Navigating Diagnostic Ambiguity and Histologic Uncertainty in HCC

For oncology clinicians, the Liver Imaging-Reporting and Data System (LI-RADS) provides a robust framework for the noninvasive diagnosis of hepatocellular carcinoma (HCC). However, clinical practice frequently presents scenarios that test the boundaries of these standardized algorithms, particularly when faced with the absence of confirmed cirrhosis or the diagnostic ambiguity of mixed histology.

In a virtual Case-Based Roundtable event, Mark Yarchoan, MD, associate professor at Johns Hopkins Hospital, and clinical experts explored the complexities of managing suspected HCC when "characteristic" imaging meets atypical patient backgrounds. The conversation highlights a critical diagnostic pivot: while LI-RADS 5 status typically bypasses the need for tissue acquisition, the lack of a cirrhosis diagnosis or the presence of metabolic dysfunction-associated steatohepatitis (MASH) may necessitate a biopsy to exclude more aggressive entities like mixed HCC-cholangiocarcinoma.

Beyond the initial diagnosis, the experts delineate the multifaceted factors that drive frontline therapy selection. From the prognostic weight of vascular invasion and extrahepatic spread to the evolving role of alpha-fetoprotein (AFP) as a biomarker in the second-line setting, this dialogue provides a practical roadmap for refining the management of primary liver malignancies.

Register today to join a Case-Based Roundtable near you.

CASE SUMMARY

• 62-year-old man presents to primary care physcian with complaints of early satiety, mild discomfort in right upper quadrant, and increasing fatigue (last 2-3 months)
• Unintentional weight loss (–6 lbs) in the last 1-2 months
• Initial workup by physician:
  • Label showed mild anemia, elevated liver enzymes: ALT 65 U/L, AST 72 U/L, AFP 850 ng/mL
  • Abdominal ultrasound revealed multiple liver lesions, referral for further workup
• Referral to GI/oncology:
  • MRI of the abdomen demonstrated 2 lesions in the right lobe (4-5 cm), mild portal hypertension, splenomegaly
  • Chest CT showed multiple bilateral lung nodes, measuring 1-2 cm, ECOG performance status 1
• Diagnosis: Metastatic unresectable HCC, nonviral etiology, Child-Pugh class A, BCLC stage C

Mark Yarchoan, MD: The guidelines state that if you have a definitive LI-RADS 5 HCC, you do not necessarily need a biopsy. However, in this specific case, you technically wouldn’t make a LI-RADS 5 diagnosis. According to the guidelines, because this patient does not have confirmed cirrhosis, you are supposed to obtain a biopsy, even if the lesion appears characteristic with appropriate arterial enhancement and delayed washout.

Now, this patient does have risk factors; they have diabetes, so they likely have undiagnosed MASH. They have characteristic lesions and a high AFP. While I don’t think it would be wrong to just get started with treatment, I would probably biopsy this. I have seen cases of mixed HCC-cholangiocarcinoma with high AFPs. In fact, I saw a patient today with an AFP of 12,000 ng/mL who actually has a cholangiocarcinoma. These tests are not perfect, so while I'm not saying it's wrong to begin treatment immediately, a biopsy provides certainty.

I think no one knows for sure how to manage mixed HCC-cholangiocarcinoma, but those of us who treat this daily generally lean toward chemotherapy. Usually, the cholangiocarcinoma is more aggressive. We would typically start with gemcitabine/cisplatin or gemcitabine/cisplatin/pembrolizumab [Keytruda]. But again, it’s not wrong to just get started.

DISCUSSION QUESTION

When you see a patient with HCC, what specific factors in your initial evaluation really drive your decision on how to treat them?

Vijay Rao, MD: Generally, the patient's performance status is critical. The presence of vascular invasion vs extrahepatic spread is also a major factor. Knowing their viral status [hepatitis B/C] helps as well. Finally, identifying whether they have cirrhosis or esophageal varices is vital; in my opinion, those are all the most important factors to consider.

DISCUSSION QUESTION

How do AFP levels factor into your thinking on this case?

Rao: Where that really plays a role for me is in the second-line setting, where ramucirumab [Cyramza] is an option for patients with high AFP, typically 400 ng/mL. But otherwise, for initial treatment, I don't really make a distinction based on the AFP level.

Yarchoan: AFP is definitely a poor prognostic marker; it tends to portend more aggressive disease. But I agree with you—it may not ultimately change your first-line treatment choice.

Melhem Jabbour, MD: When you have liver disease with significantly abnormal liver function tests [LFTs], is that a factor that would make you hesitate to use certain therapies? For example, would you use durvalumab [Imfinzi] or just monitor them? If their LFTs are high due to cirrhosis, do you monitor from there and still proceed? I know this can affect our decision—for instance, if I have a Child-Pugh B7 or B8 patient with slightly high LFTs, would you still give immunotherapies to those patients?

DISCUSSION QUESTION

How does the presence/absence of liver disease/risk factors influence prognosis and treatment selection?

Yarchoan: I think this is an evolving area. We can all agree that patients who have decompensated liver function with HCC do very poorly. It is actually shocking how poorly Child-Pugh B and C patients do.

In my mind, I make a distinction between patients who had very severe cirrhosis before they developed liver cancer vs patients who have a "normal" liver that just happens to be filled with cancer. If your bilirubin is 4 mg/dL because of biliary dilation and infiltrative disease, and your albumin is 2 g/dL because you’re sick, not eating, and have a high tumor burden, those are patients I tend to treat aggressively. However, when they have significant ascites and encephalopathy and the underlying liver function is highly cirrhotic, I tend to be more nervous about giving systemic therapy. Personally, I generally treat most Child-Pugh B patients with immunotherapy. I reserve bevacizumab [Avastin] mostly for Child-Pugh A in my own practice.

I think that's a gray zone. It really depends on whether I believe I can improve their liver function by treating the cancer. If I think they have a normal liver that’s just filled with tumor, I’m more aggressive. It is very helpful if you have LFTs from a year ago; if their liver function was normal then, I'm much more aggressive.

So, a little bit of history: when I came out of fellowship—I’m a laboratory-based researcher—I tried to pick the tumor type that had the least going on so I could focus on lab work. That was HCC, because we only had 1 drug, sorafenib ([exavar], taken twice a day. It prolonged life by about 2 to 3 months, and that was really all we had. I remember my first clinic where I saw 5 patients with HCC, and the discussion for every single one was sorafenib or hospice.

We are obviously in a very different era now. The next drug we got was lenvatinib [Lenvima] in the frontline setting. It failed to beat sorafenib in terms of overall survival but was noninferior, and it certainly beat sorafenib in secondary end points like progression-free survival, which was essentially doubled.¹

Register today to join a Case-Based Roundtable near you.

DISCLOSURES: Yarchoan receives grant/research support (to Johns Hopkins) from Bristol-Myers Squibb, Exelixis, Incyte and Genentech; receives honoraria from Exelixis, AstraZeneca, Replimune, Hepion, Lantheus, Genentech and Incyte; is the co-inventor of a patent related to neoantigen vaccines; and is a co-founder with equity of Adventris Pharmaceuticals.

REFERENCE

1.Kudo M, Finn R, Qin S, et al. Lenvatinib versus sorafenib in first-line treatment of patients with unresectable hepatocellular carcinoma: a randomised phase 3 non-inferiority trial. The Lancet, Volume 391, Issue 10126, 1163 – 1173.