The American College of Gastroenterology recently released a new clinical guideline on focal liver lesions. GEN’s Sarah Tilyou spoke with lead author Catherine Frenette, MD, a transplant hepatologist with Family Health Centers of San Diego and the executive director of global medical affairs for liver diseases at Gilead Sciences, about the guideline and its implications for GI practice.
GEN: What prompted the guideline?
Dr. Frenette: The last liver lesions guideline was in 2017, so it’s been a little while. Since that time, a lot of research has been done and some new discoveries were made, especially related to adenomas, so that guideline was becoming a little bit outdated. This new guideline is meant to be very practical and easy to use in the clinic to help with the patient sitting in front of you. That’s the goal: to make the life of the gastroenterologist and their advanced practice providers better and easier, so their patients get the best care possible.
GEN: What’s new in the guideline that clinicians need to know?
Dr. Frenette: As we’ve learned, many of these lesions can be diagnosed without biopsy, so one change that we really needed to make to the guideline was to include more information about the role of radiological testing in this setting. We discussed what imaging can be done to make the diagnosis and when to biopsy and when not to biopsy. With adenomas, in particular, there have been changes related to subtyping, with different subtypes having different risk profiles in terms of cancer development or hemorrhage development, and they’re treated differently because of that.
Another new aspect of this guideline is that we provide a flow-chart for liver lesions to help clinicians diagnose and manage these lesions. When they’re in clinic seeing a patient with a liver lesion, instead of trying to dig through and figure out what’s going on, they can use the flow-chart as a reference to guide them on the recommended steps and management. We didn’t write the guidelines to be read from beginning to end. We wanted to provide a flow-chart reference to direct clinicians to the specific areas in the guideline that are relevant to a particular case. We wrote each section to be able to stand on its own, so that practitioners could read the area that they are in need of for that particular day.
GEN: How might the guideline change practice?
Related Content
Dr. Frenette: The No. 1 one goal of the guideline is to emphasize the importance of subtyping adenomas so that you know how to treat them. That’s really important, so you aren’t putting patients through unnecessary procedures or just monitoring something that should be resected. Another important area is related to referral. As liver lesion experts, we are often asked whether a patient should be referred to a specialty center or can be taken care of by their gastroenterologist. We emphasized throughout the guideline which types of lesions require referral and need to be reviewed by a multidisciplinary tumor board, and which patients can stay with their gastroenterologist and don’t need to see a specialist.
GEN: What were some of the hottest points of debate among the guideline panel, and how did you resolve them?
Dr. Frenette: The problem with some liver lesions is that there are just not a lot of data around them. So, often our approach boils down to expert opinions. For instance, adenomas in men are known to have a higher chance of having a beta catenin mutation, and that results in a higher chance of malignancy. The recommendation has always been, if it’s an adenoma in a man, then it has to be resected. But there are still questions: What do you do with lesions that are small? Can you ablate them? Would that be good enough? Do you need to follow them after you ablate them? Is there still a risk if you biopsy and there is no beta catenin mutation? We don’t know. There are just not much data out there. So, we had a lot of discussion about that, so that we could make the best recommendations possible.
Another big discussion point was transplant in the case of fibrolamellar hepatocellular carcinoma (HCC). These patients are known to have a different cancer than HCC that develops in a patient with cirrhosis or hepatitis B, and they respond differently to treatments. Liver transplant is a great option for HCC that develops in a patient with cirrhosis or hepatitis B (assuming the tumor is within transplant criteria), but we aren’t as confident about which patients with fibrolamellar carcinoma, if any, will have good outcomes with liver transplant. Again, there really are just not that much data. So, we spent a lot of time discussing what to recommend in those cases, and how to word it correctly for the guideline.
The members of the guideline committee have taken care of many patients with these lesions, and without strong data in some areas, we had to apply expert opinion and common sense. If you’ve got a lesion in a man that looks like an adenoma that’s not resectable, and you biopsy it and there is no beta catenin, maybe it is acceptable to just follow it. There are no data to support that, but you’ve got to take a commonsense approach in some of these cases.
There also was a lot of discussion around hemangiomas, which are known to be benign. Historically, they have been imaged using contrast-enhanced, cross-sectional imaging such as CT or MRI to ensure the diagnosis is correct. There are some data to suggest that you often can make that diagnosis based on ultrasound. Many of the hepatologists were apprehensive about that, but the radiologist on the panel provided data showing that ultrasound often is sufficient to diagnose a hemangioma, particularly if the lesion is small and has certain criteria. This can avoid the need for additional imaging in many cases, and save the patient some unnecessary testing.
Another area is the use of contrast-enhanced ultrasound (CEUS). There are some areas where CEUS can be very, very helpful and there are data supporting it. So, we had a lot of discussion about whether we recommend it in certain diagnostic quandaries. However, because it’s not routinely available in most of the country, we kept the wording around CEUS relatively soft, suggesting it as an option rather than firmly recommending that it be used. We had to balance the value of CEUS in this setting with the practical likelihood of it being available. We expect that it will expand in use in the next years, and the next iteration of the guideline likely will have much stronger language around the use of CEUS in the diagnosis and management of liver lesions.
GEN: What are the biggest remaining gaps in evidence?
Dr. Frenette: One piece that we did not talk a lot about in the guideline is the use of artificial intelligence in radiology and liver lesion diagnosis. We’re seeing a lot of data, and I think that’s a huge gap that we need to understand and figure out.
We’re also seeing AI being used in in pathology and biopsy reading, with data coming, but it’s still a huge gap. I don’t think AI will ever fully replace an expert who looks at this every single day, but we need to determine the scenarios in which AI potentially could be useful, making what we do more efficient and more streamlined. That’s a place that we need a lot more research.
Another big data gap relates to less invasive treatment options that could be as good as historical treatments, such as resection or transplant. We know that something such as ablation for a benign liver lesion can have less morbidity for the patient, and be less costly, but we don’t know whether long-term outcomes in patients who undergo ablation for some of these lesions are as good as has been seen with resection.
The last area that needs more data is when and how to use MRI and CT. When is it the right time to use MRI versus CT, with the pluses and minuses of both. CT is cheaper; it’s shorter. MRI has no radiation and often has a little bit better resolution. When do you make that decision? And with MRI, how do you choose between standard contrast agents versus hepatobiliary contrast agents? Should every liver MRI have had a biliary contrast? We don’t have enough data to say that’s what we confidently should recommend.
There’s plenty to work on over the next five years—there always is. The more we learn, the more we want to know.
This article is from the October 2024 print issue.