1087 Clinical Experience with the Calypso® 4D Localization System in Prostate Cancer Patients: Implantation, Tolerance, Migration, Localization and Real Time Tracking P. Kupelian, 1 T. Willoughby, 1 D. Litzenberg, 2 H. Sandler, 2 M. Roach, 3 L. Levine, 4 M. van Waardenburg, 4 A. Cunningham, 4 S. Meeks 1 1 Radiation Oncology, M. D. Anderson Cancer Center Orlando, Orlando, FL, 2 Radiation Oncology, University of Michigan, Ann Arbor, MI, 3 Radiation Oncology, University of California San Francisco, San Francisco, CA, 4 Calypso Medical Technologies Inc., Seattle, WA Purpose/Objective: To study the implantation and long-term tolerance of Beacon® transponders, and localization and tracking in patients with localized prostate cancer. Materials/Methods: Calypso 4D Localization System (“Calypso System”, Calypso Medical, Seattle, WA) is an investigational patient positioning device including markers (transponders: 8.5  1.85 mm, glass-encapsulated electrical circuits used for permanent implantation), and external electronics that continuously localize the transponders using electromagnetic signals at a 10 Hz rate. Three transponders were implanted with 14G needles at the apex, left and right base under transrectal ultrasound guidance in 20 patients. Adverse event checks were conducted at Day 0 (day of implant), 4, 7, 10, 14 (simulation); Fraction 4, 10, 15, 20; and 1, 6, and 12 months post- RT. CTs or pair of kV X-rays were acquired at Day 0, 4, 7, 10, 14; at Fractions 1, 2, 3, 4, 10, 15, 20; and 1 month post-RT to document transponder geometry and compute inter-transponder distance, a migration indicator. Results: The implantation procedures completed without any major complications. Of 60 transponders, one was presumably implanted in the urethra and subsequently voided (patient was asymptomatic) and one migrated cephalad approx. 3 cm from the apex prior to Day 4 presumably within the venous plexus (the transponder subsequently remained stable). In 2 cases, there was a minimal change in urinary status (IPSS scores) prior to RT. All acute bleeding (hematuria, hematospermia, and/or hematochezia) resolved by Day 10 (n=7). To date, 5/20 subjects have completed the 12-month post-RT follow-up. Dysuria was reported in 11 patients after RT. One case of prostatitis was reported at 6 months, and one patient with a very large prostate gland had urinary obstruction requiring surgical intervention at 8 months post-RT that was considered unrelated to the transponders. Analysis of inter-transponder distances indicated stable geometry in 15/20 patients by Day 4, and in all patients by Day 14. Distances at 4-week post-RT again showed stable geometry. In 11/20 patients, localization and tracking were performed (1 patient not localizable and 8 not evaluated). Comparison of the Calypso System coordinates with transponder locations on X-rays showed average (+SD) 3D difference between the two systems was 1.5  0.9 mm. Finally, during tracking, 2/11 patients showed significant organ motion (1 cm) over 8 minutes. Longitudinal and vertical motion followed the same pattern of change, with no significant motion in the lateral direction. The excursions persisted for over one minute. Conclusions: Transrectal implantation of transponders for use with the Calypso System has an acceptable adverse event profile that appears consistent with gold marker implantation. Care should be taken to implant transponders within the prostate gland, off mid-line, to avoid migration. Real-time target position monitoring is possible for the first time without requiring X-rays for imaging. Tracking the prostate showed significant motion over time periods of interest in radiation therapy in some patients. Real-time tracking of organ motion will play an important role in modern external beam radiotherapy. 1088 Assessing the Utility of a 2-Year Post Treatment Biopsy for Forecasting Eventual Biochemical Failure and Survival W. Vance, S. Tucker, R. de Crevoisier, D. Kuban, R. Cheung Radiotherapy, UT MD Anderson Cancer Center, Houston, TX Purpose/Objective: The utility of routine biopsy after prostate cancer radiotherapy (RT) in predicting outcome is controversial. We evaluated the prognostic value of biopsy in predicting biochemical failures (BF) and overall survival (OS). Materials/Methods: 189 patients underwent planned 2-year post-RT prostate biopsies on UTMDACC trial 93– 001. These patients were treated with external beam RT from 1993 to 1998. We used ASTRO and current nadir + 2 (CN+2) BF definitions. Evidence of metastasis and use of salvage hormone were also considered BF. Patients with biochemical failures prior to the biopsy were excluded. The relationship between Gleason score, 1992 AJCC stage, pre-treatment PSA, risk group (low-risk disease = PSA  10 and Gleason score  6 and AJCC stage T2a or lower disease; high-risk disease = PSA  20 or Gleason score  8 or AJCC stage T3 or higher; intermediate-risk disease = all others), RT dose (protocol arm 1=70Gy, arm 2= 78Gy), and 2-year post-RT biopsy and the time to BF and OS were analyzed with Kaplan Meier, Cox multivariate and Recursive Partitioning Analyses (RPA). The analysis was performed with biopsy time as time zero, and was repeated with time zero at end of RT. Results: There were 171 patients available for analysis. 90 patients were in protocol arm 1 and 81 were in arm 2. 33 were low-risk, 84 were intermediate-risk, and 54 were high-risk. Two-year biopsy results were: normal (N = 81), atypical cells (N = 32), treatment effect (TE) (N = 43), and Gleason 6 –9 adenocarcinoma (N = 15). Time to CN+2 failure after biopsy was not significantly related to protocol arm (p = 0.506), but was related to risk group (p = 0.004). CN+2 failure was also significantly associated with biopsy results (p = 0.029). The main difference was between biopsy (normal or atypical cells) and biopsy (TE or histology 6 –9). The recursive partitioning analysis produced: low/int risk and normal or atypical cells vs low/int risk and TE or histology 6 –9 vs high risk with any biopsy results. The Cox PH model identified that high-risk and biopsy histology 6 –9 were independent risk factors for CN+2 failure after biopsy, with hazard ratios 3.0 (95% CI 1.5 to 5.9) and 2.9 (95% CI 1.1 to 7.4), respectively. Measuring CN+2 failure from the end of RT instead of from biopsy had essentially no effect on the results. Neither protocol arm (p = 0.289), nor risk group (p = 0.082), nor biopsy results (p = 0.406) was significantly associated with time to ASTRO failure. Biopsy results were also not significant when dichotomized. Neither multivariate analysis (RPA and Cox PH) identified any factors related to ASTRO failure in these 171 patients. None of these conclusions changed if ASTRO failure was measured from end of RT instead. There were 19 deaths in this cohort of 171 patients. There was no effect of protocol arm (p = 0.682), risk group (p = 0.128) or biopsy results (p = 0.152) when all categories were considered, but there were some significant effects when the data were dichotomized. Patients with biopsy histology 6 –9 had significantly worse survival from biopsy (p = 0.034), and patients with intermediate or high risk disease had marginally worse survival than patients with low risk disease (p = 0.055). In multivariate S197 Proceedings of the 47th Annual ASTRO Meeting