What is the best approach for primary prevention?
1.Recommend non-selective beta-blockers (NSBB) or varicose vein ligation (VBL), along with propranolol as first-line therapy, or VBL if there is a contraindication to NSBB. the choice of NSBB or VBL should be based on the patient’s wishes (1a, level A).
2. Carvedilol or nadolol is recommended as an alternative to propranolol (1b, level A).
3.Dose of drug use
Propranolol: 40 mg bid, the maximum dose tolerated or 320 mg at a heart rate of 50-55 beats/min (Class 1b, Level A).
Nadolol: 40 mg qd, maximum tolerated dose or 240 mg at 50-55 bpm (1b, Class A).
Carvedilol: initial dose of 6, 25 mg qd, with increments of 12, 5 mg after one week if tolerated or if heart rate decreases to 50-55 beats/min (1b, Level A).
If spontaneous bacterial peritonitis, renal impairment, or hypotension develops, discontinuation of NSBB is recommended (level 2b, B).
4. If NSBB is contraindicated or not tolerated, VBL is recommended (level 1a, A).
Which patients need to be followed for the presence of variceal bleeding?
Endoscopy is recommended for all patients with cirrhosis at the time of first diagnosis (grade a, A). For patients on NSBB, there is no indication for repeat endoscopy.
How often does a patient with cirrhosis need an endoscopy?
1. If no varices are found at the first endoscopy, repeat endoscopy is recommended every 2-3 years (level 2a, B).
2.If grade I varices are found at the time of diagnosis of cirrhosis, patients are advised to repeat endoscopy (grade 2a, B) every year.
3. If there is sufficient evidence of disease progression, clinicians may modify the interval of repeat endoscopy according to the actual situation. Endoscopy should also be performed in the decompensated phase of cirrhosis (grade 2a, B).
Which patients need primary prevention?
Primary prevention is indicated regardless of the severity of liver disease if grade I varices with red signs or grade 2-3 varices are found at the time of diagnosis of cirrhosis (grade 1a, A).
Treatment not recommended
1. Proton pump inhibitors are not recommended unless the patient has concurrent ulcer disease (1b, level B).
2. Isosorbide mononitrate alone is not recommended for primary prevention (1b, Level A), and there is insufficient evidence to recommend its use in combination with NSBB (1b, Level A).
3.Shunt surgery or transjugular intrahepatic portosystemic shunt (TIPSS) is not recommended as primary prevention (1a, level A).
4. Sclerotherapy is not recommended as a primary prevention treatment (1a, Level A).
Areas for further research
1.The effect of NSBB on patients with cirrhosis without varices, especially carvedilol.
2, The role of NSBB in patients with the presence of small varicose veins, especially carvedilol.
3.Comparison of carvedilol and propranolol in primary prevention.
4, Identification of new drugs available for primary prevention and subsequent clinical evaluation, such as statins.
Quality indicators
1. Proportion of patients with cirrhosis who underwent endoscopic screening for varices at diagnosis (1a, level A) Numerator: number of patients with newly diagnosed cirrhosis who underwent endoscopy before or within 6 months after diagnosis. Denominator: Patients with newly diagnosed cirrhosis.
2. Proportion of newly diagnosed cirrhotic patients with combined grade I varices, red signs or grade 2-3 varices who underwent primary prevention. Numerator: Number of newly diagnosed cirrhotic patients with grade I varicose veins, red signs or grade 2-3 varicose veins undergoing primary prevention. Denominator: Total number of newly diagnosed cirrhotic patients with grade I, red or grade 2-3 varicose veins.
Treatment of active variceal bleeding
Most studies have reported an average 6-week mortality rate of up to 20% after the first variceal bleed, with survival rates having improved significantly since the 1980s, when mortality during hospitalization was 40-50%, and more recently in the UK, where the current outcome is 15%
Drug treatment
The two main drugs that have been used to control acute variceal bleeding are vasopressin or its derivatives (alone or in combination with nitroglycerin) and growth inhibitors or their derivatives. Terlipressin is the only drug that has been shown to reduce mortality in placebo-controlled clinical trials.
However, in a systematic review and a recent large RCT comparing the efficacy of terlipressin, growth inhibitors, and octreotide in the treatment of variceal bleeding, no difference was found. Prophylactic application of antibiotics may also reduce mortality.
1.Vasopressin
Vasopressin can reduce the flow of the portal vein and portal collateral circulation shunt and decrease the varicose vein pressure. However, it has obvious adverse effects on the body circulation, such as increasing peripheral resistance, reducing cardiac output, heart rate and coronary blood flow.
Clinical trials have shown that vasopressin reduces variceal bleeding treatment failure rates compared to non-aggressive treatment (OR = 0,22), but has no effect on survival. A meta-analysis comparing the efficacy of sclerotherapy with vasopressin yielded similar findings.
2. Vasopressin versus nitroglycerin
Three clinical trials comparing the efficacy of vasopressin alone and vasopressin in combination with nitroglycerin suggested that the combination reduced the rate of treatment failure for gastrointestinal bleeding (OR = 0, 39), but had no survival benefit.
3. Terlipressin
Terlipressin is a synthetic analogue of vasopressin, which is converted to vasopressin in vivo and has a systemic vasoconstrictor effect, with subsequent effects on portal vascular dynamics.
A meta-analysis showed that terlipressin reduced failure to control bleeding and increased survival, while terlipressin, vasopressin, endoscopic therapy, and balloon compression did not differ in terms of treatment failure and survival. Terlipressin combined with VBL will be discussed in the section on endoscopic therapy combined with pharmacological therapy.
The recommended dose of terlipressin is 2 mg IV every 4 h. However, because of its peripheral vasoconstrictive effect and the pain it causes in the hands and feet, many providers have reduced the dose to 2 mg q6 h. The Baveno V consensus recommends 5 days of IV application, but because it does not significantly improve survival, many providers discontinue it once bleeding has stopped for practical purposes.
A randomized clinical trial has shown that in esophageal variceal bleeding, when terlipressin and VBL are combined, discontinuing the former after 24 hours of hemostasis is as effective as discontinuing the former after 72 hours of hemostasis.
If patients cannot tolerate terlipressin or if terlipressin is not available in some countries, its replacement can be considered.
4.Growth inhibition and octreotide
Growth inhibitor can selectively cause visceral vasoconstriction, reduce portal venous pressure and decrease portal venous blood flow. Octreotide is an analogue of growth inhibitor. The mechanism of action of the above two drugs is not clear. It increases vasodilation by inhibiting glucagon and thus, rather than having a direct vasoconstrictive effect; it also decreases postprandial intestinal congestion.
The effects of octreotide on hepatic and circulatory hemodynamics are transient and therefore require continuous dosing, starting with a 50ug loading dose followed by 25-50ug/h. Growth inhibitors are administered intravenously with a 250ug loading dose followed by 250 mg/h.
A meta-analysis showed that growth inhibitors and octreotide were equivalent to terlipressin in the treatment of acute variceal bleeding, and a large RCT by Seo et al. comparing the efficacy of these three drugs found no difference in treatment success, rebleeding rates, or mortality among the three.
Low systolic blood pressure on admission, high creatinine levels, active bleeding on emergency gastroscopy, gastric varices, and Child-c were all independent risk factors predicting 5-day treatment failure.
5. Antibiotics
A meta-analysis showed that the application of antibiotics capable of covering gram-negative bacteria improved survival. Studies have shown that antibiotics can reduce bacterial infections and early rebleeding.
Therefore, short-term intravenous antibiotic application should be the standard measure in all patients with cirrhotic variceal bleeding, regardless of positive evidence of infection. Third-generation cephalosporins such as ceftriaxone (1 g qd iv) are superior to oral norfloxacin in reducing sepsis in gram-negative bacteria. However, the local spectrum of drug-resistant bacteria and available drugs should be fully considered when choosing antibiotics.
6.Proton pump inhibitors
An RCT comparing proton pump inhibitors with vasoconstrictors in patients with acute varicose veins in the short term after hemostasis found no difference in bleeding and survival rates despite the larger ulcer size in the latter. 50% of patients had ascites and proton pump inhibitors had an increased risk of spontaneous bacterial peritonitis.
Endoscopic treatment
Endoscopy should be performed within 24 hours of admission or even earlier if there is significant bleeding, and the above opinion is based on a low level of evidence.
Many guidelines and reviews recommend that endoscopy should be performed within 12 hours, however, the only trial that examined the prognostic impact of the timing of endoscopy found no advantage to performing endoscopy within 12 hours of admission.
The most appropriate timing is after adequate resuscitation and medication, by an experienced endoscopist, in a well-equipped endoscopy center, and with airway protection. Airway protection is necessary when the risk of aspiration is high, which allows the endoscopist to perform a thorough evaluation, including adequate aspiration of the clot and necessary therapeutic means such as pressure plugs.
The endoscopic team must include a nurse experienced in endoscopic treatment of varicose veins and an endoscopist skilled in the application of endoscopic lancing devices and compression therapy.
1. Varicose vein ligation
This technique was developed from the elastic ligatures used to treat internal hemorrhoids and was first used in humans in 1988. A meta-analysis including 7 clinical trials comparing VBL with sclerotherapy for acute variceal bleeding showed that the former reduced the rate of variceal rebleeding (OR = 0,47), mortality (OR = 0,67), lower incidence of esophageal strictures, and fewer treatments required to make the varices disappear.
2. Sclerotherapy
Sclerotherapy has been replaced by VBL and should no longer be the standard of care for acute variceal bleeding.
3.Other endoscopic treatment measures
An RCT study showed no advantage of cyanoacrylate binders over VBL and a high risk of embolization and rebleeding.
A small study including 7 patients showed no rebleeding within 24 hours and no patient death within 15 days with hemostatic powder spraying for acute variceal bleeding.
4. Endoscopic treatment combined with drug therapy
Meta-analysis has shown that endoscopic treatment (VBL or sclerotherapy) combined with medication is effective in early control of bleeding and has a high rate of 5-day hemostasis, but has no effect on survival.
Two RCTs have compared the efficacy of VBL in combination with pharmacotherapy or sclerotherapy in combination with pharmacotherapy, with one clinical trial using vasopressin showing a high rate of hemostasis at 72 hours and fewer complications, and the other clinical trial showing a low rate of failure to control acute bleeding and fewer serious complications in the former. Overall survival was similar in both clinical trials.
Balloon compression
Balloon compression is very effective, controlling bleeding in more than 90% of patients, although about 50% of patients rebleed after balloon deflation and 15-20% develop serious complications, such as esophageal ulceration and aspiration pneumonia. Nevertheless, it may be a life-saving treatment for uncontrollable variceal haemorrhage where other treatments have failed.
Placement of an appropriate SengstakenCBlakemore tube buys time for resuscitation and transport, and allows for repeat endoscopy or interventional shunts provided that hemodynamics are stable. Esophageal balloon inflation is rarely required and it should not be used alone, but only in precursors where the gastric balloon is properly placed and the pressure is appropriate but there is still persistent bleeding.
Endoscopic placement of a SengstakenCBlakemore tube or placement guided by a guidewire may be able to reduce the risk of complications such as esophageal stricture.
Removable esophageal stents
The SX-Ella Danis is a removable laminated metal mesh stent that is placed endoscopically in the lower esophagus and has no effect on fundic variceal bleeding. These stents can be left in place for up to 2 weeks, unlike the SengstakenCBlakemore tube, which must be removed after a maximum of 24-48 hours. There are no clinical trials comparing the efficacy of these devices with balloon compression.
Transjugular intrahepatic portosystemic shunts
Several uncontrolled clinical trials have studied the role of simple salvage TIPSS in acute variceal bleeding, and a meta-analysis of 15 studies showed that it controlled bleeding at 90-100%, rebleeding at 6-16%, and mortality at 15% (30 days)-75% (during hospitalization), but note that in most of the trials included in the meta-analysis, sclerotherapy was the first-line endoscopic treatment in most of the trials included in the meta-analysis.
A long-term follow-up study of the efficacy of TIPSS versus H-type portosystemic shunts in patients who failed nonsurgical treatment suggested that the latter was effective in reducing portal pressure and had a lower failure rate, but did not improve survival in patients other than Child A and B patients.
A recent RCT comparing the efficacy of emergency portal shunts with TIPSS alone for the treatment of acute esophageal variceal bleeding in patients with nonselective cirrhosis showed that the former was superior to the latter in improving long-term control of bleeding, hepatic encephalopathy, and prognosis. More research is needed regarding the promotion of overlapping stents before the procedure is widely used for the treatment of acute variceal bleeding.
Evidence suggests that overmolded TIPSS stents should be used in practice rather than bare metal stents. Randomized controlled trials have shown that overmolded stents have a higher patency rate than bare stents and reduce the risk of hepatic encephalopathy, but there is no significant difference in survival impact between the two.
However, evidence from two RCT studies suggests that stratification of cirrhotic patients by HVPG, Child class, and active bleeding requires early implementation of TIPSS for eligible patients rather than as a remedy.
Monescillo et al. randomized patients with acute esophageal variceal bleeding with HVPG ≥ 20 mm Hg within 24 hours of admission to two groups, one receiving TIPSS and one receiving standard treatment, sclerotherapy.
When treatment failure was defined as failure to control acute bleeding and/or early rebleeding, the former significantly reduced treatment failure (12% vs. 50%) and improved survival (62% vs. 35%).
Garcia-Pagan et al. randomized patients with active bleeding Child B cirrhosis or Chlid C (score less than 14) to the TIPSS within 72 hours group and the standard treatment group receiving VBL and medication, which had a lower risk of treatment failure (3% vs 50%) and improved 1-year survival (86% vs 61%) with no increased risk of hepatic encephalopathy. There was no increased risk of hepatic encephalopathy.
In addition, a recent observational study of early TIPSS did not yield such high survival rates, with an 11-year survival rate of 67%, similar to that of patients receiving endoscopic therapy and medication alone.
Therefore, larger multicenter randomized controlled studies are needed to further evaluate the efficacy of early TIPSS. To prevent rebleeding, it is important to distinguish between salvage TIPSS and early TIPSS.
Liver transplantation
Liver transplantation may only be appropriate for patients who are awaiting liver transplantation for bleeding, but there are no studies comparing the efficacy of VBL or TIPSS with emergency liver transplantation. However, it is a rare option for patients because of the tight liver supply and other factors. There are no clinical trials on liver transplantation for uncontrolled/active bleeding.