Synopsis  

Variceal bleeding remains one of the most dreadful complications of portal hypertension, causing high mortality. In the last few decades major advances have been made in the treatment of variceal bleeding including endoscopic hemostasis, the use of safe vasoactive agents, and radiological shunting. Among these options, endoscopic variceal ligation stands out as the most promising therapy and hence should be considered a first-line treatment. In many cases, combinations of these treatment modalities offer the best results to control bleeding and improve survival.

 Background   

Hepatic blood flow is normally about 1.5 liters/min, representing 20–25% of cardiac output. Approximately two-thirds of this flow come from the portal venous system and, with obstruction to this blood flow, the pressure in the portal venous system builds up rapidly. When the pressure of the portal vein exceed 12 mmHg, variceal bleeding may occur. Variceal hemorrhage is the most serious complication of portal hypertension and constitutes about one-fifth of the mortality in cirrhosis. The mortality rate of variceal hemorrhage without treatment could be as high as 70%.

 National history of variceal bleeding   

Thirty percent of patients with compensated cirrhosis and 60% of patients with decompensated cirrhosis have gastroesophageal varices at the time of presentation [1,2]. Among cirrhotic patients who have no esophageal varices on diagnosis, the incidence of new varices ranges between 5% and 23% after 1 year of follow-up [3–5]. In a recent prospective study, D'Amico et al. reported that the incidence of varices in newly diagnosed compensated cirrhosis is around 4.5% per year and mortality from bleeding 0.5% per year [6].

The risk of developing varices parallels with the severity of liver disease as reflected by Child's grading. It is further suggested that platelet count also predicts the development of varices [6]. The rate of growth of varices from small to large size in compensated patients is faster than the rate of de novo appearance of varices. Among patients with cirrhosis and esophageal varices, the incidence of first variceal hemorrhage ranges from 20 to 40% within 2 years. A recent meta-analysis studying the effects of prophylactic beta-blocker treatment revealed that the mean weighted bleeding rate of the control group (without treatment) at 2 years is around 24% [7].

Interestingly, only one-third of patients with esophageal varices will bleed from their varices in their lifetime [8]. It also means that many patients with gastroesophageal varices may never bleed. Better understanding of risk factors for gastroesophageal variceal bleeding and the capability of predicting which patients with varices will bleed are thus important to plan timely treatment.

Mechanism of bleeding   

Bleeding from gastroesophageal varices is now widely accepted as a phenomenon of 'explosion' instead of 'erosion'. Thus the risk of hemorrhage is related to the size of the varices, wall thickness, and intravariceal pressure according to LePlace's law. Several groups have confirmed that variceal size is the most important prognostic factor for variceal bleeding [9,10].

Variceal stigmata   

Red color signs, which include cherry-red spots and red wale markings, are also associated with more advanced grade of varices and higher risk of hemorrhage (Fig. 1). These signs are thought to represent focal weakness or 'blowouts' in the variceal wall. Fibrin clots or 'white nipple sign' are occasionally seen over the variceal columns which have recently bled (Fig. 2).

Risk of bleeding   

Beside gross appearance of varices, patients with decompensated cirrhosis would definitely fare worse than those with compensated liver function. Alcoholic hepatitis and portal vein thrombosis are other poor prognostic indicators. The risk of bleeding from varices is correlated to the Child–Pugh's score, not just because of more severe portal hypertension, but also because it is related to other factors such as nutritional deficiency, coagulopathy, and increased fibrinolysis in advanced liver disease. Recently, an association between bacterial infection and failure to control bleeding has been observed and a causal relationship has been implicated [11].

Meta-analysis of prophylactic antibiotics in variceal bleeding in cirrhotics shows reduced morbidity and mortality when infection is brought under control [12]. A diurnal periodicity of variceal bleeding has also been observed. Bleeding episodes occur more frequently in the early mornings and late evenings [13], probably as a result of hyperdynamic blood flow in the portal system after meals.

Prognostic indices   

Several prognostic indices have been developed to predict which patients with esophageal varices are likely to bleed. Some of them are based on clinical parameters [14,15], others are combined with endoscopic features [8,16], and the rest use a combination of endoscopic features, biochemical parameters, and echo-Doppler ultrasound findings [17]. The most widely used index is still the North Italian Endoscopic Club index (NIEC index) [8].

This index is based on (1) severity of liver disease (Child–Pugh class), (2) size of varices, and (3) presence of red markings on the varices (Fig. 3). Base on this index, cirrhotic patients have been classified into six risk classes, each with a prediction rate of bleeding (Fig. 4). The NIEC index has been prospectively validated on independent series. Yet, with the best prognostic index, one could only predict less than 40% of variceal bleeding. Obviously, some factors predisposing to variceal bleeding remain to be discovered.

Rebleeding   

Most variceal bleeding temporarily stops by the time the patient arrives at the hospital. Without proper treatment, however, recurrent bleeding occurs in 30–40% within the next 2–3 days, and up to 60% within 1 week [18,19]. The risk of recurrent bleeding is presumably related to changes in hemodynamics of the portal system, including an increase in portocollateral resistance after hypotension, increased splanchnic blood flow stimulated by blood in the gut, and an increase in portal venous pressure as a result of overzealous volume expansion during resuscitation. Two cohort studies have showed that after the index bleeding, mortality is highest in the first 5 days and returns to baseline levels by 3–4 months [18,19]. This is the critical time window for optimal treatment to improve survival of variceal bleeders.

 Endoscopy: general   

The objectives of endoscopy in gastrointestinal bleeding are to:

• identify the source of bleeding;
  
• assess the risk of recurrent hemorrhage;
  
• attempt to control the bleeding and prevent recurrent hemorrhage.
  

It is important to identify the source of bleeding before implementing therapy. About one-third of patients with a known history of portal hypertension actually bleed from non-variceal sources such as peptic ulcers, portal hypertensive gastropathy, and Mallory–Weiss tear. In patients with decompensated liver disease, light sedation is preferred to general anesthesia.

The judicious use of endotracheal intubation is recommended during emergency endoscopy to prevent aspiration of blood. A therapeutic endoscope with a working channel of 3.2 mm would allow for adequate suction. In the presence of a large quantity of blood and clots in the stomach obscuring the view, a large 6 mm-channel endoscope could be used. In cases of blood or food material sticking to the mucosa, a water-pik system would provide a forceful water jet for washing.

Occasionally, an over-tube in the stomach will be needed to carry out proper gastric lavage. Evacuating blood from the stomach not only allows for better vision, but also helps in reducing the subsequent risk of hepatic encephalopathy in cirrhotic patients with liver dysfunction.

 Endoscopic treatments   

Several methods have been developed and used in the management of esophageal and gastric varices. The field continues to evolve, focusing on methods which are easy to use and effective, and which have minimal short- and long-term adverse effects.

Endoscopic injection sclerotherapy (EST)   

More than half a century after it was introduced by Crafoord and Frenckner [20], EST remains the primary treatment for bleeding esophagogastric varices in many parts of the world. It is used both in the control of acute bleeding and in elective obliteration of varices. Ironically, the mechanism of action of EST is not entirely clear despite its popularity. The hemostatic effect is unlikely to be due to acute venous thrombosis induced by sclerosants. It has been shown that blood flow in the varices is much faster than that in peripheral varicosities.

Furthermore, when sclerosants are injected around the varices, acute hemostasis can still be achieved in the same rate [21]. Autopsy studies indicate that venous thrombosis, mucosal ulceration and acute inflammatory reactions can be found in the injection site as early as 2 days after sclerotherapy [22]. A multitude of actions might be operating in controlling bleeding by EST.

EST technique   

Endoscopists vary in their precise methods for EST. Differences include the choice of sclerosants (type, concentration, single or in combination), number of treatment sessions, time interval between sessions, site and volume of injections, and the use of an over-tube. Injections may be directed into the veins (intravariceal injection) or into the esophageal wall adjacent to the varices (paravariceal injection), but most endoscopists favor the former technique.

In practice, however, a combination of intravariceal and paravariceal injections are given irrespective of the target chosen by endoscopists. Most endoscopists use a free-hand method injecting sclerosant to the source of bleeding or variceal column that shows a fibrin clot. After that, systematic injections to each variceal column should be given, starting from the esophagogastric junction and up until it reaches mid esophagus.

Sclerosants   

Many sclerosing agents have been used but there have been few direct comparisons between them. Sclerosants are either chemical irritants such as fatty acids (e.g. sodium morrhuate, ethanolamine oleate, polidocanol) or dehydrating agents (e.g. sodium tetradecyl sulfate, ethanol, hypertonic glucose, and phenol). Most endoscopists accept that they have similar hemostatic effects in controlling variceal hemorrhage. The choice remains largely a personal preference and depends on sclerosant availability in different countries.

Depending on the size of the varix, 1–2 ml aliquots are injected into each column at esophagogastric junction. A total volume of 20 ml can be injected at each session. Sodium tetradecyl sulfate, especially used in high concentrations, may be associated with more severe esophageal ulceration and strictures compared with polidocanol or ethanolamine oleate. Bleeding may be provoked by needle pricking into the vascular channels of variceal columns. It is advisable to advance the endoscope into the stomach using the body of the endoscope to compress on the bleeding spots. Bleeding usually stops within 1–2 min. Afterwards more injections can be given.

Accessory devices   

Many accessories have been developed for use with flexible endoscopes, e.g. flexible over-tubes (Kitano tube) and balloon catheters, to provide variceal compression during EST. Because of the additional discomfort to patients when using these devices and the high success of free-hand injection, they are seldom used nowadays.

Post-EST care   

If there is no evidence of further bleeding, feeding can be resumed after several hours of observation. Recent studies support the use of prophylactic antibiotics to prevent bacteremia and associated peritonitis after EST [12,13]. Patients may be given sucralfate or acid suppressants after receiving EST to prevent extensive esophageal ulceration. Further injection is then continued at regular intervals until all varices are obliterated. The appropriate interval for follow-up EST after control of the initial hemorrhage remains somewhat arbitrary. Most centres adopt a weekly or biweekly treatment schedule. EST should be performed according to the patient tolerance, response to therapy, and development of ulcers or other complications.

EST—proof of value?   

Although EST has been used as the standard treatment for decades, few data compare this treatment vs. sham to support its efficacy. The best evidence for the value of EST comes only recently from the Veteran Affairs Cooperative Variceal Sclerotherapy Group [23]. In this study, EST stopped hemorrhage from actively bleeding esophageal varices more effectively than sham EST with medical therapy (91% vs. 60%), and significantly improved survival (75% vs. 51%). EST was compared with medical treatment with or without balloon tamponade in four other trials. Barsoum et al. randomized 100 patients and found that EST is superior to balloon tamponade in reducing both in-hospital rebleeding and death [24].

In the Copenhagen Esophageal Varices Sclerotherapy Project, no significant difference in the initial control of hemorrhage was reported between EST and balloon tamponade [25]. Rebleeding within the first 40 days was similar in both treatment groups. After 40 days, a significant reduction in rebleeding associated with EST was noted (15% vs. 31%). The beneficial effect of EST was apparent in the long-term following of patients after eradication of varices. Paquet et al. [26] and Moreto et al. [27] recruited smaller numbers of patients and reported favorable improvement in control of bleeding with EST. In summary, EST is superior to balloon tamponade in the acute control of bleeding and prevention of recurrent hemorrhage from esophageal varices.

Endoscopic variceal ligation (EVL)   

Endoscopic variceal ligation was first introduced by Stiegmann and Goff in 1986 [28,29]. The basic principle of variceal banding involves placement of elastic O-ring ligatures, causing strangulation of the varix. The varix is sucked into the ligation chamber until it produces a red-out appearance. The trip wire is pulled to retract the inner cylinder and release the O ring. A strong suction is required and complete red-out of the endoscopic view is mandatory before firing the rubber band. Treatment should begin with ligation of most distal variceal columns in the esophagus just above the esophagogastric junction.

In the treatment of actively bleeding varices or varices with stigmata of hemorrhage, the bleeding point should be handled first. In situations where the exact bleeding point cannot be seen, it is recommended that the scope be first passed beyond the gastroesophageal junction and worked proximally towards the lower esophagus. Because only a single band can be fired at one time in the first-generation banding device, an over-tube across the cricopharyngeus sphincter is required to facilitate multiple passages of endoscope.

The use of an over-tube can lead to serious injury. Pinching of esophageal wall in the gap between the over-tube and the endoscope may lead to severe esophageal laceration and perforation. The original banding device also created a tunnel view of the intestinal lumen limiting thorough examination prior to therapy. A number of modifications have since been made.

Multi-fire devices   

A number of multi-fire band ligating devices are now available, obviating the need of an over-tube. The Speedband® (Microvasive) and the Six-shooters® (Wilson-Cook) allow applications of five and six rubber bands, respectively (Fig. 5). The suction drum is now made of transparent plastics to improve the field of vision (Fig. 6). These multi-fire banding devices have largely replaced the single-fire banding device for convenience and safety. They shorten the duration in each treatment session and reduce patient discomfort [30,31]. After a few sessions of EVL, varices become smaller and scarring of the esophageal wall starts to develop. Further banding may become difficult.

Recurrence   

The other potential problem with EVL is early recurrence of varices. Many studies have reported early recurrence of esophageal varices within 6–12 months after obliteration. EVL obliterates varices by pinching on the mucosa and submucosa of the esophageal wall. The periesophageal varices and perforating veins, which are the determinant factor for recurrence, are not occluded by EVL. Because of these problems, a combined treatment of EVL and EST has been proposed (see below).

Comparing EST and EVL   

At least 11 studies have compared EVL with EST [32–42] (Fig. 7). Most studies reported that EVL is a safer and quicker technique with fewer episodes of rebleeding and complications. The early meta-analysis by Laine and Cook demonstrated reduction in rebleeding rate (odds ratio 0.52, 95% CI 0.37–0.74), overall mortality (odds ratio 0.67, 95% CI 0.46–0.98), and mortality caused by rebleeding (odds ratio 0.49, 95% CI 0.24–0.996) after EVL as compared to EST [43]. The number of EVL sessions required to achieve variceal obliteration was also lower. The frequency of esophageal strictures was less with EVL.

Although attractive because of a faster eradication rate and fewer complications, EVL is associated with a higher variceal recurrence. The difference in variceal recurrence can be explained by the fact that the ulcers caused by banding are usually superficial and induce very limited scarring. Since the high portal pressure is not affected by endoscopic treatment and the perforating veins are not obliterated by EVL, the residual esophageal varices reappear over a period of months or even weeks. In almost all studies comparing the long-term effects of EST vs. EVL, recurrence of esophageal varices were found more frequently in the EVL-treated patients (30–48%) compared to the EST-treated patients (8–30%) (Fig. 7).

On the contrary, rebleeding from gastric varices may be more common among those treated by EST than by EVL (37% vs. 8%). The overall survival rates are similar in both treatment modalities and it is determined by hepatic reserve rather than by the method of endoscopic hemostasis. A comparison between EST and EVL is summarized in Fig. 8.

Cyanoacrylate (Histoacryl®) injection   

Endoscopic injection of the varices with N-butyl-2-cyanoacrylate (Histoacryl®) was first introduced by Soehendra in 1995 [44]. The tissue adhesive coagulates almost instantaneously when in contact with blood. In order to avoid blockage of the working channel of the endoscope and to allow fluoroscopic monitoring of the injection, cyanoacrylate is mixed with radiological contrast (lipoidol) prior to injection. Before the injection, the working channel of the endoscope and the injector probe are flushed with 1–2 ml of lipoidol to prevent channel blockage.

A mixture of 0.5 ml cyanoacrylate and 0.7 ml lipoidol is injected under fluoroscopic control. A serpentine structure demonstrated on fluoroscopy confirms successful intravariceal injection (Fig. 9). The injected submucosal vessels hardened and bleeding ceased (Fig. 10). The needle must be thoroughly rinsed with distilled water before retracting into the working channel. Suction should be prohibited at this time to avoid adhesive clogging the suction channel of the endoscope. A cast will slough off from the injection site within 1–2 weeks leaving an ulcer.

Cyanoacrylate is very effective in controlling actively spurting variceal hemorrhage. Non-randomized studies have shown that it is superior to other sclerosants in acute hemostasis [45,46]. Cyanoacrylate also offers the best results in the treatment of gastric varices [47]. In patients with hepatocellular carcinoma complicated with variceal bleeding, a single injection of cyanoacrylate has been shown to be as effective as weekly injection of sodium tetradecyl sulfate [48].

Complications of EST and EVL   

Both EST and EVL are associated with certain procedure-related complications. Almost all patients experience chest pain and transient dysphagia, and many have low-grade fever after EST. These adverse events are likely to be related to chemical and mechanical irritation producing muscle spasm and mucosal ulceration of the esophagus. They usually resolve spontaneously. More serious complications, e.g. recurrent bleeding from esophageal ulcer, perforation, and mediastinitis, develop in up to 20% of patients after EST. Aspiration pneumonia is not uncommon in agitated patients receiving EST. Bacteremia and bacterial peritonitis are not uncommon in both EST and EVL.

The only serious complication associated with EVL is related to the use of the over-tube. Severe esophageal ulcerations heal with extensive fibrosis, and stricture of esophagus occurs only with EST. Portal vein thrombosis, brain abscess, spinal cord paralysis, and pericarditis have been reported. Organ infarction and multiple septic emboli have been reported, associated with the use of cyanoacrylate injection [49,50]. Acid-suppressing agents and sucralfate have been used to minimize the effect of acid reflux and to accelerate ulcer healing. Prophylactic antibiotics have been shown to reduce bacteremia and bacterial peritonitis.

Detachable mini-snare   

The detachable mini-snare was initially designed for endoscopic resection of elevated mucosal lesion to minimize the risk of bleeding. The snare has a maximum diameter of 4 cm when fully opened. Yoshida first used the detachable snare for the ligation of gastric varices and reported excellent results [51]. Yet the technique has not gained popularity as gastric varices are usually quite big and positioning of the snare is often difficult. With modification of the design, the detachable snare can now be used for the treatment of esophageal varices [52]. This method of variceal ligation has the advantages of (1) obviating the use of an over-tube, (2) producing a clear endoscopic image, and (3) offering an unlimited number of ligations.

Technique   

A transparent ligation chamber with a rim on the inside is fitted onto the tip of an endoscope. The detachable mini-snare is inserted into the working channel of the endoscope until it opens at the rim of the ligation chamber. The endoscope is positioned to obtain an en face view of a variceal column. The varix is then sucked into the ligation chamber and the mini-snare manoeuvred to surround the varix. Ligation of the varices is performed by sliding the stopper forward (Fig. 11). When the loop is tightened around the base of the varix, suction can be released and the mini-snare detached from the snare. The ligated portion of the varix appears as a 'polyp' with a short string attached to the base (Fig. 12). Additional ligations are carried out by reloading the mini-snare with the endoscope remaining in the esophagus.

Results   

A recent prospective randomized study showed that mini-snares performed equally well when compared to Speedband®, a multiple variceal ligator [53]. However, the use of detachable snares for variceal ligation requires experienced endoscopy assistance and the procedure time is longer than for the preloaded multiple banding devices. It may not be the best option during the active phase of bleeding, especially when skilled technicians are unavailable.

 Treatment of ACUTE variceal hemorrhage   

In patients with acute variceal bleeding, the goal of treatment is to restore hemodynamic stability, stop hemorrhage, prevent complications, including early rebleeding, and decrease mortality. Over the last few decades, pharmacological therapy, endoscopic treatment, surgical devascularization or bypass, and radiological shunting have been developed for the optimal treatment of acute variceal bleeding. A combination of these therapies, in conjunction with supportive therapy and correction of coagulopathy, probably achieves the best results in the management of this deadly complication of portal hypertension.

Pharmacological treatments   

Vasopressin was the first vasoactive agent used in the control of acute variceal bleeding. While most studies confirmed that vasopressin is effective in the control of acute bleeding, its use has been limited by the frequent side-effects due to non-selective vasoconstriction. Unlike vasopressin, somatostatin and its analogs have selective vasoconstrictive activity in the splanchnic circulation, thus reducing the risk of systemic side-effects. Although the mechanism of action is not entirely clear, somatostatin and its analogs have been shown to be able to reduce portal pressure and blood flow in collaterals [54,55], leading to control of variceal hemorrhage.

Four trials compared somatostatin [56–59] and three compared octreotide against endoscopic sclerotherapy in the treatment of acute bleeding esophageal varices [60–62]. All studies showed similar rates of initial control of bleeding, recurrent bleeding, and death with pharmacological or endoscopic control of acute variceal bleeding. In a recent meta-analysis combining data collected from 13 randomized trials, octreotide has been shown to improve control of esophageal variceal hemorrhage compared to alternative therapies. Its effect is comparable to immediate EST but with fewer major complications than vasopressin/terlipressin [63]. Vasoactive agents provide an effective stop-gap treatment when endoscopic therapy is not readily available.

Combined endoscopic and pharmacological therapy   

Since both endoscopic therapy and vasoactive drugs are effective in the acute control or prevention of recurrent bleeding in esophageal varices, it would be logical to combine the two treatment modalities. The optimal approach is to administer vasoactive drugs at the earliest possible time for patients suspected to have variceal bleeding. Levacher et al. used terlipressin plus nitroglycerin started before hospitalization [64]. The medication was given by the emergency medical team at patients' homes before they were brought to hospital. Control of variceal bleeding was significantly improved. When somatostatin and vapreotide were given prior to EST, variceal bleeding was controlled before endoscopic therapy [65,66].

Endoscopic therapy becomes easier as the endoscopist is working in a 'dry field'. Recurrent bleeding after endoscopic therapy can therefore be reduced. The lowering of portal venous pressure by early administration of vasoactive drugs thus holds the key to success of subsequent endoscopic control of variceal bleeding. In the study using terlipressin [64], mortality due to bleeding in the first 2 weeks was reduced in the group who received terlipressin prior to endoscopic treatment. Similarly, a vasoactive agent seems to produce an adjunctive effect in combination with EVL in the treatment of acute variceal bleeding [67]. The combined endoscopic and pharmacological therapy leads to less rebleeding, thus requiring less balloon tamponade and retreatment with endoscopy. However, no improvement on mortality was demonstrated, which appeared to be related primarily to severity of hepatic decompensation.

Surgery   

Occasionally, hemostasis fails despite the combination of endoscopic and pharmacological therapy. Shunting of the hypertensive portal venous system to the normotensive systemic venous circulation would be inevitable. Surgical treatments for variceal bleeding include direct esophageal devascularization of the lower esophagus and the proximal stomach as well as a variety of surgical shunting procedures. Simple surgical devascularization with esophageal transaction is an effective means of controlling acute bleeding but long-term results are poor. Shunt operations, including selective and non-selective shunts that divert all or part of portal blood flow into the inferior vena cava, offer an effective control of bleeding but also lead to severe hepatic encephalopathy.

Tipss   

In the last decade, transjugular intrahepatic portosystemic stent shunt (TIPSS) has been developed as a radiological portocaval shunt. It has the advantage of avoiding a major operation and the risk of general anesthesia in patients with poor hepatic reserve and hemodynamic instability. Despite high success in controlling acute hemorrhage, TIPSS has failed to demonstrate improvement in mortality. One of the major problems with TIPSS is the development of hepatic encephalopathy, which should be expected in any shunting procedure.

A number of prognostic indicators after TIPSS have been under investigation but none has high success in predicting outcome. This includes a poor APACHE score, presence of hyponatremia and Child C liver disease, hepatic encephalopathy before TIPSS, presence of ascites, and serum albumin. TIPSS should not be used as a routine form of therapy for variceal bleeding, yet it is highly efficacious as a rescue therapy should endoscopic therapy fail (Fig. 13). TIPSS should be reserved for the subset of patients who continue to bleed or develop recurrent bleeding after endoscopic therapy.

Comparing TIPSS with endoscopic treatments   

TIPSS has been compared with endoscopic therapy as therapy for variceal bleeding in 11 studies involving 811 patients. Endoscopic treatments in these trials were: EST in 5 studies [68–72], EVL in 3 studies [73–75], and EST or EVL plus propranolol in 3 studies [76–78]. In a meta-analysis pooling data from the above studies, variceal rebleeding was significantly more frequent with endoscopic therapies (47%) compared with TIPSS (19%) (OR, 3.8; 95% CI, 2.8–5.2) [79).

Consensus approach to acute bleeding   

Based on existing evidence, a clinical algorithm is suggested below (Fig 14). During initial resuscitation with fluid and blood replacement, cirrhotic patients with upper gastrointestinal bleeding should be given intravenous injection/infusion of vasoactive agents. Emergency endoscopy should be arranged if the patient remains hemodynamically unstable. After endoscopic confirmation of the source of bleeding, EVL and/or injection EST should be offered immediately. Vasoactive agents should be continued for 5 days to prevent early rebleeding and prophylactic antibiotics are also recommended. When initial hemostasis is achieved, the patient can be scheduled for an endoscopic obliteration program with weekly or biweekly banding or sclerotherapy until all varices are obliterated.

Patients who fail to respond to endoscopic therapy or suffer from recurrent bleeding during the acute phase should be offered balloon tamponade before a second session of therapeutic endoscopy. Vasoactive therapy should be continued. If secure hemostasis still cannot be achieved, TIPSS or surgical shunting (when TIPSS is not available or not feasible) offers the best salvage therapy. Patients undergoing TIPSS or shunt surgery should be closely monitored for shunt patency and mental status. Liver transplantation would be the last resort, when all treatment options have failed.

Combined endoscopic therapies vs. single therapy   

Combination of EVL and EST has been proposed to produce synergistic effects in initial hemostasis and long-term therapy for esophageal varices. The two endoscopic therapies work on different mechanisms and may produce additive effects. However, the literature has produced conflicting results as to the therapeutic effect of combined endoscopic therapy. Some studies have combined EVL and EST during the same session (synchronous treatment). Others use the two therapies sequentially (metachronous treatment).

Synchronous combinations   

Among those using a synchronous approach, some injected adjacent to a ligated varix, while others injected into the adjacent mucosa or ligate on the injected varix. While achieving a marginally higher variceal eradication rate, Laine et al. reported that synchronous treatment is associated with a higher complication rate and required a higher number of sessions with the combination therapy compared with EVL alone [80]. The rebleeding rate and mortality were similar to monotherapy. Saeed et al. reported that combined treatment had not achieved any superiority to EVL alone as the former required more treatment sessions, associated with higher complication rates and higher mortality [81]. Umehara et al. also noticed that the combined approach had resulted in more complications but a significantly lower rate of recurrence of varices [82].

Hou described the so-called 'sandwich' technique using a pneumoactive ligation device and demonstrated a marginal benefit of combined treatment in reducing variceal recurrence [83]. When EST is compared against combination therapy, combination therapy requires fewer treatment sessions and produces fewer complications than EST alone [84,85]. Thus patients undergoing synchronous endoscopic therapy may fare worse than those receiving EVL alone but probably do better those receiving EST alone.

Metachronous combinations   

On the other hand, there might be an advantage to perform metachronous endoscopic therapy. Bhargava et al. reported higher eradication and lower rebleeding with combination therapy [86]. Lo et al. reported similar treatment sessions and eradication rate for both treatments. However, mortality, rebleeding, and variceal recurrence were lower with combination therapy than with EVL alone [87]. Cheng confirmed that recurrence of varices is more common in patients receiving single endoscopic therapy [88]. Garg et al. compared EST with combination therapy. Again eradication rate and number of treatment sessions were similar, but complications and rebleeding rate were higher with EST alone [89].

Masumoto et al. conducted a study comparing EVL, EST, and EVL followed by EST. The number of patients in each study arm was small. Their study indicated that combined endoscopic therapy is superior to either EST or EVL alone, requiring fewer treatment sessions and causing fewer complications. Recurrence of varices was highest among those who received EVL alone [90]. Based on existing data from the literature, the initial hemostasis of EVL is comparable to combined therapy. However, combining EST with EVL may reduce risk of recurrence varices and rebleeding at a cost of a higher complication rate.

 Prophylactic treatment of esophageal varices   

Detection and surveillance   

All patients with confirmed diagnosis of portal hypertension, with or without cirrhosis, should receive endoscopic screening for varices. Based on studies of natural history of esophageal varices, patients with no varices need to have follow-up endoscopy every 3–4 years. Patients with esophageal varices should be assessed by a predictive scoring system, e.g. NIEC score. Low-risk patients should be re-examined yearly for progression of esophageal varices. Moderate- to high-risk patients should be offered prophylactic therapy. Although beta-blocker treatment remains the mainstay for prophylaxis against variceal bleeding, only one-third of patients will respond to this medication and it is not tolerated by all.

Endoscopic prophylaxis   

There has long been controversy over the role of endoscopic treatment in primary prophylaxis against first bleed in selected high-risk patients. In the 1980s, prophylactic EST was tested as a preventative for variceal bleeding. Three initial controlled trials reported that prophylactic EST significantly reduces the risk of first variceal bleeding and improves survival [91–93]. However, in subsequent trials, prophylactic EST has not shown survival benefit. EST may provoke bleeding. Indeed, in two American studies, the mortality was higher in the EST group than the control group [94,95]. While EST may produce more harm than good to patients with cirrhosis, EVL has been shown to be safer and associated with fewer procedure-related complications.

Three trials [96–98] compared endoscopic EVL against no treatment and two compared it against propranolol therapy in the prevention of variceal bleeding [99,100] (Fig. 14). The results of these studies showed a significant decrease in the risk of bleeding and the mortality was also significantly reduced in two studies. In a recently published meta-analysis based on four trials recruiting 283 patients, it has been estimated that, compared to no treatment, EVL reduces the risk of variceal bleeding by 64% and mortality by 45% [101].

When compared to beta-blocker, EVL reduces the risk of variceal bleeding by 52% but fails to reduce mortality. With these promising results, prophylactic EVL could be considered for selected patients with high risk of variceal bleeding who cannot tolerate pharmacological prophylaxis. A clinical algorithm in the management of varices without bleeding is shown in Fig. 16.

 Gastric varices   

Compared to esophageal varices, the natural history of, and optimal treatment for, gastric varices are less clear. Gastric varices tend to bleed less frequently than esophageal varices but bleeding tends to be more severe. Mortality in those who bled from gastric varices is substantial.

According to Sarin, gastric varices can be categorized into two groups according to their locations in the stomach and relationship with esophageal varices: (1) gastroesophageal varices (GOVs), found associated with esophageal varices, and (2) isolated gastric varices (IGV), found independently of esophageal varices [102]. Each of the subtypes can be further classified into Type 1 and 2 based on their locations in the stomach (Fig. 16).

Type 1 GOVs are continuous with esophageal varices, extending 2–5 cm beyond the gastroesophageal junction along the lesser curvature of the stomach, while Type 2 GOVs extend into the fundus of stomach. Type 1 IGV refer to isolated fundal gastric varices and Type 2 refers to isolated 'ectopic' varices occurring in the distal portion of the stomach, including the duodenum. This classification of gastric varices has prognostic implications and may be helpful in management planning. Type 1 GOVs are the commonest type, accounting for 75% of cases. Their rate of bleeding is only 11.8%. Their treatment is identical to that of esophageal varices and they are usually responsive to endoscopic therapy.

Type 2 GOVs tend to bleed more frequently (55%). IGV1 has the highest rate of bleeding at 78% and carries substantial mortality. EST has been described in the treatment of bleeding gastric varices with an overall initial control of 80% [103,104]. In addition, rebleeding after endoscopic sclerotherapy for bleeding gastric varices is common, especially in IGV1. EVL is less effective in bleeding gastric varices. Deployment of bands can be difficult with the endoscope in a retroflexed position. Band retention is a problem and a slipped band may precipitate early rebleeding. Histoacryl injection is the most widely accepted therapy for gastric varices.

In a non-randomized series, Oho et al. compared injections using cyanoacrylate to ethanolamine in 53 patients [35]. Their study showed that the rate of initial control of bleeding is significantly higher in patients treated with cyanoacrylate. In-hospital mortality was significantly higher in the ethanolamine-treated group. In a recent study comparing endoscopic banding ligation with cyanoacrylate injection, 60 patients with gastric varices were randomized to receive the two treatments [105].

Compared to EVL, cyanoacrylate injection is found to have better initial hemostasis and less recurrent bleeding. Blood transfusion is less and mortality is improved with cyanoacrylate injection of gastric varices. In general, the lower rate of hemostasis and higher rate of recurrent bleeding in gastric varices with endoscopic treatment would imply that alternate therapies such as surgery or TIPSS are more often needed.

 Endoscopic ultrasonography in variceal hemorrhage   

Endoscopic ultrasound (EUS) has been applied to patients with variceal hemorrhage since 1990 [106]. Using EUS, a large part of the portal venous system can be visualized, including the portal vein, splenic vein, superior mesenteric vein, and the azygos vein. However, when used in detecting esophageal varices, EUS was found to be inferior to conventional upper endoscopy [107,108]. This was mainly due to the large-bore echoendoscope of the first generation, and the use of water-filled balloons, which compress variceal columns.

With the improvement in technology, and the development of catheter probes using higher ultrasound frequency [109,110], it is also found that EUS can detect vascular channels both inside and outside the esophageal lumen. EUS can also be used to map vessels in the adjacent tissues outside the esophageal wall (periesophageal varices), and collateral veins in the mediastinum running longitudinally alone with the esophagus (paraesophageal varice), as well as the perforators that connect the submucosal and peripheral vascular channels [111,112] (Fig. 17).

Paraesophageal varices and perforating veins are reported in around 80% of patients with esophageal varices, and their presence has been associated with poor response to endoscopic therapy and early recurrence of varices [113,114].

It has also been suggested that patients undergoing EVL are more likely to have multiple paraesophageal varices of considerable size compared to those who received EST [115,116]. EUS-guided sclerotherapy using color Doppler has been attempted to improve results of EST [117]. It has been suggested that after endoscopic therapy has achieved obliteration of submucosal varices, EUS examination of the lower esophagus is warranted to identify those who had perforating veins and peri-/paraesophageal varices. Small volume injections of sclerosants may be given to seal these perforators to prevent early recurrence of varices and rebleeding, followed by close monitoring of these patients. The development of color Doppler EUS has also allowed the study of blood flow hemodynamics in the portal venous system. Cirrhotic patients are found to have an enlarged azygos vein and thoracic ducts [118].

Azygos vein blood flow has been found to be reduced after endoscopic therapy [119]. The use of azygos blood flow measurement in the clinical management of variceal hemorrhage remains to be defined.

EUS and gastric varices   

Unlike esophageal varices, EUS has been found to be very useful in detecting gastric varices as it can easily distinguish vascular structure from other submucosal lesions in the gastric fundus (Fig. 17) [107,108]. EUS is also a useful tool in monitoring the effects of EST as it confirms obliteration of vascular channel in the submucosa of the stomach. It has been reported that cyanoacrylate injection monitored by EUS reduces recurrence of gastric varices [120].

 Outstanding issues and future trends   

It is still impossible to predict rebleeding after endoscopic therapy. The endoscopic appearance of the varices and clinical parameters have not been able to accurately identify high-risk patients. Future research should be directed to non-invasive measurement of blood flow and pressure of varices using EUS and Doppler ultrasound to monitor therapy of esophageal and gastric varices. An endoscope which could serve both diagnostic (including detection of blood flow) and therapeutic purposes is very much needed.

 References  

1 Lebrec, D, DeFleury, P & Rueff, B et al. Portal hypertension, size of esophageal varices, and risk of gastrointestinal bleeding in alcoholic cirrhosis. Gastroenterology 1980; 79: 1139. PubMed

2 GarciaTsao, G, Groszmann, RJ & Fisher, RL et al. Portal pressure, presence of gastroesophageal varices and variceal bleeding. Hepatology 1985; 5: 419. PubMed

3 Christensen, E, Fauerholdt, L & Schlichting, P et al. Aspects of natural history of gastrointestinal bleeding in cirrhosis and the effect of prednison. Gastroenterology 1981; 81: 944. PubMed

4 Cales, P, Desmorat, H & Vinet, JP et al. Incidence of large esophageal varices in patients with cirrhosis: application to prophylaxis of first bleeding. Gut 1990; 31: 1298. PubMed

5 D'Amico, G, Pasta, LV & Izzini, GB et al. Natural history of cirrhosis. J Hepatol 1990; 13: S22.

6 D'Amico, G, Pasta, L & Madonia, S et al. The incidence of esophageal varices in cirrhosis. Gastroenterology 2001; 120: A2.

7 D'Amico, G, Pagliaro, L & Bosch, J. The treatment of portal hypertension: a meta-analytic review. Hepatology 1995; 22: 332. PubMed

8 North Italian Endoscopic Club for the Study and Treatment of Esophageal Varices. Prediction of the first variceal hemorrhage in patients with cirrhosis of the liver and esophageal varices. N Engl J Med 1998; 319: 983.

9 Boyer, TD. Portal hypertensive hemorrhage: pathogenesis and risk factors. Semin Gastrointest Dis 1995; 6: 125. PubMed

10 Paquet, KJ. Prophylactic endoscopic sclerosing treatment of the esophageal wall in varices: a prospective controlled trial. Endoscopy 1982; 14: 4. PubMed

11 Goulis, J, Armonis, A, Patch, D, Sabin, C, Greenslade, L & Burroughs, AK. Bacterial infection is independently associated with failure to control bleeding in cirrhotic patients with gastrointestinal hemorrhage. Hepatology 1998; 27: 1207. PubMed CrossRef

12 Bernard, B, Nguyen, KE, Nguyen, KE, Ololon, P & Poynard, T. Antibiotic prophylaxis for the prevention of bacterial infection in cirrhotic patients with gastrointestinal bleeding: a meta-analysis. Hepatology 1999; 29: 1655. PubMed

13 McCormick, PA, Jenkins, SA & McIntyre, N et al. Why portal hypertensive varices bleed and bleed: a hypothesis. Gut 1995; 36: 100. PubMed

14 Snady, H & Feinman, L. Prediction of variceal hemorrhage: a prospective study. Am J Gastroenterol 1988; 83: 519. PubMed

15 Zoli, M, Merkel, C & Magalotti, D et al. Evaluation of a new endoscopic index to predict first bleeding from the upper gastrointestinal tract in patients with cirrhosis. Hepatology 1996; 24: 1047. PubMed

16 Kleber, G, Sauerbruch, T & Ansari, H et al. Prediction of variceal hemorrhage in cirrhosis: a prospective follow-up study. Gastroenterology 1991; 100: 1331.

17 Siringo, S, Bolondi, L & Gaiani, S et al. Timing of the first variceal hemorrhage in cirrhotic patients: prospective evaluation of Doppler flowmetry, endoscopy and clinical parameters. Hepatology 1994; 20: 66. PubMed CrossRef

18 Graham, DY & Smith, JL. The course of patients after variceal hemorrhage. Gastroenterology 1981; 80: 800. PubMed

19 Burroughs, AK, Mezzanotte, G, Phillips, A, McCormick, PA & McIntyre, N. Cirrhotics with variceal hemorrhage: the importance of the time interval between admission and the start of analysis for survival and rebleeding rates. Hepatology 1989; 9: 801. PubMed

20 Crafoord, C & Frenckner, P. New surgical treatment of varicous veins of the esophagus. Acta Otolaryngol 1939; 27: 422.

21 Paquet, K. Endoscopic paravariceal injection sclerotherapy of the esophagus—indications, techniques, complications: results of a period of 14 years. Gastrointest Endosc 1983; 29: 310. PubMed

22 Ayres, S, Goff, J & Warren, G et al. Esophageal ulceration and bleeding after flexible fiberoptic esophageal vein sclerosis. Gastroenterology 1982; 83: 131. PubMed

23 Hartigan, PM, Gehbard, RL & Gregorty, PB et al. Sclerotherapy for actively bleeding esophageal varices in male alcoholics with cirrhosis. Veterans Affairs Cooperative Variceal Sclerotherapy Group. Gastrointest Endosc 1997; 46: 1. PubMed

24 Basorum, M. Tamponade and injection sclerotherapy in the management of bleeding esophageal varices. Br J Surg 1982; 69: 76. PubMed

25 The Copenhagen Esophageal Varices Sclerotherapy Project. Sclerotherapy after first variceal hemorrhage in cirrhosis: a randomized multicenter trial. N Engl J Med 1984; 311: 1594–600. PubMed

26 Paquet, KJ & Feussner, H. Endoscopic sclerotherapy and esophageal balloon tamponade in acute hemorrhage from esophago-gastric varices: a prospective controlled randomized trial. Hepatology 1985; 5: 580. PubMed

27 Moreto, M. & Bernal, A et al. A randomized trial of tamponade or sclerotherapy as immediate treatment for bleeding esophageal varices. Surg Gynecol Obstet 1988; 167: 331. PubMed

28 Stiegmann, GV, Cambre, T & Sun, JH. A new endoscopic elastic band ligating device. Gastrointest Endosc 1986; 32: 230–3. PubMed

29 Stiegmann, GV & Goff, JS. Endoscopic esophageal varix ligation: preliminary clinical experience. Gastrointest Endosc 1988; 34: 113–18. PubMed

30 Saeed, ZA, Khan, AA, Alam, A, Shafqat, F, Butt, A & Hassan, W. A therapeutic shoot-out for esophageal varices: head to head multi-randomized comparison of 'single-shot' ligation versus the Saeed six-shooter. Gastrointest Endosc 1996; 43: A216.

31 Hoepffner, N, Domschke, W & Foerster, EC. Multiple band ligator for esophageal varices: first clinical experience. Gastrointest Endosc 1996; 43: A187.

32 Stiegmann, GV, Goff, JS & Michaletz-Onody, PA et al. Endoscopic sclerotherapy as compared with endoscopic ligation for bleeding esophageal varices. N Engl J Med 1992; 326: 1527–32. PubMed

33 Gimson, AE, Ramage, JK & Panos, MZ et al. Randomized trial of variceal banding ligation versus injection sclerotherapy for bleeding esophageal varices. Lancet 1993; 342: 391–4. PubMed CrossRef

34 Laine, L, El-Newihi, HM & Migikovsky, B et al. Endoscopic ligation compared with sclerotherapy for the treatment of bleeding esophageal varices. Ann Intern Med 1993; 119: 1–7. PubMed

35 Lo, GH, Lai, KH & Cheng, JS et al. A prospective randomized trial of sclerotherapy versus ligation in the management of bleeding esophageal varices. Hepatology 1995; 22: 466–71. PubMed CrossRef

36 Hou, MC, Lin, HC & Kuo, BI et al. Comparison of endoscopic variceal injection sclerotherapy and ligation for the treatment of esophageal variceal hemorrhage: a prospective randomized trial. Hepatology 1995; 21: 151–22.

37 Lo, GH, Lai, KH & Cheng, JS et al. Emergency banding ligation versus sclerotherapy for the control of active bleeding from esophageal varices. Hepatology 1997; 25: 1101–4. PubMed CrossRef

38 Avgerinos, A, Armonis, A & Manolakopoulos, S et al. Endoscopic sclerotherapy versus variceal ligation in the long-term management of patients with cirrhosis after variceal bleeding: a prospective randomized study. J Hepatol 1997; 26: 1034–41. PubMed CrossRef

39 Masci, E, Stigliano, R & Mariani, A et al. Prospective multicenter randomised trial comparing banding ligation with sclerotherapy of esophageal varices. Hepatogastroenterology 1999; 46 (27): 1769–73. PubMed

40 Hou, MC, Lin, HC & Kuo, BIT et al. The rebleeding course and long-term outcome of esophageal variceal hemorrhage after ligation: comparison with sclerotherapy. Scan J Gastroenterol 1999; 34: 1071–6. CrossRef

41 Sarin, SK, Govil, A & Jain, AK et al. Prospective randomized trial of endoscopic sclerotherapy versus variceal band ligation for esophageal varices: influence on gastropathy, gastric varices and variceal recurrence. J Hepatol 1997; 26: 826–32. PubMed CrossRef

42 Baroncini, D, Milandri, GL & Borione, D et al. A prospective randomized trial of sclerotherapy versus ligation in the treatment of bleeding esophageal varices. Endoscopy 1997; 29: 235–40. PubMed

43 Laine, L & Cook, D. Endoscopic ligation compared with sclerotherapy for treatment of esophageal variceal bleeding: a meta-analysis. Ann Intern Med 1995; 123: 280–7. PubMed

44 Binmoeller, KF & Soehendra, N. 'Superglue': the answer to variceal bleeding and fundal varices? Endoscopy 1995; 27: 392. PubMed

45 Oho, K, Iwao, T, Sumino, M, Toyonaga, A & Tanikawa, K. Ethanolamine oleate versus butyl cyanoacrylate for bleeding gastric varices: a nonrandomized study. Endoscopy 1995; 27: 349–54. PubMed

46 Feretis, C, Tabakopoulos, D, Benakis, P, Xenofontos, M & Golematis, B. Endoscopic hemostasis of esophageal and gastric variceal bleeding with Histoacryl. Endoscopy 1990; 22: 282–4. PubMed

47 Iwase, H, Maeda, O, Shimada, M, Tsuzuki, T, Peek, RM Jr & Nishio, Y, et al. Endoscopic ablation with cyanoacrylate glue for isolated gastric variceal bleeding. Gastrointest Endosc 2001; 53: 585–92. PubMed CrossRef

48 Sung, JJY, Yeo, W & Suen, R et al. Cyanoacrylate vs sodium tetradecyl sulfate for the injection of bleeding varices in patients with hepatocellular carcinoma. Gastrointest Endosc 1998; 47: 235–9. PubMed

49 Roesch, W & Rexroth, G. Pulmonary, cerebral and coronary emboli during bucrylate injection of bleeding fundic varices. Endoscopy 1998; 30: S89–90. PubMed

50 Turler, A, Wolff, M, Dorlars, D & Hirner, A. Embolic and septic complications after sclerotherapy of fundic varices with cyanoacrylate. Gastrointest Endosc 2001; 53: 228–30. PubMed CrossRef

51 Yoshida, T, Hayashi, N, Suzumi, S, Miyazaki, S, Terai, S & Itoh, T et al. Endoscopic ligation of gastric varices using a detachable snare. Endoscopy 1994; 26: 502–5. PubMed

52 Sung, JJY & Chung, SCS. The use of a detachable mini-loop for the treatment of esophageal varices. Gastrointest Endosc 1998; 47: 178–81. PubMed

53 Shim, CS, Cho, JY & Park, YJ et al. Mini-detachable snare ligation for the treatment of esophageal varices. Gastrointest Endosc 1999; 50: 673–6. PubMed

54 Sonneburg, GE, Keller, U & Perruchoud, A et al. Effect of somatostatin on splanchnic hemodynamics in patients with cirrhosis of the liver and in normal subjects. Gastroenterology 1981; 80: 526–32. PubMed

55 Bosch, J, Kravetz, D & Rodes, J. Effects of somatostatin on hepatic and systemic hemodynamics in patients with cirrhosis of the liver: comparison with vasopressin. Gastroenterology 1981; 80: 518–25. PubMed

56 Shields, R, Jenkins, SA & Baxter, JN et al. A prospective randomized controlled trial comparing the efficacy of somatostatin with injection sclerotherapy in the control of bleeding esophageal varices. J Hepatol 1992; 16: 128–37. PubMed

57 Planas, R, Quer, JC & Boix, J et al. A prospective randomized trial comparing somatostatin and sclerotherapy in the treatment of acute variceal bleeding. Hepatology 1994; 20: 370–5. PubMed CrossRef

58 Di Febo, G, Siringo, S & Vacirca, M et al. Somatostatin and urgent sclerotherapy in active oesophageal variceal bleeding. Gastroenterology 1990; 98: 583A.

59 Escorsell, A, Bordas, JM & Del Arbol, LR et al. Randomized controlled trial of sclerotherapy versus somatostatin in the prevention of early rebleeding following acute variceal hemorrhage in patients with cirrhosis. Variceal Bleeding Study Group. J Hepatol 1998; 29: 779–88. PubMed CrossRef

60 Sung, JJY, Chung, SCS & Lai, CW et al. Octreotide infusion or emergency sclerotherapy for variceal hemorrhage. Lancet 1993; 342: 637–41. PubMed CrossRef

61 Jenkins, SA, Shield, RA & Davies, M et al. A multicenter randomized trial comparing octreotide and injection sclerotherapy in the management and outcome of acute variceal hemorrhage. Gut 1997; 41: 526–33. PubMed

62 Bildozola, M, Kravetz, D & Argonz, J et al. Efficacy of octreotide and sclerotherapy in the treatment of acute variceal bleeding in cirrhotic patients: a prospective, multicenter and randomized clinical trial. Scand J Gastroenterology 2000; 35: 419–25.

63 Corely, DA, Cello, JP, Adkisson, W, Ko, WF & Kerlikowske, K. Octreotide for acute esophageal variceal bleeding: a meta-analysis. Gastroenterology 2001; 120: 946–54. PubMed

64 Levacher, S, Letoumelin, P & Pateron, D et al. Early administration of terlipressin and glyceryl trinitrate to control acute upper gastrointestinal bleeding in cirrhotic patients. Lancet 1995; 346: 865–8. PubMed CrossRef

65 Averginos, A, Nevens, F & Raptis, S et al. Early administration of somatostatin and efficacy of sclerotherapy in acute esophageal variceal bleeds: the European Acute Bleeding Oesophageal Variceal Episodes (ABOVE) randomised trial. Lancet 1997; 350: 1495–9. PubMed CrossRef

66 Cales, P, Masliah, C & Bernard, B et al. Early administration of vapreotide in variceal bleeding in patients with cirrhosis. N Engl J Med 2001; 344: 23–8. PubMed

67 Sung, JJY, Chung, SCS & Yung, MY et al. Prospective randomised study of effect of octreotide on rebleeding from oesophageal varices after endoscopic banding ligation. Lancet 1995; 346: 1666–9. PubMed CrossRef

68 Cabrera, J, Maynar, M & Branados, R et al. Transjugular intrahepatic portosystemic shunt versus sclerotherapy in the elective treatment of variceal hemorrhage. Gastroenterology 1996; 110: 832. PubMed

69 Sanyal, AJ, Freedman, AM & Luketic, VA et al. Transjugular intrahepatic portosystemic shunt compared with sclerotherapy for the prevention of recurrent variceal hemorrhage: a randomized controlled trial. Ann Intern Med 1997; 126: 849. PubMed

70 Cello, JP, Ring, EJ & Olcott, E et al. Endoscopic sclerotherapy compared with percutanous transjugular intrahepatic portosystemic shunt after initial sclerotherapy in patients with acute variceal hemorrhage: a randomized controlled trial. Ann Intern Med 1997; 126: 858. PubMed

71 Merli, M, Riggio, O & Capocaccia, L et al. TIPS vs sclerotherapy for prevention of variceal rebleeding: results of a randomized controlled trial. Gastroenterology 1998; 110: A1265.

72 Carcia-Villareal, L, Martinez-Lagares, F & Sierra, A et al. TIPS vs sclerotherapy for the prevention of variceal rebleeding: preliminary results of a randomized study. Hepatology 1999; 24: 208A.

73 Jalan, R, Forrest, EH & Stanley, AJ et al. A randomized trail comparing transjugular intrahepatic portosystemic stent-shunt with variceal band ligation in the prevention of rebleeding from esophageal varices. Hepatology 1997; 26: 1115–22. PubMed

74 Pomier-Layrargues, G, Dufresne, MP & Bul, B et al. TIPS versus endoscopic variceal banding in the prevention of variceal rebleeding in cirrhotic patients: a comparative randomized clinical trial (interim analysis). Hepatology 1997; 26 (Suppl.): 137A. CrossRef

75 Sauer, P, Benz, C & Theilmann, L et al. Transjugular intrahepatic portosystemic shunt (TIPS) vs endoscopic banding in the prevention of variceal rebleeding: final results of a randomized study. Gastroenterology 1998; 114: A1334.

76 Groupe d'Etude des Anastomoses Intra-Hepatiques (Toulouse BLNCPE). TIPS vs sclerotherapy + propranolol in the prevention of variceal rebleeding: preliminary results of a multicenter randomized trial. Hepatology 1995; 22 (Suppl.): 299A.

77 Sauer, P, Theilmann, L & Benz, C et al. Transjugular intrahepatic portosystemic stent shunt vs sclerotherapy in the prevention of variceal rebleeding: a randomized study. Gastroenterology 1996; 110: A1313.

78 Rossle, M, Deibert, P & Haage, K et al. Randomized trial of transjugular intrahepatic portosystemic shunt versus endoscopy plus propranolol for prevention of variceal rebleeding. Lancet 1997; 349: 1043. PubMed

79 Papatheodoridis, GV, Goulis, J & Leandro, G et al. Transjugular intrahepatic portosystemic shunt compared with endoscopic treatment for prevention of variceal rebleeding: a meta-analysis. Hepatology 1999; 30: 612–22. PubMed

80 Laine, L, Stein, C & Sharma, V. Randomized comparison of ligation versus ligation plus sclerotherapy in patients with bleeding esophageal varices. Gastroenterology 1996; 110: 529–33. PubMed

81 Saeed, ZA, Stiegmann, GV & Ramirez, FC et al. Endoscopic variceal ligation is superior to combined ligation and sclerotherapy for esophageal varices. Hepatology 1997; 25: 71–4. PubMed

82 Umehara, M, Onda, M & Tajiri, T et al. Sclerotherapy plus ligation versus ligation for the treatment of esophageal varices: a prospective randomized study. Gastrointest Endosc 1999; 50: 7–12. PubMed

83 Hou, MC, Chen, WC & Lin, HC et al. A new 'sandwich' method of combined endoscopic variceal ligation and sclerotherapy alone in the treatment of esophageal variceal bleeding: a randomized trial. Gastrointest Endosc 2001; 53: 572–8. PubMed CrossRef

84 Iso, Y, Kawanaka, H & Tomikawa, M et al. Repeated injection sclerotherapy is preferable to combined therapy with variceal ligation to avoid recurrence of esophageal varices: a prospective randomised trial. Hepatogastroenterology 1997; 44: 467–71. PubMed

85 Nishikawa, Y, Hosokawa, I & Doi, T et al. Evaluation of endoscopic injection sclerotherapy with and without ligation for the treatment of esophageal varices. J Gastroenterol 1999; 34: 158–62.

86 Bhargava, DK & Pokharna, R. Endoscopic variceal ligation versus endoscopic variceal ligation and endoscopic sclerotherapy: a prospective randomized study. Am J Gastroenterol 1997; 92: 950–3. PubMed

87 Lo, GH, Lai, KH & Cheng, JS. The additive effect of sclerotherapy to patients receiving repeated endoscopic variceal ligation: a prospective randomized trial. Hepatology 1998; 50: 369–73.

88 Cheng, YS, Shan, P & Lien, GS et al. Adjuvant sclerotherapy after ligation for the treatment of varices: a prospective randomised long term study. Gastrointest Endosc 2001; 53: 566–71. PubMed

89 Garg, PK, Joshi, YK & Tandon, RK. Comparison of endoscopic variceal ligation sclerotherapy with sequential endoscopic band ligation plus low-dose sclerotherapy for secondary prophylaxis of variceal hemorrhage: a prospective randomized study. Gastrointest Endosc 1999; 50: 369–73. PubMed

90 Masumoto, H, Toyonaga, A & Oho, K et al. Ligation plus low-volume sclerotherapy for high-risk esophageal varices: comparison with ligation therapy or sclerotherapy alone. J Gastroenterol 1998; 33: 1–5. PubMed

91 Paquet, KJ. Prophylactic endoscopic sclerosing treatment of the esophageal wall in varices: a prospective controlled randomised trial. Endoscopy 1982; 14: 4–5. PubMed

92 Witzel, L, Wolbergs, E & Merki, H. Prophylactic endoscopic sclerotherapy of esophageal varices: a prospective controlled study. Lancet 1985; 1: 773–5. PubMed CrossRef

93 Piai, G, Cipolletta, L & Claar, M. Prophylactic endoscopic sclerotherapy of high risk esophageal varices: results of a multicenter controlled trial. Hepatology 1988; 8: 1495–500. PubMed

94 The Veterans Affairs Cooperative Variceal Sclerotherapy Group. Prophylactic sclerotherapy for esophageal varices in men with alcoholic liver disease: a randomised single-blind, multicenter clinical trial. N Engl J Med 1991; 324: 1779. PubMed

95 Santangelo, WC, Dueno, MI, Estes, BL & Krejs, GJ. Prophylactic sclerotherapy of large esophageal varices. N Engl J Med 1988; 13: 149.

96 Sarin, SK, Guptan, RK, Jain, AK & Sundaram, KR. A randomized controlled trial of endoscopic variceal band ligation for primary prophylaxis of variceal bleeding. Eur J Gastroenterol Hepatol 1996; 8: 337–42. PubMed

97 Lay, CS, Tsai, YT & Teg, CY et al. Endoscopic variceal ligation in prophylaxis of first variceal bleeding in cirrhotic patients with high-risk esophageal varices. Hepatology 1997; 25: 1346–50. PubMed CrossRef

98 Lo, GH, Lai, KH & Cheng, JS et al. Prophylactic banding ligation of high-risk oesophageal varices in patients with cirrhosis: a prospective randomized trial. J Hepatol 1999; 31: 451–6. PubMed CrossRef

99 De, BK, Ghoshal, UC & Das, T et al. Endoscopic variceal ligation for primary prophylaxis of esophageal variceal bleed: preliminary report of a randomized controlled trial. J Gastroenterol Hepatol 1999; 14: 220–4. PubMed CrossRef

100 Sarin, SK, Lamba, GS & Kumar, M et al. Comparison of endoscopic ligation and propranolol for the primary prevention of variceal bleeding. N Engl J Med 1999; 340: 988–93. PubMed CrossRef

101 Imperiale, TF & Chalasani, N. A meta-analysis of endoscopic variceal ligation for primary prophylaxis of esophageal variceal bleeding. Hepatology 2001; 33: 802–7. PubMed CrossRef

102 Sarin, SK, Lahoti, D & Sazena, SP et al. Prevalence, classification and natural history of gastric varices: a long term follow up study in 568 portal hypertensive patients. Hepatology 1992; 6: 1343–9.

103 Trudeau, W & Prindiville, T. Endoscopic injection sclerosis in bleeding gastric varices. Gastrointest Endosc 1985; 32: 264–8.

104 Yassin, YM, Eita, MS & Hussein, AMT. Endoscopic sclerotherapy for bleeding gastric varices. Gut 1985; 26: A1105.

105 Lo, GH, Lai, KH, Cheng, JS, Chen, MH & Chiang, HT. A prospective, randomized trial of butyl cyanoacrylate injection versus band ligation in the management of bleeding gastric varices. Hepatology 2001; 33: 1060–4. PubMed CrossRef

106 Caletti, GC, Brocchi, E, Baraldini, M, Ferrari, A, Gilbiaro, M & Barbara, L. Assessment of portal hypertension by endoscopic ultrasonography. Gastrointest Endosc 1990; 36: S21–7.

107 Caletti, GC, Brocchi, E & Ferrari, A et al. Value of endoscopic ultrasound in the management of portal hypertension. Endoscopy 1992; 24: 342–6. PubMed

108 Tio, L, Kimming, S & Rauw, E et al. Endosonography of gastroesophageal varices: evaluation and follow-up of 76 cases. Gastrointest Endosc 1995; 42: 145–50. PubMed

109 Kishimoto, H, Sakai, M & Kajiyama, T et al. Miniature ultrasonic probe evaluation of esophageal varices after endoscopic variceal ligation. Gastrointest Endosc 1995; 42: 256–60. PubMed

110 Miller, LS, Schiano, TD & Adrain, A et al. Comparison of high-resolution endoluminal sonography to video endoscopy in the detection and evaluation of esophageal varices. Hepatology 1996; 24: 552–5. PubMed

111 Burtin, P, Cales, P & Oberti, F et al. Endoscopic ultrasonographic signs of portal hypertension in cirrhosis. Gastrointest Endosc 1996; 44: 257–61. PubMed

112 Irisawa, A, Obara, K & Sato, Y et al. EUS analysis of collateral veins inside and outside esophageal wall in portal hypertension. Gastrointest Endosc 1999; 50: 374–80. PubMed

113 Leung, VKS, Sung, JJY & Ahuja, A et al. Large para-esophageal varices on endosonography predicts recurrent oesophageal sclerotherapy for oesophageal varices. Gastroenterology 1997; 112: 1811. PubMed

114 Lo, GH, Lai, KH & Cheng, JS et al. Prevalence of paraesophageal varices and gastric varices in patients achieving variceal obliteration by banding ligation and by injection sclerotherapy. Gastrointest Endosc 1999; 49: 428–36. PubMed

115 Irisawa, A, Saito, A & Obara, K et al. Endoscopic recurrence of esophageal varices is associated with the specific EUS abnormalities: severe peri-esophageal collateral veins and large perforating veins. Gastrointest Endosc 2001; 53: 77–84. PubMed CrossRef

116 Suzuki, T, Matsutani, S & Umebara, K et al. EUS changes predictive for recurrence of esophageal varices in patients treated by combined endoscopic ligation and sclerotherapy. Gastrointest Endosc 2000; 52: 611–17. PubMed CrossRef

117 Lahoti, S, Catalano, MF & Alcocer, E et al. Obliteration of esophageal varices using EUS-guided sclerotherapy with color Doppler. Gastrointest Endosc 2000; 53: 331–3.

118 Lee, YT, Sung, JJY & Yung, MY et al. The use of color Doppler EUS in assessing azygos blood flow for patients with portal hypertension. Gastrointest Endosc 1999; 50: 47–52. PubMed

119 Kassem, AM, Salama, ZA & Zakaria, MS et al. Endoscopic ultrasonographic study of the azygos vein before and after endoscopic obliteration of esophagogastric varices by injection sclerotherapy. Endoscopy 2000; 32: 630–4. PubMed CrossRef

120 Lee, YT, Chan, FKL & Ng, EKW et al. Endosonography guided injection of cyanoacrylate for bleeding gastric varices. Gastrointest Endosc 2000; 52: 168–74. PubMed

Copyright © Blackwell Publishing, 2005