Sebastian Weber, Michael Looby, Thomas Dumortier
Advanced Quantitative Sciences, Novartis Pharma
Objectives: The immunosuppresive drug tacrolimus (TAC) is key to prevent organ rejection in patients post liver transplantation. However, a severe side effect of TAC is nephrotoxicity. Clinically it is known that the impairment of the renal function is both acute and chronic [1,2]. The main objectives of this work are to establish (i) a model describing kidney degradation and (ii) evaluate acute (reversible) and chronic (irreversible) progression.
Methods: From the study [3] the TAC reference treatment arm with N=240 patients were included in the analysis data set. Their TAC trough concentrations were monitored over a period of up to 2 years post transplantation. Due to the high variability of TAC trough concentrations, bid oral dosing was continuously monitored and adapted by physicians based on signs of acute rejection (AR). The target regimen was 8-12 ng/mL for the first 4 months and was reduced to 6-10 ng/mL thereafter. Glomerular filtration rate (GFR) based on creatinine clearance was used as a surrogate for kidney function. The PK/PD model was developed using FOCE interaction in NONMEM 7.2.
Results: The PK was adequately described by a one-compartment model (similar to [4]) with linear elimination and included a dose-adjusted relative bioavailability. The link between the PD and GFR was modeled with an inhibitive Hill function. The Hill function mediated suitably parametrized a direct acute effect of TAC PK (E1), a time-delayed acute TAC effect via an effect compartment (E2) and a chronic GFR impairment by the total accumulated exposure (E3). The model including E2 and E3 performed best in terms of the objective function value (OFV) to describe the data when tested by a LRT against other models involving a combination of E1/E2/E3. While a contribution of a chronic kidney damage was inferred with a high precision, the increase in time of the chronic effect in relation to the acute kidney impairement showed a large uncertainty (RSE of 140%).
Conclusions: We have implemented a PK/PD model, which describes the time-course of TAC on GFR. Despite the extensive noise present in these measurements, an acute and a chronic contribution to nephrotoxic kidney degradation was observable on a time scale of 10 months – although the conversion rate of acute to chronic kidney damage was only obtainable from the data with a large uncertainty. While the large uncertainties can limit the applicability of the model, the established PK/PD model will be valuable to plan future studies with TAC treatment as reference therapy.
[1] Naesens, M. et al., CJASN, 4 (2), p. 481, 2009
[2] Staatz, C. E. and Tett, S. E., Clin Pharmacokinet, 43 (10), p. 623, 2004
[3] De Simone et al., Am J Transplant. 12(11), 3008, 2012
[4] Antignac, M. et al., Eur J Clin Pharmacol, 64, p. 409, 2005
Reference: PAGE 22 () Abstr 2857 [www.page-meeting.org/?abstract=2857]
Poster: Other Drug/Disease Modelling