Establishing cut-off points with clinical relevance for bcl-2, cyclin D1, p16, p21, p27, p53, Sox11 and WT1 expression in glioblastoma - a short report

• Published in: Cellular oncology (Dordrecht) 2018-04, Vol.41 (2), p.213-221
• Main Authors: Camacho-Urkaray, Emma, Santos-Juanes, Jorge, Gutiérrez-Corres, Francisco Borja, García, Beatriz, Quirós, Luis M., Guerra-Merino, Isabel, Aguirre, José Javier, Fernández-Vega, Iván
• Format: Article
• Language: English
• Subjects: Bcl-2 protein; Biomarkers; Biomedical and Life Sciences; Biomedicine; Cancer Research; Cyclin D1; Cyclin D1 - metabolism; Cyclin-Dependent Kinase Inhibitor p16 - metabolism; Cyclin-dependent kinase inhibitor p21; Cyclin-Dependent Kinase Inhibitor p21 - metabolism; Cyclin-dependent kinase inhibitor p27; Female; Glioblastoma; Glioblastoma - metabolism; Humans; Immunohistochemistry; Male; Middle Aged; Mutation; Oncology; p53 Protein; Pathology; Proteins; SOXC Transcription Factors - metabolism; Survival; Tissue Array Analysis; Tumor Suppressor Protein p53 - metabolism; WT1 Proteins - metabolism
• ISSN: 2211-3428; 2211-3436
• DOI: 10.1007/s13402-017-0362-4

Purpose Glioblastoma (GBM) ranks among the most challenging cancers to treat and there is an urgent need for clinically relevant prognostic and diagnostic biomarkers. Here, we set out to investigate the expression of eight proteins (bcl-2, cyclin D1, p16, p21, p27, p53, Sox11 and WT1) in GBM with the specific aim to establish immunohistochemistry cut-off points with clinical relevance. Methods Immunohistochemistry (IHC) was used to examine protein expression in 55 surgical GBM specimens using H-scores, and IHC cut-off points were established using the Cutoff Finder web platform. Protein co-expression and its correlation with histopathological features were assessed, and cases were classified according to IDH1 mutation status. Survival curves were determined using Kaplan-Meier analyses. Results Clinical and molecular parameters found to be correlated with overall survival (OS) were tumor size ( r  = −0.278; p  = 0.048), p53 ( r  = −0.452; p  = 0.001), p16 ( r  = 0.351; p  = 0.012) and Sox11 ( r  = 0.324; p  = 0.020). In addition, we found that tumor size correlated with cyclin D1 ( r  = −0.282; p  = 0.037), p53 ( r  = 0.269; p  = 0.041), Sox11 ( r  = −0.309; p  = 0.022) and WT1 ( r  = −0.372; p  = 0.003). Variables found to be significantly associated with IDH1 mutation status were OS ( p   20 ( p  = 0.003), Sox11 score ≤ 40 ( p