Daniel M. Keller, PhD

November 16, 2012

BOSTON — Patients in Taiwan who receive nucleoside analogues to treat hepatitis B virus (HBV) infection after liver resection for hepatocellular carcinoma (HCC) have a lower risk for tumor recurrence than patients who do not receive the drugs, according to a new study.

Surgery is the standard curative treatment for HCC, but disease-free survival is often unsatisfactory because of the high incidence of HCC recurrence, Chun-Ying Wu, MD, PhD, MPH, from the division of gastroenterology at the School of Medicine, National Yang-Ming University, in Taipei, Taiwan, told delegates here at The Liver Meeting 2012: American Association for the Study of Liver Diseases (AASLD) 63rd Annual Meeting.

The study findings were published in the November 14 issue of JAMA to coincide with the presentation.

HBV viral load is the most clinically controllable factor of the many that affect tumor recurrence. Dr. Wu previously showed that a viral load of more than 106 copies/mL conferred a 2.5-fold increased risk for HCC recurrence, compared with a lower load.

He and his colleagues therefore investigated the association between nucleoside analogue antiviral drug administration and the risk for tumor recurrence after curative surgery in HBV-infected patients.

In this nationwide cohort study, they examined the Taiwan National Health Insurance Research Database and identified 100,938 patients newly diagnosed with HCC. Patients were excluded if they were infected with hepatitis C virus or another form of hepatitis, had received antiviral drugs for more than 3 months before liver resection or for less than 3 months after resection, had received other liver therapies, or had other tumors.

The investigators identified 4569 eligible HBV patients who had undergone curative liver resection between October 2003 and September 2010. Of these, 4051 did not receive antiviral therapy (untreated group) and 518 were treated for HBV infection (with lamivudine, entecavir, or telbivudine for at least 90 days). The groups were similar for most demographic characteristics, including age (about 54.5 years) and sex (about 83% male). About half the patients used aspirin or nonsteroidal anti-inflammatory drugs (NSAIDs), and 3.3% to 3.9% used statins.

More patients in the untreated group than in the treated group had a major resection (35.3% vs 26.8%; P < .001). The mean follow-up was shorter in the untreated group than in the treated group (2.18 vs 2.64 years; P < .001), and the proportion of patients with cirrhosis was lower in the untreated group (38.7% vs 48.6%; P < .001).

In the treated group, the mean interval to the start of antiviral therapy was 1.19 years, and the mean duration of therapy was 1.45 years.

Antiviral Treatment Cut Risk for HCC Recurrence by One Third

During the observation period of the study, 43.6% of patients in the untreated group and 20.5% in the treated group developed HCC recurrence. After adjustment for competing mortality (death before HCC recurrence), the 6-year cumulative incidence of HCC recurrence in the treated group was 45.6% (95% confidence interval [CI], 36.5% - 54.6%) and in the untreated group was 54.6% (95% CI, 52.5%-56.6%; P < .001). The unadjusted number needed to treat with nucleoside analogues to prevent 1 HCC recurrence within 6 years is 12.

The 6-year risk for overall mortality was also lower in the treated group (29.0%; 95% CI, 20.0% - 38.0%) than in the untreated group (42.4%; 95% CI, 40.0% - 44.7%; P < .001). The unadjusted number of patients with HBV after liver resection needed to treat to prevent 1 death within 6 years was 8.

In a multivariate analysis, treatment with nucleoside analogues was associated with a 33% reduction in the risk for HCC recurrence (hazard ratio [HR], 0.67; 95% CI, 0.55 - 0.81; P < .001). Interestingly, a lower risk for recurrence was associated with the use of statins (HR, 0.68; 95% CI, 0.53 - 0.87; P = .002) or aspirin/NSAIDs (HR, 0.80; 95% CI, 0.73 - 0.88; P < .001). Each incremental year of nucleoside analogue use was associated with a 41% further risk reduction (HR, 0.59; 95% CI, 0.51 - 0.68; P < .001).

Cirrhosis was a predictor of higher risk for HCC recurrence. If cirrhosis was present, risk for recurrence was 21% higher than if it was absent (HR, 1.21; 95% CI, 1.04 - 1.30; P = .011).

Stratified analyses showed that nucleoside analogue treatment benefited all subgroups in terms of lowered risk for HCC recurrence. The benefit was statistically significant or trended toward benefit by age, sex, extent of resection, diabetes, statin use, aspirin/NSAID use, and metformin use. Antiviral treatment was only beneficial in patients who did not have cirrhosis.

Dr. Wu said the study was limited by its retrospective nature, and that causation between use of nucleoside analogues and reduced HCC recurrence risk cannot be inferred. In addition, lack of information on such things as family history, personal habits, diet, body mass index, and HBV DNA viral load could have confounded the results. Furthermore, some patients might have obtained nucleoside analogues on their own, or patients prescribed these drugs might not have been adherent, both of which could have skewed the results.

Some Reservations

In an accompanying editorial, Anna Lok, MD, from the division of gastroenterology and hepatology at the University of Michigan Health System in Ann Arbor, writes that curative resection is the treatment of choice for HCC, but only for solitary tumors and in the absence of cirrhosis. These criteria are met in only about 5% of cases in Western countries but in 40% of cases in Asia because of the high prevalence HCC caused by HBV, which can occur in the absence of cirrhosis. Adjuvant therapies before or after resection have been ineffective in preventing recurrences.

Early HCC recurrences are often metastases from the primary tumor; later ones are usually second primary tumors associated with underlying liver disease. As Dr. Wu noted, high serum HBV DNA correlates with recurrence, which suggests that therapy to reduce HBV replication lowers the risk for recurrence, as this study demonstrated.

Dr. Lok notes that several meta-analyses, systematic reviews, and randomized controlled trials of interferon or lamivudine have shown the benefit of antiviral therapy in preventing a first HCC in people chronically infected with HBV. Other trials of interferon and various antiviral drugs to prevent HCC recurrence after resection have yielded disappointing or mixed results, but the trials were often small and/or of short duration.

The strengths of this study include the large number of patients and lack of selection bias because the database they were drawn from covers more than 99% of Taiwan's population. The database was unlikely to miss many hospital admissions, HCC diagnoses, or prescriptions, Dr. Lok explains.

The limitations include the fact that it is not clear whether all patients had a truly "curative" resection. Also, the definition of recurrence of HCC after resection required hospitalization, so nonhospitalized patients with advanced disease and those receiving radiofrequency ablation on an outpatient basis could have been missed, she notes.

She proposes that studies be done "with longer duration of treatment and better characterization of HBV replication status and liver disease...to determine the magnitude of benefit and to clarify whether nucleoside/nucleotide analogues should be administered to all or a subset of patients after curative treatment for HBV-related HCC."

Robert Perrillo, MD, outgoing chair of the Hepatitis B Special Interest Group of the AASLD, senior hepatologist at the Baylor University Medical Center in Dallas, Texas, and member of the National Institute of Diabetes and Digestive and Kidney Diseases Hepatitis B Research Network, told Medscape Medical News that this study is significant because of the large numbers of patients, "but there are some reservations you have about data collection" in a retrospective study.

"There are several pieces of information that make this study very believable," he said. "You can argue about the percent benefit, but I think there is a reduction in cancer once you already have an established cancer [if you reduce the viral load] because those people are at risk for a second cancer.... This is a good population to target."

Dr. Perrillo noted that high HBV viremia is one of several factors affecting recurrence. "The recurrence of cancer is due to not only the virus, but presumably to innate barriers to cancer development, in that the people who have cancer may have alteration in natural killer cell activity, or they might be more genomically unstable and the virus may randomly integrate into a site and activate a proto-oncogene," he explained.

"People who have 1 primary [cancer] have a greater chance of developing another because you don't change their immunologic reaction with cancer cells; you therefore would think that people who have reduced levels of virus and are treated after resection might do better.... One of the real promotive factors for cancer is the level of the viral DNA," he said. "You want to treat the factors that you can to prevent that from occurring, and one obvious factor is the viral level."

There was no commercial funding for the study. Dr. Wu has disclosed no relevant financial relationships. Dr. Lok reports serving as an advisor to Arrowhead, Bayer, Bristol-Myers Squibb, Gilead, GlaxoSmithKline, and Roche; and receiving research grants from Bristol-Myers Squibb, Gilead, and GlaxoSmithKline. Dr. Perrillo reports being on the speakers bureau for some of the manufacturers of drugs used to treat HBV infection.

JAMA. 2012;308(18):1906-1913, 1922-1924. Abstract, Editorial

The Liver Meeting 2012: American Association for the Study of Liver Diseases (AASLD) 63rd Annual Meeting. Abstract 166. Presented November 12, 2012.