Meningiomas sometimes appear to stop growing after attaining a large size. Commonly used exponential growth models do not reflect this phenomenon. We attempted to find the best curve to simulate their growth. Fifty-two patients with meningioma were followed up for 3.1-21.7 years (mean 7.5 years) with four or more imaging studies each. Thirty-one patients had asymptomatic tumors. The other 21 patients with residual or recurrent tumor were followed up after surgery. Time-volume curves for each tumor were plotted. Nonlinear regression analyses were performed against power, exponential, logistic, and Gompertzian curves. Time-volume curves corresponded to the Gompertzian and logistic growth curves better than to power or exponential curves. When simulating time-volume curves with Gompertzian curves, the majority of benign meningiomas began to slow their growth before patient age of 80 years. Twenty-three of 31 asymptomatic meningiomas had already passed the inflection point before diagnosis. In contrast, this happened less frequently in symptomatic tumors. Especially, all six atypical meningiomas continued to grow quasi-exponentially. Sigmoid curves that approach a plateau were better descriptors of the growth of benign meningiomas than were curves of unlimited growth. However, atypical meningiomas were unlikely to slow their growth.

Although the expression O6-methylguanine-DNA methyltransferase (MGMT) is an important hallmark for decision of nitrosourea chemotherapy for glioma patients, no immunohistochemical method for analysis of MGMT has been standardized yet. Gliomas usually contain non-neoplastic cells even deep in the tumor. It is not known which of these components expresses MGMT. To clarify this point, we investigated MGMT expression in the non-neoplastic cells in autopsy and surgical specimens by immunohistochemistry. High grade gliomas were also studied to find a cut-off point for treatment decision. MGMT immunohistochemistry in the normal brain or brain with non-neoplastic disease revealed nuclear staining in some endothelial cells, inflammatory cells, ependymal cells, astrocytes and oligodendroglias. Some cells were double stained with CD68 (macrophages or microglias). The neurons were consistently MGMT-negative. High grade gliomas always contained an MGMT-positive non-neoplastic component. Although, the endothelial cells were easily distinguished from the neoplastic cells, other cells were often mistaken for tumor cells. The population of MGMT-positive non-neoplastic cells was usually less than 10%. We set a cut off-point at 10% between the positive and negative groups because the statistical difference in the overall survival was most distinct at this value. In 51 high grade glioma patients, who received both radiotherapy and chemotherapy with nimustine (ACNU), the median overall survival of the MGMT-negative group (23 months) was significantly longer than that of the MGMT-positive group (14 months) (P < 0.009). Multivariate analysis revealed that the negative MGMT expression was a significant prognostic variable next to the degree of surgical removal for the overall survival. In the MGMT-positive group, addition of platinum-based chemotherapy did not improve the survival.

The World Health Organization (WHO) grading system for meningioma is helpful for predicting aggressive subtypes. However, even benign meningiomas sometimes show relatively rapid growth and may recur after total removal. We attempted to find histopathological features that would be valuable for predicting recurrence or regrowth of WHO grade I meningiomas. We investigated 135 benign meningiomas, of which 120 were totally removed (Simpson's grade I-III). The median follow-up period was 9.7 years (1-21 years). The recurrence rate in the patients with total removal was 7.5% at 10 years and 9.3% at 20 years. The univariate analysis revealed that MIB-1 index (>or=2%), existence of mitosis, absence of calcification, and paucity of fibrosis significantly correlated with recurrence. On the other hand, the histological features of sheet-like growth, prominent nucleoli, and necrosis did not correlate with recurrence, because they were relatively rare in grade I tumors. Multivariate analysis revealed that high MIB-1 index and absence of calcification significantly correlated with recurrence. The patients with recurrent or residual tumors did not always receive adjuvant treatment. Including subtotally treated tumors, the retreatment rate was 9.8% at 10 years and 25.6% at 20 years. MIB-1 index and Simpson's grade significantly correlated with retreatment in both univariate and multivariate analyses.

We analyzed the relation between meningioma and the brain in 50 surgical cases. So-called capsule formation was seen in 20 meningiomas, of which 13 were categorized as thin and 7 as thick. In 21 meningiomas the arachnoid membrane was intact, and 10 meningiomas had no underlying arachnoid membrane. The other 19 tumors showed partial disruption of the arachnoid membrane. The degree of arachnoid disruption correlated with the tumor grade, perifocal edema, pial blood supply on angiography, and tumor size. The existence of brain invasion correlated with the tumor grade and partially with tumor size. In case of invasive tumor, GFAP-positive cells were found deep in the tumor, usually in contact with blood vessels. The axons in gliotic brain often showed degenerative changes such as ballooning or varicose swelling. Meningiomas were usually demarcated by a basement membrane that was collagen type 4 (Col4)-positive. However, atypical and anaplastic meningiomas usually lacked Col4 staining at the interface. In two benign meningiomas that looked like an invasive growth, Col4 staining was seen above the brain. A pia mater-like structure covered the tumor surface in both cases. We could not demonstrate a relation between the expression of matrix metalloproteinase (MMP)-2 or MMP-9 and arachnoid disruption or brain invasion.

No abstract