Multi-omics integration reveals a six-malignant cell maker gene signature for predicting prognosis in high-risk...

Neuroblastoma is the most common extracranial solid tumor of childhood, arising from the sympathetic nervous system. High-risk neuroblastoma (HRNB) remains a major therapeutic challenge with low survival rates despite the intensification of therapy. This study aimed to develop a malignant-cell marker gene signature (MMGS) that might serve as a prognostic indicator in HRNB patients.

Multi-omics datasets, including mRNA expression (single-cell and bulk), DNA methylation, and clinical information of HRNB patients, were used to identify prognostic malignant cell marker genes. MMGS was established by univariate Cox analysis, LASSO, and stepwise multivariable Cox regression analysis. Kaplan-Meier (KM) curve and time-dependent receiver operating characteristic curve (tROC) were used to evaluate the prognostic value and performance of MMGS, respectively. MMGS further verified its reliability and accuracy in the independent validation set. Finally, the characteristics of functional enrichment, tumor immune features, and inflammatory activity between different MMGS risk groups were also investigated.

We constructed a prognostic model consisting of six malignant cell maker genes (MAPT, C1QTNF4, MEG3, NPW, RAMP1, and CDT1), which stratified patients into ultra-high-risk (UHR) and common-high-risk (CHR) group. Patients in the UHR group had significantly worse overall survival (OS) than those in the CHR group. MMGS was verified as an independent predictor for the OS of HRNB patients. The area under the curve (AUC) values of MMGS at 1-, 3-, and 5-year were 0.78, 0.693, and 0.618, respectively. Notably, functional enrichment, tumor immune features, and inflammatory activity analyses preliminarily indicated that the poor prognosis in the UHR group might result from the dysregulation of the metabolic process and immunosuppressive microenvironment.

This study established a novel six-malignant cell maker gene prognostic model that can be used to predict the prognosis of HRNB patients, which may provide new insight for the treatment and personalized monitoring of HRNB patients.