Archives

  • 2022-08
  • 2022-07
  • 2022-06
  • 2022-05
  • 2022-04
  • 2021-03
  • 2020-08
  • 2020-07
  • 2020-03
  • 2019-11
  • 2019-10
  • 2019-09
  • 2019-08
  • 2019-07
  • br The nomogram is generated

    2019-10-07


    The nomogram is generated using the R software's “rms” package. The nomogram concordance index (C-index) was generated from a fitted multivariate Cox regression analysis of all patients. The larger C-index, the more accurate the predictions are. The nomogram was plotted in combination with independent prognostic factors in multi-variate analysis. And the total score for each patient can be calculated utilizing the generated nomogram. Then, the total score can be used to predict the rate of 3- or 5-year survival (Tian et al., 2017).
    2.5. Functional enrichment analysis
    Utilizing the miRNA target prediction tool TarBase v.8, the target genes of miRNAs were predicted. Gene Ontology (GO) and KEGG pathway enrichment analysis were performed in DIANA TOOLS
    (http://diana.imis.athena-innovation.gr/DianaTools/index.php) (Vlachos and Hatzigeorgiou, 2017).
    3.1. Differentially expressed miRNAs
    A total of 561 differentially expressed miRNAs were included in our study according to the selection criteria. Among these differentially expressed miRNA, 405 miRNAs were up-regulated, and 156 miRNAs were down-expressed. As shown in Heatmap [Fig. S1], red represents higher miRNA MPP+ Iodide levels and green represents lower expression levels. The volcano plot shows the distribution of the differentially expressed miRNA in Fig. S2.
    3.2. The predictive 4-miRNA signature in the training set
    Through the univariate Cox regression analysis in the training set,
    3.3. The evaluation of the 4-miRNA signature in the training set
    Calculated by the above risk score formula, all 265 patients in the training set will obtain a risk score, and the patients will be divided into 132 high-risk and 133 low-risk groups according to the cutoff value. The K-M survival curve results showed that EC patients with high-risk scores had significantly worse OS than EC patients with lower risk scores (HR = 4.35, 95% CI 1.99–9.54) [Fig. 1A]. The 5-year survival rates for the low-risk group and the high-risk group were 90.5% and 68.9%, respectively. The time-dependent ROC curves showed sensi-tivity and specificity of 5-year OS prediction by the 4-miRNA prognostic model, and it yielded AUC was 0.727 [Fig. 1B], indicating that the prognostic model can generally predict the prognosis of EC patients in training set.
    3.4. The validation of the 4-miRNA signature in the testing and entire set
    To verify the robustness of the survival prediction by 4-miRNA biomarker in EC patients, the predictive power of 4-miRNA biomarkers was tested in the testing set (n = 265) and the entire set (n = 530). Applying the same prognostic model and cutoff value from the training set, the patients were divided into 128 high risk and 136 low-risk groups in the testing set. In the testing set, the OS of EC patients with high-risk scores was still significantly lower than patients with risk scores (HR = 3.46, 95% CI 1.87–6.39) [Fig. 2A]. The 5-year survival
    Table 2
    Information of the four miRNAs associated with the OS.
    Similarly, Spindle prognostic model also showed consistent results in the entire set. For all EC patients with survival data in TCGA, the Kaplan-Meier curves for the high-risk and low-risk score are shown in Fig. 2C. Patients expressing the high risk-score exhibited poorer OS than patients expressing the low risk-score (HR = 3.78, 95% CI 2.33–6.12). The AUC for the ROC curves in the entire set was 0.704 [Fig. 2D]. The results of revalidation show that the 4-miRNA prognostic model is robust and accurate in EC patients.
    3.5. The independence of 4-miRNA signature for survival prediction
    The prognosis efficiency of 4-miRNA signature was compared with clinicopathological variables for the ROC curves of 5-year OS. The re-sults showed that the 4-miRNA signature had an AUC of 0.7, while the histological types, age, stage, and grade AUC were 0.623, 0.578, 0.693, and 0.655, respectively. It means that 4-miRNA signature has a good prognostic potential compared with traditional clinical variables. In addition, subgroup analyses were performed and analyzed for the ac-curacy of the 4-miRNA prognosis model in subgroup patients with different clinical characteristics. This prognosis model was validated in patients of different ages (≥65 vs < 65), and the results showed that 4-miRNA biomarkers were effective in young (HR = 5.96, 95% CI 2.71–13.11) [Fig. 3A] and elderly patients (HR = 2.33, 95% CI