Portal hypertension and variceal bleeding in patients with liver cancer: Evidence gaps for prevention and management

HCC portal hypertension imaging score derived from CT predicts re-bleeding and mortality after acute variceal bleeding

BACKGROUND/PURPOSE

Risk factors for re-bleeding and death after acute variceal bleeding (AVB) in cirrhotic HCC patients are not fully understood.We aimed to (1) explore how the combination of high-risk esophageal varices, HCC status, and portal vein tumor thrombus (i.e., HCC Portal Hypertension Imaging Score [HCCPHTIS]) helps predict increased risk of variceal re-bleeding and mortality; (2) assess predictability and reproducibility of the identified variceal re-bleeding rules.

METHODS

This prospective study included 195 HCC patients with first-time AVB and liver cirrhosis, and conducted multivariable Cox regression analysis and Kaplan-Meier analysis. Receiver operating characteristic curve analysis was calculated to find the optimal sensitivity, specificity, and cutoff values of the variables. The reproducibility of the results obtained was verified in a different but related group of patients.

RESULTS

56 patients (28.7%) had re-bleeding within 6 weeks; HCCPHTIS was an independent risk factor for variceal re-bleeding after AVB (Odd ratio, 2.330; 95% confidence interval: 1.728-3.142, p < 0.001). The positive predictive value of HCCPHTIS cut off value > 3 was 66.2%, sensitivity 83.9%, and specificity 82.3%. HCCPHTIS area under the curve was higher than Child-Pugh score (89% vs. 75%, p < 0.001). 74(37.9%) death occurred within 6 weeks; HCCPHTIS > 4 was associated with increased risk of death within 6 weeks after AVB (p < 0.001).

CONCLUSION

HCCPHTIS > 3 is a strong predictor of variceal re-bleeding within the first 6 weeks. However, patients with HCCPHTIS > 4 were at increased risk of death within 6 weeks.

Endoscopic variceal ligation versus propranolol for the primary prevention of oesophageal variceal bleeding in patients with hepatocellular carcinoma: an open-label, two-centre, randomised controlled trial

OBJECTIVE

This randomised trial aimed to address whether endoscopic variceal ligation (EVL) or propranolol (PPL) is more effective at preventing initial oesophageal variceal bleeding (EVB) in patients with hepatocellular carcinoma (HCC).

DESIGN

Patients with HCC and medium-to-large oesophageal varices (EVs) but without previous EVB were randomised to receive EVL (every 3-4 weeks until variceal eradication) or PPL (up to 320 mg daily) at a 1:1 ratio. Long-term follow-up data on EVB, other upper gastrointestinal bleeding (UGIB), non-bleeding liver decompensation, overall survival (OS) and adverse events (AEs) were analysed using competing risk regression.

RESULTS

Between June 2011 and April 2021, 144 patients were randomised to receive EVL (n=72) or PPL (n=72). In the EVL group, 7 patients experienced EVB, and 30 died; in the PPL group, 19 patients had EVB, and 40 died. The EVL group had a lower cumulative incidence of EVB (Gray's test, p=0.009) than its counterpart, with no mortality difference (Gray's test, p=0.085). For patients with Barcelona Clinic Liver Cancer (BCLC) stage A/B, EVL was better than PPL in reducing EVB (p<0.001) and mortality (p=0.003). For patients beyond BCLC stage B, between-group outcomes were similar. Other UGIB, non-bleeding liver decompensation and AEs did not differ between groups. A competing risk regression model confirmed the prognostic value of EVL.

CONCLUSION

EVL is superior to PPL in preventing initial EVB in patients with HCC. The benefits of EVL on EVB and OS may be limited to patients with BCLC stage A/B and not to those with BCLC stage C/D.

TRIAL REGISTRATION NUMBER

NCT01970748.

Management of varices but not anticoagulation is associated with improved outcome in patients with HCC and macrovascular tumour invasion

BACKGROUND & AIMS

The value of bleeding prophylaxis and anticoagulation in patients with hepatocellular carcinoma (HCC) and macrovascular tumour invasion (MVI) is unclear. We evaluated the impact of anticoagulation on thrombosis progression, bleeding events, and overall mortality, and assessed the efficacy of adequate management of varices as recommended for patients with cirrhosis.

METHODS

HCC patients with MVI who had Child-Turcotte-Pugh A-B7 were included between Q4/2002 and Q2/2022. Localization of the tumour thrombus and changes at 3-6 months were evaluated by two radiologists. Univariable and multivariable logistic/Cox regression analyses included time-dependent variables (i.e., anticoagulation, systemic therapy, non-selective beta blocker treatment).

RESULTS

Of 124 patients included (male: n = 110, 89%), MVI involved the main portal vein in 47 patients (38%), and 49 individuals (40%) had additional non-tumorous thrombus apposition. Fifty of 80 patients (63%) with available endoscopy had varices. Twenty-four individuals (19%) received therapeutic anticoagulation and 94 patients (76%) were treated with effective systemic therapies. The use of therapeutic anticoagulation did not significantly affect the course of the malignant thrombosis at 3-6 months. Systemic therapy (aHR: 0.26 [95%CI: 0.16-0.40]) but not anticoagulation was independently associated with reduced all-cause mortality. In patients with known variceal status, adequate management of varices was independently associated with reduced risk of variceal bleeding (aHR: 0.12 [95%CI: 0.02-0.71]). In the whole cohort, non-selective beta blockers were independently associated with reduced risk of variceal bleeding or death from any cause (aHR: 0.69 [95%CI: 0.50-0.96]).

CONCLUSION

Adequate bleeding prophylaxis and systemic anti-tumour therapy but not anticoagulation were associated with improved outcomes in patients with HCC and MVI.