Systematic Construction and Validation of a Metabolic Risk Model for Prognostic Prediction in Acute Myelogenous Leukemia

Establishment and validation of a gene mutation-based risk model for predicting prognosis and therapy response in acute myeloid leukemia

Background

Acute myeloid leukemia (AML) is a malignant clonal proliferative disease of hematopoietic system. Despite tremendous progress in uncovering the AML genome, only a small number of mutations have been incorporated into risk stratification and used as therapeutic targets. In this research, we performed to construct a predictive prognosis risk model for AML patients according to gene mutations.

Methods

Next-generation sequencing (NGS) technology was utilized to detect gene mutation from 118 patients. mRNA expression profiles and related clinical information were mined from TCGA and GEO databases. Consensus cluster analysis was applied to obtain molecular subtypes, and differences in clinicopathological features, prognosis, and immune microenvironment of different clusters were systematically compared. According to the differentially expressed genes (DEGs) between clusters, univariate and LASSO regression analysis were applied to identify gene signatures to build a prognostic risk model. Patients were classified into high-risk (HR) and low-risk (LR) groups according to the median risk score (RS). Differences in prognosis, immune profile, and therapeutic sensitivity between two groups were analyzed. The independent predictive value of RS was assessed and a nomogram was developed.

Results

NGS detected 24 mutated genes, with higher mutation frequencies in CBL (63 %) and SETBP1 (49 %). Two clusters exhibited different immune microenvironments and survival probability (p = 0.0056) were identified. A total of 444 DEGs were screened in two clusters, and a mutation-associated risk model was constructed, including MPO, HGF, SH2B3, SETBP1, HLA-DRB1, LGALS1, and KDM5B. Patients in LR had a superior survival time compared to HR. Predictive performance of this model was confirmed and the developed nomogram further improved the applicability of the risk model with the AUCs for predicting 1-, 3-, 5-year survival rate were 0.829, 0.81 and 0.811, respectively. HR cases were more sensitive to erlotinib, CI-1040, and AZD6244.

Conclusion

These findings supplemented the understanding of gene mutations in AML, and constructed models had good application prospect to provide effective information for predicting prognosis and treatment response of AML.

Identification of an optimized glycolytic-related risk signature for predicting the prognosis in breast cancer using integrated bioinformatic analysis

Aberrant metabolic disorders and significant glycolytic alterations in tumor tissues and cells are hallmarks of breast cancer (BC) progression. This study aims to elucidate the key biomarkers and pathways mediating abnormal glycolysis in breast cancer using bioinformatics analysis. Differential genes expression analysis, gene ontology analysis, Kyoto encyclopedia of genes and genomes analysis, gene set enrichment analyses, and correlation analysis were performed to explore the expression and prognostic implications of glycolysis-related genes. We effectively integrated 4 genes to construct a prognostic model of shorter survival in the high-risk versus low-risk group. The prognostic model showed promising predictive value and may be an integral part of the prognosis of BC. The survival analysis and receiver operating characteristic curves suggested that the signature showed a good predictive performance in both the The Cancer Genome Atlas training set and 2 gene expression omnibus validation sets. Multivariable analysis demonstrated that the 4-gene signature had an independent prognostic value. Furthermore, all calibration curves exhibited robust validity in prognostic prediction. We established an optimized 4-gene signature to clarify the connection between glycolysis and BC, and offered an attractive platform for risk stratification and prognosis predication of BC patients.

Lipid profile as a novel prognostic predictor for patients with acute myeloid leukemia

Purpose

This study investigated the relationship between serum lipid levels and clinical outcomes in acute myeloid leukemia (AML) by establishing a predictive risk classification model.

Method

A total of 214 AML patients who were pathologically diagnosed and treated with standard induction chemotherapy at Sun Yat-Sen University Cancer Center were included. The patients were randomly divided into the training (n = 107) and validation (n=107) cohorts. Univariate and multivariate Cox analyses were used to assess the value of triglyceride (TG), Apolipoprotein B (Apo B), Apo Apolipoprotein A-I (Apo A-I), cholesterol (CHO), and high-density lipoprotein (HDL) as prognostic factors for AML.

Results

After a series of data analyses, a five-factor model was established to divide the patients into high- and low-risk groups. Kaplan-Meier survival analysis showed that the high-risk group had a poor prognosis (P<0.05). The area under the curve of the novel model for five-year OS was 0.737. A nomogram was constructed to integrate the model with age and the 2017 ELN cytogenetic classification, with the merged model showing improved accuracy with an area under the curve of 0.987 for five-year OS.

Conclusion

A novel model was constructed using a combination of the serum lipid profile and clinical characteristics of AML patients to enhance the predictive accuracy of clinical outcomes. The nomogram used the lipid profile which is routinely tested in clinical blood biochemistry and showed both specific prognostic and therapeutic potential.

DNA damage to bone marrow stromal cells by antileukemia drugs induces chemoresistance in acute myeloid leukemia via paracrine FGF10-FGFR2 signaling

Chemoresistance remains a major challenge in the current treatment of acute myeloid leukemia (AML). The bone marrow microenvironment (BMM) plays a complex role in protecting leukemia cells from chemotherapeutics, and the mechanisms involved are not fully understood. Antileukemia drugs kill AML cells directly but also damage the BMM. Here, we determined antileukemia drugs induce DNA damage in bone marrow stromal cells (BMSCs), resulting in resistance of AML cell lines to adriamycin and idarubicin killing. Damaged BMSCs induced an inflammatory microenvironment through NF-κB; suppressing NF-κB with small molecule inhibitor Bay11-7082 attenuated the prosurvival effects of BMSCs on AML cell lines. Furthermore, we used an ex vivo functional screen of 507 chemokines and cytokines to identify 44 proteins secreted from damaged BMSCs. Fibroblast growth factor-10 (FGF10) was most strongly associated with chemoresistance in AML cell lines. Additionally, expression of FGF10 and its receptors, FGFR1 and FGFR2, was increased in AML patients after chemotherapy. FGFR1 and FGFR2 were also widely expressed by AML cell lines. FGF10-induced FGFR2 activation in AML cell lines operates by increasing P38 MAPK, AKT, ERK1/2, and STAT3 phosphorylation. FGFR2 inhibition with small molecules or gene silencing of FGFR2 inhibited proliferation and reverses drug resistance of AML cells by inhibiting P38 MAPK, AKT, and ERK1/2 signaling pathways. Finally, release of FGF10 was mediated by β-catenin signaling in damaged BMSCs. Our data indicate FGF10-FGFR2 signaling acts as an effector of damaged BMSC-mediated chemoresistance in AML cells, and FGFR2 inhibition can reverse stromal protection and AML cell chemoresistance in the BMM.

Risk stratification of acute myeloid leukemia: Assessment using a novel prediction model based on ferroptosis-immune related genes

In acute myeloid leukemia (AML), the link between ferroptosis and the immune microenvironment has profound clinical significance. The objective of this study was to investigate the role of ferroptosis-immune related genes (FIRGs) in predicting the prognosis and therapeutic sensitivity in patients with AML. Using The Cancer Genome Atlas dataset, single sample gene set enrichment analysis was performed to calculate the ferroptosis score of AML samples. To search for FIRGs, differentially expressed genes between the high- and low-ferroptosis score groups were identified and then cross-screened with immune related genes. Univariate Cox and LASSO regression analyses were performed on the FIRGs to establish a prognostic risk score model with five signature FIRGs (BMP2, CCL3, EBI3, ELANE, and S100A6). The prognostic risk score model was then used to divide the patients into high- and low-risk groups. For external validation, two Gene Expression Omnibus cohorts were employed. Overall survival was poorer in the high-risk group than in the low-risk group. The novel risk score model was an independent prognostic factor for overall survival in patients with AML. Infiltrating immune cells were also linked to high-risk scores. Treatment targeting programmed cell death protein 1 may be more effective in high-risk patients. This FIRG-based prognostic risk model may aid in optimizing prognostic risk stratification and treatment of AML.

Development and validation of a prognostic scoring model to risk stratify childhood acute myeloid leukaemia

To create a personal prognostic model and modify the risk stratification of paediatric acute myeloid leukaemia, we downloaded the clinical data of 597 patients from the Therapeutically Applicable Research to Generate Effective Treatments (TARGET) database as a training set and included 189 patients from our centre as a validation set. In the training set, age at diagnosis, -7/del(7q) or -5/del(5q), core binding factor fusion genes, FMS-like tyrosine kinase 3-internal tandem duplication (FLT3-ITD)/nucleophosmin 1 (NPM1) status, Wilms tumour 1 (WT1) mutation, biallelic CCAAT enhancer binding protein alpha (CEBPA) mutation were strongly correlated with overall survival and included to construct the model. The prognostic model demonstrated excellent discriminative ability with the Harrell's concordance index of 0.68, 3- and 5-year area under the receiver operating characteristic curve of 0.71 and 0.72 respectively. The model was validated in the validation set and outperformed existing prognostic systems. Additionally, patients were stratified into three risk groups (low, intermediate and high risk) with significantly distinct prognosis, and the model successfully identified candidates for haematopoietic stem cell transplantation. The newly developed prognostic model showed robust ability and utility in survival prediction and risk stratification, which could be helpful in modifying treatment selection in clinical routine.

Deconvolution of Bulk Gene Expression Profiles with Single-Cell Transcriptomics to Develop a Cell Type Composition-Based Prognostic Model for Acute Myeloid Leukemia

Acute myeloid leukemia (AML) is one of the malignant hematologic cancers with rapid progress and poor prognosis. Most AML prognostic stratifications focused on genetic abnormalities. However, none of them was established based on the cell type compositions (CTCs) of peripheral blood or bone marrow aspirates from patients at diagnosis. Here we sought to develop a novel prognostic model for AML in adults based on the CTCs. First, we applied the CIBERSORT algorithm to estimate the CTCs for patients from two public datasets (GSE6891 and TCGA-LAML) using a custom gene expression signature reference constructed by an AML single-cell RNA sequencing dataset (GSE116256). Then, a CTC-based prognostic model was established using least absolute shrinkage and selection operator Cox regression, termed CTC score. The constructed prognostic model CTC score comprised 3 cell types, GMP-like, HSC-like, and T. Compared with the low-CTC-score group, the high-CTC-score group showed a 1.57-fold [95% confidence interval (CI), 1.23 to 2.00; p = 0.0002] and a 2.32-fold (95% CI, 1.53 to 3.51; p < 0.0001) higher overall mortality risk in the training set (GSE6891) and validation set (TCGA-LAML), respectively. When adjusting for age at diagnosis, cytogenetic risk, and karyotype, the CTC score remained statistically significant in both the training set [hazard ratio (HR) = 2.25; 95% CI, 1.20 to 4.24; p = 0.0119] and the validation set (HR = 7.97; 95% CI, 2.95 to 21.56; p < 0.0001]. We further compared the performance of the CTC score with two gene expression-based prognostic scores: the 17-gene leukemic stem cell score (LSC17 score) and the AML prognostic score (APS). It turned out that the CTC score achieved comparable performance at 1-, 2-, 3-, and 5-years timepoints and provided independent and additional prognostic information different from the LSC17 score and APS. In conclusion, the CTC score could serve as a powerful prognostic marker for AML and has great potential to assist clinicians to formulate individualized treatment plans.

Identification of a metabolic signature to predict overall survival for colorectal cancer

PURPOSE

Metabolic genes are associated with the occurrence and development of tumors. Metabolic-related risk models have showed partly prognostic predictive ability in cancers. However, the correlation between metabolic-related genes (MRGs) and the outcome of colorectal cancer is still poorly understood.

PATIENTS AND METHODS

TCGA database is used as the training cohort; while GSE39582 is the verification cohort. The least absolute shrinkage and selection operator Cox regression analysis were utilized to identify the MRGs and establish a genetic risk scoring model. A nomogram by integrating MRGs risk scores with TNM stage was constructed. The potential biological mechanisms were explored using gene set enrichment analysis. Associations of the signature with immune cell infiltrations and the tumor mutation burden (TMB) were also uncovered by Spearman rank test.

RESULTS

A six-gene metabolic signature was identified. Based on the risk scoring model with the signature, patients were divided into two groups (high-risk versus low-risk). The overall survival (OS) duration of patients with high-risk were quite shorter than those of low-risk patients (TCGA: p < .001, GSE39582: p < .001). Metabolic-related pathways were major enriched in low-risk group, while the high-risk group exhibited multiple immune-related pathways. Moreover, our signature was more linear dependent with antigen-presenting cell than effector immune cells, and a positive correction were seen between our signature and TMB.

CONCLUSION

Our research has discovered a six-gene metabolic signature to predict the OS of colorectal cancer. These genes may play significant roles in colorectal cancer regulating tumor microenvironment and serving as potential biomarkers for anti-cancer therapy.

Construction and validation of a metabolic risk model predicting prognosis of colon cancer

Metabolic genes have played a significant role in tumor development and prognosis. In this study, we constructed a metabolic risk model to predict the prognosis of colon cancer based on The Cancer Genome Atlas (TCGA) and validated the model by Gene Expression Omnibus (GEO). We extracted 753 metabolic genes and identified 139 differentially expressed genes (DEGs) from TCGA database. Then we conducted univariate cox regression analysis and Least Absolute Shrinkage and Selection Operator Cox regression analysis to identify prognosis-related genes and construct the metabolic risk model. An eleven-gene prognostic model was constructed after 1000 resamples. The gene signature has been proved to have an excellent ability to predict prognosis by Kaplan-Meier analysis, time-dependent receiver operating characteristic, risk score, univariate and multivariate cox regression analysis based on TCGA. Then we validated the model by Kaplan-Meier analysis and risk score based on GEO database. Finally, we performed a weighted gene co-expression network analysis and protein-protein interaction network on DEGs, and Kyoto Encyclopedia of Genes and Genomes pathways and Gene Ontology enrichment analyses were conducted. The results of functional analyses showed that most significantly enriched pathways focused on metabolism, especially glucose and lipid metabolism pathways.