# The Development and Assessment of a Unique Disulfidptosis-Associated lncRNA Profile for Immune Microenvironment Prediction and Personalized Therapy in Gastric Adenocarcinoma

## TL;DR

This study identifies a new lncRNA-based model to predict outcomes and guide treatment in gastric cancer patients by analyzing disulfidptosis-related factors and immune responses.

## Contribution

A novel disulfidptosis-associated lncRNA risk model is developed for gastric adenocarcinoma prognosis and personalized therapy.

## Key findings

- A three-lncRNA risk model (AC107021.2, AC016394.2, AC129507.1) was validated for independent prognostic prediction in GC patients.
- High-risk patients showed increased immune cell infiltration and specific pathway enrichments like Wnt signaling and cAMP cascades.
- Low-risk patients exhibited higher tumor mutational burden and better response to drugs like gemcitabine and veliparib.

## Abstract

Background: Long non-coding RNAs (lncRNAs) are crucial factors affecting the occurrence, progression, and prognosis of gastric carcinoma (GC). The accumulation of disulfide bonds to excessive levels in cells expressing high SLC7A11 triggers disulfidptosis, which functions as a regulated form of cellular death. Research has demonstrated that upregulated SLC7A11 is common in human cancers, but the effect of disulfidptosis on GC remains unclear. Identifying lncRNAs associated with disulfidptosis (drlncRNAs) and establishing a prognostic risk profile holds considerable importance for advancing GC research and treatment. Methods: Clinical records and transcriptomic datasets from individuals with GC were acquired from The Cancer Genome Atlas (TCGA) repository. A three-drlncRNA risk model was built using three common regression analysis methods. Then we used receiver operating characteristic (ROC) curves, independent prognostic analysis, and additional statistical approaches to assess the precision of the model. This investigation additionally encompassed Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis, immune cell infiltration evaluation, and pharmacological sensitivity predictions. To further investigate immunotherapy response disparities between patient cohorts with elevated- and reduced-risk scores, analyses of tumor mutational burden (TMB), tumor immune dysfunction and exclusion (TIDE), and microsatellite instability (MSI) were implemented. Results: We constructed a unique model composed of three drlncRNAs (AC107021.2, AC016394.2, and AC129507.1). Its independent prognostic capability for GC patients was validated through both single-variable and multivariable Cox regression analyses. GO and KEGG pathway assessments revealed predominant enrichment within the elevated-risk cohort, particularly in pathways involving sulfur compound interactions, traditional Wnt signaling mechanisms, cell-substrate adherens junctions, and cAMP signaling cascades, among others. Tumor microenvironment (TME) evaluation demonstrated elevated ImmuneScores, StromalScores, and ESTIMATEScores within the high-risk patient population. Concurrently, this elevated-risk cohort exhibited enhanced immune cell infiltration patterns, whereas the reduced-risk group displayed superior expression of immune checkpoints (ICPs). Additional investigations revealed that patients categorized into the reduced-risk classification possessed greater tumor mutational burden, increased MSI-high proportions, and diminished tumor immune dysfunction and exclusion scores compared to their high-risk counterparts. Pharmacological sensitivity assessments confirmed the superior efficacy of several therapeutic agents, including gemcitabine and veliparib (ABT.888), in patients with lower risk classifications. Conclusions: Our established risk stratification system demonstrates independent prognostic predictive capacity while offering personalized clinical intervention guidance for individuals diagnosed with GC.