Endoscopic treatments for portal hypertension
- 247 Downloads
Acute esophageal variceal hemorrhage is a dreaded complication of portal hypertension. Its management has evolved rapidly in recent years. Endoscopic therapy is often employed to arrest bleeding varices as well as to prevent early rebleeding. The combination of vasoconstrictor and endoscopic therapy is superior to vasoconstrictor or endoscopic therapy alone for control of acute esophageal variceal hemorrhage. After control of acute variceal bleeding, combination of banding ligation and beta-blockers is generally recommended to prevent variceal rebleeding. To prevent the catastrophic event of acute variceal bleeding, endoscopic banding ligation is an important tool in the prophylaxis of first bleeding. Endoscopic obturation with cyanoacrylate is usually utilized to arrest acute gastric variceal hemorrhage as well as to prevent rebleeding. It can be concluded that endoscopic therapies play a pivotal role in management of portal hypertensive bleeding.
KeywordsEsophageal varices Sclerotherapy Banding ligation Gastric varices Endoscopic obturation
Many etiologies are associated with development of portal hypertension. Collaterals generally develop to decompress the elevated pressure and blood flow. Collaterals that develop and protrude to the lumen of the gastrointestinal tract become so-called varices, most commonly encountered growing around the lower esophagus and cardia. Esophageal and gastric varices can be easily detected by endoscopy. Acute esophageal variceal hemorrhage is one of the most dreaded complications of portal hypertension, with mortality of up to 40% and rebleeding rate of 60% in survivors [1, 2, 3]. Fortunately, the mortality of acute esophageal variceal hemorrhage has decreased to 20% in recent years . Other than hemodynamic and clinical parameters, endoscopic findings of variceal sizes and red color signs are important predictive factors for variceal rupture. Hence, preventing first variceal bleeding, treatment of acute variceal bleeding, as well as prevention of variceal rebleeding are all integral parts of the management of complications related to portal hypertension . Endoscopic therapy plays a pivotal role in management of all three aspects of variceal bleeding. This article focuses on the role of endoscopy in management of portal hypertension.
Timing of endoscopy for portal hypertensive bleeding
It is generally agreed that cirrhotic patients suffering from acute upper gastrointestinal hemorrhage should receive endoscopy to confirm the bleeding sources . Prior to endoscopic therapy, resuscitation is required and vasoconstrictors should be instituted. However, the most appropriate time to perform endoscopic examination remains undetermined. One study showed that patients presenting with stable vital signs and no signs of active bleeding can be managed with vasoconstrictors and receive endoscopic examination during working hours of endoscopists . On the other hand, some studies have shown that patients with portal hypertension presenting with hematemesis should receive emergency endoscopic therapy to achieve a higher hemostatic rate [7, 8]. The latest Baveno Consensus on portal hypertension suggests that endoscopy should be performed within 12 h for cirrhotic patients with acute upper gastrointestinal bleeding .
Modalities of endoscopic therapy for esophageal variceal bleeding
Endoscopic injection sclerotherapy (EIS) has been widely employed to arrest acute variceal bleeding for more than three decades [9, 10]. The volume of sclerosant in each injection and the interval of treatment vary greatly among operators, and no standard EIS method could be established . The advantages of EIS include easy technique and highly effective hemostatic rate. Before the advent of endoscopic variceal ligation (EVL), EIS was the most widely used endoscopic therapy in management of acute variceal bleeding. The hemostatic rate achieved by emergency EIS ranged between 60 and 100%.
Comparison between EIS and EVL
For acute variceal bleeding
Active variceal bleeding
Technically more difficult
Similar efficacy or inferior
Similar efficacy or superior
Very early rebleeding rate
Acute impact on portal pressure
For prevention of rebleeding
Days to weeks
Treatment sessions required for
Rates of variceal obliteration
Post treatment ulcers
Less common but deeper
More common but shallower
Post treatment ulcer bleeding
Similar or worse
Similar or better
Prophylaxis of first bleeding
Injection of tissue adhesive agent Histoacryl, a method widely used to arrest gastric variceal bleeding , has also been employed to treat acute esophageal variceal hemorrhage with some success . Use of tissue adhesive in esophageal variceal hemorrhage is not common, possibly due to the high effectiveness of EVL.
Acute esophageal variceal hemorrhage
Comparison between vasoconstrictors and EIS
Emergency EIS has been compared with vasopressin, terlipressin, octreotide, and somatostatin. A meta-analysis of six trials comparing EIS with vasopressin or terlipressin showed a significant advantage of EIS in controlling bleeding. However, the difference between EIS and somatostatin or octreotide in controlling of bleeding was not significant . Of note was that EIS was associated with more complications than vasoconstrictors. Thus, it has been suggested that vasoconstrictors could replace EIS as first-line management of variceal bleeding . Another meta-analysis including 15 studies comparing EIS with vasoconstrictors showed that the efficacy of EIS was a median of 83%. Failure to control bleeding was less frequent with EIS with a difference of 3.4%, and fewer deaths with EIS with a difference of 4.3%. However, EIS was associated with more frequent complications, with a difference of 8.8% .
Comparison between balloon tamponade and sclerotherapy
A controlled study to compare the relative efficacy of hemostasis between EIS and balloon tamponade was performed by Paquet et al. in 1985 . That study showed that hemorrhage was definitively controlled in 52% of patients in the balloon tamponade group and in 90% of patients in the EIS group (p < 0.01). The frequency of complications was similar in the two groups. This study proved the superiority of EIS over balloon tamponade in terms of control of hemorrhage and improvement of survival. Nowadays, balloon tamponade is rarely used as first-line therapy for esophageal variceal hemorrhage.
Comparison between EVL and EIS
Comparison between vasoconstrictors and EVL
Since vasoconstrictors have been proven to be a viable substitute for EIS, it is useful to know whether vasoconstrictors can also replace EVL as the therapy of choice for initial control of acute esophageal variceal hemorrhage. Only one study has compared EVL and somatostatin in the control of acute esophageal variceal hemorrhage. This controlled study included 62 patients in the EVL group and 63 patients in the somatostatin group. Three patients (4.8%) in the ligation group and 20 patients (31.7%) in the somatostatin group encountered treatment failure (p = 0.0001). The results demonstrated that EVL was superior to somatostatin in the control of hemorrhage without enhancing adverse events . No study has compared EVL with other vasoconstrictors alone in management of acute esophageal variceal bleeding. Nowadays, neither vasoconstrictors nor endoscopic therapy alone is advised for management of acute esophageal variceal bleeding.
Comparison between vasoconstrictors plus endoscopic therapy versus endoscopic therapy alone
Banares performed a meta-analysis including 939 patients from eight trials to evaluate the benefits of combination of endoscopic therapy with vasoconstrictors . The results showed that combined treatment improved initial control of bleeding and 5-day hemostasis as compared with endoscopic therapy alone, but without improvement in survival. Severe adverse events were similar in both groups. The reported efficacy of EIS alone with median of 69% in these studies was lowest compared with other studies. Thus, the benefit of combined therapy has been regarded as an overestimation .
Comparison between vasoconstrictors plus EVL versus vasoconstrictor alone
The role of endoscopic therapy has been questioned in patients with no active bleeding at endoscopy. A controlled study  showed that 48-h hemostasis was achieved in 91% in the terlipressin group and 98% in the combined EVL and terlipressin group (p = 0.20). However, very early rebleeding within 48–120 h occurred in seven patients (15%) in the terlipressin group but not in any patients (0%) in the combined group (p = 0.006). Treatment failure was 24% in the terlipressin group and 2% in the combined group (p = 0.002). This study showed that EVL is still required in patients with inactive variceal bleeding at endoscopy.
Comparison between EVL plus vasoconstrictor and EIS plus vasoconstrictor
Villanueva performed a trial to compare emergency EVL plus somatostatin (89 patients) versus EIS plus somatostatin (90 patients) in patients with acute esophageal variceal hemorrhage . Therapeutic failure occurred in 21 patients treated with sclerotherapy (24%) and in 9 treated with EVL (10%). Failure to control bleeding occurred in 15 versus 4%, respectively (p = 0.02). Serious-side effects occurred in 13% in EIS group and 4% in EVL group (p = 0.04). Six-week survival probability was better with EVL (p = 0.01). This study proved that use of EVL instead of EIS as emergency endoscopic therapy added to somatostatin for treatment of acute variceal bleeding significantly improves efficacy and safety.
Prevention of esophageal variceal rebleeding
Comparison between EVL and EIS
Laine performed a meta-analysis of seven studies and showed that EVL is more effective than EIS for reducing the rebleeding rate . The number of endoscopic treatment sessions required to achieve variceal obliteration was lower with ligation than with sclerotherapy. Another meta-analysis including 13 articles performed by de Franchis in 1999 revealed a strong benefit for EVL in decreasing variceal rebleeding (odds ratio 0.46) and similar survival between patients treated with EIS and those with EVL . The mean number of sessions required to achieve variceal obliteration was reduced from 5.4 in patients receiving EIS to 3.6 in patients receiving EVL. This may significantly shorten the duration required to achieve variceal obliteration. Recurrent variceal bleeding reduced from 8–53% in the EIS groups to 2–36% for patients receiving EVL. The high incidence of bacterial infections, esophageal stricture, and ulcer bleeding associated with EIS was significantly reduced with EVL . EVL is currently considered the endoscopic treatment of choice for patients to prevent esophageal variceal rebleeding . The main disadvantage of EVL is possibly higher frequency of recurrent varices .
Comparison between EIS and medical therapy
Both beta-blockers and EIS were important and popular modalities for prevention of variceal rebleeding during the 1980s. Studies comparing EIS and beta-blockers were widely performed. A meta-analysis of nine trials comparing beta-blockers with EIS showed a significant reduction of rebleeding in favor of EIS . However, significantly more complications were encountered in patients receiving EIS, while survival was similar between both therapies. On the other hand, Villanueva et al. showed that combination of nadolol and isosorbide mononitrate was superior to EIS in the reduction of variceal rebleeding .
Comparison between EVL and drug therapy
Six controlled trials comparing the combination of nadolol and nitrate with EVL in the prevention of variceal rebleeding have been reported [42, 43, 44, 45, 46, 47]. Meta-analysis of the first four studies showed similar survival between pharmacologic therapy and EVL , suggesting that either medication with beta-blockers plus nitrate or EVL can be used to prevent esophageal variceal rebleeding. Long-term observation by us demonstrated that EVL was definitely superior to combination drug therapy in preventing rebleeding from esophageal varices. However, possibly owing to systemic effects, survival in patients treated with combination drug therapy appeared to be better than those treated with repeated EVL without drug therapy .
Comparison between EVL plus EIS and EVL or EIS alone
Combination of EVL and EIS to eradicate esophageal varices has ignited enthusiasm among endoscopists. Paraesophageal varices have been shown to be obliterated by EIS but not by EVL . The combination of EIS and EVL can be synchronous or metachronous. Singh et al. performed a meta-analysis including seven studies assessing the efficacy of combining EVL and EIS . Combination therapy achieved significantly higher variceal obliteration than EVL alone, however the overall rebleeding rate was similar between patients receiving EVL alone and combination therapy, i.e., 22 versus 20%, respectively. Survival was also similar between both methods of treatment. Moreover, combination therapy lengthened the treatment time and was associated with higher incidence of esophageal stricture. Thus, synchronous combination of EVL and EIS has been discarded.
On the other hand, a few trials showed that metachronous combination therapy of EIS and EVL could reduce variceal recurrence or even reduce incidence of variceal rebleeding as compared with treatment with EVL or EIS alone [52, 53]. Thus, metachronous combination therapy of EIS and EVL is favored over their synchronous combination. To reduce variceal recurrence, repeated EVL followed by microwave or argon plasma has also been shown to be a feasible method [54, 55].
Combination of EVL and drug therapy
It is reasonable to combine a local therapy, i.e., EVL, and a systemic therapy, beta-blockers, to reduce esophageal variceal rebleeding. The first study with a combination of nadolol, sucralfate, a mucosal protector, and EVL compared with EVL alone was carried out by us. The results showed that combination therapy was superior to EVL alone in terms of variceal rebleeding rates (12 versus 29%) and variceal recurrence (26 versus 50%) . A similar study by de la Pena also consistently showed that combination of EVL and beta-blockers was superior to EVL alone in reducing variceal rebleeding and prevention of variceal recurrence . A meta-analysis including 1860 patients from 23 studies with the combination of EIS or EVL and beta-blockers was performed . The results found that combination of either EIS or EVL and drug therapy reduced overall rebleeding and variceal rebleeding more than either therapy alone, with similar survival. On the other hand, it is interesting to evaluate whether EVL could enhance the efficacy of beta-blockers plus isosorbide mononitrate (ISMN) in the prevention of variceal rebleeding. Current studies showed that addition of EVL to β-blockers and ISMN achieves a nonsignificant decrease of rebleeding with no effect on mortality [59, 60].
Comparison between endoscopic therapy and TIPSS
Transjugular intrahepatic portosystemic stent shunt (TIPSS) has been widely employed in management of variceal bleeding in the last two decades [61, 62]. Meta-analyses disclosed that variceal rebleeding was significantly reduced by TIPSS (19%) as compared with endoscopic therapy (47%) . However, TIPSS is associated with significantly higher incidence of hepatic encephalopathy, and survival was similar between both therapies. Thus, TIPSS is not favored as first-line therapy for preventing variceal rebleeding but rather as rescue therapy for endoscopic and pharmacological treatment failure .
Primary prophylaxis for first bleeding from esophageal varices
It is estimated that esophageal varices may be noted in 50% of cirrhotic patients. Since variceal bleeding may lead to high incidence of mortality and rebleeding, primary prophylaxis of first bleeding from esophageal varices is mandatory in patients with high-risk esophageal varices.
Comparison between EIS and no treatment
Regarding modalities to prevent first bleeding from esophageal varices, EIS and portacaval shunt operation were widely adopted from 1980 to 1990. In 1991, a study from the USA showed that EIS as compared with sham therapy may be associated with enhanced mortality for unknown reasons and suggested that EIS should not be performed until variceal bleeding. Since then, EIS has rarely been utilized to prevent first variceal bleeding .
Comparison between EVL and no treatment
EVL has been compared with no treatment for prophylaxis of first variceal bleeding. Imperiale performed a meta-analysis of five trials comparing EVL with untreated control groups; the relative risks of first variceal bleed, bleed-related mortality, and all-cause mortality were 0.36, 0.20, and 0.55, respectively. These results showed that EVL is definitely superior to no treatment in the reduction of first variceal bleeding and mortality in cirrhotic patients . Currently, it is regarded as unethical to leave high-risk varices untreated.
Comparison between EVL and beta-blockers
Both beta-blockers and banding ligation are widely employed to prevent first variceal bleeding [69, 70, 71, 72]. A meta-analysis showed that, among 283 subjects from four trials comparing EVL with beta-blocker, the relative risk of first variceal bleed was 0.48, but there was no effect on either bleed-related mortality or all-cause mortality. It is thus suggested that prophylactic ligation should be considered for patients with large esophageal varices who cannot tolerate beta-blockers . Khuroo performed another meta-analysis and showed that EVL compared with beta-blockers significantly reduced rates of first gastrointestinal bleed by 31% and first variceal bleed by 43%. All-cause and bleed-related deaths were unaffected. Severe adverse events were significantly less in EVL compared with beta-blockers .
On the other hand, Tripathi et al. demonstrated that carvedilol was superior to EVL in the prevention of first variceal bleeding, i.e., 10 versus 23%, with similar survival . It is generally recommended that beta-blockers be used as first-line therapy for high-risk esophageal varices, with EVL being reserved for patients who cannot tolerate or with contraindications to beta-blockers .
Comparison between EVL and EIS
Only one Czech study has compared EVL and EIS in the prophylaxis of first bleeding from high-risk varices . Akin to studies on prevention of rebleeding, the number of sessions required for variceal eradication was lower in EVL group. EIS group was associated with lower rate of variceal recurrence and higher incidence of complications. The variceal bleeding rate and survival were similar in the two groups. Since EIS is associated with higher incidence of complications, it is not recommended to prevent first variceal bleeding.
Combination of beta-blockers and banding ligation
Combination of EVL and beta-blockers has been recommended as standard therapy for prevention of variceal rebleeding. However, the role of combination therapy is not yet established in primary prophylaxis of first bleeding. Two controlled studies have compared the combination of EVL and beta-blockers with either EVL or beta-blockers alone in preventing first variceal bleeding [76, 77]. However, no significant benefit was noted in the combination therapy.
Though the incidence of gastric variceal bleeding is lower than that of esophageal variceal bleeding, rupture of gastric varices usually results in more severe hemorrhage and higher mortality than esophageal varices . Optimal therapy for bleeding gastric varices was not established until recent years.
Comparison between EIS and glue injection for gastric variceal hemorrhage
Endoscopic injection sclerotherapy was the earliest available endoscopic method. Three studies have compared EIS using ethanolamine or absolute alcohol with Histoacryl (glue) in treatment of acute gastric variceal bleeding . The results showed that glue injection was superior to EIS in the control of acute bleeding and prevention of rebleeding. As the rebleeding rate may be as high as 89%, sclerotherapy has been abandoned by most endoscopists .
Comparison between EVL and glue injection for gastric variceal hemorrhage
Three controlled trials including 194 patients with acute gastric variceal bleeding have been carried out [80, 81, 82, 83]. Active bleeding control was achieved in 94% of patients in the cyanoacrylate glue injection group, compared with 80% in the EVL group (p = 0.03). The rebleeding rate was comparable in type 2 gastroesophageal varices (GOV2) between the two interventions, but glue injection was superior in reducing rebleeding of type 1 gastroesophageal varices (GOV1, 26 versus 48%) and isolated gastric varices (IGV1, 17 versus 86%). There were no differences in complications or mortality between the two interventions. Complications of glue injection include thrombosis, bacterial infection, ulcer bleeding, and inadvertent clotting of injection needles . Endoscopic obturation with glue injection has gained worldwide popularity except in the USA. GOV1 varices are usually treated as esophageal varices with EVL . However, glue injection was proven to be more effective [85, 86]. Combination of endoscopic therapy and beta-blockers to prevent gastric variceal rebleeding is not proven to be superior to endoscopic therapy alone . Nevertheless, the combination is still favored for patients with concomitant gastric and esophageal varices .
Comparison between glue injection and TIPSS
Two cohort and one prospective study have compared cyanoacrylate versus TIPSS in treatment of gastric variceal hemorrhage [88, 89, 90]. Meta-analysis showed that the pooled incidence of variceal rebleeding in the TIPSS group was significantly lower than that in the EIS group. However, the risk of hepatic encephalopathy was significantly increased in the TIPSS group . Thus, glue injection instead of TIPSS is considered as the first choice in patients to prevent gastric variceal rebleeding.
Thrombin injection, using either bovine or human thrombin to convert fibrinogen to fibrin clot, has been reported mostly in European countries . A 5-ml solution containing 5000 units of thrombin can clot a liter of blood within 60 s. The efficacy of thrombin injection in hemostasis ranged from 70 to 100%, and rebleeding rates ranged from 0 to 27%. Associated side-effects were rarely reported. In view of the technical ease, effectiveness, and safety of thrombin injection, it is believed to be promising for treatment of bleeding gastric varices.
Comparison between glue injection and beta-blockers in primary prophylaxis of gastric variceal bleeding
Primary prophylaxis of first bleeding from gastric varices has rarely been investigated. Mishra et al. performed a controlled trial in patients with large gastric varices which had never bled. The actuarial probability of bleeding from gastric varices over median follow-up of 26 months was lower and the survival was higher in the cyanoacrylate group as compared with the no-treatment group. It appeared that glue injection is effective in primary prophylaxis of high-risk gastric variceal bleeding . Since the sample size was very small, more studies are still required.
Ectopic variceal bleeding
Patients with portal hypertensive may develop varices along the gastrointestinal tract other than gastroesophageal varices. Because the incidence of ectopic variceal bleeding is appreciably low as compared with bleeding arising from gastroesophageal varices, no modalities have been well documented to be both effective and safe in the management of ectopic variceal bleeding. Endoscopic therapies including EIS, EVL, argon plasma coagulation, and endoscopic glue obturation have been reported to be effective in treating duodenal, small bowel, colon, and rectal varices [94, 95]. A retrospective analysis from Japan showed that EIS may be associated with lower incidence of rebleeding as compared with EVL in the management of rectal variceal bleeding .
Portal hypertensive gastropathy (PHG) bleeding
Portal hypertensive gastropathy refers to changes in the mucosa of the stomach in patients with portal hypertension. Portal hypertensive gastropathy may appear as snake skin mosaic pattern, black–brown spots, or even active hemorrhage . Endoscopic therapy of varices may accentuate PHG [97, 98]. Treatment of PHG is aimed at reducing hepatic venous pressure gradients, most often by pharmacologic means, but may require shunt procedures in severe cases. Endoscopic treatment of acute bleeding secondary to PHG is generally ineffective. Endoscopic therapy with argon plasma coagulation (APC) or coagulation therapy may be considered on an individual basis .
Hemospray is a proprietary mineral blend powder developed specifically for endoscopic hemostasis. Several case reports have shown successful hemostasis of PHG bleeding by use of Hemospray .
Endoscopy is nowadays of paramount importance in assessing the severity of gastroesophageal varices as well as gastropathy for patients with portal hypertension. Endoscopic examination is still required to confirm the presence of varices. A wide armamentarium can be employed to manage variceal bleeding. EVL has currently replaced EIS as endoscopic therapy of choice in the arresting of active variceal bleeding as well as inactive variceal bleeding at endoscopy. Among patients with acute variceal bleeding and endoscopic treatment failure, rescue therapy with balloon tamponade, self-expanding metal stent, or TIPSS may be indicated, depending on available expertise and facility. For prevention of esophageal variceal rebleeding, EVL plus beta-blockers is regarded as the first choice. TIPSS should be considered for patients with repeated rebleeding despite endoscopic therapy. To prevent first esophageal variceal bleeding, EVL is a good alternative to beta-blockers.
Regarding treatment of gastric variceal bleeding, glue injection has been proved superior compared with EIS or EVL in arresting acute bleeding as well as preventing rebleeding. Glue injection may also be applied in the prevention of first bleeding from gastric varices.
In summary, endoscopic examinations and endoscopic therapy have played important roles in portal hypertensive bleeding over the last three decades and may continue to do so in the future.
Compliance with Ethical Standards
This study was not funded.
Conflict of Interest
Gin-Ho Lo declares that he has no conflicts of interest.
This article does not contain any studies with human participants or animals performed by the author.
- 32.Laine L, Planas R, Nevens F, Rafael Bañares, David Patch Jaime Bosch. Treatment of the acute bleeding episode. In: de Franchis R. Portal hypertension IV: proceedings of the fourth Baveno International Consensus Workshop. Blackwell Publishing 2006; p 217–242Google Scholar
- 45.Romero G, Kravetz D, Argonz J, Vulcano C, Suarez A, Fassio E, et al. Comparative study between nadolol and 5-isosorbide mononitrate vs. endoscopic band ligation plus sclerotherapy in the prevention of variceal rebleeding in cirrhotic patients: a randomized controlled trial. Aliment Pharmacol Ther. 2006;24:601–611CrossRefPubMedGoogle Scholar
- 46.Sarin SK, Wadhawan M, Gupta R, Shahi H. Evaluation of endoscopic variceal ligation (EVL) versus propanolol plus isosorbide mononitrate/nadolol (ISMN) in the prevention of variceal rebleeding: comparison of cirrhotic and noncirrhotic patients. Dig Dis Sci. 2005;50:1538–1547CrossRefPubMedGoogle Scholar
- 48.Garcia-Pagan JC, Bosch J. Endoscopic band ligation in the treatment of portal hypertension. NCP Gastroenterol Hepatol. 2005;2:526–535Google Scholar
- 61.Sanyal AJ, Freedman AM, Luketic VA, Purdum PP, Shiffman ML, Tisnado J, et al. Transjugular intrahepatic portosystemic stent shunt compared with endoscopic therapy for the prevention of recurrent variceal hemorrhage: a randomized, controlled trial. Ann Intern Med. 1997;126:849–857CrossRefPubMedGoogle Scholar
- 78.Sarin SK, Primignani M, Agarwal SR. Gastric varices. In: de Franchis R, editor. Portal hypertension III. In: Proceedings of the third Baveno International Consensus Workshop on definitions, methodology and therapeutic strategies. Oxford: Blackwell, Science 2001; 76–96Google Scholar
- 88.Mahadeva S, Bellamy MC, Kessel D, Davies MH, Millson CE. Cost-effectiveness of N-butyl-2-cyanoacrylate (histoacryl) glue injections versus transjugular intrahepatic portosystemic shunt in the management of acute gastric variceal bleeding. Am J Gastroenterol. 2003;98:2688–2693CrossRefPubMedGoogle Scholar