Login to Access Video or Poster Abstract: MP28-09
Sources of Funding: None

Introduction

We investigated the correlation between serum dehydroepiandrosterone (DHEA) levels and Gleason scores at biopsy.

Methods

We analyzed data of 203 men with a total PSA below 10 ng/ml who underwent initial systematic prostate needle biopsy for suspected prostate cancer. Peripheral blood DHEA concentrations were determined by liquid chromatography with two serially linked mass spectrometers (LC-MS/MS). Blood levels of DEHA was compared with pathological findings by multivariate analyses.

Results

Median values of patients&[prime] age, PSA and prostate volume measured by ultrasound were 68 years, 5.5 ng/mL, and 31.2 mL, respectively. Benign prostate was diagnosed in 118 patients (58.1%) and prostate cancer was diagnosed in 85 (41.9%) patients, including 31 (15.3%) patients with a Gleason score of 6, 36 (17.7%) patients with a Gleason score of 7, and 18 (8.9%) patients with a Gleason score of 8-10. Median values of DHEA in blood was 1654.7 pg/mL. In multivariate analysis, PSA (p=0.009), prostate volume (p=0.022), and serum DHEA (p=0.041) were significant predictors of benign prostate or prostate cancer with a Gleason score of ≤7. The DHEA cutoff point to predict benign prostate or cancer with a Gleason score ≤7 was established at 2179 pg/mL, with sensitivity, specificity, positive predictive value, and negative predictive value of 33.0%, 100%, 100%, and 12.7%, respectively. The area under the receiver operating characteristics curve (AUROC) for predicting benign prostate or cancer with a Gleason score ≤7 of the base model (age, PSA, and prostate volume) was 0.74. The AUROC of the base model plus DHEA was 0.81. Addition of DHEA to the base model significantly improved AUROC compared to base model._x000D_

Conclusions

We confirmed that high DHEA blood levels can predict benign prostate or prostate cancer with a Gleason score ≤7 in men with PSA levels below 10 ng/mL. Serum DHEA could help to distinguish the patients with more indolent disease, which would be useful for selecting suitable patients for active surveillance form those with more aggressive disease.

Funding

None