SUMMARY BackgroundAnti-viral therapy in chronic hepatitis B (CHB) is associated with a reduced risk of hepatocellular carcinoma (HCC) primary described in patients with cirrhosis.

Our goal was to examine rates and predictors for hepatocellular carcinoma (HCC) surveillance adherence and persistency, since studies of such adherence and persistency in patients with chronic hepatitis (CHB) are currently limited.Consecutive CHB patients (N = 1329) monitored for ≥1 year at 4 US clinics from January 1996 to July 2013 were retrospectively studied. Surveillance adherence was evaluated based on the American Association for the Study of Liver Diseases guidelines. Kaplan–Meier method was used to analyze surveillance persistency of 510 patients who had initially fair adherence (having at least annual surveillance imaging with further follow-up).Mean age was 48, with the majority being male (58%), Asian (92%), foreign-born (95%), and medically insured (97%). Patients with cirrhosis and those seen at university liver clinics were more likely to have optimal HCC surveillance than those without cirrhosis and those seen at community clinics (38.4% vs 21.6%, P <0.001 and 33.5% vs 14.4%, P < 0.001, respectively). HCC diagnosed in optimally adherent patients trended toward smaller tumor size (P < 0.08). On multivariate analysis also inclusive of age, sex, clinical visits, cirrhosis, clinic setting and antiviral therapy use, strong independent predictors for having at least annual imaging were a history of more frequent clinical visits (odds ratio [OR] = 2.5, P < 0.001) and university-based care (OR = 5.2, P < 0.001). Even for those with initially fair adherence, persistency dropped to 70% at 5 years.Adherence and persistency to HCC surveillance in CHB patients is generally poor. More frequent clinic visits and university-based settings were significant and strong predictors of at least annual HCC surveillance adherence.

For chronic hepatitis B (CHB), alanine aminotransferase (ALT) ≥2 × upper limit of normal (ULN) is often used as a major criteria to initiate treatment in absence of cirrhosis, though patients with lower ALT may not be free from future risk of hepatocellular carcinoma (HCC). We aimed to examine the effect of antiviral therapy on HCC incidence based on ALT levels.We performed a retrospective study on 3665 patients consisting of United States and Taiwanese REVEAL-HBV cohort who were consecutive, treatment-naïve, noncirrhotic CHB patients aged ≥40 years. Patients were categorized by ALT cutoffs (≥2 × ULN vs <2 × ULN) and subgrouped by treatment status. Kaplan–Meier and Cox proportional hazards models were used to calculate cumulative incidence and hazard ratio (HR) of HCC adjusting for REACH-B scores.A total of 202 patients developed HCC. Antiviral treatment significantly reduced HCC risk: HR 0.24, 95% confidence interval 0.10–0.58; P = 0.001. HCC incidence per 100,000 person-years was significantly higher in untreated versus treated patients, even for those with ALT < 2 × ULN: 314.46 versus 0 per 100,000 person-years, P = 0.0042. For patients with Hepatitis B Virus (HBV) Deoxyribonucleic Acid (DNA) ≥ 2000 IU/mL, the number-needed-to-treat (NNT) were 15 and 14 to prevent 1 incident HCC at year 10 for patients with ALT < 2 × ULN and ≥2 × ULN, respectively.After adjustment by REACH-B score, antiviral treatment significantly decreased HCC incidence even in patients with ALT < 2 × ULN. NNT to prevent 1 incident HCC after 10 years of therapy was low (14–15) in patients with mildly elevated HBV DNA ≥ 2000 IU/mL regardless of ALT levels.