|Home | About | Journals | Submit | Contact Us | Français|
While meningiomas are known as slow-growing extra-cerebral neoplasms, the subgroup of secretory meningiomas with histologically benign characteristics tend to cause disproportional peritumoral edema, frequently leading to severe medical and neurological complications in postoperative management. Among 1,484 meningiomas that were resected at our institution between 1990 and 2007, 44 (3%) patients were found to have the histological diagnosis of a secretory meningioma. The clinical course, radiological appearance, and histopathological features were retrospectively analyzed to examine the specifics of these benign lesions. Meningiomas were located at the convexity (n = 14), the cranial base (18), and the sphenoid ridge (12). A severe, nearly hemispheric perifocal edema disproportional to tumor size was seen on preoperative MR imaging in 18 (41%) patients. Following surgical resection, the postoperative course was uneventful in 29 patients. In 15 patients, severe peritumoral edema continued or even increased on postoperative CT imaging. Six patients showed midline shift and clinical worsening necessitating respirator-assisted ventilation and intracranial pressure monitoring. An association between the extent of brain edema and number of periodic acid Schiff–positive pseudopsammomas was found (p < 0.02). Further, the size of the edema correlated with the number of immunohistochemically detected cells expressing carcinoembryonic antigen (CEA) and cytokeratin (CK) (p < 0.01). Mean MIB-1 (Ki-67 antigen) proliferation index was 3.0% (range, 0%–17%) and did not correlate with edema or tumor recurrence. Secretory meningiomas are frequently associated with severe peritumoral edema. The extent of edema correlates with immunohistochemically detected expression of CEA and CK. Extended perifocal edema in meningiomas is an unusual finding and should alert the neurosurgeon that surgery may aggravate edema excessively, leading to a life-threatening postoperative situation.
Secretory meningioma is a rare benign subtype of meningioma1–3 containing glandular lumina with secretory globules. These so-called pseudopsammomas are suspected to be the cause of an unusually severe peritumoral edema,2–7 but its exact pathogenesis is still unknown. Life-threatening complications are seen more frequently than expected in the first postoperative days, making intensive care medicine in some of these patients mandatory.
The present study is the largest single-center experience correlating the clinical and histopathological features of secretory meningiomas.
Among 1,484 meningiomas resected at our institution between 1990 and 2007, 44 (3.0%) patients were diagnosed to have a secretory meningioma. CT and MR images were retrospectively analyzed to determine the tumor size and the extent of perifocal edema in the pre- and postoperative course. Perifocal edema was correlated to the tumor size using a modified version of the classification of Buhl et al.5 Edema smaller or equal in size to the tumor was classified as grade I. In grade II, the extent of edema exceeded the size of the tumor, and a nearly hemispheric edema was defined as grade III.
The histopathological samples were assessed according to WHO criteria8 by two neuropathologists. For routine diagnostics, formalin-fixed and paraffin-embedded material was stained with hematoxylin and eosin, periodic acid Schiff (PAS), and Elastica van Gieson stain. In addition, the tissue was investigated immunohistochemically for expression of epithelial membrane antigen and Ki-67 antigen (MIB-1). Sections were further labeled with antibodies against carcinoembryonic antigen (CEA) and broad-spectrum cytokeratin (CK). Bound antibodies were visualized using the alkaline phosphatase–anti-alkaline phosphatase method. Fast red was applied as chromogen. Content of PAS-positive pseudopsammomas and expression of CEA and CK were arbitrarily graded in five levels: 1, scattered pseudopsammomas and CEA/CK-positive cells within the tumor; 2, loosely scattered small groups of pseudopsammomas and CEA/CK- positive cells; 3, focally up to 10% of tumor area occupied by pseudopsammomas and CEA/CK-positive cells; 4, focally up to 20% of tumor area occupied by pseudopsammomas and CEA/CK-positive cells; 5, focally up to 30% of tumor area occupied by pseudopsammomas and CEA/CK-positive cells. In addition to the secretory differentiation of the tumor, other growth patterns of the meningiomas that were observed in the histopathological assessment (e.g., meningothelial or fibroblastic) were included to give a more complete description of the morphological features (Table 1). Histopathological aspects were correlated with the clinical and radiological course and were statistically evaluated using Kendall tau-B and Spearman rho nonparametric methods (bivariate correlations).
The extent of edema was correlated with the histopathological findings and the postoperative clinical course.
Secretory meningiomas were confirmed in 39 female and 5 male (8:1 ratio) patients. Mean age was 58.0 years (range, 35–78 years). None of the patients previously had a meningioma or was treated for another brain tumor. The tumors were located at the convexity (14), cavernous sinus (4), middle cranial base (11), sphenoid wing (12), olfactory grove (2), and foramen magnum (1). Clinical data are summarized in Table 1.
Perifocal edema equal to or smaller than the size of the tumor (grade I) was radiologically seen in 6 (14%) patients; a perifocal edema exceeding the tumor size (grade II) was observed in 3 (7%) patients, and a severe, nearly hemispheric edema (grade III) was diagnosed in 18 (41%) patients. No perifocal edema was seen in 17 (38%) patients. Complete surgical resection was achieved in 38 (86%) patients (Simpson grade I and II). Tumors were incompletely resected in six patients due to infiltration of the cavernous sinus. Statistically complete resection was significantly more often accomplished in tumors of the convexity than in tumors located at the base of the brain (p < 0.01).
The postoperative course was uneventful in 29 patients. These patients were observed in the intensive care unit (ICU) for 12 h and mobilized on the first postoperative day. Fifteen patients experienced progressive neurological symptoms or even loss of consciousness during the postoperative period. Perifocal edema exceeding the tumor size or reaching nearly hemispheric size (grade II and III) was seen in all of these 15 patients already preoperatively. Postoperative CT imaging in 6 of 15 patients showed constant or increasing edema-related mass effect necessitating aggressive antiedematous therapy with respirator-assisted ventilation and intracranial pressure monitoring for 3–11 days. In these six cases, patient age (mean, 56 years) and location of the tumor (convexity, n = 2; cranial base, n = 2; sphenoid wing, n = 2) were unsuspicious. Perifocal edema exceeding the tumor size (grade II) was observed in all six cases, but its postoperative aggravation was not predictable on admission. An illustrative case is shown in Figs. 1 and and2.2. No deaths were seen in our study group. Statistically, the extent of resection correlated with severity of postoperative brain edema (p < 0.02).
Histopathological investigation showed secretory meningiomas with PAS-positive pseudopsammomas in all cases (Table 1). There were no histological features of malignancy. CEA and CK expression were also found in all cases (Fig. 3). Statistical evaluation revealed a highly significant correlation between CK and CEA expression and PAS-positive pseudopsammomas (p < 0.001, CEA vs. PAS; p < 0.001, CK vs. PAS; p < 0.001, CK vs. CEA), as well as a correlation between brain edema and PAS, CK, and CEA, respectively (p < 0.02, edema vs. PAS; p < 0.01, edema vs. CEA; p < 0.01, edema vs. CK). There was no correlation between secretory parameters and MIB-1 proliferation index, which ranged from 0% to 17%, with a mean of 3.0%.
Follow-up ranged from 2 weeks to 8 years, with a medium time of 3.8 years. Seventeen patients were lost to follow-up or had inaccessible archive data. In two patients the tumor recurred after 2 and 6 years, respectively, most likely due to incomplete resection because specific histopathological features could not be seen in these cases. No deaths related to the meningioma were seen during follow-up.
Apart from one study of 31 patients9 and another two analyses of neurofibromatosis type 2 mutations in 14 and 33 patients,10,11 only small series (n < 14) on secretory meningiomas have been published so far.1–16 The prevalence is reported in the literature to range between 1.2% and 9.3%.2–5 A female:male ratio of 11:0 reported by Buhl et al.4 has not been confirmed by other series. However, a female predominance seems to be a characteristic feature of this subtype and may be supported by the finding that secretory meningiomas show high rates (33%–100%) of progesterone and estrogen receptors.2–5
Our retrospective analyses of 44 secretory meningioma patients showed a strong correlation between the extent of edema and the secretory component. In 18 (41%) patients, a severe peritumoral edema on T2-weighted images was found involving the whole hemisphere, which is comparable to the observations made by Buhl et al.4 (45%), Probst-Cousin et al.3 (64%), Colakoglu et al.2 (60%), and Tirakotai et al.12 (43%). Only 14% (6 of 44) of the secretory meningiomas reported here showed edema smaller than or equal to the size of the tumor, whereas 67% of all meningioma patients in our cohort (1,484 cases) presented with no edema (234 patients) or with edema smaller than or equal to the size of the tumor. Immunohistochemical analyses in our study showed a significantly higher expression of CK and CEA in patients with large edema. PAS was more than twice as high in the edema grade III patients as in the group with mild or moderate edema (grade I or II), suggesting a close correlation between peritumoral edema and CEA-, CK-, and PAS-positive pseudopsammomas.
Although all cases showed specific features of secretory meningioma upon histopathological workup, not all patients had severe peritumoral edema requiring ICU therapy in their postoperative period. In our series, 29 of 44 patients, including patients with grade II and III edema (n = 6), had an uneventful postoperative course. Therefore, it still remains unclear why some tumors of this meningioma subgroup are associated with a life-threatening edema and others are not. Furthermore, secretory meningiomas of the skull base may react in a different way because in our study, partial tumor resection in six patients with meningiomas of the cavernous region was followed by an uneventful postoperative course.
Compromised or thrombosed veins due to surgical retraction of the brain could be excluded as the cause for edema. Also, venous compression, which is known to cause peritumoral edema, is expected to resolve and decrease with the resection of the tumor and so seems unlikely to be the cause for increased swelling in secretory meningioma. In secretory meningioma, pseudopsammomas are associated not only with CEA and CK expression but also with increased numbers of mast cells.12 The latter may release substances leading to vascular dilatation and increased permeability. Therefore, it might be the surgical preparation of the tumor itself that by activation of mast cells triggers the release of vascular mediators (e.g., histamine), resulting in brain edema. Hence, mast cells seem to be a promising subject for future studies.
In summary, the diagnosis of secretory meningioma should be established as early as possible so that clinicians can be on guard against a potential stormy postoperative course. Neuroradiological signs of a peritumoral edema exceeding the tumor size together with a meningeal origin of the tumor should trigger intraoperative frozen sections even if the diagnosis of meningioma is not in doubt. The subtype of secretory meningioma can be diagnosed by frozen sections with high accuracy. Awareness in the clinical course enables the postoperative caregivers to brace for extended sedation, antiedematous therapy, and a lower threshold for postoperative imaging.
Secretory meningiomas are characterized by clinical and histopathological features that differ from the overall group of meningiomas. Reviewing the literature, a female predominance is evident (female:male ratio, 8:1). In 41%–64% of the cases, a suspicious disproportion between small tumor size and extended perifocal edema is observed that may justify an aggressive antiedematous therapy before surgery in order to prevent postoperative aggravation of edema. Pseudopsammomas and expression of CEA and CK in secretory meningiomas are closely related to cytotoxic edema leading to reduced compliance of the brain and life-threatening postoperative complications in some patients. Apart from this acute perioperative problem, secretory meningiomas show a prognosis comparable to the overall group of benign meningiomas.