Effect of Urate-Lowering Therapy on Cardiovascular and Kidney Outcomes

Visual Abstract

Abstract

Background and objectives

Several clinical practice guidelines noted the potential benefits of urate-lowering therapy on cardiovascular disease and CKD progression; however, the effect of this regimen remains uncertain. In this systematic review, we aimed to evaluate the efficacy of urate-lowering therapy on major adverse cardiovascular events, all-cause mortality, kidney failure events, BP, and GFR.

Design, setting, participants, & measurements

We systematically searched MEDLINE, Embase, and the Cochrane databases for trials published through July 2020. We included prospective, randomized, controlled trials assessing the effects of urate-lowering therapy for at least 6 months on cardiovascular or kidney outcomes. Relevant information was extracted into a spreadsheet by two authors independently. Treatment effects were summarized using random effects meta-analysis.

Results

We identified 28 trials including a total of 6458 participants with 506 major adverse cardiovascular events and 266 kidney failure events. Overall urate-lowering therapy did not show benefits on major adverse cardiovascular events (risk ratio, 0.93; 95% confidence interval, 0.74 to 1.18) and all-cause mortality (risk ratio, 1.04; 95% confidence interval, 0.78 to 1.39) or kidney failure (risk ratio, 0.97; 95% confidence interval, 0.61 to 1.54). Nevertheless, urate-lowering therapy attenuated the decline in the slope of GFR (weighted mean difference, 1.18 ml/min per 1.73 m² per year; 95% confidence interval, 0.44 to 1.91) and lowered the mean BP (systolic BP: weighted mean difference, −3.45 mm Hg; 95% confidence interval, −6.10 to −0.80; diastolic BP: weighted mean difference, −2.02 mm Hg; 95% confidence interval, −3.25 to −0.78). There was no significant difference (risk ratio, 1.01; 95% confidence interval, 0.94 to 1.08) in the risk of adverse events between the participants receiving urate-lowering therapy and the control group.

Conclusions

Urate-lowering therapy did not produce benefits on the clinical outcomes, including major adverse cardiovascular events, all-cause mortality, and kidney failure. Thus, there is insufficient evidence to support urate lowering in patients to improve kidney and cardiovascular outcomes.

Introduction

Urate-lowering therapy has been recommended for hyperuricemia, including gout flares, tophi, and nephrolithiasis, and several clinical practice guidelines, including those of the European League against Rheumatism, the British Society for Rheumatology, the American College of Rheumatology, and Kidney Disease Improving Global Outcomes (KDIGO), have noted the potential benefits of urate-lowering therapy on cardiovascular disease and CKD (1–4). Hyperuricemia is common in the general population, and epidemiologic studies and prospective observational studies have shown that elevated serum urate concentrations are associated with worse outcomes in cardiovascular disease and kidney disease (5). However, there is considerable controversy regarding the use of agents to lower serum urate in people to reduce the risk of cardiovascular events or delay the progression of CKD (6,7). A meta-analysis of 992 participants with stages 3–5 CKD and hyperuricemia showed that urate-lowering therapy improved eGFR and reduced serum creatinine, indicating that urate-lowering therapy might delay the progression of CKD (8). A more recent meta-analysis of 1211 participants further demonstrated that urate-lowering therapy reduced the incidence of kidney failure events and cardiovascular events in adults with CKD (9), whereas another meta-analysis concluded that urate-lowering therapy might make little or no difference in the incidence of kidney failure or death in participants with or without CKD, and the evidence for cardiovascular disease was poor (10). The findings of these meta-analyses may be influenced by a small sample size and the heterogeneity of patient population selection. Recently, the results of several randomized controlled studies evaluating the effects of urate-lowering agents have been published (11–13). Accordingly, a reappraisal of the current evidence is required. We performed this systematic review and meta-analysis to evaluate the efficacy of urate-lowering therapy on the incidence of major adverse cardiovascular events, all-cause mortality, and kidney failure events, and influence of GFR, proteinuria, and BP among adults in randomized controlled trials.

Materials and Methods

Data Sources, Search Strategy, and Selection Criteria

This systematic review was performed according to a prespecified protocol that had been registered in PROSPERO (registration no. CRD42018104227). The review was carried out according to the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (14). A comprehensive literature search was performed using the following data sources: MEDLINE by Ovid (1946 to July 1, 2020), EMBASE (1974 to July 1, 2020), and the Cochrane Library database (Cochrane Central Register of Controlled Trials; no date restriction). Details on the search strategy are reported in Supplemental Table 1. The ClinicalTrials.gov website was also searched for randomized trials that were registered as completed but not yet published.

Studies were eligible for inclusion if they were randomized controlled trials; compared a urate-lowering agent with placebo or no study medication for at least 6 months; and reported clinical outcomes, including major adverse cardiovascular events, mortality, kidney failure events, or the change in GFR or BP. There was no age restriction. We excluded those trials that only compared a urate-lowering agent with another urate-lowering agent but without a non-urate-lowering therapy control arm.

Data Extraction, Quality Assessment, and Outcome Estimation

Published reports were obtained for each eligible trial, and relevant information was extracted into a spreadsheet. The sought data included study characteristics (design, method of randomization, and withdrawals/dropouts); baseline patient characteristics (age, sex, comorbidity diseases, serum urate level, CKD or non-CKD, mean eGFR, and mean proteinuria or albuminuria); type of drug used; the dose of drug; follow-up duration; number of major adverse cardiovascular events, deaths, and kidney failure events; and changes in GFR, proteinuria, BP, and serum urate concentrations. When the required quantitative data were not provided in relevant articles, we used Engauge Digitizer software (http://engauge-digitizer.findmysoft.com/) to extract exact numbers from published figures.

We evaluated all potentially relevant sources of bias using the Cochrane Collaboration risk of bias tool. The following domains were evaluated: random sequence generation; allocation concealment; blinding of the patient, investigator, and outcome assessors; reporting bias; attrition bias; and any other potential sources of bias, such as those related to trial designs or the risk for contamination or crossover between the groups. We summarized both individual and aggregate risk of bias data for the included studies. The literature search, study selection, data extraction, and quality assessment were undertaken independently by two authors (Q.C. and Z.W.) using a standardized approach according to the predefined protocol. The disagreement was resolved by consensus or by a discussion with a third author (J.L.).

For efficacy outcomes, we collected data for major adverse cardiovascular events, all-cause deaths, and kidney failure events. Major adverse cardiovascular events were defined as a composite including cardiovascular death, myocardial infarction, ischemic stroke, and heart failure or comparable definitions used by individual authors. Kidney failure events included a 30% decrease in eGFR, doubling of serum creatinine level, or kidney failure, as defined by the authors of each study during the follow-up period. We also collected data on changes in GFR per year, proteinuria, and BP from baseline to the end of follow-up. For safety, considering that we conducted a broad analysis of the effects of urate-lowering drugs, it was difficult to make a summary evaluation of specific adverse reactions, and we recorded the number of all adverse reactions reported in each study and conducted a comprehensive effect evaluation. An adverse event is defined as any unfavorable symptom or sign (including any abnormal laboratory findings) observed in a patient regardless of the relationship to the study treatment.

Data Synthesis and Analyses

All outcome measures were determined using the random effects model in the meta-analysis, as described by DerSimonian and Laird (15), and sensitivity analysis was also performed in the fixed effect model (16). For dichotomous outcomes, the results were expressed as risk ratios (RRs) with 95% confidence intervals (95% CIs). For continuous outcomes, differences were calculated by weighted mean differences.

We estimated the percentage of variability across studies attributable to heterogeneity beyond chance using the I² statistic. Publication bias for the efficacy outcome was assessed using the Egger test. Sensitivity analyses were conducted by the removal of any study to assess the stability of our meta-analytic findings. We explored potential heterogeneity by comparing summary results obtained from subsets of studies grouped by type of urate-lowering drug, duration of follow-up, type of controls, study quality, ethnic regions, or baseline characteristics of the enrolled population, including baseline kidney function and presence or absence of gout, as well as the percentage reduction in serum urate. Between-subgroup heterogeneity was assessed by the chi-squared test (17). A metaregression analysis was performed to examine the relationship between serum urate reduction and efficacy outcomes. A two-sided P<0.05 was considered to be statistically significant, and statistical analyses were performed using RevMan version 5.0.16 (The Nordic Cochrane Centre; Cochrane Collaboration) and STATA version 14.0 (Stata Corp. LP).

Results

Search Results and Characteristics of Selected Studies

A total of 12,296 records were identified by searching electronic databases. After duplicates were removed and titles and abstracts were screened, we retrieved 276 full-text articles for further assessment. Finally, 28 trials (11–13,18–42) enrolling a total of 6458 participants fulfilled the inclusion criteria of this meta-analysis. The identification process for eligible studies is shown in Figure 1.

Figure 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow diagram showing the selection of studies.

The baseline and critical characteristics of the included studies are shown in Table 1 and Supplemental Table 2. Eleven trials enrolled participants with hyperuricemia or gout, one trial recruited people with normal serum urate, and the remaining trials recruited participants irrespective of their serum urate levels. Fifteen trials enrolled 2476 (38%) participants with CKD, and six trials recruited 965 (15%) participants with cardiovascular disease or stroke. The mean age was 60 years (SD of 9 years), and 71% were men. Baseline serum urate was 8.1 mg/dl (SD of 1.5 mg/dl). Overall, five different urate-lowering agents were studied, of which allopurinol or oxypurinol was studied in 16 trials, febuxostat was studied in nine trials, topiroxostat was studied in one trial, and nonxanthine oxidase inhibitors (lesinurad or pegloticase) were studied in three trials. Nineteen trials were placebo controlled, and there was no study medication in the control arm of the remaining ten trials. The mean follow-up was 68 weeks (median, 48 weeks; interquartile range, 24–96 weeks), and it ranged from 6 to 84 months. The percentage reduction in serum urate concentration was reported in 24 trials, and the mean percentage reduction is 34% (median, 38%; interquartile range, 29%–43%), ranging from 5% to 49%. The weighted mean difference in serum urate levels during follow-up between active and control treatments across all trials was −2.53 mg/dl (95% CI, −2.81 to −2.26) (Supplemental Figure 1).

Table 1.

Characteristics of included trials and patients

Study	Inclusion Criteria	No.	Treatment Group, mg/d	Control Group	Follow-Up Time	CKD Participants, no. (%)	Age, yr, mean ± SD	Reduction in Urate, %
PERL 2020	Type 1 diabetes; serum uric acid at least 4.5 mg/dl; eGFR 40.0–99.9 ml/min per 1.73 m2; evidence of diabetic kidney disease	530	Allopurinol: 100–400	Placebo	164 wk	530 (100)	51±11	36
CKD-FIX 2020	Adults with CKD stage 3 or 4; UACR≥265.2 mg/g or decrease in eGFR≥3.0 ml/min per 1.73 m2 in the preceding ≤12 mo; no history of gout	369	Allopurinol: 100–300	Placebo	104 wk	369 (100)	62±13	35
FREED 2019	Age ≥65 yr; hyperuricemia (serum uric acid 7.0–9.0 mg/dl); at risk for cerebral or cardiorenovascular disease	1070	Febuxostat: 10–40	Nonfebuxostat	36 mo	707 (66)	76±7	40
UPWARD 2018	Age 20–75 yr; diabetic kidney disease; gout or hyperuricemia; UACR 45–300 mg/g; eGFR≥30 ml/min per 1.73 m2	65	Topiroxostat: 40–160	Placebo	28 wk	65 (100)	61±10	41
FEATHER 2018	Age ≥20 yr; hyperuricemia (serum uric acid 7.0–10.0 mg/dl); stage 3 CKD; no history of gout	441	Febuxostat: 10–40	Placebo	108 wk	443 (100)	65±12	46
Mukri et al. (20)	Age 18–75 yr; CKD (eGFR 15–60 ml/min per 1.72 m2); HbA1c<8%; asymptomatic hyperuricemia (serum uric acid ≥400 µmol/L); on optimal tolerated dose of antiproteinuric agents	93	Febuxostat: 40	No treatment	6 mo	93 (100)	66±8	39
LIGHT 2017	Age 18–85 yr; body mass index <45 kg/m2; gout	214	Lesinurad: 400	Placebo	6 mo	126 (59)	54±12	29
Golmohammadi et al. (22)	Age ≥18 yr; serum uric acid ≥6 mg/dl; CKD (eGFR 15–60 ml/min per 1.73 m2)	196	Allopurinol: 100	Placebo	12 mo	196 (100)	N/A	21
Dalbeth et al. (23)	Age ≥18 yr; serum uric acid ≥7.0 mg/dl; gout	314	Febuxostat: 40–80	Placebo	24 mo	225 (72)	51±12	35
Xiao et al. (24)	Age 43–71 yr; class II–III heart failure; normal serum uric acid	125	Allopurinol: 300	Standard drug therapy	18 mo	0	52±14	5
Saag et al. (25)	Men (age ≥18 yr) or postmenopausal women (age ≥45 yr); gout; serum uric acid ≥7.0 mg/dl; BP<160/95 mm Hg; eGFR 15–50 ml/min per 1.73 m2	95	Febuxostat: 40–80	Placebo	12 mo	95 (100)	66±11	44
Beddhu et al. (26)	Serum uric acid ≥5.5 mg/dl in men and ≥4.6 mg/dl in women; adults with type 2 diabetes and kidney disease	80	Febuxostat: 80	Placebo	24 wk	80 (100)	68±10	46
Tani et al. (27)	Serum uric acid ≥7.0 mg/dl; hypertensive outpatients	70	Febuxostat: starting at 10	Nonfebuxostat	6 mo	26 (37)	67±12	29
Sircar et al. (28)	Age 18–65 yr; serum uric acid ≥7 mg/dl; eGFR 15–60 ml/min per 1.73 m2	93	Febuxostat: 40	Placebo	6 mo	93 (100)	57±15	42
Liu et al. (29)	Age <70 yr; asymptomatic hyperuricemia (serum uric acid 7.0–8.0 mg/dl); T2DM; diabetic kidney disease (UAER<20 μg/min)	152	Allopurinol: starting at 100	Conventional treatments	3 yr	152 (100)	50±11	32
Goicoechea et al. (30)	Age ≥18 yr; CKD (eGFR 15–60 ml/min per 1.73 m2); serum uric acid ≥6 mg/dl	113	Allopurinol: 100	Standard treatment	84 mo	113 (100)	72±10	15
EXACT-HF 2015	Symptomatic heart failure due to left ventricular systolic dysfunction; serum uric acid ≥9.5 mg/dl; at least one additional high-risk marker	253	Allopurinol: 300–600	Placebo	24 wk	N/A	63±16	44
Yood et al. (32)	Age ≥18 yr; hyperuricemic (serum uric acid ≥8 mg/dl); symptomatic gout; conventional therapy is contraindicated or has been ineffective; CKD	103	Pegloticase: 8 (mg/2–4 wk)	No treatment	86 wk	103 (100)	62	N/A
Higgins et al. (33)	Aged ≥18 yr; ischemic stroke or transient ischemic attacks	69	Allopurinol: 300	Placebo	1 yr	0	68±10	28
Szwejkowski et al. (34)	Type 2 diabetes mellitus; BP<150/90 mm Hg; left ventricular hypertrophy	59	Allopurinol: 600	Placebo	9 mo	N/A	65±9	45
Rekhraj et al. (35)	Hospital cardiopathy; BP<150/90 mm Hg; left ventricular hypertrophy	60	Allopurinol: 600	Placebo	9 mo	0	65±7	46
Shi et al. (36)	Hyperuricemia (serum uric acid ≥7 mg/dl in men or ≥6 mg/dl in women); IgA nephropathy	40	Allopurinol: 100–300	Usual therapy	6 mo	N/A	40±10	28
Sundy et al. (37)	Age ≥18 yr; hyperuricemic (serum uric acid ≥8 mg/dl); symptomatic gout; conventional therapy is contraindicated or has been ineffective	212	Pegloticase: 8 (mg/2–4 wk)	Placebo	25 wk	60 (24)	47±16	38
Kao et al. (38)	CKD 3; left ventricular hypertrophy	53	Allopurinol: 300	Placebo	9 mo	53 (100)	72±7	41
Schumacher et al. (39)	Age 18–85 yr; gout; hyperuricemia (serum uric acid ≥8.0 mg/dl); normal or impaired kidney function	1072	Febuxostat: 80–240/allopurinol: 300	Placebo	28 wk	40 (4)	52±13	49
OPT-CHF 2008	Age 18–85 yr; symptomatic heart failure; left ventricular ejection fraction ≤40%	405	Oxypurinol: 600	Placebo	24 wk	N/A	65±13	26
Siu et al. (41)	Kidney disease; hyperuricemic (serum uric acid ≥7.60 mg/dl)	51	Allopurinol: 100–300	Usual therapy	12 mo	51 (100)	48±17	40
Gibson et al. (42)	Primary gout; treated with colchicine 0.5 mg twice per day	59	Allopurinol: 200	No allopurinol	2 yr	N/A	49±12	30

PERL, Preventing Early Renal Function Loss; CKD-FIX 2020, Controlled trial of slowing of Kidney Disease progression From the Inhibition of Xanthine oxidase; UACR, urinary microalbumin-creatinine ratio; FREED 2019, Febuxostat for Cerebral and CaRdiorenovascular Events PrEvEntion StuDy; UPWARD 2018, Uric acid-lowering and renoprotective effects of topiroxostat, a selective xanthine oxidoreductase inhibitor, in patients with diabetic nephropathy and hyperuricemia: a randomized, double-blind, placebo-controlled, parallel-group study; FEATHER 2018, Febuxostat Versus Placebo Randomized Controlled Trial Regarding Reduced Renal Function in Patients With Hyperuricemia Complicated by Chronic Kidney Disease Stage 3; HbA1c, hemoglobin A1C; LIGHT 2017, Lesinurad Monotherapy in Gout Subjects Intolerant to Xanthine Oxidase Inhibitors study; N/A, not reported; T2DM, type 2 diabetes mellitus; UAER, urinary albumin excretion rate; EXACT-HF 2015, Xanthine Oxidase Inhibition for Hyperuricemic Heart Failure Patients trial; OPT-CHF 2008, Efficacy and Safety Study of Oxypurinol Added to Standard Therapy in Patients with New York Heart Association Class III-IV Congestive Heart Failure.

Efficacy Outcomes

Major Adverse Cardiovascular Events.

A total of 15 trials with 5327 participants and 506 major adverse cardiovascular events were recorded, and urate-lowering therapy did not reduce the risk of major adverse cardiovascular events (RR, 0.93; 95% CI, 0.74 to 1.18; heterogeneity I² =33%) (Figure 2) compared with placebo or no study medication.

Figure 2.

Urate-lowering therapy did not show benefits on major adverse cardiovascular events, all-cause mortality, or kidney failure events. The random effects model was used for the effect, and a fixed effects analysis is also shown. 95% CI, 95% confidence interval.

In the subgroup analysis of major adverse cardiovascular events, there seemed to be no apparent heterogeneity of effects between all different subgroups (Figure 3). We did not find statistically significant subgroup heterogeneity between different percentage reduction in serum urate (P=0.57) (Figure 3). Unadjusted regression did not show a clear linear relationship between urate reduction and cardiovascular benefit (P=0.69).

Figure 3.

No apparent heterogeneity of effects between all different subgroups was found in the subgroup analysis of major adverse cardiovascular events. ^aHyperuricemia is defined by urate concentrations in baseline above 7.0 mg/dl (420 µmol/L) as measured by automated enzymatic (uricase) laboratory methods. ^bPatient populations with CKD were determined according to whether the CKD group was included in the inclusion criteria. ^cWhite race is defined as >70% White people of the population. Two studies (Xanthine Oxidase Inhibition for Hyperuricemic Heart Failure Patients [EXACT-HF] trial and Higgins et al. [33]) were not analyzed in subgroups of races because no one particular race was >70% of the population or there were unavailable characteristics of the population. ^dOne year is the median follow-up time of all included trials. ^eThe median percentage reduction of serum uric acid in the treatment arm of all included trials is 38%.

All-Cause Mortality.

Thirteen trials with 4228 participants reported 164 deaths events, and there was no statistically significant difference for all-cause mortality in the urate-lowing therapy intervention group relative to placebo or usual care (RR, 1.04; 95% CI, 0.78 to 1.39; heterogeneity I² =0%) (Figure 2).

Kidney Outcomes

Overall, there were eight trials with 3087 participants, and 266 kidney failure events were recorded, which indicated that urate-lowing therapy had no significant effect (RR, 0.97; 95% CI, 0.61 to 1.54; heterogeneity I² =52%) (Figure 2) on preventing kidney failure events. There was no statistically significant interaction with the follow-up time (P=0.16) or CKD progression risk (P=0.28) (Supplemental Figure 2).

Urate-lowering therapy attenuated the decline in the slope of GFR compared with the control group (19 trials, 3934 participants; weighted mean difference, 1.18 ml/min per 1.73 m² per year; 95% CI, 0.44 to 1.91; heterogeneity I² =69%) (Figure 4). Subgroup analysis showed that the benefits were more obvious in trials with short follow-up (<2 years; P=0.002) (Supplemental Figure 3) and trials without high risk of bias (P=0.002; most of them were with shorter follow-up time or were not analyzed as intention to treat) (Supplemental Figure 3). Proteinuria reduction was also evaluated, and no benefit of urate-lowing therapy was observed in urinary microalbumin-creatinine ratio (weighted mean difference, −8.05 mg/g; 95% CI, −29.39 to 13.30), urinary albumin excretion rates (weighted mean difference, −1.34 μg/min; 95% CI, −13.93 to 11.25), urinary proteinuria-creatinine ratio (weighted mean difference, 0.56 mg/mmol; 95% CI, −13.92 to 15.03), or proteinuria (weighted mean difference, −0.10 g/d; 95% CI, −0.89 to 0.69) (Figure 4).

Figure 4.

Urate-lowering therapy attenuated the decline in the slope of GFR (milliliters per minute per 1.73 m² per year) and lowered the mean BP (millimeters of mercury), but had no effect on urinary microalbumin-creatinine ratio (UACR; milligrams per gram), urinary albumin excretion rates (UAERs; micrograms per minute), urinary proteinuria-creatinine ratio (UPCR; milligrams per millimole), or proteinuria (grams per day). A random effects model and weighted mean difference were used to pool the effect.

Blood Pressure

The significant antihypertensive effect of urate-lowering therapy was observed in systolic BP (13 trials, 2179 participants; weighted mean difference, −3.45; 95% CI, −6.10 to −0.80; heterogeneity I² =94%) (Figure 4) and diastolic BP (13 trials, 2179 participants; weighted mean difference, −2.02; 95% CI, −3.25 to −0.78; heterogeneity I² =50%) (Figure 4). Subgroup analysis demonstrated that the effect size was much more in trials with hyperuricemic participants (P=0.02) or trials of Asians (P=0.04) (Supplemental Figure 4). Unadjusted regression showed that urate-lowering therapy produced the most benefit when urate concentrations were improved (P=0.05) (Supplemental Figure 5).

Adverse Effects

All reported adverse events were summarized, and no significant difference (21 trials, 5511 participants; RR, 1.01; 95% CI, 0.94 to 1.08; heterogeneity I² =36%) (Supplemental Figure 6) in risk was observed in the urate-lowering participants. Potential harms of treatment are attributed to adverse drug reactions, mostly including skin rash, arthralgia, gastrointestinal symptoms, and elevation of liver function enzyme.

Study Quality, Sensitivity Analyses, and Publication Bias

The reported trial quality varied substantially (Supplemental Figures 7 and 8). Among included trials, 64% adequately generated their randomization sequence, 43% adequately concealed allocation, 68% blinded patients and caregivers, and 89% completely reported results. However, detection bias (75%) and other bias (29%) were unclear or high. The overall results of sensitivity analysis were not changed significantly after excluding every single study (Supplemental Figure 9). Formal statistical testing illustrated that there was no evidence of publication bias (Supplemental Figure 10).

Discussion

Individuals with hyperuricemia are at higher risk of cardiovascular events and progression of CKD, but, to date, the effect of urate lowering has not been established. In this sizeable quantitative overview, including 28 trials with 6458 individuals recording 506 cardiovascular events, 164 deaths, and 266 kidney failure events, we showed that urate-lowering therapy did not reduce the risk of major cardiovascular events, death, or kidney failure. Although urate-lowering therapy improved the GFR decline slope and lowered the BP levels, these effects did not translate to clinical outcomes in the follow-up, including kidney or cardiovascular events. Thus, there is insufficient evidence to support that urate lowering improves kidney or cardiovascular outcomes, including in CKD.

Hyperuricemia has been a concern since 1965, when it was shown to be closely related to coronary artery disease and hypertension (43). Many observational studies have shown that patients with gout have a higher risk of cardiovascular death, whereas patients with cardiovascular or kidney disease mostly have elevated serum urate levels and a higher risk of gout (44). The association of serum urate between cardiovascular and kidney diseases and the possible benefits of urate-lowering therapy have been explored over the past decades. To date, the role of urate-lowering therapy in diseases other than hyperuricemia has not been established. A recent Cochrane meta-analysis, including 1187 participants, showed insufficient evidence to support an effect on cardiovascular markers by urate-lowering therapy (10). The Febuxostat for Cerebral and CaRdiorenovascular Events PrEvEntion StuDy included 1070 participants with a high risk of cardiovascular disease and found that febuxostat treatment significantly reduced major cardiovascular and kidney events (hazard ratio, 0.75; 95% CI, 0.59 to 0.95), with cardiovascular and kidney events included. However, the most frequent event consituted was kidney proteinuria progression, which compromised the overall effect of urate-lowering therapy on cardiovascular or kidney failure (13). In this systematic review with >6000 participants, we demonstrated that the use of urate-lowering therapy did not reduce the risk of cardiovascular events, death, and kidney failure. Although some studies reported potential clinical benefits of kidney or cardiovascular benefits, most were limited to trials with short follow-up or low quality. Trials with high quality, including the newly released Preventing Early Renal Function Loss (PERL) study and the Controlled trial of slowing of Kidney Disease progression From the Inhibition of Xanthine oxidase (CKD-FIX), failed to show any benefits on the kidney progression or cardiovascular outcomes.

Hyperuricemia is common in people with CKD and prevails with the worsening of kidney function. Many epidemiologic studies have demonstrated that high serum uric acid concentrations are independently associated with the development of new-onset CKD or the progression to kidney failure (6). Although several trials have evaluated the effect of urate lowering, the KDIGO guideline suggested that there is insufficient evidence to support or refute the use of agents to lower serum urate concentrations in people with CKD to delay the progression of CKD. This meta-analysis with 3934 participants suggests that urate-lowering therapy could slow the decline of GFR, but this effect is mainly driven by the trials with short follow-up or low quality. The two recent trials, CKD-FIX with 369 patients with stage 3 or 4 CKD and the PERL study with 530 individuals with type 1 diabetes and early to moderate kidney disease, did not show benefits on the decline rate of GFR after >2 years of allopurinol therapy. Our analysis, including 266 kidney failure events, showed that urate-lowering therapy did not reduce the risk of kidney failure events (RR, 0.97; 95% CI, 0.61 to 1.54). Taken together, there is insufficient evidence to support the use of urate-lowering agents to prevent kidney progression in patients with kidney disease.

In this study, we have shown moderate benefits of urate-lowering therapy on BP control. However, this BP-lowering effect could not translate to clinical benefits of cardiovascular events—although this inconsistency might be due to a short follow-up in most trials that was hard for observing the effect for major cardiovascular events. The two recent high-quality trials, the PERL study and CKD-FIX, did not find any effect on BP control. Thus, it is still unclear whether the antihypertension effect found in this study could be biased. It still needs a large, high-quality trial with a long-term follow-up to evaluate the effect of urate-lowering agents on these clinical outcomes.

Our study provides comprehensive evidence of urate-lowering therapy for cardiovascular and kidney outcomes in participants with hyperuricemia. The research benefits from a large volume of data that was able to be included and the rigorous methodology used. Our study does, however, have several limitations. Importantly, the overall follow-up time of included studies in this overview was limited, which reduced our review's power to evaluate the hard end point including cardiovascular or kidney failure events. Also, the trials involved in this meta-analysis have significant heterogeneity related to the level of kidney function, underlying disease, and other conditions. Short-term trials or those that showed only minor progression in the control group could hinder the ability to detect a difference. Finally, there was large and significant variability in the quality of the trials, and also the usage of renin-angiotensin-aldosterone system inhibitors in these studies. Even CKD-FIX had a significant dropout rate (25%–30%) that could have confounded results. Large and well-controlled trials with long-term follow-up are still needed.

In conclusion, this meta-analysis suggests that urate-lowering therapy was associated with the reduction of BP and slowing the decline of GFR. Nevertheless, the therapy did not produce benefits on clinical outcomes, including major adverse cardiovascular events, all-cause mortality, and kidney failure. Thus, there is insufficient evidence to support urate lowering in patients with hyperuricemia, including in patients with kidney disease, and more well-controlled trials with long-term follow-up are needed.

Disclosures

S.V. Badve reports receiving personal fees from Amgen Australia, personal fees and nonfinancial support from Bayer Australia (fees paid to institution), and personal fees from Pfizer Australia (fees paid to institution), outside the submitted work. All remaining authors have nothing to disclose.

Funding

This work was funded by National Key Research and Development Program of China grant 2018YFC1314004. S.V. Badve reports receiving grants from the National Health and Medical Research Council of Australia (APP1043203) during the conduct of the study. J. Lv was supported by National Natural Science Foundation of China grants 81270795 and 81874401.

Supplemental Material

This article contains the following supplemental material online at http://cjasn.asnjournals.org/lookup/suppl/doi:10.2215/CJN.05190420/-/DCSupplemental.

Supplemental Figure 1. Forest plot showing the effects of urate-lowering or nonurate-lowering regimens on the concentrations of serum urate.

Supplemental Figure 2. The subgroup analysis for the effects of urate-lowering therapy on kidney failure events.

Supplemental Figure 3. Subgroup analysis of follow-up time for the effects of urate-lowering therapy on GFR.

Supplemental Figure 4. The subgroup analysis for the effects of urate-lowering therapy on mean arterial pressure.

Supplemental Figure 5. A clear relationship between reduction in concentration of uric acid and the antihypertensive effects of urate-lowering therapy was observed.

Supplemental Figure 6. The effects of urate-lowering or nonurate-lowering regimens on adverse events.

Supplemental Figure 7. Risk of bias graph showing review authors’ judgments about each risk of bias item presented as percentages across all included studies.

Supplemental Figure 8. Risk of bias summary showing review authors’ judgments about each risk of bias item for each included study.

Supplemental Figure 9. Sensitivity analysis by omitting each study in turn.

Supplemental Figure 10. Funnel plots for publication bias.

Supplemental Table 1. Search strategy for data sources.

Supplemental Table 2. Characteristics of included trials and patients.