BACKGROUND AND OBJECTIVE:Biochemical recurrence (BCR) after radical prostatectomy (RP) is a heterogeneous disease state in prostate cancer with multiple treatment options. Improved risk stratification could enable more personalized decision-making. We developed and validated a digital pathology-based multimodal artificial intelligence (MMAI) model to predict outcomes in post-RP BCR patients undergoing salvage therapy. METHODS:An MMAI model was trained to predict distant metastasis (DM) using prostate histopathology image features and clinical variables (pathologic grade group, pathologic T stage, prostate-specific antigen level before salvage radiotherapy [SRT], age, and surgical margin). The locked model was validated in 533 patients from NRG/RTOG 9601 and 0534 treated with SRT ± hormone therapy (HT), using Cox regression and time-dependent area under the receiver operating characteristic curve. KEY FINDINGS AND LIMITATIONS:With a median follow-up of 9.3 yrs, MMAI score was significantly associated with DM (subdistribution hazard ratio = 2.17 per standard deviation [95% confidence interval 1.65-2.85]; p < 0.001) and remained independently prognostic after adjusting for clinical variables and treatment. The 10-yr time-dependent area under the receiver operating characteristic curve for MMAI was 0.74 compared with 0.68 for a clinical nomogram. Binary risk categorization demonstrated higher 10-yr DM incidence in the MMAI high-risk (25%) than in the low-risk (8.8%) group. The absolute reduction in 10-yr DM incidence with HT plus SRT versus SRT alone was 21% in the high-risk group versus 2.5% in the low-risk group. Limitations include the use of archived trial cohorts. CONCLUSIONS AND CLINICAL IMPLICATIONS:The post-RP MMAI model provides individualized risk estimates after SRT ± HT and may support shared decision-making about salvage treatment. External and prospective validation are ongoing.

PURPOSE:Long-term androgen deprivation therapy (ADT) improves survival in men with high-risk localized prostate cancer (PCa) receiving radiotherapy (RT). Predictive biomarkers are needed to guide ADT duration. METHODS:A multimodal artificial intelligence (MMAI)-derived predictive biomarker was trained for long-term (LT) versus short-term (ST) ADT using pretreatment digital prostate biopsy images and clinical data (age, prostate-specific antigen, Gleason, and T stage) from six NRG Oncology phase III randomized radiotherapy trials. The novel MMAI-derived biomarker was developed to predict the differential benefit of LT-ADT on the primary end point, distant metastasis (DM). MMAI predictive utility was validated on a seventh randomized trial, RTOG 9202 (N = 1,192), which randomly assigned men to RT + ST-ADT (4 months) versus RT + LT-ADT (28 months). Fine-Gray and cumulative incidence analyses for DM, and secondarily, death with DM, were performed. Deaths without DM were treated as competing risks. RESULTS:In the validation cohort (median follow-up, 17.2 years), LT-ADT significantly improved DM from 26% to 17% (subdistribution hazard ratio [sHR], 0.64 [95% CI, 0.50 to 0.82], P < .001). A significant biomarker-treatment predictive interaction was observed (P = .04) for DM, whereby MMAI biomarker-positive men (n = 785, 66%) had reduced DM with LT-ADT versus ST-ADT (sHR, 0.55 [95% CI, 0.41 to 0.73], P < .001), whereas no treatment benefit was observed for MMAI biomarker-negative men (n = 407; sHR, 1.06 [95% CI, 0.61 to 1.84], P = .84). The estimated 15-year DM risk difference between RT + LT-ADT and RT + ST-ADT was 14% in MMAI biomarker-positive men and 0% in MMAI biomarker-negative men. The MMAI biomarker was also prognostic for DM, irrespective of treatment (sHR, 2.35 [95% CI, 1.72 to 3.19], P < .001). CONCLUSION:To our knowledge, the MMAI model is the first validated predictive biomarker to guide ADT duration with RT in localized/locally advanced PCa. Approximately one third of men with high-risk PCa could safely be spared the additional 24 months of ADT and the associated morbidity.

PURPOSE Artificial intelligence (AI) tools could improve clinical decision making or exacerbate inequities because of bias. African American (AA) men reportedly have a worse prognosis for prostate cancer (PCa) and are underrepresented in the development genomic biomarkers. We assess the generalizability of tools developed using a multimodal AI (MMAI) deep learning system using digital histopathology and clinical data from NRG/Radiation Therapy Oncology Group PCa trials across racial subgroups. METHODS In total, 5,708 patients from five randomized phase III trials were included. Two MMAI algorithms were evaluated: (1) the distant metastasis (DM) MMAI model optimized to predict risk of DM, and (2) the PCa-specific mortality (PCSM) MMAI model optimized to focus on prediction death in the presence of DM (DDM). The prognostic performance of the MMAI algorithms was evaluated in AA and non-AA subgroups using time to DM (primary end point) and time to DDM (secondary end point). Exploratory end points included time to biochemical failure and overall survival with Fine-Gray or Cox proportional hazards models. Cumulative incidence estimates were computed for time-to-event end points and compared using Gray's test. RESULTS There were 948 (16.6%) AA patients, 4,731 non-AA patients (82.9%), and 29 (0.5%) patients with unknown or missing race status. The DM-MMAI algorithm showed a strong prognostic signal for DM in the AA (subdistribution hazard ratio [sHR], 1.2 [95% CI, 1.0 to 1.3]; P = .007) and non-AA subgroups (sHR, 1.4 [95% CI, 1.3 to 1.5]; P < .001). Similarly, the PCSM-MMAI score showed a strong prognostic signal for DDM in both AA (sHR, 1.3 [95% CI, 1.1 to 1.5]; P = .001) and non-AA subgroups (sHR, 1.5 [95% CI, 1.4 to 1.6]; P < .001), with similar distributions of risk. CONCLUSION Using cooperative group data sets with a racially diverse population, the MMAI algorithm performed well across racial subgroups without evidence of algorithmic bias.

PURPOSE Current clinical risk stratification methods for localized prostate cancer are suboptimal, leading to over- and undertreatment. Recently, machine learning approaches using digital histopathology have shown superior prognostic ability in phase III trials. This study aims to develop a clinically usable risk grouping system using multimodal artificial intelligence (MMAI) models that outperform current National Comprehensive Cancer Network (NCCN) risk groups. MATERIALS AND METHODS The cohort comprised 9,787 patients with localized prostate cancer from eight NRG Oncology randomized phase III trials, treated with radiation therapy, androgen deprivation therapy, and/or chemotherapy. Locked MMAI models, which used digital histopathology images and clinical data, were applied to each patient. Expert consensus on cut points defined low-, intermediate-, and high-risk groups on the basis of 10-year distant metastasis rates of 3% and 10%, respectively. The MMAI's reclassification and prognostic performance were compared with the three-tier NCCN risk groups. RESULTS The median follow-up for censored patients was 7.9 years. According to NCCN risk categories, 30.4% of patients were low-risk, 25.5% intermediate-risk, and 44.1% high-risk. The MMAI risk classification identified 43.5% of patients as low-risk, 34.6% as intermediate-risk, and 21.8% as high-risk. MMAI reclassified 1,039 (42.0%) patients initially categorized by NCCN. Despite the MMAI low-risk group being larger than the NCCN low-risk group, the 10-year metastasis risks were comparable: 1.7% (95% CI, 0.2 to 3.2) for NCCN and 3.2% (95% CI, 1.7 to 4.7) for MMAI. The overall 10-year metastasis risk for NCCN high-risk patients was 16.6%, with MMAI further stratifying this group into low-, intermediate-, and high-risk, showing metastasis rates of 3.4%, 8.2%, and 26.3%, respectively. CONCLUSION The MMAI risk grouping system expands the population of men identified as having low metastatic risk and accurately pinpoints a high-risk subset with elevated metastasis rates. This approach aims to prevent both overtreatment and undertreatment in localized prostate cancer, facilitating shared decision making.

Background: Accurate risk stratification is critical to guide management decisions in localized prostate cancer (PCa). Previously, we had developed and validated a multimodal artificial intelligence (MMAI) model generated from digital histopathology and clinical features. Here, we externally validate this model on men with high-risk or locally advanced PCa treated and followed as part of a phase 3 randomized control trial. Objective: To externally validate the MMAI model on men with high-risk or locally advanced PCa treated and followed as part of a phase 3 randomized control trial. Design, setting, and participants: Our validation cohort included 318 localized high-risk PCa patients from NRG/RTOG 9902 with available histopathology (337 [85%] of the 397 patients enrolled into the trial had available slides, of which 19 [5.6%] failed due to poor image quality). Outcome measurements and statistical analysis: Two previously locked prognostic MMAI models were validated for their intended endpoint: distant metastasis (DM) and PCa-specific mortality (PCSM). Individual clinical factors and the number of National Comprehensive Cancer Network (NCCN) high-risk features served as comparators. Subdistribution hazard ratio (sHR) was reported per standard deviation increase of the score with corresponding 95% confidence interval (CI) using Fine-Gray or Cox proportional hazards models. Results and limitations: The DM and PCSM MMAI algorithms were significantly and independently associated with the risk of DM (sHR [95% CI] = 2.33 [1.60-3.38], p < 0.001) and PCSM, respectively (sHR [95% CI] = 3.54 [2.38-5.28], p < 0.001) when compared against other prognostic clinical factors and NCCN high-risk features. The lower 75% of patients by DM MMAI had estimated 5- and 10-yr DM rates of 4% and 7%, and the highest quartile had average 5- and 10-yr DM rates of 19% and 32%, respectively (p < 0.001). Similar results were observed for the PCSM MMAI algorithm. Conclusions: We externally validated the prognostic ability of MMAI models previously developed among men with localized high-risk disease. MMAI prognostic models further risk stratify beyond the clinical and pathological variables for DM and PCSM in a population of men already at a high risk for disease progression. This study provides evidence for consistent validation of our deep learning MMAI models to improve prognostication and enable more informed decision-making for patient care.

299 Background: Recently, an MMAI prognostic biomarker, ArteraAI Prostate, was trained and validated in localized prostate cancer to more accurately risk stratify patients for multiple endpoints compared to NCCN risk groups (Esteva et al., 2022). Prognostication within an NCCN risk group remains clinically important given the multiple treatment decisions required within each risk group (e.g., radiotherapy dose or hormone therapy use). Herein, we validated the MMAI biomarker in high-risk prostate cancer where an increasing number of therapeutic decisions is required. Methods: This study leveraged histopathology image and clinical data from patients with at least one high-risk feature (HRF; cT3-cT4, Gleason 8-10, PSA > 20 ng/mL, primary Gleason pattern 5) from six NRG/RTOG phase III randomized trials (n=1,088). Patients from two trials not part of the initial MMAI biomarker training/validation (RTOG 0521 [n=344] and 9902 [n=318]) and the MMAI validation cohort (RTOG 9202, 9408, 9413, and 9910 [n=426]) were included. Fine-Gray, cumulative incidence, and time dependent area under the curve (tdAUC) analyses were performed for time to distant metastasis (DM) and prostate cancer-specific mortality (PCSM) for standard clinicopathologic variables (age, PSA, Gleason score, T-stage, number of HRFs) and the MMAI model, as a continuous score (per standard deviation increase) and categorically by quartile. Death from other causes were treated as competing risks. Results: The analyzed cohort had a median follow-up of 10.4 years. Median PSA was 21 ng/mL, 60% had Gleason 8-10 disease, 37% had cT3-T4 disease, and 20% were African American. On univariable analysis, the MMAI model was significantly associated with DM (subdistribution hazard ratio [sHR] 2.05, 95% CI 1.74-2.43, p<0.001) and PCSM (sHR 2.04, 95% CI 1.73-2.42, <0.001). On multivariable analysis, the MMAI model, adjusting for either age, PSA, Gleason score, T-stage, or number of HRFs, was the only variable significantly associated with DM. TdAUC was highest for the MMAI biomarker for both 5-year DM (0.71), compared to PSA (0.56), Gleason score (0.61), T-stage (0.63), or number of HRFs (0.64), and for 5-year PCSM (0.75), compared to clinicopathologic variables (range 0.53-0.63). The estimated 10-year DM and 15-year PCSM rates for MMAI quartile 1 vs 4 were 8% vs 31% and 8% vs 34%, respectively. Conclusions: Our novel MMAI prognostic biomarker was successfully validated across six phase III randomized trials with long-term follow-up to be independently prognostic over standard clinical and pathologic variables for men with high-risk prostate cancer. Despite all patients having high-risk disease, the MMAI biomarker identified those with highly variable risks for DM and PCSM. This tool can help enable personalized, shared decision making for patients and providers.

Bladder normal tissue and tumors

Bladder cancer outcome and subtype classification by gene expression

PAM classifier of pT1 versus pTa tumors.

Clinical Information for all Tumors

Prostate cancer is the most frequent cancer in men and a leading cause of cancer death. Determining a patient's optimal therapy is a challenge, where oncologists must select a therapy with the highest likelihood of success and the lowest likelihood of toxicity. International standards for prognostication rely on non-specific and semi-quantitative tools, commonly leading to over- and under-treatment. Tissue-based molecular biomarkers have attempted to address this, but most have limited validation in prospective randomized trials and expensive processing costs, posing substantial barriers to widespread adoption. There remains a significant need for accurate and scalable tools to support therapy personalization. Here we demonstrate prostate cancer therapy personalization by predicting long-term, clinically relevant outcomes using a multimodal deep learning architecture and train models using clinical data and digital histopathology from prostate biopsies. We train and validate models using five phase III randomized trials conducted across hundreds of clinical centers. Histopathological data was available for 5654 of 7764 randomized patients (71%) with a median follow-up of 11.4 years. Compared to the most common risk-stratification tool-risk groups developed by the National Cancer Center Network (NCCN)-our models have superior discriminatory performance across all endpoints, ranging from 9.2% to 14.6% relative improvement in a held-out validation set. This artificial intelligence-based tool improves prognostication over standard tools and allows oncologists to computationally predict the likeliest outcomes of specific patients to determine optimal treatment. Outfitted with digital scanners and internet access, any clinic could offer such capabilities, enabling global access to therapy personalization.

BACKGROUND: Androgen deprivation therapy (ADT) with radiotherapy can benefit patients with localized prostate cancer. However, ADT can negatively impact quality of life, and there remain no validated predictive models to guide its use. METHODS: We used digital pathology images from pretreatment prostate tissue and clinical data from 5727 patients enrolled in five phase 3 randomized trials, in which treatment was radiotherapy with or without ADT, as our data source to develop and validate an artificial intelligence (AI)–derived predictive patient-specific model that would determine which patients would develop the primary end point of distant metastasis. The model used baseline data to provide a binary output that a given patient will likely benefit from ADT or not. After the model was locked, validation was performed using data from NRG Oncology/Radiation Therapy Oncology Group (RTOG) 9408 (n=1594), a trial that randomly assigned men to radiotherapy plus or minus 4 months of ADT. Fine–Gray regression and restricted mean survival times were used to assess the interaction between treatment and the predictive model and within predictive model–positive, i.e., benefited from ADT, and –negative subgroup treatment effects. RESULTS: Overall, in the NRG/RTOG 9408 validation cohort (14.9 years of median follow-up), ADT significantly improved time to distant metastasis. Of these enrolled patients, 543 (34%) were model positive, and ADT significantly reduced the risk of distant metastasis compared with radiotherapy alone. Of 1051 patients who were model negative, ADT did not provide benefit. CONCLUSIONS: Our AI-based predictive model was able to identify patients with a predominantly intermediate risk for prostate cancer likely to benefit from short-term ADT. (Supported by a grant [U10CA180822] from NRG Oncology Statistical and Data Management Center, a grant [UG1CA189867] from NCI Community Oncology Research Program, a grant [U10CA180868] from NRG Oncology Operations, and a grant [U24CA196067] from NRG Specimen Bank from the National Cancer Institute and by Artera, Inc. ClinicalTrials.gov numbers NCT00767286, NCT00002597, NCT00769548, NCT00005044, and NCT00033631.)

Outcome based on groups of tumors identified by hierarchical clustering.

Log2 ratios for 80 bladder tumors (UR), 3 normal bladder samples, and 9 bladder cancer cell lines.

Supplementary Data from Protein Acetylation and Histone Deacetylase Expression Associated with Malignant Breast Cancer Progression

5001 Background: Androgen deprivation therapy (ADT) improves survival and reduces risk of metastasis in men with high-risk localized prostate cancer (PC) receiving radiotherapy (RT). Predictive biomarkers are needed to guide ADT duration to maximize benefits and minimize risks. We sought to train and validate the first predictive biomarker for long-term (LT) vs short-term (ST) ADT using multiple phase III NRG Oncology randomized trials. Methods: Pre-treatment prostate biopsy slides were digitized from six phase III NRG/RTOG randomized trials of men receiving RT +/- ADT. The artificial intelligence (AI)-derived clinical and histopathological predictive biomarker was trained on RTOG 9408, 9413, 9902, 9910, and 0521 to predict differential benefit of LTADT on distant metastasis (DM). After the AI biomarker was locked, it was validated on RTOG 9202, which randomized men to RT + STADT (4 mo) vs LTADT (28 mo). The predictive utility of the AI biomarker was evaluated for the primary and secondary endpoints of DM and PC-specific mortality (PCSM), respectively, for ADT duration with Fine-Gray interaction models. Event rates were estimated by the cumulative incidence method. Deaths from other causes were treated as competing risks. Results: The AI-derived biomarker was trained on 2,641 men (median follow-up of 9.8 years, IQR [8.2, 11.5]) and validated on 1,192 men from RTOG 9202 (median follow-up of 17.2 years, IQR [9.1, 19.6]), where 80% had at least one high/very high (H/VH) risk feature (cT3-4, Gleason 8-10, PSA > 20, or primary Gleason pattern 5). Consistent with published results, LTADT significantly improved DM (subdistribution HR [sHR] 0.64, 95% CI 0.50-0.82, p < 0.001) in the validation cohort. The AI biomarker was prognostic for DM (sHR 2.35, 95% CI 1.72-3.19, p < 0.001). A significant biomarker-treatment interaction was observed (p = 0.04), in which AI-biomarker (+) men (n = 785, 66%) had reduced DM with LTADT (sHR 0.55, 95% CI 0.41-0.73, p < 0.001), but no benefit was observed (sHR 1.06, 95% CI 0.61-1.84, p = 0.84) for AI-biomarker (-) men (n = 407, 34%). The 10-year DM rate difference between RT + LTADT vs RT + STADT was 13% in AI-biomarker (+) men vs 2% in AI-biomarker (-) men. Similar trends were observed for PCSM outcomes. Risk classification (NCCN intermediate [n = 221, 43% (+)] vs other H/VH risk [n = 954, 71% (+)]) was prognostic but not predictive of LTADT benefit. Conclusions: We have successfully validated the first predictive biomarker of LTADT benefit with RT in localized high-risk PC using an AI-derived digital pathology-based platform in the phase III NRG/RTOG 9202 trial. The predictive AI biomarker identified 34% of men that could derive similar benefit with STADT, avoiding the side effects of prolonged ADT, and 43% of intermediate risk men who would benefit from LTADT.

cDNA clones identified as differentially expressed in superficial (pTa and pT1) versus muscle-invasive (pT2-pT4) tumors by both SAM and maxT.

108 Background: Artificial intelligence (AI) tools can display racial bias as a result of existing systemic health inequities and biased datasets. We have previously developed multi-modal AI (MMAI) prognostic models based on digital pathology images from five phase III randomized radiotherapy prostate cancer trials that outperform NCCN risk groups for prediction of distant metastasis (DM), biochemical failure (BF), prostate cancer-specific mortality (PCSM) and all-cause mortality (OS). In this study, we assessed the algorithmic fairness of the locked MMAI models between African American (AA) and non-AA populations in the five randomized trials. Methods: Patients enrolled in NRG/RTOG 9202, 9408, 9413, 9910, and 0126 with digitized biopsy histopathology slides were included in this study. The locked MMAI models were applied, and subgroup analyses were conducted by comparing distributions of clinical variables and MMAI scores (medians for continuous variables and proportions for categorical variables reported), and evaluating MMAI models’ prognostic ability among AA and non-AA men. The performance of the models were compared using DM as the primary endpoint and secondary endpoints of BF, PCSM, OS (death without an event as a competing risk) with Fine-Gray or Cox Proportional Hazards models. Either Kaplan Meier or cumulative incidence estimates were computed and compared using log-rank or Gray’s test. Results: This study included 5,624 men: 932 (17%) AA, 4503 (80%) white, and 189 (3%) other races. AA had younger median age (69 vs 71 year [yr]), higher median baseline PSA (12 vs 10 ng/mL), more T1-T2a (62% vs 57%), more Gleason < 7 (42% vs 36%) and 8-10 (15% vs 12%), and more NCCN low and high risk (12% vs 10% and 41% vs 33%). AA and non-AA had estimated 5-yr BF rates 27% and 27%, 5-yr DM rates 5% and 5%, 10-yr PCSM 5% and 7%, and 10-yr OS 58% and 60%, respectively. The median (interquartile range) score of the model optimizing for 5-yr DM (5-yr DM MMAI) was 0.044 (0.037–0.059) in AA and 0.043 (0.036–0.057) in non-AA. Similarly, all other MMAI models had differences in the medians between AA and non-AA ranging from 0.001 to 0.02. For all endpoints, the 5-yr DM MMAI model showed strong prognostic signal (hazard ratio [HR] per one standard deviation increase: 1.6 for DM, 1.4 for BF, 1.6 for PCSM and 1.3 for OS, all p-values < 0.001) and had comparable trends within AA vs. non-AA in the entire cohort (e.g., HR for DM 1.4 vs 1.6). Similar results were observed for the MMAI model optimizing for 10-yr PCSM. Conclusions: To our knowledge, this represents the first comparative analyses of a digital pathology AI prognostic model in AA vs. non-AA prostate cancer patients. The prognostic performance of the AI models was found to be comparable between subgroups. Our data supports the use of these models across racial groups, though further validation in AA cohorts is ongoing.

222 Background: Prognostication in localized prostate cancer is reliant on non-specific tools, an issue that leads to the over- and under-treatment of patients. Various tissue-based molecular biomarkers have attempted to fill this unmet need, but most lack prospective randomized trial validation. Herein, we train and validate prognostic biomarkers in localized prostate cancer using five phase III randomized trials, by leveraging multi-modal deep learning on digital histopathology. Methods: Histopathology image data was generated from pre-treatment biopsy slides in five NRG Oncology phase III randomized radiotherapy prostate cancer trials (RTOG 9202, 9408, 9413, 9910, and 0126). The trials were randomly split into training (80%) and validation (20%) cohorts. A multi-modal artificial intelligence (MMAI) architecture was developed to take clinicopathologic and image-based (histopathology) data as input and predict binary outcomes. Using this architecture, various models were trained to predict relevant clinical endpoints: biochemical recurrence (BCR), distant metastasis (DM), prostate cancer-specific survival (PCaSS), and overall survival (OS). These models were then validated for measures of prognostic discrimination using the time-based area under the curve (AUC) method. Results: Clinicopathologic and histopathology image data was available for 5,654 of 7,957 eligible patients (71.1%), yielding 16.1 TB of data from 16,204 histopathology slides of pretreatment biopsy samples. After training the models, locking them, and evaluating them on the validation cohort, we found that the MMAI prognostic model had superior discrimination compared to the NCCN model (PSA, T-stage, and Gleason score) for 5-year DM (AUC of 0.84 vs 0.73), 5-year BCR (AUC of 0.69 vs 0.58), 10-year PCaSS (AUC of 0.79 vs 0.66), and 10-year OS (AUC of 0.65 vs 0.58). Within each of the individual trials in the validation cohort, the MMAI-model had superior performance compared to NCCN risk groups for all clinical endpoints. Conclusions: This represents the first ever development and validation of prognostic biomarkers in localized prostate cancer using multiple large phase III clinical trials. We have successfully validated that our MMAI-prognostic biomarkers are superior to standard clinical and pathologic variables in identifying future BCR, DM, PCaSS, and OS. This massively scalable technology is feasible and can help personalize the management of prostate cancer patients. Funding: This project was supported by grants U10CA180868 (NRG Oncology Operations), U10CA180822 (NRG Oncology SDMC), UG1CA189867 (NCORP), U24CA196067 (NRG Specimen Bank) from the National Cancer Institute (NCI).