Association of geographic clustering of cutaneous T-cell lymphoma in the state of Georgia with environmental exposure to benzene and trichloroethylene.

Abstract

1551 Background: Geographic clustering of CTCL has been recently reported in large registries, but its association with environment factors is unknown. Benzene and trichloroethylene (TCE) are two common carcinogenic environmental toxins associated with hematological cancers. We investigated associations between geographic clustering of CTCL incidence in the state of Georgia with benzene and TCE exposure. Methods: We obtained county-level incidence of CTCL within Georgia from the Georgia Cancer Registry between 1999-2015. To account for the demographic structure in each county, standardized incidence ratios (SIR) were calculated by dividing the observed number of cases of CTCL in Georgia by the expected number of cases using national incidence rates by age, sex, and race. Using spatial analyses, we assessed for population-adjusted county-level clustering of SIRs. We also recorded county-level exposure concentration of benzene and TCE between 1996-2014 from the EPA’s National Air Toxics Assessment database. Linear regression analyses on CTCL incidence were performed comparing SIRs to exposure levels of benzene and TCE by county. Results: Our analyses demonstrated significant geographic clustering of CTCL in Georgia (Moran’s I statistic 0.0991, p-value = 0.022). Local spatial tests revealed several statistically significant hot spots throughout Georgia, particularly around Atlanta. This clustering was strongly correlated with benzene (R2 0.0824, p-value 0.0006) and TCE (R2 0.0614, p-value 0.0016) exposure concentration. Among the four most populous counties in Georgia (Cobb, Dekalb, Fulton, and Gwinnett) CTCL incidence was 1.7 to 2.7 times higher than the average county, and benzene and TCE exposure concentration was 3.0 to 6.3 times higher. Conclusions: These results demonstrate non-random geographic clustering of CTCL incidence in Georgia. This is the first analysis to correlate geographic clustering of CTCL with environmental toxic exposures, demonstrating a statistically significant correlation between environmental exposure to benzene and TCE and CTCL incidence within Georgia. [Table: see text]