CARES: Mortality Risk Hinted for Patients at High CV Risk Receiving Gout Drug Febuxostat

March 13, 2018

ORLANDO — The risk for cardiovascular (CV) events was similar whether patients at high cardiovascular (CV) risk in a randomized "noninferiority" trial took febuxostat (Uloric, Takeda Pharmaceuticals) or allopurinol for the treatment of gout.

That primary outcome — in a postmarket safety study stipulated by US regulators when febuxostat was approved in 2009 — was tempered by a secondary finding of significantly increased all-cause mortality, driven by CV deaths, in patients taking the newer drug compared with mortality in those taking allopurinol.

The comparison of the two xanthine oxidase inhibitors, called Cardiovascular Safety of Febuxostat and Allopurinol in Patients with Gout and Cardiovascular Comorbidities (CARES), followed 6190 patients who took the drugs for an average of 2 and a half years.

CARES was published March 12 in the New England Journal of Medicine, with William B White, MD, University of Connecticut, Farmington, as lead author, to coincide with his presentation of the study here at the American College of Cardiology (ACC) 2018 Annual Scientific Session.

Dr William B White

Enrolled patients were required to have a CV history that included coronary disease, cerebrovascular or peripheral vascular disease, or diabetes with small-vessel disease.

That entry requirement was reflected in a high rate of the composite primary endpoint, explained White in an interview. That endpoint consisted of cardiac mortality, nonfatal myocardial infarction (MI), nonfatal stroke, and cardiac ischemia requiring urgent surgery, which reached 10.8% for patients getting febuxostat and 10.4% for those treated with allopurinol.

The adjudicated rates of sudden cardiac death were 2.7% for febuxostat and 1.8% for allopurinol, respectively, and rates of CV death were 4.3% and 3.2% (P = .03), respectively.

A similar pattern of outcomes was seen in both the trial's primary modified intention-to-treat (mITT) analysis and an "as-treated" analysis.

Mechanism Unknown

White also pointed to an unexplained quirk of the results, the combination of a noninferior primary endpoint for febuxostat that included CV death but secondarily a jump in all-cause mortality and especially CV death.

Adjudication of clinical events showed no significant differences between the treatments in nonfatal events, including MI, revascularization, arrhythmias, or heart failure hospitalization.

The mortality signal's cause has been elusive, White said. Extensive preclinical studies didn't turn up any possible proarrhythmic, prothrombotic, atherosclerotic, or cardiac functional effects that might account for it.

Despite no differences in individual nonfatal events and given little understanding of the mechanism of harm, "the results of CARES raise continued concerns regarding the safety of febuxostat," said Michelle L O'Donoghue, MD, MPH, Brigham and Women's Hospital, Boston, Massachusetts.

Although the trial met its noninferiority endpoint, "The focus must still be on the potential for risk with febuxostat at this time," O'Donoghue, who isn't connected with the trial, said as a discussant at one of White's presentations on CARES.

FDA Cautions on the Mortality Hazard

The main CARES outcomes message had been announced in preliminary form in November 2017 in a safety communication from the US Food and Drug Administration (FDA), which cautioned about an excess of "heart-related deaths and death from all causes" with febuxostat.

The FDA alert also pointed out that there was no such similar increase in risk for the primary endpoint.

What the agency will do with the new CARES data is unknown, White told | Medscape Cardiology. Options include a new warning in the labeling in addition to the one already there, or pulling febuxostat from the market.

A safe alternative to allopurinol for gout would be welcome. The many gout patients with renal disease don't tolerate it, White observed, and it's associated with sensitivity skin reactions, especially in African Americans.

"Those would be populations in which we'd always thought febuxostat would be a better idea than allopurinol."

Of note, patients with more advanced kidney disease tended to have better clinical outcomes with febuxostat than those without kidney disease, White said. "That is, they did not have a significant increase in the cardiovascular mortality that we saw in the population as a whole."

Indeed, observed Eileen Handberg, PhD, University of Florida, Gainesville, there are compelling reasons to keep febuxostat available, given the lack of more definitive evidence of a significant hazard.

Many drugs with useful roles have shown signals of risk, Handberg, who isn't with the CARES trial, told | Medscape Cardiology.

If regulators were to pull them from the market on the basis of such signals, "then you'd take away a patient's ability to make an informed decision with their provider. The rheumatologists deal with a very chronically ill group of patients, and if either drug is doing well by them, are we going to remove it?" she said.

"Let's better understand the signal before we make that decision." If the risk can be identified, she said, then it can be up to patients and providers to decide whether to accept it.

Handberg also noted that the high rate at which patients dropped out of CARES, about 45% in both the febuxostat and allopurinol groups, "shows that this type of patient isn't very satisfied with either therapy."

North American Trial

CARES assigned the patients with gout and a CV history to receive one or the other agent, in double-blind fashion, at 320 centers in the United States, Canada, and Mexico; treatment continued for up to 6 years.

The 3098 patients in the febuxostat group received the drug at 40 to 80 mg/day regardless of renal function. The 3092 patients assigned to allopurinol received 200  to 600 mg/day depending on serum urate levels.

The two groups were similar with respect to baseline CV medications and prevalence of prior use of gout drugs, about 62%; 4.2% in both groups had previously used febuxostat.

Gout flare-ups during follow-up were about as common in one group as the other: 0.68 and 0.63 per patient-year for febuxostat and allopurinol, respectively.

Table. Hazard Ratios (95% CIs) for Clinical Outcomes, Febuxostat vs Allopurinol

Endpoints Modified ITT Analysis On-Treatment Analysis
Primary endpoint 1.03 (0.87-1.23)a 1.00 (0.82 - 1.22)b
Death from any cause 1.22 (1.01 - 1.47) 1.26 (0.93 - 1.72)
CV death 1.34 (1.03 - 1.73) 1.49 (1.01 - 2.22)
Nonfatal MI 0.93 (0.72 - 1.21) 0.87 (0.66 - 1.34)
Nonfatal stroke 1.01 (0.73 - 1.41) 0.94 (0.66 - 1.34)
Urgent revascularization due to unstable angina 0.86 (0.59 - 1.26) 1.00 (0.66 - 1.52)
a P = .002 for noninferiority.

b97.0% CI.


The high rate of study drug discontinuation in CARES makes interpretation of the trial a challenge, O'Donoghue said. "The on-treatment analysis here is important because many of these deaths were occurring off of study drug."

Point estimates for both all-cause mortality and CV death were "very comparable" in the mITT and on-treatment analyses. The trial was "reassuring for cardiovascular events overall, and in nonfatal events we did not see any directional signal."

White said the apparently bidirectional outcomes for CV death in the composite primary endpoint vs the isolated secondary endpoint in both the mITT and on-treatment analyses suggests that the effect is somehow real and not by chance.

CARES was funded by Takeda Development Center Americas. White discloses receiving consulting fees or honoraria from Novartis. O'Donoghue reports receiving research grants from Amgen, Eisai, Janssen Pharmaceuticals, The Medicines Company, and Merck. Handberg discloses receiving research grants from Aastrom Biosciences, Amorcyte, Biocardia, Ccapricor, Cytori Therapeutics, Direct Flow Medical, Everyfit, Medtronic, Merck, Mesoblast, PCORI, and Sanofi Aventis; and unspecified relationships with Amgen, AstraZeneca, Boerhinger Ingleheim, Daiichi Sankyo, Gilead Sciences, Ionis, and Relypsa.

American College of Cardiology (ACC) 2018 Annual Scientific Session. Abstract 408-08. Presented March 12, 2018. 

N Engl J Med. Published online March 12, 2018. Full text

Follow Steve Stiles on Twitter: @SteveStiles2. For more from, follow us on Twitter and Facebook.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.