Acute variceal bleeding

<p class="article-intro">Acute variceal bleeding (AVB) is one of the major causes of death in cirrhotic patients. The management during the acute phase and the secondary prophylaxis are well defined. Recent recommendations (2015 Baveno VI expert consensus) are available and should be followed for an optimal management, which must be performed in emergency in a liver or general intensive care unit.</p> <p class="article-content"><div id="keypoints"> <h2>Keypoints</h2> <ul> <li>AVB patients should be managed in emergency in an intensive care unit. A restrictive transfusion strategy should be applied.</li> <li>Initial management always includes vasoactive drugs, antibiotic prophylaxis and emergency EBL.</li> <li>Secondary prophylaxis depends on the severity of underlying cirrhosis. TIPS placement should be discussed in severe patients.</li> <li>Liver transplantation should be discussed when liver function is deteriorated.</li> </ul> </div> <h2>Introduction</h2> <p>Acute variceal bleeding (AVB) is one of the major causes of death in cirrhotic patients.¹ It is also the major cause of upper gastrointestinal (GI) bleeding in cirrhotic patients, accounting for 70 % of cases.² The mortality during the first episode is estimated to 15-20 % .3 It is high in severe patients (Child Pugh C), around 30 % , but is very low in patients with compensated cirrhosis (Child Pugh A).³ The main predictors of bleeding in clinical practice are: large versus small varices, red wale marks, Child Pugh C versus Child Pugh A-B.⁴<br /><br /> In recent years, improvements have been made concerning the management of acute variceal bleeding (AVB), leading to a better prognosis.^{5, 6} Recent recommendations (2015 Baveno VI expert consensus) summarize the most important aspects.⁶ In this review, we will discuss Baveno VI conclusions and more recent data in order to provide guidance for an optimal management of AVB.</p> <h2>Management of acute bleeding</h2> <p><strong>Unspecific measures</strong><br /> Resuscitation<br /> Restoring mean arterial pressure (MAP) allowing an adequate tissue perfusion is of paramount importance. Prolonged hypovolemia favors kidney failure, bacterial infections and increases mortality. ^{7, 8} Conversely, portal pressure follows a linear correlation with MAP. As such, excessive MAP could promote bleeding.^{9, 10} The optimal MAP is not well defined in this context, but a target of around 65mm- Hg can reasonably be extrapolated from recommendations established for septic or hemorrhagic shock in trauma patients.^{11, 12} Transfusion of packed red blood cells should be performed with a restrictive hemoglobin target of 7g/dl.^{13, 6} A liberal transfusion strategy (hemoglobin target of 9g/dl) has been associated with a poorer prognosis in Child Pugh A and B patients. This effect was not observed in the most severe patients (Child Pugh C), either because of a lack of power of the study, or because of a more severe hypovolemia in these patients. Transfusion strategy should be adapted to specific situations such as high cardiovascular risk. The effect of fresh frozen plasma administration has never been evaluated during AVB and no recommendations are available. It is important to note that prothrombin rate and INR (international normalized ratio) should be considered as markers of liver function and not of coagulation disorders. Correcting them is not part of the management of AVB. Platelet transfusion is usually recommended when the platelet count falls below 30.000/mm.⁶ No specific data are available for AVB patients. Two studies have evaluated the effect of recombinant factor VII administration during AVB, but no significant results have been obtained considering the whole population. ^{14, 15} A potential efficacy in the subgroup of patients with severe cirrhosis (Child Pugh C) with active bleeding was noted in post-hoc analyses. This therapeutic is not recommended.<br /><br /> Prior to endoscopy<br /> In the absence of contraindications (QT prolongation), infusion of erythromycin 250mg, as a prokinetic agent, should be administered in order to improve stomach clearance and thus facilitate endoscopy.6 Gastric lavage has shown no superiority compared to erythromycin or both therapeutics. ¹⁶ <br /><br /><strong>Specific measures</strong><br /> The management of AVB associates vasoactive drugs, antibiotic prophylaxis and endoscopic band ligations (EBL).⁶<br /><br /> Vasoactive drugs<br /> Three types of drugs are available: somatostatin, somatostatin analogs such as octreotide, and terlipressin. All these drugs induce splanchnic vasoconstriction and reduce portal pressure. The choice depends on availability, cost and contraindications. Somatostatin has an effect on splanchnic hemodynamic through splanchnic arterial vasoconstriction, leading to a decrease of portal pressure (reflected by wedged hepatic pressure).^{13, 17, 18} This effect has been shown during AVB.^{19, 20} Somatostatin, early administered before endoscopy, tested against placebo, has led to fewer cases of active bleeding during endoscopy and fewer hemorrhagic recurrences. ^{18, 21} In practice, somatostatin or octreotide should be administered as early as possible with bolus infusion of 250&micro;g and 50&micro;g followed by continuous infusion of 250&micro;g/h and 50&micro;g/h, respectively.^{22, 23} Terlipressin is a vasopressin analog. It is an arterial vasoconstrictor with splanchnic and general effect. It should not be used in high cardiovascular risk patients. Early administration of terlipressin against placebo during AVB has led to an improved survival.^{24, 25} Recently, a randomized trial including more than 1.000 patients has compared the administration of somatostatin, terlipressin or octreotide after EBL and antibiotic prophylaxis. Efficacy (hemorrhage control at day 5) was similar in the 3 groups.²⁶ More adverse effects were observed in the terlipressin group, especially hyponatremia. In practice, there is no recommendation on which drug to use and the first available should be preferred.<br /><br /> Antibiotic prophylaxis<br /> Without antibiotic prophylaxis, 30-40 % of cirrhotic patients presenting with upper GI bleeding will develop a bacterial infection within one week after admission.²⁷ These infections originate from bacterial translocation towards mesenteric lymph nodes, blood stream and potentially ascites. Translocation is facilitated by GI bacterial overgrowth, higher intestinal barrier permeability and altered immunity during cirrhosis. Most infections are due to gram-negative bacteria (GNB), anaerobic bacteria are rarely isolated. All causes of upper GI bleeding (directly related to portal hypertension or not) are concerned. Infections are associated with a higher rate of early rebleeding and a higher mortality. ^{28, 29} The risk of infection rises with the severity of cirrhosis.³⁰ A systematic antibiotic prophylaxis during upper GI bleeding leads to fewer infections and a lower short term mortality, which seems to be the consequence of a lower rate of early rebleeding.³¹ For these reasons, antibiotic prophylaxis is recommended for 7 days in all cirrhotic patients admitted for upper GI bleeding.⁶ Oral norfloxacin, 400mg twice daily, is recommended. Its activity against GNB and its weak intestinal absorption makes it a treatment of choice. Other quinolones, such as ciprofloxacin or ofloxacin, can be used when the oral route is impossible, as well as amoxicillin-clavulanic acid or third generation cephalosporins. GNB resistance to quinolones is a rising concern. More recently, a study has compared antibiotic prophylaxis with IV ceftriaxone or oral norfloxacin in severe patients (Child Pugh B or C). The use of ceftriaxone was associated with a lower rate of suspected or proven infections.³² The authors attributed the results to the high rate of quinolone resistant bacteria. Nevertheless, this conclusion must be balanced for two reasons. The rate of quinolone resistant bacteria in the hospital where the study was performed was unknown. Infections were also more frequent in the quinolone group than in most studies previously published. The applicability of these results in other centers remains uncertain. Recently, in Child Pugh A patients only, a non-randomized study showed that infections occurred at a similar frequency (1 % versus 2 % ) with or without antibiotic prophylaxis.33 The advantage of antibiotic prophylaxis in these patients is thus a matter of debate, especially at a time when use of antibiotics should be cautious due to the spread of multi-resistant bacteria. Further, randomized studies will address this question and eventually lead to avoid antibiotic prophylaxis in Child Pugh A patients.<br /><br /> Endoscopic treatment<br /> Endoscopic diagnosis of variceal bleeding relies on the presence of large varices and red wale marks or active bleeding (fig. 1). The presence of blood in the stomach without any other cause but large varices is also possible. It is recommended to perform upper GI endoscopy as soon as possible (within 12 hours) after initial resuscitation. Sclerotherapy is the oldest AVB endoscopic treatment. Due to almost constant formation of an ulcer on the site of injection, sometimes responsible of a hemorrhage, sclerotherapy should be abandoned. EBL, the actual treatment of choice, should be performed during initial endoscopy.³⁴ Minor complications can occur, such as dysphagia or chest pain. Post-banding ulcer bleeding can occur around day 7 and be clinically significant. EBL has never been compared to the absence of endoscopic treatment. However, it has been compared to sclerotherapy in several studies including 2 meta-analyses, all in favor of less rebleeding and less side effects when performing EBL.^35-37</p> <p><img src="/custom/img/files/files_datafiles_data_Zeitungen_2017_Leading Opinions_Innere_1702_Weblinks_s66_fig1.jpg" alt="" width="2150" height="649" /><br /><br /> Balloon tamponade<br /> Balloon tamponade through a nasogastric tube equipped with inflatable balloons (esophagus and stomach), usually named &ldquo;Blakemore&rsquo;s tube&rdquo;, is effective to control variceal bleeding.^{38, 39} It is associated with severe complications such as necrosis or perforation of the esophagus and aspiration pneumonia. Endotracheal intubation to protect airways is thus necessary. Hemorrhage recurrence occurs in more than 50 % of cases after deflation. It is thus only a temporary solution when bleeding is not controlled. Recently, esophageal stents have been developed to replace &ldquo;Blakemore&rsquo;s tube&rdquo;.⁴⁰ They offer the advantages to allow oral intakes after the acute phase, and to be removable. A recent randomized controlled trial comparing esophageal stenting and balloon tamponade during refractory AVB showed a better control of bleeding and less adverse events when using esophageal stents.⁴¹ Survival was better but did not reach significance. The Baveno VI recommendations stated that esophageal stenting should be preferred to balloon tamponade.<br /><br /> Emergency portosystemic shunts<br /> Surgical portosystemic anastomoses, once the rescue treatment for refractory AVB, are almost abandoned. Nowadays, intrahepatic shunting through transjugular insertion of a stent, commonly named TIPS (transjugular intrahepatic portosystemic shunt) is the treatment of choice for refractory AVB as well as for secondary prophylaxis in severe patients as we will further develop. TIPS insertion during refractory AVB is effective to control bleeding in almost all cases.⁴² Nevertheless, mortality after one year remains high (around 50 % ), even since the use of cov-ered stents, with a low risk of thrombosis leading to recurrence of portal hypertension. Elevated mortality is due to other complications of severe cirrhosis such as infections, kidney failure and encephalopathy. In our experience, making the decision of inserting a rescue TIPS should take account of the practicability of a liver transplantation (LT) in the near future.<br /><br /> Systematic portosystemic shunts for severe patients<br /> TIPS were initially dedicated to the rescue treatment of refractory bleeding. There are now data suggesting that the most severe patients benefit of a systematic TIPS insertion after an initial bleeding episode, commonly named &ldquo;early TIPS&rdquo;.^{43, 44} After initial management (vasoactive drugs and emergency endoscopic variceal ligation [EVL]), patients were randomized to receive combined therapy (NSBB and EVL) or a covered TIPS within 3 days after initial bleed. The patients in the TIPS group were more often free from rebleeding (97 % vs. 50 % ) and had a higher one year survival rate (86 % vs. 61 % ). Although external validation studies confirmed the lower rebleeding rate, the benefit on survival is still a matter of debate.⁴⁵ A possible explanation is that patients in the NEJM study were highly selected (63 patients recruited over 3 years in 9 centers), thus not fully comparable to the ones in validation studies. However, a recent meta-analysis was in favor of a benefit of survival.⁴⁶ The Baveno VI recommendations state that TIPS should be inserted in Child Pugh C or B with active bleeding at initial endoscopy. Availability of TIPS remains a practical issue. Liver transplantation should rapidly be discussed in these patients, as the deterioration of liver function and occurrence of other complications of cirrhosis become the forefront. Management of AVB in Child Pugh C 14 or 15 patients is still a matter of debate. These patients were excluded from the &ldquo;early TIPS&rdquo; studies. In our experience,⁴⁷ an &ldquo;early TIPS&rdquo; strategy is a solution if the patient can have a quick access to LT, as liver failure and encephalopathy are likely to keep deteriorating after TIPS.<br /><br /> Prevention of hepatic encephalopathy (HE)<br /> Therapeutic action to prevent HE in the context of digestive bleeding is recommended. Data are available from a prospective randomized controlled trial including 70 patients admitted for AVB with no signs of HE. Patients were allocated to receive lactulose or placebo.⁴⁸ The proportion of patients developing HE was significantly lower in the lactulose group (14 % vs. 40 % , p=0,03). Rifaximin is a non-absorbable antibiotic used in the prevention of recurrence of HE. Its use has never been studied in the context of digestive bleeding and is not recommended at the moment.<br /><br /> Proton pump inhibitors (PPI)<br /> It is recommended not to use PPI during AVB. A randomized trial has studied the effect of PPI administration after EBL.⁴⁹ The size of post banding ulcers was smaller in the PPI group but their number and clinical repercussions were similar. In addition, there is a growing amount of data showing that the use of PPI in cirrhotic patients favors bacterial infections including spontaneous peritonitis.^{50, 51}</p> <h2>Secondary prophylaxis</h2> <p><strong>Non selective &szlig;-blockers (NSBB) and EBL</strong><br /> Due to a high risk of recurrent bleeding, it is essential to initiate a secondary prophylaxis after the first episode. EBL associated with NSBB is the first line strategy recommended. NSBB (propranolol and nadolol) effect consist in lowering portal pressure through a bradycardia that leads to a decrease in cardiac output (anti- &szlig;1 action), and a splanchnic vasoconstriction (anti-&szlig;2 action).⁵² Carvedilol, a potent a- and &szlig;-blocker, has never been tested adequately in secondary prophylaxis. Its use is thus not recommended. Repeated EBL, usually every 2 to 4 weeks, leads to a disappearance of varices, and lowers the risk of bleeding. The interval between may eventually lower. A recent trial compared EBL sessions at an interval of 1 or 2 weeks.⁵³ Varices eradication was obtained quicker in the 1 week group, without additional complications. Nevertheless, the number of sessions necessary, rebleeding rate and mortality were similar in both groups. Many studies have compared the use of &szlig;-blockers associated to EVL versus one of these treatments in a single therapy.^{54, 55, 56} The results of a recent meta- analysis, combining the data of 23 trials, show a decrease of the risk of recurrent bleeding with a double treatment, without a positive impact on survival.⁵⁷ However, survival was better in patients treated with NSBB (single or combined therapy) compared to patients treated with EBL alone. Hence, EBL therapy alone should never be recommended, and in case of any contraindications to &szlig;-blockers, TIPS should be discussed. Few severe patients (Child Pugh C) were included in these studies and it is uncertain whether the results are applicable to these patients.<br /><br /> NSBB in patients with refractory ascites<br /> Several studies have raised a concern about the safety of NSBB in patients with ascites, as a higher mortality was observed in patients treated with such drugs. The detrimental effect of NSBB appears to be more pronounced in patients who experienced an episode of spontaneous peritonitis. ^{58, 59} It has, however, not been observed in all the studies addressing this question⁶⁰ and remains a matter of debate. Baveno VI&rsquo;s guidelines⁶ recommend to use NSBB with caution in patients with refractory ascites and discontinue the treatment if poorly tolerated: systolic blood pressure &lt;90mmHg, hyponatremia &lt;130mmol/l, episode of acute kidney injury. NSBB can be restarted with titration if a precipitating factor for such an event has been identified and corrected. The Baveno VI&rsquo;s guidelines recommend TIPS insertion in secondary prophylaxis if NSBB cannot be used.<br /><br /> <strong>TIPS in secondary prophylaxis</strong><br /> Extending the &ldquo;early TIPS&rdquo; criteria after an AVB episode is a much debated issue. The Baveno VI&rsquo;s guidelines recommend TIPS if NSBB cannot be used or if bleeding reoccurs in spite of a well conducted combined therapy. Recently, two randomized studies addressed this question, using covered TIPS.^{61, 62} They compared TIPS insertion versus a combined therapy in the first case, and with an initially drug only treatment in the second case. Indeed, hepatic venous pressure gradient (HVPG) was systematically measured before and after starting a combination of NSBB and nitrates. Patients without hemodynamic response were switched to EVL. In both studies, there was significantly less recurrence of bleeding in TIPS groups, but survival rates where similar, as well life quality assessments. Worth notice, in the second study, rebleeding rate was similar when comparing TIPS and hemodynamic responders to NSBB.<br /> As well as in studies evaluating the effects of NSBB and EBL, few Child Pugh C patients where included and data are missing for such patients. Hence, TIPS should not be recommended at first line for secondary prophylaxis.</p> <div id="fazit&quot;"> <h2>Conclusion</h2> The specific management of AVB is well defined (fig. 2): vasoactive drugs, antibiotic prophylaxis and EBL. For severe patients (Child Pugh C &lt;14 or B with active bleeding), a TIPS should be discussed in order to be placed within 72h. TIPS should also be inserted when hemorrhage reoccurs in spite of a well conducted secondary prophylaxis. LT should be discussed in the rare cases when TIPS doesn&rsquo;t control bleeding and when liver function is deteriorated. However, some questions are still under debate: Which patients really take advantage of an &ldquo;early TIPS&rdquo; strategy? How to improve TIPS availability? Is antibiotic prophylaxis necessary in Child Pugh A patients?</div> <div>&nbsp;</div> <div><img src="/custom/img/files/files_datafiles_data_Zeitungen_2017_Leading Opinions_Innere_1702_Weblinks_s66_fig2.jpg" alt="" width="1422" height="1294" /></div></p> <p class="article-footer"> <a class="literatur" data-toggle="collapse" href="#collapseLiteratur" aria-expanded="false" aria-controls="collapseLiteratur" >Literatur</a> <div class="collapse" id="collapseLiteratur"> <p><strong>1</strong> D&rsquo;Amico G et al: J Hepatol 2006; 44(1): 217-31 <strong>2</strong> Rudler M et al: Aliment Pharmacol Ther 2012; 36(2): 166-72 <strong>3</strong> Merli M et al: J Hepatol 2003; 38(3): 266-72 <strong>4</strong> Carbonell N et al: Hepatology 2004; 40(3): 652-9 <strong>5</strong> Rudler M: M&eacute;thodes de traitement des varices oesophagiennes. Encyclop&eacute;die m&eacute;dicale 2016 <strong>6</strong> de Franchis R et al: J Hepatol 2015; 63(3): 743-52 <strong>7</strong> Osman D et al: R&eacute;animation 2012, sous presse <strong>8</strong> Cardenas A et al: Hepatology 2001; 34(4 P t 1): 671-6 <strong>9</strong> Castaneda B et al: Hepatology 2000; 31(3): 581-6 <strong>10</strong> de Franchis R et al: Clin Liver Dis 2001; 5(3): 645-63 <strong>11</strong> Dellinger RP et al: Crit Care Med 2008; 36(1): 296-327 <strong>12</strong> Rossaint R et al: Crit Care 2010 14(2): R52 <strong>13</strong> Villanueva C et al: N Engl J Med 2013; 368(1): 11-21 <strong>14</strong> Bosch J et al: Gastroenterology 2004; 127: 815-30 <strong>15</strong> Bosch J et al: Hepatology 2008; 47(5): 1604-14 <strong>16</strong> Pateron D et al: Ann Emerg Med 2011; 57(6): 582-9 <strong>17</strong> Bosch J et al: Gastroenterology 1981; 80(3): 518-25 <strong>18</strong> Cirera I et al: Hepatology 1995; 22(1): 106-11 <strong>19</strong> Villanueva C et al: Gastroenterology 2001; 121(1): 110-7 <strong>20</strong> Avgerinos A et al: Lancet 1997; 350(9090): 1495-9 <strong>21</strong> Burroughs AK et al: Gastroenterology 1990; 99(5): 1388-95 <strong>22</strong> Sung JJ et al: Lancet 1995; 346(8991-8992): 1666-9 <strong>23</strong> Corley DA et al: Gastroenterology 2001; 120(4): 946-54 <strong>24</strong> Escorsell A et al: J Hepatol 1997; 26(3): 621-7 <strong>25</strong> Levacher S et al: Lancet 1995; 346(8979): 865-8 <strong>26</strong> Seo YS et al: Hepatology 2014; 60(3): 954-63 <strong>27</strong> Bernard B et al: Gastroenterology 1995; 108(6): 1828-34 <strong>28</strong> Pauwels A et al: Hepatology 1996; 24: 802-6 <strong>29</strong> Goulis J et al: Hepatology 1998; 27(5): 1207-12 <strong>30</strong> Hou MC et al: Hepatology 2004; 39(3): 746-53 <strong>31</strong> Bernard B et al: Hepatology 1999; 29(6): 1655-61 <strong>32</strong> Fernandez J et al: Gastroenterology 2006; 131(4): 1049-56 <strong>33</strong> Tandon P et al: Clin Gastroenterol Hepatol 2015; 13(6): 1189-96 <strong>34</strong> Augustin S et al: Am J Gastroenterol 2011; 106(10): 1787-95 <strong>35</strong> Stiegmann GV et al: N Engl J Med 1992; 326: 1527-32 <strong>36</strong> Laine L et al: Ann Intern Med 1993; 119: 1-7 <strong>37</strong> Villanueva C et al: J Hepatol 2006; 45(4): 560-7 <strong>38</strong> Sarin SK et al: Ann R Coll Surg Engl 1984; 66(1): 30-2 <strong>39</strong> Avgerinos A et al: Scand J Gastroenterol Suppl 1994; 207: 11-6 <strong>40</strong> Wright G et al: Gastrointest Endosc 2010; 71(1): 71-8 <strong>41</strong> E scorsell &Agrave; et al: Hepatology 2016; 6 3(6): 1957-67 <strong>42</strong> Azoulay D et al: J Hepatol 2001; 35(5): 590-7 <strong>43</strong> Monescillo A et al: Hepatology 2004; 40: 793-801 <strong>44</strong> Garc&iacute;a-Pag&aacute;n JC et al: N Engl J Med 2010; 362(25): 2370-9 <strong>45</strong> Rudler M et al: Aliment Pharmacol Ther 2014; 40(9): 1074-80 <strong>46</strong> Deltenre P et al: Eur J Gastroenterol Hepatol 2015; 27(9): e1-9 <strong>47</strong> Rudler M et al: Transpl Int 2013; 26(6): E50-1 <strong>48</strong> Sharma P et al: J Gastroenterol Hepatol 2011; 26(6): 996-1003 <strong>49</strong> Shaheen NJ et al: Hepatology 2005; 41(3): 588-94 <strong>50</strong> Bajaj JS et al: Am J Gastroenterol 2009; 104(5): 1130-4 <strong>51</strong> Trikudanathan G et al: Int J Clin Pract 2011; 65(6): 674-8 <strong>52</strong> Groszmann RJ et al: N Engl J Med 2005; 353: 2254-61 <strong>53</strong> Sheibani S et al: Randomized trial of 1-week versus 2-week intervals for endoscopic ligation in the treatment of patients with esophageal variceal bleeding. Hepatology 2016; 64(2): 549-55 <strong>54</strong> Garc&iacute;a- Pag&aacute;n JC et al: Gut 2009; 58(8): 1144-50 <strong>55</strong> Kumar A et al: Gastroenterology 2009; 137(3): 892-901, 901.e1 <strong>56</strong> Gonzalez R et al: Ann Intern Med 2008; 149: 109-22 <strong>57</strong> Puente A et al: Liver Int 2014; 34(6): 823-33 <strong>58</strong> Serst&eacute; T et al: Hepatology 2010; 52(3): 1017-22 <strong>59</strong> Mandorfer M et al: Gastroenterology 2014; 146(7): 1680-90.e1 <strong>60</strong> Leithead JA et al: Gut 2015; 64(7): 1111-9 <strong>61</strong> Holster IL et al: Hepatology 2016; 63(2): 581-9 <strong>62</strong> Sauerbruch T et al: Gastroenterology 2015; 149(3): 660-8.e1</p> </div> </p>