Daniel M. Keller, PhD

November 28, 2012

BOSTON — Long-term therapy with nucleoside or nucleotide analogues (NUCs) does not eliminate the risk for hepatocellular carcinoma (HCC) in patients with genotype D hepatitis B virus (HBV) infection and pre-existing cirrhosis who are negative for hepatitis B e antigen (HBeAg).

HBV genotype D is common in India, the Mediterranean region, and the Middle East; in the United States, its frequency depends on ethnicity.

Adriano Pellicelli, MD, from the liver unit at Azienda Ospedaliera San Camillo Forlanini in Rome, Italy, told delegates here at The Liver Meeting 2012: American Association for the Study of Liver Diseases 63rd Annual Meeting that NUCs can stop the progression of chronic HBV infection, the most common cause of HCC, but previous studies have indicated that HCC occurs despite durable suppression of HBV.

To evaluate the risks for and predictors of HCC in white HBeAg-negative genotype D patients treated for the first time with NUCs for 18 months or more, Dr. Pellicelli and colleagues in the Italian CLEO Group (Gruppo Epatologi Ospedalieri) performed a retrospective and prospective study of 246 patients from 10 liver units. Patients were excluded if they had previous HCC, developed HCC in the first 18 months of NUC treatment, or had co-infection with hepatitis C virus, hepatitis D virus, or HIV.

Patients were followed from January 1998 to December 2011, or until the occurrence of HCC.

Patients had HBV DNA levels evaluated every 4 to 6 months and underwent routine laboratory tests every 4 months. Patients with cirrhosis had liver ultrasound tests every 6 months, and patients with chronic hepatitis B infection had ultrasounds at least yearly. A virologic response was defined as an undetectable level of HBV DNA.

The median duration of NUC therapy was 64 months (range, 18 - 168 months) in the 162 patients with chronic HBV and 62 months (range, 18 - 142 months) in the 72 patients with cirrhosis.

Cirrhosis and Older Age Predict HCC

Over a median follow-up of 6.8 years, 22 patients developed HCC (12 patients with decompensated cirrhosis were excluded from the analysis). There were 2 (1.2%) cases of HCC in the chronic HBV group, for an incidence of 0.36 cases/100 person-years. There were 20 (27.8%) cases of HCC in the group with compensated cirrhosis (4.8 cases/100 person-years).

Cirrhosis and older age predicted HCC. On multivariable Cox regression analysis, the risk for HCC in patients with cirrhosis was 13-fold greater than that in patients with chronic hepatitis B only (hazard ratio [HR], 13.3; 95% confidence interval [CI], 3.9 - 44.9; P < .01). In addition, the risk for HCC was greater for patients older than 60 years than in younger patients (HR, 4.85; 95% CI, 1.7 - 13.1; P < .01).

However, HCC risk did not correlate with virologic response or with resistance to NUCs. Although the HRs actually were less than 1 for no virologic response and for NUC resistance, neither difference was statistically significant.

Dr. Pellicelli noted that "independent hepatocellular risk factors for patients with chronic hepatitis B infection treated with nucleoside/nucleotide analogues are liver cirrhosis and age greater than 60 years despite a virologic response to therapy." He added that "long-term therapy with a nucleoside does not eliminate hepatocellular carcinoma risk in HBeAg-negative genotype D patients with liver cirrhosis."

In responding to a question from the audience, Dr. Pellicelli explained that the researchers have not yet analyzed whether the risk for HCC decreases in patients whose fibrosis has regressed after treatment.

"That's an extremely important point," session moderator David Wong, MD, clinical director of the Liver Centre at the Toronto Western Hospital and director of hepatology education the University of Toronto in Ontario, Canada, who was not involved with the study, told Medscape Medical News. "This study highlights something that's been said at this meeting over and over and over — with hepatitis B, the risk really is liver cancer, and you do not eliminate the risk of liver cancer no matter what you use.... We are not yet in a position to predict who is or who is not still at risk for liver cancer from this talk."

Dr. Wong referred to another study presented at the meeting in which fibrosis was gauged using FibroScan. That study showed a persistently high risk if the reading was in the advanced or cirrhotic range prior to treatment. The risk decreased if the FibroScan values improved during treatment. For patients who started off with lower readings, there were very few cancers, he noted.

"Personally, I believe it's the fibrosis, not the virus. The virus allows you to get to fibrosis, but it's the fibrosis that remains your risk for cancer," Dr. Wong explained. He recommends that every hepatology practice have a FibroScan machine. "I don't know how you can actually practice hepatology without assessing fibrosis, and you can't biopsy everyone."

He said that his center tries to scan every patient once a year to stratify them by their risk for cancer and to see if they are responding to antiviral therapy.

There was no commercial support for this study. Dr. Pellicelli and Dr. Wong have disclosed no relevant financial relationships.

The Liver Meeting 2012: American Association for the Study of Liver Diseases (AASLD) 63rd Annual Meeting. Abstract 21. Presented November 11, 2012.