A Comprehensive Review of NSAID Cardiovascular Toxicity

This led to scrutiny of other COX-2 inhibitory agents. Several other first- and second-generation COX-2 inhibitors, such as etorocoxib, lumiracoxib, parecoxib and valdecoxib, were associated with a similar cardiovascular risk.^14-16

These studies demonstrated the possible role of COX-2 inhibition as a mechanism for cardiovascular risk. Subsequently, nonselective inhibitors of COX were also shown to be associated with cardiovascular toxicity risk.

A position statement by American Heart Association summarized the risk of myocardial infarction associated with NSAIDs.¹⁷ The position statement highlighted the cardiovascular toxicity of COX-2 inhibitors; even among nonselective COX inhibitors, there was an increased risk of MI with drugs that preferentially inhibited COX-2, such as diclofenac, which features a 60% increased risk. Notably, naproxen was not associated with an increased risk of MI in clinical trials or observational studies.

This seminal manuscript had two direct impacts. First, it led to rapid decline in the use of diclofenac, in addition to COX-2 inhibitors. Second, several other studies started examining the risk of MI with other nonselective inhibitors.

Among the COX-2 inhibitors, celecoxib has not been associated with the risk of MI. In 2016, the PRECISION trial (a non-inferiority trial) assessed cardiovascular outcomes among naproxen, ibuprofen and celecoxib users, and it did not show an increased risk of MI in celecoxib compared to the other groups.¹⁸ It also showed reduced GI toxicity with celecoxib. However, the majority of the patients did not receive more than 200 mg/day of celecoxib.

Such results were also noted in the Standard Care versus Celecoxib Outcome Trial.¹⁹ Hence, at lower doses, celecoxib may be a safe anti-inflammatory drug.²⁰

Nonselective Inhibitors & MI Risk Over the past decade, several investigators have evaluated the risk of MI with other nonselective agents. Such studies have been conducted using large population databases, including the UK-CPRD database, the UK-THIN database, the Finnish national database, the Dutch-PHARMO database and the Canadian, Danish and Taiwanese national databases.^7,21-24 In a meta-analysis and systematic review of these studies in 2013, the NSAIDs etodolac, indomethacin and diclofenac were associated with a risk of MI. The study also demonstrated a dose response relationship for diclofenac and COX-2 inhibitors, but these drugs remained unsafe even at low doses. Naproxen was again noted to be safe both at low and high doses.^25-27

The meta-analysis also assessed the risk associated with (then) less frequently used drugs, such as meloxicam, and noted increased risk. Meloxicam is an increasingly popular analgesic today, and to evaluate whether its use conferred an increased risk of MI, we conducted a study using the UK-THIN database. We demonstrated a significantly increased risk of MI associated with meloxicam use, similar to that of diclofenac.