Purpose: Liver cancer (hepatocellular carcinoma (HCC)) incidence and mortality rates are increasing in the United States. New York City (NYC) has a high burden of liver cancer risk factors, including hepatitis C (HCV) and hepatitis B (HBV) infection, which disproportionately affect persons of low socioeconomic position. Identifying neighborhoods with HCC disparities is essential to effectively define targeted cancer control strategies.
Methods: New York State Cancer Registry data from 1 January 2001 through 31 December 2012 were matched with NYC HCV and HBV surveillance data. HCC data were aggregated to NYC Zip Code Tabulation Areas (ZCTAs). Moran's I cluster analysis, Poisson regression, and geographically weighted Poisson regression were used to identify hotspots in HCC incidence and to examine the spatial associations with viral hepatitis rates, poverty, and uninsured status.
Results: Among NYC residents, 8,827 HCC cases were diagnosed during 2001-2012. Significant clustering was detected in the HCC rates (Moran's I = 0.25) with the strongest clustering found in HCC patients with comorbid HCV infection (Moran's I = 0.47). Poverty and uninsured status were associated (p < 0.05) with increased rates of HCC patients with HBV or HCV infection. Neighborhoods with high rates of HCC without viral hepatitis infection had lower rates of poverty and uninsured status.
Conclusions: The geographic variation in HCC highlights the need for neighborhood-targeted interventions to address risk factors and barriers to care. The clusters of HCC by viral hepatitis status may serve as a basis for healthcare policymakers and practitioners to prioritize neighborhoods for cancer screening and control efforts.
Keywords: Geographically weighted regression; Hepatitis B; Hepatitis C; Hepatocellular carcinoma; Liver cancer; Moran’s I.