Comparison of glomerular filtration rate estimating equations in Indian geriatric patients with cancer

e24025 Background: Accurate estimation of glomerular filtration rate (eGFR) is essential for dose calculation of certain anticancer drugs. It is particularly important in older cancer patients because of age-related compromise in renal function. 99m Tc-DTPA plasma clearance for GFR measurement poses practical difficulties in the elderly and is expensive. Therefore, we investigated the predictive accuracy of eight commonly used eGFR equations against 99m Tc-DTPA plasma clearance as the gold standard in Indian geriatric oncology population. Methods: A cross-sectional study was conducted in the geriatric oncology clinic of Tata Memorial Hospital, India, from June 2021 to November 2022. Age, sex, weight, BUN and serum creatinine were obtained from the hospital records. Serum cystatin C was estimated using sandwich ELISA (Sigma, Germany). GFR was estimated using Cockcroft-Gault (CG); 6 variable MDRD; 4 variable MDRD; CKD-EPI creatinine; CKD-EPI cystatin; CKD-EPI creatinine-cystatin; BIS 1; and BIS 2 equations. GFR was also measured by 99m Tc-DTPA plasma technique (Gates’ method). The agreement between different eGFR methods and the measured GFR was analyzed using Bland-Altman test. Results: Out of 311 enrolled patients, 276 patients having full covariate data required for estimation of eGFR by all eight equations were included in the analysis. Majority were males (N = 214). Median age (range) of the patients was 68 (60 – 88) years. Median of measured GFR using 99m Tc-DTPA was 71 (17 – 141) ml/min. The performance i.e., bias and 95% Limit of Agreement (LoA) of eGFR equations against measured GFR is shown in table 1. Overall, the BIS 2 equation with least bias of 0.854 and narrowest 95% LoA (-31.23 to 32.94) had the highest agreement with ‘gold standard’. In fact, BIS 2 performed consistently well across age (60-70, 71-80, 81-90) and BMI (≤18.5, 18.6-24.9, ≥25) groups (data not shown). BIS 2 was followed by CKD-EPIcr-cys and CKD-EPIcys with best estimates of bias and 95% LoA (Table 1). Interestingly, all three equations have cystatin C as a covariate. 4vMDRD had the highest bias (-24.08) and least agreement with DTPA (-86.52 – 38.36). Post hoc analysis showed that 59/115 (51.3%) patients who were cisplatin-ineligible (GFR < 60) based on CG were cisplatin-eligible based on BIS 2, while 19/161 (11.8%) patients who were cisplatin-eligible based on CG were actually ineligible as per BIS 2. Conclusions: Equations derived using cystatin C in general, and BIS 2 in particular, have highest predictive accuracy for GFR estimation in older cancer patients. [Table: see text]