Development of a closed loop whole body (WB) physiologically based pharmacokinetic model (PBPK) of beta-blockers in the rat
S.Y. Amy Cheung (1), Malcolm Rowland (2), Leon Aarons (2), Trudy Rodgers (1), Iva Gueorguieva (3)
(1)Centre for Applied Pharmacokinetic Research,(2)School of Pharmacy and Pharmaceutical Sciences, The University of Manchester, Oxford Road, Manchester, M13 9PL, United Kingdom,(3)Eli Lilly & Co,Erl Wood Manor, Sunninghill Road,GU20 6PH,Windlesham,England
Objective: To develop a closed loop (CDL) WBPBPK model to investigate the in vivo pharmacokinetics of some beta-blockers from rat data. Some tissues exhibit slow equilibration with blood levels which suggests permeability limited (PML) kinetics. Therefore tissue models representing this process must be developed.
Method: 12 tissues (adipose, bone, brain, gut, heart, kidney, liver, lung, muscle, skin, testes & thymus) and arterial blood of rats for 15 beta-blocker compounds (R- and S- acebutolol, betaxolol, bisoprolol, metoprolol, oxprenolol, propranolol, pindolol & S- timolol) were previously collected. Open loop (OPL) modelling [1], also called the forcing function approach [2], allows the estimation of partition coefficients (Kp) in a PBPK model for individual well-stirred tissues by using the known arterial concentration-time profile as a driving function together with the physiological information of tissue volumes and tissue blood flow rate, assuming perfusion limited (PFL) kinetics. This results in a 1 compartment tissue model. For certain tissues, PML kinetics is modelled by 2 compartments with the rate of exchange between vasculature and tissue controlled by permeability (PST). These findings necessitated the incorporation of certain PML tissues into the closed loop WBPBPK model of each compound. The estimated Kp values with its %SE from the OPL modelling was then used together with the prior function subroutine [3] in NONMEM to develop a WBPBPK model. In the CDL modelling, all tissue Kps of a compound are estimated simultaneously.
Results: The Kp estimates for S-acebutalol are compared to the steady state (SS) values in the table below as an example. The CDL PML Kp estimates more closely reflect the SS estimates than the CDL PFL estimates. The difference in objective function (OBJ FN) values indicates the model fit of the permeability incorporated WBPBPK model is also better than the perfusion limited model.
|
Tissues |
SS |
CDL PFL |
CDL PML |
|
Lung |
6.07 |
5.81 |
5.31 |
|
Gut |
90.4 |
0.127 |
150 |
|
Gut PST |
NA |
NA |
1.78 |
|
Thymus |
4.35 |
2.94 |
2.83 |
|
Liver |
24.6 |
1.72 |
9.78 |
|
Brain |
0.356 |
0.202 |
0.165 |
|
Heart |
4.29 |
7.10 |
6.89 |
|
Kidneys |
31.4 |
13.2 |
12.8 |
|
Skin |
2.45 |
8.50 |
7.69 |
|
Muscle |
4.06 |
5.53 |
4.81 |
|
Adipose |
0.772 |
0.553 |
0.470 |
|
Testis |
2.50 |
3.25 |
3.39 |
|
Testis PST |
NA |
NA |
0.0513 |
|
Bone |
0.0436 |
0.422 |
0.0373 |
|
OBJ FN |
NA |
1240 |
461 |
Conclusion: These results show the importance of using the information obtained in OPL modelling in the development of a WBPBPK model using the prior function subroutine in NONMEM. This also demonstrates that PML kinetics must be incorporated into the model.
Reference:
1. Gueorguieva, I., et al. (2004) J. Pharmacokinet. Pharmacodyn. 31(4) 269 - 297
2. Foster, D. M., (1998) Adv. Exp. Med. Biol. 445: 59-78
3. Langdon, G., (2007) Eur. J. Clin. Pharmacol. 63: 485-498
