Hyperuricemia and chronic kidney disease: an enigma yet to be solved

Abstract

The role of uric acid (UA) on the pathogenesis and progression of chronic kidney disease (CKD) remains controversial. Experimental and clinical studies indicate that UA is associated with several risk factors of CKD including diabetes, hypertension, oxidative stress, and inflammation and hyperuricemia could be considered as a common dominator linking CKD and cardiovascular disease. Notably, the impact of serum UA levels on the survival of CKD, dialysis patients, and renal transplant recipients is also a matter of debate, as there are conflicting results from clinical studies. At present, there is no definite data whether UA is causal, compensatory, coincidental or it is only an epiphenomenon in these patients. In this article, we attempt to review and elucidate the dark side of this old molecule in CKD and renal transplantation.

• Chronic kidney disease
• dialysis
• hyperuricemia
• mortality
• renal transplantation

Introduction

Uric acid (UA) is the end product of purine nucleotides metabolic breakdown. Xanthine oxidase (XO), which is a form of xanthine oxidoreductase, catalyzes the final oxidation of hypoxanthine and xanthine to UA.1

The kidneys eliminate approximately 70% of the UA load while 30% is eliminated by the gastrointestinal tract.1 UA is filtered by the glomeruli and the amount of UA excretion depends on its reabsorption and secretion mainly at the proximal tubule. Above 5–10% of the filtered UA is excreted in the urine, while 90% is reabsorbed.1

Hyperuricemia (HUA) is the result of the balance between UA production and excretion. In more than 90% of the cases, HUA is due to impaired renal excretion of UA.1 The prevalence of HUA is increased as the glomerular filtration rate (GFR) declines, ranging from 40% to 70% in chronic kidney disease (CKD) stages 1–5.2,3 Decreased UA filtration due to reduced GFR, impaired UA tubular secretion caused by tubulointerstitial disease, and use of diuretics and cyclosporine (CsA) contribute to increased prevalence of HUA in CKD patients.4 In terms of ethnic and gender differences, Asian patients with end-stage renal disease (ESRD) present higher rates of HUA5, and men have a higher risk of HUA and gout in the general population.6

Data from several studies have associated HUA with increased incidence and/or progression of CKD. Diabetes is one of the leading causes of CKD and HUA has been associated with the pathogenesis of the microvascular complications of diabetes.7 Findings from animal studies revealed a causal role of persistent HUA in the pathogenesis and development of hypertension8–10 which, in turn, is implicated in the progression of CKD and cardiovascular (CV) mortality.

UA has been associated with oxidative stress, inflammation, and endothelial dysfunction11 (Table 1). The impact of UA in all-cause and CV mortality of dialysis patients' remains controversial.12,13 Controversial also remains the role of UA in renal transplantation.7

Table 1. Risk factors for CKD progression associated with hyperuricemia.

• Diabetes

• Metabolic syndrome

• Hypertension

• Oxidative stress

• Inflammation

• Endothelial dysfunction

• CsA nephrotoxicity

CKD, chronic kidney disease; CsA, cyclosporine A.

Relationship of HUA – chronic kidney disease incidence and progression

Several studies with conflicting results in some instance have investigated the relationship between HUA and CKD. Some studies reported an independent positive association of HUA with incident CKD,14–19 progression of CKD, and ESRD.2,20–27 Interestingly, a positive association of HUA, IgA nephropathy, and lupus nephritis has also been reported.28,29 The majority of all these studies has been recently described elsewhere.30,31

On the contrary, this positive association between HUA and CKD progression has not been supported by other studies. A cohort study including 840 participants with CKD stages 3 and 4 reported that HUA was not an independent risk factor for CKD progression, but was correlated with increased risk for all-cause mortality and CV mortality.3 In accordance, a prospective 7 years follow-up study including 227 Caucasian patients with non-diabetic CKD, revealed that HUA did not predict CKD progression.32

These contradictory results may be explained by the differences in the case mix and the follow-up period as well as gender differences (higher incidence of HUA in men), and different thresholds of HUA in these studies. Thus, some authorities support that UA is not a real risk factor for CKD. In addition, given the decreased clearance of UA, even in the early stages of CKD, HUA might be seen as an epiphenomenon and not a real cause of CKD. In this respect, HUA may just reflect the severity of CKD (advanced stage).

Data from animal models with HUA caused by the administration of low doses of oxonic acid (uricase inhibitor), suggested a causal role for UA in the progression of CKD.8 A renal vascular disease characterized by renal cortical vasoconstriction, afferent arteriolopathy, and glomerular hypertension has been observed in this animal model.8,9,33 In addition, light microscopy revealed tubular dilatation, renal cortex atrophy, and to some extent glomerulosclerosis.33 Notably, activation of renin–angiotensin system (RAS), enhanced oxidative stress, alterations in endothelin-1 expression, and increased expression of cyclooxygenase-2 (COX2) have been proposed to explain endothelial and vascular abnormalities observed in animals with HUA.7 In support, administration of febuxostat which is a XO inhibitor alleviated systemic and glomerular hypertension and blockage of RAS and prevented pre-glomerular vasculopathy in animal models with HUA.34,35 Interestingly, a higher grade of proteinuria, increased blood pressure, and higher serum creatinine levels have been observed in rats with CKD (remnant kidney model) and HUA compared with remnant kidney rats without HUA.36 Moreover, renal hypertrophy, glomerulosclerosis, and interstitial fibrosis were more pronounced in hyperuricemic rats.36

Notably, findings from animal models are usually specie specific and may provide some evidence, but do not necessarily translate into the human underlying disease process. So, extrapolation of experimental animal findings to human disease is not always straightforward.

An interesting approach to investigate the role of UA on CKD progression is to evaluate the effect of decreasing UA levels on CKD. Several studies reported a beneficial effect of UA lowering on CKD progression and UA has been proposed as a promising target of therapy in CKD patients.37–43 Recently, Omori et al.37 reported that in animal models with obstructive nephropathy, febuxostat inhibited renal interstitial inflammation and fibrosis suggesting a therapeutic value in slowing CKD progression. In accordance, Sakai et al.42 reported that febuxostat by lowering UA levels in CKD patients may suppress the progression of CKD. Notably, data from a recent meta-analysis supported that allopurinol may slow the progression of CKD.43 However, in another meta-analysis including eight trials, an uncertain effect of allopurinol has been reported.44

Finally, the relative importance of all these associations is yet controversial. One difficulty is to determine whether UA per se is a risk factor for the development and progression of CKD. Given the high incidence of UHA due to decreased clearance of UA in CKD, it could be assumed that the increased levels of UA in CKD present only as an epiphenomenon. In this respect, CKD could be considered the cause of HUA and not the reverse. Moreover, the independence of a risk factor is not necessarily equal to causality and a risk factor may be independent and not causal, when the real causal parameter is not included in the statistical analysis.45 In the same way, a risk factor may be not independent but it could still be causal.45 It could be assumed that HUA is an indirect cause of kidney disease, but the difficulty to clarify the role of UA as an independent risk factor for CKD is mainly due to the complex correlations and interactions between UA and conventional risk factors for CKD. However, the elucidation of the relationship between UA and CKD progression remains an evolving area of research.

Gouty nephropathy and CKD

Findings from early studies in rats provided evidence that HUA in combination with hyperuricosuria may lead to kidney disease through intraluminal deposition of crystals in the collecting duct of the nephron in a manner simulating gouty arthropathy.46–48 UA crystals adhere to the surface of renal epithelial cells and may induce an acute local inflammatory response.49 Interestingly, UA induces phenotypic transition of renal tubular cells suggesting a novel potential mechanism of CKD.50 Apart from the increased risk of developing kidney stones, such effects seem to contribute to a decrease in GFR aggravated by the presence of hypertension.51 In the early 1960s, Talbott and Terplan,52 reported that 10–20% of patients with untreated gout developed ESRD. Moreover, renal biopsies in these patients revealed tubulointerstitial fibrosis, arteriosclerosis, and glomerulosclerosis.52 In subsequent years, the causal role of HUA and gout in kidney disease onset and progression had been the subject of a large debate. Some investigators suggested that the histological findings revealed in renal biopsy specimens in patients with HUA such as tubulointerstitial fibrosis, arteriosclerosis, and glomerulosclerosis were similar with those observed in hypertensive renal disease.45 These non-specific alterations were also present in advanced age or in patients with asymptomatic HUA. Moreover, given that patients with HUA are usually hypertensive, it is difficult to attribute such non-specific lesions to UA or to hypertension. Notably, Nickeleit and Mihatsch53 concluded that renal tophi hardly lead to terminal irreversible renal failure based on autopsy cases with gout. Additionally, gout is correlated with several other risk factors including diabetes, advanced age, and male sex. Thus, it is difficult to dissect the causal effect of UA from the synergic effect of the other risk factors. Therefore, the term gouty nephropathy has been considered as a misnomer by some authorities and, in the past, it was removed from the list of causes of CKD in the classical nephrology textbooks.

HUA, diabetes, and diabetic nephropathy

Diabetic nephropathy is one of the leading causes of ESRD. HUA promotes insulin resistance and concomitant hyperinsulinemia.54,55 Zhang et al.56 reported that HUA contributes to abnormal glucose metabolism by increasing oxidative stress and by inhibiting the function of pancreatic β cells. Kidney tubular function seems to be influenced by insulin metabolic signaling and renal UA clearance decreases with reducing insulin-mediated glucose disposal.57 Obesity predisposes to type 2 diabetes and in obese patients, especially in those with increase high-fructose diet, an increase hepatic production of UA has been reported, suggesting a potential link among obesity, diabetes, and UA.58

Data from animal studies are limited. Kosugi et al.59 reported that treatment of diabetic mice with a allopurinol for 8 weeks, reduced albuminuria, and ameliorated tubulointerstitial injury. The authors concluded that the protective mechanism was the reduction of local inflammation. Kim et al.60 found that lowering UA with allopurinol retarded diabetic nephropathy in animal models with type 2 diabetes, through attenuation of the transforming growth factor-β1 and Smad pathway-induced profibrogenic tubular changes.

Data from clinical studies supported the concept that HUA is a risk factor for diabetic nephropathy. Hovind et al.61 reported that baseline UA levels predicted persistent macro-albuminuria in patients with type 1 diabetes. In accordance, findings from the Coronary Artery Calcification in Type 1 Diabetes Study supported that high serum UA levels predict the development of albuminuria.62 Moreover, the authors reported that every 1 mg/dL increase in serum UA levels may lead to an 80% increased risk for micro- or macroalbuminuria.62 Ficociello et al.63 have found an association between serum UA levels and early GFR loss in 355 participants with type 1 diabetes. In an observational, cross sectional study, high UA serum levels were associated with a decline in GFR in type 1 diabetic patients with normo- or microalbuminuria.64 Similar, serum UA levels were correlated with micro and macro-vascular complications in type 2 diabetic patients.65 Data from a another study, including 1449 type 2 diabetic patients with preserved kidney function and a 5 years follow-up period, reported that the presence of HUA doubled the risk of incident CKD. The authors concluded that HUA is an independent risk factor for the development of CKD in type 2 diabetic patient.66 Additionally, Kim et al.67 reported that increased levels of UA may predict the occurrence of advanced CKD in 512 type 2 diabetics with preserved renal function. Interestingly, Bo et al.68 reported that HUA in type 2 diabetes is correlated with early onset or increased progression of diabetic nephropathy, whereas the opposite occurs with hypouricemia. In another study from Japan including 1213 type 2 diabetic patients, a positive independent association among HUA, coronary heart disease, and renal dysfunction was reported, supporting the hypothesis that HUA is implicated in the pathogenesis of the micro- and macro-vascular diabetic complications.69 In accordance, Tanaka et al.70 reported that HUA was correlated with increased risk of macrovascular disease in diabetic CKD patients.

The impact of lowering UA on the progression of diabetic nephropathy is challenging, but few studies have been conducted in this field. Momeni et al.71 investigated the effect of allopurinol administration in 40 patients with type 2 diabetic nephropathy. The authors reported a decrease in proteinuria after 4 months of treatment.71 However, the small size, the short follow-up period, and the lack of assessment of renal function do not permit safe conclusions. According to the Reduction in Endpoints with the Angiotensin Antagonist Losartan (RENAAL) study, an angiotensin II antagonist (Losartan), apart its effect on lowering blood pressure and proteinuria, has also an UA lowering effect suggesting that the renoprotective effect of losartan may be partially attributed to decreased serum UA levels.72 Moreover, the renoprotective effect of losartan was decreased from 22% to 17% after adjusting for the effect of allopurinol on serum UA levels, suggesting that UA may partially contribute to diabetic nephropathy progression.72

HUA, hypertension, and CKD

Several studies reported that HUA predicts a higher risk of hypertension, independent of other risk factors.73–83 Interestingly, HUA is associated even with prehypertension, especially when microalbuminuria is present.84,85 In the majority of these studies, HUA precedes the onset of hypertension indicating that it is not a result of hypertension. However, Forman et al.86 reported that HUA was not a risk factor for incident hypertension among 750 participants who developed hypertension after the age of 60 years. It is worth noting that HUA is not usually associated with secondary hypertension, impoverishing the hypothesis that HUA is an epiphenomenon of hypertension.87 Interestingly, HUA is not only associated with hypertension but also with its complications including left ventricular hypertrophy.88,89 However, Tsioufis et al.90 did not confirm this finding.

Findings from several studies provided a direct evidence that UA is implicated in the pathogenesis of hypertension.8--10,91 Additionally, the microvascular changes caused by UA remain after correction of blood pressure indicating that UA contributes to microvascular disease independently of hypertension.92 Interestingly, data from in vitro studies reported that UA induces proliferation of vascular smooth-muscle cells, local inflammation, oxidative stress, endothelial dysfunction, and local RAS activation.93

Recently, Sheikhbahaei et al.94 reported a synergistic effect between UA and various components of metabolic syndrome including hypertension and diabetes, on CKD odds. In this respect, it could be assumed that UA contributes to CKD development and progression through conventional CKD risk factors such as diabetes and hypertension, indicating an indirect effect. In support, data from a recent prospective cohort study involving 2601 individuals reported that increased serum UA levels were independently associated with a faster decline of GFR and incident CKD and this association was stronger in hypertensive individuals. The authors concluded that hypertension mediates the correlation between high serum UA levels and CKD.95

Data from experimental and clinical studies reported that lowering UA by using XO inhibitors may lead to a decrease in blood pressure, supporting a direct effect of UA.8,96–98 However, Gois and Souza99 reported that there is insufficient evidence of an antihypertensive effect of allopurinol.

Finally, based on the experimental data, there is evidence of a direct action of UA on hypertension-induced kidney injury through RAS activation, inflammation, endothelial dysfunction, and oxidative stress. Thus, the majority of patients with hypertension and HUA should present CKD, but actually this does not occur. Probably, UA in some instance mediates and contributes to CKD development and progression through hypertension-induced renal injury suggesting an indirect action on the pathogenesis of kidney disease. Another hypothesis is that CKD results in HUA, which, in turn, through hypertension-induced changes and endothelial dysfunction, aggravates the decline in renal function. Nevertheless, all these theories remain to be proven.

HUA, oxidative stress, and inflammation

Oxidative stress is implicated both in CKD and in CV disease (CVD) progression and it is characterized by increased production of free radicals and reactive oxygen species (ROS).100 On one hand, in humans, XO does not permit further oxidation of UA.100 During this process, ROS are generated which play a crucial role in the enhanced vascular oxidative stress observed in both CKD and CVD. Based on the clinical data that HUA may contribute to CKD and CVD progression, some therapeutic strategies are oriented on the inhibition of the increased XO activity. It could be assumed that the potential benefit of allopurinol is not due to its UA lowering effect, but to an antioxidant action attributed to the decreased XO activity.11 Moreover, findings from animal studies supported that UA itself may promote endothelial dysfunction by inhibition of nitric oxide function.11

On the other hand, UA may have a protective role under certain conditions. Nieto et al.101 reported that UA plays a protective role as an antioxidant in atherosclerosis and this is a probable compensatory mechanism in order to counteract oxidative damage attributed to atherosclerosis. In accordance, both in vitro and in vivo studies support that UA is an important antioxidant by acting as a free radical scavenger and a potent chelator of transitional metal ions.102,103 Kurra et al.104 reported that HUA reduced oxidative stress in animal models with renal insufficiency. However, the real determinative factors of UA action on oxidative process have not been elucidated. Lippi et al.105 in an attempt to explain the paradox action of UA reported that in certain situations, the antioxidant compounds may become pro-oxidant, especially when they are at supranormal levels in blood. Thus, it could be postulated that UA concentration determines its double controversial action. Urate at low concentrations may have a beneficial action as an antioxidant; while at higher concentrations, it may promote oxidative stress reflecting a rather maladaptive/compensatory mechanism in order to slow the progression of kidney damage. However, a threshold has not been yet specified. Another hypothesis is that UA may have different biochemical effects under different pathological or physiological conditions and in different cellular environments. Recently, Johnson et al.31 reported that on one hand UA is a pro-oxidant in the intracellular environment by stimulating NADPH oxidases, while outside the cell is a potent antioxidant. On the other hand, it could be hypothesized that XO activity and not UA itself plays a crucial role on oxidative process but all these hypothesis remain to be proven.

Both CKD and CVD are characterized by low-grade inflammation. Inflammation may be considered as a potential underlying mechanism linking HUA, CKD, and CVD. In vascular smooth muscle cells, UA promotes inflammation by activating NF-κB and by enhancing serum C-reactive protein (CRP) expression.106,107 Moreover, HUA was associated with several circulating inflammatory cytokines.108 Interestingly, C reactive protein (CRP) in combination with HUA may interact on the development of albuminuria and concomitant CKD progression on type 2 diabetic patients.109 A significant positive correlation between CRP and serum UA levels was observed in 337 CKD patients.110 In this regard, it could be assumed that HUA is a marker of inflammation associated with both CKD and CVD.

Several studies provided evidence that UA mediates endothelial dysfunction contributing to CKD and CVD progression. Endothelial dysfunction with HUA in the early stages of CKD is associated with micro-inflammation and oxidative stress.111 More attention has been given to the xanthine oxidoreductase system since allopurinol may prevent the progression of atherosclerosis by reducing endothelial dysfunction.112,113 However, data from other studies did not support a direct effect of UA on endothelial dysfunction.114

HUA and CV mortality

According to the Hemodialysis (HEMO) study, CVD is the leading cause of death in dialysis patients.115 Moreover, there is a reciprocal relationship between CVD and CKD, and UA might be considered as a common dominator linking these two clinical conditions.

Similar to the general population, the evidence for a strong correlation between UA and CVD mortality in CKD patients is conflicting. Data from epidemiological studies in CKD patients revealed a direct linear or “J-shaped” correlation with all cause or CV mortality.3,116,117 However, according to other studies, this correlation diminished after adjustment for GFR or proteinuria.118,119 Moreover, Madero et al.3 found that CKD patients with high serum UA concentration had a higher risk for all-cause and CV mortality. Similar, Kanbay et al.120 reported that HUA predicts CV events in advanced CKD stages. In accordance, Liu et al.116 suggested that HUA is a risk factor for all-cause and CV events but not for renal outcomes in stages 3–5 CKD patients. Additionally, HUA independently correlates with CV mortality and this correlation is stronger with decreasing GFR.121 Interestingly, Chung et al.122 revealed that HUA is an independent risk factor for all-cause mortality in CKD patients undergoing percutaneous coronary intervention but not in those without CKD. Notably, allopurinol administration in hypertensive CKD patients prevents CV and all-cause mortality.123

In the hemodialysis (HD) population, the data are more contradictory. Hsu et al.12 have found a “J-shaped” association between UA and all-cause mortality in 146 HD patients. Recently, Lee et al.124 reported that serum UA levels were independently associated with valvular and vascular calcifications in 81 HD patients. Additionally, Antunovic et al.125 reported that HUA may predict all-cause and CV mortality in HD patients. In support, Lobo et al.126 reported a positive association of UA levels and several markers of inflammation suggesting a potential role of HUA in the pathogenesis of atherosclerosis in HD patients. In addition, Cohen et al.127 in a cohort of 259,209 patients from the United States Renal Data System evaluated the incidence of gout and its relationship with mortality. The authors calculated that an episode of gout was independently correlated with a 1.5-fold increase in mortality risk. The authors proposed inflammation and atherosclerosis as the underlying mechanisms of the increased mortality.127

On the contrary, data from a retrospective study reported that low serum UA concentration in HD patients with hypoalbuminemia is a mortality risk factor.128 Another large retrospective study involving 5827 HD patients reported that higher serum UA levels were correlated with lower risk of all-cause and CVD mortality.13 These results might be explained by the reverse epidemiology. Interestingly, from ancient times, it was well known that gout was a disease of the Emperors of the Byzantium suggesting their better social-economic status compared with malnourished people.129 In this respect, HUA may reflect a better nutritional status in HD patients and low UA levels may represent a form of protein energy wasting syndrome.

Contradictory results have been reported also in peritoneal dialysis (PD) patients. Data from two studies revealed an association of HUA with all-cause and CVD mortality.130,131 However, in a recent observational study including 2264 PD patients, only a weak correlation of HUA with all-cause mortality could be detected.132 The lack of large prospective studies in this field does not lead to reliable conclusions.

HUA and renal transplantation

HUA occurs at the early post-transplant period. The use of diuretics and calcineurin inhibitors, reduced GFR, advanced recipient age, obesity, and pre-existing history of HUA are common risk factors.133 CsA is the most common risk factor and conversion of CsA to tacrolimus or everolimus immunosuppression may decrease serum UA levels.134,135

The impact of HUA on graft survival is controversial. Some studies indicate that HUA is only a marker of graft dysfunction,136–139 whereas some other studies suggest that UA is an independent risk factor for graft dysfunction.140–146 Recently, Hart et al.147 reported that HUA correlated with tubular atrophy and interstitial fibrosis in chronic allograft nephropathy. Similar to the experimental findings, HUA in humans may exacerbate CsA nephrotoxicity.148 Additionally, CsA-induced hypertension and endothelial dysfunction may contribute to graft dysfunction.149 Interestingly, an association among HUA, inflammation, graft dysfunction, and CV events has been reported.142

While post-transplant gout attacks should be treated, the treatment of asymptomatic HUA remains controversial.133 Minimization of diuretics and CsA and avoidance of purine-rich foods and alcohol are effective strategies to decrease UA concentrations. Considering the link of UA with diabetes and metabolic syndrome, hypertension and CVD risk, minimizing these risk factors may be proven beneficial for graft function.

Conclusion

After so many years of scientific research, the real significance of this old molecule is widely debated and it remains obscure whether UA is causal, compensatory, coincidental, or it is only an epiphenomenon of CKD. In accordance, nebulous also remains the relationship of UA with CV mortality. As HUA is usually associated with a better nutritional status, whereas hypo-uricemia with malnutrition, it could be speculated that the alterations of UA metabolism in advanced CKD and its reported associations may be just a reflection of different population characteristics, comorbidities, or social-economic status (Figure 1). More prospective randomized and well-designed trials with a large number of participants are needed to clarify the exact relationship among UA, CKD, and CV mortality.

Figure 1. The complex interplay among hyperuricemia, chronic kidney disease (CKD), and cardiovascular (CV) mortality.

Declaration of interest

The authors report that they have no conflicts of interest. The authors alone are responsible for the content and writing of the paper.