Table of contents

Original articles

113 Does the addition of LH activity to FSH make gonadotrophins more superior? A systematic review and meta-analysis

Hesham Al-Inany, Amr Wahba, Hatem Abu Hashim, Human Fatemi and Ahmed Abousetta

121 A comparative study between cabergoline, coasting, and step-down regimens in the prevention of severe OHSS and their correlation

with pregnancy rate in intracytoplasmic sperm injection cycles

Hoda Abdelaal, Olfat N. Riade and Mostafa Mahmoud

126 Antral follicular count is a better indicator for the assessment of ovarian reserve

Ahmed Soliman, Sherif Sharkawy, Tamer Abdelmoniem, Hesham Al-Inany and Odette Wahba

133 Comparison between transfrontal and axial three-dimensional ultrasound acquisition in the visualization of midline structures

of the fetal brain

Sherif M.M. Negm and Rasha A. Kamel

138 Glycosylated hemoglobin level at 34 weeks in insulin-controlled diabetic pregnancies, relation to fetal outcome

Hoda Abdelaal, Mohamed A. Kasem, Abdelmeguid Ali and Alaa M. Abd Elazem

142 Evaluation of the atherogenic role of lipoproteins and oxidized low-density lipoprotein in pre-eclampsia

Yasser H. Shaaban, Fatma Aletebi, Manal Helal and Walid Saber

Evidence Based Women’s Health Journal

Vol 2 No 4 November 2012

Editorial board

Editor-in-Chief

Hesham Al-Inany, EgyptAhmed Rezk, Egypt

Deputy Editor

Ahmed Abou-Setta, Canada

Editors

Karim Abd-El-Maeboud, EgyptKonstantin Boyarsky, RussiaCarlo Alviggi, ItalySaad Amer, UKMahnaz Ashrafi, IranNazar Amso, UKHamdy Azab, KuwaitOsama Azmy, EgyptHassan Boudrar, MoroccoYacoub Khalaf, UKWael Madkour, UEAMohamed Bedaiwy, USAEbtihag Shafik, EgyptMostafa Metwally, UK

Nardo Luciano, UKIoannis E. Messinis, GreecePanagiotis Peitsidis, GreeceNikolaos Polyzos, BelgiumHazem Al-Rumaih, KuwaitSandro Gerli, ItalyAhmed Sekotory Ahmed, UKOsama Shaeer, EgyptAhmed Shahin, M.D, EgyptKhaldoun Sharif, JordanAiman Smadi, JordanRobabeh Taheripanah, IranHans Verstraelen, BelgiumMohamed Shereen, Egypt

Trainee Editors

Amr Wahba, EgyptFady Kamel, UKMona Moghazy, UK

Managing Editor:

Tamer Abdelmoniem, Kuwait

Evidence Based Women’s Health Journal

Vol 2 No 4 November 2012

Information for authors

Policy and ethicsThe work described in your article must have been carried out in accordance with The Code of Ethics of the World Medical

Association (Declaration of Helsinki) for experiments involving humans http://www.wma.net/en/30publications/10policies/b3/

index.html; the EU Directive 2010/63/EU for animal experiments http://ec.europa.eu/environment/chemicals/lab_animals/

legislation_en.htm; Uniform Requirements for manuscripts submitted to Biomedical journals http://www.icmje.org. This must be

stated at an appropriate point in the article.

Layout of the manuscript:The manuscript should be submitted in a Microsoft Word, RTF, or Wordperfect document file format.

Page layout: the text should be on A4 size paper and the margins of the paper should be as follows: 4 cm for the upper and left

margins and 2 cm for the lower and right margins.

Line spacing and font style: The text should be double-spaced and an 11 point font should be used. Italics and bolds should also be

used rather than underlining, except for URL links.

Manuscript language: The manuscript should be written in English.

Structure of the manuscript: The manuscript should be composed of the following: a Title page, an Abstract and the Main

Manuscript.

Title Page: This should contain the following: the title of the manuscript, a list of all authors with their suffixes and affiliations, the

contact details for the nominated corresponding author including email address, telephone number and mailing address.

Abstract: This should be a separate page and should follow these subheadings:

Objectives: Describe the aim of the study or investigation.

Study design: Describe the type of study and its design.

Patients and Methods: Describe the setting of the study and explain how it was conducted.

Results: Describe the outcome of the study and explain its statistical significance.

Conclusion: Provide a summary of the main findings of the study.

Keywords: Please include a maximum of five keywords related to the manuscript topic.

Main manuscript: This should be structured around these subheadings: Introduction, Patients and Methods, Results, Discussion,

Conclusion, Acknowledgements and References.

Introduction: Please keep this brief.

Patients and Methods: A detailed description of the setting of the study and the observational and analytical methods used.

Acknowledgements: The acknowledgements section should contain two distinct statements:

1. Assistance with the study. Acknowledgements should be made only to those who have made a substantial contribution to the

study. Authors are responsible for obtaining written permission from people acknowledged by name in case readers infer their

endorsement of data and conclusions.

2. Conflict of interest and sources of funding. You must make reference to all relevant conflicts of interest and sources of funding

under a separate sub-heading. If there are no conflicts of interest or sources of funding please state: There are no conflicts of interest.

References: References should be numbered consecutively in the order in which they first appear in the text. They should be

assigned Arabic numerals, which should be given in brackets, e.g. [17]. References should include the names of all authors when six

or fewer; when seven or more, list only the first six names and add et al. References should also include full title and source

information. Journal names should be abbreviated as in MEDLINE (NLM Catalog, http://www.ncbi.nlm.nih.gov/nlmcatalog).

E.g.:

[1] Sackett DL, Haynes RB, Guyatt GH and Tugwell P. Clinical Epidemiology. A basics science for clinical medicine, 2nd edition.

Boston: Little Brown and Company; 1991.

[2] Gdoura R, Keskes-Ammar L, Bouzid F, Eb F, Hammami A, et al. Chlamydia trachomatis and male infertility in Tunisia. Eur

J Contracept Reprod Health Care. 2001 Jun; 6(2):102-7.

Evidence Based Women’s Health Journal

Vol 2 No 4 November 2012

Tables and figures:Each table should be typed on a separate sheet in double spacing. Tables should not be submitted as photographs. Each table

should be assigned an Arabic numeral, e.g. (Table 3) and a brief title. Vertical rules should not be used. Place explanatory matter in

footnotes, not in the heading. Explain in footnotes all non-standard abbreviations that are used in each table. Identify statistical

measures of variations, such as standard deviation and standard error of the mean. Be sure that each table is cited in the text. If

you use data from another published or unpublished source, obtain permission and acknowledge the source fully. References to

figures should be made in order of appearance in the text and should be in Arabic numerals in parentheses, e.g. (Fig. 2). Most file

formats are accepted, but TIFF and EPS files, with fonts embedded, are preferred. If scanned, line art should be at a resolution of

800 dpi, and halftones and colour at 300 dpi. All colour values should be CMYK. If hard copies are submitted they should have a

label pasted to the back bearing the figure number, the title of the paper, the author’s name and a mark indicating the top of the

figure. If photographs of people are used, their identities must be obscured or the picture must be accompanied by written consent

to use the photograph. If a figure has been published before, the original source must be acknowledged and written permission

from the copyright holder for both print and electronic formats should be submitted with the material. Permission is required

regardless of authorship or publisher, except for documents in the public domain. Figures may be reduced, cropped or deleted at

the discretion of the editor.

Word limits:The title should contain a maximum of 50 words and the running title a maximum of 25 words. The abstract should not exceed 250

words and the main manuscript without including tables, references and the acknowledgements should not exceed 3000 words.

Reporting clinical trialsAll randomised controlled trials submitted for publication should include a completed Consolidated Standards of Reporting Trials

(CONSORT) flow chart. Please refer to the CONSORT statement website at http://www.consort-statement.org for more

information.

All correspondence should be sent electronically to the Editor:Dr. Hesham Al-Inany

Mobile: + 0111 222 0298

E-mail: kaainih@yahoo.com

Focus and Scope

Evidence Based Women’s Health Society Journal publishes peer reviewed papers reporting original research with high quality

evidence including narrative reviews, systematic reviews, randomized controlled trials as well as cross sectional studies. The journal

will only consider original papers, contributed solely to the Evidence Based Women’s Health Society Journal. These articles aim to

provide up-to-date high quality evidence on all topics pertinent to women’s health including reproductive physiology and

pathology, endocrinology, andrology, gametogenesis, fertilization, embryo development, implantation, infertility treatment,

pregnancy, genetics, genetic diagnosis, contraception, infectious disease, oncology, surgery, ethics and psycho-social issues. The

highest scientific and editorial standard is maintained throughout the journal along with a rapid rate of publication.

Evidence Based Women’s Health Journal

Vol 2 No 4 November 2012

Does the addition of LH activity to FSH make gonadotrophins

more superior? A systematic review and meta-analysisHesham Al-Inanya,b, Amr Wahbaa,b, Hatem Abu Hashima,b, Human Fatemia,b

and Ahmed Abousettaa,b

aDepartment of Obstetrics & Gynecology, Facultyof Medicine, Cairo University, Cairo and bDepartmentof Obstetrics & Gynecology, Faculty of Medicine,Mansoura University, Mansoura, Egypt

Correspondence to Hesham Al-Inany, MD,Abdelhadi St, Manial, 11451 Cairo, EgyptTel: + 011 1222 0298;e-mail: kaainih@yahoo.com

Received 23 April 2012Accepted 12 June 2012

Evidence Based Women’s Health Journal

2012, 2:113–120

Background

The contribution of luteinizing hormone (LH) activity to follicle-stimulating hormone

(FSH) in ovarian stimulation in improving the outcome of IVF/intracytoplasmic sperm

injection (ICSI) has been an area of major debate.

Objective

To systematically locate, review, and analyze the best available evidence on the value of

additional LH activity in ovarian stimulation in IVF and/or ICSI irrespective of the source

of LH.

Design

Systematic review of properly randomized trials comparing FSH only [as recombinant

FSH (recFSH)] vs. LH-containing protocols (derived either from a urinary source or

developed by recombinant technology) in women undergoing IVF and/or ICSI with

desensitization achieved either by long GnRH agonist or GnRH antagonist protocols.

A meticulous search was carried out using electronic databases and hand searches

of the literature.

Results

Thirty-one trials were identified. Only nine trials reported on live birth rate and ovarian

hyperstimulation (OHSS) rates. Pooling of the trials showed that the live birth rate did

not show a statistically significant difference [recFSH (304/1120; 27.14%) vs.

FSH/LH (324/1110; 29.19%) (P = 0.29; odds ratio (OR) = 0.90, 95% confidence

interval (CI) = 0.75–1.09)]. OHSS rates also did not show a significant difference

[recFSH (34/1888; 1.80%) vs. FSH/LH (29/1843; 1.57%) (P = 0.79; OR = 1.08,

95% CI = 0.63–1.83)]. The clinical pregnancy rate significantly favored additional LH

activity [(recFSH (748/2758; 27.1%) vs. FSH/LH (838/2772; 30.2%), P = 0.008;

OR = 0.85 95% CI = 0.76–0.96)], whereas FSH-only protocols (recFSH) yielded a

higher number of retrieved oocytes compared with LH-containing protocols (FSH/LH)

(P = 0.002; mean difference = 1.25, 95% CI = 0.48–2.02).

Conclusion

LH activity is not paramount in ovarian stimulation.

Keywords:

follicle-stimulating hormone, intracytoplasmic sperm injection, in-vitro fertilization,

luteinizing hormone, meta-analysis, ovulation stimulation, systematic review

Evid Based Women Health J 2:113–120& 2012 Evidence Based Women’s Health Journal2090-7265

IntroductionGonadotrophins differ in terms of the source from which

they are extracted (either from urine or from recombinant

DNA technology) and in their formulation [either

containing both follicle-stimulating hormone (FSH) and

luteinizing hormone (LH) or either FSH or LH only].

The composition as well as the source of gonadotrophins

could have an impact on the outcome of IVF. Human

menopausal gonadotropins (hMG) have long been the

classic and principal urinary gonadotrophins used for

ovarian stimulation since the development of ART

technology in the 1970s. It is composed of equal amounts

of FSH and LH (75 U for each per ampoule). However,

since the 1980s, a variety of generations of urinary hMG

have been developed with the intention of eliminating

most or all of the LH content [1]. During the mid-1990s,

recombinant FSH (recFSH) was produced from hamster

ovarian cell cultures and this step was considered

a landmark in the production of gonadotrophins [2].

With the development of such gonadotrophins devoid of

LH activity (namely recFSH and HP-FSH), questions

about the role of adding LH activity to FSH in controlled

ovarian hyperstimulation (COH) in IVF and/or intracyto-

plasmic sperm injection (ICSI) on treatment outcome have

been raised and have been a topic for intense research.

Previous systematic reviews have primarily focused on

comparing the effectiveness of different sources of

gonadotropins. Urinary hMG was compared with recFSH

in two recent meta-analyses, both of which showed

Original article 113

2090-7265 & 2012 Evidence Based Women’s Health Journal DOI: 10.1097/01.EBX.0000419241.48874.fe

a better outcome in terms of live birth rate when hMG

was used for ovarian stimulation compared with recFSH

in the GnRH agonist (GnRHa) long protocol [3,4]. Two

other meta-analyses compared urinary HP-FSH versus

(recFSH) and showed similar pregnancy and live birth

rates, with no difference in the ovarian hyperstimulation

(OHSS) rate [5,6].

However, there is still a need for a better understanding

of the differential effects of different gonadotrophin

preparations and of the impact of the administration of

LH activity during COH.

To extract the evidence on the impact of COH with LH-

containing gonadotrophins on live birth rate and OHSS

rate, the current meta-analysis and systematic review was

carried out to compare FSH-only protocols (recFSH)

with LH-containing protocols irrespective of the source

of gonadotrophins (either hMG or recFSH + recLH) and

irrespective of the protocol of desensitization (either the

GnRHa long protocol or the GnRH antagonist protocol).

Materials and methodsCriteria for considering studies for this review

All published manuscripts and abstracts describing

randomized trials and reporting data that compared

outcomes for women undergoing IVF/ICSI and rando-

mized to undergo ovarian stimulation with FSH-only

protocols (recFSH) or LH-containing protocols (either

hMG or recFSH + recLH) were sought in all languages.

Search strategy for the identification of studies

Meticulous computerized searches (last performedy)

were carried out using MEDLINE (1978 to present),

EMBASE (1980 to present), the Cochrane Central

Register of Controlled Trials (CENTRAL), and the Trial

Register of Controlled Trials (e.g. http://www.controlled-trials.com). Furthermore, the reference lists of all known

primary studies, review articles, citation lists of relevant

publications, abstracts of major scientific meetings (e.g.

ESHRE and ASRM), and included studies were exam-

ined to identify additional relevant citations. Finally,

ongoing and unpublished trials were sought by contacting

experts in the field and commercial entities.

Methods of the review

A standardized data extraction form was developed and

piloted for consistency and completeness. Trials were

considered for inclusion and trial data were extracted.

Data management and statistical analyses were carried

out using the ‘Review Manager 4.3’ (The Cochrane

Collaboration). Individual outcome data were included in

the analysis if they fulfilled the prestated criteria. Where

possible, data were extracted to allow for an intention-to-

treat analysis, defined as including all randomized cycles

in the denominator. If data from the trial reports were

insufficient or missing, the investigators of individual

trials were contacted by e-mail for additional information

to carry out analysis on an intention-to-treat basis.

For the meta-analysis, the number of participants

experiencing the event in each group of the trial was

determined by visual inspection of the outcome tables

and using the w2-test for heterogeneity. In addition, the

I2-test was used to attempt quantification of any apparent

inconsistency. The I2-test is a statistical measure used to

quantify heterogeneity. It describes the percentage of the

variability in effect estimates that is because of hetero-

geneity rather than sampling error (chance). An I2-value

greater than 50% may be considered to represent

substantial heterogeneity.

Types of outcomes measures

The primary outcomes for this systematic review were

live birth rate and OHSS rate. The secondary outcomes

were clinical and ongoing pregnancy rates. In addition,

cycle characteristics (number of oocytes retrieved,

treatment duration, amount of FSH, estradiol on day of

hCG, progesterone on day of hCG, rate of poor

responders) were also evaluated. The trials included

were analyzed for the following characteristics: method of

randomization, presence or absence of blinding to

treatment allocation, quality of allocation concealment,

number of patients randomized, whether an intention-to-

treat analysis was carried out, and the duration, timing,

and location of the study.

Comparison of results

For the meta-analysis, the number of participants

experiencing the event was recorded. Data were ex-

tracted to allow for an intention-to-treat analysis, defined

as including in the denominator all randomized cycles.

Meta-analysis of dichotomous data was carried out using

the Mantel–Haenszel method using a random-effects

model, and the odds ratio (OR) and 95% confidence

interval (CI) were evaluated. Meta-analysis of continuous

data was carried out using the inverse variance method

utilizing a random-effects model. The mean difference

(MD) and 95% CI were evaluated. P values were

presented for further confirmation of the results. The

standardized mean difference was used when multiple

scales were provided (E2). A P value of less than 0.05 was

considered to be significant. The results were pooled

using a fixed-effects model only after confirming that

marked statistical heterogeneity was not present (i.e. the

observed treatment effects in individual trials were not

statistically significantly different from the overall pooled

estimate of the treatment effect).

Participants

Randomized women undergoing IVF/ICSI and COH with

either FSH-only protocols (recFSH) or LH-containing

protocols (either hMG or recFSH + recLH). We ex-

cluded all nonrandomized trials, trials using LH priming

(LH only then FSH only), and trials using HP-FSH

instead of recFSH (as the HP-FSH contains some LH

activity).

114 Evidence Based Women’s Health Journal

ResultsThirty-one randomized-controlled trials (RCTs) were

identified. Twenty-one trials compared the use of recFSH

vs. hMG/HP-hMG whereas 10 RCTs compared recFSH

vs. recFSH + recLH. The method of fertilization used

was either IVF or ICSI. These included trials enrolled

4571 participants, which yielded an adequate power for

meta-analysis.

Nine trials only reported on live birth rate and OHSS

rate. Data were extracted to allow for an intention-to-

treat analysis. Pooling of the included trials showed that

live birth rate was not statistically different between both

groups, although it showed trends toward improvement

with LH-containing protocols [recFSH (304/1120;

27.14%) vs. FSH/LH (324/1110; 29.19%) (P = 0.29;

OR = 0.90, 95% CI = 0.75–1.09)] (Fig. 1).

The rate of OHSS was also not statistically different

between FSH-only protocols and LH-containing proto-

cols [recFSH (34/1888; 1.80%) vs. FSH/LH (29/1843;

1.57%) (P = 0.79; OR = 1.08, 95% CI = 0.63–1.83)]

(Fig. 2).

Eighteen trials assessed the ongoing pregnancy rate,

which showed no significant difference between both

groups [(recFSH (544/2385; 22.81%) vs. FSH/LH (589/

2363; 24.93%) (P = 0.31; OR = 0.93, 95% CI =

0.81–1.07)] (Fig. 3).

The clinical pregnancy rate (CPR) significantly favored

additional LH activity [(recFSH (748/2758; 27.1%) vs.

FSH/LH (838/2772; 30.2%) (P = 0.008; OR = 0.85, 95%

CI = 0.76–0.96)], whereas the number of oocytes was

higher with FSH-only protocols (recFSH vs. FSH/LH,

Figure 1

Live birth rate.

Figure 2

Ovarian hyperstimulation rate.

Questioning the value of LH activity Al-Inany et al. 115

P = 0.002; MD = 1.25; 95% CI = 0.48–2.02). Subgroup

analysis of CPR showed a significant difference in favor of

hMG, but not recFSH + recLH. Therefore, LH activity

per se could not be the underlying mechanism accounting

for this significant difference (Figs 4 and 5).

DiscussionLH is a glycoprotein that plays a crucial role in

folliculogenesis during the natural ovarian cycles. It

exerts its activity in theca cells, inducing androgen

production, the substrate of estradiol, as well as in

granulosa cells, where it regulates, together with FSH,

the local production of various molecules (inhibin B and

growth factors) that are important in promoting follicular

maturation [7].

Despite this clear physiological role of LH in folliculo-

genesis, the role of exogenous LH activity in ovarian

stimulation has been the subject of intense debate [8,9].

Premature luteinization and spontaneous ovulation re-

sulting from premature LH surges presented significant

challenges to early ovarian stimulation protocols until the

introduction of GnRHa, which significantly reduced the

incidence of LH surges and increased the pregnancy rates

by overcoming this unwanted phenomenon [10].

That early, researchers did not pay much attention to the

low levels of LH resulting from this desensitization, as at

that time, ovarian stimulation was carried out using

gonadotrophins with LH activity. Nonetheless, with the

introduction of gonadotrophins devoid of LH activity in

the 1990s [11–13], the interest in the role of LH during

ovarian stimulation started to grow.

The perplexity created by the contradictory data on the

essential physiological role of LH and its debatable

therapeutic role has further motivated researchers to

assess the importance of LH activity during follicular

development and its potential value during ovarian

stimulation for IVF by comparing FSH-only protocols

with LH-containing protocols in RCTs.

For FSH-only protocols, recFSH is the purest form of

FSH that contains no LH activity, whereas LH activity

can be provided as hMG (from urinary source) or as

recombinant LH (recLH).

Comparing recFSH versus hMG (urinary source LH)

Earliest meta-analysis that compared recFSH and hMG

treatment in normogonadotrophic patients undergoing a

GnRHa long protocol was carried out by Van Wely et al. [14]

that included four RCTs. The authors failed to show any

statistically significant difference in the main outcome

measures (ongoing pregnancy rates and live birth rate).

As more RCTs addressing the same question were carried

out, systematic reviews and meta-analyses were subse-

quently published summarizing the extracted data [3,4].

The study carried out by Coomarasamy et al. [3] included

seven RCTs [15–20](European and Israeli Study Group

on highly purified menotropin vs. recombinant FSH,

2002) with a total of 2159 patients. In that meta-analysis,

live birth rates were significantly increased by 4% in

patients treated with hMG compared with those treated

with recFSH (95% CI: 0.4–7.5).

A subsequent systematic review and meta-analysis by

Al-Inany and colleagues included two more studies [21,22]

Figure 3

Ongoing pregnancy rate.

116 Evidence Based Women’s Health Journal

and a further 689 patients and confirmed the above

finding [4] as live birth rates were found to be significantly

increased by 3.4% (95% CI: 0.3–6.5) in women randomized

to receive hMG compared with those randomized to receive

recFSH after downregulation with a long GnRHa protocol.

The evaluation of the secondary outcome measures

including the occurrence of OHSS and of multiple

pregnancies did not show significant differences in both

meta-analyses [3,4]. However, in the meta-analysis by Al-

Inany and colleagues, the duration of ovarian stimulation

was significantly shorter in women receiving hMG

compared with those who received recFSH (weighted

MD: – 1.21 days, 95% CI: – 1.35 to – 1.06) after

downregulation with a long GnRHa protocol.

Whether the GnRH antagonist protocol could yield similar

or different results compared with the GnRH agonist long

protocol was addressed in an RCT by Bosch et al. [23], who

evaluated the IVF outcome in 280 patients stimulated with

hMG (n = 140) vs. recFSH (n = 140). In this study, the

number of oocytes retrieved was significantly reduced in

patients treated with hMG compared with those treated

with recFSH [(mean ± SD) 11.3 ± 6.0 vs. 14.4 ± 8.1,

respectively; difference: – 3.1 cumulus–oocyte complexes,

95% CI: – 4.9 to – 1.3]. However, the delivery rates were

not significantly different between patients randomized

to receive hMG vs. those randomized to receive recFSH

(34.3 vs. 31.4%, respectively; risk difference: + 2.9, 95%

CI: –8.1 to + 13.7) [23].

recFSH versus recFSH + recLH

Several researchers have investigated the impact of recLH

addition during ovarian stimulation on the outcome of

IVF/ICSI. A recent meta-analysis has summarized the data

Figure 4

Clinical pregnancy rate.

Questioning the value of LH activity Al-Inany et al. 117

extracted from seven RCTs [24–31] including a total of 701

patients to determine whether the addition of recLH

increases the live birth rate in IVF patients treated with

FSH and GnRHa [32]. No statistically significant differ-

ences in the live birth rates were observed between

patients who received recLH and those who did not when

the seven eligible RCT were pooled (95% CI: – 7.7 to

+ 4.85, P = 0.65; heterogeneity P = 0.32, fixed-effects

model). The result did not change in all subgroup analyses

carried out on the basis of the type of GnRH analogue used

for LH surge inhibition, the time or dose of recLH

addition, or patient’s age.

Consistent with the above findings, the results of a large

RCT carried out by Nyboeandersen and colleagues in

which 526 women were randomized to be stimulated with

either recFSH alone (n = 261) or recFSH with the

addition of recLH (n = 265) from day 6 of stimulation

onwards showed that the ongoing pregnancy rates at

week 10–12 were not significantly different between the

groups (28.7 vs. 27.2%, respectively; RD: + 1.5%, 95%

CI: – 6.1 to + 9.2).

recFSH versus LH-containing protocols (either hMG or

recFSH + rec LH)

In our current meta-analysis, we aimed to evaluate the role

of exogenous LH in ovarian stimulation irrespective of its

source, that is, whether extracted from urine (in the form

of hMG) or manufactured by recombinant DNA technology

(recLH), and also irrespective of the mode of desensitiza-

tion, that is, whether achieved by the long GnRHa protocol

or by the rapidly acting GnRH antagonist protocol.

The results of this systematic review and meta-analysis

failed to show any significant difference in the major

outcome measures (live birth rate and OHSS) when

comparing COH using FSH-only protocols with LH-

containing protocols, suggesting that LH is not of

paramount importance in ovarian stimulation. Moreover,

LH-containing protocols showed a detrimental effect on

the number of oocytes retrieved, which was higher with

FSH-only protocols than with LH-containing protocols.

Although in the current review the CPR favored

additional LH activity significantly, this can be ques-

tioned because the subgroup analysis of CPRs showed a

significant difference in favor of hMG, but not in favor of

recFSH + recLH. Therefore, LH activity per se may not

be the underlying mechanism accounting for this

significant difference.

In this respect, it is worth noting that hMG differs

fromrecFSH not only in the LH activity content but also in

the nature of the FSH contained. Indeed, the human-

derived FSH contained in hMG carries a moreextensive

glycosylation and is closer in its isoformdistribution to the

pituitary FSH. In contrast, recFSH carries a rodent-type

glycosylation that is simpler in structure and that includes a

lower number of sialic acid residues. The glycosylation and

sialilation status of FSH affects its half-life [33] and a

variety of FSH biological properties [34]. It is therefore

possible that some of the differences found between hMG

and recFSH are actually because of the different nature of

the FSH component. Our findings that demonstrated

better pregnancy rates in hMG protocols and not in

recFSH + recLH protocols further endorse this interpreta-

tion. Accordingly, possible differences in the FSH compo-

nent should be taken into due account in future works

investigating the value of the LH supplementation.

FSH isoforms: a new theory for the endocrine aspect

of follicular and oocyte growth

The clinical efficacy of commercially available gonadotro-

phin preparations has been the subject of an intense debate

primarily focused on the origin of FSH activity (urine vs.

recombinant derived) and whether the preparation in-

cluded LH like activity. FSH isoform composition has

received little or no attention, and is usually considered to

Figure 5

Number of oocytes retrieved.

118 Evidence Based Women’s Health Journal

have a negligible effect on clinical effectiveness. However,

in a study carried out by Andersen et al. [35], it was

reported that the FSH isoform profile of commercial

gonadotrophin preparations is of clinical importance and

should be taken into account when evaluating endocrine

aspects and efficacy for each preparation.

FSH exists as a family of isohormones with distinct

oligosaccharide structures, and the FSH isoform mixtures

released change during the follicular phase of the

menstrual cycle. The different isoforms exert a number

of different and divergent biological effects. Exposure of

cumulus–oocyte complexes to less-acidic FSH isoforms in

a pulse-like manner results in a rapid pattern of cAMP

accumulation exceeding that seen with acidic isoforms.

It appears that pulsatile and intermittent release of less-

acidic/short-living FSH isoforms is sufficient to induce

biological responses while allowing the granulosa cell

FSH receptors to regain responsiveness to further FSH

stimulation. This may explain the elevation of progester-

one in recombinant forms (less-acidic isoforms) as these

isoforms are more potent than their counterparts of

urinary origin. Future studies should evaluate the

difference in the premature increase in progesterone in

recFSH and urinary FSH cycles [36].

We can firmly conclude from the current available

evidence at hand that the addition of LH activity to

FSH in ovarian stimulation irrespective of its source, that

is, whether urinary or recombinant, and irrespective of

the pituitary desensitization protocol, that is, whether a

GnRHa long protocol or a GnRH antagonist, did not show

any superiority over FSH-only protocols in terms of the

live birth rate, OHSS rate, and ongoing pregnancy rates.

However, additional LH activity provided as hMG

showed higher CPRs, although it did not when provided

as recLH, indicating that factors other than LH activity

per se may be responsible for this significant difference.

FSH-only protocols yielded more number of retrieved

oocytes than LH-containing protocols.

AcknowledgementsConflicts of interestThere are no conflicts of interest.

References1 Zafeiriou S, Loutradis D, Michalas S. The role of gonadotropins in follicular

development and their use in ovulation induction protocols for assisted re-production. Eur J Contracept Reprod Health Care 2000; 5:157–167.

2 Out HJ, Driessen SGAJ, Mannaerts BMJL, Bennink HJTC. Recombinantfollicle-stimulating hormone (follitropin beta, Puregon) yields higher preg-nancy rates in in vitro fertilization than urinary gonadotropins. Fertil Steril1997; 68:138–142.

3 Coomarasamy A, Afnan M, Cheema D, Van Der Veen F, Bossuyt PMM, VanWely M. Urinary hMG versus recombinant FSH for controlled ovarian hy-perstimulation following an agonist long down-regulation protocol in IVF orICSI treatment: a systematic review and meta-analysis. Hum Reprod 2008;23:310–315.

4 Al-Inany HG, Abou-Setta AM, Aboulghar MA, Mansour RT, Serour GI. Effi-cacy and safety of human menopausal gonadotrophins versus recombinantFSH: a meta-analysis. Reprod BioMed Online 2008; 16:81–88.

5 Moustafa M, Abdelwahed A, Abosekena I, Abdelazim S, Abou-Setta A, Al-Inany H. IVF outcomes with either highly purified FSH vs recombinant FSH indown-regulated normogonadotrophic women: a prospective comparative

study in a developing country and meta-analysis. Open Women’s Health J2009; 3:11–15.

6 Al-Inany HG, Abou-Setta AM. Are all human-derived follicle-stimulating hor-mone products the same? A systematic review and meta-analysis using di-rect and adjusted indirect analyses, to determine whether Fostimon is moreefficient than Metrodin-HP. Gynecol Endocrinol 2012; 28:94–101.

7 Huang ZH, Clayton PE, Brady G, Morris ID. Insulin-like growth factor-I geneexpression in human granulosa-lutein cells. J Mol Endocrinol 1994; 12:283–291.

8 Levy DP, Navarro JM, Schattman GL, Davis OK, Rosenwaks Z. The role ofLH in ovarian stimulation: exogenous LH: let’s design the future. Hum Reprod2000; 15:2258–2265.

9 Kolibianakis EM, Zikopoulos K, Schiettecatte J, Smitz J, Tournaye H, CamusM, et al. Profound LH suppression after GnRH antagonist administration isassociated with a significantly higher ongoing pregnancy rate in IVF. HumReprod 2004; 19:2490–2496.

10 Hughes EG, Fedorkow DM, Daya S, Sagle MA, Van de Koppel P, Collins JA.The routine use of gonadotropin-releasing hormone agonists prior to in vitrofertilization and gamete intrafallopian transfer: a meta-analysis of randomizedcontrolled trials. Fertil Steril 1992; 58:888–896.

11 Devroey P, van Steirteghem A, Mannaerts B, Bennink HC. Successful in-vitrofertilisation and embryo transfer after treatment with recombinanthuman FSH. Lancet 1992; 339:1170.

12 Porchet HC, Le Cotonnec J-Y, Canali S, Zanolo G. Pharmacokinetics ofrecombinant human follicle stimulating hormone after intravenous, in-tramuscular, and subcutaneous administration in monkeys, and comparisonwith intravenous administration of urinary follicle stimulating hormone. DrugMetab Dispos 1993; 21:144–150.

13 Le Cotonnec J-Y, Porchet HC, Beltrami V, Khan A, Toon S, Rowland M.Clinical pharmacology of recombinant human follicle-stimulating hormone(FSH). I. Comparative pharmacokinetics with urinary human FSH. Fertil Steril1994; 61:669–678.

14 Van Wely M, Kwan I, Burt AL, Thomas J, Vail A, Van der Veen F, Al-Inany HG.Recombinant versus urinary gonadotrophin for ovarian stimulation in assistedreproductive technology cycles. Cochrane Database Syst Rev 2011;2:CD005354.

15 Gordon UD, Harrison RF, Fawzy M, Hennelly B, Gordon AC. A randomizedprospective assessor-blind evaluation of luteinizing hormone dosage andin vitro fertilization outcome. Fertil Steril 2001; 75:324–331.

16 Ng EHY, Lau EYL, Yeung WSB, Ho PC. HMG is as good as recombinanthuman FSH in terms of oocyte and embryo quality: a prospectiverandomized trial. Hum Reprod 2001; 16:319–325.

17 Westergaard LG, Erb K, Laursen SB, Rex S, Rasmussen PE. Human me-nopausal gonadotropin versus recombinant follicle-stimulating hormone innormogonadotropic women down-regulated with a gonadotropin-releasinghormone agonist who were undergoing in vitro fertilization and in-tracytoplasmic sperm injection: a prospective randomized study. Fertil Steril2001; 76:543–549.

18 Balasch J/, Penarrubia J, Fabregues F, Vidal E, Casamitjana R, Manau D,et al. Ovarian responses to recombinant FSH or HMG in normogonado-trophic women following pituitary desensitization by a depot GnRH agonistfor assisted reproduction. Reprod BioMed Online 2003; 7:35–42.

19 Kilani Z, Dakkak A, Ghunaim S, Cognigni GE, Tabarelli C, Parmegiani L,Filicori M. A prospective, randomized, controlled trial comparing highly pur-ified hMG with recombinant FSH in women undergoing ICSI: ovarian re-sponse and clinical outcomes. Hum Reprod 2003; 18:1194–1199.

20 Andersen AN, Devroey P, Arce J-C. Clinical outcome following stimulationwith highly purified hMG or recombinant FSH in patients undergoing IVF:a randomized assessor-blind controlled trial. Hum Reprod 2006; 21:3217–3227.

21 Rashidi BH, Sarvi F, Tehrani ES, Zayeri F, Movahedin M, Khanafshar N. Theeffect of HMG and recombinant human FSH on oocyte quality: a randomizedsingle-blind clinical trial. Eur J Obstet Gynecol Reprod Biol 2005; 120:190–194.

22 Hompes PGA, Broekmans FJ, Hoozemans DA, Schats R. Effectiveness ofhighly purified human menopausal gonadotropin vs. recombinant follicle-stimulating hormone in first-cycle in vitro fertilization-intracytoplasmic sperminjection patients. Fertil Steril 2008; 89:1685–1693.

23 Bosch E, Vidal C, Labarta E, Simon C, Remohi J, Pellicer A. Highly purifiedhMG versus recombinant FSH in ovarian hyperstimulation with GnRH an-tagonists – a randomized study. Hum Reprod 2008; 23:2346–2351.

24 Sills ES, Levy DP, Moomjy M, McGee M, Rosenwaks Z. A prospective,randomized comparison of ovulation induction using highly purified follicle-stimulating hormone alone and with recombinant human luteinizing hormonein in-vitro fertilization. Hum Reprod 1999; 14:2230–2235.

25 Balasch J, Creus M, Fabregues F, Civico S, Carmona F, Puerto B, et al. Theeffect of exogenous luteinizing hormone (LH) on oocyte viability: evidencefrom a comparative study using recombinant human follicle-stimulating hor-mone (FSH) alone or in combination with recombinant LH for ovarian sti-mulation in pituitary-suppressed women undergoing assisted reproduction.J Assist Reprod Genet 2001; 18:250–256.

26 Balasch J, Vidal E, Penarrubia J, Casamitjana R, Carmona F, Creus M, et al.Suppression of LH during ovarian stimulation: analysing threshold values andeffects on ovarian response and the outcome of assisted reproduction in

Questioning the value of LH activity Al-Inany et al. 119

down-regulated women stimulated with recombinant FSH. Hum Reprod2001; 16:1636–1643.

27 Humaidan P, Bungum M, Bungum L, Andersen CY. Effects of recombinantLH supplementation in women undergoing assisted reproduction with GnRHagonist down-regulation and stimulation with recombinant FSH: anopening study. Reprod BioMed Online 2004; 8:635–643.

28 Sauer MV, Thornton MH II, Schoolcraft W, Frishman GN. Comparative effi-cacy and safety of cetrorelix with or without mid-cycle recombinant LH andleuprolide acetate for inhibition of premature LH surges in assisted re-production. Reprod BioMed Online 2004; 9:487–493.

29 Griesinger G/, Schultze-Mosgau A, Dafopoulos K, Schroeder A, Schroer A,von Otte S, et al. Recombinant luteinizing hormone supplementation to re-combinant follicle-stimulating hormone induced ovarian hyperstimulation in theGnRH-antagonist multiple-dose protocol. Hum Reprod 2005; 20:1200–1206.

30 Fabregues F, Creus M, Penarrubia J, Manau D, Vanrell JA, Balasch J. Effectsof recombinant human luteinizing hormone supplementation on ovarian sti-mulation and the implantation rate in down-regulated women of advancedreproductive age. Fertil Steril 2006; 85:925–931.

31 Tarlatzis B, Tavmergen E, Szamatowicz M, Barash A, Amit A, Levitas E,Shoham Z. The use of recombinant human LH (lutropin alfa) in the late

stimulation phase of assisted reproduction cycles: a double-blind, rando-mized, prospective study. Hum Reprod 2006; 21:90–94.

32 Kolibianakis EM, Kalogeropoulou L, Griesinger G, Papanikolaou EG, Papa-dimas J, Bontis J, Tarlatzis BC. Among patients treated with FSH and GnRHanalogues for in vitro fertilization, is the addition of recombinant LH asso-ciated with the probability of live birth? A systematic review and meta-ana-lysis. Hum Reprod Update 2007; 13:445–452.

33 Wide L, Eriksson K, Sluss PM, Hall JE. Serum half-life of pituitary gonado-tropins is decreased by sulfonation and increased by sialylation in women. JClin Endocrinol Metab 2009; 94:958–964.

34 Nayudu PL, Vitt UA, Barrios De Tomasi J, Pancharatna K, Ulloa-Aguirre A.Intact follicle culture: what it can tell us about the roles of FSH glyco-forms during follicle development. Reprod BioMed Online 2002; 5:240–253.

35 Andersen CY, Westergaard LG, van Wely M. FSH isoform composition ofcommercial gonadotrophin preparations: a neglected aspect? ReprodBiomed Online 2004; 9:231–236.

36 Arey BJ, Lopez FJ. Are circulating gonadotropin isoforms naturally occurringbiased agonists? Basic and therapeutic implications. Rev Endocr MetabDisord 2011; 12:275–288.

120 Evidence Based Women’s Health Journal

A comparative study between cabergoline, coasting, and

step-down regimens in the prevention of severe OHSS

and their correlation with pregnancy rate in intracytoplasmic

sperm injection cyclesHoda Abdelaal, Olfat N. Riade and Mostafa Mahmoud

Department of Obstetrics & Gynecology, Faculty ofMedicine, Cairo University, Cairo, Egypt

Correspondence to Hoda Abdelaal, Department ofObstetrics and Gynecology, Faculty of Medicine,Cairo University, Cairo, EgyptTel: +01 222 848 993;e-mail: hodatolba@gmail.com

Received 20 November 2011Accepted 16 January 2012

Evidence Based Women’s Health Journal

2012, 2:121–125

Objective

To compare coasting, cabergoline, and step-down methods in the prevention of severe

ovarian hyperstimulation syndrome (OHSS) and their effects on various outcome

measures in intracytoplasmic sperm injection cycles.

Materials and methods

Patients at risk of developing OHSS were randomized into three groups: group A,

which included 16 patients who had undergone coasting by withholding human

menopausal gonadotropin while continuing the gonadotropin-releasing hormone

analogue till the E2 decreased below 3000 pg/ml before triggering, group B

(cabergoline group), which included 28 patients who received 0.5 mg/day for

8 days starting from the day of human chorionic gonadotropin, and group C

(step-down group), which included 30 patients who had decreased the dose of

human menopausal gonadotropin by 1 ampoule to decrease E2 below 3000 pg/ml

then triggering by 10 000 IU of human chorionic gonadotropin and oocytes retrieved

after 34–36 h later.

Results

The number of oocytes retrieved, M2, and number of fertilized oocytes were higher in

the cabergoline group than the other two groups (Po0.05), but the clinical pregnancy

rate was higher in the coasting group (50%) compared with the other two groups and

the lowest in the cabergoline group (14.3%), with no statistical significance between

the three groups in the development of severe OHSS (P = 0.41).

Conclusion

Coasting may have a higher pregnancy rate and may represent a higher preventive

method for the development of severe OHSS than the other two methods.

Keywords:

cabergoline, coasting, intracytoplasmic sperm injection, ovarian hyperstimulation

syndrome, step-down

Evid Based Women Health J 2:121–125& 2012 Evidence Based Women’s Health Journal2090-7265

IntroductionOvarian hyperstimulation syndrome (OHSS) is an iatro-

genic complication of ovarian stimulation for assisted

reproduction technology and other infertility treatments;

usually, it develops several days after oocytes are retrieved

or after assisted ovulation. This syndrome is characterized

by ovarian enlargement because of multiple ovarian cysts

and acute fluid shift into the extravascular space including

ascites, hemoconcentration, hypovolemia, and electrolyte

imbalance [1].

Grades of OHSS are mild, including grade I, II, moderate

grade III, and severe OHSS grade IV, V [2]. Its patho-

genesis is related to increased vascular permeability in the

region surrounding the ovaries and their vasculature [3].

b-Human chorionic gonadotropin (HCG) and its analogs

have been implicated in the past. Vasoactive substances

such as interleukins, tumor necrosis factor-a endothelial,

and vascular endothelial growth factor (VEGF) secreted by

the ovaries have been implicated in the increase in vascular

permeability [4].

VEGF not only stimulates new blood vessel development

in the ovary but also induces vascular hyper permeability

by interacting with the VEGF receptor [5]. The rates of

OHSS are as follows: mild, 8–23%, moderate, 1–7%, and

severe, 0.25–5% [6]. Once the syndrome occurs, little can

be done to alter the course of events and only supportive

measures can be carried out; however, a definite and

useful method is cycle cancelation or no HCG adminis-

tration. Many other methods have been used to prevent

the syndrome. The appropriate protocol and gonadotro-

pin needs to be chosen [7]. Unilateral ovarian follicular

aspiration before the administration of HCG was

successful in some investigations [8]. In another study,

aspiration of half of the follicles before the administration

of HCG reduced the risk of severe OHSS [9].

Original article 121

2090-7265 & 2012 Evidence Based Women’s Health Journal DOI: 10.1097/01.EBX.0000419242.56497.0c

Another method for the prevention of OHSS is the

coasting approach withholding gonadotropin administra-

tion, which has been used in ovulation induction cycles to

prevent excessive response [10]. Recently, a dopamine-

agonist, cabergoline, has been used successfully for the

prevention of severe OHSS (anti-VEGF) [11]. The step-

down human menopausal gonadotropin (HMG) protocol

seems rational as it mimics the natural ovarian cycles

when ultrasound monitoring (usually by stimulation day

7) reveals the advancement of a reasonable number of

developing follicles [12]. The impact of this kind of

management of patients with risk factors is that earlier

close monitoring of serum E2 and ultrasound could be

valuable in determining early evolution of OHSS [13].

Approach might be viewed as further development of

conventional ‘coasting’, yet avoiding the risk of abrupt E2

fall with reduction of oocyte retrieval rate and embryo

quality [14].

The aim of this study was to determine the effect of

cabergoline on the cycle outcome and prevent severe OHSS

and to compare it with other popular methods such as

the coasting method and the step-down method (decreasing

the dose of HMG) in preventing severe OHSS and in the

detection of the cycle outcome and pregnancy rate.

Patient and methodsThe study was approved by the ethical committee of

research in Kasr El-Aini Hospital, Cairo University. This

was a prospective cohort study of couples who were

subjected to intracytoplasmic sperm injection cycles in

the in-vitro fertilization (IVF) unit in Kasr El-Aini

Hospital, Cairo University. Seventy-four women at risk

of developing OHSS were included in the study. This

included young women below 35 years of age, PCO

patients, those with E2 more than 4000 pg/ml during the

follow-up of induction, women with a necklace sign

(antral follicle count 420) before induction, and women

with preovulatory follicular number more than 20 follicles

on both sides or a large number of follicles more than

15 mm during the course of induction.

All women were undergoing the long protocol for

stimulation of cycle. They were downregulated with a

gonadotropin-releasing hormone analogue (GnRH-a)

subcutaneously (Decapeptyl 0.1 mg daily from day 20 in

the previous cycle of stimulation). Then, daily adminis-

tration of HMG was carried out intramuscularly along

with the (GnRH-a). E2 was administered for women at

risk for OHSS to determine the level and counsel the

patient to complete the cycle or not. Women who were at

risk of developing OHSS and did not want to cancel their

cycle were divided into three groups.

Group A, the coasting group, included 16 women, for

whom we continued the GnRH-a, but withheld the

gonadotropins until serum E2 started to decrease, with

the mean duration of coasting being 2–3 days. Then HCG

was administered to this group of women when E2 was

below 3000 pg/ml. Group B, the cabergoline group,

included 28 women who received 0.5 mg/day for 8 days

starting from the day of HCG administration. Group C,

the step-down group, included 30 women in whom the

dose of gonadotropins was decreased by 1 ampoule until

the E2 started to decrease. In all three groups, when at

least three follicles of at least 18 mm in size were

achieved, the HCG dose (10 000 IU) was administered

for triggering.

After the administration of HCG, all three groups were

followed up every 48 h for the detection of severe OHSS;

Hb, hematocrit value, and kidney functions were

determined by follow-up ultrasound for all patients.

The embryo transfer procedure was carried out for all

women using the same method and luteal support was

carried out for all women using progesterone in oil

(800 mg into two divided doses) for 14 days. Clinical

pregnancy was detected by the presence of a gestational

sac or cardiac pulsation 3 weeks after ET.

Data were statistically described in terms of the mean

(SD), median and range, or frequencies (number of

cases) and percentages when appropriate. Comparison of

numerical variables between the study groups was carried

out using the Kruskal–Wallis test, with the Mann–

Whitney U-test for independent samples carried out for

post-hoc multiple two-group comparisons. For comparison

of categorical data, the w2-test was carried out. The exact

test was used when the expected frequency was less than

5. P-values less than 0.05 were considered statistically

significant. All statistical calculations were carried out

using computer programs statistical package for the social

science (SPSS Inc., Chicago, Illinois, USA) version 15 for

Microsoft Windows.

ResultsThe study was carried out between March 2010 and August

2011; we studied 74 patients at the IVF center in Kasr El-

Aini Hospital, Cairo University. The patients were divided

into three groups: 28 patients in the cabergoline group, 30

patients in the step-down group, and 16 patients in the

coasting group, and were evaluated. The mean age of the

patients was 27.4 ± 6.0 years in the coasting group,

29.4 ± 3.7 years in the cabergoline group, and 30.1 ± 5.1

years in the step-down group (P = 0.122).

The mean duration of infertility was 5.41 ± 3.7 years in

the coasting group, 6.0 ± 4.18 years in the cabergoline

group, and 5.2 ± 5.05 years in the step-down group

(P = 0.762). The mean BMI was 27.9 ± 3.1 in the

coasting group, 32.6 ± 6.0 in the cabergoline group, and

in 32.8 ± 3.7 the step-down group (P = 0.001).

Antral follicle count was 17.88 ± 2.14 in the coasting

group, 17.84 ± 2.34 in the cabergoline group, and

18.2 ± 2.18 in the step-down group (P = 0.806). The

number of eight-cell embryos was 1.06 ± 2.331 in the

coasting group, 1.07 ± 1.783 in the cabergoline group, and

2.05 ± 3.349 (P = 0.005) in the step-down group. The

mean number of retrieved oocytes was 14.41 + 5.173 in

the coasting group and 21.50 ± 8.307 in the cabergoline

122 Evidence Based Women’s Health Journal

group (the P-value between these two groups was 0.003),

with a statistically significant difference. The mean

number of retrieved oocytes in the step-down group was

15.73 ± 6.264 (P-value between the step-down and the

coasting group was 0.563), with no statistical significance.

The P-value between the cabergoline and the step-down

group was 0.002, with the overall P-value equal to 0.001.

The mean number of MII was 7.50 ± 2.129 in the coast-

ing group and 13.86 ± 5.536 in the cabergoline group,

with a statistical significance between the coasting and

the cabergoline group (P = 0.001), and the mean number

of MII in the step-down group was 11.27 ± 4.763, with no

statistical significance between the coasting and the step-

down group (P = 0.005), and also no statistical signifi-

cance between the cabergoline and the step-down group

(P = 0.130); the overall P-value is equal to 0.001.

The mean number of fertilized oocytes was 5.65 ± 2.936 in

the coasting group and 5.00 ± 4.225 in the cabergoline

group (P = 0.001). The mean number of fertilized oocytes

was 9.47 + 5.355 in the step-down group (P = 0.007

between the coasting and the step-down group,

P = 0.4115 between the cabergoline and the step-down

group, and overall P = 0.002). The chemical pregnancy rate

was eight women (50%) in the coasting group, four women

(14.3%) in the cabergoline group, and 10 women (33.3%) in

the step-down group; the overall P-value was 0.038.

The number of women who developed severe OHSS was

zero in the coasting group, two women in the cabergoline

group, and one woman in the step-down group; the

overall P-value was 0.481. The mean E2 on the day of

HCG was 4512 ± 752 in the coasting group, 4396 ± 1906

in the cabergoline group, with no statistical significance

between them (P-value = 0.687), and 4120 ± 2978 in the

step-down group, with no statistical significance between

the coasting and the step-down groups (P = 0.064) and

also no statistical significance between the cabergoline

and the step-down groups (P = 0.171); the overall P-value

was 0.156 (Table 1).

DiscussionOne of the major, dangerous complications of ovarian

stimulation in IVF-intracytoplasmic sperm injection women

is OHSS, which may be life threatening. There are some

measures for the prevention of OHSS; the most effective

method of prevention is cancelation of the cycle, but in

most of the IVF centers, the couple refuses the cancelation

because of financial and emotional causes. Another method

for the prevention of OHSS is freezing all embryos and

postponing embryo transfer to the subsequent cycle [15];

also, another approach for the prevention of OHSS is

follicular aspiration before the administration of HCG [8].

There are other methods for reducing the risk for severe

OHSS; one of these methods is coasting, which is

withholding the administration of gonadotropin and post-

poning the HCG injection while continuing the GnRH

agonist till the E2 level increases to less than 3000 pg/ml.

Coasting causes atresia of a large number of small follicles,

and thus leads to a decrease in the serum E2 level and

vasoactive mediators [16]. A decrease in the serum level of

FSH concentrations downregulates the LH receptors of

the follicles; thus, this results in the maturation of a fewer

number of oocytes by HCG and leads to a reduction in the

number of oocytes retrieved [16].

In our study, we used cabergoline for the prevention of

severe OHSS, which has been used successfully in high-

risk patients for OHSS. Cabergoline inactivates the

VEGFR-2; this growth factor is very important in

endometrial angiogenesis during the beginning of the

secretory phase [17]. Moreover, VEGF expression is

important in the growth and development of the

endometrium with advancing stages of the menstrual

cycle. VEGF expression has been especially correlated

with implantation of the embryo [17]. It has been shown

that pharmacological doses of cabergoline can inhibit

VEGF-mediated microvascular permeability, proliferation,

and migration of endothelial cells in vitro by inducing

endocytosis of VEGF-R2 [17]. Another study has shown

that changes in VEGF-R2 induced by low-dose cabergo-

line reversed the occurrence of increased vascular

permeability without altering angiogenesis [18]. Cabergo-

line used for the prevention of OHSS reduced neither the

rate of pregnancy nor the implantation rate [18].

In our study, we found that the number of retrieved

oocytes in the cabergoline group was higher than that of

the coasting group (P = 0.003); also, the number of MII

and number of fertilized oocytes in the cabergoline group

Table 1 Patient demographic data

Coasting Cabergoline Step-down P-value

Age 27.4 ± 6 29.4 ± 3.7 30.1 ± 5.1 0.122Duration of infertility 5.41 ± 3.7 6 ± 4.18 5.2 ± 5.05 0.762BMI 27.9 ± 3.1 32.6 ± 6.0 32.8 ± 3.27 0.001AFC 17.88 ± 2.14 17.84 ± 2.34 18.2 ± 2.18 0.806E2 of day HCG 4512 ± 752 4396 ± 1906 4120 ± 2978 0.156Outcome measures

Number of retrieved oocytes 14.41 ± 5.17 21.5 ± 8.307 15.73 ± 6.26 0.001Number of MII 7.5 ± 2.129 13.86 ± 5.536 11.27 ± 4.763 0.001Number of fertilized oocytes 5.56 ± 2.93 10.07 ± 4.22 9.47 ± 5.355 0.002Number of eight-cell embryos 1.06 ± 2.33 1.07 ± 1.78 3.53 ± 4.34 0.005Chemical pregnancy rate 8 (50%) 4 (14.3%) 10 (33.3%) 0.038Rate of OHSS 0 2 1 0.481

AFC, antral follicle count; HCG, human chorionic gonadotropin; OHSS, ovarian hyperstimulation syndrome.

Cabergoline for the prevention of severe OHSS Abdelaal et al. 123

were higher than those of the coasting group (P = 0.001

and 0.002, respectively). The eight-cell embryos in the

coasting group (1.06 ± 2.331) were less than those in the

cabergoline group (1.07 ± 1.783) (P = 0.511).

Also, the chemical pregnancy rate in the coasting group

(50%) was higher than that of the cabergoline group

(14.3%) (P = 0.016); although the number of oocytes

retrieved was affected by coasting, the endometrial

receptivity was affected in the cabergoline group and

thus the clinical pregnancy rate was lower in the

cabergoline group.

These results are in agreement with those of Aflatoonian

et al. [19], as they found that the number of retrieved

oocytes, number of MII, eight-cell embryos, and the

number of patients who developed severe OHSS is higher

in the cabergoline group than in the coasting group.

However, our results differ from theirs in terms of the

pregnancy rate as they found that the pregnancy rate in

the cabergoline group was higher than that of the coasting

group.

In terms of the development of severe OHSS in the

coasting group, no cases were reported, but in the

cabergoline group, two women developed severe OHSS,

with no statistical significance between them (P = 0.48).

The third method we used in this study was the step-

down of the HMG by 1 ampoule in suspicious cases for

OHSS and follow-up by E2 every other day to reach less

than 3000 pg/ml to yield HCG. The step-down in

induction of ovulation will allow more follicles to undergo

atresia, thus reducing the overall number of follicles

capable of secretory activity by the time of HCG

administration; thus, this leads to a reduction in the rate

of OHSS [20]. On comparison of the coasting and the

step-down group, we found that the number of retrieved

oocytes in the step-down group was higher than that of

the coasting group, but with no statistical significance

between them (P = 0.563), and the number of MII in the

step-down group was higher than that of the coasting

group (P = 0.005) (positive statistical significance). Also,

the number of fertilized oocytes in the step-down group

was higher than that of the coasting group, with no

statistical significance (P = 0.007). The number of eight-

cell embryos in the step-down group was higher than that

of the coasting group, with no statistical significance

(P = 0.009).

Also, in terms of the eight-cell embryo number in each

group, there was a significant difference among the three

groups, with an overall P-value equal to 0.005.

The chemical pregnancy rate in the step-down group was

33.3% (10 cases), which is lower than the chemical

pregnancy rate in the coasting group of 50% (eight cases)

(P = 0.270). In the step-down group, only one woman

developed severe OHSS, but in coasting group, none of

the women developed severe OHSS, with no statistical

significance between them.

When we compared the results of the step-down group

and the cabergoline group, it was found that the number

of retrieved oocytes in the step-down group was lower

than that of the cabergoline group (P = 0.002), but with

no statistical significance between both groups. The

number of MII and number of fertilized oocytes

(P = 0.130 and 0.415, respectively) in the step-down

group were lower than those of the cabergoline group.

The chemical pregnancy rate was 33.3% (10 cases) in the

step-down group and 14.3% (four cases) in the cabergo-

line group (P = 0.090). There was no statistical signifi-

cance in the occurrence of severe OHSS among both

groups.

ConclusionWe found that the number of retrieved oocytes, MII, and

fertilized oocytes were higher in the cabergoline group

than the other two groups, but the chemical pregnancy

rate in the coasting group was higher than that of

the other two groups, and the cabergoline group had the

lowest pregnancy rate, with no difference in the

development of severe OHSS.

AcknowledgementsConflicts of interestThere are no conflicts of interest.

References1 Myrianthefs P, Ladakis C, Lappas V, Pactitis S, Carouzou A, Fildisis G,

Baltopoulos G. Ovarian hyperstimulation syndrome (OHSS): diagnosis andmanagement. Intensive Care Med 2000; 26:631–634.

2 Golan A, Ron-El R, Herman A, Soffer Y, Weinraub Z, Caspi E. Ovarian hy-perstimulation syndrome: an update review. Obstet Gynecol Surv 1989;44:430–440.

3 Morris RS, Paulson RJ. Ovarian derived prorenin-angiotensin cascade inhuman reproduction. Fertil Steril 1994; 62:1105–1114.

4 Elchalal U, Schenker JG. The pathophysiology of ovarian hyperstimulationsyndrome – views and ideas. Hum Reprod 1997; 12:1129–1137.

5 Verheul HM, Hoekman K, Jorna AS, Smit EF, Pinedo HM. Targeting vascularendothelial growth factor blockade: ascites and pleural effusion formation.Oncologist 2000; 5 (Suppl 1): 45–50.

6 Brinsden PR, Wada I, Tan SL, Balen A, Jacobs HS. Diagnosis, prevention andmanagement of ovarian hyperstimulation syndrome. Br J Obstet Gynaecol1995; 102:767–772.

7 Delvinge A, Rozenberg S. Epidemiology and prevention of ovarian hyper-stimulation syndrome (OHSS): a review. Hum Reprod Update 2002; 8:559–577.

8 Egbase PE, Makhseed M, Sharhan MA, Grudzinskas JG. Timed unilateralovarian follicular aspiration prior to administration of human chorionic gona-dotropin for the prevention of severe ovarian hyperstimulation syndrome in in-vitro fertilization: a prospective randomized study. Hum Reprod 1997; 12:2603–2606.

9 Aflatonian A, Karymsadeh-Mibodi MA, Dehghani-Firozabadi R, Panah RT,Kalantar SM, Arjmand MHA, Solimani M. The role of aspiration ofhalf of ovarian follicles prior to administration of hCG or GnRH-a forprevention of severe OHSS in ART programs. Middle East Fertil Soc J 2000;5:73–75.

10 Schenker JG. Prevention and treatment of ovarian hyperstimulation syn-drome. Hum Reprod 1993; 8:653–659.

11 Alverez C, Bosch E, Melo MAB, Fernandez-San Chez M, Munoz EA, RemohiJ, et al. The dopamine agonist cabergoline prevents moderate-severe earlyovarian hyperstimulation (OHSS) in high-risk patients. Hum Reprod 2006;21:196.

12 Aboulghar MA, Mansour RT. Ovarian hyperstimulation syndrome: classifi-cations and critical analysis of preventive measures. Hum Reprod Update2003; 9:275–289.

13 Schoot DC, Van Dessel TJ, Hop WC, De Jong FH, Fauser BC. Initial estradiolresponse predicts outcome of exogenous gonadotropins using a step-down

124 Evidence Based Women’s Health Journal

dose regimen for induction of ovulation in polycystic ovary syndrome. FertilSteril 1995; 64:1081–1087.

14 Levinsohn-Tavor O, Friedler S, Schachter M, Raziel A, Strassburger D, Ron-El R. Coasting – what is the best formula? Hum Reprod 2003; 18:937–940.

15 Tiitinen A, Husa L-M, Tulppala M, Simberg N, Seppala M. The effect ofcryopreservation in prevention of ovarian hyperstimulation syndrome. Br JObstet Gynaecol 1995; 102:326–329.

16 Enskog A, Henriksson M, Unander M, Nilsson L, Brannstrom M. Prospectivestudy of the clinical and laboratory parameters of patients in whom ovarianhyperstimulation syndrome developed during controlled ovarianhyperstimulation for in vitro fertilization. Fertil Steril 1999; 71:808–814.

17 Moller B, Rasmussen C, Lindblom B, Olovsson M. Expression of angiogensgrowth factors VEGF, FGF-2, EGF and their receptors in normal human

endometrium during menstrual cycle. Exp Clin Endocrinol Diabetes 2003;111:33–40.

18 Alvarez C, Gonzelarg S, Crespo J, Simon C, Remohi J, Pellicer A. Ca-bergoline (Cb2) administration to prevent ovarian hyperstimulation syndrome(OHSS) in IVF-ICII does not reduce implantation and pregnancy rates. FertilSteril 2006; 86:s337.

19 Aflatoonian A, Ghandi S, Tabibnejad N. Comparison of coasting withCabergoline administration for prevention of early severe OHSS in ARTcycles. Iranian J Reprod Med 2008; 6:51–55.

20 Homburg R, Levy T, Ben-Rafael Z. A comparative prospective study ofconventional regimen with chronic low-dose administration of follicle-stimu-lating hormone for anovulation associated with polycystic ovary syndrome.Fertil Steril 1995; 63:729–733.

Cabergoline for the prevention of severe OHSS Abdelaal et al. 125

Antral follicular count is a better indicator for the assessment of

ovarian reserveAhmed Solimana, Sherif Sharkawya, Tamer Abdelmoniemb, Hesham Al-Inanya

and Odette Wahbac

aDepartment of Obstetrics and Gynecology, CairoUniversity, bDepartment of Obstetrics and Gynecology,Al-Salama Hospital, Kuwait and cGenomic MedicineUnit, Cairo University, Cairo, Egypt

Correspondence to Ahmed Soliman, MD, Departmentof Obstetrics and Gynecology, Faculty of Medicine,Cairo University, Cairo, EgyptTel: + 01 009 811 128;e-mail: a.soliman17@yahoo.com

Received 3 May 2012Accepted 28 June 2012

Evidence Based Women’s Health Journal

2012, 2:126–132

Objectives

To assess the ovarian reserve using the antral follicular count (AFC) versus the basal

serum anti-Mullerian hormone (AMH), follicle-stimulating hormone (FSH), E2, and

serum progesterone according to the ovarian response and the chemical pregnancy

rate.

Material and methods

Eighty-six women undergoing Intra Cytoplasmic Sperm Injection or IVF for male factor,

tubal factor, or unexplained infertility were studied using a long luteal phase protocol

for induction of ovulation. All patients were subjected to measurements of basal serum

FSH, E2, progesterone, and AMH and also assessment of AFC. The ovarian response

was determined by the number of oocytes retrieved. The correlation between different

tests and outcomes was determined.

Results

There was a significant statistical difference between the good and the poor

responders in terms of the basal serum FSH level, serum AMH, and AFC (6.45 ± 2.27

and 12.95 ± 6.69 IU/ml), (0.692 ± 0.819 and 0.227 ± 0.505 ng/ml), and (11.76 ± 5.3

and 5.71 ± 3.3 follicles), respectively. Receiver-operator characteristic curve showed

that both FSH and AMH have limited value in predicting the ovarian response, with

a sensitivity and specificity of 9.26 IU/ml, 70 and 91.2%, likelihood ratio 7.927, and

0.11 ng/ml, 73.33 and 72.85%, likelihood ratio 2.7, respectively, but the AFC has

a high value in predicting the ovarian response; the cut-off value, sensitivity, and

specificity were 8, 85.71 and 73.46%, likelihood ratio 3.22. All the investigated tests

showed limited value in predicting pregnancy.

Conclusion

All three basal serum FSH, AMH, and AFC seem to predict the ovarian reserve but the

AFC is superior to FSH and AMH, and it is less invasive and cheaper.

Keywords:

Anti-Mullerian hormone, ICSI, infertility, IVF, ovarian reserve, ovarian response

Evid Based Women Health J 2:126–132& 2012 Evidence Based Women’s Health Journal2090-7265

IntroductionOvarian reserve (OR), a term that has evolved in the era

of assisted reproductive technology (ART), refers to the

residual oocyte–granulosa cell repertoire that, at any

given age, is available for procreation. Both quantitative

and qualitative deteriorations in the oocyte complement,

and therefore a waning OR, are recognized phenomena

associated with advancing age. Although a decline in OR

thus accompanies chronological aging, there is an

acceleration in this process [1].

Diminished OR is defined as a decrease in the

reproductive potential of a woman’s remaining oocytes

to the point at which the probability of producing a viable

gestation is low. The age-related decline in reproductive

potential is the principal factor contributing toward the

increase in the prevalence of infertility in western

society [2]. A cut-off level of less than six antral follicles

used by Kwee et al. [3] had a sensitivity of 41% and

a specificity of 95%, with a prevalence of 27% for a poor

response. In the population studied after ovarian

hyperstimulation in an IVF treatment, the accuracy was

89% [3].

OR tests can identify the high-risk population of poor

response, and hence yield information that affects

prognosis, counseling, and treatment decisions. Abnormal

tests in older women can help to persuade patients to

abandon plans to pursue aggressive and costly treatment.

On the other hand, mild abnormal test in young women

can help to convince them to do just the opposite and to

take the fullest advantage of a rapidly closing window of

opportunity [4].

Over the past two decades, a number of so-called ovarian

reserve tests (ORTs) have been designed to determine

oocyte reserve and quality and have been evaluated for

their ability to predict the outcome of IVF in terms of

oocyte yield and the occurrence of pregnancy. Many of

126 Original article

2090-7265 & 2012 Evidence Based Women’s Health Journal DOI: 10.1097/01.EBX.0000419240.48874.b7

these tests have become part of the routine diagnostic

procedure for infertility patients who undergo assisted

reproductive techniques. The unifying goals are tradi-

tionally to determine how a patient will respond to

stimulation and their chances of pregnancy [5].

Several tests have been proposed to define the OR status

of individual patients and predict fertility outcomes for

assisted reproduction and these include age, history of

canceled cycles, clinical symptoms (e.g. oligomenorrhea),

basal blood tests [follicle-stimulating hormone (FSH),

luteinizing hormone (LH), anti-Mullerian hormone

(AMH), estradiol (E2), FSH/LH ratio, inhibin B, proges-

terone (P4), P4 : E2 ratio, testosterone, vascular endothelial

growth factor (VEGF), insulin-like growth factor-1 : insulin-

like growth factor binding protein-1 ratios (IGF-1 : IGFBP-1

ratio)]. Dynamic tests (clomiphene citrate challenge test

(GnRH agonist stimulation test, exogenous FSH ovarian

reserve test). Ultrasound tests [ovarian volume, antral

follicle count (AFC), ovarian stromal blood flow] [6,7].

Materials and methodsPatients

This is a prospective cross sectional study carried out at

the IVF center of the Egypt International Hospital

between January 2007 and February 2009, and included

86 patients who would be undergoing IVF or ICSI for the

first time with the following inclusion criteria: age from

31 to 49 years, cause of infertility male factor, tubal factor,

or unexplained infertility, basal FSH level below 15 mIU/ml,

both ovaries well visualized by transvaginal sonography

without any cysts, no criteria of polycystic ovary, no history

of ovarian surgery, no history of endocrinological disorder,

and no history, symptoms, or signs of endometriosis. The

study was discussed with the patient before her participa-

tion and consent was taken.

Study protocol

All patients of the study were recruited on days 2–5 of the

cycle; before the ovulation protocol was carried out, they

were subjected to the following: full history taking and

a thorough clinical examination, transvaginal sonography

for the assessment of the AFC, both ovaries, and the

uterus, and to exclude any pelvic pathology. For ultra-

sonographic evaluation, we used transvaginal 7.5 MHz

(Sonoace 9900; Medison, Seoul, Korea).

Measurements of basal serum FSH, E2, progesterone,

and AMH were carried out. Also, routine laboratory tests

such as complete blood test and assessment of liver and

kidney functions were carried out.

Induction of ovulation was carried out using the long

luteal phase protocol. All patients received Decapeptyl

(GnRHa) 0.1 mg subcutaneous injection daily at 8 p.m.

starting from day 20 of the cycle and for 2 weeks, after

insuring downregulation by serum E2 level; human

menopausal gonadotropins (hMG) injection was adminis-

tered at a starting dose of 150–300 IU/day for 7 days and

then adjusted according to the response. The patient was

monitored every other day for folliculometry. Human

chorionic gonadotropin (hCG) was administered if the

mean follicular diameter was 18–20 mm at a dose of

10 000 IU intramuscular injection; if the mean diameter

of the follicles was less than 10 mm after 14 days, the

cycle was canceled. Ovum pick up wascarried out after

36 h of hCG and embryo transfer was carried out on day 2

or 3. Pregnancy was confirmed by a positive qualitative

B-hCG test 2 weeks after embryo transfer.

Hormone assays

Serum FSH, E2, and AMH were measured on cycle day 3,

whereas progesterone was measured on cycle day 1. Both

E2 and progesterone were measured again on the day of

hCG administration. Serum FSH was measured using

radioimmunoassay (RIA) (Gambyt-CR; Diagnostic Pro-

ducts Corporation, Los Angeles, California, USA). The

interassay coefficient of variation (CV) of FSH was 3.7%.

The intra-assay CV was 4.7%.

Serum AMH was measured using two enzyme-linked

immunosorbent assays. In the traditional assay, Immulon

two plates (Dynatech Corp., Chantilly, Virginia, USA)

were coated with monoclonal antibody 10.6 (Dr R. Cate,

Cambridge, Massachusetts, USA), raised against recom-

binant human AMH by an overnight incubation at room

temperature. Sera were assayed at 1 : 4, 1 : 8, 1 : 16, and

1 : 32 dilutions in PBS containing 1% BSA. As the second

antibody, we used polyclonal antibody L40, consisting of

an immunoglobulin G fraction isolated by affinity

chromatography on protein A-Sepharose from the serum

of a rabbit immunized with recombinant human AMH.

L40 was added to the wells at 1 mg/ml in PBS–1% BSA for

1 h at room temperature. Subsequently, an alkaline

phosphatase-labeled goat anti-rabbit immunoglobulin G

antibody (Jackson ImmunoResearch Laboratories Inc.,

West Grove, Pennsylvania, USA) was added and incu-

bated for a further hour before the reaction was visualized

with an MRX spectrophotometer (Dynatech Corp.) at

405 nm, using paranitrophenyl phosphate (Sigma Chimie,

Saint-Quentin-Fallavier, France) as a substrate.

Estradiol levels were measured by RIA. Duplicate

aliquots (0.2 ml) of plasma were fixed with 0.1 ml of a

1.5 mol/l sodium carbonate solution, pH 10.5, and

extracted with 10 vol of diethyl ether. The mixture was

frozen, the ether was decanted and evaporated, and the

residue was dissolved in PBS and incubated with tritium-

labeled E2 (Amersham International, Little Chalfont,

Buckinghamshire, UK) and sheep anti-E2 antiserum

(Bioclin Services, Cardiff, Wales, UK). Bound and free

E2 were separated using dextran-coated charcoal. The

interassay CVs for two control plasmas were 9.5 and 6.6%

for E2 concentrations of 174 and 409 pmol/l, respectively.

Serum progesterone was measured using RIA (Gambyt-

CR; Diagnostic Products Corporation). The intra-assay

and interassay CV were 9.6 and 4.9%.

Outcome measures

The main outcome of the study was the ovarian response

and was determined according to the number of oocytes

retrieved; five or more was considered to indicate a good

response and less than five was considered a poor

AFC as indicator for ovarian reserve Soliman et al. 127

response. The secondary outcome was the chemical

pregnancy rate (PR).

Statistical analysis

Data were statistically described in terms of range,

mean ± SD, median, frequencies (number of cases),

and relative frequencies (percentages) when appropriate.

Comparison of quantitative variables between different

groups in the present study was carried out using the

Mann–Whitney U-test for independent samples. For

comparison of categorical data, the w2 test was carried

out. Exact test was used when the expected frequency

was less than 5. Accuracy was represented using the terms

sensitivity and specificity. Receiver-operator characteris-

tic (ROC) analysis was used to determine the optimum

cut-off value for the diagnostic markers studied. Correla-

tions between various variables were determined using

Pearson’s moment correlation and Spearman’s rank

correlation equations. A probability value (P value) less

than 0.05 was considered statistically significant. All

statistical calculations were carried out using computer

programs Microsoft Excel version 7 (Microsoft Corpora-

tion, New York, USA), SPSS (Statistical Package for the

Social Science; SPSS Inc., Chicago, Illinois, USA), and

the ArcusQuickStat (Biomedical version; Addison Wesley

Longman Ltd, USA) statistical program.

ResultsAll the 86 patients were eligible for analysis; the patients

were categorized according to the response into good

responders, who retrieved five or more oocytes, or poor

responders, who retrieved less than five, and according to

the result of the B-hCG test as a positive test (chemical

pregnancy) or a negative test. 81.4% (70 patients) were

good responders and 18.6% (16 patients) were poor

responders, and the chemical PR was 33.72% (29

patients).

Table 1 shows a comparison of both groups in terms of the

general variables and it indicates that there was a

statistically significant difference between both groups

in terms of the patient age, duration of infertility, and

number of hMG ampoules used. As the poor responders

were older than the good responders and they had more

long-standing infertility, they required more hMG am-

poules to be stimulated.

In terms of the other variables, when we compared both

groups, it was found that there was a statistically

significant difference in the basal serum FSH level (Pvalue 0.000), serum AMH (P value 0.001), and basal AFC

(P value 0.003), whereas there was no statistically

significant difference between both groups in the basal

levels of E2 and progesterone. Measurement of the E2

and progesterone levels on the day of hCG administration

indicated that there was a statistically significant

difference between both groups in the E2 level, whereas

the progesterone level was insignificant. Table 2 shows

a comparison between both groups in terms of these

variables.

On analyzing the previous data on the ROC curve

characteristic, we found that basal AFC was superior to

basal serum FSH and serum AMH in the prediction of

the ovarian response; the best cut-off value, sensitivity,

specificity, and likelihood ratio for each test are presented

in Table 3.

The secondary outcome was the chemical PR, which was

33.72% (29 patients). Table 4 presents a comparison of

the group with a positive pregnancy test and those with a

negative pregnancy test in terms of the basal serum FSH,

Table 1 Background characteristics of the study population

Variables Total (n = 86) Good responders (n = 70) Poor responders (n = 16) P value

Mean age ± SD 31.41 ± 5.7 29.68 ± 4.391 38.44 ± 4.77 0.0000 (S)Mean BMI 29.65 ± 4.41 28.52 ± 5.5 29.9 ± 4.1 NSDuration of infertility (years) 6.34 ± 4.6 5.43 ± 3.7 10.27 ± 5.7 0.001 (S)Type of infertility

Primary 53 (61.62%) 14 (16.27%)Secondary 17 (19.767%) 2 (2.32%)

Mean number of hMG ampoules used 38.69 ± 14.2 49.56 ± 19.42 0.02 (S)Mean duration of stimulation 11.28 ± 5.65 10.69 ± 2.49 NS

hMG, human menopausal gonadotropin; S, significant.

Table 2 Hormonal assay of follicle-stimulating hormone, Estradiol, progesterone, anti-Mullerian hormone

Variables Total (n = 86) Good responders (N = 70) Poor responders (N = 16) P value

Basal serum FSH ± SD (IU/ml) 8 ± 4.46 6.45 ± 2.27 12.95 ± 6.69 0.000 (S)Basal serum E2 ± SD (pg/ml) 26.1 ± 30.71 26.67 ± 33.26 23.75 ± 16.66 NSSerum E2 level on day of hCG (pg/ml) 22 394.43 ± 1714.92 2580 ± 1689.62 874.2 ± 1025.84 0.000 (S)Value of change in E2 level (pg/ml) 2147.7 ± 1809.9 2347.69 ± 1811.84 987.8 ± 1440 0.03 (S)Basal serum progesterone level (ng/ml) 0.554 ± 0.487 0.587 ± 0.504 0.318 ± 0.268 NSSerum progesterone level on day of hCG (ng/ml) 0.964 ± 0.668 0.927 ± 0.565 1.118 ± 1.02 NSBasal serum AMH (ng/ml) 0.613 ± 0.819 0.692 ± 0.819 0.277 ± 0.505 0.001 (S)Basal AFC 10.94 ± 5.5 11.76 ± 5.3 5.71 ± 3.3 0.003 (S)

AFC, antral follicular count; AMH, anti-Mullerian hormone; FSH, follicle-stimulating hormone; hCG, human chorionic gonadotropin; S, significant.

128 Evidence Based Women’s Health Journal

E2, progesterone, AMH, and AFC. The data showed

that none of these tests showed a statistically significant

difference between both groups, except in the basal

FSH (6.65 ± 2.31 and 9.23 ± 5.23) and the basal AFC

(8 ± 5.3 and 8.77 ± 4.48), respectively. After plotting

the data on the ROC curve, it was found that none

of the tests used had a high power of prediction or

correlation with pregnancy. The data are shown in Table 5

(Figs 1–3).

DiscussionIn this study, we used different tests for the prediction of

the ovarian response in 86 patients undergoing IVF or

ICSI for the first time; we found that age showed a strong

relationship with the response as the mean age for the

good responders and the poor responders was (68 ± 4.391

and 38.44 ± 4.77), respectively. In a study carried out by

Van Voorhis [8], it was concluded that age is a well-

established predictor for the OR. And the good Odds

Ratio can be suggested by age less than 38 years at the

time of treatment, as the decay rate of resting follicles

accelerates from 38 years onwards [9,10].

Also, a relatively longer duration of infertility was

observed in poor responders, 10.27 ± 5.7 years, whereas

in good responders, it was 5.43 ± 3.7 years. This can be

attributed to the fact that patients with a poor OR would

have undergone more trials of ICSI with poor results,

thus increasing the duration of infertility. As reported by

Hammadieh et al. [11], increasing duration of infertility is

associated with a decrease in the probability of sponta-

neous pregnancy in untreated infertile couples (15.3% for

1–3 years, 14% for 4–6 years, 12.9% for 7 – 9 years, 12.4%

for 10–12 years, and 8.6% for more than 12 years).

The number of hMG ampoules used in both good and

poor responders showed a statistically significant differ-

ence (38.69 in good responders and 49.56 in poor

responders), and this can be considered a logical result

in poor responders. A small number of randomized-

controlled trials and retrospective studies have evaluated

the effectiveness of high-dose FSH regimens beyond

Table 4 Pregnancy outcome according to different parameters

Variables Positive pregnancy test (N = 29) Negative pregnancy test (N = 57) P value

Basal serum FSH (IU/ml) 6.65 ± 2.31 9.23 ± 5.23 0.001 (S)Basal serum E2 (pg/ml) 18.94 ± 12.79 34.06 ± 43.86 NSSerum E2 on day hCG injection (pg/ml) 2424.05 ± 2129 1967.19 ± 137794 NSBasal progesterone level (ng/ml) 0.485 ± 0.47 0.613 ± 0.538 NSSerum progesterone level on day hCG injection (ng/ml) 0.913 ± 0.7 0.939 ± 0.636 NSBasal serum AMH (IU/ml) ± SD 0.764 ± 0.86 0.58 ± 0.904 NSBasal AFC ± SD 8 ± 5.3 8.77 ± 4.48 0.002 (S)

AFC, antral follicular count; AMH, anti-Mullerian hormone; FSH, follicle-stimulating hormone; hCG, human chorionic gonadotropin; S, significant.

Table 5 Accuracy of different test in predicting pregnancy

Variables Cut-off value Sensitivity (%) Specificity (%) Likelihood ratio Area under ROC curve (%)

Basal FSH 4.935 IU/ml 33.33 95.45 7.32 64.58o75Basal E2 47.32 pg/ml 100 15.38 1.18 54.35o75Basal AMH 0.3 ng/ml 65.38 63.88 1.81 46.84o75Basal AFC 9.048 91.3 59.1 2.23 70.45o75

AFC, antral follicular count; AMH, anti-Mullerian hormone; FSH, follicle-stimulating hormone; ROC, receiver-operator characteristic.

Table 3 Accuracy of different parameters

Data set Cut-off value Sensitivity (%) Specificity (%) Likelihood ratio Area under ROC curve (%)

Basal serum FSH 9.26 IU/ml 70 91.2 7.927 74.26o75Basal serum AMH 0.11 ng/ml 73.33 72.85 2.7 71.1o75AFC 8.022 85.7 73.469 3.22 85.5475

AFC, antral follicular count; AMH, anti-Mullerian hormone; FSH, follicle-stimulating hormone; ROC, receiver-operator characteristic.

Figure 1

ROC plot for FSH in predicting poor response

0.00 0.25 0.50 0.75 1.000.00

0.25

0.50

0.75

1.00Sensitivity

1-Specificity

9.25832

FSH

Receiver-operator characteristic plot for basal follicle-stimulatinghormone in predicting a poor response.

AFC as indicator for ovarian reserve Soliman et al. 129

a 300 IU threshold. The results of these studies have

shown these approaches to be of little or no clinical

benefit, although both the costs of treatment and the

side-effects were higher. Other studies have examined

regimens of 450 IU FSH or even higher [12]. The

number of days of stimulation by hMG in patients

showed an insignificant statistical difference: 11.28 days

in good responders and 10.69 days in poor responders.

Lengthy, expensive regimens are not free from short-term

and long-term risks and complications.

In our study, the result of ovarian stimulation was

expressed as the number of oocytes retrieved: greater

than or equal to five was considered a good response and

less than five was considered a poor response for both

groups [13]. Seventy patients produced five or more

oocytes (81.4%) whereas 16 patients produced less than

five oocytes (18.6%). However, 29 patients (33.72%)

achieved chemical pregnancy whereas 57 (66.28%) did

not.

The AFC was studied as an ORT. OR may be defined as

the total number of follicles that can be stimulated under

maximal ovarian stimulation with FSH. ORTs are

considered to indirectly reflect the size of the cohort of

small antral follicles (2–10 mm in diameter) in the ovary.

Thus, a decrease in follicle number increases the risk of a

poor response after ovarian hyperstimulation in patients

with IVF at an older age [3,14].

In our study, AFC was obviously lower in poor responders,

5.71, whereas it was 11.76 in good responders, with a

statistically significant difference (P value 0.003) (the

mean AFC in all patients was 10.94). Taking P value less

than 0.05, AFC was considered significant for a poor

response. In ROC curve analysis, area under the curve is

greater than 75% (85.56), that is the value of testing AFC

in predicting a response is high. With an optimum cut-off

point = 8, the highest sensitivity and specificity that

could be obtained were 85.71 and 73.46%, respectively.

The likelihood ratio was 3.22. The mean value of AFC on

cycle day 3 in pregnant patients was 8 ranging from 5 to

30. In nonpregnant patients, the mean value was 8.77,

ranging from 3 to 17. There was a statistically significant

difference (P value 0.002). Area under the curve was less

than 75% (70.45), that is the value of testing AFC in

predicting pregnancy was low. At a cut-off value of 9.048,

the sensitivity was 91.3% and specificity was 59.1%. The

likelihood ratio was 2.23. And this agrees with that more

number of decreases implantation rates [15]. Another

study showed that a low AFC less than 10 antral follicles

with a diameter measuring 2–10 mm was found to

indicate reduced OR and decreased chance for pregnancy

after ART [16].

The mean basal FSH among patients was 8 IU/ml. In good

responders, it was 6.45 ± 2.27 IU/ml, whereas in poor

responders it was 12.95 ± 6.69 IU/ml, with a statistically

significant difference when considering P value less than

0.05 as significant. In ROC curve analysis, area under the

curve was less than 75% (74.26), that is the value of testing

FSH in predicting response is low. With an optimum cut-

off point = 9.26 IU/ml, the highest sensitivity and specifi-

city that could be obtained were 70 and 91.2%, respectively.

The likelihood ratio was 7.927. The mean value of basal

serum FSH in pregnant patients was 6.65 ± 2.31 IU/ml,

ranging from 2.31 to 11.64 IU/ml. In nonpregnant patients,

the mean value was 9.23 ± 5.23 IU/ml, ranging from 5.23 to

15 IU/ml. There was a statistically significant difference

(P value 0.001). In ROC curve analysis, area under the

curve was less than 75% (64.85), that is the value of testing

FSH in predicting pregnancy is low. With an optimum cut-

off point = 4.94 IU/ml, the highest sensitivity and specifi-

city that could be obtained were 33.33 and 95.46%,

respectively. The likelihood ratio was 7.32.

The literature has consistently suggested that FSH

testing is the best marker for assessing OR and for

predicting the response to super ovulation, with a good

correlation with PRs [17,18] and had significant correla-

tions with FSH-hMG administered, total number of

oocytes retrieved, conception rate, and the possibility of

developing ovarian hyperstimulation syndrome [19,20].

A cut-off point of 5.25 mIU/ml for day 3 FSH was

suggested for predicting OR [20].

Figure 2

ROC plot for AFC in predicting poor response

0.00 0.25 0.50 0.75 1.000.00

0.25

0.50

0.75

1.00Sensitivity

1-Specificity

AFC 8.022

Receiver-operator characteristic plot for antral follicle count in theprediction of a poor response.

Figure 3

ROC plot for AMH in predicting poor response

0.00 0.25 0.50 0.75 1.000.00

0.25

0.50

0.75

1.00Sensitivity

1-Specificity

AMH

.1053

Receiver-operator characteristic plot for basal anti-Mullerian hormonelevel in the prediction of a poor response.

130 Evidence Based Women’s Health Journal

However, an elevated FSH value does not necessarily

mean that a woman will be unable to conceive a

pregnancy naturally. There are many examples of patients

with elevated FSH values who were subsequently able to

conceive without treatment. The ORTs are better used as

a predictor of patients who are not likely to benefit from

ART; young women even with elevated FSH may still

have a favorable IVF outcome [8].

The mean value of serum estradiol days 1–3 was 26.1 pg/ml

in all patients, 26.67 pg/ml in good responders, 23.75 pg/ml

in poor responders, with a statistically insignificant

difference. The mean value of basal serum estradiol days

1–3 in pregnant patients was 18.94 pg/ml, ranging from 5 to

47 pg/ml. In nonpregnant patients, the mean value was

34.06 pg/ml, ranging from 10 to 189 pg/ml. There was a

statistically insignificant difference (P value 0.177). In

ROC curve analysis for the ability of basal day 3 serum

estradiol to predict pregnancy, the area under the curve was

less than 75% (54.35), that is the value of testing estradiol

in predicting pregnancy is low. At a cut-off value of

47.32 IU/ml, the sensitivity was 100% and specificity was

15.38%. The likelihood ratio was 1.18.

In a study carried out to test the hypothesis that elevated

E2 levels on day 3 of IVF cycles without a GnRH agonist

are associated with reduced oocyte numbers and PRs, the

ongoing PR per retrieval for patients with E2 levels less

than 30 pg/ml was significantly higher than that for

patients with E2 levels 31–75 pg/ml. There were no

pregnancies if the E2 level was greater than 75 pg/ml.

The mean number of oocytes per retrieval was signifi-

cantly lower in patients from the E2 groups with E2

greater than 60 pg/ml compared with patients in groups

with E2 less than 60 pg/ml. Day 3 FSH and E2 levels were

also evaluated simultaneously. In patients with the lowest

levels of FSH and E2, the PR was the highest. No

pregnancies occurred if the FSH level was greater than

17 mIU/ml (conversion factor to SI unit, 1.00) and the E2

level was greater than 45 pg/ml on day 3 [21].

A systematic analysis of all studies testing basal serum

estradiol as an ORT concluded that the clinical applic-

ability for basal estradiol as a test before starting IVF is

prevented by a very low predictive accuracy, both for poor

response and for nonpregnancy [5].

The mean value of serum estradiol day hCG injection was

2239.43 pg/ml in all patients: 2580 pg/ml in good respon-

ders and 874.2 pg/ml in poor responders, with a significant

statistical difference. The mean value of the basal serum

estradiol day hCG injection in pregnant patients was

2424.05 pg/ml, ranging from 500 to 10 020 pg/ml. In

nonpregnant patients, the mean value was 1967.19 pg/ml,

ranging from 124 to 3806 pg/ml. There was a statistically

insignificant difference (P value 0.497). Serum estradiol on

day of hCG injection reflects the number and health of

oocytes, and showed a statistical difference in predicting

the response however, it couldn’t prove that value.

The mean value of serum progesterone days 1–3 was

0.554 ng/ml in all patients, 0.587 ng/ml in good responders

and 0.318 ng/ml in poor responders, with a statistically

insignificant difference (P value 0.244). The mean value

of serum progesterone day 1–3 in pregnant patients was

0.485 ng/ml, ranging from 0.1 to 2.1 ng/ml. In nonpreg-

nant patients, the mean value was 0.613 ng/ml, ranging

from 0.1 to 1.9 ng/ml. There was a statistically insignif-

icant difference (P value 0.244). The mean value of

serum progesterone day hCG was 0.924 ng/ml in all

patients, 0.927 ng/ml in good responders and 1.118 ng/ml

in poor responders, with a statistically insignificant

difference (P value 0.655). The mean value of serum

progesterone day hCG injection in pregnant patients was

0.913 ng/ml, ranging from 0.3 to 2.9 ng/ml. In nonpreg-

nant patients, the mean value was 0.939 ng/ml, ranging

from 0.3 to 2.5 ng/ml. There was a statistically insignif-

icant difference (P value 0.821).

In a study carried out by Silverberg et al. [22], it was

reported that serum progesterone (P4) levels greater than

0.9 ng/ml on the day of hCG administration were

associated with decreased PRs in IVF/embryo transfer

cycles. Two critical breakpoints were identified: 0.4 and

0.9 ng/ml. Clinical pregnancies occurred in nine of 18

patients in patients with P4 less than 0.4 ng/ml compared

with 11 of 81 patients with P4 between 0.4 and 0.9 ng/ml

(P = 0.001) and zero of 14 patients with P4 more than or

equal to 0.9 ng/ml (P = 0.001). These findings indicate

that even modest increases in serum P4 levels (40.4 ng/ml)

are associated with reduced PRs in IVF/ET cycles. In

addition, it appears that the mechanism may not exclu-

sively involve poor oocyte quality [22].

The mean value of serum AMH day 3 was 0.613 ± 0.819

ng/ml in all patients, 0.692 ± 0.819 ng/ml in good responders

0.277 ± 0.505 ng/ml in poor responders, with a statistically

significant difference (P value 0.001). In ROC curve

analysis, area under the curve was less than 75% (71.71),

that is the value of testing AMH in predicting response was

low. At a cut-off value of 0.11 ng/ml, the sensitivity was

73.33% and the specificity was 72.85% the likelihood ratio

was 2.7. The mean value of basal serum AMH day 3 in

pregnant patients was 0.764 ± 0.86 IU/ml, ranging from 0.05

to 4 IU/ml. In nonpregnant patients, the mean value was

0.58 ± 0.904 IU/ml, ranging from 0.05 to 3.8 IU/ml. There

was a statistically insignificant difference (P value 0.22). In

ROC curve analysis area, under the curve was less than 75%

(46.84), that is the value of testing AMH in predicting

pregnancy was low. At a cut-off value of 0.3 ng/ml, the

sensitivity was 65.38% and specificity was 63.88%. The

likelihood ratio was 1.81. Serum AMH day 3 was useful in

predicting response; however, it was not significant in

predicting pregnancy.

In a study carried out by Gnoth et al. [23] that included

316 patients, a total of 132 oocyte retrievals were

performed. A calculated cut-off level less than 1.26 ng/ml

AMH alone detected poor respondents (o4 oocytes) with a

sensitivity of 97%, and there was a 98% correct prediction

of a normal response in COH if the levels were above this

threshold. With levels less than 0.5 ng/ml, a correct

prediction of a very poor response (o2 oocytes) was

possible in 88% of cases. Levels of AMH greater than

0.5 ng/ml were not significantly correlated with clinical PRs.

AFC as indicator for ovarian reserve Soliman et al. 131

Their results showed that age and AMH levels are superior

parameters predicting ovarian response [23].

To conclude, basal AFC on day (2–5) before down-

regulation has a high predictive value in the prediction of

OR and it is a cheap and noninvasive test. In addition, age

is one of the best markers for ovarian reserve and for us it

seems that the age of 38 years is the cut-off point.

Moreover, a longer duration of infertility may be

indicative of a poor OR.

AcknowledgementsConflicts of interestThere are no conflicts of interest.

References1 Galey-Fontaine J, Cedrin-Durnerin I, Chaıbi R, Massin N, Hugues J-N. Age

and ovarian reserve are distinct predictive factors of cycle outcome in lowresponders. Reprod BioMed Online 2005; 10:94–99.

2 Sharara FI, Scott RT Jr. Assessment of ovarian reserve and treatment of lowresponders. Infertility Reprod Med Clin N Am 1997; 8:501–522.

3 Kwee J, Elting ME, Schats R, McDonnell J, Lambalk CB. Ovarian volume andantral follicle count for the prediction of low and hyper responders within vitro fertilization. Reprod Biol Endocrinol 2007; 5.

4 Speroff L, Fritz MA. Clinical gynecologic endocrinology and infertility. 7th ed.USA: Lippincott Williams and Wilkins; 2005.

5 Broekmans FJ, Kwee J, Hendriks DJ, Mol BW, Lambalk CB. A systematicreview of tests predicting ovarian reserve and IVF outcome. Hum ReprodUpdate 2006; 12:685–718.

6 Johnson NP, Mak W, Sowter MC. Surgical treatment for tubal disease inwomen due to undergo in vitro fertilisation. Cochrane Database Syst Rev2004; 3:CD002125.

7 Lass A. Assessment of ovarian reserve – is there a role for ovarian biopsy?Hum Reprod 2001; 16:1055–1057.

8 Van Voorhis BJ. Outcomes from assisted reproductive technology. ObstetGynecol 2006; 107:183–200.

9 Kupesic S, Kurjak A. Predictors of IVF outcome by three-dimensional ultra-sound. Hum Reprod 2002; 17:950–955.

10 Pellicer A, Ardiles G, Neuspiller F, Remohı J, Simon C, Bonilla-Musoles F.Evaluation of the ovarian reserve in young low responders with normal basallevels of follicle-stimulating hormone using three-dimensional ultra-sonography. Fertil Steril 1998; 70:671–675.

11 Hammadieh N, Olufowobi O, Afnan M. Factors affecting the outcome ofin vitro fertilization treatment. Prog Obstet Gynecol 2005; 16:259–276.

12 Siristatidis CS, Hamilton MP. What should be the maximum FSH dose inIVF/ICSI in poor responders? J Obstet Gynecol 2007; 27:401–405.

13 Ficicioglu C, Kutlu T, Baglam E, Bakacak Z. Early follicular antiMullerian hormoneas an indicator of ovarian reserve. Fertil Steril 2006; 85:592–596.

14 Goverde AJ, McDonnell J, Schats R, Vermeiden JPW, Homburg R, Lambalk CB.Ovarian response to standard gonadotrophin stimulation for IVF isdecreased not only in older but also in younger women in couples with idio-pathic and male subfertility. Hum Reprod 2005; 20:1573–1577.

15 Verberg MFG, Eijkemans MJC, Macklon NS, Heijnen EMEW, Baart EB,Hohmann FP, et al. The clinical significance of the retrieval of a low number ofoocytes following mild ovarian stimulation for IVF: a meta-analysis. HumReprod Update 2009; 15:5–12.

16 Frattarelli JL, Lauria-Costab DF, Miller BT, Bergh PA, Scott RT Jr. Basal antralfollicle number and mean ovarian diameter predict cycle cancellation andovarian responsiveness in assisted reproductive technology cycles. FertilSteril 2000; 74:512–517.

17 Scott RT. Evaluation and treatment of the low responder patient. In:Gardner DK, Weissman A, Howles CM, et al., editors. Textbook of assistedreproductive techniques: laboratory and clinical perspectives. London, UK:Martin Dunitz; 2002.

18 Toner JP, Philput CB, Jones GS, Muasher SJ. Basal follicle-stimulating hor-mone level is a better predictor of in vitro fertilization performance than age.Fertil Steril 1991; 55:784–791.

19 Barnhart K, Coutifaris C. The use of ultrasound in the evaluation and treat-ment of the infertile woman. In: Benson CB, Arger PH, Bluth EI, editors.Ultrasonography in obstetrics and gynecology: a practical approach. NewYork: Thieme; 2000. pp. 965–971.

20 Onagawa T, Shibahara H, Ayustawati, Machida S, Hirano Y, Hirashima C,et al. Prediction of ovarian reserve based on day-3 serum follicle stimulatinghormone concentrations during the pituitary suppression cycle using a go-nadotropin releasing hormone agonist in patients undergoing in vitro fertili-zation-embryo transfer. Gynecol Endocrinol 2004; 18:335–340.

21 Licciardi FL, Liu H-C, Rosenwaks Z. Day 3 estradiol serum concentrations asprognosticators of ovarian stimulation response and pregnancy outcome inpatients undergoing in vitro fertilization. Fertil Steril 1995; 64:991–994.

22 Silverberg KM, Burns WN, Olive DL, Riehl RM, Schenken RS. Serum pro-gesterone levels predict success of in vitro fertilization/embryo transfer inpatients stimulated with leuprolide acetate and human menopausal gonado-tropins. J Clin Endocrinol Metab 1991; 73:797–803.

23 Gnoth C, Schuring AN, Friol K, Tigges J, Mallmann P, Godehardt E.Relevance of anti-Mullerian hormone measurement in a routine IVF program.Hum Reprod 2008; 23:1359–1365.

132 Evidence Based Women’s Health Journal

Comparison between transfrontal and axial three-dimensional

ultrasound acquisition in the visualization of midline structures

of the fetal brainSherif M.M. Negm and Rasha A. Kamel

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

Correspondence to Sherif M.M. Negm, El-Safa MedicalTower, 1 Higaz Square, El-Mohandessein, Giza, EgyptTel/fax: + 20 02 33386408;e-mail: sherifnegm@hotmail.com

Received 27 June 2012Accepted 15 July 2012

Evidence Based Women’s Health Journal

2012, 2:133–137

Objective

To compare the visualization of the midline structures of the fetal brain as well as

visualization of the fastigium of the fourth ventricle and the primary and secondary

vermian fissures obtained by three-dimensional (3D) multiplanar reconstruction of

volumes acquired from the axial plane with transfrontal 3D acquisition.

Study design

A prospective observational study.

Patients and methods

A total of 127 patients with a normal fetal anomaly scan between 18 and 24 weeks

participated in this study. Fetal brain volumes for the multiplanar evaluation were

obtained with the transcerebellar plane as the initial plane of acquisition, with the

incident ultrasound beam making an angle of about 451 with the cerebral midline. For

the transfrontal acquisition, the plane of the midsagittal fetal facial profile was obtained

with the ultrasound beam aligned with the frontal suture so as to utilize the metopic

suture as an acoustic window.

Results

The acquisition of the fetal brain in the axial plane was successful in 122 cases

(96.1%), whereas the transfrontal acquisition was successful in 106 cases (83.4%),

with a statistically significant difference between the two methods (P = 0.002).

Visualization of the median plane of the fetal brain by 3D multiplanar reconstruction

was adequate in 99 out of the 122 (81.1%) volumes, whereas 94 out of the 106

(88.7%) transfrontal acquisitions resulted in adequate midline images; the difference

between the two acquisition methods was not statistically significant (P = 0.12). There

was no statistically significant difference between the two acquisition methods in the

visualization of the fastigium of the fourth ventricle or the primary and secondary

vermian fissures, which were adequately visualized in 58/122 (47.5%) of the 3D

multiplanar reconstructed images and in 62/106 (60.8%) of the transfrontally acquired

volumes (P = 0.09).

Conclusion

Images of the midsagittal plane of the fetal brain obtained by 3D multiplanar

reconstruction of volumes acquired from axial plane are easier to acquire than the 3D

transfrontal approach and result in comparable image quality, with adequate

visualization of the cerebral midline as well as the main landmarks of the cerebellar

vermis.

Keywords:

fetal brain, midsagittal plane, multiplanar, three dimensional, three-dimensional

ultrasound, transfrontal

Evidence Based Women’s Health Journal 2:133–137& 2012 Evidence Based Women’s Health Journal2090-7265

IntroductionThe standard axial ultrasound views used to evaluate the

fetal brain do not adequately display the midline cerebral

structures, most notably the corpus callosum and the

cerebellar vermis [1], resulting in a low detection rate of

congenital anomalies involving these important struc-

tures [2,3]. Many authors have thus advocated the use of

the transfrontal plane, which is a sagittal section of

the fetal head using the metopic suture as an acoustic

window, to directly visualize the facial profile and

the midline fetal cerebral structure at the same

time [4–6]. However, this scanning plane is difficult to

obtain at midgestation as most fetuses are in a horizontal

line and thus obtaining the proper plane requires

considerable ability and, occasionally, transvaginal exam-

ination [7].

One of the many advantages of three-dimensional (3D)

ultrasound is its ability to manipulate a volume and obtain

planes that are different from the one used to acquire it.

For this reason, 3D ultrasound has been used in several

Original article 133

2090-7265 & 2012 Evidence Based Women’s Health Journal DOI: 10.1097/01.EBX.0000419244.02239.14

studies to obtain reconstructed midline views of the fetal

brain from volumes acquired from axial planes [8–10].

The aim of this study is to compare the visualization of

the midline structures of the fetal brain, obtained by 3D

multiplanar reconstruction of volumes acquired from axial

plane with transfrontal 3D acquisition.

Patients and methodsIn this prospective observational study, unselected single-

ton pregnancies between 18 and 24 weeks underwent a

fetal anomaly scan by one investigator only (S.M.M.N.)

according to the practice guidelines issued by the

International Society of Ultrasound in Obstetrics and

Gynecology [11]. Patients with medical disorders, multi-

fetal pregnancies as well as pregnancies with obvious fetal

structural anomalies were excluded from the study. Healthy

patients with an apparently normal anomaly scan were

asked to participate in this study. The departmental ethics

committee approved the study and an informed consent

was obtained from all recruited patients. Ultrasound

examinations were carried out with a Voluson 730 Pro

(GE Healthcare Austria GmbH & Co., Zipf, Austria) using

a transabdominal multifrequency volume probe (RAB

4–8L, GE Healthcare, Austria GmbH & Co., Zipf, Austria).

Fetal brain volumes for the multiplanar evaluation were

obtained with the fetal head in the axial plane according to

the method suggested by Fratelli et al. [12]. Briefly, after

optimization of the two-dimensional (2D) image with

harmonics set to medium or high, with the transcerebellar

plane as the initial axial plane of acquisition, the volume

was acquired with the incident ultrasound beam making an

angle of about 451 with the cerebral midline so as to avoid

acoustic shadowing of the skull base on the posterior fossa.

The 3D volume box was adjusted to include the entire

fetal head, the acquisition quality was set to maximum, the

sweep angle was set to 60–701, and the upper reference

line was set close to the anterior parietal bone. With this

method, boxes A, B, and C of the acquired multiplanar

images always corresponded to the axial, coronal, and

sagittal planes, respectively.

For the transfrontal acquisition, we used the method

described by Vinals et al. [13]. The plane of the midsagittal

fetal facial profile was obtained with the ultrasound beam

aligned with the frontal suture so as to utilize the metopic

suture as an acoustic window. Gentle manipulation of the

fetal head by the examiner was carried out when required

to bring the fetal head into a favorable position. Similar to

Figure 1

A three-dimensional multiplanar reconstructed images of the fetal brain acquired in the axial plane. The midsagittal view of the fetal brain is shown inthe C plane image (bottom left image), showing good visualization of the midline structures the fetal brain. The axial plane (A) shown in the upper leftimage is the acquisition plane. CM, cisterna magna; CSP, cavum septum pellucidum; F, fastigium of the fourth ventricle; PF, primary cerebellar fissure;SF, secondary cerebellar fissure; V, cerebellar vermis; 3V, third ventricle.

134 Evidence Based Women’s Health Journal

the previously described method, the 2D ultrasound image

was optimized so as to visualize the corpus callosum and

the cerebellar vermis with harmonics set to medium or

high. The 3D render box was adjusted to include the

entire fetal head, the acquisition quality was set to

maximum, and the sweep angle was set at 301.

For each fetus, at most two acquisitions per 3D technique

were obtained and the volumes with the least acoustic

shadowing were stored. The stored images were later

analyzed offline by either of the two authors (S.M.M.N.

or R.A.K.) on a computer using dedicated software (4D

view; GE Healthcare). Both authors have several years of

experience in the field of fetomaternal ultrasound and are

Consultants at the Fetal Medicine Unit at Cairo

University. The volumes were manipulated so as to

display the clearest image of the sagittal midline plane of

the fetal brain, mainly the corpus callosum, the cavum

septum pellucidum, the third ventricle, the cerebellar

vermis, and the cisterna magna. We defined the

visualization of the midsagittal plane as adequate when

all the anatomical landmarks were identified clearly

according to the guidelines set by the International

Society of Ultrasound in Obstetrics and Gynecology for

carrying out the fetal neurosonogram [14]. The ability to

visualize the main landmarks of the cerebellar vermis,

mainly the fastigium of fourth ventricle and the primary

and secondary vermian cerebellar fissures, was also

compared between images obtained by the two methods.

Statistical analysis

Normally distributed parameters are presented as mean ±

SD and non-normally distributed values are presented as

median (range). Normality of data were tested for all

continuous variables using the Shapiro–Wilk test. Catego-

rical variables were compared using the w2 or Fisher’s exact

test according to cell size. Nonparametric (continuous)

variables were compared using the Mann–Whitney U-test.

A P value less than 0.05 was considered as statistically

significant. Statistical analysis was carried out using

Microsoft Excel 2010 (Microsoft Corporation, Redmond,

Washington, USA) and Arcus Quickstat Biomedical version

1.0 (Research Solutions, Cambridge, UK).

ResultsA total of 127 patients were examined in this prospective

observational study. The patients had a mean age of 30.2

years (SD, 3.8; 95% CI, 29.5–30.9), a mean BMI of 24.6

(SD, 1.9; 95% CI, 24.2–24.9), and a mean gestational age of

21.3 weeks (SD, 1.4; 95% CI, 21.0–21.5). Ninety-seven

fetuses (76.4%) were in cephalic presentation, whereas the

remaining 30 (23.6%) were in breech presentation. The

acquisition of the fetal brain in the axial plane was

successful in 122 cases (96.1%), whereas the transfrontal

acquisition was successful in 106 cases (83.4%), with a

statistically significant difference between the two meth-

ods (P = 0.002). Table 1 shows that none of the patient or

the fetal characteristics had a statistically significant effect

on the successful acquisition in either method.

Both the 3D reconstructed images from the axially acquired

images as well as the transfrontal 3D images provided

detailed visualization of the midline structures of the fetal

brain (Figs 1 and 2) in the majority of cases. Visualization of

the median plane of the fetal brain by 3D multiplanar

reconstruction was considered adequate in 99 out of the

122 (81.1%) volumes acquired from the axial plane,

whereas 94 out of the 106 (88.7%) transfrontal acquisitions

resulted in adequate midline images; the difference

between the two acquisition methods was not statistically

significant (P = 0.12). There was no statistically significant

difference between the two acquisition methods in the

visualization of the fastigium of the fourth ventricle as well

as the primary and secondary vermian fissures, which were

adequately visualized in 58/122 (47.5%) of the 3D multi-

planar reconstructed images from the axial volumes and in

62/106 (60.8%) of the transfrontally acquired volumes

(P = 0.09). Images were considered adequate if the corpus

callosum could be seen clearly as an anechoic curved

structure with an echogenic defining contour overlying and

clearly differentiated from the sonolucent cavum septum

pellucidum. The contents of the posterior fossa, mainly the

cerebellar vermis as an echogenic well-defined structure

separated from the occipital bone by the sonolucent

cisterna magna, should also be seen. In the cases with

inadequate visualization of the midline cerebral structures,

shadowing by the cranial bones was the reason for the

poor image quality, with poor demarcation of the corpus

callosum from the underlying cavum septum pellucidum

resulting in a single comma-shaped anechoic structure or an

ill-defined posterior fossa.

DiscussionOne of the many advantages of 3D ultrasound is the ability

to store the acquired volumes for later offline analysis or

even send them to an expert in a different site for analysis

by telemedicine. With the aid of specialized software such

as 4D view, which we used in the current study, the

operator can navigate through a multiplanar display of the

volume to construct virtual 2D planes, which are difficult to

Table 1 Effect of the demographic variables on successful axial and transfrontal acquisition

Axial acquisition Transfrontal acquisition

Successful Failed P value Successful Failed P value

Patient age [mean (SD)] 30.1 (3.7) 30.0 (3.6) 0.54 30.0 (3.5) 30.2 (3.7) 0.67BMI [mean (SD)] 24.2 (1.9) 24.6 (2.0) 0.51 24.1 (1.6) 24.7 (1.8) 0.39Gestational age [mean (SD)] 21.1 (1.4) 21.4 (1.6) 0.35 21.0 (1.4) 21.2 (1.5) 0.56Cephalic presentation [N (%)] 94 (96.9%) 3 (3.1%) 0.37 82 (84.5%) 15 (15.5%) 0.55Breech presentation [N (%)] 28 (93.3%) 2 (6.7%) 24 (80%) 6 (20%)

Transfrontal vs. axial 3D of fetal brain Negm and Kamel 135

acquire using conventional 2D ultrasound such as the

midsagittal plane of the fetal brain. The use of postproces-

sing software such as speckle reduction filters (SRI) can

help reduce the tissue speckle pattern and enhance the

contrast resolution of the image.

In the present study, we were able to obtain a satisfactory

axial view of the fetal head in order to acquire a 3D

multiplanar volume in the majority of cases (96.1%),

whereas the direct midsagittal view required to acquire a

3D transfrontal volume was obtained in only 83.4% of

cases (P = 0.002), mainly because of the unfavorable

position of the fetal head that could not be manipulated

to carry out a proper acquisition, whereas the axial plane,

similar to the transthalamic plane used to measure fetal

head diameters, was more readily imaged.

Although images with greater detail could have been

obtained by transvaginal 3D acquisition as shown by Malinger

and Zakut [1], our goal was to try and incorporate fetal

neurosonography into the routine transabdominal midtrimes-

teric fetal anomaly scan. Furthermore, the transvaginal route

would not have been possible in patients with a breech fetus.

Fratelli et al. [12] studied the fetal brain using 3D

ultrasound using volumes acquired in the axial view with

an angle of 451 between the incident ultrasound beam and

the cerebral midline to avoid acoustic shadowing by the

skull base on the posterior fossa, which was the method

that we adopted in the present study. Their results showed

adequate visualization of the midsagittal plane of the fetal

brain in 83–87% of examined volumes. They also found

good interobserver agreement, with a k value of 0.70.

These results are similar to those found in the present

study, in which 81.1% of successfully acquired axial volumes

resulted in adequately visualized images of the midsagittal

plane of the fetal brain. Similar results were obtained by

Correa et al. [9], who studied the role of 3D ultrasound in

the assessment of the fetal brain and found that acceptable

cerebral multiplanar images were obtained in 92% of the

cases and that the median sagittal plane could be visualized

adequately in 85% of analyzed volumes.

Visentin et al. [15] used conventional 2D ultrasound by the

transfrontal view of the fetal head in midtrimester fetuses

to visualize the fetal brain in the midsagittal plane. They

were able to successfully obtain a transfrontal view in 89%

of cases, with good visualization of the midline structures of

the fetal brain.

Vinals et al. [13] compared visualization of the midline

cerebral structures obtained by 3D multiplanar imaging

using an axial acquisition plane with transfrontal 3D

acquisition. Measurement of the corpus callosum and

Figure 2

A three-dimensional transfrontal image of the fetal brain. The midsagittal view of the fetal brain is show in the A plane image (upper left image), whichis the same as the original acquisition plane. CM, cisterna magna; CSP, cavum septum pellucidum; F, fastigium of the fourth ventricle; PF, primarycerebellar fissure; SF, secondary cerebellar fissure; V, cerebellar vermis; 3V, third ventricle.

136 Evidence Based Women’s Health Journal

cerebellar vermis and visualization of the fourth ventricle

and the main vermian fissures were also compared. In their

study, the midsagittal plane could be obtained in 88 and

87% of the 3D multiplanar and transfrontal images,

respectively. They found a good correlation between the

two methods in the measurement of the corpus callosum

and cerebellar vermis diameters. However, the images

obtained by transfrontal acquisition were clearer and more

sharply defined than those obtained by the 3D multiplanar

method. Also, the primary and secondary fissures of the

cerebellar vermis could be detected in 26 and 13%,

respectively, of multiplanar images compared to 79 and

52%, respectively, of transfrontal analyses (Po0.0001).

They attributed the poorer quality of the 3D multiplanar

images mainly to acoustic shadowing by the petrous part of

the temporal bone, which frequently obscured the fetal

brain stem, fourth ventricle, and cerebellar vermis.

The results of Vinals et al. [13] are different from our

findings, where we found no statistically significant

difference between the two methods in the image quality

of the midsagittal plane of the fetal brain (P = 0.12) or in

the visualization of the primary and secondary vermian

fissures (P = 0.09). Our better results in terms of the

midsagittal images obtained by 3D multiplanar acquisi-

tion could be attributed to our use of the acquisition

method proposed by Fratelli et al. [12], which involves

using the transcerebellar plane instead of the transtha-

lamic plane as the initial plane of acquisition and with the

incident ultrasound beam making an angle of 451 with the

cerebral midline, thus minimizing the acoustic shadowing

of the skull base on the structures of the posterior fossa.

ConclusionImages of the midsagittal plane of the fetal brain obtained

by 3D multiplanar reconstruction of volumes acquired from

the axial plane are easier to acquire than the 3D

transfrontal approach and result in comparable image

quality, with adequate visualization of the cerebral midline

as well as the main landmarks of the cerebellar vermis.

AcknowledgementsConflicts of interestThere are no conflicts of interest.

References1 Malinger G, Zakut H. The corpus callosum: normal fetal development

as shown by transvaginal sonography. Am J Roentgenol 1993; 161:1041–1043.

2 Monteagudo A. Fetal neurosonography: should it be routine? Should it bedetailed. Ultrasound Obstet Gynecol 1998; 12:1–5.

3 Pilu G, Visentin A, Valeri B. The Dandy–Walker complex and fetal sono-graphy. Ultrasound Obstet Gynecol 2000; 16:115–117.

4 Visentin A, Pilu G, Falco P, Bovicelli L. The transfrontal view: a new approachto the visualization of the fetal midline cerebral structures. J Ultrasound Med2001; 20:329–333.

5 D’Addario V, Pinto V, Di Cagno L, Pintucci A. The midsagittal view of the fetalbrain: a useful landmark in recognizing the cause of fetal cerebral ven-triculomegaly. J Perinat Med 2005; 33:423–427.

6 Achiron R, Achiron A. Development of the human fetal corpus callosum:a high-resolution, cross-sectional sonographic study. Ultrasound ObstetGynecol 2001; 18:343–347.

7 Benn PA, Gainey A, Ingardia CJ, Rodis JF, Egan JFX. The fetal cerebellarvermis: normal development as shown by transvaginal ultrasound. PrenatDiagn 2001; 21:687–692.

8 Vinals F, Munoz M, Naveas R, Shalper J, Giuliano A. The fetal cerebellarvermis: anatomy and biometric assessment using volume contrastimaging in the C-plane (VCI-C). Ultrasound Obstet Gynecol 2005; 26:622–627.

9 Correa FF, Lara C, Bellver J, Remohı J, Pellicer A, Serra V. Examination of thefetal brain by transabdominal three-dimensional ultrasound: potential forroutine neurosonographic studies. Ultrasound Obstet Gynecol 2006;27:503–508.

10 Pilu G, Segata M, Ghi T, Carletti A, Perolo A, Santini D, et al. Diagnosis ofmidline anomalies of the fetal brain with the three-dimensional median view.Ultrasound Obstet Gynecol 2006; 27:522–529.

11 Salomon LJ, Alfirevic Z, Berghella V, Bilardo C, Hernandez-Andrade E, JohnsenSL, et al. Practice guidelines for performance of the routine mid-trimester fetalultrasound scan. Ultrasound Obstet Gynecol 2011; 37:116–126.

12 Fratelli N, Taddei F, Prefumo F, Franceschetti L, Farina G, Frusca T. Inter-observer reproducibility of transabdominal 3-dimensional sonography of thefetal brain. J Ultrasound Med 2009; 28:1009–1013.

13 Vinals F, Munoz M, Naveas R, Giuliano A. Transfrontal three-dimensionalvisualization of midline cerebral structures. Ultrasound Obstet Gynecol2007; 30:162–168.

14 Paladini D, Malinger G, Monteagudo A, Pilu G, Timor-Tritsch I, Toi A. Sono-graphic examination of the fetal central nervous system: guidelines for per-forming the ‘basic examination’ and the ‘fetal neurosonogram’. UltrasoundObstet Gynecol 2007; 29:109–116.

15 Visentin A, Pilu G, Falco P, Bovicelli L. The transfrontal view: a new approachto the visualization of the fetal midline cerebral structures. J Ultrasound Med2001; 20:329–333.

Transfrontal vs. axial 3D of fetal brain Negm and Kamel 137

Glycosylated hemoglobin level at 34 weeks in insulin-controlled

diabetic pregnancies, relation to fetal outcomeHoda Abdelaala, Mohamed A. Kasema, Abdelmeguid Alia

and Alaa M. Abd Elazemb

aDepartment of Obstetrics & Gynecology, Faculty ofMedicine, Cairo University, Cairo and bShebeen ElkomTeaching Hospital, Egypt

Correspondence to Hoda Abdelaal, Department ofObstetrics and Gynecology, Faculty of Medicine, CairoUniversity, Cairo, EgyptTel: + 012 2284 8993;e-mail: hodatolba@gmail.com

Received 10 November 2011Accepted 12 December 2011

Evidence Based Women’s Health Journal

2012, 2:138–141

Objective

To determine the role of glycosylated hemoglobin at 34 weeks’ gestation and onwards

in the prediction of fetal outcome in insulin-controlled diabetic pregnancies.

Design

Cohort study.

Setting

Kasr Al-Aini Hospital.

Participants and methods

A total of 50 singleton pregnant women on insulin treatment with gestational age

between 34 and 36 weeks were included. All women underwent ultrasound

fetometery, and glycosylated hemoglobin level at 34 weeks, amniotic fluid index,

placental maturation, and biophysical profile were determined. Doppler ultrasound was

also carried out. Follow-up was carried out for all women until delivery and neonates

were examined for body weight, APGAR scoring at 1 and 5 min, and blood glucose

level (adverse if <45 mg/dl).

Results

There was a statistically significant correlation between HbA1c and BMI, amniotic fluid

index, and neonatal outcomes. HbA1c of 7 or higher was found to be a cutoff value for

the prediction of prematurity, with area under curve of 91.7% (Fig. 2).

Conclusion

HbA1c may be a useful marker for prematurity in pregnant diabetic women and may

correlate with fetal outcome. For the antenatal care of diabetic mothers, it is

recommended is to maintain HbA1c less than 7% decrease fetal adverse outcome.

Keywords:

APGAR, glycosylated hemoglobin, maturity, pregnancy

Evid Based Women Health J 2:138–141& 2012 Evidence Based Women’s Health Journal2090-7265

IntroductionGestational diabetes (GDM) represents a significant health

risk for the mother and offspring during pregnancy, delivery,

and throughout life. Offspring of women with GDM are at

an increased risk for neonatal complications, in particular,

macrosomia, large for gestational age, respiratory distress,

prematurity, hypoglycemia, polyhydramnios, and death [1].

These offspring are also at a high risk for obesity, insulin

resistance, and type 2 diabetes (T2DM) [2]. Thus, the

diagnosis and appropriate management of GDM has the

potential to markedly reduce neonatal and maternal

morbidity and the burden of T2DM.

Glycosylated hemoglobin, as measured by hemoglobin A1C

(HbA1c), is used as a marker of glycemic control in patients

with type 1 diabetes (T1DM) and T2DM [2]. Notably, the

American Diabetic Association has recently recommended

that A1C higher than 6.5% may be used as a diagnostic

measure for T2DM and that women with A1C higher than

6.5% at their first prenatal visit be diagnosed with T2DM

rather than GDM [3]. In the nonpregnant diabetic

population, sustained high A1C is associated with increased

complications of diabetes [4]. Therefore, optimal A1C levels

in pregnancy should be defined by the level of increased risk

for adverse pregnancy outcomes. Elevated A1C may be an

attractive option for the identification of pregnant women at

a high risk for GDM, postpartum T2DM, and macrosomia or

delivering an large for gestational age infant. The aim of the

present study is to determine a role of glycosylated

hemoglobin at 34 weeks’ gestation and onwards in the

prediction of fetal outcome in insulin-controlled diabetic

pregnancies.

Patients and methodsThis prospective cohort study included 50 pregnant women

recruited from the Obstetric Outpatient Clinic and Depart-

ment of Obstetrics & Gynecology at Kasr Al-Aini Hospital.

We included singleton pregnant women on insulin treatment

with gestational age at the time of inclusion between 34–36

weeks. We excluded women on diet control or those with

medical disorders not related to diabetes such as Systemic

Lupus Erythematous, thyroid disorders, hypertension, and

preeclampsia.

138 Original article

2090-7265 & 2012 Evidence Based Women’s Health Journal DOI: 10.1097/01.EBX.0000419243.94615.e9

All women underwent ultrasound fetometery, and glycosy-

lated hemoglobin level at 34 weeks, amniotic fluid index,

placental maturation, and biophysical profile were deter-

mined. Doppler ultrasound was also carried out. Follow-up

was carried out for all women until delivery and neonates

were examined for body weight, APGAR scoring at 1 and

5 min, adverse neonatal outcome (when the APGAR score is

<7 at 5 min), death either intrauterine or early after birth,

neonatal weight more than 4.5 kg, presence of signs of respi-

ratory distress syndrome, and blood glucose less than 45 mg/dl.

Doppler study

Uterine, umbilical, and middle cerebral artery Doppler was

carried out by measurement of Doppler indices. The

pulsatility index (PI) was used as the test Doppler

parameter because PI describes the shape of the velocity

wave form much better. (a) Umbilical artery Doppler: All

patients were placed in a semirecumbent position with a left

lateral tilt, and then the uterine content was scanned to

select an area of amniotic cavity with several loops of cord.

When the screen showed at least three successive waveforms

of similar height, the image was frozen and the Doppler

umbilical artery pulsatility index was estimated. (b) Uterineartery Doppler: Doppler examination of the uterine arteries

was performed at the same time. Color Doppler was used to

identify the apparent crossing of the uterine and iliac vessels.

Uterine artery velocimetry was recorded just cranial of the

vessel ‘crossing’. Three even subsequent blood flow velocity

waveforms were used to calculate PI and to analyze the

presence or absence of an early diastolic notch.

Glycosylated hemoglobin

Whole blood was collected by a standard procedure

(1 cm). Heparin or edeta was used as an anticoagulant.

HbA1c was stable for 7 days at 2 – 81C. After preparing

the hemosylate, where the labile fraction is eliminated,

hemoglobins are retained by cation exchange resin,

HbA1c is specifically eluted after washing the hemoglo-

bin Aa + b fraction, and quantified by direct photometric

reading at 415 nm. Reference values were as follows:

(1) HbA1co6.3%: very good glycemic control.

(2) HbA1c between 6.3 and 7.1%: good glycemic control.

(3) HbA1c between 7.1 and 9%: poor glycemic control.

(4) HbA1c49%: poor glycemic control.

Statistical analysis

The data were coded, entered, and processed on a computer

using SPSS (version 16, IBM, New York, USA). Qualitative

data are presented as number and percentages whereas

quantitative data are presented as mean and SD. Student’s

t-test was used to assess the statistical significance of

the difference between two population means in a study

involving independent samples. A P-value was considered

significant if less than 0.05.

The Pearson correlation coefficient was used to determine

the relation between the quantitative parameters. Logistic

regression analysis was used to assess the effect of

quantitative parameter on an outcome. ROC curve was used

to assess the sensitivity and specificity of factor in predicting

the outcome.

ResultsThe present study included women ranging in age from 19 to

35 years, mean ± SD (26.94 ± 4.497); the patient’s BMI

ranged from 21 to 33, mean ± SD (25.04 ± 3.110). Table 1

shows other characteristics of our participants. No statistically

significant relation was found between HbA1c and age of the

studied patients, whereas there was a statistically significant

correlation between HbA1c and BMI, amniotic fluid index,

and neonatal outcomes (Table 2). HbA1c of 7 or higher was

found to be a cutoff value for prediction of prematurity with

an area under curve of 91.7% (Fig. 1); also, this cutoff value

was used for the prediction of polyhydramnios, macrosomia,

neonatal hypoglycemia (Fig. 2), respiratory distress syndrome,

and cesaerean section rate (with area under the curve of 77.5,

72.6, 98.3, 98.3, and 87.1, respectively).

DiscussionIn the USA, the prevalence of GDM may range from 1 to

14% of pregnancies [3]. GDM generally has few symptoms

and it is most commonly diagnosed by screening during

pregnancy. Diagnostic tests detect inappropriately high

levels of glucose in blood samples [5].

Glycosylated hemoglobin, as measured by HbA1c, is used as

a marker of glycemic control in patients with T1DM and

Table 1 Descriptive statistics for the parameters studied

Minimum Maximum Mean SD

Insulin dose 30 90 41.80 13.92FW (kg) 1.6 4.9 3.09 0.81AFI 6 28 17.40 5.47BPD 4 8 7.40 1.01UTA PI 0.6 1.2 0.78 0.14MCA PI 1.4 1.95 1.81 0.17UMA PI 0.81 1.25 0.92 0.13HbA1c 5.8 13 8.17 1.98FBS 70 135 96.46 19.762hPPS 120 280 186.08 59.72APGAR1 4 8 6.65 1.41APGAR5 5 10 7.87 1.84Birth weight (kg) 2 5.2 3.30 0.74Glucose 25 85 49.06 20.36Delivery age (weeks) 35 38 36.69 1.19

2hPPS, two hours post prandial blood sugar; AFI, amniotic fluid index;BPD, biparietal diameter; FBS, fasting blood sugar; FW, fetal weight;MCA, middle cerebral artery; PI, pulsatility index; UMA, umbilical artery;UTA, uterine artery.

Table 2 Correlation between HbA1c with age, BMI, AFI, BMI

fetal weight, birth weight, neonatal glucose, and APGAR5

HbA1c

R P-value

Age 0.244 0.088BMI 0.432** 0.002AFI 0.301* 0.034FW (kg) 0.426** 0.003Birth weight (kg) 0.308* 0.033Neonatal glucose – 0.552** 0.000APGAR5 – 0.689** 0.000

AFI, amniotic fluid index; FW, fetal weight.*Correlation is significant at the 0.05 level (two-tailed).**Correlation is significant at the 0.01 level (two-tailed).

Glycosylated Hb and fetal outcome Aal et al. 139

T2DM [2]. In the present study, we aimed to evaluate its

use as a marker in gestational diabetic women at 34 weeks

and follow them up. Interestingly, the present study

observed a positive relation between glycosylated hemoglo-

bin and adverse fetal outcome; that is, when HbA1c more

than 7% at 34 weeks, there was increase in the frequency of

adverse fetal outcomes including respiratory distress syn-

drome, macrosomia, polyhydramnios, prematurity, and neo-

natal hypoglycemia [1].

The results of the current study showed that the

sensitivity of HbA1c more than 7% for predicting

polyhydramnios was 100% and the specificity was

57.14%. The incidence of large for gestational age was

38% and this was in agreement with the result of Yang

et al. [6], who reported a 40% in Pregnancy Outcomes of

Births in Nova Scotia from 1988 to 2002 in Women with

and Without Pregestational Diabetes.

The study also found a highly statistically significant

positive relation between HbA1c with fetal weight and a

significant positive relation with birth weight (P = 0.003

and 0.033, respectively), which was in agreement with

Katon et al [7], who reported that the association of A1C

and birth weight was stronger when antepartum A1C was

measured after a diagnosis of GDM (at 34 weeks).

The results of the current study showed that the sensitivity

of HbA1c more than 7% for predicting macrosomia was 82%

and the specificity was 63%. In this study, the incidence of

hypoglycemia was 40% and there was a highly statistically

significant relation (Po0.005) between HbA1c and

neonatal glucose level, yielding an area under the curve

of 98.3. This was in agreement with Arumugam and Abdul

Majeed [8], who reported that HbA1c levels in late

pregnancy are good predictors of hypoglycemia in the

newborn, yielding an area under the curve of 0.99. In

simple terms, what this means is that if there were two

babies who were randomly selected, one with hypoglycemia

and one without, the probability that the hypoglycemic

neonate would have shown an abnormally high maternal

HbA1c would be around 99% [9]. In this study, there was

no statistically significant relation between HbA1c and

Doppler and this in agreement with Pietryga et al. [10], who

concluded that there was no correlation between long-term

maternal glycemic control (HbA1c) and changes in blood

flow velocity in placental circulation in pregnancy compli-

cated by diabetes mellitus.

ConclusionHbA1c may be a useful marker for prematurity in

pregnant diabetic women. For the antenatal care of

diabetic mothers, we recommend maintaining HbA1c less

than 7% to decrease fetal adverse outcome. For further

confirmation of the above results, we need to study a

larger number of patients.

AcknowledgementsConflicts of interestThere are no conflicts of interest.

References1 Vohr BR, Boney CM. Gestational diabetes: the forerunner for the develop-

ment of maternal and childhood obesity and metabolic syndrome? J MaternFetal Neonatal Med 2008; 21:149–157.

2 Goldstein DE, Little RR, Lorenz RA, Malone JI, Nathan D, Peterson CM,Sacks DB. Tests of glycemia in diabetes. Diabetes Care 2004; 27:1761–1773.

Figure 1

HbA1C

0 20 40 60 80 100

0

20

40

60

80

100

100-Specificity

Sens

itivi

ty

ROC curve for the prediction of prematurity with a cutoff point of 7 andarea under curve of 91.7.

Figure 2

HbA1C

0 20 40 60 80 100

0

20

40

60

80

100

100-Specificity

Sens

itivi

ty

ROC curve for the prediction of neonatal hypoglycemia with a cutoffpoint of 7 and area under curve of 98.3.

140 Evidence Based Women’s Health Journal

3 American Diabetes Association. Diagnosis and classification of diabetesmellitus. Diabetes Care 2011; 34 (Suppl 1): S62–S69.

4 Nathan DM, Kuenen J, Borg R, Zheng H, Schoenfeld D, Heine RJ. Translatingthe A1C assay into estimated average glucose values. Diabetes Care 2008;31:1473–1478.

5 Crowther CA, Hiller JE, Moss JR, McPhee AJ, Jeffries WS, Robinson JS.Effect of treatment of gestational diabetes mellitus on pregnancy outcomes.N Engl J Med 2005; 352:2477–2486.

6 Yang J, Cummings EA, O’Connell C, Jangaard K. Fetal and neonataloutcomes of diabetic pregnancies. Obstet Gynecol 2006; 108 (3 I):644–650.

7 Katon J, Williams MA, Reiber G, Miller E. Antepartum A1C, maternal diabetesoutcomes, and selected offspring outcomes: an epidemiological review.Paediatr Perinat Epidemiol 2011; 25:265–276.

8 Arumugam K, Abdul Majeed N. Glycated haemoglobin is a good predictor ofneonatal hypoglycaemia in pregnancies complicated by diabetes. Malays JPathol 2011; 33:21–24.

9 Obuchowski NA. Receiver operating characteristic curves and their use inradiology. Radiology 2003; 229:3–8.

10 Pietryga M, Brazert J, Wender-Ozegowska E, Dubiel M, Gudmundsson S.Placental Doppler velocimetry in gestational diabetes mellitus. J Perinat Med2006; 34:108–110.

Glycosylated Hb and fetal outcome Aal et al. 141

Evaluation of the atherogenic role of lipoproteins and oxidized

low-density lipoprotein in pre-eclampsiaYasser H. Shaabana, Fatma Aletebib, Manal Helald and Walid Saberc

Departments of aClinical and Chemical Pathology,bInternal Medicine, cObstetrics and Gynecology,Faculty of Medicine, Cairo University, Cairo, Egypt anddDepartment of Obstetrics and Gynecology, KingAbdulAziz University, Saudi Arabia

Correspondence to Yasser H. Shaaban, MD,Department of Chemical Pathology, Faculty ofMedicine, Cairo University, Cairo, EgyptE-mail: yassershaaban@maktoob.com

Received 16 March 2012Accepted 14 April 2012

Evidence Based Women’s Health Journal

2012, 2:142–146

Objectives

The aim of this study was to evaluate the levels of lipid parameters and oxidized

low-density lipoprotein (ox-LDL) and compare them in normal and pre-eclamptic

pregnant women as well as to determine the relative risk of developing pregnancy-

induced hypertension disorders in patients in whom ox-LDL levels are elevated.

Participants and methods

Singleton pregnant women (18–38 years old) were recruited and divided into two

groups: 26 women with pre-eclampsia and 21 women with normal uncomplicated

pregnancies. Blood and urine samples were taken from all the participants; different

lipid parameters including triglycerides, cholesterol, high-density lipoprotein

cholesterol, and low-density lipoprotein cholesterol were determined. In addition,

urine protein and ox-LDL were assessed.

Results

The two groups studied were matched in terms of maternal age, BMI, and gestational

age. Pre-eclamptic patients had significantly higher blood pressure compared with the

women with normal pregnancies (control). Of the measured lipid parameters,

significantly higher values of triglycerides, very low-density lipoprotein-C, and ox-LDL

were found in the women in the pre-eclampsia group compared with the control group

(P = 0.003 and 0.049, respectively). None of the measured lipid parameters showed

a significant association when correlated to ox-LDL. The calculated cut-off for ox-LDL

in our study was 36 ng/ml. Patients with high ox-LDL levels during pregnancy had 1.7

times higher risk for developing pre-eclampsia than women with normal pregnancies.

Conclusion

Dyslipidemia and oxidative stress may play an important role in the pathogenesis of

pregnancy-induced hypertensive disorders. Thus, our findings may be relevant for

understanding the pathophysiology of pre-eclampsia and showed that there is a need

for further research work to establish the role of lipid-modifying regimens, lifestyle

modifications, and antioxidant therapy to reduce the risk of developing such disorders.

Keywords:

dyslipedemia, lipids, oxidized low-density lipoprotein, pre-eclampsia, pregnancy

Evid Based Women Health J 2:142–146& 2012 Evidence Based Women’s Health Journal2090-7265

IntroductionPre-eclampsia is a common multisystem pregnancy dis-

order in which the diagnosis is made on the basis of the

presence of hypertension and proteinuria, affecting 3–5%

of all pregnant women. Pre-eclampsia occurs de novo in

late pregnancy after the 20th week of gestation, and severe

pre-eclampsia may be a serious threat to the mother

and the fetus. This disorder is a major cause of prenatal

and maternal morbidity and mortality worldwide [1].

The etiology of pre-eclampsia is still unknown and mul-

tiple factors have been implicated in its pathogenesis.

Metabolic syndrome, especially insulin resistance, and

pre-eclampsia have been well examined in previous re-

search [2,3]. Insulin resistance is common in pregnancy,

as are aberrant lipid metabolism, higher fasting glucose

level, and increased insulin secretion. These metabolic

changes are likely to have evolved to fulfill the metabolic

demands of a growing fetus, but these changes are aug-

mented in pre-eclampsia [4].

Previous studies have shown that in pre-eclampsia,

plasma lipids increase markedly beyond the levels found

in normal pregnancies [5,6]. It has been proposed that

such lipid changes may play a role in the endothelial

damage characteristic of pre-eclampsia. Given the phy-

siological role of gestational hyperlipidemia in supplying

both cholesterol and triglycerides to the rapidly develop-

ing fetus, it is conceivable that pregnancies complicated

by intrauterine growth retardation show abnormal lipo-

protein metabolism in an attempt to compensate for the

placental insufficiency. This mechanism has been pro-

posed to be responsible for the higher triglyceride concen-

trations observed in women with pre-eclampsia [7].

Diffuse vascular endothelial dysfunction, secondary to

oxidative stress, is an important pathological characteristic

142 Original article

2090-7265 & 2012 Evidence Based Women’s Health Journal DOI: 10.1097/01.EBX.0000415828.84622.0c

of pre-eclampsia. Oxidative conversion of low-density

lipoproteins (LDLs) into oxidized-LDL (ox-LDL) is

considered an important step in the transformation of

macrophages into lipid-laden foam cells destined to

develop into early atherosclerotic-like lesions [8].

Small dense LDL is also increased in the plasma of

women with pre-eclampsia. Small dense LDLs are more

susceptible to oxidation, resulting in the generation of ox-

LDL, which can bind to the lectin-like oxidized-LDL

receptor-1 (LOX-1) on endothelial cells. LOX-1 is a type

II membrane protein cell surface receptor found on

endothelial cells, vascular smooth muscle cells, and

monocyte macrophages. LOX-1 is expressed in athero-

sclerotic lesions in humans and has also been shown to be

elevated in hypertensive rats [9]. LOX-1 is responsible

for the binding, uptake, and degradation of ox-LDL. One

recent study has reported elevated LOX-1 expression in

the placenta of women with pre-eclampsia. However, the

expression, regulation, and significance of LOX-1 in the

maternal systemic vasculature of pre-eclampsia remain

unknown [10].

The aims of the present study are to measure the serum

levels of cholesterol, triglyceride, high-density lipoprotein

(HDL), LDL, and ox-LDL in pre-eclamptic pregnancy

and to compare them with those in normal pregnancy.

Participants and methodsThe present study was conducted from December 2009

to January 2011 among women with pre-eclampsia ac-

cording to the 10 current American College of Obste-

tricians and Gynecologists guidelines (American College

of Obstetricians and Gynecologists, 1996), which defined

pre-eclampsia as sustained pregnancy-induced hyperten-

sion with proteinuria. Pre-eclampsia was defined as per-

sistent (Z6 h) blood pressure of at least 140/90 mmHg

occurring after 20 weeks of gestation. Proteinuria was

defined as a urine protein concentration of 30 mg/dl or

higher (or 1 + on a urine dipstick or higher) in at least

two random specimens collected at least 4 h apart.

Two groups of 26 women with pre-eclampsia and 21 con-

trols with uncomplicated pregnancies matched for age,

gestational age, weight, and BMI were recruited. We ex-

cluded women with pregnancies associated with diabetes

mellitus, hepatic, renal, metabolic, cardiovascular dis-

eases, chronic or transient hypertension, thyrotoxicosis,

and hemophilia from our study. Also, pregnant women

with a previous history of gestational diabetes or those on

chronic medications were excluded from the study.

Data and sample collection

During participants’ antenatal care visits, we administered

structured interview questionnaires to collect information

on maternal sociodemographic, medical, reproductive, and

lifestyle characteristics. All interviews were conducted in

Arabic. BMI and the measure of adiposity were calculated

as weight (kg) divided by height (m2). Nonfasting blood

samples were collected in dry 10 ml vacutainer tubes

during the visit. Specimens were centrifuged at 3000g;

serum was separated and frozen at – 701C until analysis.

Analytical methods

Serum cholesterol and triglyceride concentrations were

measured enzymatically using assays on a Hitachi 917

chemistry analyzer (Roche Diagnostics, Berlin, Germany)

with a colorimetric procedure. HDL-C was assayed using

a homogenous enzymatic assay on a Hitachi 917

chemistry analyzer (Roche Diagnostics). LDL-cholesterol

(LDL-C) was calculated using the Friedewald formu-

la [11]. From the previous assays, the HDL/LDL and

cholesterol/HDL ratios were calculated. Plasma ox-LDL

concentrations were measured using commercial ELISA

assay kits (Immundiagnostik AG, Bensheim, Germany).

Statistical analysis

All analyses were performed using SPSS statistical soft-

ware version 10 (IBM, Chicago, Illinois, USA). All contin-

uous variables were presented as mean ± SD. All reported

confidence intervals were calculated at the 95% level.

Differences between groups were evaluated using Stu-

dent’s unpaired t-test and, when a variable was not

normally distributed, the Mann–Whitney U-test and

Wilcoxon W-tests were used. The relationship between

the variable was explored using the Pearson correlation

test. The association of ox-LDL with the maternal plasma

lipid profile was estimated using Spearman’s correlation

coefficients. The cutoff for ox-LDL was estimated using

the receiver operating characteristic (ROC) curve, area

under the curve, and the best sensitivity/specificity was

determined to calculate the cutoff. Logistic regression

procedures were used to calculate odds ratios and 95%

confidence intervals (CIs). To estimate the relative

association between pre-eclampsia and levels of maternal

ox-LDL, we categorized each study group into Healthy and

Cases on the basis of the level of the ox-LDL cut-off value

determined by the ROC curve.

ResultsThe demographic and clinical characteristics of pre-

eclampsia cases and controls are shown in Table 1. The

two groups studied were matched for maternal age

(18–38 years old), BMI, and gestational age at sampling.

Systolic and diastolic blood pressures were significantly

higher in women with pre-eclampsia (mean 163 and 105,

respectively) than those in the control group (mean 122

and 77, respectively) (Po0.001). Infants of pre-eclamp-

tic women had significantly lower Apgar scores for their

pregnancy outcomes when compared with those of

women with normal pregnancies.

Pregnant women with a complication of pre-eclampsia

had significantly higher mean triglyceride and very low-

density lipoprotein-C (VLDL-C) concentrations, relative

to the control group (P = 0.003 and 0.049, respectively).

In contrast, there were no significant differences in the

mean concentrations of total cholesterol, HDL-C and

LDL-C, and risk ratios between the cases and the

Oxidized-LDL in pre-eclampsia Shaaban et al. 143

controls (P40.05) as determined using the t-test as

shown in Table 2. Significantly higher ox-LDL values

were found in the pre-eclampsia group compared with

the control group using the Mann–Whitney U-test

(P = 0.044).

We next examined the correlations between maternal

plasma ox-LDL concentrations and lipid profile for pre-

eclampsia cases and controls Table 3. Ox-LDL concen-

trations were nonsignificantly positively correlated

with cholesterol (r = 0.043) and triglycerides (r =

0.072) (P40.05) in both pre-eclampsia cases and the

controls. Ox-LDL concentrations were nonsignificantly

inversely correlated with HDL-C (r = – 2.26) and LDL-C

(r = – 0.023) (P40.05). The calculated cut-off value for

ox-LDL was 36 ng/ml.

Ox-LDL levels were higher in women with pre-eclampsia

than those of the women in the control group. The overall

risk odds ratio was 1.750 (CI ranging from 0.364 to 8.424);

the ratio was 1.375 for the cases, with CI ranging from

0.547 to 3.459, and 0.786 for the healthy controls, with CI

ranging from 0.405 to 1.525 (Table 4 and Fig. 1).

DiscussionExperimental models indicated that increased ox-LDL is

capable of inducing many endothelial changes of poten-

tial relevance to pre-eclampsia [12]. However, it remains

unclear whether circulating ox-LDL is increased in pre-

eclampsia. Increased autoantibodies to an epitope of ox-

LDL, which is considered an indirect marker of ox-LDL,

have been found in women with pre-eclampsia relative to

those with a normal pregnancy [13], although negative

reports also exist [14,15]. The role and status of serum

lipids and ox-LDL in pregnant women are still being

discussed. The aims of the present study are to measure

the serum levels of cholesterol, triglyceride, HDL, LDL,

and ox-LDL in women with pre-eclampsia and to com-

pare these with those of women with normal pregnancies.

In the present study, there was no significant difference

(P40.05) in the serum levels of cholesterol, HDL-C,

LDL-C, and risk ratios between women with pre-

eclamptic pregnancies and those with normal pregnancies.

Our results are in agreement with those of some studies

[6,16], but not with other research works that have

reported significant increases in cholesterol, HDL-C, and

LDL-C [17,18]. These findings may be because of

differences in races and nutrition status. It should be

noted that all blood samples were taken from preeclamp-

tic women at their entry to Obstetrics and Gynecology

department to avoid any possible effect of medications on

the results of our study.

Significantly higher values (Po0.05) triglycerides and

VLDL-C were found in the pre-eclampsia group when

compared with normal pregnant women. Some previous

studies showed that the most significant damage in the

lipid profile in normal pregnancy is serum hypertriglycer-

idemia, which may be as high as two- to three-fold in the

third trimester compared with the levels in nonpregnant

women. The principal modulator of this hypertriglycer-

idemia is estrogen as pregnancy is associated with hyper-

estrogenemia. Estrogen induces hepatic biosynthesis of

endogenous triglycerides, which is carried by VLDL-C.

This process may be modulated by hyperinsulinism found

in pregnancy. Serum triglyceride and VLDL-C concentra-

tions were also found to increase significantly more in

toxemia of pregnancy in our study, which is in agreement

with the findings of many workers [17,19]. The above-

mentioned associations along with increased endothelial

triglyceride accumulation may result in endothelial cell

dysfunction in gestosis. Increased triglycerides, found in

pregnancy-induced hypertension, is likely to be deposited

in predisposed vessels, such as the uterine spiral arteries,

and contributes to endothelial dysfunction, both directly

and indirectly, through the generation of small, dense

LDL. Moreover, this hypertriglyceridemia may be asso-

ciated with hypercoagulability [20].

The values of ox-LDL in our study were significantly

increased (Po0.05) in pre-eclamptic women than in

normal pregnancies. Oxidation of LDL is commonly

believed to be essentially involved in the pathogenesis of

atherosclerosis, which is known to be the primary cause

of cardiovascular disease and stroke. Recent studies

have reported that ox-LDL impairs trophoblast invasion

in vitro, suggesting its role in the regulation of uterine

spiral artery remodeling in vivo [21]. Reduced trophoblast

invasion and atherosclerotic placental changes, described

as ‘acute atherosis’, are hallmarks in the pathogenesis of

pre-eclampsia and intrauterine growth retardation. Some

efforts have been made to evaluate ox-LDL levels in PE,

which have yielded inconsistent data: although in two

studies the ox-LDL concentration was found to be

significantly increased in comparison with normal healthy

pregnant controls [22,23], one research group did not find

a significant difference, but reported a 2.9-fold increased

risk for PE in women who had elevated ox-LDL

levels [24]. Only the group of Raijmakers et al. [25]

reported lower ox-LDL concentrations in PE.

None of the lipid and lipoprotein parameters showed

a significant association (P40.05) with ox-LDL, while

total cholesterol, triglycerides, and VLDL-C showed posi-

tive correlations with ox-LDL, maternal serum LDL-C

and HDL-C showed negative correlations. In one re-

search work [24], lipid parameters showed significant

Table 1 Demographic and clinical data of the studied groups

Mean ± SD

Items

Pre-eclampsia

groupControlgroup

P value(t-test)

Maternal age (years) 34.2 ± 2.3 32.1 ± 2.1 0.78BMI (kg/m2)Gestational (age/weeks) 27.2 ± 5.4Apgar score (min)

1 5.1 ± 1.2 NA5 7.8 ± 1.4 NA

Systolic BP (mm/Hg) 163.7 ± 13.8 122 ± 3.5 o0.001Diastolic BP (mm/Hg) 105 ± 10 77 ± 4.3 o0.001

BP, blood pressure.

144 Evidence Based Women’s Health Journal

associations with ox-LDL; this controversy results with

our findings may have been because of difference in

number of participants. The ROC curve constructed for

the estimation of the cut-off for ox-LDL in pre-eclampsia

was 36 ng/ml in our study.

In the present study, it was found that women with

increased ox-LDL concentrations had a 1.7-fold increased

risk of pre-eclampsia, compared with women who had

lower ox-LDL levels. Our results are in agreement with

those of some research works [24,26] and some results of

autoantibodies to ox-LDL (an indirect marker of in-vivo

ox-LDL) from cross-sectional case–control studies [13,27].

However, there has been only one published negative

report on maternal plasma ox-LDL concentrations in

relation to the risk of developing pre-eclampsia [25]. In

that study, women with pre-eclampsia had lower ox-LDL

concentrations than gestational-age-matched normoten-

sive pregnant controls. The authors attributed this inverse

association to the higher levels of autoantibodies to

ox-LDL, which might be responsible for the rapid

clearance of ox-LDL.

ConclusionCompared with normal pregnant women with similar

gestational age and BMI, with no significant difference in

maternal age, we observed that pre-eclamptic women had

higher blood pressure and increased triglycerides, VLDL-

C, and ox-LDL during pregnancy. Increased ox-LDL in

pregnancy is associated with a higher risk of devolving

pregnancy-induced hypertensive diseases. Our results

confirm the role of dyslipedemia and oxidative stress in

the pathogenesis of pregnancy-induced hypertensive

disorders. There is a need for further research work to

establish the role of diet, lifestyle modifications, and

antioxidant therapy to reduce the risk of development of

such disorders.

Table 3 Correlations of oxidized low-density lipoprotein with

different lipid parameters in both the groups (Spearman’s

correlation)

Oxidized-LDLn = 27

Parameters Correlation coefficient r P value

Total cholesterol 0.043 0.830Triglycerides 0.072 0.719HDL-C – 2.26 0.258LDL-C – 0.023 0.908

C, cholesterol; HDL, high-density lipoprotein; LDL, low-densitylipoprotein.

Table 4 Risk estimation by odds ratios and 95% confidence

interval of pre-eclampsia according to oxidized low-density

lipoprotein subgroups

95% confidence interval

Groups Odds ratio Lower Upper

Pre-eclampsia/control 1.750 0.364 8.424For cohort odds Cases

(pre-eclampsia/control)1.375 0.547 3.459

For cohort odds Healthy(pre-eclampsia/control)

0.786 0.405 1.525

Figure 1

Receiver operating characteristic (ROC) curve for the oxidized low-density lipoprotein. Area under the curve = 0.267, 95% confidenceinterval = 0.05–0.484, Cutoff (ox-LDL) = 36 ng/ml (40% sensitivity and50% specificity).

Table 2 Lipid parameters and oxidized low-density lipoprotein values of the studied groups

Mean ± SD

Items Pre-eclampsia group Control group P value (t-test)

Total cholesterol (mg/dl) 179.25 ± 55.243 186.64 ± 45.502 0.718Triglycerides (mg/dl) 207.69 ± 84.467 107.64 ± 69.618 0.003VLDL (mg/dl) 41.538 ± 16.893 21.528 ± 13.924 0.061HDL-C (mg/dl) 44.31 ± 9.555 46.45 ± 5.484 0.510LDL-C (mg/dl) 89.56 ± 51.673 118.64 ± 40.562 0.131Cholesterol/HDL ratio 4.044 ± 5.782 4.018 ± 8.297 0.621LDL/HDL ratio 2.02 ± 5.408 2.554 ± 7.396 0.172Ox-LDL (ng/ml) 12.50b ± 198.654 0.00b ± 108.838 0.044a

C, cholesterol; HDL, high-density lipoprotein; LDL, low-density lipoprotein; ox-LDL, oxidized-LDL; VLDL, very low-density lipoprotein.aMann–Whitney U-test.bMedian concentrations.

Oxidized-LDL in pre-eclampsia Shaaban et al. 145

AcknowledgementsConflicts of interestThere are no conflicts of interest.

References1 Bonney EA. Preeclampsia: a view through the danger model. J Reprod

Immunol 2007; 76:68–74.

2 Villa PM, Laivuori H, Kajantie E, Kaaja R. Free fatty acid profiles in pre-eclampsia. Prostaglandins Leukot Essent Fatty Acids 2009; 81:17–21.

3 Soonthornpun K, Soonthornpun S, Wannaro P, Setasuban W, Thamprasit A.Insulin resistance in women with a history of severe pre-eclampsia. J ObstetGynaecol Res 2009; 35:55–59.

4 Bodnar LM, Ness RB, Markovic N, Roberts JM. The risk of preeclampsiarises with increasing prepregnancy body mass index. Ann Epidemiol 2005;15:475–482.

5 Sattar N, Bendomir A, Berry C, Shepherd J, Greer IA, Packard CJ. Lipo-protein subfraction concentrations in preeclampsia: pathogenic parallels toatherosclerosis. Obstet Gynecol 1997; 89:403–408.

6 Kaaja R, Tikkanen MJ, Viinikka L, Ylikorkala O. Serum lipoproteins, insulin andurinary prostanoid metabolites in normal and hypertensive pregnant women.Obstet Gynecol 1995; 85:353–356.

7 Sattar R, Gaw A, Packard CJ, Greer IA. Potential pathogenic roles ofaberrant lipoprotein and fatty acid metabolism in pre-eclampsia. Br J ObstetGynaecol 1996; 103:614–620.

8 Sanchez SE, Williams MA, Muy Mvera M, Qiu C, Vadachkoria S, Bazul V.A case-control study of oxidized low density lipoproteins andpreeclampsia risk. Gynecol Endocrinol 2005; 21:193–199.

9 Lee H, Park H, Kim YJ, Kim HJ, Ahn YM, Park B, et al. Expression of lectin-likeoxidized low-density lipoprotein receptor-1 (LOX-1) in human preeclampticplacenta: possible implications in the process of trophoblast apoptosis.Placenta 2005; 26:226–233.

10 Sankaralingam S, Xu Y, Sawamura T, Davidge ST. Increased lectin-like oxidizedlow-density lipoprotein receptor-1 expression in the maternal vasculature of wo-men with preeclampsia role for peroxynitrite. Hypertension 2009; 53:270–277.

11 Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration oflow-density lipoprotein cholesterol in plasma, without use of the preparativeultracentrifuge. Clin Chem 1972; 18:499–502.

12 Simon BC, Cunningham LD, Cohen RA. Oxidized low density lipoproteinscause contraction and inhibit endothelium-dependent relaxation in the pigcoronary artery. J Clin Invest 1990; 86:75–79.

13 Ware Branch D, Mitchell MD, Miller E, Palinski W, Witztum JL. Pre-eclampsiaand serum antibodies to oxidised low-density lipoprotein. Lancet 1994;343:645–646.

14 Bowen RS, Moodley J, Dutton MF, Fickl H. Antibodies to oxidised low-density lipoproteins and cardiolipin in pre-eclampsia and eclampsia. J ObstetGynaecol 2002; 22:123–126.

15 Kurki T, Ailus K, Palosuo T, Ylikorkala O. Antibodies to oxidized low-densitylipoprotein, cardiolipin and phosphatidyl serine fail to predict the risk ofpreeclampsia. Hypertens Pregnancy 1996; 15:251–256.

16 Baksu B, Baksu A, Davas I, Akyol A, Gulbaba G. Lipoprotein(a) levels inwomen with pre-eclampsia and in normotensive pregnant women. J ObstetGynaecol Res 2005; 31:277–282.

17 Bayhan G, Kocyigit Y, Atamer A, Atamer Y, Akkus Z. Potential atherogenicroles of lipids, lipoprotein(a) and lipid peroxidation in preeclampsia. GynecolEndocrinol 2005; 21:1–6.

18 Van den Elzen HJ, Wladimiroff JW, Cohen Overbeek TE, Bruijn AJ, Grobbee DE.Serum lipids in early pregnancy and risk of pre-eclampsia. Br J Obstet Gynaecol1996; 103:117–122.

19 Cekmen MB, Erbagci AB, Balat A, Duman C, Maral H, Ergen K, et al. Plasmalipid and lipoprotein concentrations in pregnancy induced hypertension. ClinBiochem 2003; 36:575–578.

20 Kokia E, Barkai G, Reichman B, Segal P, Goldman B, Mashiach S. Maternalserum lipid profile in pregnancies complicated by hypertensive disorders.J Perinat Med 1990; 18:473–478.

21 Pavan L, Hermouet A, Tsatsaris V, Therond P, Sawamura T, Evain Brion D,et al. Lipids from oxidized low-density lipoprotein modulate human tropho-blast invasion: Involvement of nuclear liver X-receptors. Endocrinology 2004;145:4583–4691.

22 Ramsay JE, Ferrell WR, Crawford L, Wallace AM, Greer IA, Sattar N.Divergent metabolic and vascular phenotypes in pre-eclampsia and in-trauterine growth restriction: relevance of adiposity. J Hypertens 2004;22:2177–2183.

23 Kim YJ, Park H, Lee HY, Ahn YM, Ha EH, Suh SH, et al. Paraoxonase genepolymorphism, serum lipid and oxidized low-density lipoprotein in pre-eclampsia. Eur J Obstet Gynecol Reprod Biol 2007; 133:47–52.

24 Qiu C, Phung TTT, Vadachkoria S, Muy Rivera M, Sanchez SE, Williams MA.Oxidized low-density lipoprotein (oxidized LDL) and the risk of preeclampsia.Physiol Res 2006; 55:491–500.

25 Raijmakers MTM, Van Tits BJH, Hak Lemmers HLM, Roes EM,Steegers EAP, Peters WHM. Low plasma levels of oxidized low densitylipoprotein in preeclampsia. Acta Obstet Gynecol Scand 2004; 83:1173–1177.

26 Pecks U, Caspers R, Schiessl B, Bauerschlag D, Piroth D, Maass N, et al.The evaluation of the oxidative state of low-density lipoproteins in intrauterinegrowth restriction and preeclampsia. Hypertens Pregnancy 2012; 31:156–165.

27 Uotila J, Solakivi T, Jaakkola O, Tuimala R, Lehtimaki T. Antibodiesagainst copper-oxidised and malondialdehyde-modified low densitylipoproteins in pre-eclamptic pregnancies. Br J Obstet Gynaecol 1998;105:1113–1117.

146 Evidence Based Women’s Health Journal