129
Clinicn Chimica Acta, 103 (1980) 129-133 @ Elsevier/North-Holland Biomedical Press
CCA 1311
SPECIFIC QUANTITATION OF PLASMA MEDROXYPROGESTERONE ACETATE BY GAS CHROMATOGRAPHY/MASS SPECTROMETRY
G. PHILLIPOU
* and R.G. FRITH
Endocrine Laboratory, Department of Obstetrics and Gynaecology, Hospital, Woodville, South Australia 5011 (Australia)
The Queen Elizabeth
(Received July 30th, 1979)
Summary Investigation of various derivatives (0-methyloxime, trifluoroenol acetate and t-butyldimethylsilyl enol ether) coupled with the efficient preparation of a trideuterated analogue of medroxyprogesterone acetate, has allowed the development of a rapid and accurate assay for its quantitation in plasma by isotope dilution gas chromatography/mass spectrometry. High oral doses (2.8 g weekly) of medroxyprogesterone acetate are shown to lead consistently to plasma levels of <16 ng * ml-i. Introduction (17a-acetoxy-6ar-methyl-4-pregnene-3,20Medroxyprogesterone acetate dione, MPA II) is a synthetic progestin which has been shown to be effective both in the control of fertility [l] and the management of endometrial adenocarcinoma [ 21. Although plasma levels have been conveniently measured by radioimmunoassay techniques [ 3-81, the specificity of several of these procedures has been questioned [ 6-81. Alternatively, while two almost identical methods based on electron-capture gas chromatography (GC) have been reported [ 9,101, they are limited in their sensitivity, due to interference from endogenous cholesterol. In this communication we report our investigations aimed at establishing a rapid and accurate assay of plasma MPA based on stable-isotope dilution gas chromatography-mass spectrometry (GC/MS). Materials and methods Medroxyprogesterone (17a-hydroxy-6or-methyl-4-pregnene-3,2Odione, MP I) and MPA were purchased from Steraloids (Wilton, NH, U.S.A.). [‘H6]Acetic * To whom correspondence
should
be addressed.
130
anhydride (>99% 2H6) was obtained from Aldrich (WI, U.S.A.). [2H3]Medroxyprogesterone acetate (III) MP (0.1 g) and Amberlite IR-120 resin (0.1 g) in [ 2H6]acetic anhydride (1 ml) were stirred at room temperature for three days. The mixture was then poured into water and extracted with ether. The ethereal extract was washed with saturated aqueous sodium bicarbonate, brine and then dried (anhy. Na2S04). Removal of the ether followed by preparative thin layer chromatography (TLC) (chloroform/acetone, 9 : 1, v/v) and recrystallization (ether : n-hexane) gave [‘H3]MPA (60%) m.p. 204-205°C [elD + 58°C (lit [ 111 207-209’C, [e]D + 61°C) homogeneous by TLC and GC. Mass spectrometric analysis showed it to be 5.5% ‘Hz and 94.5% 2H3. Extraction procedure This was essentially as described by Kaiser et al. [9], except that in most cases only 0.2-0.5 ml of plasma was used. Derivatisa tion The preparation of the perfluoroacyl and t-butyldimethylsilyl ether derivatives of MPA were as described previously [9,12].
(BDMS) enol
Gas chromatography/mass spectrometry GC/MS was conducted using a Hewlett-Packard 5992B GC/MS. Conditions: injection port, 250°C; 2 m X 2 mm ID glass column packed with 1.6% OV-101 at 25O”C, helium flow, 25 ml - minei; selected-ion monitoring m/z 482.3 and 485.3 dwell time, 500 m/s. Results and discussion While it has been reported that the tertiary 17a-hydroxyl group of a steroid can be acetylated readily at room temperature using acetic anhydridelptoluenesulphonic acid [ 131, we found it unsuited to MP, due to extensive enol acetylation. Consequently, a method was developed in which a very weak, acidic catalyst was employed [14] and which is described fully in the experimental section. Under these conditions competitive formation of the enol acetate was only a minor reaction. MPA has been previously derivatised and chromatographed as its heptafluoroenol butyrate or O-methyl oxime [9,15]. We found that it also readily formed the BDMS enol ether in the usual manner [ 121. The mass spectra of these derivatives all show strong molecular ions and a fragmentation pattern associated with the loss of acetic acid followed by expulsion of the -COCHs side-chain [16] (Scheme). The intense fragment ions at M-103 cannot be used for quantitation, since they are common to both MPA and [‘HJ]MPA derivatives. A major problem in the GC/MS quantitation of plasma MPA is the interference by cholesterol which has the same nominal mass and is present in gmol quantities. Choice of either the trifluoroenol acetate or BDMS enol ether derivative of MPA likewise forms a derivative of the same mass for cholesterol. While this occurrence is not true for the O-methyl oxime, the signal noise at M’ and the chromatographic properties are inferior to those of the two previous derivatives.
131 1'
1'
33 I
R=H
I[
R =
OCOCH,
m
R
OCOC’H,
q
SCHEME
Since the BDMS ether of cholesterol has a negligible molecular ion typical, in general, of steroid BDMS ethers [17], and the BDMS enol ether of MPA possesses an intense molecular ion, it was felt that interference from cholesterol would not be a problem. When [ ‘H,]MPA was derivatised to its respective BDMS enol ether however, complete loss of the label occurred, presumably by some mechanisms of transacetylation with the potassium acetate catalyst. Treatment of cholesterol propionate under these conditions [ 121 did not result in formation of cholesterol acetate. It appears that this type of transacetylation is influenced by the presence of the 20-0~0 group, which is being investigated in more detail. Previously, the interference of cholesterol heptafluorobutyrate eluting in a high relative concentration before the respective MPA derivative on OV-17 phase in GC procedures, imposed a restriction on the sensitivity of 20 ng * ml-l when applied to human plasma [lo]. If OV-101 is used as the GC phase, however, perfluoroacyl derivatives of MPA have a shorter retention time than the corresponding cholesterol derivative. Under these conditions, MPA is readily resolved and measured in the presence of cholesterol (Fig. 1). It is apparent from Fig. 1 that formation of the trifluoroacetate derivative of MPA also produces an enol isomer. The level of this isomer, however, was consistently <8%, and did not vary appreciably between derivatisations. Selected-ion monitoring of the molecular ions for MPA (m/z 482) and [*H3]MPA (m/z 485) trifluoroacetates using OV-101 as a GC phase allowed quantitation of plasma MPA with a lower detection limit of
132
Fig. 1. GC resolution of MPA (A) and cholesterol (B) as their trifluoroacetate en01 isomer of MPA while peak (D) is a rearrangement product of cholesterol. m/z 482 + 485 for a biological sample.
derivatives. Peak (C) is an The trace is a summation of
plasma level8 strongly suggest that the drug is either poorly absorbed or rapidly metabolized, an argument which has been previously advocated by Martin and Adlercreutz [ 71. Substantial investigation on the MPA remission of endometial adenocarcinema has suggested a therapeutic range of >90 ng - ml-1 of MPA-related material [2]. The assessment of plasma MPA levels by a very unspecific RIA procedure [ 31, however, has severely restricted the general applicability of this research to most laboratories. The low plasma levels (Table I) found in this
TABLE
I
PLASMA
LEVELS
a ’ d d d d
a Intramuscular.1 c oral,
IN PATIENTS
WITH CARCINOMA
46 33 14.6 10.2 10.1 3.6 15.7
X 900 ma weekly, 6 days after dose. 3 X 800 mg weekly, 2 days after last dose. 3 X 100 mg daily, for two months previously. sampled at 10.00 a.m. after dose at 8.00 am. 4 X 100 mg dally, 1-2 weeks previously, sampled at 10.00 a.m. after dose at 8.00 a.m.
b Intramuscuti, d oral,
* ml-l)
Plasma level
Patient male male b female female female female female
OF MPA (ng
133
study - on three times the oral dose suggested by Bonte [Z] - implicate a possible important role for MPA metabolites. In conclusion, it appears that the described GC/MS procedure fulfills the criterion as both a rapid and accurate assay of plasma MPA, suitable for assessment of patients on long-term treatment for carcinoma. Acknowledgement The assistance of Dr. P. Weir (Mercy Maternity providing plasma samples is gratefully acknowledged.
Hospital,
Melbourne)
in
References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Nash, B.A. (1975) Contraception 12, 377-393 Bonte. J., Decoster, J.M., Ide. P. and BiBiet. G. (1978) Gynaecol. Oncol. 6. 60-75 Cornette, J.C.. Kirkton, K.T. and Duncan, G.W. (1971) J. Clin. Endocrinol. 33.459-466 Royer. M.E., Ko, H., CampbelL J.A., Murray, H.C.. Evans, J.S. and Kaiser, D.G. (1974) Steroids 23. 713-730 Hiroi, M.. Stanczyk. F.Z., Goebelsmann. M., Brenner. P.F., Lumkin. M.E. and Mitchell, D.R. Jr. (1975) Steroids 26. 373586 Shrimanker, K., Saxena, B.N. and Fotherby. K. (1978) J. Ster. Biochem. 359-363 Martin, F. and Adlercreutz, H. (1977) In Pharmacology of Steroid Contraceptive Drugs (Garattini, S. and Berendes. H.W., eds.), p. 99. Raven Press, New York Laatikainen, T., Nieminen, U. and Adlercreutz, H. (1979) Acta Obstet. Gynecol. Scand. 58.95-99 Kaiser, D.G., Carlson. R.G. and Kirton, K.T. (1974) J. Pharm. Sci. 63, 420425 Rossi. E., Pascale, A., Negrini, P., Frigerio, A. and Castegnaro, E. (1979) J. Chromatogr. 169. 416421 The Merck Index. Merck and Co. Inc., Rahway, NJ, U.S.A. (1968) p. 648. 8th Edn Blair, I.A. and PhiBipou. G. (1978) J. Chromatogr. Sci. 16.201-203 Minion. H.. Wilson, E., Wendler, N.L. and Tishler, M. (1952) J. Am. Chem. Sot. 74. 5394-5396 Suzuki, T., Yamado, N., Wada, T.. Sawai. N. and Chuma. K. (1962) Chem. Abstr. 58.91896 Adlercreutz, H., Nieminen. U. and Ervast. H.S. (1974) J. Ster. Biochem. 5. 619-626 FehIhaber. H.W., Lenoir. D. and Welzel, P. (1975) Adv. Mass Spectrom. 5.689-692 PhiIIipou, G.. Bigham, D.A. and Seamark, R.F. (1975) Steroids 26, 516-524