Markers of Ovarian Reserve

The most widely used endocrine marker for ovarian reserve is the early follicular phase FSH
level.
[10]
Early follicular !asal" FSH level has !een shown to !e an independent predictor of #$F
outcome%
[&'%&(]
and is a stron)er predictor of cycle cancellation due to poor response and the
num!er of oocytes collected at pick*up. #n contrast% a)e appears to !e more closely related to the
chance of implantation and on)oin) pre)nancy.
[+0%+(%&,]
This was illustrated !y the findin) that
women older than -0 years with normal !asal FSH levels . 1' /01" may demonstrate lower
cancellation rates% !ut the implantation rate per em!ryo and the on)oin) pre)nancy rates are
lower than those o!served in youn) women with elevated !asal FSH levels.
[&2]
3oreover% recent
studies have indicated that youn) women with hi)h FSH levels demonstrate lower num!ers of
)rowin) follicles and a hi)h pro!a!ility of cycle cancellation% !ut can achieve )ood on)oin)
pre)nancy rates if oocytes and em!ryos are o!tained.
[&2]
The com!ined and independent effects
of increased !asal FSH levels and a)e on #$F outcomes can !e illustrated )raphically
[&4]

The independent and com!ined affect of a)e and !asal follicle*stimulatin) hormone FSH" levels
on delivery rate per cycle 5" derived from a theoretical model is illustrated. #n youn) women% the
success rate is )reater than that in older women unless the !asal FSH is hi)hly elevated.
6dapted from Toner 78. 3odest follicle*stimulatin) hormone elevations in youn)er women9 warn
!ut don:t dis;ualify. Fertil Steril +00-<291-4&*1-4'.

=iscrepancies !etween the predictive value of FSH for ovarian response to ovarian stimulation
and for on)oin) pre)nancy illustrate the limitations of tests of ovarian responsiveness as markers
of ovarian reserve. 6lthou)h response to stimulation reflects the ;uantity% pre)nancy rate reflects
!oth the ;uantity and ;uality of the oocytes.
[-0]
The num!er of follicles or oocytes o!tained after
ovarian hyperstimulation is therefore merely a surro)ate for ovarian reserve. This differentiation
was emphasi>ed !y the conclusions of a recent meta*analysis% in which !asal FSH levels showed
a moderate predictive performance for poor response !ut a low predictive performance for
nonpre)nancy
[-1]
. 6 recent retrospective analysis showin) that on)oin) pre)nancy rates of +25
may !e achieved in re)ularly cyclin) women with !asal FSH levels a!ove 1' /01 su))ests that
;uality can compensate for ;uantity of oocytes when the latter is diminished in youn)er women.
[-+]
#t is therefore su))ested that youn) patients should not !e e?cluded from #$F treatment purely on
the !asis of increased !asal FSH levels. @aution is certainly re;uired when interpretin) !asal
FSH levels in a clinical conte?t. #nconsistencies may arise from the wide interindividual variation
that e?ists in follicular phase FSH concentrations in the normo*ovulatory cycle.
[-&]
3oreover% cycle*
to*cycle variations in !asal FSH limit the relia!ility of a sin)le measurement. Ahen more than one
measurement is made% it appears that the hi)hest value is of most pro)nostic si)nificance
Beceiver operatin) characteristic curves of studies reportin) on the performance of !asal follicle*
stimulatin) hormone FSH" to predict poor ovarian response 6" and nonpre)nancy C". The
curves indicate that althou)h FSH is a moderate predictor for ovarian response% it is a poor
predictor for nonpre)nancy.

Estradiol is fre;uently measured at the same time as !asal FSH. Ahen elevated on cycle day &% it
appears to predict poor response to ovarian stimulation for 6BT% even when !asal FSH levels are
normal.
[-(]
#t is su))ested that elevated early follicular phase estradiol levels may indicate an
inappropriately advanced sta)e of follicular development% consistent with ovarian a)in). However%
it may simply reflect the presence of functional ovarian cysts.
[-,]
#nhi!in C is an alternative endocrine marker that has !een postulated. #t has !een su))ested that
measurin) inhi!in C concentrations would provide a more direct assessment of ovarian reserve%
)iven that in contrast to FSH" it is directly produced !y developin) early antral follicles.
[-2%-4]

#nhi!ins are dimeric )lycoproteins consistin) of an D su!unit linked throu)h disulfide !indin) with
either a E6 inhi!in 6" or EC inhi!in C" su!unit. #nterest in inhi!in as a marker of ovarian a)in) has
revived in recent years with the development of en>yme*linked immunosor!ent assays capa!le of
distin)uishin) !etween these two heterodimers.
['0]
#t is now clear that elevated early follicular
phase FSH concentrations are associated with a si)nificant decrease in inhi!in.
['1]
#nitial studies
reported an association !etween diminished ovarian response durin) #$F and decreased serum
inhi!in C levels.
['+%'&]
However% later studies were una!le to confirm these findin)s
[-2%'-%'']
and studies
of the value of measurin) inhi!in C on cycle day ' showed discordant results.
['(%',]
Becently% it was
demonstrated in a multivariate analysis that addition of !asal FSH and inhi!in C levels to a
lo)istic model includin) ultrasound characteristics can improve the performance of the model.
['2]
The ultrasound parameter in ;uestion was the antral follicle count. The num!er of antral follicles
in the early follicular phase assessed !y ultrasound
['4]
has !een found to decrease with advancin)
a)e and this is thou)ht to represent ovarian a)in). Becent data indicate that the num!er of antral
follicles present on cycle day & provides a !etter sin)le pro)nostic indicator for poor response
durin) #$F than the patient:s a)e or any other endocrine marker.
['2%(0]
Fvarian volume% which partly
reflects the num!er of ovarian follicles% has also !een shown to decrease with a)e%
[(1]
and several
studies have su))ested a role for this parameter as a marker of ovarian reserve.
[(+%(&%(-]
Several challen)e tests have !een desi)ned that offer the theoretical !enefit of measurin)
ovarian reserve in a dynamic% and perhaps more clinically representative way. The clomiphene
citrate challen)e test @@@T" involves the administration of 100 m) clomiphene citrate on days '
to 4% and the measurement of FSH levels on days & and 10. 6n a!normal test is defined as an
a!normally hi)h FSH on day 10.
[('%((]
#t has !een su))ested that the @@@T may !e !etter than
!asal FSH for predictin) infertility treatment outcome !ecause two levels of FSH are o!tained%
and the addition of clomiphene citrate may serve to reveal women who mi)ht not !e detected !y
!asal FSH screenin) alone.
[(,]
#n a recent meta*analysis% !asal FSH and the @@@T were found to
!e of similar value in predictin) a clinical pre)nancy in women under)oin) infertility treatment.
[(2]

Aith either test a normal result was of little predictive value% !ut an a!normal result predicted poor
outcome from infertility treatment. Given that the @@@T offers no clear advanta)e compared with
a sin)le !asal FSH measurement% and is itself associated with potential adverse effects% !asal
FSH is preferred.
The e?o)enous FSH ovarian reserve test was desi)ned to screen for )ood or poor response in
#$F.
[(4]
The chan)e in estradiol levels in response to &00 / of FSH administered on cycle day &
provides the dynamic component to this test and has recently !een shown to demonstrate less
intercycle varia!ility than !asal FSH levels.
[,0]
However% the value of this parameter as an
independent varia!le predictive of #$F outcome has not !een evaluated fully. 6 test similar in
concept is the GnBH a)onist stimulation test% which evaluates the chan)e in estradiol levels after
+ or & days in response to the flare effect of GnBH a)onist administration.
[,1]
The a!ility of this test
to differentiate !etween normal and diminished ovarian reserve appears to !e limited.
[(']
The availa!ility and application of so many tests of ovarian reserve serves to illustrate the lack of
a sin)le relia!le techni;ue. However% a recent addition to the list of promisin) candidates for
predictin) ovarian response is antimHllerian hormone 63H"% a mem!er of the transformin)
)rowth factor*E superfamily. 63H is produced in the ovary !y )ranulosa cells of )rowin) preantral
and small antral follicles. #n a recent study in youn) normo*ovulatory women% serum
concentrations of 63H decreased over time% whereas other markers associated with ovarian
a)in) did not chan)e
.
Given that 63H concentrations correlate well with antral follicle count and
a)e% and less stron)ly with inhi!in C and FSH levels% it mi)ht constitute a sensitive marker for
ovarian a)in).
[,+%,&]
#ndeed% it has !een shown that poor response in #$F is associated with
decreased 63H levels.
[,-]
Becently% prediction of poor response to ovarian stimulation !ased on
inhi!in C and !asal FSH concentrations was improved after inclusion of 63H in a multivariate
re)ression analysis.
[,']