Image_1_Prognostic and Predictive Value of an Immunoscore Signature in Glioblastoma Multiform.tif (326.99 kB)
Download file

Image_1_Prognostic and Predictive Value of an Immunoscore Signature in Glioblastoma Multiform.tif

Download (326.99 kB)
figure
posted on 10.11.2020, 08:41 authored by Xiangjun Tang, Pengfei Xu, Ann Chen, Gang Deng, Shenqi Zhang, Lun Gao, Longjun Dai, Qianxue Chen
Background

Although increasing evidence shows that immune infiltration plays an essential role in glioblastoma (GBM), current prognostic indicators do not accurately represent the risk of immune cells infiltration in patients. It is therefore critical to identify new prognostic markers for GBM. Here, we investigated the effectiveness of using immunoscore to improve risk stratification and prediction of prognosis in GBM patients receiving chemotherapy.

Methods

Using mRNA microarrays and CIBERSORT, we analyzed 22 types of immune cell fractions in 517 GBM samples and characterized an immunoscore using the least absolute shrinkage and selection operator (LASSO) Cox regression model based on the fraction of immune cell types and patients’ overall survival. The prognostic and predictive accuracy of immunoscore was verified in the validation cohort and the entire cohort.

Results

Using the LASSO model, an immunoscore was developed to classify patients into High and Low immunoscore groups in the training cohort (P < 0.0001) based on the fraction of eight immune cell types. The immunoscore performance was validated in the validation cohort (P < 0.0001) and the entire cohort (P < 0.0001). Furthermore, a nomogram comprising age, IDH1 status, and immunoscore was generated to predict one- and three-year survival rates in the training cohort. The predictive value of the immunoscore was also confirmed in the validation cohort and the entire cohort (C-index: 0.66, 0.67, and 0.68, respectively). In addition, we concluded that patients in the low-immunoscore group may benefit from adjuvant chemotherapy for GBM.

Conclusion

Immunoscore, an immune-infiltration-based signature, is a reliable prognostic and predictive tool for GBM.

History

References