Selective Nonsteroidal Glucocorticoid Receptor Modulators for the Inhaled Treatment of Pulmonary Diseases

Article abstract of DOI:10.1021/acs.jmedchem.7b01215

A class of potent, nonsteroidal, selective indazole ether-based glucocorticoid receptor modulators (SGRMs) was developed for the inhaled treatment of respiratory diseases. Starting from an orally available compound with demonstrated anti-inflammatory activity in rat, a soft-drug strategy was implemented to ensure rapid elimination of drug candidates to minimize systemic GR activation. The first clinical candidate 1b (AZD5423) displayed a potent inhibition of lung edema in a rat model of allergic airway inflammation following dry powder inhalation combined with a moderate systemic GR-effect, assessed as thymic involution. Further optimization of inhaled drug properties provided a second, equally potent, candidate, 15m (AZD7594), that demonstrated an improved therapeutic ratio over the benchmark inhaled corticosteroid 3 (fluticasone propionate) and prolonged the inhibition of lung edema, indicating potential for once-daily treatment.

Full text of DOI:10.1021/acs.jmedchem.7b01215

Subscriber access provided by University of Sussex Library
Article
Selective Non-steroidal Glucocorticoid Receptor Modulators
for the Inhaled Treatment of Pulmonary Diseases
Martin Hemmerling, Stinabritt Nilsson, Karl Edman, Stefan Eirefelt, Wayne Russell, Ramon Hendrickx,
Eskil Johnsson, Carina Kärrman-Mårdh, Markus Berger, Hartmut Rehwinkel, Anna Abrahamsson, Jan
Dahmén, Anders R. Eriksson, Balint Gabos, Krister Henriksson, Nafizal Hossain, Svetlana Ivanova,
Anne-Helene Jansson, Tina J. Jensen, Anders Jerre, Henrik Johansson, Tomas Klingstedt, Matti Lepistö,
Martin Lindsjö, Irene Mile, Grigorios Nikitidis, John Steele, Ulrika Tehler, Lisa Wissler, and Thomas Hansson
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.7b01215 • Publication Date (Web): 22 Sep 2017
Downloaded from http://pubs.acs.org on September 23, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
Pharmaceutical Sciences
Wissler, Lisa; PEPPAREDSLEDEN 1,
Hansson, Thomas; Medicines for Malaria Venture; AstraZeneca AB
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 47
1
2
3
4
5
6
7
8
9
Selective Non-steroidal Glucocorticoid Receptor
Modulators for the Inhaled Treatment of
Pulmonary Diseases
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Martin Hemmerling,*† Stinabritt Nilsson,∆ Karl Edman,‡ Stefan Eirefelt,∆ Wayne Russell,∆
Ramon Hendrickx,† Eskil Johnsson,† Carina Kärrman Mårdh,† Markus Berger,# Hartmut
Rehwinkel,# Anna Abrahamsson,† Jan Dahmén,∆ Anders R. Eriksson,† Balint Gabos,∆ Krister
Henriksson,∆ Nafizal Hossain,∆ Svetlana Ivanova,∆ Anne-Helene Jansson,∆ Tina J. Jensen,†
Anders Jerre,∆ Henrik Johansson,∆ Tomas Klingstedt,∆ Matti Lepistö,† Martin Lindsjö,¶ Irene
Mile,∆ Grigorios Nikitidis,∆ John Steele,† Ulrika Tehler,¶ Lisa Wissler,‡ and Thomas
Hansson†
†Respiratory, Inflammation & Autoimmunity, Innovative Medicines and Early Development Biotech
Unit, AstraZeneca, Pepparedsleden 1, Mölndal 43183, SE
‡Discovery Sciences, Innovative Medicines and Early Development Biotech Unit, AstraZeneca,
Pepparedsleden 1, Mölndal 431 83, SE
¶Pharmaceutical Sciences, Innovative Medicines and Early Development Biotech Unit, AstraZeneca,
Pepparedsleden 1, Mölndal 43183, SE
#Medicinal Chemistry Berlin, Drug Discovery, Pharmaceuticals, Bayer AG, Berlin 13353, DE
∆AstraZeneca R&D Lund, Scheelevägen 1, Lund 22187, Sweden, SE
ACS Paragon Plus Environment

Page 3 of 47
Journal of Medicinal Chemistry
1
2
3
ABSTRACT
4
5
6
7
8
9
A class of potent, non-steroidal, selective indazole ether based glucocorticoid receptor
modulators (SGRMs) was developed for the inhaled treatment of respiratory diseases. Starting
from an orally available compound with demonstrated anti-inflammatory activity in rat, a soft
drug strategy was implemented to ensure rapid elimination of drug candidates and minimize
systemic GR activation. The first clinical candidate 1b (AZD5423) displayed potent inhibition
of lung edema in a rat model of allergic airway inflammation following dry powder inhalation
combined with a moderate systemic GR-effect, assessed as thymic involution. Further
optimization of inhaled drug properties provided a second, equally potent, candidate 15m
(AZD7594) that demonstrated an improved therapeutic ratio over the benchmark inhaled
corticosteroid 3 (fluticasone propionate) and prolonged inhibition of lung edema, indicating
potential for once-daily treatment.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
INTRODUCTION
In 1948 the introduction of glucocorticoids (GCs) into clinical practice constituted a
breakthrough in anti-inflammatory therapy.1 Further developments led to synthetic steroids
with improved therapeutic profiles that are, to date, the mainstay of inflammatory and
autoimmune disease therapy.2 GCs act through binding to the glucocorticoid receptor (GR),
triggering the up- and down-regulation of gene transcription and ultimately modulation of the
immune response and energy homeostasis.3 The latter GR function is associated with clinical
adverse effects and limits the prolonged use of traditional steroidal GR agonists.4 The search
for glucocorticoids with reduced adverse effects led to the discovery of non-steroidal GR
modulators, which have the potential to dissociate the desired anti-inflammatory activity from
metabolic adverse effects at a mechanistic level through differential DNA binding and co-
regulator recruitment. For pulmonary diseases such as asthma and chronic obstructive
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 47
1
2
3
4
5
6
7
8
9
pulmonary disease (COPD), an inhalation approach may limit systemically driven adverse
effects.5 Whilst inhalation delivery effectively reduces circulating drug levels, chronic
treatment with inhaled corticosteroids (ICS) can still lead to suppression of the hypothalamic-
pituitary-adrenal (HPA) axis, an indicator of systemic GR activation.6 Documented adverse
effects associated with ICS include growth retardation in children, bone mineral loss and
development of type II diabetes in the elderly population.7 In addition, patients with severe
asthma and COPD often exhibit reduced responsiveness to current ICS therapy and more
efficacious alternatives are highly anticipated.8 Herein we describe the medicinal chemistry
program leading to the discovery of two inhaled non-steroidal clinical drug candidates, 1b and
15m, aimed at the control of pulmonary diseases with an inflammatory component such as
asthma and COPD.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
We recently reported the discovery of indazole ethers as a novel class of highly potent non-
steroidal GR modulators covering a wide range of the physicochemical property space and
thus with the potential to be optimized for different routes of administration.9
In this series, compound 1a demonstrated excellent oral efficacy in a rat arthritis model but
also exhibited several properties consistent with the inhalation design guidelines for
pulmonary targeting via inhaled administration.10 Overall, 1a possesses suitable solid state
properties such as crystallinity and low hygroscopicity to facilitate dry powder inhalation
(DPI) formulations. More specifically, 1a displayed picomolar cellular potency
(transrepression activity) in a range similar to current ICS (Figure 1), preferred to overcome
dose limitations set by the device capacity.11 In addition, compound 1a exhibited high
nonspecific binding to rat plasma protein (99.99 % bound) coupled to a high steady state
volume of distribution in rat (Vss 13.5 L/kg).9 As large volumes result when nonspecific
binding to lipids exceeds the nonspecific binding to plasma protein,12 it was hypothesized that
high tissue affinity would contribute to increased local drug levels following inhalation. We
ACS Paragon Plus Environment

Page 5 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
concluded that compound 1a was a suitable starting point for a chemistry program to identify
an inhaled, selective non-steroidal GR modulator (SGRM) to provide a safer treatment of
asthma than current ICS. For the preclinical evaluation, a precedented methodology was
adopted to assess the therapeutic ratio in the rat, derived from the inhibition of allergen-
induced lung edema and the systemic involuting effect on the thymus.13
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1: GR ligands with picomolar transrepression potency
CHEMISTRY
The general synthetic strategy for the indazole series follows the route outlined in Schemes 1
and 2. To facilitate a building block approach, the basic indazole ether core was assembled
by an ether formation, fusing a 5-iodo-indazole with a substituted norephedrine 7. The
modified norephedrines were available through the Weinreb-Nahm ketone synthesis reacting
the Weinreb amide of a suitably protected L-alanine derivative (4) with a freshly-prepared
aryl-Grignard reagent to form the phenylalkyl ketone 5.14 A stereoselective Meerwein-
Ponndorf-Verley reduction15 provided the carbamoyl-protected substituted norephedrines 6 as
single stereoisomers that following deprotection yielded the corresponding (1R,2S)-
norephedrines 7 having the same absolute configuration as natural (-)-norephedrine. The
enantiomers with (1S,2R)-configuration are accessible from D-alanine applying the same
diasteroselective route. Specific rotations obtained for the (1R,2S) intermediate 7c were
similar to the reported value for (-)-norephedrine16, supporting the assignment of absolute
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 47
1
2
3
4
5
6
7
8
9
configuration. The opposite, positive rotation was measured for the (1S,2R) enantiomer ent-
7c. To access the syn-diastereomers with (S,S) and (R,R) configuration, a published method
was applied to transform the tert-butyl carbamoyl-protected amino alcohols in 6c to 4-methyl-
5-aryl-substituted oxazolidine-2-one 9 with concomitant inversion of the benzylic
stereocenter.16 Proton NMR comparison with the 4,5-cis-substituted oxazolidin-2-one 8,
formed with conservation of both stereocenters, confirmed the relative configuration of the
diastereomers. Hydrolysis of the oxazolidinone with retention of stereoconfiguration released
the norpseudoephedrine 10. The final steps to the putative GR-ligands are shown in Scheme
2. The indazole ether core was synthesized by an Ullmann coupling of a substituted
norephedrine 7 and an iodoindazole to produce the primary amine intermediates 11 and 12.17
Further conversion to carboxamides using standard amide coupling conditions provided the
final indazole ether ligands 1 and 14 or the isobutyl ester intermediate 13. Saponification and
a second amide coupling yielded the extended indazole ether ligands 15.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 1: Diastereoselective synthesis of phenylpropanolamine analogues
ACS Paragon Plus Environment

Page 7 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Reagents and conditions: a) i-PrMgCl·LiCl, 0-10°C; b) i-PrMgCl, THF, 0°C – rt; c) aq. HCl
(1.5 N), EtOAc, 10 °C; d) Al(OiPr)3 (0.2 equiv.), i-PrOH (11 equiv.), toluene, 50°C; e) HCl;
f) triphosgene, THF, -15 °C to rt; g) MeSO2Cl, Et3N, THF, 50 °C; h) LiOH, THF/H2O, 90
°C.
Scheme 2: Synthesis of indazole ethers
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 47
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For specified substituents NR4R5 refer to Table 2. Reagents and conditions: a) 1-(4-
fluorophenyl)-5-iodo-indazole (16), CuI (0.05 equiv.), Cs2CO3 (2.2 equiv.), n-PrCN, 125 °C;
b) [R1C(O)]2O; NEt3; c) R1CO2H, NEt3, T3P, MeCN; d) (acetoxy)acylchloride, DIPEA, THF;
e) NH4OH, MeOH; f) isobutyl 3-(5-iodoindazol-1-yl)benzoate (17), CuI (0.05 equiv.) Cs2CO3
(2.2 equiv.), n-PrCN, 110 °C; g) 2,2-difluoropropionic acid, HATU, DIPEA, DCM; h) LiOH,
THF/H2O (1:1); i) HNR4R5, HBTU, DIPEA, DMF.
RESULTS AND DISCUSSION
Inhaled administration of GR modulators aims to minimize adverse effects, and the improved
safety profile of current marketed agents is largely attributable to reduced systemic
concentrations compared to oral treatment.7a However, once absorbed into the systemic
circulation, rapid inactivation is important to reduce the risk of side effects. The most-
prescribed ICS are all antedrugs where metabolically labile motifs are appended to the
otherwise stable steroidal core.10a Three prominent examples, budesonide (2), fluticasone
propionate (3) and fluticasone furoate, achieve hepatic deactivation through oxidative
ACS Paragon Plus Environment

Page 9 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
degradation of labile substituents at the 16α, 17α or 17β position of the familiar tetracyclic
steroid core18 and a similar strategy was considered for the evolution of the starting point 1a
which was characterized by high resistance to hepatic degradation and high systemic
stability.9 Analysis of the binding conformation of 1a and 2 reveals that parts of the indazole
ether scaffold superimpose, or are in close proximity with the substituents of the steroidal
positions 17 and 16 (Figure 2). From a design perspective, it was plausible to introduce
metabolically labile substituents either as modified amide residues, or in the meta- and para-
positions of the norephedrine aryl group, designed to mimic the 17β-group and the 16α,17α-
group respectively. The metabolic clearance was assessed in human hepatocytes, with the
resulting data used for prediction of in vivo hepatic extraction.19
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2: Superimposition of the GR complexes of 1a (magenta) and 2 (blue); protein not
shown (RMSD 0.44 for 984 main chain atoms). Steroid sites with precedent for metabolic
deactivation are marked with yellow circles.
Designing in metabolic lability needed to be achieved within the established SAR of GR
agonism, as subnanomolar potency is commonplace for commercial ICS. To this end it was
critical to monitor potency and also crystallinity of the new ligands, since both properties play
a pivotal role in successful inhaled delivery and efficacy. Cellular potency was assessed both
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 47
1
2
3
as repression of AP-1 dependent transcription (Cell TR) and as inhibition of LPS-stimulated
4
5
6
TNF-α production in human peripheral blood mononuclear cells (PBMC).
7
8
9
The first metabolic inactivation option, the introduction of amides, was limited by the
structure-activity relationship of the series. As reported in a recent study on the indazole ether
series, small substituents including cyclopropyl, tert-butyl and trifluoromethyl amides with
low conversion in hepatocytes were preferred for potency.9 Therefore, the plan to achieve
metabolic inactivation focused on the exploration of the norephedrine phenyl moiety where
lipophilic extensions were well tolerated.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 1. In vitro properties of
N
-fluoro-phenyl indazole ethers in comparison with the
ICS 2 and 3
1
14
Human hepatic
clearancee
Chrom
logD
Cpd
R1
R2
Cell TRa,b
Human PBMCa,d
DSCf
IC50
(nM)
Eff.
IC50
(nM)
Eff.
in vitro
CLint
in vivo
EH
mp
(°C)
(%)c
(%)c
0.13
0.003
(2)
0.50
0.27
(2)
1a
1b
CF3
CF3
H
92
83
82
<2.1
16
<0.2
0.7
121
6.1
6.0
5.0
0.049
0.019
(2)
0.16
0.10
(6)
MeO
114
82
8.00
3.28
(2)
4.60
5.50
(9)
1c
CF3
CF3
CF3
OH
122
89
82
-
33
<2.7
NT
0.8
0.3
-
NA
NA
NA
464
255 (3)
(ClogP
3.7)
1d
OS(O)2OH
NT
0.029
0.014
(2)
0.073
0.075
(6)
14a
105
83
5.7
5.5
0.020
0.002
(2)
0.080
0.015
(2)
14b
MeCF2
121
84
28
0.8
117
ACS Paragon Plus Environment

Page 11 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
0.032
0.002
(2)
0.101
0.016
(2)
14c
14d
14e
2
MeCF2
MeCF2
MeCF2
108
117
114
111
113
87
70
67
79
85
14
20
15
20
33
0.6
0.7
0.7
0.7
0.8
121
NA
NA
255
280
5.7
5.4
5.4
3.4
4.9
0.156
0.11
(2)
0.322
0.10
(4)
0.066
0.024
(2)
0.132
0.004
(3)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
0.50
0.23)
(6)
1.57
0.94
(12)
0.040
0.002
(8)
0.044
0.045
(9)
3
aValue as mean SD (number of replicates). bTransrepression activity (TR), assessed as
inhibition of AP-1 mediated transcription.cEfficacy, maximum response relative to
Dexamethasone (100 %). dLPS-induced TNF-
α release in peripheral blood mononuclear cells
(PBMCs). eThe intrinsic clearance (Clint ; µL/min/106 cells) in human hepatocytes was scaled
to give an estimate for the in vivo hepatic clearance and the extraction ratio (EH) at hepatic
blood flow of 20 mL/min/kg. fMelting point of crystalline material only, assessed by
differential scanning calorimetry. NA, not applicable. NT, not tested.
This alternative approach built on the introduction of a metabolically labile methoxy group to
the norephedrine phenyl. An anisole derivative was expected to be metabolically stable in
organ tissue but could serve as a substrate for cytochrome-mediated degradation in the liver.
In principle, oxidative demethylation and subsequent conjugation of the resulting phenol by
secondary metabolism should render compounds inactive at the GR. In practice, the
introduction of a methoxy group provided compound 1b which combined excellent potency in
both the reporter gene cell assay and in human PBMCs with rapid conversion in hepatocytes,
corresponding to a predicted extraction in man of 67% of liver blood flow (Table 1).
Summarizing this initial effort to tailor the properties of the lead compound 1a, few examples
were found that met the in vitro requirements for functional cellular potency combined with
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 47
1
2
3
4
5
6
7
8
9
rapid degradation in hepatocytes. Between these 1b was the only compound that could be
crystallized (mp 82 °C), indicating the potential for dry powder inhalation. The systemic rat
pharmacokinetics of compound 1b indicated that the clearance from circulation (CL 38
mL/min/kg) was well predicted from clearance in rat hepatocytes. The large volume of
distribution Vss (23 L/kg) indicated a high tissue affinity of 1b and led to an extended terminal
half-life of 4.7 hours. A high fraction of the oral dose was absorbed from the gastrointestinal
tract and led to a moderate oral bioavailability of 37 % (Table 4).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
To validate the soft-drug principle the metabolic fate of 1b was studied in hepatocytes from
different species. In human and dog hepatocytes the GR-inactive sulfate 1d was the only
major metabolite, suggesting that efficient deactivation of 1b could be achieved. However, in
rat hepatocytes the active phenol 1c (rat PBMC IC50 0.56 nM) was predominantly formed and
is expected to provide a minor contribution to the systemic GR activity in the rat but not in
dog or man (Supplementary Information, Scheme S1).
Having demonstrated satisfactory plasma pharmacokinetics, the inhalation properties of 1b
were investigated in advanced dry powder inhalation systems which allow for pulmonary
particle distribution by active breathing.20 A delayed distribution into the systemic circulation
is desirable to prolong the pulmonary residence time of the drug to favor the local over
systemic effects.10a, 10d, 21 In our hands, the mean retention times (MRT) in rat lung for
established ICS were found to be 0.74 to 6.5 hours for 2 and 3 respectively. Inhaled 1b
displayed a mean retention time of 1.5 hours, similar to the faster absorbed 2 (Table 5) which
supported further evaluation of its anti-inflammatory efficacy in a rat PD model of allergic
asthma.22 Following dry powder inhalation 1b potently inhibited Sephadex-induced lung
edema in a dose-dependent manner with an ED50 of 1.2 µg/kg, 4-fold more potent than the
comparator 3 in this model (Table 6).
ACS Paragon Plus Environment

Page 13 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
As a sensitive marker of systemic GR agonism, thymic involution was evaluated in the same
Sephadex-challenged animals, measured as an ED25 value: reduction of thymus weight by 25
%.23 In this model, both 1b and 3 exerted a similar effect on thymus weight. Thus, 1b
displayed a 4-5-fold improvement in therapeutic ratio (TR) in rat over the benchmark ICS 3
when using the formula TR = ED25 thymus involution / ED50 lung edema (Table 6).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
We therefore decided to nominate 1b for clinical evaluation in asthma patients as a
development candidate drug, 2,2,2-trifluoro-N-((1R,2S)-1-((1-(4-fluorophenyl)-1H-indazol-5-
yl)oxy)-1-(3-methoxyphenyl)propan-2-yl)acetamide (AZD5423).24
With 1b progressing into clinical studies, a follow up program set out to expand the indazole
ether series and deliver a second candidate drug. The key achievements of 1b, in particular
picomolar cellular potency and efficient deactivation in human hepatocytes, needed to be
retained but there was still scope to improve the lung retention. A compound with slower
distribution from lung to circulation could provide two benefits. First, a reduction in peak
systemic drug levels and second, an extended duration of efficacy in lung tissue, thereby
providing a once-daily dosing platform. An additional design ambition was to avoid active rat
metabolites and thereby simplify the efficacy and safety evaluations.
The first modifications of 1b built on the methoxyphenyl motif which provided the rat-
specific phenol metabolite 1c. Guided by previous indazole ether SAR, electron-rich bicyclic
residues (R2 in 14, Table 1) were expected to maintain high potency. Annulation into a cyclic
ether motif would preserve the metabolically-labile site, as hepatic degradation would be
expected to occur α- to the ether oxygen, resulting in more polar, ring-opened metabolites
with reduced GR activity.
Example 14a, containing a dihydro-1,4-benzodioxane R2 group, showed a modest two-fold
increase in potency over the first candidate 1b in both of the cellular transrepression assay and
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 47
1
2
3
4
5
6
7
8
9
in PBMCs. A further change from the trifluoroacetamide to the 2,2-difluoropropionamide
introduced an additional dipole and provided the equipotent and crystalline 14b and 14c (a
dihydrobenzofuran analogue). Importantly, the accelerated clearance in human hepatocytes
indicated that the benzodioxane heterocycle could enhance the soft-drug potential. Use of the
other isomeric and slightly more polar benzodioxanes at R2 was less promising, giving
reduced potencies and amorphous products, examples 14d and 14e.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Compound 14b showed improved potency and intrinsic crystallinity compared to 1b and was
thus further profiled for rat pharmacokinetics in vivo. Following iv administration, plasma PK
of the two compounds was similar, although the oral bioavailability of 14b decreased to 11%
(Table 4). Following intratracheal administration (it), the lung PK showed rapid absorption of
14b with a slightly improved lung retention after 4 hours (4.5% of dose, Table 5) compared to
the first candidate 1b (1% of dose).
Although 14b represented progress, its high lipophilicity and intrinsic crystallinity were
similar to the first selected candidate 1b and resulted in an almost identical pharmacokinetic
profile. To obtain a truly differentiated drug candidate, the target profile was revised in two
ways. First, to significantly delay absorption from the lung, slow dissolution driven by high
crystallinity was prioritized. Secondly, an increase in polarity would reduce permeation
through lung tissue, lowering the volume of distribution to reduce exposure in peripheral
tissue. As more drastic structural changes would be needed to achieve this, a structure-guided
approach was adopted to identify opportunities for addition of polar interactions within the
GR binding pocket.
The X-ray structure of 1b in complex with the GR LBD and a NCOA2 peptide (residues 740-
753) confirmed that 1b exploits all regions of the receptor used by traditional, steroidal
glucocorticoids (Figure 3A). As such, the ligand binding pocket offers limited scope to
significantly modify the chemical scaffold. However, in common with compound 1a and
ACS Paragon Plus Environment

Page 15 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
other GR ligands displaying a similar indazole motif,9, 25 the fluorophenyl group of 1b
rearranged the gatekeeper residues Gln570 and Arg611, expanding the binding pocket
between helices 3 and 5. As a result, the ligand-binding pocket is fused with a large cavity
beneath Trp577 (Figure 3B). This cavity is lined with hydrophilic residues and in the crystal
structure is filled with ordered water molecules and an ethylene glycol derived from the
crystallization conditions. This presented an opportunity to extend the indazole ether core to
exploit the solvated cavity with new ligands of increased size and polar surface area – in line
with the revised target profile. As an added benefit, the additional polar groups could facilitate
intermolecular interactions in the crystalline state and consequently increase the melting
point.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. GR LBD (green) in complex with compound 1b (magenta). Putative hydrogen
bonds are marked with dashed lines. (A) Compound 1b places the amide within hydrogen
bonding distance of Asn564 and Glu642 of helices 3 and 7, respectively. The fluorophenyl
motif is inserted in the induced pocket in between helices 3 and 5. (B) The meta position of
the fluorophenyl motif is directed towards a large cavity beneath Trp577.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 47
1
2
3
Based on crystallographic data, the familiar phenyl indazole moiety was extended from the
4
5
6
meta-position of the N-phenyl group. By analogy to published GR modulators,25d
a
7
8
9
carboxamide was selected as a polar linker to position side chains along the preferred vector
into the induced, solvent stabilized cleft (Figure 3). Amino acid amides were initially explored
as polar substituents to form amides of the general structure 15, exemplified by the prolines
15a to 15d (Table 2). The D-prolinamide residue proved to be the preferred stereoisomer and
excellent potency was achieved for all aryl substituents A to C. However, the reduced
lipophilicity limited intrinsic clearance in hepatocytes. Furthermore, the polar prolinamide
functionality increased solubility and derailed our attempts to reduce aqueous solubility by
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
increasing crystallinity and melting point. Switching to the related γ-butyrolactone instead of
a proline in 15e and 15f resulted in a breakthrough, providing compounds that were easily
crystallized, melting in the region of 200°C - an increase of about 100°C compared to the first
candidate 1b. Importantly, for this matched pair of butyrolactones the bicyclic aryl variant B
in 15f improved potency by an order of magnitude over its partner 15e. In human hepatocytes,
15f was more rapidly cleared than the more-polar prolines 15c and 15d, reaching the target of
> 60% of predicted hepatic extraction (EH). Overall, 15f achieved all of the a priori target
criteria and is a key bridgehead for the design of improved analogues.
Table 2: Carboxamide extended indazole ethers
Human hepatic
clearancee
Chrom
logD
Cpd
Ar
R
Cell TRa,b
Human PBMCa,d
DSCf Solubilityg
Eff.
(%)c
IC50
(nM)
Eff.
in vitro
CLint
in vivo
EH
mp
(°C)
S / C
(µM)
IC50 (nM)
(%)c
ACS Paragon Plus Environment

Page 17 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
0.21
0.08 (3)
1.1
0.61(2)
15a
15b
15c
15d
15e
15f
A
A
B
C
A
B
107
86
78
NT
80
61
63
66
<2.7
NT
11
<0.3
NA
0.6
A
36 / NA
85 / NA
3.3
3.3
2.9
2.9
3.9
3.6
17.0
1.88 (2)
NT
NA
NA
NA
221
183
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
0.12
0.008
(2)
0.25
0.18
(2)
99
47 / NA
0.15
0.078
(2)
0.61
0.10
(2)
85
9.7
NT
17
0.6
19 / NA
13.4
5.95 (2)
18 0.76
(4)
92
NA
0.7
2.0 / 0.12
0.16 / 0.22
0.93
0.33
(2)
0.92
0.13 (2)
101
2.4 0.62
(2)
5.0
(1)
15g
15h
15i
A
B
B
103
99
58
69
70
<2.7
23
<0.3
0.7
140
180
215
1.0 / 0.34
3.8
3.6
3.6
1.4
0.39
(6)
0.21
0.11 (8)
0.01 / 0.68
0.17
0.035
(10)
0.54
0.32
(8)
<0.01 /
<0.1
102
22
0.7
0.17
0.041
(4)
0.037
0.003(2)
<0.1 /
0.23
15j
B
C
104
103
72
70
21
21
0.7
0.7
127
158
3.7
3.6
0.062
0.055
(5)
0.38
0.23
(6)
<0.1 /
0.02
15k
0.051
0.004
(2)
0.66
0.038
(6)
15l
15m
15n
B
B
C
105
107
104
78
75
72
30
29
26
0.8
0.8
0.8
175
177
192
1.3 / 0.09
3.7
3.7
3.6
0.056
0.021
(4)
0.43
0.16
(6)
<0.1 /
0.02
0.19
0.018
(3)
0.625
0.062
(4)
0.11 / 0.03
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 47
1
2
3
4
5
6
7
8
For annotations a, b, c, d, e, f refer to Table 1. gSolubility in 0.1M phosphate buffer (pH 7.4); S: at
20 °C, starting from 10 mM DMSO stock solution; C: at 23 °C, starting from crystalline solid.
NA, not applicable. NT, not tested.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The sulfolane amides 15h and 15i maintained the high-melting crystalline profile and further
improved potency, with a preference for the R-isomer in 15i. Both isomers were rapidly
cleared by human hepatocytes and compound 15i was progressed as a potential clinical
candidate for further profiling. The pyridylmethyl amide pair 15j and 15k, differentiated only
by the regiochemistry of the norephedrine aryl group both improved on potency over the
sulfolane 15i and retained rapid clearance in hepatocytes. Whilst measured lipophilicity was
similar for the regioisomeric pair, the significant difference in melting temperatures led to a
very different solubility. As a result, 15j and 15k, provided an opportunity to study the impact
of differentiated solid-state properties on in vivo pharmacokinetics and pharmacology, and
both compounds were subjected to additional in vivo studies. From the triad of N-
tetrahydrofuran-3-yl amides 15l-15n, the epimers 15l and 15m could not be differentiated in
vitro. Both compounds displayed the highest observed potencies in human and rat PBMCs
and provided an increased melting point, indicating a higher intrinsic crystallinity. 15m was
arbitrarily selected for further progression. The benzodioxane regioisomer 15n showed
reduced potency and was not progressed.
As recently reported, indazole ether based GR modulators are in general highly selective.9 In
contrast to marketed ICS that typically display nanomolar affinities at the progesterone (PR)
and mineralocorticoid receptor (MR),10a the indazole ether based GR modulators improve PR
selectivity by at least 100-fold and show no binding to MR up to the limit of binding assay
detection (Table 3). Substantially improved steroid hormone receptor selectivity could
ACS Paragon Plus Environment

Page 19 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
potentially contribute to clinical safety and differentiates selective GR modulators from
conventional steroids.26
6
7
8
Table 3. Hormone receptor selectivitya, b
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Cpd
GR (µM)
PR (µM)
MR (µM)
>5 (7)
AR (µM)
>10 (4)
ER
α
(µM)
ERβ (µM)
1b
0.0009
0.0002 (2)
1.76 0.63 (3)
>10 (3)
>10 (4)
>10 (2)
>10 (6)
>10 (2)
>10 (3)
>10 (3)
>5 (2)
>10 (3)
>10 (4)
>10 (2)
>10 (6)
>10 (2)
>10 (3)
>10 (3)
>5 (2)
15i
15j
15k
15m
2
0.0015
0.0005 (5)
>10 (5)
8.28 1.45 (3)
>10 (6)
>10 (6)
>10 (3)
>10 (7)
>10 (3)
>10 (4)
>10 (3)
0.0010
0.0011 (6)
0.0008
0.0001 (6)
>10 (6)
0.0009
0.0007 (3)
>10 (3)
>10 (3)
0.0038c
( 0.0019) n=7
0.028 0.0065
(6)
0.014 0.010
(6)
0.71 0.17 (4)
>10 (3)
3
0.0009
0.0009 (333)
0.021 0.014
(6)
0.149 0.166
(4)
FF
0.0008
0.0003 (5)
0.021 0.007
(4)
0.166 0.016
(2)
1.88 0.28 (2)
avalues as mean SD (number of replicates). bfor assay descriptions refer to the
Supplementary Information. cassessed in Fluorescence Polarization assay format. FF;
fluticasone furoate.
The primary goal of the follow up program was to identify a candidate with a differentiated
PK profile leading to improved lung targeting. For the shortlisted examples of the subseries
15, rapid clearance from circulation was similar to that observed for the first drug candidate
1b. Furthermore, compounds 15i – 15m demonstrated much reduced volumes of distribution
(Vss), shortening half-lives substantially. Most critically, the significant bioavailability
observed for 1b did not translate to 15i – 15m, confirming that this subclass did have
improved PK properties, resulting in reduced systemic exposure (Table 4).
Table 4. Plasma PK in rata
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 47
1
2
3
4
5
6
7
8
9
Cpd
Rat Hep CLint
Pred. in vivo CLb
CL [iv]
Vss [iv]
L/kg
23
t1/2 [iv]
(h)
F [po]
(%)
37
(µL/min/106 cells) (mL/min/kg)
(mL/min/kg)
1b
14b
15i
15j
15k
15m
3
13
22
31
25
25
34
28
34
43
48
45
45
50
46
38
42
4.7
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
10
7.4
11
102
68
2.8
0.3
<0.1
0.16
0.3
0.97
0.81
0.16
31
0.2
73
0.1
30
0.1
<0.1
<0.1
103
3.3
aMale Han Wistar rat; bPredicted hepatic clearance in the rat, based on intrinsic clearance in rat hepatocytes19
For all selected examples with the generic structure 15, the combined analysis of the
pharmacokinetic profiles after intravenous and inhaled administration indicated an absorption
rate-limited elimination.7a, 10d This was demonstrated by the prolonged terminal lung half-
lives observed after DPI (Table 5), which were invariably 1-2 orders of magnitude larger than
the corresponding plasma half-lives obtained after iv administration (Table 4). The underlying
mechanism is likely a dissolution-limited absorption, because experimental solubility in
human intestinal fluid (HIF), rather than lipophilicity, correlates with the observed lung
retention (MRT) (Table 5). Slow particle dissolution, particularly in the less-perfused regions
of the central lung, was previously shown to drive a favorable concentration gradient from the
lung epithelium to the circulation, improving the lung selectivity.27
The lipophilic examples 1b and 14b displayed similar terminal half-lives in lung and plasma
after DPI and iv administration respectively (Table 4 and 5). This indicates that for these
compounds only a transient lung selectivity was achieved through the inhalation route.
Table 5. Inhalation PK in rata
ACS Paragon Plus Environment

Page 21 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
c
c
Cpd
Sol [HIF]b
(µM)
12
% dose in lungc
AUC0-∞
t1/2
MRTc
(h)
Cmax [plasma]
at 4h (%)
at 24h (%)
(µM*h)
18
(h)
7.5
8.9
179
4.9
9.0
17
(nM)
122
30
1b
14bd
15i
15j
15k
15m
3
0.7
6.2
93
0 - 0.4
0.8
1.5
8.8
210
2.6
10
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
NT
12
0.30
2.1
91
5200
26
4.9
21
4.1
41
0.18
8.6
1.30
0.60
0.62
190
740
200
12
73
32
23
7.4
24
43
1.9
4.5
6.0
aDry powder inhalation; normalized lung dose 0.1 µmol/kg. bThermodynamic solubility in
human intestinal fluid at 37°C. cIn total lung; t1/2, apparent terminal half-life in the total lung;
MRT, mean residence time. dit instillation of the lung dose. NT, not tested.
The inhaled anti-inflammatory potential of the compounds 15i, 15k and 15m was evaluated in
the rat model of Sephadex induced airway inflammation. Compound 15j shares the generic
structure 15 but displayed similar lung PK properties to the first candidate 1b and was
included to allow a pharmacological assessment of the chemotype.
Table 6. PD effects in rat model of Sephadex induced airway inflammation
Sephadex-induced airway inflammation model (rat)
Cpd
Rat PBMCa
Lung edema
Thymus weight
involution ED25
Therapeutic ratio
ED25 / ED50
Exp.
nb
inhibition ED50
nM (% eff.)c
0.14 (78)
µg/kg (CI 95%)d
1.3 (0.30, 5.6)
µg/kg (CI 95%)d
4.6 (1.7, 12)
(CI 95%)c
1b
3.7 (0.64, 21)
2
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 47
1
2
3
4
5
6
7
8
9
15i
15j
15k
15m
3
0.34 (95)
0.29 (92)
0.10 (100)
0.30 (95)
0.14 (81)
2.4 (0.60, 9.3)
>130 (NA)
12 (7.1, 19)
27 (15, 46)
40 (17, 92)
3.9 (1.9, 8.1)
>54 (NA)
1
1
1
1
3
8.7 (4.7, 16)
2.7 (0.96, 7.6)
0.8 (0.45, 1.5)
6.0 (1.5, 23)
1.4 (0.61, 3.1)
10 (3.0, 33)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
50 (18, 140)
0.7 (0.14, 3.1)
aValue as mean of at least two experiments. bnumber of experiments; each experiment
included a minimum of 18 treated animals, divided between 3-7 dose groups with at least 4
animals in each dose group, and 12-20 untreated animals, divided betweet 2 groups: native
and challenged. cEfficacy, maximum response relative to Dexamethasone (100 %). dCI, 95%
confidence interval.
All of the profiled indazole ethers were potent with ED50 values below 10 µg/kg in the rat
Sephadex PD model, comparable to, or even superior to the benchmark ICS 3. Neither
potency in rat PBMCs nor total compound levels in rat lung (AUC) would predict the same
efficacy ranking as observed in the PD model, which is likely to be a complex result of these
and additional parameters. However, the systemic GR effect, measured by thymus involution,
correlated with peak levels of total drug in blood (Table 5 and 6). As a consequence, the
observed therapeutic ratio (TR) is largely dominated by the lung retention of the drug.
Based on the results obtained in the rat Sephadex model, the pyridylmethyl amide ligands
(15j, 15k) were parked due to inferior potency and therapeutic ratio. Nonetheless, this
matched pair with very similar molecular, structural and in vitro pharmacological properties
demonstrated that divergent solid-state properties do translate into differentiated inhalation
PK and PD. Ultimately, 15i which displayed the highest therapeutic ratio was discarded in
favor of 15m which most efficiently inhibited rat lung edema. This selection was driven by a
theoretical safety risk of 15i associated with long retention exceeding 24 hours in lung of
ACS Paragon Plus Environment

Page 23 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
presumably undissolved material that could be compounded by repeat dosing. The therapeutic
ratio obtained was 70-fold higher for 15m than for 3. The experimental settings presented
here do not allow a differentiation between the pharmacokinetic and mechanistic drivers of
the observed therapeutic ratio.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Dose-response study of 15m in a 4-day rat model of Sephadex-induced airway
inflammation. The anti-inflammatory effect was assessed as inhibition of the induced lung
edema, determined as reduction of left lung lobe weight (A). The systemic GR agonism was
assessed as thymus weight reduction (involution) in the same animals (B). The data points
shown represent single animals, the bars indicate the mean and the standard deviation. DPI
inhalation with 15m (black dots) reduced the edema by 39% to 109% (p<0.001) with a
significant effect already at the lowest dose (0.6 µg/kg, p<0.002). Significant effects of 15m
were observed on thymus weight only at the highest dose (47.9 µg/kg, p=0.041). Two doses
of 3 (grey dots) were included as control.
The slow dissolution of inhaled 15m, resulting in slow absorption into the systemic
circulation, could potentially prolong the anti-inflammatory effect in the airways. To
experimentally address the duration of the effect, the Sephadex rat PD model was modified in
two ways. The time span between administration of the drug and Sephadex-challenge was
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 47
1
2
3
4
5
6
7
8
9
increased to either 9 or 24 hours to investigate the potential for once-daily treatment. This
experimental design did not support multiple dosing and a single dose was given (DPI). To
obtain significant results, the dose was increased to 30 µg/kg (approx. 3-fold over ED90 of
15m). As a head-to-head benchmark, an approximately equal dose of the ICS 3 was profiled
in the same experiments.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Compound 15m achieved full inhibition when pre-dosed 9 h prior to challenge and still
preserved a significant effect, reducing lung edema by 76%, when dosed 24 h prior to
challenge. In this model, 15m demonstrated superiority over 3, which showed only partial
effect when pre-dosed 9 h and had no significant effect when dosed 24 h prior to challenge.
The superior performance of 15m compared with 3 is not necessarily related to the longer
sustained drug release into the lung tissue but could be explained by better potency in the rat
model. Nonetheless, the 10-fold lower systemic activity observed for inhaled 15m suggests
that equal or higher doses than used for 3 might be applied without increasing the risk of
adverse effects.
Figure 5: Duration of effect for inhibiting Sephadex-induced lung edema. 15m (black dots)
showed full inhibition effect when predosed 9 h prior to challenge (p<0.001) and 76 %
inhibition when dosed 24 h prior to challenge (p<0.001). 3 (grey dots) inhibited the challenge-
ACS Paragon Plus Environment

Page 25 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
induced lung edema by 61 % when dosed 9 h prior to challenge (p<0.01) but no significant
inhibitory effect when dosed 24 h prior to challenge.
6
7
8
9
As compound 15m displayed only a moderate effect on thymus size in rat, it can be concluded
that the distribution of 15m and any potential active metabolites to the thymus is less efficient
compared to FP. This can be rationalized by the improved lung PK profile of 15m which led
to a slower absorption and subsequently lower plasma concentrations of the parent. In
addition, the higher polarity of 15m results in a drastically reduced volume of distribution (Vss
0.16 L/kg), reducing peripheral tissue affinity. Coupled to this is the rapid clearance of 15m
from circulation by efficient degradation and elimination. To investigate the metabolic fate of
the parent compound, 14C-radiolabelled 15m was prepared and profiled in human, rat and dog
hepatocytes. Multiple degradation pathways were found to be active, generating a complex
mixture of metabolites which was similar across the three species. Hepatic degradation
proceeded predominantly by oxidation of the heterocycle of the benzodioxane moiety,
followed by ring opening and glucuronidation. Another major pathway led to the cleavage of
the indazole ether pharmacophore to release a 5-hydroxy-indazole fragment (Supplementary
information, Table S4). Based on our understanding of the series SAR, it is very likely that
GR activity is eliminated or greatly reduced in all major metabolites. Furthermore, unlike the
first candidate 1b, the risk of generating a species-specific active metabolite in a preclinical
safety species was successfully mitigated. In summary, Compound 15m combined excellent
potency and a favorable therapeutic ratio in rat and was nominated as a second clinical
candidate, 3-(5-((1R,2S)-2-(2,2-difluoropropanamido)-1-(2,3-dihydrobenzo[b][1,4]dioxin-6-
yl)propoxy)-1H-indazol-1-yl)-N-((R)-tetrahydrofuran-3-yl)benzamide (AZD7594).28
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
To conclude the structural investigation, the stereoisomers of candidate 15m were prepared
and tested for transrepression activity (Table 7). In the central 1-aryl-2-amino-propanoxy
moiety the (1R,2S) configuration is clearly preferred, whilst the configuration of the 3-
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 47
1
2
3
4
5
6
7
8
aminotetrahydrofuran had no influence on potency. For the other stereoisomers only the anti-
stereoisomeric pair with (1S,2S) configuration retained limited activity with the repression
potency reduced by more than 100-fold.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 7. Transrepression activity of stereoisomers of 15m
1R
2S
3'R
15m
Cpd
15m
15l
18
19
20
21
22
23
Configuration 1R,2S,3’R
1R,2S,3’S
1S,2R,3’R 1S,2R,3’S 1S,2S,3’R 1S,2S,3’S
1R,2R,3’R
>1000 (2)
1R,2R,3’S
>1000 (2)
Cell TR IC50
%-Efficacy
0.055
0.021 (4)
0.051
0.004 (2)
>1000 (2)
6.6
>1000 (2)
9.3
12.7 4.0
(2)
24.3 5.3
(2)
107
105
100
99
35
30
Conclusion
A class of highly potent, non-steroidal GR modulators, highly selective across the steroid
hormone receptor family, served as the starting point for optimization into two inhaled drug
candidates for the treatment airway inflammation. Initially, a soft drug strategy was developed
to ensure rapid clearance from systemic circulation. This approach delivered the first inhaled
candidate 1b which was more-rapidly absorbed than the benchmark ICS 3 (fluticasone
propionate), and possessed high availability in lung tissue during the first two hours following
inhaled administration to the rat. Compared to 3, compound 1b demonstrated marginally
improved potency in inhibition of lung edema in a rat model of airway inflammation.
Further optimization increased lung targeting: prolonging exposure in lung tissue whilst
limiting systemic levels. A revised strategy aimed to retard drug absorption via a
differentiated profile featuring decreased dissolution rate, increased polarity and increased
ACS Paragon Plus Environment

Page 27 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
molecular size. This delivered the second inhalation candidate drug 15m which achieved the
requisite low solubility profile but retained the potency of the first candidate and
demonstrated a robust local anti-inflammatory effect in rat lung airways at the same dose. The
prolonged retention in lung tissue was balanced by limited systemic exposure, minimizing
GR-driven effect on rat thymus. As a consequence, the therapeutic ratio of 15m was increased
over the first candidate 1b and the comparator ICS 3 by an order of magnitude. Head-to-head
with 3 at equivalent doses, 15m displayed more enduring efficacy, indicating potential for
once-daily dosing.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Although the development of 1b was halted due to lack of efficacy in a phase 2 study in
COPD patients,29 15m has recently demonstrated dose-dependent efficacy in a phase 2 lung
function study in asthmatics.30
EXPERIMENTAL SECTION
Synthesis. General conditions. Commercial solvents and reagents were used without
further purification. Anhydrous solvents were dried over molsieves (4Å). Water was filtered
and demineralized (Milli-Q). Purification by flash chromatography was carried out using
silica gel 60 (0.040-0.063 mm, Merck, or using prepacked Biotage KP-SIL or Isolute
columns. Purification by HPLC was performed on a Gilson HPLC system with the following
columns: Kromasil KR-100-5-C18 (250 x 20 mm), Kromasil 100-10-C18 (250 x 50 mm);
gradient of acetonitrile / water (containing 0.1% TFA). Alternatively XBridge Prep C18, 5
(150x30 mm) or XTerra Prep MS C18, 5 m (50x19 mm); gradient of acetonitrile / water
containing 2 mL 28% aq. NH4OH. In NMR spectra the chemical shifts are internally
ꢀm
ꢀ
δ
referenced to the residual solvent peak. Low resolution LCMS (APCI); Column Waters
Symmetry C18 2.1 x 30 mm; flow: 0.7 mL/min; eluents A H2O, B acetonitrile with both
solvents containing 0.1% TFA; gradient 15% to 95% B in 2.7 min, 95% B for 0.3 min; UV
detection at 254 nM. Purity was determined by HPLC: Column Kromasil C18, 5µm 3.0 x 100
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 47
1
2
3
4
5
mm; eluents A H2O, B acetonitrile with both solvents containing TFA (0.1%); gradient B 10-
100% in 20 min; flow 0.6 mL/min; detection at 220, 252 and 280 nM. All test compounds
6
7
8
were of
≥95% purity. Chiral purity was determined using the following systems: Chiral SFC:
9
LUX Amylose-1 (AD) 3µm, 0.46 x 15 cm; 40% EtOH / DEA (100/0.5) in CO2, 120 bar, 3.5
mL/min. diode array detection 215-330 nm. Chiral HPLC, method A: Chiralpak IA 5µm;
method B: Chiralpak IB 5µm; both method A and B 4.6x250mm, eluent i-hexane / DCM /
MeOH 68:30:2; flow 1 mL/min; UV=254 nm. Chiral HPLC, method C: Chiralpak IB 5µm,
4.6x100 mm; eluent iso-Hexane / DCM / MeOH / EtOH 65:29:2:4, flow 1 mL/min; UV=254
nm. Chiral HPLC, method D: Chiralpak AS-H 5 µm, 4.6 x 50mm, eluent i-hexane / i-PrOH
95:5, flow 1 mL/min; UV=220 nm. Chiral HPLC, method E: Chiralpak AD-H 5 µm, 4.6 x
250 mm, eluent i-hexane / EtOH 80:20, flow 1 mL/min, UV=220 nm.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2,2,2-Trifluoro-N-((1R,2S)-1-((1-(4-fluorophenyl)-1H-indazol-5-yl)oxy)-1-(3-
methoxyphenyl)propan-2-yl)acetamide (1b). Prepared essentially as described for 13b,
starting from 11b (21.5 g, 55 mmol), triethylamine (23 mL, 165 mmol) and trifluoroacetic
anhydride (23 mL, 165 mmol). Flash chromatography on silica gel (20% EtOAc in isohexane)
afforded 1b (22.0 g, 82%) as a yellow foam. The material was crystallized from MTBE /
isohexane, assisted by seed crystals, to yield 1b as a white solid (14.0 g, 52%), mp = 82 °C.
20
LCMS m/z 488 (MH+). [
MHz, DMSO-d6)
α]
D -44° (c 1.1, MeOH). HPLC tR 14.23, >99%. 1H NMR (600
δ
9.52 (d, J = 8.3 Hz, 1H), 8.17 (s, 1H), 7.71 – 7.76 (m, 2H), 7.69 (d, J = 9.2
Hz, 1H), 7.35 – 7.42 (m, 2H), 7.26 (t, J = 7.9 Hz, 1H), 7.21 (dd, J = 2.3, 9.2 Hz, 1H), 7.13 (d,
J = 2.3 Hz, 1H), 6.99 (d, J = 7.7 Hz, 1H), 6.96 (s, 1H), 6.84 (dd, J = 2.4, 8.2 Hz, 1H), 5.27 (d,
J = 6.3 Hz, 1H), 4.21 – 4.3 (m, 1H), 3.72 (s, 3H), 1.33 (d, J = 6.8 Hz, 3H). 13C NMR (151
MHz, DMSO-d6) δ 160.20 (d, J = 243.6 Hz), 159.10, 155.50 (q, J = 36.3 Hz), 152.74, 139.66,
136.07 (d, J = 2.6 Hz), 135.01, 134.09, 129.41, 125.26, 123.92 (d, J = 8.5 Hz), 119.83,
118.91, 116.34 (d, J = 22.8 Hz), 115.79 (q, J = 288.4 Hz), 113.15, 112.46, 111.41, 104.21,
ACS Paragon Plus Environment

Page 29 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
81.23, 54.89, 50.68, 15.15. Chiral HPLC (method E) tR = 5.5 min, 100% (1R,2S enantiomer),
tR = 7.1 min, 0% (1S,2R-enantiomer). Chiral Purity >99.9 % ee.
6
7
8
9
(S)-Tert-butyl 1-(3-methoxyphenyl)-1-oxopropan-2-ylcarbamate (5b). Essentially
prepared as described for 5c, starting from (3-methoxyphenyl)magnesium bromide, 1M in
THF (276 mL, 276 mmol). Yield after recrystallisation from methyl tert-butyl ether (490 mL)
58.9 g (84%) as a colorless crystalline solid. LCMS m/z 180 (MH+-BOC). 1H-NMR (400
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
MHz, CDCl3): δ 7.55 (d, J = 7.6 Hz, 1H), 7.49 (t, J = 1.9 Hz, 1H), 7.39 (t, J = 8.0 Hz, 1H),
7.15 (dd, J = 8.2, 2.0 Hz, 1H), 5.56 (d, J = 6.9 Hz, 1H), 5.28 (quint, J = 7.2 Hz, 1H), 3.86 (s,
3H), 1.47 (s, 9H), 1.41 (d, J = 7.1 Hz, 3H). Chiral HPLC (method D) tR = 2.36 min, 99.95%
(S-enantiomer), tR = 1.70 min, 0.05% (R-enantiomer). Chiral Purity 99.9 % ee.
(S)-Tert-butyl 1-(2,3-dihydrobenzo[b][1,4]dioxin-6-yl)-1-oxopropan-2-ylcarbamate (5c).
Magnesium turnings (0.46 g, 18.9 mmol) and one small iodine crystall were suspended in dry
THF (3 mL) under an argon atmosphere. A solution of 6-bromo-2,3-
dihydrobenzo[b][1,4]dioxine (4.03 g, 18.7 mmol) in dry THF (17 mL) was added. The
mixture was gently heated to initiate the reaction, after a few minutes the iodine color
vanished, the external heating was removed and continued stirring for one hour resulted in a
green solution. GC/MS analysis of a small sample, quenched with iodine in THF, indicated a
conversion of 75% of the starting bromide to the Grignard reagent to form a solution of (2,3-
dihydrobenzo[b][1,4]dioxin-6-yl)magnesium bromide in THF (0.7 M, 20 mL, 14.0 mmol) that
was used directly in next step. In a separate vessel, a suspension of (S)-tert-butyl 1-
(methoxy(methyl)amino)-1-oxopropan-2-ylcarbamate (4) (3.0 g, 12.9 mmol) in THF (30 mL)
was cooled to -15 ºC and isopropyl magnesium chloride (2 M in THF, 6.5 mL, 13.0 mmol)
was added keeping the temperature below -10 ºC. The solid started to dissolve, the reaction
temperature was allowed to reach 0 ºC and the freshly prepared solution of (2,3-
dihydrobenzo[b][1,4]dioxin-6-yl)magnesium bromide, 0.7 M in THF (20 mL, 14.0 mmol)
ACS Paragon Plus Environment