Do Brain Cancer Rates Correlate with Ambient PM-Levels or with Hazardous Air Pollutant (HAP) Concentrations ?

PETER A. VALBERG (1), Chris M. Long (1)

(1) Gradient LLC, Cambridge, MA

     Abstract Number: 115
     Last modified: November 4, 2009

Abstract
The factors causing age-adjusted brain cancer mortality rates to vary considerably across the United States are unknown. No obvious patterns in demography, diet, health care, or occupation seem adequate to explain a four-fold variation. No specific lifestyle factors are known to play a part in brain cancer risk. We thus investigated geographical patterns of brain cancer mortality and their correlation with patterns of ambient-air pollution, either hazardous air pollutants (HAPs) or particulate matter (PM).

The National Cancer Institute (NCI) maintains brain-cancer mortality statistics on a nationwide basis, and for the 2001-2005 period, there were 16,080 brain cancer deaths nationwide. The nationwide average mortality rate was 4.4 per 100,000 per year, but numerous county-specific rates showed statistically significant differences, varying from more than twice this rate to about half this rate. Aside from spontaneously occurring DNA damage, the only established risk factors for brain cancer are high-dose ionizing radiation (e.g., therapeutic radiation) and inherited cancer-susceptibility genes. Numerous hypothetical risk factors have been suggested, including several chemicals, but no consistent effects have been found in analyses of risk for glioma in people exposed to chemicals via dietary intake or inhalation. Consistent with these study findings, none of the HAPs are IARC Group 1 or Group 2 carcinogens on the basis of brain cancer risk. Animal studies have suggested that nanoparticles in ambient air might reach the brain via direct translocation along olfactory nerve tracts (Elder et al, 2006), but a connection to brain cancer risk is unclear. Another hypothesis suggests viruses (e.g., SV-40, Epstein-Barr, Miller, 2009) may be risk factors for brain cancer; viruses have been found in human brain tumors, and elevated risk is seen in people with significant human-to-human contact (doctors, teachers) or human-to-animal contact (butchers, dairy farmers, livestock managers).

We analyzed patterns of brain cancer mortality against patterns of PM, or individual HAP, concentrations. For our analyses, we relied upon the PM measurement data from the USEPA nationwide database on PM air pollution (Aerometric Information Retrieval System) and, for HAPs, projections of local, ambient-air concentrations made as part of the National Air Toxics Assessment (NATA) studies. In the NATA, USEPA models air concentrations of HAPs based on known emissions from a national-scale air toxics database. Even though these concentrations are for time periods more recent than might be desired for studying etiology of brain cancer, they may be reasonable surrogates predictive of longer-term exposures. We investigated several Group 1 chemical carcinogens present in ambient air (e.g., benzene, vinyl chloride), and, because viruses are a Group 1 carcinogen, we also examined correlations between dairy-cow populations and brain cancer mortality. We report the strength of evidence regarding the correlation with null, elevated, or reduced brain-cancer mortality rates.

Elder et al. 2006. Translocation of inhaled ultrafine manganese oxide particles to the central nervous system. EHP 114:1172.
Miller G. 2009. Brain cancer. A viral link to glioblastoma? Science 323:30.