Synthesis of the anti-prostate cancer drug abiraterone acetate

Article abstract of DOI:10.3987/COM-17-13839

Abiraterone acetate is used for the treatment of castration-resistant prostate cancer. Abiraterone acetate was synthesized from dehydroepiandrosterone via a three-step reaction including the formation of tosylhydrazone, cross-coupling reaction and acetylation, and a two-step purification including column chromatography and recrystallization with an overall yield of 51.9%. Here, an improved procedure for the preparation of abiraterone acetate is described. This synthetic process is of easy operation and low cost, which is suitable for industrialization.

Full text of DOI:10.3987/COM-17-13839

HETEROCYCLES, Vol. 96, No. 3, 2018
461
HETEROCYCLES, Vol. 96, No. 3, 2018, pp. 461 - 469. © 2018 The Japan Institute of Heterocyclic Chemistry
Received, 16th November, 2017, Accepted, 5th February, 2018, Published online, 15th February, 2018
DOI: 10.3987/COM-17-13839
SYNTHESIS
OF
THE
ANTI-PROSTATE
CANCER
DRUG
ABIRATERONE ACETATE
Siyue Ma,a Jianheng Li,a* Huayang Tang,a and Feng Xub
aSchool of Pharmacy and Key laboratory of Pharmaceutical Quality Control of
Hebei Province, Hebei University, No.180, WuSi Road, Baoding, 071002, PR
China, E-mail: lijianheng@hbu.cn. bHebei Zhi Tong Bio-Pharmaceutical
Company
Abstract – Abiraterone acetate is used for the treatment of castration-resistant
prostate
cancer.
Abiraterone
acetate
was
synthesized
from
dehydroepiandrosterone via a three-step reaction including the formation of
tosylhydrazone, cross-coupling reaction and acetylation, and a two-step
purification including column chromatography and recrystallization with an
overall yield of 51.9%. Here, an improved procedure for the preparation of
abiraterone acetate is described. This synthetic process is of easy operation and
low cost, which is suitable for industrialization.
3β-Acetoxy-17-(pyrid-3-yl)androsta-5,16-diene (abiraterone acetate, 1) is
a
prodrug for
17-(pyrid-3-yl)androsta-5,16-dien-3β-ol (abiraterone, 2) which is used for the treatment of metastatic
castration-resistant prostate cancer (Figure 1). It is marketed as Zytiga in the United States and is a potent
inhibitor of human cytochrome P45017α (steroidal 17α-hydroxylase-C17,20-lyase).1,2
N
N
O
O
HO
1
2
Figure 1. Abiraterone acetate (1) and abiraterone (2)

462
HETEROCYCLES, Vol. 96, No. 3, 2018
The original method3 to synthesize 1 adopts the Suzuki coupling reaction of 17-enol triflate (3), derived
from 3-acetate of dehydroepiandrosterone and triflic anhydride, with diethyl(pyrid-3-yl)borane (4)
(Figure 2). This route is associated with high production cost, and formation of by-product such as
androsta-3,5,16-trien-17-yl triflate. Previously, one improvement4 was made to adopt the triethylamine
(TEA) rather than the expensive base 2,6-di-tert-butyl-4-methylpyridine (DTBMP) during the reaction
between triflic anhydride and dehydroepiandrosterone acetate (6, R=Ac) in order to reduce the production
of by-products. In this method, though dehydroepiandrosterone acetate (6, R=Ac) could be crystallized
and removed from the crude target compound 1 by recrystallization without using column
chromatography, the yield and the purity are much lower.
N
OTf
Pd catalyst
Na2CO3
N
+
O
O
BEt2
O
O
3
4
1
Figure 2. One method of synthesis of abiraterone acetate (1)
The key step of another route reported5 is the Suzuki coupling reaction based on
17-iodo-androsta-5,16-dien-3β-ol (5) with 4. However, coupling with the iodide was much slower, which
requires 4 days to complete this reaction, and it is hard to obtain the pure compound 1, the impurities in
the crude product only could be separated via reverse phase column chaomatography (Figure 3).
I
Ac2O
Pd catalyst
Na2CO3
+
4
2
1
pyridine
HO
5
Figure 3. Another method of synthesis of abiraterone acetate (1)
In another reported route,6 5 was also coupled with 3-bromopyridine (8) with the existence of
butyllithium. Cryogenic conditions are indispensable for reducing by-products. In addition, another
drawback of this method is the use of column chromatography for purification.
The N-sulfonylhydrazones have been applied widely in organic synthesis for over sixty years, especially

HETEROCYCLES, Vol. 96, No. 3, 2018
463
used in the preparation of diazo compound through Bamford-Stevens reaction.7 Over the last decade, the
palladium-catalyzed cross-coupling between sulfonylhydrazones and aryl halides has been established
and applied widely in organic synthesis.8 Based on our interest in the functionalization of
N-sulfonylhydrazones,9 we described a protocol to access 2 (Figure 4). Our aim was to develop a simple
and efficient new method for preparation of abiraterone acetate (1). This method was sensible for its
simple industrial art, moderate operating condition and high yield.
O
NNHTs
NH2NHTs, MeOH
reflux
RO
HO
6
7
Pd catalyst
base ligands
DMAP, AcCl
2
1
Br
N
8
Figure 4. The new method of synthesis of abiraterone acetate (1)
Differed from other reported routes, the present method involves Bamford-Stevens reaction using
dehydroisoandrosterone-17-N-toluenesulfonylhydrazone (7) as an intermediate compound instead of 5 or
3. The rationable for choosing 7 over 5 and 3 was due to the reduction of by-product, these known forms
of impurity such as androsta-3,5,16-trien-17-yl triflate and bis-steroidal compound were not formed due
to the totally different mechanism. Furtherly, the cost advantage of 3-bromopyridine is obviously over
diethyl(pyrid-3-yl)borane in the two reported methods.
On the basis of our experience and related literature, we began the study in examining the reaction
between 7 and 8. In an initial attempt, the target product 2 was obtained in 21.7% yield when using
Pd(OAc)2 as the catalyst and LiOtBu as the base and Xphos as the ligand. By contrast, the reaction time
of this present mehtod is quite longer than these reported routes, it requires 18 hours at 110 °C to
complete this reaction as compared with the 5 hours required when the compound 3 was used. Even
worse, what indicated the very low conversion rate was that there was still a lot of ingredient left. This
result was actually unacceptable and we almost gave this method up. However, to our delight, the yield
was obviously increased to 32.8% when using Pd(PPh3)4 as the catalyst.

464
HETEROCYCLES, Vol. 96, No. 3, 2018
The strong base and prolonged reaction time required for the cross-coupling reaction had enabled a
side-reaction to occur between the product 2 and the ingredient compound 8 to form a by-product. The
by-product was compound 9 (Figure 5). Whereas column chromatography on silica-gel of crude 2 afford
pure 9 which was eluted first. Compound 9 was not able to be removed completely by recrystallization.
Fortunately, the yield of this impurity was commoly under 10%, and it was easily reomved from the
crude compound 2 via column chromatography.
N
N
9
Figure 5. The by-product compound 9
During our development work, for the preparation of 2 using palladium-catalyzed cross-coupling of 7
with 8, our emphasis was on selection of solvent, base, ligand and catalyst.
Catalysts. Different palladium catalysts such as Pd(PPh3)2Cl2, palladium(II) acetylacetonate, and
palladium(II) acetate and several others were attempted for this cross-coupling reaction, and the results
are summarized in Table 1. As mentioned earlier, when employing Pd(OAc)2 as a catalyst the target
compound 2 could be isolated in 21.7% yield (entry 1), and Pd(PPh3)4 could increase the yield to 32.8%
(entry 3). Compared with Pd(OAc)2, Pd(PPh3)2Cl2 could slightly increase the yield and reduce reaction
time (entry 2). While the reaction efficiency dropped acutely when PdCl2(dppf) as a catalyst was
employed (entry 5). However, the reaction with palladium hydroxide and palladium chloride did not go to
completion though we increased the loading of catalyst and reaction time (entries 4 and 7). Either
Pd(acac)2 or PdCl2(MeCN)2 resulted in an acceptable yield (entries 6 and 8). In addition, the reaction
gave the best total yields of 51.9% when Pd2(dba)3 was used as a catalyst.
Table 1. Effect of catalysts for cross-coupling reactionα
Entry
Catalyst
Pd(OAc)2
Pd(PPh3)2Cl2
Pd(PPh3)4
PdCl2
Quantity (equiv) Time (h)
Yield (%) Remainder (%) Impurity (%)
1
2
3
4
0.01
0.01
0.01
0.03
18
20
17
30
21.7
28.6
32.8
-
68.6
60.4
57.2
90.7
7.8
8.3
8.5
7.3

HETEROCYCLES, Vol. 96, No. 3, 2018
465
5
6
7
8
9
PdCl2(dppf)
Pd(acac)2
0.02
0.01
0.03
0.01
0.01
26
20
30
24
24
16.2
39.6
-
73.9
50.2
89.9
56.6
38.2
6.7
7.7
8.1
6.8
8.4
Pd(OH)2
PdCl2(MeCN)2
31.7
Pd2(dba)3
51.9
αStandard conditions: 1,4-dioxane (6 vols), LiOtBu (0.7 equiv), Xphos (0.02 equiv), 7 (1.1 equiv), 8 (1.3
equiv), Reaction temperature (110 °C)
Solvents. Different solvents were screened for this cross-coupling reaction, and the resluts are
summarized in Table 2. The reported solvent10 for this type of reaction was 1,4-dioxane, and it was found
to be optimal among the represent solvents (entry 1). To our disappointment, the reaction failed to be
conducted when using toluene as a solvent, probably owing to the poor solubility (entry 2). MeCN
resulted in a terrible yield (entry 3), and THF furnished the product in a low due to the low reaction
temperature of 75 °C (entry 4). Only did DME provide the final product an acceptable yield (entry 5).
Bases. Base played a significant role in Bamford-Stevens reaction. We tried our best to select the most
suitable one for this reaction among the represent bases. The results are also summarized in Table 2. The
reported base for this type of reaction is LiOtBu,11 and it really contributed to the reaction (entry 1).
However, more by-product was obtained with the attempt to increase the loading of LiOtBu to 1 equiv.
Compared with LiOtBu, NaOtBu and KOtBu failed to provide 2 in acceptable yields. Moreover, the yield
of the impurity compound 9 obviously increased for these extreme alkalinty (entries 6 and 7). Worse, the
reaction failed to be conducted when using Cs2CO3 or piperidine as the additive (entries 8 and 9).
Table 2. Effect of solvents and basesα
Entry
Solvent
1,4-dioxane
toluene
Bases
LiOtBu
LiOtBu
LiOtBu
LiOtBu
LiOtBu
NaOtBu
KOtBu
Cs2CO3
Temp (°C) Time (h) Yield (%) Remainder (%) Impurity (%)
1
2
3
4
5
6
7
8
9
110
110
95
24
20
30
20
24
24
24
24
24
51.9
-
38.2
100
77.3
100
55.2
59.6
58.5
100
100
9.4
-
MeCN
13.7
-
5.6
-
THF
75
DME
95
36.3
24.7
23.6
-
6.2
13.7
16.4
-
1,4-dioxane
1,4-dioxane
1,4-dioxane
110
110
110
110
1,4-dioxane piperidine
-
-
αStandard conditions: Pd2(dba)3 (0.01 equiv), Xphos (0.02 equiv), 7 (1.0 equiv), 8 (1.2 equiv)

466
HETEROCYCLES, Vol. 96, No. 3, 2018
Ligands. Various kind of ligands were added in the reaction including Xphos, Sphos, dppm and so on
(Table 3). To our delight, about half of the represent ligands exhibited good reactivity (entries 1-3, 7, 9
and 11). The yield was signficantly diminished when using Binap or dppb as a ligand (entries 5 and 8).
However, Dave Phos, DpePhos and dppm gave the same terrible results (entries 4, 6 and 10).
Table 3. Effect of ligandsa
Entry
Ligand
Xphos
JohnPhos
Sphos
Time (h)
24
Yield (%)
51.9
42.7
32.3
-
Remainder (%)
38.2
Impurity (%)
1
2
9.4
5.5
6.7
6.7
9.6
8.4
6.5
7.8
7.2
9.3
8.5
24
51.2
3
20
59.6
4
Dave Phos
Binap
24
92.4
5
22
14.6
-
74.7
6
DpePhos
XantPhos
dppb
24
90.3
7
22
35.2
11.3
40.9
-
55.6
8
24
80.8
9
dppe
24
51.4
10
dppm
24
88.3
11
dppp
24
29.7
60.1
αStandard conditions: 1,4-dioxane (6 vols), LiOtBu (0.7 equiv), Pd2(dba)3 (0.02 equiv), 7 (1.0 equiv), 8
(1.3 equiv), Reaction temperature (110 °C)
The optimal condition has been selected among those various reaction conditions. The target product was
obtained in 51.9% when using Pd2(dba)3 as a catalyst and using 1,4-dioxane as a solvent together with
using LiOtBu and Xphos as base and ligand, the temperature was at 110 °C during this reaction.
In summary, a simple and efficient method for the preparation of abiraterone acetate with improved yield
and less by-product is reported. And we have also discovered the optimized conditions. We anticipate that
this approach will find applications in the preparation of abiraterone acetate. Futher improvement of this
catalytic reaction is underway in our laboratory.
EXPERIMENTAL
General Experimental Procedure for the Preparation of Abiraterone Acetate. Reagents were uesd as such
1
without purification. H NMR (600 MHz) and 13C NMR (150 MHz) were recorded using a Bruker
spectrometer. The chemical shift data are reported as δ (ppm) using tetramethylsilane as internal standard.
Mass spectra were recorded using an Agilent 1200-6320 Ion Trap XCT instrument. HPLC analysis was

HETEROCYCLES, Vol. 96, No. 3, 2018
467
performed on a Shimadzu instrument with a UV detector (210 nm) using a Shimadzu Wondasil C18 (250
mm x 4.6 mm, 5 μm) column. Mobile phase A [prepared by mixture of MeOH and MeCN in the ratio
(80:20 v/v)]: mobile phase B [water], flow 1 mL/min, gradient 65:35 (0-25 min), 90:10 (25-40 min),
65:35 (40-50 min), 65:35 (50-60 min).
Dehydroepiandrosterone-17-N-tosylhydrazone (7)
Into a 100 mL round-bottomed flask, fitted with a magnetic stirrer bar, was placed p-toluenesulfonyl
hydrazide (0.931 g, 5 mmol) and dehydroepiandrosterone (6, R=H) (1.442 g, 5 mmol) and MeOH (40
mL), followed by 3 mol·L-1 sulfuric acid (3 M, 0.05 mL). The temperature was raised to reflux to get a
clear solution, after stirring for 5 h, the color of solution turned to light green. Then water (40 mL) was
added over a period of 10 min, then the reaction mixture was cooled to room temperature and poured into
cold water (100 mL), and filtered. The precipitate was collected, washed with a mixture solution of water
(40 mL) and MeOH (10 mL) for two times. The product was dried in oven at 70 °C for 2 h, to give 2.17 g
1
(91.6%) of 7 as light green powders. mp 176-178 °C. H NMR (600 MHz, CDCl3): δ 1.05 (s, 3H,
19-CH3), 1.07 (s, 3H, 18-CH3), 2.26-2.30 (m, 2H, 16-2H), 2.42 (s, 24-CH3), 3.51-3.58 (m, 1H, 3α-H),
5.39 (d, 1H, J=4.8 Hz, 6-H), 7.29 (d, 2H, J=7.8 Hz, Ts2-H, Ts6-H), 7.82 (d, 2H, J=8.4 Hz, Ts3-H, Ts4-H).
13C NMR (150 MHz, CDCl3): δ 16.56, 19.37, 20.47, 21.58, 23.39, 25.98, 31.22, 31.30, 31.58, 33.69,
36.62, 37.20, 42.21, 44.74, 50.29, 53.60, 71.62, 120.92, 128.01, 128.28, 129.31, 129.94, 135.56, 141.07,
143.75, 171.63. ESI HRMS: calcd for C26H36N2O3S+Na 479.2339, found 479.2332.
17-(Pyrid-3-yl)androsta-5,16-dien-3β-ol (2)
A stirred solution of 7 (1.89 g, 4 mmol) in 1,4-dioxane (50 mL) in a 100 mL round-bottomed flask was
purged with nitrogen and tris(dibenzylideneacetone)dipalladium(0) catalyst (0.0405 g, 0.044 mmol) and
2-dicyclohexylphosphino-2′,4′,6′-triisopropylbiphenyl ligand (0.0385 g, 0.076 mmol) were added. After
sitrring for 10 min, to the resultant rufous solution was added lithium tert-butoxide (0.2227 g, 2.78 mmol)
base, then sitrring for 5 min and 8 (0.5 mL, 5.31 mmol) was added. The flask was fitted with a reflux
condenser and these were purged with nitrogen. The mixture was heated to the refluxing temperature
(110 °C) in an oil bath pan with stirring for 24 h. The reaction was completed and the solution turned to
orange from dark brown. The reaction mixture was cooled to room temperature. Then EtOAc (20 mL)
and cold water (20 mL) was added slowly at low temperature (10 °C) into the reaction mixtures. The
organic phase was separated and washed with water (3 x 10 mL), then the constituents were isolated and
purified by column chromatography, with the ratio of MeOH to CH2Cl2 was 1:100 in developing agent.
The solvent was removed by vacuum distillation and the purified product was dried with anhydrous
sodium sulfate. The yellow powder was collected in a little beaker to give crude 2 (0886 g, 62.9%) as pale
yellow crystals, mp 213-215 °C, lit.12 212-215 °C. 1H NMR (600 MHz, CDCl3): δ 0.95 (s, 3H, 19-CH3),
0.97 (s, 3H, 18-CH3), 3.41-3.63 (m, 1H, 3α-H), 5.32 (d, 1H, J=4.8 Hz, 6-H), 5.90 (s, 1H, 16-H), 7.12-7.18

468
HETEROCYCLES, Vol. 96, No. 3, 2018
(m, 1H, Py5-H), 7.58 (d, 1H, J=8.0 Hz, Py4-H), 8.37 (d, 1H, J=4.4 Hz, Py6-H), 8.54 (s, 1H, Py2-H). 13C
NMR (150 MHz, CDCl3): δ 16.56, 19.33, 19.41, 20.37, 30.47, 31.46, 31.62, 31.80, 35.28, 36.66, 37.22,
42.30, 47.35, 51.80, 57.57, 71.50, 121.24, 123.05, 129.28, 133.05, 133.78, 141.13, 147.67, 151.67. ESI
HRMS: calcd for C24H31NO+H 350.2478, found 350.2477.
3β-Acetoxy-17-(pyrid-3-yl)androsta-5,16-diene (1)
A stirred solution of the product from the foregoing reaction (0.752 g, 2.05 mmol) in dry CH2Cl2 (30 mL)
in
a
100 mL round-bottomed flask was added triethylamine (0.5 mL, 2.76 mmol),
4-dimethylaminopyridine (1.37 mg, 0.01 mmol). Acetyl chloride (0.25 mL, 3.15 mmol) was added at
20-30 °C over a period of 15 min. The mixture stirred at room temperature for 2 h. After reaction
completed, water (50 mL) was added and stirred for 10 min, and the pH was adjusted to 6-7 with 0.1
mol·L-1 aqueous NaHCO3. The organic layer was separated, washed with water (3 x 20 mL), and then
concentrated. To the product was added anhydrous MeOH (30 mL), and the solution was heated to reflux
and treated with activated carbon (3 g) for 2 h. The reaction mixture was then cooled to room temperature
and stirred for 2 h. Then the product was obtained through filtering and washed with 80% EtOH and dried
under vacuum for 6 h to give pure 1 (0.745 g) as white crystrals. The total productivity of pure 1 was
51.9% (based on 6: R=H), mp 143-145 °C, lit.13 144-145 °C. 1H NMR (600 MHz, CDCl3): δ 1.05 (s, 3H,
19-CH3), 1.08 (s, 3H, 18-CH3), 2.04 (s, 3H, CH3CO2), 4.58-4.66 (m, 1H, 3α-H), 5.42 (d, 1H, J=4.76 Hz,
6-H), 5.99 (s, 1H, 16-H), 7.22 (dd, 1H, J1=4.8 Hz, J2=7.8 Hz, Py5-H), 7.64 (d, 1H, J=7.9 Hz, Py4-H),
8.46 (d, 1H, J=4.6 Hz, Py6-H), 8.63 (s, 1H, Py2-H). 13C NMR (150 MHz, CDCl3): δ 16.56, 19.29, 20.48,
21.44, 27.72, 30.37, 31.78, 35.27, 36.64, 37.02, 38.17, 47.35, 50.34, 57.46, 73.82, 122.16, 122.89, 132.87,
133.64, 140.62, 147.82, 147.96, 151.66, 170.38. ESI HRMS: calcd for C26H33NO2+Na 392.2584, found
392.2575. The spectroscopic data of the final product from this procedure were identical with those
reported for the product obtained by the route previously described.14
REFERENCES
1. B. Dong, L. Fan, Y. Wang, C. Chi, X. Ma, R. Wang, W. Cai, X. Shao, J. Pan, Y. Zhu, X. Shangguan,
Z. Xin, J. Hu, X. Xie, X. Kang, L. Zhou, and W. Xue, Prostate, 2017, 77, 1373.
2. J. S. de Bono, M. R. Smith, F. Saad, D. E. Rathkopf, P. F. A. Mulders, E. J. Small, N. D. Shore, K.
Fizazi, P. De Porre, T. Kheoh, J. Li, M. B. Todd, C. J. Ryan, and T. W. Flaig, Eur. Urol., 2017, 71,
656.
3. R. D. Bruno, T. S. Vasaitis, L. K. Gediya, P. Purushottamachar, A. M. Godbole, Z. Ates-Alagoz, A.
M. H. Brodie, and V. C. O. Njar, Steroids, 2011, 76, 1268.
4. N. J. Hunt, US, US 7700766, 2010.
5. A. P. Encabo, J. Á. Turiel Hernández, F. J. Gallo Nieto, A. L. Bonde-Larsen, and C. M. Sandoval

HETEROCYCLES, Vol. 96, No. 3, 2018
469
Rodríguez, EP, EP2607371 A1, 2013.
6. M. Mizhiritskii, E. Marom, and S. Rubnov, US, US 20140228417 A1, 2014.
7. X. Zhao, J. Jing, K. Lu, Y. Zhang, and J. Wang, Chem. Commun., 2010, 46, 1724.
8. Y. Xia, F. Hu, Y. Xia, Z. Liu, F. Ye, Y. Zhang, and J. Wang, Synthesis, 2016, 49, 1073.
9. Z.-S. Chen, X.-H. Duan, L.-Y. Wu, S. Ali, K.-G. Ji, P.-X. Zhou, X.-Y. Liu, and Y.-M. Liang, Chem.
Eur. J., 2011, 17, 6918.
10. Z. Shao and H. Zhang, Chem. Soc. Rev., 2012, 41, 560.
11. Q. Xiao, J. Ma, Y. Yang, Y. Zhang, and J. Wang, Org. Lett., 2009, 11, 4732.
12. T. Cristiano and M. Alpegiani, US, US 20170044207 A1, 2017.
13. A. N. Balaev, A. V. Gromyko, and V. E. Fedorov, Pharm. Chem. J., 2016, 50, 404.
14. M. K. Madhra, H. M. Sriram, M. Inamdar, M. K. Sharma, M. Prasad, and S. Joseph, Org. Process
Res. Dev., 2014, 18, 555.