Search

Background: Black patients were severely under-represented in the clinical trials that led to the approval of immune checkpoint inhibitors (ICIs) for all cancers. The aim of this study was to characterise the effectiveness and safety of ICIs in Black patients., Methods: We did a retrospective cohort study of patients in the US Veterans Health Administration (VHA) system's Corporate Data Warehouse containing electronic medical records for all patients who self-identified as non-Hispanic Black or African American (referred to as Black) or non-Hispanic White (referred to as White) and received PD-1, PD-L1, CTLA-4, or LAG-3 inhibitors between Jan 1, 2010, and Dec 31, 2023. Effectiveness outcomes were overall survival, time to treatment discontinuation, and time to next treatment. The safety outcome was the frequency of immune-related adverse events; assessed among a random sample of 1000 Black patients and 1000 White patients, 892 pairs were matched on the basis of baseline characteristics using 1:1 exact matching without replacement. After manual chart review, patients who did not receive ICI therapy or who had inadequate follow-up were excluded. The adjusted effect of race on each effectiveness outcome was assessed in the whole ICI-treated cohort with propensity-weighted Cox regression with robust standard errors. Immune-related adverse events outcomes were analysed in the random matched sample with multivariable Cox regression, adjusting for baseline characteristics., Findings: We identified 26 398 patients, of whom 4943 (18·7%) patients were Black, 21 455 (81·3%) were White, 895 (3·4%) were female, 25 503 (96·6%) were male, 11 859 (45%) had non-small-cell lung cancer, and 26 045 (98·7%) received PD-1 or PD-L1 inhibitors. As of data cutoff (Aug 28, 2024), median follow-up was 40·3 months (95% CI 38·3-42·3) for Black patients and 43·9 months (43·0-45·1) for White patients. Compared with White patients, Black patients had longer time to treatment discontinuation (2-year unadjusted rates 10·7% [95% CI 9·8-11·7] for Black patients vs 8·6% [8·2-9·0] for White patients; adjusted hazard ratio [HR] 0·91, 95% CI 0·87-0·95, p<0·0001), similar time to next treatment (23·5% [22·3-24·8] for Black patients vs 25·6% [25·0-26·2] for White patients; 1·00, 0·95-1·05, p=0·96), and slightly improved overall survival (36·5% [35·2-38·1] for Black patients vs 36·5% [35·8-37·1]; 0·95, 0·90-0·99, p=0·036). 1710 patients (n=862 Black and n=848 White) were analysed for safety outcomes. Compared with White patients, Black patients had a reduced risk of all-grade immune-related adverse events (unadjusted 2-year rate 33·1% [95% CI 28·9-37·1] vs 44·1% [95% CI 39·1-48·7]; adjusted HR 0·75, 95% CI 0·62-0·90, p=0·0026), immune-related adverse events requiring treatment with systemic steroids (0·61, 0·46-0·81, p=0·00051), and immune-related adverse events resulting in permanent ICI discontinuation (0·58, 0·44-0·78, p=0·00024). In exploratory analyses of irAE subtypes, a significant risk reduction in Black patients was found for colitis (0·46, 0·27-0·76, p=0·0026) and hyperthyroidism or hypothyroidism (0·63, 0·44-0·90, p=0·011), and no significant differences were found for any other immune-related adverse event subtypes analysed. Similar results were found in analyses using a steroid-based definition of immune-related adverse events among the entire ICI-treated cohort., Interpretation: Compared with White patients, Black patients had similar ICI effectiveness and lower toxicities among those treated in the national VHA system, potentially reflecting an important difference in the therapeutic ratio (ratio of benefit to harm) of ICIs. Our findings of decreased toxicity among Black patients require further investigation to assess their generalisability., Funding: Million Veteran Program, Office of Research and Development, Veterans Health Administration and the LUNGevity foundation., Competing Interests: Declaration of interests We declare no competing interests., (Copyright Published by Elsevier Ltd.)

Background: The cumulative, health system-wide survival benefit of immune checkpoint inhibitors (ICIs) is unclear, particularly among real-world patients with limited life expectancies and among subgroups poorly represented on clinical trials. We sought to determine the health system-wide survival impact of ICIs., Methods: We identified all patients receiving PD-1/PD-L1 or CTLA-4 inhibitors from 2010 to 2023 in the national Veterans Health Administration (VHA) system (ICI cohort) and all patients who received non-ICI systemic therapy in the years before ICI approval (historical control). ICI and historical control cohorts were matched on multiple cancer-related prognostic factors, comorbidities, and demographics. The effect of ICI on overall survival was quantified with Cox regression incorporating matching weights. Cumulative life-years gained system-wide were calculated from the difference in adjusted 5-year restricted mean survival times., Results: There were 27,322 patients in the ICI cohort and 69,801 patients in the historical control cohort. Among ICI patients, the most common cancer types were NSCLC (46%) and melanoma (10%). ICI demonstrated a large OS benefit in most cancer types with heterogeneity across cancer types (NSCLC: adjusted HR [aHR] 0.56, 95% confidence interval [CI] 0.54-0.58, p < 0.001; urothelial: aHR 0.91, 95% CI 0.83-1.01, p = 0.066). The relative benefit of ICI was stable across patient age, comorbidity, and self-reported race subgroups. Across VHA, 15,859 life-years gained were attributable to ICI within 5-years of treatment, with NSCLC contributing the most life-years gained., Conclusion: We demonstrated substantial increase in survival due to ICIs across a national health system, including in patient subgroups poorly represented on clinical trials., (© 2024 The Author(s). Cancer Medicine published by John Wiley & Sons Ltd.)

In recent years, data-driven, deep-learning-based models have shown great promise in medical risk prediction. By utilizing the large-scale Electronic Health Record data found in the U.S. Department of Veterans Affairs, the largest integrated healthcare system in the United States, we have developed an automated, personalized risk prediction model to support the clinical decision-making process for localized prostate cancer patients. This method combines the representative power of deep learning and the analytical interpretability of parametric regression models and can implement both time-dependent and static input data. To collect a comprehensive evaluation of model performances, we calculate time-dependent C-statistics [Formula: see text] over 2-, 5-, and 10-year time horizons using either a composite outcome or prostate cancer mortality as the target event. The composite outcome combines the Prostate-Specific Antigen (PSA) test, metastasis, and prostate cancer mortality. Our longitudinal model Recurrent Deep Survival Machine (RDSM) achieved [Formula: see text] 0.85 (0.83), 0.80 (0.83), and 0.76 (0.81), while the cross-sectional model Deep Survival Machine (DSM) attained [Formula: see text] 0.85 (0.82), 0.80 (0.82), and 0.76 (0.79) for the 2-, 5-, and 10-year composite (mortality) outcomes, respectively. In addition to estimating the survival probability, our method can quantify the uncertainty associated with the prediction. The uncertainty scores show a consistent correlation with the prediction accuracy. We find PSA and prostate cancer stage information are the most important indicators in risk prediction. Our work demonstrates the utility of the data-driven machine learning model in prostate cancer risk prediction, which can play a critical role in the clinical decision system., (© 2022. The Author(s).)