Abstract
Background
Whether a treat-to-target urate management strategy improves kidney outcomes in patients with chronic kidney disease (CKD) and asymptomatic hyperuricemia initiating urate-lowering therapy (ULT) remains uncertain, despite current guidelines.
Methods
This target trial emulation (TTE) study included 2092 adults with stage G3–G4 CKD and asymptomatic hyperuricemia who initiated ULT at Zhejiang Provincial People’s Hospital between 2018 and 2024. Using a clone-censor-weight (CCW) approach to address biases, we compared a treat-to-target strategy (targeting serum urate <360 μmol/L within 183 days) versus a non-treat-to-target strategy. Patients were followed for up to 3 years. The primary outcome was a composite kidney outcome: end-stage kidney disease (ESKD) or ≥40% eGFR decline.
Results
The 3-year absolute risk of composite kidney outcome was 33.3% (95% CI, 30.1%–36.6%) under the treat-to-target strategy and 35.8% (95% CI, 30.7%–40.5%) under the non-treat-to-target strategy, yielding a risk difference of −2.5% (95% CI, −6.5% to 2.5%) and a risk ratio of 0.93 (95% CI, 0.83–1.08). Subgroup analyses revealed no heterogeneity, and sensitivity analyses yielded consistent results.
Conclusion
A treat-to-target urate management strategy was not associated with a reduced risk of composite kidney outcomes compared to a non-treat-to-target strategy. Among patients with stage G3-G4 CKD and asymptomatic hyperuricemia in whom clinicians decide to start ULT, these findings do not support pushing to achieve this intensive urate target for improving renal outcomes. As findings are largely based on a febuxostat-dominant real-world setting, extrapolation to other ULT agents may be limited.
Keywords: Urate-lowering therapy, treat-to-target strategy, chronic kidney disease, asymptomatic hyperuricemia, trial emulation
1. Introduction
Chronic kidney disease (CKD) prevention has emerged as a critical public health priority worldwide. Although numerous observational studies have identified an association between hyperuricemia and the onset and progression of CKD [1–3], evidence from randomized clinical trials (RCTs) has not consistently supported this relationship. The CKD-FIX trial [4], which enrolled patients with stage G3-G4 CKD, and the PERL study [5], which included patients with stage G1-G3 CKD and type 1 diabetes, both failed to demonstrate renoprotective benefits of allopurinol. Similarly, the FEATHER trial [6], conducted among patients with stage G3-G4 CKD and asymptomatic hyperuricemia, showed no beneficial effect of febuxostat on slowing CKD progression. Consequently, the 2024 Kidney Disease: Improving Global Outcomes (KDIGO) guidelines remain inconclusive regarding the recommendation of urate-lowering therapy (ULT) for the purpose of delaying adverse renal outcomes [7].
In contrast, several RCTs conducted in China have suggested that ULT may improve renal outcomes in patients with asymptomatic hyperuricemia and coexisting CKD [8,9]. The 2019 Chinese guidelines recommended initiating ULT in patients with asymptomatic hyperuricemia accompanied by kidney disease [10]. More recently, the 2024 Update of Chinese Guidelines for Diagnosis and Treatment of Hyperuricemia and Gout further advised that patients with CKD and asymptomatic hyperuricemia should achieve a target serum uric acid (SUA) level below 6 mg/dL (360 μmol/L) [11,12]. Febuxostat, currently endorsed as a first-line ULT agent for patients with advanced CKD [11], has demonstrated favorable efficacy and safety profiles and is widely adopted by nephrologists in China. Prior evidence has indicated that febuxostat confers superior renoprotective effects compared with allopurinol in patients with stage G3a CKD and asymptomatic hyperuricemia [13]. Although ULT is an established intervention for gout prevention and management, whether intensive urate-lowering strategies improve renal outcomes, particularly among individuals with preexisting kidney impairment, remains uncertain. The optimal target SUA concentration for maximal renoprotection has yet to be determined. A recent target trial emulation (TTE) study supported optimizing ULT to achieve the target serum urate level (TSUL) in patients with gout and stage G3 CKD [14], whereas a multicenter RCT from Japan failed to demonstrate that intensive ULT improved albuminuria or estimated glomerular filtration rate (eGFR) in patients with CKD [15]. Given the limited relevant evidence from Chinese populations, further investigation is warranted.
TTE is a methodological approach that constructs an RCT-like framework from observational data, simulating an ideal trial design that could be implemented in practice. By explicitly specifying the time configuration, eligibility criteria, intervention strategies, outcome definitions, and analytical protocols, TTE aims to minimize biases such as immortal time bias and prevalent user bias, thereby enhancing the interpretability and causal validity of observational findings [16,17].
While our previous study evaluated the initial clinical decision of whether to start ULT [18], the present study addresses the subsequent sequence in clinical decision-making: whether to pursue a strict treat-to-target strategy among patients who have already initiated ULT. Therefore, we conducted this study using the TTE framework and data from Zhejiang Provincial People’s Hospital to emulate an RCT evaluating the effect of a treat-to-target urate management strategy on kidney outcomes among patients with CKD and asymptomatic hyperuricemia who initiated ULT.
2. Methods
2.1. Data source
This study utilized data obtained from the electronic health records of Zhejiang Provincial People’s Hospital. Ethical approval was granted by the Ethics Committee of Zhejiang Provincial People’s Hospital (Approval Number: QT2025212). The study was conducted in accordance with the principles of the Declaration of Helsinki. Written informed consent was obtained from all participants for the use of their clinical data in this research. All personal identifiers have been removed to protect patient privacy.
2.2. Study design
We applied a TTE framework to examine the effect of a treat-to-target urate management strategy (aiming for a TSUL <360 μmol/L within 183 days of ULT initiation) on the risk of a composite kidney outcome among patients with CKD and asymptomatic hyperuricemia.
To address immortal time bias and prevalent user bias, we implemented the clone-censor-weight (CCW) method within a new-user design [17,19]. Accordingly, the analysis was restricted to patients receiving their first recorded ULT prescription. A visual representation of the CCW method is shown in Figure S1, which includes five illustrative examples of the cloning and censoring process. The study design schema, including definitions of eligibility, treatment strategy, follow-up, and endpoints, is presented in Table S1.
2.2.1. Clone
Every eligible patient was cloned at time zero (defined as the initiation of ULT). Each patient’s clones were then simultaneously assigned to both emulated treatment arms: the treat-to-target strategy and the non-treat-to-target strategy. This procedure ensures that the two strategy cohorts are exactly identical regarding all baseline characteristics.
2.2.2. Censor
Following the generation of cloned datasets, artificial censoring was applied, meaning that clones who deviated from their assigned treatment protocol were censored. Specifically, clones in the treat-to-target arm who failed to reach TSUL within grace period (183 days) and clones in the non-treat-to-target arm who attained TSUL within this period were censored at the time of deviation. Importantly, patients who experienced outcome events, loss to follow-up, or death during the 183-day grace period prior to TSUL attainment were retained in both treatment arms until the time of the event. Because their clinical trajectory up to that point was compatible with both strategies, this prospective handling ensures that group assignment is not conditioned on future events, thereby explicitly preventing immortal time and selection biases.
2.2.3. Weight
To address selection bias from artificial censoring, inverse probability of censoring weights (IPCW) were applied. Weights were estimated using pooled logistic regression models fitted at daily intervals, incorporating baseline and time-varying covariates as well as a cubic term for time, and truncated at the 95th percentile to limit the influence of extreme values. Applying these truncated weights yielded a balanced pseudo-population that effectively recovered the causal estimand.
2.3. Eligibility criteria
Eligible participants were adults aged ≥18 years who initiated ULT between January 1, 2018, and December 31, 2024, with documented diagnoses of both stage G3–G4 CKD and hyperuricemia (serum uric acid [SUA] > 420 μmol/L) prior to enrollment. Stage G3–G4 CKD was defined as an eGFR of 15–60 mL/min/1.73 m2 documented on at least two occasions separated by >90 days within a one-year period, or the presence of at least one ICD-10 diagnostic code for CKD stage 3 or 4. The index date was defined as the date of the first ULT prescription (including febuxostat, benzbromarone, and allopurinol) following the hyperuricemia and CKD diagnosis. Patients were excluded if they had missing data for SUA or eGFR, or a documented history of gout, end-stage kidney disease (ESKD), or cancer prior to the index date. Following application of these criteria, a total of 2092 patients were included in the final analytical cohort (Figure 1).
Figure 1.
Flow diagram of the included patients.
2.4. Treatment strategies
We compared two urate management strategies in patients with stage G3–G4 CKD and asymptomatic hyperuricemia. The ‘treat-to-target strategy’ arm was defined as attaining TSUL within a 183-day grace period after ULT initiation. Patients who did not achieve TSUL during the grace period were considered the ‘non-treat-to-target strategy’ arm.
2.5. Follow-up and outcomes
Follow-up commenced on the index date and continued until the occurrence of an outcome event, loss to follow-up, death, the end of the study period (December 31, 2024), or three years from baseline, whichever occurred first. The primary outcome was a composite kidney endpoint, defined as progression to ESKD or a ≥ 40% sustained decline in eGFR relative to baseline. ESKD was defined as an eGFR <15 mL/min/1.73 m2 or initiation of kidney replacement therapy (KRT).
2.6. Covariates
Selection of baseline and time-varying covariates for adjustment was guided by directed acyclic graphs (DAGs) constructed to elucidate presumed causal relationships (Figure S2). Baseline covariates included demographic characteristics (age and sex), lifestyle factors (smoking tobacco and drinking alcohol), comorbidities (atrial fibrillation, chronic obstructive pulmonary disease [COPD], dementia, depression, hypertension, diabetes, fracture, myocardial infarction, osteoporosis, pneumonia or infection, stroke, and varicose veins), medications (antidiabetic, antihypertensive, and anticoagulant agents, nonsteroidal anti-inflammatory drugs [NSAIDs], aspirin, nitrates, and diuretics [including loop, potassium-sparing, and thiazide diuretics]), and healthcare utilization (outpatient visits, hospitalizations, and emergency encounters). Laboratory values (SUA and eGFR) were obtained from the most recent measurements within 3 months before the index date. All other covariates were ascertained during the 12 months preceding the index date. The eGFR was calculated from serum creatinine using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation [20]. Time-varying covariates, updated throughout follow-up, included lifestyle factors, comorbidities, and medication use.
2.7. Statistical analysis
Baseline and time-varying covariates before and after weighting were summarized as mean (SD) for continuous variables and frequencies (percentages) for categorical variables. Standardized mean differences (SMDs) were calculated to evaluate covariate balance between treatment groups. An SMD <0.15 was considered indicative of adequate balance [21,22].
The effect of the treat-to-target strategy on composite kidney outcome was estimated using weighted pooled logistic regression, adjusted for baseline and time-varying covariates as well as a cubic term for time. This approach has been demonstrated to yield hazard ratio (HR) estimates comparable to those obtained from Cox proportional hazards models [23,24]. Predicted probabilities derived from this model were used to generate cumulative incidence curves for each treatment strategy. 3-year absolute risks, risk differences, and risk ratios with corresponding 95% confidence intervals (CIs) were calculated using nonparametric bootstrapping with 1000 resamples. HRs were also approximated from odds ratios obtained from the pooled logistic regression. Statistical significance was defined as a 95% CI for the HR excluding 1 or a 95% CI for the risk difference excluding 0.
To address missing data, variables exceeding 20% missingness were excluded. Body mass index (BMI) was removed from analysis due to high missing rates (51.7%), as weight was typically measured only during hospitalizations. Smoking and alcohol consumption, with missing rates of 14.1% and 13.9% respectively, were included using a ‘missing’ category.
2.7.1. Subgroup and sensitivity analyses
Subgroup analyses were stratified by age, sex, UA, eGFR, and diabetes to assess for potential effect modification. Additionally, seven sensitivity analyses were performed to test the robustness of the results. In the first two sensitivity analyses, the grace period was varied to 90 and 365 days to examine whether the association between the treat-to-target strategy and kidney outcome remained consistent across different time windows. Third, analyses were repeated using untruncated IPCW weights. Fourth, a more stringent serum urate target of 300 μmol/L was applied. Fifth, to mitigate concerns regarding reverse causation, analyses were replicated in patients with stage G2-G4 CKD and asymptomatic hyperuricemia. Sixth, a positive control analysis was undertaken to examine the association between urinary albumin-to-creatinine ratio (UACR) and kidney outcome. We expected to observe the well-established association between elevated UACR and adverse kidney outcome. Failure to detect this relationship would suggest residual confounding by unmeasured factors, thereby undermining the credibility of the primary findings. For this analysis, patients were classified according to the maximum UACR level during the grace period into A1/A2 (normal to moderately increased) or A3 (severely increased) categories. Crucially, this positive control analysis strictly applied the identical CCW framework and IPCW adjustments used in the primary analysis to ensure methodological consistency. Seventh, given the distinct pharmacological mechanisms of different ULT agents, we conducted an additional sensitivity analysis restricted entirely to patients initiating febuxostat (the predominant agent in our cohort) to evaluate whether ULT class heterogeneity influenced our findings.
Data extraction and management were performed using Navicat Premium 17. All statistical analyses were conducted using R version 4.3.3 (R Foundation for Statistical Computing).
3. Results
3.1. Patient characteristics
The primary analysis included 2092 participants (mean [SD] age, 70.5 [15.7] years; 1347 men [64.4%] and 745 women [35.6%]). The mean (SD) serum urate and eGFR values were 562.2 (105.0) μmol/L and 41.3 (11.8) mL/min/1.73 m2, respectively. Hypertension (79.1%) and diabetes (35.9%) were the most prevalent comorbidities recorded at baseline. Regarding concomitant medications, 1477 patients (70.6%) received antihypertensive agents and 823 (39.3%) received loop diuretics (Table 1). Among the 2092 ULT initiators, febuxostat was the most frequently prescribed agent (1852 [88.5%]), followed by benzbromarone (220 [10.5%]) and allopurinol (20 [1.0%]). The median (IQR) time from the index date to TSUL achievement was 22 (5–71) days, with 1370 patients (65.5%) attaining TSUL within the 183-day grace period. Following IPCW adjustment, baseline and time-varying characteristics were well balanced between groups (Tables S2 and S3), with the distribution of IPCW weights presented in Table S4. Figure 2 illustrates mean SUA trajectories during follow-up, stratified by treatment strategy (treat-to-target vs. non-treat-to-target).
Table 1.
Baseline characteristics of included patients.
| Characteristic | Study population (N = 2092) |
|---|---|
| Demographic information | |
| Age, Mean (SD), y | 70.5 (15.7) |
| Female, n (%) | 745 (35.6) |
| Male, n (%) | 1347 (64.4) |
| Lifestyle factors | |
| Smoking tobacco, n (%) | |
| No | 1284 (61.4) |
| Yes | 513 (24.5) |
| Missing | 295 (14.1) |
| Drinking alcohol, n (%) | |
| No | 1371 (65.5) |
| Yes | 431 (20.6) |
| Missing | 290 (13.9) |
| Serum urate, Mean (SD), μmol/L | 562.2 (105.0) |
| eGFR, Mean (SD), mL/min/1.73 m2 | 41.3 (11.8) |
| Comorbidity, n (%) | |
| Atrial fibrillation | 333 (15.9) |
| COPD | 75 (3.6) |
| Dementia | 103 (4.9) |
| Depression | 30 (1.4) |
| Hypertension | 1655 (79.1) |
| Diabetes | 752 (35.9) |
| Fracture | 108 (5.2) |
| Myocardial infarction | 87 (4.2) |
| Osteoporosis | 287 (13.7) |
| Pneumonia or infection | 399 (19.1) |
| Stroke | 437 (20.9) |
| Varicose veins | 22 (1.1) |
| Medication, n (%) | |
| Antidiabetic medicine | 662 (31.6) |
| Antihypertensive medicine | 1477 (70.6) |
| Anticoagulants | 555 (26.5) |
| NSAIDs | 432 (20.7) |
| Aspirin | 535 (25.6) |
| Nitrates | 439 (21.0) |
| Loop diuretics | 823 (39.3) |
| Potassium-sparing diuretics | 641 (30.6) |
| Thiazide diuretics | 315 (15.1) |
| Health care utilization, Mean (SD)* | |
| Outpatient | 1.4 (2.5) |
| Hospitalization | 1.9 (3.2) |
| Emergency | 0.2 (0.5) |
eGFR: estimated glomerular filtration rate; NSAID: nonsteroidal anti-inflammatory drug.
*Frequency during the past year.
Figure 2.
Time course changes in mean serum urate concentrations after index date. M: month. Note: Mean ± 95% CI is calculated using a two-step method: within-individual averages are computed for each time period, and the overall mean and 95% CI are then calculated from these averages.
3.2. Composite kidney outcome
There were 339 and 181 cases of the composite kidney outcome in the treat-to-target and non-treat-to-target strategy groups, respectively. The 3-year absolute risk was 33.3% (95% CI, 30.1%–36.6%) under the treat-to-target strategy and 35.8% (95% CI, 30.7%–40.5%) under the non-treat-to-target strategy. There was a small reduction in absolute risk difference of −2.5% (95% CI, −6.5% to 2.5%) and risk ratio of 0.93 (95% CI, 0.83–1.08), albeit being statistically non-significant. Similarly, the HR was 0.91 (95% CI, 0.79–1.10), showing high consistency with the risk ratio (Table 2 and Figure 3).
Table 2.
Composite kidney outcome in patients with stage 3–4 CKD and asymptomatic hyperuricemia after clone-censor-weight.
| Treatment | No. of patients | No. of patient-days | No. of outcomes | 3-year absolute risk (%) (95% CI) | 3-year risk difference (%) (95% CI) | 3-year risk ratio (95% CI) | Hazard ratio (95% CI) |
|---|---|---|---|---|---|---|---|
| Composite kidney outcome | |||||||
| Non-treat-to-target strategy | 2092 | 371,178 | 181 | 35.8 (30.7–40.5) | Reference | Reference | Reference |
| Treat-to-target strategy | 2092 | 668,946 | 339 | 33.3 (30.1–36.6) | −2.5 (−6.5 to 2.5) | 0.93 (0.83–1.08) | 0.91 (0.79–1.10) |
CKD: chronic kidney disease.
Figure 3.
Cumulative incidence curve for the composite kidney outcome.
3.3. Subgroup and sensitivity analyses
No significant heterogeneity in the effect of the treat-to-target strategy on the composite kidney outcome was observed across subgroups stratified by age, sex, UA, eGFR, and diabetes (Figure 4). The results of all sensitivity analyses and weighted cumulative incidence curves are presented in Tables S5–S11 and Figure S3, respectively. In sensitivity analyses adjusting the grace period to 90 and 365 days, the 3-year risk differences between the treat-to-target and non-treat-to-target groups were −1.0% (95% CI, −6.5% to 3.7%) and −0.5% (95% CI, −6.2% to 5.6%), respectively. Consistent with the primary analysis, the analysis using untruncated weights yielded a 3-year risk difference of −5.6% (95% CI, −18.8% to 7.6%). Similarly, evaluating a stricter treat-to-target strategy (aiming for a target of 300 μmol/L) resulted in a risk difference of −2.2% (95% CI, −6.2% to 3.1%). In the sensitivity analysis expanding the cohort to include stage G2-G4 CKD patients, the 3-year risk difference was 0.9% (95% CI, −2.3% to 3.4%). In the positive control analysis, elevated UACR levels were significantly associated with an increased risk of the composite kidney outcome, with a 3-year risk difference of 10.3% (95% CI, 3.8%–16.9%) for the UACR A3 group relative to the A1/A2 group. Crucially, to address potential heterogeneity introduced by the distinct mechanisms of different ULT agents, a sensitivity analysis restricted entirely to febuxostat initiators (N = 1852) was conducted. This analysis yielded a 3-year risk difference of −1.1% (95% CI, −4.8% to 4.1%) and an HR of 0.96 (95% CI, 0.84–1.17), demonstrating that our primary findings remained robust even when evaluating a single ULT class.
Figure 4.
Subgroup analysis after using the clone-censor-weight method across categories of age, sex, UA, eGFR, diabetes.
4. Discussion
In this TTE study of a large hospital-based cohort, a treat-to-target urate management strategy (aiming for TSUL <360 μmol/L within 183 days) compared with a non-treat-to-target strategy was not associated with a reduced risk of the composite renal outcome (ESKD or substantial eGFR decline) among patients with stage G3–G4 CKD and asymptomatic hyperuricemia who had initiated ULT. A series of sensitivity analyses yielded consistent findings, including those employing a more stringent serum uric acid target (<300 μmol/L) and those extending the study population to patients with stage G2–G4 CKD. Importantly, while current Chinese guidelines often recommend a target SUA of <360 μmol/L or even a more stringent target of <300 μmol/L for this population, our real-world findings challenge this routine recommendation. These results suggest that among patients in whom clinicians decide to start ULT, pushing to achieve such intensive urate targets does not confer additional renoprotective benefits.
Several factors strengthen the robustness of our findings. First, we conducted a positive control analysis based on the original study design. The results demonstrated that achieving UACR category A2/A3 compared with A1 during the grace period was associated with a 10.3% increased risk of the 3-year composite kidney outcome (95% CI, 3.8%–16.9%). This finding aligns with robust prior epidemiological evidence demonstrating that albuminuria levels are associated with CKD progression over time [25,26], thereby validating that our analytical pipeline (including data extraction, variable definitions, and statistical modeling) possessed adequate sensitivity to detect true signals, which increases confidence in our negative findings. Second, as shown in Figure 2, there was clear separation in SUA levels between the treat-to-target arm and the non-treat-to-target arm during follow-up, with levels fluctuating around 300–360 μmol/L and 420 μmol/L, respectively, thereby mitigating the possibility of false-negative conclusions attributable to insufficient between-group differences in SUA levels. Third, potential confounders influencing kidney function progression, including smoking status, alcohol consumption, hypertension, and diabetes, were well-balanced between groups both at baseline and throughout the post-baseline period. Fourth, 1618 patients (77.3%) achieved the TSUL of less than 360 μmol/L following ULT during the study follow-up period, consistent with clinical expectations. Fifth, multiple sensitivity analyses (including alternative study populations, varying grace periods, and use of untruncated weights) yielded results consistent with the primary analysis. Notably, to address potential heterogeneity introduced by pooling different ULT agents, a sensitivity analysis restricted entirely to febuxostat users showed consistent results, confirming that our findings were not driven by pharmacological differences between xanthine oxidase inhibitors and uricosuric agents.
Our findings align with results from several major RCTs, including the CKD-FIX, PERL, and FEATHER trials [4–6], which demonstrated no significant renoprotective benefit from ULT. Our study extends these observations by applying a TTE framework to a diverse, real-world population of new ULT users. Unlike randomized trials, which often necessitate strict exclusion criteria, our cohort reflects actual clinical practice where treatment initiation is frequently driven by physician judgment regarding disease severity. Consequently, our study population likely bore a higher burden of comorbidities and baseline risk compared with selected participants in prior trials. The persistence of negative findings in this higher-risk, real-world setting provides robust evidence that a treat-to-target urate management strategy does not retard the progression of CKD. Our results are also consistent with those reported by Badve et al. [15], who conducted a randomized trial among patients with CKD stage 3 and asymptomatic hyperuricemia and demonstrated no significant difference in kidney function improvement between intensive and standard urate management groups. However, several important distinctions exist between our study and theirs that may enhance the validity of our conclusions. First, the relatively small between-group difference in SUA concentrations in the Badve trial (change of −2.10 mg/dL in the standard group vs −2.90 mg/dL in the intensive group) may have limited the study’s ability to detect a positive effect. In contrast, our study demonstrated more substantial separation in SUA levels during follow-up (300–360 μmol/L in the treat-to-target arm vs 420 μmol/L in the non-treat-to-target arm). Second, the Badve trial [15] used topiroxostat, whereas our study primarily utilized other xanthine oxidase inhibitors (predominantly febuxostat). Furthermore, the Badve study had a follow-up period of only 1 year, excluded patients with overt proteinuria, enrolled participants with relatively preserved kidney function (mean eGFR, 51.6 ± 11.0 vs 41.3 ± 11.8 mL/min/1.73 m2 in our study), and used change in UACR rather than hard endpoints such as ESKD as the primary outcome, which may limit generalizability. Our study addressed these limitations by incorporating a composite outcome encompassing both ESKD and rapid kidney function decline, thereby providing a more comprehensive assessment.
Given the discrepancy between our findings and the current Chinese guidelines, which recommend ULT for CKD patients with asymptomatic hyperuricemia, it is essential to reconcile our results with previously published Chinese studies. For instance, Yang et al. and Siu et al. demonstrated that febuxostat or allopurinol could slow eGFR decline in CKD stages 3–4 over a 12-month period [9,27]. Similarly, Wen et al. [8] reported short-term (24-week) improvements in renal function among patients with diabetic nephropathy. These divergent results, compared to our study, may be attributed to population heterogeneity, differences in age distribution, varying baseline renal function, and diverse outcome definitions. Specifically, the study by Yang et al. involved a significantly younger population (mean age around 57 years) and a disproportionately high percentage of males (80.9% in the febuxostat group), which may limit the generalizability of their findings to a broader CKD cohort. Furthermore, while they observed a positive effect on eGFR decline ≥50%, their analyses for eGFR decline ≥30% and dialysis initiation remained non-significant, with wide confidence intervals suggesting limited statistical robustness. Regarding the study by Wen et al. their focus was restricted to CKD stage 3 patients with diabetic nephropathy, whereas diabetic patients comprised only 30% of our study population. The reliability of their conclusions is further limited by the small sample size (N = 38), the brief 6-month follow-up, and the use of surrogate endpoints (eGFR changes) instead of hard clinical events, which may yield overoptimistic or unstable results. While the earlier study by Siu et al. employed rigorous hard endpoints (e.g. ESKD requiring dialysis, death), their use of allopurinol as the primary intervention contrasts with our cohort, where febuxostat was the predominant ULT prescription; this may introduce differences due to the distinct pleiotropic effects of these drug classes. Additionally, our results are highly consistent with a recent large-scale analysis from the China Renal Data System (CRDS), provided that drug-specific effects and patient phenotypes are considered [28]. While the CRDS study overall associated ULT with a reduced risk of CKD progression, a more nuanced subgroup analysis revealed that this benefit was primarily driven by allopurinol and benzbromarone. Crucially, febuxostat—which accounted for approximately 90% of the ULT prescriptions in our cohort—showed only a non-significant trend toward renoprotection in the CRDS study, directly aligning with our observations. Furthermore, a fundamental distinction exists in the study populations. The CRDS cohort included patients with symptomatic gout (9.4%), whereas our study focused exclusively on asymptomatic hyperuricemia. Given that patients with gout inherently carry a higher risk of renal progression and a greater burden of intrarenal crystal deposition, they may derive more benefit from ULT than the purely asymptomatic population investigated here. Therefore, the apparent discrepancy in overall conclusions between the two studies is likely explained by the predominance of febuxostat and the absence of gouty patients in our study.
Collectively, current evidence suggests that uric acid (UA) may function as a bystander rather than a direct mediator in CKD progression [29]. Mechanistically, asymptomatic hyperuricemia differs fundamentally from gout, which can be explained by the crystal-deposition hypothesis. Recent experimental data suggest that asymptomatic hyperuricemia differs fundamentally from gout in its renal impact. Elevated serum UA levels alone appear insufficient to accelerate kidney function decline [30]. Instead, renal injury likely requires the precipitation of monosodium urate (MSU) crystals within the tubulointerstitium, a process often facilitated by urinary acidification or specific dietary factors rather than high serum levels per se. According to this hypothesis, UA crystals (rather than soluble UA) act as the primary trigger for M1-like macrophage infiltration and the formation of granulomatous interstitial nephritis, which subsequently drives progressive fibrosis. Notably, soluble UA may even act as an intrinsic negative regulator of innate immunity via the SLC2A9 transporter, potentially exerting an immunomodulatory effect in acute inflammation [31,32]. Additionally, a rigorous post-hoc analysis of gout patients enrolled in the CARES trial suggested that more stringent UA control was associated with reduced risk of CKD progression, implying that patients with established crystal deposition may derive greater benefit from intensive urate-lowering strategies [33]. Taken together, the current body of evidence does not support pushing for strict urate targets solely for the purpose of preserving kidney function in patients with CKD and asymptomatic hyperuricemia in whom clinicians have decided to start ULT.
The principal strength of this study is the application of a TTE framework to real-world clinical data to simulate a RCT comparing treat-to-target versus non-treat-to-target strategies for delaying CKD progression while mitigating potential biases inherent in observational research. Additionally, the implementation of multiple sensitivity analyses strengthens the robustness of our findings.
This study has several limitations. First, and most notably, BMI was excluded from our models due to a high rate of missing data (51.7%). Because obesity is a well-established and important confounder for CKD progression, residual confounding from unmeasured or poorly measured factors, including BMI, remains a nontrivial limitation despite our use of IPCW to balance observed covariates. Second, the exclusion of patients with missing baseline SUA or eGFR measurements may limit the generalizability of our findings to patients with poor treatment adherence or more severe disease. Third, this study was conducted using data exclusively from Zhejiang Provincial People’s Hospital, which precludes capture of follow-up information for patients who may have received care at other institutions, potentially resulting in information loss. Fourth, the pooling of different ULT agents (febuxostat, benzbromarone, and allopurinol) as a single exposure is a design constraint. These drugs have different mechanisms (xanthine oxidase inhibition versus uricosuric effects) and possibly different non-urate effects on the kidney, potentially introducing heterogeneity. Although our subgroup sensitivity analysis restricted to febuxostat users partially addressed this concern, it is important to clarify that our results are largely based on a febuxostat-dominant real-world practice setting. Therefore, to avoid overgeneralization to ‘ULT overall’, extrapolation of these findings to other agents or classes may be limited. Future studies utilizing multicenter collaborations and larger real-world datasets, such as health insurance claims databases, are warranted to validate these findings.
5. Conclusion
In this TTE study, a treat-to-target urate management strategy (aiming for a serum urate level <360 μmol/L) was not associated with a reduced risk of the composite kidney outcome (ESKD or rapid kidney function decline) when compared to a non-treat-to-target strategy among patients with stage G3–G4 CKD and asymptomatic hyperuricemia who had initiated ULT. Crucially, these findings suggest that among patients in whom clinicians decide to start ULT, pushing to achieve this intensive urate target does not improve renal outcomes.
Supplementary Material
Acknowledgments
We are grateful to all participants from Zhejiang Provincial People’s Hospital for their participation in this study.
Funding Statement
This study was not supported by any sponsor or funder.
Ethical approval
This study was approved by the Ethics Committee of Zhejiang Provincial People’s Hospital on 19 September 2025 (No: QT2025212).
Disclosure statement
No potential conflict of interest was reported by the author(s).
Data availability statement
The data that support the findings of this study are not publicly available due to local laws related to data confidentiality but are available from the corresponding author Qiang He (email address: qianghe1973@126.com) upon reasonable request.
References
- 1.Su X, Xu B, Yan B, et al. Effects of uric acid-lowering therapy in patients with chronic kidney disease: a meta-analysis. PLOS One. 2017;12(11):e0187550. doi: 10.1371/journal.pone.0187550. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Sato Y, Feig DI, Stack AG, et al. The case for uric acid-lowering treatment in patients with hyperuricaemia and CKD. Nat Rev Nephrol. 2019;15(12):767–775. doi: 10.1038/s41581-019-0174-z. [DOI] [PubMed] [Google Scholar]
- 3.Borghi C, Agabiti-Rosei E, Johnson RJ, et al. Hyperuricaemia and gout in cardiovascular, metabolic and kidney disease. Eur J Intern Med. 2020;80:1–11. doi: 10.1016/j.ejim.2020.07.006. [DOI] [PubMed] [Google Scholar]
- 4.Badve SV, Pascoe EM, Tiku A, et al. Effects of allopurinol on the progression of chronic kidney disease. N Engl J Med. 2020;382(26):2504–2513. doi: 10.1056/nejmoa1915833. [DOI] [PubMed] [Google Scholar]
- 5.Doria A, Galecki AT, Spino C, et al. Serum urate lowering with allopurinol and kidney function in type 1 diabetes. N Engl J Med. 2020;382(26):2493–2503. doi: 10.1056/NEJMoa1916624. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kimura K, Hosoya T, Uchida S, et al. Febuxostat therapy for patients with stage 3 CKD and asymptomatic hyperuricemia: a randomized trial. Am J Kidney Dis. 2018;72(6):798–810. doi: 10.1053/j.ajkd.2018.06.028. [DOI] [PubMed] [Google Scholar]
- 7.Kidney Disease: Improving Global Outcomes (KDIGO) CKD Work Group . KDIGO 2024 clinical practice guideline for the evaluation and management of chronic kidney disease. Kidney Int. 2024;105(4):S117–S314. doi: 10.1016/j.kint.2023.10.018. [DOI] [PubMed] [Google Scholar]
- 8.Wen H, Yongling Z, Shuying Z, et al. Effect of febuxostat on renal function in patients from South China with CKD3 diabetic nephropathy. J Bras Nefrol. 2020;42(4):393–399. doi: 10.1590/2175-8239-JBN-2019-0091. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Yang H, Li R, Li Q, et al. Effects of febuxostat on delaying chronic kidney disease progression: a randomized trial in China. Int Urol Nephrol. 2023;55(5):1343–1352. doi: 10.1007/s11255-022-03437-5. [DOI] [PubMed] [Google Scholar]
- 10.Chinese Medical Association Chinese Society of Endocrinology . Guideline for the diagnosis and management of hyperuricemia and gout in China (2019). Chin J Endocrinol Metab. 2020;36(1):1–13. doi: 10.3760/cma.j.issn.1000-6699.2020.01.001. [DOI] [Google Scholar]
- 11.Sun M, Lyu Z, Wang C, et al. 2024 update of Chinese guidelines for diagnosis and treatment of hyperuricemia and gout part i: recommendations for general patients. Int J Rheum Dis. 2025;28(7):e70375. doi: 10.1111/1756-185x.70375. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Li C, Sun M, Liu Z, et al. 2024 update of chinese guidelines for management of hyperuricemia and gout part II: recommendations for patients with common comorbidities. Int J Rheum Dis. 2025;28(8):e70402. doi: 10.1111/1756-185x.70402. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Kohagura K, Satoh A, Kochi M, et al. Urate-lowering drugs for chronic kidney disease with asymptomatic hyperuricemia and hypertension: a randomized trial. J Hypertens. 2023;41(9):1420–1428. doi: 10.1097/HJH.0000000000003484. [DOI] [PubMed] [Google Scholar]
- 14.Wang Y, Dalbeth N, Terkeltaub R, et al. Target serum urate achievement and chronic kidney disease progression in patients with gout and kidney disease. JAMA Intern Med. 2025;185(1):74–82. doi: 10.1001/jamainternmed.2024.6212. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Yamamoto T, Kasahara M, Ueshima K, et al. Multicenter randomized controlled trial of intensive uric acid lowering therapy for CKD patients with hyperuricemia: TARGET-UA. Clin Exp Nephrol. 2024;28(8):764–772. doi: 10.1007/s10157-024-02483-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Matthews AA, Danaei G, Islam N, et al. Target trial emulation: applying principles of randomised trials to observational studies. BMJ. 2022;378:e071108. doi: 10.1136/bmj-2022-071108. [DOI] [PubMed] [Google Scholar]
- 17.Hernán MA, Wang W, Leaf DE.. Target trial emulation: a framework for causal inference from observational data. JAMA. 2022;328(24):2446–2447. doi: 10.1001/jama.2022.21383. [DOI] [PubMed] [Google Scholar]
- 18.Liu Q, Liu T, Fang Z, et al. The effect of febuxostat on kidney outcomes in patients with chronic kidney disease and asymptomatic hyperuricemia: a target trial emulation. Am J Nephrol. 2025:1–11. doi: 10.1159/000550047. [DOI] [PubMed] [Google Scholar]
- 19.Hernán MA. How to estimate the effect of treatment duration on survival outcomes using observational data. BMJ. 2018;360:k182. doi: 10.1136/bmj.k182. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Levey AS, Stevens LA, Schmid CH, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009;150(9):604–612. doi: 10.7326/0003-4819-150-9-200905050-00006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Stuart EA. Matching methods for causal inference: A review and a look forward. Stat Sci. 2010;25(1):1–21. doi: 10.1214/09-STS313. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Austin PC. Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity-score matched samples. Stat Med. 2009;28(25):3083–3107. doi: 10.1002/sim.3697. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Hernán MA, Brumback B, Robins JM.. Marginal structural models to estimate the causal effect of zidovudine on the survival of HIV-positive men. Epidemiology. 2000;11(5):561–570. doi: 10.1097/00001648-200009000-00012. [DOI] [PubMed] [Google Scholar]
- 24.Ngwa JS, Cabral HJ, Cheng DM, et al. A comparison of time dependent Cox regression, pooled logistic regression and cross sectional pooling with simulations and an application to the Framingham Heart Study. BMC Med Res Methodol. 2016;16(1):148. doi: 10.1186/s12874-016-0248-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Coresh J, Heerspink HJL, Sang Y, et al. Change in albuminuria and subsequent risk of end-stage kidney disease: an individual participant-level consortium meta-analysis of observational studies. Lancet Diabetes Endocrinol. 2019;7(2):115–127. doi: 10.1016/S2213-8587(18)30313-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Heerspink HJL, Collier WH, Chaudhari J, et al. A meta-analysis of albuminuria as a surrogate endpoint for kidney failure. Nat Med. 2026;32(1):281–287. doi: 10.1038/s41591-025-04057-z. [DOI] [PubMed] [Google Scholar]
- 27.Siu YP, Leung KT, Tong MKH, et al. Use of allopurinol in slowing the progression of renal disease through its ability to lower serum uric acid level. Am J Kidney Dis. 2006;47(1):51–59. doi: 10.1053/j.ajkd.2005.10.006. [DOI] [PubMed] [Google Scholar]
- 28.Nie S, Zhou S, Chen R, et al. Urate-lowering therapy and kidney outcomes in patients with chronic kidney disease and hyperuricemia. Signal Transduct Target Ther. 2025;10(1):399. doi: 10.1038/s41392-025-02497-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Stack AG. Can we crystallize the role of urate-lowering treatment in chronic kidney disease? Kidney Int. 2025;107(3):394–396. doi: 10.1016/j.kint.2024.11.025. [DOI] [PubMed] [Google Scholar]
- 30.M S HP, M Q, K N, et al. Only hyperuricemia with crystalluria, but not asymptomatic hyperuricemia, drives progression of chronic kidney disease. J Am Soc Nephrol. 2020;31(12):2773–2792. doi: 10.1681/ASN.2020040523. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Ma Q, Honarpisheh M, Li C, et al. Soluble uric acid is an intrinsic negative regulator of monocyte activation in monosodium urate crystal-induced tissue inflammation. J Immunol. 2020;205(3):789–800. doi: 10.4049/jimmunol.2000319. [DOI] [PubMed] [Google Scholar]
- 32.Mulay SR, Eberhard JN, Pfann V, et al. Oxalate-induced chronic kidney disease with its uremic and cardiovascular complications in C57BL/6 mice. Am J Physiol Renal Physiol. 2016;310(8):F785–F795. doi: 10.1152/ajprenal.00488.2015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Ghang B, Park J, Lee JS, et al. Post-hoc analysis of the CARES trial suggests delayed progression of chronic kidney disease in patients with gout during urate-lowering therapy. Kidney Int. 2025;107(3):521–529. doi: 10.1016/j.kint.2024.10.022. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are not publicly available due to local laws related to data confidentiality but are available from the corresponding author Qiang He (email address: qianghe1973@126.com) upon reasonable request.