A NEW INDEX BASED ON THE DIFFERENCE BETWEEN CREATININE-BASED EGFR AND CYSTATIN-C EGFR ASSOCIATES WITH THE PHYSICAL COMPONENT OF THE SF-36 AND PREDICTS CARDIOVASCULAR EVENTS AND PROGRESION TO KIDNEY FAILURE IN CKD.

Abstract Background and Aims Analyses in a large clinical trial in hypertensive patients (SPRINT) and in the Cardiovascular Health study cohort coherently showed that the difference between the GFR as estimated by serum cystatin and creatine (eGFRDiff) associates with frailty and predicts a lower risks for adverse outcomes including frailty, falls, cardiovascular events and mortality. Whether eGFRdiff in CKD patients associates with frailty metrics like the physical component summary (PCS) of the SF-36 and cardiovascular (CV) and kidney outcomes has not been studied. Method In a cohort of 757 with stages 2-5 CKD we tested the relationship between eGFRDiff with PCS (n=582) and in the whole cohort investigated the relationship between the same indicator with the incidence rate of two combined endpoint: non fatal CV events/death and renal events (dialysis/transplantation/eGFR reduction >30%)/death over a median follow up of 35.6 months (inter-quartile range 21.3-36.2months). Results The eGFRDiff was strongly related to PCS (rho=0.25, P<0.001) and physical functioning (rho=0.26, P<0.001) emerged as the strongest correlate of eGFRDiff among PCS sub-domains. At baseline, eGFRDiff had a median value of 7.0 ml/min/1.73m2 (interquartile range: 2.2-13.0 ml/min/1.73m2) and h P=0.001), serum phosphate (β=-0.11, P=0.003), and BMI (β=-0.08, P=0.043) had an independent relationship with the eGFRDiff. During follow-up, 118 patients had fatal (n=29) and non fatal (n=89) CV events and 13 died of causes other than CV. Overall, 131 patients had the combined endpoint of non fatal CV event/death. Furthermore, 246 patients had renal outcomes and 276 patients had the combined end-point renal events/death. On univariate Cox regression analyses, 1 unit increase in eGFRdiff associated with a 3% reduction of the HR of non fatal CV events/death (hazard ratio: 0.965, 95% CI: 0.951-0.978, P<0.001) and a 2% reduction of the HR of renal events/death ( 0.979, 95% CI: 0.969-0.990, P<0.001). Data adjustment for potential confounders (age, gender, smoking, diabetes, cardiovascular comorbidities, antihypertensive therapy, BMI, systolic BP, haemoglobin, albumin total cholesterol, 24h urinary protein, phosphate, and C-reactive protein) did not materially modify the eGFRdiff- non fatal CV events-death (HR: 0.958, 95% CI: 0.939-0.977, P<0.001) as well as the eGFRdiff- renal events-death (HR: 0.984, 95% CI: 0.971-0.997, P=0.014). Conclusion The eGFRDiff, a new biomarker of frailty, associates with the Physical Component summary of SF36 and predicts cardiovascular events, progression to kidney failure and death independently of other risk factors in CKD patients. Considering eGFRDiff as a marker of patients’ functional status may be helpful to nephrologists as an indicator of poor prognosis.