This study showed that a higher percentage of CD133+ cells in cancer nests was strongly associated with the lower 5-year survival rate in colon cancer patients with stage IIIB, a locally advanced disease among which most of patients would die from metastasis in spite of adjuvant chemotherapy, implying that the overpopulation hypothesis of cancer stem cell seems reasonable as CD133 is a putative marker of colon cancer stem cells.
The evidence concerning the correlation of the percentage of CD133+ tumor cells with the prognosis of patients was scarce as a few of observations were reported [43–46]. Recently the relationship between CD133 expression and prognosis in colorectal carcinomas was examined. Horst reported that CD133 expression is an independently prognostic marker whereas this kind of correlation was not observed by Kojima. [40, 41] The discrepancy might derived from inadequate patient quantity and the mixed tumor stage. For example, in Kojima's study, a total of 189 patients consisted of 106 cases of colon cancers and 83 cases of rectal cancers with TNM stages varying from I to VI, that is, one group of patients with a definite stage contained only 20 or 30 cases of colon or rectal cancer patients, respectively. Similar situation existed in Horst's study . To narrow the heterogeneity of patients and make the results more reproducible this study included 104 cases of colon carcinoma patients with stage IIIB. The results showed that CD133+ cancer cells contributed to the progression of colon cancer, arguing the Hosrt's observation.
The discrepancy concerning the pattern and the frequency of CD133 expression in colon cancer also existed between the studies mentioned above and this study. Horst and Kojima reported that CD133 antigen, stained with antibodies from Miltenyi Biotech, Sata Cruz Biotechnology, or Cell signaling, was localized exclusively on the glandular-luminal surface of colorectal cancer. Staining of the CD133 was observed neither on the budding cancer nest nor on poorly differentiated cancer cells [40, 41]. However, in this study, being stained with antibodies from Abcam CD133 expression existed not only on the apical membrane but also on basal surface of tumor cells, both on the budding cancer nest (the invasive front) and on the poorly differentiated cancer cells, although the intensity of staining was weaker. This pattern of CD133 expression might be more likely consistent with the hypothesis that CD133+ cancer cells would reveal a more aggressive phenotype. Since the intensity of CD133 is cell cycle-dependent, among which the least CD133 immunoreactive cells are in the G0/G1 portion, and the increased CD133+ cells is correlated with increased DNA content, and cancer cells is relatively arrested in the invasive front, so, attenuated expression of CD133 occurred in the invasive front (budding)[47, 48]. As for the frequency of CD133+ cells in colorectal cancers the discrepancy also existed. In Kojima's study CD133 expression was detected in only 29 of the 189 tumors (15.3%). Of these, 21 tumors (11.1%) showed CD133 over-expression among which CD133 positive area occupied more than 10% of the entire tumor tissue. Otherwise, in Horst's study tumors with more than 50% of CD133+ tumor cells exist in 20 out of 79 colorectal cancers (25.3%) . In this study, the percentage of CD133+ cells varying from 5% to 25% existed in 23 cases (22.1%), from 26% to 50% in 12 cases (11.5%), and more than 50% in 7 cases (6.7%). Therefore, it is reasonable to infer that the heterogeneous patterns and frequencies of CD133 expression in colon cancer derived from the specificity of antibody clones used. In the future, more attention should be paid to the specificity of CD133-targeting antibodies, the standardization of the CD133 positive cells classification system, and homogeneity of tissues.
Recently the representative of CD133 as marker of colon cancer stem cells was questioned. On the one hand, CD133+ colon cancer cells revealed 'stem-like' characteristics, and stem cells marked by CD133 was susceptible to transformation into tumors. On the other hand, CD133 expression was detected not only on cancer cells, but also on the luminal layer of epithelium of digestion duct, on the mature epithelium of the pancreatic duct, on the proximal tubules of the kidney, and on the lactiferous ducts of the mammary gland [50–52]. Furthermore, both CD133+ and CD133- metastatic colon cancer cells initiated tumors. Additionally, CD44+ cancer cells rather than CD133+ cells have an increased tumorigenicity. Those data pointed that CD133 should not be a unique marker for colon cancer stem cells. It is less likely that a known marker for colon cancer stem cells, such as CD44, CD166, EpCAM, and Lgr5, has the potential just like Pten-related pathway in leukemia, which could distinguish hematopoietic stem cells from leukemia-initiating cells [54–57]. Collectively, a combination of cell surface markers is need for the definition of colon cancer stem cells [58–60]. This study implied that, given that CD133 may not represent all the entire cancer stem cells, it is still a useful biomarker as CD133+ cells is more aggressive than CD133- partners in colon cancer.