• 
• 

Three Part Question

In patients undergoing [cardiac surgery] is [clopidogrel superior to aspirin and a proton-pump inhibitor] for the prevention of [gastrointestinal bleeding or complications or peptic ulcers]?

Clinical Scenario

You are seeing a 72 year old man 8 days post-CABG. He had some melaena on day 2 and endoscopy showed a duodenal ulcer, which was injected. He has had no more symptoms or signs of continued bleeding and you would like to resume anti-platelet therapy. You have heard that clopidogrel is better for your stomach and decide to restart this instead of aspirin, but later that day the endoscopist reviews the patient and suggests that aspirin and lansoprazole would be safer and cheaper too. You resolve to check this in the literature.

Search Strategy

Medline 1966-November 2006 using the OVID interface.
[clopidogrel.mp] AND [exp Aspirin/ OR aspirin.mp] AND [exp Gastrointestinal Hemorrhage/ OR gastrointestinal bleed$.mp OR exp Peptic Ulcer Hemorrhage/ OR exp Peptic Ulcer Perforation/ OR exp Duodenal Ulcer/ OR exp Peptic Ulcer Hemorrhage/ OR exp Peptic Ulcer/ OR Gastrointestinal ulcer.mp OR exp Stomach Ulcer/ ]]

Search Outcome

A total of 40 papers were found. All relevant papers had their reference list crosschecked. From this search six papers were deemed to represent the best evidence on the topic.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Chan et al, 2005, Hong Kong	320 patients with previous GI bleeding on cardiovascular aspirin dose, and endoscopically proven ulcer healing, H. Pylori negative, no other antiplatelet treatment. Follow-up 12 months. Two groups: 159 patients 80mg aspirin plus 20mg esomeprazole twice daily versus 161 patients 75mg clopidogrel daily plus placebo.	Double-Blind PRCT ( Level 1b)	Peptic ulcer bleeding	13 patients in clopidogrel group (OR 8.6, CI 4.1-13.1) versus 1 patient in aspirin plus esomeprazole group (OR 0.7, CI 0-2.0) p=0.001, NNT 13)	No patients included with low-dose aspirin without prophylaxis. Results possibly not applicable to all patients in view of population characteristics. Study drugs were repacked maybe influencing uptake and resorption.
			Lower GI bleeding	7 patients in clopidogrel group (OR 4.6, CI 1.3-7.9) versus 7 patients in aspirin plus esomeprazole group (OR 4.6, CI 1.3-7.9) (p=0.98)
			Extra-intestinal bleeding	3 patients in clopidogrel group No patient in aspirin plus esomeprazole group
			Mortality	8 patients in clopidogrel group 4 patients in aspirin plus esomeprazole group
Ibanez et al, 2006, Spain and Italy	9841 study participants. Two groups: 2783 patients admitted for upper GI bleeding of which 546 patients on antiplatelet therapy alone, 130 patients on PPIs alone, 27 patients on antiplatelet therapy and PPIs. 7058 controls matched to centre, date of admission, gender, age. Of these 804 patients on antiplatelet therapy alone, 278 patients on PPIs alone, 49 patients on antiplatelet therapy and PPIs.	Case-Control Study ( Level 3b)	Upper GI bleeding	Antiplatelet drugs cause 14.5% of all upper GI bleedings467 patients and 642 controls on aspirin (OR 4.0, CI 3.2-4.9) versus 15 patients and 29 controls on aspirin and PPI (OR 1.1, CI 0.5-2.6)467 patients and 642 controls on aspirin (OR 4.0, CI 3.2-4.9) versus 15 patients and 29 controls on aspirin and PPI (OR 1.1, CI 0.5-2.6) As well decrease PPI's non-aspirin anti-platelet drugs from OR 2.1 (1.5-2.9) to OR 0.9 (0.4-2.3). 12 patients and 17 controls on clopidogrel (OR 2.3, CI 0.9-6.0). 12 patients and 17 controls on dipyridamole (OR 0.9, CI 0.4-2.0) 10 patients and 10 controls on indobufen (OR 3.8, CI 1.2-12.2) 52 patients and 55 controls on ticlopidine (OR 3.1, CI 1.8-5.1). Dose related increased risk on ticlopidine. 28 patients and 66 controls on triflusal (OR 1.8, CI 0.9-3.5)
Ng et al, 2004, Hong Kong	129 patients on low-dose aspirin (80-160mg daily) with active minor peptic ulcer disease. Two groups: All patients received 20mg omeprazole daily. 69 patients received 75mg clopidogrel daily and 60 patients continued their daily low-dose aspirin.	Single-Blind PRCT ( Level 1b)	Endoscopically proven ulcer healing within 8 weeks of treatment.	No evidence of bleeding, perforation or obstruction in either group. 4 patients with ongoing minor peptic ulcer disease or erosions in clopidogrel group versus 3 patients in the aspirin group.	Endpoint of 8 weeks may not be sufficient. Small numbers, and underpowered for treatment effect (minimum of 400 patients for each arm). Patients with high risk peptic ulcer disease were excluded.
Lai et al, 2006, Hong Kong	170 patients on low-dose aspirin with previous ulcer bleeding, negative for H. pylori, proven ulcer healing. Two groups: 86 patients received 100mg aspirin plus 20mg esomeprazole daily and 84 patients received 75mg clopidogrel daily.	Double-Blind PRCT ( level 1b)	Recurrent peptic ulcer complications within 12 months follow-up.	No patient in aspirin plus esomeprazole group had complications (CI 0) 9 patients in clopidogrel group (OR 13.6 CI 6.3-20.9) (p=0.0019)	Results possibly not applicable to all patients in view of population characteristics.
Ng et al, 2003, Hong Kong	70 patients previous history of non-aspirin peptic ulcer disease or aspirin related GI complications (dyspepsia or peptic ulcer) received 75mg clopidogrel daily. Follow-up 12 months.	Retrospective cohort study (Level 2b)	Re-occurrence of ulcer complications	9 patients (12%) with GI bleeding and 1 patient with perforated peptic ulcer. No patient was on maintenance PPI therapy. Clopidogrel increased risk in patients with previous GI bleeding versus patients without previous GI bleeding (22% versus 0%, p=0.007, OR 1.3, CI 1.1-1.5).	Small numbers, and underpowered for treatment effect. No control group. No standardised use of acid suppression because of retrospective study. Results possibly not applicable to all patients in view of population characteristics.
CAPRIE Study Organisation, 1996, USA + Europe,	19185 patients with previous ischaemic stroke, MI or atherosclerotic peripheral arterial disease. Two groups: 9586 patients received 325mg aspirin daily. 9599 patients received 75mg clopidogrel daily.	Multicentre Double-blind PRCT (level 1b)	Occurrence of ischaemic stroke, MI, vascular death	2800 outcome events. 1171 patients on clopidogrel and 1236 patients on aspirin with 1 event of whom 158 patients on clopidogrel and 182 patients on aspirin had >1 event. RR reduction for clopidogrel 8.7% (CI 0.3-16.5, p=0.043) Annual risk for clopidogrel 5.32% versus aspirin 5.83%.
			Any bleeding	890 patients (9.27%) on clopidogrel and 890 patients (9.28%) on aspirin (RR 1.00, CI 0.92-1.09).
			GI bleeding	191 patients (1.99%) on clopidogrel versus 255 patients (2.66%) on aspirin (RR 1.34, CI 1.11-1.61, p<0.05)

Comment(s)

Chan et al. performed a randomized controlled trial of 320 patients which was published in the New England Journal of Medicine. They showed that in patients with previous upper gastrointestinal (GI) bleeding there was a significant risk reduction if aspirin plus a proton-pump inhibitor (PPI) was used compared to clopidogrel alone (p=0.001, NNT 13). There was no risk reduction in lower GI bleeding which is likely to have been due to the effects of PPI's being confined to the upper GI. Chan et al. demonstrated significant superiority in upper GI bleeding prevention for aspirin plus a proton-pump inhibitor versus clopidogrel alone. Ibanez et al. performed a case control study of 9841 patients. This study, comparing patients with previous GI bleeding with matched controls, demonstrated an increased risk of upper GI bleeding in patients on aspirin (OR 4.0, CI 3.2-4.9) versus clopidogrel (OR 2.3, CI 0.9-6.0). There was a decreased risk of upper GI bleeding in patients on aspirin plus a PPI (OR 1.1, CI 0.5-2.6) versus patients on aspirin alone (OR 4.0, CI 3.2-4.9). PPI used simultaneously with a non-aspirin anti-platelet drug, showed a similar decrease in risk profile from OR 2.1 (1.5-2.9) to OR 0.9 (0.4-2.3). Therefore an increased risk of aspirin alone versus clopidogrel was demonstrated, but a decreased risk for aspirin plus a PPI. The risk of upper GI bleeding seems to be lowest with aspirin plus a PPI. The study from Ng et al. included 129 patients on low-dose aspirin with minor peptic ulcer disease. The trial compared aspirin in combination with PPI versus a group which discontinued aspirin and replaced this with clopidogrel plus PPI. The end-point was endoscopically proven ulcer healing within 8 weeks. There was no evidence of peptic ulcer complication in either group. Three patients in the aspirin arm had ongoing minor peptic ulcer disease compared to 4 patients in the clopidogrel group. This study suggests that both aspirin and clopidogrel plus a PPI are equally safe considering peptic ulcer complications. This study was unfortunately small, excluded high risk patients and only measured up to 8 weeks treatment effect and a bigger trial is therefore needed to verify the results with a longer follow-up time. Lai et al. included 170 patients on low-dose aspirin in their study with previous peptic ulcer bleeding. The patients were randomized into two groups. One group received aspirin plus a PPI and the other group received clopidogrel alone. None of the patients in the aspirin arm had peptic ulcer complications versus 9 patients in the clopidogrel group (OR 13.6 (CI 6.3-20.9), p=0.0019). In this trial aspirin plus a PPI was demonstrably superior to clopidogrel alone. Ng et al. retrospectively studied 70 patients with either a previous history of non-aspirin related peptic ulceration or aspirin-related GI complications (dyspepsia or peptic ulcer). All 70 participants were given clopidogrel. None of the patients on were on maintenance PPI treatment. Nine (12%) patients had GI bleeding and 1 patient had a perforated ulcer. This study demonstrated a significant increase in peptic ulcer bleeding in patients with a history of GI bleeding versus patients without a history of peptic ulcers whilst on clopidogrel (p=0.007; OR 1.3, CI 1.1-1.5). The CAPRIE-Trial involved 19185 patients with previous ischaemic stroke, MI or atherosclerotic peripheral disease. One arm of the study received aspirin and the other clopidogrel. The end-point was occurrence of ischaemic stroke, MI or vascular death. The outcome was a significant risk reduction for clopidogrel versus aspirin of 8.7% (p=0.043; CI 0.3-16.5). The annual risk for complications was 5.32% (clopidogrel) and 5.83% (aspirin) respectively. GI bleeding was found in 191 patients (1.99%) on clopidogrel versus 255 (2.66%) on aspirin (RR 1.34, CI 1.11-1.61, p<0.05). Therefore, they demonstrated a significant reduction of GI bleeding with clopidogrel in patients with previous cardiovascular events.

Editor Comment

(CI: 95% Confidence Interval, OR: Odds Ratio, NNT: Number Needed To Treat, RR: Relative Risk)

Clinical Bottom Line

Clopidogrel causes fewer gastrointestinal complications than aspirin in those patients with no previous history of gastric or duodenal ulceration with a number needed to treat of around 200 to prevent an episode of bleeding per year. However in those patients with a previous history of gastrointestinal complications, clopidogrel alone is not a safer alternative than aspirin alone. Either aspirin or clopidogrel combined with a proton pump inhibitor are equally effective for these patients.

References

1. Chan FKL, Ching JYL, Hung LCT et al. Clopidogrel versus aspirin and esomeprazole to prevent recurrent ulcer bleeding. N Engl J Med 2005;352:238-44.
2. Ibanez L, Vidal X, Vendrell L et al. On behalf of the Spanish-Italian collaborative group for the epidemiology of gastrointestinal bleeding. Upper gastrointestinal bleeding associated with antiplatelet drugs. Aliment Pharmacol Ther 2006;23:235–242.
3. Ng FH, Wong BCY, Wong SY et al. Clopidogrel plus omeprazole compared with aspirin plus omeprazole for aspirin-induced symptomatic peptic ulcers/erosions with low to moderate bleeding/re-bleeding risk - a single-blind, RCT. Aliment Pharmacol Ther 2004;9:359–365.
4. Lai KC, Chu KM, Hui WM et al. Esomeprazole with aspirin versus clopidogrel for prevention of recurrent gastrointestinal ulcer complications. Clinical Gastroenterology and Hepatology 2006;4(7):860-865.
5. Ng FH, Wong SY, Chang CM. High incidence of clopidogrel-associated gastrointestinal bleeding in patients with previous peptic ulcer disease. Aliment Pharmacol Ther 2003:18:443–449.
6. CAPRIE Study Organisation. Steering Committee: Gent M, Beaumont D, Blanchard J et al. A randomised, blinded, trial of clopidogrel versus aspirin in patients at risk of ischaemic events (CAPRIE). Lancet 1996;348:1329–39.