Is better treatment reducing liver cancer among people with hepatitis B?

The incidence of hepatocellular carcinoma (HCC) appears to be decreasing and mortality improving among people with chronic hepatitis B treated with suppressive antiviral therapy, according to studies presented at the recent 2016 AASLD Liver Meeting in Boston. However, liver cancer remains a major indication for liver transplants and has a negative effect on survival of people with hepatitis B.

Over years or decades chronic hepatitis B virus (HBV) infection can lead to advanced liver disease including cirrhosis and HCC, and it is a leading cause of liver cancer worldwide. Antiviral therapy using nucleoside/nucleotide analogues such as entecavir (Baraclude) or tenofovir disoproxil fumarate (Viread) is the mainstay of treatment for chronic hepatitis B. While these drugs can suppress HBV replication during therapy – and can thereby reduce the risk of liver disease progression – they usually do not lead to a cure and long-term treatment is generally required.

Yao-Chun Hsu of E-Da Hospital and colleagues looked at changes in the risk of HCC over time among people with chronic hepatitis B in Taiwan who received optimal antiviral therapy.

This retrospective cohort study looked at more than 65,000 Taiwanese residents who received entecavir or tenofovir DF between 2008 and 2013, identified in Taiwan’s national healthcare database. People treated for less than three months, those previously exposed to other nucleoside/nucleotide analogues such as lamivudine (Epivir), those with pre-existing malignancies and those who developed HCC within three months of starting therapy were excluded. The final analysis included 27,820 eligible people. Nearly three-quarters were men, the median age was 48 years, a third had cirrhosis and 7% had co-infection with hepatitis C.

A total of 802 people developed HCC over a median follow-up period of 29 months. Cumulative incidence rates during the first, second and third years were 1.82%, 3.05% and 4.06%, respectively.

HCC incidence declined significantly with more time on treatment (adjusted incidence rate ratio [IRR] 0.73, or about a 27% reduction per year). Cirrhosis was the strongest predictor of developing HCC in a multivariate analysis (adjusted hazard ratio [HR] 4.91), followed by male sex (adjusted HR 1.73), older age (adjusted HR 1.65 per decade), diabetes (adjusted HR 1.25) and hepatitis C co-infection (adjusted HR 1.23).

"The incidence rate of HCC decreased year by year in chronic hepatitis B patients on entecavir or tenofovir," the researchers concluded. They suggested their study could inform a surveillance strategy to more intensively monitor those at greatest risk, as well as identify modifiable factors to reduce liver cancer risk.

In a related study, George Papatheodoridis of the University of Athens Medical School and colleagues assessed HCC incidence beyond the first five years on entecavir or tenofovir DF. The researchers had previously shown that HCC could still develop within the first five years of treatment despite viral suppression.

This analysis included more than 1900 Caucasian adults with chronic hepatitis B at ten centres in Europe. About 70% were men, the mean age was 53 years, 18% were HBeAg-positive, most reported little alcohol use and just over a quarter had compensated cirrhosis. People with pre-existing HCC, decompensated cirrhosis, or hepatitis C or HIV co-infection were excluded.

Participants were treated with entecavir or tenofovir DF for at least a year; about 40% had previously used other antivirals and about 20% had used pegylated interferon. Of these, 1205 people who did not develop HCC within the first five years on therapy were followed for at least an additional five years.

HCC was diagnosed in 5% of participants within the first five years of therapy, and in 17 of the 1205 people (1.4%) still at risk beyond the first five years. Nine people with cirrhosis (2.8%) and eight without cirrhosis (0.9%) developed liver cancer after five years on treatment. The annual HCC incidence rate was 1.22% within the first five years and 0.73% after five years.

There was no difference in pre- and post-five year HCC incidence among people without cirrhosis, but the risk dropped significantly after five years for people with cirrhosis (from 3.21% to 1.57%). All HCC cases diagnosed after five years occurred in people who were older than 50 when they started treatment, and being over 50 was a significant predictor of late HCC.

"The difference in the HCC incidence rates between initially cirrhotic and non-cirrhotic patients becomes less pronounced after year five, when older age, especially age >50 years, represents the main risk factor for HCC development," the researchers concluded.

Dr Papatheodoridis and his team also looked at survival of people with and without cirrhosis in this cohort of 1951 people on long-term entecavir or tenofovir DF. All were treated for at least one year (median six years). Follow-up continued to April 2016.

Over a median follow-up period of six years, 37 people without cirrhosis (2.7%) and 44 people with cirrhosis (8.4%) died from any cause; 10 (0.7%) and 23 (4.4%) of these deaths were due to liver-related causes. Thirty-seven people without cirrhosis (2.7%) and 80 people with cirrhosis (15.2%) developed HCC, and eight (0.6%) and nine (1.7%), respectively, received liver transplants.

Cumulative survival probability for the whole patient population at 1, 3, 5, 8 and 10 years was high: 99.7%, 97.8%, 95.9%, 94.1% and 94.1%, respectively. However, survival rates were significantly higher for people without cirrhosis (100%, 98.5%, 97.3%, 96.2% and 96.2%) compared to those with cirrhosis (99.1%, 95.9%, 92.8%, 89.3% and 89.3%).

Development of liver cancer was the major factor affecting overall mortality – and was in fact the only factor affecting liver-related mortality in this cohort, according to the researchers.

Among people with HCC, 45.9% of people without cirrhosis and 32.5% of those with cirrhosis either died from a liver-related cause or got a liver transplant. In contrast, this occurred in just 0.01% of people without cirrhosis and 1.3% of people with cirrhosis without HCC. Having HCC gave a hazard ratio of 5.47 – more than five times higher risk of all-cause death – but after adjusting for HCC in a multivariate analysis, cirrhosis had a much smaller though still significant effect (HR 1.08).

Finally, Kellie Young of the Santa Clara Valley Medical Center in California, Robert Wong of the Alameda Health System and colleagues used data from the United Network for Organ Sharing registry to evaluate trends in liver transplant wait-list registrations, survival while wait-listed, and the likelihood of receiving transplants among adults with chronic hepatitis B in the US.

This retrospective study looked at approximately 6700 people (about 80% men) waitlisted during three time periods:

• Era 1: 1992-1996, before treatment with nucleoside/nucleotide analogues
• Era 2: 1997-2004, lamivudine and adefovir (Hepsera) available
• Era 3: 2005-2015, current therapies available (entecavir starting in 2005 and tenofovir DF in 2008).

The number of waitlisted individuals more than doubled from Era 1 (about 900) to Era 2 (about 2800), but then stabilised in Era 3 (about 3000). The proportion of white individuals fell over time (from nearly two-thirds to a third) while the proportion of Asians rose (from about a quarter to about half); black and Hispanic people accounted for a small proportion of transplant candidates across time.

Overall, about a quarter of waitlisted individuals had HCC. But the number of candidates with liver cancer rose steadily, from just 5% in Era 1 to 15% in Era 2 and 39% in Era 3, although HCC dipped somewhat in the last two years. The proportion of Asians with HCC reached two-thirds in Era 3. Dr Wong suggested this might be because HBV genotypes found in Asia (B and C) may be more likely to cause liver cancer.

During Era 1, 0.9% of waitlisted individuals died and 46.6% received transplants within a year; the average time to death on the waitlist was 1432 days and the mean time to transplantation was 273 days. During Era 2, 8.4% died and 40.3% got transplants within a year; the mean times to death and transplantation were 569 and 311 days, respectively. And during Era 3, 6.2% died while waiting and 47.5% got transplants; the mean times to death and transplantation were 350 and 178 days.

The likelihood of dying while on the waitlist was significantly higher in Era 1 compared to Era 2 (HR 4.55), but fell from Era 2 to Era 3 (HR 3.63 vs Era 1). Waitlist mortality was affected by both the number of people who died and time to death on the transplant list. Era 1 had both the smallest number of deaths and the longest mean time to death, which the researchers said might be due to selection bias in favour of healthier individuals, as transplant outcomes were poor for people with hepatitis B prior to the 1990s.

The researchers noted that studies have shown that MELD scores at the time of transplantation have increased in recent years, suggesting people on the waitlist are now sicker. But the decline in the likelihood of death on the waitlist from Era 2 to Era 3 may reflect the improvement in hepatitis B treatment. A subgroup analysis of Era 3 showed that survival increased from 2005-2007 to 2008-2011 (HR 0.77) and 2012-2015 (HR 0.61), even after controlling for disease severity.

The likelihood of receiving a transplant also fell from 2005 to 2015. "These results we believe are a testament to the improved HBV therapy as patients on the waitlist may have delayed progression of their liver disease and thus a delayed need for transplant or longer time on the transplant list," the researchers said. They also found that survival after transplantation improved in each successive era, and also within Era 3.

"While current therapies are effective in suppressing HBV and reducing risk of cirrhosis and hepatocellular carcinoma, our current study demonstrates that HBV-related hepatocellular carcinoma is still a major concern," Dr Wong said in an AASLD press release. "[W]ith the many potential HBV therapies on the horizon, it will be interesting to understand what treatment endpoints are most effective at reducing HBV disease progression and hepatocellular carcinoma (e.g. viral suppression, normalization of alanine aminotransferases and HBV surface antigen loss)."