Using a nationwide population database, we investigated the association between the use of aspirin and HCC risk in HCV carriers. This is the first study to find that HCV carriers who used aspirin had a lower risk of HCC than HCV carriers who did not use aspirin (adjusted HR = 0.56, 95% CI = 0.43–0.72, p < 0.001; Table 2). Furthermore, the male HCV carriers had a 1.71-fold higher risk of HCC than the female HCV carriers (adjusted HR = 1.71, 95% CI = 1.35–2.18, p < 0.001; Table 2), although aspirin treatment significantly reduced the HCC risk in both genders (female: adjusted HR = 0.51, 95% CI = 0.35–0.76, p < 0.001; male: adjusted HR = 0.59, 95% CI = 0.42–0.83, p < 0.01; Table 3). The Kaplan-Meier curves showed that the HCV carriers treated with aspirin had a lower cumulative incidence rate of HCC than those not treated with aspirin (p < 0.0001; Figure 2). In considering the duration of aspirin use, the lowest adjusted HR for developing HCC was 0.33 (95% CI = 0.18–0.61, p < 0.001),, which was found in those aspirin users with a duration of aspirin use of 1–2 years. Overall, the results of our study demonstrate that aspirin treatment can decrease the occurrence of HCC in HCV carriers.
Taiwan is one of the HBV endemic areas and has a substantial proportion of patients with dual HBV and HCV infection . To prevent interference from HBV infections, we excluded HBV carriers in order to purify our study group. Since interferon or direct-acting anti-viral treatment might totally cure an HCV infection , our study also excluded those patients who received interferon or direct-acting anti-viral treatment before the index date. This guaranteed that the investigated patients were HCV carriers while also ensuring that any lowering of the HCC occurrence was due only to the use of aspirin.
The HCC risk in the individuals with the comorbidity of hypertension was similar to that in those without hypertension. As shown in Table 2, the comorbidity of hypertension was not associated with an increased occurrence of HCC in our study population. It is interesting, however, to note that the HCC risk among participants receiving anti-hypertensive agents was 1.94-fold higher than that among those not receiving such agents (95% CI = 1.45–2.6, p < 0.001; Table 2). Although our present findings cannot explain this conflict, Ho et al.  found that the use of angiotensin-converting enzyme inhibitors and angiotensin II receptor blockers was associated with higher HCC occurrence in patient subgroups consisting of patients with no cirrhosis, no DM, and no hyperlipidemia. This issue requires further study in the future.
The real mechanism by which aspirin reduces HCC risk in HCV carriers is not well understood. A previous study, however, revealed different pathways for the viral and non-viral carcinogenesis of HCC. In an HBV transgenic mice model, it was found that platelets facilitate immune-mediated liver damage through the accumulation of HBV-specific cytotoxic T lymphocytes (CTLs) . Sitia et al.  further found that aspirin decreased T-cell mediated inflammation, liver fibrosis, and progression to HCC in this HBV transgenic mice model. Whether the HBV-related carcinogenesis of HCC and the effect of aspirin on immunomodulation are the same as in HCV is still under investigation. However, aspirin has been found to have antioxidative and antiviral activity in HCV-expressing cells through Cu/Zn superoxide dismutase (SOD1) induction  and the downregulation of inducible nitric oxide synthase (iNOS) . Moreover, Trujillo-Murillo et al.  found that acetylsalicylic acid decreases HCV replication via the inhibition of COX–2 expression through the activation of p38 and mitogen-activated protein kinase/extracellular signal-regulated kinase kinase 1/2 (MEK1/2). In a more recent study, Yin and Zhang  investigated the effects of aspirin on the blocking of HCV entry. They discovered that aspirin degrades claudin–1, an HCV receptor, through the proteasome degradation pathway and inhibits the entry of all genotypes of HCV pseudoparticles . Taken together, past studies suggest that aspirin could reduce HCV expression and prevent HCV re-infection in patients with chronic HCV infection through COX–2-dependent and COX–2-independent pathways .
We found that aspirin lowered the HCC incidence rate in HCV carriers over the first 10 years of aspirin treatment, including a 67% reduction in the occurrence rate in cases in which the duration of aspirin treatment was less than one year (Table 4; Figure 2). Moreover, the cumulative HCC incidence rate became even lower over time. Our hypothesis is that aspirin could relieve chronic inflammation via the inhibition of the cyclooxygenase enzyme, thus lowering the occurrence of HCC at first [31–32]. However, the inhibition of the cyclooxygenase enzyme also restrains the immune system, such that an HCV infection may become uncontrolled [33–34]. Due to the resulting accumulation of liver cell damage and the duplication of HCV, the HCC incidence rate then gradually becomes higher. Nonetheless, regular close follow-up visits and aspirin treatment might help lower the incidence rate of HCC in HCV carriers, especially over the first 10 years of aspirin treatment. That said, further clinical trials are warranted to clarify the preventive effects of aspirin against HCC risk in HCV carriers.
The present study, which was based on data from National Health Insurance program in Taiwan, had the following advantages: First, the large national sample population can be taken as representative of the entire population of Taiwan from 2000 to 2013; nearly all patients in Taiwan could get the proper medical care; and only a few people would take over-the-counter drugs themselves. Hence, we could retrieve the details of patients’ information and medical records from the NHIRD. Second, cancers are categorized as serious diseases in Taiwan, such that cancer patients can apply for a “catastrophic illness card” as part of the NHI program. Most such patients’ medical visits are free when they visit outpatient departments or are admitted to a hospital, and almost all cancer patients will receive medical treatment. The present study, therefore, could identify HCV carriers and HCC patients accurately. Third, we could rather precisely collect information for 16,466 HCC patients to perform a 1:1 propensity score matching by age, sex, comorbidities, drugs, treatment course of aspirin, and index year. Due to the gross sample size and detailed/accurate matching, we could then precisely analyze the association of aspirin use with HCC risk in the HCV carriers.
There were some limitations to this study. First, the study was a retrospective study. As such, we could not trace the HCV titer in these patients and thus could not prove our hypothesis that the HCV titer might flare up because of the restrained immune response owing to aspirin. Second, there were no data regarding the aspirin doses per day of the patients, because definite data in that regard were not included in the database. However, the prescribed dose in Taiwan is usually 100 mg per day. Third, alcohol consumption also leads to a higher incidence of HCC among HCV-infected individuals. However, the effects of alcohol consumption could not be measured in this retrospective study.