Further observations on adrenal cortical function during pregnancy

Vol. 78, No. 1 July, 1959

OBSTETRICS

AND

GYNECOLOGY Obstetrics

FURTHER FUNCTION

OBSERVATIONS ON ADRENAL DURING PREGNANCY*

M.D., NICHOLAS P. CHRISTY, M.D.,** M.B.,*** ELEANOR Z. WALLACE, M.D.,**** W. E. L. GORDON, M.B.,****+ NEW YORK, N. Y.

JOSEPH

W.

CORTICAL

JAILER,

DONALD LONGSON, AND

(From the Departments of Medicine and Obstetrics Surgeons, Columbia University; the Presbyterian State University of New York, College of Medicine

and Gynecology, Hospital; and the at New York City)

College of Physicians and Department of Medicine,

A

NUMBER of workers have reported that adrenal hypertrophy occurs in certain species of animals during pregnancy.l So far as the human being is concerned, Venning2 was the first to report the finding of elevated glucocorticoids, as detected by bioassay, in the urine of pregnant women. This observation was confirmed by other investigators3% 4 who used chemical nlethods for measurement of urinary metabolites of adrenal cortical steroids. On t,he other hand, Whitelcy and Stone? observed no significant increase in adrenal weight during the progress of gestation in women, and Dohriner and co-worker@ could find no increase in the Cl9 urinary metabolites of hydrocortisone during the course of human pregnancy. With the development of methods for the determination of plasma sttaroids of adrenal cortical origin, increased plasma levels of free hydrocortisone were found in plasma of women during the third trimester.7-12 In -ent ent

*4ided by grants from the National Institutes of Health. **John and Mary R. Markle Scholar in Medical Science. ***Work performed during the tenure of a Schering Fellowship in Endocrinology. address, University of Manchester, United Kingdom. ****Work performed during the tenure of a Damon Runyon Research Fellowship. address, State University of New York, College of Medicine at New York City. *****Trainee, National Institute of Arthritis and Metabolic Diseases.

PresPres-

.,L

JAILkX

P:T

Ala

Am. J. Ohst. & G,nec 1111). I’)r’l

;uJdit,ion, increased responsiveness of plasma. cortic&croids t,o intravenous A( 17’~ was sJlown to occur during the gravitl state, the magnitude of r~~sJ)o~ls~* IJeing similar to that seen in patients with (Cushin,“.ls syndronlc: tlr1e ttl I,ilatf~r~iIl adrenal cortical hyperplasia.ll These steroidal findings, the suggestive cvidencc for decreased gJuCos(’ tolerance, and the amelioration of such diseases as rheumatoid arthrit,is’ observed in pregnancy have suggested the existence of a “hyperadrenal stat,c.” \Vith the a(lvcnt of more specific chemical and physiological techniques, the hyperadrcnalisnt o I’ mechanisms responsible for the apparent untlcrlying pregnancy are becoming clea,rer. The purpose of this report is to dascribc it \rariety of steroid studies which may tlefine more precisely the true Functiozr;~I level of the adrenal cortex in pregnancy and t,o correJate these observations with investigations previously reported in the lit,crature. Material

and Methods

All subjects studied were patients in the (Ibstctrical or Gynecological None of the pregnant wome~l ,Services of the Sloane Hospital for Women. showed evidence of pre-eclampsia or diabetes mellitus. Hospitalized patients were admitted for elective cesarean section or premature rupture of the membranes. The hospitalized nonpregnant women were admitted for elective gynecological surgery. Analytic Methods.-The plasma and urinary 17-hydroxycorticosteroids are present in the free form or as conjugate#s with glucuronic acid. Unless specifically stated to the contrary, all values for plasma 17-OH-corticosteroids are given as free steroid, while those for urinary 17..OH-corticosteroids are expressed as total steroid (free plus glucuronidc, the latter being the fraction of steroid released after beta-glucuronidase hydrolysis). Both the plasmit and urinary 17-OH-corticosteroid levels WCPCdetermined by the Peterson modification14 of the Silber-Porter method.‘” E’or partial identification of plssma hydrocortisone, quantitatively the chief circulating component of the free 17-OH-corticosteroid fraction, pools of normal and pregnancy plasma were chromatographed after extraction in the Zaffaroni chloroform-methanol-formamide system.16 The ultraviolet light-absorbing areas on paper corresponding to hydrocortisone were cluted and quantit,ated by ultraviolet light absorption at 240 rnp in a Bcc?kman DU spectrophototneter. P~ocetIu~es.-,4drcllal cortical rcsponsc was assessed by means of a 4 hour intravenous infusion of 25 1.11. of ACTHIL with blood specimens for plasma 17-OH-corticosteroid determination being drawn at the beginning and end of the infusion. An effort was then made to simulate the conditions obtained during the intravenous ACTH test so that an appraisal could be made of the effects on steroid response to ACTH of the delayed rate of hydrocortisone metabolism which is charact,eristic of pregnancy.17, I* The water-soluble hemisuccinate and phosphate derivatives of hydrocortisone (100 mg.) were admjnjster&l intravenously in :500 ml. of glucose and wa.tcr to normal and pregnant subjects over a 4 hour pcriotl (the same period as was used for the jntravenous ACTH tests). J’Ja,sma 17-OH-cortSicostcroid levels were measured in bJood specimens drawn at the start and end of the infusion.

Volume 78 Number I

ADRENAL

CORTICAL

FUNCTION

DURING

PREGNANCY

3

Results The average free plasma 17-OH-corticosteroid level for 50 nonpregnant subjects was 16 + 6.9 ,pg per 100 ml.,lO whereas the average value in 27 women studied during the third trimester was 37 t 7.6 pg per 100 ml. The values differed significantly (P
TABLE

I

Normal

NO.

50

Pregnancy, third

and

07 trimester

II.

EFFECT LEVELS

s.It.*

16.0

6.9

11.0

3i.n

7.G

23.0

(P
OF INTRAVENOUS IN THE FIRST AND

same

in

normal

First

Second

I

and

in

CORTICOTROPHIN UPON PLASMA SECOND TRIMESTERS OF NORMAL PLASMA

PATIENT

PLASMA 17.OH-W AFTERCHROMATOCRAPHY (@G %,i

-

*Difference is significant +Recovery of steroid is 62%, respectively).

TABLE

pLASMA l7-OH-CS (SILBER-PORTER) (PG %o)

BEFORE

A('TH

17.OH-CS I

pregnant

plasma

(69%

IT-OH-CORTICOSTEROII) PREGNANCY

(/.&G 'j?,) AFTER

ACTH

Trimester.Va McK Cl Al Trimester.Hi Ro Ca To Lo Ma

E 26 24 34 19 37 26 32 40

51 45 45 46 56 42 61 7.3 74 ‘ii

A single plasma 17-OH-corticosteroid level may not in itself be a very Response of plasma steroid level to accurate reflection of adrenal function. maximal ACTH stimulation under standard conditions has been shown to be a fairly reliable index of adrenal cortical activity.ll, l8 Studies of adrenal cortical response showed that plasma corticoid rise after ACTH in 4 women in the first trimester remained within normal limits (Table II). Study of 6 women in the second trimester showed a gradually increasing response with some normal and some exaggerated corticoid rises after ACTH (Table II). In the third trimester, an apparent hyperresponse was found in all 9 patients This excessive reaction was reminiscent of that seen studied (Table III). in patients with Cushing’s syndrome associated with bilateral adrenal hyperplasia, and was suggestive of an increased sensitivity of the adrenal to of prednisone partially suppressed the ACTH.*O However, the administration exaggerated response in 2 pregnant women (Table III) but did not materially

t)J; J~l~c~c~tJiSUllc~ reduce it in ( lushing ‘s syndromc.‘“~ ‘I The degree of sLLpl)ressiun observetl in pregnamy was intermctliatc t)ctwccn that of the nor111a1 sl~ll,jec~l (in whom suppression of steroid response IO A(:TH ai’t,er prednisonc is tInI t,kcbfl1 and the patient with Cushing’s syndroruc~.

-

____-. ___-.

PLASMA NO.

PATIEiYT

Bo Re Br

_

-

POST

17-OH-CS

iluG

%)

DAYS

I'ARTUM 1

-. In the study in which we attempted to reproduce approximately the conditions of the ACTH test with exogenous steroid (see Materials and Methods), 20 normal and 20 pregnant subjects received infusions of 100 mg. of hydra cortisone over 4 hours. Fig. 1 shows the results of plasma 17-OH-corticosteroid measurement at zero and 4 hours. The rat,c of plasma 17-OH-corticosteroid rise was greater in pregnant than in normal women. Fig. 2 shows the results

Volume 78 Yomtwr 1

AI~ltI:NdI~

CORTICAL

FUNCTION

DURIKG

PREGNANCY

5

of the -1-hour ACTII test in the third trimester for comparison. Again, the rate of plasma 1.7~@I-I-corticosteroid rise> was greater than normal. In both st,utlies the rate of st,croid incrcasr in prcgnaney was significantly in excess of 110r1wLl (P
010 PR(GNANT l

-a NORh4AL

4

0

TIME,HOURS Fig. I.-Rate of rise of plasma 17-OH-corticosteroid hours of 100 mg. hydrocortisone. Rate of increase pregnant than in normal subjects (P
of

level following the infusion over steroid level is significantly greater

4 in

both urinary and plasma determinations should be used ideally. In the present study, 24 hour urine specimens were obtained from 10 normal and 10 pregnant subjects on the day before and the day of standard ACTH tests. Total (free and conjugated) urinary 17-OH-corticosteroids were determined in these specimens (Table V) . In spite of the fact that the resting plasma levels were elevated and the response to ACTH exaggerated, response of the total The urinary corticosteroids was not excessive when compared to the normal. significant elevation of control urinary corticoids (before ACTH) over the normal may be the result of nonspecific Porter-Silber chromogens, the levels In patients with Cushof which are not affected by ACTH administration. ing’s syndrome, by contrast, the excessive response of plasma 17-OH-corticosteroid,s to ACTH is reflected in an exaggerated response of urine corticoids.22

JAILER

6

ET

AL.

It should be re-emphasized that the plasma corticosteroids as determined while the urine measurement, inby the method used’j are not conjugated, To explain the apparent COW cludes total (free plus glucuronide) steroid. tradiction between exaggerated free plasma steroid response and normal total urine steroid response a study was made of the proportion of steroid in the free and in the conjugated (glucuronide j fraction of urine after an intraThe hormone was administered venous load of 100 mg. of hydrucortisoncl. in an intravenous injection over a :! tllinutc pcriotl to 6 normal and 6 pregl’rine sp(4lttens were collrctcd ov(‘r it nant women in the thirtl trimester. PREGNANCY,THIRD

TRIMESTER

100 95

s :: <

75

$

65

’ i2 5 5 &

60

70

55 50

8 F 0:

40

45

s

35

i-5 5

30

a

25

3

20

0=

I5 to 5

Fig. L-Plasma 17-OH-corticosteroid response to infusion hour period. Sold lines represent the 9 pregnant women tested, steroid response in 40 normal individuals.*0 Rate of increase of icantly greater in pregnant than in normal subjects (P
of 25 shaded plasma

1.U. ACTH over a 4 area indicates plasma steroid level iS Signif-

period of 8 hours. In the normal individuals, the ratio of urinary corticosteroid glucuronide to free steroid was 3:l. In the pregnant women the ratio was 2:l. This observation suggested that a greater percentage of exogenous (and by extension, of endogenous) hytlrocortisone escaped conjugation in late pregnancy. This finding confirmed studies of Migeon and associatesG7 and may explain the elevated urinary glucocorticoid level found in pregnancy by bioassay.? In the bioassay used by Venningz free corticosteroids are not removed prior to hydrolysis, and these free steroids (e.g., hydrocortisone) are probably more potent. than the conjugated derivatives.

Volume 78 Number I

ADRENAL

CORTICAL

FUNCTION

DURING

PREGNANCY

7

Tt is possible that the increase in biologically active glucocorticoid described’ may bc related to the increased proportion of free steroid excreted in the urine near term. TABLE

V.

EFFECT OF 4 HOUR INTRAVENOUS CORTICOSTEROID VALUES IN NORMAL THIRD URINARY NORMAL ACTH

BEFORE 6 9

Averwe* grouts

I?'-OXI-CORTICOSTEROIDS

urinary but

(MG.

PREGNANT

AFTER

ACTH

BEFORE

ACTH

WOMEN I

AFTER

ACTH

10 9

13

11 17 5 13

13 20 7 19 11 13 14

17

ii 11 7 10

steroid values before levels after ACTH do

IT-OH-

PER 24 HOURS)

SUBJECTS

13 18 14 13 22 10 17 17 17 18 16

5 G 8 4 G 4 3 3 6 *Average (P
INFUSION OF CORTICOTROPHIN ON URINARY SUBJECTS vs. PREGNANT WOMEN IN THE TRIMESTER

ACTH are significantly not differ significantly.

different

in

the

two

Comment From the experimental data presented above certain conclusions emerge: (1) the level of free plasma corticoids is distinctly above normal in late pregnancy ; (2) plasma corticoid levels after ACTH stimulation are also greater than normal; (3) in contrast, total (free and conjugated) urinary corticosteroid values in late pregnancy are somewhat greater than those found in nonpregnant subjects, but after ACTH stimulation these values are no higher than normal. The proportion of free urinary corticosteroid appears to be increased in pregnancy. It may be possible to reconcile some of these apparent contradictions. The slower than normal rate of disappearance of hydrocortisone from plasma in pregnant women17g I8 may account satisfactorily, or at least in part, for the excessive plasma steroid response to ACTH. Hydrocortisone secreted by the ACTH-stimulated adrenal might be expected to persist for a longer period in the plasma compartment, giving the illusion of increased adrenal response to corticotrophin. Since free hydrocortisone levels in plasma are elevated, the proportion of free corticosteroid in urine is increased, but this increase is presumably insufficient to raise the urinary value of total (free plus conjugated) corticoid after ACTH to an excessive degree. However, the slower rate of hydrocortisone disappearance from pregnancy plasma does not appear to account entirely for the elevated resting plasma steroid levels observed during gestation,‘-12 since in the presence of sevcse hepatic diseasel”! I8 and in in which hydrocortisone clearance from plasma is also hypothyroidism,‘3 delayed, normal plasma 17-OH-corticosteroid levels are generally found. In addition, cirrhotic and hypothyroid patients usually show normal (not excessive) plasma steroid response to ACTH.l’

it is generally accepted that one or the mechanisms controllillg th(’ ITlease of ACTJI froln the anterior pituitary is t.he It~rl of hwdat,ill~ iI(Ll’tlr1a.l In man it has been shown that, the AC’I’TI cclnl.(~~~f01’ ~~cJ(K[ glucoeorticoid.2~ is elevated in the absence of adrenal corticostthroitk (i.c., in p:Ltieltt.S with Addison’s diseasez5) and that this high ACTH level can be reduced by steroitl It is conccivablc that the steroidal influenw 011 AWII I’(‘a,dIninistratioa.*” lease is altered during pregnancy in swh a way that the high l)lilsItt;l stcrc)i(l level does not suppress ACT11 production as it dew in the normal. Iinde: normal conditions one would expect the high plas~w\ corticoitl It~vrl to rcwllt in ACTH suppression so that the stimulus to the adrenal cortex woulrl be rctluccd and the level of plasma stcroitl ultitrlatc~ly brought b;~(*k lo Ilc~l’lll~l~. IfAvctl. In pregnancy the “feed-back” mc~hanism ~rlay he “WI.” at ;1n alJl~or?rl:rl Some support for this assumption ~nay be found I,y analogy in thca oI)s(‘r\.iltion that pituitary melanophorotrophic hortnonc~ (MSH) has ~J~TII r.ctl)t~t.i(.(I 1.0 be increased occasionally in the blood of pregnant wotnen.27 :\lll~c~uglr high levels of blood ACTH have not brcn consixttwt Iy I’ountl in ~)r*~~yn;~r~~~~. the analogy seems pertinent in that MST-1 secretion, like ACTJJ rc~lc~a~c~;“~ (‘iIt also be suppressed by corticosteroid aclniinistratic,n.” There is no direct evidence to explain how sncl~ a11 alterat,ion in homrostasis might, come about in pregnancy. It is pc*rhaps 13~l(~vant to draw ;I [Jil t’2Ilt’l between pregnancy and the results of administration of estrogenic hormones. In the latter situation two groups of workers have observed levels of ~~I~KwI~I free 17-OH-corticosteroids as high as OI higher than those fouri(l ii, gestntion.*g* ” Since circulating estrogen is demonstrably increased in pr~gnnnc) it is tempting to raise the question of whether this increase in PlldOger~otls estrogen is the cause of the increased blood corticosteroid level. Mucdh more evidence is required to substantiate this hgpothcsis, especially in viet$. (II’ thy fact that a very large number of hormonal and ot,hev metabolic changrs OCCIII~ during gestation. Any one of these ma.y finally btl found to b(a 1not’(L clirl?c*tlS related to the elevation of plasma corticosteroids than the increase in (:sttwgetl secretion. In searching for an explanation for the possibly alter4 homeostn,tic~ wl;ltionship between the adrenal tortes and the anterior pituitary in I)1*r#ll>lll(‘y, it is perhaps pertinent to recall that pregnant women exhibit 110 cl~~~-(~ul evidence of hyperadrenocorticism despite the observed high levels of plasnla hydrocortisone. Similarly, patients trea.ted with c&rogen have high plasma corticosteroid 1evelP~ 3o but show no clinical signs or symptoms of (JsctJkqive adrenal cortical activity. Further, pregnancy and the administration OE estrogen are both associated with definite incrclasos in the blootl I(BV~~Iok’ protein-bound iodine,31, 32 but there is no clinical cvidt~nce of thyrotusic*osis under either of these conditions. It is possible that illthough the If~tc~ls 01 both thyroxine and hydrocortisone are elevated, t,hcsscIIO~IIIOI~CLS a~ tl(,t (.olI1pletely active biologically in the physical state in which they exist during pr(dg:‘nanry and during the administration of estrogens. F:strogenic hc~rmom4“ and pregnancy”’ have been shown to bring about an illcrease in serum illyV roxine-binding protein, a finding which may explain in part the hint] I,Ovt~ls

Volume 78 Number

I

ADRENAL

CORTICAL

FUNCTION

DURING

PREGNANCY

9

of PDI. At this time, analogous changes in protein binding of corticosteroids during pregnancy are controversial. Daughaday35 claimed that corticosteroid binding in pregnancy does not differ from the normal, but Slaunwhite and Sandberg, using a different method, observed increased binding of hydrocortisone to protein in pregnant women. There is further circumstantial evidence in support of the hypothesis that in pregnancy (and in estrogen-treated subjects) adrenal cortical steroids may not be able to exert their full biological effects. Curry and Beaton have reported a marked resktance to certain biochemical effects of administered cortisone in pregnant rats. The placenta appears to be essential for maintenance of this cortisone resistance. Pregnant women have also been reported to show cortisone resistance.38s 39 In studies to be published, Cutman and Jailer40 have observed that treatment of mice with diethylstilbestrol prevents the deposition of liver glycogen which usually accompanies the administration of cortisone to fasted animals. From these oh8servations it is perhaps reasonable to advance tentatively the following hypothesis: that although free plasma hydrocortisone is increased during pregnancy or estrogen administration the biologically effective level is within normal limits. This assumption would account for the lack of real symptoms and signs of hypercorticism in these two states. It might also account for the apparent failure to inhibit pituitary ACTH release presumed to exist during pregnancy and after estrogen administration. One would then be confronted with the apparent paradox of seemingly high levels of plasma hydroeortisonc, perhaps 50 per cent of which are relatively inert biologically, so that the hormone is not capable of cserting either its inhibitory effects upon the cells of the adenohypophysis or its widespread metabolic effects upon other tissues of the body.

Summary 1. h’ree plasma hydrocortisonc of pregnancy.

levels are increased in the third

trimrst,er

3, There is an apparent exaggeration of free plasma Ii’-CII-corticostcroid response to ACTH in the third trimester which disappears by the end of the first week post partum. The excessive response may be partly accounted for by a delay in the clearance of hydrocortisone from plasma in late pregnancy. 3, Free urinary corticoid values are increased in late pregnancy, but the response to ACTH of total urinary corticoids (free plus glucuronides) is no greater than normal. L$. The hypothesis is advanced that there may he in late pregnancy a resistance to the effects of hydrocortisone upon various tissues (e.g., the anterior pituitary), perhaps because some physical factor, as yet unknown, renders a portion of the circulating corticosteroid biologically inert. JV~ wish to express our for generous supplies of steroid valuable technical assistance.

gratitude hormones

to Merck & CO., Inc. and of corticotro@ia.

and to the Miss Elsie

Upjohn 6. Ewen

Company rendered

JAILER

10

ET

AL.

References 1. Smith, 2. 3. 4. 5. 6. 7. 8. 9I. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 30. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 3X. 39. 40.

P. E;. : Endocrinology 56: 271, 1955. Venning, E. H. : Endocrinology 39: 203,1946. M. D. Thesis, Univ. of Cambridge, 1956. Mills, I. H.: Appleby, J. I., and Norymberski, J. K.: J. Endocrinol. 15: 310, 1957. Whiteley, H. J., and Stoner, H. B.: J. Endocrinol. 14: 325, 1957. Dobriner, K., Lieherman, S., Rhoads, C. P., and Taylor, H. C., *Jr.: The Normal a!111 Pathological Physiology of Pregnancy, Baltimore, 1948, Williams and Wilkins Company, p. 75. Gemzell, C. A.: J. Clin. Endocrinol. 13: 898, 1953. Bayliss, R. I. S., Browne, J. C!. M., Round, B. P., and Steinbeck, A. W.: Lancet 1: 62, 1955. Robinson, H. J., Bernhard, W. G., Grubin, H., Wanner, H., Sewekow, G. W., and Silber. R. H.: J. Clin. Endocrinol. 15: 317, 1955. J. Clin. Endocrinol. 15: 1073, 1955. Wallace, E. Z., Christy, N. P., and Jailer, J. W.: Christy, N. P., Wallace, E. Z., and Jailer, J. W.: .I, Clin. Invest. 34: 899, 19%. Assali, N. S., Garst, J. B., and Voskian, J.: .J. Lab. & Clin. Med. 46: 385, 1955. Hench, P. S.: Proc. Staff Meet. Mayo Clin. 13: 161, 1938. J. B., Guerra, S. L., Brotlie, B. B., and Buuilrl. .I. .J.: Peterson, R. E., Wyngaarden, J. Clin. Invest. 34: 1779, 1955. Silber, R. H., and Porter, C. C.: J. Biol. Chern. 210: 923, 1954. Burton, R. R., Zaffaroni, A., and Keutmann, E. H.: J. Biol. Chem. 188: 763, ll)51. Migeon, C. J., Bertrand, J., and Wall, P. E.: J. Clin. Invest,. 36: 1350, 1957. Christy, N. P., Wallace, E. Z., Gordon, W. E. L., and Jailer, J. W.: .J. (Olin. Invest. 38: 299, 1959. Eik-Nes, K., Sandberg, A. A., Nelson, D. H., Tyler, F. II., and Samuels, I,. ‘l’.: .J. Clin. Invest. 33: 1503, 1954. Christy, N. P., Wallace, E. Z., and Jailer, J. W.: J. Cliu. E;u,locrinol. 16: 11159, 1956. Christy, N. P., Longson, D., and Jailer, J. W.: Am. J. Med. 23: 910, 1957. Laidlaw, J. C., *Tankins, D., Reddy, W. J., Harrison, J. H., an(1 Thorn, 0. W.: .J. Clill. Endocrinol. 14: 781, 1954. Levin, &f. S., and Daughaday, W. H.: J. Clin. Endocrinol. 15: 13!)!), 1955. Sayers, G.: Physiol. Rev. 30: 241, 1950. Taylor, A. B., Albert, A., and Sprague, R. G.: l