Long non-coding RNA urothelial carcinoma–associated 1 as a tumor biomarker for the diagnosis of urinary bladder cancer

Abstract

To investigate the diagnostic value of urothelial carcinoma–associated 1 as a urine biomarker in urinary bladder cancer patients by performing a comprehensive meta-analysis. A comprehensive literature search was conducted by the databases PubMed, Embase, Cochrane Library, China Knowledge Resource Integrated, and Web of Science. The quality of eligible studies was scored with the Quality Assessment of Diagnostic Accuracy Studies. The bivariate meta-analysis model was used to pool the sensitivity, specificity, likelihood ratio, and diagnostic odds ratio. Receiver operating characteristic curves and hierarchical summary receiver operating characteristic models were employed to check the overall test performance in this meta-analysis. Seven publications involving 678 patients and 563 controls were included in this meta-analysis. The pooled sensitivity was 0.84 (95% confidence interval: 0.80–0.88), specificity was 0.87 (95% confidence interval: 0.75–0.94), positive likelihood ratio was 6.5 (95% confidence interval: 3.10–13.62), negative likelihood ratio was 0.18 (95% confidence interval: 0.13–0.25), and diagnostic odds ratio was 36 (95% confidence interval: 13–99). The area under the summary receiver operating characteristic curve was 0.89 (95% confidence interval: 0.86–0.91). Our results indicated that urothelial carcinoma–associated 1 was a potential diagnostic biomarker with good specificity and sensitivity in urinary bladder cancer. Further prospective studies with larger cohorts are necessary to evaluate the diagnostic accuracy of urothelial carcinoma–associated 1 for urinary bladder cancer.

Keywords

Bladder cancer long non-coding RNA urothelial carcinoma–associated 1 diagnosis biomarker

Introduction

Urinary bladder cancer (UBC) is one of the most common malignancies with the 4th and 10th leading cause of cancer-related deaths in males and females, respectively.¹ In 2015, 74,000 new UBC cases in the United States with a male/female sex ratio of 2.5:1 were reported.² The histologic subtype of UBC is quite diverse, such as urothelial carcinoma, squamous cell carcinoma, adenocarcinoma, and small cell carcinoma. The most common subtype of UBC is transitional cell cancer or more properly urothelial cell carcinoma.³ However, the outcome of UBC patients is not favorable. The majority of patients with UBC could be diagnosed at early stages, but up to 50% cases recur after treatment, and 15%–40% cases progress to muscle-invasive diseases.⁴ Screening cystoscopy, voided urinary cytology (VUC), and random bladder biopsies remain the most commonly used methods for initial diagnosis of UBC.⁵ As their complex preparatory procedures, high cost or unreliable result of single approach; for example, the sensitivity of cytology was 0.51.⁶ Thus, an urgent requirement is needed to identify an alternative urine biomarker to the current diagnostic system.

Long non-coding RNAs (lncRNAs) are longer than 200 nucleotides encoding no protein, which are involved in a variety of biological functions and are dysregulated in various cancers.⁷ The aberrant expression of lncRNAs is associated with a broad range of cellular processes, such as cell growth, survival, migration, invasion, and differentiation.^8,9 In recent years, increasing studies have highlighted the role of lncRNAs in tumorigenesis and suggested that this group of genes might be used as biomarkers in cancer.¹⁰

The human urothelial carcinoma–associated 1 (UCA1), a 2314-bp lncRNA located on human chromosome 19,¹¹ was upregulated in many cancers, such as hepatocellular cancer, colorectal cancer, gastric cancer, esophageal squamous cell carcinoma, and lung cancer.^12–16 Especially, it was first identified in bladder cancer and thought to be involved in embryogenesis and UBC progression.¹⁷ So, we analyzed transcriptome profiling in human bladder cancer tissues and normal tissue from The Cancer Genome Atlas (TCGA), aiming to validate UCA1 differentially expressed in UBC.

We wonder whether UCA1 differentially expressed in urine sediments so that it could be used as a novel urine biomarker for the detection of UBC. As we found that recent studies have assessed the diagnostic value of UCA1 in UBC, the inconsistencies or discrepancies about diagnostic accuracy exist.^6,17–22 For example, Wang et al. showed that the specificity was 91.8% in the diagnosis of bladder cancer, but Milowich et al. showed the specificity was 64.6%. Therefore, we performed this meta-analysis to evaluate the diagnostic efficiency of UCA1 in patients with UBC.

Materials and methods

Identification of differentially expressed genes associated with bladder cancer

For identification of differentially expressed genes, we used edgeR algorithm²³ based on transcriptome profiling downloaded from TCGA. The Benjamini–Hochberg (BH) multiple testing method was used to adjust the p value. The thresholds were the false discovery rate (FDR) < 0.05 and |log2FC| > 2.

Publication search

PubMed, Embase, Cochrane Library, China Knowledge Resource Integrated (CNKI), and Web of Science databases were searched (up to 14 August 2016) for relevant articles that estimated the diagnostic value of UCA1 in UBC. The search terms for the literature retrieval were “UCA1” or “urothelial cancer associated 1” or “CUDR” or “UCAT1” or “LINC00178” or “onco-lncRNA-36” or “NCRNA00178” and “Bladder Neoplasm” or “Bladder Tumor” or “Bladder Cancer” or “Bladder carcinoma” or “Malignant Tumor of Urinary Bladder.” Reference lists of relevant articles were also reviewed to identify potential eligible studies.

Inclusion and exclusion criteria

All the eligible studies included in this meta-analysis had to meet the following criteria: (1) UBC was diagnosed by histopathological confirmation, (2) the expression of UCA1 in urine sediments was tested, (3) healthy individuals or patients with benign disease were included in the control group, and (4) sufficient data in the studies were enough to construct a two-by-two table. Publications were excluded based on any of the following criteria: (1) unqualified data, (2) duplicate publications, and (3) animal research, letters, editorials, expert opinions, case reports, or laboratory articles. The retrieved literature were independently assessed by two reviewers according to the preestablished criteria, and discrepancies were solved through discussions and consensus.

Data extraction and quality assessment

Data were retrieved from the eligible studies independently by three investigators, according to the inclusion and exclusion criteria. Disagreements were discussed with the corresponding author in conference. The characteristics of data extracted for this systematic review included first author, publication year, country of publication, cancer type, sample type, test method, sample size, reference test, cutoff value, and data for two-by-two tables (true positive (TP), false positive (FP), true negative (TN), and false negative (FN)).

The Quality Assessment of Diagnostic Accuracy Studies (QUADAS) list was used to systematically assess the quality of the articles included in this diagnostic meta-analysis.²⁴ Specifically, 14 items from the QUADAS list were applied to each article, with an answer of “yes,” “no,” or “unclear.” The answer “yes” obtained a score of 1, whereas “no” or “unclear” gained a score of 0 with a total score of 14.

Statistical analysis

Data analysis was performed using Stata 13.0 software.²⁵ For this diagnostic meta-analysis, we retrieved TP, FP, TN, and FN as bivariate data directly or through recalculation on the basis of relative data from each eligible study.²⁶ Subsequently, we calculated the sensitivity, specificity, diagnostic odds ratio (DOR), positive likelihood ratio (PLR), and negative likelihood ratio (NLR) and constructed their forest plots. Then, we generated summary receiver operator characteristic (SROC) curves analysis and calculated the area under the ROC curves (AUC) to evaluate the associations of lncRNA UCA1 and UBC.²⁷ In case of difference from different diagnostic threshold effect in the traditional SROC, we plotted study-specific sensitivity and specificity by applying the hierarchical summary ROC model (HSROC).²⁸

To detect cutoff threshold effects, the Spearman correlation coefficient was employed to assess the relationship between sensitivity and specificity. We used Cochran’s Q and I² statistics to evaluate the heterogeneity across this study, I² more than 50% indicated the existence of significant heterogeneity.²⁹ Random-effects model was employed when heterogeneity existed; otherwise, the fixed-effects model was applied.

Results

UCA1 expression levels in human bladder cancer tissue

An analysis was performed in the whole-genome gene expression profiles including 410 tumor tissues and 19 normal tissues from TCGA using edgeR algorithm. When comparing cancer tissues with normal tissues in the data set, we found 1156 differentially expressed genes being assigned for bladder cancer (Figure 1(a)) including UCA1. The cutoff value of the adjusted p values was set to 0.05 and the minimum fold change (FC) to 2. We identified that the expression level of UCA1 in bladder cancer tissues is higher than in normal bladder tissues (Figure 1(b)).

Figure 1.

(a) Hierarchical clustering of differentially expressed genes: rows represent genes and columns represent samples. For a gene, red represents higher expression level, blue represents lower expression level, and white represents medial expression level for all samples. For a sample, red means tumor tissues and blue means tumor adjacent tissues from patients with UBC. (b) Relative expression of UCA1 in the tumor tissue compared to normal tissues.

Literature search

A total of 5758 studies were initially retrieved. After excluding duplicates, 2011 articles were excluded. After screening the title and abstract carefully, 3703 articles were excluded for unrelated to UCA1, bladder cancer, reviews, letters, or not original articles. After further inspection of the full articles, 37 articles were excluded for lacking available diagnostic outcome or sufficient clinical data in their records. Finally, 7 articles were considered eligible for this meta-analysis upon the given criteria. The whole procedure of this study selection is shown in Figure 2.

Figure 2.

Flow diagram of study selection process based on inclusion and exclusion criteria.

Study characteristics and quality assessment

Seven publications involving 1241 cases and 563 controls were analyzed.^6,17–22 In these selected studies, all of the individuals had been histopathologically confirmed. And, the 563 controls included healthy volunteers and patients with benign disease that had never been diagnosed with malignancy. The studies enrolled in this systemic review were conducted in China, Belgium, Egypt, and India. These included publications published between 2006 and 2015 and investigated the diagnostic value of UCA1 as a urine biomarker for patients with UBC. Furthermore, all the studies used real-time quantitative polymerase chain reaction (RT-qPCR) to assess the expression of UCA1 in urine samples. The main characteristics of the eligible studies are presented in Table 1. We used the QUADAS assessment tool to assess the quality of the seven articles with 14 items. The analysis scores are shown in Table 2 with the total point 14.

Table 1.

Main characteristics of the eligible studies.

Study ID	First author	Year	Journal	Country	Cancer type	Sample	Test method	Reference test	Cutoff	Quality score	Cases/control	TP	FP	FN	TN
1	Wang¹⁷	2006	Clinical Cancer Research	China	Bladder TCC	Urine samples	RT-qPCR	Histology	ROC	10	179/85	76	7	18	78
2	Milowich¹⁹	2015	Springerplus	Belgium	Bladder cancer	Urine samples	RT-qPCR	Histology	NA	12	79/49	25	18	5	31
3	Eissa⁶	2015	Tumor Biology	Egypt	Bladder cancer	Urine samples	RT-qPCR	Histology	ROC	11	210/116	86	4	8	112
4	Eissa²⁰	2015	Biomarkers	Egypt	Bladder cancer	Urine samples	RT-qPCR	Histology	ROC	11	184/45	124	3	15	42
5	Srivastava²¹	2014	Tumor Biology	India	Bladder cancer	Urine samples	RT-qPCR	Histology	ROC	11	191/74	93	15	24	59
6	Fang²²	2012	J Mol Diagn Ther	China	Bladder cancer	Urine samples	RT-qPCR	Histology	NA	11	74/50	21	20	3	30
7	Zhang¹⁸	2012	Natl Med J China	China	Bladder cancer	Urine samples	RT-qPCR	Histology	ROC	11	324/144	152	11	28	133

NA: not available; TCC: transitional cell carcinoma; RT-qPCR: real-time quantitative polymerase chain reaction; ROC: receiver operating characteristic curve; TP: true positive; FP: false positive; TN: true negative; and FN: false negative.

Table 2.

QUADAS assessment for the eligible studies.

	1	2	3	4	5	6	7
1. Was the spectrum of patients representative of the patients who will receive the test in practice?	Y	Y	Y	Y	Y	Y	Y
2. Were selection criteria clearly described?	N	Y	Y	Y	Y	Y	Y
3. Is the reference standard likely to correctly classify the target condition?	Y	Y	Y	Y	Y	Y	Y
4. Is the time period between reference standard and index test short enough to be reasonably sure that the target condition did not change between the two tests?	Y	Y	Y	Y	Y	Y	Y
5. Did the whole sample or a random selection of the sample receive verification using a reference standard?	Y	Y	Y	Y	Y	Y	Y
6. Did patients receive the same reference standard regardless of the index test result?	Y	Y	Y	Y	Y	Y	Y
7. Was the reference standard independent of the index test (i.e. the index test did not form part of the reference standard)?	Y	Y	Y	Y	Y	Y	Y
8. Was the execution of the index test described in sufficient detail to permit replication of the test?	Y	Y	Y	Y	Y	Y	Y
9. Was the execution of the reference standard described in sufficient detail to permit its replication?	Y	Y	Y	Y	Y	Y	Y
10. Were the index test results interpreted without knowledge of the results of the reference standard?	U	U	U	U	U	U	U
11. Were the reference standard results interpreted without knowledge of the results of the index test?	U	U	U	U	U	U	U
12. Were the same clinical data available when test results were interpreted as would be available when the test is used in practice?	Y	Y	Y	Y	Y	Y	Y
13. Were uninterpretable/intermediate test results reported?	U	Y	U	U	U	U	U
14. Were withdrawals from the study explained?	Y	Y	Y	Y	Y	Y	Y

QUADAS: Quality Assessment of Diagnostic Accuracy Studies.

N for “no,” Y for “yes,” U for “unclear.”

Data analysis

The forest plot showed the sensitivity ranged from 0.79 to 0.91 (pooled: 0.84, 95% confidence interval (CI): 0.80–0.88), while specificity ranged from 0.60 to 0.97 (pooled: 0.87, 95% CI: 0.75–0.94). The heterogeneity existed in pooled sensitivity and specificity in the seven studies (I² = 39.45% and I² = 91.55%, respectively), suggesting significant heterogeneity in sensitivity and specificity (Figure 3).

Figure 3.

Forest plots of sensitivity and specificity for UCA1 test accuracy in the diagnosis of bladder cancer.

Since compared with the specificity and sensitivity, likelihood ratios (LRs) are considered to be more comprehensive and clinically meaningful. We calculated the PLR and NLR for the diagnostic performance of UCA1 (Figure 4(a)). The pooled PLR was 6.5 (95% CI: 3.10–13.62), indicating that the UBC patients have more than a six-fold probability to express UCA1 in comparison with healthy individuals. The pooled NLR was 0.18 (95% CI: 0.13–0.25), which means that the probability of individuals with UBC was 18% when the UCA1 test is negative. PLR > 10 or NLR < 0.1 suggests high diagnostic accuracy (Figure 4(b)). It was also noted that DOR was 36 (95% CI: 13–99), indicating that UCA1 can be used as a good biomarker for UBC diagnosis.

Figure 4.

(a) Forest plots of positive and negative likelihood ratios for UCA1 test accuracy in the diagnosis of bladder cancer and (b) pooled LNR and LPR of the diagnostic studies of UCA1 test.

The estimates of sensitivity and specificity of all the included studies were also shown in a SROC curve, together with the summary ROC point (pooled sensitivity against pooled specificity). The value of AUC was near to 1, which was considered statistically significant. The area under the summary ROC curve (AUC) was 0.89 (95% CI: 0.86–0.91). The 95% prediction contour and 95% confidence contour were also plotted (Figure 5(a)).

Figure 5.

(a) Summary receiver operating characteristic curves for UCA1 in the diagnosis of bladder cancer and (b) hierarchical summary receiver operating characteristics (HSROC) curve for UCA1 in the diagnosis of bladder cancer.

The HSROC curve of these selected studies was consistent with the results from the bivariate model. The summary operating point estimate of sensitivity and specificity was 0.84 (95% CI: 0.80–0.88) and 0.87 (95% CI: 0.75–0.94), respectively. The value of β was 2.48, and the p value was 0.013. This result showed that the HSROC was asymmetrical. The value of γ, which helps distinguish UBC patients from healthy individuals, was 4.32 (95% CI: 2.59–6.05). The 95% prediction region and 95% confidence region were also plotted. This result indicates that UCA1 is a relatively accurate diagnostic urine marker for UBC (Figure 5(b)).

To validate the stability of outcomes in this meta-analysis, sensitivity analysis showed no significant change after omitting any of the included studies (Table 3).

Table 3.

The influence of each trial for the outcome of the meta-analysis.

Study ID	First author	DOR	95% CI
1	Wang¹⁷	34	11-109
2	Milowich¹⁹	46	16-134
3	Eissa⁶	26	10-64
4	Eissa²⁰	29	10-84
5	Srivastava²¹	43	14-130
6	Fang²²	47	17-133
7	Zhang¹⁸	32	10-104
	Combined	36	13-99

DOR: diagnostic odds ratio; CI, confidence interval.

Threshold effect and heterogeneity

The threshold effect was the major cause of heterogeneity, which leads to differences in sensitivity and specificity. In this meta-analysis, the Spearman correlation coefficient of logarithm sensitivity and 1-specificity was −0.46 with p = 0.30 (p > 0.05), suggesting that the heterogeneity was not result of the threshold effect. The I² of the heterogeneity test was 87%, indicating significant heterogeneity. Because there were only seven articles included, meta-regression analysis and subgroup analysis cannot be used to search sources of heterogeneity.

Discussion

Recent studies have revealed several novel diagnostic biomarkers for UBC, including UCA1 as a urine biomarker. UCA1 is an lncRNA,¹⁰ which is localized to nuclear speckles. The UCA1 transcript is a non-coding RNA with more than 1400 nt, which expressed from chromosome 19p13.12.³⁰ For the reason that UCA1 expression sequences were conserved across several species, it had been indicated to play a regulatory role in the molecular biology of cancer and other human diseases.¹⁰ Now, with the widespread application of new technologies, a growing number of studies have proven that UCA1 is upregulated in several cancers. Abnormal UCA1 expression was strongly correlated to clinicpathologic characteristics, including lymph node metastasis, chemoresistance, and overall survival.^31,32 Recently, UCA1 has been found to be required for mitotic progression and controls cell cycle progression in human cells.^30,33 In bladder cancer, UCA1 level was upregulated compared with adjacent normal tissues. Yang et al.³⁴ found that UCA1 regulates cell cycle progression through cAMP response element–binding protein (CREB) via PI3K-AKT–dependent signaling pathways. Zhen et al.³⁵ demonstrated that UCA1 has an oncogenic role in promoting bladder cancer progression. Wang et al.¹⁰ showed that UCA1 expression increases invasion and drug resistance of bladder cancer cells.

The UCA1 has attracted the researchers’ attention not only because of the application as biomarkers for cancer prognosis and predictors for metastasis³³ but also because of the potential use as a valuable diagnostic urine biomarker. In early studies, the associations between UBC and several biomarkers such as nuclear matrix protein 22 (NMP22),³⁶ microRNA-96 (miR-96),³⁷ soluble fragments of cytokeratin 19 (CYFRA 21-1)³⁸ were identified. However, the diagnostic value of UBC was unsatisfactory (the specificity or specificity was low). The studies included in our meta-analysis used RT-qPCR to test urine UCA1 for detecting UBC, which was a non-invasive way with good sensitivity and specificity. In combination with cytology, it may improve the detection of UBC with a non-invasive diagnostic test and may increase the interval between cystoscope and replace cystoscopies altogether.²¹ Wang et al.¹⁷ showed that a UCA1 assay was highly specific (91.8%, 78 of 85) and very sensitive (80.9%, 76 of 94) in the diagnosis of bladder cancer. However, Milowich et al.¹⁹ indicated that the efficiency of the UCA1 test for detecting primary and recurring bladder cancer was unsatisfactory, the sensitivity was 83.9%, and the specificity was 64.6%. Thus, the present meta-analysis was performed to provide an integrated and up-to-date evaluation of the diagnostic and clinical values of UCA1 as a urinary marker for UBC.

In this meta-analysis, the combined sensitivity and specificity were 0.84 (95% CI: 0.80–0.88) and 0.87 (95% CI: 0.75–0.94), respectively, which implied good sensitivity and specificity. The DOR value is defined as the ratio of TP odds to FP odds reflecting the extent of the association between the diagnostic results and the disease. The DOR value of UCA1 was 36, suggesting that UCA1 can serve as a promising diagnostic biomarker for UBC. In addition, LR combines the stability of sensitivity and specificity to provide an omnibus index of test performance and is more clinically valuable than DOR.³⁹ PLR greater than 10 or NLR less than 0.1 generate large and often conclusive changes from pre-test to post-test probability. Due to the threshold effect among studies, the SROC curve appears to show a better method to assess the summary diagnostic accuracy of the discrimination between cases and controls rather than pooled sensitivity, specificity, PLR, or DOR. In this study, the AUC was 0.89, suggesting a relatively high level of diagnostic accuracy.

Heterogeneity is a potential problem that can influence the incorporation effect and the interpretation of the meta-analysis results.⁴⁰ Although we have set strict inclusion and exclusion criteria to gain eligible studies, heterogeneity still exists due to the existence of potential confounding factors. A prime cause of heterogeneity is the threshold effect, which arises because some studies did not supply the cutoff values. So, Spearman rank correlation test was employed to analyze the threshold effect, and the Spearman correlation coefficient was −0.46 with p = 0.30 (p > 0.05). This result demonstrated that threshold effect was not a certain source of heterogeneity. As there were only seven articles included, meta-regression analysis and subgroup analysis cannot be used. Therefore, the other factors such as ethnicity and test method were not evaluated as sources of heterogeneity.

Zhen et al.³⁵ also conducted an impressive meta-analysis on the diagnostic value of UCA1 for UBC. They reported a pooled sensitivity of 83% and a pooled specificity of 83%. Their results gave support to our conclusion. In comparison with their included data, our data of included original studies were only from screening groups (data of follow-up were not included), which aim to present UCA1 as a diagnostic test in a screening population (for detecting primary BC). Moreover, in this present meta-analysis, HSROC curve, sensitivity analysis, and Spearman rank correlation were performed.

Several limitations in the present meta-analysis should be emphasized. First, only seven studies were included in this meta-analysis; Deeks’ funnel plots cannot be used to assess the presence of publication bias in this meta-analysis, it has weakened the reliability of this meta-analysis’ results to some extent. So, larger-size and better design studies are needed to confirm our results. Second, not all of the studies reported the cutoff values of UCA1. Further comprehensive studies are needed to solve this problem. Third, only publications in English or Chinese were brought into, researches in other languages might be missed. Fourth, the blind detection and blind judgment methods can minimize the tendencies of diagnostic trials; however, all the studies did not report using blind detection.

Conclusion

Our comprehensive analysis validates that UCA1 may act as a clinical potential urine biomarker in UBC diagnosis with a high AUC value. Furthermore, combining UCA1 with other conventional tests may improve the detection of UBC.

Footnotes

Acknowledgements

C.Z. conceived and designed the experiments. Z.W. and X.W. performed the experiments, analyzed the data, and drafted the paper. D.Z., Y.Y., and L.C. contributed to the reagents/materials/analysis tools.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Ethical standard

As the study was a systemic review of published data and meta-analysis of the pooled data, we did not apply for the approval of Institutional Review Board.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by the National Natural Science Foundation of China (grant numbers 81171996 and 81272289), the Wujieping Medical Foundation (grant number 320.6750.13252), and Heilongjiang Province Health and Family Planning Commission Science Foundation (grant number 2014-287).

References

Chen

Chong

Yin

. Molecular events are associated with resistance to vinblastine in bladder cancer. Cell Mol Biol 2015; 61: 33–38.

Siegel

Miller

Jemal

. Cancer statistics, 2015. CA Cancer J Clin 2015; 65: 5–29.

Chalasani

Chin

Izawa

. Histologic variants of urothelial bladder cancer and nonurothelial histology in bladder cancer. Can Urol Assoc J 2009; 3: S193–S198.

Amin

Smith

Reuter

. Update for the practicing pathologist: the international consultation on urologic disease-European association of urology consultation on bladder cancer. Mod Pathol 2015; 28: 612–630.

Siegel

Zou

. Cancer statistics, 2014. CA Cancer J Clin 2014; 64: 9–29.

Eissa

Matboli

Essawy

. Integrative functional genetic-epigenetic approach for selecting genes as urine biomarkers for bladder cancer diagnosis. Tumour Biol 2015; 36: 9545–9552.

Kapranov

Cheng

Dike

. RNA maps reveal new RNA classes and a possible function for pervasive transcription. Science 2007; 316: 1484–1488.

Wilusz

Sunwoo

Spector

. Long noncoding RNAs: functional surprises from the RNA world. Genes Dev 2009; 23: 1494–1504.

Xue

. An expanding universe of the non-coding genome in cancer biology. Carcinogenesis 2014; 35: 1209–1216.

10.

Wang

Xie

. UCA1, a non-protein-coding RNA up-regulated in bladder carcinoma and embryo, influencing cell growth and promoting invasion. FEBS Lett 2008; 582: 1919–1927.

11.

Xue

Chen

. Urothelial cancer associated 1: a long noncoding RNA with a crucial role in cancer. J Cancer Res Clin Oncol 2016; 142: 1407–1419.

12.

Wang

Ying

. Upregulated lncRNA-UCA1 contributes to progression of hepatocellular carcinoma through inhibition of miR-216b and activation of FGFR1/ERK signaling pathway. Oncotarget 2015; 6: 7899–7917.

13.

Bian

Jin

Zhang

. LncRNA-UCA1 enhances cell proliferation and 5-fluorouracil resistance in colorectal cancer by inhibiting miR-204-5p. Sci Rep 2016; 6: 23892.

14.

Guo

. Increased urothelial cancer associated 1 is associated with tumor proliferation and metastasis and predicts poor prognosis in colorectal cancer. Int J Oncol 2015; 47: 1329–1338.

15.

Zhang

. Overexpression of long non-coding RNA UCA1 predicts a poor prognosis in patients with esophageal squamous cell carcinoma. Int J Clin Exp Pathol 2014; 7: 7938–7944.

16.

Zheng

Dai

. Aberrant expression of UCA1 in gastric cancer and its clinical significance. Clin Trans Oncol 2015; 17: 640–646.

17.

Wang

Zhang

Wang

. Rapid identification of UCA1 as a very sensitive and specific unique marker for human bladder carcinoma. Clin Cancer Res 2006; 12: 4851–4858.

18.

Zhang

Hao

Zhang

. Evaluation of novel gene UCA1 as a tumor biomarker for the detection of bladder cancer. Zhonghua Yi Xue Za Zhi 2012; 92: 384–387.

19.

Milowich

Le Mercier

De Neve

. Diagnostic value of the UCA1 test for bladder cancer detection: a clinical study. Springerplus 2015; 4: 349.

20.

Eissa

Matboli

Essawy

. Rapid detection of urinary long non-coding RNA urothelial carcinoma associated one using a PCR-free nanoparticle-based assay. Biomarkers 2015; 20: 212–217.

21.

Srivastava

Singh

Rath

. Appraisal of diagnostic ability of UCA1 as a biomarker of carcinoma of the urinary bladder. Tumour Biol 2014; 35: 11435–11442.

22.

Fang

Fan

. Establishment of method about detecting UCA1 mRNA through real-time fluorescence quantitative PCR reverse transcription polymerase chain reaction and its clinical assessment. J Mol Diag Ther 2012; 4: 171–176.

23.

Robinson

McCarthy

Smyth

. edgeR: a bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 2010; 26: 139–140.

24.

Whiting

Rutjes

Reitsma

. The development of QUADAS: a tool for the quality assessment of studies of diagnostic accuracy included in systematic reviews. BMC Med Res Methodol 2003; 3: 25.

25.

Zamora

Abraira

Muriel

. Meta-DiSc: a software for meta-analysis of test accuracy data. BMC Med Res Methodol 2006; 6: 31.

26.

Reitsma

Glas

Rutjes

. Bivariate analysis of sensitivity and specificity produces informative summary measures in diagnostic reviews. J Clin Epidemiol 2005; 58: 982–990.

27.

Moses

Shapiro

Littenberg

. Combining independent studies of a diagnostic test into a summary ROC curve: data-analytic approaches and some additional considerations. Stat Med 1993; 12: 1293–1316.

28.

Rutter

Gatsonis

. A hierarchical regression approach to meta-analysis of diagnostic test accuracy evaluations. Stat Med 2001; 20: 2865–2884.

29.

Dinnes

Deeks

Kirby

. A methodological review of how heterogeneity has been examined in systematic reviews of diagnostic test accuracy. Health Technol Assess 2005; 9: 1–113, iii.

30.

Xie

Wang

. Cellular localization and tissue expression pattern of UCA1, a non-coding RNA. Nan Fang Yi Ke Da Xue Xue Bao 2010; 30: 57–60.

31.

Yang

Jiang

Wan

. UCA1 functions as a competing endogenous RNA to suppress epithelial ovarian cancer metastasis. Tumour Biol 2016; 37: 10633–10641.

32.

Zhang

Cao

Zhang

. UCA1 overexpression predicts clinical outcome of patients with ovarian cancer receiving adjuvant chemotherapy. Cancer Chemother Pharmacol 2016; 77: 629–634.

33.

Kaneko

Ito

Ono

. Gene expression profiling reveals upregulated UCA1 and BMF in gallbladder epithelia of children with pancreaticobiliary maljunction. J Pediatr Gastroenterol Nutr 2011; 52: 744–750.

34.

Yang

Wang

. Long non-coding RNA UCA1 regulated cell cycle distribution via CREB through PI3-K dependent pathway in bladder carcinoma cells. Gene 2012; 496: 8–16.

35.

Zhen

Hua

Liu

. Inhibition of long non-coding RNA UCA1 by CRISPR/Cas9 attenuated malignant phenotypes of bladder cancer. Oncotarget 2017; 8: 9634–9646.

36.

Huber

Schwentner

Taeger

. Nuclear matrix protein-22: a prospective evaluation in a population at risk for bladder cancer. Results from the UroScreen study. BJU Int 2012; 110: 699–708.

37.

Wang

Luo

. hsa-miR-96 up-regulates MAP4K1 and IRS1 and may function as a promising diagnostic marker in human bladder urothelial carcinomas. Mol Med Rep 2012; 5: 260–265.

38.

Fatela-Cantillo

Fernandez-Suarez

Menendez

. Low utility of CYFRA 21-1 serum levels for diagnosis and follow-up in bladder cancer patients. J Clin Lab Anal 2005; 19: 167–171.

39.

Gallagher

. Clinical utility of likelihood ratios. Ann Emerg Med 1998; 31: 391–397.

40.

Lijmer

Bossuyt

Heisterkamp

. Exploring sources of heterogeneity in systematic reviews of diagnostic tests. Stat Med 2002; 21: 1525–1537.