Search

12 Background: Increased use of prostate-specific antigen (PSA) has introduced an increase in PCa incidence and a lead time and stage migration at diagnosis, altering the natural history of PCa. Contemporary PCa patients are likely younger and have smaller tumor burden at diagnosis. We investigated if changes in the PCa landscape have altered the course of low-risk localized PCa. Methods: In the Danish Prostate Cancer Registry (DaPCaR), patients diagnosed from 1995 to 2011 with localized (T1-2, N0/X, M0) PCa with Gleason score ≤ 6 were identified. Patients were stratified into three periods of diagnosis; 1995-2000 (period 1), 2001-2005 (period 2) and 2006-2011 (period 3). Competing risk analysis treating PCa and other-cause death as competing events was performed. Results: Of the 5,660 patients identified, 35.9% had undergone radical prostatectomy (RP). From period 1 to period 3, the median age at diagnosis decreased from 72.2 to 66.0 years and the median PSA decreased from 16.2 to 8.6 ng/mL. From period 1 to period 3, the 5-year risk of PCa-death decreased from 14.3% (95% CI: 12.1-16.4%) to 1.3% (95% CI: 0.83-1.7%), p < .0.0001 and the risk of other cause death decreased from 18.1% (95% CI: 15.8-20.5%) to 7.2% (95% CI: 6.2-8.2), p = 0.0001. In patients undergoing RP, the 5-year risk of PCa-death decreased from 0.67% (95% CI: 0.67-2.0%) for patients diagnosed in period 1 to 0.45% (95% CI: 0.0055-0.84), for patients diagnosed in period 3, p = 0.92. For patients not undergoing RP, the 5-year risk of PCa death decreased from 16.6% (95% CI: 14.1-19.1) to 2.0% (95% CI: 1.2-2.7%), p < 0.0001. Conclusions: In a nationwide cohort of patients with low risk localized PCa, the 5-year risk of PCa-death significantly decreased when comparing patients diagnosed during 2006-2011 to those diagnosed during 1995-2000. This was mainly driven by patients not undergoing RP. In the most recently diagnosed group, the difference in 5-year PCa-death between patients undergoing RP and not undergoing RP was small (0.45% vs. 2.0%). Our data demonstrate that the impact of PSA induced lead-time and stage migration has diminished the absolute effect of RP on the risk of 5-year PCa-death because contemporary low-risk localized patients have a significantly better prognosis.

217 Background: Prostate specific antigen (PSA) based screening increases the number of men diagnosed with early localized prostate cancer (PCa). Further, curatively intended therapies have been demonstrated to reduce PCa mortality in randomized trials. However, controversy exists, and the overall impact on PCa mortality is less clear. Men who eventually die from PCa may constitute a subgroup with either adverse histopathological characteristics and/or clinically advanced disease at diagnosis. However, the clinical characteristics at diagnosis for men who eventually die from PCa are largely unknown. We retrieved clinical characteristics of all men dying from PCa in Denmark in an 18-year period. Methods: All men who died of PCa during the period 1995 to 2013 were identified in the Danish Causes of Death Registry. Age, Gleason score (GS), tumor stage classification, and PSA were retrieved from the Danish Prostate Cancer Registry (DaPCaR). For validation, manual revision of patient charts was performed. Patients were divided into three clinical phenotypes: distant metastatic disease, locally advanced/N+ disease, and localized disease. Patients with localized disease were further grouped according to GS and PSA. Results: A total of 19,487 men died of PCa in the period 1995-2013. In total, 46.7%, 16.8% and 25.1% of men presented with distant metastatic disease, locally advanced/N+ disease or localized disease, respectively. Among men with localized disease, 85.1% had GS ≥ 7 and only 2.1% (0.5% of all men dying from PCa) only, presented with low risk (PSA < 20 and GS ≤ 6) localized disease at the time of diagnosis. Conclusions: The majority of men (63.5%) who died from PCa had either locally advanced/N+ or M+ disease at diagnosis. Among men with localized PCa at diagnosis, the majority of men subsequently dying from PCa had either PSA > 20 ng/ml and/or adverse histopathological characteristics with Gleason score ≥ 7. A total of 94.5% of patients dying from PCa had either metastatic, locally advanced/N+, and/or GS ≥ 7 disease. Patients with localized disease, PSA < 20 ng/ml and GS ≤ 6 amounted for only 0.5% of all patients dying from PCa.

171 Background: The risk of prostate cancer (PCa)-death in men diagnosed with metastatic (M+) PCa is high. During the past decade, new life-prolonging therapies have been approved for the treatment of advanced PCa. Even though demonstrated in randomized clinical trials, the impact of these advancements on mortality of men with newly diagnosed M+ PCa has not been described in a nation-wide setting. Methods: In the Danish Prostate Cancer Registry (DaPCaR), all men diagnosed with M+ PCa in Denmark from 1995 to 2011 were identified. Patients were grouped according to the year of diagnosis; 1995-2000, 2001-2005 and 2006-2011. In a competing risk setting, the 5-year cumulative incidences of PCa, other-cause, and overall death were calculated. Multivariate cause-specific Cox analysis was performed. Results: A total of 1,892 (1995-2000), 2,329 (2001-2005), and 2,653 (2006-2011) men were included (total: 6,874). Patient characteristics at diagnosis showed essential differences as median age and median PSA decreased by 1.0 year (74.1 to 73.1) and 134 ng/mL (276 to 142), respectively, in the period studied. The 5-year PCa-specific mortality decreased by 17.0% from 72.8% (1995-2000) (95%CI: 70.8% – 74.8%) to 55.8% (2006-2011) (95%CI: 53.9% – 57.7%), p < 0.0001. The 5-year other-cause mortality increased by 5.7% from 11.4% (95%CI: 9.9% – 12.8%) to 17.1% (95%CI: 15.6 – 18.6), p < 0.0001. The risk of PCa-death decreased for patients diagnosed in 2000-2005; HR: 0.69 (95%CI 0.61-0.79) and for patients diagnosed in 2006-2011; HR: 0.53 (95%CI 0.47-0.61) compared to patients diagnosed in 1995-2000, when adjusting for age, PSA, and Gleason score (GS) in the statistical analysis. Conclusions: A significant reduction in 5-year PCa-specific mortality was observed in a nationwide cohort of patients diagnosed with M+ PCa since 1995. Changes in age and PSA at diagnosis suggest that lead-time introduced by increased PSA use may have affected the results. However, in multivariate analysis, a significant reduction in hazard of almost 50% was observed when adjusting for age, PSA, and GS. Only minor changes in other cause mortality were found, which suggests that the improvement to a large extend can be credited to improved management of men with advanced PCa.

475 Background: TOP1 inhibitor treatment is frequently being used in combination therapy of metastatic CRC. This study aims to reveal whether TOP1 gene copy number associates with patient prognosis, since such a relationship may have significant implications for future studies aiming at validating the predictive value of TOP1 gene copy number. Methods: The study included TOP1 and CEN-20 FISH analyses (DAKO A/S Denmark) on FFPE tissue sections from 154 stage III CRC patients who did not receive adjuvant chemotherapy. TOP1 gene copy number, CEN-20 copy number and the TOP1/CEN-20 ratios were analyzed and correlated to overall survival (OS), to time to recurrence (TTR) of patients with CRC and to local recurrence (LR) in patients with rectal cancer (RC). Results: TOP1 copy number counts and the TOP1/CEN-20 ratios, age, gender and primary tumor location were separately added into a multivariate analysis as continuous variables. For OS and LR, TOP1 copy number was significant and the ratio was borderline significant with higher copy number associated with longer OS or longer time to LR. When the patients were dichotomized using the TOP1 median copy number, we found that patients with high TOP1 copy number in their tumor cells had a significant longer OS (HR: 0.68; 95% CI: 0.47-0.98; p = 0.04) compared to patients with low TOP1 copy number. Using the median TOP1/CEN-20 ratio to dichotomize, no significant differences were observed between patients with levels above or below the median ratio number for OS (HR: 0.76; 95% CI: 0.53-1.10; p = 0.14). TOP1 copy number divided the RC patients into two groups with a trend towards a significant difference in time to LR (HR: 0.56; 95% CI: 0.27-1.16; p = 0.11) with higher copy number. If the median ratio was used, a significant association with longer time to LR (HR: 0.43; 95% CI: 0.20-0.92; p = 0.03) was found. No significant associations between TOP1 copy number or ratio and TTR were observed. Conclusions: Increased TOP1 copy number is associated with longer OS in CRC patients and fewer LR in RC patients. Thus, future studies analyzing the association between TOP1 copy number and response to therapy in CRC patients should be planned in such a way that a prognostic and a predictive value of TOP1 can be separated.

3629 Background: Cytotoxic effect of 5-FU is mediated through competitive binding to thymidylate synthetase (TS), leading to thymine depletion. 5-FU is catabolized by dihydropyrimidine dehydrogenas...