Abstract
Introduction
Nondiabetic kidney disease (NDKD) is a major contributor to chronic kidney disease worldwide. Although renin-angiotensin system inhibitors (RASi) remain standard therapy, sodium-glucose cotransporter-2 inhibitors (SGLT2i), non-steroidal mineralocorticoid receptor antagonists (non-steroidal MRAs), and glucagon-like peptide-1 receptor agonists are now recommended in clinical guidelines. However, direct comparative evidence to guide their prioritization in NDKD is limited.
Methods
We performed a systematic review and Bayesian network meta-analysis of randomized controlled trials enrolling adults with NDKD. The primary endpoint was the annualized change in estimated glomerular filtration rate (eGFR) slope; secondary endpoints included urine protein-to-creatinine ratio (UPCR), blood pressure, and biochemical safety parameters.
Results
Eighteen trials were included. SGLT2i, as add-on to RASi, produced the greatest preservation of kidney function compared with RASi alone (mean difference [MD] 12.1 mL/min/1.73 m2/year; 95% credible interval [CrI] 3.9 to 20.4). Non-steroidal MRA achieved the largest reduction in proteinuria versus RASi (UPCR MD −0.4 g/g; 95% CrI: −1.0 to −0.2). SGLT2i also yielded the greatest systolic blood pressure reduction versus RASi (MD −9.1 mm Hg vs. RASi; 95% CrI: −11.7 to −6.5). Surface-under-the-cumulative-ranking probabilities identified SGLT2i as most effective for eGFR slope preservation (99.8%) and non-steroidal MRA as most effective for proteinuria reduction (98.8%). Safety analyses showed a mild, expected hemoglobin decrease with SGLT2i and a nonsignificant creatinine rise with non-steroidal MRA.
Conclusions
SGLT2i and non-steroidal MRA confer distinct renal benefits in NDKD – SGLT2i primarily preserve kidney function, whereas non-steroidal MRAs most effectively reduce proteinuria. These results support a phenotype-driven therapeutic framework and highlight the need for head-to-head trials to validate precision-guided treatment strategies in NDKD.
Keywords: Sodium-glucose cotransporter-2 inhibitors, Non-steroidal mineralocorticoid receptor antagonists, Renin-angiotensin system inhibitors, Glucagon-like peptide-1 receptor agonists, Nondiabetic kidney disease
Introduction
Nondiabetic kidney disease (NDKD) comprises a diverse spectrum of disorders – including IgA nephropathy, focal segmental glomerulosclerosis, and hypertensive nephrosclerosis – that together account for a major share of the global chronic kidney disease (CKD) burden [1, 2]. These conditions often progress silently to kidney failure, and their increasing recognition within mixed diabetic and nondiabetic cohorts has complicated both diagnosis and management [3].
For more than 2 decades, renin-angiotensin system inhibitors (RASi) have remained the cornerstone of therapy in proteinuric NDKD. By lowering intraglomerular pressure through efferent arteriolar dilation, RASi slow disease progression and continue to be recommended as first-line therapy in the 2024 KDIGO CKD guideline update [4]. Yet despite widespread use, residual risk remains substantial – many patients experience persistent proteinuria or ongoing estimated glomerular filtration rate (eGFR) decline even under maximal RAS inhibition [5, 6].
To address this therapeutic gap, the 2024 KDIGO framework now supports a multi-pillar approach integrating newer renoprotective agents. Sodium-glucose cotransporter-2 inhibitors (SGLT2i) are recommended for all patients with CKD and albuminuria irrespective of diabetes status, reflecting their hemodynamic and cardiovascular benefits [7, 8]. Non-steroidal mineralocorticoid receptor antagonists (non-steroidal MRAs), particularly the non-steroidal agent finerenone, are endorsed for patients with residual proteinuria and elevated cardiovascular risk owing to their anti-inflammatory and antifibrotic properties [9]. Glucagon-like peptide-1 receptor agonists (GLP-1 RAs) are suggested for those with obesity or metabolic risk, offering modest renal and cardiovascular benefit [10].
However, this broadened therapeutic landscape leaves a key clinical question unresolved: how should these agents be prioritized or sequenced in nondiabetic CKD? NDKD encompasses biologically distinct phenotypes, some dominated by hemodynamic stress, others by inflammatory or fibrotic injury, raising the possibility that each drug class exerts optimal benefit in specific disease contexts. No head-to-head randomized controlled trials (RCTs) have directly compared RASi, SGLT2i, non-steroidal MRA, and GLP-1 RA in a well-defined NDKD population, leaving clinicians without clear evidence to guide therapy selection beyond general guideline recommendations.
To fill this gap, we conducted a systematic review and Bayesian network meta-analysis (NMA) of RCTs in adults with NDKD. By focusing on an exclusively nondiabetic population and using the annualized eGFR slope as a standardized indicator of long-term renal decline, this analysis provides a quantitative framework for comparing modern renoprotective therapies. Our findings aimed to clarify the relative efficacy and mechanistic profiles of these drug classes and to inform a more phenotype-driven, precision-based approach to treatment in NDKD.
Methods
Protocol and Reporting Guidelines
This NMA followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Network Meta-Analyses (PRISMA-NMA) guidelines [11]. The study protocol was prospectively registered in PROSPERO (CRD420251087200). One amendment was made during the review process to refine subgroup definitions.
Search Strategy
A comprehensive search of PubMed, Cochrane Library, Web of Science, Scopus, and Embase was performed from database inception to March 2025. We identified RCTs evaluating four major renoprotective drug classes – SGLT2i, RASi, non-steroidal MRA, and GLP-1 RAs – in adults with NDKD. For each class, both drug names and class descriptors were included: dapagliflozin, empagliflozin, canagliflozin (SGLT2i); lisinopril, enalapril, losartan, valsartan, sacubitril/valsartan (RASi); non-steroidal MRAs; and semaglutide and related GLP-1 RAs. Disease-related terms encompassed “non-diabetic kidney disease,” “glomerulonephritis,” “IgA nephropathy,” “focal segmental glomerulosclerosis,” and “hypertensive nephropathy,” along with corresponding MeSH terms. Two reviewers independently screened titles and abstracts, assessed full texts for eligibility, and resolved disagreements by discussion or third-party adjudication. Reference lists of eligible studies and prior meta-analyses were also searched manually.
Study Selection
Eligible studies met all of the following criteria: (1) peer-reviewed RCTs enrolling adults (≥18 years) with NDKD; (2) comparison of at least one of the four drug classes (SGLT2i, RASi, non-steroidal MRA, GLP-1 RA) versus placebo or active control; (3) reporting of at least one predefined outcome (eGFR, proteinuria, blood pressure, or biochemical markers). The primary analysis was restricted to trials including exclusively nondiabetic participants or reporting separate data for NDKD subgroups. “Nondiabetic” was defined as the absence of physician-diagnosed diabetes, typically supported by HbA1c <6.5% and/or fasting glucose <7.0 mmol/L. Trials with mixed populations lacking separable data were excluded. In total, 18 RCTs were included in the primary NMA.
Data Extraction and Quality Assessment
Two reviewers independently extracted study characteristics, patient demographics, baseline renal function, treatment details, and quantitative outcome data (means, standard deviations, and sample sizes). When eGFR slope was not directly reported, it was calculated as mean eGFR change divided by study duration (years). Risk of bias was assessed using the Cochrane Risk of Bias 2 (RoB 2) tool [12], evaluating randomization, deviations from intended interventions, missing data, outcome measurement, and selective reporting. Discrepancies were resolved by consensus.
Outcome Selection
The primary outcome was the annualized eGFR slope (mL/min/1.73 m2/year). Secondary outcomes included urine protein-to-creatinine ratio (UPCR), systolic blood pressure (SBP), diastolic blood pressure (DBP), serum creatinine, hemoglobin, and potassium. Although hard clinical endpoints such as end-stage kidney disease, sustained 40% eGFR decline, or renal death represent the definitive measures of kidney disease progression, they were rarely reported in the included trials, particularly those with shorter follow-up durations or those evaluating newer agents in nondiabetic populations. Therefore, this NMA focuses on eGFR slope and proteinuria as well-established surrogate endpoints for CKD progression, which have been robustly validated against hard clinical outcomes in large-scale cohort studies and clinical trials.
Statistical Analysis
A Bayesian random-effects NMA was conducted in R (version 4.3.1; R Foundation for Statistical Computing, Vienna, Austria) using the gemtc package (version 1.0-2) for Bayesian NMA and the rjags package to interface with JAGS for Markov chain Monte Carlo simulations. Treatment effects for continuous outcomes were expressed as mean differences (MDs) with 95% credible intervals (CrIs) [13]. The primary outcome was the annualized eGFR slope (mL/min/1.73 m2/year). Secondary outcomes included UPCR, SBP, DBP, serum creatinine, hemoglobin, and potassium. Relative efficacy was ranked using surface-under-the-cumulative-ranking (SUCRA) probabilities. Model convergence was confirmed by trace plots and Brooks-Gelman-Rubin statistics approaching 1.0. Heterogeneity was quantified using the I2 statistic, and node-splitting methods were used to assess local inconsistency. Sensitivity analyses included (1) fixed-effects modeling and (2) restriction to studies with biopsy-proven NDKD. Potential publication bias was evaluated visually using comparison-adjusted funnel plots, acknowledging limited power for some comparisons due to small sample sizes. In many included trials, especially those of newer agents, treatment effects were estimated relative to RASi-based standard care rather than in the absence of background therapy. However, because background RASi use was not consistently reported across all studies, this represents a source of clinical heterogeneity and should be considered when interpreting indirect comparisons.
Results
Literature Search and Included Studies
The study selection process is illustrated in the PRISMA flow diagram (shown in Fig. 1). A total of 4,025 records were retrieved from PubMed (n = 1,647), Cochrane Library (n = 1,321), Web of Science (n = 93), Scopus (n = 612), and Embase (n = 352). After removing 614 duplicates, 3,411 unique records were screened by title and abstract. Of these, 3,167 were excluded for the following reasons: non-randomized design (n = 983), unsuitable population (n = 786), review or meta-analysis (n = 722), inappropriate design (n = 564), or nonhuman study (n = 112). A total of 244 full-text articles were reviewed in detail, and 226 were excluded for not reporting outcomes of interest. Ultimately, 18 RCTs met the inclusion criteria and were incorporated into the NMA. These comprised seven SGLT2i studies, seven RASi studies, three GLP-1 RA studies, and four non-steroidal MRA studies [14–31].
Fig. 1.
PRISMA flow diagram of study selection. Flowchart showing the systematic search and screening process for RCTs included in the NMA. Numbers at each stage correspond to records identified, screened, excluded, and included.
Study Characteristics and Quality Assessment
The 18 included RCTs encompassed diverse geographic regions across Asia, Europe, Australia, and the Middle East, alongside several multinational consortia, underscoring the global relevance of the analysis. Participant characteristics varied widely across studies (online suppl. Table S1, S2; for all online suppl. material, see https://doi.org/10.1159/000552201). Mean participant age ranged from <45 years in younger cohorts to approximately 67 years in older populations. The proportion of male participants ranged from ∼25% to >70%, reflecting heterogeneous recruitment. Baseline kidney function spanned the spectrum of NDKD, with mean eGFR values from ∼15 mL/min/1.73 m2 in advanced CKD to >110 mL/min/1.73 m2 in early disease. HbA1c levels were generally <6.0%, confirming the nondiabetic status of included participants. By intervention class, seven studies evaluated SGLT2i, seven evaluated RASi, three investigated GLP-1 receptor agonists, and four assessed mineralocorticoid receptor antagonists. The SGLT2i trials covered a wide CKD spectrum and included landmark multinational studies such as EMPA-KIDNEY. Methodological quality was appraised using the Cochrane RoB 2 tool (shown in Fig. 2). Most studies demonstrated low overall risk of bias, particularly in blinding and outcome assessment domains. Several older RASi trials showed unclear risk in random sequence generation and allocation concealment due to incomplete reporting. Publication bias was examined using comparison-adjusted funnel plots (shown in online suppl. Fig. S1), which indicated low risk for all outcomes except serum potassium. The GRADE framework further supported the overall moderate-to-high quality of evidence across efficacy and safety outcomes (shown in online suppl. Table S3).
Fig. 2.
Risk-of-bias assessment for included studies. Summary of authors’ judgments across five Cochrane RoB 2 domains for the 18 included RCTs. Green = low risk; yellow = some concerns/unclear risk.
NMA Outcomes
The evidence networks for each outcome are presented in the corresponding figures. All networks were anchored by a common placebo comparator, permitting both direct and indirect comparisons across the four major renoprotective drug classes. Results are reported by outcome domain.
Renal Function
The evidence networks for renal function outcomes are presented in Figures 3a and 4a. All networks were connected through shared comparators, permitting indirect comparisons across the four drug classes.
Fig. 3.
Effects of pharmacologic classes on eGFR slope. a Network geometry of treatment comparisons. b Forest plot showing MDs with 95% CrIs. c SUCRA probabilities for treatment ranking. SGLT2i showed the most favorable profile for eGFR preservation, followed by non-steroidal MRA.
Fig. 4.
Effects of pharmacologic classes on UPCR. a Network geometry of available comparisons. b Forest plot showing MD (95% CrI). c SUCRA probabilities for treatment ranking. Non-steroidal MRA demonstrated the greatest reduction in proteinuria, followed by SGLT2i.
eGFR Slope. The evidence network for renal function outcomes (shown in Fig. 3a) was well connected through shared comparators, allowing both direct and indirect treatment contrasts across all four drug classes. In the Bayesian NMA, SGLT2i (administered as add-on therapy to RASi) produced the greatest attenuation of kidney function decline. Compared with RASi alone, SGLT2i slowed the eGFR decline by an MD of 12.1 mL/min/1.73 m2 per year (95% CrI, 3.9 to 20.4) (shown in Fig. 3b). The effect of non-steroidal MRA versus RASi was smaller and statistically significant (MD −8.9 mL/min/1.73 m2 per year; 95% CrI, −12.8 to −4.9). On the contrary, GLP-1 RA showed no significant difference compared with RASi (MD 2.1; 95% CrI, −0.9 to 5.2). The network league table (shown in Fig. 3a) confirmed these relationships. Ranking analyses based on SUCRA probabilities indicated that SGLT2i had the highest probability of being the most effective therapy for preserving eGFR slope (99.8%), followed by GLP-1 RA (71.9%), RASi (47.1%), placebo (30.8%), and non-steroidal MRA (0.4%) (shown in Fig. 3c). It should be noted that SUCRA values describe relative ranking probabilities rather than absolute effect magnitudes. Model convergence for the Bayesian analysis was confirmed by trace plots and Gelman-Rubin shrink factors approaching 1.0. Online supplementary Figure S2 demonstrated stable trace lines and unimodal posterior distributions, supporting the reliability of the eGFR slope estimates.
UPCR. For proteinuria reduction, non-steroidal MRA demonstrated the strongest effect. Compared with RASi, non-steroidal MRA treatment significantly lowered UPCR (MD −0.4 g/g; 95% CrI, −1.0 to −0.2) (shown in Fig. 4a, b). SGLT2i, relative to RASi, showed a modest, nonsignificant trend toward lower UPCR (MD −0.0 g/g; 95% CrI, −0.1 to 0.1). RASi alone reduced UPCR significantly versus placebo (MD −0.5 g/g; 95% CrI, −0.7 to −0.3). Ranking probabilities based on SUCRA values indicated that non-steroidal MRA had the highest probability of being the most effective therapy for proteinuria reduction (98.8%), followed by GLP-1 RA (61.4%), SGLT2i (41.7%), RASi (32.2%), and placebo (15.8%) (shown in Fig. 4c). Convergence of the Bayesian safety model was supported by inspection of trace plots and Gelman-Rubin shrink factors approaching 1.0, with stable trace lines and unimodal posterior distributions shown in online supplementary Figure S3. Together with eGFR results, these findings highlight a clear dissociation of therapeutic effects: SGLT2i ranked highest for long-term kidney function preservation, whereas non-steroidal MRAs were most effective for reducing proteinuria.
Blood Pressure
The effects of pharmacologic interventions on SBP and DBP are summarized in online supplementary Figures S6 and S4, respectively.
DBP. The effects of each pharmacologic class on diastolic pressure are shown in Figure S4. Overall, treatment-related changes in DBP were small and statistically inconsistent across interventions. SGLT2i showed a nonsignificant reduction compared with RASi (MD −3.1 mm Hg; 95% CrI, −9.0 to 2.8), and no other pairwise comparisons reached statistical significance (shown in Fig. S4A, B). Ranking probabilities based on SUCRA values yielded a counterintuitive hierarchy, with RASi ranking highest for DBP reduction (76.1%), followed by non-steroidal MRA (60.6%), GLP-1 RA (50.8%), RASi (34.1%), and placebo (28.3%) (shown in Fig. S4C). This pattern likely reflects uniformly modest treatment effects and wide uncertainty intervals rather than meaningful clinical differences among interventions. Model convergence for the Bayesian DBP analysis was confirmed by trace plots and Gelman-Rubin shrink factors, with online supplementary Figure S5 demonstrating stable trace lines and unimodal posterior distributions.
SBP. As shown in online supplementary Figure S6, SGLT2i, which were used as add-on therapy to RASi, produced the largest and statistically significant reduction in SBP (MD −9.1 mm Hg vs. RASi; 95% CrI, −11.7 to −6.5). In contrast, non-steroidal MRA showed no significant effect versus SGLT2i (MD 3.8 mm Hg; 95% CrI, 1.4 to 6.1), and RASi alone had minimal effect versus placebo (MD −1.3 mm Hg; 95% CrI, −3.5 to −0.9). The network league table (shown in online suppl. Fig. S6A) corroborated these findings, showing consistent advantages for SGLT2i over other active comparators. SUCRA rankings placed SGLT2i as most likely to be the top therapy for SBP reduction (88.6%), followed by non-steroidal MRA (77.9%), GLP-1 RA (47.5%), RASi (31.8%), and placebo (4.2%) (shown in online suppl. Fig. S6C). Model convergence for the Bayesian SBP analysis was confirmed by trace plots and Gelman-Rubin shrink factors, with online supplementary Figure S7 demonstrating stable trace lines and unimodal posterior distributions. Thus, SGLT2i were associated with a robust and clinically relevant reduction in SBP, whereas effects on diastolic pressure were modest and inconsistent across all drug classes.
Biochemical and Safety Indicators
The comparative effects on key biochemical and safety markers are summarized below.
Serum Creatinine. The NMA showed no statistically significant differences in serum creatinine change among active treatments (shown in online suppl. Fig. S8A). Non-steroidal MRA produced a modest, nonsignificant increase compared with RASi (MD 8.2 µmol/L; 95% CrI, −3.0 to 19.4), whereas SGLT2i showed minimal change (MD 0.6 µmol/L; 95% CrI, 0.2 to 1.1). RASi alone had a small, nonsignificant difference versus placebo (MD −6.9 µmol/L; 95% CrI, −14.3 to 0.4) (shown in online suppl. Fig. S8B). SUCRA probabilities ranked RASi highest for favorable creatinine profile (73.7%), followed by SGLT2i (72.0%), placebo (33.1%), and non-steroidal MRA (21.3%) (shown in online suppl. Fig. S8C). Model convergence for the Bayesian serum creatinine analysis was confirmed by trace plots and Gelman-Rubin shrink factors (online suppl. Fig. S9).
Hemoglobin. Across trials, hemoglobin changes were minor (shown in online suppl. Fig. S10A). SGLT2i caused a small but statistically significant reduction compared with RASi (MD −4.6 g/L; 95% CrI, −7.7 to −1.4), consistent with the known hemodilution effect of this class. RASi showed a larger decline versus placebo (MD −15.4 g/L; 95% CrI, −19.9 to −10.9), while non-steroidal MRA showed no significant difference relative to RASi (MD −2.5 g/L; 95% CrI, −5.3 to 0.3) (shown in online suppl. Fig. S10B). SUCRA rankings suggested RASi were associated with the smallest overall hemoglobin reduction (75.2%), followed by placebo (72.4%), SGLT2i (37.6%), and non-steroidal MRA (14.8%) (shown in online suppl. Fig. S10C). Model convergence for the Bayesian hemoglobin analysis was confirmed by trace plots and Gelman-Rubin shrink factors (online suppl. Fig. S11).
Serum Potassium. Changes in serum potassium were minimal and clinically insignificant (shown in online suppl. Fig. S12A). All MDs overlapped zero, for example, SGLT2i vs. RASi (MD 0.00 mmol/L; 95% CrI, −0.00 to 0.10) and non-steroidal MRA vs. RASi (MD 0.10 mmol/L; 95% CrI, 0.10 to 0.20) (shown in online suppl. Fig. S12B). SUCRA values showed negligible separation among interventions: non-steroidal MRA (94.4%), placebo (46.4%), SGLT2i (41.6%), and RASi (17.6%) (shown in online suppl. Fig. S12C). Model convergence for the Bayesian serum potassium analysis was confirmed by trace plots and Gelman-Rubin shrink factors (online suppl. Figure S13).
In summary, biochemical safety profiles varied modestly by drug class. SGLT2i produced a mild, expected hemoglobin decrease without affecting potassium; non-steroidal MRAs were associated with a nonsignificant creatinine rise; and RASi showed moderate hemoglobin reduction. Across all agents, effects on serum potassium were minimal.
Discussion
This updated NMA provides a comprehensive evaluation of four major renoprotective drug classes in NDKD. By standardizing background therapy and restricting inclusion to NDKD populations, the analysis delineates distinct efficacy profiles across classes. SGLT2i most effectively preserved kidney function by slowing eGFR decline, whereas non-steroidal MRA achieved the greatest reduction in proteinuria. This divergence indicates complementary therapeutic mechanisms that could inform a more personalized treatment framework for NDKD.
The superior preservation of eGFR slope with SGLT2i underscores their predominant action on the hemodynamic axis of CKD progression. By enhancing tubuloglomerular feedback and lowering intraglomerular pressure, SGLT2i mitigate mechanical nephron stress – particularly relevant to hyperfiltration phenotypes such as obesity-related glomerulopathy or hypertensive nephrosclerosis [32, 33]. In contrast, the pronounced antiproteinuric effect of non-steroidal MRAs reflects their impact on the inflammatory and fibrotic axis of disease. By blocking mineralocorticoid receptor-mediated pathways of inflammation, oxidative stress, and fibrosis, non-steroidal MRAs address structural kidney injury, providing greater benefit in proteinuric or fibrotic disease states [34, 35]. Together, these findings suggest that hemodynamic and fibrotic pathways represent distinct yet targetable mechanisms of renal decline in NDKD.
These findings call for a fundamental rethinking of how renoprotective therapies are deployed in NDKD. Rather than applying a uniform treatment sequence to all patients, a phenotype-guided strategy may offer greater precision and efficacy. In individuals whose disease is driven primarily by hemodynamic stress, such as those with hyperfiltration or hypertensive nephrosclerosis, SGLT2i could serve as the therapeutic foundation, mitigating intraglomerular pressure and preserving filtration capacity. In contrast, patients with inflammatory or fibrotic phenotypes marked by persistent proteinuria may benefit more from mineralocorticoid receptor antagonists, which target the cellular pathways of inflammation and fibrosis that underlie structural progression.
Safety considerations further support this individualized approach. The modest hemoglobin reduction observed with SGLT2i represents an expected hemodilution effect, while the nonsignificant creatinine rise with non-steroidal MRA likely reflects transient hemodynamic adjustment rather than nephrotoxicity [36, 37]. Neither signal warrants withholding therapy, though both merit clinical monitoring.
Future studies should include head-to-head trials directly comparing SGLT2i and non-steroidal MRA in NDKD populations, stratified by hemodynamic and fibrotic biomarkers. Integrating multi-omic profiling and computational modeling could refine phenotype identification and guide mechanism-based therapy selection, advancing precision nephrology.
This study’s strengths include adherence to PRISMA-NMA standards, a stringent focus on nondiabetic CKD populations, and a network structure reflecting modern combination therapy paradigms. Nonetheless, several limitations should be acknowledged. Indirect comparisons introduce potential residual confounding, particularly across trials conducted in different eras. While background therapy differences were addressed in the network design, unmeasured covariates may persist. The use of surrogate endpoints, including eGFR slope and proteinuria, limits direct inference regarding hard clinical outcomes such as end-stage kidney disease, sustained major eGFR decline, or renal death. This reflects a broader evidence gap in NDKD, as most existing trials were not designed or powered to capture event-driven renal outcomes. Future long-term, event-driven RCTs are needed to validate these comparative findings.
Finally, the smaller number of non-steroidal MRA and GLP-1 RA trials necessitates cautious interpretation of their relative rankings. These limitations simultaneously highlight key directions for future work: well-powered, prospective, head-to-head trials; incorporation of molecular and imaging biomarkers for phenotyping; and integration of computational approaches for individualized therapy optimization. Moreover, evidence for the non-steroidal MRA node remains limited and is derived predominantly from finerenone trials in IgA nephropathy, which may challenge the transitivity of indirect comparisons and restrict the generalizability of these findings across the broader NDKD population.
Another important limitation of this study is the substantial variation in follow-up duration across the included RCTs, ranging from short crossover studies to multiyear trials. This issue is particularly relevant for eGFR slope, because short-term studies may predominantly capture acute hemodynamic or treatment initiation effects, whereas longer trials are more likely to reflect sustained changes in the rate of chronic kidney function decline. As a result, eGFR slope estimates may not be fully comparable across studies with markedly different durations. This variation in follow-up may also have contributed to between-study heterogeneity and may have challenged the comparability and transitivity underlying indirect comparisons in the network; therefore, our findings should be interpreted with caution. By the way, evidence for the non-steroidal MRA node was derived exclusively from finerenone trials; therefore, these findings should not be interpreted as applying to steroidal MRAs.
Conclusion
In summary, this analysis reveals distinct mechanistic and clinical signatures of contemporary renoprotective therapies in NDKD. SGLT2i predominantly preserve kidney function through hemodynamic modulation, whereas non-steroidal MRAs most effectively reduce proteinuria through antifibrotic mechanisms. These findings support a phenotype-driven therapeutic hierarchy rather than a universal sequencing strategy and underscore the need for prospective trials to validate this precision-guided framework.
Statement of Ethics
This study is a network meta-analysis of previously published literature. Therefore, ethical approval and patient consent were not required for this work.
Conflict of Interest Statement
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Funding Sources
This work was supported by the High-level Medical Team Project in Baoan, Shenzhen (202401), and the 2024 High-quality Development Research Project of Shenzhen Bao’an Public Hospital (BAGZL2024037). These funding bodies supported the research activities related to this study. However, they had no role in the design of the study, the collection, analysis, or interpretation of data, the conception, planning, or writing of the manuscript, or the decision to publish the results.
Author Contributions
Rulong Chen: conceptualization, methodology, writing – original draft, investigation, formal analysis, and data curation; Jinxin Zhang: investigation; Jiating Chen: methodology and investigation; Tingfei Xie and Yunpeng Xu: supervision; Zhaoyong Hu: conceptualization, investigation, and writing – review and editing; Jihong Chen: funding acquisition, conceptualization, and project administration.
Funding Statement
This work was supported by the High-level Medical Team Project in Baoan, Shenzhen (202401), and the 2024 High-quality Development Research Project of Shenzhen Bao’an Public Hospital (BAGZL2024037). These funding bodies supported the research activities related to this study. However, they had no role in the design of the study, the collection, analysis, or interpretation of data, the conception, planning, or writing of the manuscript, or the decision to publish the results.
Data Availability Statement
The data that support the findings of this study are not publicly available due to institutional data protection policies, but are available from the corresponding author upon reasonable request. Access to the data will be granted for noncommercial research purposes after review of a short research proposal and, where applicable, the signing of a data use agreement.
Supplementary Material.
Supplementary Material.
References
- 1. Wanner C, Schuchhardt J, Bauer C, Lindemann S, Brinker M, Kong SX, et al. Clinical characteristics and disease outcomes in non-diabetic chronic kidney disease: retrospective analysis of a US healthcare claims database. J Nephrol. 2023;36(1):45–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Wang J, Han Q, Zhao L, Zhang J, Wang Y, Wu Y, et al. Identification of clinical predictors of diabetic nephropathy and non-diabetic renal disease in Chinese patients with type 2 diabetes, with reference to disease course and outcome. Acta Diabetol. 2019;56(8):939–46. [DOI] [PubMed] [Google Scholar]
- 3. Anders HJ, Huber TB, Isermann B, Schiffer M. CKD in diabetes: diabetic kidney disease versus nondiabetic kidney disease. Nat Rev Nephrol. 2018;14(6):361–77. [DOI] [PubMed] [Google Scholar]
- 4. Lin H, Geurts F, Hassler L, Batlle D, Mirabito Colafella KM, Denton KM, et al. Kidney angiotensin in cardiovascular disease: formation and drug targeting. Pharmacol Rev. 2022;74(3):462–505. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Danser AH. The increase in renin during renin inhibition: does it result in harmful effects by the (pro)renin receptor? Hypertens Res. 2010;33(1):4–10. [DOI] [PubMed] [Google Scholar]
- 6. Xu Y, Fu EL, Trevisan M, Jernberg T, Sjölander A, Clase CM, et al. Stopping renin-angiotensin system inhibitors after hyperkalemia and risk of adverse outcomes. Am Heart J. 2022;243:177–86. [DOI] [PubMed] [Google Scholar]
- 7. Singhal R, Hechanova LA. SGLT2 inhibitors: benefits for CKD and cardiovascular disease in type 2 diabetes. Curr Cardiol Rep. 2022;24(3):183–9. [DOI] [PubMed] [Google Scholar]
- 8. Preda A, Montecucco F, Carbone F, Camici GG, Lüscher TF, Kraler S, et al. SGLT2 inhibitors: from glucose-lowering to cardiovascular benefits. Cardiovasc Res. 2024;120(5):443–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Zachariah T, Radhakrishnan J. Potential role of mineralocorticoid receptor antagonists in nondiabetic chronic kidney disease and glomerular disease. Clin J Am Soc Nephrol. 2024;19(11):1499–512. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Zhang S, Sidra F, Alvarez CA, Kinaan M, Lingvay I, Mansi IA. Healthcare utilization, mortality, and cardiovascular events following GLP1-RA initiation in chronic kidney disease. Nat Commun. 2024;15(1):10623. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Moher D, Liberati A, Tetzlaff J, Altman DG; PRISMA Group . Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009;6(7):e1000097. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Higgins JP, Altman DG, Gøtzsche PC, Jüni P, Moher D, Oxman AD, et al. The cochrane collaboration’s tool for assessing risk of bias in randomised trials. BMJ. 2011;343:d5928. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. 2003;327(7414):557–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Cherney DZI, Dekkers CCJ, Barbour SJ, Cattran D, Abdul Gafor AH, Greasley PJ, et al. Effects of the SGLT2 inhibitor dapagliflozin on proteinuria in non-diabetic patients with chronic kidney disease (DIAMOND): a randomised, double-blind, crossover trial. Lancet Diabetes Endocrinol. 2020;8(7):582–93. [DOI] [PubMed] [Google Scholar]
- 15. ElSharkawy M, Emara A, Ahmed MM, Ghonamy E, Teama NM. Clinical value of adding Dapagliflozin in patients with nephrotic syndrome. Int Urol Nephrol. 2024;56(11):3617–25. [DOI] [PubMed] [Google Scholar]
- 16. Hammad H, Shaaban A, Philips MV, Fayed A, Abdelaziz TS. Effect of sodium-glucose transporter 2 inhibitor empagliflozin on proteinuria and kidney function progression in patients with non-diabetic glomerulonephritis: a pilot superiority randomized controlled trial. Int Urol Nephrol. 2023;55(9):2321–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Nielsen SF, Duus CL, Buus NH, Bech JN, Mose FH. Randomized, placebo-controlled trial on the renal and systemic hemodynamic effects of empagliflozin. Kidney Int Rep. 2025;10(1):134–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Uchiyama K, Kamano D, Nagasaka T, Hama EY, Shirai R, Sumura R, et al. Open-label, randomized, controlled, crossover trial on the effect of dapagliflozin in patients with ADPKD receiving tolvaptan. Kidney Int Rep. 2025;10(4):1063–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Ihle BU, Whitworth JA, Shahinfar S, Cnaan A, Kincaid-Smith PS, Becker GJ. Angiotensin-converting enzyme inhibition in nondiabetic progressive renal insufficiency: a controlled double-blind trial. Am J Kidney Dis. 1996;27(4):489–95. [DOI] [PubMed] [Google Scholar]
- 20. Ishimitsu T, Kameda T, Akashiba A, Takahashi T, Ando N, Ohta S, et al. Effects of valsartan on the progression of chronic renal insufficiency in patients with nondiabetic renal diseases. Hypertens Res. 2005;28(11):865–70. [DOI] [PubMed] [Google Scholar]
- 21. Ishimitsu T, Akashiba A, Kameda T, Takahashi T, Ohta S, Yoshii M, et al. Benazepril slows progression of renal dysfunction in patients with non-diabetic renal disease. Nephrology. 2007;12(3):294–8. [DOI] [PubMed] [Google Scholar]
- 22. Lee YJ, Cho S, Kim SR, Jang HR, Lee JE, Huh W, et al. Effect of losartan on proteinuria and urinary angiotensinogen excretion in non-diabetic patients with chronic kidney disease. Postgrad Med J. 2011;87(1032):664–9. [DOI] [PubMed] [Google Scholar]
- 23. Renke M, Lizakowski S, Tylicki L, Rutkowski P, Knap N, Heleniak Z, et al. Aliskiren attenuates oxidative stress and improves tubular status in non-diabetic patients with chronic kidney disease-Placebo controlled, randomized, cross-over study. Adv Med Sci. 2014;59(2):256–60. [DOI] [PubMed] [Google Scholar]
- 24. Shen PC, He LQ, Yang XJ, Cao HX. Renal protection of losartan 50 mg in normotensive Chinese patients with nondiabetic chronic kidney disease. J Investig Med. 2012;60(7):1041–7. [DOI] [PubMed] [Google Scholar]
- 25. Tang SCW, Chan KW, Ip DKM, Yap DYH, Ma MKM, Mok MMY, et al. Direct renin inhibition in non-diabetic chronic kidney disease (DRINK): a prospective randomized trial. Nephrol Dial Transplant. 2021;36(9):1648–56. [DOI] [PubMed] [Google Scholar]
- 26. Apperloo EM, Gorriz JL, Soler MJ, Cigarrán Guldris S, Cruzado JM, Puchades MJ, et al. Semaglutide in patients with overweight or obesity and chronic kidney disease without diabetes: a randomized double-blind placebo-controlled clinical trial. Nat Med. 2025;31(1):278–85. [DOI] [PubMed] [Google Scholar]
- 27. Lincoff AM, Brown-Frandsen K, Colhoun HM, Deanfield J, Emerson SS, Esbjerg S, et al. Semaglutide and cardiovascular outcomes in obesity without diabetes. N Engl J Med. 2023;389(24):2221–32. [DOI] [PubMed] [Google Scholar]
- 28. Colhoun HM, Lingvay I, Brown PM, Deanfield J, Brown-Frandsen K, Kahn SE, et al. Long-term kidney outcomes of semaglutide in obesity and cardiovascular disease in the SELECT trial. Nat Med. 2024;30(7):2058–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Mårup FH, Thomsen MB, Birn H. Additive effects of dapagliflozin and finerenone on albuminuria in non-diabetic CKD: an open-label randomized clinical trial. Clin Kidney J. 2024;17(1):sfad249. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Wang QR, Wu L, Huang J, Pan H, Wang XF, Li L, et al. Efficacy and safety of finerenone in IgA nephropathy: an observational multicentre study. Clin Kidney J. 2025;18(5):sfaf125. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Guo Y, Zhang X, Zhao S, Wang Y, Wang L, Li Q, et al. Effect and safety of finerenone in patients with IgA nephropathy. J Inflamm Res. 2025;18:9755–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Cinti F, Laborante R, Cappannoli L, Morciano C, Gugliandolo S, Pontecorvi A, et al. The effects of SGLT2i on cardiac metabolism in patients with HFpEF: fact or fiction? Cardiovasc Diabetol. 2025;24(1):208. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Yau K, Cherney DZI, van Raalte DH, Wever BE. Kidney protective mechanisms of SGLT2 inhibitors: evidence for a hemodynamic effect. Kidney Int. 2024;105(6):1168–72. [DOI] [PubMed] [Google Scholar]
- 34. Agarwal R, Kolkhof P, Bakris G, Bauersachs J, Haller H, Wada T, et al. Steroidal and non-steroidal mineralocorticoid receptor antagonists in cardiorenal medicine. Eur Heart J. 2021;42(2):152–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Hu Y, Li C, Li S, Xu L, Wang Y, Dong B. Finerenone in chronic kidney disease: a comprehensive review of mechanisms, efficacy and safety. Diabetes Metab Res Rev. 2025;41(6):e70075. [DOI] [PubMed] [Google Scholar]
- 36. Ahwin P, Martinez D. The relationship between SGLT2 and systemic blood pressure regulation. Hypertens Res. 2024;47(8):2094–103. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Agarwal R, Joseph A, Anker SD, Filippatos G, Rossing P, Ruilope LM, et al. Hyperkalemia risk with finerenone: results from the FIDELIO-DKD Trial. J Am Soc Nephrol. 2022;33(1):225–37. [DOI] [PMC free article] [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 institutional data protection policies, but are available from the corresponding author upon reasonable request. Access to the data will be granted for noncommercial research purposes after review of a short research proposal and, where applicable, the signing of a data use agreement.




