pursuing the holy grail of predicting and verifying tissue ... · atp adp + pi 2 failure rate and...
TRANSCRIPT
Jashvant (Jash) Unadkat Milo Gibaldi Endowed Professor
Dept. of Pharmaceutics School of Pharmacy
University of Washington Seattle, WA
Pursuing the holy grail of predicting and verifying tissue drug concentrations: A proteomics and PET
imaging approach
https://sop.washington.edu/people/jashvant-unadkat/
ATP ADP + Pi
2
Failure Rate and Reasons for Failure in Drug Development
Smietana et al., Nature Reviews Drug Discovery 15, 379–380 (2016) Harrison, Nature Reviews Drug Discovery 15, 817–818 (2016) SSX
ATP ADP + Pi
One Possible Reason for Lack of Drug Efficacy & Safety
• Unable to measure or predict tissue conc. of drugs
• Unbound plasma conc. ≠ unbound tissue conc. if transporters are involved, i.e. asymmetry between blood and tissue drug conc. (e.g. liver:blood due to OATPs)
3 SSX
ATP ADP + Pi
4 Eyal, Hsiao & Unadkat Pharmacol. Ther., 2009
Asymmetry In Brain:Blood Drug Conc. Due to Blood-Brain Barrier Efflux Transporters
SSX
5
P-gp at the Rat BBB Results in Asymmetry in Brain:Blood Conc. of Verapamil (P-gp substrate)
0%
500%
1000%
1500%
0 5 10 15 20Blood CsA concentraion (μM)
Human PET study
Mouse KO study
Perce
nt inc
rease
in bra
in: bl
ood
[3 H]-ra
dioac
tivity
Hsiao et al J Pharmacol Exp Ther. 2006 SSX
6 Eyal et al., Clin Pharmacol Ther. 2010
Time (min)
0 10 20 30 40 50
Brain
11C-
radio
activ
ity / D
ose (
L-1)
0.01
↑~2-fold
Time (min)
0 10 20 30 40 50
Plas
ma 11
C-rad
ioacti
vity /
Dos
e (L-1
)
0.01
0.1Plasma
Brain
- CsA + CsA
+ CsA -CsA
Asymmetry in Drug Conc. at the Human Brain:Blood Barrier: P-gp Efflux of 11C-Verapamil
11C-verapamil AUCbrain:AUCblood (20 min) - 0.42±0.04 SSX
ATP ADP + Pi
Asymmetry in Tissue:Blood Drug Conc. - Liver
7
CLbile CLsin
CLsef
Bile Blood
[D] [D]
CLmet
e.g. OATPs
e.g. P-gp, BCRP, MRP2
SSX
ATP ADP + Pi
Asymmetry in Liver Tissue:Plasma Conc. when Clsin is the Rate-Determining Step
i.e. CLmet+bile >> CLsef
0
0.5
1
1.5
2
0 10 20 30 40 50 60
Con
cent
ratio
n (m
g/L
)
Time (min)
0
0.2
0.4
0.6
0 10 20 30 40 50 60
Con
cent
ratio
n (m
g/L
)
Time (min)
CLins CLefs
CLmet CLbile
CLins CLefs
CLmet CLbile
0
0.5
1
1.5
2
0 10 20 30 40 50 60
Con
cent
ratio
n (m
g/L
)
Time (min)
0
0.2
0.4
0.6
0 10 20 30 40 50 60
Con
cent
ratio
n (m
g/L
)
Time (min)
CLins CLefs
CLmet CLbile
0
0.5
1
1.5
2
0 10 20 30 40 50 60
Con
cent
ratio
n (m
g/L
)
Time (min)
0
1
2
3
0 10 20 30 40 50 60
Con
cent
ratio
n (m
g/L
)
Time (min)
Inhibition of sinusoidal uptake CL
Inhibition of metabolism and biliary efflux CL
No inhibition
Assumptions: > Liver is the only
eliminating organ > 90% inhibition
8
Liver
Blood
Patilea-Vrana & Unadkat. CPT. 2016
↑
↔
↔
↑
ATP ADP + Pi
Asymmetry in Hepatic:Blood Conc. of 11C-Rosuvastatin in the Rat
Coronal 2 min SUV images of 11C-Rosuvastatin
He et al., Mol Pharm., ‘14
Bile Blood
ROS
Hepatocyte
OATPs BCRP/MRP2
-RIF +RIF
9 SSX
Hepatic Uptake and Biliary Excretion of 11C-Rosuvastatin in the Rat
He et al., Mol Pharm., ‘14 10 SSX
Changes in Rat Liver Exposure to Rosuvastatin ± Rifampin
11 He et al., Mol Pharm. 2014
2.3 fold increase in RSV plasma AUC but NO significant increase in RSV liver AUC
Rat Blood Rat Liver
% o
f Dos
e
% o
f Dos
e
Time (min)
Time (min)
OATPs
Bile Blood
ROS
Hepatocyte
BCRP/MRP2
SSX
ATP ADP + Pi
12
Biodistribution of 11C-rosuvastatin in humans
Billington et al., In progress SSX
How Can we Predict Tissue Drug Conc. in Humans?
• PET imaging (MRI and other imaging modalities do not have the required sensitivity): – Requires sophisticated equipment and
radiochemistry – Costly (about $20-40K/experiment/subject)
• Therefore we need alternative methods that will allow us to predict tissue conc. of drugs in humans 13
APPLICATION OF PROTEOMICS DATA TO PREDICT PK AND TISSUE CONC. OF DRUGS
14
ATP ADP + PiATP ADP + Pi
Hypothesis: Predict transporter-mediated in-vivo CL and tissue concentration of drugs in humans from in-vitro studies
15
Predict in-vivo CL and tissue conc. using transporter scaling factor
1. transporter-mediated drug CL
2. transporter abundance using quantitative proteomics
3. Obtain transporter scaling factor
Transporter-expressing cells
Verify predictions using PET imaging
ATP ADP + Pi
Hepatic Uptake and Biliary Excretion of 11C-Rosuvastatin in the Rat
Boston 16
Coronal 2 min SUV images of 11C-Rosuvastatin
He et al., Mol Pharm., ‘14
Bile Blood
ROS
Hepatocyte
OATPs BCRP/MRP2
-RIF +RIF
ATP ADP + Pi
0
20
40
60
80
Clea
ranc
e (m
L/m
in/k
g bo
dy w
eigh
t)
Previous PET imagingstudy
SCRH not corrected forOatp expression
p<0.05
CLs,uptake CLs,efflux CLbile
Successful prediction of the hepatobiliary clearance of rosuvastatin using cell lines, sandwich-cultured rat hepatocytes and quantitative proteomics
CLs,uptake; sinusoidal uptake CLs,efflux; sinusoidal efflux CLbile; canalicular efflux
0
2
4
6
8
10
Oat
p ex
pres
sion
(fmol
/µg
mem
bran
e pr
otei
n)
SD rat liverSCRH
p<0.05
p<0.05 p<0.05
Ishida et al., DMD, 2018 17
0
20
40
60
80
100
120
CLs,
upta
ke(m
L/m
in/k
g bo
dy w
eigh
t)
ATP ADP + Pi
Rat Hepatic Rosuvastatin Conc. well Predicted
18 Ishida et al., DMD 2018
0
500
1000
1500
2000
2500
3000
3500
0 3 6 9 12 15
Hepa
tic c
once
ntra
tion
(kBq
/mL)
Time (min)
Observed valuepredicted value95% CI of observed data
Can Rosuvastatin Hepatobiliary CL and Hepatic Conc. be Predicted in Humans?
19
ATP ADP + PiATP ADP + Pi
Total transporter abundance in suspended (SH), plated (PH), sandwich-cultured (SCH) hepatocytes and liver tissue
20
Kumar et al., Unpublished data
Data removed as not published
ATP ADP + PiATP ADP + Pi
Plasma membrane transporter abundance in suspended (SH), plated (PH), sandwich-cultured (SCH) hepatocytes cf liver tissue
Kumar et al., Unpublished data
21
Data removed as not published
ATP ADP + PiATP ADP + Pi
22
Transporter-expressing cells better predict in-vivo (IV ) human rosuvastatin hepatic uptake clearance than hepatocytes
RSV uptake CL scaled on the basis of hepatocellularity *(Classical method)
*88 mg protein/g human liver
*Assuming sinusoidal uptake is the rate determining step in
RSV plasma CL
Patilea-Vrana G and Unadkat JD, Clin Pharmacol Ther, 2016
Martin et al., Clin Ther. 2003 Oct;25(10):2553-63. Karlgren et al., J Med Chem. 2012 May 24;55(10):4740-63.
Data removed as not published
ATP ADP + Pi
Metformin renal clearance is reasonably well-predicted using OCT2 expressing
cells
Kumar et al., DMD 2018 23
Observed metformin renal secretory clearance in humans: 432 (range 215-643) mL/min
0
100
200
300
400
500
600
700
Total abundance PM abundance PM abundanceand PM potential
Met
form
in re
nal s
ecre
tory
cl
eara
nce
(mL/
min
)
215
643
432
ATP ADP + Pi
• Predicting tissue concentration and therefore efficacy and toxicity of a drug is the next frontier in ADME research
• The hepatic ECL model clarifies when transporters will or will not affect the systemic and tissue PK of a drug
• Tissue conc. measurement is possible using PET. However, this method cannot be routinely applied
• IVIVE using transfected cells and quantitative transporter proteomics is a promising technique to predict tissue drug conc.
• These predictions should be validated using PET imaging probes that interrogate multiple drug transporters
24
Summary
ATP ADP + Pi
Bhagwat Prasad Anand Deo 25
YuYang Jiake He
Major Contributors
Sarah Billington Li Wang Gabriela Patilea-Vrana
Kazuya Ishida Vineet Kumar SSX
ATP ADP + Pi
26
Genentech, Merck, Biogen, Gilead, BMS, Takeda, Pfizer
SSX
ATP ADP + PiOther Collaborators
Dept. of Radiology: Jeanne Link, David Mankoff, Todd Richards, Janet Eary, Satoshi Minoshima, Ken Maravilla, Mark Muzi, Steve Shoner, David Lewis, Jean Lee and the PET suite team Dept. of Medicine: Ann Collier and her team; Scott Lee and his team Dept. of Anesthesiology: Karen Domino, Matthew Pennington Dept. of Pharmaceutics: Bhagwat Prasad, Edward Kelly, Carol Collins, Joanne Wang Kidney Research Institute: Jonathan Himmelfarb Univ. of Greifswald: Stefan Oswald and team Children’s Mercy Hospitals: Steven Leeder and team Acknowledgement: NIH P01DA032507, MH63641, P50 HD44404, RR 00166, HD47892, AG031485, RC1NS068904, Solvo Biotech., UWRAPT funded by Genentech, Merck, Biogen, Gilead, BMS, Takeda, Pfizer
27 SSX
ATP ADP + Pi
SSX 28