C. Sweeney

BACKGROUND: Currently, there is no universally accepted prognostic classification for patients (pts) with metastatic hormone sensitive prostate cancer (mHSPC) treated with androgen deprivation therapy (ADT). Subgroup analyses demonstrated that pts with low volume (LV), per CHAARTED trial definition, mHSPC, and those who relapse after prior local therapy (PLT) have longer overall survival (OS) compared to high volume (HV) and de-novo (DN), respectively. Using a hospital-based registry, we aimed to assess whether a classification based on time of metastatic disease (PLT vs DN) and disease volume (LV vs HV) are prognostic for mHSPC pts treated with ADT. METHODS: A retrospective cohort of consecutive patients with mHSPC treated with ADT between 1990 and 2013 was selected from the prospectively collected Dana-Farber Cancer Institute database and categorized as DN or PLT and HV or LV, at time of ADT start. Primary and secondary endpoints were OS and time to castration-resistant prostate cancer (CRPC), respectively, which were measured from date of ADT start using Kaplan-Meier method. Multivariable Cox proportional hazards models using known prognostic factors was used. RESULTS: The analytical cohort consisted of 436 patients. The median OS and time to CRPC for PLT/LV were 92.4 (95%CI: 80.4-127.2) and 25.6 (95%CI: 21-35.7) months and 43.2 (95%CI: 37.2-56.4) and 12.2 (95%CI: 9.8-14.8) months for DN/HV, respectively, whereas intermediate values were observed for PLT/HV and DN/LV. A robust gradient for both outcomes was observed (Trend test P < 0.0001) in the four groups. In a multivariable analysis, DN presentation, HV, and cancer-related pain were independent prognostic factors. CONCLUSIONS: In our hospital-based registry, time of metastatic presentation and disease volume were prognostic for mHSPC pts treated with ADT. This simple prognostic classification system can aid patient counseling and future trial design.

4594 Background: Elevated markers of bone turnover at baseline are prognostic for poorer cancer outcomes. Suppressed levels on therapy may be prognostic for better ADT efficacy in controlling prostate cancer. Methods: Markers of bone turnover were assessed at baseline and 6 months in a prospective prostate cancer trial of patients with bone metastases commencing ADT. Association with time to castration resistant prostate cancer (CRPC) and skeletal related event (SRE) was evaluated using a MVA model with a Cox regression analysis with covariates of treatment assignment to 30 mg/day of risedronate versus placebo and known prognostic variables, PSA nadir < 0.2 ng/ml vs. > 0.2 ng/ml and extent of metastases. Results: Between 12/03 and 8/05, 63 patients were enrolled with median follow-up of 39.7 months. BTO markers were available for 50 patients at baseline and 40 patients at 6 months. The median values at baseline and 6 months were urine n-telopeptide (nmolBCE/mmol Creat) - 42.65/27.40; urine total DPD (nmol/mmol Creat) - 9.84/8.30; bone specific alkaline phosphatase (ng/mL) - 20.35/9.77 and serum osteocalcin (μg/mL) -19.43/16.82. On MVA, 6 month BTO markers which were all below the baseline median BTO marker level (p=0.014), nadir PSA below 0.2 ng/ml (p=0.042) and lower volume of metastases at baseline (p=0.033), were associated with longer time to SRE. 6 month BTO biomarkers below the baseline median and PSA nadir < 0.2ng/ml were the factors associated with longer time to CRPC (p=0.026 and p=0.058 respectively) and 6 month BTO markers below the baseline median was the only factor weakly associated with OS (p=0.092) in this study in men on ADT for metastatic prostate cancer with bone involvement. Conclusions: Lower bone turnover markers at 6 months of ADT is prognostic for better outcomes with ADT.

51 Background: Significant variation in response duration and overall survival exists among prostate cancer (PCa) patients treated with ADT. Germ-line SNPs affecting function of genes critical to hormone synthesis, transport, metabolism, binding sites, and degradation may contribute to variability in clinical outcomes observed in PCa patients treated with ADT. Methods: Between 1/07 and 10/08, all PCa patients seen in the Indiana University Simon Cancer Center oncology clinics were approached for recruitment to the Prostate Cancer Genetic Risk Evaluation of SNPs Study (PROGRESS). Participants completed a demographic and clinical questionnaire and provided a peripheral blood sample. Only patients with confirmed ADT initiation dates were included in this analysis. Germ-line DNA was analyzed for SNP genotyping on a 128-SNP chip using a TaqMan OpenArray GT Kit (Applied Biosystems). The chip included genes critical to hormone signaling, transport, and elimination pathways with minor allele frequencies &gt; 5%. Patients were followed for progression-free survival (PFS) and overall survival (OS) endpoints. Univariable analyses were performed to identify significant associations between SNP genotype, clinical parameters, and PFS and OS outcomes. Results: 107 patients with PCa initiated on ADT enrolled. Demographics included: age (median)–69 yrs, prostate specific antigen (PSA) (median)–28.0 ng/ml, PSA doubling time (median)–4.9 months, biochemical/metastatic–25%/75%, concurrent anti-androgen therapy–44%. No clinical parameters were associated with PFS and OS. Significant SNP associations with PFS and OS are summarized in the Table. Conclusions: Interpatient differences in hormone pathway germ-line SNPs may contribute to variability in clinical outcomes in patients treated with ADT. [Table: see text] [Table: see text]