2013

http://informahealthcare.com/gyeISSN: 0951-3590 (print), 1473-0766 (electronic)

Gynecol Endocrinol, 2013; 29(9): 826–829! 2013 Informa UK Ltd. DOI: 10.3109/09513590.2013.813466

PCO

Anti-Mullerian hormone and antral follicle count for predictionof ovarian stimulation response in polycystic ovary syndrome

Akmal El-Mazny and Nermeen Abou-Salem

Department of Obstetrics and Gynecology, Faculty of Medicine, Cairo University, Cairo, Egypt

Abstract

Objective: To evaluate the ability of a combination of multiple ovarian reserve markers topredict ovarian stimulation response in polycystic ovary syndrome (PCOS).Methods: On cycle Day 3 of 75 infertile patients with PCOS, serum follicle stimulating hormone(FSH), luteinizing hormone (LH), and anti-Mullerian hormone (AMH) were measured, and antralfollicle count (AFC) and ovarian volume (OV) were evaluated by transvaginal sonography (TVS).All patients underwent the same mild ovarian stimulation protocol using clomiphene citrateand highly purified FSH. Ovulation was monitored by TVS and confirmed by midluteal serumprogesterone level.Results: AMH, AFC, and ‘‘ovulation index’’ [OI, serum AMH (ng/ml)�bilateral AFC] weresignificantly lower in the ovulatory group (n¼ 57, 76%) compared with the anovulatory group,whereas LH, FSH, LH/FSH ratio, and OV were not significantly different. Using receiver-operatingcharacteristic curve analysis, the OI at a cutoff value of ‘‘85’’ had a sensitivity of 73.7% and aspecificity of 72.2% in the prediction of ovulation, with an area under the curve of 0.733.Patients with OI585 had significantly higher ovulation rate (p50.001).Conclusion: The OI, combining both AMH and AFC, is a potentially useful predictor of theoutcome of ovarian stimulation in PCOS.

Keywords

Anti-Mullerian hormone, antral follicle count,ovarian reserve, ovulation index, polycysticovary syndrome

History

Received 19 November 2012Revised 12 May 2013Accepted 6 June 2013Published online 15 July 2013

Introduction

Polycystic ovary syndrome (PCOS) is the most common cause ofanovulatory infertility, affecting �5–10% of women in theirreproductive age [1]. PCOS is clinically diagnosed when at leasttwo of the following three features are present: chronic oligo- oranovulation, biochemical hyperandrogenemia or hyperandrogen-ism, and polycystic ovarian morphology in ultrasound examin-ation [2]. Although the primary defect in PCOS is unclear, themain aim of treatment of anovulatory infertility is the induction ofmono-ovulatory cycles so that pregnancy can be achieved [3].

Ovarian reserve represents the remaining population ofprimordial and resting follicles, and can be defined as thequantity and quality of the follicles present in the ovary at a giventime that can be stimulated into dominant follicle growth [4].Therefore, accurate evaluation of ovarian reserve may allowindividualized patient management and prediction of ovarianstimulation outcome [5]. Traditionally, basal (Day 3) serumfollicle stimulating hormone (FSH) level, and antral follicle count(AFC) and ovarian volume (OV) assessed by transvaginalsonography (TVS) are often used as tests of ovarian reserve [6].

Anti-Mullerian hormone (AMH) is a dimeric glycoproteinmember of the transforming growth factor-b subfamily [7]. AMHis produced by the granulosa cells of pre-antral and small antralfollicles. Follicular growth is modulated by AMH, which inhibitsrecruitment of follicles from the primordial pool by modifying the

FSH sensitivity of those follicles. As a follicle matures, AMHproduction disappears allowing the follicle to complete thedevelopment process [8]. Since AMH levels reflect the numberof developing follicles, their measurement may be used as amarker of ovarian follicle impairment in PCOS and hence ovarianreserve [9].

The aim of this study was to evaluate the ability of acombination of multiple ovarian reserve markers to predictovarian stimulation response in PCOS.

Methods

This prospective clinical trial was conducted at the Department ofObstetrics and Gynecology, Kasr El-Ainy Teaching Hospital,Faculty of Medicine, Cairo University, during the period fromOctober 2010 to March 2012. The study protocol was approved bythe Scientific Research Committee and informed consent wasobtained from each patient.

Seventy-five infertile patients with PCOS, based on ESHRE/ASRM criteria [2], were enrolled in the study. At least two of thefollowing three features needed to be present: (i) oligo- oranovulation: 56 menstrual cycles per year; (ii) clinical and/orbiochemical signs of hyperandrogenism: hirsutism (Ferriman–Gallwey score 48), severe persistent acne, and/or total testoster-one level40.8 ng/ml; and (iii) sonographic evidence of PCOS: atleast 1 ovary containing 12 or more peripheral follicles measuring2–9 mm in diameter and/or OV of at least 10 cm3 [10].

The inclusion criteria for the study were: (i) age between25 and 35 years; (ii) primary or secondary infertility of at least2 years duration; (iii) body mass index (BMI) between 19 and30 kg/m2; (iv) both ovaries are present; and (v) adequate

Address for correspondence: Akmal El-Mazny, MD, FICS, Departmentof Obstetrics and Gynecology, Faculty of Medicine, Cairo University,Cairo, Egypt. Tel: 002-01001454576. Fax: 002-02-23643554. E-mail:dr_akmalelmazny@yahoo.com

Gyn

ecol

End

ocri

nol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

ichi

gan

Uni

vers

ity o

n 10

/17/

14Fo

r pe

rson

al u

se o

nly.

visualization of ovaries on TVS. The exclusion criteriawere: (i) the presence of other factors of infertility; (ii) otherovarian pathology; (iii) previous ovarian surgery; (iv) endocrineor medical disorders; and (v) current hormonal or ovulationinduction therapy.

On Day 3 of the cycle (or progestin-induced withdrawalbleeding), venous blood samples were obtained for FSH, lutein-izing hormone (LH), and AMH measurement; and basal TVSusing a 7.5-mHz transducer (Toshiba Famio-5 SSA-510A;Toshiba Medical Systems Corporation, Tochigi, Japan) wasperformed for bilateral AFC (510 mm in diameter) and summedOV calculation using the formula of ellipsoid(0.523� length�width� height).

All patients underwent the same mild ovarian stimulationprotocol. Clomiphene citrate [CC (clomid; Aventis Pharma,Cairo, Egypt)] 100 mg/day was administered orally from Day 3to Day 7 of the cycle. Highly purified human urofollitropin [HP u-FSH (fostimon; Institut Biochimique SA, Lugano, Switzerland)]75 IU/day was administered IM from Day 6 to Day 10.Transvaginal folliculometry was performed on Day 9 of thecycle and every other day. Human chorionic gonadotropin[hCG (choriomon; Institut Biochimique SA, Lugano,Switzerland)] 10 000 IU was administered IM when at least oneleading follicle reached �18 mm in diameter.

Ovulation (main outcome measure) was monitored by TVSand confirmed by midluteal (Day 21) serum progesterone level of�10 ng/ml. Secondary outcome measures included the day ofhCG administration, number of mature follicles (�18 mm),endometrial thickness (ET) and estradiol (E2) level as determinedon day of hCG administration, and midluteal serum progesteronelevel. The ovarian reserve markers were compared between theovulatory and the anovulatory group, and their possible predictivevalue for ovulation was calculated.

Blood samples were immediately centrifuged and serawere separated and stored at �70 �C until subsequent analysis.Serum AMH levels were measured using a specific enzyme-linked immunosorbent assay [11] with kit supplied by DSL (MIS/AMH ELISA; Diagnostic Systems Laboratories, Inc., Webster,TX). The procedure and interpretation of the assay wereperformed according to the manufacturer’s protocol. The sensi-tivity of the assay was 0.017 ng/ml and the intra- and interassaycoefficients of variation were 4.6 and 8.0%, respectively.Serum FSH, LH, E2, and progesterone levels were measuredusing enzyme-linked chemiluminescence immunometricassay on Immulite autoanalyzer [12] with kit suppliedby DPC (IMMULITE; Diagnostic Product Corporation, LosAngeles, CA).

Statistical analysis

Continuous data were expressed as mean�SD, and werecompared using Student’s t-test. Categorical data were expressedas number and percentage, and were compared using Fisher’sexact test. Receiver-operating characteristic (ROC) curve analysiswas used to evaluate the optimal cutoff value of ovarian reservemarkers for prediction of ovulation, based on an equivalentsensitivity and specificity, and the highest value of the area underthe curve (AUC). A p value 50.05 was considered statisticallysignificant. The SPSS (Statistical Package for the Social Science;SPSS Inc., Chicago, IL) was used for data analyses.

Results

A total of 75 eligible patients with PCOS completed the studyprotocol, 57 (76%) responded to ovarian stimulation (group I:ovulatory group), whereas 18 (24%) did not respond (group II:anovulatory group). The clinical characteristics: age, duration of

infertility, and BMI were not significantly different between thetwo groups. However, group I had significantly lower AMH leveland AFC (p50.05, for each) compared with group II, whereas LHand FSH levels, LH/FSH ratio, and OV were not significantlydifferent (Table 1).

To evaluate the ability of a combination of ovarian reservemarkers to predict ovulation, a single index including AMH andAFC was hypothesized. This novel ‘‘ovulation index’’ (OI) wascalculated as serum AMH level (ng/ml)� bilateral AFC. The OIwas significantly lower in group I (79.6� 2.8) than in group II(87.9� 3.1), p50.001 (Table 1). Using ROC curve analysis, theOI at a cutoff value of ‘‘85’’ had a sensitivity of 73.7% and aspecificity of 72.2% in the prediction of ovulation, with an AUCof 0.733, p50.001 (Figure 1).

Patients with OI585 had significantly lower AMH level andAFC (p50.05, for each), and significantly higher ovulation rate(p50.001) than those with OI� 85, whereas the clinical charac-teristics: LH and FSH levels, LH/FSH ratio, and OV were notsignificantly different. The day of hCG administration, number ofmature follicles (�18 mm), ET and E2 level on day of hCGadministration, and midluteal serum progesterone level were alsonot significantly different between the ovulatory cases in the twogroups (Table 2).

Discussion

In the present study, we have evaluated multiple ovarian reservemarkers, including FSH, AMH, AFC, and OV, in 75 infertilepatients with PCOS before undergoing ovarian stimulation.Our results showed that AMH level and AFC were significantlylower in women with PCOS who responded to ovarian stimulationcompared with those who did not (p50.05, for each), whereasLH and FSH levels, LH/FSH ratio, and OV were not significantlydifferent.

Earlier data showed that in women with PCOS, serum andfollicular AMH levels are higher than in normal controls [13,14].Subsequent data confirmed this finding and indicated that theseelevated AMH levels are related to increased number of smallantral follicles seen in PCOS [15–17]. Therefore, serum AMHvalues could be a precise, subsidiary diagnostic marker of the

Table 1. Characteristics of PCOS patients according to ovarian response.

VariableGroup I:ovulatory

Group II:anovulatory p Value

No. of patients (n) 57 18Age (years) 28.5� 1.7 27.8� 2.5 NSDuration of

infertility (years)2.9� 0.6 3.1� 0.7 NS

BMI (kg/m2) 26.3� 1.9 25.7� 2.4 NSLH (mlU/ml) 10.1� 2.4 9.2� 2.3 NSFSH (mIU/ml) 4.8� 1.9 4.9� 1.6 NSLH/FSH ratio 2.2� 0.3 2.1� 0.1 NSAMH (ng/ml) 6.5� 1.2 7.3� 1.3 50.05AFC (n) 11.2� 2.6 13.1� 3.2 50.05OV (cm3) 22.4� 3.9 22.7� 3.8 NSDay of hCG

administration12.1� 1.6 NA NA

Follicles �18 mm (n) 1.7� 0.7 NA NAET on day of

hCG (mm)9.2� 1.5 NA NA

E2 on day ofhCG (pg/ml)

170.6� 32.8 NA NA

Progesterone(ng/ml)

12.4� 1.4 NA NA

OI [AMH(ng/ml)�AFC]

79.6� 2.8 87.9� 3.1 50.001

NA¼ not applicable; NS¼ not significant.

DOI: 10.3109/09513590.2013.813466 Ovulation index in PCOS 827

Gyn

ecol

End

ocri

nol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

ichi

gan

Uni

vers

ity o

n 10

/17/

14Fo

r pe

rson

al u

se o

nly.

syndrome, particularly in cases in which the TVS examination isnot feasible [9,18].

This increase in AMH has been implicated in the pathogenesisof PCOS. High serum AMH levels in PCOS lower the folliclesensitivity to circulating FSH, thus preventing follicle selectionand resulting in follicle arrest at the small antral phase, withfailure of dominance. Furthermore, AMH inhibits the productionof aromatase, which is activated by FSH action on granulosa cells,resulting in reduction of follicular production of E2. The resultinglow levels of E2 may also contribute to the failure of follicleselection [19].

Our results provide an addition to the available evidencesupporting the correlation between high AMH levels andpoor ovarian response to treatment. El-Halawaty et al. [20]found that AMH level, at cutoff value of 1.2 ng/ml, could beused to predict response to CC in obese women with PCOS.They also noted that FSH concentration was not significantlydifferent between women who responded to CC and those whodid not. Furthermore, Amer et al. [21] concluded that pre-treatment circulating AMH level, at cutoff value of 7.7 ng/ml,

seems to be a good predictor of the ovarian response tolaparoscopic ovarian diathermy in PCOS.

Several previous studies on ovarian reserve in women withoutPCOS undergoing IVF treatment have shown that low AMHlevels are associated with poor ovarian response to stimulation[22–24], whereas, on the opposite end of the spectrum, high AMHlevels are associated with ovarian hyperstimulation syndrome[25]. Therefore, it was possible to hypothesize that normal AMHlevels are necessary to achieve optimal ovarian response tostimulation [21]. Furthermore, the consistency of serum AMHlevels throughout the entire menstrual cycle, unlike FSH, makes itan attractive marker of response to treatment [9].

In this study, we have for the first time evaluated the ability ofa combination of ovarian reserve markers to predict ovulation inwomen with PCOS undergoing ovarian stimulation. The OI,calculated as serum AMH level (ng/ml)� bilateral AFC, wassignificantly lower in the ovulatory group (79.6� 2.8) than in theanovulatory group (87.9� 3.1), p50.001. Using ROC curveanalysis, the OI at a cutoff value of ‘‘85’’ had a sensitivity of73.7% and a specificity of 72.2% in the prediction of ovulation,with an AUC of 0.733, p50.001. Patients with OI 585 hadsignificantly higher ovulation rate (p50.001) than those withOI� 85. However, our results may be limited by the relativelysmall sample size and potential discrepancies in sample size ofthe two groups.

In conclusion, the OI, combining both AMH and AFC, wasfound to be a potentially useful predictor of the outcome ofovarian stimulation in PCOS. We have also identified a cutoffvalue of ‘‘85’’, above which the chances of ovulation may besignificantly reduced. Assessment of OI could help in selection oftreatment and counseling of patients about the expected outcomeof ovulation induction. Further studies are planned to investigatethe possible prognostic value of OI in women with PCOSundergoing IVF treatment, and of a ‘‘modified’’ OI in womenwithout PCOS.

Declaration of interest

The authors report no conflicts of interest. The authors alone areresponsible for the content and writing of this article.

References

1. Speroff L, Fritz M. Clinical gynecologic endocrinology andinfertility. 7th ed. New York, NY: Lippincott Williams & Wilkins;2005:485–513.

2. Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus WorkshopGroup. Revised 2003 consensus on diagnostic criteria and long termhealth risks related to polycystic ovary syndrome (PCOS).Hum Reprod 2004;19:41–7.

3. ESHRE Capri Workshop Group. Mono-ovulatory cycles: a key goalin profertility programmes. Hum Reprod Update 2003;9:263–74.

4. Broekmans FJ, Kwee J, Hendriks DJ, et al. A systematic review oftests predicting ovarian reserve and IVF outcome. Hum ReprodUpdate 2006;12:685–718.

5. Devroey P, Fauser BC, Diedrich K; Evian Annual Reproduction(EVAR) Workshop Group 2008. Approaches to improve thediagnosis and management of infertility. Hum Reprod Update2009;15:391–408.

6. Roudebush WE, Kivens WJ, Mattke JM. Biomarkers of ovarianreserve. Biomark Insights 2008;3:259–68.

7. La Marca A, Volpe A. Anti-Mullerian hormone (AMH) in femalereproduction: is measurement of circulating AMH a useful tool?Clin Endocrinol 2006;64:603–10.

8. Visser JA, de Jong FH, Laven JS, Themmen AP. Anti-Mullerianhormone: a new marker for ovarian function. Reproduction 2006;131:1–9.

9. Karkanaki A, Vosnakis C, Panidis D. The clinical significance ofanti-Mullerian hormone evaluation in gynecological endocrinology.Hormones 2011;10:95–103.

Figure 1. ROC curve of OI for prediction of ovulation.

Table 2. Comparison of patients’ characteristics according to OI.

Variable OI585 OI� 85 p Value

No. of patients (n) 47 28Age (years) 28.7� 1.5 28.5� 2.1 NSDuration of

infertility (years)2.9� 0.5 2.7� 0.9 NS

BMI (kg/m2) 26.1� 2.1 26.5� 1.9 NSLH (mlU/ml) 10.1� 2.3 9.3� 2.4 NSFSH (mIU/ml) 4.9� 1.8 4.6� 1.5 NSLH/FSH ratio 2.1� 0.4 2.2� 0.3 NSAMH (ng/ml) 6.7� 0.9 7.3� 1.1 50.05AFC (n) 11.3� 2.8 12.9� 3.1 50.05OV (cm3) 21.9� 3.8 23.4� 3.9 NSDay of hCG administration* 11.8� 1.7 12.4� 1.5 NSFollicles �18 mm (n)* 1.8� 0.8 1.6� 0.5 NSET on day of hCG (mm)* 9.4� 1.5 8.9� 1.4 NSE2 on day of hCG (pg/ml)* 171.6� 31.5 168.9� 35.1 NSProgesterone (ng/ml)* 12.5� 1.4 12.3� 1.5 NSOvulation rate (n) 42 (89.4%) 15 (53.6%) 50.001

NS¼ not significant. *In ovulatory cases only.

828 A. El-Mazny & N. Abou-Salem Gynecol Endocrinol, 2013; 29(9): 826–829

Gyn

ecol

End

ocri

nol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

ichi

gan

Uni

vers

ity o

n 10

/17/

14Fo

r pe

rson

al u

se o

nly.

10. Balen AH, Laven JS, Tan SL, Dewailly D. Ultrasound assessmentof the polycystic ovary: international consensus definitions.Hum Reprod Update 2003;9:505–14.

11. Freour T, Mirallie S, Bach-Ngohou K, et al. Measurement of serumanti-Mullerian hormone by Beckman Coulter ELISA and DSLELISA: comparison and relevance in assisted reproduction technol-ogy (ART). Clin Chem Acta 2007;375:162–4.

12. Babson AL. The Cirrus IMMULITE automated immunoassaysystem. J Clin Immunoassay 1991;14:84–8.

13. Fallat ME, Siow Y, Marra M, et al. Mullerian inhibiting substance infollicular fluid and serum: a comparison of patients with tubalfactor infertility, polycystic ovary syndrome and endometriosis.Fertil Steril 1997;67:962–5.

14. Cook C, Siow Y, Bremer AG, Fallat ME. Relation betweenMullerian inhibiting substance and other reproductive hormones inuntreated women with PCOS and normal women. Fertil Steril 2002;77:141–6.

15. Pigny P, Merlen E, Robert Y, et al. Elevated serum level of AMH inpatients with polycystic ovary syndrome: relationship to the ovarianfollicle excess and to the follicular arrest. J Clin Endocrinol Metab2003;88:5957–62.

16. Laven JS, Mulders AG, Visser JA, et al. Anti-Mullerian hormoneserum concentrations in normoovulatory and anovulatory women ofreproductive age. J Clin Endocrinol Metab 2004;89:318–23.

17. Begawy AF, El-Mazny AN, Abou-Salem NA, El-Taweel NE.Anti-Mullerian hormone in polycystic ovary syndrome and normo-ovulatory women: correlation with clinical, hormonal and ultra-sonographic parameters. Middle East Fertil Soc J 2010;15:253–8.

18. Pigny P, Jonard S, Robert Y, Dewailly D. Serum AMH as a surrogatefor antral follicle count for definition of polycystic ovary syndrome.J Clin Endocrinol Metab 2006;91:941–5.

19. Grossman M, Nakajima S, Fallat M, Siow Y. Mullerian inhibitingsubstance inhibits cytochrome P450 aromatase activity in humangranulosa lutein cell culture. Fertil Steril 2008;89:1364–70.

20. El-Halawaty S, Rizk A, Kamal M, et al. Clinical significance ofserum concentration of anti-Mullerian hormone in obese womenwith polycystic ovary syndrome. Reprod Biomed Online 2007;15:495–9.

21. Amer SA, Li TC, Ledger WL. The value of measuring anti-Mullerian hormone in women with anovulatory polycystic ovarysyndrome undergoing laparoscopic ovarian diathermy. Hum Reprod2009;24:2760–6.

22. Seifer DB, MacLaughlin DT, Christian BP, et al. Early follicularserum Mullerian-inhibiting substance levels are associated withovarian response during assisted reproductive technology cycles.Fertil Steril 2002;77:468–71.

23. Nardo LG, Gelbaya TA, Wilkinson H, et al. Circulating basal anti-Mullerian hormone levels as predictor of ovarian response in womenundergoing ovarian stimulation for in vitro fertilization. Fertil Steril2009;92:1586–93.

24. Broer S, Mol BW, Dolleman M, et al. The role of anti-Mullerianhormone assessment in assisted reproductive technology outcome.Curr Opin Obstet Gynecol 2010;22:193–201.

25. Nakhuda GS, Chu MC, Wang J, et al. Elevated serum MIS levelsmay be a marker for ovarian hyperstimulation syndrome in normalwomen undergoing IVF. Fertil Steril 2006;85:1541–3.

DOI: 10.3109/09513590.2013.813466 Ovulation index in PCOS 829

Gyn

ecol

End

ocri

nol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

ichi

gan

Uni

vers

ity o

n 10

/17/

14Fo

r pe

rson

al u

se o

nly.