Limited sampling strategies (LSS) have been proposed for predicting total exposure of tacrolimus, a widely used immunosuppressant in transplantation. This study aims to validate the equation developed by Wong et al for estimation of the tacrolimus area-under-the-concentration-over-12-hour-curve (AUC0-12) and to assess the effects of hemoglobin and duration of tacrolimus therapy on predictive performance of the equation in adult kidney transplant recipients.
Seven time point blood concentration profiles were collected from 31 stable kidney transplant recipients who received oral tacrolimus twice daily. The chemiluminescent microparticle immunoassay method was used to determine the tacrolimus concentration. Measured AUC0-12 (AUCm) was calculated by the linear trapezoidal rule. Predicted AUC0-12 (AUCp) was calculated using the equation that used tacrolimus concentrations measured at 2 hours (C2) and 4 hours (C4) after dose: 16.2 + 2.4(C2) + 5.9(C4). Predictive performance of the equation was determined by calculating bias and precision. Agreement between AUCp and AUCm was assessed. The effects of hemoglobin and duration of tacrolimus therapy on bias and precision were also evaluated.
The median (interquartile range) of AUCm was 133.00 (98.25, 185.70) ng·h·mL−1. The AUCp well correlated with the AUCm (r2 = 0.962, P < 0.001). The equation had a mean percentage prediction error of −2.22% (95% CI, −5.14 to 0.71), mean absolute percentage prediction error of 6.67% (95% CI, 4.92–8.42), and root mean squared error (%CV) of 14.08 (10.29%) ng·h·mL−1. A Bland–Altman plot showed good agreement between AUCp and AUCm with a mean bias of −5.43 ng·h·mL−1 (95% CI, −10.28 to −0.59). The hemoglobin level and duration of tacrolimus therapy did not influence the predictive performance of the equation.
The equation had low bias and high precision in predicting the AUC0-12 of tacrolimus. The equation is a simple and reliable tool for estimating tacrolimus exposure.