The association between hyperuricemia, gout, and deterioration of renal function is well established.1-5 Analysis of the National Health and Nutrition Examination Survey III (2007–2008) data revealed that of the estimated 8.3 million US citizens with gout, 71% also have stage 2 or higher chronic kidney disease (CKD).6,7 In healthy adults, an annual reduction in glomerular filtration rate (GFR) of 0.8 to 1.3 mL/min is expected, while in untreated hyperuricemic adults, the annual decline in estimated GFR (eGFR) is 2.5 mL/min/1.73 m2 per year.8,9
Febuxostat, a selective xanthine oxidase (XO) inhibitor, is an effective and well- tolerated therapy for the management of hyperuricemia in patients with gout.10 Recently, in a cohort of 116 subjects with hyperuricemic gout (baseline serum uric acid [SUA] levels > 8 mg/dL), it was revealed that ≥ 5 years of urate-lowering therapy (ULT) with febuxostat was associated with more stable and even improved renal function.11 The US Food and Drug Administration–approved doses of febuxostat, 40 or 80 mg daily, are more efficacious than the more commonly prescribed doses of allopurinol (≤ 300 mg daily)12 in subjects with mild or moderate renal impairment.13 In addition, with febuxostat treatment, dose adjustment is not needed in patients with mild-to-moderate renal impairment.13 This post hoc analysis was designed to evaluate the effect of febuxostat treatment for ≤ 48 months based on eGFR in a much larger population of subjects with hyperuricemic gout than previously explored, and to quantify the preservation of eGFR associated with the extent of change in SUA level achieved with long-term use.
Subjects completing either the Febuxostat Versus Allopurinol Controlled Trial (FACT; 12-month duration)14 or Allopurinol- and Placebo-Controlled, Efficacy Study of Febuxostat (APEX; 6-month duration)15 phase 3 randomized trials were invited to enroll in the long-term, open-label Febuxostat/Allopurinol Comparative Extension Long-Term (EXCEL) study,16 and received ULT for 28 weeks to 4 years. A full description of study design and criteria for subject enrollment has been previously reported by Becker et al.16 Baseline data were collected at enrollment in the APEX and FACT trials. Exclusion criteria for the previous 2 trials included secondary hyperuricemia, a history of xanthinuria, pregnancy or lactation, serum creatinine (SCr) level > 2 mg/dL in the APEX trial and > 1.5 mg/dL in the FACT trial, other significant medical conditions that would interfere with treatment safety or compliance, or known intolerance to allopurinol. All protocols, informed consent/privacy authorization forms (the Health Insurance Portability and Accountability Act form in the United States and the Personal Information Protection and Electronic Documents Act form in Canada), and subject information forms were approved by an institutional review board or an independent ethics committee.
The FACT14 and APEX15 trials were comparative, randomized studies with subjects receiving placebo, febuxostat (80, 120, or 240 mg), or allopurinol (100 or 300 mg). Subjects who completed the FACT or APEX trials were then invited to enroll in the open-label EXCEL study. Under the initial EXCEL study protocol, 351 subjects were assigned to receive febuxostat 80 mg16; however, the EXCEL protocol was amended to include the febuxostat 120 mg and allopurinol treatment groups. Subjects enrolled after this protocol amendment (n = 735) were randomized in a 2:2:1 ratio to receive either febuxostat 80 mg, febuxostat 120 mg, or allopurinol 300 or 100 mg, regardless of treatment received during the FACT or APEX trials. Selection of allopurinol doses was based on renal function: 300 mg if SCr level was ≤ 1.5 mg/dL and 100 mg if SCr level was > 1.5 to ≤ 2 mg/dL. After initial treatment assignment, subjects were permitted to switch between either doses of febuxostat and/or allopurinol during the first 6 months of treatment to achieve and maintain an SUA level ≥ 3 to < 6 mg/dL, due to an adverse event (AE), or at the investigator’s discretion. Any subject with 3 consecutive SUA levels > 6 mg/dL was to be withdrawn from the study and was considered to have experienced therapeutic failure. Subjects were to be on a stable daily maintenance dose of febuxostat 80 mg, febuxostat 120 mg, or allopurinol (300 or 100 mg, based on renal function) by the end of the first 6 months of treatment. Of the 1086 subjects who enrolled in the EXCEL study, 145 initially received allopurinol and 941 received febuxostat (80 or 120 mg). As subjects were able to switch between study drugs and doses during the first 6 months of the EXCEL study to achieve and maintain an SUA level ≥ 3 to < 6 mg/dL, only data from the 551 subjects who received any dose of febuxostat (no placebo or allopurinol) throughout the duration of the phase 3 trials (FACT or APEX and EXCEL) were included for the current analysis.
During the first 2 months of the EXCEL study, subjects were provided with either naproxen 250 mg twice daily or colchicine 0.6 mg once daily for gout flare prophylaxis. Subjects were not permitted to take losartan during the study. Subjects receiving febuxostat were permitted to take thiazide diuretics and valsartan hydrochlorothiazide. Vital signs, SUA level, laboratory tests, AEs, and concomitant medications were assessed every 2 months and at the final visit.
In this EXCEL analysis, eGFRs were calculated using the Modification of Diet in Renal Disease (MDRD),17 Cockcroft-Gault (CG),18 and the newly developed Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI)19 equations.
The long-term effects of SUA reduction on renal function (using MDRD-calculated eGFR) were sequentially evaluated by dividing the study cohort into 5 groups representing the spectrum of chronic reduction of SUA from baseline levels needed to achieve the goal range of ≥ 3 to < 6 mg/dL. Specifically, the study groups, based on mean reduction of SUA from baseline level, were: A) ≤ 3 mg/dL; B) > 3 to ≤ 4 mg/dL; C) > 4 to ≤ 5 mg/dL; D) > 5 to ≤ 6 mg/dL; and E) > 6 mg/dL. In addition, the relationship between average SUA level on treatment (SUA level < 4 mg/dL, ≥ 4 to < 5 mg/dL, ≥ 5 to < 6 mg/dL, and ≥ 6 mg/dL) and mean change in eGFR was examined.
The use of concomitant medications known to affect renal function was evaluated during the study observation period. Specifically, we assessed the balance in the quantitative use of nonsteroidal anti-inflammatory drugs (NSAIDs), angiotensin-converting enzyme (ACE) inhibitors, and angiotensin receptor blockers (ARBs) in study groups A through E.
All baseline data reported were collected at the beginning of the FACT or APEX trials. A mathematical model was developed to predict the deterioration of renal function in this cohort should hyperuricemia be left untreated. The model determined the relationship between the change in SUA from baseline level and the change in eGFR using the MDRD equation from baseline over time. For each year, the eGFR and SUA level selected for model analysis were those nearest to the year interval (eg, for year 1, the eGFR and SUA level nearest to 365 days from the first dose of study drug taken). Average SUA level on treatment was determined from post-baseline SUA level. The predictors (time [in year], categorized SUA change from baseline level [study groups A through E], SUA change from baseline level, and baseline eGFR [< 60 vs ≥ 60 mL/min]) were evaluated for a relationship to eGFR change from baseline using a repeated-measures linear model. Backward selection (
Of the 1086 subjects who had previously completed either the FACT (52 weeks)14 or APEX (28 weeks)15 trial and enrolled in the EXCEL study,16 only 551 received febuxostat (any dose) monotherapy. Subjects in this study cohort who completed the EXCEL study received febuxostat for ≤ 48 months. There were 207 (37.6%) subjects who prematurely discontinued treatment. The most common reasons for premature discontinuation were: lost to follow-up (n = 45; 8.2%), AEs (n = 43; 7.8%), personal reasons (n = 41; 7.4%), therapeutic failure (n = 20; 3.6%), and other (eg, site closure, withdrawn consent, or noncompliance; n = 48; 8.7%). These discontinuation rates did not differ from those reported for the entire EXCEL study population,16 except for therapeutic failure (6.4% of the total EXCEL population discontinued due to therapeutic failure). The majority of subjects were Caucasian (79.3%) and aged 45 to 65 years (60.1%); 357 (64.8%) had a body mass index of ≥ 30 kg/m2; 528 (95.8%) were men; and 236 (42.8%) had hypertension (Table 1). Mean baseline SUA level was 9.8 mg/dL (range, 7.4–14.9 mg/dL), and mean duration of gout was 11 years prior to study entrance.
Baseline Demographics, Characteristics, and Medical Histories by Change From Baseline in SUA Levela
|Variable||Group A n = 53||Group B n = 87||Group C n = 137||Group D n = 134||Group E n = 140||All Subjects N = 551|
|Male, n (%)||53 (100.0)||85 (97.7)||133 (97.1)||129 (96.3)||128 (91.4)||528 (95.8)|
|Race, n (%)|
|Asian||1 (1.9)||4 (4.6)||4 (2.9)||6 (4.5)||2 (1.4)||17 (3.1)|
|Black||8 (15.1)||5 (5.7)||8 (5.8)||8 (6.0)||19 (13.6)||48 (8.7)|
|Caucasian||33 (62.3)||73 (83.9)||113 (82.5)||109 (81.3)||109 (77.9)||437 (79.3)|
|Hispanic||8 (15.1)||5 (5.7)||10 (7.3)||6 (4.5)||5 (3.6)||34 (6.2)|
|Other||3 (5.7)||0||2 (1.5)||5 (3.7)||5 (3.6)||15 (2.7)|
|Age, y, mean ± SD||47.0 ± 9.53||50.7 ± 11.49||51.9 ± 10.26||51.9 ± 11.13||52.2 ± 13.61||51.3 ± 11.59|
|≥ 30 kg/m2, n (%)||36 (67.9)||55 (63.2)||92 (67.2)||73 (54.5)||101 (72.1)||357 (64.8)|
|Mean, kg/m2 ± SD||33.2 ± 5.54||32.7 ± 6.01||32.9 ± 5.78||31.7 ± 5.63||33.3 ± 6.05||32.7 ± 5.84|
|Alcohol use, n (%)||37 (69.8)||58 (66.7)||94 (68.6)||90 (67.2)||82 (58.6)||361 (65.5)|
|Years with gout, mean ± SD||9.8 ± 7.90||9.6 ± 7.07||11.2 ± 8.80||11.3 ± 8.95||12.0 ± 10.65||11.0 ± 9.04|
|Tophi present, n (%)||5 (9.4)||15 (17.2)||21 (15.3)||26 (19.4)||33 (23.6)||100 (18.1)|
|SUA level, mg/dL, mean ± SD||9.0 ± 1.19||9.0 ± 0.95||9.4 ± 0.88||9.9 ± 0.88||10.9 ± 1.29||9.8 ± 1.26|
|Medical history, n (%)|
|Cardiovascular disease||0||8 (9.2)||11 (8.0)||17 (12.7)||23 (16.4)||59 (10.7)|
|Diabetes||4 (7.5)||5 (5.7)||4 (2.9)||7 (5.2)||12 (8.6)||32 (5.8)|
|Hypertension||26 (49.1)||34 (39.1)||50 (36.5)||56 (41.8)||70 (50.0)||236 (42.8)|
aCFB SUA level to average SUA level on treatment: group A = CFB ≤ 3 mg/dL; group B = CFB > 3 to ≤ 4 mg/dL; group C = CFB > 4 to ≤ 5 mg/dL; group D = CFB > 5 to ≤ 6 mg/dL; group E = CFB ≥ 6 mg/dL.
At baseline, based on eGFR calculated by the MDRD equation,17 43.7% (n = 241) of subjects had normal renal function, defined as an eGFR ≥ 90 mL/min; 49.2% (n = 271) had mild renal impairment (eGFR ≥ 60 to < 90 mL/min); and 6.9% (n = 38) had moderate renal impairment (eGFR ≥ 30 to < 60 mL/min). One subject had a baseline eGFR < 30 mL/min and was included in the moderate renal impairment group. Baseline renal function, calculated by the CG equation,18 distributed 52.8%, 36.3%, and 10.9% of subjects into normal, mildly impaired, and moderately impaired renal function groups, respectively, although this distribution shifted to 47.4%, 44.6%, and 8%, respectively, when calculated by the CKD-EPI equation.19 Measures of renal function (eGFR predicted by MDRD, CG, and CKD-EPI equations, and as determined by SCr, averaged across all subjects) are presented in Table 2.
SUA Level, SUA Percent Change From Baseline Averaged Across All Subjects, and Renal Function Summary Statistics
|Time Points||SUA, mg/dL||% Change From Baseline in SUA Level||eGFra, mL/min||eGFrb, mL/min||eGFrc, mL/min||SCr, mg/dL|
|Baseline (n = 551)||9.8 (1.40)||92.1 (27.42)||87.9 (23.57)||87.4 (19.30)||1.0 (0.22)|
|Month 6 (n = 551)||4.5 (1.65)||−53.2 (17.41)||93.7 (27.75)||89.9 (23.22)||88.7 (19.62)||1.0 (0.23)|
|Month 12 (n = 534)||4.8 (1.65)||−50.5 (17.28)||95.8 (30.86)||92.6 (27.80)||90.3 (19.65)||1.0 (0.22)|
|Month 18 (n = 505)||4.8 (1.37)||−50.0 (14.87)||92.7 (27.56)||89.2 (23.35)||88.1 (19.56)||1.0 (0.26)|
|Month 24 (n = 462)||4.7 (1.34)||−50.9 (14.20)||90.7 (26.86)||87.2 (22.34)||86.1 (18.65)||1.0 (0.28)|
|Month 30 (n = 422)||4.7 (1.35)||−51.2 (14.55)||89.2 (24.88)||85.6 (20.35)||85.1 (18.30)||1.0 (0.21)|
|Month 36 (n = 388)||4.7 (1.34)||−50.8 (14.86)||86.1 (23.59)||82.4 (18.66)||82.7 (18.55)||1.1 (0.23)|
|Month 42 (n = 217)||4.7 (1.21)||−51.3 (13.46)||84.1 (23.09)||80.9 (18.32)||81.0 (17.98)||1.1 (0.21)|
|Month 48 (n = 88)||4.5 (1.17)||−53.2 (12.43)||82.6 (23.09)||78.8 (18.79)||78.7 (17.77)||1.1 (0.20)|
aCalculated by the Cockcroft-Gault equation.
bCalculated by the Modification of Diet in Renal Disease equation.
cCalculated by the Chronic Kidney Disease Epidemiology Collaboration equation.
Table 3 provides the mean change in eGFR from baseline for each of the 5 groups, A through E, of chronic SUA level reduction. Greater sustained decreases in SUA levels were associated with less renal function decline (
Overall mean change in eGFR for all subjects compared with expected declines in normouricemic and untreated hyperuricemic individuals based on Kuo et al9 (SDs for all febuxostat-treated subjects can be found in Table 3).
Change in eGFra From Baseline by Change From Baseline in SUAb, Averaged Across All Subjects
aCalculated by the Modification of Diet in Renal Disease equation.
bCFB SUA level to average SUA level on treatment: group A = CFB ≤ 3 mg/dL; group B = CFB > 3 to ≤ 4 mg/dL; group C = CFB > 4 to ≤ 5 mg/dL; group D = CFB > 5 to ≤ 6 mg/dL; group E = CFB ≥ 6 mg/dL.
Concomitant medication use was assessed for each subject throughout their participation in the EXCEL study. Chronic use of NSAIDs was exclusionary in the APEX and FACT trials and carried over into the EXCEL study. Therefore, the use of NSAIDs was acute in nature, with short therapeutic courses administered almost entirely for the management of gout flares. The average duration of NSAID therapy was approximately 5 days. Treatment response for groups A through E showed the mean duration of NSAID therapy to be comparable. Angiotensin-converting enzyme inhibitors and ARBs were used by 30% and 14% of EXCEL subjects in the current analysis, respectively. Usage was chronic and stable throughout the study. The indication for ACE inhibitors or ARBs at entry into the study was for the management of hypertension and/or for renal glomerular protection. The use profiles for both types of therapeutic agents were comparable between treatment response groups A through E.
Rates of most frequently reported AEs and all serious AEs in this cohort were reflective of those observed in the total EXCEL study population.16 The most common AE leading to premature discontinuation was investigator-defined liver function abnormalities (n = 6). Eight subjects in this cohort died during the study. Five deaths were attributed to cardiovascular events sustained by subjects with extensive histories of cardiovascular disease. Two subjects died from cancer and 1 subject died from postsurgical sepsis.
Multiple studies, both prospective population cohort studies and retrospective database analyses, have established the association between hyperuricemia, gout, and renal disease.3-5,9,20,21 In the largest retrospective database analysis to date, a Taiwanese cohort of 63 785 subjects without gout was followed for 12 years. Compared with normouricemic subjects (mean SUA level, 5.6 ± 1.3 mg/dL; n = 51 916), subjects with hyperuricemia at baseline (SUA level > 7.7 mg/dL for men, > 6.6 mg/dL for women; mean SUA level, 8.5 ± 3.5 mg/dL; n = 11 869), after multivariate adjustment, had an odds ratio for an accelerated eGFR decline of ≥ 3 mL/min/1.73 m2 per year of 1.28 (95% CI, 1.23–1.33;
Our analysis of 551 subjects with hyperuricemia and gout effectively treated with febuxostat, the largest such analysis to date, provides additional supportive evidence in humans to underscore the beneficial effects on renal function of reducing SUA levels in this study population. The greater the chronic reduction in SUA levels in hyperuricemic subjects with gout, the less the decline in eGFR. The data presented here support earlier published results.11 In the smaller Febuxostat Open-Label Clinical Trial of Urate-Lowering Efficacy and Safety (FOCUS),22 116 subjects received a febuxostat 40-, 80-, or 120-mg daily dose for ≤ 5 years. The observed maintenance or improvement in eGFR was inversely correlated with the extent of SUA level reduction from baseline.11 The previously reported FOCUS data indicated that for every 1-mg/dL decrease in SUA level, the projected improvement in eGFR from the untreated value was 1 mL/min per year. In our current analysis of 551 EXCEL study subjects who received various doses of febuxostat for 28 weeks to 4 years, our statistical model projected that for every 1-mg/dL SUA level decrease from baseline, the projected improvement in eGFR was 1.15 mL/min per year compared with no SUA level decrease in individuals with gout.
A similar inverse relationship between the extent of SUA level decline after initiation of ULT with allopurinol and stabilization of renal function was reported by Goicoechea et al23 in a cohort of patients with CKD who did not have gout. Patients with baseline eGFR < 60 mL/min (no selection for baseline SUA level) were randomized to receive either allopurinol 100 mg daily (n = 57) or continue with standard of care only (control; n = 56) for 24 months. In the control group, after 24 months, mean eGFR decreased by 3.3 ± 1.2 mL/min, whereas it increased by 1.3 ± 1.3 mL/min in the allopurinol group. The ability of allopurinol to impede renal disease progression was found to be independent of age, sex, concurrent diabetes, C-reactive protein, albuminuria, and use of renin-angiotensin system blockers.23 Furthermore, when the change in SUA level from baseline to 24 months was plotted against the change in eGFR, a significant inverse correlation between changes was identified (
The mechanisms by which chronic SUA level reduction contributed to the beneficial effect on renal function noted in our study are not currently known; however, some potential contributing mechanisms are apparent and a synopsis of these contributing factors is warranted. In ischemic and inflammatory vascular disease, there is an important up-regulation of the production of XO on the surface of endothelial cells. This enriched activity of the vascular-associated XO is known to: 1) use molecular oxygen as hypoxanthine is converted to xanthine and xanthine to uric acid, thereby potentially decreasing tissue oxygenation; 2) increase the production of reactive oxygen species, leading to oxidative stress; 3) increase degradation of adenosine triphosphate to adenosine monophosphate, thereby decreasing local energy resources; and 4) increase catabolism of nitric oxide with an associated decrease in vascular dilatory response.24,25 Chronic inhibition of XO during the course of our investigation is likely to have inhibited much of these activities, thereby providing a mechanism that improved function within the renal vasculature.
Elevated SUA level is associated with significant elevation of blood pressure in human adults and children in a dose-dependent manner.26-29 In adolescents, primary hypertension can be reversed by SUA level reduction through XO inhibition.30 In our study, there was no protocol-mandated approach to monitoring blood pressure, and managing elevated blood pressure was at the discretion of the investigators. Nonetheless, the effect of XO inhibition may have had a beneficial effect on the renal function of the study participants.
Our subjects had a decade-long history of gout prior to participation in the investigation and their average baseline SUA level was approximately 10 mg/dL, a value that substantially exceeds the biologic solubility of SUA (6.8 mg/dL) at core body temperature. Therefore, it is possible that micro-tophi of monosodium urate crystals may have chronically accumulated within the renal tissues and that the long-term use of febuxostat in our study progressively mobilized the latter urate deposits, resulting in improvement in renal function.
The benefit on renal function reported for this cohort of the EXCEL study was particularly manifest in the first 2 years of gout treatment; however, it was not sustained, but was followed thereafter by a lesser quantitative benefit (Figure 2), wherein the slope of the curve of renal function was comparable with the rate of renal function decrease manifest in non-hyperuricemic adults, as reported by Kuo et al.9 In years 3 and 4 of our study, there were fewer subjects than in the earlier years; therefore, robust prospective studies will be needed to further characterize the quantitative effect of gout treatment on the long-term pattern of renal function benefits.
Both the current analysis and our earlier study of the FOCUS data11 are post hoc in nature, which is an important limitation. There is a paucity of reported studies examining the impact of ULT on renal function in subjects with gout, whereas there are considerably more studies demonstrating the benefits of ULT on renal function in hyperuricemic subjects with CKD and/or hypertension who did not have gout.31-35 In addition, we did not explore the impact of long-term ULT on other factors that could also influence eGFR, such as change in weight or hypertensive status. It is possible that such changes could confound the results. Prospective trials are needed to confirm our results and those of others reporting the renal benefits of lowering SUA level with long-term ULT in patients with hyperuricemia and gout.
In the context of renal function, particularly hypertension, multiple therapeutic agents are known to affect SUA levels. In a retrospective, nested, case-controlled claims database analysis of 24 678 cases with incidental gout and a random sample of 50 000 matched controls, Choi et al36 evaluated the risk of gout, stratified by hypertension or no hypertension, and the use of antihypertensive medications. Multivariate analysis revealed that a decreased risk for incidental gout among subjects with hypertension was associated with the use of calcium channel blockers and losartan, but was increased for subjects taking diuretics, β-blockers, ACE inhibitors, and non-losartan ARBs.36 Results were similar among those subjects without hypertension. In our analysis, use of such agents, along with NSAIDs, was comparable across treatment response groups A through E, and their use was at the discretion of the investigators. Therefore, any impact on SUA level by use of such medications would be expected to be comparable as well.
The results of this analysis have important clinical implications that need to be confirmed in future prospective investigations. The findings of this largest-to-date, multi-year study of the impact of febuxostat on the progression of renal function in carefully treated subjects with gout further supports earlier reported results in a smaller cohort of subjects with gout.11 The current data underscore the importance of maintaining a biochemically appropriate target SUA level of < 6 mg/dL in the long-term management of gout, as recommended by currently published guidelines.37,38 We cannot conclude from our study design that, in patients with gout, the target SUA level should also be < 6 mg/dL to beneficially impact renal function. Nonetheless, it appears to be a reasonable goal until further studies clarify the most desirable target SUA level(s) for ensuring renal function benefit in the long-term management of patients with gout. The clinical approach used in our subjects appears to have provided important benefit in the progression of renal function compared with the available data defining renal function progression in non-treated subjects with gout, as well as in subjects with hyperuricemia who do not have gout. In the long-term management of subjects with gout who were treated with febuxostat (at least during the study period of some 4 years in the currently reported analysis), it appears that for every 1-mg/dL reduction in SUA level, there is the preservation of 1.15 mL/min of eGFR. Impaired renal function is one of the most common comorbidities of gout7; therefore, these important new observations in the management of gout require clinical attention and prospective study confirmation.
The FACT (NCT# 00102440), APEX (NCT# 00174915), and EXCEL (NCT# 00175019) trials were completely funded by TAP Pharmaceutical Products, Inc, which is now a part of Takeda Global Research