Discovery of novel hepatoselective HMG-CoA reductase inhibitors for treating hypercholesterolemia: A bench-to-bedside case study on tissue selective drug distribution

Article abstract of DOI:10.1016/j.bmcl.2010.11.103

The design of drugs with selective tissue distribution can be an effective strategy for enhancing efficacy and safety, but understanding the translation of preclinical tissue distribution data to the clinic remains an important challenge. As part of a discovery program to identify next generation liver selective HMG-CoA reductase inhibitors we report the identification of (3R,5R)-7-(4-((3-fluorobenzyl)carbamoyl)-5-cyclopropyl-2-(4-fluorophenyl) -1H-imidazol-1-yl)-3,5-dihydroxyheptanoic acid (26) as a candidate for treating hypercholesterlemia. Clinical evaluation of 26 (PF-03491165), as well as the previously reported 2 (PF-03052334), provided an opportunity for a case study comparison of the preclinical and clinical pharmacokinetics as well as pharmacodynamics of tissue targeted HMG-CoA reductase inhibitors.

Full text of DOI:10.1016/j.bmcl.2010.11.103

Bioorganic & Medicinal Chemistry Letters 21 (2011) 2725–2731
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of novel hepatoselective HMG-CoA reductase inhibitors
for treating hypercholesterolemia: A bench-to-bedside case study on tissue
selective drug distribution
⇑
Jeffrey A. Pfefferkorn , John Litchfield, Richard Hutchings, Xue-Min Cheng, Scott D. Larsen, Bruce Auerbach,
Mark R. Bush, Chitase Lee, Noe Erasga, Daniel M. Bowles, David C. Boyles, Gina Lu, Catherine Sekerke,
Valerie Askew, Jeffrey C. Hanselman, Lisa Dillon, Zhiwu Lin, Andrew Robertson, Karl Olsen, Carine Boustany,
Karen Atkinson, Theunis C. Goosen, Vaishali Sahasrabudhe, Jonathan Chupka, David B. Duignan, Bo Feng,
Renato Scialis, Emi Kimoto, Yi-An Bi, Yurong Lai, Ayman El-Kattan, Rebecca Bakker-Arkema, Paul Barclay,
Erick Kindt, Vu Le, Jaap W. Mandema, Mark Milad, Bradley D. Tait, Robert Kennedy, Bharat K. Trivedi,
Mark Kowala
Pfizer Global Research & Development, Eastern Point Road, Groton, CT 06340, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
The design of drugs with selective tissue distribution can be an effective strategy for enhancing efficacy
and safety, but understanding the translation of preclinical tissue distribution data to the clinic remains
an important challenge. As part of a discovery program to identify next generation liver selective HMG-
CoA reductase inhibitors we report the identification of (3R,5R)-7-(4-((3-fluorobenzyl)carbamoyl)-5-
cyclopropyl-2-(4-fluorophenyl)-1H-imidazol-1-yl)-3,5-dihydroxyheptanoic acid (26) as a candidate for
treating hypercholesterlemia. Clinical evaluation of 26 (PF-03491165), as well as the previously reported
2 (PF-03052334), provided an opportunity for a case study comparison of the preclinical and clinical
pharmacokinetics as well as pharmacodynamics of tissue targeted HMG-CoA reductase inhibitors.
Ó 2010 Elsevier Ltd. All rights reserved.
Received 8 October 2010
Revised 17 November 2010
Accepted 22 November 2010
Available online 26 November 2010
Keywords:
HMG-CoA reductase inhibitor
Statin
Hypercholesterolemia
Liver selective
Hepatoselective
The tissue distribution of a drug can have significant impact on
both its efficacy and safety and as a result there is interest in the
development of small molecule therapeutics with selective tissue
distribution profiles.1 One area of particular focus is the design of
hepatically-targeted agents given the liver’s key role in metabolic
regulation and the fact that it is the principal tissue affected by dis-
eases such as hepatitis B and C viruses as well as hepatocellular
carcinoma. Liver-targeted drug discovery approaches have been
previously reported for diverse mechanistic classes such as HMG-
CoA reductase inhibitors (i.e., statins),2 glucocorticoid antagonists,3
thyroid hormone receptor agonists,4 glucokinase activators,5
stearoyl-CoA desaturase inhibitors,6 hepatitis B and C antivirals,7
caspase inhibitors8 and nucleoside oncolytics.9 Among these exam-
ples, several strategies have been employed to achieve hepatose-
lectivity, including: (a) optimizing molecules for recognition and
active uptake via liver specific transporters such as the organic
anion transporting polypeptide (OATP) transporters and organic
cation transporters (OCT); (b) synthesizing conjugates with liver
targeting motifs like bile acids or statins; and (c) designing
prodrugs that undergo liver specific metabolic activation.2–4,7,9,10
A common issue when using either active transport processes or
metabolic activation of prodrugs to achieve selective liver distribu-
tion is understanding the value and confidence of preclinical
in vitro and in vivo models for predicting human drug distribution.
In this report we describe a case study on the discovery of novel
hepatoselective HMG-CoA reductase inhibitors, which provided
an opportunity to compare preclinical transporter and drug distri-
bution data with observed clinical pharmacokinetic and pharmaco-
dynamic data.
Next generation HMG-CoA reductase inhibitors with enhanced
hepatoselectivity relative to current statins offer promise for help-
ing patients achieve increasingly aggressive LDL-reduction goals
while minimizing the risk of myalgia, the muscle pain and weak-
ness sometimes associated with statin therapy.2 While the overall
incidence of statin-induced myalgia is rather low (2–7% of
patients), the probability of occurrence increases with drug dose,
and it can be a key factor in reducing patient compliance with
⇑
Corresponding author. Tel.: +1 860 686 3421; fax: +1 860 715 4608.
E-mail address: jeffrey.a.pfefferkorn@pfizer.com (J.A. Pfefferkorn).
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.11.103

2726
J. A. Pfefferkorn et al. / Bioorg. Med. Chem. Lett. 21 (2011) 2725–2731
in Figure 1. As previously reported, these efforts led to the discov-
O
OH OH
ery of pyrazole-based PF-03052334 (2) as a HMG-CoA inhibitor
candidate.2e To further improve upon the hepatoselectivity of 2,
we now report an extension of this strategy to imidazole-based
HMG-CoA reductase inhibitors (i.e., 3, Fig. 1). The synthesis and
structure–activity studies of this imidazole-based series leading
to clinical candidate selection will be described. Subsequently,
we examine the preclinical transporter interactions, hepatocyte
uptake, tissue distribution and biomarker response of this candi-
date as well as PF-03052334 (2) relative to their observed clinical
data.
Scheme 1 illustrates the general synthesis of the inhibitors eval-
uated in these studies. A complete description of this synthesis has
also been recently reported.14 Briefly, condensation of glycine
benzyl ester hydrochloride (4, Scheme 1) with benzophenone
imine afforded diphenyl imine 5 in high yield. Acylation of glycine
diphenyl imine 5 with various acid chlorides (R1COCl) at low
temperature provided amine 6, which upon treatment with 4-flu-
orobenzoyl chloride underwent conversion to keto amide 7. Treat-
ment of keto amide 7 with amine 815 at elevated temperature in
the presence of AcOH and catalytic TsOH resulted in cyclization
to form imidazole 9. Subsequent hydrogenolysis of 9 to the corre-
sponding carboxylic acid 10 and CDI-mediated coupling with
amines (R2NH2) provided amide 11. Deprotection of the acetonide
of 11 with TFA or HCl (aq) in MeOH followed by hydrolysis with
NaOH provided the final product 13 as a sodium salt. The route
outlined in Scheme 1 was utilized, with appropriate substitution
of alternative building blocks, to prepare the analogs described in
Table 1.
CO2H
N
H
N
1: atorvastatin
F
O
O
OH OH
CO2H
N
H
N
N
2: PF-03052334
F
R1
OH OH
CO2H
N
H
N
R2
N
3: Imidazole Template
F
Figure 1. Structures of HMG-CoA reductase inhibitors.
aggressive cholesterol treatment regimens.11 Statin-induced myal-
gia is thought to involve, in part, inhibition of HMG-CoA reductase
in non-hepatic tissues (particularly muscle) and accordingly it is
hypothesized that its occurrence can be reduced by selectively tar-
geting HMG-CoA reductase inhibitors to hepatic tissues and limit-
ing peripheral exposure.11 Moreover, it has been shown that this
statin hepatoselectivity can be achieved by reducing inhibitor lipo-
philicity to minimize passive permeability into peripheral tissues
while concurrently optimizing the molecules as substrates for ac-
tive transport into hepatocytes.12 In particular the OATP family of
membrane transporters has been reported to be important for
the selective liver distribution of statins as evidenced, in part, by
the fact that patients with genetic variants in such transporters
are at increased risk of statin-induced myalgia.13
During structure–activity studies, compounds were initially
screened in an HMG-CoA reductase enzyme inhibition assay from
a rat microsomal preparation.16 Potent analogs were then evalu-
ated for inhibition of cholesterol synthesis in both rat hepatocyte
and rat myocyte cell lines with the ratio of these values (myocyte
IC50/hepatocyte IC50
) utilized as an assessment of cellular
hepatoselectivity. Compounds were then evaluated in an in vivo
hamster model measuring acute inhibition of cholesterol synthe-
sis. In these experiments, animals were dosed po with 10 mg/kg
inhibitor at t = 0 h then given an intraperitoneal injection of 14C
sodium acetate at t = 2 h and blood samples were obtained at
t = 4 h and analyzed for 14C cholesterol levels. These cholesterol
levels were then compared to untreated control animals to deter-
mine percent inhibition of acute cholesterol synthesis. In several
cases, in order to obtain a further assessment of a compound’s
As part of a discovery program to identify novel statins with
best-in-class tolerability and efficacy through increased hepatose-
lectivity, we have investigated replacement of the pyrrole core of
atorvastatin (1) with alternative five-membered heterocycles that
offer reduced substitution and inherent lipophilicity as illustrated
O
R1
Ph
Ph
OBn
a
b
c
HClH2N
OBn
OBn
Ph
N
5
Ph
N
6
O
N
4
O
O
O
O
8
CO2tBu
H2N
R1
R1
O
O
O
OH OH
O
O
O
R1
O
CO2tBu
O
R2
N
X
Z
N
H
OBn
N
N
g
8, d
N
H
F
7
F
9 X = OBn
F
11: Z = OtBu
e
f
h
13: Z = O-Na+
10: X = OH
11: X = NHR2
Scheme 1. General synthesis of substituted imidazoles as HMG-CoA reductase inhibitors. Reagents and conditions: (a) benzophenone imine, CH2Cl2, 20 °C, 8 h, 99%; (b) (i)
KOtBu, R1COCl, THF, ꢀ78 °C; (ii) HCl (aq); (c) 4-fluorobenzoyl chloride, Et3N, CH2Cl2, 0 °C, 0.5 h; (d) 8, xylene, AcOH, TsOH (cat), reflux, 16 h; (e) 10% Pd/C, H2, MeOH, 25 °C, 4 h;
(f) R2NH2, CDI, DMF, 0–20 °C, 2 h; (g) TFA, CH2Cl2, 0 °C, 0.5 h or HCl (aq), MeOH, 45 °C, 12 h; (h) NaOH (aq), THF, 50 °C, 2 h.

J. A. Pfefferkorn et al. / Bioorg. Med. Chem. Lett. 21 (2011) 2725–2731
2727
Table 1
Structure and biological activity of substituted imidazole analogs 14–33
R1
OH OH
O
O
R4
O-Na+
N
N
H
R3
N
R2
F
R1
R2
R3
R4
HMG-CoAa IC50 (nM)
Cellular inhibition of cholesterol synthesisb
In vivo inhibition of cholesterol
synthesisb
Log Dc
(pH 7.4)
Hepatocycte
IC50 (nM)
Myocycte
IC50 (nM)
Cell
Selectivity
(10 mg/kg)
t = 2–4 h
t = 3–5 h
2
1.0
3.1
94
123
26
7.9
0.9
0.3
—
—
—
730
250
—
—
—
810
830
—
—
—
ꢀ74%
ꢀ76%
—
—
—
—
—
—
—
0.24
1.01
—
—
—
Rosuvastatin
14
15
16
17
18
Me
Et
nPr
iPr
cPr
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
—
0.3
0.2
3030
632
10,100
3160
ꢀ53%
ꢀ7%
0.64
0.03
1.4
ꢀ58%
—
19
20
21
cPr
cPr
cPr
Me
H
H
H
Me
H
H
H
Me
1.7
7.2
4.9
0.1
0.3
0.1
178
532
4850
1780
1770
48,500
ꢀ79%
ꢀ54%
—
ꢀ36%
ꢀ50%
0%
0.45
0.47
0.44
22
23
24
cPr
cPr
cPr
OMe
H
H
H
OMe
H
H
H
OMe
1.2
1.2
7.0
0.2
0.2
0.1
6400
538
10,700
32,000
2960
107,000
ꢀ60%
ꢀ53%
—
—
0.02
0.11
0.16
ꢀ42%
—
25
26
27
28
29
cPr
cPr
cPr
cPr
cPr
F
H
F
H
F
H
H
F
H
F
3.4
1.2
1.5
1.2
1.2
0.7
0.1
ND
0.5
1620
444
27,200
2780
4870
2310
4440
—
5560
24,350
—
0%
0.17
0.51
0.08
—
H
H
F
ꢀ68%
ꢀ43%
ꢀ4%
ꢀ8%
—
—
—
ꢀ46%
F
H
0.2
—
30
31
cPr
cPr
Me
OMe
F
F
H
H
4.1
3.7
0.1
0.2
224
890
2240
4450
—
ꢀ36%
—
0%
0.69
—
32
33
cPr
cPr
C(O)NH2
SO2NMe2
H
H
H
H
7.2
4.9
0.4
0.5
>100,000
>100,000
>250,000
>200,000
—
—
0%
0%
ꢀ1.1
ꢀ0.46
a
b
c
In vitro assay values reported as the mean SEM of n P2 independent determinations.
Data represent the mean of the percent change from the control group (n = 8/group).
Log D determined by shake flask method.
duration of action, the experiment was repeated with later
timepoints in separate animal groups with drug dosing at t = 0 h,
14C sodium acetate injection at t = 3 h and cholesterol level determi-
nation at t = 5 h.
optimal with substitution in either the para or meta position
(19 > 20 > 21). Third, efforts to combine optimal substituents
(4-Me, 4-OMe, 3-F) from the monosubstituted analogs into disub-
stituted analogs such as 30 and 31 generally did not afford additive
potency or efficacy. Finally, as shown for 32 and 33 introduction of
increasingly polar substituents resulted in analogs with signifi-
cantly increased cellular hepatoselectivity but no in vivo efficacy
presumably due to the very poor passive membrane permeability
of these analogs as suggested by their significantly lower measured
Log D values.
Among the imidazole analogs evaluated, compound 26 was
selected for further characterization based on its favorable balance
of hepatoselectivity and in vivo efficacy. Specifically, the cellular
hepatoselectivity (4440ꢂ) of 26 offered a fivefold improvement
relative to either rosuvastatin or 2 and the compound had compa-
rable acute in vivo efficacy for inhibition of cholesterol synthesis.
To further characterize its efficacy, 26 was then evaluated in a
7-day LDL lowering dose–response study conducted in chow-fed
guinea pigs as shown in Figure 2.16 For comparison, rosuvastatin
and pravastatin were included as benchmarks representing the
current standard of care statins with the highest levels of hepa-
toselectivity. As illustrated, the ED50 of 26 was 5.9 mg/kg with an
Emax similar to rosuvastatin, but less than that of compound 2.
The comparative pharmacokinetics and tissue distribution
properties of 2 and 26 were next characterized in preclinical spe-
cies. Plasma and tissue protein binding properties were character-
ized as shown in Table 2 wherein compound 2 was found to be
approximately 10-fold more protein bound than 26.
As shown in Table 1, initial structure–activity efforts focused on
evaluating the effect of the R1 imidazole substituent. As shown for
compounds 14–18, inhibitory potency increased with increasing
substituent size with cPr > iPr > nPr > Et ꢁ Me. Among these ana-
logs, both 17 (R1 = iPr) and 18 (R1 = cPr) offered sufficient in vitro
potency to warrant selectivity and efficacy testing. Encouragingly,
the cellular hepatoselectivity ratios for 17 (10,100ꢂ) and 18
(3160ꢂ) were significantly greater than that of 2 (810ꢂ) or
rosuvastatin (830ꢂ); however, on the downside both 17 and 18
demonstrated reduced acute in vivo efficacy relative to these same
benchmarks.
We next examined the effects of substituents on the benzyl
amide with the goal of increasing efficacy while maintaining cellu-
lar hepatoselectivity. Given the moderate improvement in vitro
potency offered by 18 over 17, these studies were conducted in
the R1 = cPr series. As highlighted in Table 1, several trends were
observed during these studies. First, introduction of a 4-Me (19),
3-Me (20), 3-OMe (23) or 3-F (3) offered optimal effects on magni-
tude and duration of in vivo inhibition of cholesterol synthesis
relative to the unsubstituted case (i.e., 18) while still maintaining
the hepatoselectivity. Second, for cases where a complete ortho,
meta, and para substitution pattern was explored (e.g., 19, 20,
21) cellular hepatoselectivity tended to be optimal with ortho sub-
stitution (21 > 19, 20) but, in contrast, in vivo efficacy tended to be

2728
J. A. Pfefferkorn et al. / Bioorg. Med. Chem. Lett. 21 (2011) 2725–2731
20
0
be 0.04- and 1.7-fold of the parent acid concentration in dog and
rat, respectively.
To better understand the underlying transport mechanism
responsible for the distribution of these HMG-CoA reductase inhib-
itors we examined their uptake in sandwich-cultured rat and hu-
man hepatocyctes as summarized Tables 4 and 5, respectively. In
rat hepatocyctes, the active uptake rates were similar for the three
molecules and uptake was inhibited by rifamycin SV as a pan-OATP
inhibitor. In human hepatocyctes, the active uptake rates were
similar for 2 and rosuvastatin but somewhat lower for 26. Similar
to rat hepatocytes, the active uptake of all three compounds was
significantly inhibited in the presence of rifamycin SV. To further
understand the mechanism of hepatic active uptake, the three
HMG-CoA reductase inhibitors were then evaluated as a potential
substrate for rat (Table 4) and human (Table 5) hepatic OATP trans-
porters. In these assays, individual rat (rOatp1a1, rOatp1a4 and
rOatp1b2) and human transporters (hOATP1B1, hOATP1B3 and
hOATP2B1) were over expressed in CHO-cells or HEK293 cells
enabling determination of the fold difference in substrate uptake
relative to the wild type cell line for each transporter.18 Rosuvast-
atin/pravastatin and midazolam served as positive and negative
controls, respectively, in each assay. Both 26 and 2 were found to
be substrates for rOatp1a1, rOatp1a4 and rOatp1b2 transporters
as well as human hOATP1B1 and hOATP1B3. Compound 26 also
appeared to be a weak substrate for hOATP2B1. Qualitatively, 2
and 26 were found to undergo similar uptake via the rat transport-
ers, but 26 was suggested to be a stronger substrate for the corre-
sponding human transporters. Moreover, in the human hepatocyte
uptake assay (Table 5, left side) uptake of 26 was found to be 97%
active versus 78% active for 2, further suggesting that 26 was
potentially a better substrate for human hepatic uptake.
Having determined 2 and 26 to be in vitro substrates for hepatic
transporters, we next evaluated their in vivo liver distribution and
pharmacokinetic/pharmacodynamic (PK/PD) profiles. Oral dosing
of 2, 26 and rosuvastatin to rats followed by sacrifice and determi-
nation of free plasma and liver drug concentration at regular inter-
vals (t = 0, 0.5, 1, 2, 3, 4, 7 and 24 h) enabled calculation of liver/
plasma AUC(0–24h) ratios. As summarized in Figure 3, all three
exhibited enhanced liver-to-plasma drug concentrations. The
liver-to-plasma ratios were 41ꢂ, 128ꢂ and 30ꢂ for 26, 2 and rosu-
vastatin, respectively. Given that a similar direct determination of
liver drug exposures during clinical studies was not feasible, we
sought a circulating biomarker measurement of hepatic inhibition
of HMG-CoA reductase that could be used as a surrogate using a
PK/PD relationship. Mevalonate (MVA) was selected for this
purpose since it is the direct product of HMG-CoA reductase.19
The liver is the major site of MVA production with approximately
1% reaching systemic circulation. These plasma MVA concentra-
tions have been shown to closely correlate with hepatic cholesterol
-20
-40
-60
-80
ED50
2
3.6 mg/kg
5.9 mg/kg
13.5 mg/kg
73.3 mg/kg
26
Rosuvastatin
Pravastatin
0
1
10
100
Dose
(mg/kg)
Figure 2. Dose–response effect of 26 on plasma low density lipoprotein (LDL) in the
Guinea pig. Animals were dosed with 2, 26, pravastatin or rosuvastatin for seven days.
All groups were sacrificed 2 h post dose on the final day of dosing. Data represent the
mean SEM of the percent change from the control group (n = 8/group).
Table 2
Protein binding properties of selected HMG-CoA inhibitors
Compd
Rat plasma
fu
Rat liver
Dog plasma
Human plasma
fu
fu
fu
2
26
0.016
0.15
Rosuvastatin 0.06
0.052
0.23
0.17
0.032
0.43
ND
0.051
0.29
0.17
Plasma protein binding value determined at 0.5
values determined at 1 M.
lg/mL. Liver tissue protein binding
l
As summarized in Table 3, rat and dog pharmacokinetics stud-
ies revealed both 2 and 26 to have moderate systemic plasma
clearance, low volume of distribution and relatively low bioavail-
ability (particularly for 26) consistent with the preclinical profiles
of other statins. Evaluation of clearance mechanisms in rat re-
vealed biliary excretion of parent as the primary clearance path-
way for both compounds with negligible contributions from
renal clearance or oxidative metabolism. Separately, since certain
statins are known to form side-chain lactones (i.e., closure of C-3
hydroxyl with carboxylate) under physiological conditions, during
these rat and dog PK studies of 2 and 26 plasma samples of were
concurrently analyzed for the corresponding lactones.17 While
such lactones are intrinsically inactive against HMG-CoA reduc-
tase, there is a possibility that circulating levels of these more per-
meable lactone forms may distribute to non-hepatic tissues where
reconversion to the active acid form could occur thus reducing the
apparent hepatoselectivity of an inhibitor.17 In both rat and dog
there were no detectable lactone concentrations found for com-
pound 26. For compound 2 plasma lactone levels were found to
Table 3
Preclinical pharmacokinetics studies for 2 and 26 in rat and dog
a
a
Compd
Dose (mg/kg)
AUC(0–24) (ng * h/mL)
Cmax (ng/mL)
Cl (mL/min/kg)
T1/2 (h)
Vdss (L/kg)
%F
2
2
Rat iv
Rat po
1
10
10
6.4
—
1.3
27
—
0.34
—
0.14
—
—
6%
2
2
Dog iv
1
5
28
64
—
51.5
19
—
0.78
—
0.10
—
—
44%
Dog pob
26
26
Rat iv
Rat po
1
10
83
2.7
—
2.0
39
—
0.43
—
0.17
—
—
0.3%
26
26
Dog iv
1
5
293
51
—
5.4
25
—
0.27
—
0.10
—
—
2.2%
Dog poc
Pharmacokinetic parameters expressed as geometric mean of n = 2 animals per group unless otherwise noted.
a
AUC(0–24) and Cmax reported as free drug concentration using plasma protein binding values determined in Table 3.
n = 3.
n = 1.
b
c

J. A. Pfefferkorn et al. / Bioorg. Med. Chem. Lett. 21 (2011) 2725–2731
2729
Table 4
Evaluation of transporter mediated uptake of 2, 26 and rosuvastatin in rat hepatocytes and transfected cells expressing rat transporters18
Compound
Concn (
l
M)
Rat hepatocycte uptake ratesa
Rat transporters fold uptakeb
rOatp1a4 rOatp1b2
Rate (pmol/min/mg protein)
% Inhibitablec
rOatp1a1
2
2
1
10
14
—
98%
—
3.6
4.2
2.7
3.0
2.9
2.6
26
26
1
10
17
—
100%
—
5.8
4.4
4.4
7.3
1.8
1.7
Rosuvastatind
Rosuvastatin
1
10
24
—
97%
—
8.2
5.3
8.6
17.3
—
—
Pravastatind
Pravastatin
1
10
—
—
—
—
2.4
3.9
1.5
1.8
3.6
7.0
Midazolame
Midazolam
1
10
—
—
—
—
1.6
1.7
0.93
0.82
1.1
1.3
a
b
c
Determined in rat hepatocytes cultured in sandwich configuration using BioCoat plates with Matrigel overlay.
Conducted in transfected CHO or HEK cells.
Inhibition by pan OATP inhibitor rifamycin SV.
Positive control.
d
e
Negative control.
Table 5
Evaluation of transporter mediated uptake of 2, 26 and rosuvastatin in human hepatocytes and transfected cells expressing human transporters
Compound
Concn (
lM)
Human hepatocycte uptake ratesa
Rate (pmol/min/mg protein)
% Inhibitablec
Human transporters fold uptakeb
hOATP1B3
hOATP1B1
hOATP2B1
2
2
1
10
7.8
—
78%
—
21.6
18.4
11.1
5.0
2.1
1.4
26
26
1
10
1.5
—
97%
—
49.0
84.9
34.5
36.4
7.1
6.8
Rosuvastatind
Rosuvastatin
1
10
8.0
—
85%
—
14.4
69.6
3.7
11.1
5.5
6.6
Pravastatind
Pravastatin
1
10
—
—
—
—
73.7
223
13.0
32.8
—
—
Midazolame
Midazolam
1
10
—
—
—
—
0.6
0.9
0.9
1.1
0.8
0.7
a
b
c
Determined in human hepatocytes cultured in sandwich configuration using BioCoat plates with Matrigel overlay.
Conducted in transfected CHO or HEK cells.
Inhibition by pan OATP inhibitor rifamycin SV.
Positive control.
d
e
Negative control.
mevalonic acid were concurrently measured at each time point.
Figure 4 illustrates the time course for both plasma drug (Panel
A) and MVA (Panel B) concentrations during this experiment.
While compound 26 exhibited the lowest plasma concentrations
it also offered the most significant and sustained reduction in the
MVA biomarker relative to either 2 or rosuvastatin with a notable
47% MVA reduction even at 24 h post dose. This property of 26 is
further illustrated in PK/PD analysis (Panel C) where plotting plas-
ma drug exposure versus MVA reduction suggests that 26 achieves
comparable levels of MVA reduction relative to 2 or rosuvastatin
but does so at lower systemic drug concentrations. This observa-
tion, coupled with the data of a more sustained MVA reduction ob-
served with 26 in time course studies (i.e., Fig. 4, Panel B), may
suggest that, in rat, 26 achieves a more favorable balance of hepatic
uptake and efflux relative to the other inhibitors. Interestingly, this
PK/PD data (Fig. 4, Panel C) which indicates 26 to be more function-
ally hepatoselective than 2 somewhat contrasts the previous liver/
plasma drug distribution data ( Fig. 3) which indicated that 2
achieved a greater (128ꢂ vs 41ꢂ) liver-to-plasma free drug ratios.
Based on promising efficacy and preclinical safety data, both 2
(PF-03052334)21 and 26 (PF-03491165) were advanced to Phase
1 ascending single dose studies in healthy volunteers to evaluate
the human safety and pharmacokinetics of these novel statins. This
dose escalation also offered an opportunity to evaluate human
1000
100
10
1
Compound 2
Compound 26
Rosuvastatin
Figure 3. Liver-to-plasma distribution of 2 (25 mg/kg), 26 (10 mg/kg) and rosu-
vastatin (10 mg/kg) after single oral dose in Sprague-Dawley rats. Ratios were
determined based on AUC(0–24h) of free drug concentrations in liver and plasma.
Time points were samples at t = 0, 0.5, 1, 2, 3, 4, 7 and 24 h with n = 3 animals per
time point.
synthesis making it an attractive circulating biomarker for evaluat-
ing the hepatic action of HMG-CoA reductase inhibitors.20 To
establish a preclinical PK/PD relationship, during the drug tissue
distribution studies described in Figure 3, the plasma levels of

2730
J. A. Pfefferkorn et al. / Bioorg. Med. Chem. Lett. 21 (2011) 2725–2731
(B) Timecourse of Plasma Mevalonate (MVA) Reduction
100
10
0
20
(A) Timecourse of Plasma Drug Levels
Compound 2 (25 mg/kg)
Compound 26 (10 mg/kg)
Rosuvastatin (10 mg/kg)
1
40
0.1
60
0.01
0.001
80
100
0
4
8
12
16
20
24
0
4
8
12
16
20
24
Time (h)
Time (hr)
(C) PK/PD for Plasma Drug Concentration vs. MVA Reduction
0
20
40
60
80
100
0.0001
0.01
1
100
10000
[Plasma,free Drug] / IC50
Figure 4. Plasma drug levels (A), plasma mevalonate (MVA) changes (B) and exposure effect relationships (C) for 2, 26 and rosuvastatin in Sprague-Dawley rats.
pharmacodynamics via mevalonate (MVA) changes as summarized
in Table 6. Both compounds were well tolerated, and demonstrated
relatively dose-proportional increases in drug exposure. Inhibitor 2
had a terminal half-life of 6–9 h with 12–16% of drug excreted in
the urine at 24 h post dose. Inhibitor 26 had a terminal half-life
of 4–7 h with <1% of drug excreted in the urine at 24 h post dose.
At Cmax plasma lactone levels of 2 averaged 0.08-fold of the parent
acid concentration whereas plasma lactone levels of 26 averaged
twofold of the parent acid concentration.
MVA biomarker measurements revealed dose-dependent de-
creases both for 2 and 26. For example, at a dose of 120 mg, 26
achieved a plasma Cmax of 3.31 ng/ml resulting in a decrease in
plasma mevalonate AUC(0–24h) of 46% whereas a dose of 80 mg of
2 achieved plasma Cmax of 3152 ng/ml resulting in a decrease in
Table 6
Human pharmacokinetics and mevalonate (MVA) changes for Phase 1 single ascending dose studies of 2 and 26 in human volunteers
Plasma exposures total drugb
Plasma exposures free drugc
Mean% change plasma MVA relative to placebo
controld
a
Compd Dose (mg)
N
t1/2 (h)
*
*
Cmax (ng/mL) AUC(0-1) (ng h/mL) Cmax (ng/mL) AUC(0-1) (ng h/mL) AUC(0–24h)
AUC(0–72h)
2
6
1
5
20
40
80
160
6
6
6
6
5
6
2.55 (33)
4.54 (31)
5.27 (42)
8.57 (46)
6.28 (23)
8.87 (45)
33.5 (27)
196 (27)
740 (28)
1770 (24)
4480 (12)
6810 (27)
78.5 (30)
434 (32)
1930 (21)
5110 (31)
11,800 (9)
25,000 (23)
1.71
10.0
37.7
90.3
228
347
4.00
22.1
98.4
261
602
1280
ꢀ16 27
ꢀ13 32
ꢀ8 25
+3 14
ꢀ17 11
ꢀ3 17
ꢀ29 21
ꢀ24 22
ꢀ10 21
ꢀ13 10
ꢀ19 16
ꢀ17 19
ꢀ30 31
ꢀ28 37
ꢀ34
7
ꢀ41 14
ꢀ65
5
ꢀ55 16
3
10
40
120
240
320
6
6
6
6
5
6
—
—
0.0552 (245)
0.355 (202)
2.25 (94)
3.31 (90)
9.27 (58)
—
2.71
6.27 (83)
12.5 (91)
27.2 (58)
36.0 (79)
0.016
0.103
0.653
0.960
2.69
—
ꢀ5 23
ꢀ25 10
ꢀ39 19
0.786
1.81
3.63
7.89
10.4
3.22 (39)
3.07 (69)
4.06
ꢀ46
9
ꢀ56 12
ꢀ61 11
7.36 (156) 44.1 (121)
12.8
a
b
c
Terminal half-life (t1/2) reported as arithmetic mean (%CV).
Cavg and AUC(0-1) values reported as arithmetic mean (%CV) of total drug levels for parent carboxylic acid.
Cavg and AUC(0-1) values reported as free drug levels for parent carboxylic acid using protein binding values reported in Table 3.
d
Percent change from placebo standard deviation of plasma mevalonic acid.

J. A. Pfefferkorn et al. / Bioorg. Med. Chem. Lett. 21 (2011) 2725–2731
2731
Sandberg, J.; Österlund, M.; Barkhem, T.; Höglund, M.; Wang, J.; Fung, S.;
Wilcox, D.; Nguyen, P.; Jakob, C.; Hutchins, C.; Färnegardh, M.; Kauppi, B.;
Öhman, L.; Jacobson, P. B. J. Med. Chem. 2004, 47, 4213; (b) Link, J. T.; Sorensen,
B. K.; Lai, C.; Wang, J.; Fung, S.; Deng, D.; Emery, M.; Carroll, S.; Grynfarb, M.;
Goos-Nilson, A.; von Geldern, T. Bioorg. Med. Chem. Lett. 2004, 14, 4173.
4. Boyer, S. H.; Jiang, H.; Jacintho, J. D.; Reddy, M. V.; Li, H.; Godwin, J. L.; Schulz,
W. G.; Cable, E. E.; Hou, J.; Wu, R.; Fujitaki, J. M.; Hecker, S. J.; Erion, M. D. J. Med.
Chem. 2008, 51, 7075.
5. Bebernitz, G. B.; Beaulieu, V.; Dale, B. A.; Deacon, R.; Duttaroy, A.; Gao, J.;
Grondine, M. S.; Gupta, R. C.; Kakmak, M.; Kavana, M.; Kirman, L. C.; Liang, J.;
Maniara, W.; Munshi, S. S.; Schuster, F. F.; Stams, T.; Denny, I. S.; Taslimi, P. M.;
Vash, B.; Caplan, S. L. J. Med. Chem. 2009, 52, 6142.
6. (a) Koltun, D. O.; Zilbershtein, T. M.; Migulin, V. A.; Vasilevich, N. I.; Parkhill, E.
Q.; Glushkov, A. I.; McGregor, M. J.; Brunn, S. A.; Chu, N.; Hao, J.; Mollova, N.;
Leung, K.; Chisholm, J. W.; Zablocki, J. Bioorg. Med. Chem. Lett. 2009, 19, 4070;
(b) Koltun, D. O.; Vasilevich, N. I.; Parkhill, E. Q.; Glushkov, A. I.; Zilbershtein, T.
M.; Mayboroda, E. I.; Boze, M. A.; Cole, A. G.; Henderson, I.; Zautke, N. A.; Brunn,
S. A.; Chu, N.; Hao, J.; Mollova, N.; Leung, K.; Chisholm, J. W.; Zablocki, J. Bioorg.
Med. Chem. Lett. 2009, 19, 3050.
7. For a review, see: Erion, M. D.; Bullough, D. A.; Lin, C.-C.; Hong, Z. Curr. Opin.
Invest. Drugs 2006, 7, 109.
8. Hoglen, N. C.; Chen, L.-S.; Fisher, C. D.; Hirakawa, B. P.; Groessl, T.; Contreras, P.
C. J. Pharmacol. Exp. Ther. 2004, 309, 634.
9. Boyer, S. H.; Sun, Z.; Jiang, H.; Esterbrook, J.; Gómez-Galeno, J. E.; Craigo, W.;
Reddy, K. R.; Ugarkar, B. G.; MacKenna, D. A.; Erion, M. D. J. Med. Chem. 2006, 49,
7711.
10. (a) Wess, G.; Kramer, W.; Schubert, G.; Enhsen, A.; Baringhaus, K.-H.; Glombik,
H.; Müllner, S.; Bock, K.; Kleine, H.; John, M.; Neckermann, G.; Hoffman, A.
Tetrahedron Lett. 1993, 34, 819; (b) Wess, G.; Kramer, W.; Han, X. B.; Bock, K.;
Enhsen, A.; Glombick, H.; Baringhaus, K.-H.; Böger, G.; Urmann, M.; Hoffman,
A.; Falk, E. J. Med. Chem. 1994, 37, 3240; (c) Verdine, G. L.; Norimine, Y.; Gude-
Rodrigues, L. WO 2007005941 A2.
11. (a) Brukert, E.; Hayem, G.; Dejager, S.; Yau, C.; Begaud, B. Cardiovasc. Drugs Ther.
2005, 19, 403; (b) Baer, A. N.; Wortmann, R. L. Curr. Opin. Rheumatol. 2007, 19,
67; (c) Tiwari, A.; Bansal, V.; Chugh, A.; Mookhtiar, K. Expert Opin. Drug Saf.
2006, 5, 651; (d) Athar, H.; Shah, A. R.; Thompson, P. D. Future Lipidol. 2006, 1,
143; (e) Franc, S.; Dejager, S.; Bruckert, E.; Chauvenet, M.; Giral, P.; Turpin, G.
Cardiovasc. Drugs Ther. 2003, 17, 459.
12. (a) Hamelin, B. A.; Turgeon, J. Trends Pharm. Sci. 1998, 19, 26; (b) Rosenson, R.
S.; Tangney, C. C. JAMA 1998, 279, 1643; See also: (c) Schachter, M. Fundam.
Clin. Pharmacol. 2004, 19, 117; (d) Ho, R. H.; Tirona, R. G.; Leake, B. F.; Glaeser,
H.; Lee, W.; Lemke, C. J.; Wang, Y.; Kim, R. B. Gastroenterology 2006, 130, 1793.
13. For a reviews, see: (a) Rodrigues, A. C. Expert Opin. Drug Metab. Toxicol. 2010, 6,
621; (b) Niemi, M. Clin. Pharm. Ther. 2010, 87, 130; For a representative
example of genetic variant data, see: SEARCH Collaborative Group N. Eng. J.
Med. 2008, 359, 789.
14. Bowles, D. M.; Bolton, G. L.; Boyles, D. C.; Curran, T. T.; Hutchings, R. H.; Larsen,
S. D.; Miller, J. M.; Park, W. K. C.; Ritsema, K. G.; Schineman, D. C. Org. Process
Res. Dev. 2008, 12, 1183.
15. Baumann, K. L.; Butler, D. E.; Deering, C. F.; Mennen, K. E.; Millar, A.; Nanninga,
T. N.; Palmer, C. W.; Roth, B. D. Tetrahedron Lett. 1992, 33, 2283.
16. For a full description of in vitro and in vivo biological assays as well as
experimental conditions, see Ref. 2e.
plasma mevalonate AUC(0–24h) of 65%. For comparison, in a sepa-
rate study, once daily dosing of 10 mg of rosuvastatin for 14 days
in healthy volunteers resulted in a 29.9% decrease in plasma mev-
alonate AUC(0–24h) accompanied by a 41.3% reduction in LDL-C with
a plasma Cmax of 4.58 ng/ml for this benchmark statin.22
From the perspective of drug distribution, while 2 and 26
achieved similar pharmacodynamic effects (i.e., MVA reductions)
they were achieved at significantly different systemic drug expo-
sures based on either free or total drug as shown in Table 6. For
example, at the highest doses of each compound which offered
comparable MVA reductions, the systemic free drug concentrations
of 2 at Cmax were 27-fold higher than for 26. This may suggest that
in humans 26 undergoes more efficient hepatic extraction than 2
and also has sufficient liver residency time to induce a sustained
reduction in hepatic cholesterol synthesis. Presumably, the lower
systemic drug levels observed for 26 may offer the potential for side
effects reduced effects, such as myalgia, that occur as a result of
non-hepatic inhibition of HMG-CoA reductase. This clinical obser-
vation of efficacy at lower systemic drug concentrations of 26 ver-
sus 2 was consistent with the preclinical PK/PD trend observed in
rat (Fig. 4, Panel C) as well as the in vitro transporter data suggesting
that 26 (vs 2) was a more effective substrate for the human
hOATP1B1, hOATP1B3 and hOATP2B1 transporters (Table 5). Inter-
estingly, the sandwich-cultured human hepatocyte uptake data did
not predict the increased hepatic uptake rate of 26 versus 2.
We have described the structure–activity studies leading to the
identification of hepatoselective HMG-CoA reductase inhibitor 26
as an early clinical candidate for treating hypercholesterolemia.
The comparative preclinical profiling of imidazole-based 26 versus
pyrazole-based inhibitor 2 highlighted interesting differences in
transporter interactions, tissue distribution and lactone formation
between these structurally similar candidates. The prediction of
tissue selective drug distribution remains an important drug dis-
covery challenge, and in the current case study preclinical hepatic
transporter assays and a PK/PD analysis of systemic drug concen-
tration versus hepatic pharmacodynamic effects offered insight
useful for predicting the rank order (26 > 2) of hepatoselectivity
observed in early clinical studies of these candidates.
References and notes
17. Duggan, D. E.; Vickers, S. Drug Metab. Rev. 1990, 22, 333.
1. For reviews, see: (a) Lin, J. H. Curr. Drug Metab. 2006, 7, 39; (b) Lanao, J. M.;
Fraile, M. A. Curr. Pharm. Des. 2005, 11, 3829; Mizuno, N.; Niwa, T.; Yotsumoto,
Y.; Sugiyama, Y. Pharmacol. Rev. 2003, 55, 425.
18. Kalgutkar, A. S.; Feng, B.; Nguyen, H. T.; Frederick, K. S.; Campbell, S. D.; Hatch,
H. L.; Bi, Y. A.; Kazolias, D. C.; Davidson, R. E.; Mireles, R. J.; Duignan, D. B.; Choo,
E. F.; Zhao, S. X. Drug Metab. Dispos. 2007, 35, 2111.
2. For a review, see: (a) Pfefferkorn, J. A. Curr. Opin. Invest. Drugs 2009, 10, 245; For
representative examples, see: (b) Koga, T.; Shimada, Y.; Kuroda, M.; Tsujita, Y.;
Hasegawa, K.; Yamazaki, M. Biochim. Biophys. Acta 1990, 1045, 115; (c) Roth, B.
D.; Bocan, T. M.; Blankey, C. J.; Chucholowski, A. W.; Creger, P. L.; Creswell, M.
W.; Ferguson, E.; Newton, R. S.; O’Brien, P.; Picard, J. A.; Roark, W. H.; Sekerke,
C. S.; Sliskovic, D. R.; Wilson, M. W. J. Med. Chem. 1991, 34, 463; (d) Ahmad, S.;
Madsen, C. S.; Stein, P. D.; Janovitz, E.; Huang, C.; Ngu, K.; Bisaha, S.; Kennedy, L.
J.; Chen, B. C.; Zhao, R.; Sitkoff, D.; Monshizadegan, H.; Yin, X.; Ryan, C. S.;
Zhang, R.; Giancarli, M.; Bird, E.; Chang, M.; Chen, X.; Setters, R.; Search, D.;
Zhuang, S.; Nguyen-Tran, V.; Cuff, C. A.; Harrity, T.; Darienzo, C. J.; Li, T.; Reeves,
R. A.; Blanar, M. A.; Barrish, J. C.; Zahler, R.; Robl, J. A. J. Med. Chem. 2008, 51,
2722;; (e) Pfefferkorn, J. A.; Choi, C.; Larsen, S. D.; Auerbach, B.; Hutchings, R.;
Park, W.; Askew, V.; Dillon, L.; Hanselman, J. C.; Lin, Z.; Lu, G. H.; Robertson, A.;
Sekerke, C.; Harris, M. S.; Pavlovsky, A.; Bainbridge, G.; Caspers, N.; Kowala, M.;
Tait, B. D. J. Med. Chem. 2008, 51, 31.
19. McTaggart, F.; Buckett, L.; Davison, R.; Holdgate, G.; McCormick, A.; Schneck,
D.; Smith, G.; Warwick, M. Am. J. Cardiol. 2001, 87, 23B.
20. (a) Parker, T. S.; McNamera, D. J.; Brown, C.; Garrigano, O.; Kolb, R.; Batwin, H.;
Ahrens, E. H. Proc. Natl. Acad. Sci. 1982, 79, 3037; (b) Parker, T. S.; McNamara, D.
J.; Brown, C. D.; Kolb, R.; Ahrens, E. H.; Alberts, A. W.; Tobert, J.; Chen, J.; De
Schepper, P. J. J. Clin. Invest. 1984, 74, 795; (c) Yoshida, T.; Honda, A.; Tanaka, N.;
Matsuzaki, Y.; He, B.; Osuga, T.; Kobayashi, N.; Ozawa, K.; Miyazaki, H. J.
Chromatogr., B 1993, 613, 185.
21. Pfefferkorn, J. A.; Choi, C.; Larsen, S. D.; Auerbach, B.; Askew, V.; Dillon, L.;
Hanselman, J. C.; Kowala, M.; Lin, Z.; Lu, G. H.; Robertson, A.; Sekerke, C.;
Bakker-Arkema, R.; Barclay, P.; Kindt, E.; Le, V.; Mandema, J. W.; Milad, M.
Pyrazoles as Hepatoselective HMG-CoA Reductase Inhibitors: Discovery and Early
Clinical Evaluation of PF-3052334 for the Treatment of Hypercholesterolemia; 31st
National Medicinal Chemistry Symposium; Pittsburgh, PA. June 15–19, 2008.
22. Martin, P. D.; Mitchell, P. D.; Schneck, D. W. Br. J. Clin. Pharmacol. 2002, 54,
472.
3. (a) von Geldern, T. W.; Tu, N.; Kym, P. R.; Link, J. T.; Jae, H.-S.; Lai, C.; Apelqvist,
T.; Rhonnstad, P.; Hagberg, L.; Koehler, K.; Grynfarb, M.; Goos-Nilsson, A.;