Comparison of pharmacodynamic properties of various estrogen formulations

Comparison of pharmacodynamic properties of various estrogen formulations C.

ANN

MASHCHAK

ROGER10

A.

RYOKO

DOZONO-TAKANO

PETER

EGGENA

ROBERT PAUL

LOB0

M.

NAKAMURA

F. BRENNER

DANIEL

R.

Los

California

Angeles,

MISHELL,

JR.

A group of 23 healthy postmenopausal women received one or more Pweek courses of daily administration of the following estrogen preparations: piperazine estrone sulfate (Ogen), 0.3, 0.625, 1.25, 2.5, and 5.0 mg; micronized estradiol (Estrace), 1, 2, and 10 mg; conjugated estrogens (Premarin), 0.3, 0.625, 1.25, and 2.5 mg; ethinyl estradiol (Estinyl), 10 and 20 w; and diethylstilbestrol, 0.1 and 0.5 mg. Each dosage of each formulation was ingested by three women. In those women who received more than one dosage, each course was separated by a drug-free interval of at least 4 weeks. Pretreatment and posttreatment levels of follicle-stimulating hormone (FSH), luteinizing hormone (LH), corticosteroid-binding globulin-binding capacity, sex hormone-binding globulin-binding capacity, angiotensinogen, estrone, and estradiol were determined. The relative potency of these five estrogen formulations was determined by parallel line analysis for each of these responses, except LH. On a weight basis, piperazine estrone sulfate and micronized estradiol were equipotent for all responses. Conjugated estrogens suppressed FSH in a fashion equipotent to that of the other nonsynthetic estrogens; however, for all three hepatic parameters, the response was exaggerated twofold to threefold. The synthetic estrogens, diethylstilbestrol and ethinyl estradiol, were relatively more potent on a weight basis for every response and produced the most marked response (fourfold to eighteenfold in excess of their FSH suppression) for the hepatic parameters. (AM. J. OBSTET. GYNECOL. 144:511, 1962.)

ALT H ou G H several different oral estrogen formulations are available for the treatment of postmenopausal estrogen deficiency, there is inadequate comparative information on their pharmacologic effects

From the Department of Obstetrics and Gynecology, University of Southern California School of Medicine, and Women’s Hospital, Los Angeles County-University Southern California Medical Center.

of

Supported by grants from the Ford Foundation (Grant No. 690.065OA) and the American Heart Association and research funds from the Veteralls Administration. Presented in part at the Twenty-seventh Annual Meeting of the Society for Gynecologic Investigation, in Denver, Colorado, March 19-22, 1980. Reprint requests: Daniel R. M&hell, Jr., M.D., Women’s Hospital, 1240 North Mission Road, Los Angeles, California 90033. 0002-9378/82/210511+08$00.80/0

0

1982 The

C. V. Mosby

Co

and their differences in potency. Some comparative evaluation of conjugated estrogens and ethinyl estradiol has been reported, ‘3 ’ but the evaluation of other compounds which are currently used clinically, such as piperazine estrone sulfate, micronized estradiol, and diethylstilbestrol (DES), is still very limited. Estrogen potency can be determined and expressed in terms of several parameters. Estrogens have been shown to produce a dose-related suppression of gonadotropins in postmenopausal women? The hepatic production of two globulins, sex hormone-binding globulin and angiotensinogen, is known to be increased by estrogen in a dose-related manner.” 4 We previously demonstrated that another globulin produced by the liver, corticosteroid-binding globulin, is increased in a dose-related fashion after the administration of different doses of conjugated estrogens to postmenopausal 511

512

Mashchak et al.

November Am.

100

I. IYSZ

J. Obstet.

Gvnecol.

m=a6.1 r =0.95

60

I o.bl

0.1

ESTROGEN

IlO

Id.0

mg

Fig. 1. The log dose-SHBG-BC response curves of piperazine estrone sulfate (o), conjugated estrogens (m), micronized estradio! (v), diethylstilbestrol (7) and ethinyl estradiol (0). women.’ These responses can be separated into two categories, systemic and hepatic. With oral administration, greater concentrations of estrogens reach the liver because of the enterohepatic blood flow, whereas most other organs, such as the pituitary gland, receive estrogens at the lower concentrations present in the systemic circulation. In this study, various dosages of five estrogen formulations were administered to postmenopausal women. Serum levels of follicle-stimulating hormone (FSH), luteinizing hormone (LH), corticosteroid-binding globulin-binding capacity (CBG-BC), sex hormone-binding globulin-binding capacity (SHBG-BC), angiotensinogen, estrone (E,), and estradiol (E2) were measured before and after treatment to compare the relative potencies of these formulations.

Material and methods The study group consisted of 23 healthy postmenopausal women who were at least 1 year beyond a natural (19 subjects) or a surgical (4 subjects) menopause and had serum FSH levels that ranged from 27.5 to 142.4 mIU/ml. Most subjects received more than one dosage regimen, up to a maximum of five. No subject had received exogenous steroid hormones during the 4 weeks prior to any treatment period. The mean age of the subjects who were naturally menopausal was 57 years (range, 51 to 69 years), and that of those who were surgically menopausal was 49 years (range, 25 to 56 years). For the entire group, the mean weight was 63 kg (range, 50 to 123 kg). The mean height was 159 cm (range, 150 to 165 cm). The following formulations

0.1 mg - EQUIVALENTS

l:o OF

PIPEAAZINE

10.0 ESTRONE

SULFATE

Fig. 2. The correlation of the SHBGBC response with all doses of piperazine estrone sulfate (0) from 1.25 mg, conjugated estrogens (9) from 0.625 mg, micronized estradiol (v) from 1.0 mg, DES (v) from 0.1 mg, and ethinyl estradiol (0) from 0.01 mg with each test dose expressed as milligram equivalents of-piperazine estrone sulfate. In this and the-subsequent figures, m is the slope of the regression line and r is the correlation coefficient. and dosages were investigated: conjugated estrogens (Premarin*), 0.3, 0.625, 1.25, and 2.5 mg; piperazine estrone sulfate (Ogent), 0.3, 0.625, 1.25, 2.5, and 5.0 mg; micronized estradiol (Estracet), 1. 2, and 10 mg; ethinyl estradiol (Estinylg), 0.010 and 0.020 mg: and DES,// 0.1 and 0.5 mg. Each dose of the five formulations was administered every morning for 2 weeks to three subjects. All serum samples were obtained in the morning after the subjects had fasted overnight. The pretreatment samples were obtained just prior to ingestion of the first tablet, and posttreatment samples were obtained 24 hours after the eleventh and fourteenth tablets. The blood was allowed to clot; the serum was separated and stored at -20” C until the assays were performed. Determinations of serum FSH, LH. CBG-BC, SHBG-BC, angiotensinogen, E,, and E2 were performed by previously reported radioimmunoassay methods.4. 6-R Statistical analysis. Data sets were analyzed by Rankit analysis to determine whether they followed Gaussian or log-Gaussian distributions. Where applicable, “Ayerst Laboratories, New York, New York. tAbbott Laboratories, North Chicago. Illinois. $Mead Johnson Pharmaceutical Division, Evansdie, diana. &Schering Corporation, Kenilworth, New Jersey. (IEli Lilly & Company, Indianapolis, Indiana.

In-

Pharmacodynamics

Table

I. Pharmacodynamics

Estrogen

j7repuration

Piperazine estrone sulfate (mg) 0.3 0.6 1.25 2. 5

5.0 Co$;gated

estrogen

513

preparations

A SHBG-BC (nW (mean k SEM) -5.7 3.0 10.3 46.7 65.3

formulations

t 2.8 r 2.3 + 7.0 ” 18.4 k 23.9

A CBG-BC (~14 (mean 2 SEM)

Angiotensinogen (nd4 (mean ‘- SEM)

1.0 ? 1.2 -2.4 2 0.4 -0.3 r 0.5 2.7 k 2.6 6.5 +- 2.2

4,600 6,000 9,500

FSH

suf@ssian (W (man f SEM) 6.3 r 10.7 27.8 f 5.2 29.3 -e 7.3 51.0 2 13.3 62.9 t 3.4

QNS QNS

5 1,500 -c 1,400 ” 1,200

estrogens (mg) 6.7 2 3.8 38.0 2 20.0 54.0 T 16.0 104.3 2 26.8

0.625 1.25 2.5

Micronized 1.0

of various

of estrogen

-2.6 0.6 3.3 8.6

f k k k

0.3 0.3 0.4 1.6

6,100 10,000 11,300

900 400

19.9 18.6 55.0 51.5

k 6.8 +- 20.6 2 9.8 * 7.9

QNS 2 2 2

0

estradiol (mg)

2.0

10.0 DES (mg) 0.1 0.5

Ethinyl estradiol (mg) 0.01 0.02

15.7 62.7 90.0

+ 8.1 f 25.4 2 7.0

0.7 k 1.6 5.0 f 1.4 18.2 + 1.1

2,900 6,400 8,400

2 500 + 200 -c 3,300

34.7 54.0 77.0

f 8.4 + 10.7 2 5.5

45.7 109.7

+ 11.6 f 23.1

8.1 f 20.3 f

1.1 1.2

7,700 9,100

2 2,600 r 1,800

23.7 38.7

+ k

71.3 110.7

c 11.8 k 29.8

12.3 5 1.8 12.5 f 3.4

7,500 9,200

2 900 f 3,900

42.6 41.8

_’ 10.4 2 5.3

2.0 9.9

QNS, Quantity not sufficient. the Gaussian distributions were treated by normal parametric methods and the log-Gaussian distributions were treated by log-parametric methods, i.e., t tests and regression analyses. The linear response curve to the log dose for each of the five estrogen preparations was calculated by the least-squares regression method. The four parameters of estrogenicity analyzed in this fashion were serum FSH, CBG-BC, SHBG-BC, and angiotensinogen. For each of these four parameters, the relative estrogen potencies were estimated and a composite regression line was calculated by normalizing parallel responses of the conjugated estrogens, micronized estradiol, ethinyl estradiol, and DES to the piperazine estrone sulfate regression line.gr ‘” For each parameter of estrogenicity, a composite log dose-response curve for serum E, and E, levels with the three nonsynthetic estrogen formulations was calculated.

Results SHBG-BC. Baseline SHBG-BC values were log normally distributed. The mean baseline serum SHBG-BC level was 46 nM; 95% confidence limits, 11.5 to 200 nM. When the dose of each estrogen was plotted against the change in SHBG-BC, a log dose-related increase was demonstrated for the administration of all five preparations (Fig. 1). The log dose-response curves of all five compounds were sufficiently parallel to allow normalization to the piperazine estrone sulfate curve in the following fashion. A relative potency factor for each preparation was determined by the dis-

tance from the reference curve of piperazine estrone sulfate. The potency of micronized estradiol was found to be the same as that of piperazine estrone sulfate, whereas the potency of conjugated estrogens was calculated to be 3.2-fold greater. The synthetic estrogens were markedly more potent. DES was calculated to be 28.4 fold, and ethinyl estradiol was 600-fold, more potent than piperazine estrone sulfate. All estrogen doses were reexpressed in milligram equivalents of piperazine estrone sulfate by multiplying the actual milligram dose by the potency factor. Then a composite regression line was constructed by plotting the estrogen dose in milligram equivalents of piperazine estrone sulfate against the increase in serum SHBG-BC (Fig. 2). This line had a slope (m) of 98.3 nM per milligram equivalent of piperazine estrone sulfate and a correlation coefficient (r) of 0.95. A plot (not shown) of the estrogen dose to the SHBG-BC response demonstrated that the beginning of maximization of response under the conditions of this study occurred with doses greater than 8 mg equivalents of piperazone estrone sulfate. The greatest increase in SHBG-BC was achieved after administration of 0.02 mg of ethinyl estradiol. The intersection of the composite regression line with the abscissa-zero response of SHBG-BC-at the 1 mgequivalent of piperazine estrone sulfate quantifies the lower limit of sensitivity of the SHBG-BC response for these drugs. Doses equal to or less than 1.25 mgequivalents of piperazine estrone sulfate failed to stimulate significant increases in serum SHBG-BC.

514

Mashchak et al.

20

m=15.1 r = 0.98

m =6869

f

r =0.93

Oil 1:o mg -EQUIVALENTS

rd.0 OF

PIPERAZINE

rob.0 ESTRONE

mg - EQUIVALENTS

ll0 OF PIPERAZINE

Id.0 ESTRONE

SULFATE

SULFATE

Fig. 3. The correlation of the CBG-BC response with all doses of piperazine estrone sulfate (e) from 1.25 mg. conjugated estrogens (m) from 0.625 mg, micronized estradiol (D) from 1.0 mg, and DES (v) from 0.1 mg with each test dose expressed as milligram equivalents of piperazine estrone sulfate.

Fig. 4. The correlation of the angiotensinogen response with all doses of piperazine estrone sulfate (a) from 1.25 mg. conjugated estrogens (B) 0.625 mg, micronized estradiol (‘7) from 1.0 mg, DES (T) from 0.1 mg, and ethinyl estradiol (c) from 0.01 mg with each test dose expressed as milligram equivalents of piperazine estrone sulfate.

CBGBC. Baseline serum CBG-BC values followed a Gaussian distribution. The mean baseline serum CBG-BC level was 17.3 k 3.5 pg/dl (mean + SD). A log dose-related increase in serum CBG-BC was demonstrated with the administration of all estrogens, except ethinyl estradiol (Table I). The relative potencies for CBG-BC were calculated and expressed in the same way that they were for SHBG-BC. A composite regression line with a slope of 15.1 pg of CBG-BC per deciliter per milligram equivalent of piperazine estrone sulfate was determined (Fig. 3). The correlation for the composite line was 0.98. The potency of micronized estradiol was 1.9-fold greater than that of piperazine estrone suIfate. Conjugated estrogens were 2.5-fold more potent. The potency of the synthetic estrogen DES was seventyfold that of piperazine estrone sulfate. Although a dose-related response was not demonstrated for ethinyl estradiol, significant (p < 0.05) and nearly equal increases in serum CBGBC were seen with 0.010 and 0.020 mg (9.3 and 11.5 pg/dl, respectively). By estimation, the relative potency of ethinyl estradiol appears to be about l,OOO-fold greater than that of piperazine estrone sulfate. The largest increase in CBG-BC was produced by the administration of 0.5 mg of diethylstilbestrol (35 mg equivalents of piperazine estrone sulfate). A plot of the estrogen dose to the CBG-BC response (not shown) suggested that this value approached a maximized response. The intersection of the composite regression line with the abscissa-zero response of CBG-BC-placed the lower limit of sensitivity at 1.7 mg equivalents of piperazine

estrone sulfate. Doses equal to or less than 1.7 mg equivalents of piperazine estrone sulfate failed to produce a significant increase in serum CBG-BC. Angiotensinogen. Baseline angiotensinogen levels were distributed in a Gaussian fashion. The mean baseline serum angiotensinogen level was 3,400 * 1,600 rig/ml (mean + SD). A log dose-related increase in serum angiotensinogen was demonstrated with all five preparations (Table I). A composite regression line was derived from the relative potencies which had been estimated and expressed just as they had been fol SHBG-BC. The slope (m) of the regression line was 6,869 ng of angiotensinogen per milliliter per milligram equivalent of piperazine estrone sulfate, and its correlation coefficient (r) was 0.93 (Fig. 4). Micronized estradiol and piperazine estrone sulfate were estimated to be equipotent. The potency of the conjugated estrogens was 3.5-fold greater than that of piperazine estrone sulfate. The synthetic estrogens were more potent. DES was calculated to be thirteenfold, and ethinyl estradiol was 232-fold more potent than piperazine estrone sulfate. Although the linear portion of the log dose-response curve began at approximately 0.625 mg equivalents of piperazine estrone sulfate, the calculated angiotensinogen response exceeded the normal baseline range only with administration of 2.2 mg equivalents or greater. FSH. Baseline serum FSH levels were distributed in a Gaussian fashion. The mean baseline serum FSH level was 74 +. 30.0 mILJ (mean k SD). A log dose-related suppression of serum FSH, expressed as the percent-

Pharmacodynamics

of estrogen

formulations

515

100 m=338 r=0.99

P v) iz

/

80

5000

8 8

60

2 Q

m v

40

/

6

Ip

l

.

4000

y 0 F 8

3000 2000

l

.

20

0

.

$$ 0

l

/

1000

.

l I

1.0

0.1

10.0 mg - EOUIVALENTS

mg - EQUIVALENTS

OF

PIPERAZINE

ESTRONE

OF PIPERAZINE

ESTRONE

SULFATE

SULFATE

5. The correlation of FSH suppression with all doses of piperazine estrone sulfate (0) from 0.3 mg, conjugated estrogens (m) from 0.3 mg, micronized estradiol (V) from 1.0 mg, and DES (v) from 1 mg with each test dose expressed as milligram equivalents of piperazine estrone sulfate.

Fig. 6. Correlation of estrone levels produced by all doses of piperazine estrone sulfate (0) from 0.625 mg, conjugated estrogens (m ) from 0.625 mg, and micronized estradiol (v) from 1.0 mg with their potency determined by FSH suppression and expressed in milligram equivalents of piperazine estrone sulfate.

age of baseline, was demonstrated with the administration of each of the formulations, except ethinyl estradiol (Table I). A composite regression line was derived from relative potencies which had been calculated and expressed just as they had been for SHBG-BC. A composite regression line with a slope (m) of 40.8 mIU FSH/ml per milligram equivalents of piperazine estrone sulfate was determined (Fig. 5). The correlation coefficient (r) for the composite line was 0.95. The relative potencies of the natural estrogen preparations piperazine estrone sulfate, micronized estradiol, and conjugated estrogens were nearly equivalent (1 .O, 1.3, and 1.4, respectively). The potency of diethylstilbestrol was 3.8fold greater than that of the reference preparation piperazine estrone sulfate. Although a doserelated response was not demonstrated for the two doses of ethinyl estradiol tested, a significant (p < 0.05) and nearly equal suppression of serum FSH was seen with each dose (42.6% and 41.8% of the pretreatment levels, respectively). The estimated relative potency of ethinyl estradiol appears to be at least eightyfold to 200-fold greater than that of piperazine estrone sulfate. A plot of the estrogen dose to the FSH response (not shown) suggested that maximization of response began with the administration of the 5 mg dose of piperazine estrone sulfate. The lowest dose of piperazine estrone sulfate which produced a significant suppression of FSH (p < 0.05) was 0.625 mg. LH. Baseline serum LH values were distributed in a log normal fashion. The mean baseline serum LH level

was 65 mIU/ml, with 95% confidence level of 36 to 112 mIU/ml. Although the mean posttreatment LH value of 54 mIU/ml was significantly less than baseline (t test, p < O.OOl), a dose-response relationship with serum LH was not demonstrated for any of the estrogen preparations. Composite. A comparison of the sensitivity of the FSH, CBG-BC, SHBG-BC, and angiotensinogen response systems was made in the following manner. First, the estimate of the maximum response was based on the plateau levels demonstrated in a dose-response curve. Then, with use of probit analysis, the 50% response was determined and used to rank the systems according to their sensitivity. The FSH and angiotensinogen systems were the most sensitive, each requiring only 1.7 mg equivalents of piperazine estrone sulfate to produce a 50% response. SHBG-BC was less sensitive, requiring 4.4 mg equivalents for a 50% response: and CBG-BC was the least sensitive, requiring 8.5 mg equivalents. El and E2. The mean baseline serum E, level was 30 pgiml (95% confidence level 25 to 45 pg/ml). The baseline serum Ez level was less than 20 pg/ml in all but two patients, whose baseline levels were 25 and 27 pg/ml. All nonsynthetic estrogen preparations, conjugated estrogens, piperazine estrone sulfate, and micronized estradiol exhibited linear dose-response relationships with respect to serum E, and E, levels. Serum E, and E, levels did not change from baseline after the synthetic preparations were administered. Adminis-

Fig.

Mashchak et al.

516

Table

II. Relative

potency

estimated

according

to four

specific

parameters

of estrogenicity Serum

Estrogen

preparation

Serum FSH

Piperazine estrone sulfate Micronized estradiol Conjugated estrogens DES Ethinvl estradiol

Serum CEG-3C

1.1 1.3 1.4 3.8 (80-200)*

1.0 1.9 2.5 (lglo)*

Serum

SHBG-BC

1.0 1.0 3.2 28 614

angiotensinogen

1.0 0.7 3.5

l?l L) -. ‘1‘)-

*Estimate in the absence of parallelism.

tration of equivalent doses (milligram for milligram) of each of the nonsynthetic preparations produced comparable serum El levels and comparable E2 levels, as well. The E,IE, ratio was greater than 1 after all doses of these nonsynthetic estrogens. The potency determined by the relative suppression of FSH correlated best (0.99) with serum E, levels (Fig. 6). The serum El level correlated nearly as well (0.97) with the potency determined for CBG-BC, but less well (0.78) with that determined for SHBG-BC and least well (0.58) with that determined for angiotensinogen. Serum E, did not correlate as well as El with any of the parameters (Fig. 7).

Comment With piperazine estrone sulfate used as the standard, an estimate of relative potency of conjugated estrogens, micronized estradiol, DES, and ethinyl estradiol was made in four different response systems: SHBG-BC, CBG-BC, angiotensinogen, and FSH. The basis for these estimations was the parallelism of the individual log dose-response curves for each estrogen preparation, confirmed by the high correlation of the composite regressions with linearity (0.95, 0.98, 0.93, 0.95, respectively). Under usual circumstances, log dose response curves are based on a minimum of three dosage levels. In this study, response curves for ethinyl estradioi and diethylstilbestrol were estimated from only two dosage levels because multiple dosages were not commercially available for these compounds. These estimates were accepted because the persistent parallelism strongly suggested that the five estrogen compounds were all subsets of one estrogen family in which each estrogen affected target tissues in a qualitatively similar manner. This concept is given additional support when the available FSH response data of several investigators, reexamined in the same fashion as in this study, showed the same parallelism. With their FSH responses expressed in terms of a percentage decrease from baseline, log-dose response curves were calculated for the piperazine estrone sulfate data of Isaacs and Harvard,” conjugated estrogen data of Geola and associates’ and ethinyl estradiol data of

Wise and associates’* and also of Franchimont and associates.” A relative potency factor was determined and used to normalize each of these data sets to oul piperazine estrone sulfate log dose-response curve. The composite regression line of our study was recalculated with the addition of the data of these investigators (Fig. 8). The slope (m) and correlation of this augmented regression line (40.6 and 0.95, respectively) were strikingly similar to those of this study alone (-10.8 and 0.95, respectively). In our study, the lack of parallelism with ethinyl estradiol prevented accurate estimates of potency from being made in the CBG-BC and FSH systems. This situation may have been due to inadequate sampling ot may have indicated that the portion of the log doseresponse curve examined was unsuitable for analysis. More test dosages, and, possibly, more subjects, will be required to answer this question. Piperazine estrone sulfate was chosen as the reference standard for normalization because it is a single, nonsynthetic estrogen, and it was found to be least potent on a weight basis. Expression of each dose of the various estrogen preparations in milligram equivalents of piperazine estrone sulfate facilitated the evaluation of each of the response systems over a greater range of responses. The Y-week treatment period was chosen because we had previously found that an increase in CBG-BC to static levels was reached in 10 days of daily estrogen treatments.5 Others have shown that, with ethinyl estradiol administration, maximum FSH suppression also occurred in a lo-day period.“, I2 Direct comparison of any other data with ours in order to estimate relative potency (vide infra) requires that the duration of treatment be the same or that static steady-state response levels be reached in each case. In order to obtain the steady-state nadir of serum E, and El. all samples were obtained 24 hours after the previously ingested tablet. The relative potencies with respect to piperazine estrone sulfate are listed in Table II. These data make it possible to rank the estrogen preparations from the

Volume Number

144 5

Pharmacodynamics

of estrogen

formulations

517

m;;62.0 r.rO.97

J

0 2 mg

4

- EOUIVALENTS

6

8

I 10

OF PIPERAZINE

12

ESTRONE

a ZR

14

SULFATE

Fig. 7. Correlation of estradiol levels produced by all doses of piperazine estrone sulfate (0) from 0.625 mg, conjugated estrogens (m) from 0.625 mg, and micronized estradiol (v) from 1.0 mg with their potency determined by FSH suppression and expressed in milligram equivalents of piperazine estrone sulfate. least to the most potent. This ranking order remains the same for all four parameters, although the numerical potency estimates differ, depending on the response measured. The induction of SHBG synthesis by estrogen is a well-known phenomenonP Prior to this study, a doserelated response in postmenopausal women had been shown for conjugated estrogens and ethinyl estradiol by Geola and associates* and Frumar and associates.2 Information about the steady-state response is gained when our data are looked at along with theirs. In their studies, the maximum plateau response after 6 weeks of ethinyl estradiol treatment occurred at a lower dose (10 pg) than after our 2-week treatment period (20 pg). These findings suggest that the steady-state levels of SHBG-BC, unlike those for CBG-BC, are not reached in 2 weeks of estrogen treatment. On the other hand, evidence is present to support the occurrence of maximum suppression of FSH levels within a 2-week treatment period. The relative potency of conjugated estrogens to suppress FSH in our P-week treatment regimen (1.4) was not different from t.he relative potency calculated from reexamination of Geola and associates’ &week data (1.5).’ The lack of a dose-related FSH response for 10 and 20 pg dosages of ethinyl estradiol in this study is considered to have been due to inadequate sampling, rather than a plateauing of the response. Wise and associates’z have convincingly shown that the log dose response of FSH to ethinyl estradiol is linear between 2 and 50 pg. For micronized estradiol and diethylstilbestrol, previously published FSH response data are not available for comparison. Although significant suppression of LH was demonstrated for all treatment regimens combined, an LH

1.0

10.0

OF PIPERAZINE

ESTRONE

0.1 mg - EQUIVALENTS

SULFATE

Fig. 8. The correlation of FSH suppression with all doses of piperazine estrone sulfate, conjugated estrogens, micronized estradiol, and DES presented in this study (o), as well as by piperazine estrone sulfate reported by Isaacs and Havard (v), conjugated estrogens reported by Geola and associates (v), and

ethinyl

Franchimont are expressed sulfate.

estradiol

reported

by Wise

and

associates

(B) and

and associates (0). All doses of all preparations as milligram

equivalents

of piperazine

estrone

dose response to estrogen administration was not demonstrated in this study. However, Geola and associates’ were able to demonstrate a dose-response relationship for conjugated estrogens and LH. In their study, in contrast to ours, each successive dosage level was administered to the same patients. Isaacs and Havard” were able to show LH suppression with both 1.5 and 3.0 mg of piperazine estrone sulfate (17% and 22%, respectively), but they failed to find a significant difference in the LH suppression of these two dosages. In our study, as well as that of Isaacs and Havard, the different dosages were taken by different groups of postmenopausal women. The marked interpatient variability of LH may have obscured the dose-response relationship which was noted by Geola and associates. Our failure, and that of Geola and associates, as well, to demonstrate a dose-response relationship for ethinyl estradiol and LH may have been due to selection of doses that were so potent that they produced plateau range responses. Ten micrograms or more appear to cause a maximum decrease in LH. Milligram for milligram, the three nonsynthetic estrogen preparations (piperazine estrone sulfate, conjugated estrogens, and micronized estradiol) produced similar serum E, and Ez levels. This finding supports the supposition that these three compounds have similar gastrointestinal absorption, and that the steady-state condition for El and Ez was reached for each. The E,

518

Mashchak

November Am. J. Obstet.

et al.

and E2 levels correlated equivalents of piperazine system,

and

almost

best with

potency

estrone

as well

(milligram

sulfate)

in

the

in

the

CBG-BC

FSH

system.

However, the E, and Ez levels correlated less well in the SHBG-BC system, and least well in the angiotensinogen system. The increased potency of the conjugated estrogens and

upon

Ea levels

ferences. component trogens,

the appears

hepatic

systems

to account

for

to their

most

of

these

E, dif-

This finding suggests that some estrogenic of this formulation, perhaps the equine esis a more

effective

stimulant

system than are E, and EB. Conclusion. On a weight trone

relative

sulfate

and

micronized

of

the

hepatic

basis, the piperazine estradiol

es-

which contain only those estrogens that occur naturally in humans, are nearly equipotent to each other in all systems measured. They also have the least exaggerated potency in the hepatic systems relative to their ability to suppress FSH, a systemic parameter. Alconjugated

estrone

sulfate but also 20% to 35% equine

fate,

were

equipotent

estrogens, to

the

which other

contain nonsynthetic

not

equilin

tions for

are the

only

sulestro-

gens in the FSH system, they were twofold to threefold more potent in all the hepatic systems. Not only were ethinyl estradiol and DES relatively more potent on a weight basis in every system, but their hepatic potency

well, fourfold to eighteenfold in excess of their estimate by the systemic FSH suppression reThis exaggerated hepatic response suggests, aldoes not prove. that these synthetic preparaless desirable postmenopausal

forms

of

women.

estrogen

replacement

Examination

of‘ the

dose-response curves for the hepatic parameters, particularly angiotensinogen. identifies the lowest doses without undesirable hepatic side effects to be 1.25 mg of piperazine estrone sulfate, 1.O mg of micronized estradiol,

formulations,

though

was, as potency sponse. though

1, I982 Gynecol.

and

0.3

mg

of conjugated

estrogens.

Data

for

the selection of the optimum dosage range for estrogen replacement therapy are not presented here. Determination of the optimum dosage requires similar analysis of parameters, with quantification of the various desirable effects of estrogen on bone metabolism, the lower genital tract tissues. and vasomotor flushes. Nevertheless, it would appear that, of the five formulations tested, the three natural compounds would be preferable because of their decreased hepatic effects. If the dosages listed above relieve adverse symptoms, the dosages

for

long-term

replacement

therapy

should

not

these because of possible adverse hepatic effect which may lead to the development or aggravation of hypertension in some women. exceed

REFERENCES 1. Geola, F. L., Frumar, A. M., Tataryn, I. V., Sambhi, M. P., Davidson, B. J., and Hershman, J. M.: Biological effects and dose response of conjugated estrogens, in Abstracts, Sixty-second Annual Meeting of the Endocrine Society, Washington, D. C., 1980, p. 244. (Abst. No. 678.) 2. Frumar, A. M., Ceola, F., Tataryn, I. V., Lu, K. H., and Judd, H. L.: Biological effects of estrogen at different sites of action in post-menopausal women, in Scientific Abstracts, Twenty-sixth Annual Meeting of the Society for Gynecologic Investigation, San Diego, California, March 21-24, 1979, p. 66. (Abst. No. 109.) 3. Franchimont, P., Legros, J. J., and Meurice, J.: Effect of several estrogens on serum gonadotropin levels in postmenopausal women, Horm. Metab. Res. 4~2881972. 4. Anderson, D. C.: Sex-hormone-binding globulin, Clin. Endocrinol. (Oxf.), 3:69, 1974. 5. Moore, D. E., Kawagoe, S., Davajan, V., Nakamura, R. M., and Mishell, D. R., Jr.: An in viva system in man for quantitation of estrogenicity. Il. Pharmacologic changes in binding capacity of serum corticosteroid-binding globulin induced by conjugated estrogens, mestranol, and ethinyl estradiol, AM. J. OBSTET. GYNECOL. 130:475, 1978. P., Barrett, J. D., Wiedeman, C. E., and Sambhi, 6. Eggena, M. P.: The validity of comparing the measurements of angiotensin I generated in human plasma by radioimmunoassay and bioassay, J. Clin. Endocrinol. Metab. 39:865, 1975.

7. Moore, D. E., Kawagoe, S., Davajan, V., Mishell, D. R., Jr., and Nakamura, R. M.: An in vivo system in man for quantitation of estrogenicity. I. Physiologic changes in binding capacity of serum corticosteroid-binding globulin, AM. J. OBSTET. GYNECOL. 130:475, 1978. 8. Mishell, D. R., Jr., Nakamura, R. M., Crosignani, P. G., Stone, S., Kharma, K., Nagata, Y., and Thorneycroft, I. H.: Serum gonadotropin and steroid patterns during the normal menstrual cycle, AM. J. OBSTET. GYNECOL. 111:60, 1971. 9. Delaunois, A. L.: International Encyclopedia of Pharmacology and Therapeutics, Section 7, Biostatistics in Pharmacology, New York, 1972, vol. 2, Pergamon Press, p. 988. 10. Nakamura, R. M., Nagata, Y., Osborn, C., and Mishell, D. R., Jr.: Use of a postmenopausal serum pool as a reference preparation for RIA of gonadotrophins, Acta Endocrinol. (Copenh.) 75:478, 1978. C. W.: Effect of piperazine 11. lsaacs, A. J., and Havard, oestrone sulfate on serum oestrogen and gonadotrophin levels in post-menopausal women, Clin. Endocrinol. (Oxf.) 9:297, 1978. 12. Wise, A. J., Gross, M. A., and Schalch, D. S.: Quantitative relationships of the pituitary-gonadal axis in postmenopausal women, J. Lab. Clin. Med. 81:28, 1973.