Nephrology · Medical Literature
Comparative efficacy of lipid-lowering therapies on the cardio-renal-metabolic axis in diabetic kidney disease: a Bayesian network meta-analysis addressing residual CRM risk.
Abstract
In diabetic kidney disease (DKD), dyslipidemia accelerates renal decline and heart failure risk, making Cardio-Renal-Metabolic (CRM) risk reduction critical. The bidirectional relationship between cardiac and renal dysfunction further highlights the necessity of lipid-lowering therapy as a common intervention pathway; however, its comparative efficacy remains unclear. To compare the Cardio-Renal-Metabolic (CRM) outcomes of various lipid-lowering agents. We searched PubMed, Embase, Web of Science, and Cochrane Library for Randomized Controlled Trials (RCTs) up to May 10, 2025; a Bayesian network meta-analysis compared their effects. From 20 RCTs, Cerivastatin best reduced Total Cholesterol (TC) (Mean Difference (MD): -94.03, 95% CI: -185.37 to -2.16), while Simvastatin best lowered Low-Density Lipoprotein Cholesterol (LDL-C) (MD: -56.05, 95% CI: -101.64 to -11.66). For cardiorenal outcomes, Atorvastatin, Rosuvastatin, and Fenofibrate potentially improved Urine Albumin-to-Creatinine Ratio (UACR). Atorvastatin (MD: -3.19, 95% CI: -5.12 to -1.27) and Fenofibrate (MD: -1.44, 95% CI: -2.78 to -0.09) most robustly reduced cardiovascular event rates (CVER) in hyperlipidemic DKD. Lipid-lowering agents have divergent effects; Atorvastatin and Fenofibrate best reduce CVER despite variable renal effects, supporting a phenotype-driven CRM strategy prioritizing residual cardiovascular risk reduction. https://www.crd.york.ac.uk/PROSPERO/view/, identifier CRD420251049719.
Clinical relevance
Among available lipid-lowering therapies (statins, fibrates, and other agents), which agents provide the greatest cardio-renal-metabolic (CRM) benefit—across lipid, renal, and cardiovascular event outcomes—in adults with diabetic kidney disease (DKD) and hyperlipidemia? __SNAPSHOT__ {"studyDesign":"Systematic review with Bayesian network meta-analysis (NMA) of randomized controlled trials","setting":"Multi-database search (PubMed, Embase, Web of Science, Cochrane Library); international RCTs; literature search up to May 10, 2025","sampleSize":"20 RCTs included; total participant number not reported in the abstract","population":"Adults with diabetic kidney disease (DKD) and hyperlipidemia; specific CKD staging, eGFR thresholds, and baseline lipid criteria not reported in the abstract","intervention":"Various lipid-lowering agents including statins (cerivastatin, simvastatin, atorvastatin, rosuvastatin, lovastatin) and fibrates (fenofibrate), assessed individually via network comparisons","comparator":"Indirect comparisons via Bayesian network meta-analysis; specific active comparators and placebo arms across included RCTs; placebo or standard care assumed as network anchor but not explicitly stated in abstract","primaryOutcome":"Cardio-Renal-Metabolic (CRM) outcomes including lipid parameters (TC, LDL-C, TG, HDL-C), renal markers (eGFR, UACR, serum creatinine, 24-hour urine protein), and cardiovascular event rate (CVER); no single pre-specified primary endpoint declared in the abstract","followUp":"Not reported","fundingSource":"Not reported"}
Key findings
Across 20 RCTs analyzed by Bayesian NMA, an efficacy hierarchy was established across CRM domains. For lipid-lowering potency: cerivastatin ranked highest for total cholesterol (TC) reduction (MD: −94.03 mg/dL; 95% CI: −185.37 to −2.16), and simvastatin ranked highest for LDL-C reduction (MD: −56.05 mg/dL; 95% CI: −101.64 to −11.66). Both CIs are wide and approach the null boundary, raising uncertainty about the precision of these estimates. For cardiovascular event rate (CVER) reduction—the most clinically relevant endpoint—atorvastatin (MD: −3.19; 95% CI: −5.12 to −1.27) and fenofibrate (MD: −1.44; 95% CI: −2.78 to −0.09) demonstrated statistically significant reductions; fenofibrate's CI also approaches the null. For renal outcomes, atorvastatin, rosuvastatin, and fenofibrate were described as 'potentially improving' UACR, while lovastatin appeared advantageous for eGFR and triglyceride reduction; however, no quantitative effect sizes for these renal outcomes are reported in the abstract. A NNT for CVER reduction cannot be calculated from the abstract, as event rate denominators and absolute risk data are not reported. The abstract does not report adverse event data in detail. The authors additionally characterized fenofibrate's transient serum creatinine elevation as a likely benign hemodynamic effect ('Fenofibrate Paradox') rather than nephrotoxicity, though this interpretation is not independently validated within this NMA's dataset as reported.
Practice implications
This NMA establishes that lipid-lowering agents have divergent effects across CRM outcomes in DKD: agents with the strongest surrogate lipid-lowering efficacy (simvastatin for LDL-C, cerivastatin for TC) are not the same agents with the strongest cardiovascular event reduction (atorvastatin, fenofibrate). This dissociation between lipid biomarker effects and clinical event outcomes is clinically important and consistent with prior literature. The finding that atorvastatin and fenofibrate most robustly reduce CVER supports a phenotype-driven CRM strategy that prioritizes residual cardiovascular risk, particularly in patients with predominant hypertriglyceridemia and impaired HDL metabolism characteristic of CKD-associated dyslipidemia. The suggestion that lovastatin may preferentially improve eGFR warrants cautious interpretation as no quantitative data are provided in the abstract. The 'Fenofibrate Paradox' interpretation—attributing Scr elevation to a reversible hemodynamic mechanism—is clinically relevant for practitioners who have historically withheld fenofibrate in CKD; however, this mechanistic claim is not prospectively tested within this NMA. The findings nominally apply to hyperlipidemic DKD patients included in the constituent RCTs; the severity of CKD (eGFR staging), type of DKD, background therapy (e.g., RAS blockade, SGLT2i, GLP-1 RA), and proportion with established CVD are not reported and may substantially affect generalizability. Patients with advanced CKD (eGFR <30 mL/min/1.73 m²), those on dialysis, and those on contemporary multi-drug CRM regimens are likely underrepresented. Critically, only an abstract and partial full-text were available for this review; detailed SUCRA rankings, consistency analyses, heterogeneity assessments, and GRADE certainty ratings cannot be fully evaluated. Practice change on the basis of this NMA alone is premature. __LIMITATIONS__ Only 20 RCTs included; total sample size and individual trial sizes not reported, raising concerns about power and precision—several key CIs approach the null boundary.,Cerivastatin is no longer clinically available (withdrawn in 2001 due to fatal rhabdomyolysis risk), severely limiting the practical relevance of its top ranking for TC reduction.,Bayesian NMA relies on indirect comparisons across trials with heterogeneous populations, interventions, doses, follow-up durations, and outcome definitions; consistency and heterogeneity statistics are not reported in the abstract.,No single pre-specified primary outcome is declared; evaluation across multiple CRM endpoints without explicit hierarchical testing raises concerns about multiplicity.,UACR, eGFR, TC, and LDL-C are surrogate or intermediate endpoints; hard outcomes (ESKD, MACE, all-cause mortality, cardiovascular death) are not individually reported.,Absolute risk data, NNT, and NNH are not calculable from the abstract; clinical magnitude of CVER reductions expressed as mean differences requires context of baseline event rates.,Background therapy (RAS inhibitors, SGLT2i, GLP-1 RA, finerenone) is not reported; modern CRM combination therapy may substantially alter the incremental benefit of lipid-lowering agents.,CKD stage, eGFR distribution, and albuminuria severity of included participants are not described, limiting stratification and generalizability.,Funding source is not reported; potential conflicts of interest and sponsorship bias cannot be assessed.,Only abstract-level and partial full-text data were available for this summary; GRADE certainty of evidence ratings and full sensitivity analyses are not assessable. __BOTTOM_LINE_CATEGORY__ potentially_practice_informing __BOTTOM_LINE__ This Bayesian NMA of 20 RCTs in hyperlipidemic DKD patients demonstrates that agents with the greatest LDL-C or TC lowering potency (simvastatin, cerivastatin) are not the same agents with the greatest cardiovascular event rate reduction; atorvastatin and fenofibrate showed the most statistically significant CVER benefit and together with lovastatin (for eGFR/TG) represent a phenotype-driven CRM hierarchy. These findings support individualized lipid-lowering selection based on CRM risk profile rather than a uniform LDL-C–centric strategy, but should be interpreted cautiously given wide confidence intervals, surrogate-dominant endpoints, the clinical unavailability of cerivastatin, unreported background therapy, and the inability to fully assess evidence quality from the abstract alone. Confirmation via prospective RCTs incorporating contemporary multi-drug CRM regimens, hard clinical endpoints, and clearly defined CKD stages is needed before this hierarchy can be translated into a protocol change.