NSBBs decrease cardiac output (β-1 receptor blockade) and promote vasoconstriction of the mesenteric vasculature (β-2 receptor blockade) resulting in decreased portal venous inflow and consequently portal pressure. The most commonly used NSBBs include propranolol and nadolol [1, 14]. Recently carvedilol, non-cardioselective β-blocker with α-1 blocking properties has been used with promising results [15] (see Table 9.3).

The goal of treatment with NSBBs is to decrease the HVPG to less than 12 mmHg which reduces the risk of hemorrhage and improves survival [16]. A reduction of 10–20 % of HVPG baseline decreases the risk of the first variceal hemorrhage [17, 18]. NSBBs are associated with a median reduction of HVPG of 15 %, with 37 % of patients being responders defined as a reduction of HVPG to < 12 mmHg and/or > 20 % from baseline [19]. The reduction of HVPG with NSBBs might have additional benefits beyond the prevention of variceal bleeding , such as decreased frequency of development of ascites, spontaneous bacterial peritonitis, and death [17, 20].

Routine measurement of HVPG to guide β-blockade therapy is not regularly practiced and is limited to specialized referral centers. For this reason, surrogates markers such as a decrease in heart rate of 25 % below baseline, heart rate of 55–60 beats per min or maximally tolerated dose are followed as guides for medication adjustment [1, 14]. The risk of bleeding recurs when NSBBs are stopped, so in general, therapy should be continued indefinitely [21].

NSBBs have common side effects that include lightheadedness, fatigue, erectile dysfunction, and shortness of breath. Relative contraindications for the initiation of NSBBs include sinus bradycardia, relative hypotension, and insulin-dependent diabetes. Absolute contraindications are severe obstructive pulmonary disease, heart failure, aortic valve disease, heart block, and peripheral arterial insufficiency [14, 22].

The use of NSBBs for primary prophylaxis reduces the incidence of variceal hemorrhage. Meta-analysis has compared NSBBs versus placebo. These studies have shown that the bleeding rate in the placebo group was 25 %, as compared to 15 % in the treatment group over a 2-year period [23].

NSBBs for patients with small varices are associated with a nonstatistical reduction in the incidence of the first variceal hemorrhage. The bleeding rate was reduced from 7 % in the untreated patients to 2 % in those on NSBBs therapy over a 2-year period [23].

A multicenter, single-blinded trial compared the use of nadolol against placebo in the evolution of varices . The treated group showed slower progression to large varices (11 vs. 37 %) at 3 years without survival differences. The benefit of β-blockade was an expression of the time patients remained with small sized varices [24].

Special attention should be placed to identify patients with small varices at high risk of bleeding: Child–Pugh class B/C cirrhosis or varices with red wale marks in whom treatment with NSBBs is indicated to prevent the first bleeding episode. In the absence of high-risk features, treatment can be considered for small varices, but its long-term benefit has not been fully established and the benefit may be outweighed by medication side effects. If no treatment is given in this setting, reassessment should be performed in 2 years unless hepatic decompensation occurs (see Table 9.1) [1].

A large meta-analysis that included more than 1100 patients with medium/large varices showed that NSBB therapy reduced the risk of the first variceal bleeding to 14 % as compared to 30 % in the placebo group. This study also showed benefit in mortality in the pharmacologically treated group. The number needed to treat to prevent one bleeding episode was ten [23, 25].

In general, NSBBs are the first-line therapy which can be switched to EVL if not tolerated, ineffective, or contraindicated. Up to 15 % of patients have relative contraindications to the use of NSBBs. Some of the side effects disappear with continued use or dose reduction. Treatment withdrawal occurs in approximately 15 % (10 % for nadolol and 17 % for propranolol) [26]. The selection of treatment modality should consider patient preferences, local resources and contraindication, or adverse effects to therapy [27].

In 2010, a prospective trial reported increased mortality when utilizing NSBB in refractory ascites [28]. Another study described worse outcomes (hemodynamic compromise, hepatorenal syndrome, and reduced transplant free survival) when NSBB were continued after an episode of spontaneous bacterial peritonitis [29]. The concept of a therapeutic window for NSBB for variceal hemorrhage prevention has been hypothesized. The opening point is still debated, but it is believed to be the initiation of gut bacterial translocation in the presence of ascites (beginning of hepatic decompensation). The closing point occurs when the cardiac compensatory reserve is lost (end-stage cirrhosis) ; hypotension and decrease end organ perfusion (refractory ascites, hepatorenal syndrome, spontaneous bacterial peritonitis) [30]. This concept is still evolving and should be interpreted with caution.

EVL involves suctioning of a varix into the endoscope channel followed by deployment of a rubber band, so that the tissue undergoes necrosis and fibrosis. Bands should be placed in areas that show evidence of recent bleeding (red whale sign or nipple sign) or staring at the gastroesophageal junction and moving proximally every 2 cm in the distal esophagus. If EVL is the treatment chosen, then repeat EGD should be performed every 2–4 weeks until obliteration. The first surveillance EGD should occur in 1–3 months after obliteration and then every 6–12 months to evaluate for recurrence [1, 14].

Although, the literature suggests a slight advantage of EVL in the prevention of the first variceal bleed, well-designed meta-analysis have shown no difference when comparing NSBBs to EVL [31]. No difference in mortality has been shown [32, 33]. Meta-analyses have demonstrated that in high-risk varices (large with or without red wale signs), EVL is associated with a slight but significant decrease risk of the first variceal hemorrhage episode when compared to NSBBs, but without impact on mortality. The rate of adverse events from EVL is lower than NSBB therapy but usually more serious including bleeding from ligation induced esophageal ulcers [34, 35] (see Fig. 9.2).

Endoscopic sclerotherapy alone or the combined use of NSBBs and nitrates for primary prophylaxis has yielded mixed results. Shunt surgery is effective in preventing the first bleeding but it is associated with higher mortality and hepatic encephalopathy. None of these therapies are recommended for this purpose [1, 36–38].

The importance of aggressive secondary prophylaxis lies in that the absence of therapy is associated with a recurrence rate of 60 % at 1 year and a mortality of 33 % [23, 39]. The use of NSBBs, EVL, or shunt procedures reduces significantly the rebleeding rate. All patients who are potential transplant candidates and that have recovered from variceal bleeding should be referred to a transplant center for evaluation [1].

NSBBs alone reduce the rebleeding risk from 63 to 42 % (number needed to treat; NNT 5) when compared to placebo. The mortality in these comparison groups was also reduced from 27 to 20 % (NNT 14) [23]. Besides NSBBs, nitrates have systemic vasodilating effects and reduce portal pressure through splanchnic vasoconstriction as a result of hypotension rather than from intrahepatic vasodilation. Nitrates (isosorbide mononitrate) are not recommended as monotherapy, but when used in combination with NSBBs might have synergistic effects in the reduction of portal pressure [40].

Some randomized clinical trials support a reduction in rebleeding with the combined use of isosorbide–mononitrate and NSBBs, but when pooled data have been assessed in meta-analyses this has not been supported [41–43]. Compared with endoscopic therapy , there may be a survival advantage in using isosorbide mononitrate and NSBBs, but long-term and better designed studies are still needed to validate its routine use [41]. In clinical practice, this form of combination therapy is associated with frequent side effects that mandate discontinuation of therapy [23, 42].

The combination of NSBBs with EVL is the standard of care for secondary prophylaxis . The rebleeding risk is reduced from 38 to 47 % when using EVL alone to 14–23 % for EVL and NSBBs combined [44, 45]. The rebleeding risk is also reduced when combination therapy is compared to medical therapy alone. There is no mortality benefit when EVL and NSBBs are used [46].