Mean residence time as estimated from cropped and truncated moments

For non-compartmental concentration-time curve analysis the mean residence time (MRT) is usually estimated using numerical summation and log-linear extrapolation. However, when the initial ear of the curve is cropped and the terminal tail is truncated the MRT can be estimated without extrapolation using numerical methods alone. This paper is to evaluate the error of the cut MRT derived from cropped and truncated moments. The cut MRT is estimated from exclusively measured but non extrapolated concentrations. Mono-exponential kinetics after single intravenous dosing is considered. Two conditions define the acceptable cutoff error of 5% or less: For cropping, the first concentration must be measured immediately after dosing at time t1 that is 0.14 times the elimination half-life (t1 = 0.14 x T1/2). For truncation, the concentration must be observed for a near 2 log decline period (tn) or, exactly, 6.51 times the elimination half-life (tn = 6.51 x T1/2). From data of previous studies, the MRT truncation error of only 0.39% for methylprednisolone (CAS 2375-3-3) was estimated with an observation period of 10 times one half-life (24 h = 10.4 x 2.3 h) but 4.3% for cyclosporin (CAS 59865-13-3), with an investigation interval of only 2 times one half-life (12 h = 2 x 6.1 h). For the complex kinetics of raloxifene (CAS 82640-04-8) it was seen that the cropping and truncation errors are non additive.

If a reliable extrapolation is impossible, the cut MRT can be estimated only from measured concentrations. However, the cropping and the truncation maneuvers should be considered simultaneously to counterbalance the cutoff errors.