The usefullness of simulation techniques for investigating inter individual pharmacodynamic variability of neuromuscular blocking agents

F. Lagneau, M. Sénéchal, M. Tod, J. Marty

Anaesthesia and Intensive Care Unit, Hôpital Beaujon, Université Paris VII, France.

Background and Goal of study: to compare the effects of single bolus infusions of pancuronium (PAN, n=10, 0.1 mg/kg) or cisatracurium (CIS, n=23, 0.15 mg/kg) in patients (normal renal function) undergoing abdominal surgery to those extracted from literature and to partially explain observed differences by simulations.

Materials and Methods: Anaesthesia was induced by thiopental/fentanyl and maintained by O2/NO2/isoflurane (0.5±0.1%) inhalation. The curare was selected depending on duration of surgical procedure. Onset time, duration of action (T0-25) and recovery index (T25-75) were recorded by acceleromyography at the adductor pollicis. Observed data were compared with 1) simulated data using Stanpump software for patients receiving PAN taking account of age and body weight 2) simulated data in 1000 health patients (70 kg) receiving 0.15 mg/kg of CIS using a bicompartmental model with central and peripheral elimination and effect compartment (1). This simulation was performed by implementing the corresponding model in the ADAPT II software and allowing a 10% coefficient of variation for transfer and elimination microconstants and central and peripheral compartment volumes.

Results and discussion: Results are expressed in mean±sd except for ADAPT II simulation (mean, range)

PAN	Onset (s)	T0-25 (min)	T25-75 (min)
Observed	448±192	121±70	41±22
Literature	240±23*	90±12*	NA
Stanpump	427±52	124±23	94±4**

*lit vs obs (p<0.05) ; **Stanpump vs obs (p<0.05)

CIS	Onset (s)	T0-25 (min)	T25-75 (min)
Observed	272±80	71±16	22±19
Literature	204±35*	55±10*	14±6**
ADAPT II	108 (NA)	40(26-95)	80 (28-120)

*lit vs obs (p<0.05) ; **lit vs obs (p=0.051)

Conclusions :Differences between observed data and literature may be explained by a moderate inter individual pharmacokinetic variability in patients extracted from the same population. Taking account of specific covariables as performed by Stanpump simulator could improve pharmacodynamic prediction.

Reference: (1) Bergeron et al, Anesthesiology, 2001