Serum inhibin A, inhibin B, and pro-�Clevels are altered after surgically orpharmacologically induced menopause

Luigi Cobellis, M.D., Stefano Luisi, M.D., Ilaria Pezzani, M.D.,Fernando M. Reis, M.D., Ph.D., Vincenzo De Leo, M.D., and Felice Petraglia, M.D.

Chair of Obstetrics and Gynecology, University of Siena, Siena, Italy

Objective: To evaluate the course of changes in serum inhibin A, inhibin B, and pro-�C levels in women withsurgically or pharmacologically induced menopause.

Design: Longitudinal study.

Setting: Academic Health Center of Siena, Siena, Italy.

Patient(s): Four groups of women were studied: [1] surgical menopause including bilateral oophorectomy(n � 15), [2] amenorrhea induced by GnRH-analogue for treatment of endometriosis (n � 13), [3] amenorrheainduced by antineoplastic chemotherapy before (n � 15) and after chemotherapy (n � 13), and [4] controlphysiological menopause (n � 67).

Intervention(s): Collection of blood specimens.

Main Outcome Measure(s): Serum inhibin A, inhibin B, and pro-�C concentrations were measured by usingspecific two-site ELISAs.

Result(s): Following oophorectomy, serum inhibin A, inhibin B, and pro-�C levels were decreased on thefirst postoperative day; on the fifth postoperative day they were still significantly reduced. Women withamenorrhea induced by GnRH-analogue treatment exhibited serum inhibin A and pro-�C levels that weresignificantly higher than those observed in physiological menopause. Patients undergoing antineoplasticchemotherapy had higher serum inhibin A levels than those in physiological menopause, whereas inhibin Band pro-�C levels did not differ. During the course of chemotherapy, median serum inhibin A concentrationswere similar to those of patients evaluated after the suspension of treatment. In postmenopause, inhibin A, andinhibin B levels were low, whereas levels of pro-�C were still detectable.

Conclusion(s): Circulating levels of inhibin A, inhibin B, and pro-�C are reduced after oophorectomy.Women with amenorrhea induced by GnRH-analogue treatment or by antineoplastic chemotherapy stillproduce inhibin A and pro-�C. This probably reflects a residual ovarian function and hormone synthesis.Therefore, the ovary may be a source of pro-�C after menopause; significant amounts of pro-�C are presentin circulation after natural menopause, but not after oophorectomy. (Fertil Steril� 2002;77:745–9. ©2002 byAmerican Society for Reproductive Medicine.)

Key Words: Inhibin A, inhibin B, pro-�C, menopause, surgical menopause, GnRH-analogue, chemotherapy

The inhibins were first characterized fortheir inhibitory effect on pituitary follicle-stimu-lating hormone (FSH) secretion and are now re-garded as members of the transforming growthfactor-� (TGF-�) superfamily (1, 2). The inhib-ins are dimers composed by a common �-subunitcoupled to one of the two �-subunits (�A and�B), resulting in inhibin A (��A) and inhibin B(��B). In the circulation, both inhibin A andinhibin B are detectable as well as a precursor ofthe �-chain named pro-�C. Pro-�C contains thepro-region of the �-subunit and the monomericpro-� subunit is biologically inactive.

Over the last decade, the development of atwo-site enzyme-linked immunosorbent assay(ELISA) for the inhibin family has allowed abetter understanding of these proteins in humanendocrinology (3). Circulating inhibin levelschange during a woman’s reproductive years;inhibin A is released from dominant folliclesand the corpus luteum, whereas inhibin B se-cretion is a product of developing preantral andsmall antral follicles (4). Furthermore, otherparacrine/autocrine factors modulate inhibinsecretion, and inhibin A and inhibin B releaseare differently regulated by FSH and IGF-I (5).

Received April 23, 2001;revised and acceptedSeptember 27, 2001.Reprint requests: FelicePetraglia, M.D., Chair ofObstetrics andGynecology, University ofSiena, Policlinico LeScotte, Viale Bracci, 53100Siena, Italy (FAX: �39-0577-233454; E-mail:petraglia@unisi.it).

FERTILITY AND STERILITY�VOL. 77, NO. 4, APRIL 2002Copyright ©2002 American Society for Reproductive MedicinePublished by Elsevier Science Inc.Printed on acid-free paper in U.S.A.

0015-0282/02/$22.00PII S0015-0282(01)03234-4

745

The reduction of ovarian reserve is also related to a decreaseof the inhibins; low levels of both inhibin A and inhibin Bare typical of postmenopausal women (6). Indeed, womenwith premature ovarian failure have very low serum levels ofinhibins (7), and low inhibin A and inhibin B levels aretypical of older fertile women (8). Moreover, low inhibin Blevels 3 days after ovulatory stimulation protocol are predic-tors of a low ovarian reserve and a reduced reproductiveoutcome (9, 10). The function of pro-�C has not yet beenascertained and no significant studies have been performed.

A previous study indicated a decrease of serum inhibin Aand inhibin B levels accompanies the menopausal transition(11), but no studies have been conducted on circulatinginhibin A, inhibin B, and pro-�C after surgical oophorec-tomy, after the use of GnRH-analogue, or after chemother-apy for oncological treatments (conditions that mimic amenopausal status).

Therefore, the aim of our study was to evaluate the timecourse of serum inhibin A, inhibin B, and pro-�C changesfollowing surgical menopause, the use of GnRH-analogue,or antineoplastic chemotherapy in fertile women. Womenexhibiting physiological menopause served as a controlgroup.

MATERIALS AND METHODS

PatientsThe study population consisted of four groups of women

as follows.

Group 1: women undergoing surgical menopause (n �15). In all cases, hysterectomy and bilateral oophorectomyhad been performed for treatment of uterine fibroids associ-ated with ovarian cysts. The median age was 49 years(interquartile range 48 to 50 years).

Group 2: younger women (n � 13) (median age � 31years, interquartile range: 29 to 33 years) in whom amenor-rhea was induced by GnRH-analogues for treatment of en-dometriosis. All patients had menstrual cycles before theonset of treatment.

Group 3: fertile women with breast cancer and/orHodgkin disease who experienced secondary amenorrheaduring antineoplastic chemotherapy. They were evaluatedduring the course (n � 15, median age � 28 to 42 years) andafter (n � 13) chemotherapy treatment.

Group 4: healthy women (n � 40) who had experienceda natural menopause (lasting 3 to 10 years). The median timeelapsed since the last menopausal period was 6 years and themedian age was 57 years. All of the women had beenreferred to our menopause center to monitor their bonemetabolism. Exclusion criteria were malignant conditions.

All participants provided a blood sample by venipuncturebetween 08:00 and 10:00 A.M. and serum was immediatelyseparated and stored at �20°C. In group 1, serial blood

samples were taken the day before the operation and 1, 3,and 5 days after the surgery.

The study was approved by the institutional review boardof the Academic Health Center of Siena and informed con-sent was obtained from the participants.

Serum Hormone MeasurementsInhibin A, inhibin B, and pro-�C concentrations were

measured in all serum samples using specific two-siteELISAs purchased from Serotec (Oxford, UK). Standardsand samples were diluted as appropriate and mixed with anequal volume of distilled water containing 6% (w/v) SDS.Samples were then boiled for 3 minutes, and after coolingfreshly prepared hydrogen peroxide solution was added.Denatured and oxidized samples were loaded onto antibody-coated microtiter plates containing the specific primary an-tibodies for each assay.

Inhibin A AssayPlates were incubated at room temperature for 2 hours.

After washing with a phosphate buffer, 50 �L of alkalinephosphatase–conjugated extravidin was added, and plateswere incubated for 1 hour. The plates were read at 490 nm onan automated ELISA plate reader (BRIO: Basic Radim Im-munoassay Operator, Radim spa, Pomezia, Italy).

The inhibin A detection limit was 4 pg/mL serum, withintra-assay and interassay coefficients of variation (CVs) forquality control samples of 4.0% and 8.0%, respectively.Cross-reactions for each assay with the various inhibin-related proteins were �0.5%.

Inhibin B AssayPlates were incubated overnight at room temperature.

After washing with enzyme immunoassay (EIA) wash buffer[0.1 mol/L Tris-HCl, 0.15 mol/L NaCl, 10% (wt/vol) BSA,5% (vol/vol) Triton X-100, and 0.1% (wt/vol) sodium azide,pH 7.5], 50 �L of alkaline phosphatase–conjugated Fabmouse anti-human inhibin �-subunit was added, and plateswere incubated for 3 hours. Plates were washed and boundalkaline phosphatase was quantitated adding 50 �L of sub-strate solution. After 1 hour of incubation, 50 �L of ampli-fier solution was added and the plates were read at 490 nm.

The assay detection limit for inhibin B was less than 5pg/mL. Within-plate and between-plate CVs were 6% and8%, respectively. Cross-reactions for the assay with thevarious proteins of the inhibin-related family were �0.1%.

Pro-�C AssayPlates were incubated overnight at 4°C. After washing

with EIA wash buffer, 50 �L of alkaline phosphatase–conjugated Fab mouse anti-human inhibin �-subunit wasadded, and plates were incubated for 1 hour. Plates werewashed and bound alkaline phosphatase was identified byadding 50 �L of substrate solution. After 2 hours of incu-bation, 50 �L of amplifier solution was added and the plateswere read at 490 nm.

746 Cobellis et al. Inhibin-related proteins and menopause Vol. 77, No. 4, April 2002

The assay detection limit for pro-�C was less than 2pg/mL. Within-plate and between-plate coefficients of vari-ation were 6.5% and 8%, respectively. Cross-reactions withthe various proteins of the inhibin-related family were�0.1%.

Statistical AnalysisFor statistical purposes, samples that were below the

specific assay detection limit were assumed to be equal to thedetection limit of the assay.

Because the hormone concentrations deviated from nor-mal distribution with significant skewness and kurtosis, theresults are presented as medians and interquartile ranges(IQR). Differences between groups were verified withKruskal-Wallis analysis of variance (ANOVA) followed byDunn’s test for nonparametric multiple comparisons. Fried-man’s ANOVA was applied to the repeated measures ob-tained before and after surgical intervention. Linear correla-tion was tested by calculating the Spearman’s rankcorrelation coefficient. P�.05 was considered to be statisti-cally significant.

RESULTS

Surgical MenopauseOophorectomy reduced circulating levels of inhibin A,

inhibin B, and pro-�C. Preoperatively, circulating inhibin Alevels had a median of 28.5 pg/mL with a range of 15.9–39.0pg/mL. However, 5-day postoperative levels of inhibin Awere significantly reduced to a median of 22.9 pg/mL, witha range of 14.1–28.2 pg/mL (P�.05). Inhibin B levels de-creased from a median of 14.3 pg/mL with a range of8.3–19.6 pg/mL to a 5-day postoperative level of 8.1 pg/mLwith a range of 6.4–11.3 pg/mL (P�.001). The circulatingserum level of pro-�C prior to oophorectomy was 144.9pg/mL with a range of 82.7–236.5 pg/mL, and decreased to12.4 pg/mL with a range of 7.2–26.3 pg/mL 5 days after theprocedure (P�.0001). The circulating levels inhibin A, in-hibin B, and pro-�C for each patient are depicted in Figure1. These curves clearly illustrate that the levels of all threeproteins fell after oophorectomy, reaching very low concen-trations up to the fifth postoperative day.

Medical MenopauseWomen who exhibited amenorrhea that was attributed to

the treatment of endometriosis by a GnRH-analogue hadsignificantly higher circulating inhibin A levels than did thewomen observed after physiological menopause (P�.001;Fig. 2A). Also, pro-�C levels were significantly higher inwomen receiving GnRH-analogue as compared to the con-trol group (P�.01; Fig. 2C), but inhibin B levels remainedunchanged (see Fig. 2B).

Patients undergoing antineoplastic chemotherapy had el-evated serum inhibin A levels (Fig. 2A), but inhibin B andpro-�C levels remained unchanged (Fig. 2B and 2C). After

chemotherapy there was no significant difference in circu-lating levels of all three proteins (Fig. 2A–C).

DISCUSSION

This is the first report that examines the time course ofserum concentration changes of inhibin-related proteins in

F I G U R E 1

Time course of serum inhibin A, inhibin B, and pro-�C levelsfollowing surgical menopause. Blood samples were collectedthe day before and on days 1, 3, and 5 after oophorectomy.The curves represent individual data (n � 15). The changes inall three proteins over time were statistically significant(Friedman’s ANOVA).

Cobellis. Inhibin-related proteins and menopause. Fertil Steril 2002.

FERTILITY & STERILITY� 747

surgical and pharmacological menopause, compared with ahealthy menopausal control group.

Oophorectomy induced a rapid decrease in all three pro-teins, which all were still significantly reduced 5 days afterthe operation. In contrast, pharmacological intervention forendometriosis with a GnRH-analogue, which caused amen-orrhea, significantly elevated inhibin A and pro-�C levelsbut had no effect on inhibin B. Fertile women who exhibitedbreast cancer and/or Hodgkin disease who experienced a

secondary amenorrhea due to antineoplastic chemotherapywere also tested and none of the proteins were changed,either during or after chemotherapy treatment.

Previous reports have indicated that serum inhibin A andinhibin B levels are low or almost undetectable in postmeno-pausal women and that pro-�C concentrations remain mea-surable in intact postmenopausal women (6, 11, 12). It isinteresting that we observed that pro-�C was still detectablein the circulation even some years after the last menses; andin the group of surgical menopause patients, the inhibin-related protein levels fell within few days after oophorec-tomy, reaching the lowest amounts by the fifth day after thesurgical procedure. These results indicate that the ovary isthe main source of dimeric inhibins and pro-�C in women.

Endometriosis is characterized by normal inhibin A andinhibin B serum levels despite the production and secretionof both dimers at endometriotic implants, suggesting a roleof these proteins in local cellular modulation and differenti-ation (13, 14). Our observations do not confirm previousresults by Lockwood et al. (15), who suggested that pituitarydesensitization, in preparation of ovarian stimulation, re-duced more inhibin A than inhibin B levels and left pro-�Clevels unaltered. The increase in circulating levels of inhibinA and pro-�C supports the hypothesis of a different endo-crine regulation of these proteins and of a residual ovarianactivity.

Antineoplastic chemotherapies administrated for lym-phoma or breast cancer result in an ovarian toxicity in fertilewomen inducing a decrease of inhibin A, inhibin B, andpro-�C levels. A percentage of 48% to 77% of womentreated with chemotherapy for lymphoma exhibited amenor-rhea or ovarian failure (16).

The inhibin-related protein levels after the suspension oftreatment are similar to those during chemotherapy. Serumlevels of inhibin A was higher in menopause, but duringGnRH-analogue treatment, the gonadal suppression obtainedwith antineoplastic drugs suggests a residual secretion ofinhibin A and pro-�C, demonstrating an ovarian activity inwhich granulosa cells or other ovarian compartments arefunctioning. A different source of active proteins in responseto chemotherapy/GnRH-analogue administration may repre-sent another hypothesis.

A physiological role of the �-chain precursors is currentlyunder investigation. Recently, several reports have suggestedalterations of pro-�C levels during physiological and patho-physiological condition (8, 17, 18). Muttukrisha et al. (8)have demonstrated that pro-�C is mainly a luteal product,reaching the highest concentrations during the luteal phase.However, in aging fertile women, pro-�C levels increase,suggesting that this rise is related to ovarian aging (8).

The detection of high concentrations of pro-�C in patientswith epithelial ovarian cancer suggests that this protein maybe implicated during ovarian tumorigenesis. Serum pro-�C

F I G U R E 2

(A) Serum inhibin A, (B) inhibin B, and (C) pro-�C levels inwomen treated with GnRH-analogue (GnRH-a); women un-dergoing antineoplastic chemotherapy (CT during chemo-therapy); women who had finished antineoplastic chemother-apy (CT after chemotherapy); and women who hadexperienced physiological menopause (menopause). Dataare reported as single points and median values are indicated(horizontal bars). *P�.05 vs. menopause group (Dunn’s test).

Cobellis. Inhibin-related proteins and menopause. Fertil Steril 2002.

748 Cobellis et al. Inhibin-related proteins and menopause Vol. 77, No. 4, April 2002

concentrations were found elevated in 41% of women withepithelial ovarian cancer (17) and were twofold to tenfoldhigher than the concentrations in healthy postmenopausalwomen (18). In the latter report, the investigators also de-termined the pro-�C levels in peritoneal fluid, showing anelevated level in both benign and in malignant epithelialneoplasms. If the postmenopausal ovary continues to pro-duce pro-�C, as suggested by these findings, this mightcontribute to the intriguing association between menopause,which is a state of relative ovarian inactivity, and a higherincidence of ovarian tumors.

In conclusion, surgical menopause induces a drasticchange in circulating hormones and this is reflected by thedecrease in the inhibin-related proteins. Furthermore, themenopausal ovary is a source of �-chain during naturalmenopause, but this effect is absent after oophorectomy.During pharmacological treatment of endometriosis, inhibinA and pro-�C were increased in significant amounts, indi-cating residual ovarian activity despite granulosa suppres-sion.

References1. Vale W, Rivier C, Hsueh A, Campen C, Meunier H, Bicsak T, et al.

Chemical and biochemical characterization of the inhibin family pro-tein hormones. Recent Prog Horm Res 1988;44:1–34.

2. Burger HG. Clinical utility of inhibin measurements. J Clin EndocrinolMetab 1993;76:1391–6.

3. Bhasin S, de Krester DD. Measurement of circulating inhibin levels:revisiting the inhibin hypothesis. J Clin Endocrinol Metab 1996;81:1318–20.

4. Petraglia F, Zanin E, Faletti A, Reis FM. Inhibins: paracrine andendocrine effects in female reproductive function. Curr Opin ObstetGynecol 1999;11:241–7.

5. Welt CK, Schneyer AL. Differential regulation of inhibin B and inhibinA by follicle-stimulating hormone and local growth factors in humangranulosa cells from small antral follicles. J Clin Endocrinol Metab2001;86:330–6.

6. Burger HG, Dudley EC, Hopper JL, Shelley JM, Green A, Smith A, etal. The endocrinology of the menopausal transition: a cross-sectional

study of a population-based sample. J Clin Endocrinol Metab 1995;80:3537–45.

7. Petraglia F, Hartmann B, Luisi S, Florio P, Kirchengast S, Santuz M, etal. Low levels of serum inhibin A and inhibin B in women withhypergonadotropic amenorrhea and evidence of high levels of activin Ain women with hypothalamic amenorrhea. Fertil Steril 1998;70:907–12.

8. Muttukrishna S, Child T, Lockwood GM, Groome NP, Barlow DH,Ledger WL. Serum concentrations of dimeric inhibins, activin A,gonadotrophins and ovarian steroids during the mestrual cycle in olderwomen. Hum Reprod 2000;15:549–56.

9. Welt CK, McNicholl DJ, Taylor AE, Hall JE. Female reproductiveaging is marked by decreased secretion of dimeric inhibin. J ClinEndocrinol Metab 1999;84:105–11.

10. Creus M, Penarrubia J, Fabregues F, Vidal E, Carmona F, CasamitjanaR, et al. Day 3 serum inhibin B and FSH and age as predictors ofassisted reproduction treatment outcome. Hum Reprod 2000;15:2341–6.

11. Burger HG, Dudley EC, Hopper JL, Groome N, Guthrie JR, Green A,et al. Prospectively measured levels of serum follicle-stimulating hor-mone, estradiol, and the dimeric inhibins during the menopausal tran-sition in a population-based cohort of women. J Clin Endocrinol Metab1999;84:4025–30.

12. Groome NP, Illingworth PJ, O’Brien M, Priddle J, Weaver K, McNeillyAS. Quantification of inhibin pro-alpha C-containing forms in humanserum by a new ultrasensitive two-site enzyme-linked immunosorbentassay. J Clin Endocrinol Metab 1995;80:2926–32.

13. Florio P, Luisi S, Vigano P, Busacca M, Fadalti M, Genazzani AR, etal. Healthy women and patients with endometriosis show high concen-trations of inhibin A, inhibin B, and activin A in peritoneal fluidthroughout the menstrual cycle. Hum Reprod 1998;13:2606–11.

14. Reis FM, Di Blasio AM, Florio P, Ambrosini G, Di Loreto C, PetragliaF. Evidence for local production of inhibin A and activin A in patientswith ovarian endometriosis. Fertil Steril 2001;75:367–73.

15. Lockwood GM, Muttukrishna S, Groome NP, Knight PG, Ledger WL.Circulating inhibins and activina A during GnRH-analogue down-regulation and ovarian hyperstimulation with recombinant FSH forin-vitro fertilization-embryo transfer. Clin Endocrinol (Oxf) 1996;45:741–8.

16. Kreuser ED, Felsenberg D, Behles C, Seibt-Jung H, Mielcarek M, DiehlV, et al. Long-term gonadal dysfunction and its impact on bone min-eralization in patients following COPP/ABVD chemotherapy forHodgkin’s disease. Ann Oncol 1992;3(Suppl 4):105–10.

17. Menon U, Riley SC, Thomas J, Bose C, Dawnay A, Evans LW, et al.Serum inhibin, activin and follistatin in postmenopausal women withepithelial ovarian carcinoma. BJOG 2000;107:1069–74.

18. Ala-Fossi SL, Maenpaa J, Blauer M, Tuohimaa P, Punnonen R. InhibinA, B and pro-alphaC in serum and peritoneal fluid in postmenopausalpatients with ovarian tumors. Eur J Endocrinol 2000;142:334–9.

FERTILITY & STERILITY� 749