Accuracy of portal and forearm blood flow measurements in the assessment of the portal pressure response to propranolol

BACKGROUND/AIMS: The portal pressure response to propranolol varies significantly in individual patients with cirrhosis. At present, propranolol responders can be identified only by measuring the hepatic venous pressure gradient. The aims of this study were: 1) to investigate whether the noninvasive monitoring of portal blood flow by pulsed Doppler ultrasound and forearm blood flow by strain-gauge plethysmography can predict the hepatic venous pressure gradient response to propranolol in patients with cirrhosis, and 2) to analyze the factors that may influence this response. METHODS: Hemodynamic measurements were undertaken in 80 patients with cirrhosis before and after receiving propranolol (0.15 mg/kg i.v., n = 60) or placebo (n = 20). RESULTS: No changes were observed in the placebo group. Propranolol lowered (p < 0.01) hepatic venous pressure gradient from 17.6 +/- 3.8 to 14.7 +/- 3.8 mmHg, portal blood flow from 1122 +/- 363 to 897 +/- 332 ml/min and forearm blood flow from 7.52 +/- 3.1 to 6.12 +/- 2.3 ml/min%. Changes in hepatic venous pressure gradient were correlated (p < 0.01) with those of portal blood flow (r = 0.82) and forearm blood flow (r = 0.54). The reduction in hepatic venous pressure gradient was > 20% in 23 patients ("responders"). The accuracy of portal Doppler flowmetry in identifying responders was higher than that of forearm plethysmography (88.3 vs. 68.3%, p < 0.05). Multivariate analysis proved that previous variceal bleeding was the only factor independently associated with a lack of response to propranolol (relative risk 3.42, 95% CI 1.5-7.4, p < 0.01). Hepatic venous pressure gradient reduction by propranolol was higher in non-bleeders than in bleeders (-19.9 +/- 9.4 vs. -11.3 +/- 8.6%, p < 0.01). CONCLUSIONS: Portal Doppler ultrasound can be used as a reliable surrogate indicator of the hepatic venous pressure gradient response to acute propranolol administration. In addition, our study indicates that this response is mainly influenced by previous variceal hemorrhage.