A multiplex model of combining gene-based, protein-based, and metabolite-based with positive and negative markers in urine for the early diagnosis of prostate cancer.

BACKGROUND

Multiplex urine-based assay emerged outperforms single biomarker (e.g., prostate-specific antigen, PSA) for predicting prostate cancer (CaP), whereas its combined mode has to be fully optimized. Our aim is to determine whether a strategy of combining gene-based, protein-based, metabolite-based with positive, negative makers in urine could optimize a multiplex model for detecting CaP.

METHODS

Using quantitative PCR, Western blot, and liquid chromatography-mass spectrometry, expression patterns of PCA3, TMPRSS2: ERG, Annexin A3, Sarcosine, and urine PSA were evaluated in urine samples from 86 untreated patients with CaP and 45 patients with no evidence of malignancy. Multivariate logistic regression analysis was used to generate a final model and receiver-operating characteristic (ROC) analysis and special bootstrap software to assess diagnostic performance of tested variables.

RESULTS

The expression patterns of PCA3, TMPRSS2: ERG, Annexin A3, Sarcosine, and a panel including these biomarkers were significant predictors of CaP both in patients with PSA 4-10 ng/ml and in all patients (all P < 0.05). Employing ROC analysis, the area under the curves of the panel in these both cohorts were 0.840 and 0.856, respectively, which outperform that of any single biomarker (PCA3: 0.733 and 0.739; TMPRSS2: ERG: 0.720 and 0.732; Annexin A3: 0.716 and 0.728; Sarcosine: 0.659 and 0.665, respectively).

CONCLUSIONS

Compared with single biomarker, the multiplex model including PCA3, TMPRSS2: ERG, Annexin A3 and Sarcosine adds even more to the diagnostic performance for predicting CaP. Further validation experiments and optimization for the strategy of constructing this model are warranted.