Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Prognostic Value of CD133 and SOX2 in Advanced Cancer

Prognostic Value of CD133 and SOX2 in Advanced Cancer Hindawi Journal of Oncology Volume 2019, Article ID 3905817, 12 pages https://doi.org/10.1155/2019/3905817 Research Article 1 2 1 1 1 1 Susu Han , Tao Huang , Xing Wu, Xiyu Wang, Shanshan Liu, Wei Yang, 1 1 1 Qi Shi, Hongjia Li, and Fenggang Hou Shanghai Municipal Hospital of Traditional Chinese Medicine, Shanghai University of Traditional Chinese Medicine, Zhijiang Road, Shanghai , China e Affiliated Hospital of Ningbo University, Ningbo, Zhejiang , China Correspondence should be addressed to Susu Han; anyasue@163.com, Tao Huang; huangtao334@163.com, and Fenggang Hou; fghou555@126.com Received 27 August 2018; Revised 24 October 2018; Accepted 26 November 2018; Published 1 January 2019 Academic Editor: Akira Hara Copyright © 2019 Susu Han et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Background. eTh prognostic value of CD133 and SOX2 expression in advanced cancer remains unclear. This study was first conducted to investigate the association between CD133 or SOX2 positivity and clinical outcomes for advanced cancer patients. Methods. Hazard ratios (HRs) with 95% confidence intervals (95% CIs) were calculated to evaluate the correlation between CD133 or SOX2 positivity and overall survival (OS), disease-free survival (DFS), progression-free survival (PFS), cancer-specific survival (CSS), or recurrence-free survival (RFS) from multivariable analysis. Trial sequential analysis (TSA) was also performed. Results.13 studies with 1358 cases (CD133) and vfi e studies with 433 cases (SOX2) were identified. CD133 positivity was correlated with worse CSS and OS, but there was no correlation between CD133 positivity and DFS. SOX2 positivity was associated with poor DFS and RFS but was not linked to PFS. Stratified analysis by study source showed that only CD133 positivity can decrease OS for Chinese patients. Stratified analysis by treatment regimens indicated that CD133 positivity was linked to poor OS in patients treated with adjuvant therapy. TSA showed that additional studies were necessary. Conclusions. CD133 and SOX2 might be associated with worse prognosis in advanced cancer. More prospective studies are strongly needed. Impact. CD133 and SOX2 may be promising targeted molecular therapy for advanced cancer patients. 1. Introduction proliferation and differentiation, and resistance to conven- tional treatments like chemotherapy or radiation [6, 7]. Recently, some stem cell markers have been described, Cancer is still one of the most threatening diseases worldwide [1]. Although surgery, chemotherapy, and/or radiotherapy suchas CD44,CD166,EpCAM,CD133,and SOX2 [8–10]. have greatly improved the clinical survival for early cancer CD133, also named as prominin-1, is a member of pentas- patients, therapies for patients with advanced or metastatic pan transmembrane cell surface glycoproteins [11, 12]. Sex- cancer still have a major challenge [2]. Improvements in determining region Y-box protein 2 (SOX2), a High Mobility the treatment of advanced or metastatic cancer patients Group (HMG) domain transcription factor, is involved in (surgical technique, chemotherapy, radiotherapy, targeted the regulation of stem cells self-renewal and pluripotency molecular therapy, and immunotherapy regimens) have [13]. CD133 expression has been reported and contributes to extended patients’ median survival, but such as 5-year overall malignant transformation and chemo- and radioresistance survival is still poor [3–5]. us Th the development of new and [14].SOX2hasbeen studiedinsometypes of humancancers novel therapeutic regimens for advanced or metastatic cancer and facilitates tumor initiation and progression [15–17]. Some meta-analyses investigated the prognostic value of CD133 and patients is important. Increasing evidence has been suggested regarding cancer SOX2 expression in some human cancers [18–21], but the stem cells (CSCs) in various cancers. The major character- prognostic significance of CD133 and SOX2 expression in istics of CSCs are the capability of self-renewal, unlimited advanced cancer patients remains unclear and unknown. 2 Journal of Oncology Toourknowledge,the expressionofCD133 andSOX2is studies using univariable analysis, done in the present meta- hitherto undescribed in advanced cancer by a meta-analysis. analysis. REMARK scores can be used and evaluated for To clarify the correlation between the expression of stem cell sensitivity analyses. eTh following information was extracted markers (CD133 and SOX2) and the prognosis in advanced from eligible studies: rfi st author’s name, publication year, or metastatic cancer patients, we investigated the relationship study population, study source, mean or median age, type between the expression of these two markers and survival of of cancer, detection method, therapy regime, study design, the samples. sample type, cut-off value, median or mean follow-up period, survival rate, adjusted variables, and clinical outcomes, etc. All authors resolved the discrepancy when information was 2. Materials and Methods controversial. .. Literature Selection. The present meta-analysis was .. Data Analysis. To estimate the effect of CD133 or SOX2 reported in accordance with the Preferred Reporting Items expression status on advanced cancer survival (OS, DFS, PFS, for Systematic reviews and Meta-Analyses (PRISMA) guide- CSS,RFS,orMFSofmultivariableanalysis),theresultwithan line [22]. eTh potential studies were identified through HR>1 demonstrated an unfavorable prognosis, whereas an searching online databases including PubMed, EMBASE, HR<1 stood for a good prognosis. The Cochran’s Q statistic EBSCO, Web of Science, and Cochrane Library before April was used to evaluate heterogeneity among the included stud- 2018 without language restrictions. eTh main key words ies [26]. The random-effects model (DerSimonian-Laird) was and search items were “CD133 OR PROM1 OR prominin- used in the meta-analysis (heterogeneity: P< 0.1) [27, 28]. For 1 OR AC133 antigen OR SOX2 OR Sex determine region the results (> seven studies) with substantial heterogeneity, Y-box 2 OR SRY box-2 OR SRY-Related HMG-Box Gene subgroup analyses based on tumor type, study source, sur- 2”, “metastatic OR advanced OR metastasized OR recur- vival rate, sample type, age (years), testing method, and study rent”, “cancer OR tumor OR carcinoma OR neoplasm”, and center design were performed to explain the potential het- “survival OR outcome OR prognosis”. Additional potential erogeneity and different strength of the association between articles were also manually searched by the reference lists of subgroups. If all relevant P values of heterogeneity were theeligiblestudies. greaterthan0.1 amongdieff rentsubgroups,itindicates the source of heterogeneity from a subgroup variable. eTh Egger’s and Begg’s funnel plots were used to evaluate publication .. Eligibility Criteria. Papers identiefi dfor theinclusion bias [29, 30]. Pooled data were analyzed using Stata software, criteria in this study for the current analysis were as follows: version 12.0 (Stata Corp., College Station, TX, USA). (1) studies reported the patients with advanced, metastatic, or recurrent cancer; (2) studies investigated the prognostic value of expression of CD133 or SOX2; (3) studies presented . Trial Sequential Analysis. In the meta-analysis involving sufficient data on hazard ratio (HR) with 95% confidence a small number of participants, random errors can lead interval (CI) from multivariable analysis for overall survival to spurious results [31, 32]. Trial sequential analysis (TSA) (OS), disease-free survival (DFS), progression-free survival was conducted to control random errors and to estimate (PFS), cancer-specific survival (CSS), relapse/recurrence-free therequiredstudy population [33].Theoptimal apriori survival (RFS), or metastasis-free survival (MFS); (4) unclear anticipated information size (APIS) method was set in our data (HR with 95% CI) such as only P value with HR or study. We calculated diversity-adjusted TSA based on the 95% CI, survival data calculated based on the described relative risk reduction (RRR) of 20%, the prespecified type method [23, 24], or contacting the corresponding author via I error of 5%, and the type II error (20% or 10%). We also email to request the available information. If two or more calculated diversity-adjusted TSA based on a RRR of 15%, papers used the overlapping or same cancer samples, only the prespecified type I error ( 𝛼) of 5%, and a type II error the study with the largest patient numbers or the most recent (𝛽) of 20%. Monitoring boundaries are applied to decide article was selected. Case report, reviews, animal studies, whether a clinical trial could be terminated early. When unrelated articles, or survival data using univariable analysis the cumulative Z curve was more than the trial sequential were excluded. monitoring boundary or required information size (RIS) boundary, it suggested the firm evidence. Otherwise, more clinical studies are needed. Meta-analysis of HR estimates . . Data Extraction and Study Assessment. The methodology was performed using Stata software, version 12.0 (Stata Corp., of each eligible study was conducted following REMARK College Station, TX, USA) and R software, version 3.4.2 (eTh guidelines (Reporting Recommendations for Tumor Marker R Foundation for Statistical Computing, Vienna, Austria). Prognostic Studies) [25]. 20 criteria were listed in REMARK; each itemhadscores0,1,and 2, withamaximal score of 40 (Table S1). The value was 2 scores when each item 3. Results was clearly described in the article, 1 score when each item was incompletely defined, and 0 score when each item .. Study Characteristics. Flowchart describing the study was not defined or not applicable. We did not define a selectionprocessisshowninFigure1.Aeft rthe described threshold for the REMARK score of study quality because inclusion criteria, 18 eligible studies involving 1791 advanced multivariable survival measures are more valuable than cancer patients were selected for the current meta-analysis Journal of Oncology 3 2263 Articles identified via online databases 3 Additional papers identified via hand-searching 1018 Articles aer duplicates removed 911 Articles excluded Irrelevant title or abstract Not human samples 107 Full-text articles assessed for eligibility Articles excluded 67Univariate analysis or not advanced cancer 22 No available prognostic outcomes 18 Studies reporting multivariate analysis 13 Studies regarding CD133 5 Studies regarding SOX2 Figure 1: Flow chart for identification of eligible studies. [34–51]. Of these studies, 13 studies published from 2006 . . Subgroup and Sensitivity Analyses of CD Positive to 2017 (one prospective study and 12 retrospective studies) Expression in OS. We summarized the results of the subgroup evaluated the prognostic role of CD133 positivity [35, 37, analyses among several related clinical parameters (tumor 38, 40, 42, 44–51], including 1358 cases. Five studies (one type, study source, survival rate, sample type, age (years), prospective study and four retrospective studies) assessed testing method, study center design, treatment regimens, theprognosticroleofSOX2positivity[34,36, 39,41, 43], andsamplesize) forOSinTable 2. All P values of hetero- including 433 cases. The mean REMARK scores were 21, with geneity were not more than 0.1 between different subgroups; a range from 12 to 28. Most studies (78%) reported patients subgroup analyses did not find the potential sources of treated with adjuvant therapy. All articles published were heterogeneity. from 2006 to 2017, and six studies were conducted in China, Based on tumor type, significant difference was not found six studies in Japan, one study in Korea, and the remaining in 848 patients with colorectal cancer (six studies: HR = 1.27, vfi e studies in Europe. eTh characteristics of the eligible 95% CI = 0.64-2.50, P =0.493), 152patientswithovarian studies using multivariable analysis are listed in Table 1 and cancer (two studies: HR = 3.27, 95% CI = 0.43-25.03, P = Table S2. 0.254), and 32 patients with melanoma (one study: HR = 1.1, 95% CI = 0.34-3.8). There was statistical significance in .. Association between CD Positive Expression and the patients with 50 cancer patients with bone metastases (one Prognosis. The pooled data from two studies involving 176 study: HR = 9.73, 95% CI = 1.08-87.49) and 100 patients with advanced cancer patients showed that CD133 positive expres- gastric cancer (one study: HR = 2.097, 95% CI = 1.003-4.383). sion was associated with a worse cancer-specific survival Subgroup analysis by treatment regimens indicated that (CSS) (HR = 3.70, 95% CI = 1.09-12.54, P = 0.036) (Figure 2). CD133 positivity was slightly linked to poor OS in patients Data from vfi e studies involving 729 patients with advanced treated with adjuvant therapy (4 studies with 309 cases: HR = cancer demonstrated no association between CD133 positive 1.91, 95% CI = 1.08-3.39, P = 0.026). Subgroup analysis of study expression and DFS (HR = 1.62, 95% CI = 0.80-3.26, P =0.178) source showed that only Chinese with CD133 positivity was (Figure 2). significantly correlated with a worse OS (four studies with 579 11 studies with 1182 cases were included in the final cases: HR = 2.12, 95% CI = 1.35-3.33, P = 0.001). Subgroup analysis of CD133 positivity and OS. Data showed that CD133 analysis of survival rate indicated that CD133 positivity was positivity was slightly correlated with an unfavorable OS (HR significantly related to a less than 3-year OS (two studies = 1.57, 95% CI = 0.99-2.51, P = 0.057) (Figure 3). with 79 cases: HR = 10.92, 95% CI = 2.44-48.96, P =0.002). 4 Journal of Oncology Table 1: Main characteristics of studies included in the meta-analysis. Testing First author Study source Age Cancer type Study design Specimen type Cases Survival rate Outcomes Therapy method The Cancer with bone Retrospective, Part (adjuvant Mehra 2006 NA NASBA Blood 50 < 3years OS Netherlands metastases multicentre therapy) Advanced colon Retrospective, Adjuvant Li 2009 China NA IHC Tissue 104 5 years OS carcinoma single-center chemotherapy Flow Retrospective, Neoadjuvant Fusi 2011 Germany 54 cytometry Metastatic melanoma Blood 32 NA OS single-center chemotherapy analysis Colorectal liver Retrospective, Surgery and Pilati 2012 Italy 63 qRT-PCR Blood 50 3 years CSS metastasis single-center chemotherapy Colorectal cancer Retrospective, Sakai 2012 Japan NA IHC Tissue 92 3 years OS, DFS Surgery with liver metastasis single-center Advanced serous Retrospective, Adjuvant Qin 2012 China NA IHC Tissue 123 NA OS ovarian cancer multicentre chemotherapy Advanced gastric Retrospective, Surgery and adjuvant Lee 2012 Korea 61.5 IHC Tissue 100 5 years OS, DFS cancer single-center chemotherapy Advanced rectal Prospective, Surgery and Sprenger 2013 Germany 63 IHC, blind Tissue 126 NA CSS, DFS adenocarcinoma multicentre radiochemotherapy Yamamoto Colorectal cancer Retrospective, Surgery and Japan NA IHC, blind Tissue 103 5 years OS 2014 liver metastasis single-center chemotherapy Epithelial ovarian cancer with central Retrospective, Surgery and adjuvant Liu 2014 China 57 IHC, blind Tissue 29 < 3years OS nervous system single-center therapy metastasis Colorectal cancer Retrospective, Surgery and adjuvant Kazama 2015 Japan 67.1 IHC with lymph node Tissue 138 > 5years OS single-center chemotherapy metastasis Colorectal cancer Kishikawa Retrospective, Surgery and adjuvant Japan 59.4 IHC with synchronous Tissue 88 NA OS, DFS 2016 single-center chemotherapy liver metastases Advanced colorectal Retrospective, Surgery and adjuvant Pei 2016 China NA IHC, blind Tissue 323 NA OS, DFS cancer single-center chemotherapy Breast cancer with Retrospective, Huang 2014 China NA IHC, blind Tissue 107 NA DFS NA axillary lymph nodes multicentre Advanced cervical Retrospective, Shen 2014 China 51 IHC, blind squamous cell Tissue 132 5 years PFS Radiotherapy multicentre carcinoma Lung squamous cell Udagawa carcinoma with Retrospective, Japan 66 IHC Tissue 113 NA RFS Surgery 2015 lymph node single-center metastasis Advanced small-cell Prospective, Sodja 2016 Slovenia 65 qRT-PCR Blood 50 NA DFS, PFS Chemotherapy lung cancer single-center Yamawaki Advanced Retrospective, Japan NA IHC Tissue 31 NA PFS NA 2017 endometrial cancer single-center NA: not applicable; NASBA: nuclear acid sequence-based amplification; IHC: immunohistochemistry; qRT-PCR: Real-Time Quantitative PCR; OS: overall survival; DFS: disease-free survival; PFS: progression-free survival; CSS: cancer-specific survival; RFS: recurrence-free survival (RFS). Journal of Oncology 5 Study % ID HR (95% CI) Weight CSS Pilati 2012 2.61 (1.94, 3.60) 18.15 Sprenger 2013 11.30 (1.38, 92.77) 5.22 Subtotal (I-squared = 45.2%, p = 0.177) 3.70 (1.09, 12.54) 23.37 DFS Sakai 2012 1.01 (0.53, 1.92) 15.35 Lee 2012 3.41 (1.69, 6.88) 14.79 Sprenger 2013 3.97 (1.49, 9.13) 12.82 Kishikawa 2016 0.53 (0.31, 0.91) 16.33 Pei 2016 1.86 (1.22, 2.84) 17.34 Subtotal (I-squared = 84.9%, p = 0.000) 1.62 (0.80, 3.26) 76.63 Overall (I-squared = 83.9%, p = 0.000) 1.94 (1.11, 3.41) 100.00 NOTE: Weights are from random effects analysis .05 .1 .5 1 1 2 10 20 Figure 2: Forest plot for the correlation between CD133 positive expression and cancer-specific survival (CSS) and disease-free survival (DFS). Stratified analysis by age demonstrated that CD133 positivity SOX2 positivity and PFS (three studies with 213 cases: HR = was significantly associated with shorter OS in patients aged 1.77, 95% CI = 0.82-3.80, P = 0.145) (Figure 4). more than 60 years (two studies with 238 cases: HR = 2.09, 95% CI = 1.20-3.64, P =0.009).Signicfi antdieff rencewas . Publication Bias. Publication bias was detected for OS not noted between other subgroup analyses (sample type, and DFS of CD133 positive expression. No evidence of study center design, and sample size) and CD133 positivity publication bias was noted using Egger’s test (P =0.564> (Table 2). 0.05) and Begg’s test (P =0.876> 0.05) in OS (Figure S1). Sensitivity analysis was performed by omitting an indi- Moreover, we did not find publication bias for DFS of CD133 vidual study by turn to detect the robustness of the result. The positive expression (P> 0.1) (Figure S1). result showed that two studies conducted by Yamamoto 2014 et al. [42] and Kishikawa 2016 et al. [37] in Japan significantly . . TSA. When the prespecified type I error 𝛼 (5%), a aeff cted thepooledHRvalue,withthesignicfi antHR (2.02, RRR of 20%, and a type II error 𝛽 of 20% (80% power) 95% CI = 1.56-2.60, P< 0.001) and no evidence of heterogene- were set, the TSA showed that cumulative Z curve did ity (P = 0.413). not cross the sequential monitoring boundary for CSS and OS of CD133 positive expression (Figure 5). For DFS of .. Association between SOX Positive Expression and the SOX2 positivity,cumulativeZcurvewas notmorethanthe Prognosis. SOX2 positivity was associated with worse DFS sequential monitoring boundary (Figure S2A). For positive (two studies with 157 cases: HR = 3.08, 95% CI = 1.76-5.40, P< resultsofOSofCD133 positivity amongsubgroups,the TSA 0.001) and RFS (one study with 113 cases: HR = 1.736, 95% CI = also demonstrated that cumulative Z curve did not cross the 1.055-2.901, P = 0.033), but no relationship was found between trial sequential monitoring boundary (Table 2). 6 Journal of Oncology Study % ID HR (95% CI) Weight Mehra 2006 9.73 (1.08, 87.49) 3.42 Li 2009 2.68 (1.37, 5.26) 10.90 Fusi 2011 1.10 (0.34, 3.80) 7.26 Lee 2012 2.10 (1.00, 4.38) 10.42 Qin 2012 1.43 (0.85, 2.42) 11.96 Sakai 2012 2.03 (0.94, 4.37) 10.19 Liu 2014 12.08 (1.55, 94.16) 3.78 Yamamoto 2014 0.32 (0.13, 0.81) 9.08 Kazama 2015 2.08 (0.94, 5.06) 9.66 Kishikawa 2016 0.48 (0.25, 0.90) 11.13 Pei 2016 2.08 (1.27, 3.39) 12.19 Overall (I-squared = 72.6%, p = 0.000) 1.57 (0.99, 2.51) 100.00 NOTE: Weights are from random effects analysis .05 .1 .5 1 1 2 10 20 Figure 3: Forest plot for the correlation between CD133 positive expression and overall survival (OS). When the type I error of 5%, a RRR of 20%, and a type II maybeassociatedwithpooroverall survival innonsmall- cell lung cancer [55], worse prognosis in patients with error of 10% (90% power) were used, TSA also demonstrated glioblastoma [20], and reduced overall survival in colorectal that the cumulative Z curve did not reach the sequential cancer [56]. SOX2 expression may be correlated with better monitoring boundary between CD133 positivity and CSS overallsurvivalinnonsmallcelllungcancer[21], butworse and OS (Figure 6). The TSA showed that the cumulative Z overall survival in head and neck cancer [57]. However, some curve did not cross the trial sequential monitoring boundary results were contradictory, for example, patients with CD133- between SOX2 positivity and DFS (Figure S2B). positive is correlated with a better prognosis in colorectal liver When the type I error of 5%, type II error of 20%, and a metastasis [42]. Patients with CD133-positive are associated more conservative RRR of 15% were set, the results remained with an unfavorable prognosis in advanced colorectal cancer consistent, and the TSA also showed that cumulative Z [35]. eTh conventional prognostic factors such as tumor stage curve did not reach the trial sequential monitoring boundary or grade could not well predict clinical outcome based on (Figure 7 and Figure S2C). an individual basis [58]. To date, there are still no eecti ff ve markers available for the prognosis of patients with advanced cancer. er Th efore, it remains important to better understand 4. Discussion the characteristics of CSCs, CD133, and SOX2 for valuable CSCs, a small subpopulation of tumor cells, drive the growth therapeutic and prognostic targets in clinical practice to and progression of cancers [52]. More importantly, CSCs predict disease outcomes in advanced or metastatic cancer are considered to be involved in chemotherapy/radiotherapy patients. In our meta-analysis, we have attempted to estimate resistance, metastasis, and postoperative recurrence [53, 54]. the prognostic effect of CSCs, CD133, and SOX2 using Some meta-analyses showed that CD133 was a biomarker multivariable analysis in patients with advanced or metastatic of putative CSCs in many solid tumors and its positivity cancer. Journal of Oncology 7 Table 2: Subgroup analyses of CD133 positivity in overall survival (OS). Factors Subgroups Studies HR with 95% CI Heterogeneity (P)PvalueCases TSA Tumor type Colorectal cancer 6 1.27 (0.64-2.50) < 0.001 0.493 848 Ovarian cancer 2 3.27 (0.43-25.03) 0.049 0.254 152 Melanoma 1 1.1 (0.34-3.8) NA > 0.05 32 Cancer with bone metastases 1 9.73 (1.08-87.49) NA < 0.05 50 More Gastric cancer 1 2.097 (1.003-4.383) NA < 0.05 100 More Study source Japanese 4 0.90 (0.36-2.26) 0.001 0.823 421 Chinese 4 2.12 (1.35-3.33) 0.154 0.001 579 More Others 3 2.07 (0.90-4.78) 0.229 0.089 182 Survival rate 5 years 3 1.26 (0.38-4.17) 0.001 0.703 307 < 3 years 2 10.92 (2.44-48.96) 0.888 0.002 79 More Others 6 1.38 (0.85-2.26) 0.01 0.197 796 Sample type Tissue 9 1.51 (0.92-2.49) < 0.001 0.103 1100 Blood 2 2.68 (0.33-21.83) 0.088 0.358 82 Age (years) > 60 2 2.09 (1.20-3.64) 0.989 0.009 238 More ≤ 60 3 1.40 (0.31-6.27) 0.01 0.661 149 NA 6 1.64 (0.92-2.91) 0.003 0.093 795 Treatment regimens Adjuvant therapy 4 1.91 (1.08-3.39) 0.173 0.026 309 More Surgery and adjuvant therapy 6 1.34 (0.62-2.93) < 0.001 0.457 781 Testing method Blind 3 1.64 (0.32-8.37) < 0.001 0.553 455 NA 8 1.62 (1.00-2.63) 0.005 0.048 727 More Study center design Multicentre 2 2.74 (0.46-16.19) 0.097 0.266 173 Single-center 8 1.54 (0.85-2.79) < 0.001 0.154 977 NA 1 1.1 (0.34-3.8) NA > 0.05 32 Sample size ≥ 100 6 1.58 (0.97-2.57) 0.007 0.066 891 < 100 5 1.93 (0.66-5.65) 0.001 0.232 291 HR: hazard ratio; 95% CI: 95% confidenceinterval;NA:not applicable;TSA:trialsequential analysis. Chemotherapy and radiotherapy are major treatment subgroup analyses. eTh removal of the study by Yamamoto strategies to eliminate cancer cells, but chemoresistance, 2014 [42] used blinding of the detection and the removal radioresistance, and cancer recurrence are major obstacles ofthestudy by Kishikawa2016[37]did notreportblinding for the long-term survival of cancer patients [59, 60]. Recent of the detection (Table 1). We did not n fi d that the possible studies show that CSCs are resistant to chemotherapy and factors and reasons can influence the pooled HR of OS radiotherapy and targeting CSCs may become a promising in CD133. Because these two retrospective studies [37, 42] opportunity to cure patients with cancer [54, 61]. eTh studies reported that CD133 positivity was linked to favorable OS. of 78% (14 studies) reported patients with adjuvant therapy SOX2 positivity was related to shorter DFS (HR = 3.08, P< such as chemotherapy and radiotherapy in this meta-analysis. 0.001) and RFS (HR = 1.736, P = 0.033), but SOX2 positivity According to a comprehensive analysis of published studies was not correlated with PFS (HR = 1.77, P =0.145). In (CD133: 13 studies with 1358 patients and SOX2: vfi e studies addition, no publication bias was observed in OS and DFS with 433 patients). We found that patients with CD133- of CD133. These positive results were further proven by TSA, positive advanced cancer was correlated with poorer CSS and the data suggested that additional clinical trials were (HR = 3.70, P =0.036)and showedatrend towardspoor needed to confirm these conclusions. OS (HR = 1.57, P =0.057), butnorelationshipwas reported We further performed subgroup analyses of CD133 between CD133 positivity and DFS (HR = 1.62, P = 0.178). For expression stratified by cancer type, study source, survival the analyses of CD133 in OS, we performed sensitivity and rate,sampletype,age(years),testingmethod,studycenter 8 Journal of Oncology Study % ID HR (95% CI) Weight DFS Huang 2014 2.92 (1.23, 6.93) 14.22 Sodja 2016 3.20 (1.54, 6.71) 16.06 Subtotal (I-squared = 0.0%, p = 0.871) 3.08 (1.76, 5.40) 30.28 RFS Udagawa 2015 1.74 (1.05, 2.90) 19.63 Subtotal (I-squared = .%, p = .) 1.74 (1.05, 2.88) 19.63 PFS Shen 2014 2.29 (1.01, 5.20) 14.88 Sodja 2016 1.05 (0.94, 1.18) 24.15 Yamawaki 2017 3.49 (1.14, 10.71) 11.07 Subtotal (I-squared = 73.8%, p = 0.022) 1.77 (0.82, 3.80) 50.10 Overall (I-squared = 78.4%, p = 0.000) 2.06 (1.24, 3.41) 100.00 NOTE: Weights are from random effects analysis .05 .1 .5 1 1 2 10 20 Figure 4: Forest plot for the association between SOX2 positivity and disease-free survival (DFS), relapse/recurrence-free survival (RFS), and progression-free survival (PFS). design, and sample size in OS. Subgroup analysis by cancer center design,andsamplesize. WefurtherusedTSA to type showed that CD133 expression was associated with achieve more meaningful results among different subgroups. shorterOSincancerwithbonemetastasesandgastriccancer TSA suggested that the available sample data were insucffi ient but no relationship in colorectal cancer, ovarian cancer, to draw rfi m conclusions regarding the expression of CD133 and melanoma. Stratified analysis by study source indicated to OS. that only CD133 positivity could significantly reduce OS in Our meta-analysis had some limitations. First, the num- Chinese patients (HR = 2.12, P =0.001), suggestingthat ber of the included studies was not very large and some of CD133 may play a more important role in the prognosis of these eligible studies had small sample sizes. TSA confirmed advanced cancer for Chinese. Stratified analysis by survival that cumulative Z curve did not cross the sequential monitor- rate showed that only CD133 positivity might significantly ing boundary. u Th s, more trials are needed for more reliable decrease OS in patients with< 3-year survival rate (HR = results. Second, studies were mainly conducted in China, 10.92, P = 0.002), which suggested that the expression of Japan, and Europe; thus, other study sources (USA) are lack- CD133 may be correlated with shorter OS within 3 years. ing. Third, most studies were of retrospective design; only two Subgroup analysis by age indicated that only CD133 positivity studies were of prospective design. Additional prospective can significantly shorten OS in patients aged more than 60 clinical studies (such as blinded detection of CD133 and SOX2 years (HR = 2.09, P = 0.009), suggesting that CD133 may expression) are essential to obtain more firm results in differ- play a more key role in the prognosis for elderly patients. ent cancer types, such as colorectal, lung, breast, and head- However, no significant difference was found between CD133 neck cancer. Finally, there was considerable heterogeneity in positivity and other subgroups such as sample type, study this meta-analysis. Although we analyzed several factors that Journal of Oncology 9 APIS = 2543 APIS = 2543 RRR = 20% (alpha = 5%, power = 80%) RRR = 20% (alpha = 5%, power = 80%) 8 8 Sequential monitoring boundary 6 6 Sequential monitoring boundary 4 4 2 2 Cumulative Z curve Cumulative Z curve 0 0 0 500 1000 1500 2000 2500 0 500 1000 1500 2000 2500 Information size Information size CD133: overall survival (OS), cumulative Z curve did not CD133: cancer-specific survival (CSS), cumulative Z curve cross the sequential monitoring boundary did not cross the sequential monitoring boundary Figure 5: Trial sequential analysis (TSA) for cancer-specific survival (CSS) and overall survival (OS) of CD133 positive expression ( 𝛼 =5%, 𝛽 = 20%, and the relative risk reduction (RRR) = 20%). APIS = 3404 APIS = 3404 RRR = 20% (alpha = 5%, power = 90%) RRR = 20% (alpha = 5%, power = 90%) 8 8 Sequential monitoring boundary Sequential monitoring boundary 6 6 4 4 2 2 Cumulative Z curve Cumulative Z curve 0 0 0 1000 2000 3000 4000 0 1000 2000 3000 4000 Information size Information size CD133: overall survival (OS), cumulative Z curve did not CD133: cancer-specific survival (CSS), cumulative Z curve cross the sequential monitoring boundary did not cross the sequential monitoring boundary Figure 6: Trial sequential analysis (TSA) for cancer-specific survival (CSS) and overall survival (OS) of CD133 positive expression ( 𝛼 =5%, 𝛽 = 10%, and the relative risk reduction (RRR) = 20%). may influence heterogeneity, these variables could not clearly OS. Subgroup analysis by tumor type showed that CD133 explain the sources of heterogeneity. u Th s, clinical practice positivity was linked to worse OS in cancer with bone should interpret our results with caution. metastases and gastric cancer. Subgroup analysis by study To conclude, our meta-analysis showed that CD133- source demonstrated that only CD133 positivity was related positive expression may be associated with worse CSS and to poor OS for Chinese. Subgroup analysis by survival rate 10 Journal of Oncology APIS = 4776 APIS = 4776 RRR = 15% (alpha = 5%, power = 80%) RRR = 15% (alpha = 5%, power = 80%) 8 8 Sequential monitoring boundary Sequential monitoring boundary 6 6 4 4 2 2 Cumulative Z curve Cumulative Z curve 0 0 0 1000 2000 3000 4000 5000 0 1000 2000 3000 4000 5000 Information size Information size CD133: overall survival (OS), cumulative Z curve did not CD133: cancer-specific survival (CSS), cumulative Z curve cross the sequential monitoring boundary did not cross the sequential monitoring boundary Figure 7: Trial sequential analysis (TSA) for cancer-specific survival (CSS) and overall survival (OS) of CD133 positive expression ( 𝛼 =5%, 𝛽 = 20%, and the relative risk reduction (RRR) = 15%). showed that CD133 positivity was correlated with a less and Fenggang Hou contributed to data analyses and the than 3-year OS. Subgroup analysis by age demonstrated that interpretation and completion of the gfi ures and tables. All the expression of CD133 was associated with shorter OS in authors read and approved the n fi al manuscript. patients> 60 years. SOX2 positivity may be related to poor DFS and RFS. Further TSA suggested the need for addi- Acknowledgments tional clinical studies. Herein, more high-quality prospective studiesare essentialtoobtainmorereliableevidenceand Thisresearchwassupportedbygrantsfromthe Natural help stratify advanced cancer patients who can benefit from Science Foundation of China (81473624) and the Shanghai different therapies. Science and Technology Innovation Action Plan Project (No. 16401970500-3). Data Availability Supplementary Materials ed Th ata usedtosupportthefindingsofthisstudy are included within the article. Supplementary . Table S1: REMARK guidelines. Supplementary . Table S2: Detailed characteristics of studies Disclosure included in the meta-analysis. SusuHanand TaoHuangareco-rfi st authorsofthisstudy. Supplementary . Figure S1: Publication bias using Egger’s and Begg’s tests for overall survival (OS) and disease-free Conflicts of Interest survival (DFS) of CD133 positive expression. Supplementary . Figure S2: Trial sequential analysis (TSA) The authors declare no conflicts of interest. for disease-free survival (DFS) of SOX2 positivity. Authors’ Contributions References Susu Han, Tao Huang, and Fenggang Hou contributed to [1] L. A. Torre, F. Bray, R. L. Siegel, J. Ferlay, and J. Lortet- the conception and design of this research. Susu Han, Xing Tieulent, “Global cancer statistics, 2012,” CA: A Cancer Journal Wu, Xiyu Wang, Shanshan Liu, Wei Yang, and Qi Shi for Clinicians,vol.65,no.2,pp. 87–108,2015. contributed to the drafting of the article and final approval of the submitted version. Susu Han, Tao Huang, Xing Wu, [2] A.Urruticoechea,R.Alemany,J.Balart,A. Villanueva,F. Xiyu Wang, Shanshan Liu, Wei Yang, Qi Shi, Hongjia Li, Vinals, ˜ and G. Capella, ´ “Recent advances in cancer therapy: An Journal of Oncology 11 overview,” Current Pharmaceutical Design,vol.16, no.1,pp. 3– [21] Y.Chen,Y.Huang,Y.Huang etal.,“eP Th rognosticValue of 10, 2010. SOX2 Expression in Non-Small Cell Lung Cancer: A Meta- Analysis,” PLoS ONE,vol.8,no.8, p.e71140,2013. [3] S. G. C. Kroeze, C. Fritz, M. Hoyer et al., “Toxicity of concurrent stereotactic radiotherapy and targeted therapy or immunother- [22] A. Liberati, D. G. Altman, J. Tetzlaff et al., “The PRISMA apy:Asystematicreview,” Cancer Treatment Reviews,vol.53,pp. statement for reporting systematic reviews and meta-analyses of 25–37, 2017. studies that evaluate health care interventions: explanation and elaboration,” PLoS Medicine, vol. 6, no. 7, Article ID e1000100, [4] R. L. Siegel, L. Sahar, K. M. Portier, E. M. Ward, and A. Jemal, “Cancer death rates in US congressional districts,” CA: A Cancer Journal for Clinicians,vol.65, no.5,pp.339–344, 2015. [23] D. G. Altman and J. M. Bland, “How to obtain the confidence interval from a P value,” British Medical Journal,vol.343,Article [5] D. M. Pardoll, “The blockade of immune checkpoints in cancer ID d2090, 2011. immunotherapy,” Nature Reviews Cancer,vol.12,no.4,pp. 252– 264, 2012. [24] J. F. Tierney, L. A. Stewart, D. Ghersi, S. Burdett, and M. R. Sydes, “Practical methods for incorporating summary time-to- [6] M.R.Doherty,J.M.Smigiel, D.J. Junk,and M. W. Jackson, event data into meta-analysis,” Trials,vol.8,article 16,2007. “Cancer stem cell plasticity drives therapeutic resistance,” Can- cers,vol.8,no. 1,pp.1–13, 2016. [25] L.M.McShane, D. G.Altman,W.Sauerbrei,S.E.Taube, M. Gion, and G. M. Clark, “Reporting Recommendations for [7] M. Derwahl, “Linking stem cells to thyroid cancer,” e Journal Tumor Marker Prognostic Studies (REMARK),” JNCI: Journal of Clinical Endocrinology & Metabolism,vol.96,no.3,pp.610– of the National Cancer Institute,vol.97, no.16, pp.1180–1184, 613, 2011. [8] S. Han, W. Yang, S. Zong et al., “Clinicopathological, prognostic [26] E. Zintzaras and J. P. A. Ioannidis, “HEGESMA: genome search and predictive value of CD166 expression in colorectal cancer: A meta-analysis,” Oncotarget ,vol.8,no. 38,pp. 64373–64384, meta-analysis and heterogeneity testing,” Bioinformatics,vol.21, no.18, pp.3672-3673,2005. [27] E. Evangelou and J. P. A. Ioannidis, “Meta-analysis methods for [9] S.Han,S.Zong, Q.Shietal.,“Is Ep-CAM Expression a Diagnostic and Prognostic Biomarker for Colorectal Cancer? genome-wide association studies and beyond,” Nature Reviews Genetics,vol.14,no.6,pp. 379–389, 2013. A Systematic Meta-Analysis,” EBioMedicine,vol.20,pp.61–69, [28] R. DerSimonian and R. Kacker, “Random-eeff cts model for meta-analysis of clinical trials: an update,” Contemporary Clini- [10] A. R. Safa, “Resistance to cell death and its modulation in cancer stem cells,” Critical Reviews in Oncogenesis,vol.21,no.3-4,pp. cal Trials, vol. 28, no. 2, pp. 105–114, 2007. 203–219, 2016. [29] M. Egger, G. D. Smith, M. Schneider, and C. Minder, “Bias [11] D. Corbeil, C. A. Fargeas, and W. B. Huttner, “Rat prominin, like in meta-analysis detected by a simple, graphical test,” British its mouse and human orthologues, is a pentaspan membrane Medical Journal, vol. 315, pp. 629–634, 1997. glycoprotein,” Biochemical and Biophysical Research Communi- [30] C. B. Begg and M. Mazumdar, “Operating characteristics of a cations,vol.285,no.4, pp.939–944,2001. rank correlation test for publication bias,” Biometrics,vol.50, [12] A. H. Yin, S. Miraglia, E. D. Zanjani et al., “AC133, a novel no. 4, pp. 1088–1101, 1994. marker for human hematopoietic stem and progenitor cells,” [31] B. Miladinovic, R. Mhaskar, I. Hozo, A. Kumar, H. Mahony, and Blood,vol.90, no.12, pp.5002–5012,1997. B. Djulbegovic, “Optimal information size in trial sequential [13] S.Masui,Y.Nakatake, Y. Toyookaetal.,“Pluripotencygoverned analysis of time-to-event outcomes reveals potentially incon- by Sox via regulation of Oct / expression in mouse embryonic clusive results because of the risk of random error,” Journal of stem cells,” Nature Cell Biology,vol.9,no.6, pp.625–635,2007. Clinical Epidemiology,vol.66, no.6,pp.654–659,2013. [14] J.-X. Cheng, B.-L. Liu, and X. Zhang, “How powerful is CD133 [32] K. o Th rlund, P. J. Devereaux, J. Wetterslev et al., “Can trial as a cancer stem cell marker in brain tumors?” Cancer Treatment sequential monitoring boundaries reduce spurious inferences Reviews,vol.35, no.5,pp. 403–408,2009. from meta-analyses?” International Journal of Epidemiology, vol. 38, no. 1, pp. 276–286, 2009. [15] N. R. Murray, V. Justilien, and A. P. Fields, “SOX2 Determines Lineage Restriction: Modeling Lung Squamous Cell Carcinoma [33] J. Brok, K. or Th lund, C. Gluud, and J. Wetterslev, “Trial in the Mouse,” Cancer Cell,vol.30, no.4,pp.505–507, 2016. sequential analysis reveals insufficient information size and potentially false positive results in many meta-analyses,” Journal [16] S. B. Keysar, P. N. Le, B. Miller et al., “Regulation of head and of Clinical Epidemiology,vol.61, no.8,pp.763–769, 2008. neck squamous cancer stem cells by PI3K and SOX2,” Journal of the National Cancer Institute,vol.109,no. 1,2017. [34] K. Yamawaki, T. Ishiguro, Y. Mori et al., “Sox2-dependent inhi- bition of p21 is associated with poor prognosis of endometrial [17] S. Boumahdi,G.Driessens,G.Lapouge etal., “SOX2controls tumour initiation and cancer stem-cell functions in squamous- cancer,” Cancer Science,vol.108,no.4, pp.632–640,2017. cell carcinoma,” Nature,vol.511, no.7508, pp.246–250,2014. [35] Q. Pei, H. Zhu, F. Tan et al., “Intravascular emboli is an [18] Z.Fan,M.Li,X.Chenetal.,“Prognostic ValueofCancerStem independent risk factor for the prognosis of stage III colorectal cancer patients aer ft radical surgery,” Oncotarget ,vol.7,no. 35, Cell Markers in Head and Neck Squamous Cell Carcinoma: a Meta-analysis,” Scientific Reports,vol.7,no.1,2017. pp. 57268–57276, 2016. [19] B. Cheng, G. Yang, R. Jiang et al., “Cancer stem cell markers [36] E. Sodja, M. Rijavec, A. Koren, A. Sadikov, P. Koroˇsec, and T. predict a poor prognosis in renal cell carcinoma: A meta- Cufer,“ep Th rognosticvalue ofwholebloodSOX2,NANOG analysis,” Oncotarget ,vol.7,no. 40,pp. 65862–65875, 2016. and OCT4 mRNA expression in advanced small-cell lung cancer,” Radiology and Oncology, vol. 50, no. 2, pp. 188–196, [20] W.Zhang,H.Chen,S.Lv,andH.Yang,“HighCD133Expression Is Associated with Worse Prognosis in Patients with Glioblas- toma,” Molecular Neurobiology,vol.53,no.4,pp. 2354–2360, [37] J. Kishikawa, S. Kazama, K. Oba et al., “CD133 Expression at 2016. theMetastaticSitePredicts Patients’OutcomeinColorectal 12 Journal of Oncology Cancer with Synchronous Liver Metastasis,” Annals of Surgical [52] N.A.Lobo,Y.Shimono,D.Qian,andM. F. Clarke,“eTh biology Oncology,vol.23, no.6,pp.1916–1923,2016. of cancer stem cells,” Annual Review of Cell and Developmental Biology,vol.23,no.1,pp.675–699,2007. [38] S. Kazama, J. Kishikawa, K. Yasuda et al., “CD133 Expression in Lymph Node Metastases Is Associated with Tumor Aggres- [53] K. M. Govaert, B. L. Emmink, M. W. Nijkamp et al., “Hypoxia aer ft liver surgery imposes an aggressive cancer stem cell siveness During Lymph Node Metastasis in Colorectal Cancer,” Anticancer Reseach, vol. 35, no. 12, pp. 6599–6605, 2015. phenotype on residual tumor cells,” Annals of Surgery,vol.259, no. 4, pp. 750–759, 2014. [39] H. Udagawa, G. Ishii, M. Morise et al., “Comparison of the [54] R. Morrison, S. M. Schleicher, Y. Sun et al., “Targeting the expression levels of molecular markers among the peripheral Mechanisms of Resistance to Chemotherapy and Radiotherapy area and central area of primary tumor and metastatic lymph with the Cancer Stem Cell Hypothesis,”JournalofOncology,vol. node tumor in patients with squamous cell carcinoma of the 2011, Article ID 941876, 13 pages, 2011. lung,” Journal of Cancer Research and Clinical Oncology,vol.141, no. 8, pp. 1417–1425, 2015. [55] E. Chen, Z. Zeng, B. Bai, J. Zhu, and Z. Song, “eTh prognostic value of CSCs biomarker CD133 in NSCLC: A meta-analysis,” [40] B.-L. Liu, S.-J. Liu, A. Baskys et al., “Platinum sensitivity and Oncotarget ,vol.7,no. 35,pp. 56526–56539, 2016. CD133 expression as risk and prognostic predictors of central nervoussystemmetastasesinpatientswithepithelialovarian [56] K. Wang, J. Xu, J. Zhang, and J. Huang, “Prognostic role of cancer,” BMC Cancer,vol.14,no.1,2014. CD133expressionincolorectalcancer:ameta-analysis,” BMC Cancer,vol.12,no.1,2012. [41] L.Shen,X.Huang,X.Xie,J.Su,J. Yuan,andX. Chen,“High Expression of SOX2 and OCT4 Indicates Radiation Resistance [57] W.Li,B.Li, R.Wang,D.Huang,W.Jin,andS.Yang,“SOX2as and an Independent Negative Prognosis in Cervical Squamous prognostic factor in head and neck cancer: A systematic review Cell Carcinoma,” Journal of Histochemistry & Cytochemistry, and meta-analysis,” Acta Oto-Laryngologica,vol.134,no.11, pp. vol.62,no. 7,pp.499–509,2014. 1101–1108, 2014. [42] S. Yamamoto, K. Tanaka, K. Takeda et al., “Patients with CD133- [58] A. Bamias, M. Chorti, C. Deliveliotis et al., “Prognostic signif- negative colorectal liver metastasis have a poor prognosis aer ft icance of CA 125, CD44, and epithelial membrane antigen in hepatectomy,” Annals of Surgical Oncology,vol.21,no.6,pp. renalcellcarcinoma,” Urology,vol.62, no.2,pp. 368–373, 2003. 1853–1861, 2014. [59] K. Nurgali, R. T. Jagoe, and R. Abalo, “Editorial: Adverse effects [43] Y.-H. Huang, M.-H. Luo, Y.-B. Ni et al., “Increased SOX2 of cancer chemotherapy: Anything new to improve tolerance expression in less differentiated breast carcinomas and their and reduce sequelae?” Frontiers in Pharmacology,vol.9,2018. lymph node metastases,” Histopathology,vol.64, no.4,pp. 494– [60] G. Delaney, S. Jacob, C. Featherstone, and M. Barton, “The 503, 2014. role of radiotherapy in cancer treatment: estimating optimal [44] T. Sprenger, L.-C. Conradi, T. Beissbarth et al., “Enrichment utilization from a review of evidence-based clinical guidelines,” of CD133-expressing cells in rectal cancers treated with pre- Cancer, vol. 104, no. 6, pp. 1129–1137, 2005. operative radiochemotherapy is an independent marker for [61] P. S. Kharkar, “Cancer stem cell (CSC) inhibitors: a review metastasis and survival,” Cancer,vol.119,no.1,pp.26–35,2013. of recent patents (2012-2015),” Expert Opinion on erapeutic [45] H. H. Lee, K. J. Seo, C. H. An, J. S. Kim, and H. M. Jeon, Patents,vol.27,no.7,pp. 753–761, 2017. “CD133 expression is correlated with chemoresistance and early recurrence of gastric cancer,” Journal of Surgical Oncology,vol. 106, no.8,pp. 999–1004, 2012. [46] Q. Qin, Y. Sun, M. Fei et al., “Expression of putative stem marker nestin and CD133 in advanced serous ovarian cancer,” Neoplasma,vol.59,no.3,pp.310–315,2012. [47] N. Sakai, H. Yoshidome, T. Shida et al., “CXCR4/CXCL12 expression profile is associated with tumor microenvironment and clinical outcome of liver metastases of colorectal cancer,” Clinical & Experimental Metastasis,vol.29, no.2,pp. 101–110, [48] P. Pilati, S. Mocellin, L. Bertazza et al., “Prognostic value of putative circulating cancer stem cells in patients undergoing hepatic resection for colorectal liver metastasis,” Annals of Surgical Oncology,vol.19, no.2,pp.402–408,2012. [49] A. Fusi, U. Reichelt, A. Busse et al., “Expression of the stem cell markers nestin and CD133 on circulating melanoma cells,” Journal of Investigative Dermatology,vol.131,no.2,pp.487–494, [50] C. Li, B. Li, Y. Liang et al., “Higher percentage of CD133+ cells is associated with poor prognosis in colon carcinoma patients with stage IIIB,” Journal of Translational Medicine,vol.7,no. 1, p. 56, 2009. [51] N. Mehra, M. Penning, J. Maas et al., “Progenitor marker CD133 mRNA is elevated in peripheral blood of cancer patients with bone metastases,” Clinical Cancer Research,vol.12,no.16, pp. 4859–4866, 2006. MEDIATORS of INFLAMMATION The Scientific Gastroenterology Journal of World Journal Research and Practice Diabetes Research Disease Markers Hindawi Hindawi Publishing Corporation Hindawi Hindawi Hindawi Hindawi www.hindawi.com Volume 2018 http://www www.hindawi.com .hindawi.com V Volume 2018 olume 2013 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 International Journal of Journal of Immunology Research Endocrinology Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 Submit your manuscripts at www.hindawi.com BioMed PPAR Research Research International Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 Journal of Obesity Evidence-Based Journal of Journal of Stem Cells Complementary and Ophthalmology International Alternative Medicine Oncology Hindawi Hindawi Hindawi Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2013 Parkinson’s Disease Computational and Behavioural Mathematical Methods AIDS Oxidative Medicine and in Medicine Neurology Research and Treatment Cellular Longevity Hindawi Hindawi Hindawi Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Oncology Hindawi Publishing Corporation

Prognostic Value of CD133 and SOX2 in Advanced Cancer

Loading next page...
 
/lp/hindawi-publishing-corporation/prognostic-value-of-cd133-and-sox2-in-advanced-cancer-mTy500dQJ8
Publisher
Hindawi Publishing Corporation
Copyright
Copyright © 2019 Susu Han et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
ISSN
1687-8450
eISSN
1687-8469
DOI
10.1155/2019/3905817
Publisher site
See Article on Publisher Site

Abstract

Hindawi Journal of Oncology Volume 2019, Article ID 3905817, 12 pages https://doi.org/10.1155/2019/3905817 Research Article 1 2 1 1 1 1 Susu Han , Tao Huang , Xing Wu, Xiyu Wang, Shanshan Liu, Wei Yang, 1 1 1 Qi Shi, Hongjia Li, and Fenggang Hou Shanghai Municipal Hospital of Traditional Chinese Medicine, Shanghai University of Traditional Chinese Medicine, Zhijiang Road, Shanghai , China e Affiliated Hospital of Ningbo University, Ningbo, Zhejiang , China Correspondence should be addressed to Susu Han; anyasue@163.com, Tao Huang; huangtao334@163.com, and Fenggang Hou; fghou555@126.com Received 27 August 2018; Revised 24 October 2018; Accepted 26 November 2018; Published 1 January 2019 Academic Editor: Akira Hara Copyright © 2019 Susu Han et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Background. eTh prognostic value of CD133 and SOX2 expression in advanced cancer remains unclear. This study was first conducted to investigate the association between CD133 or SOX2 positivity and clinical outcomes for advanced cancer patients. Methods. Hazard ratios (HRs) with 95% confidence intervals (95% CIs) were calculated to evaluate the correlation between CD133 or SOX2 positivity and overall survival (OS), disease-free survival (DFS), progression-free survival (PFS), cancer-specific survival (CSS), or recurrence-free survival (RFS) from multivariable analysis. Trial sequential analysis (TSA) was also performed. Results.13 studies with 1358 cases (CD133) and vfi e studies with 433 cases (SOX2) were identified. CD133 positivity was correlated with worse CSS and OS, but there was no correlation between CD133 positivity and DFS. SOX2 positivity was associated with poor DFS and RFS but was not linked to PFS. Stratified analysis by study source showed that only CD133 positivity can decrease OS for Chinese patients. Stratified analysis by treatment regimens indicated that CD133 positivity was linked to poor OS in patients treated with adjuvant therapy. TSA showed that additional studies were necessary. Conclusions. CD133 and SOX2 might be associated with worse prognosis in advanced cancer. More prospective studies are strongly needed. Impact. CD133 and SOX2 may be promising targeted molecular therapy for advanced cancer patients. 1. Introduction proliferation and differentiation, and resistance to conven- tional treatments like chemotherapy or radiation [6, 7]. Recently, some stem cell markers have been described, Cancer is still one of the most threatening diseases worldwide [1]. Although surgery, chemotherapy, and/or radiotherapy suchas CD44,CD166,EpCAM,CD133,and SOX2 [8–10]. have greatly improved the clinical survival for early cancer CD133, also named as prominin-1, is a member of pentas- patients, therapies for patients with advanced or metastatic pan transmembrane cell surface glycoproteins [11, 12]. Sex- cancer still have a major challenge [2]. Improvements in determining region Y-box protein 2 (SOX2), a High Mobility the treatment of advanced or metastatic cancer patients Group (HMG) domain transcription factor, is involved in (surgical technique, chemotherapy, radiotherapy, targeted the regulation of stem cells self-renewal and pluripotency molecular therapy, and immunotherapy regimens) have [13]. CD133 expression has been reported and contributes to extended patients’ median survival, but such as 5-year overall malignant transformation and chemo- and radioresistance survival is still poor [3–5]. us Th the development of new and [14].SOX2hasbeen studiedinsometypes of humancancers novel therapeutic regimens for advanced or metastatic cancer and facilitates tumor initiation and progression [15–17]. Some meta-analyses investigated the prognostic value of CD133 and patients is important. Increasing evidence has been suggested regarding cancer SOX2 expression in some human cancers [18–21], but the stem cells (CSCs) in various cancers. The major character- prognostic significance of CD133 and SOX2 expression in istics of CSCs are the capability of self-renewal, unlimited advanced cancer patients remains unclear and unknown. 2 Journal of Oncology Toourknowledge,the expressionofCD133 andSOX2is studies using univariable analysis, done in the present meta- hitherto undescribed in advanced cancer by a meta-analysis. analysis. REMARK scores can be used and evaluated for To clarify the correlation between the expression of stem cell sensitivity analyses. eTh following information was extracted markers (CD133 and SOX2) and the prognosis in advanced from eligible studies: rfi st author’s name, publication year, or metastatic cancer patients, we investigated the relationship study population, study source, mean or median age, type between the expression of these two markers and survival of of cancer, detection method, therapy regime, study design, the samples. sample type, cut-off value, median or mean follow-up period, survival rate, adjusted variables, and clinical outcomes, etc. All authors resolved the discrepancy when information was 2. Materials and Methods controversial. .. Literature Selection. The present meta-analysis was .. Data Analysis. To estimate the effect of CD133 or SOX2 reported in accordance with the Preferred Reporting Items expression status on advanced cancer survival (OS, DFS, PFS, for Systematic reviews and Meta-Analyses (PRISMA) guide- CSS,RFS,orMFSofmultivariableanalysis),theresultwithan line [22]. eTh potential studies were identified through HR>1 demonstrated an unfavorable prognosis, whereas an searching online databases including PubMed, EMBASE, HR<1 stood for a good prognosis. The Cochran’s Q statistic EBSCO, Web of Science, and Cochrane Library before April was used to evaluate heterogeneity among the included stud- 2018 without language restrictions. eTh main key words ies [26]. The random-effects model (DerSimonian-Laird) was and search items were “CD133 OR PROM1 OR prominin- used in the meta-analysis (heterogeneity: P< 0.1) [27, 28]. For 1 OR AC133 antigen OR SOX2 OR Sex determine region the results (> seven studies) with substantial heterogeneity, Y-box 2 OR SRY box-2 OR SRY-Related HMG-Box Gene subgroup analyses based on tumor type, study source, sur- 2”, “metastatic OR advanced OR metastasized OR recur- vival rate, sample type, age (years), testing method, and study rent”, “cancer OR tumor OR carcinoma OR neoplasm”, and center design were performed to explain the potential het- “survival OR outcome OR prognosis”. Additional potential erogeneity and different strength of the association between articles were also manually searched by the reference lists of subgroups. If all relevant P values of heterogeneity were theeligiblestudies. greaterthan0.1 amongdieff rentsubgroups,itindicates the source of heterogeneity from a subgroup variable. eTh Egger’s and Begg’s funnel plots were used to evaluate publication .. Eligibility Criteria. Papers identiefi dfor theinclusion bias [29, 30]. Pooled data were analyzed using Stata software, criteria in this study for the current analysis were as follows: version 12.0 (Stata Corp., College Station, TX, USA). (1) studies reported the patients with advanced, metastatic, or recurrent cancer; (2) studies investigated the prognostic value of expression of CD133 or SOX2; (3) studies presented . Trial Sequential Analysis. In the meta-analysis involving sufficient data on hazard ratio (HR) with 95% confidence a small number of participants, random errors can lead interval (CI) from multivariable analysis for overall survival to spurious results [31, 32]. Trial sequential analysis (TSA) (OS), disease-free survival (DFS), progression-free survival was conducted to control random errors and to estimate (PFS), cancer-specific survival (CSS), relapse/recurrence-free therequiredstudy population [33].Theoptimal apriori survival (RFS), or metastasis-free survival (MFS); (4) unclear anticipated information size (APIS) method was set in our data (HR with 95% CI) such as only P value with HR or study. We calculated diversity-adjusted TSA based on the 95% CI, survival data calculated based on the described relative risk reduction (RRR) of 20%, the prespecified type method [23, 24], or contacting the corresponding author via I error of 5%, and the type II error (20% or 10%). We also email to request the available information. If two or more calculated diversity-adjusted TSA based on a RRR of 15%, papers used the overlapping or same cancer samples, only the prespecified type I error ( 𝛼) of 5%, and a type II error the study with the largest patient numbers or the most recent (𝛽) of 20%. Monitoring boundaries are applied to decide article was selected. Case report, reviews, animal studies, whether a clinical trial could be terminated early. When unrelated articles, or survival data using univariable analysis the cumulative Z curve was more than the trial sequential were excluded. monitoring boundary or required information size (RIS) boundary, it suggested the firm evidence. Otherwise, more clinical studies are needed. Meta-analysis of HR estimates . . Data Extraction and Study Assessment. The methodology was performed using Stata software, version 12.0 (Stata Corp., of each eligible study was conducted following REMARK College Station, TX, USA) and R software, version 3.4.2 (eTh guidelines (Reporting Recommendations for Tumor Marker R Foundation for Statistical Computing, Vienna, Austria). Prognostic Studies) [25]. 20 criteria were listed in REMARK; each itemhadscores0,1,and 2, withamaximal score of 40 (Table S1). The value was 2 scores when each item 3. Results was clearly described in the article, 1 score when each item was incompletely defined, and 0 score when each item .. Study Characteristics. Flowchart describing the study was not defined or not applicable. We did not define a selectionprocessisshowninFigure1.Aeft rthe described threshold for the REMARK score of study quality because inclusion criteria, 18 eligible studies involving 1791 advanced multivariable survival measures are more valuable than cancer patients were selected for the current meta-analysis Journal of Oncology 3 2263 Articles identified via online databases 3 Additional papers identified via hand-searching 1018 Articles aer duplicates removed 911 Articles excluded Irrelevant title or abstract Not human samples 107 Full-text articles assessed for eligibility Articles excluded 67Univariate analysis or not advanced cancer 22 No available prognostic outcomes 18 Studies reporting multivariate analysis 13 Studies regarding CD133 5 Studies regarding SOX2 Figure 1: Flow chart for identification of eligible studies. [34–51]. Of these studies, 13 studies published from 2006 . . Subgroup and Sensitivity Analyses of CD Positive to 2017 (one prospective study and 12 retrospective studies) Expression in OS. We summarized the results of the subgroup evaluated the prognostic role of CD133 positivity [35, 37, analyses among several related clinical parameters (tumor 38, 40, 42, 44–51], including 1358 cases. Five studies (one type, study source, survival rate, sample type, age (years), prospective study and four retrospective studies) assessed testing method, study center design, treatment regimens, theprognosticroleofSOX2positivity[34,36, 39,41, 43], andsamplesize) forOSinTable 2. All P values of hetero- including 433 cases. The mean REMARK scores were 21, with geneity were not more than 0.1 between different subgroups; a range from 12 to 28. Most studies (78%) reported patients subgroup analyses did not find the potential sources of treated with adjuvant therapy. All articles published were heterogeneity. from 2006 to 2017, and six studies were conducted in China, Based on tumor type, significant difference was not found six studies in Japan, one study in Korea, and the remaining in 848 patients with colorectal cancer (six studies: HR = 1.27, vfi e studies in Europe. eTh characteristics of the eligible 95% CI = 0.64-2.50, P =0.493), 152patientswithovarian studies using multivariable analysis are listed in Table 1 and cancer (two studies: HR = 3.27, 95% CI = 0.43-25.03, P = Table S2. 0.254), and 32 patients with melanoma (one study: HR = 1.1, 95% CI = 0.34-3.8). There was statistical significance in .. Association between CD Positive Expression and the patients with 50 cancer patients with bone metastases (one Prognosis. The pooled data from two studies involving 176 study: HR = 9.73, 95% CI = 1.08-87.49) and 100 patients with advanced cancer patients showed that CD133 positive expres- gastric cancer (one study: HR = 2.097, 95% CI = 1.003-4.383). sion was associated with a worse cancer-specific survival Subgroup analysis by treatment regimens indicated that (CSS) (HR = 3.70, 95% CI = 1.09-12.54, P = 0.036) (Figure 2). CD133 positivity was slightly linked to poor OS in patients Data from vfi e studies involving 729 patients with advanced treated with adjuvant therapy (4 studies with 309 cases: HR = cancer demonstrated no association between CD133 positive 1.91, 95% CI = 1.08-3.39, P = 0.026). Subgroup analysis of study expression and DFS (HR = 1.62, 95% CI = 0.80-3.26, P =0.178) source showed that only Chinese with CD133 positivity was (Figure 2). significantly correlated with a worse OS (four studies with 579 11 studies with 1182 cases were included in the final cases: HR = 2.12, 95% CI = 1.35-3.33, P = 0.001). Subgroup analysis of CD133 positivity and OS. Data showed that CD133 analysis of survival rate indicated that CD133 positivity was positivity was slightly correlated with an unfavorable OS (HR significantly related to a less than 3-year OS (two studies = 1.57, 95% CI = 0.99-2.51, P = 0.057) (Figure 3). with 79 cases: HR = 10.92, 95% CI = 2.44-48.96, P =0.002). 4 Journal of Oncology Table 1: Main characteristics of studies included in the meta-analysis. Testing First author Study source Age Cancer type Study design Specimen type Cases Survival rate Outcomes Therapy method The Cancer with bone Retrospective, Part (adjuvant Mehra 2006 NA NASBA Blood 50 < 3years OS Netherlands metastases multicentre therapy) Advanced colon Retrospective, Adjuvant Li 2009 China NA IHC Tissue 104 5 years OS carcinoma single-center chemotherapy Flow Retrospective, Neoadjuvant Fusi 2011 Germany 54 cytometry Metastatic melanoma Blood 32 NA OS single-center chemotherapy analysis Colorectal liver Retrospective, Surgery and Pilati 2012 Italy 63 qRT-PCR Blood 50 3 years CSS metastasis single-center chemotherapy Colorectal cancer Retrospective, Sakai 2012 Japan NA IHC Tissue 92 3 years OS, DFS Surgery with liver metastasis single-center Advanced serous Retrospective, Adjuvant Qin 2012 China NA IHC Tissue 123 NA OS ovarian cancer multicentre chemotherapy Advanced gastric Retrospective, Surgery and adjuvant Lee 2012 Korea 61.5 IHC Tissue 100 5 years OS, DFS cancer single-center chemotherapy Advanced rectal Prospective, Surgery and Sprenger 2013 Germany 63 IHC, blind Tissue 126 NA CSS, DFS adenocarcinoma multicentre radiochemotherapy Yamamoto Colorectal cancer Retrospective, Surgery and Japan NA IHC, blind Tissue 103 5 years OS 2014 liver metastasis single-center chemotherapy Epithelial ovarian cancer with central Retrospective, Surgery and adjuvant Liu 2014 China 57 IHC, blind Tissue 29 < 3years OS nervous system single-center therapy metastasis Colorectal cancer Retrospective, Surgery and adjuvant Kazama 2015 Japan 67.1 IHC with lymph node Tissue 138 > 5years OS single-center chemotherapy metastasis Colorectal cancer Kishikawa Retrospective, Surgery and adjuvant Japan 59.4 IHC with synchronous Tissue 88 NA OS, DFS 2016 single-center chemotherapy liver metastases Advanced colorectal Retrospective, Surgery and adjuvant Pei 2016 China NA IHC, blind Tissue 323 NA OS, DFS cancer single-center chemotherapy Breast cancer with Retrospective, Huang 2014 China NA IHC, blind Tissue 107 NA DFS NA axillary lymph nodes multicentre Advanced cervical Retrospective, Shen 2014 China 51 IHC, blind squamous cell Tissue 132 5 years PFS Radiotherapy multicentre carcinoma Lung squamous cell Udagawa carcinoma with Retrospective, Japan 66 IHC Tissue 113 NA RFS Surgery 2015 lymph node single-center metastasis Advanced small-cell Prospective, Sodja 2016 Slovenia 65 qRT-PCR Blood 50 NA DFS, PFS Chemotherapy lung cancer single-center Yamawaki Advanced Retrospective, Japan NA IHC Tissue 31 NA PFS NA 2017 endometrial cancer single-center NA: not applicable; NASBA: nuclear acid sequence-based amplification; IHC: immunohistochemistry; qRT-PCR: Real-Time Quantitative PCR; OS: overall survival; DFS: disease-free survival; PFS: progression-free survival; CSS: cancer-specific survival; RFS: recurrence-free survival (RFS). Journal of Oncology 5 Study % ID HR (95% CI) Weight CSS Pilati 2012 2.61 (1.94, 3.60) 18.15 Sprenger 2013 11.30 (1.38, 92.77) 5.22 Subtotal (I-squared = 45.2%, p = 0.177) 3.70 (1.09, 12.54) 23.37 DFS Sakai 2012 1.01 (0.53, 1.92) 15.35 Lee 2012 3.41 (1.69, 6.88) 14.79 Sprenger 2013 3.97 (1.49, 9.13) 12.82 Kishikawa 2016 0.53 (0.31, 0.91) 16.33 Pei 2016 1.86 (1.22, 2.84) 17.34 Subtotal (I-squared = 84.9%, p = 0.000) 1.62 (0.80, 3.26) 76.63 Overall (I-squared = 83.9%, p = 0.000) 1.94 (1.11, 3.41) 100.00 NOTE: Weights are from random effects analysis .05 .1 .5 1 1 2 10 20 Figure 2: Forest plot for the correlation between CD133 positive expression and cancer-specific survival (CSS) and disease-free survival (DFS). Stratified analysis by age demonstrated that CD133 positivity SOX2 positivity and PFS (three studies with 213 cases: HR = was significantly associated with shorter OS in patients aged 1.77, 95% CI = 0.82-3.80, P = 0.145) (Figure 4). more than 60 years (two studies with 238 cases: HR = 2.09, 95% CI = 1.20-3.64, P =0.009).Signicfi antdieff rencewas . Publication Bias. Publication bias was detected for OS not noted between other subgroup analyses (sample type, and DFS of CD133 positive expression. No evidence of study center design, and sample size) and CD133 positivity publication bias was noted using Egger’s test (P =0.564> (Table 2). 0.05) and Begg’s test (P =0.876> 0.05) in OS (Figure S1). Sensitivity analysis was performed by omitting an indi- Moreover, we did not find publication bias for DFS of CD133 vidual study by turn to detect the robustness of the result. The positive expression (P> 0.1) (Figure S1). result showed that two studies conducted by Yamamoto 2014 et al. [42] and Kishikawa 2016 et al. [37] in Japan significantly . . TSA. When the prespecified type I error 𝛼 (5%), a aeff cted thepooledHRvalue,withthesignicfi antHR (2.02, RRR of 20%, and a type II error 𝛽 of 20% (80% power) 95% CI = 1.56-2.60, P< 0.001) and no evidence of heterogene- were set, the TSA showed that cumulative Z curve did ity (P = 0.413). not cross the sequential monitoring boundary for CSS and OS of CD133 positive expression (Figure 5). For DFS of .. Association between SOX Positive Expression and the SOX2 positivity,cumulativeZcurvewas notmorethanthe Prognosis. SOX2 positivity was associated with worse DFS sequential monitoring boundary (Figure S2A). For positive (two studies with 157 cases: HR = 3.08, 95% CI = 1.76-5.40, P< resultsofOSofCD133 positivity amongsubgroups,the TSA 0.001) and RFS (one study with 113 cases: HR = 1.736, 95% CI = also demonstrated that cumulative Z curve did not cross the 1.055-2.901, P = 0.033), but no relationship was found between trial sequential monitoring boundary (Table 2). 6 Journal of Oncology Study % ID HR (95% CI) Weight Mehra 2006 9.73 (1.08, 87.49) 3.42 Li 2009 2.68 (1.37, 5.26) 10.90 Fusi 2011 1.10 (0.34, 3.80) 7.26 Lee 2012 2.10 (1.00, 4.38) 10.42 Qin 2012 1.43 (0.85, 2.42) 11.96 Sakai 2012 2.03 (0.94, 4.37) 10.19 Liu 2014 12.08 (1.55, 94.16) 3.78 Yamamoto 2014 0.32 (0.13, 0.81) 9.08 Kazama 2015 2.08 (0.94, 5.06) 9.66 Kishikawa 2016 0.48 (0.25, 0.90) 11.13 Pei 2016 2.08 (1.27, 3.39) 12.19 Overall (I-squared = 72.6%, p = 0.000) 1.57 (0.99, 2.51) 100.00 NOTE: Weights are from random effects analysis .05 .1 .5 1 1 2 10 20 Figure 3: Forest plot for the correlation between CD133 positive expression and overall survival (OS). When the type I error of 5%, a RRR of 20%, and a type II maybeassociatedwithpooroverall survival innonsmall- cell lung cancer [55], worse prognosis in patients with error of 10% (90% power) were used, TSA also demonstrated glioblastoma [20], and reduced overall survival in colorectal that the cumulative Z curve did not reach the sequential cancer [56]. SOX2 expression may be correlated with better monitoring boundary between CD133 positivity and CSS overallsurvivalinnonsmallcelllungcancer[21], butworse and OS (Figure 6). The TSA showed that the cumulative Z overall survival in head and neck cancer [57]. However, some curve did not cross the trial sequential monitoring boundary results were contradictory, for example, patients with CD133- between SOX2 positivity and DFS (Figure S2B). positive is correlated with a better prognosis in colorectal liver When the type I error of 5%, type II error of 20%, and a metastasis [42]. Patients with CD133-positive are associated more conservative RRR of 15% were set, the results remained with an unfavorable prognosis in advanced colorectal cancer consistent, and the TSA also showed that cumulative Z [35]. eTh conventional prognostic factors such as tumor stage curve did not reach the trial sequential monitoring boundary or grade could not well predict clinical outcome based on (Figure 7 and Figure S2C). an individual basis [58]. To date, there are still no eecti ff ve markers available for the prognosis of patients with advanced cancer. er Th efore, it remains important to better understand 4. Discussion the characteristics of CSCs, CD133, and SOX2 for valuable CSCs, a small subpopulation of tumor cells, drive the growth therapeutic and prognostic targets in clinical practice to and progression of cancers [52]. More importantly, CSCs predict disease outcomes in advanced or metastatic cancer are considered to be involved in chemotherapy/radiotherapy patients. In our meta-analysis, we have attempted to estimate resistance, metastasis, and postoperative recurrence [53, 54]. the prognostic effect of CSCs, CD133, and SOX2 using Some meta-analyses showed that CD133 was a biomarker multivariable analysis in patients with advanced or metastatic of putative CSCs in many solid tumors and its positivity cancer. Journal of Oncology 7 Table 2: Subgroup analyses of CD133 positivity in overall survival (OS). Factors Subgroups Studies HR with 95% CI Heterogeneity (P)PvalueCases TSA Tumor type Colorectal cancer 6 1.27 (0.64-2.50) < 0.001 0.493 848 Ovarian cancer 2 3.27 (0.43-25.03) 0.049 0.254 152 Melanoma 1 1.1 (0.34-3.8) NA > 0.05 32 Cancer with bone metastases 1 9.73 (1.08-87.49) NA < 0.05 50 More Gastric cancer 1 2.097 (1.003-4.383) NA < 0.05 100 More Study source Japanese 4 0.90 (0.36-2.26) 0.001 0.823 421 Chinese 4 2.12 (1.35-3.33) 0.154 0.001 579 More Others 3 2.07 (0.90-4.78) 0.229 0.089 182 Survival rate 5 years 3 1.26 (0.38-4.17) 0.001 0.703 307 < 3 years 2 10.92 (2.44-48.96) 0.888 0.002 79 More Others 6 1.38 (0.85-2.26) 0.01 0.197 796 Sample type Tissue 9 1.51 (0.92-2.49) < 0.001 0.103 1100 Blood 2 2.68 (0.33-21.83) 0.088 0.358 82 Age (years) > 60 2 2.09 (1.20-3.64) 0.989 0.009 238 More ≤ 60 3 1.40 (0.31-6.27) 0.01 0.661 149 NA 6 1.64 (0.92-2.91) 0.003 0.093 795 Treatment regimens Adjuvant therapy 4 1.91 (1.08-3.39) 0.173 0.026 309 More Surgery and adjuvant therapy 6 1.34 (0.62-2.93) < 0.001 0.457 781 Testing method Blind 3 1.64 (0.32-8.37) < 0.001 0.553 455 NA 8 1.62 (1.00-2.63) 0.005 0.048 727 More Study center design Multicentre 2 2.74 (0.46-16.19) 0.097 0.266 173 Single-center 8 1.54 (0.85-2.79) < 0.001 0.154 977 NA 1 1.1 (0.34-3.8) NA > 0.05 32 Sample size ≥ 100 6 1.58 (0.97-2.57) 0.007 0.066 891 < 100 5 1.93 (0.66-5.65) 0.001 0.232 291 HR: hazard ratio; 95% CI: 95% confidenceinterval;NA:not applicable;TSA:trialsequential analysis. Chemotherapy and radiotherapy are major treatment subgroup analyses. eTh removal of the study by Yamamoto strategies to eliminate cancer cells, but chemoresistance, 2014 [42] used blinding of the detection and the removal radioresistance, and cancer recurrence are major obstacles ofthestudy by Kishikawa2016[37]did notreportblinding for the long-term survival of cancer patients [59, 60]. Recent of the detection (Table 1). We did not n fi d that the possible studies show that CSCs are resistant to chemotherapy and factors and reasons can influence the pooled HR of OS radiotherapy and targeting CSCs may become a promising in CD133. Because these two retrospective studies [37, 42] opportunity to cure patients with cancer [54, 61]. eTh studies reported that CD133 positivity was linked to favorable OS. of 78% (14 studies) reported patients with adjuvant therapy SOX2 positivity was related to shorter DFS (HR = 3.08, P< such as chemotherapy and radiotherapy in this meta-analysis. 0.001) and RFS (HR = 1.736, P = 0.033), but SOX2 positivity According to a comprehensive analysis of published studies was not correlated with PFS (HR = 1.77, P =0.145). In (CD133: 13 studies with 1358 patients and SOX2: vfi e studies addition, no publication bias was observed in OS and DFS with 433 patients). We found that patients with CD133- of CD133. These positive results were further proven by TSA, positive advanced cancer was correlated with poorer CSS and the data suggested that additional clinical trials were (HR = 3.70, P =0.036)and showedatrend towardspoor needed to confirm these conclusions. OS (HR = 1.57, P =0.057), butnorelationshipwas reported We further performed subgroup analyses of CD133 between CD133 positivity and DFS (HR = 1.62, P = 0.178). For expression stratified by cancer type, study source, survival the analyses of CD133 in OS, we performed sensitivity and rate,sampletype,age(years),testingmethod,studycenter 8 Journal of Oncology Study % ID HR (95% CI) Weight DFS Huang 2014 2.92 (1.23, 6.93) 14.22 Sodja 2016 3.20 (1.54, 6.71) 16.06 Subtotal (I-squared = 0.0%, p = 0.871) 3.08 (1.76, 5.40) 30.28 RFS Udagawa 2015 1.74 (1.05, 2.90) 19.63 Subtotal (I-squared = .%, p = .) 1.74 (1.05, 2.88) 19.63 PFS Shen 2014 2.29 (1.01, 5.20) 14.88 Sodja 2016 1.05 (0.94, 1.18) 24.15 Yamawaki 2017 3.49 (1.14, 10.71) 11.07 Subtotal (I-squared = 73.8%, p = 0.022) 1.77 (0.82, 3.80) 50.10 Overall (I-squared = 78.4%, p = 0.000) 2.06 (1.24, 3.41) 100.00 NOTE: Weights are from random effects analysis .05 .1 .5 1 1 2 10 20 Figure 4: Forest plot for the association between SOX2 positivity and disease-free survival (DFS), relapse/recurrence-free survival (RFS), and progression-free survival (PFS). design, and sample size in OS. Subgroup analysis by cancer center design,andsamplesize. WefurtherusedTSA to type showed that CD133 expression was associated with achieve more meaningful results among different subgroups. shorterOSincancerwithbonemetastasesandgastriccancer TSA suggested that the available sample data were insucffi ient but no relationship in colorectal cancer, ovarian cancer, to draw rfi m conclusions regarding the expression of CD133 and melanoma. Stratified analysis by study source indicated to OS. that only CD133 positivity could significantly reduce OS in Our meta-analysis had some limitations. First, the num- Chinese patients (HR = 2.12, P =0.001), suggestingthat ber of the included studies was not very large and some of CD133 may play a more important role in the prognosis of these eligible studies had small sample sizes. TSA confirmed advanced cancer for Chinese. Stratified analysis by survival that cumulative Z curve did not cross the sequential monitor- rate showed that only CD133 positivity might significantly ing boundary. u Th s, more trials are needed for more reliable decrease OS in patients with< 3-year survival rate (HR = results. Second, studies were mainly conducted in China, 10.92, P = 0.002), which suggested that the expression of Japan, and Europe; thus, other study sources (USA) are lack- CD133 may be correlated with shorter OS within 3 years. ing. Third, most studies were of retrospective design; only two Subgroup analysis by age indicated that only CD133 positivity studies were of prospective design. Additional prospective can significantly shorten OS in patients aged more than 60 clinical studies (such as blinded detection of CD133 and SOX2 years (HR = 2.09, P = 0.009), suggesting that CD133 may expression) are essential to obtain more firm results in differ- play a more key role in the prognosis for elderly patients. ent cancer types, such as colorectal, lung, breast, and head- However, no significant difference was found between CD133 neck cancer. Finally, there was considerable heterogeneity in positivity and other subgroups such as sample type, study this meta-analysis. Although we analyzed several factors that Journal of Oncology 9 APIS = 2543 APIS = 2543 RRR = 20% (alpha = 5%, power = 80%) RRR = 20% (alpha = 5%, power = 80%) 8 8 Sequential monitoring boundary 6 6 Sequential monitoring boundary 4 4 2 2 Cumulative Z curve Cumulative Z curve 0 0 0 500 1000 1500 2000 2500 0 500 1000 1500 2000 2500 Information size Information size CD133: overall survival (OS), cumulative Z curve did not CD133: cancer-specific survival (CSS), cumulative Z curve cross the sequential monitoring boundary did not cross the sequential monitoring boundary Figure 5: Trial sequential analysis (TSA) for cancer-specific survival (CSS) and overall survival (OS) of CD133 positive expression ( 𝛼 =5%, 𝛽 = 20%, and the relative risk reduction (RRR) = 20%). APIS = 3404 APIS = 3404 RRR = 20% (alpha = 5%, power = 90%) RRR = 20% (alpha = 5%, power = 90%) 8 8 Sequential monitoring boundary Sequential monitoring boundary 6 6 4 4 2 2 Cumulative Z curve Cumulative Z curve 0 0 0 1000 2000 3000 4000 0 1000 2000 3000 4000 Information size Information size CD133: overall survival (OS), cumulative Z curve did not CD133: cancer-specific survival (CSS), cumulative Z curve cross the sequential monitoring boundary did not cross the sequential monitoring boundary Figure 6: Trial sequential analysis (TSA) for cancer-specific survival (CSS) and overall survival (OS) of CD133 positive expression ( 𝛼 =5%, 𝛽 = 10%, and the relative risk reduction (RRR) = 20%). may influence heterogeneity, these variables could not clearly OS. Subgroup analysis by tumor type showed that CD133 explain the sources of heterogeneity. u Th s, clinical practice positivity was linked to worse OS in cancer with bone should interpret our results with caution. metastases and gastric cancer. Subgroup analysis by study To conclude, our meta-analysis showed that CD133- source demonstrated that only CD133 positivity was related positive expression may be associated with worse CSS and to poor OS for Chinese. Subgroup analysis by survival rate 10 Journal of Oncology APIS = 4776 APIS = 4776 RRR = 15% (alpha = 5%, power = 80%) RRR = 15% (alpha = 5%, power = 80%) 8 8 Sequential monitoring boundary Sequential monitoring boundary 6 6 4 4 2 2 Cumulative Z curve Cumulative Z curve 0 0 0 1000 2000 3000 4000 5000 0 1000 2000 3000 4000 5000 Information size Information size CD133: overall survival (OS), cumulative Z curve did not CD133: cancer-specific survival (CSS), cumulative Z curve cross the sequential monitoring boundary did not cross the sequential monitoring boundary Figure 7: Trial sequential analysis (TSA) for cancer-specific survival (CSS) and overall survival (OS) of CD133 positive expression ( 𝛼 =5%, 𝛽 = 20%, and the relative risk reduction (RRR) = 15%). showed that CD133 positivity was correlated with a less and Fenggang Hou contributed to data analyses and the than 3-year OS. Subgroup analysis by age demonstrated that interpretation and completion of the gfi ures and tables. All the expression of CD133 was associated with shorter OS in authors read and approved the n fi al manuscript. patients> 60 years. SOX2 positivity may be related to poor DFS and RFS. Further TSA suggested the need for addi- Acknowledgments tional clinical studies. Herein, more high-quality prospective studiesare essentialtoobtainmorereliableevidenceand Thisresearchwassupportedbygrantsfromthe Natural help stratify advanced cancer patients who can benefit from Science Foundation of China (81473624) and the Shanghai different therapies. Science and Technology Innovation Action Plan Project (No. 16401970500-3). Data Availability Supplementary Materials ed Th ata usedtosupportthefindingsofthisstudy are included within the article. Supplementary . Table S1: REMARK guidelines. Supplementary . Table S2: Detailed characteristics of studies Disclosure included in the meta-analysis. SusuHanand TaoHuangareco-rfi st authorsofthisstudy. Supplementary . Figure S1: Publication bias using Egger’s and Begg’s tests for overall survival (OS) and disease-free Conflicts of Interest survival (DFS) of CD133 positive expression. Supplementary . Figure S2: Trial sequential analysis (TSA) The authors declare no conflicts of interest. for disease-free survival (DFS) of SOX2 positivity. Authors’ Contributions References Susu Han, Tao Huang, and Fenggang Hou contributed to [1] L. A. Torre, F. Bray, R. L. Siegel, J. Ferlay, and J. Lortet- the conception and design of this research. Susu Han, Xing Tieulent, “Global cancer statistics, 2012,” CA: A Cancer Journal Wu, Xiyu Wang, Shanshan Liu, Wei Yang, and Qi Shi for Clinicians,vol.65,no.2,pp. 87–108,2015. contributed to the drafting of the article and final approval of the submitted version. Susu Han, Tao Huang, Xing Wu, [2] A.Urruticoechea,R.Alemany,J.Balart,A. Villanueva,F. Xiyu Wang, Shanshan Liu, Wei Yang, Qi Shi, Hongjia Li, Vinals, ˜ and G. Capella, ´ “Recent advances in cancer therapy: An Journal of Oncology 11 overview,” Current Pharmaceutical Design,vol.16, no.1,pp. 3– [21] Y.Chen,Y.Huang,Y.Huang etal.,“eP Th rognosticValue of 10, 2010. SOX2 Expression in Non-Small Cell Lung Cancer: A Meta- Analysis,” PLoS ONE,vol.8,no.8, p.e71140,2013. [3] S. G. C. Kroeze, C. Fritz, M. Hoyer et al., “Toxicity of concurrent stereotactic radiotherapy and targeted therapy or immunother- [22] A. Liberati, D. G. Altman, J. Tetzlaff et al., “The PRISMA apy:Asystematicreview,” Cancer Treatment Reviews,vol.53,pp. statement for reporting systematic reviews and meta-analyses of 25–37, 2017. studies that evaluate health care interventions: explanation and elaboration,” PLoS Medicine, vol. 6, no. 7, Article ID e1000100, [4] R. L. Siegel, L. Sahar, K. M. Portier, E. M. Ward, and A. Jemal, “Cancer death rates in US congressional districts,” CA: A Cancer Journal for Clinicians,vol.65, no.5,pp.339–344, 2015. [23] D. G. Altman and J. M. Bland, “How to obtain the confidence interval from a P value,” British Medical Journal,vol.343,Article [5] D. M. Pardoll, “The blockade of immune checkpoints in cancer ID d2090, 2011. immunotherapy,” Nature Reviews Cancer,vol.12,no.4,pp. 252– 264, 2012. [24] J. F. Tierney, L. A. Stewart, D. Ghersi, S. Burdett, and M. R. Sydes, “Practical methods for incorporating summary time-to- [6] M.R.Doherty,J.M.Smigiel, D.J. Junk,and M. W. Jackson, event data into meta-analysis,” Trials,vol.8,article 16,2007. “Cancer stem cell plasticity drives therapeutic resistance,” Can- cers,vol.8,no. 1,pp.1–13, 2016. [25] L.M.McShane, D. G.Altman,W.Sauerbrei,S.E.Taube, M. Gion, and G. M. Clark, “Reporting Recommendations for [7] M. Derwahl, “Linking stem cells to thyroid cancer,” e Journal Tumor Marker Prognostic Studies (REMARK),” JNCI: Journal of Clinical Endocrinology & Metabolism,vol.96,no.3,pp.610– of the National Cancer Institute,vol.97, no.16, pp.1180–1184, 613, 2011. [8] S. Han, W. Yang, S. Zong et al., “Clinicopathological, prognostic [26] E. Zintzaras and J. P. A. Ioannidis, “HEGESMA: genome search and predictive value of CD166 expression in colorectal cancer: A meta-analysis,” Oncotarget ,vol.8,no. 38,pp. 64373–64384, meta-analysis and heterogeneity testing,” Bioinformatics,vol.21, no.18, pp.3672-3673,2005. [27] E. Evangelou and J. P. A. Ioannidis, “Meta-analysis methods for [9] S.Han,S.Zong, Q.Shietal.,“Is Ep-CAM Expression a Diagnostic and Prognostic Biomarker for Colorectal Cancer? genome-wide association studies and beyond,” Nature Reviews Genetics,vol.14,no.6,pp. 379–389, 2013. A Systematic Meta-Analysis,” EBioMedicine,vol.20,pp.61–69, [28] R. DerSimonian and R. Kacker, “Random-eeff cts model for meta-analysis of clinical trials: an update,” Contemporary Clini- [10] A. R. Safa, “Resistance to cell death and its modulation in cancer stem cells,” Critical Reviews in Oncogenesis,vol.21,no.3-4,pp. cal Trials, vol. 28, no. 2, pp. 105–114, 2007. 203–219, 2016. [29] M. Egger, G. D. Smith, M. Schneider, and C. Minder, “Bias [11] D. Corbeil, C. A. Fargeas, and W. B. Huttner, “Rat prominin, like in meta-analysis detected by a simple, graphical test,” British its mouse and human orthologues, is a pentaspan membrane Medical Journal, vol. 315, pp. 629–634, 1997. glycoprotein,” Biochemical and Biophysical Research Communi- [30] C. B. Begg and M. Mazumdar, “Operating characteristics of a cations,vol.285,no.4, pp.939–944,2001. rank correlation test for publication bias,” Biometrics,vol.50, [12] A. H. Yin, S. Miraglia, E. D. Zanjani et al., “AC133, a novel no. 4, pp. 1088–1101, 1994. marker for human hematopoietic stem and progenitor cells,” [31] B. Miladinovic, R. Mhaskar, I. Hozo, A. Kumar, H. Mahony, and Blood,vol.90, no.12, pp.5002–5012,1997. B. Djulbegovic, “Optimal information size in trial sequential [13] S.Masui,Y.Nakatake, Y. Toyookaetal.,“Pluripotencygoverned analysis of time-to-event outcomes reveals potentially incon- by Sox via regulation of Oct / expression in mouse embryonic clusive results because of the risk of random error,” Journal of stem cells,” Nature Cell Biology,vol.9,no.6, pp.625–635,2007. Clinical Epidemiology,vol.66, no.6,pp.654–659,2013. [14] J.-X. Cheng, B.-L. Liu, and X. Zhang, “How powerful is CD133 [32] K. o Th rlund, P. J. Devereaux, J. Wetterslev et al., “Can trial as a cancer stem cell marker in brain tumors?” Cancer Treatment sequential monitoring boundaries reduce spurious inferences Reviews,vol.35, no.5,pp. 403–408,2009. from meta-analyses?” International Journal of Epidemiology, vol. 38, no. 1, pp. 276–286, 2009. [15] N. R. Murray, V. Justilien, and A. P. Fields, “SOX2 Determines Lineage Restriction: Modeling Lung Squamous Cell Carcinoma [33] J. Brok, K. or Th lund, C. Gluud, and J. Wetterslev, “Trial in the Mouse,” Cancer Cell,vol.30, no.4,pp.505–507, 2016. sequential analysis reveals insufficient information size and potentially false positive results in many meta-analyses,” Journal [16] S. B. Keysar, P. N. Le, B. Miller et al., “Regulation of head and of Clinical Epidemiology,vol.61, no.8,pp.763–769, 2008. neck squamous cancer stem cells by PI3K and SOX2,” Journal of the National Cancer Institute,vol.109,no. 1,2017. [34] K. Yamawaki, T. Ishiguro, Y. Mori et al., “Sox2-dependent inhi- bition of p21 is associated with poor prognosis of endometrial [17] S. Boumahdi,G.Driessens,G.Lapouge etal., “SOX2controls tumour initiation and cancer stem-cell functions in squamous- cancer,” Cancer Science,vol.108,no.4, pp.632–640,2017. cell carcinoma,” Nature,vol.511, no.7508, pp.246–250,2014. [35] Q. Pei, H. Zhu, F. Tan et al., “Intravascular emboli is an [18] Z.Fan,M.Li,X.Chenetal.,“Prognostic ValueofCancerStem independent risk factor for the prognosis of stage III colorectal cancer patients aer ft radical surgery,” Oncotarget ,vol.7,no. 35, Cell Markers in Head and Neck Squamous Cell Carcinoma: a Meta-analysis,” Scientific Reports,vol.7,no.1,2017. pp. 57268–57276, 2016. [19] B. Cheng, G. Yang, R. Jiang et al., “Cancer stem cell markers [36] E. Sodja, M. Rijavec, A. Koren, A. Sadikov, P. Koroˇsec, and T. predict a poor prognosis in renal cell carcinoma: A meta- Cufer,“ep Th rognosticvalue ofwholebloodSOX2,NANOG analysis,” Oncotarget ,vol.7,no. 40,pp. 65862–65875, 2016. and OCT4 mRNA expression in advanced small-cell lung cancer,” Radiology and Oncology, vol. 50, no. 2, pp. 188–196, [20] W.Zhang,H.Chen,S.Lv,andH.Yang,“HighCD133Expression Is Associated with Worse Prognosis in Patients with Glioblas- toma,” Molecular Neurobiology,vol.53,no.4,pp. 2354–2360, [37] J. Kishikawa, S. Kazama, K. Oba et al., “CD133 Expression at 2016. theMetastaticSitePredicts Patients’OutcomeinColorectal 12 Journal of Oncology Cancer with Synchronous Liver Metastasis,” Annals of Surgical [52] N.A.Lobo,Y.Shimono,D.Qian,andM. F. Clarke,“eTh biology Oncology,vol.23, no.6,pp.1916–1923,2016. of cancer stem cells,” Annual Review of Cell and Developmental Biology,vol.23,no.1,pp.675–699,2007. [38] S. Kazama, J. Kishikawa, K. Yasuda et al., “CD133 Expression in Lymph Node Metastases Is Associated with Tumor Aggres- [53] K. M. Govaert, B. L. Emmink, M. W. Nijkamp et al., “Hypoxia aer ft liver surgery imposes an aggressive cancer stem cell siveness During Lymph Node Metastasis in Colorectal Cancer,” Anticancer Reseach, vol. 35, no. 12, pp. 6599–6605, 2015. phenotype on residual tumor cells,” Annals of Surgery,vol.259, no. 4, pp. 750–759, 2014. [39] H. Udagawa, G. Ishii, M. Morise et al., “Comparison of the [54] R. Morrison, S. M. Schleicher, Y. Sun et al., “Targeting the expression levels of molecular markers among the peripheral Mechanisms of Resistance to Chemotherapy and Radiotherapy area and central area of primary tumor and metastatic lymph with the Cancer Stem Cell Hypothesis,”JournalofOncology,vol. node tumor in patients with squamous cell carcinoma of the 2011, Article ID 941876, 13 pages, 2011. lung,” Journal of Cancer Research and Clinical Oncology,vol.141, no. 8, pp. 1417–1425, 2015. [55] E. Chen, Z. Zeng, B. Bai, J. Zhu, and Z. Song, “eTh prognostic value of CSCs biomarker CD133 in NSCLC: A meta-analysis,” [40] B.-L. Liu, S.-J. Liu, A. Baskys et al., “Platinum sensitivity and Oncotarget ,vol.7,no. 35,pp. 56526–56539, 2016. CD133 expression as risk and prognostic predictors of central nervoussystemmetastasesinpatientswithepithelialovarian [56] K. Wang, J. Xu, J. Zhang, and J. Huang, “Prognostic role of cancer,” BMC Cancer,vol.14,no.1,2014. CD133expressionincolorectalcancer:ameta-analysis,” BMC Cancer,vol.12,no.1,2012. [41] L.Shen,X.Huang,X.Xie,J.Su,J. Yuan,andX. Chen,“High Expression of SOX2 and OCT4 Indicates Radiation Resistance [57] W.Li,B.Li, R.Wang,D.Huang,W.Jin,andS.Yang,“SOX2as and an Independent Negative Prognosis in Cervical Squamous prognostic factor in head and neck cancer: A systematic review Cell Carcinoma,” Journal of Histochemistry & Cytochemistry, and meta-analysis,” Acta Oto-Laryngologica,vol.134,no.11, pp. vol.62,no. 7,pp.499–509,2014. 1101–1108, 2014. [42] S. Yamamoto, K. Tanaka, K. Takeda et al., “Patients with CD133- [58] A. Bamias, M. Chorti, C. Deliveliotis et al., “Prognostic signif- negative colorectal liver metastasis have a poor prognosis aer ft icance of CA 125, CD44, and epithelial membrane antigen in hepatectomy,” Annals of Surgical Oncology,vol.21,no.6,pp. renalcellcarcinoma,” Urology,vol.62, no.2,pp. 368–373, 2003. 1853–1861, 2014. [59] K. Nurgali, R. T. Jagoe, and R. Abalo, “Editorial: Adverse effects [43] Y.-H. Huang, M.-H. Luo, Y.-B. Ni et al., “Increased SOX2 of cancer chemotherapy: Anything new to improve tolerance expression in less differentiated breast carcinomas and their and reduce sequelae?” Frontiers in Pharmacology,vol.9,2018. lymph node metastases,” Histopathology,vol.64, no.4,pp. 494– [60] G. Delaney, S. Jacob, C. Featherstone, and M. Barton, “The 503, 2014. role of radiotherapy in cancer treatment: estimating optimal [44] T. Sprenger, L.-C. Conradi, T. Beissbarth et al., “Enrichment utilization from a review of evidence-based clinical guidelines,” of CD133-expressing cells in rectal cancers treated with pre- Cancer, vol. 104, no. 6, pp. 1129–1137, 2005. operative radiochemotherapy is an independent marker for [61] P. S. Kharkar, “Cancer stem cell (CSC) inhibitors: a review metastasis and survival,” Cancer,vol.119,no.1,pp.26–35,2013. of recent patents (2012-2015),” Expert Opinion on erapeutic [45] H. H. Lee, K. J. Seo, C. H. An, J. S. Kim, and H. M. Jeon, Patents,vol.27,no.7,pp. 753–761, 2017. “CD133 expression is correlated with chemoresistance and early recurrence of gastric cancer,” Journal of Surgical Oncology,vol. 106, no.8,pp. 999–1004, 2012. [46] Q. Qin, Y. Sun, M. Fei et al., “Expression of putative stem marker nestin and CD133 in advanced serous ovarian cancer,” Neoplasma,vol.59,no.3,pp.310–315,2012. [47] N. Sakai, H. Yoshidome, T. Shida et al., “CXCR4/CXCL12 expression profile is associated with tumor microenvironment and clinical outcome of liver metastases of colorectal cancer,” Clinical & Experimental Metastasis,vol.29, no.2,pp. 101–110, [48] P. Pilati, S. Mocellin, L. Bertazza et al., “Prognostic value of putative circulating cancer stem cells in patients undergoing hepatic resection for colorectal liver metastasis,” Annals of Surgical Oncology,vol.19, no.2,pp.402–408,2012. [49] A. Fusi, U. Reichelt, A. Busse et al., “Expression of the stem cell markers nestin and CD133 on circulating melanoma cells,” Journal of Investigative Dermatology,vol.131,no.2,pp.487–494, [50] C. Li, B. Li, Y. Liang et al., “Higher percentage of CD133+ cells is associated with poor prognosis in colon carcinoma patients with stage IIIB,” Journal of Translational Medicine,vol.7,no. 1, p. 56, 2009. [51] N. Mehra, M. Penning, J. Maas et al., “Progenitor marker CD133 mRNA is elevated in peripheral blood of cancer patients with bone metastases,” Clinical Cancer Research,vol.12,no.16, pp. 4859–4866, 2006. MEDIATORS of INFLAMMATION The Scientific Gastroenterology Journal of World Journal Research and Practice Diabetes Research Disease Markers Hindawi Hindawi Publishing Corporation Hindawi Hindawi Hindawi Hindawi www.hindawi.com Volume 2018 http://www www.hindawi.com .hindawi.com V Volume 2018 olume 2013 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 International Journal of Journal of Immunology Research Endocrinology Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 Submit your manuscripts at www.hindawi.com BioMed PPAR Research Research International Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 Journal of Obesity Evidence-Based Journal of Journal of Stem Cells Complementary and Ophthalmology International Alternative Medicine Oncology Hindawi Hindawi Hindawi Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2013 Parkinson’s Disease Computational and Behavioural Mathematical Methods AIDS Oxidative Medicine and in Medicine Neurology Research and Treatment Cellular Longevity Hindawi Hindawi Hindawi Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018

Journal

Journal of OncologyHindawi Publishing Corporation

Published: Jan 1, 2019

References