MO155: The Importance of Measured GFR in Clinical Practice: An Old Knowledge for Nephrologists, a New Challenge for Oncologists and Surgeons

Abstract

BACKGROUND AND AIMS In daily clinical practice, an accurate assessment of renal function is of paramount importance in several categories of patients where the adequate medical or surgical treatment depends on the values of glomerular filtration rate (GFR). In particular, both oncological and urological patients require a personalized medical approach able to avoid misleading errors which could dramatically shorten their lifespan. Unfortunately, the most used method to measure GFR in these groups of patients is represented by the estimated glomerular filtration rate (eGFR), which harbours a significant error in comparison to gold standards methods (mGFR). The aim of this study was to determine the extent of the error of eGFR compared to the mGFR in a consecutive prospective cohort of oncological patients affected by urological malignancies. METHOD A total consecutive cohort of 352 patients enrolled in a single tertiary institution between 2018 and 2021 was collected in order to compare the most common eGFR formulas used by physicians (Cockroft-Gault, MDRD, CKD-EPI based on serum creatinine and/or serum cystatin and the new eGFR equation based on creatinine and cystatin without race adjustment) with the most widespread mGFR method (Iohexol Plasma Clearance). The study cohort was composed by 188 oncological patients affected by different types of urological malignancies (cases) before surgical operation and/or medical treatment and 164 non-oncological patients (controls) matched for baseline clinical variables and GFR. The agreement between eGFR and mGFR was evaluated using bias (as median of difference), precision (as interquartile range of difference—IQR) accuracy (as P30) and total deviation index (TDI). The differences between cohorts were evaluated with Fisher's exact test and Chi-squared test for ordinal characteristics and Wilcoxon rank sum test for continuous variables. Data analysis was performed using programming language R and Python. RESULTS Clinical data were as follows: median age 68 (IQR: 20.68), M/F ratio 3.19, median BMI 24.9 (IQR: 0.003, 24.938). A total of 61.2% of patients had hypertension, and 14.3% were diabetics. The median creatinine value in the overall population was 1.48 mg/dL (IQR: 0.45, 1.48); the median cystatin value was 1.26 mg/dL (IQR: 0.42, 1.26). Based on iohexol plasma clearance, 3.13% of patients were classified in CKD stage 1, 23.30% in stage 2, 29.55% in stage 3a, 27.27% in stage 3b, 14.77% in stage 4 and 1.99% in stage 5. The two matched cohorts of oncological and non-oncological patients displayed no statistical differences in terms of clinical variables and agreement parameters (TDI, CCC and P30). Surprisingly, both groups harboured a non-negligible error in each CKD class with a huge discrepancy between the eGFR formulas and the gold standard method (Figures 1 and 2), suggesting the great relevance of mGFR in the clinical decision-making algorithm, both in oncological and in non-oncological patients. CONCLUSION We observed that the error in the classification of CKD stages using eGFR formulas was very common, both in oncological and in nephrological patients, with a poor agreement with mGFR in all CKD classes. Therefore, mGFR remains a crucial tool for a correct clinical decision in all onconephrological patients who require surgery or potential nephrotoxic agents.