Who CARES? Getting to the Heart of Allopurinol and Febuxostat CV Outcomes

Author(s)
Sophia Dietrich, PharmD
Michael W. Nagy, PharmD, BCACP

Reviewed By
Stefanie Nigro, PharmD, BCACP
M. Shawn McFarland, PharmD, BCACP

Citation
White WB, Saag KG, Becker MA, et al. Cardiovascular safety of febuxostat or allopurinol in patients with gout. N Engle J Med. 2018 Mar 29; 378(13): 1200-1210.

The Problem

Should we target uric acid levels when working to reduce cardiovascular risk? Cardiovascular (CV) disease remains the leading cause of death with many contributing risk factors, including hyperuricemia.¹ Evidence suggests the elevations in uric acid levels is associated with and can lead to worse outcomes for individuals with CV disease and heart failure.^2-4 According to the Preventative Cardiology Information System Database Cohort Study every 1 mg/dL increase in serum urate concentration increases the relative risk of all-cause death by 39%.³ The Cardiovascular Safety of Febuxostat and Allopurinol in Patients with Gout and Cardiovascular Morbidities (CARES) trial was conducted to evaluate whether febuxostat was noninferior to allopurinol with regard to CV events in patients with gout and CV disease.^2,4   During the early clinical trials, there was a suggestion that febuxostat might modestly increase the risk of cardiovascular events which prompted the FDA to require this CV safety trial.²

What’s Known

Gout is a form of arthritis that results from the oversaturation of monosodium urate crystals occurring at uric acid serum concentrations in excess of 6.8 mg/dL.^5,6 Addressing hyperuricemia through lifestyle modifications and pharmacotherapy are the cornerstone of gout prevention and treatment. If urate lowering pharmacotherapy is indicated, a xanthine oxidase inhibitor (XOI), such as allopurinol, is first line and should be dosed to target urate concentration of less than 6 mg/dL. In patients with gout who develop skin rash to allopurinol or where allopurinol does not lower urate concentrations sufficiently, febuxostat is an appropriate alternative agent.

Given the connection between serum urate levels and cardiotoxicity, there is support for the hypothesis that urate lowering therapies may have cardioprotective effects.^2-4  A meta-analysis conducted in 2018 demonstrated a decrease in incident perioperative myocardial infarction (MI) following coronary artery bypass grafting (CABG) with allopurinol use.⁷  Other studies concluded that uric acid lowering therapy may be beneficial in the management of CV disease.^8,9 Uric acid activates inflammatory cytokines in the vascular endothelium which is correlated to worse clinical outcomes in chronic heart failure (CHF) patients. This observation prompted the LEAF-CHF trial which is currently exploring the potential protective effects of febuxostat in patients with gout and CHF.¹⁰  A pilot study in patients with hypertension and hyperuricemia suggested that febuxostat may improve the renin-angiotensin-aldosterone system (RAAS) and renal function.¹¹  While intriguing, these studies are not CV outcomes trials – prompting the FDA to require the manufacturer to conduct the CARES study.¹²

What’s New

CARES was a 32-month randomized, double-blinded, multicenter, noninferiority trial conducted between April 2010 and May 2017.² A total of 6,190 patients randomized to receive either febuxostat (N=3,098) or allopurinol (N=3,092). Febuxostat was initiated at 40 mg daily and titrated to 80 mg based on uric acid levels, while allopurinol was initiated at a dose based on renal function and titrated to achieve a uric acid level < 6mg/dL. All patients were initiated on colchicine 0.6 mg daily for gout flare prophylaxis. Those with intolerable side effects to colchicine were given naproxen and lansoprazole.

The primary endpoint of the study was major cardiovascular adverse effects (MACE) defined as the composite of CV death, nonfatal myocardial infarction, nonfatal stroke, or urgent revascularization due to unstable angina. Secondary endpoints included the individual components of the primary end point, the composite of CV death, nonfatal stroke and myocardial infarction, as well as death from any cause.

The baseline characteristics of the febuxostat and allopurinol groups were similar (Table 1). The use of cardioprotective medications at baseline and during trial were similar between groups including aspirin (~60%), clopidogrel (~18%), beta blockers (~58%), lipid lowering agents (~74%), and renin-angiotensin inhibiting agents (69%).

Table 1: Baseline Characteristics in CARES^2,3

Characteristic	Febuxostat (N=3098)	Allopurinol (N=3092)
Median age – yr	64.0	65.0
Age ≥65 yr – %	48.9	51.3
Male sex – %	84.1	83.8
Duration of gout – yr	11.8±11.4	11.9±11.2
Baseline serum urate level – mg/dL	8.7±1.7	8.7±1.7
Presence of tophi – %	21.6	21.0
Body-mass index	33.6±7.0	33.4±6.9
Race or ethnic group – no. (%)
White	69.7	69.2
Black	17.8	19.2
Asian	3.0	3.1
American Indian or Alaska Native	8.5	7.6
Other	1.0	1.0
Cardiovascular risk factors and history – no. (%)
Diabetes mellitus with small-vessel disease	38.5	69.2
Hypertension	92.4	92.2
Dyslipidemia	86.4	87.4
Myocardial Infarction	38.6	39.8
Coronary revascularization	36.4	38.2
Congestive heart failure	20.1	20.4
Stroke	14.8	13.3
Peripheral vascular disease	13.3	12.1
Stage of chronic kidney disease – no./total no. (%)
Stage 1 or 2	47.1	47.2
Stage 3	52.9	52.8

Although the CARES trial was not assessing the urate lowering efficacy of febuxostat compared to allopurinol, there was a slightly greater number of patients in the febuxostat group that had a serum urate level of less than 6.0 mg.

The rate of discontinuation was high but similar in both groups. In total, 56.6% of patients discontinued the trial prematurely (57.3% in the febuxostat group and 55.9%  in the allopurinol group).

The primary endpoint occurred in 10.8% of patients in the febuxostat group and 10.4% of patients assigned to allopurinol [HR 1.03; 97% CI 0.87-1.23; P=0.002 for noninferiority]. However, death from any cause and CV death occurred more frequently in the febuxostat group. See Table 2.

Table 2: Primary and Secondary Outcomes of CARES^2,3

	Endpoint
	Primary endpoint	CV death	Nonfatal MI	Nonfatal stroke	All-Cause Mortality
Febuxostat -%	10.8	4.3	3.6	2.3	7.8
Allopurinol -%	10.4	3.2	3.8	2.3	6.4
Hazard ratio for Febuxostat Group (95% CI)	1.03 (0.87 -1.23)	1.34 (1.3-1.73)	0.93 (0.72-1.21)	1.01 (0.73-1.41)	1.22 (1.01-1.47)
P-value for superiority test	0.66	0.03	0.61	0.94	0.04

Abbreviations: CV, cardiovascular; MI, myocardial infarction

Our Critical Appraisal

CARES was the largest and longest randomized controlled trial to investigate the potential long-term CV effects of urate lowering therapy in patients with gout. Additionally, it was the first to enroll patients with a history of major CV disease and first trial designed to evaluate MACE. The urate lowering therapy was titrated appropriately based on uric acid concentrations with allopurinol doses ranging from 200 to 600 mg daily.  Although the primary composite outcome showed noninferiority between febuxostat and allopurinol for the composite MACE, the secondary individual endpoints of CV and all-cause mortality demonstrated inferiority for febuxostat. The difference in all-cause mortality was driven by the difference in CV morality. In a subgroup analysis the authors reported that patients who used NSAIDs or did not use low-dose aspirin had higher rates of CV mortality with febuxostat. We know NSAIDS increase risk of MACE and there may be an interaction between NSAIDS and febuxostat. Additionally, aspirin use is well known to reduce CV mortality with established CV disease.

Two recent meta-analyses reviewed the CV benefits of febuxostat versus allopurinol or placebo and found similar findings when including the CARES trial data.^13-14 When CARES was removed from the meta-analysis, the results for CV mortality were not significantly different.¹³ While one might be tempted to conclude that febuxostat is harmful, if we dive deeper into prior literature, both allopurinol and febuxostat seem to improve CV outcomes when compared to placebo.^7-9,15 Since a placebo controlled trial would have been unethical and CARES was a noninferiority trial intended to compare febuxostat and allopurinol, it would be inappropriate to conclude that febuxostat increases the risk of CV death.

The CARES study has several limitations and the results shouldn’t be extrapolated to other populations. Noninferiority trials should establish a non-inferiority margin based on an effect size that is deemed clinically relevant.  Unfortunately, the authors of CARES did not explain how they selected the noninferiority margin. Also, in noninferiority trials, an intention to treat analysis can increase the risk of wrongly claiming noninferiority. Thus, a per protocol sensitivity analysis should be included and is often the preferred method of analysis.  A per protocol analysis was not reported in the CARES manuscript. It is important to note that most deaths (90%) occurred after discontinuation of study drugs (see supplement tables 9 and 12 in the original study).  The high rates of treatment discontinuation and lost to follow-up may have also biased the results. Additionally, patients in CARES had a high baseline CV risk but there were relatively low rates of preventative pharmacotherapy use (aspirin, lipid-lowering treatment, beta blockers).  The results can’t be extrapolated to patients at low CV risk who do not have established CV disease. Finally, this study did not enroll patients with a creatinine clearance of < 30 mL/min.

The Bottom Line

Gout increases the risk of CV disease and in prior studies both allopurinol and febuxostat seem to improve CV outcomes when compared to placebo. CARES found febuxostat to be inferior to allopurinol in terms of CV death, however the study had several limitations. Given its lower cost, allopurinol should remain the first line agent for urate lowering therapy. However, in patients who cannot tolerate allopurinol or when allopurinol does not lower urate levels sufficiently, febuxostat remains a reasonable alternative. CARES does raise an important question – does febuxostat increase the risk of CV related death? Or does it decrease the risk of CV related death but to a lesser extent than allopurinol?  Hopefully, the ongoing FAST and FREED trials will provide us with definitive answers.^15,16

The Key Points

• In patients with a history of CV disease, use of allopurinol or febuxostat for urate lowering therapy had similar impact on MACE. However, allopurinol was superior in preventing CV mortality.
• The CARES trial has several limitations but raises an important question about whether febuxostat increases the risk of CV related death or decreases the risk of CV related death to a lesser extent than allopurinol. Trials are currently underway to evaluate this question.
• Allopurinol should remain the standard of care when urate lowering therapy is indicated. However, febuxostat would be an appropriate alternative in those who cannot tolerate allopurinol or who do not achieve their target serum uric acid concentration.

FINAL NOTE: This program will be available for recertification credit through the American Pharmacists Association (APhA) Ambulatory Care Review and Recertification Program. To learn more, visit https://www.pharmacist.com/ambulatory-care-review-and-recertification-activities.