Prognostic scoring system for osteosarcoma using network-regularized high-dimensional Cox-regression analysis and...

Area of Science:

• Oncology
• Genomics
• Bioinformatics

Background:

• Personalized genomic medicine increasingly relies on prognostic stratification using gene signatures in cancer.
• Existing prognostic models for osteosarcoma often overlook the integration of gene networks with clinical data.

Purpose of the Study:

• To develop a novel prognostic scoring system for osteosarcoma patients by incorporating gene networks and clinical data.
• To enhance the accuracy of patient prognosis prediction and identify potential therapeutic targets for osteosarcoma.

Main Methods:

• Utilized messenger RNA (mRNA) expression and clinical data from Gene Expression Omnibus for osteosarcoma patients.
• Employed Network-Regularized high-dimensional Cox-regression analysis for variable selection, integrating gene network information from pathway databases.
• Calculated a risk score based on regression coefficients and mRNA expression, evaluating discriminatory power using log-rank tests, C-index, and AUC values.

Main Results:

• The novel prognostic model demonstrated significant discriminatory power in predicting overall survival for osteosarcoma patients.
• Achieved an optimal C-index of 0.967, with time-dependent AUC values ranging from 0.953 to 1.000, indicating high predictive accuracy.
• Identified potential therapeutic targets, including BACE2 and ING2 knockdown, and linifanib treatment, for improving osteosarcoma patient prognosis.

Conclusions:

• The developed prognostic scoring system offers a more accurate prediction of patient prognosis in osteosarcoma.
• This novel approach facilitates the selection of appropriate therapeutic alternatives, advancing personalized medicine for osteosarcoma.
• The study highlights the importance of integrating gene networks into prognostic models for improved clinical outcomes.