72. Diagnostics and therapy of gastrointestinal bleeding (incl. variceal bleeding)

Page created on October 15, 2021. Last updated on April 18, 2022 at 15:06

Gastrointestinal bleeding

Introduction

Gastrointestinal bleeding can range from occult (no symptoms, only found on occult blood test) to severe and life-threatening. It may manifest as haematemesis, haemodynamic instability, melena, or haematochezia.

Acute GI bleeding requires hospitalisation and urgent assessment and treatment. Mortality is high, 5 – 20%. Upper GI bleeding is 5x more frequent than lower GI bleeding.

Etiology

Upper GI bleeding (mostly haematemesis):

  • Variceal bleeding
    • Due to portal hypertension, mostly due to cirrhosis
  • Non-variceal upper GI bleeding
    • Peptic ulcer
    • Mallory-Weiss syndrome
    • Oesophagitis
    • Erosive gastritis

Melena:

  • Peptic ulcer
  • Oesophagitis
  • Gastritis
  • Oesophageal varices
  • Angiodysplasia

Haematochezia:

  • Diverticulosis (most common cause overall)
  • Inflammatory bowel disease
  • Infectious colitis (inflammatory diarrhoea)
  • Colorectal cancer
  • Angiodysplasia

Clinical features

The first evaluation should assess for features of severe bleeding and haemodynamic instability.

  • Signs of severe but partially compensated bleeding:
    • Postural hypotension
    • Tachycardia
    • Signs of vasoconstriction
  • Signs of haemorrhagic shock
    • Pulse rate > systolic blood pressure

Haematemesis refers to vomiting of fresh blood, clotted blood, or coffee grounds-like material. Haematochezia refers to fresh or clotted blood per rectum and is typically a sign of lower GI tract bleeding, while melena refers to passage of black tarry stool and is typically a sign of upper GI tract bleeding. However, large (> 1L) upper GI bleeding may cause haematochezia as well.

Haematemesis mostly occurs in large bleedings.

Diagnosis and evaluation

In case upper GI bleeding is suspected, due to haematemesis or signs of haemodynamic instability, urgent upper endoscopy is indicated. Colonoscopy is the investigation of choice for haematochezia. Both colonoscopy and upper endoscopy allow for both diagnosis and, in some cases, treatment of the bleeding.

Obtaining the patient’s bleeding parameters, medication list, and previous history are important to determine the source. As examples, anticoagulants can be reversed, known cirrhosis increases the likelihood of the source being oesophageal varices, and so on.

Aspiration of gastric contents through a nasogastric tube allows for quick differentiation between upper and lower GI bleeding. The aspirate is bloody in case of upper bleeding and clear in case of lower.

The Blatchford scoring system may be used to assess the risk of an upper GI bleeding requiring intervention. It’s based on vital parameters, comorbidities, lab tests, and clinical features. It’s used to identify patients who are low-risk (score ≤1) and can be treated outpatient.

Treatment

In case of acute bleeding, urgent assessment and stabilisation are crucial. If the patient is haemodynamically unstable they must be stabilised first. This involves monitoring, replacing lost fluids, and blood transfusion. Blood transfusion is indicated at haemoglobin level < 70 g/L, with a target of 70 – 90 g/L.

Then, we should determine the source of the bleeding, stop it, treat the underlying condition, and prevent recurrent bleeding.

Variceal bleeding

Introduction and epidemiology

Bleeding from oesophageal varices is a severe complication of portal hypertension. It’s a common complication of cirrhosis, affecting up to 60% of patients with decompensated cirrhosis.

1/3 of patients with cirrhosis develop variceal bleeding in the first 2 years. The mortality rate is very high, up to 40%.

Clinical features

Variceal bleeding presents as sudden severe upper GI bleeding in a patient with known liver disease or long-standing risk factors for liver disease.

Treatment

Variceal upper GI bleeding is severe and rarely stops spontaneously. They may also rebleed.

Pharmacological therapy of variceal bleeding includes vasoactive drugs, which either inhibit release of vasodilative hormones or induce vasoconstriction directly. In either case, the splanchnic blood flow is reduced, which decreases the bleeding. The first line is terlipressin, second line somatostatin or octreotide. All of these drugs decrease bleeding, but only terlipressin decreases mortality.

Antibiotic prophylaxis with ceftriaxone for 7 days is also indicated due to high risk of infections.

Balloon tamponade may be used to stop the bleeding temporarily. This is only a short-term solution as rebleeding often occurs once the balloon is deflated and removed.

Definite treatment of variceal bleeding is either band ligation or sclerotherapy. These should be combined with pharmacological therapy (terlipressin).

Band ligation is generally the first choice of the two. It involves endoscopically placing small elastic bands around varices. This is similar to the procedure used for haemorrhoids.

Sclerotherapy involves endoscopic injection of a sclerosing solution into the varices.

Primary prevention

All patients who receive the diagnosis of cirrhosis should undergo upper GI endoscopy to look for oesophageal varices.

  • No varices
    • -> repeat procedure in 2 years if cirrhosis is compensated
    • -> repeat procedure in 1 year if cirrhosis is decompensated
  • Grade I varices
    • -> repeat procedure in 1 year
  • Grade II or III varices
    • -> propranolol as prophylaxis
    • If intolerant to propranolol, prophylactic band ligation

Secondary prevention

After one episode of oesophageal bleeding, interventions to prevent rebleeding are indicated. These include both:

  • Pharmacological therapy with propranolol or carvedilol
  • Band ligation
    • Repeated every 1 – 4 weeks until varices are eradicated

If bleeding occurs despite the aforementioned measures, placement of a transjugular intrahepatic portosystemic shunt (TIPS) is indicated.

Non-variceal upper GI bleeding

Introduction and epidemiology

Non-variceal upper GI bleeding is also a common and high mortality, up to 15%. 70% of cases stop bleeding spontaneously, while the remaining either rebleed in 1 – 3 days or continuously bleed. These bleedings most frequently originate from ulcers.

Etiology

The most important risk factors for ulcer bleeding are:

  • NSAID and aspirin use
  • H. pylori infection
  • Alcohol
  • Anticoagulants and antiplatelets

Classification

Peptic ulcers are classified according to the Forrest classification. It’s based on the ulcer’s endoscopic morphology and is used to guide whether patients require inpatient care or not.

  • Stage I – active haemorrhage
  • Stage II – evidence of recent haemorrhage
  • Stage III – clean-based ulcer

The risk of recurring haemorrhage is highest in stage I and lowest in stage III. Forrest I – IIb is high risk and require inpatient treatment, while Forrest IIc – III is low risk and can be treated outpatient.

Treatment

Upper endoscopy is important to evaluate the morphology of the ulcer. If there is active bleeding, a non-bleeding visible vessel, or an adherent clot to the ulcer, endoscopic therapy with IV PPI is indicated. If there is a flat spot or clean based ulcer, no endoscopic therapy is necessary, only oral PPI. Endoscopic haemostasis may be achieved with:

  • Electrocoagulation
  • Heat
  • Laser
  • Clipping
  • Banding
  • Injection of adrenaline, sclerosing agents, or fibrin glue

Surgical therapy is indicated if endoscopic therapy fails, if the ulcer is large (> 2 cm), if there is perforation, or if the bleeding is so large that the patient requires many units of blood (4 – 6). However, > 90% of gastrointestinal bleedings are managed without surgery. For surgical therapy, see topic B16 of surgery-traumatology.

3 thoughts on “72. Diagnostics and therapy of gastrointestinal bleeding (incl. variceal bleeding)”

  1. Hey!
    Just a minor correction. The Blatchford score isn’t used to assess the risk of bleeding. It is used in patients who are already bleeding to assess whether or not they will need medical intervention.
    Thanks for all you do.

Leave a Reply to Nikolas Cancel reply

Inputting your name is optional. All comments are anonymous.