P O Sepúlveda1, X Mora. 1. Universidad del Desarrollo, Santiago de Chile, Chile. pasevou@mi.cl
Abstract
BACKGROUND: The first order plasma-effect-site equilibration rate constant (k(e0)) links the pharmacokinetics (PK) and pharmacodynamics (PD) of a given drug. This constant, calculated for each specific PK drug model, allowed us to predict the course of the effect in a target controlled infusion (TCI). The PK-PD model of propofol, published by Schnider et al., calculated a k(e0) value of 0.456min(-1) and a corresponding time to peak effect (t peak) of 1.6min. The aim of this study was to reevaluate the k(e0) value for the predicted Schnider model of propofol, with data from a complete effect curve obtained by monitoring the bispectral index (BIS). METHODS: The study included 35 healthy adult patients (18-90 years) scheduled for elective surgery with standard monitoring and using the BIS XP(®) (Aspect), and who received a propofol infusion to reach a plasma target of 12 μg/ml in 4min. The infusion was then stopped, obtaining a complete effect curve when the patient woke up. The Anestfusor™ (University of Chile) software was used to control the infusion pumps, calculate the plasma concentration plotted by Schnider PK model, and to store the BIS data every second. Loss (LOC) and recovery (ROC) of consciousness was assessed and recorded. Using a traditional parametric method using the "k(e0) Objective function" of the PK-PD tools for Excel, the individual and population k(e0) was calculated. Predictive Smith tests (Pk) and Student t test were used for statistical analysis. A P<.05 indicated significance. RESULTS: The evaluation included 21 male and 14 female patients (18 to 90 years). We obtained 1,001 (±182) EEG data and the corresponding calculated plasma concentration for each case. The population k(e0) obtained was 0.144min(-1) (SD±0.048), very different from the original model (P<.001). This value corresponds with a t peak of 2.45min. The predictive performance (Pk) for the new model was 0.9 (SD±0.03), but only 0.78 (SD±0.06) for the original (P<.001). With a baseline BIS of 95.8 (SD±2.34), the BIS at LOC was 77.48 (SD±9.6) and 74.65(SD±6.3) at ROC (P=.027). The calculated Ce in the original model at LOC and ROC were 5.9 (SD±1.35)/1.08 μg/ml (SD±0.32) (P<.001), respectively, and 2.3 (SD±0.63)/2.0 μg/ml (SD±0.65) (NS) for the new model. The values between LOC/ROC were significantly different between the 2 models (P<.001). No differences in k(e0) value were found between males and females, but in the new model the k(e0) was affected by age as a covariable (0.26-[age×0.0022]) (P<.05). CONCLUSIONS: The dynamic relationship between propofol plasma concentrations predicted by Schnider's pharmacokinetic model and its hypnotic effect measured with BIS was better characterized with a smaller k(e0) value (slower t½k(e0)) than that present in the original model, with an age effect also not described before.
BACKGROUND: The first order plasma-effect-site equilibration rate constant (k(e0)) links the pharmacokinetics (PK) and pharmacodynamics (PD) of a given drug. This constant, calculated for each specific PK drug model, allowed us to predict the course of the effect in a target controlled infusion (TCI). The PK-PD model of propofol, published by Schnider et al., calculated a k(e0) value of 0.456min(-1) and a corresponding time to peak effect (t peak) of 1.6min. The aim of this study was to reevaluate the k(e0) value for the predicted Schnider model of propofol, with data from a complete effect curve obtained by monitoring the bispectral index (BIS). METHODS: The study included 35 healthy adult patients (18-90 years) scheduled for elective surgery with standard monitoring and using the BIS XP(®) (Aspect), and who received a propofol infusion to reach a plasma target of 12 μg/ml in 4min. The infusion was then stopped, obtaining a complete effect curve when the patient woke up. The Anestfusor™ (University of Chile) software was used to control the infusion pumps, calculate the plasma concentration plotted by Schnider PK model, and to store the BIS data every second. Loss (LOC) and recovery (ROC) of consciousness was assessed and recorded. Using a traditional parametric method using the "k(e0) Objective function" of the PK-PD tools for Excel, the individual and population k(e0) was calculated. Predictive Smith tests (Pk) and Student t test were used for statistical analysis. A P<.05 indicated significance. RESULTS: The evaluation included 21 male and 14 female patients (18 to 90 years). We obtained 1,001 (±182) EEG data and the corresponding calculated plasma concentration for each case. The population k(e0) obtained was 0.144min(-1) (SD±0.048), very different from the original model (P<.001). This value corresponds with a t peak of 2.45min. The predictive performance (Pk) for the new model was 0.9 (SD±0.03), but only 0.78 (SD±0.06) for the original (P<.001). With a baseline BIS of 95.8 (SD±2.34), the BIS at LOC was 77.48 (SD±9.6) and 74.65(SD±6.3) at ROC (P=.027). The calculated Ce in the original model at LOC and ROC were 5.9 (SD±1.35)/1.08 μg/ml (SD±0.32) (P<.001), respectively, and 2.3 (SD±0.63)/2.0 μg/ml (SD±0.65) (NS) for the new model. The values between LOC/ROC were significantly different between the 2 models (P<.001). No differences in k(e0) value were found between males and females, but in the new model the k(e0) was affected by age as a covariable (0.26-[age×0.0022]) (P<.05). CONCLUSIONS: The dynamic relationship between propofol plasma concentrations predicted by Schnider's pharmacokinetic model and its hypnotic effect measured with BIS was better characterized with a smaller k(e0) value (slower t½k(e0)) than that present in the original model, with an age effect also not described before.