An improved synthesis of etravirine

Article abstract of DOI:10.1021/op9003289

Etravirine (1) is a novel diarylpyrimidine non-nucleoside reverse transcriptase inhibitor and has recently been approved by the U.S. Federal Drug Administsration for the treatment of AIDS. Its reported synthesis is fraught with many difficulties, the foremost being the poor yield and long reaction time required at the aminolysis stage. We attributed this problem to the presence of a bromide group adjacent to the reaction site of the advance intermediate (6). In order to circumvent this issue, we proposed to defer the installation of the bromide group at a later stage, preferably after aminolysis. Indeed, this protocol has worked well. However, in the process of installation of diarylether and diarylamine functionalities at appropriate positions, we had to reverse the sequence of displacement reactions of the dichloride intermediate (9) with 3,5-dimethyl-4-hydroxybenzonitrile (5) and 4-aminobenzonitrile (3). The classical bromination led to the completion of etravirine synthesis.

Full text of DOI:10.1021/op9003289

Organic Process Research & Development 2010, 14, 657–660
An Improved Synthesis of Etravirine
Shashikant Joshi,* Golak C. Maikap,* Shyam Titirmare, Ashok Chaudhari, and Mukund K. Gurjar
Emcure Pharmaceuticals Ltd., R&D Centre, 12/2, F-II Block, M.I.D.C. Pimpri, Pune - 411018, Maharashtra, India
Abstract:
initially reacted with 4-aminobenzonitrile (3) in the presence
of diisopropylethylamine in refluxing dioxane to give the
diarylamine derivative (4) whose reaction with 4-hydroxy-3,5-
dimethylbenzonitrile (5) in the presence of NaH in N-meth-
ylpyrrolidone solvent provided the intermediate (2) (Scheme
1).
In another synthesis of etravirine (1),1,6,7 4-guanidinoben-
zonitrile (7) was cyclized with diethylmalonate in the presence
of sodium ethoxide in ethanol to give 4-(4,6-dihydroxypyrimi-
dine-2-yl-amino)benzonitrile (8) which was subsequently treated
with POCl3 to form the corresponding dichloro derivative 9
(86%). The bromination of 9 with bromine and sodium
bicarbonate in aqueous methanol afforded 4-(5-bromo-4,6-
dichloropyrimidin-2-yl-amine)benzonitrile (10) which, when
condensed with a sodium salt of cyano-2,6-dimethylphenolate
(11) in the presence of N-methylpyrrolidone and dioxane, gave
the intermediate (6). Aminolysis of 6 with ammonia in
isopropanol solvent gave etravirine (1) with ∼10% overall yield
(Scheme 2).
Etravirine (1) is a novel diarylpyrimidine non-nucleoside reverse
transcriptase inhibitor and has recently been approved by the U.S.
Federal Drug Administsration for the treatment of AIDS. Its
reported synthesis is fraught with many difficulties, the foremost
being the poor yield and long reaction time required at the
aminolysis stage. We attributed this problem to the presence of a
bromide group adjacent to the reaction site of the advance
intermediate (6). In order to circumvent this issue, we proposed
to defer the installation of the bromide group at a later stage,
preferably after aminolysis. Indeed, this protocol has worked well.
However, in the process of installation of diarylether and diary-
lamine functionalities at appropriate positions, we had to reverse
the sequence of displacement reactions of the dichloride intermedi-
ate (9) with 3,5-dimethyl-4-hydroxybenzonitrile (5) and 4-ami-
nobenzonitrile (3). The classical bromination led to the completion
of etravirine synthesis.
Introduction
The critical step of aminolysis of 6 under ammonia at high
pressure was not satisfactory because of low yield (41%) and
long reaction time of approximately 4 days. The low yield and
long reaction period could be attributed to the presence of the
bromine atom situated adjacent to the amination site. We argued
that this issue could be resolved if we avoided the placement
of bromine in 6 prior to the aminolysis reaction. Once the amino
group is in place, then bromine could be installed by many
conventional methods which are at our disposal.
The non-nucleoside reverse transcriptase inhibitors (NNRTI)
play a pivotal role as key components of highly active
antiretroviral therapy.1-4 NNRTIs have an ability to target an
allosteric binding pocket on the reverse transcriptase (RT)
enzyme. This property allows them to be useful as broad-
spectrum agents against human immunodeficiency virus (HIV)
RT mutations. The diarylpyrimidine-based NNRTIs constitute
the second-generation drugs and are useful for treatment of HIV-
infected patients with NNRTI-resistant viruses. Etravirine (TMC
125),2 4-[[6-amino-5-bromo-2-[(4-cyanophenyl)amino]-4-pyri-
midinyl]oxy]-3,5-dimethylbenzonitrile, was approved in 2008
by the U.S. Federal Drug Administration for use in combination
therapy along with other antiretroviral agents in adult patients
with multidrug-resistant HIV infections.
Results and Discussion
In accordance with our proposed strategy, the reaction of
the trichloride 12 with 4-aminobenzonitrile (3) in the presence
of diisopropylethylamine (DIEA) was performed. The expected
product 13 was carefully analyzed with respect to the position
of aniline placement. The NOE studies of the derived product
(1) Ludovici, D.W.; Kukla, M. J.; Grous, P. G.; Krishnan, S.; Andries, K.;
de Be´thune, M. P.; Azijn, H.; Pauwels, R.; De Clercq, E.; Arnold, E.;
Janssen, P. A. J. Bioorg. Med. Chem. Lett. 2001, 11, 2235–2239.
(2) Chem. Eng. News 2008, 86, (33), 29-36.
(3) Bernard, E. J. Etravirine (TMC125, Intelence) granted accelerated approval
in US. Aidsmap News; January 21, 2008; http://www.aidsmap.com/en/
news/46E8FC38-3BB8-4DDD-9FEC-D6BFD5DD69AA.asp.
(4) Guillemont, J.; Pasquier, E.; Palandjian, P.; Vernier, D.; Gaurrand, S.;
Lewi, P. J.; Heeres, J.; de Jonge, M. R.; Koymans, L., M. H.; Daeyaert,
F. F. D.; Vinkers, M. H.; Arnold, E.; Das, K.; Pauwels, R.; Andries,
K.; de Be´thune, M.-P.; Bettens, E.; Hertogs, K.; Wigerinck, P.;
Timmerman, P.; Janssen, P. A. J. J. Med. Chem. 2005, 48, 2072–79.
(5) De Corte, B.; De Jonge, M. R.; Heeres, J.; Ho, Y.; Janssen, P. A. J.;
Kavash, R. W.; Koymans, L. M. H.; Kukla, M. J.; Ludovici, D. W.;
Van Aken, K. J. A. (Janseen Pharmaceutica NV). U.S. Pat. 6,878,717,
2005.
The procedure to prepare etravirine (1) is disclosed in
Scheme 15,6 in which 5-bromo-2,4,6-trichloropyrimidine (2) was
(6) Davies, S. L.; Castaner, J.; Silvestre, J. S.; Bayes, M. Drugs Future
2005, 30 (5), 462–468.
* Corresponding authors. Telephone: 020-30613000. Fax: 020-30613056.
E-mail: golak.maikap@emcure.co.in.
(7) Bauer, V. J.; Safir, S. R. J. Med. Chem. 1966, 9, 244–246.
10.1021/op9003289  2010 American Chemical Society
Published on Web 04/21/2010
Vol. 14, No. 3, 2010 / Organic Process Research & Development
•
657

Scheme 1
Scheme 2
Scheme 3
tion reactions (Scheme 4). Compound 12 was first reacted with
phenol derivative 5 under similar conditions to give the biaryl
ether derivative (15). The structure was scrupulously analyzed
by spectroscopic methods. Conversion of 12 to 15 was
optimized in the laboratory w.r.t. solvent ratio, reaction tem-
perature, molar ratio of compound 5, and isolation temperature.
The second substitution reaction with aniline derivative (3) using
potassium tert-butoxide as a base and N-methylpyrrolidone as
a solvent gave 16 with a yield of 58%. The major byproduct
formed in this step was the isomer 16a as shown in Scheme 4.
Experiments were carried out in the laboratory to avoid the
formation of 16a by lowering the reaction temperature to -10
to -15 °C, using different solvents such as DMSO, DMF, bases
such as sodium hydroxide, potassium hydroxide, potassium
carbonate, sodium methoxide, diisopropyl ethyl amine, pyridine,
and triethylamine. The aminolysis of 16 went smoothly at
120-125 °C in 8-12 h to give the amine (17) in good yield
(82%). Compound 16 contained about 2-3% of 16a prior to
aminolysis which was also converted to the corresponding
amino compound during aminolysis. Aminolysis compound of
16a was removed during isolation of 17 by crystallization from
clearly indicated that 13 was not formed. However, NOE studies
revealed the structure 14 to be the product (Scheme 3). Other
spectroscopic studies substantiated the structure of 14. This
observation was contrary to the reported synthesis of etravirine
(Scheme 1). The difference between the two precursors 2 and
12 was the presence of a bromide group in structure 2. It became
obvious for us to reverse the sequence of nucleophilic substitu-
658
•
Vol. 14, No. 3, 2010 / Organic Process Research & Development

Scheme 4
1,4-dioxane. Since the aminolysis product of 16a is absent in
17, the formation of the bromo isomer during bromination is
ruled out. Aminolysis reaction condition was optimized by using
different solvents such as methanol, water, and 1,4-dioxane.
Reactions were also carried out at 100 and 140 °C. Finally,
bromination of 17 was conducted in the presence of liquid
bromine in CH2Cl2 at 0-5 °C to give etravirine (1) in 85%
yield. For bromination of 17, we also tried solvents such as
acetic acid or propionic acid, but better results were obtained
in methyl dichloride. The spectral and analytical data of 1 were
in agreement with reported data. The results were reproduced
by scaling up the process from 50-g to 1.2-kg scale.
Calcd for C13H9Cl2ON3: C, 53.06; H, 3.06; N, 14.28. Found:
C, 53.09; H, 3.07; N, 14.31%.
4-[[6-Chloro-2-[(4-cyanophenyl)amino]-4-pyrimidinyl]oxy]-
3,5-dimethylbenzonitrile (16). To a solution of compound 15
(1500 g, 5.1 mol) and 4-aminobenzonitrile (602.0 g, 5.1 mol)
in N-methylpyrrolidone (7.5 L) at 0-5 °C was added potassium
tert-butoxide (1144.0 g, 10.2 mol) over a period of 2 h. The
reaction was allowed to attain RT followed by careful addition
of cold water (30 L). The reaction mixture was filtered; the
residue was suspended in water (15.0 lit.) and acidified to pH
6-7 using conc. HCl. The reaction mixture was filtered and
washed with 1.5 L of water. Wet cake was dried at 60-65 °C
until water content was <1.0%. Crude product containing about
20-25% 16a, was purified by ethyl acetate treatments (2 ×
4.5 L) at 65-70 °C followed by filtration at 10-15 °C and
washing the cake with 1.5 L of chilled ethyl acetate. The wet
cake was finally dried at 55-60 °C under vacuum (600-700
mm of Hg) to give 16 (1112 g, 58% yield, >95% purity and
16a 2-3% by HPLC); mp 278.5-280.5 °C, IR (KBr): 3352,
2229, 1522, 1458, 1418, 1365, 1331, 1290, 1223, 1195, 1137,
Experimental Section
1
The H NMR spectra were recorded in DMSO-d6 on a
Varian 400 MHz; the chemical shifts were reported in δ ppm
relative to TMS. The FT-IR spectra were recorded in the solid
state KBr dispersion using Perkins Elmer FT-IR spectropho-
tometer. The mass spectra were recorded on Applied Biosystems
spectrometer. The melting points were determined by using
Buchi apparatus .The solvents and reagents were used without
purification. N-Methylpyrrolidone with moisture content less
than 0.1% was used in the process.
1
1108 cm-1; MS(m/z): 376, 378; H NMR (DMSO-d6, 400
MHz): δ 2.12 (s, 6 H), 6.92 (s, 1 H), 7.44 -7.50 (br s, 4 H),
7.78 (s, 2 H), 10.55 (br s, 1 H); 13C NMR (DMSO-d6, 100
MHz): 15.9, 98.1, 103.8, 109.1, 118.8, 119.5, 132.9, 133.0,
143.7, 153.4, 158.4, 161.6, 169.2. Anal. Calcd for C20H14ClON5:
C, 63.91; H, 3.72; N, 18.64. Found: C, 63.99; H, 3.64; N,
18.56%.
4-[(2,6-Dichloro)-4-pyrimidinyloxy]-3,5-dimethylbenzoni-
trile (15). 2,4,6-Trichloropyrimidine (1200 g, 6.54 mol), 3,5,-
dimethyl-4-hydroxybenzonitrile (961.0 g, 6.54 mol), N,N-
diisopropylethylamine (1014 g, 7.85 mol), and 1,4-dioxane (3.6
L) were heated at 70 °C for 2 h. The reaction mixture was
brought to 10-15 °C and filtered, and the residue was washed
with 600 mL of chilled 1,4-dioxane. The wet cake was treated
with 3.6 L of water at 25-30 °C, filtered, and dried at 55-60
°C under vacuum (600-700 mm of Hg) to give 15 (1539 g,
80% yield with >98% purity by HPLC); mp: 208-210 °C; IR
(KBr): 2227, 1559, 1545, 1476, 1406, 1338, 1304, 1264, 1189,
4-[[6-Amino-2-[(4-cyanophenyl)amino]-4-pyrimidinyl]oxy]-
3,5-dimethylbenzonitrile (17). A mixture of 16 (1100 g, 2.93
mol), 25% aq ammonia (6.6 L), and 1,4-dioxane (11.0 L) was
heated in an autoclave at 120 °C for 12 h. To the reaction
mixture, was added 1.65 L of water slowly at 50-60 °C, cooled
to 5-10 °C, and filtered. The residue was washed with 275
mL of chilled 1,4-dioxane and dried at 55-60 °C to give 17
(860 g, 82% yield, >98% purity by HPLC); mp 284.5-287
°C, IR (KBr): 3505, 3366, 2225, 1611, 1532, 1507, 1456, 1408,
1
1134, 1106, 980 cm-1; MS (m/z): 394, 396, 398; H NMR
(DMSO-d6, 400 MHz): δ 2.11 (s, 6 H), 7.60 (s, 1 H), 7.73 (s,
2 H); 13C NMR (DMSO-d6, 100 MHz): 15.6, 106.8, 109.3,
116.8, 118.3, 132.3, 132.9, 152.0, 158.5, 162.1, 169.3. Anal.
1
1375, 1244, 1206, 1176 cm-1; MS (m/z): 357; H NMR
(DMSO-d6, 400 MHz): δ 2.12 (s, 6 H), 5.45 (s, 1 H), 6.78 (br
Vol. 14, No. 3, 2010 / Organic Process Research & Development
•
659

s, 2 H), 7.46 (d, 2 H), 7.64 (d, 2 H), 7.72 (s, 2 H), 9.56 (s, 1H):
13C NMR (DMSO-d6, 100 MHz): 15.8, 78.6, 101.6, 108.0,
117.9, 118.7, 119.6, 132.4, 132.5, 132.9, 145.3, 153.9, 158.9,
166.4, 168.1. Anal. Calcd for C20H16ON6: C, 67.41; H, 4.49;
N, 23.59. Found: C, 67.39; H, 4.43; N, 23.51%.
Etravirine (1). To a cooled solution of 17 (850 g, 2.387
mol) in DCM (6.8 L) at 0-5 °C was added bromine solution
(401 g, 2.5 mol in 1.7 L of DCM). The reaction was stirred at
this temperature for 4 h, diluted with water (6.8 L), and basified
with 4 M NaOH solution at pH 9-10. At this point, sodium
metabisulphite solution (42.5 g in 170 mL water) was added.
The pH of the reaction was maintained between 7.5-8.5 over
a period of 1 h by adding 4 M NaOH solution. The solid was
filtered, washed with water (12 L), and dried at 55-60 °C
temperature to get crude etravirine.
The crude product was dissolved in 15.9 L of acetone at
50-55 °C and treated with 11 g of activated charcoal. After
charcoal clarification, 12.5 L of acetone was distilled out, and
the residue was cooled to 5-10 °C and filtered. Wet cake was
washed with 1.0 L of chilled acetone and finally dried at 55-60
°C under vacuum (600-700 mm of Hg) to give 1 (836 g, 80%
yield with 99.80% purity by HPLC); mp 255-257 °C (lit.5 mp
255-256 °C); IR (KBr): 3484, 3380, 3349, 2223, 1523, 1505,
1454, 1403, 1313, 1243, 1200, 1174, 1139, 1059, 1005 cm-1;
MS (m/z): 435, 437; 1H NMR (DMSO-d6, 400 MHz): δ 2.12
(s, 6 H),7.11 (br s, 2 H),7.42 (d, 2 H), 7.54 (d, 2 H), 7.74 (s, 2
H), 9.57 (br s, 1 H); 13C NMR (DMSO-d6, 100 MHz): 15.9,
74.6, 102.4, 108.6, 118.4, 119.0, 119.8, 132.7, 133.1, 144.9,
154.4, 156.8, 162.7, 163.75.
2,6-Dichloro,4-[N-(4′-cyanophenyl)amino]pyrimidine (14).
2,4,6-Trichloropyrimidine (5.0 g, 27.25 mmol), 4-aminoben-
zonitrile (2.66 g, 22.5 mmol), N,N-diisopropylethylamine (5.26
g), and isopropanol (25 mL) were stirred at RT for 24 h, filtered,
and washed with water and cold IPA to give 14, (3.2 g, 45.0%);
mp 293-295 °C; IR (KBr): 3316, 3212, 3112, 2230, 1633,
1558, 1510, 1449, 1274, 1240, 1180, 1120 cm-1; MS (m/z):
1
263, 265, 267; H NMR (DMSO-d6, 400 MHz): δ 6.89 (s, 1
H), 7.78 (d, 2 H), 7.84 (d, 2 H), 10.66 (b s, 1 H); 13C NMR
(DMSO-d6, 100 MHz): 105.1, 105.3, 118.9, 119.9, 133.3, 142.5,
158.5, 161.6. Anal. Calcd for C11H6Cl2N4: C, 49.81; H, 2.26;
N, 21.13. Found: C, 49.82; H, 2.28; N, 20.92%.
Conclusion
In conclusion, we have successfully developed a simple,
cost-effective, and industrially scalable process for the synthesis
of etravirine.
Acknowledgment
We thank Mr. Samit Mehta, Emcure Pharmaceuticals Ltd.,
for his encouragement and support.
Supporting Information Available
Mass, IR, 1H and 13C NMR spectra, and HPLC chromato-
grams. This material is available free of charge via the Internet
at http://pubs.acs.org.
Received for review December 17, 2009.
OP9003289
660
•
Vol. 14, No. 3, 2010 / Organic Process Research & Development