4-Aminoindazolyl-dihydrofuro[3,4-d]pyrimidines as non-covalent inhibitors of mutant epidermal growth factor receptor tyrosine kinase

Article abstract of DOI:10.1016/j.bmcl.2015.11.078

The treatment of epidermal growth factor receptor (EGFR)-driven non-small cell lung cancers with the T790M resistance mutation remains a significant unmet medical need. We report the identification of 4-aminoindazolyl-dihydrofuro[3,4-d]pyrimidines as non-covalent inhibitors of EGFR, with excellent activity against the T790M resistance double mutants and initial single activating mutants. Using an optimization strategy focused on structure-based design and improving PK properties through metabolite identification, we obtained advanced leads with high oral exposure.

Full text of DOI:10.1016/j.bmcl.2015.11.078

Accepted Manuscript
4-Aminoindazolyl-dihydrofuro[3,4-d]pyrimidines as non-covalent inhibitors of
mutant epidermal growth factor receptor tyrosine kinase
Emily J. Hanan, Matt Baumgardner, Marian C. Bryan, Yuan Chen, Charles Eigenbrot, Peter
Fan, Xiao-Hui Gu, Hank La, Shiva Malek, Hans E. Purkey, Gabriele Schaefer, Stephen
Schmidt, Steve Sideris, Ivana Yen, Christine Yu, Timothy P. Heffron
PII:
S0960-894X(15)30286-9
DOI:
http://dx.doi.org/10.1016/j.bmcl.2015.11.078
BMCL 23334
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
27 October 2015
17 November 2015
20 November 2015
Please cite this article as: Hanan, E.J., Baumgardner, M., Bryan, M.C., Chen, Y., Eigenbrot, C., Fan, P., Gu, X-H.,
La, H., Malek, S., Purkey, H.E., Schaefer, G., Schmidt, S., Sideris, S., Yen, I., Yu, C., Heffron, T.P., 4-
Aminoindazolyl-dihydrofuro[3,4-d]pyrimidines as non-covalent inhibitors of mutant epidermal growth factor
receptor tyrosine kinase, Bioorganic & Medicinal Chemistry Letters (2015), doi: http://dx.doi.org/10.1016/j.bmcl.
2015.11.078
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Bioorganic & Medicinal Chemistry Letters
4-Aminoindazolyl-dihydrofuro[3,4-d]pyrimidines as non-covalent inhibitors of
mutant epidermal growth factor receptor tyrosine kinase
Emily J. Hanan*a, Matt Baumgardnera, Marian C. Bryana, Yuan Chena, Charles Eigenbrota, Peter Fana,
Xiao-Hui Gub, Hank Laa, Shiva Maleka, Hans E. Purkeya, Gabriele Schaefera, Stephen Schmidta, Steve
Siderisa, Ivana Yena, Christine Yua and Timothy P. Heffrona
aGenentech, Inc., 1 DNA Way, South San Francisco, CA, 94080, United States
bPharmaron Beijing CO. Ltd, No. 6 Taihe Road BDA, Beijing, 100176, China
ARTICLE INFO
ABSTRACT
Article history:
Received
Revised
Accepted
Available online
The treatment of epidermal growth factor receptor (EGFR)-driven non-small cell lung cancers
with the T790M resistance mutation remains a significant unmet medical need. We report the
identification of 4-aminoindazolyl-dihydrofuro[3,4-d]pyrimidines as non-covalent inhibitors of
EGFR, with excellent activity against the T790M resistance double mutants and initial single
activating mutants. Using an optimization strategy focused on structure-based design and
improving PK properties through metabolite identification, we obtained advanced leads with
high oral exposure.
Keywords:
dihydrofuropyrimidine
epidermal growth factor receptor
EGFR inhibitor
2009 Elsevier Ltd. All rights reserved.
Non-covalent
Non-small cell lung cancers (NSCLC) bearing activating
mutations in the tyrosine kinase domain of the epidermal growth
factor receptor (EGFR) are sensitive to treatment with the
tyrosine kinase inhibitors (TKIs) erlotinib, gefitinib, and
afatinib.1-5 These activating mutations, most commonly the point
mutation L858R or deletions within exon 19 (e.g. residues 746-
750), increase EGFR-driven cell proliferation and survival.6-8
Unfortunately, successful treatment of EGFR-mutated NSCLC
with first or second generation TKIs9 is followed by an acquired
resistance, with approximately 60% of that resistance arising
from a secondary mutation within the EGFR kinase domain,
substitution of the gatekeeper residue Thr790 with methionine
(T790M).10-15 Advances in third-generation EGFR inhibitors have
demonstrated the ability to overcome this T790M-mediated
resistance. Third-generation inhibitors AZD9291 and CO1686
both form a covalent bond with Cys797 in the active site and
have activity against the resistance mutations (T790M/L858R,
TMLR; T790M/del(746-750), TMdel) as well as the two initial
activating mutations.16-21 However, recent reports of the C797S
mutation emerging in the clinic conferring resistance to
AZD9291 support the continued need for a T790M inhibitor that
does not depend on a covalent interaction with Cys797.22, 23 We
report here the identification of 4-aminoindazolyl-
dihydrofuro[3,4-d]pyrimidines as non-covalent inhibitors of the
T790M containing EGFR mutants as well as the initial single
activating mutants.
activity in a cell-based assay24 (H1975 pEGFR IC50 = 0.60 M).
While possessing excellent biochemical activity for an initial hit,
several parameters required optimization. Compound 1 is
moderately stable in human liver microsomes (HLMs) and
human hepatocytes (HHeps) and is stable in rat liver microsomes
and hepatocytes (RLMs and RHeps). However, it is rapidly
cleared in rats in vivo (Cl = 86 mL/min/kg) and exhibits low oral
bioavailability (F = 13%, Table 2). The high clearance of 1 raised
concerns of underprediction of clearance by in vitro metabolic
stability data for this chemical series. Permeability was moderate
as measured in Madin-Darby
cells, with minimal active efflux. We hypothesized that this low
oral exposure was result of pre-systemic metabolism.
canine kidney (MDCK)
a
Therefore, our optimization program focused on identification of
inhibitors with retained potency and improved oral PK
properties.
Compounds described in Tables 1-4 were prepared as shown
in Scheme 1.25 Non-commercial 3-aminoindazoles were obtained
through reaction of ortho-fluorobenzonitriles with hydrazine.
SnAr reaction of these aminoindazoles with 2,4-dichloro-7,7-
dimethyl-5,7-dihydrofuro[3,4-d]pyrimidine26
afforded
the
penultimate intermediates. Suzuki coupling with aryl boronates
provided the final test compounds 1-28.
Metabolite identification studies carried out using plasma
samples from rats dosed intravenously with 1 indicated that
oxidation of the indazole ring was a primary circulating
metabolite. We were also aware of the high potential of indazole
systems to be subject to N-glucuronidation. We therefore set out
A high-throughput screen of our internal library identified 4-
aminoindazolyl-dihydrofuro[3,4-d]pyrimidine 1 as
a potent
TMLR inhibitor with a Ki of 2 nM (Table 1) and moderate

to identify indazole ring modifications or replacements that might
mitigate these metabolic liabilities with compounds described in
Table 1. In light of the underprediction of in vivo clearance by in
vitro systems for 1, we considered low turnover in LMs and Heps
as a minimum criterion for achieving low to moderate in vivo
clearance.
Table 1. Indazole modification
R1
HN
N
O
N
N
NH
Cmpd
R1
EGFR
(TMLR) Ki
(nM)
HLM / RLM
Clhep
Cmpd
R1
EGFR
(TMLR) Ki
(nM)
HLM/RLM Clhep
(mL/min/kg)a
(mL/min/kg)a
H
N
H
N
N
N
1
11
2
7 / 14
8
7 / 17
F
H
N
H
N
N
2
3
12
13
N
N
844
389
4 / 21
4 / 19
17
6 / 17
5 / 24
F
H
N
H
N
CF3
N
N
N
Cl
134
Cl
H
N
H
N
4
5
6
14
15
16
N
>2000
17
8 / 16
>2000
7 / 22
5 / 23
13 / 20
CF3
H
N
H
N
Cl
12 / 34
18 / 36
2
H
N
S
N
N
N
747
16
Cl
H
N
F
7
8
9
17
18
19
>2000
>2000
3
4 / 8
5
9 / 21
7 / 16
7 / 22
N
N
S
F
H
N
H
N
Cl
N
N
12 / 14
7 / 19
4
F
N
H
N
H
N
Cl
F
N
N
16
F
H
N
N
10
20
F
N
2
11 / 21
57
11 / 23
F
aHuman and rat liver microsome predicted hepatic clearance.27

Cl
N
NH
NH
N
N
N
N
N
O
O
O
Ar
H
B
X
N
Cl
HN
N
HN
N
NH2NH2
N
R
R
R
R
nBuOH, 100 oC
DIPEA, DMF, 60-80 o
C
Pd(dppf)Cl2, K2CO3
dioxane/water
130 oC, uW
O
O
N
H2N
Cl
Ar
X = F, Cl
Scheme 1. Synthesis of 4-aminoindazolyl-dihydrofuro[3,4-d]pyrimidines
Table 2. in vitro properties and PK parameters
Cmpd
HHep / RHep
Clhep
MDCK AB
Papp (x10-6
cm/sec)b
MDCK
Rat plasma Cl
(mL/min/kg)
Rat Vss (L/kg)
Rat POb %F
Rat POd
AUCtotal
(uM*hr)
Rat POd
AUCfree
(uM*hr)
BA/AB efflux
(mL/min/kg)a
ratio
1
11 / 9
6.7
2.4
86e
81f
6.0
6.8
2.4
1.2
3.6
5.5
1.5
2.3
3.3
13%
5%
0.37
0.05
11
13
14
15
18
20
27
28
12 / 13
11 / 16
9 / 14
8.5
1.3
1.5
1.6
1.1
27 f
91 f
74 f
60 f
60 f
38 f
36 f
0.36
0.03
10.3
9.5
14 / 18
7 / 16
7.3
15 / 18
15 / 17
17 / 18
12.9
8.4
0.9
0.7
117%
62%
7.5
3.4
0.23
0.09
aHuman and rat hepatocyte predicted hepatic clearance.27
bApparent permeability (Papp) from the apical (A) to basolateral (B) direction in MDCK cells.
cratio of Papp BA to Papp AB in MDCK cells.
dPO dosed at 5.0 mg/kg as a suspension in MCT.
eIV dosed as a solution in 10/40/50 ethanol/PEG400/citrate pH 5 buffer
fIV dosed as a solution in 50/35/15 Cremophor EL/NMP/Miglyol®
We determined that the truncated pyrazole 2 lost most of the
activity of the indazole, and could not be recovered through the
use of dichloro-substitution as a phenyl isostere (compound 3).
Saturation of the phenyl ring, which would be desirable to reduce
planarity and improve overall physiochemical properties was also
not tolerated (compound 4). Indole 5 was slightly less active than
the indazole (17 nM vs. 2 nM), however benzoisothiazole 6 lost
nearly 400-fold activity. These data suggest that both hydrogen
bond donor and acceptor of the indazole are important for
binding, with a more significant contribution from the hydrogen-
bond donor. Benzothiazole 7 improved HLM and RLM stability,
but was completely inactive.
18 resulted in only a slight reduction in clearance (Cl = 60
mL/min/kg). While 7-trifluoromethyl substitution (13) afforded a
reduction in plasma clearance to 27 mL/min/kg, unfortunately,
oral bioavailability remained low (F = 5%). Conversely, 6-
trifluoromethyl substitution (14) resulted in high clearance.
Metabolite ID studies were carried out in rat and human
hepatocytes for compounds 11, 15 and 13. Phase I and phase II
metabolites were observed for all compounds, with
glucuronidation as a major metabolite in both rat and human
hepatocytes (site of glucuronidation not determined). Oxidative
metabolism occurred largely on the dihydrofuryl moiety, with
oxidation on the indazole ring to only a minor degree, indicating
that the strategy of introducing electron withdrawing substituents
had been successful in reducing oxidative metabolism on the
indazole ring. While the 7-trifluoromethyl substitution was not
sufficient to entirely block glucuronidation, the decreased in vivo
clearance of 13 relative to 6-trifluoromethyl 14 suggests that the
steric hindrance of the 7-trifluoromethyl group may be able to
reduce the rate of indazole N-glucuronidation. To avoid indazole
N-glucuronidation altogether, the indazole N-methylated
compound was prepared (20), but found to have a 19-fold
reduction in activity, and remained only moderately stable in
both rat and human LM and Heps, with high clearance in vivo
(Cl = 60 mL/min/kg) (Tables 1 and 2). The potential decrease in
glucuronidation may be counteracted by the introduction of a
metabolic soft-spot (potential for oxidative N-demethylation). In
general, all compounds tested displayed higher in vivo clearance
than predicted by in vitro microsomal or hepatocyte systems;
however, good passive permeability was maintained.
We attempted to modulate metabolism of the indazole ring
through the introduction of an additional heteroatom to reduce
electron density, however pyrazolopyrimidine 8 lost all activity.
Substitution of the indazole ring with electron-withdrawing
substituents was more fruitful; compounds 9, 10, 11, and 12
showed that fluoro substitution was tolerated with minimal loss
of activity, with substitution at the 6- or 7- position affording
compounds equipotent with starting point 1.
Unfortunately, mono-fluoro substitution at any position failed
to provide an improvement in HLM or RLM stability. Larger
groups such as chloro or trifluoromethyl were moderately
tolerated only at the 7-position (13 and 15), with larger drops in
potency observed for substitution at the 6-position (14 and 16).
No improvement in HLM or RLM stability was observed with
either chloro or trifluoromethyl substitution. Di-substitution with
electron withdrawing groups was evaluated (compounds 17-19)
and while activity was retained, clearance predicted by HLM and
RLM remained in the moderate to high range.
To guide further optimization of this series, we sought an X-
ray structure with TMLR, resulting in the structure with 15
shown in Figure 1. The pyrazole ring functions as the hinge
binding element, adopting a position consistent with hydrogen-
bonds with main chain atoms of Gln791 (C=O) and Met793
(NH). The dihydrofuropyrimidine core is situated in the center of
the active site, with the gem-dimethyl groups occupying the
Several compounds were profiled in vivo in rat PK studies to
determine if their in vitro profiles correlated with in vivo findings
(Table 2). Compounds with single fluoro or chloro substitution
(11 and 15) were observed to have high clearance similar to the
starting point 1 and above expected clearance based on liver
microsome and hepatocyte incubations. Di-substituted compound

lipophilic ribose pocket and lipophilic contacts between the
pyrimidine ring and the side chain of Val726. Intriguingly, the
methionine gatekeeper is not in the extended conformation
observed for other inhibitors as previously published (see
supplemental Figure S1).28 Instead, the Met790 sidechain is
folded down to form a three-dimensional lipophilic pocket that is
filled by the indazole, with the aromatic ring sandwiched
between the sidechains of Met790 and Lys745. This is
reminiscent of the binding site observed for erlotinib in wild-type
EGFR, where the shorter Thr790 side chain forms the boundary
of a similar pocket (see supplemental Figure S2).29 Finally, the
indazole N1 nitrogen is positioned consistent with a hydrogen-
bonding interaction with Glu762. While the indazole N2 nitrogen
does not make any direct interactions with the protein, it resides
in a polar environment and there is the possibility of water-
mediated interactions not visible at this resolution (2.8 Å). In
light of this binding mode and the tight lipophilic pocket
occupied by the indazole phenyl ring, the limited tolerance
observed for substitution or addition of polar heteroatoms on the
indazole ring is easily rationalized.
all resulted in a substantial reduction in potency, consistent with
the calculations, and with no concomitant improvement in LM or
Hep stability.
Table 3. Linker modification
NH
N
X1
N
O
N
N
NH
Cmpd
X1
EGFR
(TMLR) Ki
(nM)
HLM/RLM
Clhep
HHep/RHep
Clhep
(mL/min/kg)a
(mL/min/kg)b
21
22
23
NCH3
O
519
85
16 / 26
10 / 32
13 / 36
16 / 33
C=O
104
aHuman and rat liver microsome predicted hepatic clearance.27
bHuman and rat hepatocyte predicted hepatic clearance.27
We therefore turned our attention to the pyrazole hinge binder.
Triazole 24 lost nearly all TMLR activity; despite providing the
required complement of hydrogen-bonding partners, lone-pair
repulsion between the triazole and pyrimidine nitrogens may
prevent the desired co-planar conformation. Thiazole 25 is well
tolerated (TMLR Ki = 5.6 nM), suggesting that an aromatic CH
hydrogen bond donor is sufficient to retain activity.30 Pyrazole
regioisomer 26 also offers a CH hydrogen bond donor but is
much less active at 175 nM, possibly due to the low-energy
tautomer placing the hydrogen bond acceptor at the wrong
location. However, this pyrazole regioisomer exhibits
significantly improved stability in both HLMs and RLMs. N-
methylated pyrazole 27 enforces the desired tautomer and regains
5-fold activity. An X-ray structure of 27 with TMLR confirms
the expected binding mode and pyrazole positioning consistent
with an aromatic CH hydrogen bond with Gln791 (C=O) (See
supplemental Figure S3).
Compound 27 has reduced HLM stability relative to 26, but
retains the good stability in RLMs seen for the des-methyl
pyrazole. When tested in vivo in rats, 27 was found to have
moderate clearance (Cl = 38 mL/min/kg) and excellent oral
bioavailability (F ~ 100%) with an oral AUC 20-fold higher than
that of compound 1 (unbound AUC 5-fold higher than compound
1, Table 2). These results implicate glucuronidation of the
pyrazole hinge-binder as a likely driver of the low oral
bioavailability observed for compounds 1 and 13. While 27 has
an appealing in vivo exposure profile, the 34 nM Ki translated
into only marginal activity in a cell-based assay (H1975 pEGFR
IC50 = 3 µM).
A combination of difluoromethyl pyrazole ring substitution
for electronic and steric deactivation together with an electron-
withdrawing substituent on the indazole ring afforded compound
28. This compound exhibits biochemical activity similar to the
starting point 1. The indazole and pyrazole substitution are
expected to have an impact on both oxidative and conjugative
metabolism; while 28 has only moderate stability in human and
rat microsomes in vitro, gratifyingly, when tested in vivo 28
exhibits significantly reduced clearance (Cl = 36 mL/min/kg),
high oral bioavailability (F = 62%) and satisfactory oral AUC.
Metabolite identification experiments conducted on rat urine and
bile samples from animals dosed intravenously with 28
confirmed that oxidative and glucuronide metabolites are
predominant excretion products, with no 28 excreted unchanged.
Figure 1. a) X-ray structure of 15 complexed with TMLR; b) ligand-
protein interaction map. PDB accession 5EDQ.
Given that the NH linker of 15 does not appear to be making
any interactions, we attempted to remove that possible site where
phase II conjugation could occur. Conformational analysis in
Macromodel v9.9 (Schrödinger, LLC, New York, NY) suggested
that there was not a replacement for the –NH– linkage that gave
the same conformation between rings, although several provided
angles within 20-30° of the crystal conformation and 0.5-2.3
kcal/mol of their global minimum conformations (see
Supplemental Table S1). As shown in Table 3, N-methylation or
replacement with an ether or ketone linkage (compounds 21-23)

This observation confirms the primary role of metabolism as the
main elimination pathway, as opposed to active transport or
direct excretion of the parent molecule. Interestingly, the
increased bioavailability for 28 relative to 13 is not predicted
based on similar in vitro metabolic stability or MDCK
permeability for these two compounds (Table 2). This may be
due to contribution of non-hepatic metabolism (e.g.,
glucuronidation in the gut) or a difference in absorption due to
dissolution or solubility. While the two compounds have similar
kinetic solubility (3-5 M), thermodynamic solubility was not
measured. Compound 28 also exhibits moderate activity in a cell-
based assay (H1975 pEGFR IC50 = 0.37 µM)24, slightly improved
relative to the starting point 1, thereby achieving our initial goal
of an improved pK profile without sacrificing activity.
While these compounds were generally equipotent against the
wild-type enzyme and the activating mutants in a biochemical
assay, the activating mutants’ decreased affinity for ATP is
expected to afford selectivity in a cell based setting, similar to
that observed for erlotinib.31,32 However, the T790M resistance
mutation restores ATP affinity to a similar level as wtEGFR, and
increased biochemical selectivity for TMLR and TMdel relative
to wtEGFR would be desirable to prevent wtEGFR-related
toxicities.33 Biochemical selectivity against a larger panel of
kinases was also determined for several Table 5 compounds.
Compound 1 displayed good selectivity, inhibiting only 14 out of
296 kinases at > 80% when tested at 100 nM, approximately
100x the EGFR Ki. Tested at a higher concentration of 1.0 M
against an abbreviated panel, compound 28 inhibited 4 out of 16
kinases at > 80%. See Supplementary Tables S3-4 for kinase
panel data.
Table 4. Hinge binder modification
NH
R2
N
Table 5. Biochemical activity against wild-type EGFR and
EGFR mutants
HN
N
Cmpd
wt-
EGFR
(L858R)
Ki (nM)
EGFR
(del746-
750) Ki
(nM)
EGFR
(TMLR)
Ki (nM)
EGFR
(TMdel)
Ki (nM)
EGFR
Ki
(nM)
N
O
R3
1
0.6
1.5
1.1
6.5
0.5
0.2
2.0
1.6
1.3
21.3
2.5
95
Cmpd
R2
R3
EGFR
(TMLR) Ki
(nM)
HLM/RLM
Clhep
15
27
28
1.3
0.4
2.1
(mL/min/kg)a
< 0.5
3.4
< 0.5
1.7
34.3
9.0
N
24
25
26
27
H
H
H
H
732
5.6
9 / 7
N
erlotinib 0.1
0.3
0.1
123
N
H
N
AZD-
9291
3.5
3.2
2.2
13 / 24
3 / 9
S
175
34.3
In conclusion, beginning with a potent high-throughput
screening hit, we identified a novel series of non-covalent 4-
aminoindazolyl-dihydrofuro[3,4-d]pyrimidines as inhibitors of
EGFR. An optimization strategy driven by metabolite
identification, PK, and structure-based design provided advanced
compounds with improved oral PK properties. Their excellent
activity against the resistance T790M double mutants as well as
the single activating mutants renders these molecules promising
leads for the treatment of EGFR-mutant NSCLC.
N
N
N
H
10 / 9
N
28
N
Cl
9.0
10 / 22
F
N
H
F
aHuman and rat liver microsome predicted hepatic clearance.27
Acknowledgments
Several key compounds from this chemical series were tested
against all relevant EGFR mutants and the wild-type protein
(wtEGFR) as shown in Table 5. All compounds tested displayed
an intriguing pan-inhibitor profile, strongly inhibiting both single
activating mutations and the T790M double mutants. The
biochemical activity of compounds in Table 5 compares
favorably to other pan-mutant inhibitors currently in clinical
trials, including AZD9291, suggesting the potential of this series
of inhibitors to provide a clinical benefit without a requisite
covalent interaction with Cys797. The ability of these
compounds to inhibit both Thr790- and Met790-containing
proteins may be rationalized by their binding mode. As shown in
Figures 1 and S2, the indazole ring makes favorable lipophilic
contacts with the Met790 sidechain that would not be available
with the Thr790 sidechain. Similar to structures with different
classes of inhibitors (e.g. Supplemental Figure S1), an apo
TMLR structure places the Met790 sidechain in an extended
conformation incompatible with inhibitor binding (see
Supplemental Figure S4). The possible energy penalty associated
with Met790 sidechain folding to accommodate the inhibitors
may explain the similar Ki values across both Thr790 and
Met790 proteins.
The authors thank Zhibo Hua, Chengling Pang, Huojian
Wang, Jianfeng Xin and Tieguang Yao for the synthesis of some
test compounds, the analytics groups at Genentech and
Pharmaron for purification support, and Teresa Mulder and Hoa
Le for their contributions to metabolite ID studies. We also thank
Robert Heald, Eileen Seward, and Bryan Chan for expert advice
and discussion.
Supplementary Material
Supplementary material associated with this article may be
found, in the online version, at XXX. PDB accession codes are as
follows: 15/TMLR, 5EDQ; 27/TMLR, 5EDR; apo-TMLR,
5EDP.
References and notes
1.
Lynch, T. J.; Bell, D. W.; Sordella, R.; Gurubhagavatula,
S.; Okimoto, R. A.; Brannigan, B. W.; Harris, P. L.; Haserlat, S. M.;
Supko, J. G.; Haluska, F. G.; Louis, D. N.; Christiani, D. C.;
Settleman, J.; Haber, D. A. N. Eng. J. Med. 2004, 350, 2129.

2.
Paez, J. G.; Janne, P. A.; Lee, J. C.; Tracy, S.; Greulich,
20.
Tjin Tham Sjin, R.; Lee, K.; Walter, A. O.; Dubrovskiy,
H.; Gabriel, S.; Herman, P.; Kaye, F. J.; Lindeman, N.; Boggon, T.
J.; Naoki, K.; Sasaki, H.; Fujii, Y.; Eck, M. J.; Sellers, W. R.;
Johnson, B. E.; Meyerson, M. Science 2004, 304, 1497.
A.; Sheets, M.; Martin, T. S.; Labenski, M. T.; Zhu, Z.; Tester, R.;
Karp, R.; Medikonda, A.; Chaturvedi, P.; Ren, Y.; Haringsma, H.;
Etter, J.; Raponi, M.; Simmons, A. D.; Harding, T. C.; Niu, D.;
Nacht, M.; Westlin, W. F.; Petter, R. C.; Allen, A.; Singh, J. Mol.
Cancer Ther. 2014, 13, 1468.
3.
Pao, W.; Miller, V.; Zakowski, M.; Doherty, J.; Politi, K.;
Sarkaria, I.; Singh, B.; Heelan, R.; Rusch, V.; Fulton, L.; Mardis, E.;
Kupfer, D.; Wilson, R.; Kris, M.; Varmus, H. Proc. Natl. Acad. Sci.
U.S.A. 2004, 101, 13306.
21.
Janne, P. A.; Yang, J. C.; Kim, D. W.; Planchard, D.; Ohe,
Y.; Ramalingam, S. S.; Ahn, M. J.; Kim, S. W.; Su, W. C.; Horn, L.;
Haggstrom, D.; Felip, E.; Kim, J. H.; Frewer, P.; Cantarini, M.;
Brown, K. H.; Dickinson, P. A.; Ghiorghiu, S.; Ranson, M. N. Eng.
J. Med. 2015, 372, 1689.
4.
Jackman, D. M.; Miller, V. A.; Cioffredi, L.-A.; Yeap, B.
Y.; Janne, P. A.; Riely, G. J.; Ruiz, M. G.; Giaccone, G.; Sequist, L.
V.; Johnson, B. E. Clin. Cancer Res. 2009, 15, 5267.
5.
Rosell, R.; Moran, T.; Queralt, C.; Porta, R.; Cardenal, F.;
22.
Thress, K. S.; Paweletz, C. P.; Felip, E.; Cho, B. C.;
Camps, C.; Majem, M.; Lopez-Vivanco, G.; Isla, D.; Provencio, M.;
Insa, A.; Massuti, B.; Gonzalez-Larriba, J. L.; Paz-Ares, L.; Bover,
I.; Garcia-Campelo, R.; Moreno, M. A.; Catot, S.; Rolfo, C.;
Reguart, N.; Palmero, R.; Sanchez, J. M.; Bastus, R.; Mayo, C.;
Bertran-Alamillo, J.; Molina, M. A.; Sanchez, J. J.; Taron, M. N.
Eng. J. Med. 2009, 361, 958.
Stetson, D.; Dougherty, B.; Lai, Z.; Markovets, A.; Vivancos, A.;
Kuang, Y.; Ercan, D.; Matthews, S. E.; Cantarini, M.; Barrett, J. C.;
Janne, P. A.; Oxnard, G. R. Nat. Med. 2015, 21, 560.
23.
Yu, H. A.; Tian, S. K.; Drilon, A. E.; Borsu, L.; Riely, G.
J.; Arcila, M. E.; Ladanyi, M. J. Am. Med. Assoc. Oncol. 2015, 1,
982.
6.
Sordella, R.; Bell, D. W.; Haber, D. A.; Settleman, J.
24.
Measurement of reduction in autophosphorylation of
Science 2004, 305, 1163.
EGFR in H1975 cells, a TMLR-driven cell line. A large shift in
biochemical to cell potency is expected, driven by the low TMLR Km
for ATP (~1-2 M).
7.
8.
Gazdar, A. F.; Minna, J. D., PLoS medicine 2005, 2, e377.
Sharma, S. V.; Bell, D. W.; Settleman, J.; Haber, D. A.
Nat. Rev. Cancer 2007, 7, 169.
25.
See Supplementary Material for representative synthetic
9.
First generation TKIs are non-covalent inhibitors (e.g.
procedures.
erlotinib and gefitinib); second generation TKIs are covalent
inhibitors (e.g. afatinib).
26.
Cohen, F.; Bergeron, P.; Blackwood, E.; Bowman, K. K.;
Chen, H.; DiPasquale, A. G.; Epler, J. A.; Koehler, M. F. T.; Lau, K.;
Lewis, C.; Liu, L.; Ly, C. Q.; Malek, S.; Nonomiya, J.; Ortwine, D.
F.; Pei, Z.; Robarge, K. D.; Sideris, S.; Trinh, L.; Truong, T.; Wu, J.;
Zhao, X.; Lyssikatos, J. P. J. Med. Chem. 2011, 54, 3426.
10.
Kobayashi, S.; Boggon, T. J.; Dayaram, T.; Janne, P. A.;
Kocher, O.; Meyerson, M.; Johnson, B. E.; Eck, M. J.; Tenen, D. G.;
Halmos, B. N. Eng. J. Med. 2005, 352, 786.
11.
Pao, W.; Miller, V. A.; Politi, K. A.; Riely, G. J.; Somwar,
27.
Obach, R. S.; Baxter, J. G.; Liston, T. E.; Silber, B. M.;
R.; Zakowski, M. F.; Kris, M. G.; Varmus, H. PLoS Med. 2005, 2,
e73.
Jones, B. C.; MacIntyre, F.; Rance, D. J.; Wastall, P. J. Pharmacol.
Exp. Ther. 1997, 283, 46.
12.
Balak, M. N.; Gong, Y.; Riely, G. J.; Somwar, R.; Li, A.
28.
Hanan, E. J.; Eigenbrot, C.; Bryan, M. C.; Burdick, D. J.;
R.; Zakowski, M. F.; Chiang, A.; Yang, G.; Ouerfelli, O.; Kris, M.
G.; Ladanyi, M.; Miller, V. A.; Pao, W. Clin. Cancer Res. 2006, 12,
6494.
Chan, B. K.; Chen, Y.; Dotson, J.; Heald, R. A.; Jackson, P. S.; La,
H.; Lainchbury, M. D.; Malek, S.; Purkey, H. E.; Schaefer, G.;
Schmidt, S.; Seward, E. M.; Sideris, S.; Tam, C.; Wang, S.; Yeap, S.
K.; Yen, I.; Yin, J.; Yu, C.; Zilberleyb, I.; Heffron, T. P. J. Med.
Chem. 2014, 57, 10176.
13.
Kosaka, T.; Yatabe, Y.; Endoh, H.; Yoshida, K.; Hida, T.;
Tsuboi, M.; Tada, H.; Kuwano, H.; Mitsudomi, T. Clin. Cancer Res.
2006, 12, 5764.
29.
Stamos, J.; Sliwkowski, M. X.; Eigenbrot, C. J. Biol.
14.
H.; Ahn, M. J.; Park, K. Mol. Cancer Ther. 2012, 11, 784.
15. Yu, H. A.; Arcila, M. E.; Rekhtman, N.; Sima, C. S.;
Kim, Y.; Ko, J.; Cui, Z.; Abolhoda, A.; Ahn, J. S.; Ou, S.
Chem. 2002, 277, 46265.
30.
Pierce, A. C.; Sandretto, K. L.; Bemis, G. W. Proteins
2002, 49, 567.
Zakowski, M. F.; Pao, W.; Kris, M. G.; Miller, V. A.; Ladanyi, M.;
31.
Carey, K. D.; Garton, A. J.; Romero, M. S.; Kahler, J.;
Riely, G. J. Clin. Cancer Res. 2013, 19, 2240.
Thomson, S.; Ross, S.; Park, F.; Haley, J. D.; Gibson, N.;
16.
Sequist, L. V.; Besse, B.; Lynch, T. J.; Miller, V. A.;
Sliwkowski, M. X. Cancer Res. 2006, 66, 8163.
Wong, K. K.; Gitlitz, B.; Eaton, K.; Zacharchuk, C.; Freyman, A.;
Powell, C.; Ananthakrishnan, R.; Quinn, S.; Soria, J. C. J. Clin.
Oncol. 2010, 28, 3076.
32.
H.; Meyerson, M.; Eck, M. J. Cancer Cell 2007, 11, 217.
33. Yun, C. H.; Mengwasser, K. E.; Toms, A. V.; Woo, M. S.;
Yun, C. H.; Boggon, T. J.; Li, Y.; Woo, M. S.; Greulich,
17.
Walter, A. O.; Sjin, R. T.; Haringsma, H. J.; Ohashi, K.;
Greulich, H.; Wong, K. K.; Meyerson, M.; Eck, M. J. Proc. Natl.
Sun, J.; Lee, K.; Dubrovskiy, A.; Labenski, M.; Zhu, Z.; Wang, Z.;
Sheets, M.; St Martin, T.; Karp, R.; van Kalken, D.; Chaturvedi, P.;
Niu, D.; Nacht, M.; Petter, R. C.; Westlin, W.; Lin, K.; Jaw-Tsai, S.;
Raponi, M.; Van Dyke, T.; Etter, J.; Weaver, Z.; Pao, W.; Singh, J.;
Simmons, A. D.; Harding, T. C.; Allen, A. Cancer Discov. 2013, 3,
1404.
Acad. Sci.U.S.A. 2008, 105, 2070.
18.
Cross, D. A.; Ashton, S. E.; Ghiorghiu, S.; Eberlein, C.;
Nebhan, C. A.; Spitzler, P. J.; Orme, J. P.; Finlay, M. R.; Ward, R.
A.; Mellor, M. J.; Hughes, G.; Rahi, A.; Jacobs, V. N.; Red Brewer,
M.; Ichihara, E.; Sun, J.; Jin, H.; Ballard, P.; Al-Kadhimi, K.;
Rowlinson, R.; Klinowska, T.; Richmond, G. H.; Cantarini, M.; Kim,
D. W.; Ranson, M. R.; Pao, W. Cancer Discov. 2014, 4, 1046.
19.
Finlay, M. R. V.; Anderton, M.; Ashton, S.; Ballard, P.;
Bethel, P. A.; Box, M. R.; Bradbury, R. H.; Brown, S. J.;
Butterworth, S.; Campbell, A.; Chorley, C.; Colclough, N.; Cross, D.
A. E.; Currie, G. S.; Grist, M.; Hassall, L.; Hill, G. B.; James, D.;
James, M.; Kemmitt, P.; Klinowska, T.; Lamont, G.; Lamont, S. G.;
Martin, N.; McFarland, H. L.; Mellor, M. J.; Orme, J. P.; Perkins, D.;
Perkins, P.; Richmond, G.; Smith, P.; Ward, R. A.; Waring, M. J.;
Whittaker, D.; Wells, S.; Wrigley, G. L. J. Med. Chem. 2014, 57,
8249.

Graphical abstract