# Unveiling fatty acid subtypes: immunometabolic interplay and therapeutic opportunities in gastric cancer

## TL;DR

This study identifies a fatty acid metabolism gene signature that predicts gastric cancer prognosis and treatment response, offering new therapeutic insights.

## Contribution

A novel fatty acid metabolism-based gene signature was developed for gastric cancer prognosis and treatment stratification.

## Key findings

- A risk model using fatty acid metabolism genes showed strong predictive performance (AUC of 0.86) for gastric cancer prognosis.
- High-risk patients responded better to embelin and imatinib, while low-risk patients showed higher response to immune checkpoint blockade.
- Transcriptomic and metabolomic integration revealed distinct fatty acid profiles between risk groups in gastric cancer.

## Abstract

The goal of this study was to develop a predictive signature using genes associated with fatty acid metabolism to evaluate the prognosis of individuals with gastric cancer (GC).

A total of 24 prognostic-related genes were identified by intersecting differentially expressed genes with 525 fatty acid metabolism (FAM) -related genes and applying a univariate Cox proportional hazards model. By performing consensus clustering of 24 genes associated with FAM, two distinct clusters of GC patients were identified. Subsequently, a risk model was constructed using 39 differentially expressed mRNAs from the two clusters through a random forest model and univariate Cox regression.

An R package, “GCFAMS”, was developed to assess GC patients’ prognosis based on FAM gene expression. The low-risk group exhibited a more favorable prognosis compared to the high-risk group across various datasets (P < 0.05). The model demonstrated strong predictive performance, with AUC values of 0.86, 0.623, and 0.508 for 5-year survival prediction in the training and two validation datasets. The high-risk group displayed lower IC50 values for embelin and imatinib, suggesting the potential efficacy of these drugs in this subgroup. Conversely, the low-risk group demonstrated an elevated response to immune checkpoints blockade therapy and a higher immunophenoscore, which was further validated in additional cancer cohorts. Public data from single-cell RNA sequencing confirmed that the characterized genes were predominantly expressed in endothelial cells and fibroblasts. Furthermore, the integration of transcriptomics and metabolomics revealed notable variations in fatty acid levels between the clusters, underscoring the clinical relevance of our fatty acid metabolism signature in shaping the metabolic profiles of GC patients.

This developed FAM signature demonstrated potential as a biomarker for guiding treatment and predicting prognosis in GC.