Background/aims: Cirrhotic patients chronically treated with beta-blockers who achieve a decrease of hepatic venous pressure gradient (HVPG) > or =20% from baseline or to < or =12 mmHg have a marked reduction of first bleeding or re-bleeding. However, two HVPG measurements are needed to evaluate response. This study was aimed at investigating the predictive role of acute HVPG response to i.v. propranolol for bleeding and survival.
Methods: We retrospectively studied 166 cirrhotic patients with varices with HVPG response to i.v. propranolol (0.15 mg/kg). All patients subsequently received non-selective beta-blockers to prevent first bleeding (n=78) or re-bleeding (n=88).
Results: Thirty-seven patients developed a portal hypertension-related bleeding over 2 years of follow-up. Decrease (12%) in HVPG was the best cut-off for bleeding risk discrimination. This parameter was used to classify patients in responders (n=95) and non-responders (n=71). In primary prophylaxis (54 responders vs. 24 non-responders) the actuarial probability of bleeding was half in responders than in non-responders (12% vs. 23% at 2 years; ns). In secondary prophylaxis (41 responders vs. 47 non-responders) a good hemodynamic response was also significantly and independently associated with a 50% decrease in the probability of re-bleeding (23% at 2 years vs. 46% in non-responders; p=0.032) and a better survival (95% vs. 65%; p=0.003).
Conclusion: The evaluation of acute HVPG response to i.v. propranolol before initiating secondary prophylaxis for variceal bleeding is a useful tool in predicting the efficacy of non-selective beta-blockers. If adequately validated, this might be a more cost-effective strategy than the chronic evaluation of HVPG response and might be useful to guide therapeutic decisions in these patients.