*2.2. Specific Management of Acute Esophageal VH*

Standard therapy for acute VH includes intravenous splanchnic vasoconstrictors and placement of rubber bands around esophageal varices, especially the one that is expected to be the source of bleeding [5–8]. Endotracheal intubation to protect the airway system may be considered in patients with massive bleeding prior to endoscopy. However, whether intubation is really protective or increases the risk of respiratory infections is unclear [20], therefore, it cannot be recommended for every patient.

#### 2.2.1. Intravenous Splanchnic Vasoconstrictors

Three intravenous splanchnic vasoconstrictors are available: terlipressin, somatostatin or octreotide. These drugs exert their action by reducing splanchnic blood flow, therefore lowering portal pressure [35]. They are very effective and a recent meta-analysis clearly demonstrated that the use of vasoconstrictors is associated with a significantly higher probability of bleeding control and a lower seven-day mortality [36]. As a proof of concept, treatment with vasoconstrictors alone was previously found to control bleeding in >80% of patients [37]. It is most likely the widespread adoption of these drugs, together with the optimization of general medical care, that has significantly lowered the VH-related short-term mortality in the recent years [38].

A vasoconstrictor shall be initiated as soon as possible and early administration is associated with improved survival [5–8]. A placebo-controlled trial in which terlipressin was administered during the ambulance transfer showed that such early timing of administration was associated with increased probability to control of bleeding and survival in the treatment arm [39].

In clinical practice, the choice among these three intravenous vasoconstrictors is dictated by local availability and cost [40]. Recommended dose for terlipressin of 2 mg/4 h during the first 48 h, followed by 1 mg/4 h. If terlipressin is contraindicated, somatostatin is an alternative and should be administered as a continuous infusion of 250 mg/h (that can be increased up to 500 mg/h), with an initial bolus of 250 mg. The recommended dose of octreotide is a continuous infusion of 50 mg/h with an initial bolus of 50 mg [5–8].

Vasoconstrictors should be continued up to five days after the confirmation of VH because the risk of rebleeding during this time is particularly high [5–8]. However, as vasoconstrictors may be associated with potentially serious adverse events, the feasibility of a shorter administration (i.e., 24–48 h vs. 3–5 days) has been considered. In a recent meta-analysis, although the risk of 42-day mortality was not significantly different between one to three and five days, risk stratification was missing [41]. It may be that Child A patients could receive a shorter duration of therapy, whereas all others would require five days, but further studies are required to answer this question.

In summary, guidelines recommend that an intravenous splanchnic vasoconstrictor shall be initiated as soon as possible, prior to diagnostic endoscopy, and be administered for three to five days [5–8].

#### 2.2.2. Endoscopic Therapy

Once hemodynamic stability has been reached, an upper GI endoscopy shall be performed to determine the cause of bleeding and to provide specific treatment [42]. Early data and one relatively recent retrospective study suggested that endoscopy within 12 h from the index event might be associated with reduced rates of recurrent bleeding and mortality [43]. On the other hand, in a larger multicenter study including 1373 patients with cirrhosis and VH, endoscopy within 24 h from admission was associated with lower mortality in patients with Child A or B cirrhosis (OR: 0.38, 95% CI: 0.16–0.86; *p* = 0.020) and in those with systolic blood pressure <90 mmHg (OR: 0.053, 95% CI: 0.006–0.51; *p* = 0.011) [44]. In contrast, performance of endoscopy within either 6 or 12 h was not associated with a further reduction in mortality compared with endoscopy within 24 h. Interestingly, the association between endoscopy within 24 h and reduced mortality was seen in Child A and B patients, but not in the overall group including also Child C [44].

This notwithstanding, current guidelines recommend that once hemodynamic stability has been achieved, endoscopy should be performed as early as possible, and within 12 h since presentation [5–8]. When VH is confirmed, either by the presence of a bleeding varix, a clot, or a "white nipple" over the varix, or when varices are the only abnormality observed that would explain the hemorrhage, all esophageal varices should be ligated, particularly the one that is considered the source of hemorrhage. Endoscopic variceal ligation (EVL) should be performed within the same endoscopy session. EVL is more effective than sclerotherapy, is associated with fewer adverse effects, and does not lead to further increase in portal hypertension [45]. Therefore, sclerotherapy should be restricted to the rare cases in whom ligation is not technically feasible. Hemostatic powder applied endoscopically may be considered as a rescue therapy, however, few data exist and its applicability remains to be determined [46,47].

In patients with uncontrolled bleeding, guidelines recommend placement of balloon tamponade (Sengastaken-Blackemore or Minnesota tubes) [5–8]. However, tamponade carries a high risk of complications, particularly respiratory infection [20], and shall be considered only as a temporary (maximum 24 h) bridge to TIPS [5]. Recent data suggest that placement of a self-expandable esophageal metal stent (placed orally or endoscopically) may be associated with greater bleeding control and lower adverse events compared to balloon [48]. As these stents may remain in place for up to 7 days, this would allow more time to plan for a definitive treatment. Per the last Baveno consensus, if available, the application of stents may be considered as a preferred alternative compared with balloons [5].
