The First Asymmetric Pilot-Scale Synthesis of TV-45070

Article abstract of DOI:10.1021/acs.oprd.7b00237

TV-45070 is a small-molecule lactam containing a chiral spiro-ether that has been reported as a potential topical therapy for pain associated with the Na_v1.7 sodium ion channel encoded by the gene SCN9A. A pilot-scale synthesis is presented that is highlighted by an asymmetric aldol coupling at ambient temperature, used to create a quaternary chiral center. Although only a moderate ee is obtained, the removal of the undesired isomer is achieved through preferential precipitation of a near racemic mixture from the reaction, leaving the enantiopure isomer in solution. Cyclization to form the final API uses an uncommon diphenylphosphine-based leaving group which proved successful on the neopentyl system when other traditional leaving groups failed.

Full text of DOI:10.1021/acs.oprd.7b00237

Article
pubs.acs.org/OPRD
The First Asymmetric Pilot-Scale Synthesis of TV-45070
Joseph A. Sclafani,*,†,§ Jian Chen,†,∥ Daniel V. Levy,†,§ Harlan Reese,†,⊥ Mina Dimitri,†
Partha Mudipalli,† Michael Christie,† Christopher J. Neville,‡ Mark Olsen,‡ and Roger P. Bakale†,#
†Chemical Process Research and Development, ‡Analytical Research and Development, Teva Branded Pharmaceutical Products R&D
Inc., 383 Phoenixville Pike, Malvern, Pennsylvania 19355, United States
S
* Supporting Information
ABSTRACT: TV-45070 is a small-molecule lactam containing a chiral spiro-ether that has been reported as a potential topical
therapy for pain associated with the Nav1.7 sodium ion channel encoded by the gene SCN9A. A pilot-scale synthesis is presented
that is highlighted by an asymmetric aldol coupling at ambient temperature, used to create a quaternary chiral center. Although
only a moderate ee is obtained, the removal of the undesired isomer is achieved through preferential precipitation of a near
racemic mixture from the reaction, leaving the enantiopure isomer in solution. Cyclization to form the final API uses an
uncommon diphenylphosphine-based leaving group which proved successful on the neopentyl system when other traditional
leaving groups failed.
racemic drug substance prior to SMB chromatography.
Coupling with 6 was found to be necessary to enhance
solubility for the SMB resolution. Although the early synthesis
proved vital to provide quantities of API for clinical trials,
significant losses late in the synthesis highlighted the need for
an asymmetric route.
INTRODUCTION
■
Postherpetic neuralgia (PHN) is a rare disorder that is defined
as significant pain or abnormal sensation 120 days or more after
the presence of the initial rash caused by shingles. This pain
persists after the healing of the associated rash. Generally, this
affliction occurs in older individuals and individuals suffering
from immunosuppression. There are about one million cases of
shingles in the US per year, of which 10−20% will result in
PHN.1 Topical analgesics such as lidocaine2 and capsaicin3 are
traditionally used to treat this disorder. Both lidocaine and TV-
45070 have a mechanism of action that involves the inhibition
of voltage-gated sodium ion channels.
TV-45070 (formerly XEN-402) was in-licensed by Teva
from Xenon Pharmaceuticals and is reported to be an
antagonist of the Nav1.7 sodium ion channel protein. It is
currently in Phase II clinical trials for PHN. Interestingly, the
loss of function of the Nav1.7 sodium ion channel was reported
to result in the inability to experience pain as a hereditary trait
in certain individuals.4 Primary erythromelalgia5 is another rare
disease where alterations in Nav1.7 or mutations in the
corresponding encoding gene SCN9A have been reported to
result in chronic burning pain that can last for hours or even
days. Thus, compounds which regulate this protein have
potential therapeutic value as analgesics for chronic pain.
The first-generation process route6 (Scheme 1) used for
Phase I clinical supplies began with the coupling of
commercially available isatin and sesamol mediated by
isopropylmagnesium chloride to give 2 in quantitative yield.
The removal of the hydroxyl group in 2 was achieved in good
yield using triethylsilane in neat trifluoroacetic acid. The
addition of formaldehyde promoted by NaOH was not found
to be chemoselective, giving racemic 4 that included hemi-
aminal formation from the addition of a second equivalent of
formaldehyde. In situ removal of the hydroxymethyl group on
the oxindole nitrogen, preceded by Mitsunobu cyclization,
provided 5 in 49% yield and 89.4 area percent (A%) purity.
Coupling of 5 with 6 was mediated by cesium carbonate and
required large-scale chromatography to purify the crude
A first-generation asymmetric route7 performed on kilo lab
scale focused on generating the desired S-isomer (1) directly,
thus avoiding the need for SMB chromatography (Scheme 2).
This synthesis again utilized the common raw materials isatin
and sesamol to build the core of the final drug substance.
However, this route featured an asymmetric alkylation
promoted by the Lygo phase-transfer catalyst,8 which provided
the desired center in 90% ee, when performed at −20 °C. No
attempts with the allylated Corey variant8c of the catalyst were
reported. To accommodate the asymmetric step, several
protecting groups were added to the synthesis, resulting in an
increase in the number of steps. This route also introduced the
bromomethylfuran late in the synthesis, allowing for the
possibility of genotoxic impurity 6 in the final API.
Nonetheless, this route demonstrated that enrichment in ee
was possible over several steps, ultimately leading to API of an
enantiopurity which was comparable to the earlier SMB
process. Thus, from this work it seemed possible that even
moderate ee in the asymmetric step could be upgraded through
crystallization. Additional improvements included the use of a
phosphine which could be removed by treatment with aqueous
acid thus facilitating chromatographic removal of impurities in
the Mitsunobu step. Encouraged by this work, a shorter
asymmetric route was planned that could similarly benefit from
enantio-enrichment through successive crystallizations.
Received: July 10, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/acs.oprd.7b00237
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Scheme 1. Early-Phase Synthesis of TV-45070
Scheme 2. First-Generation Asymmetric Route
protic sites in both starting materials, this amount of metalating
reagent seemed to be insufficient and suggested that it was
merely acting as a base. A quick base screen demonstrated that
the chemistry could be performed in tetrahydrofuran (THF) or
RESULTS AND DISCUSSION
■
While generating initial quantities of 2, it was noted that only a
slight excess of one equivalent of Grignard reagent was
necessary to complete the reaction. Given the number of
B
DOI: 10.1021/acs.oprd.7b00237
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
dimethylformamide (DMF) using a slight excess of potassium
carbonate (Scheme 3). Although it was necessary to use
to be too labile, reverting to 3 upon prolonged exposure to air.
As a result, it was not investigated further. The use of benzyl
bromide with 3 resulted in exclusively C-benzylation adjacent
to the carbonyl rather than substitution on the phenol.
Scheme 3. Potassium Carbonate-Promoted Coupling of
Sesamol and Isatin
Unfortunately, in the case of Sharpless ligands, poorer ee was
observed and longer reaction times, despite a cleaner impurity
profile. Although no ee was observed when catalyst 2213 was
employed with 3, the substitution of the phenol gave
encouraging improvement (Table 2). In general, the TES
protecting group gave inferior results when compared to the
TBDMS protected substrate. Carbinolamine formation with
formaldehyde continued to be observed in all reactions.
Hoping to avoid hemiaminal formation, the functionalization
of the oxindole with 6 was performed earlier in the synthesis.
This also avoided the potential trace contamination by
genotoxic raw material 6 in the final step. The incorporation
of 6 also negated the need for a nitrogen protecting group,
which would ultimately require a separate step for removal. As
thiourea catalysts 21 and 22 (Figure 2) are proposed to bind to
formaldehyde through hydrogen bonding14 in addition to
functioning as chiral bases, hemiaminal formation could
potentially impact the chiral event. When employing 18 as a
substrate, a significant improvement in ee was observed at 25
°C (Table 3). When performed in acetonitrile, a modest
improvement with both 21 and 22 was observed with trace
amounts of impurities. However, when slurried in heptane, a
significant increase in ee was observed. By lowering the
concentration in heptane, the ee fell to levels comparable to
those observed with solvents where the product was more
soluble. This suggested that, by limiting the amount of product
in solution, the competitive hydrogen bonding of the installed
hydroxymethyl group with formaldehyde was reduced. This in
turn could enhance the coordination of formaldehyde with the
catalyst giving greater stereocontrol.
Having obtained moderate ee under our best conditions, we
sought to enrich the amount of preferred S-isomer, either by
preferential crystallization15 of racemic material or through
direct isolation of enantiopure product. Gratifyingly, when
dissolved in methanol at elevated temperature, cooling
selectively precipitated product measuring 5% ee. The
remaining mother liquors contained the S-isomer in 98−99%
ee, and the product could be isolated through further addition
of water after the filtration of near racemic material.
Obtaining sufficient quantities of enriched 19 allowed for the
study of conditions to synthesize the spiro furan ring as the final
step. Deprotection of 19 was found to be best performed in
acetonitrile using catalytic 47% HBr in water at 37 °C. These
conditions avoided retro aldol and gave peak-to-peak
carbonate with a small particle size for best results, this
improvement removed a costly, moisture-sensitive base from
the synthesis. This also provided a safer process which avoided
the emission of propane.
With 2 in hand, preliminary efforts to synthesize 3 focused
on using a cosolvent in an attempt to reduce the amount of
TFA called for in the former synthesis (Scheme 1). However,
efforts to alter the conditions resulted in poor reactivity due
either to the insolubility of triethylsilane9 or competitive
formation of dimer impurity 15 (Figure 1). This impurity
Figure 1. Dimeric impurity.
proved difficult to remove by crystallization and initial
quantities of 3 were produced using neat acid to focus on
screening of conditions for an asymmetric aldol reaction.
A number of catalysts10,11 in the literature have been
reported to provide good to excellent ee under noncryogenic
conditions using formaldehyde as an electrophile. Employing
Sharpless ligands12 as catalysts gave encouraging results under
mild conditions, using unprotected 3 (Table 1). However,
significant levels of impurities were observed when long
reaction times occurred, despite performing the reactions at
lower temperatures. In addition, products associated with
hemiaminal formation were again observed.
Hoping to tune the enantioselectivity through substitution of
the phenol, several derivatives were prepared with a focus
mainly on silyl protecting groups. The TMS group was found
Table 1. Asymmetric Aldol Reaction with Protected Phenol 3
entry
catalyst
temp. (°C)
time (h)
3 (A%)
4a (A%)
4 (A%)
impurities (A%)
ee of 4 (%)
1
2
3
4
5
(DHQ)2PHAL
(DHQ)2PHAL
(DHQ)2Pyr
(DHQ)2Pyr
22
20
0
22
64
18
64
18
0.9
5.4
2.3
0.7
26
36
41
52
34
31
61
7
2.1
46.6
20.7
64.4
43
46
51
25
0
20
0
25
9
20
43
C
DOI: 10.1021/acs.oprd.7b00237
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Table 2. Asymmetric Aldol Reaction with Sharpless Ligands and Phenol-Protected Intermediates
a
entry
P
catalyst
solvent
time (h)
23 (A%)
23a (A%)
ee of 23 (%)
1
TBDMS
(DHQ)2PHAL
CH2Cl2
CH2Cl2
THF
113
90
90
18
18
18
18
21
21
18
114
18
30
41
35
87
77
91
92
92.7
71.5
82
8
62
47
27
27
13
33
32
10
25
17
30
35
2
2
(DHQ)PHN
3
(DHQ)PHN
56
4
21
CH2Cl2
THF
9
5
21
13
6
21
toluene
CH3CN
toluene
THF
3.9
3.2
3.2
25.9
7.1
91
7
21
8
22
9
22
10
11
12
22
CH2Cl2
CH2Cl2
CH2Cl2
TES
(DHQ)2PHAL
21
88
2.7
26
a
The remainder is SM.
be the bromomethyl compound. This intermediate converted
to the API upon addition of triethylamine, and product could
be isolated through solvent exchange into methanol followed
by addition of water. However, closer inspection of this
intermediate by mass spectral analysis and NMR revealed an
oxidized version of intermediate A as the phosphine oxide when
isolated after chromatography. Use of other solvents with 24
did not reveal a bromomethyl intermediate when used with
PBr3, and conversion was hindered. The examination of the ee
of the isolated API revealed scrambling at the quaternary
center. The mechanism for the scrambling was not elucidated
and remains unclear. Due to the complexity of the intermediate
and loss of stereochemical integrity, this initial approach was
quickly abandoned.
Figure 2. Thiourea catalysts for asymmetric aldol.
Table 3. Asymmetric Aldol Reaction with Thiourea Catalysts
with Oxindole-Substituted and Phenol-Protected
Intermediate Using Paraformaldehyde and 5% Catalyst
Recognizing that a phosphorus-containing leaving group
could allow for ring closure to occur, several reagents used in
ligand synthesis were examined. Diethyl chlorophosphate,
dichlorophenyl phosphine, and diethylchlorophosphite did
not provide sufficient conversion or a good impurity profile.
However, chlorodiphenylphosphine16 completely consumed 24
when heated to 37 °C and provided the product 1 without an
intermediate observed by HPLC. No scrambling of ee was
observed with this leaving group, and no base was needed to
complete the conversion. Byproducts observed from the
reaction were a mixture of diphenylphosphine oxide,
diphenylphosphinic acid, and diphenylphosphine. To remove
the stench associated with trace amounts of diphenylphosphine,
30% hydrogen peroxide in water was added to the reaction
mixture prior to workup which oxidized the malodorous
byproduct to diphenylphosphinic acid.
b
entry catalyst
solvent
time (h) 18 (A%) 19 (A%) ee (%)
1
2
3
4
5
6
7
8
22
22
22
22
21
21
21
21
CH3CN
CH3CN
heptane
heptane
heptane
CH3CN
THF
18
18
18
18
18
18
40
18
30
28
5
70
72
95
71
98
97.3
82
68
45
50
73
54
58
36
41
54
a
a
29
2
2.7
18
32
toluene
a
b
Performed using 30 mL/g vs 10 mL/g. Integration of product and
starting material in all cases.
Plant-Scale Synthesis. With a route in hand, optimization
resulted in a concise plant process (Scheme 5), which benefited
from telescoping of several steps. For the synthesis of 16, it was
found that the installation of 6 could be combined with the
stepwise addition of sesamol in one vessel. This avoided the use
of Cs2CO3 previously reported on scale to install the furan.
Under the optimized conditions, 1.01 equiv of 6 was added to a
slurry of K2CO3 and isatin in DMF at 48 3 °C followed by
heating for 1 h. Higher equivalents of isatin resulted in the
competitive formation of 2 which complicated workup. Once
the consumption of isatin was confirmed by in-process control
conversion in the HPLC to 24. The closure of 24 to provide 1
proved to be problematic. Eager to avoid the use of Mitsunobu
reagents which complicated isolation through the generation of
colored byproducts, an intramolecular Williamson ether
approach was tried. All attempts to cleanly synthesize chloro
or mesylate intermediates were unsuccessful due to the
congested neopentyl center in 24. Traditional reagents such
as thionyl chloride failed. Only in one case was the API
synthesized in good yield using PBr3 in THF (Scheme 4)
through an intermediate observed in HPLC, initially thought to
D
DOI: 10.1021/acs.oprd.7b00237
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Scheme 4. Phosphorous Tribromide-Promoted Cyclization
Scheme 5. Optimized Plant Synthesis
(IPC), a DMF solution of sesamol was added at 25−35 °C.
The product 16 was then isolated directly from the reaction
mixture through pH adjustment with aqueous acetic acid and
addition of i-PrOH. Seeding was necessary to avoid gumming
of the product as it precipitated. Product 16 from the two-step
sequence was isolated in 94−97% yield with a typical assay of
97−98 wt % and purity of 99 area percent (A%) by HPLC.
In the following step employing triethylsilane, most attempts
at using cosolvents failed. Thus, optimization of the silane
reduction conditions from the first asymmetric synthesis
became a priority. Methylene chloride, although undesirable
as a halogenated, class 2 solvent, had the advantage of poorly
solubilizing both starting material and product, thus minimizing
the formation of 15. Optimized conditions were 10 volumes of
CH2Cl2, 1.3 equiv of triethylsilane, and 6 equiv of TFA. To
avoid the formation of dimer 15, the slow addition of TFA at 0
°C, followed by warming to 25 °C and subsequent stirring at
that temperature, gave clean, complete conversion. Isolation
involved washing with water to remove TFA, followed by a
partial solvent exchange into heptane (ca. 90:10 heptane−
CH2Cl2) to give the product in 91% yield and 99.8 A% purity.
The protection of 17 was straightforward in THF with
TBDMSCl and triethylamine. To drive the reaction to
completion, 1.2 equiv of silyl reagent was needed resulting in
an indole silyl ether side product. This disilylated side product
was cleanly converted to the desired product upon pH
adjustment with HCl. The isolation of 18 was successfully
completed in 90% yield and 99.7 A% purity by filtration after
solvent exchange into i-PrOH.
The key asymmetric conversion of 18 to 19 was found to
perform best with a mixture of solid paraformaldehyde and
aqueous formaldehyde using only 1.1% of 2217 at 25 °C. High
agitation in n-heptane was needed. However, on plant scale,
additional time was called for (47 h) to reach 95% conversion
after which the product was isolated by filtration. On a small
scale, an increase in temperature allowed for shorter reaction
time but resulted in lower enantioselectivity. After drying the
crude mixture, which measured 70.5% ee, the solids were
heated in methanol at 62 °C and cooled to 25 °C, where the
near racemic product (ca. 5−6% ee) precipitated and was
collected by filtration. Unlike the SMB process, racemic 19
could be recycled by conversion back to 18 through a separate
retroaldol reaction. Recharging of the filtered solution to the
cleaned reactor, followed by the addition of water, precipitated
the product in 99% ee and 58% yield.
The final step of the synthesis could be successfully
telescoped when performed in acetonitrile. Although the
amount of HBr was not found to be critical, the amount of
water associated with the reagent was such that undercharging
of water below 0.5 equiv stalled the deprotection due to
incomplete formation of disiloxane. The removal of the
byproduct disiloxane was achieved through extraction of the
acetonitrile solution with cyclohexane. With the proper amount
of water charged, the Karl Fischer (KF) titration measurement
at the end of the reaction after extraction with cyclohexane
measured ca. 0.1% water, allowing for the addition of
chlorodiphenylphosphine without significant quenching of the
reagent. Heating at 37 °C furnished the product after
E
DOI: 10.1021/acs.oprd.7b00237
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
subsequent addition of chlorodiphenylphosphine. Isolation
involved the addition of hydrogen peroxide, which was
confirmed to be completely consumed by IPC with peroxide
paper before continuing with the workup. Solvent exchange
into methanol followed by seeding gave the product in 81%
yield and 99.6 A% purity with an ee greater than 99.9%.
an Agilent Technologies 6520 QTOF accurate mass system
coupled with an Agilent 1260 Infinity HPLC.
3-Hydroxy-3-(6-hydroxy-1,3-benzodioxol-5-yl)-1-[[5-
(trifluoromethyl)-2-furyl]methyl]indoline-2-one (16). A
200 L Hastelloy reactor was inerted, then charged with isatin
(9.0 kg, 61.2 mol, LR), powdered anhydrous potassium
carbonate (13.5 kg, 97.7 mol, 1.6 equiv), and N,N-
dimethylformamide (DMF) (33.6 kg, 4 vol). The resulting
slurry was kept under nitrogen and heated to 45 °C over 25
min. Neat 6 (14.45 kg, 63.1 mol, 1.03 equiv) was added over
about an hour such that the internal temperature was kept
between 45 and 54 °C. Once the addition was complete, the
reaction mixture was stirred at 48 °C for an hour and then
analyzed by HPLC for completeness. The reaction mixture was
then cooled to 30 °C, and a solution of sesamol (9.3 kg, 67.3
mol, 1.1 equiv) in DMF (8.5 kg, 1 vol) was added over 76 min
such that the temperature was maintained between below 30
°C. The reaction mixture was stirred at 30 °C for 3 h. An in-
process control (IPC) sample was removed to determine
reaction completeness by HPLC. The contents of the 200 L
reactor were transferred to a 400 L glass-lined jacketed reactor
with retreat-curve agitator.
CONCLUSION
■
In summary, despite the use of a protecting group, the
optimized asymmetric plant synthesis represents the shortest
synthesis of TV-45070 (XEN-402) to date. No chromatography
was used for any of the steps, and all intermediates were
isolated through direct isolation. This route uses mild, process-
friendly reagents in the first step, avoids the need for high
molecular weight bases, and quickly builds molecular complex-
ity. The modified reduction step addresses issues with dimer
formation and optimizes the amount of TFA. The use of the
TBDMS group on the phenol was shown to greatly improve
the enantiomeric excess in the subsequent aldol chemistry. The
key asymmetric step was successfully performed using a low
loading of 1.1% of thiourea catalyst. This asymmetric step did
not require cryogenic conditions and produced product in high
purity and moderate ee, which could be then upgraded to >98%
ee through recrystallization in methanol. The final stage of the
synthesis combined deprotection of the TBDMS group and
cyclization in one vessel. The use of the diphenylphosphino
group offered a cleaner alternative to the Mitsunobu reaction,
avoiding the use of azodicarboxylate reagents in the final step.
The overall yield for the five-step sequence was 35.3% from
isatin, compared with 9% in the first generation and 11.4% for
the racemic synthesis. This overall yield could be increased
further through the recycling of the racemic intermediate.
The 200 L reactor was rinsed with isopropanol (45.4 kg, 6.4
vol), and the rinse solution was added to the 400 L reactor. The
contents of the 400 L reactor were then heated to 53 °C over
about 0.5 h. Water (212 kg, 23.6 vol) was added over 54 min
while keeping the reaction temperature between 53 and 55 °C.
Next a slurry of 16 seeds (0.53 kg, 2 wt %) in water (1.35 kg)
was charged and the mixture agitated for an additional 30 min.
The contents of the reactor was cooled to 40 °C prior to
charging acetic acid (3.6 kg, 59.3 mol, 1 equiv) in water (124
kg, 13.8 vol) over 89 min while maintaining the temperature
between 41 and 44 °C. The reaction solution was cooled to 20
°C over an hour to crystallize a tan solid. The slurry was aged at
20 °C overnight. The solid was collected on an Aurora filter/
dryer, washed with water (81.3 kg, 9 vol), and then dried at 60
5 °C under a reduced pressure of 80 Torr for 3 days to afford
16 (25.5 kg, 94.1% yield) as a brown solid with purity 98.9 A%
EXPERIMENTAL SECTION
■
General Experiment. All reactions were carried out under
an atmosphere of dry nitrogen unless otherwise specified. All
reagents and solvents were used as received without further
purification unless otherwise specified. For all steps, reactions
were monitored, and the purity was assessed by HPLC, using
an Agilent 1100 or 1200 series instrument (Table 4). Diluent:
1
(HPLC-UV at 230 nm) and 97.0% assay. H NMR (DMSO,
400 MHz) δ 9.13 (s, 1H), 7.25 (s, 1H), 7.24−7.18 (m, 2H),
7.00 (d, J = 7.8 Hz, 1H), 6.90 (m, 2H), 6.58 (m, 2H),6.23 (s,
1H), 5.90 (d, J = 6.4 Hz, 1H), 5.90 (d, J = 6.4 Hz, 1H), 4.99 (s,
2H). 13C NMR (100 MHz, DMSO-d6,): 176.37, 153.69,
148.11, 146.72, 142.82, 139.48 139.30 (q, JCF = 42.4 Hz),
132.34, 128.61. 123.53, 122.18, 119.66, 119.03 (q, JCF = 266.4
Hz), 114.03 (q, JCF = 3.2 Hz), 109.08, 108.28, 106.68, 100.70,
97.35, 74.50, 36.28.
3-(6-Hydroxy-1,3-benzodioxol-5-yl)-1-[[5-(trifluoro-
methyl)-2-furyl]methyl]indolin-2-one (17). Solid 16 (18.7
kg, 43.1 mol, LR) was charged to a 400 L glass-lined reactor.
Dichloromethane (253 kg, 10.0 vol) was added, and the
Table 4. HPLC Gradient Program
time (min)
%A
%B
initial
7
90
50
50
20
90
90
10
50
50
80
10
10
14
17
17.1
22
resultant slurry was cooled to 0
3 °C over 16 min.
Triethylsilane (6.64 kg, 1.30 equiv) was added over 2 min,
keeping the internal temperature at 0 3 °C. Trifluoroacetic
acid (30.0 kg, 259 mol, 6.0 equiv) was charged in one
continuous stream to the slurry over 1 h, maintaining the
reaction temperature at ≤10 °C. A mild exotherm was noted.
The internal temperature was raised to 25 3 °C for 1 h to
complete the reaction as determined by an in-process test (0.75
A% 16 remaining, HPLC-UV).
acetonitrile. Mobile phase A: 0.1% TFA/water. Mobile phase B:
0.1% TFA/acetonitrile. Column: Zorbax Bonus-RP 150 × 4.6
mm; 3.5 μm packing. Column temperature: 40 °C. Detector
wavelengths: 230 and 250 nm. Injection volume: 5 μL. Flow
rate: 1.0 mL/min.
1H NMR (400 MHz) were recorded on a Bruker Avance II
spectrometer. Chemical shifts are reported in ppm (δ units)
downfield of internal tetramethylsilane [(CH3)4Si]; 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. LC/MS was performed on
DI water (190 kg, 10 vol) was charged to the reactor. The
mixture was stirred at 25 °C for 5 min. The stirring was
stopped, and the contents of the reactor held at 25 °C for 16
F
DOI: 10.1021/acs.oprd.7b00237
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
min, allowing the phases to separate. The lower organic layer
was dropped to drums, and the upper aqueous layer was
removed and discarded. The organic layer was returned to the
reactor, and then the DI water washing was repeated two more
times. The final wash was with a 10% brine solution (209 kg, 11
vol) .
3-[6-[tert-Butyl(dimethyl)silyl]oxy-1,3-benzodioxol-5-
yl]-3-(hydroxymethyl)-1-[[5-(trifluoromethyl)-2-furyl]-
methyl]indolin-2-one (19). A 400 L glass-lined vessel was
charged with 18 (19.2 kg, 36.1 mol, LR), 22 (238 g, 0.399 mol,
0.011 equiv), and paraformaldehyde (1.08 kg, 36.1 mol, 1.0
equiv). n-Heptane (144 kg, 11 vol) was added by vacuum, and
the suspension was agitated (123 rpm) at 25 °C for 20 min.
Aqueous formaldehyde (37% in H2O, 3.23 kg, 39.7 mol, 1.1
equiv) was added in one portion. The reaction was stirred at 25
°C for 47 h, at which point 3.0 A% of 18 remained. The crude
product was isolated at 25 °C using an Aurora filter. The initial
filtration took 16 min. The crude product was then washed with
n-heptane (19.8 kg, 1.5 vol) over 12 min. The solids were dried
on the Aurora filter under vacuum at 45 °C for 27 h to give
19.5 kg (96% yield) of crude 19. The remaining water content
was measured to be 0.6% by Karl Fischer (KF) titration, and
the chiral purity of the solids was 70.5% ee.
The crude product was then charged back into the reactor
and suspended in a premixed solution of methanol (91.2 kg, 6.0
vol) and acetic acid (5.03 kg, 0.25 vol). The slurry was stirred at
20 °C for 30 min followed by heating to 50 °C and stirring for
1 h. The slurry was cooled back to 20 °C and stirred for
approximately 20 h at 75 rpm. The racemic solids were filtered
using an Aurora filter, and the vessel and solids were washed
with methanol (30.3 kg) over 25 min. The filtrate was then
transferred into a clean 400 L glass-lined vessel through a 1 μm
cartridge filter to remove any trace racemic solids.
The organic layer was concentrated over 4.3 h at reduced
pressure (267 mbar) and an internal temperature of 10−13 °C.
Approximately 150 L of distillate was collected, resulting in a
reaction volume of 35 L. During this period the reactor internal
temperature was in the range of 10−13 °C. Heptane (104 kg, 8
vol) was next added, and the suspension was stirred for 10 min.
1
Analysis by H NMR indicated that a target level of 9.7 wt %
DCM in heptane was obtained.
After cooling to 0 °C the solid was collected by filtration.
The off-white crystalline product was washed with heptane (44
kg) and dried at 50−55 °C for 24 h to an LOD of 0.28%. The
yield was 16.4 kg, 91.1%, and chemical purity measured 100 A%
(HPLC-UV, 230 nm). 1H NMR (CDCl3, 400 MHz) δ 8.49 (s,
1H), 7.38 (t, J = 7.8 Hz, 1H), 7.32 (d, J = 7.5 Hz, 1H), 7.2 (t, J
= 7.5 Hz, 1H), 7.06 (d, J = 7.8 Hz, 1H), 6.72, (d, J = 3.5 Hz,
1H), 6.59 (s, 1H), 6.35, (d, J = 3.5 Hz, 1H), 6.34 (s, 1H), 5.86
(d, J = 16.5 Hz, 1H), 5.86 (d, J = 16.5 Hz, 1H), 5.09 (s, 1H),
4.94 (s, 2H). 13C NMR (100 MHz, DMSO-d6): 175.80, 153.59,
149.98, 146.77, 142.39, 139.69, 139.37 (q, JCF = 42.1 Hz),
129.92, 127.44, 123.64, 122.25, 120.29, 118.96 (q, JCF = 266.8
Hz), 115.38, 113.99 (q, JCF = 2.9 Hz), 109.38, 108.41, 100.75,
97.77, 47.05, 36.16.
DI water (15.4 kg, 0.8 vol) was then added over at least 20
min, giving crystal growth. After a 1 h hold, the remaining DI
water (49.9 kg, 2.6 vol) was added over 20 min. The batch was
stirred for 2 h and the solids isolated by filtration at 22 °C. The
reactor and solids were washed with a premixed solution of
methanol (17.0 kg, 1.12 vol) and DI water (9.22 kg, 0.48 vol).
The overall time for the filtration was 21 min. The solids were
dried on an Aurora filter under vacuum at 45 °C for 91 h to
give 0.18% remaining H2O by KF titration. Overall, 11.8 kg
(58.0%) of 19 was isolated as a light tan solid with an HPLC
chemical purity of 100.0 A% and chiral purity of 99.2% ee. An
3-[6-[tert-Butyl(dimethyl)silyl]oxy-1,3-benzodioxol-5-
yl]-1-[[5-(trifluoromethyl)-2-furyl]methyl]indolin-2-one
(18). A 200 L Hastelloy reactor was charged with 17 (16.2 kg,
38.8 mol, LR) and tert-butyldimethylsilyl chloride (7.60 kg, 50.4
mol, 1.3 equiv). Tetrahydrofuran (57.7 kg, 4 vol) was next
added to the vessel by vacuum. The contents were agitated at
120 rpm, while a solution formed. Next, triethylamine (8.63 kg,
85.3 mol, 2.2 equiv) was added over 14 min, giving a slight
exotherm from 20 to 30 °C. The reaction was stirred for 20 h
after which no 17 was detected by IPC through HPLC. A
solution of conc. HCl (36.5−38%, 4.01 kg, 1.05 equiv) and
sodium chloride (8.02 kg) in DI water (43.3 kg, 2.7 vol) was
added over 50 min to adjust the pH to below 1. The
temperature of the batch was maintained between 11 and 16 °C
during the addition. The batch was then stirred at 22 °C for 30
min. The top layer was sampled, and no bis-silylated side
product was detected by HPLC. The bottom layer was then
removed and discarded to waste. Next, solvent exchange was
performed into isopropanol and the temperature lowered to 10
°C. Seeds (40 g) were added, and crystallization was observed
after 37 min at 10 °C. The mixture was then stirred for 2 h. DI
water (13.0 kg, 0.8 vol) was next added over 25 min. After a 15
min hold, additional DI water (27.5 kg, 1.7 vol) was added over
28 min allowing the temperature to rise to 25 °C. The slurry
was stirred for 21 h and then isolated using an Aurora filter.
The solids were washed with isopropanol−DI water (1:1, 48.6
L, 3.0 vol) and dried on the Aurora filter below 30 °C for 22 h.
The in-process control showed a weight loss of 4.5% by TGA.
The solids were transferred into the tray dryer and dried for 26
h at 28−30 °C. TGA analysis showed 0.07% weight loss at this
point. This gave 18.2 kg (88%) of 18 as a white solid with 100
A% purity. 1H NMR (DMSO, 400 MHz) δ 7.32 (d, J = 7.5 Hz,
1H)7.18 (m, 1H), 7.14 (d, J = 7.9 Hz, 1H), 7.00 (m, 2H), 6.77
(d, J = 3.5 Hz, 1H), 6.58 (s, 1H), 6.02 (s, 1H), 5.95 (d, J = 10.1
Hz, 2H), 5.07 (m, 3H), 0.96 (s, 9H), 0.69 (s, 6H).
1
HPLC assay showed 99.9 wt % purity. H NMR (CDCl3, 400
MHz) δ 7.23 (d, J = 7.6 Hz, 1H), 7.19 (s, 1H), 6.99 (t, J = 7.6
Hz, 1H), 6.94 (d, J = 7.4 Hz, 1H), 6.91 (d, J = 7.8 Hz, 1H),
6.71 (d, J = 3.5 Hz, 1H), 6.37 (s, 1H), 6.36 (d, J = 3.5 Hz, 1H),
5.94 (d, J = 11.1 Hz, 1H), 5.94 (d, J = 11.1 Hz, 1H), 5.45 (d, J
= 16.3 Hz, 1H), 4.54 (d, J = 16.3 Hz, 1H), 4.24 (t, J = 11.3 Hz,
1H), 3.75 (d, J = 11.2 Hz, 1H), 2.59 (d, J = 10.1 Hz, 1H), 0.69
(s, 9H), −0.05 (s, 6H). 13C NMR (100 MHz, DMSO-d6):
176.83, 153.63, 147.94, 146.33, 142.50, 140.59, 139.32 (q, JCF
=
41.8 Hz), 132.13, 127.09, 123.36, 121.90, 120.24, 119.00 (q, JCF
= 266.8 Hz), 113.88 (q, JCF = 3.0 Hz), 109.37, 108.84, 108.14,
101.17, 99.52, 66.09, 55.57, 36.37, 25.93, 18.64, −3.79, −3.92.
(S)-1′-[(5-Methyl-2-furyl)methyl]spiro[6H-furo[3,2-f ]-
[1,3]benzodioxole-7,3′-indoline]-2′-one (1). To a 200 L
glass-lined reactor was charged 19 (10.7 kg, 19.0 mol, LR),
acetonitrile (33.1 kg, 4 vol), and 47−49% HBr in water (258
mL, 0.13 equiv). The batch was then heated to 37 3 °C over
20 min. Upon reaching temperature, the batch was stirred for 6
h. After this period an IPC was performed indicating no starting
material remaining by HPLC. The batch was then cooled to 20
3 °C and extracted with cyclohexane (8.35 kg, 1 vol) to
remove disiloxane. Next an IPC was performed for KF, and the
batch was found to contain 0.1% water.
Next chlorodiphenylphosphine (4.6 kg, 20.9 mol, 1.1 equiv)
was added and the temperature raised to 37 3 °C. The batch
G
DOI: 10.1021/acs.oprd.7b00237
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
was agitated at this temperature for 21 h. An IPC was
performed, and the remaining amount of 24 was measured to
be 0.15 A% by HPLC. After cooling to 20 5 °C, a solution of
30% hydrogen peroxide in water (0.65 kg, 0.3 equiv) in
acetonitrile (8.3 kg, 0.5 vol) was charged such that the
temperature did not rise above 35 °C. The batch was then
cooled to 20 5 °C and an IPC to measure peroxide showed
none detected.
Next, using an Aurora filter, the precipitated diphenylphos-
phinic acid was separated from the reaction mixture. The
resulting solution containing the product was recharged to the
reactor, and solvent was exchanged for methanol. An IPC
showed no acetonitrile detected by GC.
Once the solvent exchange was completed, the temperature
of the batch was raised to 65 5 °C and cooled to 50 5 °C
where IPC was performed and showed the concentration was
227 mg/mL. Estimating the volume of the batch at 38 L, an
additional amount of methanol (12 kg) was added to adjust the
concentration to the nominal range of 135−163 mg/mL.
To begin the crystallization of the product, the batch was
cooled to 35 3 °C, and TV-45070 seeds (110 g) were added.
After agitating for 1 h, DI water (12.7 kg, 1.19 vol) was added
while maintaining the temperature at 35 °C. The batch was
then cooled to 22 3 °C over 30 min and agitated at that
temperature for 12.5 h.
⊥H.R.: Dermira Inc., Menlo Park, California 94025, United
States.
#R.P.B.: Receptos Inc., 3033 Science Park Rd., San Diego,
California 92121, United States.
Notes
The authors declare no competing financial interest.
REFERENCES
■
(1) (a) Wasner, G.; Fleetwood-Walker, S. M.; Garry, E. M.; Abbadie,
C.; Johnson, R. W.; Baron, R. Rev. Analg. 2007, 9, 71−86. (b) Johnson,
R. W.; Rice, A. S. C. N. Engl. J. Med. 2014, 371, 1526−1533.
(2) Demant, D. T.; Lund, K.; Finnerup, N. B.; Vollert, J.; Maier, C.;
Segerdahl, M. S.; Jensen, T. S.; Sindrup, S. H. Pain 2015, 156, 2234−
2344.
(3) Burness, C. B.; McCormack, P. L. Drugs 2016, 76, 123−134.
(4) Cox, J. J.; Reimann, F.; Nicholas, A. K.; Thornton, G.; Roberts,
E.; Springell, K.; Karbani, G.; Jafri, H.; Mannan, J.; Raashid, Y.; Al-
Gazali, L.; Hamamy, H.; Valente, E. M.; Gorman, S.; Williams, R.;
McHale, D. P.; Wood, J. N.; Gribble, F. M.; Woods, C. G. Nature
2006, 444, 894−898.
(5) Tang, Z.; Chen, Z.; Tang, B.; Jiang, H. Orphanet J. Rare Dis. 2015,
10, 127.
(6) Cadieux, J.-J.; Chafeev, M.; Chowdhury, S.; Fu, J.; Jia, Q.; Abel,
S.; El-Sayed, E.; Huthmann, E.; Isarno, T. Synthetic Methods For
Spiro-Oxindole Compounds. U.S. Patent 8,445,696, May 21, 2013.
(7) Sun, S.; Fu, J.; Chowdhury, S.; Hemeon, I. W.; Grimwood, M. E.;
Mansour, T. S. Asymmetric Syntheses of Spiro-Oxindole Compounds
Useful As Therapeutic Agents. U.S. Patent 9,487,535, Nov 08, 2016.
(8) (a) Lygo, B.; Wainwright, P. G. Tetrahedron Lett. 1997, 38,
8595−8598. (b) Lygo, B.; Andrews, B. I. Acc. Chem. Res. 2004, 37,
518−525. (c) Corey, E. J.; Xu, F.; Noe, M. C. J. Am. Chem. Soc. 1997,
119, 12414−12415.
(9) (a) Carey, F. A.; Tremper, H. S. J. Am. Chem. Soc. 1968, 90,
2578−2583. (b) Carey, F. A.; Tremper, H. S. J. Am. Chem. Soc. 1969,
91, 2967−2972. (c) Carey, F. A.; Tremper, H. S. J. Org. Chem. 1971,
36, 758−761.
The product was then filtered using an Aurora filter, and the
cake was washed with a preprepared 23% aqueous methanol
solution (4 vol) followed by water (18 kg). The solids were
agitated in the wash solvent while on the filter prior to
application of vacuum. The solids were then dried under
vacuum (50−80 mmHg) for 3.5 days. After that period, the
product was found to have LOD of 0.087%. The final weight
was 6.59 kg (80.7% yield). The purity was 99.6 A% with 100%
1
ee. H NMR (DMSO, 400 MHz) δ 7.32 (t, J = 7.7 Hz, 1H),
7.20 (m, 3H), 7.07 (t, J = 7.3 Hz, 1H), 6.77 (d, J = 3.3 Hz, 1H),
6.72 (s, 1H), 6.10 (s, 1H), 5.94 (d, J = 9.1 Hz, 1H), 5.94 (d, J =
9.1 Hz, 1H), 5.13 (d, J = 16.5 Hz, 1H), 5.02 (d, J = 16.5 Hz,
1H), 4.82 (d, J = 9.5 Hz, 1H), 4.73 (d, J = 9.5 Hz, 1H). 13C
NMR (100 MHz, DMSO-d6): 176.48, 155.28, 153.02, 148.40,
141.80, 141.51, 139.54 (q, JCF = 41.9 Hz), 131.63, 128.79,
123.64, 123.29, 119.69, 118.92 (q, JCF = 266.4 Hz), 114.01 (q,
JCF = 2.9 Hz) 109.86, 109.21, 102.55, 101.44, 93.31, 79.52,
57.41, 36.44.
(10) For a recent review of aldol reactions with formaldehyde see
Meninno, S.; Lattanzi, A. Chem. Rec. 2016, 16, 2016−2030.
(11) For examples of aldol reactions with formaldehyde see:
(a) Shen, K.; Liu, X.; Wang, W.; Wang, G.; Cao, W.; Li, W.; Hu,
X.; Lin, L.; Feng, X. Chem. Sci. 2010, 1, 590−595. (b) Kobayashi, S.;
Kokubo, M.; Kawasumi, K.; Nagano, T. Chem. - Asian J. 2010, 5, 490−
492. (c) Boeckman, R. K., Jr.; Miller, J. R. Org. Lett. 2009, 11, 4544−
4547.
(12) Ogawa, S.; Shibata, N.; Inagaki, J.; Nakamura, S.; Toru, T.;
Shiro, M. Angew. Chem., Int. Ed. 2007, 46, 8666−8669.
(13) For examples of thiourea catalysts with oxindoles see: (a) Liu,
X.-L.; Liao, Y.-H.; Wu, Z.-J.; Cun, L.-F.; Zhang, X.-M.; Yuan, W.-C. J.
Org. Chem. 2010, 75, 4872−4875. (b) Bui, T.; Syed, S.; Barbas, C. F.,
III J. Am. Chem. Soc. 2009, 131, 8758−8759.
ASSOCIATED CONTENT
■
S
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.oprd.7b00237.
́ ́ ́ ́
(14) (a) Tarkanyi, G.; Kiraly, P.; Varga, S.; Vakulya, B.; Soos, T.
Chem. - Eur. J. 2008, 14, 6078−6086. (b) Singh, R. P.; Foxman, B. M.;
Deng, L. J. Am. Chem. Soc. 2010, 132, 9558−9560. (c) Connon, S. J.
Chem. Commun. 2008, 2499−2510. (d) Fukata, Y.; Miyaji, R.;
Okamura, T.; Asano, K.; Matsubara, S. Synthesis 2013, 45, 1627−
Spectral data of selected intermediates and of the final
compound (PDF)
́ ́
1634. (e) Vakulya, B.; Varga, S.; Csampai, A.; Soos, T. Org. Lett. 2005,
7, 1967−1969. (f) Okino, T.; Hoashi, Y.; Takemoto, Y. J. Am. Chem.
Soc. 2003, 125, 12672−12673.
(15) For reviews on crystallization-based separation of enantiomers
see: (a) Collet, A. Enantiomer 1999, 4, 157−172. (b) Lorenz, H.;
Capla, F.; Polenske, D.; Elsner, M. P.; Seidel-Morgenstern, A. J. Chem.
Technol. Metall. 2007, 42, 5−16. (c) Wang, Y.; Chen, A. M. Org.
Process Res. Dev. 2008, 12, 282−290.
(16) Shintou, T.; Mukaiyama, T. J. Am. Chem. Soc. 2004, 126, 7359−
7367.
(17) For a large scale synthesis of 22 and the corresponding quinine-
based derivative, see: Wang, Y.; Milkiewicz, K. L.; Kaufman, M. L.; He,
L.; Landmesser, N. G.; Levy, D. V.; Allwein, S. P.; Christie, M. A.;
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: jasclafan@yahoo.com.
ORCID
Joseph A. Sclafani: 0000-0003-4126-9463
Present Addresses
§J.A.S., D.V.L.: Incyte Corporation, 1801 Augustine Cut-Off,
Wilmington, Delaware 19803, United States.
∥J.C.: Celgene Corporation, 86 Morris Ave., Summit, New
Jersey 07901, United States.
H
DOI: 10.1021/acs.oprd.7b00237
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Olsen, M. A.; Neville, C. J.; Muthukumaran, K. Org. Process Res. Dev.
2017, 21, 408−413.
I
DOI: 10.1021/acs.oprd.7b00237
Org. Process Res. Dev. XXXX, XXX, XXX−XXX