Surveillance for Hepatocellular Carcinoma

Richard S. Finn, MD: You mention the word “surveillance.” Liver cancer is an interesting disease because like a lot of cancers, I guess it’s curable. But you commented on people who are at high risk. Should we be checking alpha-fetoprotein in these people? Is that how we identify this disease early, or any thoughts on that?

Amit Singal, MD: With this identifiable target population and the vast difference in prognosis between early detection and advanced detection, despite advances in the advanced stage, it’s really important that we find this early so we can apply curative treatments, like you said. Surveillance is typically recommended using abdominal ultrasound with or without alpha-fetoprotein. The data suggest that alpha-fetoprotein actually does add something when you use it with ultrasound. Current recommendations, at least based on the data we have, suggest you should be using the two in combination every 6 months in any patient with cirrhosis.

Richard S. Finn, MD: Reena, you’re a hepatologist, and you’re seeing these patients before necessarily the oncology community does. What about for patients who are obese from fatty liver disease and NASH [nonalcoholic steatohepatitis]? Is ultrasound the perfect test for them?

Reena Salgia, MD: That’s a great point, Rich. You’re absolutely right. In that patient population, we’re certainly challenged with the performance characteristics of ultrasound, in the obese male population, the growing fatty liver population, and any time a patient has ascites as well, these are limitations to the characteristics of ultrasound.

Richard S. Finn, MD: One of the things that I find frustrating, though I think I understand, is the US Preventive Services Task Force [USPSTF], they make the recommendations for screening, colonoscopy, mammogram. Still, screening for liver cancer is not something they’ve recommended. Reena or Amit, can you speak to the reason for that?

Amit Singal, MD: Rich, the data for surveillance are good, but not optimal. What we have in the setting of cirrhosis is that we have many cohort studies, and largely retrospective cohort studies with limitations, like lead time and length time biases. When a randomized-controlled trial was attempted in the past, it actually failed to accrue and had to be shut down. I think without this level 1 data, the USPSTF has been reluctant to put forward a recommendation for HCC [hepatocellular carcinoma] surveillance. I think we’ve seen this in other cancers, where these recommendations are somewhat rushed out, then the USPSTF has to pull back. I think that given this controversy, for example, with prostate cancer screening, breast cancer screening in younger women, we’re finding that these organizations would rather have the data be there up front so they can make a recommendation, and everyone can feel truly confident that that is the right thing to do.

Transcript Edited for Clarity