Although survival rates for people with multiple myeloma (MM) have significantly improved over the last 15 years due to advances in treatment, relapse occurs in nearly everyone with the disease. This creates a need for risk stratification to find reliable prognostic markers that can help provide a more accurate prognosis for patients with MM and guide their clinical treatment.

A team of researchers based in China set out to develop a risk score model that could help improve the accuracy of risk stratification. Their findings were published in Frontiers in Genetics.

The researchers developed their model based on myeloma gene expression profiles from 3 independent data sets: GSE6477, GSE13591, and GSE24080. They found a 4-gene model was not as valid as a 5-gene model in terms of effectively predicting a prognostic outcome. So, they built their risk score model using EPAS1, ERC2, PRC1, CSGALNACT1, and CCND1 genes, which were shown on multivariate Cox regression analysis to be significantly associated with prognosis in patients with MM.


Continue Reading

These genes are considered protective genes, except for PRC1, which was deemed a harmful gene.

The researchers validated the predictive ability of the model by analyzing 3 testing data sets: MMRF, GSE2658, and GSE136337. Patients were categorized as high risk or low risk based on their median risk score. The low-risk group was found to have longer overall survival than the high-risk group.

“The risk score model was confirmed to be an independent prognostic factor in multiple analyses that included genetic factors and clinical factors,” the researchers explained. “Compared with other prognostic models, our model predicted survival outcomes effectively and were applied to predict the prognosis of patients with high-risk ISS/R-ISS stage or patients without high-risk factors innovatively.”

The large sample sizes were advantageous for this research, but the researchers also suggested that there may be some biases between the different platforms from the multiple data sets. This study was a retrospective study, and they called for future prospective studies to confirm the results.

Reference

Chen X, Liu L, Chen M, et al. A five-gene risk score model for predicting the prognosis of multiple myeloma patients based on gene expression profiles. Front Genet. Published online November 30, 2021. doi:10.3389/fgene.2021.785330

This article originally appeared on Oncology Nurse Advisor