Peptic ulcer is one of the most common causes of upper gastrointestinal bleeding (UGIB) and often the most common cause of hospitalization for non-variceal bleeding. The author has recently read the Guidelines for the Management of Patients with Bleeding Ulcers issued by the American College of Gastroenterology (ACG) and is deeply impressed by articles 13-19 related to endoscopic treatment, which are translated for the reference of medical professionals.
What kind of patients need endoscopic treatment?
1. Those with jet bleeding, active bleeding, and visible exposed vessels should receive endoscopic treatment (highly recommended High-quality evidence) (Figure 1)
2. Patients with adherent blood clots that cannot be removed by aggressive irrigation may be considered for endoscopic treatment. Patients with potentially high bleeding risk clinical features (e.g., elderly, concurrent disease, inpatients with concurrent bleeding) benefit (conditional recommendation, low level of evidence).
3. Patients with a clean ulcer base or a black base do not need to undergo endoscopic treatment (strong recommendation, high quality evidence).
Summary of the evidence
The results of the meta-analysis of ulcerated patients with active bleeding (jet and active oozing) showed that the likelihood of further bleeding was significantly reduced with endoscopic treatment compared with no endoscopic treatment (RR = 0.29,0.29 – 0.43) and the NNT (number of cases required to get one more effective treatment) was only 2. The number of cases requiring urgent intervention and surgery was also significantly reduced. A meta-analysis of ulcer patients with visible exposed vessels also showed that endoscopic treatment not only significantly reduced the chance of further bleeding (RR = 0.49, 0.40 C 0.59; NNT = 5), but also reduced the need for emergency intervention and surgery.
Although jet bleeding and bleeding were combined in the randomized study and meta-analysis described above, the chance of further bleeding was lower in patients with bleeding. Nevertheless, the fact that 39% of patients with oozing bleeding still rebleed after conservative treatment provides strong support for endoscopic treatment. The results of endoscopic treatment are better in patients with high-risk signs of bleeding. In a randomized cohort study applying high-dose PPI with placebo control, the rebleeding rate after 72 h of endoscopic treatment was lower in the oozing group (4.9%) than in the jet bleeding group (22.5%), the clot group (17.7%), and the bare vessel group (11.3%).
A randomized meta-analysis showed that patients with clot adherence did not benefit significantly from endoscopic treatment (RR = 0.31, 0.06C1.77). There was significant variability across studies in this meta-analysis. Two US studies reported a significant advantage of endoscopic hemostatic treatment, with a rebleeding rate of 3 vs. 35% for the mixed sample compared with drug treatment. The other studies from Europe and Asia did not mention the superiority of endoscopic treatment.
One of these studies followed the now recommended treatment (aggressive endoscopic irrigation, endoscopic treatment followed by continuous intravenous infusion after the first dose of PPI in high doses) and found no rebleeding in the control group of 24 patients with clot adhesions treated with medication only. The reasons for the apparently inconsistent results of the above studies are not clear to us, and possible explanations are the different severity of coexisting comorbidities (the US study was mainly from tertiary referral centers), the different etiology of ulcer disease (H. pylori ulcers are very common except in the US), and the different efficacy of treatment with PPIs (higher in Asian H. pylori-positive patients).
Patients with a clean ulcer base or a black base rarely rebleed (45) and so will not benefit from endoscopic treatment.
What should be done for endoscopic treatment?
1. Epinephrine injection therapy should not be used alone. If used, it should be combined with a second hemostatic therapy (strongly recommended, high quality evidence).
2. Thermal coagulation therapy with bipolar electrocoagulation or thermal probe or injection of tissue sclerosing agent (e.g. anhydrous ethanol) is recommended to stop bleeding. These methods of hemostasis may reduce further bleeding, the need for surgery, and reduce mortality (strongly recommended, high quality evidence).
3. The application of titanium clips to stop bleeding is recommended because they reduce the possibility of further bleeding and surgical procedures. However, the application of titanium clips alone has various results compared with other treatments, for which there is a lack of good studies (conditional recommendation, low quality evidence).
4. In a subset of patients with active bleeding ulcers, thermal coagulation or epinephrine injection with other hemostatic methods can sometimes provide better initial hemostasis than titanium clips and sclerotherapy alone (conditional recommendation, low- to moderate-quality evidence).
Summary of Evidence
The initial study termination point in the UGIB study was to prevent further active bleeding patients and to prevent rebleeding in patients with initial hemostasis and in patients without active bleeding symptoms. Endoscopic hemostatic therapy showed good efficacy in randomized studies with thermal coagulation therapy (bipolar electrocoagulation, thermal probe coagulation, monopolar electrocoagulation, argon ion coagulation, laser), injectable therapy (epinephrine, sclerosing agents such as anhydrous ethanol, ethoxysclerol, and ethanolamine), thrombin or fibrin glue (thrombin plus fibrinogen), and titanium clips.
A randomized study found that endoscopic injection of epinephrine was effective in patients with active bleeding to achieve initial hemostasis, but there was no significant difference when compared with other endoscopic hemostatic methods. A meta-analysis of treatment with epinephrine injection alone and other hemostatic methods alone (bipolar electrocoagulation, titanium clips, and fibrin glue were applied as controls for the other three methods) and surgical treatment found that epinephrine injection alone was less effective in preventing further bleeding (RR = 1.72, 1.08C2.78; NNT = 9).
In addition, epinephrine plus another endoscopic treatment (e.g., bipolar electrocoagulation, sclerotherapy, and titanium clips) was significantly more effective (RR = 0.34, 0.23C0.50; NNT = 5) than epinephrine alone in reducing further bleeding and surgical risk. However, if a second endoscopy is performed to repeat treatment of high-risk lesions, there is no significant superiority of endoscopic combination therapy compared with epinephrine alone.
A meta-analysis of 15 randomized trials found that the application of bipolar electrocoagulation, with thermal probe coagulation treatment, significantly achieved initial hemostasis and reduced the likelihood of further bleeding (RR = 0.44, 0.36 C 0.54; NNT = 4), surgical risk, and mortality (RR = 0.58, 0.34C0.98; NNT = 33) than without endoscopic treatment. No significant differences have been found between the two thermal coagulation treatments. The term “multipolar electrocoagulation” has been used in some studies. Both multipolar and bipolar thermal probes are able to stop bleeding by bipolar electrocautery, and the difference in the names of the two probes is only in the structure of the thermal probe tip. Therefore, most meta-analyses have combined multipolar and bipolar thermal probes.
Two smaller studies suggest that endoscopic epinephrine injection combined with bipolar electrocoagulation is more effective than bipolar electrocoagulation alone, which is less effective when compared with other studies. A large sample of high-quality studies found that injectable thrombin combined with thermocoagulation was not more effective than thermocoagulation alone. Thus, although the limited findings suggest that the application of epinephrine injection after thermal coagulation therapy is more effective than thermal coagulation therapy alone, there is insufficient information to suggest that just as monopolar thermal coagulation therapy alone should not be used, so should other thermal coagulation devices.
However, once specific signs of recent hemorrhage (SRH) are present, some people empirically preemptively inject epinephrine before applying other endoscopic treatments. It is claimed that epinephrine injection for active bleeding will slow or even stop the bleeding, thus improving the endoscopic view for the next treatment. In addition, an advance injection of epinephrine for adherent clots that are still not removed by aggressive irrigation may reduce severe bleeding during clot removal if removal is required.
Three randomized meta-analysis studies on anhydrous ethanol found that endoscopic sclerotherapy also significantly reduced further bleeding (RR = 0.56, 0.38C0.83; NNT = 5), surgical risk, and mortality when compared with controls not treated endoscopically. Because sclerotherapy can cause mucosal tissue necrosis after sclerotherapy injection, the volume of sclerosing agent must be strictly controlled, and sclerotherapy alone is not the best choice for active bleeding.
When comparing patients with active bleeding treated with anhydrous ethanol sclerotherapy with those without any treatment, the anhydrous ethanol group could achieve a hemostasis rate of 46% compared with only 8% in the control group (64). In patients with active bleeding, it is reasonable to give epinephrine injections before applying sclerotherapy, but there are no randomized studies in which sclerotherapy alone was applied as a control.
In a comparative study of thermal coagulation versus sclerotherapy, the study did not find significant differences in further bleeding, surgery, and mortality between the two, despite less need for acute interventions (surgery, repeat endoscopic therapy, interventional radiation) and less tendency for further bleeding after thermal coagulation (RR = 0.69, 0.47C1.01) (64).
Titanium clips, although not yet compared with controls without endoscopic treatment, were able to reduce the rate of further bleeding and surgery when compared with endoscopic epinephrine injection and water injection. When compared with other standard treatments (thermal coagulation or sclerotherapy, with or without epinephrine), titanium clips were less effective in stopping bleeding than thermal coagulation (thermal probe), but there was no significant difference for other treatment outcomes such as further bleeding. However, there were differences in these studies, for example, one study found titanium clips to be effective, while the other two studies found that titanium clip therapy was poor at preventing further bleeding when comparing control treatments.
Therefore we need more information on the application of titanium clips alone in the management of acute UGIB. One reason for these differences is due to different endoscopists and the other is the different models of titanium clips. The newer the titanium clip the easier it is to apply, and titanium clips come in different sizes, hardness, depth of mucosa clamped, and retention time after clamping; however, this was not well observed in randomized studies. Another theoretical advantage of titanium clamping is that it does not cause tissue damage, unlike thermal coagulation therapy and sclerotherapy, and is therefore preferred by patients receiving antithrombotic therapy and retreatment for rebleeding.
We do not recommend the application of laser, monopolar electrocoagulation, argon ion coagulation, and thrombin injection, fibrin glue injection as first-line treatment, although these treatments have shown some effectiveness in randomized studies because of the lack of strong evidence, the potential slightly higher risk and side effects, the ease of application and/or the price (64).
Endoscopic hemostatic treatment techniques Endoscopic hemostasis is usually applied at the site of bleeding to force it to stop, blocking or occluding the vessels at the base of the ulcer in the vicinity of the SRH thus preventing rebleeding. In two studies of patients treated for clot adhesion, after removal of the clot (e.g., biopsy forceps, manipulator forceps, endoscopic tip) and application of thermal coagulation therapy, epinephrine was applied endoscopically to the four corners of the ulcer, and epinephrine saline solution (1:10,000 or 1:20,000 in saline) was typically injected 0.5 C 2 ml per spot into the bleeding ulcer base signs within and around the mucosa.
Although good results have been reported with high-dose epinephrine injections alone (e.g., 30C45 ml), no studies have yet suggested the ideal dose of epinephrine needed in combination with other treatments. We recommend continuous endoscopic injection until active bleeding slows or stops, or there are no signs of mucosal bleeding, at the base of the ulcer and the four corners immediately adjacent to the SRH.
Anhydrous ethanol is usually injected at 0.1C0.2 ml per site because excessive doses can lead to tissue damage, so the dosage is limited to 1C2 ml . 5% ethanolamine is applied at 0.5C1.0 ml per site, and the dosage for the treatment of ulcer bleeding has been reported to vary widely from 0.5C14 ml in randomized studies.
Bipolar electrocoagulation should be performed so that the end of the inner lens is as close as possible to the bleeding ulcer, with good application of the largest probe (3.2 mm) placed in the largest area of the cut, or at an angle to maintain a certain ratio of compression/maximum pressure. The recommended mode of electrocoagulation is ~ 15 W , 8C10 s. Multiple electrocoagulations are performed at the base of the ulcer and around the SRH until the bleeding stops, the vessels are flattened and the base turns white. The recommended thermal probes are all uniform mode of 30 J. Titanium clips need to be clamped to the bleeding site to close the underlying artery on both sides of the SRH.