Ledenko, Matthew

AbstractBackgroundThe highly variable occurrence of primary liver cancers across the United States emphasize the relevance of location‐based factors. Social determinants such as income, educational attainment, housing, and other factors may contribute to regional variations in outcomes. To evaluate their impact, this study identified and analyzed clusters of high mortality from primary liver cancers and the association of location‐based determinants with mortality across the contiguous United States.MethodsA geospatial analysis of age‐adjusted incidence and standardized mortality rates from primary liver cancers from 2000 to 2020 was performed. Local indicators of spatial association identified hot‐spots, clusters of counties with significantly higher mortality. Temporal analysis of locations with persistent poverty, defined as high (>20%) poverty for at least 30 years, was performed. Social determinants were analyzed individually or globally using composite measures such as the social vulnerability index or social deprivation index. Disparities in county level social determinants between hot‐spots and non‐hot‐spots were analyzed by univariate and multivariate logistic regression.ResultsThere are distinct clusters of liver cancer incidence and mortality, with hotspots in east Texas and Louisiana. The percentage of people living below the poverty line or Hispanics had a significantly higher odds ratio for being in the top quintile for mortality rates in comparison to other quintiles and were highly connected with mortality rates. Current and persistent poverty were both associated with an evolution from non‐hotspots to new hotspots of mortality. Hotspots were predominantly associated with locations with significant levels of socioeconomic vulnerability or deprivation.ConclusionsPoverty at a county level is associated with mortality from primary liver cancer and clusters of higher mortality. These findings emphasize the importance of addressing poverty and related socio‐economic determinants as modifiable factors in public health policies and interventions aimed at reducing mortality from primary liver cancers.

INTRODUCTION: Poverty traps, locations with multigenerational poverty, result from structural and economic factors that can affect health of residents within these locations. The aim of this study was to define poverty traps within the contiguous United States and their impact on outcomes from liver diseases or cancers. METHODS: A systematic census-tract level analysis was used to spatially define regions that encompassed poverty traps. Clusters of prevalent poverty and mortality from chronic liver diseases or liver cancers were identified. Temporal trends and the relationship between race and ethnicity, type of space and escape from poverty traps on disease mortality within hot spots were determined. RESULTS: The proportion of census tracts enduring multigenerational poverty within counties was strongly associated with mortality from liver disease or cancer. There was a highly significant clustering of persistent poverty and increased mortality. Hot spots of high-mortality areas correlated with factors related to income, ethnicity, and access to health care. Location or noneconomic individual factors such as race and ethnicity were important determinants of disparities within hot spots. Distinct groups of poverty traps were defined. The highly characteristic demographics and disease outcomes within each of these groups underscored the need for location-specific interventions. DISCUSSION: Poverty traps are a major and important spatially determined risk factor for mortality from liver diseases and cancers. Targeted location-specific interventions and economic development aimed at addressing the underlying causes of poverty and enhancing prosperity will be required to reduce mortality from liver diseases within poverty traps.

Background and Aims: Cyanobacteria are commonly found in water bodies and their production of hepatotoxins can contribute to liver damage. However, the population health effects of cyanobacteria exposure (CE) are unknown. Our objectives were to determine the effect of chronic exposure to cyanobacteria through proximity to water bodies with high cyanobacteria counts on the incidence and mortality of liver cancers, as well as to identify location-based risk factors. Approach and Results: Across the contiguous United States, regions with high cyanobacteria counts in water bodies were identified using satellite remote sensing data. The data were geospatially mapped to county boundaries, and disease mortality and incidence rates were analyzed. Distinctive spatial clusters of CE and mortality related to liver diseases or cancer were identified. There was a highly significant spatial association between CE, liver disease, and liver cancer but not between CE and all cancers. Hot spots of CE and mortality were identified along the Gulf of Mexico, eastern Texas, Louisiana, and Florida, and cold spots across the Appalachians. The social vulnerability index was identified as a major location-based determinant by logistic regression, with counties in the fourth or fifth quintiles having the highest prevalence of hot spots of CE and mortality from liver cancer. Conclusions: These findings emphasize the importance of environmental exposure to cyanobacteria as a location-based determinant of mortality from liver cancer. Public health initiatives addressing CE may be considered to reduce mortality, particularly in areas of high social vulnerability.

Abstract IntroductionLiver disease such as cirrhosis is known to cause changes in the composition of volatile organic compounds (VOC) present in patient breath samples. Previous studies have demonstrated the diagnosis of liver cirrhosis from these breath samples, but studies are limited to a handful of discrete, well-characterized compounds.MethodsWe utilized VOC profiles from breath samples from 46 individuals, 35 with cirrhosis and 11 healthy controls. A deep-neural network was optimized to discriminate between healthy controls and individuals with cirrhosis.ResultsA 1D convolutional neural network (CNN) was accurate in predicting which patients had cirrhosis with an AUC of 0.90 (95% CI: 0.75, 0.99). Shapley Additive Explanations characterized the presence of discrete, observable peaks which were implicated in prediction, and the top peaks (based on the average SHAP profiles on the test dataset) were noted.ConclusionCNNs demonstrate the ability to predict the presence of cirrhosis based on a full volatolomics profile of patient breath samples. SHAP values indicate the presence of discrete, detectable peaks in the VOC signal.