Mats O. Karlsson1, Tarja Naukkarinen1,2 and Hans Lennernds1
1 Div of Pharmacokinetics, Faculty of Pharmacy ,Uppsala University, Sweden 2 Orion Corp., Helsinki, Finland
Patients developing ParkinsonÃs disease respond well to treatment with LD, but its short half-life necessitates frequent dosing and often leads to poor control (so-called on-off phenomena). Simultaneous administration of DDC inhibitors, that partly inhibit LD elimination, is standard therapy. The other major route of LD elimination, degradation by COMT, can also be inhibited. Using data in healthy volunteers and patients, we have developed a population PK/PD model for a therapy with LD, a DDC inhibitor (carbidopa) and a COMT inhibitor (entacapone). The model was developed to qualitatively and quantitatively describe the system and to aid optimisation of this triple therapy.
PK data from a volunteer study and PK and PD data from a patient study were available. All studies were of the cross-over type, where the dose of COMT inhibitor was varied over five dose levels (0 to 400 mg entacapone). LD/carbidopa were given as fixed doses in the volunteer study, whereas individualised doses were used in the patient study. PK data were available for entacapone, carbidopa, LD and LD metabolites 3-O-methyl-dopamine (3-OMD) and DOPAC. PD data in patients came from a performance score (Unified ParkinsonÃs Disease Rating Scale, UPDRS). A decrease of 20% in UPDRS was considered an adequate response.
The analysis was performed sequentially. One variable at a time was analysed, conditioned upon the (linearly interpolated) observations of other variables. In the volunteer study the following analyses were performed: (i) 3-OMD conditioned on LD and entacapone, (ii) DOPAC conditioned on LD, carbidopa and entacapone, (iii) LD conditioned on carbidopa and entacapone, (iv) carbidopa, and (v) entacapone. Interactions in the disposition were modelled as concentration-dependent and mechanistic (via COMT or DDC), whereas interactions in the absorption were modelled as dose dependent and empirical. In multiple drug systems with interactions, parameter correlations within an individual between drugs need to be considered. Such correlations were investigated with particular emphasis on coincidence of absorption. For that purpose flexible absorption models, comparable for LD, carbidopa and entacapone, were used. The pharmacodynamic data were analysed using an inhibitory sigmoidal Emax model with an effect-compartment.
The model indicated that the two pathways for systemic elimination of LD were of similar importance in the absence of inhibitors. Whereas DDC inhibition could be described by a single homogeneous enzyme model, a more complex model was required for COMT. Entacapone influenced the absorption of both LD (slower absorption) and carbidopa (lower bioavailability). There was a considerable correlation in the absorption rates between the three drugs. The maximal decrease UPDRS was about 50%, with a relatively steep (slope=2.9) relationship with the effect-compartment concentration of LD. The half-life of the effect equilibration was 0.4 hours and the EC50 was 1060 ng/ml.
Using the developed model the effects of different dosing strategies were investigated. Specifically, (i) changing size of entacapone and/or carbidopa doses, and (ii) the addition of a slow-release component to entacapone and/or carbidopa. For each scenario the following was evaluated: (i) time-course of LD elimination, (ii) relative size of elimination routes, (iii) time-course of LD concentration, (iv) time-course of UPDRS, and (v) maximal dosing interval possible while maintaining adequate response (“on-time”), and (vi) maximal effect-site concentration.
Reference: PAGE 5 (1996) Abstr 562 [www.page-meeting.org/?abstract=562]
Poster: oral presentation