Objective: The aim of this study was to analyze tumour gene expression profiles of progressive and non-progressive T1G3 bladder cancer (BC) patients to develop a gene expression signature to predict tumour progression.
Methods: Retrospective, multicenter study of 96 T1G3 BC patients without carcinoma in situ (CIS) who underwent a transurethral resection. Formalin-fixed paraffin-embedded tissue samples were collected. Global gene expression patterns were analyzed in 21 selected samples from progressive and non-progressive T1G3 BC patients using Illumina microarrays. Expression levels of 94 genes selected based on microarray data and based on literature were studied by quantitative polymerase chain reaction (qPCR) in an independent series of 75 progressive and non-progressive T1G3 BC patients. Univariate logistic regression was used to identify individual predictors. A variable selection method was used to develop a multiplex biomarker model. Discrimination of the model was measured by area under the receiver-operating characteristic curve. Interaction networks between the genes of the model were built by GeneMANIA Cytoscape plugin.
Results: A total of 1294 genes were found differentially expressed between progressive and non-progressive patients. Differential expression of 15 genes was validated by qPCR in an additional set of samples. A five-gene expression signature (ANXA10, DAB2, HYAL2, SPOCD1, and MAP4K1) discriminated progressive from non-progressive T1G3 BC patients with a sensitivity of 79% and a specificity of 86% (AUC = 0.83). Direct interactions between the five genes of the model were not found.
Conclusions: Progressive and non-progressive T1G3 bladder tumours have shown different gene expression patterns. To identify T1G3 BC patients with a high risk of progression, a five-gene expression signature has been developed.
Keywords: Biomarkers; Bladder cancer; Gene expression signature; Prognosis; Progressive disease.
Copyright © 2016 Elsevier Ltd. All rights reserved.