Another agent increasingly used due to its proven GI safety is nabumetone. Aside from a single Finnish observational study that demonstrates its safety with the heart, no other studies assess its cardiovascular risk.22
Heart Failure
Several clinical trials and observational studies have noted an increased risk of hospitalization for heart failure due to any NSAID use. A meta-analysis of more than 600 clinical trials conducted by Coxib and the traditional NSAID Trialists (CNT) Collaboration showed more than twofold increased risk of hospitalization due to heart failure across all NSAIDs assessed.27 In 2016, Arfè et al. assessed the hospitalization risk due to heart failure associated with individual NSAIDs using five population-based healthcare databases. Notably, the risk estimates for heart failure followed the same trend as that of MI. Coxibs, such as rofecoxib and etoricoxib, diclofenac, indomethacin and piroxicam doubled this risk. Naproxen was associated with 16% increased risk of hospitalization due to heart failure, and celecoxib was not associated with this risk.28
NSAIDs may impair renal function in patients with decreased effective circulating volume by inhibiting prostaglandin synthesis. This leads to reduced glomerular filtration and sodium and water retention, which ultimately leads to decompensated heart failure in these patients. Even the selective COX-2 inhibitors have similar effects on the renal function.29
Stroke
NSAIDs can be associated with risk of ischemic stroke, but this risk has been less well described than heart failure and myocardial infarction. Although the APPROVe trial had very few stroke events, it was posited the risk would be similar to that of myocardial infarction. In fact, a meta-analysis of randomized trials did not show any increased risk of stroke with NSAID use.27 However, several observational trials have consistently shown increased risk with highly selective COX-2 inhibitors (except celecoxib, for which there are very divergent results) ranging from 30% to more than twofold increased risk. Similarly, not enough data are available regarding stroke risk and other NSAIDs.30
Conventionally, NSAIDs have been thought to be associated with hemorrhagic stroke due to inhibitory effect on the platelets, similar to the GI bleeding risk associated with NSAIDs. However, some studies suggest no association between NSAID use and hemorrhagic stroke.
Venous Thromboembolism
Although data for the association between NSAID use and thromboembolism are not as robust, there is growing evidence to suggest NSAID use can be associated with the risk of deep vein thrombosis, as well as pulmonary embolism. In fact, in meta-analysis conducted by Ungprasert et al. in 2015, use of any NSAID was associated with 80% increased odds of venous thromboembolism.31 We assessed the risk of thromboembolism with individual NSAIDs in a population-based cohort of osteoarthritis patients (to minimize confounding by indication). The risk was most noted with use of COX-2 inhibitors (including celecoxib) and nonselective COX inhibitor preferentially inhibiting COX-2 (i.e., diclofenac, meloxicam). Naproxen did not increase the risk of thromboembolism.32
Common Study Pitfalls
Aside from the initial randomized clinical trials assessing COX-2 inhibitors, diclofenac, ibuprofen and naproxen, most of the evidence for other medications come from observational studies, many of which were conducted in large population databases or claims databases. These studies afford large sample sizes representative of general population, which allows evaluation of less frequently used NSAIDs, such as meloxicam, indomethacin, etolodolac, nabumetone, etc.
