Think before or sink after: choosing an appropriate nsaid by balancing gastrointestinal and cardiovascular risks

Think Before or Sink After: Choosing an
Appropriate NSAID by Balancing
Gastrointestinal and Cardiovascular Risks
Jyh-Ming Liou,1,2 Ming-Shiang Wu,1* Jaw-Town Lin1,3
Nonsteroidal anti-inflammatory drugs (NSAIDs) are introduced to the market with popularity be- have become one of the most important drugs cause they are associated with fewer gastrointesti- for relieving pain and reducing inflammation. nal (GI) complications than non-selective NSAIDs.
It was reported that more than 111,000,000 pre- Unfortunately, rofecoxib and valdecoxib were scriptions for NSAIDs were prescribed in the withdrawn from the market because of their seri- United States at a cost of $5 billion each year.1 ous cardiovascular risks. Emerging data suggest Most NSAIDs derive their antipyretic, analgesic, that nonselective NSAIDs, except for naproxen, and anti-inflammatory effects through inhibi- are also associated with increased risks for cardio- tion of cyclooxygenase (COX), including COX-1 vascular events.8 Thus, the choice of appropriate and COX-2, and hence the synthesis of prosta- glandins. NSAIDs and other analgesic/antipyretic drugs such as acetaminophen might also act through inhibition of a newly discovered cyclooxy- genase isoenzyme, COX-3, to reduce pain and pos- sibly fever.2 In recent decades, aspirin has also been widely used for the prevention of cardiovas- cular disease. In addition, COX-2 inhibitors have GI intolerance is the most common adverse ef- the potential for chemoprevention of gastric and fect of NSAIDs, and a meta-analysis has revealed that NSAIDs increased the risk of dyspepsia by However, the anti-inflammatory and analgesic 36%.9 In the United States, NSAID-associated effects of NSAIDs are not obtained without cost.
upper GI adverse events are estimated to result in NSAIDs are an important cause of bleeding pep- 103,000 hospitalizations and 16,500 deaths per tic ulcers in countries where the prevalence of year.10 NSAID users had a 12.5 per 1000 person- Helicobacter pylori infection is decreasing. A recent years excess rate of ulcer-related hospitalization study from Taiwan revealed that about 57.2% of compared to non-NSAID users.11 A nationwide patients with bleeding peptic ulcer reported re- study of mortality associated with hospital admis- cent use of NSAIDs or antiplatelet agents.7 NSAIDs sion due to severe GI events in Spain revealed that that are selective for COX-2 inhibition (coxibs) the death rate attributed to NSAID/aspirin use was 1Department of Internal Medicine, College of Medicine, and 2Graduate Institute of Epidemiology, College of PublicHealth, National Taiwan University, Taipei, and 3Department of Internal Medicine, E-Da Hospital/I-Shou University,Kaohsiung, Taiwan.
*Correspondence to: Professor Ming-Shiang Wu, Departments of Internal Medicine and Primary Care Medicine,
National Taiwan University, 7 Chung-Shan South Road, Taipei 100, Taiwan.
E-mail: mingshiang@ntu.edu.tw
15.3 deaths/100,000 NSAID/aspirin users, and hypothesis that anti-inflammatory effects are usu- that up to one third of all NSAID/aspirin deaths ally due to COX-2 inhibition and that adverse can be attributed to low-dose aspirin use.12 In an effects usually occur because of COX-1 inhibi- endoscopic evaluation of patients who had contin- tion, selective COX-2 inhibitors (celecoxib, rofe- uously used NSAIDs over the previous 6 months, coxib, valdecoxib, etc.) were developed to reduce gastroduodenal ulcers were detected in 24% of NSAID-associated GI toxicities.19 Several clinical patients, and approximately 1–2% of NSAID users trials showed a 41–57% reduction in the rate of developed ulcer-related complications (bleeding, GI toxicities with the use of selective COX-2 in- perforation, obstruction) annually.13,14 Notably, hibitors.20 However, the VIGOR trial raised the the majority of patients with NSAID-related GI issue of the cardiovascular safety of the coxibs complications did not have preceding abdomi- after a statistically insignificant increase in the nal symptoms. The first sign of an ulcer was a life- incidence of myocardial infarctions in patients threatening complication in 58.2% of patients on rofecoxib was found.20 The first well-known cardiovascular toxicities of coxibs arose from che- Nonselective NSAIDs increase the risk of GI moprevention trials for colorectal polyps, when bleeding not only in the upper GI tract but also large doses and longer durations of treatment in the lower GI tract. Wilcox and Clark reported were needed. In the APPROVe trial, a significantly that the odds ratio (OR) for NSAID-associated increased risk (relative risk, 1.97) of cardiovascu- upper and lower GI bleeding were 3.2 and 2.6, lar events as compared to placebo was shown.21 respectively.17 NSAID-related GI toxicities are re- The results led to the withdrawal of rofecoxib in lated to direct irritation of GI mucosa and re- 2004. A meta-analysis including 17 case-controlled duction of protective prostaglandins through the studies and six cohort studies revealed a dose- inhibition of COX-1. The risk of peptic ulcer related cardiovascular risk for rofecoxib.8 The varies according to duration of therapy, dosage ORs for the cardiovascular risks of rofecoxib of drugs, and type of NSAID.10–18 Weil et al re- were 1.33 and 2.19 for dosages of ≤ 25 mg/day ported that the OR for duodenal ulcer bleeding and > 25 mg/day, respectively.8 Celecoxib at the also increased with dosage of aspirin and other usual doses was not associated with an elevated NSAIDs.18 Longer duration of therapy was also risk of vascular occlusion, with a relative risk of associated with increased risk of GI toxicity.
1.06 (95% confidence interval [CI], 0.91–1.23).8 Indomethacin, ketoprofen and piroxicam appear This meta-analysis also raised serious questions to be associated with the highest prevalence of about the increased cardiovascular risks of di- GI toxicity, whereas ibuprofen and diclofenac clofenac, with a relative risk of 1.40 (95% CI, 1.16–1.70). Naproxen was not associated with an increased risk of cardiovascular events, with a relative risk of 0.97 (95% CI, 0.87–1.07).8 Taken together, the cardiovascular toxicities seem to vary depending on type, dosage and duration of NSAID treatment. Intriguingly, geographic or The discovery of COX-2 provided the basis for ethnic differences may also play a role. A recent the development of selective COX-2 inhibitors.
population-based analysis in Taiwanese adults COX-1 is a constitutively expressed enzyme which with long-term (≥ 180 days) use of NSAIDs re- mediates the synthesis of thromboxane A2 in ported that no significant differences in the risk platelets and the production of protective pros- of treatment-related cardiovascular events were taglandin of the gastric endothelium. COX-2 cat- observed between groups treated with nonselec- alyzes prostaglandin synthesis in the inflammatory tive NSAIDS (etodolac, nabumetone, ibuprofen, cells and leads to inflammation. Based on the taking various NSAIDs showed that serious upper GI complications were reduced by 40% among patients who received misoprostol compared to Owing to the potential serious GI complications those who received placebo (OR, 0.6; 95% CI, associated with nonselective NSAIDs, several 0.364–0.982).25 However, misoprostol was poorly strategies have been used to decrease the risks, tolerated because of diarrhea and related prob- including: (1) use of selective COX-2 inhibitors; lems.25 H2-blockers have also been used to pre- (2) eradication of H. pylori infection; and (3) co- vent NSAID-related ulcers. However, a randomized prescription of gastroprotective agents. Selective controlled trial revealed that omeprazole healed COX-2 inhibitors are indeed effective for reduc- and prevented ulcers more effectively than did ing GI adverse events, but their cardiovascular toxicities might restrict their utilization, espe- cially when large-dose and long-term use are Which strategy would be better in the
required. The other two strategies deserve further
prevention of GI toxicity?
In very high risk patients (e.g. previous ulcer bleeding induced by nonselective NSAIDs), the Eradication of H. pylori
most effective prevention strategy would be com- In a meta-analysis of 25 observational studies bination therapy. Chan et al found that celecoxib (8843 patients), Huang et al found that H. pylori plus esomeprazole was better than celecoxib alone infection and NSAIDs increase the risk of peptic for the prevention of recurrent ulcer bleeding and ulcers independently and have synergistic ef- could reduce the recurrent bleeding rate to 0%.27 fects.23 Compared with H. pylori-negative indi- A recent population-based, matched case-control viduals not taking NSAIDs, the ORs of ulcer were analysis consisting of 1382 NSAID/COX-2 users 18.1 for H. pylori-positive non-NSAID users, 19.4 with upper GI complications and 33,957 controls for H. pylori-negative NSAID users, and 61.1 for compared the effects of different strategies. The H. pylori-positive NSAID users.23 Another meta- results demonstrated that all of the commonly ac- analysis by Vergara and colleagues found that the cepted gastroprotective strategies (PPI, COX-2 in- incidence of peptic ulcer in the overall popula- hibitors, low-dose/high-dose misoprostol), either tion receiving NSAIDs was reduced after H. pylori alone or in combination, can reduce the risk of eradication (7.4%), as compared to the control upper GI complications in NSAID users.28 This group (13.3%).24 Sub-analyses further showed a study also confirmed that the combination of significant reduction in the risk of ulcer for non- COX-2 inhibitors with PPIs offers the greatest risk NSAID users (OR, 0.26) but not for NSAID users reduction. However, it was shown that celecoxib may be superior to the combination of nonselec- tive NSAIDs with a PPI. In contrast to this study, Co-prescription of gastroprotective agents
Chan et al found that celecoxib was as effective as The commonly used gastroprotective agents in- diclofenac plus omeprazole in patients with a re- clude misoprostol, H2-blocker, and proton pump cent history of ulcer bleeding.29 The probability inhibitors (PPI). Prophylactic use of antacids will of recurrent bleeding was 4.9% in the celecoxib not only not reduce the risk of GI toxicities, but group and 6.4% in the diclofenac plus omepra- may even mask the symptoms of subsequent se- zole group during the 6-month period.29 Another rious GI complications such as bleeding.16 Miso- important question is: would clopidogrel be bet- prostol is a synthetic prostaglandin E1 analog used ter than PPI plus low-dose aspirin in preventing in the prevention of NSAID-induced peptic ulcers.
recurrent bleeding? Chan et al reported that the A double-blind randomized controlled trial in cumulative incidence of recurrent bleeding was 8843 patients with rheumatoid arthritis who were 8.6% in the group treated with clopidogrel and 0.7% in the group treated with aspirin plus es- duration of NSAID use, when in the first few omeprazole during the 12-month period; they months of NSAID use, concomitant use of anti- concluded that the latter strategy is better.30 coagulants and corticosteroids, and other debili- Next, is co-prescription of PPI or H. pylori tating diseases.10–18 Cardiovascular risk may be eradication more effective in the prevention of assessed by the Framingham risk-score calculator, recurrent upper GI bleeding? Chan et al found which estimates a patient’s 10-year risk of devel- that for H. pylori-infected low-dose aspirin users, oping myocardial infarction and coronary death.
eradication therapy is equivalent to co-prescription The patient’s age, sex, smoking status, total and of PPI in the prevention of recurrent upper GI high-density lipoprotein cholesterol levels, sys- bleeding.31 However, for other NSAID users who tolic blood pressure, and whether or not the were H. pylori-infected, co-prescription of PPI is patient is on antihypertensive treatment are more effective than eradication therapy in the pre- vention of recurrent bleeding.31 A meta-analysis The general principles are that COX-2 in- including two randomized controlled trials also hibitor is preferred for patients with high GI risk, revealed that co-prescription of PPI might be more whereas naproxen is preferred for patients with effective (0%) than H. pylori eradication (2.6%) high cardiovascular risk. Gastroprotective agents in the prevention of recurrent ulcer bleeding.24 are recommended for patients receiving naproxen However, whether co-prescription of PPI plus or with high GI risk. The recommended strate- H. pylori eradication would be better than co- gies according to cardiovascular and GI risks are prescription of PPI alone in the primary prophy- shown in the Table.1,32 For patients with low car- laxis of ulcer bleeding remains unknown.
diovascular risk, the choices of NSAIDs and gas- troprotective agents can be managed according Strategies for choosing appropriate NSAID
to their GI risk. Patients with low GI risk (without according to cardiovascular and GI risks
GI risk factors as described above) can be treated It is recommended that a patient’s cardiovascular with nonselective NSAIDs alone. For patients with and GI risks be evaluated before the prescription medium GI risk, either COX-2 inhibitor alone or of NSAIDs.1,32 Risk factors for the development nonselective NSAID plus a PPI or misoprostol is of serious NSAID-related GI events include old appropriate. For patients with high GI risk, COX-2 age (> 60–65 years), history of peptic ulcer dis- inhibitor plus a PPI or misoprostol is recom- ease, H. pylori infection, higher dose or longer mended. Among patients with high cardiovascular Table. Recommendations on the use of NSAIDs according to CV and GI risks1,32
NSAIDs = nonsteroidal anti-inflammatory drugs; CV = cardiovascular; GI = gastrointestinal; COX-2 = cyclooxygenase-2. risk, naproxen plus a PPI or misoprostol are pre- ferred if they have low/medium GI risk. Low- dose COX-2 inhibitor alone is also acceptable 1. Graham DY, Chan FK. NSAIDs, risks, and gastroprotective for patients with low GI risk. For patients with strategies: current status and future. Gastroenterology 2008; high cardiovascular risk and high GI risk, it is rec- 2. Chandrasekharan NV, Dai H, Roos KL, et al. COX-3, a ommended that NSAIDs and COX-2 inhibitors cyclooxygenase-1 variant inhibited by acetaminophen and be avoided if possible. If anti-inflammatory ther- other analgesic/antipyretic drugs: cloning, structure, and apy is necessary, the choice of NSAID should be expression. Proc Natl Acad Sci USA 2002;99:13926–31.
based on the relative importance of the GI and 3. Arber N, Eagle CJ, Spicak J, et al. Celecoxib for the pre- cardiovascular risks of an individual patient. There vention of colorectal adenomatous polyps. N Engl J Med is currently no evidence to recommend which 4. Bertagnolli MM, Eagle CJ, Zauber AG, et al. Celecoxib for combination would be better for this group of the prevention of sporadic colorectal adenomas. N Engl J patients, but co-prescription of a PPI or miso- prostol is suggested. If the cardiovascular risk 5. Chang YJ, Wu MS, Lin JT, et al. Induction of cyclooxygenase-2 outweighs the GI risk, then naproxen is preferred overexpression in human gastric epithelial cells by Heli- cobacter pylori involves TLR2/TLR9 and c-Src-dependent over low-dose COX-2 inhibitor and vice versa.
nuclear factor-kappaB activation. Mol Pharmacol 2004; Randomized controlled trials are warranted to test the safety and efficacy of these recommendations 6. Chang YJ, Wu MS, Lin JT, et al. Helicobacter pylori-induced for patients with high cardiovascular and high invasion and angiogenesis of gastric cancer is mediated by COX-2 induction through TLR2/TLR9 and promoter regulation. J Immunol 2005;175:8242–52.
Liu NJ, Lee CS, Tang JH, et al. Outcomes of bleeding pep- Perspectives
tic ulcers: a prospective study. J Gastroenterol Hepatol There remain some unresolved questions regard- ing the prevention of NSAID-related GI toxici- 8. McGettigan P, Henry D. Cardiovascular risk and inhibition ties. First, whether or not routine screening for of cyclooxygenase: a systematic review of the observational and treatment of H. pylori infection before the studies of selective and nonselective inhibitors of cyclooxy- genase 2. JAMA 2006;296:1633–44.
use of NSAID is effective in the primary prophy- 9. Straus WL, Ofman JJ, MacLean C, et al. Do NSAIDs cause laxis of upper GI bleeding remains unknown dyspepsia? A meta-analysis evaluating alternative dyspep- because most of the previous studies used reduc- sia definitions. Am J Gastroenterol 2002;97:1951–8.
tion in recurrent bleeding (secondary prophy- 10. Singh G, Triadafilopoulos G. Epidemiology of NSAID in- laxis) as the end point. Second, whether or not duced gastrointestinal complications. J Rheumatol Suppl the application of the recommended strategies as 11. Smalley WE, Ray WA, Daugherty JR, et al. Nonsteroidal shown in the Table will reduce the occurrence anti-inflammatory drugs and the incidence of hospitaliza- (primary prophylaxis) of NSAID-related compli- tions for peptic ulcer disease in elderly persons. Am J cations is also not known. Third, some host genetic factors (such as variant CYP2C9*3 allele) 12. Lanas A, Perez-Aisa MA, Feu F, et al. A nationwide study have been reported to increase host susceptibility of mortality associated with hospital admission due to severe gastrointestinal events and those associated with to NSAID-related ulcer bleeding in Western pop- nonsteroidal antiinflammatory drug use. Am J Gastroenterol ulations.33,34 However, the variant CYP2C9*3 al- lele is very rare in Chinese populations.35 Further 13. Geis GS, Stead H, Wallemark CB, et al. Prevalence of mu- studies are warranted to identify the susceptible cosal lesions in the stomach and duodenum due to chronic genes in Chinese populations. If such genes can use of NSAID in patients with rheumatoid arthritis or osteoarthritis, and interim report on prevention by miso- be identified, personalized selection of NSAIDs prostol of diclofenac associated lesions. J Rheumatol Suppl and gastroprotective agents based on pharma- cogenomic approaches might become feasible in 14. Silverstein FE, Graham DY, Senior JR, et al. Misoprostol reduces serious gastrointestinal complications in patients with rheumatoid arthritis receiving nonsteroidal anti- with rheumatoid arthritis receiving nonsteroidal anti- inflammatory drugs. A randomized, double-blind, placebo- inflammatory drugs. A randomized, double-blind, placebo- controlled trial. Ann Intern Med 1995;123:241–9.
controlled trial. Ann Intern Med 1995;123:241–9.
15. Armstrong CP, Blower AL. Non-steroidal anti-inflammatory 26. Yeomans ND, Tulassay Z, Juhász L, et al. A comparison of drugs and life threatening complications of peptic ulcera- omeprazole with ranitidine for ulcers associated with non- steroidal antiinflammatory drugs. Acid Suppression Trial: 16. Singh G, Ramey DR, Morfeld D, et al. Gastrointestinal Ranitidine versus Omeprazole for NSAID-associated Ulcer tract complications of nonsteroidal anti-inflammatory drug Treatment (ASTRONAUT) Study Group. N Engl J Med treatment in rheumatoid arthritis. A prospective observa- tional cohort study. Arch Intern Med 1996;156:1530–6.
27. Chan FK, Wong VW, Suen BY, et al. Combination of a 17. Wilcox CM, Clark WS. Association of nonsteroidal anti- cyclo-oxygenase-2 inhibitor and a proton-pump inhibitor inflammatory drugs with outcome in upper and lower for prevention of recurrent ulcer bleeding in patients at gastrointestinal bleeding. Dig Dis Sci 1997;42:985–9.
very high risk: a double-blind, randomised trial. Lancet 18. Weil J, Colin-Jones D, Langman M, et al. Prophylactic aspirin and risk of peptic ulcer bleeding. BMJ 1995;310: 28. Targownik LE, Metge CJ, Leung S, et al. The relative effi- cacies of gastroprotective strategies in chronic users of 19. Vonkeman HE, Brouwers JR, van de Laar MA. Under- nonsteroidal anti-inflammatory drugs. Gastroenterology standing the NSAID related risk of vascular events. BMJ 29. Chan FK, Hung LC, Suen BY, et al. Celecoxib versus di- 20. Curfman GD, Morrissey S, Drazen JM. Expression of con- clofenac and omeprazole in reducing the risk of recurrent cern: Bombardier et al., “Comparison of upper gastroin- ulcer bleeding in patients with arthritis. N Engl J Med testinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis.” N Engl J Med 2000;343:1520–8. 30. Chan FK, Ching JY, Hung LC, et al. Clopidogrel versus as- N Engl J Med 2005;353:2813–4.
pirin and esomeprazole to prevent recurrent ulcer bleed- 21. Bresalier RS, Sandler RS, Quan H, et al. Cardiovascular ing. N Engl J Med 2005;352:238–44.
events associated with rofecoxib in a colorectal adenoma 31. Chan FK, Chung SC, Suen BY, et al. Preventing recurrent chemoprevention trial. N Engl J Med 2005;352:1092–102.
upper gastrointestinal bleeding in patients with Helicobacter 22. Huang WF, Hsiao FY, Wen YW, et al. Cardiovascular pylori infection who are taking low-dose aspirin or naproxen.
events associated with the use of four nonselective NSAIDs N Engl J Med 2001;344:967–73.
(etodolac, nabumetone, ibuprofen, or naproxen) versus a 32. Chan FK. The David Y. Graham lecture: use of nonsteroidal cyclooxygenase-2 inhibitor (celecoxib): a population-based antiinflammatory drugs in a COX-2 restricted environment.
analysis in Taiwanese adults. Clin Ther 2006;28:1827–36.
Am J Gastroenterol 2008;103:221–7.
23. Huang JQ, Sridhar S, Hunt RH. Role of Helicobacter pylori 33. Fries S, Grosser T, Price TS, et al. Marked interindividual infection and non-steroidal anti-inflammatory drugs in variability in the response to selective inhibitors of peptic-ulcer disease: a meta-analysis. Lancet 2002;359: cyclooxygenase-2. Gastroenterology 2006;130:55–64.
34. Pilotto A, Seripa D, Franceschi M, et al. Genetic sus- 24. Vergara M, Catalán M, Gisbert JP, et al. Meta-analysis: ceptibility to nonsteroidal anti-inflammatory drug-related role of Helicobacter pylori eradication in the prevention gastroduodenal bleeding: role of cytochrome P450 2C9 of peptic ulcer in NSAID users. Aliment Pharmacol Ther polymorphisms. Gastroenterology 2007;133:465–71.
35. Ma J, Yang XY, Qiao L, et al. CYP2C9 polymorphism in 25. Silverstein FE, Graham DY, Senior JR, et al. Misoprostol non-steroidal anti-inflammatory drugs-induced gastropathy.
reduces serious gastrointestinal complications in patients

Source: http://ir.lib.isu.edu.tw/retrieve/63210/10690.pdf