Development of a new sensitive and specific time-resolved fluoroimmunoassay (TR-FIA) of chlormadinone acetate in the serum of treated menopausal women.

Article abstract of DOI:10.1016/S0039-128X(02)00050-8

We describe the development of a serum chlormadinone acetate (CMA) time-resolved fluoroimmunoassay (TR-FIA). We prepared haptens (3-CMO-chlormadinone acetate and 6-chloropregna-4,6-dien-17,20-diol-3-one-20-hemisuccinate), biotinylated tracers (3(biotinylaminopropylamido) 3-CMO-chlormadinone acetate and 3-(6-chloropregna-4,6-dien-17,20-diol-3-one-20-hemisuccinylamino)1-biotinylaminopropane), and immunogens necessary for eliciting two antibodies (anti-chlormadinone acetate 3-CMO/BSA and anti-chlormadinone 20-hemisuccinate/BSA). The specificity of the assay was rigorously studied to eliminate possible interference by polar metabolites of CMA, particularly 17 alpha-acetoxy-6-chloro-3beta-hydroxypregna-4,6-diene-20-one (3beta-hydroxy metabolite), employing an easy-to-use ethylene glycol chromatographic step prior to immunoassay, so as to separate the polar metabolites, in particular the 3beta-hydroxy-CMA metabolite, from the intact CMA. The choice of the anti-CMA antibody was guided by the high assay sensitivity obtained with the anti-CMA 3-CMO/BSA antibody. The detection limit was 51pg/ml. Interassay reproducibility CVs were between 2.6 and 4.5%. This TR-FIA thus appeared to be a sensitive, specific, precise, and consequently well-suited method for measurement of serum CMA during a pharmacokinetic study in women.

Full text of DOI:10.1016/S0039-128X(02)00050-8

Steroids 67 (2002) 1045–1055
Development of a new sensitive and specific time-resolved
fluoroimmunoassay (TR-FIA) of chlormadinone acetate
in the serum of treated menopausal women
a,b,∗
a
c
d
Jean Fiet
, Frank Giton , Jack Auzerie , Hervé Galons
a
Laboratoire de Biologie Hormonale, H oˆ pital Saint-Louis, 1 avenue Claude Vellefaux, 75475 Paris Cedex 10, France
b
Laboratoire de Biochimie, Faculté de Pharmacie, 75006 Paris, France
ITEC Service, 3 avenue George Clémenceau, Cenonn, 33150 Bordeaux, France
Laboratoire de Chimie Organique, Faculté de Pharmacie, 75006 Paris, France
c
d
Received 27 March 2002; received in revised form 13 May 2002; accepted 24 May 2002
Abstract
We describe the development of a serum chlormadinone acetate (CMA) time-resolved fluoroimmunoassay (TR-FIA). We prepared
haptens (3-CMO-chlormadinone acetate and 6-chloropregna-4,6-dien-17,20-diol-3-one-20-hemisuccinate), biotinylated tracers (3(biotiny-
laminopropylamido) 3-CMO-chlormadinone acetate and 3-(6-chloropregna-4,6-dien-17,20-diol-3-one-20-hemisuccinylamino)1-biotinyl-
aminopropane), and immunogens necessary for eliciting two antibodies (anti-chlormadinone acetate 3-CMO/BSA and anti-chlormadinone
2
0-hemisuccinate/BSA). The specificity of the assay was rigorously studied to eliminate possible interference by polar metabolites of CMA,
particularly 17␣-acetoxy-6-chloro-3␤-hydroxypregna-4,6-diene-20-one (3␤-hydroxy metabolite), employing an easy-to-use ethylene gly-
col chromatographic step prior to immunoassay, so as to separate the polar metabolites, in particular the 3␤-hydroxy-CMA metabolite, from
the intact CMA. The choice of the anti-CMA antibody was guided by the high assay sensitivity obtained with the anti-CMA 3-CMO/BSA
antibody. The detection limit was 51 pg/ml.
Interassay reproducibility CVs were between 2.6 and 4.5%. This TR-FIA thus appeared to be a sensitive, specific, precise, and conse-
quently well-suited method for measurement of serum CMA during a pharmacokinetic study in women.
©
2002 Elsevier Science Inc. All rights reserved.
Keywords: Time-resolved fluoroimmunoassay; Celite chromatography; Chlormadinone acetate; Steroid
1
. Introduction
injection into rabbits, in order to develop anti-CMA anti-
bodies. Biotinylated CMA and biotinylated chlormadinone
were also synthesised. The specificity of this assay is guar-
anteed by a chromatographic separation step on ethylene
glycol/celite columns, which separate intact CMA from its
polar metabolites. The accuracy of the assay, in serum was
checked by dilution tests, as well as by overload tests. The
sensitivity of the assay is at least equal to 50 pg/ml of serum.
This immunoassay is now being applied to the measure-
ment of CMA in the sera of CMA + 17␤ estradiol-treated
menopausal women.
Chlormadinone acetate (CMA) (17␣-acetoxy-6-chloro-
,6-pregnadiene-3,20-dione) is a synthetic progestin used
4
as an oral contraceptive. After oral intake, the measurement
of CMA in plasma has been carried out using a gas–liquid
chromatographic method. However, this method is very
insensitive and requires a large volume of plasma [1]. Al-
though it is very probable that assay of this drug has also
been carried out in serum or plasma by gas chromatography–
mass spectrometry, to our knowledge, no paper has reported
on this subject.
We present here for the first time, a serum immunoassay
for CMA, using time-resolved fluoroimmunoassay (TR-
FIA) methodology. This implied the synthesis of CMA
haptens, followed by coupling to BSA (immunogens), then
2
. Materials and methods
2.1. Steroids
∗
2.1.1. Chlormadinone acetate and most of the reagents
CMA and most of the reagents were obtained
from Sigma–Aldrich (Saint-Quentin Fallavier, France).
Corresponding author. Tel.: +33-1-42-49-92-87;
fax: +33-1-42-49-42-80.
E-mail address: jean.fiet@libertysurf.fr (J. Fiet).
0
039-128X/02/$ – see front matter © 2002 Elsevier Science Inc. All rights reserved.
PII: S0039-128X(02)00050-8

1
046
J. Fiet et al. / Steroids 67 (2002) 1045–1055
CMA purity was >99%. Solvents were obtained from
SDS.
2.1.2.2. 3(Biotinylaminopropylamido)carboxymethyloxy-
chlormadinone acetate (4). Coupling to the biotinylation
reagent (3) was performed as previously described [2] to
give 4 (75% yield); mp: 185–192 C. H NMR (CDCl3):
δ (ppm) 0.70 (s, 3H, 18-CH3); 0.90 and 1.0 (2s, 3H, E
and Z 19-CH3); 1.98 and 2.05 (2s, 2 × 3H, 20-CH3 and
CH3COO); 4.22 and 4.45 (2m, 2H, CHN); 4.55 and 4.6 (2s,
2H, 2 CH2–CO); 4.80 and 5.60 (2 broad s, 2s, NH-biotin);
6.0 and 6.15 (2d, 1H, E and Z 7-H); 6.40 and 7.00 (2s, 1H,
E and Z 3-H); 6.50 and 6.80 (2t, 2H, 2 NH–CH2).
◦
1
2
.1.2. Preparation of 3-CMO-chlormadinone
acetate (hapten), and 3(biotinylaminopropylamido)
3
-CMO-chlormadinone acetate (tracer) (Scheme 1)
.1.2.1. Carboxymethyloxy-6-chloro-17-acetoxypregna-
2
4
,6-dien-3,20-dione
(3-carboxymethyloxychlormadinone
acetate or 3-CMO-chlormadinone acetate) (2). To a solu-
tion of 1 (0.404 g, 1 mmol) in methanol was added aminooxy
acetic acid hemihydrochloride (0.11 g, 1 mmol) and sodium
acetate (0.136 g, 1 mmol). The solution was stirred for 24 h
at RT. After concentration in vacuo, the solid was taken
with water 5 ml and acidified to pH 2 with 1N HCl. The
CMOs were extracted with ethyl acetate (3 × 20 ml). The
organic layer was dried on Na2SO4 and concentrated in
vacuo to afford 2 (90% yield) as a 70:30 mixture of E–Z
2.1.3. Preparation of
6-chloro-17-acetoxypregna-1,4,6-trien-3,20-dione
(precursor for tritiated chlormadinone acetate synthesis),
and chlormadinone synthesis
(non-acetylated chlormadinone) (Scheme 2)
2.1.3.1. 6-Chloro-17-acetoxypregna-1,4,6-trien-3,20-dione
(5). A mixture of CMA 1, 0.2 g and 2,3-dichloro-5,6-
dicyano-1,4-benzoquinone (DDQ) 0.165 g in 5 ml toluene
◦
stereoisomers; mp: 185–190 C.
1H NMR (CDCl3): δ (ppm) 0.73 (s, 3H, 18-CH3); 0.95
and 1.01 (2s, 3H, Z and E 19-CH3); 1.97 and 2.03 (2s, 1.98
and 2.03, CH3CO and CH3COO); 4.55 and 4.65 (2s, 2H,
CH2, Z 30% and E 70% CMO); 6.02 and 6.11 (2d, 1H, E
and Z 7-H); 6.45 and 7.02 (2s, 1H, E and Z 3-H).
◦
was heated at 90 C for 36 h. The solution was washed with
5 ml 1N NaOH then twice with water, dried and concen-
trated in vacuo. The precursor 5 (65% yield) was separated
by column chromatography using toluene–ethyl acetate 1:1
Scheme 1. Preparation of the hapten 2 and the tracer 4.

J. Fiet et al. / Steroids 67 (2002) 1045–1055
1047
Scheme 2. Preparation of the trienone 5, the 3-β-hydroxyderivative 6 and of chlormadinone 8.
◦
1
as eluent; mp: 170–176 C. H NMR (CDCl3): δ (ppm)
in vacuo and neutralised by addition of a few drops of 1N
HCl. After extraction by ethyl acetate the organic layer was
0
2
4
.78 (s, 3H, CH3); 1.30 (s, 3H, CH3); 2.02 and 2.04 (2s,
× 3H, 20-CH3 and CH3COO); 6.12 (d, J = 2.5 Hz, 1H,
-H); 6.26 (dd, J = 11 Hz and J = 2.5 Hz, 1H, 2-H); 6.58
◦
dried and evaporated. Yield: 85%, mp: 210–215 C.
1
H NMR (CDCl3): δ (ppm) 0.72 (s, 3H, 18-CH3); 1.10
(
s broad, 1H, 7-H); 7.03 (d, 1H, J = 11 Hz, 1-H).
(s, 3H, 19-CH3); 2.20 (s, 3H, 20-CH3); 6.25 (m, 2H, 3-H
and 7-H).
2
2
.1.3.2. 6-Chloro-3β-hydroxy-17-acetoxypregna-4,6-dien-
0-one (6) and 6-chloro-17-acetoxypregna-4,6-dien-3β-20-
2.1.4. Preparation of 6-chloropregna-4,6-dien-17,20-
diol-3-one-20-hemisuccinate (hapten), and
3-(6-chloropregna-4,6-dien-17,20-diol-3-one-20-
hemisuccinylamino)1-biotinylaminopropane (tracer)
(Scheme 3)
diol (7). A solution of CMA 1, 0.3 g in methanol 5 ml was
cooled to 0 C and sodium borohydride 0.038 g in 0.8 ml
◦
H2O was added. The solution was stirred at RT for 30 min.
Excess of reductive agent was destroyed by addition of a
few drops of acetic acid. The mixture was concentrated and
extracted with methylene chloride and water. The organic
layer was dried and evaporated. The alcohols 6 (50% yield)
and 7 (12% yield) were separated by column chromatogra-
phy using toluene–ethyl acetate 7:3 as eluent.
2.1.4.1. 17-Acetoxy-6-chloro-3-ethylendioxypregna-4,6-
dien-20-one (9). A mixture of CMA (0.808 g, 2 mmol)
ethylene glycol 0.2 ml and paratoluene sulfonic acid 0.05 g
in toluene–cyclohexane 50:50 was refluxed for 4 h in a Dean
Stark apparatus. After cooling, the mixture was taken with
ethyl acetate washed with 5 ml 1N Na2CO3 then with water
20 ml. After drying, the solvent was evaporated and diox-
olane 9 (83% yield) was separated from unreacted 1 (5%
yield) on a column chromatography toluene–AcOEt 7:3 as
Compound 6: mp: 220–224 C. 1H NMR (CDCl3): δ
◦
(ppm) 0.65 (s, 3H, 18-CH3); 0.95 (s, 3H, 19-CH3); 2.0 (s,
3
H, CH3COO); 2.15 (s, 3H, CH3); 4.26 (t, 1H, 3-H); 5.82
(d, 1H, 7-H); 6.05 (s, 1H, 3-H).
Compound 7: mp: 215–218 C. 1H NMR (CDCl3): δ
◦
◦
1
(ppm) 0.75 (s, 3H, 18-CH3); 0.92 (s, 3H, 19-CH3); 1.08 (d,
eluent; mp: 180–185 C. H NMR (CDCl3): δ (ppm) 0.60
(s, 3H, 18-CH3); 0.95 (s, 3H, 19-CH3); 2.00 and 2.05 (2s,
2 × 3H, CH3COO and 21-CH3); 4.00 (m, 4H, CH2–CH2
dioxolane); 5.90 (s, d, 2H, H-3 and H-7).
1
H, 20-CH3); 3.90 (m, 1H, 20-H); 4.20 (m, 1H, 3-H); 5.75
(d, 1H, 7-H); 6.02 (s, 1H, 3-H).
2
.1.3.3. 6-Chloro-17-hydroxypregna-4,6-dien-3,20-dione
(non-acetylated chlormadinone) (8). To a solution of CMA
2.1.4.2. 3-Ethylendioxy-6-chloropregna-4,6-dien-17,20
(0.404 g, 1 mmol) in 10 ml methanol, was added 1 ml of 1N
triol (10). To a suspension of 0.2 g AlLiH in THF 5 ml
4
sodium hydroxide and 0.120 mg of sodium perchlorate. The
solution which was stirred for 2 h at RT was concentrated
was added 9 (0.675 g, 1.5 mmol). After 2 h refluxing, the
mixture was cooled to 0 C, AcOEt 5 ml and then 50 ml cold
◦

1
048
J. Fiet et al. / Steroids 67 (2002) 1045–1055
Scheme 3. Preparation of the hapten 12 and the biotinylated tracer 13.
water were added. The mixture was extracted with AcOEt
washed with 1N HCl dried and evaporated. The hapten 12
3
× 50 ml. The organic layer was washed with water, dried
(45% yield) was purified by recrystallisation from AcOEt;
◦
1
◦
1
and evaporated to give 10 (90% yield); mp: 145–150 C. H
NMR (CDCl3): δ (ppm) 0.68 and 0.78 (2s, 3H, 18-CH3);
0
mp: 223–228 C. H NMR (CDCl3): δ (ppm) 0.70 and
0.80 (2s, 3H, 18-CH3); 1.05 (s, 3H, 19-CH3); 2.60 (m, 4H,
COCH2CH2CO); 5.05 (m, 1H, 20-H); 6.26 (s, 1H, 3-H and
d, 1H, 7-H).
.95 (s, 3H, 19-CH3); 1.12 (d, 3H, 21-CH3); 3.80 (m, 1H,
0-H); 4.00 (m, 4H, CH2–CH2 dioxolane); 5.82 (s, 1H,
2
H-3); 5.90 (d, 1H, H-7).
2.1.4.5. 3-(6-Chloropregna-4,6-dien-17,20-diol-3-one-20-
2
.1.4.3. 6-Chloropregna-4,6-dien-17,20 triol-3-one (11).
Deprotection was achieved by stirring the ethylene ketal 10
0.41 g, 1 mmol) in 20 ml 1N HCl–MeOH 50:50. After con-
hemisuccinylamino)1-biotinylaminopropane (13). The
acylation of the biotinylation reagent by hapten 12 was
achieved as previously described [2] to afford tracer 13
(
◦
1
centration in vacuo, the mixture was extracted with AcOEt.
The organic layer was washed with 5N NaHCO3, then with
water, dried and evaporated to afford 11 (95% yield); mp:
(38% yield); mp: 185–192 C. H NMR (CDCl3): δ (ppm)
0.70 and 0.80 (s, 3H, 18-CH3); 4.20 and 4.38 (2m, 2H,
2CHN); 5.03 (m, 1H, CH–N); 5.03 (m, 1H, 20-H); 5.75
and 5.85 (2s, 2H, NHCONH); 6.15 (s, 1H, 3-H); 6.30 (d,
1H, 7-H); 7.25 and 7.42 (2t, 4H, 2NHCOCH2).
◦
1
1
61–164 C. H NMR (CDCl3): δ (ppm) 0.72 and 0.81 (2s,
3
H, 18-CH3); 3.82 and 3.98 (2q, 1H, 20-H); 6.25 (s, 1H,
H-3); 6.30 (d, 1H, H-7).
2.2. Preparation of tritiated chlormadinone acetate and
2
.1.4.4. 6-Chloropregna-4,6-dien-17,20-diol-3-one-20-
tritiated chlormadinone
hemisuccinate (12). To a solution of diol 11 in CH2Cl2
was added succinic anhydride and DMAP. The mixture
was stirred at room temperature for 24 h. The reaction was
monitored by TLC. After completion, the solution was
Tritiated CMA was prepared from 6-chloro-17-acetoxy-
pregna-1,4,6-trien-3,20-dione (Amersham TRQ10081); spe-
cific activity = 35 Ci/mmol.

J. Fiet et al. / Steroids 67 (2002) 1045–1055
1049
We prepared tritiated chlormadinone (non-acetylated
chlormadinone) by hydrolysing CMA according the same
procedure as described in Section 2.1.3.
Fluorometer: Time-resolved fluorescence was measured
with a 1234 Delfia Fluorimeter (Perkin-Elmer, Wallac 91945
Courtaboeuf, France). A model 1296-024 platewash device
(
Perkin-Elmer, Wallac) was also used.
2
3
.2.1. Synthesis of 6-chloro-17-hydroxypregna-4,6-dien-
,20-dione (non-acetylated chlormadinone) (8)
2.6. CMA standard solutions
This reagent was purified by celite/ethyleneglycol chro-
matography.
The CMA mother solution (CMA of certified purity: anal-
yse certificate of CMA Sigma Ref C-5145, purity >99%)
and mother solution of metabolite 17␣-acetoxy-6-chloro-
2
.3. Preparation of the immunogens
5-pregnene-3␤-ol-20-one, containing 10 mg of steroid per
The two haptens, CMA-3-CMO, and CMA-20-HS, were
100 ml of pure ethanol, were used to prepare daughter solu-
tions 100 times less concentrated. Alcoholic standard solu-
tions were stored at 4 C, and used to prepare the aqueous
standards necessary for establishing immunocompetition
standard curves. The quantities of CMA (expressed in pg)
put into the wells to establish the standard curves were:
600, 300, 150, 75, 37.5, 18.75, 9.37, and 4.69, which cor-
respond to 1.48, 0.74, 0.37, 0.18, 0.092, 0.046, 0.023, and
coupled to a carrier, bovine serum albumin (BSA), by
the mixed anhydride method of Erlanger et al. [3]. After
coupling, the immunogens were dialysed overnight, then
◦
◦
lyophilised and kept at 4 C.
2
.4. Preparation of anti-CMA-3-CMO/BSA and
CMA-20-HS/BSA antisera
0.011 pmol.
Four New Zealand white female rabbits were immu-
nised with CMA-3-CMO/BSA immunogen and three others
were immunised with CMA-20-HS/BSA, according to the
Vaitukaitis et al. method [4].
Immunogens (0.5 mg) were first injected into each an-
imal with complete Freund’s adjuvant, then, with incom-
plete Freund’s adjuvant every month. Blood was drawn
from the ear marginal vein. Antisera were separated, and
the titers, i.e. dilutions of the antisera which bound 50%
of the biotinylated conjugate tracers, were studied at each
bleeding.
2.7. Biotinylated solutions (tracer solutions)
Stock solutions of biotinylated-3-CMA and 17-CMA
were prepared by, respectively, dissolving 15 and 10.8 mg
each of these conjugates in 50 ml of pure ethanol. Two bi-
otinylated tracer intermediary solutions (SI) were obtained
by diluting the two stock solutions 1/200 in ethanol. Stock
◦
and intermediary ethanol solutions were kept at 4 C. The
stock and intermediate alcoholic biotinylated solutions have
◦
been stored so far for 9 months at 4 C without loss of
activity.
2
.5. Special reagents, devices, and apparatus for TR-FIA
(described previously) [5]
2.8. Coating and saturation of microtiter plates
Microtiter plate wells were coated by adding 250 ␮l of
goat anti-rabbit antibody, diluted 1000-fold in coating buffer.
After washing three times with a washing solution, free well
sites were saturated by adding 300 ␮l of the saturation so-
lution. The microtitration plates were covered with sealing
Microtitration plates: 12 × 8 NUNC (Ref 1244-550 Wal-
lac).
Goat anti-rabbit antibody from Valbiotech (75010
France) (4 mg/2 ml) for coating.
Washing solution: physiological saline solution with
Tween 20 (0.1%).
◦
tape (Corning 430454) and kept at 4 C.
Europium-labelled streptavidin: 0.1 mg/ml, 2.5 ml (Ref
244-360, Wallac), diluted 1/1000 in the following buffer:
BSA 5 g/l + Tween 40, in physiologic saline solution + Na
azide (0.5%), diethylene triamine penta-acetic acid 15 mg/l,
adjusted to pH 7.8 with Tris–HCl.
1
2.9. Procedure for CMA assay in the serum of
CMA-treated patients
2.9.1. Steroid extraction step from serum
Coating buffer: 0.05 M, pH 9.6, anhydrous Na carbonate
Na2CO3) 1.55 g, Na hydrogen carbonate (NaHCO3) 2.97 g,
In order to monitor losses occurring during the extraction,
chromatography, and redissolution steps before immunoas-
say, we added 0.1 ml of a aqueous tritiated H CMA solution
(containing 1500 dpm) to 1 ml of serum in polypropylene
tubes (Sarstedt Ref 55515). After 30 min of incubation with
intermittent shaking at room temperature, extraction was car-
ried out with 8 ml of a mixture of ethyl acetate/cyclohexane
50/50 (v/v), by mixing for 2 min with a multivortex. Cen-
trifugation at 2500 t/min, freeze the aqueous inferior phase,
and transfer the upper organic layer to 13 ml glass tubes
(
3
qsp 1000 ml distilled water, adjusted to pH 9.6 with diluted
HCl.
Saturation solution and assay buffer: pH 7.4, 0.05 M:
disodium hydrogen phosphate (Na2HPO4·2H2O) 7.2 g,
sodium di-hydrogen phosphate (NaH2PO4·2H2O) 1.3 g,
distilled water qsp 1000 ml. Dissolve bovine serum albu-
min 15 g (Sigma A-9647) and Na azide 2 g (Merck, Ref
◦
1
06-688), kept at 4 C.

1
050
J. Fiet et al. / Steroids 67 (2002) 1045–1055
◦
(
CCML Ref 1610010R0299). Evaporate at 37 C under a
tated 350 rpm at laboratory room temperature for 120 min.
The immunoreaction was stopped by washing the wells
three times with washing solution. A solution of Europium-
labelled streptavidin 200 ␮l were then added and the plates
were agitated for 20 min at 350 rpm before washing three
times. The Europium was dissociated by adding 200 ␮l of
enhancement solution to each well, then agitated 350 rpm
for 20 min before measuring fluorescence.
stream of filtered air. We obtained a dry extract containing
all the steroids.
2
.9.2. Chromatographic separation step
The chromatographic step was carried out using 6 ml
polypropylene minicolumns (Ref 57026 Supelco) packed
with a mixture of celite/ethylene glycol (0.7 g/0.35 ml) [6].
We used a particular celite, i.e. ‘celite analytical filter aid’
with very small pore size (median pore size 2.5 ␮m) (from
Lompoc, CA, USA). This celite is conditioned by treat-
ing with cyclohexane extensive washing, drying and kept at
00 C, before packing into chromatographic minicolumns.
The dry extract was redissolved in isooctane (1.5 ml), vor-
3
. Results
3.1. Chemistry
-Carboxymethyloxychlormadinone acetate was obtained
◦
1
3
texed for 30 s, ultrasonicated for 5 min, and vortexed again
for 30 s.
using (aminoxy)acetic acid hemihydrochloride and sodium
acetate in methanol. Coupling with biotinylaminopropylam-
monium was achieved by the mixed anhydride method lead-
ing to the tracer 4.
This organic solution of steroids was layered onto treated
chromatographic minicolumns and passed through with the
aid of negative air pressure (2 kg Pa), using a Visiprep appa-
ratus (Supelco, Saint-Quentin Fallavier, France) (Ref 57250-
U). One millilitre of pure isooctane was added, which passed
through the minicolumns. This volume was not collected
The reduction of the 3-keto-group of 1 was achieved by
using sodium borohydride, leading to alcohol 6 which was
separated by column chromatography from the diol 7.
The trienone 5 precursor for tritiation and non-acylated
chlormadinone 8 were obtained by classical methods.
The hemisuccinate of the 17-OH of 8 could not be ob-
tained. This led us to prepare a hemisuccinate of the alcohol
obtained, by reduction of 20-OH. The dioxolane 9 formed
by protection of the 3-ketone was reduced into diastero-
isomeric diols 10. Elimination of the protecting group led
to 11 which gave, upon acylation with succinic anhydride,
the hapten 12. Finally, 12 was coupled to the biotinylation
reagent 3 leading to the tracer 13.
(
fraction 1, Fig. 1).
We then added 4 ml of pure isooctane, this volume was
passed through the minicolumns, collected and evaporated
to dryness. It was then redissolved in 1 ml of BSA/phosphate
buffer with the aid of vortexing and ultrasons. This aqueous
solution contained the CMA without CMA polar derivatives.
The solution was immunoassayed by TR-FIA.
2
.9.3. Competition time-resolved fluoroimmunoassay
of CMA
The assays were done in duplicate. The assay was carried
out by adding 100 ␮l of the CMA chromatographic eluates
of samples (in duplicate) or standards (in duplicate) to each
coated well, followed by 50 ␮l of, the appropriate dilution of
the labelled biotine tracer, followed by 50 ␮l of anti-CMA-
rabbit antiserum. Plates were incubated while being agi-
3.2. Characterisation of the anti-CMA antisera
3.2.1. Titers
Dilutions of the anti-CMA-3-CMO/BSA and anti-CMA-
17-HS/BSA antisera used in the TR-FIA were, respectively,
Fig. 1. Ethylene glycol celite chromatographic diagram of CMA ( ), 17␣-acetoxy-6-chloro-5-pregnene-3␤-ol-20-one (3␤-hydroxy CMA metabolite:
and non-acetylated chlormadinone ( ) with mean percent of recoveries in 10 experiences.
)

J. Fiet et al. / Steroids 67 (2002) 1045–1055
1051
−
4
0
1
0
.5 × 10 , corresponding to a final dilution of 0.125 ×
(0.2)/63.1% (1.3), 6.6% (0.3)/50.1% (0.9), 3.7% (0.1)/38.4%
(0.9) of the binding of the zero standard (Fig. 2).
−4
−4
0
, and 0.8 × 10 , corresponding to a final dilution of
−4
.2 × 10
.
3.3.2. Fluorescence units
3
.2.2. Specificity
Using the anti-CMA-3-CMO/BSA and CMA-20-HS/BSA
antibodies, the percentages of cross-reactivity were, respec-
tively, 100 and 100% for CMA; 100 and 0.1% for 17␣-
acetoxy-6-chloro-5-pregnene-3␤-ol-20-one (CMA metabo-
lite); 3.89 and <0.01% for cyproterone acetate; <0.01 and
% for delta-4-androstenedione; <0.01 and 0.87% for pro-
gesterone; <0.01 and <0.1% for pregnenolone and <0.01
and 2.19% for 17-OH progesterone.
Fluorescence units range between 165,000 and 5882 in
the CMA standard curve established with the anti-CMA-3-
CMO/BSA antiserum and between 79,500 and 31,546 in the
CMA curve established with the anti-CMA-20-HS/BSA.
3.3.3. The lowest detection limits for CMA on the standard
1
curves [7]
The lowest detection limits for CMA on the standard
curves (the least quantity of CMA that significantly dis-
placed the tracers, i.e. equivalent to B0 or mean binding
of tracer in the absence of standard—2S.D.) were found to
be 1.65 and 11.5 pg per well on the standard curves estab-
lished with the anti-CMA-3-CMO/BSA antiserum and with
the anti-CMA-20-HS/BSA antiserum, respectively.
3
3
.3. Characterisation of the CMA standard curves
.3.1. Percentages of mean bindings
Percentages of mean bindings (S.D.) (n = 12) of the
tracers as reported in the CMA standard curves established
with the anti-CMA-3-CMO/BSA antiserum and with the
anti-CMA-20-HS/BSA antiserum for each standard (0,
3.3.4. The sensitivity of the CMA standard curves
The sensitivity of the CMA standard curves was estab-
lished as the quantities of CMA per well which displaced 20,
50 and 80% of the tracers. With the anti-CMA-3-CMO/BSA
antiserum, these displacements were obtained with 9.38,
21.81 and 87.03 pg per well of standard CMA.
4
.69, 9.37, 18.75, 37.50, 75, 150, 300 and 600 pg) were,
respectively, 100% (1.1)/100% (2.2), 90.9% (0.8)/99.5%
1.4), 80% (0.8)/96.5% (1.6), 59.9% (1.3)/91.5% (2.8),
9.1% (0.5)/84.6% (1.3), 22.7% (0.2)/74.9% (1.1), 12.4%
(
3
Fig. 2. Dose–response curve of the CMA TR-FIA using two anti-CMA antibodies. Mean ± 2S.D. (n = 12).

1
052
J. Fiet et al. / Steroids 67 (2002) 1045–1055
3
3
On the standard curve established with the anti-CMA-
0-HS/BSA antiserum, the same displacement of the corre-
sponding tracer was obtained with much higher quantities of
CMA, 63.64, 329.8 and more than 600 pg per well of CMA.
overloaded by H tritiated non-acetylated CM tracer. H
tritiated non-acetylated CM was not eluted in either the
first fraction of 6 ml of pure isooctane, or in the second
isooctane/ethyl acetate mixture (95/5, v/v) 6 ml fraction.
The tritiated CM was eluated in a third, more polar fraction
consisting of a mixture of isooctane/ethyl acetate (85/5,
v/v), mean recovery = 85% CV = 7.3% (n = 10) (Fig. 1).
2
3
1
.4. Celite chromatographic separation of CMA from
7α-acetoxy-6-chloro-5-pregnene-3β-ol-20-one and from
non-acetylated chlormadinone (CM)
The chromatographic elution profile of CMA was
established using H tritiated CMA as described in
4. Validation of the CMA assay method using the
CMA-3-CMO/BSA antiserum
3
Section 2.1.3.2. After extraction of serum overloaded with
3
H CMA, we redissolved the dry extract in 1.5 ml of isooc-
4.1. Specificity of the CMA TR-FIA
tane. This volume is layered onto the chromatographic col-
umn and penetrate into the chromatographic celite–ethylene
glycol phase with the aid of a slight negative pressure
Using celite chromatography [6,8], the two main metabo-
lites of CMA, 17␣-acetoxy-6-chloro-5-pregnene-3␤-ol-20-
one, and non-acetylated CM were eluted at different times,
following CMA chromatographic elution (Fig. 1); conse-
quently, they do not interfere in the CMA assay.
(2 kg Pa). We added 1 ml of pure isooctane, which is not
collected. Further addition of pure isooctane 4 ml, ml by
ml is collected ml per ml and the radioactivity measured
3
in each ml of eluate. It can be seen that almost all the H
Endogenous steroids which might be eluted in the CMA
isooctane elution fraction: pregnenolone, progesterone, and
tritiated CMA was eluated in these 4 ml (Fig. 1).
4
During the assays the entire 4 ml of isooctane were added
at one time, and these 4 ml collected in the same tube. This
volume is evaporated to dryness, redissolved in phosphate
gelatin buffer before TR-FIA, as described previously in
ꢀ -androstenedione, did not cross-react at all with the anti-
CMA-3-CMO/BSA. Moreover, serum levels of these ovarian
and adrenal steroids are rather low in menopausal women.
3
Section 2.1.3. The individual recovery of H CMA was de-
4
4
.2. Accuracy
termined on each serum sample, and after TR-FIA, we took
into account the individual recovery of each sample to cal-
culate the serum CMA concentrations. The mean recovery
.2.1. Recovery experiments
Recovery experiments were carried out with serum sam-
3
of the H CMA added to 300 different menopausal women
ples from untreated menopausal women overloaded with
three different quantities of CMA (Table 1). For the three
CMA overloads, 250, 1500, and 3000 pg/ml, percentages
of recovery were, respectively, 98–110, 95.2–106, and
serum samples was 85%, with a CV = 6.2% (n = 300).
“
We investigated possible interference by the main
metabolite of CMA, the 3␤-hydroxylated derivative of
CMA”—17␣-acetoxy-6-chloro-5-pregnene-3␤-ol-20-one-
in the CMA TR-FIA, by establishing the chromatographic
immune profile of this metabolite as follows: the sera of
several untreated menopausal women were overloaded by
high levels of 17␣-acetoxy-6-chloro-5-pregnene-3␤-ol-20-
one (20 ng/ml), then extracted and chromatographed with
pure isooctane, as just previously reported. After layering
95.6–101.7%.
4.2.2. Dilution test
The measurements of CMA levels was carried out in sera
taken at several times after a single oral ingestion of CMA by
menopausal women. CMA were assayed in undiluted sera,
and in sera diluted 1/2, 1/4, 1/8 (Table 2).
Nine serum samples containing various CMA concentra-
tions serially diluted with the serum of untreated menopausal
1
.5 ml of 17␣-acetoxy-6-chloro-5-pregnene-3␤-ol-20-one
isooctane solution, onto the top of the chromatographic
minicolumn, we applied a slight vacuum and then 1.5 ml
of the compound was passed through the column. Then we
added 6 ml of pure isooctane, followed by 6 ml of a mixture
of isooctane/ethyl acetate (95/5, v/v), and collected each
eluate ml. Each eluate is evaporated to dryness and redis-
solved in phosphate gelatin buffer and immununoassayed
by TR-FIA, using anti-CMA-3-CMO/BSA and biotinylated
CMA. No immunoassayable steroid could be detected in
the 4 ml elution fraction of CMA. The immunoassayable
Table 1
Measurements of CMA in menopausal women’s sera overloaded with
three quantities (250, 1500, and 3000 pg/ml) of CMA
Level 1
Level 2
Level 3
2
275
248
53
1463
1592
1481
1421
1428
1463
2931
2933
2868
2976
3092
2914
2
2
2
45
55
50
1
7␣-acetoxy-6-chloro-5-pregnene-3␤-ol-20-one was col-
lected in the isooctane/ethyl acetate (95/5, v/v) eluate, with
a mean recovery of 85% CV = 7.1% (n = 10) (Fig. 1).
The possible interference of non-acetylated CM was
investigated using menopausal women’s serum samples
254 ± 10.7
1474 ± 61.8
2952 ± 76.7
Means of levels, S.D.

J. Fiet et al. / Steroids 67 (2002) 1045–1055
1053
Table 2
Dilution test
4.7. Stability of CMA in samples of menopausal
woman’s sera
Dilution S1
S2
S3
S4
S5
S6
S7
S8
S9
The TR-FIA of the sera 1 week after receiving of samples
and 3 months later gave no statistically different results.
Equally the thawing and freezing carried out five times on
eight sera do not influence the initial CMA results.
1
1
1
1
/1
/2
/4
/8
1632 2960 2817 1037 345 240 129 162 108
1373 2813 2527 1016 328 222 105 146 90
1382 2907 2470 977 324 238 ND 157 ND
1555 2690 2679 1085 ND ND ND ND ND
CMA levels in pure and diluted sera obtained from blood samples taken
at five periods of the pharmacokinetic study following the oral ingestion
of CMA. Mean ± S.D. (CV).
4
.8. CMA serum level variations after oral intake
of CMA + estradiol
In this context, the peak CMA concentration and the time
of peak concentration measured in one patient were, respec-
tively, 3 ng/ml and 2.5 h.
women, revealed the measured values on diluted sera to be
between 84 and 107% of the CMA values measured in the
undiluted sera.
4
.3. Detection limit
5. Discussion
The detection limit on the standard curve depends on
Synthesis of the CMA-3-CMO and CM-20-hemisuccinate
haptens, followed by coupling to BSA permitted us to pro-
duce two kind of antisera of different specificity and affinity,
using the corresponding synthesised biotinylated tracers, as
those previously described in [9–14].
Compared to GLC determination of CMA in plasma [1],
the sensitivity of the serum CMA TR-FIA is considerably
higher, since 51 pg of CMA may be measured in 1 ml of
serum with a CV of 4%.
Nevertheless, measurement of CMA in tissues and edi-
ble muscle by GC/MS, has been published [15,16]. Accord-
ing to the authors, the limit of detection was, “in the most
favourable case,” in the order of 1 ng/g.
the lowest quantity which may be statistically measured
on the standard curve of the CMA TR-FIA. It was found
to be 1.65 pg per well. The detection limit of CMA in
serum depends on the methodology used. With 1 ml of ex-
tracted serum, 1 ml of phosphate gelatin buffer used to redis-
solve steroid after chromatography, 0.1 ml of aqueous chro-
matographic solution assayed by TR-FIA (as described in
Sections 2.1–2.9), and a mean recovery of 85% as found,
the calculated detection limit was 19.4 pg/ml.
4
.4. The quantification limit
The metabolism of the CMA was studied in vitro in
the human and in male rat liver, yielding 17␣-acetoxy-6-
chloro-3␤-hydroxypregna-4,6-diene-20-one as the major
metabolite [17].
The quantification limit or ‘lower limit of quantification’
was determined in the serum of menopausal women over-
loaded with low quantities of CMA. We overloaded serum
with 50 pg of CMA/ml and assayed this serum 10 times,
according to the method described in this paper, specifi-
cally, 1 ml of extracted serum and 1 ml of phosphate gelatin
buffer, to redissolve the steroid after the chromatographic
step. We obtained a mean concentration of 51 pg/ml ± 2.27
In vivo, in male rabbits, it has been shown that fol-
3
lowing H tritiated CMA oral administration, intact
3
3
H tritiated CMA and H tritiated polar metabolites
of which 17␣-acetoxy-6-chloro-3␤-hydroxypregna-4,6-
diene-20-one (3␤-hydroxy metabolite) and 17␣-acetoxy-6-
chloro-2␣-hydroxy-4,6-pregnadiene-3,20-dione (2␣-hydro-
xymetabolite) were eliminated in the urine of the animals
(
CV = 4%). The quantification limit was 51 pg/ml.
4
.5. Reproducibility
[18]. In the reported publication, separation of intact CMA
from its polar metabolites was carried out by a gradient elu-
tion chromatography on celite using isooctane + ethylene
glycol as the eluting solvents. In this paper, intact CMA
was the first tritiated steroid to be eluted from the chro-
matography column and was well separated from the polar
metabolites of the 3␤-hydroxymetabolite and the 2-hydroxy
metabolite.
The CV of the results of all the duplicates of the assayed
CMA levels were between 0 and 5.4% with a mean of 0.91%
(
n = 300).
Interassay precision: CV were 4.5, 3.6 and 2.6%, respec-
tively, for CMA levels of 250, 1500 and 3000 pg/ml (n = 8),
in sera assayed in each run of assays.
An in vivo study of the metabolism of H and 14C CMA in
baboons showed a moderate deacylation of the 17␣-acetate,
with a predominant elimination of glucosiduronate conju-
gates in urine and bile [19].
3
4
.6. Stability of the tracer ‘biotine-3-CMO-CMA’
The stock solutions and intermediary alcoholic biotiny-
◦
lated solutions have been stored for 9 months so far at 4 C,
without loss of activity.
In our experience [6,12,20,21] of the separation of exoge-
nous and endogenous steroids on celite column chromato-

1
054
J. Fiet et al. / Steroids 67 (2002) 1045–1055
graphy using a mixture of isooctane + ethyl acetate of in-
creasing polarity prior to RIA, it was easy to separate intact
CMA from its major metabolite, 17␣-acetoxy-6-chloro-3␤-
hydroxypregna-4,6-diene-20-one (3␤-hydroxy metabolite),
and from the non-acetylated CMA (CM) (Fig. 1).
This chromatographic elution step prior to CMA im-
munoassay is of the utmost importance, because it avoids
obtaining CMA serum levels that are too high and inaccu-
rate, particularly in the last samples of the pharmacokinetic
profile.
References
[1] Bopp RJ, Murphy HW, Nash JF, Novotny CR. GLC determination
of chlormadinone acetate in plasma. J Pharm Sci 1972;61:1441–4.
[2] Boudi A, Fiet J, Galons H. Preparation and acylation of a new
biotinylation reagent. Synth Commun 1999;29:3137–41.
[
3] Erlanger BF, Borek F, Beiser SM, Liberman S. Steroid–protein
conjugates. Preparation and characterisation of conjugates of bovine
serum albumin with progesterone. J Biol Chem 1957;234:1090–4.
4] Vaitukaitis J, Robbins JB, Nieschlag E, Ross GT. A method for
producing specific antisera with small doses of immunogen. J Clin
Endocrinol Metab 1971;33:988–91.
[
[
The percentage of recovery of tritiated CMA added to
each serum sample before extraction was rather elevated
5] Fiet J, Giton F, Boudi A, Boudou Ph, Soliman H, Villette JM,
et al. Development of a sensitive and specific new plasma 4-
androstene-3,17 dione time-resolved fluoroimmunoassay (TR-FIA)
Steroids 2001;66:609–14.
(
85%), and reproducible (CV = 6%).
The choice of the anti-CMA antibodies developed was
[
6] Fiet J, Gosling JP, Galons H, Boudou Ph, Aubin Ph, Bélanger A, et
al. Coordinated radioimmunoassays for eight plasma steroids relevant
to the investigation of hirsutism and acne in women. Clin Chem
guided by the considerably greater affinities of the CMA-3-
CMO/BSA antiserum for CMA, compared to 20-HS-BSA
antiserum (Fig. 2).
1994;40:2296–305.
The 100% cross-reactivity of the CMA-3-CMO/BSA an-
tiserum with 17␣-acetoxy-6-chloro-3␤-hydroxypregna-4,6-
diene-20-one (3␤-hydroxy metabolite) was not a problem,
because this metabolite was separated from intact CMA with
excellent resolution.
Although eluted in the CMA isooctane chromatographic
fraction, the apolar endogenous steroids (pregnenolone, pro-
gesterone, and androstenedione) could not interfere in our
CMA immunoassay because CMA-3-CMO/BSA antiserum
do not cross-react with these endogenous steroids.
The high specificity and sensitivity of the method allowed
us to attain high accuracy, as proved by the overload test,
and the dilution tests.
[7] Abraham GE. Radioimmunoassay of steroids in biological materials.
Acta Endocrinol (kbh) 1974;183(Suppl):1–41.
[
8] Abraham GE, Buster JE, Lucas LA, Corrales PC, Teller RC.
Chromatographic separation of steroid hormones for use in
radioimmunoassay. Anal Lett 1972;5:509–17.
[
9] Gomez-Sanchez CE, Leon LM, Gomez-Sanchez EP. Biotin-
hydrazide derivatives for the development of steroid enzyme-linked
immunoassays. J Steroid Biochem Mol Biol 1992;43:523–7.
[10] Boudi A, Giton F, Galons H, Eulry B, Villette JM, Soliman H,
et al. Development of a plasma 17␣-hydroxyprogesterone time-
resolved fluorescence immunoassay involving a new biotinylated
tracer. Steroids 2000;65:103–8.
[
11] Fiet J, Boudi A, Giton F, Villette JM, Boudou Ph, Soliman
H, et al. Plasma 21-deoxycortisol: comparison of time-
a
resolved fluoroimmunoassay using a biotinylated tracer with a
radioimmunoassay using 125 iodine. J Steroid Biochem Mol Biol
Indeed, interassay precision with a CV < 4.2% in three
quality control sera, respectively, containing 250, 1500, and
2000;72:55–60.
[
[
[
12] Fiet J, Giton F, Boudi A, Boudou Ph, Vexiau P, Galons H, et al.
Highly sensitive and specific time-resolved fluoroimmunoassay (TR-
FIA) of cyproterone acetate and free cyproterone. Steroid Biochem
Mol Biol 2001;75:315–22.
13] Fiet J, Giton F, Boudou P, Villette JM, Soliman H, Morineau G, et
al. A new specific and sensitive time-resolved fluoroimmunoassay
of 11-deoxycortisol in serum. J Steroid Biochem Mol Biol 2001;77:
3
000 pg/ml may be considered very satisfactory.
The lowest CMA plasma level, which we took as the
detection limit, was 51 pg/ml with a CV of 4%.
Thus, this TR-FIA appears to be an interesting new tech-
nique in the field of endogenous and now exogenous steroid,
such as CMA-immunoassay, using a biotinylated steroid as
the tracer [9–14] and Europium-labelled streptavidin, for
fluorimetric detection, which allows higher sensitivity and
better reproducibility, compared with RIA methods [9–14].
This TR-FIA is now being applied to the measurement
of serum CMA levels in menopausal women after an oral
intake of CMA (2 mg) + estradiol (2 mg).
143–50.
14] Fiet J, Giton F, Boudi A, Boudou Ph, Soliman H, Villette JM,
et al. Plasma 17-OH pregnenolone: comparison of a time-resolved
fluoroimmunoassay using a new tracer 17-OH pregnenolone-3-
oxyacetyl-biotine with a radioimmunoassay using 125I 17-OH
pregnenolone-3-hemisuccinate-histamine. Steroids 2001;66:81–6.
15] Daeseleire EA, De Guesquiere A, Van Peteghem CH. Derivatization
and gas chromatographic–mass spectrometric detection of anabolic
steroid residues isolated from edible muscle tissues. J Chromatogr
[
The first results found with a CMA Cmax of 3.5 ng/ml
and a time of peak concentration of 2.5 h in a menopausal
woman after oral intake of CMA (2 mg) + estradiol (2 mg),
should be considered in comparison with published results
obtained by GLC in the plasma of women (n = 10) treated
with CMA (2 mg) without estradiol, Cmax = 1.98 ng/ml, and
time of peak 2.5 h [22].
1991;562:673–9.
[
[
[
[
16] Daeseleire EA, De Guesquiere A, Van Peteghem CH. Multiresidue
analysis of anabolic agents in muscle tissues and urine of cattle by
GC-MS. J Chromatogr Sci 1992;30:409–14.
17] Handy RW, Palmer KH, Wall ME, Piantadosi C. The metabolism
of antifertility steroids. The in vitro metabolism of chlormadinone
acetate. Drug Metab Disposition 1974;2:214–20.
18] Abe T, Kambegawa A. Urinary and biliary metabolites of 17␣-
acetoxy-6-chloro-4,6-pregnadiene-3,20-dione in the rabbit. Chem
Pharm Bull 1974;22:2824–9.
19] Honjo H, Ishihara M, Osawa Y, Kirdani RY, Sandberg AA. The
metabolic fate of chlormadinone acetate in the baboon. Steroids
1976;27:79–98.
Acknowledgments
We thank Perkin-Elmer for having made a Delfia 1234
Fluorometer available to us.

J. Fiet et al. / Steroids 67 (2002) 1045–1055
1055
[
20] Fiet J, Gourmel B, Villette JM, Brerault JL, Julien R, Cathelineau
G, et al. Simultaneous radioimmunoassay of androstenedione,
dehydroepiandrosterone and 11␤-hydroxyandrostenedione in plasma.
Hormone Res 1980;13:133–49.
in adrenal late-onset 21-hydroxylase deficiency suggest a mild
defect of the mineralocorticoid pathway. J Clin Endocrinol Metab
1989;68:542–7.
[22] Nash JF, Bopp RJ, Shreve RW, Scholz NE. A study of chlor-
madinone acetate concentrations in human plasma following the oral
administration of tablets. Curr Ther Res Clin Exp 1971;13:407–11.
[21] Fiet J, Gueux B, Raux-Demay MC, Kuttenn F, Vexiau P, Brerault
JL, et al. Increased plasma 21-deoxycorticosterone (21-DB) levels