Long-term tenofovir reduces liver cancer risk in people with hepatitis B

Long-term treatment with tenofovir (Viread) lowered the risk of developing hepatocellular carcinoma among people with chronic hepatitis B, with a notable divergence from expected rates after about  five years, according to a report presented at the 48^th International Liver Congress (EASL 2013) this week in Amsterdam.

Over years or decades, chronic hepatitis B virus (HBV) infection can cause advanced liver disease including cirrhosis and hepatocellular carcinoma (HCC), a type of primary liver cancer.

W Ray Kim from the Mayo Clinic in Rochester, Minnesota, evaluated the long-term effect of antiviral therapy using tenofovir on HCC incidence over time, using a predictive model know as REACH-B.

Prior research has shown that an older nucleoside analogue, lamivudine (Epivir or 3TC), reduces liver cancer risk by approximately 50% among chronic hepatitis B patients who achieve good viral suppression. HBV easily develops resistance to lamivudine, however, compromising its long-term effectiveness. Tenofovir has a higher barrier to resistance and can keep HBV suppressed for years.

Kim explained that they selected REACH-B, which estimates the risk of developing HCC over ten years, because it is more "granular" than other commonly used models and provides the most detailed interim predictions. The model takes into account patient age, sex, ALT level, hepatitis B 'e' antigen (HBeAg) status and HBV DNA level. It has been validated for both non-cirrhotic and cirrhotic patients.

This analysis included more than 600 participants in two large, multinational, placebo-controlled phase III tenofovir trials: Gilead Sciences' Study 102 (for HBeAg-negative patients) and Study 103 (for HBeAg-positive patients). In both trials participants were randomly assigned to treatment with tenofovir or placebo for one year, after which they could elect to continue on open-label tenofovir for eight years. Development of HCC was not a study endpoint but it was captured as an adverse event.

In this retrospective analysis, investigators estimated REACH-B scores using baseline data collected for the trials. They looked at new cases of HCC developing at any time since study randomisation and calculated a standardised incidence ratio of observed versus expected number of cases for patients with and without cirrhosis.

About three-quarters of the 634 participants were men and the mean age was approximately 40 years. About 60% were white and 30% were Asian. Just under 25% had cirrhosis, and this group was a bit older; 40% of cirrhotic patients and 60% without cirrhosis were HBeAg-positive. The mean HBV DNA level at study entry was approximately 7.7 log and about half had HBV genotype D.

A total of 14 new cases of HCC emerged during seven years of follow-up, nine in the HBeAg-negative study and five in the HBeAg-positive study. Four cases were reported during the first two years, four cases during the third year, two cases each during years 4 and 5, and one case each during years 6 and 7.

As expected, patients with cirrhosis were more likely to develop liver cancer, with an incidence rate of 4.5 at seven years – about twice the rate for non-cirrhotics. Six of the 14 people who developed HCC had cirrhosis, despite cirrhotics accounting for just one-quarter of the study population.

Looking at non-cirrhotic patients, the observed number of cases began to diverge from the rising curve of predicted cases after about 2 years, with the difference first reaching statistical significance between years 5 and 6. By year 7 the standardised incidence ratio was 0.45, indicating a 55% decrease from the predicted number of cases.

Among people with cirrhosis, development of HCC more closely followed the rising curve of expected cases. A combined analysis showed the same pattern as the non-cirrhotic group, but with a standardised incidence ratio of 0.5, or a 50% reduction.

The incidence of HCC in patients on tenofovir in these two studies was "lower than predicted by the REACH-B model", the investigators concluded. "Potent antiviral therapy with [tenofovir] may reduce risk of HCC."

"In non-cirrhotic patients, the effect of [tenofovir] becomes noticeable at approximately two years of therapy and became significant (55% reduction) at six years," they added. The tenofovir effect "was less pronounced in cirrhotic patients", but the number of people with cirrhosis in this analysis was small and the study therefore "may not rule out longer-term benefits".