Start Selective and Rigidify: The Discovery Path toward a Next Generation of EGFR Tyrosine Kinase Inhibitors

Article abstract of DOI:10.1021/acs.jmedchem.9b01169

The epidermal growth factor receptor (EGFR), when carrying an activating mutation like del19 or L858R, acts as an oncogenic driver in a subset of lung tumors. While tumor responses to tyrosine kinase inhibitors (TKIs) are accompanied by marked tumor shrinkage, the response is usually not durable. Most patients relapse within two years of therapy often due to acquisition of an additional mutation in EGFR kinase domain that confers resistance to TKIs. Crucially, oncogenic EGFR harboring both resistance mutations, T790M and C797S, can no longer be inhibited by currently approved EGFR TKIs. Here, we describe the discovery of BI-4020, which is a noncovalent, wild-type EGFR sparing, macrocyclic TKI. BI-4020 potently inhibits the above-described EGFR variants and induces tumor regressions in a cross-resistant EGFR^{del19 T790M C797S} xenograft model. Key was the identification of a highly selective but moderately potent benzimidazole followed by complete rigidification of the molecule through macrocyclization.

Full text of DOI:10.1021/acs.jmedchem.9b01169

Subscriber access provided by LUNDS UNIV
Article
Start selective and rigidify: The discovery path towards
a next generation of EGFR tyrosine kinase inhibitors.
Harald Engelhardt, Dietrich Böse, Mark Petronczki, Dirk Scharn, Gerd Bader, Anke Baum, Andreas Bergner,
Eugene Chong, Sandra Döbel, Georg Egger, Christian Engelhardt, Peter Ettmayer, Julian E. Fuchs, Thomas
Gerstberger, Nina C. Gonnella, Andreas Grimm, Elisabeth Grondal, Nizar Haddad, Barbara Hopfgartner,
Roland Kousek, Mariusz Krawiec, Monika Kriz, Lyne Lamarre, Joyce Leung, Moriz Mayer, Nitinchandra D
Patel, Biljana Peric Simov, Jonathan T. Reeves, Renate Schnitzer, Andreas Schrenk, Bernadette Sharps,
Flavio Solca, Heinz Stadtmüller, Zhulin Tan, Tobias Wunberg, Andreas Zöphel, and Darryl B. McConnell
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b01169 • Publication Date (Web): 05 Nov 2019
Downloaded from pubs.acs.org on November 5, 2019
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 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 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.
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 101
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Signaling
Lamarre, Lyne; Boehringer Ingelheim RCV GmbH & Co KG, Cancer Cell
Signaling
Leung, Joyce; Boehringer Ingelheim Pharmaceuticals Inc, Chemical
Development
Mayer, Moriz; Boehringer Ingelheim RCV GmbH und Co KG, Medicinal
Chemistry
Patel, Nitinchandra; Boehringer Ingelheim Pharmaceuticals Inc, Chemical
Development
Peric Simov, Biljana; Boehringer Ingelheim RCV GmbH & Co KG,
Medicinal Chemistry
Reeves, Jonathan; Boehringer Ingelheim, Chemical Development
Schnitzer, Renate; Boehringer Ingelheim RCV GmbH & Co KG, Cancer
Cell Signaling
Schrenk, Andreas; Boehringer Ingelheim RCV GmbH & Co KG, Cancer
Cell Signaling
Sharps, Bernadette; Boehringer Ingelheim RCV GmbH & Co KG, Cancer
Cell Signaling
Solca, Flavio; Boehringer Ingelheim RCV GmbH & Co KG, Cancer
Pharmacology & Translational Research
Stadtmüller, Heinz; Boehringer Ingelheim RCV GmbH & Co KG, Medicinal
Chemistry
Tan, Zhulin; Boehringer Ingelheim, Chemical Development
Wunberg, Tobias; Boehringer Ingelheim RCV GmbH & Co KG, Medicinal
Chemistry
Zöphel, Andreas; Boehringer Ingelheim RCV GmbH & Co KG, Medicinal
Chemistry
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
McConnell, Darryl; Boehringer Ingelheim RCV GmbH & Co KG, Medicinal
Chemistry
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 101
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Start selective and rigidify: The discovery path
towards a next generation of EGFR tyrosine kinase
inhibitors.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
1
1
1
1
Harald Engelhardt ‡*, Dietrich Böse †‡, Mark Petronczki , Dirk Scharn , Gerd Bader ,
1
1
2
1
1
Anke Baum , Andreas Bergner , Eugene Chong , Sandra Döbel , Georg Egger ,
1
1
1
1
Christian Engelhardt , Peter Ettmayer , Julian E. Fuchs , Thomas Gerstberger , Nina
2
1
1
2
1
Gonnella , Andreas Grimm , Elisabeth Grondal , Nizar Haddad , Barbara Hopfgartner ,
1
2
1
1
2
Roland Kousek , Mariusz Krawiec , Monika Kriz , Lyne Lamarre , Joyce Leung ,Moriz
1
2
1
2
Mayer , Nitinchandra D. Patel , Biljana Peric Simov , Jonathan T. Reeves , Renate
1
1
1
1
1
Schnitzer , Andreas Schrenk , Bernadette Sharps , Flavio Solca , Heinz Stadtmüller ,
2
1
1
1
Zhulin Tan , Tobias Wunberg , Andreas Zoephel , Darryl B. McConnell
1Boehringer Ingelheim RCV GmbH & Co KG, Dr-Boehringer-Gasse 5-11, 1120 Vienna,
Austria
ACS Paragon Plus Environment
1

Page 3 of 101
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2Boehringer Ingelheim Pharmaceuticals, Inc., 900 Ridgebury Road, Ridgefield,
Connecticut 06877, United States
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment
2

Journal of Medicinal Chemistry
Page 4 of 101
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ABSTRACT
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The epidermal growth factor receptor (EGFR), when carrying an activating mutation like
del19 or L858R, acts as an oncogenic driver in a subset of lung tumors. While tumor
responses to tyrosine kinase inhibitors (TKIs) are accompanied by marked tumor
shrinkage, the response is usually not durable. Most patients relapse within two years of
therapy often due to acquisition of an additional mutation in EGFR kinase domain that
confers resistance to TKIs. Crucially, oncogenic EGFR harboring both resistance
mutations, T790M and C797S, can no longer be inhibited by currently approved EGFR
TKIs. Here, we describe the discovery of BI-4020, which is a non-covalent, wild-type
EGFR sparing, macrocyclic TKI. BI-4020 potently inhibits the above described EGFR
variants and induces tumor regressions in a cross-resistant EGFRdel19 T790M C797S xenograft
model. Key was the identification of a highly selective but moderately potent
ACS Paragon Plus Environment
3

Page 5 of 101
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
benzimidazole followed by complete rigidification of the molecule through
macrocyclization.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment
4

Journal of Medicinal Chemistry
Page 6 of 101
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
INTRODUCTION
Kinases are one of the most successfully pursued drug target families with more than
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
50 kinase inhibitors approved by the FDA to date. Despite this success, discovering
kinase inhibitors with a sufficiently high degree of kinome selectivity and hence tolerability
remains a significant challenge to medicinal chemists.2, 3 This challenge is exacerbated
in oncology by the emergence of drug resistance which requires next generation inhibitors
that are able to inhibit both primary and resistant oncogenic clones while maintaining
kinome selectivity.4-7 New approaches for the discovery of highly selective kinase
inhibitors have emerged such as the screening of libraries of kinase inhibitors with intrinsic
selectivity profiles8, 9 and macrocyclization of kinase inhibitors.10, 11
The epidermal growth factor receptor (EGFR) is a receptor tyrosine kinase that
transduces mitogenic signals. Mutations in the EGFR gene are found in approximately 12
%
to 47 % of non-small cell lung cancer (NSCLC) tumors with adenocarcinoma
12
histology. The two most frequent EGFR alterations found in NSCLC tumors are short
in-frame deletions in exon 19 (del19) of the EGFR gene and L858R, a single missense
ACS Paragon Plus Environment
5

Page 7 of 101
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
3
mutation in exon 21. These two mutations cause ligand-independent EGFR activation
M+
and are collectively referred to as EGFR . Del19 and L858R mutations in EGFR
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
sensitize NSCLC tumors to the treatment with EGFR tyrosine kinase inhibitors (TKIs).
Clinical experience shows that the objective response rate in lung cancer patients
carrying tumors with EGFRM+ is approximately 60-85 % in 1 line with the 1 , 2 and 3
st
st
nd
rd
generation EGFR TKIs erlotinib, gefitinib, afatinib and osimertinib (structures see Figure
).14-17 These responses demonstrate that EGFRM+ NSCLC cells and tumors depend on
1
oncogenic EGFR activity for survival and proliferation, establishing del19 or L858R
mutated EGFR variants as a validated drug targets and predictive biomarkers for the
treatment of NSCLC.
While tumor responses are accompanied by marked tumor shrinkage in patients, the
response is usually not durable, and most patients relapse within 9.2 to 14.7 months of
treatment with 1 and 2 generation EGFR TKIs.14-18 The most prominent molecular
st
nd
resistance mechanism associated with disease progression in 50% to 70% of patients
st
nd
treated with 1 and 2 generation EGFR inhibitors is the acquisition of a secondary
ACS Paragon Plus Environment
6

Journal of Medicinal Chemistry
Page 8 of 101
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
mutation in EGFR, namely the gatekeeper mutation T790M.19, 20 This mutation attenuates
st
nd
the inhibitory activity of 1 and 2 generation TKIs in cellular assays. EGFR wild-type
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
rd
sparing covalent 3 generation EGFR TKIs, such as osimertinib, have been developed
that effectively inhibit the primary EGFR mutations del19 and L858R independent of the
presence or absence of the secondary T790M resistance mutation.21, 22
st
nd
rd
Figure 1. Overview of selected approved 1 , 2 and 3 generation EGFR TKIs.
While osimertinib in 2nd line treatment of EGFR M T790M-positive NSCLC patients
+
demonstrates clear clinical efficacy23, 24 disease progression occurs after an average
24
nd
duration of 10 months. Recent data suggest that 20-40% of 2 line patients relapsing
ACS Paragon Plus Environment
7

Page 9 of 101
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
on osimertinib TKI25-29 acquire a tertiary EGFR mutation at the cysteine residue (C797S)
required for covalent inhibitor binding.21, 22 Crucially, the EGFR del19/L858R T790M
C797S cis mutant kinase variant that emerges in 2nd line patients progressing on
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
osimertinib25-29 can no longer be inhibited by any of the approved 1 , 2 or 3 generation
st
nd
rd
EGFR TKIs.27 While the EGFR resistance mutation spectrum after 1 line osimertinib
st
treatment has not been extensively studied, first available data also suggest the
30
emergence of the C797S mutation. Hence, the next generation of EGFR TKIs should
be (a) highly active against most common EGFR activation mutation del19, (b) not be
susceptible to the most common resistance mutations T790M and/or C797S, (c) have low
wt
inhibition of EGFR to avoid on-target toxicities such as rash and diarrhea and (d) be
highly selective across the human kinome.
ACS Paragon Plus Environment
8

Journal of Medicinal Chemistry
Page 10 of 101
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
3
Figure 2. Overview of selected published EGFR inhibitors with inhibition data (R1 = 31b ;
37
35
39
43
42
R2 = JND3229 ; R3 = B30 ; R4 = 42 ; R5 = 21 and R6 = 19 ). Compound 1 serves
as starting point for this work and BI-4020 demonstrates the superiority of the compounds
reported by this work compared to state of the art.
Recently, a number of potent EGFR inhibitors able to overcome EGFR T790M and
C797S resistance mutations were reported31-44 but these compounds are equipotent or
ACS Paragon Plus Environment
9

Page 11 of 101
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
wt
31-38
more active on EGFR (see R1-3 in Figure 2).
Conversely, some cases of good
selectivity for mutated versus wild-type EGFR have been reported but with low overall in
vitro potency (see R4 and R6 in Figure 2).40-42 The non-covalent inhibitor R543 has
demonstrated acceptable p-EGFR biomarker modulation in-vivo in a NCI-H1975
xenograft model (EGFRL858R T790M) after high oral dose administration, but the compound
displays an insufficient EGFR wt-sparing ratio. Allosteric EGFR inhibitors which overcome
T790M and C797S resistance and display EGFR wt-sparing behavior have been
reported,45-49 but these compounds are only active on the L858R mutant but not on the
most prevalent del19 mutant.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
In this paper, we describe the discovery of BI-4020, the first compound to exhibit a next
generation EGFR TKI profile, wt-sparing, capable of inhibiting complex EGFR mutations
containing T790M and/or C797S modifications, including a triple cis mutation in the
EGFRdel19 background. Key to the discovery was the identification of a highly selective
but moderately potent benzimidazole by screening our corporate selective kinase inhibitor
library followed by complete rigidification of the molecule through macrocyclization. BI-
ACS Paragon Plus Environment
10

Journal of Medicinal Chemistry
Page 12 of 101
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
020 is a non-covalent, macrocyclic TKI which inhibits not only the triple mutant EGFRdel19
T790M C797S
variant but also the double mutant EGFRdel19 T790M and primary mutant
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
EGFRdel19 whilst sparing activity against EGFR (see Figure 2 and Table 4). BI-4020
shows high potency on EGFR mutant cells, high kinome selectivity and good DMPK
properties which led to tumor regressions in the human PC-9 (EGFRdel19 T790M C797S) triple
mutant NSCLC xenograft model in mice.
wt
ACS Paragon Plus Environment
11

Page 13 of 101
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
RESULTS AND DISCUSSION
st
Selectivity 1 : Screening for a highly selective rather than a highly active starting point.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
At the outset of the project we decided to start a screening campaign to find chemical
starting points which provided selectivity for the mutated EGFR proteins over EGFRwt
protein rather than a high activity. We hypothesized that it would be easier to improve
wt
upon the activity of an EGFR selective molecule than it would be to rationally design in
EGFRwt selectivity of a potent but unselective inhibitor. A structure-based design
approach to generate selectivity for the EGFR primary activation mutation L858R and
wt
del19 over EGFR was further ruled out as no structural differences between the binding
pockets from the publicly available x-ray structures of the EGFRL858R and EGFR proteins
wt
were observable.
Instead, we resolved to design a focused in-house library based on Boehringer
Ingelheim kinase selectivity panel data obtained with previous research compounds from
various projects. Focusing on selective kinase inhibitors where more than 10 mg solid
was in stock and that fulfilled minimal drug like properties (heavy atoms ≤ 34 or soluble
ACS Paragon Plus Environment
12

Journal of Medicinal Chemistry
Page 14 of 101
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
at pH 6.8), we compiled a list of 1,826 kinase inhibitors from 40 past kinase and 47 non-
kinase projects. A compound, which qualifies as a selective kinase inhibitor needs to be
tested against at least 30 kinases and should be active in at least 2 assays (single
concentration measurement using a biochemical kinase assay; compound need to
display a value a value blow 50% of control) but should not hit more than 10% of all tested
kinases. Interestingly, this Selective Kinase Inhibitor (SKIny) Screening Library covered
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
269 kinases.
Despite the fact that we were most interested in the more frequent appearing EGFRdel19
T790M C797S
mutation, the triple mutant EGFRL858R T790M C979S protein was used for the
biochemical screening assay, because we were not able to obtain sufficient high quality
amounts of the EGFRdel19 T790M C797S protein. In parallel to an activity screening, we
wt
conducted a selectivity screening using the EGFR protein in order to directly assess the
WT-sparing effect of the hits found in the screen. The aminobenzimidazole 1 was
identified as a promising hit from the screen as it showed reasonable affinity to mutated
L858R T790M C797S
del19 T790M C797S
EGFR (IC EGFR
= 2,100 nM); IC EGFR
= 250 nM) and
5
0
50
wt
had no measurable affinity to EGFR (see Table 1). In order to assess the overall kinase
ACS Paragon Plus Environment
13

Page 15 of 101
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
5
0
selectivity of 1 we performed a kinase selectivity assay with 238 kinases and found that
only 5% of the screened kinases were inhibited above 70%. In order to determine the
overall kinase selectivity for further compounds more quickly, we have defined a diverse
set of 30 kinases representing the kinome. For key compounds 1, 3 and BI-4020 we have
confirmed the results of the small panel with a larger set of kinases (see Table 1 and 2).
Gratifyingly, we were also already able to measure biomarker modulation in an EGFRdel19
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
T790M C797S
dependent BaF3 cell line with an IC = 790 nM.
50
Figure 3. X-ray structure of ligand 1 (PDB: 6S9B) bound to the ATP pocket of the
EGFRL858R T790M protein.
ACS Paragon Plus Environment
14

Journal of Medicinal Chemistry
Page 16 of 101
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
EGFRL858R protein was used for protein crystallography as EGFRdel19 protein has proven
particularly recalcitrant to x-ray crystallography with no structures published. Compound
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
was successfully crystallized with EGFRL858R with the gatekeeper mutation T790M
(Figure 3). From the x-ray structure a two-point hinge binding motive can be seen, with
the benzimidazole N-H and carbonyl group binding to Met793. The isobutyl hydroxy group
fills the sugar pocket51 of the kinase while the pyridine substituent in R -position
2
(nomenclature see Figure 4) binds to the Lys745 in the phosphate region of the pocket.
No clash between the bulky Met790 gatekeeper residue and the benzimidazole is
observed. Important for our further optimization was the observation that the torsion angle
between the two aromatic rings of 1 is ~40°, which was found to be a low energy
conformation based on a QM-torsion angle scan (see Figure 4). Therefore, the dihedral
angle was scanned in 10° steps while optimizing all other degrees of freedom within the
ligand on B97XD/cc-pVDZ level of theory.52
A close analog (2) missing the fluorine atom in the R position (see Table 1) was found
1
L858R T790M
to show a similar activities in the biochemical and biomarker assays (IC EGFR
50
C797S
del19 T790M C797S
= 2,700 nM; IC EGFR
= 173 nM). Due to synthetic accessibility, we
50
ACS Paragon Plus Environment
15

Page 17 of 101
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
decided to move on with our further optimization program without a substituent in the R1
position.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
In the next step, we decided to address the hydroxyl group of the side chain of Thr854
via a nitrogen acceptor from the X-position of the central aromatic ring system (see Figure
4). In addition, the newly designed ligands should still have a torsion angle of 30-40° at a
low energy conformation to enable a proper interaction with Lys754. In order to access
both requirements we performed a QM-torsion angle scan for the bis aryl system of the
newly designed compounds 3, 3a and 4 (Figure 4).
ACS Paragon Plus Environment
16

Journal of Medicinal Chemistry
Page 18 of 101
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4. QM-torsion angle scans for compounds 1, 3, 3a, 4 and 5. The baseline energy
is set for each compound individually (to zero), and not deducted from the lowest energy
of all compounds.
The pyridyl nitrogen of the R moiety of compound 3 de-symmetrizes the bis aryl system
2
further and therefore a lager energy difference between the minima (30° and 160°) in
comparison to compound 1 is observed (see Figure 4). In case of the minimum at 30°,
both pyridyl nitrogens of compound 3 are located at the same side, which leads to an
unfavorable dipole-dipole interaction (overall dipole: 3.87 Debye) compared to the
ACS Paragon Plus Environment
17

Page 19 of 101
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
minimum at 160° for which the pyridyl nitrogens are located at the opposite side leading
to partial dipole cancellation (overall dipole: 2.35 Debye).
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The local minimum at 30° of compound 3 would allow the ligand to interact via the
pyridyl nitrogen of the R moiety with Lys745. However, this local minimum is
2
approximately 1 kcal/mol higher than the absolute minimum, which is located at 160°.
Therefore, an energy penalty has to be paid for the non-optimal dihedral angle, but a gain
in binding affinity was expected due to the interaction with Thr854. Interestingly, the
benzimidazole 3 showed a factor of 3-8 improved activity in the biochemical assays (IC50
EGFRL858R T790M C797S = 353 nM; IC EGFR
del19 T790M C797S
= 57 nM) and demonstrated an
50
even more pronounced gain in potency in the cellular biomarker modulation assay, while
at the same time maintaining the low affinity for wild type EGFR.
For compound 4 a similar situation was observed. A local minimum was found at a
torsion angle of 0°, while the absolute minimum is approximately 1 kcal/mol lower and is
located at 180°. Again, an energy penalty must be paid for the non-optimal dihedral angle,
but on the other hand a gain in binding affinity could be expected due to the interaction
with Thr854. Also, for compound 4 a slight improvement of activity was observed in
ACS Paragon Plus Environment
18

Journal of Medicinal Chemistry
Page 20 of 101
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
biochemical and an even more pronounced gain in the cellular biomarker modulation
assay.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
We investigated several different aromatic ring systems for the R position with the goal
2
to identify motifs, which would bind with the absolute minimum conformation regarding
the dihedral angle of the bis aryl system to EGFR. One possible solution is shown by
compound 3a, which bears a 4-methoxy-pyrid-3-yl moiety at the R position. The QM-
2
torsion angle scan (see Figure 4) for the bis aryl system showed high energy for dihedral
angle 160-180° due to unfavorable interactions of the methoxy group and the facing
pyridyl nitrogen, but an absolute minimum at 30°. The unfavorable dipole-dipole
interaction between the two ring nitrogen atoms at 30° is also given for compound 3a, but
the energy penalty is compensated by the favorable electrostatic interactions of the
methoxy group with the facing proton (see Supplementary Figure 1). Therefore the energy
at 30° of ligand 3a is lower than for compound 3. This optimized binding conformation of
ligand 3a translates in a 30-fold improved biochemical activity based of the EGFRL858R
T790M C797S
protein compared to ligand 3 and 4. The same trend can be seen comparing
the biochemical EGFRdel19 T790M C797S data. This improved biochemical activity lead to a
ACS Paragon Plus Environment
19

Page 21 of 101
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
3
5-fold improved p-EGFRdel19 T790M C797S biomarker modulation and in a 5-fold improved
anti-proliferative effect in BaF3 cells compared to ligand 3 and 4. However, the compound
displayed a clearly lower general kinase selectivity compared to compound 3 and 4 (36%
hits; see Table 1). These data show that activity can be massively gained by optimal
interaction with Lys745, but at the expense of reduced kinase selectivity.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
We hypothesized, that better kinase selectivity could be achieved, if the pocket around
Lys745 (phosphate region) is addressed by an aromatic system which bears no acceptor.
Prerequisite for this strategy was that the envisioned R moiety does not clash with
2
Lys745, which could be achieved with a dihedral angle of <20°.
To probe the hypothesis, we synthesized benzimidazole 5, which does not contain a
hydrogen bond acceptor atom in the second aromatic ring and carries only a symmetric
phenyl ring in the R -position. Therefore almost no energy difference is seen between
2
the minima (20° and 160°; see Figure 4). To our delight, we found that, while the
biochemical and cellular potency against EGFR mutated cells was basically unchanged
compared to compound 3, the selectivity was significantly improved, as it is evident from
the not observable cytotoxic activity against EGFR independent cells. At the same time,
ACS Paragon Plus Environment
20

Journal of Medicinal Chemistry
Page 22 of 101
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the kinase selectivity was found to be very high, as only mutant EGFR was hit amongst
a panel of 30 kinases. Furthermore, benzimidazole 5 also showed an excellent wt-sparing
effect based on the comparison of the anti-proliferative effect on EGFRdel19 T790M C797S
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
wt
versus EGFR in BaF3 cell lines (see Table 1).
Figure 5. X-ray structure of compound 5 (PDB: 6S9C) in complex with EGFRL858R T790M
.
The x-ray structure of compound 5 with EGFRL858R T790M shows (see Figure 5), that the
compound binds to the protein with the expected dihedral angle of the bis aryl system of
approximately 20°. In addition, the distance to the Lys745 and Glu762 is sufficiently large
(3.5 Å) to avoid a clash.
ACS Paragon Plus Environment
21

Page 23 of 101
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 1. SAR for the open-chain aminobenzimidazoles 1-5.
R1
O
H
N
X
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
N
N
R2
OH
b
BaF3 cells
BaF3 cells
a
Kinase sel. :
%
Biochemical kinase assays : IC50 [nM]
c
d
hit (#total) biom. : IC50 [nM] anit-proliferation : IC50 [nM]
d19 TM CS*
LR TM CS**
w t
d19 TM CS*
d19 TM CS*
w t
#
R1
F
X
R2
EGFR
EGFR
EGFR
@ 1 µM
p-EGFR
EGFR
EGFR EGFR indep.

3
% (30)
1
CH
250 (1.13)
173 (1.17)
57 (1.13)
2,100 (1.08) >100,000
2,700 (1.38) >100,000
353 (1.42) >100,000
790 (1.09)
nd
nd
nd
nd
nd
5
% (238)
N
N


2
3
H
H
H
CH
N
13% (30)
1,400 (1.32)
35 (1.27)
nd
1
0% (30)
3,600
(1.22)
3,500
(1.35)
270 (1.21)
51 (1.18)
1
1% (238)
N

7
00
990
(1.06)
O
#
3
a
N
≤ 3
12 (1.07)
10,000
36% (30)
10% (30)
0% (30)
1.0 (1.08)
(1.04)
N

5
(
,100
1.45)
8,900
(1.12)
4
5
H
H
N
N
72 (1.12)
35 (1.15)
519 (1.14) >100,000
267 (1.18) >100,000
43 (1.56)
22 (1.27)
333 (1.24)
N N

150 (1.29) >10,000
>25,000
Data are reported as geometric mean of at least 3 independent measurements and in
a
brackets is the geometric standard deviation as multiplicative factor given; biochemical
kinase assay using different EGFR variants, which measures the inhibitory effect of
compounds on the phosphorylation activity of EGFR enzyme forms on poly-GT substrate
in the presence of ATP; b kinase selectivity was determined using the “SelectScreen
Kinase Profiling Services” from ThermoFisher; c measurement of inhibition of the p-
EGFRdel19 T790M C797S signal in BaF3 cells after 4 h compound treatment; dBaF3 cells
bearing EGFRdel19 T790M C797S protein were treated for 3 days with compound, the effect on
ACS Paragon Plus Environment
22

Journal of Medicinal Chemistry
Page 24 of 101
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
proliferation was measured; nd means not determined; *EGFRd19 TM CS means EGFRdel19
T790M C797S; **EGFRLR TM CS means EGFR L858R T790M C797S; #activity of compound at or below
assay wall (assay wall around 3 nM due to enzyme concentration/activity and substrate
concentration).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Entropy driven optimization: Rigidification through macrocyclization
Compound 5 represented a selective but not yet sufficiently potent compound (Table 1,
11
Entry 6). In analogy to the discovery of lorlatinib, we chose reduction of ligand entropy
through macrocyclization as the strategy to improve potency.53-55 We pursued both
theoretical and experimental methods to assist the design of cyclized compounds.
Initially, as for the computational conformation analysis of ligand 5, the MOE module
56
LowModeMD , followed by semi empirical (MOPAC, AM1) energy minimization of all the
conformers in vacuo was employed for the analysis and selection of low-energy
conformers. In the course of the project, we then complemented this approach with
57
enhanced sampling molecular dynamics methods in explicit solvation . These findings
were supported by a detailed 2D NMR analysis of compound 5 which showed ROE
correlations between protons, as indicated in Scheme 1. The ROEs support the presence
ACS Paragon Plus Environment
23

Page 25 of 101
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
of both active and inactive conformational states in the population observed in solution.
This was further corroborated by QM-DFT 13C, H and partial 15N chemical shift
predictions which support a Boltzmann-weighted energy distribution, with significant
population of active and inactive conformers.57, 58 We therefore expected a significant
improvement of binding affinity to EGFRdel19 T790M C979S if the molecule could be
constrained in its “active” binding conformation.
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
O
O
N
H
N
H
N
N
N
N
N
N
5
5
"
OH
OH
active" conformation
"inactive" conformation
Scheme 1. 2-Dimensional representation of the most prevalent conformations of ligand 5
as computationally predicted and detected in solution by 2D-NMR. Important ROE-
correlations are displayed as arrows.
Starting from ligand 5 as a template for the active conformation, several different
macrocyclic variants with varying linker lengths and attachment points were designed.
The macrocycles were then virtually subjected to conformational searches (MOE
ACS Paragon Plus Environment
24

Journal of Medicinal Chemistry
Page 26 of 101
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
LowModeMD),56 followed by AM1 energy minimization of all conformers. Low energy
conformations were compared with the binding conformation of amino benzimidazole 5
and macrocyclic ligand 6, with a C H O linker was found to display a good overlap with
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
10
the desired “active” conformation (see Figure 6).
Figure 6. Overlay of x-ray-based structure of compound 5 (orange) and calculated
minimum conformation structure of macrocyclic ligand 6 (green).
The first macrocyclic ligand 6 displayed a biochemical activity in the EGFRdel19 T790M
C797S
58
assay at the assay wall , which hampers the evaluation of the activity improvement.
Fortunately, the biochemical activity of macrocycle 6 in the EGFRL858R T790M C797S assay is
still above the assay wall (see Supplementary Table 8, entry 1) and therefore a 17 fold
improved biochemical activity over the open chain ligand 5 can be determined. The gain
ACS Paragon Plus Environment
25

Page 27 of 101
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
in biochemical activity translates in a 5-fold improvement of cellular potency over the open
chain ligand 5 as is evident from the inhibition of the p-EGFR signal and antiproliferative
effect on EGFRdel19 T790M C979S dependent BaF3 cells (see Table 2, entry 1). Noteworthy
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
wt
for compound 6 was a favorable EGFR and general kinase selectivity (0/30 kinases
50
were hit @ 1µM) observed.
Figure 7. X-ray structure of compound 6 (PDB: 6S9D) in complex with EGFRL85R T790M
.
From the protein x-ray structure (see Figure 7), it is evident that the observed potency
improvement resulted mostly from the conformational restrictions as no new or changed
ACS Paragon Plus Environment
26

Journal of Medicinal Chemistry
Page 28 of 101
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
interactions with the EGFR protein are visible. Also, the observed binding conformation
of ligand 6 is in excellent agreement with the computationally predicted minimum
conformation.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
From the predicted minimum energy conformation of ligand 6 it was expected that the
bis aryl torsion angle in the macrocyclic ligand should be constrained to approx.~40°, a
value which is very close to the observed angle in the x-ray structure (39°) (see Figure
7). This indicated that a re-installment of a hydrogen bond acceptor in the second
aromatic ring should lead to a favorable hydrogen bond formation to Lys745. Therefore,
we decided to further improve the potency of the molecule by implementing a polar
interaction with Lys745, as it was seen in ligands 1-4. We chose the N-methylpyrazole
from ligand 4 rather than the pyridine moiety of compound 3, because the metabolic
stability predicted from human hepatocytes was superior for ligand 4 (%QH: 12%) over
ligand 3 (%QH: 34%). Additionally, a fluorine atom was reinstalled in the -position (R1)
of the pyridine nitrogen to fill the observed small sub-pocket in the protein x-ray structure
of 6 with the EGFR protein (see Figure 7). We hypothesized that this substituent might
ACS Paragon Plus Environment
27

Page 29 of 101
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
be in part responsible for the excellent general kinase selectivity of ligand 1. These
structural changes resulted in ligand 7, which showed at least an additional 5-fold
improved biochemical activity in the EGFRL858R T790M C797S assay. It is noteworthy that
macrocycle 7 also shows activity around the assay wall in this biochemical assay (see
Supplementary Table 8, entry 2). This prevents the use of biochemical EGFR assays to
determine further potency improvements. Also, a 5-fold improved antiproliferative effect
against EGFRdel19 T790M C797S dependent cells of ligand 7 versus 6 can be seen. At the
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
wt
same time, we observed a sub-micromolar antiproliferative effect against EGFR and
EGFR independent cells for the first time. However, the ratio of the IC values remained
50
wt
del19 T790M C797S
unchanged between ligand 6 (IC EGFR / IC EGFR
= 48) and ligand 7
50
50
wt
del19 T790M C797S
(
IC EGFR / IC EGFR
= 45) suggesting a generally improved kinase
50
50
activity instead of a specific improvement for mutant EGFR. The size of the substituent at
the R -position was increased from fluorine to methyl which gave rise to compound 8.
1
del19 T790M C797S
While the cellular potency of aminobenzimidazole 8 (IC EGFR
= 5 nM)
50
del19 T790M C797S
was found to be unchanged compared to ligand 7 (IC EGFR
= 6 nM), the
50
selectivity for EGFR mutant dependent cells over EGFRwt dependent cell or EGFR
ACS Paragon Plus Environment
28