Febuxostat After CARES: The Uric Acid Stone Former With Coronary Disease
His allopurinol option closed on genetic grounds years ago. Febuxostat is the only other xanthine oxidase inhibitor — in a patient whose coronary disease is exactly the population CARES raised its warning about.
Frank O., a 58-year-old high school chemistry teacher, had a myocardial infarction four years ago and has been on secondary-prevention therapy since, including a statin and low-dose aspirin. He has recurrent gout and, over the past three years, two uric acid kidney stones, the second requiring ureteroscopy six months ago. His 24-hour urine at that time showed a low pH of 5.2 and hyperuricosuria at 850mg/day — a pattern consistent with both stones and confirmed on stone analysis as pure uric acid. Allopurinol, the usual first-line urate-lowering agent, was ruled out two years ago after a pretreatment HLA-B*5801 test came back positive, a genetic marker that carries a substantial risk of severe hypersensitivity reaction to the drug; his rheumatologist did not attempt it.
That leaves febuxostat as the only other xanthine oxidase inhibitor available to him, and Frank's own coronary disease is not incidental to that choice — it is exactly the population the CARES trial enrolled when it found a higher rate of cardiovascular and all-cause death with febuxostat compared with allopurinol in patients with established cardiovascular disease, the finding that produced the drug's current boxed warning. The literature has moved since 2018: the FAST trial, a similarly sized study run across the UK, Denmark and Sweden and designed specifically to re-examine febuxostat's cardiovascular safety with far lower loss to follow-up than CARES managed, found no excess cardiovascular mortality. But FAST is not a bigger, better CARES, and the difference cuts against Frank: it randomized 6,128 to CARES's 6,190, and only about a third of its participants had established cardiovascular disease at entry, which was CARES's entry requirement. Frank sits inside 100% of the population CARES studied and inside roughly a third of FAST's — so the reassuring trial is the one that describes him least well, which is the actual question in front of the room and not a hypothetical extrapolation.
The two trials disagree in a way worth being precise about: a follow-up analysis of the CARES data itself found that a substantial share of the excess deaths occurred only after patients had already discontinued either drug, not while actively taking it — a pattern that motivated FAST's investigators to test the same question with tighter on-treatment ascertainment. That doesn't make CARES's original result wrong so much as it makes FAST a methodological refinement rather than a second opinion — though FAST bought that cleaner follow-up open-label, after a lead-in phase that screened out the least adherent patients before randomization. The FDA's boxed warning was written against the earlier analysis and has not been revised.
In clinic, choosing a urate-lowering strategy
Allopurinol closed as an option on independent genetic grounds — that part isn't up for debate. The real comparison is febuxostat versus nothing, and FAST is a meaningful update on that question: it was built specifically to re-examine CARES's own signal, held onto far more of its patients through follow-up, and found no excess mortality. I'd start febuxostat.
I've read FAST too, and it's genuinely reassuring — for the patients it actually enrolled. Only about a third of FAST's participants came in with established cardiovascular disease. CARES required it of all of them. Frank is not a marginal member of the CARES population, he is a core one, and he is a minority member of FAST's. Citing the trial with the better follow-up doesn't help if its population isn't his. The FDA reviewed that same reassessment literature and chose not to withdraw the boxed warning. I'd rather manage his stone risk through alkalinization alone and leave his urate-lowering therapy off the table for now.
A successor trial addressing a predecessor's limitations is genuinely useful evidence, but one trial reversing a regulatory warning isn't the same as the warning being withdrawn. Until it is, I want the more conservative reading applied to a patient with his specific history, not the more optimistic one.
I don't think alkalinization alone is actually the safe default it sounds like. His gout has been recurrent for years, and untreated gout carries its own real inflammatory and cardiovascular burden over time — this isn't a choice between a risky drug and a risk-free abstention, it's a choice between two different cardiovascular exposures. I'd start febuxostat at the lowest effective dose with cardiology monitoring built in from day one, rather than either extreme.
Agreed: start febuxostat at low dose with cardiology monitoring, alongside potassium citrate to alkalinize his urine independently of the urate-lowering strategy.
Not agreed: the cardiologist signed off on the plan under monitoring but stated directly that she would have preferred alkalinization alone if his gout burden had been milder; the urologist and clinical pharmacologist did not share that reservation.