Browse Articles
Discover research articles across all indexed journals
Evaluating the rates of surgical overtreatment of prostate cancer.
357 Background: Overtreatment of prostate cancer is a public health concern and a barrier to prostate cancer screening. In 2012 the US Preventative Task Force (USPTF) recommended against PSA screening citing overtreatment of indolent disease, and its associated morbidity, as a major limitation to the safety of widespread screening. In 2018 the USPTF updated their recommendation to PSA screening should only be pursued after an individualized discussion with the patient, emphasizing the risk of overtreatment. We sought to evaluate the rates of surgical overtreatment of prostate cancer over the last two decades. Methods: We evaluated the proportion of radical prostatectomies (RP) with pathologic Grade Group 1 (pGG1), as a surrogate for clinically insignificant disease, between 2012 and 2024 within the Michigan Urological Surgery Improvement Collaborative (MUSIC) and between 2010 and 2020 within the Surveillance Epidemiology and End Results (SEER) registry. MUSIC served as a statewide sample with a focus on urology outcomes, while SEER served as a national sample. Within MUSIC, pGG1, as a binary variable, was modeled in a mixed effects logistic model, against time while controlling for preoperative PSA, age, race, proportion and number of biopsy cores positive, and biopsy grade. We also assessed the proportion of these pGG1 RPs with preoperative higher-risk features: PSA ≥10, >50% of biopsy cores positive, or clinical Grade Group 2 on biopsy. Presence of any of these risk features was modeled as a binary variable against time, controlling for age and race. Results: Among 23,370 RPs in MUSIC and 162,558 RPs in SEER, the proportion of pGG1 decreased over the time periods examined. The proportion of patients undergoing RP with pGG1 on final pathology decreased from 21% to 2.7% from 2012 to 2024 within MUSIC and decreased from 32% to 7.8% from 2010 and 2020 within SEER. Within MUSIC, time was independently associated with a decrease in the proportion of RPs that were pGG1 (OR 0.52 per 5-years, 95% CI 0.47-0.57). During these time periods, the proportion of patients undergoing pGG1 RP with preoperative PSA ≥10 increased from 6.0% to 13% within MUSIC and increased from 8.1% to 10% within SEER. The proportion of pGG1 RPs with >50% cores positive on preoperative biopsy increased from 3.6% to 19% within MUSIC and from 11.8% to 15% within SEER. Time was independently associated with an increase in the proportion of pGG1 RP that had any pre-operative higher-risk features (OR 2.25 per 5 years, 95% CI 2.07-2.46). Conclusions: The rate of surgical overtreatment of low-risk prostate cancer has profoundly declined over the last two decades. A lower proportion of RPs are pGG1 and the pGG1 RPs being performed are more likely to have higher-risk preoperative features. Proportion of prostatectomies that were pGGG1. 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 MUSIC - - 21% 18% 19% 16% 12% 8.4% 8.4% 7.2% 5.5% 5.6% 5.5% 3.7% 2.7% SEER 32% 31% 26% 23% 20% 16% 13% 11% 10% 8.8% 7.8% - - - -
Biomarker analysis in localized urothelial cancer DUART patients post-treatment.
859 Background: Patients with localized unresectable or cisplatin-ineligible urothelial cancer (UC) have limited treatment options. Biomarker identification can guide targeted therapies. In the DUART study, pre-treatment immune cell subsets were significantly linked to disease control. Our planned correlative aim was to evaluate the same biomarkers using post-adjuvant treatment (post-Rx) peripheral blood mononuclear cells (PBMCs). Methods: This was a prospective, multi-institutional study BTCRC-GU15-023. Our N=16 all had valid Post-Rx values and disease control status. Eligibility criteria: >18yrs, advanced/unresectable UC, and available tumor specimen. All received concurrent durvalumab and radiation therapy followed by adjuvant durvalumab in a Phase II study. Blood samples were taken at pretreatment, 12 weeks, and post-Rx. Biomarkers were detected using multicolor flow cytometry-based analysis of PBMCs to detect T lymphocyte subsets and by dimensionality reduction using FlowJo. Correlative objective: Evaluate post-Rx time points for biomarkers that contribute to disease progression or response. Two-sample T-tests were used to study the association. All tests were two-sided and the statistical significance level used was 0.05. Results: Standard flow cytometry analysis revealed a statistically significant increase in ICOS+ CD4 and CD8 T cells in post-Rx samples among patients with progression-free survival at one year. In addition, responder patients (CR/PR/SD, n=12) showed a significant decrease in CD8 central memory T cells compared to progressors (PD, n=4) in post-Rx samples. Although not statistically significant, additional trends were noted, including decreased PD-1+ CD4 T cells in responder patients, decreased CD4 T effector memory RA+ (TEMRA), and increased CD4 naïve T cells in responder patients. There was a slight increase in interferon gamma-producing CD8 T cell subsets in responder patients and a significant decrease in central memory CD8 T cells. tSNE analysis revealed similar trends in the data, including increased naïve CD4 T cells in responder patients and slight increases in some cytokine-producing CD8 T cell subsets. Conclusions: Our small cohort demonstrates some significant differences in post-Rx T cell populations linked to therapy response, and further evaluation in a larger cohort of patients is needed. The identification of predictive biomarkers could help a more personalized therapeutic approach.
Circulating kidney injury molecule-1 (KIM-1) in association with kidney injury biomarkers and outcomes in metastatic renal cell carcinoma.
582 Background: Kidney injury molecule-1 (KIM-1) is overexpressed in clear cell and papillary renal cell carcinoma (RCC) and in proximal tubular kidney injury. While circulating KIM-1 is a minimally invasive biomarker for RCC, it is unknown whether kidney disease, a common comorbidity among RCC patients, impacts the association of KIM-1 with RCC outcomes. We evaluated the association between KIM-1 and outcomes in metastatic RCC after adjustment for multiple kidney injury biomarkers using plasma proteomics. Methods: Plasma samples from patients with metastatic clear cell and papillary RCC were obtained prior to 1 st line systemic therapy. Samples were analyzed using a high-throughput aptamer-based proteomics assay (SomaLogic), and results were log-transformed for analysis. Clinical and laboratory data, as well as cancer outcomes, were retrospectively curated. Spearman’s ρ was used to evaluate correlations between circulating KIM-1 and kidney injury biomarkers (cystatin C, TNFR1, TNFR2, eGFR). Cox regression analyses were used to evaluate the association between KIM-1 as a continuous variable and overall survival (OS) and progression-free survival (PFS), after adjusting for kidney injury biomarkers. Performance of KIM-1 tertiles versus IMDC risk groups for prognosticating OS was evaluated using the C-index. Results: Among 210 patients, higher baseline KIM-1 was associated with worse PFS (p = 0.004) and OS (p < 0.001) in univariate Cox regression analysis (Table). The prognostic value of KIM-1 was consistent across clear cell and papillary RCC (p-value for interaction = 0.98). KIM-1 remained prognostic for PFS (p = 0.01) and OS (p < 0.001) after multivariable adjustment for kidney injury biomarkers and eGFR (Table). Median follow-up was 22.1 months. Kidney injury biomarkers (cystatin C, TNFR1, TNFR2, eGFR) were correlated with each other but not with plasma KIM-1. KIM-1 tertiles (high/medium/low) were more prognostic for OS than the IMDC risk groups (C-index, KIM-1 0.63 vs. IMDC groups 0.58) and the addition of KIM-1 to the IMDC model improved its performance (C-index, KIM-1 + IMDC groups 0.64). Conclusions: Plasma KIM-1 was associated with PFS and OS in metastatic clear cell and papillary RCC. Plasma KIM-1 was not correlated with kidney injury biomarkers, suggesting that at least in metastatic RCC, circulating KIM-1 derives predominantly from tumor rather than benign kidney. The addition of KIM-1 improves IMDC model performance and may be useful for risk prognostication in RCC. Association of KIM-1 with PFS and OS in metastatic RCC. Multivariable models are adjusted for kidney injury markers (cystatin C, TNFR1, TNFR2) and eGFR. log KIM-1 HR (95% CI) p-value OS (univariate) 1.4 (1.2 – 1.7) <0.001*** PFS (univariate) 1.2 (1.1 – 1.4) 0.004** OS (multivariate) 1.4 (1.2 – 1.6) <0.001*** PFS (multivariate) 1.2 (1.1 – 1.3) 0.01**
Efficacy of <sup>177</sup> Lu-PSMA-617 with or without ARPIs for the treatment of mCRPC: VISION secondary analysis.
121 Background: In the Phase 3 VISION trial, treatment with the prostate-specific membrane antigen (PSMA)-targeted radioligand therapy [ 177 Lu]Lu-PSMA-617 ( 177 Lu-PSMA-617) + protocol-permitted standard of care (SoC) significantly improved median radiographic progression-free survival (rPFS; 8.7 vs 3.4 months [mo]) and overall survival (OS; 15.3 vs 11.3 mo) in patients with metastatic castration-resistant prostate cancer (mCRPC) vs SoC alone. However, it is unclear how the addition of androgen receptor pathway inhibitors (ARPIs) affected 177 Lu-PSMA-617 treatment outcomes. Here we compared the efficacy and safety of 177 Lu-PSMA-617 in patients treated with vs without concomitant ARPIs. Methods: In VISION, adult patients with PSMA-positive mCRPC previously treated with ≥1 prior ARPI and 1–2 taxane chemotherapy regimens were randomized 2:1 to receive 177 Lu-PSMA-617 (7.4 GBq Q6W, 4–6 cycles) + SoC (which included ARPIs) vs SoC alone. In this secondary analysis, we assessed baseline characteristics, OS, rPFS, prostate-specific antigen (PSA) response rate, duration of PSA response, PSA-PFS, and safety among patients in the intervention arm who were treated with vs without concomitant ARPIs. Results: In total, 289 patients were treated with concomitant ARPIs and 262 patients were not (N=551). Imbalances between the groups were observed in some baseline characteristics: patients treated with concomitant ARPIs were younger, had lower ECOG performance scores, lower mean PSA and lactate dehydrogenase levels, and fewer prior taxane chemotherapy regimens. Baseline mean standardized [ 68 Ga]Ga-PSMA-11 uptake values and mean hemoglobin levels were well-balanced. A statistically significant difference in median OS was observed in patients treated with concomitant ARPIs vs those without (17.8 vs 12.4 mo; HR, 0.72; 95% CI, 0.58–0.89; nominal P=0.001). No statistically significant between-group differences were observed for rPFS, PSA response rate, duration of PSA response or PSA-PFS. The proportion of adverse events reported was similar in patients with or without concurrent ARPI, except for the number of adverse events leading to interruption of best supportive care/SoC, which was higher in patients treated with concurrent ARPIs. Conclusions: OS was significantly different in patients treated with concomitant ARPIs vs those without, while other efficacy endpoints were not. No new safety signals were observed for 177 Lu-PSMA-617 with concomitant ARPI. These findings may be confounded by varied exposure to ARPIs, or differing patient characteristics; patients who received concomitant ARPIs may represent a more favorable treatment population than those who did not. These findings should be considered as hypothesis-generating, and a multivariate analysis has been planned to ascertain the influence of these potential biases. Clinical trial information: NCT03511664 .
Phase 2 trial of immuno-ablation with intrabladder injection of N-803 or intravesical N-803 plus BCG for intermediate-risk non-muscle invasive papillary bladder cancer.
TPS905 Background: Response rates (55-65%) with bacillus Calmette-Guerin (BCG) monotherapy in papillary non-muscle invasive bladder cancer (NMIBC) are lower than rates (70-75%) for carcinoma in situ (CIS) disease, pointing to an unmet need for efficacious treatment of papillary NMIBC. We previously reported a complete response (CR) rate of 71% with combined intravesical use of the interleukin-15 (IL-15) superagonist fusion molecule N-803 (nogapendekin alfa inbakicept, NAI; ANKTIVA) and BCG in QUILT 3.032 cohort A study participants with BCG-unresponsive bladder CIS +/- Ta/T1 papillary disease and a disease-free survival (DFS) probability of 55.4% at 12 months for cohort B participants with BCG-unresponsive high-grade Ta/T1 papillary NMIBC. Based on these and other findings, the combination of N-803 plus BCG was recently approved by the FDA for BCG-unresponsive bladder CIS +/- Ta/T1 papillary disease. In a planned study, the efficacy of intrabladder N-803 immunoablative monotherapy by peritumoral injection of N-803 will be compared to intravesical N-803 plus BCG in patients with intermediate-risk BCG-naïve papillary NMIBC. The rationale for injection is based on findings from pre-clinical murine tumor models, the hypothesis that N-803 half-life will be extended by injection, and the suitability of injection for papillary disease. Methods: In the open-label phase 2 randomized study QUILT-215, adult participants with histologically-confirmed BCG-naïve intermediate-risk papillary Ta/T1 NMIBC will be enrolled into either Arm A or B. Up to 20 participants will be enrolled in each arm. In the first treatment period, Arm A participants will receive 400 µg N-803 monotherapy every 3 weeks by peritumoral intrabladder delivery and Arm B participants will receive 400 µg N-803 plus 50 mg BCG weekly for 6 weeks by intravesical delivery. The initial response assessment will be at 3 months. The second treatment period would commence at the end of month 3 and continue through month 15, with treatment depending upon the month 3 response. The primary endpoint is the CR (absence of any grade papillary NMIBC or CIS disease) rate at month 3 as determined by Investigator assessment of cystoscopy, cytology, and biopsy. The CR rate will be summarized by the number and percent and exact 95% CI using the Clopper-Pearson method. Secondary endpoints are progression-free survival (PFS), time to disease progression, overall survival (OS), disease-specific survival (DSS), duration of response (DOR) and cystectomy avoidance rate, to be analyzed using Kaplan-Meier methods. Safety will be assessed, including adverse events (AEs) and serious AEs (SAEs).
Assessing global disparities in clinical trial availability for renal cell carcinoma (RCC).
449 Background: Clinical trials are the cornerstone for developing novel therapies and diagnostic tools. Due to advancements from clinical trials, the median overall survival (OS) of patients with metastatic RCC has increased 4-fold in the last 20 years (Lancet 2024). However, clinical trial availability remains disproportionately concentrated in high-income regions, limiting the inclusion of a diverse patient population (JCO CCI 2020). We aimed to assess the current state of the global availability of clinical trials for RCC. Methods: Using the National Clinical Trials database, we identified all clinical trials enrolling patients with RCC from 6/1/2019 to 6/1/2024. We initially screened all trials and excluded those involving non-interventional studies, pediatric patients, and non-RCC neoplasms. Per trial, we annotated which countries had at least one active site. We classified them based on income using the World Bank Ranking (WBR) into high-income (HICs), upper-middle-income (UMICs), lower-middle-income (LMICs), and low-income (LICs) countries. Additionally, we registered information on RCC type, sponsor, phase, cancer stage, and primary and secondary endpoints. We used descriptive statistics to summarize the characteristics of each trial. The association between RCC trial availability and WBR was assessed using the Kruskal-Wallis test. We applied Poisson regression analysis to evaluate the association between clinical trial availability and incidence, mortality, WBR, health expenditure, and gross national income (GNI). Results: Out of 558 identified RCC trials, 357 met the eligibility criteria and 201 were exclude. The median number of countries per trial was 2.6. Most trials were conducted exclusively in HICs (76%), with fewer trials available in UMICs (23%), LMICs (3%), and none in LICs. The majority of trials included patients with clear cell RCC (52%) and metastatic disease (80%). Furthermore, 47% of trials were sponsored by academic institutions and 81% were early-phase trials. We found that the WBR was significantly associated with RCC clinical trial availability (p < .001). UMICs, LMICs, and LICs had significantly lower odds of hosting RCC clinical trials than HICs (OR 0.2, 0.05, and 0.013, respectively). Moreover, trials in non-HICs were more frequently funded by pharma (64% vs 40%), were part of multinational trials (45% vs 16%) and were late-phase trials (25% vs 8%). Poisson regression analysis revealed that GNI, health expenditure, and mortality rates were significantly associated with the number of clinical trials in a country. Conclusions: RCC clinical trials are disproportionately concentrated in HICs, with a direct association between GNI, health expenditures and trial availability. Expanding access to a broader range of trials, including early-phase and academic-sponsored studies, in underserved regions will promote more equitable advancements for all RCC patients.
Transcriptional profiling of patients with metastatic hormone-sensitive prostate cancer to uncover specific signatures linked to the transition from androgen-dependent to androgen-independent phenotype.
227 Background: Metastatic hormone-sensitive prostate cancers (mHSPC) are generally sensitive to androgen deprivation therapy (ADT). However, resistance often occurs, leading to disease progression. Understanding the transcriptomic changes underpinning the shift to antiandrogen resistance is crucial for identification of novel therapeutic vulnerabilities. Methods: Sixty patients with mHSPC undergoing ADT plus androgen receptor pathway inhibitors (ARPI) were enrolled. We defined non-responder (NR) patients (13/60) with biochemical/ radiological progression within 6 months of treatment and responder (R) (47/60) as those with stable disease or partial/complete response and eventually disease progression after 6 months. Total RNA was extracted from archived Formalin-Fixed Paraffin-Embedded (FFPE) basal prostate biopsies tissue samples using Maxwell RSC RNA FFPE Kit, and RNA was profiled using NanoString Tumor Signaling 360 Panel. Results: To identify gene signatures associated with response to ADT + ARPI, we analyzed differentially expressed genes between NR and R patients, through Rosalind platform Gene Set Analysis. Notably, NR patients exhibited significant upregulation of genes linked to cell cycle progression and DNA replication. The mitotic kinases AURKB and PLK1 were the most markedly upregulated genes in NR versus R samples (p=0.02 and p=0.01, respectively), suggesting enhanced cell cycle progression despite antiandrogen therapy. Coherent with these results, hyper-expression of KIF23 and CDC20 in NR vs R samples further confirmed aberrant cell cycle progression. Importantly, NR biopsies also exhibited increased expression of metastasis-promoting genes, such as HGF (p= 6.53 e-3 ), which may influence therapy response. Kaplan-Meier survival analysis reinforced the clinical significance of these findings, revealing that patients with hyper-expression of AURKB, KIF23, or CDC20, had shorter progression-free survival (PFS) compared to those with lower expression levels (p=0.0037, p=0.0007 and p=0.0289, respectively). These results underscore the potential of these molecular markers as predictive tools for resistance to ADT + ARPI and as potential targets for future therapeutic interventions. Conclusions: This study provides new insights into the transcriptomic landscape of mHSPC and identifies key gene signatures responsible for resistance to standard-of-care. Validation in larger cohorts is essential to confirm the predictive value of these genes and to identify patients eligible for novel combinatory treatments aimed at delaying this critical phenomenon.
Utilization, healthcare expenditures, and patient costs of definitive treatment modalities for localized prostate cancer in the United States.
370 Background: Radical prostatectomy (RP) and radiotherapy are standard-of-care treatments for localized prostate cancer. We studied the contemporary trends in utilization and costs of prostatectomy and radiotherapeutic modalities in the U.S. Methods: We analyzed the MarketScan Medicare Supplemental and Coordination of Benefits (MDCR) and Commercial Claims and Encounters (CCAE) Databases (the latter includes patients insured by employer-sponsored plans). We identified men with non-metastatic prostate cancer treated with RP, external beam radiation therapy (EBRT), brachytherapy (BT), EBRT combined with BT (EBRT+BT), stereotactic body radiation therapy (SBRT), or proton beam therapy (PBT) between 2009 and 2021. Year-to-year utilization of each treatment was compared using Kendall’s Tau-b test. Total healthcare and patient out-of-pocket payments within 12 months following treatment initiation were compared among treatment modalities using the Kruskal-Wallis test. Results: In MDCR, we identified 44,937 patients who received treatment with either RP (n=12,879), EBRT (n=26,193), BT (n=926), EBRT+BT (n=4,706), PBT (n=57), or SBRT (n=176). Between 2009 and 2021, the proportion of patients treated with EBRT increased from 52.5% to 62.2% (trend p<0.001), SBRT increased from 0.4% to 0.5% (p<0.001), BT decreased from 3.1% to 1.0% (p<0.001), and EBRT+BT decreased from 14.8% to 6.8% (p<0.001). There was no significant difference in the proportion of patients treated with RP (29.1% vs 29.4%; p=0.82) or PBT (0.1% vs 0.1%; p=0.93). In CCAE, we identified 75,626 patients treated with either RP (n=50,278), EBRT (n=16,985), BT (n=1,243), EBRT+BT (n=6,811), PBT (n=91), or SBRT (n=217). EBRT increased from 20.0% to 24.9% (p<0.001), SBRT increased from 0.1% to 0.8% (p<0.001), BT decreased from 2.5% to 0.7% (p<0.001), and EBRT+BT decreased from 10.6% to 7.4% (p<0.001). There was no significant difference in the proportion of patients treated with RP (66.8% to 66.1%; p= 0.82) or PBT (0.1% vs 0.1%; p=0.76). In both cohorts, PBT had the highest 12-month total and patient costs, while BT had the lowest total cost and SBRT had the lowest patient cost (Table). Conclusions: There is increasing use of EBRT and SBRT in the US. Despite BT being the least costly, its utilization, alone or in combination with EBRT, has declined. Distribution of inflation-adjusted total payments 12 months post diagnosis. Payment MDCR CCAE RP EBRT BT EBRT+ BT PBT SBRT P RP EBRT BT EBRT+ BT PBT SBRT P Total ($) Mean (STD) 30,149 (9,217) 56,199 (19,551) 26,241 (7,920) 38,676 (11,089) 69,719 (17,998) 47,861 (17,377) < 0.001 35,175 (4,800) 66,940 (11,397) 27,377 (3,707) 45,875 (6,620) 102,312 (12,090) 52,544 (8,027) <0.001 Patient ($) Mean (STD) 1,102 (255) 1,285 (294) 1,048 (231) 1,287 (299) 1,044 (237) 992 (325) <0.001 1,965 (671) 1,786 (636) 1,554 (569) 2,031 (695) 1,466 (477) 1,524 (536) <0.001
Presurgical treatment for inferior vena cava tumor thrombus in patients with renal cell carcinoma: A scoping review.
602 Background: To date, there is no evidence-based standard of care for the management of patients with tumor-related thrombus (TT) in the inferior vena cava (IVC) associated with renal cell carcinoma (RCC).The aim of our study is to analyze published papers on presurgical treatment and their effect on reducing the TT length, level and deintensification of surgical approach. Methods: MEDLINE, Embase, and Cochrane Central were systematically searched using PRISM-Scr checklist for studies evaluating presurgical therapies intended to decrease the TT length, and level in RCC patients as well as surgical outcome. Patient were reviewed to have the following outcomes: clinically significant (>20%) reduction in TT length, reduction in TT level by Mayo classification and deintensification of planned surgical intervention. Results: After comprehensive review of 1175 studies by PRISMA-Scr protocol, 21 studies fulfilled prespecified criteria that included 1 phase II clinical trial, 3 case series and 16 retrospective cohort studies. Among 216 patients reviewed, 178 (82.4%) had clear cell RCC, 106 (49.07%) had stage T3b and 53 (24.53%) had nodal involvement. Presurgical regimes included Thyrosine Kinase Inhibitors (TKI) (81%), TKI-Immune check point inhibitor (ICI) combination (6%), ICI monotherapy (3%), Stereotactic radiation (4%) and others (6%). Significant decrease in length in TT was reported in 40.29%, 42.86% and 84.62% of patients treated with TKI, ICI and TKI-ICI combination, respectively. Levels of TT were decreased in 30.32%, 46.15% and 14.29% patients treated with TKI, ICI and TKI-ICI combination, respectively. Among the patients reported to have less invasive surgical approach following presurgical treatment (n=34), 29.41% avoided cardiac bypass. Conclusions: Published studies have reported benefits of presurgical treatment in reducing length and level of TT as well as improving surgical outcome. However, significant heterogeneity exists and prospective studies are required to draw a meaningful conclusion.
Survival outcomes among metastatic castration-resistant prostate cancer patients treated with androgen receptor pathway inhibitor and docetaxel.
57 Background: Real-world evidence on treatment patterns and real-world overall survival (rwOS) are limited in metastatic castration-resistant prostate cancer (mCRPC) patients treated with prior androgen receptor pathway inhibitor (ARPI) and docetaxel treatment. The objective of the study was to describe distribution of ARPI switch or cabazitaxel and their rwOS in patients with mCRPC settings after ARPI and docetaxel treatment. Methods: The study analyzedConcertAI Patient 360 data with a prostate cancer diagnosis in US, derived from electronic health records and claims data between 1990 and 2023. Among the eligible adult mCRPC patients, the study population included patients who subsequently received ARPI switch or cabazitaxel in mCRPC setting. The patient number and distribution (%) by treatments and subgroups were summarized. The study used the Kaplan-Meier method to calculate rwOS (median, months; interquartile range, IQR) and rwOS rate (%) at 12, 24 and 36 months from the start date of ARPI switch or cabazitaxel (Index Date). Results: Among 945 mCRPC patients with prior ARPI and docetaxel treatment (median age: 72 years; white: 79%; baseline PSA ≥ 100 ng/mL: 41%), cabazitaxel was more commonly used (59%) compared to ARPI switch (abiraterone: 22%; enzalutamide: 19%). The rwOS was 13 months, and rwOS rates of the overall patients were 53%, 29% and 21% for at 12, 24 and 36 months, respectively. Patients with liver metastasis had 9 months of median rwOS, while patients without liver metastasis had 14 months. Conclusions: The relatively short OS demonstrated the unmet medical need in mCRPC patients who received multiple treatments including ARPI and taxane-based chemotherapy, warranting the development of more effective and innovative treatments extending patient survival. Overall survival in metastatic castration-resistant prostate cancer patients treated with androgen receptor pathway inhibitor and docetaxel. Total (N=945) Months (IQR) Median OS in total patients (Death, n=770; 81%) 13 (6 - 29) Median OS in subgroup patients by liver metastasis No 14 (6 - 30) Yes 9 (5 - 20) OS Rate % 12 months 53 24 months 29 36 months 21 IQR= interquartile range; OS=overall survival.
Boosting survival in advanced urothelial carcinoma: The power of CBM588 probiotic and pembrolizumab combination.
774 Background: Immune checkpoint inhibitors (ICIs) is now standard cares for locally advanced or metastatic Urothelial Carcinoma (UC). One of the probiotics, CBM588 (MIYAIRI 588), is widely used in clinical practice and regulates the intestinal microbiota. Recently, the relationship between ICIs and probiotics in cancer therapy has attracted attention with prolonged prognosis reported in lung and renal cancers when combined with CBM588. However, there are no reports on the usefulness in UC. Methods: We retrospectively evaluated the effect of pembrolizumab plus CBM588 in patients with unresectable UC treated with pembrolizumab as second-line therapy at our institution, with and without CBM588. The analysis was divided into two groups: those taking CBM588 (n=33) and those not taking CBM588 (n=11) at the time of pembrolizumab initiation. Progression-free survival (PFS) and overall survival (OS) were the primary endpoints. Results: The median follow-up was 12.0 months in the pembrolizumab alone group versus 12.2 months in the CBM588 group. The median PFS was 3.9 months in the pembrolizumab alone group versus 10.2 months in the CBM588 group. On multivariate analysis, the hazard ratio for the CBM588 arm was 0.074. The median OS was 12.2 months in the CBM588 arm versus 12.0 months in the pembrolizumab alone arm. On multivariate analysis, the hazard ratio for the CBM588 arm was 0.105 (95% CI 0.022-0.533, P=0.0056). Conclusions: The combination of CBM588 with pembrolizumab may prolong progression in UC. Although the sample size and follow-up period are limited, the positive outcomes observed emphasize the necessity for further studies to confirm these findings and investigate the underlying causes. Multivariate analysis of PFS and OS. Multivariate Analysis-PFS Multivariate Analysis-OS Variables Groups Hazard ratio 95% CI P-value Hazard ratio 95% CI P-value Treatment PEM reference reference PEM with CBM588 0.074 0.016-0.340 0.0008 ** 0.105 0.022-0.533 0.0056 * Age 0.947 0.901-0.995 0.030 * 0.992 0.944-1.043 0.738 Sex male reference reference female 1.410 0.546-3.643 0.4786 0.717 0.292-1.985 0.513 ECOG-PS 0 - 1 reference reference 2 - 4 0.338 0.136-0.837 0.0191 * 1.067 0.400-2.712 0.894 Tumor location (primary) Lower urinary tract reference reference Upper urinary tract 1.455 0.577-3.662 0.426 0.666 0.228-1.986 0.461 T stage (primary) 0 - 2 reference reference 3 3.634 1.340-9.835 0.011 * 1.210-8.763 0.019 * unknown 7.623 1.291-45.00 0.025 * 6.368 1.069-37.93 0.042 * N stage 0 reference reference 1 0.976 0.380-2.505 0.958 1.328 0.521-3.389 0.552 M stage 0 reference reference 1 1.407 0.298-6.652 0.666 1.579 0.313-7.944 0.579 Lung metastasis 1 0.871 0.297-2.55 0.801 0.497 0.151-1.633 0.249 Liver metastasis 1 5.465 1.516-19.70 0.0094 * 2.889 0.735-11.34 0.129 Bone metastasis 1 1.841 0.608-5.57 0.280 1.447 0.445-4.703 0.539 IrAE 1 0.133 0.03-0.586 0.0077 ** 0.277 0.056-1.359 0.114 *p < 0.05; **p < 0.008.
Validation of new prognostic factors for relapse in patients with clinical stage I seminoma.
625 Background: Patients with clinical stage I (cSI) seminomatous testicular germ cell tumors (GCT) have a relapse risk of 5 - 30% under surveillance after orchiectomy, depending on tumor size and rete testis infiltration. Recently, two studies introduced new prognostic risk models to improve clinical decision-making: the Danish Testicular Cancer database (DaTeCa) and an Individual Patient Data Analysis by the European Association of Urology (EAU) Testicular Cancer Guidelines Panel and Guidelines Office. This study aimed to validate these new prognostic risk models in an independent cohort of cSI seminoma patients. Methods: We included patients with unilateral cSI seminoma (retroperitoneal lymph nodes <10 mm in short-axis diameter), normalized serum tumor markers post-orchiectomy (β-human chorionic gonadotropin (β-hCG) and lactate dehydrogenase (LDH)), and no adjuvant therapy. All patients had at least 12 months of follow-up. Cox regression analysis was applied to evaluate the proposed prognostic factors and relapse probabilities were estimated using the Kaplan-Meier method. Results: Among 139 patients, 25 (18%) relapsed within a median follow-up of 37 months (CI 47.9–63.1). In multivariate analysis, only rete testis infiltration was confirmed as an independent predictor of relapse (p=0.039). The other prognostic factors of DaTeCa (lymphovascular invasion, elevated pre-orchiectomy β-hCG and LDH) and EAU (tumor size and lymphovascular invasion) could not be confirmed. The 5-year relapse risk according to DaTeCa risk groups was consistent with our cohort (no risk factors: 4% vs. 6%; all four risk factors: 67% vs. 62%). Similarly, the EAU classification showed comparable 5-year relapse risks for low (13% vs. 8%) and intermediate (22% vs. 20%) groups. However, the high-risk group showed a greater discrepancy (67% vs. 44%), although our study included only a limited number of high-risk patients (n=6, 4.3%). Conclusions: The risk classification models from DaTeCa and EAU demonstrated good overall reproducibility in our cohort. These models may aid in identifying seminoma patients at high-risk for relapse who might be candidates for adjuvant therapy. However, the proposed prognostic factors were mainly not independently confirmed in the multivariate analysis in our study cohort.
Interpretable machine learning for prostate biopsy: Cohort study.
333 Background: The objective of this study is to develop prediction models and provide diagnostic results for prostate biopsy among patients. Methods: This paper presents an improved tree model, called the Light gradient boosting machine model, for building a prostate cancer (PCa) risk prediction model. The research data used in this study (Waves 2008-2022) are obtained from 1987 patients who underwent transperineal biopsy at West China Hospital. All patients had biopsy indications (high levels of prostate specific antigen (PSA), suspect lesions in magnetic resonance imaging (MRI) suspect lesions). A total of 13 baseline variables were considered as candidate features, and eight machine learning models were compared in the experiment. Shapley additive explanations (SHAP) was used for explaining the prediction model. Results: The Light gradient boosting machine model demonstrated the best performance, achieving an area under the receiver operating characteristic (ROC) curve of 0.76 (CatBoost=0.71, KNN=0.65, LDA=0.61, CART=0.67, SVM=0.64, NB=0.55, XGBoost=0.71). According to the interpretable analysis, the apparent diffusion coefficient (ADC) of MRI was identified as the most important feature in the prediction model. Other important features included prostate volume, red blood cell count, neutrophil count, and platelet count. Conclusions: This study suggests that optimal models show promise in screening out patients at high risk for PCa in prostate biopsies. The decisions regarding prostate biopsies should specifically focus on the ADC of MRI, prostate volume, and blood routine indices such as RBC count, platelet count, and neutrophil count. Baseline characteristic of all patients. Mean (SD) Pca (n=1001) Non-Pca (n=986) P Age (years) 68.68(9.29) 67.99(9.32) 0.901 Blood routine RBC (10 12 /L) 4.66(0.62) 4.65(0.61) 0.56 Hemoglobin (g/L) 141.46(19.72) 141.84(17.21) 0.149 Platelet (10 9 /L) 188.49(64.65) 193.46(69.12) 0.02 WBC (10 9 /L) 6.76(6.36) 6.6(1.95) 0.56 Neutrophil count (10 9 /L) 4.15(1.52) 4.19(1.71) 0.026 Lymphocyte count (10 9 /L) 1.89(5.73) 1.75(0.76) 0.405 Monocyte count (10 9 /L) 0.48(0.16) 0.47(0.17) 0.626 Eosinophil count (10 9 /L) 0.16(0.16) 0.15(0.16) 0.781 Basophil Cell Count (10 9 /L) 0.25(1.21) 0.36(1.47) <0.001 PSA (ng/ml) 65.02(346.16) 11.32(23.01) <0.001 Prostate volume (ml) 43.96(27.66) 51.24(26.85) 0.007 PSAD (ng/ml 2 ) 1.82(12.88) 0.28(0.544) 0.001 ADC 733.13(172.09) 756.93(181.31) 0.004
Serum biomarkers in metastatic castration-resistant prostate cancer (mCRPC) patients receiving [ <sup>177</sup> Lu]Lu-PSMA-617 therapy: Post hoc analysis of a phase II clinical trial.
190 Background: [ 177 Lu]Lu-PSMA-617 ( 177 Lu-PSMA-617) has been recently approved for metastatic castration-resistant prostate cancer (mCRPC). Several clinical trials have assessed the efficacy of 177 Lu-PSMA-617, but only half of the eligible patients seem to obtain a clear benefit from it. Currently, there is a lack of validated biomarkers that could help find the best candidates for this radioligand therapy. Methods: In the IRST185.03, a single-arm monocentric phase II clinical trial, 142 mCRPC patients received 177 Lu-PSMA-617 at a dose of 3.7-5.5 Gbq every 8-12 weeks for a maximum of 4 cycles. Baseline blood samples were collected within 7 days from the first dose of 177 Lu-PSMA-617. Hemoglobin (Hb), alkaline phosphatase (ALP), prostate-specific antigen (PSA), carcinoembryonic Antigen (CEA), chromogranin A (CgA), and neuron-specific Enolase (NSE), as well as neutrophil-to-lymphocyte ratio (NLR) and platelet-to-lymphocyte ratio (PLR), were analyzed at both uni- and multivariate levels to evaluate their correlation to progression-free survival (PFS). Results: At the time of data cut-off, after a median follow-up time of 44.0 months, the median PFS in the global cohort was 7.0 months. At univariate analysis, among serum biomarkers, lower PLR (cut-off: 210, HR: 0.66, 95% CI: 0.44-0.99, p=0.04), lower ALP (cut-off: upper limit of normal, HR: 0.51, 95% CI: 0.34-0.75, p<0.01), lower PSA (cut-off: median value, HR: 0.51, 95% CI: 0.35-0.74, p<0.01), and lower CEA (cut-off: upper limit of normal, HR: 0.55, 95% CI: 0.32-0.95, p=0.03) were associated with longer PFS. Among clinical characteristics, the previous use of docetaxel (HR: 1.48, 95% CI: 1.04-2.10, p=0.03) was associated with shorter PFS. At multivariate analysis, only PSA and CEA remained prognostic for PFS (p=0.01 and 0.02. respectively). Conclusions: High PSA and CEA levels appeared the only serum predictors of poor clinical outcome to 177 Lu-PSMA-617 therapy in mCRPC. The role of CEA as potential biomarker associated with 177 Lu-PSMA-617 needs to be confirmed since it could be linked to aggressive prostate cancer variant. Clinical trial information: 2016-002732-32 .
Association between vascular patterns of clear cell renal cell carcinoma and immune cell infiltration and therapeutic response.
575 Background: Clear cell renal cell carcinoma (ccRCC) is a highly vascularized tumor with histological heterogenous appearance. We explored whether vascular pattern (VP) are predictive to response of therapy. Methods: We defined three categories of VP in the publicly available TCGA cohort: high-branching (HB), low-branching (LB) and sinusoid. VP-based gene signatures were generated by integrating the transcriptomes of matched ccRCC samples. We developed a learning-based (AI) algorithm for the classification of VP based on CD31 immunohistochemistry and analyzed histology specimens from patients treated with tyrosine kinase inhibitors +/- immunotherapy (TKI) (University Clinic Dresden, retrospective cohort n=38, and NivoSwitch trial NCT: NCT03013946 , n=38 respectively) as well as the publicly available transcriptome datasets from two phase III clinical trials (JAVELIN Renal 101; IMmotion 151). Multiplex immunofluorescence (CODEX) was used for spatial mapping. Outcome measures employed KM-plots and log-rank analyses. Results: We identified a trajectory from a HB to LB vascular phenotype, paralleled by a decline in the expression of proximal tubule cell lineage traits. Applying the VP gene signatures to the transcriptome datasets from JR101 and IM151 we found that the progression-free survival (PFS) benefit from adding immunotherapy to anti-angiogenic therapy (IMmotion151: Bevacizumab; JAVELIN Renal 101: Axitinib) was limited to low-branching ccRCC (IM151: HR (95% CI): 0.64 (0.51-0.80), JR101: HR (95% CI): 0.45 (0.33-0.61, both p < 0.001), respectively) and linked to an immune cell infiltrated microenvironment. Patients with a HB ccRCC demonstrated prolonged PFS in the Dresden TKI cohort (p = 0.026) and in the NivoSwitch trial (p = 0.01). CODEX analysis using the SPACEc pipeline revealed profound differences in cellular neighborhoods, characterized by dense immune cell infiltration of ccRCC with LB compared to HB vascular patterns. Using 30 ccRCC patient-derived organoid in air-liquid interface (ALI) cultures, we confirmed the association between LB pattern and higher T cell infiltration, resulting in reduced viability of LB ccRCC organoids under immune-stimulating conditions. Conclusions: Vascular pattern can reliably predict therapy response in advanced ccRCC. Tumors with high branching phenotype respond better to anti-angiogenic TKI therapy. Low branching pattern was associated with improved response to immunotherapy.
Systemic therapy clinical trial participation in patients with bladder and kidney cancers.
451 Background: Patient participation in clinical trials has led to numerous treatment advances in renal cell carcinoma (RCC) and urothelial carcinoma (UC) over the past decade. The rate of patient participation in RCC and UC trials and factors influencing participation are unknown. This study evaluates patient participation rates in RCC and UC clinical trials at a major United Kingdom cancer centre. Methods: All referrals to St Bartholomew’s Hospital (SBH) Genitourinary Cancer Department between Jan 2020 to Sept 2022 were reviewed. Patients with RCC or UC of any stage were included. Dates of consultation and follow up visits were cross-referenced with a list of systemic therapy clinical trials open at SBH from Jan 2020 to Oct 2024. The proportion of patients who a) had a trial available to them, b) entered trial screening, and c) were eligible for a trial, was determined. Multilevel mixed-effects logistic regression models were used to assess the likelihood of clinical trial screening and enrolment with adjustment for relevant baseline variables (age, cancer type, gender, line of therapy, and performance status [PS]). Results: 403 patients were included in the analysis: 215 RCC (44% stage I-III and 60% had or developed metastatic disease) and 188 UC (41% stage I-III and 69% had or developed metastatic disease). 63% (254/403) of patients had at least one eligibility opportunity to be screened for a trial during the follow up. 40% (161/403) consented to trial screening, and 30% (118/403) were enrolled into at least one trial. The table shows trial availability, screening, and enrolment by line of therapy (rates were similar between RCC and UC, data not shown). Variables associated with increased odds of entering trial screening were line of therapy (second line odds ratio (OR) 8.6 (2.3-31.8), p<0.01, third line OR 3.4 (1.3-9.0) p=0.02, compared to adjuvant) and UC vs RCC trials OR 2.4 (1.3-4.3) p<0.01. Poor PS decreased the odds of entering trial screening (OR 0.21 (0.1-0.5) p<0.01). Gender and age were not associated with screening rates. No variables were associated with trial enrolment after a patient had consented to screening. Conclusions: At a major UK clinical trial centre, 40% of patients with RCC or UC entered clinical trial screening and 30% participated. Most patient characteristics were not associated with increased screening except for PS. Screening rates were higher in later line treatment studies. The effect of ethnicity and randomisation will be presented at the meeting. These data highlight patient willingness to screen for trials when they are available. Clinical trial availability, screening rates, and enrollment rates by line of therapy for patients with RCC and UC. Neo/Adjuvant 1L 2L 3/4L Trial Available 45% (86/248) 58% (151/259) 34% (35/104) 66% (40/60) Screened 56% (48/86) 52% (78/151) 91% (32/35) 83% (33/40) Enrolled 73% (35/48) 68% (53/78) 63% (20/32) 82% (27/33)
Treatment of low-grade intermediate-risk non-muscle-invasive bladder cancer with UGN-102: Results of the phase 3 ATLAS and ENVISION studies.
777 Background: The ATLAS and ENVISION phase 3 studies (NCT04688931/NCT05243550) treated patients with low-grade intermediate-risk non-muscle-invasive bladder cancer (LG-IR-NMIBC) with UGN-102, a reverse thermal hydrogel containing mitomycin. Primary efficacy and safety results were previously reported for the individual studies; here we report additional data. Methods: In the randomized controlled ATLAS study, patients with newly diagnosed or recurrent LG-IR-NMIBC were randomized to 6 weekly intravesical instillations of UGN-102 (n=142) or transurethral resection of bladder tumor (TURBT) (n=140). In the single-arm ENVISION study, patients with recurrent LG-IR-NMIBC received 6 weekly intravesical instillations of UGN-102 (n=240). In both trials, patients were examined for recurrence of bladder cancer using cystoscopy, urine cytology testing, and for-cause biopsy at 3 months, and in those with a complete response (CR), at regular intervals thereafter. Duration of response (DOR) was calculated using the Kaplan–Meier method. Results: In both studies (ATLAS, ENVISION) patients were mostly aged ≥65 years (60%, 68%), white (99%, 98%), and male (70%, 61%). In the ATLAS study, 89 patients (63.6%) in the TURBT arm and 92 patients (64.8%) in the UGN-102 arm had a CR at 3 months. Among patients achieving a CR at 3 months in ATLAS, 79.7% in the UGN-102 arm vs 67.7% in the TURBT arm remained event free 12 months later; the majority with recurrence had low-grade (LG) disease (Table). In ENVISION, 79.6% had a CR at 3 months, and 82.3% remained event free 12 months later, again with the majority of recurrence being LG disease (Table). Median DOR was not estimable in any arm due to low recurrence rates, with a between arms hazard ratio in ATLAS of 0.46 (95% CI 0.24–0.86) favoring the UGN-102 arm. The most common adverse event with UGN-102 in both studies was dysuria, occurring in 30.4% in ATLAS and 22.5% in ENVISION. Conclusions: In both studies, CR rate in patients initially treated with UGN-102 was robust, with the majority of patients remaining event-free at 12 months follow-up. These results demonstrate that treatment with UGN-102 results in a high and clinically meaningful durable CR rate in patients with newly diagnosed or recurrent LG-IR-NMIBC. UGN-102 may represent a valuable non-surgical treatment option for these patients. Clinical trial information: NCT04688931 / NCT05243550 . ATLAS ENVISION UGN-102 TURBT UGN-102 CR at 3 months 92/142 (64.8%) 89/140 (63.6%) 191/240 (79.6%) CR rate (95% CI) 64.8 (56.3–72.6) 63.6 (55.0–71.5) 79.6 (73.9–84.5) Follow-up time for DORMedian (95% CI) 12.45 (12.02–14.13) 12.16 (11.89–12.75) 13.86 (12.19–14.52) Any recurrence 18/92 (19.6%) 24/89 (27.0%) 33/191 (17.3%) LG disease 15/92 (16.3%) 17/89 (19.1%) 27/191 (14.1%) HG disease 3/92 (3.3%) 6/89 (6.7%) 4/191 (2.1%) Death 0 1/89 (1.1%) 2/191 (1.0%) HG, high grade.
Application of canary histology classifier in prostate biopsies for risk stratification.
412 Background: Nguyen et al recently proposed an outcome-based radical prostatectomy (RP) histologic classifier (PMID 38828674, “Canary histology risk group”), combining all high-risk carcinoma histology patterns into “unfavorable histology,” versus absence as “favorable histology.” We examine whether this classifier, with additional stratification of favorable, is applicable to prostate biopsies for risk stratification. Methods: Archival databases of one single institution were searched for patients with long term clinical follow-up after RP and available prostate biopsy slides. Most recent biopsies underwent blinded re-review. Canary risk group was assigned based on highest risk patterns in any core. Unfavorable histology encompasses all Gleason pattern 5, large cribriform/intraductal carcinoma (>0.25 mm), anastomosing cords of carcinoma, grade 3 stromogenic carcinoma, and complex intraluminal papillary architecture. To capture patterns potentially predictive of unsampled unfavorable histology, “biopsy borderline” combines small cribriform/glomeruloid (≤0.25 mm), simple glomerulations, predominance of extensive poorly formed glands, equivocal small anastomosing cords, grade 2 stromal response, and epithelial complexity associated with mucin (beyond mucinous fibroplasia). “Biopsy favorable” is defined as void of any unfavorable or biopsy borderline histology. Biopsy and RP Gleason Grade Group (GG), pT stage, and pN stage were recorded from original reports. Biochemical recurrence (BCR) was defined as consecutive PSA of ≥0.2 ng/mL 8 weeks after RP or receipt of salvage treatment. RP stage and GG were analyzed using chi-square. BCR-free survival was analyzed using Kaplan-Meier curve. Results: 360 patients were identified (median, IQR): Age at diagnosis 61 years (56, 66), PSA at diagnosis 5.38 ng/mL (4.2, 7.6), post-RP follow-up 9 years, (7, 12). The patients were classified as histologically unfavorable (n=63; 17.8%), biopsy borderline (n=92; 25.6%), and biopsy favorable (n=205; 56.9%). The three-tiered classification is significantly associated with earlier BCR (log-rank p-value <.01), RP GG (<.01), pT stage (<.01), and pN stage (<.01) (Table). Conclusions: Unfavorable histology on biopsy was associated with highest risk of BCR after RP, compared to “biopsy borderline” and “biopsy favorable” groups. This proof-of-principle study demonstrates the three-tiered histology-based classifier could be applicable in biopsies for risk stratification. Canary Histology Risk Group N (%) GG3+ at RP pT3a at RP pT3b at RP pN1 BCR event within 7 years of RP Unfavorable 63 (18%) 46 (73%) 30 (48%) 11 (17%) 11 (17%) 53 (84%) Biopsy Borderline 92 (26%) 21 (23%) 32 (35%) 5 (5%) 1 (1%) 60 (63%) Biopsy Favorable 205 (57%) 22 (11%) 69 (34%) 3 (1%) 0 (0%) 110 (54%) All patients 360 (100%) 89 (25%) 131 (36%) 19 (5%) 12 (3%) 223 (62%)
A prospective, single-center, single-arm clinical study of adjuvant toripalimab combined with axitinib in non-clear renal cell carcinoma (nccRCC) patients with high-risk recurrence factors (IUNU-RC-102).
TPS613 Background: Non-clear cell renal cell carcinoma (nccRCC) accounts for approximately 25–30% of all renal cell carcinoma (RCC) cases and is associated with a poor prognosis. While early-stage treatment typically involves surgery, there is a significant risk of postoperative recurrence. Once recurrence or distant metastasis occurs, management becomes more challenging. Current clinical trial results for metastatic advanced nccRCC have shown variability, and no standard treatment protocol has been established. For advanced clear cell renal cell carcinoma (ccRCC), however, the combination of immunotherapy and targeted therapy has become the standard of care. The RENOTORCH study demonstrated the efficacy and safety of toripalimab combined with axitinib as a first-line treatment for advanced ccRCC[1]. This study aims to investigate the efficacy and safety of toripalimab combined with axitinib as adjuvant therapy in postoperative patients with high-risk nccRCC. Methods: IUNU-RC-102 (NCT05768464) is a single-center, prospective, single-arm clinical trial. The study plans to enroll 30 patients. Key inclusion criteria include: a clinical diagnosis of high-risk recurrent nccRCC following nephrectomy, age between 18 and 75 years, an ECOG performance status of 0–1, no prior systemic therapy for renal cancer, and confirmed pathological histology of nccRCC following partial or radical nephrectomy, with exclusions for clear cell carcinoma, chromophobe carcinoma, and eosinophilic renal cell carcinoma. All participants will receive axitinib (5 mg twice daily) for 4 weeks, along with toripalimab (240 mg intravenously on the first day of each 21-day cycle). The treatment duration will extend up to 52 cycles. The study follows Simon’s minimax two-stage design. In the first stage, three patients will be enrolled and will receive two treatment cycles to assess safety. If dose-limiting toxicity (DLT) occurs in one or more of the three patients, the study will be paused. Investigators will then determine whether to terminate or amend the study protocol after further deliberation. In the second stage, the primary endpoint will be 2-year disease-free survival (DFS). Secondary endpoints will include 3-year overall survival (OS), disease recurrence-specific survival (DRSS), and safety. The first participant was enrolled in IUNU-RC-102 in October 2023. We anticipate that the combination of toripalimab and axitinib will offer a promising new treatment option for these patients. Clinical trial information: NCT05768464 .
Automated risk stratification in localized prostate cancer using an AI-assisted framework.
341 Background: Specialist clinicians (oncologists, urologists, radiation oncologists) read through free text reports of prostate MRI and biopsy reports to risk stratify patients during initial consultation of localized PCa. This laborious and time-consuming process is repeated for each day during the clinic. Herein, we propose a hybrid stratification framework utilizing a combination of large language model (LLM) and rule-based programming to automate this process. Methods: This retrospective study included patients with localized PCa (2004-2024) presenting to care at Mayo Clinic with at least one positive prostate biopsy and MRI report available. The state-of-the-art LLM – GPT4 – was utilized using a structured zeroshot prompt to extract relevant phenotypic variables (clinical T stage, Gleason patterns, grade group, extent of specimen/cores involved, prostate size/volume) from unstructured MRI and biopsy reports. Using a rule-based algorithm, the extracted variables and PSA at PCa diagnosis were used to categorize each patient into one of the six NCCN risk categories (very low [VL], low [L], favorable intermediate [FInt], unfavorable intermediate [UFInt], high risk (HR), very high risk [VHR]). Prompts were iteratively developed using ~5% of the total dataset and validated on ~10% of the dataset. Final performance was assessed against a held-out expert annotated test dataset using evaluation metrics (accuracy, F1-score, precision, recall). Additionally, manual annotation was conducted by two independent novice reviewers to compare machine annotated and novice-human annotated risk categorization. Results: A total of 397 patients were included in the evaluation. The median age at diagnosis was 64.8 (IQR: 59.7-68.5); majority of the men were White (n: 366; 92%) and non-Hispanic (n: 371; 94%). The most prevalent risk category was UFint (n: 151; 38%) followed by FInt (n: 80; 20.2%), HR (n: 76; 19.1%), VHR (n: 64; 16.1%), L (21: 5.3%), and VL (n: 5; 1.3%). Novice reviewers achieved an accuracy of 79%, precision of 80%, recall of 82% and a F1-score of 0.81. The hybrid stratification framework using GPT4 achieved an accuracy of 89%, precision of 89%, recall of 88% and a F1-score of 0.88. Among 43 errors by GPT4 risk stratification, majority of the errors were due to miscategorization of clinical T-stage (n: 21; 49%) followed by erroneous number of positive cores (n: 13; 30%). However, additional evaluation showed that GPT4 achieved numerically higher performance (accuracy: 90%; precision: 91%; recall: 90%; F1: 0.90) compared to novice reviewers (accuracy: 77%; precision: 89%; recall: 77%; F1: 0.82) for ascertaining clinical T-stage. Conclusions: Large language models exhibited superior performance than novice clinicians for risk stratification in patients with localized PCa. As the next step, we intend to validate our framework for automated risk classification in a prospective manner.