Naproxen CV Safety vs Ibuprofen Challenged in WHI Analysis

July 08, 2014

DALLAS, TX — Contrary to a lot of observational data and common belief even among clinicians, regular intake of naproxen entails more cardiovascular risk than ibuprofen, according to an analysis of nonsteroidal anti-inflammatory drug (NSAID) use among participants in the Women's Health Initiative (WHI)[1]. Regular intake of ibuprofen, in fact, did not seem to involve any elevation in CV risk, defined as CV death, nonfatal MI, or nonfatal stroke.

The analysis, published today in Circulation: Cardiovascular Quality and Outcomes, also saw a complicated relationship between aspirin and most other NSAIDs. It suggested that concomitant aspirin may alleviate the increased CV risk long associated with selective inhibitors of cyclooxygenase-2 (COX-2) such as celecoxib (Celebrex, Pfizer), but not the risk from non–COX-selective NSAIDs with proportionally more selectivity for COX-2 than COX-1, such as naproxen.

It's underappreciated that non–COX-selective NSAIDs vary in their relative inhibition of COX-1 vs COX-2 and therefore that non–COX-selective NSAIDs can tend toward greater selectivity for COX-2 than COX-1 and vice versa[2], the study's lead author Dr Anthony A Bavry (University of Florida, Gainesville) told heartwire . "Many investigators and a lot of writing" hold naproxen to be safer than ibuprofen in terms of CV risk, perhaps because they perceive naproxen as primarily COX-1 selective, "but in fact that's not true."

The non–COX-selective diclofenac also leans toward COX-2 selectivity, but its associated CV-risk increase didn't reach significance in the analysis from Bavry and colleagues.

In February of this year, a panel consisting of two FDA advisory committees voted against the agency's contention that there may be data enough to justify changing naproxen's labeling to reflect a lower cardiovascular risk profile compared with other nonaspirin NSAIDs, as reported by heartwire . The FDA cited a recent high-profile meta-analysis[3] that suggested the CV-risk elevation with high-dose diclofenac and possibly ibuprofen rivaled that of the COX-2-selective inhibitors (also called coxibs) and that high-dose naproxen entailed less CV risk than other nonaspirin NSAIDs.

The panel's rejection of support for the proposed label change, Bavry said, "ended up being a wise decision."

As a group, nonselective NSAIDs that lean toward COX-1 selectivity, including ibuprofen, ketoprofen, and indomethacin, did not elevate CV risk, according to the analysis. "As such, these agents may be safer from a cardiovascular perspective for long-term use among postmenopausal women," the group writes.

Indeed, said Bavry, the analysis and its conclusions apply only to postmenopausal women. "These findings might apply to men, but that's entirely speculative."

"The FDA Was Right"

Dr Garret A FitzGerald (University of Pennsylvania, Philadelphia), in an email to heartwire , agreed that the FDA panel made the right decision in not endorsing the label change, but not for Bavry's reasons. FitzGerald was a member of the Coxib and Traditional NSAID Trialists' (CNT) Collaboration that authored the May 2013 meta-analysis suggesting less CV risk with naproxen.

The CNT analysis, he said, signaled diclofenac as having "a CV risk similar to the coxibs, which fits its selectivity. Naproxen seemed less harmful than the coxibs in some respects, which again fits its pharmacology, assuming it behaves like aspirin in some individuals, due to its kinetics. However, both dose and individual kinetics can change that picture, so I think the FDA was right not to change things at a population level."

Although the CNT analysis had limitations, he said, "the source data, individual patient-level data from randomized controlled trials, was more robust than the source data here." The current analysis, he said, "is just another retrospective trawl through observational data collected for other reasons. There is barely a positive signal for celecoxib and rofecoxib, so the ability to discriminate among other NSAIDs is highly questionable."

The CNT analysis, he said, included fewer data for ibuprofen than for either diclofenac or naproxen and so couldn't say much conclusively about ibuprofen's relative CV safety compared with those other drugs.

Speaking with heartwire , Dr Elliot Antman (Brigham and Women's Hospital, Boston, MA) said the current study "extends observations about the risk of NSAIDs to postmenopausal women. That is helpful and confirmatory. The study also supports the concern that the more COX-2-selective a set of drugs is, the more there is risk." Antman was the lead author of a 2007 American Heart Association guidance on NSAID use.

On the other hand, he said, the study isn't sufficiently powered to compare specific drugs with each other for CV risk. It compared NSAIDs collectively in the three broad categories and extrapolated "a lot" about naproxen, for example.

"It is an overstatement, in my opinion, to actually draw out any inferences about individual drugs," Antman said.

He also questioned the nature of the three NSAID categories. With respect to COX-2 selectivity, he said, "ibuprofen is virtually the same as naproxen. So there's no good biologic basis to believe that there would be a substantial difference between ibuprofen and naproxen" with respect to COX-2–related CV risk.

Also, the risk of "bias and confounding is certainly possible here and highly likely." On the basis of what is currently known about the NSAIDs, Antman said, the message from the 2007 document remains true today: "use the safest drug, at the lowest dose that works, for the shortest period of time, and do not take them chronically."

Bavry pointed out, in fact, that in his cardiovascular patients who have chronic pain syndromes "and are regular users of NSAIDs, I always talk to them about the association of NSAIDs and potential risk of adverse cardiovascular events. To try to minimize their use of an NSAID, I advise my patients to consider the use of acetaminophen, use a smaller dose, or use NSAIDs less frequently."

NSAID Use and Risks in the WHI

The current study of 160 801 postmenopausal women categorized them as regular users vs nonusers of nonaspirin NSAIDs and analyzed NSAID use as a time-dependent covariate, which measured NSAID exposure but allowed for cases in which they were started and stopped intermittently. Over an 11-year follow-up, >53 000 women reported regular NSAID use at some time.

In adjusted analysis, regular NSAID use was associated with a 10% rise in risk for CV death, nonfatal MI, or nonfatal stroke (p<0.001) vs no regular NSAID use. But the hazard ratio (HR) for CV risk varied by each of the three NSAID categories used by Bavry et al: group 1 (COX-2 selective, that is, the coxibs); group 2 (non–COX-selective agents with greater COX-2 than COX-1 selectivity, including naproxen and diclofenac); and group 3 (non–COX-selective agents with greater COX-1 than COX-2 selectivity, including ibuprofen and ketoprofen).

CV risk went up a "modest" 13% (p=0.004) with group 1 agents and 17% (p<0.001) with group 2 agents; there was no significant elevation for group 3 agents.

Hazard Ratio* (HR, 95% CI) for CV Death, Nonfatal MI, or Nonfatal Stroke vs No NSAID Use, by NSAID Category, in WHI

NSAID category HR (95% CI) p
All nonaspirin NSAIDs 1.10 (1.06–1.15) <0.001
Overall group 1 1.13 (1.04–1.23) 0.004
Celecoxib only 1.13 (1.01–1.27) 0.031
Overall group 2 1.17 (1.10–1.24) <0.001
Naproxen only 1.22 (1.12–1.34) <0.001
Overall group 3 1.01 (0.95–1.07) 0.884
Ibuprofen only 1.00 (0.93–1.07) 0.996
*Adjusted for age, ethnicity, education, income, US region of residence, body-mass index, systolic BP, physical activity, history of rheumatoid arthritis, hypertension, high cholesterol, diabetes, MI, stroke or transient ischemic attack, heart failure, CABG or PCI, peripheral arterial disease, smoking status, menopausal hormone use, statin use, acetaminophen use, and aspirin use
Group 1: COX-2 selective; Group 2: non–COX-selective but COX2>COX1 inhibition; Group 3: non–COX-selective NSAIDs but COX-1>COX2 inhibition

A subgroup analysis by aspirin use, write the authors, "largely mirrored the main study findings, except for concomitant users of selective COX-2 inhibitors. Among aspirin users, concomitant selective COX-2 inhibitors were no longer associated with increased hazard for cardiovascular events, although they were still associated with increased hazard for cardiovascular events among nonaspirin users."

Also among aspirin users, Bavry said to heartwire , ibuprofen still remained neutral with respect to CV risk: there was still no evidence of harm with nonselective agents that tended toward more COX-1 selectivity, such as ibuprofen. But there was still evidence of harm among aspirin users using nonselective agents that leaned more toward COX-2 selectivity, such as naproxen.

All that "makes sense, because COX-2 [receptors are] not on platelets, so COX-2 inhibitors and aspirin are not competing for the same [receptor] sites," he said. "That's different from aspirin [effects] with nonselective NSAIDs. They're both competing to bind to the platelet [receptor] sites."

At one level, therefore, concomitant aspirin seems to provide some protection against the adverse CV effects of COX-2–selective NSAIDs, but that doesn't mean such coadministration should be recommended, Bavry said.

"I wouldn't go that far, because COX-2 inhibitors were designed to minimize gastrointestinal toxicity, and regular use of aspirin is well known to be associated with GI toxicity and bleeding." Their coadministration would presumably restore some of the bleeding risk the coxibs were intended to minimize, he said.

But in contrast, he added, "if you concomitantly use naproxen and aspirin, there is still the same degree of [cardiovascular] risk, if not more, than in individuals who are just naproxen users [without] aspirin on top of that."

Bavry is a contractor for the American College of Cardiology's Cardiosource. FitzGerald discloses consulting for Lilly. Antman is president-elect of the American Heart Association and had no other disclosures.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.