Kidney function estimated by creatinine and cystatin C and adverse cardiovascular outcomes in patients with atrial fibrillation

The association of impaired kidney function with ischemic and bleeding events in atrial fibrillation (AF) patients is important for treatment decisions regarding anticoagulation. We aimed to compare the associations and discriminatory performance of kidney function estimations by cystatin C (eGFRcy), creatinine (eGFRcr), both biomarkers combined (eGFRc2), and the ratio of eGFRcy/eGFRcr (eGFRra) with ischemic events and major bleedings [1, 2]. We combined data from the prospective Swiss AF Study [3] and the Beat- AF Study [4]. All different estimated glomerular filtration rate (eGFR) models were calculated using the sex and age-specific CKD-EPI formula [5]. The ischemic outcome was major cardiovascular events (MACE), defined as a composite of ischemic stroke, systemic embolism (SE), myocardial infarction (MI), and cardiovascular (CV) death. The bleeding outcome was defined by the criteria of the International Society on Thrombosis and Hemostasis for major bleeding [6]. Detailed outcome definitions and methods are provided in the supplement. The studies were approved by the lead ethics commission, and all study centers received approval from the individual regional ethical committees. ClinicalTrials.gov Identifier: NCT02105844. We included 3867 participants: mean age 71 ± 10 years, 28% female, and 49% paroxysmal AF (Table S1). Mean eGFRcr, eGFRcy, and eGFRc2 were 63 ± 20, 64 ± 26, and 64 ± 23 mL/min/1.73 m2, respectively. Over a median follow-up of 6 years, 738 MACE (3.22/100 patient-years; 487 CV deaths, 206 ischemic strokes/SE, and 155 MI) occurred with increasing incidence rates over decreasing eGFRc2 categories (1.10–11.45 from lowest to highest) (Table S2). Cumulative incidence curves of MACE and major bleeding stratified by combined eGFR levels are shown in Fig. 1a. For MACE, the hazard ratio (HR) (95% confidence interval [CI]) per 10 mL/min/1.73 m2 decrease of eGFRc2, eGFRcr, and eGFRcy were 1.16 (95% CI 1.11–1.21, p < 0.001), 1.09 (1.04–1.14, p < 0.001), and 1.18 (1.13–1.23, p < 0.001), respectively, and per 1-point decrease in eGFRra 3.46 (2.42, 4.93, p < 0.001) (Fig. 1b, Table S3). Fig. S1 shows the HRs over continuous eGFR levels. Besides CV death, no individual MACE component was associated with any eGFR calculation (Table S3). The Akaike information criterion (AIC) fIor MACE was lower for eGFRcy (10930.58), eGFRc2 (10949.01), and eGFRra (10938.92) compared to eGFRcr (10975.97). The bleeding outcome occurred in 423 participants (1.85/100 patient-years) with increasing incidence rates over decreasing eGFRc2 categories (0.80–4.57 from lowest to highest) (Table S2). Cumulative incidence curves for major bleeding stratified by eGFRc2, eGFRcr, eGFRcy, and eGFRra are shown in Fig. 1a and Figs. S2–S4. For major bleeding, HRs (95% CI) were significant for eGFRcy (1.11 [1.05–1.17], p < 0.001), eGFRc2 (1.09 [1.03–1.16], p = 0.005), and eGFRra (2.10 [1.34–3.30], p = 0.001), but not for eGFRcr (1.04 [0.98–1.10], p = 0.24). The AC for major bleeding were lower for eGFRcy (6533.73), eGFRc2 (6539.60), and eGFRra (6536.40) compared to eGFRcr (6546.23). Results were consistent in sensitivity analyses adding lifestyle factors (Table S4). In our prospective cohort of AF patients, 19.0% experienced an ischemic and 10.9% a major bleeding event over 6 years. Patients with reduced kidney function below 30 mL/min/1.73 m2 had an excess risk of ischemic and major bleeding events. Only cystatin c-based eGFR estimations and eGFR-ratio were consistently associated with both ischemic and bleeding events. In a post hoc study from ARISTOTLE, cystatin C–based equations showed less variation and higher prognostic value for death and major bleeding, but ischemic stroke/embolism was not investigated [7]. We expand these prior findings to a clinically useful ischemic endpoint and show that cystatin C–based equations might be preferential in AF patients for treatment decisions. Potential mechanisms for improved diagnostic accuracy and stronger associations by cystatin C may include its metabolism. It is continually synthesized in all nucleated cells, not actively secreted, and due to its low molecular weight, it is almost completely filtered in the glomerulus and catabolized in the proximal tubule [8, 9]. This leads to a higher stability in serum concentration. In contrast, creatinine is primarily released by muscles, partially actively secreted by the proximal tubule and by the gastrointestinal system. Thus, its levels may be unstable and result in an over-/underestimation of kidney function [10]. A mismatch between the eGFRcy and eGFRcr can be quantified by the eGFRra, which may indicate a shrunken pore syndrome [2]. Limitations of our study include biomarker measurements at baseline only, limited generalizability to other patient populations, and the observational design. In conclusion, patients with AF and reduced kidney function had an excess risk of ischemic and major bleeding events compared to patients with normal kidney function. Cystatin C–based GFR estimations and the calculated GFR ratio were consistently associated both with ischemic and bleeding events, whereas creatinine-based GFR estimations were only associated with ischemic events. Moreover, cystatin C-based GFR estimations outperformed creatinine-based estimations in terms of discriminatory performance for both endpoints. David Conen received speaker fees from Servier, and consulting fees from Roche Diagnostics and Trimedics, all outside of the current work. Giorgio Moschovitis has received consultant fees for taking part to advisory boards from Novartis and Astra Zeneca outside of the submitted work. Leo H. Bonati received grants from the Swiss National Science Foundation (PBBSB-116873, 33CM30-124119, 32003B-156658, 32003B-197524); Berne, Switzerland), The Swiss Heart Foundation (Berne, Switzerland, and the University of Basel [Basel, Switzerland]). LHB has received an unrestricted research grant from Astra Zeneca, and consultancy or advisory board fees or speaker's honoraria from Amgen, Bayer, Bristol-Myers Squibb, and Claret Medical, and travel grants from AstraZeneca and Bayer. Michael Coslovsky reports no disclosures. Michael Kühne reports grants from the Swiss National Science Foundation (Grant numbers 33CS30_148474, 33CS30_177520, 32473B_176178, 32003B_197524), the Swiss Heart Foundation, the Foundation for Cardiovascular Research Basel and the University of Basel, grants from Bayer, grants from Pfizer, grants from Boston Scientific, grants from BMS, grants from Biotronik, grants and personal fees from Daiichi Sankyo. Nicolas Rodondi received a grant from the Swiss Heart Foundation. Philipp Krisai reports speaker fees BMS/Pfizer. Grants from the Swiss National Science Foundation, Swiss Heart Foundation, Foundation for Cardiovascular Research Basel, Machaon Foundation. Stefanie Aeschbacher received speaker fee from Roche Diagnostics. Stefan Osswald received a grant from the Swiss National Science Foundation. Andreas S. Mueller reports fellowship and training support from Biotronik, Boston Scientific, Medtronic, Abbott/St. Jude Medical, and Biosense Webster; speaker honoraria from Biosense Webster, Medtronic, Abbott/St. Jude Medical, AstraZeneca, Daiichi Sankyo, Biotronik, MicroPort, Novartis, and consultant honoraria for Biosense Webster, Medtronic, Abbott/St. Jude Medcal, and Biotronik. Tobias Reichlin has received research grants from the Swiss National Science Foundation, the Swiss Heart Foundation, and the sitem insel support fund, all for work outside the submitted study. Speaker/consulting honoraria or travel support from Abbott/SJM, Astra Zeneca, Brahms, Bayer, Biosense-Webster, Biotronik, Boston-Scientific, Daiichi Sankyo, Medtronic, Pfizer-BMS and Roche, all for work outside the submitted study. Support for his institution's fellowship program from Abbott/SJM, Biosense-Webster, Biotronik, Boston-Scientific and Medtronic for work outside the submitted study. The other authors report no conflicts. Swiss National Science Foundation (grant numbers 33CS30_148474, 33CS30_177520, 32473B_176178, and 32003B_197524), the Swiss Heart Foundation, the Foundation for Cardiovascular Research Basel (FCVR), and the University of Basel. The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.