Search

Chemotherapy is an integral part of the treatment of castration resistant prostate cancer. With the introduction of new drugs the need to identify nonresponders is increasing. To our knowledge there are no prognostic parameters to date for use upon the initiation of any treatment.cfDNA was isolated from a serum specimen before chemotherapy. Its value was correlated to recurrence-free and overall survival using Kaplan-Meier curves. Univariate and multivariate Cox regression analysis was performed to identify independent predictors.Of 59 men 48 (81.4%) had a measurable prostate specific antigen decrease from baseline. Median followup was 15.0 months (range 2.4 to 58.4). The median cfDNA concentration in all men in this study was 27.71 ng/ml (mean 32.64). A threshold of 55.03 ng/ml was significantly associated with a poor prostate specific antigen response of less than 30% (p = 0.005). On univariate and multivariate analysis circulating cfDNA was an independent predictor of overall survival (HR 0.36, 95% CI 0.13-0.97, p = 0.044 and HR 0.34, 95% CI 0.12-0.91, p = 0.032, respectively). Limitations of the study are its retrospective character, and first and second line therapies.Our trial shows that the cfDNA concentration before therapy may be a useful predictive and prognostic biomarker for prostate specific antigen response and survival.

e584 Background: Taxane-based chemotherapy is one of the first-line life prolonging therapeutic approaches in metastatic castration-resistant prostate cancer (mCRPCA). Until now, there is no valid surrogate marker directly reflecting therapeutic response. Prostate specific antigen (PSA) and imaging studies are used to monitor therapy. The aim of this study was to assess whether circulating cell-free DNA (cfDNA) is suitable as a predictive and prognostic marker. Methods: cfDNA was isolated in 74 patients with mCRPCA receiving taxane based chemotherapy. Median serum cfDNA - concentration was correlated with localization of metastasis and tumor burden (bone, lymph node and visceral). In addition we examined the change of cfDNA - concentration under treatment. The median serum cfDNA - concentration at the time of 1st, 5th and 10th cycle of chemotherapy was correlated with PSA response (>80%, >50%, 80%, 41 patients with a PSA decrease >50%, 12 patients with a PSA decrease 80% was 34.96 ng/ml, in patients with a PSA decrease > 50% 34.96 ng/ml and in patients with a PSA decrease < 30% 38.475 ng/ml. The Median cfDNA concentration in patients with a PSA rise was 31.25 ng/ml. Median cfDNA concentration in patients with lymph node metastasis was 33.78 ng/ml, with bone metastasis < 3 34.24 ng/ml, bone metastasis >3 31.97 ng/ml and with visceral metastasis 34.57 ng/ml. There was no significant correlation between median cfDNA - concentration and type of metastasis (cfDNA vs. PSA bone metastasis 3 P = 0.472, cfDNA vs. PSA lymph node metastasis P = 0.423). There was no significant change of median cfDNA concentration under treatment ( P > 0.05). Conclusions: Although cfDNA-concentration is a prognostic and predictive marker for survival and therapy response as shown recently, our findings suggest that cfDNA as a surrogate marker is unsuitable.