Recategorization of Non-Aspirin Nonsteroidal Anti-inflammatory Drugs According to Clinical Relevance: Abandoning the Traditional NSAID Terminology

Non-aspirin nonsteroidal anti-inflammatory drugs (NSAIDs) are frequently used to treat pain, fever, and inflammation. Historically, NSAIDs have been categorized as traditional NSAIDs and newer cyclooxygenase (COX)-2 inhibitors (coxibs). However, traditional NSAIDs also inhibit the COX-1 and COX-2 enzyme isoforms to a varying degree. This diversity of COX-1 and COX-2 selectivity within the class of traditional NSAIDs has proven clinically important, with evidence accumulating on the cardiovascular risks associated with selective COX-2 inhibition. Thus, the relative COX-2 selectivity of traditional NSAIDs correlates with their cardiovascular risk profile, being more favourable for non-selective NSAIDs, such as naproxen and low-dose ibuprofen, and less favourable for more COX-2 selective agents, such as diclofenac. To enhance clinically relevant terminology, we advocate categorizing all non-aspirin NSAIDs—including traditional NSAIDs—according to their relative COX-1 and COX-2 selectivity as either COX-1 inhibitors, non-selective NSAIDs, or COX-2 inhibitors. We further recommend subcategorizing COX-2 inhibitors as newer COX-2 inhibitors (coxibs) or older COX-2 inhibitors. Finally, we recommend examining the effects of the individual NSAIDs included in each of the proposed categories. Adhering to these recommendations will align future studies, advance interpretation of COX-specific adverse cardiovascular effects, and provide better guidance to clinicians prescribing NSAIDs.

Nonsteroidal anti-inflammatory drugs (NSAIDs) include aspirin and non-aspirin NSAIDs. Non-aspirin NSAIDs were introduced in the early 1960s, and their use to treat pain, fever, and inflammation rapidly increased during the 1970s.

Today, approximately 15% of the general population redeems 1 or more NSAID prescriptions annually. NSAIDs function by inhibiting the cyclooxygenase (COX) enzymes catalyzing the conversion of arachidonic acid into thromboxane A₂ and different prostaglandins with diverse biological effects.  The 2 major COX enzyme isoforms are COX-1, which is continuously expressed in most tissues, and COX-2, which is upregulated in response to inflammatory processes. The analgesic and antipyretic effects of NSAIDs derive mainly from inhibition of COX-2 and the resulting decrease in proinflammatory prostaglandins E₂ and I₂ (prostacyclin),  whereas the gastrointestinal complications (eg, gastric or duodenal bleeding or perforation) derive mainly from inhibition of COX-1 and the resulting inhibition of gastrointestinal protective prostaglandins E₂ and I₂.

The increased risk of gastrointestinal complications associated with COX-1 inhibition promoted the development of newer COX-2 inhibitors (also known as coxibs) in the late 1990s. These agents were anticipated to provide comparative analgesic, antipyretic, and anti-inflammatory effects as the existing (traditional) NSAIDs but with fewer gastrointestinal side effects.  Several coxibs were, however, withdrawn because of an increased thromboembolic risk,  which was biologically explained by an altered equilibrium between the prothrombotic thromboxane A2 and the antithrombotic prostacyclin in favour of platelet aggregation, increased atherogenesis in early stages of atherosclerosis, and elevated systolic blood pressure.

Non-aspirin NSAIDs have, therefore, historically been categorized as traditional NSAIDs or coxibs. Consequently, studies have typically applied the same categorization without distinguishing between the COX-1 and COX-2 selectivity of individual traditional NSAIDs. Nonetheless, traditional NSAIDs also inhibit the COX-1 and COX-2 enzyme isoforms to a varying degree (Fig. 1). This diversity of COX-1 and COX-2 selectivity within the class of traditional NSAIDs has proved clinically important, with evidence accumulating on the cardiovascular risks associated with selective COX-2 inhibition.  Hence, the relative COX-2 selectivity of traditional NSAIDs also correlates with their cardiovascular risk profile, being more favourable for non-selective NSAIDs, such as naproxen and low-dose ibuprofen (< 1200 mg per day), and less favourable for older COX-2 inhibitors: in particular, diclofenac.

Thus, diclofenac has a COX-2 selectivity similar to that of celecoxib (Fig. 1) and has been shown in meta-analyses of both randomized clinical trials and observational studies to increase the risk of adverse cardiovascular events more than other traditional NSAIDs and similar to several coxibs. To illustrate, the NSAID Trialists Collaboration summarized data from 754 randomized clinical trials and found a similarly increased rate of adverse cardiovascular events associated with diclofenac (1.41-fold) and coxibs (1.37-fold) but no increased rate associated with naproxen. Both the European Society of Cardiology and the American Heart Association have acknowledged the link between the COX-2 selectivity of individual NSAIDs and their associated cardiovascular risk. Consequently, both societies recommend using non-selective NSAIDs (naproxen ≤ 500 mg per day or ibuprofen ≤ 1200 mg per day) before COX-2 inhibitors (diclofenac or coxibs) to treat musculoskeletal symptoms in patients with—or at high risk of —adverse cardiovascular events.