Variceal endoscopic hemostasis


31
Variceal endoscopic hemostasis


Patrick McKiernan, Lauren Johanson, and Mike Thomson


Portal hypertension and variceal formation


Portal hypertension is defined as a hepatic venous pressure gradient (HVPG) of >5 mmHg. It is caused by increased portal blood flow and/or increased portal resistance. In children, portal hypertension is predominantly caused by intrahepatic disease, but important presinusoidal causes include extrahepatic portal vein obstruction (EHPVO) or congenital hepatic fibrosis [1]. Where HVPG is >10 mmHg esophageal varices may form, and bleeding may occur when the HVPG is >12 mmHg. Although mortality following the first bleed is low when specialist medical care is available (<1%) [2], recurrent variceal hemorrhage has a mortality of up to 8% [3]. Variceal bleeding may precipitate the need for earlier liver transplantation.


Diagnosis, classification, and risk stratification of varices


Upper gastrointestinal endoscopy remains the reference for diagnosis of esophageal varices (Figure 31.1). There are no evidence‐based pediatric guidelines for the management of portal hypertension and hence the timing of the first endoscopy is debatable. In our unit, the development of splenomegaly is an indication of clinically significant portal hypertension. At endoscopy, the size and distribution of varices in addition to red color changes should be recorded.

Photos depict (a and b) large esophageal varices before and after band ligation.

Figure 31.1 Large esophageal varices before and after band ligation.


Table 31.1 Classification system for esophageal varices












Grade I Small and nontortuous varices
Grade II Tortuous varices but occupying <1/3 of the distal esophageal radius
Grade III Large and tortuous varices covering >1/3 of the distal esophageal radius

Currently, a three‐grade classification system (Table 31.1) is used in the UK and Europe. This has been shown to have reasonable interobserver repeatability [4].


A retrospective study has shown that high‐risk varices in children can be considered as one of the following patterns: grade 3 esophageal varices, grade 2 esophageal varices with red color changes and/or gastric varices along the cardia [5].


Primary prophylaxis


There are limited studies addressing primary prophylaxis of variceal hemorrhage in children. There has been one randomized controlled trial to date, which used sclerotherapy [6]. Duché et al. demonstrated that primary prophylaxis by a skilled operator was safe in 36 children with high‐risk varices due to biliary atresia [7], and a number of uncontrolled studies have shown that endoscopic variceal ligation (EVL) is safe and effective. In 2015, a group of experts could not recommend routine primary prophylaxis in children because of a lack of definitive proof of efficacy and safety of current techniques [8].


Acute bleeding


Initial management should be according to an institutional protocol and usually involves treatment with octreotide and antibiotics. Therapeutic endoscopy should be performed under general anesthetic, as soon as possible once the patient is hemodynamically stable. ESGE‐ESPGHAN guidance is that endoscopy should occur within 12 hours of an acute upper gastrointestinal bleed if the child requires ongoing circulatory support, if they have known esophageal varices, or if a large hematemesis or melena has occurred [9]. If bleeding continues, the procedure should be carried out immediately. If a bleeding varix is identified, this should be treated first. Where no bleeding point is found, treatment should usually be commenced, so long as no other potential site for bleeding exists and varices are high risk.


Endoscopic sclerotherapy is well established and the choice of sclerosant appears unimportant, with 5% ethanolamine oleate used widely. The injection is carried out via the endoscopic operating channel using single‐use injectors. Endoscopes with working channels down to 2.2 mm may be used. Injections are started just above the gastroesophageal junction and may be intra‐ or paravariceal. All varices at this level should be treated, with 1–3 mL of sclerosant being used per injection, then repeated if necessary to all varices 3–5 cm caudally. No more than 10–15 mL of sclerosant should be used per session, depending on patient size. Varices more than 5 cm from the gastroesophageal junction should not be treated unless actively bleeding.


Endoscopic ultrasound (EUS) may help to target sclerotherapy with histoacryl glue in the gastric fundus (Figure 31.2). It affords a real‐time image of the gastrointestinal wall and feeding blood vessels, treatment of which may reduce variceal recurrence. After sclerotherapy, Doppler may be used to confirm variceal ablation.


Sclerotherapy is very effective, with control of acute bleeding in 90% of cases. Complications are more likely following emergency sclerotherapy, occurring in approximately 18% [10], and include esophageal ulceration, bleeding, mediastinitis, esophageal perforation, chylothorax, pneumothorax, and stricture formation. There is a procedural mortality of approximately 1%.

Photos depict (a) EUS varices preinjection. (b) EUS varices with needle.

Figure 31.2 (a) EUS varices preinjection. (b) EUS varices with needle.


Band ligation (EVL) was first reported in humans in 1989. In this technique, a hollow cylinder with prestretched rubber rings is attached to the front of an endoscope. The variceal column is directly sucked into the hollow cylinder and the band is released around the base of the varix (Figure 31.3). Strangulation of the varix stops acute bleeding. Over the next few days, ischemic necrosis of the mucosa and submucosa develops and the rings are sloughed, leaving shallow mucosal ulceration. Epithelialization occurs within 2–3 weeks and the submucosal vascular layers are replaced by maturing scar tissue by eight weeks [11].


An initial control endoscopy is carried out to recognize bleeding points and document the distance to the gastroesophageal junction to prevent inadvertent gastric banding. The apparatus is loaded onto the endoscope, ensuring a secure fit, and then repassed as far as the gastroesophageal junction. The bander is maneuvered to the most distal variceal column and suction applied (see Figure 31.3). When the varix fills the cylinder, the tripwire is pulled and the endoscope pulled back. Care must be taken, especially in small children, that the esophageal wall is not aspirated into the cylinder. Subsequent bands are applied in a proximal direction to variceal columns in the lower 5–6 cm of the esophagus. Up to 10 bands can be used per session but we usually restrict this to four, as more is often uncomfortable. Some manufacturers insert a colured band to remind the operator when a single band remains. EVL is much better tolerated than sclerotherapy [12]; 10–15% of children complain of transient retrosternal pain but esophageal stricture has not been reported in the pediatric literature.

Photo depicts varices being banded.

Figure 31.3 Varices being banded.


Where feasible, EVL is the preferred technique for managing acute bleeding. A major advantage is that once the bleeding point is ligated, bleeding control with improved visibility is immediate. EVL can be used in children <10 kg so long as the loaded bander can be passed through the cricopharynx. However, it is not possible to pass the banding equipment in all children, so sclerotherapy should continue to be available.


Following endoscopic therapy, patients should fast for at least two hours and solid feeding should be withheld until liquids are tolerated. Sucralfate should be given for five days as this appears to decrease the risk of early rebleeding [13]. Proton pump inhibitors and antibiotics probably do not affect rebleeding rate. Clipping may be used occasionally but is not routinely recommended [14]. Post‐Kasai variceal banding is reported [15].


Hemospray®


Hemostatic spray is a simple technique in which powder is sprayed through the endoscope onto actively bleeding lesions where it forms a mechanical hemostatic barrier. This has been effective in the treatment of diffuse portal hypertensive gastropathy. It has also been applied directly to actively bleeding varices with good short‐term effect. While not a definitive treatment, this appears safe and easy to administer, particularly if experienced interventional endoscopists are not immediately available [16] (see Chapter 30).


Self‐expanding metal stents


This technique has recently been described in adult practice for resistant bleeding. The stent is deployed in the lower esophagus with or without direct endoscopic visualization and provides immediate hemostasis while allowing oral intake [17]. The stent can be left in place for 14 days, to facilitate definitive treatment, and subsequently removed endoscopically. There is no reported pediatric experience with the technique yet, but it would appear to be a useful addition to our therapeutic options.


Secondary prophylaxis


Following a variceal bleed in children, rebleeding rates are as high as 80% [18,19], hence all should receive secondary prophylaxis. Where the underlying liver disease is compensated, this will often be endoscopic but many techniques are available.


Where feasible, EVL is the preferred endoscopic method and is safer and more effective than sclerotherapy [12,20,22]. EVL should be undertaken every 2–4 weeks until varices are ablated. Subsequent follow‐up endoscopies are recommended at 6–12‐monthly intervals with recurrent varices being ablated where possible. EVL can safely be undertaken as a day case.


Gastric varices


The Sarin classification is used to describe the location of gastric varices [23]. Type 1 gastroesophageal (GEV) varices are continuous with esophageal varices, extending for 2–5 cm below the gastroesophageal junction along the lesser curve; type 2 varices extend beyond the gastroesophageal junction towards the greater curve into the fundus (Figure 31.4). Type 1 isolated gastric varices (IGV) are gastric fundal varices, whereas type 2 IGV are ectopic varices occurring elsewhere in the stomach or the first part of the duodenum.


Although gastric varices are less likely to bleed than esophageal varices, the mortality and rebleeding rates are higher where they do [24]. Type 2 GEV bleed most frequently and the risk of death is greatest from IGV bleeds [25].


Conventional endoscopic treatment with banding or sclerotherapy is not effective. In adults, endoscopic cyanoacrylate (CYA) injection is first‐line treatment for the acute and chronic management of bleeding gastric varices [26,27]. CYA is a tissue adhesive agent that rapidly hardens upon contact with blood and intravariceal injection causes rapid variceal occlusion (see Figure 31.2).


Prior to the procedure, the endoscopic tip is coated with silicone oil and the instrument channel flushed with oil to reduce the risk of glue adherence and endoscope damage. The injection needle is primed with either sterile water or saline, depending on the product. CYA is injected directly into the varix in 0.5–1 mL aliquots. The injection time depends on the agent used and whether it has been diluted. Traditionally, probing with a blunt‐tipped instrument was used to confirm variceal obliteration. EUS, where available, is now the preferred method (see Figure 31.2). A case study of 21 children with active gastric variceal bleeding showed primary hemostasis of 96% [28]. CYA may be successfully used in infants weighing less than 10 kg [29].

Photo depicts varices with EUS probe.

Figure 31.4 Varices with EUS probe.


The major limitation of injecting this glue is not embolization or local complications for the child, but damage to the biopsy channel of the endoscope itself. If any glue leaks from the injecting endoscopic needle (which is the same as a standard injecting needle) then this will be blocked, thus rendering the endoscope totally unusable. This dictates that the operator must employ great care when performing this procedure. Hence the tip of the catheter is not removed through the biopsy channel and is cut with scissors before removal to prevent the glue accidentally being extruded into the channel on catheter removal (Figure 31.5).


Acute injection of thrombin into gastric varices is described in small series but is generally not common practice. In adult patients who have undergone thrombin injections, hemostasis in the acute setting has been successful in up to 92% of cases. Patients usually receive 1–4 sessions of thrombin, with a mean total dose of approximately 10 mL for variceal eradication. Dry thrombin for reconstitution can be kept in the endoscopy unit fridge for acute use and technique is identical to other injection techniques.

Photo depicts cutting the catheter after glue injection prior to removal to protect the endoscope biopsy channel.

Figure 31.5 Cutting the catheter after glue injection prior to removal to protect the endoscope biopsy channel.

Dec 15, 2022 | Posted by in GASTROENTEROLOGY | Comments Off on Variceal endoscopic hemostasis

Full access? Get Clinical Tree

Get Clinical Tree app for offline access