Toward a better follow-up of ovarian

recovery in young women after

chemotherapy with a hypersensitive
antim
ullerian hormone assay
Christine Decanter, M.D.,a Maeliss Peigne, M.D.,a Audrey Mailliez, M.D.,b Franck Morschhauser, M.D.,c
Audrey Dassonneville,d Didier Dewailly, M.D.,a and Pascal Pigny, Ph.D.d
a
Service de Gynecologie Endocrinienne et Me decine de la Reproduction, Ho
^ pital Jeanne de Flandre, Centre Hospitalier
Re  partement de se
 gional Universitaire; b De  nologie, Centre Oscar Lambret; c Service des maladies du sang, Ho ^ pital
Huriez, Centre Hospitalier Re  gional Universitaire; and d Laboratoire de Biochimie & Hormonologie, Centre de Biologie
Pathologie, Centre Hospitalier Re  gional Universitaire, Lille, France

Objective: To evaluate the utility of a hypersensitive assay for measuring low antim ullerian hormone (AMH) levels in young cancer
patients during the ovarian recovery phase of their chemotherapy.
Design: Retrospective study.
Setting: Academic medical center.
Patient(s): Fifty-eight samples drawn at least 3 months after the end of chemotherapy in 30 women having either breast cancer
(n 13) or hematologic malignancies (n 17) were selected to constitute two equally size groups: amenorrhea (n 30 samples)
or spontaneous cycle (n 28 samples).
Intervention(s): None.
Main Outcome Measure(s): Serum AMH levels were measured by a conventional AMH ELISA (EIA AMH/MIS) and a hypersensitive
ELISA (PicoAMH, AnshLabs) on the same sample.
Result(s): Using a conventional assay, serum AMH was detectable (R3 pmol/L) in 6.7% and in 10.7% of the samples corresponding to
amenorrheic or cycling patients, respectively (nonsignicant). By contrast, with PicoAMH, serum AMH was detectable (R0.07 pmol/L)
in 71.4% of the samples from cycling women vs. 16.7% of the samples from amenorrheic patients. Multivariate regression analysis
showed that among putative contributors, only the menstrual status (r 0.307) and serum FSH level (r 0.546) were independently
correlated to a detectable serum AMH with the picoAMH assay exclusively.
Conclusion(s): The picoAMH assay, allowing measurement of very low AMH concentrations in
human serum, should rene postchemotherapy ovarian follow-up in young women. (Fertil Use your smartphone
Steril 2014;102:4837. 2014 by American Society for Reproductive Medicine.) to scan this QR code
Key Words: AMH, ovarian follicles, chemotherapy, follow-up, cancer and connect to the
discussion forum for
this article now.*
Discuss: You can discuss this article with its authors and with other ASRM members at http://
fertstertforum.com/decanterc-ovarian-recovery-chemotherapy-hypersensitive-amh-assay/ * Download a free QR code scanner by searching for QR
scanner in your smartphones app store or app marketplace.

A
ntim
ullerian hormone (AMH) hormone is produced by granulosa roles seems to be the inhibition of
is a biomarker that can be cells of preantral and small antral the recruitment of growing follicles
measured easily in blood sam- follicles until they reach a diameter from the primordial follicle pool (3,
ples and that strongly correlates with of approximately 8 mm, just before 4). Therefore, AMH is increasingly
the number of growing follicles within the follicle selection for dominance well-validated as a marker of the
the ovary (1, 2). Antim ullerian (3, 4). One of its main physiologic functional ovarian reserve and is
extensively used as a predictive
Received February 7, 2014; revised and accepted May 7, 2014; published online June 18, 2014. marker of IVF success (1, 2).
C.D. has nothing to disclose. M.P. has nothing to disclose. A.M. has nothing to disclose. F.M. has
nothing to disclose. A.D. has nothing to disclose. D.D. has nothing to disclose. P.P. has nothing
Young women treated by
to disclose. chemotherapy for cancer usually lose
Reprint requests: Pascal Pigny, Ph.D., Laboratoire de Biochimie & Hormonologie, Centre de Biologie their growing follicles, either tempo-
Pathologie, Centre Hospitalier Re gional Universitaire, Lille 59037, France (E-mail: pascal.
pigny@chru-lille.fr). rarily or permanently, depending on
several factors, mainly age and type
Fertility and Sterility Vol. 102, No. 2, August 2014 0015-0282/$36.00
Copyright 2014 American Society for Reproductive Medicine, Published by Elsevier Inc.
and dose of chemotherapy (5, 6) but
http://dx.doi.org/10.1016/j.fertnstert.2014.05.014 also individual susceptibility and

VOL. 102 NO. 2 / AUGUST 2014 483

ORIGINAL ARTICLE: FERTILITY PRESERVATION
pretreatment ovarian follicle content (710). The recovery of
normal menstrual cycles is a late marker of the reinitiation of
follicle growth after chemotherapy, and studies have shown
that follicular depletion may occur despite maintenance of a
regular menstrual cycle (11, 12). Conversely, in healthy
cycling women, it is now well established that AMH
becomes undetectable approximately 5 years before the
occurrence of natural menopause (13, 14). The question of
fertility capacity after cancer treatment is becoming more
and more important for young survivors. In this setting, it is
indeed especially crucial to have a reliable biomarker
allowing earlier prediction of ovarian function recovery. In
this respect, AMH measurements might offer the possibility
of a more accurate assessment of the follicular depletion
intensity (15).
We thus initiated in 2006 a systematic follow-up of AMH
levels before, during, and after chemotherapy in our young
lymphoma and breast cancer patients (16). To date we have
used a commercially available immunoassay (EIA AMH/
MIS; Immunotech) with a detection limit of 0.93 pmol/L
and a functional sensitivity (limit of quantitation) of approx-
imately 2.5 pmol/L. The results of this follow-up in this
specic population highlighted that the mean AMH serum
level was closely related to follicular depletion and recovery
and varied according to the type of chemotherapy (16). How-
ever, on an individual basis the reliability and sensitivity of
this marker was questioned by the frequent observation of
women regaining normal menstrual cycles despite having
undetectable serum AMH levels (7, 16, 17).
We therefore hypothesized that these women had some
amount of circulating AMH that we were unable to measure
with the available immunoassay. To test this hypothesis we
used the recently released hypersensitive picoAMH assay
(AnshLabs) to assess serum AMH in a cohort of young lym-
phoma or breast cancer patients, at least 3 months after the
end of chemotherapy, regardless of their menstrual status.
Each sample was tested with both the picoAMH and EIA
AMH/MIS assays.
MATERIALS AND METHODS
Patients
From our biobank of blood samples, we selected 58 serum
samples stored at 80 C collected from 3 to 24 months after
the end of chemotherapy in 30 women regardless of their
menstrual status. These 30 young patients were prospectively
recruited before the initiation of chemotherapy for lymphoma
(n 17) or for early breast cancer (n 13), to be followed-up
regarding their ovarian reserve by repeated AMH measure-
ments during and after treatment. All the patients were
treated with an alkylating regimen that is notoriously gona-
dotoxic. The mean age at sampling for the whole population
was 30.7 years, ranging from 16 to 41 years. The mean age for
lymphoma patients was 28.3 years, ranging from 16 to
38 years. The mean age of breast cancer patients was
33.5 years, ranging from 27 to 41 years. All the breast cancer
patients were treated by the FEC-T protocol (three cycles of
uorouracil 500 mg/m2, epirubicin 100 mg/m2, and cyclo-
phosphamide 500 mg/m2, followed by three cycles of doce-
484
taxel 100 mg/m2). Lymphoma patients were assigned to an
R-CHOP (adriamycin 50 mg/m2, cyclophosphamide 750 mg/
m2, vincristine 1.4 mg/m2, and prednisone 40 mg/m2),
hyper-CVAD (adriamycin 75 mg/m2, cyclophosphamide
1,200 mg/m2, bleomycin 10 mg/m2, and vindesin 2 mg/m2),
or BEACOPP (adriamycin 25 mg/m2, cyclophosphamide
650 mg/m2, etoposide, procarbazine 100 mg/m2, and vincris-
tine 1.4 mg/m2) protocol, depending on the extent and
severity of the disease. Serum samples for AMH assays were
obtained and frozen for further assays according to a stan-
dardized protocol that we previously reported (16). This
prospective, observational study was approved by the institu-
tional review board of the Lille University Hospital, and each
patient gave her informed consent. We selected our samples
to constitute two equally sized groups, according to the
menstrual status of the patients at the time of sampling: either
amenorrhea (n 30) or spontaneous cycle (n 28). In
patients who were at least at 12 months into follow-up
(n12) only one sample per patient was used because we
consider that the ovarian recovery phase might have occurred
within this delay. In patients who were only at 3 or
6 months of their follow-up (n 18), several samples per
patient were used 3, 6, 9, 12, 18, and/or 24 months after the
end of chemotherapy. Therefore, the distribution of the sam-
ples was as follows: 3 months: 9; 6 months: 13; 9 months: 11;
12 months: 18; 18 months: 2; and 24 months after the end of
chemotherapy: 5. Consequently, the total number of samples
per patients varied from 1 to 4. This parameter was included in
the multivariate analysis (see below), to detect any potential
patient effect that might have biased the results.
Assays
Antim ullerian hormone was measured by a conventional
assay, EIA AMH/MIS (A11893; Immunotech, Beckman
Coulter), and a hypersensitive assay, picoAMH (AL-124i;
AnshLabs) on the same serum sample by the same operator.
Both assays measure the promature AMH complex and are
standardized against recombinant human AMH. The
dynamic ranges of the standard curve are 3150 pmol/L for
the EIA AMH/MIS and 0.076.5 pmol/L for the picoAMH
assay. The between-run reproducibility was determined by
measuring either two quality control samples (mean concen-
trations, 0.69 and 2.0 pmol/L) for the picoAMH assay or three
quality control samples (mean concentrations, 9.8, 18.6, and
37.1 pmol/L) for the EIA AMH/MIS in duplicate in 3 or 156 in-
dependent assays, respectively. The coefcients of variation
(CVs) are 1.38% and 3.84% for the two levels assessed with
the picoAMH assay (preliminary data) and 14%, 13%, and
12.6% for the three levels assessed with the EIA AMH/MIS
assay, respectively. The limit of quantication (LoQ) that cor-
responds to the lowest concentration that can be measured
with an acceptable performance (i.e., CV %20%) was deter-
mined from the precision prole curve (CV(%) f ([AMH])
in pmol/L) built from three independent series. Limit of quan-
tication is 2.5 pmol/L for the EIA AMH/MIS and 0.034 pmol/
L for the picoAMH assay. For this study, results under the rst
calibrator value (i.e., <0.07 [picoAMH] or 3.0 pmol/L [EIA
AMH/MIS]) were expressed as undetectable.
VOL. 102 NO. 2 / AUGUST 2014

Statistical Analysis
Proportions of patients with detectable AMH by the different
assays were compared by the c2 test. Multivariate regression
analysis was performed to determine which parameters inde-
pendently contributed to a detectable AMH level. All statisti-
cal analyses were performed with SPSS 15.0 software. A P
value of < .05 was considered statistically signicant.
RESULTS
Fifty-eight samples corresponding to 30 patients were evalu-
ated. As shown in Table 1, the rate of measurable AMH levels
with the EIA AMH/MIS assay was not signicantly higher in
samples from patients with normal cycles than from amenor-
rheic ones. Conversely, the difference was much greater and
signicant with the picoAMH assay (Fig. 1). Accordingly, in
the 28 serum samples from patients who regained normal cy-
cles, AHM was much more frequently detectable with the
latter assay, the difference being highly signicant (Table 1).
By multivariate regression analysis, we tested the serum
AMH level as a binary variable (i.e., not detectable 0;
detectable 1) vs. a set of putative independent contributors:
menstrual status, age at sampling, time lapse between sam-
pling and end of chemotherapy, sample number by patient,
and serum FSH level. With the hypersensitive picoAMHassay,
this analysis showed that both the menstrual status (i.e.,
amenorrhea 0 and presence of normal cycles 1) and
serum FSH level were signicantly and independently associ-
ated to detectable AMH, positively and negatively, respec-
tively (r 0.307, P .023 and r 0.546, P .0001,
respectively). With the EIA AMH/MIS assay, none of these
parameters was signicantly associated to a detectable
AMH level.
By univariate analysis, no correlation was found between
the absolute values of AMH (when detectable) and FSH serum
levels with the picoAMH assay (n 26). This analysis was not
possible with the EIA AMH/MIS assay because of an insuf-
cient number of observations (n 6).
Of the ve patients who were amenorrheic despite detect-
able serum AMH with the picoAMH assay, three recovered
spontaneous cycles 3 months later. Conversely, all the
patients who menstruated despite undetectable AMH levels
with the picoAMH assay (n 8) remained menstruating
3 months later. Therefore the positive predictive value of a
detectable AMH for predicting recovery of menstrual cycles
increased from 60% (3 of 5) with the EIA AMH/MIS to a
TABLE 1
Frequency of detectable AMH with two different assays in
postchemotherapy women, according to the menstrual status.
Amenorrheic Normally cycling
patients patients
Parameter (n [ 30 samples) (n [ 28 samples) c2 test
Conventional assay 2/30 (6.7) 3/28 (10.7)
Ultrasensitive assay 5/30 (16.7) 20/28 (71.4)
c2 test n.s. P< .0001
Note: Values are number (percentage). n.s. nonsignicant.
Decanter. AMH levels detection after chemotherapy. Fertil Steril 2014.
VOL. 102 NO. 2 / AUGUST 2014
Fertility and Sterility
FIGURE 1
1.5
pico AMH (pmol/L)
1.0
0.5
0.0
g
lin
he
yc
rr
no
C
ne
m
A
Distribution of the serum picoAMH values obtained in samples from
amenorrheic patients (n 30) or from cycling patients (n 28). The
red horizontal axis corresponds to the lowest detectable
concentration (0.07 pmol/L) that we set.
Decanter. AMH levels detection after chemotherapy. Fertil Steril 2014.
maximal value of 92% (20 3 of 25) with the picoAMH assay.
Among the 18 patients who were sampled at different times of
their follow-up, we observed in only three cases a very slight
increase of AMH levels with the ultrasensitive assay (but not
with the conventional one), followed 3 or 6 months later by a
reversion to undetectable values although these patients were
still regularly menstruating. For 11 of the 15 other patients
(61% of the whole), we observed a transition of serum AMH
from undetectable to detectable values that preceded or
coincided with the return of menstrual activity.
DISCUSSION
Since 2010 two ELISA kits have been available for the mea-
surement of AMH in blood samples. The Gen II assay devel-
oped by Beckman Coulter in 2010 on the basis of the former
DSL kit recognizes both human, rat, and mouse AMH and
has a functional sensitivity ranging from 1.14 to 1.50 pmol/
L according to the available data (18) (manufacturer data).
The EIA AMH/MIS assay that we have used since 2003 was
developed earlier by Immunotech (19). It detects only human
AMH and has an LoQ of approximately 2.5 pmol/L. Therefore,
both assays are not able to detect low or very low amounts of
AMH in blood. Interestingly, we had the opportunity to focus
on these low concentrations with the new picoAMH assay
developed by AnshLabs, which is a hypersensitive assay
whose dynamic range is 0.076.5 pmol/L and LoQ is 0.034
pmol/L. Here we report that this gain in sensitivity has clinical
relevance in a population of young female cancer patients
with low serum AMH concentrations.
n.s. To date the early systematic follow-up of young cancer
P< .0001 patients undergoing chemotherapy revealed that in a large
proportion of patients, AMH levels remained undetectable af-
ter the end of treatment, irrespective of their menstrual status
(7, 16, 17). The results of the present study conrm this
485
ORIGINAL ARTICLE: FERTILITY PRESERVATION
nding, suggesting that the EIA AMH/MIS assay is not
sensitive enough to detect very low AMH levels during the
postchemotherapy phase. As shown by the present data, this
is particularly relevant in the group of patients having
recovered spontaneous menstrual cycles. In this group the
serum AMH level was undetectable in 90% of the samples
with the conventional EIA AMH/MIS assay, whereas this
rate was much lower with the picoAMH assay (28.6%). To
our knowledge this is the rst report showing such a
signicant difference in young postchemotherapy patients.
Concerning our amenorrheic patients, the rate of detectable
AMH with the picoAMH assay (17%) was very close to the
one recently reported with the same assay in older
reproductive-age breast cancer patients (20). Likewise,
another recent study using a different ultrasensitive Anshlab
AMH assay with a higher LoQ (0.2 pmol/L) conrmed the ex-
istence of detectable AMH concentrations in 57 early breast
cancer patients after chemotherapy (21). The present study
thus conrms that a hypersensitive assay can allow a better
follow-up of AMH during the ovarian recovery phase.
Although this study was not designed to test the clinical rele-
vance of such AMH variations after the end of chemotherapy,
it has to be noted that in 61% of patients who were sampled
several times, the transition of serum AMH from undetectable
to detectable values preceded or coincided with the return of
the menstrual activity.
Having a hypersensitive assay is very useful in this specic
cancer population with extremely low serum AMH levels. It
should improve our knowledge about the degree of follicle
loss and the dynamics of follicular recovery according to the
different types of gonadotoxic protocols (1517). However,
the clinical relevance of having detectable or undetectable
AMH levels after the cancer treatment in terms of fertility
and reproductive lifespan remains to be investigated
prospectively. Indeed, a recent study shows a good successful
pregnancy rate in childhood lymphoma survivors despite low
AMH concentrations (22), whereas another one highlights
high incidence of infertility and early menopause in patients
previously treated by chemotherapy (23). Improvement of
AMH assay sensitivity may also benet other situations such
as premature ovarian failure, poor ovarian response in IVF
programs, and serum AMH follow-up after chemotherapy dur-
ing childhood. Because AMH is detectable in girls of all ages,
being able to detect very low AMH levels may help to identify
young girls who will potentially require puberty induction (24).
In IVF programs, a sensitive AMH assay may discriminate
between patients with a negligible or reduced chance to
get proper criteria for triggering (25). Last, it could also be use-
ful in the evaluation of the ovarian reserve after auto-
transplantation of frozenthawed ovarian cortical tissue: it
has been recently shown that in several cases AMH remains
undetectable with a conventional assay despite the resolution
of normal menstrual function (26).
In conclusion, these preliminary results conrm that the
improvement of AMH assay sensitivity is timely and would
allow a better follow-up of the ovarian follicular content in
young cancer patient. Longitudinal studies are urgently
needed to investigate the clinical relevance of such dynamics
of AMH levels.
486
Acknowledgments: The authors thank AnshLabs for
providing the AL-124 picoAMH kits, Ms. S. Ogez and E. d'Or-
azio for patients care, and our colleagues who referred their
patients for ovarian follow-up.
REFERENCES
1. La Marca A, Sighinol G, Radi D, Argento C, Baraldi E, Artenisio AC, et al.
AMH as a predictive marker in assisted reproductive technology. Hum Re-
prod Update 2010;16:11330.
2. Dewailly D, Andersen CY, Balen A, Broekmans F, Dilaver N, Fanchin R, et al.
The physiology and clinical utility of AMH in women. Hum Reprod Update
2014;20:37085.
3. Weenen C, Laven JS, Von Bergh AR, Craneld M, Groome NP,
Visser JA, et al. AMH expression pattern in the human ovary: potential
implications for initial and cyclic follicle recruitment. Mol Hum Reprod
2004;10:7783.
4. Jeppensen JV, Anderson RA, Kelsey TW, Christiansen SL, Kristelsen SG,
Jayaprakasan K, et al. Which follicles make the most AMH in humans? Evi-
dence for an abrupt decline in AMH production at the time of follicle selec-
tion. Mol Hum Reprod 2013;19:51927.
5. Roness H, Gavish Z, Cohen Y, Meirow D. Ovarian follicle burnout a universal
phenomenon? Cell Cycle 2013;12:32456.
6. Morgan S, Anderson RA, Gourley C, Wallace WH, Spears N. How do
chemotherapeutic agents damage the ovary? Hum Reprod Update
2012;18:52535.
7. Anderson RA, Cameron DA. Pretreatment serum AMH predicts long term
ovarian function and bone mass after chemotherapy for early breast cancer.
J Clin Endocrinol Metab 2011;96:133643.
8. Anderson RA, Rosendahl M, Kelsey TW, Cameron DA. Pretreatment AMH
predicts for loss of ovarian function after chemotherapy for early breast can-
cer. Eur J Cancer 2013;49:340411.
9. Dillon KE, Sammel MD, Prewitt M, Ginsberg JP, Walker D, Mersereau JE,
et al. Pretreatment AMH levels determine rate of posttherapy ovarian
reserve recovery: acute changes in ovarian reserve during and after chemo-
therapy. Fertil Steril 2013;99:47783.
10. Rosendahl M, Andersen MT, Ralkiaer E, Kjeldsen L, Andersen MK,
Andersen CY. Evidence of residual disease in cryopreserved ovarian cortex
from female patients with leukemia. Fertil Steril 2010;94:218690.
11. Bath LE, Wallace WH, Shaw MP, Fitzpatrick C, Anderson RA. Depletion of
ovarian reserve in young women after treatment for cancer in childhood:
detection by AMH, inhibin B and ovarian ultrasound. Hum Reprod 2003;
18:236874.
12. Partridge AH, Ruddy KJ, Gelber S, Shapira L, Abusief M, Meyer M, et al.
Ovarian reserve in women who remain premenopausal after chemotherapy
for early stage breast cancer. Fertil Steril 2010;94:63844.
13. Sowers MR, Eyvazzadeh AD, McConnell D, Yosef M, Jannausch ML,
Zhang D, et al. AMH and inhibin B in the denition of ovarian aging and
the menopause transition. J Clin Endocrinol Metab 2008;93:347883.
14. Broer SL, Eijkemans MJ, Scheffer GJ, Van Rooij IA, de Vet A, Themmen AP,
et al. AMH predicts menopause: a long term follow up study in normoovu-
latory women. J Clin Endocrinol Metab 2011;96:25329.
15. Peigne M, Decanter C. Serum AMH level as a marker of acute and long term
effects of chemotherapy on the ovarian follicular content: a systematic re-
view. Reprod Biol Endocrinol 2014;12:26.
16. Decanter C, Morschhauser F, Pigny P, Lefebvre C, Gallo C, Dewailly D. AMH
follow-up in young women treated by chemotherapy for lymphoma: prelim-
inary results. Reprod Biomed Online 2010;20:2805.
17. Anderson RA, Themmen AP, Al Qahtani A, Groome NP, Cameron DA. The
effects of chemotherapy and long term gonadotrophin suppression on the
ovarian reserve in premenopausal women with breast cancer. Hum Reprod
2006;21:258392.
18. Kumar A, Kaira B, Patel A, Mc David L, Roudebush WE. Development of a
second generation AMH ELISA. J Immunol Methods 2010;362:519.
19. Long WQ, Ranchin V, Pautier P, Belville C, Denizot P, Cailla H, et al. Detec-
tion of minimal levels of serum AMH during follow up of patients with
VOL. 102 NO. 2 / AUGUST 2014

ovarian granulose cell tumor by means of a highly sensitive ELISA. J Clin En-
docrinol Metab 2000;85:5404.
20. Chung K, Stanczyk F, Ma L, Sammel M, Haunschild C, Su I. Can ultrasensitive
AMH assays detect ovarian function in older reproductive-aged women with
breast cancer? (abstract P55). Fertil Steril 2013;100:S164.
21. Fleming R, Fairbairn C, Lucas D, Gaudoin M, Anderson RA. Analysis of two
assays for the measurement of AMH in women with low ovarian reserve (ab-
stract P48). Fertil Steril 2013;100:S162.
22. Hamre H, Kiserud CE, Rudd E, Thorsby PM, Fossa SD. Gonadal function and
parenthood 20 years after treatment for childhood lymphoma: a cross-
sectional study. Pediatr Blood Cancer 2012;59:2717.
VOL. 102 NO. 2 / AUGUST 2014
Fertility and Sterility
23. Letourneau JM, Ebbel EE, Katz PP, Katz A, Ai WZ, Chien AJ, et al. Pretreat-
ment fertility counseling and fertility preservation improve quality of life in
reproductive age women with cancer. Cancer 2012;118:17107.
24. Brougham MF, Crofton PM, Johnson EJ, Evans N, Anderson RA, Wallace WH.
AMH is a marker of gonadotoxicity in pre- and postpubertal girls treated for
cancer: a prospective study. J Clin Endocrinol Metab 2012;97:205967.
25. Nelson SM. AMH: is the writing on the wall for antral follicle count? Fertil
Steril 2013;99:15634.
26. Greve T, Schmidt KT, Kristensen SG, Ernst E, Andersen CY. Evaluation of the
ovarian reserve in women transplantated with frozen and thawed ovarian
cortical tissue. Fertil Steril 2012;97:13948.
487