TL;DR
This study develops a novel metabolic-related gene pair (MRGP) signature to predict prognosis in IDH-wildtype glioblastoma (IDH<sup>WT</sup> GBM), a particularly aggressive form of brain cancer lacking reliable biomarkers. The signature highlights the crucial role of ABCA1 expression in tumor-associated macrophages (TAMs) as a key prognostic indicator, offering a potential avenue for personalized treatment strategies.
What Problem Does This Paper Address?
Glioblastoma (GBM) is a devastating brain cancer, and its prognosis is heavily influenced by the mutation status of the isocitrate dehydrogenase (IDH) gene. While IDH-mutant gliomas have a better prognosis and are treated differently, the vast majority of GBMs (over 90%) are IDH-wildtype (IDH<sup>WT</sup>). These IDH<sup>WT</sup> GBMs are highly aggressive and currently lack robust molecular biomarkers to guide clinical management and predict treatment response, especially for patients undergoing chemoradiotherapy. Existing prognostic signatures often suffer from limitations like small discovery datasets, batch effects from merging diverse data, and lack of validation. This paper aims to address this critical unmet need by developing and validating a reliable prognostic signature for IDH<sup>WT</sup> GBM patients.
Methods
The researchers employed a multi-stage approach to develop and validate their prognostic signature:
- Meta-analysis and Data Integration: They integrated gene expression profiles from four public glioblastoma datasets (one TCGA and three CGGA). To overcome platform and batch effects, they utilized a novel "relative ranking" approach on metabolism-related genes (MRGs) within individual samples.
- MRGP Signature Development: They generated over 135,000 metabolism-related gene pairs (MRGPs) for each sample. Using LASSO Cox proportional hazards regression on a training dataset, they identified a signature composed of 21 MRGPs.
- Prognostic Model and Risk Score: A risk score was calculated for each patient based on the coefficients of the 21 MRGPs. An optimal cut-off value was determined using a 1-year time-dependent ROC curve to stratify patients into high- and low-risk groups.
- Validation: The signature's prognostic power was validated in independent datasets, including an external validation set and across different GBM subtypes and treatment regimens (chemotherapy and radiotherapy).
- Functional and Immune Analyses: Gene Ontology (GO) analysis and Gene Set Enrichment Analysis (GSEA) were performed to understand the biological processes associated with the MRGP signature. Furthermore, they investigated tumor purity and immune cell infiltration using ESTIMATE, xCell, and CIBERSORT.
- Biological Validation: The expression of key MRGs, particularly ABCA1 in tumor-associated macrophages (TAMs), was biologically validated in orthotopic GBM models and cell lines.
Datasets
The study utilized the following datasets:
- Training and Validation Datasets: A combined meta-dataset of 377 primary IDH<sup>WT</sup> GBM patients from four public datasets:
- Chinese Glioma Genome Atlas (CGGA) - one microarray dataset and two RNA-seq datasets.
- The Cancer Genome Atlas (TCGA) - one RNA-seq dataset.
- Patients were randomly divided into a training set (n=204) and a validation set (n=173).
- External Validation Dataset: GSE7696 (n=70).
- Single-cell RNA-seq Dataset: GEO: GSE131928 for analyzing the expression distribution of signature genes in different cell populations within tumor tissue.
- Cell Lines: Human GBM cell lines (U87-MG, U251-MG) and control cell lines (HeLa S3, HEK293, HA1800) were used for in vitro validation.
Key Insights and Ideas
- Metabolic Reprogramming as a Prognostic Driver: The study strongly reinforces the idea that metabolic alterations are not just hallmarks of cancer but can be precisely quantified and used for prognostic stratification, even in aggressive tumors like IDH<sup>WT</sup> GBM.
- Gene Pairs as Robust Biomarkers: The use of gene pairs and a relative ranking approach proved effective in mitigating batch effects and technical variability, leading to a more robust and generalizable signature.
- TAMs and ABCA1: A Critical Link: The most striking insight is the identification of ABCA1 expression on tumor-associated macrophages (TAMs) as a critical prognostic factor. This points to the immune microenvironment, specifically macrophage lipid metabolism, as a key player in GBM progression.
- Immune Microenvironment Differences: High-risk patients exhibited distinct immune profiles, characterized by increased monocytes and decreased activated dendritic cells and γδ T cells, suggesting a more immunosuppressive tumor microenvironment.
- Translational Potential: The signature's ability to stratify patients based on treatment response (chemo/radiotherapy) and its validation in cell lines suggest a direct translational potential for clinical decision-making.
Key Results
- A prognostic signature consisting of 21 metabolic-related gene pairs (MRGPs) was developed and validated.
- This MRGP signature was an independent prognostic factor for overall survival in IDH<sup>WT</sup> GBM patients, outperforming another existing 9-gene signature.
- The signature accurately stratified patients into high- and low-risk groups across multiple datasets (C-index ranging from 0.648 to 0.801).
- The high-risk group showed significantly enriched immune-related biological processes, including humoral immune response and phagocytosis.
- ABCA1 expression in TAMs was found to increase with tumor progression and positively correlated with TAM population size, suggesting its role in promoting an inflammatory phenotype.
- Modulating ABCA1 expression (e.g., with lovastatin) in vitro and in vivo showed potential in shifting TAM polarization towards an inflammatory phenotype and controlling tumor growth.
Limitations and Future Work
- Focus on ABCA1: While the study identified 5 MRGs significantly changed in TAMs, it focused on ABCA1 for biological validation due to the complexity of characterizing all prognostic MRGs in TAMs. Future work could explore the roles of other identified MRGs.
- In Silico and Pre-clinical Validation: While robust, the study's validation relied heavily on in silico analyses and pre-clinical models. Clinical validation in prospective patient cohorts would be the next crucial step.
- Tumor Purity and Immune Infiltration: While differences in immune infiltration were observed, the study noted no significant difference in overall tumor purity between risk groups. Further investigation into the specific cellular composition and interactions within the tumor microenvironment is warranted.
- Translational Application: The authors suggest that the MRGP signature could be used for individualized management of primary IDH<sup>WT</sup> GBM patients, implying a need for further development and implementation into clinical practice.
Summary updated