#2295 Can we detect the evolution of GFR with estimated GFR in patients with ADPKD?

Abstract

Abstract Background and Aims ADPKD affects about 10% of the patients with CKD. In the disease the formation of cysts replace normal renal parenchyma leading to renal function loss. The evolution of the disease is irregular: some patients lose renal function very fast while others remain stable or have a slow loss. This may have consequences in clinical care. In day-to-day practice, GFR is evaluated by means of formulas, which are algorithms that use creatinine and others markers like age and gender to estimate GFR. However, the accuracy and precision of eGFR in reflecting real GFR in ADPKD is not clear, being this the objective of our study. Method To analyse the capacity of eGFR by creatinine-based formulas in predicting mGFR changes over time. We evaluated patients with ADPKD, who underwent mGFR and eGFR by a group 5 equations at baseline, month 6 and then annually. mGFR was evaluated by the clearance of iohexol using dried blood spots (iohexol-DBS) and eGFR with the following creatinine-based equations: Cockcroft-Gault, aMDRD, CKD-EPI, FAS and EKFC. We calculated renal function decline (GFR decline) using repeated measured with mGFR and eGFR. All subjects have at least 3 repeated evaluations of renal function. The agreement between mGFR and eGFR declines was analysed by the concordance correlation coefficient (CCC). Also, we evaluated the capacity of eGFR decline in detecting diverse degree of GFR decline (a) rapid progressors: mGFR of −5 ml/min/y; (b) moderate progressors: −2 to −5 ml/min/y; (c) stable: > −2 ml/min/y. Results We evaluated 111 (43; 37% men); age: 42 y ± 14, with a 5 ± 1 evaluations of GFR over time (mGFR and eGFR). The average GFR decline was: −2,8 ml/min/year (mGFR); −3,1 (Cockcroft-Gault), −2,9 (aMDRD), −3,0 (CKD-EPI), −2,8 (FAS) and −2,6 (EFKC) According with mGFR decline, 21 patients (19%) were rapid progressors; 46 (41%) moderate progressors and 44 (40%) stable. The agreement between eGFR decline and mGFR decline was poor, reflected by a low CCC: 0.37 (Cockcroft-Gault); 0.76 (aMDRD); 0.80 (CKD-EPI), 0.77 (FAS); 0.81 (EFKC). In general, eGFR decline failed to detect progression of renal function over time. On average, 43% of the rapid progressors, 53% of the moderate progressors and 54% of the stable patients were not detected by any formula, respectively. Finally, a total of 12% and 6% of those stable or moderate progressors were falsely classified as rapid progressors. Conclusion eGFR is not able to reflect properly renal function changes over time in patients with ADPKP in about half of the patients. A relevant number of cases are erroneously considered as rapid progressors or even stable patients. This may have clinical consequences. Whenever possible, mGFR could be considered in this population.