Manufacturing Development and Genotoxic Impurity Control Strategy of the Hedgehog Pathway Inhibitor Vismodegib

Article abstract of DOI:10.1021/acs.oprd.6b00208

The development work toward the robust and efficient manufacturing process to vismodegib, the active pharmaceutical ingredient (API) in Erivedge, is described. The optimization of the four-stage manufacturing process was designed to produce the API with the required critical quality attributes: (1) the selective catalytic hydrogenation reduction of the nitro compound 3 to the corresponding aniline 4 while minimizing the formation of potential genotoxic (mutagenic) impurities; (2) the control of the polymorphic phase and multipoint specification for particle size distribution.

Full text of DOI:10.1021/acs.oprd.6b00208

Subscriber access provided by CORNELL UNIVERSITY LIBRARY
Full Paper
Manufacturing Development and Genotoxic Impurity Control
Strategy of the Hedgehog Pathway Inhibitor Vismodegib
Remy Angelaud, Mark Reynolds, Cadapakam J. Venkatramani, Scott Savage, Huldreich Trafelet,
Thomas Landmesser, Peter Demel, Michael Levis, Olivier Ruha, Baerbel Rueckert, and Heinz Jaeggi
Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/acs.oprd.6b00208 • Publication Date (Web): 27 Jul 2016
Downloaded from http://pubs.acs.org on July 28, 2016
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.
Organic Process Research & Development 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 35
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Manufacturing Development and Genotoxic
Impurity Control Strategy of the Hedgehog Pathway
Inhibitor Vismodegib
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
Remy Angelaud,*,† Mark Reynolds, † Cadapakam Venkatramani, ‡ Scott Savage, † Huldreich
Trafelet,§ Thomas Landmesser, § Peter Demel, § Michael Levis, § Olivier Ruha,§ Baerbel
Rueckert§ and Heinz Jaeggi§
†Small Molecule Process Chemistry, ‡Small Molecule Analytical Chemistry, Genentech, A
Member of the Roche Group, 1 DNA Way, South San Francisco, CA 94080, USA
§Siegfried AG, Untere Brühlstrasse 4, CHꢀ4800 Zofingen, Switzerland
ACS Paragon Plus Environment
1

Organic Process Research & Development
Page 2 of 35
1
2
3
4
Table of Contents Graphic
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
ACS Paragon Plus Environment
2

Page 3 of 35
Organic Process Research & Development
1
2
3
4
ABSTRACT
5
6
7
8
9
The development work towards the robust and efficient manufacturing process to Vismodegib
the active pharmaceutical ingredient (API) in Erivedge® is described. The optimization of the 4ꢀ
stage manufacturing process was designed to produce the API with the required critical quality
attributes: 1) the selective catalytic hydrogenation reduction of the nitro compound 3 to the
corresponding aniline 4 while minimizing the formation of potential genotoxic (mutagenic)
impurities; 2) the control of the polymorphic phase and multipoint specification for particle size
distribution.
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
KEYWORDS. Vismodegib, nitro aromatic reduction, particle size distribution, genotoxic
impurity, hedgehog.
ACS Paragon Plus Environment
3

Organic Process Research & Development
Page 4 of 35
1
2
3
4
INTRODUCTION
5
6
7
8
9
The Hedgehog1 signalling pathway transmits information to embryonic cells and is required for
proper development. This signalling regulates embryonic development, ensuring that tissues
reach their correct size and location, maintain tissue polarity and cellular content. In the skin, the
Hedgehog pathway is critical for regulating hair follicle and sebaceous gland development.
Hedgehog signalling normally remains inactive in most adult tissues. Inappropriate reactivation
of the Hedgehog pathway has been linked to several human cancers, most often caused by
exposure to UV radiation. In basal cell carcinoma (BCC), abnormal Hedgehog pathway
signalling is the key molecular driver of the disease and more than 90% of BCCs have abnormal
activation of Hedgehog pathway signalling. Vismodegib, the active pharmaceutical ingredient
(API) in Erivedge® is the first inhibitor2 of the Hedgehog pathway to be approved for the
treatment of metastatic or locally advanced BCC and represents an important treatment option
for patients where surgery is not recommended.
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
Vismodegib 1 is synthesized in a 4ꢀstep manufacturing process from two designated starting
materials, 2ꢀ(2ꢀchloroꢀ5ꢀnitrophenyl)pyridine 3 and 2ꢀchloroꢀ4ꢀ(methylsulfonyl)benzoic acid 2
(Scheme 1). The development and optimization towards the manufacturing process faced two
major hurdles: (1) the control of the numerous potential residual genotoxic (mutagenic)
impurities (GTIs) in the API introduced by the nitro starting material 3; (2) the control of the
polymorph phase and the particle size distribution (PSD) of the API since it impacted the
dissolution of the corresponding capsule drug product. Vismodegib is assigned as a
Biopharmaceutics Classification System (BCS) Class 2 compound with low solubility and high
permeability.
ACS Paragon Plus Environment
4

Page 5 of 35
Organic Process Research & Development
1
2
3
4
Scheme 1. Vismodegib Synthesis
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
RESULTS AND DISCUSSION
Starting Materials Manufacturing
Starting material acid 2 is commercially available and was directly purchased in >99.5 A%
HPLC purity with a reproducible impurity profile. Nitro starting material 3 is custom made and
two manufacturing routes to this product were developed.3 In the first route,4 iodonitrobenzene 7
is produced from the corresponding nitroaniline 6 via a Sandmeyer reaction followed by
crystallization from water. Metal halogen exchange on 2ꢀbromopyridine with isopropyl Grignard
followed by transmetalation with ZnCl2 furnishes the corresponding organozinc reagent which is
used directly in the subsequent Negishi palladium catalyzed Negishi crossꢀcoupling with aryl
iodide 7 in THF to provide 3 after treatment with activated charcoal to scavenge residual
palladium, in >99.5A% HPLC (Scheme 2).
ACS Paragon Plus Environment
5

Organic Process Research & Development
Page 6 of 35
1
2
3
4
Scheme 2. Negishi Route to 3
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
In the second route5, 2ꢀ(2ꢀchloroꢀ5ꢀnitrophenyl)pyridine 3 is synthesized through the sequence
depicted in Scheme 3, in which 2ꢀchloroꢀ5ꢀnitroacetophenone intermediate 9 is obtained by
Knoevenagel condensation of the corresponding acyl chloride of 2ꢀchloroꢀ5ꢀnitrobenzoic acid 8
with dimethyl malonate followed by decarboxylation and subsequent purification by
crystallization in methanol/water. Condensation of 1,1,3,3ꢀtetramethoxypropane and N,N’ꢀ
dimethyl urea provides pyrimidinium chloride 10 after crystallization from 2ꢀpropanol. Biphenyl
3 is finally obtained by condensation of 9 and oxoꢀdihydropyrimidinium chloride in presence of
ammonium acetate in acetic acid in >99.5A% purity by HPLC after crystallization from
acetone/water.
ACS Paragon Plus Environment
6

Page 7 of 35
Organic Process Research & Development
1
2
3
4
Scheme 3. Pyrimidinium Route to 3
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
Both starting materials 2 and 3 are well characterized and their respective manufacturing
processes routinely produce them in high purity with no single impurities >0.2A% HPLC.
However, toxicology assessment of nitro aromatic precursors 6, 7, 8, 9 and nitro starting material
3 were performed using Quantitative StructureꢀActivity Relationship ((Q)SAR)6 methodologies
that predict the outcome of a bacterial mutagenicity assay. All five nitro compounds gave
structural alerts for predicted mutagenicity and were then submitted to a biological assay to
assess their mutagenic potential (Ames test)7. Although starting material 3 was Amesꢀnegative,
nitroaromatic intermediates 6, 7, 8 and 9 were found to be Amesꢀpositive and are therefore
considered genotoxic impurities (GTIs) in the process. In order to set appropriate specifications
for these four GTIs, a more thorough process understanding was needed and their respective
purging will be discussed later in this paper.
ACS Paragon Plus Environment
7

Organic Process Research & Development
Page 8 of 35
1
2
3
4
Manufacturing Process Optimization
5
6
7
8
9
The original Discovery route8 to Vismodegib involved the following 2ꢀstep process: reduction of
3 with iron filings in acetic acid to the corresponding aniline 5, followed by amide coupling with
acid 2 mediated by HBTU/HOAt. Because of the large waste generation and its poor scalability
this early route was not suitable for production and we decided to develop and optimize a more
efficient route as follows (Scheme 1): 1) Stage 1: Catalytic hydrogenation of the nitro starting
material 3; 2) Stage 2: Activation of acid 2 to the corresponding acyl chloride; 3) Stage 3:
Standard SchottenꢀBaumann type acylation; 4) Stage 4: Control of the physical properties
(polymorphic form, residual solvents and PSD).
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
Stage 1: Hydrogenation of 3 to amine Amine 4
Selective catalytic hydrogenation of nitroaromatic compounds to the corresponding aromatic
amines is a wellꢀknown process in the literature and numerous examples have been reported9.
The initial process development on the Stage 1 hydrogenation involved an exploratory
experimental design evaluating the effect of catalyst type (Pd(OH)2/C, Pd/C, and Pt/C), pressure
and temperature of the reaction (Table 1).
Table 1. Effect of catalyst, H2 pressure and temperature on reduction of 3
Entrya
Catalyst
Pressure Temperature
4 (A%)e 3 (A%)
(psi)
(°C)
ACS Paragon Plus Environment
8

Page 9 of 35
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
1
2
Pd(OH)2/Cb
Pd(OH)2/C
Pd(OH)2/C
Pd(OH)2/C
Pd/Cc
110
50
80
80
25
25
80
80
25
25
80
80
25
25
92.2
93.5
26.9
22
ND
ND
65
3
110
50
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
4
68.7
ND
ND
36.7
81.1
ND
ND
ND
ND
5
110
50
95.4
96.2
59.5
14.5
93.6
95.6
99.0
96.7
6
Pd/C
7
Pd/C
110
50
8
Pd/C
9
Pt/Cd
110
50
10
11
12
Pt/C
Pt/C
110
50
Pt/C
aReactions run on 2.3 g (9.8 mmol) of 3 in THF (25 mL). b20% Pd(OH)2/C (50% wet; 2.5 wt%).
c5% Pd/C (60% wet; 10 wt%). 5% Pt/C (60% wet; 10 wt%). Area % by HPLC10 after 20 h
d
e
reaction time.
Analysis of the results in Table 1 showed that in THF, 5% Pt/C was the optimal catalyst (entry 9ꢀ
12). Poor conversion to 4 was observed at 25 °C with Pd(OH)2 (entries 3 and 4) and with Pd/C
(entries 7 and 8) and the reactions had to be heated to go to completion. This had a negative
impact on the impurity profile of the reaction. In order to further understand the process, we
identified the main impurities generated during the screen in Table 1 and a general trend was
observed (Scheme 4). When the hydrogenation is performed at room temperature/low pressure,
an incomplete reduction generates hydroxylamine 11 which then generates11 12 which in turn, is
further reduced to aza and hydrazine byꢀproducts 13 and 14, respectively. On the other hand,
ACS Paragon Plus Environment
9

Organic Process Research & Development
Page 10 of 35
1
2
3
4
when the hydrogenation is performed at high temperature/high pressure, overꢀreduction leads to
5
6
desꢀchloro amine 15 and piperidine 16.
7
8
9
Scheme 4. Underꢀ and overꢀreduced impurities generated during hydrogenation of 3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
With the impurity markers in hand, we found out that hydroxylamine 11 coꢀeluted with desired
amine 4 with the HPLC method used at the time of this development and thus, conversions for
entries 9ꢀ12 (Table 1, Pt/C) were deceptively high. Furthermore, 11 was found to be Amesꢀ
positive12 and is considered a GTI. Therefore a new HPLC method for reaction monitoring was
implemented and we designed another screen. A medium temperature (60 °C) and pressure (88
psi) were chosen in order to minimize overꢀ and underꢀreduction. Methanol and ethanol were
chosen as solvents based on solubility studies and to facilitate the work up. Acid additives (HCl
and AcOH) were also tested and 1% Pt +2% V/C catalyst was introduced based on literature
precedent13. The results with acetic acid are reported in Table 2.
ACS Paragon Plus Environment
10

Page 11 of 35
Organic Process Research & Development
1
2
3
4
Table 2. Effect of catalyst, solvent and acetic acid on reduction of 3
5
6
7
8
Entrya
Catalyst
Solventb
4
11
12
13
14
15
16
(A%)c (A%) (A%) (A%) (A%) (A%) (A%)
9
1
2
3
4
5
6
1%Pt+2%V/C
MeOH
96.7
ND
ND
ND
80
0.51
0.34
0.31
0.49
0.30
0.37
0.32
0.24
0.34
0.33
0.19
0.26
1.51
1.26
1.08
0.27
3.03
1
0.38
0.16
0.19
1.81
0.41
8.64
ND
ND
ND
ND
0.19
ND
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%Pt+2%V/C MeOH/AcOH 97.4
1%Pt+2%V/C EtOH/AcOH 97.4
5%Pd/C
5%Pt/C
MeOH/AcOH 91.3
MeOH/AcOH 94.0
25
20%Pd(OH)2/C MeOH/AcOH 83.9
25
aReactions run on 1 g (4.3 mmol) of 3 in 10 mL of solvent at 60 °C and 88 psi for 6 hours.
bROH/AcOH 9:1. cArea % by HPLC after 3 h reaction time.
The results of these screening experiments show that the 1% platinum/2% vanadium on carbon
catalyst was the superior catalyst (entries 1ꢀ3) giving > 97% conversion to 4 and very low levels
of overꢀ and underꢀreduced impurities. MeOH/AcOH (9:1) was the best solvent system based on
the level of overꢀreduced products 15 and 16 (entry 2). A final screen was then performed in
order to define optimum ranges for temperature, concentration, molar equivalents of AcOH
while narrowing catalyst loading ranges and monitoring reaction completion and formation of
desꢀchloro impurity 15 and hydroxylamine 11 (Table 3).
ACS Paragon Plus Environment
11

Organic Process Research & Development
Page 12 of 35
1
2
3
4
Table 3. Optimum ranges screen for reduction of 3
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
MeOH AcOH
4
15
11
Temp
(°C)
Catalyst Amount Amount
(wt%)
(g/g)
(equiv)
Entry
1
(A%)
0.05
0.35
0.13
0.35
0.03
0.08
0.07
0.27
ND
25
25
25
25
25
25
25
25
60
60
60
60
60
60
60
60
40
40
40
40
40
40
40
55
55
3
7
3
7
3
7
3
7
3
7
3
7
3
7
3
7
5
4
6
5
5
5
5
5
3
4
4
8
8
4
4
8
8
4
4
8
8
4
4
8
8
6
6
6
5
7
6
6
6
4
1.50
1.50
1.50
1.50
6
15.59
93.17
49.69
95.48
8.59
0.03
0.12
0.08
ND
0.02
0.01
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
0.03
2
3
4
5
6
6
43.64
21.88
95.90
7.95
7
6
8
6
9
1.50
1.50
1.50
1.50
6
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
95.97
67.41
97.40
6.60
0.32
0.07
0.17
ND
6
96.50
88.63
97.05
97.05
73.11
97.20
98.82
96.94
96.77
87.11
95.01
15.59
0.22
0.08
0.21
0.17
0.12
0.16
0.16
0.15
0.18
ND
6
6
3.70
3.70
3.70
3.70
3.70
1.60
5.90
3.70
1.50
ND
0.35
aReactions run on 1 g (4.3 mmol) of 3 at 88 psi for 3 hours.
From these experiments, it is clear that when using the 1% platinum/2% vanadium on carbon
catalyst in MeOH/AcOH:
ACS Paragon Plus Environment
12

Page 13 of 35
Organic Process Research & Development
1
2
3
4
5
6
ꢀ Formation of 11 (GTI) is only seen in the low temperature experiments (entries 1ꢀ3) and
even then is not formed to any significant extent (entry 2)
7
8
9
ꢀ Amounts of catalyst < 5% gave low levels of conversion within the investigated reaction
time of 3 hours (entries 1, 5, 9, 13 and 25)
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
ꢀ Increased reaction temperatures gave higher levels of the desꢀchloro impurity 15 (entries
10, 14, 16 and 25)
ꢀ The impact of acetic acid concentration is small within the investigated range
ꢀ The level of 15 is highest when the conversion to product 4 is also highest (entries 4, 8,
16, 17, 20 and 21)
ꢀ Entry 24 shows the best compromise between conversion and impurity profile with 95%
conversion and not detectable amounts of both underꢀreduced 11 and overꢀreduced 15
Based on these results, the manufacturing parameters for the hydrogenation were defined as
followed: MeOH (9 wt)14, AcOH (1 wt), 1% Pt + 2% V/C (0.05 wt), temperature 50ꢀ55 °C
(target: 51 °C), pressure 60ꢀ88 psi (4ꢀ6 bar) for approximatively three hours15. The higher
percentage of methanol relative to the screened conditions ensured good dissolution of the
starting material. With MeOH/AcOH chosen as the solvent system for the hydrogenation
reaction, experiments were performed to optimize the isolation of 4 directly from the reaction
after removal of the catalyst by filtration. Isolation of the freebase was accomplished by
distillation of a portion of the MeOH/AcOH and addition of an aqueous base. Sodium carbonate
was used initially, but as this produced significant offꢀgassing, sodium hydroxide and aqueous
ammonia were investigated. While both allowed isolation of the freebase directly from the
reaction mixture, aqueous ammonia proved better at rejecting residual platinum and vanadium to
<2 ppm and so this was chosen for scaling up. Thus after filtration of the catalyst, the batch was
ACS Paragon Plus Environment
13

Organic Process Research & Development
Page 14 of 35
1
2
3
4
5
6
7
8
9
concentrated down under reduced pressure to 50% of its original volume. A 1:1 mixture of
water and aqueous ammonia (25 wt%) is added at 10ꢀ15 °C until the pH is >10. The resulting
slurry is aged for one hour, centrifuged and the solid washed with water until the pH of the
washing liquor is <7.5. Although wet 4 can be vacuum dried at up to 50 °C, the next stage of the
synthesis is run in a biphasic mixture of water/THF and so there is no need for drying. Thus, in
the current process intermediate 4 is isolated as a waterꢀwet solid, stored at ≤ –15 °C16 used as is
in Stage 3 with a loss on drying analysis performed to determine the appropriate amount to
charge in Stage 3. This step is routinely run on 18 kg scale of 3 and produces amine 4 in 95ꢀ98 %
yield (corrected for residue on drying) with >99.9 A% purity by HPLC. Table 4 summarizes the
operating ranges and criticality assessment for the Stage 1 hydrogenation.
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 4. Stage 1 operating ranges comparison and criticality assessment
Normal
Key for
Critical
Operating Yield or Process
Parameter
1% Pt+2% V/C (~50% wet)
Methanol
Target
0.05 wt
9.0 wt
1.0 wt
4 wt
Range
± 5%
Purity Parameter
Yesa
No
Yesa
No
± 5%
Acetic Acid
± 5%
No
No
Water
± 5%
No
No
25% Aqueous NH4OH
Reaction temperature
Reaction pressure
4 wt
± 5%
No
No
51 °C
5 bar
50–55 oC
Yes
No
Yes
No
~5 bar
aThe catalyst type is a critical material attribute
ACS Paragon Plus Environment
14

Page 15 of 35
Organic Process Research & Development
1
2
3
4
Stage 2: Activation of 2 to acid chloride 5
5
6
7
8
9
Standard conditions to form the acid chloride (SOCl2, catalytic DMF in CH2Cl2) were chosen to
initiate process development of this step and screening DOEs to better understand the reagent
stoichiometry and to identify regions of optimum reaction space were performed initially.
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. Effect of SOCl2 stoichiometry and reaction concentration on the conversion to 5
SOCl2
(molar equiv) (vol)
CH2Cl2
10
Entrya
(A%)b
1
2
3
4
5
6
7
1.9
1.5
1.5
1.7
1.7
1.2
1.2
6.0
11.5
6.0
99
99
99
99
99
99
99
11.5
6.0
11.5
6.0
aReactions run at a scale of 1 g of 2 for
b
6 hours at 40 °C. Conversion to 5 is
monitored by HPLC through derivatization
to the corresponding methyl ester 10.17
First we studied the impact of SOCl2 and DMF stoichiometry on the conversion to 5 in DCM at
40 °C and it was clear that the ratio of SOCl2 / DMF has no impact18 on the conversion (>99%)
and that the reaction performs well with as little as 0.05 weight equivalent of DMF and 1.9 molar
ACS Paragon Plus Environment
15

Organic Process Research & Development
Page 16 of 35
1
2
3
4
5
6
7
8
9
equivalent of SOCl2. Next, we studied the impact of the molar equivalents of SOCl2 and the
reaction concentration on the conversion to 5 while holding temperature (40 °C) and equivalents
of DMF (0.05 wt) constant (Table 5). The results demonstrated that the reaction could be
successfully run with 1.2 molar equivalent of SOCl2, 0.05 weight of DMF in 6 volumes of
CH2Cl2. While this firstꢀgeneration process performed well at 20 kg scale, we opted to use of
toluene instead of CH2Cl2 for safety (vide infra), environmental reasons and for isolation
purposes since 5 crystallizes out from this solvent.19
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In toluene, the acid chloride formation is run as a suspension at a calculated concentration of
18.7 wt%. Despite the low solubility of starting material 2 in the reaction medium it completely
dissolved as it was consumed during the reaction to the acid chloride intermediate 5. The low
end of the reaction temperature range in toluene is limited to 65 °C for safety reasons, ensuring
that the reagents are consumed faster than they are charged, avoiding a potential adiabatic
exothermic runaway reaction if a buildꢀup were to occur. Thus, the reaction is routinely run at 70
°C. The feed rate of the thionyl chloride has been set at >45 minutes for safety reasons, as stated
above. Additionally, a large volume of HCl and SO2 byꢀproduct gases have to be removed from
the production vessel (191 L/kg of 2). This gas flow has to be low enough to ensure that there is
no build up of pressure in the vessel, and this is ensured using the addition rate described above.
After the reaction is complete, the solution is partially concentrated at atmospheric pressure
(~111 °C). This serves mainly to remove excess thionyl chloride and HCl and SO2 reaction
gases. The distillation is controlled by the amount of distillate (~0.5 wt) per kilogram of starting
material. To test whether the acid chloride intermediate 5 remains stable in the reaction solution,
the reaction mixture was held at reflux at two different concentrations (prior and after distillation
for 24 hours). During these experiments, the only observed impurity was a small amount of the
ACS Paragon Plus Environment
16

Page 17 of 35
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
N,Nꢀdimethylamide impurity 17 (Table 5). This impurity has never been detected in normal
processing and can be purged during the isolation of acid chloride 5 so it is not of concern. The
mixture is then cooled20 to 10 °C and aged for 2 hours, to enable full crystallization of 5,
centrifuged, and the solid washed with toluene (1.2 vol). Due to its reactive nature, the isolation
of 5 requires significant care and it had been dried originally at 40 °C under reduced pressure.
However, because of its reactive and corrosive nature, metals were quickly corroded in the
drying equipment. Furthermore, in order to reduce the potential for hydrolysis during isolation,
drying and handling, it was decided not to incorporate a drying step and to store acid chloride 5
as a tolueneꢀwet cake cold21 (2–8 °C) in sealed polyethylene bags purged with an inert gas
(nitrogen or argon) and used asꢀis in the amide coupling Stage 3. On scale, this step is routinely
run on 23 kg of 2 and produce acid chloride 5 in 92ꢀ95 % yield (corrected for residue on drying)
with >99.8 A% purity by HPLC. Table 6 summarizes the operating ranges and criticality
assessment for the Stage 2 acid chloride formation.
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. Stage 2 operating ranges and criticality assessment
Normal
Key for
Critical
Process
Operating Yield or
Parameter
Target
0.6 wt
Range
± 5%
Purity
No
Parameter
Thionyl
No
Chloride
DMF
0.05 wt
4.22 wt
70 °C
± 5%
± 5%
± 5 oC
No
No
No
No
No
No
Toluene
Reaction
Temperature
ACS Paragon Plus Environment
17

Organic Process Research & Development
Page 18 of 35
1
2
3
4
The only impurity detected at this Stage is the starting acid 2 in <0.2 A% HPLC and it is purged
5
6
at the next stage.
7
8
9
Stage 3: Amidation to Crude Vismodegib 1
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
Since we wanted to carry out the Stage 3 with acid chloride 5 and amine 4 as a Schottenꢀ
Baumann acylation, THF was chosen as a solvent since it offered high solubility of both starting
materials (>15 wt% at 20 °C) and because we could isolate crude Vismodegib directly from THF
/ water mixture through crystallization. Our main area of development focused on the type and
stoichiometry of the inorganic base and thus we designed the experiments in Table 7 with the
following assumptions: use of acid chloride 5 in 1.1 molar equivalent excess to ensure that all
amine 4 is consumed and the slight excess of 5 will be hydrolyzed to the corresponding acid and
purged with the basic aqueous waste stream as a salt. The solvent ratio (THF/water) was
designed so amine 4 is fully soluble at the beginning of the reaction. Before addition of 5, the
THF/water ratio is 60:40 and solubility of 4 is ~7.5 wt% and well in range of the solubility22 of
~5.3 wt% in 50:50 THF/water at 20 °C.
Table 7. Impact of type and stoichiometry of the base on the SchottenꢀBaumann acylation
Entrya
Base
Molar
equivalent
1 (A%) 4 (A%) Yieldb (%)
ACS Paragon Plus Environment
18

Page 19 of 35
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
1
2
3
4
5
Na2CO3
Na2CO3
Na2CO3
K2CO3
K2CO3
0.60
0.66
0.72
0.66
0.72
94.4
94.5
98.4
99.8
99.8
5.4
5.4
1.4
ND
ND
83
96
87
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
101
99
aReactions run on ~72 g (0.35 mol) of 4, in 14 volumes of THF/water (60:40), cooling to 5 °C,
adding 1.1 molar equivalent of 5 in 11 volumes of THF, then warming to 20 °C for 1ꢀ2 hours.
bCrude yield not corrected for purity.
The results from Table 7 screen clearly shows that potassium carbonate is a more robust base for
the amidation, giving higher yields and purities (entries 4 and 5). Also, purging of excess acid 2
seems to be more efficient with potassium carbonate presumably because of the higher solubility
of the potassium salt of 2 vs the corresponding sodium salt in water (entries 4ꢀ5 vs 1ꢀ3). In order
to ensure good purging of the potassium salt of 2, a pH check to verify that the aqueous layer is
on the basic side (pH > 7) was implemented. Thus, 0.66 molar equivalent of K2CO3 was chosen
based on the resulting yield and purity (entry 4). For the isolation, the aqueous layer is drained
after reaction completion (<1 A% of 4 remaining) and the THF is removed by distillation under
atmospheric pressure with concomitant addition of water. During development, it was found that
the formation of an oily product layer was inevitable so seeding was implemented to prevent
spontaneous crystallization of the supersaturated solution. The solvent exchange is performed in
several phases: first 6 volumes of water (half of the total amount of water) are added directly
after the phase split is executed; the THF is distilled off while the second half of water is added
concomitantly to the distillation in such a way that the volume of the distillation is kept constant.
The batch is then seeded with Vismodegib 1 and the remaining THF is distilled off until a
boiling point of 82 °C is reached. At this temperature point, the ratio water/THF is 97.5:2.5
(verified by KFꢀtitration). The resulting slurry is then cooled to 15 °C (the solubility of 1 is <0.2
ACS Paragon Plus Environment
19

Organic Process Research & Development
Page 20 of 35
1
2
3
4
5
6
7
8
9
wt% between 10 and 30°C) and the solid isolated by centrifugation and washed with water.
Similarly to Stages 1 and 2, the solid is stored “wet”23 at ꢀ15 °C so no drying is necessary. On
scale, this step is routinely run on 15 kg of 4 and 23.5 kg of 5 to produce Vismodegib 1 in 97ꢀ
99% yield (corrected for residue on drying) with >99.9A% HPLC purity. Table 8 summarizes the
operating ranges and criticality assessment for the Stage 3 amide formation.
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 8. Stage 3 operating ranges comparison and criticality assessment
Normal
Key for
Critical
Process
Operating Yield or
Parameter
Acid 3
Target
1.1
Range
± 1%
Purity
Yes
Parameter
No
(molar equivalents)
THF (wt)
5.3
± 5%
± 5%
No
No
No
Potassium
0.66
Yes
Carbonate
(molar equivalents)
Water (wt)
4.3
± 5%
No
No
No
No
Reaction
2 °C
0–5 °C
Temperature
Final Temperature 20 °C 15–30°C
No
No
No
No
Vismodegib Seed 0.013
Crystals
± 1%
(molar equivalents)
This stage does not generate any impurities >0.05 A%. The acylated derivatives of amines byꢀ
product 11, 14, 15 and 16 from Stage 2 do not present any structural alert from in silico
assessment and are therefore not considered GTIs. The acid chloride 5 is considered a GTI and
its purging and fate will be discussed in the genotoxic assessment section.
ACS Paragon Plus Environment
20

Page 21 of 35
Organic Process Research & Development
1
2
3
4
Stage 4: Control of Vismodegib Physical Properties (polymorphic phase and PSD)
5
6
7
8
9
This Stage was designed to deliver the API with the required critical quality attributes: 1) the
desired polymorph form of the API (polymorph B) with acceptable levels of residual solvents
and 2) the required particle size distribution (PSD). Polymorph B is the thermodynamically most
stable polymorph observed and is the only form used in clinical trials. However, a less
thermodynamically stable form (Form A) has been observed during development but has never
been produced on scale. Figure 1 shows a differential scanning calorimetry (DSC) trace
containing a mixture of polymorph Form A (melting point, 177 °C) and Form B (melting point,
187 °C). The formation of Form B is controlled by seeding the crystallization step with the
appropriate polymorphic form and by drying to ensure low solvate and thus low residual solvent
levels. If a mixture of Form A and B is used for seeding, the crystallization also furnishes a
mixture of both forms.
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. DSC of form A and B mixture
Vismodegib has poor solubility24 in most organic solvents and eventually methyl isobutyl ketone
(MIBK) was chosen because it offered high solubility of 1 at reflux (11.8 wt%, 117 °C) and
ACS Paragon Plus Environment
21

Organic Process Research & Development
Page 22 of 35
1
2
3
4
5
6
7
8
9
down to <1 wt% at 5 °C. During process development, control of residual solvent levels was a
significant challenge due to MIBK solvate formation, because a significant amount of MIBK25
remained trapped in the isolated Vismodegib even after prolonged drying. As indicated by the
metastable zone width measurements (MSZW) in pure MIBK (Figure 2), a 9.5 wt%
concentration of 1 in MIBK was chosen to ensure a large thermodynamic solubility window
temperature range (~80 – 105 °C) and to eliminate the risk of spontaneous crystallization during
the polish filtration required prior final isolation of the API. The optimal seeding temperature
was set around 95 °C for this concentration.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Metastable Zone Width Measurements of Vismodegib
MSZW GDC-0449 in pure MIBK (0946M715)
MSZW of 1 in pure MIBK
13.00%
12.00%
11.00%
10.00%
9.00%
8.00%
7.00%
6.00%
60°C
70°C
80°C
90°C
100°C
110°C
120°C
Temperatur
Solubility
Spontaneous crystallization
MSZW of 1 in MIBK + 0.5% water
MSZW GDC-0449 in MIBK, 0.5%water (0950M705)
13.00%
12.00%
11.00%
10.00%
9.00%
8.00%
7.00%
6.00%
88°C
115°C
86°C
111°C
83°C
104°C
79°C
99°C
74°C
92°C
60°C
70°C
80°C
90°C
Temperature
Solubility curve Spontaneous crystallization
100°C
110°C
120°C
ACS Paragon Plus Environment
22

Page 23 of 35
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Several crystallization experiments were performed to better understand the impact of
temperature cycling and cooling rate on residual MIBK (Table 9).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 9. Impact of temperature cycling and cooling rate on residual MIBK in 1
Final
Residual
Temperature MIBK
Entrya
Notes
(°C)
20
20
20
5
(ppm)g
2144
1
2
3
4
5
6
Cycled onceb
Cycled twicec
9107
Seeded at 85 °Cd
Stepwise crystallizatione
5 hr hold at seeding
Linear coolingf
19,796
54,220
28,306
9,164
5
5
aExperiments run on ~15 g of 1 (9.5 wt% in MIBK), aged for ~10 hours at the final temperature, isolated by
filtration and dried overnight at 50 °C. bSeeded at 95 °C, aged for 2 hours, then cooled stepwise to 20 °C in 5 hours,
then back to 95 °C. cSeeded at 95 °C, aged for 2 hours, then cooled stepwise to 20 °C in 5 hours, then back to 95 °C
d
and repeated. Started at 115 °C, cooled to, seeded at 85 °C, aged for 2 hours, then cooled stepwise to 20 °C in 5
hours. eSeeded at 95 °C, aged for 2 hours, then cooled stepwise to 20 °C in 5 hours, fSeeded 95 °C, aged for 2 hours,
then cooled linearly to 5 °C in 1 hour. gResidual solvent levels determined by GC analysis.
Results from Table 9 showed that variation of cooling rates26 (entries 4 and 6), seeding
temperature (entry 3), hold time after seeding and temperature cycling (entries 1 and 2) produce
1 with unacceptable high level of residual MIBK. Drying at higher temperature (90 °C) did not
further reduce the amount of residual MIBK. Since we could not observe a significant difference
in XRPD between Vismodegib reference standard and samples containing >10,000 ppm of
MIBK, it was assumed that the MIBK solvate was a channel solvate. In order to test this
hypothesis, we ran new crystallization experiments with water as additive. We were pleased to
ACS Paragon Plus Environment
23

Organic Process Research & Development
Page 24 of 35
1
2
3
4
5
6
7
8
9
find that by using MIBK with 0.5 wt% of water in the crystallization we could isolate the API
with very low level of MIBK. As illustrated in Figure 1, the MSZW of Vismodegib in MIBK
plus 0.5 wt% of water shows solubility curves almost identical to pure MIBK making the 95 °C
seeding temperature still a robust set point. While the crystallization with water significantly
reduced residual MIBK, drying at 90 °C under vacuum was required to reach the low acceptable
levels. On production, this process always reduced MIBK to <2000 ppm.
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 the final manufacturing process, wet crude 1 is dissolved in MIBK and the solvent is
azeotropically distilled at atmospheric pressure (115 °C) to remove residual water from the Stage
3 intermediate. After cooling to 105 °C, fresh dry MIBK is added and the solution is polishꢀ
filtered and the volume adjusted by distillation to reach a concentration of ~ 9.5 wt%. Cooled to
95 °C, 0.5 % water relative to MIBK is added, then seeded with Form B and aged for 2 hours.
The slurry is cooled to 5 °C at 18 °C/hour and aged before isolation by centrifugation. The cake
is dried at 65 °C for 5 hours and then at 90 °C for at least 6 hours until residual solvents meet
specifications. On production scale, this step is routinely run on 20 kg of crude 1 to produce
Vismodegib 1 in 92ꢀ97% yield with >99.95 A% purity by HPLC and the desired polymorph
form B.
Finally, development was undertaken to develop milling parameters to provide API with a
particle size range as required for the Drug Product (formulated capsules) manufacturing. Per
ICH Q6A, the API particle size distribution is considered a critical quality attribute since it
impacts the dissolution of the corresponding formulated capsules. The correlation of the PSD of
the API and the dissolution profile of the capsules is illustrated in Figure 3, in which bigger
particles significantly slow down the time needed for the capsules to reach 85% dissolution.
ACS Paragon Plus Environment
24

Page 25 of 35
Organic Process Research & Development
1
2
3
4
Figure 3. Correlation between PSD of the API and capsule dissolution profiles
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
Thus, a multipoint specification for the API particle size distribution was established to ensure
Drug Product quality as follows: d(v,0.1) = 1 ꢁm to 15 ꢁm, d(v,0.5) = 20 ꢁm to 70 ꢁm and
d(v,0.9) = 100 ꢁm to 200 ꢁm. Vismodegib is milled using a hammer mill and the milling
parameters were systematically investigated with the goal to provide a robust process meeting
the above multipoint specification of the PSD. Experiments were performed to assess the impact
of the hammer speed on the PSD on different lots of unmilled Vismodegib and the results are
illustrated in Figure 4.
ACS Paragon Plus Environment
25

Organic Process Research & Development
Page 26 of 35
1
2
3
4
Figure 4. Correlation of Hammer Speed on Particle Size
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
Hammer Speed (rpm)
The data shows a good correlation between the hammer speed and the final mean PSD27, with a
faster hammer speed producing a smaller mean PSD. For production, a hammer speed in the
range of 3750 ꢀ 4250 rpm produces a very reproducible PSD meeting the multipoint PSD
specification and ensures Drug Product quality and manufacturability.
Genotoxic Impurity Control Strategy
ICH M728 limits the exposure of genotoxic impurities (GTIs) in patients to a maximum of 1.5
micrograms (ꢁg)/day (Threshold of Toxicological Concern or TTC) corresponding to a
theoretical 10ꢀ5 excess lifetime risk of cancer. Based on the TTC and on a Vismodegib dose of
150 mg/day, a maximum daily exposure can be calculated: 1.5 ꢁg/150 mg = 10 ppm. Therefore,
any GTI must have a 10 ppm maximum limit (acceptance criterion) in the API specifications. To
ACS Paragon Plus Environment
26

Page 27 of 35
Organic Process Research & Development
1
2
3
4
5
6
ensure that the 10 ppm limit in Vismodegib will be met, a better understanding of the fate and
purging of the GTI’s in the process for each identified GTI was needed.
7
8
9
First, we addressed the fate of the nitroaromatic GTIs 6 and 7 originated from the Negishi route
(Scheme 2) and nitroaromatics 8 and 9 originated from the pyridinium route to 3 (Scheme 3). We
designed spiking experiments with all 4 GTIs in the Stage 1 (hydrogenation) to assess the
purging capability of the process. Thus in different experiments, 1 wt% (10,000 ppm) of each 6,
7, 8 and 9 were spiked in the starting material 3 and submitted to the Stage 1 process. A GCꢀMS
method was developed and validated for the quantitation of 6, 7 and 9, while a LCꢀMS was
developed for 8. Both methods with selective ion monitoring29 have a limit of quantitation <4
ppm. In all isolated samples of 4, none of the GTIs were present >4 ppm, giving a purging factor
for each GTI of >2,500. Based on this data, impurities 6, 7, 8 and 9 were specified at <0.1% in
the starting material nitroaromatic 3. A similar experiment was run with hydroxylamine 11, but
this time 1 wt% of 11 was spiked into amine intermediate 4 and submitted to the Stage 3 process
(amide coupling). The resulting isolated Vismodegib was submitted for analysis by LCꢀMS and
11 was not detected >1 ppm, giving a purging factor of >10,000. This purging factor justifies the
specification for 11 of <0.50% in the amine 4 and guarantee that the 10 ppm acceptance criterion
in Vismodegib will be met. Finally, to better understand the fate of the last GTI, the kinetic
hydrolysis of acid chloride 5 at 60 °C was measured in a RCꢀ1 calorimeter by recording the
reaction heat flow. The calculated data according to a pseudo first order reaction kinetic predicts
hydrolysis times of 4 ꢀ 15 minutes for reaching 1 ppm of the initial acid chloride concentration.
Thus we can assume that less than 1 ppm of 5 will be left at the end of the Stage 3 workꢀup,
which requires at least 5 hours on scale, and therefore no specification is needed for residual 5.
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
27

Organic Process Research & Development
Page 28 of 35
1
2
3
4
Scheme 5. Vismodegib Manufacturing Process
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
In summary, we have developed a robust and efficient 4ꢀstage manufacturing process to
Vismodegib, the active pharmaceutical ingredient in Erivedge®. The process as illustrated in
Scheme 5 routinely produces Vismodegib in >80% overall yield and with >99.95% purity on 20
kg scale, with the desired polymorphic phase and the specified particle size distribution. A
control strategy in which genotoxic impurities are specified at the different stages of the
manufacturing process was implemented and ensures that the 10 ppm limit for GTIs in the API
will be met.
Experimental Section
General. Unless otherwise noted, all reactions were run under a nitrogen atmosphere, and
solvents and reagents were without further purification. 1H NMR (400 MHz) and 13C NMR (100
MHz) were recorded on a Bruker Avance 3 spectrometer. Chemical shifts are reported in ppm (δ
ACS Paragon Plus Environment
28

Page 29 of 35
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
units) downfield of internal tetramethylsilane [(CH3)4Si] or residual CHCl3; coupling constants
are reported in hertz (Hz). Multiplicities are as follows: s = singlet, d = doublet, t =triplet, q =
quartet, dd = doublet of doublets, m = multiplet. Melting points were measured by differential
scanning calorimetry (DSC, Büchi Bꢀ540). Compounds 2 (CAS#53250ꢀ83ꢀ2), 6 (CAS#6283ꢀ25ꢀ
6), 7 (CAS#74534ꢀ15ꢀ9), 8 (CAS#2516ꢀ96ꢀ3), 9 (CAS#23082ꢀ50ꢀ0), 10 (CAS#623927ꢀ88ꢀ8)
and 15 (CAS#15889ꢀ32ꢀ4) are commercially available.
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
Stage
1,
Synthesis
of
4-chloro-3-(pyridin-2-yl)aniline
4.
A
mixture
of
2ꢀ(2ꢀchloroꢀ5ꢀnitrophenyl)pyridine 3 (18.0 kg, 76.7 mol, 1 wt) and a 50 wt% wet catalyst with
approximately 1% platinum/2% vanadium on carbon (0.90 kg, 0.05 wt) in methanol (162 kg,
9 wt) and acetic acid (18 kg, 1 wt) is hydrogenated at a temperature of 50ꢀ55 °C under a pressure
of ~5.0 bar (~75 psi) for 3 hours (until hydrogen uptake is complete, as determined by a flow
of ≤ 50 L/hour). After cooling to 20 °C, the catalyst is removed by filtration. The remaining
solution is concentrated (~50%) under vacuum at a jacket temperature of a maximum of 60 °C
and then cooled to approximately 13 °C. A 1:1 mixture of water (72 kg) and 25 wt% aqueous
ammonia (73 kg) is added at 15 °C over 1.5 hours (until the pH is ≥ 10). The resulting
suspension is aged for at least 1 hr at 8ꢀ10 °C and centrifuged. The solid is washed with water
(397 kg) until the pH of the washing liquor is ≤ 7.5 and 15.9 kg of 4 is obtained (97.9% residue
on drying, 99.2% corrected yield, >99.9A% by HPLC). The wet 4 intermediate, is stored at ≤ –
15 °C in polyethylene bags placed in a fiberboard drum and used as is in Stage 3. 1H NMR (400
MHz, CDCl3) δH 8.72ꢀ8.68 (ddd, J = 4.9, 1.8, 1.0 Hz, 1H), 7.74 (td, J = 7.6, 1.8 Hz, 1H), 7.66
(dt, J = 7.9, 1.1 Hz, 1H), 7.27 (ddd, J = 7.5, 4.9, 1.3 Hz, 1H), 7.23 (d, J = 8.5 Hz, 1H), 6.93 (d, J
ACS Paragon Plus Environment
29