Background
Hyperuricemia has been reported to be a risk factor for kidney dysfunction [
1‐
5]. The mechanism underlying the progression of kidney dysfunction by activation of xanthine oxidase (XO) has been reported. Activation of XO increases reactive oxygen species (ROS) production and oxidative stress, as well as uric acid (UA) production, and causes vascular damage and organ dysfunction. Indeed, ROS is known to inactivate nitric oxide production and activates renin-angiotensin system, resulting in endothelial dysfunction and/or tubular injuries [
6‐
8]. Alternatively, UA salts activate inflammasomes, triggering inflammation and leading to the development of tubular injuries [
9‐
11]. Antihyperuricemic XO inhibitors, including febuxostat and allopurinol, are believed to suppress renal dysfunction via oxidative stress reduction and suppression of endothelial dysfunction and tubular injuries. However, some studies reported that hyperuricemia is not significantly related to the progression of kidney dysfunction in patients with advanced Chronic Kidney Disease (CKD) [
12,
13], and that a renoprotective effect via lowering of the serum UA level was not observed in CKD patients [
14]. Thus, the renoprotective effect of XO inhibitors is controversial, and further investigation is required.
We often find that CKD patients treated with antihyperuricemics show improved or maintained estimated glomerular filtration rate (eGFR) along with a reduction in serum UA level, whereas we also found cases showing none of the renoprotective effects. Since various risk factors for kidney dysfunction are involved in CKD progression, these factors could make it difficult to detect febuxostat-induced renoprotection.
In this study, we investigated whether antihyperuricemic therapy using febuxostat exerts renoprotective effects in CKD patients and which patient characteristics influence these renoprotective effects.
Methods
Study design and patients
This was a single-center, retrospective, observational study. Two-hundred thirty-two CKD patients underwent antihyperuricemic therapy using febuxostat at the Department of Nephrology in Shinshu University Hospital between May 2011 and October 2014. Patients were defined as having CKD if eGFR was <60 ml/min/1.73m
2 and/or urine protein was >0.5 mg/day, or both urinary blood and protein were present over 3 months. We excluded patients with acute kidney injuries (AKI), acute nephritis, or rapidly progressive glomerulonephritis from the current study. Fifteen patients discontinued the febuxostat treatment owing to adverse events during the follow-up period, and 19 patients required renal replacement therapy. Follow-up clinical data in 20 patients could not be obtained owing to change in hospital or discontinuation of attendance. A total of 178 cases could be analyzed. No patients died during the follow-up period. Among these 178 patients, 118 did not previously use any antihyperuricemic agents and 60 were switched from allopurinol or benzbromarone to febuxostat. In Japan, a low initial dosage (10 mg/day) is recommended (maximum dose 60 mg/day). Moreover, febuxostat is known to show an increased AUC in the patients with severe kidney dysfunction (GFR < 30 ml/min) [
15], and careful administration is necessary for these patients. For these reasons, febuxostat administration commenced at a dose of 10 mg/day in the antihyperuricemic-naïve group, and 10 or 20 mg/day in the group that was switched from other agents. The respective attending physicians determined dosage adjustments based on serum UA levels and its safety during the study period. All treatments patients received are considered standard care for their condition. Patient characteristics and clinical data were tabulated using medical record information. Changes in eGFR and serum UA levels at 6 months after febuxostat treatment initiation were analyzed statistically.
Statistical analyses
To examine the effect of febuxostat therapy on changes in serum UA levels and eGFR, those parameters were measured at the time of febuxostat prescription, as well as 3 months and 6 months after therapy initiation. The correlation between mean serum uric acid level after treatment (mUA) and the change of eGFR after 6 months (6mΔeGFR) was analyzed using Spearman’s correlation. mUA was calculated as the mean serum UA level at 3 and 6 months. 6mΔeGFR was defined as the difference obtained by subtracting eGFR at treatment initiation from eGFR at 6 months. To investigate the effect of antihyperuricemic therapy for renal function lowering in CKD patients, we performed the correlation and ROC analyses using mUA and 6mΔeGFR data. Based on a calculated optimal cutoff level estimating progression of kidney dysfunction and standard goal of UA lowering therapy, the ΔeGFR was analyzed by dividing patients into mUA < 6.0 mg/dl and mUA ≥ 6.0 mg/dl groups.
The correlation between mUA and 6mΔeGFR was analyzed in all patients or in patient subgroups based on characteristics at the start of the therapy. Each patient’s clinical data were collected from their medical records. The patient characteristics examined were sex, age (< 70 vs. ≥ 70 years), systolic blood pressure (< 130 vs. ≥ 130 mmHg), the presence or absence of an abnormal cholesterol level, the presence or absence of diabetes mellitus, eGFR (< 30 vs. ≥ 30 ml/min/1.73 m2), urine protein level (< 0.5 vs. ≥ 0.5 g/gCre), and antihyperuricemic naive vs. agent switching group. The patients who were using a lipid-lowering drug or showed an abnormal level of LDL or HDL cholesterol (LDL cholesterol ≥140 mg/dl, HDL cholesterol <40 mg/dl) were defined as having symptoms of abnormal cholesterol levels. The patients who were using an antidiabetic or presented above 6.2% of HbA1c (NGSP) in a blood examination were defined as diabetes mellitus.
The subgroups, in which a significant correlation between mUA and 6mΔeGFR was detected, were then divided into the mUA < 6.0 mg/dl group and the mUA ≥ 6.0 mg/dl group to compare ΔeGFR. We also analyzed the other subgroups, which were separated based on numbers of risk factors.
Mann-Whitney U test was used to examine the statistical significance in continuous variables, and Fisher’s exact test was used to examine category variables. Wilcoxon signed-rank test was used to examine the statistical significance in time course changes in continuous variables. SPSS Statistics v.22.0 J (IBM Corp., Armonk, NY, USA) analysis software was used in this study. Differences were considered statistically significant at p < 0.05.
Discussion
Attempts to control serum UA levels using previous antihyperuricemic medicines, including benzbromarone and allopurinol, in patients with kidney dysfunction, have proved difficult. Since benzbromarone is an agent that enhances urinary excretion of UA, its antihyperuricemic effect is attenuated by kidney dysfunction. Another XO inhibitor, allopurinol, is mainly excreted by urine; therefore, its serum concentration and adverse events are increased in the patients with kidney dysfunction, requiring dose reduction from an early phase of CKD and resulting in poor control of serum UA levels. Since febuxostat is excreted via dual excretion pathways, urine 49.1% and fecal 44.9% [
17], febuxostat can exert strong antihyperuricemic effect without serious adverse events even for patients with kidney dysfunction without changing dosage [
18,
19]. Although careful febuxostat administration is necessary for patients with severe kidney dysfunction (GFR < 30 ml/min), its metabolic characteristics are advantageous compared to those of allopurinol. Indeed, this study demonstrates that febuxostat therapy is effective in controlling serum UA levels in patients at any CKD stage. The percentage of patients for whom febuxostat could be administered continuously for 6 months was high, as the incidence of serious adverse events was very low. In the very small number of CKD patients who developed an adverse event, the symptoms improved rapidly after drug discontinuation. These results are consistent with reports stating that febuxostat treatment exerts a more potent antihyperuricemic effect with an equivalent incidence of adverse events compared to allopurinol in patients with kidney dysfunction [
18,
19]. Therefore, these results indicate that febuxostat treatment is useful for lowering serum UA levels in CKD patients.
Many studies have reported that hyperuricemia is a risk factor for the progression of kidney dysfunction in the general population, diabetes mellitus, AKI and CKD patients [
1‐
5]. Therefore, antihyperuricemic therapy is expected to contribute to kidney protection. In this study, we showed that the decline of eGFR was remarkable in CKD patients with poor control of UA level (mUA ≥ 6.0 mg/dl), whereas it was likely to be ameliorated by febuxostat therapy in patients with good control of UA level (mUA < 6.0 mg/dl). In addition, we detected an inverse correlation between mUA and 6mΔeGFR, suggesting that further decreases in mUA after febuxostat administration contribute to suppression of CKD-induced eGFR decline. These findings support the hypothesis that lowering serum UA level is renoprotective. These results are consistent with previous studies reporting that antihyperuricemic therapy in patients with CKD ameliorates eGFR reduction [
20,
21]. However, the renoprotective effects of antihyperuricemic therapy in CKD patients are controversial in several small-scale randomized controlled trials (RCTs) and various follow-up studies involving CKD patients. Two placebo-controlled small-scale RCTs in which allopurinol or febuxostat were used reported that CKD-induced eGFR decline was suppressed more in the antihyperuricemic group than in the placebo group [
20,
21]. Levy DG et al. performed antihyperuricemic therapy for 3 years using allopurinol, febuxostat, or probenecid for patients with hyperuricemia, and found that renal dysfunction could be suppressed in patients with controlled UA levels at <6 mg/dL [
16]. In our study, we also found that deterioration of kidney function could be suppressed in patients with UA levels controlled at lower than 6 mg/dl, supporting the above reports. In contrast, a RCT involving 40 patients with IgA nephropathy and CKD stage 1–3 reported no renoprotective effects of a 6-month allopurinol therapy course [
14]. The lack of consensus among many clinical studies regarding a renoprotective effect by antihyperuricemic therapy may be due to the differences of patient characteristics in each study.
The current subgroup analyses demonstrated that patients with parameters including male, < 70 years, systolic blood pressure < 130 mmHg, no abnormal cholesterol levels, or no diabetes mellitus are more likely to show an inverse correlation between mUA level and 6mΔeGFR, as well as amelioration of eGFR decline by UA lowering, than their opposite subgroups. This result suggests that the renoprotective effects of febuxostat differ among CKD patients. The factors masking the renoprotective effects detected in this study, including aging, hypertension, hyperlipidemia, and diabetes mellitus, are important vascular risk factors causing endothelial dysfunction and arteriosclerosis. XO activation and hyperuricemia are also thought to be vascular injury factors. XO activation increases oxidative stress, suppresses nitric oxide production, and activates the renin-angiotensin system [
6‐
8]. Hyperuricemia is reported to induce arteriolopathy of preglomerular vessels, which impairs endothelial function and the autoregulatory response of afferent arterioles, resulting in glomerular hypertension [
8]. Antihyperuricemic therapy using febuxostat is expected to improve these vascular events induced by XO activation and hyperuricemia; however, its vascular protective effects may be masked by other vascular risk factors. The current results demonstrate that eGFR is likely to decrease despite optimal control of the UA level in CKD patients with various vascular risk factors, and that coexisting vascular risk factors exert additive masking effects against febuxostat renoprotection. To detect the reliable renoprotective effects of antihyperuricemic therapy, clinical studies in which control and interventional groups have small number of background vascular risk factors will be needed in the future.
The current findings do not negate the therapeutic significance of febuxostat treatment for CKD patients with various risk factors such as diabetes mellitus. Zoppini et al. reported that hyperuricemia is an independent risk factor for CKD progression in patients with type 2 diabetes [
2]. We believe that antihyperuricemic therapy is an important interventional strategy to protect kidney function even in patients with diabetes mellitus, because it is expected to ameliorate additional vascular damage induced by hyperuricemia.
Limitations of this study include a small number of subjects from a short duration, single-center retrospective follow-up study. The analyses with small n-number and short-duration follow-up might show borderline significance values, especially in subgroup analyses. However, we believe that the findings of the current study would provide useful information for clinical studies in the future, by identifying the patient characteristics influencing the renoprotective effects of febuxostat therapy. Further clarification of the differences in the renoprotection effects by antihyperuricemic therapy among various patient subgroups will require future examination in large-scale, extended-duration, multi-center prospective studies. Second, the dosage of febuxostat in Japan differs from that in other countries due to safety assessment and the dose ranging studies. The proportions of patients with serum UA < 6.0 mg/dl were 40 mg, 56%; 80 mg, 76%; and 120 mg, 94% in the dose ranging study in USA [
22], whereas they were 20 mg, 46.5%; 40 mg, 82.9%; 60 mg, 83.3%; and 80 mg, 87.8% in Japan [
23]. These findings suggest that low-dose febuxostat therapy can lower serum UA levels in Japanese patients to a greater extent than in European and Americans patients. From the results of the APEX, FACT, and CONFIRMS studies [
17,
18,
24], febuxostat dose was set at 40–80 mg/day in USA and 80–120 mg/day in Europe, whereas the initial dose was set at 10 mg/day, with careful gradual dose increase of febuxostat (maximum 60 mg/day), in Japan based on the results of the allopurinol-controlled comparative studies and a long-term study [
25‐
27]. This suggested a possible difference in antihyperuricemic effect due to factors such as genetic differences, body mass index, or dietary habits. Therefore, the result of the present study was not directly applicable to patients of other countries. Third, the current study included only a small of number female patients, in whom a high number of adverse events were reported, making the evaluation of the clinical effects of febuxostat therapy in female patients inconclusive. Therefore further research for female patients is necessary. In addition, factors not examined in this study might have influenced the study results; therefore, careful interpretation of the results would be required.