Clinical significance of CD133 and Nestin in astrocytic tumor: The correlation with pathological grade and survival

Abstract Background We aimed to investigate the interaction between CD133 and Nestin and further assessed the correlation of CD133 and Nestin with clinicopathological characteristics and survival in patients with astrocytic tumor. Methods Totally 127 patients with astrocytic tumor underwent surgical resection were enrolled. Patients’ age, gender, and World Health Organization (WHO) grade were recorded, and the survival data were extracted from the follow‐up records. The expressions of CD133 and Nestin in astrocytic tumor tissues were analyzed by immunohistochemistry assay. The WHO grade I and II astrocytic tumors were defined as low‐grade astrocytic tumors (LGA), the WHO grade III and IV astrocytic tumors were defined as high‐grade astrocytic tumors (HGA). Results There were 79 (62.2%), 34 (26.8%), 14 (11.0%), and 0 (0.0%) patients with CD133 negative, low, moderate, and high expression, respectively; 7 (5.5%), 47 (37.0%), 20 (15.7%), 53 (41.7%) patients with Nestin negative, low, moderate, high expression, respectively. CD133 and Nestin were both correlated with advanced WHO grade but not with age or gender, and positive correlation was observed between CD133 and Nestin. For survival, both CD133 and Nestin were correlated with unfavorable overall survival (OS), and further analysis illustrated that Nestin but not CD133 independently predicted poor OS. Subgroup analysis also revealed that Nestin but not CD133 negatively associated with shorter OS in LGA patients, while both CD133 and Nestin were correlated with poor OS in HGA patients. Conclusion CD133 and Nestin present as potential biomarkers for advanced pathological grade and poor survival in patients with astrocytic tumor.


| INTRODUC TI ON
Astrocytic tumor is the most common glial tumor of the central nervous system, accounting for 13%-26% of intracranial tumors. 1 It is characterized by fast and invasive growth as well as no obvious boundary with normal brain tissue, which leads to short course of the disease and poor prognosis. 2 Currently, the treatment for astrocytic tumor relies on surgery, radiotherapy and chemotherapy. 1 Surgical resection is theoretically a cure for astrocytic tumor; however, due to the infiltration and frequent recurrence of the tumor, surgical resection is incomplete and ineffective in a proportion of cases, or even hard to be carried out when tumor grows in certain important areas such as brain stem. 3 Radiotherapy is the common therapy for almost all types of glial tumors including astrocytic tumor. However, only a few subtypes present high sensitivity to radiotherapy and the effect of radiation-induced necrosis on brain function should not be underestimated. [4][5][6] In addition, chemotherapy is limited to blood-brain barrier and side effects. Therefore, it is of clinical significance to explore potential biomarkers for disease monitoring and prognosis prediction, which would assist with treatment of astrocytic tumor.
In recent years, researchers discover the existence of cancer stem cells that have self-renewal and immense differentiation ability in astrocytic tumor. These cells are insensible and highly resistant to current radiotherapy and chemotherapy and contribute to the tumor metastasis and recurrence. 7,8 Identification of cancer stem cells has been realized via a number of markers, and undoubtedly, the cancer stem cell markers provide a new direction for the treatment of astrocytic tumor and the prevention of recurrence. 9 CD133, a brain tumor stem cell marker, is shown to be upregulated in recurrent pilocytic astrocytoma, and suppression of CD133 increases cell chemosensitivity via inhibiting PI3K-Akt-NF-κB signaling mediators and MDR1. 10 In addition, the intermediate filament protein Nestin, which is critical in central nervous system development, is correlated with higher WHO grade and poor prognosis in patients with astrocytic tumor. 11,12 However, current studies investigating the correlation of CD133 and Nestin with clinicopathological features and prognosis are mostly based on general glioma patients or astrocytic tumor patients in a specific pathological stage. Therefore, in this study, we evaluated the interaction between CD133 and Nestin and further assessed the correlation of CD133 and Nestin with clinicopathological characteristics and survival in patients with astrocytic tumor.

| Patients and tumor specimens
A total of 127 patients with astrocytic tumor who underwent surgical resection at our hospital between January 2005 and December 2011 were reviewed and analyzed in this study. All reviewed patients had (a) newly diagnosed as primary astrocytic tumor by pathological examination, (b) sufficient available tumor tissue for immunohistochemical analysis, (c) detailed clinical and follow-up data, and (d) without neoadjuvant therapy before surgery. All 127 human astrocytic tumor specimens were removed from surgical resection, confirmed by the pathology, fixed in the 10% neutral formaldehyde and embedded in paraffin wax. This study was approved by the Institutional Review Board of Shenzhen Nanshan People's Hospital. Written informed consents or verbal agreements with tape recording were collected from reviewed patients or their guardians.

| Data collection
The clinical data of studied cases were collected from the medical records, including age, gender and World Health Organization (WHO) grade. The assessment of WHO grade was based on the 2000 WHO classification of nervous system tumors, 13 and the WHO grade I and II astrocytic tumors were defined as low-grade astrocytic tumors (LGA), the WHO grade III and IV astrocytic tumors were defined as high-grade astrocytic tumors (HGA). The postoperative treatments including chemotherapy, radiotherapy (50-60Gy/5W), and immunotherapy were administered to patients based on the clinical status. All patients were followed up by telephone or clinic visit, with a median follow-up of 24.0 months (ranging from 2 to 113 months), and the survival data were extracted from the follow-up records, which were used to calculate the overall survival (OS).

| Immunohistochemistry (IHC) assay
The expression of CD133 and Nestin in astrocytic tumor tissue was analyzed by IHC staining, which was performed using the Rabbit anti-human CD133 monoclonal antibody (Cell Signaling Technology, Inc) and the Mouse anti-human Nestin monoclonal antibody (Santa Cruz Biotechnology, Inc) served as primary antibody, respectively.
IHC assay was carried out as follows: (a) formalin-fixed paraffin-embedded astrocytic tumor tissues were cut into 4 µm sections, then were baked in oven at 60℃ for 30 minutes; (b) deparaffinizing in xylene and rehydration in a graded series of ethanol were preformed; (c) for antigen retrieval, the sections were placed in 0.01 mol/L citrate buffer (pH 6.0) and boiled for 20 minutes; (d) to block endogenous peroxidase, each section was incubated at room temperature with 3% H 2 O 2 for 10 minutes; (e) for immunoreaction, the section was incubated with primary antibody (Rabbit anti-human CD133 monoclonal antibody or Mouse anti-human Nestin monoclonal antibody) at 37℃ for 2 hours (negative control was incubated with phosphate buffer saline (PBS) instead of primary antibody); (f) after incubated with the primary antibody, the section was incubated with horseradish peroxidase-conjugated goat-anti-rabbit/goat-antimouse immunoglobulin G antibody (Santa Cruz Biotechnology, Inc); (g) newly prepared diaminobenzidine (DAB) solution were added to each section for visualization; (h) the section was washed using distilled water, followed by the counterstaining with hematoxylin for 1 minute; and (i) after routine dehydration, transparency, drying and sealing, the section was observed on the microscope.

| Statistical analysis
Data were presented as count (percentage). Correlation analysis was performed using the Chi-square test or Spearman's rank correlation test. OS was calculated from primary surgery until death or date of censoring, and the OS was illustrated using the Kaplan-Meier curves and compared by the log-rank test. Factors predicting OS were assessed by the use of univariable and multivariable Cox's proportional hazard regression model analyses. A P value < .05 was considered significant. All data analysis was performed using the SPSS 22.0 statistical software (IBM), and the graph was made using the GraphPad Prism 7.02 software (GraphPad Software Inc).

| Correlation of CD133 with Nestin
In order to determine the interaction between CD133 and Nestin, Spearman's rank correlation test was performed, which observed that CD133 expression was positively correlated with Nestin expression in astrocytic tumor tissue (P < .001, r = .299) ( Table 2).

| Correlation of CD133 and Nestin with OS
Patients with CD133 moderate expression presented with the lowest OS, followed by patients with CD133 low expression, and patients with CD133 negative, further analysis illustrated that CD133 was negatively correlated with OS (P < .001) (Figure 2A) with OS (P < .001) ( Figure 2B).

| Correlation of CD133 and Nestin with OS in LGA patients and HGA patients
The correlation of CD133 and Nestin with OS were then evaluated separately in LGA patients and HGA patients. In LGA patients, CD133 was not correlated with OS (P = .775) ( Figure 3A), whereas Nestin was negatively correlated with OS (P = .004) ( Figure 3B).

| Factors affecting OS in LGA patients and HGA patients
Univariate Cox's regression showed that higher Nestin expres-

| D ISCUSS I ON
There are a small population of highly tumorigenic cancer stem cells existed in a variety of malignancies including astrocytic tumor. 7 The presence of cancer stem cells drives the invasiveness of tumor and resistance to therapies, which is crucial for tumor metastasis and relapse. 8 The isolation of cancer stem cells is achieved via specific stem cell-related surface antigens, and the abundance of these cancer stem cell markers are shown to be closely related with cancer development and progression. 9 For instance, the transmembrane glycoprotein CD133 has been used to identify cancer stem cells in different solid tumors including brain, lung, gastric, and liver cancers. 16 In brain tumors, CD133 is first used to identify cancer stem cells in pediatric samples of glioma and medulloblastoma, and CD133 positive tumor cells are more aggressive and with higher capacity to self-renew. 9,17 And in astrocytic tumor, CD133 expression is correlated with higher histological grade and larger tumor size. 18   Abbreviations: CI, confidence; HGA, high-grade astrocytic tumor (WHO grade III ~ IV); HR, hazard ratio; OS, overall survival; WHO, World Health Organization. a CD133 and Nestin expression were included in the Cox's regression in the form of ordered categorical variable, which were encoded as: negative (0%) = 0, low (0%-30%) = 1, moderate (30%-60%) = 2, high (≥60%) = 3.
needed to further validate our findings. Additionally, due to relatively early patient enrollment time, WHO classification 2000 was used for tumor evaluation, which was a bit old compared with the newly launched WHO classification 2016. However, due to lack of cytogenic and molecular genetics information, re-evaluation of the tumors using the 2016 classification was not feasible. And the correlation of IDH mutation with CD133 and Nestin as well as survival was not assessed.
In conclusion, CD133 and Nestin are correlated with advanced WHO grade and poor survival in patients with astrocytic tumor, which benefits the exploration of prognostic factor for astrocytic tumor.

This study supported by Knowledge Innovation Basic Research
Project of Shenzhen Science and Technology Innovation Committee (JCYJ20140416094330210).

CO N FLI C T S O F I NTE R E S T
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.