A Practical Synthesis of the Platelet Fibrinogen Antagonist, Elarofiban

Article abstract of DOI:10.1021/op034103o

Elarofiban is a novel, nonpeptide, orally active fibrinogen receptor antagonist useful for the treatment of platelet mediated thrombotic disorders (Costanzo, M. J.; Hoekstra, W. J.; Maryanoff, B. E. WO, 97/41102,1997). Herein we describe the process research that was carried out for the synthesis of elarofiban that eventually led to the development of a safe and cost-effective commercial scale process.

Full text of DOI:10.1021/op034103o

Organic Process Research & Development 2003, 7, 866−872
A Practical Synthesis of the Platelet Fibrinogen Antagonist, Elarofiban
Judith H. Cohen,* Mary Ellen Bos,| Sergio Cesco-Cancian,† Bruce D. Harris,§ John T. Hortenstine,§ Michael Justus,‡
Cynthia A. Maryanoff,§ John Mills,§ Stefan Muller,‡ Armin Roessler,‡ Lorraine Scott,§ Kirk L. Sorgi,†
Frank J. Villani, Jr.,§ Robin R. H. Webster,† and Christian Weh‡
Johnson & Johnson Pharmaceutical Research & DeVelopment LLC, Drug EValuation, Chemical & Pharmaceutical
DeVelopment Department, Spring House, PennsylVania 19477, U.S.A., Johnson & Johnson Pharmaceutical Research &
DeVelopment LLC, Drug EValuation, Chemical & Pharmaceutical DeVelopment Department, Raritan, New Jersey 08869,
U.S.A., and Cilag AG Pharmaceutical and Chemical Operations, Hochstrasse 201, 8205 Schaffhausen, Switzerland.
Abstract:
Elarofiban is a novel, nonpeptide, orally active fibrinogen
receptor antagonist useful for the treatment of platelet mediated
thrombotic disorders (Costanzo, M. J.; Hoekstra, W. J.;
Maryanoff, B. E. WO, 97/41102, 1997). Herein we describe the
process research that was carried out for the synthesis of
elarofiban that eventually led to the development of a safe and
cost-effective commercial scale process.
cally enriched â-amino ester 3 was prepared in low overall
yield via enzymatic resolution of 3-phenylacetylamino-3-
pyridin-3-ylpropionic acid; however, this procedure was both
volume inefficient and expensive; (2) Although the coupling
of â-amino ester 3 and N-boc-(R)-nipecotic acid (4) pro-
ceeded in good yield, the use of expensive 2-[1H-benzo-
triazol-1-yl]-1,1,3,3-tetramethyluronium hexafluorophosphate
(HBTU) coupling reagent was required to prevent racem-
ization. Furthermore, purification of the resulting product 5
was difficult as it is a thick oil thus requiring column
chromatography; (3) In the final coupling step, HBTU was
again required to avoid epimerization and the product 7 was
purified by column chromatography; (4) Elarofiban was
isolated as an amorphous dihydrochloride salt in low overall
yield (approximately 7%); and (5) In addition to these issues,
we needed to eliminate the use of hazardous solvents such
as chloroform and dioxane. Therefore, a strong need for the
development of a more efficient synthesis existed.
Synthesis Strategy. As shown in Scheme 2, there are
two coupling strategies possible for assembling elarofiban.
Route A, used in the original Drug Discovery synthesis, was
problematic on large-scale production for reasons mentioned
above. After preliminary evaluation, we determined Route
B to be the more efficient one to prepare large quantities of
drug substance for two main reasons. First, the starting
materials and intermediates are crystalline solids easily
purified by recrystallization, thus avoiding column chroma-
tography. Second, the most expensive starting material,
methyl (S)-3-amino-3-(3-pyridyl)propionate is used at the end
of the process which lowers the cost of production. Herein,
we describe the process research that was carried out for
the synthesis of elarofiban using Route B, which eventually
Introduction
Integrins are heterodimeric molecules composed of R and
â subunits which combine to form ligand-specific receptors.1-3
Antagonists of the glycoprotein GPIIb/IIIa integrin (of the
RIIbâ3 subunit found on the surface of platelets) were shown
to be effective as platelet aggregation inhibitors.4 As part of
our research towards the synthesis of orally active, nonpep-
tide GPIIb/IIIa antagonists, we identified a novel series of
nipecotamide analogues which are potentially useful in the
treatment of platelet-mediated thrombotic disorders such as
re-occlusion of an artery following thrombolytic therapy,
acute myocardial infarction, and unstable angina.2 The lead
compound in this series, elarofiban, was chosen for further
development because it can be dosed both intravenously and
orally, has a good duration of action, and demonstrated
excellent safety characteristics.
To prepare the necessary supplies of drug substance to
support both toxicological and clinical studies, we needed
to develop an efficient large-scale process. We began by
evaluating the existing synthetic method developed by Drug
Discovery (Scheme 1),5 which presented many challenges
for the preparation of material in large quantity. Specifically,
we needed to address the following areas: (1) Enantiomeri-
* Author for correspondence. E-mail: jcohen@prius.jnj.com. Telephone:
(215) 628-5059. Fax: (215) 628-7067.
† Johnson & Johnson Pharmaceutical Research & Development LLC, Raritan,
NJ.
§ Johnson & Johnson Pharmaceutical Research & Development LLC, Spring
House, PA.
‡ Cilag AG Pharmaceutical and Chemical Operations.
| Current address: 19502 Encino Gap, San Antonio, TX 78259.
(1) Costanzo, M. J.; Hoekstra, W. J.; Maryanoff, B. E. WO, 97/41102, 1997.
(2) Shinagawa, S.; Kanamaru, T.; Asai, M.; Okazaki, H. J. Med. Chem. 1987,
30, 1458.
(3) Scott, V. R.; Boehme, R.; Matthews, T. R. Antimicrob. Agents Chemother.
1988, 32, 1154.
(4) Kawabata, N.; Inamoto, Y.; Sakane, K.; Iwamoto, T.; Hashimoto, S. J.
Antibiot. 1992, 45, 513.
(5) Hoekstra, W. J.; Maryanoff, B. E.; Damiano, B. P.; Andrade-Gordon, P.;
Cohen, J. H.; Costanzo, M. J.; Haertlein, B. J.; Hecker, L. R.; Hulshizer, B.
L.; Kauffman, J. A.; Keane, P.; McComsey, D. F.; Mitchell, J. A.; Scott,
L.; Shah, R. D.; Yabut, S. C. J. Med. Chem. 1999, 42, 5254.
866
•
Vol. 7, No. 6, 2003 / Organic Process Research & Development
10.1021/op034103o CCC: $25.00 © 2003 American Chemical Society
Published on Web 10/01/2003

Scheme 1 a
a
Reagents and conditions: (a) CH2(CO2H)2, NH4OAc, EtOH; (b) PhCH2COCl, NEt3, aq Me2CO; (c) penicillin amidase, chromatography; (d) 6N HCl, reflux; (e)
HCl, MeOH; (f) HBTU, HOBT, chromatography; (g) LiOH, aq THF; (h) HCl, dioxane, chromatography.
bonyl)amino-3-(3′-pyridyl)propionic acid)8 afforded inter-
Scheme 2
mediate 12. The ester was saponified using LiOH, and finally
the Cbz group was removed via catalytic hydrogenation to
afford crude elarofiban which was purified via recrystalli-
zation from n-BuOH.
While this process was suitable for the preparation of the
initial supplies of drug substance, it identified other issues
that needed to be addressed for the production of larger
quantities of material. Specifically: (1) the use of expensive
PtO2 catalyst and formation of a dimeric byproduct during
the protection step in the synthesis of 3-(N-benzyloxycar-
bonyl-4-piperidyl)propionoic acid (9); (2) development of a
more cost-effective method to prepare the enantiomerically
pure â-amino ester 3; (3) the use of enantiomerically enriched
R-ethyl nipecotate tartrate since there was only one com-
mercial supplier and the material was very expensive; and
(4) increase in the overall yield of the process to make it
more cost-efficient.
led to the development of a safe and cost-effective com-
mercial scale process.6
Synthesis of 3-(N-Benzyloxycarbonyl-4-piperidyl)pro-
pionoic (9). In 1966, DeGraw and Kennedy reported the
synthesis of acid 9 via catalytic hydrogenation of 4-pyr-
idinelacrylic acid with PtO2 in 2N HCl followed by protection
of the resulting amine using benzyl chloroformate (CbzCl).7
However, upon scale-up, we encountered several problems
with isolation of the water-soluble acid 9 from the large
volume of aqueous solvent, thus requiring an exhaustive
extractive workup. We therefore modified this procedure so
the intermediate 3-(4-piperidyl)propionic acid HCl salt was
isolated from the reduction and then treated with N-
(benzyloxycarbonyloxy)succinimide (CbzOSu) and triethyl-
amine in EtOAc to eliminate the use of aqueous reaction
conditions. However, when this sequence was used on a 1.2
kg scale, reaction of 9 with CbzOSu followed by N-
Results and Discussion
Initially, we needed to prepare 1 kg of drug substance
for early toxicological and clinical studies in a very short
time frame. Therefore, we concentrated on preparation of
the required drug substance without optimization of reaction
parameters, using the process outlined in Scheme 3. R-Ethyl
nipecotate (8) was obtained from the commercially available
tartrate salt and reacted with acid 9 (prepared according to
a modified literature procedure).7 The resulting ester 10 was
saponified with LiOH to afford the corresponding crystalline
acid 11. Coupling of acid 11 with methyl (S)-3-amino-3-
(3′-pyridyl)propionate dihydrochloride (3) (prepared via
classical resolution of the corresponding 3-(tert-butoxycar-
(8) Boesch, H.; Cesco-Cancian, S.; Hecker, L. R.; Hoekstra, W. J.; Justus, M.;
Maryanoff, C. A.; Scott, L.; Shah, R. D.; Solms, G.; Sorgi, K. L.; Stefanick,
S. M.; Thurnheer, U.; Villani, F. J., Jr.; Walker, D. G. Org. Process Res.
DeV. 2001, 5, 23.
(6) Cohen, J. H.; Justus, M.; Maryanoff, C. A.; Rossler, A.; Schroder, F.; Sorgi,
K. L.; Villani, F. J., Jr.; Weh, C. U.S. Patent 6,515,130, B1, 2003.
(7) DeGraw, J. I.; Kennedy, J. G. J. Heterocycl. Chem. 1966, 3, 90.
Vol. 7, No. 6, 2003 / Organic Process Research & Development
•
867

Scheme 3 a
a
Reagents and conditions: (a) K2CO3; (b) DCC, MeCN, HOBT; (c) LiOH, aq THF; (d) Pd/C, H2, MeOH; (e) recrystallization (n-BuOH).
hydroxysuccinimide resulted in formation of an activated
ester, which during the long reaction time, was coupled to a
molecule of 3-(4-piperidyl)propionic acid, leading to byprod-
uct 14. To minimize formation of this difficult-to-remove
by-product, we stopped the reaction prior to completion and
isolated the acid 9 as a thick oil in approximately 60% overall
yield with only a trace amount of 14 present (Scheme 4).
Scheme 5
Scheme 4
Synthesis of Methyl (S)-3-Amino-3-(3′-pyridyl)propi-
onate 3. While the initial process8 to prepare â-amino ester
3 via classical resolution of the corresponding 3-[(tert-
butoxy)carbonyl]amino-3-(3′-pyridyl)propionic acid with
(-)-ephedrine afforded the desired material, we needed to
develop a more cost-efficient, large-scale process which
would not require an expensive protection/deprotection
sequence. We surveyed the literature and found several
methods reported for the synthesis of â-3-pyridyl alanine
esters;10-13 however, for reasons of cost, low overall yield,
low reaction temperature, or general patent issues, none of
these procedures would be useful as a large-scale process.
As part of our process research, we evaluated different
approaches for the preparation of 3 and thus developed
several methods, including both an asymmetric synthesis14
and a classical resolution. However, we found the more cost-
effective process was classical resolution of the correspond-
For large-scale production, we still needed to improve
this process to remove the expensive PtO2 catalyst as well
as increase the overall yield. After much investigation, we
found that reduction under neutral conditions using a Rh/
Al2O3 catalyst9 proved very effective (Scheme 5). This
allowed isolation of 3-(4-piperidyl)propionic acid (15) as a
crystalline solid in quantitative yield after precipitation from
acetonitrile. The acid was then suspended with Ca(OH)2 in
aqueous acetonitrile and treated with CbzCl to afford the
calcium salt 16 in 95% overall yield. This improved process
led to an increase in overall yield from 60 to 95%, used less
expensive reagents, and allowed isolation of the product as
a crystalline calcium salt.
(10) Bovy, P. R.; Rico, J. G.; Lindmark, R. J.; Rogers, T. E.; Tjoeng, F. S.;
Zablocki, J. A. U.S. Patent 5,254,573, 1993.
(11) Davis, F. A.; Reddy, R. T.; Reddy, R. E. J. Org. Chem. 1992, 57, 6387.
(12) Behling, J. R.; Boys, M. L.; Cain-Janicki, K. J.; Colson, P. J.; Doubleday,
W. W.; Duran, J. E.; Farid, P. N.; Knable, C. N.; Muellner, F. W.; Nugent,
S. T.; Topgi, R. S. U.S. Patent 5,840,961, 1998.
(13) Jiang, J.; Schumacher, K. K.; Joullie, M. M.; Davis, F. A.; Reddy, R. E.
Tetrahedron Lett. 1994, 35, 2121.
(14) Zhong, H. M.; Cohen, J. H.; Abdel-Magid, A. F.; Kenney, B. D.; Maryanoff,
C. A.; Shah, R. D.; Villani, F. J., Jr.; Zhang, F. Tetrahedron Lett. 1999, 40,
7721.
(9) Freifelder, M. U.S. Patent 3,159,639, 1964.
868
•
Vol. 7, No. 6, 2003 / Organic Process Research & Development

Scheme 6 a
a
Reagents and conditions: (a) NaOMe, toluene, methyl acetate, 65 °C, 13 h, extraction and azeotropic removal of water, then HCl 0-10 °C; (b) toluene, NaOAc,
MeOH, acetic acid, 65 °C, NH3, 4 h; (c) NaBH4, THF, HOAc, -5 °C, 5 h; MeOH, -5 °C, 30 min, followed by HCl, 5 °C; (d) MeCN, NEt3, 35 °C, 2 h, then
EtOH/H2O, (+)-tartaric acid, RT.
Scheme 7
ing racemic â-amino ester because the racemate was easily
prepared in high yield/purity and the enantiomerically pure
resolving agent, (+)-tartaric acid, was commercially available
and inexpensive (Scheme 6).
The synthesis of â-ketoester 17 was reported in the
literature by treatment of methyl nicotinate with a base
cost, (()-ethyl nipecotate was resolved using L-(+)-tartaric
(sodium hydride15 or sodium ethoxide16,17), followed by
acid as shown in Scheme 7. According to the procedure by
addition of methyl acetate. We found sodium methoxide was
Akkerman et al.,18 the resolution was carried out using 1
a better base for this step, leading to increased yield and
equiv of L-(+)-tartaric acid in EtOH. However, it required
purity in addition to being safer to use on large scale.
five recrystallizations from ethanol to obtain >99% de in
Isolation of the â-ketoester via precipitation of the hydro-
31% overall yield using approximately 45 L of solvent per
chloride salt from the toluene phase after extractive workup
kilogram of material. We developed a better alternative by
afforded 17 in 85% yield. Treatment with gaseous ammonia
carrying out the resolution in aqueous i-PrOH which provided
in a mixture of toluene and i-PrOH at elevated temperature
the (+)-tartrate salt in >99% de after only 1 slurry requiring
under acid catalysis afforded 18, which was crystallized
15.5 L of solvent per kilogram of material with an increased
directly from the reaction mixture in excellent yield (89%)
overall yield (36%).
and purity (99.5%). The reduction of enamine 18 to give
Synthesis of Elarofiban. With the preparation of the key
â-amino ester (()-3 was carried out using either catalytic
starting materials and intermediates in place, we completed
hydrogenation (Pd/C, AcOH) or borohydride reagents (NaBH4/
the synthesis of elarofiban as shown in Scheme 8. Typically,
AcOH/THF). However, it was necessary to carry out the
the formation of an amide bond is carried out by reacting a
hydrogenation under water-free conditions since wet catalysts
carboxylic acid with an amine (either in its free form or as
led to lower yields resulting from partial hydrolysis of the
a salt) in the presence of a coupling agent and base. When
enamine as was evident by the formation of the correspond-
the amine used is in the form of a carboxylate salt, a separate
ing â-hydroxy ester after reduction. Thus, it was more
step is added in which the amine salt is converted to a free
efficient to carry out the reduction using NaBH4 in the
amine prior to coupling, as the carboxylate anion may
presence of AcOH. After quenching the reaction with
interfere with the coupling, resulting in undesired side
methanol and subsequent treatment with gaseous HCl, the
products.19 However, for the preparation of elarofiban, we
racemic dihydrochloride salt (()-3 was obtained in 74%
were able to take advantage of the unique physical-chemical
overall yield and then easily converted to the free amine via
properties of calcium tartrate to eliminate this extra step.
treatment with triethylamine in acetonitrile prior to the
By mixing the calcium salt 16 with R-(-)-ethyl nipecotate
classical resolution. The optimal condition for the resolution
(+)-tartrate in the presence of Ca(OH)2 in calcium tartrate
in terms of yield and throughput was achieved using 0.25
precipitated from the aqueous THF reaction mixture and was
equiv of (+)-tartaric acid in a mixture of EtOH/water. This
easily removed via filtration. The filtrate, which contains no
precipitated the diastereomerically enriched hemi-tartrate salt,
tartrate anion, was then treated with DCC/HOBT to afford
which was then purified by slurrying in EtOH/water to afford
the coupled ester without racemization or formation of any
19 in 28% yield and >98% de.
by-products or racemization. The ester was not isolated but
Synthesis of R-(-)-Ethyl Nipecotate (+)-Tartrate.
converted directly to acid 11 upon hydrolysis with LiOH
Although R-ethyl nipecotate was commercially available as
and acidification in excellent yield (92%) and purity (>98%
the (+)-tartrate salt, it was very expensive. Thus, to minimize
(18) Akkerman, A. M.; de Jongh, D. K.; Veldstra, H. Recl. TraV. Chim. Pays-
(15) Wenkert, E.; Orito, K.; Simmons, D. P. J. Org. Chem. 1983, 48, 5006.
(16) Strong, F. M.; McElvain, S. M. J. Am. Chem. Soc. 1933, 55, 816.
(17) Stein, M. L.; Burger, A. J. Am. Chem. Soc. 1957, 79, 154.
Bas 1951, 70, 899.
(19) Bodansky, M. Principles of Peptide Synthesis, 2nd ed.; Springer-Verlag:
Berlin, Germany, 1993.
Vol. 7, No. 6, 2003 / Organic Process Research & Development
•
869

Scheme 8 a
a
Reagents and conditions: (a) Ca(OH)2, DCC, HOBT, aq THF; (b) LiOH, aq THF; (c) phosphate buffer, aq THF, Ca(OH)2, DCC 0-5 °C; (d) Pd/C, H2, MeOH,
2 h; (e) n-BuOH.
ee). Acid 11 was then reacted with the â-amino ester hemi-
tartrate salt 19 using similar reaction conditions20 as the
previous coupling step to yield fully protected elarofiban (12).
Finally, the ester was hydrolyzed with aqueous LiOH, the
Cbz group was removed via catalytic hydrogenation, and the
crude product was recrystallized from n-BuOH to afford pure
elarofiban in 58% overall yield for the four steps.
3-(4-Piperidyl)propionic Acid (15). A suspension of
3-(4-pyridine)acrylic acid (18 kg, 2.7 mol) in 75 kg of water
was neutralized (pH 7.5) with 25% aq ammonia (6.8 kg). A
slurry of Rh/Al2O3 (0.9 kg) in 5 kg of water was added, and
the reaction mixture was hydrogenated under 3-3.5 bar of
pressure at 85-95 °C. When no further change in pressure
was observed (∼8 h), the mixture was cooled to 25-35 °C.
The catalyst was filtered and washed with 4.0 kg of water.
The filtrate was concentrated under vacuum at 80-90 °C,
and the product began to precipitate. Acetonitrile (116 kg)
was added, and then the mixture was again concentrated (ca.
50%) under vacuum to remove residual ammonia. Additional
acetonitrile (57.1 kg) was added to aid in crystallization, and
the mixture was stirred for 1-4 h at ambient temperature.
The product was isolated via filtration and dried under
vacuum at 45-55 °C to afford 19.1 kg (100%) of 15, which
was identical to the literature.9
(R)-(-)-Ethyl Nipecotate (+)-Tartrate (20). L-(+)-
Tartaric acid (47.74 g, 318 mmol) was suspended in 265 g
of i-PrOH 16.91 g of water and heated to 60-65 °C to afford
a homogeneous solution. The mixture was treated with (()-
ethyl nipecotate (50 g, 318 mmol) and heated to 70-75 °C
for 20-30 min. The mixture was then cooled to 60 °C, and
seed crystals of (R)-(-)-ethyl nipecotate-L-(+)-tartrate (25
mg, 0.08 mmol) were added. Upon cooling to ambient
temperature, a white solid precipitated which was isolated
(62.0 g, 94.8% de) and washed twice with a mixture of
i-PrOH (21.05 g) and water (1.34 g). The crude product was
slurried in a mixture of i-PrOH (188 g) and water (12 g),
heated to 73-77 °C for 10-20 min, and then cooled to
ambient temperature. The product was isolated by filtration
and washed twice with a mixture of i-PrOH (21.05 g) and
water (1.34 g) to afford 35.2 g (36%) of 20 as a white
powder, which was identical to the literature:18 mp 155-
156 °C. By HPLC analysis this material was greater than
98.8% de.21
Conclusion
In summary, we have developed a very efficient com-
mercial scale process to prepare elarofiban which offers
several advantages over the original Drug Discovery syn-
thesis. This new synthetic strategy avoids the use of
chromatography for isolation/purification by using crystalline
starting materials and intermediates, reduces the overall
number of steps by introducing a novel procedure for free
amine liberation from the corresponding tartrate salts em-
ploying Ca(OH)2 followed by direct coupling with carboxylic
acids under mild conditions to produce the amide products
with no racemization or formation of by-products, eliminates
all hazardous solvents, reduces the overall cost by using less
expensive reagents, and improves the overall yield.
Experimental Section
General Procedures. Melting points were determined
with a Thomas-Hoover capillary melting point apparatus and
are uncorrected. 1H NMR spectra were recorded on a Bruker
AM 300, 300 MHz spectrometer. The chemical shifts are
expressed as δ units with Me4Si as the internal standard
(multiplicities in 1H NMR are referred to as: s for singlets,
d for doublets, t for triplets, q for quartets, and m for
multiplets). All reactions were carried out under a nitrogen
or an argon atmosphere. Solvents and reagents were obtained
from commercial sources and used without further treatment
or purification.
(20) To obtain a high yield for this coupling reaction, the pH needs to be
maintained between 6.5 and 7.0. Initially this was done via constant addition
of either base or acid throughout the reaction. However, to simplify this
process, a buffer system was developed which maintained the desired pH
and thus afforded 12 in high yield and purity.
(R)-1-[3-(1-Benzyoxycarbonyl-4-piperidyl)-propionyl]-
3-piperidinecarboxylic Acid (11). Compound 15 (5.3 g,
(21) Rustum, A. M. J. Chromatogr., A 1995, 696, 75.
870
•
Vol. 7, No. 6, 2003 / Organic Process Research & Development

33.9 mmol) and calcium hydroxide (4.0 g, 53.7 mmol) were
suspended in 16 g water/65 g acetonitrile and then cooled
to 0-10 °C. Benzyl chloroformate (6.8 g, 39.6 mmol) was
added within 30 min, and the reaction was stirred at 0-5
°C for 2 h. The product precipitated during the reaction and
was isolated (as a mixture with calcium hydroxide and
calcium chloride) by filtration to afford 3-(N-benzyloxycar-
bonyl-4-piperidyl)propionic acid calcium salt (16). Crude 16
(21.9 g, 32.2 mmol), 20 (21.7 g, 70.8 mmol), and hydroxy
benzyltriazole (HOBT) (1.30 g, 9.65 mmol) were suspended
in water (40 g)/THF (80 g), and the resulting suspension
was adjusted to pH 7 with Ca(OH)2. The precipitated calcium
tartrate was collected by filtration and washed with 10 g of
THF. The filtrate was cooled to 0-5 °C and treated with a
solution of DCC (19.9 g, 96.5 mmol) in 40 g of THF. The
mixture was warmed slowly to ambient temperature, and
N,N′-dicyclohexylurea (DCU) precipitated. After 4 h, the
DCU was removed by filtration and washed with 8 g of THF.
The filtrate was again cooled to 0-5 °C and treated with a
solution of lithium hydroxide (6.67 g, 159.0 mmol) in 60.38
g of water. The resulting pale yellow solution was warmed
to ambient temperature. After 3 h, ethyl acetate (45.4 g) was
added and the pH was adjusted to 4.0 with ca. 18.6 g of
concentrated HCl. DCU precipitated and was filtered from
the mixture. The aqueous layer was separated and washed
twice with 31.8 g of ethyl acetate. The combined organic
layers were washed twice with 50 mL of saturated brine,
the ethyl acetate layer was separated, and the solvent was
removed by distillation under vacuum at or below 55 °C.
tert-Butyl methyl ether (MTBE) (70.8 g) was added, and
the suspension was stirred for 30 min at 45-50 °C, then
cooled to ambient temperature and stirred for 1 h until
crystallization was complete. The product was filtered,
washed with 6.3 g of MTBE, and then dried under vacuum
at 40-50 °C to afford 11.9 g (87%) of 11: mp 134-135
Ja ) 1.5 Hz, Jb ) 4.6 Hz), 8.82 (1H, d, J ) 2.3 Hz). MS
(ESI) m/z: 179 (MH+).
Methyl 3-Amino-3-(3-pyridyl)propanoate Dihydro-
chloride (()-3. Procedure A. Glacial acetic acid (526.9 g,
8.78 mol) was added dropwise at -5 °C to a suspension of
18 (0.45 mol) and sodium borohydride (44.3 g, 1.17 mol) in
THF (500 g), and the resulting reaction mixture was stirred
at -5 to 0 °C. After 5 h, methanol (600 g) was added
dropwise to the solution followed by gaseous HCl (163 g,
4.47 mol) after an additional 0.5 h. After 8 h, the white
precipitate was collected by filtration and dried at 40 °C to
yield 101.6 g (89%) of (()-3 as a white crystalline solid.
Procedure B. Dry palladium on charcoal (0.54 g,
manufactured by Degussa, 5% Pd/C) was added to a solution
of 18 (5.4 g, 30 mmol) in dry acetic acid (13 g) in a 450-
mL Pyrex high-pressure bottle and hydrogenated at 3-3.2
bar. After 1.5-2 h, the catalyst was filtered and washed with
20 g of i-PrOH until the wash solvent was no longer yellow.
Gaseous HCl (10.6 g, 0.3 mol) was bubbled through the
stirred filtrate at 5-15 °C, and the resulting suspension was
cooled to 0-5 °C for 2 h. The product was isolated by
filtration, washed with 5 g of i-PrOH, and dried at 45 °C to
yield 5.95 g (78.4%,) of (()-3 as a white crystalline solid:
mp 187.5-189 °C. 1H NMR (D2O) δ (ppm): 3.19 (1H, dd),
3.28 (1H, dd), 3.62 (3H, s), 5.05 (1H, t), 8.08 (1H, m),
8.66 (1H, m), 8.80 (1H, m), 8.91 (1H, s). MS (ESI) m/z:
181 (MH+).
Methyl (S)-3-Amino-3-(3-pyridyl)propanoate Hemi-
tartrate (19). (()-3 (150 g, 0.563 mol) was suspended in
acetonitrile (425 g) and treated with triethylamine (125.3 g,
1.239 mol) at 35 °C or less. The reaction was stirred for 2
h at 20 °C, then cooled to 5 °C. After 0.5 h, the resulting
triethylamine hydrochloride was removed via filtration and
washed with 50 g of acetonitrile. The filtrate was evaporated
to dryness under vacuum to afford crude methyl 3-amino-
3-(3-pyridyl)propanoate. The free base (ca. 105 g) was
dissolved in 80 g of ethanol and treated with a solution of
(+)-tartaric acid (21.1 g, 0.141 mol) in 80 g EtOH/5 g water.
The mixture was stirred for 4 h at ambient temperature and
then cooled slowly to 10-15 °C for an additional 2 h. The
precipitate was collected via filtration and washed with 30
g of ethanol. The crude hemi-tartrate salt was slurried at 35-
40 °C for 2 h in a mixture of 150 g of ethanol and 4.6 g of
water and then cooled to ambient temperature. The resulting
precipitate was isolated via filtration, washed with 30 g of
ethanol, and dried under vacuum to afford 35.5 g (28%) of
1
°C. H NMR (DMSO) δ (ppm): 0.9-1.1 (2H, m), 1.31-
1.75 (8H, m), 1.86-2.00 (1H, m), 2.20-2.50 (3H, m), 2.65-
2.85 (2H, m), 2.94-3.05 (1H, m), 3.25-3.35 (1H, m), 3.70-
3.84 (1H, m), 4.34-4.43 (1H, m), 3.95-4.02 (2H, m), 5.05
(2H, s), 7.27-7.40 (5H, m), 12.40 (1H, s). MS (ESI) m/z:
403 (MH+).
Methyl 3-Amino-3-(3-pyridyl)-2-propenoate (18). Meth-
yl nicotinoyl acetate (88 g, 0.5 mol) was dissolved in toluene
(200 g), i-PrOH (200 g), and formic acid (98-100%, 1.22
g, 0.03 mol) and heated to 60-65 °C. Gaseous ammonia
(23 g, 1.35 mol) was bubbled through the solution for 15
min, and the resulting white suspension was stirred at 65 °C
until a homogeneous solution formed. The solution was
stirred for 2 h at 65 °C and then was concentrated (ca. 200
g) at 65 °C. The residue was cooled to -5 °C with stirring,
and methyl 3-amino-3-(3-pyridyl)-2-propenoate crystallized
as colorless needles. The process of reducing the volume to
50% followed by cooling was repeated three times with the
mother liquors. Filtration, washing with toluene, and drying
at 30 °C resulted in 77.74 g (88.8%) of 18 as colorless
crystals: mp 118-120 °C. 1H NMR (DMSO) δ (ppm): 3.61
(3H, s), 4.87 (1H, s), 7.50 (1H, dd, Ja ) 4.6 Hz, Jb ) 8.0
Hz), 8.01 (1H, dt, Ja ) 8.0 Hz, Jb ) 4.6 Hz), 8.68 (1H, dd,
1
19 as a white crystalline solid: mp 139-141 °C. H NMR
(DMSO) δ (ppm): 2.90 (2H, m), 3.55 (3H, s), 3.97 (1H, s),
4.45 (1H, t, J ) 8.0 Hz), 7.39 (1H, dd, Ja ) 4.6 Hz, Jb )
8.0 Hz), 7.89 (1H, dt, Ja ) 8.0 Hz, Jb ) 3.8 Hz), 8.49 (1H,
dd, Ja ) 1.5 Hz, Jb ) 4.6 Hz), 8.62 (1H, d, J ) 2.3 Hz). MS
(ESI) m/z: 181 (MH+ free base).
[S-(R*,S*)]-â-[[[1-[1-Oxo-3-(4-piperidinyl)propyl]-3-pi-
peridinyl]carbonyl]amino]-3-pyridine Propanoic Acid
(Elarofiban). 11 (60 kg, 149 mol), 19 (41.8 kg, 164 mol),
and HOBT (1.98 kg, 14.8 mol) were suspended in a cold
solution (0-5 °C) of KH2PO4 (7.4 kg, 42.6 mol) and
Na2HPO4 (4.3 kg, 30 mol) in water (95 kg)/THF (55 kg).
Vol. 7, No. 6, 2003 / Organic Process Research & Development
•
871

The pH was adjusted to 6.0-6.4 using calcium hydroxide,
and the resulting suspension was cooled to 0-5 °C. The
reaction mixture was treated with a solution of DCC (37.2
kg, 180.3 mol) in THF (110 kg) and stirred for 1 h at 0-5
°C, then warmed to ambient temperature. After 4 h, the
suspension was cooled to 0-5 °C and ethyl acetate (2 kg)
was added. The precipitate (a mixture of DCU and calcium
tartrate) was removed via filtration and washed with pre-
cooled THF (60 kg). The organic layer was separated,
washed with 5% NaHCO3 (50 kg), and then concentrated at
40-50 °C. The residual oil was dissolved in THF (50 kg)
and evaporated to dryness to afford 12 as an oil. Crude 12
was dissolved in 163 kg of THF at 45 °C, cooled to 0-5
°C, and treated with a solution of lithium hydroxide
monohydrate (14.3 kg, 340.8 mol) in 151 kg of water. The
resulting pale yellow solution was stirred for 2 h at ambient
temperature, and then the pH was adjusted to 4.1 with HCl
(36-38%, 38 kg). NaCl (7.2 kg) was added, and the layers
were separated. The organic layer was washed twice with a
solution of 36.4 kg of NaCl in 72.6 kg of water. The organic
layer was concentrated under vacuum, and the resulting oil
was dissolved in 75 kg of THF. (This process was repeated
until a water content of <2% was achieved.) The precipitated
inorganic salts were removed and washed with 9 kg of THF.
The filtrate was evaporated under vacuum at 45 °C to afford
13 as an oil. Crude 13 was dissolved in 312 kg of methanol,
treated with a suspension of 60 kg of methanol and 15 kg of
Pd/C (wet), and then hydrogenated under pressure (2-3 bar)
with stirring at 38-42 °C. When the hydrogenation was
finished, the catalyst was filtered through Hyflo SuperCel
and washed with 39 kg of methanol. The filtrate was
concentrated under reduced pressure at 40-50 °C, and the
resulting crude oil was dissolved in 756 kg of n-butyl alcohol
and heated to 75-85 °C for 15-20 min, then cooled to 20-
30 °C. tert-Butylamine (0.7 kg) was added, and the mixture
was cooled to 0-5 °C for an additional hour. The precipitate
was isolated, washed with 102 kg of MTBE, and dried under
vacuum at 60-80 °C to yield 36 kg (58%) of elarofiban as
a white crystalline solid: mp 157-159 °C. 1H NMR
(DMSO-d6) δ (ppm): 1.2-2.0 (m, 12 H), 2.3 (m, 3 H), 2.5
(m, 1 H), 2.8 (m, 5 H), 3.2 (d, J ) 8 Hz, 2 H), 3.8 (m, 2 H),
4.2 (m, 2 H), 5.2 (m, 1 H), 7.8 (t, J ) 4 Hz, 1 H), 8.3 (t, J
) 4 Hz, 1 H), 8.5 (m, 1 H), 8.7 (m, 1 H), 8.9 (m, 2 H); MS
m/e: 417 (MH+).
Received for review July 25, 2003.
OP034103O
872
•
Vol. 7, No. 6, 2003 / Organic Process Research & Development