Identification and validation of a histone modification-related gene signature to predict the prognosis of multiple myeloma.

Abstract

Background: Multiple myeloma (MM) is an incurable plasma cell malignancy with high heterogeneity. Current staging systems, including the International Staging System (ISS) and Revised ISS (R-ISS), have limited prognostic accuracy. Given the role of histone modifications in MM progression, we developed a histone modification-related (HMR) prognostic model to improve MM risk stratification.

Methods: Gene expression and mutation data were downloaded from the Gene Expression Omnibus database and the Cancer Genome Atlas. Prognostic HMR-related genes were identified through a combination of univariate Cox regression, least absolute shrinkage and selection operator Cox regression, and random survival forest analysis. The genes identified were then used to construct the HMR risk score model using multivariate Cox regression. The model was validated using Kaplan-Meier survival, time-dependent receiver operating characteristic curves analysis. A nomogram combining the HMR score with clinical features was developed. Functional enrichment, immune infiltration, somatic mutation, and drug sensitivity analysis were conducted to explore the biological relevance of the model.

Results: Seven HMR genes with prognostic significance were identified. The HMR risk score stratified patients into high-risk and low-risk groups, with significant survival differences. The model demonstrated favorable predictive performance, and was shown to be an independent prognostic factor. The nomogram showed good calibration and discriminative ability, offering a practical tool for individual patient risk assessment. Functional analysis revealed that the HMR risk score is associated with dysregulated cell cycle progression, proliferation, and immunosuppression in MM, which may contribute to disease progression and drug resistance. Moreover, drug sensitivity analysis indicated potential associations between the HMR score and response to specific therapeutic agents, highlighting its potential role in guiding personalized treatment.

Conclusion: We developed an HMR gene signature that has potential for prognostic prediction and may help guide personalized treatment strategies in MM.