Introduction: Area under the concentration-time curve (AUC) has been advocated as a better parameter to monitor cyclosporine A than trough concentrations. Up to now, more than 100 equations to estimate AUC using a limited sampling strategy have been published, but not all have been validated.
Material and methods: Eight equations for AUC0-12h and two for AUC0-8h were validated. Concentrations of cyclosporine A were analyzed by high-performance liquid chromatography (HPLC) and a specific radioimmunoassay (RIA) method. Forty male renal transplant patients were included in the study. Blood samples were taken predose and at 0.5, 1, 1.5, 2, 3, 5, 8, and 12 hours after the morning dose when the patient was in steady state. The percentage prediction error (%pe) was used for an assessment of the performance of the equations. Mean %pe less than ± 15% and absolute %pe less than 30% in 95% of predictions were considered to be acceptable. Other possibilities such as %pe less than 25%, 20%, and 15% were also tested.
Results: Eight equations for AUC0-12h met the requirements using both assays, six in the HPLC set only and four in the RIA set only. The highest precision was obtained with AUC0-12h = 123.792 + 1.165*C1h + 3.021*C3h + 7.33*C8h proposed by de Mattos et al. The mean %pe was 1% ± 8% (-16 to 19) for HPLC (values given as mean ± standard deviation [range]) and -1 ± 5 (-17 to 10) for RIA. Mean absolute %pe was 7 ± 5 (0.0 to 19) for HPLC and 4 ± 4 (0.0 to 17) for RIA. For clinical use, the most suitable equation was AUC0-12h = 363.078 + 8.77*C1h + 3.07*C3h proposed by Wacke et al, which produced the second lowest %pe and used two sampling points in the period of 1 to 3 hours after dose. The mean %pe was -7 ± 10 (-25 to 25) for HPLC and 2.3 ± 6 (-10 to 17) for RIA. Mean absolute %pe was 10 ± 7 (0.4 to 25) for HPLC and 5 ± 4 (0.0 to 17) for RIA. The equation: AUC0-8h = 55.37 + 2.89*C0h + 1.08*C1h0.9*C2h + 2.23*C3h proposed by Foradori et al met the criteria with 95% of prediction with absolute %pe less than 15% in the HPLC set and 10% in the RIA set.
Conclusion: The validation of equations is of major importance for prediction precision, whereas the analytical method for limited sampling strategy proposals had no influence. Because of the wide interassay variability, it is also important to know which analytical method was used for AUC calculation when interpreting the results.