Ovarian-pituitary hormone interactions during the perimenopause

Moturifus, I1 (1989) 331-339 Elscvier Scientific Publishers Ireland

331 Ltd.

MAT 00566

Ovarian-pituitary the S.E. Shideler’,

ne interactions

during

G.W. DeVane’, P.S. Ka?ral, E. Benirschke’ and B.L. Lasley*

‘Department of Obsfetrfcs and Gynecclog~. University of Fkwida CoIlege of Medicine, Gainesville, FL, *In.Mlute forEnvirunmenral Healfit Research, Universiry of Cal~ornia, D&s, CA 9S616 and ‘Universi?y of California. Sun Diego. CA 92112 (U.S.A.) (Received

29 June

198% Revlon

received

11 August

1989; accepted

20 September

1989)

that occur at the onset of the perimmopause. daily urine and from 5 peri-menopausal womer for 3 or 4 consecutive cycles. Estrone conjugate and prz4nsnediol-J-gluroronide concentrations were determined for urine samples. Circulating luteiuizing hormone. follicle stimulating hormone, progesterone, estradiol and estrone mucentrations were derermined in serum wmpks. Two of the 5 women exoerienced irregular menstrual intervals during the study period. One of there subjects experienced a prolonged intermenstrual interval. Three other womer. exhibited ~XWeotly regular ovulatory menstrual cycles. To describe

the hormone

chnnges

random blood samples were collected

The prolonged intermenstrual interval of one women exhibiting irre~elar menstrudl inteirals Was associated with low uriwry estrogen levels in th: eariy to;iic:d;r phase of the affected cycle. f01lo~ed by increased gonadotrophin level; and increased estrogen levels that rose to exceed nonoel cycle coocentrations by 2. or 3.fold. Iucrezsed estrogen levels were lollowed by deelining gonadotrophin levels, minimal progesterone production, and, ultimately, vaginal bleeding. These data suggest that there are some forms of menstrual variability at the time of the wimenopause associated with lowered early folliculzr phase estrogen levels. Reduced negative feedback and subsquently increased gonadotrophin levels may have stimulated estrogen production which may have suppressed gonadotrophin secretion and lowered estrogen exretion, resulting in the observed oscillations between episodes of hypo-and hypaertrogenism. (Key words:

Perimenopause,

Steroids,

Gonadotropins)

Introdnction The interval immediately preceding the menopause is characterized by variability in menstrual cycle length it: many women [1,2]. Previous endocrine studies suggest that during this transitional period, normal urinary estrogen levels corrrFponaence

k?: B.L. Lasley.

0378-5122/89/$03.50 0 1989 Elsevier Scien!ifir Printed and Published in 1r;land

F Ahlishers Ireland

L:G

332

simultaneously with higher than normal levels of gonadotrophins, especially luteinizing hormone (LH). and, to a lesser extent, follicle stimulating hormone $SH) [3,4]. However, more recent studies [5] have demonstrated that estradioi (EJ secretion during both the follicular and luteal phases is lower in ovulatory pcrimenopausal women than in normal younger women. Furthermore, serum FSH levels are consistently elevated in pcri-menopausal women, ever,when LH levels are within the normaI range for cycle phases [S-Q]. In this sttidy, longitudinal measurements of urinary ovarian steroids in 5 perimenopausal women during several consecutive menstrual cycles of variable length were performed. Selected measurements of circulating LH, FSH, estrone (E,), E,, and progesterone (P) were also performed to substantiate the daily urinary profiles and to assess the feedback relationship between the ovary and pituitary. The purpose of the study was to clarify further the sequential changes in steroid and gonadotrophin hormone production that are associated with menstrual cycles of variable length during the perimenopause by observing the endocrine events associated with menstrual cycles of variable length during the perimenopause. The present study extends the information in previous reports, which relied upon serum samples collected once [g--10] or twice a week [ll], or less frequently, [12] by employing daily urine evaluations to accuratelyassess cyclic changes in hormonal patterns and demonstrate the utility of urine evaluations for studying the longitudinal changes associated with the perimenopause. occur

Matedals and methods Patients Five healthy female subjects between the ages of 42 and 47 were rerrGtcd aa control subjects for a study approved by the Investigational Review Board of the University of Plorlda College of Medicine. (The subjects !!ere paid for taking pati in the study). Subiects were not taking any medications, leading unusual life. styles, on special diets, and recalled regular menstrual cycles (25-29 days) for 6 mth preceeding the study. All 5 subjects were selected on the basis of subjective assessment and recall of recently irregnlar patterning, i.e., long or short cycle intervals, and/or the recent, irregular onset of vasomotor symptoms, such as hot flushes. They collected early morning urine specimens daily for three or four menstrual cycles. During the second and fourth cycles, biood samples were obtained 3 times per week.

Urinary measurements of steroid conjugates were performed as previously reported [13]. Briefly, the combined concentrations of urinary estrone-3-stilfate and cstrone-3-glucoronide were evaluated by direct immucoassay using an antiserum that cross-reacted equally with both estrone conjugates (E, Conj). Pregnanediol-3-glucuronide (PdG) concen:.ations wzre measured by direct immunoassay using antiserua and labelled PdC purchased from the Courtauid Institute (London). All concentrations of urinary steroid conjugates were related

333

to creathrine and expressed as mean ratios. The urinary creatirdne levels were measured by rhe Taussky mahod 1141. Radioimmunoassay measurements of serum LH, WH. E,, Es and P were performed by DeVane et al. US]. This group reported that the radioimmunoassay Iower and upper limits of serum LH concentrations for normally cycling women were 5-25 mIU/ml, with mid-cycle levels reaching 100 mlU/ml. In contrast, LH concentrations for menopausal women were & mRJ/ml or more. Similarly the expected range for serum P§:I concentrations for normally cycling women was 2 -20 mIU/nl, and equal to or greater than 40 rrlU/ml for menopausal women. The normal range of estradiol levels for cycling women was given as being between 50 and 200 pg/ml 1151.

During the first cycle of a subject aged 42 yr (Kg.), uriuary E, Conj levels were erratic throughout the Is-day foliizular phase pig. 1). This was followed by

K.B. ~42Y/Ol 240

b 200 B -. P .* s

I

8

IO

w iO0 60

6

40

4

20

2 10

LO

30

40

30

60

70

DAYS Fig. 1. Levels of B, Conj and PdG indexed to creatiine concentration collected over 71 days in subject K.B. Cwarian steroid and gonadotrophin levels are illustrated during the second intermenstrual cyc!e. The notched rectangx~ icwsent menstrual bleeding intervak

334

a PdG increase of 12 days duration, with levels consistent with adequate corpus luteum function [13]. During the second intermenstrual cycle, blood hormone measurements revealed normal gonadotrophin concentrations. A mid-cycle LH surge was not observed. presumably because of infrequent sampling. During a relatively short follicular phase (9 days), serum E, levels rose to a midcycle value of 290 pg/ml, the timing of which coincided with the urinary E, Conj peak which was followed by a PdG rise. The urinary steroid profile for this cycle was within the range found in fertile women during conception cycles [13]. Decreased E, Conj excretion with a luteal phase PdG elevation of 12 days duration was observed in the final cycle. One subject aged 45 yr (M.P., Fig. 2) experienced 2 normal ovulatory cycles out of 3, as inferred from urinary steroid metaholite profiles. At the onset of the second intermenstrual cycle, blood hormone levels during early and mid fohicular phase revealed slightly greater than normal LH and FSH concentrations (27 and 21 mIU/ml, respectively). At mid-cycle, increased E2 levels were followed by a rise of P values. During the third intermenstrual cycle, low levels of y, Conj were measured for 20 days, followed by a gradual increase in the E, Conj profile for 40 days that reached peak values of 250 ng/mg Cr. Consistently low PdG values

DAYS Fig. 2. Levels of E, Conj and PM indexed to creatinineconcentrationcdlcctcd over 125 days in subject M.P. Ovarian steroid and gonadotrophin levels depicted during the sec.md cycle. The solid recm&s representmensmnl bleeding episodes.

335

throughout this interval indicated this was an anovulatory cycle. A heavy vaginal bleeding episode ended this intermenstrual interval. The urinary steroid metabolite profile of the first menstrual cycle of a second subject aged 45 yr (A.P., Fig. 3) reflected an ovulatory pattern characterized by a short follicular phase length of 10 days, followed by a substantial PdG elevation of 13 days duration, reaching levels consistent with adequate corpus luteum function. During the second intermenstrual interval, baseline urinary E, Conj levels were low, approximating pre-menarchial levels of less than or equal to 20 ng/mg Cr 114, and remained low for 20 days with elevated serum LH levels (40 mIU/ ml) consistent with menopausal levels, and FSH concentrations were 22 mIU/ml. During this time, mildly elevated LH and FSH levels rose gradually to levels of

Fig. 3. L.ev& of E, Conj and PdG indexed to creatinine conctntiation collected during 4 intennensrrual cycles in subjxt A.P. The solid rectangles represent menstrual bleeding episodes. Ovarian steroid and gonadotrophin levels are denicted during the second md iourth cycle studied.

336 125 m.iUiml and 59 mIU/ml, respectively. The elevated gonadotrophies were associated with a progressive increase in urinary E, Conj and serum E, and E, concentrations that persisted for 28 days. The magnitude of this increase, determined by E, Conj levels, was greater than that of the previous cycles. Moreover, in comparison with the other subjects, serum concentrations of E, and E, were also greater in this individual. Concommitant serum progesterone hnd urinary PdG levels increased only minimally, indicating ovulation did not occur. The prolonged, high, circulating estrogen was associated u ith supressed gonadotrophin secretion. This intermenstrual interval lasted 48 days, resulting in menstruation on day 71 of the study. The subsequent 28-day intermenstrual interval was not monitored. The fourth intermenstrual interval (or third monitored cycle) revealed a normal urinary hormone profife and exhibited a iollicular phase of 12 days and lutaal phase of I3 days. The fifth and final intermeustrual interval monitored lasted 43 days. As in the second interval, early follicular phase cs!rogen levels were very low and associated with elevated gonadotrophin levels. In spite of serum LH and FSH levels consistent with ovarian failure [17], ovulation appeared to occur, as indicated by all steroid indices, including PdG excretion. Following ovulation, urinary estrogen concentrations remained elevated, trs did serum estrogen levels, particularly E,. The luteal phase was characterized by a short interval (9 days) of increased PdG and serum progesterone. This interval ended with menstruation at the time of progesterone withdrawal, 13 days postovulation.

Diszussion Thsse data provide ,abstrvations of daily ovarian steroid excretion in successive menstrual cycles during the perimenopause transition period and describe cyclic hormonal events during this time. The concurrent, Iess frequent analyses of serum hormones substantiate urinary steroid measurements and document circuIOlting pituitarr gonadotrophin concentrations. The paired serum and urinary ster( id values we.re found to provide the same qualitative information regarding ovarian function and allowed the documentation of gonadotrophin “escape” in which E, Conj. lnrels below 20 ng/mg Cr represent a rate of estrogen production that was insufficient for adequate negative feedback. In contrast, the serum levels were not as helpful in elucidating this dynamic, since circulating estrogerl levels (< 20 pg/ml) at this time were near or below the limit of sensitivity of the assay. The dramatic and sustained release of estrogen during prolonged intermenstrua.1 intervals han not been previc dv re,nxtcd. This j-&tern appears !o be a direct result of endogenous hypeztimulation by gonadotrophins in response to previous low estrogen production which may be characteristic of some perimenopausal women [8,12]. The finding of an LH/FSH ratio greater than I in some peri-mmopausal women suggests that variability exists in gonadotrophin response to reduced ovarian function. Other investigators [6,8,9,1 l] observed ekvatcd FSX amccntrations and normal LM levels in perimenopausai women, while the present subjects always demonstrated higher LH than FSH ievels. Apparently, irr the early perimenopause, the capacity of residual follicles to pro-

337

duce E, in response to a wide range of ratios of gonadotrophins is retained. Moreover, estrogen feedback mechanisms appear to be intact, since gonadotrophin levels fall in response to prolonged increases in Ez. Certainly, other factors such as inbibm may play a role in determining gonadotroping ievels, however, tbe present data are consistent with an earlier report [lZ] which indicated that if inhibin is involved, it is either less active than estrogen, or its secretion is parallel to that of estrogen. Other investigators have reported the endocrine characteristics of women with idiopathic premature ovarian failure ]lS]. In spitz of increased LH and FSH levels, occasional follicular activity and prolorged secretion of E,. similar to the hormone secretion profile obtained from one VO.l...en (A.D.) in this study, have been documented. Measurements e!f pregnanediol, estrogen. FflH and LH leveis in urine collected at weekly intervzrs from pre- and peri-menopausal women also have been reparted ]19,20]. In long interval cycies, high gonadotrophin excretion has been cbserved, followed in a reciprocal fashion by elevations in urinary estrogens which, typically, led to vaginal bleeding. Interestingly, excretion patterns of many peri-menopausal women studied [19] suggest that estrogen hypersccretion may not be a rare finding among women undergoingthis transi-

tion. The present data confirm that women undergoing the perimenopausal transition may exhibit lengthened intermenstrual intervals associated with increased estrogen production. The possible association between such chronic, unopposed, high estrogen secretion and breast cancer in peri-menopausal women continues to be explored [21-231. In addition, high levels of unopposed estrogen appear to have a direct bearing on the degree and severitv of per&menopausal dysfunctional uterin bleeding [ll]. The variability of vasomotor symptoms observed to precede lhe menopause may also be related to irregular estrogen production. It is likely that this elevaeed estrogen secretion relates to the frequent endometrial abnormalities associated with this group of women. During the perimenopause, the endometrium frequently displays histological abnormalities, ns,mely, cystic glandular changes, infolding of epithelial tissues into glands and true adenomatous hyperplasia. The precise pathogenesis of endometrial hgperplasis and cancer have been elusive but further studies using methods described here may reveal useful information in this area. In some peri-menopausal women, the basal estrogen production rate may be difficult to r~sess by measuring circuiating estradiol levels because icvels are at, or near, the limit of detectability of most assay systems presently employed. The present dam demonstrate that rrrinary measurements of steroid conjugates accu:a:ely refiect the clculating concentrations of the active steroids even at low levels. Furthermore, serial urinary monitoring provides a practical method for the direct longitudinal ssessment of ovarian function through successive menstrual intervals, revealing those that are characteristic of the premer;jpausal transition period, i.e., variable in length with decreased luteal function [S]. Thus. this approach allows for prospective evaluations of perimenopausal women and may be used for therapeutic intervention to prevent repetitive episodes of endogenous hy&restrogenicity.

338

Acknowledgment The authors express their appreciation to Pam Carroll for the preparation of the manuscript and to Cheryl Miller, R.N. for excellent clinical assistance. We also appreciate the helpful comments of d. W. Overstreet and N. Santaro. This work was supported by NM grant RR/HD 01337-01, NIH grant ES0 4699-2, NICHD Clinical Investigator Award HD 00473, and NSF-BNS-8719698. References I 2

3 4 5 6 7 g 9 10 II li 13 I,1 I5 16

17 18 18 20

T’reloar AE. Ew,nton RE, Behn 86, Brown BW. Variation of the human menstrual cycle through reproductive life. lo* J Fertil 1967; IZ: 77-126. Vollman RF. The menstrual cycle. Philadelphia: Saunders. 1977: p. 138. Papaoicoiaou AD, Loraine JA, Dove GA, Loudon NB. Hormone cxcretic?. patterns in perimmopausal women. J Obstet Gyttaecol Br Commonw I%% 76: 308-316. Adatnopoulos DO, Lorainc JA, Dove. GA. Endoerinologic studies in women approaching the menopause. J Obstet Gytwcol Br Commonw 1971; 78: 62-79. Sherman BA, Korenman SG. Hormonal characteristics of the human menstrual cycle throughow reproductive bre. J Clin Invest 1975: 55: 6PP-7%. Shennatt EA. West JH. Korcnman SC. The menopusal transition: Analysis of LH. FSH, estradi01, and progesterone concetttratioos during metwcual cy:les of older women. J Clin Endocrine Metab1976 42: 629-634. Shnttmo BA, Wallace RB, Treloar AE. The menopausal tmnsition: Endocrinological and epidemiologiral considerations. J Biosoc Sci 1979; 6: 19-23. Metcalf MG. Donald RA, Livesey JH. Pituitary-ovarian function in normal women during the menopausal transition. Clio Endocrinol 1981: 14: 245-255. Reyes FI. Winter JD, Faimao C. Pituitary-ovarian relationships preceding the menopause. Am J Obstet Gyoecol 1977; 129: 557. Baltinger CB, Browning MCK, Smith AHW. Hormcne protile. and pqchological symptoms in perimmopausal woosen. Maturitas 1987; 9: 235-251. Abe T, Yauiaya Y, Wada Y. Suzuki M. Pltzitary-ovarian relationships in women apprcwhing the menopause. Mao&as i983; 5: 31-37. Longcope C. Franz C. Morel10 C, Baker R, Johnston CC, Jr. Steroid and gonadotropin levels in women during the peri-menopausal years. Maturitas 1986: 8: 18%196. Lasley BL. Stahenfeldt GH. Overstreet JW, Hanson FW, Czekala N, Munro C. Urinary hormone levels at the time of ovulation and implantation. Fenil Steril 1985; 43: 861-867. Taussky HH. A microcolori~~etric det:mroirz!io? cf rr:: ‘ine in urine by the Jaffe reaction. J Biol Chem IP54; 208: 853-861. DcVane GW. Vere Mt. Buhi WC, Kalra PS. Opioid peptiUes in prec.locyesis. Obstet Gynecol 1985; 65: 153-188. DeVaue GW, Soto D, Lasley BL. Daily urinary estrogen and progesterone metaboiites at the time of mesxrche. Phoeui., Arizona: Proceedings of the Society for Gynecdogic Investigation. 1985; March 21. .._ Goldenberg RL. Grodin JM, Rodbard D. Ross GT. Gonadotrophins in worn’& ~i!h ~menorrhea. Am J Obstet Gyoecol 1983: 116: 1003--1012. Rebar RW. Erickson GR. Yen SSC. ldiooathic oremature ovarian f !ilure: Clinical and endowi-? . rharacteris;ics. Fertil St&l 1982; 3:: 35-41. Metcalf MO. Donald RA, Livesay JH. Classification of menstrual cycles in pre and perimeno,msal women. J Endowinol 1981; 91: l-10. Met&f MG. Incidence of ovulatory cycles in worn” approaching the menopause. J Biosoc Sci 1979; 1t : 3P-48.

21 22 23

Metcalf MO, %iacktie ;A. MenstnmJ cycle and exposure to oeatro~cns unopposed by praestetonc: Rekvarxe to studies on breast cancer risk factors. J Endocrinol 1935: 104~ 137--141. Wallace RB. Sherman BM, Ban JA. Leper JP, Tretoar AE. Menstrual cyck patterns 2nd breast cancer risk factors. Cancer Rs 1978; 38: 4021-4024. Reed MJ. Bermek PA, James VHT. t??%qx~ produaion and met&&m in perimmopau~ women. Maturitas 1986; 8: 29-34.