PRECISION Upends Assumptions on NSAIDs and the Heart

Despite long-standing concerns to the contrary, the sole remaining selective COX-2 inhibitor marketed in the U.S., celecoxib, does not confer excess cardiovascular risk compared with two widely used nonselective NSAIDs. In fact, celecoxib is associated with fewer cardiovascular adverse events than both naproxen and ibuprofen and appears to have lower rates of gastrointestinal (GI) and renal adverse events and perhaps of all-cause mortality.

So reported investigators with the 24,000-patient PRECISION trial, who presented their long-awaited results as a late-breaking trial at the American Heart Association’s Scientific Sessions 2016. The study was simultaneously published online by the New England Journal of Medicine.

“PRECISION shows unequivocally that the cardiovascular risk of the selective COX-2 inhibitor, celecoxib, was not greater than that of the older nonselective NSAIDs, naproxen and ibuprofen,” says Cleveland Clinic Cardiovascular Medicine Chairman Steven Nissen, MD, who chaired the PRECISION study and presented the results. “In fact, celecoxib was arguably the safest of the three NSAIDs studied.”

He adds that while the trial has broad implications, it’s imperative that clinicians provide patients with proper context to allay unjustified fears from overinterpreting its findings. “Patients need to know this study included only individuals taking chronic daily prescription-strength NSAIDs for at least 18 months,” Dr. Nissen notes. “We did not study the occasional use of over-the-counter NSAIDs for periodic aches and pains.”

PRECISION backstory

PRECISION was a 10-year study prompted by the 2004 market withdrawal of the selective COX-2 inhibitor rofecoxib (Vioxx) after disclosure of its association with increased risk of myocardial infarction (MI) and stroke. In response, the FDA required Pfizer to conduct a long-term safety study of its COX-2 inhibitor celecoxib (Celebrex; also available in generic form since 2014) as a condition of keeping the drug on the market.

“Most experts believed celecoxib was probably somewhat worse than nonselective NSAIDs for cardiovascular safety,” says Dr. Nissen, “but we thought it was important to perform the study. When you jump to conclusions, you make mistakes.”

Study design and milestones

The resulting PRECISION trial was directed by the Cleveland Clinic Coordinating Center for Clinical Research (C5 Research) and enrolled patients at 926 centers around the world. Its primary objective was to assess celecoxib for noninferiority versus naproxen and ibuprofen using the Antiplatelet Trialists’ Collaboration (APTC) end point of cardiovascular death (including hemorrhagic stroke), nonfatal MI and nonfatal stroke.

Secondary objectives included comparative safety assessment of celecoxib versus the two other NSAIDs for all-cause mortality and major cardiovascular, GI and renal adverse events.

To ensure robust assessment, the researchers prespecified achieving noninferiority on both intention-to-treat and on-treatment analyses.

Eligible patients were individuals with osteoarthritis or rheumatoid arthritis with established cardiovascular disease or at increased cardiovascular risk who required NSAID therapy for six months or more.

Of 31,857 patients screened, 24,081 were enrolled and randomized to one of the NSAIDs, each of which had a specified prescription-strength starting dose that could be increased for unrelieved symptoms to a prespecified maximum. Mean actual daily dosages were as follows:

• Celecoxib, 104 mg twice daily
• Ibuprofen, 681 mg three times daily
• Naproxen, 426 mg twice daily

Mean drug exposure was 20.3 months, and mean follow-up was 34.1 months. Key patient characteristics included the following:

• Mean age, 63 years
• 64 percent female
• 90 percent had osteoarthritis, 10 percent had rheumatoid arthritis
• Slightly under one-quarter had a prior cardiovascular event
• Slightly more than one-third had diabetes

Major findings

The primary analysis showed celecoxib to meet all specified noninferiority criteria vis-à-vis the other two NSAIDs on the APTC end point. “The lowest cardiovascular event rates were for celecoxib, the highest were for ibuprofen, and naproxen was intermediate,” says Dr. Nissen. “Although differences between individual agents were not statistically significant, the noninferiority test for celecoxib was highly significant, at the P < .001 level. This was an unequivocal result.”

Although the secondary analyses were not adjusted for multiplicity and should be viewed as hypothesis-generating, they yielded a number of notable findings, including the following:

• Relative to celecoxib, naproxen was associated with significantly higher rates of GI adverse events and a borderline significant increase in all-cause mortality in intention-to-treat analyses as well as significantly higher rates of all-cause mortality and major GI and renal events in on-treatment sensitivity analyses.
• Relative to celecoxib, ibuprofen was associated with significantly higher rates of GI and renal adverse events in intention-to-treat analyses as well as significantly higher rates of major adverse cardiovascular events, cardiovascular death, all-cause mortality, and major GI and renal events in on-treatment sensitivity analyses.

Results were consistent regardless of patients’ baseline aspirin use, and the GI safety differences emerged despite prophylactic use of esomeprazole across the study.

Surprises abound

Dr. Nissen says a few of PRECISION’s findings are big surprises, starting with the result on the primary end point. “Like everyone else, I thought celecoxib probably would prove to be a bit worse than the other NSAIDs in cardiovascular risk,” he notes. “The fact that it went in the other direction was very surprising.”

He also cites the significant increase in renal events with ibuprofen as unanticipated, along with the signal for increased all-cause mortality with naproxen in the intention-to-treat analysis.

Implications for practice

One thing that won’t surprise Dr. Nissen is patient concern in the wake of media reports on this study, since 100 million NSAID prescriptions were written in the U.S. in 2013.

He advises clinicians to put such concerns in proper context for patients: “These data do not provide conclusive evidence on the safety of intermittent NSAID treatment or the use of low-dose over-the-counter NSAID preparations. This must be emphasized to patients, who should not be afraid to occasionally use these drugs.” He adds that PRECISION’s results cannot be extrapolated to the more than 20 other available NSAIDs not included in the study.

Beyond that advice, Dr. Nissen sees a few practical takeaways from the trial. “I think the message for rheumatologists is that perhaps the safest of these agents for arthritis patients is celecoxib, which is now available in generic form at reduced cost. For cardiologists, I think it’s now difficult to recommend high-dose ibuprofen for a patient with heart disease, as opposed to over-the-counter doses or other preparations.”

And he thinks PRECISION holds an overarching lesson for the medical community, in light of concerns about equipoise from the study’s conception. “This trial demonstrates the hazards of prejudgment about therapies’ risks and benefits based on expectations and indirect methods,” he says. “It reminds us that we may arrive at erroneous conclusions when we fail to follow a systematic, unbiased approach to scientific and public health questions.”