Practice-Changing PRECISION Subanalyses

By M. Elaine Husni, MD, MPH

PRECISION trial results have challenged many assumptions about the use of nonselective NSAIDs versus selective COX-2 inhibitors.

NSAIDs are normally classified by the relative selectivity of COX-1 and COX-2 enzymes. Nonselective NSAIDs, such as naproxen and ibuprofen, inhibit both COX-1 and COX-2. Selective NSAIDs, such as celecoxib, are COX-2-specific and were developed to spare COX-1 inhibition to allow more gastrointestinal (GI) protection.

Many osteoarthritis (OA) and rheumatoid arthritis (RA) patients rely on celecoxib, the only COX-2 inhibitor still marketed in the U.S., but we assumed that they faced a greater risk for cardiovascular (CV) disease.

PRECISION data tell us that’s not so. PRECISION found celecoxib to be as safe as naproxen and ibuprofen in terms of CV risk. In patients with OA, celecoxib carries less CV risk than ibuprofen and similar risk to naproxen, and less GI risk than ibuprofen and naproxen. Celecoxib was similar to both ibuprofen and naproxen in all-cause mortality. In patients with RA, the study found no difference in the rates of major CV and renal adverse events among the three drugs but found a doubling of all-cause mortality in patients who used naproxen vs. celecoxib.

Since the original results were published, many subanalyses have dissected the data for its relevance to particular populations or disease states. I find the below studies of particular relevance for practicing rheumatologists who frequently help patients balance the benefit of arthritis medications with the risk of comorbidities.

Aspirin coadministration and CV prevention in arthritis patients that use NSAIDs

An important PRECISION substudy from our colleagues at Cleveland Clinic shows that adding aspirin attenuates celecoxib’s safety advantage over the nonselective NSAIDs naproxen and ibuprofen, but that celecoxib with aspirin still has an equal or better safety profile (in regards to GI and renal events) relative to both agents. The study evaluated the trial’s on-treatment population for both OA and RA, which consisted of 11,018 patients taking concomitant aspirin and 12,935 patients not on aspirin. Propensity score weighting was used to adjust for baseline characteristics, thereby increasing the validity of comparisons.

Another substudy published in Rheumatology tested the hypothesis that RA patients have a different risk-benefit profile for the use of aspirin in secondary CV risk prevention. Of 1,852 subjects with RA in PRECISION, 540 reported using low-dose aspirin for CV prevention and 1,312 did not. We observed major NSAID toxicity in 79 (6.0 percent) of non-aspirin users and 37 (6.9 percent) aspirin users (P = 0.50). Thus, in the RA population, low-dose aspirin users experienced the same rate of primary outcome as non-aspirin users. The risk of major adverse CV event was similar as well.

These findings highlight the importance of appropriately counseling arthritis patients on safety profiles, especially when they are taking multiple medications. Remember:

• There were very few CV events observed in arthritis patients on the studied NSAIDs over 18 months.
• Selective NSAIDs did not indicate a higher CV risk than nonselective NSAIDs.
• Using aspirin may decrease the CV safety advantage of selective NSAIDs — although there was no difference between aspirin users and non-aspirin users in the RA population.

GI safety in arthritis patients

Another subanalysis examined the overall GI safety of celecoxib, ibuprofen and naproxen in arthritis patients on concomitant esomeprazole and low-dose aspirin or corticosteroids. Our randomized, double-blind controlled trial published in Alimentary Pharmacology & Therapeutics found that celecoxib had a safer GI profile overall compared with ibuprofen or naproxen for patients with RA and OA. The primary endpoints were clinically significant GI events (CSGIEs), including bleeding, obstruction, perforation events from the stomach downward or symptomatic ulcers, and iron deficiency anemia.

Patients received 100 to 200 mg celecoxib twice daily (N = 8,072), 600 to 800 mg ibuprofen three times daily (N = 8,040) or 375 to 500 mg naproxen twice daily (N = 7,969) as well as 20 to 40 mg esomeprazole for gastroprotection. CSGIEs occurred in 0.34, 0.74 and 0.66 percent of patients receiving celecoxib, ibuprofen and naproxen, respectively. There also was less iron deficiency anemia in patients on celecoxib than those on naproxen or ibuprofen. Helicobacter pylori status was also studied but did not influence the outcome.

Interestingly, concomitant corticosteroid use increased total GI events and CSGIEs. Our data show that CSGIEs are infrequent in patients with OA and RA taking NSAIDs plus esomeprazole, but celecoxib has better overall GI safety than ibuprofen or naproxen at these doses regardless of concurrent low-dose aspirin or corticosteroid use.

As rheumatologists, we can be reassured that patients with OA and RA who may need prolonged use of NSAIDs for joint pain have relatively infrequent CSGIEs. These results allow us to consider celecoxib for patients at higher risk of CSGIEs.

Tailoring treatment for patients with multisystemic disease or high risk

These subgroup analyses are hypothesis generating rather than definitive, but they help address gaps in our knowledge related to chronic NSAID use in patients with OA and RA. First, the overall data allow us a more tailored approach to treatment for patients using selective and nonselective NSAIDs, especially when the need may be chronic. Second, subanalyses allow us to offer a more nuanced approach for patients with comorbid GI and CV disease and/or increased risk, such as the elderly population or those on dual therapy with low-dose aspirin or corticosteroids.

Although several PRECISION substudies have generated interesting data, those mentioned above are particularly relevant to our daily practice as clinical rheumatologists and should inform our interactions with and recommendations for patients.

Dr. Husni is Director of the Arthritis and Musculoskeletal Treatment Center and Endowed Chair of Translational Functional Medicine Research at Cleveland Clinic.