Unveiling tumor senescence-driven prognostic heterogeneity via MALISS in stage II/III colorectal cancer.

MALISS was developed as a 30-gene prognostic signature using a CoxBoost-Lasso algorithm applied to transcriptomic data from 1296 patients with stage II/III colorectal cancer.

• The model was derived from transcriptomic data encompassing 1296 patients total
• The CoxBoost-Lasso algorithm was selected for final model construction among machine learning approaches
• The final signature comprises 30 genes focused on immunosenescence
• The model was validated across multiple independent cohorts beyond the discovery dataset

MALISS effectively stratified stage II/III CRC patients into high- and low-risk groups with distinct progression-free survival outcomes.

• Patients were dichotomized into high-risk and low-risk groups based on MALISS scores
• The signature demonstrated significant differences in progression-free survival between risk groups
• Validation was performed across multiple independent cohorts to confirm prognostic performance
• The signature addressed the prognostic heterogeneity challenge in stage II/III CRC clinical management

NR1D2 was identified as a key gene in MALISS that promotes tumor migration through cellular senescence.

• Functional analysis identified NR1D2 as a biologically significant gene within the MALISS signature
• NR1D2 was shown to promote tumor migration via a cellular senescence mechanism
• This finding links immunosenescence biology directly to tumor invasive behavior in CRC
• NR1D2's role was identified through functional analysis of the signature genes

The high-risk MALISS group was characterized by a unique mutational landscape, an altered tumor microenvironment, and differential drug sensitivity compared to the low-risk group.

• High-risk patients exhibited a distinct mutational landscape relative to low-risk patients
• The tumor microenvironment composition differed between high- and low-risk groups
• Differential drug sensitivity was observed between risk groups, suggesting potential therapeutic implications
• These characteristics collectively defined the biological distinctiveness of the high-risk group

A prognostic nomogram integrating MALISS with clinical biomarkers demonstrated improved predictive performance over MALISS alone.

• The nomogram combined MALISS scores with clinical biomarkers to enhance prognostication
• Integration of clinical variables with the molecular signature improved predictive performance
• The nomogram was developed to facilitate clinical translation of the MALISS tool
• This combined model was designed to address the prognostic heterogeneity in stage II/III CRC management