Outcomes of Black men with prostate cancer treated with radiation therapy in the Veterans Health Administration.

BACKGROUND: Population-based studies demonstrate that Black men in the United States have an increased risk of death from prostate cancer. Determinants of racial disparities are multifactorial, including socioeconomic and biologic factors. METHODS: The authors conducted a pooled analysis of patients derived from 152 centers within the Veterans Health Administration. The cohort included men who had nonmetastatic prostate diagnosed between 2001 and 2015 and received definitive radiation therapy. The primary endpoint was prostate cancer-specific mortality (PCSM). Secondary endpoints included all-cause mortality (ACM) and the time from a prostate-specific antigen level ≥4 ng/mL to biopsy and radiation therapy. A Cox regression model was performed to adjust for differences between clinical parameters. RESULTS: Among the 31,131 patients included in the cohort, 9584 (30.8%) were Black. The 10-year cumulative incidence of death from prostate cancer was lower in Black men compared with White men (4.0% vs 4.8%; P = .004). In a competing risk model, Black race was associated with a decreased risk of PCSM (subdistribution hazard ratio, 0.79; 95% CI, 0.69-0.92; P = .002). Similarly, the 10-year cumulative incidence of death from any cause was lower in Black men (27.6% vs 31.8%; P < .001). In multivariable analysis, Black men had a 10% decreased risk of ACM (hazard ratio, 0.90; 95% CI, 0.85-0.95; P < .001). CONCLUSIONS: The current results indicate relatively lower PCSM and ACM among Black men who were included in a large Veterans Health Administration cohort and received radiation therapy as primary treatment for nonmetastatic prostate cancer. There is an ongoing need to continue to understand and mitigate the factors associated with disparities in health care outcomes.

Quality of care at safety-net hospitals and the impact on pay-for-performance reimbursement.

BACKGROUND: Pay-for-performance reimbursement ties hospital payments to standardized quality-of-care metrics. To the authors' knowledge, the impact of pay-for-performance reimbursement models on hospitals caring primarily for uninsured or underinsured patients remains poorly defined. The objective of the current study was to evaluate how standardized quality-of-care metrics vary by a hospital's propensity to care for uninsured or underinsured patients and demonstrate the potential impact that pay-for-performance reimbursement could have on hospitals caring for the underserved. METHODS: The authors identified 1,703,865 patients with cancer who were diagnosed between 2004 and 2015 and treated at 1344 hospitals. Hospital safety-net burden was defined as the percentage of uninsured or Medicaid patients cared for by that hospital, categorizing hospitals into low-burden, medium-burden, and high-burden hospitals. The authors evaluated the impact of safety-net burden on concordance with 20 standardized quality-of-care measures, adjusting for differences in patient age, sex, stage of disease at diagnosis, and comorbidity. RESULTS: Patients who were treated at high-burden hospitals were more likely to be young, male, Black and/or Hispanic, and to reside in a low-income and low-educated region. High-burden hospitals had lower adherence to 13 of 20 quality measures compared with low-burden hospitals (all P < .05). Among the 350 high-burden hospitals, concordance with quality measures was found to be lowest for those caring for the highest percentage of uninsured or Medicaid patients, minority patients, and less educated patients (all P < .001). CONCLUSIONS: Hospitals caring for uninsured or underinsured individuals have decreased quality-of-care measures. Under pay-for-performance reimbursement models, these lower quality-of-care scores could decrease hospital payments, potentially increasing health disparities for at-risk patients with cancer.

Survival of African American and non-Hispanic white men with prostate cancer in an equal-access health care system.

BACKGROUND: African American (AA) men in the general US population are more than twice as likely to die of prostate cancer (PC) compared with non-Hispanic white (NHW) men. The authors hypothesized that receiving care through the Veterans Affairs (VA) health system, an equal-access medical system, would attenuate this disparity. METHODS: A longitudinal, centralized database of >20 million veterans was used to assemble a cohort of 60,035 men (18,201 AA men [30.3%] and 41,834 NHW men [69.7%]) who were diagnosed with PC between 2000 and 2015. RESULTS: AA men were more likely to live in regions with a lower median income ($40,871 for AA men vs $48,125 for NHW men; P < .001) and lower high school graduation rates (83% for AA men vs 88% for NHW men; P < .001). At the time of diagnosis, AA men were younger (median age, 63.0 years vs 66.0 years; P < .001) and had a higher prostate-specific antigen level (median, 6.7 ng/mL vs 6.2 ng/mL; P < .001), but were less likely to have Gleason score 8 to 10 disease (18.8% among AA men vs 19.7% among NHW men; P < .001), a clinical T classification ≥3 (2.2% vs 2.9%; P < .001), or distant metastatic disease (2.7% vs 3.1%; P = 0.01). The 10-year PC-specific mortality rate was slightly lower for AA men (4.4% vs 5.1%; P = .005), which was confirmed in multivariable competing-risk analysis (subdistribution hazard ratio, 0.85; 95% CI, 0.78-0.93; P < .001). CONCLUSIONS: AA men diagnosed with PC in the VA health system do not appear to present with more advanced disease or experience worse outcomes compared with NHW men, in contrast to national trends, suggesting that access to care is an important determinant of racial equity.

Palliative Radiotherapy Versus Esophageal Stent Placement in the Management of Patients With Metastatic Esophageal Cancer.

BACKGROUND: Patients with advanced esophageal cancer often experience pain and dysphagia, yet the optimal palliative management remains unclear. This retrospective study evaluated outcomes and adverse effects of palliative radiotherapy (RT) compared with esophageal stenting among a cohort of U.S. veterans with metastatic esophageal cancer. PATIENTS AND METHODS: We identified 1,957 veterans in the United States with metastatic esophageal cancer who received palliative RT to the esophagus or esophageal stenting, and assessed the risks of severe adverse effects, including esophageal fistula formation, perforation, obstruction, hemorrhage, and esophagitis. We determined palliative efficacy by evaluating pain and dysphagia scores before and after intervention. Multivariable analyses were used to control for potential confounding factors. RESULTS: In our cohort, 1,593 patients underwent RT and 364 underwent esophageal stenting. The cumulative incidence of any severe adverse effect at 6 months was higher among patients who received stents compared with those who received RT (21.7% vs 12.4%; P<.0010). In multivariable analysis, patients who received stents had an increased risk of any severe adverse effect, including fistula, perforation, and hemorrhage (all P<.0500). Multivariable analysis also showed that, compared with stenting, RT was associated with more rapid and durable pain relief (P<.0010) with no difference in relief of dysphagia over time when accounting for pretreatment dysphagia scores (P=.1029). CONCLUSIONS: Compared with esophageal stenting, RT was associated with a decreased risk of adverse effects, greater pain relief, and equivalent relief of moderate to severe dysphagia over time. Unmeasured patient- or tumor-related factors could have influenced the choice of intervention, thereby impacting our study outcomes. To our knowledge, this is the largest study to date analyzing the comparative risks and benefits of palliative RT and esophageal stenting among patients with metastatic esophageal cancer.

Androgen deprivation therapy and depression in men with prostate cancer treated with definitive radiation therapy.

BACKGROUND: There is no consensus on the association between the use of androgen deprivation therapy (ADT) and the risk of developing depression. This study investigated the association between ADT use and the development of depression, outpatient psychiatric services, inpatient psychiatric services, and suicide in a homogeneous group of men with prostate cancer (PC) treated with definitive radiation therapy (RT) after controlling for multiple sources of selection bias. METHODS: This was a retrospective, observational cohort study of 39,965 veterans with PC who were treated with definitive RT and were diagnosed by the US Department of Veterans Affairs health care system between January 1, 2001, and October 31, 2015. Exposure was ADT initiation within 1 year of the PC diagnosis. The primary outcome was new development of depression. Secondary outcomes were outpatient psychiatric use, inpatient psychiatric use, and suicide. RESULTS: During follow-up, 934 patients were newly diagnosed with depression, 7825 patients used outpatient psychiatric services, 358 patients used inpatient psychiatric services, and 54 patients committed suicide. In the multivariable competing risks regression model, ADT was associated with the development of depression (subdistribution hazard ratio [SHR], 1.50; 95% confidence interval [CI], 1.32-1.71; P < .001). ADT was also associated with outpatient psychiatric utilization (SHR, 1.21; 95% CI, 1.16-1.27; P < .001). Finally, ADT was not associated with inpatient psychiatric utilization or suicide. CONCLUSIONS: An increase in the risk of depression and the use of outpatient psychiatric services was observed in a large cohort of men with PC who received ADT with definitive RT. These results may provide further evidence for the long-term risks of ADT for psychiatric health in the treatment of PC.

Combined Androgen Blockade in Localized Prostate Cancer Treated With Definitive Radiation Therapy.

BACKGROUND: The addition of androgen deprivation therapy to radiation therapy (RT) improves survival in patients with intermediate- and high-risk prostate cancer (PCa), but it is not known whether combined androgen blockade (CAB) with a gonadotropin-releasing hormone agonist (GnRH-A) and a nonsteroidal antiandrogen improves survival over GnRH-A monotherapy. METHODS: This study evaluated patients with intermediate- and high-risk PCa diagnosed in 2001 through 2015 who underwent RT with either GnRH-A alone or CAB using the Veterans Affairs Informatics and Computing Infrastructure. Associations between CAB and prostate cancer-specific mortality (PCSM) and overall survival (OS) were determined using multivariable regression with Fine-Gray and multivariable Cox proportional hazards models, respectively. For a positive control, the effect of long-term versus short-term GnRH-A therapy was tested. RESULTS: The cohort included 8,423 men (GnRH-A, 4,529; CAB, 3,894) with a median follow-up of 5.9 years. There were 1,861 deaths, including 349 resulting from PCa. The unadjusted cumulative incidences of PCSM at 10 years were 5.9% and 6.9% for those receiving GnRH-A and CAB, respectively (P=.16). Compared with GnRH-A alone, CAB was not associated with a significant difference in covariate-adjusted PCSM (subdistribution hazard ratio [SHR], 1.05; 95% CI, 0.85-1.30) or OS (hazard ratio, 1.02; 95% CI, 0.93-1.12). For high-risk patients, long-term versus short-term GnRH-A therapy was associated with improved PCSM (SHR, 0.74; 95% CI, 0.57-0.95) and OS (SHR, 0.82; 95% CI, 0.73-0.93). CONCLUSIONS: In men receiving definitive RT for intermediate- or high-risk PCa, CAB was not associated with improved PCSM or OS compared with GnRH alone.

A Population-Based Study of Morbidity After Pancreatic Cancer Diagnosis.

BACKGROUND: Pancreatic cancer is an aggressive disease characterized by early and relentless tumor spread, thus leading healthcare providers to consider it a "distant disease." However, local pancreatic tumor progression can lead to substantial morbidity. This study defines the long-term morbidity from local and nonlocal disease progression in a large population-based cohort. METHODS: A total of 21,500 Medicare beneficiaries diagnosed with pancreatic cancer in 2000 through 2011 were identified. Hospitalizations were attributed to complications of either local disease (eg, biliary disorder, upper gastrointestinal ulcer/bleed, pain, pancreas-related, radiation toxicity) or nonlocal/distant disease (eg, thromboembolic events, cytopenia, dehydration, nausea/vomiting/motility problem, malnutrition and cachexia, ascites, pathologic fracture, and chemotherapy-related toxicity). Competing risk analyses were used to identify predictors of hospitalization. RESULTS: Of the total cohort, 9,347 patients (43.5%) were hospitalized for a local complication and 13,101 patients (60.9%) for a nonlocal complication. After adjusting for the competing risk of death, the 12-month cumulative incidence of hospitalization from local complications was highest in patients with unresectable disease (53.1%), followed by resectable (39.5%) and metastatic disease (33.7%) at diagnosis. For nonlocal complications, the 12-month cumulative incidence was highest in patients with metastatic disease (57.0%), followed by unresectable (56.8%) and resectable disease (42.8%) at diagnosis. Multivariable analysis demonstrated several predictors of hospitalization for local and nonlocal complications, including age, race/ethnicity, location of residence, disease stage, tumor size, and diagnosis year. Radiation and chemotherapy had minimal impact on the risk of hospitalization. CONCLUSIONS: Despite the widely known predilection of nonlocal/distant disease spread in pancreatic cancer, local tumor progression also leads to substantial morbidity and frequent hospitalization.

The impact of traveling distance and hospital volume on post-surgical outcomes for patients with glioblastoma.

BACKGROUND: High-volume hospitals are associated with improved outcomes in glioblastoma (GBM). However, the impact of travel burden to high-volume centers is poorly understood. We examined post-operative outcomes between GBM patients that underwent treatment at local, low-volume hospitals with those that traveled long distances to high-volume hospitals. METHODS: The National Cancer Database was queried for GBM patients that underwent surgery (2010-2014). We established two cohorts: patients in the lowest quartile of travel distance and volume (Short-travel/Low-Volume: STLV) and patients in the highest quartile of travel and volume (Long-travel/High-Volume: LTHV). Outcomes analyzed were 30-day, 90-day mortality, overall survival, 30-day readmission, and hospital length of stay. RESULTS: Of 35,529 cases, STLV patients (n = 3414) traveled a median of 3 miles (Interquartile range [IQR] 1.8-4.2) to low-volume centers (5 [3-7] annual cases) and LTHV patients (n = 3808) traveled a median of 62 miles [44.1-111.3] to high-volume centers (48 [42-71]). LTHV patients were younger, had lower Charlson scores, largely received care at academic centers (84.4% vs 11.9%), were less likely to be minorities (8.1% vs 17.1%) or underinsured (6.9% vs 12.1), and were more likely to receive trimodality therapy (75.6% vs 69.2%; all p < 0.001). On adjusted analysis, LTHV predicted improved overall survival (HR 0.87, p = 0.002), decreased 90-day mortality (OR 0.72, p = 0.019), lower 30-day readmission (OR 0.42, p < 0.001), and shorter hospitalizations (RR 0.79, p < 0.001). CONCLUSIONS: Glioblastoma patients who travel farther to high-volume centers have superior post-operative outcomes compared to patients who receive treatment locally at low-volume centers. Strategies that facilitate patient travel to high-volume hospitals may improve outcomes.

Clinical Impact of Local Progression in Pancreatic Cancer.

Background: The high prevalence of distant metastatic disease among patients with pancreatic cancer often draws attention away from the local pancreatic tumor. This study aimed to define the complications and hospitalizations from local versus distant disease progression among a retrospective cohort of patients with pancreatic cancer. Methods: Records of 298 cases of pancreatic cancer treated at a single institution from 2004 through 2015 were retrospectively reviewed, and cancer-related symptoms and complications requiring hospitalization were recorded. Hospitalizations related to pancreatic cancer were attributed to either local or distant progression. Cumulative incidence analyses were used to estimate the incidence of hospitalization, and multivariable Fine-Gray regression models were used to identify factors predictive of hospitalizations. Results: The 1-year cumulative incidences of hospitalization due to local versus distant disease progression were 31% and 24%, respectively. Among 509 recorded hospitalizations, leading local etiologies included cholangitis (10%), biliary obstruction (7%), local procedure complication (7%), and gastrointestinal bleeding (7%). On multivariable analysis, significant predictors of hospitalization from local progression included unresectable disease (subdistribution hazard ratio [SDHR], 2.42; P<.01), black race (SDHR, 3.34; P<.01), younger age (SDHR, 1.02 per year; P=.01), tumor in the pancreatic head (SDHR, 2.19; P<.01), and larger tumor size (SDHR, 1.13 per centimeter; P=.02). Most patients who died in the hospital from pancreatic cancer (56%) were admitted for complications of local disease progression. Conclusions: Patients with pancreatic cancer experience significant complications of local tumor progression. Although distant metastatic progression represents a hallmark of pancreatic cancer, future research should also focus on improving local therapies.

Management of glioblastoma at safety-net hospitals.

BACKGROUND: Safety-net hospitals (SNHs) provide disproportionate care for underserved patients. Prior studies have identified poor outcomes, increased costs, and reduced access to certain complex, elective surgeries at SNHs. However, it is unknown whether similar patterns exist for the management of glioblastoma (GBM). We sought to determine if patients treated at HBHs receive equitable care for GBM, and if safety-net burden status impacts post-treatment survival. METHODS: The National Cancer Database was queried for GBM patients diagnosed between 2010 and 2015. Safety-net burden was defined as the proportion of Medicaid and uninsured patients treated at each hospital, and stratified as low (LBH), medium (MBH), and high-burden (HBH) hospitals. The impact of safety-net burden on the receipt of any treatment, trimodality therapy, gross total resection (GTR), radiation, or chemotherapy was investigated. Secondary outcomes included post-treatment 30-day mortality, 90-day mortality, and overall survival. Univariate and multivariate analyses were utilized. RESULTS: Overall, 40,082 GBM patients at 1202 hospitals (352 LBHs, 553 MBHs, and 297 HBHs) were identified. Patients treated at HBHs were significantly less likely to receive trimodality therapy (OR = 0.75, p < 0.001), GTR (OR = 0.84, p < 0.001), radiation (OR = 0.73, p < 0.001), and chemotherapy (OR = 0.78, p < 0.001) than those treated at LBHs. Patients treated at HBHs had significantly increased 30-day (OR = 1.25, p = 0.031) and 90-day mortality (OR = 1.24, p = 0.001), and reduced overall survival (HR = 1.05, p = 0.039). CONCLUSIONS: GBM patients treated at SNHs are less likely to receive standard-of-care therapies and have increased short- and long-term mortality. Additional research is needed to evaluate barriers to providing equitable care for GBM patients at SNHs.

Breast Cancer Presenting With Intravascular Tumor Emboli in Axillary Soft Tissue: Recurrence Risk and Radiation Therapy Outcomes.

Purpose: Intravascular tumor emboli in axillary soft tissue (ITE) is a rare pathologic finding in breast cancer and is associated with higher axillary nodal disease burden. The independent prognostic and predictive value of this entity is unknown, as is the role of radiation therapy for ITE. Methods and Materials: We analyzed a prospectively maintained database of breast cancer patients treated from 1992 to 2020. Patients with ITE were matched to those without (1:2) based on propensity scores to control for potential confounding factors. Locoregional (LRR) and distant recurrence (DR) were evaluated using competing risks methods accounting for death as a competing event. Overall survival (OS) and disease-free survival (DFS) were evaluated by Cox regression models. Among patients with ITE, we also evaluated whether RT improved outcomes. Results: Among 2377 total patients, 129 had ITE, of whom 126 were propensity score matched to 252 without ITE. Median follow-up from time of surgery was 5.5 years (IQR 2.3, 9.7). There were no statistically significant differences in the 5-year incidence of LRR between groups (5.4% [95% CI, 1.6%-13%] with ITE vs 10% [95% CI, 6.7%-15%] without, P = .53) or DR (24% [95% CI, 15% 35%] with ITE vs 21% [95% CI, 16%-27%] without, P = .51). Five-year OS and DFS did not differ between groups (P > .9 for both comparisons, patients with ITE vs without ITE). In analyzing the effect of RT among patients with ITE, receipt of RT was associated with significantly improved DFS (HR, 0.34, 95% CI, 0.12-0.93, P = .04). Conclusions: Patients with ITE do not exhibit significantly worse LRR, DR, DFS, or OS compared with a propensity-score-matched cohort without ITE. However, among patients with ITE, those who received RT demonstrated significantly improved DFS. Larger studies with longer follow-up are needed to evaluate the prognostic and predictive implications of ITE.

Split course palliative radiotherapy for advanced lung cancer with 3D planning based analysis of outcome: a retrospective review.

BACKGROUND: Durable palliation of advanced lung cancer is a common objective for radiation oncologists. However, there is no consensus on how to deliver the radiation course. Herein we report our experience of using split course radiotherapy and our assessment of outcomes based on planning from three-dimensional (3D) simulation before each treatment course. METHODS: All lung cancer patients from 2006-2020 were identified. Of these, 52 patients received a split course treatment of 50-60 Gy in 18-25 fractions intended to provide durable palliation for disease not amenable to curative therapy. Treatment involved 3D planning with repeat computed tomography (CT) simulation prior to the second course. Survival and symptomatic response were analyzed via chart review. We categorized rapid responders versus non-rapid responders from the initial radiation course based on ≥30% gross tumor volume (GTV) reduction at the second CT simulation. We evaluated the impact of response on overall survival and palliative response. RESULTS: Among our cohort treated with split course palliative radiotherapy, 33 (63%) had a rapid response to initial treatment. There was no difference in survival between groups [hazard ratio (HR) =1.30, P=0.47]. There was no significant difference in palliative response rates between rapid and non-rapid responders. On multivariable analysis, only female sex (HR =0.26, P<0.01) and receipt of systemic therapy following radiotherapy (HR =0.19, P<0.01) were associated with improved survival. CONCLUSIONS: There is currently significant practice pattern variability for palliative lung radiotherapy. Split course palliative radiation of 50-60 Gy in 18-25 fractions represents an option to consider for patients with advanced lung cancer who do not undergo definitive therapy and may benefit from a higher dose regimen. Our retrospective review suggests that rapid tumor response in a split course model does not predict survival or symptomatic response. Prospective studies are needed to further define which lung cancer patients may benefit from higher dose regimens.

Impact of Radiation on Cardiovascular Outcomes in Older Resectable Esophageal Cancer Patients With Medicare.

OBJECTIVES: Preoperative radiotherapy improves outcomes for operable esophageal cancer patients, though the proximity of the heart to the esophagus puts patients at risk of radiation-induced cardiovascular disease. This study characterizes the impact of radiotherapy and different radiation techniques on cardiovascular morbidity among a cohort of esophageal cancer patients. MATERIALS AND METHODS: We identified 1125 patients aged 65 and older diagnosed between 2000 and 2011 with esophageal cancer who received surgery alone, or surgery preceded by either preoperative chemotherapy or preoperative chemoradiation from the Surveillance Epidemiology and End Results (SEER)-Medicare database. We used Medicare claims to identify severe perioperative and late cardiovascular events. Multivariable logistic regression and Fine-Gray models were used to determine the effect of presurgery treatment on the risk of perioperative and late cardiovascular disease. RESULTS: Preoperative chemotherapy or chemoradiation did not significantly increase the risk of perioperative cardiovascular complications compared with surgery alone. Patients treated with preoperative chemoradiation had a 36% increased risk of having a late cardiovascular event compared with patients treated with surgery alone (subdistribution hazard ratio [SDHR]: 1.36; P=0.035). There was no significant increase in late cardiovascular events among patients treated with preoperative chemotherapy (SDHR: 1.18; P=0.40). Among patients treated with preoperative chemoradiation, those receiving intensity modulated radiotherapy had a 68% decreased risk of having a late cardiovascular event compared with patients receiving conventional radiation (SDHR: 0.32; P=0.007). CONCLUSIONS: This study demonstrates an increased risk of cardiovascular complications among operative esophageal cancer patients treated with preoperative chemoradiation, though these risks might be reduced with more cardioprotective radiation techniques such as intensity modulated radiotherapy.

Evaluation of the Use of Cancer Registry Data for Comparative Effectiveness Research.

Importance: Researchers often analyze cancer registry data to assess for differences in survival among cancer treatments. However, the retrospective, nonrandomized design of these analyses raises questions about study validity. Objective: To examine the extent to which comparative effectiveness analyses using observational cancer registry data produce results concordant with those of randomized clinical trials. Design, Setting, and Participants: In this comparative effectiveness study, a total of 141 randomized clinical trials referenced in the National Comprehensive Cancer Network Clinical Practice Guidelines for 8 common solid tumor types were identified. Data on participants within the National Cancer Database (NCDB) diagnosed between 2004 and 2014, matching the eligibility criteria of the randomized clinical trial, were obtained. The present study was conducted from August 1, 2017, to September 10, 2019. The trials included 85 118 patients, and the corresponding NCDB analyses included 1 344 536 patients. Three Cox proportional hazards regression models were used to determine hazard ratios (HRs) for overall survival, including univariable, multivariable, and propensity score-adjusted models. Multivariable and propensity score analyses controlled for potential confounders, including demographic, comorbidity, clinical, treatment, and tumor-related variables. Main Outcomes and Measures: The main outcome was concordance between the results of randomized clinical trials and observational cancer registry data. Hazard ratios with an NCDB analysis were considered concordant if the NDCB HR fell within the 95% CI of the randomized clinical trial HR. An NCDB analysis was considered concordant if both the NCDB and clinical trial P values for survival were nonsignificant (P ≥ .05) or if they were both significant (P < .05) with survival favoring the same treatment arm in the NCDB and in the randomized clinical trial. Results: Analyses using the NCDB-produced HRs for survival were concordant with those of 141 randomized clinical trials in 79 univariable analyses (56%), 98 multivariable analyses (70%), and 90 propensity score models (64%). The NCDB analyses produced P values concordant with randomized clinical trials in 58 univariable analyses (41%), 65 multivariable analyses (46%), and 63 propensity score models (45%). No clinical trial characteristics were associated with concordance between NCDB analyses and randomized clinical trials, including disease site, type of clinical intervention, or severity of cancer. Conclusions and Relevance: The findings of this study suggest that comparative effectiveness research using cancer registry data often produces survival outcomes discordant with those of randomized clinical trial data. These findings may help provide context for clinicians and policy makers interpreting observational comparative effectiveness research in oncology.

Tobacco smoking and death from prostate cancer in US veterans.

BACKGROUND: Cigarette smoking is a risk factor for mortality in several genitourinary cancers, likely due to accumulation of carcinogens in urine. However, in prostate cancer (PC) the link has been less studied. We evaluated differences in prostate cancer-specific mortality (PCSM) between current smokers, past smokers, and never smokers diagnosed with PC. METHODS: This was a retrospective cohort study of PCSM in men diagnosed with PC between 2000 and 2015 treated in the US Veterans Affairs health care system, using competing risk regression analyses. RESULTS: The cohort included 73,668 men (current smokers: 22,608 (30.7%), past smokers: 23,695 (32.1%), and never smokers: 27,365 (37.1%)). Median follow-up was 5.9 years. Current smoker patients were younger at presentation (median age current: 63, never: 66; p < 0.001), and had more advanced disease stage (stage IV disease current: 5.3%, never: 4.3%; p < 0.04). The 10-year incidence of PCSM was 5.2%, 4.8%, and 4.5% for current, past, and never smokers, respectively. On competing risk regression, current smoking was associated with increased PCSM (subdistribution hazard ratio: 1.14, 95% confidence interval: (1.05-1.24), p = 0.002), whereas past smoking was not. Hierarchical regression suggests that this increased risk was partially attributable to tumor characteristics. CONCLUSIONS: Smoking at the time of diagnosis is associated with a higher risk of dying from PC as well as other causes of death. In contrast, past smoking was not associated with PCSM suggesting that smoking may be a modifiable risk factor. PC diagnosis may be an important opportunity to discuss smoking cessation.

Associations among statins, preventive care, and prostate cancer mortality.

BACKGROUND: Increasing evidence indicates an association between statins and reduced prostate cancer-specific mortality (PCSM). However, significant bias may exist in these studies. One particularly challenging bias to assess is the healthy user effect, which may be quantified by screening patterns. We aimed to evaluate the association between statin use, screening, and PCSM in a dataset with detailed longitudinal information. METHODS: We used the Veterans Affairs Informatics and Computing Infrastructure to assemble a cohort of patients diagnosed with prostate cancer (PC) between 2000 and 2015. We collected patient-level demographic, comorbidity, and tumor data. We also assessed markers of preventive care utilization including cholesterol and prostate specific antigen (PSA) screening rates. Patients were considered prediagnosis statin users if they had at least one prescription one or more years prior to PC diagnosis. We evaluated PCSM using hierarchical Fine-Gray regression models and all-cause mortality (ACM) using a cox regression model. RESULTS: The final cohort contained 68,432 men including 40,772 (59.6%) prediagnosis statin users and 27,660 (40.4%) nonusers. Prediagnosis statin users had higher screening rates than nonusers for cholesterol (90 vs. 69%, p < 0.001) and PSA (76 vs. 67%, p < 0.001). In the model which excluded screening, prediagnosis statin users had improved PCSM (SHR 0.90, 95% CI 0.84-0.97; p = 0.004) and ACM (HR 0.96, 95% CI 0.93-0.99; p = 0.02). However, after including cholesterol and PSA screening rates, prediagnosis statin users and nonusers showed no differences in PCSM (SHR 0.98, 95% CI 0.91-1.06; p = 0.59) or ACM (HR 1.02, 95% CI 0.98-1.05; p = 0.25). CONCLUSION: We found that statin users tend to have more screening than nonusers. When we considered screening utilization, we observed no relationship between statin use before a prostate cancer diagnosis and prostate cancer mortality.

Testosterone therapy does not increase the risks of prostate cancer recurrence or death after definitive treatment for localized disease.

BACKGROUND: The safety of testosterone therapy (TT) after definitive treatment for localized prostate cancer remains undefined. We analyzed the risks of biochemical recurrence and mortality in men receiving TT after treatment for localized prostate cancer. METHODS: Cohort analysis using the national US Veterans Affairs Informatics and Computing Infrastructure. We identified 69,984 patients with localized prostate cancer diagnosed from 2001 to 2015 treated with surgery or radiation. We coded receipt of TT after treatment as a time-dependent covariate; used the National Death Index to identify cause of death; and defined biochemical recurrence as PSA > 0.2 ng/mL after surgery and nadir + 2 ng/mL after radiation. We analyzed recurrence and mortality using cumulative incidence curves, Fine-Gray competing risk regression, and Cox regression. RESULTS: This cohort included 28,651 surgery patients and 41,333 radiation patients, of whom 469 (1.64%) and 543 (1.31%), respectively, received TT with a median follow-up of 6.95 years. Comparing testosterone users to nonusers, there were no between-group differences in biochemical recurrence, prostate cancer-specific mortality, or overall mortality after surgery [hazard ratios (HR): 1.07; HR: 0.72 (p = 0.43); and HR: 1.11 (p = 0.43), respectively] or radiation [HR: 1.07; HR: 1.02 (p = 0.95); and HR: 1.02 (p = 0.86), respectively]. Limitations included lack of detailed data on TT duration and serum testosterone concentrations. CONCLUSIONS: In this multi-ethnic national cohort, TT did not increase the risks of biochemical recurrence or prostate cancer-specific or overall mortality after surgery or radiation. These data suggest that TT is safe in appropriate men after definitive treatment of localized prostate cancer.

Cost-Effectiveness of Chimeric Antigen Receptor T-Cell Therapy in Pediatric Relapsed/Refractory B-Cell Acute Lymphoblastic Leukemia.

BACKGROUND: Chimeric antigen receptor T-cell (CAR-T) therapy is a promising new class of cancer therapy but has a high up-front cost. We evaluated the cost-effectiveness of CAR-T therapy among pediatric patients with relapsed/refractory B-cell acute lymphoblastic leukemia (B-ALL). METHODS: We built a microsimulation model for pediatric patients with relapsed/refractory B-ALL receiving either CAR-T therapy or standard of care. Outcomes included costs, quality of life (health utility), complications, and survival. We measured cost-effectiveness with the incremental cost-effectiveness ratio (ICER), with ICERs under $100 000 per quality-adjusted life-year (QALY) considered cost effective. One-way and probabilistic sensitivity analyses were used to test model uncertainty. RESULTS: Compared to standard of care, CAR-T therapy increased overall cost by $528 200 and improved effectiveness by 8.18 QALYs, resulting in an ICER of $64 600/QALY. The model was sensitive to assumptions about long-term CAR-T survival, the complete remission rate of CAR-T patients, and the health utility of long-term survivors. The base model assumed a 76.0% one-year survival with CAR-T, although if this decreased to 57.8%, then CAR-T was no longer cost effective. If the complete remission rate of CAR-T recipients decreased from 81% to 56.2%, or if the health utility of disease-free survivors decreased from 0.94 to 0.66, then CAR-T was no longer cost effective. Probabilistic sensitivity analysis found that CAR-T was cost effective in 94.8% of iterations at a willingness to pay of $100 000/QALY. CONCLUSION: CAR-T therapy may represent a cost-effective option for pediatric relapsed/refractory B-ALL, although longer follow-up of CAR-T survivors is required to confirm validity of these findings.

Association between Radical Prostatectomy and Survival in Men with Clinically Node-positive Prostate Cancer.

Evidence supporting radical prostatectomy (RP) for men with clinically node-positive (cN+) prostate cancer (PC) is limited. In a US national database, we identified 741 men with cN+ nonmetastatic PC diagnosed during 2000-2015 who underwent definitive local therapy with RP (n=78), radiotherapy (RT) with neoadjuvant androgen deprivation therapy (ADT) (n=193), or nondefinitive therapy with ADT alone (n=445) or observation (n=25). We compared PC-specific mortality (PCSM) and all-cause mortality (ACM) using multivariable Fine-Gray competing risk regression and Cox regression, respectively. Compared to nondefinitive therapy, RP was associated with significantly better PCSM (subdistribution hazard ratio [SHR] 0.32, 95% confidence interval [CI] 0.16-0.66; p=0.002) and ACM (HR 0.36, 95% CI 0.21-0.61; p<0.001). Compared to RT, RP was not associated with a significant difference in PCSM (SHR 0.47, 95% CI 0.19-1.17; p=0.1) or ACM (HR 0.88, 95% CI 0.46-1.70; p=0.71). These data suggest that RP is associated with favorable survival outcomes that appear to be superior to those for patients who did not receive definitive therapy and comparable to those for patients receiving definitive ADT/RT. Randomized trials of surgery with multimodal therapy are needed. PATIENT SUMMARY: We found that in clinically node-positive prostate cancer, radical prostatectomy was associated with a cancer-specific and overall survival benefit compared to nondefinitive therapy. Randomized clinical trials are required to determine the best treatment approach in this patient population.

Association of Treatment With 5α-Reductase Inhibitors With Time to Diagnosis and Mortality in Prostate Cancer.

Importance: 5α-Reductase inhibitors (5-ARIs), commonly used to treat benign prostatic hyperplasia, reduce serum prostate-specific antigen (PSA) concentrations by 50%. The association of 5-ARIs with detection of prostate cancer in a PSA-screened population remains unclear. Objective: To test the hypothesis that prediagnostic 5-ARI use is associated with a delayed diagnosis, more advanced disease at diagnosis, and higher risk of prostate cancer-specific mortality and all-cause mortality than use of other or no PSA-decreasing drugs. Design, Setting, and Participants: This population-based cohort study linked the Veterans Affairs Informatics and Computing Infrastructure with the National Death Index to obtain patient records for 80 875 men with American Joint Committee on Cancer stage I-IV prostate cancer diagnosed from January 1, 2001, to December 31, 2015. Patients were followed up until death or December 31, 2017. Data analysis was performed from March 2018 to May 2018. Exposures: Prediagnostic 5-ARI use. Main Outcomes and Measures: The primary outcome was prostate cancer-specific mortality (PCSM). Secondary outcomes included time from first elevated PSA (defined as PSA≥4 ng/mL) to diagnostic prostate biopsy, cancer grade and stage at time of diagnosis, and all-cause mortality (ACM). Prostate-specific antigen levels for 5-ARI users were adjusted by doubling the value, consistent with previous clinical trials. Results: Median (interquartile range [IQR]) age at diagnosis was 66 (61-72) years; median [IQR] follow-up was 5.90 (3.50-8.80) years. Median time from first adjusted elevated PSA to diagnosis was significantly greater for 5-ARI users than 5-ARI nonusers (3.60 [95% CI, 1.79-6.09] years vs 1.40 [95% CI, 0.38-3.27] years; P < .001) among patients with known prostate biopsy date. Median adjusted PSA at time of biopsy was significantly higher for 5-ARI users than 5-ARI non-users (13.5 ng/mL vs 6.4 ng/mL; P < .001). Patients treated with 5-ARI were more likely to have Gleason grade 8 or higher (25.2% vs 17.0%; P < .001), clinical stage T3 or higher (4.7% vs 2.9%; P < .001), node-positive (3.0% vs 1.7%; P < .001), and metastatic (6.7% vs 2.9%; P < .001) disease than 5-ARI nonusers. In a multivariable regression, patients who took 5-ARI had higher prostate cancer-specific (subdistribution hazard ratio [SHR], 1.39; 95% CI, 1.27-1.52; P < .001) and all-cause (HR, 1.10; 95% CI, 1.05-1.15; P < .001) mortality. Conclusions and Relevance: Results of this study demonstrate that prediagnostic use of 5-ARIs was associated with delayed diagnosis and worse cancer-specific outcomes in men with prostate cancer. These data highlight a continued need to raise awareness of 5-ARI-induced PSA suppression, establish clear guidelines for early prostate cancer detection, and motivate systems-based practices to facilitate optimal care for men who use 5-ARIs.