Cystatin C Predicts Adverse Outcomes In Heart Failure With Preserved Ejection Fraction

Background Serum creatinine (Cr) is used to estimate glomerular filtration rate (GFR). Cr levels vary with lean muscle mass and diet, confounding measurement of renal function. Cystatin C (CysC) is an alternative marker used to estimate GFR which is not subject to these sources of variability. In conditions where frailty and low muscle mass are common, such as heart failure with preserved ejection fraction (HFpEF), Cr may overestimate GFR, leading to under detection of renal disease. Chronic kidney disease (CKD) is a common co-morbidity in HFpEF and portends worse prognosis; however, there are currently minimal data on CysC in HFpEF. Methods Patients enrolled in the Johns Hopkins HFpEF Program from July 2014 to March 2020 were included for study. Patients underwent a comprehensive clinical evaluation and echocardiography. Estimated GFR was calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equations for Cr (CKD-EPIcr) and CysC (CKD-EPIcys). Distributions of GFR were compared with a Wilcoxon matched-pairs test. Trends in clinical and echocardiographic variables with CysC were assessed using multivariate linear regression. Cox proportional hazard models examined the association of CysC with the combined outcome of heart failure hospitalization and all-cause mortality. Results The study cohort included 190 patients. The average age was 64 years; 65% of patients were female and 59% were black. Median follow-up was 28 months. The median estimated GFR by CKD-EPIcr and CKD-EPIcys were 54.0 mL/min/1.73m2 and 46.7 mL/min/1.73m2, respectively (Figure A and C; z=3.81, p<0.001). As compared to CKD-EPIcr, CKD-EPIcys reclassified 74 patients (39% of cohort) to a higher stage of CKD (Figure B and D). CysC was significantly associated with greater left ventricular mass index (β =7.16 [0.23-14.01], p=0.043) and E/e’ ratio (β=2.28 [0.33-4.22], p=0.022) after adjustment for GFR by CKD-EPIcr. CysC was predictive of the combined outcome in the univariate model and after adjustment for relevant medical comorbidities and estimated GFR by CKD-EPIcr (Figure). Conclusion In this prospective study, CysC estimated a lower median GFR in HFpEF patients as compared to Cr. CysC was predictive of worse diastolic function and adverse outcomes, even after accounting for GFR by CKD-EPIcr and relevant comorbidities. These results suggest CysC provides clinical and prognostic information beyond traditional Cr-based assessment of renal function in HFpEF. Further investigation of the use of CysC in clinical practice is warranted. Serum creatinine (Cr) is used to estimate glomerular filtration rate (GFR). Cr levels vary with lean muscle mass and diet, confounding measurement of renal function. Cystatin C (CysC) is an alternative marker used to estimate GFR which is not subject to these sources of variability. In conditions where frailty and low muscle mass are common, such as heart failure with preserved ejection fraction (HFpEF), Cr may overestimate GFR, leading to under detection of renal disease. Chronic kidney disease (CKD) is a common co-morbidity in HFpEF and portends worse prognosis; however, there are currently minimal data on CysC in HFpEF. Patients enrolled in the Johns Hopkins HFpEF Program from July 2014 to March 2020 were included for study. Patients underwent a comprehensive clinical evaluation and echocardiography. Estimated GFR was calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equations for Cr (CKD-EPIcr) and CysC (CKD-EPIcys). Distributions of GFR were compared with a Wilcoxon matched-pairs test. Trends in clinical and echocardiographic variables with CysC were assessed using multivariate linear regression. Cox proportional hazard models examined the association of CysC with the combined outcome of heart failure hospitalization and all-cause mortality. The study cohort included 190 patients. The average age was 64 years; 65% of patients were female and 59% were black. Median follow-up was 28 months. The median estimated GFR by CKD-EPIcr and CKD-EPIcys were 54.0 mL/min/1.73m2 and 46.7 mL/min/1.73m2, respectively (Figure A and C; z=3.81, p<0.001). As compared to CKD-EPIcr, CKD-EPIcys reclassified 74 patients (39% of cohort) to a higher stage of CKD (Figure B and D). CysC was significantly associated with greater left ventricular mass index (β =7.16 [0.23-14.01], p=0.043) and E/e’ ratio (β=2.28 [0.33-4.22], p=0.022) after adjustment for GFR by CKD-EPIcr. CysC was predictive of the combined outcome in the univariate model and after adjustment for relevant medical comorbidities and estimated GFR by CKD-EPIcr (Figure). In this prospective study, CysC estimated a lower median GFR in HFpEF patients as compared to Cr. CysC was predictive of worse diastolic function and adverse outcomes, even after accounting for GFR by CKD-EPIcr and relevant comorbidities. These results suggest CysC provides clinical and prognostic information beyond traditional Cr-based assessment of renal function in HFpEF. Further investigation of the use of CysC in clinical practice is warranted.