hum. reprod.-2014-bermejo-1271-8.pdf

ORIGINAL ARTICLE Reproductive biology

The impact of using the combined oralcontraceptivepill for cycle schedulingongene expression related to endometrialreceptivityAlfonso Bermejo1,*, Carlos Iglesias1, Mara Ruiz-Alonso2, David Blesa3,Carlos Simon3, Antonio Pellicer4, and Juan Garca-Velasco11Instituto Valenciano de Infertilidad, Av. Del Talgo 68 (28023), Madrid, Spain 2IVIOMICS, Paterna, Valencia, Spain 3Fundacion IVI, Universidadde Valencia, Valencia, Spain 4Instituto Valenciano de Infertilidad, Universidad de Valencia, Valencia, Spain

*Correspondence address. Tel: +34-91-180-29-00; Fax: +34-91-180-29-10; E-mail: [email protected]

Submitted on November 13, 2013; resubmitted on March 4, 2014; accepted on March 7, 2014

study question: Does the combined oral contraceptive pill (COCP) change endometrial gene expression when used for cycleprogramming?

summary answer: COCP used for scheduling purposes does not have a signicant impact on endometrial gene expression related toendometrial receptivity.

what is known already: Controversy exists around COCP pretreatment for IVF cycle programming, as some authors claim that itmight be detrimental to the live birth rate. Microarray technology applied to the study of tissue gene expression has previously revealed the be-havior of genes related to endometrial receptivity under different conditions.

study design, size, and duration: Proof-of-concept study of 10 young healthy oocyte donors undergoing controlled ovarianstimulation (COS) recruited between June 2012 and February 2013.

participants/materials, setting, and methods: Microarray data were obtained from endometrial biopsies from 10young healthy oocyte donors undergoing COS with GnRH antagonists and recombinant FSH. In group A (n 5), COCP pretreatment wasused for 1216 days, and stimulation began after a 5-day pill-free interval. Stimulation in group B (n 5) was initiated on cycle day 3 after a spon-taneous menses. Endometrial biopsies were collected 7 days after triggering with hCG.

main results and the role of chance:No individual genes exhibited increasedor decreasedexpression (fold change (FC).2)in patientswith priorCOCPtreatment (groupA) comparedwith controls (groupB).However, the results of the functional analysis showeda totalof 11 biological processes that were signicantly enriched in group A compared with group B (non-COCP).

limitations, reasons for caution: The Endometrial Receptivity Array (ERA) has only been validated on endometrial samplesobtained in natural cycles and after hormonal replacement treatment (HRT). Therefore, itwasnot possible in this study to classify the endometrialsamples as receptive or non-receptive.We used the ERA to focus on 238 genes that are intimately related to endometrial receptivity, thus sim-plifying the analysis and understanding of the data.

wider implications of the findings: Cycle scheduling is common in IVF units and is used to avoidweekend retrievals and/or todistribute evenly theworkload for better efciency. Our failure to detect any relevant changes in the genes related to thewindow of implantationwhencycleswereprogrammedwithCOCPpretreatment suggests that, despite controversial clinical results in previous studies, theuseofCOCPsin this way does not affect uterine receptivity adversely.

study funding/competing interest(s): Funding for this study was provided by an unrestricted grant from Merck Sharp &Dohme. C.S. and A.P. are co-inventors (with Patricia Diaz-Gimeno) of the Endometrial Receptivity Array and hold the patent. The otherauthors have no conicts of interest to declare.

& The Author 2014. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved.For Permissions, please email: [email protected]

Human Reproduction, Vol.29, No.6 pp. 12711278, 2014

Advanced Access publication on April 4, 2014 doi:10.1093/humrep/deu065

by guest on Decem

ber 22, 2015http://hum

rep.oxfordjournals.org/D

ownloaded from

trial registration number: EudraCT registration number is 2011-003250-34.

Key words: endometrial receptivity / contraceptive pill / GnRH antagonist / gene expression / cycle programming

IntroductionFor IVF applications, cycle planning provides improved organization anddistribution of laboratory work at assisted reproduction centers. Cycleplanning evenly distributes the number of oocyte retrievals throughoutthe week, which leads to cost savings by reducing the requirement forweekend personnel (Zorn et al., 1987).

Protocols that includeGnRHantagonists (GnRHant) have been asso-ciated with a reduction in ovarian hyperstimulation syndrome, shorterstimulation times and lower gonadotrophin dosages. The GnRHantprotocols have achieved pregnancy rates similar to those achievedwith longer protocols (Devroey et al., 2009; Al-Inany et al., 2011).However, GnRHant protocols have the drawback that the stimulationcycle must begin with spontaneous menstruation. In contrast, the longprotocol allows one to delay the onset of stimulation once pituitary de-sensitization has been achieved. To plan the start of a GnRHant proto-col, different approaches have been described: a exible start onDays 2or 3, a delay or advance of the hCG trigger or steroid pretreatment(Barmat et al., 2005; Tremellen and Lane, 2010; Cedrin-Durnerinet al., 2012).

Over the last few years, controversy has arisen regarding whetherpretreatment with a combined oral contraceptive pill (COCP) mayhave an impact on the IVF cycle outcome (Kolibianakis et al., 2006;Rombauts et al., 2006; Bellver et al., 2007). A recent meta-analysissuggested that COCPs had a negative impact on IVF results (Griesingeret al., 2010); however, not all authors agree (Garca-Velasco et al.,2011). This discrepancy prompted us to further evaluate endometrialreceptivity in IVF cycles where COCPs were administered as apretreatment.

Over the last decade, with the arrival of microarray technology, investi-gations have focusedon the genes related toendometrial receptivity duringnatural and stimulated cycles. The Endometrial ReceptivityArray (ERA) is adiagnostic tool that has been validated for analyzing the expression of 238genes that are intimately related to endometrial receptivity (Garrido-Gomez et al., 2013). Microarray technology has provided the ability tostudy, over a reduced period of time, the behavior of thousands of genesin a specic tissue under various conditions. Thus, this technology hasbeen applied to the evaluation of broids (Horcajadas et al., 2008a,b)and polyps (Liu et al., 2012), the premature elevation of progesterone(Labarta et al., 2011) and the impact of ovarian stimulation on endometrialreceptivity (Horcajadas et al., 2008a,b).

In the case of cycle planning with GnRHant, a microarray analysis canallow the assessment of endometrial receptivity from a molecular ex-pression standpoint. This information can be linked to whether steroidpretreatment applied prior to the start of the cyclemight have amolecu-lar effect on endometrial receptivity and whether steroid pretreatmentmight inuence the clinical results of IVF. In the present study, wecompared endometrial gene expression between women treatedwith COCP and those allowed to undergo spontaneous menstruationwithout COCP.

Materials andMethods

Study designThis was a single-center, proof-of-concept study conducted in auniversity-afliated private infertility clinic between June 2012 and February2013. The study was designed to compare the gene expression prole inthe endometrium between two groups: A, the study group (ve women)who received a COCP during the cycle prior to stimulation to schedulethe cycle starting on Day 1 (D1) of menses, and B, the control group (vewomen) who began stimulation directly with a spontaneous menseswithout prior COCP priming.

Study population and randomizationA total of 10 volunteers from our egg donor program were included in thestudy. The inclusion criteria were as follows: Caucasian; age 1835 years;a regular menstrual cycle (2535 days); normal basal serum levels of folliclestimulating hormone (FSH), luteinizing hormone (LH; both,10 UI/ml) andestradiol (E2;,60 pg/ml); normalkaryotype;bodymass index1825 kg/m2;normal cervical smear from the past year; and a normal vaginal ultrasound.The exclusion criteria were the presence of the following: endometriosis(AFS classication stage .2); polycystic ovary syndrome (according to therevised Rotterdam criteria); or an intrauterine device within the last3 months.

All patients agreed to participate in the study and signed the informedconsent form on the hCG triggering day. The Institutional Review Boardand the Institutional Ethics Committee approved the study. The EudraCTregistration number is 2011-003250-34.

Study groups and stimulation protocolPatients from group A received a COCP (Microgynon30, Bayer, Berlin,Germany) during the cycle prior to stimulation. The dosing consisted of 1tablet per day for 1216 days to schedule the cycle starting on D1 ofmenses. In groupA, stimulation began after a 5-day pill-free interval. Patientsfrom group B began stimulation directly onD2/D3 of the spontaneousmen-strual cycle without prior COCP priming.

The stimulation started with a xed dose of 150 IU recombinant FSHadministered subcutaneously (sc; Puregon, MSD, Madrid, Spain) for 4days. After that, doses were adjusted according to the follicular response,as visualized on ultrasound. To avoid a premature increase in LH, a dailydose of 0.25 mg GnRHant (Orgalutran, MSD, Madrid, Spain) was adminis-tered sc on xed D5 until the day of triggering.

Final oocytematuration was induced as soon as three follicles with ameandiameter .17 mm were observed on ultrasound. This was designated dayLH+0. For triggering, patients received 250 mg recombinant HCG sc (Ovi-trelle, Merck-Serono, Geneva, Switzerland). Follicle aspiration was con-ducted 36 h after triggering.

For luteal phase support, patients in both groups applied 200 mgvaginal micronized progesterone (Progefk, Efk, Madrid, Spain) every12 h.Thismedicationwascontinueduntil thedayof thebiopsies in all patients(LH+7).

1272 Bermejo et al.

by guest on Decem



ownloaded from

Hormonal and statistical analysisDonors were tested for progesterone (P4), LH and estradiol (E2) concentra-tions on the day of recombinant HCG administration (LH+0) and on daysLH+2, LH+4 and LH+6. Serum E2, P4 and LHwere evaluated using a com-mercially available,microparticleenzyme immunoassay kit (Abbott Laborator-ies, Abbott Park, IL, USA). Plasma samples fromall donorswere centrifuged at200 g for 5 min, and the supernatant was stored at 2808C until assayed.

The mean hormonal concentrations were compared with a one-way ana-lysis of variance or KruskalWallis test based on the normality of the distri-bution. Signicance was set at P, 0.05. The MannWhitney U-test wasused for pairwise comparisons of values that were not normally distributed,which yields results identical to the KruskalWallis test for two independentsamples. If multiple comparisons were performed, Bonferroni correction formultiple comparisons was used to analyze the differences between the pro-tocols. All statistical analyses were performed with the SPSS (StatisticalPackage for the Social Sciences) 18.0 software package (SPSS, Inc.).

Sample attainmentAt 7days after triggering (LH+7), an endometrial biopsywas obtained understerile conditions with a Pipelle catheter (Pipelle de Cornier, Prodimed,Neuilly-en-Thelle, France) from the uterine fundus of each woman.Samples were immediately placed in cryovials, homogenized in 1.5 mlRNA Later solution (QIAGEN, Barcelona, Spain), vigorously shaken for afew seconds and stored at room temperature for subsequent ERA analysis.All samples were processed within 4872 h.

Sample labeling and microarray hybridizationTotal RNA was extracted with the QIAGEN RNeasy mini Kit (QIAGEN,Chatsworth, CA, USA). Approximately 12 mg of total RNA per mg ofendometrial tissue was obtained. The RNA quality was assessed by loading300 ng of total RNA onto an RNA LabChip and running the chip on anAgilent A2100 Bioanalyzer (Agilent Technologies, Inc., Santa Clara, CA,USA). Good-quality RNA samples, with an RNA Integrity Number equalto or greater than seven, were a prerequisite for ERA analysis.

A total of 10 separate hybridizationswere performed. Sample preparationand hybridization were adapted from the Agilent technical manual (one-color). In short, rst-strand cDNA was reverse transcribed from 200 ng oftotal RNAwith T7-Oligo(dT) Promoter Primers. Samples were transcribedin vitro and labeled with Cy-3 with the Low Input Quick Amp Labeling kit(Agilent Technologies, Inc., SantaClara,CA,USA). The labeling reaction typ-ically yielded 45 mg of complementary RNA (cRNA) with a specic activitygreater than six. Fragmented cRNA sampleswere hybridized onto the custo-mized ERA array by incubating at 658C for 17 h with constant rotation. Themicroarray was then washed in two 1-min steps and two washing buffers(Agilent Technologies, Inc., Santa Clara, CA, USA). The microarrays withhybridized cRNAs were scanned with an Axon 4100A scanner (MolecularDevices, Sunnyvale, CA, USA), and the data were extracted using theGenePix Pro 6.0 software (Molecular Devices, Sunnyvale, CA, USA).

Endometrial receptivity array analysisERA gene expression values were pre-processed and normalized, and theendometrial receptivity status was diagnosed with the ERA computationalpredictor (Daz-Gimeno et al., 2011). The ERA test indicated whether theendometrial samples were Receptive (R) or Non-receptive (NR), with anassociated diagnostic probability. However, the ERA test has only been vali-dated on endometrial samples obtained in natural cycles and after hormonalreplacement treatment (HRT). Therefore, in this study, we used the ERA tofocus on 238 genes that are intimately related to endometrial receptivity. Theaccuracy and consistency of the ERA diagnostic tool has been demonstratedto be superior to endometrial histology. The results were completely

reproducible in a second ERA test that was applied 2940 months afterthe rst ERA test (Daz-Gimeno et al., 2013).

Data processing and data analysisAll analyses were performed using the Babelomics web-based suite (Medinaet al., 2010). The Agilent array background was corrected with robust multi-array averaging (RMA) methodology. The intensity signal was standardizedacross arrays with the quantile normalization algorithm.

Thedataobtained fromgroupAwere comparedwith thedata fromgroupBto evaluate the effects of COCP pretreatment. For all comparisons, weassessed the differential gene expression with limma moderated t-statistics.Standard microarray analysis techniques were used to perform one test foreach gene (or a probe-set) in the microarray and for comparisons. Thus, foreach gene, a t-test statistic is reported togetherwith its corresponding P-value.

The raw P-values had to be corrected for multiple testing to minimize theamount of false positives in the study. In this analysis,weused the convention-al multiple testing P-value correction procedures proposed by BenjaminiHochberg (Benjamini et al., 2001) to derive adjusted P-values.

The gene Set Analysis was performed for each of the comparisonsexplored in the study. We used the logistic regression models described inprevious studies (Montaner andDopazo, 2010; Sartor et al., 2010) to identifyfunctional blocks that were enriched in either of the conditions. In this study,we used the functional blocks described in the GO Biological Process data-base (Dennis et al., 2003). The conventional multiple testing P-value correc-tion procedure proposed by BenjaminiYekutieli (Benjamini and Yekutieli,2001) was used to derive the adjusted P-values.

When performing the array analysis, we opted for a two-way approach toovercome type II error: (a) instead of using triplicates as usual, we chose vesamples per group tominimize inter-patient variation andobtain a homogen-ous group, and (b) to obtain as much information as possible from eachsample, we decided that they should be tested individually.

ResultsEpidemiological characteristics and thebaselineovarian stimulationpara-meters for each group are summarized in Table I. There were no signi-cant differences in age, body mass index or the number of priordonations.With respect to ovarian stimulation, therewere no signicantdifferences in the gonadotrophin doses received, the number of mature

........................................................................................

Table I Baseline characteristics and parameters ofovarian stimulation.

A (COCP,n5 5)

B (non-COCP,n 5 5)

P-value

Age (years) 23.0+1.4 23.8+1.6 ns

BMI (Kg/m2) 20.9+0.9 20.7+1.2 ns

Total dose FSH (IU) 1485+314 1580+218 ns

Follicles .15 mmtriggering day

12.6+2.4 14.4+2.7 ns

No. of oocytesretrieved

20.8+4.6 19+2.7 ns

Days of stimulation 10.4+0.7 9.6+0.6 ns

E2 on day oftriggering (pg/ml)

2110+350 1975+280 ns

Values represent means+ SD.COCP, combined oral contraceptive pill; BMI, body mass index; FSH, folliclestimulating hormone; IU, international units; E2, plasma estradiol; ns, not signicant.

Oral contraception and endometrial gene expression 1273

by guest on Decem



ownloaded from

oocytes retrieved, the pre-puncture estradiol levels or the number ofdays of stimulation.

The serum levels of estradiol, LH andprogesterone are summarized inFig. 1 for each of the three donor sampling points: the day of oocyte re-trieval (LH + 2), 48 h later (LH + 4) and prior to endometrial biopsy(LH+7). No signicant differences were found between group A andthe non-COCP group (B).

Differential gene expressionAll samples were adequately hybridized and pre-analyzed. Data normal-ization reduced the technical artifacts and boosted detection of the bio-logical signal.

We analyzed 238 genes in the ERA and compared genetic expressionlevels between the two groups.Only one gene, tyrosine-3monooxygen-ase (TH), exhibited increased differential expression (fold change (FC).2) in patients with prior COCP treatment (group A) compared withthe patients without prior cycle planning (group B). However, the differ-ence was not statistically signicant (adjusted P-value .0.05). The heatmap plot (Fig. 2) showed a similar gene prole for samples withingroupsA (COCP) andB (non-COCP). Theprincipal component analysis(PCA) plot is shown in Fig. 3.

The results obtained from the ERA bioinformatic predictor do nothave clinical value in our study because the results and predictionshave only been validated in natural or hormone replacement therapycycles.Assuming this limitation, all analyzed sampleswere classiedas re-ceptive by the ERA predictor.

Functional analysis of gene expressionFunctional blocks from theGOBiological Process databasewere used inthis analysis (Dennis et al., 2003). The GO database is unique in that thefunctions or blocks of genes are organized into aDirected Acyclic Graph(DAG) structure. However, this procedure makes many of the signi-cantly enriched functions redundant. We displayed only the signicant,non-redundant terms (functions); that is, among the signicant terms,we selected the more specic terms.

The results showed that 11 biological processes were signicantlyenriched samples from group A (COCP) compared with those fromgroup B (non-COCP). These processes were dened as follows: re-sponse to metal ion (regulation of changes or cellular activities in re-sponse to stimulation with a metal ion); organ morphogenesis (thedevelopment of different organs upon stimulation from specic

Figure 1 Mean (n 5) serum levels of luteinizing hormone (LH,A),estradiol (E2,B) and progesterone (P4,C) at each of the three samplingpoints (LH+2, LH+4 and LH+7, considering day of triggering as dayLH+0). Blue and red lines represent group A (combined oral contra-ceptive pill, COCP) and B (controls, non-COCP) respectively.

Figure 2 Gene clustering by Pearsons correlation. A1A5 and B1B5 represent samples from group A (combined oral contraceptive pill,COCP) and group B (controls, non-COCP), respectively. On the leftof the heat map is the color-coding of gene transcripts: red signiesup-regulation, blue down-regulation and white unchanged.

1274 Bermejo et al.

by guest on Decem



ownloaded from

activities); response to oxygen-containing compounds (regulation ofcellular activity in response to an oxygenated compound); tissue devel-opment (regulationof tissue development until the denitive structure isachieved); cellular response to organic substance (regulation of the cel-lular response to certain molecular stimuli); single organism signaling(cell signalingprocesses); cell communication (mediationof intercellularor extracellular matrix communications); primary metabolic process(normal cellular catabolic or anabolic processes); cellular metabolicprocess (chemical processes by which a cell transforms substances);organic substancemetabolic process (chemical reactions for processingmolecules with a carbon backbone); and regulation of biological pro-cesses (control of genetic expression, protein changes and molecularinteractions). The affected biological processes, the implicated genesand their respective P-values are summarized in Table II.

DiscussionIn the past few years, the use of COCPs for GnRHant cycle planning haselicited substantial controversy (Kolibianakis et al., 2006;Rombauts et al.,2006; Bellver et al., 2007). It has been proposed thatCOCPs could nega-tively affect clinical results due to the possible effect of progestogen onendometrial receptivity (Griesinger et al., 2010). Additionally, it was sug-gested thatCOCPpretreatmentmaystrongly suppress LHsecretion andimpair folliculogenesis in subsequent cycles when only rFSHwas used. Inthe present study, we examined the expression of genes responsible forendometrial receptivity inpatients subjected to controlledovarian stimu-lation (COS) under a GnRHant protocol in the presence or absence of

COCP treatment. The take home message of this study was that wedid not observe signicant differences in endometrial gene expressionrelated to endometrial receptivity in these patients, regardless of previ-ous COCP treatment.

When analyzing our results, only one gene, TH, showed an increasedexpression in the COCP group compared with patients without priorcycle scheduling, but this increase was not signicant (adjusted P-value.0.05). This enzyme catabolizes tyrosine conversion to phenylalanine,an essential amino acid that acts as a dopamine precursor. TH geneup-regulationwithin thewindowof implantation has beendescribedpre-viously by other authors (Daz-Gimeno et al., 2011) under natural condi-tions.

As expected, the secretory phase is marked bymore gene expressionchanges due to the presence of the progesterone peak. These changescoincide with the time that the endometrium becomes receptive. Thegenes that areoverexpressed in themid-secretory versus theearly secre-tory phase have known functions that are involved in implantation prep-aration, and they are implicated in processes related to cell adhesion,metabolism, response to external stimuli, signaling, immune response,cell communication and negative regulation of proliferation and develop-ment (Talbi et al., 2006).

Eleven biological processes were signicantly up-regulated in womenpretreated with COCP compared with non-COCP. Three of them arerelated tometabolic processes (involving primary, cellular and organic sub-stances), and one (single organism signaling) was related to signaling func-tions. All of these processes are enriched, which occurs in a natural cycleduring the window of implantation (Ruiz-Alonso et al., 2012). Other keyprocesses for embryo implantation, such as cell adhesion, innate immuneresponse and the response to stress or wounding, were not modied inthe endometrial samples from the COCP group. Functional analysis ofthe included gene in the ERA revealed that the most signicant annota-tion is the presence of 19 processes including ones relating to theimmune response, cytoskeleton proteins needed for remodeling theendometrium, oxidoreductase activity, carbohydrate binding and recep-tor binding (Daz-Gimeno et al., 2011). None of these 19 biological

Figure3 Normalizeddata fromPrincipalComponentAnalysis (PCA)plots. PCAwas done on the 10 samples from the 2 groups of this study.The samples are represented spatially according to the gene expressionvalue for the 238 Endometrial Receptivity Array (ERA) genes. Thegroups are represented by different colors: Group A (combined oralcontraceptive pill, COCP), red; Group B (non-COCP), blue. PC1 andPC2 are rst and second principal components that explain thehighest variability that separates samples in the space.

........................................................................................

Table II GO terms related to differentially expressedgenes.

GO term Name Genes P-value

GO:0010038 Response to metal ion 11 0.033

GO:0009887 Organ morphogenesis 23 0.023

GO:1901700 Response to oxygen-containingcompound

29 0.024

GO:0009888 Tissue development 37 0.011

GO:0071310 Cellular response to organicsubstance

40 0.019

GO:0044700 Single organism signaling 110 0.024

GO:0007154 Cell communication 114 0.032

GO:0044238 Primary metabolic process 131 0.023

GO:0044237 Cellular metabolic process 127 0.033

GO:0071704 Organic substance metabolicprocess

137 0.023

GO:0050789 Regulation of biological process 158 0.023


by guest on Decem



ownloaded from

processes that characterize the receptive endometriumwere affected inthe study group (Table II).

The estrogen and progestogen in COCP have been shown to affectFSH and LH secretion, respectively (Tsai and Yen, 1971; Andersonet al., 1990). However, the estrogen doses in the COCPs used in differ-ent studies have greatly varied from 15 to 50 mg. In addition, differentstudies have varied the COCP-free interval prior to the start of stimula-tion between 2 and 5 days, with variation of prior treatment durationbetween 10 and 21 days (Cedrin-Durnerin et al., 2007; Smulders et al,2010). These variables could have an impact on clinical results. Previousstudies have demonstrated that after a 5-day pill-free interval, the endo-crine prole of women pretreated with a COCP resembled that ofwomen allowed to undergo spontaneous menstruation (Cedrin-Durnerin et al., 2007).

The rst studies to assess the effects of the use of COCP in cycle plan-ning with GnRHant suggested that COCP had a possible benecial effecton clinical results (Biljan et al., 1998; Fukuda et al., 2000). This result con-trastedwith later studies,whichsuggested thatCOCPhadanegativeeffectonpregnancy rates (Pinkas et al., 2008;Meldrum et al., 2009).However, insome studies, thewashout periodwas only 2 days. A short washout inter-val has been linked to higher early pregnancy loss, due to the continued,strong COCP-mediated suppression of endogenous gonadotrophin se-cretion during the earlywashout period (Griesinger et al., 2010). The vari-ability in the type of COCP, the duration of treatment and the duration ofthe pill-free interval has generated heterogeneous results and confusionabout the effects of COCP on IVF outcome.

A turning point occurred with the meta-analysis published in 2010(Griesinger et al., 2010). That meta-analysis included six previousstudies and reported that the group that receivedpriorCOCP treatmentshowed a signicant decrease in the rate of ongoing pregnancy, anincreased duration of stimulation, and greater consumption of gonado-trophins compared with those that did not receive COCP. Thus, itwas proposed that the progestogen in COCPs might have a negativeeffect on endometrial receptivity.

A randomized controlled study fromour group (Garca-Velasco et al.,2011) evaluated the clinical results from 115 women who had receivedprior treatment with COCP before starting a GnRHant protocol aspart of their IVF/ICSI cycle. These results were compared with theresults from113patientswho received stimulation under a long protocolwithout prior COCP treatment. We found no signicant differences inthe pregnancy, implantation or live birth rates between the groups.

The number of days of COCP administration is relevant becauseCOCP suppresses pituitary and ovarian activity. Prolonged COCP ad-ministration was associated with a greater suppression of activity (vanHeusden et al., 2002). Therefore, more stimulation days and higherFSH doses would be required throughout the COS cycle, which wouldhave a negative economic impact on the cost of the cycle. In thepresent study, we administered COCP for between 12 and 16 days,which we considered sufcient to restrain FSH levels without excessiveinhibition. In this series and in our previous study (Garca-Velascoet al., 2011), COCP pretreatment for 1216 days did not increase thetotal dose of gonadotrophins required compared with non-COCPpatients.

High progesterone levels at the beginning of the cycle, which couldarise from the progestogen content of the COCP, have been linked toa decreased probability of pregnancy; however, this effect remains con-troversial because recent publications have not conrmed those ndings

(Fatemi et al., 2013). Interestingly, in addition to the deleterious effects ofprogesterone at the beginning of the cycle, elevated progesterone levelsat the end of the stimulation may have a negative effect on endometrialreceptivity; this was recently studied using microarray technology. Im-portant differences were observed in the genetic expression patternsof patients with progesterone concentrations above or below 1.5 ng/mlon the day of follicular maturation (Labarta et al., 2011).

In the present study, we used a 30-mg dose of COCP and implemen-ted a 12- to 16-day treatment period with a 5-day washout period. Wedid not observe differences in the gene expression patterns at the endo-metrial level that could affect endometrial receptivity. Thus, we con-cluded that the 10 samples were nearly identical in this respectregarding the endometrial gene expression of the genes related to thewindow of implantation.

All participants received the sameprotocol to avoid problemswith re-producibility or cycle-to-cycle variations and to increase the consistencyof the gene expression patterns for assessing endometrial receptivity. Arecent report has demonstrated the safety and reproducibility of ERAendometrial microarrays and compared the results with classical histo-logical analyses. Duplicate biopsies were obtained under the same con-ditions from sevenpatients, with intervals betweenbiopsies ranging from29 to 40months. They demonstrated 100% reproducibility between thetwo samples (Daz-Gimeno et al., 2013).

We can conclude that the informationwe discovered regarding endo-metrial gene expression usingmicroarray analysis suggests that the use ofCOCP as a pretreatment did not negatively impact the prole of endo-metrial gene expression for women undergoing IVF with the GnRHantprotocol.WhenCOCPswere used for a maximum of 16 days and a pill-free interval was allowed for 5 days, endometrial gene expression wasnot modied.

Authors rolesA.B. designed the study, conducted the data analysis and interpretationand wrote the manuscript. He is the corresponding author for thereviewing procedure. M.R.-A. and D.B. processed the endometrialsamples and performed the microarray technology, data analysis andmanuscript writing. C.I. participated in donor recruitment and careduring the medical treatment and performed the endometrial biopsiesfor sample collection. A.P. and C.S. contributed to the review of themanuscript. J.G.-V. participated in the study design and reviewed the in-tellectual content of the manuscript.

FundingFunding for this study was provided by an unrestricted grant fromMerckSharp & Dohme.

Conict of interestC.S. and A.P. are co-inventors (with Patricia Diaz-Gimeno) of the Endo-metrial ReceptivityArray and hold the patent.Noneof the other authorshas any conicts of interest to declare.

1276 Bermejo et al.

by guest on Decem



ownloaded from

ReferencesAl-InanyHG,YoussefMA,AboulgharM,BroekmansF, SterrenburgM, Smit J,Abou-Setta AM. Gonadotrophin-releasing hormone antagonists forassisted reproductive technology. Cochrane Database Syst Rev 2011;5:CD001750.

AndersonRE, SteinA, PaulsonRJ, Stanczyk FZ, VijodAG, LoboRA. Effects ofnorethindrone on gonadotropin and ovarian steroid secretion when usedfor cycle programming during in vitro fertilization. Fertil Steril 1990;54:96101.

Barmat L, Chantilis S, Hurst B, Dickey R. A randomized prospective trialcomparing gonadotropin releasing hormone (GnRH) antagonist/recombinant follicle-stimulating hormone (rFSH) versus GnRH agonist/rFSH in women pretreated with oral contraceptives before in vitrofertilization. Fertil Steril 2005;83:321330.

Bellver J, Albert C, Labarta E, Pellicer A. Early pregnancy loss in womenstimulated with gonadotropin releasing hormone antagonist protocolsaccording to oral contraceptive pill pretreatment. Fertil Steril 2007;87:10981101.

Benjamini Y, Yekutieli D. The control of the false discovery rate in multipletesting under dependency. Ann Stat 2001;4:11651188.

Benjamini Y, Drai D, Elmer G, Kafka N, Golani I. Controlling the falsediscovery rate in behavior genetics research. Behav Brain Res 2001;125:279284.

Biljan M, Mahutte N, Dean N, Hemmings R, Bissonnette F, Tan SL. Effects ofpre-treatment with oral contraceptive on the time required to achievepituitary suppression with gonadotropin-releasing hormone analoguesand on subsequent implantation and pregnancy rates. Fertil Steril 1998;70:10631169.

Cedrin-Durnerin I, Bstandig B, Parneix I, Bied-Damon V, Avril C,Decanter C, Hugues JN. Effects of oral contraceptive, syntheticprogestogen or natural estrogen pre-treatments on the hormonalprole and the antral follicle cohort before GnRH antagonist protocol.Hum Reprod 2007;22:109116.

Cedrin-Durnerin I, Guivarch-Leveque A, Hugues JN. Pretreatment withestrogen does not affect IVF-ICSI cycle outcome compared with nopretreatment in GnRH antagonist protocol: a prospective randomizedtrial. Fertil Steril 2012;97:13591364.

Dennis G Jr, Sherman BT, Hosack DA, Yang J, Gao W, Lane HC,Lempicki RA. DAVID: Database for Annotation, Visualization, andIntegrated Discovery. Genome Biol 2003;4:P3.

Devroey P, Aboulghar M, Garcia-Velasco JA, Griesinger G, Humaidan P,Kolibianakis E, Ledger W, Tomas C, Fauser BC. Improving thepatients experience of IVF/ICSI: a proposal for an ovarian stimulationprotocol with GnRH antagonist co-treatment. Hum Reprod 2009;24:764774.

Daz-Gimeno P, Horcajadas JA, Martnez-Conejero JA, Esteban FJ, Alama P,Pellicer A, Simon C. A genomic diagnostic tool for human endometrialreceptivity based on the transcriptomic signature. Fertil Steril 2011;95:5060.

Daz-Gimeno P, Ruiz-Alonso M, Blesa D, Bosch N, Martnez-Conejero JA,Alama P, Garrido N, Pellicer A, Simon C. The accuracy andreproducibility of the endometrial receptivity array is superior tohistological dating as diagnostic method for the endometrial factor. FertilSteril 2013;99:508517.

Fatemi HM, Doody K, Griesinger G, Witjes H, Mannaerts B. High ovarianresponse does not jeopardize ongoing pregnancy rates and increasescumulative pregnancy rates in a GnRH-antagonist protocol. Hum Reprod2013;28:442452.

FukudaM,FukudaK,AndersenCY,ByskovAG.Doesanovulation inducedbyoral contraceptives favor pregnancy during the following two menstrualcycles? Fertil Steril 2000;73:742747.

Garca-Velasco JA, Bermejo A, Ruiz F, Martnez-Salazar J, Requena A,Pellicer A. Cycle scheduling with oral contraceptive pills in the GnRHantagonist protocol vs the long protocol: a randomized, controlled trial.Fertil Steril 2011;96:590593.

Garrido-Gomez T, Ruiz-Alonso M, Blesa D, Diaz-Gimeno P, Vilella F,Simon C. Proling the gene signature of endometrial receptivity: clinicalresults. Fertil Steril 2013;99:10781185.

Griesinger G, Kolibianakis EM, Venetis C, Diedrich K, Tarlatzis B. Oralcontraceptive pretreatment signicantly reduces ongoing pregnancylikelihood in gonadotropin-releasing hormone antagonist cycles: anupdated meta-analysis. Fertil Steril 2010;94:23822384.

Horcajadas JA, Goyri E, Higon MA, Martnez-Conejero JA, Gambadauro P,Garca G, Meseguer M, Simon C, Pellicer A. Endometrial receptivity andimplantation are not affected by the presence of uterine intramuralleiomyomas: a clinical and functional genomics analysis. J Clin EndocrinolMetab 2008a;93:34903498.

Horcajadas JA, Mnguez P, Dopazo J, Esteban FJ, Domnguez F, Giudice LC,Pellicer A, Simon C. Controlled ovarian stimulation induces a functionalgenomic delay of the endometrium with potential clinical implications.J Clin Endocrinol Metab 2008b;93:45004510.

Kolibianakis EM, Papanikolau EG, Camus M, Tournaye H, VanSteirteghem AC, Devroey P. Effect of oral contraceptive pillpretreatment on ongoing pregnancy rates in patients stimulated withGnRH antagonists and recombinant FSH for IVF. A randomizedcontrolled trial. Hum Reprod 2006;21:352357.

Labarta E, Martnez-Conejero JA, Alama P, Horcajadas JA, Pellicer A,Simon C, Bosch E. Endometrial receptivity is affected in women withhigh circulating progesterone levels at the end of the follicular phase: afunctional genomics analysis. Hum Reprod 2011;26:18131825.

Liu Y, Zhang Y, Fu J, TanW. Inammation-related gene expression proles ofendocervical polyps. J Interferon Cytokine Res 2012;32:191197.

Medina I, Carbonell J, Pulido L, Madeira SC, Goetz S, Conesa A, Tarraga J,Pascual-Montano A, Nogales-Cadenas R, Santovo J. Babelomics: anintegrative platform for the analysis of transcriptomics, proteomics andgenomic data with advanced functional proling. Nucleic Acids Res 2010;38:W210W213.

Meldrum D, Scott R, Levy M, Alper M, Noyes N. Oral contraceptivepretreatment in women undergoing controlled ovarian stimulation inganirelix acetate cycles may, for a subset of patients, be associatedwith low serum luteinizing hormone levels, reduced ovarian responseto gonadotropins, and early pregnancy loss. Fertil Steril 2009;91:19631965.

Montaner D, Dopazo J. Multidimensional gene set analysis of genomic data.PLoS One 2010;5:e10348.

PinkasH, SapirO,AvrechO, Ben-HaroushA, Ashkenazi J, Fish B, Farhi J. Theeffect of oral contraceptive pill for cycle scheduling prior to GnRHantagonist protocol on IVF cycle parameters and pregnancy outcome.J Assist Reprod Genet 2008;25:2933.

Rombauts L, Healy D, Norman RJ. Comparative randomized trial to assessthe impact of oral contraceptive pretreatment on follicular growth andhormone proles in GnRH antagonist-treated patients. Hum Reprod2006;13:235245.

Ruiz-AlonsoM,BlesaD, SimonC.The genomics of the humanendometrium.Biochim Biophys Acta 2012;1822:19311942.

Sartor MA, Mahavisno V, Keshamouni VG, Cavalcoli J, Wright Z,Karnovsky A, Kuick R, Jagadish HV, Mirel B, Weymouth T. ConceptGen:a gene set enrichment and gene set relation mapping tool. Bioinformatics2010;26:456463.

Smulders B, van Oirschot S, Farquhar C, Rombauts L, Kremer J. Oralcontraceptive pill, progestogen or estrogen pre-treatment for ovarianstimulation protocols for women undergoing assisted reproductivetechniques. Cochrane Database Syst Rev 2010; CD006109.


by guest on Decem



ownloaded from

Talbi S, Hamilton AE, Vo KC, Tulac S, Overgaard MT, Dosiou C, Le Shay N,NezhatCN,KempsonR, LesseyBA et al.Molecular phenotyping of humanendometrium distinguishes menstrual cycle phases and underlyingbiological processes in normo-ovulatory women. Endocrinology 2006;147:10971121.

TremellenKP, LaneM.Avoidanceofweekendoocyte retrievals duringGnRHantagonist treatment by simple advancement or delay of hCGadministration does not adversely affect IVF live birth outcomes. HumReprod 2010;25:12191224.

Tsai CC, Yen SC. The effect of ethinyl-estradiol administration during earlyfollicular phase on the gonadotrophin levels and ovarian function. J ClinEndocrinol Metab 1971;33:917923.

van Heusden AM, Coelingh Bennink HJ, Fauser BC. FSH and ovarianresponse: spontaneous recovery of pituitary-ovarian activity during thepill-free period vs exogenous recombinant FSH during high-dosecombined oral contraceptives. Clin Endocrinol (Oxf) 2002;56:509517.

Zorn JR, BoyerP,GuichardA.NeveronSunday: programming for IVF-ETandGIFT. Lancet 1987;1:385386.

1278 Bermejo et al.

by guest on Decem



ownloaded from

/ColorImageDict > /JPEG2000ColorACSImageDict > /JPEG2000ColorImageDict > /AntiAliasGrayImages false /CropGrayImages true /GrayImageMinResolution 150 /GrayImageMinResolutionPolicy /OK /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 175 /GrayImageDepth -1 /GrayImageMinDownsampleDepth 2 /GrayImageDownsampleThreshold 1.50286 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages false /GrayImageAutoFilterStrategy /JPEG2000 /GrayACSImageDict > /GrayImageDict > /JPEG2000GrayACSImageDict > /JPEG2000GrayImageDict > /AntiAliasMonoImages true /CropMonoImages true /MonoImageMinResolution 1200 /MonoImageMinResolutionPolicy /OK /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 175 /MonoImageDepth 4 /MonoImageDownsampleThreshold 1.50286 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict > /AllowPSXObjects true /CheckCompliance [ /None ] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile (None) /PDFXOutputConditionIdentifier () /PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped /False

/Description >>> setdistillerparams> setpagedevice

hum. reprod.-2014-bermejo-1271-8.pdf

Documents