Beta-Blockers to Prevent Esophageal Varices — An Unfulfilled Promise
http://www.100md.com
《新英格兰医药杂志》
Variceal bleeding is the most feared complication of portal hypertension in patients with cirrhosis and is a major cause of death. Treatments frequently include endoscopic banding or sclerotherapy of esophageal varices in parallel with splanchnic vasoconstrictors and intensive medical care.1 Alternative interventions, such as decompressive surgical shunts, transjugular intrahepatic systemic shunts, and liver transplantation, are needed in patients with esophageal variceal bleeding refractory to endoscopic intervention or patients with bleeding from gastric or ectopic varices. However, the choice of therapy is based on the presence or absence of decompensated liver disease and is often influenced by local expertise. For example, patients with compensated cirrhosis can frequently be treated with surgical shunts or transjugular intrahepatic systemic shunts, whereas liver transplantation should be offered to those with decompensated cirrhosis, who are at unacceptably high risk for complications and death from such procedures.2 Unfortunately, among patients with decompensated cirrhosis, long-term survival is improved only by liver transplantation.
Although the risk of bleeding is low in patients without varices (1 to 2 percent per year) or with small varices (5 percent per year), small varices have a median rate of progression to larger varices of 12 percent per year, and large varices, in turn, have an annual risk of bleeding of 15 percent.3 The lifetime risk of at least one bleeding episode is 30 percent among patients with cirrhosis and varices, with the greatest risk among those with large varices and decompensated liver disease. Since a single episode of uncontrolled variceal bleeding has an immediate rate of death of 5 to 8 percent and a six-week mortality rate of 50 percent, preventing varices from bleeding is of paramount importance.4 Primary prophylaxis, or the prevention of a first episode of variceal bleeding in a patient with cirrhosis with known varices, is often managed with nonselective beta-blockers such as propanolol, nadolol, or timolol.5 The mechanism of action of these medications is complex and varies according to dose. Low doses cause unopposed splanchnic vasoconstriction, leading to reduced portal blood flow and, therefore, decreased portal hypertension. As the dose is increased, portal flow is further reduced by the negative effect of these medications on cardiac output. Furthermore, there is also evidence that these medications decrease the development of portosystemic collaterals, an important precursor to the formation of varices.
Since the original study by Lebrec et al. in 1981,6 several randomized trials and cost-effectiveness analyses have conclusively demonstrated that nonselective beta-blockers are the medications of first choice for primary prophylaxis against variceal bleeding in patients with varices.7,8 More recently, randomized, controlled trials have reported a lower rate of variceal bleeding among patients treated with endoscopic banding — a procedure usually reserved for acutely bleeding varices or secondary prophylaxis — than among patients treated with beta-blockers, although no effect on mortality was noted.9,10 However, these studies are limited by their short follow-up and small size. Until these issues are addressed in future studies, gastroenterologists will continue to choose nonselective beta-blockers for primary prophylaxis in most patients and reserve endoscopic interventions for primary prophylaxis in patients who cannot tolerate beta-blockers or in those with large varices who live far from medical care.
A potential extension of the finding that beta-blockers are beneficial in patients with varices would be the use of beta-blockers to prevent variceal formation in patients with cirrhosis, an intuitively attractive proposition, given the high lifetime risk of bleeding once large varices develop. In this issue of the Journal, Groszmann et al. report the results of a randomized, double-blind, placebo-controlled study evaluating the effectiveness of timolol for the primary prevention of varices in patients with cirrhosis.11 This study, conducted at four centers renowned for their expertise in portal hypertension, randomly assigned 213 patients with cirrhosis to receive timolol or placebo. The primary objectives of the study were to determine whether timolol prevented gastroesophageal varices and whether baseline and serial measurements of the hepatic venous pressure gradient (HVPG) could be used to predict the development of varices.
The investigators reported that timolol not only did not prevent gastroesophageal varices but also was associated with a higher incidence of adverse events, many of which were serious. They also demonstrated that the risk of variceal formation was decreased among patients with a baseline HVPG of less than 10 mm Hg and among patients whose HVPG decreased by more than 10 percent per year, irrespective of timolol use.
The relatively low baseline portal pressures in the study population may have played a role in the failure to demonstrate a benefit of timolol. The median value for HVPG in both study groups was only 11.7 mm Hg, suggesting that patients were at low risk for varices. Portal hypertension, defined by an HVPG exceeding 5 mm Hg, is usually not clinically significant until the gradient exceeds 12 mm Hg, at which point variceal bleeding and ascites are more likely to develop. Conversely, in patients with bleeding varices or refractory ascites treated by transjugular intrahepatic systemic shunts, a successful outcome is defined by a reduction of the HVPG to less than 12 mm Hg. However, not all patients whose HVPG exceeds 12 mm Hg have varices, and the relationship between a rising gradient and the development of varices remains poorly understood. To complicate matters further, the HVPG is also increased by alcohol use and worsening liver function and can improve with abstinence from alcohol use.12
The inclusion of 14 patients (27 percent) who had been abstinent from alcohol use for less than one month and 5 patients with alcoholic hepatitis may also have contributed to the study's negative findings. Since alcohol consumption was not permitted during the study, the cessation of alcohol use among some patients may have decreased their HVPG and risk of varices. It would have been prescient to use a conventional period of abstinence, such as six months, to avoid the confounding effects of alcohol cessation on the development of varices. Another important point to emphasize is that although the investigators reported that similar percentages of patients in each group reached the primary end point, small varices accounted for 83 percent of these end points in both groups, and small varices are unlikely to bleed and are of less clinical significance than large varices.
The choice of timolol for this study was interesting, since prior trials of portal hypertension have evaluated nadolol or propranolol. Although studies have demonstrated the potent negative inotropic effects of timolol in healthy volunteers and the acute portal-pressure–lowering effects in patients with cirrhosis, there have been no studies of its effectiveness with long-term use in patients with cirrhosis. The high frequency of adverse events (48 percent in the timolol group, as compared with 32 percent in the placebo group) is of great concern, particularly since prior randomized trials comparing propranolol or nadolol with placebo have reported adverse events in fewer than 17 percent of patients receiving beta-blockers.13
What role, if any, do nonselective beta-blockers have in preventing variceal formation in patients with cirrhosis? The negative findings of the study by Groszmann et al. clearly do not support the empirical use of these medications in patients who have cirrhosis without varices, since the risks far outweigh the benefits. This conclusion may surprise clinicians, particularly since cost-effectiveness analyses have historically promoted the use of these drugs, further reinforcing the importance of well-conducted clinical trials.8 However, a very important and novel finding of this study is the pivotal role of HVPG monitoring as a diagnostic and prognostic tool in patients with cirrhosis. Unfortunately, HVPG monitoring is rarely used outside the realms of clinical research. Gastroenterologists often use vital signs to monitor the effectiveness of beta-blockade or use upper endoscopy to screen patients for varices. The findings of this study, together with the report by Targownik et al. demonstrating the cost-effectiveness of HVPG monitoring,14 should encourage physicians to consider this procedure as an integral part of their clinical armamentarium in the treatment of patients with cirrhosis and to view it as complementary to rather than competing with upper endoscopy. Until we can accurately predict when varices will develop, the results of this important trial suggest that nonselective beta-blockers should not be used to prevent varices in patients. However, these findings should not affect current recommendations to use beta-blockers, alone or in conjunction with endoscopic banding, to prevent bleeding in patients with established esophageal varices.15
Source Information
From the University of Nebraska Medical Center, Omaha.
References
Harry R, Wendon J. Management of variceal bleeding. Curr Opin Crit Care 2002;8:164-170.
Wright AS, Rikkers LF. Current management of portal hypertension. J Gastrointest Surg 2005;9:992-1005.
Zoli M, Merkel C, Magalotti D, et al. Natural history of cirrhotic patients with small esophageal varices: a prospective study. Am J Gastroenterol 2000;95:503-508.
Graham DY, Smith JL. The course of patients after variceal hemorrhage. Gastroenterology 1981;80:800-809.
Bosch J, Albrades J. Variceal bleeding: pharmacological therapy. Dig Dis 2005;23:18-29.
Lebrec D, Poynard T, Hillon P, Benhamou J-P. Propranolol for prevention of recurrent gastrointestinal bleeding in patients with cirrhosis: a controlled study. N Engl J Med 1981;305:1371-1374.
Talwalkar JA, Kamath PS. An evidence-based medicine approach to beta-blocker therapy in patients with cirrhosis. Am J Med 2004;116:759-766.
Saab S, DeRosa V, Nieto J, Durazo F, Han S, Roth B. Costs and clinical outcomes of primary prophylaxis of variceal bleeding in patients with hepatic cirrhosis: a decision analytic model. Am J Gastroenterol 2003;98:763-770.
Sarin SK, Lamba GS, Kumar M, Misra A, Murthy NS. Comparison of endoscopic ligation and propranolol for the primary prevention of variceal bleeding. N Engl J Med 1999;340:988-993.
Jutabha R, Jensen DM, Martin P, Savides T, Han SH, Gornbein J. Randomized study comparing banding and propranolol to prevent initial variceal hemorrhage in cirrhotics with high-risk esophageal varices. Gastroenterology 2005;128:870-881.
Groszmann RJ, Garcia-Tsao G, Bosch J, et al. Beta-blockers to prevent gastroesophageal varices in patients with cirrhosis. N Engl J Med 2005;353:2254-2261.
Vorobioff J, Groszmann RJ, Picabea E, et al. Prognostic value of hepatic venous pressure gradient measurements in alcoholic cirrhosis: a 10-year prospective study. Gastroenterology 1996;111:701-709.
Bernard B, Lebrec D, Mathurin P, Opolon P, Poynard T. Beta-adrenergic antagonists in the prevention of gastrointestinal rebleeding in patients with cirrhosis: a meta-analysis. Hepatology 1997;25:63-70.
Targownik LE, Spiegel BM, Dulai GS, Karsan HA, Gralnek IM. The cost-effectiveness of hepatic venous pressure gradient monitoring in the prevention of recurrent variceal hemorrhage. Am J Gastroenterol 2004;99:1306-1315.
de Franchis R. Evolving consensus in portal hypertension: report of the Baveno IV consensus workshop on methodology of diagnosis and therapy in portal hypertension. J Hepatol 2005;43:167-176.(Sandeep Mukherjee, M.D., )
Although the risk of bleeding is low in patients without varices (1 to 2 percent per year) or with small varices (5 percent per year), small varices have a median rate of progression to larger varices of 12 percent per year, and large varices, in turn, have an annual risk of bleeding of 15 percent.3 The lifetime risk of at least one bleeding episode is 30 percent among patients with cirrhosis and varices, with the greatest risk among those with large varices and decompensated liver disease. Since a single episode of uncontrolled variceal bleeding has an immediate rate of death of 5 to 8 percent and a six-week mortality rate of 50 percent, preventing varices from bleeding is of paramount importance.4 Primary prophylaxis, or the prevention of a first episode of variceal bleeding in a patient with cirrhosis with known varices, is often managed with nonselective beta-blockers such as propanolol, nadolol, or timolol.5 The mechanism of action of these medications is complex and varies according to dose. Low doses cause unopposed splanchnic vasoconstriction, leading to reduced portal blood flow and, therefore, decreased portal hypertension. As the dose is increased, portal flow is further reduced by the negative effect of these medications on cardiac output. Furthermore, there is also evidence that these medications decrease the development of portosystemic collaterals, an important precursor to the formation of varices.
Since the original study by Lebrec et al. in 1981,6 several randomized trials and cost-effectiveness analyses have conclusively demonstrated that nonselective beta-blockers are the medications of first choice for primary prophylaxis against variceal bleeding in patients with varices.7,8 More recently, randomized, controlled trials have reported a lower rate of variceal bleeding among patients treated with endoscopic banding — a procedure usually reserved for acutely bleeding varices or secondary prophylaxis — than among patients treated with beta-blockers, although no effect on mortality was noted.9,10 However, these studies are limited by their short follow-up and small size. Until these issues are addressed in future studies, gastroenterologists will continue to choose nonselective beta-blockers for primary prophylaxis in most patients and reserve endoscopic interventions for primary prophylaxis in patients who cannot tolerate beta-blockers or in those with large varices who live far from medical care.
A potential extension of the finding that beta-blockers are beneficial in patients with varices would be the use of beta-blockers to prevent variceal formation in patients with cirrhosis, an intuitively attractive proposition, given the high lifetime risk of bleeding once large varices develop. In this issue of the Journal, Groszmann et al. report the results of a randomized, double-blind, placebo-controlled study evaluating the effectiveness of timolol for the primary prevention of varices in patients with cirrhosis.11 This study, conducted at four centers renowned for their expertise in portal hypertension, randomly assigned 213 patients with cirrhosis to receive timolol or placebo. The primary objectives of the study were to determine whether timolol prevented gastroesophageal varices and whether baseline and serial measurements of the hepatic venous pressure gradient (HVPG) could be used to predict the development of varices.
The investigators reported that timolol not only did not prevent gastroesophageal varices but also was associated with a higher incidence of adverse events, many of which were serious. They also demonstrated that the risk of variceal formation was decreased among patients with a baseline HVPG of less than 10 mm Hg and among patients whose HVPG decreased by more than 10 percent per year, irrespective of timolol use.
The relatively low baseline portal pressures in the study population may have played a role in the failure to demonstrate a benefit of timolol. The median value for HVPG in both study groups was only 11.7 mm Hg, suggesting that patients were at low risk for varices. Portal hypertension, defined by an HVPG exceeding 5 mm Hg, is usually not clinically significant until the gradient exceeds 12 mm Hg, at which point variceal bleeding and ascites are more likely to develop. Conversely, in patients with bleeding varices or refractory ascites treated by transjugular intrahepatic systemic shunts, a successful outcome is defined by a reduction of the HVPG to less than 12 mm Hg. However, not all patients whose HVPG exceeds 12 mm Hg have varices, and the relationship between a rising gradient and the development of varices remains poorly understood. To complicate matters further, the HVPG is also increased by alcohol use and worsening liver function and can improve with abstinence from alcohol use.12
The inclusion of 14 patients (27 percent) who had been abstinent from alcohol use for less than one month and 5 patients with alcoholic hepatitis may also have contributed to the study's negative findings. Since alcohol consumption was not permitted during the study, the cessation of alcohol use among some patients may have decreased their HVPG and risk of varices. It would have been prescient to use a conventional period of abstinence, such as six months, to avoid the confounding effects of alcohol cessation on the development of varices. Another important point to emphasize is that although the investigators reported that similar percentages of patients in each group reached the primary end point, small varices accounted for 83 percent of these end points in both groups, and small varices are unlikely to bleed and are of less clinical significance than large varices.
The choice of timolol for this study was interesting, since prior trials of portal hypertension have evaluated nadolol or propranolol. Although studies have demonstrated the potent negative inotropic effects of timolol in healthy volunteers and the acute portal-pressure–lowering effects in patients with cirrhosis, there have been no studies of its effectiveness with long-term use in patients with cirrhosis. The high frequency of adverse events (48 percent in the timolol group, as compared with 32 percent in the placebo group) is of great concern, particularly since prior randomized trials comparing propranolol or nadolol with placebo have reported adverse events in fewer than 17 percent of patients receiving beta-blockers.13
What role, if any, do nonselective beta-blockers have in preventing variceal formation in patients with cirrhosis? The negative findings of the study by Groszmann et al. clearly do not support the empirical use of these medications in patients who have cirrhosis without varices, since the risks far outweigh the benefits. This conclusion may surprise clinicians, particularly since cost-effectiveness analyses have historically promoted the use of these drugs, further reinforcing the importance of well-conducted clinical trials.8 However, a very important and novel finding of this study is the pivotal role of HVPG monitoring as a diagnostic and prognostic tool in patients with cirrhosis. Unfortunately, HVPG monitoring is rarely used outside the realms of clinical research. Gastroenterologists often use vital signs to monitor the effectiveness of beta-blockade or use upper endoscopy to screen patients for varices. The findings of this study, together with the report by Targownik et al. demonstrating the cost-effectiveness of HVPG monitoring,14 should encourage physicians to consider this procedure as an integral part of their clinical armamentarium in the treatment of patients with cirrhosis and to view it as complementary to rather than competing with upper endoscopy. Until we can accurately predict when varices will develop, the results of this important trial suggest that nonselective beta-blockers should not be used to prevent varices in patients. However, these findings should not affect current recommendations to use beta-blockers, alone or in conjunction with endoscopic banding, to prevent bleeding in patients with established esophageal varices.15
Source Information
From the University of Nebraska Medical Center, Omaha.
References
Harry R, Wendon J. Management of variceal bleeding. Curr Opin Crit Care 2002;8:164-170.
Wright AS, Rikkers LF. Current management of portal hypertension. J Gastrointest Surg 2005;9:992-1005.
Zoli M, Merkel C, Magalotti D, et al. Natural history of cirrhotic patients with small esophageal varices: a prospective study. Am J Gastroenterol 2000;95:503-508.
Graham DY, Smith JL. The course of patients after variceal hemorrhage. Gastroenterology 1981;80:800-809.
Bosch J, Albrades J. Variceal bleeding: pharmacological therapy. Dig Dis 2005;23:18-29.
Lebrec D, Poynard T, Hillon P, Benhamou J-P. Propranolol for prevention of recurrent gastrointestinal bleeding in patients with cirrhosis: a controlled study. N Engl J Med 1981;305:1371-1374.
Talwalkar JA, Kamath PS. An evidence-based medicine approach to beta-blocker therapy in patients with cirrhosis. Am J Med 2004;116:759-766.
Saab S, DeRosa V, Nieto J, Durazo F, Han S, Roth B. Costs and clinical outcomes of primary prophylaxis of variceal bleeding in patients with hepatic cirrhosis: a decision analytic model. Am J Gastroenterol 2003;98:763-770.
Sarin SK, Lamba GS, Kumar M, Misra A, Murthy NS. Comparison of endoscopic ligation and propranolol for the primary prevention of variceal bleeding. N Engl J Med 1999;340:988-993.
Jutabha R, Jensen DM, Martin P, Savides T, Han SH, Gornbein J. Randomized study comparing banding and propranolol to prevent initial variceal hemorrhage in cirrhotics with high-risk esophageal varices. Gastroenterology 2005;128:870-881.
Groszmann RJ, Garcia-Tsao G, Bosch J, et al. Beta-blockers to prevent gastroesophageal varices in patients with cirrhosis. N Engl J Med 2005;353:2254-2261.
Vorobioff J, Groszmann RJ, Picabea E, et al. Prognostic value of hepatic venous pressure gradient measurements in alcoholic cirrhosis: a 10-year prospective study. Gastroenterology 1996;111:701-709.
Bernard B, Lebrec D, Mathurin P, Opolon P, Poynard T. Beta-adrenergic antagonists in the prevention of gastrointestinal rebleeding in patients with cirrhosis: a meta-analysis. Hepatology 1997;25:63-70.
Targownik LE, Spiegel BM, Dulai GS, Karsan HA, Gralnek IM. The cost-effectiveness of hepatic venous pressure gradient monitoring in the prevention of recurrent variceal hemorrhage. Am J Gastroenterol 2004;99:1306-1315.
de Franchis R. Evolving consensus in portal hypertension: report of the Baveno IV consensus workshop on methodology of diagnosis and therapy in portal hypertension. J Hepatol 2005;43:167-176.(Sandeep Mukherjee, M.D., )