Patients with chronic kidney disease (CKD) face increased risk of developing heart failure. There is a strong and independent association between lower estimated glomerular filtration rate (eGFR) and risk of heart failure events. There are variations in eGFR among individual patients depending on whether creatinine or cystatin C is used for estimation of GFR. Cystatin C had been shown to have stronger and more linear associations with incident heart failure compared with creatinine; however, creatinine-based eGFR (eGFRcr) is more widely used in clinical practice than cystatin C-based eGFR (eGFRcys).
Previous studies with population-based cohorts or clinical trials have found associations of the difference in GFR estimates (eGFRdiff, defined as eGFRcys minus eGFRcr). According to Debbie C. Chen, MD, and colleagues, the associations of eGFRdiff with hospitalizations for heart failure have not been examined in a cohort of participants with established CKD. There are also only few available data on the clinical interpretation of changes in eGFRdiff over time.
Dr. Chen et al conducted a prospective cohort study to answer three questions: (1) Among individuals without prevalent heart failure at baseline, is baseline eGFRdiff independently associated with incident heart failure hospitalization? (2) Does the inclusion of time-updated measures yield stronger associations between eGFRdiff and incident heart failure? (3) Is widening or narrowing of eGFRdiff over time independently associated with incident heart failure? Results were reported in the American Journal of Kidney Diseases [2022;80(6):762-771].
The study cohort included adults with CKD and without prevalent heart failure who enrolled in the CRIC (Chronic Renal Insufficiency Cohort) study. The study exposure was the difference in GFR estimates (eGFRdiff [eGFRcys minus eGFRcr]). The outcome of interest was incident heart failure hospitalization. The associations of baseline, time-updated, and slope of eGFRdiff with incident heart failure were investigated using Fine-Gray proportional subhazards regression.
A total of 4512 CRIC study participants were included in the analysis. Of those, 44% (n=1981) were female, 42% (n=1906) were non-Hispanic Black, and 11% (n=447) were Hispanic. Mean age was 59.4 years, eGFRcys was 55 mL/min/1.73 m2 , and mean eGFRcr was 49 mL/min/1.73 m2. Baseline eGFRdiff ranged from –52 to 65 mL/min/1.73 m2.
A total of 2977 patients had a midrange baseline eGFRdiff (eGFRcys similar to eGFRcr); 340 had a negative baseline eGFRdiff (eGFRcys lower than eGFRcr); and 1195 had a positive baseline eGFRdiff (eGFRcys higher than eGFRcr). Those in the negative eGFRdiff group were generally older and had the highest prevalence of diabetes and cardiovascular disease compared with the other two groups. Those in the eGFRdiff positive group were younger and had lower prevalence of baseline comorbidities and medication use.
Twelve percent (n=532) of the study participants developed incident heart failure, with median time to incident heart failure hospitalization of 3.5 years. Following adjustment for demographics and eGFR, there was an association between each 15-mL/min/1.73 m2 lower baseline eGFRdiff and a 56% higher risk of incident heart failure. After additional multivariable adjustment, the association was attenuated to 20%. Participants with negative eGFRdiff had the highest crude rate of incident heart failure, and those with positive eGFRdiff had the lowest incidence of heart failure. Compared with the midrange eGFRdiff group, the positive and negative eGFRdiff groups were not statistically significantly associated with multivariable-adjusted risk of incident heart failure.
Following multivariable adjustments accounting for time-updated eGFRdiff and covariates, there was an association between each 15 mL/min/1.73 m2 lower baseline eGFRdiff and 36% higher risk of incident heart failure. Compared with the group with midrange eGFRdiff, the adjusted risk of incident heart failure was higher among participants in the group with negative eGFRdiff (subdistribution hazard ratio [HR], 1.99; 95% CI, 1.39-2.86), and those in the positive eGFRdiff group had a lower adjusted risk of incident heart failure (subdistribution HR, 0.67; 95% CI, 0.49-0.91). There was modest attenuation with further adjustment for markers of nutritional status and inflammation.
Slopes of eGFRdiff were derived using a median of four annual eGFRdiff values. The mean annual change in eGFRdiff was –0.4 mL/min/1.73 m2. Following adjustment for baseline eGFRdiff, in multivariable models including adjustment for baseline eGFRdiff, each 1-SD lower eGFRdiff slope was associated with 37% higher risk of incident heart failure. Participants with faster declines in eGFRcys relative to eGFRcr had higher risk of incident heart failure (HR, 1.49; 95% CI, 1.19-1.85) compared with those in whom eGFRcys and eGFRcr declined in parallel.
The researchers cited some limitations to the study findings, including entry into the CRIC study being determined by eGFRcr, the use of the absolute rather than relative difference between eGFRcys and eGFRcr, not calibrating cystatin C measures to a traceable international standard, lack of sufficient data on ejection fraction at the time of hospitalization for heart failure, and using a joint model to obtain within-participant estimates of eGFRdiff slope.
In summary, the authors said, “Our study showed that large differences between eGFRcys and eGFRcr convey important prognostic information regarding risk of incident heart failure hospitalization. During longitudinal follow-up, steeper declines in eGFRcys than eGFRcr portend higher risk of heart failure events. Thus, in patients with CKD, annual measures of serum creatinine and cystatin C and separate reporting of both eGFRcys and eGFRcr could optimize the assessment of heart failure risk.”
Takeaway Points
- Researchers reported results of a prospective cohort study designed to assess the clinical implications of differences in estimated glomerular filtration rate based on cystatin C (eGFRcys) and eGFR based on creatinine (eGFRcr) on the risk of heart failure in individuals with chronic kidney disease.
- The study cohort included participants in the CRIC (Chronic Renal Insufficiency Cohort) study, a multicenter observational cohort study that enrolled 5499 adults from seven clinical centers in the United States.
- Among participants with large differences between eGFRcys and eGFRcr, there was an increased risk of incident heart failure; diverging slopes between eGFRcys and eGFRcr over time were also independently associated with the risk of incident heart failure.
