A 2.2 cm lesion and a portal vein thrombus in Child-Pugh B cirrhosis
A 58-year-old man with hepatitis C-related cirrhosis, sustained virological response 6 years after sofosbuvir/daclatasvir, attends six-monthly surveillance. He is Child-Pugh B (score 7: bilirubin 1.6 mg/dL, albumin 31 g/L, INR 1.4, small ascites controlled with spironolactone). Platelets are 72,000/µL, AFP 28 ng/mL, MELD 11. Ultrasound shows a 2.2 cm hypoechoic lesion in segment VII, and multiphase CT shows arterial hyperenhancement with portal venous washout and a capsule, consistent with LI-RADS 5. There is also partial thrombus in the right posterior portal vein branch. Upper endoscopy shows small varices without red signs. He asks whether he can have 'that immunotherapy' he read about. There is no history of bleeding, encephalopathy or alcohol use, and he has no other comorbidity.
1. Does this lesion require biopsy? Justify your answer, including when you would biopsy.
Read the worked answer
No biopsy is needed. In a patient with cirrhosis, a lesion of 1 cm or more showing arterial hyperenhancement, non-peripheral washout in the portal venous or delayed phase, and a capsule is diagnostically HCC by non-invasive criteria, corresponding to LI-RADS 5, with a positive predictive value above 95%. Biopsy adds risk - bleeding, tract seeding, and a small false-negative rate from sampling error - without changing management in this case.
Biopsy is reserved for specific situations: a lesion that is not typical on good-quality multiphase imaging (LI-RADS 4 or indeterminate), a patient without cirrhosis where the differential includes adenoma or metastasis, a non-hypervascular lesion, or a clinical trial requiring histology. If the imaging were atypical, a second contrast study (MRI with hepatobiliary contrast) is often more informative than immediate biopsy. AFP is not diagnostic at 28 ng/mL and should not be used to override imaging.
One caveat: the presence of portal vein thrombus does not change the diagnosis of the liver lesion but raises the question of tumour thrombus, which affects staging. Adequate imaging technique is essential: multiphase CT or MRI with thin slices, arterial and portal phases timed correctly, and review by an experienced liver radiologist. The patient should be discussed in a multidisciplinary meeting before treatment.
2. How do you determine whether the portal vein thrombus is bland or neoplastic, and why does it matter?
Read the worked answer
The distinction changes stage and treatment. A bland thrombus in a cirrhotic patient may be managed with anticoagulation and does not by itself exclude curative therapy, whereas tumour thrombus means macrovascular invasion, which is stage B or C disease and generally excludes transplantation and makes resection much less likely to be curative.
The imaging signs of tumour thrombus are enhancement of the thrombus in the arterial phase, contiguous extension from the tumour into the vein, expansion or distension of the vein, neovascularity within the thrombus, and sometimes direct invasion of the vein wall. Bland thrombus tends to be non-enhancing, may be associated with portal hypertension and sluggish flow, and does not expand the vessel. A thrombus in a branch adjacent to a segment VII tumour is suspicious but not proof; MRI with contrast, and possibly contrast-enhanced ultrasound, improve confidence. Serum AFP is supportive but not decisive, and PET-CT is unreliable in cirrhosis.
If imaging remains equivocal and the answer determines whether he is offered transplant or resection, image-guided biopsy of the thrombus can be considered, accepting bleeding risk with platelets of 72,000/µL. In this patient the thrombus is segmental in the right posterior branch, which already limits transplant eligibility regardless of its nature, so the practical decision is between liver-directed therapy and systemic therapy.
3. Given Child-Pugh B7, a single 2.2 cm lesion and segmental portal vein thrombus, what treatment options are reasonable, and what is your view on atezolizumab-bevacizumab?
Read the worked answer
This is early-stage tumour bulk with borderline liver function and probable macrovascular invasion, so no single answer is automatically correct. Resection would be considered only if the future liver remnant is adequate and portal hypertension is not severe (platelets 72,000/µL and small varices suggest clinically significant portal hypertension), and it is a stretch in Child-Pugh B. Transplantation is excluded by the Child-Pugh score and the thrombus. Ablation is attractive for a 2.2 cm lesion if it can be reached safely, but it does not address the vein.
Reasonable options are therefore image-guided ablation or transarterial chemoembolisation if liver function is preserved, stereotactic radiotherapy for a small lesion, or systemic therapy. Transarterial chemoembolisation in Child-Pugh B carries a real risk of decompensation; the decision should be individualized with bilirubin, ascites control and performance status. If the thrombus proves neoplastic, systemic therapy becomes the leading option.
Atezolizumab-bevacizumab is a standard first-line regimen for advanced HCC, but the pivotal trial excluded Child-Pugh B patients, so efficacy and safety in this man are uncertain and the evidence is observational. Bleeding risk must be addressed before bevacizumab: endoscopic band ligation of varices with red signs and avoidance of bevacizumab for at least 6 months after a bleed. Immunotherapy also raises questions about autoimmune toxicity in a cirrhotic patient. My approach would be multidisciplinary review, treatment of varices if indicated, and a candid discussion of benefit and uncertainty rather than simply acceding to his request.