Which nonsteroidal anti-inflammatory drugs (NSAIDs) have the lowest cardiovascular (CV) risk or lowest gastrointestinal (GI) risk?

Comment by InpharmD Researcher

Pooled data have determined celecoxib and naproxen to have the lowest risk of serious cardiovascular (CV) events, while ibuprofen, rofecoxib, and diclofenac have been associated with the highest risk. Based on a large case-control study, the risk of upper gastrointestinal (GI) bleeding is lowest with diclofenac. COX-1 selective agents (e.g., piroxicam, ketorolac) tend to have a higher risk for bleeding than nonselective or COX-2 selective NSAIDs. In general, the bleeding risk is higher when NSAIDs are combined with anticoagulants, but the bleeding risk for individual NSAIDs with concomitant anticoagulant use needs further evaluation. The overall risk for CV or GI events may also depend on the dose and duration of individual agents.

Background

The primary concern for bleeding risk for non-steroidal anti-inflammatory (NSAID) agents is mainly via the inhibition of cyclooxygenase-1 (COX-1) within the gastrointestinal (GI) tract, although COX-2 may also be involved in gastric healing. All NSAIDs have been demonstrated to increase the risk of upper GI complications to varying extents. For example, celecoxib and ibuprofen are considered lower risk for upper GI complications, while ketorolac and piroxicam are considered the highest. The higher-risk agents typically have greater selectivity for COX-1. Higher doses are also associated with increased inhibition of COX-1 and COX-2, which have been linked to an increased risk of GI-related complications compared to lower doses. [1], [2], [3]

Concomitant use of NSAIDs with other medications and how their effects on GI bleeding rates have also been investigated. The combination of antithrombotics and NSAIDs and subsequently increased risk of upper GI bleeding is well established. Randomized trials observing the addition of antithrombotic therapy to low-dose aspirin observed an increased risk of GI bleeding by approximately 2-folds. Suggested management involves proper education, prescription and over-the-counter (OTC) NSAIDs assessment, and minimizing exposure to NSAIDs or recommending a COX-2-specific agent. Gastroprotective agents may also be added to the regimen. [3], [4]

In a 2007 meta-analysis, researchers pooled data to provide an educational update on the current evidence of NSAIDs' efficacy and adverse effects. They determined which NSAIDs are more effective and associated with fewer adverse effects. Of the adverse effects, GI incidences were among the most serious complications of NSAID use. Moreover, recent studies indicated that some NSAIDs are associated with a higher GI risk than others. In terms of risk variation, data showed that ibuprofen has the lowest risk among NSAIDs, while diclofenac and naproxen have intermediate risks, and piroxicam and ketorolac carry the greatest risk (Table 3). However, the advantages of “low-risk” options may be lost at higher doses. Additionally, it was noted that COX-2 inhibitors could be used as an alternative analgesic to decrease the risk of gastrointestinal events. [5]

A 2011 network meta-analysis compared available evidence on the CV safety of NSAIDs. A total of 31 trials (N= 11,6429 patients with more than 115,000 patient-years of follow-up) were included for analysis. Patients were allocated to naproxen, ibuprofen, diclofenac, celecoxib, etoricoxib, rofecoxib, lumiracoxib, or placebo. Compared to placebo, rofecoxib had the highest risk of myocardial infarction (rate ratio [RR] 2.12; 95% credibility interval [CrI] 3.56), whereas etoricoxib had the lowest (RR 0.75; 95 CrI 0.23 to 2.39). Ibuprofen had the highest risk of stroke (RR 3.36; 95% CrI 1.00 to 11.60), whereas rofecoxib had the lowest (RR 1.07; 95% CrI 0.60 to 1.82). Etoricoxib had the highest risk of CV death (RR 4.07; 95% CrI 1.23 to 15.70), and naproxen had the lowest (RR 0.98; 95% CrI 0.41 to 2.37). For the composite of non-fatal myocardial infarction, non-fatal stroke, or CV death, ibuprofen had the highest risk (RR 2.26; 95% CrI 1.11 to 4.89), whereas celecoxib had the lowest risk (RR 1.43; 95% CrI 0.94 to 2.16). Overall, the authors determined that, in general, naproxen seems to be the safest analgesic for patients with osteoarthritis concerning CV safety, but this advantage has to be weighed against GI toxicity and the need for a concomitant prescription of a proton pump inhibitor in many patients. [6]

A 2006 meta-analysis of controlled observational studies compared the risk of serious CV events with individual NSAIDs and COX-2 inhibitors. A total of 17 case-control and 6 cohort studies were included for analysis. Of COX-2 inhibitors, celecoxib was determined to have the lowest risk of serious CV events (risk ratio [RR] 1.06; 95% confidence interval [CI] 0.91 to 1.23). Rofecoxib with a dose > 25 mg/day had the highest risk of serious CV events (RR 2.19; 95% CI 1.64 to 2.91). Of the nonselective NSAIDs, naproxen had the lowest risk of serious CV events (RR 0.97; 95% CI 0.87 to 1.07), and diclofenac had the highest risk (RR 1.40; 95% CI 1.16 to 1.70). [7]

References:

[1] Brune K, Patrignani P. New insights into the use of currently available non-steroidal anti-inflammatory drugs. J Pain Res. 2015;8:105-118. Published 2015 Feb 20. doi:10.2147/JPR.S75160
[2] Hunter TS, Robison C, Gerbino PP. Emerging evidence in NSAID pharmacology: important considerations for product selection. Am J Manag Care. 2015;21(7 Suppl):S139-S147.
[3] Laine L. NSAID-associated gastrointestinal bleeding: assessing the role of concomitant medications. Gastroenterology. 2014;147(4):730-733. doi:10.1053/j.gastro.2014.08.021
[4] Vazquez SR. Drug-drug interactions in an era of multiple anticoagulants: a focus on clinically relevant drug interactions. Hematology Am Soc Hematol Educ Program. 2018 Nov 30;2018(1):339-347. doi: 10.1182/asheducation-2018.1.339
[5] Ong CK, Lirk P, Tan CH, Seymour RA. An evidence-based update on nonsteroidal anti-inflammatory drugs. Clin Med Res. 2007;5(1):19-34. doi:10.3121/cmr.2007.698
[6] Trelle S, Reichenbach S, Wandel S, et al. Cardiovascular safety of non-steroidal anti-inflammatory drugs: network meta-analysis. BMJ. 2011;342:c7086. Published 2011 Jan 11. doi:10.1136/bmj.c7086
[7] McGettigan P, Henry D. Cardiovascular risk and inhibition of cyclooxygenase: a systematic review of the observational studies of selective and nonselective inhibitors of cyclooxygenase 2. JAMA. 2006;296(13):1633-1644. doi:10.1001/jama.296.13.jrv60011

Literature Review

A search of the published medical literature revealed 3 studies investigating the researchable question:

Which nonsteroidal anti-inflammatory drugs (NSAIDs) have the lowest cardiovascular (CV) risk or lowest gastrointestinal (GI) risk?

Please see Tables 1-3 for your response.


 

Risk of upper gastrointestinal ulcer bleeding associated with selective cyclo-oxygenase-2 inhibitors, traditional non-aspirin non-steroidal anti-inflammatory drugs, aspirin, and combinations

Design

Hospital-based, case-control study with prospective case ascertainment and retrospective data collection

N= 8,309

Objective

To determine the risk of peptic ulcer upper gastrointestinal bleeding (UGIB) associated with the use of coxibs, traditional non-steroidal anti-inflammatory drugs (NSAIDs), aspirin, or combinations of these drugs in clinical practice

Study Groups

Cases (n= 2,777)

Controls (n= 5,532)

Inclusion Criteria

Aged 20–85 years; free of liver disease, coagulation disorders or malignancies for the previous 5 years

Exclusion Criteria

Any other cause of bleeding (gastroesophageal varices, vascular lesions, tumors, Mallory–Weiss, associated coagulopathy, and esophagitis); patients with unreliable sources of information; in-hospital bleeding patients

Methods

Cases and controls were collected through a network of general hospitals in Spain. Patients and controls were interviewed by the same person (a gastroenterologist or gastroenterology trainee) in each participating center, in general, within 48 h of admission. To avoid bias, interviewers were not involved in the design, were not the investigators at each center of the study, and were unaware of the actual objectives of the study.

Duration

Between 2001 and 2004

Outcome Measures

Risk of UGIB according to individual NSAIDs, NSAID use with anticoagulants, half-life and daily dose, and timing, NSAID use with low-dose aspirin, coxib use with low-dose aspirin

Baseline Characteristics

 

Cases (n= 2,777)

Controls (n= 5,532)

 

Mean age, years

61  61  

Female

28% 48%  

History of ulcer complication

19% 4.9%  

Smoking status

Current smoker

Previous smoker

 

29.1%

31.5%

 

21.6%

26%

 

Aspirin use

Current

Past 

 

746 (26.9%)

90 (3.2%)

 

524 (9.5%)

334 (6.0%)

 

NSAID use

Current

Past 

 

657 (23.7%)

103 (3.7%)

 

511 (9.2%)

328 (5.9%)

 

Coxib use

Current

Past 

 

34 (1.2%)

9 (0.3%)

 

67 (1.2%)

25 (0.5%)

 

Anticoagulant  use

Current

Past 

 

179 (6.4%)

2 (0.1%)

 

205 (3.7%)

10 (0.9%)

 

Other antiplatelet agents (clopidogrel or ticlopidine)

3.9% 1.5%  

Results

Endpoint

Cases (n= 2,777)

Controls (n= 5,532)

Adjusted condition relative risk of UGIB (95% confidence interval [CI])**

Individual NSAIDs*

Non-use

Ibuprofen

Diclofenac 

Aceclofenac

Naproxen

Piroxicam

Indomethacin

Meloxicam

Ketorolac

Lornoxicam

Ketoprofen

Other NSAIDs

 

2,017

174

126

31

80

98

20

20

24

9

14

17 

 

4,693

162

140

52

46

32

14

13

7

6

5

 

Reference

4.1 (3.1 to 5.3)

3.1 (2.3 to 4.2)

2.6 (1.5 to 4.6)

7.3 (4.7 to 11.4)

12.6 (7.8 to 20.3)

9.0 (3.9 to 20.7)

9.8 (4.0 to 23.8)

14.4 (5.2 to 39.9)

7.7 (2.4 to 24.4)

8.6 (2.5 to 29.2)

13.8 (4.2 to 44.8)

Anticoagulants and NSAIDs

Non-use

NSAID use only

Dicumarinics only

Combination


1,784

622

144

35


4,500

504

196

9


Reference

5.0 (4.2 to 5.9)

2.8 (2.1 to 3.7)

19.3 (8.2 to 45.3)

NSAID plasma half-life

All NSAID dose

<12 H

≥12 H

Low-medium daily dose

<12 H

≥12 H

High daily dose

<12 H

≥12 H

 

 

406

207

 

250

58 

 

156

149

 

 

388

95

 

275

41

 

113

54 

 

 

4.0 (3.3 to 4.9)

9.4 (6.9 to 12.7)

 

3.5 (2.8 to 4.5)

6.4 (3.9 to 10.6)

 

5.0 (3.6 to 6.9)

12.4 (8.1 to 18.8)

Coxib use

Non-use

Current (0-7 days)

Celecoxib

Rofecoxib

Past (≥8 days)

 

2,734

34

13

23

 

5,440

67

31

36

25 

 

Reference

1.5 (0.9 to 2.4)

1.0 (0.4 to 2.1)

2.1 (1.1 to 4.0)

1.0 (0.4 to 2.3)

Aspirin use

Non-use

Current (0-7 days)

Past (≥8 days)


1,941

746

90


4,674

524

334


Reference

5.3 (4.5 to 6.3)

0.7 (0.6 to 1.0)

NSAID and low-dose aspirin

55

27

12.7 (7.0 to 23.0)

Coxib and low-dose aspirin

6

4

14.5 (3.3 to 63.9)

*Estimates of relative risk were calculated for individual NSAID drugs with five or more exposed
NSAIDs include flurbiprofen, meclofenamate, morniflumateniflumic acid, nimesulide, and tenoxicam.
**Adjusted for age, sex, calendar semester, ulcer history, nitrates, anticoagulants, antiplatelets, acid-suppressing drugs, coxib, and aspirin use.

Adverse Events

See results

Study Author Conclusions

Coxib use presents a lower RR of UGIB than non-selective NSAIDs. However, when combined with low-dose aspirin, the differences between non-selective NSAIDs and coxibs tend to disappear. Treatment with either non-aspirin antiplatelet or cardioprotective aspirin has a similar risk of UGIB.

InpharmD Researcher Critique

Despite considering adjustments for the potential confounders, this study was limited based on subjective data from interviewees and may be subject to recall bias. Additionally, Helicobacter pylori infection assessment was not included in the protocol.



References:

Lanas A, García-Rodríguez LA, Arroyo MT, et al. Risk of upper gastrointestinal ulcer bleeding associated with selective cyclo-oxygenase-2 inhibitors, traditional non-aspirin non-steroidal anti-inflammatory drugs, aspirin and combinations. Gut. 2006;55(12):1731-1738. doi:10.1136/gut.2005.080754

 

Patients With Atrial Fibrillation Taking Nonsteroidal Anti-Inflammatory Drugs and Oral Anticoagulants in the ARISTOTLE Trial

Design

Post hoc analysis of the randomized controlled ARISTOTLE trial

N= 17,423

Objective

To examine the use of nonsteroidal anti-inflammatory drug (NSAIDs) in patients with atrial fibrillation (AF) anticoagulated with apixaban or warfarin and to explore bleeding, thrombotic, heart failure hospitalizations, and all-cause mortality outcomes by type of oral anticoagulant (OAC) agent

Study Groups

Apixaban (n= 9,120)

Warfarin (n= 9,081)

Inclusion Criteria

Patients with AF; anticoagulated with apixaban or warfarin in ARISTOTLE trial without severe renal (creatine clearance ≤ 30 mL/min) or liver disease

Exclusion Criteria

Prosthetic heart valves, use of aspirin >165 mg daily or clopidogrel, and stroke within 7 days of randomization, chronic liver disease or severe kidney disease

Methods

Eligible patients with AF and at least 1 additional risk factor for stroke were randomly assigned to receive apixaban 5 mg twice daily or dose-adjusted warfarin (international normalized ratio [INR], 2.0–3.0). A reduced dose of apixaban 2.5 mg or apixaban placebo twice daily was administered to 4.7% of patients with 2 or more of the following criteria: age ≥80 years, weight ≤60 kg, or serum creatinine concentration ≥1.5 mg/dL.

Duration

Median follow-up: 1.8 years

Outcome Measures

Primary: major bleeding

Secondary: clinically relevant nonmajor (CRNM) bleeding, gastrointestinal (GI) bleeding

Baseline Characteristics

 

Baseline NSAID (n= 832)

Incident NSAID (n= 2,185)

Never NSAID (n= 14,406)  

Median age (interquartile range [IQR]), years

70 (64, 77)

70 (63, 75)

70 (62, 76)

 

Female

37.0% 34.7% 34.8%  

Medical history

Peripheral arterial disease

Prior myocardial infarction

Prior clinically relevant or spontaneous bleeding

 

4.3%

12.4%

24.5%

 

4.7%

11.8%

21.0%

 

4.6%

14.4%

15.6% 

 

Prior use of vitamin K antagonists (> 30 days)

37.1%  40.2% 43.0%  

HAS-BLED score distribution

Mean

≤ 1

2

≥ 3

 

2.5 ± 1.0

121 ± 14.5

324 ± 38.9

387 ± 46.5

 

1.8 ±1.0

841 ± 38.5

823 ± 37.7

521 ± 23.8

 

1.7 ± 1.0

6,347 ± 44.1

5,129 ± 35.6

2,930 ± 20.3

 

Randomized to apixaban

429 (51.6%)

1,048 (48.0%)

7,242 (50.3%)

 
 

Baseline NSAID (n= 832)

Incident NSAID (n= 2,185)

   

NSAID

Ibuprofen

Diclofenac

Cox II inhibitor

Naproxen

Meloxicam

Celecoxib

 

170 (14.3%)

179 (15.0%)

172 (14.4%)

124 (10.4%)

88 (7.4%)

82 (6.9%)

 

521 (18.4%)

697 (24.6%)

6 (0.2%)

149 (5.3%)

171 (6.0%)

183 (6.5%)

   

Days on NSAID during follow-up

687.2 ± 351.2

141.6 ± 251.9 

   

No. of unique prescribed NSAIDs during follow-up

1

2

3

4

 

529 (63.6%)

161 (19.4%)

77 (9.3%)

65 (7.8%)

 

1,466 (67.1%)

450 (20.6%)

139 (6.4%)

130 (5.9%)

   

Oral antiplatelet at randomization

Aspirin

Clopidogrel

273 (32.8%)

260 (31.3%)

12 (1.4%)

783 (35.8%)

730 (33.4%)

43 (2.0%)

   

HAS-BLED score: Hypertension, Abnormal Renal/Liver Function, Stroke, Bleeding History or Predisposition, Labile INR, Elderly, Drugs/ Alcohol Concomitantly

Results

Endpoint

Apixaban (n= 9,120)

Warfarin (n= 9,081)

Hazard ratio (95% Confidence interval)

p-value

Major bleeding

NSAID user

Non-NSAID user

 

20 (2.74%)

296 (2.12%)

 

29 (4.24%)

410 (2.99%)

 

0.64 (0.36 to 1.14)

0.71 (0.61 to 0.82) 

0.75

Major or CRNM bleed

NSAID user

Non-NSAID user

 

36 (5.01%)

556 (4.06%)

 

55 (8.35%)

778 (5.82%)

 

0.60 (0.39 to 0.91)

0.70 (0.63 to 0.78)

0.49

GI bleeding

NSAID user

Non-NSAID user

 

6 (0.81%)

94 (0.67%)

 

8 (1.15%)

102 (0.74)

 

0.70 (0.24 to 2.02)

0.91 (0.69 to 1.20)

0.64

Adverse Events

See results

Study Author Conclusions

New use of NSAID during the trial was associated with major and CRNM bleeding, but not with gastrointestinal bleeding. The use of NSAIDs did not appear to be associated with a significant change in the safety or efficacy of apixaban relative to warfarin. Although the interaction between treatment arm and NSAID use over time was not statistically significant, patients on warfarin had an increased risk of major bleeding and gastrointestinal bleeding with NSAID use. In patients on apixaban, no such increase in risk was observed.

InpharmD Researcher Critique

Given the post hoc analysis nature, this study was subject to selection bias and confounding. Moreover, data on the indication or dosage of NSAIDs was lacking and discontinuation of NSAIDs was self-reported.



References:

Dalgaard F, Mulder H, Wojdyla DM, et al. Patients with atrial fibrillation taking nonsteroidal anti-inflammatory drugs and oral anticoagulants in the aristotle trial. Circulation. 2020;141(1):10-20. doi:10.1161/CIRCULATIONAHA.119.041296

 

Relative risk of gastrointestinal complications with NSAIDs, relative to ibuprofen or non-use (95% confidence interval)

Drug

Case-control studies

Cohort study

Case-control

Nonuse

---

---

1.0

Ibuprofen

1.0

1.0

2.1 (0.6 to 7.1)

Fenoprofen

1.6 (1.0 to 2.5)

3.1 (0.7 to 13)

---

Aspirin

1.6 (1.3 to 2.0)

---

---

Diclofinac

1.8 (1.4 to 2.3)

1.4 (0.7 to 2.6)

2.7 (1.5 to 4.8)

Sulindac

2.1 (1.6 to 2.7)

---

---

Diflusinal

2.2 (1.2 to 4.1)

---

---

Naproxen

2.2 (1.7 to 2.9)

1.4 (0.9 to 2.5)

4.3 (1.6 to 11.2)

Indomethacin

2.4 (1.9 to 3.1)

1.3 (0.7 to 2.3)

5.4 (1.6 to 18.9)

Tolmetin

3.0 (1.8 to 4.9)

---

---

Piroxicam

3.8 (2.7 to 5.2)

2.8 (1.8 to 4.4)

9.5 (6.5 to 13.8)

Ketoprofen

4.2 (2.7 to 6.4)

1.3 (0.7 to 2.6)

3.2 (0.9 to 11.9)

Azopropazone

9.2 (2.0 to 21)

4.1 (2.5 to 6.7)

---

Ketorolac

---

---

24.7 (9.6 to 63.5)

 

References:

Adapted from: Ong CK, Lirk P, Tan CH, Seymour RA. An evidence-based update on nonsteroidal anti-inflammatory drugs. Clin Med Res. 2007;5(1):19-34. doi:10.3121/cmr.2007.698