# Identification and validation of pyroptosis patterns in AML via comprehensive bioinformatics analysis

## TL;DR

This study explores how pyroptosis patterns in AML patients affect survival and immune responses, identifying new biomarkers for better treatment strategies.

## Contribution

The study introduces a novel classification of AML based on pyroptosis patterns and links these patterns to immune activity and genomic alterations.

## Key findings

- Nine pyroptosis-related genes were identified as significant prognostic markers in AML.
- Two pyroptosis subtypes were classified, with ELANEhigh showing better survival and ELANElow linked to poor outcomes.
- ELANElow subtype is associated with an immunologically active microenvironment and specific cell death pathways.

## Abstract

Pyroptosis, a lytic inflammatory cell death mechanism, plays dual roles in tumorigenesis, but its clinical relevance in acute myeloid leukemia (AML) remains poorly understood. Through an integrative analysis of 40 pyroptosis-related genes in newly diagnosed AML patients (TCGA, n = 151) and healthy controls (GTEx, n = 386), we identified 32 genes with aberrant expression. Among these, 9 genes were found to be significant prognostic markers, including ELANE (protective), and CASP1, CHMP4B, BAK1, and CHMP2A (risk), which retained their prognostic significance after adjusting for age and gender. Using unsupervised nonnegative matrix factorization (NMF) on TCGA data, we classified AML into two pyroptosis patterns: the ELANEhigh subtype, associated with favorable survival, and the ELANElow subtype, which was enriched in poor karyotypes and adverse outcomes. This classification was validated in an independent cohort (GSE10358, n = 91). Single-cell RNA sequencing data (GSE116256, n = 15) revealed that the ELANElow subtype is characterized by an immunologically active microenvironment, marked by an expansion of cytotoxic T cells and naive CD4 + /CD8 + T cells. Factor analysis revealed associations between pyroptosis patterns and other forms of cell death, including ferroptosis, autophagy, and apoptosis, as well as with karyotype, leukemia stemness, and TP53/FLT3-ITD mutations. Prognostic immune gene sets enriched in the ELANElow subtype were associated with interferon signaling and ubiquitin-mediated degradation pathways. Furthermore, protein–protein interaction (PPI) network analysis identified three sub-networks and nine key hub genes. This study integrates gene expression data from newly diagnosed AML patients, revealing the heterogeneity of pyroptosis patterns within the population. It highlights the potential links between distinct pyroptosis patterns, the immune microenvironment, various cell death pathways, leukemia stemness, and genomic alterations, offering novel biomarkers and therapeutic targets for risk stratification and immunomodulatory interventions in AML.

The online version contains supplementary material available at 10.1007/s12672-025-02298-5.