The reversal of rocuronium-induced neuromuscular block by the cyclodextrin ORG25969: model development and validation
B. Ploeger(2), N. Houwing(1), A. Bom(3), M.P. van Iersel(1), D. Zollinger(1), M. Danhof(2)
(1) NV Organon, Oss, The Netherlands; (2) LAP&P Consultants BV, Leiden, The Netherlands; (3) NV Organon, Newhouse, United Kingdom
Objectives: ORG25969 is a chemically optimized cyclodextrin designed to selectively bind to rocuronium, leading to a reversal of rocuronium induced neuromuscular block (NMB). It is hypothesized that this reversal is due to a rapid decrease in the unbound rocuronium concentration. This hypothesis was evaluated and validated using a model-based analysis, in which the in vitro dissociation constant determined by isothermal microcalorimetry was used as a predictor of the complexation process in vivo.
Methods: A population approach was applied using NONMEM, to evaluate if the complexation between ORG25969 and rocuronium could be described best either under the assumption of instantaneous equilibrium between unbound rocuronium (steady state model) or by assuming hysteresis between the association of rocuronium to, or dissociation from, the ORG25969-rocuronium complex (dynamic model). First, a PK-PD model for rocuronium alone was developed. After fixing the parameters of this model to the individual-specific posthoc estimates, both interaction models were optimized using data from a phase I study, in which 10 male subjects received 0.1-8 mg/kg ORG25969 3 minutes after 0.6 mg/kg rocuronium. Subsequently, the models were applied to predict the observed change in rocuronium-induced NMB after ORG25969 administration. The model with the greatest predictability was further validated using a predictive check of another study (n=99), in which ORG25969 (0-8 mg/kg) was administered 3, 5 or 15 minutes after 0.6 mg/kg rocuronium.
Results: As evaluated with a posterior predictive check, the dynamic interaction model describes the observed PK and PD data after administration of ORG25969 considerably better as compared to the steady state interaction model. Subsequently, external model validation with a predictive check confirmed that the optimized dynamic interaction model with the in vitro kd of 0.1 µM could adequately predict the observed increase in the total rocuronium concentration and reversal of rocuronium-induced NMB.
Conclusions: Model based evaluation showed that the reversal of rocuronium-induced NMB after ORG25969 administration is due to a rapid decrease in the unbound rocuronium concentration. As the model was validated on an external dataset, the model can be applied for simulation of clinically relevant questions, such as selecting the dose schedule for optimal reversal of rocuronium-induced NMB.