Racial disparities in stage at bladder cancer diagnosis in the US Veterans Affairs healthcare system.

OBJECTIVE: To describe patient characteristics and pathological stage at bladder cancer (BCa) diagnosis in a diverse population within a national, equal-access healthcare system. METHODS: This retrospective cohort study identified 15 966 men diagnosed with BCa in the Veterans Affairs (VA) healthcare system from 2000 to 2020. The primary outcome was pathological stage at diagnosis, determined by index transurethral resection of bladder tumour. Logistic regression was used to assess the relationship between race and stage. Competing risk models tested the association between race and BCa-specific mortality with cumulative incidence estimates. RESULTS: Of 15 966 BCa patients, 12 868 (81%), 1726 (11%), 493 (3%) and 879 (6%) were White, Black, Hispanic and Other race, respectively. Black patients had significantly higher muscle-invasive bladder cancer (MIBC) rates than White patients (35% vs 32%; P = 0.009). In multivariable analysis, the odds of presenting with MIBC did not differ significantly between Black and White patients (odds ratio [OR] 1.10, 95% confidence interval [CI] 0.98-1.22) or between Hispanic patients (OR 0.82, 95% CI 0.67-1.01) and White patients. Compared to White patients, Black patients had a similar risk of BCa-specific mortality (hazard ratio [HR] 0.89, 95% CI 0.75-1.06), whereas Hispanic patients had a lower risk (HR 0.56, 95% CI 0.38-0.82). CONCLUSIONS: Black patients presented with the highest rates of de novo MIBC. However, in a large, equal-access healthcare system, this did not result in a difference in BCa-specific mortality. In contrast, Hispanic patients had lower risks of MIBC and BCa-specific mortality.

Association of Age with Non-muscle-invasive Bladder Cancer: Unearthing a Biological Basis for Epidemiological Disparities?

BACKGROUND: Age disparity in patients with non-muscle-invasive bladder cancer (NMIBC) exists. Whether this is due to differences in adequate cancer care or tumour biology is unclear. OBJECTIVE: To investigate age disparities in NMIBC using the Surveillance, Epidemiology, and End Results (SEER)-Medicare and UROMOL datasets. DESIGN, SETTING, AND PARTICIPANTS: The SEER-Medicare data were used to identify patients with clinical stage Ta, Tis, and T1 NMIBC between 2005 and 2017 (n = 32 225). Using the UROMOL cohort (n = 834), age disparities across transcriptomic, genomic, and spatial proteomic domains were assessed. OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS: For the SEER-Medicare data, multivariable competing-risk regression was used to examine the association between age and recurrence, progression, and bladder cancer-specific mortality (BCSM). For the UROMOL cohort, multivariable general linear model and multinomial logistic regression were performed to evaluate the association between age and tumour biology. RESULTS AND LIMITATIONS: An analysis of the SEER-Medicare cohort revealed 5-yr recurrence rates of 55.2%, 57.4%, and 58.9%; 5-yr progression rates of 25.6%, 29.2%, and 36.9%; and 5-yr BCSM rates of 3.9%, 5.8%, and 11.8% in patients aged 66-70, 71-80, and ≥81 yr, respectively. After multivariable adjustment, age ≥81 yr was associated with a higher risk of recurrence (hazard ratio [HR] 1.07, 95% confidence interval [CI] 1.03-1.12; p = 0.001), progression (HR 1.32, p < 0.001), and BCSM (HR 2.58, p < 0.001). UROMOL2021 transcriptomic class 2a was most frequently observed in patients with advanced age (34.0% in ≥76 yr vs 21.6% in ≤65 yr; p = 0.004), a finding confirmed on multivariable analysis (risk ratio [RR] 3.86, p = 0.002). UROMOL2021 genomic class 3 was observed more frequently in patients aged ≥76 yr (4.9% vs 24.2%; p = 0.001). Limitations include the definitions used for recurrence and progression, which may lead to under- or overestimation of true rates. CONCLUSIONS: Among SEER-Medicare patients with NMIBC, advanced age is associated with inferior oncological outcomes. These results reflect age-related molecular biological differences observed across transcriptomic and genomic domains, providing further evidence that innate tumour biology contributes to observed disparities in NMIBC outcomes. PATIENT SUMMARY: Older patients with non-muscle-invasive bladder cancer have worse oncological outcomes than younger patients. Some of this age disparity may be due to differences in tumour biology.

Standardization of the evaluation and surveillance of patients with BCG unresponsive high grade non-muscle invasive bladder cancer clinical trials.

There are multiple ongoing and planned clinical trials that are evaluating novel therapies to treat patients with BCG-unresponsive high grade nonmuscle invasive bladder cancer (NMIBC). Importantly, there is considerable variation in surveillance strategies between these clinical trials, specifically with regards to the use of advanced imaging, enhanced cystoscopy, and mandatory biopsies, which could impact landmark efficacy assessments of investigational agents. To present guideline recommendations for the standardization of cystoscopic evaluation, surveillance, and efficacy assessments for patients with BCG-unresponsive NMIBC participating in clinical trials. On September 29, 2023 at the annual meeting of the International Bladder Cancer Network, a breakout session was convened, during which representatives from various disciplines discussed potential guidance statements with opportunity for discussion and comment. A set of statements regarding use of white light and enhanced cystoscopy were developed to help guide a pragmatic approach to surveillance and efficacy assessments of patients in clinical trials. The use of "for cause" and "mandatory" biopsies was also addressed. A standard approach to evaluation of patients within the context of clinical trials is necessary to accurately assess the efficacy of novel agents, especially within single arm trials that lack an appropriate comparator. Additionally, the utilization and timing of mandatory biopsies is critical, as these biopsies may impact both disease evaluations and the determination of duration of response.

A SEER-Medicare Based Quality Score for Patients With Metastatic Upper Tract Urothelial Carcinoma.

BACKGROUND: Population-based studies evaluating outcomes for metastatic upper tract urothelial carcinoma (mUTUC) are sparse and rarely capture both patients with de novo (synchronous) metastases and those who progress to metastatic disease (metachronous). Herein we evaluated the outcomes and costs associated with synchronous and metachronous mUTUC, utilizing a novel Methodology. Additionally, we created a guideline-based quality score to improve care in this space. PATIENTS AND METHODS: We identified all patients with mUTUC aged 66 years and older included in the SEER-Medicare linked database between 2004 and 2012. Achievement of 3 quality criteria was assessed: (1) cancer-specific survival (CSS)>12 months; (2) receipt of systemic therapy; (3) receipt of hospice/palliative care. Total healthcare and out-of-pocket costs were evaluated. Regression analyses were performed to assess characteristics associated with quality criteria and total healthcare costs. RESULTS: Of the 1223 patients identified, at least one quality criterion was met in just 40.2% and only 54 patients (4.4%) received palliative care. In multivariable analysis, patients with synchronous mUTUC (OR:0.55, 95%CI:0.41-0.72), and at least 3 comorbidities (OR:0.68, 95%CI:0.47-0.98) were less likely to achieve at least 1 quality criterion. Meeting at least 1 quality criterion was associated with increased costs ($94,677, 95%CI:87,702-101,652 versus $63,575, 95%CI:59,598-67,552). CONCLUSIONS: Less than half of patients with mUTUC met at least 1 quality criterion. Quality score achievement was associated with a modest increase in total healthcare spending. These findings not only provide guidance for future study of rare diseases using secondary data, but also highlight inadequacies in the current management of mUTUC.

Background and Update for ECOG-ACRIN EA8212: A Randomized Phase 3 Trial of Intravesical Bacillus Calmette-Guérin (BCG) Versus Intravesical Docetaxel and Gemcitabine Treatment in BCG-naïve High-grade Non-muscle-invasive Bladder Cancer (BRIDGE).

EA8212 BRIDGE is a phase 3 randomized trial comparing BCG vs GemDoce for BCG naïve high-risk non-muscle-invasive bladder cancer. This article provides an explanation for the rationale of the clinical trial and details the study design.

Exposure to Agent Orange and Risk of Bladder Cancer Among US Veterans.

Importance: To date, limited data exist regarding the association between Agent Orange and bladder cancer, and the Institute of Medicine concluded that the association between exposure to Agent Orange and bladder cancer outcomes is an area of needed research. Objective: To examine the association between bladder cancer risk and exposure to Agent Orange among male Vietnam veterans. Design, Setting, and Participants: This nationwide Veterans Affairs (VA) retrospective cohort study assesses the association between exposure to Agent Orange and bladder cancer risk among 2 517 926 male Vietnam veterans treated in the VA Health System nationwide from January 1, 2001, to December 31, 2019. Statistical analysis was performed from December 14, 2021, to May 3, 2023. Exposure: Agent Orange. Main Outcomes and Measures: Veterans exposed to Agent Orange were matched in a 1:3 ratio to unexposed veterans on age, race and ethnicity, military branch, and year of service entry. Risk of bladder cancer was measured by incidence. Aggressiveness of bladder cancer was measured by muscle-invasion status using natural language processing. Results: Among the 2 517 926 male veterans (median age at VA entry, 60.0 years [IQR, 56.0-64.0 years]) who met inclusion criteria, there were 629 907 veterans (25.0%) with Agent Orange exposure and 1 888 019 matched veterans (75.0%) without Agent Orange exposure. Agent Orange exposure was associated with a significantly increased risk of bladder cancer, although the association was very slight (hazard ratio [HR], 1.04; 95% CI, 1.02-1.06). When stratified by median age at VA entry, Agent Orange was not associated with bladder cancer risk among veterans older than the median age but was associated with increased bladder cancer risk among veterans younger than the median age (HR, 1.07; 95% CI, 1.04-1.10). Among veterans with a diagnosis of bladder cancer, Agent Orange was associated with lower odds of muscle-invasive bladder cancer (odds ratio [OR], 0.91; 95% CI, 0.85-0.98). Conclusions and Relevance: In this cohort study among male Vietnam veterans, there was a modestly increased risk of bladder cancer-but not aggressiveness of bladder cancer-among those exposed to Agent Orange. These findings suggest an association between Agent Orange exposure and bladder cancer, although the clinical relevance of this was unclear.

Robot-assisted Radical Cystectomy Versus Open Radical Cystectomy: A Systematic Review and Meta-analysis of Perioperative, Oncological, and Quality of Life Outcomes Using Randomized Controlled Trials.

CONTEXT: Differences in recovery, oncological, and quality of life (QoL) outcomes between open radical cystectomy (ORC) and robot-assisted radical cystectomy (RARC) for patients with bladder cancer are unclear. OBJECTIVE: This review aims to compare these outcomes within randomized trials of ORC and RARC in this context. The primary outcome was the rate of 90-d perioperative events. The secondary outcomes included operative, pathological, survival, and health-related QoL (HRQoL) measures. EVIDENCE ACQUISITION: Systematic literature searches of MEDLINE, Embase, Web of Science, and clinicaltrials.gov were performed up to May 31, 2022. EVIDENCE SYNTHESIS: Eight trials, reporting 1024 participants, were included. RARC was associated with a shorter hospital length of stay (LOS; mean difference [MD] 0.21, 95% confidence interval [CI] 0.03-0.39, p = 0.02) than and similar complication rates to ORC. ORC was associated with higher thromboembolic events (odds ratio [OR] 1.84, 95% CI 1.02-3.31, p = 0.04). ORC was associated with more blood loss (MD 322 ml, 95% CI 193-450, p < 0.001) and transfusions (OR 2.35, 95% CI 1.65-3.36, p < 0.001), but shorter operative time (MD 76 min, 95% CI 39-112, p < 0.001) than RARC. No differences in lymph node yield (MD 1.07, 95% CI -1.73 to 3.86, p = 0.5) or positive surgical margin rates (OR 0.95, 95% CI 0.54-1.67, p = 0.9) were present. RARC was associated with better physical functioning or well-being (standardized MD 0.47, 95% CI 0.29-0.65, p < 0.001) and role functioning (MD 8.8, 95% CI 2.4-15.1, p = 0.007), but no improvement in overall HRQoL. No differences in progression-free survival or overall survival were seen. Limitations may include a lack of generalization given trial patients. CONCLUSIONS: RARC offers various perioperative benefits over ORC. It may be more suitable in patients wishing to avoid blood transfusion, those wanting a shorter LOS, or those at a high risk of thromboembolic events. PATIENT SUMMARY: This study compares robot-assisted keyhole surgery with open surgery for bladder cancer. The robot-assisted approach offered less blood loss, shorter hospital stays, and fewer blood clots. No other differences were seen.

The Impact of Blue Light Cystoscopy Use Among Nonmuscle Invasive Bladder Cancer Patients in an Equal Access Setting: Implications on Recurrence and Time to Recurrence.

INTRODUCTION: Prior studies suggest that white light cystoscopy (WLC) alone can fail to detect cases of non-muscle invasive bladder cancer (NMIBC) vs. blue light cystoscopy (BLC). We describe bladder cancer outcomes and the impact of BLC among NMIBC patients in an equal access setting. MATERIALS AND METHODS: We assessed 378 NMIBC patients within the Veterans Affairs system that had a CPT code for BLC from December 1, 2014 to December 31, 2020. We determined recurrence rates and time to recurrence prior to BLC (ie, after previous WLC if available) and following BLC. We used the Kaplan-Meier method to estimate event-free survival and Cox regression to determine association between BLC and recurrence, progression, and overall survival; and further, whether these outcomes differed by race. RESULTS: Of 378 patients with complete data, 43 (11%) were Black and 300 (79%) White. Median follow-up was 40.7 months from bladder cancer diagnosis. Median time to first recurrence following BLC was longer vs. WLC alone (40 [33-NE] vs. 26 [17-39] months). Recurrence risk was significantly lower following BLC (Hazard Ratio [HR] 0.70; 95% Confidence Interval [CI], 0.54-0.90). There was no significant difference in recurrence (HR 0.69; 95% CI, 0.39-1.20), progression (HR 1.13; 95% CI, 0.32-3.96), and overall survival (HR 0.74; 95% CI, 0.31-1.77) following BLC by Black vs. White race. CONCLUSION: In this study from an equal access setting in the VA, we observed significantly decreased recurrence risk and prolonged time interval to recurrence following BLC vs. WLC alone. There was no difference in bladder cancer outcomes by race.

Cost-Effectiveness and Economic Impact of Bladder Cancer Management: An Updated Review of the Literature.

Bladder cancer remains one of the costliest malignancies to manage. We provide a narrative review of literature assessing the economic burden and cost-effectiveness of bladder cancer treatment and surveillance. This is an update to a previous review and focuses on data published within the past 10 years. We queried PubMed and MEDLINE for all bladder cancer cost-related literature between 2013 and 2023. After initial screening, 117 abstracts were identified, 50 of which were selected for inclusion in our review. Management of disease recurrence and treatment complications contributes significantly to the high cost of care. High-value interventions are therefore treatments that improve recurrence-free and overall survival at minimal additional toxicity. De-escalation of surveillance and diagnostic interventions may help to reduce costs in this space without compromising oncologic control. The persistently rising cost of novel cancer drugs undermines their value when only modest gains in efficacy are observed. Multiple cost-effectiveness analyses have been published and are useful for contextualizing the cost, efficacy, and impact on quality of life that interventions have in this population. Further cost-effectiveness work is needed to better characterize the impact that treatment costs have on patients' financial well-being and quality of life.

Association Between Antibiotic Prophylaxis Before Cystectomy or Stent Removal and Infection Complications: A Systematic Review.

CONTEXT: Patients undergoing radical cystectomy frequently suffer from infectious complications, including urinary tract infections (UTIs) and surgical site infections (SSIs) leading to emergency department visits, hospital readmission, and added cost. OBJECTIVE: To summarize the literature regarding perioperative antibiotic prophylaxis, ureteric stent usage, and prevalence of infectious complications after cystectomy. EVIDENCE ACQUISITION: A systematic review of PubMed/Medline, EMBASE, Cochrane Library, and reference lists was conducted. EVIDENCE SYNTHESIS: We identified 20 reports including a total of 55 306 patients. The median rates of any infection, UTIs, SSIs, and bacteremia were 40%, 20%, 11%, and 6%, respectively. Perioperative antibiotic prophylaxis differed substantially between reports. Perioperative antibiotics were used only during surgery in one study but were continued over several days after surgery in all other studies. Empirical use of antibiotics for 1-3 d after surgery was described in 12 studies, 3-10 d in two studies, and >10 d in four studies. Time to stent removal ranged from 4 to 25 d after cystectomy. Prophylactic antibiotics were used before stent removal in nine of 20 studies; two of these studies used targeted antibiotics based on urine cultures from the ureteric stents, and the other seven studies used a single shot or 2 d of empirical antibiotics. Studies with any prophylactic antibiotic before stent removal found a lower median percentage of positive blood cultures after stent removal than studies without prophylactic antibiotics before stent removal (2% vs 9%). CONCLUSIONS: We confirmed a high proportion of infectious complications after cystectomy, and a heterogeneous pattern of choice and duration of antibiotics during and after surgery or stent removal. These findings highlight a need for further studies and support quality prospective trials. PATIENT SUMMARY: In this review, we observed wide variability in the use of antibiotics before or after surgical removal of the bladder.

Bladder preservation after neoadjuvant therapy - 2021 IBCN updates part 1.

The morbidity associated with radical cystectomy (RC) for muscle-invasive bladder cancer (MIBC) has fueled investigations into the feasibility of bladder preservation strategies after a favorable clinical response to neoadjuvant therapy (NAT). Identifying optimal candidates for bladder preservation is predicated on our ability to identify tumors with inherent cisplatin sensitivity and accurately stage patients before and after NAT. In the present review, we evaluate the accuracy and limitations of contemporary staging modalities and investigate clinical outcomes in patients with MIBC who were managed with bladder preservation after NAT. Lastly, we discuss the predictive role of cisplatin-sensitizing DNA damage response (DDR) gene alterations as a foundational component to current prospective clinical trials evaluating bladder preservation in the setting of MIBC.

New Mental Health Diagnosis as a Prognostic Factor for Muscle-Invasive Bladder Cancer.

BACKGROUND: To examine differences in survival outcomes for muscle-invasive bladder cancer patients stratified by new mental health diagnosis. METHODS: Using the Surveillance, Epidemiology, and End Results (SEER)-Medicare data, we identified patients diagnosed with muscle-invasive bladder cancer between 2008 and 2014. Our primary outcome was cancer-specific and overall hazards of mortality. As a secondary outcome, we reported predictors of developing a new mental health diagnosis after bladder cancer diagnosis. We used Cox proportional hazards models to determine the impact of palliative care and mental health diagnoses on survival outcomes after adjusting for grade, stage, comorbidity index, and baseline demographics. RESULTS: Of the 3794 patients who met inclusion criteria, 1193 (31%) were diagnosed with a mental health illness after their bladder cancer diagnosis during the 6 years in the study period. The most common diagnoses were depression (13%), alcohol and drug abuse (12%), and anxiety (11%). Patients with a post-bladder cancer mental health diagnosis had a 57% higher hazard of overall mortality (HR 1.57, P = .048) and an 80% higher hazard of bladder cancer-specific mortality (HR 1.81, P = .037) CONCLUSIONS: New mental health diagnoses are associated with worse survival in patients with muscle invasive bladder cancer. This suggests that a multimodal approach to bladder cancer treatment should include addressing the non-oncologic needs of the patient to optimize survival outcomes.

Competing mortality risk from second primary malignancy in bladder cancer patients following radical cystectomy: Implications for survivorship.

INTRODUCTION: Muscle-invasive bladder cancer (BC) often occurs in patients with competing mortality risks, while also being associated with the highest rate of second primary nonurothelial cancers (SNUC) of all solid malignancies. We investigated the incidence, risk factors, and timing of SNUC as a competing mortality risk factor in patients with BC who were treated with curative intent radical cystectomy (RC). METHODS: We performed a retrospective cohort study assessing patients who underwent RC for cT2-4 N0M0 BC from January 1, 2005 to December 31, 2018 at a single, high volume tertiary care referral center. The Fine-Gray multivariable regression model was used to evaluate predictive factors for SNUC. Cumulative incidence of mortality (CIM) was estimated with modified Kaplan-Meier analysis. RESULTS: The median follow-up time for the 693 patients who underwent RC was 3.7 years (interquartile range [IQR] 1.9-5.9 years). SNUC developed in 85 (12.3%) patients at a median 3.0 years post-RC (IQR 1.2-5.5 years). On multivariable analysis, the only significant predictor for developing SNUC was freedom from BC recurrence or metastasis (HR 1.54, 95% CI 1.12-1.76, P = 0.019). The most common SNUCs were primary lung cancer (24, 3.2% of cohort) and colon cancer (9, 1.3% of cohort). BC surveillance imaging diagnosed SNUC in 35/52 (67.3%) patients with solid-organ visceral primaries. The overall mortality rate for any SNUC was 38.8%, with the 3 most lethal cancer types being pancreatic, lung, and colon (62.5%, 54.2%, and 44.4% mortality, respectively). The incidence of SNUC uniformly increased postoperatively, with a cumulative incidence of 22.1% (95% CI, 16.8-27.9%) at 12-years post-RC. 163 patients (23.5%) died from BC, 33 patients (4.8%) died from SNUC, and 94 patients (13.6%) died from other causes. While the CIM for BC plateaued around 5-years post-RC at 24%, the incidence of other-cause mortality uniformly rose throughout the postoperative period. By post-RC year 9 there was no significant difference in CIM between BC (CIM 27.2%, 95% CI, 23.5-31.1%) and other-causes (CIM 20.0%, 95% CI, 15.8-24.6%). CONCLUSIONS: The cumulative incidence of SNUC at 12-years post-RC was 22%, with the majority identified on BC surveillance imaging. While BC mortality plateaued around 5-years post-RC, mortality related to SNUC or other causes rose steadily in the postoperative period. These data have clinical significance with regards to patient counseling, survivorship and oncologic surveillance in the highly comorbid muscle-invasive BC population.

25 years International Bladder Cancer Network (IBCN): The past, the present, and the future.

In 1997 an international group of scientists organized a meeting in Barcelona, Spain, to discuss the use of biomarkers in the management of patients with bladder cancer. This meeting was the offspring of an - initially informal - group that finally resulted in the foundation and incorporation of the International Bladder Cancer Network (IBCN) e.V. in 2005. Over the years the group has supported several research initiatives and generated several recommendations on the use of biomarkers in the diagnosis and treatment of bladder cancer. Meeting quality was generated by inviting experts presenting state-of-the-art lectures or work in progress reports, interdisciplinarity and the limited number of participants supporting an open and personal exchange resulted in a format increasingly attracting participants from all over the world. The recent limitations caused by the Covid-19 pandemic were partially met by organizing several well attended webinars. The future challenge is to maintain the IBCN meeting spirit despite an increasing interest of the scientific community and industrial partners to participate. However, the integration of and interaction between increasingly more specialized disciplines is a challenge that can be better catalyzed by an international multidisciplinary network than mostly national professional associations.

Epidemiology, Screening, and Prevention of Bladder Cancer.

CONTEXT: Bladder cancer (BC) represents a significant health problem due to the potential morbidity and mortality associated with disease burden, which has remained largely unaltered over time. OBJECTIVE: To provide an expert collaborative review and describe the incidence, prevalence, and mortality of BC and to evaluate current evidence for BC screening and prevention. EVIDENCE ACQUISITION: Data on the estimated incidence and mortality of BC for 2020 in 185 countries were derived from the International Agency for Research on Cancer GLOBOCAN database. A review of English-language articles published over the past 5 yr was conducted using PubMed/MEDLINE to identify risk factors in addition to contemporary evidence on BC screening and prevention. EVIDENCE SYNTHESIS: BC is the tenth most common cancer worldwide, with 573 278 cases in 2020. BC incidence is approximately fourfold higher in men than women. Tobacco smoking remains the principal risk factor, accounting for approximately 50% of cases. There is insufficient evidence to recommend routine BC screening. However, targeted screening of high-risk individuals (defined according to smoking history or occupational exposure) may reduce BC mortality and should be the focus of prospective randomized trials. In terms of disease prevention, smoking cessation represents the most important intervention, followed by a reduction in exposure to occupational and environmental carcinogens. CONCLUSIONS: BC confers a significant disease burden. An understanding of BC epidemiology and risk factors provides an optimal foundation for disease prevention and the care of affected patients. PATIENT SUMMARY: Bladder cancer is the tenth most common cancer worldwide and is approximately four times more common among men than among women. The main risk factors are tobacco smoking, followed by exposure to carcinogens in the workplace or the environment. Routine screening is not currently recommended, but may be beneficial in individuals at high risk, such as heavy smokers. Primary prevention is extremely important, and smoking cessation represents the most important action for reducing bladder cancer cases and deaths.

Identification of Patients With Metastatic Prostate Cancer With Natural Language Processing and Machine Learning.

PURPOSE: Understanding treatment patterns and effectiveness for patients with metastatic prostate cancer (mPCa) is dependent on accurate assessment of metastatic status. The objective was to develop a natural language processing (NLP) model for identifying patients with mPCa and evaluate the model's performance against chart-reviewed data and an International Classification of Diseases (ICD) 9/10 code-based method. METHODS: In total, 139,057 radiology reports on 6,211 unique patients from the Department of Veterans Affairs were used. The gold standard was metastases by detailed chart review of radiology reports. NLP performance was assessed by sensitivity, specificity, positive predictive value, negative predictive value, and date of metastases detection. Receiver operating characteristic curves was used to assess model performance. RESULTS: When compared with chart review, the NLP model had high sensitivity and specificity (85% and 96%, respectively). The NLP model was able to predict patient-level metastasis status with a sensitivity of 91% and specificity of 81%, whereas sensitivity and specificity using ICD9/10 billing codes were 73% and 86%, respectively. For the NLP model, date of metastases detection was exactly concordant and within < 1 week in 55% and 58% of patients, compared with 8% and 17%, respectively, using the ICD9/10 billing codes method. The area under the curve for the NLP model was 0.911. A limitation is the NLP model was developed on the basis of a subset of patients with mPCa and may not be generalizable to all patients with mPCa. CONCLUSION: This population-level NLP model for identifying patients with mPCa was more accurate than using ICD9/10 billing codes when compared with chart-reviewed data. Upon further validation, this model may allow for efficient population-level identification of patients with mPCa.