Impact of Frailty on Performance of the Kidney Failure Risk Equation Using Creatinine- and Cystatin C-Based eGFR

Background: There is a high prevalence of frailty in people with chronic kidney disease (CKD). Frailty is associated with increased risk of mortality. Risk predictions equations, such as the kidney failure risk equation (KFRE) can guide management. Consideration of the competing risk of mortality on risk of future need for kidney replacement therapy is important. Decline in muscle mass in frail people may impact accuracy of creatinine-based eGFR. Using Cystatin C based eGFR (eGFRcys) or combined creatinine and Cystatin C eGFR (eGFRcrcys), may improve risk stratification of future kidney failure. We assessed the impact of frailty on performance of KFRE using eGFRcr, eGFRcys and eGFRcrcys in the UK Biobank research cohort. Methods: Adults with CKD G3-5 (eGFR<60mL/min/1.73m2 by any of eGFRcr, eGFRcys and eGFRcrcys) were studied. Frailty was assessed by the Fried, Rockwood and laboratory frailty indices. The outcome was kidney failure, i.e.long-term dialysis or kidney transplantation. KFRE performance at 2- and 5-years was assessed. Results: 24,489 participants were studied; 8,533 (34.8%) had frailty by at least one frailty measure. The frail group had a mean age of 62.4 years (SD 5.7), 57% were female and mean eGFRcr 67.7ml/min/1.73m2 (SD 17.7) compared to mean age 63.1 years (SD 5.6), 52% female and mean eGFRcr 64.5 ml/min/1.73m2 (SD 14.8) in the non-frail group (all p<0.001). The frail group had higher discrepancy between eGFRcys and eGFRcr compared to the non-frail (-15.8 vs -6.9 ml/min/1.73m2, p<0.001). There were 312 (1.27%) kidney failure and 1,471 (6.01%) death events within 5-years. Discrimination power of KFRE was good in individuals with and without frailty (AUC ≥0.87 across all frailty measures and eGFR equations). A higher risk of kidney failure was demonstrated in frail compared to non-frail groups with increased divergence between Kaplan-Meier and Aalen-Johansen cumulative incidence curves suggesting greater impact of the competing risk of death in the frail group. Conclusion: KFRE adequately predicts kidney failure risk in individuals with frailty. GFR estimated using equations containing Cystatin C does not impact performance. Further work developing models considering competing risk of death may refine the risk of kidney failure in individuals with frailty. Funding: Private Foundation Support