Bleeding Peptic Ulcer in the Elderly

Drugs Aging 2007; 24 (10): 815-828THERAPY IN PRACTICE 1170-229X/07/0010-0815/$44.95/0

© 2007 Adis Data Information BV. All rights reserved.

Bleeding Peptic Ulcer in the ElderlyRisk Factors and Prevention Strategies

Angelo Zullo, Cesare Hassan, Salvatore M.A. Campo and Sergio Morini

Gastroenterology and Digestive Endoscopy, “Nuovo Regina Margherita” Hospital, Rome, Italy

Contents

Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8151. The Disease Burden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8162. Risk Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 816

2.1 Role of NSAIDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8172.2 Role of Selective Cyclo-Oxygenase (COX)-2 Inhibitors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8192.3 Role of Helicobacter pylori . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8202.4 H. pylori and NSAIDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8212.5 H. pylori and Selective COX-2 Inhibitors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8212.6 Role of Other Drugs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 821

3. Prevention Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8223.1 Management of NSAIDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8223.2 Management of H. pylori Infection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 824

4. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 825

Peptic ulcer bleeding is a frequent and dramatic event with both a highAbstractmortality rate and a substantial cost for healthcare systems worldwide. It has beenfound that age is an independent predisposing factor for gastrointestinal bleeding,with the risk increasing significantly in individuals aged >65 years and increasingfurther in those aged >75 years. Indeed, bleeding incidence and mortality aredistinctly higher in elderly patients, especially in those with co-morbidities.NSAID therapy and Helicobacter pylori infection are the most prevalentaetiopathogenetic factors involved in peptic ulcer bleeding. The risk of bleedingseems to be higher for NSAID- than for H. pylori-related ulcers, most likelybecause the antiplatelet action of NSAIDs impairs the clotting process. NSAIDusers may be classified as low or high risk, according to the absence or presence ofone or more of the following factors associated with an increased risk of bleeding:co-morbidities; corticosteroid or anticoagulant co-therapy; previous dyspepsia,peptic ulcer or ulcer bleeding; and alcohol consumption. Different types ofNSAIDs have been associated with different bleeding risk, but no anti-inflam-matory drug, including selective cyclo-oxygenase (COX)-2 inhibitors, is com-pletely safe for the stomach. Furthermore, even low-dose aspirin (acetylsalicylic

816 Zullo et al.

acid) [<325 mg/day] and a standard dose of non-aspirin antiplatelet treatment(clopidogrel or ticlopidine) have been found to cause bleeding and mortality. Noclear risk factor favouring H. pylori-related ulcer bleeding has been identified.Peptic ulcer bleeding prevention remains a challenge for the physician, but dataare now available on use of a safer and cheaper strategy for both low- and high-risk patients. Unfortunately, despite the fact that several society and nationalguidelines have been formulated, these are poorly followed in clinical practice.Proton pump inhibitor (PPI) or misoprostol therapy and H. pylori eradication inNSAID-naive patients are the most commonly proposed strategies. SelectiveCOX-2 inhibitor therapy in high-risk patients has also been suggested, butconcerns over the possible cardiovascular adverse effects of some of these agentsshould be taken into account. Moreover, switching to selective COX-2 inhibitorsin patients with previous bleeding is not completely risk free, and concomitant PPItherapy is also needed. H. pylori eradication is mandatory in all patients withpeptic ulcer, and such an approach has been found to be significantly superior toPPI maintenance therapy. H. pylori eradication is frequently achieved withsequential therapy in elderly patients with peptic ulcer.

In conclusion, upper gastrointestinal bleeding is a dramatic event with a highmortality rate, particularly in the elderly. Some effective preventative strategiesare now available that should be implemented in clinical practice.

1. The Disease Burden at an increased risk of developing several gastroduo-denal diseases, including peptic ulcer.[9] This largely

Acute upper gastrointestinal bleeding (UGB) is aresults from both a reduction in defensive mecha-

very common condition, with an estimated inci-nisms in the gastrointestinal mucosa and a higher

dence as high as 40–150 per 100 000 cases annuallyprevalence of aggressive factors in the elderly.[10-12]

in different countries, and highest rates in areas ofIn addition, a dramatically higher mortality rate due

low socioeconomic status.[1-3] A definite cause forto gastrointestinal bleeding has been observed in

UGB may be identified in nearly 80% of cases.aged patients compared with young people.[1] In-

Peptic ulcer (35–70%) and gastroduodenal erosionsdeed, UGB mortality has been reported to be virtual-

(8–15%) are the most frequently encountered dis-ly 0% for patients aged <40 years and >30% for

eases, overall accounting for >75% of cases.[3,4]nonagenarian patients.[13] Moreover, the mortality

Undeniably, UGB is a dramatic event resulting in afor UGB is strongly associated with the presence of

high mortality rate, ranging from 0.9% to 26.5%.[1,5]significant co-morbidities, which are generally more

UGB is also a frequent cause of hospitalisationcommon in geriatric patients.[14-16]

worldwide. It has been estimated that there are107 000 hospital admissions for UGB each year in 2. Risk Factorsthe US,[6] implying a substantial burden on health-care resources.[7] Different risk factors have been identified for

Several studies have found that UGB is a much peptic ulcer bleeding. Among these, NSAIDs andmore common condition in the elderly than in young Helicobacter pylori infection are clearly the mostpatients.[1,2,8] Geriatric patients are well known to be prevalent causes of UGB in elderly patients.

© 2007 Adis Data Information BV. All rights reserved. Drugs Aging 2007; 24 (10)

Bleeding Peptic Ulcer in the Elderly 817

NSAIDs are commonly used in geriatric patients for 2.1 Role of NSAIDs

several purposes,[17] and H. pylori prevalence isNSAIDs are among the most frequently pre-distinctly higher in aged patients compared with

scribed drugs worldwide, being used by >30 millionyoung people, at least in developed countries.[7]

people daily. Such drugs are more frequently pre-Epidemiological studies have shown that, on the one

scribed in aged patients because of the high preva-hand, the admission rate for peptic ulcer is decreas- lence of predisposing conditions, such as chronicing overall, especially in young patients, but, on the musculoskeletal diseases (osteoarthritis, rheumatoidother hand, the hospitalisation rate for both gastric arthritis, osteoporosis), persistent non-malignantand duodenal ulcer bleeding is increasing, particu- pain (neuropathies, migraine) and malignant pain inlarly in aged patients.[8] A decreasing H. pylori this population.[18,19] Moreover, elderly patients takeinfection rate could be involved in the reduction of low-dose NSAIDs for cardiovascular prevention.[20]

It has been calculated that approximately 15% ofpeptic ulcer incidence in young patients, whereaspatients aged >65 years take NSAIDs in New Zea-the wide use of NSAIDs and the increasing aging ofland,[21] as do up to 25% of this age group in Italy.[14]the general population could explain the higher fre-NSAID consumption is also likely to be largelyquency of hospital admissions due to UGB.underestimated in the elderly, since several patients

It has been well documented that gastrointestinaluse these drugs as over-the-counter therapy.[22] It is

mucosal damage depends on a break in the normalnoteworthy that several UGB episodes specifically

balance between aggressive and protective factors. occur in elderly patients taking over-the-counterIn elderly patients, aggressive factors act on an NSAIDs, most likely because of a lack of adequatealready compromised gastroduodenal mucosa. In- gastric protection. In an observational study per-deed, some studies have found that different protec- formed in the US, not only were 63 of 81 (75%)tive factors and repair mechanisms are impaired in bleeding patients taking NSAIDs, but of these, 60aged patients.[10,11] Several alterations of the gas- (95%) patients were taking such drugs as over-the-

counter treatment.[4]troduodenal mucosa have been found in aged pa-The role of NSAIDs in UGB is widely recog-tients, including reductions in: (i) mucus-secreting

nised. It has been calculated that the attributable riskcells; (ii) prostaglandin production; (iii) mucosalin the general population is as high as 38%.[2] Insurface hydrophobicity; (iv) basal and stimulatedaddition, up to 46–75% of bleeding ulcers,bicarbonate secretion; (v) expression of trefoil fac-25–36.2% of deaths and 29% of hospital admissionstors; (vi) glutathione concentration; (vii) epithelialfor peptic ulcer are attributable to NSAID use.[1,4,23]

cell proliferation; (viii) gastric emptying; andThe incidence of UGB is dramatically higher in

(ix) gastric blood flow.[10,11] All of these factors,aging NSAID users compared with young patients,

through different mechanisms, act either by protect-and it further increases with aging. It has been

ing the gastroduodenal mucosa from the develop- calculated that the UGB risk increases from 1.65 perment of new lesions or by favouring the repair of 100 000 in patients aged <65 years to 5.7 in thosealready established erosions or ulcers. Taken togeth- aged >65 years and 12.7 in those aged >75er, these observations suggest that a high prevalence years.[24,25] Besides age, other factors have beenof risk factors and a vulnerable gastroduodenal mu- involved in increasing the risk of UGB in NSAIDcosa lead to an increased risk of UGB in elderly users. It has been observed that the presence ofpatients. relevant co-morbidities significantly increases the


818 Zullo et al.

risk of both UGB and mortality.[1] It is known that table I. In a large epidemiological study, the risk ofco-morbidities are present in the majority of elderly bleeding in patients aged >60 years hospitalised forpatients, as suggested by the observation that UGB appeared to be low for ibuprofen and91–95% of patients aged >75 years take three to diclofenac, intermediate for piroxicam and naproxeneight drugs daily.[14-16] It has been also found that and high for ketoprofen and azapropazone.[34] It hasconcomitant administration of either corticosteroid also been found that 30-day mortality for UGB dueor antiplatelet treatment further increases the UBG to NSAID use is lower with ibuprofen, intermediaterisk in NSAID users.[24,25] Similarly, anticoagulant with diclofenac and high with naproxen.[35]

therapy, which is mainly used in aged patients forDuration of NSAID therapy is also important. Of

non-valvular atrial fibrillation, venous thromboem-note, it has been found that the risk of gastrointesti-

bolic disease, ischaemic heart disease and mechani-nal complications is doubled with short- (<1 month)

cal heart valves,[26] significantly increases the inci-compared with long-term NSAID therapy (ORs:

dence of UGB when NSAIDs are administered.[24] A4.14–8.9 vs 1.71–4.4, respectively).[2,32] A meta-history of uncomplicated peptic ulcer before NSAIDanalysis found that the relative risk of developingtherapy has also been associated with a higher UGBgastroduodenal lesions is 8 within the first 30 daysincidence, and the risk is markedly increased inof NSAID therapy, 3.31 with therapy of 30–90 dayspatients with a previous bleeding peptic ulcer (oddsand 1.92 with longer therapy.[25] This phenomenonratio [OR]: 13).[27] The role of dyspeptic symptomsis probably due to an adaptation process, occurringbefore NSAID therapy is controversial, although thein response to some NSAIDs, which involves in-available data suggest an increased risk.[2,27-29]

creased salivary and gastric production of cytokinesSome studies have also found that alcohol con- (epidermal growth factor, transforming growth fac-

sumption increases the risk of bleeding in NSAID tor, fibroblast growth factor) that stimulate ulcerusers.[28,30] UGB incidence is significantly higher in healing.[36,37] However, adaptive processes in theboth low- (OR: 2.8) and high-dose (OR: 7) aspirin gastric mucosa within the first month of treatment(acetylsalicylic acid) users, particularly for heavy

with NSAIDs are known to be significantly lessdrinkers (>21 drinks/week), whilst the risk for UGB

functional in older than in young patients.[38] Anoth-among ibuprofen users was estimated to be 2.7,

er possible explanation for the early increase inirrespective of alcohol intake.[30] Finally, a recent

events is the presence of pre-existing lesions/ulcersstudy enrolling 2777 patients with UGB found that

that were not recognised.smoking (current or past) significantly increasesUGB risk (OR: 1.53–2.32).[29] However, these datacontrast with the results of a previous study in whichpeptic ulcer bleeding was found not to be associatedwith smoking.[31]

Although different types of NSAIDs have beenassociated with different UGB risk, no anti-inflam-matory drug is completely safe for the stomach.Some studies have assessed the relative risk of dif-ferent NSAIDs for gastrointestinal complica-tions.[2,29,32,33] The relative UGB risks associatedwith the most commonly used NSAIDs are shown in

Table I. Relative upper gastrointestinal bleeding risk for the mostcommonly used NSAIDs[2,21,26,32,34]

Drug Odds ratio

Aceclofenac 1.4–2.6

Aspirin (acetylsalicylic acid) [>300 mg/day] 3.3–8

Diclofenac 3.1–8.3

Ibuprofen 2–4.1

Indometacin 4.9–11.3

Ketoprofen 5.4–23.7

Ketorolac 4.8–24.7

Naproxen 3.1–14.9

Nimesulide 2.6–13.8

Piroxicam 5.4–19.1



Although a dose-related risk has been generally rofecoxib, valdecoxib, lumiracoxib, etoricoxib) areobserved, UGB can occur even with low-dose as effective as non-selective NSAIDs in improving(75–325mg) aspirin, which is widely used for car- physical function in patients with osteoarthritis anddiovascular prevention.[39] Some large, double-blind rheumatoid arthritis, including elderly patients.[46] Atrials consistently found that the risk of bleeding was review of a published study found that 3 months ofmore than twice (OR: 2.2–3.4) that of placebo fol- celecoxib (100 or 200 mg/day) therapy relieves painlowing aspirin 75 mg/day therapy.[39,40] Moreover, a and improves quality of life in patients aged >75recent nationwide study performed in Spain showed years with osteoarthritis.[47] Similar results havethat up to one-third of all NSAID/aspirin deaths been observed with daily use of the other selectivecould be attributed to low-dose aspirin use.[1] COX-2 inhibitors in low doses.[45] However, an in-

Two recent studies have also found that the UGB creased incidence of cardiovascular adverse eventsrisk with a standard dose of non-aspirin antiplatelet following use of some selective COX-2 inhibitorstreatment (clopidogrel or ticlopidine) is similar to has been a matter for concern, such that boththat with aspirin 100 mg/day (ORs: 2.7–4 vs rofecoxib and valdecoxib have been recently with-2.3–3.1, respectively).[27,29] Finally, there has been drawn from the market.an attempt to reduce the toxic effect of NSAIDs on

Despite their equivalent efficacy to NSAIDs, se-the gastric mucosa by modifying the drug formula-

lective COX-2 inhibitors have a lower potential totion or route of administration. Unfortunately, buf-

injure the gastroduodenal mucosa, a property whichfered or enteric formulations of NSAIDs,[40] as well

depends on their selective inhibition of COX-2.[45]

as use of different routes of administration (oral vsFor this reason, use of selective COX-2 inhibitors is

rectal), do not reduce the risk of UGB.[41] Indeed, itincreasing in clinical practice, particularly in the

is well known that NSAID-related mucosal damageelderly and in patients who previously complained

is due to both topical and systemic effects of theof gastrointestinal adverse effects with use of stan-

drug.[42]

dard NSAIDs.[46] Unfortunately, even these drugsSome investigators have proposed stratifying

are not completely safe with respect to the gas-NSAID users according to the presence of different

troduodenal mucosa and, as reported for standardrisk factors: class I (low risk): absence of risk; class

NSAIDs, gastrointestinal adverse events followingII (moderate risk): presence of one to two risk fac-

selective COX-2 inhibitor therapy are more frequenttors; class III (high risk): presence of more thanin aged than in young patients.[29] Three large stud-three risk factors; and class IV (very high risk):ies, CLASS (Celecoxib Long-term Arthritis Safetyhistory of previous peptic ulcer bleeding.[43] TheStudy),[48] VIGOR (Vioxx GI Outcomes Re-incidence of UGB has been shown to be 0.8%, 2%,search)[49] and TARGET (Therapeutic Arthritis Re-8.6% and 18%, respectively, in these four classes.[44]

search and Gastrointestinal Event Trial),[50] showedthat the incidence of UGB is nearly 2-fold lower2.2 Role of Selective Cyclo-Oxygenasewith celecoxib or rofecoxib therapy compared with(COX)-2 Inhibitorsstandard NSAIDs, and 4-fold lower with lumiracox-ib compared with high-dose ibuprofen, but a definiteSince no NSAID is free of risk of gastrointestinalrisk for UGB is still detectable with selectiveadverse effects, a novel drug class, i.e. the selectiveCOX-2 inhibitors. Indeed, in the CLASS study, overCOX-2 inhibitors, was developed for long-term paina period of 3 days to 6 months, the incidence of ulcerrelief and anti-inflammatory purposes.[45] There iscomplications was 0.76% with celecoxib therapyconsiderable evidence that these drugs (celecoxib,


820 Zullo et al.

and 1.45% with ibuprofen or diclofenac.[48] In con- that switching to selective COX-2 inhibitors in suchpatients is not completely risk free.[54]trast, other recent, very large, observational studies

found a 2-fold increase in the risk of bleeding with2.3 Role of Helicobacter pylorirofecoxib therapy compared with controls, but no

significantly increased UGB risk was observed withThe aetiopathogenetic role of H. pylori infectioncelecoxib.[29,51] However, another recent cohort

in peptic ulcer disease is widely recognised.[55] In-study performed in Denmark that enrolled 7232deed, virtually all duodenal and gastric ulcers, afterpatients with peptic ulcer bleeding (mean age: 74exclusion of NSAID-related lesions, are due toyears; range: 62–82), found that the 30-day mortali-H. pylori. Moreover, H. pylori eradication definitelyty with celecoxib, when administered concomitantlychanges the natural history of peptic ulcer disease,with other drugs, was significantly higher (OR: 2)with lesion recurrence being prevented in all suc-

compared with controls.[35] A recent study, thecessfully cured patients.[56] H. pylori is highly preva-

MEDAL (Multinational Etoricoxib and Diclofenaclent in the elderly, and peptic ulcer disease is much

Arthritis Long-term) trial, which enrolled 34 701more frequent in aged than in young patients, most

patients with osteoarthritis or rheumatoid arthritis,likely because of their compromised gastroduodenal

found a lower gastrointestinal adverse effect ratemucosa (see section 1).[10,11] H. pylori eradication

with etoricoxib than with diclofenac (0.67% vshas been shown to dramatically reduce both duode-

0.97%, respectively), but the risk of complicatednal and gastric ulcer recurrence in elderly pa-

upper gastrointestinal events (bleeding, perforation)tients.[57]

did not significantly differ between the two groupsAlthough every peptic ulcer is a potential cause

(0.30% vs 0.32%, respectively).[52] Duration of se-of UGB, lesions exclusively associated with H. py-

lective COX-2 inhibitor therapy may also be anlori infection seem to cause bleeding less frequently

important factor in the gastrointestinal safety ofthan those related to NSAIDs. The antiplatelet ac-

these drugs. Indeed, the significant advantage oftion of NSAIDs could play a role, since ulcer bleed-

selective COX-2 inhibitor therapy compared with ing even from small lesions is presumably facilitatedibuprofen or diclofenac observed within 6 months by impairment of clotting. Nonetheless, a definitehas been shown to disappear over the subsequent rate of UGB due to H. pylori-related ulcers has been12–15 months.[53]

reported. It has been estimated that approximatelyIt has been widely observed that concomitant use 1–3% of patients with duodenal ulcer who are not

of a selective COX-2 inhibitor and a traditional receiving antisecretory therapy are likely to haemor-NSAID or aspirin is associated with a relative risk of rhage during each year of follow-up, resulting in aUGB higher than would be expected (OR: 14.5) cumulative risk of haemorrhage after 5 years offrom consideration of the additive effects of the approximately 10–14%.[58]

single molecules, and exceeding that of NSAIDs In a study that enrolled 81 patients with UGBplus aspirin therapy (OR: 12.7).[29] Therefore, con- conducted in the US,[4] 10% of peptic ulcers werecomitant therapy with aspirin or NSAIDs signifi- exclusively associated with H. pylori infection. Incantly undermines the safety of selective COX-2 another study performed in Spain that enrolled 92inhibitors. Finally, some observations suggest that patients with peptic ulcer bleeding, UGB due to aselective COX-2 inhibitor therapy does not com- peptic ulcer was related to H. pylori infection in aspletely prevent UGB recurrence in patients with a many as 56% of patients.[59] It has been also reportedprevious episode of peptic ulcer bleeding, indicating that 22% of 7232 patients with peptic ulcer bleeding



had never used NSAIDs.[35] However, H. pylori NSAIDs; OR: 6.13 for H. pylori plus NSAIDs).[62] Itinfection was not assessed in this cohort study, and it should be noted that both factors have been detectedmay be presumed that a large proportion of these in 35% of UGB episodes.[4,59] Another study foundpatients were currently infected. Other important that H. pylori infection was an independent riskdata on the role of H. pylori infection in duodenal or factor for bleeding even in low-dose aspirin usersgastric ulcer bleeding arise from studies of bacterial (OR: 4.7).[66] Finally, a synergistic effect has beeneradication performed to prevent rebleeding. It has described in a large, case-control study performed inbeen found that H. pylori eradication is superior to Denmark.[31] This study found that the presence ofboth no antisecretory treatment and maintenance H. pylori infection increases the risk (OR: 1.81) ofantisecretory therapy in patients with UGB.[60] bleeding in elderly NSAID users, and that 24% of all

UGB events in these patients are attributable to thisUnfortunately, in contrast with NSAIDs, no spe-infection.cific studies have focused on assessment of risk

factors for UGB in peptic ulcer associated with2.5 H. pylori and Selective COX-2 InhibitorsH. pylori infection. However, it could be hy-

pothesised that increased age, presence of co-mor- Only scanty data are available on the possiblebidities and anticoagulant therapy could play roles relationship between H. pylori infection and selec-in this setting. tive COX-2 inhibitor therapy in peptic ulcer devel-

opment and, to our knowledge, no data exist with2.4 H. pylori and NSAIDs respect to UGB. Celecoxib therapy has been found

to significantly increase the incidence of peptic ulcerAlthough several studies have investigated the

in infected versus uninfected patients (12.9% vspossible interaction between H. pylori infection and

2.9%, respectively).[67] After rofecoxib therapy, theNSAID therapy in the pathogenesis of UGB, the

duodenal ulcer incidence in patients with H. pyloriavailable data are still controversial.[61] A meta-

did not differ from that in patients taking naproxen,analysis of 463 trials found a 61-fold increased risk

and a 3.5-fold increase in gastric events was ob-of peptic ulcer onset in NSAID users with H. pylori

served in infected patients with a previous pepticinfection compared with NSAID never users and

ulcer compared with uninfected patients.[68]uninfected patients.[62] However, these two factors,which are both very common in elderly patients, 2.6 Role of Other Drugshave been found to play alternatively protective,

Besides NSAIDs, other drugs have been associat-independent or additive effects in relation to causinged with a significant risk of development of UGBUGB. Although H. pylori has been significantlydevelopment, such as spironolactone, antidepres-associated with an increased prevalence of bothsants and calcium channel antagonists; conversely,gastric and duodenal ulcer in elderly NSAIDHMG-CoA reductase inhibitors (statins) appear tousers,[63] the same investigators found that the pres-exert a protective role.ence of H. pylori infection significantly reduced

(OR: 0.2) patients’ risk of gastric ulcer bleeding.[64] A population-based, case-control study enrollingA similar finding was also observed in a study 306 345 patients recently performed in The Nether-conducted in Turkey (OR: 0.09).[65] On the other lands found that spironolactone therapy was associ-hand, a meta-analysis found that these two factors ated with a 2.7-fold increased risk of a gastrointesti-independently increased the risk of UGB via an nal event (peptic ulcer or bleeding), and that the riskadditive effect (OR: 1.79 for H. pylori, OR: 4.85 for was greater as the dosage increased (5.1-fold with


822 Zullo et al.

>0.5 of defined daily dose).[69] This finding was not In a recent retrospective study of 10 288 patientswith acute coronary syndromes receiving aspirin,observed with other diuretics, such as furosemide orthe rate of UGB was shown to be significantly loweramiloride. A possible interpretation of this observa-in statin users than in control patients (1% vs 1.8%,tion is that the stomach and the duodenum expressrespectively; OR: 0.68).[72] The investigators sug-mineralocorticosteroid receptors, which are impli-gested that this protective effect could be related tocated in fibrous tissue formation, and, in particular,an increased systemic prostaglandin production in-the healing of gastric or duodenal erosions and ul-duced by statins, which may protect the gastriccers. Thus, aldosterone receptor antagonists such asmucosa and prevent gastrointestinal bleeding.spironolactone could impair healing of gastric or

duodenal erosions, resulting in the formation of3. Prevention Strategies

gastroduodenal ulcers and bleeding.[69]

Given the frequent incidence of UGB, especiallyA recent comprehensive review of the literaturein elderly patients, management of this conditionfound that use of selective serotonin reuptake inhibi-remains a challenge for physicians in clinical prac-tors (SSRIs) is associated with an increased risk oftice and has a significant economic impact onUGB, a finding that has been attributed to decreasedhealthcare resources. Moreover, the continuing highserotonin levels in platelets and consequent defec-mortality rate of the disorder raises a number oftive aggregation.[70] However, only a few epidemio-ethical problems, given that UGB episodes and re-

logical studies have provided weak evidence to sup-lated deaths are preventable, at least in a subgroup of

port the hypothesis of a link between these drugs andpatients. Indeed, different risk factors for UGB have

UGB at a population level. Nevertheless, concomi-been clearly identified and, consequently, more ef-

tant use of NSAIDs or aspirin with SSRIs greatlyforts should be made to improve prevention strate-

increases the risk of UGB, especially in the elderly gies. Obviously, some of these factors are not modi-or in those with a history of peptic ulcer.[70]

fiable, such as age or co-morbidities, but NSAIDAlthough several studies have evaluated the po- use, H. pylori infection and some habits (alcohol,

tential role of calcium channel antagonist therapy in smoking) can be more appropriately managed.UGB, data are still conflicting. In a cohort study

3.1 Management of NSAIDsenrolling 5888 elderly patients it was observed thatUGB tended to be more frequent with calcium chan- NSAIDs are the most common cause of UGB innel antagonist therapy than with other antihyperten- the elderly and, therefore, should be used only whensive agents (OR: 1.37).[71] This finding was con- really necessary. Although no NSAID is risk free,firmed in a large, retrospective study.[72] In contrast, some are certainly less damaging than others to thein another trial, a lower UGB risk was observed with gastroduodenal mucosa. Selecting a less toxicuse of calcium channel antagonists compared with drug – administered at the lowest effective dose – isACE inhibitors.[73] Moreover, the UGB risk was a simple and clearly effective approach. Moreover,lower in NSAID users treated with calcium channel before starting NSAID therapy, the presence of riskantagonists compared with placebo, suggesting even factors (class risk stratification; see section 2.1)a protective effect.[74] In another study, calcium should be clearly considered. In addition, use of achannel antagonist therapy was significantly associ- gastroprotective drug should be considered whenated with lower, but not upper, gastrointestinal opportune. Unfortunately, this approach seems to bebleeding.[71] rare in clinical practice. Indeed, a large retrospective



study of 69 648 NSAID users conducted in The Despite this facility, it is disappointing to note thatthe vast majority of elderly NSAID users in Italy doNetherlands found that as many as 86.6% of patientsnot receive adequate gastroprotective therapy inin class risk II and 81.2% in class risk III did notclinical practice.[14] Guideline recommendationsreceive any gastroduodenal protection in clinicalshould be stringently implemented.practice.[75] Similarly, in a prospective observational

study performed in Italy, only 124 of 779 (15.9%) What is the best gastroprotective therapy forelderly patients treated with NSAIDs actually re- UGB in these elderly patients taking NSAIDs? Un-ceived adequate gastroduodenal protective ther- deniably, proton pump inhibitor (PPI) or misopros-apy.[14] Is this still acceptable behaviour? In reality, tol co-therapy are the most effective gastroduodenaleach physician should be required to not only avoid protective strategies for high-risk NSAIDinappropriate prescription of NSAIDs and provide users.[81,82] Such an approach has been also suggest-adequate gastroduodenal protection in the elderly, ed for elderly patients.[17,32] Misoprostol therapy isbut also to inform patients of the risk of over-the- slightly cheaper than PPI therapy (€1 vs ≈€1.5counter use of such drugs. daily, respectively, in Italy), but it requires taking

four tablets daily, it can cause abdominal crampsAs far as the timing of gastroduodenal protectiveand diarrhoea in many patients and it does nottherapy is concerned, it is of paramount importanceprevent dyspeptic symptoms, which occur in >25%to note that gastroduodenal damage develops earlyof NSAID users.[83] Therefore, long-term compli-(<1 month) in NSAID therapy.[2,32] In addition, itance with this therapy would be expected to be morehas been observed that 40–58% of UGB episodesdifficult compared with that of a standard regimenrelated to use of these drugs occur without previous(1 tablet/day) of PPI in the morning, especially indyspeptic symptoms.[76,77] Therefore, it is of para-elderly patients who are already taking severalmount importance that patients with increased riskdrugs.[5,14,16] Moreover, PPI therapy appears to beof developing UGB promptly receive adequate gas-superior to misoprostol in terms of preventing duo-troduodenal protective therapy, commencing at thedenal ulcers.[84] In a trial enrolling 3111 elderlysame time as NSAID therapy. This procedure haspatients focused on gastroprotection, it was foundbeen specifically endorsed in several society andthat PPI therapy was significantly associated with anational treatment guidelines, including US, UK andreduced risk of peptic ulcer in both short- (OR: 0.70)Dutch guidelines.[3,78,79] Regrettably, these welland long-term (OR: 0.32) NSAID users, and that therecognised recommendations appear to be followednumber needed to treat to avoid one peptic ulcer wasonly rarely in clinical practice, as reported in The3 in both short- and long-term users.[33] On theNetherlands.[75] According to a recent study con-contrary, concomitant treatment with histamine H2ducted in Turkey, only 2% of high-risk NSAIDreceptor antagonists was associated with a signifi-users actually received adequate gastroprotec-cantly higher risk of peptic ulcer both in short- (OR:tion.[80] In Italy, protective therapy is dispensed10.9) and long-term (OR: 6.26) NSAID users.[33]

without charge for every high-risk patient (age >75Unsatisfactory efficacy of H2 receptor antagonistyears, or younger if concomitant corticosteroid ortherapy, even when the dose is doubled, has beenanticoagulant therapy are being taken or the patientconfirmed in other trials.[29,43]

has had a previous ulcer resistant to H. pylori eradi-cation). This approach would appear to be safe, Some aspects should be considered when choos-notwithstanding the fact that the risk of UGB is ing a PPI to treat elderly patients. In two comprehen-already significantly increased after 65 years of age. sive reviews of the literature it has been found that


824 Zullo et al.

omeprazole, pantoprazole and lansoprazole are ef- 3.2 Management of H. pylori Infection

fective; however, they also strongly interfere withThere is no doubt that all patients with H. pylori-cytochrome 450 enzyme 2C19, resulting in possible

positive peptic ulcers must be treated for the infec-interactions with other drugs frequently used bytion. Indeed, bacterial eradication strongly preventsaged patients.[85,86] Rabeprazole and esomeprazoleboth ulcer recurrence and complications.[55,56] Ahave better pharmacokinetic and pharmacodynamicmeta-analysis of nine case-control trials which en-profiles and a more rapid onset of action thanrolled a total of 825 patients found that H. pyloriomeprazole, pantoprazole and lansoprazole.[85,86]

eradication significantly reduced the recurrence ofAnother suggested approach could be to switch UGB compared with either no antisecretory treat-

NSAIDs to selective COX-2 inhibitor therapy in ment (4.6% vs 23.7%, respectively; OR: 0.18) orhigh-risk patients when other cardiovascular contra- maintenance antisecretory therapy (1.6% vs 5.6%,indications are absent.[45] A systematic review of the respectively; OR: 0.25).[60] Moreover, when onlyliterature found that the cost of therapy would be patients with documented successful H. pylori eradi-lower when high-risk patients were treated with cation were included, the re-bleeding rate was as

low as 0–1%.[60,88]selective COX-2 inhibitors, whilst no advantage wasSeveral studies have assessed the role of H. pylo-observed when treating low-risk patients.[87] There-

ri eradication in the prevention of peptic ulcer orfore, it has been suggested that NSAIDs should beulcer complications in NSAID users. Data from fiveused in low-risk patients and selective COX-2 inhib-well designed trials that enrolled a total of 939itors in high-risk patients. However, high-risk pa-patients were systematically evaluated in a recenttients with a prior gastrointestinal event did notmeta-analysis.[89] It was found that H. pylori eradi-experience a significant reduction in the incidencecation significantly reduced the risk of development

of new ulcer complications with use of selectiveof both peptic ulcer (OR: 0.43) and ulcer bleeding

COX-2 inhibitor therapy.[68] Moreover, in patients(OR: 0.13) in these patients, although it was inferior

with a previous UGB, celecoxib therapy did notto PPI maintenance therapy in terms of reducing

significantly reduce bleeding recurrence compared ulcer bleeding (OR: 7.43). Moreover, sub-analysiswith omeprazole and diclofenac therapy (4.9% vs showed that the risk of peptic ulcer was significantly6.4%, respectively).[54] Finally, 30-day mortality reduced only in NSAID-naive patients (OR: 0.26),rates following UGB related to either selective not in those already taking anti-inflammatory ther-COX-2 inhibitors (celecoxib, rofecoxib) or standard apy (OR: 0.95).[89]

NSAIDs (ibuprofen, diclofenac, naproxen) therapy The available data suggest that three factorsare not significantly different.[35] Further data are should be taken into account before prescribingtherefore needed in relation to this approach. NSAID therapy in high-risk patients: type of drug

Table II. Suggested gastroduodenal protective strategies before commencing NSAID therapy in high-risk patients

Drug Duration (months) PUD history Strategy

NSAID or aspirin (acetylsalicylic acid) <3 No PPI co-therapy

NSAID or aspirin >3–6 No Test and treat for Helicobacter pylori

Aspirin >3–6 Yes Test and treat for H. pylori

NSAID >3–6 Yes Test and treat for H. pylori plus PPI maintenancetherapy

PPI = proton pump inhibitor; PUD = peptic ulcer disease.



(NSAID/aspirin), history of the patient (previous the physician, but considerable evidence relating toulcer/ulcer complication) and duration of therapy.[90] use of a safer and cheaper strategy for both low- andA reasonable approach that takes into account these high-risk patients is now available. Although severalthree factors is outlined in table II. High-risk pa- society and national guidelines for prevention oftients already receiving NSAIDs should receive PPI UGB have been formulated, there is evidence thatco-therapy. these are rarely followed in clinical practice in dif-

ferent countries. This situation is ethically unaccept-How should H. pylori infection be cured in elder-able in medical care. Eliminating ignorance is prob-ly patients? Several studies have shown that theably the best and cheapest way to stop the bleeding!H. pylori eradication rate following standard triple

therapies is decreasing worldwide.[91] We have pro-Acknowledgementsposed a novel 10-day sequential regimen, consisting

of 5 days of dual therapy (PPI plus amoxicillin) No sources of funding were used to assist in the prepara-followed by 5 days of triple therapy (PPI, clarithro- tion of this article. The authors have no conflicts of interest

that are directly relevant to the content of this article.mycin and tinidazole), which has been associatedwith a very high eradication rate in adult and paedia-

Referencestric patients,[91,92] including those harbouring clar-1. Lanas A, Perez-Aisa MA, Feu F, et al. A nationwide study ofithromycin-resistant strains.[93] We also conducted a

mortality associated with hospital admission due to severetrial of 179 elderly patients (mean age: 69.5 years; gastrointestinal events and those associated with nonsteroidal

antiinflammatory drug use. Am J Gastroenterol 2005; 100:range: 65–83) with duodenal ulcer and H. pylori1685-93

infection, in which the same sequential therapy 2. Laporte JR, Ibanez L, Vidal X, et al. Upper gastrointestinalbleeding associated with the use of NSAIDs. Drug Saf 2004;achieved a bacterial eradication rate significantly27: 411-20higher than that of standard triple therapy (94.4% vs

3. Palmer KR. Non-variceal upper gastrointestinal haemorrhage:80%, respectively), with similarly high compliance guidelines. Gut 2002; 51 Suppl. IV: 1-6

4. Cebollero-Santamaria F, Smith J, Gioe S, et al. Selective outpa-and a low incidence of adverse effects.[94] Thesetient management of upper gastrointestinal bleeding in the

results suggest that this sequential regimen could be elderly. Am J Gastroenterol 1999; 94: 1242-7used as first-line therapy for H. pylori infection in 5. Sandel MH, Kolkman JJ, Kuiper EJ, et al. Nonvariceal upper

gastrointestinal bleeding: differences in outcome for patientsaged patients.admitted to internal medicine and gastroenterological services.Am J Gastroenterol 2000; 95: 2357-62

6. Wolfe MM, Lichtenstein DR, Singh G. Gastrointestinal toxicity4. Conclusionof nonsteroidal antiinflammatory drugs. N Engl J Med 1999;340: 1888-99

UGB is a dramatic event with both a high mortal- 7. Marshall JK, Collins SM, Gafni A. Prediction of resource utili-zation and case cost for acute nonvariceal upper gastrointesti-ity rate and a substantial cost for healthcare systemsnal hemorrhage at a Canadian community hospital. Am Jworldwide. Several studies have found that age is anGastroenterol 1999; 94: 1841-6

independent predisposing condition for gastrointes- 8. Kang JY, Elders A, Majeed A, et al. Recent trends in hospitaladmissions and mortality rates for peptic ulcer in Scotlandtinal bleeding, the risk being significantly increased1982–2002. Aliment Pharmacol Ther 2006; 24: 65-79

in patients aged >65 years and further increased in 9. Kemppainen H, Raiha I, Sourander L. Clinical presentation ofpeptic ulcer in the elderly. Gerontology 1997; 43: 283-8those aged >75 years. Moreover, UGB incidence

10. Pilotto A. Aging and the gastrointestinal tract. Ital J Gastroenter-and mortality are distinctly higher in elderly patientsol Hepatol 1999; 31: 137-53

with co-morbidities. NSAID therapy and H. pylori 11. Newton JL. Effect of age-related changes in gastric physiologyon tolerability of medications for older people. Drugs Aginginfection are the main aetiopathogenetic factors in-2005; 22: 655-61

volved in the development of UGB. Preventing 12. Marshall BJ. Helicobacter pylori. Am J Gastroenterol 1994; 89bleeding from peptic ulcers remains a challenge for Suppl. 1: S116-28


826 Zullo et al.

13. Fries JF, Williams CA, Bloch DA, et al. Nonsteroidal antiin- inflammatory drugs: a case-control study. Gastroenterologyflammatory drug-associated gastropathy: incidence and risk 1999; 116: 1305-9factor models. Am J Med 1991; 104: 213-22 32. Pilotto A, Franceschi M, Leandro G, et al. The risk of upper

14. Pilotto A, Franceschi M, Leandro G, et al. NSAID and aspirin gastrointestinal bleeding in elderly users of aspirin and otheruse by the elderly in general practice: effect on gastrointestinal non-steroidal anti-inflammatory drugs: the role of gastropro-symptoms and therapies. Drugs Aging 2003; 20: 701-10 tective drugs. Aging Clin Exp Res 2003; 15: 494-9

15. Nygaard HA, Naik M. Drug use in homes for the aged: a 33. Pilotto A, Franceschi M, Leandro G, et al. Proton-pump inhibi-comparison between mentally intact and mentally impaired tors reduce the risk of uncomplicated peptic ulcer in elderlyresidents. Aging Clin Exp Res 1999; 11: 186-93 either acute or chronic users of aspirin/non-steroidal anti-

16. Doucet J, Chassagne P, Trivalle C, et al. Drug-drug interactions inflammatory drugs. Aliment Pharmacol Ther 2004; 20: 1091-related to hospital admission in older adults: a prospective 7study of 1000 patients. J Am Geriatr Soc 1996; 44: 944-8 34. Langman MJS, Weil J, Wainwright P, et al. Risks of bleeding

17. Næsdal J, Brown K. NSAID-associated adverse effects and acid peptic ulcer associated with individual non-steroidal anti-in-control aids to prevent them: a review of current treatment flammatory drugs. Lancet 1994; 343: 1075-8options. Drug Saf 2006; 29: 119-32 35. Thomsen RW, Riis A, Christensen S, et al. Outcome of peptic

18. Davis MP, Srivastava M. Demographics, assessment and man- ulcer bleeding among users of traditional non-steroidal anti-agement of pain in the elderly. Drugs Aging 2003; 20: 23-57 inflammatory drugs and selective cyclo-oxygenase-2 inhibi-

19. Sarchielli P, Mancini ML, Calabresi P. Practical considerations tors. Aliment Pharmacol Ther 2006; 24: 1431-8for the treatment of elderly patients with migraine. Drugs 36. Konturek KJ. Role of growth factors in gastroduodenal protec-Aging 2006; 23: 461-89 tion and healing of gastric ulcer. Gastroenterol Clin North Am

20. Mah´e I, Leizorovicz A, Caulin C, et al. Aspirin for the preven- 1990; 19: 41-65tion of cardiovascular events in the elderly. Drugs Aging 2003; 37. Kelly SM, Jenner JR, Dickinson RJ, et al. Increased gastric juice20: 999-1010 epidermal growth factor after non-steroidal antiinflammatory

21. Savage RL, Moller PW, Ballantyne CL, et al. Variation in the drug ingestion. Gut 1994; 35: 611-4risk of peptic complications with nonsteroidal antiinflammato-

38. Lipscomb GR, Campbell F, Rees WDW. The influence of age,ry drug therapy. Arthritis Rheum 1993; 36: 89-94

gender, Helicobacter pylori and smoking on gastric mucosal22. Wilcox CM, Cryer B, Triadafilopoulos G. Patterns of use and

adaptation to nonsteroidal antiinflammatory drugs. Alimentpublic perception of over-the-counter pain relievers: focus on

Pharmacol Ther 1997; 11: 907-12nonsteroidal antiinflammatory drugs. J Rheumatol 2005; 32:

39. Laine L. Review article: gastrointestinal bleeding with low-dose2218-24aspirin: what’s the risk? Aliment Pharmacol Ther 2006; 24:

23. FDA Drug Bull 1989; 19: 3-4897-908

24. Hansen JM, Hallas J, Lauritsen JM, et al. Non-steroidal anti-40. Sorensen HT, Mellemkjaer L, Blot WJ, et al. Risk of upperinflammatory drugs and ulcer complications: a risk factor

gastrointestinal bleeding associated with use of low-dose aspi-analysis for clinical decision-making. Scand J Gastroenterolrin. Am J Gastroenterol 2000; 95: 2218-241996; 31: 126-30

41. Ligumski M, Sestrieri M, Karmeli F, et al. Rectal administration25. Gabriel SE, Jaakkimainen L, Bombardier C. Risk for seriousof nonsteroidal antiinflammatory drugs. Gastroenterologygastrointestinal complications related to use of nonsteroidal1990; 98: 1245-9anti-inflammatory drugs: a meta-analysis. Ann Intern Med

42. Hunt RH, Bazzoli F. Review article: should NSAID/low-dose1991; 115: 787-96aspirin takers be tested routinely for H. pylori infection and26. Sebastian JL, Tresch DD. Use of oral anticoagulants in oldertreated if positive? Implications for primary risk of ulcer andpatients. Drugs Aging 2000; 16: 409-35ulcer relapse after initial healing. Aliment Pharmacol Ther27. Ibanez L, Vidal X, Vendrell L, et al. Upper gastrointestinal2004; 19 Suppl. 1: 9-16bleeding associated with antiplatelet drugs. Aliment

43. Chan FKL, Graham DY. Review article: prevention of non-Pharmacol Ther 2006; 23: 235-43steroidal anti-inflammatory drug gastrointestinal complica-28. Hawkey CJ. Risk of ulcer bleeding in patients infected withtions: review and recommendations based on risk assessment.Helicobacter pylori taking non-steroidal anti-inflammatoryAliment Pharmacol Ther 2004; 19: 1051-61drugs. Gut 2000; 46: 310-1

44. Silverstein FE, Graham DY, Senior JR, et al. Misoprostol29. Lanas A, Garcia-Rodriguez LA, Arroyo MT, et al. Risk of upperreduces serious gastrointestinal complications in patients withgastrointestinal ulcer bleeding associated with selective cyclo-rheumatoid arthritis receiving nonsteroidal anti-inflammatoryoxygenase-2 inhibitors, traditional non-aspirin non-steroidaldrugs: a randomized, double-blind, placebo-controlled trial.anti-inflammatory drugs, aspirin and combinations. Gut 2006;Ann Intern Med 1995; 123: 241-955: 1731-8

45. Savage R. Cyclo-oxygenase-2 inhibitors. When should they be30. Kaufman DW, Kelly JP, Wiholm BE, et al. The risk of acuteused in the elderly? Drugs Aging 2005; 22: 185-200major upper gastrointestinal bleeding among users of aspirin

and ibuprofen at various levels of alcohol consumption. Am J 46. National Institute for Clinical Excellence. Guidance on the useGastroenterol 1999; 94: 3189-96 of selective inhibitors, celecoxib, rofecoxib, meloxicam and

31. Aalykke C, Lauritsen JM, Hallas J, et al. Helicobacter pylori etodolac for osteoarthritis and rheumatoid arthritis: technologyand risk of ulcer bleeding among users of nonsteroidal anti- no. 27. London: Government Publication, 2001



47. Clemett D, Goa KL. Celecoxib: a review of its use in osteoar- 62. Huang JQ, Sridhar S, Hunt RH. Role of Helicobacter pylorithritis, rheumatoid arthritis and acute pain. Drugs 2000; 59: infection and non-steroidal anti-inflammatory drugs in peptic957-80 ulcer disease: a meta-analysis. Lancet 2002; 359: 14-22

63. Pilotto A, Franceschi M, Leandro G, et al. The effect of48. Silverstein FE, Faich G, Goldstein JL, et al. GastrointestinalHelicobacter pylori infection on NSAID-related gas-toxicity with celecoxib vs nonsteroidal anti-inflammatorytroduodenal damage in the elderly. Eur J Gastroenteroldrugs for osteoarthritis and rheumatoid arthritis: the CLASSHepatol 1997; 9: 951-6study: a randomized controlled trial. JAMA 2000; 284: 1247-

64. Pilotto A, Leandro G, Di Mario F, et al. Role of Helicobacter55pylori infection on upper gastrointestinal bleeding in the elder-49. Bombardier C, Laine L, Reicin A, et al. Comparison of upperly: a case-control study. Dig Dis Sci 1997; 42: 586-91gastrointestinal toxicity of rofecoxib and naproxen in patients

65. Okan A, Tankurt E, Aslan BU, et al. Relationship between non-with rheumatoid arthritis. N Engl J Med 2000; 343: 1520-30steroidal anti-inflammatory drug use and Helicobacter pylori

50. Schnitzer TJ, Burmester GR, Mysler E, et al. Comparison ofinfection in bleeding or uncomplicated peptic ulcers: a case-

lumiracoxib with naproxen and ibuprofen in the Therapeuticcontrol study. J Gastroenterol Hepatol 2003; 18: 18-25

Arthritis Research and Gastrointestinal Event Trial (TAR-66. Lanas A, Fuentes J, Benito R, et al. Helicobacter pylori in-GET), reduction in ulcer complications: randomised controlled

creases the risk of upper gastrointestinal bleeding in patientstrial. Lancet 2004; 364: 665-74taking low-dose aspirin. Aliment Pharmacol Ther 2002; 16:

51. Mamdani M, Rochon PA, Juurlink DN, et al. Observational 779-86study of upper gastrointestinal haemorrhage in elderly patients

67. Goldstein JL, Correa P, Zhao WW, et al. Reduced incidence ofgiven selective cyclo-oxygenase-2 inhibitors or conventional

gastroduodenal ulcers with celecoxib, a novel cyclooxyge-non-steroidal anti-inflammatory drugs. BMJ 2002; 325: 624-9

nase-2 inhibitor, compared to naproxen in patients with arthri-52. Cannon CP, Curtis SP, Fitzgerald GA, et al. Cardiovascular tis. Am J Gastroenterol 2001; 96: 1019-27

outcomes with etoricoxib and diclofenac in patients with os- 68. Laine L, Bombardier C, Hawkey CJ, et al. Stratifying the risk ofteoarthritis and rheumatoid arthritis in the Multinational Etor- NSAID-related upper gastrointestinal clinical events: results oficoxib and Diclofenac Arthritis Long-term (MEDAL) pro- a double-blind outcomes study in patients with rheumatoidgramme: a randomised comparison. Lancet 2006; 368: 1771- arthritis. Gastroenterology 2002; 123: 1006-1281 69. Verhamme KMC, Mosis G, Dieleman JP, et al. Spironolactone

53. Juni P, Rutjes AWS, Dieppe PA. Are selective COX 2 inhibitors and risk of upper gastrointestinal events: population basedsuperior to traditional non steroidal anti-inflammatory drugs? case-control study. BMJ. Epub 2006 Jul 13Adequate analysis of the CLASS trial indicates that this may 70. Yuan Y, Tsoi K, Hunt RH. Selective serotonin reuptake inhibi-not be the case. BMJ 2002; 324: 1287-8 tors and risk of upper GI bleeding: confusion or confounding?

54. Chan FKL, Hung LCT, Suen BY, et al. Celecoxib versus Am J Med 2006; 119: 719-27diclofenac and omeprazole in reducing the risk of recurrent 71. Kaplan RC, Heckbert SR, Koepsell TD, et al. Use of calciumulcer bleeding in patients with arthritis. N Engl J Med 2002; channel blockers and risk of hospitalized gastrointestinal tract347: 2104-10 bleeding. Arch Intern Med 2000 Jun 26; 160 (12): 1849-55

72. Atar S, Cannon CP, Murphy SA, et al. Statins are associated55. Vaira D, Gatta L, Ricci C, et al. Helicobacter pylori: diseases,with lower risk of gastrointestinal bleeding in patients withtests and treatment. Dig Liver Dis 2001; 33: 788-94unstable coronary syndromes: analysis of the Orbofiban in56. Bazzoli F, Bianchi Porro G, Maconi G, et al. Treatment ofPatients with Unstable coronary Syndromes-Thrombolysis InHelicobacter pylori infection: indications and regimens: anMyocardial Infarction 16 (OPUS-TIMI 16) trial. Am Heart Jupdate. Dig Liver Dis 2002; 34: 70-832006; 151: 1-6

57. Pilotto A, Franceschi M, Di Mario F, et al. The long-term73. Leenen FH, Nwachuku CE, Black HR, et al. Clinical events inclinical outcome of elderly patients with Helicobacter pylori-

high-risk hypertensive patients randomly assigned to calciumassociated peptic ulcer disease. Gerontology 1998; 44: 153-8channel blocker versus angiotensin-converting enzyme inhibi-

58. Mignon M, Penston JG, Deltenre M, et al. Natural history of tor in the Antihypertensive and Lipid-Lowering treatment toduodenal ulcer disease: Are we at a turning point? Gastroenter- prevent Heart Attack Trial. Hypertension 2006; 48: 374-84ol Int 1994; 7: 95-113

74. Celis H, Thijs L, Staessen JA, et al. Interaction between nonster-59. Gisbert JP, Gonzalez L, de Pedro A, et al. Helicobacter pylori oidal anti-inflammatory drug intake and calcium-channel

and bleeding duodenal ulcer: prevalence of the infection and blocker-based antihypertensive treatment in the Syst-Eur trial.role of non-steroidal anti-inflammatory drugs. Scand J Gas- J Hum Hypertens 2001; 15: 613-8troenterol 2001; 36: 717-24 75. Sturkenboom MCJM, Burke TA, Dieleman JP, et al. Underu-

60. Gisbert JP, Khorrami S, Carballo F, et al. Meta-analysis: tilization of preventive strategies in patients receivingHelicobacter pylori eradication therapy vs. antisecretory non- NSAIDs. Rheumatology 2003; 42 Suppl. 3: 23-31eradication therapy for the prevention of recurrent bleeding 76. Armstrong CP, Blower AL. Non-steroidal anti-inflammatoryfrom peptic ulcer. Aliment Pharmacol Ther 2004; 19: 617-29 drugs and life threatening complications of peptic ulceration.

Gut 1987; 28: 527-3261. Fiorucci S, Romano M. Helicobacter pylori and non-steroidalanti-inflammatory drugs. Dig Liver Dis 2000; 32 Suppl. 3: 77. Laszlo A, Kelly JP, Kaufman DE, et al. Clinical aspects of upperS211-3 gastrointestinal bleeding associated with the use of nonster-


828 Zullo et al.

oidal antiinflammatory drugs. Am J Gastroenterol 1998; 93: 88. Labenz J, Borsch G. Role of Helicobacter pylori eradication in721-5 the prevention of peptic ulcer bleeding relapse. Digestion

1994; 55: 19-2378. Recommendations for the medical management of osteoarthritisof the hip and knee: 2000 update. American College of Rheu- 89. Vergara M, Catalan M, Gisbert JP. Meta-analysis: role ofmatology Subcommittee on Osteoarthritis Guidelines. Arthri-

Helicobacter pylori eradication in the prevention of peptictis Rheum 2000; 43: 1905-15

ulcer in NSAID users. Aliment Pharmacol Ther 2005; 21:79. Moens HJ, van Croonenborg JJ, Al MJ, et al. Guideline: NSAID

1411-8use and the prevention of gastric damage. Ned Tijdschr

90. Sung JJY. Should we eradicate Helicobacter pylori in non-Geneeskd 2004; 148: 604-8steroidal anti-inflammatory drug users? Aliment Pharmacol80. Dincer D, Duman A, Dikici H, et al. NSAID-related upperTher 2004; 20 Suppl. 2: 65-70gastrointestinal bleeding: are risk factors considered during

prophylaxis? Int J Clin Pract 2006; 60: 546-8 91. Zullo A, Vaira D, Vakil N, et al. High eradication rates of81. Lanas A, Ferrandez A. NSAID-induced gastrointestinal dam- Helicobacter pylori with a new sequential treatment. Aliment

age: current clinical management and recommendations for Pharmacol Ther 2003; 17: 719-26prevention. Chin J Dig Dis 2006; 7: 127-33

92. Francavilla R, Lionetti E, Castellaneta SP, et al. Improved82. Micklewright R, Lane S, Linley W, et al. Review article: efficacy of 10-day sequential treatment for Helicobacter pylori

NSAIDs, gastroprotection and cyclo-oxygenase-II-selectiveeradication in children: a randomized trial. Gastroenterologyinhibitors. Aliment Pharmacol Ther 2003; 17: 321-322005; 129: 1414-9

83. Hawkey CJ. Non-steroidal anti-inflammatory drugs: who93. De Francesco V, Margiotta M, Zullo A, et al. Clarithromycin-should receive prophylaxis? Aliment Pharmacol Ther 2004; 20

resistant genotypes and eradication of Helicobacter pylori.Suppl. 2: 59-64Ann Intern Med 2006; 144: 94-10084. Hawkey CJ, Karrasch JA, Szczepanski L, et al. Omeprazole

compared with misoprostol for ulcers associated with nonster- 94. Zullo A, Gatta L, De Francesco V, et al. High rate ofoidal antiinflammatory drugs. N Engl J Med 1998; 338: 727- Helicobacter pylori eradication with sequential therapy in34

elderly patients with peptic ulcer: a prospective controlled85. Thjodleifsson B. treatment of acid-related diseases in the elderly study. Aliment Pharmacol Ther 2005; 21: 1419-24

with emphasis on the use of proton pump inhibitors. DrugsAging 2002; 19: 911-27

86. Klotz U. Pharmacokinetic considerations in the eradication of Correspondence: Dr Angelo Zullo, Gastroenterologia ed En-Helicobacter pylori. Clin Pharmacokinet 2000; 38: 243-70

doscopia Digestiva, Ospedale Nuovo Regina Margherita,87. Rahme E, Barkun AN, Adam V, et al. Treatment costs to

Via E. Morosini, 30, 00153 Rome, Italy.prevent or treat upper gastrointestinal adverse events associat-ed with NSAIDs. Drug Saf 2004; 27: 1019-42 E-mail: [email protected]


Bleeding Peptic Ulcer in the Elderly

Documents

Transcript of Bleeding Peptic Ulcer in the Elderly