Chronic Kidney Disease

Definition

Chronic kidney disease (CKD) is defined by structural or functional kidney abnormalities persisting ≥3 months, characterized by reduced estimated glomerular filtration rate (eGFR <60 ml/min/1.73m²) and/or markers of kidney damage (albuminuria ≥30 mg/g, hematuria, structural abnormalities). CKD affects approximately 800 million adults globally (2023), more than doubling since 1990. While type 2 diabetes is the most common single cause, more than half of all CKD cases occur in persons without diabetes. CKD is a major independent risk factor for cardiovascular disease, with cardiovascular mortality exceeding the risk of progression to end-stage kidney disease (ESKD). sources/Finerenone-FIND-CKD-NEJM-2026 — very high

Key Concepts

Staging and Screening (ESC 2026)

• Stage by CGA — Cause, GFR, Albuminuria — displayed as the KDIGO heatmap (I A). G5 is persistent eGFR <15 and is now termed kidney failure, replacing "end-stage kidney disease". Albuminuria categories are A1 <3, A2 3–30 and A3 >30 mg/mmol (A3 ≈ ≥0.3 g/day). (sources/CVD-CKD-ESC-2026, rating: very high)
• Relative kidney-failure risk by heatmap colour: moderate 3–20-fold, high 20–70-fold, very high >70-fold, independent of age and sex.
• Screen every patient with cardiovascular disease — hypertension, CAD, HF, PAD or stroke — with eGFR plus a spot morning uACR (I C). Repeat annually in diabetes or high risk, and within days to weeks when a new or unexpected abnormality appears (to catch AKI). Creatinine-based eGFR with a validated equation and two values ≥3 months apart to establish chronicity (I C); cystatin C-based eGFR when creatinine is unreliable, such as at extremes of muscle mass or with trimethoprim-like tubular secretion interference (IIa A).
• The 2009 CKD-EPI equation without a race coefficient is "an optimal compromise for European populations"; the 2021 US-derived equation overestimates GFR in large parts of Europe.
• Kidney Failure Risk Equation in G3–G5 (I A): >3%–5% five-year risk prompts nephrology referral, >40% two-year risk prompts dialysis-access planning. The heatmap is for population triage; KFRE gives individual absolute risk.
• Referral triggers: KDIGO very-high risk, eGFR <30, uACR ≥70 mg/mmol, uACR ≥30 mg/mmol with haematuria, a ≥2-fold uACR rise from a baseline ≥30 mg/mmol, or metabolic complications of severe CKD.
• Dose drugs by absolute (de-indexed) GFR — eGFR × BSA/1.73 — at very high or very low body surface area; Cockcroft-Gault and 24-hour collections overestimate GFR through tubular secretion. CKD is itself a risk factor for QT prolongation.

Cardiovascular Risk in CKD

• CKD is a cardiovascular risk state; CVD is a kidney risk state. In a meta-analysis of 114 cohorts (~28 million people) eGFR <60 and/or albuminuria independently raised MI, stroke, HF, AF, PAD and cardiovascular mortality, with the highest risk at eGFR <30 and A3. Relative risks are similar regardless of age, sex, race and diabetes status. (sources/CVD-CKD-ESC-2026, rating: very high)
• Incident CVD predicts kidney failure most strongly in the first three months: 2–4-fold after CAD, stroke or AF, and 45-fold after incident heart failure.
• Risk categories: "relatively very high" = eGFR <30 with any albuminuria, or eGFR 30–44 with uACR 3–30, or uACR >30 at any eGFR (I B); "relatively high" = eGFR 30–44 with uACR <3, or eGFR 45–59 with uACR 3–30 (I B).
• Use SCORE2 or SCORE2-OP with the CKD Add-On in CKD without type 2 diabetes or ASCVD (I B), SCORE2-Diabetes in type 2 diabetes (I B) and SMART2 in established ASCVD (I B). Do not sum separate ASCVD and heart-failure estimates, and scores do not predict kidney failure.
• Routine antiplatelet therapy at eGFR <60 without symptomatic ASCVD is not recommended — III B1.

Epidemiology and Prognosis

• ~800 million adults affected globally in 2023; prevalence has more than doubled since 1990. sources/Finerenone-FIND-CKD-NEJM-2026 — very high
• Type 2 diabetes is the most common cause, but >50% of cases are in persons without diabetes. sources/Finerenone-FIND-CKD-NEJM-2026 — very high
• Progressive loss of kidney function and adverse cardiovascular outcomes occur in many persons with CKD despite guideline-recommended therapy, particularly those with severely increased albuminuria. sources/Finerenone-FIND-CKD-NEJM-2026 — very high
• Cardiovascular disease is the leading cause of death in CKD; the risk of CV death exceeds the risk of progressing to ESKD at most stages.

Pathophysiology

• Mineralocorticoid receptor overactivation: Implicated in CKD progression through increased sodium and fluid retention and activation of proinflammatory and profibrotic processes. This pathway is independent of aldosterone levels — cortisol can also activate the MR in states of tissue damage and oxidative stress. sources/Finerenone-FIND-CKD-NEJM-2026 — very high
• Glomerular hyperfiltration: Compensatory hyperfiltration in remaining nephrons leads to progressive glomerulosclerosis, proteinuria, and further nephron loss.
• Tubulointerstitial fibrosis: Final common pathway of all progressive CKD, regardless of etiology.

KDIGO 2024 Guideline-Directed Medical Therapy

• RAS inhibitors (ACEi/ARB): Established first-line therapy; reduce proteinuria and slow eGFR decline by 0.5–1.0 ml/min/1.73m²/year. Renoprotective in both diabetic and non-diabetic CKD. sources/Finerenone-FIND-CKD-NEJM-2026 — very high
• SGLT2 inhibitors: Class I recommendation; slow eGFR decline; reduce ESKD and CV death. Benefit across diabetic/non-diabetic CKD, all eGFR strata ≥20, and all albuminuria categories (see concepts/SGLT2-Inhibitors-in-CKD). sources/dapagliflozin-dapackd-nejm-2020 — very high; sources/empagliflozin-empakidney-nejm-2025 — very high
• Nonsteroidal MRA (finerenone): FIND-CKD (NEJM 2026) establishes finerenone as the third pillar of CKD pharmacotherapy — slows eGFR decline by 0.7 ml/min/1.73m²/year and reduces kidney/CV events (HR 0.77; P=0.04) in non-diabetic CKD. Previously established in diabetic CKD (FIDELIO-DKD, FIGARO-DKD). (See entities/Finerenone). sources/Finerenone-FIND-CKD-NEJM-2026 — very high

eGFR Slope as a Surrogate Endpoint

• eGFR slope (annual rate of change) is increasingly used as a primary endpoint in CKD trials.
• A meta-analysis of 66 RCTs found that a 0.75 ml/min/1.73m² attenuation of eGFR decline predicted a 23% reduction in adverse kidney outcomes. sources/Finerenone-FIND-CKD-NEJM-2026 — very high
• The two-slope model (acute effect up to month 3 + chronic slope thereafter) separates the initial hemodynamic effect from the disease-modifying effect on kidney function decline.
• Finerenone causes an acute eGFR increase (opposite to SGLT2i acute dip), attributed to reduced intraglomerular pressure through efferent arteriole dilation — the total eGFR slope may therefore underestimate the true chronic benefit.

Albuminuria as a Therapeutic Target and Surrogate

• Albuminuria (UACR) is both a marker of kidney damage and an independent risk factor for CKD progression and CV events.
• Finerenone reduced UACR by 41.3% vs 9.1% at 6 months (relative reduction 35.4%); 56.0% achieved ≥30% UACR reduction with finerenone vs 24.4% with placebo (OR 3.99). sources/Finerenone-FIND-CKD-NEJM-2026 — very high
• Early UACR reduction is a validated surrogate for long-term kidney outcomes.

Hyperkalemia Management in CKD Patients on RAAS + MRA

• Hyperkalemia is the principal adverse effect of dual RAS + MRA blockade.
• FIND-CKD: any hyperkalemia 17.0% finerenone vs 13.3% placebo; K⁺ >5.5 in 18.8% vs 12.2%; K⁺ >6.0 in 4.3% vs 2.3%.
• Serious hyperkalemia events (<1% hospitalization, 1.5% vs 0.1% permanent discontinuation, no fatal events). sources/Finerenone-FIND-CKD-NEJM-2026 — very high
• Protocol-mandated K⁺ monitoring and dose adjustment algorithm successfully prevented serious outcomes.
• In contrast, BARACK-D trial: >50% of CKD patients discontinued spironolactone (steroidal MRA) within 6 months, primarily due to safety criteria including hyperkalemia — finerenone's tolerability profile appears superior to steroidal MRAs in CKD. sources/Finerenone-FIND-CKD-NEJM-2026 — very high

Non-Diabetic CKD — Specific Considerations

• Causes: Hypertension/nephrosclerosis, chronic glomerulonephritis (IgA nephropathy, FSGS, membranous), tubulointerstitial disease, autosomal dominant polycystic kidney disease (ADPKD), lupus nephritis, ANCA-associated vasculitis, and other rare diseases.
• FIND-CKD population: Patients with ADPKD, lupus nephritis, and ANCA vasculitis were EXCLUDED; the trial population mainly represents hypertensive/glomerular CKD. Efficacy in excluded etiologies is unknown.
• Evidence base historically smaller than diabetic CKD: RAS inhibitors (landmark trials including AASK, REIN) and SGLT2i (DAPA-CKD 32.5% non-diabetic; EMPA-KIDNEY) established efficacy before FIND-CKD.

Three-Pillar Pharmacotherapy for Non-Diabetic CKD

• All three classes have shown benefit in both diabetic and non-diabetic CKD.
• Whether triple therapy (RASi + SGLT2i + finerenone) provides additive benefit over dual therapy is plausible but not yet proven in a dedicated RCT.

Four-Pillar Cardiorenal Pharmacotherapy (ESC 2026)

The 2026 ESC/ERA guideline formalises a fourth pillar and attaches an initiation threshold to each, taken directly from trial eligibility criteria. (sources/CVD-CKD-ESC-2026, rating: very high)

• SGLT2 inhibitors reduce kidney disease progression by ~40% irrespective of baseline eGFR, with a class effect across agents; there is no effect on MI or stroke, and the cardiovascular benefit is largely heart-failure hospitalisation and cardiovascular or sudden death. Continue below eGFR 20 until dialysis begins.
• Finerenone: FIDELITY kidney composite HR 0.77 (0.67–0.88), cardiovascular composite HR 0.86 (0.78–0.95) driven by heart-failure hospitalisation HR 0.78. Initiate when potassium is ≤5.0 mmol/L.
• GLP-1 receptor agonists: FLOW primary kidney composite HR 0.76 (0.66–0.88), eGFR slope reduced by about a third, MACE HR 0.82 (0.68–0.98). Not tested in non-diabetic CKD.
• A <30% eGFR dip on initiation is expected and acceptable — it is haemodynamic, not nephron injury. Check eGFR and potassium 1–4 weeks after starting or increasing an ACE inhibitor, ARB or finerenone, but no extra biochemical monitoring is routinely needed for SGLT2 inhibitors or GLP-1 receptor agonists. Stop ACE inhibitor, ARB or finerenone at potassium ≥6.0; apply sick-day rules during intercurrent illness and restart promptly.
• Co-prescription is encouraged on complementary mechanisms — SGLT2 inhibitor plus MRA lowers albuminuria more than either alone, and simultaneous empagliflozin and finerenone initiation is safe, with a larger acute dip.

Cardiac Troponin in CKD

• Troponin is three times more likely to be elevated in CKD: cTnT above the 99th percentile URL in 70% versus 24% of patients without CKD, and cTnI in 47% versus 16%. The frequency rises six-fold, from 10% at eGFR ≥90 to 66% at eGFR <30 mL/min/1.73 m², and CKD patients are on average 20 years older. (sources/5thMI-AHA-2026, rating: very high)
• Half of CKD patients with elevated troponin have injury without infarction, yet the final diagnosis is primary MI more often than in non-CKD patients (24% versus 11%). Elevated troponin should not be dismissed as reduced clearance. (sources/5thMI-AHA-2026)
• Clearance is dual: troponin is cleared renally at normal concentrations and hepatically at high ones, so halving eGFR doubles basal troponin. Specificity and positive predictive value of the 99th percentile are correspondingly lower in CKD; higher thresholds or larger absolute deltas would improve PPV but would need per-stage, per-assay values that do not yet exist. (sources/5thMI-AHA-2026)

Contradictions / Open Questions

• eGFR slope vs clinical outcomes: FIND-CKD primary endpoint (eGFR slope) was positive while the kidney composite narrowly missed significance (P=0.06). This reflects the trial being powered for the more sensitive eGFR slope endpoint, not clinical events. Whether the 0.7 ml/min/1.73m²/year eGFR benefit will translate into clinically meaningful reductions in ESKD over longer follow-up remains to be confirmed. sources/Finerenone-FIND-CKD-NEJM-2026 — very high
• Finerenone vs steroidal MRAs in non-diabetic CKD: No head-to-head comparison exists. BARACK-D data suggest steroidal MRAs (spironolactone) are poorly tolerated in CKD, but efficacy relative to finerenone cannot be inferred from cross-trial comparisons. sources/Finerenone-FIND-CKD-NEJM-2026 — very high
• Triple therapy (RASi + SGLT2i + finerenone): Only 17% of FIND-CKD participants were on SGLT2i at baseline. The hypothesis that finerenone provides additive benefit on top of SGLT2i + RASi dual therapy is plausible based on complementary mechanisms (hemodynamic + metabolic + anti-fibrotic) but was not directly tested. sources/Finerenone-FIND-CKD-NEJM-2026 — very high
• Generalizability to all CKD etiologies: ADPKD, lupus nephritis, and ANCA vasculitis were excluded. sources/Finerenone-FIND-CKD-NEJM-2026 — very high
• Hyperkalemia risk in real-world practice: FIND-CKD had protocol-directed monitoring and dose adjustment. Real-world hyperkalemia rates with less intensive monitoring could be higher, particularly in patients with lower eGFR or higher baseline K⁺. sources/Finerenone-FIND-CKD-NEJM-2026 — very high
• CV outcomes underpowered: CV composite HR 0.60 (95% CI 0.27–1.33) is directionally favorable but non-significant. A larger/longer trial is needed to establish CV benefit in non-diabetic CKD. sources/Finerenone-FIND-CKD-NEJM-2026 — very high
• Trial-eligibility thresholds are presented as treatment thresholds. The eGFR and uACR cut-offs in the four-pillar table (eGFR 20, 25, 30; uACR 10, 20, 30 mg/mmol) are enrolment criteria and health-economic analyses rather than biological boundaries, and the guideline concedes that within-person variability can misclassify patients without offering a firm rule. (sources/CVD-CKD-ESC-2026, rating: very high)
• The guideline contradicts itself on two points. The narrative extends GLP-1 receptor agonists to eGFR 25–89 with uACR ≥10 while the recommendation table narrows the band; and the key messages call it "reasonable" to start SGLT2 inhibitors below eGFR 20 while the recommendation table and the heart-failure section decline to recommend initiation there.
• CKD in HFpEF vs CKD without HF: Finerenone reduced kidney events in CKD trials (FIDELIO-DKD, FIGARO-DKD, FIND-CKD) but the kidney composite was numerically worse in FINEARTS-HF (HR 1.33) — the nephroprotective effect may be albuminuria/CKD-dependent and absent in HFpEF patients with low-level albuminuria. See entities/Finerenone for full discussion. sources/finerenone-hfpef-fineartshf-nejm-2024 — very high

Connections

• Related to entities/Finerenone — third pillar of CKD pharmacotherapy
• Related to concepts/SGLT2-Inhibitors-in-CKD — second pillar; complementary mechanism (hemodynamic vs anti-fibrotic)
• Related to concepts/Mineralocorticoid-Receptor-Antagonists-Post-MI — broader MRA evidence framework
• Related to concepts/Cardiorenal-Syndrome — bidirectional HF-CKD interaction
• Related to concepts/Aldosterone-Synthase-Inhibitors — alternative aldosterone pathway strategy
• Related to entities/HFpEF — CKD is a common comorbidity in HFpEF

Sources

• sources/Finerenone-FIND-CKD-NEJM-2026
• sources/dapagliflozin-dapackd-nejm-2020
• sources/empagliflozin-empakidney-nejm-2025
• sources/finerenone-hfpef-fineartshf-nejm-2024
• sources/5thMI-AHA-2026
• sources/CVD-CKD-ESC-2026