Analysis of neural stem cell immunophenotypic composition with tissue microarray method and high-risk groups for recurrence in pilocytic astrocytomas

Aim and hypothesis: There is growing evidence for the role of stem cells in some primarycentral nervous system tumors. In order to determine the presence and the prognosticsignificance of stem cell markers in pilocytic astrocytoma, we have analyzed the expressionof a set of antibodies including CD133, CD34, p75NTR, Sox-2, in a large series of pilocyticastrocytomas. In addition, we have analyzed the expression of glial (olig-2, GFAP, vimentin)and neuronal (neurofilament, synaptophysin, and NeuN)) markers to determine a relationshipbetween the above groups of markers. We have also analyzed some of the tumor-associatedantigens, such as EGFR, p53, bcl-2, and Ki-67 in this group of tumors. The results were thencorrelated with clinical findings in order to determine an association between the markerstatus of tumors and their clinical characteristics in order to determine biologically importantfactors.Method: The tissue microarray blocks and slides were prepared from 62 pilocyticastrocytoma cases. Fetal and adult normal cerebral and cerebellar brain sections from autopsycases were used as controls. All microarrays were stained with antibodies against CD133,CD34, p75NTR, bcl-2, EGFR, Neurofilament, GFAP, vimentin, synaptophysin, Sox-2, Olig-2, p53, Ki-67 and NeuN using standard immunohistochemical techniques. Fifteen randomcases were also analyzed for 1p/19q deletion by flourescent in situ hybridization.Results: There were 62 patients with a median age of 84 months (range 14-193 months).Femal/male ratio was equal. The tumor localizations were as follows: posterior fossa, 68%(n=42); supratentorial, 29% (n=18); spinal cord, 3% (n=2). Gross total resection wasperformed in 32 cases (52%) and the remainder of the cases underwent subtotal resection.There were nine recurrences (15%) with a mean recurrence free period of 12 months. Allcases strongly expressed vimentin and GFAP was diffusely positive in all but four cases inwhich the staining was focal. There was synaptophysin positivity in 25, neurofilamentpositivity in 12 and NeuN positivity in two cases. No tumor was positive for EGFR, p53,p75NTR and CD34 protein. Bcl-2 was found positive in six cases (16%). The mean Ki-67labeling index for all tumors was 2% (range 0-8%). CD133 was detected in only one case.The mean percentage of Sox-2 positive cells was 45% (range 0-90%) and Olig-2 positive cellswas 35% (range 0-90%). There were no deletions on 1p/19q analysis by flourescent in situhybridization.In univariate/bivariate analysis, recurrence negatively correlated with extent ofsurgical resection and age at initial diagnosis, positively correlated with Ki-67 labeling index.Patients younger than 5 years of age had shorter recurrence free period even among patientswith subtotal resection. Tumors with a Ki-67 labeling index higher than 2% also had a shorterrecurrence free period. There was no statistically significant correlation between theexpression of immunohistochemical markers and recurrence free survival.Conclusion: The positive staining with neuronal markers highlight the challenges oftenexperienced in practical surgical pathology in the diagnosis of pilocytic astrocytomas. Whilewe believe most such staining were either due to incorporated normal brain tissue,immunopositivity of neuronal markers in pilocytic astrocytomas requires further scrutiny. Thepresence of a non-specific stem cell marker; sox-2 in pilocytic astrocytoma also suggests thatsome tumor cells may have stem-cell like properties, or may have been derived from sox-2positive stem cells. Olig-2 positivity along with GFAP and vimentin simply implies a strongglial differentiation, and absence of a so-called oligodendroglial differentiation, especially intumors that resemble oligodendroglioma, is further supported by the absence of 1p/19qdeletion. p53, EGFR, p75NTR, CD133 antigens don?t seem to have a role in the biologicalbehaviour of pilocytic astrocytomas. Subtotal or incomplete surgical resection still seems tobe a major risk for recurrence, so are age less than 5 years at diagnosis and Ki67 labelingindex greater than 2%. Other immunohistochemical markers tested in this study were notassociated with recurrence-free or overall survival. Tissue microarray method providedbenefits in this study; enabled the use of a large immunohistochemical panel and in situhybridization, and can be recommended for similar studies.Key words: Pilocytic astrocytoma, neural stem cell, recurrence, tissue microarray method