pre-RT PSA level available; the mean PSADT was 8.6 months. No androgen deprivation was delivered at the time of salvageradiotherapy. The median follow-up time was 32 months (range: 1-187).

Results: For the entire cohort, the 5 and 7-year bRFS rates after salvage prostate fossa radiotherapy were 54% and 50%,respectively. The 5 and 10-year overall survival rates were 90% and 80%, respectively. The 5 and 10-year clinical disease-freesurvival rates were 89% and 70%, respectively. On univariate analysis, initial biopsy Gleason score, clinical T-stage, SM status,SV status, and pre-RT PSA were significant predictors of biochemical failure after RT. Age, race, initial PSA level, ECE, LN,PSADT, RT dose, and RT technique were not. The 5-year bRFS rates for patients with negative SM versus positive SM were42% and 62%, respectively (p�0.002). The 5-year bRFS rates for patients with positive SV versus negative SV were 20% and64%, respectively (p�0.001). The 5-year bRFS rates for patients with a preRT PSA �1.0 versus �1.0 were 58% and 37%,respectively (p�0.079). On multivariate analysis, SM status, SV status, and pre-RT PSA (continuous variable) were the onlysignificant predictors of failure. Initial biopsy Gleason score and clinical T-stage were not. On the basis of the independentpredictors of biochemical failure after RT, two risk groups were identified: favorable (positive surgical margins, negativeseminal vesicles, and preRT PSA �1.0), and unfavorable (negative surgical margins or positive seminal vesicles or preRT PSA�1.0). The 5-year bRFS rates for patients favorable versus unfavorable risk were 74% and 41%, respectively (p�0.001, Fig.1).

Conclusions: Patients experiencing biochemical failure after radical prostatectomy have a high likelihood of achievingbiochemical control after radiation therapy to the prostatic fossa if they had surgical margin involvement, no seminal vesicleinvolvement, and a pre-radiation PSA level 1.0 ng/mL or less.

2079 PSA Kinetics Following IMRT for Prostate Cancer: The Impact of Daily Stereotactic UltrasoundTargeting

M. Fuss,1 S. Cavanaugh,1,2 L. Voeltz,2 C. Fuller,1,2 B. Salter,1,3 T. Herman1

1Radiation Oncology, University of Texas Health Science Center at San Antonio, San Antonio, TX, 2Division ofRadiological Sciences, University of Texas Health Science Center at San Antonio, San Antonio, TX, 3Cancer Therapy andResearch Center, San Antonio, TX

Purpose/Objective: In a recent analysis of a large prostate cancer patient database from the Cleveland Clinic, the probabilityof long-term biochemical control has been linked to the event and the time to fall below PSA cutoff values of 1.0, 0.5 and 0.2ng/ml (submitted as an abstract to ASTRO 2003). Early fall below the chosen cutoff PSA values was significantly correlatedwith reduced risk for biochemical failure. Thus, despite the widely accepted conclusion that PSA half-life may not be aprognosticator and early nadir may rather predict unfavorable outcomes, an analysis of early PSA kinetics, and more specificallythe fall below such favorably low PSA cutoff values might yield prognostically significant outcome information. This mayprovide a means to assess potential improvements in outcomes following new or modified treatment regimens when long-termoutcome data are not yet available. Daily stereotactic ultrasound targeting for external beam radiotherapy of prostate cancer hasbeen accepted as a routine component of prostate cancer radiotherapy and is generally considered to improve outcomes by moreprecise radiation dose delivery. However, biochemical or survival outcome data confirming this hypothesis are not yet available.The present study evaluated the kinetics of early PSA response in a prostate cancer patient population treated by IMRT withor without daily stereotactic ultrasound target localization during the treatment course.

Materials/Methods: Inclusion criteria for the present study were: (a) tomotherapeutic IMRT treatment with curative intent forclinically localized prostate cancer; (b) minimum treatment dose 70 Gy (mean total dose 75.1 Gy, range 70.2 to 80 Gy); (c) atleast one follow-up PSA value; (d) minimum clinical follow-up 6 months. Of 60 patients treated between 2/99 and 3/02, 38patients were aligned for their IMRT treatment using the BAT daily stereotactic ultrasound target localization device (NomosCorp., Sewickley, PA). We determined the time to reach PSA cutoff values of �1.0, �0.5 and �0.2 ng/ml, the percentage ofpatients below the respective cutoff values during follow-up and the correlation of PSA response with use of the BAT device.We also analyzed the impact of covariates such as initial PSA, Gleason score, tumor stage, and hormone treatment.

Results: During limited follow-up (mean 655 days, range 309 to 1437 days), 47, 38, and 24 patients reached PSA cutoff valuesof 1.0, 0.5 and 0.2 ng/ml, respectively. In the cohort of patients treated by IMRT with ultrasound targeting, 84.2% and 68.4%reached PSA cutoff values of 1.0 and 0.5 ng/ml compared with 68.2% and 54.5% of patients treated without BAT. In patientstreated with ultrasound targeting, Kaplan Meier curves to reach PSA cutoff values were statistically significantly different forPSA �1.0 (P�0.001, logrank) and PSA �0.5 (P�0.005, logrank), as was mean time to fall below PSA 1.0 and 0.5 (P�0.006and P�0.044), respectively. Kaplan-Meier comparison of subpopulations, defined by chosen covariates showed no significantdifference except use of daily ultrasound targeting (logrank).

S390 I. J. Radiation Oncology ● Biology ● Physics Volume 57, Number 2, Supplement, 2003

Conclusions: Post treatment PSA kinetic differed significantly in patients treated by IMRT with or without daily ultrasoundprostate targeting. A significantly higher percentage of patients treated using ultrasound targeting reached prognosticallyfavorable PSA cutoff levels during limited follow-up. Thus, the present study indicates that more precise daily prostate targetingachieves a prognostically significant treatment goal earlier in follow-up. The observed higher percentage of patients withfavorable interim PSA suggest improved outcomes following IMRT with daily ultrasound targeting. This conclusion issupported by data derived from a long-term follow-up database confirming the prognostic significance of reaching PSA valuesbelow 1.0, 0.5 and 0.2 ng/ml, and proving that reaching those cutoff values early in follow-up yields significantly betteroutcomes.

2080 Long-Term Results of Radiotherapy for Early-Stage Testicular Seminoma

A.M. Garcia-Serra, R.A. Zlotecki, R.J. Amdur

Radiation Oncology, University of Florida, Gainesville, FL

Purpose/Objective: To evaluate the outcome and assess the rate of complications in patients treated with adjuvant radiotherapyfor early-stage testicular seminoma over a 35-year period at a single institution.

Materials/Methods: An outcome analysis for 73 male patients treated at the University of Florida with adjuvant radiotherapy(RT) following orchiectomy was performed. All patients had the classical seminoma histology and no patient receivedchemotherapy as part of the initial treatment. The majority of patients were Marsden Stage I (57of 73). Few patients were StageII (IIA, 14 of 73 and IIB, 2 of 73). All patients received radiotherapy to the paraortic (PAN) and pelvic fields. Approximatelyone-third of the patients (25 of 73) received prophylactic RT to mediastinal and supraclavicular fields. Data in regards topotential radiotherapy late effects were collected by reviewing the patient medical records as well as by telephone survey.

Results: The median follow-up time was 15 years. Median RT dose to the PAN and pelvic fields was 25 Gy in 17 fractions.Local control and ultimate local control at 20 years were 95% and 97%, respectively. Cause-specific survival at 20 years is 96%.There were three recurrences in total; one of the three patients was salvaged with chemotherapy. In a subset analysis of 41patients, the development of peptic ulcer disease was rare (2%, 1 of 41) but 27% (11 of 41) complained of gastrointestinal refluxsymptoms at some point after RT. At least 78% (32 of 41) fathered a child either before or after the RT. There was noappreciable increased rate of cardiovascular complications in those patients treated with prophylactic RT to the mediastinal andsupraclavicular regions. The rate of non-testicular second malignancy for the overall group was 8% (6 of 73). Only one patient(1of 73) developed a subsequent contralateral testicular malignancy.

Conclusions: Patients treated with adjuvant low-dose RT to the paraortic and pelvic fields for treatment of early-stageseminoma enjoy an excellent long-term cure rate (96% at 20 years). Long-term sequelae from treatment remains acceptable withminimal risk of radiotherapy complications.

2081 Preliminary Report of a Multicenter Spanish Trial (GICOR 05) of Risk-Adapted Androgen AblationCombined with Dose-Escalation 3D Conformal Therapy for Prostate Cancer: Impact on Early Toxicity

A. Zapatero,1 P. Alcantara,1 F. Valcarcel,2 A. de la Torre,2 F.A. Calvo,4 R. Algas,3 A. Bejar,5 J. Maldonado,6 S. Villa7

1Radiation Oncology, Hospital Universitario la Princesa, Madrid, Madrid, Spain, 2Radiation Oncology, Hospital Puerta deHierro, Madrid, Spain, 3Radiation Oncology, Hospital Gregorio Maranon, Madrid, Spain, 4Radiation Oncology, HospitalClinico de Valencia, Madrid, Spain, 5Radiation Oncology, Hospital Reina Sofia de Cordoba, Madrid, Spain, 6RadiationOncology, Hospital General Vall D‘Hebron, Madrid, Spain, 7Radiation Oncology, Institut Catala Oncologia, Madrid,Spain

Purpose/Objective: A multicenter prospective study was conducted to determine the impact on biochemical control andsurvival of risk-adapted androgen ablation (AA) in combination with dose escalation 3D conformal radiotherapy (3D CRT) forlocalized prostate cancer. This is a preliminary report of early toxicity encountered on the GICOR 05 study.

Materials/Methods: Between October 1999 and October 2001, 396 eligible localized prostate cancer patients were enrolled ina prospective multicenter study. One hundred and seventy low risk patients (defined as having T1c stage or T2 stage with PSA�20 ng/ml and Gleason score �7) were treated with 3D-CRT alone. Seventy-three intermediate risk patients (T2 stage withPSA 20-40 ng/ml or Gleason score � 7) were allocated to receive short-term AA, 2 to 6 months prior and during 3D CRT. Onehundred and fifty high-risk patients (T2 stage with Gleason score �7 and PSA 20-40 ng/ml, or T3 stage or Gleason score �7or PSA �40 ng/ml) received long-term AA during and 2 years following 3D CRT. All AA consisted of an LHRH agonist witha non-steroidal anti-androgen. Stratification was performed into 3 groups according to radiation dose level (group I: � 70 Gy;group II: 70-72 Gy; group III: � 72 Gy). RTOG toxicity scores were used. Results Three hundred and eighty seven cases wereanalyzable for early toxicity. Acute toxicity was low with 86% of patients having either no or grade 1 rectal toxicity and 79%having no or grade 1 bladder toxicity. Fifty-four (14%) patients had grade � 2 acute rectal toxicity and 73 (19%) had grade �2 bladder toxicity. Few patients (1.3% and 1.8%) experienced grade 3 rectal or bladder toxicity respectively, and there wereno grade 4 or 5 toxicities. On univariate analysis, factors associated with increased risk of early grade 2 rectal toxicity werepatient risk group (p�0.03), use of AA (p�0.04), radiation dose (p�0.02), inclusion of seminal vesicles on planning targetvolume (PTV) (p�0.04) and inclusion of pelvic lymph nodes on PTV (p�0.006). Factors associated with increased risk ofgrade 2 bladder toxicity were radiation dose (p�0.005), T stage (p�0.03) and inclusion of pelvic lymph nodes on PTV(p�0.01). Multivariate analysis (MVA) showed that the inclusion of lymph nodes on PTV was the only significant factorassociated with higher risk of grade 2 acute rectal toxicity (p�0.003; OR: 2.4; CI: 1.3-4.4). For grade 2 acute bladdercomplications the significant covariates were radiation dose (p�0.04; OR: 2.2; CI: 1.0-4.9) and inclusion of lymph nodes onPTV (p�0.01; OR: 2.1; CI:1.2-3.8). There was not evidence in MVA that AA had a significant impact on rectal or bladdertoxicity.

S391Proceedings of the 45th Annual ASTRO Meeting