We read with interest the article by Asahina et al., which clearly demonstrated a higher incidence of hepatocellular carcinoma (HCC) in patients with higher levels of α-fetoprotein (AFP) after interferon-based antiviral therapy. There was a surprisingly high incidence of HCC (almost 50%) in patients who achieved sustained virological response (SVR) but whose postinterferon (IFN) AFP levels were higher than 20 ng/mL.
There are two distinct patterns of HCC development after SVR. In one pattern, HCC develops after the eradication of hepatitis C virus (HCV). This pattern is associated with the residual potential for hepatocarcinogenesis after SVR, which may be signaled by elevated AFP after IFN. The other pattern involves HCC that was too minute to be detected before and just after IFN treatment, but grew enough to be visualized on imaging studies during post-SVR follow-up. Previous studies of HCC tumor volume doubling time suggest that some minute HCC tumors would take several years to be detected by imaging modalities. Some patients who achieved SVR, therefore, might have had minute, undetectable HCC at the time of SVR. Although the authors described excluding patients with HCC based on imaging studies, such modalities always have limitations in their ability to detect minute HCC (for example, <5 mm in diameter). In particular, the ability of imaging modalities to detect minute HCC was unsatisfactory during the earlier part of the study period (1990s).
Their Fig. 2F suggests a specific feature in the cumulative incidence curves for HCC based on post-IFN AFP levels in patients with SVR. Among patients who did not achieve SVR, the incidence of HCC continued to increase gradually according to the number of years after SVR. This includes patients with high post-IFN AFP levels, whose HCC incidence curves were similar to incidence curves stratified by post-IFN ALT levels. In contrast, in patients with SVR, cumulative HCC incidence curves according to post-IFN AFP levels were different. The incidence of HCC in patients with post-IFN AFP ≥20 ng/mL and with post-IFN AFP ≥10 ng/mL and <20 ng/mL increased rapidly until 3 to 4 years after SVR, with only a few patients developing HCC thereafter. This feature could be due to the detection of preexisting minute HCC after SVR. Was the post-IFN AFP elevation observed in these patients a marker of enhanced hepatocarcinogenesis or an existing HCC?
It will be difficult to determine whether elevated levels of AFP were produced by HCC without visible HCC on imaging studies. It would be interesting to check the fucosylated fraction of AFP (AFP-L3), a more specific marker that has been reported as a marker of minute HCC, if post-IFN AFP-L3 data were available.
Hidenori Toyoda, M.D., Ph.D.
Takashi Kumada, M.D., Ph.D.
Toshifumi Tada, M.D.
Department of Gastroenterology
Ogaki Municipal Hospital
Ogaki, Gifu, Japan