Sufentanil disposition in patients undergoing renal transplantation: influence of choice of kinetic model.

We studied the disposition of sufentanil in 10 patients undergoing renal transplantation, and compared the data with those from eight healthy anaesthetized patients undergoing lower abdominal surgery. Patients received sufentanil 2.5 micrograms kg-1 as part of a balanced anaesthetic technique. Central venous blood samples were collected at intervals up to 600 min, and plasma sufentanil concentrations assayed by radioimmunoassay. Pharmacokinetic parameters were calculated from drug concentration-time profiles by extended least squares non-linear regression analysis (ELSFIT) and by a model independent (MI) approach using AUC and its first moment, AUMC. There were no differences (based on MI results) between the two groups for elimination half-life (T1/2 beta) (renal failure: 188 min; anaesthetized controls: 195 min), clearance (CI) (1030 and 1093 ml min-1) and apparent volume of distribution at steady state (Vss) (223.0 and 215.3 litre). Sufentanil binding to plasma proteins was 91.4% in the renal patients and 92.2% in the healthy group (ns). Comparison of kinetic methods showed significant correlation of the individual estimates for T1/2 beta, CI and Vss (P less than 0.01). The mean absolute differences between methods were: T1/2 beta 2.7 min (95% limits: -26.2 to 31.5), CI 36.5 ml min-1 (-5.5 to 78.4), Vss -18.4 litre (-47.7 to 10.9). When the mean estimate for the two methods [ELSFIT + MI)/2) was compared with the difference, there was no correlation for the estimates of CI and Vss+. MI tended to overestimate clearance and underestimate volume of distribution. There was a significant relationship between estimates for elimination half-life, with a slope greater than zero.