Abstract To assess risk factors that contribute to lung cancer burden in the Abramson Cancer Center (ACC) catchment area, we integrated geospatial data of exposure to pollutants from publicly available EPA and NASA datasets. The study area covers the 421 zip codes that make up the 12 counties of the catchment area from which most of the ACC patients come. The counties include 5 that surround Philadelphia, 6 in New Jersey, and 1 in Delaware. Environmental exposure data, sourced from US-EPA Air Quality System (AQS) Data Mart, were focused on air pollutants since air pollution is recognized by the International Agency on Cancer (IARC) as a Group 1 human carcinogen. Exposomics data included: hourly, daily, and annual (1980 -2018) PM2.5, PM10, NO2; Hazardous Air Pollutants (HAPS); Volatile Organic Compounds (VOCs); (Air Quality Index) AQI; NONOxNOy monitoring; and annual Toxic Release Inventory (TRI) air emissions by chemical classifier and point source (1987 -2017). Annual NASA satellite-derived grids were incorporated for PM2.5 (1998-2016; 1 km resolution) and NOx (1997 - 2012; 10 km resolution). ESRI’s ArcGIS was used to develop programming scripts to automate the process of data integration, geocoding, and classifying chemical parameters by (1) status as a lung carcinogen with sufficient evidence of lung carcinogenesis; (2) status as one of the priority 16 EPA polycyclic aromatic hydrocarbons, as a surrogate marker of exposure to carcinogens; (3) status in the IARC rankings for Cancer Group; (4) status as a component of diesel exhaust; and (5) status as a VOC. 1-km search radius kernel density grids were generated for each air pollutant. We sliced the density estimates into ordinal rankings ranging from “10 = high” to “1 = low.” A hazard index may be generated by summing data layers of cumulative environmental exposomics in a process called map algebra. Spatial sorting and merging of exposome releases by facility, year, chemical and zip code concentration allow for addressing “low-hanging fruit” through summary statistics. Although the focus of this investigation is on lung cancer, the utility of the methodology may be applied to probe exposures related to other cancers. Incorporating more years or larger geographic areas of study may make exploring the risk of exposure possible for less prevalent cancers. In future studies, we are conducting statistical analysis to determine whether geocoded exposure data predict lung cancer risks in those vulnerable zip codes using electronic health record data of geolocations of lung cancer patients. This novel approach will help determine whether geocoded exposomics data are associated with cancer incidence. The hazard index was used to identify zip codes that are the most vulnerable to carcinogen exposure. Zip codes 19720, 19061, 08066, 08027, 19153, and 19145 scored highest on the hazard index based on cumulative exposure. (Supported by P30-CA-016520 and P30-ES013508.) This abstract is also being presented as Poster A08. Citation Format: Thomas P. McKeon, Vicky Tam, Wei-Ting Hwang, Paul Wileyto, Karen Glanz, Trevor M. Penning. Geocoding and integrating multiple environmental exposomics sources: Assessing population hazard to lung carcinogens in 421 zip codes of a cancer center catchment area [abstract]. In: Proceedings of the AACR Special Conference on Modernizing Population Sciences in the Digital Age; 2019 Feb 19-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2020;29(9 Suppl):Abstract nr PR06.
Read full abstract