Prognostic Values of Long Noncoding RNA linc00152 in Various Carcinomas: An Updated Systematic Review and Meta‐Analysis

Abstract

Dysregulation of the long noncoding RNA linc00152 has been reported in various solid tumors. Here, we performed a synthetic analysis to clarify the clinical value of linc00152 as a prognostic indicator in malignant tumors. Article collection was conducted using several electronic databases, including PubMed, Web of Science, Medline, OVID, and Embase (up to February 13, 2018). The meta‐analysis comprised nine original studies and 808 total patients. The application of a random‐effects model revealed significant positive association between high expression level of linc00152 and lymph node metastasis (odds ratio [OR] = 2.93, 95% confidence interval [CI]: 1.88–4.57, p < .0001; I² = 48.8, p = .119) and negative association with low‐grade cancer (OR = 2.43, 95% CI: 1.51–3.92; I² = 61.7, p = .033), while with tumor recurrence (hazard ratio [HR] = 3.32, 95% CI: 1.98–5.57, p < .0001; I² = 0, p = .451) by fixed‐effects model as the low heterogeneity. As demonstrated via the application of the fixed‐effects model, Linc00152 overexpression is positively related to poor overall survival (pooled HR = 1.98, 95% CI: 1.70–2.31, p < .0001; I² = 0%, p = .756) and poor disease‐free survival (HR = 1.66, 95% CI: 1.20–2.29, p < .0001; I² = 75.8%, p = .042) in human solid cancers. Statistically significant associations were additionally found with cancer type, sample size, and follow‐up time. In conclusion, linc00152 is of potential value as a novel biomarker of lymph node metastasis and prognosis in human cancer.

Implications for Practice

linc00152 is of potential value as a novel biomarker of lymph node metastasis and prognosis in human cancer.

Short abstract

This review considers the question of the prognostic value of linc00152 and whether it can be used as a novel biomarker of lymph node metastasis in patients with cancer.

Introduction

Rationale

With increasing incidence and a high risk for mortality, cancer is becoming the leading cause of death and continues to be a major health problem worldwide. It is expected that 1,688,780 new cases of cancer and 600,920 cancer deaths will occur in 2017 in the U.S. According to the literature, there were 4,292,000 new cancer cases and 2,814,000 cancer deaths in 2015 in China 1, 2. Although clinical treatments such as surgery, chemotherapy, radiotherapy, and targeted therapy may improve prognosis 3, the 5‐year survival rate of patients with cancer remains low as a result of malignant progression 4 and a lack of effective early diagnostic methods. Therefore, the identification of potential diagnostic and prognostic biomarkers for monitoring cancer progression is critical.

Long noncoding RNAs (lncRNAs) are mRNA‐like transcripts with a length of more than 200 nucleotides that lack an open reading frame 5. These transcripts were initially considered genomic “noise” 6. However, increasing evidence suggests that lncRNAs act as critical regulators in diverse diseases and cellular processes, such as the regulation of gene expression and post‐translational processing 7, 8, 9. Recent studies on lncRNAs have been reported for a variety of tumors 10, 11, 12, 13, and several have shown promising prognostic markers 14, 15, 16.

Long intergenic noncoding RNA 00152 (linc00152), an 828‐bp lncRNA that maps to chromosome 2p11.2, was first described as being highly expressed in gastric tissues and cells by Pang et al. in 2014 17. Subsequently, numerous studies have reported that linc00152 is independently associated with multiple types of cancer and is upregulated in many cancers, such as gallbladder cancer (GBC) 18, gastric cancer (GC) 19, renal cell carcinoma (RCC) 20, 21, colon cancer (CC) 22, lung adenocarcinoma (LAD) 23, 24, hepatocellular carcinoma (HCC) 25, and tongue squamous cell carcinoma (TSCC) 26. Moreover, high expression of linc00152 is significantly related to lymph node metastasis (LNM), higher tumor node metastasis (TNM) stage, and poor survival 18, 19, 20, 21, 22, 23, 24, 25, 26. However, because these individual studies provide discrete examples and are limited by sample size, no consensus has been reached on the prognostic value of linc00152 in patients with cancer.

Objectives

This meta‐analysis was conducted to elucidate the prognostic value of linc00152 as a novel candidate biomarker in patients with malignant tumors.

Research Question

What is the prognostic value of linc00152 and can it be used as a novel biomarker of lymph node metastasis in patients with cancer?

Materials and Methods

Participants, Interventions, and Comparators

Our goal was to analyze studies that included participants with various types of cancer, and the inclusion criteria were as follows: (a) articles that explored the association between linc00152 expression and cancer prognosis; (b) studies with participants divided into high and low linc00152 expression groups; (c) articles that described related clinicopathologic parameters such as LNM, TNM stage, tumor recurrence, and tumor size; and (d) the inclusion of sufficient data for the computation of odds ratios (ORs) and corresponding 95% confidence intervals (CIs). Exclusion criteria were as follows: (a) duplicate publications; (b) reviews, letters, case reports, and non‐human subject research; and (c) articles without usable data and articles that were not published in English.

Systematic Review Protocol

This systematic review was performed according to the Preferred Reporting Items for Systematic Reviews and Meta‐Analyses guidelines (Moher et al., 2009). Ethical approval was not required as this study was a systematic review.

Search Strategy

According to the standard meta‐analysis guidelines, relevant articles were searched and accessed through PubMed, Web of Science, Medline, OVID, and Embase, using MeSH. Key terms including the following: “linc00152,” “LncRNA linc00152,” “Long intergenic non‐coding RNA 00152,” “lncRNA,” “noncoding RNA,” “cancer,” “tumor,” “carcinomas,” “neoplasms,” “prognostic,” “prognosis,” “outcome,” “survival,” or “recurrence.” These terms were used to maximize the probability of finding the appropriate articles. The literature search was performed to include articles published until February 13, 2018. The reference lists of retrieved articles were also searched manually to ensure the inclusion of eligible studies. Any conflicts regarding the inclusion and exclusion of articles were resolved through group discussion.

Quality Assessment of Primary Studies

Two investigators independently conducted the quality assessment of this study. We assessed the quality of all eligible studies according to the Newcastle‐Ottawa Scale (NOS), including the following domains: Adequacy of case definition, Number of cases, Representativeness of the cases, Ascertainment of exposure, Ascertainment of detection method, Ascertainment of cutoff, Assessment of outcome, and Adequate follow‐up. A higher score indicates better methodological quality.

Data Sources, Studies Sections, and Data Extraction

Data extraction was performed according to the inclusion and exclusion criteria. Disagreements were resolved by discussing with authors W.H., L.Y., and A.T. The following information was recorded: (a) first author, (b) publication date, (c) country of origin, (d) cancer type, (e) sample size, (f) tumor stage, (g) linc00152 expression detection method, (h) cutoff value, (i) number of patients with high and low linc00152 expression, (j) hazard ratios (HRs) and corresponding 95% CIs for overall survival (OS) and disease‐free survival (DFS), and (k) follow‐up interval. Moreover, clinicopathologic parameters (LNM, TNM stage, tumor recurrence, and tumor size) were recorded from all retrieved articles. The survival data were directly applied if an article stated the detailed HRs and 95% CIs for survival; otherwise, the Engauge Digitizer 4.1 (http://digitizer.sourceforge.net/) was used to extract survival data from Kaplan‐Meier curves.

Target Gene Prediction and Signal Pathway Network Construction

The genes related to linc00152 were extracted from the MEM‐Multi Experiment Matrix database; following this, the gene ontology analysis (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis were carried out and the signal pathway network was constructed by Cytoscape. Human RNA‐binging protein‐LINC00152 interactions were extracted from the starBase v2.0 database.

Statistical Analysis

Study quality was assessed according to the NOS. All statistical analyses were conducted using Stata statistical software version 12.0 (Stata Corporation, College Station, TX). A chi‐square‐based Q test and I² statistics were performed to measure the heterogeneity of the eligible studies 27. The fixed‐effects model was adopted in the absence of heterogeneity (p > .1 and I² <50%), and the random‐effects model was applied if heterogeneity was determined (p < .1 and I² >50%) 28, 29. Egger's test was applied to evaluate potential publication bias, and the stability of the results was assessed through a sensitivity analysis. p values <.05 represented statistical significance

Results

Flow Diagram of the Studies Retrieved for the Review

A flow diagram of studies retrieved for the review is shown in Figure 1.

Figure 1.

The flow diagram of this meta‐analysis.

Study Selection and Characteristics

As shown in the flow diagram (Fig. 1), a total of 331 articles were identified by initial screening, of which 78 duplicate reports were excluded. After screening the titles and abstracts of the articles, nonrelevant studies and non‐human subject studies were excluded, and 32 potentially eligible articles were selected. After further assessment of the 32 studies, 23 articles were excluded owing to lack of usable data or information regarding LNM or survival outcomes. Finally, a total of nine articles met the selection criteria 18, 19, 20, 21, 22, 23, 24, 25, 26.

The studies, which were published between 2016 and 2017, comprised data for a total of 808 patients; the mean patient sample size was 89.78 (range, 35–182). Among the nine studies, one focused on CC, one on GBC, one on GC, one on clear cell RCC, two on LAD, one on HCC, one on RCC, and one on TSCC. The expression of linc00152 was detected by real‐time quantitative polymerase chain reaction (RT‐qPCR) and standardized to that of glyceraldehyde‐3‐phosphate dehydrogenase or β‐actin. All patients described in the retrieved articles were divided into two groups based on high or low expression of linc00152. The main features of the eligible studies are shown in Table 1. The NOS confirmed that all studies were of good quality (supplemental online Table 1).

Table 1.

Characteristics of studies in this meta‐analysis

Study	Year	Country	Cancer type	Total no.	Tumor stage	Detection method	Cutoff	linc00152 expression				Survival analysis	Multivariate analysis	HR statistic	HR (95% CI)	Follow‐up period, months
								High expression	High with LNM	Low expression	Low with LNM
Yue	2016	China	CC	134	NR	RT‐qPCR	Median	96	NR	38	NR	OS/DFS	Yes	Rep	3.67 (1.64–8.21)	80
Cai	2016	China	GBC	35	I–IV	RT‐qPCR	Median	18	NR	17	NR	OS	NR	SC	2.09 (1.30–3.36)	<30
Chen	2016	China	GC	97	I–IV	RT‐qPCR	Median	49	43	48	34	OS/DFS	Yes	Rep	1.659 (1.008–2.731)	60
Wu	2016	China	ccRCC	77	I–IV	RT‐qPCR	Median	38	NR	39	NR	OS	Yes	Rep	2.577 (1.233–5.387)	>60
Chen	2017	China	LAD	60	I–III	RT‐qPCR	Median	30	25	30	13	OS/PFS	NR	SC	2.09 (1.34–3.25)	30
Deng	2017	China	HCC	77	I–IV	RT‐qPCR	NR	39	NR	38	NR	OS	NR	SC	1.67 (1.25–2.24)	60
Feng	2017	China	LAD	101	NR	RT‐qPCR	Mean	51	NR	50	NR	OS	NR	SC	2.27 (1.40–3.68)	60
Wang	2017	China	RCC	45	I–IV	RT‐qPCR	Mean	27	19	28	7	OS	NR	SC	2.04 (1.25–3.33)	60
Yu	2017	China	TSCC	182	I–IV	RT‐qPCR	Median	95	45	87	28	OS/RFS	NR	SC	2.04 (1.33–3.13)	>60

Abbreviations: CC, colon cancer; ccRCC, clear cell renal cell carcinoma; CI, confidence interval; DFS, disease‐free survival; GBC, gallbladder cancer; GC, gastric cancer; HCC, hepatocellular carcinoma; HR, hazard ratio; LAD, lung adenocarcinoma; LNM, lymph node metastasis; NR, no report; OS, overall survival; PFS, progression‐free survival; RCC, renal cell carcinoma; Rep, report; RFS, recurrence‐free survival; RT‐qPCR, real‐time quantitative polymerase chain reaction; SC, survival curve; TSCC, tongue squamous cell carcinoma.

Synthesized Findings

Association Between linc00152 and Clinicopathological Features

Data from four articles for a total of 384 patients were selected for analysis of the association between linc00152 and LNM in a random‐effects model (I² = 48.8%, p = .119). The OR was 2.93 with a 95% CI of 1.88–4.57 (p < .0001; supplemental online Fig. 1). The analysis suggested that high linc00152 expression indicates a higher propensity to develop LNM.

An association between high expression of linc00152 and advanced clinical stage was also found (OR = 2.43, 95% CI: 1.51–3.92, p < .0001) through application of the random‐effects model (I² = 61.7%, p = .033; Table 2). The heterogeneity of this analysis was significantly diminished after the study by Cai et al. 18 was excluded (I² = 0, p = .629), without affecting the results (OR = 3.12, 95% CI: 1.88–5.18, p < .0001; supplemental online Fig. 2).

Table 2.

Results of subgroup analysis of pooled hazard ratios of overall survival of patients with overexpression of linc00152

Stratified analysis	No. of studies	No. of patients	Pooled HR (95% CI)	p value	Heterogeneity
					I2, %	p value	Model
Cancer type
Digestive system	5	525	1.88 (1.55– 2.28)	<.0001	0.0	.422	Fixed effects
Urogenital system	2	122	2.19 (1.45– 3.29)	<.0001	0.0	.598	Fixed effects
Respiratory system	2	161	2.17 (1.56– 3.03)	<.0001	0.0	.814	Fixed effects
Sample size
>100	3	417	2.30 (1.70– 3.11)	<.0001	0.0	.446	Fixed effects
<100	6	391	1.88 (1.57– 2.25)	<.0001	0.0	.832	Fixed effects
Follow‐up time
≥5	7	713	1.95 (1.64– 2.32)	<.0001	0.0	.561	Fixed effects
<5	2	95	2.10 (1.51– 2.90)	<.0001	0.0	1.000	Fixed effects

Abbreviations: CI, confidence interval; HR, hazard ratio.

A total of 137 patients in two studies were included to examine the relationship between linc00152 expression levels and tumor size. An association was found between high expression of linc00152 and tumor size (>5 cm; OR = 5.12, 95% CI: 2.35–11.57) in a fixed‐effects model (I² = 0, p = .719; supplemental online Table 2).

Two studies reporting on a total of 316 patients were selected for analysis of the association between linc00152 and tumor recurrence. The fixed‐effects model was adopted as a result of low heterogeneity (I² = 0, p = .451). linc00152 expression was significantly associated with tumor recurrence (OR = 3.32, 95% CI: 1.98–5.57, p < .0001; supplemental online Table 2).

Association Between linc00152 and Disease‐Free Survival

Two studies reporting a total of 231 patients were selected for analysis of the association between linc00152 and DFS. The random‐effects model was adopted as a result of significant heterogeneity (I² = 75.8%, p = .042). In this analysis, linc00152 expression was significantly associated with DFS (HR = 1.66, 95% CI: 1.20–2.29, p < .0001; supplemental online Fig. 3).

Association Between linc00152 and Overall Survival

To evaluate the association between the expression level of linc00152 and OS in patients with cancer, data for pooled HRs and 95% CIs for OS were collected from the nine studies with a total of 808 patients. There was a significant association between high linc00152 expression and poor OS in patients with cancer (pooled HR = 1.98, 95% CI: 1.70–2.31, p < .0001) through the application of the fixed‐effects model (I² = 0%, p = .756; Fig. 2).

Figure 2.

Forest plot of the pooled HRs of elevated linc00152 expression for overall survival in different cancer types.

Abbreviations: CI, confidence interval; HR, hazard ratio.

Although there was no significant interstudy heterogeneity, a subgroup analysis via the fixed‐effects model was performed according to cancer type, sample size, and follow‐up interval (Table 2). High expression of linc00152 was significantly associated with shorter OS in patients with cancers of the digestive system (HR = 1.88, 95% CI = 1.55–2.28, p < .0001), urogenital system (HR = 2.19, 95% CI: 1.45–3.29, p < .0001), and respiratory system (HR = 2.17, 95% CI: 1.56–3.03, p < .0001; supplemental online Fig. 4). There was a significant association between high linc00152 expression and OS in studies of all sample sizes (n <100: HR = 1.88, 95% CI: 1.57–2.25, p < .0001; n >100: HR = 2.30, 95% CI: 1.70–3.11, p < .0001; supplemental online Fig. 5). Furthermore, high expression of linc00152 was significantly associated with poor OS of patients with cancer regardless of follow‐up interval (≥5 years: HR = 1.95, 95% CI: 1.64–2.32, p < .0001; <5 years: HR = 2.10, 95% CI: 1.51–2.90, p < .0001; supplemental online Fig. 6; Table 2).

Risk of Bias

A funnel plot and Egger's test were performed to assess publication bias. The shape of the funnel plot was roughly symmetrical, without obvious evidence of asymmetry (Fig. 3). Because of the small sample size used to determine LNM, TNM, DFS, and other factors, publication bias analysis was not conducted in these groups.

Figure 3.

Funnel plot of the publication bias for the analysis of the independent role of linc00152 in overall survival in the different cancer types.

Sensitivity Analysis

To evaluate the impact of a single study on the overall meta‐analysis, sensitivity analysis was performed by omitting each eligible study at a time. When each study was sequentially excluded, the results of the analysis were not significantly affected (supplemental online Fig. 7).

Analysis of linc00152‐Related Genes

We selected the top 100 coexpressed genes of linc00152 from the MEM‐Multi Experiment Matrix database and their correlations were analyzed (supplemental online Fig. 8). We found that galectin 1 (LGALS1), annexin A1 (ANXA1), and annexin A2 (ANXA2) ranked as the top three predicted target genes in terms of the p value in supplemental online Figure 7, which indicates that they were significantly related to linc00152 gene expression. We then analyzed the GO and KEGG pathway (Fig. 4; supplemental online Table 3). Signal pathway network construction is shown in Figure 5. Furthermore, we analyzed the proteins that might interact with linc00152 (supplemental online Tables 4, 5).

Figure 4.

GO terms and the KEGG pathway. (A): GO enrichment of target genes in biological process ontology. (B): GO enrichment of target genes in cellular component ontology. (C): GO enrichment of target genes in molecular function ontology (p < .05). (D): The top 10 pathways related to the differentially expressed genes by the KEGG database analysis.

Abbreviations: BP, biological process; CC, cellular component; GO, gene ontology analysis; KEGG, Kyoto Encyclopedia of Genes and Genomes; MF, molecular function.

Figure 5.

Differentially expressed gene interaction network analysis. Green nodes represent target genes and blue nodes represent the related pathway. As indicated in red, linc00152 localized at the center of the network.

Discussion

Summary of Main Findings

This meta‐analysis found that overexpression of linc00152 was related to poor clinical outcomes in patients with cancer. We investigated the association between the expression of linc00152 and clinicopathological features, and our meta‐analysis showed that overexpression of linc00152 was associated with LNM (OR = 2.93, 95% CI: 1.88–4.57, p < .0001) and advanced clinical stage (OR = 2.43, 95% CI: 1.51–3.92, p < .0001). However, there was extensive heterogeneity between linc00152 expression level and clinical stage. Because of the heterogeneity, which may have affected the results of the meta‐analysis, we performed a sensitivity analysis, and the existing heterogeneity significantly decreased. This analysis showed that the heterogeneity was significantly diminished after the 2016 study by Cai et al. was excluded (I² = 0, p = .629), and the results were not altered (OR = 3.12, 95% CI: 1.85–5.18, p < .0001). This means that this study had a significant effect on heterogeneity when assessing the relationship of linc00152 with clinical stage. We believe that the heterogeneity may have been caused by the following factors: (a) the sample size and sample collection of the study; (b) the different reaction conditions or reaction system procedures of RT‐qPCR; and (c) the difference in cutoff value.

When the association between linc00152 and tumor size was explored, we found that high expression of linc00152 was related to tumor size (>5 cm; OR = 5.12, 95% CI: 2.35–11.57, p < .001). Moreover, high expression of linc00152 was also positively correlated with tumor recurrence (OR = 3.32, 95% CI: 1.98–5.57, p < .0001). We also found that higher expression of linc00152 was positively related to poor DFS (HR = 1.66, 95% CI: 1.20–2.29, p < .0001). Furthermore, we analyzed the relationship between high expression of linc00152 and OS and found that patients with high linc00152 expression had a shorter OS (HR = 1.98, 95% CI: 1.70–2.31, p < .0001). A subgroup analysis indicated that there was a significant relationship between linc00152 and OS in tumors in the digestive system (HR = 1.88, 95% CI = 1.55–2.28, p < .0001), urogenital system (HR = 2.19, 95% CI: 1.45–3.29, p < .0001), and respiratory system (HR = 2.17, 95% CI: 1.56–3.03, p < .0001). A significant association was also found between linc00152 expression and OS, regardless of sample size or median follow‐up interval. Taken together, our data indicate that high expression of linc00152 serves as a potential predictive factor for lymph node metastasis and poor prognosis in various human cancers.

Research has shown that linc00152 is upregulated in a variety of malignant tumors and is thought to act as a novel oncogene in tumorigenesis and progression, promoting tumor cell proliferation and metastasis, blocking apoptosis, and decreasing survival 18, 19, 20, 21, 23. Although the role of linc00152 in tumorigenesis is still uncertain, a number of studies have highlighted important features of linc00152. Chen et al. 19 found that high expression of linc00152 in GC was significantly associated with tumor invasion depth, LNM, higher TNM stage, and poor survival. They also reported that linc00152 promoted cell cycle dysregulation via downregulation of p15 and p21 through binding to enhancer of zeste homolog 2 (EZH2). Cai et al. 18 found linc00152 may be acting as a competing endogenous RNA (ceRNA) for microRNA (miR)‐138, which can significantly stimulate GBC cell invasion, migration, and epithelial‐mesenchymal transition (EMT) via the linc00152/miR‐138/hypoxia inducible factor‐1a pathway. Moreover, in RCC, high expression of linc00152 is significantly related to advanced TNM stage, LNM, and poor OS 21. It has also been shown that linc00152 promotes RCC progression through suppression of p16 by increasing the presence of EZH2, LSD1 zinc finger family protein, and histone H3 lysine 27 methylation at the p16 promoter, as well as through interacting with miR‐205 20, 21. Furthermore, linc00152 was also found to stimulate tumor progression in CC, where it conferred resistance to oxaliplatin (L‐OHP)‐induced apoptosis. linc00152 is believed to function as a ceRNA to regulate miR‐193‐3p and erb‐b2 receptor tyrosine kinase 4 (ERBB4) expression, as downregulating ERBB4 expression causes decreased serine/threonine kinase (AKT) phosphorylation, which would result in decreased L‐OHP resistance. Because AKT is the most frequently hyperactivated signaling pathway in most tumors, and this pathway mediates tumor cell survival, the linc00152/miR‐193a‐3p/ERBB4/AKT signaling axis might be a mechanism of drug resistance in CC and therefore a potential therapeutic target 22, 30, 31. In HCC, linc00152 expression was positively related to the expression of hepatitis B virus X protein, and EMT in hepatocellular carcinoma cells might be stimulated by high expression of linc00152 via its binding to EZH2 25. Moreover, the transcription of IL‐24 could be inhibited by linc00152 via interaction with EZH2 23, 24. Downregulating linc00152 expression could normalize the expression of oncogenes such as mitogen activated protein kinase p38a (p38a); signal transducer and activator of transcription 1 (STAT1);signal transducer and activator of transcription 3 (STAT3); cAMP responsive element binding protein 1 (CREB1); cyclin E1 (CCNE1); transcriptional regulator Myc‐like (c‐MYC).

To further explore the value of linc00152, we performed the target gene prediction and signal pathway analysis of linc00152 using the starBase v2.0 database and the MEM‐Multi Experiment Matrix database. Moreover, the GO and KEGG were carried out. The result of our analysis showed that LGALS1, ANXA1, and ANXA2, which have an important roles in multiple tumors including clear cell renal cell carcinoma 32, pancreatic cancer 33, acute lymphoblastic leukemia 34, breast cancer 35, and oral squamous cell carcinoma 36, were significantly related to linc00152 gene expression (P < 4.12 × 10⁻⁶⁰). Following this, we performed the GO and KEGG pathway analysis, and the results showed that linc00152 was significantly related to negative regulation of apoptotic processes. linc00152 was also associated with extracellular exosomes and had phospholipase inhibitor activity. For the KEGG pathway, linc00152 was particularly related to proteoglycans in cancer. Furthermore, we constructed a signal pathway network by using Cytoscape (Fig. 5) and detected the potential RNA‐binding protein associated with linc00152 by using starBase v2.0 (supplemental online Table 4).

Limitations

Several limitations in this meta‐analysis should be pointed out. First, there were only nine studies in this analysis, most of which had a small sample size. Second, the scope of our research results might be limited because all the studies are from China. Third, the accuracy of the results might be affected by estimating the HR and 95% CIs from Kaplan‐Meier curves for some studies. Fourth, many of the articles were positive results, as negative results are much more difficult to publish. Fifth, the different cutoff values of linc00152 in the high and low expression groups might cause heterogeneity. Sixth, more experiments are required to validate the target genes.

Conclusion

Overexpression of linc00152 was significantly related to LNM, TNM stage, and poor prognosis. Our results showed that linc00152 might act as a novel biomarker of LNM and may be used as a prognostic indicator for patients with cancer. However, studies with larger sample sizes and other ethnic groups need to be performed to affirm its prognostic significance in human cancers.

Author Contributions

Conception/design: Aifa Tang

Collection and/or assembly of data: Han Wang, Yang Liu

Manuscript writing: Han Wang, Yang Liu

Final approval of manuscript: Han Wang, Yang Liu, Aifa Tang

Disclosures

The authors indicated no financial relationships.

Supporting information

See http://www.TheOncologist.com for supplemental material available online.

Supplemental Figures

Supplemental Tables