Liver cancer remains a major challenge in the U.S. with over 41,000 new cases and 30,000 deaths annually.
Hepatocellular carcinoma (HCC), the primary form of liver cancer, develops within the context of progressive
chronic liver disease (CLD), including non-alcoholic fatty liver disease (NAFLD). The etiology of HCC is not fully
understood and up to one-third of U.S. patients have no known risk factors, suggesting yet unrecognized causes.
Cyanobacteria are ubiquitous in terrestrial and aquatic ecosystems and include a wide range of species
producing liver toxins with tumor-promoting and potentially carcinogenic properties, including microcystin (MC),
nodularin (NOD), cylindrospermopsin (CYN) —as well as the largely uncharacterized anabaenopeptins (AB). To
date, there is very little direct knowledge of the potential role of cyanobacteria and cyanotoxins in the etiology of
HCC. Our novel work from communities in the U.S. Pacific, Northeast U.S., and Central America have yielded
compelling evidence supporting the role of cyanobacteria and cyanotoxin exposure in the development of NAFLD
and HCC. Our highly novel overall hypothesis posits that cyanobacteria and cyanotoxin exposure increases the
risk of HCC independently and/or in interaction with metabolic risk factors. We propose a nested case-control
investigation of incident HCC cases (n=1,194) and matched controls (n=1,194) and incident CLD cases (n=824)
and matched controls (n=824) with prediagnostic exposure data and biospecimens (blood, urine, oral samples,
tumor tissue) from four well-characterized U.S. cohorts collectively comprised of residents of 30 U.S. states and
5 racial/ethnic groups. The oral bacterial microbiome, with a focus on cyanobacteria, will be characterized using
16S rRNA Illumina MiSeq in prediagnostic oral DNA samples from HCC (n=334) and CLD (n=164) cases and
controls and evaluated for their association with HCC and CLD risk (Aim 1). MC/NOD, CYN, and AB will be
measured by direct competitive ELISA in prediagnostic samples from HCC cases (serum n=458, urine n=216,
oral samples n=201), CLD cases (serum n=440, urine n=220, oral samples n=164), and controls and evaluated
for their association with HCC and CLD risk; a subset will be evaluated by liquid chromatography–mass
spectrometry for distinction of MC and NOD and main congeners. (Aim 2). Signatures of cyanotoxin exposure in
HCC tumors will be evaluated by Nanostring gene expression (n=98) (Aim 3). Our proposed study is highly
significant and has critical and immediate implications for public health in the U.S. and globally and may inform
future policies regarding environmental surveillance of and testing for cyanotoxins as well as strategies to
mitigate human exposure.