|
|
| |||||||||||||||||||||||||||||||||
Background Patients who have bleeding from esophageal varices are at high risk for rebleeding and death. We compared the efficacy and safety of endoscopic sclerotherapy with the efficacy and safety of nadolol plus isosorbide mononitrate for the prevention of variceal rebleeding.
Methods Eighty-six hospitalized patients with cirrhosis and bleeding from esophageal varices diagnosed by endoscopy were randomly assigned to treatment with repeated sclerotherapy (43 patients) or nadolol plus isosorbide-5-mononitrate (43 patients). The primary outcomes were rebleeding, death, and complications. The hepatic venous pressure gradient was measured at base line and after three months.
Results Base-line data were similar in the two groups, and the median follow-up was 18 months in both. Eleven patients in the medication group and 23 in the sclerotherapy group had rebleeding. The actuarial probability of remaining free of rebleeding was higher in the medication group for all episodes related to portal hypertension (P = 0.001) and variceal rebleeding (P = 0.002). Four patients in the medication group and nine in the sclerotherapy group died (P = 0.07 for the difference in the actuarial probability of survival). Seven patients in the medication group and 16 in the sclerotherapy group had treatment-related complications (P = 0.03). Thirty-one patients in the medication group underwent two hemodynamic studies; 1 of the 13 patients with more than a 20 percent decrease in the hepatic venous pressure gradient had rebleeding, as compared with 8 of the 18 with smaller decreases in the pressure gradient (P = 0.04 for the actuarial probability of rebleeding at two years).
Conclusions As compared with sclerotherapy, nadolol plus isosorbide mononitrate significantly decreased the risk of rebleeding from esophageal varices.
Recent hemodynamic studies have shown that isosorbide mononitrate combined with propranolol results in a greater reduction in portal pressure than that achieved with propranolol alone13 and is also effective in patients who do not have a response to therapy with propranolol alone.14 We compared the efficacy and safety of endoscopic sclerotherapy with the efficacy and safety of nadolol plus isosorbide mononitrate for the prevention of variceal rebleeding. Nadolol was used instead of propranolol because it is not metabolized by the liver and needs to be administered only once a day.15,16,17
Methods
Selection of Patients
From August 1991 to January 1994, 765 patients were admitted to our hospital because of upper gastrointestinal bleeding. Of these patients, 121 met the following eligibility criteria: cirrhosis, emergency endoscopy performed within the first four hours after admission, and endoscopic evidence of a hemorrhage from esophageal varices. The diagnosis of cirrhosis was verified by clinical, biochemical, and echographic findings.
Thirty-five of the 121 patients were excluded for the following reasons: age under 18 years (2 patients), a Child–Pugh score higher than 12 points (11), advanced hepatocellular carcinoma (3), lung cancer (1), previous sclerotherapy (8), the failure of medical therapy to control the index bleeding (7), and refusal to participate in the study (3).
Randomization and Treatment
On the fifth hospital day, 86 patients were randomly assigned to one of two treatment groups with the use of opaque sealed envelopes that contained the treatment assignments, which were derived from computer-generated random numbers. Randomization was stratified according to the severity of liver failure, as assessed by the Child–Pugh classification (class A and B or class C), and the presence or absence of a history of variceal bleeding. Written informed consent was obtained from all the patients or their next of kin, and the trial was approved by the ethics committee of our hospital.
In the medication group, continuous pharmacologic therapy was started immediately after randomization. Nadolol was given orally at an initial dose of 80 mg once daily. The dose was subsequently adjusted over a period of five days until the resting heart rate had been reduced by 25 percent or was 55 beats per minute. Oral isosorbide mononitrate was started immediately after the dose of nadolol had been adjusted. The dose was increased progressively up to a dose of 40 mg twice a day over a period of one week, unless side effects appeared. Compliance was assessed at each follow-up visit by careful questioning of the patient and his or her relatives.
In the sclerotherapy group, sclerotherapy was performed by injecting 5 percent ethanolamine into all esophageal varices at different levels within the lower 4 cm of the esophagus, up to a total dose of 10 to 20 ml. Sclerotherapy sessions were carried out on days 0 (the day of randomization), 4, 10, and 30, and then monthly until the varices were eradicated. After eradication, follow-up endoscopic studies (with further sclerotherapy sessions if the varices reappeared) were performed at three months and subsequently every six months.
Follow-Up and End Points
The study continued until seven months after the enrollment of the last patient. The primary end points were rebleeding, complications, and death.
Rebleeding was defined as any episode of hematemesis or melena (or both) occurring during follow-up and was evaluated by emergency endoscopy. In both groups, the index hemorrhage, as well as rebleeding, was treated with octreotide or emergency sclerotherapy, or both. We calculated the rebleeding index for each patient by dividing the months of follow-up by the number of rebleeding episodes plus 1, as suggested in a consensus workshop.18 This index reflects the time free of rebleeding during follow-up. Treatment failure was defined as the occurrence of two or more rebleeding episodes that required the transfusion of at least 2 units of red cells or continued hemorrhage despite medical treatment, requiring the transfusion of 4 or more units of red cells. For patients in whom the study treatment failed, decisions about alternative treatment were made individually.
Hemodynamic Studies
Hemodynamic studies were performed before randomization and again three to four months after the start of the drug treatment or after the completion of the sclerotherapy sessions. After an overnight fast, a venous catheter with an introducer sheath was placed in the right femoral vein by the Seldinger technique and used to advance, under fluoroscopic guidance, a 7-French balloon-tipped catheter (Meditech) into the right main hepatic vein and a Swan–Ganz catheter (Abbott Laboratories) into the pulmonary artery. Portal pressure was measured as the hepatic venous pressure gradient. Cardiopulmonary pressures and cardiac output were also measured. All measurements were performed in triplicate with the use of a previously calibrated strain-gauge transducer, with the midaxillary line as the zero reference point.
Statistical Analysis
In calculating the sample size, we assumed a 50 percent rebleeding rate in the sclerotherapy group.19 Forty-two patients were required in each treatment group to detect a difference of at least 30 percent, with alpha and beta values of 0.05 and 0.2, respectively.
The statistical analysis was performed according to an intention-to-treat strategy. Qualitative variables were compared with Fisher's exact test. Student's t-test was used to compare the mean values for continuous variables, and the Wilcoxon rank-sum test was used for skewed or ordinal data.20 Actuarial probabilities were calculated by the Kaplan–Meier method and compared with the log-rank test. Data were censored at the time of death or at the last visit. The Cox proportional-hazards model was used to identify the subgroup of variables that best explained the differences in survival and rebleeding rates.21 All P values were two-tailed, and values of less than 0.05 were considered to indicate statistical significance.22 Calculations were performed with the BMDP statistical package.
Results
Forty-three patients were randomly assigned to each treatment group. Base-line data were similar in the two groups, and the median follow-up in both groups was 18 months (Table 1 and Table 2). Three patients in the medication group had contraindications to the use of nadolol and received only isosorbide mononitrate. There were no patients with contraindications to isosorbide mononitrate or sclerotherapy. Only one patient (in the medication group) did not comply with the study protocol.
|
|
Data on rebleeding are presented in Table 3 and Figure 1. The likelihood of rebleeding was significantly lower in the medication group than in the sclerotherapy group (P = 0.001) (Figure 1). Similarly, the actuarial probability of remaining free of rebleeding at two years was higher in the medication group than in the sclerotherapy group when the data were analyzed according to the Child–Pugh class: class A, 67 percent versus 48 percent; class B, 63 percent versus 49 percent; and class C, 60 percent versus 43 percent (P = 0.002).
|
|
The likelihood of treatment failure at two years was significantly lower in the medication group (10 percent) than in the sclerotherapy group (54 percent, P<0.001). Treatment failed in three patients in the medication group: two patients subsequently received portacaval shunts, and one, who had end-stage liver disease, received no further treatment. In the sclerotherapy group, treatment failed in 16 patients: 7 were subsequently treated with nadolol and isosorbide mononitrate, 3 received portacaval shunts, 2 received transjugular intrahepatic portosystemic shunts, and the remaining 4 received no other treatment.
Survival
The actuarial probability of survival was higher in the medication group than in the sclerotherapy group, although the difference was not significant (P = 0.07) (Figure 2). Nine patients in the sclerotherapy group and four in the medication group died. Seven patients (five in the sclerotherapy group and two in the medication group) died of liver failure. Death was related to rebleeding in two patients (both in the sclerotherapy group) and to hepatocellular carcinoma in three (two in the sclerotherapy group and one in the medication group). One patient in the medication group died of a stroke.
|
Complications and Subsidiary Outcomes
Seven patients in the medication group had treatment-related complications, as compared with 16 in the sclerotherapy group (P = 0.03). In the sclerotherapy group, seven patients had bleeding esophageal ulcers, two had aspiration pneumonia, two had esophageal strictures, two had pleural effusions, and three had septic complications. In the medication group, three patients had resting heart rates of less than 50 beats per minute, two had weakness and dyspnea, one had bronchospasm, and one had impotence. Nadolol had to be discontinued in four of the patients (those with bradycardia or bronchospasm). None of the complications were fatal.
At the index endoscopy, four patients in each group had moderate portal hypertensive gastropathy. At the last endoscopy performed during follow-up, 6 patients in the medication group and 31 in the sclerotherapy group had moderate or severe portal hypertensive gastropathy (P<0.001).
Kidney function varied in similar fashion in the two groups, and laboratory values remained within the normal range (Table 2). During follow-up, ascites developed in 15 patients in the medication group (6 of whom had no history of ascites) and in 18 patients in the sclerotherapy group (9 of whom had no history of ascites).
Hemodynamic Measurements
Two hemodynamic studies were performed in 31 patients in each group; the second study could not be carried out in the remaining patients because of previous treatment failure or refusal of consent.
Continued drug therapy, but not sclerotherapy, significantly reduced the hepatic venous pressure gradient (Table 2). The gradient decreased to less than 12 mm Hg in nine patients in the medication group and in two patients in the sclerotherapy group (P = 0.04). The hepatic venous pressure gradient decreased by more than 20 percent from the base-line value in 13 patients in the medication group and in 4 in the sclerotherapy group (P = 0.02). In two patients (one in each group), the gradient fell below 12 mm Hg, but the value was not more than 20 percent below the base-line value.
Bleeding did not recur in any of the patients who had a hepatic venous pressure gradient below 12 mm Hg. In the medication group, bleeding recurred in 1 of the 13 patients who had a decrease of more than 20 percent from the base-line value and in 8 of 18 with a 20 percent or smaller decrease; the actuarial probability of rebleeding at two years in these two groups was 33 and 47 percent, respectively (P = 0.04). In the sclerotherapy group, bleeding recurred in none of the 4 patients with a decrease of more than 20 percent in the hepatic venous pressure gradient but in 15 of the 27 with a 20 percent or smaller decrease; the actuarial probability of rebleeding at two years was 0 and 68 percent, respectively (P = 0.07).
Discussion
Our results suggest that after an acute episode of variceal bleeding has been controlled, therapy with nadolol plus isosorbide mononitrate has substantial advantages over injection sclerotherapy. The rebleeding rates in our sclerotherapy group were similar to those previously reported at our center19,23 and elsewhere.3,4 However, rebleeding was significantly less common with combined-medication therapy, whether all the episodes related to portal hypertension are considered or only those due to esophageal varices. Many factors related to rebleeding were also significantly improved (Table 3). The value of combined therapy with beta-blockers and isosorbide mononitrate was also demonstrated in a trial comparing this treatment with shunt placement for patients in Child–Pugh class A or B and sclerotherapy for those in class C.24 Similar results were obtained with the medical and invasive treatments.24
The benefits of combined-drug therapy in our study were not due to differences between the groups in factors that are important in predicting recurrent variceal hemorrhage, such as the severity of liver disease, the interval from the index bleeding episode to the start of the analysis, the treatment used to stop acute episodes of bleeding, or whether or not the patient has abstained from the use of alcohol.1,25,26 Both groups were well matched with respect to these prognostic variables.
The pharmacologic treatment of portal hypertension is based on the assumption that a sustained reduction in portal pressure reduces the incidence of variceal hemorrhage.27 A hepatic venous pressure gradient of 12 mm Hg has been identified as the threshold for the development of variceal hemorrhage.28 Once the gradient has been reduced to a level below 12 mm Hg, patients are protected from a first variceal hemorrhage and have a significant increase in survival.29 However, such a reduction in the hepatic venous pressure gradient is achieved in only a minority of patients.29 After an episode of variceal bleeding, a reduction to a level more than 20 percent below the base-line value, even if the final value is above 12 mm Hg, significantly decreases the risk of rebleeding.30 We found that the proportion of patients with a decrease of more than 20 percent in the base-line hepatic venous pressure gradient and the proportion with values that were reduced to a level below 12 mm Hg were significantly higher in the medication group than in the sclerotherapy group. In the third month of follow-up, the hepatic venous pressure gradient was an independent predictor of rebleeding. The greater efficacy of medical therapy may be related to the sustained reduction in portal pressure. Our results also indicate that hemodynamic measurements may be useful in identifying patients with a poor response to combined pharmacologic treatment, who may benefit from alternative therapy.
We also found that the incidence of treatment-related complications was significantly lower in the patients treated with nadolol plus isosorbide mononitrate than in those undergoing sclerotherapy. In patients with cirrhosis, nitrates may have deleterious effects on kidney function and on the control of ascites.31,32,33 As in previous studies,34,35 however, our results suggest that the combination of a beta-blocker and isosorbide mononitrate does not impair renal function or the control of ascites. It has also been suggested that the prevalence of portal hypertensive gastropathy may be related to sclerotherapy.36 In accordance with this proposed association, we found that the patients treated with sclerotherapy had a significantly higher frequency of portal hypertensive gastropathy than those treated with medications.
In conclusion, we found that therapy with nadolol plus isosorbide mononitrate, as compared with sclerotherapy, significantly decreased the incidence of variceal rebleeding and treatment-related complications. There was also a trend toward improved survival, although the difference between the groups was not significant.
Supported in part by a grant from the Fundació Investigació Sant Pau.
Source Information
From the Department of Gastroenterology, Hospital de la Santa Creu i Sant Pau, Barcelona, Spain.
Address reprint requests to Dr. Balanzó at the Servei de Patologia Digestiva, Hospital de la Santa Creu i Sant Pau, Avgda. Sant Antoni M. Claret, 167, 08025 Barcelona, Spain.
References
| |||||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
