Neutral 5-substituted 4-anilinoquinazolines as potent, orally active inhibitors of erbB2 receptor tyrosine kinase

Article abstract of DOI:10.1016/j.bmcl.2007.08.073

Neutral 5-substituted 4-anilinoquinazolines addressed high in vivo clearance and phospholipidosis associated with previous basic compounds. A representative compound 8a inhibited tumor growth in a mouse xenograft model when co-administered with the cytoch

Full text of DOI:10.1016/j.bmcl.2007.08.073

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 17 (2007) 6326–6329
Neutral 5-substituted 4-anilinoquinazolines as potent, orally
active inhibitors of erbB2 receptor tyrosine kinase
Peter Ballard,a Bernard C. Barlaam,b Robert H. Bradbury,a,* Allan Dishington,a
Laurent F. A. Hennequin,b D. Mark Hickinson,a Ian M. Hollingsworth,a
Jason G. Kettle,a Teresa Klinowska,a Donald J. Ogilvie,a Stuart E. Pearson,a
James S. Scott,a Abid Suleman,a Robin Whittaker,a Emma J. Williams,a
Robin Wooda and Lindsay Wrighta
aCancer and Infection Research, AstraZeneca, Mereside, Alderley Park, Macclesfield SK10 4TG, UK
bAstraZeneca, Centre de Recherches, Z.I. la Pompelle, BP1050, 51689 Reims Cedex 2, France
Received 26 July 2007; revised 29 August 2007; accepted 30 August 2007
Available online 2 September 2007
Abstract—Neutral 5-substituted 4-anilinoquinazolines addressed high in vivo clearance and phospholipidosis associated with pre-
vious basic compounds. A representative compound 8a inhibited tumor growth in a mouse xenograft model when co-administered
with the cytochrome P450 inhibitor 1-aminobenzotriazole (ABT), and data are consistent with pharmacology primarily reflecting
inhibition of erbB2 receptor tyrosine kinase.
Ó 2007 Elsevier Ltd. All rights reserved.
With introduction into clinical practice of the EGFR
SO2Me
(erbB1) inhibitors erlotinib and gefitinib, and the mono-
clonal antibody to erbB2 trastuzumab, blockade of the
EGFR family signaling pathway has emerged as a lead-
ing approach to selective targeting of proliferating tu-
mor cells.1 The dual EGFR-erbB2 receptor tyrosine
kinase inhibitor lapatinib 12 (Fig. 1) has recently been
approved in combination with capecitabine by the
FDA for treatment of advanced metastatic breast cancer
and is also undergoing clinical trials for treatment of a
number of other tumor types. Other dual EGFR-erbB2
inhibitors in earlier stage trials include1 AEE788, BMS-
599626, and the irreversible inhibitors canertinib, HKI-
272 and BIBW 2992. Although an erbB2-selective inhib-
itor may offer more flexibility in combination therapy,1
CP-724714 2 is the only selective small-molecule inhibi-
tor reported to have entered the clinic.
O
F
HN
Cl
HN
N
O
N
1
O
O
OMe
N
O
HN
HN
2
N
O
N
N
N
HN
Cl
5
4
N
3
N
All of these small-molecule inhibitors are based on bind-
ing of a 4-anilinoquinazoline or related template at the
Figure 1. Lead compound 3 and examples of erbB2 receptor tyrosine
kinase inhibitors undergoing clinical trials: lapatinib 1; CP-724714 2.
Keywords: Anilinoquinazoline; erbB2-selective inhibitor; Tumor gro-
wth; Cytochrome p450 inhibitor.
kinase ATP site.1 We have previously described work
leading to the 5-substituted 4-anilinoquinazoline 33
(Fig. 1) a potent, selective, and orally active inhibitor
*
Corresponding author. Tel.: +44 1625 515708; fax: +44 1625
510097; e-mail: Rob.Bradbury@astrazeneca.com
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.08.073

P. Ballard et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6326–6329
6327
of erbB2. Work since published on 5-substituted quina-
zolines as inhibitors of EGFR4 and c-Src5 strongly sup-
ports the hypothesis that the 5-substituent occupies the
ribose binding site of the kinase.
O
N
Cl
F
HN
N
(a), (b)
N
We were unable, however, to progress compound 3 be-
yond pre-clinical models for reasons attributable to the
basic nature of the 5-substituent. Oral administration
of 3 in rat toxicology studies6 resulted in mild to severe
effects in liver, lung, and spleen characteristic of phos-
pholipidosis, a toxicity known to be associated with cat-
ionic amphiphilic drugs7 (3: pKa 8.1, ClogP 5.1,
logD > 3.5). In addition, a range of basic analogs gener-
ally showed moderate to high clearance in rat and dog.6
We envisaged that hydroxyacylation of the basic nitro-
gen could circumvent these issues and also introduce a
pendant hydroxyl group that could interact with resi-
dues located in the ribose pocket. In this publication,
we describe how this work led to erbB2 inhibitors de-
void of phospholipidosis and with reduced clearance.
HO
O
N
4
O
N
Cl
O
HN
N
N
5a (
5b (
R
S
) enantiomer
) enantiomer
Scheme 1. Synthesis of compounds 5a–b. Reagents and conditions:
(a) (R) or (S) prolinol, NaH (2.5 equiv), 15-crown-5, DMA, 95 °C;
(b) HOCH2CO2H, HATU, i-Pr2NEt, DMF, 20 °C.
OH
Cl
OH
H
N
F
HN
N
O
HN
N
Cl
(a)
R
N
7a R = H
Illustrative compounds prepared during the course of
this work are listed in Table 1, and synthetic routes
are outlined in Schemes 1–3.8 Measured logD values
for these compounds lie in the range 3–3.5 and led to
comparable physicochemical properties to compound 3
(data not shown).
N
7b R = Me (R )
6
7c R = Me (S )
O
O
HO
N
N
HN
N
Cl
8a R = H
O
(b), (c)
In extension of previous work,3 initial variants 5a–b
were prepared (Scheme 1) from fluoroquinazoline 43
by a sequence involving displacement with the alkoxide
derived from (R) or (S) prolinol followed by acylation of
the pyrrolidine nitrogen with glycolic acid.
R
N
8b R = Me (
8c R = Me (
R )
S
)
Scheme 2. Synthesis of compounds 8a–c. Reagents and conditions:
(a) MeNHCH(R)CH2OH,9,10 NaH (2.5 equiv), 15-crown-5, DMA,
110 °C; (b) HOCH2CO2H, HATU, i-Pr2NEt, DMF, 20 °C; (c) 2-
picolyl chlorideÆHCl, K2CO3, 18-crown-6, DMF, 20 °C.
Since displacement reactions of fluoroquinazoline 4 with
acyclic amino alkoxides gave low and variable yields, an
alternative route was used (Scheme 2) to synthesize ana-
logs 8a–c. In contrast with 4, base-promoted displace-
ment of the corresponding phenol 69 with the
appropriate amino alcohol10,11 gave intermediates 7a–c
in high yield, and N-acylation followed by O-alkylation
then provided 8a–c.
OH
R
R
N
O
HN
N
Cl
+
6
(a)
R
N
OH
N
9a R = Me (
R
S
)
)
10a R = Me (
R
S
)
)
9b R = Me (
10b R = Me (
As part of a wider body of work,8 this more versatile
route enabled late-stage variation of both the acyl and
extended aniline3 moieties. We speculate that under
strongly basic conditions the anilinoquinazoline NH of
substrate 4 is deprotonated, thereby deactivating the
adjacent 5-fluorine atom to nucleophilic displacement,
O
O
HO
N
N
HN
N
Cl
O
(b) - (d)
N
11a R = Me (
R
S
)
)
11b R = Me (
Table 1. Inhibition data3 versus erbB2 and EGFR kinase and erbB2
cellular autophosphorylation for 5a–b, 8a–c, 11a–b
Scheme 3. Synthesis of compounds 11a–b. Reagents and conditions:
(a) NaH (2.5 equiv), 15-crown-5, DMA, 110 °C; (b) (Ph3P)3RhCl,
MeCN/H2O, microwave, 110 °C; (c) HOCH2CO2H, HATU, i-Pr2NEt,
DMF, 20 °C; (d) 2-picolyl chlorideÆHCl, K2CO3, 18-crown-6, DMF,
20 °C.
a
a
a
Compound
Erb B2 IC50
(nM)
EGFR IC50
(nM)
Clone 24 IC50
(nM)
5a
<2
2
97 ( 64)
327 ( 226)
35 ( 6)
147 ( 79)
259 ( 53)
43 ( 20)
11 ( 5)
5b
8ab
8b
4 ( 2)
<2
whereas for substrate 6 it is the phenol which is depro-
tonated, leaving the free anilinoquinazoline NH to assist
the neighboring 5-fluoro leaving group.
3 ( 1)
8c
19 ( 10)
<2
22 ( 10)
2580 ( 1350)
2 ( 1)
191 ( 137)
131 ( 48)
11 ( 0.5)
291 ( 144)
11a
11b
a n P 2, standard error is given in parentheses.
As the chiral amino alcohol precursors of analogs 11a–b
were not readily accessible, a modified route was em-
b IC50 > 1000 nM (Kdr, Src).

6328
P. Ballard et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6326–6329
Table 4. Plasma concentrations for 8a after dosing in female mice at
50 mg/kg p.o., and ratios of free plasma levels to IC50 value in the
Clone 24 and BT474C cellular assays
ployed (Scheme 3) whereby precursor 6 underwent
base-promoted displacement with N-protected amino
alcohols 9a–b, readily obtained by ytterbium triflate
catalyzed ring opening12 of the corresponding chiral
oxirane with N-methylallylamine. De-allylation of inter-
mediates 10a–b and functionalization as detailed earlier
then gave 11a–b.
Time (h)
2
4
6
16
24
Plasma concentration 8aa (lM) 0.35 0.09 0.02 NDb NDb
Plasma concentration 8ac (lM) 16.5 13.8 6.3 0.026 0.007
c,d
Ratio free 8a/Clone 24 IC50
c,d
Ratio free 8a/BT474C IC50
7.6
2.5
6.4 2.9 0.012 0.003
2.1 0.97 0.004 0.001
The compounds listed in Table 1 were evaluated in
erbB2 and EGFR kinase assays and in a cellular erbB2
autophosphorylation assay (‘Clone 24’) as cited previ-
ously.3 Evolution of this series of compounds led to neu-
tral acyclic variants 8a–b and 11a with enhanced cellular
potency relative to compound 3.3 Significant potency
differences were also seen between pairs of enantiomers
(8b vs 8c, 11a vs 11b). Compound 8a showed a clean
selectivity profile versus in-house and external kinase
panels (data not shown).13 Thus, for example, 8a gave
selectivity ratios of >333 versus the tyrosine kinases
Kdr and Src (Table 1, footnote b).
a Without co-administered ABT, AUC 0.89 lM h.
b Not detected.
c With ABT co-administered at 100 mg/kg, AUC 99 lM h.
d Measured free drug in mouse plasma is 2.0% total concentration.
Control
1.6
ABT 100 mg/kg
1.4
8a 50 mg/kg once daily with ABT
8a 50 mg/kg twice daily with ABT
1.2
1
36%
0.8
0.6
0.4
0.2
Antiproliferative effects were determined in assays using
BT474C and KB cell lines (Table 2), known to respond,
respectively, to stimulation by both erbB2 and EGFR,
and EGFR alone.3 Good separation was seen between
antiproliferative activity and effect on basal growth in
KB cells, but in contrast with lead compound 33 a less
marked difference was observed between the magnitude
of the antiproliferative effect in the two cell lines. As dis-
cussed below, however, we believe that antitumor effects
in vivo continue to primarily reflect inhibition of erbB2.
112%
25
0
0
5
10
15
20
Days of dosing
Figure 2. Inhibition of growth of BT474C xenograft15 in athymic mice
dosed orally with compound 8a and ABT.
and no evidence of phospholipidosis was seen in 14-
day rat toxicology studies at up to 600 mg/kg q.d.
Pharmacokinetic parameters were determined in mouse,
rat, and dog, the species used for in vivo models and pre-
clinical toxicology studies. Based on these parameters,
compound 8a (Table 3) was selected for further evalua-
tion. In contrast to compound 3, the neutral 5-substitu-
ent in 8a led to low to moderate clearance in rat and
dog. Oral exposure in rat scaled well to higher doses,
Compound 8a and analogs were, however, more highly
cleared and poorly bioavailable in mouse, and conse-
quently at an oral dose of 50 mg/kg minimal exposure
was seen (Table 4). To increase exposure, the compound
was co-administered with pre-dosed 1-aminobenzotriaz-
ole (ABT), an irreversible inhibitor of cytochrome P450
metabolizing enzymes known to reduce P450-mediated
clearance in vivo.14 This regimen gave approximately
100-fold increase in exposure, and calculation of mouse
plasma concentrations showed the free levels of 8a to
compare favorably with IC50 values in the Clone 24
and BT474C assays up to the 6-h time point (Table 4).
Table 2. Cellular antiproliferative data3 for 8a and 11a
a
a
Compound
BT474C IC50
(nM)
KB IC50
(nM)
KB basala
(nM)
8a
11a
130 ( 20)
34 ( 17)
621 ( 350)
141 ( 44)
8145 ( 3770)
6936 ( 3500)
Compound 8a was evaluated at 50 mg/kg p.o. q.d. and
b.i.d. for inhibition of growth of a BT474C tumor xeno-
graft in a 28-day study in athymic mice.15 No significant
reduction in body weight was observed for any animal
group in the study. As shown in Figure 2, compound
8a gave statistically significant inhibition with both dos-
ing schedules, and twice daily dosing gave better effi-
cacy, in accord with more prolonged cover relative to
cellular IC50.16 Analysis of excised tumor samples con-
firmed significant inhibition of phospho-erbB2 (e.g.
70% inhibition 1 h after the last dose at 50 mg/kg b.i.d.).
a n P 2, standard error is given in parentheses.
Table 3. Mouse, rat and dog pharmacokinetic parameters for 8a
Species
Vdss
(l/kg)
Cl (ml/min/kg) AUCd
Bioavailability %
(lM h)
Mousea 0.7
Ratb
Dogc
72
10
10
0.02
5.5
32
4
0.9
1.0
32
51
a Immunocompetent female Alderley Park mice dosed at 2 mg/kg iv
and 5 mg/kg p.o.
Although the BT474C mouse xenograft model responds
to both erbB215 and EGFR inhibition,6 the modest
activity of compound 8a in the KB cell assay (Table 2)
is consistent with xenograft activity primarily reflecting
inhibition of erbB2. In accord with this proposal, 8a
b Female Alderley Park rats dosed at 2 mg/kg iv and 5 mg/kg p.o.
c Mean values for male Alderley Park beagle administered doses of
50 mg iv and 250 mg p.o.
d AUC0–24 from p.o. dosing.

P. Ballard et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6326–6329
6329
Klinowska, T.; Leach, A. G.; Morgentin, R.; Pass, M.;
Ogilvie, D. J.; Olivier, A.; Warin, N.; Williams, E. J.
Bioorg. Med. Chem. Lett. 2006, 16, 1633.
showed no growth or phospho-EGFR inhibition in
mouse and rat LoVo xenograft models17,18 at oral doses
of 50 mg/kg b.i.d. and 300 mg/kg q.d., respectively (data
not shown).
5. Hennequin, L. F.; Allen, J.; Breed, J.; Curwen, J.;
Fennell, M.; Green, T. P.; Lambert van der Brempt, C.;
Morgentin, R.; Norman, R. A.; Olivier, A.; Otterbein,
L.; Ple, P. A.; Warin, N.; Costello, G. J. Med. Chem.
2006, 49, 6465.
6. Unpublished data.
7. Kacew, S.; Reasor, M. J.; Ruben, Z. Drug Metab. Rev.
1997, 29, 355.
8. (a) For experimental procedures see: Bradbury, R. H.;
Hennequin, L. F. A.; Kettle, J. G.; Barlaam, B. C. PCT
Int. Appl. WO 2005026151; (b) Bradbury, R. H.; Henne-
quin, L. F. A.; Kettle, J. G. PCT Int. Appl. WO
2005051923.
In summary, replacement of the basic 5-substituent in this
anilinoquinazoline subseries with a neutral hydroxya-
cylamino moiety circumvented high in vivo clearance
and phospholipidosis attributable to the basicity of the
5-substituent, and a representative compound 8a inhib-
ited tumor growth in a mouse BT474C xenograft model
when co-administered with ABT. Further evolution of
5-substituted 4-anilinoquinazolines as inhibitors of erbB2
receptor tyrosine kinase will be reported in due course.
9. Readily obtained by reaction of 4-chloro-5-fluoroquinazo-
line3 and 2-chloro-4-aminophenol in refluxing isopropanol.
10. Beckett, A. H.; Harper, N. J. J. Chem. Soc. 1957, 858.
11. Chiacchio, U.; Casuscelli, F.; Corsaro, A.; Librando, V.;
Rescifina, A.; Romeo, R.; Romeo, G. Tetrahedron 1995,
51, 5689.
12. Chini, M.; Crotti, P.; Favero, L.; Macchia, F.; Pineschi,
M. Tetrahedron Lett. 1994, 35, 433.
13. The activity of compound 9a was assessed at a concen-
tration of 10 lM against a panel of 39 recombinant
protein kinases at Km ATP concentration for each
enzyme.
Acknowledgments
We thank the following for technical assistance: Sarah
Beck, Rowena J. Callis, Sara Davenport, Frederic Duc-
rozet, Paul M. Farrington, Janet A. Jackson, Helen S.
Ross, Robin Smith, Cath B. Trigwell, John P. Vincent,
Lisa C. Woods.
References and notes
14. Balani, S. K.; Li, P.; Nguyen, J.; Cardoza, K.; Zeng, H.;
Mu, D.-X.; Wu, J.-T.; Gan, L.-S.; Lee, F. W. Drug Metab.
Disp. 2004, 32, 1092.
15. Baselga, J.; Norton, L.; Albanell, J.; Kim, Y. M.;
Mendelsohn, J. Cancer Res. 1998, 58, 2825.
16. Simeoni, M.; Magni, P.; Cammia, C.; De Nicolao, G.;
Croci, V.; Pesenti, E.; Germani, M.; Poggesi, I.; Rocchetti,
M. Cancer Res. 2004, 64, 1094.
17. Wakeling, A. E.; Guy, S. P.; Woodburn, J. R.; Ashton, S.
E.; Curry, B. J.; Barker, A. J.; Gibson, K. H. Cancer Res.
2002, 62, 5749.
18. Stragand, J. J.; Drewinko, B.; Henderson, S. D.; Grossie,
B.; Stephens, L. C.; Barlogie, B.; Trujillo, J. M. Cancer
Res. 1982, 42, 3111.
1. For a comprehensive review of the medicinal chemistry,
structural biology, and clinical evaluation of EGFR family
kinase inhibitors, see Wallace, E. M.; Yeh, T. C.; Laird, E.
R.; Blake, J. F.; Lyssikatos, J. In Topics in Medicinal
Chemistry (Cancer); Bradbury, R. H., Ed.; Springer:
Heidelberg, 2007; p 83.
2. Lackey, K. Curr. Top. Med. Chem. 2006, 6, 435.
3. Ballard, P.; Bradbury, R. H.; Hennequin, L. F. A.;
Hickinson, D. M.; Johnson, P. D.; Kettle, J. G.; Kli-
nowska, T.; Morgentin, R.; Ogilvie, D. J.; Olivier, A.
Bioorg. Med. Chem. Lett. 2005, 15, 4226.
4. Ballard, P.; Bradbury, R. H.; Harris, C. S.; Hennequin, L.
F. A.; Hickinson, M.; Johnson, P. D.; Kettle, J. G.;