Swami, Umang

PURPOSE Bone metastases (BM) occur in approximately 30% of patients with metastatic renal cell carcinoma (mRCC) and are associated with poor survival and symptomatic skeletal events (SSEs). Radium-223, an alpha-emitting bone-seeking radioisotope, and cabozantinib, a tyrosine kinase inhibitor, have demonstrated activity in BM. RADICAL (Alliance A031801; ClinicalTrials.gov identifier: NCT04071223 ) evaluated cabozantinib ± radium-223 in mRCC with BM. METHODS This phase II trial enrolled patients with mRCC of any histology with ≥1 BM. Patients were randomly assigned 1:1 to cabozantinib ± radium-223, stratified by osteoclast-targeted therapy (OTT), prior therapy, opioid use, and International mRCC Database Consortium (IMDC) risk. The primary end point was SSE-free survival (SSE-FS). Secondary end points included safety, objective response rate (ORR), progression-free survival, and overall survival (OS). A target of 124 evaluable patients was planned, with a prespecified interim futility analysis at 50% of expected SSE-FS events; the trial would stop if the stratified hazard ratio (sHR) > 1.0. RESULTS The prespecified interim futility analysis was conducted after 90 patients were enrolled and crossed the futility boundary, leading to closure at 98 patients. The final analysis included all 98 patients. Median age was 63 years, 82.7% had clear cell histology, and 79.6% were using an OTT. IMDC risk was favorable (18.4%), intermediate (67.3%), and poor (14.3%). Median follow-up was 13.1 months. Median SSE-FS for cabozantinib with radium-223 versus cabozantinib was 16.7 versus 17.6 months (sHR, 1.46 [90% CI, 0.86 to 2.51]). Median OS was 28.3 versus 19.7 months (sHR, 1.40 [95% CI, 0.70 to 2.79]). ORR was 19.4% versus 25.0% ( P = .78). Grade ≥3 adverse events were similar across arms (69.6% v 75.5%). CONCLUSION Radium-223 did not improve SSE-FS when added to cabozantinib. The combination demonstrated a manageable safety profile.

This cohort study examines patterns of bone-modifying agent use from 2013 to 2024 among men with metastatic castration-resistant prostate cancer.

21 Background: Large language models (LLMs) are becoming an increasing source of medical information for both patients and providers. Clinical decision making is occasionally complex and may use factors outside of traditional guidelines. We conducted a comparative performance evaluation of frontier models on standardized genitourinary (GU) oncology knowledge assessments derived from the American Society of Clinical Oncology Self-Evaluation Program 2025 (ASCO SEP 2025) in addition to challenging questions developed by GU oncology experts. Methods: Three frontier LLMs were selected: ChatGPT-5.2 (OpenAI, San Francisco, CA, USA), Claude Opus 4.6 (Anthropic, San Francisco, CA, USA), and Gemini 3 Pro (Google DeepMind, Mountain View, CA, USA). Model selection was based on publicly reported performance on “Humanity’s Last Exam,” a benchmark of advanced scientific reasoning (Scale AI leaderboard Jan-Feb 2026). The OpenEvidence (OpenEvidence Inc., MA, USA) search engine was also selected for evaluation. All models were asked 145 ASCO SEP 2025 GU oncology questions and 12 real world challenging scenarios using an uniform prompt. Subject matter experts peer reviewed LLM generated responses to challenging questions and credit was given only for correct rationale and option selection. Results: All models demonstrated high and comparable accuracy on Standardized knowledge assessment, with Gemini 3 Pro leading numerically at 97.24% (95% CI, 93.1% – 99.0%), OpenEvidence followed with an accuracy of 94.48% (95% CI, 89.5% – 97.3%), while Chat GPT -5.2 and Claude OPUS 4.6 scored 92.41% (95%, CI 86.9% – 95.7%) and 79.31% (95% CI, 71.9% – 85.1%) respectively. On challenging real world scenarios, ChatGPT-5.2 demonstrated the highest numerical accuracy at 100% (95% CI, 75.8% – 100%), Gemini 3 Pro and Open Evidence scored 75.0% (95% CI, 46.8%–91.1%) each, followed by Claude Opus 4.6 at 66.7% (39.1–86.3). Conclusions: All current frontier AI models appear to provide accurate medical information and rationale on a large majority of guideline based medical knowledge assessments for GU oncology. Frontier AI models also appear to have emerging competence in highly challenging real-world scenarios. AI based tools may become a valuable and reliable tool for patients and providers. Comparative LLM performance. Model ASCO SEP 2025 % (95% CI) Real World Scenarios % (95% CI) Gemini 3 Pro 97.24% (93.1–99.0) 75.0% (46.8–91.1) OpenEvidence 94.48% (89.5–97.3) 75.0% (46.8–91.1) ChatGPT-5.2 92.41% (86.9–95.7) 100% (75.8–100) Claude Opus 4.6 79.31% (71.9–85.1) 66.67% (39.1–86.3)

57 Background: Cardiovascular disease (CVD) is the leading non-cancer cause of death among men with prostate cancer (PC), particularly during androgen-deprivation therapy. While CVD risk-prediction models are well established in the general population, barriers to integration within oncology workflows remain poorly defined. Guided by the PRISM and RE-AIM frameworks, this study evaluated multilevel determinants influencing adoption, feasibility, and sustainability of CVD risk-assessment tools in PC care. Methods: A national, cross-sectional survey of 45 physicians (medical oncology 49%, urology 38%, radiation oncology 13%) from 18 U.S. states was conducted. Respondents represented early- (27%), mid- (36%), and late-career (38%) clinicians across academic (87%) and community (13%) settings. Quantitative items mapped to PRISM/RE-AIM domains assessed adaptability, cost/resource feasibility, effectiveness, equity, and sustainability; open-ended responses explored contextual barriers and facilitators. Descriptive statistics summarized quantitative data. Results: Physicians reported moderate-to-high readiness to adopt a CVD risk-assessment tool, emphasizing workflow alignment and EHR integration over algorithmic complexity. Table 1 summarizes domain-specific findings. Ninety-three percent of respondents favored lab-independent models that minimize workflow disruption. Conclusions: Pre-implementation analysis highlighted that structural barriers, rather than algorithmic or statistical limitations, remain the primary obstacles to adopting CVD risk-assessment tools in prostate-cancer care. Physicians emphasized the need for EHR-integrated, non-laboratory-based, and time-efficient tools that align with routine oncology workflows and minimize cross-specialty friction. Addressing these practical and organizational determinants may be pivotal to achieving equitable, scalable, and sustainable cardio-oncology implementation across diverse practice settings. PRISM and RE-AIM constructs evaluation results. Construct Likely/Very Likely (%) Interpretation Perceived Effectiveness 87% Broad confidence in clinical utility Equity in Patient Reach 75% Strong perceived equity in reach Equity in Effectiveness 73% Expected to perform consistently across populations Adaptability to Workflow 71% High compatibility with existing clinic workflows Cost/Resource Feasibility 65% Feasible for resource-constrained settings Likelihood of Adoption 60% Moderate readiness for implementation Equity in Adoption 58% Equitable uptake possible despite resource differences Reach to Target Population 57% Moderate likelihood of reaching most patients Sustainability Over Time 51% Moderate long-term feasibility Consistency of Implementation 42% Need for standardized protocols

Abstract Background Clival metastases are a rare and clinically aggressive manifestation of advanced prostate cancer, associated with cranial nerve palsy and poor survival. The molecular features of prostate cancers giving rise to clivus metastases remain unknown. Patients and methods We performed a multi-center retrospective study across six institutions, identifying prostate cancer patients with radiographically confirmed clival metastases and available next-generation sequencing (NGS) data. Baseline characteristics and clinical outcomes were collected. Genomic alterations from tissue- and/or blood-based assays were aggregated at the patient level and compared with a publicly available metastatic castration-resistant prostate cancer (mCRPC) cohort (SU2C/PCF). Results Fifty-nine patients with clival metastases contributed 87 molecular assays. More than half of patients had Gleason grade group 5 cancer and presented with de novo metastatic (M1) disease. The median interval from initial prostate cancer diagnosis to clival metastasis was 71.4 months (95% CI, 42.0-101.7), while median overall survival following clival involvement was only 15.3 months (95% CI, 6.9-22.8). Compared with the SU2C/PCF mCRPC cohort, clival metastases showed significant enrichment of BRAF and CHEK2 alterations as well as homologous recombination repair (HRR) with relative depletion of AR-related, PI3K pathway, and G2–M pathway alterations. Conclusion Prostate cancers giving rise to clival metastases exhibit a distinct molecular profile enriched for DNA damage–repair and RAF kinase alterations, suggesting unique metastatic biology and potential therapeutic vulnerabilities.

Abstract Background Socioeconomic position (SEP) is a fundamental social determinant of health (SDoH) contributing to observed disparities in cancer and cardiovascular (CV) care among individuals with prostate cancer (PC). Understanding the influence of SEP on CV health is essential for addressing outcome inequities within this population. Methods We conducted a retrospective study utilizing Surveillance, Epidemiology, and End Results (SEER)-Medicare to evaluate CV outcomes in patients ≥65 years with PC from 2009 to 2017. We studied the impact of SEP on cardiovascular events (CVEs), cardiovascular mortality (CVm), PC-specific mortality (PCsm), and all-cause mortality. CVE, CVm, and PCsm were analyzed using competing risk models and all-cause mortality with Cox proportional hazards models. A subgroup analysis was performed in Non-Hispanic Black and Non-Hispanic White individuals. Results We included 141 242 patients of whom 76 844 were in high SEP areas (45.6%) and 64 398 in low SEP ones (54.4%). Both groups had a mean age of 72 years; patients in low SEP areas were 67.5% Non-Hispanic White, 34.4% having diabetes, 73.3% hypertension, and 67.7% dyslipidemia versus the high SEP group with 85.6% Non-Hispanic White, 30.4% diabetes, 69.9% hypertension, and 70.7% dyslipidemia. Patients in low SEP areas had higher risks of CVm (subdistribution hazard ratio [sHR] = 1.18, 95% CI = 1.12 to 1.25), PCsm (sHR = 1.12, 95% CI = 1.06 to 1.17), CVE (sHR = 1.07, 95% CI = 1.04 to 1.09), and all-cause mortality (HR 1.16, 95% CI = 1.13 to 1.20). Accentuated results were shown in the Non-Hispanic Black subgroup. Conclusion Adverse SEP plays a significant role in shaping outcomes among older patients with PC, contributing to elevated risks of adverse CV and survival outcomes.

ImportancePatients with metastatic castration–resistant prostate cancer (mCRPC) harboring BRCA1/2 alterations are eligible to receive poly(adenosine diphosphate-ribose) polymerase (PARP) inhibitors as single agents or in combination with an androgen receptor pathway inhibitor after these agents showed survival improvement in their landmark clinical trials. However, data are limited regarding the uptake of PARP inhibitors in these patients.ObjectiveTo investigate the use of PARP inhibitors in patients with mCRPC harboring BRCA1/2 alterations.Design, Setting, and ParticipantsThis retrospective cohort study used the deidentified Flatiron-Health electronic health record–derived database of US community and academic practices to extract patient-level data. The data cutoff date was May 31, 2024. Patients with mCRPC with evidence of harboring BRCA1/2 alterations and alive after August 15, 2020 (ie, 3 months after the approval of the first PARP inhibitor, rucaparib, in mCRPC) were included. Statistical analysis was performed from September 2024 to May 2025.ExposuresAge, race and ethnicity, insurance status, and practice type at the time of mCRPC diagnosis.Main Outcomes and MeasuresThe receipt of PARP inhibitors. Multivariable logistic regression was conducted to assess the association between the exposures and the main outcome.ResultsOf 24 105 patients with metastatic prostate cancer, 443 male patients (median [IQR] age, 72 [65-79] years) had mCRPC with BRCA1/2 alterations and were eligible and included in our analysis. Of these patients, 227 (51.2%) received a PARP inhibitor, whereas 216 (48.8%) did not. Compared with patients covered by a commercial health plan, those covered by Medicare were significantly more likely to receive a PARP inhibitor (odds ratio, 1.91; 95% CI, 1.02-3.66; P = .047). The odds of receiving a PARP inhibitor were not significantly higher in patients treated in community practice compared with those treated in academic centers (odds ratio, 1.64; 95% CI, 1.00-2.70; P = .05).Conclusions and RelevanceThis retrospective cohort study of patients with mCRPC and evidence of BRCA1/2 alterations found that approximately half of patients did not receive PARP inhibitors despite evidence of survival improvement in this population. These findings highlight the need to increase awareness of the survival data and access to life-prolonging therapies in patients with mCRPC.

5064 Background: FoundationOneCDx (F1CDx) supports two FDA-approved biomarkers to guide treatment decisions for ICI for patients with mCRPC: MSI status and TMB. MSI-H and TMB-H (10+ mut/MB) have strongly overlapping prevalence. We sought to better characterize ICI outcome associations of TMB-H / non-MSI-H population and relative effectiveness of taxanes & ICI among patients who received these agents in sequence. Methods: Following a prespecified analysis plan, this study used the nationwide (US-based) de-identified Flatiron Health-Foundation Medicine mCRPC clinico-genomic database (FH-FMI CGDB), with data originating from ~280 US cancer clinics (~800 sites of care). Inclusion criteria included patients with mCRPC treated with single-agent anti-PD(L)1 therapy in the FH network between 1/1/2011 - 3/30/2024. This study used the MSI and TMB algorithms from the tissue based F1CDx. Time to next treatment (TTNT) and OS were assessed with Kaplan-Meier plots and in multivariable Cox models adjusted for ECOG performance score, socioeconomic status, prior treatment history, and baseline PSA. Among patients who received taxanes in a prior line of therapy, the effectiveness (TTNT1 vs. TTNT2) of taxane and ICI were compared. Results: Among 2995 prostate cancer tissue specimens in the database, 95 (3.1%) were MSI-H and 142 (4.7%) were TMB-H. 94 (3.1%) were MSI-H & TMB-H, 1 was MSI-H & TMB-L, and 48 (1.6%) were TMB-H & not MSI-H. Among these, 84 patients with mCRPC were treated with ICI and met inclusion criteria, including MSI-H & TMB-H (n = 30), non-MSI-H & TMB-H (n = 8), and non-MSI-H and TMB-L (n = 46). The respective median TTNT on ICI was 8.0 vs. 9.6 vs. 3 months. The respective median OS from initiation of ICI was 10.9 vs. not reached vs. 4.4 months. In multivariable models evaluating ICI only, compared to non-MSI-H & TMB-L, the MSI-H & TMB-H group had more favorable TTNT (HR: 0.20, 95%CI: 0.10 – 0.41, p < 0.001) and OS (HR: 0.33, 95%CI: 0.16 – 0.70, p = 0.004), and the non-MSI-H & TMB-H group also had more favorable TTNT (HR: 0.13, 95%CI: 0.04 – 0.45, p = 0.002) and OS (HR: 0.20, 95%CI: 0.05 – 0.75, p = 0.017). 50 of the 84 (60%) patients treated with ICI had prior mCRPC taxane treatment. Better TTNT2 on subsequent ICI vs. prior taxane was observed for MSI-H (HR: 0.49, 95%CI: 0.23 – 1.01, p = 0.051) and TMB-H (0.54, 95%CI: 0.30 – 0.98, p = 0.044), but the opposite was true for non-MSI-H and TMB-L subgroups, with significant treatment interactions for each (p = 0.0018, p = 0.00052). Conclusions: TMB-H (4.7%) is more prevalent than MSI-H (3.1%) by F1CDx in mCRPC. Non-MSI-H / TMB-H (1.6%) in the routine practice cohort have similar outcome associations on ICI to MSI-H. Both MSI-H and TMB-H by F1CDx are predictive of differential benefit for ICI vs. taxanes in later mCRPC treatment lines.

5076 Background: C and Lu are both life-prolonging therapies for pts with mCRPC after progression on androgen receptor pathway inhibitor (ARPI) and docetaxel (D). Herein, our objective was to assess the comparative effectiveness of C vs. Lu in pts with mCRPC with prior progression on D and ARPI in real-world pts in the USA. Methods: A de-identified nationwide Flatiron Health electronic health record (EHR)-derived database was used to extract pt-level data. Eligibility: pts with mCRPC who progressed on D and ARPI and received for the first-time single-agent C or Lu. Pts who received alternative Lu or C in a later line of therapy (LOT) were excluded. Endpoints: real-world time to next therapy (rwTTNT) and real-world overall survival (rwOS). These were summarized via Kaplan-Meier survival estimates with a 95% confidence interval (CI) and compared in the context of propensity score (PS) matching weighted analysis using the Cox proportional hazard model. PS model included the following covariates: age, race-ethnicity, socioeconomic status, treatment year, Gleason score, ECOG performance status, log2PSA, alkaline phosphatase, hemoglobin, creatinine, LOT, insurance, and practice type. All the covariates achieved balance after PS matching weighting. Results: Among 24,105 pts with metastatic prostate cancer in the dataset, 1,445 met the eligibility criteria and were included (1,227 treated with C, 218 treated with Lu). In C cohort: median age was 73 (IQR 67 – 78), 66.9% had Gleason score ≥ 8, and median LOT was 4 (IQR 3-4). In Lu cohort: median age was 75 (IQR 67.25 – 80), 68.1% had Gleason score ≥ 8, and median LOT was 4 (IQR 3-5). In PS matching weighting analysis, there was evidence that pts receiving Lu had a significantly improved rwTTNT (median 8.3 months [mo], 95% CI 6.3 – 9.8) compared to those receiving C (median 4.7 mo, 95% CI 4.2 – 5.4) (HR 0.49, 95% CI 0.37 – 0.63, p < 0.001), which persisted after adjusting for covariates (HR 0.43, 95% CI 0.32 – 0.57, p < 0.001). Additionally, there was evidence that rwOS was longer in pts receiving Lu (median 10.3, 95% CI 8.9 – 12.8) compared to those receiving C (median 8.8 mo, 95% CI 6.6 – 10.8) (HR 0.69, 95% CI 0.51 – 0.92, p = 0.01). This improvement in rwOS with Lu compared to C persisted after adjusting for covariates (HR 0.61, 95% CI 0.44 – 0.84, p < 0.001). Conclusions: This is the largest real-world data to date assessing the comparative effectiveness of Lu vs. C, and it showed significantly longer rwTTNT and rwOS with Lu compared to C in pts with mCRPC pretreated with ARPI and D. Limitations: retrospective nature, selection bias, missingness, lack of randomization, etc. and residual confounding in real-world datasets. Upon external validation, these findings could guide treatment selection in the clinic and design of clinical trials.

727 Background: Despite the availability of targeted therapies like erdafitinib for pts with aUC with susceptible genomic alterations, the rate of NGS testing remains low at 32% in the United States (Hage Chehade...Swami, JAMA Network Open, 2024). Furthermore, limited data exist regarding the timing of NGS testing in relation to the time of death. Herein, we sought to assess the rates of NGS testing at the end of life in pts with aUC in a real-world patient population. Methods: A de-identified nationwide Flatiron Health electronic health record (EHR)-derived database was used to extract patient-level data. Inclusion criteria: pts who received NGS testing (blood/tissue) and who had a date of death recorded. The analytic cohort included pts diagnosed with aUC between 1/1/2011 and 11/14/2022 who received first NGS testing results between 5/3/2012 and 12/21/2022. The time difference between each pt’s first NGS result delivered and death was calculated, and thereafter, pts were categorized into 3 categories: pts receiving results > 3 months (mo) before death, within 3 mo of death, or after death. The frequency and percentages of the 3 categories were reported, as well as by year of death. The percentages in each category were compared between 2019 (year of erdafitinib approval) and 2022 (last year in dataset) to evaluate changes in NGS testing trend. Results: Among 12,157 pts with aUC in the dataset, 1,637 had both a recorded date of death and underwent NGS testing, and were eligible and included in final analysis. The median age was 72 (IQR 66 – 79). The majority were male (73%), White non-Hispanic (76%), and treated in community practice (86%). 68.2% received NGS results > 3 mo before death (n = 1,117), 28.3% received results within 3 mo of death (n = 464), and 3.4% of NGS results were reported after death (n = 56). In pts who died in 2019 versus 2022, the percentage of pts receiving NGS results > 3 mo before death increased from 61% to 75%, while the percentage of pts receiving results within 3 months decreased from 33% to 22%, and the percentage of NGS results reported after death declined from 5.2 to 2.7%. Baseline characteristics (including race-ethnicity, insurance plan, practice type [academic vs. community]) in these categories will be presented at the meeting. Conclusions: Our findings suggest that a sizeable number of pts with aUC (∼ 32%) received their first NGS testing results towards the end of life, highlighting the need to integrate genomic testing earlier during the course of the disease to optimize benefit with targeted therapies. These results could inform clinical practice and therapeutic decisions in the clinic. CHC, YJ: Equal contribution; US, NA: Senior authors.