Development of a Factory Process for Omecamtiv Mecarbil, a Novel Cardiac Myosin Activator

Article abstract of DOI:10.1021/acs.oprd.9b00200

The development of a factory process to manufacture the novel cardiac myosin activator omecamtiv mecarbil (1) is described. Omecamtiv mecarbil is prepared via the convergent synthesis and coupling of two key fragments, aniline 2 and carbamate 4-HCl, which serves as a masked isocyanate. To enable practical access to aniline 2, reduction of the corresponding nitroaromatic was designed to control potential mutagenic impurities. Key to the efficient preparation of 2 was the benzylic bromination of 8 followed by selective debromination of a gem-dibromide byproduct and subsequent alkylation with 5-phosphate. Overall, the longest linear sequence consists of six steps, including a final salt formation step to afford the drug substance in 55% overall yield. Because of poor performance of the original free-base form of the drug substance in modified-release formulations, an improved dihydrochloride hydrate form was developed to aid drug product performance and manufacturability.

Full text of DOI:10.1021/acs.oprd.9b00200

Article
pubs.acs.org/OPRD
Cite This: Org. Process Res. Dev. XXXX, XXX, XXX−XXX
Development of a Factory Process for Omecamtiv Mecarbil, a Novel
Cardiac Myosin Activator
Seb Caille,*,† Alan M. Allgeier,† Charles Bernard,† Tiffany L. Correll,⊥ Andrew Cosbie,†
Richard D. Crockett,† Sheng Cui,‡ Margaret M. Faul,† Karl B. Hansen,‡ Seth Huggins,† Neil Langille,‡
Steven M. Mennen,‡ Bradley P. Morgan,§ Henry Morrison,∥ Alexander Muci,§ Karthik Nagapudi,∥
Kyle Quasdorf,† Krishnakumar Ranganathan,† Philipp Roosen,† Xianqing Shi,† Oliver R. Thiel,‡
Fang Wang,⊥ Justin T. Tvetan,† Jacqueline C. S. Woo,† Steven Wu,⊥ and Shawn D. Walker*,‡
†Drug Substance Technologies, Process Development, Amgen Inc., One Amgen Center Drive, Thousand Oaks, California 91320,
United States
‡Drug Substance Technologies, Process Development, Amgen Inc., 360 Binney Street, Cambridge, Massachusetts 02142, United
States
§Research & Nonclinical Development, Cytokinetics, 280 East Grand Avenue, South San Francisco, California 94080, United States
∥Drug Product Technologies, Process Development, Amgen Inc., One Amgen Center Drive, Thousand Oaks, California 91320,
United States
⊥Attribute Sciences, Amgen, Inc., One Amgen Center Drive, Thousand Oaks, California 91320, United States
S
* Supporting Information
ABSTRACT: The development of a factory process to manufacture the novel cardiac myosin activator omecamtiv mecarbil (1)
is described. Omecamtiv mecarbil is prepared via the convergent synthesis and coupling of two key fragments, aniline 2 and
carbamate 4-HCl, which serves as a masked isocyanate. To enable practical access to aniline 2, reduction of the corresponding
nitroaromatic was designed to control potential mutagenic impurities. Key to the efficient preparation of 2 was the benzylic
bromination of 8 followed by selective debromination of a gem-dibromide byproduct and subsequent alkylation with 5-
phosphate. Overall, the longest linear sequence consists of six steps, including a final salt formation step to afford the drug
substance in 55% overall yield. Because of poor performance of the original free-base form of the drug substance in modified-
release formulations, an improved dihydrochloride hydrate form was developed to aid drug product performance and
manufacturability.
KEYWORDS: cardiac myosin activator, omecamtiv mecarbil, masked isocyanate, benzylic bromination, nitro reduction,
urea formation
synthesis of the unsymmetrical urea core of the target from
stable precursors as well as engineering of the drug substance
form and particle properties to aid drug product manufactur-
ability and performance.
INTRODUCTION
■
Heart failure is a degenerative and terminal disease that affects
over 25 million people globally.1 The high prevalence, high
mortality rates, and significant health resource utilization
associated with heart failure underscore the need for new
medicines to treat this serious illness.2 Omecamtiv mecarbil
(1) is a first-in-class direct activator of cardiac myosin, the
motor protein of the cardiac sarcomere that transfers chemical
energy into the mechanical force required for heart
contraction.3 By promotion of the force-generating cycle of
myosin, cardiac contractility is improved without altering
cardiac myocyte intracellular calcium cycling.4 Accordingly, 1
is being evaluated as a potential treatment for chronic heart
failure with the goal of establishing a new continuum of care
for patients in both the hospital and outpatient settings. To
support the drug development program, including pivotal
Phase 3 studies, practical access to multihundred-kilogram
quantities of 1 were required. This article describes the design
of a robust synthetic process to prepare omecamtiv mecarbil
and highlights the challenges and opportunities at the drug
substance−drug product interface. This includes a convergent
As shown in Scheme 1, 1 possesses an unsymmetrical urea
core containing both aryl and heteroaryl components. Thus, a
convergent route to 1 would comprise the synthesis and
coupling of two key fragments, aniline 2 and isocyanate 3. To
enable large-scale manufacture of aniline 2, an efficient
synthesis starting from cheap and readily available precursors
was needed. Key to this approach were (a) the development of
a reliable benzylic bromination process for 8 that overcame the
challenges associated with overbromination, (b) the develop-
ment of a selective nitro reduction process ensuring control of
potential mutagenic impurities in the drug substance, and (c)
Special Issue: Honoring 25 Years of the Buchwald-Hartwig
Amination
Received: May 2, 2019
© XXXX American Chemical Society
A
DOI: 10.1021/acs.oprd.9b00200
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Scheme 1. Key Intermediates in the Retrosynthesis of Omecamtiv Mecarbil
identification of a suitable salt form of 5 to improve its stability.
Additionally, the poor stability and physiochemical properties
of isocyanate 3 necessitated the development of a stable
“masked isocyanate” 4 for use in a urea formation step. Lastly,
the drug substance final form, crystallization, and milling
process needed to be designed to ensure adequate drug
product performance in controlled-release formulations. This
paper describes how each of these challenges was addressed to
develop a factory process to manufacture omecamtiv mecarbil.5
1 of the sequence employed a relatively expensive aryl chloride
raw material, and the cyanation process required stoichio-
metric nickel, long reaction times (days), and elevated
temperatures (200 °C) to proceed. In the subsequent
reductive amination step, small levels of over-reduction and
polyalkylation side products were observed, necessitating a
tedious sequence of salt formation and salt-break crystallization
steps to purge these process impurities. The isocyanate starting
material 3 was not practical for long-term use since it was
found to be an unstable oil that degraded to a mixture of dimer
and trimer impurities during storage (Figure 1).7 The
RESULTS AND DISCUSSION
■
Initial Clinical Manufacturing Route. The six-step
synthesis of omecamtiv mecarbil used to supply early clinical
studies is depicted in Scheme 2. In the first step of the process,
Scheme 2. Initial Clinical Manufacturing Route to
Omecamtiv Mecarbil
Figure 1. Degradation impurities that accumulate in isocyanate 3
during storage.
formation of these insoluble impurities also meant that reagent
3 needed to be filtered prior to use. Furthermore, because of
the potential mutagenicity of the isocyanate and its late-stage
introduction in the process, an additional recrystallization step
from alcohol solvent was added as a control measure to destroy
any residual isocyanate. In practice, this involved dissolution of
1 in methanol (50 volumes), polish filtration, a solvent switch
to ethanol (10 volumes) via distillation, and cooling to
complete the crystallization. Although this typically afforded 1
with excellent purity (>99.5 LC area %), the material was
isolated as a mixture of small needle-shaped particles and
agglomerated bundles, which resulted in long filtration and
drying times.8
Development of an Improved Process for Aniline 2. A
more robust and cost-effective synthesis of aniline 2 was
devised that started from the relatively inexpensive and readily
available nitrotoluene 8. It was proposed that radical
bromination of 8 followed by alkylation with methyl
piperazine-1-carboxylate (5) and subsequent nitro reduction
would represent an attractive route for large-scale production.
The benzylic bromination was attempted using a variety of
brominating reagents [N-bromosuccinimide (NBS), Br2,
dibromodimethylhydantoin], solvents (AcOH, MeCN, IPAC,
CCl4, cyclohexane) and initiators (PhCO2−O2CPh, AIBN),
and the conditions shown in Scheme 3 were found to be
optimal. Thus, semibatch addition of NBS to 8 in acetic acid
(5 volumes) in the presence of 3% benzoyl peroxide at 83 °C
nickel-mediated cyanation of 7 afforded benzonitrile 9.
Treatment of 9 with DIBAl-H followed by reductive amination
with 5 in the presence of sodium triacetoxyborohydride
(STAB)6 afforded aniline 2. Compound 2 was initially isolated
as the corresponding dihydrochloride salt followed by a salt-
break step to crystallize the desired free-base form. Finally,
coupling of aniline 2 with isocyanate 3 generated the
unsymmetrical urea core of omecamtiv mecarbil. Although
this fit-for-purpose process was suitable to supply early clinical
development, it was not optimal for the long-term manufacture
of 1, and a more robust and efficient process was desired. Step
B
DOI: 10.1021/acs.oprd.9b00200
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Scheme 3. Kilo-Lab Process for 14-HCl
for 12 h afforded a mixture of monobrominated product 12
and dibrominated product 13 in an ∼2:1 ratio. Because of the
highly corrosive nature of the bromination mixture, it was
essential to ensure that there were no exposed metal parts in
the reactor (see Figure 2).9 At the end of the reaction,
Figure 3. Structure of dialkylation impurity 15.
Scheme 4. Preparation of 5-Phosphate for Use in the
Factory Process
Figure 2. Corrosion of a Hastelloy impeller (left) and C22 coupon
(right) by the bromination reaction mixture.
via reaction with 1.1 equiv of methyl chloroformate followed
by pH adjustment and washing to remove excess piperazine as
well as bis(carbamate) side product 16 before final
crystallization of 5-phosphate. The process was readily
scalable, and 5-phosphate was prepared in 51% yield with
>99 LC area % purity. By the use of this salt in the optimized
alkylation process, 14-HCl was successfully prepared in 81%
overall yield with >99 LC area % purity (Scheme 5).
phosphorous acid (0.1 equiv) was charged to quench residual
peroxide. Addition of water and toluene was followed by
separation of the phases. The toluene phase was washed with
aqueous sodium hydroxide to remove acetic acid and provide a
solution of benzylic bromides 12 and 13 in toluene. The
mixture of bromides could be employed directly in the
alkylation step, as only monobromide 12 reacted with 5 to
afford nitropiperazine 14. The latter was isolated as a
crystalline hydrochloride salt from a mixture of toluene and
isopropanol. The crystallization process was effective in
rejecting process impurities (including the dibromide),
providing 14-HCl in 63% overall yield with >99% purity.
Although further reaction optimization was unable to reduce
the level of gem-dibromide 13, selective debromination of this
material was successfully achieved via treatment of the mixture
of 12 and 13 with diethyl phosphite.10 Thus, the crude
polybrominated mixture in toluene was treated with diethyl
phosphite (0.46 equiv) in methanol in the presence of
diisopropylethylamine for 3 h at 40 °C to afford a >95%
assay yield of the desired monobromide 12.
Scheme 5. Plant Process for 14-HCl
During the course of development, it was found that
commercial batches of starting material 5 contained varying
amounts of residual piperazine and that the levels of this
impurity increased during storage. To avoid the formation of
impurity 15 (Figure 3), which could not be purged through
crystallization of 14-HCl, the amount of piperazine in 5 had to
be controlled to a low level (<0.5%). To address this issue, 5-
phosphate, a stable crystalline sesquiphosphate salt of 5, was
developed (Scheme 4). This salt was prepared as a hydrate
with high purity (containing <5000 ppm piperazine) and could
be stored at room temperature and used directly in the
alkylation process. This material was prepared from piperazine
With nitroaryl 14-HCl in hand, attention was focused on the
development of a scalable catalytic hydrogenation process to
prepare aniline 2.11 Nitro hydrogenations proceed via a
stepwise process [nitro (14) → nitroso (17) → hydroxylamine
(18) → aniline (2)], and in addition to the intermediates
formed along the major pathway, several disproportionation
side products [including azoxy (19), azo (20), and hydrazo
C
DOI: 10.1021/acs.oprd.9b00200
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
(21)] can be generated before final reduction to the aniline
(Scheme 6). As these intermediates and side products are
product and mother liquor were found to be consistently low,
leaching of iron from the Pd(Fe)/C catalyst occurred, and dark
product solutions were obtained that necessitated extra cake
washing to prevent contamination and discoloration of the
product. Furthermore, in this case the iron-modified catalyst
was not effective at preventing accumulation of hydroxylamine
18 (observed to be over 50% by LC when 14 was completely
consumed), and an alternative system was desired. The
crystallization from methanol/water was challenging, requiring
the use of concentrated brine as an antisolvent for good
product recovery and thorough washing of the filter cake to
remove residual NaCl. Finally, although it was operationally
convenient to use the salt 14-HCl directly in the reduction
process, the acid promoted the formation of aminophenol
impurity 23, presumably via Bamberger rearrangement of
hydroxylamine intermediate 18.14
Scheme 6. Intermediates and Side Products Formed during
Hydrogenation of Nitro 14
Additional experimentation revealed that the free base 14
was a suitable substrate for the hydrogenation process and that
in this case catalyst modifiers/additives were not needed to
ensure clean conversion to 2 with low levels of process
impurities. Accordingly, a Pd/C catalyst (Escat 1421) was
selected to replace the specialty Pd(Fe)/C catalyst for the
reduction step. As an improvement to the original crystal-
lization of 2 from an aqueous mixture, a robust orthogonal
crystallization process from toluene/heptane was developed,
enabling better rejection of impurities. To further streamline
operations, toluene solvent was also back-integrated into the
hydrogenation step. Ethanol cosolvent was added to ensure
miscibility of the water generated during the reaction and
prevent clustering of the catalyst (with water) along the walls
of the vessel. Thermal hazards evaluation confirmed that the
reaction was highly exothermic (ΔHrxn = −566 kJ/mol) with
an adiabatic temperature rise of 78 °C. The catalyst loading
and operating temperature were found to be the most
important parameters influencing the hydrogenation reaction
kinetics. Since the rate of the competing debenzylation (to give
22) also increased at higher temperatures, maintaining the
batch temperature below 25 °C was preferred. As shown in
Figure 4, the reaction progress in a small-scale run was
monitored in situ via ReactIR, which demonstrated that 14 was
consumed after ∼5 h to afford ∼35% yield of hydroxylamine
18 and ∼65% of 2, with the additional hold time serving to
complete the conversation to 2. For the optimized plant
process (Scheme 8), a salt-break procedure was developed that
considered potential mutagenic impurities, the reaction design
and isolation must ensure that they are controlled to low levels
in the final drug substance.12
Screening of hydrogenation conditions revealed that the
reaction proceeded in a variety of solvents using Pd, Pt, or Ni
catalysts. The nitroso (17) and hydroxylamine (18) inter-
mediates and the azoxy (19) and azo (20) side products were
typically reduced smoothly under the hydrogenation con-
ditions, and the residual hydrazo (21) and debenzylation (22)
side products were the materials commonly carried into the
isolation step. Fortunately, 22 was an oil that was well rejected
in the downstream crystallization steps, and 21 was shown to
be nonmutagenic in a bacterial reverse mutation test (in vitro
Ames test). For the first kilo-lab campaign, a Pd(Fe)/C
catalyst was selected because of its successful performance in
reducing intermediate hydroxylamine levels in a previous
development program.13 Nitro salt 14-HCl was used directly
in the hydrogenation process in a 1:1 methanol/water mixture
with 1% catalyst at 60 psi H2. High conversion was achieved
after several hours at 20−30 °C, but the batch was aged
overnight (19 h) to ensure complete consumption of the
reaction intermediates. Although the nitro−HCl starting
material had poor solubility in this solvent system, the
corresponding aniline salt was fully soluble, and after filtration
to remove the catalyst, the pH of the filtrate was adjusted to
crystallize 2. By this process (Scheme 7), the aniline was
isolated in 87% yield on a 9.1 kg scale and contained <1 ppm
residual palladium. However, before moving to a larger scale,
several issues needed to be addressed to improve the process
performance and efficiency. Although the Pd levels in both the
Scheme 7. Catalytic Hydrogenation: Lab to Kilo-Lab
Process for Aniline 2
Figure 4. Reaction profile for the reduction of 14 with 0.7% Pd/C at
15 °C and 60 psi H2.
D
DOI: 10.1021/acs.oprd.9b00200
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
clean solution of 1 in >95% assay yield. Importantly, excess
iPr2NEt was necessary for high conversion and purity, serving
to both neutralize the HCl introduced with the carbamate and
promote isocyanate release via E1cb elimination of phenol.17
The crystallization of 1 was completed by charging water (5
volumes), seeding, and adding additional water (10 volumes)
over 3 h (Scheme 10), thus completing a practical isolation of
the free-base form of omecamtiv mecarbil.18 In contrast to the
fine particles obtained in the initial synthetic process, in this
case agglomeration of the product needles produced large oval
particles during the addition of water antisolvent (Figure 5).
Formation of these large agglomerates (∼200−2000 μm in
length) led to fast filtration and drying times, affording
crystalline free base 1 with >99.9 LC area % purity in 82%
yield.
Design of Drug Substance Final Form and Bulk
Properties. During the drug development program, five
different modified release (MR) oral formulations, an
immediate release oral formulation, and an intravenous (IV)
formulation were contemporaneously evaluated. Because of the
potential risk of excessive pharmacological effects (e.g.,
myocardial ischemia or infarction) at high plasma concen-
trations of omecamtiv mecarbil, the MR formulations were
developed with the goal of preserving bioavailability while
lowering Cmax and the peak to trough ratio as well as
attenuating the influence of gastrointestinal pH on drug release
and absorption. These formulations spanned a range of
delivery technologies, including swellable core tablets (osmotic
pumps), matrix tablets, and multiparticulate capsules.19 Given
the challenges in developing the various drug products, our
objective was to reduce any complexity introduced by the drug
substance across the formulations. Therefore, the drug
substance needed to be delivered with a consistent
morphology, controlled particle size distribution (PSD), low
aspect ratio, and suitable bulk flow properties and had to be
stable during manufacture of the various tablets/capsules.20
However, MR matrix tablet prototypes revealed that the
free-base form of omecamtiv mecarbil reacted with acidic
excipients in the formulations (including citric or fumaric
acid), generating high levels of amorphous citrate/fumarate
salts.21 To avoid undesired chemistry and mixtures of forms in
the tablets, salt screening was conducted to identify an
alternative drug substance form that would be stable during
drug product manufacture and storage. As a result, the
omecamtiv mecarbil dihydrochloride salt hydrate form was
identified and subsequently selected for long-term develop-
ment. The salt readily crystallized in high yield from
isopropanol/water mixtures as long rodlike particles. Poly-
morph screening on the dihydrochloride salt revealed only one
Scheme 8. Catalytic Hydrogenation: Plant Process for
Aniline 2
consisted of treatment of 14-HCl with aqueous NaOH in
toluene, separation of the phases, and telescoping of the
organic phase into the hydrogenation step. The reduction was
performed in toluene/ethanol (6 volumes) at 15−20 °C for 12
h using 0.7% Pd/C at 60 psi H2. Upon reaction completion,
the catalyst was removed via filtration, and the filtrate was
concentrated by distillation to remove ethanol and water. The
batch was seeded at 35 °C and cooled to 20 °C and heptane
was added to complete the crystallization of 2, which was
isolated in 90% yield with 99 LC area % purity.
Development of a Masked Isocyanate and Urea
Formation. After evaluating a variety of carbamates as
potential isocyanate equivalents, we selected phenyl carbamate
salt 4-HCl as affording the best balance of stability, ease of
preparation, and reactivity in the subsequent coupling step.15
This material was prepared by a one-pot direct drop isolation
process from aminopyridine 6 and phenyl chloroformate
(Scheme 9).16 Aminopyridine 6 was dissolved in acetonitrile
(15 volumes) at 20 °C, and phenyl chloroformate was added
(over ∼3 h) at a rate low enough to maintain the batch
temperature below 30 °C. Reactive crystallization of 4-HCl
was well-controlled by the feed rate, and the product was
isolated in excellent yield after filtration. By means of this
process, 4-HCl was prepared in 95% yield with >98 LC area %
purity. Importantly, in contrast to the unstable isocyanate
reagent employed previously, the crystalline carbamate 4-HCl
was stable for over 2 years at room temperature and for over 6
months under the accelerated conditions of 40 °C and 75%
relative humidity.
The urea coupling between aniline 2 and carbamate 4-HCl
to generate omecamtiv mecarbil was studied extensively by
varying the solvent, base, temperature, operating mode (batch
vs semibatch, workup vs direct isolation), and crystallization
performance. The coupling process proceeded as a slurry-to-
slurry transformation under many of the conditions examined,
with poorly controlled reactive crystallizations of the product.
The heavy precipitation of salts and product in many solvents
also resulted in thick mixtures that were difficult to stir.
However, it was found that heating a mixture of 2 (1.0 equiv),
4-HCl (1.1 equiv), and iPr2NEt (1.2 equiv) as the base in 1:1
THF/acetonitrile (5 volumes) at 55 °C for 20 h afforded a
Scheme 9. Preparation of Carbamate 4-HCl
E
DOI: 10.1021/acs.oprd.9b00200
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Scheme 10. Urea Formation and Crystallization of Omecamtiv Mecarbil (1)
line PSD analysis via laser light diffraction (Figure 7). The
rotor tip speed was kept constant during scale-up to achieve a
consistent terminal particle size. After wet milling, the slurry
was filtered, washed, and dried. Water levels in the filter cake
consistent with the monohydrate form could be maintained by
using a final wash with 5% v/v water in isopropanol and
vacuum drying. By means of this process, omecamtiv mecarbil
dihydrochloride hydrate was prepared in 93% yield with
>99.9% purity (Scheme 11).
Figure 5. Agglomeration of free base 1 needles during crystallization:
(a) PVM image after 1 h of water addition; (b) PVM image after 3 h
of water addition; (c) microscope image of the isolated product at
10× magnification, showing ∼750 μm long particles.
CONCLUSIONS
■
A safe, efficient and robust process to manufacture the novel
cardiac myosin activator omecamtiv mecarbil has been
developed.23 The longest linear sequence consists of six
steps, including a final salt formation step to afford the drug
substance in ∼55% overall yield. Relative to the initial route
used to supply preclinical and early clinical studies, the E-factor
for the optimized process was reduced from 308 to 72 (>75%
reduction).24 Additionally, as a result of the amorphous
conversion of the free-base form in the presence of excipients,
a more stable dihydrochloride hydrate form of omecamtiv
mecarbil was developed. Overall, the process has been
successfully implemented on a large scale to afford >1
metric ton of drug substance to support registrational studies.
Scheme 12 summarizes the plant process to prepare
omecamtiv mecarbil dihydrochloride hydrate.
hydrated form that was nonhygroscopic and existed as a
channel hydrate as per single-crystal X-ray analysis. Impor-
tantly, no omecamtiv mecarbil dihydrochloride hydrate form
conversion or erosion of crystallinity was observed during MR
matrix tablet manufacturing.
A scalable salt formation process was developed that
involved treatment of a solution of omecamtiv mecarbil in
isopropanol with aqueous HCl (2.2 equiv), followed by polish
filtration, cooling, and addition of more isopropanol to
complete the dihydrochloride hydrate crystallization (Scheme
11). Wet milling with an in-line high-shear rotor−stator mixer
Scheme 11. Omecamtiv Mecarbil (1) Dihydrochloride
Hydrate Salt Formation and Isolation
EXPERIMENTAL SECTION
■
Manufacture of 5-Phosphate. A 6000 L glass-lined
jacketed reactor set at 20 °C under a nitrogen atmosphere and
vented through a scrubber (containing 5 N NaOH) was
charged with piperazine (450 kg, 5224 mol, 1.0 equiv)
followed by water (1800 L, 4 volumes). The suspension was
heated to 35−40 °C and agitated until dissolved. The resulting
solution was cooled to 20 5 °C, and methyl chloroformate
(543 kg, 5746 mol, 1.1 equiv) was added over 5 h so as to
maintain the internal temperature below 30 °C. The reaction
mixture was washed twice with dichloromethane (2 × 450 L, 2
volumes) to remove 16. The pH of the aqueous layer was
adjusted with 30 wt % aqueous sodium hydroxide (557 kg,
4179 mol, 0.8 equiv) to a pH of 9.8. Sodium chloride (450 kg,
1 kg/kg of piperazine) was charged, and the mixture was
agitated until the solids were dissolved. 5 was extracted from
this aqueous solution (containing unreacted piperazine) three
times with dichloromethane (3 × 1800 L, 12 volumes). The
combined organic extracts were then solvent-switched to
methyl tert-butyl ether (MTBE) by concentration under
reduced pressure (300−400 mbar) to 2.5 volumes while the
internal batch temperature was maintained at ≤25 °C followed
by charging of MTBE (2028 L, 4.5 volumes). This
concentration and MTBE addition process was repeated. A
final batch concentration to 2.5 volumes was followed by
was implemented to reduce the aspect ratio of the crystals,
improve the bulk properties, and aid downstream drug product
processing (Figure 6).22 During development runs, in-line
focused beam reflectance measurement (FBRM) was used to
monitor the progress of the milling and the number of cycles
required to pass in-process-test acceptance criteria. Slurry
samples were collected periodically to profile the milling by off-
Figure 6. Crystallization of omecamtiv mecarbil dihydrochloride
hydrate (a) before and (b) after wet milling.
F
DOI: 10.1021/acs.oprd.9b00200
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Figure 7. Crystallization of omecamtiv mecarbil dihydrochloride hydrate. A good correlation between in-line FBRM chord length measurements
(mean chord length in green) and off-line PSD (dark-red dots) was observed.
Scheme 12. Factory Process To Prepare Omecamtiv Mecarbil Dihydrochloride Hydrate
charging of MTBE (1350 L, 3.0 volumes). The mixture was
then transferred to an 8000 L glass-lined jacketed reactor
through a lenticular filter to remove any salts that precipitated
during the solvent switch. To a separate vessel, MTBE (1575
L, 3.5 volumes) and 85 wt % aqueous phosphoric acid (211 kg,
1828 mol, 0.35 equiv) were charged. The batch in the 8000 L
equiv), benzoyl peroxide (14.1 kg, 44.0 mol, 0.03 equiv,
containing 25 wt % water), and acetic acid (450 L, 2.0
volumes) was heated at 83 °C under an atmosphere of
nitrogen for 1.5 h. A suspension of N-bromosuccinimide (408
kg, 2292 mol, 1.58 equiv) and acetic acid (792 L, 3.52
volumes), prepared in a separate vessel, was added in three
portions (approximately 1/4, 1/4, and 1/2 of the total mass)
over 5 h at 83 °C. Acetic acid (225 L, 1.0 volume) was added,
rinsing any residual N-bromosuccinimide into the reactor. The
reaction mixture was agitated at 83 °C for 12 h. A solution of
H3PO3 (17.0 kg, 145 mol, 0.1 equiv, containing 30 wt % water)
and acetic acid (22.5 L, 0.1 volume), prepared in separate
vessel, was added, and the mixture was agitated for 2.5 h. The
reaction mixture was cooled to 25 °C, and deionized (DI)
water (1240 L, 5.5 volumes) and toluene (1800 L, 8 volumes)
were charged. The biphasic mixture was agitated (30 min), and
the layers were separated. Aqueous 1.6 N NaOH (1600 L, 7.1
volumes) was added to the organic layer at a rate allowing the
batch temperature to stay under 25 °C, and the pH of the
resultant aqueous phase was measured to be 12.5. The biphasic
mixture was filtered through a sparkler filter, and the layers
were separated. To the organic layer at 20 °C was added
diisopropylethylamine (99.35 kg, 769 mol, 0.53 equiv)
followed by methanol (20 L, 494 mol, 0.34 equiv), and the
mixture was heated to 40 °C. A solution of diethyl phosphite
reactor was heated to 40
solution in MTBE was added over 3 h, maintaining an internal
temperature of 40 5 °C. The resulting suspension was
5 °C, and the phosphoric acid
cooled to 20 5 °C over 2 h and agitated for an additional 1 h.
The product was isolated in a centrifuge, washing the cake with
MTBE (900 L, 2 volumes) and deliquoring. The wet cake was
dried in an agitated contact dryer under vacuum (10−50
mbar) with a 30 °C jacket for 16 h to afford 573.7 kg (51%
yield) of the product 5-phosphate as white solid with >99 LC
1
area % purity. H NMR (400 MHz, DMSO-d6) δ 7.42 (br s,
3H) 3.60 (s, 3H) 3.40−3.48 (m, 4H) 2.79−2.85 (m, 4H);
13C{1H} NMR (101 MHz, DMSO-d6) δ 155.0, 52.4, 43.5,
42.3; melting point = 95−97 °C.
Manufacture of 14-HCl. In a 6000 L glass-lined reactor
(containing no exposed metal parts) equipped with a reflux/
return condenser and a scrubber charged with a 5 N NaOH
solution, a mechanically stirred mixture of 8 (225.0 kg, 1450
mol, 1.0 equiv), N-bromosuccinimide (25.8 kg, 145 mol, 0.1
G
DOI: 10.1021/acs.oprd.9b00200
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
(91.5 kg, 667 mol, 0.46 equiv) in methanol (82 L, 2024 mol,
1.4 equiv) was prepared and added to the reaction mixture at
40 °C over 2.5 h at such a rate that the batch temperature
stayed at 40 5 °C. The contents were stirred for a period of 5
h at 40 °C until complete conversion to 12 was achieved.
Characterization data for 12: 1H NMR (400 MHz, chloroform-
d) δ 7.98−8.05 (m, 1H), 7.69−7.76 (m, 1H), 7.28−7.34 (m,
1H), 4.56 (s, 2H); 13C{1H} NMR (100 MHz, chloroform-d) δ
153.3, 137.7, 136.5, 128.5, 126.1, 124.5, 23.6; melting point =
53−55 °C; HRMS (ESI-TOF) m/z calcd for C7H6BrFNO2 [M
+ H]+ 233.9566, found 233.9561.
In a 6700 L glass-lined reactor equipped with a reflux/return
condenser and scrubber charged with a 5 N NaOH solution, a
mechanically stirred mixture of 5-phosphate (336 kg, 1538
mol, 1.06 equiv) and methanol (360 L, 1.6 volumes) was
heated to 40 °C. To a separate 3000 L reactor containing a
toluene solution (9 volumes) of 12 (prepared in the previous
step) at 25 °C under an atmosphere of nitrogen were added
diisopropylethylamine (433 kg, 3350 mol, 2.31 equiv) and
methanol (45 L, 0.2 volumes). This mixture was added to the
reactor containing the methanol solution of 5-phosphate over
1 h, followed by a vessel rinse with methanol (180 L, 0.8
volume) and toluene (225 L, 1.0 volume). The combined
mixture was agitated at 25 °C for 6.5 h, and analysis of a
sample confirmed completion of the alkylation reaction. The
reaction mixture was treated with water (1125 L, 5.0 volumes),
and the biphasic mixture was agitated for 30 min, after which
the layers were separated and the aqueous phase was discarded.
The organic layer was washed twice with aqueous NH4Cl (20
wt %, 1240 L, 5.5 volumes; prepared from 225 kg of NH4Cl
and 1125 L of DI water), and the biphasic mixture was agitated
for 30 min. The layers were separated, and the organic layer
was washed with aqueous NaHCO3 (9 wt %, 1175 L, 5.2
volumes; prepared from 101 kg of NaHCO3 and 1125 L of DI
water). The organic layer was then filtered through a 5 μm
Teflon in-line cartridge filter into a 6700 L glass-lined reactor,
followed by a vessel and filter rinse with toluene (225 L, 1.0
volume).
Manufacture of 2. To a 3600 L jacketed reactor were
added 14-HCl (340 kg, 1.00 equiv) and toluene (1360 L, 4.0
volumes). To the suspension was added 4 wt % aqueous
sodium hydroxide solution (1161 kg, 1.1 equiv). The mixture
was stirred for >15 min, resulting in a clear biphasic mixture.
Agitation was stopped, and the layers were allowed to settle.
The organic layer was washed twice with 6.5 wt % aqueous
sodium chloride (1091 kg, 1.2 equiv). The resulting organic
layer was concentrated to 3.8 volumes under reduced pressure
and transferred through an in-line filter into a 4000 L Hastelloy
hydrogenator, removing precipitated salts. The 3600 L reactor
and lines were rinsed with an additional portion of toluene
(340 L, 1.0 volume). To the reaction mixture were added 5.0
wt % palladium on carbon (2.38 kg, Escat 1421, approximately
50% water) and ethanol (340 L, 1.0 volume). The hydro-
genator was purged three times with nitrogen, cooled to 10 °C,
and then pressurized to 60
reaction mixture was stirred at 15 5 °C while a hydrogen
pressure of 60 5 psig was maintained until reaction was
5 psig with hydrogen. The
deemed complete, waiting for 2 h after hydrogen uptake ceased
(typically 10−15 h). At the end of the reaction, the
hydrogenator was purged with nitrogen twice (pressurized to
60 10 psig and then vented to atmospheric pressure). The
crude reaction mixture was filtered through a sparkler filter
followed by a 0.45 μm cartridge filter in series into a 3600 L
reactor. The hydrogenator, filter, and lines were washed with
an additional aliquot of toluene (340 L, 1.0 volume). The
reaction mixture was concentrated under reduced pressure
(typically 50−300 mbar) at a batch temperature of 50 5 °C
until the total reaction volume was approximately 820 L (2.4
volumes). The batch was cooled to 38 3 °C and seeded with
2.7 kg of 2. The resulting slurry was stirred at 35 5 °C for 1 h
and then cooled to 20 5 °C over 1 h. Heptane (1900 L, 5.6
volumes) was added over 3 h, and the batch was allowed to stir
at 20 5 °C for >1 h. The reaction slurry was filtered in an
agitated filter dryer, and the cake was displacement-washed
with a mixture of toluene (204 L, 0.6 volumes) and heptane
(476 L, 1.4 volumes) followed by heptane (680 L, 2.0
volumes). The cake was dried under vacuum at 50 °C for >12
h until sample dryness was confirmed by LOD analysis.
Product 2 (249 kg, 90% yield) was isolated with 99.0 wt %
purity by HPLC and 98.9 LC area % purity (containing <1
To a mechanically stirred toluene solution of free base 14
(prepared as described above) at 22 °C in a 6700 L reactor
under an atmosphere of nitrogen, isopropanol (2183 L, 9.7
volumes) and DI water (93 L, 0.41 volume) were charged. The
mixture was heated to 55 °C, and 20% of the 32 wt % aqueous
HCl (41.76 kg, 367 mol, 0.25 equiv) was charged. The
contents were agitated for 15 min, and 14-HCl seed (15 kg, 45
mol, 0.03 equiv) was charged as a suspension in isopropanol
(45 L, 0.2 volumes). The mixture was agitated for 30 min, and
the remaining 32 wt % aqueous HC1 (162.4 kg, 1425 mol,
0.983 equiv) was added over a period of 4 h. The mixture was
stirred at 55 °C for 1 h, cooled to 20 °C in a linear manner
over 3 h, and agitated at this temperature for 10 h. The
suspension of the product was filtered in an agitated filter
dryer. The filter cake was washed twice with isopropanol (1260
L, 5.6 volumes) and dried to constant weight under vacuum at
a jacket temperature of 30−50 °C (42 h), and 14-HCl (400.5
kg, 81.3% corrected yield) was isolated with 100 wt % and 99.7
1
ppm Pd). H NMR (400 MHz, MeOH-d4) δ 6.86 (t, J = 7.7
Hz, 1H), 6.76 (td, J = 8.2, 1.6 Hz, 1H), 6.65 (m, 1H), 3.67 (s,
3H), 3.57 (d, J = 1.4 Hz, 2H), 3.44−3.48 (m, 4H), 2.44−2.48
(m, 4H); 13C{1H} NMR (100 MHz, MeOH-d4) δ 157.7,
153.2, 150.9, 137.0, 124.9, 121.3, 117.7, 56.5, 53.7, 53.4, 44.8;
melting point = 89−91 °C; HRMS (ESI-TOF) m/z calcd for
C13H19FN3O2 [M + H]+ 268.1456, found 268.1448.
Manufacture of 4-HCl. An 8000 L glass-lined, jacketed
reactor set at 20 °C under a nitrogen atmosphere and vented
through a scrubber (containing 5 N NaOH) was charged with
6 (125 kg, 1156 mol, 1.0 equiv) followed by acetonitrile (1875
L, 15 volumes). Phenyl chloroformate (190 kg, 1214 mol, 1.05
equiv) was added via a pump over 3 h so as to maintain the
internal temperature at or below 30 °C. The reaction mixture
was agitated for 3 h at 20 5 °C and then filtered using an
agitated filter dryer. The mother liquor was forwarded to a
glass-lined jacketed reactor for quenching of excess phenyl
chloroformate with NaOH. The product cake was washed
twice with acetonitrile (2 × 250 L, 4 volumes) and dried under
vacuum at 35 °C for 24 h to afford 290 kg (95% yield) of 4-
1
LC area % purity. H NMR (300 MHz, DMSO-d6) δ 8.17−
8.34 (m, 1H), 7.98−8.15 (m, 1H), 7.44−7.63 (t, J = 8 Hz,
1H), 4.47 (s, 2H), 3.52−3.66 (m, 8H), 3.25 (br s, 3H);
13C{1H} NMR (75 MHz, DMSO-d6) δ 155.7, 154.7, 152.2,
139.8, 137.4, 127.9, 125.1, 119.6, 52.8, 51.4, 50.3; melting
point = 248−250 °C; HRMS (ESI-TOF) m/z calcd for
C13H17FN3O4 [M + H]+ 298.1203, found 298.1198.
1
HCl as a solid with 100 wt % and 99.6 LC area % purity. H
H
DOI: 10.1021/acs.oprd.9b00200
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
NMR (400 MHz, DMSO-d6) δ 11.24 (s, 1H), 8.81 (s, 1H),
8.41 (d, J = 8.8 Hz, 1H), 7.85 (d, J = 8.8 Hz, 1H), 7.48−7.44
(m, 2H), 7.32−7.26 (m, 3H), 2.69 (s, 3H); 13C{1H} NMR
(100 MHz, DMSO-d6) δ 151.7, 150.0, 147.5, 136.1, 133.8,
130.0, 129.5, 127.8, 125.9, 121.7, 18.6; melting point = 224−
226 °C; HRMS (ESI-TOF) m/z calcd for C13H13N2O2 [M +
H]+ 229.0972, found 229.0961.
D2O) δ 8.87 (s, 1H), 8.22 (d, J = 8.9 Hz, 1H), 7.83 (t, J = 7.5
Hz, 1H), 7.71 (d, J = 8.8 Hz, 1H), 7.35−7.29 (m, 2H), 4.48 (s,
2H), 4.24 (br s, 2H), 3.73 (s, 3H), 3.31 (br s, 6H), 2.68 (s,
3H); 13C{1H} NMR (150 MHz, D2O) δ 156.8, 154.2, 153.9,
147.8, 136.3, 136.1, 130.1, 129.4, 128.0, 127.2, 125.5, 125.1,
116.1, 53.5, 50.9, 40.5, 18.2; melting point = 252−254 °C.
ASSOCIATED CONTENT
* Supporting Information
Manufacture of Omecamtiv Mecarbil (1). To a 12 000
L glass-lined reactor were charged 2 (222 kg, 829 mol, 1.0
equiv), 4-HCl (242 kg, 912 mol, 1.1 equiv), tetrahydrofuran
(554 L, 2.5 volumes), acetonitrile (554 L, 2.5 volumes), and
diisopropylethylamine (129 kg, 998 mol, 1.2 equiv). The batch
was heated at 57 °C for 21 h. Water (1109 L, 5.0 volumes) was
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.oprd.9b00200.
1
Copies of H and 13C spectra (PDF)
added, maintaining the batch temperature at 55
5 °C,
AUTHOR INFORMATION
Corresponding Authors
*E-mail: scaille@amgen.com.
*E-mail: walkers@amgen.com.
followed by seeding with 1 (16.8 kg, 42 mol, 0.05 equiv).
Water (2218 L, 10.0 volumes) was added to the reactor over 3
h. The resulting suspension was cooled to 22 °C and filtered in
an agitated filter dryer. The filter cake was washed with a
mixture of acetonitrile (166 L, 0.75 volumes) and water (499
L, 2.25 volumes) followed by three cake washes of acetonitrile
(3 × 665 L, 9 volumes). The cake was dried under vacuum
(10−50 mbar) with a jacket temperature of 35−40 °C for 17 h.
Omecamtiv mecarbil (1) (287 kg, 82% yield) was isolated as a
white solid with 99.6 wt % purity by HPLC and 99.9 LC area
% purity. 1H NMR (400 MHz, DMSO-d6) δ 9.10 (s, 1H), 8.57
(d, J = 2.3 Hz, 1H), 8.46 (d, J = 2.5 Hz, 1H), 8.03 (t, J = 7.3
Hz, 1H), 7.82 (dd, J = 8.4, 2.6 Hz, 1H), 7.17 (d, J = 8.5 Hz,
1H), 7.09 (t, J = 7.9 Hz, 1H), 7.01 (t, J = 6.8 Hz, 1H), 3.54−
3.60 (m, 6H), 3.31−3.30 (m, 4H), 2.34−2.42 (m, 6H);
13C{1H} NMR (101 MHz, DMSO-d6) δ 155.0, 152.3, 151.9,
151.1. 139.2, 133.6, 127.3, 127.2, 125.8, 124.1, 124.0, 123.8,
122.8, 119.5, 54.5, 52.1, 43.4, 23.2; melting point = 197−202
°C; HRMS (ESI-TOF) m/z calcd for C20H25FN5O3 [M + H]+
402.1941, found 402.1936.
■
ORCID
Seb Caille: 0000-0001-5434-5483
Steven M. Mennen: 0000-0003-1905-5530
Shawn D. Walker: 0000-0001-8827-4679
Notes
The authors declare no competing financial interest.
REFERENCES
■
(1) Savarese, G.; Lund, L. H. Global Public Health Burden of Heart
Failure. Card. Failure Rev. 2017, 3, 7−11.
(2) (a) Heidenreich, P. A.; Albert, N. M.; Allen, L. A.; Bluemke, D.
A.; Butler, J.; Fonarow, G. C.; Ikonomidis, J. S.; Khavjou, O.;
Konstam, M. A.; Maddox, T. M.; Nichol, G.; Pham, M.; Pina, I. L.;
̃
Trogdon, J. G. Forecasting the impact of heart failure in the United
States: a policy statement from the American Heart Association. Circ.:
Heart Failure 2013, 6, 606−619. (b) Malinowski, J. T.; St. Jean, D. J.,
Jr. Next-generation small molecule therapies for heart failure: 2015
and beyond. Bioorg. Med. Chem. Lett. 2018, 28, 1429−1435.
Manufacture of Omecamtiv Mecarbil (1) Dihydro-
chloride Hydrate. 1 (218 kg, 542 mol, 1.0 equiv) was
charged to a 12 000 L glass-lined reactor, followed by water
(333 L, 1.35 volumes) and 2-propanol (566 L, 2.60 volumes).
The slurry was agitated and heated to approximately 40 °C,
whereupon aqueous 6 N HCl (199 L, 1192 mol, 2.2 equiv)
was charged to the slurry, resulting in a colorless homogeneous
solution. The solution was heated to between 55 and 60 °C
and transferred through an in-line filter to a 6000 L glass-lined
reactor preheated to 55 °C. A rinse of the reactor and filter
consisting of 2-propanol (72 L, 0.33 volume) and water (72 L,
0.33 volume) was also transferred to the 6000 L reactor. The
batch was cooled to 45 °C, and omecamtiv mecarbil (1)
dihydrochloride hydrate seed (8 kg, 3.7 wt %) was charged.
The mixture was aged at 45 °C for 0.5 h, and then 2-propanol
(846 L, 3.9 volumes) was added over 3 h. A heat cycle (heating
to 55 °C for 0.5 h followed by cooling to 45 °C over 1 h) was
conducted, and the mixture was agitated at 45 °C for an
additional 2 h. 2-Propanol (1603 L, 7.37 volumes) was added
over 4 h, and the mixture agitated at 45 °C for 1 h and cooled
to 20 °C over 2 h. The batch was recirculated through a wet
mill for three to four batch turnovers to reduce the particle
aspect ratio. The slurry was filtered in an agitated filter dryer,
and the cake was washed twice with 95:5 v/v 2-propanol/water
solution (2 × 435 L, 4 volumes). The cake was dried under
vacuum with a jacket temperature of 35−40 °C. Omecamtiv
mecarbil (1) dihydrochloride hydrate (248 kg, 93% yield) was
isolated as a colorless crystalline solid with 100 wt % purity by
(3) (a) Malik, F. I.; Hartman, J. J.; Elias, K. A.; Morgan, B. P.;
Rodriguez, H.; Brejc, K.; Anderson, R. L.; Sueoka, S. H.; Lee, K. H.;
Finer, J. T.; Sakowicz, R.; Baliga, R.; Cox, D. R.; Garard, M.; Godinez,
G.; Kawas, R.; Kraynack, E.; Lenzi, D.; Lu, P. P.; Muci, A.; et al.
Cardiac myosin activation: a potential therapeutic approach for
systolic heart failure. Science 2011, 331, 1439−1443. (b) Morgan, B.
P.; Muci, A.; Lu, P. P.; Qian, X.; Tochimoto, T.; Smith, W. W.;
Garard, M.; Kraynack, E.; Collibee, S.; Suehiro, I.; Tomasi, A.; Valdez,
S. C.; Wang, W.; Jiang, H.; Hartman, J.; Rodriguez, H. M.; Kawas, R.;
Sylvester, S.; Elias, K. A.; Godinez, G.; Lee, K.; Anderson, R.; Sueoka,
S.; Xu, D.; Wang, Z.; Djordjevic, N.; Malik, F. I.; Morgans, D. J., Jr.
Discovery of omecamtiv mecarbil the first, selective, small molecule
activator of cardiac myosin. ACS Med. Chem. Lett. 2010, 1, 472−7.
(4) Although inotropes such as dobutamine, dopamine, and
milrinone increase cardiac contractility, they do so by activating
second messenger pathways (β-receptor, cAMP, and protein kinase A
pathways, respectively) that raise the levels of cardiac calcium
concentrations in the sarcomere. Activation of the second messenger
pathways increases both oxygen consumption and heart rate, can
produce cardiac arrhythmias, and is associated with higher rates of
mortality in heart failure patients that limit their utility. For more
information, see: (a) Kass, D. A.; Solaro, R. J. Circulation 2006, 113
(2), 305−315. (b) Cohn, J. N.; Goldstein, S. O.; Greenberg, B. H.;
Lorell, B. H.; Bourge, R. C.; Jaski, B. E.; Gottlieb, S. O.; McGrew, F.;
DeMets, D. L.; White, B. G. N. Engl. J. Med. 1998, 339 (25), 1810−
1816. (c) Packer, M.; Carver, J. R.; Rodeheffer, R. J.; Ivanhoe, R. J.;
DiBianco, R.; Zeldis, S. M.; Hendrix, G. H.; Bommer, W. J.; Elkayam,
U.; Kukin, M. L. N. Engl. J. Med. 1991, 325 (21), 1468−1475.
(5) (a) Caille, S.; Quasdorf, K.; Roosen, P.; Shi, X.; Cosbie, A.;
Wang, F.; Wu, Z.; Neergunda, A.; Quan, B. P.; Guan, L.. WO 2019/
1
HPLC and >99.9 LC area % purity. H NMR (500 MHz,
I
DOI: 10.1021/acs.oprd.9b00200
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
006231 A1. (b) Cui, S.; Morrison, H.; Nagapudi, K.; Walker, S.;
Bernard, C.; Hansen, K. B.; Langille, N. F.; Allgeier, A. M.; Mennen,
S.; Woo, J.; Morgan, B. P.; Muci, A. WO 2014/152270 A1.
(6) Gribble, G. W. The Synthetic versatility of acyloxy borohydrides.
Org. Process Res. Dev. 2006, 10, 1062−1075.
(7) Isocyanates are known to undergo self-condensation reactions to
form cyclic dimers (uretidinediones) and trimers (isocyanurates).
See: (a) Buckles, R. E.; McGrew, L. A. Kinetic Study of the
Dimerization of Phenyl Isocyanate. J. Am. Chem. Soc. 1966, 88, 3582−
3586. (b) Malwitz, N.; Kresta, J. E. Amine Catalysis of the Aryl
Isocyanate Reaction. J. Cell. Plast. 1988, 24, 495−523. (c) Nawata, T.;
Kresta, J. E.; Frisch, K. C. J. Cell. Plast. 1975, 11, 267. (d) Wirpsza, Z.
Polyurethanes: Chemistry, Technology and Application; Ellis Horwood:
London, 1993.
(8) (a) A 42 kg batch of drug substance produced via this process
required a filtration time of 3 days and a drying time of over 1 week.
(b) For a detailed summary of the solid-state characteristics of the
anhydrous free-base form of omecamtiv mecarbil and comparison to a
hemihydrate form, see: Morrison, H.; Quan, B. P.; Walker, S. D.;
Hansen, K. B.; Nagapudi, K.; Cui, S. Appearance of a New Hydrated
Form during Development: A Case Study in Process and Solid-State
Optimization. Org. Process Res. Dev. 2015, 19, 1842−1848. (c)
Prolonged refluxing of omecamtiv mecarbil in EtOH during the
recrystallization can lead to degradation, which lowers the yield. For
example, characterization studies demonstrated that heating ome-
camtiv mecarbil in ethanol at reflux for 6 h resulted in an erosion of
the solution chemical purity from >99% to 86% by LC analysis..
(9) Reaction vessels for the bromination process were vented
through a circulating scrubber containing aqueous 5 N NaOH.
(10) (a) Hirao, T.; Masunaga, T.; Ohshiro, Y.; Agawa, T. Reduction
of gem-Dibromides with Diethyl Phosphite. J. Org. Chem. 1981, 46,
3745−3747. (b) Liu, P.; Chen, Y.; Deng, J.; Tu, Y. An Efficient
Method for the Preparation of Benzylic Bromides. Synthesis 2001,
2001, 2078−2080.
(11) For recent lead references on methods for the reduction of
nitroarenes, see: (a) Shen, M.; Liu, H.; Yu, C.; Yin, Z.; Muzzio, M.; Li,
J.; Xi, Z.; Yu, Y.; Sun, S. J. Am. Chem. Soc. 2018, 140, 16460−16463.
(b) Song, J. J.; Huang, Z. F.; Pan, L.; Li, K.; Zhang, X. W.; Wang, L.;
Zou, J. J. Review on selective hydrogenation of nitroarene by catalytic,
photocatalytic and electrocatalytic reactions. Appl. Catal., B 2018,
227, 386. (c) Blaser, H. U.; Steiner, H.; Studer, M. Selective Catalytic
Hydrogenation of Functionalized Nitroarenes: An Update. Chem-
CatChem 2009, 1, 210. (d) Tafesh, A. M.; Weiguny, J. A review of the
selective catalytic reduction of aromatic nitro compounds into
aromatic amines, isocyanates, carbamates, and ureas using CO.
Chem. Rev. 1996, 96, 2035. (e) Kasparian, A. J.; Savarin, C.; Allgeier,
A. M.; Walker, S. D. Selective Catalytic Hydrogenation of Nitro
Groups in the Presence of Activated Heteroaryl Halides. J. Org. Chem.
2011, 76, 9841−9844.
(15) (a) For the synthesis of ureas from phenyl carbamates see:
Thavonekham, B. A practical synthesis of ureas from carbamates.
Synthesis 1997, 1997, 1189. Adamiak, R. W.; Stawinski, J. Tetrahedron
Lett. 1977, 18, 1935. (b) For the use of isopropenyl carbamates see:
Gallou, I.; Eriksson, M.; Zeng, X.; Senanayake, C.; Farina, V. Practical
Synthesis of Unsymmetrical Ureas from Isopropenyl Carbamates. J.
Org. Chem. 2005, 70, 6960−6963. (c) For the use of N-
hydroxysuccinimide carbamates, see: Morpurgo, M.; Bayer, E. A.;
Wilchek, M. N-Hydroxysuccinimide carbonates and carbamates are
useful reactive reagents for coupling ligands to lysines on proteins. J.
Biochem. Biophys. Methods 1999, 38, 17−28. (d) For the use of tert-
butoxycarbamates, see: Lamothe, M.; Perez, M.; Colovray-Gotteland,
V.; Halazy, S. Synlett 1996, 1996, 507. (e) For a discussion of the
relative reactivities of various carbamates in producing unsymmetrical
ureas as well as the electrochemical preparation of trifluoroethylcar-
bamates, see: Matsumura, Y.; Satoh, Y.; Onomura, O.; Maki, T. J. Org.
Chem. 2000, 65, 1549−1551.
(16) For a paper highlighting similar examples of the bottom-up
approach to process development, including several direct isolation
processes, see: Chen, C.-K.; Singh, A. K. Org. Process Res. Dev. 2001,
5, 508−513. Also see: Anderson, N. G. Practical Process Research &
Development; Academic Press: San Diego, CA, 2000.
(17) Ilieva, S.; Nalbantova, D.; Hadjieva, B.; Galabov, B. Aminolysis
of Phenyl N-Phenylcarbamate via an Isocyanate Intermediate: Theory
and Experiment. J. Org. Chem. 2013, 78, 6440−6449.
(18) After addition of water at 55 °C, excess carbamate 4 is
hydrolyzed to aminopyridine 6 and phenol.
(19) (a) Herrlich, S.; Spieth, S.; Messner, S.; Zengerle, R. Osmotic
micropumps for drug delivery. Adv. Drug Delivery Rev. 2012, 64,
1617−1627. (b) Kumar, S.; Jeet, K.; Baldi, A. Recent Technological
Advancements in Multiparticulate Formulations: The Smart Drug
Delivery Systems. Asian J. Pharm. 2015, 9, S13−S25. (c) Dash, T. R.;
Verma, P. Matrix Tablets: An Approach towards Oral Extended
Release Drug Delivery. Int. J. Pharm. Res. Rev. 2013, 2, 12−24.
(20) For a study correlating the drug substance particle size
distribution with other bulk properties (e.g., surface area and powder
flow) to enable a drug product formulation process, see: Morrison, H.
G.; Tao, W.; Trieu, W.; Walker, S. D.; Cui, S.; Huggins, S.; Nagapudi,
K. Correlation of Drug Substance Particle Size Distribution with
Other Bulk Properties to Predict Critical Quality Attributes. Org.
Process Res. Dev. 2015, 19, 1076−1081.
(21) The crystalline and amorphous contents in the tablet
formulations were quantified by solid-state 19F NMR spectroscopy.
(22) For some lead references on the application of wet milling to
pharmaceuticals, see: (a) Harter, A.; Schenck, L.; Lee, I.; Cote, A.
High-Shear Rotor-Stator Wet Milling for Drug Substances: Expanding
Capability with Improved Scalability. Org. Process Res. Dev. 2013, 17,
1335−1344. (b) Acevedo, D.; Kamaraju, V. K.; Glennon, B.; Nagy, Z.
K. Modeling and Characterization of an in Situ Wet Mill Operation.
Org. Process Res. Dev. 2017, 21, 1069−1079.
(23) For an alternative Pd-catalyzed synthesis of unsymmetrical
ureas, including omecamtiv mecarbil, see: (a) Breitler, S.; Oldenhuis,
N. J.; Fors, B. P.; Buchwald, S. L. Synthesis of Unsymmetrical
Diarylureas via Pd-Catalyzed C-N Cross-Coupling Reactions. Org.
Lett. 2011, 13, 3262−3265. (b) Vinogradova, E. V.; Fors, B. P.;
Buchwald, S. L. Palladium-Catalyzed Cross-Coupling of Aryl
Chlorides and Triflates with Sodium Cyanate: A Practical Synthesis
of Unsymmetrical Ureas. J. Am. Chem. Soc. 2012, 134, 11132−11135.
(24) Environmental (E) factor = (kg of waste)/(kg of product). See:
Sheldon, R. A. The E factor: fifteen years on. Green Chem. 2007, 9,
1273−1283.
(12) The ICH M7 TTC-based acceptable intake of a mutagenic
impurity present in pharmaceuticals for long-term (>10 years)
treatment is 1.5 μg/day. See: http://www.ich.org/fileadmin/Public_
Web_Site/ICH_Products/Guidelines/Multidisciplinary/M7/M7_
R1_Addendum_Step_4_31Mar2017.pdf.
(13) Iron, vanadium, and other additives have been reported to
reduce the accumulation of hydroxylamine intermediates in nitro
group reductions. In the screening for this project, an Fe-doped
catalyst was used instead of adding Fe powders or salts to existing
catalyst systems. For a lead reference on metal modifiers used in
catalytic supported hydrogenations, see: Baumeister, P.; Blaser, H.-U.;
Studer, M. Strong reduction of hydroxylamine accumulation in the
catalytic hydrogenation of nitroarenes by vanadium promoters. Catal.
Lett. 1997, 49, 219−222.
(14) For early accounts of the Bamberger rearrangement, see:
(a) Bamberger, E. Ber. Dtsch. Chem. Ges. 1894, 27, 1548−1557.
(b) Robertson, G. R.; Evans, R. A. J. Org. Chem. 1940, 5, 142−145.
Although this was a minor impurity, it carried forward through the
urea formation step into the final drug substance at ∼0.1%.
J
DOI: 10.1021/acs.oprd.9b00200
Org. Process Res. Dev. XXXX, XXX, XXX−XXX