin (DADLE) with mu suppressor present. Binding parameters were estimated by homologous displacement assays followed by analysis using the LIGAND program. Sigma binding was detected in 15 of 16 tumors examined with very high levels (pmol/mg protein) found in a brain metastasis from an adenocarcinoma of lung and a human neuroblastoma (SK-N-MC) passaged in nude mice. kappa opioid receptor binding was detected in 4 of 4 glioblastoma multiforme specimens and 2 of 2 human astrocytoma cell lines tested but not in the other brain tumors analyzed.

Multifocal glioblastoma with liver metastases in the absence of surgery. Case report.

A patient with two intracerebral glioblastomas of differing histology with metastases to the liver in the absence of surgery is reported. The gliomatous nature of the lesions was confirmed by staining for glial fibrillary acidic protein. Histological and immunohistochemical evidence suggests that the metastases arose from the more poorly differentiated of the intracerebral tumors. One of the intracerebral tumors had enhanced expression of the ras p21 oncogene as compared to the other tumors and as compared to nonmalignant brain tissue from this patient.

Dual lineage of astrocytomas.

Experimental observations suggest that type 1 and type 2 astrocytes have distinct lineages distinguished, respectively, by the absence or presence of the A2B5 antigen. To investigate the hypothesis that the strikingly different biological behaviors of low-grade and high-grade astrocytomas might be related to cell lineage, 38 astrocytomas (grades I through IV), were examined for the presence of the A2B5 antigen, glial fibrillary acidic protein (GFAP), and galactocerebroside (GC), a marker for oligodendrocytes. Twenty-nine of the tumors (76%) were composed of moderate or high densities of GFAP+ and unlabeled cells, and were devoid of A2B5+ and GC+ neoplastic cells. Only four tumors had high densities (75% to 100%) of A2B5+ cells, three of which were grade I tumors; two tumors, also Grade I, contained moderate densities (20% to 25%) of GC+ cells that had cytologic features of astrocytoma or oligodendroglioma. The findings suggest that most cerebral astrocytomas lack the A2B5 antigen and thus may represent neoplasia along the type 1 astrocyte lineage. In contrast, A2B5+ lineage among astrocytomas, ie, neoplasia with differentiation toward type 2 astrocytes and GC+ oligodendrocytes, is less common and may be correlated with prognostically favorable histopathologic features.

Quantification of the c-myc oncoprotein in human glioblastoma cells and tumor tissue.

The identification of a quantifiable oncoprotein marker in glial cells could lead to its use as an aid in the diagnosis, grading, and treatment of tumours of glial origin. In this study, monoclonal antibodies to the c-myc oncoprotein were used in conjunction with immunofluorescence microscopy, flow cytometry, and immunoblot analysis to quantitate and characterize the expression of this oncoprotein in neoplastic and benign cultured glial cells and brain-tumor tissue. Flow cytometric analysis revealed that the c-myc oncoprotein was highly expressed in neoplastic cell lines and in glioblastoma tumor specimens. In contrast, anti-c-myc oncoprotein staining was not present in a non-neoplastic glial cell line or in a benign brain tissue specimen. Immunoblot analysis revealed two distinct c-myc oncoprotein bands, having molecular weights of 64 and 75 kD. Densitometric determinations of the relative levels of the 64-kD protein were in good agreement with the determinations made by flow cytometry. Flow cytometry was also used to relate the quantity of the c-myc oncoprotein present in the cells to their cell cycle phase. In the malignant cultured cells, the protein underwent an approximate twofold increase as the cells progressed from G1/G0 to G2/M in the cell cycle. The present results suggest that the c-myc oncoprotein may prove to be a useful marker for the proliferation status and/or malignancy of glial cells.

Cytologic composition of the untreated glioblastoma with implications for evaluation of needle biopsies.

To study the progression of astrocytic neoplasms and to provide practical information about the topography of the glioblastoma multiforme, the distribution of eight defined cell types was mapped from whole brain sections of 18 glioblastomas studied postmortem. Based on the densities and topographic distributions of well-differentiated and anaplastic cells, three principal categories of neoplasms were defined. In one group of three cases, multifocal glioblastomas appeared to be emerging in the background of a better differentiated, and presumably precursory, astrocytic neoplasm. In another group of nine cases, the neoplasms were more intimate mixtures of well and poorly differentiated cells. The third group of five cases was composed of neoplasms that were largely undifferentiated without a component of better differentiated cells. The study suggests that progression from a better differentiated neoplasm to one composed largely of undifferentiated cells is common in the fibrillary astrocytic neoplasms. Although some glioblastomas appear largely undifferentiated and consistent with the de novo appearance of overt malignancy, the size of these neoplasms and the patterns of necrosis leave open the possibility that a preexisting better differentiated neoplasm had been obliterated by necrosis and the overgrowth of the anaplastic component. The potential topographic variation of cellular constituency in a glioblastoma underscores the care that must be exercised in utilization of the needle biopsy technique in the diagnosis and grading of astrocytic neoplasms.

Flow-cytometric DNA analysis and immunohistochemical measurement of Ki-67 and BUdR labeling indices in human brain tumors.

Cell proliferation potential was assessed by measuring the labeling indices of the monoclonal antibody Ki-67 and of 5-bromodeoxyuridine (BUdR), and the cellular deoxyribonucleic acid (DNA) content in 48 human brain tumors. The diagnostic and prognostic value of flow-cytometric DNA analysis was also evaluated using ethanol-fixed paraffin-embedded BUdR-labeled specimens; these were the same specimens as were used for measuring the BUdR and Ki-67 labeling indices. Both the Ki-67 and the BUdR labeling indices correlated with the degree of malignancy estimated from conventional histological preparations. The Ki-67 labeling index was 1.7 times greater than the BUdR labeling index. The relationship of DNA aneuploidy to the labeling indices or to morphology in cases of glioma was examined. All of the tumors with an aneuploid line corresponded to malignant glioma classified by histological criteria, although malignant glioma did not always show DNA aneuploidy. In addition, the cases with aneuploid lines showed high BUdR and Ki-67 labeling indices. The cell kinetic data, which indicate the biological character of tumors, allowed prediction of the prognosis of the patients with gliomas. In contrast, despite the presence of an aneuploid line, three of 13 meningiomas showed a benign histological pattern without an aggressive clinical course, and neither the Ki-67 nor the BUdR labeling index was high. These results indicate an unequivocal relationship between DNA aneuploidy and clinical behavior; in general, both labeling indices may prove to be objective indicators of the outcome of patients with brain tumors.

Immunohistochemical demonstration of immunoglobulins and albumin in human brain tumors.

Routinely processed biopsy material, including 56 gliomas of varying malignancy, 10 meningiomas, 10 brain metastases and 12 brain abscesses, was examined for the presence and distribution of IgG, IgA, IgM, IgD and albumin using the unlabeled antibody peroxidase-antiperoxidase technique. In all specimens the deposition of stained immunoglobulins (Ig) was strictly associated with that of albumin even on cell surfaces. Thus there was no evidence for specific membrane binding or cytotoxicity. The interstitial proteins demonstrated are most likely derived from the plasma by blood-brain barrier breakdown which occurs in nearly all tumors and abscesses. Obvious intracellular staining for Ig and albumin was seen in glioma cells and astrocytes only. This is suggested to be due to active protein uptake as a specific feature of astrocyte differentiation which decreases with malignancy and is lost in glioblastomas. Evidence for local Ig production was found in 8 out of 10 metastases with striking IgG- and IgA-positive plasma cells within lymphocytic infiltrations and in one meningioma showing conspicuous plasma cells components. No glioma contained Ig-bearing plasma cells, though round cell infiltrations were present in 64% of the unselected cases. The significance of these findings regarding the immunological situation in brain tumors is briefly discussed.

Intermediate filament expression in astrocytic neoplasms.

Immunohistochemical analysis of 30 paraffin-embedded astrocytic neoplasms was performed to correlate the expression of intermediate filament proteins with histologic subtype. Each tumor was studied with monoclonal antibodies to keratin, vimentin, desmin, 200-kd neurofilament protein, and glial fibrillary acidic protein (GFAP). Immunoreactivity with the anti-keratin monoclonal antibodies AE1 and AE3 was demonstrated in 24 cases (80%) including 4 of 6 (66%) well-differentiated astrocytomas (WDAs), 10 of 12 (83%) anaplastic astrocytomas (ANAs), and 10 of 12 (83%) glioblastomas multiforme (GBMs). These cases were further studied with the monoclonal antikeratin antibodies 34 beta E12 and 34 beta H11. Of the 24 AE1/AE3-positive cases, 14 (58%) reacted with 34 beta E12. None of the cases was reactive with 34 beta H11. Vimentin expression was demonstrated in 24 cases (80%), including 2 of 6 (33%) WDAs, 11 of 12 (92%) ANAs, and 11 of 12 (92%) GBMs. Coexpression of keratin and vimentin was observed in 20 cases (67%), including 2 of 6 WDAs, 9 of 12 (75%) ANAs, and 9 of 12 (75%) GBMs. Immunoreactivity with GFAP antibody was present in all 30 (100%) cases, but none of the tumors was reactive with antibodies to desmin or 200-kd neurofilament protein. These findings demonstrate that expression of both keratin and vimentin intermediate filaments is common in astrocytic neoplasms regardless of histologic grade.

Endogenous sugar receptor pattern in human glioblastomas and gangliocytomas studied by histochemical application of biotinylated (neo)glycoproteins and affinity chromatography.

Biotinylation of chemically glycosylated bovine serum albumin, yielding a panel of neoglycoproteins, and of desialylated, naturally occurring glycoproteins allowed to systematically evaluate presence and distribution of various types of endogenous sugar receptors in the sections of human glioblastomas and gangliocytomas by a routine histochemical procedure. Pronounced cytoplasmic staining with markers, carrying constituents of natural glycoconjugates, e.g. for beta-galactoside-specific receptors, contrasted with the different intensities, noticed for alpha- and beta-glucoside-specific receptors. Significant qualitative differences between the two tumor types were detected with N-acetyl-D-galactosamine- and sialic acid-carrying probes. Nuclear staining with only a part of the applied panel underscored the specificity of the protein-carbohydrate interaction. Fine structural features of the synthetic neoglycoproteins, e.g. the mode of coupling of the carbohydrate moiety to the protein, were found to exert a significant influence on their suitability as histochemical markers. On the basis of the histochemical results, exemplary biochemical analysis of certain classes of endogenous sugar receptors by affinity chromatography and subsequent gel electrophoresis, namely of beta-galactoside-, alpha-fucoside-, alpha-mannoside- and alpha-glucoside-specific proteins, revealed presence and characteristics of respective sugar receptors that can contribute to the histochemical staining. Similar extent of histochemical staining with the respective probes notwithstanding, the different tumor types exhibited qualitative differences in the expression of individual endogenous sugar receptors. The combined histochemical and biochemical analysis is supposed to be of conspicuous value for biological and clinical investigations on endogenous sugar receptors.

Renin in glioblastoma multiforme and its role in neovascularization.

Significant proliferation of capillaries with hyperplastic vascular endothelium is one of the characteristic histologic features of glioblastoma multiforme (GBM). It has been shown that the renin-angiotensin II cascade stimulates new vessel formation. The presence of renin in several types of highly vascularized neoplasm suggests that it may also be implicated in the mechanism of tumor angiogenesis. In order to study the possible relationship of renin to GBM, immunohistochemical search for human renin was carried out in ten instances of such a tumor. Eight of these cases demonstrated renin-containing neoplastic astrocytes, whereas seven cases of reactive gliosis and six cases of low-grade astrocytoma revealed no renin-containing cells. The immunostaining was not present after preabsorption of the renin antiserum with pure human renin or substitution of preimmune serum for the specific renin antiserum. Because it has also been demonstrated that a product of renin, angiotensin II, has angiogenic properties, it seems reasonable to postulate that renin, through angiotensin II, may play a role in the mechanism of GBM-associated neovascularization.

Correlation of DNA content and histology in prognosis of astrocytomas.

Thirty-two cases of astrocytoma were analyzed for DNA content and cell-cycle proliferation features by flow cytometry using paraffin-embedded tissue. The findings were correlated with histologic grading and survival. Abnormal DNA (aneuploidy or elevated G2-M fraction greater than or equal to 7%) was present in 18 cases (56%). Glioblastoma multiforme (GBM) had 11 of 16 (69%), anaplastic astrocytomas (ANA) 7 of 11 (64%), and low-grade (LG) neoplasms 0 of 5 cases with abnormal DNA content. Short-term survival (less than or equal to 26 months) occurred in all 16 patients with GBM (100%), 7 of 11 patients with ANA (64%), and 1 of 5 patients with LG neoplasms (20%). Seventeen of 18 patients (94%) with abnormal DNA content were short-term survivors (P less than 0.0002). Abnormal DNA content was found in 17 of 24 short-term survivors (71%), whereas histologic grading identified 16 of 24 such cases (67%). A combination of grading and abnormal DNA content identified 22 of 24 (92%) of the poor survival cases. DNA content was most useful in the anaplastic group. Six of seven cases (86%) with abnormal DNA content had short survival (P less than 0.055), and three of four (75%) with normal DNA content had long survival. DNA analysis combined with histologic grading improves prognosis designation.

[Immunohistochemical study of S-phase cells in human gliomas].

Intermediate filament proteins are cytoskeletal components in most vertebrate eukaryotic cells and some of these proteins are recognized markers of cell differentiation. To investigate the expression of intermediate filament proteins of the S-phase cells in human glial tumors, we have examined fourteen patients with benign and malignant gliomas by immunohistochemical study using in vivo labeling with bromodeoxyuridine (BrdU). Five glioblastoma multiforme, five anaplastic astrocytoma, three fibrillary astrocytoma and one gemistocytic astrocytoma were studied. All patients were given intravenous infusion of BrdU (10 mg/kg) one hour before craniotomy for labeling the S-phase cells of the tumors. Surgical specimens were immersed in 70% ethanol, and embedded in paraffin. Four micron sections were immunostained with anti-BrdU monoclonal antibody (Mab) and anti-vimentin Mab by avidin-biotin complex (ABC) method, and anti-glial fibrillary acidic protein (GFAP) serum by peroxidase-antiperoxidase (PAP) method. All sections (except for case 4) were double-labeled with anti-BrdU Mab and anti-GFAP serum, or with anti-BrdU Mab and anti-vimentin Mab. The population of BrdU-labeled cells (i.e. S-phase cells), and double-labeled cells were analyzed. The proportions of BrdU-labeled cells ranged from 6.1% to 17.0% (average 11.1%) in glioblastoma multiforme, from 3.5% to 15.6% (average 8.8%) in anaplastic astrocytoma, and from 2.0% to 2.8% (average 2.5%) in fibrillary astrocytomas. One gemistocytic astrocytoma showed S-phase fraction of 1.7%. Two recurrent cases of anaplastic astrocytoma showed higher S-phase fractions than other non-recurrent cases of anaplastic astrocytoma.(ABSTRACT TRUNCATED AT 250 WORDS)

Histological grading, DNA content, cell proliferation and survival of patients with astroglial tumors.

Light microscopy, image cytometry (ICM), and flow cytometry (FCM) were used to study the degree of differentiation, DNA content, and S-phase of astrocytomas and glioblastoma multiforme in 102 patients. The postoperative real survival time (RST) was also studied. Using ICM, 62 astrocytomas were investigated. Grade I astrocytomas were composed of DNA-diploid cell lines, while grade III and glioblastoma multiforme consisted predominantly of DNA-aneuploid lines. Moderately differentiated astrocytomas were divided as follows: 14 DNA-diploid and 18 DNA-aneuploid. Forty astrocytomas were studied by FCM. Using the DNA index (DI) value, cases with abnormal DNA cell lines were established in all astrocytomas, with their number increasing in grades II and III astrocytomas. FCM indicated the same subdivision of moderately differentiated astrocytomas: 12 with DNA-diploid and 12 with DNA-aneuploid stem lines. Patients with DNA-diploid cell lines in the astrocytomas and low S-fraction survived longer than patients with abnormal DNA cell populations and higher S-fraction. The results from this study indicate that, together with the degree of differentiation of astroglial tumors, the appearance of cell lines with abnormal DNA value and higher S-fractions also have prognostic value.

[Factors affecting the prognosis of brain tumors].

A variety of tumors with different histologic types are included in a group of brain tumors. Although each histologic type of tumor has its own range of malignancy, the prognosis seems to be affected by several clinical, histologic and cell-biological factors. For example, relative survival rate of patients with glioblastoma is lower if the patient is older than 50 or 60 years. The leptomeningeal dissemination of glioma cells is a sign of poor prognosis. The presence of necrotic foci in the astrocytic tumors suggests shorter astrocytic tumors suggests shorter survival. Using a monoclonal antibody to bromodeoxyuridine (BrdU), the growth activity of the tumor can be estimated by BrdU labeling index (BrdU-LI, %). Higher BrdU-LI is correlated with more malignant histologic features in astrocytic tumors. In meningiomas, higher BrdU-LI is correlated with a more frequent or rapid recurrence of the tumor. The significance of growth factor receptors and oncogene of growth factor receptors and oncogene products as a cell-biologic marker of malignancy was investigated with an immunohistochemical method. Transferrin receptor was demonstrated in all tumors, and epidermal growth factor in about 40% of astrocytic tumors. The immunoreaction to c-myc oncogene product was detected in most astrocytic tumors; with higher intensity in anaplastic astrocytomas and glioblastomas than in low-grade astrocytomas. The role of these markers in the prognosis of brain tumors is, however, still unclear. Total or subtotal resection of glioblastoma results in longer resection of glioblastoma results in longer survival. Both postoperative radiotherapy and chemotherapy are effective. However, maintenance of chemotherapy longer than longer than 2 years does not significantly improve the prognosis.

[Interleukin 1 and gliomas: immunohistochemical and immunocytochemical examinations].

Interleukin 1 is a monokine produced by macrophage and has an ability to activate thymocytes. In addition to the immunological regulatory effect, interleukin 1 has attracted a great deal of investigators as a new peptide hormone that was secreted by many cells and has a various physiological activities. In central nervous systems, interleukin 1 promotes the glial cells proliferations on the injured brain and the fetal brain. The cell sources of interleukin 1 in central nervous systems are considered to the microglial cells. On gliomas, Lachmann and Dinarello reported growth promoting effect of IL-1 on U-373 MG human glioblastoma cell. The authors investigated the roles and effects of IL-1 on the growth of gliomas using recombinant human IL-1 beta and anti-HuIL-1 beta monoclonal antibody. On Immunohistochemistry, paraffin sections of 10 cases of gliomas were stained with immunoperoxidase method using anti-human IL-1 beta and anti-GFAP mouse monoclonal antibody. All astrocytomas examined and 2 of 4 glioblastomas were stained by anti-IL-1 beta. The origin of IL-1 that was stained by immunoperoxidase staining is unknown. The authors think it that IL-1 existed in glioma cells were secreted by microglial cells or that the glioma cells themselves secreted IL-1. In either case, IL-1 must be related to the growth of glioma in situ. On immunocytochemistry, U-373 MG human glioblastoma cells purchased from ATCC were incubated on cover-slip with 0 and 10 U/ml of rHuIL-1 beta for 3 or 7 days. The cells were stained with immunoperoxidase method using anti-GFAP monoclonal antibody.(ABSTRACT TRUNCATED AT 250 WORDS)