Upper GI Hemorrhage

Chapter 28


Upper GI Hemorrhage


Brad S. Moffat, Sarah Knowles and Neil G. Parry


Overview


Gastrointestinal (GI) bleeding can be a very challenging problem in the critically ill patient. Unlike many other sources of bleeding, the GI tract source is not readily compressible. It is not surprising, therefore, that patients who present in hemorrhagic shock or require ICU admission from upper GI bleeding have a mortality rate of up to 30%.1, 2


GI bleeding is a unique issue in the ICU patient because it can not only be the primary reason for ICU admission, but it can also present as a common complication in critically ill patients. This chapter will discuss both the patient who presents with massive upper GI hemorrhage requiring ICU admission, as well as the patient already admitted to ICU who develops an acute upper GI bleed as a complication of their stay. While the pathogenesis differs, the management is essentially the same. In this chapter, special aspects of Upper GI hemorrhage are presented to enhance the intensivist’s capability in managing these cases.


Introduction


GI bleeding is classified by the anatomic source of hemorrhage. Upper GI bleeding refers to a source in the foregut, proximal to the ligament of Treitz. The three main structures in the foregut are the esophagus, stomach, and duodenum; Each having its own blood supply and causes for bleeding.


Bleeding from the upper GI tract can present along a spectrum from very slow, intermittent bleeding with mild anemia, to brisk life-threatening exsanguination. Most are managed medically or endoscopically by gastroenterology and/or general surgery. A small minority will present with massive bleeding requiring ICU admission.


Patients with hemodynamic derangements secondary to GI bleeding should be admitted to the ICU for close monitoring and aggressive resuscitation. Early involvement of a surgeon and a gastroenterologist is mandatory.


Epidemiology


The incidence of GI bleeding requiring hospitalization, in the USA, is approximately 170 cases per 100,000 adults per year and accounts for 1–2% of all acute admissions.3 An upper GI source is the cause in 80% of cases.2 The majority of hospitalized GI bleeds will stop spontaneously or return to normal after minimal transfusion. About 15% of cases will continue to bleed despite initial resuscitative measures and may require admission to ICU.4


Admission to the ICU is also an independent risk factor for developing an upper GI bleed. Various factors contribute to a process known as stress ulceration in the stomach. Since the advent of acid suppression therapy, the overall risk of upper GI bleeding secondary to stress ulcers is relatively low at about 1 in 1,000. However, ventilation for greater than 48 hours with concomitant coagulopathy increases that risk to 3%, even with acid suppression therapy.5, 6


Risk Factors


Several factors have been associated with increased risk of upper GI bleeding and certain risk factors predispose specific sites to bleeding. The most pervasive risk factor in the general population is the use of non-steroidal anti-inflammatory medications. They inhibit endogenous prostaglandins, primarily in the stomach, which predispose the gastric mucosa to acid injury and consequent ulceration and bleeding. The risk increases substantially with prolonged usage and at high doses.


Helicobacter pylori infection is another common risk factor. H. pylori infection significantly increases the risk for ulceration thereby increasing the risk of associated bleeding as compared to the non-infected population. This tends to result in duodenal ulceration but can also affect the distal stomach (antrum).


Several scoring systems have been devised to predict the risk of rebleeding following intervention and the subsequent associated mortality. They have also been used to predict high risk patients who likely need ICU admission. The AIMS65 score uses five criteria (albumin, INR, mental status, hypotension, age > 65) to estimate mortality. This scoring system precedes the age of ubiquitous endoscopy and has been well validated.7 If all five risk factors are present, the mortality rate is 25%. More recent scoring systems, such as Rockall and Blatchford incorporate biochemistry, comorbidities, and endoscopic findings to predict mortality and rebleed rates.8


A less common but important risk factor for upper GI bleeding is gastroesophageal varices. They tend to form around the gastroesophageal junction and when they bleed, they can be very difficult to control. Variceal bleeding occurs in 25–40% of patients with cirrhosis.2 The variceal location, size, and severity of liver disease are all predictors of bleeding risk.


Pathophysiology


Upper GI bleeding emanates from the foregut which consists of the esophagus, stomach, and duodenum. Each of these structures has unique physiology which can contribute to different patterns of bleeding (Table 1).



Esophagus


The most common source of major bleeding in the esophagus is variceal.2 Esophageal varices form at the lower end of the esophagus, typically within 5–10 centimeters of the gastro esophageal junction. While they can develop throughout the GI tract, they are most prone to bleeding in the esophagus due to their close proximity to the mucosal surface and repeated minor trauma (food boluses and acid reflux). Less common causes of esophageal bleeding include: esophageal tears (Mallory–Weiss tears), mucosal bleeding from acid reflux and neoplasms.


Stomach


The most common cause of bleeding in the stomach is peptic ulcer disease.9 A peptic ulcer forms when the normal mucosal physiology is disrupted and acid in the stomach is allowed to penetrate into the submucosal and muscular layers resulting in an ulcer. Peptic ulcers in the duodenum form more readily as it lacks the normal defenses of the stomach. However, peptic ulcers that form in the stomach are more prone to bleeding so the rates of bleeding from each are about equal accounting for up to 40% of all upper GI bleeding.2


NSAIDs inhibit prostaglandins, which play a key role in mucous secretion, and thus chronic NSAID use disrupts the normal protective layer of the stomach which predisposes to peptic ulceration.


Any insult which causes inflammation (e.g., H. Pylori) of the stomach can lead to peptic ulceration. H. pylori infection is implicated in 60% of gastric peptic ulcers.2 H. pylori colonize the mucosa below the mucous layer and lead to chronic inflammation and gastric ulceration. The chronic inflammation also leads to increased production of gastrin which also increases acid production. Other causes of bleeding from the stomach include: arterial or venous malformations (Dieulefoy lesion, gastric varices, gastric antral vascular ectasia) and neoplasms.


Duodenum


Like the stomach, the most common cause of bleeding in the duodenum is from peptic ulceration. The mucosa of the duodenum is more susceptible to acid injury than the stomach. Conditions which increase the acid load in the duodenum, such as chronic gastritis or H. pylori infection can easily lead to ulceration. H. pylori is responsible for approximately 90% of duodenal ulcers.


Due to its fixed location in the retroperitoneum, ulcers in the duodenum are at higher risk for eroding directly into surrounding blood vessels, most commonly, the gastroduodenal artery. Bleeding from this artery can be brisk and result in life threatening hemorrhage. Other causes of bleeding from the duodenum include: biliary or pancreatic sources (hemobilia, hemosuccus pancreaticus), aortoenteric fistula (a complication of aortic aneurysm repair) or neoplasms.


Stress ulceration


Stress ulceration is a unique entity and is considered separately from peptic ulcer disease. It is thought to be primarily the consequence of a low flow state to the stomach mucosa and is caused by an imbalance between increased acid production and impaired mucosal protection in the setting of reduced gastric blood flow, mucosal ischemia and reperfusion injury.


The systemic inflammatory response exhibited by critically ill patients appears to increase acid production and impair mucous secretion in the stomach. This was first observed in the 1970’s with specific patterns of stress ulcer formation among head injured (Cushing’s ulcers) and burn patients (Curling’s ulcers). The stomach lining also appears to be susceptible to mucosal ischemia during shock states and exhibits early mottling and microcirculatory insufficiency which further predisposes it to acid damage.


Acid suppression therapy is recommended for all patients admitted to the ICU and typically consists of H2 blockers or proton pump inhibitors. This gut prophylaxis has resulted in a dramatic decrease in stress ulcer complications, principally bleeding. Two major risk factors have repeatedly been shown to increase the risk of stress ulceration and bleeding in the ICU patient: Prolonged mechanical ventilation, and coagulopathy (Table 2). Patients with these risk factors should be closely monitored for signs of upper GI hemorrhage.


Clinical Presentation


The majority of patients presenting to hospital with GI bleeding can be managed on the ward. These patients may only exhibit clinical symptoms of anemia or overt signs of GI bleeding such as melena, blood per rectum, or hematemesis.



Table 2. Risk factors for stress ulcer bleeding.5,6































Risk Factor


Odds Ratio/Relative Risk of Bleeding


Respiratory Failure (Mechanical Ventilation > 48 hours)


OR = 15.6 (P < 0.001)


Coagulopathy


OR = 4.3 (P < 0.001)


Renal Failure


RR = 1.16 (P = 0.023)


Hypotension


3.7 ++


Sepsis


2.0 ++


Hepatic Failure


1.6 ++


Enteral Nutrition **


RR = 0.30 (P = 0.004)


Stress Ulcer Prophylaxis with Ranitidine **


RR = 0.39 (P = 0.024)


** Protective against stress ulcers.
++ Not statistically significant with multivariable regression.

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Apr 19, 2017 | Posted by in CRITICAL CARE | Comments Off on Upper GI Hemorrhage

Full access? Get Clinical Tree

Get Clinical Tree app for offline access