QD Clinic: Allopurinol for All Gout? Not Quite Save

Hey everyone, my name is Brian Jaros. I am one of the rheumatologists at Northwestern University in Chicago. And today I'll be talking about a gout case that's a little bit bread and butter, but I think has some really important learning pearls that I wanted to discuss with you all.

So this is a 71-year-old male. He has a past medical history of high blood pressure, a remote history of MSSA bacteremia which ended up causing necrotizing glomerulonephritis, a long-standing history of untreated gout at an outside hospital or institution, but presented to our hospital for three days of worsening joint pain. Was in his usual state of health when he developed very quick onset pain and swelling in his right ankle similar to what he associated with prior gout flares but without a particular trigger.

On exam he was actually found to be febrile to 102°F. His other vitals were normal. But he had an appreciable tophus over his left elbow as well as over the right patellar tendon, kind of confirming this suspicion for long-standing gout. He had moderate warmth, tenderness, swelling of the right ankle that he was complaining about, but also actually of the right wrist and the right elbow. And his labs showed a uric acid of 10, a very elevated CRP to 95, and ESR of 42.

So although the suspicion was high for gout, given the fevers and his history of bacteremia, he did undergo arthrocentesis of the right ankle which showed a massive number of white blood cells — 120,000. It did confirm gout crystals, and cultures of that synovial fluid but also of the blood were negative at 48 hours. So ultimately he was treated with colchicine and a prednisone taper with good success.

When we were considering initiating allopurinol, we obtained additional history from him that he was actually of Han Chinese ancestry, and so initiation of allopurinol was deferred in favor of testing for HLA-B*58:01. Two weeks later this came back as positive.

So that's one of the things I wanted to talk about today, and take kind of a step back to talk about HLA-B*58:01. I'll call it B-58 for the purposes of this video, so you don't have to keep listening to me say the full name. But refreshing ourselves on what HLA even is — HLA stands for human leukocyte antigen and refers to a complex gene family that encodes proteins that are on most of our cell surfaces, and they help our immune system distinguish between our own body and foreign invaders, so kind of a mechanism to help regulate against autoimmunity. There are three major genes, HLA-A, B, and C, and they contain different alleles or arrangements that are referred to by a numbering system. Most of us in rheumatology are probably familiar, for example, with HLA-B27. These specific alleles can then confer individual risk for certain conditions, certain autoimmune diseases, or even drug hypersensitivity.

And so why is B-58 relevant to gout? Well, this allele is actually associated with an increased risk of hypersensitivity reaction to allopurinol. This is one of the more potentially serious adverse events of allopurinol use. A 2025 meta-analysis by Fam et al. looked at over 13,000 patients from 24 case-control studies and found that carrying the HLA-B*58:01 allele was associated with a massively increased risk of severe cutaneous reactions to allopurinol, with an odds ratio of 117.6 compared to people without the allele. So obviously this is a really major risk factor, and it's relevant to us and our patients because these cutaneous reactions are more than just a little rash. They can really develop into serious manifestations — things like Stevens-Johnson syndrome, DRESS — which as many of you know carries significant morbidity and actually even mortality.

The allele is more frequent in certain ancestral populations, specifically those of South and East Asian descent — populations like Han Chinese, Korean, Thai, Filipino — as well as people of African descent. And so as a result, the ACR 2020 clinical practice guidelines for gout actually conditionally recommend testing for this allele in at-risk people of Southeast Asian descent and in African-American patients prior to starting allopurinol. They determined that testing in these at-risk populations is probably a cost-effective strategy, but actually conditionally recommend against universal testing in all populations, probably due to a lower prevalence of the allele and an unclear cost-benefit.

So in our patient we found this allele; we know he's at higher risk for allopurinol-associated cutaneous reaction. So how do we proceed? Clearly he needs long-term treatment for tophaceous gout, and the consideration here is febuxostat. Febuxostat is a non-purine xanthine oxidase inhibitor, similar but not identical in mechanism to allopurinol, and is effective to reduce uric acid levels in gout. The typical starting dose is around 40 mg and you can
uptitrate this to the FDA dose of 80 milligrams, but some countries use all the way up to 120 milligrams daily. So the question would be if this works like allopurinol and is effective, why is this also not a first-line option that we typically use for gout.

And the answer is there's a concern about additive cardiovascular risk with use of febuxostat. So in early trials of febuxostat there was maybe a modest signal of a higher rate of cardiovascular events compared to allopurinol use in gout patients. And so the FDA actually mandated a postmarketing study of febuxostat, which came out in 2018, and that was called the CARES study, and this sought to examine whether febuxostat was non-inferior to allopurinol with regards to MACE safety in gout. It was a massive study. Over 6,000 patients with gout and with a history of cardiovascular disease were randomized to either febuxostat or allopurinol with a primary endpoint being a composite of MACE essentially.

And in brief summary, the median follow-up was 32 months and febuxostat actually was non-inferior to allopurinol with regards to the primary composite MACE endpoint. However, when they broke that composite into individual components, febuxostat actually demonstrated a higher rate of cardiovascular death. So 4.3% in the febuxostat group versus 3.2% in the allopurinol group, and then this drove a higher all-cause mortality compared to allopurinol. So as a result, an additional black box warning was added by the FDA for febuxostat for overall mortality in addition to cardiovascular risk, and emphasized judicious use in select populations really when allopurinol could not be used.

Some context and important things to note about the trial and criticisms would be that there was a very very high rate of discontinuation of drug — over 50% of people — and also a large number were lost to follow-up, around 45%. It's also worth noting that while the patients had a history of cardiovascular disease, so were a high-risk population, many of the patients were still not on risk-optimizing medications, for example like lipid-lowering treatment. And mechanistically, it really remained unclear why febuxostat specifically would increase cardiovascular death compared to allopurinol.

So do we have other data that replicates this finding? The answer is we have other data but actually challenges this finding. So in 2020 a European group published the FAST trial, which was prospective and randomized. It was open label but blinded to the endpoint and non-inferiority. I won't go — for time's sake — into the whole ins and outs of this study, but it's worth noting that these were patients with cardiovascular risk factors but not necessarily disease. So a little bit of a lower-risk population compared to CARES. It was similarly a huge study with over 6,000 patients, a long follow-up median treatment of 3.6 years. And in their study, febuxostat was non-inferior both for the primary composite endpoint but also showed no increased cardiovascular death or all-cause mortality, which was a difference compared to the finding in CARES. And notably in FAST there was a much much lower loss-to-follow-up rate — so only about 5.8% of patients were lost to follow-up.

I'll also mention a systematic meta-analysis in 2021 by Gao et al. also found no difference between febuxostat and allopurinol in terms of cardiovascular mortality.

So at this point clinicians are a little bit left to make determinations about whether febuxostat really carries this increased cardiovascular risk or not. My current practice is to follow the FDA guidelines to limit febuxostat use to those who are intolerant or who fail allopurinol, and I will use it cautiously in patients who have a CV history with a lot of effort to control any risk factors that are modifiable.

I would also point out that we have emerging data that pretty solidly suggests that gout flares themselves are a risk factor for cardiovascular events. So ultimately we need to get adequate control in these people regardless of the agent, and if they can't take allopurinol then I think febuxostat really remains a viable option. And so in our patient, he was started on febuxostat 40 milligrams, ultimately uptitrated to 80 milligrams daily, now with a serum uric acid of 4.1, which is below our goal of five for tophaceous gout, and he's doing well without further flares.

So I hope this was an interesting case — a little bit of a review of HLA-B*58:01 and the risk for allopurinol hypersensitivity, and also of febuxostat dosing, safety considerations, when to use it, and what precautions we might take. Thanks for listening and take care everybody.