Double-blind randomized controlled trial comparing terlipressin and somatostatin for acute variceal hemorrhage. Variceal Bleeding Study Group

GASTROENTEROLOGY 1996;111:1291–1299

LIVER, PANCREAS, AND BILIARY TRACT Double-Blind Randomized Controlled Trial Comparing Terlipressin and Somatostatin for Acute Variceal Hemorrhage ˜ ARES,§ RAMON PLANAS,x JAUME BOSCH,* and FAUST FEU,* LUIS RUIZ DEL ARBOL,‡ RAFAEL BAN MEMBERS OF THE VARICEAL BLEEDING STUDY GROUP *Liver Unit, Hospital Clinic i Provincial, Department of Medicine, University of Barcelona, Barcelona; ‡Department of Gastroenterology, Hospital Ramo´n y Cajal, University of Alcala´ de Henares, Madrid; §Department of Gastroenterology, Hospital General Gregorio Maran ˜o´n, Universidad Complutense, Madrid; and xDepartment of Gastroenterology, Hospital Universitari Germans Trias i Pujol, Badalona, Spain

Background & Aims: Terlipressin and somatostatin decrease portal pressure and have been used to treat variceal hemorrhage, but they have not been adequately compared. The aim of this study was to compare the efficacy and safety of these drugs in the treatment of variceal bleeding in cirrhotic patients. Methods: Of 161 patients with variceal bleeding, 80 were randomized to receive (double-blind) intravenous terlipressin (2 mg/4 h) and 81 to receive somatostatin (continuous infusion of 250 mg/h after an intravenous injection of 250 mg). Success of therapy was defined as a 24-hour bleeding-free period within 48 hours from randomization. Results: Success of therapy was similar with terlipressin (80%) and somatostatin (84%). In patients with Child’s class A and B disease, terlipressin was effective in 52 of 60 (87%) and somatostatin in 48 of 55 (87%). Success rates in class C were 60% and 77% (P Å 0.33). No differences were observed in rebleeding rates (30% vs. 28.4%) and 6-week mortality rates (13 vs. 13 patients). Incidence of side effects was significantly higher in the terlipressin group (38.8% vs. 23.5%; P Å 0.042). Severe side effects requiring intervention occurred in 5 of 80 and 4 of 81 patients, respectively. Conclusions: Terlipressin and somatostatin are highly effective as first-line treatment of variceal hemorrhage in cirrhotic patients. The low incidence of severe side effects suggests that drug therapy may be maintained for longer periods to prevent early rebleeding.

V

ariceal hemorrhage is a leading cause of death in patients with cirrhosis. Despite the advances introduced in the treatment of this complication, especially endoscopic sclerotherapy, the mortality rate remains high, averaging 35% in recent series.1,2 Because of the high mortality rate and the difficulties to provide specialized treatments such as sclerotherapy at admission, there has been an increased effort in the search of effective pharmacological agents for the treatment of variceal hemorrhage.3 / 5E13$$0011

10-02-96 22:14:19

gasa

Terlipressin (triglycyl-lysine vasopressin or glypressin) is a long-acting vasopressin derivative with vasoconstricting activity that is slowly transformed to vasopressin by enzymatic cleavage by tissue peptidases after intravenous administration.4,5 Terlipressin therapy has lower cardiovascular toxicity than vasopressin infusion.6 Moreover, contrary to vasopressin, terlipressin has no fibrinolytic effects.7 Hemodynamic studies have shown that terlipressin causes a marked and sustained reduction of portal pressure and portal-collateral blood flow.8 – 10 Because of its prolonged biological activity, terlipressin can be administered as intravenous injections every 4 hours.11 Randomized controlled trials have shown that terlipressin is more effective than placebo,12–14 vasopressin,15–17 or vasopressin plus nitroglycerin6 in the control of variceal bleeding and as effective as balloon tamponade18,19 with a lower incidence of side effects. More importantly, terlipressin is the only drug that has been shown to reduce mortality from variceal hemorrhage in placebocontrolled trials.14 These findings have been confirmed by meta-analysis.20 Somatostatin was introduced for the treatment of acute variceal hemorrhage because of its ability to decrease portal pressure without significant effects on the systemic hemodynamics.21 Several randomized clinical trials have shown that somatostatin may be as effective as vasopressin,22 – 27 balloon tamponade,28,29 and endoscopic sclerotherapy30,31 and that its use is not associated with significant side effects.3 However, placebo-controlled trials yield conflicting results; both significant benefit32 and lack of effect33,34 have been reported. These conflicting results may be related, at least partly, to the dose and schedule of somatostatin administration. Somatostatin boluses are known to cause dramatic, but transient, deAbbreviation used in this paper: CI, confidence interval. q 1996 by the American Gastroenterological Association 0016-5085/96/$3.00

WBS-Gastro

1292 FEU ET AL.

GASTROENTEROLOGY Vol. 111, No. 5

creases in portal pressure and gastroesophageal collateral blood flow, but these effects are much less evident during continued infusions.21,35 It has therefore been suggested that giving repeated boluses of somatostatin at the start of therapy may increase the number of patients in whom bleeding is controlled.35 To date, terlipressin and somatostatin have been only compared in a small series of patients. The preliminary report of this study suggested that both drugs had similar efficacy.36 The present study aimed to provide an objective, double-blind comparison of the efficacy and safety of terlipressin and somatostatin in the treatment of acute hemorrhage from ruptured esophagogastric varices in a large series of patients with cirrhosis of the liver.

Patients and Methods Selection of Patients The study was conducted in four teaching hospitals in Spain with extensive experience in the treatment of the complications of portal hypertension: Hospital Clinic i Provincial, Barcelona (Coordinating Center); Hospital Ramo´n y Cajal, Madrid; Hospital Gregorio Maran˜o´n, Madrid; and Hospital Germans Trias i Pujol, Badalona. Patients with histologically proven or clinically suspected liver cirrhosis admitted to the participating hospitals because of hematemesis and/or melena were included in the study if they met the following criteria: (1) clinical evidence of bleeding (hematemesis and/or melena) during the previous 24 hours; (2) endoscopically proven hemorrhage from esophagogastric varices, with bleeding considered to be variceal when emergency endoscopy (within 6 hours of admission) showed active bleeding from a varix (oozing or spurting), stigmata of recent hemorrhage (clot or ‘‘white nipple’’ over a varix), or fresh blood in the stomach and esophagogastric varices as the only potential source of bleeding; (3) age between 18 and 75 years; (4) no previous randomization in this study (therefore, patients, not episodes, could be analyzed); and (5) no previous use of vasopressin and/or somatostatin to control the bleeding episode. Patients referred from other hospitals could be included if they met the above-mentioned criteria and had not undergone endoscopic sclerotherapy in the previous 5 days. All patients gave written informed consent to participate in the study. The study protocol was approved by the ethical committee of each participating hospital. Exclusion criteria were a history of severe cardiovascular disease, including acute myocardial infarction, atrioventricular block, heart failure, chronic peripheric ischemia, and arterial hypertension (defined by a systolic blood pressure of ú170 mm Hg and/or a diastolic blood pressure of ú100 mm Hg); a known hypersensitivity to any of the study drugs; chronic renal failure; ongoing treatment for bronchial asthma; and body weight of õ40 kg.

Randomization Randomization was performed immediately after diagnostic endoscopy. Randomization was generated by computer

/ 5E13$$0011

10-02-96 22:14:19

gasa

in blocks of 8 patients, independently for each participating hospital. Patients were randomized to groups receiving either active terlipressin plus placebo of somatostatin or active somatostatin plus placebo of terlipressin under double-blind conditions. The study drugs and the randomized boxes containing the medication for the entire bleeding episode, prepared for each of the four participating centers, were provided by Ferring AB (Malmo¨, Sweden). Each box included ampules of active terlipressin and placebo of somatostatin or active somatostatin and placebo of terlipressin. Thus, patients were given terlipressin and somatostatin ampules, but only one of them was active drug. Placebo ampules were indistinguishable from those containing the active drug. At randomization, patients were stratified into two groups according to the severity of liver failure, which was determined by a modification of the Child’s classification based on the presence or absence of jaundice, ascites, encephalopathy, and wasting. Patients with none or one of these features were considered at low risk, and patients with two or more were considered at high risk. This classification was preferred to the Child– Pugh classification because the four variables were clinically available and did not cause any delay in randomization.

Clinical Assessment and General Management Before starting the trial medication, clinical history and findings of physical examination, electrocardiography, chest radiography, and complete laboratory analysis were obtained. Laboratory determinations included hematologic parameters (hematocrit and hemoglobin values, red and white blood cell counts, platelet count, and prothrombin activity), blood chemistry (concentrations of total protein, plasma albumin, plasma g-globulin, total bilirubin, alanine aminotransferase, aspartate amino transferase, alkaline phosphatase, g-glutamyltransferase, blood urea nitrogen, creatinine, uric acid, cholesterol, triglycerides, blood glucose, and sodium and potassium in plasma), and urinalysis. Patients were closely monitored during the whole period of the study. Whenever possible, the patients were maintained in intensive care units (three of the four participating Centers have Gastrointestinal Bleeding Units). A nasogastric tube was placed after emergency endoscopy for examination of the gastric content every hour during the study period. Blood pressure and heart and breath rate were recorded hourly. Hematocrit was determined every 6 hours. Hypovolemia was corrected using packed red cells, fresh frozen plasma, or plasma expanders. Blood transfusion was aimed at increasing hematocrit up to only 0.28–0.30. In addition, all patients received oral lactulose as prophylaxis of hepatic encephalopathy and nonabsorbable oral antibiotics for prevention of infections of intestinal origin.37

Treatment Protocols The study drugs were administered for a maximum of 48 hours. Terlipressin was administered as intravenous injections of 2 mg every 4 hours. Somatostatin was administered

WBS-Gastro

November 1996

TERLIPRESSIN AND SOMATOSTATIN IN BLEEDING VARICES 1293

as a continuous intravenous infusion of 250 mg/h after an initial bolus injection of 250 mg. Three additional boluses of somatostatin (or of placebo in the terlipressin group) were allowed during the study period: two boluses could be administered during the first 6 hours after starting therapy if bleeding continued, and another bolus could be administered during treatment if there was a reactivation of the hemorrhage without reaching failure criteria.

Definitions of Success and Failure of Therapy According to the guidelines set at a consensus meeting held in Baveno in 1990,38 success of therapy was defined as having achieved a 24-hour bleeding-free period within the first 48 hours after randomization shown by the following criteria: absence of hematemesis, absence of signs of hypovolemia (systolic blood pressure of õ80 mm Hg and heart rate of ú120 bpm), absence of a decrease in hematocrit of ú8 points, and absence of fresh blood in gastric aspirates within this 24-hour period. The criteria for defining the failure of therapy were hematemesis or presence of fresh blood in six consecutive hourly nasogastric aspirates with signs of hypovolemia (systolic blood pressure of õ80 mm Hg and heart rate of ú120 bpm), need to transfuse 6 or more units of blood in a period of 6 hours to maintain hemodynamic stability, and continued bleeding after the first 24 hours of therapy. When failure occurred, patients underwent alternative therapy to control the bleeding using either balloon tamponade, sclerotherapy, transjugular intrahepatic portosystemic shunting, surgery, or a combination of them. Time zero (for all analysis) was defined as the hour of admission to the hospital or the hour of first manifestation of hemorrhage in patients bleeding while in the hospital for other reasons. The trial medication was stopped after meeting success or failure criteria, if the patient died, if severe side effects were detected, or if the patient withdrew consent to continue in the study. After control of bleeding, patients underwent elective therapy to prevent recurrent bleeding using either endoscopic sclerotherapy, pharmacological therapy, or surgery. Rebleeding was defined as any further manifestation of hemorrhage after achieving 24 hours’ control of bleeding. Death related to bleeding was defined as any death occurring within 6 weeks of time zero.38

Sample Size Calculation The sample size calculation was performed according to the reported results using the trial medications. The expected hemostatic efficacy was 79% for terlipressin12 – 15 and 58% for somatostatin.22 – 25,32,33 The sample size was calculated to have enough power to detect differences in efficacy and side effects of ú21%. Assuming an a error of 5% and a b error of 20% in a two-tailed comparison, a total of 75 patients were required in each treatment group.

/ 5E13$$0011

10-02-96 22:14:19

gasa

Patients Between October 1991 and November 1992, 218 cirrhotic patients were admitted due to acute variceal bleeding in the four participating hospitals. Fifty-seven of these patients were not included in the study because of the following reasons: previous treatment of the present bleeding episode with vasopressin, somatostatin, and/or endoscopic sclerotherapy (n Å 19); previous randomization for this study (n Å 15); severe cardiovascular disease (n Å 8); age older than 75 years (n Å 11); chronic renal failure (n Å 3); and massive bleeding resulting in death before randomization in 1 patient. Thus, a total of 161 patients were included in the study: 51 at the Hospital ClıB nic i Provincial, Barcelona; 43 at the Hospital Ramo´n y Cajal, Madrid; 35 at the Hospital Gregorio Maran˜o´n, Madrid; and 32 at the Hospital Germans TrıB as i Pujol, Badalona.

Follow-up A complete evaluation, including physical examination, electrocardiography, and laboratory determinations (hematologic parameters, blood chemistry, and urinalysis), was performed 24 hours after stopping the trial medication. The outcome of patients until 6 weeks after randomization was recorded.38

Statistical Analysis Statistical analysis was performed according to an intention-to-treat basis. The BMDP statistical package (BMDP Statistical Software Inc., Los Angeles, CA) was used for this analysis. Data are reported as means { SD. Categorical variables were analyzed using the two-tailed Fisher’s Exact Test, and continuous variables were compared with the unpaired Student’s t test (or the nonparametric Mann–Whitney rank sum test for unpaired data). The Kaplan–Meier method was used for rebleeding and survival analysis at 6 weeks. Actuarial curves were compared with the log rank test. Stepwise logistic regression was used to identify independent predictive variables for failure of treatment and death. Odds ratios and the 95% confidence intervals (CIs) are shown. Survival was evaluated at 6 weeks.36 Statistical significance was established at a P value of õ0.05.

Results Eighty patients were randomized to receive terlipressin and 81 to receive somatostatin. As shown in Table 1, there were no significant differences between the groups regarding demographic data, severity of liver failure, endoscopic findings, and severity of bleeding. No differences were found in the time elapsed from first manifestation of bleeding to admission (11.3 { 18 and 8.7 { 12 hours for terlipressin and somatostatin, respectively; P Å 0.97, Mann–Whitney rank sum test) and from admission to randomization (4.2 { 2.7 and 4.1 { 3.4 hours, respectively; P Å 0.28, Mann–Whitney). WBS-Gastro

1294 FEU ET AL.

GASTROENTEROLOGY Vol. 111, No. 5

Table 1. Characteristics of Patients at Admission

Sex (M/F) Age (yr )a Cause of cirrhosis (alcoholic/nonalcoholic) Previous variceal bleeding (n) Ascites Encephalopathy Bilirubin (mg/dL)a Plasma albumin (g/L)a Child–Pugh class (A/B/C) Child–Pugh scorea Risk group (low/high) Source of bleeding Esophageal varices Gastric varices Active bleeding at endoscopy Transfusion before randomization (U)a SBP (mm Hg )a DBP (mm Hg )a Heart rate (bpm)a Hypovolemic shock Hematocrita

Terlipressin (n Å 80)

Somatostatin (n Å 81)

P value

58/22 58 { 12

61/20 56 { 12

0.72 0.29

43/37 28 33 6 3.0 { 2.6 30.3 { 6.1 22/38/20 8.0 { 1.9 52/28

44/37 24 37 11 3.5 { 3.6 30.5 { 5.7 14/41/26 8.5 { 2.0 52/29

1.0 0.50 0.63 0.30 0.36 0.87 0.26 0.11 1.0

77 3 35

75 6 35

0.49

{ { { { 6 27.9 { 0.6 119 68 98

1.1 23 14 17 7.0

0.7 { 114 { 68 { 96 { 13 27.5 {

1.0 1.2 27 14 18 7.2

0.61 0.27 0.99 0.56 0.14 0.74

SBP, systolic blood pressure; DBP, diastolic blood pressure. a Results are expressed as the mean { SD.

during the treatment period (1.8 { 1.5 and 1.9 { 1.7 U for terlipressin and somatostatin, respectively; 95% CI, 00.6 to 0.4; P Å 0.69). In the multivariate analysis (including the following variables: age, sex, cause of cirrhosis, active alcoholism, time elapsed from first manifestation of bleeding to admission and from admission to randomization, hypovolemic shock, transfusional requirements, Child–Pugh score, source of bleeding, active bleeding at endoscopy, treatment received, and hospital of inclusion), only the presence of active bleeding during emergency endoscopy (odds ratio, 1.62; 95% CI, 1.04–2.53; P Å 0.03) and the Child–Pugh score (odds ratio, 1.81; 95% CI, 1.15– 2.85; P Å 0.01) were selected as independent predictive factors of failure of treatment. The treatment received had no predictive value. In the 16 patients in whom terlipressin failed to stop bleeding, the hemorrhage was arrested by balloon tamponade in 6 patients, by sclerotherapy in 7, and after conservative therapy in 3. In the 13 patients in whom somatostatin failed to stop bleeding, hemorrhage was stopped by balloon tamponade in 4 patients, sclerotherapy in 4, a portacaval shunt in 1, and conservative therapy in 4.

Control of Bleeding

Rebleeding

Control of bleeding, defined as obtaining a 24hour bleeding-free period within 48 hours of treatment, was achieved in 64 of 80 patients (80%) in the terlipressin group and in 68 of 81 patients (84%) in the somatostatin group (P Å 0.54). The mean interval to cessation of bleeding in successfuly treated patients was similar in both groups (Table 2). In the subgroup of patients with active bleeding during emergency endoscopy, success of therapy was achieved in 23 of 35 patients (66%) in the terlipressin group and in 29 of 35 patients (83%) in the somatostatin group (P Å 0.17). Additional boluses of somatostatin were deemed necessary in 40 patients (50%) in the somatostatin group: in 30 patients (75%) during the first 6 hours after starting treatment and in 10 patients (25%) after the first 6 hours. Therapy was successful in 28 of the 40 patients (70%) receiving additional somatostatin boluses. Among the 46 patients (57%) in the terlipressin group who received placebo boluses, therapy was successful in 32 patients (70%). The efficacy of therapy was influenced by the severity of liver failure. In patients with Child–Pugh class A and B disease, the success rate was very high (87% in each group), whereas in Child–Pugh class C, the efficacy of treatment was slightly lower (60% vs. 77%; P Å 0.33) (Table 2). There were no statistically significant differences between the groups regarding transfusion requirements / 5E13$$0011

10-02-96 22:14:19

gasa

The patients were followed up for 6 weeks after randomization. During this period, the hemorrhage recurred in 22 of 64 patients (34%) successfully treated with terlipressin and in 21 of the 68 patients (31%) successfully treated with somatostatin (P Å 0.71). Most of these episodes occurred during the first week after control of bleeding. At this time, the cumulative probability of rebleeding was 23% in the terlipressin group

Table 2. Results of Treatment Terlipressin (n Å 80) Success of treatment Overall Child–Pugh A / B Child–Pugh C Mean time to cessation of bleeding (h)a Transfusion during treatment (U)a Alternative treatment Balloon tamponade Sclerotherapy Surgery Rebleeding at 6 wk Mortality at 6 wk a

64/80 (80%) 52/60 (87%) 12/20 (60%)

P value

68/81 (84%) 48/55 (87%) 20/26 (77%)

0.54 1.0 0.33

5.7 { 4.9

4.7 { 5.9

0.24

1.8 { 1.5

1.9 { 1.7

0.69

6 7 0 24/80 (30%) 13/80 (16%)

4 4 1 23/81 (28.4%) 13/81 (16%)

0.53 0.37 1.0 0.86 1.0

Results are expressed as the mean { SD.

WBS-Gastro

Somatostatin (n Å 81)

November 1996

TERLIPRESSIN AND SOMATOSTATIN IN BLEEDING VARICES 1295

Table 3. Adverse Events Probably Related to Therapy

Figure 1. Actuarial probability of remaining free of rebleeding after successful control of hemorrhage. Most episodes of recurrent variceal hemorrhage occurred in the first week. At this time, the cumulative probability of remaining free of rebleeding was 74% in the somatostatin group and 73% in the terlipressin group.

and 24% in the somatostatin group (Figure 1). Among patients in whom the initial treatment failed and who required alternative therapy, 2 in each group rebled. Thus, the incidence of rebleeding was 30% in the terlipressin group and 28.4% in the somatostatin group (P Å 0.86). Survival No significant differences in survival at 6 weeks were found between the treatment groups: 13 patients in each group died during this period. The cumulative probability of mortality at 6 weeks was 16% in each group (Figure 2). Five patients in the terlipressin group and 2 in the somatostatin group died during the treatment period (P Å 0.28). Most deaths occurred in patients with Child–Pugh class C disease (terlipressin, 6 of 13 deaths; somatostatin, 10 of 13 deaths). The actuarial

Cardiovascular side effects Ventricular fibrillation PSVT Bradycardia (õ55 bpm) Hypertension (SBP ú 160 mm Hg) Other side effects Hyponatremia (õ130 mEq/L) Renal failure (creatinine, ú1.5 mg/dL) Hyperglycemia (ú300 mg/dL) Hypoglycemia (symptomatic) Skin rash Abdominal cramps and diarrhea

Terlipressin (n Å 80)

Somatostatin (n Å 81)

1 1 10 11

0 0 7 5

5

3

0 1 0 0 2

1 1 1 1 0

PSVT, transient paroxysmal supraventricular tachycardia; SBP, systolic blood pressure.

probability of survival at 6 weeks in patients with Child– Pugh’s A and B class was 88% in the terlipressin group and 94% in the somatostatin group. Corresponding figures in Child C class were 70% and 61.5%, respectively. One patient in the terlipressin group was lost to followup after discharge from hospital at day 13 after randomization, and 3 patients in the somatostatin group were lost to follow-up at days 13, 16, and 21 after randomization. All of them belonged to Child–Pugh class A and B. The causes of death were similar in both groups: variceal bleeding (7 patients in the terlipressin group vs. 5 patients in the somatostatin group), liver failure (4 vs. 5 patients), and infections (2 vs. 3 patients). On multivariate analysis (including the same variables used for failure of treatment), only the Child–Pugh score was found to be an independent predictive factor of mortality at 6 weeks (odds ratio, 2.30; 95% CI, 1.45–3.67; P õ 0.001). Adverse Events

Figure 2. Actuarial probability of survival. No differences were found between the two treatment groups in 6-week mortality (16% in each group).

/ 5E13$$0011

10-02-96 22:14:19

gasa

The incidence of side effects was significantly higher in the terlipressin group (31 patients, 38.8%) than in the somatostatin group (19 patients, 23.5%; P Å 0.042). Cardiovascular events were also significantly more frequent in the terlipressin group than in the somatostatin group (28.8% vs. 14.8%; P Å 0.037). This higher number of cardiovascular events in patients receiving terlipressin was mainly caused by a higher, but not significant, incidence of bradycardia and hypertension (Table 3). Bradycardia (heart rate, õ55 bpm) was observed in 10 patients in the terlipressin group and in 7 in the somatostatin group (P Å 0.45) and was asymptomatic in all patients. No patient required specific treatment or stopping trial medication because of bradycardia. Arterial hypertension, defined as a systolic blood pressure WBS-Gastro

1296 FEU ET AL.

GASTROENTEROLOGY Vol. 111, No. 5

of ú160 mm Hg, was observed in 11 patients in the terlipressin group and in 5 patients in the somatostatin group (P Å 0.12). Specific treatment for hypertension was given to 4 of the 11 patients receiving terlipressin and to 4 of the 5 patients receiving somatostatin. In addition, there was one case of ventricular fibrillation in a young alcoholic patient receiving terlipressin who recovered with medical management and withdrawal of terlipressin. Hyponatremia (serum sodium concentration of õ130 mEq/L) was observed in 5 patients receiving terlipressin and in 3 patients treated with somatostatin (P Å 0.49). These complications caused no deaths.

Discussion An increasing number of publications suggest that different pharmacological agents may be useful in the treatment of acute bleeding episodes caused by esophageal varices.3 However, the role of drugs in the treatment of variceal bleeding remains controversial, mainly because of the small number of patients included in many of these studies and the use of different definitions for success and failure of therapy. The interpretation of the results of different trials and their comparison with previous studies is therefore difficult. Consequently, a consensus meeting, held in Baveno, Italy, in 1990,38 set up a series of guidelines to design clinical studies for the treatment of bleeding esophageal varices. The present study was designed following these guidelines, such as the definitions of bleeding, success of therapy, time zero, rebleeding, and death and the time frames for rebleeding and death. Other salient features of the current study are that it was performed and analyzed under strict doubleblind conditions (thus avoiding investigator bias) and that it included a large number of patients, having enough power to detect the expected differences between treatments. In fact, our study is the largest randomized trial of any pharmacological therapy for bleeding varices published so far. The results of the present study confirm that terlipressin is indeed highly effective in the treatment of acute variceal bleeding (80%), a finding that is in agreement with previous studies using this drug.6,12 – 19,36 Interestingly, the success rate for terlipressin in these studies is quite homogeneous despite substantial differences in design and assessment of treatment effect. Thus, terlipressin is unique in uniformly showing a high efficacy in the treatment of bleeding esophageal varices as well as improving survival.20,39 The success rate obtained with somatostatin was similar to that achieved with terlipressin. However, the success rate was higher than expected from the results of previous studies using this drug.22 – 29,32 – 34 It is possible / 5E13$$0011

10-02-96 22:14:19

gasa

that the good results are partly a result of the administration of additional boluses of somatostatin during the first hours of treatment. These were scheduled because previous hemodynamic studies35 showed that bolus injections of somatostatin cause rapid and marked reductions in portal pressure and azygos blood flow (averaging 52% and 45%, respectively), which were more pronounced than the effects observed during continuous infusions (reductions of 13% and 23%). Forty of 81 patients (50%) in the somatostatin group received additional boluses during the first 6 hours of therapy, and control of bleeding was achieved in 70% of these patients. However, placebo boluses were used with a similar frequency in the terlipressin group, which makes it difficult to assess the real contribution of repeated somatostatin boluses in its efficacy. Whichever the mechanism, such a high efficacy of somatostatin was also observed in a recently published randomized controlled trial comparing somatostatin and sclerotherapy in the treatment of acute variceal bleeding that used additional boluses of somatostatin.31 The beneficial effect of both drugs was observed not only in the overall series of patients but also in patients with poor prognostic indicators, such as active bleeding at endoscopy or Child–Pugh class C.40,41 In fact, 35 patients in each therapeutic group had active bleeding at emergency endoscopy, which was completely arrested in 23 patients treated with terlipressin and in 29 patients receiving somatostatin. Burroughs et al.32 also found that among patients with active bleeding at endoscopy, somatostatin was effective in 22 of 32 patients (70%) compared with only 11 of the 30 patients (37%) receiving placebo. Similarly, in the study by Planas et al.,31 therapy was successful in the subset of patients with active bleeding at endoscopy in 13 of 18 patients (72%) receiving somatostatin and in 13 of 17 patients (76%) treated with sclerotherapy. Therefore, the present findings suggest that both terlipressin and somatostatin may be used as first-line therapy in the treatment of variceal bleeding, even when active bleeding is observed at endoscopy. The efficacy of therapy was influenced, as expected, by the severity of liver failure. Both terlipressin and somatostatin achieved a very high success rate in patients with Child–Pugh class A and B disease (87%). In Child– Pugh class C, the efficacy was slightly lower than in patients with good liver function. Nevertheless, even in these high-risk patients, success of therapy was quite high: 60% in the terlipressin group and 77% in the somatostatin group (P Å 0.33). This is in accordance with a study of So¨derlund et al.14 in which the 3 patients in whom terlipressin did not arrest variceal bleeding belonged to the Child C class. Experimental studies have shown that portal hypertension is associated with a deWBS-Gastro

November 1996

TERLIPRESSIN AND SOMATOSTATIN IN BLEEDING VARICES 1297

creased vascular responsiveness to vasoconstrictor agents, including vasopressin.42,43 It is therefore possible that patients with very poor liver function, who have a more pronounced alteration of the hemodynamic state, may be less responsive to terlipressin therapy. The high efficacy obtained in this study in controlling the hemorrhage was associated with a remarkably low mortality rate (cumulative probability of mortality at 6 weeks was 16% in each group). It is unlikely that this low mortality rate was caused by selection of ‘‘good’’ patients because the percentage of patients with Child– Pugh class C in our study was similar to that in previous trials with higher mortality rates.6,25,29,33 Actually, even in patients with class C, the 6-week cumulative probability of mortality was low (30% in the terlipressin group and 38.5% in the somatostatin group). This is in accordance with the findings of So¨derlund et al.14 and of a recent meta-analysis20 showing a lower hospital mortality for patients treated with terlipressin than in those receiving placebo. In addition, a recently reported ambulance trial also showed that survival rates were higher in patients treated with terlipressin at home and during transfer to hospital than in those receiving placebo.44 In contrast with more invasive therapies, such as balloon tamponade and emergency sclerotherapy, the incidence of severe side effects associated with drug therapy was low. The trial medication had to be withdrawn only in 1 patient receiving terlipressin, a figure that clearly differs from the high incidence of severe side effects observed with vasopressin infusion requiring end of therapy in about 25% of patients.3 Although cardiovascular events such as bradycardia or hypertension were frequent, especially in the terlipressin group, they were mild in most cases, and only 4 patients in each group required specific treatment because of them (high systolic blood pressure in all instances). Therefore, there were no significant differences in severe side effects between the two groups, in agreement with previous studies using terlipressin6,12 – 15,45 and somatostatin.22 – 33 The low incidence of severe side effects suggests that both drugs may be administered for longer periods of time after the initial control of bleeding to prevent early variceal rebleeding, which occurred in nearly 30% of patients during the first week of admission. Although it could be argued that side effects are more frequent and severe during prolonged administration of these drugs, Burroughs et al.32 did not observe side effects during a 5-day somatostatin infusion. Similarly, Fiaccadori et al.45 treated a large series of patients with terlipressin for 7 days without severe complications. Therefore, the results of the present study show that both terlipressin and somatostatin have a similar high / 5E13$$0011

10-02-96 22:14:19

gasa

efficacy in controlling acute variceal hemorrhage and maintaining a low mortality rate. These good results compare very well with those obtained with emergency sclerotherapy. In fact, in a recent meta-analysis of 11 studies published in full, the efficacy of emergency sclerotherapy was 85%,20 which is similar to that observed in the current study using either terlipressin or somatostatin. However, it should be remembered that the frequency and severity of the complications caused by emergency sclerotherapy is much greater than that observed in this study, being 18% for severe side effects, 15% of these resulting in death.20 The results of this study therefore suggest that both drugs may be used as first-line treatment for acute variceal bleeding. Sclerotherapy could be delayed in many patients or used only in patients in whom pharmacological therapy fails to control variceal bleeding. Moreover, the low incidence of severe side effects suggests that both drugs may be maintained for longer periods to prevent early variceal rebleeding. Terlipressin and somatostatin are not available in some countries. Our findings with terlipressin cannot be extrapolated to vasopressin, which, as mentioned, has been proven to be less effective and more toxic. Similarly, our findings using somatostatin should not be extrapolated to octreotide, a drug that has also been proposed for the treatment of variceal bleeding but for which there is limited evidence.39

References 1. Burroughs AK, Bosch J. Clinical manifestations and management of bleeding episodes in cirrhotics. In: McIntyre N, Benhamou JP, Bircher J, Rizzeto M, Rode´s J, eds. Oxford textbook of clinical hepatology. Oxford, England: Oxford University, 1991:408–425. 2. Burroughs AK, McCormick PA. Natural history and prognosis of variceal bleeding. Baillie`re´s Clin Gastroenterol 1992;6:437– 450. 3. Feu F, D’Amico G, Bosch J. The acute bleeding episode: advances in drug therapy. In: Arroyo V, Bosch J, Rode´s J, eds. Treatments in hepatology. Barcelona, Spain: Masson S.A., 1995:9–22. 4. Pliska V, Chard T, Rudinger J, Forsling ML. In vivo inactivation of synthetic hormonogens of lysine vasopressin: N-glycyl-glycylglycyl-[8-lysine] vasopressin in the cat. Acta Endocrinol 1976;81: 474–481. 5. Forsling ML, Aziz LA, Miller M, Davis R. Conversion of triglycylvasopressin to lysine-vasopressin in man. J Endocrinol 1980; 85:237–244. 6. D’Amico G, Traina M, Vizzini G, Tine` F, Politi F, Montalbano L, Luca A, Pasta L, Pagliaro L, Morabito A, and The Liver Study Group of V. Cervello Hospital. Terlipressin or vasopressin plus transdermal nitroglycerin in a treatment strategy for digestive bleeding in cirrhosis. A randomized clinical trial. J Hepatol 1994; 20:206–212. 7. Douglas JG, Forrest JAH, Prowse CV, Cash JD, Finlayson NDC. Effects of lysine vasopressin and glypressin on the fibrinolytic system in cirrhosis. Gut 1979;20:565–567. 8. Merkel C, Bianco S, Bolognesi M, Caregaro L, Amodio P, Sacerdoti D, Gatta A. Effect of triglycyl-lysin-vasopressin on quantitative liver function tests in patients with cirrhosis. Am J Gastroenterol 1992;87:1580–1586.

WBS-Gastro

1298 FEU ET AL.

GASTROENTEROLOGY Vol. 111, No. 5

9. Moreau R, Soubrane O, Hadengue A, Sogni P, Gaudin P, Kleber G, Lebrec D. Effets hemodynamiques de l’administration de terlipressine seule ou associee a la nitroglycerine chez les malades atteints de cirrhose. Gastroenterol Clin Biol 1992;16:680–686. 10. Escorsell A, Feu F, GarcıB a-Paga´n JC, Luca A, Trejo R, Bosch J, Rode´s J. Long-term reduction of portal pressure and collateral blood flow after single IV injection of 1 and 2 mg of glypressin in patients with cirrhosis. A double blind controlled study (abstr). J Hepatol 1994;21(Suppl):S134. 11. Nilsson G, Lindblom P, Ohlin M, Berling R, Vernersson E. Pharmacokinetics of terlipressin after single I.V. doses to healthy volunteers. Drugs Exp Clin Res 1990;6:307–314. 12. Walker S, Stiel A, Raedsch R, Kommerell B. Terlipressin in bleeding esophageal varices: a placebo-controlled, double-blind study. Hepatology 1986;6:112–115. 13. Freeman JG, Cobden I, Record CO. Placebo-controlled trial of terlipressin (Glypressin) in the management of acute variceal bleeding. J Clin Gastroenterol 1989;11:58–60. 14. So¨derlund C, Magnusson I, To¨rngren S, Lundell L. Terlipressin (triglycyl-lysine vasopressin) controls acute bleeding esophageal varices. A double-blind, randomized, placebo-controlled trial. Scand J Gastroenterol 1990;25:622–630. 15. Freeman JG, Cobden I, Lishman AH, Record CO. Controlled trial of terlipressin (‘‘glypressin’’) versus vasopressin in the early treatment of esophageal varices. Lancet 1982;2:66–68. 16. Desaint B, Florent C, Levy VG. A randomized trial of triglycyl-lysine vasopressin versus lysine vasopressin in active cirrhotic variceal hemorrhage. In: Lebrec D, Blei AT, eds. Vasopressin analogs and portal hypertension. Paris: John Libbey Eurotext, 1987:155– 157. 17. Chiu KW, Sheen IS, Liaw YF. A controlled study of glypressin versus vasopressin in the control of bleeding from oesophageal varices. J Gastroenterol Hepatol 1990;5:549–553. 18. Colin R, Giuli N, Caernicow P, Ducrotte P, Lerebours E. Prospective comparison of glypressin, tamponade and their association in the treatment of bleeding esophageal varices. In: Lebrec D, Blei AT, eds. Vasoanalogs and portal hypertension. Paris: John Libbey Eurotext, 1987:149–153. 19. Fort E, Sautereau D, Silvain C, Ingrand P, Pillegand B, Beauchant M. A randomized trial of terlipressin plus nitroglycerin vs. balloon tamponade in the control of acute variceal hemorrhage. Hepatology 1990;11:678–681. 20. D’Amico G, Pagliaro L, Bosch J. The treatment of portal hypertension. A meta-analytic review. Hepatology 1995;22:332–354. 21. Bosch J, Kravetz D, Rode´s J. Effects of somatostatin on hepatic and systemic hemodynamics in patients with cirrhosis of the liver: comparison with vasopressin. Gastroenterology 1981;80: 518–525. 22. Kravetz D, Bosch J, Tere´s J, Bruix J, Rimola A, Rode´s J. Comparison of intravenous somatostatin and vasopressin infusions in treatment of acute variceal hemorrhage. Hepatology 1984;4: 442–446. 23. Jenkins SA, Baxter JN, Corbett W, Devitt P, Ware J, Shields R. A prospective randomised controlled clinical trial comparing somatostatin and vasopressin in controlling acute variceal haemorrhage. BMJ 1985;290:275–278. 24. Bagarani M, Albertini V, Anza M, Barlattani M, Bracci F, Cucchiara G, Gizzonio D, Grassini G, Mari T, Procacciante F. Effect of somatostatin in controlling bleeding from esophageal varices. Ital J Surg Sci 1987;17:21–26. 25. Cardona C, Vida F, Balanzo´ J, Cusso X, Farre A, Guarner C. Eficacia terape´utica de la somatostatina versus vasopresina ma´s nitroglicerina en la hemorragia activa por varices esofagoga´stricas. Gastroenterol Hepatol 1989;12:30–34. 26. Hsia HC, Lee FY, Tsai YT, Lee SD, Lai HK, Lin WJ, Lay CS, Wang SS, Lo KJ. Comparison of somatostatin and vsopressin in the

/ 5E13$$0011

10-02-96 22:14:19

gasa

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

control of acute esophageal variceal hemorrhage. A randomized, controlled study. Chin J Gastroenterol 1990;7:71–78. Saari A, Kivilaakso E, Inberg M, Paakkonen M, Lahtineh J, Hockerstedt K, Schroder T. Comparison of somatostatin and vasopressin in bleeding esophageal varices. Am J Gastroenterol 1990; 85:804–807. Jaramillo JL, de la Mata M, Min ˜o G, Costa´n G, Go´mez-Camacho F. Somatostatin versus Sengstaken balloon tamponade for primary haemostasia of bleeding esophageal varices. J Hepatol 1991; 12:100–105. Avgerinos A, Klonis C, Rekoumis G, Gouma P, Papadimitriou N, Raptis S. A prospective randomized trial comparing somatostatin, balloon tamponade and the combination of both methods in the management of acute variceal haemorrhage. J Hepatol 1991;13:78–83. Shields R, Jenkins SA, Baxter AN, Kingsnorth S, Ellenbogen CA, Makin I, Gilmore AI, Morris D, West CR. A prospective randomised controlled trial comparing the efficacy of somatostatin with injection sclerotherapy in the control of bleeding oesophageal varices. J Hepatol 1992;16:128–137. Planas R, Quer JC, Boix J, Canet J, Armengol M, Cabre´ E, Pintanel T, Humbert P, Oller B, Broggi MA, Gassull MA. A prospective randomized trial comparing somatostatin and sclerotherapy in the treatment of acute variceal bleeding. Hepatology 1994;20: 370–375. Burroughs AK, McCormick PA, Hughes MD, Sprengers D, D’Heygere F, McIntyre N. Randomized, double-blind, placebo-controlled trial of somatostatin for variceal bleeding. Emergency control and prevention of early variceal rebleeding. Gastroenterology 1990; 99:1388–1395. Valenzuela JE, Schubert T, Fogel MR, Strong RM, Levine J, Mills PR, Fabry TL, Taylor LW, Conn HO, Posillico JT, a Multicenter Study Group. A multicenter, randomized, double-blind trial of somatostatin in the management of acute hemorrhage from esophageal varices. Hepatology 1989;10:958–961. Gotzsche PC, Gjorup I, Bonnen H, Brahe NEB, Becker U, Burcharth F. Somatostatin v placebo in bleeding oesophageal varices: randomised trial and meta-analysis. BMJ 1995;310:1495– 1498. Cirera I, Feu F, Luca A, GarcıB a-Paga´n JC, Ferna´ndez M, Escorsell A, Bosch J, Rode´s J. Effects of bolus injections and continuous infusions of somatostatin and placebo in patients with cirrhosis. A double-blind hemodynamic investigation. Hepatology 1995;22: 106–111. Walker S, Kreichgauer HP, Bode JC. Terlipressin vs. somatostatin in bleeding esophageal varices: a controlled, double-blind study. Hepatology 1992;15:1023–1030. Rimola A, Bory F, Tere´s J, Pe´rez-Ayuso RM, Arroyo V, Rode´s J. Oral, nonabsorbable antibiotics prevent infection in cirrhotics with gastrointestinal hemorrhage. Hepatology 1985;3:463– 467. De Franchis R, Pascal JP, Ancona E, Burroughs AK, Henderson M, Fleig W, Groszmann R, Bosch J, Sauerbruch T, Soederlund C, Lebrec D, Soerensen TIA, Pagliaro L. Definitions, methodology and therapeutic strategies in portal hypertension. A Consensus Development Workshop, Baveno, Lake Maggiore, Italy, April 5 and 6, 1990. J Hepatol 1992;15:256–261. Groszmann RJ, Bendtsen F, Bosch J, Fleig W, Grace N, Lebrec D, Merkel C, Vinel JP. Baveno II consensus statements: drug therapy for portal hypertension. In: De Franchis R, ed. Portal hypertension. II. Proceedings of the Second Baveno International Consensus Workshop on Definitions, Methodology and Therapeutic Strategies. Oxford, England: Blackwell Scientific, 1995:98– 99. Burroughs AK, McCormick PA. Natural history and prognosis of variceal bleeding. Baillie`re´s Clin Gastroenterol 1992;6:437– 450.

WBS-Gastro

November 1996

TERLIPRESSIN AND SOMATOSTATIN IN BLEEDING VARICES 1299

41. Pagliaro L, D’Amico G, Pasta L, Politi F, Vizzini G, Traina M, Madonia S, Luca A, Guerrera D, Puleo A, D’Antoni A. Portal hypertension in cirrhosis: natural history. In: Bosch J, Groszmann RJ, eds. Portal hypertension. Pathophysiology and treatment. Oxford, England: Blackwell Scientific, 1994:72–92. 42. Kravetz D, Cummings SA, Groszmann RJ. Hyposensitivity to vasopressin in a hemorrhaged-transfused rat model of portal hypertension. Gastroenterology 1987;93:170–175. 43. Mesh CL, Joh T, Korthuis RJ, Granger DN, Benoit JN. Intestinal vascular sensitivity to vasopressin in portal hypertensive rats. Gastroenterology 1991;100:916–921. 44. Levacher S, Letoumelin P, Pateron D, Blaise M, Lapandry C, Pourriat JL. Early administration of terlipressin plus glyceryl trinitrate to control active upper gastrointestinal bleeding in cirrhotic patients. Lancet 1995;346:865–868. 45. Fiaccadori F, Pedreti G, Biraghi M, Arcidiacono R. Terlipressin and endoscopic sclerotherapy control variceal bleeding and pre-

/ 5E13$$0011

10-02-96 22:14:19

gasa

vent early rebleeding in cirrhotic patients. Current Ther Res 1993; 54:1–10. Received July 26, 1995. Accepted July 8, 1996. Address requests for reprints to: Jaume Bosch, M.D., Liver Unit, Hospital Clinic I Provincial, Villarroel 170, 08036 Barcelona, Spain. Fax (34) 3-451-5522. Supported in part by grants 94/0757, 94/0734, and 94/0068 from the Fondo de Investigaciones Sanitarias and by Ferring AB (Malmo¨, Sweden). The Variceal Bleeding Study group included A. Mas, A. Escorsell, J. C. CarcıB a-Paga´n, J. M. Bordas, J. M. Salmero´n, and J. Rode´s, Hospital Clinic i Provincial, Barcelona; A. Monescillo, M. Vicente, and A. GarcıB a-Plaza, Hospital Ramo´n y Cajal, Madrid; M. Casado, G. Clemente, P. Menchen, and E. Cos, Hospital General Gregorio Maran ˜o´n, Madrid; and J. C. Quer, J. Boix, P. Humbert, and M.A. Gassull, Hospital Universitari Germans Trias i Pujol, Badalona, Spain.

WBS-Gastro