Background
Both the number of cancer patients and the mortality rate are disturbingly increasing. Cancer has become a common disease that is seriously detrimental to human health, which is a significant cause of death in many countries around the world. Despite the dramatic developments in the diagnosis and therapy of tumors over the past few decades, the overall survival (OS) of patients remains unsatisfactory. Tumor markers play a significant role in monitoring and treating tumors. However, fewer tumor markers were used in clinical diagnosis. Therefore, it is urgent need to discover molecular biomarkers to improve the sensitivity and specificity for the detection and prognosis for cancer.
With the development of high-throughput sequencing technology, an increasing number of long non-coding RNAs (lncRNAs) have been gradually discovered and become the hotspot of research. LncRNAs, which cannot encode proteins, are important members of the non-coding RNA family. The biological functions of lncRNAs are still in its infancy and no definitive conclusion has been reached on its function and clinical significance of lncRNA. Recently, accumulating evidences have indicated that lncRNAs were closely related to initiation and progression of human diseases, especially cancer. LncRNAs could be used as a carcinogen or suppressor gene in the development and progression of cancer.
X-inactive specific transcript (XIST) is a kind of lncRNA derived from XIST gene that only expressed from the inactive X chromosome [
1,
2]. Many clinical studies have clarified that the expressions of lncRNA XIST not only played an important role in the differentiation, proliferation, and genome maintenance of cells, but also with the development and progression of cancer [
3]. For instance, the perturbation of lncRNA XIST expression related to metastasis and recurrence in a variety of cancers, including bladder cancer [
4], nasopharyngeal carcinoma (NPC) [
5], pancreatic cancer (PC) [
6], colorectal cancer (CRC) [
7‐
9], glioma [
10,
11], prostate cancer (PCa) [
12], ovarian cancer, gastric cancer (GC) [
13,
14], hepatocellular carcinoma (HCC) [
15,
16], and non-small cell lung cancer (NSCLC) [
17,
18]. Nevertheless, the consistency and magnitude of the prognostic impact of lncRNA XIST remains enigmatic, and the prognostic value of lncRNA XIST expression in different tumor types remains still controversial. To verify its clinical relevance, we integrated all published evidence systematically in this meta-analysis to reveal the prognostic value of lncRNA XIST in various types of solid tumors.
Materials and methods
Search strategy
A systematic review of the literature was conducted according to the PRISMA guidelines. PubMed, EMBase, Web of Science, Cochrane library were searched to evaluate the impact of lncRNA XIST expression on survival in solid tumors. The following search terms included: “Long non coding RNA XIST” OR “Long Noncoding RNA XIST” OR “long non-coding RNA XIST” OR “lncRNA XIST” OR “X-inactive specific transcript” (all fields) AND “Prognosis” OR “Prognoses” OR “Prognostic” OR “Outcome” OR “survival” (all fields) AND “Neoplasia” OR “Neoplasias” OR “Neoplasm” OR “Tumor” OR “Cancer” OR “tumour” OR “carcinoma” (all fields). Moreover, the literature has been tracked to determine more relevant studies.
Selection criteria
All collected studies were included in this meta-analysis according to the criteria as follows: (1) lncRNAncRNA XIST expression was detected only in solid tumors, not including hematologic malignancies; (2) investigation of the association between lncRNA XIST expression and survival outcome were represented in overall survival; (3) reporting sufficient data to estimate the hazard ratio (HR) and 95% confidence interval (CI) according to lncRNA XIST expression; (4) lncRNA XIST expression was detected by quantitative reverse transcription PCR (qRT-PCR) in OS tissues; (5) not a review, meta-analysis, case reports, duplicate publications.
Data extraction and quality assessment
Data extraction of literature was as follows: first author, publication year, country of origin, cancer type, sample size, number of patients in high and low lncRNA XIST expression groups, the detection method, and the cut-off, survival analysis, the HR and 95% CI. If HR was provided in the study, we extracted them directly. Otherwise, survival data were extracted from the original study data (Kaplan–Meier curves or required data) using the software Engauge Digitizer 4.1 and calculated by Tierney. The quality of included studies was evaluated by two investigators independently according to the Newcastle–Ottawa Quality Assessment Scale (NOS). Furthermore, two investigators could resolve their differences by consensus or in discussions with a third investigator. The lowest and highest scores were scored at 0 and 9, respectively, and a study with a score greater than 6 or higher was considered a high-quality study.
Statistical analysis
HR with 95% CI was estimated to evaluate the effective value of lncRNA XIST expression on prognosis in solid tumors. The high expression and low expression of lncRNA XIST was defined according to the cut-off values provided in the article. The heterogeneity of pooled results was evaluated using Cochran’s Q test and Higgins I-squared statistic. A statistically significant heterogeneity was defined as p < 0.10 or I2 > 50%, where a random-effect was applied. Otherwise a fixed-effect model was used. Subgroup analysis was used to further explore possible sources of heterogeneity. The stability of the results was assessed using a sensitivity analysis. The possibility of publication bias was also assessed using Begg’s test. All data were analyzed using STATA software version 12.0 (Stata Corporation, College Station, TX, USA), and a p value less than 0.05 was considered as statistically significant.
Discussion
LncRNA XIST was a product of the XIST gene located in the X inactivation center [
14], which was the first regulatory RNAs discovered to be involved in the formation of the inactive X chromosome [
19]. When an X chromosome was inactivated in female animal, lncRNA XIST diffuses throughout the X chromosome, ultimately resulting in inactivation of the X chromosome [
20]. Moreover, lncRNA XIST could play a dosage compensation role in female animal cells. In other words, the phenotypes determined by the gene on the X chromosome were equally expressed in XY males and XX females [
15].
Aberrant expression of lncRNA XIST has been detected in many diseases. It played an important role in proliferation, migration and invasion in cancer cells in vitro and in vivo, which indicated that XIST exerted an essential role on the occurrence and development of various tumors. Differentially expressed lncRNAs could act as oncogenes or tumor suppressors to improve cancer diagnosis, discover potential treatment targets, and improve prognosis. Although many studies found that the high expression of lncRNA XIST was closely related to the prognosis of a variety of tumors, the results of the studies were quite different. It was reported that the high expression of lncRNA XIST was a risk factor for the prognosis of cancers, while some reports indicated that the high expression of lncRNA XIST was a beneficial factor in the prognosis of cancers.
For instance, Chen et al. and Wu et al. demonstrated that knockdown of lncRNA XIST suppressed cells proliferation, migration and invasion in vitro as well as tumorigenesis and metastasis in vivo in GC (2016) and ESCC (2017), respectively. Moreover, they all found that an inverse relationship between lncRNA XIST and miR-101, and knockdown of lncRNA XIST exerted its tumor-suppressive effects at least in part through regulating miR-101 to modulate EZH2 expression [
14,
21]. Meanwhile, a study from Ma et al. showed that lncRNA XIST promoted cell cycle progression from the G1 phase to the S phase and protected cells from apoptosis, which contributed to GC cell growth. XIST was responsible for GC cell proliferation and invasion through the miR-497/MACC1 axis [
13]. Furthermore, Temozolomide (TMZ) was the most commonly used alkylating agent in glioma chemotherapy. The data from Du et al. revealed that XIST knockdown could sensitize TMZ-resistant glioma cells to TMZ. XIST inhibited miR-29c expression by direct targeting in TMZ-resistant glioma cells [
11]. In summary, it implicated that overexpression of lncRNA XIST was associated with adverse prognosis and could be used as an independent prognostic factor.
Contrary to the above tumors, increasing evidence demonstrated that XIST could also act as tumor suppressors, and played important roles in the initiation and progression of multiple cancers [
12,
15,
22,
23]. For example, Kobayashi et al. observed in 2016 that the 4-year overall survival rates of patients with CSCC were 87.1 and 54.4% in the high and low XIST expression groups, respectively [
22]. The results suggest that XIST could be a potential biomarker or therapeutic target for OS. However, the effect aberrant XIST expression on the prognosis of patients was still controversial in HCC and osteosarcoma. Recently, a study from Ma et al. showed that patients with JPX/XIST overexpression in HCC had longer survival times than those with low expression [
15], contrary to previous research from Kong et al. [
16]. Furthermore, a study from Zhang et al. revealed that lncRNA XIST regulated PDCD4 expression by interacting with miR-21-5p and inhibits osteosarcoma cell growth and metastasis [
23]. While a study from Li et al. suggested that lncRNA XIST had a tumor promoter effect, and thus, to be a predictor of outcome in patients with osteosarcoma [
24].
To get more accurate evidence to prove the relationship between the high expression of lncRNA XIST and the prognosis of cancers, relevant studies have been comprehensively retrieved and analyzed. Furthermore, the regulatory mechanism involved in lncRNA XIST was complex. And there was a lack of systematic research for effect of lncRNA XIST expression on tumor prognosis. Therefore, we conducted a meta-analysis to evaluate the potential value of lncRNA XIST as a novel biomarker for predicting tumor prognosis, which provided a reference for the follow-up study.
In this study, high expression of lncRNA XIST in cancer tissue was associated with poor prognosis in cancer patients (HR = 1.54, 95% CI 1.07–2.23, p = 0.021), with heterogeneity in the data (I2 > 50%). Numerous studies have shown that lncRNAs were involved in the regulation of protein-coding genes at the transcriptional and post-transcriptional levels, and it also influenced the signaling pathway pathways both intracellular as well as in organism development, thus affecting cell growth, apoptosis, and metastasis. Based on the above, deregulations of lncRNA could be a major cause of disease in human-complex diseases, including tumors. It indicated that it might serve as a negative prognostic marker for solid tumors.
Subgroup analysis and sensitivity analysis were used to investigate whether the heterogeneity of the data affected the interpretation of the analysis results. The association between lncRNA XIST overexpression and worse OS was statistically significant in digestive system tumors (HR = 1.67, 95% CI 1.11–2.51, p = 0.014, random-effect). These results indicated that the adverse prognostic effect of high lncRNA XIST remained substantial in digestive system cancers. In the meantime, lncRNA XIST overexpression was associated with a poor prognosis, which was statistically significant when the patients’ number is greater than 65 [patients’ number > 65 (HR = 1.75, 95% CI 1.24–2.47, p = 0.001, random-effect)], multivariate analysis (HR = 2.39, 95% CI 1.28–4.46, p = 0.006, random-effect), and reported in text (HR = 2.50, 95% CI 1.49–4.18, p = 0.000, random-effect).
However, there were several limitations in this paper. First of all, the different thresholds of lncRNA XIST expression were different in different studies that could not reach a uniform standard. Second, HR and 95% CI in some studies could not be obtained directly from the original literature, but HR estimates were derived from their survival curves, which might affect the results of this study. Furthermore, the limited number of included studies, all from Asians, and the small sample size (1290 cases in total) might diminish the reliability of the results. In the future, the studies of high-quality samples needed to be further confirmed.