- Email: [email protected]
Long non-coding RNA CRNDE as a potential prognostic biomarker in solid tumors: A meta-analysis
Clinica Chimica Acta 481 (2018) 99–107
Contents lists available at ScienceDirect
Clinica Chimica Acta journal homepage: www.elsevier.com/locate/cca
Review
Long non-coding RNA CRNDE as a potential prognostic biomarker in solid tumors: A meta-analysis ⁎
Chaojie Lianga,1, Bing Zhangb,1, Hua Gea, Yingchen Xua, Guangming Lia, , Jixiang Wua, a b
T
⁎
Department of General Surgery, Beijing Tongren Hospital, Capital Medical University, Beijing 100730, China Department of Tumor Center, Beijing Tongren Hospital, Capital Medical University, Beijing 100730, China
A R T I C L E I N F O
A B S T R A C T
Keywords: LncRNA CRNDE Neoplasm Prognosis Meta-analysis
Background and aim: Long non-coding RNA colorectal neoplasia differentially expressed (CRNDE) has been demonstrated to be highly expressed in many malignant tumors; however, the role of CRNDE in cancer remains undetermined because of limitations in sample size. We conducted a meta-analysis to assess the role of CRNDE in cancer. Methods: PubMed, Medline, Cochrane Library, Web of Science, EMBASE database, Ovid, Chinese CNKI, and Chinese WanFang database were systematically searched. The relation between CRNDE and the clinicopathological characteristics and prognosis of patients with cancer was determined using pooled odds ratios (ORs) and hazard ratios (HRs) with 95% confidence interval (CI). Results: Thirteen studies with 1570 patients were included. The pooled results indicated that high CRNDE expression was related to lymph node metastasis (YES vs. NO: OR = 3.50, 95% CI = [1.05, 8.09]) and TNM stage (I + II vs. III + IV: OR = 0.26, 95% CI = [0.18, 0.37]) but not to gender, tumor size, and differentiation. High CRNDE expression indicated poor overall survival (OS) (HR = 2.06, 95% CI = [1.66, 2.47]). CRNDE could be an independent predictive factor for OS (HR = 1.62, 95% CI = [1.15, 2.08]) in patients with cancer. Conclusion: Therefore, high CRNDE expression was associated with advanced clinicopathological characteristics, and CRNDE could be used as a reliable prognostic biomarker in human cancer. However, more high-quality studies with a large sample size are needed to support this meta-analysis.
1. Introduction In 2015, nearly 4 million cases and 3 million recent deaths in China were attributed to cancer, one of the most deadly diseases worldwide [1]. In the United States, about 1,700,000 new cancer patients and 600,000 cancer-related deaths were reported, according to the American Cancer Society [2]. Despite technological advances and scientific research in cancer, the clinical outcome of cancer remains poor. One of the main reasons is the lack of effective biomarkers to diagnose cancer at an early stage, rendering necessary the identification of potential diagnostic and prognostic biomarkers for cancer patients [3–5]. Long non-coding RNAs (LncRNA), which were initially considered as junk or transcriptional noise, measuring > 200 nucleotides in length. Accumulating evidence indicates that lncRNAs play a vital role in the progression and development of cancer [6,7] and participate in numerous biological processes, including the regulation of the proliferation, invasion, and metastasis of cancer. Studies also indicate that lncRNAs can be used as prognostic biomarkers for cancer [8]. ⁎
1
Long non-coding RNA colorectal neoplasia differentially expressed (CRNDE), which is located on chromosome 16, was initially identified as an lncRNA in colorectal cancer (CRC) and reported to be highly elevated in CRC [9]. In further studies, CRNDE was also elevated in many malignant tumors, such as gastric cancer (GC) [10], papillary thyroid cancer [11], cervical cancer [12], and pancreatic cancer (PC) [13]. Other studies indicated that the expression of lncRNA CRNDE was correlated with the clinicopathological features and prognosis of patients with cancer [14,15]. Huan et al. [16] found that high CRNDE expression was associated with tumor size and TNM stage but not with age and histological grade in breast cancer (BC). Ding et al. [17] demonstrated that CRNDE upregulation in CRC was related to tumor size and tumor stage but not correlated with the depth of tumor, age, and gender. Du et al. [10] reported that high CRNDE expression in GC was related to lymph node metastasis, depth of invasion, and TNM stage but not to tumor size and differentiation. Wang et al. [13] indicated that high CRNDE expression in PADC was associated with TNM stage, tumor size, differentiation, and lymph node metastasis. However, most studies on CRNDE are limited by
Corresponding authors at: Department of General Surgery, Beijing Tongren Hospital, Capital Medical University, No. 2, Chongwenmennei Street, Dongcheng, Beijing 100730, China. E-mail addresses: [email protected] (G. Li), [email protected] (J. Wu). Are equal to this study.
https://doi.org/10.1016/j.cca.2018.02.039 Received 17 January 2018; Received in revised form 28 February 2018; Accepted 28 February 2018 Available online 01 March 2018 0009-8981/ © 2018 Elsevier B.V. All rights reserved.
Clinica Chimica Acta 481 (2018) 99–107
C. Liang et al.
the sample size and discrete outcome, and no meta-analysis has thus far been conducted to assess the role of CRNDE in cancer. Thus, we conducted this meta-analysis to explore the relationship between CRNDE and clinicopathological features and prognosis in cancer.
The exclusion criteria were as follows: (1) Reviews, case reports, letters, editorials, meta-analysis, and conference reports; (2) studies focused on the molecular structure and function of CRNDE; and (3) non-human studies and duplicate publications.
2. Materials and methods
2.2. Data extraction and quality assessment
2.1. Search strategy and study selection
Two researchers (Chaojie Liang and Bing Zhang) searched and assessed the literature independently in accordance with the criteria. The extracted data included the name of the first author, the year of publication, country/region, number of cases, clinicopathological features, and overall survival (OS). The quality of the literature was assessed using the Newcastle–Ottawa–Scale (NOS) criteria. Articles with a NOS score ≥ 6 were considered high-quality; otherwise, they were regarded as low-quality.
The studies included in this meta-analysis were searched in PubMed, Medline, Cochrane Library, Web of Science, the EMBASE database, Ovid, Chinese CNKI, and the Chinese WanFang database. The research ended on January 10, 2018, and was not limited by inadequate data. The key words used for the searches were as follows: “CRNDE” or “colorectal neoplasia differentially expressed” or “lncRNA CRNDE” and “cancers” or “neoplasm.” The inclusion criteria were as follows: (1) LncRNA CRNDE expression was detected in human solid cancer tissues, and patients were grouped according to lncRNA CRNDE expression; (2) The relationship between lncRNA CRNDE expression and clinicopathological features and prognosis was described; and (3) The hazard ratios (HRs) with 95% confidence interval (CI) were provided or could be calculated using the survival curve.
2.3. Statistical analysis All statistical data were calculated using STATA 14.2. The HRs with 95% confidence intervals (CIs) were extracted from a Kaplan–Meier curve by using Engauge Digitizer 10.0. The association between CRNDE expression and clinicopathological characteristics—gender (male vs. female), tumor size (> 5 cm vs. ≤5 cm), differentiation (low vs. high
Fig. 1. Flow diagram of study selection.
100
Clinica Chimica Acta 481 (2018) 99–107
C. Liang et al.
3. Results
and moderate), lymph node metastasis (yes vs. no), distant metastasis (YES vs. NO), and TNM stage (I + II vs. III + IV)—were assessed by pooled odds ratios (ORs) with 95% CIs. The relationship between high CRNDE expression and OS was calculated using HRs with 95% CI. I2 and Q tests were performed to calculate heterogeneity when I2 > 50%. Either the random effects model or the fixed effects model was used. If the pooled OR HR with 95%CI did not overlap 1, the pooled results were significant. The source of heterogeneity and the stability of results were assessed by sensitivity analysis and subgroup analysis, and Begg's test was used to assess the potential publication bias.
3.1. Study identification and characteristics Figs. 1 and 2 show that 1570 patients and 13 studies (2 glioma cancer, 1 ovarian cancer, 1 cervical cancer, 4 CRC, 1 BC, 1 PC, 1 lung adenocarcinoma, and 1 non-small lung cancer) [10,13,16–26] were included in this meta-analysis in accordance with the criteria. The study samples ranged from 58 to 251, and the studies were published from 2015 to 2017. A total of 10 studies were published in English and 2 in Chinese; most
Fig. 2. Forest plot of studies evaluating the relationship between CRNDE expression and clinicopathological features. A. gender; B. tumor size; C. differentiation; D. lymph node metastasis; E. TNM stage.
101
Clinica Chimica Acta 481 (2018) 99–107
C. Liang et al.
Table 1 Characteristics of studies included in the meta-analysis. Study
Year
Country
Cancer type
Detection method
Sample size
LncRNA CRNDE expression High
Lukasz Han Y Jing SY Han P Liu T Huan JL Ding J Du DX Jiang HJ Wang G Kiang Zhang M Li XH
2015 2015 2016 2017 2016 2017 2017 2017 2017 2017 2017 2017 2017
Poland China China China China China China China China China China China China
OC CC Glioma CRC CRC BC CRC GC CRC PC Glioma LAD NSCLC
RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR
135 87 164 64 142 103 80 118 251 58 165 66 137
Survival information
NOS score
OS OS OS OS OS OS NA OS OS OS OS OS OS
8 6 8 8 8 6 6 7 7 6 6 6 7
Low
Total
LNM
HTS
LD
TS > 5
M
Total
LNM
HTS
LD
TS > 5
M
118 51 83 32 71 44 40 61 120 38 83 36 83
NA 39 NA 24 NA NA NA 35 53 26 NA 28 45
NA NA 12 NA NA 26 17 24 NA 15 NA 13 NA
NA 25 NA 12 NA NA NA 39 19 20 NA 12 40
NA 32 NA 24 NA NA 25 24 77 NA NA NA NA
NA NA 55 23 NA NA 25 21 73 24 NA 23 40
17 36 81 32 71 59 40 57 131 20 82 30 66
NA 12 NA 8 NA NA NA 26 67 12 NA 8 32
NA NA 71 NA NA 18 23 37 NA 23 NA 23 NA
NA 26 NA 20 NA NA NA 22 101 18 NA 24 37
NA 19 NA 8 NA NA 15 37 43 NA NA NA NA
NA NA 28 16 NA NA 15 40 47 14 NA 13 43
LNM: lymph node metastasis, DM: distant metastasis, HTS: high TNM stage (III/IV), T1/2: depth of infiltration (T1/2), LD: low differentiation. TS > 5: tumor size > 5 cm, M: male. Table 2 Survival data of studies included in the meta-analysis. Study
Year
Country
Cancer type
Detection method
Sample size
Survival information
HR statistic
NOS score
Data in paper Data in paper Data in paper/survival curve Data in paper Data in paper Survival curve Survival curve Data in paper Data in paper Survival curve Survival curve Survival curve Survival curve Data in paper/survival curve
8 8 6 8 8 6 6 7 8 6 6 6 6 7
Univariate analysis Multivariate analysis Lukasz1 Lukasz2 Han Y Jing SY Liu T1 Han P Huang JL Du DX Jiang HJ Kiang Kiang Wang G Zhang M Li XH
2015 2015 2015 2016 2016 2017 2017 2017 2017 2017 2017 2017 2017 2017
Poland Poland China China China China China China China China China China China China
OC OC CC Glioma CRC CRC BC GC CRC Glioma Glioma PC LAD NSCLC
RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR
103 135 87 164 142 64 103 118 251 101 64 58 66 137
NA NA 2.239(1.095–4.593)(E) 2.236(1.452–7.125) 2.906(1.452–7.125) 1.850(1.032–6.558)(E) 1.770(1.011–3.768)(E) 2.663(1.458–5.532) 2.062(1.291–3.173) 1.780(1.252–2.532)(E) 2.491(1.018–6.093)(E) 1.676(1.011–6.509)(E) 2.516(1.319–4.801)(E) 3.411(1.517–7.670)(E)
6.072(1.814–20.32) 5.437(1.699–17.4) 4.702(3.542–12.618) 1.589(1.034–6.873) 1.589(1.034–3.684) NA NA 2.453(1.321–4.777) 1.693(1.047–2.738) NA NA NA NA 4.032(2.541–10.350)
OC: ovarian cancer; CC: cervical cancer; CRC: colorectal cancer; BC: breast cancer; GC: gastric cancer; PADC: pancreatic cancer; LAD: lung adenocarcinoma; NSCLC: non-small lung cancer. Table 3 LncRNA CRNDE clinicopathological features for cancers. Heterogeneity Clinicopathological features
No. of studies
No. of patients
Pooled OR (95% CI)
PHet
I2 (%)
p value
Model used
Gender Tumor size Differentiation Lymph node metastasis TNM stage
8 5 8 7 6
950 576 884 791 588
1.07 1.96 1.24 3.50 0.26
0.349 0.007 < 0.001 0.001 0.799
10.5 71.4 76.6 74.7 0.0
0.366 0.053 0.530 < 0.001 < 0.001
Fixed Random Random Random Fixed
[0.83, [0.99, [0.64, [1.87, [0.18,
1.39] 3.86] 2.40] 6.56] 0.37]
Random, random-effects model; Fixed, fixed-effects model; OR, odds ratio; CI, confidence interval.
gender (male: female, OR = 0.87, 95% CI = [0.66, 1.17], p = 0.366, Fig. 2A), and 5 studies with 576 patients demonstrated that CRNDE upregulation was not associated with tumor size (> 5 cm vs. ≤5 cm: OR = 1.96, 95% CI = [0.99, 3.86], p = 0.053, Fig. 2B). Similarly, the pooled results from 8 studies with 884 patients showed that CRNDE expression was not correlated with differentiation (low: moderate + high, OR = 1.24, 95% CI = [0.64, 2.40], p = 0.530, Fig. 2C). However, the pooled results suggested that high CRNDE expression was associated with lymph node metastasis (YES:NO, OR = 3.50, 95% CI = [1.87, 6.56], p < 0.001, Fig. 2D) and advanced TNM stage (I + II: III + IV, OR = 0.26, 95% CI = [0.18, 0.37], p < 0.001,
studies were from China, and 1 study was from Poland. All studies scored ≥6 according to the NOS score criteria and thus were considered highquality. The clinicopathological characteristics and survival information obtained from the studies are summarized in Tables 1 and 2. 3.2. Association between CRNDE and clinicopathological characteristics As shown in Table 3, the pooled ORs with 95% CI were calculated to assess the relationship between CRNDE and clinicopathological features in patients with cancer. The pooled results from 8 studies with 950 patients indicated that the high CRNDE expression was not related to 102
Clinica Chimica Acta 481 (2018) 99–107
C. Liang et al.
Table 4 Subgroup analysis of overall survival by tumor type, NOS score, sample size. Subgroups Tumor type Female reproductive system Nervous system Digestive system Respiratory system Sample size ≤110 > 110 NOS score ≤7 >7
No. of studies
No. of patients
Pooled HR (95% CI)
PHet
I2 (%)
p value
2 3 5 2
190 333 629 203
1.95 1.84 2.26 2.74
3.03) 2.45) 2.95) 4.25)
0.679 0.834 0.841 0.622
0.0 0.0 0.0 0.0
0.12 < 0.001 < 0.001 < 0.001
6 6
547 908
2.17 (1.24, 3.10) 2.26 (1.64, 2.88)
0.963 0.823
0.0 0.0
< 0.001 < 0.001
9 3
798 557
1.98 (1.50, 2.46) 2.28 (1.52, 3.04)
0.986 0.604
0.0 0.0
< 0.001 < 0.001
(0.87, (1.24, (1.57, (1.22,
Abbreviations: OR = hazard ratio, CI = confidence interval, n = number of sample size.
Fig. 3. Begg's publication bias plots for CRNDE-related studies: A. gender; B. tumor size; C. differentiation; D. lymph node metastasis; E. TNM stage.
103
Clinica Chimica Acta 481 (2018) 99–107
C. Liang et al.
Fig. 4. Forest plots and Begg's publication bias plots of studies evaluating the relation between CRNDE expression and overall survival (OS) rate (A) and independent predictive factor for OS (B).
low: HR = 2.28, 95% CI = [1.52, 3.04], p < 0.001, Fig. 5C) (n ≤ 7: high vs. low: HR = 1.98, 95% CI = [1.50, 2.46], p < 0.001, Fig. 5C). The robustness of the pooled results was evaluated by sensitivity analysis, and the results (Fig. 6A) were not significantly affected when any individual study was removed, which indicated that the results were reliable.
Fig. 2E). Therefore, this meta-analysis indicated that the high CRNDE expression was correlated with advanced clinicopathological features. 3.3. Prognostic value of CRNDE for OS of cancer The prognostic value of CRNDE for OS of cancer included 11 studies (Kiang conducted 2 trials) and 1355 patients. The pooled results indicated that high CRNDE expression was correlated with poor OS (high vs. low: HR = 2.06, 95% CI = [1.66, 2.47], p < 0.001, Fig. 4A). We then conducted a subgroup analysis for OS according to cancer type, NOS score, and sample size. As shown in Table 4, the subgroup for cancer type indicated that high CRNDE expression was a strong prognostic biomarker in cancer of the nervous system (high vs. low: HR = 1.84, 95% CI = [1.24, 2.45], p < 0.001, Fig. 5A), digestive system (high vs. low: HR = 2.26, 95% CI = [1.57, 2.95], p < 0.001, Fig. 5A), and respiratory system (high vs. low: HR = 2.74, 95% CI = [1.22, 4.25], p < 0.001, Fig. 5A). However, the CRNDE expression was not associated with OS in female reproductive cancer (high vs. low: HR = 1.95, 95% CI = [0.87, 3.03], Fig. 5A). We conducted a subgroup analysis according to sample size, and the threshold was 110 patients. The pooled results showed that CRNDE upregulation was correlated with poor OS in both groups (n ≥ 100: high vs. low: HR = 2.26, 95% CI = [1.64, 2.88], p < 0.001, Fig. 5B) (n < 100: high vs. low: HR = 1.92, 95% CI = [1.39, 2.46], p < 0.001, Fig. 5B). We also conducted a subgroup analysis according to NOS score; similarly, the pooled results suggested that high CRNDE expression was significantly correlated with poor OS in both groups (n > 7: high vs.
3.4. Independent prognostic value of CRNDE in cancer The independent prognostic value of CRNDE in cancer were calculated based on the multivariate analysis in 7 studies (Lukasz conducted 2 trials) with 1342 patients. The pooled HRs showed that CRNDE expression was an independent prognostic factor for OS in patients with cancer (HR = 1.62, 95% CI = [1.15, 2.08], I2 = 0, Fig. 4B), and the results were also reliable according to the sensitivity analysis (Fig. 6B). 3.5. Begg's funnel plot analysis Begg's funnel plot was conducted to evaluate publication bias. Figs. 3 and 4 show no publication bias for gender (p = 0.536), tumor size (p = 0.086), differentiation (p = 0.108), TNM stage (p = 0.268), overall survival (p = 1), and independent factor for OS (p = 0.266), However, the publication bias existed for lymph node metastasis (p = 0.016) and sensitivity analysis was conducted to determine whether individual studies influenced the pooled OR. And the results showed no study substantially influenced the pooled OR, which suggested that more studies needed to be included for further research. 104
Clinica Chimica Acta 481 (2018) 99–107
C. Liang et al.
Fig. 5. Forest plots of subgroup analysis for OS: subgroup analysis by tumor type (A), subgroup analysis by sample size (B), and subgroup analysis by Newcastle–Ottawa–Scale score (C).
Fig. 6. Sensitivity analysis for overall survival (OS) (A) and independent predictive factor for OS (B).
105
Clinica Chimica Acta 481 (2018) 99–107
C. Liang et al.
4. Discussion
CRNDE overexpression is significantly associated with increased lymph node metastasis and advanced TNM stage; moreover, high CRNDE expression was related to poor OS and the expression of CRNDE could be an independent factor for OS in patients with cancer. Thus, lncRNA CRNDE could be used as a potential prognostic biomarker for patients with cancer. However, more studies with a large sample size and high quality need to be conducted to elucidate the role of lncRNA CRNDE.
An increasing number of studies demonstrated that the abnormal expression of lncRNAs was associated with the development and progression of cancer and significantly related to clinicopathological features and prognosis. CRNDE was initially identified as an lncRNA in CRC. Further studies also found that CRNDE was overexpressed in several types of tumors and played an oncogenic role in tumors. In gliomas [27], CRNDE overexpression promoted cell growth and migration in vitro and in vivo by regulating the mTOR signal pathway. Zheng et al. [28] demonstrated that the upregulation of CRNDE expression could promote the proliferation, migration, and invasion of glioma stem cells by regulating miR-186 expression. In addition, miR186 could bind to XIAP and the PAK7 3′UTR region to regulate the expression of caspase3, BAD, cyclin D1, and MARK2. In addition, Zheng et al. [29] suggested that CRNDE could play an oncogenic role by attenuating the miR-384/PIWL4/STAT3 axis. Kiang [23] confirmed that CRNDE depletion inhibited glioma growth by activating EGFR signaling. In gallbladder cancer, Shen et al. [30] indicated that CRNDE promoted the carcinogenesis of gallbladder cancer by activating the PI3K-AKT pathway. In a further study, they found that CRNDE could act as a scaffold to recruit DMBT1 and c-IAP1. In renal cell carcinoma (RCC) [31], CRNDE promoted cell proliferation and growth in RCC in vivo and in vitro and regulated the cell cycle transition of RCC by activating Wnt/β-catenin signaling via modulation of CCND1 and CCNE1 expression. Wnt/β-catenin signaling activated by CRNDE was also observed in colorectal carcinoma [19,32] and BC [16]. In hepatocellular carcinoma [33], CRNDE could promote the proliferation, migration, and invasion of HCC by negatively regulating miR-384 expression, which accelerated NF-κB and p-AKT expression. Han et al. [19] indicated that in CRC, CRNDE could negatively regulate miR-181a-5p expression, which contributed to the progression and chemoresistance of CRC. Gao et al. [34] found that CRNDE promoted metastasis and oxaliplatin resistance of CRC cells via the miR-136/E2F1 axis. Hu et al. [35] demonstrated that in GC, CRNDE promoted proliferation via the miR-145/E2F3 axis. Du et al. [10] reported that CRNDE could be used as a novel tumor promoter by modulating the PI3K/AKT signaling pathway. Other similar studies in PC [13], cervical cancer [12], osteosarcoma [36], and papillary thyroid cancer [11] indicated that CRNDE promoted the progression of cancer and could be a therapeutic target for cancer intervention. LncRNA CRNDE was recently found to be related to the clinicopathological parameters and prognosis of patients with cancer. We conducted a comprehensive meta-analysis to investigate the role of CRNDE in patients. Thirteen studies with 1570 patients were enrolled in this meta-analysis. The pooled results indicated that CRNDE overexpression was significantly correlated to poor prognosis and could be used as an independent prognostic biomarker for patients with cancer. Moreover, the relation between CRNDE and clinicopathological features was evaluated. The pooled results indicated that CRNDE was related to lymph node metastasis and advanced TNM stage; however, no association was determined between CRNDE and gender, tumor size, and differentiation. This meta-analysis included several limitations: (1) most of the included studies were from China so that the results may be suitable for China or Asia only. Thus, a large sample size with high-quality studies from other regions should be included to support the conclusion; (2) No specific HRs were found in some articles; thus, we extracted the data through the K–M curves, which could lead to errors; (3) Some articles with different results may have not been published, which could lead to publication bias; (4) The heterogeneity of meta-analysis about the relationship between lncRNA CRNDE and clinicopathological features was significant, which might be caused by insufficient data and variation in tumor type. Thus, the results of this meta-analysis should be confirmed in further studies. Regardless of these limitations, the conclusion reveals that lncRNA
Acknowledgments Funding This study was supported by Beijing Municipal Administration of Hospital Clinical Medicine Development of Special Funding Support (Grant no. ZYLX201612), Capital Foundation of Medical Development (shoufa2016-22053), and Foundation of Medical Development (shoufa2016-2-2053), Beijing Tongren Hospital Municipal Administration of Hospital Clinical Medicine Development of Special Funding Support (trzdyxzy201705). Conflicts of interest The authors report no conflicts of interest in this work. References [1] W. Chen, R. Zheng, P.D. Baade, S. Zhang, H. Zeng, F. Bray, et al., Cancer statistics in China, 2015, CA Cancer J. Clin. 66 (2016) 115–132. [2] R.L. Siegel, K.D. Miller, A. Jemal, Cancer statistics, 2017, CA Cancer J. Clin. 67 (2017) 7–30. [3] M.A. Avila, C. Berasain, B. Sangro, J. Prieto, New therapies for hepatocellular carcinoma, Oncogene 25 (2006) 3866–3884. [4] K. Yang, W. Park, S.J. Huh, D.S. Bae, B.G. Kim, J.W. Lee, Clinical outcomes in patients treated with radiotherapy after surgery for cervical cancer, Radiat. Oncol. J. 35 (2017) 39–47. [5] J. Ferlay, I. Soerjomataram, R. Dikshit, S. Eser, C. Mathers, M. Rebelo, et al., Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012, Int. J. Cancer 136 (2015) E359–86. [6] J.E. Wilusz, H. Sunwoo, D.L. Spector, Long noncoding RNAs: functional surprises from the RNA world, Genes Dev. 23 (2009) 1494–1504. [7] T.R. Mercer, M.E. Dinger, J.S. Mattick, Long non-coding RNAs: insights into functions, Nat. Rev. Genet. 10 (2009) 155–159. [8] Y.C. Xu, C.J. Liang, D.X. Zhang, G.Q. Li, X. Gao, J.Z. Fu, et al., LncSHRG promotes hepatocellular carcinoma progression by activating HES6, Oncotarget 8 (2017) 70630–70641. [9] L.D. Graham, S.K. Pedersen, G.S. Brown, T. Ho, Z. Kassir, A.T. Moynihan, et al., Colorectal neoplasia differentially expressed (CRNDE), a novel gene with elevated expression in colorectal adenomas and adenocarcinomas, Genes Cancer 2 (2011) 829–840. [10] D.X. Du, D.B. Lian, B.H. Amin, W. Yan, Long non-coding RNA CRNDE is a novel tumor promoter by modulating PI3K/AKT signal pathways in human gastric cancer, Eur. Rev. Med. Pharmacol. Sci. 21 (2017) 5392–5398. [11] H. Sun, L. He, L. Ma, T. Lu, J. Wei, K. Xie, et al., LncRNA CRNDE promotes cell proliferation, invasion and migration by competitively binding miR-384 in papillary thyroid cancer, Oncotarget 8 (2017) 110552–110565. [12] Y. Meng, Q. Li, L. Li, R. Ma, The long non-coding RNA CRNDE promotes cervical cancer cell growth and metastasis, Biol. Chem. 399 (2017) 93–100. [13] G. Wang, J. Pan, L. Zhang, Y. Wei, C. Wang, Long non-coding RNA CRNDE sponges miR-384 to promote proliferation and metastasis of pancreatic cancer cells through upregulating IRS1, Cell Prolif. 50 (2017). [14] B. Yu, Q. Du, H. Li, H.Y. Liu, X. Ye, B. Zhu, et al., Diagnostic potential of serum exosomal colorectal neoplasia differentially expressed long non-coding RNA (CRNDE-p) and microRNA-217 expression in colorectal carcinoma, Oncotarget 8 (2017) 83745–83753. [15] X.X. Liu, H.P. Xiong, J.S. Huang, K. Qi, J.J. Xu, Highly expressed long non-coding RNA CRNDE promotes cell proliferation through PI3K/AKT signalling in non-small cell lung carcinoma, Clin. Exp. Pharmacol. Physiol. 44 (2017) 895–902. [16] J. Huan, L. Xing, Q. Lin, H. Xui, X. Qin, Long noncoding RNA CRNDE activates Wnt/ beta-catenin signaling pathway through acting as a molecular sponge of microRNA136 in human breast cancer, Am. J. Transl. Res. 9 (2017) 1977–1989. [17] J. Ding, J. Li, H. Wang, Y. Tian, M. Xie, X. He, et al., Long noncoding RNA CRNDE promotes colorectal cancer cell proliferation via epigenetically silencing DUSP5/ CDKN1A expression, Cell Death Dis. 8 (2017) e2997. [18] L.M. Szafron, A. Balcerak, E.A. Grzybowska, B. Pienkowska-Grela, A. Podgorska, R. Zub, et al., The putative oncogene, CRNDE, is a negative prognostic factor in ovarian cancer patients, Oncotarget 6 (2015) 43897–43910. [19] P. Han, J.W. Li, B.M. Zhang, J.C. Lv, Y.M. Li, X.Y. Gu, et al., The lncRNA CRNDE promotes colorectal cancer cell proliferation and chemoresistance via miR-181a-5pmediated regulation of Wnt/beta-catenin signaling, Mol. Cancer 16 (2017) 9.
106
Clinica Chimica Acta 481 (2018) 99–107
C. Liang et al.
regulating miR-186, Oncotarget 6 (2015) 25339–25355. [29] J. Zheng, X. Liu, P. Wang, Y. Xue, J. Ma, C. Qu, et al., CRNDE promotes malignant progression of glioma by attenuating miR-384/PIWIL4/STAT3 Axis, Mol. Ther. 24 (2016) 1199–1215. [30] S. Shen, H. Liu, Y. Wang, J. Wang, X. Ni, Z. Ai, et al., Long non-coding RNA CRNDE promotes gallbladder carcinoma carcinogenesis and as a scaffold of DMBT1 and CIAP1 complexes to activating PI3K-AKT pathway, Oncotarget 7 (2016) 72833–72844. [31] K. Shao, T. Shi, Y. Yang, X. Wang, D. Xu, P. Zhou, Highly expressed lncRNA CRNDE promotes cell proliferation through Wnt/beta-catenin signaling in renal cell carcinoma, Tumour Biol. 37 (2016) 15997–16004. [32] B. Yu, X. Ye, Q. Du, B. Zhu, Q. Zhai, X.X. Li, The long non-coding RNA CRNDE promotes colorectal carcinoma progression by competitively binding miR-217 with TCF7L2 and enhancing the Wnt/beta-catenin signaling pathway, Cell. Physiol. Biochem. 41 (2017) 2489–2502. [33] Z. Chen, C. Yu, L. Zhan, Y. Pan, L. Chen, C. Sun, LncRNA CRNDE promotes hepatic carcinoma cell proliferation, migration and invasion by suppressing miR-384, Am. J. Cancer Res. 6 (2016) 2299–2309. [34] H. Gao, X. Song, T. Kang, B. Yan, L. Feng, L. Gao, et al., Long noncoding RNA CRNDE functions as a competing endogenous RNA to promote metastasis and oxaliplatin resistance by sponging miR-136 in colorectal cancer, Oncol. Targets Ther. 10 (2017) 205–216. [35] C.E. Hu, P.Z. Du, H.D. Zhang, G.J. Huang, Long noncoding RNA CRNDE promotes proliferation of gastric cancer cells by targeting miR-145, Cell. Physiol. Biochem. 42 (2017) 13–21. [36] Z. Li, Y. Tang, W. Xing, W. Dong, Z. Wang, LncRNA, CRNDE promotes osteosarcoma cell proliferation, invasion and migration by regulating Notch1 signaling and epithelial-mesenchymal transition, Exp. Mol. Pathol. 104 (2017) 19–25.
[20] T. Liu, X. Zhang, Y.M. Yang, L.T. Du, C.X. Wang, Increased expression of the long noncoding RNA CRNDE-h indicates a poor prognosis in colorectal cancer, and is positively correlated with IRX5 mRNA expression, Oncol. Targets Ther. 9 (2016) 1437–1448. [21] S.Y. Jing, Y.Y. Lu, J.K. Yang, W.Y. Deng, Q. Zhou, B.H. Jiao, Expression of long noncoding RNA CRNDE in glioma and its correlation with tumor progression and patient survival, Eur. Rev. Med. Pharmacol. Sci. 20 (2016) 3992–3996. [22] H. Jiang, Y. Wang, M. Ai, H. Wang, Z. Duan, H. Wang, et al., Long noncoding RNA CRNDE stabilized by hnRNPUL2 accelerates cell proliferation and migration in colorectal carcinoma via activating Ras/MAPK signaling pathways, Cell Death Dis. 8 (2017) e2862. [23] K.M. Kiang, X.Q. Zhang, G.P. Zhang, N. Li, S.Y. Cheng, M.W. Poon, et al., CRNDE expression positively correlates with EGFR activation and modulates glioma cell growth, Target. Oncol. 12 (2017) 353–363. [24] M. Zhang, C. Gao, Y. Yang, G. Li, J. Dong, Y. Ai, et al., Long noncoding RNA CRNDE/PRC2 participated in the radiotherapy resistance of human lung adenocarcinoma through targeting p21 expression, Oncol. Res. (2017) (Epub ahead of print). [25] Y. Han, M. Lao, Q. Lei, Y. Li, Expression and clinical significance of long non-coding RNA CRNDE in cervical cancer, Chin. J. Clin. Oncol. 42 (2015) 705–708 (In Chinese). [26] X. Li, D. Tian, W. Tang, X. Dan, S. Hu, Z. Xiao, Expression and clinical significance of lncRNA CRNDE in patients with non-small cell lung cancer, Chin. J. Cancer Biother. 24 (2017) 889–893 (In Chinese). [27] Y. Wang, Y. Wang, J. Li, Y. Zhang, H. Yin, B. Han, CRNDE, a long-noncoding RNA, promotes glioma cell growth and invasion through mTOR signaling, Cancer Lett. 367 (2015) 122–128. [28] J. Zheng, X.D. Li, P. Wang, X.B. Liu, Y.X. Xue, Y. Hu, et al., CRNDE affects the malignant biological characteristics of human glioma stem cells by negatively
107
Contents lists available at ScienceDirect
Clinica Chimica Acta journal homepage: www.elsevier.com/locate/cca
Review
Long non-coding RNA CRNDE as a potential prognostic biomarker in solid tumors: A meta-analysis ⁎
Chaojie Lianga,1, Bing Zhangb,1, Hua Gea, Yingchen Xua, Guangming Lia, , Jixiang Wua, a b
T
⁎
Department of General Surgery, Beijing Tongren Hospital, Capital Medical University, Beijing 100730, China Department of Tumor Center, Beijing Tongren Hospital, Capital Medical University, Beijing 100730, China
A R T I C L E I N F O
A B S T R A C T
Keywords: LncRNA CRNDE Neoplasm Prognosis Meta-analysis
Background and aim: Long non-coding RNA colorectal neoplasia differentially expressed (CRNDE) has been demonstrated to be highly expressed in many malignant tumors; however, the role of CRNDE in cancer remains undetermined because of limitations in sample size. We conducted a meta-analysis to assess the role of CRNDE in cancer. Methods: PubMed, Medline, Cochrane Library, Web of Science, EMBASE database, Ovid, Chinese CNKI, and Chinese WanFang database were systematically searched. The relation between CRNDE and the clinicopathological characteristics and prognosis of patients with cancer was determined using pooled odds ratios (ORs) and hazard ratios (HRs) with 95% confidence interval (CI). Results: Thirteen studies with 1570 patients were included. The pooled results indicated that high CRNDE expression was related to lymph node metastasis (YES vs. NO: OR = 3.50, 95% CI = [1.05, 8.09]) and TNM stage (I + II vs. III + IV: OR = 0.26, 95% CI = [0.18, 0.37]) but not to gender, tumor size, and differentiation. High CRNDE expression indicated poor overall survival (OS) (HR = 2.06, 95% CI = [1.66, 2.47]). CRNDE could be an independent predictive factor for OS (HR = 1.62, 95% CI = [1.15, 2.08]) in patients with cancer. Conclusion: Therefore, high CRNDE expression was associated with advanced clinicopathological characteristics, and CRNDE could be used as a reliable prognostic biomarker in human cancer. However, more high-quality studies with a large sample size are needed to support this meta-analysis.
1. Introduction In 2015, nearly 4 million cases and 3 million recent deaths in China were attributed to cancer, one of the most deadly diseases worldwide [1]. In the United States, about 1,700,000 new cancer patients and 600,000 cancer-related deaths were reported, according to the American Cancer Society [2]. Despite technological advances and scientific research in cancer, the clinical outcome of cancer remains poor. One of the main reasons is the lack of effective biomarkers to diagnose cancer at an early stage, rendering necessary the identification of potential diagnostic and prognostic biomarkers for cancer patients [3–5]. Long non-coding RNAs (LncRNA), which were initially considered as junk or transcriptional noise, measuring > 200 nucleotides in length. Accumulating evidence indicates that lncRNAs play a vital role in the progression and development of cancer [6,7] and participate in numerous biological processes, including the regulation of the proliferation, invasion, and metastasis of cancer. Studies also indicate that lncRNAs can be used as prognostic biomarkers for cancer [8]. ⁎
1
Long non-coding RNA colorectal neoplasia differentially expressed (CRNDE), which is located on chromosome 16, was initially identified as an lncRNA in colorectal cancer (CRC) and reported to be highly elevated in CRC [9]. In further studies, CRNDE was also elevated in many malignant tumors, such as gastric cancer (GC) [10], papillary thyroid cancer [11], cervical cancer [12], and pancreatic cancer (PC) [13]. Other studies indicated that the expression of lncRNA CRNDE was correlated with the clinicopathological features and prognosis of patients with cancer [14,15]. Huan et al. [16] found that high CRNDE expression was associated with tumor size and TNM stage but not with age and histological grade in breast cancer (BC). Ding et al. [17] demonstrated that CRNDE upregulation in CRC was related to tumor size and tumor stage but not correlated with the depth of tumor, age, and gender. Du et al. [10] reported that high CRNDE expression in GC was related to lymph node metastasis, depth of invasion, and TNM stage but not to tumor size and differentiation. Wang et al. [13] indicated that high CRNDE expression in PADC was associated with TNM stage, tumor size, differentiation, and lymph node metastasis. However, most studies on CRNDE are limited by
Corresponding authors at: Department of General Surgery, Beijing Tongren Hospital, Capital Medical University, No. 2, Chongwenmennei Street, Dongcheng, Beijing 100730, China. E-mail addresses: [email protected] (G. Li), [email protected] (J. Wu). Are equal to this study.
https://doi.org/10.1016/j.cca.2018.02.039 Received 17 January 2018; Received in revised form 28 February 2018; Accepted 28 February 2018 Available online 01 March 2018 0009-8981/ © 2018 Elsevier B.V. All rights reserved.
Clinica Chimica Acta 481 (2018) 99–107
C. Liang et al.
the sample size and discrete outcome, and no meta-analysis has thus far been conducted to assess the role of CRNDE in cancer. Thus, we conducted this meta-analysis to explore the relationship between CRNDE and clinicopathological features and prognosis in cancer.
The exclusion criteria were as follows: (1) Reviews, case reports, letters, editorials, meta-analysis, and conference reports; (2) studies focused on the molecular structure and function of CRNDE; and (3) non-human studies and duplicate publications.
2. Materials and methods
2.2. Data extraction and quality assessment
2.1. Search strategy and study selection
Two researchers (Chaojie Liang and Bing Zhang) searched and assessed the literature independently in accordance with the criteria. The extracted data included the name of the first author, the year of publication, country/region, number of cases, clinicopathological features, and overall survival (OS). The quality of the literature was assessed using the Newcastle–Ottawa–Scale (NOS) criteria. Articles with a NOS score ≥ 6 were considered high-quality; otherwise, they were regarded as low-quality.
The studies included in this meta-analysis were searched in PubMed, Medline, Cochrane Library, Web of Science, the EMBASE database, Ovid, Chinese CNKI, and the Chinese WanFang database. The research ended on January 10, 2018, and was not limited by inadequate data. The key words used for the searches were as follows: “CRNDE” or “colorectal neoplasia differentially expressed” or “lncRNA CRNDE” and “cancers” or “neoplasm.” The inclusion criteria were as follows: (1) LncRNA CRNDE expression was detected in human solid cancer tissues, and patients were grouped according to lncRNA CRNDE expression; (2) The relationship between lncRNA CRNDE expression and clinicopathological features and prognosis was described; and (3) The hazard ratios (HRs) with 95% confidence interval (CI) were provided or could be calculated using the survival curve.
2.3. Statistical analysis All statistical data were calculated using STATA 14.2. The HRs with 95% confidence intervals (CIs) were extracted from a Kaplan–Meier curve by using Engauge Digitizer 10.0. The association between CRNDE expression and clinicopathological characteristics—gender (male vs. female), tumor size (> 5 cm vs. ≤5 cm), differentiation (low vs. high
Fig. 1. Flow diagram of study selection.
100
Clinica Chimica Acta 481 (2018) 99–107
C. Liang et al.
3. Results
and moderate), lymph node metastasis (yes vs. no), distant metastasis (YES vs. NO), and TNM stage (I + II vs. III + IV)—were assessed by pooled odds ratios (ORs) with 95% CIs. The relationship between high CRNDE expression and OS was calculated using HRs with 95% CI. I2 and Q tests were performed to calculate heterogeneity when I2 > 50%. Either the random effects model or the fixed effects model was used. If the pooled OR HR with 95%CI did not overlap 1, the pooled results were significant. The source of heterogeneity and the stability of results were assessed by sensitivity analysis and subgroup analysis, and Begg's test was used to assess the potential publication bias.
3.1. Study identification and characteristics Figs. 1 and 2 show that 1570 patients and 13 studies (2 glioma cancer, 1 ovarian cancer, 1 cervical cancer, 4 CRC, 1 BC, 1 PC, 1 lung adenocarcinoma, and 1 non-small lung cancer) [10,13,16–26] were included in this meta-analysis in accordance with the criteria. The study samples ranged from 58 to 251, and the studies were published from 2015 to 2017. A total of 10 studies were published in English and 2 in Chinese; most
Fig. 2. Forest plot of studies evaluating the relationship between CRNDE expression and clinicopathological features. A. gender; B. tumor size; C. differentiation; D. lymph node metastasis; E. TNM stage.
101
Clinica Chimica Acta 481 (2018) 99–107
C. Liang et al.
Table 1 Characteristics of studies included in the meta-analysis. Study
Year
Country
Cancer type
Detection method
Sample size
LncRNA CRNDE expression High
Lukasz Han Y Jing SY Han P Liu T Huan JL Ding J Du DX Jiang HJ Wang G Kiang Zhang M Li XH
2015 2015 2016 2017 2016 2017 2017 2017 2017 2017 2017 2017 2017
Poland China China China China China China China China China China China China
OC CC Glioma CRC CRC BC CRC GC CRC PC Glioma LAD NSCLC
RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR
135 87 164 64 142 103 80 118 251 58 165 66 137
Survival information
NOS score
OS OS OS OS OS OS NA OS OS OS OS OS OS
8 6 8 8 8 6 6 7 7 6 6 6 7
Low
Total
LNM
HTS
LD
TS > 5
M
Total
LNM
HTS
LD
TS > 5
M
118 51 83 32 71 44 40 61 120 38 83 36 83
NA 39 NA 24 NA NA NA 35 53 26 NA 28 45
NA NA 12 NA NA 26 17 24 NA 15 NA 13 NA
NA 25 NA 12 NA NA NA 39 19 20 NA 12 40
NA 32 NA 24 NA NA 25 24 77 NA NA NA NA
NA NA 55 23 NA NA 25 21 73 24 NA 23 40
17 36 81 32 71 59 40 57 131 20 82 30 66
NA 12 NA 8 NA NA NA 26 67 12 NA 8 32
NA NA 71 NA NA 18 23 37 NA 23 NA 23 NA
NA 26 NA 20 NA NA NA 22 101 18 NA 24 37
NA 19 NA 8 NA NA 15 37 43 NA NA NA NA
NA NA 28 16 NA NA 15 40 47 14 NA 13 43
LNM: lymph node metastasis, DM: distant metastasis, HTS: high TNM stage (III/IV), T1/2: depth of infiltration (T1/2), LD: low differentiation. TS > 5: tumor size > 5 cm, M: male. Table 2 Survival data of studies included in the meta-analysis. Study
Year
Country
Cancer type
Detection method
Sample size
Survival information
HR statistic
NOS score
Data in paper Data in paper Data in paper/survival curve Data in paper Data in paper Survival curve Survival curve Data in paper Data in paper Survival curve Survival curve Survival curve Survival curve Data in paper/survival curve
8 8 6 8 8 6 6 7 8 6 6 6 6 7
Univariate analysis Multivariate analysis Lukasz1 Lukasz2 Han Y Jing SY Liu T1 Han P Huang JL Du DX Jiang HJ Kiang Kiang Wang G Zhang M Li XH
2015 2015 2015 2016 2016 2017 2017 2017 2017 2017 2017 2017 2017 2017
Poland Poland China China China China China China China China China China China China
OC OC CC Glioma CRC CRC BC GC CRC Glioma Glioma PC LAD NSCLC
RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR RT-qPCR
103 135 87 164 142 64 103 118 251 101 64 58 66 137
NA NA 2.239(1.095–4.593)(E) 2.236(1.452–7.125) 2.906(1.452–7.125) 1.850(1.032–6.558)(E) 1.770(1.011–3.768)(E) 2.663(1.458–5.532) 2.062(1.291–3.173) 1.780(1.252–2.532)(E) 2.491(1.018–6.093)(E) 1.676(1.011–6.509)(E) 2.516(1.319–4.801)(E) 3.411(1.517–7.670)(E)
6.072(1.814–20.32) 5.437(1.699–17.4) 4.702(3.542–12.618) 1.589(1.034–6.873) 1.589(1.034–3.684) NA NA 2.453(1.321–4.777) 1.693(1.047–2.738) NA NA NA NA 4.032(2.541–10.350)
OC: ovarian cancer; CC: cervical cancer; CRC: colorectal cancer; BC: breast cancer; GC: gastric cancer; PADC: pancreatic cancer; LAD: lung adenocarcinoma; NSCLC: non-small lung cancer. Table 3 LncRNA CRNDE clinicopathological features for cancers. Heterogeneity Clinicopathological features
No. of studies
No. of patients
Pooled OR (95% CI)
PHet
I2 (%)
p value
Model used
Gender Tumor size Differentiation Lymph node metastasis TNM stage
8 5 8 7 6
950 576 884 791 588
1.07 1.96 1.24 3.50 0.26
0.349 0.007 < 0.001 0.001 0.799
10.5 71.4 76.6 74.7 0.0
0.366 0.053 0.530 < 0.001 < 0.001
Fixed Random Random Random Fixed
[0.83, [0.99, [0.64, [1.87, [0.18,
1.39] 3.86] 2.40] 6.56] 0.37]
Random, random-effects model; Fixed, fixed-effects model; OR, odds ratio; CI, confidence interval.
gender (male: female, OR = 0.87, 95% CI = [0.66, 1.17], p = 0.366, Fig. 2A), and 5 studies with 576 patients demonstrated that CRNDE upregulation was not associated with tumor size (> 5 cm vs. ≤5 cm: OR = 1.96, 95% CI = [0.99, 3.86], p = 0.053, Fig. 2B). Similarly, the pooled results from 8 studies with 884 patients showed that CRNDE expression was not correlated with differentiation (low: moderate + high, OR = 1.24, 95% CI = [0.64, 2.40], p = 0.530, Fig. 2C). However, the pooled results suggested that high CRNDE expression was associated with lymph node metastasis (YES:NO, OR = 3.50, 95% CI = [1.87, 6.56], p < 0.001, Fig. 2D) and advanced TNM stage (I + II: III + IV, OR = 0.26, 95% CI = [0.18, 0.37], p < 0.001,
studies were from China, and 1 study was from Poland. All studies scored ≥6 according to the NOS score criteria and thus were considered highquality. The clinicopathological characteristics and survival information obtained from the studies are summarized in Tables 1 and 2. 3.2. Association between CRNDE and clinicopathological characteristics As shown in Table 3, the pooled ORs with 95% CI were calculated to assess the relationship between CRNDE and clinicopathological features in patients with cancer. The pooled results from 8 studies with 950 patients indicated that the high CRNDE expression was not related to 102
Clinica Chimica Acta 481 (2018) 99–107
C. Liang et al.
Table 4 Subgroup analysis of overall survival by tumor type, NOS score, sample size. Subgroups Tumor type Female reproductive system Nervous system Digestive system Respiratory system Sample size ≤110 > 110 NOS score ≤7 >7
No. of studies
No. of patients
Pooled HR (95% CI)
PHet
I2 (%)
p value
2 3 5 2
190 333 629 203
1.95 1.84 2.26 2.74
3.03) 2.45) 2.95) 4.25)
0.679 0.834 0.841 0.622
0.0 0.0 0.0 0.0
0.12 < 0.001 < 0.001 < 0.001
6 6
547 908
2.17 (1.24, 3.10) 2.26 (1.64, 2.88)
0.963 0.823
0.0 0.0
< 0.001 < 0.001
9 3
798 557
1.98 (1.50, 2.46) 2.28 (1.52, 3.04)
0.986 0.604
0.0 0.0
< 0.001 < 0.001
(0.87, (1.24, (1.57, (1.22,
Abbreviations: OR = hazard ratio, CI = confidence interval, n = number of sample size.
Fig. 3. Begg's publication bias plots for CRNDE-related studies: A. gender; B. tumor size; C. differentiation; D. lymph node metastasis; E. TNM stage.
103
Clinica Chimica Acta 481 (2018) 99–107
C. Liang et al.
Fig. 4. Forest plots and Begg's publication bias plots of studies evaluating the relation between CRNDE expression and overall survival (OS) rate (A) and independent predictive factor for OS (B).
low: HR = 2.28, 95% CI = [1.52, 3.04], p < 0.001, Fig. 5C) (n ≤ 7: high vs. low: HR = 1.98, 95% CI = [1.50, 2.46], p < 0.001, Fig. 5C). The robustness of the pooled results was evaluated by sensitivity analysis, and the results (Fig. 6A) were not significantly affected when any individual study was removed, which indicated that the results were reliable.
Fig. 2E). Therefore, this meta-analysis indicated that the high CRNDE expression was correlated with advanced clinicopathological features. 3.3. Prognostic value of CRNDE for OS of cancer The prognostic value of CRNDE for OS of cancer included 11 studies (Kiang conducted 2 trials) and 1355 patients. The pooled results indicated that high CRNDE expression was correlated with poor OS (high vs. low: HR = 2.06, 95% CI = [1.66, 2.47], p < 0.001, Fig. 4A). We then conducted a subgroup analysis for OS according to cancer type, NOS score, and sample size. As shown in Table 4, the subgroup for cancer type indicated that high CRNDE expression was a strong prognostic biomarker in cancer of the nervous system (high vs. low: HR = 1.84, 95% CI = [1.24, 2.45], p < 0.001, Fig. 5A), digestive system (high vs. low: HR = 2.26, 95% CI = [1.57, 2.95], p < 0.001, Fig. 5A), and respiratory system (high vs. low: HR = 2.74, 95% CI = [1.22, 4.25], p < 0.001, Fig. 5A). However, the CRNDE expression was not associated with OS in female reproductive cancer (high vs. low: HR = 1.95, 95% CI = [0.87, 3.03], Fig. 5A). We conducted a subgroup analysis according to sample size, and the threshold was 110 patients. The pooled results showed that CRNDE upregulation was correlated with poor OS in both groups (n ≥ 100: high vs. low: HR = 2.26, 95% CI = [1.64, 2.88], p < 0.001, Fig. 5B) (n < 100: high vs. low: HR = 1.92, 95% CI = [1.39, 2.46], p < 0.001, Fig. 5B). We also conducted a subgroup analysis according to NOS score; similarly, the pooled results suggested that high CRNDE expression was significantly correlated with poor OS in both groups (n > 7: high vs.
3.4. Independent prognostic value of CRNDE in cancer The independent prognostic value of CRNDE in cancer were calculated based on the multivariate analysis in 7 studies (Lukasz conducted 2 trials) with 1342 patients. The pooled HRs showed that CRNDE expression was an independent prognostic factor for OS in patients with cancer (HR = 1.62, 95% CI = [1.15, 2.08], I2 = 0, Fig. 4B), and the results were also reliable according to the sensitivity analysis (Fig. 6B). 3.5. Begg's funnel plot analysis Begg's funnel plot was conducted to evaluate publication bias. Figs. 3 and 4 show no publication bias for gender (p = 0.536), tumor size (p = 0.086), differentiation (p = 0.108), TNM stage (p = 0.268), overall survival (p = 1), and independent factor for OS (p = 0.266), However, the publication bias existed for lymph node metastasis (p = 0.016) and sensitivity analysis was conducted to determine whether individual studies influenced the pooled OR. And the results showed no study substantially influenced the pooled OR, which suggested that more studies needed to be included for further research. 104
Clinica Chimica Acta 481 (2018) 99–107
C. Liang et al.
Fig. 5. Forest plots of subgroup analysis for OS: subgroup analysis by tumor type (A), subgroup analysis by sample size (B), and subgroup analysis by Newcastle–Ottawa–Scale score (C).
Fig. 6. Sensitivity analysis for overall survival (OS) (A) and independent predictive factor for OS (B).
105
Clinica Chimica Acta 481 (2018) 99–107
C. Liang et al.
4. Discussion
CRNDE overexpression is significantly associated with increased lymph node metastasis and advanced TNM stage; moreover, high CRNDE expression was related to poor OS and the expression of CRNDE could be an independent factor for OS in patients with cancer. Thus, lncRNA CRNDE could be used as a potential prognostic biomarker for patients with cancer. However, more studies with a large sample size and high quality need to be conducted to elucidate the role of lncRNA CRNDE.
An increasing number of studies demonstrated that the abnormal expression of lncRNAs was associated with the development and progression of cancer and significantly related to clinicopathological features and prognosis. CRNDE was initially identified as an lncRNA in CRC. Further studies also found that CRNDE was overexpressed in several types of tumors and played an oncogenic role in tumors. In gliomas [27], CRNDE overexpression promoted cell growth and migration in vitro and in vivo by regulating the mTOR signal pathway. Zheng et al. [28] demonstrated that the upregulation of CRNDE expression could promote the proliferation, migration, and invasion of glioma stem cells by regulating miR-186 expression. In addition, miR186 could bind to XIAP and the PAK7 3′UTR region to regulate the expression of caspase3, BAD, cyclin D1, and MARK2. In addition, Zheng et al. [29] suggested that CRNDE could play an oncogenic role by attenuating the miR-384/PIWL4/STAT3 axis. Kiang [23] confirmed that CRNDE depletion inhibited glioma growth by activating EGFR signaling. In gallbladder cancer, Shen et al. [30] indicated that CRNDE promoted the carcinogenesis of gallbladder cancer by activating the PI3K-AKT pathway. In a further study, they found that CRNDE could act as a scaffold to recruit DMBT1 and c-IAP1. In renal cell carcinoma (RCC) [31], CRNDE promoted cell proliferation and growth in RCC in vivo and in vitro and regulated the cell cycle transition of RCC by activating Wnt/β-catenin signaling via modulation of CCND1 and CCNE1 expression. Wnt/β-catenin signaling activated by CRNDE was also observed in colorectal carcinoma [19,32] and BC [16]. In hepatocellular carcinoma [33], CRNDE could promote the proliferation, migration, and invasion of HCC by negatively regulating miR-384 expression, which accelerated NF-κB and p-AKT expression. Han et al. [19] indicated that in CRC, CRNDE could negatively regulate miR-181a-5p expression, which contributed to the progression and chemoresistance of CRC. Gao et al. [34] found that CRNDE promoted metastasis and oxaliplatin resistance of CRC cells via the miR-136/E2F1 axis. Hu et al. [35] demonstrated that in GC, CRNDE promoted proliferation via the miR-145/E2F3 axis. Du et al. [10] reported that CRNDE could be used as a novel tumor promoter by modulating the PI3K/AKT signaling pathway. Other similar studies in PC [13], cervical cancer [12], osteosarcoma [36], and papillary thyroid cancer [11] indicated that CRNDE promoted the progression of cancer and could be a therapeutic target for cancer intervention. LncRNA CRNDE was recently found to be related to the clinicopathological parameters and prognosis of patients with cancer. We conducted a comprehensive meta-analysis to investigate the role of CRNDE in patients. Thirteen studies with 1570 patients were enrolled in this meta-analysis. The pooled results indicated that CRNDE overexpression was significantly correlated to poor prognosis and could be used as an independent prognostic biomarker for patients with cancer. Moreover, the relation between CRNDE and clinicopathological features was evaluated. The pooled results indicated that CRNDE was related to lymph node metastasis and advanced TNM stage; however, no association was determined between CRNDE and gender, tumor size, and differentiation. This meta-analysis included several limitations: (1) most of the included studies were from China so that the results may be suitable for China or Asia only. Thus, a large sample size with high-quality studies from other regions should be included to support the conclusion; (2) No specific HRs were found in some articles; thus, we extracted the data through the K–M curves, which could lead to errors; (3) Some articles with different results may have not been published, which could lead to publication bias; (4) The heterogeneity of meta-analysis about the relationship between lncRNA CRNDE and clinicopathological features was significant, which might be caused by insufficient data and variation in tumor type. Thus, the results of this meta-analysis should be confirmed in further studies. Regardless of these limitations, the conclusion reveals that lncRNA
Acknowledgments Funding This study was supported by Beijing Municipal Administration of Hospital Clinical Medicine Development of Special Funding Support (Grant no. ZYLX201612), Capital Foundation of Medical Development (shoufa2016-22053), and Foundation of Medical Development (shoufa2016-2-2053), Beijing Tongren Hospital Municipal Administration of Hospital Clinical Medicine Development of Special Funding Support (trzdyxzy201705). Conflicts of interest The authors report no conflicts of interest in this work. References [1] W. Chen, R. Zheng, P.D. Baade, S. Zhang, H. Zeng, F. Bray, et al., Cancer statistics in China, 2015, CA Cancer J. Clin. 66 (2016) 115–132. [2] R.L. Siegel, K.D. Miller, A. Jemal, Cancer statistics, 2017, CA Cancer J. Clin. 67 (2017) 7–30. [3] M.A. Avila, C. Berasain, B. Sangro, J. Prieto, New therapies for hepatocellular carcinoma, Oncogene 25 (2006) 3866–3884. [4] K. Yang, W. Park, S.J. Huh, D.S. Bae, B.G. Kim, J.W. Lee, Clinical outcomes in patients treated with radiotherapy after surgery for cervical cancer, Radiat. Oncol. J. 35 (2017) 39–47. [5] J. Ferlay, I. Soerjomataram, R. Dikshit, S. Eser, C. Mathers, M. Rebelo, et al., Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012, Int. J. Cancer 136 (2015) E359–86. [6] J.E. Wilusz, H. Sunwoo, D.L. Spector, Long noncoding RNAs: functional surprises from the RNA world, Genes Dev. 23 (2009) 1494–1504. [7] T.R. Mercer, M.E. Dinger, J.S. Mattick, Long non-coding RNAs: insights into functions, Nat. Rev. Genet. 10 (2009) 155–159. [8] Y.C. Xu, C.J. Liang, D.X. Zhang, G.Q. Li, X. Gao, J.Z. Fu, et al., LncSHRG promotes hepatocellular carcinoma progression by activating HES6, Oncotarget 8 (2017) 70630–70641. [9] L.D. Graham, S.K. Pedersen, G.S. Brown, T. Ho, Z. Kassir, A.T. Moynihan, et al., Colorectal neoplasia differentially expressed (CRNDE), a novel gene with elevated expression in colorectal adenomas and adenocarcinomas, Genes Cancer 2 (2011) 829–840. [10] D.X. Du, D.B. Lian, B.H. Amin, W. Yan, Long non-coding RNA CRNDE is a novel tumor promoter by modulating PI3K/AKT signal pathways in human gastric cancer, Eur. Rev. Med. Pharmacol. Sci. 21 (2017) 5392–5398. [11] H. Sun, L. He, L. Ma, T. Lu, J. Wei, K. Xie, et al., LncRNA CRNDE promotes cell proliferation, invasion and migration by competitively binding miR-384 in papillary thyroid cancer, Oncotarget 8 (2017) 110552–110565. [12] Y. Meng, Q. Li, L. Li, R. Ma, The long non-coding RNA CRNDE promotes cervical cancer cell growth and metastasis, Biol. Chem. 399 (2017) 93–100. [13] G. Wang, J. Pan, L. Zhang, Y. Wei, C. Wang, Long non-coding RNA CRNDE sponges miR-384 to promote proliferation and metastasis of pancreatic cancer cells through upregulating IRS1, Cell Prolif. 50 (2017). [14] B. Yu, Q. Du, H. Li, H.Y. Liu, X. Ye, B. Zhu, et al., Diagnostic potential of serum exosomal colorectal neoplasia differentially expressed long non-coding RNA (CRNDE-p) and microRNA-217 expression in colorectal carcinoma, Oncotarget 8 (2017) 83745–83753. [15] X.X. Liu, H.P. Xiong, J.S. Huang, K. Qi, J.J. Xu, Highly expressed long non-coding RNA CRNDE promotes cell proliferation through PI3K/AKT signalling in non-small cell lung carcinoma, Clin. Exp. Pharmacol. Physiol. 44 (2017) 895–902. [16] J. Huan, L. Xing, Q. Lin, H. Xui, X. Qin, Long noncoding RNA CRNDE activates Wnt/ beta-catenin signaling pathway through acting as a molecular sponge of microRNA136 in human breast cancer, Am. J. Transl. Res. 9 (2017) 1977–1989. [17] J. Ding, J. Li, H. Wang, Y. Tian, M. Xie, X. He, et al., Long noncoding RNA CRNDE promotes colorectal cancer cell proliferation via epigenetically silencing DUSP5/ CDKN1A expression, Cell Death Dis. 8 (2017) e2997. [18] L.M. Szafron, A. Balcerak, E.A. Grzybowska, B. Pienkowska-Grela, A. Podgorska, R. Zub, et al., The putative oncogene, CRNDE, is a negative prognostic factor in ovarian cancer patients, Oncotarget 6 (2015) 43897–43910. [19] P. Han, J.W. Li, B.M. Zhang, J.C. Lv, Y.M. Li, X.Y. Gu, et al., The lncRNA CRNDE promotes colorectal cancer cell proliferation and chemoresistance via miR-181a-5pmediated regulation of Wnt/beta-catenin signaling, Mol. Cancer 16 (2017) 9.
106
Clinica Chimica Acta 481 (2018) 99–107
C. Liang et al.
regulating miR-186, Oncotarget 6 (2015) 25339–25355. [29] J. Zheng, X. Liu, P. Wang, Y. Xue, J. Ma, C. Qu, et al., CRNDE promotes malignant progression of glioma by attenuating miR-384/PIWIL4/STAT3 Axis, Mol. Ther. 24 (2016) 1199–1215. [30] S. Shen, H. Liu, Y. Wang, J. Wang, X. Ni, Z. Ai, et al., Long non-coding RNA CRNDE promotes gallbladder carcinoma carcinogenesis and as a scaffold of DMBT1 and CIAP1 complexes to activating PI3K-AKT pathway, Oncotarget 7 (2016) 72833–72844. [31] K. Shao, T. Shi, Y. Yang, X. Wang, D. Xu, P. Zhou, Highly expressed lncRNA CRNDE promotes cell proliferation through Wnt/beta-catenin signaling in renal cell carcinoma, Tumour Biol. 37 (2016) 15997–16004. [32] B. Yu, X. Ye, Q. Du, B. Zhu, Q. Zhai, X.X. Li, The long non-coding RNA CRNDE promotes colorectal carcinoma progression by competitively binding miR-217 with TCF7L2 and enhancing the Wnt/beta-catenin signaling pathway, Cell. Physiol. Biochem. 41 (2017) 2489–2502. [33] Z. Chen, C. Yu, L. Zhan, Y. Pan, L. Chen, C. Sun, LncRNA CRNDE promotes hepatic carcinoma cell proliferation, migration and invasion by suppressing miR-384, Am. J. Cancer Res. 6 (2016) 2299–2309. [34] H. Gao, X. Song, T. Kang, B. Yan, L. Feng, L. Gao, et al., Long noncoding RNA CRNDE functions as a competing endogenous RNA to promote metastasis and oxaliplatin resistance by sponging miR-136 in colorectal cancer, Oncol. Targets Ther. 10 (2017) 205–216. [35] C.E. Hu, P.Z. Du, H.D. Zhang, G.J. Huang, Long noncoding RNA CRNDE promotes proliferation of gastric cancer cells by targeting miR-145, Cell. Physiol. Biochem. 42 (2017) 13–21. [36] Z. Li, Y. Tang, W. Xing, W. Dong, Z. Wang, LncRNA, CRNDE promotes osteosarcoma cell proliferation, invasion and migration by regulating Notch1 signaling and epithelial-mesenchymal transition, Exp. Mol. Pathol. 104 (2017) 19–25.
[20] T. Liu, X. Zhang, Y.M. Yang, L.T. Du, C.X. Wang, Increased expression of the long noncoding RNA CRNDE-h indicates a poor prognosis in colorectal cancer, and is positively correlated with IRX5 mRNA expression, Oncol. Targets Ther. 9 (2016) 1437–1448. [21] S.Y. Jing, Y.Y. Lu, J.K. Yang, W.Y. Deng, Q. Zhou, B.H. Jiao, Expression of long noncoding RNA CRNDE in glioma and its correlation with tumor progression and patient survival, Eur. Rev. Med. Pharmacol. Sci. 20 (2016) 3992–3996. [22] H. Jiang, Y. Wang, M. Ai, H. Wang, Z. Duan, H. Wang, et al., Long noncoding RNA CRNDE stabilized by hnRNPUL2 accelerates cell proliferation and migration in colorectal carcinoma via activating Ras/MAPK signaling pathways, Cell Death Dis. 8 (2017) e2862. [23] K.M. Kiang, X.Q. Zhang, G.P. Zhang, N. Li, S.Y. Cheng, M.W. Poon, et al., CRNDE expression positively correlates with EGFR activation and modulates glioma cell growth, Target. Oncol. 12 (2017) 353–363. [24] M. Zhang, C. Gao, Y. Yang, G. Li, J. Dong, Y. Ai, et al., Long noncoding RNA CRNDE/PRC2 participated in the radiotherapy resistance of human lung adenocarcinoma through targeting p21 expression, Oncol. Res. (2017) (Epub ahead of print). [25] Y. Han, M. Lao, Q. Lei, Y. Li, Expression and clinical significance of long non-coding RNA CRNDE in cervical cancer, Chin. J. Clin. Oncol. 42 (2015) 705–708 (In Chinese). [26] X. Li, D. Tian, W. Tang, X. Dan, S. Hu, Z. Xiao, Expression and clinical significance of lncRNA CRNDE in patients with non-small cell lung cancer, Chin. J. Cancer Biother. 24 (2017) 889–893 (In Chinese). [27] Y. Wang, Y. Wang, J. Li, Y. Zhang, H. Yin, B. Han, CRNDE, a long-noncoding RNA, promotes glioma cell growth and invasion through mTOR signaling, Cancer Lett. 367 (2015) 122–128. [28] J. Zheng, X.D. Li, P. Wang, X.B. Liu, Y.X. Xue, Y. Hu, et al., CRNDE affects the malignant biological characteristics of human glioma stem cells by negatively
107