Cyclosporine A monitoring--how to account for twice and three times daily dosing.

Cyclosporine A (CsA) dose-interval pharmacokinetic profiles, performed 1-4 years post-transplantation, were collected from 74 renal transplanted children. Forty patients were on three times daily dosing (t.i.d.) and 34 on twice daily dosing (b.i.d.). Regression models for prediction of area under the curve (AUC) using 1-3 concentration time points as independent variables were developed. With similar weight-adjusted single doses (mg kg(-1)) of CsA, t.i.d. dosing resulted in a trough-concentration (C0) similar to that from b.i.d. dosing, but a 30% lower 2 h post-dose concentration (C2). For b.i.d. dosing the relationship between C0 and AUC was poor (r2=0.23) and the prediction error was large (5.8+/-33.5%). For t.i.d. dosing the relationship was better (r2=0.79), but prediction error was still large (4.5+/-24.9%). For C2 relationships were similar to those for the b.i.d. (r2=0.59) and t.i.d. (r2=0.63) groups, but explained modestly the variations of AUC (prediction error=2.6+/-16.8% b.i.d., 4.8+/-23.2% t.i.d.). Both C0 and C2 are useful monitoring methods when CsA is administered t.i.d. If the aim is similar specified daily drug exposure, the target C2 should be roughly 30% smaller in t.i.d. dosing than in b.i.d. dosing and the target C0 could be similar. The prediction error of AUC can be large in individual patients when using single time-point determinations, however. The use of multiple time points reduces the variation, but is less feasible.