Xanthine oxidoreductase (XOR) is the rate-limiting enzyme in purine catabolism and converts hypoxanthine to xanthine, and xanthine into uric acid. When concentrations of uric acid exceed its biochemical saturation point, crystals of uric acid, in the form of monosodium urate, emerge and can predispose an individual to gout, the commonest form of inflammatory arthritis in men aged over 40 years. XOR inhibitors are primarily used in the treatment of gout, reducing the formation of uric acid and thereby, preventing the formation of monosodium urate crystals. Allopurinol is established as first-line therapy for gout; a newer alternative, febuxostat, is used in patients unable to tolerate allopurinol. This review provides an overview of gout, a detailed analysis of the structure and function of XOR, discussion on the pharmacokinetics and pharmacodynamics of XOR inhibitors–allopurinol and febuxostat, and the relevance of XOR in common comorbidities of gout.
Allopurinol is an effective inhibitor of xanthine oxidoreductase (XOR), thereby reducing the formation of uric acid, the end-product of the pathway of purine synthesis and breakdown (Figure 1). Uric acid has a pKa of about 5.8 and the main form in plasma is the anion, urate.
Pathways of synthesis and breakdown of purines
Allopurinol is a purine analogue and was synthesized and studied by Nobel Prize winners in Medicine, Dr George Hitchings and Dr Gertrude Elion in their programme to discover therapies for cancer . Serendipitously, the substantial plasma urate lowering effect was observed and was approved for the treatment of gout by the FDA in 1966. This drug is now established as the first-line urate lowering therapy (ULT) for the treatment of gout, reducing the recurrence of acute attacks and decreasing the body load of monosodium urate, which is associated with inflammation, tissue deposits and organ damage [2,3]. Gout is prevalent, increasingly so, including in the developing world, so that there is considerable morbidity and socio-economic impact as well as substantial world-wide use of allopurinol [4,5].
Alternate ULT includes uricosuric agents probenecid, benzbromarone and, recently, lesinurad, but the use of these is minor in comparison with allopurinol (Figure 2) . Uricosuric efficacy is hampered to a variable extent in renal impairment, common in gout patients, and by adverse effects, principally formation of renal stones, most commonly upon initiation of therapy. Serious hepatotoxicity led to the withdrawal of benzbromarone in some markets; however, it continues to have a useful, if limited role in patients unable to tolerate other ULT . More recently an alternate XOR inhibitor has been introduced into clinical practice, namely febuxostat, the initial rationale being for its use in patients intolerant of allopurinol (Figure 2) . The advent of this inhibitor of XOR has been helpful in gaining new insights into XOR biology. ULT is extremely effective for most patients with gout yet, as noted, the prevalence of gout is great and increasing. This commentary examines this problem from a broad perspective but also with a focus upon recent insights into the structure and functions of XOR and its inhibition that is of potential relevance for gout and its treatment and for some of the common co-morbidities of gout.
Chemical structures of xanthines, major XOR inhibitors and their metabolites
The predominant risk factor for gout is an elevated plasma urate concentration, indicative of an increased total body load of uric acid. Hyperuricaemia is primarily due to reduced renal clearance of urate, a result of reduced glomerular filtration rate in those with renal impairment and/or reduced fractional clearance of urate related to the balance struck between the reabsorption and secretory transporter proteins responsible for urate transfer across the tubule [9,10] (Figure 3). Several transporters are involved in the renal handling of urate with resorption of urate via URAT1 and GLUT9 among the most significant [11,12]. An important contribution from reduced efflux transport of urate across the gut has recently been established, implicating the variants of the ABCG2 transporter [10,13].
The renal tubular handling of urate and oxypurinol
Gout first presents with acute, extremely painful inflammatory arthritis, most commonly at the base of the big toe. This is most likely to occur in middle-aged males who may have a family history of gout and who have risk factors for accelerated cardiovascular disease, namely obesity, dyslipidaemia and type II diabetes mellitus .
Women, post-menopause become susceptible to gout, indicating a protective role of female hormones, the mechanism likely related to oestrogen acting on urate renal tubular transport . Male gout sufferers are more likely to drink alcohol, especially beer, and partake of foods rich in purines, such as shell-fish and white bait .
Fructose found in soft drinks that are consumed in increasing quantities, has contributed to rising average plasma urate concentrations and the risk of gout . Plasma urate concentrations between attacks of gout are elevated above the most commonly reported upper limit of the normal range of 0.42 mmol/l. This concentration is also the approximate solubility limit of urate although there are factors other than concentration that influence the solubility of urate such as protein binding . The distribution of plasma urate concentrations in the population is approximately normal but with a skew to the right or higher concentrations . Overall, the risk of attacks of acute gout increases with increasing plasma concentration of urate [4,19].
The diagnosis of gout is made by examination of synovial fluid from an affected joint using polarized light microscopy. Identification of the characteristic needle-shaped, negatively birefringent crystals engulfed by phagocytes makes the diagnosis absolute (Figure 4). In order to prevent recurrence of attacks that are extremely painful, disruptive for significant periods of time e.g., weeks, and diminish the quality of life of the sufferer substantially, the plasma urate must be lowered and maintained below a target concentration by the use of regular ULT. If this is accomplished, recurrence of attacks is extremely unlikely and the long-term complications of uncontrolled gout of joint damage and deformity along with organ damage, especially chronic renal impairment, will not occur. With widespread and inexpensive availability of allopurinol and with other ULT options, the investigation and understanding of the increasing prevalence and, therefore, significant personal and societal impact of gout is an increasingly important imperative [20–22].
Appearance of monosodium urate crystals under polarizing light microscopy
STRUCTURE AND FUNCTIONS OF XANTHINE OXIDOREDUCTASE
XOR is found throughout the body including the blood stream but is most abundant in liver and intestine . As discussed above, inhibition of XOR is the preferred mechanism of lowering plasma urate and thereby reducing the risk of recurrence of gout.
XOR catalyses the oxidation of hypoxanthine to xanthine and thence to uric acid (Figure 1). Other substrates include 1-methylxanthine (1-MX), which can be used as a probe for the activity of XOR in vivo. Caffeine is metabolized to 1-MX . Using the urinary molar ratios of substrate, 1-MX and product, 1-methyluric acid (1-MU) according to the relationship 1-MU/(1-MX+1-MU), variations of up to 10-fold in enzyme activity have been observed. Distributions have generally been unimodal and ‘normal’ but some studies have observed bimodal distributions. Poor metabolizers of allopurinol in the range 4–11% of these populations have been identified [25,26].
A Nigerian population study revealed about 8% with higher XOR activity than average . Functional variations in XOR activity due to a range of single nucleotide polymorphisms (SNPs) in the XOR gene have been identified . The responsible gene is located on chromosome 2. For example, SNPs Ile703Val and His1221Arg were associated with double XOR activity . Levels of activity due to genetic variants as well as possibly external influences such as age, gender, smoking and diet may influence concentrations of urate but also other products of XOR activity involved in inflammation. Carcassi et al. (1969) found evidence of increased XOR enzyme activity in gouty subjects as compared with controls using liver biopsy specimens supporting the possibility of up-regulating the production of enzyme .
The molecular weight of XOR is 290 kDa and it is homodimeric, comprising two identical and independent sub-units (Figure 5). These contain two iron–sulfur centres (2Fe–2S) at the N-terminal end, a flavin adenine dinucleotide (FAD) domain centrally and a molybdopterin co-factor (mo-co) binding site at the C-terminus. The oxidative hydroxylation of substrates including xanthine and hypoxanthine occurs at the molybdenum site and reducing equivalents thus introduced are transferred via the two Fe–S centres to the FAD domain of the enzyme, where NAD+ or O2 reduction occurs (Figure 5).
Structural components of the XOR enzyme
One form of xanthine oxidase (XO) is produced by enzymatic oxidation of thiol groups via a two-step process. Firstly, oxidation of the sulfhydryl elements on the two cysteines results in a disulfide bond. The slower, second step involves oxidation of two more cysteine residues. This step brings the molybdenum and FAD domains of each of the respective enzyme sub-units together. The oxidation of the four cysteines produces XO that can be converted back to XDH (Figure 5) [31,32]. Thus there is bi-directional inter-conversion between XDH and XO.
Another form of XO can be formed in an irreversible reaction. Under hypoxic conditions, XO may be produced by calcium-dependent, proteolytic attack on the intermediate, single disulfide-bond product of the rapid oxidation of XOR (Figure 5).
Enzymatic studies on the mechanism of action of XOR are usually conducted with commercial XO from cow's milk. A complication is that salicylic acid is added to commercial XO because salicylic acid stabilizes the enzyme . The extent to which the added salicylic acid affects the molecular mechanism of action of XO is unknown.
Aldehyde oxidase (AO) is an enzyme that utilizes oxygen to oxidize aldehydes with the production of a carboxylic acid and hydrogen peroxide. AO is closely related to XOR and an important enzyme in the metabolism of allopurinol. Like XOR, AO is also a molybdenum-containing enzyme very similar to XOR in its protein structure. However, the substrate specificities of XOR and AO differ at both the FAD and molybdenum centres, and AO catalyses oxidation of aldehydes better than XOR whereas XOR catalyses the oxidation of purines, xanthine and hypoxanthine more efficiently . AO also oxidizes many drugs including several containing a heterocyclic structure, including allopurinol.
ULT INHIBITION OF XOR
Allopurinol and oxypurinol
Allopurinol is an analogue of hypoxanthine and was thought initially to be a simple competitor for xanthine at the molybdenum centre of XOR with a Ki of 1.9×10−7 M . However, in vitro, the binding is much stronger due to a time-dependent process that leads to oxypurinol, the active oxidation product of allopurinol, being bound slowly to the reduced molybdenum site. Reduced molybdenum is only transiently present during the oxidation of oxypurinol. The molybdenum–oxypurinol complex releases oxypurinol with a half-life of 300 min in the presence of oxygen . Thus, allopurinol has been described as a ‘suicide inhibitor’ of XOR to reflect the tight binding and slow release of oxypurinol. However, this finding may have little relevance to the mechanism of action of these drugs as a crossover study has shown that allopurinol and preformed oxypurinol have very similar hypouricaemic actions in patients .
Allopurinol has not been as effective in reducing or inhibiting vascular inflammation as anticipated given its effectiveness in inhibiting the formation of uric acid and expected inhibition of the production reactive oxygen species (ROS) via XOR (see below). There is a further factor that limits the ability of allopurinol to reduce vascular inflammation. Glycosaminoglycan (GAG)-bound XOR is relatively resistant to inhibition by oxypurinol due to limited access to the enzyme. The binding affinity (Ki) values of oxypurinol for free XOR and GAG-bound XOR, respectively, were 230 and 405 nM (Table 1) . This discrepancy could even be greater in conditions of inflammation and/or hypoxia where there is more immobilization of XOR by endothelial GAG (Figure 6). Further, complete inhibition of cell-endothelial bound XOR is not achieved by oxypurinol at concentrations (200 μM) well above those achievable clinically (100 μM) . These data could indicate a lessened effect of allopurinol, not only on blocking urate production but also production or ROS, during inflammatory flares, due to the diminished ability to inhibit GAG and endothelial-bound XOR.
|Form of XO||Febuxostat||Allopurinol||Febuxostat||Oxypurinol|
|Soluble XOR||1.8 nM||2.9 μM||0.96||230|
|Glycosaminoglycan bound-XO||9.4 nM||64 μM||0.92||405|
|Form of XO||Febuxostat||Allopurinol||Febuxostat||Oxypurinol|
|Soluble XOR||1.8 nM||2.9 μM||0.96||230|
|Glycosaminoglycan bound-XO||9.4 nM||64 μM||0.92||405|
Immobilization of endothelial XO by GAG
The Ki values of febuxostat for free and GAG-immobilized XOR are 0.96 and 0.92 nM, respectively, revealing equivalent inhibition of both forms of the enzyme (Table 1) . Malik et al.  also demonstrated that complete inhibition of endothelial-bound XOR is achieved by 25 nM febuxostat, a concentration that is attainable clinically.
The mode of inhibition of XOR differs markedly from allopurinol. Febuxostat displays mixed competitive and non-competitive inhibition of XDH and XO and the febuxostat–XOR complex is independent of the redox-state of the enzyme at the molybdenum site [39,40]. Essentially, the drug blocks access to the molybdenum site through electrostatic chemical reactions in the channel approaching the active site and no ROS are produced. Finally, febuxostat is a selective inhibitor for XOR having no effect at concentrations up to 100 μM on other key enzymes of purine and pyrimidine metabolism such as hypoxanthine–guanine phosphoribosyltransferase . At the same time, febuxostat is considerably more potent than allopurinol with a Ki of 0.12 nM versus 700 nM for allopurinol at pH 8.2, a ratio of over 5000-fold . These contrasts in the mechanism of inhibition of ULT by allopurinol and febuxostat are of considerable interest but the clinical consequences, if any, remain to be elucidated.
PHARMACOKINETICS AND METABOLISM OF XOR INHIBITORS
The oral availability of allopurinol is almost complete but it is approximately 80% converted to the active metabolite, oxypurinol . With the higher molecular mass of oxypurinol, approximately 90 mg oxypurinol is formed from 100 mg allopurinol . The oxidation of allopurinol to oxypurinol may be catalysed by XDH and/or AO (Figure 5). As AO is not inhibited by allopurinol or oxypurinol, it is most likely that AO is largely responsible for the oxidation of allopurinol to oxypurinol .
Interestingly, mutation of either of the two key amino acid residues for purine substrates in the active site of XOR transforms the substrate specificity of XOR to one similar to that of AO, and the transformed XOR is no longer subject to time-dependent inhibition by allopurinol .
Oxypurinol has a long half-life of elimination of around 23 h allowing effective therapy with once daily dosing . This long half-life means that concentration effect relationships can be examined easily as the peak to trough concentration range over a dosing interval is not excessive and average steady-state concentrations a good representation of exposure of tissues to the drug. There is a good correlation between allopurinol daily dose and steady-state concentrations of oxypurinol, although there is considerable intersubject variability (Figure 7) .
The linear relationship between allopurinol dose and plasma oxypurinol concentrations
Clearance of oxypurinol is largely by the kidneys via glomerular filtration. There is active reabsorption across the renal tubule, via a number of transporter proteins, but predominantly via URAT1 (Figure 3) . Oxypurinol does not compete for reuptake with urate and, therefore, it is not uricosuric  but its reuptake is inhibited by probenecid and other uricosuric drugs resulting in increased excretion and lower plasma concentrations of oxypurinol [46,47]. Despite the lower plasma concentrations of oxypurinol, uricosuric drugs increase the efficacy of allopurinol (see Pharmacodynamics below).
Most of the variability in pharmacokinetics is linked to renal function variability, and clearance of oxypurinol is lower in those with reduced glomerular filtration [48–50]. However as recent work has shown, prediction of the appropriate dose to achieve a target plasma urate is less affected by renal function than body weight and diuretic use; the lack of effect of renal function on dose prediction being due to the common renal tubular transport mechanisms for oxypurinol and urate [50,51]. Thus, urate is retained by decreasing renal function with consequent greater retention. Decreasing renal function also leads to greater accumulation of oxypurinol but this is required to reduce the otherwise higher plasma concentrations of urate .
Febuxostat pharmacokinetics have been elucidated mainly in healthy subjects with fewer comprehensive pharmacokinetic studies in gout patients . The drug is a weak acid (pKa 3.3) whose unionized form is highly lipid soluble and, as such, is expected to have good oral availability, be highly protein bound and have a relatively small volume of distribution. During chronic dosing the oral availability is around 85% .
The range of concentrations during a dosing interval from peak to the 24 h trough is very large, of the order of 100-fold , in obvious contrast with oxypurinol. Despite this fluctuation, febuxostat is administered once daily. The plasma concentrations of febuxostat are best described by a two-compartment model. The initial and terminal half-lives are approximately 2 and 9 h, respectively. The apparent oral clearance (CL/F) is about 11 l/h . Clearance is hepatic by oxidation and acyl-glucuronidation . The apparent volume of distribution is small to moderate (50 litres), as predicted. The pharmacokinetics are unaffected in mild to moderate hepatic failure . The influence of renal failure on the clearance is contentious in renal impairment with either no consistent change in plasma concentrations [55,56] or decreased clearance being reported .
PHARMACODYNAMICS OF XOR INHIBITORS
The reduction in plasma urate concentrations during allopurinol treatment is related to several factors. Not surprisingly, these include the allopurinol and plasma oxypurinol concentrations . Higher baseline, pre-treatment urate concentrations require larger doses of allopurinol for reduction in plasma urate concentrations to target levels . A resistant plasma level of urate also affects the quantitative response to allopurinol. For gout patients, the resistant level is about 0.2 mmol/l, but may be lower in healthy subjects.
Despite the hypouricaemic efficacy of allopurinol, it often does not reduce plasma urate to target levels. In efforts to understand why success rates for eliminating recurrent acute gout and tophaceous gout were poor, it was discovered by numerous researchers that doses of allopurinol were too low to achieve the necessary low concentrations of urate [59–61].
The widespread use of insufficient dose rates was the result of effective education following the influential work of Hande et al. . These investigators were intent on reducing the risk of allopurinol hypersensitivity syndrome (AHS). These very serious reactions occur in about 0.1% of patients exposed to the drug and have mortality rates around 20–30% . Hande et al. found that plasma oxypurinol concentrations were raised in patients with renal impairment and, in their patient cohort, serious hypersensitivity reactions appeared to be related to excessive dosing for the degree of renal impairment and linked high concentrations of plasma oxypurinol.
The work of Hande et al. led to authoritative guideline advice to limit dosage in patients with renal impairment because of the potential risk of AHS [59,62]. However, the risk conferred by concomitant renal impairment and high concentrations of oxypurinol is still unresolved [63–65]. Whether commencing allopurinol at lower doses in individuals with renal impairment and increasing the dose slowly can mitigate the risk of AHS is under active study as is the influence of renal impairment per se . However, it is now very clear, if the Hande guidelines are followed, thereby restricting the doses of allopurinol in patients with renal impairment, as is still commonly the case in practice, a considerable proportion of these patients do not achieve a sufficiently low plasma urate concentration to control their gout . Also, it is now well proven that maintaining plasma urate concentrations consistently below 0.30 mmol/l in the patients with tophaceous deposits of urate, or below 0.36 mmol/l in those with recurrent acute gout attacks, will eliminate gout in virtually all patients [2,59,67].
Febuxostat is clearly an effective hypouricaemic agent. The maximal effect of febuxostat is about an 80% reduction in plasma urate in healthy subjects whereas the ED50, the dose of febuxostat that reduces urate by 50% of the maximum, is about 30–40 mg [32,66]. Doses in excess of 120 mg do not provide significant additional reduction in plasma urate in healthy subjects [68,69].
In gouty patients, the percentage reductions of plasma urate achieved are somewhat lower than those seen in healthy subjects. The maximal effect of febuxostat is about a 66% reduction in plasma urate  whereas the ED50 (30–40 mg daily) is similar to that seen in healthy subjects [71,72]. The smaller percentage effect in patients with gout correlates with the higher baseline concentrations of urate in these patients  and is seen with allopurinol, as noted above. As is the case with allopurinol, the hypouricaemic effect of febuxostat does not appear to be influenced significantly by decreased renal function [55–57].
INSUFFICIENT ULT DOSE RATES
Resistance of XOR to inhibition?
In clinical research studies designed to understand apparent failures of allopurinol therapy in gout patients, studies of dose, plasma concentration and response rates for this drug have been conducted in volunteers and patients. In order to assess the activity of XDH in vivo, xenobiotic substrates for XDH have been administered, for example and as noted previously, 1-MX. The ratio of the product 1-methylurate (1-MU) to 1-MX in urine was used as a measure of the inhibition of XDH. Different methods of delivering 1-MX have been investigated including direct administration intravenously and following administration of oral theophylline or caffeine, both metabolized to 1-MX [74–76]. The relationship between plasma oxypurinol concentrations and XDH inhibition as estimated by the 1-MU/1-MX ratio when volunteers were dosed to steady-state with 50, 100, 300 and 600 mg allopurinol/day each for a week was hyperbolic and the inhibition was rapidly reversible upon cessation of the allopurinol (Figure 8). The 50% and 90% effective inhibitory oxypurinol concentrations were 1.4±0.5 and 4.1±2.0 mg/l, respectively , substantially less than the purported therapeutic concentration range for oxypurinol in gout patients of 30–100 mmol/l (5–15 mg/l) .
A hyperbolic relationship between 1-MU/1-MX ratio measured in urine and plasma oxypurinol concentrations
The relationship between allopurinol daily dose, plasma oxypurinol concentrations and inhibition of XDH as reflected in plasma urate concentrations has been described in detail in work undertaken in healthy volunteers . They were dosed with allopurinol 50, 100, 300, 600 and 900 mg/day for 1 week at each dose, these taken in random order. There was a linear relationship between dose and plasma oxypurinol concentrations (Figure 7). This result was surprising at the time as the oxidation of allopurinol to oxypurinol was thought to be catalysed by XOR exclusively, but the enzyme was also inhibited by both allopurinol and oxypurinol. The expectation was that there would be a less than proportional increase in plasma oxypurinol concentrations with increasing allopurinol dose. However, from studies on the very few individuals documented to have neither, either or both of functioning XOR and the closely related AO enzyme, Reiter et al. and Ichida et al. determined that the metabolism of allopurinol to oxypurinol was catalysed at least in part by AO [78,79].
A classical inhibition, sigmoid EMAX model was proposed for this dose, concentration and response data. An exponent that reflected the steepness of the curve provided satisfactory fits to the data and an IC50 in these normal volunteers was estimated to be an oxypurinol concentration of 37 ± 8 μmol (5.6 mg/l), somewhat lower than concentrations observed clinically . Studies in patients were less clear but a putative therapeutic range for plasma oxypurinol of 5–15 mg/l was proposed by Emmerson et al. .
A new insight into dosage requirements was obtained recently when an empirical model constructed from allopurinol dose, plasma urate concentrations and renal function data from gout patients followed longitudinally was found to provide good predictions of doses required to reach target plasma urate . There was close agreement (Figure 9) between the observed plasma urate concentrations and the concentrations of plasma urate predicted from the modified inhibition EMAX model:
where D is the dose of allopurinol, ID50 is the dose of allopurinol that reduces the inhibitable urate (UP−UR) by 50%; UP, pre-treatment plasma urate concentration; UR, resistant urate concentration and UT, plasma urate concentration during treatment with allopurinol.
The relationship between observed plasma urate concentrations and plasma concentrations of urate predicted from a modified EMAX model
Surprisingly, renal function was not influential in the model but baseline plasma urate concentrations before commencing ULT were. The model also incorporated an uninhibitable, or ‘resistant’ urate concentration which could not be lowered further with increased allopurinol dosage. These observations were supported by Wright et al.  who constructed an allopurinol pharmacokinetics-pharmacodynamics (PK-PD) model. They found that renal function had a minor influence whereas body size and concomitant diuretic use were significant co-variates . In retrospect, this may not appear to be surprising, but the observation had not been made previously. The most important observation from this work was confirmation that many patients need substantially higher doses and concentrations of oxypurinol than suggested from the normal volunteer studies to bring plasma urate concentrations to below a target value associated with prevention of attacks of gout (Table 2). And, as noted previously, inter alia, emerging studies indicated that the relationship between plasma oxypurinol concentrations and risk of AHS was not close as had been suggested by the work of Hande et al. [62,65].
|Pre-treatment plasma urate (mmol/l)||Predicted allopurinol dose (mg/day) to achieve EULAR target (0.36 mmol/l)||Predicted allopurinol dose (mg/day) to achieve BSR target (0.30 mmol/l)|
|Pre-treatment plasma urate (mmol/l)||Predicted allopurinol dose (mg/day) to achieve EULAR target (0.36 mmol/l)||Predicted allopurinol dose (mg/day) to achieve BSR target (0.30 mmol/l)|
ADHERENCE TO ULT
The major problem in the therapy of gout is insufficient adherence to pharmacotherapy to enable plasma urate concentrations to be maintained below at least 0.36 mmol/l [80–82]. The condition can be extremely sensitive to fluctuations in plasma urate concentrations. Acute attacks occur during fluctuations of plasma urate either when they are rising or falling. A patient who forgets to take the drug for a few days may suffer an acute attack as a result and the attack may also be exacerbated as a result of recommencing the drug. Many patients are not appraised well enough of this risk and often, in consequence, abandon the therapy as ineffective . Establishment of therapy with allopurinol or any ULT is a risky time for exacerbating the rate of attacks. Patients need to understand this and clinicians need to introduce these drugs slowly, start with lower doses in those with renal impairment, increase the dose at two to four weekly intervals until target urate concentrations are reached and provide pharmacotherapy prophylaxis for acute attacks with anti-inflammatory drugs or colchicine for the first 6 months of therapy .
The basis for the susceptibility to acute attacks during increases or decreases in plasma urate concentrations especially when instituting ULT is unknown but speculated to be related to mobilization of tissue urate. It seems unlikely that this effect would relate directly to aspects of the interaction of ULT with XOR.
XOR, INFLAMMATION AND CARDIOVASCULAR RISK
ULT reduces the risk of attacks of gout and thereby inflammatory reactions. Inflammation is an increasingly considered putative contributor to the pathophysiology of chronic processes underpinning such conditions as ischaemic cardiovascular disorders and even obesity. ROS are a significant component of inflammatory processes and, therefore, the role of XOR is an important consideration, not only because of its contribution to ROS production but also because uric acid itself may have anti- or pro-oxidant roles. Importantly, the true clinical significance of these processes and biochemical phenomena remain to be elucidated.
The inflammatory reaction of acute gout in patients actually taking ULT is well known, either during institution of therapy, change in dosage or because the dose has been insufficient to reduce plasma urate concentrations to a safe level. An unanswered question is whether the intensity of an inflammatory reaction differs in those taking allopurinol or febuxostat, or uricosuric agents, given the mechanistic contrasts between these drugs with respect to ROS production, although studies from Ogino et al. and George et al. suggest less effect of uricosuric drugs than allopurinol on endothelial function despite similar reductions in urate concentrations and improved endothelial function with allopurinol that did not correlate with urate reduction [84,85].
Gout attacks are highly inflammatory due to activation of the ‘inflammasome’ in response to urate crystals . The inflammasome describes the localized and self-limiting reaction that is due to activation of the innate immune response with ultimate production of interleukin-1β. There are recognized genetic traits that predispose individuals to activate the inflammasome, notably genetic variants of the NOD-like receptor pyrin containing 3 (NLRP3) component . The contributions of XOR and the effects of drugs that reduce urate concentrations by different mechanisms and with varying effects on ROS production upon this complex process are by no means understood.
Production of ROS, especially the highly reactive superoxide free radical as well as hydroxyl radical and hydrogen peroxide, by XOR in its various forms, viz XDH and XO, may contribute to inflammation as noted. The XOR gene is up-regulated in inflammatory states driven by a range of pro-inflammatory cytokines including TNF, IL-1 and IL-6 . Irreversible formation of XO due to calcium-dependent, proteolytic attack on XOR occurs in the presence of pro-inflammatory mediators and areas of low oxygen tension typical of some inflammatory sites . XO requires molecular oxygen to oxidize the purines hypoxanthine and xanthine in the process producing superoxide free radicals. Additionally, in inflammatory states, there is an enhanced degradation of energy-rich adenosine phosphates with increased formation of hypoxanthine, xanthine and thus plasma urate. It has been postulated that hyperuricaemia can then act as a compensatory mechanism, or antioxidant, to protect the body from the flux of pro-oxidants associated with the inflammatory state (see below). However, XOR may contribute to inflammation further by reducing the nitrite ion (NO2−) to nitric oxide (•NO) in low oxygen, low pH conditions typical of inflammation . Indeed, administering the nitrite ion has been examined as a therapeutic source of •NO in inflammatory states to counteract the negative vascular consequences associated with inflammation. As Kelley points out, this presents a dilemma–on the one hand inhibition of XOR might be reducing ROS production but on the other also the production of •NO . Considerably more work is needed before accepting that this potentially deleterious effect of XOR inhibition has relevance for inflammatory states in humans however .
Given the concentrations of urate seen in man, a physiological protective role against oxidative stress has been proposed. Some studies link low plasma urate concentrations to an increased risk of neurodegenerative disorders most notably Parkinson's Disease, the suggestion being that there is a loss of protection of critical neural tissue against oxidative stress. Observational studies have not linked prolonged low urate concentrations as a result of ULT to other chronic degenerative conditions however. The role of uric acid in pathophysiology is controversial because it acts not only as antioxidant and free radical scavenger but also a pro-oxidant [32,90]. Although, it is certain that XOR is important for endothelial function as noted, there are other important mechanisms involved such as ROS production by NADPH oxidase and NO synthase, and human clinical studies that can control for these factors and have sufficient power to show the effects of ULT inhibition are needed.
Reperfusion following ischaemia is an important process that contributes to some forms of inflammation and is particularly relevant to transplantation and various cardiovascular conditions. XOR appears to be central to this process that is characterized by production of free radicals and ROS largely by XO. The main ROS produced are superoxide (O2−) and hydrogen peroxide and largely from the XO species of the enzyme. However, recent work from Lee et al.  has established that XDH can also generate significant amounts of superoxide in the presence of xanthine and that NAD+ and NADH inhibited this process in a dose-dependent manner. Tissue hypoxia is characterized by low levels of NAD+ and NADH and under these conditions XDH generates increased amounts of O2− . Thus, XOR does not need to be in the XO state to contribute to ROS production in tissue ischaemic situations. Upon reperfusion of organs, for example after thrombosis removal or dissolution post myocardial infarction, there is a flush of ROS that put re-perfused and transplanted organs at serious risk of tissue damage .
Calcium-dependent proteases are induced by ischaemia and as noted previously, irreversibly convert XDH to XO. Allopurinol has been investigated as an inhibitor of ROS production–potentially relevant in reperfusion injury pathologies seen also in heart failure and organ preservation for transplantation . Resupply of molecular oxygen on reperfusion is now the source of ROS. Although there is considerable soluble XOR circulating in the blood, as noted above, this enzyme is also located on the vascular endothelium bound via sulfated glycosaminoglycan interactions . The placement of XDH/XO in this location puts the vascular tree and organs at great risk from free radical attack. XDH/XO in this location is less sensitive to allopurinol than is circulating enzyme but this differential effect does not apply to febuxostat. Release of XOR in large amounts during tissue injury increases the risk of ROS-driven damage to other organs not originally damaged, another impetus to examining the efficacy of XOR inhibition in these conditions .
Gout is commonly associated with insulin resistance and type II diabetes mellitus, obesity, accelerated cardiovascular and peripheral vascular disease and renal impairment [95,96]. Importantly, inflammation and involvement of ROS are features of these conditions as they are of gout attacks. There has long been interest in a potential role for ULT in ischaemic cardiovascular disorders [95–98]. However, human studies of the efficacy of allopurinol in improving health outcomes in patients with acute and chronic cardiac, cerebral and peripheral vascular conditions have produced mixed results . Indeed, large registration and extension studies suggested an increased risk for cardiovascular conditions in patients initiated on febuxostat . However, there was no benefit for clinical outcomes in a 24-week, placebo-controlled trial of oxypurinol in 405 patients with cardiac failure due to systolic dysfunction . Recently, Kim et al. , using a US-based insurance claims data-base in an observational study, compared over 24,000 pairs of hyperuricaemic gout patients, well-matched in all respects except for initiation of treatment with XOR inhibitors. A standard, composite cardiovascular outcome was used but no contrast in this outcome was discerned.
The progression of chronic kidney disease, even in individuals with normal plasma urate concentrations and without gout, may be reduced with allopurinol therapy and large, controlled trials of ULT intervention studies are underway to test this promising approach for this prevalent yet underappreciated problem [101,102]. The rationale for maintaining plasma urate concentrations below at least 0.36 mmol/l in gout patients would be significantly strengthened with the availability of strong evidence of the beneficial effects of allopurinol or other ULT on renal function.
Our understanding of the cellular and molecular mechanisms of XDH inhibitors has increased considerably. Their effects beyond reducing plasma urate concentrations and thereby reducing the risk of recurrent attacks of acute gout and long-term damage from chronic gout revolve around their ability to modulate inflammation, ischaemia reperfusion processes and conditions where ROS may play a pathophysiological role. We are only at the beginning of understanding how ULT may contribute to the treatment and progress of these complex, often co-existent chronic conditions of gout and the clinical components of the metabolic syndrome, namely type II diabetes mellitus, accelerated cardio, cerebrovascular and peripheral vascular disease and obesity. Much further work examining the wide manifestations of the clinical conditions of gout and associated, comorbid conditions, their pathophysiology and their responses to ULT with different suites of pharmacological actions now constitutes an important research agenda in the quest to improve human health .
Richard Day's research work is supported by National Health & Medical Research Council Program [grant number APP 1054146].
allopurinol hypersensitivity syndrome
apparent oral clearance
dose at which 50% of maximum effect occurs
flavin adenine dinucleotide
dose at which 50% of maximum inhibition occurs
nicotinamide adenine dinucleotide
reactive oxygen species
single nucleotide polymorphism
urate lowering therapy