Treating GI Disease: Minimizing Risks and Enhancing Benefits

From the 45th Annual Northern Michigan Summer Conference: Update on Common Clinical Concerns in Primary Care, sponsored by the University of Michigan Medical School, Ann Arbor

James M. Scheiman, MD, Professor, Department of Internal Medicine, Division of Gastroenterology, University of Michigan Medical School, Ann Arbor

Educational Objectives

The goal of this program is to reduce cardiovascular (CV) and gastrointestinal (GI) risks associated with use of non­steroidal anti-inflammatory drugs (NSAIDs). After hearing and assimilating this program, the clinician will be better able to:

1.   Review data about cyclooxygenase (COX)-2 inhibitors and CV risk.

2.   Identify patients who may be at increased risk for GI complications with NSAID use.

3.   Compare efficacy and outcomes of COX-2 inhibitors to nonselective NSAIDs.

4.   Describe the effect of adding proton pump inhibitors (PPIs) to NSAID therapy.

5.   Select appropriate candidates for intravenous (IV) PPI administration for management of upper GI bleeding.

Faculty Disclosure

In adherence to ACCME Standards for Commercial Support, Audio-Digest requires all faculty and members of the planning committee to disclose relevant financial relationships within the past 12 months that might create any personal conflicts of in­terest. Any identified conflicts were resolved to ensure that this educational activity promotes quality in health care and not a proprietary business or commercial interest. For this program, the following has been disclosed: Dr. Scheiman is a consultant for AstraZeneca, Bayer, NiCox, Novartis, Pfizer, Pozen, and Takeda. He has also received speaker’s honoraria from Takeda. The planning committee reported nothing to disclose. In his lecture, Dr. Scheiman presents information that is related to off-la­bel or investigative use of a therapy, product, or device.

Acknowledgements

Dr. Scheiman spoke in Bellaire, MI, at the 45th Annual Northern Michigan Summer Conference: Update on Common Clin­ical Concerns in Primary Care, presented June 22-26, 2009, by the University of Michigan Medical School. The Audio-Di­gest Foundation thanks Dr. Scheiman and the University of Michigan Medical School for their cooperation in the production of this program.

Reducing Cardiovascular and GI Hazards of  NSAIDs

Nonsteroidal anti-inflammatory drugs (NSAIDs) and adverse cardiovascular (CV) outcomes: Adenomatous Polyp Prevention on Vioxx (APPROVe) trial  —  looked at rofecoxib (Vioxx) in secondary prevention of colon pol­yps; showed likelihood of adverse thrombotic cardiovascular event twice as high with rofecoxib, compared to pla­cebo; rofecoxib removed from market; celecoxib  —  Adenoma Prevention with Celecoxib (APC) trial looked at doses of 200 to 400 mg bid, and showed ³2.6-fold increase in risk for major fatal or nonfatal CV event, compared to placebo; other trial saw no adverse CV outcomes with 400 mg once daily; Alzheimer’s Disease Anti-inflamma­tory Prevention Trial (ADAPT) showed no CV risk with 200 mg once daily; data conflicting; important to consider dose and baseline CV risk before evaluating outcome; strong dose effect in patients at high and moderate risk, small dose effect in low-risk patients; deaths from CV causes in APC trial  —  in nonaspirin users, hazard ratio 2.4 (higher in aspirin users); although celecoxib may not interact with aspirin (ie, allows aspirin to bind to platelet), no evidence that celecoxib does not have CV risk; consider dose, patient, and frequency of medication

Effect of ibuprofen on aspirin: study showed ibuprofen and aspirin swallowed “close in proximity” reduces anti­platelet effect of aspirin; retrospective analysis of >7000 patients found those taking ibuprofen and aspirin had higher mortality, compared to those taking aspirin alone or aspirin with other NSAID; in patients taking aspirin for CV benefits, simultaneous ibuprofen use not recommended; Food and Drug Administration (FDA) suggests doses can be separated “by a number of hours” (no supportive data)

Naproxen: potent cyclooxygenic (COX-1) inhibitor; Antiplatelet Trialists’ Collaboration showed COX-2 inhibitors associated with higher risk for CV events compared to naproxen; COX-2 inhibitors had same CV risk as non-naproxen NSAIDs (eg, high-dose ibuprofen; indirect evidence that naproxen did not increase risk); Prospective Randomized Evaluation of Celecoxib Integrated Safety versus Ibuprofen Or Naproxen (PRECISION) study look­ing at celecoxib (100-200 mg bid), naproxen (375-500 mg bid), or ibuprofen (600-800 mg tid) with esomeprazole and low-dose aspirin under way

Conclusions: COX-2 inhibitors associated with increased CV risk in randomized controlled trials due to multifacto­rial mechanisms, eg, ibuprofen can raise blood pressure, leading to adverse CV effects; use COX-2 selective inhib­itors with caution in patients at risk for CV disease; no evidence that aspirin abrogates CV effect (adding aspirin to COX-2 inhibitor results in risk similar to traditional NSAID; risk additive); other nonselective NSAIDs can in­crease CV risk (naproxen appears to be exception; many guidelines suggest naproxen as medication of choice for patients with underlying CV risk); weigh risks and benefits

Gastrointestinal (GI) Side Effects

Ulcers: endoscopically diagnosed ulcers markers for serious adverse events; patients complain of dyspeptic symp­toms or present with serious bleeding; death rate 1 in 10 for patients with ulcer complications; multiple medical complications get worse when GI bleeding develops; most deaths from CV effects; in patient with dyspepsia, must reduce risk of developing ulcer and ulcer complications

Aspirin: in Danish study of 28,000 patients, those on low doses (£150 mg) had nearly 3-fold increase in risk for GI adverse event (6-fold risk in patients taking low-dose aspirin and traditional NSAID); before prescribing NSAID, ask whether patient taking aspirin; appropriate use  —  reduces mortality in patients using for secondary prophylaxis (ie, previous cardiac or cerebrovascular event, or diabetes [debatable]); be cautious with other antiplatelet agents (indicated for specific clinical outcomes); for primary prevention  —  decide on low-dose aspirin based on tradeoff between CV risk, GI risk, and risk for hemorrhagic stroke (no intervention); according to United States Preventive Services Task Force guidelines, aspirin should be used in healthy men and women (although benefits to women modest) with CV risk at 10 yr of ³6% (according to American Heart Association, ³10%); calculate CV risk before prescribing aspirin (calculation includes age, sex, cholesterol levels, smoking status, blood pressure, and medica­tions; calculator available online); data show only »50% of those with high CV risk taking aspirin

NSAID-associated GI complications: risk factors  —  previous complicated ulcer (relative risk [RR] increased nearly 14-fold); multiple NSAIDs (eg, NSAID and aspirin) approaches risk of patient with previous complicated ulcer (higher than risk for patient with previous uncomplicated ulcer); ask about use of over-the-counter (OTC) NSAIDs; use of anticoagulant or antiplatelet drug; use of selective serotonin reuptake inhibitors (SSRIs); advanced age; AP­PROVe trial  —  compared to placebo, rofecoxib shown to significantly increase RR (4- to 5-fold) in patients with symptomatic (ie, clinical event) or complicated ulcer; COX-2 inhibitors should not be considered similar to placebo (patient’s profile determines outcome); Celecoxib Long-term Arthritis Safety Study (CLASS)  —  saw significant re­duction in GI events in nonaspirin users; no significant benefit in aspirin users; modest numeric differences favor­ing celecoxib

Dyspepsia: common reason for stopping NSAID; endoscopic findings often do not correlate with symptoms; not risk factor for serious ulcer complications; most patients with ulcer complications have no previous symptoms; risk for dyspepsia slightly reduced with COX-2 inhibitors, but still greater than that of placebo; compared to nonselective NSAIDs, COX-2 inhibitors equally efficacious and lead to fewer (»50%) GI ulcers and complications if patient not taking aspirin or any other agent that inhibits COX-1 (eg, OTC NSAID); “coxibs” associated with marginally better dyspepsia rates or need for additional medications for GI symptoms; treating GI symptoms  —  no studies support acid suppression; study of esomeprazole in symptomatic patients taking nonselective or COX-2 selective NSAIDs saw 50% of patients symptomatic in either arm on placebo over time (additional 20%-30% patients became asymp­tomatic when proton pump inhibitor [PPI] added)

Helicobacter pylori: H pylori causes inflammation (mechanism by which ulcers develop) and affects gastrin levels and acid secretion; NSAIDs do not cause inflammation; NSAIDs cause direct chemical injury and interfere system­ically with factors of GI function; look for H pylori in patients with history of ulcer (patients still at increased risk with NSAIDs, even after eradication of H pylori); patients with previous ulcer bleeding who require NSAID should be tested and treated for H pylori, and remain on PPI to reduce recurrence of bleeding; ask patients about NSAID use and ulcer history; no good evidence that eradicating H pylori beneficial in patients without risk factors

Evidence that PPIs reduce risk for ulcers and bleeding: high-risk patients have significant risk of rebleeding on low-dose aspirin (risk can be reduced >90% with PPI); according to AHA guidelines, clopidogrel should be used in patients with aspirin intolerance; study  —  patients who bled on low-dose aspirin switched to clopidogrel; at 1 yr, 9% rebled (can be reduced to <1% if less expensive antiplatelet drug and PPI used); risk factors  —  patients with GI bleeding in setting of percutaneous coronary intervention (PCI) had 20-fold higher mortality at 30 days; risks for bleeding  —  sick patients with acute myocardial infarction; cardiac arrest; inotropic support

Algorithm: when starting antiplatelet therapy, ask about GI risk factors (eg, history of ulcer bleeding or any ulcer dis­ease [test and treat for H pylori]); in patients with GI bleeding on dual antiplatelet or anticoagulant therapy, reason­able to use gastroprotection; reasonable to use PPI in patients with multiple risk factors (eg, dual antiplatelet therapy and advanced age)

Alternatives to COX-2 inhibitors: in high-risk patients with history of GI bleeding, outcomes similar between COX-2 inhibitor alone (celecoxib), and PPI (omeprazole) plus nonselective NSAID (diclofenac); either option rea­sonable; studies  —1) study compared esomeprazole to placebo in patients on long-term arthritis therapy (COX-2 inhibitor or nonselective NSAID); patients had risk factors, but not previous ulcer bleeding; PPI reduced risk in low- and high-risk GI patients; high placebo rates showed COX-2 inhibitors carry risk in high-risk patients; 2) sub­sequent studies showed benefit of combining COX-2 inhibitor with PPI (nonselective NSAID and PPI similar to COX-2 inhibitor alone; benefit of adding gastroprotection to COX-2 inhibitor); 3) H pylori eradicated in patients on arthritis therapy who bled; patients randomized to esomeprazole and celecoxib, or celecoxib alone; no recurrent events at 1 yr in esomeprazole and celecoxib group (9% in all patients [lower if patients not on aspirin])

Lower GI events: eg, diverticular bleeding; some studies compare COX-2 inhibitors to naproxen and PPI (clinical significance unclear); studies show many mucosal breaks with COX-2 inhibitor use, compared to nonselective NSAID use; NSAIDs and aspirin can predispose patients to lower GI bleeding (clinical significance unclear); in ab­sence of aspirin, COX-2 inhibitors likely to cause lower rates of GI bleeding compared to traditional NSAIDs due to absence of antiplatelet effect; PPI will not affect small bowel injury or colonic injury; studies under way

Management: patients with no CV and low GI risk, not on aspirin  —  nonselective NSAID reasonable; COX-2 spe­cific inhibitor also reasonable (more expensive); patients with no CV risk and increased GI risk, not on aspirin  —  COX-2 selective inhibitors useful (similar GI outcomes with nonselective NSAID and PPI); patients with previous GI bleeding  —  COX-2 inhibitor and PPI recommended; aspirin users  —  naproxen useful; naproxen associated with GI risk (reasonable to add PPI); aspirin users with GI risk  —  consider CV risk (naproxen with PPI recommended); COX-2 inhibitor recommended for serious previous GI bleeding

Improving Outcomes from Upper GI Bleeding

Upper GI bleeding: most commonly caused by ulcer disease; prognostic factors  —  predict complications and mor­tality; clinical factors as or more important than endoscopic factors; advanced age; need for transfusion; shock at presentation; other serious medical problems; endoscopic findings (eg, major stigmata of hemorrhage; ongoing or bright red bleeding); hypotension; elevated prothrombin time; erratic mental status (sign of shock)

Types of ulcers: clean-based ulcer  —  most common; risk for rebleeding <5%; PPI therapy sufficient; flat spot  —  slightly increased risk for bleeding; adherent clots —fresh, red clot; associated with higher rebleeding risk; remove clot and treat endoscopically; with active bleeding, mortality can be as high as 11%

Initial management of upper GI bleeding: large-bore intravenous (IV) access; hemodynamic assessment for ortho­stasis; appropriate laboratory testing for coagulopathy and associated medical conditions; type and cross; aggres­sive resuscitation; admit high-risk patients to intensive care unit (ICU); admit low-risk patients to medical ward (check for orthostatic hypotension and other high-risk clinical factors (eg, 12-unit bleed); endoscopic therapy  —  should be performed urgently (ie, within 12 hr of presentation); patients with clean-based ulcer may be discharged home; patients with pumping vessel should remain in ICU and be treated endoscopically; using 2 techniques for en­doscopic hemostasis more effective than using 1 (injection not effective as monotherapy; also use cautery and/or clipping); dual endoscopic treatment reduces rebleeding by 40% and surgery by 33%; IV PPI with endoscopic ther­apy significantly reduced likelihood of recurrent bleeding

Acid suppression: for prophylaxis of stress gastritis (ICU condition where patients develop ischemic injury to fore­gut and bleed), target pH >4; to affect coagulation, target pH 5 to 7 (consider using high doses of oral PPI or contin­uous infusion of H₂-blocker); studies showed H₂-blockers not effective for acute bleeding, and do not reduce number of blood transfusions, need for surgery, or survival; H₂-blockers similar to placebo in many clinical trials; tolerance  —  with repeated doses over time, acid suppression with PPIs remains steady or increases; with H₂-block­ers, acid suppression decreases over time; PPI doses required for acid suppression decreased, whereas pH effect of H₂-blockers disappeared, even with increased doses; continuous infusion  —  half-life of PPI in bloodstream (given orally or as IV bolus), 1 hr; PPI binds at each pulse of infusion pump, leading to effective acid suppression; com­pared to placebo, older studies showed 80 mg of IV omeprazole at infusion rate of 8 mg/hr associated with signifi­cant improvement in recurrent bleeding and need for transfusion; oral PPI  —  studies looking at oral PPI (40 mg bid for 5 days) found bleeding did not stop in 20% to 33% of patients (studies done in populations with high H pylori, so data may not be generalizable to US population); IV omeprazole study  —  >700 patients with GI bleeding; pa­tients who underwent endoscopic treatment given IV omeprazole then switched to oral PPI after 3 days, or placebo; all patients received oral omeprazole (20 mg qd); in first 3 days, patients treated with IV PPI and endoscopy (com­pared to endoscopy and placebo) had significant reductions in endoscopic treatment, transfusions, and length of stay (no significant change after 3 days)

Use of IV PPIs: analysis of IV pantoprazole showed efficacy only in subgroup of high-risk stigmata patients; global study comparing IV esomeprazole to placebo  —  nearly 50% of patients >65 yr of age; some patients had comorbid­ities (eg, 5% presented in shock), visible vessels, or ulcers (most <2 cm); PPI highly effective (>50% reduction) in reducing risk for rebleeding for up to 30 day (most effects occur within first 3 days); dose effect important in acid suppression; IV PPI before endoscopy  —  patients who presented with overt signs of upper GI bleeding given IV omeprazole or placebo; study showed twice as many patients who did not receive IV PPI before endoscopy needed endoscopic therapy; twice as many actively bleeding ulcers in placebo group; speaker recommends starting IV in­fusion in emergency department; no need to continue IV infusion for rest of patient's hospital stay; IV PPI shown cost-effective

Restarting low-dose aspirin: if concerned about rebleeding, should be able to restart aspirin same day as endoscopic therapy; stopping aspirin for prolonged time in setting of GI bleeding risky; half-life of aspirin, 7 to 10 days

Summary: recommended endoscopic therapy includes injection and thermal therapy or clipping; PPI useful as adju­vant; PPI before endoscopy worthwhile; do not prolong restarting of low-dose aspirin

Suggested Reading

Becker MC et al: PRECISION Investigators. Rationale, design, and governance of Prospective Randomized Evaluation of Cele­coxib Integrated Safety versus Ibuprofen Or Naproxen (PRECISION), a cardiovascular end point trial of nonsteroidal antiinflamma­tory agents in patients with arthritis. Am Heart J 157:606, 2009; Bhatt DL et al: American College of Cardiology Foundation Task Force on Clinical Expert Consensus Documents. ACCF/ACG/AHA 2008 expert consensus document on reducing the gastrointesti­nal risks of antiplatelet therapy and NSAID use: a report of the American College of Cardiology Foundation Task Force on Clinical Expert Consensus Documents. J Am Coll Cardiol 52:1502, 2008; Bhatt DL et al: American College of Cardiology Foundation Task Force on Clinical Expert Consensus Documents. ACCF/ACG/AHA 2008 expert consensus document on reducing the gastrointesti­nal risks of antiplatelet therapy and NSAID use: a report of the American College of Cardiology Foundation Task Force on Clinical Expert Consensus Documents. Circulation 118:1894, 2008; Chan FK et al: Management of patients on nonsteroidal anti-inflamma­tory drugs: a clinical practice recommendation from the First International Working Party on Gastrointestinal and Cardiovascular Ef­fects of Nonsteroidal Anti-inflammatory Drugs and Anti-platelet Agents. Am J Gastroenterol 103:2908, 2008; Elnachef N et al: Changing perceptions and practices regarding aspirin, nonsteroidal anti-inflammatory drugs, and cyclooxygenase-2 selective nonste­roidal anti-inflammatory drugs among US primary care providers. Aliment Pharmacol Ther 28:1249, 2008; FitzGerald GA: COX-2 in play at the AHA and the FDA. Trends Pharmacol Sci 28:303, 2007; Lau JY et al: Effect of intravenous omeprazole on recurrent bleeding after endoscopic treatment of bleeding peptic ulcers. N Engl J Med 343:310, 2000; Lau JY et al: Omeprazole before endos­copy in patients with gastrointestinal bleeding. N Engl J Med 356:1631, 2007; Ng FH et al: Management and outcome of peptic ul­cers or erosions in patients receiving a combination of aspirin plus clopidogrel. J Gastroenterol 43:679, 2008; Scheiman JM: Balancing risks and benefits of cyclooxygenase-2 selective nonsteroidal anti-inflammatory drugs. Gastroenterol Clin North Am 38:305, 2009; Solomon DH et al: Subgroup analyses to determine cardiovascular risk associated with nonsteroidal antiinflammatory drugs and coxibs in specific patient groups. Arthritis Rheum 59:1097, 2008; Sung JJ et al: Intravenous esomeprazole for prevention of peptic ulcer re-bleeding: rationale/design of Peptic Ulcer Bleed study. Aliment Pharmacol Ther 27:666, 2008; Vardeny O et al: Cyclooxygenase-2 inhibitors, nonsteroidal anti-inflammatory drugs, and cardiovascular risk. Cardiol Clin 26:589, 2008.