Citations

59 Background: There are several real-world data sources for prostate cancer (PC), each with its own limitations; the most notable limitation is an incomplete picture of the multidisciplinary clinical management of PC. The PRECISION data platform, the largest, most comprehensive, and continually updated real-world evidence source on patients with advanced PC to date, tries to address this limitation and shed light on PC management in both urology and oncology practices in both academic and community settings in the US. The platform includes a wide variety of PC-specific data variables among patients with metastatic hormone-sensitive PC (mHSPC) and metastatic castration-resistant PC (mCRPC) treated across the US. This study aims to provide an overview of the characteristics, treatment patterns, and clinical outcomes among patients with mHSPC and mCRPC included in PRECISION to date. Methods: A retrospective cohort study that included patients with a diagnosis of mHSPC or mCRPC (index date) in PRECISION between 2018 and 2024 (cohorts could overlap). For each cohort, patient characteristics and treatment patterns were evaluated descriptively. Results: The PRECISION platform currently includes 72,855 and 33,538 patients with mHSPC and mCRPC, respectively. Among the mHSPC cohort, the mean ± standard deviation (SD) age was 74.5±9.4 years; 67% and 33% were treated at oncology and urology centers, respectively; 40% and 60% were from academic and community settings, respectively. The average ± SD time from mHSPC to mCRPC was 11.2±11.7 months. Overall, 77.6% of patients were treated with androgen-deprivation therapy (ADT), of whom 64.7% had ADT combined with androgen receptor pathway inhibitors (ARPIs) including abiraterone acetate, enzalutamide, and apalutamide. Among the mCRPC cohort, the mean ± SD age was 74.1±9.3 years; similar treatment setting patterns as mHSPC were observed. Overall, 74.6% initiated a first-line (1L) therapy on/after mCRPC diagnosis; 52.1% had 1L ARPIs, 16.2% 1L immunotherapy, and 4.6% 1L taxane chemotherapy. In addition, 2.2% of patients received the recently US FDA-approved therapy, lutetium-177 vipivotide tetraxetan, at some point during their mCRPC disease duration. Conclusions: This analysis provides key insights into the patients included in the PRECISION data platform. This multidisciplinary dataset provides a 70/30 split between oncology and urology centers, and a 40/60 split between academic and community settings. The observed clinical profile, treatment patterns, and outcomes comprise the most comprehensive representation of current treatment practice in the US. Future analyses of clinical outcomes with existing and novel therapies could contribute significantly to our understanding of opportunities to improve the treatment of patients with mHSPC and mCRPC in the US.

56 Background: The broad therapeutic arsenal for mCRPC, with overlapping indications, makes selecting the optimal treatment sequence challenging in clinical practice. The aim of this study was to evaluate the prostate-specific antigen (PSA) responses and overall survival (OS) among patients with mCRPC receiving 177 Lu-PSMA-617 based on prior ARPI and taxane treatment exposure. Methods: This retrospective, observational study used real-world data from the PRECISION data platform, a harmonized dataset of advanced prostate cancer patients in the US from diverse clinical settings. Patients with mCRPC who had received ≥1 dose of 177 Lu-PSMA-617 between March 23, 2022 and July 31, 2024, after prior treatment with ≥1 ARPI and ≥1 taxane, were included. To assess PSA response, the baseline PSA value obtained within 90 days before starting 177 Lu-PSMA-617 was compared with the on-treatment PSA value obtained ≥28 days after 177 Lu-PSMA-617 initiation. The number of patients achieving PSA reductions ≥50% (PSA50) and ≥80% (PSA80) were estimated, with patients classified into two cohorts based on prior ARPI and taxane exposure: (1) prompt cohort: after only 1 ARPI and 1 taxane, and (2) deferred cohort: >1 ARPI and/or >1 taxane. Patient characteristics and treatment patterns were evaluated descriptively. OS from 177 Lu-PSMA-617 initiation was estimated using Kaplan–Meier methodology. Results: A total of 431 patients were included: 177 patients in the prompt cohort and 254 patients in the deferred cohort. Of patients in the deferred cohort, 185 (72.8%) received ≥1 ARPI and 1 taxane, 32 (12.6%) received 1 ARPI and ≥1 taxane, and 37 (14.6%) received ≥1 ARPI and ≥1 taxane. Median (interquartile range age was 73 (67–78) years and most demographic characteristics were balanced between groups. Patients in the deferred cohort had higher median PSA levels at baseline (63.1 ng/mL vs 25.2 ng/mL). PSA50 was observed in 18/33 (55%) patients in the prompt cohort vs 28/66 (42%) in the deferred cohort. PSA80 was observed in 13/33 (39%) patients in the prompt cohort vs 21/66 (31%) in the deferred cohort. Across both cohorts, median OS was 18.0 months (95% confidence interval [CI]: 15.1 months–not reached [NR]). Median OS was NR in patients in the prompt cohort vs 15.1 months in patients in the deferred cohort (95% CI: 11.6 months–NR). Conclusions: In this large real-world study, prompt initiation of 177 Lu-PSMA-617 (after only 1 ARPI and 1 taxane) showed superior clinical outcomes versus deferred initiation of 177 Lu-PSMA-617. Patients were more likely to achieve a PSA reduction ≥50% and ≥80% and had longer OS with prompt initiation of 177 Lu-PSMA-617. These results may help to guide treatment sequencing in clinical practice.

410 Background: Inflammasomes are multiprotein complexes that regulate inflammation-associated signaling pathways. Although inflammation plays a crucial role in cancer cell proliferation, the specific role of inflammasomes in prostate cancer (PCa) remains underexplored. This study aims to elucidate the expression of inflammasome-related genes in PCa and assess their association with clinical outcomes. Methods: De-identified transcriptome data from a cohort of 52,266 radical prostatectomy (RP) samples tested (2016-2024) with the Decipher prostate genomic classifier (Veracyte, San Diego, CA) were retrieved from the GRID registry (NCT02609269). Expression analysis focused on 33 genes involved in inflammatory pathways. Outcomes analysis was conducted on a retrospective cohort of 855 patients treated with RP (META855), using Cox regression analysis, adjusting for baseline pathological characteristics. Results: Analysis of baseline gene expression in the GRID RP cohort revealed that most genes exhibit low baseline expression, whereas HSP90AB1, APP, TXN, and TXNIP demonstrate strong expression signals. Higher expression of most genes was associated with higher rate of Gleason grade group 4 or 5 in the GRID RP cohort. On survival analysis of the META855 cohort, higher expression (top 25%) of AIM2 and HSP90AB1 were associated with worse metastasis-free survival (p<0.05 for both). Conversely, both high and low expression levels of NLRP3 were associated with better metastasis-free survival outcomes following RP compared to average expression (p<0.05). On multivariable Cox regression analysis, adjusting for grade group, seminal vesicle involvement, lymph node involvement, and margin status, higher expression of AIM1 (HR 1.75) and HSP90AB1 (HR 1.60) were significantly associated with shorter time to metastasis following RP (p<0.05 for both). Conclusions: There is a molecular heterogeneity within pro-inflammatory genes among patients with PCa.Our findings showed thathigher expression of HSP90AB1 is linked to poorer oncological outcomes, whereas both high and low expression levels of NLRP3 are associated with better outcomes following RP.Further refinement is required to build a robust signature, along with external validation of these findings in other cohorts.

95 Background: Most pts with prostate cancer with bone metastasis experience symptomatic skeletal events (SSEs) [PMID: 23884473], leading to increased morbidity and mortality. BMAs have been approved since 2007 and are recommended in mCRPC to prevent SSEs in pts with bone metastases or a high risk of osteoporotic fracture. Herein, we sought to assess the real-world trends of BMA administration in pts with mCRPC. Methods: We used the de-identified nationwide Flatiron Health electronic health record (EHR)-derived database to extract pt-level data. Inclusion criteria: pts with mCRPC with available date of diagnosis of mCRPC and information on receipt of BMA. Pts were categorized into 2 cohorts based on BMA receipt (e.g., bisphosphonates, denosumab, teriparatide, romosozumab) or not. Treatment trends of BMA by year of mCRPC diagnosis were summarized using frequency and percentages. Baseline characteristics at the time of mCRPC diagnosis (age, race-ethnicity, practice type, insurance) were compared between cohorts using the Wilcoxon rank-sum and Chi-squared tests. Results: Of 24,105 pts with metastatic prostate cancer in the dataset, 14,112 with mCRPC were eligible and included. 7,990 (56.6%) received BMA while 6,122 (43.4%) did not. In BMA receipt cohort: median age was 75 (IQR 67 – 81), majority were White (61%), treated in community practice (87%) and had commercial health plan (80%). In BMA non receipt cohort: median age was 75 (IQR 68 – 82), majority were White (61%), treated in a community practice (78%), and had commercial health plan (76%). The use of BMA decreased over time from 56% in 2013 to 46% in 2024 (Table). Statistically significant differences existed between cohorts in race-ethnicity, practice type, and insurance (p < 0.001) and will be presented at the meeting. Conclusions: Despite recommendations to use BMAs in pts with mCRPC and bone metastasis or high risk of osteoporotic fracture, our findings reveal a low utilization rate of BMAs in mCRPC setting with a notable decline over time. This highlights the need to incorporate the use of BMAs in clinical practice and improve access to these agents. Trends by year of mCRPC diagnosis in receipt of the first BMA and frequency and percentage of each agent. Year 2013N = 549 2014N = 1035 2015N = 1223 2016N = 1345 2017N = 1441 2018N = 1346 2019N = 1469 2020N = 1380 2021N = 1352 2022N = 1393 2023N = 1196 2024N = 383 Receipt of BMA, n (%) 316 (56) 674 (65) 792 (65) 837 (62) 886 (61) 766 (57) 816 (56) 766 (56) 692 (51) 713 (51) 556 (46) 176 (46) Denosumab, n (%) 190 (60.1) 450 (66.8) 557 (70.3) 599 (71.6) 606 (68.4) 533 (70) 526 (64.5) 488 (63.7) 410 (59.2) 414 (58.1) 290 (52.2) 94 (53) Zoledronic acid, n (%) 120 (38) 220 (32.6) 228 (28.7) 233 (27.8) 270 (30.5) 225 (29) 281 (34.4) 273 (35.6) 277 (40) 295 (41.4) 258 (46.4) 82 (47) Teriparatide, other bisphosphonates, n (%) 6 (1.9) 4 (0.6) 7 (1) 5 (0.6) 10 (1.1) 8 (1) 9 (1.1) 5 (0.7) 5 (0.7) 4 (0.5) 8 (1.4) 0

90 Background: Poly(ADP) ribose polymerase inhibitors and pembrolizumab are approved for patients with metastatic prostate cancer (mPC) with susceptible genomic alterations based on NGS testing. We recently reported that the rate of NGS testing in mPC remains low at 29.3% in the United States (1). Herein, we investigated the rates of NGS testing at the end of life in a real-world patient population. Methods: This study utilized the nationwide Flatiron Health electronic health record (EHR)-derived de-identified database. Eligibility: diagnosis of mPC, with information on receipt of NGS testing (blood/tissue) and recorded date of death. The time between each patient’s first NGS testing result date and date of death was measured and categorized into 3 groups: NGS test results delivered more than 3 months (mo) before death, within 3 mo of death, and delivered after death. Frequencies and percentages of these 3 categories were reported, also by the year of death. Results: Out of 24,105 patients with mPC in the database, 3,397 had both a recorded date of death and underwent NGS testing and were eligible and included. The median age was 74 (IQR 67–80), 70% were White non-Hispanic, and 83% were treated in a community-based practice. In these patients, 2,901 (85.4%) NGS results were reported more than 3 mo before death, 457 (13.5%) within 3 mo of death, and 39 (1.1%) after death. From 2015 to 2024, the percentage of patients receiving their result more than 3 mo before death increased from 27% to 91%, while those receiving it within 3 mo before death decreased from 55% to 8.5%, and the percentage of NGS results reported after death declined from 18% to 0.6%. Baseline characteristics (including race-ethnicity, insurance plan, practice type [academic vs. community]) in these categories will be presented at the meeting. Conclusions: Despite the availability of life-prolonging targeted therapies based on NGS testing results for patients with mPC, in ∼14% of patients, the first NGS testing results were reported within 3 mo before death or after death. The lack of timely NGS testing limits patients' access to life-prolonging personalized treatments in earlier settings, resulting in missed opportunities for improved outcomes. 1. Hage Chehade, Swami, JAMA Network Open, 2024.

866 Background: Emerging data suggests that Language Models (LLMs) provide an unprecedented opportunity to extract critical data from the unstructured pathology reports. However, the performance of state of art (SOTA) models on clinically meaningful tasks such as assessment of pathologic responses in patients receiving neoadjuvant chemotherapy (NAC) in muscle invasive bladder cancer(MIBC) is not well known. Methods: This retrospective cohort included patients with pathologically confirmed Muscle-Invasive Bladder Cancer from the year 2018 to 2024. Selected patients included those who underwent Neoadjuvant chemotherapy followed by definitive cystectomy at Mayo Clinic. Gold standard labels were manually curated by trained clinicians. The state of the art (SOTA) LLM – GPT4 – was utilized using a structured zero-shot prompt to extract relevant phenotypic variables such as tumor site, tumor histology and additional histological variants if any, presence of in-situ / invasive carcinoma, presence of lamina propria / muscularis propria / lympho-vascular / perineural invasion, and T-stage from unstructured pathology reports. Structured prompts were iteratively developed for each data variable and validated using ~5% of the total dataset. The extracted variables were manually compared and labelled as True positive / True negative / False positive / False negative. Final performance was assessed against a held-out expert annotated test dataset using evaluation metric (accuracy). Results: A total of 200 reports from 99 patients were extracted by the LLM. Of the 200 reports, 41 duplicate / other pathology reports (e.g. – autopsy, cholecystectomy, etc.) were removed during final assessment. Significant characteristics such as tumor site, histology, and additional histology variant had an accuracy of 89%, 87%, and 91% respectively. Notably, accuracy of detection of in-situ versus invasive carcinoma was 70% versus 91%. Specific variables such as lamina propria / muscularis propria invasion / T-stage / lympho-vascular invasion / perineural invasion were answered with the accuracy of 75%, 75%, 67%, 90%, and 94% respectively. Conclusions: LLMs have the potential to extract critical data from unstructured pathology reports and automate assessment of pathologic responses in MIBC patients receiving NAC. Iterative improvements in performance can be achieved by improving prompt design using expert clinical input. This process can be potentially scaled to automate cancer registries.

831 Background: Long non-coding RNA (lncRNA)-based genomic profiling has suggested utility to identify a distinct tumor subgroup corresponding to a favorable prognosis in patients with bladder cancer. Here, we further evaluate a genomic classifier in a cohort of patients undergoing radical cystectomy (RC). Methods: Transcriptome-wide expression profiling using Decipher Bladder was performed on TURBT samples from a cohort of patients with high grade, clinically organ confined (cTa-T2N0M0) UC who subsequently underwent RC without any neoadjuvant therapy (n=226). LncRNA-based luminal favorable status was determined using a previously developed genomic classifier. The primary endpoint was overall survival (OS) following surgery. Secondary endpoints included cancer-specific mortality and upstaging at RC. Results: In the study, 134 patients were clinical NMIBC (cTa/Tis/T1) and 92 patients were cT2. We identified 60 patients with luminal favorable subtype, all of which showed robust gene expression patterns associated with less aggressive bladder cancer biology. On MVA, patients with the luminal favorable subtype (vs. without) were significantly associated with lower odds of upstaging to pT3+ disease (OR [95% CI] 0.32 [0.12-0.82], p = 0.02), any upstaging (OR [95% CI] 0.41 [0.20-0.83], p = 0.01), and any upstaging and/or pN+ (OR [95% CI] 0.50 [0.25-1.00], p = 0.05). Luminal favorable bladder cancer was significantly associated with better OS (HR 0.33 [95% CI 0.15-0.74], p=0.007). Conclusions: This study validates the performance of the genomic classifier for identifying urothelial carcinomas with a luminal favorable subtype, harboring less aggressive tumor biology.

428 Background: Prostate cancer management presents a significant healthcare burden, with the need to efficiently triage patients for treatment. Our objective is to leverage large language models to predict physician-recommended treatment plans from unstructured clinical notes. By accurately predicting treatment plans, we aim to risk stratify and triage patients effectively, thereby optimizing the allocation of physician resources. Methods: 448 unstructured initial urology consultation patient notes following first positive prostate cancer biopsy were identified. The recommended and final treatments received were manually annotated to establish ground truth labels (Table 1). The dataset was split 80:20 for training and testing, preprocessed to remove plan sections and formatted into question-answer (QA) format. A domain-specific large language model (LLM) inspired by GPT and a specialized tokenizer (PCa- LLM) for prostate cancer terminology were developed. QA models were built using the PCa-LLM and compared with those using GPT-2 as the backbone to predict recommended and final treatments. Results: For the physician-recommended treatment plans, our LLM (PCa-LLM) showed superior performance with higher AUROC scores for curative vs. non-curative treatments (0.78 vs. 0.65), chemo-hormonal vs. other non-curative treatments (0.89 vs. 0.65), and surveillance vs. all other treatments (0.72 vs. 0.70), while both models achieved the same high AUROC of 0.99 for chemo-hormonal vs. all other treatments. For final treatments, PCa-LLM demonstrated better AUROC for curative vs. non-curative treatments (0.77 vs. 0.74) and chemo-hormonal vs. other non-curative treatments (0.71 vs. 0.66), while GPT2 outperformed PCa-LLM for surveillance vs. all other treatments (0.78 vs. 0.70). Both models achieved an AUROC of 0.99 for chemo-hormonal vs. all other treatments. Conclusions: PCa-LLM accurately predicted most treatment categories better than GPT2, with higher AUROC scores, and can be utilized to triage prostate cancer patients using initial consultation notes. Task Physician -Recommended Treatment Plan Final Treatment Received Curative Prostatectomy/Radiation 228 230 Non-Curative Focal Therapy 22 15 Active Surveillance 40 45 Chemo-hormonal 30 30 Model Predictions (AUROC) GPT2 PCa-LLM GPT2 PCa-LLM Curative vs. non-curative 0.65 0.78 0.74 0.77 Chemohormonal vs. other non-curative 0.65 0.89 0.66 0.71 Chemohormonal vs. all other 0.99 0.99 0.99 0.99 Surveillance vs. other non-curative 0.64 0.60 0.67 0.59 Surveillance vs. all other 0.70 0.72 0.78 0.70

425 Background: Prostate cancer (PC) ranks among the most prevalent cancers in men, with incidence rising due to factors like aging populations and enhanced screening techniques. Effective data extraction from clinical records is crucial for advancing research and improving patient care; however, traditional manual methods can be time-consuming and susceptible to mistakes. This study seeks to evaluate RadOnc-GPT against human annotation for extracting essential clinical information from prostate cancer clinical records. Methods: RadOnc-GPT is a retrieval augmented generation (RAG)-based chatbot with access to a variety of patient data including treatment details, radiology reports, and clinical notes. RadOnc-GPT uses Open AI’s gpt-4o model. Patients who received curative intent radiotherapy as initial treatment for PC in our institution between 2014 and 2023 were randomly selected for evaluation. Relevant Clinical factors were extracted by a human expert annotator on our on-site RedCap database. The extraction of the same variables was performed using RadOnc-GPT. A comparison between the ground truth (human annotation) and RadOnc-GPT, to evaluate the performance of the AI system, was carried out including accuracy, recall and F1 score. Results: A total of 382 PC patients were included in the cohort for data extraction. The clinical variables extracted were the treatment intent, NCCN risk category, T and N stage, Gleason score, radiation volume, and Androgen Deprivation therapy (ADT) use. The overall performance of RadOnc-GPT was excellent, with the highest accuracy achieved for treatment intent (99.7%), Gleason score (98.4%), and Treatment Volume (90.1-99.5%). Conclusions: RadOnc-GPT performed excellent at extracting relevant clinical details from the electronic medical record and radiation oncology information systems with dramatic reduction in time and cost compared with manual review. LLMs could be highly useful for large-scale data extraction in PC research, offering a promising solution for automating the accurate data extraction process. Correct (n) Total (n) Accuracy Recall F1 score RT with curative intent 381 382 99.7 99.9 99.9 Prostate Cancer Risk Category 345 382 90.3 90.4 95.1 Primary tumor staging (T) at diagnosis 368 382 96.3 96.3 97.9 Regional nodal status (N) at diagnosis 352 382 92.1 92.2 95.9 Gleason Score Primary Secondary Total 376372376 382382382 98.497.498.4 98.597.498.5 99.298.799.2 Treated Areas Prostate Pelvic Nodes Para-Aortic Nodes Distant Site 380344378372 382382382382 99.590.19997.4 99.490.199.197.4 99.79599.598.7 ADT Use with RT 365 382 95.5 95.7 97.8

424 Background: Advances in prostate cancer (PC) care have led to a 98% 10-year survival rate. Prostate Specific Antigen (PSA) is the primary method of follow-up after definitive radiotherapy (RT), with benign PSA ‘bounce’ occurring in up to 30% of patients, causing uncertainty and anxiety. Distinguishing between PSA bounce and biochemical recurrence (BCR) remains challenging. The Phoenix Criteria is the standard definition of BCR since 2006, defined as PSA nadir + 2 ng/dl, but does not consider patient specific information. Our study aims to explore factors associated with bounce and BCR and develop a predictive tool to differentiate between these two events. Methods: PC patients who received RT at our institution between 2006 and 2022 were identified, including only patients with a PSA increase during follow-up. Exclusion criteria included palliative RT, prostatectomy, or insufficient follow-up information. Patients’ clinical data was collected, and PSA increase was classified as bounce or BCR. Variables were compared using Kruskal Wallis or Chi-square. Multivariate logistic regression was performed in the subset of patients with absolute PSA value >2 (n= 581). Results: Of 1783 patients reviewed due to PSA rise after treatment, 694 met the inclusion criteria. There were 213 (30.7%) with PSA bounce and 481 (69.3%) with recurrence. Prior to meeting the Phoenix Criteria, 20 (4.2%) had biopsy-proven recurrence and 18 (3.7%) imaging-proven recurrence. The median delta PSA was 0.8 ng/mL for bounce vs 5 ng/mL for recurrence (p<0.001). The median time from treatment to event was 14.4 months for bounce vs 45.3 months for BCR (p<0.001). Recurrence was significantly associated (p<0.001) with older age at diagnosis, N1 stage, higher risk category, Gleason score, and higher baseline PSA. In the subgroup of patients without ADT, recurrence was associated with a lower median first PSA post-RT (1.9 vs. 2.4, p=0.046) and a lower median nadir before event (0.6 vs. 1.8, p<0.001). Multivariate logistic regression achieved an accuracy of 0.94, with a sensitivity of 0.83, specificity of 0.98, and an area under the curve (AUC) of 0.97. Conclusions: Our results highlight significant differences in clinical factors between bounce and recurrence in post-RT patients. The multivariate logistic regression differentiated the events with 94% accuracy for patients with a rising PSA value >2 subgroup. We highlight the possible improvement that could be accomplished by personalizing a PSA threshold for recurrence versus the Phoenix definition. This tool offers patients and providers a highly specific tool to differentiate between PSA recurrence and bounce, and may reduce stress and unnecessary testing, or help diagnosis recurrence faster. Next steps include prospective validation of the model to assess the impact on clinical practice costs, patient experience, and cancer outcomes.