Purpose/Objective(s): We sought to assess the utility of docetaxel administered concurrently with salvage radiation therapy (SRT) following postprostatectomy biochemical failure (BF). Methods and materials: Men with postprostatectomy BF were accrued on a single-arm phase 2 clinical trial. SRT doses ranged from 64.8 to 70.2 Gy and were delivered in 1.8-Gy fractions to the prostate bed alone as the clinical target volume with a +1-cm uniform planning target volume expansion. The primary endpoint was progression-free survival at 4 years compared with the Stephenson nomogram estimate. Kaplan-Meier methods were used to assess late toxicity, BF, and distant metastases. An unplanned matched-pair analysis was performed with 19 patients treated with SRT alone. Results: Nineteen men were accrued and treated. Median follow-up was 4.8 years. Median pre-RT prostate-specific antigen level was 0.7 ng/mL (interquartile range, 0.4-1.3 ng/mL). All 8 cycles of docetaxel were completed in 17 (89%) patients. Acute grade 1-4 toxicities were observed in 79%, 50%, 58%, and 11%, respectively. A total of 68% of acute grade 1 toxicities were related to fatigue, urinary, or bowel symptoms. For grade 2 toxicities, 76% were related to neutropenia, fatigue, or urinary symptoms. Acute grade 3 and 4 toxicities were most commonly neutropenia (84% and 100%, respectively). All late toxicities were grade 1 to 2 with 89% related to bowel or urinary function. Predicted 4-year progression-free survival was 39% and observed was 42% (90% confidence interval [CI], 24-60). Matched-pair analysis demonstrated no significant improvement in BF (P = .96, hazard ratio, 0.98; 90% CI, 0.4-2.3) or distant metastases (P = .09; hazard ratio, 0.3; 90% CI, 0.07-1.2), and no difference between late bowel (P = .60) or urinary toxicity (P = .41). Conclusions: Docetaxel can safely be administered concurrently with SRT without significantly impacting posttreatment toxicity. Neutropenia was the most significant acute toxicity. Given the small sample size, no clear clinical benefit was observed. Larger studies are needed to determine the efficacy of concurrent docetaxel in this setting.

Purpose/Objective(s)Low-risk PCa patients have excellent outcomes, with treatment selection often determined by patient preference and perceived impact of treatment modality on QOL. Acute urinary symptoms are common during external beam therapy, while chronic symptoms have been linked to urethral dose. Since most low risk PCa occurs in the peripheral zone, we hypothesized that US-IMRT could improve urinary toxicity while maintaining biochemical control.Materials/MethodsLow-risk PCa patients (Gleason < or = 6, PSA < or = 10, and T1c) with no visible lesion within 5 mm of the prostatic urethra on MRI were randomized to US-IMRT or standard (S) IMRT. The PTV was the MRI-defined peripheral zone + 3-5 mm for US-IMRT and the prostate + 3 mm for S-IMRT. The PTV was prescribed 75.6 Gy in 41 fractions (max < or = 115%; min > or = 93%). The prostatic urethral planning organ-at-risk volume (U-PRV) was the urethra + 3 mm, limited to 65 Gy to the proximal and 74 Gy to the distal U-PRV. Fiducial markers were used for image guidance. The expanded prostate cancer index (EPIC) QOL questionnaire was completed at baseline, during treatment, and at each follow-up. The primary endpoint was the change in EPIC Urinary Function Score at 3 months. To demonstrate a 9 point improvement (1/2 SD) in urinary QOL at 3 months (80% power using a two-sided t-test at a 0.05 significance level) would require 63 pts per arm. Secondary endpoints included PSA failure (nadir + 2) (BF), time to progression, and overall survival.ResultsFrom 6/04-11/06, 16 evaluable pts were enrolled with median follow-up of 4.7 years. An unplanned futility analysis concluded that continued accrual was unlikely to demonstrate a difference in the primary endpoint. There was no difference in baseline EPIC urinary QOL (85.4 and 86.8) in the control and experimental arms, respectively. The mean change in EPIC urinary QOL at 3 months was +0.7 in the S-IMRT arm and -1.4 in the US-IMRT arm. There were no differences between arms in the remaining 4 EPIC domains. Median PSA nadir was higher in the US-IMRT arm (0.7 vs. 1.3, p = 0.05). Three pts with US-IMRT and 0 pts with S-IMRT experienced BF, with 2 year BF 25% in the US-IMRT arm and 0% in the S-IMRT arm (p = 0.08, log-rank; HR 7.8, 95% CI 0.8 - 75). All 3 failures underwent biopsy (2.2, 3.1, and 4.4 years since end of RT) and 2 were positive (one with and one without treatment effect). Both biopsy proven recurrences were contralateral to the original site of PCa. No distant failures or deaths occurred.ConclusionsUS-IMRT failed to improve urinary QOL, but resulted in a higher PSA nadir and inferior biochemical control. The contralateral local failures, despite low-risk features and MRI screening, should serve as a cautionary tale for focal PCa treatments. Purpose/Objective(s)Low-risk PCa patients have excellent outcomes, with treatment selection often determined by patient preference and perceived impact of treatment modality on QOL. Acute urinary symptoms are common during external beam therapy, while chronic symptoms have been linked to urethral dose. Since most low risk PCa occurs in the peripheral zone, we hypothesized that US-IMRT could improve urinary toxicity while maintaining biochemical control. Low-risk PCa patients have excellent outcomes, with treatment selection often determined by patient preference and perceived impact of treatment modality on QOL. Acute urinary symptoms are common during external beam therapy, while chronic symptoms have been linked to urethral dose. Since most low risk PCa occurs in the peripheral zone, we hypothesized that US-IMRT could improve urinary toxicity while maintaining biochemical control. Materials/MethodsLow-risk PCa patients (Gleason < or = 6, PSA < or = 10, and T1c) with no visible lesion within 5 mm of the prostatic urethra on MRI were randomized to US-IMRT or standard (S) IMRT. The PTV was the MRI-defined peripheral zone + 3-5 mm for US-IMRT and the prostate + 3 mm for S-IMRT. The PTV was prescribed 75.6 Gy in 41 fractions (max < or = 115%; min > or = 93%). The prostatic urethral planning organ-at-risk volume (U-PRV) was the urethra + 3 mm, limited to 65 Gy to the proximal and 74 Gy to the distal U-PRV. Fiducial markers were used for image guidance. The expanded prostate cancer index (EPIC) QOL questionnaire was completed at baseline, during treatment, and at each follow-up. The primary endpoint was the change in EPIC Urinary Function Score at 3 months. To demonstrate a 9 point improvement (1/2 SD) in urinary QOL at 3 months (80% power using a two-sided t-test at a 0.05 significance level) would require 63 pts per arm. Secondary endpoints included PSA failure (nadir + 2) (BF), time to progression, and overall survival. Low-risk PCa patients (Gleason < or = 6, PSA < or = 10, and T1c) with no visible lesion within 5 mm of the prostatic urethra on MRI were randomized to US-IMRT or standard (S) IMRT. The PTV was the MRI-defined peripheral zone + 3-5 mm for US-IMRT and the prostate + 3 mm for S-IMRT. The PTV was prescribed 75.6 Gy in 41 fractions (max < or = 115%; min > or = 93%). The prostatic urethral planning organ-at-risk volume (U-PRV) was the urethra + 3 mm, limited to 65 Gy to the proximal and 74 Gy to the distal U-PRV. Fiducial markers were used for image guidance. The expanded prostate cancer index (EPIC) QOL questionnaire was completed at baseline, during treatment, and at each follow-up. The primary endpoint was the change in EPIC Urinary Function Score at 3 months. To demonstrate a 9 point improvement (1/2 SD) in urinary QOL at 3 months (80% power using a two-sided t-test at a 0.05 significance level) would require 63 pts per arm. Secondary endpoints included PSA failure (nadir + 2) (BF), time to progression, and overall survival. ResultsFrom 6/04-11/06, 16 evaluable pts were enrolled with median follow-up of 4.7 years. An unplanned futility analysis concluded that continued accrual was unlikely to demonstrate a difference in the primary endpoint. There was no difference in baseline EPIC urinary QOL (85.4 and 86.8) in the control and experimental arms, respectively. The mean change in EPIC urinary QOL at 3 months was +0.7 in the S-IMRT arm and -1.4 in the US-IMRT arm. There were no differences between arms in the remaining 4 EPIC domains. Median PSA nadir was higher in the US-IMRT arm (0.7 vs. 1.3, p = 0.05). Three pts with US-IMRT and 0 pts with S-IMRT experienced BF, with 2 year BF 25% in the US-IMRT arm and 0% in the S-IMRT arm (p = 0.08, log-rank; HR 7.8, 95% CI 0.8 - 75). All 3 failures underwent biopsy (2.2, 3.1, and 4.4 years since end of RT) and 2 were positive (one with and one without treatment effect). Both biopsy proven recurrences were contralateral to the original site of PCa. No distant failures or deaths occurred. From 6/04-11/06, 16 evaluable pts were enrolled with median follow-up of 4.7 years. An unplanned futility analysis concluded that continued accrual was unlikely to demonstrate a difference in the primary endpoint. There was no difference in baseline EPIC urinary QOL (85.4 and 86.8) in the control and experimental arms, respectively. The mean change in EPIC urinary QOL at 3 months was +0.7 in the S-IMRT arm and -1.4 in the US-IMRT arm. There were no differences between arms in the remaining 4 EPIC domains. Median PSA nadir was higher in the US-IMRT arm (0.7 vs. 1.3, p = 0.05). Three pts with US-IMRT and 0 pts with S-IMRT experienced BF, with 2 year BF 25% in the US-IMRT arm and 0% in the S-IMRT arm (p = 0.08, log-rank; HR 7.8, 95% CI 0.8 - 75). All 3 failures underwent biopsy (2.2, 3.1, and 4.4 years since end of RT) and 2 were positive (one with and one without treatment effect). Both biopsy proven recurrences were contralateral to the original site of PCa. No distant failures or deaths occurred. ConclusionsUS-IMRT failed to improve urinary QOL, but resulted in a higher PSA nadir and inferior biochemical control. The contralateral local failures, despite low-risk features and MRI screening, should serve as a cautionary tale for focal PCa treatments. US-IMRT failed to improve urinary QOL, but resulted in a higher PSA nadir and inferior biochemical control. The contralateral local failures, despite low-risk features and MRI screening, should serve as a cautionary tale for focal PCa treatments.

4585 Background: Prostate cancer (PCa) recurrence can be identified by biochemcial failure (BF) many years prior to clinical recurrence. A short IBF (<18 months) has been proposed as a surrogate for cause specific survival (CSS). Since dose-escalated RT influences both the rate and timing of BF we sought to validate the association between short IBF and both CSS and overall survival (OS).METHODSFrom 1998-2008, 718 patients (median age 69.3 (IQR: 63-74) were treated with definitive external beam radiation therapy (RT) with or without androgen deprivation therapy (ADT) at the University of Michigan (minimum dose 75 Gy, median follow-up 64 months). BF was defined by the Phoenix definition (nadir +2 ng/ml). Log-rank and Cox-proportional hazard regressions were used to assess the association of BF, the IBF, and other clinical factors with CSS and OS.RESULTSThere were 149 BFs (21%), and no low-risk patients had a short IBF while for intermediate- and high-risk disease it accounted for 14% and 40% of all BFs, respectively. BF had no significant impact on OS (p=0.36) even in patients with NCCN high-risk disease (p=0.8). However, in those with recurrence a short IBF predicted decreased CSS (p<0.0001 HR: 5.6 [2.4-13.0]) and OS (p<0.0001, HR: 4.8 [2.3-10.3]). Further, the 10-year rate of OS was 74% in those without BF, compared to 77% in those with a long IBF (p=0.1, HR: 0.7 [0.4-1.1]) and only 33% in those with a short IBF (p<0.0001, HR: 3.7 [2.3-5.9]). On multivariate analysis, after accounting for age, co-morbid illness, PSA, T-stage, Gleason score, and the use of ADT, a short IBF increased the risk of prostate cancer death (p<0.0001, HR: 18.1 [8.4-39]) and all cause mortality (p=0.0027, HR: 1.5 [1.2-2.1]) while late BF did not.CONCLUSIONSWe validated the relationship between the IBF and CSS in patients treated with dose-escalated RT. The IBF is associated with OS while BF by itself is not. Prompt salvage therapy should be considered for all patients with a short IBF while those with a long IBF might be appropriately managed with a more conservative approach. Further evaluation of the IBF as a surrogate end-point for clinical trials is warranted.

15 Background: The division of Gleason score (GS) into three categories (2-6, 7, 8-10), may not fully utilize its prognostic power as shown by recent reports demonstrating that the presence of Gleason Pattern 5 (GP5) is a strong adverse prognostic factor. Therefore, we analyzed clinical outcomes for patients treated with dose-escalated radiation therapy (RT) based upon the presence or absence of GP5 within the biopsy specimens. Methods: Clinical outcomes were analyzed for 718 men treated for localized prostate cancer with definitive external beam RT to at least 75 Gy. We assessed the impact of GP5 as well as pre-treatment and treatment related factors on freedom from biochemical failure (FFBF), freedom from metastasis (FFM), cause-specific survival (CSS), and overall survival (OS). Results: Median follow-up was 64 months. At biopsy, 89% of patients had no GP5 while 11% had GP5. There was no difference in age, co-morbid illness, clinical T-stage, PSA, or the use or duration of androgen deprivation therapy (ADT) between GS8 without GP5 and GS8-10 with GP5. The presence of GP5 predicted lower FFM (p<0.002 [HR: 3.4{1.7-7.1}]), CSS (p<0.0001 [HR: 12.9 {5.4-31}]), and OS (p<0.0001 [HR: 3.6 {2.0-6.5}]) when compared to GS8 (without GP5). Ten-year FFM, CSS, and OS were 89%, 98%, and 57% for those with Gleason 8 prostate cancer without GP5 as compared to 61%, 55%, and 31% for those with GP5. In addition, both FFM and CSS were strongly influenced by ADT use concurrent with RT. On multivariate analysis GP5 was the strongest prognostic factor for all clinical end-points. Conclusions: The presence of GP5 predicts for worse clinical behavior, which therefore needs to be accounted for by risk stratification schemes. Further intensification of local and/or systemic therapy may be appropriate for such patients. No significant financial relationships to disclose.

Purpose/Objective(s)In randomized trials, dose escalated radiation therapy (RT) improved prostate cancer (PCa) biochemical control without improvements in metastasis free (MF) or cancer specific survival (CSS). Additional randomized trials demonstrated improved MF and CSS for advanced PCa treated with conventional dose RT with androgen deprivation therapy (ADT). However, the role for and optimal duration of ADT when delivering dose escalated RT is unknown.Materials/MethodsPretreatment and treatment related variables as well as outcome were evaluated using a prospectively maintained IRB approved data base limited to patients treated to ≥75 Gy. Based upon previously published studies and prevailing clinical practice ADT use was analyzed by duration: none, short term ADT, ((STAD),<2 years), and long term ADT ((LTAD), ≥ 2 years). Differences between treatment groups were evaluated with t-test and chi-squared tests while survival analysis was performed using log-rank and Cox proportional hazards analysis.ResultsFrom 1988-2008 a total of 715 men were treated with dose escalated RT (77 Gy (IQR: 75.8-77.8)) with median follow-up 64 months. ADT was utilized in 286 (40%) of patients with STAD in 29% (6.1 months (IQR: 5.0-7.4)) and LTAD in 11% (27.5 months (IQR: 25.0-30.6)). There were no differences in age, co-morbid illness, or RT techniques based upon ADT duration. Mean PSA was 10.5, 16.8, and 27.7 ng/mL in those with no ADT, STAD, and LTAD, respectively (p < 0.001). Similarly, cT3-T4 disease was present in 7%, 13%, and 32% (p < 0.001); while Gleason ≥ 8 disease was present in 4%, 28%, and 83% of cases based on ADT use (p < 0.0001). This gave NCCN high risk disease in 12%, 50%, and 98% of patients based upon ADT use (p < 0.0001). At 10-years NCCN low, intermediate, and high risk groups had MF of 98%, 84%, and 75% (p < 0.0001) and CSS of 99%, 94%, and 81% (p < 0.0001), respectively. Given the close association between risk groups and ADT use on univariate analysis, neither ADT use nor duration were prognostic for MF or CSS. Multivariate analysis revealed lower MF based upon: increasing PSA (p < 0.003), advanced T-stage (p < 0.05), and higher Gleason score (p < 0.0001). While both STAD (HR: 0.52 (95% CI: 0.26-0.79)) and LTAD (HR: 0.25 (95% CI: 0.11 - 0.56)) predicted for improved MF (p < 0.0001). Finally, after accounting for clinical and treatment related factors only Gleason score (p < 0.0001) and LTAD (HR: 0.35 (95% CI: 0.12 - 1.03), p < 0.06) were prognostic for CSS.ConclusionsThese results using dose escalated RT confirm the conclusions from previous trials which identified improvement in MF and CSS with LTAD as compared to STAD for high risk or locally advanced PCa treated with conventional dose RT. Purpose/Objective(s)In randomized trials, dose escalated radiation therapy (RT) improved prostate cancer (PCa) biochemical control without improvements in metastasis free (MF) or cancer specific survival (CSS). Additional randomized trials demonstrated improved MF and CSS for advanced PCa treated with conventional dose RT with androgen deprivation therapy (ADT). However, the role for and optimal duration of ADT when delivering dose escalated RT is unknown. In randomized trials, dose escalated radiation therapy (RT) improved prostate cancer (PCa) biochemical control without improvements in metastasis free (MF) or cancer specific survival (CSS). Additional randomized trials demonstrated improved MF and CSS for advanced PCa treated with conventional dose RT with androgen deprivation therapy (ADT). However, the role for and optimal duration of ADT when delivering dose escalated RT is unknown. Materials/MethodsPretreatment and treatment related variables as well as outcome were evaluated using a prospectively maintained IRB approved data base limited to patients treated to ≥75 Gy. Based upon previously published studies and prevailing clinical practice ADT use was analyzed by duration: none, short term ADT, ((STAD),<2 years), and long term ADT ((LTAD), ≥ 2 years). Differences between treatment groups were evaluated with t-test and chi-squared tests while survival analysis was performed using log-rank and Cox proportional hazards analysis. Pretreatment and treatment related variables as well as outcome were evaluated using a prospectively maintained IRB approved data base limited to patients treated to ≥75 Gy. Based upon previously published studies and prevailing clinical practice ADT use was analyzed by duration: none, short term ADT, ((STAD),<2 years), and long term ADT ((LTAD), ≥ 2 years). Differences between treatment groups were evaluated with t-test and chi-squared tests while survival analysis was performed using log-rank and Cox proportional hazards analysis. ResultsFrom 1988-2008 a total of 715 men were treated with dose escalated RT (77 Gy (IQR: 75.8-77.8)) with median follow-up 64 months. ADT was utilized in 286 (40%) of patients with STAD in 29% (6.1 months (IQR: 5.0-7.4)) and LTAD in 11% (27.5 months (IQR: 25.0-30.6)). There were no differences in age, co-morbid illness, or RT techniques based upon ADT duration. Mean PSA was 10.5, 16.8, and 27.7 ng/mL in those with no ADT, STAD, and LTAD, respectively (p < 0.001). Similarly, cT3-T4 disease was present in 7%, 13%, and 32% (p < 0.001); while Gleason ≥ 8 disease was present in 4%, 28%, and 83% of cases based on ADT use (p < 0.0001). This gave NCCN high risk disease in 12%, 50%, and 98% of patients based upon ADT use (p < 0.0001). At 10-years NCCN low, intermediate, and high risk groups had MF of 98%, 84%, and 75% (p < 0.0001) and CSS of 99%, 94%, and 81% (p < 0.0001), respectively. Given the close association between risk groups and ADT use on univariate analysis, neither ADT use nor duration were prognostic for MF or CSS. Multivariate analysis revealed lower MF based upon: increasing PSA (p < 0.003), advanced T-stage (p < 0.05), and higher Gleason score (p < 0.0001). While both STAD (HR: 0.52 (95% CI: 0.26-0.79)) and LTAD (HR: 0.25 (95% CI: 0.11 - 0.56)) predicted for improved MF (p < 0.0001). Finally, after accounting for clinical and treatment related factors only Gleason score (p < 0.0001) and LTAD (HR: 0.35 (95% CI: 0.12 - 1.03), p < 0.06) were prognostic for CSS. From 1988-2008 a total of 715 men were treated with dose escalated RT (77 Gy (IQR: 75.8-77.8)) with median follow-up 64 months. ADT was utilized in 286 (40%) of patients with STAD in 29% (6.1 months (IQR: 5.0-7.4)) and LTAD in 11% (27.5 months (IQR: 25.0-30.6)). There were no differences in age, co-morbid illness, or RT techniques based upon ADT duration. Mean PSA was 10.5, 16.8, and 27.7 ng/mL in those with no ADT, STAD, and LTAD, respectively (p < 0.001). Similarly, cT3-T4 disease was present in 7%, 13%, and 32% (p < 0.001); while Gleason ≥ 8 disease was present in 4%, 28%, and 83% of cases based on ADT use (p < 0.0001). This gave NCCN high risk disease in 12%, 50%, and 98% of patients based upon ADT use (p < 0.0001). At 10-years NCCN low, intermediate, and high risk groups had MF of 98%, 84%, and 75% (p < 0.0001) and CSS of 99%, 94%, and 81% (p < 0.0001), respectively. Given the close association between risk groups and ADT use on univariate analysis, neither ADT use nor duration were prognostic for MF or CSS. Multivariate analysis revealed lower MF based upon: increasing PSA (p < 0.003), advanced T-stage (p < 0.05), and higher Gleason score (p < 0.0001). While both STAD (HR: 0.52 (95% CI: 0.26-0.79)) and LTAD (HR: 0.25 (95% CI: 0.11 - 0.56)) predicted for improved MF (p < 0.0001). Finally, after accounting for clinical and treatment related factors only Gleason score (p < 0.0001) and LTAD (HR: 0.35 (95% CI: 0.12 - 1.03), p < 0.06) were prognostic for CSS. ConclusionsThese results using dose escalated RT confirm the conclusions from previous trials which identified improvement in MF and CSS with LTAD as compared to STAD for high risk or locally advanced PCa treated with conventional dose RT. These results using dose escalated RT confirm the conclusions from previous trials which identified improvement in MF and CSS with LTAD as compared to STAD for high risk or locally advanced PCa treated with conventional dose RT.

e15033 Background: The treatment of biochemically recurrent prostate cancer continues to be controversial. While androgen deprivation therapy with an LHRH agonist is the current standard of care, it is associated with a growing list of potential toxicities. Sequential androgen blockade (SAB), with a 5-alpha reductase inhibitor and a nonsteroidal antiandrogen, has been shown to decrease PSA and delay disease progression in patients with biochemical recurrence. Methods: We performed a retrospective review of 61 prostate cancer patients treated at a single institution with a combination of oral finasteride 5mg and bicalutamide 50 mg. 22 patients were treated with continuous SAB and 39 were treated intermittently based on PSA. 55 patients had a rising PSA after definitive treatment; 6 patients had no local definitive treatment. In addition to local treatments, over 50% had received other treatments for their prostate cancer: 24 had received LHRH agonists prior to SAB (duration 4-54 months), 13 had prior chemotherapy and 4 had been on vaccine trials. 14 patients had evidence of metastatic disease prior to initiation of SAB (4 with lympadenopathy, 7 with bone metastases; 2 with nodal and bone disease; 1 with lung metastases). Kaplan-Meier methods were used to determine duration of response. Results: Excluding 5 patients with undetectable PSAs at the time of SAB initiation, median PSA decline at 3 months was 90%. With a median follow up of 44 months (range 3-149 months), median duration of response to SAB was 72.5 months (95% CI 38, 87). Response was noted across all Gleason scores with median duration in Gleason 7, 8 and 9 disease of 72.5, 72.5 and 82.3 months respectively. Currently 35 patients remain on SAB with median follow up of 29 months. 21 patients treated with SAB were later treated with LHRH analogs; median response duration to subsequent LHRH in this group was 24 months (95% CI 6, 47). Treatment was well tolerated. Conclusions: In this heterogeneous group of prostate cancer patients, SAB was well tolerated with a median duration of biochemical control of just over 6 years. Our observations suggest that further study of this combination in randomized trials may be warranted. No significant financial relationships to disclose.

Purpose/Objective(s)Numerous models have been derived to predict outcomes for prostate cancer. The CAPRA score was developed to be more accurate than D'Amico classification yet easier to use than paper or computer-based nomograms. It has been validated for biochemical failure (BF) in multiple data sets following surgical treatment, as well as in a multi-institutional set for metastasis (METS), cause specific (CSS), and overall survival (OS). We evaluated the CAPRA in a cohort of patients treated with dose-escalated RT.Materials/MethodsPretreatment and treatment-related variables were evaluated using a prospectively maintained IRB-approved database limited to patients treated with dose-escalated RT (median 77 Gy (IQR: 75.8-77.8)). The Phoenix definition was utilized for BF. The CAPRA score was determined in 668 men, yielding a 0-9 score based upon age, PSA, Gleason score, and T-stage. Differences between treatment groups were evaluated with t-test and chi-squared tests, while survival analysis was performed using log-rank and Cox proportional hazards analysis. The discriminatory power of the CAPRA score was evaluated using the c-index.ResultsMedian follow-up was 63 months. The median CAPRA score was 3 (IQR: 2 - 5) with 20% of men with a CAPRA score of 6-9, which was higher compared to the CAPRA derivation set (median: 2 (IQR: 1-3)) containing only 5% of men with a CAPRA score of 6-9 (p < 0.0001). All elements composing the CAPRA score except for age were associated with BF (p < 0.0001). Freedom from BF at 5-years varied from 96% (95% CI 94-98) with a CAPRA of 0-1 to 34% (95% CI:19-49) with a CAPRA score of 9 (p < 0.0001), with the risk of BF lower compared to the derivation set across all CAPRA scores (p < 0.0001) Each CAPRA point increased the risk of: BF (HR: 1.42 (1.31-1.53), p < 0.0001), METS (HR: 1.41 (1.26-1.59), p < 0.0001), CSS (HR: 1.53 (1.29-1.81), p < 0.0001), and OS (HR: 1.18 (1.08-1.28), p < 0.002). Overall, the CAPRA was highly discriminating for each of these end-points with c-indices of: BF 0.70 (0.54-0.85), METS 0.66 (0.42-0.87), CSS 0.68 (0.33-0.96), and OS 0.83 (0.68-0.95). Overall 40% of men were treated with ADT, which correlated with increasing CAPRA score (14% for CAPRA 0-1 and 92% for CAPRA 9, p < 0.0001). When risk of METS was evaluated, CAPRA score increased risk of METS (HR: 1.52 (95% CI: 1.33-1.73), p < 0.0001) while ADT use decreased risk (HR: 0.52 (95% CI: 0.29-0.94), p < 0.03).ConclusionsThe CAPRA score is an easily used and robust prognostic tool. Given the increasing risk across all CAPRA scores when comparing data sets care must be taken to evaluate the full range of CAPRA scores lest bias is introduced against those with the highest scores. Finally, a limitation of the CAPRA score is the absence of treatment-related variables as observed by the impact of ADT on METS for patients treated with dose-escalated RT. Purpose/Objective(s)Numerous models have been derived to predict outcomes for prostate cancer. The CAPRA score was developed to be more accurate than D'Amico classification yet easier to use than paper or computer-based nomograms. It has been validated for biochemical failure (BF) in multiple data sets following surgical treatment, as well as in a multi-institutional set for metastasis (METS), cause specific (CSS), and overall survival (OS). We evaluated the CAPRA in a cohort of patients treated with dose-escalated RT. Numerous models have been derived to predict outcomes for prostate cancer. The CAPRA score was developed to be more accurate than D'Amico classification yet easier to use than paper or computer-based nomograms. It has been validated for biochemical failure (BF) in multiple data sets following surgical treatment, as well as in a multi-institutional set for metastasis (METS), cause specific (CSS), and overall survival (OS). We evaluated the CAPRA in a cohort of patients treated with dose-escalated RT. Materials/MethodsPretreatment and treatment-related variables were evaluated using a prospectively maintained IRB-approved database limited to patients treated with dose-escalated RT (median 77 Gy (IQR: 75.8-77.8)). The Phoenix definition was utilized for BF. The CAPRA score was determined in 668 men, yielding a 0-9 score based upon age, PSA, Gleason score, and T-stage. Differences between treatment groups were evaluated with t-test and chi-squared tests, while survival analysis was performed using log-rank and Cox proportional hazards analysis. The discriminatory power of the CAPRA score was evaluated using the c-index. Pretreatment and treatment-related variables were evaluated using a prospectively maintained IRB-approved database limited to patients treated with dose-escalated RT (median 77 Gy (IQR: 75.8-77.8)). The Phoenix definition was utilized for BF. The CAPRA score was determined in 668 men, yielding a 0-9 score based upon age, PSA, Gleason score, and T-stage. Differences between treatment groups were evaluated with t-test and chi-squared tests, while survival analysis was performed using log-rank and Cox proportional hazards analysis. The discriminatory power of the CAPRA score was evaluated using the c-index. ResultsMedian follow-up was 63 months. The median CAPRA score was 3 (IQR: 2 - 5) with 20% of men with a CAPRA score of 6-9, which was higher compared to the CAPRA derivation set (median: 2 (IQR: 1-3)) containing only 5% of men with a CAPRA score of 6-9 (p < 0.0001). All elements composing the CAPRA score except for age were associated with BF (p < 0.0001). Freedom from BF at 5-years varied from 96% (95% CI 94-98) with a CAPRA of 0-1 to 34% (95% CI:19-49) with a CAPRA score of 9 (p < 0.0001), with the risk of BF lower compared to the derivation set across all CAPRA scores (p < 0.0001) Each CAPRA point increased the risk of: BF (HR: 1.42 (1.31-1.53), p < 0.0001), METS (HR: 1.41 (1.26-1.59), p < 0.0001), CSS (HR: 1.53 (1.29-1.81), p < 0.0001), and OS (HR: 1.18 (1.08-1.28), p < 0.002). Overall, the CAPRA was highly discriminating for each of these end-points with c-indices of: BF 0.70 (0.54-0.85), METS 0.66 (0.42-0.87), CSS 0.68 (0.33-0.96), and OS 0.83 (0.68-0.95). Overall 40% of men were treated with ADT, which correlated with increasing CAPRA score (14% for CAPRA 0-1 and 92% for CAPRA 9, p < 0.0001). When risk of METS was evaluated, CAPRA score increased risk of METS (HR: 1.52 (95% CI: 1.33-1.73), p < 0.0001) while ADT use decreased risk (HR: 0.52 (95% CI: 0.29-0.94), p < 0.03). Median follow-up was 63 months. The median CAPRA score was 3 (IQR: 2 - 5) with 20% of men with a CAPRA score of 6-9, which was higher compared to the CAPRA derivation set (median: 2 (IQR: 1-3)) containing only 5% of men with a CAPRA score of 6-9 (p < 0.0001). All elements composing the CAPRA score except for age were associated with BF (p < 0.0001). Freedom from BF at 5-years varied from 96% (95% CI 94-98) with a CAPRA of 0-1 to 34% (95% CI:19-49) with a CAPRA score of 9 (p < 0.0001), with the risk of BF lower compared to the derivation set across all CAPRA scores (p < 0.0001) Each CAPRA point increased the risk of: BF (HR: 1.42 (1.31-1.53), p < 0.0001), METS (HR: 1.41 (1.26-1.59), p < 0.0001), CSS (HR: 1.53 (1.29-1.81), p < 0.0001), and OS (HR: 1.18 (1.08-1.28), p < 0.002). Overall, the CAPRA was highly discriminating for each of these end-points with c-indices of: BF 0.70 (0.54-0.85), METS 0.66 (0.42-0.87), CSS 0.68 (0.33-0.96), and OS 0.83 (0.68-0.95). Overall 40% of men were treated with ADT, which correlated with increasing CAPRA score (14% for CAPRA 0-1 and 92% for CAPRA 9, p < 0.0001). When risk of METS was evaluated, CAPRA score increased risk of METS (HR: 1.52 (95% CI: 1.33-1.73), p < 0.0001) while ADT use decreased risk (HR: 0.52 (95% CI: 0.29-0.94), p < 0.03). ConclusionsThe CAPRA score is an easily used and robust prognostic tool. Given the increasing risk across all CAPRA scores when comparing data sets care must be taken to evaluate the full range of CAPRA scores lest bias is introduced against those with the highest scores. Finally, a limitation of the CAPRA score is the absence of treatment-related variables as observed by the impact of ADT on METS for patients treated with dose-escalated RT. The CAPRA score is an easily used and robust prognostic tool. Given the increasing risk across all CAPRA scores when comparing data sets care must be taken to evaluate the full range of CAPRA scores lest bias is introduced against those with the highest scores. Finally, a limitation of the CAPRA score is the absence of treatment-related variables as observed by the impact of ADT on METS for patients treated with dose-escalated RT.