Development of a gram-scale synthesis of pbrm, an irreversible inhibitor of 17beta-hydroxysteroid dehydrogenase type 1

Article abstract of DOI:10.1021/acs.oprd.8b00402

Efforts toward the development of a reliable gram scale synthesis of PBRM, a potent and selective steroidal covalent inhibitor of 17β-hydroxysteroid dehydrogenase type 1 (17β-HSD1), are described. Among the three synthetic routes (C-E) developed herein, route E is the most efficient one with only 6 chemical steps from commercially available estrone, and an overall yield of 13% leading to PBRM with a high HPLC grade purity (99.7%) after recrystallization. Important improvements have been achieved in this sequence from previous reported routes (A and B). Notably, we used a palladium catalyzed Suzuki-Miyaura cross-coupling reaction to rapidly install the requested C3 chain on estrone. Also, catalytic hydrogenation of the C16-enone was shortened by half using Pearlman's catalyst. Finally, we used a selective bromination through deoxygenation of alcohol at the last step of the sequence to provide PBRM without dehydration of its carboxamide functionality, a persistent problem observed in other routes. Crystals of PBRM were also obtained from recrystallization in acetonitrile and submitted to x-ray analysis, which confirmed the PBRM structure. This work now makes it possible to start a proof-of-principle in a non-human primate model for the treatment of endometriosis, while supporting its future pharmacological development.

Full text of DOI:10.1021/acs.oprd.8b00402

Subscriber access provided by UNIV OF MISSOURI ST LOUIS
Full Paper
Development of a Gram-Scale Synthesis of PBRM, an Irreversible
Inhibitor of 17beta-Hydroxysteroid Dehydrogenase Type 1
René Maltais, and Donald Poirier
Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/acs.oprd.8b00402 • Publication Date (Web): 19 Sep 2019
Downloaded from pubs.acs.org on September 19, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 32
Organic Process Research & Development
1
2
3
4
5
Organic Process Research and Development (second revised version)
6
7
8
9
Development of a Gram-Scale Synthesis of PBRM, an Irreversible Inhibitor of 17beta-
Hydroxysteroid Dehydrogenase Type 1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
René Maltais† and Donald Poirier*†,‡
†
Laboratory of Medicinal Chemistry, Endocrinology and Nephrology Unit, CHU de Québec –
Research Center, Québec, QC, Canada
‡ Department of Molecular Medicine, Faculty of Medicine, Université Laval, Québec, QC, Canada
(*) Corresponding Author:
Donald Poirier
Laboratory of Medicinal Chemistry
CHU de Québec – Research Center (CHUL, T4-42)
2705 Laurier Boulevard
Québec, QC, G1V 4G2, Canada
Tel.: 1-418-654-2296; Fax: 1-418-654-2298; E-mail: donald.poirier@crchul.ulaval.ca
1
ACS Paragon Plus Environment

Organic Process Research & Development
Page 2 of 32
1
2
3
4
5
Table of Contents:
6
7
8
9
Synthetic routes
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
O
OH
Route A
7% (10 steps)
6-17% (8 steps)
Route B
O
H
H
Route C
Route D
10% (9 steps)
31% (6 steps; purity : 95.2%)
13% (6 steps; purity : 99.7%)
NH2
H
H
H
H
HO
Br
Route E
(1 g = 2 $)
Estrone
PBRM
2
ACS Paragon Plus Environment

Page 3 of 32
Organic Process Research & Development
1
2
3
ABSTRACT
4
5
6
7
8
9
Efforts toward the development of a reliable gram scale synthesis of PBRM, a potent and
selective steroidal covalent inhibitor of 17β-hydroxysteroid dehydrogenase type 1 (17β-HSD1), are
described. Among the three synthetic routes (C-E) developed herein, route E is the most efficient one
with only 6 chemical steps from commercially available estrone, and an overall yield of 13% leading to
PBRM with a high HPLC grade purity (99.7%) after recrystallization. Important improvements have
been achieved in this sequence from previous reported routes (A and B). Notably, we used a palladium
catalyzed Suzuki-Miyaura cross-coupling reaction to rapidly install the requested C3 chain on estrone.
Also, catalytic hydrogenation of the C16-enone was shortened by half using Pearlman’s catalyst.
Finally, we used a selective bromination through deoxygenation of alcohol at the last step of the
sequence to provide PBRM without dehydration of its carboxamide functionality, a persistent problem
observed in other routes. Crystals of PBRM were also obtained from recrystallization in acetonitrile
and submitted to x-ray analysis, which confirmed the PBRM structure. This work now makes it
possible to start a proof-of-principle in a non-human primate model for the treatment of endometriosis,
while supporting its future pharmacological development.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
KEYWORDS: Covalent inhibitor, 17beta-hydroxysteroid dehydrogenase type 1, steroid, multigram
synthesis
3
ACS Paragon Plus Environment

Organic Process Research & Development
Page 4 of 32
1
2
3
1. INTRODUCTION
4
5
6
7
8
9
PBRM, formally named as 3-{[(16β,17β)-3-(2-bromoethyl)-17-hydroxyestra-1,3,5(10)-trien-16-
yl]methyl}benzamide (Figure 1), is the first selective covalent inhibitor of 17β-hydroxysteroid
dehydrogenase type 1 (17β-HSD1).1 Since this enzyme is involved in the last step of estrogen
biosynthesis by converting the weak estrogen estrone (E1) into the highly potent estrogen estradiol
(E2), the use of a 17β-HSD1 inhibitor for the treatment of estrogen-dependent diseases (EDDs) has
been identified as a promising therapeutic approach.2 Clinical validation of this therapeutic target
remains to be done for the treatment of EDDs including breast cancer,3 endometrial cancer,4 and
endometriosis.5 PBRM has been identified as a promising lead candidate with desirable properties such
as strong inhibitory activity on 17β-HSD1 (in vitro and in vivo), good oral bioavailability, and low
toxicity profile in rodent.6 A successful proof-of-principle study on a xenograft model with estrogen-
sensitive human breast cancer (T-47D cells) in nude mice demonstrated the potential of such a selective
covalent inhibitor to treat EDDs.7 As the next step of its therapeutic validation, we are now interested
in studying the potency of PBRM for the treatment of endometriosis.8 Importantly, PBRM is an
irreversible inhibitor that requires the presence of a histidine (His) residue for an alkylation on 17β-
HSD1.9 Considering the presence of a glycine rather that His at key position 221 in rodent, this animal
model is inadequate for efficiency studies. Fortunately, monkeys represent a valuable model to
evaluate PBRM efficiency, since they possess the same His as humans.10 Moreover, PBRM has been
shown to be a potent inhibitor of 17β-HSD1 in ovarian tissue from monkeys with a comparable level of
inhibition as those measured in human ovarian tissues.9b Thus, in order to validate the potency of
PBRM in a relevant non-human primate endometriosis model such as the baboon,11 the elaboration of a
synthetic route allowing the preparation of a gram-scale quantity was absolutely necessary.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
18
OH
12
11
1'
17
2''
1''
O
1
4
13
14
9
8
3''
10
16
15
2
NH2
6''
4''
7
Br
5
3
5''
6
Figure 1. Carbon numbering and molecular structure of PBRM, a steroid derivative inhibiting 17β-
HSD1.
Two chemical syntheses of PBRM have been initially reported (Scheme 1, routes A and B).1
The first one (route A), developed in a medicinal chemistry context, leads to milligram quantities and
4
ACS Paragon Plus Environment

Page 5 of 32
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
allowed the in vitro assessment of PBRM inhibitory activities. This initial route consisted in 10 steps
from commercially available E1 and resulted in an overall yield of 7% (Table 1).1a To obtain a few
grams of PBRM and to proceed to an in vivo proof-of-principle on a breast cancer xenograft model in
mice,7 this first synthesis was later optimized and shortened to 8 steps (route B, scale-up 1) with an
overall yield of 17%.1b However, major limitations were encountered with these two routes, including
variable yields, incomplete and/or unselective bromination, as well as laborious final product
purification. These problems forced us to redesign the synthesis to provide an optimal route. We
present herein, the optimization efforts that led to a shorter synthesis of 6 steps that provided the
required quantity of PBRM. This work represents a critical step to start our proof-of-principle study in
baboons, and consequently to promote the use of PBRM as a new drug to treat endometriosis and other
EDDs.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
O
O
O
O
O
a
b
c
H
H
H
H
H
H
H
H
H
H
H
H
HO
TfO
3
2
1
E1
d
O
k
O
H
O
O
O
H
H
O
g
l
H
H
H
HO
NH2
4
H
H
H
H
H
H
HO
HO
e, f
O
10
11
9
O
H
h, i
H
H
BnO
OH
OH
5
O
O
i
H
H
g
NH2
NH2
H
H
H
H
HO
BnO
h
O
8
7
O
H
NH2
H
H
j
BnO
6
OH
O
H
Route A:
Route B:
NH2
H
H
Br
PBRM
Scheme 1. Synthesis of PBRM via route A (10 steps) and route B (8 steps) from E1. Reagents and
conditions: (a) Trifluoromethanesulfonic anhydride, TEA, DCM, rt; (b) CH2=CHSnBu3, Pd(PPh3)4,
LiCl, THF, rt (route A) or Potassium vinyltrifluoroborate, PdCl2, Cs2CO3, PPh3, THF/H2O (9:1), 80 °C
(route B); (c) Ethylene glycol, p-TSA, Dean-Stark, toluene, reflux; (d) i) BH3-(CH3)2S, THF, -78 °C-rt;
ii) NaHCO3, H2O2, rt, 3 h; (e) NaH, benzylbromide, DMF, 0 °C-rt, 18 h; (f) HCl, MeOH, rt, 5 h; (g) 3-
5
ACS Paragon Plus Environment

Organic Process Research & Development
Page 6 of 32
1
2
3
4
5
Formyl-benzamide, KOH, EtOH, reflux, 0.5 h; (h) NaBH4, MeOH, DCM, rt, 1 h; (i) H2, 10% Pd/C,
MeOH, rt, 36 h; (j) PPh3, CBr4, DCM, 0 °C to rt; (k) Oxone, NaHCO3, acetone/ACN (1:2), rt; (l) Pd/C
(10%), ammonium formate, MeOH, 70 °C.
6
7
8
9
2. RESULTS AND DISCUSSION
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Limitations of synthetic route B and attempts for optimization
The major issue encountered with the last reaction step strongly limited the conversion of route
B into a reliable scale-up synthesis. The selective bromination of the primary alcohol 8 into PBRM
gave unreproducible yields varying from 30-73%, but most of the time below 50%. Different aspects of
this reaction were found to be problematic. First, the solubility of alcohol 8 in dichloromethane (DCM),
the usual solvent used for the Appel reaction,12 was very limited and forced us to increase the number
of equivalents of PPh3 and CBr4 to drive the reaction to completion. Under this large reagent excess,
the carboxamide functionality on the 16β-side chain was found vulnerable to dehydration leading to the
corresponding nitrile. Second, 17β-OH was also found to react in those conditions and to form the
corresponding secondary bromide as a minor side product (10-15%). As a consequence of reagent
excess, PBRM was difficult to isolate from the triphenylphosphine oxide formed as a by-product. Their
very similar mobility on silica gel led to multiple laborious flash chromatography purifications.
Therefore, we explored different bromination methods as alternatives to the Appel reaction. The first
method tested was based on the use of (chloro-phenylthio-methylene)dimethylammonium chloride
(CPMA) in the presence of tetrabutylammonium bromide (TBAB), which has been reported for the
selective bromination of primary alcohol.13 Interestingly, CPMA provided a good yield (70%) for
halogenation of 8 selectively as opposed to the secondary 17-beta alcohol. Unfortunately, the presence
of a chlorination compound, found in a ratio 1:3 versus the bromide (PBRM), disqualified this reaction
as both compounds were found to be inseparable by chromatography. Varying the TBAB equivalent,
the temperature and the concentration did not alter this ratio. Two other bromination assays, using N-
bromosuccinimide and triphenylphosphine in DMF14 or tetraethylammonium halide and Et2NSF2BF4
(XtalFluor-E) in DCM,15 also gave unsatisfactory results by leading to carboxamide dehydration.
Overall, route B from batch 1 and 2, was found to be more straightforward than route A, but provided
variable overall yields (6-17%, Table 1). The direct bromination of the primary alcohol in presence of
the carboxamide remains the major limitation to its translation to a reliable and scalable synthesis of
PBRM and resulted in variable yields.
6
ACS Paragon Plus Environment

Page 7 of 32
Organic Process Research & Development
1
2
3
4
Table 1. Comparison of the five synthetic route efficiencies for the synthesis of PBRM.
5
6
7
Synthetic
route
Number
of steps
Starting
material (g) product (g)
Final
Overall yield
(%)
HPLC purity
(%)
8
9
(Estrone)
1.6
(PBRM)
0.2
A1a
10
8
7
98.5
98.5
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
B1b
8.6
2.5
3.7
17
(batch 1)
B
8
33.6
6
96.7
(batch 2)
C
D
E
9
6
6
60.0
38.0
30.0
11.3
23.5
9.2 (6.9*)
10
31
17 (13*)
97.6
95.2
95.1 (99.7*)
* Obtained after recrystallization from 9.2 g of PBRM in ACN.
Synthetic Route C
Synthesis of PBRM via route B was successful on a relatively small scale (< 5 g), but with
important limitations related to the last synthetic step (bromination). Consequently, we redesigned the
synthetic sequence, seeking to avoid the bromination in the presence of the carboxamide functionality
(Scheme 2). Thus, we preserved the first four steps of route B and replaced the carboxamide by a nitrile
group, which would be transformed back to primary amide at the last step of the sequence using mild
hydration conditions.16 Prior starting the sequence, we first tried to optimize the reactions needed for
the synthesis of alcohol 10, with a particular interest for the opening of epoxide 9 (Scheme 1). In fact,
we observed variable yields (31-71%) for the epoxide opening in the scale-up process necessary to
obtain a larger quantity of 10. The formation of the corresponding methyl-ether (20-25%), coming from
the competition of methanol over hydrogen for the nucleophilic attack, was found to be in large part
responsible for this significant lower yield. To avoid this side reaction, we tried to replace the methanol
by less nucleophilic hindered alcohols such as isopropanol and tert-butyl-alcohol, but without
significant improvement. Change in reaction temperature or reagent concentration was also found
ineffective to avoid this by-product formation. Nevertheless, we obtained a sufficient quantity of
alcohol 10 (46 g) to engage next step of the sequence. Indeed, alcohol 10 underwent aldolization with
3-CN-benzaldehyde using NaOH solution in EtOH at room temperature instead of KOH solution in
refluxing EtOH as reported in routes A and B. This strategy limited the hydrolysis of nitrile to
carboxamide and provided an acceptable yield (> 53%) of 12. The subsequent reduction of the enone to
allylic alcohol 13 proceeded in high yield as previously reported for routes A and B. The
7
ACS Paragon Plus Environment

Organic Process Research & Development
Page 8 of 32
1
2
3
4
5
6
7
8
9
stereoselective catalytic hydrogenation of the allylic alcohol 13 to 16β-benzyl-CN derivative 14 was
however a potential issue because the nitrile group is known to be sensitive to catalytic hydrogenation,
thus forming the corresponding primary amine via an imine intermediate.17 In fact, the nitrile function
was partially hydrogenated during the reaction time (96 h) necessary to complete the double bond
reduction of the 17β-allylic alcohol 13. As result, a moderated yield (47%) of the saturated alcohol 14
was obtained. The amine by-product was however easily removed by a simple filtration on silica gel.
Unexpectedly, the hydrogenation of the allylic alcohol was not completely stereoselective and gave a
85:15 mixture of 16β/16α benzyl-CN side-chain, as determined by NMR using the representative 18-
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
13
CH3 and 17-CH signals. The key H and C NMR signals for each of the four possible diastereomers
at C16/C17-stereogenic centers were previously identified by us.18 Fortunately, this 16α-benzyl-CN
isomer was efficiently removed by chromatography in subsequent steps of purification, and no trace of
this isomer was found in the final PBRM product. The subsequent bromination of alcohol 14 gave, as
expected, a good yield (85%) of compound 15. Finally, the conversion of the nitrile to corresponding
carboxamide by a palladium-dichloride catalyzed hydration with acetamide16 provided a satisfactory
90% of PBRM in a pure form after flash chromatography. A quantity of 11.3 g of PBRM was thus
generated in a high purity level (97.6%) using route C (Table 1).
O
O
O
O
a
H
H
c
d
b
H
H
E1
H
H
H
H
H
H
H
H
(60 g)
HO
TfO
O
2
1
(Filtration SiO2)
(Filtration SiO2)
9
(No chromatography)
(No chromatography)
10
46.2 g
HPLC purity = 88.8%
Y = 72%
82.7 g
HPLC purity = 99.1%
Y = 92%
48.3 g
HPLC purity = N/A
Y = 99% (crude)
46.1 g
HPLC purity = 75.0%
Y = 71%
e
OH
OH
O
H
g
f
CN
H
H
H
H
CN
H
H
CN
H
H
HO
HO
HO
14 (Filtration SiO2)
13 (No chromatography)
(Filtration SiO2)
12
16.1 g
HPLC purity = 77.0%
Y = 47%
31.5 g
HPLC purity = 80.6%
Y = 83%
30.5 g
HPLC purity = 79.0%
Y = 53%
h
OH
OH
O
H
H
i
CN
H
H
NH2
H
H
Br
Br
15 (Chromatography)
PBRM
(Chromatography)
14.2 g
HPLC purity = 84.8%
Y = 85%
11.3 g
HPLC purity = 97.6%
Y = 90%
8
ACS Paragon Plus Environment

Page 9 of 32
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Scheme 2. Gram scale synthesis of PBRM via route C (9 steps) from E1. Reagents and conditions:
(a) Trifluoromethanesulfonic anhydride, TEA, DCM, rt; b) Potassium vinyltrifluoroborate, PdCl2,
Cs2CO3, PPh3, THF/H2O (9:1), 80 °C; (c) Oxone, NaHCO3, acetone/ACN (1:2); (d) Pd/C (10%),
ammonium formate, MeOH, 70 °C (e) 3-CN-benzaldehyde, NaOH, EtOH, rt; (f) NaBH4, MeOH,
DCM, rt, 3 h; (g) H2, 10% Pd/C, EtOH, rt, 96 h; (h) PPh3, CBr4, DCM, 0 °C to rt; (i) PdCl2, acetamide,
THF/H2O, rt, 24 h. (*) Yields for this reaction varied from 37 to 71%, based on different assays..
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthetic Route D
The variable yields for the formation and opening of epoxide 9 and the moderate yields for the
aldolization and catalytic hydrogenation steps lead us to doubt that route E could be an efficient
synthetic strategy for scale-up. For this reason, we had considered using a straightforward method to
install the bromoethyl side chain directly at the C3 position of the E1 scaffold. At first sight, the use of
a Suzuki-Miyaura palladium catalyzed cross-coupling reaction,19 between estrone-triflate (1) and
alkoxyethylboron coupling partner, represented an attractive option.20 Unfortunately, the synthetic
methods available for the preparation of the alkoxyethylboron compounds were quite limited at that
time and were reported only on very small scales.21 As a possible alternative, Fleury-Brégeot et al.22
reported the successful synthesis of different potassium alkoxyethyltrifluoroborates via a copper-
catalyzed borylation. This method would simplify the synthesis protocol and solve stability issues
encountered with alkoxyethylboron derivatives.22 Therefore we revisited the initial sequence based on a
Suzuki-Miyaura reaction and elaborated a new synthetic strategy (route D, Scheme 3).
As validation step, we first performed the Suzuki-Miyaura coupling with estrone-triflate (1) on
a small scale. Since Fleury-Brégeot et al.,22 reported coupling only with aryl halides, we used the
conditions published by Molander et al.,23 who reported different examples of cross-coupling reactions
of potassium alkyltrifluoroborate with aryl triflates. We obtained a similar coupling yield (67%) as
those published (66-75%) with the exception that a small amount (15%) of the aryl-triflate reductive
product (3-deoxy-E1). Inseparable at this stage, the mixture was found to be separable later in the
sequence of reactions by chromatography. We then proceeded to this coupling reaction on a larger
quantity with estrone-triflate (30 g) and potassium (2-benzyloxyethyl)trifluoroborate (21 g) with
comparable yield to small scale assay. The 2-benzyloxyethyl)trifluoroborate reagent being successfully
prepared using the published procedure.22 With a multigram quantity of 5 in hands, we then followed
the previously reported sequence of reactions1b to 1) introduce the 16β-benzylcarboxamide chain via an
aldol condensation to provide 6, 2) reduce the enone to allylic alcohol 7, and 3) hydrogenate the double
9
ACS Paragon Plus Environment

Organic Process Research & Development
Page 10 of 32
1
2
3
4
5
6
7
8
bond to benzylcarboxamide derivative 8. This sequence of reactions had the advantage to remove the
benzyl ether protecting group during the palladium catalytic hydrogenation of the allylic alcohol 7. On
this larger scale, we however observed that the reaction time to complete the reduction of the allylic
alcohol was found to be much longer than observed in smaller scale (from 1 to 9 days).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
O
O
O
O
H
H
H
a
b
H
H
NH2
H
H
H
H
BnO
BnO
TfO
(No chromatography)
6
(Filtration SiO )
1
5
2
60 g (from route C)
39.0 g
55.9 g
HPLC purity = 91.9%
Y = 62%
HPLC purity = 72.5%
Y = 85%
c
OH
OH
OH
d
e
O
H
O
H
H
X
H
H
NH2
NH2
H
H
H
H
Br
HO
BnO
8
(Chromatography)
(Filtration SiO2)
(Filtration SiO2)
7
PBRM
(X = CONH2)
29.4 g
46.4 g
23.5 g
f
HPLC purity = 81.6%
Y = 86%
HPLC purity = 69.9 %
Y = 80%
HPLC purity = 95.2%
Y = 81%
(X = CN)
15
Scheme 3. Gram scale synthesis of PBRM via route D (6 steps) from E1. Reagents and conditions:
(a) Potassium (2-benzyloxyethyl)trifluoroborate, PPh3, PdCl2, Cs2CO3, THF/H2O (9:1), 90 °C; (b) 3-
Carboxamide-benzaldehyde, NaOH (10%), EtOH/DCM (9:1), (c) NaBH4, MeOH, DCM, rt, 1 h; (d)
H2, 10% Pd/C, EtOH, rt, 9 days; (e) PPh3, CBr4, THF, 0 °C to rt; (f) PdCl2, acetamide, THF/H2O, rt, 24
h.
Even though intermediate 8 was obtained in good overall yield, we had not resolved the
problem regarding the last bromination step. The very low solubility of diol 8 in DCM was a possible
cause of this reactivity problem, because of incomplete and slow conversion of the alcohol was
observed. Consequently, we investigated the tolerance of the Appel reaction for other solvents than
DCM. Among aprotic solvents that allow complete dissolution of 8 at 0 °C, THF was selected as an
alternative to DCM. To our knowledge, no examples of Appel’s reaction were reported using THF. We
performed the bromination of 8 using conventional conditions, exchanging DCM for anhydrous THF.
Interestingly, we obtained PBRM in good and reproducible yields with a rapid and complete
conversion of the starting alcohol. However, as observed with the classic Appel reaction conditions,24
the formation of the by-product 15 (35%) coming from carboxamide dehydration was also obtained.
Fortunately, this nitrile side product was easily separated from PBRM by flash chromatography and
10
ACS Paragon Plus Environment

Page 11 of 32
1
Organic Process Research & Development
2
3
4
5
successfully recycled to PBRM using the palladium-catalyzed hydration conditions previously used in
route C.16 During this last flash chromatography, the 3-deoxy-E1 by-product present during all the
sequence starting from the Suzuki-Miyaura reaction was easily removed.
6
7
8
9
Another challenge for obtaining PBRM was the removal of PPh3O by-product formed during
the Appel reaction. Solvent systems usually used in normal-phase flash chromatography (ex:
EtOAc/hexanes; DCM/MeOH, acetone/hexanes, etc.) lead to laborious separation of PBRM from
PPh3O. These two compounds have a similar mobility on SiO2 and were only separated following
successive flash chromatographic steps on normal phase with a high ratio of silica gel:product (100:1)
and this separation can be only viable in a relatively small scale (< 3 g of PBRM).1a In an ultimate
attempt, we investigated the capacity of unusual solvent systems to provide a clean separation and to
simplify the final purification process of PBRM. As result, a ternary solvent system
(diethylether/DCM/MeOH: 48/48/4) provided an efficient separation of PPh3O from PBRM in a single
chromatography step, being efficient even on a larger scale chromatographic column (> 10 g of
PBRM). A potential alternative to avoid flash chromatography would have be to remove PPh3O by
precipitation with zinc chloride,25 but this interesting new methodology was found only partially
effective in our case.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Analysis of the final PBRM indicated the presence of a small amount (4.6%) of the
corresponding chloride analog. This compound, also present from route C, but in a smaller proportion
(1%), could not be removed either by flash chromatography or by crystallization. However, the
presence of chloride derivative is not a problem at this stage of the proof-of-principle in monkeys,
because its pharmacology is closely related to PBRM, showing lower inhibition of 17β-HSD1 than
PBRM but without estrogenic properties.1a Even if we did not yet identify the exact source of this side
product, we suspected an halogen exchange with the PdCl2 used during the recycling step of the
benzonitrile derivative 15 to PBRM, or it could also be the result of a contamination of the CBr4
reagent with chloride impurities.
Synthetic Route E
Route D has considerably improved the first syntheses of PBRM, but some problems remain,
such as: 1) the number of silica gel filtrations (5 for 6 steps) needed during the sequence of reactions, 2)
the long time needed for the catalytic hydrogenation reaction (9 days), and 3) the unselective final
bromination of primary alcohol vs the carboxamide and 17-OH functions. We thus engaged additional
efforts toward optimization (route E, Scheme 4) to address these weaknesses.
11
ACS Paragon Plus Environment

Organic Process Research & Development
Page 12 of 32
1
2
3
4
5
6
7
8
9
At the first step, the purification of estrone-triflate (1) was performed by trituration with
hexanes to remove trimethylamine salt residues. This procedure gave a comparative high-grade purity
(HPLC purity = 99.1%) to route D, but the yield of the reaction was slightly lower (83% vs 95%). The
condition of the next reaction, based on a palladium catalyzed Suzuki-Miyaura cross-coupling, was
unchanged from route D, but the SiO2 filtration was replaced by a trituration of crude compound 5 with
toluene, which removed a large amount of the E1 impurity, known to be almost insoluble in that
solvent.26 Efforts to recrystallize this crude compound 5 (HPLC purity = 63.3%) were unsuccessful
because this hydrophobic steroid derivative (cLog P = 5.8) was an amorphous solid and found to be
soluble in almost all organic solvents, greatly limiting the possibility of recrystallization. Compound 5
was used as such for the next aldolization reaction, where a smaller amount of 3-formyl carboxamide
was used (1.1 vs 2.0 eq in scale-up 1) to limit the quantity of 3-formyl-carboxamide remaining.
Purification of 6 (HPLC purity = 64.3%) by recrystallization was also found to be unsuccessful,
considering the large solubility of this intermediate in organic solvents. Crude compound 6 was then
submitted to a stereoselective reduction of the C17-carbonyl with NaBH4 at -40 °C rather than at room
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
temperature, to maximize the selectivity formation of 17β-OH. H NMR analysis of the resulting
alcohol 7 showed no trace of 17α-OH reduction product by NMR analysis.18 Once again,
recrystallization attempts of crude compound 7 (HPLC purity = 64.9%) were ineffective, presumably
due to its hydrophobic nature (cLog P = 6.6). Thus, at this stage, all purification attempts from
intermediates 5-7 by recrystallization have failed, probably related to high hydrophobicity, high
solubility in organic solvent and the amorphous solid state of the compounds.
The next objective was to reduce the time necessary for catalytic hydrogenation, which required
over a week to be completed. We thus changed palladium 10% on charcoal (Pd/C) for palladium
hydroxide (20%) on charcoal (Pearlman’s catalyst),27 a catalyst reported as a useful alternative when
classical Pd/C conditions are inefficient.28 The new conditions were beneficial and shortened the
reaction time by more than an half (4 days instead of 9). At this stage, the more polar nature of
compound 8 (cLog P = 4.1) offered a better opportunity for recrystallization of the crude mixture than
previous intermediates (only 72.0% of 8 by HPLC, with 8.3% of 3-OH and 6.8% of 3-H analogs).
Among the solvents screened for recrystallization (ACN, MeOH, EtOH, i-PrOH, 1-butanol, THF,
EtOAc, AcOH, DCE, toluene, and acetone), ACN and acetone offered the best properties. Indeed, the
compound 8 mixture was almost insoluble at room temperature in those two solvents, and completely
soluble around their refluxing temperature. Unfortunately, despite variation of solution concentration
(1-5%), duration and temperature of recrystallization, we never observed the formation of crystals.
12
ACS Paragon Plus Environment

Page 13 of 32
1
Organic Process Research & Development
2
3
4
5
6
7
8
9
Consequently, we proceeded to the bromination of 8, thus generating crude PBRM, and to see if
separation of the impurities could be possible. Using the Appel reaction conditions of route D, we
observed that it was impossible to separate the 3-H analog by-product from PBRM in all eluent systems
tried. Also, recrystallization assays of this mixture of PBRM and 3-H analog in ACN led to crystals,
but co-crystallization with 3-H impurity was obtained (88% PBRM and 6.0% of 3-H analog).
Therefore, two chromatographies using different systems (1: DCM/MeOH and 2: EtOAc 100%) were
needed to remove the two major impurities (3-OH and 3-H analogs) and obtain 8 in a pure form (HPLC
purity = 98.7%). These two chromatographic steps were however harmful on the yield compared to
route D (69 vs 86%).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
O
O
O
O
c
H
a
H
b
H
E1
NH2
H
H
H
H
H
H
BnO
TfO
BnO
(30 g)
5
(No chromatography)
(No chromatography)
6
(No chromatography)
1
29.0 g
HPLC purity: 63.3%
Y = 58%
27.1 g
HPLC purity: 64.3%
Y = 71%
37.1 g
HPLC purity: 99.1%
Y = 83%
d
OH
OH
OH
e
f
O
O
O
H
H
H
NH2
NH2
NH2
H
H
H
H
H
H
HO
Br
BnO
PBRM
(Chromatographies)
(No chromatography)
7
8
24.7 g
HPLC purity: 64.9%
Y = 92%
9.3 g
HPLC purity: 98.7%
Y = 69%
(Chromatography)
9.2 g
HPLC purity: 95.1%
Y = 83%
(Recrystallization)
6.9 g
HPLC purity: 99.7%
Y = 62%
Scheme 4. Gram scale synthesis of PBRM via route E (6 steps) from E1. Reagents and conditions:
(a) Trifluoromethanesulfonic anhydride, TEA, DCM, rt; (b) Potassium (2-
benzyloxyethyl)trifluoroborate, PPh3, PdCl2, Cs2CO3, THF/H2O (9:1), 90 °C; (c) 3-Carboxamide-
benzaldehyde, NaOH (10%), EtOH/DCM (9:1); (d) NaBH4, MeOH, DCM, -40 °C rt, 1 h; (e) H2, 10%
Pd(OH)2/C, EtOH, rt, 4 days; (f) PPh3, TBAI, 1,2-dibromoethane, 60°C.
Finally, to fix the unselective bromination reaction of primary alcohol 8 over carboxamide and
17β-OH functionalities, we investigated several other conditions not yet tried. Our attention was first
focused on the use of trichlorotriazine (TCT), which in addition to NaBr in DMF, was reported for the
bromination of alcohols.29 However, since a residual chloride derivative (20%) was observed in the
reported methodology, we adapted the protocol by running the reaction in anhydrous THF in the
presence of LiBr. The high solubility of this salt in THF increased the equivalents of bromine ions in
13
ACS Paragon Plus Environment

Organic Process Research & Development
Page 14 of 32
1
2
3
4
5
6
7
8
9
solution vs the use of NaBr. As a result, we obtained PBRM with only a slight contamination by the
chloride analog (3%), but in a moderate yield (35%) because we observed 40% of the nitrile by-product
resulting from carboxamide dehydratation. Thus, this reaction led to the same problem observed with
other methodologies used previously, but without presence of PPh3O. Another promising reaction we
tried was the visible-light-mediated photocatalytic alcohol activation with Ru(bpy)3Cl2, which converts
alcohols to the corresponding bromides using CBr4 in DMF.30 Unfortunately, the elimination product
(nitrile) was also observed (35%) in addition to PBRM. In an ultimate attempt, we tried the reaction
developed by Chen et al.31 and reporting a bromination from deoxygenation of alcohols using
tetrabutylammonium iodide and triphenylphosphine in heated dibromoethane. For the first time in all
our bromination assays, we were delighted to see no trace of nitrile by-product, thus leading to a good
yield of PBRM (83%) after purification by chromatography (HPLC purity = 95.1%). Even if PPh3O
was formed during the reaction, this method eliminates the need for a supplementary step to recycle the
nitrile by-product into PBRM and also eliminate the formation of the inseparable chloride analog
observed in route D.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Recrystallization of PBRM
Considering the advantages of recrystallization in the development of a drug candidate,
especially regarding chemistry, manufacturing and control (CMC).32 we were highly interested in
obtaining PBRM in a crystal form. Identification of a valuable solvent of recrystallization was thus
explored by screening the solubility of PBRM in a range of organic solvents, including DCM, DMF,
EtOAc, ACN, diethylether, methanol, toluene, and acetone. Among all these, two solvents (acetone and
ACN) stood out and allowed to recrystallize PBRM with a lower recrystallization yield (42%) from
acetone and with a good recrystallization yield (75%) from ACN. In route E, these later conditions
were used to recrystallize PBRM in 74% yield with a high purity of 99.7%. X-ray analysis of crystals
(Fig. 2 and Supporting Information) confirmed the PBRM molecular structure.
14
ACS Paragon Plus Environment

Page 15 of 32
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Figure 2. Molecular structure of PBRM obtained from X-ray analysis.
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3. CONCLUSIONS
The development of an efficient chemical route that affords a gram quantity of the selective
covalent 17β-HSD1 inhibitor PBRM has been successfully achieved. In fact, the redesign of synthetic
routes A and B previously reported for the preparation of PBRM on small scales led to new routes C, D
and E for gram scale purposes. Route E emerged as the most convenient one taking into account the
synthesis length (6 steps), the overall yield of 13%, and the purity level of final compound obtained
from recrystallization (HPLC purity = 99.7%). This new synthesis represents a significant
improvement compared to the ten steps required in the initial route A and the overall yield of 7%. The
success of route E was based on two strategic modifications of the initial synthetic sequence. First, a
Suzuki-Miyaura reaction was used as a key step to rapidly install the C3 side chain on the estrane
scaffold. Second, the use of a recent methodology for bromination through deoxygenation of alcohol
using a PPh3/BrCH2CH2Br/TBAI reagent system provided selectivity over carboxamide functionality.
Overall, the multigram quantity of PBRM obtained from novel synthetic routes C-E will now enable to
start a proof-of-principle study on primates to assess its therapeutic potential to treat endometriosis. At
the same time, the obtainment of a more reliable synthesis of PBRM supports its further
pharmacological development toward the treatment of estrogen-dependent diseases.
4. EXPERIMENTAL SECTION
General
15
ACS Paragon Plus Environment

Organic Process Research & Development
Page 16 of 32
1
2
3
Reagents and solvents
4
5
6
7
8
9
Estrone was purchased from Zhejiang Xianju Pharmaceuticals Co. (Xiangju, Zhejiang, China).
Triphenylphosphine, oxone and sodium borohydride were purchased from Sigma-Aldrich Canada Ltd.
(Oakville, ON, Canada). Palladium 10% on charcoal was purchased from either BeanTown Chemicals
(Hudson, NH, USA) or Acros Organic (Belgium Town, WI, USA). Trifluoroacetic anhydride and
acetamide were purchased from Oakwood Chemicals (Estill, SC, USA). Carbon tetrabromide was
purchased from either TCI (Portland, OR, USA) or Oakwood Chemicals. Cesium carbonate was
purchased from Chem-Impex (Wood Dale, IL, USA). Ammonium formate and triethylamine was
acquired from Alfa Aesar (Tewksbury, MA, USA). Palladium dichloride was obtained from Aurum
Pharmatech Inc (Franklin Park, NJ, USA). Anhydrous acetonitrile (ACN), dichloromethane (DCM),
diethyl ether, dimethylformamide (DMF), isopropanol, and tetrahydrofuran (THF) were obtained from
Sigma-Aldrich, but the usual solvents ethyl acetate (EtOAc), dichloromethane (DCM), hexanes,
methanol (MeOH), and ethanol (EtOH), were purchased from Fisher Scientific (Montréal, QC,
Canada).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Purification
Thin-layer chromatography (TLC) and flash-column chromatography were performed on 0.20-
mm silica gel 60 F254 plates (E. Merck; Darmstadt, Germany) and with 230-400 mesh ASTM silica
gel 60 (Silicycle, Québec, QC, Canada), respectively. HPLC purities were determined with a Shimadzu
apparatus using a Shimadzu SPD-M20A photodiode array detector, an Alltima HP C18 reversed-phase
column (250 mm x 4.6 mm, 5 µm), and a solvent gradient of MeOH:H2O. Semi-preparative HPLC
purifications were performed to separate chloride analog from PBRM using an Alltima HP C18 (250
mm x 10 mm; 5 µm) with a solvent gradient from 70:30 MeOH/H2O to 100% MeOH on 60 min at a
flow rate of 10 mL/min. The wavelength of the UV detector was selected at maximal compound
absorbance.
Characterization
1
Nuclear magnetic resonance (NMR) spectra were recorded at 400 MHz for H and 100.6 MHz
13
for C on a Bruker Avance 400 digital spectrometer (Billerica, MA, USA). The chemical shifts (δ)
were expressed in ppm and referenced to chloroform (7.26 and 77.0 ppm) or acetone (2.05 and 28.9
ppm) for 1H and 13C NMR, respectively. Infrared (IR) spectra were recorded on a MB 3000 ABB FTIR
spectrometer (Québec, QC, Canada), and only the significant bands are reported in cm-1. Low-
16
ACS Paragon Plus Environment

Page 17 of 32
1
Organic Process Research & Development
2
3
4
5
6
7
8
9
resolution mass spectra (LRMS) were recorded on a Shimadzu apparatus (Kyoto, Japan) equipped with
a turbo ion-spray source. High-resolution mass spectra (HRMS) were provided by Pierre Audet at the
Chemistry Department of Université Laval (Québec, Qc, Canada). Melting points were determined on a
Gallenkamp melting point apparatus (England). The chemical names of the steroid derivatives were
generated with ACD/Laboratories (chemist version) software (Toronto, ON, Canada).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthetic Routes A and B
The experimental details and compound characterization (NMR, IR, MS) for the synthesis of
PBRM according to routes A and B were previously reported.1
Synthetic Route C
17-oxoestra-1,3,5(10)-trien-3-yl trifluoromethanesulfonate (1)
In a three-neck round-bottom flask (3 L) under an argon atmosphere were added estrone (0.22
mol, 60.0 g), DCM (1.5 L) and TEA (0.79 mol, 108 mL). The solution was cooled to 0 °C and
trifluoromethanesulfonic anhydride (0.35 mmol, 49.1 mL) was slowly added over a period of 15 min.
The solution was stirred at 0 °C for 1 h and then at room temperature overnight. The resulting solution
was poured into water, washed with brine, dried over sodium sulfate, filtered and evaporated under
reduced pressure. The crude compound (105 g) was purified using a 2 L filter loaded with 1.5 L of
silica gel which was beforehand solvated with EtOAc/hexanes (1/9). The compound was solubilized in
DCM (150 mL) and gently deposited on the top of silica gel in order to be adsorbed. At this point,
portions of eluent (20 x 450 mL (EtOAc/hexanes: 1/9) were passed through silica gel and harvested
using vacuum Erlenmeyer (500 mL). The fractions 6-26, containing the desired compound as identified
by molybdate indicator on TLC plate (Rf = 0.49; EtOAc/hexanes: 3/7), were combined and evaporated
under reduced pressure to give estrone-triflate (1) (82.7 g, 92%). HPLC purity = 99.1%. NMR data
agree with those reported in literature.33
3-ethenylestra-1,3,5(10)-trien-17-one (2)
In a three-neck round-bottom flask (2 L) under an argon atmosphere were added estrone-triflate
(1) (0.103 mol, 41.4 g), THF (225 mL) and H2O (25 mL). The solution was then degassed for 15 min
17
ACS Paragon Plus Environment

Organic Process Research & Development
Page 18 of 32
1
2
3
4
5
6
7
8
9
with bubbling argon. Potassium vinyltrifluoroborate (0.123 mol, 16.5 g), cesium carbonate (0.31 mol,
100.5 g), triphenylphosphine (6.2 mmol, 1.62 g) and palladium dichloride (2.1 mmol, 0.365 g) were
added under a stream of argon. The solution was degassed with bubbling argon again for 15 min under
vigorous stirring. The heterogeneous mixture was gradually heated to 80 °C and then stirred for 4 days.
The resulting black solution was filtered over a celite pad (1 L) and washed with EtOAc (1 L). The
organic phase was washed with brine, dried over sodium sulfate, filtered and evaporated under reduced
pressure. The crude compound (65.5 g) was purified using a 2 L filter loaded with 1.5 L of silica gel
which was previously solvated with EtOAc/hexanes (3/7). The compound was solubilized in DCM
(150 mL) and gently deposited on the top of the silica gel to be adsorbed. At this point, portions of
eluent (10 x 450 mL of EtOAc/hexanes: 3/7) were passed through silica gel and harvested using
vacuum Erlenmeyers (500 mL). The fractions 2-11, containing the desired compound as identified by
molybdate indicator on TLC plate (Rf = 0.45; EtOAc/hexanes: 3/7), were combined and evaporated
under reduced pressure to give 2 (23.1 g). This reaction was repeated a second time using the same
conditions to give a total quantity of 46.2 g (Yield = 72%). HPLC purity = 88.2%. NMR data agree
with those reported in literature.1b
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3-(oxiran-2-yl)estra-1,3,5(10)-trien-17-one (9)
In a 5 L round-bottom flask, the 3-vinyl-deoxyestrone (2) (0.082 mol, 23.1 g) was dissolved
using a mixture of acetone (500 mL) and ACN (1 L), followed by the addition of an aqueous solution
of NaHCO3 (10%) (1 L). Oxone (0.50 mol, 76.0 g) was then added by small portions over 5 min and
the solution was vigorously stirred for an additional 40 min. The resulting solution was poured into
water (1 L) and extracted with EtOAc (1 L). The organic layer was then washed with brine, dried over
sodium sulfate, filtered and evaporated under reduced pressure. The reaction was repeated a second
time with another 23.1 g portion of compound 2 to give a total quantity of 48.3 g of crude oxirane 9
(Yield = 99%). This compound was used for the next step without further purification. NMR data agree
with those reported in literature.1b
3-(2-hydroxyethyl)estra-1,3,5(10)-trien-17-one (10)
In a 2 L flask, dried beforehand under an argon atmosphere, was added ammonium formate
(0.81 mol, 50.8 g) and palladium on charcoal (10%) (24.0 g). Isopropanol (550 mL) was added to those
solids and the resulting suspension was heated 5 min at 40 °C under an argon atmosphere. The crude
oxirane 9 (24.0 g) was dissolved in isopropanol (350 mL) and the solution was rapidly added using an
18
ACS Paragon Plus Environment

Page 19 of 32
1
Organic Process Research & Development
2
3
4
5
6
7
8
9
addition funnel and stirred at 60 °C for 1 h. The solution was allowed to return at room temperature and
then filtered over celite (1 L). The celite pad was washed with a mixture of DCM/MeOH (4 L) and the
organic solvent was evaporated under reduced pressure. The crude mixture was triturated with DCM,
filtered and evaporated under reduced pressure to give 23.1 g of crude ethyl alcohol derivative 10. This
compound was used for the next step without further purification. The reaction was repeated a second
time with another portion (24.0 g) of compound 9 to give a total quantity of 46.1 g of crude alcohol 10
(Yield = 71%). HPLC purity = 75.0%. NMR data agree with those reported in literature.1b
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3-{(E)-[(16E)-3-(2-hydroxyethyl)-17-oxoestra-1,3,5(10)-16-ylidene]methyl}benzonitrile (12)
To a solution of crude alcohol 10 (23.0 g) in EtOH (595 mL) were added 3-cyano-benzaldehyde
(0.268 mol, 35.2 g) and an aqueous solution of NaOH (10%) (105 mL). The solution was stirred for 15
min at room temperature and then poured into water. The resulting solution was neutralized to pH 7
with an aqueous solution of HCl (10%) and then extracted two times with EtOAc. The organic layer
was washed with brine, dried with sodium sulfate, filtered and evaporated under reduced pressure. The
reaction was repeated a second time with another portion (23.0 g) of alcohol 10. The combined crude
compound of both assays (118 g) was purified using a 2 L filter loaded with 1.5 L of silica gel which
was beforehand solvated with EtOAc/hexanes (3/7). The compound was solubilized in DCM (150 mL)
and gently deposited on the top of silica gel in order to be absorbed. At this point, portions of different
eluents (10 x 450 mL of EtOAc/hexanes (3/7)), 10 x 450 mL of EtOAc/hexanes (4/6) and 5 x 450 mL
of EtOAc/hexanes (1/1)), were passed through silica gel and harvested using vacuum Erlenmeyers (500
mL). The fractions 10-22, containing the desired compound (Rf = 0.26; EtOAc/Hexanes (1/1)) as
identified by molybdate indicator on TLC plate, were combined and evaporated under reduced pressure
to give compound 12 (30.5 g, 53%). IR (NaCl film): 3420 (OH), 2230 (CN), 1720 (C=O), 1628 (C=C).
1H NMR (CDCl3): 1.01 (s, 18-CH3), 1.23-2.61 (unassigned CH and CH2), 2.83 (t, J = 6.5 Hz,
CH2CH2OH), 2.95 (m, 6-CH2 and 1H of 15-CH2), 3.87 (broad t, J = 6.0Hz, CH2CH2OH), 7.00 (s,
4-CH), 7.04 (d, J = 8.0 Hz, 2-CH), 7.26 (d, J = 7.6 Hz, 1-CH), 7.41 (s, 1'-CH), 7.54 (t, J = 7.8 Hz, 5''-
CH), 7.64 (d, J = 7.8 Hz, 6''-CH), 7.74 (d, J = 7.9 Hz, 4''-CH), 7.83 (s, 2''-CH). 13C NMR (CDCl3):
14.4, 25.7, 26.9, 29.1, 29.3, 31.6, 37.8, 38.7, 44.3, 47.9, 48.5, 63.6, 113.1, 118.4, 125.6, 126.5,
129.6, 129.7, 130.4, 132.1, 132.9, 134.4, 136.0, 136.5, 136.9, 137.7, 138.5, 208.7. LRMS for
C28H30NO2 [M + H]+ 412.2 m/z. HPLC purity = 79.0%.
19
ACS Paragon Plus Environment

Organic Process Research & Development
Page 20 of 32
1
2
3
3-{(E)-[(16E,17)-17-hydroxy-3-(2-hydroxyethyl)estra-1,3,5(10)-16ylidene]methyl}benzonitrile (13)
4
5
6
7
8
9
To a solution of compound 12 (0.074 mol, 30.5 g) in a mixture of MeOH/DCM (9/1) (1.3 L) at
0 °C was added NaBH4 (0.25 mol, 9.5 g) by small portions over 15 min. The solution was stirred for 1
h at 0 °C before returning to room temperature and being stirred for an additional 2 h. The resulting
solution was concentrated to a volume of approximately 500 mL and then poured into water to be
extracted four times with EtOAc (4 x 500 mL). The organic layer was washed with brine, dried over
sodium sulfate, filtered and evaporated under reduced pressure to give crude compound 13 (31.5 g,
83%), which was used for the next step without further purification. IR (NaCl film): 3387 (OH), 2230
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
(CN). H NMR (CDCl3): 0.74 (s, 18-CH3), 1.35-2.50 (unassigned CH and CH2), 2.75 (dd, J1 =
5.7 Hz, J2 = 16.9 Hz, 1H of 15-CH2 ), 2.82 (t, J = 6.5 Hz, CH2CH2OH), 2.92 (m, 6-CH2),
3.86 (broad q, J = 6.2Hz, CH2CH2OH), 4.17 (d, J = 8.7 Hz, OH), 4.76 (d, J = 4.2 Hz, OH),
6.54 (d, J = 2.4 Hz, 1'-CH), 6.99 (s, 4-CH), 7.03 (d, J = 8.0 Hz, 2-CH), 7.27 (d, J = 6.1 Hz, 1-CH),
13
7.46 (m, 5''-CH and 6''-CH), 7.59 (d, J = 7.6 Hz, 4''-CH), 7.69 (s, 2''-CH). C NMR (CDCl3): 11.0,
26.0, 27.4, 29.4, 30.6, 36.3, 38.0, 38.6, 43.3, 44.3, 47.6, 63.6, 84.9, 112.5, 119.0, 121.3, 125.5, 126.4,
129.2, 129.7 (2C), 131.2, 132.6, 135.8, 136.7, 138.1, 139.0, 149.0. LRMS for C28H30NO [M - H2O +
H]+ 396.3 m/z. HPLC purity = 80.6%.
3-{[(16,17)-17-hydroxy-3-(2-hydroxyethyl)estra-1,3,5(10)-16-yl]methyl}benzonitrile (14)
To a solution of crude compound 13 (10.5 g) in anhydrous EtOH (500 mL) in a three-neck
round-bottom flask (2 L), was added palladium on charcoal 10% (2.1 g) at room temperature under an
argon atmosphere. The solution was submitted to vacuum and purged with hydrogen for 2 min. This
operation was repeated three times. The three necks of the round-bottom flask were connected with
hydrogen balloons which were periodically refilled, as needed. The solution was stirred for 96 h under
1
a hydrogen atmosphere, until no trace of the olefin signal ( = 6.58 ppm, 1'-CH) was present in H
NMR spectrum. The resulting solution was filtered over a celite pad (500 mL), washed with MeOH (1
L), and evaporated under reduced pressure to give 10.1 g of crude compound 14. In fact, this reaction
was performed in parallel (3 x 10.1 g of 13) in three 2 L three-neck round-bottom flasks to increase the
contact surface, maximize the hydrogen saturation of the solutions and thus accelerate the rate of the
reaction. The combined crude compound (30.3 g) was purified using a 2 L filter loaded with silica gel
(1 L), which was solvated beforehand with EtOAc/hexanes (1/1). The compound was solubilized in
DCM (150 mL) and gently deposited on the top of the silica gel to be adsorbed. Portions of eluent (6 x
450 mL of EtOAc/hexanes (1/1)) were then passed through silica gel and harvested using vacuum
20
ACS Paragon Plus Environment

Page 21 of 32
1
Organic Process Research & Development
2
3
4
5
6
7
8
9
Erlenmeyers (500 mL). The fractions 2-7, containing the desired compound (Rf = 0.25; EtOAc/hexanes
(1/1)) as identified by molybdate indicator on a TLC plate, were combined and evaporated under
reduced pressure to give compound 14 (17.0 g). At this point, we proceeded to a NMR analysis and
observed that compound 14 contained a small proportion of 17-ketone (15%), which was surprisingly
formed during the reaction, and the isomer 16α-benzyl-CN (15%). We thus submitted the crude
mixture to a NaBH4 reduction in MeOH to regenerate the 17β-alcohol from 17-ketone by-product. The
usual work-up for NaBH4 reduction, as reported previously for the synthesis of 13, was performed and
lead to crude compound 14 (16.1 g, 47%). IR (NaCl film): 3410 (OH), 2230 (CN). 1H NMR (CDCl3):
0.87 (s, 18-CH3), 1.05-2.54 (unassigned CH and CH2), 2.84 (m, CH2CH2OH and 6-CH2), 3.15 (d,
J = 8. 8 Hz, 1’-CH2 ), 3.85 (m, CH2CH2OH and OH), 6.95 (s, 4-CH), 7.01 (dd, J1 = 1.6 Hz, J2
= 8.0 Hz, 2-CH), 7.24 (d, J = 7.9 Hz, 1-CH), 7.39 (t, J = 7.6 Hz, 5''-CH), 7.47 (m, 4''-CH and 6''-CH),
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
13
7.52 (s, 2''-CH). C NMR (CDCl3): 12.5, 25.9, 27.3, 29.3, 32.0, 37.2, 37.5, 37.9, 38.5, 41.5, 44.2,
44.3, 48.6, 63.5, 81.8, 112.1, 119.0, 125.4, 126.2, 128.9, 129.4, 129.5, 132.2, 133.4, 135.6, 136.7,
138.3, 143.7. LRMS for C28H32NO [M - H2O + H]+ 398.0 m/z. HPLC purity = 77.0%.
3-{[(16,17)-3-(2-bromoethyl)-17-hydroxyestra-1,3,5(10)-16-yl]methyl}benzonitrile (15)
To a solution of crude compound 14 (16.1 g) in anhydrous THF (750 mL) at 0 °C was added
triphenylphosphine (0.078 mol, 20.4 g) and carbon tetrabromide (0.78 mol, 25.8 g). The solution was
stirred at 0 °C for 5 min and then allowed to return to room temperature and stirred for 70 min. The
resulting solution was poured into water, extracted with EtOAc, washed with brine, dried over sodium
sulfate, filtered and evaporated under reduced pressure to give 54.4 g of crude material. Purification by
flash chromatography using EtOAc/hexanes (2:8) as eluent give compound 15 (14.2 g, 85%). IR (NaCl
1
film): 3479 (OH), 2230 (CN). H NMR (CDCl3): 0.87 (s, 18-CH3), 1.03-2.57 (unassigned CH and
CH2), 2.84 (m, 6-CH2), 3.10 (t, J = 7.8 Hz, CH2CH2Br), 3.15 (dd, J1 = 3. 3 Hz, J2 = 12.2
Hz, 1’-CH2 ), 3.55 (t, J = 7.8 Hz, CH2CH2Br), 3.87 (d, J = 9.3 Hz, OH), 6.93 (s, 4-CH),
6.99 (d, J = 8.0 Hz, 2-CH), 7.24 (d, J = 8.1 Hz, 1-CH), 7.38 (t, J = 7.6 Hz, 5''-CH), 7.48 (m, 4''-CH
13
and 6''-CH), 7.52 (s, 2''-CH). C NMR (CDCl3): 12.6, 26.0, 27.3, 29.4, 32.1, 32.9, 37.3, 37.6, 38.0,
39.0, 41.6, 44.3, 44.4, 48.7, 81.9, 112.3, 119.1, 125.6, 125.9, 129.0, 129.2, 129.5, 132.3, 133.4, 136.2,
136.9, 138.9, 143.7. LRMS for C28H33BrNO [M + H]+ 476.2 and 478.2 m/z. HPLC purity = 84.8%.
3-{[(16β,17β)-3-(2-bromoethyl)-17-hydroxyestra-1,3,5(10)-trien-16-yl]methyl}benzamide (PBRM)
21
ACS Paragon Plus Environment

Organic Process Research & Development
Page 22 of 32
1
2
3
4
5
6
7
8
9
To a solution of compound 15 (0.03 mol, 14.2 g) in a THF/H2O (3:1) mixture (150 mL) was
added acetamide (0.24 mol, 13.9 g) and PdCl2 (2.92 mmol, 518 mg) under an argon atmosphere. The
solution was stirred at room temperature for 24 h and then poured into water and extracted three times
with EtOAc. The combined organic phases were washed with brine, dried over sodium sulfate, filtered
and evaporated under reduced pressure to give 16.4 g of crude material. Purification by flash
chromatography using EtOAc/hexanes (8/2) gave 11.3 g (90%) of PBRM. NMR data agree with those
reported in literature.1a HPLC purity = 97.6%.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthetic Route D
Potassium (2-benzyloxyethyl)trifluoroborate
In a 500 mL flask was successively added bis(pinacolato)biboron (88.5 g, 0.35 mol), lithium
methoxide (17.7 g, 0.47 mol), triphenylphosphine polymer-bound loaded at 1.5 mmol/g (20.0 g, 30
mmol), and CuI (4.42 g, 23 mmol). The flask was purged with argon, followed by the addition of
degassed anhydrous DMF (1.2 L) and benzyl 2-bromoethyl ether (50 g, 0.23 mol). The solution was
bubbled with argon for an additional 10 min, and then vigorously stirred at room temperature for 20 h
under an argon atmosphere. The resulting solution was diluted with DCM, filtered over celite, and
evaporated under reduced pressure. The crude mixture was diluted with diethyl ether and washed with
water. The organic layer was washed with brine, dried over sodium sulfate, and evaporated under
reduced pressure. The resulting crude mixture was treated with a solution of potassium hydrogen
fluoride in THF/H2O (1000/200 mL) and stirred overnight at room temperature. Evaporation of
THF/H2O under reduced pressure gave a crude mixture that was triturated with diethyl ether. The
resulting white solid was washed with diethyl ether and dried under reduced pressure to give 44.8 g
(80%) of potassium (2-benzyloxyethyl)trifluoroborate. NMR data agree with those reported in
literature.22
3-[2-(benzyloxy)ethyl]estra-1,3,5(10)-trien-17-one (5)
In a Schlenk tube (300 mL) was added estrone-triflate (1) (0.075 mol, 30.0 g), cesium carbonate
(0.224 mol, 72.9 g), PdCl2 (1.5 mmol, 265 mg), triphenylphosphine (4.5 mmol, 1.17 g) and potassium
(2-benzyloxyethyl)trifluoroborate (0.09 mol, 21.7 g), which was prepared beforehand following the
Fleury-Brégeot methodology.23 To the solids, under an argon atmostphere was added 100 mL of a
22
ACS Paragon Plus Environment

Page 23 of 32
1
Organic Process Research & Development
2
3
4
5
6
7
8
9
degassed solution of THF/H2O (9/1) and the solution was heated at 90 °C and vigorously stirred for 24
h. The solution was cooled, poured into water, and extracted with EtOAc. The organic phase was
washed with brine, dried with sodium sulfate, filtered and evaporated under reduced pressure. The
reaction was repeated a second time from 30 g of 1. The crude material from both reactions were
combined (60.5 g) and purified using a 2 L filter loaded with silica gel (1.5 L), which was beforehand
solvated with EtOAc/hexanes (1/9). The compound was solubilized in DCM (150 mL) and gently
deposited on the top of silica gel in order to be adsorbed. At this point, portions of eluents (15 x 450
mL of EtOAc/hexanes: 1/9) have been passed through silica gel and harvested using vacuum
Erlenmeyers (500 mL). The fractions 4-15, containing the desired compound (Rf = 0.50;
EtOAc/hexanes (1:9)) as identified by molybdate indicator on TLC plate, were combined and
evaporated under reduced pressure to give compound 5 (39.0 g, 67%), including 15% of 3-deoxy-
estrone, which was inseparable at this stage. HPLC purity = 91.9%. NMR data of both compounds
agree with those reported in literature.1a, 33
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3-{(E)-[(16E)-3-[2-(benzyloxy)ethyl]-17-oxoestra-1,3,5(10)-trien-16-ylidene]methyl}benzamide (6)
To a solution of compound 5 (0.098 mol, 38.0 g) in a mixture of EtOH/DCM (9/1) (1 L) was
added 3-formylbenzamide (0.195 mol, 29.2 g) and an aqueous solution of NaOH (10%) (150 mL). The
solution was stirred at room temperature for 5 h. The resulting solution was then poured into water (2
L) and neutralized at pH 7 with an aqueous solution of HCl (10%) before being extracted with EtOAc
(5 x 500 mL). The combined organic phases were washed with brine, dried over sodium sulfate,
filtered and evaporated under reduced pressure to give 55.9 g of crude compound 6 (Yield = 85%).
This compound was used for the next step without further purification. HPLC purity = 72.5%.
3-{(E)-[(16E,17)-3-[2-(benzyloxy)ethyl]-17-hydroxyestra-1,3,5(10)-trien-16
ylidene]methyl}benzamide (7)
To a solution of crude compound 6 (55.9 g) in a mixture of MeOH/DCM (9/1) (1 L) was added
NaBH4 (0.45 mol, 16.9 g) at 0 °C. The solution was stirred for 30 min at 0 °C and allowed to return to
room temperature and stirred for an additional 2 h. The resulting solution was concentrated under
reduced pressure to a volume of about 500 mL and then poured into water (2 L). The aqueous solution
was extracted with EtOAc (3 x 400 mL), washed with brine, dried over sodium sulfate, filtered and
evaporated under reduced pressure to give 56.0 g of crude compound. The crude material was purified
23
ACS Paragon Plus Environment

Organic Process Research & Development
Page 24 of 32
1
2
3
4
5
6
7
8
9
using a 2 L filter loaded with silica gel (1.5 L), which was solvated beforehand with EtOAc/hexanes
(7/3). The compound was solubilized in DCM (200 mL) and gently deposited on the top of the silica
gel to be adsorbed. At this point, portions of eluent (30 x 450 mL of EtOAc/hexanes (7/3)) have been
passed through silica gel and harvested using vacuum Erlenmeyers (500 mL). The fractions 8-27,
containing the desired compound (Rf = 0.17; EtOAc/hexanes (7/3)) as identified by molybdate
indicator on TLC plate, were combined and evaporated under reduced pressure to give compound 7
(46.4 g, 80%). The corresponding 3-deoxy-estrone derivative (15%) was present, but inseparable at this
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
stage. IR (NaCl film): 3366 and 3209 (OH and NH2), 1666 (C=O, amide). H NMR (CDCl3): 0.74 (s,
18-CH3), 1.35-2.45 (unassigned CH and CH2), 2.79 (dd, J1 = 5.8 Hz, J2 = 17.0 Hz, 1H of
15-CH2 ), 2.88 (broad t, J = 7.2 Hz, CH2CH2O and 6-CH2), 3.69 (t, J = 7.2 Hz, CH2CH2OBn),
4.15 (s, 17α-CH), 4.54 (s, OCH2Ph), 5.80-6.20 (2 broad s, CONH2), 6.59 (d, J = 2.0 Hz,
1'-CH), 6.97 (s, 4-CH), 7.02 (d, J = 7.9 Hz, 2-CH), 7.23 (d, J = 8.0 Hz, 1-CH), 7.32 (m, OCH2Ph),
7.42 (t, J = 7.7 Hz, 5''-CH), 7.56 (d, J = 7.8 Hz, 6''-CH), 7.61 (d, J = 7.7 Hz, 4''-CH), 7.87 (s, 2''-CH).
13C NMR (CDCl3): 11.1, 26.1, 27.5, 29.4, 30.6, 35.8, 36.3, 38.1, 43.3, 44.4, 47.7, 71.3, 72.9, 84.9,
122.3, 124.9, 125.3, 126.3, 127.4, 127.5, 127.6 (2C), 128.3 (3C), 128.6, 129.5, 131.5, 133.5, 136.2,
136.4, 137.9, 138.4, 147.6, 169.6. LRMS for C35H40NO3 [M + H]+ 522.3 m/z. HPLC purity = 69.9%.
3-{[(16β,17β)-17-hydroxy-3-(2-hydroxyethyl)estra-1,3,5(10)-trien-16-yl]methyl}benzamide (8)
In a 2 L three-neck round-bottom flask, compound 7 (15.4 g) obtained above was dissolved in
anhydrous EtOH (1 L) and palladium on charcoal 10% (3.2 g) was added under an argon atmosphere.
The solution was submitted to vacuum and then purged with hydrogen for 2 min. This operation was
repeated three times. The three necks of the round-bottom flask were connected with hydrogen
balloons which were periodically refilled, as needed. The solution was then stirred for 9 days under
hydrogen atmosphere, until no trace of the olefin signal ( = 6.58 ppm, 1'-CH) was present in 1H NMR
spectrum. The resulting solution was filtered over a celite pad (500 mL), washed with MeOH (1 L), and
evaporated under reduced pressure to give crude compound 8. In fact, this reaction was performed in
parallel (3 x 15.4 g of 7) in three 2 L three-neck round-bottom flasks to increase the contact surface,
maximize the hydrogen saturation of the solutions, and thus accelerate reaction rate. Crude compound
from the three reactions were combined. At this point, we proceeded to a NMR analysis and observed
that crude compound 8 contained a small proportion of 17-ketone (15%). We thus submitted the crude
compound to a NaBH4 reduction in MeOH in order to regenerate the 17β-alcohol from 17-ketone side
24
ACS Paragon Plus Environment

Page 25 of 32
1
Organic Process Research & Development
2
3
4
5
6
7
8
9
product. The usual work-up for NaBH4 reduction was performed as reported previously for the
synthesis of 13. The crude material (40.6 g) was then divided in two portions (20.3 g) and purified by
flash chromatography using EtOAc/hexanes (8/2) to give a total of 29.4 g (86%) of compound 8,
including 7% of the 16α-benzylbenzamide isomer which was found inseparable at this stage. NMR data
agree with those reported in literature.1a HPLC purity = 81.6%.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3-{[(16β,17β)-3-(2-bromoethyl)-17-hydroxyestra-1,3,5(10)-16-yl]methyl}benzamide (PBRM)
To a solution of compound 8 (0.012 mol, 5.0 g) obtained above in anhydrous THF (50 mL) at 0
°C was added triphenylphosphine (0.023 mol, 6.06 g) and carbon tetrabromide (0.023 mol, 7.68 g). The
solution was stirred at 0 °C for 10 min, and then allowed to return to room temperature and stirred for 1
h. The resulting mixture was poured into water and extracted three times with EtOAc. The organic
phase was washed with brine, dried over sodium sulfate, filtered, and evaporated under reduced
pressure to give 15.6 g of crude material. Purification by flash chromatography using diethyl
ether/DCM/MeOH (48/48/4) as eluent provided PBRM (2.5 g) and a less polar mixed fraction (3.0 g)
containing in part the nitrile derivative 15, a by-product coming from dehydration of carboxamide, and
CBr4/CHBr3 by-product. To regenerate PBRM from 15, this mixed fraction (3.0 g) was submitted
overnight to hydration conditions using PdCl2 (0.92 mmol, 160 mg) and acetamide (74.4 mmol, 4.4 g)
in a mixture of THF/H2O (3/1) (50 mL). Purification by flash chromatography using
diethylether/DCM/MeOH (48:48:4) as eluent provided an additional 1.4 g of PBRM, which was
combined with the 2.5 g obtained previously to give a total of 3.9 g (68%) of PBRM. The entire
process was repeated with two larger portions of compound 8 (12.25 g) and gave reproducible yields
(9.5 g and 9.8 g) of PBRM. NMR data agree with those reported in literature.1a The PBRM portions
originating from the bromination assays performed (3.9 + 9.5 + 9.8 = 23.5 g (81%)) were combined,
showing an HPLC purity of 95.2 %. An impurity representing 4.6% was identified as (3-{[(16β,17β)-3-
(2-chloroethyl)-17-hydroxyestra-1,3,5(10)-16-yl]methyl}benzamide after isolation from a semi-
preparative HPLC chromatography on a small portion (200 mg) of the final PBRM batch. NMR data of
this chloro derivative agree with those reported in literature.1b
Synthetic Route E
17-oxoestra-1,3,5(10)-trien-3-yl trifluoromethanesulfonate (1)
25
ACS Paragon Plus Environment

Organic Process Research & Development
Page 26 of 32
1
2
3
4
5
6
7
8
9
In a three-neck round-bottom flask (3 L) under an argon atmosphere were added estrone (0.11
mol, 30.0 g), DCM (750 mL) and TEA (0.40 mol, 54 mL). The solution was cooled to 0 °C and
trifluoromethanesulfonic anhydride (0.18 mmol, 24.6 mL) was slowly added over a period of 15 min.
The solution was stirred at 0 °C for 1 h and then at room temperature overnight. The resulting solution
was poured into water, washed with brine, dried over sodium sulfate, filtered and evaporated under
reduced pressure. The crude compound (51 g) was purified by successive washing with hexanes (5 x 1
L). The combined solutions were evaporated under reduced pressure to give estrone-triflate (1) (37.1 g,
83%). NMR data were identical to those reported for 1 in route C.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3-[2-(benzyloxy)ethyl]estra-1,3,5(10)-trien-17-one (5)
In a Schlenk tube (300 mL) was added estrone-triflate (1) (0.091 mol, 37.0 g), cesium carbonate
(0.28 mol, 89.9 g), PdCl2 (1.8 mmol, 326 mg), triphenylphosphine (7.4 mmol, 1.92 g) and potassium
(2-benzyloxyethyl)trifluoroborate (0.11 mol, 29.0 g), which was prepared beforehand following the
Fleury-Brégeot methodology.22 To the solids, under an argon atmosphere was added 120 mL of a
degassed solution of THF/H2O (9/1) and the solution was heated at 90 °C and vigorously stirred for 24
h. The solution was cooled, poured into water, and extracted with EtOAc. The organic phase was
washed with brine, dried with sodium sulfate, filtered and evaporated under reduced pressure to give
29.0 g (52%). HPLC purity = 63.3%. This compound was used for the next step without further
purification. NMR data were consistent to those reported for compound 5 in route D.
3-{(E)-[(16E)-3-[2-(benzyloxy)ethyl]-17-oxoestra-1,3,5(10)-trien-16-ylidene]methyl}benzamide (6)
To a solution of compound 5 (74.6 mmol, 29.0 g) in a mixture of EtOH/DCM (85/15) (595 mL)
was added 3-formylbenzamide (82 mmol, 12.3 g) and an aqueous solution of NaOH (10%) (105 mL).
The solution was stirred at room temperature for 5 h. The resulting solution was then poured into water
(2 L) and neutralized at pH 7 with an aqueous solution of HCl (10%) before being extracted with
EtOAc (5 x 500 mL). The combined organic phases were washed with brine, dried over sodium sulfate,
filtered and evaporated under reduced pressure to give 27.1 g of crude compound 6 (Yield = 71%).
This compound was used for the next step without further purification. HPLC purity = 64.9%. NMR
data were consistent to those reported for compound 6 in route D.
26
ACS Paragon Plus Environment

Page 27 of 32
Organic Process Research & Development
1
2
3
3-{(E)-[(16E,17)-3-[2-(benzyloxy)ethyl]-17-hydroxyestra-1,3,5(10)-trien-16
ylidene]methyl}benzamide (7)
4
5
6
7
8
9
To a solution of crude compound 6 (27.1 g) in a mixture of MeOH/DCM (9/1) (600 mL) at -40
°C was added NaBH4 (104 mmol, 3.95 g) at 0 °C. The solution was stirred at -40 °C for 2 h. The
resulting solution was concentrated under reduced pressure to a volume of about 500 mL and then
poured into water (2 L). The aqueous solution was extracted with EtOAc (3 x 400 mL), washed with
brine, dried over sodium sulfate, filtered and evaporated under reduced pressure to give 24.7 g of crude
compound. This compound was used for the next step without further purification. HPLC purity =
69.9%. NMR data were identical to those reported for compound 7 in route D.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3-{[(16β,17β)-17-hydroxy-3-(2-hydroxyethyl)estra-1,3,5(10)-trien-16-yl]methyl}benzamide (8)
In a 2 L three-neck round-bottom flask, compound 7 (24.7 g) obtained above was dissolved in
anhydrous EtOH (500 mL) and palladium hydroxide (20% on charcoal) (3.0 g) was added under an
argon atmosphere. The solution was submitted to vacuum and then purged with hydrogen for 2 min.
This operation was repeated three times. The three necks of the round-bottom flask were connected
with hydrogen balloons which were periodically refilled, as needed. The solution was then stirred for 4
days under hydrogen atmosphere, until no trace of the olefin signal ( = 6.58 ppm, 1'-CH) was present
1
in H NMR spectrum. The resulting solution was filtered over a celite pad (250 mL), washed with
MeOH (500 L), and evaporated under reduced pressure to give crude compound 8. At this point, we
proceeded to a NMR analysis and observed that crude compound 8 contained a small proportion of 17-
ketone (15%). We thus submitted the crude compound to a NaBH4 reduction in MeOH in order to
regenerate the 17β-alcohol from 17-ketone side product. The usual work-up for NaBH4 reduction was
performed as reported previously for the synthesis of 13. The crude material (20.2 g) purified by flash
chromatography using DCM/MeOH (9/1) to give a total of 13.4 g of compound 8, but still containing
3-OH by-product (3-{[(16β,17β)-3,17-dihydroxyestra-1,3,5(10)-trien-16-yl]methyl}benzamide). A
second flash chromatography using EtOAc provide 9.3 g of pure compound 8. NMR data were
consistent to those reported for compound 8 in route D. HPLC purity = 98.7%.
3-{[(16β,17β)-3-(2-bromoethyl)-17-hydroxyestra-1,3,5(10)-16-yl]methyl}benzamide (PBRM)
27
ACS Paragon Plus Environment

Organic Process Research & Development
Page 28 of 32
1
2
3
4
5
6
7
8
9
To compound 8 (21.4 mmol, 9.3 g) in a 500 mL flask was added triphenylphosphine (38.5
mmol, 10.1 g), tetrabutylammonium iodide (38.5 mmol, 14.2 g) and 1,2-dibromoethane (250 mL). The
solution was stirred at 60 °C for 1 h. The resulting solution was evaporated under reduced pressure.
Purification by flash chromatography using diethyl ether/DCM/MeOH (49:49:2) provided PBRM (8.8
g, 83%) showing an HPLC purity of 95.1%. Recrystallization from acetonitrile provided 6.9 g (74% of
recrystallization yield) of PBRM with a high HPLC purity (99.7%). The impurity (0.26%)
corresponding to PBRM-O, oxidized form at position C17, and identified by LC-MS from previous
data.6 NMR data were consistent to those reported for PBRM in previous routes.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
PBRM recrystallization assays and X-ray analysis
Recrystallization in acetonitrile
To 200 mg of PBRM at room temperature was added 10 mL of ACN. The suspension was
heated at reflux for 5 min and cooled to room temperature. The solution was left to rest at room
temperature for 45 min and cooled at 4 °C overnight. Crystals were filtered, washed with cold ACN
and dried over vacuum to give 151 mg of PBRM (75% of recovery). M.p. = 183-184°C. 1H NMR was
identical to the starting PBRM.
Recrystallization in acetone
To 100 mg of PBRM at room temperature was added 5 mL of acetone. The suspension was
heated at reflux for 5 min and cooled to room temperature. The solution was left to rest at room
temperature for 3 days and cooled at 4 °C for a week. Crystals were filtered, washed with cold acetone,
1
and dried over vacuum to give 42 mg of PBRM (42% of recovery). M.p. = 183-184°C. H NMR was
identical to the starting PBRM.
X-ray analysis
Two PBRM crystals were analyzed by the X-ray Diffraction Laboratory (Department of
Chemistry, Université de Montréal, Montréal, QC, Canada) and details of x-ray analysis and data were
provided in the Supporting Information.
ASSOCIATED CONTENT
28
ACS Paragon Plus Environment

Page 29 of 32
1
Organic Process Research & Development
2
3
Supporting Information
4
5
6
7
8
1) 1H NMR spectra of PBRM from routes B (batch 2), C, D, and E. 2) HPLC chromatograms of PBRM
from routes B (batch 2), C, D, and E. 3) X-ray analysis of PBRM from route D. This material is
available free of charge via internet at http://pubs.acs.org.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
AUTHOR INFORMATION
Corresponding Author
Phone: 1-418-654-2296. Fax: 1-418-654-2298. E-Mail: Donald.poirier@crchul.ulaval.ca
ORCID
Donald Poirier: 0000-0002-7751-3184
René Maltais: 0000-0002-0394-1653
Author Contributions
RM performed the laboratory work and participated in the writing of the manuscript; DP contributed to
the writing. Both have approved the final version of the manuscript.
Funding
We thank the Canadian Institutes of Health Research (CIHR) for an operating grant.
Notes
The authors declare the following competing financial interest (s): RM and DP have ownership interest
in a patent application.
ACKNOWLEDGMENTS
We are grateful to Mr. Thierry Maris (Université de Montréal, Montréal, QC, Canada) for x-ray
analysis, and to Mrs. Micheline Harvey for careful reading of this manuscript.
REFERENCES
(1) (a) Maltais, R.; Ayan, D.; Poirier, D., Crucial role of 3-bromoethyl in removing the estrogenic
activity of 17β-HSD1 inhibitor 16β-(m-carbamoylbenzyl)estradiol. ACS Med. Chem. Lett. 2011, 2,
29
ACS Paragon Plus Environment