Introduction

An estimated 2–3% of the population has an incidental asymptomatic meningioma in autopsy studies [14, 20]. With the wider use of CT and MRI, many meningiomas are discovered as incidental findings during investigation for unrelated symptoms [9, 13, 16, 17, 20, 21]. The growth potential of meningiomas varies. Some meningiomas remain unchanged in size for a long period, whereas others grow rapidly [15]. Sex, age, initial tumor size, and calcification were reported to be related to the tumor growth judging from follow-up scans [9, 13, 16, 17, 21].

The nuclear antigen Ki-67 expressed by proliferating cells has become available for routinely processed paraffin section. The MIB-1 antibody detects an epitope on the Ki-67 antigen, a nuclear protein present only during active phase of the cell cycle (G1, S, G2, and M) [2]. Several studies investigated how Ki-67 labelling indices could help to predict recurrence [1, 3, 4, 6–8, 11, 12, 18, 19]. An increased MIB-1 staining index was highly correlated with a shorter tumor volume doubling time [12]. In the cases that showed an MIB-1 of ≥ 3%, the tumor volume doubling time was <2 years. Nakaguchi et al. [12] found the formula which can calculate tumor doubling time (Td) from the MIB-1 staining index at surgery: log Td=31.4–0.14×MIB-1 SI (R2=0.556). The time interval to the next recurrence for recurrent meningiomas is associated with the MIB-1 staining index. Meningiomas with MIB-1 staining index of 3% or higher had a significantly higher tendency of recurrence [11]. Although these cell kinetics methods are valuable for growth potential, they can be applied only after the verification of pathology [15].

The purpose of this study was to clarify the clinical and radiological features related to meningioma proliferation using the MIB-1 staining index. Sex, age, calcification, edema, symptom, size, and shape of tumor were already reported to be related to the tumor growth and MIB-1 staining index [1, 3, 4–8, 11–13, 16–19, 21]. The results, however, were inconsistent because of a lack of sufficient case number and inadequate statistical analysis. Independent risk factors for high growth rate should be determined. We analyzed the relationship of MIB-1 staining indices to the characteristics of 342 consecutive patients with meningioma surgically removed between 1995 and 2004 by logistic regression analysis.

Materials and methods

Three hundred and forty-two patients with meningiomas were surgically treated in our department of neurosurgery between 1995 and 2004. Radiological features were analyzed by CT scans and/or MRI. Location of the tumor was classified as follows: convexity, falx, parasagittal, sylvian fissure, tentorial, ventricular, foramen magnum, olfactory groove, petroclival, petrous, sphenoid ridge, and tuberculum sellae. The latter six locations were considered as skull base. The tumor volume was calculated using the formula: length × depth × width × 0.5 [9]. When patients had multiple meningiomas, only the largest tumor was included. On the basis of conventional CT and bone window CT, patients were divided into two groups according to the low density area around the tumor and calcification in the tumor. A low density area due to surgical scar was not included in edema in patients with recurrent meningioma.

The tumors were histologically classified according to the World Health Organization classification of tumors [10]. An avidin-biotin immunoperoxidase or simple stain MAX-peroxidase (Nichirei, Tokyo) technique was used to perform MIB-1 monoclonal antibody (DAKO, Denmark) assay in selected sections of each case. All tissue sections were examined at high-power magnification (×400). The number of cells stained positively with MIB-1 and the total number of tumor cells were counted in several representative fields containing more than 1,000 cells. Their ratio was indicated as the MIB-1 staining index (%).

Statistical analysis

All data were stored on a personal computer and analyzed using commercially available statistical software (SPSS version 12.0, SPSS Inc.). Chi-squired analysis was used to compare the MIB-1 staining index to characteristics of patients with meningioma. All variables were included in a logistic regression model to determine which variables were independently associated with a high MIB-1 staining index (≥3.0). Significance was judged at a value of p<0.05 for all analyses.

Results

Table 1 shows the characteristics and MIB-1 staining index of the 342 patients. One hundred and forty-nine of patients with meningioma were ≥60 in age; 89 male; 48 recurrent; 203 symptomatic; 157 at the skull base; 124≥20 cm3 in volume; 24 multiple; 136 with edema; 117 with calcification. We compare these characteristics to the MIB-1 staining index. We divided them into two groups: <3.0 and ≥3.0 [11]. The MIB-1 staining index in 100 of 342 meningiomas was > 3.0. Sex (p=0.0001), recurrence (p=0.0001), symptomatic (p=0.013), volume (p=0.014), edema (p=0.001), and calcification (p=0.0001) were correlated with the MIB-1 staining index by chi-square test; age, skull base, and multiple were not.

Table 1 Characteristics and MIB-1 staining index in 342 meningiomas
Full size table

Meningothelial, transitional, and fibrous meningiomas were the three major subtypes, and they accounted for about three fourth of the total. Two hundred and ninety-six meningiomas belonged to grade I; 28 grade II; and 18 grade III. MIB-1 staining index in 56 of 296 grade I meningiomas was ≥ 3.0; that in 27 of 28 grade II; and that in 17 of 18 grade III, respectively (Table 2).

Table 2 Histological subtypes and MIB-1 staining index of 342 meningiomas
Full size table

Logistic regression analysis demonstrated that male (odds ratio [OR], 2.374, p=0.003), recurrence (OR, 7.574, p=0.0001), skull base (OR, 0.540, p=0.035), calcification (OR, 0.498, p=0.019) were independent risk factors for a high MIB-1 staining index (≥ 3.0); age, symptomatic, volume, multiple, and edema were not (Table 3).

Table 3 Logistic regression analysis for factors independently related to MIB-1 staining index
Full size table

Discussion

We analyzed the relationship of the MIB-1 staining indices to the characteristics of 342 consecutive patients with meningioma surgically removed between 1995 and 2004 by logistic regression analysis. Logistic regression analysis demonstrated that male (odds ratio [OR], 2.374, p=0.003), recurrence (OR, 7.574, p=0.0001), skull base (OR, 0.540, p=0.035), calcification (OR, 0.498, p=0.019) were independent risk factors for a high MIB-1 staining index (≥ 3.0); age, symptomatic, volume, multiple, and edema were not.

The relationship between the growth rate or MIB-1 staining index and age has been controversial: a higher MIB-1 staining index and higher growth rate were observed for younger patients [11, 13, 21]; but not in other reports [1, 12, 15, 19]. Our series of 342 patients with meningioma showed no relation.

It is well known that atypical and anaplastic meningiomas are predominant in males [10]. Matsuno et al. [11] reported that the mean MIB-1 staining index in 50 male patients was 5.5%, whereas that in 77 female patients was 2.7%. Our findings show that male (odds ratio [OR], 2.374, p=0.003) was an independent risk factor for a high MIB-1 staining index. We also found a higher MIB-1 staining index in males even in grade I meningioma (MIB-1 staining index in 32 of 226 females, and in 24 of 70 males was ≥ 3.0, p=0.0001, chi-squire test).

Recurrence (OR, 7.574, p=0.0001) was the most significant independent risk factor for a high MIB-1 staining index (≥ 3.0). Therefore, we propose prompt management for recurrent meningiomas. In most of the recurrent meningiomas, the MIB-1 staining index was higher at the time of recurrence than at the time of initial surgery [1, 11, 19]. Changes in histological morphology and malignant transformation are also known in meningiomas.

Although there is a significant difference in the MIB-1 staining index between symptomatic and non-symptomatic meningiomas by chi-square test, symptomatic is not an independent risk factor for a high MIB-1 staining index. Meningiomas commonly present with seizure disorders, and are associated with location, perilesional edema, and convexity location. Symptoms and signs of elevated intracranial pressure could be due to the large size of meningioma itself, or to the pronounced cerebral swelling resulting from reactive vasogenic edema. Focal neurological deficits caused by meningiomas are generally related to direct local brain, cranial-nerve compression, and can be predicted from the site of origin of the tumor [20]. Thus, symptomatic meningioma may not be related to a high MIB-1 staining index. The growth rate of incidental meningioma may be similar to that of symptomatic meningioma.

Our results demonstrated that skull base (OR, 0.540, p=0.035) is an independent risk factor for a high MIB-1 staining index. No relationship has been reported between the MIB-1 staining index and the location of meningiomas [15, 19]. In general, the surgical risk for meningiomas is higher in skull base. A low proliferative potential in skull base meningiomas should be taken into consideration especially when treating elderly patients with asymptomatic meningiomas [9, 16].

Although there was a significant difference of the MIB-1 staining index in tumor size by chi-square test [15], tumor size is not an independent risk factor for a high MIB-1 staining index. The tumor volume is associated with the annual growth rate but not with doubling time [13, 16, 21]. Assuming that a tumor shows a constant relative volume increase, larger tumors will show higher growth rates if the annual volume increase is expressed in absolute values. Large tumors should be carefully observed even though the initial volume is not a risk factor for a high MIB-1 index, and it is not correlated with doubling time.

The etiology of peritumoral brain edema associated with meningiomas is multifactorial. Factors that may influence the etiology of peritumoral edema include tumor size, histological subtypes, vascularity, venous stasis, and brain invasion [1, 4]. Ide et al. [4] found a significant correlation of both the MIB-1 staining index and tumor size with the extent of edema. A high MIB-1 staining index itself did not seem to be directly responsible for perifocal edema, since our logistic regression analysis demonstrated that edema is not an independent risk factor for a high MIB-1 index [15].

Tumors with calcification grew significantly less than those without calcification [9, 13, 16]. Absence of calcification on CT correlated strongly with doubling time [5]. Diffusely calcified meningiomas had a low mean MIB-1 staining index of 0.57%. Focally calcified tumors showed a relatively low proliferative potential (0.92%) compared with that of noncalcified tumors (1.75%) [15]. The results are always consistent when the relationship between calcification and proliferative potential or growth rate is compared. We also confirmed this characteristic in meningioma (OR, 0.498, p=0.019).

In conclusion, male, recurrence, non-skull base, absence of calcification are independent risk factors for a high proliferative potential. These should be taken into consideration when managing meningiomas.