A first-generation multiplex biomarker analysis of urine for the early detection of prostate cancer

Abstract

Although prostate-specific antigen (PSA) serum level is currently the standard of care for prostate cancer screening in the United States, it lacks ideal specificity and additional biomarkers are needed to supplement or potentially replace serum PSA testing. Emerging evidence suggests that monitoring the noncoding RNA transcript PCA3 in urine may be useful in detecting prostate cancer in patients with elevated PSA levels. Here, we show that a multiplex panel of urine transcripts outperforms PCA3 transcript alone for the detection of prostate cancer. We measured the expression of seven putative prostate cancer biomarkers, including PCA3, in sedimented urine using quantitative PCR on a cohort of 234 patients presenting for biopsy or radical prostatectomy. By univariate analysis, we found that increased GOLPH2, SPINK1, and PCA3 transcript expression and TMPRSS2:ERG fusion status were significant predictors of prostate cancer. Multivariate regression analysis showed that a multiplexed model, including these biomarkers, outperformed serum PSA or PCA3 alone in detecting prostate cancer. The area under the receiver-operating characteristic curve was 0.758 for the multiplexed model versus 0.662 for PCA3 alone (P = 0.003). The sensitivity and specificity for the multiplexed model were 65.9% and 76.0%, respectively, and the positive and negative predictive values were 79.8% and 60.8%, respectively. Taken together, these results provide the framework for the development of highly optimized, multiplex urine biomarker tests for more accurate detection of prostate cancer.

Figure 1

Characterization of candidate urine-based biomarkers of prostate cancer. A to C, qPCR was performed on WTA cDNA from urine obtained from patients presenting for needle biopsy or prostatectomy. Biomarker expression in patients with negative needle biopsies (green) or patients with prostate cancer (PCa; positive needle biopsy or prostatectomy; red) is shown. Normalization was performed using −ΔC_t, with PCA3 normalized to urine PSA expression as performed previously (25). AMACR, ERG, GOLPH2, SPINK1, and TFF3 were normalized to the average of urine sediment PSA and GAPDH expression. TMPRSS2:ERG gene fusion expression was dichotomized as positive or negative. The −ΔC_t values of genes that were not significant predictors of prostate cancer by univariate analysis (see Table 1) are shown in A, and the expression of those that were significant predictors is shown in B and C. P values from the univariate analysis for the detection of prostate cancer are indicated. D, ROC curves for individual variables for the diagnosis of prostate cancer. AUCs for GOLPH2, PCA3, SPINK1, and serum PSA are 0.664, 0.661, 0.642, and 0.508, respectively.

Figure 2

A multiplexed set of urine biomarkers outperforms PCA3 alone in the detection of prostate cancer. A, multivariate regression analysis resulted in a multiplexed model, including SPINK1, PCA3, GOLPH2, and TMPRSS2:ERG as significant predictors of prostate cancer (see Table 1). ROC analysis was then performed based on the predicted probabilities derived from the final model. The multiplexed model (red) showed significantly greater AUC than PCA3 (blue) alone (0.758 versus 0.662; P = 0.003) for the detection of prostate cancer. The point on the ROC curve with the maximum sum of sensitivity (Sens) and specificity (Spec) is indicated by the dashed line, and the positive (PPV) and negative (NPV) predictive values are given. B, as in A, except LOOCV strategy was used to generate unbiased AUCs. The AUC for the LOOCV multiplex model is significantly better than LOOCV of PCA3 alone (0.736 versus 0.645; P = 0.006).