Estrone Sulfate, Estrone, Estradiol and Estriol Plasma Levels in Human Pregnancy D. L. LORIAUX, H. J. RUDER, D. R. KNAB, AND M. B. LIPSETT Reproduction Research Branch, National Institute of Child Health and Human Development, Institutes of Health and the National Naval Medical Center, Bethesda, Maryland 20014 ABSTRACT. Estrone sulfate, estrone, estradiol and estriol were measured through the course of pregnancy in 85 women. Of the unconjugated estrogens, estradiol had the highest plasma concentration, being approximately twice that of estrone and four times that of estriol. Estrone sulfate level was twice that of estradiol. All estrogens

T

OTAL estrogen concentrations in blood and urine rise markedly during human pregnancy. Because of limitations in sensitivity, earlier blood estrogen measurements have generally included both free and conjugated estrogens. Prior to the advent of the radioligand assay for estrogens, only a few reports of unconjugated estrogens in late pregnancy plasma were available (1-5). The radioligand assay has facilitated the study of plasma estrogens and systematic studies of plasma estrone ( E l ) (6), estradiol (E2) (7-9,34), estriol (E3) (10,11), and E l , E2 and E3 (12) have recently appeared. Our finding that estrone sulfate1 (ElS) is the most abundant estrogen in blood (13) coupled with the known interaction between fetal sulfokinases and placental sulfatases led us to study estrone sulfate and the classical estrogens mentioned above through the course of pregnancy. Estrone sulfate and unconjugated estrone, estradiol, and estriol were measured between the 8th and 36th week of pregnancy in a group of 85 women, each of whom was sampled once in this time period.

Received May 25, 1972. 1 Systematic names of steroids used in this study are: Estrone sulfate = 1,3,4(10), estratriene-17-on3-yl-sulfate.

National

measured showed the same pattern of increase with time. When plotted against fetal and placental weights, the increase in plasma concentration of all estrogens correlated better with fetal weight than with placental weight. ( / Clin Endocrinol Metab 35: 887, 1972)

Materials and Methods All patients were drawn from the outpatient obstetrical service of the National Naval Medical Center, Bethesda, Maryland. Gestational age was estimated from the last known menstrual period. No attempt was made to standardize the time of phlebotomy. The measurements of estrone,2 estradiol and estrone sulfate have been described previously (13). All estrogens were measured using an antibody generated against estrone-17-oxime coupled with bovine serum albumin through a mixed anhydride reaction. The various estrogens were isolated using an LH-20 Sephadex column, and the bound and free fractions were separated using dextran coated charcoal. The measurement of estriol was performed in an identical fashion using a later fraction from the LH-20 sephadex column (20-25 ml) to isolate estriol from the other estrogens. Additionally, estriol 1-2, 3H (40 mCi/mM) was used as the trace material for the assay. The apparent association constant, k, for estriol with the antiserum employed was 2.63 X 10°. Water and charcoal stripped plasma gave blank values that could not be differentiated from zero. The association constant, k, would predict a sensitivity limit of 11 pg per assay tube (14). The assay, as constructed, would then have a detection limit of 14 pg/ml. Intra-assay variation over the The authors wish to acknowledge the expert technical assistance of Mr. Louis Brown.

887

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 22 January 2017. at 04:09 For personal use only. No other uses without permission. . All rights reserved.

LORIAUX, RUDER, KNAB AND LIPSETT

888

JCE & M Vol35

1972 No 6

working range of the standard curve was % (IS). Inter-assay variation was 27%. Plasma estriol was undetectable in follicular and luteal phase samples from 4 normally cycling women. Estrone levels were plotted against length of gestation using the logarithms of all values to generate non-linear regression lines of best fit.3 Results The primary data are presented in Figs. 1 through 4. Each black point is the mean of values obtained from 2-10 women. The open WEEKS OF PREGNANCY circles represent calculated points along the line of best fit. The coefficients of cor- FIG. 2. Plasma estradiol in pregnancy. Symbols as relation ranged from 0.81 (E3)-0.89, indi- in Fig. 1. cating that a single exponential function was adequate to explain most of the variation of estrogen concentration with time. Assuming this function to accurately represent the data, the rate of increase of each estrogen was approximately the same from the eighth to the 36th week (Table 1). Thus, the ratios of _ 8.0 the concentration of each estrogen derived from the line of best fit, for 36th/8 week, 36/16, and 36/24 weeks were the same. Unconjugated estradiol averaged 800 pg/ ml at eight weeks of pregnancy, two to three times the peak estradiol concentration \\ WEEKS OF PREGNANCY

3. Plasma estriol in pregnancy. Symbols as in Fig. 1.

FIG.

12.0

10.0

e OT

8 0

I8

z