Advances in gout: some answers, more questions

In a previous issue of the journal, Becker and colleagues present efficacy and safety data from a large study comparing febuxostat to allopurinol. The study showed non-inferiority of febuxostat 40 mg/day in lowering serum urate compared to allopurinol 200 to 300 mg/day. More importantly, the study showed a similar frequency of important cardiovascular adverse events (cardiovascular death, nonfatal myocardial infarction and nonfatal stroke) for febuxostat 40 mg/day (0%), febuxostat 80 mg/day (0.4%) and allopurinol groups (0.4%). Other cardiac adverse event rates (unstable angina, coronary revascularization, cerebral revascularization, transient ischemic attack, venous and peripheral arterial vascular thrombotic event, congestive heart failure, and arrhythmia) were also similar for febuxostat 40 mg/day (1.3%), febuxostat 80 mg/day (1.2%) and allopurinol groups (0.9%). A meta-analysis of safety data from published studies is presented.

For many decades, urate-lowering therapy consisted of allopurinol, a xanthine-oxidase inhibitor and uricosurics, such as probenecid, sulfi npyrazone and benzbromarone. Febuxostat, another xanthine inhibitor, was approved in 2009 in the US, Canada, Japan and the European Union for the treatment of hyperuricemia in gout patients. Gout is associated with high medical comorbidity and defi cits in health-related quality of life. In a previous issue of Arthritis Research and Th erapy, Becker and colleagues [1] describe a non-inferiority randomized controlled trial (RCT) of febuxostat compared to allopurinol in 2,269 patients with gout. Several fi ndings of this important study merit discussion.
Th e cornerstone of gout treatment is achievement of a target serum urate <6 mg/dl. Th is therapeutic goal is based on the solubility of urate at 37°C (6.8 mg/dl), levels below which have been associated with lower risk of gout fl ares and tophi [2][3][4]. A target serum urate was achieved in signifi cantly more patients in the febuxostat 80 mg group (67%) compared to the 40 mg group (45%) and the allopurinol group (42%; 300 mg daily or 200 mg daily in moderate renal impairment); the 40 mg febuxostat dose was not inferior to allopurinol [5]. First, the superiority of febuxostat 80 mg/day to allopurinol 300 mg/day in achiev ing serum urate <6 mg/dl in this study confi rms similar fi ndings from previous RCTs of febuxostat using this or higher doses of febuxostat (120 and 240 mg daily) [5][6][7]. Th is study provides evidence for non-inferiority of a febuxostat 40 mg dose compared to allopurinol in achieving a target serum urate <6 mg/dl. One important fact to bear in mind, however, is that the allopurinol dose should be optimally titrated between 100 and 800 mg/day in clinical settings (or even higher doses in some cases according to experts) to achieve a target serum urate <6 mg/dl, and all febuxostat studies to date have used 200 or 300 mg allopurinol as the comparator. Th erefore, these studies do not answer an important question: is febuxostat superior to a titrated dose of allopurinol in the treatment of hyperuricemia in patients with gout? We know now that allopurinol doses of ≤300 mg/day fail to adequately treat hyperuricemia in 50% of gout patients [8]. It should be noted that, in practice, many physicians do not adequately titrate allopurinol and fail to follow serum urate to achieve the target level [9][10]. Th us, there is limited 'real world logic' in such a study design strategy.
Second, this study provides safety data as one of its important secondary outcomes. Th is fi nding has high relevance since cardiovascular adverse outcomes were an early safety signal in the fi rst large febuxostat RCT, with four deaths in the two febuxostat groups and no deaths in the allopurinol group [5]. In the current study, febuxostat was well-tolerated and not associated with signifi cantly more cardiovascular adverse events compared to allopurinol [1]. Cardiovascular adverse events as defi ned by the Adjudicated Antiplatelet Trialists Collaboration (APTC; cardiovascular death, nonfatal myocardial infarction, and nonfatal stroke) occurred in 0% of patients in the febuxostat 40 mg group, 0.4% in the febuxostat 80 mg group and 0.4% in the allopurinol group, without a

Abstract
In a previous issue of the journal, Becker and colleagues present effi cacy and safety data from a large study comparing febuxostat to allopurinol. The study showed non-inferiority of febuxostat 40 mg/day in lowering serum urate compared to allopurinol 200 to 300 mg/ day. More importantly, the study showed a similar frequency of important cardiovascular adverse events (cardiovascular death, nonfatal myocardial infarction and nonfatal stroke) for febuxostat 40 mg/day (0%), febuxostat 80 mg/day (0.4%) and allopurinol groups (0.4%). Other cardiac adverse event rates (unstable angina, coronary revascularization, cerebral revascularization, transient ischemic attack, venous and peripheral arterial vascular thrombotic event, congestive heart failure, and arrhythmia) were also similar for febuxostat 40 mg/day (1.3%), febuxostat 80 mg/day (1.2%) and allopurinol groups (0.9%). A meta-analysis of safety data from published studies is presented. statistically signifi cant diff erence; non-APTC events (unstable angina, coronary revascularization, cerebral revascularization, transient ischemic attack, venous and peripheral arterial vascular thrombotic event, congestive heart failure, and arrhythmia) occurred in 1.3%, 1.2% and 0.9%, respectively. A review of RCTs of febuxostat showed that cardiovascular adverse events were reported in 1.3 to 2% of febuxostat-treated patients versus 0.4 to 0.9% of allopurinol-treated patients, and death occurred in 0 to 1% of febuxostat-treated patients versus 0 to 0.4% of allopurinol-treated patients (Table 1). Th ese diff erences were not statistically signifi cantly diff erent in each study.
A meta-analysis of data provided in Table 1 was performed by combining the 40 mg, 80 mg and 120 mg daily doses of febuxostat into a single category. Th e relative risk of cardiovascular adverse events (APTC and non-APTC combined) in febuxostat compared to allopurinol (200/300 mg) was 1.7 (95% confi dence interval (CI) 0.4, 7.0) and compared to placebo was 2.5 (95% CI 0.3, 19.4), with neither statistically signifi cant (since the 95% CI included one). Th e relative risks for mortality and serious adverse events was 1.0 (95% CI 0.1, 12.8) and 0.9 (95% CI 0.5, 1.7), respectively, for febuxostat compared to allopurinol; for febuxostat compared to placebo, these were not estimable (no deaths occurred in either group) and 2.5 (95% CI 0.7, 9.0), respectively. Although none of these comparisons were statistically signifi cant, there was a non-signifi cant trend towards more serious adverse events in febuxostat compared to placebo, but no diff erence when compared to allopurinol. A limitation of this meta-analysis is the presence of signifi cant heterogeneity (as assessed by I 2 statistic of >50%): cardiovascular adverse events (I 2 , 48%), mortality (I 2 , 58%) and serious adverse events (I 2 , 41%). A 5-year open-label study [3] of patients taking febuxostat 40 to 120 mg found 18% (21 of 116) suff ered serious adverse events and 5% (6 of 116) suff ered cardiovascular adverse events. Another limitation is that outcomes were assessed at diff erent time points for adverse events due to varying lengths of RCTs.
Th e data regarding safety presented in this study are important. Associations of hyperuricemia with cardiovascular risk are well known [11]. In addition, recent evidence also suggests that gout is an independent risk factor for overall cardiovascular mortality [12] and for myocardial infarction, after adjusting for hyperuricemiaassociated risk [13]. So, why would febuxostat, a medication that lowers serum urate (a cardiac risk factor), One out of 134 (1%) had cardiovascular adverse events, 5 out of 134 (4%) had serious adverse events and none of the patients died in the 240 mg febuxostat group in the study by Schumacher and colleagues [7]. Adjudicated Antiplatelet Trialists Collaboration (APTC) events were defi ned as one or more of the following: cardiovascular death, nonfatal myocardial infarction, and nonfatal stroke. Non-APTC cardiovascular events included unstable angina, coronary revascularization, cerebral revascularization, transient ischemic attack, venous and peripheral arterial vascular thrombotic event, congestive heart failure, and arrhythmia. a Serious adverse events were defi ned as follows in each study: no defi nitions were provided in the Schumacher and colleagues [7], Becker and colleagues [6] or the current study [1]; Becker and colleagues [5] defi ned a serious adverse event as ' …an event that was life-threatening or that resulted in death, hospitalization or prolongation of hospitalization, persistent disability or incapacity, or a congenital anomaly or birth defect. ' increase cardiac adverse events? Th ere are perhaps two important issues that need to be addressed before we seek answers to this question: fi rst, is the cardiovascular risk increased with febuxostat treatment compared to allopurinol or other urate-lowering agents, as was suspected with the initial RCT [5]? And second, is uratelowering below a certain level (<4 or <3 mg/dl or some threshold) undesirable since lowering urate too much or too rapidly may create oxidative stress that could predispose to cardiovascular outcomes? We need welldesigned large database or registry post-marketing surveillance studies with validated, adjudicated cardiovascular outcomes to answer these important questions.
While randomized studies of febuxostat with safety as the primary outcome would also be helpful, clinical trials often are of insuffi cient size, duration of follow-up, or have patients with insuffi cient generalizability to address such questions. It is possible that 'J' or 'U' shaped curves similar to ones observed for blood pressure and stroke risk [14] and diastolic blood pressure and cardiovascular outcomes in patients with coronary artery disease [15] may apply to serum urate and cardiovascular risk as well.
In this exciting era of availability of new gout treatments and high-quality epidemiological and outcomes studies in gout, several existing questions will (need to) be answered and several new ones will likely emerge.

Competing interests
There are no fi nancial confl icts related to this work. No fi nancial support was received for writing this editorial. JAS has received speaker honoraria from Abbott; research and travel grants from Allergan, Takeda, Savient, Wyeth and Amgen; and consultant fees from Savient, URL pharmaceuticals and Novartis.