Clinical significance of a new early diagnostic model for bladder cancer based on genome-wide microarray profiling of serum exosomal lncRNAs

Abstract Purpose The aim of our report was to recognize bladder cancer (BC)-specific serum exosome-derived long noncoding RNAs (lncRNAs) profile for early diagnosis of BC. Methods Potential BC-specific exosomal lncRNA indicators were discerned by genome-wide microarray profiling analysis of serum exosomes from 10 healthy participants and 10 early-stage BC patients (Ta and T1), followed by multi-stage validation through quantitative real-time PCR (qRT-PCR) in BC cells, culture solution as well as 200 serum specimens and 50 tissue specimens from non-muscle-invasive bladder cancer (NMIBC) patients. The diagnostic panel was established using logistic regression and evaluated by receiver-operating characteristic (ROC) curve. Results In the training stage, a diagnostic panel was constructed based on three up-regulated exosomal lncRNAs (G023016, RP11-553N19.1 and LINC0087) in NMIBC patients compared with healthy controls, yielding an area under ROC curve (AUC) of 0.827. We verified tumor-derived origin of these three lncRNAs which existed steadily in serum because of being enclosed in exosomes. The three-lncRNA panel was demonstrated to perform well in terms of NMIBC diagnosis, revealing AUC values of 0.809 and 0.812 respectively in the following expanded validation stage and double-blind stage which was demonstrated to be significantly superior to that of urine cytology in double-blind stage (AUC = 0.630) (P < 0.0001). Moreover, serum exosome-derived G023016 significantly associated with tumor grade and TNM stage (P = 0.006 and P < 0.001, respectively), and LINC0087 significantly associated with TNM stage (P = 0.023). Conclusion The three-exosomal lncRNA signature could function as qualified blood-based non-invasive indicator for early diagnosis of BC.