Elder K. Human preimplantation embryo selection

Подождите немного. Документ загружается.

HUMAN PREIMPLANTATION EMBRYO SELECTION

perifollicular blood flow and oocyte competence

were related as indicated by embryo development

in vitro and outcome after transfer.

34–39,40,41

In two

large-scale studies in 1997, Chiu et al

and Van

Blerkom et al

characterized individual perifollicu-

lar blood flow rates and the proportion of the fol-

licular circumference in which flow was detected in

several hundred individual follicles. Both studies

came to similar conclusions: perifollicular velocity

was follicle specific; for the ‘typical’ infertile patient

undergoing COHS for IVF, follicles with moderate-

to-high flow indices were often in the minority; and

MII oocytes aspirated from comparatively high flow

follicles had a significantly higher level of compe-

tence when compared with siblings derived from

follicles with perifollicular flow that was low-to-

undetectable, or involved only a portion of the

circumference of the follicle.

Van Blerkom et al

examined the ploidy of MII

oocytes derived from high and low flow follicles

and reported that a significantly higher frequency of

aneuploidy (monosomy and trisomy) occurred in

oocytes aspirated from fully grown follicles with low

or no detectable perifollicular blood flow, and that

after fertilization, oocytes recovered from such folli-

cles showed poor performance in vitro, as indicated

by high frequencies of asymmetric cleavages, toxic

levels of fragmentation and premature arrest. Other

studies have confirmed the follicle-specific nature of

perifollicular velocity and the significant improve-

ment in outcome when oocytes aspirated from high

flow follicles are identified and selected for transfer

as embryos.

36–41

Independent support for IVF out-

come findings was reported in a study of over 700

women undergoing COHS and intrauterine insem-

ination.

Two conclusions applicable for decision

making in infertility treatment can be derived from

these findings.

First, for IVF, treatment should be

halted when perifollicular flow characteristics indi-

cate that few, if any, Graafian follicles have devel-

oped a perifollicular vascular network (during the

follicular phase) that can support high velocity

blood flow. Second, for intrauterine insemination

(IUI), treatment should be halted when multiple

fully grown follicles display high flow characteristics,

owing to the increased probability that the induced

ovulation will produce multiple high-competence

oocytes and multifetal pregnancies.

Measurements of perifollicular blood flow in cycles

of COHS and IVF are made on each day during the

final stages of the follicular phase, or in most pro-

grams, only once on the day of ovulation induction.

This provides an overview of how many follicles in

high, moderate, or low flow categories exist, and can

be used in planning the retrieval such that the

number and location of follicles in each class can be

determined prior to aspiration. This protocol is of

particular relevance when numerous follicles occur

and the oocytes are aspirated under conventional

ultrasound guidance, in a sequence that is based

only on their proximity to each other. Even when

there are large numbers of follicles, oocytes that are

distinguished by their follicular Doppler ultrasound

characteristics can be pooled separately, and their

performance after IVF followed without adding

significant complexity to the laboratory routine of

embryo inspection. Figures 24.1A–D show repre-

sentative color pulsed Doppler images taken at the

time of ovulation induction (following COHS for

IVF) and presented here in gray-scale. Typically,

differential perifollicular blood flow rates are seen

between follicles on the same ovary, and in these

cross-sectional views, high flow at specific regions

around the follicular wall appears in white (see arrows

in Figure 24.1A–C). In contrast, follicles with low

perifollicular blood flow show either focal regions

of detectable flow (asterisk in Figure 24.1B) or no

detectable signal (asterisks in Figure 24.1D). The

presence, absence, or focal nature of a Doppler signal

is confirmed for each follicle by rotating the ultra-

sound probe to survey the entire detectable perimeter

of the follicular wall.

Collectively, results from the above studies of hun-

dreds of stimulated cycles are consistent with respect

to oocyte selection in IVF, as follows: (1) the extent

and velocity of perifollicular blood flow are follicle

specific, and adjacent follicles of comparable size

can display very different characteristics that are not

predicted by conventional ultrasonography; (2) the

number of high flow follicles can be ovary specific,

with most or all residing on one ovary; (3) with the

same protocol of COHS, the number of ‘high flow’



HPE_Chapter24.qxp 7/13/2007 4:54 PM Page 306

OOCYTE SELECTION IN CONTEMPORARY CLINICAL IVF

follicles is cycle specific and does not necessarily

improve by altering stimulation regimen;

and

(4) some infertile patients and women of advanced

reproductive age repeatedly fail to develop any

moderate-to-high flow follicles, regardless of stimu-

lation protocol. In cases where there is a reduced

probability of competent gametes being aspirated

or spontaneously ovulated, with a past history of

repeated IUI or IVF failure, Doppler ultrasonographic

evidence could be used to determine whether treat-

ment should be continued, or whether alternatives

with higher probability outcomes should be consid-

ered, such as IVF with donor oocytes.

This notion was recently supported by Shrestha

et al,

who conducted a retrospective longitudinal

study of perifollicular blood flow rates during the fol-

licular phase of COHS in IVF cycles. Patient responses

could be classified into two groups based on sequen-

tial power Doppler ultrasonographic assessments:

the so-called ‘good beginners’, defined by cycles in



CPA

- 1

- 2

- 3

- 4

Figure 24.1 (A)–(D) Pulsed Doppler ultrasound images of fully grown human follicles in four women after controlled ovarian

hyperstimulation for in vitro fertilization. Although the original follicular scans involve the entire circumference of the

follicular wall, they are shown here in midline. The original color images that depict differential perifollicular blood flow velocity

were transformed into gray-scale with the most intense flow shown in white and noted by arrows. Follicles characterized by a pattern

of perifollicular blood flow that was of low velocity, or that was discontinuous along the follicular wall when viewed in three

dimensions, are indicated by asterisks (see text for details). (E) and (F) Immunofluorescent images of a 24 hour human cumulus

granulosa cell culture showing translocation of the hypoxia-inducible transcription factor HIF1␣ from the cytoplasm ‘C’, in (E) to

the nucleus ‘N’, in (F), within 15 minutes of transfer from normal culture conditions (95% air, 5% CO

) to medium pre-equilibrated

in an atmosphere containing 1% O

(see text for details).

HPE_Chapter24.qxp 7/13/2007 4:54 PM Page 307

HUMAN PREIMPLANTATION EMBRYO SELECTION

which at least one small or medium size follicle(s)

exhibited high grade blood flow rates in the early

follicular phase; and the so-called ‘poor beginners’,

distinguished by the absence of moderate-to-high

grade flow in any small or medium size follicles dur-

ing the early follicular phase. Although individual

oocytes and transferred embryos were not correlated

with specific follicles, the authors reported that

perifollicular blood flow rates in the early follicular

phase were associated with graded follicle-specific

flow rates at ovulation induction and with outcome

after embryo transfer in IVF. In both the early and

late follicular phase, clinical pregnancy rates were

significantly higher in the ‘good beginner’ group

compared with those classified as ‘poor beginners’,

and notably, this included relatively older patients

(median age 40 years) who achieved a comparatively

high pregnancy rate (37%) and low pregnancy loss

rate (11%).

The pattern of ‘good beginners’ was so distinct

that Shrestha et al

proposed that it may be possible

to use Doppler analysis of perifollicular blood flow

patterns to identify good and poor prognosis IVF

cycles at an early stage of COHS; if the pattern is

suboptimal, this could provide a biological justifi-

cation for early cancellation. In contrast, Palomba

et al

reported that power Doppler assessments do

not seem to be clinically useful for oocyte selection

in IVF cycles for young non-obese infertile women

with low basal FSH levels. In this study, oocytes were

chosen for insemination based on perifollicular vas-

cularity, and the outcome was compared with a con-

trol group where ‘standard’ morphological criteria

alone were used to select oocytes for insemination

and embryos for transfer. While this study con-

firmed previous findings that poor perifollicular

flow characteristics and poor embryo performance

in vitro were closely related,

this method of follic-

ular analysis was problematic for these investigators,

due to the higher number of oocytes with optimal

quality (and vascularization) that were available for

insemination. As they noted,‘in almost all cases’ the

embryos transferred had developed from oocytes

obtained from follicles with high-grade vasculariza-

tion, which led to the suggestion that more refined

ultrasound criteria may be required to distinguish

between high-grade (i.e. high velocity) follicles if

this is to be used for the assessment of oocyte and

embryo competence.

Merce et al

offers a more sanguine conclusion

concerning the clinical usefulness of follicular imag-

ing, by demonstrating that three dimensional (3D)

ultrasonography combined with power Doppler

angiography on the day of ovulation induction

predicted gestation in 76% of IVF patients. As they

note, 3D Doppler angiography depicts all the ovar-

ian and follicular blood vessels and can be used to

derive a vascularization index and a blood flow index

from the whole ovarian volume, as well as high-

resolution vascular and flow imaging of individual

follicles. With this higher resolution approach to

Doppler imaging of stimulated ovaries, Merce et al

reported a pregnancy rate of 30% when no embryos

from high-grade (i.e. flow) follicles were transferred,

but the miscarriage rate was 67%. In contrast, trans-

fer of a single embryo from a high-grade follicle was

associated with a pregnancy rate of 35% and a mis-

carriage rate of zero. Whether the inclusion of 3D

power Doppler angiography provides the ability to

discriminate between follicles classified as high grade

by 2- and 3D color pulsed Doppler ultrasonogra-

phy, as described above, remains to be determined.

However, the very high miscarriage rate observed by

Merce et al

with embryos derived from low grade

follicles is consistent with the very high frequency

of aneuploidy reported by Van Blerkom et al

for

oocytes aspirated from such (low flow) follicles.

The preponderance of evidence from color, power,

and most recently 3D Doppler imaging of ovary and

follicle-specific perifollicular blood flow and vascu-

larity demonstrates a clear association with devel-

opmental competence and outcome after IVF. The

clinical usefulness of this protocol has been demon-

strated with patients classified as ‘poor beginners’,

or at the terminal stage of the follicular phase in

IUI

and IVF cycles,

34–41,43–46

when only low-grade

follicles (low-to-undetectable perifollicular flow) are

detectable. Taken together, these outcome-based find-

ings support a conclusion that was advanced by Chui

et al

and Van Blerkom et al

nearly a decade ago;

namely, that perifollicular blood flow rates are a very

positive non-invasive indicator of oocyte and embryo



HPE_Chapter24.qxp 7/13/2007 4:54 PM Page 308

OOCYTE SELECTION IN CONTEMPORARY CLINICAL IVF

competence. While not absolute, Doppler imaging

should have a central role in the decision making

process as to whether a treatment cycle should be

continued or stopped, and for some patients,

whether alternatives to assisted reproduction with

patient oocytes need to be considered.

At present, the utility of Doppler ultrasound meas-

urements is controversial for the so-called ‘poor res-

ponder’ class of infertility patients, i.e. women who

develop comparatively few antral follicles and have

relatively low serum estradiol levels after COHS.

Pan et al

suggested that the reduced sensitivity of

poor responders to gonadotropin stimulation may

be associated with changes in ovarian stromal blood

flow detectable by 3D power Doppler ultrasonogra-

phy. These investigators reported marked differences

in blood flow patterns and velocities between

normoresponders and poor responders that were

reflected by their outcome after IVF. Because estro-

gen is known to bind to its receptors on vessels and

initiate the release of nitric oxide, which increases

vasodilation and blood flow,

48,49

Pan et al

pro-

posed that an estrogen deficit in poor responders

may be one of the factors contributing to signifi-

cantly reduced rates of perifollicular blood flow. In

contrast, Kan et al

reported no significant differ-

ences in implantation, clinical pregnancy, and live

birth rates among poor responders, with and with-

out high-grade perifollicular vascularity; patients in

the former category tended to have higher multiple

and live birth rates, and a lower frequency of mis-

carriage. However, too few women were included in

this study to achieve statistical significance, and no

correlations between specific follicles, oocytes, and

embryos were made. The poor responder is not only

common in assisted reproduction, but is also one

that is often difficult to treat because of the high rate

of recurrence despite different protocols of endocrine

management. Additional Doppler studies of the

poor responder are warranted, and, in particular,

may determine if the presence of a single high-grade

follicle could be grounds for a measure of optimism

with respect to outcome.

While the above evidence clearly shows the im-

portance of ovarian and follicular imaging in con-

temporary assisted reproduction in general, and in

clinical IVF in particular, a question that remains to

be answered is whether this technology should be

extended to fertile women who may be at high risk

for ovulating aneuploid MII oocytes. These patients

include women of advanced reproductive age and

those who chronically miscarry. It might also include

women of any age with severe anxiety owing to a

previous conception that was aneuploid. Doppler

imaging of ovaries in natural menstrual cycles may

be of significant value in this regard, if it can be

shown to identify cycles in which conception should

be avoided. Validation for its application in these

instances will require confirmation that the occur-

rence of low-grade follicles is associated with an unac-

ceptably high probability that the corresponding

oocyte is chromosomally abnormal.

MECHANISMS OF REGULATION

OF INTRAFOLLICULAR OXYGEN

AND PERIFOLLICULAR BLOOD FLOW

Although Doppler ultrasonography and power angi-

ography can identify perifollicular blood flow char-

acteristics and vascularization, that are unique to

each follicle, these imaging and analytical methods

should not be considered to be absolute predictors

of competence. Oocytes with chromosomal defects

and embryos that prematurely arrest or develop

abnormally in vitro can originate from follicle(s)

judged as high grade,

35,40,42

albeit at a much lower

frequency than their low grade/flow counterparts.

The relationship between high competence for the

oocyte and perifollicular blood flow is also not under-

stood. However, when the dissolved oxygen content

of follicular fluid was measured at ovum retrieval,

high frequencies of aneuploidy were detected in

oocytes that originated from poorly oxygenated

follicles that also showed poor flow characteristics

during the follicular phase.

A preliminary study of

the average intracellular pH (pHi) in MII oocytes of

normal appearance obtained from follicles on the

same ovary showed that pHi levels (~6.8–6.9) in

oocytes derived from low flow follicles with a dis-

solved oxygen content ⱕ1%, were below those

measured in sibling oocytes (~7.2) that had been



HPE_Chapter24.qxp 7/13/2007 4:54 PM Page 309

HUMAN PREIMPLANTATION EMBRYO SELECTION

aspirated from high flow follicles with dissolved oxy-

gen contents ⱖ3%.

35,51

Although these early findings

require confirmation, they suggest that a slight reduc-

tion in pHi could result from inadequate cytoplasmic

buffering capacity or inability to effectively metabo-

lize lactic acid.

Gaulden

proposed that slight reductions in

human oocyte pHi could retard or perturb the normal

dynamics of microtubule polymerization/depoly-

merization, leading to asymmetric and malformed

meiotic spindles with a reduced ability to capture,

retain, or normally segregate chromosomes. In her

morphological studies of Graafian follicles in women

of advanced reproductive age (who were more prone

to have children with Down’s syndrome), she observed

that the perifollicular capillary bed in older women

was underdeveloped compared with the situation in

younger women, and that it tended to be further

from the approximate center of the follicle. This too

requires confirmation, but the suggestion is supported

by coincident immunofluorescent studies of spindle

microtubules and DNA fluorescent staining of MII

chromosomes in MII oocytes obtained from fully

grown follicles with poor blood flow characteristics

and low intrafollicular oxygenation. These studies

demonstrated abnormalities in spindle organization

and morphology, and high frequencies of chro-

mosomal malalignment and displacement, respec-

tively.

35,41

In a study of naturally cycling women

ⱖ40 years of age, Battaglia et al

reported that about

80% of MII oocytes of normal appearance after aspi-

ration from single dominant follicles showed spindle

abnormalities that would be likely to result in aneu-

ploid embryos if fertilized. The apparent physiolog-

ical relationship between oocyte pHi and levels of

intrafollicular oxygen needs to be investigated in

detail, especially with respect to metabolic pathways

that may contribute to the acidification of the

ooplasm in a low oxygen environment. In this regard,

a recent mathematical modeling of oxygen con-

centrations in bovine and murine cumulus–oocyte

complexes concluded that the cumulus cells proba-

bly remove little oxygen, suggesting that the concen-

tration of dissolved oxygen measured in follicular

fluids reflects what is experienced by the oocyte

itself.

If confirmed by experimental studies, follicle-

specific differences in oxygen content measured in

human follicular aspirates at ovum retrieval

35,38

could

reflect the unique capacity of each follicle to respond

to hypoxia

and upregulate angiogenesis in the

perifollicular capillary bed, as discussed below. For

the corresponding oocyte, these differences may

have development consequences for competence at

the molecular, cellular, and nuclear levels.

35,51

The above findings suggest a physiological basis

for compromised oocyte competence that correlates

with follicle-specific information obtained by non-

invasive Doppler studies that can be undertaken rou-

tinely during the follicular phase. The results obtained

to date strongly suggest that MII oocytes from mod-

erate-to-high flow follicles have a significantly higher

bias for normal competence than do those obtained

from follicles with low or no detectable perifollicu-

lar flow. Given the reported correlations between

outcome and oocyte/embryo selection schemes that

have incorporated perifollicular blood flow charac-

teristics, it is curious that Doppler imaging has not

become widespread in contemporary clinical IVF.

For many programs, the added expense, time, and

operator expertise associated with real-time Doppler

ultrasound imaging could be prohibitory. In other

instances, current laboratory paradigms of selection

that include embryo morphology and performance

are satisfactory, and the potential benefits of follicu-

lar monitoring are not deemed sufficient to warrant

change. However, recent restrictions on the num-

ber of oocytes that can be inseminated or embryos

transferred, without the option of cryopreservation

or preimplantation genetic diagnosis, have placed

greater importance on oocyte selection and the need

to improve the sensitivity of the criteria used. As a

result, the recent re-investigations of Doppler ultra-

sound assessments of perifollicular blood flow have

largely validated previous findings with respect to

the occurrence and distribution of high and low

flow follicles, and more importantly, have been suf-

ficiently related to outcome, on a per oocyte basis, to

seriously warrant their inclusion in selection schemes

for IVF. The use of this non-invasive technology in

clinical IVF may be encouraged by enhancements in

resolution and imaging capability incorporated into

newer generation ultrasound instruments, including



HPE_Chapter24.qxp 7/13/2007 4:54 PM Page 310

OOCYTE SELECTION IN CONTEMPORARY CLINICAL IVF

those specific for gynecological use that offer views

of perifollicular blood flow and vascularity in three

dimensions. While the above findings support the

notion that rates of perifollicular blood flow reflect

differences in the underlying biology of the follicle

that appear to be of direct relevance for the develop-

mental competence of the oocyte, several impor-

tant questions remain. For example, if the velocity

of perifollicular blood flow indicates normal follic-

ular function, why do rates differ between follicles,

and most importantly, what specific intrafollicular

influences could have a positive or negative impact

on this activity? Since the follicle is both a source

of, and a ‘sink’ for a wide variety of bioactive mole-

cules and growth factors, the rate of transfollicular

trafficking between the systemic circulation and

the intrafollicular milieu might be higher in follicles

with a well-developed perifollicular microvascula-

ture, consistent with a comparably high velocity flow.

This could explain the association between higher

levels of intrafollicular oxygenation and perifollicu-

lar blood flow rates. However, it should be emphasized

that expansion of human perifollicular microvas-

culature under gonadotropin stimulation has not

been demonstrated directly, but rather, inferred from

upregulated blood flow rates

and associations with

the dissolved oxygen content of preovulatory follicu-

lar fluid. In this regard, increased blood flow resulting

from the increased patency of pre-existing perifollicu-

lar vessels could be a direct effect of estrogen synthe-

sized by the mural granulosa, a cellular layer that is

usually separated from the capillary bed by the follic-

ular wall, and a space that is measured in microns.

Increased vascularization of the existing bed by angio-

genic induction is another means by which perifollic-

ular blood flow and capillary permeability may be

significantly increased over a relatively short period of

time during follicular stimulation, as discussed below.

ANGIOGENIC FACTORS IN FOLLICULAR

FLUID

Human follicular fluid is a ‘factorologists’ dream

owing to the wide range of growth and other regu-

latory molecules that are synthesized by follicular

granulosa cells or which enter the follicle from exter-

nal sources (for review see references 56–58). With

respect to perifollicular blood flow, however, the

production of potent angiogenic factors by granu-

losa cells, such as vascular endothelial growth factor

(VEGF) and leptin may be particularly relevant in

understanding follicle-specific blood flow rates.

58–63

Analysis of these angiogenic factors in human follic-

ular fluid has shown that levels are follicle specific

and has led to the notion that their export may

induce a rapid expansion of the perifollicular vascu-

lature by stimulating angiogenesis from existing

endothelial cells.

35,38

VEGF is also expressed by

thecal cells, and this has a proximal source which could

stimulate the expansion of perifollicular endothelial

cells.

VEGF expression is regulated by hypoxia in

general

and by the hypoxia inducible transcription

factor (HIF-1) pathway in particular.

66,67

This path-

way can in turn be upregulated by certain reactive

oxygen species normally produced by mitochondria

(e.g. superoxide),

which are the organelles involved

in sensing oxygen at the cellular level.

The possibility that the oxygen sensing ability of

mural and cumulus granulosa cell mitochondria is

upregulated during gonadotropin (FSH) stimulation

can be tested.

According to this hypothesis, diffu-

sion of oxygen into the developing follicle from the

existing perifollicular capillary bed is first ‘sensed’

by mitochondria in the gonadotropin-stimulated

mural granulosa layer, which responds by increasing

levels of superoxide that is converted to peroxide

within the cytosol. Bell et al

proposed that peroxide

upregulates HIF-1 expression, which in turn upreg-

ulates the expression of a cohort of regulatory genes,

including VEGF. Indeed, the expression of this tran-

scription factor has been detected in mural and

cumulus granulosa cells in human preovulatory fol-

licles, and differences in mRNA levels between high

and low flow follicles on the same ovary have been

reported.

The HIF-1 pathway is commonly used

by cells that recognize and respond to hypoxia by

upregulating levels of angiogenic factor production.

It seems likely that such a mechanism may be

involved in increasing perifollicular vascularity or

the patency of the existing vasculature, or both, in

order to increase intrafollicular oxygen.



HPE_Chapter24.qxp 7/13/2007 4:54 PM Page 311

HUMAN PREIMPLANTATION EMBRYO SELECTION

The speed with which HIF-1␣ translocates from

the cytoplasm to the nucleus in human cumulus cells

cultured in a hypoxic environment is illustrated in

Figure 24.1E and F. Figure 24.1E shows the typical

pattern of diffused cytoplasmic HIF-1␣ immuno-

staining in cumulus cells cultured on glass coverslips

in a standard atmosphere of 95% air 5% CO

,24

hours after retrieval from a high flow follicle.Within

15 minutes following transfer to medium that had

been pre-equilibrated in an atmosphere containing

1% O

(Figure 24.1F), intense nuclear HIF-1␣ im-

munofluorescence is detected, with virtually no sig-

nal evident in the cytoplasm. A similar response is

observed with granulosa cells derived from the

mural layer, and from cells obtained from follicles

with low or poor perifollicular circulation (Antczak

and Van Blerkom, unpublished). When returned to

normal atmospheric conditions, cytoplasmic HIF-

1␣ immunofluorescence becomes pronounced as the

nuclear signal diminishes. These findings confirm

previous molecular analyses demonstrating the pres-

ence of HIF-1 mRNA in human granulosa cells

and provide direct evidence for the expression of a

hypoxia-sensing mechanism in the human ovarian

follicle.

The association between follicle-specific

differences in perifollicular blood flow, the dissolved

oxygen content of follicular fluid, and levels of

angiogenic factor biosynthesis that may be regu-

lated by the HIF pathway warrants further investi-

gation. At present, it is an intriguing possibility that

failure to express HIF-1␣, or an inability to translo-

cate it to the nucleus, may contribute to the occur-

rence of developmentally incompetent oocytes in

women who rarely, if ever, develop follicles with

high perifollicular blood flow characteristics.Whether

age-associated mitochondrial dysfunctions compro-

mise their ability to sense or respond to hypoxia also

warrants further study.

Ovarian pathologies, such as ovarian hyperstimu-

lation syndrome (OHSS), are associated with intrafol-

licular levels of angiogenic promoters (e.g. VEGF)

that are often significantly higher than levels measured

in the follicular fluids of normal women undergoing

COHS.

The inability to downregulate angiogenic

factor expression can lead to excessive perifollicular

vascular permeability, which results in capillary leak-

age and the third-spacing of fluids that is character-

istic of this abnormal response to gonadotropin

stimulation, especially in women with polycystic

ovarian syndrome. Unusually high VEGF levels have

been measured in the follicular fluids of young

infertility patients (non-PCOS) and women of

advanced reproductive age who showed no indica-

tions of OHSS during or after COHS for IVF.

While the etiology of this atypical expression pat-

tern is unknown, it may result from an angiogenic

receptor defect that renders perifollicular endothelial

cells refractory to this promoter(s) and consequently,

the affected follicle(s) may be unable to upregulate

oxygenation or downregulate HIF-1 activity (and the

attendant expression of other growth factors regu-

lated by this factor pathway).

41,51

How each follicle recognizes and responds to

hypoxia and the extent to which, if any, this has an

impact on the viability of the oocyte remains to be

investigated. However, many of the same factors are

involved in the postovulatory development of a well-

vascularized corpus luteum. In this respect, follicle

competence is a periodic process that begins during

folliculogenesis and continues after ovulation.

56,59–61

At present, quantification of the levels of angiogenic

factors such as VEGF or leptin on a follicle-specific

basis to assess oocyte and embryo competence, is is

not advisable. Van Blerkom et al

reported that fol-

licles with different perifollicular blood flow rates

and intrafollicular dissolved oxygen contents tend

to be associated with certain ranges of VEGF con-

centrations, but the interfollicular levels in each group

(low-to-moderate-to-high flow) were not consid-

ered sufficiently significant or sensitive to use as

an independent indicator of oocyte competence. A

similar conclusion has been reported for normo-

responders

and poor responders.

However,

when taken together, these findings begin to offer a

biological context for follicle-specific differences that

may be related to fertility in general, and the mater-

nal-age related reduction in fertility, in particular.

They also provide a molecular basis for investigat-

ing the mechanisms by which the potential influences

of certain intrafollicular signaling pathways may



HPE_Chapter24.qxp 7/13/2007 4:54 PM Page 312

OOCYTE SELECTION IN CONTEMPORARY CLINICAL IVF

have adverse or beneficial effects on oocyte com-

petence.

FOLLICULAR BIOCHEMISTRY – MINING

FOLLICULAR FLUID FOR COMPETENCE

RELATED DETERMINANTS

The expression levels of certain regulatory genes

appears to be follicle specific,

and this emphasizes

a central issue in human follicular biology: the res-

ponse of each follicle to stimulation can be unique,

and is often unpredictable in terms of whether or

how it may affect the developmental competence of

its oocyte. However, the identification of specific

follicular components that can be correlated with

competence and combined with findings from non-

invasive measures, such as Doppler ultrasound, would

go a long way towards providing objective and stan-

dard parameters for oocyte/embryo selection in IVF.

Follicular fluid is a complex substance, and investi-

gation of the different growth factors, cytokines and

other bioactive and regulatory molecules that may

influence oocyte developmental competence adds

its own layer of complexity to the clinical IVF labo-

ratory: each assay must be validated, standardized,

and quality controlled. If certain molecules are found

to be competence related, the results need to be

available in a timely manner for oocyte selection,

i.e. before the consequences of the natural decay in

competence defeat the purpose of establishing a

biochemical basis for selection (e.g. spontaneous

degradation of normal meiotic spindle organization

and chromosomal segregation capacity).

Thus,

the choice of factors for analysis, either singly or in

combination, must meet both stringent laboratory

criteria and, more importantly, have a demonstrable

association with outcome. It should also be noted

that regulatory factors detected in follicular fluid may

largely represent paracrine and autocrine processes

that reflect the heterogeneous activities of the somatic

cell component of the follicle; influences on oocyte

competence, if any, may be indirect.

Since the time that clinical IVF became an estab-

lished treatment for infertility, a sizable literature

has developed on follicular biochemistry and its

relationship to outcome. However, as is often the

case, contradictory and equivocal results populate

this landscape, and upon closer examination, many

promising leads for oocyte and embryo selection are

not confirmed, or their suggested sensitivity is not val-

idated (for review see reference 57). However, there

is growing evidence that quantification of certain

bioactive molecules may provide a positive bias for

selection with sensitivity levels that warrant incor-

poration into the clinical IVF laboratory.

CORTISOL : CORTISONE RATIO

The ratio of cortisol to cortisone stands out as one

of the more consistent and reproducible markers of

oocyte and embryo competence, the expression of

which is upregulated by LH and hCG,

which upreg-

ulate expression of the enzyme 11␤-hydroxysteroid

dehydrogenase (11␤-HSD), which converts the active

glucocorticoid cortisol into cortisone, an inactive

steroid. Keay et al

reported that significantly higher

pregnancy rates occurred at higher intrafollicular

cortisol : cortisone ratios, and suggested that the

presence of the active form in follicular fluid may

influence the final stages of preovulatory oocyte

maturation. The increased levels of cortisol in folli-

cles from which competent oocytes were retrieved

may reflect an early progesterone effect that results

from the initial stages of transformation of granu-

losa cells to luteal cells. Progesterone inhibits 11␤-

HSD activity, and therefore inhibits the conversion

of cortisol to cortisone.

In a study of 80 IVF cycle

outcomes,

a cortisol : cortisone ratio of ⬎11.2 was

associated with pregnancy, while unsuccessful pro-

cedures occurred with ratios of ⬍11.2. In support of

a developmentally significant relationship between

cortisol/cortisone levels and competence, Michael

observed that other follicular factors suggested to

predict outcome actually mediate levels of 11␤-HSD

expression and activity. On a follicle-specific basis,

these can indirectly determine the ratio of the active

to inactive forms of this steroid.

Higher cortisone levels appear to be associated

with preovulatory luteinization of the mural granulosa,



HPE_Chapter24.qxp 7/13/2007 4:54 PM Page 313

HUMAN PREIMPLANTATION EMBRYO SELECTION

which may influence the production of regulatory

factors within the follicle that promote oocyte

competence during the terminal stages of oogenesis.

Michael

offered the following explanation as to how

these steroid hormones could exert an influence on

the oocyte: at the end of the follicular phase, higher

cortisol levels resulting from the inhibition of

11␤-HSD activity by progesterone and other folli-

cle-specific agents, may facilitate the closure of com-

municating (gap) junctions between the oocyte and

cumulus/corona transzonal processes (for review see

reference 4). After closure, the oocyte is no longer

under direct maternal regulation, and the process of

preovulatory nuclear (i.e. meiotic) and cytoplasmic

maturation required to produce a competent gamete

is initiated. This could explain why competent oocytes

may be more likely to originate from follicles in

which progesterone-mediated luteinization begins

prior to ovulation. Whether the cortisol : cortisone

ratio influences perifollicular blood flow rates and

levels of angiogenic promoters synthesized by follic-

ular granulosa cells warrants investigation. However,

the 11␤-HSD-cortisol : cortisone pathway appears

to be a common denominator in numerous interac-

tions between important regulatory factors (inhibitors

and promoters) in follicular fluid the levels of which

have been associated with competence.

If the superior outcomes observed at the high

ratios are confirmed in large-scale studies, and the

mechanisms by which oocyte competence is influ-

enced clearly identified, the inclusion of this assay

in the clinical IVF laboratory will merit very serious

consideration. However, as with any assay that

requires follicular specificity, including those

described below, modifications in the protocol of

aspiration would have to be incorporated such that

the integrity of each aspirate is maintained in order

to detect unique chemical characteristics. This may

prove to be difficult to apply on a routine basis when

many follicles exist, and, because of high patient load,

oocyte retrievals are done rapidly and in a sequential

manner, based on proximity and without intervening

washes of the aspiration equipment. Thus, to warrant

use in the clinical IVF program, the burden of proof

will require unambiguous evidence that is based on

outcome.

PROTEIN GROWTH FACTORS

ANTI-MULLERIAN HORMONE

Protein markers of competence would seem to be a

‘laboratory-friendly’ approach to oocyte selection if

they can be detected and quantified by immunolog-

ically based ‘kits’that are currently available. Despite

the diversity of protein growth factors, cytokines,

and hormones identified in follicular fluid, includ-

ing those known to be products of granulosa cells,

few of the suggested protein markers of competence

(e.g. VEGF), whether assessed individually or col-

lectively, have proved to be sufficiently sensitive and

specific to permit definitive determinations of com-

petence. However, one regulatory protein, Mullerian

inhibiting hormone (homologous to its male coun-

terpart, anti-Mullerian hormone, or AMH), a dimeric

glycoprotein member of the transforming growth

factor (TGF)-␤ superfamily, currently shows prom-

ise in this regard. This factor is expressed by granu-

losa cells when the primordial follicle enters into the

growth phase, and is upregulated at the transcrip-

tional and translational levels in large pre-antral

and small antral follicles.

Several studies indicate

that serum AMH is an independent indicator of the

ovarian reserve (ovarian aging) that seems to effec-

tively predict a patient’s unique response to COHS

in general, and in IVF cycles, in particular.

77–79

Whether AMH is an indicator of oocyte compe-

tence is controversial, since in some animal systems

its expression is markedly downregulated in FSH-

sensitized follicles that are in the growing antral and

preovulatory stages.

However, AMH is detected in

the serum and follicular fluids of women during

COHS and in preovulatory follicles for at least 34

hours after the administration of an ovulatory dose

of hCG.

80,81

Results from several recent clinical studies indi-

cate that serum AMH levels measured before,

during follicular stimulation,

or on the day of ovu-

lation induction with hCG,

can be highly predic-

tive of the ovarian reserve and follicular response to

stimulation, measured by the number of preovula-

tory follicles that develop, and normal stage-appro-

priate embryo development in vitro. AMH levels



HPE_Chapter24.qxp 7/13/2007 4:54 PM Page 314

OOCYTE SELECTION IN CONTEMPORARY CLINICAL IVF

measured in follicular fluid at ovum retrieval

also seem to predict normal embryo development

in vitro.

These results strongly suggest that deter-

minations of AMH may be a significant predictor of

oocyte competence. Indeed, Eldar-Geva et al

con-

cluded that of all the protein factors currently sug-

gested to be markers of competence, AMH may be

the only reliable predictor of outcome that has

currently been identified in follicular fluid.

The finding that AMH levels are follicle specific

and correlate with outcome when combined with

levels measured in serum during the follicular and

early luteal phase

is an important one, because it

strongly suggests that measurements of AMH in

fluid aspirates from stimulated follicles may be an

indication of normal intrafollicular conditions that

produce a competent oocyte. Confirmation of this

very promising lead will require following embryo

development in vitro and fate after transfer on a

follicle-specific basis. Follicle specific differences in

cortisol : cortisone ratios, levels of HIF1 expression,

and perifollicular blood flow patterns have each been

suggested to indicate the development of a healthy

follicle with a competent oocyte; whether these are

related to AMH concentrations warrants detailed

investigation. A strong association between perifol-

licular velocity characteristics, AMH levels measured

in the corresponding follicular fluid, and outcome

after embryo transfer, could lead to a common assess-

ment for most infertility patients undergoing COHS

and IVF, namely, Doppler ultrasonography as the

primary non-invasive means of predicting, whether

competent oocytes exist prior to retrieval, and in

which follicles they are most likely to reside.

INHIBINS

Inhibins, specifically inhibin B, are another class

of protein factors the follicle-specific levels of which

have been related to competence. These heterodi-

meric glycoproteins in the TGF-␤ superfamily con-

sist of a common ␣-subunit and one of two possible

␤-subunits (A and B) linked by disulfide bonds.

Inhibins are secreted by granulosa cells, but inhibin

B is expressed only in developing preantral and antral

follicles. Inhibin B levels increase quickly during the

early follicular phase, peak at the early to mid-fol-

licular phase, and then progressively decline until

just after the LH surge. A transient spike in the level

of inhibin B is detected 2 days after the LH surge,

but then declines for the remainder of the luteal

phase.

Inhibin B is part of the hypothalamic–

pitutary–ovarian feedback mechanism that inhibits

FSH secretion as follicles develop during the follicu-

lar phase.

Quantitative assays initially measured both inhibin

A and B in the follicular fluid of spontaneously cycling

women,

and high levels were correlated with the

fertilizing capacity of the corresponding oocytes.

Later assays distinguished A from B, and serum

inhibin B levels in cycles of COHS were reported to

positively correlate with the number of follicles and

oocytes obtained for IVF, as well as with the risk of

developing ovarian hyperstimulation syndrome.

Follicular fluid inhibin B levels were also correlated

with oocyte quality and normal embryo development

on days 2 and 3 of culture.

87,88

Inhibin B is produced

by the cumulus granulosa and its occurrence in

serum at high levels on the day of follicular aspira-

tion appears to positively correlate with outcome in

cycles of COHS and IVF.

Inhibin B assays tend to

be difficult to perform on a routine basis and as a

result, it remains to be determined whether the

results are more predictive with respect to outcome

than assays of AMH.

OTHER CONSIDERATIONS OF FOLLICULAR

BIOLOGY RELATED TO COMPETENCE

EMPTY FOLLICLE SYNDROME

The current emphasis on the identification of

bioactive molecules in follicular fluid is related to

the need for unambiguous criteria for gamete selec-

tion in IVF.

However, IVF cycles occur in which no

oocytes are recovered. This phenomenon, termed

‘empty follicle syndrome’, is applied when no oocytes

are found despite repeated, vigorous, and meticu-

lous aspiration of follicles. Generally, the rate of

follicular growth during the follicular phase is nor-

mal, and serum estradiol levels are stage-appropriate



HPE_Chapter24.qxp 7/13/2007 4:54 PM Page 315