MiR-24-3p and various cancers: From a meta-analysis view


 Background: A growing number of researches suggests that microRNAs (miRNAs) as oncogene or tumor suppressor genes play a fundamental role in various kinds of cancers. Among them, miR-24-3p, as a star molecule, is widely studied. However, the prognostic value of miR-24-3p is unclear and controversial. We conducted this meta-analysis to evaluate the prognostic value of miR-24-3p in a variety of cancers by integrated existing articles from four databasesMethods: PubMed, Embase, Web of Science, and Cochrane Library (last update in March 2020) were searched for approach literature. Hazard ratios (HRs) and odds ratios (ORs) were used to evaluate the association between miR-24-3p expression levels and prognostic value or clinicopathological characteristics, respectively.Results: A total of 15 studies from 14 literature were finally qualified and concluded in the present meta-analysis. A significantly worse overall survival was observed in higher expression of miR-24-3p cancer group for OS(Overall survival) of log rank tests and cox multivariate regression by fixed effects model. Also, we found a significant correlation between elevated miR-24-3p levels to RFS (Recurrence-free survival) and DFS(Disease free survival). In addition, the pooled odds ratios (ORs) showed that evaluated miR-24-3p was also associated with the larger tumor size (≥5cm) and advanced TNM stage (Ⅲ and Ⅳ).Conclusion: Built on the above findings, elevated expression levels of miR-24-3p may serve as a promising biomarker used to predict the worse prognosis of cancer patients.

tests. P < 0.05 and / or I 2 > 50% were de ned as signi cant heterogeneity and random effects model was further to used. In addition, sensitivity analysis was used to evaluate the contribution of each study to the pooled HR and we could further to estimate the stability of the consequence. Finally, we evaluate the potential publication bias by funnel plot, Begg's test and Egger's test. P < 0.05 was known as obvious publication bias [32].
3.1. The association between miR-24-3p expression levels and the overall survival (OS) Ten enrolled articles including eleven studies and 1212 patients were used to investigate the correlation between miR-24-3p expression levels and the OS by using log rank tests and presented the data of univariate. Generally, a signi cant correlation between miR-24-3p levels and OS (HR = 1.609, CI: 1.291-2.004, Figure. 2a). However, an obvious heterogeneity was also observed (I 2 = 85.20%, P = 0.000, Table 2). Hence, the random effects model was followed in succession but the signi cance was disappeared (HR = 1.507, CI: 0.810-2.803, Table 2), indicating that the heterogeneity signi cantly in uenced the consequence.
In order to explore the source of the heterogeneity, subgroup analyses were applied by factors including population (Asian(Chinese) and Non-Asian), sample size (≥ 100 and < 100), NOS scores (≥ 8 and < 8), specimen (tissue and non-tissue) tumor category 1 (solid tumor and non-solid tumor), tumor category 2 (digestive system and non-digestive system) and tumor (esophageal cancer, osteosarcoma, lung cancer, gastric cancer, colorectal cancer, ALL, AML and hepatocellular carcinoma). As a consequence, the heterogeneity was controlled successfully in six subgroups and all them have signi cant correlations: 1).
The patients of lung cancer (HR = 3.274, CI: 1.422-7.539, I 2 = 0.000%, P = 0.698). In addition, signi cant correlations are also observed in the study of NOS score less than 8 by random effects model, which were consistent with the signi cance of the results by xed effects model (Table 2). Moreover, signi cant correlations were observed between miR-24-3p expression levels and OS in the studies with the population derived from Asian(Chinese) (HR = 1.381, CI: 1.219-2.004), solid tumor (HR = 1.448, CI: 1.131-1.852), digestive system (HR = 1.705, CI: 1.291-2.253) and non-digestive system (HR = 1.461, CI: 1.021-2.090) by xed effects model, while there were no signi cances identi ed in these groups when the random effects model was applied ( Gao Y et al. [29] HR = 0.285, CI: 0.139-0.584). Due to insu cient data, the consequence was lack of e ciency and the heterogeneity was also signi cant (I 2 = 93.00%, P = 0.000). Therefore, more relevant studies are required to perform the analysis. Built on the above consequences, meta regression was further used, but there was no meaningful contribution identi ed to impact on the heterogeneity (Table 2). Subsequently, the sensitivity analysis was performed, but there was also no positive consequence ( Figure. 2c). Next, funnel plots, Begg's test and Egger's test were implemented to assess the potential publication bias and two studies as the outliers were identi ed eventually ( Figure. 2d) (Liu et al. [41] and Gao et al. [29] ). After removing them, dramatically decline of the heterogeneity was observed (I 2 = 34.60%, P = 0.141) in the overall analysis, and the signi cance of the prognostic effects of miR-24-3p expression was still obvious ( Figure.

The independent role of miR-24-3p expression levels as a prognostic indicator
Five studies containing 775 patients implemented the cox multivariate regression to assess the prognostic value of miR-24-3p expression levels in carcinoma patients by adjusting other factors. The signi cant correlation of miR-24-3p expression levels to the OS (HR = 2.384, CI: 1.813-3.134) was observed by xed effects model. However, the heterogeneity was relatively obvious (I 2 = 82.30%, P = 0.000, Table 3) and the signi cance was vanished by random effects model (HR = 1.994, CI: 0.991-4.015). Homoplastically, Subgroup analyses were applied to reduce the heterogeneity. As a result, the homogeneity was reached within the studies of sample size greater than or equal to 100 (I 2 = 0.000%, P = 0.861), NOS less than 8 (I 2 = 45.50%, P = 0.176) and the patients of hepatocellular carcinoma (I 2 = 45.50%, P = 0.176). And the signi cant association was identi ed between miR-24-3p expression levels and OS with the sample size greater than 100 (HR = 3.369, CI: 2.414-4.701), NOS less than 8 (HR = 3.041, CI: 2.150-4.300) and the patients of hepatocellular carcinoma (HR = 3.041, CI: 2.150-4.300). In addition, the signi cant correlations were identi ed between miR-24-3p expression levels to the OS in the population from Asian(Chinese) (HR = 2.373, CI: 1.813-3.134), the specimen derived from tissue (HR = 2.448, CI: 1.804-3.323) and NOS larger than or equal 8 by xed effects model, which become to no signi cance within those subgroups by random effects model (Table 3) (Table 3). But the sensitivity analysis suggested that Gao et al. [29] has signi cant impact on the result (Figure. 4c). The heterogeneity was vanishing (I 2 = 0.000%, P = 0.591, Figure. 4b)by removing this outlier and the correlation of miR-24-3p expression levels to the OS was also signi cant (HR = 3.039, CI: 2.268-4.074, Figure. 4b). Finally, funnel plots, Begg's test (P = 0.734) and Egger's test (P = 0.460) indicated that there was no bias. But, the number of enrolled studies was few, more data are needed to reinforce this result.

The correlation of miR-24-3p expression levels to the RFS /DFS
Except OS as a prognostic indicator, RFS and DFS are also be accepted as an evaluation criterion. Here, four studies reported RFS including 393 patients applied log rank tests, while only one also utilized cox multivariate regression. After pooling the HR, we observed a signi cant association between miR-24-3p expression levels to the RFS of log rank tests (HR = 2.315, CI: 1.491-3.594, gure. 5a) by xed effects model. However, the heterogeneities were quite obvious (I 2 = 66.70%, P = 0.290, Table 4). The random effects model was further implemented but the signi cance was disappeared (HR = 1.814, CI: 0.741-4.440), indicating that the heterogeneity in uenced the consequences signi cantly. Furthermore, owing to limited number of statistics from cox multivariate regression, the sensitivity analysis and publication bias were only applied to analysis with data extracted from log rank tests. The sensitivity analysis result indicated that no studies had signi cant in uence on the consequent (Figure. 5c). However, the investigation of potential publication bias identi ed an outlier ( Figure. 5d, Wang S et al. [40]). After deleting this study, the heterogeneity was obvious declined (I 2 = 45. 30%, P = 0.161) and the signi cance of correlation between miR-24-3p expression levels and the RFS was not altered (HR = 2.575, CI: 1.642-4.029, Figure. 5b). Due to the limit included studies, more data are needed in order to enhance the result. In addition, there were only two studies containing 226 patients revealed the DFS statistics and almost no heterogeneity in both log rank tests and cox multivariate regression (I 2 = 3.600%, P = 0.309, I 2 = 0.000%, P = 0.330, respectively, Table 4) by used a xed effects model. We also observed signi cant strong correlation between miR-24-3p expression levels to the DFS of both log rank tests (HR = 2.361, CI: 1.390-4.012) and cox regression tests (HR = 2.313, CI: 1.315-4.067) by xed effects model.

Correlations between miR-24-3p levels and clinicopathological features among various carcinomas
Six studies containing 536 patients investigated the correlation of miR-24-3p expression levels to different clinical characteristics. As showed in Table 5 There were no heterogeneity in the analysis of age (I 2 = 0.000%, P = 0.525) and gender (I 2 = 0.000%, P = 0.842), but the heterogeneity of lymph node metastasis and TNM stage were obviously (I 2 = 70.90%, P = 0.064; I 2 = 85.50%, P = 0.000, respectively). In order to decrease the heterogeneity, sensitivity analysis and publication bias were further investigated to each of them. As a result, an outlier was found (Liu et al. [36]) in the TNM stage. After removing the outlier, the heterogeneity was dramatically decreased from 85.50-0.000% and the associations between high miR-24-3p expression levels to advanced TNM stage were signi cant (OR = 2.328, CI: 1.490-3.637). (Figure.

Discussion
It is of great importance to explore prognostic biomarkers with the patients of carcinoma as speci c biomarkers can further help to directly stratify patients and effectively guide clinical decision-making. MiR-24-3p, as an oncogene or tumor suppressor, plays a key role in the occurrence, progression and metastasis of human carcinoma was realized by more and more researchers gradually [24,25]. Quan et al. [45] had made a meta-analysis to research the correlation between miR-23a/24 − 2/27a cluster with human cancers, but they only had limited data to draw a conclusion that high expression levels of miR-23a/24 − 2/27a indicated a worse prognosis and no further analyzing the correlation between miR-24-3p expression levels to the clinicopathological characteristics.
Subsequently, more and more studies which focusing on the miR-24-3p expression levels with cancer progression, metastasis and prognosis of patients were carried on [4][5][6][7][14][15][16]. Thus, the exact role of miR-24-3p on the clinical prognosis of patients in various human carcinomas still need to further investigate. In this meta-analysis, total 15 studies including 1518 people were recruited. Among them, ten studies containing 1212 patients provided the statistics of the OS by log rank tests. By the pooling strategy, we know that the elevated miR-24-3p expression levels were linked to worse prognosis of cancer patients. Subsequently, several approaches were put in place to investigate the heterogeneity. First, subgroup analyses were conducted to identify the potential sources of heterogeneity. We found that the heterogeneity was achieved within the non-Asian population, non-tissue, the studies of sample size greater than or equal 100, hematologic tumor, hepatocellular carcinoma and lung cancer. But it was not controlled in other groups, such as the Asian population, studies of sample sizes less than to 100, solid tumor and so on. Second, the sensitivity analysis was used, but no outlier was identi ed to impact on the pooled results signi cantly. Third, two studies were identi ed as outliers by publication bias evaluation (Liu et al. [41] and Gao et al. [29]). After retrieving the data of outliers, we found that the specimen recruited in them were all from Asian (Chinese), tissue and the sample size less than 100 which all have strong heterogeneity. Besides, Liu et al. [41] was the sole study that focused on the Osteosarcoma and Gao et al. [29] had an opposite conclusion with Kerimis et al. [38] who also investigated the miR-24-3p expression levels to colorectal cancer. After removing those two outliers, greatly declined of the heterogeneity was observed. Built on the mentioned above, these two studies could be the major sources of heterogeneity. However more relevant data are needed to further investigate because of the limit number of studies. There were ve studies including 775 patients obtained the data of HRs by cox multivariate regression. Cox multivariate regression has been known as an effective approach because it can evaluate the contribution of each factor independently by adjusting other factors [46]. Thus, the consequences by cox multivariate regression are always considered as independent effects of each factor on the clinical outcome. As a result, we found that the signi cance was inconsistent among different effects model. This phenomenon suggested that the heterogeneity was relatively obvious and the consequences were instable. Through the subgroup analyses, we found that the heterogeneity was declined in hepatocellular carcinoma and achieved in the studies of sample size larger than or equal 100. In addition, the sensitivity analysis identi ed one outlier, Gao et al. [29] who has an opposite conclusion with Kerimis et al [38]. After removing this study, the heterogeneity had been signi cantly vanished. High miR-24-3p expression had a signi cantly worse survival and there was no publication bias. Thus, the power of miR-24-3p expression levels might serve as an independent prognostic indicator and we need more data to reinforce this conclusion. Also, we detected additional indexes such as RFS and DFS. MiR-24-3p expression levels were deemed to be signi cantly associated with DFS of statistics extracted from both log rank tests and cox regression analysis. For the RFS of cancer patients, only the xed effects model revealed a signi cant correlation between miR-24-3p expression with this prognostic index and the heterogeneity was palpable. We identi ed an outlier (Wang et al. [40]) through publication bias evaluation. After removing this study, the heterogeneity was declined and the signi cance of association between miR-24-3p expression levels and the RFS was not altered.
As for the clinicopathological parameters, six studies including 536 patients had evaluated the association of miR-24-3p expression levels to the distinctive clinical parameters. The over-expression of miR-24-3p was found to be signi cantly related to larger tumor size by xed effects model. Moreover, we found a signi cant heterogeneity between miR-24-3p expression levels to TNM stage. Appling sensitivity analyses, we identi ed one study (Liu et al. [36]) which had greatly impact on the result for the TNM stage. After removing this study, the heterogeneity completely disappeared, the association between miR-24-3p expression levels to the TNM stage was also signi cant. In addition, there were only two studies about the lymph node metastasis are enrolled and the conclusion might be not reliable. The analyzes of clinical features of a de nite carcinoma should be normalized for the cut-off values, the feature categories and so on, to enrich the enrolled cases and characteristics for the meta-analysis.
As far as we know, this meta-analysis was the most comprehensive and systematic one to explore the correlation between the miR-24-3p expression levels with the prognosis of cancer patients in depth. Subgroup analysis, meta regression, sensitivity analysis and publication bias had been used to investigate the possible source of the heterogeneity to the greatest extent [47]. In spite of this, several aws were hard to avoid in this meta-analysis. First, inevitable limitation from insu cient data in this analysis (only 15 studies with 1518 patients). Second, the cut-off values of the miR-24-3p expression levels were not exactly among those studies, thus, the accuracy of prognostic results may be in uenced. Third, part of HRs was calculated from the survival curves which may cause some bias. Four, the number of recruited studies for DFS, RFS and clinicopathological features analyses were relatively insu cient. Taking above reasons into account, we need better designed and large sample size studies for further research before applying miR-24-3p as a prognostic biomarker of tumor in clinical applications.

Conclusions
The over expression of miR-24-3p was an underlying risk of poor prognosis in various human carcinomas, especially in hepatocellular carcinoma and lung cancer. As for other types of carcinomas, the results are not yet stable and more studies including normalized research conditions are required to further identify miR-24-3p prognostic values. In addition, high miR-24-3p expression levels were linked to the progression of cancers, developing more malignant behavior, such as larger tumor sizes and the advanced TNM stages.miR-24-3p expression levels could serve as a potential prognostic marker of human carcinoma. The authors declare that all data supporting the ndings of this study are available within the article and the enrolled articles for meta-analysis.

Abbreviations
The datasets generated and/or analysed during the current study are available in the PubMed, Embase, Web of science (WOS) and Cochrane library repository. The ow chart of the meta-analysis The independent role of miR-24-3p as a prognostic indicator for (A) overall survival, (B) overall survival without outliers, and (C) sensitivity analysis, (D) publication bias evaluation