Hyposensitivity to vasopressin in patients with hepatitis B–related cirrhosis during acute variceal hemorrhage



It has been suggested that vasopressin given during hemorrhage may be less effective than when given during a stable state in a portal-hypertensive rat model. This study was designed to evaluate the hemodynamic response to vasopressin infusion in 25 HBsAgpositive cirrhotic patients. Nine patients had active variceal hemorrhage before vasopressin infusion, and the other 16 patients were in a stable condition at the time of infusion. The two groups of patients were similar in baseline values except that a higher heart rate was found in patients with hemorrhage (96 ± 20 vs. 73 ± 10 beats/min, mean ± S.D., p < 0.01). Thirty minutes after vasopressin infusion (0.66 units/min), hepatic venous pressure gradient significantly decreased in both bleeding and stable patients (from 21 ± 9 to 18 ± 9 mm Hg, p < 0.05; and from 18 ± 4 to 8 ± 3 mm Hg, p < 0.000 1, respectively). However, the decrease of hepatic venous pressure gradient was less obvious in bleeding patients as compared with stable patients (4 ± 3 vs. 9 ± 2 mm Hg, p < 0.0001). A significant reduction of hepatic venous pressure gradient after vasopressin infusion was found in five bleeding patients without shock (from a median of 16 mm Hg [range = 12 to 26] to 11 mm Hg [range = 6 to 18], p < 0.05), but not in four bleeding patients with shock (from 28 [range = 15 to 36] to 25 [range = 18 to 33] mm Hg, p > 0.05). Vasopressin infusion resulted in a significantly lower heart rate and cardiac output with a concomitantly higher systemic vascular resistance in both groups of patients (p < 0.05). Mean arterial pressure increased after vasopressin infusion in stable patients (p < 0.000 1), but it remained unchanged in bleeding patients (p > 0.05). These findings confirm that vasopressin given during acute variceal hemorrhage, especially in cirrhotic patients with shock, may be less effective than when given during a stable state. (HEPATOLOGY 1991;13:407–412.)