A comparison of measured and estimated glomerular filtration rate (GFR) for carboplatin dose calculation in women with ovarian cancer (OC).

Abstract

e17051 Background:Carboplatin based chemotherapy is widely used to treat women with Ovarian cancer (OC) in the adjuvant, neoadjuvant and palliative settings. Carboplatin is normally dosed by the Calvert formula: [GRF+25] x AUC, with a target AUC (area under the plasma concentration curve) of 5-6 mg/ml/min. GFR (glomerular filtration rate) can be precisely measured with radionuclide based ligands such as 51Cr-ethylenediamine tetraacetic acid (51Cr-EDTA), but in clinical practice is often estimated using formulae incorporating serum creatinine, weight and age, such as the Cockcroft–Gault (CG) or Wright (WR) formulae. It is unknown, however, if these formulae allow a sufficiently accurate dosing of carboplatin in OC patients. Methods:We retrospectively compared measured (EDTA-based) and estimated (CG and WR formulae) GFR and the resulting carboplatin doses, in women with stage 1c to IV OC treated with carboplatin AUC 6 at our centre between 2015 and 2016. All patients (pts) had concurrent 51Cr-EDTA GFR and serum creatinine measurement before starting carboplatin. Accuracy and bias of The WR and CG formulae were analysed using mean absolute percentage error (MAPE) and mean percentage error (MPE). Results:92 pts were included in the study (median age = 69 years (r. 25 -89), median serum creatinine 63.5 umol/L (r. 38 – 168)). MPE was -37.3 and -26.0 and MAPE was 38.0 and 32.2 for the WR and CG formulae, respectively. If the WR or CG formulae had been used instead of EDTA to calculate the dose of carboplatin, the discrepancy would have been at least 10% in 77.1% and 69.8% of pts, respectively; a discrepancy of 50% or more would have occurred in 10% and 13% of pts. Conclusions:Neither the WR nor CG formulae precisely predict renal function in ovarian cancer pts. They often overestimate GFR, meaning that if these formulae are used instead of isotope-based measurement of GFR, most pts will receive a higher dose of carboplatin. Our findings are relevant for clinical practice and for the design and interpretation of clinical trials, where the choice between EDTA- or formulae- based calculations of carboplatin dose are often left to investigators’ preference.