This Issue Views 892 Citations 0 26 Invited Commentary March 18, 2019 Aspirin Use and Risk of Hepatocellular Carcinoma in Hepatitis B Rena K. Fox, MD1; Tamar H. Taddei, MD2,3; David E. Kaplan, MD4,5 Article Information JAMA Intern Med. 2019;179(5):640-641. doi:10.1001/jamainternmed.2018.8314 related articles icon Related Articles 1 Chronic hepatitis B virus (HBV) and hepatitis C virus are the primary underlying causes of HCC, with contributions from alcohol, tobacco, aflatoxin, obesity, and metabolic syndrome. In patients with chronic hepatitis C virus, the risk of HCC can be reduced rapidly with short-term, direct-acting antiviral medications. In chronic HBV, oncogenesis starts with viral DNA integration into the genome of host hepatocytes that induces a cascade of host responses, including cell injury, necrosis, and angiogenesis, reflected clinically by the levels of serum alanine aminotransferase and HBV DNA, both of which are predictors of HCC. 2 Despite the potential chemopreventive effect of NA, only patients with HBV and cirrhosis and/or with active hepatitis (high alanine aminotransferase levels, high-level HBV DNA) are candidates for NA treatment, since clinical benefits have not been demonstrated for patients without active hepatitis (alanine aminotransferase levels within the reference range). Even for treatment candidates, additional barriers to receipt of NAs include expense and disparities in access to care. Because 248 million people with chronic HBV exist worldwide,3 other, less costly chemoprevention strategies are being examined, including nonpharmacologic (coffee, tea, fruits, vegetables) and pharmacologic (statins, metformin, and aspirin) agents. 4 Two prospective US cohort studies including 133 371 persons found that regular use of standard-dose aspirin at least 2 or more times per week was associated with a 49% reduced risk of HCC. Benefits appeared with aspirin use for 5 years or more in patients with and without cirrhosis.5 A prospective trial showed that daily low-dose aspirin therapy was associated with a significantly reduced HCC risk (hazard ratio, 0.39; 95% CI, 0.17-0.91), but neither a dose-response nor duration-response relationship was observed.6 These existing studies are encouraging but problematic owing to the variable dosage of aspirin, variable duration of aspirin, variable causes of liver disease, variable patient populations, confounding by indication, and incomplete adjustment for multiple confounders, such as age, sex, cirrhosis status, alcohol use, and current or previous use of HBV antiviral therapy. 7 aim to examine the use of aspirin and HCC risk, but without some of the limitations of the previous studies.8 The authors take advantage of a national medical database in Taiwan with approximately 23 million people. The retrospective cohort design always introduces biases, such as selection and observational bias, but studies from a database this large with well-validated data on prescription history and disease diagnoses can illuminate clinically important associations. Unlike the prior studies, the investigators were able to identify a relatively homogeneous cohort of patients with a single liver disease (chronic HBV) and further identify those who were receiving consecutive aspirin therapy. These patients were propensity matched to similar patients with chronic HBV who were not receiving aspirin therapy, including matching for important factors known to influence HCC risk, such as use of NA treatment and presence of cirrhosis. Results confirmed several established associations in the field: the strongest independent predictor of HCC was cirrhosis, and the strongest independent factor to reduce the risk of HCC was the use of NAs. The aspirin findings are promising. Most importantly, aspirin was associated with a reduced HCC risk, specifically for patients with HBV who were not receiving NA therapy. The study highlights that at least 2 years of aspirin therapy were needed to lower the 5-year incidence of HCC. However, for patients already receiving NA therapy or those with cirrhosis, aspirin was not significantly associated with any further reduced HCC risk. 8 Such a low frequency appears conspicuous given that the majority of the cohort consisted of men with a mean age of 58.7 years. These data suggest that either prescribing patterns for aspirin strongly differ in Taiwan compared with the United States or there is bias in ascertainment or selection. The authors report the frequency of some of the possible conditions that may warrant aspirin therapy in the aspirin users from this cohort, but do not report the frequency of these conditions in the nonusers. In the United States, a major reason for daily low-dose aspirin therapy is primary and secondary prevention of atherosclerotic disease. In this aspirin cohort, 98.0% were using a low dose (<100 mg/d) and most (82.2%) had 1 or more risk factors for cardiovascular disease, but a significant plurality had already been diagnosed with coronary artery disease (41.2%) or cerebrovascular disease (37.6%). This finding raises the concern that cardiovascular disease–associated death could have been a significant competing risk for HCC in the population. In addition, for patients with chronic liver disease, higher serum cholesterol levels are associated with less advanced stages of hepatic fibrosis, and therefore hyperlipidemia as an indication for aspirin use could be a sign that the aspirin users had less severe liver disease than the non-aspirin users. This potential for confounding by indication could predefine the aspirin users as having a lower risk for liver morbidity and HCC, independent of aspirin prescription.
