Abstract
Aims
Differences between cystatin C-based and creatinine-based glomerular filtration rate estimations (eGFRdiff) have been associated with adverse outcomes. We aimed to investigate the prevalence of eGFRdiff and its association with clinical outcomes and brain lesions in patients with atrial fibrillation (AF).
Methods and results
We analysed 3804 patients with AF enrolled in two prospective, multicentre cohort studies. The eGFRdiff was calculated by ‘cystatin C-based eGFR—creatinine-based eGFR’. Patients were stratified into three groups: eGFRdiff >10, −10 to 10, and <−10 mL/min/1.73 m². Associations with high-sensitivity troponin T (hs-TnT), NT-proBNP, major adverse cardiovascular (CV) events (stroke/systemic embolism, myocardial infarction and CV death), first heart failure (HF) hospitalization, and magnetic resonance imaging (MRI)-detected brain lesions were assessed. The mean age was 71 years, 28% were female, and 49% had paroxysmal AF. An eGFRdiff of >10, −10 to 10, and <−10 mL/min/1.73 m2 was present in 17.7%, 71.7%, and 10.6% of the patients, respectively. Incidence rates per 100 patient-years increased across these groups for major CV events (MACEs: 1.7, 3.9, 6.8) and HF hospitalization (1.2, 3.5, 6.1). Similarly, the prevalence of MRI-detected brain infarcts (28%, 39%, 42%) and moderate-to-severe white matter hyperintensities (38%, 55%, 63%) increased with greater eGFRdiff. The levels of hs-TnT and NT-proBNP were inversely correlated with eGFRdiff. In multivariable-adjusted linear models, lower eGFRdiff was associated with MACE (P = 0.006) and HF hospitalization (P = 0.015).
Conclusion
A substantial proportion of patients with AF exhibit clinically relevant eGFRdiff. More negative eGFR difference was associated with elevated cardiac biomarkers, greater burden of ischaemic brain lesions, and increased risk of CV events and HF hospitalization.
ClinicalTrials.gov Identifier
Keywords: atrial fibrillation, cystatin C, creatinine, glomerular filtration rate, biomarker
Graphical Abstract
Graphical Abstract.
What’s new?
Nearly 30% of patients with atrial fibrillation exhibit clinically relevant differences between creatinine-based and cystatin C-based glomerular filtration rate estimates (eGFRdiff).
A more negative eGFRdiff identifies a high-risk phenotype with four-fold higher rates of major cardiovascular events and five-fold higher heart failure hospitalization.
eGFR discrepancies correlate with elevated cardiac biomarkers and increased burden of MRI-detected brain lesions.
Current reliance on creatinine-based estimates alone may lead to inappropriate anticoagulant dosing in a substantial proportion of atrial fibrillation patients.
Introduction
About one-third of patients with atrial fibrillation (AF) have at least moderate renal impairment,1 which associates with increased risks of ischaemic events, heart failure (HF) hospitalization, and all-cause mortality.2–4 Kidney function in clinical practice is typically assessed using estimated glomerular filtration rate (eGFR) with creatinine as the standard biomarker. While creatinine-based eGFR may be sufficiently accurate in most clinical scenarios, it has several limitations such as dependency on nutritional status, body composition, and muscle mass, especially in elderly or comorbid patients.5,6 In contrast, cystatin C is less affected by these variables, and combined creatinine and cystatin C equations for eGFR are generally considered more accurate and robust.7
Intraindividual differences between creatinine-based and creatinine and cystatin C-based eGFR are common in chronic kidney disease (CKD).8 Moreover, eGFR differences have been associated with an increased risk of mortality in patients with CKD and HF.8–10 Specifically, when creatinine-based eGFR exceeds cystatin C-based estimates, this may reflect overestimation of renal function in advanced HF, with potential implications for drug dosing and prognostication.9 As HF remains the leading cause of death among patients with AF and considering the importance of appropriate dosing of direct oral anticoagulants (DOACs),11,12 discrepancies in eGFR estimation may have important clinical implications in the ever growing population.
In this extended analysis of a previous reported finding,4 we investigated the prevalence and extent of differences in eGFR between creatinine- vs. cystatin C-based equations, and their association with clinical outcomes and brain MRI-detected brain lesions in a large, multicentre cohort of patients with AF.
Methods
Study design
For this analysis, we combined data from two prospective, observational multicentre cohort studies, the Swiss Atrial Fibrillation Study (Swiss-AF) and the Basel Atrial Fibrillation Study (Beat-AF). Details of each study have been previously described.13–15 Briefly, both studies enrolled patients with documented AF and had similar study designs. In Swiss-AF, patients aged ≥65 years were enrolled, with a sub-population (10%) of patients <65 years to study health-related work absences, whereas there was no age restriction in Beat-AF. Main exclusion criteria in both studies were any acute illness in the last 4 weeks, non-sustained secondary forms of AF (e.g. due to surgery or infection), and inability to provide written informed consent. The screening and recruitment were performed on in- and outpatients in the participating hospitals and by contacting general practitioners. The studies were approved by the lead ethics committee, and all study centres received approval from the individual regional ethical committees.
Overall, the Swiss-AF study enrolled 2415 patients in 14 study centres between 2014 and 2017, and the Beat-AF study enrolled 1545 in 7 centres in Switzerland between 2010 and 2014. Of the 3960 enrolled patients, 156 had missing values for either biomarker or had no follow-up data, and were excluded from this analysis, which left 3804 individuals for the current analysis. The study flow is shown in Supplementary material online, Figure S1.
Biomarkers and GFR estimation
Non-fasting venous blood samples were collected at enrolment. Blood samples were centrifuged, aliquoted in cryotubes, and stored in a biobank at −80°C. Biomarkers were measured after the enrolment of the last patient. Cystatin C and creatinine levels were measured on a Roche Cobas analyser (Roche, Switzerland). Nine biomarker values were below the lower limit of detection (0.4 mg/L for cystatin C, 5 μmol/L for creatinine) and were set to half of the biomarker specific cut-off. Two participants exceeded the upper limit of quantification for cystatin C, so the values were set to the specific lower limit cut-off of 6.8 mg/L.
We used the latest sex- and age-specific Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) formula to estimate a creatinine-based GFR (eGFRcrea) and a creatinine and cystatin C-based GFR (eGFRcrea-cys) (see Supplementary material online, Table S1).7 The difference in eGFR (eGFRdiff) was calculated as eGFRcrea-cys—eGFRcrea. Accordingly, higher eGFR by creatinine compared to creatinine and cystatin C results in a negative eGFRdiff and vice versa.
Clinical outcomes
The two main clinical outcomes were major CV events (MACEs) and HF hospitalizations. The major CV event was defined as a composite of any stroke, systemic embolism, myocardial infarction, and cardiovascular (CV) death. Secondary outcomes included the individual components of MACE, all-cause death, and major bleeding. Detailed outcome definitions are provided in the Supplementary material. All clinical outcomes were assessed yearly by telephone, clinical visit and/or chart review, and were blindly validated by two independent clinicians. In case of disagreement, a third reviewer was consulted.
Brain magnetic resonance imaging
Brain magnetic resonance imaging (MRI) at study enrolment was available for 1722 patients from the Swiss-AF study. 693 patients from the Swiss-AF study did not undergo MRI, mainly due to cardiac devices or claustrophobia. No brain MRI was performed in Beat-AF. Details of the standardized brain MRI protocols were previously published.16,17 Ischaemic brain infarcts were classified as large noncortical or cortical infarcts (LNCCIs) or small noncortical infarcts (SNCIs). Noncortical infarcts >20 mm were defined as large noncortical infarcts, while cortical infarcts were defined as hyperintense lesions on FLAIR involving the cortex, regardless of the size. The SNCIs were defined as lesions measuring ≤20 mm on FLAIR without involving the cortex. White matter hyperintensities (WMHs) were graded according to the Fazekas scale. A score ≥2 was defined as at least moderate disease and considered relevant for this analysis.18 Microbleeds were identified as nodular, strongly hypointense lesions on T2*weighted or susceptibility-weighted imaging.
Other study variables
Patient demographics, past and current medical history, lifestyle factors, and medication were assessed by trained study personnel using standardized case report forms. At baseline, blood pressure was measured three times. The mean of all measurements was calculated and used for this analysis. A 12- or 16-lead electrocardiogram was performed, and the rhythm was defined as sinus rhythm, AF, or any other rhythm. The AF type classification was performed using the 2010 Guidelines of the Task Force for the Management of Atrial Fibrillation of the European Society of Cardiology.19
Statistical analysis
Baseline characteristics and MRI data are shown overall and stratified by three eGFRdiff categories: <−10, −10 to 10, and >10 mL/min/1.73 m2. These categories were chosen based on prior evidence in patients with HF.9 Continuous data are presented as mean ± standard deviation or median (interquartile range) and compared with one-way ANOVA or Kruskal–Wallis tests, in case strongly skewed. Categorical data are presented as numbers (percentages) and compared with Pearson’s Chi-square tests.
The distribution of eGFRdiff was plotted for the whole study population. Cystatin C, creatinine, and their respective eGFR are plotted against NT-proBNP and hs-TnT levels. Incidence rates of the main and secondary outcomes were calculated as the number of events per 100 patient-years. Similarly, the Kaplan–Meier curves for each main outcome are stratified by the three predefined eGFRdiff groups and group comparisons were performed using the log-rank test. In addition, we used eGFRdiff as a continuous and categorial variable in two different Cox proportional hazards models to estimate hazard ratios of clinical outcomes. We used eGFRdiff as a categorical variable using the predefined three-group classification. Categorial models were constructed with the eGFRdiff group −10 to 10 mL/min/1.73 m2 as the reference group. For continuous models, we used natural cubic splines as the associations of eGFRdiff and the main outcomes were not linear. First, we used age- and sex-adjusted models. Second, we additionally adjusted for diabetes, arterial blood pressure, history of stroke/TIA, and smoking status. The proportional hazards assumption was tested using the Schoenfeld residuals. Additionally, interaction terms between eGFR difference (modelled continuously and categorically) and HF status at baseline were included in Cox proportional hazards models for the main outcomes to assess potential effect modification. As we found no significant interaction (P for interaction 0.45 for MACE, 0.61 for HF hospitalization), we did not perform stratified analyses in patients with and without HF at baseline. For the two main clinical outcomes, we estimated the sensitivity and specificity of our pre-specified eGFRdiff cut-offs (−10 and 10 mL/min/1.73 m2, respectively) and calculated an optimal cut-off using the Youden index for different follow-up periods. We additionally performed time-dependent calculations of the area under the curve (AUC) at 10 years for MACE and HF hospitalization using inverse probability of censoring weights (IPCW) for eGFRdiff, eGFRcrea vs. eGFRcrea-cysc and investigated the increase in AUC by adding eGFRdiff to eGFRcrea or eGFRcrea-cys, respectively, in the multi-adjusted models.
All analyses were performed using R (version 4.4.2), and a two-sided P-value of 0.05 was considered statistically significant. No correction for multiple testing was done.
Results
Baseline characteristics
Baseline characteristics are shown in Table 1. The mean age of the population was 71 ± 10 years, 2743 (72%) were male. The mean CHA2DS2-VASc score was 3.2 ± 1.7 and 1862 (49%) patients had paroxysmal AF. Baseline characteristics stratified by study cohort are shown in Supplementary material online, Table S1.
Table 1.
Baseline characteristics stratified by the combined eGFR categories
| Overall | eGFR difference (mL/min/1.73 m2) | P-value | |||
|---|---|---|---|---|---|
| <−10 | −10 to 10 | >10 | |||
| N | 3804 | 403 | 2727 | 674 | |
| Age (years) | 71 (10) | 74 (9) | 73 (9) | 64 (10) | <0.001 |
| Sex (male) | 2743 (72%) | 283 (70%) | 1914 (70%) | 546 (81%) | <0.001 |
| BMI (kg/m2) | 27.4 (4.7) | 27.9 (5.5) | 27.4 (4.7) | 27.1 (4.2) | 0.481 |
| Systolic BP (mmHg) | 134 (19) | 133 (21) | 134 (19) | 134 (17) | 0.577 |
| Diastolic BP (mmHg) | 78 (12) | 76 (12) | 77 (12) | 80 (11) | <0.001 |
| Heart rate (bpm) | 66 (58; 78) | 70 (61; 80) | 66 (59; 78) | 64 (57; 75) | <0.001 |
| Active smoking | 298 (7.9%) | 46 (11%) | 198 (7.3%) | 54 (8.0%) | 0.015 |
| CHA2DS2-VASc Score | 3.2 (1.7) | 3.8 (1.6) | 3.4 (1.7) | 2.1 (1.6) | <0.001 |
| AF type | <0.001 | ||||
| Paroxysmal | 1862 (49%) | 155 (39%) | 1311 (48%) | 396 (59%) | |
| Non-paroxysmal | 1939 (51%) | 247 (61%) | 1414 (52%) | 278 (41%) | |
| Rhythm at baseline | <0.001 | ||||
| SR | 1871 (50%) | 131 (33%) | 1288 (47%) | 452 (67%) | |
| AF | 1466 (39%) | 214 (54%) | 1082 (40%) | 170 (25%) | |
| Other | 441 (11%) | 50 (13%) | 343 (13%) | 48 (7%) | |
| Medical history | |||||
| Prior pulmonary vein isolation | 807 (21%) | 54 (13%) | 545 (20%) | 208 (31%) | <0.001 |
| Coronary artery disease | 1017 (27%) | 119 (30%) | 792 (29%) | 106 (16%) | <0.001 |
| Prior stroke/TIA | 647 (17%) | 87 (22%) | 474 (17%) | 86 (13%) | <0.001 |
| Hypertension | 2626 (69%) | 303 (75%) | 1946 (71%) | 377 (56%) | <0.001 |
| Heart failure | 903 (24%) | 135 (33%) | 685 (25%) | 83 (12%) | <0.001 |
| Diabetes | 614 (16%) | 86 (21%) | 460 (17%) | 68 (10%) | <0.001 |
| Prior myocardial infarction | 562 (15%) | 58 (14%) | 448 (16%) | 56 (8.3%) | <0.001 |
| Prior major/clinically relevant non-major bleeding | 501 (13%) | 72 (18%) | 387 (14%) | 42 (6.2%) | <0.001 |
| Prior systemic embolism | 190 (5.0%) | 30 (7.4%) | 135 (5.0%) | 25 (3.7%) | 0.024 |
| Biomarker and eGFR | |||||
| Creatinine (mg/dL) | 1.10 (0.94; 1.31) | 0.98 (0.86; 1.13) | 1.13 (0.96; 1.37) | 1.05 (0.93; 1.19) | <0.001 |
| Cystatin C (mg/dL) | 1.12 (0.93; 1.41) | 1.41 (1.23; 1.70) | 1.17 (1.00; 1.44) | 0.80 (0.72; 0.89) | <0.001 |
| eGFRcrea (mL/min/1.73 m2) | 66.88 (53.17; 80.87) | 73.05 (61.65; 85.74) | 63.02 (49.35; 77.45) | 76.11 (64.04; 87.10) | <0.001 |
| eGFRcrea-cys (mL/min/1.73 m2) | 67.98 (51.21; 84.66) | 58.22 (46.78; 70.38) | 63.73 (47.96; 78.55) | 93.49 (81.11; 104.27) | <0.001 |
| eGFRdiff (mL/min/1.73 m2) | 0.73 (−4.90; 7.42) | −13.50 (−16.73; −11.57) | −0.06 (−3.96; 4.03) | 14.89 (12.16; 19.09) | <0.001 |
| Hs-TnT (ng/L) | 12.97 (8.42; 21.15) | 18.40 (12.32; 33.01) | 13.79 (9.13; 22.27) | 8.37 (6.06; 11.76) | <0.001 |
| NT-proBNP (ng/L) | 571 (176; 1;343) | 1;086 (433; 1;871) | 650 (218; 1;448) | 178 (80; 558) | <0.001 |
| Medication | |||||
| β Blockers | 2617 (69%) | 295 (73%) | 1910 (70%) | 412 (61%) | <0.001 |
| ACE inhibitors | 1099 (29%) | 124 (31%) | 829 (30%) | 146 (22%) | <0.001 |
| Calcium antagonist | 759 (20%) | 567 (21%) | 91 (23%) | 101 (15%) | 0.001 |
| ARB | 1153 (30%) | 863 (32%) | 133 (33%) | 157 (23%) | <0.001 |
| Digoxin | 158 (4%) | 106 (4%) | 35 (9%) | 17 (3%) | <0.001 |
| Diuretics | 1597 (42%) | 1232 (45%) | 244 (61%) | 121 (18%) | <0.001 |
| DOAK | 1342 (35%) | 160 (40%) | 1010 (37%) | 172 (26%) | <0.001 |
| Vitamin K antagonists | 1856 (49%) | 192 (48%) | 1321 (48%) | 343 (51%) | 0.448 |
| Antiplatelet therapy | 820 (22%) | 101 (25%) | 593 (22%) | 126 (19%) | 0.045 |
| AAD, Class Ic and III | 891 (23%) | 47 (12%) | 612 (22%) | 232 (34%) | <0.001 |
Data are presented as mean (standard deviation), median (interquartile range), or numbers (percentage).
AAD, antiarrhythmic drugs; ACE, angiotensin converting enzyme; AF, Atrial fibrillation; ARB, angiotensin-II receptor blocker; BMI, body mass index; BP, blood pressure; DOAC, direct oral anticoagulant; ECG, electrocardiogram; eGFR, estimated glomerular filtration rate; Hs-TNT, high-sensitivity troponin T; SR, sinus rhythm; TIA, transient ischaemic attack; VKA, vitamin K antagonist.
The eGFRdiff showed a bell-shaped distribution (Figure 1) and the median eGFRdiff was 0.73 (−4.90; 7.42) mL/min/1.73 m2. Among the 3804 patients included, 403 (10.6%) had an eGFRdiff <−10 mL/min/1.73 m2, 2727 (71.6%) had an eGFRdiff from −10 to 10 mL/min/1.73 m2 and 674 (17.7%) had an eGFRdiff >10 mL/min/1.73 m2. Patients in the lowest eGFRdiff group were older, more often female, had a higher CHA2DS2-VASc score, more often non-paroxysmal AF and a greater burden of comorbidities (Table 1).
Figure 1.
Distribution of estimated glomerular filtration rate differences.
MRI detected brain lesions
Results of 1722 patients with available brain MRI are shown in Table 2 and Figure 2. Overall, 38% of patients had evidence of any SNCI and/or LNCCI, 53% had at least moderate WMH (Fazekas ≥2) and 22% had microbleeds. Across the three eGFR groups the brain lesion prevalence increased from 28% to 42% (P = 0.006) for SNCI and/or LNCCI and from 38% to 63% (P < 0.001) for WMH (Fazekas ≥2). Prevalent microbleeds were not different between groups (18%, 22% and 25%; P = 0.2).
Table 2.
Prevalence of brain lesions overall and according to estimated GFR difference group
| Overall | eGFR difference (mL/min/1.73 m2) | P-value | |||
|---|---|---|---|---|---|
| <−10 | −10 to 10 | >10 | |||
| N | 1722 | 220 | 1286 | 216 | |
| SNCI and/or LNCCI | 647 (38%) | 91 (42%) | 495 (39%) | 61 (28%) | 0.006 |
| LNCCI | 392 (23%) | 57 (26%) | 298 (23%) | 37 (17%) | 0.066 |
| SNCI | 381 (22%) | 47 (21%) | 300 (24%) | 34 (16%) | 0.038 |
| WMH (Fazekas ≥2) | 918 (53%) | 139 (63%) | 698 (55%) | 81 (38%) | <0.001 |
| Microbleeds | 365 (22%) | 54 (25%) | 274 (22%) | 37 (18%) | 0.17 |
Values are presented as count (percentage), P-values were calculated Pearson’s chi-square tests.
eGFRdiff = eGFRcrea-cys – eGFRcrea.
LNCCI, large non-cortical and cortical infarct; SNCI, small non-cortical infarct; WMH, white matter hyperintensity.
Figure 2.
Prevalence of brain magnetic resonance imaging lesions by estimated glomerular filtration rate difference groups. eGFRdiff = eGFRcrea-cys – eGFRcrea. LNCCI, large non-cortical and cortical infarct; SNCI, small non-cortical infarct; WMH, white matter hyperintensity.
eGFR difference and cardiac biomarkers
Biomarker levels of hs-TnT and NT-proBNP increased continuously across decreasing eGFRdiff (Figure 3). Across the three eGFR groups hs-TnT levels were 8.4 (6.1; 11.8), 13.8 (9.1; 22.3) and 18.4 (12.3; 33.0) ng/L and NT-proBNP levels were 178 (90; 558), 650 (218; 1448) and 1086 (433; 1871) ng/L, respectively (Table 1). The relation of creatinine and cystatin C with these two biomarkers is shown in Supplementary material online, Figures S2 and S3.
Figure 3.
Relation of estimated glomerular filtration rate difference with high-sensitivity troponin T (left panel) and NT-proBNP (right panel).
eGFR difference and adverse outcomes
During a median follow-up of 8 years, 985 MACE and 831 HF events occurred. The annual incidence rates increased from 1.69 to 3.90 to 6.78 for MACE and from 1.24 to 3.48 to 6.07 for HF events across >10, −10 to 10, and <−10 mL/min/1.73 m2 eGFRdiff groups (Figure 4). Cumulative incidence curves for the two main outcomes are shown in Figure 5. The respective curves of the <−10 mL/min/1.73 m2 groups increased substantially more compared to the other two eGFRdiff groups.
Figure 4.
Association of estimated glomerular filtration rate difference with cardiovascular events. Model 1 was adjusted for age and sex. Model 2 was adjusted for age, sex, diabetes, arterial blood pressure, history of stroke/transient ischaemic attack, and smoking status. eGFRdiff = eGFRcrea-cys – eGFRcrea.
Figure 5.
Incidence curves for major cardiovascular events (left panel) and first heart failure hospitalization (right panel) according to estimated glomerular filtration rate difference groups. eGFRdiff = eGFRcrea-cys – eGFRcrea.
In Cox-proportional hazards regression models, eGFRdiff categories were significantly associated with both primary outcomes (Figure 4). Using the eGFRdiff group −10 to 10 mL/min/1.73 m2 as the reference, patients with an eGFRdiff <−10 mL/min/1.73 m2 had an adjusted hazard ratio (95% confidence interval) of 1.50 (1.26; 1.79; P < 0.001) for MACE and of 1.49 (1.23; 1.81, P < 0.001) for HF hospitalizations. Similarly, eGFRdiff was significantly associated with both outcomes when used as a continuous variable (Figure 6). The results were consistent for CV death, all-cause death, and major bleeding, but not for stroke or systemic embolism and myocardial infarction (Figure 4). In sensitivity and specificity analyses, the <−10 mL/min/1.73 m2 cut-off had a very high sensitivity to identify high-risk patients and the >10 mL/min/1.73 m2 cut-off had a very high specificity (see Supplementary material online, Table S2). The calculated optimal cut-off using the Youden index was around 0 for different follow-up periods (see Supplementary material online, Table S3).
Figure 6.
Association of estimated glomerular filtration rate difference with cardiovascular events. Model 1 was adjusted for age and sex. Model 2 was adjusted for age, sex, diabetes, arterial blood pressure, history of stroke/TIA, and smoking status. eGFRdiff = eGFRcrea-cys – eGFRcrea.
Time-dependent AUC (95% CI) at 10 years for MACE and HF hospitalization were highest for eGFRcrea-cys (0.772 and 0.803, respectively) followed by eGFRdiff (0.742 and 0.756, respectively) (see Supplementary material online, Table S4 and Supplementary material online, Figure S4). Adding eGFRdiff to eGFRcrea and eGFRcrea-cys increased all AUCs, though the increase was smaller for models with eGFRcrea-cys (see Supplementary material online, Table S5).
Discussion
In this prospective cohort of patients with AF, about 30% had a clinically relevant difference exceeding 10 mL/min/1.73 m2 between creatinine- and cystatin C-based eGFR (eGFRdiff). The eGFRdiff was strongly associated with elevated cardiac biomarkers and adverse outcomes. Patients with higher eGFRcrea compared to eGFRcrea-cys showed higher levels of hs-TnT and NT-proBNP, and an increased risk for MACE, HF events, and prevalent brain lesions. These findings suggest that eGFRdiff may identify a high-risk subset of AF patients and may have potential implications for pharmacologic management, particularly for renal function-guided drug dosing.
A considerable proportion of AF patients had a clinically relevant eGFRdiff of >10 mL/min/1.73 m2, raising concerns regarding inappropriate dosing of DOAC. These findings raise the question, which eGFR estimation method should be used in clinical practice for drug dosing in AF patients. In the landmark DOAC trials ROCKET-AF and ENGAGE AF-TIMI 48, creatinine clearance calculated by the Cockcroft–Gault equation was used for dose adjustment, while creatinine was directly used in ARISTOTLE and no dose adjustment criteria were applied in RE-LY.20–23 In clinical practice, the latest CKD-EPI equation for eGFRcrea is most commonly used for dose adjustment. While there is little difference between different iterations of the CKD-EPI equation,24 the correlation between creatinine clearance, eGFRcrea, and eGFRcrea-cys varies substantially.25,26 In a post-hoc analysis of ARISTOTLE, cystatin c-based eGFR was most robust for predicting death and bleeding.25 Our prior analysis of this cohort supports that observation: cystatin c-based eGFR was associated with both ischaemic and bleeding outcomes, while creatinine-based eGFR was only associated with ischaemic outcomes.4 These findings underscore the need for harmonization and prospective validation, ideally within large pooled datasets such as the COMBINE-AF dataset that includes all landmark DOAC trials.27
A negative eGFRdiff was most pronounced among patients with higher disease burden, as reflected by higher levels of hs-TnT and NT-proBNP. Creatinine is primarily released by muscles, partially actively secreted by the proximal tubule and excreted by the kidneys and to a lesser extent by the gastro-intestinal system.28 As patients with more advanced HF have less muscle mass in addition to potentially increased tubular creatinine secretion due to neurohumoral activation, creatinine levels might be less accurate for estimating kidney function. Moreover, a more negative eGFRdiff might also reflect greater congestion by more severe HF and worse kidney function, rather than intrinsic glomerular damage.9 Conversely, cystatin C is continually synthesized in all nucleated cells, not actively secreted and, due to its low molecular weight, almost completely filtered in the glomerulus and catabolized in the proximal tubule.28,29 This leads to a higher stability in serum concentration, a reduced sensitivity to confounding factors and a more accurate estimation of kidney function.30,31 These biochemical characteristics translate into more consistent associations of cystatin C-based eGFR with ischaemic and bleeding events, compared to creatinine-based eGFR.4
In patients with AF, our findings add to the role of renal function as part of a multidomain risk profile, rather than an isolated marker, within the CV-kidney-metabolic (CKM) framework.32,33 Patients with AF have a particularly high burden of CKM domains, that were associated both with adverse outcomes and treatment decisions.32 Specifically, AF patients with multiple CKM domains, like systolic blood pressure, HbA1c, and triglyceride levels, have highly complex clinical profiles, that influence pharmacological and lifestyle interventions.33 Pathophysiologically, cardiorenal interactions lead to sympathetic nervous system activation with vasoconstriction, an activated renin–angiotensin–aldosterone system, chronic inflammation with oxidative stress and endothelial dysfunction. These mechanisms further impair the glomerular filtration by damaging the renal microvasculature and fibrotic remodelling.34,35
Patients with an eGFRdiff <−10 mL/min/1.73 m2 had a four-fold higher risk of MACE, nearly five-fold increased HF hospitalization risk, and a greater burden of ischaemic brain lesions, compared to patients with an eGFRdiff >10 mL/min/1.73 m2. Specifically, both overt and covert ischaemic brain lesions were associated with cognitive impairment in prior analyses of our cohort.16,17 Similar observations of clinical outcomes have been shown in patients with CKD and HF.8–10 In patients with CKD a negative eGFRdiff associates with worse outcomes in 5000 In the Chronic Renal Insufficiency Cohort study and in a meta-analysis of more than 800 000 patients.8,10 Moreover, a change in eGFRdiff over time was associated with an increased risk for mortality in CKD patients.8 In the PARADIGM-HF study, a significant eGFRdiff was linked to biomarker profiles of advanced HF and poor prognosis.9 Our study extends these findings to the growing AF population, where accurate renal function assessment is essential to guide drug dosing. The clinical impact of potentially inappropriate drug dosing needs to be further investigated in other large AF patient populations. The translation of our current findings into clinical practice is currently limited by the necessity of simultaneous availability of both biomarkers.
Strength and limitations
Strengths of our study include the large, prospectively enrolled, well-defined population with standardized outcome adjudication by an independent clinical committee. The availability of biomarker and MRI data enhances mechanistic insights. The following limitations should be acknowledged. Biomarker measurements were limited to baseline precluding longitudinal assessment. Brain MRI was only available in the Swiss-AF cohort with potential selection bias. Directly measured GFR was not available, preventing direct comparison to a reference standard. While the results are generalizable to patients with AF in Switzerland, caution is warranted when extrapolating to other geographical or disease populations. The observational design limits causal inference, and residual confounding, by for example comorbidity burden or unmeasured frailty indices, cannot be excluded. Reverse causation may also be present, whereby patients with more advanced CV disease exhibit both altered biomarker profiles and greater eGFR discordance.
Conclusion
A substantial proportion of patients with AF had clinically relevant discrepancies between creatinine- and cystatin C-based eGFR estimates. A negative eGFRdiff was associated with elevated cardiac biomarkers, adverse CV outcomes, and increased burden of brain lesions. eGFRdiff may serve a marker of cardiorenal and cerebrovascular risk in AF and warrants further investigation to clarify its role in guiding treatment decisions.
Supplementary Material
Acknowledgements
None.
Contributor Information
Philipp Krisai, Cardiology/Electrophysiology Division, Department of Medicine, University Hospital Basel, Petersgraben 4, 4031 Basel, Switzerland; Cardiovascular Research Institute Basel, University Hospital Basel, University of Basel, Spitalstrasse 2, 4056 Basel, Switzerland.
Mikko Mutti, Cardiology/Electrophysiology Division, Department of Medicine, University Hospital Basel, Petersgraben 4, 4031 Basel, Switzerland; Cardiovascular Research Institute Basel, University Hospital Basel, University of Basel, Spitalstrasse 2, 4056 Basel, Switzerland.
Stefanie Aeschbacher, Cardiology/Electrophysiology Division, Department of Medicine, University Hospital Basel, Petersgraben 4, 4031 Basel, Switzerland; Cardiovascular Research Institute Basel, University Hospital Basel, University of Basel, Spitalstrasse 2, 4056 Basel, Switzerland.
Michael Coslovsky, Cardiovascular Research Institute Basel, University Hospital Basel, University of Basel, Spitalstrasse 2, 4056 Basel, Switzerland; Department of Clinical Research, University of Basel, University Hospital Basel, Spitalstrasse 8, 4056 Basel, Switzerland.
Tobias Reichlin, Department of Cardiology, Inselspital, Bern University Hospital, University of Bern, Freiburgstrasse 20, 3010 Bern, Switzerland.
Nicolas Rodondi, Institute of Primary Health Care (BIHAM), University of Bern, Mittelstrasse 43, 3012 Bern, Switzerland; Department of General Internal Medicine, Inselspital, Bern University Hospital, University of Bern, Freiburgstrasse 20, 3010 Bern, Switzerland.
Peter Ammann, Department of Cardiology, Kantonsspital St Gallen, Rorschacher Strasse 95, 9000 St Gallen, Switzerland.
Giorgio Moschovitis, Division of Cardiology, Cardiocentro Ticino Institute, Ente Ospedaliero Cantonale, Via Tesserete 48, 6900 Lugano, Switzerland.
Elia Rigamonti, Division of Cardiology, Cardiocentro Ticino Institute, Ente Ospedaliero Cantonale, Via Tesserete 48, 6900 Lugano, Switzerland.
Giulio Conte, Division of Cardiology, Cardiocentro Ticino Institute, Ente Ospedaliero Cantonale, Via Tesserete 48, 6900 Lugano, Switzerland.
Moa L Haller, Institute of Primary Health Care (BIHAM), University of Bern, Mittelstrasse 43, 3012 Bern, Switzerland; Department of General Internal Medicine, Inselspital, Bern University Hospital, University of Bern, Freiburgstrasse 20, 3010 Bern, Switzerland.
Amelie Stoecklin, Cardiology/Electrophysiology Division, Department of Medicine, University Hospital Basel, Petersgraben 4, 4031 Basel, Switzerland; Cardiovascular Research Institute Basel, University Hospital Basel, University of Basel, Spitalstrasse 2, 4056 Basel, Switzerland.
Leo H Bonati, Department of Neurology and Stroke Center, University Hospital Basel, University of Basel, Petersgraben 4, 4031 Basel, Switzerland; Research Department, Reha Rheinfelden, Salinenstrasse 98, 4310 Rheinfelden, Switzerland.
Stefan Osswald, Cardiology/Electrophysiology Division, Department of Medicine, University Hospital Basel, Petersgraben 4, 4031 Basel, Switzerland; Cardiovascular Research Institute Basel, University Hospital Basel, University of Basel, Spitalstrasse 2, 4056 Basel, Switzerland.
David Conen, Population Health Research Institute, McMaster University, 237 Barton Street East, Hamilton, ON L8L 2X2, Canada.
Felix Mahfoud, Cardiology/Electrophysiology Division, Department of Medicine, University Hospital Basel, Petersgraben 4, 4031 Basel, Switzerland; Cardiovascular Research Institute Basel, University Hospital Basel, University of Basel, Spitalstrasse 2, 4056 Basel, Switzerland.
Michael Kühne, Cardiology/Electrophysiology Division, Department of Medicine, University Hospital Basel, Petersgraben 4, 4031 Basel, Switzerland; Cardiovascular Research Institute Basel, University Hospital Basel, University of Basel, Spitalstrasse 2, 4056 Basel, Switzerland.
Data availability
The data that support the findings of this study are not openly available due to reasons of sensitivity and are available from the corresponding author upon reasonable request.
Supplementary material
Supplementary material is available at European Heart Journal Open online.
Author contributions
Philipp Krisai (Writing—review & editing [supporting]), Mikko Mutti (Writing—review & editing [supporting]), Stefanie Aeschbacher (Writing—review & editing [supporting]), Michael Coslovsky (Writing—review & editing [supporting]), Tobias Reichlin (Writing—review & editing [supporting]), Nicolas Rodondi (Writing—review & editing [supporting]), Peter Ammann (Writing—review & editing [supporting]), Giorgio Moschovitis (Writing—review & editing [supporting]), Elia Rigamonti (Writing—review & editing [supporting]), Giulio Conte (Writing—review & editing [supporting]), Moa L Haller (Writing—review & editing [supporting]), Amelie Stoecklin (Writing—review & editing [supporting]), Leo H Bonati (Writing—review & editing [supporting]), Stefan Osswald (Writing—review & editing [supporting]), David Conen (Writing—review & editing [supporting]), Felix Mahfoud (Writing—review & editing [supporting]), and Michael Kuehne (Writing—review & editing [supporting]).
Funding
The Swiss-AF study is supported by grants of the Swiss National Science Foundation (grant numbers 33CS30_148474, 33CS30_177520, 32473B_176178, and 32003B_197524), the Swiss Heart Foundation, the Foundation for Cardiovascular Research (FCVR), and the University of Basel. David Conen is a recipient of a PHRI career award.
The other authors report no conflicts.
References
- 1. Ding WY, Potpara TS, Blomström-Lundqvist C, Boriani G, Marin F, Fauchier L, Lip GYH. Impact of renal impairment on atrial fibrillation: ESC-EHRA EORP-AF Long-Term General Registry. Eur J Clin Invest 2022;52:e13745. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Carrero JJ, Trevisan M, Evans M, Svennberg E, Szummer K. Kidney function and the risk of heart failure in patients with new-onset atrial fibrillation. Int J Cardiol 2020;320:101–105. [DOI] [PubMed] [Google Scholar]
- 3. De Marchis GM, Krisai P, Werlen L, Sinnecker T, Aeschbacher S, Dittrich TD, Polymeris AA, Coslovksy M, Blum MR, Rodondi N, Reichlin T, Moschovitis G, Wuerfel J, Lyrer PA, Fischer U, Conen D, Kastner P, Ziegler A, Osswald S, Kühne M, Bonati LH. Biomarker, imaging, and clinical factors associated with overt and covert stroke in patients with atrial fibrillation. Stroke 2023;54:2542–2551. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Schweigler A, Hennings E, Aeschbacher S, Carmine D, Reichlin T, Rodondi N, Stauber A, Ammann P, Moschovitis G, Bolt L, Demarchi A, Mueller AS, Reneau D, Coslovsky M, Zuern CS, Bonati LH, Conen D, Osswald S, Kühne M, Krisai P. Kidney function estimated by creatinine and cystatin C and adverse cardiovascular outcomes in patients with atrial fibrillation. J Intern Med 2025;297:335–338. [DOI] [PubMed] [Google Scholar]
- 5. Caregaro L, Menon F, Angeli P, Amodio P, Merkel C, Bortoluzzi A, Alberino F, Gatta A. Limitations of serum creatinine level and creatinine clearance as filtration markers in cirrhosis. Arch Intern Med 1994;154:201–205. [PubMed] [Google Scholar]
- 6. Beddhu S, Samore MH, Roberts MS, Stoddard GJ, Pappas LM, Cheung AK. Creatinine production, nutrition, and glomerular filtration rate estimation. J Am Soc Nephrol 2003;14:1000–1005. [DOI] [PubMed] [Google Scholar]
- 7. Inker LA, Eneanya ND, Coresh J, Tighiouart H, Wang D, Sang Y, Crews DC, Doria A, Estrella MM, Froissart M, Grams ME, Greene T, Grubb A, Gudnason V, Gutiérrez OM, Kalil R, Karger AB, Mauer M, Navis G, Nelson RG, Poggio ED, Rodby R, Rossing P, Rule AD, Selvin E, Seegmiller JC, Shlipak MG, Torres VE, Yang W, Ballew SH, Couture SJ, Powe NR, Levey AS. New creatinine- and cystatin C-based equations to estimate GFR without race. N Engl J Med 2021;385:1737–1749. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Chen DC, Shlipak MG, Scherzer R, Bauer SR, Potok OA, Rifkin DE, Ix JH, Muiru AN, Hsu C-Y, Estrella MM. Association of intraindividual difference in estimated glomerular filtration rate by creatinine vs cystatin C and end-stage kidney disease and mortality. JAMA Netw Open 2022;5:e2148940. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Tolomeo P, Butt JH, Kondo T, Campo G, Desai AS, Jhund PS, Køber L, Lefkowitz MP, Rouleau JL, Solomon SD, Swedberg K, Vaduganathan M, Zile MR, Packer M, McMurray JJV. Importance of cystatin C in estimating glomerular filtration rate: the PARADIGM-HF trial. Eur Heart J 2023;44:2202–2212. [DOI] [PubMed] [Google Scholar]
- 10. Estrella MM, Ballew SH, Sang Y, Grams ME, Coresh J, Surapaneni A, Alencar de Pinho N, Ärnlöv J, Brenner H, Carrero J-J, Chen TK, Cohen DL, Cushman M, Gansevoort RT, Hwang S-J, Inker LA, Ix JH, Kabasawa K, Konta T, Lees JS, Polkinghorne KR, Shlipak MG, Vernooij RWM, Wheeler DC, Yadav AK, Levey AS, Eckardt K-U. Discordance in creatinine- and cystatin C-based eGFR and clinical outcomes: a meta-analysis. JAMA 2025;334:1915–1926. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Healey JS, Oldgren J, Ezekowitz M, Zhu J, Pais P, Wang J, Commerford P, Jansky P, Avezum A, Sigamani A, Damasceno A, Reilly P, Grinvalds A, Nakamya J, Aje A, Almahmeed W, Moriarty A, Wallentin L, Yusuf S, Connolly SJ. Occurrence of death and stroke in patients in 47 countries 1 year after presenting with atrial fibrillation: a cohort study. Lancet Lond Engl 2016;388:1161–1169. [DOI] [PubMed] [Google Scholar]
- 12. Van Gelder IC, Rienstra M, Bunting KV, Casado-Arroyo R, Caso V, Crijns HJGM, De Potter TJR, Dwight J, Guasti L, Hanke T, Jaarsma T, Lettino M, Løchen M-L, Lumbers RT, Maesen B, Mølgaard I, Rosano GMC, Sanders P, Schnabel RB, Suwalski P, Svennberg E, Tamargo J, Tica O, Traykov V, Tzeis S, Kotecha D. 2024 ESC Guidelines for the management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS). Eur Heart J 2024;45:3314–3414. [DOI] [PubMed] [Google Scholar]
- 13. Conen D, Rodondi N, Mueller A, Beer J, Auricchio A, Ammann P, Hayoz D, Kobza R, Moschovitis G, Shah D, Schlaepfer J, Novak J, di Valentino M, Erne P, Sticherling C, Bonati L, Ehret G, Roten L, Fischer U, Monsch A, Stippich C, Wuerfel J, Schwenkglenks M, Kuehne M, Osswald S. Design of the Swiss Atrial Fibrillation Cohort Study (Swiss-AF): structural brain damage and cognitive decline among patients with atrial fibrillation. Swiss Med Wkly 2017;147:w14467. [DOI] [PubMed] [Google Scholar]
- 14. Krisai P, Blum S, Aeschbacher S, Beer JH, Moschovitis G, Witassek F, Kobza R, Rodondi N, Moutzouri E, Mahmood A, Healey JS, Zuern CS, Kühne M, Osswald S, Conen D. Associations of symptoms and quality of life with outcomes in patients with atrial fibrillation. Heart Br Card Soc 2020;106:1847–1852. [DOI] [PubMed] [Google Scholar]
- 15. Ruperti Repilado FJ, Doerig L, Blum S, Aeschbacher S, Krisai P, Ammann P, Erne P, Moschovitis G, di Valentino M, Shah D, Schläpfer J, Stempfel S, Kühne M, Sticherling C, Osswald S, Conen D. Prevalence and predictors of atrial fibrillation type among individuals with recent onset of atrial fibrillation. Swiss Med Wkly 2018;148:w14652. [DOI] [PubMed] [Google Scholar]
- 16. Conen D, Rodondi N, Müller A, Beer JH, Ammann P, Moschovitis G, Auricchio A, Hayoz D, Kobza R, Shah D, Novak J, Schläpfer J, Di Valentino M, Aeschbacher S, Blum S, Meyre P, Sticherling C, Bonati LH, Ehret G, Moutzouri E, Fischer U, Monsch AU, Stippich C, Wuerfel J, Sinnecker T, Coslovsky M, Schwenkglenks M, Kühne M, Osswald S. Relationships of overt and silent brain lesions with cognitive function in patients with atrial fibrillation. J Am Coll Cardiol 2019;73:989–999. [DOI] [PubMed] [Google Scholar]
- 17. Kühne M, Krisai P, Coslovsky M, Rodondi N, Müller A, Beer JH, Ammann P, Auricchio A, Moschovitis G, Hayoz D, Kobza R, Shah D, Stephan FP, Schläpfer J, Di Valentino M, Aeschbacher S, Ehret G, Eken C, Monsch A, Roten L, Schwenkglenks M, Springer A, Sticherling C, Reichlin T, Zuern CS, Meyre PB, Blum S, Sinnecker T, Würfel J, Bonati LH, Conen D, Osswald S. Silent brain infarcts impact on cognitive function in atrial fibrillation. Eur Heart J 2022;43:2127–2135. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Fazekas F, Chawluk JB, Alavi A, Hurtig HI, Zimmerman RA. MR signal abnormalities at 1.5 T in Alzheimer’s dementia and normal aging. AJR Am J Roentgenol 1987;149:351–356. [DOI] [PubMed] [Google Scholar]
- 19. (EHRA) D with the special contribution of the EHRA, (EACTS) E by the EA for C-TS, Members AF; Camm AJ, Kirchhof P, Lip GYH, Schotten U, Savelieva I, Ernst S, Van Gelder IC, Al-Attar N, Hindricks G, Prendergast B, Heidbuchel H, Alfieri O, Angelini A, Atar D, Colonna P, De Caterina R, De Sutter J, Goette A, Gorenek B, Heldal M, Hohloser SH, Kolh P, Le Heuzey J-Y, Ponikowski P, Rutten FH. Guidelines for the management of atrial fibrillation: the task force for the management of atrial fibrillation of the European Society of Cardiology (ESC). Eur Heart J 2010;31:2369–2429. [DOI] [PubMed] [Google Scholar]
- 20. Patel MR, Mahaffey KW, Garg J, Pan G, Singer DE, Hacke W, Breithardt G, Halperin JL, Hankey GJ, Piccini JP, Becker RC, Nessel CC, Paolini JF, Berkowitz SD, Fox KAA, Califf RM. Rivaroxaban versus warfarin in nonvalvular atrial fibrillation. N Engl J Med 2011;365:883–891. [DOI] [PubMed] [Google Scholar]
- 21. Giugliano RP, Ruff CT, Braunwald E, Murphy SA, Wiviott SD, Halperin JL, Waldo AL, Ezekowitz MD, Weitz JI, Špinar J, Ruzyllo W, Ruda M, Koretsune Y, Betcher J, Shi M, Grip LT, Patel SP, Patel I, Hanyok JJ, Mercuri M, Antman EM. Edoxaban versus warfarin in patients with atrial fibrillation. N Engl J Med 2013;369:2093–2104. [DOI] [PubMed] [Google Scholar]
- 22. Granger CB, Alexander JH, McMurray JJV, Lopes RD, Hylek EM, Hanna M, Al-Khalidi HR, Ansell J, Atar D, Avezum A, Bahit MC, Diaz R, Easton JD, Ezekowitz JA, Flaker G, Garcia D, Geraldes M, Gersh BJ, Golitsyn S, Goto S, Hermosillo AG, Hohnloser SH, Horowitz J, Mohan P, Jansky P, Lewis BS, Lopez-Sendon JL, Pais P, Parkhomenko A, Verheugt FWA, Zhu J, Wallentin L. Apixaban versus warfarin in patients with atrial fibrillation. N Engl J Med 2011;365:981–992. [DOI] [PubMed] [Google Scholar]
- 23. Connolly SJ, Ezekowitz MD, Yusuf S, Eikelboom J, Oldgren J, Parekh A, Pogue J, Reilly PA, Themeles E, Varrone J, Wang S, Alings M, Xavier D, Zhu J, Diaz R, Lewis BS, Darius H, Diener H-C, Joyner CD, Wallentin L. Dabigatran versus warfarin in patients with atrial fibrillation. N Engl J Med 2009;361:1139–1151. [DOI] [PubMed] [Google Scholar]
- 24. Emrich IE, Pickering JW, Götzinger F, Kramann R, Kunz M, Lauder L, Papademetriou V, Böhm M, Heine GH, Mahfoud F. Comparison of three creatinine-based equations to predict adverse outcome in a cardiovascular high-risk cohort: an investigation using the SPRINT research materials. Clin Kidney J 2024;17:sfae011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Hijazi Z, Granger CB, Hohnloser SH, Westerbergh J, Lindbäck J, Alexander JH, Keltai M, Parkhomenko A, López-Sendón JL, Lopes RD, Siegbahn A, Wallentin L. Association of different estimates of renal function with cardiovascular mortality and bleeding in atrial fibrillation. J Am Heart Assoc 2020;9:e017155. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Hijazi Z, Hohnloser SH, Oldgren J, Andersson U, Connolly SJ, Eikelboom JW, Ezekowitz MD, Reilly PA, Siegbahn A, Yusuf S, Wallentin L. Efficacy and safety of dabigatran compared with warfarin in relation to baseline renal function in patients with atrial fibrillation: a RE-LY (Randomized Evaluation of Long-term Anticoagulation Therapy) trial analysis. Circulation 2014;129:961–970. [DOI] [PubMed] [Google Scholar]
- 27. Carnicelli AP, Hong H, Giugliano RP, Connolly SJ, Eikelboom J, Patel MR, Wallentin L, Morrow DA, Wojdyla D, Hua K, Hohnloser SH, Oldgren J, Ruff CT, Piccini JP, Lopes RD, Alexander JH, Granger CB. Individual patient data from the pivotal randomized controlled trials of non-vitamin K antagonist oral anticoagulants in patients with atrial fibrillation (COMBINE AF): design and rationale: from the COMBINE AF (A Collaboration between Multiple institutions to Better Investigate Non-vitamin K antagonist oral anticoagulant use in Atrial Fibrillation) Investigators. Am Heart J 2021;233:48–58. [DOI] [PubMed] [Google Scholar]
- 28. Porrini E, Ruggenenti P, Luis-Lima S, Carrara F, Jiménez A, de Vries APJ, Torres A, Gaspari F, Remuzzi G. Estimated GFR: time for a critical appraisal. Nat Rev Nephrol 2019;15:177–190. [DOI] [PubMed] [Google Scholar]
- 29. Woo K-S, Choi J-L, Kim B-R, Kim J-E, Han J-Y. Clinical usefulness of serum cystatin C as a marker of renal function. Diabetes Metab J 2014;38:278–284. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Shlipak MG, Matsushita K, Ärnlöv J, Inker LA, Katz R, Polkinghorne KR, Rothenbacher D, Sarnak MJ, Astor BC, Coresh J, Levey AS, Gansevoort RT. Cystatin C versus creatinine in determining risk based on kidney function. N Engl J Med 2013;369:932–943. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Herget-Rosenthal S, Bökenkamp A, Hofmann W. How to estimate GFR-serum creatinine, serum cystatin C or equations? Clin Biochem 2007;40:153–161. [DOI] [PubMed] [Google Scholar]
- 32. Mei DA, Imberti JF, Vitolo M, Romiti GF, Corica B, Mantovani M, Bonini N, Cherubini B, Marin F, Diemberger I, Dan GA, Potpara T, Proietti M, Lip GYH, Boriani G. Cardiovascular-kidney-metabolic domains and impact on antithrombotic treatment, integrated care and clinical outcomes in patients with atrial fibrillation: results from a prospective European registry. Eur J Intern Med 2026;145:106512. [DOI] [PubMed] [Google Scholar]
- 33. Chan Y-H, Kao Y-W, Cheng W-H, Chen S-W, Hung C-L, Romiti GF, Lip GYH, Chao T-F. Cardiovascular-kidney-metabolic interplay in patients with atrial fibrillation receiving direct oral anticoagulants. Europace 2026;28:euag028. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Xu X, Zhang B, Wang Y, Shi S, Lv J, Fu Z, Gao X, Li Y, Wu H, Song Q. Renal fibrosis in type 2 cardiorenal syndrome: an update on mechanisms and therapeutic opportunities. Biomed Pharmacother 2023;164:114901. [DOI] [PubMed] [Google Scholar]
- 35. Rangaswami J, Bhalla V, Blair JEA, Chang TI, Costa S, Lentine KL, Lerma EV, Mezue K, Molitch M, Mullens W, Ronco C, Tang WHW, McCullough PA. Cardiorenal syndrome: classification, pathophysiology, diagnosis, and treatment strategies: a scientific statement from the American Heart Association. Circulation 2019;139:e840–e878. [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 openly available due to reasons of sensitivity and are available from the corresponding author upon reasonable request.