PSOR06 Presentation Time: 12:25 PM

Purpose MRI-assisted radiosurgery (MARS) is a novel LDR brachy technique with soft tissue delineation. We report outcomes and toxicity of salvage MARS with Pd-103 after various treatments including surgery, EBRT, and brachy. Materials & Methods We reviewed those undergoing salvage MARS for intraprostatic or prostate bed recurrences using Pd-103. Intraprostatic recurrences underwent whole or focal gland MARS based on template-guided biopsies. Preop diagnostic MR obtained 2 weeks before MARS was used to contour targets and organs-at-risk (OAR) on 3D-T2W sequence including the external urethral sphincter (EUS). Treatment sim was performed with a virtual probe and needles using MIM software. Intra-op transrectal ultrasound fused MR was verified with implantation and confirmed by postop MR on day 0. Constraints were rectum V100<1 cc and a EUS V200 <0.04 cc. Urethra was avoided with planning. No patients had metastatic disease. All patients except 1 were treated with MARS without concurrent ADT. Freedom from biochemical failure (FFbF) defined as post MARS nadir + 2 ng or initiation of ADT, hormone therapy-free survival (HFS), and overall survival (OS) were derived by Kaplan-Meier estimates. Toxicity was graded by CTCAE v5.0. Statistics performed with SAS Pro 15.0. Results 31 patients were treated between 2016-2022. Median follow up was 1.6 years. Gleason grade grouping at diagnosis was: GG1 2 patients (6%), GG2 8 (26%), GG3 10 (32%), GG4 10 (32%), and GG5 1 (3%). Median age at MARS was 71 years. 4 patients (13%) had prior prostatectomy (RP) with salvage EBRT. 24 patients (77%) underwent upfront EBRT. 3 patients had prior brachy. Median PSA at MARS was 3.3 (range, 0.01-16.3). 26 patients (84%) underwent PSMA or Fluciclovine PET before MARS. 22 patients underwent focal MARS, while 9 others had whole gland MARS. 30 patients received 100Gy, while 1 received 125Gy. Median D90 was 128 Gy, and median V200 was 54%. Median rectum V100 was 0 cc, and median EUS V200 was 0.004cc. 27 patients had SpaceOAR placement at MARS (87%). At last follow-up, 3 patients (10%) experienced bF and ADT was started, 1 additional patient started ADT before meeting bF criteria. Among 3 with bF, 2 underwent upfront RP, and 1 underwent upfront LDR. 27 patients (87%) remain bF-free and off ADT. 2 patients died from comorbidities. 2 patients had PET-avid recurrences with 1 seminal vesicle plus nodal recurrence and 1 prostate bed only. No patients died of prostate cancer. 2-year FFbF and HFS was 84.4% and 83.5%, respectively. Median PSA nadir without bF was 0.2 ng/mL. Grade 3 toxicity occurred in 5 patients (16%), including 4 patients (13%) with GU events and 1 patient (3%) with a gastrointestinal (GI) event. Grade 3+ GI involved anterior rectal wall infiltration of SpaceOAR requiring a diverting colostomy. All patients with grade 3+ GU underwent upfront EBRT. Grade 3+ GU was not associated with EUS dose, SpaceOAR usage, EBRT type, or V200 dose. Conclusions In a modern salvage MARS series with Pd-103, we observed safe and tolerable treatment with acceptable toxicity. MARS allows focal brachy of intraprostatic lesions given its reliance on MR planning, unlike ultrasound in most salvage techniques. Early data suggests promising FFbF, and broader applicability of MARS regardless of upfront treatment. MRI-assisted radiosurgery (MARS) is a novel LDR brachy technique with soft tissue delineation. We report outcomes and toxicity of salvage MARS with Pd-103 after various treatments including surgery, EBRT, and brachy. We reviewed those undergoing salvage MARS for intraprostatic or prostate bed recurrences using Pd-103. Intraprostatic recurrences underwent whole or focal gland MARS based on template-guided biopsies. Preop diagnostic MR obtained 2 weeks before MARS was used to contour targets and organs-at-risk (OAR) on 3D-T2W sequence including the external urethral sphincter (EUS). Treatment sim was performed with a virtual probe and needles using MIM software. Intra-op transrectal ultrasound fused MR was verified with implantation and confirmed by postop MR on day 0. Constraints were rectum V100<1 cc and a EUS V200 <0.04 cc. Urethra was avoided with planning. No patients had metastatic disease. All patients except 1 were treated with MARS without concurrent ADT. Freedom from biochemical failure (FFbF) defined as post MARS nadir + 2 ng or initiation of ADT, hormone therapy-free survival (HFS), and overall survival (OS) were derived by Kaplan-Meier estimates. Toxicity was graded by CTCAE v5.0. Statistics performed with SAS Pro 15.0. 31 patients were treated between 2016-2022. Median follow up was 1.6 years. Gleason grade grouping at diagnosis was: GG1 2 patients (6%), GG2 8 (26%), GG3 10 (32%), GG4 10 (32%), and GG5 1 (3%). Median age at MARS was 71 years. 4 patients (13%) had prior prostatectomy (RP) with salvage EBRT. 24 patients (77%) underwent upfront EBRT. 3 patients had prior brachy. Median PSA at MARS was 3.3 (range, 0.01-16.3). 26 patients (84%) underwent PSMA or Fluciclovine PET before MARS. 22 patients underwent focal MARS, while 9 others had whole gland MARS. 30 patients received 100Gy, while 1 received 125Gy. Median D90 was 128 Gy, and median V200 was 54%. Median rectum V100 was 0 cc, and median EUS V200 was 0.004cc. 27 patients had SpaceOAR placement at MARS (87%). At last follow-up, 3 patients (10%) experienced bF and ADT was started, 1 additional patient started ADT before meeting bF criteria. Among 3 with bF, 2 underwent upfront RP, and 1 underwent upfront LDR. 27 patients (87%) remain bF-free and off ADT. 2 patients died from comorbidities. 2 patients had PET-avid recurrences with 1 seminal vesicle plus nodal recurrence and 1 prostate bed only. No patients died of prostate cancer. 2-year FFbF and HFS was 84.4% and 83.5%, respectively. Median PSA nadir without bF was 0.2 ng/mL. Grade 3 toxicity occurred in 5 patients (16%), including 4 patients (13%) with GU events and 1 patient (3%) with a gastrointestinal (GI) event. Grade 3+ GI involved anterior rectal wall infiltration of SpaceOAR requiring a diverting colostomy. All patients with grade 3+ GU underwent upfront EBRT. Grade 3+ GU was not associated with EUS dose, SpaceOAR usage, EBRT type, or V200 dose. In a modern salvage MARS series with Pd-103, we observed safe and tolerable treatment with acceptable toxicity. MARS allows focal brachy of intraprostatic lesions given its reliance on MR planning, unlike ultrasound in most salvage techniques. Early data suggests promising FFbF, and broader applicability of MARS regardless of upfront treatment.