Introduction

Distinguishing between low-grade and high-grade prostate cancer (PCa) as detected by biopsy results is very important, not only for diagnosis, but also for monitoring patients on active surveillance. However, biopsy results may underestimate the actual grade of the PCa when prostatectomy is performed. _x000D_ We developed an algorithm to predict the presence of high-grade PCa on biopsy using cell-free RNA (cfRNA) levels of UAP1, PDLIM5, IMPDH2, HSPD1, PCA3, PSA, TMPRSS2, AR, PTEN, and ERG genes in both urine and peripheral blood plasma. We show that this algorithm is highly reliable in predicting high grade (Gleason ≥ 3+4) PCa based on biopsy results in 489 patients. We also show in a prospective study that this algorithm is very reliable in an additional 300 patients who underwent prostatectomy.

Methods

Levels of cfRNA in urine and plasma from 489 patients were quantified using real-time PCR. Patients were selected randomly based on clinical suspicion of the presence of PCa. In addition, urine and blood samples from 300 patients were collected and tested prior to performing prostatectomy.

Results

Biopsy results from the first group showed Gleason 3+3 PCa in 103 patients (21%) and Gleason ≥3+4 PCa in 147 patients (30%). Patients with Gleason ≥4+3 were 55 (11%). Patients with Gleason 3+3 were grouped with non-cancer patients and considered as one group._x000D_ We used a simple algorithm incorporating prostate size, age, serum PSA along with 5 biomarkers to divide patients into two separate groups as low-risk and high-risk. Then each group was further refined and separated using a second algorithm implementing the rest of the biomarkers along with prior history of prostate biopsy. This resulted in high accuracy in predicting the presence of high-grade PCa with sensitivity between 97% and 86% and a specificity between 36% and 57%, dependent on which cut-off point was used. Sensitivity for predicting PCa Gleason ≥4+3 was between 99% and 96% and specificity between 37% and 59%. Diagnosis of Gleason ≥3+4 was missed in 1% to 3% of tested patients and of Gleason ≥4+3 in 0.2% to 1%. Testing the additional 300 samples with prostatectomy data demonstrated accurate prediction of cancer missing only 2.6% to 7% of Gleason ≥3+4 and 1 to 3% of ≥4+3, dependent on which cut-off point is used. _x000D_

Conclusions

Taking advantage of urine and plasma biomarkers as well as serum PSA and prostate size and prior history of biopsy, we were able to predict high grade prostate cancer with negative predictive value (NPV) of 97% to 90% for Gleason ≥3+4 and between 99% to 98% for Gleason ≥4+3. Furthermore, this test was further proved to be highly sensitive as confirmed by prostatectomy data.

Funding

None