Astrocytes derived from p53-deficient mice provide a multistep in vitro model for development of malignant gliomas.

Loss or mutation of p53 is thought to be an early event in the malignant transformation of many human astrocytic tumors. To better understand the role of p53 in their growth and transformation, we developed a model employing cultured neonatal astrocytes derived from mice deficient in one (p53 +/-) or both (p53 -/-) p53 alleles, comparing them with wild-type (p53 +/+) cells. Studies of in vitro and in vivo growth and transformation were performed, and flow cytometry and karyotyping were used to correlate changes in growth with genomic instability. Early-passage (EP) p53 -/- astrocytes achieved higher saturation densities and had more rapid growth than EP p53 +/- and +/+ cells. The EP p53 -/- cells were not transformed, as they were unable to grow in serum-free medium or in nude mice. With continued passaging, p53 -/- cells exhibited a multistep progression to a transformed phenotype. Late-passage p53 -/- cells achieved saturation densities 50 times higher than those of p53 +/+ cells and formed large, well-vascularized tumors in nude mice. p53 +/- astrocytes exhibited early loss of the remaining wild-type p53 allele and then evolved in a manner phenotypically similar to p53 -/- astrocytes. In marked contrast, astrocytes retaining both wild-type p53 alleles never exhibited a transformed phenotype and usually senesced after 7 to 10 passages. Dramatic alterations in ploidy and karyotype occurred and were restricted to cells deficient in wild-type p53 following repeated passaging. The results of these studies suggest that loss of wild-type p53 function promotes genomic instability, accelerated growth, and malignant transformation in astrocytes.

Gamma-radiation sensitivity and risk of glioma.

BACKGROUND: About 9% of human cancers are brain tumors, of which 90% are gliomas. gamma-Radiation has been identified as a risk factor for brain tumors. In a previous pilot study, we found that lymphocytes from patients with glioma were more sensitive to gamma-radiation than were lymphocytes from matched control subjects. In this larger case-control study, we compared the gamma-radiation sensitivity of lymphocytes from glioma patients with those from control subjects and investigated the association between mutagen sensitivity and the risk for developing glioma. METHODS: We used a mutagen sensitivity assay (an indirect measure of DNA repair activity) to assess chromosomal damage. We gamma-irradiated (1.5 Gy) short-term lymphocyte cultures from 219 case patients with glioma and from 238 healthy control subjects frequency matched by age and sex. After irradiation, cells were cultured for 4 hours, and then Colcemid was added for 1 hour to arrest cells in mitosis. Fifty metaphases were randomly selected for each sample and scored for chromatid breaks. All statistical tests were two-sided. RESULTS: We observed a statistically significantly higher frequency of chromatid breaks per cell from case patients with glioma (mean = 0.55; 95% confidence interval [CI] = 0.50 to 0.59) than from control subjects (mean = 0.44; 95% CI = 0.41 to 0.48) (P<.001). Using 0.40 (the median number of chromatid breaks per cell in control subjects) as the cut point for defining mutagen sensitivity and adjusting for age, sex, and smoking status, we found that mutagen sensitivity was statistically significantly associated with an increased risk for glioma (odds ratio = 2.09; 95% CI = 1.43 to 3.06). When the data were divided into tertiles, the relative risk for glioma increased from the lowest tertile to the highest tertile (trend test, P<.001). CONCLUSION: gamma-Radiation-induced mutagen sensitivity of lymphocytes may be associated with an increased risk for glioma, a result that supports our earlier preliminary findings.

Cyclooxygenase-2 expression in human gliomas: prognostic significance and molecular correlations.

Cyclooxygenase (COX)-2, the inducible isoform of prostaglandin H synthase, has been implicated in the growth and progression of a variety of human cancers. Although COX-2 overexpression has been observed in human gliomas, the prognostic or clinical relevance of this overexpression has not been investigated to date. In addition, no study has analyzed the relationship between COX-2 expression and other molecular alterations in gliomas. Consequently, we examined COX-2 expression by immunohistochemistry in tumor specimens from 66 patients with low- and high-grade astrocytomas and correlated the percentage of COX-2 expression with patient survival. We also analyzed the relative importance of COX-2 expression in comparison with other clinicopathological features (age and tumor grade) and other molecular alterations commonly found in gliomas (high MIB-1 level, p53 alteration, loss of retinoblastoma (Rb) protein or p16, and high bcl-2 level). Kaplan-Meier analyses demonstrated that high COX-2 expression (>50% of cells stained positive) correlated with poor survival for the study group as a whole (P < 0.0001) and for those with glioblastoma multiforme in particular (P < 0.03). Cox regression analyses demonstrated that COX-2 expression was the strongest predictor of outcome, independent of all other variables. In addition, high COX-2 expression correlated with increasing histological grade but did not correlate with positive p53 immunostaining, bcl-2 expression, loss of p16 or retinoblastoma protein expression, or high MIB-1 expression. These findings indicate that high COX-2 expression in tumor cells is associated with clinically more aggressive gliomas and is a strong predictor of poor survival.

Fibroblast growth factor receptor gene expression and immunoreactivity are elevated in human glioblastoma multiforme.

Glioblastomas were examined for abnormalities in fibroblast growth factor receptor (FGFR) expression by polymerase chain reaction and immunocytochemical analysis. Polymerase chain reaction analysis demonstrated that FGFR1 mRNA levels were significantly higher in glioblastomas than in normal brain adjacent to the tumor or in untransformed human brain. These results were consistent with immunocytochemical localization of FGFR1 protein in glioblastomas: glioblastoma cells exhibited intense FGFR1 immunoreactivity in frozen sections of tumor and low to undetectable FGFR1 immunoreactivity in adjacent normal brain or in normal white matter obtained from patients without neoplastic disease. Endothelial cells of capillaries and larger vessels within the tumor were devoid of FGFR1 immunoreactivity. All glioblastomas evaluated in the present study expressed FGFR1 mRNA and FGFR1 immunoreactivity. Examination of the FGFR1 gene by Southern blot analysis indicated that overexpression of FGFR1 mRNA in glioblastomas did not result from gene amplification. These results indicate that glioblastoma cells, in contrast to endothelial cells within the tumor, display increased levels of FGFR1. Therefore, FGFR1 signal transduction may be associated with increased autocrine growth activity of tumor cells and is probably not related to the increased endothelial cell proliferation associated with these tumors.

Expression and localization of urokinase-type plasminogen activator receptor in human gliomas.

Urokinase-type plasminogen activator receptors (uPARs) play an important role in tumor invasion by localizing degradative enzymes at the invasive zone. In the present study, we examined the presence and distribution of uPARs in human gliomas in vivo. The amounts of uPARs were measured by radioreceptor assays and Northern blotting and were significantly higher in anaplastic astrocytomas and glioblastomas than they were in normal brain tissues and low-grade gliomas. In situ hybridization was performed to investigate the cellular source of uPAR mRNA in various types of astrocytomas and normal brain tissues. uPAR mRNA was localized in astrocytoma cells and endothelial cells within tumor tissue, especially near sites of vascular proliferation and at the leading edges of tumors. uPAR mRNA was also expressed in tumor cells near necrotic areas. Expression was barely detectable in low-grade astrocytomas and normal brain tissues. These results suggest that expression of uPAR in the invading astrocytoma cells may play a significant role in the invasive behaviors of glioblastomas.

Expression and cellular localization of messenger RNA for plasminogen activator inhibitor type 1 in human astrocytomas in vivo.

We investigated the expression and cellular localization of plasminogen activator inhibitor type 1 (PAI-1) in human astrocytoma in vivo. Northern blot and densitometric quantitation of PAI-1 mRNA indicated that PAI-1 transcripts were significantly higher in human malignant astrocytomas and especially in glioblastomas than in low-grade gliomas and normal brain tissues in vivo. Using in situ hybridization with paraffin-embedded surgical specimens of human gliomas and normal brain tissues, PAI-1 mRNA was abundantly expressed in glioblastomas. PAI-1 mRNA was localized mainly in tumor cells and endothelial cells. The distribution of PAI-1 mRNA expression was particularly abundant around areas of vascular proliferation and in remnant tumor cells surrounding necrotic foci. PAI-1 mRNA was also expressed in both the tumor and endothelial cells of anaplastic astrocytomas, whereas it was not expressed or only weakly expressed in low-grade astrocytomas or normal brain tissues. These results suggest that high expression of PAI-1 is associated with the malignant progression of astrocytic tumors and that excessive PAI-1 expression might be associated with intratumoral necrosis in glioblastomas.

Expression and localization of urokinase-type plasminogen activator in human astrocytomas in vivo.

Plasminogen activators regulate a variety of processes involved in tissue morphogenesis, as well as cell differentiation, migration, and invasion. We examined the relative amounts of mRNA and protein and localization of urokinase-type plasminogen activator (uPA) in human astrocytomas in vivo. Using fibrin zymography and densitometric quantitation, we found that uPA activity was significantly higher in malignant astrocytomas, especially in glioblastomas, than it was in normal brain tissues or low-grade gliomas. The amounts of uPA mRNA, as determined by Northern blot analysis, were higher in anaplastic astrocytomas and glioblastomas than in normal brain tissues and low-grade gliomas, consistent with the amount of uPA activity. To investigate the cellular source of uPA in various tissues, we performed immunocytochemical localization of uPA protein and in situ hybridization of uPA mRNA with astrocytomas and normal brain tissues. Immunocytochemical staining for uPA showed strong immunoreactivity in the tumor cells and vasculature of glioblastomas and anaplastic astrocytomas but undetectable or very low immunoreactivity for uPA in low-grade gliomas and normal brain tissues. uPA mRNA was located in astrocytoma and endothelial cells and was heterogeneously distributed within glioblastoma, with preferential localization near vascular proliferation and at the leading edge of the tumor. uPA expression was dramatically higher in highly malignant astrocytomas, especially glioblastomas, and was correlated with malignant progression of astrocytomas.

Altered expression and distribution of heparan sulfate proteoglycans in human gliomas.

The expression of heparan sulfate proteoglycans (HSPGs) by human glioma cells was examined by biochemical and immunological methods in vitro and in vivo. Chondroitin sulfate was shown to represent the major [3H]glucosamine-labeled glycosaminoglycan synthesized by cultured normal brain cells. However, high-grade glioma-derived cells were shown to express significantly increased quantities of hyaluronic acid and heparan sulfate and approximately equal amounts of chondroitin sulfate compared with normal glial cells. To investigate further the differential expression of HSPGs, proteoglycans were isolated from glioma cells and were used as an immunogen to generate monoclonal antibodies (MAbs). One of these MAbs, 39H (an IgM), was shown to bind more to high-grade glioma-derived cells then to low-grade glioma or normal brain cells in vitro. MAb 39H was also observed to bind to isolated HSPGs but not to heparan sulfate glycosaminoglycan chains or trypsin-treated cells. Immunofluorescence staining of the cultured high-grade glioma cells revealed an intense diffuse cell surface staining pattern over the entire cell and also isolated footpads. In contrast, the low-grade tumor or normal glial cells showed a distinctive punctated staining. A similar differential staining of MAb 39H was most prominent between tissue sections of glioblastoma multiforme and anaplastic astrocytomas versus low-grade astrocytomas and normal brain. The low grade gliomas exhibited a weak punctated staining, whereas the high-grade gliomas showed significantly more intense staining, particularly along the apical regions of the cells. These results suggest that altered expression of HSPGs may be related to the malignant transformation or growth potential of glial-derived cells.

Reactivation of insulin-like growth factor binding protein 2 expression in glioblastoma multiforme: a revelation by parallel gene expression profiling.

We carried out a gene expression profiling study using cDNA array technology with 24 primary glioma tissues of low-grade (oligodendroglioma), intermediate-grade (anaplastic oligodendroglioma and anaplastic astrocytoma), and high-grade (glioblastomas multiforme) tumors and found that insulin-like growth factor binding protein 2 (IGFBP2) was consistently overexpressed only in glioblastoma multiforme. The cDNA array results were confirmed by Northern and Western blotting. The fact that the IGFBP2 gene, which is normally expressed in fetal cells and turned off in adult cells, becomes reactivated in the most advanced stage of glioma suggests that glioma progression is a result of dedifferentiation or results from a block of differentiation. Identification of IGFBP2 as a gene associated with glioma progression demonstrates the power and utility of high-throughput gene expression profiling in cancer gene discovery.

Mutagen sensitivity and risk of gliomas: a case-control analysis.

Although the risk factors contributing to the etiology of brain tumors remain largely unknown, this pilot study suggests that genetically determined sensitivity to environmental carcinogens may play a role in the pathogenesis of these tumors. In this study, we examined short-term lymphocyte cultures from 45 adult malignant glioma patients and 117 age-, sex-, and ethnicity-matched healthy controls for mutagen-induced chromatid breaks and evaluated their family history of cancer, smoking, and demographic variables to ascertain the association between mutagen sensitivity and risk of brain tumors. The mutagen selected was gamma-radiation. The mean number of induced breaks/cell was 0.72 (SD=0.45) for the cases and 0.45 (SD = 0.35) for the controls (P < 0.0001). Using the median number of induced breaks/cell in the controls as the breakpoint for defining mutagen sensitivity, we observed an unadjusted odds ratio of 5.36 (95% confidence interval = 2.12-13.69) for mutagen sensitivity and brain tumor risk and an adjusted odds ratio of 5.79 (2.26-14.83), when we controlled for epidemiological risk factors including smoking, race, income, and education. Although a larger study is needed to confirm this intriguing result, these preliminary findings suggest that increased sensitivity to radiation is an independent risk factor for gliomas.

Reduced expression of mismatch repair genes measured by multiplex reverse transcription-polymerase chain reaction in human gliomas.

Microsatellite instability (MIN) is frequently observed in hereditary nonpolyposis colon cancer and in other sporadic cancers including gliomas. Abnormalities in at least one of five mismatch repair (MMR) genes are implicated in the development of cancers in hereditary nonpolyposis colon cancer and the associated MIN. Using a newly developed multiplex reverse transcription-PCR assay, we evaluated the expression of the five known human MMR genes (hMSH2, hMLH1, hPMS1, hPMS2, and GTBP) in human gliomas by measuring simultaneously the relative levels of the transcripts. The beta-actin gene was used as an internal control for RNA degradation and DNA contamination and as a reference for quantifying the levels of their transcripts. Of the 33 gliomas examined, 42% (14) had low expression of hMSH2 (at least 4-5-fold lower than normal mean), 21% (7) had low expression of hMLH1, and 18% (6) had low expression of hPMS1 compared with the expression in the lymphocytes from 13 normal individuals. Furthermore, six of the 33 (18%) tumor samples had decreased expression of more than one MMR gene. Two of these six patients with multiple gene abnormalities had second primary cancers, and an additional patient had multifocal gliomas. Further molecular analysis of available DNA samples indicated that one of five of those tumors with aberrant expression of MMR genes had MIN, as compared with none of five tumors with normal expression. These data suggest that reduced expression of MMR genes is frequent in human gliomas and that aberrant expression of more than one MMR gene may be associated with increased risk of second primary malignancies in glioma patients.

Hypermethylation of the CpG island of p16/CDKN2 correlates with gene inactivation in gliomas.

There is considerable evidence that lack of p16 protein expression is a frequent event in human gliomas. Nevertheless, the molecular mechanisms underlying this absence of p16 protein expression are not completely understood. In some gliomas, homozygous deletions are the main cause of p16/CDKN2 gene inactivation. However, other gliomas lacking p16 expression exhibit intact p16/CDKN2 gene, suggesting that p16/CDKN2 is down-regulated at the transcriptional level. In this study we investigated whether aberrant p16/CDKN2 gene methylation correlated with absence of p16 expression in the latter group of gliomas. In a series of 27 gliomas, 12 malignant tumors exhibited loss of p16/CDKN2 expression but not gene deletion. Methylation analysis of the CpG island in the 5' region of the p16/CDKN2 gene showed that exon 1 was extensively methylated in six and partially methylated in the other six of the 12 malignant gliomas. In contrast, no methylation was observed in four other malignant gliomas and two low-grade gliomas that expressed p16 protein. These results indicate that abnormal hypermethylation of the CpG island encompassing the 5' end of the p16/CDKN2 gene may be a mechanism of transcriptional silencing in gliomas without homozygous deletions.

Increased levels of p21WAF1/Cip1 in human brain tumors.

The cdk inhibitor p21WAF1/Cip1 (p21), which can be transcriptionally activated by p53, functions to block cell cycle progression. In this study, we analysed the expression of p21 in normal and reactive brain and in gliomas of various malignancy grades. Southern blotting showed no p21 gene deletion. Western blotting and immunohistochemical assay showed that the levels of p21 protein in normal and reactive brain tissue were very low; however, p21 was elevated in a majority of gliomas tested, regardless of their malignancy grades. In glioblastoma multiforme, marked elevation of p21 was observed in samples harboring either wild-type or mutant p53. But, in anaplastic astrocytomas, the level of p21 was not elevated in samples harboring mutant-type p53. Immunohistochemical staining of paraffin-embedded astrocytomas and glioblastomas showed that tumor cells and not contaminating normal cells were positive for p21. Therefore, overexpression of p21 appears to be an early event in the development of glial neoplasms and p53-dependent p21 expression appears to be tumor grade specific.

Prognostic significance of p53 immunoreactivity in patients with glioma.

Abnormal p53 as revealed by immunostaining has been shown to be a predictor of poor outcome in a variety of malignant tumors. This study examines the relationship of p53 immunostaining and survival in 182 adult patients with gliomas. Tumor tissues obtained from patients with glioma within 4 months of initial diagnosis were investigated by immunohistochemical analysis for detection of p53 protein abnormalities using the monoclonal antibody PAb 1801. There were 122 patients with glioblastoma multiforme, 48 patients with anaplastic glioma, and 12 patients with low-grade glioma. Among these patients, 73 of those with glioblastoma multiforme, 35 with anaplastic glioma, and 6 with low-grade glioma had positive p53 immunoreactivity. Kaplan-Meier survival plots (log rank test) showed that the patients with anaplastic astrocytoma or low-grade glioma and p53-positive tumors had longer survival times compared to the patients with p53-negative tumors. No differences in survival were detected among the glioblastoma patients. Cox proportional hazards regression analysis, adjusted for age at diagnosis, showed that the p53 positivity was a significant predictor of longer survival (relative risk = 0.56; 95% confidence intervals = 0.35, 0.90; P = 0. 015) in anaplastic astrocytoma patients, but not in glioblastoma patients (relative risk = 1.03; 95% confidence intervals = 0.82, 1. 29; P = 0.80). These results suggest that anaplastic glioma patients with p53 protein alterations may have a better response to chemoradiation, possibly because the malignant cells cannot arrest in G1 to correct lethal damage induced by chemotherapy or radiotherapy.

Phase III randomized study of postradiotherapy chemotherapy with alpha-difluoromethylornithine-procarbazine, N-(2-chloroethyl)-N'-cyclohexyl-N-nitrosurea, vincristine (DFMO-PCV) versus PCV for glioblastoma multiforme.

Although the efficacy of the nitrosourea-based combination chemotherapy procarbazine, N-(2-chloroethyl)-N'-cyclohexyl-N-nitrosurea, and vincristine (PCV) has been previously demonstrated in the setting of anaplastic/intermediate-grade gliomas, the benefit for glioblastoma patients remains unproven. In the current study, we sought to determine whether the addition of alpha-difluoromethylornithine (eflornithine), an inhibitor of ornithine decarboxylase, which has shown encouraging results in the setting of recurrent glioma patients, to a nitrosourea-based therapy (PCV) would constitute a more effective adjuvant therapy in the treatment of glioblastoma multiforme patients in the postradiation therapy setting. Following conventional radiation therapy, 272 glioblastoma (GBM) patients were randomized to receive either alpha-difluoromethylornithine-PCV (DFMO-PCV; 134 patients) or PCV alone (138 patients), with survival and time to tumor progression being the primary endpoints. The starting dosage of DFMO was 3.0 g/m2 p.o. q8h for 14 days before and after treatment with N-(2-chloroethyl)-N-cyclohexyl-N-nitrosurea; PCV was administered as previously described1. Clinical and radiological (Gadolinium-enhanced MRI) follow-ups were nominally at the end of each 6 or 8 week cycle (PCV at 6 weeks; DFMO-PCV at 8 weeks). Laboratory evaluations for hematologic and other adverse effects were at 2 week intervals. There was no difference in median survival or median time-to-tumor progression between the two treatment groups, as measured from day of commencement of postradiotherapy chemotherapy [MS (months): DFMO-PCV, 10.5; Overall survival, as measured from time of tumor diagnosis at first surgery, was 13.3 and 14.2 months at the median and 6.2 and 8.7% at 5 years, respectively, for the DFMO-PCV and PCV arms. The treatment effect was unchanged after adjustment for age, performance status (KPS), extent of surgery, and other factors using the multivariate Cox proportional hazard model. Adverse effects associated with DFMO consisted of gastrointestinal (diarrhea nausea/vomiting), cytopenias, and minimal ototoxicity (limited to tinnitus) at the dose range tested. The addition of DFMO to the nitrosourea-based PCV regimen in this phase III study demonstrated no additional benefit in glioblastoma patients, underscoring the resistance of glioblastoma multiforme tumors to alkylating agents. For patients with anaplastic (intermediate grade) gliomas, in which the previously demonstrated benefit of post-radiation chemotherapy is more substantial, the evaluation of DFMO-PCV vs. PCV is still ongoing and hopefully will yield more encouraging results.

Immunohistochemical localization of plasminogen activator inhibitor type 1 in human brain tumors.

We investigated the immunohistochemical localization of plasminogen activator inhibitor type 1 (PAI-1) in surgical specimens of 41 human intracranial tumors. Malignant tumors (7 glioblastoma multiforme, 5 anaplastic gliomas, 4 malignant meningiomas, and 9 metastatic tumors) showed consistently stronger PAI-1 immunohistoreactivity than benign or low-grade tumors (5 low-grade gliomas, 8 benign meningiomas, and 3 Schwannomas). Strong PAI-1 positivity was confined to glomeruloid-shaped proliferative vessels seen in high-grade gliomas and metastatic tumors. Blood vessels near necrotic foci and some zones of necrosis also showed strong PAI-1 positivity. PAI-1 localizes in the vascular basement membrane and perivascular connective tissue, while endothelial cells themselves show weak positivity. None of the specimens showed PAI-1 reactivity within the tumor cells per se. Localization of PAI-1 in proliferating vessels suggests that PAI-1 may be involved in angiogenesis.

Genetic polymorphisms in glutathione S-transferase mu and theta, N-acetyltransferase, and CYP1A1 and risk of gliomas.

The role of genetic polymorphisms in modulating susceptibility to carcinogenic exposures has been well explored for tobacco-related neoplasms but not for other neoplasms including gliomas. It is relevant to explore these polymorphisms because certain carcinogenic exposures such as nitrosamines are implicated in the risk of gliomas. We therefore conducted a pilot case-control study to examine the role of polymorphisms in GSTM1, GSTT1, NAT2 (rapid, intermediate, and slow acetylation), and CYP1A1 and risk of glioma. Ninety patients diagnosed with glioma were ascertained as part of an ongoing genetic epidemiological study and were age, gender, and race matched with 90 healthy controls. We used PCR based methodology to determine the prevalence of the above genetic polymorphisms using sequences and PCR conditions directly adapted from studies reported previously. We calculated univariate odds ratios and performed multiple logistic regression to assess interactions between polymorphisms. We found no statistically significant associations between the null genotypes of GSTM1 and GSTT1, and CYP1A1 and risk of gliomas. However, there was an intriguing pattern with NAT2 acetylation status (odds ratios, 1.81, 1.34, and 0.61 for rapid, intermediate, and slow acetylation, respectively; P = 0.10 for trend). It is unlikely that any single polymorphism is sufficiently predictive of risk, and a panel of markers integrated with epidemiological data should be conducted on a large number of study subjects to fully understand the role of genetic polymorphisms and brain tumor risk.

Descriptive epidemiology of primary brain and CNS tumors: results from the Central Brain Tumor Registry of the United States, 1990-1994.

The Central Brain Tumor Registry of the United States (CBTRUS) obtained 5 years of incidence data (1990-1994)--including reports on all primary brain and CNS tumors--from 11 collaborating state cancer registries. Data were available for 20,765 tumors located in the brain, meninges, and other CNS sites, including the pituitary and pineal glands. The average annual incidence was estimated at 11.5 cases per 100,000 person-years. The higher incidence of tumors in male patients (12.1 per 100,000 person-years) than in female patients (11.0 per 100,000 person-years) was statistically significant (P < 0.05); the higher incidence in whites (11.6 per 100,000 person-years) compared with blacks (7.8 per 100,000 person-years) was statistically significant (P < 0.05). The most frequently reported histologies were meningiomas (24.0%) and glioblastomas (22.6%). Higher rates for glioblastomas, anaplastic astrocytomas, oligodendrogliomas, anaplastic oligodendrogliomas, ependymomas, mixed gliomas, astrocytomas not otherwise specified, medulloblastomas, lymphomas, and germ cell tumors in male than in female patients were statistically significant (P < 0.05), with relative risks (RR) ranging from 1.3 to 3.4. Meningiomas were the only tumors with a significant excess in females (RR = 0.5). We noted higher occurrence rates in whites than in blacks for the following histologies: diffuse astrocytomas, anaplastic astrocytomas, glioblastomas, oligodendrogliomas, ependymomas, mixed gliomas, astrocytomas NOS, medulloblastomas, nerve sheath tumors, hemangioblastomas, and germ cell tumors, with RRs ranging from 1.5 to 3.4. Racial differences in occurrence rates were not observed for predominately benign meningiomas or pituitary tumors. This study represents the largest compilation of data on primary brain and CNS tumors in the United States. Standard reporting definitions and practices must be universally adopted to improve the quality and use of cancer registry data.

Suppression of human glioma growth by adenovirus-mediated Rb gene transfer.

OBJECTIVE: This study was conducted to obtain evidence that restoration of the retinoblastoma protein function may have therapeutic application for gliomas. BACKGROUND: The development of glioblastoma multiforme involves progressive inactivation of several tumor suppressor genes. Abnormalities of the retinoblastoma tumor suppressor gene are found in the majority of cancers, including at least 30% of malignant gliomas. No final evidence has been produced about the role of Rb in suppressing glioma growth. METHODS: To address this question, the Ad5CMV-Rb adenovirus carrying a 3.2-kb cDNA of the Rb gene was constructed. Expression of the exogenous protein was assessed by immunoblot and immunohistochemistry analyses. Growth curve assays were used to evaluate the effect of the Rb protein on glioma cell growth. Flow-cytometry analyses were used to analyze the phenotype of the cell cycle after the transfer of Rb. Human glioma xenografts implanted subcutaneously in nude mice were used for the tumorigenicity assay. RESULTS: After the transfer of Rb, 80% of the treated cells expressed high levels of the retinoblastoma protein for at least 7 days. Within 5 days of treatment, the cells lost the neoplastic morphology and showed marked growth suppression. The majority of the Rb-expressing cells were arrested in the G1 phase of the cell cycle. In addition, the restoration of the retinoblastoma activity rendered the human glioma cells unable to form tumors in nude mice. CONCLUSIONS: These findings provide direct evidence that inactivation of the retinoblastoma protein is a critical event in gliomas, and suggest that the restoration of wild-type retinoblastoma activity in these tumors may have therapeutic utility.

Neuropathology of malignant gliomas.

Astrocytomas are the most common primary gliomas, with the highly anaplastic glioblastoma multiforme being the most frequently occurring astrocytoma. Distinctive histological features permit astrocytomas to be graded into levels of anaplasia, and these histological grades correlate with biological behavior and patient prognosis. However, there is also a strong correlation between patient age, tumor grade, and prognosis. More objective indicators of tumor proliferative potential, such as BUdR or Ki-67 LI, are currently being investigated with the hope that these will be a more accurate means of predicting patient survival. Oligodendrogliomas are a much less common primary glioma, with a generally better survival rate than astrocytomas. However, grading systems for oligodendrogliomas are not well defined. For any type of glioma, subsequent surgery after radiation therapy may be required for treatment of therapeutic effects or for therapy planning at recurrence. The histological changes observed in these post-therapy biopsy specimens or resections may be difficult to distinguish from reactive changes that can simulate recurrent tumor and vice versa.