An integrated signature of clinical metrics and immune-related genes as a prognostic indicator for ST-segment elevation myocardial infarction patient survival

Abstract

Background: The immune-inflammatory pathway plays a critical role in myocardial infarction development. However, few studies have systematically explored immune-related genes in relation to myocardial infarction prognosis using bioinformatic analysis. Our study aims to identify differentially expressed immune-related genes(DEIRGs) in ST-segment elevation myocardial infarction (STEMI) patients and investigate their association with clinical outcomes.

Materials and methods: We conducted a systematic review of Gene Expression Omnibus datasets, selecting GSE49925, GSE60993, and GSE61144 for analysis. DEIRGs were identified using GEO2R and overlapped across the chosen datasets. Functional enrichment analysis elucidated the DEIRGs' biological functions and pathways. We established an optimal prognostic prediction model using LASSO penalized Cox proportional hazards regression. The signature's clinical utility was evaluated through survival analysis, ROC curve assessment, and decision curve analysis. Additionally, we constructed a prognostic nomogram for survival rate prediction. External validation was performed using our own plasma samples.

Results: The resulting prognostic signature integrated two dysregulated DEIRGs (S100A12 and IL2RB) and two clinical variables (serum creatinine level and Gensini score). This signature effectively stratified patients into low- and high-risk groups. Survival analysis, ROC curve analysis, and decision curve analysis demonstrated its robust predictive performance and clinical utility within the first two years post-disease onset. External validation confirmed significant outcome differences between risk groups.

Conclusions: Our study establishes a prognostic signature that combines DEIRGs and clinical variables for STEMI patients. The signature exhibits promising predictive capabilities for patient stratification and survival risk assessment.

Keywords: Acute coronary syndrome; Bioinformatics; IL2RB protein; Prognosis; S100A12 protein; Transcriptome.

Fig. 1

Identification of DEIRGs from GSE61144 and GSE60993 microarray. (A) Venn diagram of 17 DEIRGs from the two microarray datasets. (B) The volcano plots of DEIRGs in the two datasets. Red indicates genes with high levels of expression, blue indicates genes with low levels of expression based on the criteria of P < 0.05 and |log(Fold Change)| > 1.0, respectively. (C) Ranking of the transcriptional difference among the DEIRGs. DEGs differential expression genes, DEIRGs differential expression of immune-related genes.

Fig. 2

GO enrichment and KEGG pathway analysis of DEIRGs. (A) Group display of biological processes, cellular components, and molecular functions of DEIRGs. (B) Network of the association between DEIRGs and GO terms and KEGG pathways. GO Gene Ontology, KEGG: Kyoto Encyclopedia of Genes and Genomes, NAD + nicotinamide adenine dinucleotide, ADP adenosine diphosphate, BP for biological process, CC for cellular component and MF for molecular function.

Fig. 3

(A) Expression correlation heatmap of DEIRGs between different prognostic groups of GSE49925. Cor correlation coefficient. (B) Co-expression heat maps between DEIRGs based on an individual level. The correlation was significantly weakened in the death group (left) compared to the survival group (right).

Fig. 4

GSE49925 validated the survival prediction of this signature. (A) Comparison of 17 DEIRGs in patients with different prognosis of GSE49925 (Wilcoxon rank sum test, *P < 0.05). (B) Comparison of risk scores between different prognostic groups. (C) A risk distribution plot intuitively shows the significant difference of survival outcomes between the two groups with different risk stratification.

Fig. 5

The signature predicts survival of STEMI patients in GSE49925. (A) Time-dependent ROC curves. TPR true positive rate, FPR false positive rate. (B) Kaplan-Meier survival curves. HR hazard ratio. (C&D) Decision curve analysis of the constructed model at one and two years of onset.

Fig. 6

Construction and performance verification of nomogram. (A) Nomogram for predicting 1-,2-, and 3-year survival in the entire cohort. (B) Calibration curves of nomogram on consistency between predicted and observed 1-, 2- and 3-year survival in the entire cohort. The grey line at 45° implicated a perfect prediction, and the actual performances of our nomogram were shown in colored lines.

Fig. 7

The results of external validation. (A) A risk distribution plot and (B) Kaplan-Meier survival curves analysis showed significant differences in the incidence of MACEs among different risk groups.

figs1

figs2