Liver cancer screening for people with hepatitis B virus (HBV) infection should be targeted at those with cirrhosis, according to the results of a systematic review and meta-analysis published in PLOS ONE. The risk of hepatocellular carcinoma (HCC) and death were 31- and 44- times higher, respectively, in people with liver cirrhosis compared to people who did not have cirrhosis. Moreover, incidence among people without cirrhosis was well below the threshold of screening cost effectiveness.
“Cost of screening HBV patients without cirrhosis may outweigh the benefits,” comment the authors. “Additionally, it is important to balance potential benefits with potential harms of screening.”
Globally, tens of millions of people are living with chronic HBV infection. Its complications include HCC and an increased mortality risk. Antiviral treatment can improve outcomes, but not everyone with HBV has access to such therapy.
Guidelines recommend that everyone with HBV should be screened for liver cancer. However, it is uncertain which patients should be prioritised for screening. Screening is considered cost-effective if annual incidence is at least 0.2 per 100 person-years.
Investigators from Denmark performed a systematic review and meta-analysis of randomised controlled trials and observational studies reporting on incidence of HCC and death in people with untreated chronic HBV infection.
A total of 68 studies met their inclusion criteria; 49 were randomised controlled trials, the remainder case-controlled studies. A total of 27,584 people were enrolled in these studies, and they provided approximately 265,000 person-years of follow-up. The median duration of follow-up was two years.
The quality of studies was evaluated as high.
Data on incidence of HCC were reported in 57 studies. The cancer was diagnosed in 1285 of 26,687 people (5%). Annual incidence was 0.88 per 100 person-years (95% CI, 0.76-0.99).
Almost a quarter of people with cirrhosis (23%) developed HCC compared to 1% of people who did not have cirrhosis. Incidence of liver cancer among people with cirrhosis was 3.16 per 100 person-years (95% CI, 2.58-3.74), some 31-fold higher than the incidence in people without cirrhosis (0.10 per 100 person-years; 95% CI, 0.02-0.18). Co-infection with hepatitis C virus, older age and elevations in liver enzymes were also associated with increased HCC risk.
Mortality was reported in 30 studies. A total of 730 of 12,511 patients (6%) died. The overall mortality incidence was 1.26 per 100 person-years (95% CI, 1.01-1.51). Over a quarter (28%) of people with cirrhosis died, compared to 2% of people who did not have cirrhosis.
The mortality incidence among people with HCC was 4.89 per 100 person-years (95% CI, 3.16-6.63), which was 44-fold higher than that seen in people without cirrhosis (0.11 per 100 person-years; 95% CI, 0.09-0.14).
Screening for HCC had no influence on mortality.
“The combined evidence stresses the importance of risk stratification in HBV,” conclude the authors. “In non-cirrhotic patients without inflammatory activity HCC screening could be futile due to the low incidence, whereas efforts should be made to detect HCC in at-risk patients with cirrhosis, HCV coinfection, old age and inflammatory activity.”
Thiele M et al. Large variations in risk of hepatocellular carcinoma and mortality in treatment naïve hepatitis B patients: systematic review with meta-analysis. PLOS ONE, 9(9): e107177, 2014.