C M Abramovich1, R A Prayson. 1. Department of Anatomic Pathology, Cleveland Clinic Foundation, Cleveland, Ohio 44195, USA.
Abstract
BACKGROUND: Predicting the behavior of meningiomas based on histopathologic features alone has remained problematic. DESIGN: This study retrospectively compared several histopathologic features and MIB-1 labeling indices (LIs) in recurrent meningiomas with those of nonrecurrent meningiomas. Six histopathologic features, including mitoses, necrosis, loss of architectural pattern, hypervascularity/hemosiderin deposition, prominent nucleoli, and nuclear pleomorphism, were compared between 32 recurrent and 27 nonrecurrent meningiomas using Fisher exact tests. MIB-1 LIs (% positive tumor cell nuclei) were compared using the Wilcoxon rank sum test. RESULTS: The patients in the recurrent group included 26 women (mean age, 55 years), who developed 1 to 5 recurrences. Time intervals to the first recurrence ranged from 5 to 183 months (mean, 55 months). The nonrecurrent group included 21 women (mean age, 56 years), with follow-up ranging from 88 to 124 months (mean, 109 months). Of the histopathologic features evaluated, statistically significant differences between the recurrent and nonrecurrent groups were found only with respect to prominent nucleoli (P =.024) and nuclear pleomorphism (P <.001), both of which were more common in the recurrent group. In the recurrent group, 9 tumors were considered malignant (defined by brain invasion or metastasis) versus 2 of the nonrecurrent meningiomas. Nineteen percent of nonrecurrent tumors versus 41% of recurrent tumors had 2 or more of the 6 histopathologic features. MIB-1 LIs in the nonrecurrent group ranged from 0 to 8.3 (mean, 1.5) and were generally lower than those in the recurrent group (range, 0-32.5; mean, 5.4); no statistical difference was identified between these groups. No statistically significant difference with regard to histology or MIB-1 LIs was noted between the initially excised recurrent tumor and the most recently resected recurrence. CONCLUSIONS: Of the histopathologic features examined, only prominent nucleoli and nuclear pleomorphism were found to be statistically more common in recurrent than nonrecurrent meningiomas. The mean MIB-1 LI was higher in the recurrent than in the nonrecurrent group, although there was no statistical difference between means and there was clear overlap with regard to MIB-1 LI ranges.
BACKGROUND: Predicting the behavior of meningiomas based on histopathologic features alone has remained problematic. DESIGN: This study retrospectively compared several histopathologic features and MIB-1 labeling indices (LIs) in recurrent meningiomas with those of nonrecurrent meningiomas. Six histopathologic features, including mitoses, necrosis, loss of architectural pattern, hypervascularity/hemosiderin deposition, prominent nucleoli, and nuclear pleomorphism, were compared between 32 recurrent and 27 nonrecurrent meningiomas using Fisher exact tests. MIB-1 LIs (% positive tumor cell nuclei) were compared using the Wilcoxon rank sum test. RESULTS: The patients in the recurrent group included 26 women (mean age, 55 years), who developed 1 to 5 recurrences. Time intervals to the first recurrence ranged from 5 to 183 months (mean, 55 months). The nonrecurrent group included 21 women (mean age, 56 years), with follow-up ranging from 88 to 124 months (mean, 109 months). Of the histopathologic features evaluated, statistically significant differences between the recurrent and nonrecurrent groups were found only with respect to prominent nucleoli (P =.024) and nuclear pleomorphism (P <.001), both of which were more common in the recurrent group. In the recurrent group, 9 tumors were considered malignant (defined by brain invasion or metastasis) versus 2 of the nonrecurrent meningiomas. Nineteen percent of nonrecurrent tumors versus 41% of recurrent tumors had 2 or more of the 6 histopathologic features. MIB-1 LIs in the nonrecurrent group ranged from 0 to 8.3 (mean, 1.5) and were generally lower than those in the recurrent group (range, 0-32.5; mean, 5.4); no statistical difference was identified between these groups. No statistically significant difference with regard to histology or MIB-1 LIs was noted between the initially excised recurrent tumor and the most recently resected recurrence. CONCLUSIONS: Of the histopathologic features examined, only prominent nucleoli and nuclear pleomorphism were found to be statistically more common in recurrent than nonrecurrent meningiomas. The mean MIB-1 LI was higher in the recurrent than in the nonrecurrent group, although there was no statistical difference between means and there was clear overlap with regard to MIB-1 LI ranges.