Title: Preferential Expression of Chemokine Receptor CXCR4 by Highly Malignant Human Gliomas and Its Association with Poor Patient Survival
Abstract: In Reply: We thank Drs. Ehtesham, Stevenson, and Thompson for their very helpful comments on our article (4). In this correspondence, we would like to provide additional analyses and clarification of the results concerning the correlation between the levels of CXCR4 expression and the degree of malignancy of human astrocytic gliomas. In our study, owing to the relatively small numbers of World Health Organization (WHO) Grade III and IV glioma patients, we grouped them together for analysis of survival time. Likewise, CXCR4-positive (with protein staining + and ++) tumors were included in a single group. WHO Grade III is referred to as anaplastic astrocytoma, and WHO Grade IV is referred to as glioblastoma (including the previously termed glioblastoma multiforme [GBM]). Both Grade III and IV tumors are highly malignant, and anaplastic astrocytomas usually recur after surgery and transform to glioblastomas. Our previous studies showed that there was no significant difference in the survival rate of patients with either Grade III or IV gliomas after aggressive treatment (1). When we analyzed Grade III and IV patients separately, as shown in Table C1, there indeed was no significant difference in their 3-year survival. However, both Grade III and IV tumor patients showed a significantly reduced rate of survival as compared with WHO Grade II tumor patients (P < 0.01). By further analyzing the clinical parameters in detail, we found that whereas tumor size exceeding 3 cm in diameter and higher WHO grade significantly reduced the rate of patient survival, age and sex had minimal influence. In addition, the postoperative survival time of the patients was found to be closely correlated with the levels of CXCR4 expression in the tumors (Table C2; Fig. C3). Also, there is a correlation between tumor size and the levels of CXCR4 expression. For instance, for CXCR4-positive (including + and ++) tumors, their mean size (maximal length) was 3.20 ± 1.08 cm, as compared with 2.50 ± 0.76 cm for CXCR4-negtive tumors.FIGURE C3: Kaplan-Meier survival curves of patients included in this study. A, comparison of survival rates of patients with CXCR4-negative tumors and patients with CXCR4-positive (+ and ++) tumors as a single group (P < 0.01). B, comparison of survival curves of patients with CXCR4-negative tumors and those with CXCR4-positive (either + or ++) tumors, respectively (P < 0.01).TABLE C2: Correlation between CXCR4 expression and pathological parameters in patients with gliomasaTABLE C1: Survival time of patients with different World Health Organization grade gliomasaThe postoperative survival time of patients with tumors expressing different levels of CXCR4 was analyzed with Kaplan-Meier and log rank tests. There were significant differences between patients with CXCR4-negative tumors and CXCR4-positive tumors, regardless of analysis with CXCR4 + and ++ as separate groups or in combination (Fig. C3). These results were consistent with the recent report by Ehtesham et al. (5), in which the CXCR4/CXCL12 axis correlated with the invasiveness of gliomas. Thus, accumulating evidence supports the conclusion that CXCR4 plays an important role in the progression of gliomas and its expression level can be considered as one of the prognostic markers for patients after surgical treatment. All patients with WHO Grade III and IV gliomas analyzed in our study received similar chemotherapy and topical irradiation after surgical removal of the primary tumors. The treatment included three to six courses of chemotherapy with oral administration of lomustine and one single dose of radiotherapy (20 Gy). Owing to the limited number of patients, it is premature to analyze the results with more detailed grouping. Such analysis is certainly important and should be performed when more cases of gliomas are available. As for the measurement of CXCR4, because of its heterogeneous expression in a single tumor or in cell lines, immunocytochemistry would better reveal the location of CXCR4 protein in tumor cells in parallel with markers of cell differentiation (2). The CXCR4 images could then be quantitatively analyzed in a fairly precise manner by using computer software. Our quantitation of CXCR4 was always performed in the presence of positive and negative controls and was very reproducible. We therefore are confident of the results presented in our article. However, we agree that, for freshly removed tumor specimens, real-time polymerase chain reaction and Western blotting to measure the CXCR4 gene and protein should also yield valuable data, which will be considered in our future studies with newly recruited patients. We also would like to take this opportunity to present the correct structures of natural nordihydroguaiaretic acid (meso-NDGA) and Nordy (dl-NDGA), which are capable of inducing glioma cell differentiation (Fig. C4) (3).FIGURE C4: Structure of natural nordihydroguaiaretic acid (meso-NDGA) (A) and Nordy (dl-NDGA) (B).In summary, we hope that our further analyses of our results and clarification of several issues raised by Ehtesham et al. will support the notion for CXCR4 as an indicator for the degree of the malignancy of gliomas and a predictor of the prognosis of patients after combined therapy. Xiu-wu Bian Hua-liang Xiao Shi-xin Yang Chongqing, China Ji Ming Wang Frederick, Maryland Disclosures This study was supported by grants from the National Natural Science Foundation of China (NSFC No. 30370552) and the International Cancer Technology Transfer Award from the International Union Against Cancer, Geneva, Switzerland (ICRETT No. 661/2002).