Irrigation during endoscopy is important not only to identify the bleeding site, but also to prepare the target tissue for intervention. Irrigation of suspected bleeding lesions is done with water or saline directed onto the lesion until adequate visualization is achieved. Most current endoscopes include ports and accessories to deliver a forward-directed water jet through the tip of the endoscope, allowing for easy and copious directed delivery of irrigating fluid. When significant air bubbles are present, addition of simethicone to the irrigating fluid in the water bottle can aide in visualization. Irrigation of an ulcer should not be deferred out of fear for provoking bleeding. Irrigation can help to identify the precise location of the SRH and can provide the operator information needed to determine which endoscopic tool or approach is optimal.

If a large volume of blood or clot is present, a standard endoscope may not provide enough suction capacity to clear the area. In these cases, a large channel or dual-channel therapeutic endoscope can be used to facilitate more effective suctioning of fluid and/or clots. A large capacity external suction device can also be attached to the biopsy port of an endoscope for patients with large amounts of intraluminal contents requiring aspiration bypassing the suction inside of the endoscope housing connected to the endoscope processor and allowing for more effective suctioning.

Several pharmacologic agents have been used in UGIB to help clear the stomach contents in an effort to facilitate improved visualization during endoscopy. Intravenous erythromycin used prior to endoscopy can help with gastric visualization. Erythromycin is a motilin-like prokinetic agent, promoting gastric contractions and subsequent gastric emptying. Erythromycin at 250 mg bolus or 3 mg/kg infusion administered over 30 min (intravenously) is effective clinically and should be administered 30–120 min prior to the anticipated endoscopy. Metoclopramide 10 mg IV has also been used as a prokinetic agent to promote clearance of gastric debris and blood from the stomach prior to endoscopy, although there is less available data concerning its efficacy. These prokinetic agents improve gastric visualization and potentially reduce the need for repeat endoscopy [26].

Injection therapy is primarily performed with dilute epinephrine, although saline alone may be used if epinephrine is not available [27]. Injection therapies induce hemostasis by producing a tamponade effect on the area and epinephrine causes vasoconstriction reducing local blood flow, although this effect is less than the primary fluid tamponade. Epinephrine is generally diluted to 1:10,000. This may be done by adding 1 mL of 1:1000 epinephrine into a syringe containing 9 mL of saline. The concentration can be further reduced in patients with serious cardiac comorbidities to 1:100,000, especially when used near the gastroesophageal junction where its use may cause more systemic cardiac effects. Patients with contraindications to epinephrine can receive saline alone as the injectate. Saline alone can be used as an injectant to produce tamponade if epinephrine is not available.

The general technique of injection is to introduce a standard endoscopic injection catheter through the working channel of the endoscope until the tip is visible. Then, the injection needle is advanced and locked in position. Different injection needles with different diameters allow for variability in the amount of force needed to introduce the needle into the submucosa. It is recommended to inject into multiple locations surrounding the bleeding lesion with a four-quadrant technique, although some lesions with active bleeding may achieve hemostasis after only one injection. If the needle is not inserted deep enough to reach the submucosal space, injected fluid will leak into the lumen when the syringe is depressed. If this happens, one can readjust the needle by pulling the injection catheter out of the mucosa and then reinserting at the same or a different location in an effort to reach the submucosa with the needle tip. Of note, some pressure is required to introduce the injection needle tip into the submucosa. Alternatively, the injection can be started with the tip of the needle in the lumen and the probe advanced during injection to find the submucosal space.

Injection can be used to help control bleeding of various etiologies, including from vessels within ulcers, vascular malformations, and Dieulafoy’s lesions, as well as when a discrete lesion is not visualized due to active bleeding. In the stomach, 8–10 mL total injection can be used in multiple injections of about 2 mL each, although there is no absolute number or volume of injections. Higher doses are more likely to cause cardiovascular side effects, and this should be kept in mind, especially when epinephrine is used near the gastroesophageal junction.

Soon following the injection, the area around the lesion will develop pallor and the hemostatic effects are seen when any active bleeding slows or stops. Use of injection monotherapy is not recommended because it is less effective than other monotherapies or combination therapies and is less durable as it is associated with higher rates of rebleeding. Rather, injection therapy is often used as a prelude to a second treatment (thermal or mechanical therapy) once bleeding has abated and visualization has been improved.

Injection therapy can be used to treat a myriad of bleeding lesions because of the short-lived effect of vasoconstriction and temporary cessation of bleeding. In situations where there is overwhelming bleeding obscuring visualization despite irrigation, injection of epinephrine can help slow bleeding and ultimately identify and thus allow treatment of the source. Injection therapy is most helpful to slow or stop bleeding; thermal and/or mechanical therapy can subsequently be applied to achieve complete and durable hemostasis.

Other injection agents include sclerosants (although these are usually used in the treatment of bleeding varices if band ligation has failed or is not available). These agents can be used for nonvariceal bleeding sources as well. Sclerosants include ethanol, polidocanol, and ethanolamine. These agents induce local inflammation and fibrosis of a bleeding vessel. Other agents that can be injected include cyanoacrylate and fibrin glues. Cyanoacrylate glue is a liquid material that transforms (polymerizes) into a solid after injection. This can be particularly useful in bleeding gastric varices, whereby the glue becomes an artificial thrombus in the varix reducing blood flow by occluding the vessel. Fibrin glue has also been used endoscopically as a form of injection therapy. Fibrinogen and factor XIII are mixed with thrombin and calcium. In this manner, the clotting cascade is activated and clotting is promoted. Cyanoacrylate and fibrin glue may be of limited availability.

Contact thermal therapies used for the control of UGIB include heater and bipolar probes, as well as monopolar therapies [27]. The technique for the use of heater and bipolar therapies involves controlling bleeding by simultaneously compressing and cauterizing a bleeding vessel, known as coaptive coagulation. Heater and bipolar therapies do not require a grounding pad applied to the patient as the electrical circuit is completed within the device itself. Once in widespread use, heater probes rarely are utilized in current practice and bipolar electrocautery devices are now commonly employed to treat GI bleeding. Heater probes are quite effective and, if available, are still an excellent choice for treating upper GI bleeds.

While the bipolar probe can be used multi-directionally, with either a perpendicular or tangential approach, a heater probe can be used perpendicularly only. The larger 10 French probes can deliver thermal energy over a larger area than the smaller 7 French probes and are felt to be more effective, although the 10 French probes require a therapeutic endoscope with a large channel size.

The technique of applying endoscopic cautery using either a bipolar or heater probe is similar for both devices. First, the probe is advanced out of the tip of the endoscope and into the lumen for a short distance, so that the tip is visualized endoscopically (Video 2.1). If the probe is too far out of the scope, it can be difficult to control and the operator will lose mechanical advantage. The endoscope should be positioned as close to the lesion as possible for the control of therapy to maximize visualization and efficiency of endoscopic maneuvers. The probe should make direct contact with the bleeding vessel and be held in place with direct pressure to ensure continued contact. Moderate-to-firm pressure is usually used in the stomach due to its relatively thick wall, and mild-to-moderate pressure is typically used in the rest of the gastrointestinal tract, such as in the small bowel or esophagus where the walls are thinner. One suggested technique uses four to six pulses of energy for approximately 10 s each, although many variations exist. No gold standard on the number and duration of therapy exists as these depend on the specifics of the lesion being treated and on its location. The endpoint for therapy is the cessation of bleeding and visible cauterization of the target lesion, which often appears flattened following successful therapy application.

Energy levels recommended are 10–15 W in the duodenum and 15–20 W in the stomach. After each round of therapy, the target lesion can be inspected for ongoing bleeding, adverse events, and the need for additional therapy. If the probe is stuck to the vessel, removal of the probe can sometimes trigger rebleeding. Some devices allow water irrigation directly through the probe to minimize the risk of the probe adhering to the coagulum and can be irrigated after each application of cautery.

Monopolar therapy can also treat bleeding vessels and has been extensively used to treat endoscopically induced bleeding, such as bleeding occurring during endoscopic submucosal dissection (ESD). However, there is much less clinical data available on the use of monopolar cautery for the control of acute UGIB. One monopolar probe is a rotatable probe with flat jaws (Coagrasper, Olympus Corporation, Center Valley, PA), used to capture and compress tissue while delivering thermal energy. The technique involved with this device is different from the coaptive coagulation technique used for the heater and bipolar probes. By using monopolar forceps, the bleeding lesion is grasped and “tented” toward the scope. Cautery is used at higher power settings, such as 50 W, for shorter durations of 1–2 s [28]. Monopolar cautery requires the use of a grounding pad, similar to that used for polypectomy. This pad should not be placed over any implanted metal, such as a joint replacement.

Argon plasma coagulation (APC) is a noncontact, superficial method of thermal therapy that induces destruction of bleeding lesions or aberrant vessels and vascular malformations. APC uses argon gas that is electrically conducted creating a high-energy plasma. It is unlike heater, bipolar, or monopolar probe therapies, as it does not touch or compress the tissue targeted for therapy. APC uses monopolar energy, and a grounding pad must be placed on the patient prior to use. The probes for APC are available in a variety of configurations, including with those tips that are end-firing and circumferential. Cautery will seek the closest mucosal surface to the probe, regardless of probe type. APC has, in general, a lesser degree of tissue penetration than other hemostatic methods.

APC technique involves passing the probe carefully through the endoscope as the probe can easily kink and advancing the probe close to the target tissue. The probe needs to be only a few millimeters away from the target but should not make contact with the mucosa. Contact of the probe may cause dissection of charged argon gas through the wall and result in perforation. Pulses of argon gas and ionization charge are controlled by a foot pedal. The lesion may be “sprayed” or “painted” with the goal being adequate treatment of the tissue, with a white charring of the superficial layer of the mucosa (Fig. 2.3).

APC can be used in the treatment of bleeding from vascular malformations, radiation-induced rectal bleeding, or gastric antral vascular ectasia (GAVE). In a lesion that spreads, such as GAVE, APC is effective in treating a large involved area. The tissue is sprayed as the APC probe or endoscope is moved along a lesion and large amount of mucosa can therefore be quickly treated. Repeated treatments are often required for cases of GAVE, with typically three to four treatment sessions required depending on the extent of GAVE. The tip of the APC probe will collect charred material if there is contact with the tissue. The probe should then be removed from the endoscope removing the charred material with gauze, following which the probe may be reintroduced through the endoscope for resuming treatment.