CKD-Heart: Heart Failure Prediction in Asian CKD Patients

1.1 Background and rationale for conducting this study

Patients with chronic kidney disease (CKD), a clinical scenario closely associated with multiple co-morbid conditions (e.g., hypertension, diabetes, and coronary artery diseases), are less likely to receive guideline-directed foundational medical therapy due to clinical concerns about hyperkalemia and hypotension, and are thereby prone to the development of heart failure (HF). In this regard, early identification and prevention of HF is important (1). To date, N-terminal pro-brain natriuretic peptide (NT-pro-BNP) has been shown to be a powerful clinical predictor of HF, though its clinical utilization is easily affected by renal insufficiency, thus limiting its use in these patients (2).

1.2 Rationale for study design, doses and control groups

Overall rationale and study population:

Recent studies have shown improvement in cardiovascular outcomes, particularly in HF, and renal protective effects of SGLT2 inhibitors (3). Early identification of HF among CKD patients in a pre-clinical HF stage may allow appropriate pharmacological interventions as early as possible to reduce future HF burden. Yet, data regarding the exact clinical characteristics and risk stratification for HF diagnosis in CKD subjects that may benefit from interventions remain largely unexplored.

Whether the use of several key baseline clinical risk factors, including baseline eGFR (using MDRD or CKD-EPI formula), Urine albumin to creatinine ratio (UACR), serum biomarkers (NT-proBNP, hs-CRP, or hs-TNI), either combined or alone, among CKD ethnic Asians at a pre-clinical HF stage (without prevalent HF) or at a time point where early therapeutic intervention may be beneficial for such patients in daily practice remains to be determined from large-scale real-world datasets. We further aim to explore the clinical cut-off of NT-proBNP in identifying incident HF in CKD patients applicable to ethnic Asians based on a large-scale CKD population.

Prior evidence: 51% of heart failure patients have an eGFR <60 mL/min/1.73 m2 and 11% have an eGFR <30 mL/min/1.73 m2. Increasing mortality was found as kidney function decreased, regardless of the presence of diabetes, age, hemoglobin levels, New York Heart Association (NYHA) class, and duration of heart failure (4). Kidney dysfunction is not only correlated with short-term and long-term outcomes in patients with heart failure but is also a predictor for the risk of heart failure (5).

Objectives:

Use several key clinical risks, for example, Framingham heart study risk factors and variables collected in the CRIC study (6), to early identify HF events among CKD patients free from prevalent HF; Explore the clinical cut-off of NT-proBNP in identifying HF in CKD patients in Taiwan.

Unique aspects: The risk factors identified by the CRIC (Chronic Renal Insufficiency Cohort) study are not commonly used for clinical daily practice, for example, insulin resistance, inflammation, and hsTnI (6, 7). We would like to use CRIC clinical scores including clinical demographics (e.g., age, blood pressure, smoking, and medical comorbidities), echo-based parameters (e.g., LVM, LVEF), biomarkers (e.g., hsTnT and NT-proBNP), and renal profiles (eGFR and UACR), to identify high-risk HF in CKD patients. Besides, the cut-offs of eGFR/UACR/hs-TnI/NT-proBNP to identify HF among large-scale Asian CKD patients are not clear.

Expected data:

Identify several key clinical risk factors that are easy to obtain for clinical practice to early identify HF in CKD among a large-scale Asian cohort.

Find out the optimal cutoff of eGFR (using different formulas, such as MDRD or CKD-EPI) and UACR in predicting HF incidence in CKD patients.

1.3 Benefit/risk and ethical assessment

Benefit:

Identification of HF predictors among CKD patients in pre-clinical HF stage (Stage A-B) prior to overt clinical HF.

Besides, eGFR/UACR, hs-TnI/NT-proBNP checkup as routine to identify HF risk may not be generalizable in asymptomatic CKD patients.

Risk:

Not applicable

Ethical assessment:

This protocol will be reviewed by IRB.

1.4 Study Design

Study group:

Retrospective study: Pooled 9000-10000 CKD patients from Mackay Memorial Hospital (2-center pooled data) Data collection range: retrospective, starting from 2011-2020 Data collection timeline: half a year

Several clinical risk factors and NT-ProBNP level at baseline and subsequent vascular events up to 5 years in a retrospective manner will be collected. Associations of these clinical risks with the development of HF and vascular events will be established.

Data collection content:

• Framingham risk factors and CRIC study-based clinical variables including age, HDL-C, total cholesterol, blood pressure, HbA1c, smoking habit, IL-6, TNF-α, insulin resistance, anemia, albuminuria or ACR (urine albumin/creatinine ratio), and hs-TnI will also be collected if available.
• Data regarding conventional echocardiography parameters for indexed LV mass and LV ejection fraction (LVEF) will also be collected if available.
• Diabetes and dialysis duration will be collected if available.
• Long-term used (i.e., with ≥ 6 months treatment duration) medications that may affect renal function and heart function will be collected (ACEi/ARB, MRA, ARNI, SGLT2i, diuretic, TZD, GLP-1 RA, and NSAID)

Follow-up: clinical events (vascular events including stroke, myocardial infarction, unstable angina or elective PCI, atrial fibrillation (AF), atrial flutter (AFL), heart failure (HF), and any cause of death event) up to 5 years in a retrospective manner following the index date of data collection for any individual.

Echocardiography Sub-study:

• Echocardiography features in symptom-free CKD of ethnic Asians
• Prevalence and prognostic utilization of diastolic dysfunction in symptom-free CKD among ethnic Asians

Inclusion criteria

• Clinical diagnosis of CKD (an estimated glomerular filtration rate (eGFR) of less than 60 mL/min/1.73 m², persisting for 3 months or more)
• Age ≥ 20 years
• Free at baseline of diagnosed HF
• Baseline patient-level clinical data available

Exclusion criteria

• Known HF hospitalization history
• Known impaired LV systolic function (LVEF<40%) by any imaging modality
• Any terminal disease other than CKD lowering the patient's life expectancy to less than two years
• eGFR ≥ 60ml/min/1.73m2