Stem Cell Marker CD133 Expression Predicts Outcome in Glioma Patients

Abstract

Evidence strongly indicates that a small fraction of cells within a tumor known as cancer stem cells (CSCs) drive tumorigenesis in solid neoplasms, including gliomas. CSCs are hypothesized not only to initiate new tumors but also to mediate recurrence by resistance to standard anticancer therapies. The CSC population has been enriched from brain and other tumors using the stem cell surface antigen CD133. CD133+ glioma cells are highly tumorigenic in animal models. For the first time, researchers at the German Cancer Research Center in Heidelberg, Germany recently showed that CD133 expression in primary tumor specimens correlates with survival and outcome measures in patients with gliomas (Clin Cancer Res 14:123–129, 2008). This group analyzed CD133 expression by immunohistochemistry in a series of surgical pathological specimens from 95 gliomas of various grades. CD133 expression and the topographical arrangement of CD133+ cells into clusters significantly increased with higher tumor grade. In 18 of 24 Grade II specimens, CD133 was either not detected or, when detected, was present in less than 1% of tumor cells. In contrast, CD133 was present in greater than 5% of tumor cells in 45 of 47 glioblastomas multiforme and present in 10 to 50% of these cells in 21 of those tumors. A high percentage of CD133+ tumor cells and the presence of clusters correlated with decreased progression-free and overall survival, independent of tumor grade, extent of surgical resection, or age. Moreover, tumor recurrence and progression from lower to higher grade were predicted by an increasing proportion of CD133+ cells within the initial specimen. CD133 was colocalized in a subset of nestin+ cells, indicating that CD133 expression may identify an increasingly tumorigenic subpopulation of CSCs. Further, CD133 was coexpressed with epidermal growth factor receptor vIII mutant, when present, proving that CD133+ cells derived from the tumor. This is one of a handful of recently published reports suggesting that a subset of cells within human gliomas express known neural stem cell markers (CD133, Sox2, Musashi-1, and others) and that such expression correlates with clinical outcomes. The CSC hypothesis dictates that targeting and eradicating these tumorigenic cells may significantly improve glioma treatment. Homing in on CSCs by these antigens may be confounded by their expression in normal neural stem cells, but work such as this may ultimately lead to better treatments and outcomes for patients.FIGURE: Differential expression of CD133 in gliomas of various grades correlates with clinical outcome. A, CD133 was not detectable in some World Health Organization (WHO) Grade II gliomas, whereas expression was detected in relatively few cells in other tumors of the same grade (B, arrows). C, WHO Grade III anaplastic astrocytoma with an increased percentage of CD133+ cells. D, WHO Grade IV glioblastoma multiforme with CD133+ cells arranged in clusters (arrowheads). E, Recurrence and malignant progression correlated with increased CD133 expression in WHO Grade II and III gliomas. F, Kaplan-Meier curves depict progression to glioblastoma multiforme and overall survival (G) in relation to higher CD133 expression in WHO Grade II gliomas. From Clin Cancer Res (14:123–129, 2008).BRIAN M. HOWARD, MD JOHN A. BOOCKVAR, MD STEM CELL RESEARCH