# Comprehensive genomic characterization of programmed cell death-related genes to predict drug resistance and prognosis for patients with multiple myeloma

## TL;DR

This study identifies six molecular subtypes of multiple myeloma based on programmed cell death genes, offering insights into prognosis, immune response, and drug resistance.

## Contribution

A novel 12-gene risk score model using PCD-related genes to predict survival and drug resistance in multiple myeloma patients.

## Key findings

- Six PCD-related molecular subtypes were identified, with three showing higher immune escape and two linked to worse prognosis.
- The C3 subtype showed activated oxidative phosphorylation and DNA repair pathways, while C2 and C4 showed apoptosis-related pathways.
- A risk score model accurately predicted overall survival, with high-risk patients showing low survival and distinct drug resistance profiles.

## Abstract

Background: Multiple myeloma (MM) is a cancer that is difficult to be diagnosed and treated. This study aimed to identify programmed cell death (PCD)-related molecular subtypes of MM and to assess their impact on patients’ prognosis, immune status, and drug sensitivity.

Methods: We used the ConsensusClusterPlus method to classify molecular subtypes with prognostically relevant PCD genes from the MM patients screened. A prognostic model and a nomogram were established applying one-way COX regression analysis and LASSO Cox regression analysis. MM patients’ sensitivity to chemotherapeutic agents was predicted for at-risk populations.

Results: Six molecular subtypes were classified employing PCD-related genes, notably, three of them had a higher tendency for immune escape and two of them were correlated with a worse prognosis of MM. Furthermore, the C3 subtype had activated pathways such as oxidative phosphorylation and DNA repair, while the C2 and C4 subtypes had activated pathways related to apoptosis. The Risk score showed that the nomogram can correctly predict the OS for MM patients, in particular, patients in the high-risk group had low overall survival (OS). Pharmacovigilance analyses revealed that patients in the high-risk and low-risk groups had greater IC50 values for the drugs SB505124_1194 and AZD7762_1022, respectively.

Conclusions: A 12-gene Risk score model developed with PCD-related genes can accurately predict the survival for MM patients. Our study provided potential targets and strategies for individualized treatment of MM.