Protecting the Stomach Without Trading Away the Heart
A single patient whose two organ systems are each arguing for a different anti-inflammatory choice. The disagreement is between the drug matched to his best-documented risk and the drug matched to his most recent one.
G.D., a 68-year-old retired locksmith, has bilateral knee osteoarthritis severe enough that acetaminophen alone no longer lets him keep up with his grandchildren the way he wants to, and his rheumatologist is ready to start a genuine anti-inflammatory regimen rather than continue managing around its edges. What complicates a straightforward NSAID choice is a chart with real weight on both sides of the risk equation: a peptic ulcer diagnosed eight years ago, successfully treated, that puts him at meaningfully elevated risk for NSAID-induced gastropathy, and a coronary stent placed three years ago for stable angina, now off dual antiplatelet therapy but still on daily aspirin, that puts him at elevated cardiovascular risk from certain anti-inflammatory choices as well. His mildly reduced renal function — an eGFR of 71, not yet a contraindication to either option but a real factor either NSAID choice will need monitored against — adds a third axis to a decision that already has his stomach and his heart pulling in different directions.
COX-2-selective agents like celecoxib spare the COX-1-mediated gastric mucosal prostaglandin production that nonselective NSAIDs suppress, meaningfully reducing ulcer and GI bleeding risk — the CONDOR trial demonstrated exactly this directly, comparing celecoxib against diclofenac-plus-PPI and finding fewer clinically significant GI events with the COX-2-selective drug alone. But COX-2 selectivity also reduces vascular endothelial prostacyclin without touching platelet thromboxane the way aspirin does, a mechanistic imbalance that raised real cardiovascular concern when rofecoxib was withdrawn from the market in 2004 over exactly this signal, and cast a shadow over the entire drug class that outlived rofecoxib itself.
The PRECISION trial, run specifically to settle whether that shadow still applied to celecoxib at moderate doses, found celecoxib noninferior to naproxen and ibuprofen on cardiovascular safety in a population enriched for cardiovascular risk — a genuinely reassuring result, though one with real caveats: high rates of study discontinuation, doses of celecoxib on the lower end of its typical range, and a trial population that, while cardiovascular-risk-enriched, wasn't specifically built around patients with a recent coronary stent the way G.D.'s history requires thinking through carefully.
Two organ systems, two different risk profiles, one drug decision
His ulcer history is the risk I know best how to characterize — confirmed, treated, and exactly the population CONDOR studied directly, comparing celecoxib against diclofenac plus a PPI and finding fewer clinically significant GI events with celecoxib alone.
PRECISION was specifically designed to answer whether celecoxib's cardiovascular safety concerns still applied at moderate doses, and it found celecoxib noninferior to naproxen and ibuprofen in a cardiovascular-risk-enriched population. I'd start celecoxib alone — it addresses his best-documented risk most directly, with real reassurance on the other side.
I take PRECISION seriously, and I'm not arguing celecoxib is unsafe in general.
What I'd weigh differently is that his coronary stent is three years old and specific — PRECISION's population was cardiovascular-risk-enriched broadly, not specifically built around patients this close to a coronary intervention, and the trial carried real discontinuation rates and used celecoxib doses on the lower end of typical range. Naproxen's reputation as the most cardiovascular-favorable nonselective NSAID comes out of the older observational and meta-analytic literature rather than out of PRECISION. If we can address his GI risk adequately with a PPI instead, I'd rather lean on that literature for a patient whose cardiac history is this specific.
I'll take the stent-specificity point, but I want to push back on the premise underneath it. PRECISION is the trial that tested naproxen's reputation head-to-head, and it didn't confirm it — cardiovascular event rates came out comparable across all three drugs, and all-cause mortality actually ran numerically higher on naproxen than on celecoxib. If we're switching for cardiovascular reasons, we should be clear we're doing it on the older literature, not on the trial we've both been citing.
There's a second thing that decides it for me, and it runs the other way from where I started: he takes aspirin every day for a stent. Nonselective NSAIDs compete with aspirin at the same COX-1 site and can blunt the irreversible platelet inhibition he's taking it for — celecoxib, being COX-2 selective, doesn't. So naproxen is defensible here, but only if we dose it apart from his aspirin and say so on the prescription rather than leaving it to chance. On that condition I'd support naproxen plus a PPI: his ulcer risk is manageable with proton-pump co-therapy, and the separation instruction is what keeps the cardiovascular argument from quietly undercutting itself.
Naproxen at the lowest effective dose started alongside daily omeprazole, prioritizing the more cardiovascular-favorable NSAID choice given his recent coronary stent, with PPI co-therapy addressing his ulcer-history GI risk directly.
Agreed cleanly; the Rheumatologist's initial preference for celecoxib wasn't judged wrong on the GI-safety data, only outweighed once the Cardiologist's more specific read of his coronary history was heard directly — both physicians treated the other's risk factor as real, not as the lesser concern.