Randomized Controlled Trial of Febuxostat Versus Allopurinol or Placebo in Individuals with Higher Urinary Uric Acid Excretion and Calcium Stones

Abstract

Background and objectives 

Higher urinary uric acid excretion is a suspected risk factor for calcium oxalate stone formation. Febuxostat, a xanthine oxidoreductase inhibitor, is effective in lowering serum urate concentration and urinary uric acid excretion in healthy volunteers and people with gout. This work studied whether febuxostat, compared with allopurinol and placebo, would reduce 24-hour urinary uric acid excretion and prevent stone growth or new stone formation.

Design, setting, participants, & measurements 

In this 6-month, double-blind, multicenter, randomized controlled trial, hyperuricosuric participants with a recent history of calcium stones and one or more radio-opaque calcium stone ≥3 mm (as seen by multidetector computed tomography) received daily febuxostat at 80 mg, allopurinol at 300 mg, or placebo. The primary end point was percent change from baseline to month 6 in 24-hour urinary uric acid. Secondary end points included percent change from baseline to month 6 in size of index stone and change from baseline in the mean number of stones and 24-hour creatinine clearance.

Results 

Of 99 enrolled participants, 86 participants completed the study. Febuxostat led to significantly greater reduction in 24-hour urinary uric acid (−58.6%) than either allopurinol (−36.4%; P=0.003) or placebo (−12.7%; P<0.001). Percent change from baseline in the size of the largest calcium stone was not different with febuxostat compared with allopurinol or placebo. There was no change in stone size, stone number, or renal function. No new safety concerns were noted for either drug.

Conclusions 

Febuxostat (80 mg) lowered 24-hour urinary uric acid significantly more than allopurinol (300 mg) in stone formers with higher urinary uric acid excretion after 6 months of treatment. There was no change in stone size or number over the 6-month period.

Introduction

The prevalence of kidney stones is increasing, regardless of sex or race. The reported prevalence of kidney stones during 1988–1994 was 5.2% (¹). During 2007–2010, this had risen to 8.8% (²), an increase of 69%. Approximately 50% of first-time stone formers will have a recurrent stone within 5 years, despite preventive therapy (³). For patients managed with minimally invasive procedures, the risk of recurrence after 5 years is 63% (⁴). There remains a need to develop medical treatment modalities that can reduce the likelihood of recurrent nephrolithiasis.

Calcium oxalate (CaOx) is the predominant component of kidney stones (⁵). Elevated urinary uric acid (uUA) or hyperuricosuria is a suspected risk factor for CaOx stone formation, because up to one third of CaOx stone formers have higher uUA excretion (^6–10). This association is further supported by studies showing that treatment with the urate-lowering drug allopurinol, a xanthine oxidoreductase inhibitor (XORI), led to reduced rates of stone recurrence (^6–8,10). Febuxostat, a newer XORI, is highly efficacious in reducing serum urate levels (sUAs) (^11–13). It also reduces uUA levels in healthy volunteers (¹⁴) and in individuals with hyperuricemic gout (¹⁵). Therefore, febuxostat may provide an alternative or superior treatment to allopurinol in stone prevention. This study evaluated treatment with febuxostat compared with allopurinol (at the approved and previously studied doses for higher uUA excretion) or placebo in the reduction of 24-hour uUA excretion in participants with higher uUA and a recent history of kidney stones.

Materials and Methods

Study Design and Participant Selection

This was a phase 2, 6-month, multisite, United States-based, randomized febuxostat and allopurinol- and placebo-controlled double blind trial (clinicaltrials.gov NCT01077284). The study was conducted in compliance with Institutional Review Board regulations, Good Clinical Practice guidelines, and the Declaration of Helsinki.

All participants signed informed consent and Health Insurance Portability and Accountability Act of 1996 forms before initiation of any study procedures. Men and women with higher uUA excretion (>700 mg/24 h) over the age of 18 years with a recent (≤5 years) history of kidney stones and one or more radio-opaque calcium kidney stone ≥3 mm in its longest in-plane diameter at screening were enrolled. Radio-opacity was judged on the abdominal scout film, taken at the beginning of the computed tomograph, which is equivalent to a plain abdominal film. Whether stones were composed of calcium oxalate or calcium phosphate could not be determined, but the former presumably constituted the majority. Because the qualifying stone was radio-opaque, uric acid composition was very unlikely. If premenopausal, women were using an accepted form of contraception and could not be pregnant or lactating. Participants were excluded if they had any of the following criteria: gout, secondary hyperuricemia, or hyperparathyroidism; higher urinary calcium excretion (>250 mg/d for women and >300 mg/d for men or >4 mg/kg per day if body weight was ≥75 kg for men and ≥62.5 kg for women); a history of xanthinuria, inflammatory bowel disease, or bowel resection; renal tubular acidosis; cancer in the past 5 years; febuxostat, allopurinol, or probenecid prescription in the past 2 years; a known hypersensitivity to allopurinol or febuxostat; any other clinically relevant conditions that would interfere with participation in the study; or use of excluded medications (salicylates>325 mg/d, azathioprine, mercaptopurine, theophylline, colchicine, pyrazinamide, sulfamethoxazole/trimethoprim, or losartan). Participants were not excluded if they were taking potassium citrate. Participants receiving diuretics were to be on a stable dose for ≥30 days before screening. Eligible individuals were randomized 1:1:1 to receive once-daily febuxostat (80 mg), allopurinol (300 or 200 mg depending on renal function; 300 mg if creatinine clearance [Ccr] was ≥60 ml/min or 200 mg if Ccr was ≥30 to <60 ml/min), or placebo.

The presence and size of kidney stones were assessed by multidetector computed tomography (MDCT) (¹⁶) at screening and the month 6 or final visit. Abdominal MDCT was done without contrast using a 16+ detector helical scanner with a 30- to 40-cm field of view and a 512×512-voxel matrix. The patient was positioned supine, and after an abdominal scout scan, axial images were acquired at 1.25-mm slice thickness during a single breath hold. Digital images or hardcopy films (if digital transfer was not available) were then sent to Perceptive Informatics, Inc. (Waltham, MA) for image quality control and assessment. Adjudicated image assessments were performed by blinded, independent radiologists.

At screening and on the day before the month 3 and 6 visits, 24-hour urine collections were performed to measure uUA and Ccr along with urinary excretion of calcium, oxalate, citrate, creatinine, magnesium, sodium, potassium, sulfate, total protein, and albumin, and to determine pH. Other clinical laboratory assessments, including sUA and hepatic enzymes, were performed at screening, day 1, month 3, and month 6. Adverse events (AEs) were recorded using Medical Dictionary for Regulatory Activities terminology.

End Points

The primary efficacy end point was the percent change from baseline to month 6 in 24-hour uUA excretion. Secondary end points were percent change from baseline to month 6 in the in-plane diameter of the largest calcium stone, change from baseline to month 6 in the number of calcium stones, and change from baseline to month 6 in 24-hour measured Ccr. Additional end points included percentage of participants with sUA<6.0 mg/dl at month 6, percent change from baseline to months 3 and 6 in sUA levels, change from baseline in estimated GFR (eGFR) as calculated by the Modification of Diet in Renal Disease Equation (¹⁷), and analysis of AEs and serious AEs.

Statistical Analyses

Efficacy analyses were performed on the full analysis set, which was defined as all randomized participants who received one or more dose of study drug and had a baseline 24-hour uUA >700 mg and one or more calcium stone ≥3 mm in the largest in-plane diameter. Safety analyses were performed on all participants who received one or more dose of study drug. Descriptive statistics were presented for all variables assessed, and all statistical tests were two-sided and conducted at the 0.05 significance level. Pairwise comparisons of febuxostat with allopurinol and placebo for the primary end point and the calcium stone-related secondary efficacy end points were made using the Wilcoxon Mann–Whitney rank sum test. For the secondary efficacy end point of change in 24-hour Ccr and additional efficacy end points of change in eGFR and percent change in sUA, pairwise comparisons between treatment groups were made using contrast statements within the framework of ANOVA with treatment group as a factor. For each primary and secondary end point, the Hochberg method for multiple comparisons was used to ensure that the overall 0.050 level of significance was maintained for pairwise comparisons of febuxostat with allopurinol and placebo.

For the primary analysis, missing 24-hour values from the month 6 visit were imputed with the baseline value if the subject dropped out from the study because of any AE. Otherwise, the missing value was imputed with the last available postbaseline value. Missing data were imputed for only two end points: uUA and measured Ccr. Month 6 values were imputed for no more than three subjects per treatment group. The imputation did not significantly influence the results. Data for 24-hour urinary solutes were summarized only for participants with sample values at baseline and month 6.

A total of 90 participants (30 per treatment group) was planned for enrollment. This sample size was considered sufficient to characterize the effect of febuxostat on 24-hour uUA excretion compared with allopurinol, and it was not designed based on formal power calculations.

Results

Participants

Of 895 individuals screened, 99 individuals were enrolled and randomized (33 per treatment group) (Figure 1). The majority of potential enrollees who failed the screening did so because 24-hour uUA levels were <700 mg. Thirteen participants prematurely discontinued: 5, 5, and 3 participants in the febuxostat, allopurinol, and placebo groups, respectively. Overall, the primary reasons for discontinuation were voluntary withdrawal (6.1%), lost to follow-up (3.0%), AEs (3.0%), and other (1.0%). Demographics and baseline characteristics are listed in Table 1. Most participants were men (85.9), white (86.9%), and over 45 years of age (61.6%). Mean body mass index (BMI) was 32.8 kg/m². Mean lifetime history of stone episodes was 10.9, mean largest stone diameter was 9.9 mm, and mean number of stones on MDCT was 5.6. Mean baseline Ccr was 147 ml/min, sUA was 6.3 mg/dl, urine calcium excretion was 272.2 mg/d, and uUA was 952.7 mg/d. No participant in the allopurinol group had moderate renal impairment; therefore, all participants in that group received 300 mg allopurinol daily. Concomitant medications were used by 81 (81.8%) participants. The most frequently used concomitant medications were analgesics (32.3%) and lipid-modifying agents (28.3%). Agents acting on the renin-angiotensin system were used by 18.2% of participants, whereas 5.1% of participants were taking a diuretic. One individual was taking potassium citrate at enrollment and throughout the duration of the study.

Figure 1: Flow of participants through the study. AE, adverse event.

Table 1: Demographics and baseline characteristics

Efficacy End Points

The percent decrease from baseline to month 6 in 24-hour uUA was significantly greater in the febuxostat group (−58.6%) compared with the allopurinol (−36.4%; P=0.003) and placebo (−12.7%; P<0.001) groups. At month 3, the percent decreases from baseline for both the febuxostat and allopurinol treatment groups were similar to those values observed at month 6 (Figure 2), suggesting that the impact of either XORI on uUA occurred primarily within the first 3 months and then remained stable.

Figure 2: Percent change from baseline to months 3 and 6 in 24-hour urinary uric acid excretion. ^a P<0.001 versus placebo. ^b P=0.008 versus allopurinol. ^c P=0.003 versus allopurinol. Error bars represent the SD for each data point.

Table 2 provides data about stone size and number at baseline and month 6 as well as the change from baseline for each of these parameters. Although the percent change from baseline to month 6 in the diameter of the largest calcium stone detected by MDCT at screening was numerically greater in the febuxostat group, there were no statistically significant differences between the treatment groups in this secondary end point. In addition, there were no significant differences or meaningful changes in the total number of calcium stones between treatment groups from baseline to month 6.

Table 2: Change from baseline in stone diameter and number

The changes from baseline to month 6 in 24-hour Ccr were −9.0, −7.7, and −19.0 ml/min for the febuxostat, allopurinol, and placebo groups, respectively. These differences were not statistically significant. No changes in serum creatinine levels in any treatment group were observed. At baseline and month 6, the average serum creatinine levels were 1.04 and 1.05 mg/dl, respectively, in the febuxostat group, 1.03 and 1.03 mg/dl, respectively, in the allopurinol group, and 0.97 and 0.97 mg/dl, respectively, in the placebo group. There were no significant differences between treatment groups in the change from baseline to month 6 in eGFR.

The proportion of participants with sUA<6.0 mg/dl at month 6 was significantly greater in the febuxostat (100%) and allopurinol (88.5%) groups compared with the placebo group (44.8%; P≤0.001 versus placebo for both febuxostat and allopurinol); the difference between febuxostat and allopurinol was not statistically significant. It is worth noting that there was no entry criterion based on baseline sUA in the study, and the overall average baseline sUA was 6.3 mg/dl. The percent changes from baseline (reported as least squares mean) to months 3 and 6 were significantly greater in the febuxostat group at either time point (−43.3% and −47.3%, respectively) compared with the allopurinol (−26.9% and −26.2%, respectively [P<0.001 for both], or placebo groups (−1.9% and −1.0%, respectively [P<0.001 for both]); the differences between allopurinol and placebo at both time points were also statistically significant (P<0.001).

Results of 24-hour urine collections are reported in Table 3. The number of determinations for some values vary slightly, because some participants had missing data. No clinically important differences were observed between treatment groups for any variable at baseline or any variable other than uUA at 6 months. The change from baseline in the ratio of 24-hour uUA/creatinine was determined in a post hoc analysis to explore the possibility that incomplete or inaccurate 24-hour urine collections may have confounded the primary end point results. The mean percent change from baseline to month 6 in the uUA/creatinine ratio closely resembled the percent change from baseline in 24-hour uUA for the febuxostat (−58.2%) and allopurinol (−40.4%) groups, whereas it was smaller in the placebo group (−0.5%). This result suggests that incomplete or inaccurate 24-hour urine collections had little impact on the primary results. At both months 3 and 6, the differences in the percent change from baseline in the uUA/creatinine ratio were statistically significant (P<0.001) for febuxostat versus allopurinol, febuxostat versus placebo, and allopurinol versus placebo.

Table 3: Urinary variables at baseline and month 6 and change from baseline to month 6

Safety

More than one half of participants reported a treatment-emergent AE (59.6%): 60.6%, 57.6%, and 60.6% in the febuxostat, allopurinol, and placebo groups, respectively. The most frequently reported treatment-emergent AEs by Medical Dictionary for Regulatory Activities High-Level Terms were musculoskeletal and connective tissue pain and discomfort (12.1%), renal lithiasis (8.1%), upper respiratory tract infections (8.1%), headache (6.1%), and edema (5.1%). Only one subject in the placebo group reported a serious AE (severe pain associated with kidney stone), which required hospitalization. Among eight reports of nephrolithiasis (four, three, and one report in the placebo, febuxostat, and allopurinol groups, respectively), including the above-mentioned AEs and one severe AE, two individuals reportedly passed stones (not the index stones based on the month 6 computed tomography), and the remaining six participants experienced symptoms consistent with but not proven to be caused by stones. No participants died during the study, and no elevated hepatic enzyme tests were reported. Three participants prematurely discontinued because of AEs: one subject in the placebo group (musculoskeletal pain) and two participants in the febuxostat group (one participant with viral hepatitis and increased blood creatinine phosphokinase levels and one participant with nausea, vomiting, and renal colic).

Discussion

This study is the first randomized, multisite, blinded clinical trial designed to evaluate the effect of treatment with the XORI drugs febuxostat and allopurinol in recurrent calcium stone formers with higher uUA excretion. Our results show that febuxostat (80 mg) once daily lowered 24-hour uUA significantly more than allopurinol (300 mg) in recurrent stone formers with higher uUA excretion. After only 6 months of treatment, we did not observe significant changes in the diameter of the largest stone (determined at baseline), suggesting that these index stones did not get bigger. In addition, only one subject reported acute nephrolithiasis requiring hospitalization, and there were no significant changes in the mean numbers of stones in any treatment arm, suggesting that no new stones large enough to enumerate on MDCT had formed. There was no difference in the number of episodes of renal colic between the three groups, although the study was not powered to detect such a difference.

In one of the only other randomized controlled trials to examine the impact of XORI therapy on CaOx stone formation in recurrent stone formers, Ettinger et al. (⁸) followed 60 hyperuricosuric and normocalciuric participants randomized to receive either allopurinol (100 mg three times daily) or placebo for 6 to 24 months. The allopurinol group experienced significant declines from baseline in sUA and uUA. New calculus events, defined as the passing of a stone or appearance of stone on x-ray after 6 months of treatment, occurred in 18 participants in the placebo group and 9 participants in the allopurinol group. The mean rate of calculus events was reduced from pretreatment levels by 63.4% and 81.2% in the placebo and allopurinol groups, respectively (P<0.001 for magnitude of change from baseline for both groups). Ettinger et al. (⁸) concluded that allopurinol could “provide clinically important protection” for those stone formers. These data suggest that longer duration of treatment with febuxostat could lead to reduced recurrence of CaOx stones in stone formers with higher uUA excretion. Although allopurinol is approved for doses up to 800 mg/d for reducing elevated sUA in gout, with dose reductions recommended in participants with renal impairment, we used the recommended dose for uUA reduction in patients with CaOx stones (¹⁸), which is the same total daily dose used in the study by Ettinger et al. (⁸). If higher doses of allopurinol were used in our study, they may have led to greater uUA reductions in that treatment group.

There is still some debate as to whether higher uUA excretion is a risk factor for CaOx stone formation. In an epidemiologic cohort analysis of 24-hour urine chemistries from 2237 stone formers and 1113 controls, there was no correlation between increasing uUA and risk of stone formation. In fact, among men, there was an unexpected observed inverse relationship between lower uUA and higher risk for stone formation (¹⁹). In another study, the 24-hour urinary excretion of urate was investigated in 467 consecutive stone formers (350 men and 117 women) and 89 healthy controls with no history of stone disease. Men had a higher uUA than women, but there was no difference between stone formers and controls (²⁰). Given this doubt and uncertainty about what today constitutes hyperuricosuria, we have referred to our study population as having higher uUA excretion.

The underlying mechanism by which uric acid may promote CaOx stone formation has been investigated. It was first thought that monosodium urate crystals served as seeds that triggered heterogeneous nucleation of CaOx crystals. Although this finding was shown in inorganic solutions designed to mimic salt concentrations in urine (²¹), it was not observed in undiluted urine (²²). Instead, evidence favors salting out of calcium oxalate (²³). Grover et al. (²⁴) showed that, in urine samples with various solute concentrations, there was a dose–response relationship between higher amounts of dissolved urate added and greater tendency of the urine to crystallize CaOx, until the saturation point of urate was reached. Analysis of resulting crystals found no evidence of urate crystals. Grover et al. (²⁴) suggested that, in individuals with high urine concentrations of calcium and oxalate, even a slight increase of dissolved urate, within the normouricosuric range, could facilitate precipitation of CaOx crystals (²⁴).

Limitations of this study include the treatment duration of 6 months. Other studies that have shown the positive impact of XORI use on recurrent stone rates treated individuals for up to 6 years (^6–8). However, in early studies by Worcester and Coe (⁶) and Coe and Kavalach (⁷), many individuals ceased to experience kidney stones within 3 months of treatment with allopurinol. We did not observe any significant changes in the mean size of the largest stones in the febuxostat group or increases in mean stone numbers, but 6 months of treatment may not have been long enough to determine if new stone formation or existing stone growth had been deterred. This study did not examine the effects of XORI use on symptomatic stone episodes. Additional long-term studies are needed to assess the effect of treatment with XORIs on reduction in the number of stones, recurrent stone formation, and clinical stone episodes.

Another limitation may have been any potential inconsistencies in pH levels across the urine samples. Traditional recommendations for urine processing for the measurement of urate include alkalinization of the urine (²⁵). In urine with low pH, urate will precipitate out of solution. If not all of the urine samples were adequately alkalinized, it could lead to an underestimation of uUA levels in our samples. Participants were not prescribed any therapy to alter urine pH as part of the study protocol, and only one subject was taking potassium citrate at enrollment and throughout the duration of the study. At baseline, mean urine pH levels were 6.0, 6.0, and 5.8 in the placebo, febuxostat, and allopurinol groups, respectively. These values are not significantly different than expected in nonstone-forming adults. Therefore, there is no reason to suspect that such an artifact leading to underestimation of uUA was different at baseline than at month 3 or 6.

There were large SDs for creatinine excretion values, indicating that the 24-hour urine collections were not done in an ideal fashion. Coefficients of variation of urine creatinine excretion do increase with BMI (²⁶); the mean baseline BMI in our population was 32.8 kg/m², and therefore, it may have also contributed to the large SD. Our analysis of urine uric acid to creatinine ratios, in part, overcomes the limitations of inaccurate or incomplete collections; despite the inaccurate or incomplete collections, the ratios were consistent with the treatment effects shown by the values for 24-hour uric acid excretion.

We cannot rule out the possibility that some of the participants’ stones were uric acid stones; however, the qualifying index stone for study enrollment and repeated measurement was radio-opaque by the equivalent of a plain x-ray. Some participants had missing data on 24-hour urine collections (Table 3). Finally, the participants may not be typical stone formers, because they had experienced, on average, more than 10 stone episodes and were seen in busy endourology practices.

In conclusion, 6 months of treatment with febuxostat (80 mg) once daily led to significantly greater declines in uUA compared with allopurinol (300 mg) once daily or placebo in hyperuricosuric people with calcium nephrolithiasis. Additional studies of greater duration with greater numbers of participants and additional doses of febuxostat (i.e., 40 mg) and allopurinol are needed to determine if reduction of uUA excretion by XORI therapy will reduce the incidence of symptomatic calcium stone recurrence.

Disclosures

D.S.G. has served as a consultant for Takeda Global Research & Development Center, Inc. D.S.G. was involved in all phases of study design and execution, was involved in writing the manuscript, and had access to the full dataset. At the time of study design and execution, P.A.M., L.G., S.C., and L.M. were employees of Takeda Global Research & Development Center, Inc.

Acknowledgments

The study was funded by Takeda Global Research & Development Center, Inc. Medical writing support was provided by Meryl Gersh of AlphaBioCom, LLC in King of Prussia, PA and funded by Takeda Pharmaceutical International, Inc.

These data were presented at the American Society of Nephrology Kidney Week, San Diego, November 2, 2012 and the American College of Rheumatology Annual Meeting, Washington, D.C., November 11, 2012.