Bladder Cancer Diagnostic Biomarkers Discovery Using Urine-based Metabolomics Method

Objective: As is known that bladder cancer(BCa) is a common malignancy worldwide and has a high probability of recurrence, so early detection is vital to improve the overall survival rate of bladder cancer patients. So far, diagnostic modalities, such as cystoscopy and urinary cytology, have limitations. There is a lack of a reliable urine-based assay for early detection of BCa. In this study, potential metabolic biomarkers have been discovered through gas chromatography-mass spectrometry based metabolomics of urine between BCa patients and healthy people. We tried to study bladder cancer through metabolomics could reveal new biomarkers that could be useful for its future prognosis, diagnosis and even therapy.Methods: This study includes two steps: Training Phase and Test Phase. During the Training Phase, urine samples were collected from 32 patients diagnosed of bladder cancer and 21 healthy controls. We applied unsupervised principal component analysis(PCA) and orthogonal partial least-squares discriminant analysis(OPLS-DA) model used as a diagnostic model to distinguish two groups. AUCs and ROC curves for differential metabolite profiling. We further constructed multivariate linear regression model using combinations of the metabolites to improve the sensitivity and specificity for early BCa determination. During the Test Phase, urine samples from 79 BC patients and 51 non-BC controls were subjected to urinary metabotyping. PCA and OPLS-DA were applied, and we tried to establish the model as a diagnostic panel to combine metabolites for BCa diagnosis.Results: In Training phase, a set of 22 candidate differential metabolites was selected based on statistical significance and fold difference. AUCs range from 0.7 to 0.8.The ROC analysis of Model1 showed significant improvement of sensitivity and specificity with AUC reaching 0.94.In Test phase, 14 differential metabolites were found between two groups, namely, Succinic acid, Glycerol 1-Phosphate,Glycine,Mimosine,D-Rhamnose,4-Aminobutyric acid, Phosphate, Kynurenic acid, Pyruvic acid, enol, 3, Phosphoglyceric acid, N-Acetyl-Aspartic acid, 2-Amino-4,6-Dihydroxypyrimidine, Methylcitric acid and Cytosine. Interestingly, thesemetabolites were the same metabotyping as Training phase. AUCs in this phase were lower than step one, the ROC analysis of Model1 showed significant improvement of sensitivity and specificity with AUC reaching 0.94.Moreover, there was significant difference in levels of 3-Phosphoglyceric acid and Cytosine for different stages and grades.Conclusion: These metabolites may have great potential to be used in the clinical diagnosis after further rigorous assessment. Our results suggest that urine metabolic profiling may have potential for early clinical diagnosis of bladder cancer. In addition to its utility, these metabolites may be used as new biomarkers as a diagnostic tool and, and may enhance our understanding of the mechanisms involved. The development of urine-based metabolomics to detect bladder cancer would be of tremendous benefit to both patients and the healthcare system.