Abstract

Osteosarcoma (OS) is a form of primary bone malignancy associated with poor prognostic outcomes. Recent work has highlighted vasculogenic mimicry (VM) as a key mechanism that supports aggressive tumor growth. The patterns of VM-associated gene expression in OS and the relationship between these genes and patient outcomes, however, have yet to be defined. Here, 48 VM-related genes were systematically assessed to examine correlations between the expression of these genes and OS patient prognosis in the Therapeutically Applicable Research to Generate Effective Treatments (TARGET) cohort. Patients were classified into three OS subtypes. Differentially expressed genes for these three OS subtypes were then compared with hub genes detected in a weighted gene co-expression network analysis, leading to the identification of 163 overlapping genes that were subject to further biological activity analyses. A three-gene signature (CGREF1, CORT, and GALNT14) was ultimately constructed through a least absolute shrinkage and selection operator Cox regression analysis, and this signature was used to separate patients into low- and high-risk groups. The K-M survival analysis, receiver operating characteristic analysis, and decision curve analysis were adopted to evaluate the prognostic prediction performance of the signature. Furthermore, the expression patterns of three genes derived from the prognostic model were validated by quantitative real-time polymerase chain reaction (RT-qPCR). VM-associated gene expression patterns were successfully established, and three VM subtypes of OS that were associated with patient prognosis and copy number variants were defined. The developed three-gene signature was constructed, which served as independent prognostic markers and prediction factors for the clinicopathological features of OS. Finally, lastly, the signature may also have a guiding effect on the sensitivity of different chemotherapeutic drugs. Overall, these analyses facilitated the development of a prognostic VM-associated gene signature capable of predicting OS patient outcomes. This signature may be of value for both studies of the mechanistic basis for VM and clinical decision-making in the context of OS patient management.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call