Significant discrepancy in cyclosporin A post-dose concentrations when analyzed with specific RIA and HPLC method

C(2) or AUC sparse sampling methods are widely recommended for therapeutic monitoring of cyclosporin A (CsA). One additional reason for promoting the C(2) sampling time in place of commonly used C(0) is that the C(2) level may actually provide more accurate measurement of parent drug concentration by immunoassays, as lower portion of metabolites has been formed 2 hours post-dose than at the steady-state trough time point. HPLC and RIA whole blood levels of CsA and its main metabolites AM1, AM9 and AM4N were compared during 12 hours profile after chronic administration. 40 stable renal transplant male patients (age 49 +/- 6 years, body weight 76 +/- 7 kg) were treated with CsA (Sandimmun Neoral, Novartis s.r.o, Prague, Czech Republic) in doses 198 +/- 56 mg twice daily. Samples were collected in steady state (after 2 weeks of regular treatment regimen) as follows: pre-dose, 0.5, 1, 1.5, 2, 3, 5, 8 and 12 hours after dose. CsA concentrations were determined both specific RIA assay (Cyclo-Trac SP Whole, Dia Sorin) and HPLC method, where concentrations of metabolites AM1, AM9 and AM4N were simultaneously analyzed. The AUC(0-12) was calculated by the linear trapezoidal rule. The percentage prediction error defined as [(RIA value-HPLC value)/HPLC value] x 100 was used for estimation of differences. C(max), t(max), and C(avg) were compared using Student's t-test. RIA produced significantly higher CsA levels than HPLC method in the period of 0.5 - 5 hours after application. The greatest differences (43 - 56%) occurred between 1 and 3 hours after dose. AUC(0-12), C(max) a C(avg) calculated from RIA results were consequently significantly higher. Only t(max) remained unchanged. The ratio of metabolites/parent drug after CsA intake is decreasing but their absolute concentrations are significantly increasing. Mean levels at C(0)/C(2) were CsA-RIA 82/612, CsA-HPLC 89/425, AM1 121/179, AM9 4.1/81.4, AM4N 9.5/21.0 ng/ml. TDM using C(2) and AUC sparse sampling may cause misleading interpretation using both methods alternately for the same patient.