Discovery and Evaluation of Clinical Candidate AZD3759, a Potent, Oral Active, Central Nervous System-Penetrant, Epidermal Growth Factor Receptor Tyrosine Kinase Inhibitor

Article abstract of DOI:10.1021/acs.jmedchem.5b01073

Recent reports suggest that an increasing number of patients with lung cancer, especially those with activating mutations of the epidermal growth factor receptor (EGFR), also present with brain metastases and leptomeningeal metastases. These patients have poor prognosis as there are no approved drugs for these indications. Available agents have poor efficacy for these patients even at well above their standard dose. Herein, we report the discovery of (4-[(3-chloro-2-fluorophenyl)amino]-7-methoxyquinazolin-6-yl (2R)-2,4-dimethylpiperazine-1-carboxylate 1m (AZD3759), an investigational drug currently in Phase 1 clinical trial, which has excellent central nervous system penetration and which induces profound regression of brain metastases in a mouse model.

Full text of DOI:10.1021/acs.jmedchem.5b01073

Subscriber access provided by UNIV OF NEBRASKA - LINCOLN
Article
Discovery and Evaluation of Clinical Candidate AZD3759, a
Potent, Oral Active, Central Nervous System-Penetrant, Epidermal
Growth Factor Receptor Tyrosine Kinase Inhibitor (EGFR TKI)
Qingbei Zeng, Jiabing Wang, Ziqiang Cheng, Kan Chen, Peter Johnström,
Katarina Varnas, David Yunzhi Li, Zhen Fan Yang, and Xiaolin Zhang
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.5b01073 • Publication Date (Web): 27 Aug 2015
Downloaded from http://pubs.acs.org on August 28, 2015
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 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Discovery and Evaluation of Clinical Candidate
AZD3759, a Potent, Oral Active, Central Nervous
SystemꢀPenetrant, Epidermal Growth Factor
Receptor Tyrosine Kinase Inhibitor (EGFR TKI)
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
Qingbei Zeng1, Jiabing Wang1, Ziqiang Cheng1, Kan Chen1, Peter Johnström2,3, Katarina
Varnäs3, David Yunzhi Li1, Zhen Fan Yang1, Xiaolin Zhang1,*
1Innovation Center China, Asia & Emerging Markets iMed, AstraZeneca Innovative Medicines
and Early Development, 199 Liangjing Road, Zhangjiang HiꢀTech Park, Shanghai 201203,
China
2AstraZeneca Translational Centre, Personal Healthcare and Biomarkers iMed, AstraZeneca
R&D, Karolinska Institutet, Department of Clinical Neuroscience, R5:U1 Karolinska University
Hospital, SEꢀ171 76 Stockholm, Sweden
3Karolinska Institutet, Department of Clinical Neuroscience, Centre for Psychiatry Research,
R5:02 Karolinska University Hospital, SEꢀ171 76 Stockholm, Sweden
ABSTRACT: Recent reports suggest that an increasing number of patients with lung cancer,
especially those with activating mutations of the epidermal growth factor receptor (EGFR), also
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 45
1
2
3
4
5
6
7
8
9
present with brain metastases (BM) and leptomeningeal metastases (LM). These patients have
poor prognosis as there are no approved drugs for these indications. Available agents have poor
efficacy for these patients even at well above their standard dose. Herein we report the discovery
of (4ꢀ[(3ꢀchloroꢀ2ꢀfluorophenyl)amino]ꢀ7ꢀmethoxyquinazolinꢀ6ꢀyl (2R)ꢀ2,4ꢀdimethylpiperazineꢀ
1ꢀcarboxylate 1m (AZD3759), an investigational drug currently in Phase 1 clinical trial, which
has excellent central nervous system (CNS) penetration and which induced profound regression
of brain metastases in a mouse model.
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
Nonꢀsmall cell lung cancer (NSCLC) accounts for more than 85% of all lung cancer cases and is
one of the leading causes of cancer death worldwide.1 It is estimated that in 2014 there were
224,210 new lung cancer cases and 159,260 NSCLC related deaths in the United States alone.2
Genetic aberrations in the tyrosine kinase domain of epidermal growth factor receptor have been
identified as one of the key drivers of NSCLC progression.3 EGFR kinase inhibitors, such as
gefitinib, erlotinib, and afatinib (Scheme 1), deliver significant patient benefit and have been
approved for the treatment of advanced NSCLC with EGFR activating mutations.4ꢀ7
An increased incidence of CNS metastases has been reported as patients live longer with
treatment and with the improvement of imaging technology and routine screening.8,9 Studies
have suggested that around 40ꢀ50% of patients with lung cancer would develop CNS metastases,
including brain metastases and leptomeningeal metastases, in their life span.10,11 Currently
approved drugs can neither effectively treat lung cancer CNS metastases nor prevent
development of the metastases at standard dosing due to their limited bloodꢀbrainꢀbarrier (BBB)
penetration.12,13 For patients with symptomatic brain metastases, selective use of radiation
therapy and surgery are still the mainstay of treatment but provide limited benefit.14,15 Given the
ACS Paragon Plus Environment
2

Page 3 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
effectiveness of EGFR TKIs in treating tumors in the lung, there is a major unmet medical need
for an agent which can effectively penetrate the CNS and treat tumors in this tissue.
Scheme 1. Approved EGFR Kinase Inhibitors
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
Cl
F
HN
N
O
O
O
N
N
Gefitinib
ꢀ RESULTS AND DISCUSSION
Discovery of compound 1m. In order to achieve efficacy in the CNS and periphery, it is
necessary to design biologically active agents with sufficient free drug exposures in both
environments. Initially, we started with the clinically proven EGFR TKI gefitinib, developed by
AstraZeneca, as the lead for the project with the objective of developing a CNS penetrable
candidate with potency and selectivity equal to or better than gefitinib.
It has been a challenging task to deliver a pharmaceutical agent that can enter the CNS.16 Efflux
transport is a major determinant of drug disposition to the CNS. ATPꢀbinding cassette (ABC)
export proteins such as Pꢀglycoprotein (Pgp) and breast cancer resistance protein (BCRP) are
localized in the luminal membrane of brain microvessel endothelial cells.17,18 Pgp and BCRP
have been shown to play an important role in limiting distribution of various drugs into the CNS.
We examined several key parameters related to BBB penetration for approved EGFR inhibitors
(see Supporting Information for experimental protocols). Table 1 summarizes physicochemical
properties, unbound brain to unbound blood ratios (Kpuu,brain), cerebrospinal fluid (CSF) to
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 45
1
2
3
4
5
6
7
8
9
unbound blood ratios (Kpuu,CSF), and efflux ratios of these compounds in MDCKIIꢀMDR1 and
MDCKIIꢀBCRP permeability experiments. All these agents are either Pgp (MDR1) or BCRP
substrates such that their Kpuu,brain or Kpuu,CSF values are low. Unfavorable rotatable bonds in
gefitinib (10 rotatable bonds) and erlotinib (11 rotatable bonds) and undesired hydrogen bond
donors in afatinib (2 hydrogen bond donors) are likely responsible for the low CNS penetration
of these agents. Recently published papers highlighted that the correct balance of permeability,
potential for active efflux, and physicochemical properties, is needed to achieve good drug
partitioning and distribution into CNS with CNS active drugs having lower molecular weight,
fewer rotatable bonds, smaller polar surface area (PSA), and fewer hydrogen bond donors than
nonꢀCNSꢀpenetrant drugs.19ꢀ23
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. Physicochemical Properties, Permeability and CNS Penetration Properties of Approved
EGFR Kinase Inhibitors
Gefitinib
446.7
62.6
10
Erlotinib
393.4
70.3
11
Afatinib
485.9
86.1
9
Molecular weight
tPSA (Å2)
Rotatable bondsa
Hydrogen bond donor number
MDCKIIꢀ
1
1
2
Papp, A to B (10ꢀ6 cm/s)
8.3
3.0
19
1.7
14
2.5
11
MDR1b
Efflux ratioc
Papp, A to B (10ꢀ6 cm/s)
Efflux ratioc
4.2
1.3
MDCKIIꢀ
BCRPd
18
5.4
4.5
2.9
0.11
0.29
59
fu,blood (%)e
fu,brain (%)f
3.3
4.1
0.39
0.021
0.088
0.54
0.0066
0.27
g
Kpuu,brain
h
Kpuu,CSF
aRotatable bond number is calculated by the equation of Nrot = Nrot,ac + Σ(size ꢀ 4 ꢀ Nring,ac ꢀNfused), where Nrot is rotatable bond number in the
molecule, Nrot,ac is the number of rotatable acyclic bonds, Size is the size of the nonꢀaromatic ring, Nrig,cyc is the number of rigid bonds within the
ring and Nfused is the number of bonds which belong to more than one ring system. bPermeability was estimated by using MDR1ꢀtransfected
MDCKII cells seeded onto polyethylene membranes in 96ꢀwell BD insert systems with test compounds at 1 µM. cEfflux ratio = Papp,B to A/ Papp,A to
.
B
dPermeability was estimated by using BCRPꢀtransfected MDCKII cells seeded onto polyethylene membranes in 96ꢀwell BD insert systems
ACS Paragon Plus Environment
4

Page 5 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
with test compounds at 1 µM. ). eFraction of unbound drug in blood was obtained with the test compound at 5 µM by equilibrium dialysis.
fFraction of unbound drug in brain homogenate was obtained with the test compound at 5 µM by equilibrium dialysis. gRatio of unbound AUC0ꢀ
24h in rat brain to that in blood at the clinically relevant dose of EGFR TKIs. hRatio of AUC0ꢀ24h in rat CSF to unbound AUC0ꢀ24h in blood at the
clinical relevant dose of EGFR TKIs, see Supporting Information for more details.
6
7
8
9
Achieving these desirable physicochemical properties while retaining good EGFR TKI potency
was the challenge. Xꢀray structure analysis of the gefitinibꢀEGFR complex indicated that the
quinazoline positions a hydrogen bond acceptor to Met793 in the hinge region of EGFR and the
anilino substituent extends into the hydrophobic pocket at the back of the ATPꢀbinding cleft.24
These structural elements, therefore, are essential to EGFR inhibitor activity and hence we
decided to retain this template but to reꢀengineer the quinazolineꢀbased molecules to deliver the
physicochemical properties needed for CNS penetration, initially by reꢀdirecting the fluoro
moiety present in the phenyl group from the para to ortho position (Scheme 2). The previous
SAR study at AstraZeneca indicated that reꢀpositioning fluorine atom present in gefitinib from
the para position to the ortho position maintained EGFR kinase inhibitory potency.25 It was
believed that orthoꢀfluorine atom could partially mask the hydrogen bond donating effect of the
NꢀH by forming an intramolecular interaction.26ꢀ28 Additionally, the linear and rotatable side
chain present in gefitinib was shortened to derive molecules with the general structure 1
(Scheme 2).
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 2. Design Strategy Starting from Gefitinib as an Initial Lead
The brain and CSF penetration of the agents was estimated by Kpuu,brain and Kpuu,CSF
,
respectively.29,30 Kpuu,brain is the ratio of unbound AUC in brain to unbound AUC in blood and
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 45
1
2
3
4
5
6
7
8
9
Kpuu,CSF is the ratio of AUC in CSF to unbound AUC in blood (see Supporting Information for
more details). Compared to gefitinib, compound 1a has lower molecular weight and polar
surface area (Table 2) and is also a more rigid molecule (seven rotatable bonds) than gefitinib
(ten rotatable bonds). As a result, CNS drugꢀlike properties, such as efflux ratio (0.33) and
permeability (Papp,A to B = 32 x 10ꢀ6 cm/s) were achieved. Encouragingly, this compound
exhibited inhibitory potency (IC50) at 3.8 nM and 6.6 nM in PCꢀ9 and H3255 cell line EGFR
phosphorylation assays, which were comparable to that of gefitinib (at 6.2 nM and 6.5 nM,
respectively). Furthermore, compound 1a demonstrated acceptable human CYP450 inhibition
profile (IC50 > 10 µM for CYP2C9, CYP2D6 and CYP3A4) and low hERG inhibition liability
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
(hERG IC50 = 17 µM, patch clamp assay). More importantly, the preferred CNS drugꢀlike
properties of the compound translated into much improved brain and CSF penetration (Kpuu,brain
and Kpuu,CSF were 0.47 and 0.76, respectively) and reasonable exposures in both brain and blood
(AUC0ꢀ7h values in brain and blood at 2 mg/kg were 1013 nM·h and 834 nM·h, respectively)
were achieved. These excellent in vitro potency and promising CNS penetration results were
achieved while maintaining a reasonable pharmacokinetic (PK) properties in rat (oral
bioavailability 44%, elimination halfꢀlife 1.1 hours, clearance 59 mL/min/kg and volume of
distribution at steady state 2.8 L/kg). With these encouraging results, further reduction in the
molecular weight and the number of rotatable bonds of 1a led to 1b. As expected, this
compound benefits from increased Kpuu,brain and Kpuu,CSF (0.96 and 0.93, respectively). In order to
examine the importance of the position of the fluorine atom, inhibitor 2 with a para substituent
was prepared. Comparison of 1b with 2 suggested that indeed the intramolecular interaction
between the fluoro and NꢀH in 1b is important for CNS penetration. However, agents 1b and 2
also showed substantially lower level of blood and brain AUCs compared with 1a. As
ACS Paragon Plus Environment
6

Page 7 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
subsequent attempts to modify either 1b or 2 did not lead to further improvement in PK
properties, 1b and 2 analogs were no longer pursued and optimization of 1a became the priority.
In order to improve the PK properties of 1a, the next effort was to modify the potentially labile
methylene group (Table 2). An initial attempt was to change the methylene unit into a carbonyl
moiety (1c) which could reduce potential oxidative liability and further decrease the bond
rotation. Unfortunately, this compound (1c) was not a CNS penetrant and was much less potent
in a cell based phosphorylation assay (L858R pEGFR IC50 = 166 nM). However, an extension
of the amide present in 1c by one atom to give the carbamate (1d) significantly improved
potency (L858R pEGFR IC50 = 4.1 nM), although the exposures of 1d in brain and plasma were
very low (Table 2). The poor equilibrium solubility (12 nM in pH 7.4 buffer) of the compound
was identified as a potential limiting factor for oral exposure. Thus, further modification were
undertaken to replace the oxygen present in the morpholine ring at the terminal position with a
basic nitrogen (1e). Indeed, the solubility of 1e (free base) was increased substantially (202 mM
in neutral buffer). With this increase in basicity (pKa = 8.4) and solubility, the exposure of 1e in
brain and blood improved dramatically.
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 2. Results from Optimizing Physicochemical Properties
Compounds
Gefitinib
--
1a
1b
1c
1d
1e
2
--
R =
Molecular weight
446.7
402.9
388.8
416.8
432.8
445.9
388.8
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 45
1
2
3
tPSA (Å2)
62.6
10
53.5
7
53.2
6
71.8
6
80.9
6
73.3
6
53.2
6
4
5
Rotatable bondsa
6
7
8
9
L858R pEGFR IC50 (nM)b
MDCKIIꢀMDR1 Papp, A to B (10ꢀ6 cm/s)c
Efflux ratiod
6.5
6.6
1.5
32
166.0
28
4.1
42
5.0
28
1.5
29
8.3
32
3.0
0.33
1013
50
0.29
85
2.9
0.58
104
5
0.37
825
31
0.18
85
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
Brain AUC0ꢀ7h (nM·h)e
CSF AUC0ꢀ7h (nM·h)e
Blood AUC0ꢀ7h (nM·h)e
fu,blood (%)g
5647f
93f
101
46
2
1
31913f
3.3
834
7.8
33
1627
17
42
722
8.4
2.5
0.34
0.52
77
5.8
2.2
0.96
0.93
3.8
2.6
1.7
3.4
2.2
0.83
0.42
0.79
fu,brain (%)h
0.39
0.021j
0.088l
3.0
8.8
i
Kpuu,brain
0.47
0.76
0.03
0.16
k
Kpuu,CSF
aRotatable bond number is calculated by the equation of Nrot = Nrot,ac + Σ(size ꢀ 4 ꢀ Nring,ac ꢀNfused), where Nrot is rotatable bond number in the
molecule, Nrot,ac is the number of rotatable acyclic bonds, Size is the size of the nonꢀaromatic ring, Nrig,cyc is the number of rigid bonds within the
ring and Nfused is the number of bonds which belong to more than one ring system. bCellular phosphorylation assay on H3255 (L858R) cell lines
c
(n = 2). MDR1ꢀtransfected MDCKII cells with test compound at 1 µM were used to estimate apical to basolateral permeability. dEfflux ratio =
P
app, B to A/ Papp, A to B.
eMale Han Wistar rats (PO, 2 mg/kg; 1% MC). fClinically equivalent dose of 20 mg/kg was orally administrated for CNS
penetration assessment of gefitinib and exposure was measured between 0.5 and 24 hr. gFraction of unbound drug in blood was obtained with the
test compound at 5 µM by equilibrium dialysis. hFraction of unbound drug in brain homogenate was obtained with the test compound at 5µM by
equilibrium dialysis. iRatio of unbound brain AUC0ꢀ7h to unbound blood AUC0ꢀ7h. Ratio of unbound brain AUC0ꢀ24h to unbound blood AUC0ꢀ24h.
j
kRatio of CSF AUC0ꢀ7h to unbound blood AUC0ꢀ7h. Ratio of CSF AUC0ꢀ24h to unbound blood AUC0ꢀ24h
.
l
Given its promising in vitro potency and good brain penetration, a rat PK study was conducted
for compound 1e and the results are summarized in Table 3. Compared to 1a, inhibitor 1e
significantly improved clearance (22 mL/min/kg) in rat. It was quickly absorbed and the oral
bioavailability was 45% while the elimination halfꢀlife was increased to 2.6 h. This improved
halfꢀlife, however, did not meet our design criteria (>4 h in rat).
Table 3. Rat PK Data for Compound 1ea:
CL
AUC0ꢀ24h
Dose (route)
Formulation
F (%)c
45
T1/2 (h)d
Tmax (h)e
0.25
ꢀꢀ
Cmax (nM)f
2020
(mL/min/k
Vss (L/kg)h
(nM·h)b
g)g
5 mg/kg
(PO)
1%
3905
ꢀꢀ
ꢀꢀ
ꢀꢀ
methylcellulose
DMSO:10%
Captisol in
1 mg/kg (iv)
1713
ꢀꢀ
2.6
ꢀꢀ
22
1.8
saline = 1:99
aMale Han Wistar rats (nonꢀfasted) were used in this study. Area under concentrationꢀtime curve from 0ꢀ24 hours. Bioavailability (%). Mean
b
c
d
e
f
g
h
elimination halfꢀlife. Time to reach maximum concentration. Peak concentration. Clearance, obtained from intravenous infusion. Volume of
distribution at steady state. All pharmacokinetic parameters were calculated with Phoenix v6.3 (Certara, St Louis, MO, USA).
ACS Paragon Plus Environment
8

Page 9 of 45
Journal of Medicinal Chemistry
1
2
3
4
Table 4. Properties of Compounds 1f-n.
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
Compounds
R =
1f
1g
1h
1i
1j
1k
1l
1m
1n
isomer 1
5
isomer 2
5
Rotatable bondsa
6
6
4
6
6
6
6
L858R pEGFR IC50
(nM)b
1.1
2.7
2.3
4.4
67
12.8
76
16.9
21.5
7.2
6.9
MDCKIIꢀMDR1 Papp,
A to B (10ꢀ6 cm/s)c
47
17
14
59
54
36
51
Efflux ratiod
0.76
2037
75
0.34
798
68
0.39
332
0.60
585f
19f
0.69
727f
28f
1.5
278
19
0.94
417
7
0.41
3692
76
0.92
1935
48
Brain AUC0ꢀ7h (nM·h)e
CSF AUC0ꢀ7h (nM·h)e
BQLg
Blood AUC0ꢀ7h
(nM·h)e
3001
2483
454
786f
740f
157
404
1231
805
fu,blood (%)h
fu,brain (%)j
5.8
2.3
4.9
2.4
13.4
1.6
2.8
1.9
2.4
1.8
3.5
1.6
6.4
1.8
4.7i
2.0i
1.3
1.3
2.3
1.5
1.6
2.6
k
Kpuu,brain
0.27
0.43
0.15
0.56
0.086
BQLg
0.51
0.90
0.74
1.6
0.84
3.5
0.28
0.28
l
Kpuu,CSF
aRotatable bond number is calculated by the equation of Nrot = Nrot,ac + Σ(size ꢀ 4 ꢀ Nring,ac ꢀNfused), where Nrot is rotatable bond number in the
molecule, Nrot,ac is the number of rotatable acyclic bonds, Size is the size of the nonꢀaromatic ring, Nrig,cyc is the number of rigid bonds within the
ring and Nfused is the number of bonds which belong to more than one ring system. bCellular phosphorylation assay on H3255 (L858R) cell lines
(n = 2). cMDR1ꢀtransfected MDCKII cells with the test compound at 1 µM were used to estimate apical to basolateral permeability. dEfflux ratio
= Papp, B to A/ Papp, A to B. .
eMale Han Wistar rats (PO, 2 mg/kg; 1% MC. fAUC0ꢀ6.5h gCompound 1h concentrations in CSF at predefined time points
were below the quantification limit (1 ng/mL). ). hFraction of unbound drug in blood was obtained with test compound at 5µM by equilibrium
i
dialysis. Average value from 4 independent tests. jFraction of unbound drug in brain homogenate was obtained with the test compound at 5 µM
by equilibrium dialysis. kRatio of unbound brain AUC0ꢀ7h to unbound blood AUC0ꢀ7h. lRatio of CSF AUC0ꢀ7h to unbound blood AUC0ꢀ7h
.
The subsequent 1e metabolite identification in rat hepatocytes suggested that piperazine
oxidation was the main metabolic pathway (see Supporting Information for more details). Thus,
introduction of substituent(s) to the piperazine ring or change of the piperazine to another moiety
was examined (Table 4). Initially, we designed compounds 1f and 1g by replacing the piperazine
with a dimethylaminopyrrolidine moiety. These pyrrolidine derivatives (1f and 1g) displayed
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 45
1
2
3
4
5
6
7
8
9
excellent potency (IC50 = 1.1 nM and 2.7 nM, respectively) in cellular EGFR phosphorylation
assays but whilst blood AUC levels of these compounds were also increased and their levels of
brain penetration were not acceptable (Kpuu,brain < 0.3). The limited brain penetration is
potentially due to the increased flexibility from the rotatable bond between the terminal
dimethylamino and pyrrolidine ring. Hence subsequently, the piperazine was substituted by a
much more rigid bridged bicyclic system. The constrained analog 1h (four rotatable bonds)
showed further reduced CNS penetration (Kpuu,brain and Kpuu,CSF < 0.1) which led us to speculate
that a balanced molecular flexibility (rotatable bond number) is needed for favorable CNS
penetration. Thus, compounds 1i-n were designed and prepared. In these compounds, the
rotatable bond numbers were retained at a range of 5ꢀ6 that is comparable to the number in 1e
while the piperazine ring was substituted by a small group. Encouragingly, most compounds
achieved Kpuu,brain and Kpuu,CSF values > 0.5. In particular, the free brain and cerebrospinal fluid
AUCs of 1m were similar to that in blood, indicating its equivalent exposures in brain, CSF and
blood.
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
With the introduction of substituents to the piperazine moiety, several compounds achieved good
CNS penetration (Kpuu,brain and Kpuu,CSF > 0.5) as well as in vitro potency. As the final selection
of the compound for progression into preꢀclinical testing would be the rat PK profile, potent
inhibitors with Kpuu,brain and Kpuu,CSF greater than 0.5 were further evaluated in rats (Table 5).
Overall, these compounds showed good oral bioavailability (>52%) except for compound 1k
(24%). Agent 1m exhibited the best pharmacokinetic profile with moderate blood clearance at
17 mL/min/kg and the volume of distribution of 5.2 L/kg. The intrinsic clearance (Clint) of 1m is
11.4 µL/min/106 cells in human hepatocytes and the predicted clearance of the agent in human is
7.7 mL/min/kg by in vitroꢀin vivo extrapolation (IVIVe)31,32. Liver blood flow (LBF) method
ACS Paragon Plus Environment
10

Page 11 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
with blood protein binding correction was applied to generate individual estimates of unbound
human hepatic clearance (CLhepatic) using available data from each preꢀclinical species.33ꢀ35 The
predicted human clearance by the LBF method from rat is 8.1 mL/min/kg which is very similar
to that derived from IVIVe. Following oral dosing in rats at 2 mg/kg, absorption of 1m was
rapid with blood Cmax of 0.58 µM achieved at 1.0 h. Subsequently, blood concentrations of 1m
declined monoꢀexponentially with mean elimination halfꢀlife of 4.3 h, which was close to the
same parameter obtained from intravenous dosing (4.1 h). The bioavailability following an oral
dose in rats was 91%. Given its promising PK properties, CNS penetration and in vitro potency,
1m was selected as the candidate for comprehensive biological profiling.
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 5. Rat PK Data of Compounds 1i-n (PO: 2 mg/kg; iv: 1 mg/kg)a:
Compounds
Oral AUC0ꢀ24h (nM·h)b
F (%)c
1i
1570
52
1j
2315
84
1k
789
24
1l
1037
92
1m
3803
91
1n
2948
83
T1/2 (h)d
0.9
28
1.3
31
1.1
26
1.3
55
4.1
17
2.4
29
CL (mL/min/kg)e
Vss (L/kg)f
2.0
2.5
1.9
3.2
5.2
2.7
aIn vivo PK was obtained via intravenous infusion route (DMSO: 10% Captisol in saline = 1:99) and oral gavage route (1% methylcellulose) in
b
c
male Wistar Han rats (nonꢀfasted). Area under concentrationꢀtime curve from 0ꢀ24 hours. Bioavailability (%). dMean elimination halfꢀlife,
obtained from intravenous infusion. eClearance, obtained from intravenous infusion. fVolume of distribution at steady state. All pharmacokinetic
parameters were calculated with Phoenix v6.3 (Certara, St Louis, MO, USA).
In vitro profile of candidate 1m. The inhibition of 1m against EGFR tyrosine kinase (EGFR
TK) wild type and mutant enzymes was tested at their corresponding Km and 2 mM ATP
concentrations (Table 6). At Km ATP concentrations, the inhibition IC50 values were 0.3, 0.2
and 0.2 nM for EGFR TK wild type, L858R mutant and Exon 19Del enzymes, respectively. In
cellular EGFR phosphorylation and proliferation studies, PCꢀ9 (Exon 19Del), H3255 (L858R),
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 45
1
2
3
4
5
6
7
8
9
and H838 (wild type EGFR) cells were used (see details in Supporting Information). The results
are shown in Table 6. Compound 1m was an equally potent inhibitor of cellular
phosphorylation or proliferation in EGFR activating mutant cell lines (PCꢀ9 and H3255 cell
lines) in a range of 7.0ꢀ7.7 nM, suggesting a high level of potency against all these clinically
relevant EGFR mutations. In cellular phosphorylation studies, 1m also demonstrated 9ꢀfold
inhibition selectivity in EGFR activating mutant cell lines over EGFR wild type cell lines (H838
cell line). This was consistent with the findings in biochemical assay when ATP concentration
was increased from Km to 2mM (the hypothesized cellular ATP concentration), suggesting ATP
competitive binding mode of 1m with its possible differential binding affinities to mutant and
wild type EGFRs. Although 1m showed some activity against pEGFR in H838 cells, as these
cells do not rely on activation of EGFR pathway for proliferation, we did not see activity of 1m
on cell proliferation of H838 cells. In contrast, 1m demonstrated inhibitory effects on both
pEGFR pathway and cell proliferation of EGFR mutation derived cells, PCꢀ9 and H3255,
suggesting these cells rely on activation of EGFR pathway for proliferation. In order to broadly
evaluate the selectivity of the compound, 1m was screened in a kinase panel and in a secondary
pharmacology panel (see details in Supporting Information). Kinase panel, screened at Millipore,
comprised of 124 recombinant protein kinases and lipid kinases. The percentage inhibition of
1m was tested at one single concentration (1 ꢁM) across each of these kinases. At this
concentration, 1m displayed less than 50% inhibition against 115 kinases and greater than 50%
inhibition against the other nine kinases including EGFR kinase in the panel. The inhibition of
those 8 off targets were 83% for EphB4, 57% for Flt, 58% for Fyn, 62% for KDR, 61% for Lck
activated, 74% for Lyn, 69% for Src, and 87% for Yes. Secondary pharmacology panel,
performed at Cerep, covered 150 distinct molecular targets. Assays were run in concentrationꢀ
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
12

Page 13 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
response mode. It was found that 147 molecular targets exhibited > 1 µM and 3 targets < 1 µM
of IC50s. These active molecular targets with less than 1 µM of IC50s were KDR (156 nM), Src
(622 nM), and D2 (797 nM). The results from these panel screening indicated that 1m was a
highly selective compound. Furthermore, 1m was neither a direct inhibitor (IC50 > 50 µM) nor a
time dependent inhibitor for CYP1A2, CYP2B6, CYP2C8, CYP2C9, CYP2C19, CYP2D6 and
CYP3A4/5 isoforms. This agent exhibited less than 3 fold of mRNA induction for CYP1A2,
CYP2B6 and CYP3A4 at the predicted maximum total concentration (0.3 µM) of the compound,
suggesting a low enzyme induction liability.36,37 Also, the candidate showed acceptable hERG
inhibition activity with IC50 of 13.3 µM in a conventional manual wholeꢀcell patch clamp study.
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 6. Compound 1m Activity in Biochemical and Cellular Assaysa
Biochemical assay
Cellular phosphorylation
Cellular proliferation
EGFR
(Exon
19Del)
IC50 at
Km of
ATP
EGFR
(Exon
19Del)
IC50 at 2
mM of
ATP
EGFR(
L858R)
IC50 at
Km of
ATP
EGFR(
L858R)
IC50 at 2
mM of
ATP
PCꢀ9
H3255
(L858R
)
pEGFR
IC50
EGFR(
wt) IC50
at Km
of ATP
(nM) (n
= 65)
EGFR(
wt) IC50
at 2 mM
of ATP
(nM) (n
= 10)
H838
(wt)
pEGFR
IC50
(nM) (n
= 13)
PCꢀ9
(Exon
19Del)
pEGFR
IC50
(nM) (n
= 12)
H3255
(L858R
) GI50
(nM) (n
= 40)
H838
(wt)
GI50
(nM) (n
= 11)
(Exon
19Del)
GI50
(nM) (n
= 40)
(nM) (n
= 65)
(nM) (n
= 10)
(nM) (n
= 12)
(nM) (n
= 14)
(nM) (n
= 5)
0.2±0.0
0.2±0.0
0.3±0.0
64.5±8.
21556±
4866
2.4±0.3
7.6±2.1
102±34
7.4±1.3
7.2±1.3
7.7±2.6
7.0±1.5
2
6
4
7
aAverage value from the n times of repeat tests.
Dog pharmacokinetics. Blood pharmacokinetic parameters of 1m in male dogs were
determined following both a single dose intravenous infusion and oral administration. These
data are summarized in Table 7. Following the IV dose in dogs, 1m blood clearance was
determined as 14 mL/min/kg and volume of distribution was 6.4 L/kg. Its elimination half life
was 6.2 h. Absorption of 1m was rapid with blood Cmax (698 nM) occurring between 0.5 h and
1.5 h. The oral bioavailability of 1m was excellent at 90%.
Table 7. Dog PK Data of 1ma
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 45
1
2
3
4
AUC0ꢀ24h
(nM·h)b
CL
Dose/route
T1/2 (h)c
F (%)d
Tmax (h)e
Cmax (nM)f
Vss (L/kg)h
(mL/min/kg)g
5
6
7
8
1 mg/kg/iv
2560
4625
6.2
5.5
ꢀꢀꢀ
ꢀꢀꢀ
ꢀꢀꢀ
14
ꢀꢀꢀ
6.4
ꢀꢀꢀ
2 mg/kg/PO
90.2
0.5ꢀ1.5
698
aIn vivo PK was obtained via intravenous infusion route (DMSO: 10% Captisol in saline =1:99) and oral gavage route (1% methylcellulose) in
9
male Beagle dog (Fed). bArea under concentrationꢀtime curve from 0ꢀ24 hours. cMean elimination halfꢀlife.. dBioavailability (%), calculated with
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
AUC0ꢀ24h and nominal dose. eTime to reach maximum concentration. fPeak concentration. Clearance. hVolume of distribution at steady state. All
g
pharmacokinetic parameters were calculated with Phoenix v6.3 (Certara, St Louis, MO, USA).
Microdosing positron emission tomography (PET) studies. Brain distribution of 1m in
monkey was investigated by microdosing positron emission tomography (PET) to substantiate
the PK data in rats.38 Radioꢀlabeled [11C]ꢀ1m (0.28 µg, 150 MBq and 0.35 µg, 155 MBq,
respectively) was injected into two male cynomolgus monkeys (PET1 and PET2, respectively).
The PET distribution volume (VT) was estimated for the whole brain region using the 2ꢀtissue
compartment model.38 The distribution volume and estimated free brain to plasma partition
coefficient (Cu,brain/Cu,plasma) are summarized in Table 10. The partition coefficient values (0.50
and 0.53 for PET1 and PET2, respectively) indicated 1m penetrated extensively into monkey
brain. PET images (Figure 1) for both monkeys also suggested that the radioꢀlabeled compound
was well distributed throughout the brain.
Table 10. Unbound Brain/plasma Ratio for [11C]ꢀ1m in Cynomolgus Monkeya
Injected mass and
radioactivity
b
c
VT (mL•cm–3)
fu,,plasma
fu, brain
Cu, brain / Cu, plasma
0.02
0.02
3.0
3.7
0.12
0.14
0.50
0.53
PET1
PET2
0.28 µg, 150 MBq
0.35 µg, 155 MBq
aPET microdosing studies were performed after a single dose of [11C]ꢀ1m (phosphate buffered saline, pH 7.4) via intravenous bolus route. bThe
unbound fraction in plasma was determined in the PET experiment using ultrafiltration. cData obtained from in vitro equilibrium dialysis of rat
brain homogenate were used as an estimate of the unbound fraction in monkey brain tissue.39
ACS Paragon Plus Environment
14

Page 15 of 45
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
Figure 1. Colorꢀcoded PET images showing distribution of radioactivity in the monkey brain following administration of [11C]ꢀ1m. Horizontal
(left panel), coronal (middle panel) and sagittal (right panel) projections for Monkey PET1 (A) and Monkey PET2 (B). Images created from
summation images from 5 to 123 minutes and image intensity is normalized for injected radioactivity.
In vivo anti-tumor efficacy study. A brain metastasis model was generated by intracerebral
injection of luciferase transfected PCꢀ9 (Exon19Del) cells. Tumor growth was monitored by
IVIS Xenogen imaging system. 1m demonstrated significant dose dependent antiꢀtumor efficacy
(~78% tumor growth inhibition at 7.5 mg/kg qd and tumor regression at 15 mg/kg qd,
respectively, 4 weeks after treatment), with <20% body weight loss, while erlotinib had limited
effect in this model (Figure 2A). At the end of the study, brain tissues were collected for
histological assessment. Significantly decreased tumor area was observed by 1m treatment at the
doses of 7.5mg/kg and 15mg/kg (Figure 2B). In addition, modulation of pEGFR was detected by
a single dose 1m at 15mg/kg 1h after dosing, which confirmed target engagement by 1m (Figure
2C).
(A)
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 45
1
2
3
4
5
6
7
8
Vehicle (n=10)
Vehicle (n=10)
3
2
Erlotinib 15mg/kg qd (n=8)
1m 7.5mg/kg qd (n=8)
1m 15mg/kg qd (n=8)
Erlonitib 15mg/kg qd (n=8)
1m 7.5mg/kg qd (n=8)
1m 15mg/kg qd (n=8)
20
10
1
0
9
0
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
-20
-30
-1
-2
0
10
20
30
40
0
10
20
30
Study Period (day)
Study Period (day)
(B)
3.E+06
2.E+06
1.E+06
0.E+00
(C)
Figure 2. A brain metastasis model was established by injection of luciferase transfected PCꢀ9 cells into the brain parenchyma. Tumor growth
was monitored by measuring bioluminescence signals. About 2ꢀ3 weeks after cell injection, animals were treated with erlotinib at 15 mg/kg qd,
ACS Paragon Plus Environment
16

Page 17 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
and 1m at 7.5 mg/kg qd and 15 mg/kg qd, respectively. (A) Significant tumor growth inhibition was observed by 1m treatment in a dose
dependent manner, compared with erlotinib. (B) Histological assessment showed doseꢀdependent reduction of tumor area with 1m treatment. (C)
6
7
Modulation of pEGFR in tumor tissue was observed by a single dose of 1m at 15mg/kg 1 h after dosing (at 1X and 20X amplification)
.
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
Synthesis Scheme 3 illustrates the synthesis of key intermediates 6 and 8. Starting material
3 was conveniently prepared according to literature procedures.40,41 The reaction of 3 and
commercially available 4 was carried out in acetonitrile without the addition of other reagents.
The resulting HCl salt of acetate 5 was treated with potassium carbonate in methanol to provide
6. Another building block, 8, was readily obtained from commercially available 7 by the reaction
with triphosgene in dichloromethane.42 Crude 6 and 8 were used for the further reactions without
purification. Other similar starting materials 18a-h and 19 were prepared by the same procedure
and the crude materials were used directly.
Scheme 3. Synthesis of Intermediate 6 and 8a
aReagents and conditions: (i) acetonitrile, reflux, 4 h; (ii) potassium carbonate, methanol, 10ꢀ15 °C, 2 h; (iii) triphosgene, pyridine,
dichloromethane, 0 °C to rt.
Scheme 4 describes the synthesis of 1a and 1c. Triflate 9, prepared from 6, was conveniently
converted to 1c under conditions described in the literature.43 Heck reaction of 9 with methyl
acrylate delivered 10 which was subjected to ozonolysis to give aldehyde 11. This agent was
then reacted with morpholine under reductive condition to afford desired compound 1a.
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 45
1
2
3
4
5
6
7
8
9
Scheme 5 depicts the synthetic route of 1b and 2. Triflate 13, prepared from commercially
available 12 under typical conditions, reacted with morpholine to provide 14. After reduction of
14, the resulting 15 was converted into a quinazoline derivative 16 by using formamidine acetate
in ethanol. The preparation of reactive intermediate 17 from 16 was accomplished with oxalyl
chloride in chloroform at an elevated temperature. With this key material, 1b and 2 were
synthesized by substitution reaction without addition of other basic reagents.
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 4. Synthesis of 1a and 1ca
aReagents and conditions: (i) Tf2O, pyridine, 5 °C to rt; (ii) CO, Pd(OAc)2, dppf, morpholine, 70 °C, overnight; (iii) methyl acrylate, Pd(OAc)2,
Et3N, dppf, Bu4NBr, DMF, 80 °C, overnight; (iv) O3, CH2Cl2, ꢀ78 °C then (CH3)2S, ꢀ78 °C to rt; (v) morpholine, NaBH3CN, acetic acid, rt.
Scheme 5. Synthesis of 1b and 2a
ACS Paragon Plus Environment
18

Page 19 of 45
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
aReagents and conditions: (i) Tf2O, pyridine, CH2Cl2, 0 °C to rt, 16 h; (ii) morpholine, CH3CN, reflux, overnight; (iii) Fe, NH4Cl, CH3OH, H2O,
80 °C, 4 h; (iv) formamidine acetate salt, C2H5OH, reflux, overnight; (v) oxalyl chloride, CHCl3, DMF, 60 °C, 2 h; (vi) 3ꢀchloroꢀ2ꢀfluoroaniline,
CH3CN, reflux, 18 h; (vii) 3ꢀchloroꢀ4ꢀfluoroaniline, CH3CN, reflux, 2 days.
Scheme 6. General Synthesis of Compounds 1d-h, 1k-l, and 1na
aReagents and conditions: (i) (C2H5)3N or K2CO3, CH2Cl2 or DMF, rt.
Scheme 7. Synthesis of 1i and 1ja
aReagents and conditions: (i) K2CO3, DMF, rt, overnight; (ii) HCl, 1,4ꢀdioxane, ethyl acetate, rt, 1 h; (iii) paraformaldehyde, NaBH3CN,
CH3CO2H, methanol, rt, overnight; (iv) chiral SFC.
Scheme 8. Synthesis of 1ma
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 45
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
aConditions and reagents: (i) potassium carbonate, DMF, rt; (ii) HCl in dioxane, methanol, rt; (iii) paraformaldehyde, sodium cyanoborohydride,
methanol, rt.
The syntheses of compounds 1d-h, 1k-l, and 1n are illustrated in Scheme 6. The carbamate
formation was carried out in basic media at room temperature. Carbamate 20 was prepared
(Scheme 7) in a similar manner. The subsequent deꢀprotection of 20 and reductive methylation
of 21 provided a racemic mixture (22). The resolution of this mixture by chiral SFC technology
delivered two optically pure isomers 1i and 1j. The preparation of compound 1m from a chiral
staring material (8) by a similar synthetic route is outlined in Scheme 8.
ꢀ CONCLUSION
By correctly balancing physicochemical properties, such as permeability, solubility and efflux
ratio, we were able to identify compound 1a. This early lead has significantly improved CNS
penetration without compromising the in vitro potency. Replacement of methylene group
present in 1a by carbamate linker led to compound 1e. This compound significantly improved
metabolic stability. The terminal basic nitrogen in 1e was found to be an essential element to
achieve good exposures in both brain and blood. By incorporating a methyl group into the
piperazine moiety, we further improved PK properties of 1e, leading to successful discovery of
clinical candidate 1m. 1m exhibited excellent free compound distribution in brain, CSF, and
blood. The extensive in vitro evaluation, including biochemical, cellular, and panel screening,
indicates that this agent is highly potent against EGFR activating mutants and highly selective
towards these targets. Importantly, 1m showed tumor regression in the mouse model with brain
metastasis. The promising data package for 1m strongly supported its selection as a drug
ACS Paragon Plus Environment
20

Page 21 of 45
Journal of Medicinal Chemistry
1
2
3
4
candidate for development. The results from additional in vivo studies are due to be published
5
6
separately.
7
8
9
ꢀ EXPERIMENTAL SECTION
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
In vivo animal model. PCꢀ9 cells (EGFR Exon19 deletion) were cultured with RPMIꢀ1640
medium supplemented with 10% fetal bovine serum (FBS) at 37 °C. In order to monitor tumor
growth in the brain, PCꢀ9 cells were stably transfected with pGL4.50[luc2/CMV/Hygro] vector
containing luciferase, and the bioluminescence signals were measured by a Xenogen imaging
system.44 A brain metastasis model was established by intracerebral (ICB) injection of PCꢀ9_Luc
cells using the method by Lal with modifications.45 In brief, after a sagittal incision over the
parietoꢀoccipital bone, a hole was punctured on the skull at 2.5 mm to the right of the bregma
and 1 mm anterior to the coronal suture. Then the syringe was placed perpendicular to the skull
through the hole and placed 3 mm deep below the skull surface, and the PCꢀ9_Luc cell
suspension was slowly injected. After injection, the sterile bone gel was applied to the hole, the
scalp was pulled back to cover the skull and the wound was closed. The mouse was gently put on
a heating pad to recover, and closely monitored after surgery. All in vivo studies were approved
by the Institutional Animal Care and Use Committee (IACUC) before the experiments were
started.
Xenograft tissues were obtained from PCꢀ9 model after treatment with 1m for assessment of
histology and pEGFR modulation. Samples were harvested following formalin fixation and
paraffin embedding (FFPE) for further study. pEGFR(Tyr1068) (DAKO X9003) IHC was
performed on 3ꢁm FFPE sections using a Ventana automation (Roche) for staining.
Chemistry, Materials, and General Methods. Unless otherwise noted, all reagents and
solvents were obtained from commercial suppliers such as Aldrich, Sigma, Fluka, Acros, EMD
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 45
1
2
3
4
5
6
7
8
9
Sciences, etc., and used without further purification. Dry organic solvents (dichloromethane,
acetonitrile, DMF, etc.) were purchased from Aldrich packaged under nitrogen in Sure/Seal
bottles. All reactions involving air or moisture sensitive reagents were performed under a
nitrogen or argon atmosphere. Silica gel chromatography was performed using prepacked silica
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
1
gel cartridges (Biotage or RediSep). H NMR spectra were recorded on a Bruker AV 400 MHz,
Bruker AV Ⅱ400 MHz, Bruker AV Ⅲ 400 MHz, or a Varian 400 MHz spectrometer at ambient
temperature. Chemical shifts are reported in parts per million (ppm, δ units). Data are reported as
follows: chemical shift, multiplicity (s = singlet, d = doublet, t = triplet, q = quartet, br = broad,
m = multiplet), coupling constants, and number of protons. Reactions were monitored using
LCMS (Shimadzu 20A) with UV detection at 220 & 254 nm and a low resonance electrospray
mode (ESI). All final compounds were purified to > 95% purity, as determined by LCMS (4
min). LCMS (4 min) methods used the following: Shimadzu 20A spectrometer, Xtimate C18 2.1
x 30 mm, 3 µm at 50 °C with a 0.8 mL/min flow rate; solvent A of 1.5 mL TFA in 4 L water,
solvent B of 0.75 mL TFA in 4 L acetonitrile; 0.0–3.0 min, 10% B; 3.0–3.5 min, 80% B; 3.51–
4.0 min, 10% B. Flow from UV detector was split (1:2) to the MS detector, which was
configured with ESI as ionizable source. All high resolution mass spectrometry (HRMS) data
were acquired on an Agilent 1100 Series (LC/MSD TOF G3251) operated in positive
electrospray ionization mode. The mobile phase: 0.1% TFA in water (solvent A) and 0.1% TFA
in acetonitrile (solvent B), using the elution gradient 10%ꢀ80% (solvent B) over 2 minutes and
holding at 80% for 1 minute, then back to 10% and holding at 10% for 1 minute at a flow rate of
1 mL/minute. Nebulization Gas: nitrogen; Time of Flight mass spectrometer with ESI source;
Positive ion mode; Drying gas (N2) flow: 12 l/min; nebulizer pressure: 20 psi, 350 Centigrade;
Capillary voltage: 3000V. The fragmentor voltage: 50V.
ACS Paragon Plus Environment
22

Page 23 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
4-(3-Chloro-2-fluorophenylamino)-7-methoxyquinazolin-6-yl acetate (5). To a suspension of
4ꢀchloroꢀ7ꢀmethoxyquinazolinꢀ6ꢀyl acetate (3) (100 g, 396 mmol) in acetonitrile (4 L) was added
3ꢀchloroꢀ2ꢀfluoroaniline (4) (57.8 g, 396.8 mmol). The reaction mixture was heated to reflux
for 4 h with stirring before being cooled to room temperature. After filtration, the cake was
washed with acetonitrile (1 L) and dried in vacuo to afford the HCl salt of 5 (143.3 g, 91%) as a
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
1
white solid. H NMR (400 MHz, DMSOꢀd6): δ 11.64 (s, 1H), 8.87 (s, 1H), 8.68 (s, 1H), 7.67ꢀ
7.62 (m, 1H), 7.52 (s, 2H), 7.39ꢀ7.33 (m, 1H), 4.01 (s, 3H), 2.39 (s, 3H). LCMS (ESI) m/z
calculated for C17H13ClFN3O3 + H+ [M+H+]: 362.1. Found: 361.8.
4-(3-Chloro-2-fluorophenylamino)-7-methoxyquinazolin-6-ol (6). Potassium carbonate (114.0
g, 826 mmol) was added to a solution of HCl salt of 5 (150.0 g, 377 mmol) in methanol (800 ml).
The reaction mixture was stirred at 10ꢀ15 °C for 2 h. After filtration, the solid was rinsed with
methanol (120 mL). The filtrate was concentrated in vacuo to a residue (~300 g) and treated with
methyl tertꢀbutyl ether (800 mL). The resulting mixture was filtered. The solid was dried at 40ꢀ
1
45 °C to afford a potassium salt of compound 6 (132.0 g, 98%). H NMR (400 MHz, DMSOꢀ
d6): δ 7.97 (s, 1H), 7.61ꢀ7.57 (m,1H), 7.24ꢀ7.11 (m, 2H), 7.29ꢀ7.25 (m,1H), 6.89 (s,1H), 6.79 (s,
1H), 3.77 (s, 3H). LCMS (ESI) m/z calculated for C15H11ClFN3O2 + H+ [M+H+]: 320.1. Found:
319.9.
(R)-tert-Butyl 4-(chlorocarbonyl)-3-methylpiperazine-1-carboxylate (8). Pyridine (18 g, 225
mmol) was added dropꢀwise to a mixture of triphosgene (23 g, 75 mmol) in dry dichloromethane
(250 mL) followed by (R)ꢀ4ꢀBocꢀ2ꢀmethylpiperazine (7) (15 g, 75 mmol) at 0 °C under nitrogen.
The reaction mixture was stirred overnight at room temperature and concentrated to afford 8 as a
yellow solid, which was used without further purification.
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 45
1
2
3
4
4-(3-Chloro-2-fluorophenylamino)-7-methoxyquinazolin-6-yl
trifluoromethanesulfonate
5
6
7
8
(9). Pyridine (5 g, 62.4 mmol) was added to a solution of 6 (5 g, 15.6 mmol) in dichloromethane
(60 mL) followed by the addition of trifluoromethanesulfonic anhydride (13.2 g, 46.8 mmol)
dropwise at 5 °C. The reaction mixture was warmed to room temperature and stirred at that
temperature overnight. The mixture was treated with water and separated. The organic layer
was dried over sodium sulfate and concentrated. The residue was purified by chromatography
using silica gel column and eluting with ethyl acetate in petroleum (1:3) to afford 9 (5.2 g, 11.53
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
1
mmol, 74%) as a yellow solid. H NMR (400 MHz, CDCl3): δ 8.81 (s, 1H), 8.42ꢀ8.39 (m, , 1H),
7.76 (s, 1H), 7.46 (s, 1H), 7.39 (s, 1H), 7.24ꢀ7.18 (m, 2H), 4.10 (s, 3H). LCMS (ESI) m/z
calculated for C16H10ClF4N3O4S + H+ [M+H+]: 452.0. Found: 451.8.
(E)-Methyl 3-(4-(3-chloro-2-fluorophenylamino)-7-methoxyquinazolin-6-yl)acrylate (10).
Methyl acrylate (2.21 g, 25.7 mmol), Pd(OAc)2 (287 mg, 1.28 mmol), 1,1'ꢀ
bis(diphenylphosphino)ferrocene (710 mg, 1.28 mmol), Bu4NBr (4.13 g, 12.8 mmol) and Et3N
(2.6 g, 25.7 mmol) were added to a solution of 9 (5.8 g, 12.8 mmol) in DMF (200 mL). The
reaction mixture was degassed and refilled with nitrogen three times. The mixture was stirred at
80 °C overnight, cooled to room temperature, and treated with ethyl acetate (500 mL) and water
(600 mL). After separation, the aqueous layer was extracted with ethyl acetate three times. The
combined extracts were washed with brine, dried over sodium sulfate, and concentrated. The
residue was purified by flash chromatography eluting with ethyl acetate in petroleum (1:2) to
afford 10 (4 g, 10.3 mmol, 80%). 1H NMR (400 MHz, CDCl3): δ 8.76 (s, 1H), 8.48ꢀ8.44 (m,
1H), 8.07ꢀ7.03 (m, 2H), 7.72 (s, 1H), 7.30 (s, 1H), 7.23ꢀ7.16 (m, 2H), 6.74 (d, J = 16.4 Hz, 1H),
4.06 (s, 3H), 3.86 (s, 3H). LCMS (ESI) m/z calculated for C19H15ClFN3O3 + H+ [M+H+]: 388.1.
Found: 387.9.
ACS Paragon Plus Environment
24

Page 25 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
4-(3-Chloro-2-fluorophenylamino)-7-methoxyquinazoline-6-carbaldehyde (11). A solution
of 10 (4 g, 10.3 mmol) in dichloromethane (100 mL) was cooled with dryꢀice bath and ozone gas
was bubbled into the solution for 15 min. The nitrogen was bubbled into the mixture to remove
O3. After addition of methyl sulfide (3.2 g, 51.9 mmol), the mixture was warmed to room
temperature and stirred at that temperature for 30 min. The solvent was removed in vacuo and
the residue was purified by flash chromatography eluting with petroleum in ethyl acetate (2: 1) to
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
1
afford 11 (1.3 g, 38% yield) as yellow solid. H NMR (400 MHz, CDCl3): δ 10.58 (s, 1H), 8.79
(s, 1H), 8.54 (s, 1H), 8.34 (t, J = 7.2 Hz, 1H), 7.50 (s, 1H), 7.25ꢀ7.18 (m, 2H), 4.12 (s, 3H).
LCMS (ESI) m/z calculated for C16H11ClFN3O2 + H+ [M+H+]: 332.1. Found: 332.1.
N-(3-Chloro-2-fluorophenyl)-7-methoxy-6-(morpholinomethyl)quinazolin-4-amine
(1a).
Morpholine (420 mg, 4.82 mmol) and acetic acid (289 mg, 4.82 mmol) were added to a solution
of 11 (800 mg, 2.41 mmol) in methanol (10 mL). The reaction mixture was stirred at room
temperature for 30 min and NaBH3CN (227 mg, 3.61 mmol) was added. The resulting mixture
was stirred at room temperature for additional 1 h and concentrated. The residue was purified by
reverse phase preparative HPLC using Agella Venusil ASB C18 150 x 21 mm, 5 µm column and
eluting with a gradient of 5ꢀ35% acetonitrile in water containing 0.05% HCl to afford HCl salt of
1a (575.2 mg, 1.31 mmol, 54%). 1H NMR (400 MHz, CD3OD): δ 8.87 (s, 1H), 8.81 (s, 1H),
7.59ꢀ7.53 (m, 2H), 7.39 (s, 1H), 7.35ꢀ3.31 (m, 1H), 4.62 (s, 2H), 4.19 (s, 3H), 4.07 (d, J = 12.8
Hz, 2H), 3.91 (t, J = 12.0 Hz, 2H), 3.54 (d, J = 12.0 Hz, 2H), 3.43ꢀ3.37 (m, 2H). 13C NMR (101
MHz, CD3OD): δ 166.1, 162.4, 153.5, 153.0, 143.6, 132.2, 131.6, ,128.3, 126.9, 126.3, 123.1,
122.6, 108.2, 100.9, 65.0, 58.2, 56.4, 53.6. HRMS(ESI): m/z calculated for C20H20ClFN4O2 + H+
[M + H+]: 403.1337. Found 403.1302
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 45
1
2
3
4
5
6
7
8
9
Methyl 4-methoxy-2-nitro-5-(trifluoromethylsulfonyloxy)benzoate (13). Pyridine (8.7 g, 110
mmol) was added to a solution of methyl 5ꢀhydroxyꢀ4ꢀmethoxyꢀ2ꢀnitrobenzoate (5 g, 22 mmol)
in dichloromethane (50 mL). The reaction mixture was cooled to 0 °C with iceꢀwater, and then
trifluoromethanesulfonic anhydride (12.4 g, 44 mmol) was added slowly to maintain the internal
temperature below 10 °C. The reaction mixture was slowly warmed to room temperature and
stirred at that temperature for 16 h. Water (100 mL) was then added and the organic layer was
separated. The aqueous layer was extracted with dichloromethane (3 x 100 mL). The combined
organic layers were washed with citric acid solution (2 x 100 mL) and brine. The organic
solution was dried over sodium sulfate and concentrated to afford 13 (5.9 g, 75%) as a yellow
solid. This material was used for the next reaction without purification. 1H NMR (400 MHz,
CDCl3): δ 7.71 (s, 1H), 7.41 (s, 1H), 4.05 (s, 3H), 3.91 (s, 3H).
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
Methyl 4-methoxy-5-morpholino-2-nitrobenzoate (14). Morpholine (0.73 g, 7.35 mmol) was
added to a solution of 13 (1 g, 2.78 mmol) in acetonitrile (20 mL). The reaction mixture was
refluxed overnight prior to the addition of a second batch of morpholine (0.5 g, 5.73 mmol). The
reaction mixture was refluxed for an additional 3 h and cooled to room temperature. The solvent
was evaporated in vacuo and the residue was purified by flash chromatography eluting with ethyl
1
acetate in petroleum (1:3) to afford 14 (800 mg, 2.7 mmol, 97%). H NMR (400 MHz, CDCl3):
δ 7.48 (s, 1H), 7.02 (s, 1H), 3.96 (s, 3H), 3.90 (s, 3H), 3.87 (t, J = 4.8 Hz, 4H), 3.21 (t, J = 4.8 Hz,
4H).
Methyl 2-amino-4-methoxy-5-morpholinobenzoate (15). Water (5 mL), ammonium chloride
(1.44 g, 27 mmol) and iron powder (753 mg, 13.5 mmol) were added to a solution of 14 (800 mg,
2.7 mmol) in methanol (15 mL). The suspension was heated to 60 °C for 4 h and cooled to room
temperature. The mixture was filtered. The filtration cake was rinsed with dichloromethane.
ACS Paragon Plus Environment
26

Page 27 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
After separation of the filtrate, the aqueous layer was extracted with dichloromethane (2 x 50
mL). The combined extracts were dried over sodium sulfate and concentrated to afford 15 (720
mg, 2.7 mmol, 100%) as a yellow solid. This product was used for the next step without
purification.
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
7-Methoxy-6-morpholinoquinazolin-4-ol (16). Formamidine acetate salt (563 mg, 5.4 mmol)
was added to a solution of 15 (720 mg, 2.7 mmol) in ethanol (20 mL). The reaction mixture was
refluxed overnight and cooled to room temperature. The solvent was evaporated in vacuo and
the residue was suspended with water (10 mL). The solid was collected by filtration and dried in
1
vacuo to afford 16 (400 mg, 1.53 mmol, 57%) as a white solid. H NMR (400 MHz, DMSOꢀd6):
δ 12.10 (br, 1H), 7.97 (s, 1H), 7.44 (s, 1H), 7.12 (s, 1H), 3.93 (s, 3H), 3.76 (t, J = 4.4 Hz, 4H),
3.03 (t, J = 4.4 Hz, 4H). LCMS (ESI) m/z calculated for C13H15N3O3 + H+ [M+H+]: 262.1.
Found: 262.1.
4-(4-Chloro-7-methoxyquinazolin-6-yl)morpholine (17). Three drops of DMF was added to a
suspension of 16 (600 mg, 2.3 mmol) in chloroform (10 mL) followed by the dropwise addition
of oxalyl chloride (870 mg, 6.9 mmol). The reaction mixture was stirred at 80 °C for 2 h and
cooled to room temperature. The solvent was evaporated in vacuo to afford 17 (644 mg, 100%)
as a brown solid, which was used for the next step directly. LCMS (ESI) m/z calculated for
C13H14ClN3O2 + H+ [M+H+]: 280.1. Found: 279.8.
N-(3-Chloro-2-fluorophenyl)-7-methoxy-6-morpholinoquinazolin-4-amine (1b). 3ꢀChloroꢀ2ꢀ
fluoroaniline (670 mg, 4.6 mmol) was added to a suspension of 17 (644 mg, 2.3 mmol) in
acetonitrile (10 mL). The reaction mixture was refluxed for 18 h and cooled to room
temperature. The solvent was evaporated in vacuo and the residue was purified by reverse phase
preparative HPLC using YMCꢀpack ODSꢀAQ 150 x 30 mm, 5 µm column and eluting with a
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 45
1
2
3
4
gradient of 35ꢀ65% acetonitrile (containing 0.05% ammonia) in water to afford 1b (446.6 mg,
5
6
1
1.15 mmol, 50%) as a white solid. H NMR (400 MHz, CD3OD): δ 8.36 (s, 1H), 7.72 (s, 1H),
7
8
9
7.58ꢀ7.54 (m, 1H), 7.44ꢀ7.41 (m, 1H), 7.25 (d, J = 9.6 Hz, 1H), 7.22 (s, 1H), 4.05 (s, 3H), 3.93ꢀ
3.90 (m, 4H), 3.23ꢀ3.21 (m, 4H). HRMS(ESI): m/z calculated for C19H18ClFN4O2 + H+ [M +
H+]: 389.1181. Found: 389.1186.
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
N-(3-Chloro-4-fluorophenyl)-7-methoxy-6-morpholinoquinazolin-4-amine (2). 3ꢀChloroꢀ4ꢀ
fluoroaniline (166 mg, 1.14 mmol) was added to a suspension of 17 (160 mg, 0.57 mmol) in
acetonitrile (3 mL). The reaction mixture was refluxed for 2 days and cooled to room
temperature. The mixture was treated with water, neutralized with sodium carbonate solution,
and extracted with ethyl acetate (3 x 100 mL). The combined organic layers were dried over
sodium sulfate and concentrated. The residue was purified by reverse phase preparative HPLC
using Agella Venusil ASB C18 150 x 25 mm, 5 µm column and eluting with a gradient of 9ꢀ79%
acetonitrile (containing 0.05% HCl) in water to afford HCl salt of 2 (110.8 mg, 0.26 mmol, 49%)
1
as a white solid. H NMR (400 MHz, CD3OD): δ 8.72 (s, 1H), 7.96ꢀ7.94 (m, 2H), 7.69ꢀ7.66 (m,
1H), 7.39 (t, J = 8.8 Hz, 1H), 7.26 (s, 1H), 4.13 (s, 3H), 3.93ꢀ3.91 (m, 4H), 3.230ꢀ3.28 (m, 4H).
13C NMR (101 MHz, CD3OD): δ 162.0, 160.4, 156.6, 149.8, 145.7, 138.0, 135.2, 128.4, 126.4,
121.9, 118.1, 112.0, 109.2, 100.6, 68.0, 57.5, 52.5. HRMS(ESI): m/z calculated for
C19H18ClFN4O2 + H+ [M + H+]: 389.1181. Found: 389.1205.
(4-(3-Chloro-2-fluorophenylamino)-7-methoxyquinazolin-6-yl)(morpholino)methanone
(1c). Pd(OAc)2 (10 mg, 0.044 mmol) and 1,1'ꢀbis(diphenylphosphino)ferrocene (24 mg, 0.044
mmol) were added to a solution of 9 (200 mg, 0.44 mmol) in morpholine (10 mL). The reaction
mixture was stirred at 70 °C under carbon monoxide atmosphere (45 psi) overnight. After being
cooled to room temperature, the solvent was evaporated in vacuo. The residue was purified by
ACS Paragon Plus Environment
28

Page 29 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
reverse phase preparative HPLC using Phenomenex Gemini C18 200 x 25 mm, 10 µm column
and eluting with a gradient of 30ꢀ50% acetonitrile in water to afford 1c (70.6 mg, 0.17 mmol,
1
39%). H NMR (400 MHz, CD3OD): δ 8.48 (s, 1H), 8.30 (s, 1H), 7.59—7.55 (m, 1H), 7.45ꢀ7.42
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
(m, 1H), 7.31 (s, 1H), 7.27ꢀ7.23 (m, 1H), 4.06 (s, 3H), 3.84ꢀ3.77 (m, 4H), 3.69ꢀ3.64 (m, 2H),
3.37ꢀ3.34 (m, 2H). 13C NMR (101 MHz, CD3OD): δ 168.6, 161.1, 160.5, 157.1, 155.8, 153.5,
129.4, 128.4, 127.9, 125.9, 124.4, 122.8, 122.7, 110.7, 107.3, 68.1, 67.8, 57.0, 43.8.
HRMS(ESI): m/z calculated for C20H18ClFN4O3 + H+ [M + H+]: 417.1130. Found: 417.1144.
4-(3-Chloro-2-fluorophenylamino)-7-methoxyquinazolin-6-yl
morpholine-4-carboxylate
(1d). Morpholineꢀ4ꢀcarbonyl chloride (18a) (250 mg, ~1.7 mmol) was added to a solution of 6
(500 mg, 1.4 mmol) and Et3N (285 mg, 2.8 mmol) in anhydrous dichloromethane (20 mL). The
reaction mixture was stirred at 25 °C overnight and concentrated under reduced pressure. The
residue was purified by reverse phase preparative HPLC using Synergi C18 200 x 25mm, 10 µm
column and eluting with 0ꢀ41% acetonitrile in water to give 1d (272 mg, 0.63 mmol, 45%) as a
yellow solid. 1H NMR (400 MHz, CD3OD): δ 8.46 (s, 1H), 8.11 (s, 1H), 7.62ꢀ7.58 (m, 1H),
7.43ꢀ7.40 (m, 1H), 7.31 (s, 1H), 7.26ꢀ7.22 (m, 1H), 4.03 (s, 3H), 3.78 (br, 6H), 3.58 (br, 2H). 13C
NMR (101 MHz, CD3OD): δ 160.1, 158.5, 156.0, 155.1, 151.0, 142.2, 129.2, 129.0, 127.8,
125.9, 122.8, 122.6, 117.4, 110.2, 108.4, 67.7, 57.1, 46.7. HRMS (ESI): m/z calculated for
C20H18ClFN4O4 + H+ [M + H+]: 433.1079. Found: 433.1106.
4-(3-Chloro-2-fluorophenylamino)-7-methoxyquinazolin-6-yl
4-methylpiperazine-1-
carboxylate (1e). 4ꢀmethylpiperazineꢀ1ꢀcarbonyl chloride (18b) (19.9 g, ~122.3 mmol) and
potassium carbonate (27.5 g, 196 mmol) were added to a solution of 6 (15.64 g, 48.9 mmol) in
DMF (270 mL). The reaction mixture was stirred at room temperature for 60 h before being
treated with water (100 mL) and extracted with CH2Cl2 (3 x 500 mL). The combined organic
ACS Paragon Plus Environment
29