PURPOSE: The purpose of this study was to determine the potential utility of a novel algorithm to calculate individual GFR values in cancer patients. Based on carboplatin AUC measurements the algorithm-based values were compared with results related to other routinely used equations. METHODS: The association between measured and predicted carboplatin AUC was examined by the Bland-Altman analysis to determine bias and precision. Based on the Calvert formula, GFR values assessed by different routes of calculation including the novel algorithm were compared with each other in individual patients. RESULTS: The mean absolute administered carboplatin dose was 498 mg and the mean measured carboplatin AUC 5.8 mg/ml × min. Compared to the novel algorithm, the degree of bias to calculate carboplatin AUC was greater with the Cockcroft-Gault, Chatelut, Hoek and Schmitt formula which includes cystatin C as a parameter. In selected patients, algorithm-based GFR values were closer to GFR according to the Calvert formula than results of other equations, including the Jelliffe formula. CONCLUSION: These results suggest that the concept of a non cystatin C-based novel algorithm including three different formulas rather than one single equation may improve accurate estimation of GFR over a broad range of constitutive values, including patients with low constitutive renal function as well as overweight patients.
