Michael J. Shulman

Change in prostate specific antigen following androgen stimulation is an independent predictor of prostate cancer diagnosis.

PURPOSE We tested the hypothesis that a single exogenous androgen injection in men with low prostate specific antigen would provoke a differential prostate specific antigen response that would correlate with the presence and volume of cancer at biopsy. MATERIALS AND METHODS Following institutional review board approval 40 men with prostate specific antigen between 2.5 and 4.0 ng/ml were given 1 intramuscular injection of 400 mg testosterone cypionate at the start of the study. Prostate specific antigen and early morning serum testosterone were measured at baseline, 48 hours, and weeks 1, 2 and 4. All men underwent 12-core transrectal ultrasound guided biopsy at week 4. RESULTS Of the 40 men 18 (45%) were diagnosed with prostate cancer. The mean change in prostate specific antigen from baseline to 4 weeks was 3.1 to 3.4 ng/ml (9.7%) in men found to have benign findings on biopsy compared to a mean increase of 2.9 to 3.8 ng/ml (29%) in those with prostate cancer (p = 0.006). The change in prostate specific antigen following androgen stimulation was significantly associated with the percent of tissue involved with cancer and it was an independent predictor of cancer diagnosis on univariate and multivariate analysis. CONCLUSIONS An increase in prostate specific antigen following androgen stimulation in men with prostate specific antigen between 2.5 and 4.0 ng/ml was highly predictive of the subsequent diagnosis of prostate cancer and it correlated with disease volume. If these findings are corroborated, prostate specific antigen provocation may become an important strategy to identify men at risk for harboring prostate cancer and minimize the number undergoing unnecessary biopsies.

Intraluminal crystalloids are highly associated with prostatic adenocarcinoma on concurrent biopsy specimens

Prostatic crystalloids are intraluminal eosinophilic structures with variable size and shape. Their presence has been described in conjunction with the occurrence of prostatic adenocarcinoma (pCA). We herein report the association of crystalloids and pCA in a prospective trial utilizing an extended multi-site transrectal ultrasound-guided (TRUS) prostate biopsy protocol. Three hundred and forty-four consecutive patients were prospectively enrolled at the Dallas Veterans Administration Hospital from November 2002 to September 2003. Indications for biopsy included a prostate-specific antigen (PSA) ⩾4 ng/ml and/or abnormal digital rectal exam. A single pathologist evaluated all biopsy cores and documented the presence or absence of significant histopathologic features. Univariate and multivariate logistic regression analysis were applied to test the association of these features with the presence of pCA on concurrent biopsy. Median number of core biopsies per patient was 12 (range 3–36). Overall cancer detection rate was 42.7%. pCA was diagnosed in 66 (81.5%) of 81 patients with crystalloids, 70 (69.3%) of 101 patients with high-grade prostatic intraepithelial neoplasia (HGPIN), and 32 (84.2%) of 38 patients with both HGPIN and crystalloids on biopsy. Multivariate analysis identified crystalloids (RR 4.53, 95% CI 2.30–8.88) and HGPIN (RR 3.20, 95% CI 1.84–5.57) as independent predictors of the presence of cancer on concurrent biopsy (P<0.001). In this prospective analysis, crystalloids were significantly associated with pCA on concurrent biopsy and more predictive of the presence of pCA than HGPIN. These findings suggest that the presence of crystalloids alone or in combination with HGPIN in prostate biopsies may be a more compelling indication for repeat biopsy than HGPIN alone.