Population Pharmacokinetic/Pharmacodynamic (PK/PD) Analysis of the Effect of Tanezumab on Overall Daily Pain Score Data in Adults with Moderate-to-Severe Pain due to Osteoarthritis of the Knee
Rujia Xie1, Rosalin Arends2, Stephen Olson2 , and Scott Marshall1
1 Pfizer, Sandwich, UK; 2 Pfizer, New London, USA
Objectives: The primary aim of this work was to characterize the exposure response (overall daily pain score, DPS) relationship across time for tanezumab a MAb for treatment of pain.
Methods: Four hundred and forty-four osteoarthritis patients were randomized to one of the following dose groups: 0, 10, 25, 50,100 and 200 mg/kg. Two doses of tanezumab or placebo were administered as a 10-min intravenous infusion 56 days apart. PK and DPS (VAS scale 0-100) data was collected for ~200 days. The PK/PD model was built according to the following steps: i) population PK model was developed, ii) individual PK parameter estimates were fixed during the PK/PD model development. The final PK/PD model was then used for simulations conducted to explore dose strategy and dose regimen.
Results: The PK of tanezumab was well described by a two-compartmental model. Body weight was found to be a significant covariate on clearance and volume but this only explained 4% of the in total 47% unexplained inter-subject variability [1].
In the PK/PD model it was assumed that the placebo and drug effects were proportional to the baseline. Placebo effect was well described by an exponential time-dependent model. The onset of placebo effect was relatively fast with an equilibrium half-life of 7.7 days (RSE 23%) after the first dose. The maximum placebo effect was estimated to be 25.2% (RSE 16%) of the baseline value.
An indirect response model was found to best characterize the delay between the response and the tanezumab concentrations. It was assumed that tanezumab inhibits the production of a pain stimulus (i.e. Nerve Growth Factor) measured by the DPS. The drug effect was characterized by an inhibitory Emax model, which was expressed as a maximal inhibiting effect (Imax) and the tanezumab concentration (IC50) to achieve half of Imax. The Imax and IC50 were estimated to be 0.538 (RSE 9.4%) and 69.3 (RSE 48%) ng/ml, respectively.
The apparent transitory dose related attenuation of DPS approximately 14 days following the first dose was well captured by an empirical model expressed by a modified Gamma distribution function. It was assumed that the reduction of Imax is related to dose described by an Emax model.
Conclusions: The DPS data was adequately described by the proposed semi-mechanistic PK/PD model. Subsequent simulations using the PK/PD model supported moving forward in tanezumab Phase 3 in osteoarthritis patients with 8 week dosage regimen at fixed doses of 2.5, 5 and 10 mg.
References:
[1]. Rosalin Arends et.al. Population PK Modeling to Support the Use of a Fixed Dosing Regimen in Phase 3 for Tanezumab, an anti-NGF humanized antibody. AAPS NBC, 2009
