Approach to upper gastrointestinal bleeding
Upper GI bleeding is the most common complication of peptic ulceration and portal hypertension.
Professor and Head of Department, Division Gastroenterology, Department of Medicine, University of Cape Town, South Africa
Corresponding author: S R Thomson (email@example.com)
Upper gastrointestinal haemorrhage has a variety of causes (Table 1) and is the commonest complication of peptic ulceration and portal hypertension. Peptic ulceration in the duodenum or stomach and oesophageal varices are the conditions most often responsible for patients who have the potential to present with life-threatening haemorrhage.1-6 The key elements of an approach to this medical and surgical emergency are outlined in Fig. 1.
Fig. 1. Key elements of an approach to gastrointestinal haemorrhage.
Resuscitation and assessment
Whatever the actual cause, the most important initial management is to assess the haemodynamic parameters and institute appropriate resuscitation measures early. In those who are haemodynamically unstable aggressive resuscitation has been shown to improve outcome. Ninety five per cent will stabilise and it is important that they are fully assessed by history and physical examination.1 , 3 , 4 This allows management to be strategised based on whether the individual has variceal or non-variceal bleeding. The former is most likely if they have stigmata of chronic liver disease and portal hypertension and the latter assumed in the absence of these findings.
A variety of clinical factors and
laboratory tests can be incorporated into risk stratification.
The Rockall system5
6 (Table 2) is the
most commonly used and is applicable for both variceal and
non-variceal bleeding. These systems predict the likelihood of
continuing to bleed or of rebleeding, and the risk of death.
These clinical factors, age over 60, and shock on admission
are highly predictive.7 Concurrent medical
therapy is particularly important as NSAIDs and
anticoagulants, which are commonly prescribed in the elderly,
have a direct deleterious effect on coagulation. The former,
which affects platelet function, increases rebleeding risk but
there is no active therapy to negate this effect in the acute
situation. Aberrant warfarin anticoagulation can and should be
corrected by blood product replacement.2
In addition, medical comorbidities also contribute to the risk
and require active management.
Monitoring and adjunctive measures
These stratification systems can also direct appropriate monitoring of high-risk patients prior to endoscopy, which is the next step. During resuscitation and stabilisation appropriate adjunctive measures are of proven benefit. In clinically high-risk patients with suspected non-variceal bleeding, intravenous PPI therapy will elevate the pH to above 6, which allows a stable clot to be formed and once formed to be less likely to lyse.8-10 Similarly, in variceal haemorrhage the use of vasopressors or somatostatin analogues, which reduce splanchnic flow, have been shown to be of benefit.2 , 3 , 11 The risk of infection, or of infection being precipitated, in patients with chronic liver disease and variceal haemorrhage, has a deleterious effect on outcome. Hence antimicrobial therapy is both rational and of confirmed benefit.2 , 3
The aim of endoscopy is to establish the cause of bleeding and, using endotherapy, control the bleeding or reduce the likelihood of further bleeding.
The patient generally should be haemodynamically stable prior to endoscopy and for the vast majority this is possible. It is important that they are transfused to at least an Hb of 8 g/dl, particularly if they were shocked on admission. It is essential that appropriate monitoring is utilised. Pulse oximetry and supplemental oxygen through nasal prongs should be used routinely, as should ECG monitoring. Those patients who are difficult to stabilise should be endoscoped in the operating theatre with the surgeon present and airway protection, otherwise adverse events are likely and the outcome of endoscopic therapy will be compromised.1-4 Conscious sedation is often helpful, but in the elderly and the potentially unstable patient it is important that it is titrated gradually to get the patient co-operative and a non-moving target without deep sedation in which the airway may be compromised.1-4
The common sites of bleeding must be visualised – that is the distal oesophagus, the incisura of the stomach, the pre-pyloric region and the first part of the duodenum. The view may not be optimal due to old or fresh blood and clots when they are present. These may prove difficult for the suction channel to cope with. In such situations it is important to have an experienced endoscopist and to carefully move past the blood rather than try to suction it away and block the endoscope. Automatic irrigation devices may be helpful to displace clots and get a better view of the bleeding site.1 , 12
The ulcers can be
classified into different categories according to the Forrest
classification (Fig. 2). This defines the risk of rebleeding
and is the determinant of which ulcers should have endotherapy
to either control the bleeding or reduce the risk of it
Fig. 2. Forrest classification of peptic ulcer bleeding related to risks of rebleeding. (NBVV = non-bleeding visible vessel.)
There are various endotherapies that have been employed to control the bleeding endoscopically. For non-variceal haemorrhage the tried and tested modality is injection with 1/10 000 solution of adrenaline.1 , 4 , 9 Use at least 10 ml. This will cause intense vasoconstriction and narrowing of the vessel lumen by a sheer volume effect. This will allow localisation of the bleeding point.
It has been shown that more lasting control can be achieved by adding one of the following two techniques – thermal coaption with a heater probe or gold probe.1 , 4 , 9 , 12 This is placed directly on the visible vessel to obliterate the bleeding vessel’s lumen by pressure and direct thermal coagulation. Alternatively, use endoscopic clipping devices which pass through the accessory channel of the endoscope and deploy a detachable clip to the vessel.1 , 4 , 8 , 9 , 12
For the majority of patients these methods should control even active bleeding 80 - 90% of the time. Duodenal ulcers are often more difficult technically than gastric ulcers as the endoscope manoeuvrability is limited in the confined space of a scarred duodenum. If there is failure to control active bleeding, if facilities and logistics allow and the patient is stable then interventional angiography may control the situation by embolotherapy with a variety of different agents. Coils are used to control the major inflow and outflow followed by chemical obliteration of the microcirculation around the area to ensure that more permanent control is obtained.13
These techniques are usually employed in the high-risk elderly patient with co-morbidities in whom surgery carries a prohibitive risk. However, there remains a place for prompt surgery as the only hope for salvage, particularly in those patients with persistent shock.7
Once controlled, the IV PPI should be continued for 48 hours9 , 10 and the ulcer then effectively managed by curative oral eradication therapy, of which the current first-line therapy is a 2-week course of amoxycillin and clarithromycin with a bd double-dose PPI.14
In patients with
variceal bleeding, the bleeding often stops in those with good
liver reserve. Once again the risk stratification allows
prognostication and the risk of rebleeding in hospital. Though
the Rockall score has been validated for variceal haemorrhage the Child-Pugh score15 (Table 3) is
most commonly used to stratify risk.
At endoscopy a full
inspection of the foregut must be undertaken. This will allow
identification of non-variceal causes of haemorrhage, which
occur in 10% of cases, and those who may be bleeding from
gastric varices or portal hypertensive gastropathy.3
16 The method of choice is banding, as shown in
Fig. 3. This has largely superseded injection sclerotherapy
with ethanolamine oleate, as it has
a higher complication rate than banding.16 Once again, those who continue to bleed are
best managed with airway control prior to endoscopy in the
operating theatre complex or the ICU. The Sengstaken-Blakemore
tube can be used when visualisation of the oesophagus is not
possible due to active bleeding. In this situation the gastric
balloon should be blown up and traction applied.17 This is usually effective but should be
followed by immediate endoscopy once deflated and removed.
Alternatively, a covered 30 mm oesophageal stent has been used as an
alternative form of tamponade18 but the interventional procedure of choice
is a TIPS shunt19 which, when placed successfully, will reduce
the portal pressure below 12 mmHg and control the bleeding. Once bleeding
has been controlled, appropriate supportive medical management
should be continued. Those patients with Child-Pugh category C
disease often have been precipitated into hepatic failure by
the bleed and require additional management for this
complication. These individuals are also likely to have a
prolonged INR and require repeated transfusion of fresh frozen
or freeze dried plasma.3
Fig. 3(A). Superiorly band, showing deployed width of 2 mm. Inferiorly band loaded on banding device width 9 mm. (B). Endoscopic view of deployed band.
Subsequent management is two-fold. Firstly, to reduce the risk of further bleeding by establishing non-selective B blockade and continued banding of the varices until they are eradicated.3 , 11 The second consideration is to establish the cause of the portal hypertension and manage it appropriately. Management at this stage is best handled by a multidisciplinary team comprising a gastroenterologist/hepatologist and a surgeon. These patients need an adaptable long-term strategy, which needs to be individualised and includes the option of liver transplantation where appropriate.3 , 11
1. ASGE Guideline: The role of endoscopy in the management of acute non-variceal upper GI bleeding. Gastrointest Endosc 2012;75:(6)1132-1138. [http://dx.doi.org/10.1016/j.gie.2012.02.033]
2. Bambha K, Kim WR, Pedersen R, et al. Predictors of early re-bleeding and mortality after acute variceal haemorrhage in patients with cirrhosis. Gut 2008;57:814-820. [http://dx.doi.org/10.1136/gut.2007.137489]
3. O’Brien J, Triantos C, Burroughs AK. Management of varices in patients with cirrhosis. Nat Rev Gastroenterol Hepatol 2013.(ahead of print).
4. Barkun AN, Bardou M, Kuipers EJ, et al. International consensus recommendations on the management of patients with non-variceal upper gastrointestinal bleeding. Ann Intern Med 2010;152:101-113. [http://dx.doi.org/10.7326/0003-4819-152-2-201001190-00009]
5. Rockall TA, Logan RFA, Devlin HB, et al. Variation in outcome after acute upper gastrointestinal haemorrhage. Lancet 1995;346:346-50. [http://dx.doi.org/10.1016/S0140-6736(95)92227-X]
6. Rockall TA, Logan RF, Devlin HB, et al. Steering Committee and Members of the National Audit of Acute Upper Gastrointestinal Haemorrhage. Incidence of and mortality from acute upper gastrointestinal haemorrhage in the United Kingdom. BMJ 1995;311:222-226. [http://dx.doi.org/10.1136/bmj.311.6999.222]
7. Bornman PC, Theodorou N, Shuttleworth RD, Essel HP, Marks IN. Importance of hypovolaemic shock and endoscopic signs in predicting recurrent haemorrhage from peptic ulceration: A prospective evaluation. BMJ 1985;291:246-248. [http://dx.doi.org/10.1136/bmj.291.6490.245]
8. Laine L, Peterson W. Bleeding peptic ulcer. N Engl J Med 1994;331:717–727. [http://dx.doi.org/10.1056/NEJM199409153311107]
9. Barkun A, Sabbah S, Enns R, et al. The Canadian Registry on Non-variceal Upper Gastrointestinal Bleeding and Endoscopy (RUGBE): Endoscopic hemostasis and proton pump inhibition are associated with improved outcomes in a real-life setting. Am J Gastroenterol 2004;99:1238-1246. [http://dx.doi.org/10.1111/j.1572-0241.2004.30272.x]
10. Sung JJ, Barkun A, Kuipers EJ, et al. Intravenous esomeprazole for prevention of recurrent peptic ulcer bleeding: A randomized trial. Ann Intern Med 2009;150:455-464. [http://dx.doi.org/10.7326/0003-4819-150-7-200904070-00105]
11. Krige JEJ, Beckingham IJ. Portal hypertension – 1: Varices. In: Beckinham I, ed. ABC of Liver, Pancreas and Gallbladder. London: British Medical Journal Publishing Group, 2001:18-21.
12. Lau JY, Sung JJ, Lam YH, et al. Endoscopic retreatment compared with surgery in patients with recurrent bleeding after initial endoscopic control of bleeding ulcers. N Engl J Med 1999;340:751-756. [http://dx.doi.org/10.1056/NEJM199903113401002]
13. Yap FY, Omene BO, Patel MN, et al. Transcatheter embolotherapy for gastrointestinal bleeding: A single center review of safety, efﬁcacy, and clinical outcomes. Dig Dis Sci 2013 (ahead of publication).
14. Malfertheiner P, The European Helicobacter Study Group (EHSG) Management of Helicobacter pylori infection the Maastricht IV/Florence Consensus Report. Gut 2012;61:641-664. [http://dx.doi.org/10.1136/gutjnl-2012-302084]
15. Pugh RN, Murray-Lyon IM, Dawson JL, Pietroni MC, Williams R. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg 1973;60(8):646-649. [http://dx.doi.org/10.1002/bjs.1800600817]
16. Krige JE, Shaw JM, Bornman PC. The evolving role of endoscopic treatment of esophageal varices. World J Surg 2005;29:966-973. [http://dx.doi.org/10.1007/s00268-005-0138-2]
17. McCormick PA, Burroughs AK, McIntyre N. How to insert a Sengstaken-Blakemore tube. Br J Hosp Med 1990;43:274-277.
18. Fabienne C, Kistlera FW, Stenzb V, Gublerc C. Treatment of esophageal variceal hemorrhage with self-expanding metal stents as a rescue maneuver in a Swiss multicentric cohort. Case Rep Gastroenterol 2013;7:97–105. [http://dx.doi.org/10.1159/000350192]
19. Garcia-Pagán JC, Pascoli MD, Caca K, et al. Early use of TIPS in patients with cirrhosis and variceal bleeding. N Engl J Med 2010;362:2370–2379. [http://dx.doi.org/10.1056/NEJMoa0910102]
Full text views: 3641