PI‐83

BACKGROUND AQUAVAN® Injection (AI) is a water‐solubleprodrug of propofol. AI population PK/PD model was developed earlier using PKand MOAA/S (Modified Observer's Assessment of Alertness/Sedation) data from acolonoscopy study. METHODS Model‐based simulations were used to design a newstudy: (i) Choose initial bolus doses that produce deep sedation (MOAA/S than 5min) in < 5% of patients; (ii) Choose titration regimens that providesedation to most patients without sedating them too deeply; (iii) Optimizeweight adjustment of dosing to maximize % of sedated patients while minimizing% of deeply sedated patients; (iv) Assess power and choose sample size withrespect to sedation success. RESULTS The following dosing paradigm should balanceefficacy and safety outcomes: weight‐proportional dosing with boundaries at 60and 90 kg (patients weighing < 60 or > 90 kg are dosed as 60 or 90 kgpatients), initial doses of 5–8 mg/kg, and supplemental doses of 25% of theinitial administered at 4 and 8 min. Initial doses should sedate patients whoare more responsive while avoiding deep sedation. Supplemental doses shouldsedate most patients who are less responsive to sedation. With 25 patients pergroup, 5 mg/kg group is differentiated from 2 and 8 mg/kg groups with 99% and79% power, respectively. CONCLUSIONS The model‐based simulations optimized thedesign with initial doses of 2, 5, 6.5, and 8 mg/kg with boundaries at 60 and90 kg. The validity of the predictions will be tested upon completion of thestudy. Clinical Pharmacology & Therapeutics (2005) 79 ,P28‐P28; doi: 10.1016/j.clpt.2005.12.104