In-vivo proliferative potential of primary human brain tumors; its correlation with histological classification and morphological features: II. Nonglial tumors

Morphological and Ultrastructural Findings of Prognostic Impact in Craniopharyngiomas

Craniopharyngioma is a slow-growing epithelial tumor with an unpredictable tendency to recur. To verify the reliability in predicting the clinical outcome, some morphological and immunohistochemical findings were analyzed in 37 primitive tumors and in 6 recurrences (one recurred twice). All the tumors were surgically excised and all recurrences exhibited an adamantinomatous pattern; mitotic rate was low (< 5 x 10 HPF) in both recurrent and in nonrecurrent tumors. Primary tumors showed a mean positivity of 1.7% (range 0.3-2%) to PCNA vs. 4.1% (range 0.3-8%) in recurrences. The MIB-1 Labelling Index was: 22.12% in primary tumors, 27.5% in recurrences, 31.3% in adult nonrecurrent tumors, and 4.1% in the pediatric tumor. CD34 labeling vessels/field was 9.3 in primary tumors and 9.91 in the recurrences; VEGF expression was higher in recurrences than in primary tumors (40 vs. 25%). Ultrastructural analysis showed fenestrated endothelium with hydropic changes in VEGF-positive vessels. Lack of clear correlations between morphological or immuno-staining patterns and behavior suggests that these features have no prognostic value in adult as well as in pediatric craniopharyngiomas. In this study, the only results that may be related to the aggressiveness of tumor is the major vascularization in the recurrent tumors in which the vessels show also ultrastructural changes.

Angiogenesis in patients with craniopharyngiomas: correlation with treatment and outcome

BACKGROUND

Craniopharyngiomas are histologically benign epithelial neoplasms of the sellar region that often exhibit aggressive and invasive growth. The authors hypothesized that tumor proliferation, spread, and recurrence are angiogenesis dependent and investigated the significance of vascularization relative to biologic behavior. To the authors' knowledge, angiogenesis for patients with craniopharyngiomas has not been examined to date.

METHODS

The authors measured microvessel densities in resected, histologically proven craniopharyngiomas using immunostains for CD-34, a monoclonal antibody that selectively recognizes endothelial cells. Both histologic types of craniopharyngiomas, adamantinomatous and papillary, were included in the study. In addition, the cellular distribution of vascular endothelial growth factor (VEGF), a strong stimulator of new vessel formation, was assessed by both immunohistochemistry and in situ hybridization for VEGF receptor 2 (VEGFR-2) mRNA expression.

RESULTS

Histologically, small numbers of capillaries were identified in temporal stroma but not in their epithelial components. Immunohistochemistry revealed strong, conclusive cytoplasmic immunoreactivity for VEGF in the epithelial cells of both adamantinomatous craniopharyngiomas and papillary craniopharyngiomas. In situ hybridization showed that VEGFR-2 mRNA was expressed widely, not only in neoplastic epithelium but also in capillary endothelium.

CONCLUSIONS

Tumors with greater microvessel density regrow more frequently compared with tumors that have lower microvessel density, suggesting that the extent of angiogenesis is of prognostic value in patients with craniopharyngioma. VEGFR-2 may act as a key modulator of VEGF activity in endothelial cells and nonendothelial cells, indicating that VEGF plays an important role in the behavior of craniopharyngiomas.

A correlative study of gliomas using in vivo bromodeoxyuridine labeling index and computer-aided malignancy grading

An in vivo bromodeoxyuridine (BrdU) labeling index (LI) was estimated in 43 cases of astrocytic tumors and mixed gliomas by one hour intra-operative intravenous infusion at a dose of 200 mg/m2 and correlated with (a) histological grading using a computer aided malignancy classifier TESTAST-268; and (b) histological typing using WHO classification. The lowest BrdU LI was seen in pilocytic and gemistocytic astrocytomas followed by astrocytomas, anaplastic astrocytomas and glioblastoma multiforme in that order. Mixed oligoastrocytomas followed the pattern of their astrocytic counterparts. Tumors of similar histological type showed different BrdU LI values especially amongst astrocytomas and glioblastomas. A statistically significant difference in the BrdU LI was also noted between the higher TESTAST grades of astrocytomas (T III and IV) versus the lower TESTAST grades (T II). Unlike earlier reports in literature, in the present study the category of BrdU LI of <1 contained no case of anaplastic astrocytoma or glioblastoma multiforme (TESTAST grades III and IV). Likewise, the category of BrdU LI >5 contained only anaplastic astrocytoma and glioblastoma multiforme (TESTAST grades III and IV). Maximum spread of cases was seen in the BrdU LI category of 1-5, not only in terms of histological types but also TESTAST grades. Thus there appeared to be a positive trend of increasing BrdU LI values both with histological types and increasing TESTAST grades. Further, an interesting observation was that by using a combination of TESTAST grades and BrdU LI, the histologically homogenous glioblastoma group could be further subdivided into 4 categories which showed a trend towards prognostic correlation. Thus, this study though preliminary with number of cases being small in some groups, highlights the possible usefulness of combined histological typing, TESTAST grading and in vivo BrdU LI for prognostication of gliomas especially glioblastoma multiforme.

Prediction of recurrence in histologically benign meningiomas: proliferating cell nuclear antigen and Ki-67 immunohistochemical study

BACKGROUND

Recurrence in individual patients after complete surgical removal of meningiomas cannot be predicted by histology alone because recurrence occurs even in histologically benign meningiomas.

METHODS

We investigated proliferating cell nuclear antigen (PCNA) and Ki-67 labeling indices of histologically benign meningiomas in 95 patients to assess their relationship to recurrence. The labeling index (LI) was expressed as the percentage of tumor cell nuclei immunoreactive for PCNA or Ki-67 to total tumor nuclei counted per section. The cases/specimens comprised the following two groups: (1) nonrecurrent group: 82 specimens from 82 patients without recurrence, (2) recurrent group: 28 specimens from 10 patients with recurrence.

RESULTS

Proliferative activities or aggressiveness do not always develop with every recurrence in recurrent meningiomas. The PCNA LI was significantly higher in the recurrent group (3.98% +/- 0.37%) than in the nonrecurrent group (0.71 +/- 0.13%) (p < 0.0001). The Ki-67 LI also was significantly higher in the recurrent group (3.15 +/- 0.40%) than in the nonrecurrent group (0.39 +/- 0.07%) (p < 0.0001). There was a good correlation between the PCNA LI and the Ki-67 LI (coefficient of correlation r = 0.79, p < 0.001).

CONCLUSIONS

The results of our study suggested that a PCNA or Ki-67 LI of more than 2% may represent an increased risk for recurrence; therefore, we suggest that radiotherapy or stereotactic radiosurgery should be considered, even for histologically benign meningiomas.

Triple approach for diagnosis and grading of meningiomas: histology, morphometry of Ki-67/Feulgen stainings, and cytogenetics

With regard to meningioma grading and the recently introduced "atypical" meningioma, we evaluated 160 cases retrospectively by conventional histology and image analysis. For that, the cell nuclei were stained with a Ki-67 (MIB1)/Feulgen-method on paraffin sections, thus enabling the assessment of both the Ki-67 proliferation index and nuclear morphometric features, such as tumour cell arrangement, nuclear pleomorphism, and cellularity. It could be demonstrated that the Ki-67 proliferation index is the most important criterion for distinguishing anaplastic meningiomas (WHO "grade" III) (mean Ki-67 index: 11%) from those of common type (WHO "grade" I) (mean Ki-67 index: 0.7%). The parameter for the "relative volume weighted mean nuclear volume" is another valuable morphometric feature. The "atypical" meningioma (WHO "grade" II) which should represent an intermediate category between common type and anaplastic meningiomas is characterized by a mean Ki-67 proliferation index of 2.1%. Common type meningiomas which comprise almost 50% of the cases of this series have a relapse rate of 9%. "Atypical" and anaplastic meningiomas recurred in 29% and 50%, respectively. Since the term "atypical" meningioma is confusing in the context of tumour grading, the term "intermediate type meningioma" is proposed. Furthermore, the results of cytogenetic analyses of 142 cases of this series were evaluated and compared with the meningioma grades. Thereby, 25 cases disclosed, independent of the typical loss of one chromosome 22, cytogenetic features assumed to be progression-associated, e.g., the gain or loss of different chromosomes and the deletion of the short arm of one chromosome 1 (hyperdiploidy, increased hypodiploidy, 1p-), when correlated to the histological and morphometric findings or the high relapse rate. For meningioma diagnosis and grading, a practical guideline is proposed based upon histology, morphometry (Ki-67), and cytogenetics.

Aggressive meningeal tumors: review of a series

A series of 25 patients with aggressive meningeal tumors was studied to determine the efficacy of various management options. The median age of the patients was 52 years, with a range of 13 to 73 years. A marked male preponderance (64%) was noted. Twenty of 25 patients experienced recurrence during a median follow-up time of 47 months. Survival and freedom from recurrence varied with histological diagnosis. Recurrence was noted sooner in patients who had received partial resections on first presentation of tumor than in those who had received total resections at first presentation. Survival time was also shorter for patients who underwent partial resections at first presentation than for patients who underwent total resections. Patients' prognoses did not improve as a result of either chemotherapy or radiotherapy. Of six patients with extracranial metastases, the median time to metastasis was 102 months, with a 5-year metastasis-free rate of 85%. The most common sites of metastasis in these six patients were lung and bone. In each tumor type, histological features used in diagnosis and radiological features studied from computerized tomography and magnetic resonance imaging were evaluated, compared, and discussed. Of eight patients studied with an in vivo bromodeoxyuridine (BUdR) labeling index (LI), seven showed an LI of 1% or more. The authors support the incorporation of the BUdR LI into the diagnostic process to provide a better estimate of the potential for tumor recurrence.