manejo de hemorragia aguda de la úlcera péptica

7/27/2019 Manejo de hemorragia aguda de la lcera pptica

1/10

review article

T h e n e w e n g l a n d j o u r n a l o fmedicine

n engl j med 359;9 www.nejm.org august 28, 2008928

Current Concepts

Management of Acute Bleedingfrom a Peptic Ulcer

Ian M. Gralnek, M.D., M.S.H.S., Alan N. Barkun, M.D., C.M., M.Sc.,and Marc Bardou, M.D., Ph.D.

From the Department of Gastroenterol-ogy and Gastrointestinal Outcomes Unit,Rambam Health Care Campus and Rappa-port Faculty of Medicine, Technion IsraelInstitute of Technology, Haifa, Israel(I.M.G.); the Divisions of Gastroenterol-

ogy and Clinical Epidemiology, McGill Uni-versity Health Centre, McGill University,Montreal (A.N.B.); and INSERM CentredInvestigations Cliniques Plurithmatique,Centre Hospitalier Universitaire du Bocageand Institut Fdratif de Recherche Sant,Sciences et Techniques de lInformationet de la Communication, Universit deBourgogne both in Dijon, France (M.B.).Address reprint requests to Dr. Gralnekat the Department of Gastroenterologyand Gastrointestinal Outcomes Unit, Ram-bam Health Care Campus, Bat Galim, Haifa31096, Israel, or at [email protected].

N Engl J Med 2008;359:928-37.Copyright 2008 Massachusetts Medical Society.

Acute upper gastrointestinal hemorrhage, which is defined asbleeding proximal to the ligament of Treitz, is a prevalent and clinicallysignificant condition with important implications for health care costs

worldwide. Negative outcomes include rebleeding and death, and many of thedeaths are associated with decompensation of coexisting medical conditions pre-

cipitated by the acute bleeding event.1 This review focuses specifically on the cur-rent treatment of patients with acute bleeding from a peptic ulcer.

Epidemiology

The annual rate of hospitalization for acute upper gastrointestinal hemorrhage inthe United States is estimated to be 160 hospital admissions per 100,000 popula-tion, which translates into more than 400,000 per year.2 In most settings, the vastmajority of acute episodes of upper gastrointestinal bleeding (80 to 90%) have non-variceal causes, with gastroduodenal peptic ulcer accounting for the majority of le-sions.3 A number of studies have suggested that the annual incidence of bleedingfrom a peptic ulcer may be decreasing worldwide,4 yet other recent population-based estimates have suggested that the incidence is about 60 per 100,000 popula-tion,5 with an increasing proportion of episodes related to the use of aspirin andnonsteroidal antiinflammatory medications. Moreover, peptic ulcer bleeding isseen predominantly among the elderly, with 68% of patients over the age of 60 yearsand 27% over the age of 80 years.6 Mortality associated with peptic ulcer bleedingremains high at 5 to 10%.1,3 Estimated direct medical costs for the in-hospital careof patients with bleeding from a peptic ulcer total more than $2 billion annually inthe United States.7

Clinical Presentation

Initial Management

Hematemesis and melena are the most common presenting signs of acute uppergastrointestinal hemorrhage. Melena is sometimes seen in patients with hemorrhagein the lower gastrointestinal tract (e.g., distal small bowel and colon) and hemato-chezia in patients with upper gastrointestinal hemorrhage.8 Appropriate hemody-namic assessment includes the careful measurement of pulse and blood pressure,including orthostatic changes, to estimate the intravascular volume status andguide resuscitative efforts. Patients who present with acute upper gastrointestinalbleeding and a substantial loss of intravascular volume have resting tachycardia(pulse, 100 beats per minute), hypotension (systolic blood pressure,


2/10

current concepts

n engl j med 359;9 www.nejm.org august 28, 2008 929

systolic blood pressure of 20 mm Hg on stand-ing).9,10 Mucous membranes, neck veins, andurine output should also be evaluated as addi-tional ways of estimating the intravascular vol-ume status.9

The first priority in treatment is correctingfluid losses and restoring hemodynamic stability.

Volume resuscitation should be initiated withcrystalloid intravenous fluids with the use oflarge-bore intravenous-access catheters (e.g., twoperipheral catheters of 16 to 18 gauge or a cen-tral catheter if peripheral access is not available).In order to maintain adequate oxygen-carryingcapacity, especially in older patients with coexist-ing cardiac illnesses, the use of supplemental oxy-gen and transfusion of plasma expanders withthe use of packed red cells should be consideredif tachycardia or hypotension is present or if thehemoglobin level is less than 10 g per deciliter.9,11

When indicated, correction of coagulopathy shouldbe undertaken.12

The insertion of a nasogastric tube may behelpful in the initial assessment of the patient(specifically, triage), although the incrementalinformation such a procedure provides remainscontroversial.10 It has been suggested that thepresence of red blood in the nasogastric aspirateis an adverse prognostic sign that may be usefulin identifying patients who require urgent endo-scopic evaluation.10,13 However, the absence ofbloody or coffee-ground material does not de-finitively rule out ongoing or recurrent bleeding,since approximately 15% of patients withoutbloody or coffee-ground material in nasogastricaspirates are found to have high-risk lesions onendoscopy.10 The use of a large-bore orogastrictube with gastric lavage (with the use of tap waterat room temperature) appears only to improvevisualization of the gastric fundus on endoscopyand has not been documented to improve the out-come.14 Intravenous erythromycin, through its ef-fect as a motilin receptor agonist, has been shown

to promote gastric motility and substantially im-prove visualization of the gastric mucosa on ini-tial endoscopy. However, erythromycin has notbeen shown to improve the diagnostic yield ofendoscopy substantially or to improve the out-come (Table 1).18

Patient Triage and Risk Stratification

With the use of clinical variables (i.e., before en-doscopy), scoring tools have been developed tofacilitate the triage of patients with acute upper

gastrointestinal hemorrhage, identify those inneed of urgent endoscopic evaluation, predict therisk of an adverse outcome, and assist in guidingtreatment.19,20 The Blatchford score, a validatedrisk-stratification tool based on clinical and lab-oratory variables, is used to predict the need formedical intervention in patients with upper gas-

trointestinal hemorrhage (Fig. 1A).16 The Blatch-ford scale ranges from 0 to 23, with higher scoresindicating higher risk. The Rockall score is prob-ably the most widely known risk-stratificationtool for upper gastrointestinal hemorrhage andhas been validated in numerous health care set-tings (Fig. 1B).17 The clinical Rockall score (i.e.,the score before endoscopy) is calculated solelyon the basis of clinical variables at the time ofpresentation. The complete Rockall score makesuse of both clinical and endoscopic criteria topredict the risks of rebleeding and death; the

scale ranges from 0 to 11 points, with higherscores indicating higher risk. The clinical Rock-all score and the Blatchford score are usefulprognostic tools in patients presenting withacute upper gastrointestinal hemorrhage, sincethe two tools have selected common features, in-cluding a determination of the patients hemody-namic status and coexisting illnesses, and mayreduce the need for urgent endoscopic evaluationin patients who are deemed to be at low risk.21Additional risk-stratification tools have been pro-posed.20 The use of such validated tools as ad-juncts to clinical evaluation and the judgment ofthe medical practitioner is encouraged in clinicalpractice.

The endoscopic appearance of a bleeding ul-cer can be used to predict the likelihood of re-current bleeding on the basis of the Forrestclassification, which ranges from IA to III. High-risk lesions include those characterized by activespurting of blood (grade IA) or oozing blood(grade IB), a nonbleeding visible vessel describedas a pigmented protuberance (grade IIA), and an

adherent clot (which is defined as a lesion that isred, maroon, or black and amorphous in textureand that cannot be dislodged by suction or force-ful water irrigation) (grade IIB) (Fig. 2A through2D). Low-risk lesions include flat, pigmentedspots (grade IIC) and clean-base ulcers (gradeIII) (Fig. 2E and 2F).8,22-24 The interobservervariation in diagnosing these endoscopic stigmatais low to moderate.25

At initial endoscopy, high-risk lesions areseen in approximately one third to one half of all

The New England Journal of Medicine

Downloaded from nejm.org on March 23, 2013. For personal use only. No other uses without permission.

Copyright 2008 Massachusetts Medical Society. All rights reserved.


3/10



Table 1. Management of Acute Bleeding from a Peptic Ulcer, According to Clinical Status and Endoscopic Findings.*

Clinical status

At presentation

Assess hemodynamic status (pulse and blood pressure, including orthostatic changes).

Obtain complete blood count, levels of electrolytes (including blood urea nitrogen and creatinine), international normalized ratio, bloodtype, and cross-match.

Initiate resuscitation (crystalloids and blood products, if indicated) and use of supplemental oxygen.Consider nasogastric-tube placement and aspiration; no role for occult-blood testing of aspirate.

Consider initiating treatment with an intravenous proton-pump inhibitor (80-mg bolus dose plus continuous infusion at 8 mg per hour)while awaiting early endoscopy; no role for H2 blocker.

Perform early endoscopy (within 24 hours after presentation).

Consider giving a single 250-mg intravenous dose of erythromycin 30 to 60 minutes before endoscopy.

Perform risk stratification; consider the use of a scoring tool (e.g., Blatchford score16 or clinical Rockall score17) before endoscopy.

At early endoscopy

Perform risk stratification; consider the use of a validated scoring tool (e.g., complete Rockall score17) after endoscopy.

Endoscopic findings

High-risk active bleeding or nonbleeding visible vessel (Forrest grade IA, IB, or IIA)

Perform endoscopic hemostasis using contact thermal therapy alone, mechanical therapy using clips, or epinephrine injection, followed bycontact thermal therapy or by injection of a second injectable agent. Epinephrine injection as definitive hemostasis therapy is not rec-

ommended. The endoscopist should use the most familiar hemostasis technique that can be applied to the identified ulcer stigma.Admit the patient to a monitored bed or ICU setting.

Treat with an intravenous proton-pump inhibitor (80-mg bolus dose plus continuous infusion at 8 mg per hour) for 72 hours after en-doscopic hemostasis; no role for H2 blocker, somatostatin, or octreotide.

Initiate oral intake of clear liquids 6 hours after endoscopy in patients with hemodynamic stability.9

Transition to oral proton-pump inhibitors after completion of intravenous therapy.

Perform testing for Helicobacter pylori; initiate treatment if the result is positive.

High-risk adherent clot (Forrest grade IIB)

Consider endoscopic removal of adherent clot, followed by endoscopic hemostasis (as described above) if underlying active bleedingor nonbleeding visible vessel is present.

Admit the patient to a monitored bed or ICU setting.

Treat with an intravenous proton-pump inhibitor (80-mg bolus dose plus continuous infusion at 8 mg per hour) for 72 hours afterendoscopy, regardless of whether endoscopic hemostasis was performed; no role for H2 blocker, somatostatin, or octreotide.

Initiate oral intake of clear liquids 6 hours after endoscopy in patients with hemodynamic stability.9

Transition to an oral proton-pump inhibitor after completion of intravenous therapy.

Perform testing for H. pylori; initiate treatment if the result is positive.

Low-risk flat, pigmented spot or clean base (Forrest grade IIC or III)

Do not perform endoscopic hemostasis.

Consider early hospital discharge after endoscopy if the patient has an otherwise low clinical risk and safe home environment.

Treat with an oral proton-pump inhibitor.

Initiate oral intake with a regular diet 6 hours after endoscopy in patients with hemodynamic stability.9

Perform testing for H. pylori; initiate treatment if the result is positive.

After endoscopy

If there is clinical evidence of ulcer rebleeding, repeat endoscopy with an attempt at endoscopic hemostasis,** obtain surgical or interven-tional radiologic consultation for selected patients.

For selected patients, discuss the need for ongoing use of NSAIDs, antiplatelet agents, and concomitant therapy with a gastroprotective agent.

* H2 blocker denotes histamine H2receptor antagonist, ICU intensive care unit, and NSAID nonsteroidal antiinflammatory drug. The use of high-dose intravenous proton-pump inhibitors while awaiting endoscopy does not appear to have an effect on outcomes for

patients, although it may be associated with a significant down-staging of endoscopic lesions. An epinephrine injection is defined as epinephrine mixed with normal saline at a ratio of 1:10,000. The use of a large, single-channel or double-channel endoscope is recommended; if contact thermal therapy is used, a large (10 French)

probe is recommended. Only omeprazole and pantoprazole have been assessed in clinical trials that were designed to reduce rates of ulcer rebleeding, surgery,

and death. This procedure is recommended only for endoscopists who are familiar with clot removal.15

** A preplanned, second-look endoscopy that is performed within 24 hours after the initial endoscopy is not recommended.





4/10

current concepts


patients,3 with rebleeding rates of 22 to 55% ifthe ulcer is left untreated endoscopically.8,22,24Additional data are needed to confirm the pos-sible improvement risk stratification provided byendoscopic Doppler ultrasonography applied di-rectly to the ulcer stigmata before and after en-doscopic hemostasis.26

Approach to Therapy

A multidisciplinary approach with timely involve-ment of a trained endoscopist and endoscopy as-sistant is widely recommended.9,19,27 Such involve-ment may entail after-hours availability, sinceearly endoscopy (performed within 24 hours af-ter presentation of the patient) is the cornerstoneof treatment for patients with acute upper gastro-intestinal hemorrhage and may improve certainoutcomes (the number of units of blood trans-

fused and the length of the hospital stay) for se-lected patients who are classif ied as being at highrisk. Early endoscopy also allows for the safe andexpedited discharge of patients who are classifiedas being at low risk and reduces the use of healthcare resources.19,28 Goals of early endoscopy areto determine the cause of bleeding, ascertain prog-nosis, and administer endoscopic therapy, if indi-cated. Treatment recommendations have focusedon the first 72 hours after presentation and en-doscopic evaluation and therapy, since this is theperiod when the risk of rebleeding is greatest(Table 1).24,29

Patients at High Risk

High-risk patients should be admitted to the hos-pital and should receive endoscopic therapy. Theyshould then be triaged to a monitored setting orintensive care unit for the first 24 hours of whatis usually at least a 3-day hospital stay.

Patients who have bleeding ulcers with high-risk stigmata as determined on endoscopy (activebleeding or a nonbleeding visible vessel) should

undergo endoscopic hemostasis, a procedure thathas been shown to decrease rates of rebleeding,the need for urgent surgery, and mortality.19,27,30Contemporary endoscopic treatments include in-jection therapy (e.g., saline, vasoconstrictors, scle-rosing agents, tissue adhesives, or a combinationthereof), thermal therapy (with the use of contactmethods, such as multipolar electrocoagulationand heater probe, or noncontact methods, suchas argon plasma coagulation), and mechanical

therapy (principally endoscopic clips). (For details,see the animation in the Supplementary Appen-dix, available with the full text of this article atwww.nejm.org.)

All methods of endoscopic hemostasis have

B Rockall Score

A Blatchford Score

At Presentation

Systolic blood pressure100109 mm Hg9099 mm Hg


5/10



been shown to be superior to no endoscopic in-tervention.19,27 Yet the addition of a second hemo-stasis approach (injectable, contact thermal ther-apy) to epinephrine injection (at a 1:10,000 ratioof epinephrine to normal saline) further reducesrebleeding rates, the need for surgery, and mor-tality,31 as compared with epinephrine injectionalone, which should be avoided.27,32 Although thesafety of injecting a sclerosant alone has beenquestioned, sclerosants only rarely cause serioustissue damage.33

A consensus statement recommends combi-nation therapy (epinephrine injection to providelocal vasoconstriction, volume tamponade, andfacilitation of a clear view of the bleeding vessel,followed by targeted contact thermal therapy),19but the superiority of combination therapy tothat of contact thermal therapy alone has been

questioned.32 Endoscopic therapy in the sub-group of high-risk patients who have an adher-ent clot has been advocated yet remains contro-versial (Table 1).15,19,34-36

The use of mechanical therapy, in particularendoscopic clips, has been qualified as promis-ing.19 The exact role of endoscopic clips remainsincompletely defined, but emerging data and pre-liminary pooled analyses suggest that clips aloneare similar to thermal therapy alone, a combina-tion of injection and contact thermal therapy,and clips followed by injection.32,37 These com-

parisons require further study. It may be that inthe future, the location and appearance of agiven bleeding lesion will determine the optimalmethod of endoscopic therapy. At present, it isprobably best for endoscopists to carry out thehemostasis technique they are most comfortableusing, since all methods have been shown to beefficacious. However, epinephrine injectionalone should not be performed. The exact rolesof newer and emerging endoscopic hemostasistechniques (including loops, cryotherapy, sutur-ing, and stapling devices) await appropriatelypowered clinical trials.

Various clinical and endoscopic factors havebeen proposed as predictors of failure of endo-scopic treatment in patients with bleeding froma peptic ulcer. These include a history of pepticulcer disease, previous ulcer bleeding, the pres-ence of shock at presentation, active bleedingduring endoscopy, large ulcers (>2 cm in diam-eter), a large underlying bleeding vessel (2 mmin diameter), and ulcers located on the lessercurve of the stomach or on the posterior or su-

perior duodenal bulb.38,39

Planned, second-look endoscopy that is per-formed within 24 hours after initial endoscopictherapy has not been recommended.19,27 Eventhough such a procedure was shown to be effica-cious in two meta-analyses,40,41 it provided onlya limited reduction in the rate of rebleeding. Also,the procedure may not be cost-effective whenmedical therapy leading to profound acid suppres-sion is used.42 Repeat endoscopy may be consid-ered on a case-by-case basis if there are clinical

l

A Spurting Blood B Oozing Blood

C Nonbleeding Visible Vessel D Adherent Clot

E Flat, Pigmented Spot F Clean Base

Figure 2. Endoscopic Stigmata of Bleeding Peptic Ulcer, Classified as High

Risk or Low Risk.

High-risks lesions are those that spurt blood (Forrest grade IA, Panel A),ooze blood (grade IB, Panel B), contain a nonbleeding visible vessel (grade IIA,Panel C), or have an adherent clot (grade IIB, Panel D). Low-risk lesions arethose that have a flat, pigmented spot (grade IIC, Panel E) or a clean base(grade III, Panel F).





6/10

current concepts


signs of recurrent bleeding or if there is uncer-tainty regarding the effectiveness of hemostasisduring the initial treatment.

Patients at Low Risk

A significant proportion of patients who are ad-mitted to the hospital with acute, nonvariceal up-

per gastrointestinal hemorrhage are at low risk forrebleeding and death.43-45 Results from random-ized and retrospective trials have shown that af-ter endoscopy, low-risk patients can be dischargedhome, depending on when the initial endoscopyis performed.46-50 A summary of selected conser-vative criteria for the care of low-risk patients ap-pears in Table 2.43,44,47-51 Low-risk patients whodo not fulfill these clinical criteria should be ad-mitted to the hospital for observation.

Medical Thera py

In the past 10 years, pharmacotherapy has fo-cused on the use of profound acid suppressionwith proton-pump inhibitors in the treatment ofpatients with nonvariceal upper gastrointestinalhemorrhage. Experimental data have shown thatgastric acid impairs clot formation, promotesplatelet disaggregation, and favors fibrinolysis.52Therefore, inhibiting gastric acid and raising theintragastric pH to 6 or more and maintaining itat that level may promote clot stability, thus de-creasing the likelihood of rebleeding. However,the goal of an intragastric pH of 6 or more istheoretical and has not been documented to be areliable proxy for clinical efficacy in the treat-ment of peptic-ulcer bleeding. Furthermore, al-though data from clinical trials support the useof a bolus followed by a continuous infusion ofproton-pump inhibitors, recent studies from NorthAmerica show that even a high-dose, continuousinfusion of proton-pump inhibitors may not sus-tain an intragastric pH of 6 or more.53 The use ofhistamine H

2receptor antagonists (H

2blockers)

in patients with peptic-ulcer bleeding has not re-sulted in a significant improvement in outcomes,54probably because of early development of phar-macologic tolerance.

Potent acid-suppressing proton-pump inhibi-tors do not induce tachyphylaxis and have hadfavorable clinical results.55 Recent meta-analysesshowed that the use of proton-pump inhibitorssignificantly decreased the risk of ulcer rebleed-ing (odds ratio, 0.40; 95% confidence interval[CI], 0.24 to 0.67), the need for urgent surgery

(odds ratio, 0.50; 95% CI, 0.33 to 0.76), and therisk of death (odds ratio, 0.53; 95% CI, 0.31 to0.91) (Fig. 3A and 3B),56,57 findings that have alsobeen confirmed in a real-world setting.3 How-ever, the reduction in mortality appears to occuronly among patients with high-risk stigmata whohave first undergone endoscopic therapy, a find-ing that supports the use of medical therapy asan adjunct to but not a replacement for endo-scopic hemostasis in such patients.34,57

There are only limited data from randomizedclinical trials in the United States that haveevaluated therapy with intravenous proton-pumpinhibitors for acute bleeding from a peptic ulcer.A recent study comparing the use of high-doseintravenous proton-pump inhibitors with that of

H2 blockers was halted prematurely because ofslow recruitment of subjects, and although therewas a trend favoring therapy with high-dose in-travenous proton-pump inhibitors, the study wasunderpowered to show any statistical differencebetween the two treatments.58 In addition, thedosing method for treatment with proton-pumpinhibitors appears to be important. A pooledanalysis of 16 randomized, controlled trials thatenrolled more than 3800 patients suggested thatintravenous bolus loading followed by continuous

Table 2. Proposed Selection Criteria for an Abbreviated Hospital Stayor Outpatient Treatment of Patients at Low Risk.*

Criteria

Age,


7/10



infusion of proton-pump inhibitors is more effec-tive than bolus dosing alone in decreasing therates of rebleeding and the need for surgery.59Therefore, it is reasonable to recommend the useof an intravenous bolus of proton-pump inhibitorsfollowed by a continuous infusion for 72 hoursafter endoscopic hemostasis, although contro-versy persists as to optimal dosing (Table 1). Theuse of high-dose intravenous proton-pump in-hibitors after endoscopic therapy has also beenshown to be more effective and less costly thanalternative approaches in a variety of clinicalsettings.60,61

The administration of high-dose intravenousproton-pump inhibitors while the patient isawaiting endoscopy does not appear to have aneffect on the outcome, even though its use may

be associated with a significant down-staging ofendoscopic lesions in other words, patientswho receive such treatment are less likely to haveendoscopic evidence of high-risk stigmata thanare patients who receive placebo (odds ratio, 0.67;95% CI, 0.54 to 0.84).62 Therefore, such patientsare less likely to need endoscopic hemostasistherapy (19.1% vs. 28.4%, P = 0.007).63 The cost-effectiveness of proton-pump inhibitors for thisindication remains somewhat controversial.64,65

The use of high-dose oral proton-pump in-

hibitors in peptic-ulcer bleeding has been shownin Asian populations to lead to reductions in therisk of rebleeding (odds ratio, 0.24; 95% CI, 0.16to 0.36), the need for surgery (odds ratio, 0.29;95% CI, 0.16 to 0.53), and the risk of death (oddsratio, 0.35; 95% CI, 0.16 to 0.74).66 However,these results may not be completely generaliz-able to North American or European popula-tions because of underlying differences in physi-ological measures, pharmacodynamic profiles(metabolism of proton-pump inhibitors throughthe cytochrome P-450 2C19 genetic polymor-phism), and prevalence rates ofHelicobacter pyloriinfection, factors that may favor the acid-sup-pressive effect of a given dose of a proton-pumpinhibitor in Asian patients.66 Additional datafrom randomized clinical trials comparing the

use of intravenous proton-pump inhibitors withthat of oral proton-pump inhibitors are requiredin Western patient populations, since high oraldoses could result in significant savings in healthcare resources.61

Somatostatin and its analogue, octreotide, in-hibit both acid and pepsin secretion while alsoreducing gastroduodenal mucosal blood flow.However, these drugs are not routinely recom-mended in patients with peptic-ulcer bleeding,since contemporary randomized, controlled trials

A

B

0.0 0.5 1.0 1.5

ControlBetter

Proton-Pump InhibitorBetter

Rebleeding

Surgery

Proton-Pump Inhibitor

no. of patients/no. of events

Control

1026/111

1081/38

1031/189

1103/71

0.0 0.5 1.5 2.01.0 2.5

ControlBetter

Proton-Pump InhibitorBetter

All patients

With endoscopic hemostasis

Without endoscopic hemostasis

Proton-Pump Inhibitor

no. of patients/no. of events

Control

Odds Ratio

Odds Ratio

1040/20

954/17

86/3

1062/38

969/32

93/6

l

Rebleeding and Surgery

Death

Group

Group

Figure 3. Effect of Proton-Pump Inhibition in Peptic-Ulcer Bleeding.Forrest plots show the efficacy of the use of proton-pump inhibitors in decreasing the rates of rebleeding and sur-gery (Panel A) and death (Panel B). Rates of death are shown for all patients and for those who have either under-gone endoscopic hemostasis or not undergone endoscopic hemostasis. The diamonds represent odds ratios (withthe size of the diamonds proportional to the number of patients), and the horizontal lines represent 95% confidenceintervals. Data are from Leontiadis et al.57





8/10

current concepts


have shown little or no benefit attributable tothem, either alone or in combination with an H

2

blocker.67 Furthermore, there are no strong datato support the adjunctive use of these drugs afterendoscopic therapy for ulcer bleeding.

Surgery and Interventional

Radiology

The drop in surgical rates to 6.5 to 7.5% has beensuggested by meta-analyses31 and national regis-try data,3 whereas epidemiologic studies havesuggested an increase in the population-basedannual incidence of emergency surgery from 5.2to 7.0 operations per 100,000 population between1987 and 1999.68 Because of a new understand-ing of peptic ulcer disease, the role of surgeryhas changed markedly within the past two de-cades and now obviates the need for routine ear-

ly surgical consultation in all patients presentingwith acute upper gastrointestinal hemorrhage.The aim of emergency surgery is no longer tocure the disease but rather to stop the hemor-rhage when endoscopic therapy is unavailable orhas failed. The role of early elective surgery is lessclear, as is the optimal surgical approach in pa-tients with acute disease.69,70 A recent cohortanalysis comparing vagotomy and drainage withvagotomy and resection procedures suggestedequivalent outcomes.71 Surgery remains an effec-tive and safe approach for treating selected pa-tients with uncontrolled bleeding (i.e., those inwhom hemodynamic stabilization cannot beachieved through intravascular volume replace-ment using crystalloid fluids or blood products)or patients who may not tolerate recurrent orworsening bleeding.72 For most patients withevidence of persistent ulcer bleeding or rebleed-ing, a second attempt at endoscopic hemostasisis often effective, may result in fewer complica-tions than surgery, and is the recommendedmanagement approach.3,73 Exceptions may include

patients with ulcers that are more than 2 cm indiameter and those who have hypotension asso-ciated with a rebleeding episode, since such pa-tients may be at increased risk for the failure ofrepeat endoscopic hemostasis.27,73

Angiography with transcatheter embolizationprovides a nonoperative option for patients inwhom a locus of acute bleeding has not been iden-tified or controlled by endoscopy. Agents such as

Gelfoam, polyvinyl alcohol, cyanoacrylic glues, andcoils are used to embolize bleeding lesions.74Primary rates of technical success range from 52to 94%, with recurrent bleeding requiring repeatedembolization procedures in approximately 10% ofpatients.75 In uncontrolled trials, successful trans-catheter embolization has been shown to signifi-

cantly reduce mortality in patients with uppergastrointestinal hemorrhage,74 although uncom-mon complications include bowel ischemia, sec-ondary duodenal stenosis, and gastric, hepatic, andsplenic infarction.75 In most institutions, radio-logic intervention is reserved for patients in whomendoscopic therapy has failed, especially if suchpatients are high-risk surgical candidates. A ret-rospective analysis showed no significant differ-ences between embolization therapy and surgeryin the incidence of recurrent bleeding (29.0% and23.1%, respectively), the need for additional sur-

gery (16.1% and 30.8%), and mortality (25.8% and20.5%), despite a more advanced age and higherprevalence of heart disease in the group receiv-ing embolization therapy.76 Although radiologicembolization may not always be a permanent cure,it may allow for the stabilization of the patientscondition until more definitive therapy is per-formed, depending on available expertise.74

Although a discussion of the long-term treat-ment of patients after acute peptic ulcer bleedingfalls outside the focus of this review, testing forand treatment ofH. pylori infection are criticalconsiderations to be addressed.77 In addition, inselected patients, evaluation for any ongoing needfor a nonsteroidal antiinflammatory or antiplate-let agent and, if such treatment is indicated, ap-propriate coadministration of a gastroprotectiveagent are important.78

Supported by an Advanced Research Career DevelopmentAward and a grant (01-191-1) from the Department of VeteransAffairs (to Dr. Gralnek) and a research scholarship from Fondsde la Recherche en Sant du Qubec (to Dr. Barkun).

Dr. Barkun reports receiving consulting fees from AstraZenecaand Abbott; and Dr. Bardou, lecture fees from AstraZeneca. Noother potential conflict of interest relevant to this article was

reported.

An animation showingendoscopic managementof acute bleeding from apeptic ulcer is availablewith the full text of thisarticle at www.nejm.org.





9/10



References

Lim CH, Vani D, Shah SG, Everett SM,1.Rembacken BJ. The outcome of suspectedupper gastrointestinal bleeding with24-hour access to upper gastrointestinalendoscopy: a prospective cohort study. En-doscopy 2006;38:581-5.

Lewis JD, Bilker WB, Brensinger C,2.Farrar JT, Strom BL. Hospitalization and

mortality rates from peptic ulcer diseaseand GI bleeding in the 1990s: relat ionshipto sales of nonsteroidal anti-inflammatorydrugs and acid suppression medications.Am J Gastroenterol 2002;97:2540-9.

Barkun A, Sabbah S, Enns R, et al.3.The Canadian Registry on NonvaricealUpper Gastrointestinal Bleeding and En-doscopy (RUGBE): endoscopic hemostasisand proton pump inhibition are associatedwith improved outcomes in a real-life set-ting. Am J Gastroenterol 2004;99:1238-46.

Targownik LE, Nabalamba A. Trends4.in management and outcomes of acutenonvariceal upper gastrointestinal bleed-ing: 1993-2003. Clin Gastroenterol Hepa-

tol 2006;4:1459-66. [Erratum, Clin Gastro-enterol Hepatol 2007;5:403.]

Lassen A, Hallas J, Schaffalitzky de5.Muckadell OB. Complicated and uncom-plicated peptic ulcers in a Danish county1993-2002: a population-based cohortstudy. Am J Gastroenterol 2006;101:945-53.

Ohmann C, Imhof M, Ruppert C, et al.6.Time-trends in the epidemiology of pepticulcer bleeding. Scand J Gastroenterol2005;40:914-20.

Viviane A, Alan BN. Estimates of costs7.of hospital stay for variceal and nonva-riceal upper gastrointestinal bleeding inthe United States. Value Health 2008;11:1-3.

Laine L, Peterson WL. Bleeding peptic8.ulcer. N Engl J Med 1994;331:717-27.

British Society of Gastroenterology9.Endoscopy Committee. Non-variceal uppergastrointestinal haemorrhage: guidelines.Gut 2002;51:Suppl 4:iv1-iv6.

Aljebreen AM, Fallone CA, Barkun AN.10.Nasogastric aspirate predicts high-riskendoscopic lesions in patients with acuteupper-GI bleeding. Gastrointest Endosc2004;59:172-8.

Hbert PC, Wells G, Blajchman MA, et11.al. A multicenter, randomized, controlledclinical trial of transfusion requirementsin critical care. N Engl J Med 1999;340:

409-17. [Erratum, N Engl J Med 1999;340:1056.]Baradarian R, Ramdhaney S, Chapal-12.

amadugu R, et al. Early intensive resusci-tation of patients with upper gastrointes-tinal bleeding decreases mortality. Am JGastroenterol 2004;99:619-22.

Corley DA, Stefan AM, Wolf M, Cook13.EF, Lee TH. Early indicators of prognosisin upper gastrointestinal hemorrhage.Am J Gast roenterol 1998;93:336-40.

Lee SD, Kearney DJ. A randomized14.controlled trial of gastric lavage prior to

endoscopy for acute upper gastrointesti-nal bleeding. J Clin Gastroenterol 2004;38:861-5.

Jensen DM, Kovacs TO, Jutabha R, et15.al. Randomized trial of medical or endo-scopic therapy to prevent recurrent ulcerhemorrhage in patients with adherentclots. Gastroenterology 2002;123:407-13.

Blatchford O, Murray WR, Blatchford16.M. A risk score to predict need for treat-ment for upper-gastrointestinal haemor-rhage. Lancet 2000;356:1318-21.

Rockall TA, Logan RF, Devlin HB,17.Northfield TC. Risk assessment after acuteupper gastrointestinal haemorrhage. Gut1996;38:316-21.

Carbonell N, Pauwels A, Serfaty L,18.Boelle PY, Becquemont L, Poupon R. Eryth-romycin infusion prior to endoscopy foracute upper gastrointestinal bleeding:a randomized, controlled, double-blindtrial. Am J Gastroenterol 2006;101:1211-5.

Barkun A, Bardou M, Marshall JK.19.Consensus recommendations for manag-

ing patients with nonvariceal upper gas-trointestinal bleeding. Ann Intern Med2003;139:843-57.

Das A, Wong RC. Prediction of out-20.come of acute GI hemorrhage: a review ofrisk scores and predictive models. Gastro-intest Endosc 2004;60:85-93.

Romagnuolo J, Barkun AN, Enns R,21.Armstrong D, Gregor J. Simple clinicalpredictors may obviate urgent endoscopyin selected patients with nonvariceal uppergastrointestinal tract bleeding. Arch InternMed 2007;167:265-70.

Forrest JA, Finlayson ND, Shearman22.DJ. Endoscopy in gastrointestinal bleed-ing. Lancet 1974;2:394-7.

Consensus conference: therapeutic en-23.doscopy and bleeding ulcers. JAMA 1989;262:1369-72.

Freeman ML, Cass OW, Peine CJ, On-24.stad GR. The non-bleeding visible vesselversus the sentinel clot: natural historyand risk of rebleeding. Gastrointest En-dosc 1993;39:359-66.

Lau JY, Sung JJ, Chan AC, et al. Stig-25.mata of hemorrhage in bleeding pepticulcers: an interobserver agreement studyamong international experts. GastrointestEndosc 1997;46:33-6.

Wong RC. Endoscopic Doppler US26.probe for acute peptic ulcer hemorrhage.

Gastrointest Endosc 2004;60:804-12.Adler DG, Leighton JA, Davila RE, et27.al. ASGE guideline: the role of endoscopyin acute non-variceal upper-GI hemor-rhage. Gastrointest Endosc 2004;60:497-504. [Erratum, Gastrointest Endosc 2005;61:356.]

Spiegel BM, Vakil NB, Ofman JJ. En-28.doscopy for acute nonvariceal upper gas-trointestinal tract hemorrhage: is soonerbetter? A systematic review. Arch InternMed 2001;161:1393-404.

Lau JY, Chung SC, Leung JW, Lo KK,29.

Yung MY, Li AK. The evolution of stig-mata of hemorrhage in bleeding pepticulcers: a sequential endoscopic study. En-doscopy 1998;30:513-8.

Cook DJ, Guyatt GH, Salena BJ, Laine30.LA. Endoscopic therapy for acute nonva-riceal upper gastrointestinal hemorrhage:a meta-analysis. Gastroenterology 1992;

102:139-48.Calvet X, Vergara M, Brullet E, Gisbert31.

JP, Campo R. Addition of a second endo-scopic treatment following epinephrineinjection improves outcome in high-riskbleeding ulcers. Gastroenterology 2004;126:441-50.

Marmo R, Rotondano G, Piscopo R,32.Bianco MA, DAngella R, Cipolletta L.Dual therapy versus monotherapy in theendoscopic treatment of high-risk bleed-ing ulcers: a meta-analysis of controlledtrials. Am J Gastroenterol 2007;102:279-89, 469.

Park WG, Yeh RW, Triadafilopoulos33.G. Injection therapies for nonvariceal

bleeding disorders of the GI tract. Gas-trointest Endosc 2007;66:343-54.

Sung JJ, Chan FK, Lau JY, et al. The34.effect of endoscopic therapy in patientsreceiving omeprazole for bleeding ulcerswith nonbleeding visible vessels or adher-ent clots: a randomized comparison. AnnIntern Med 2003;139:237-43.

Kahi CJ, Jensen DM, Sung JJ, et al. En-35.doscopic therapy versus medical therapyfor bleeding peptic ulcer with adherentclot: a meta-analysis. Gastroenterology2005;129:855-62. [Erratum, Gastroenter-ology 2006;131:980-1.]

Laine L. Systematic review of endo-36.scopic therapy for ulcers with clots: cana meta-analysis be misleading? Gastroen-terology 2005;129:2127-8.

Sung JJ, Tsoi KK, Lai LH, Wu JC, Lau37.JY. Endoscopic clipping versus injectionand thermo-coagulation in the treatmentof non-variceal upper gastrointestinalbleeding: a meta-analysis. Gut 2007;56:1364-73.

Chung IK, Kim EJ, Lee MS, et al. Endo-38.scopic factors predisposing to rebleedingfollowing endoscopic hemostasis in bleed-ing peptic ulcers. Endoscopy 2001;33:969-75.

Thomopoulos KC, Theocharis GJ, Va-39.genas KA, Danikas DD, Vagianos CE,

Nikolopoulou VN. Predictors of hemo-static failure after adrenaline injection inpatients with peptic ulcers with non-bleeding visible vessel. Scand J Gastroen-terol 2004;39:600-4.

Marmo R, Rotondano G, Bianco MA,40.Piscopo R, Prisco A, Cipolletta L. Out-come of endoscopic treatment for pepticulcer bleeding: is a second look neces-sary? A meta-analysis. Gastrointest Endosc2003;57:62-7.

Chiu PW-Y, Lau T-S, Kwong K-H, Suen41.DT-K, Kwok SP-Y. Impact of programmed





10/10

current concepts


second endoscopy with appropriate re-treatment on peptic ulcer re-bleeding:a systematic review. Ann Coll Surg HongKong 2003;7:106-15.

Spiegel BM, Ofman JJ, Woods K, Vakil42.NB. Minimizing recurrent peptic ulcerhemorrhage after endoscopic hemostasis:the cost-effectiveness of competing strat-egies. Am J Gastroenterol 2003;98:86-97.

Rockall TA, Logan RF, Devlin HB,43.Northfield TC. Selection of patients forearly discharge or outpatient care afteracute upper gastrointestinal haemorrhage:National Audit of Acute Upper Gastroin-testinal Haemorrhage. Lancet 1996;347:1138-40.

Longstreth GF, Feitelberg SP. Success-44.ful outpat ient management of acute uppergastrointestinal hemorrhage: use of prac-tice guidelines in a large patient series.Gastrointest Endosc 1998;47:219-22.

Dulai GS, Gralnek IM, Oei TT, et al.45.Utilization of health care resources forlow-risk patients with acute, nonvaricealupper GI hemorrhage: an historical co-

hort study. Gastrointest Endosc 2002;55:321-7.

Hay JA, Maldonado L, Weingarten SR,46.Ellrodt AG. Prospective evaluation of aclinical guidel ine recommending hospitallength of stay in upper gastrointestinaltract hemorrhage. JAMA 1997;278:2151-6.

Lai KC, Hui WM, Wong BC, Ching CK,47.Lam SK. A retrospective and prospectivestudy on the safety of discharging select-ed patients with duodenal ulcer bleedingon the same day as endoscopy. Gastroin-test Endosc 1997;45:26-30.

Moreno P, Jaurrieta E, Aranda H, et al.48.Efficacy and safety of an early dischargeprotocol in low-risk patients with upper

gastrointestinal bleeding. Am J Med 1998;105:176-81.

Lee JG, Turnipseed S, Romano PS, et49.al. Endoscopy-based triage significantlyreduces hospitalization rates and costs oftreating upper GI bleeding: a randomizedcontrolled trial. Gastrointest Endosc 1999;50:755-61.

Cipolletta L, Bianco MA, Rotondano50.G, Marmo R, Piscopo R. Outpatient man-agement for low-risk nonvariceal upperGI bleeding: a randomized controlled trial.Gastrointest Endosc 2002;55:1-5.

Bjorkman DJ, Zaman A, Fennerty MB,51.Lieberman D, Disario JA, Guest-WarnickG. Urgent vs. elective endoscopy for acutenon-variceal upper-GI bleeding: an effec-tiveness study. Gastrointest Endosc 2004;60:1-8.

Barkun AN, Cockeram AW, Plourde V,52.Fedorak RN. Review art icle: acid suppres-sion in non-variceal acute upper gastroin-testinal bleeding. Aliment Pharmacol Ther1999;13:1565-84.

Metz DC, Am F, Hunt B, Vakily M, Ku-53.kulka MJ, Samra N. Lansoprazole regi-mens that sustain intragastric pH >6.0:an evaluation of intermittent oral and con-

tinuous intravenous infusion dosages. Ali-ment Pharmacol Ther 2006;23:985-95.

Levine JE, Leontiadis GI, Sharma VK,54.Howden CW. Meta-analysis: the efficacyof intravenous H2-receptor antagonists inbleeding peptic ulcer. Aliment PharmacolTher 2002;16:1137-42.

Lau JYW, Sung JJY, Lee KKC, et al. Ef-55.fect of intravenous omeprazole on recur-

rent bleeding after endoscopic treatmentof bleeding peptic ulcers. N Engl J Med2000;343:310-6.

Bardou M, Toubouti Y, Benhaberou-56.Brun D, Rahme E, Barkun AN. Meta-analy-sis: proton-pump inhibition in high-riskpatients with acute peptic ulcer bleeding.Aliment Pharmacol Ther 2005;21:677-86.

Leontiadis GI, Sharma VK, Howden57.CW. Proton pump inhibitor treatment foracute peptic ulcer bleeding. CochraneDatabase Syst Rev 2006;1:CD002094.

Jensen DM, Pace SC, Soffer E, Comer58.GM. Continuous infusion of pantoprazoleversus ranitidine for prevention of ulcerrebleeding: a U.S. multicenter random-

ized, double-blind study. Am J Gastroen-terol 2006;101:1991-9, 2170.

Morgan D. Intravenous proton pump59.inhibitors in the critical care setting. CritCare Med 2002;30:Suppl:S369-S372.

Barkun AN, Herba K, Adam V, Ken-60.nedy W, Fallone CA, Bardou M. High-doseintravenous proton pump inhibition fol-lowing endoscopic therapy in the acutemanagement of patients with bleedingpeptic ulcers in the U S A and Canada:a cost-effectiveness analysis. Aliment Phar-macol Ther 2004;19:591-600.

Spiegel BM, Dulai GS, Lim BS, Mann61.N, Kanwal F, Gralnek IM. The cost-effec-tiveness and budget impact of intravenous

versus oral proton pump inhibitors inpeptic ulcer hemorrhage. Clin Gastroen-terol Hepatol 2006;4:988-97.

Dorward S, Sreedharan A, Leontiadis62.G, Howden C, Moayyedi P, Forman D.Proton pump inhibitor treatment initiatedprior to endoscopic diagnosis in uppergastrointestinal bleeding. Cochrane Data-base Syst Rev 2006;4:CD005415.

Lau JY, Leung WK, Wu JCY, et al.63.Omeprazole before endoscopy in patientswith gastrointest inal bleeding. N Engl JMed 2007;356:1631-40.

Al-Sabah S, Barkun AN, Herba K, et64.al. Cost-effectiveness of proton-pump in-hibition before endoscopy in upper gastro-intestinal bleeding. Clin GastroenterolHepatol 2008;6:418-25.

Tsoi KK, Lau JY, Sung JJ. Cost-effec-65.tiveness analysis of high-dose omeprazoleinfusion prior to endoscopy for patientspresenting with upper-GI bleeding. Gastro-intest Endosc 2008;67:1056-63.

Leontiadis GI, Sharma VK, Howden66.CW. Systematic review and meta-analysis:enhanced eff icacy of proton-pump inhibi-tor therapy for peptic ulcer bleeding inAsia a post hoc analysis from the Co-

chrane Collaboration. Aliment PharmacolTher 2005;21:1055-61.

Arabi Y, Al Knawy B, Barkun AN, Bar-67.dou M. Pro/con debate: octreot ide has animportant role in the treatment of gastro-intestinal bleeding of unknown origin?Crit Care 2006;10:218.

Paimela H, Oksala NK, Kivilaakso E.68.Surgery for peptic ulcer today: a study on

the incidence, methods and mortality insurgery for peptic ulcer in Finland be-tween 1987 and 1999. Dig Surg 2004;21:185-91.

Poxon VA, Keighley MR, Dykes PW,69.Heppinstall K, Jaderberg M. Comparisonof minimal and conventional surgery inpatients with bleeding peptic ulcer: a multi-centre tr ial. Br J Surg 1991;78:1344-5.

Millat B, Hay JM, Valleur P, Fingerhut70.A, Fagniez PL. Emergency surgical treat-ment for bleeding duodenal ulcer: over-sewing plus vagotomy versus gastric re-section, a controlled randomized trial.World J Surg 1993;17:568-73.

de la Fuente SG, Khuri SF, Schifftner71.

T, Henderson WG, Mantyh CR, Pappas TN.Comparative analysis of vagotomy anddrainage versus vagotomy and resectionprocedures for bleeding peptic ulcer dis-ease: results of 907 patients from the De-partment of Veterans Affairs NationalSurgical Quality Improvement Programdatabase. J Am Coll Surg 2006;202:78-86.

Imhof M, Ohmann C, Rher HD,72.Glutig H. Endoscopic versus operativetreatment in high-risk ulcer bleeding pa-tients results of a randomised study.Langenbecks Arch Surg 2003;387:327-36.

Lau JYW, Sung JJY, Lam Y, et al. Endo-73.scopic retreatment compared with surgeryin patients with recurrent bleeding after

initial endoscopic control of bleeding ul-cers. N Engl J Med 1999;340:751-6.

Kim SK, Duddalwar V. Failed endo-74.scopic therapy and the interventional ra-diologist: non-variceal upper gastrointes-tinal bleeding. Tech Gastrointest Endosc2005;7:148-55.

Ljungdahl M, Eriksson LG, Nyman R,75.Gustavsson S. Arterial embolisation inmanagement of massive bleeding fromgastric and duodenal ulcers. Eur J Surg2002;168:384-90.

Ripoll C, Baares R, Beceiro I, et al.76.Comparison of transcatheter arterial em-bolization and surgery for treatment ofbleeding peptic ulcer after endoscopictreatment failure. J Vasc Interv Radiol2004;15:447-50.

Suerbaum S, Michetti P.77. Helicobacterpylori infection. N Engl J Med 2002;347:1175-86.

Lanas A, Hunt R. Prevention of anti-78.inflammatory drug-induced gastrointesti-nal damage: benefits and risks of thera-peutic strategies. Ann Med 2006;38:415-28.Copyright 2008 Massachusetts Medical Society.



manejo de hemorragia aguda de la úlcera péptica

Documents