This article is more than 5 years old. Information may no longer be current.
Updates, Tips and Tricks in the Ever-Changing Landscape of Upper GI Bleeds
As technology improves and patients enter our care with more advanced bleeds and more co-morbidities than ever before, endoscopists must stay on top of the latest tips, tricks, technology and techniques to manage non-variceal upper gastrointestinal bleeding (UGIB).
These can be broken down via the three phases of endoscopy and come to the following, 1) Pre-endoscopy: determine risk and prognosis, understand when to transfuse blood products; 2) Endoscopy: know optimum timing for endoscopy, be aware of advances in endoscopic techniques, reduce rebleeding rates after endoscopic therapy; and 3) Post-endoscopy: when to appropriately restart antithrombotics.
Pre-Endoscopy: Risk Management
Before endoscopy is performed, physicians should be using risk stratification scores to identify low-risk UGIB patients who may be managed as outpatient and high-risk UGIB patients who need aggressive management. Many current guidelines recommend incorporation of risk scores, but routine clinical adoption has not yet occurred.
Of those recommended, the most widely used is the Glasgow-Blatchford Score (GBS). This is a score that includes several different categories, includes the blood urea nitrogen (BUN), hemoglobin, blood pressure, along with other markers of disease severity such as heart rate, cardiac and liver disease. Patients can score up to 23 and it can identify both low-risk and high-risk bleeders.
Patients with GBS scores of 0 to 1 are unlikely to have adverse events from the bleeding. These patients can often be managed less aggressively as outpatients. They can be discharged from the ED with scheduled follow-up in GI and an endoscopy within 48 to 72 hours. Many patients – 30% overall – who report to the ED can be managed in this manner and recent European guidelines recommended similar management for this population.
Alternatively, if patients have a GBS that is higher – 10 to 12 – they are at high risk for re-bleeding and mortality, and more likely to require an endoscopic intervention. Most recently it has been noted that the breakpoint for high-risk is 7. As a reminder, this is a dynamic score for which you use a different breakpoint depending on the outcome.
The other pre-endoscopy score widely studied in patients with UGIB is the AIMS65 score. AIMS65 stands for albumin less than 3.0 mg/dL, international normalized ratio (INR) greater than 1.5, mental status change, systolic BP less than 90 mm Hg or age greater than 65 years. It is an easy to remember score because each of these categories is 1 point where if a patient is at 2 or more, he or she is at high risk for mortality from GI bleeding.
Several different studies have shown that AIMS65 effectively predicts high risk but not low risk. In my clinical practice, I use both AIMS65 and GBS. I calculate both and correlate them with my clinical judgment to make triage and timing of endoscopy decisions.
Pre-Endoscopy: Blood Products
Basic IV fluid resuscitation is extremely important in the initial management of patients with UGIB, but blood products use has changed and overuse of blood products should be avoided.
Take packed red blood cells for example. Currently, the recommendations have a goal hemoglobin at 7 g/dL where previously the goal was 9 or 10 g/dL (and many physicians who trained in earlier eras may not be aware of this significant change in goals). This restrictive strategy improves overall survival in patients with UGIB. Too much blood results in worse outcomes for our patients with both variceal and non-variceal bleeding.
Next, we need to be aware of how to manage patients properly who have an elevated INR due to warfarin. We see many patients on antithrombotic medications with UGIB – often with elevated INRs and there are two concepts endoscopists must understand.
First, it is safe and effective to perform an endoscopy, including bleeding therapy, with an INR up to 2.5. You should not delay endoscopy in these patients.
Second, if you need to reverse the INR to get it to 2.5 or less or in someone with ongoing active bleeding, the recent recommendations all suggest using prothrombin complex concentrate (PCC). This induces rapid, low-volume and effective reversal of the INR while avoiding prolonged or hyper coagulable states, which can happen with IV vitamin K and be problematic with valvular heart disease.
Lastly, platelet transfusions can be considered for UGIB patients with a platelet count of less than 50,000. In the past, we have also considered platelet transfusions for patients who were on dual anti-platelet drugs, particularly the combination of aspirin and a thienopyridine drug such as clopidogrel. Recently, a study compared outcomes of patients on these medications with a normal platelet count who received platelet transfusion and found no benefit. We no longer recommend giving platelet transfusions in UGIB patients on dual antiplatelet agents.
Endoscopy Timing
Current guidelines all suggest that we perform endoscopy within 24 hours but, in fact, some patients do not need endoscopy within 24 hours. Those patients with a GBS of 0 to 1 have a risk level low enough to have endoscopy safely up to 72 hours after diagnosis.
But those who have higher GBS or AIMS65 scores and ongoing active bleeding would benefit most from early intervention with an early endoscopy. The challenge is to determine who needs it and when within that 24-hour window most improves their outcomes.
Patients who are initially hemodynamically unstable and are suspected to have ongoing bleeding will benefit most from early endoscopy. The other key here is adequate resuscitation.
We conducted a study last year that looked at BUN change as a predictor of outcome. Patients with increasing BUN at 24 hours had a worse outcome – particularly in mortality – a finding likely due to under resuscitation.
Two studies last year looked at mortality and timing of endoscopy. A study from Denmark showed among those patients who are sicker and have American Society of Anesthesiologists (ASA) physical statuses of 3 to 5, there was a U-shaped curve. Here, an endoscopy too early had a higher mortality as did an endoscopy too late. The other study from the Brigham and Women’s Hospital also showed urgent endoscopy is a predictor of worse outcomes in select patients with acute nonvariceal UGIB.
There was a “sweet spot” for optimal timing of endoscopy that was after resuscitation, but not too long afterward. The ideal time for endoscopy in most patients is 10 to 24 hours after admission.
Endoscopic Therapies
Currently, the standard of care in endoscopic therapy is either the use of combination therapy with injection and cautery methods or the use of endoscopic clips. However, there is still room to improve because re-bleeding occurs in 5% to 15% patients. There are concepts that can help reduce this.
There are newer over-the-scope clips (OTSC) and these have been studied both for primary control bleeding and in re-bleeding and they seem to be highly effective. There may be lesions where these larger clips should be used as first-line therapy, such as the large arteries including the gastroduodenal artery, the left gastric artery and in Dieulafoy’s lesion.
Coagulation grasping forceps that apply monopolar cautery are commonly used for people undergoing large endoscopic resection such as endoscopic submucosal bisection. The artery is grasped with the forceps and pulled toward the endoscopist and a soft coagulation current applied for 1 or 2 seconds. Studies looking at the use of monopolar cautery in patients with upper GI bleeding show it’s an effective mechanism. In my practice, I use this as a rescue method rather than a primary method, and this is an effective and easy modality to employ.
Source: John R. Saltzman, MD, FACP, FACG, FASGE, AGAF
In Europe and elsewhere, a hemostatic spray (Hemospray, Cook Medical) is used. It is more than 90% effective at stopping bleeding. Although it is approved in Europe, it’s not available in the U.S. just yet, although I am hoping that it’s available within the year.
What is good about this therapy is that it’s easy to deploy. It does not require special expertise, just that the endoscopist approach the bleeding lesion and hit a button that releases the spray. The disadvantage lies in the likely need for a second modality to prevent re-bleeding. The spray stops the bleeding but does not apply a durable therapy at the artery. It also seems to be effective in malignant bleeding and oozing, in which we do not have other good modalities.
Post-Endoscopy: Re-Bleeds
OTSCs are an effective option for patients who have re-bleeding, as they have a low rate of associated rebleeding. If a lesion is amenable to these clips, an OTSC is a good choice.
Additionally, there has been recent work looking at the role of Doppler probes. A Doppler probe can go down the instrument channel of the scope and interrogate the site of the bleeding lesion to see if there is an arterial signal. If there is an arterial signal after an endoscopic therapy has been applied, there is a significant risk for re-bleeding. However, if there is no arterial flow when you place the probe, the risk is dramatically reduced.
In a randomized controlled trial in a very high-risk group of UGIB patients reported last year, when these Doppler probes were employed, the re-bleeding reduced from 26% to 11%. These are a promising modality and can be used to help determine the endpoint of therapy. I personally use this probe in very high-risk patients such as those where I am unsure if I have applied adequate therapy or those with multiple comorbidities. These probes are also useful in lesions where you are not sure the location of the original bleed. They can help determine if a location has arteries or not. Although currently available, Doppler probes are not widely used.
Lastly, there is also an old concept that’s come back, which is called second-look endoscopy. This means routinely going back on the day after endoscopic therapy and performing another endoscopy. Overall, since we are using effective therapy, this is not needed. However, a recent study showed in high-risk patients where the endoscopist is unsure effective therapy was applied, there is a value to go back on the second day and apply additional therapy if there is still persistent high-risk stigmata of bleeding, such as a visible vessel.
Post-Endoscopy: Antithrombotics
The final concept I want to emphasize is the return of patients to antithrombotic medication as soon as possible.
If a patient is on these medications for secondary prevention, it is clear that withholding these drugs puts patients at risk for additional thrombotic events. A recent study looked at patients with peptic ulcer disease and on aspirin therapy after having a cardiovascular event. After these patients underwent endoscopic therapy for a bleeding ulcer, the researchers randomly assigned them to go back on the aspirin or receive placebo and followed them for 30 days.
When patients went back on the aspirin, there was twice as much bleeding, which is expected. However, the mortality of the patients back on aspirin was one-ninth the mortality of patients on placebo. Obviously, aspirin is extremely important in this group of patients and we would much rather have a patient come back with a re-bleed that we can almost always treat rather than have them die for lack of these medications.
The same concept is true for warfarin and other blood thinners. If you withhold them, you will reduce the re-bleed rate but increase the thrombotic rates, which range from cardiovascular event to a neurological event such as stroke. It is key to resume the antithrombotic medication at the right time.
My own clinical tactic is if I do an endoscopy on a patient with an UGIB and I find no stigmata of recent hemorrhage, I get the patients right back on their therapy. However, if I do endoscopic therapy, I will hold these medications for about 72 hours after their endoscopic therapy. They resume their antithrombotic medication typically on day 4 of the hospitalization.
Take Home
Endoscopists should be aware of the many recent advances as they manage patients with UGIB. Before endoscopy, use a risk stratification score and properly resuscitate, but do not over transfuse your patients. Perform endoscopy after adequate resuscitation and within 24 hours in all patients. When considering therapies, standard endoscopic therapies include combination therapies and cautery or endoscopic clips. To prevent re-bleeding, over-the-scope-clips are useful for large vessels and the use of Doppler probes may provide more reliable end-points for endoscopic therapy. Lastly, restart antithrombotics as soon as possible within 1 to 4 days.
- References:
- Acosta RD, Abraham NS, ASGE SOP. Gastrointest Endosc. 2016;83:3-16.
- Arima S. J Gastroenterol. 2010;45(5):501-505.
- Barkun A. Ann Intern Med. 2010;152:101-113.
- Barkun A. Gastrointest Endosc. 2013;77:692-700.
- Blatchford O. Lancet. 2000;356:1318-1321.
- Chen Y. Gastrointest Endoscopy Clin N Am. 2015;25:535-552.
- Elmunzer BJ. Am J Gastroenterol. 2008;103:2625-2632.
- Garg SK. Endoscopy International Open. 2017;05:E376-E386.
- Gralnek IM. Endoscopy. 2015;47:a1–a46.
- Jensen DM. Gastroenterology. 2017;152(6):1310-1318.
- Jensen DM. Gastrointest Endosc. 2016;83(1):129-136.
- Kirschniak A. Gastrointest Endosc. 2007;66:162-167.
- Kumar NL. Gastrointest Endosc. 2017;85(5):945-952.
- Kumar NL. Gastrointest Endosc. 2017;86(6):1022-1027.
- Laine L, Jensen D. Am J Gastroenterol. 2012;107:345-360.
- Laursen SB. Gastrointest Endosc. 2017;85(5):936-944.
- Maggio D. Can J Gastroenterol. 2013;27:454-458.
- Manno M. Surg Endosc. 2016:30(5):2026-2029.
- Manta R. Surg Endosc. 2013;27:3162-3164.
- Nayor J, Saltzman JR. Gastrointest Endosc. 2016;83(1):137-139.
- Nunoue T. J Clin Gastroenterol. 2015;49(6):472-476.
- Robertson M. Gastrointest Endosc. 2016;83:1151-1160.
- Saltzman JR. Gastrointest Endosc. 2011;74(6):1225-1229.
- Sarin N. Can J Gastroenterol. 2009;23:489-493.
- Srygley FD. JAMA. 2012;307(10):1072-1079.
- Sung JJ. Ann Intern Med. 2010;152:1-9.
- Tai CM. Am J Emerg Med. 2007;25:273-278.
- Targownik LE. Can J Gastroenterol. 2007;21:425-429.
- Tsoi KKF. Nat Rev Gastroenterol Hepatol. 2009;6:463-469.
- Villanueva C. N Engl J Med. 2013;368(1):11-21.
- Wedi E. Surg Endosc. 2017;31(1):307-314.
- Witt DM. Arch Intern Med. 2012;172:1484-1491.
- Wolf AT. Am J Gastroenterol. 2007;102:290-296.
- Zakko L. Clin Gastroenterol Hepatol. 2017;15(1):46-52.
- For more information:
- John R. Saltzman, MD, FACP, FACG, FASGE, AGAF, can be reached at Brigham and Women’s Hospital, Division of Gastroenterology, Hepatology and Endoscopy, 75 Francis St., Boston MA 02115; email: jsaltzman@bwh.harvard.edu.
Disclosure: Saltzman reports no relevant financial disclosures.