Childs, Daniel S.

TPS283 Background: Emerging evidence suggests that metastasis-directed therapy (MDT) with stereotactic body radiotherapy (SBRT) can delay systemic therapy in oligometastatic castration-sensitive prostate cancer (omCSPC). However, the optimal timing of androgen deprivation therapy (ADT) and androgen receptor pathway inhibitor (ARPI) initiation in this setting remains undefined. While deferred ADT following MDT may preserve quality of life, early short-course ADT/ARPI may eradicate micrometastatic disease and extend progression-free survival. Extracellular vesicles (EVs) carrying PSMA-positive prostate cancer (PC) signals and circulating tumor DNA (ctDNA) may serve as real-time biomarkers of MRD and treatment response. Methods: DIVINE [NCT06378866] is an open-label, phase II, randomized trial enrolling men with omCSPC defined by 1–5 metastases (≥1 extrapelvic) on conventional imaging (CT/MRI and bone scan) and testosterone >100 ng/dL. PSMA PET may be obtained for correlative purposes but does not determine eligibility. Participants (N = 220) are randomized 1:1 to: Arm A (Early systemic): SBRT plus 6 months of ADT + ARPI. Arm B (Deferred systemic): SBRT alone with observation, reserving systemic therapy until radiographic progression not amenable to further SBRT (mrPD). SBRT is delivered per institutional standard; repeat SBRT is allowed until lesions become non-addressable. The primary endpoint, modified radiographic progression-free survival (mrPFS), is the time from randomization to death or radiographic progression not amenable to further MDT, defined by PCWG3 (bone) and modified RECIST 1.1 (soft tissue) criteria on conventional imaging. Soft-tissue progression requires ≥ 20% increase in measurable disease or a new lesion; bone progression requires ≥ 2 new confirmed lesions. Secondary endpoints include OS, biochemical PFS, time to local and distant progression, and grade 3–5 adverse events during systemic therapy. Exploratory endpoints evaluate EV and ctDNA kinetics as MRD markers, correlation of PSMA-positive EV burden with outcomes, and duration of response after short-course ADT/ARPI. Correlative analyses will longitudinally quantify PSMA-positive EVs and ctDNA at baseline, on treatment, and at progression to define MRD dynamics and resistance patterns. The trial uses Pocock–Simon dynamic allocation by EV strata and metastatic burden to ensure balance across arms. The trial is open and actively enrolling across all Mayo Clinic sites; as of October 2025, 35 of the planned 220 patients have been accrued. Clinical trial information: NCT06378866 .

483 Background: Sarcomatoid dedifferentiation (sRCC) in metastatic clear-cell renal cell carcinoma (mccRCC) confers aggressive disease and poor prognosis. We maintain a living meta-analysis of first-line immune checkpoint inhibitor (ICI)–based combinations in mccRCC (Riaz et al., Eur Urol Oncol 2021), which demonstrated improved survival with ICI-based regimens over sunitinib. Here, we present an updated synthesis focused on the comparative efficacy of ICI combinations in the sarcomatoid subset. Methods: A systematic search of MEDLINE, EMBASE, and CENTRAL through October 2025 was performed as part of our living meta-analysis of first-line mccRCC therapy. Eligible phase III RCTs compared ICI combinations (IO/IO or IO/TKI) versus sunitinib, including CheckMate 214, CheckMate 9ER, KEYNOTE-426, CLEAR/KEYNOTE-581, JAVELIN Renal 101, and IMmotion151. Extracted outcomes included overall (OS) and progression-free survival (PFS), and objective (ORR), complete (CR), and partial (PR) response rates. Random-effects meta-analyses generated pooled hazard ratios (HRs) for time-to-event outcomes and risk ratios (RRs) for response rates with 95% confidence intervals. Subgroup interaction analyses compared sarcomatoid versus non-sarcomatoid subgroups, with p < 0.1 considered significant. Results: Six phase III RCTs (n = 5,121; 624 sarcomatoid) were included. Across all ICI-based combinations, ICI therapy improved outcomes versus sunitinib in both sarcomatoid and non-sarcomatoid disease. In the pooled analysis, the OS interaction favored sRCC meeting the prespecified significance threshold for subgroup analyses (HR 0.53 vs 0.68; p-interaction = 0.06 < 0.10), while no significant interaction was seen for PFS. Response outcomes showed the strongest differential benefit, particularly with IO/TKI regimens. IO/IO results were directionally consistent but limited by smaller non-sarcomatoid samples. Conclusions: ICI-based combinations significantly improve survival and response outcomes versus sunitinib in mccRCC. Using a prespecified α = 0.10 for interaction testing, the pooled analysis showed a significant OS interaction and higher response rates in sRCC, indicating enhanced ICI sensitivity. These updated meta-analysis data support ICI regimens as the preferred first-line therapy for mRCC with sarcomatoid de-differentiation. Endpoint Regimen Subgroup (S = sarcomatoid; NS = non-sarcomatoid) Effect (95% CI) p-interaction PFS IO/IO S 0.51 (0.40–0.65) – IO/TKI S vs NS 0.48 (0.36–0.64) vs 0.46 (0.39–0.52) 0.81 OS IO/IO S 0.54 (0.07–4.42) – IO/TKI S vs NS 0.51 (0.17–1.48) vs 0.68 (0.30–1.58) 0.24 ORR IO/IO S 2.93 (0.49–17.37) – IO/TKI S vs NS 2.43 (1.84–3.21) vs 1.95 (1.67–2.28) 0.03 CR IO/IO S 3.81 (0.82–17.71) – IO/TKI S vs NS 5.93 (4.73–7.42) vs 2.85 (0.10–83.73) 0.007 PR IO/IO S 3.48 (1.70–7.11) – IO/TKI S vs NS 2.10 (1.78–2.47) vs 1.75 (1.18–2.61) <0.001

464 Background: Based on LITESPARK-005 trail, Belzutifan was FDA-approved to treat advanced RCC following ICI and a VEGF-TKI. We conducted a retrospective cohort study to report the real-world outcomes of Belzutifan in pre-treated RCC patients. Methods: A retrospective chart review of all patients who received Belzutifan at the Mayo Clinic Comprehensive Cancer Center between Dec 2023 and Sep 2025 was conducted. Patients treated for sporadic RCC were eligible; those with von Hippel–Lindau disease, non-RCC pathologies (e.g., Pheochromocytoma, ovarian cancer) were excluded. Radiologic response was assessed by trained curators using RECIST-defined criteria, approximated through systematic review of radiology reports with clinician adjudication when needed. Progression-free survival (PFS), defined as the time from treatment initiation to radiographic progression or death, was estimated using the Kaplan–Meier method. Non-progressors were censored at the last radiology assessment. Patients who discontinued therapy for non-progression reasons were censored at the last radiology assessment prior to discontinuation. Any-cause hypoxia and anemia were reported using CTCAEv5 definitions. Results: A total of 158 patients were identified; 83 (52.5%) who received Belzutifan for sporadic RCC were included. The remaining were excluded for VHL disease (68, 43%) or other pathologies (7, 4.4%). Median follow up was 12 months, 67% were male, median age was 70 (range 39, 94) with 57 patients (69%) were ≥ 65. 28 patients (34%) had a poor IMDC score at baseline. The cohort received a median of 3 previous lines of therapy (range 1,8). Median treatment duration was 3 months (IQR: 1-9.5). Median PFS was 5.7 months (95% CI, 3.8–8.7). At 12 and 18 months, 23.1% and 13.9% of patients, respectively, remained alive and progression-free. Corresponding OS rates at these time points were 60.2% and 51.5%. A confirmed ORR occurred in 12 patients (14.4%). Hypoxia occurred in 44 (53%) patients, median onset 31 days (range 1–270) and 29 (34.9%) required hospitalizations. Median hemoglobin (Hb) levels at baseline were 12.65 g/dl (range, 7.2-15.5), and 68 (82%) patients developed Hb decline on Belzutifan, 27 (32.5%) required transfusions. Conclusions: In an older, high-risk cohort with higher number of prior lines of therapy, Belzutifan yielded PFS/OS and ORR landmarks comparable to the trial supporting its effectiveness in the real world. Outcomes compared with LITESPARK-005 trial. Variable Real world cohort (Mayo Clinic) LITESPARK-005, Belzutifan arm Median PFS (months, 95% CI) 5.7 (3.8-8.7) 5.6 (3.9-7) 12-month PFS rate (%, 95% CI) 23.1% (14.4-33.4) 33.4% (28.4-33.5) 18-month PFS rate (%, 95% CI) 13.9% (7.7-23.9) 24% (19-29.4) 12-month OS rate (%, 95% CI) 60.2% (48.9-70.8) 67.9% (62.9-72.4) 18-month OS rate (%, 95% CI) 51.5% (40.6-62.9) 55.2% (50-60.1) ORR (%, 95% CI) 14.4% (7.7-23.9) 21.9% (17.8-26.5)

214 Background: 177 Lu-PSMA-617 (LuPSMA) has revolutionized care for patients (pts) with metastatic castrate-resistant prostate cancer (mCRPC), but only half achieves PSA decline ≥50%. Biomarkers are urgently needed to improve pt selection. EVs and ctDNA released by cancer cells into the blood are emerging biomarkers to estimate overall tumor burden and capture tumor heterogeneity in mCRPC. Here, we analyzed pre-treatment plasma levels of tumor EVs and ctDNA in a pilot cohort of 35 pts with mCRPC treated with LuPSMA. Methods: Blood samples were collected pre-cycle 1 of LuPSMA at Tulane Cancer Center between 12/2018-1/2025. Plasma-derived small and large EVs were incubated with fluorescently conjugated antibodies and enumerated using nanoscale flow cytometry. ctDNA fraction was estimated from plasma cfDNA methylation WGS (∼10×) with ichorCNA. We also analyzed EVs from blood samples of male healthy donors (HD) and pts with localized prostate cancer (LPC) collected 1 day to 8 months after radical prostatectomy (post-RP) from 6/2020-8/2024 at Mayo Clinic as controls. Results: Median age at collection for 10 HD was 55 yrs and for 36 pts with LPC post-RP was 65 yrs. Median PSMA+ EV level was 2.0 x 10 6 /mL for post-RP and 1.4 x 10 6 /mL for HD. For 35 pts with mCRPC, 25 pts had complete data for analysis. Median PSMA+ EV level was 1.4 x 10 7 /mL, significantly higher than post-RP (p<0.001) and HD (p<0.001). At LuPSMA start, median age was 69 yrs (range 54-83), PSA was 82 ng/mL (range 0.8-503), median time from CRPC diagnosis to start was 50 months (range 6.9-285), median time from plasma collection to treatment initiation was 0.5 months (range 0-7.3). Median number of prior lines of life-extending therapy was 4 (range 1-11), including 100% androgen receptor pathway inhibitor and 75% docetaxel. Pts received median 4 cycles of LuPSMA. Median ctDNA fraction was 14.0% (range 0-51.1). Median time to reach nadir PSA for pts with any decline was 2.6 months. 15 pts (60%) achieved PSA40 (defined ≥40% decline in PSA from start to nadir) (range -99% to -42%) compared to 10 pts (40%) who did not (range -16% to rise), which predicted better median overall survival (OS) after LuPSMA start (40.7 v 6.0 months; HR 0.23, 95% CI, 0.07-0.71; p=0.0001). Pts were classified as LuPSMA responders using PSA40. Responders had a lower ratio of large to small PSMA+ EVs (0.30 v 0.53, p=0.0042) with an area under the receiver operating characteristic curve (AUC) of 0.83 (95% CI, 0.68-0.99), and a lower median ctDNA fraction (6.1% v 26.2%, p=0.0044) with an AUC of 0.83 (95% CI, 0.68-1.00). Incorporating both PSMA+ EV and ctDNA fraction together predicted PSA40 response with an AUC of 0.93 (95% CI, 0.84-1.00, p=0.0003). Conclusions: Using pre-LuPSMA plasma samples from a heavily pre-treated historical cohort, we integrated non-invasive liquid biopsy methods to successfully predict response to LuPSMA using PSA decline. Prospective validation is underway.

PURPOSE The continuous development of new imaging approaches, molecular phenotyping, genetic subtypes, prognosis assessments, and effective therapies across a range of disease states has created a need to redefine terminology and best practices for clinical trial conduct in patients with advanced prostate cancer. Methods We convened an international expert committee of diverse working groups, the Prostate Cancer Working Group 4 (PCWG4), between 2016 and 2025. Our objective was to formulate updated criteria based on emerging evidence and clinical trial data in a biomarker context to provide guidance for clinical trial design, eligibility, and end point assessments for patients with advanced prostate cancer. RESULTS PCWG4 redefines terminology around the disease state and previous therapies in a patient-centric context and terminology focused on androgen pathway modulation. We consider imaging, with a particular focus on positron emission tomography (PET)–defined disease. New recommendations are provided for disease state terminology, defining eligibility criteria, response and delay/prevent end points, intervals for reassessments including imaging, and patient-reported outcome determination. We provide recommendations in a biomarker-based context of use for the intended indication, reflective of patient benefit for specific interventions. We emphasize the need for development of validated PET imaging and molecular and phenotypic criteria as well as trial designs to appropriately risk stratify patients, predict and assess benefit, and measure post-treatment outcomes reliably in a trial framework. CONCLUSION PCWG4 updates recommendations on patient and tumor characterization, therapy development, and imaging criteria and extends guidance into earlier androgen pathway modulator–naïve/sensitive disease states to reflect an evolving, heterogeneous, and diverse patient population to optimize treatment benefits for all patients.

Background: Longitudinal genomic profiling through serial circulating tumor DNA (ctDNA) testing offers a noninvasive method to monitor clonal evolution in advanced prostate cancer. This study evaluated the frequency and nature of newly emergent and potentially actionable genomic alterations detected through serial testing in a real-world setting. Methods: We conducted a retrospective analysis of advanced prostate cancer patients who underwent multiple Guardant360 ctDNA tests between October 2020 and March 2023. The study focused on identifying new genomic alterations absent in the initial test, with particular attention to alterations relevant for on-label therapies, therapies approved in other oncologic indications (i.e., off-label), or a clinical trial. Results: Among 479 patients with at least two ctDNA tests, the median interval between the first and second evaluable tests was 207 days. New and potentially actionable alterations emerged in 57.8% of patients, including potential targets for on-label therapies (16.7%), off-label therapies (16.5%), and clinical trials (55.7%). Tumor mutational burden (TMB) increased from “low” to “high” in 11% of patients, although none had microsatellite instability or mismatch repair deficiency. In a Mayo Clinic subset, ten patients received olaparib based on treatment-emergent alterations, but none achieved a prostate-specific antigen (PSA) response. Two patients who transitioned from low to high TMB received pembrolizumab, both with progressive disease as best response. Conclusions: In a large real-world cohort, serial ctDNA testing frequently identified new alterations that were not detected at baseline and are potentially actionable therapeutic targets, highlighting the value of serial genomic profiling for capturing clonal dynamics. Additional research is needed to better establish a framework for retesting and to clarify how these results should influence subsequent treatment decisions.

Black patients with metastatic prostate cancer have higher mortality rates compared to non-Hispanic White patients. There are no data on outcomes with [177Lu]Lu-PSMA-617 (LuPSMA) across racial groups. We evaluated the association between race and outcomes with LuPSMA in a multi-institutional cohort of consecutive patients with mCRPC treated with LuPSMA. The primary outcomes were PSA-50 rate, PSA-progression-free survival (PSA-PFS), and overall survival (OS). Statistical models were adjusted for age, number of prior therapies, sites of metastases, and baseline PSA. A total of 654 patients were included; 593 (91%) were White, 45 (7%) were Black and 16 (2%) were another non-Black minority (NBM). There were no statistically significant differences in PSA-50 rates, PSA-PFS and OS between the groups. Black and White patients treated with LuPSMA had similar clinical outcomes; efforts are needed to ensure Black and NBM patients have equal access to life prolonging therapies to narrow disparities in outcomes.

5108 Background: We previously proposed a hybrid framework combining LLMs and rule based algorithm (RBA) for automating risk stratification in locPCa. Herein, we aim to validate the risk stratification agent (RSA) in a prospective cohort, develop & evaluate the treatment recommendation agent (TRA) based on NCCN guidelines, and develop an interactive interface to facilitate clinicians for accurate risk stratification and treatment recommendations at the point of care. Methods: This study included pts with locPCa (2004-2024) presenting at Mayo Clinic with at least 1 positive prostate biopsy and MRI report available. For RSA prospective validation, GPT4 extracted key phenotypic variables (PSA, T stage, prostate volume, number of cores, Gleason patterns, grade group) from unstructured MRI and biopsy reports using a zeroshot prompt. An RBA then classified pts into NCCN risk groups. The agent performance was compared with the treating clinician's documentation and evaluated against gold-standard labels manually annotated by two independent clinicians. For development of TRA, two experiments were performed using GPT4 with and without retrieval-augmented generation (RAG) to generate treatment plans. Generated treatment plans were evaluated using NCCN guideline-informed treatment decision tree based algorithm (DTA). Evaluation metrics, weighted – accuracy (acc) and F1, were computed. A clinician facing interface (lisr.org/risk) was developed to provide accurate risk stratification and treatment plans. Results: A total of 858 pts were included (500 for prospective validation, 358 for treatment recommendations). Prospective validation for RSA demonstrated a higher F1 score of 0.89 compared to the treating clinician (F1: 0.58). Treatment recommendation experiments showed that GPT4 with RAG achieved higher acc (64% full – all correct treatment options and 36% partial – at least 1 correct treatment option) compared to GPT4 alone (35% and 65%, respectively). Sensitivity analysis using DTA-informed GPT4 note generation achieved 94% full acc and 6% partial acc. GPT4 alone generated hallucinated treatment options in 71% of cases, while GPT4 with RAG reduced this to 32%. Conclusions: The multi-agent framework based on LLM and RBA achieves high accuracy in risk stratification and treatment recommendation for locPCa. A multi-agent framework with an interactive interface holds high promise to enable efficient, accurate decision-making, and improve locPCa management at the point of care Performance evaluation. RSA(acc; F1) % TRA(full acc; partial acc) % Clinician RSA* GPT4 RAG+GPT4 DTA+GPT4 Overall 60; 58 89; 89 29; 71 64; 36 94; 6 V. High 94; 33 96; 96 55; 45 69; 31 93; 7 High 46; 52 92; 89 - - - Int. Unf 80; 61 91; 90 5; 95 38; 62 93; 7 Int. Fav 57; 64 84; 87 0; 100 98; 2 100; 0 Low 57; 65 87; 80 100; 0 95; 5 90; 10 V. Low 8; 13 50; 67^ - - - *Statistical significance (p<0.01) except for ^V. Low.

261 Background: Clinicians use next-generation sequencing (NGS) to tailor targeted treatments for PCa patients. However, the results are often available in unstructured formats and are saturated with technical jargon, making them inaccessible for patients. Therefore, we aimed to assess the efficacy of LLMs to accurately extract genomic characteristics from unstructured reports and generate patient-friendly summaries for ease of variant interpretation in PCa patients. Methods: This retrospective study included patients with PCa who underwent NGS from 2023-2024. The GPT4 model was utilized using structured zero-shot prompts to extract genomic characteristics from unstructured reports in a non-decomposed manner. Prompt development was conducted iteratively using a 2% random sample of the dataset. Genomic characteristics ascertained by LLM were assessed against held out vendor-curated test dataset for evaluation. Performance was assessed using weighted evaluation metrics (precision, recall, and F1-score). Additionally, the extracted variables were organized using rule-based criterion and presented as a separate prompt to GPT4 to generate summaries for interpretation at the level of each biologically relevant variant. Mean Flesch-Kincaid scores and Vocd-D statistic with standard deviations were computed to assess readability and lexical diversity of LLM-generated variant interpretation summaries. Results: A total of 331 patients (370 NGS reports) were included in the study. Evaluation on the held-out test dataset (Table) showed that the weighted average precision ranged from 80% for extracting tumor mutational burden to 100% for extracting the CtDNA fraction from unstructured reports. The weighted average recall ranged from 85% for extracting CtDNA tumor fraction to 99% for ascertaining variant/alteration. The weighted average F1-scores varied from 0.83 for extracting tumor mutational burden to 1.0 for extracting specimen site from NGS reports. The mean Flesch-Kincaid score was 8.29 ± 1 indicating average/standard readability ease. The mean Vocd-D statistic was 7.55 ± 1 indicating focused summaries, however, with limited lexical diversity. Conclusions: This study suggests that large language models can effectively extract genomic characteristics from unstructured reports and can potentially improve genomic literacy among prostate cancer patients by providing easy-to-interpret variant summaries. However, targeted prompting may be required to increase lexical diversity for improved engagement. Performance across different genomic characteristics. Category Precision Recall F1-Score Altered gene 0.86 0.97 0.89 Variant/alteration type 0.98 0.99 0.98 Variant allele fraction 0.96 0.96 0.96 Test name/type 1 0.99 1 Specimen site 1 0.99 1 Tumor mutational burden 0.8 0.93 0.83 CtDNA tumor fraction 1 0.85 0.92 MSI status 0.99 0.97 0.98