Abstract

Three decades of clinical experience with assisted reproductive technologies (ART) led to an intimate intertwining between the responses to controlled ovarian stimulation (COS) and actual ART outcome.  In turn the functional testing that predicts COS responses – dubbed ‘ovarian reserve’ – was mistakenly seen as foretelling the degree of remaining fecundity.  The highly publicized ‘ovarian reserve’ concept has now sparked an emerging business of advanced or ‘provisional’ ART for the purpose of saving cryopreserved oocytes for later use in women whose ovarian reserve is diminished.

Oocyte preservation is certainly indicated in case of impending oocyte loss – before chemotherapy or ovarian surgery for say enometriomas.  Yet, targeting this technique toward women simply based on low AMH screen values is probably most often improper.  It is indeed important to realize and make widely known the fact that poor ovarian reserve scores – i.e. low AMH – that predict poor COS responses assess the efficacy of ART as a clinical tool, not a woman’s fecundity.

Poor ovarian responses and poor ART outcome, an enduring confusion

Over 30 years of history, in vitro fertilization (IVF) originally developed for bypassing diseased tubes causing infertility saw its indications expend to become the ultimate treatment of practically all forms of infertility.  In a semantic adjustmentmeant to reflect this broadened scope, IVF and its variant intra-cytoplasmic sperm injection (ICSI) conceived for male-factor infertility became collectively called ‘assisted reproductive technologies’ (ART).

Right frominception,it stood as evidence that ART’s resultscould be improvedby harvesting several rather than one oocyteand in turn obtaining multipleembryos. The latter enables to cherry-pick the best embryo(s) to put back and in that, optimize the implantation chances.  Looking backat the 30 year-history of ART, it stands out thatthe induction of multiple-ovulationturned out beingthe single most effective measure evertaken for improving ARToutcome. In all these years, the induction of multiple-ovulationevolved to become a true science in its own right, now known as controlled ovarian stimulation (COS).  Today COS – a process that revolves around artificially elevating the levels of circulating FSH – developedits rules, recognized pitfalls and risks, and follows sets of preferred protocols ⁷.  Yet, in spite of the great efforts taken for tailoring the doses of FSH used in COS to each woman’s needs, it is notorious that COS responses can vary greatlyfrom person to person, remaining to some degree poorly predictable ⁸.  At one end of the spectrum, certain women havestrongand possibly excessive responses.  These carry a risk of frank ovarian hyperstimulationsyndrome (OHS) – a potentially serious complication of ART – now effectively curbed by using ‘antagonist’ protocols.  At the other end of the spectrum, different women have insufficientor so-called ‘poor’ responses to COS in spite of using larger doses of FSH.  Aninternational consensus– the Bologna Conference ⁹– has defined ‘poor response’as one that is either cancelled or yields <3 oocytes.  The term ‘poor response’ underscores the concept – erroneous, as we will see – that COS responses of diminished magnitude inherently bear anominous prediction for ART outcome.  Stated differently, the diminished oocyte quantity characterizing the poor responses to COS are said to be associated with reduced oocyte quality.  While this is generally true forthe poor responses encountered in older women, it is not necessarily the case however, when the poor response to COS is due to causes other than aging, as for example endometriosis ¹⁰.

An emerging challenge in modern-day infertility management is therefore to sort out the issues of oocyte quantity –impacting on COS responses – from those of poor oocyte quality causing poor ART outcome.  Distinguishing the problems of diminished oocyte quantity and quality is the core topic of the present review.  The lines above sketched the backdrop of our discussion, one in which COS responses – oocyte quantity – have been intimately intertwined with ART outcome, a surrogate marker of oocyte quality.

Aging causes a parallel decrease in oocyte quantity and quality

Throngs of evidence indicate that COS yields dwindledown,as women are getting older ¹¹.  Data from national registries – for example, SART data reported by the CDC in the US (Fig. 1) – depicta typical age-related decrease in ART outcome ¹².  Characteristically, this phenomenonparallels the fact that oocyte harvestsbecome meager in older women.  The latter itself reflects the progressive age-dependent shrinkage of the pool of primordial follicles seen as women age (Fig. 1-A, lower curve) ¹³.  In striking contrast, the results of donor-egg ART remain constant over time(Fig. 1-A, upper curve) ¹².  Taken together these two findingsindicate that the age-related decrease in ART outcome seen in aging women stems from an age-dependent decrease in oocyte quality.  Conversely, the reproductive potential of a properly primed uterus remains practically unchanged in aging women.

Looking at the number of primordial follicles observed in the ovary as a function of age, Gougeon’s team reported that the age-related decrease is bi-modal (Fig. 1-B) ^{13 , 14}.  Strikingly, the dwindlingin the number of remaining primordial follicles parallels the decline in ART outcome.  Logically, this ledto postulatethe existence of an inherent link between oocyte quantity and quality, one that we now know is only apparent, not real.  As evident today and described in detailsbelow, the appearance of a link between oocyte quantity and quality is in fact the effect of a confounder, the aging of ovarian function.  Aging indeed affects both oocyte quantity and quality, but yet these 2 processes are not inherently linked otherwise.

Endometriosis, an age-independent process, alters oocyte quantity but not quality

Endometriosis – a disease of unknown origin –often affects ovarian functionas well ¹⁵.  This is notably the case when endometriotic cysts, or endometriomas develop and particularly, when they extend bilaterally ¹⁶.  In a recent analysis of AMH levels in women with endometriosis, the reduction in ovarian follicle ssusceptible of responding to COS primarily stemmed from surgery for endometriosis, not from the disease itself ¹⁷.  In case past surgery for endometriosis, the reduction of ovarian function can be profound, however ¹⁰.

In spite of the potentially severe alteration of ovarian function caused by endometriosis, we ¹⁰and others ^{18 , 19}have reported that ART outcome remained in generalat par with the results of age-matched controls.  The fact that poor responses to COS due to endometriosis do not necessarily carry the ominous prediction known to poor responses due to aging shattered year-long accumulations of erroneous thoughtsextrapolating oocyte quality, based on quantity assessment.  Indeed, there is now overwhelming evidence that decreased oocyte quantity due to an age-independent process such as notably endometriosis is not necessarily accompanied by a decrease in oocyte quality.  It is about time that we blow the whistle.

Ovarian reserve, a misnomer and a confusing concept

The considerations that linked COS responses to ART outcome have led to extrapolate on women’s fecundity in a highly erroneous manner, as it turns out.  Specifically, there is now evidence that COS responses and by extension ovarian reserve parameters which predict such responses ²⁰ do not reflecta woman’s fecundity.  In France, oocyte donors are mandated to be fertile and bared from receiving monetary stipends.  Hence, women in need of oocyte donation reel on motivating friends or acquaintances without further assessing their ovarian parameters and disposition to respond to COS.  Under the circumstances, a third of our oocyte donors whose mean age is of 31.8 years and are fertile (1.8 children on average) end up being poor responders.  In these women as in their infertile counterparts, low AMH levels predict the poor response to COS.  But strikingly, these poor responders – a third of the young oocyte donors self-recruited by patients – are actually fertile.  Hence as reported in throngs of publications, the markers of ovarian reserve – for example AMH – do predict ovarian responses to COS but not their fecundity.  Hence, the markers of ovarian reserve – first among them AMH – predict COS responses and the efficacy of ART, if ART is needed (as for example in case male factor infertility).  But, it is time that we realize that the ovarian reserve markers and quality of responses to COS are not inherent reflector of fecundity when taken outside of the context of ovarian aging.

The misleading concept of ‘ovarian reserve’

The magnitude of COS responses can be predicted by sets of tests that directly or indirectly assess the number of antral follicles susceptible of responding to exogenous FSH administration.  These tests include: (i) the actual counting of antral follicles identified on ultrasounds or antral follicle count (AFC) score: (ii) antimullerian hormone (AMH) levels and; (iii) day-3 or ‘baseline’ FSH levels.  Collectively, these parameters are identified as ovarian reserve testing.  Implicit in the ‘ovarian reserve ’term is the concept that such tests assess the ovarian function remaining at any given time, or an individual’s advancement on her personal ovarian aging course.  Hence for doctors and patients alike, the term of ‘ovarian reserve’ invites to believe that such testing – AMH levels notably – predicts the amount of remaining fecundity.  Reeling from the unraveling of the lingering confusion between oocyte quantity and quality addressed above, the term of‘ovarian reserve’ used for predictors of COS responses not fecundity is a true misnomer.

Indeed the confusion between ‘ovarian reserve’ and fecundity has spread widely. The highly publicized term of ‘ovarian reserve’– truly a sound bite – led for example to recommend measuring AMH in women in their thirties.  Most pernicious in this practice is the propositionto deploy advanced or ‘provisional’ ART procedures for the purpose cryopreserving oocytes for later use in case of poor ovarian reserve readings (low AMH).  In reality, low ovarian reserve testing – low AMH levels – only predicts the efficacy of ART in that patient if ART were to become necessary, not her fecundity.  Hence, it is rather the individuals who score high on ‘ovarian reserve’ testing who are more likely to benefit from advanced or‘provisional’ ART, as they can garnish large numbers of oocytes.  On the contrary, women who score low on the ‘ovarian reserve’ scale are going to be poor ART performer (limited efficacy of the ‘provisional’ freezing), but yet do not necessarily suffer from decreased fecundity.  This lingering confusion has sparked a lucrative business of advanced or ‘provisional’ ART procedures that emerged from erroneous interpretations of clinical data that merely predict COS responses, not fecundity.  The targeted population – women in their thirties– and the medical community at large should be maid aware of this confusion.

Take-home message to be passed on

Thirty years of ART experience have led to an intimate intertwining between COS responses and ART outcome.  While COS responses bear practical consequences on ART chances, we have learned that the quantitative markers of ovarian response do not reflect a woman’s inherent fecundity.  Moreover, if a poor response due to an aging process is of poor prognosis for live birth rates, the situation is drastically different when the cause of poor ovarian response is not aging, but say endometriosis.  Patients should be therefore counseled on the value of pursuing ART and guidance given as to whether the oocyte retrieval is worthwhile based not just on the number of follicles growing, but also the cause of poor response.  When poor ovarian response is due to endometriosis, we believe that the oocyte retrieval is worth undertaking even if there are as few as 2 mature follicles (>17mm) developing.  On the contrary, is not reasonable to undertake an oocyte retrieval for less than five follicles in case of poor response due to aging.

References

1.    Fauser BC, Van Heusden AM. Manipulation of human ovarian function: physiological concepts and clinical consequences. Endocrine reviews 1997;18:71-106.

2.    Zeleznik AJ. Modifications in gonadotropin signaling: a key to understanding cyclic ovarian function. Journal of the Society for Gynecologic Investigation 2001;8:S24-5.

3.    de Ziegler D, Fraisse T, de Candolle G, Vulliemoz N, Bellavia M, Colamaria S. Outlook: Roles of FSH and LH during the follicular phase: insight into natural cycle IVF. Reproductive biomedicine online 2007;15:507-13.

4.    Channing CP, Tanabe K, Jones GS, Jones HW, Jr., Lebech P. Inhibin activity of preovulatory follicles of gonadotropin-treated and untreated women. Fertility and sterility 1984;42:243-8.

5.    Bernardus RE, Jones GS, Acosta AA, et al. The significance of the ratio in follicle-stimulating hormone and luteinizing hormone in induction of multiple follicular growth. Fertility and sterility 1985;43:373-8.

6.    Jones GS, Acosta AA, Garcia JE, Bernardus RE, Rosenwaks Z. The effect of follicle-stimulating hormone without additional luteinizing hormone on follicular stimulation and oocyte development in normal ovulatory women. Fertility and sterility 1985;43:696-702.

7.    Fatemi HM, Blockeel C, Devroey P. Ovarian stimulation: today and tomorrow. Current pharmaceutical biotechnology 2012;13:392-7.

8.    Kummer NE, Feinn RS, Griffin DW, Nulsen JC, Benadiva CA, Engmann LL. Predicting successful induction of oocyte maturation after gonadotropin-releasing hormone agonist (GnRHa) trigger. Hum Reprod 2012.

9.    Ferraretti AP, La Marca A, Fauser BC, Tarlatzis B, Nargund G, Gianaroli L. ESHRE consensus on the definition of 'poor response' to ovarian stimulation for in vitro fertilization: the Bologna criteria. Hum Reprod 2011;26:1616-24.

10.  de Ziegler D, Gayet V, Aubriot FX, et al. Use of oral contraceptives in women with endometriosis before assisted reproduction treatment improves outcomes. Fertility and sterility 2010;94:2796-9.

11.  Stoop D, Ermini B, Polyzos NP, et al. Reproductive potential of a metaphase II oocyte retrieved after ovarian stimulation: an analysis of 23 354 ICSI cycles. Hum Reprod 2012;27:2030-5.

12.  Sunderam S, Kissin DM, Flowers L, et al. Assisted reproductive technology surveillance - United States, 2009. MMWR Surveill Summ 2012;61:1-23.

13.  Faddy MJ, Gosden RG, Gougeon A, Richardson SJ, Nelson JF. Accelerated disappearance of ovarian follicles in mid-life: implications for forecasting menopause. Hum Reprod 1992;7:1342-6.

14.  Gougeon A. Dynamics of follicular growth in the human: a model from preliminary results. Hum Reprod 1986;1:81-7.

15.  de Ziegler D, Borghese B, Chapron C. Endometriosis and infertility: pathophysiology and management. Lancet 2010;376:730-8.

16.  Somigliana E, Arnoldi M, Benaglia L, Iemmello R, Nicolosi AE, Ragni G. IVF-ICSI outcome in women operated on for bilateral endometriomas. Hum Reprod 2008;23:1526-30.

17.  Streuli I, de Ziegler D, Gayet V, et al. In women with endometriosis anti-Mullerian hormone levels are decreased only in those with previous endometrioma surgery. Hum Reprod 2012;27:3294-303.

18.  Al-Azemi M, Bernal AL, Steele J, Gramsbergen I, Barlow D, Kennedy S. Ovarian response to repeated controlled stimulation in in-vitro fertilization cycles in patients with ovarian endometriosis. Hum Reprod 2000;15:72-5.

19.  Surrey ES, Silverberg KM, Surrey MW, Schoolcraft WB. Effect of prolonged gonadotropin-releasing hormone agonist therapy on the outcome of in vitro fertilization-embryo transfer in patients with endometriosis. Fertility and sterility 2002;78:699-704.

20.  Andersen AN, Witjes H, Gordon K, Mannaerts B. Predictive factors of ovarian response and clinical outcome after IVF/ICSI following a rFSH/GnRH antagonist protocol with or without oral contraceptive pre-treatment. Hum Reprod 2011;26:3413-23.