Esters of levonorgestrel and etonogestrel intended as single, subcutaneous-injection, long-lasting contraceptives

Article abstract of DOI:10.1016/j.steroids.2018.07.010

An effort with the goal of discovering single-dose, long-lasting (>6 months) injectable contraceptives began using levonorgestrel (LNG)-17-β esters linked to a sulfonamide function purposed as human carbonic anhydrase II (hCA 2) ligands. One single analog from this first series showed noticeably superior anti-ovulatory activity in murine models, and a subsequent structure-activity relationship (SAR, the relationship between a compound's molecular structure and its biological activity) study based on this compound identified a LNG-phenoxyacetic acid ester analog exhibiting longer anti-ovulatory properties using the murine model at 2 and 4 mg dose than medroxyprogesterone acetate (MPA). The same ester function linked to etonogestrel (ENG) furnished a compound which inhibited ovulation at 2 mg for 60 days, the longest duration of all compounds tested at these doses. By comparison, MPA at the same dose inhibited ovulation for 32 days.

Full text of DOI:10.1016/j.steroids.2018.07.010

Accepted Manuscript
Esters of levonorgestrel and etonogestrel intended as single, subcutaneous-in-
jection, long-lasting contraceptives
Frederick A Meece, Gulzar Ahmed, Hareesh Nair, Bindu Santhamma,
Rajeshwar R Tekmal, Chumang Zhao, Nicole E Pollok, Julia Lara, Ze'ev
Shaked, Klaus Nickisch
PII:
S0039-128X(18)30133-8
https://doi.org/10.1016/j.steroids.2018.07.010
STE 8294
DOI:
Reference:
Steroids
To appear in:
Received Date:
Revised Date:
Accepted Date:
4 May 2018
18 July 2018
25 July 2018
Please cite this article as: Meece, F.A., Ahmed, G., Nair, H., Santhamma, B., Tekmal, R.R., Zhao, C., Pollok, N.E.,
Lara, J., Shaked, Z., Nickisch, K., Esters of levonorgestrel and etonogestrel intended as single, subcutaneous-
injection, long-lasting contraceptives, Steroids (2018), doi: https://doi.org/10.1016/j.steroids.2018.07.010
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Esters of levonorgestrel and etonogestrel intended as single, subcutaneous-
injection, long-lasting contraceptives
Frederick A Meecea, Gulzar Ahmeda, Hareesh Naira, Bindu Santhammaa, Rajeshwar R Tekmalb,
Chumang Zhaoa, Nicole E Polloka, Julia Laraa, Ze’ev Shakeda, Klaus Nickischa
aEvestra, Inc, 6410 Tri County Pkwy, Schertz TX, 78154
bUniversity of Texas Health Science Center San Antonio, 7703 Floyd Curl Dr., San Antonio, TX, 78229
Abstract
An effort with the goal of discovering single-dose, long-lasting (>6 months) injectable contraceptives began
using levonorgestrel (LNG)-17- esters linked to a sulfonamide function purposed as human carbonic anhydrase
II (hCA 2) ligands. One single analog from this first series showed noticeably superior anti-ovulatory activity in
murine models, and a subsequent structure-activity relationship (SAR, the relationship between a compound’s
molecular structure and its biological activity) study based on this compound identified a LNG-phenoxyacetic
acid ester analog exhibiting longer anti-ovulatory properties using the murine model at 2 and 4 mg dose than
medroxyprogesterone acetate (MPA). The same ester function linked to etonogestrel (ENG) furnished a
compound which inhibited ovulation at 2 mg for 60 days, the longest duration of all compounds tested at these
doses. By comparison, MPA at the same dose inhibited ovulation for 32 days.
Keywords: Injectable contraception, MPA, levonorgestrel, etonogestrel, prodrug
1. Introduction
As of 2011, forty million women worldwide used injectable contraception [1] which is a preferred delivery
system of contraceptive agents especially in southeastern Asian countries such as Thailand, the Philippines, and
sub-Saharan Africa. While the method remains popular with users, one obstacle to regular use is availability, due
to regional product unavailability and/or no clinical access, among other reasons [2]. The idea of injectable
contraception lasting over six months which reduces side effects and improves safety profiles could in theory
prove valuable in negating the availability issues, while benefiting women and health care systems globally.
MPA a widely used injectable contraceptive, and is commercially available as two different products- one is a
intramuscular injection containing 150 mg of MPA and the other a sub-cutaneous injection containing 104 mg of
MPA. Depo-Sub Q Provera 104 is a sub-cutaneous single-injectable formulation of the progestin
medroxyprogesterone acetate (MPA) which among other uses, provides a contraceptive effect for 13 weeks post
injection by preventing ovulation at a typical failure rate of 3% [3,4]. The approved dose for Depo Provera is 150
mg in 1 ml for intramuscular injection and 104 mg in 0.65 ml for sub-cutaneous injection-formulation [5]. There

are benefits from a solely progestin contraceptive product like Depo-Provera, including no increased risk of deep
vein thrombosis (DVT), pulmonary embolism (PE), stroke, or myocardial infarction as with estrogen-based
contraceptives, and a decreased rate of endometrial cancer and a decrease of symptoms associated with
endometriosis [6]. There are side effects caused by MPA, including weight gain, mood swings, acne and
hirsutism. In 2004, a black box warning was issued by the United States Food and Drug Administration
indicating loss of bone mineral density and recommended that patients be prescribed the drug for no longer than
two years due to this risk [7].
Levonorgestrel (LNG) is currently a very commonly prescribed progestin used in contraception. Most of the
reported side effects of MPA are not (or to a significantly lower extent) observed in women taking LNG. A
longer acting injectable contraceptive product based on LNG or ENG could result greater convenience for women
who seek injectable contraception. Risks of progestin-only contraception include possible problems with
bleeding, possibly inducing amenorrhea. This might be curtailed by a rhythmic and steady, low-dose of progestin
which would still enable contraception, provided by the single-dose, long-lasting depot of properly formulated
progestin. It would possibly be therefore very desirable to develop a long-term injectable contraceptive product
based on LNG [8].
In 1975, the World Health Organization set forth an initiative to find novel, long-lasting contraceptive agents.
Several groups discovered and published findings involving long-lasting contraceptive agents which were in
essence 17- esters of LNG and to a lesser extent, norethisterone [9,10,11]. These esters were suspended in
various media and then introduced into rats via sub-cutaneous injections. One in particular showed superior anti-
ovulation activity in murine models, which was the butyrate ester of LNG, known as levonorgestrel butyrate (LB).
This compound inhibited ovulation in the murine model for 32 days at a 2 mg dose [10]. LB underwent several
clinical trials but to this date has not been marketed [12]. A more recent study [13] found that a higher body mass
index of women resulted in earlier return to ovulation, suggesting possible inconsistencies in efficacy from
woman to woman.
Recently, our group published work regarding steroidal esters linked to sulfonamides through amino acid linkers
to improve oral bioavailability and to prevent first-pass liver metabolism [14]. These pro-drugs act as hCA 2
ligands and accumulate in red blood cells, which in effect protects the compounds from being metabolized by the
liver. Theoretically this has two main positive effects, one in being that this makes the steroids safer by curtailing
liver toxicity from metabolism and the other being a marked improvement in oral bioavailability as the compound
will be delivered straight to the vasculature upon intestinal absorption. Cleavage of the ester by endogenous
esterases then frees the active molecule to then travel on to its respective target. A hypothesis was formed
involving this same concept being applied then to sub-cutaneous injections of sulfonamide-linked esters of LNG;
from the injection site slow release of prodrug into sub-cutaneous tissue followed by uptake into capillaries and
erythrocytes laden with hCA 2 would result in ideal systemic concentration, assuming a steady state over time.

Herein, we report our work and the results regarding ovulation inhibition by a series of LNG esters and one ENG
ester.
2. Experimental Methods
2.1 General Methods
Nuclear magnetic resonance (NMR) spectra were recorded on a Bruker ARX (300 MHz) spectrometer as
deuterochloroform (CDCl3) or DMSO-d6 solutions using tetramethylsilane (TMS) as an internal standard ( = 0)
unless noted otherwise. High-performance liquid chromatography (HPLC) was performed on a Waters Alliance
2695 with a photodiode array 2996 system using a Waters XTerra® RP18 column (3.5 m, 4.6 x 150 mm). All
compounds used in anti-ovulation studies had purity than > 95% at the MAX wavelength as determined by this
system. Medium-pressure liquid chromatography (MPLC) was performed using a Isolera One from Biotage
USA® (Charlotte, NC) utilizing SiliaFlash® P60 40 – 63 m silica gel from Silicycle (Quebec, Ontario). Thin-
layer chromatography (TLC) analysis were carried out on silica gel GF (Analtech) glass plates (2.5 cm x 10 cm
with 250 m layer, pre-scored). Most chemicals and solvents were analytical grade and used without further
purification unless otherwise noted. LNG was supplied as a gift from Wyeth (DE) and ENG was purchased from
Nexconn (PRC). A Mini-Mill Pulverisette 23 purchased from Fritsch was used to mill the solid compounds for
formulation. Female Sprague-Dawley rats were purchased from Charles River (Wilmington, MA). Clean tip
sponge swabs were purchased from Texwipe (Kernsville, NC), and vaginal smears were visualized using a Nikon
Eclipse Ti-S microscope.
2.2 Chemistry
3 (bold numbers designate experimental compounds). (13S,17R)-13-ethyl-17-ethynyl-3-oxo-2,3,6,7,8,9,10,11,12,
13,14,15,16,17-tetradecahydro-1H-cyclopenta[a]phenanthren-17-yl(3-sulfamoylphenoxy)acetate (General method
#1): To an oven-dried round bottom flask was added 2 (973 mg, 2.5 mmol), 3-hydroxybenzenesulfonamide, and
4 molecular sieve. Anhydrous DMF was added (5 ml) and then cesium carbonate (814 mg, 2.5 mmol). The
mixture was allowed to stir for 24 hours, after which time an additional 0.5 equivalent of base and 0.3 equivalent
of phenol was added to fully convert the starting material by allowing to stir an additional 16 hours. The mixture
was then diluted with ice-cold saturated sodium bicarbonate and the resulting solids collected by vacuum
filtration, washed with water and allowed to air-dry. The product was isolated by subjecting the crude solid to
1
flash chromatography using a 2 to 10% gradient of acetone in DCM. H NMR (300 MHz, DMSO-d6) δ 7.52 –
7.31 (m, 4H), 7.13 (ddd, J = 7.8, 2.4, 0.9 Hz, 1H), 5.73 (s, 1H), 4.83 (s, 2H), 3.65 (s, 1H). IR (cm-1): 3324, 3294,
3231, 2934, 2866, 1758, 1742, 1658, 1343, 1209, 1158, 793, 680. MP 166.5 – 169o C.

5. (13S,17R)-13-ethyl-17-ethynyl-17-hydroxy-2,7,8,9,10,11,12, 13,14,15,16,17-dodecahydro-1H-
cyclopenta[a]phenanthren-3-yl 4-sulfamoylbenzoate. To a round bottom flask was added 1 (2.4 g, 7.7 mmol) and
cesium carbonate (3.0 g, 9.2 mmol). Anhydrous toluene (100 ml) was added, and a Dean-Stark trap was placed
atop the flask and 75 ml of toluene was removed from the trap. 4-sulfamoylbenzoyl chloride was then added (2.0
g, 9.2 mmol) and the mixture was allowed to reflux for 24 hours, upon which time the mixture was allowed to
cool and then washed with 2 M HCl. The layers were separated and extracted with ethyl acetate. The combined
organic layers were washed with saturated sodium bicarbonate, brine, and dried over sodium sulfate. The crude
was then subjected to flash chromatography using 10% acetone in DCM. A further reverse-phase medium
pressure liquid chromatography procedure was used (45-55% gradient of acetonitrile) to afford the purified
product as a foam; 1H NMR (300 MHz, DMSO-d6) δ 8.18 (d, J = 8.7 Hz, 2H), 7.97 (d, J = 8.7 Hz, 2H), 7.58 (s,
2H), 5.96 (s, 1H), 5.54 (bs, 1H), 5.29 (s, 1H), 0.919 (t, J = 6.9 Hz, 1H).
17. (13S,17R)-13-ethyl-17-ethynyl-3-oxo-2,3,6,7,8,9,10,11,12, 13,14,15,16,17-tetradecahydro-1H-
cyclopenta[a]phenanthren-17-yl 4-sulfamoylbenzoate. Formation of 4-{bis[(2,4-dimethoxy-
phenyl)methyl]sulfamoyl}benzoic acid: A round bottom flask was charged with 4.87 g (15.5 mmol) of Bis(2,4-
Dimethoxybenzyl)amine and TEA (3.9 ml, 28.1 mmol) and dissolved in THF (30 ml) and chilled to 0o C. Then
methyl 4-(chlorosulfonyl)benzoate (3.3 g, 14.1 mmol) in 45 ml THF was added dropwise to the amine solution.
The mixture was then allowed to gradually warm to room temperature while stirring overnight. The next day the
mixture was evaporated onto silica and subjected to flash chromatography using a 5 to 20% gradient of EtOAc in
DCM to obtain 5.94 g (82%) of the intermediate 4-{bis[(2,4-dimethoxyphenyl)methyl]sulfamoyl}benzoic acid
methyl ester; 1H NMR (300 MHz, CDCl3)  8.03 (d, J = 8.9 Hz, 2H), 7.70 (d, J = 8.5 Hz, 2H), 7.15 (d, J = 8.9 Hz,
2H), 6.38 (dd, J = 8.4, 2.6 Hz, 2H), 6.25 (d, J = 1.2 Hz, 2H), 4.41 (s, 4H), 3.96 (s, 3H), 3.78 (s, 6H), 3.59 (s, 6H).
The methyl ester was saponified by treating with 23 ml of 5M NaOH in 100 ml THF at reflux for 4 h to afford the
acid (isolated by removal of THF, then treating with sodium hydrogen sulfate until pH was 3 and extracting with
ethyl acetate (5.77 g recovered, 99%). 3.28 g (6.5 mmol) of this material was then placed into a round bottom
flask along with 1 (511 mg, 1.64 mmol) and DMAP (0.2 g, 1.64 mmol). The materials were dissolved in DCM
(15 ml) and then DIC added (1.0 ml, 6.5 mmol) once a homogenous solution was observed and the mixture
allowed to stir overnight. The next day the solids were removed via filtration and the crude material subjected to
flash chromatography using 3-5% acetone in DCM to give 1.16 g of the intermediate (89%). This was then
suspended in DCM (11 ml) and chilled to 0o C before addition of TFA (5.5 ml) and allowed to stir one hour at
which time TLC and HPLC indicated the conversion complete. The volatiles were then removed under vacuum
and the solid residue washed with saturated sodium bicarbonate. The crude solids were then subjected to flash
chromatography using 5-10% acetone in DCM gradient to afford the title compound in 72% yield; 1H NMR (300
MHz, DMSO-d6) δ 8.08 (d, J = 7.8 Hz, 2H), 7.96 (d, J = 7.5 Hz, 2H), 7.58 (s, 2H), 5.74 (s, 1H), 3.70 (s, 1H). IR
(cm-1): 3264, 2940, 2873, 1723, 1651, 1341, 1273, 1166, 1091, 767. MP 173.5 – 178o C.

18. (13S,17R)-13-ethyl-17-ethynyl-3-oxo-2,3,6,7,8,9,10,11,12, 13,14,15,16,17-tetradecahydro-1H-
cyclopenta[a]phenanthren-17-yl sulfamoylacetate
Formation of {bis[(2,4-dimethoxyphenyl)methyl]sulfamoyl}-acetic acid- A round bottom flask was charged with
chloro-sulfonylacetyl choride (1.2 ml, 11.3 mmol), dissolved in DCM (40 ml) and then anhydrous methanol was
added (0.46 ml, 11.3 mmol) and allowed to stir for 1 h. Then 3.74 g (11.9 mmol) of Bis(2,4-
Dimethoxybenzyl)amine and TEA (3.15 ml, 22.6 mmol) were added to the reaction mixture and allowed to stir
overnight while gradually warming to room temperature, following which time the mixture was rotovapped onto
silica and subjected to flash chromatography using ethyl acetate (0 to 8%) in DCM to afford the intermediate as
the methyl ester (1.95 g, 38% yield). The ester was saponified using LiOH-H2O (5M solution, ten equivalents) in
THF at room temp, then the THF removed, and the residue treated with sodium hydrogen sulfate to neutralize and
extracted into ethyl acetate, giving the acid in quantitative yield; 1H NMR (300 MHz, DMSO-d6)  7.04 (d, J =
8.6 Hz, 2H), 6.50 – 6.43 (m, 4H), 4.25 (s, 4H), 4.07 (s, 2H), 3.74 (s, 6H), 3.70 (s, 6H). 1.95 g (4.3 mmol) of this
material was then placed into a round bottom flask along with 1 (336 mg, 1.08 mmol) and DMAP (0.131 g, 1.08
mmol). The materials were dissolved in DCM (15 ml) and then DIC added (0.66 ml, 4.3 mmol) once a
homogenous solution was observed and the mixture allowed to stir overnight. The next day the solids were
removed via filtration and the crude material subjected to flash chromatography using 3-5% acetone in DCM.
The intermediate was then suspended in DCM, (2 ml) and chilled to 0o C before addition of TFA (1.0 ml) and
allowed to stir one hour at which time TLC and HPLC indicated the conversion complete. The volatiles were
then removed under vacuum and the solid residue washed with saturated sodium bicarbonate. The crude solids
were then subjected to flash chromatography using 5-10% acetone in DCM gradient to afford the title compound
in 94% yield; 1H NMR (300 MHz, DMSO-d6) δ 7.41 (s, 2H), 5.73 (s, 1H), 4.06 (s, 2H), 3.65 (s, 1H). IR (cm-1):
3278, 2940, 2886, 1739, 1651, 1348, 1166, 1112, 653. MP 148.5 – 152o C.
20. (13S,17R)-13-ethyl-17-ethynyl-3-oxo-2,3,6,7,8,9,10,11,12, 13,14,15,16,17-tetradecahydro-1H-
cyclopenta[a]phenanthren-17-yl 4-sulfamoylbenzoate, sodium. Compound 18 (0.3 g, 0.60 mmol) was dissolved
in THF (5 ml) and chilled to 0o C. The mixture was then treated with 2 M NaOH and stirred vigorously for 5
minutes, then the mixture was transferred to a separatory funnel and the layers separated, the aqueous was
extracted with ethyl acetate. The combined extracts were concentrated to reveal a yellowish solid (233 mg); 1H
NMR (300 MHz, DMSO-d6) δ 7.84 (d, J = 8.1 Hz, 2H), 7.76 (d, J = 8.1 Hz, 2H), 5.75 (s, 1H), 3.65 (s, 1H). IR
(cm-1): 3441, 3282, 2938, 2866, 1725, 1660, 1272, 1255, 1080, 965, 768. MP >315o C.
25. (13S,17R)-13-ethyl-17-ethynyl-3-oxo-2,3,6,7,8,9,10,11,12, 13,14,15,16,17-tetradecahydro-1H-
cyclopenta[a]phenanthren-17-yl (3-methylphenoxy)acetate (General Method #2): To a round bottom flask was
added 1 (1.41 g, 4.5 mmol), (3-methyl-phenoxy)acetic acid (3.0 g, 18.1 mmol), and DMAP (550 mg, 4.5 mmol) in
DCM (30 mL) at ambient temperature. Once a homogenous solution was observed, DIC was added (2.8 mL, 18.1
mmol). The mixture was allowed to stir overnight. The next morning the mixture was filtered, and the filtrate

was rotovaped onto silica gel and subjected to flash chromatography. The resultant foam was then crystallized
1
from methanol and DCM yielding 1.08 g of white crystal (73%). H NMR (300 MHz, CDCl3) δ 7.17 (t, J = 7.5
Hz, 1H), 6.81 (d, J = 7.2 Hz, 1H), 6.73 – 6.67 (m, 2H), 5.84 (s, 1H), 4.59 (s, 2H), 2.66 (s, 1H), 2.33 (s, 3H). IR
(cm-1): 3227, 1775, 1658, 1611, 1259, 1196, 1158, 1095, 785, 688. MP 187 – 190o C.
2.3 Formulation
Vehicle-
Component
Benzyl alcohol
Amount
1 g
Methyl cellulose
Sodium phosphate dibasic
dihydrate
1 g
0.752 g
Sodium phosphate monobasic
dihydrate
2.99 g
Deionized water
200 mL
*All solids added to water and mixed under stirring for 24 hours at ambient temperature
Mill: Fritsch pulverisette 23
Original Formulation:
A stainless steel grinding bowl with a lid and seal along with three 10 mm stainless steel balls were used to
formulate suspensions for injection. The compounds were weighed and added to the grinding bowl and steel balls.
1.5 mL of formulation vehicle was added to the grinding bowl, via pipette. The milling conditions used as listed:
20 minutes, 20 Hz. After milling, the suspension was transferred to a volumetric flask. Formulation vehicle was
used to dilute the milled suspension to produce the desired injection dose.
2.4 Biology
Long acting properties of each preparation had been determined in an estrous suppression assay using virgin,
mature (180-200g) cycling rats of Sprague-Dawley strain. Upon receipt the animals were smeared daily
(procedure described as below) and those exhibiting two consecutive cycles were used for the study. Each animal
was injected subcutaneously (s.c) with 0.5 ml of the test preparation (in the vehicle as described above) on the
same day regardless of the stage of the cycle. Each compound was initially tested in 6 rats. Every other day
excluding weekend, smears were taken starting on the day after injection and continued until such time that
cornification of vaginal epithelium was observed and cycling was re-established [10].
3. Results and discussion
3.1 Chemistry
An initial series of compounds were synthesized as shown in schemes 1-3 and equation 1 below-

Scheme 1. Synthesis of LNG-sulfamoylphenoxyacetic acid esters: (a) DMAP, toluene, reflux, 16 h; (b) Cs2CO3,
DMF, rt
The alpha-chloro LNG intermediate 2 was formed in moderate yields (5 equivalents of anhydride, 1 equivalent
of DMAP) and then the chloride displaced by the cesium phenolate in DMF. The 3-hydroxybenzenesulfonamide
reagent was synthesized as described previously [15].
Equation 1. Synthesis of 3-enol-sulfamoylaryl esters: (a) Cs2CO3, toluene or Tl2CO3, dibenzo-18-crown-6,
toluene, reflux
This analog (5, equation 1) was the main product formed using conditions described in previous articles [9,16]
regarding the use of group I metal carbonate salts to form 17- esters of LNG and norethisterone with acid
chlorides utilized as the acylating agents. There are several possible explanations for the regiochemistry being
different in this case. As the acid chlorides being used in these earlier articles are simple, aliphatic species as
opposed to aromatic sulfonamides, it is possible that under these conditions acylation of the enol tautomer is
prevalent rather than at 17-OH by the acid chloride. Another possibility is that the authors were mistaken in
assigning the structure of their acylation products. As no NMR spectral data is included in these previous articles,
this possibility cannot be verified without repeating their procedures using the aliphatic acid chlorides. In
addition to analog 5 a 4,4’-biphenylsulfonamide analog was made in the same manner.

Scheme 2. Synthesis of phenylsulfamoyl amino acid linked LNG esters: (a) DIC/DMAP/DCM, rt; (b) conc
HCl/EtOAc; (c) DIC/HOBt/DCM, rt; (d) DIEA/THF, then NH4OH (e)
Table 1. Compounds 6-9,15
Compound
6
R1
R2
-
R1-R2-propyl
7
8
Me
Me
H
H
Me
Me
9
15a
a- azetidine beta amino acid
As noted earlier, our previous work focused on steroid-sulfonamide conjugates linked using amino acids, so a
large portion of our first series also made use of this same motif. Previously described methods of forming
aliphatic esters through mixed anhydride formation [10, 17] utilizing trifluoroacetic anhydride were either
completely ineffective or created a complex, intractable mixture of products when using N-protected amino acids
as substrates. Classic Steglich-type esterification at first also proved ineffective, despite using an excess of amino
acid, and catalytic amounts of DMAP. After screening various parameters of the esterification procedure, it was
found that by using five equivalents of acid (this was later pared back to four) and one equivalent of DMAP, the
esterification proceeded well and good to excellent yields were observed. One crucial aspect to the success of this
acylation was that the carbodiimide (we mainly used N,N’-diisopropylcarbodiimide) be added last to the mixture.
The reaction concentration was also important to achieving the excellent yields, with a requirement that at least
0.1 M with respect to the steroid be maintained. The protecting group was then removed and then amide
formation using sulfamoylbenzoic acid and DIC and HOBt was performed to afford the analogs. We also used
other sulfamoyl acids to form amides, but as these were not tested in our anti-ovulation assay, they have been
omitted from this article.

Scheme 3. Synthesis of aryl and aliphatic sulfamoyl LNG esters: (a) TEA/THF or DCM, 0oC to rt; (b)
NaOH/THF, reflux; (c) 1./DIC/DMAP/DCM, rt; (d) TFA/DCM, 0oC; (e) MeOH/TEA/DCM, -78oC to rt; (f) 2M
NaOH/THF, rt
The conditions used to successfully form amino-acid esters of LNG described in scheme 2 did not work when
either an alkyl or aryl sulfamoyl carboxylic acid was the substrate. This was later attributed to the presence of the
sulfonamide group. Masking the acidic protons by making use of the bis-2,4-dimethoxy-benzylamino blocking
group [18] (scheme 3) allowed for the successful employment of the conditions described in scheme 2 and good
to excellent yields were once again observed. The protecting group was removed in the final step by treatment
with TFA. It is noted that the sulfonyl chloride carboxylic acid was protected as its methyl ester prior to chloride
displacement with bis-2,4-dimethoxybenzylamine. This allowed for much better yields and ease of isolation.
Once the conjugate was in hand, the carboxylic acid was easily attainable (work up and isolation made use of
either potassium or sodium hydrogen sulfate to bring the pH down to about 2-3, so as not to remove the bis-2,4-
dimethoxyamino group) and then used in the subsequent carbodiimide mediated esterification of LNG.
Table 2. LNG esters without sulfonamide function

Compound
R
Method
useda
Compound
R
Method
useda
*
O
*
O
O
22
2
2
2
2
2
2
36
2
2
2
2
1
1
F
*
*
*
*
*
*
*
*
*
*
23
24
25
26
27
37
38
39
40
41
N
O
F
O
O
O
MeO
O
O
O
Cl
Cl
O
SO2Me
*
*
NH
*
*
*
O
O
28
29
30
2
2
2
42
43
45
2
2
1
S
O
Cl
O
N
*
O
O
*
31
32
2
2
46
47
1
1
*
*
O
OMe
*
O
O
MeO
`
OMe
O
*
O
33
2
48
49
1
2
*
*
N
*
*
O
O
O
34
35
2
2
CN
F
a- Method 1- Displacement of Cl from intermediate 2 (see scheme 1)
Method 2- Carbodiimide mediated esterification (see equation 2)

O
O
OH
O
O
O
OH
a
44
+
O
O
Equation 2. Synthesis of phenoxyacetic acid ester of ENG: (a) DIC/DMAP/ DCM, rt
Ester formation proceeded in either one of two routes- displacement of the alpha chloride from intermediate 2
using cesium carbonate to generate the cesium phenolate salt (scheme 1) or carbodiimide-mediated coupling,
using the conditions described previously (equation 2). Formation of 42 required Boc-protection of N-
phenylglycine [19] before DIC esterification with subsequent BOC removal after esterification. The planned
analog 48 was to be a 2-pyridylphenoxyacetic acid ester of LNG, but spectroscopic analysis indicated the N-
linked amide.
3.2 Formulation of analogs used in anti-ovulation study
Compounds in a crystalline form (or amorphous when crystals were unable to be attained) were suspended in a
buffered formula (40 mg in 1.5 ml) as described in Bialy et. al. [10] with the omission of polysorbate and then
subjected to mechanical milling using 10 mm stainless steel ball bearings at a rate of 20 Hz for twenty minutes.
3.3 Biology
S-D rats (180 – 200 g) were observed for ten days prior to dosage (to ensure regular estrous cycling), then
subjected to single sub-cutaneous injection. Estrus suppression was monitored by taking vaginal swabs every
three days during a five day period (followed by two days of no observations, then a return to every three days
during a five day period) and observing the epithelial and lamina propria microscopically; smears which showed
cornified tissue indicated a return to an estrus and/or ovulating state at which time the activity was noted as
ceased.
Figure 1. Left- microscopic appearance of
murine vaginal smears at each stage of the estrous cycle. Right- description of estrogen and progesterone levels
through the estrous cycle.

Table 3. Results from anti-ovulation assay of first series of compounds (average of 4 or 5 animals per group)
Compound Structure
Rat anti-ovulatory
activity (days)
Compound
Structure
Rat anti-ovulatory
activity (days)
2 mg
4 mg
2 mg
4mg
O
O
N
O
O
O
O
O
NH2
O
S
S
21 +
7.0
34 +
3.2
O
O
3
16
1
NH2
O
O
OH
O
O
O
5
1
1
7 + 3.8
17
18
19
20
1
1
1
O
S
O
O
O
S
NH2
O
H2N
O
O
O
N
O
O
O
S
NH2
O
O
NH2
S
O
O
11
1
1
O
O
O
S
O
O
O
NH2
O
O
O
S
NH2
NH
O
O
12
O
O
O
S
O
O
N
O
NH2
O
O
O
14 +
1.2
13
1
1
1
1
O
S
NH-Na+
O
O
O
O
O
S
Ac
O
N
O
NH2
OAc
O
O
MPA (pos
control)
32 +
8.3
40 +
8.8
14
O
O
O
O
S
NH2
N
O
15
O
O
From the data listed in table 3, several aspects are made evident; all of the sulfamoyl-amino-acid esters of LNG
have arguably very poor anti-ovulatory activity in this assay. This trend was also observed with the non-amino
sulfamoyl esters of LNG. Interestingly, the sodium salt 20 did give 14 + 1.2 days of activity at the high dose but
was deemed insignificant as MPA at this dose was significantly more effective. Also of slight interest was the

mild activity at 4 mg of 3-enol-ester analog 5. Meta sulfamoylphenoxyacetic acid analog 3 was clearly the most
effective compound of this series. Curiously, this was the only analog of the series that was able to be crystallized
effectively, although when tested head to head with its amorphous form there was not much difference in duration
of ovulation suppression. The next series of compounds were based on analog 3 and set out to answer two
questions- how important was the sulfonamide function and would other substitutions on the phenoxy ring result
in better activity than 3.
Table 4. Results from anti-ovulation assay of second series of compounds.
Compound
Structure
Rat anti-ovulatory
activity (days)
Compound
Structure
Rat anti-ovulatory
activity (days)
2 mg
2 mg
4 mg
O
O
O
O
O
O
4
20 + 6.2
31
32
33
34
30 + 1.9
O
S
OMe
OMe
O
NH2
O
O
O
O
O
O
O
O
69 +
12.2
22
23
24
43 + 5.4
18 + 1.7
16 + 5.2
8 + 0
O
O
O
O
O
O
O
20 +
4.3
OMe
O
O
O
O
O
O
O
6 +
1.9
O
O
O
O
O
O
O
O
36 +
3.4
F
25
35
31 + 6.6
O
O

O
O
O
O
O
O
45 +
6.1
26
27
36
37
31 + 5.9
23 + 1.0
F
O
O
O
O
O
O
O
O
47 +
3.0
N
F
Cl
O
O
O
O
O
O
O
O
28
29
30
34 + 3.2
8 + 0.6
38
39
41 + 3.6
3 + 0.8
Cl
O
O
O
O
O
O
OMe
O
O
O
O
O
Cl
O
O
16 + 5.4
From this data (table 4) it is clear that compounds 22 and 38 were the frontrunners in the series. A quick glance at
the data indicates the necessity for the phenoxyacetic acid moiety for superior anti-ovulation activity. Although
some of the substituted phenoxyacetic acid derivatives are far less effective with regards to preventing ovulation
than 22 and 38, none of the compounds lacking the oxygen atom (23, 24, 39) give comparable anti-ovulation
activity to these two leads. The differences in anti-ovulation activity regarding the substituted phenoxyacetic
acids can be loosely described as such- in the case of methyl, methoxy, and chloro substituents, ortho substitution
resulted in the fewest days of ovulation inhibition, in contrast with ortho-fluorine substitution resulting in the
longest interval of inhibition for the fluorine substituted series. Fluorine substitution resulted in the highest
additive number of days of inhibition at the 2 mg dose, while methoxy substitution resulted in the least number of
days. Para-sulfonamide analog 4 inhibited ovulation twice as long as meta analog 3. 3-pyridyl analog 37 inhibited
ovulation over 50% less effectively than 22.
Also of interest is the comparison between analog 22 and its -methyl relative 34. It is possible that this
methyl group inhibits ester hydrolysis to the active compound, and the prodrug is eliminated intact.
Our next series of compounds explored SAR of the aromatic function in addition to other heteroatoms linking
the carboxyl group to the aromatic function besides oxygen, as well as the synthesis and determining the anti-
ovulatory activity of the phenoxyacetic acid ester of ENG.
Table 5. Results from anti-ovulation assay of third series of compounds

Compound
Structure
Rat anti-
ovulatory activity
(days), 2 mg
dose
O
O
O
O
40
41
42
20 + 3.7
31 + 7.4
13 + 0.5
O
O
O
O
O
SO2Me
O
O
O
NH
O
O
O
S
43
44
20 + 3.1
60 + 6.9
O
O
O
O
O
O
O
O
N
O
45
46
27 + 1.5
O
O
O
O
8 + 0.8
O
O
O
O
47
37 + 1.0
O

O
O
N
48
49
5 + 1.0
O
O
O
O
O
29 + 4.7
CN
O
Analysis of this raw data (table 5) make clear two concepts- none of these analogs prevent ovulation in rats for
as long as the phenoxyacetic acid analog 22, and that placing this group on ENG (analog 44) gave a compound
which gave the longest duration of anti-ovulatory activity tested to date. Of interest, ENG has a markedly reduced
duration of anti-ovulatory activity in the assay compared to LNG (8 + 2.1 vs 28 + 0.7 days at the 2 mg dose).
Changing the aromatic group while maintaining the methylene linker between acyl and aryl groups did not
result in any activity better than 22 or 44. Also, changing the aryl-linked heteroatom did not result in any
improvement of activity, although somewhat interestingly chalcogen 43 had markedly better activity than
pnictogen 42. Also of interest is the much longer duration of anti-ovulation activity in para-biphenyl analog 47 vs
meta 46. This is somewhat akin to what was observed in the second series of compounds, with para substitution
giving the superior activity, with the exception of the fluorine substituents.
4. Discussion
Analysis of this raw anti-ovulatory activity data seems to suggest a rather limited number of these ester pro-
drugs could be considered as truly superior in causing cessation of ovulation in rats. What seems to be very basic
requirements for what could be considered the most effective contraceptives of this group is the presence of an
alpha-oxygen atom, one methylene unit between the ester carbonyl function and the oxygen atom, and an
aromatic moity bound to the oxygen atom that is not a heterocycle. This rather “tight” SAR was also noted in
previous work regarding esters of LNG by Bialy et. al.[10] For instance, one of the most effective compounds
noted by these workers was levonorgestrel butyrate (LB, inhibited ovulation for 34 + 3 days at 2 mg). By
comparison, changing the number of carbon atoms in the chain by just one unit in either direction resulted in
substantially lower number of days of activity at all doses studied (propanoate ester inhibited ovulation for 17 + 1
days at 2 mg and pentanoate inhibited ovulation for 14 + 2 days at 2 mg). Ringed aliphatic esters cyclopropanoate
and cyclobutanoate gave far greater activity (37 + 2 days and 32 days both at 2 mg) than cyclopentanoate and
cyclohexanoate (4 + 1 day and 7 + 3 days both at 2 mg), again illustrating the small window of activity among
similar analogs.
Trying to rationalize the overall activity using single parameters when comparing the compounds does not
result in any noticeable trends or plausible explanations. Take for instance the idea that higher melting
compounds result in a more long-lasting depot (as more time and biological media would be required to disrupt

the crystal lattice in theory). This idea is easily nullified by comparing 22 and 44; 44 melts at 160o C while 22
melts at 206o C, yet 44 gives the longest activity of all analogs. Compound 42 was one of the highest melting
compounds listed in this article (243o C), but exhibits poor activity in comparison.
Another concept taken into account is ester hydrolysis. It was noted by Bialy et. al. that some esters with very
slow hydrolysis rates still resulted in activity. It was not mentioned whether or not these compounds were tested
for progesterone receptor binding.
As speculation regarding explanation of activity of these analogs could be extensive, a reasonable way to
explain these observations is a “perfect storm” of all parameters- ester hydrolysis, crystallinity, aqueous solubility,
particle size, and perhaps most importantly, formulation, must form and additively contribute to optimal anti-
ovulation activity. It does indeed seem that compounds 22 and 44 meet all these criteria, as these (especially 44)
have been found to have nearly double the anti-ovulatory activity of MPA in the rat model. Challenges exist to
get these compounds into the clinic for human use. Besides the obvious issues of toxicity (in this case, the parent
is an approved drug, however any intact pro-drug which did not hydrolyze and the pro-drug piece itself would
have to be subjected to tox studies), one issue that we believe to be crucial is human bioavailability. This concern
would stem from the case of LB mentioned earlier. In the 1970’s, this was arguably the WHO’s most promising
compound (again, in rats) as a single-injectable, long-term contraceptive, and has undergone human clinical trials
up to phase 3, but to this date has not been either approved or marketed. It is unclear why this might be the case,
but one cause could possibly be inconsistencies in duration of anti-ovulation from individual subject to subject. It
is quite possible that these inconsistencies (assuming this to be the case) are caused by a deficiency in the
formulation. Now that particle sizing technologies have progressed well beyond the 1970’s, our group has been
focusing on using jet-milling technology to produce spherical, tightly distributed particle sizes of each prodrug.
The tight distributions will have small (Dv 50 of 5 micron), medium (Dv 50 of 12 micron), and large (Dv 50 of 18
micron) average sized particles. Each particle set will then be tested for anti-ovulation activity in the rat model, to
determine whether an optimal Dv 50 gives even longer duration of anti-ovulation. Also, we are studying the
injection volume of the formulated pro-drugs, to determine if this plays a role in anti-ovulation duration. These
studies along with other topics such as polymorphism and ester hydrolysis studies are to be described in future
articles.
5. Summary
A series of compounds were made using a hypothesis based on previous work involving steroidal human
carbonic anhydrase ligands for improved bioavailability when probed for anti-ovulation activity in a murine
model identified only one compound with satisfactory activity. Subsequent analogs based loosely on this single
initial compound yielded phenoxyacetic acid esters of LNG and ENG which resulted in anti-ovulation activity of
substantially longer duration in comparison to MPA, and are currently undergoing further studies.

Acknowledgments
The authors thank the Bill & Melinda Gates Foundation for providing funding for this initiative through grant no.
OPP1130985.
References
1. United Nations (UN), Department of Economic and Social Affairs, Population Division. World
contraceptive use 2011. New York: UN; 2011. Available from:
http://www.un.org/esa/population/publications/contraceptive2011/contraceptive2011.htm
2. K. McKenna, J. Arcara, K. Rademacher, C. Mackenzie, F. Ngabo, E. Munyambanza, J. Wesson, E.
Tolley. Policy and programmatic considerations for introducing a longer-acting injectable contraceptive:
perspectives of stakeholders from Kenya and Rwanda. Global Health: Science and Practice 2 (2014): pgs. 459-
470.
3. J. Trussell. "Contraceptive Efficacy". In Hatcher RA, Trussell J, Stewart FH, Nelson AL, Cates Jr W,
Guest F, Kowal D. Contraceptive Technology (18th rev. ed.). New York: Ardent Media, (2004) pgs. 773–845.
4. J. Trussell. Contraceptive failure in the United States. Contraception. 70 (2004): pgs. 89–96
5. http://www.pdr.net/drug-summary/Depo-Provera-Contraceptive-Injection-medroxyprogesterone-acetate-
1373
6. C. Westhoff. Depot-medroxyprogesterone acetate injection (Depo-Provera®): a highly effective
contraceptive option with proven long-term safety. Contraception. 68 (2003): pgs. 75–87
7.https://web.archive.org/web/20051221195621/http://www.fda.gov/bbs/topics/ANSWERS/2004/ANS01325.ht
ml
8. J.E. Rosseow, G.L. Anderson, R.L. Prentice, A.Z. LeCroix, C. Kooperberg, M.L. Stefanick, R.D.
Jackson, S.A. Beresford, B.V. Howard, K.C. Johnson, J.M. Kotchen, J. Ockene. Risks and benefits of estrogen
plus progestin in healthy postmenopausal women: principal results From the Women's Health Initiative
randomized controlled trial. JAMA, 288 (2002): 321–33.
9. A. Shafiee, M. Vossoghi, F. Savabi, R. Vlahov, V. Tarpanov, M. Boshkova-Ljapova, B. Milenkov, V.
Stoilova, J. Vlahov, G. Snatzke. Long-acting contraceptive agents: cyclopropyl and cyclobutyl esters of
norethisterone. Steroids, 41 (1983), pgs. 291-307
10. G. Bialy, R.P. Blye, R.P. Enever, R.H. Naqvi, M.C. Lindberg. Long-acting contraceptive agents:
Structure activity relationships in a series of norethisterone and levonorgestrel esters. Steroids, 41 (1983), pgs.
419-439

11. P. Crabbé, S. Archer, G. Benagiano, E. Diczfalusy, C. Djerassi, J. Fried, T. Higuchi. Long-acting
contraceptive agents: design of the WHO chemical synthesis programme. Steroids, 41 (1983): pgs. 243–53
12. G. Benagiano, C. d’Arcangues, R. Harris, A. Schafer, L. Say, M. Merialdi. The special programme of
research in human reproduction: forty years of activities to archive reproductive health for all. Gynecol Obstet
Invest. 74 (2012), pgs. 190-217.
13. A.B. Edelman, G. Cherala, H. Li, D.L. Blithe, J.T. Jensen. Levonorgestrel butanoate intramuscular
injection does not reliably suppress ovulation for 90 days in obese and normal-BMI women: a pilot study.
Contraception, 95 (2017), pgs. 55-58
14. G. Ahmed, W. Elger, F. Meece, H. Nair, B. Schneider, R. Wyrwa, K. Nickisch. A prodrug design for
improved oral absorption and reduced hepatic interaction. Bioorganic Med. Chem. 25 (2017), pgs. 5569-5575
15. S.P. Sahoo, H. Koyama, D.J. Miller, N-cyclohexylaminocarbonylbenzenesulfoamide derivatives WO
2004066963
16. J.E. Herz, A.A. Leslie Gunatilaka, S. Sotheeswaran, J.V. Torres. Esterification of acid chlorides with
thallium and potassium salts of 19-norethisterone: Formation of 17-enol esters. Steroids, 40 (1982), pgs. 261-266
17. S.L. Leung, R. Karunanity, G. Becket, S.H. Yeo. Norethisterone and levonorgestrel esters: A novel
synthetic method Steroids, 46 (1985), pgs. 639-647
18. T. Reuillon, A. Bertoli, R.J. Griffin, D.C. Miller, B.T. Golding. Efficacious N-protection of O-aryl
sulfamates with 2,4-dimethoxybenzyl groups. Org. Biomol. Chem., 10 (2012), pgs. 7610-7617
19. W.S. Lee, M.J. Lee, B.J. Kim, T.C. Roh, S.H. Lee, K.D. Lee, K. You-Hui, T.H. Kwak. 1,2-
Napthoquinone derivative and method for preparing same. Pct Int Appl., 2015102371

Graphical Abstract
To create your abstract, type over the instructions in the template box below.
Fonts or abstract dimensions should not be changed or altered.
Leave this area blank for abstract info.
Esters of Levonorgestrel and Etonogestrel
Intended as Single, Subcutaneous-Injection,
Long-Lasting Contraceptives
Frederick A Meece, Gulzar Ahmed, Hareesh Nair, Bindu Santhamma, Rajeshwar R Tekmal, Chumang Zhao,
Nicole E Pollok, Julia Lara, Ze’ev Shaked, Klaus Nickisch
Evestra, Inc. 6410 Tri County Pkwy, Schertz, TX 78154
O
OH
O
R1
Formulated Sub-Q Injection,
R
R
Esterases
Compound "depot"
O
O
Months of anti-ovulatory effect due
to steady concentration of progestin

Article Highlights



First example of carbodiimide-mediated esterification of levonogestrel and etonogestrel
developed as a route to numerous analogs
First examples of sulfonamide-derived levonorgestrel analogs
An iterative drug discovery initiative which probed structure activity relationships identified two
lead compounds as effective anti-ovulation agents in the murine model

Phenoxyacetic acid ester of etonogestrel gave the longest anti-ovulation interval of any reported
subcutaneous injected contraceptive in the murine model