Screening for liver cancer should be done every six months in NAFLD patients but only in those with cirrhosis, gastroenterologists have been told.
Speaking at the Australian Gastroenterology Week (AGW) annual conference in Brisbane, Professor Jacob George, director of the Storr Liver Centre at the Westmead Institute for Medical Research, said NAFLD fitted the criteria for screening.
NAFLD was clearly a large and increasing problem predicted to be the leading cause of liver transplantation, there were suitable screening tests for liver cancer and there was benefit in detecting the disease early.
He said a recent US study of 300,000 patients and 300,000 controls had shown that patients with NAFLD had an increased risk of liver cancer (0.21 v 0.02 per 1,000 per year).
And in the world’s largest study of advanced NAFLD it was clear that those with cirrhosis were at a significantly greater risk for hepatic decompensation, hepatocellular carcinoma (HCC), and death or liver transplantation over 15-20 years of follow-up compared with patients with bridging fibrosis.
The annualised rates for HCC were 4.7 in F4-A6 patients and 1.8 in F4-A5 patients compared to 0.2 in F3 fibrosis subjects.
“So screening is worthwhile at least in this hospital population and cost effective in the CTP-A6 group,” he said.
He said that other research presented at the European Association for the Study of the Liver (EASL) 2018 conference had also confirmed that the relative risk for cirrhosis and cancer increased significantly as patients transitioned from NAFLD to NASH.
Professor George said that patients with NAFLD have a worse clinical outcome than patients with hepatitis C or B, probably because they have been diagnosed at a later stage.
“We also know that because of comorbidities, perhaps because they present later. We know that 77% of patients with NAFLD are not likely to get something treatment. So they present later, have more comorbidities, and are less likely to have curative treatment.”
He said a 2014 systematic review found the benefits of HCC surveillance in chronic liver disease were uncertain – it could detect more disease but there was little data on potential harms from screening.
“So we should be doing a randomised clinical trial and that’s exactly what we tried to do back in 2011.”
The upshot was that most patients preferred to be screened and therefore refused to be randomised into the trial. At the same time, three quarters of clinicians were screening their patients.
“The data is saying it is not possible to do a randomised controlled trial on screening right now. So then to answer the question about whether we should undertake surveillance, you really need to go onto modelling.”
A 2018 study simulating 15-year follow-up of 50-year-old patients with compensated cirrhosis found surveillance reduced the risk of all cause mortality and HCC-specific mortality.
It found the number needed to screen to reduce one all-cause and one HCC-specific death over 15 years was 28 and 18 respectively.
“So does surveillance improve early detection? Yes. Essentially the earlier you are diagnosed the higher the survival rate. Screening actually does have that benefit.”
“But patients who have succumbed to decompensation derive no benefit. Basically it’s not going to be worthwhile. We shouldn’t be screening unless they are on the transplant waiting list.”
“So based on all this data, the current recommendations are we should be screening cirrhotic patients – you should do ultrasound every six months.”
Professor George said his practice was also to include alpha fetoprotein, which improves the sensitivity of detecting early stage liver cancer over ultrasound alone.
A meta-analysis comprising more than 13,000 patients in 32 studies found the sensitivity increased from 45% to 63% with the addition of alpha fetoprotein to ultrasound.