Cancer Treatment Reviews 37 (2011) 97104

Contents lists available at ScienceDirect

Cancer Treatment Reviews

journal homepage: www.elsevierhealth.com/journals/ctrv

Controversy

Which factors should be taken into account in perimenopausal women with

early breast cancer who may become eligible for an aromatase inhibitor?
Recommendations of an expert panel
O. Ortmann a,*, O. Pagani b, A. Jones c, N. Maass d, D. Noss a, H. Rugo e, C. van de Velde f, Matti Aapro g,
R. Coleman h
a

Department of Gynaecology and Obstetrics, University of Regensburg, Germany

Institute of Oncology of Southern Switzerland, Bellinzona, Switzerland
c
Department of Clinical Oncology, Royal Free Hospital, London, UK
d
University Clinic for Gynaecology and Obstetrics, Aachen, Germany
e
University of California San Francisco Comprehensive Cancer Center, San Francisco, CA, USA
f
Department of Surgery, Leiden University Medical Centre, Leiden, The Netherlands
g
Clinique de Genolier, Genolier, Switzerland
h
Academic Unit of Clinical Oncology, Weston Park Hospital, Shefeld, UK
b

a r t i c l e

i n f o

Article history:
Received 24 March 2010
Received in revised form 21 May 2010
Accepted 25 May 2010

Keywords:
Adjuvant therapy
Aromatase inhibitors
Breast cancer
Menopause
Perimenopause
Tamoxifen

s u m m a r y
Menopausal status is a major consideration in adjuvant breast cancer therapy. The variable onset and
duration of the menopausal transition and the poor predictive value of bleeding patterns and hormone
levels mean many women fall naturally into a perimenopausal category. Women becoming amenorrhoeic during cytotoxic or endocrine treatment are also of uncertain status since ovarian function
may resume even in older patients after several months without menses. The recent St. Gallen panel
acknowledged that aromatase inhibitors (AIs) should form part of standard endocrine therapy for postmenopausal women with receptor-positive tumours. Among perimenopausal women at sufciently high
risk of recurrence, there may also be a case for adjuvant AIs either up-front or after tamoxifen. Such treatment should be initiated only after careful consideration of the patients age, menstrual history and the
effects of tamoxifen (which may make hormone levels an unreliable guide to ovarian function). In treatment-nave women whose postmenopausal status cannot be conrmed by reliable, serial hormone measurements, treatment should start with tamoxifen. Serial monitoring of hormone levels may enable an AI
to be started if postmenopausal status is conrmed. In women with treatment-induced amenorrhoea,
any decision to start an AI requires baseline hormone levels consistent with postmenopausal status;
and continuation of treatment requires that hormone levels remain postmenopausal during regular
monitoring.
2010 Elsevier Ltd. All rights reserved.

Introduction
While the median age at diagnosis of breast cancer is 61 years,
13% of cases in the US occur in women younger than 45 years and
22% between the ages of 45 and 54.1 The average age of menopause
in women of Caucasian/European origin is around 51 years.2,3
Breast cancer therefore affects a substantial number of women
who are either premenopausal or within the typical perimeno-

* Corresponding author. Address: Department of Gynaecology and Obstetrics,

University of Regensburg, Landshuter Strasse 65, D-93053 Regensburg, Germany.
Tel.: +49 941 782 7511; fax: +49 941 782 7515.
E-mail address: olaf.ortmann@klinik.uni-regensburg.de (O. Ortmann).
0305-7372/$ - see front matter 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ctrv.2010.05.005

pausal age range. In the Cancer and Menopause Study cohort of

577 women aged under 52 at the time of diagnosis of stage 0II
breast cancer, 68% were considered premenopausal, 15% perimenopausal and 9% postmenopausal.4 (The remaining 8% could
not be categorized on the available information.)
In this paper, we discuss the management of the many women
with early breast cancer who are neither clearly premenopausal
nor clearly postmenopausal. This group includes women who
are undergoing the age-related transition from regular menstruation to permanent absence of menstruation (i.e. the naturally perimenopausal) and those in whom cancer treatment has induced a
cessation of menses that may prove to be only temporary. The
importance of the area we address was recently recognised by
the St. Gallen expert panel.5 They state that in patients under

98

O. Ortmann et al. / Cancer Treatment Reviews 37 (2011) 97104

the age of 60 who are receiving AIs, it should be veried that

ovarian function suppression has achieved postmenopausal levels.
There has been considerable uncertainty about how to dene
the onset of the natural perimenopause, and its termination in true
menopause. This is understandable given that we are dealing with
a continuum, and rigid denitions are not appropriate. How to
evaluate the menopausal status of women with amenorrhoea induced by cytotoxic or endocrine agents is an equally difcult problem. It is thought that around 80% of women aged over 40 years
who are treated with adjuvant chemotherapy become at least temporarily amenorrhoeic.6 Even if menses resume, these women
experience an earlier menopause than they would have done if
not treated with chemotherapy. The management of naturally
menopausal women and those with treatment-induced amenorrhoea (TIA) is likely to be broadly similar. The difference in patients
with TIA is the relatively long period (especially in younger women) over which menses may resume and so require change in
management.
We recognise that clinicians seeking to optimise therapy in a
patient of uncertain menopausal status may consider an aromatase
inhibitor (AI) rather than tamoxifen because they feel that the benet might be greater.
Studies in which oestrogen receptor-positive patients have
been maintained on adjuvant tamoxifen or switched to anastrozole
or exemestane have shown that switching to an AI signicantly improves rates of disease-free survival (DFS).7,8 Although the switch
from tamoxifen to letrozole has not been demonstrated by intent-to-treat analysis to achieve signicantly superior DFS when
compared with 5 years tamoxifen, it achieves similar DFS to
5 years of letrozole.9 The Intergroup Exemestane Study (IES) is
the only double-blind trial to have also shown a statistically significant benet on overall survival.8 This paper does not attempt to
reconcile the differing opinions on exactly which groups of women
should be treated with AIs. It also avoids making recommendations
about the sequencing and duration of endocrine therapy. These issues have been addressed recently by the St. Gallen consensus panel.5 Our purpose is simply to advise on the steps to take before
instituting an AI in women who may be candidates for such
treatment.

Methods
On the basis of clinical relevance, a number of key unresolved
issues were identied relating to the hormonal treatment of the
perimenopausal breast cancer patient. Having identied these issues, a literature search was conducted independently of the
authors using appropriate keywords and phrases. These included
(in various combinations) aromatase inhibitors; breast cancer;
chemotherapy-induced amenorrhoea; endocrine therapy; oestradiol; oestrogen level, measurement and receptors; follicle stimulating hormone; luteinising hormone and luteinising hormone
agonists; menopausal transition; ovarian function suppression;
and peri-, post- and premenopausal. Articles were selected for consideration if they had been published within the last 10 years and
addressed one or more of these issues: differentiating between
peri- and postmenopausal breast cancer patients, hormonal measurement or other means of determining menopausal status, and
treatment recommendations for breast cancer patients of different
menopausal status. Following discussion of this evidence, the writing committee (O. Ortmann, O. Pagani, A. Jones, N. Maas, and D.
Noss) sought to achieve consensus on a series of recommendations.
These recommendations were then put to a review committee (H.
Rugo, C. van de Velde, M. Aapro, and R. Coleman) whose contributions were incorporated into the nal manuscript.

Hormonal therapy for early stage breast cancer

In the case of early stage breast cancer, surgery is likely to be
followed by some form of adjuvant systemic therapy with either
cytotoxic or endocrine agents or both. In the premenopausal patient with endocrine-sensitive disease, standard adjuvant treatment is tamoxifen plus or minus ovarian suppression.
Chemotherapy may be added if indicated. In the Cancer and Menopause Study already cited, 35% of women (who were diagnosed in
the early 1990s) received adjuvant chemotherapy, 10% adjuvant
tamoxifen, and 27% both chemotherapy and tamoxifen.4 The proportion of patients electing to receive chemotherapy may now be
considerably higher. In the Tamoxifen and EXemestane Trial
(TEXT) in which adjuvant chemotherapy was a matter of choice,
it was used in 64% of premenopausal women with endocrineresponsive early breast cancer.10 Eighty-eight percent of women
with positive nodes and 46% of those who were node-negative received chemotherapy.
Adjuvant tamoxifen reduces risk of breast cancer recurrence
within 15 years by an absolute 12% in early hormone-receptor-positive disease.11 Risk of death from breast cancer over the same
period is reduced by an absolute 9%. In postmenopausal women,
the AIs anastrozole, exemestane and letrozole, which block oestrogen production rather than activity at the receptor, achieve a further clinical benet. Postmenopausal women treated with AIs are
signicantly less likely to experience disease recurrence than women treated with tamoxifen.7,8,12 A meta-analysis presented at
the 2008 San Antonio Breast Cancer Symposium calculated that
in women with ER-positive tumours 5 years treatment with adjuvant AIs produced a statistically signicant absolute 2.7% reduction
in breast cancer recurrence when compared with 5 years tamoxifen.13 A similar meta-analysis of studies which compared a switching strategy (i.e. tamoxifen followed by an AI) against 5 years
tamoxifen concluded that switching to an AI achieved a 3.5% absolute reduction in disease recurrence and an absolute 1.6% reduction in mortality.13 In the case of both disease-free and overall
survival, differences were signicant compared with tamoxifen
alone for the full 5 years. The opposite sequence (among others)
was assessed in the BIG 1-98 trial.9 Overall, there was no clear
superiority of the sequential strategy over letrozole given for
5 years up-front.
The benet of AIs in postmenopausal patients with endocrineresponsive tumours was recognised some years ago in recommendations produced by the American Society of Clinical Oncology.14
These state that optimal adjuvant hormonal therapy for a postmenopausal woman with receptor-positive breast cancer includes
an AI as initial therapy or after treatment with tamoxifen. The
majority of the latest St. Gallen consensus panel preferred AIs as
up-front endocrine treatment particularly in patients at high risk
of relapse.5 However, the denition of postmenopausal is not always clear; and clinical trials have not been consistent in the criteria used when making this important judgment.
It is still unknown whether the use of AIs in combination with
ovarian function suppression (OFS) in premenopausal women is
superior to tamoxifen. A recent trial that investigated the effects
of zoledronic acid in addition to comparing OFS with anastrozole
against OFS plus tamoxifen as adjuvant endocrine treatment for
premenopausal women with endocrine-sensitive breast cancer
failed to demonstrate superiority for the AI combination.15 However, the trial was possibly underpowered to detect any likely difference and denitive conclusions await the ndings of the TEXT
(Tamoxifen and EXemestane) and Suppression of Ovarian Function
(SOFT) Trials. In the TEXT study, premenopausal women with steroid hormone-positive early breast cancer are being randomised to
OFS with triptorelin plus either exemestane or tamoxifen, each for
5 years. The three arms of the SOFT trial will allow comparison of

O. Ortmann et al. / Cancer Treatment Reviews 37 (2011) 97104

disease recurrence rates in premenopausal women treated with

tamoxifen alone, OFS plus tamoxifen, or OFS plus exemestane.
This paper asks whether the benet seen with AIs in the postmenopausal setting might be extended with appropriate safeguards to two further important groups of patients: women
who are undergoing a natural perimenopausal transition, and
those with treatment-induced amenorrhoea.

Dening menopausal status

Pituitary gonadotrophins stimulate ovarian steroid hormone
production (oestrogens, progesterone and androgens). Oestradiol
(E2) and progesterone act via a negative feedback loop to inhibit
pituitary gonadotrophin release. Hormone release occurs on a cyclical basis so that concentrations of follicle stimulating hormone (FSH)
peak in the mid-follicular phase and decrease during the luteal
phase, rising again shortly before menstruation. As E2 levels rise in
the late follicular phase, concentrations of luteinising hormone
(LH) surge (positive feedback) in mid cycle and then fall again during
the luteal phase (negative feedback) under the inuence of progesterone and E2. E2 and inhibin-B are under gonadotrophin inuence.
Both hormones are stimulated by FSH. This allows the preservation
of oestradiol secretion in the late reproductive phase. In contrast,
anti-Mullerian hormone (AMH), a product of small preantral follicles, is not part of the feedback system described and remains relatively constant during the normal menstrual cycle.
The decline in quality and quantity of ovarian follicles, i.e. ovarian aging, triggers the hormonal and symptomatic changes of menopausal transition. This decline accelerates around age 35 when
the number of oocytes drops to approximately 25,000 (having been
approximately 300,000 at puberty).16 AMH is one of the earliest
markers of ovarian aging since it reects the remaining pool of follicles. Compared with changes in FSH, LH, progesterone, AMH and
inhibin-B secretion, the fall in E2 levels occurs relatively close to
the menopause. Before the menopause, climacteric symptoms
and bleeding irregularities can occur as a consequence of changes
in E2 and progesterone secretion.
The menopause is commonly dened as the last menstrual
bleed or nal menstrual period (FMP). This can only be recognised
retrospectively after 12 consecutive months of amenorrhoea.17 The
FMP is typically preceded by a period of hormonal instability and
irregularity in the menstrual cycle lasting up to several years.
There is no uniform denition of when this transitional period or
perimenopause begins.18 However, various criteria (such as the
rst occurrence of more than 7 days difference in cycle length)
have been proposed. The perimenopause extends until the
12 month period of amenorrhoea has elapsed.
Prior to the FMP, bleeding patterns are highly unpredictable;
vasomotor symptoms, while common, are not invariably experienced; and there is no hormonal marker which infallibly signals
permanent cessation of menstruation.17 In 2001, the Stages of
Reproductive Aging Workshop proposed FSH as the best marker
available but did not establish levels that dened the menopause.
FSH secretion is itself variable: elevations may occur up to 10 years
before the menopausal transition.18 Furthermore, there is potential
lack of agreement between assays, and body size and age have effects independent of menstrual status.
However, Randolph et al. report that longitudinal cohort studies
carried out more recently have established that a single annual serum FSH concentration of 2039 IU/l is associated with a 24-fold
increase in likelihood that a woman will experience her FMP within 12 months of measurement.17 Values of 40 IU/l and above were
associated with a 612-fold increase in likelihood of a FMP. Levels
of FSH of 40 IU/l and above are frequently taken as good evidence
of postmenopausal status.

99

Randolph et al.17 comment that increases in FSH are seen before

decreases in serum E2 as the menopause approaches. Nevertheless,
the latter at least in population terms show a consistent and
potentially useful pattern of change across the menopause.18 While
E2 levels are essentially stable until 2 years before the FMP, they
decline 67% over the following 4 years (mean level is 64.5 pg/ml
2 years before the FMP and 21 pg/ml 2 years after the LMP).19
However, there is considerable variability of E2 and FSH both within and between individual women.
Based on changes in menstrual cycles and levels of FSH, the rst
standardised classication of STages of Reproductive Aging Workshop (STRAW) was proposed in 2001.20 This classication includes
seven stages. These run from 5 (i.e. ve stages before the FMP) to
stage +2 after FMP. Stages 5 to 3 cover the transition from early
to late reproductive life: even though menses remain regular, there
is some increase in FSH. Stage 2, the early menopausal transition
stage, begins when the difference in length between two consecutive cycles exceeds 7 days. The late menopausal transition (stage
1) occurs when the intermenstrual interval is longer than 60 days
and ends with the FMP.
These stages relate not only to changes in bleeding patterns
but also to changes in hormone levels. A prospective cohort
study showed statistically signicant differences in hormone levels between the premenopausal and early transition stage, i.e.
from stages 5 to 2. FSH levels rose from 8.8 to 9.8 mlU/mL,
and inhibin levels fell from 45.0 to 33.7 pg/mL. However, LH levels remained relatively unchanged (4.1 vs. 4.4 mlU/mL), as did
E2 levels (39.1 vs. 39.5 pg/mL). In the late transition stage and
the postmenopause, a further rise in FSH levels (from 15.3 to
17.9 mlU/mL) can be detected.19 In contrast, E2 showed elevated
average levels in late transition (42.8 pg/mL) before a rapid decline shortly before postmenopause occurred (25 pg/mL).21 Why
E2 levels remain high while those of other ovarian markers have
fallen is unclear. It is hypothesised that high FSH can trigger E2
secretion even in aging ovaries and, furthermore, that aromatase
function is increased in the late menopausal transition.22,23 Inhibin-B levels fall to low levels in the postmenopause (23.4 pg/
mL). LH levels in contrast rise slowly during the menopausal
transition, reaching moderately elevated levels in the postmenopause (8.1 mlU/mL). Serum AMH declines to very low or nondetectable levels of about 0.05 ng/ml 5 years before the FMP.24 In
the period of the perimenopause, FSH levels increase. However,
they cannot be used as a reliable predictor for the time of
FMP.18 Markers such as inhibin-A and -B and AMH could be of
potential use in better dening the stages of menopausal transition. However, their widespread use has been precluded by cost
and by the lack of sensitivity and reproducibility of available
assays.
Specic polymorphisms in genes controlling oestrogen biosynthesis and metabolism may be involved in determining age of onset of the menopause, the duration of stages in the menopausal
transition, and timing of the LMP.
According to the most recent NCCN guidelines on the management of breast cancer,25 a woman can reasonably be considered
postmenopausal if any of these conditions has been fullled: (i)
she has had prior bilateral oophorectomy; (ii) she is aged 60 years
or older; (iii) if less than 60, she has had amenorrhoea of a year or
longer in the absence of chemotherapy, tamoxifen, toremifene, or
ovarian suppression, and FSH and E2 are in the postmenopausal
range; (iv) if she is taking tamoxifen or toremifene, and aged under 60, FSH and E2 levels are in the postmenopausal range. The
guidelines note that it is not possible to determine menopausal
status when a woman is taking a GnRH agonist or antagonist. If
none of these conditions is fullled, and yet the patient has infrequent or absent menses, she should be considered pre- or
perimenopausal.

100

O. Ortmann et al. / Cancer Treatment Reviews 37 (2011) 97104

Treatment-induced amenorrhoea (TIA)

Chemotherapy-induced amenorrhoea (CIA)
In addition to direct cytotoxicity, it is known that adjuvant chemotherapy has an indirect benecial effect in hormone-receptorpositive breast cancer through the induction of partial or complete
ovarian suppression; and amenorrhoea has been associated with
improved treatment outcome in several trials.26 In a recent presentation of data from NSABP trial B-30, CIA for 6 months or longer
after completion of chemotherapy improved DFS and overall survival in women treated with both anthracycline and taxane-containing regimens.27
The denition of CIA is not consistent across the literature, and
this helps explain the wide range in reported rates among chemotherapy trials. The American College of Obstetricians and Gynecologists denes amenorrhoea as cessation of menses for 6 months.28
However, other authors have dened CIA as the cessation of menses lasting 36 months or longer, or used the criterion of menstrual
cessation lasting at least 12 months.
In order to consider treatment with an AI in women experiencing
CIA, the identication of permanent amenorrhoea is essential. Some
women may in fact regain regular menses after cessation of chemotherapy. From a pragmatic perspective, 12 months of amenorrhoea
are strongly indicative of menopause: the majority of patients who
are amenorrhoeic at 1 years follow-up experience permanent menopause (although there are exceptions, as noted below).
One review of adjuvant alkylating agent-based chemotherapy
trials found an overall mean rate of amenorrhoea of 68% among premenopausal women, but the rates reported ranged between trials
from 20% to 100%.26 It is likely that there are genuine differences
due to the constituent agents in a combination chemotherapy regimen, with alkylating agents associated with the highest rates of
CIA, as well as a clear inuence of age and individual ovarian reserve.
Not surprisingly, the likelihood of amenorrhoea is substantially
greater in older patients. Walshe et al.26 report CIA in 046% of patients under 40 years of age treated with an anthracycline plus alkylating agent, and in 65100% of women over the age of 40 receiving
the same regimen. For CMF, the incidence of amenorrhoea is 18
61% in younger women, and 6197% in elderly patients; the corresponding gures for the FEC regimen are 038% and 7388%, respectively. Shapiro and Recht29 conrmed the systematic relationship of
age with risk of CIA: women older than 40 years had a far greater risk
of developing amenorrhoea compared with younger women, irrespective of whether they had received alkylating- or anthracycline-containing chemotherapy.
Recently, several groups have reported on the risk of amenorrhoea when an anthracycline-based adjuvant treatment is followed
by a taxane. Based on a series of 159 premenopausal patients, Davis et al. concluded that patients receiving sequential anthracycline-based chemotherapy and taxane were no more likely to
experience amenorrhoea than women not receiving the taxane:
overall, approximately half of treated patients retained menstrual
function.30 Data from 166 premenopausal patients aged under 40
treated sequentially with adjuvant anthracycline and taxane chemotherapy at Memorial Sloan-Kettering Cancer Center showed
that long-term amenorrhoea (i.e. lasting at least 12 months) was
experienced by only 15%.31 This incidence of CIA was no higher
than in historical controls who did not receive taxanes.
An interaction between age and chemotherapy regimen is also
found in the study of Tham et al.32 who compared the effect of giving a taxane after doxorubicin plus cyclophosphamide (AC) against
AC therapy alone in women aged 4150 years or younger. AC followed by a taxane induced amenorrhoea in a higher proportion
of women than AC alone (64% vs. 55%; odds ratio 1.9, p = 0.05).
Overall, patients older than 40 years were associated with a 4.6-

fold increased risk of CIA, which was more likely to be irreversible.

However, in women aged 40 or younger, the addition of the taxane
had no impact on the likelihood that menses would resume.
The timing of treatment within the menstrual cycle may modulate the gonadotoxicity of chemotherapy and hence the risk of
CIA.33 In a series of 111 consecutive patients who received either
alkylating, anthracycline or taxane-containing chemotherapy, a
higher incidence of CIA was reported in patients treated in the follicular phase rather than in other phases of the menstrual cycle
(67.6% compared with 45.5%; p = 0.03).
Tham et al.32 emphasized the fact that women aged under 40 frequently resumed menses, even after 6 months and more of amenorrhoea, and urged that this possibility be taken into account when
considering appropriate subsequent hormonal therapy. This point
was reinforced by Smith et al.34 who provided important data on a
series of 45 women with high risk breast cancer (median age
47 years; range 3952 years) who received adjuvant AIs after chemotherapy-induced amenorrhoea of at least 6 months duration with
accompanying biochemical evidence of ovarian suppression. Twelve
women (27% of the series, median age 44 years, range 4050 years)
showed a resumption of ovarian function, which in one instance occurred after 59 months of amenorrhoea. In 10 cases menses restarted,
one woman became pregnant and one became biochemically premenopausal. The authors concluded that AIs may promote recovery
of ovarian function and recommended that AIs should be used in
the context of CIA only with caution and in association with serial biochemical monitoring. Monitoring should include measurement of
FSH and E2. Such measurement should be subject to careful quality
assurance (QA) and may best be undertaken in a reference laboratory
with sufcient throughput and QA to maintain a high standard.
The view of Tham et al. and Smith et al. is supported by reports
which conrm that resumption of ovarian function can also occur
after long-term anti-oestrogen therapy and is likely to be underestimated in routine clinical practice.35 AIs are known to induce ovulation and are increasingly used in the harvesting of eggs before
chemotherapy or as treatment of infertile anovulatory women.36
AIs induce ovulation by inhibiting oestrogen production: the consequent hypo-oestrogenic state leads to increased pituitary gonadotrophin synthesis which stimulates follicle formation.
Amenorrhoea during tamoxifen treatment
The impact of tamoxifen on ovarian function is less well understood, especially since the number of patients who have not also
received adjuvant chemotherapy is very small. The relative contribution of tamoxifen to CIA is debated: some studies demonstrate
an increased incidence with the addition of tamoxifen, but others
report no impact.26
Tamoxifen may increase plasma E2 concentration by interfering
with normal negative pituitary feedback mechanisms, with the
resultant increase in FSH driving ovarian steroidogenesis. However, there may be additional mechanisms, including a direct interaction of tamoxifen with granulosa cells enhancing the FSH-driven
production of E2 and modifying LH receptor expression.3739
Illustrative cases
Case 1
A 47-year old woman was diagnosed with a 2 cm grade 2 invasive ductal breast carcinoma that was pT1c pN0(sn) M0. The tumour was highly oestrogen and progesterone receptor-positive
but HER2-negative. Sentinel node biopsy was negative. Following
surgery, options for adjuvant therapy were discussed. The woman
concerned declined chemotherapy because its likely benet was
relatively small and opted for endocrine treatment alone.

O. Ortmann et al. / Cancer Treatment Reviews 37 (2011) 97104

Prior to the diagnosis of breast cancer, she had noticed that her
menstrual cycle had changed: the interval between menses was
21 days rather than the previous 28. After starting tamoxifen, her
menses became less regular and stopped 12 months after beginning adjuvant treatment.
At 2-year follow-up, a hormone prole indicated that she was
postmenopausal. The options for endocrine treatment were discussed with her again and she elected to switch to an AI. Her hormone prole 3, 6 and 12 months after starting the AI remained
postmenopausal and she completed 5 years of endocrine therapy
uneventfully.
Case 2
A woman aged 42 years underwent surgery and then adjuvant
chemotherapy for a ductal breast carcinoma that was pT2 pN1a
MO. The tumour was moderately positive for oestrogen and progesterone receptors and HER2-negative.

101

Before being diagnosed with breast cancer, she had a regular

menstrual cycle. However, menses ceased while on chemotherapy
and she developed hot ushes.
After completing chemotherapy, she was started on tamoxifen.
At 2 years, her menses had not returned but the hot ushes were
less frequent. FSH and LH were in the postmenopausal range for
the laboratory concerned. However, the level of oestradiol was
slightly higher than expected. Bone mineral density (BMD) was
normal for her age, and she was switched from tamoxifen to an
AI.
Three and six months after starting the AI, her hormone prole
was substantially unchanged. After a further 6 months, however,
she noticed vaginal spotting. Her hormone prole was again assessed and showed a rise in E2 while FSH had fallen. This patient
was therefore regarded as perimenopausal.
The options discussed with her were switching back to tamoxifen or continuing on an AI but with accompanying ovarian ablation. She had completed her family and was keen to maintain

Fig. 1. Recommendations on the assessment of menopausal status in women with early breast cancer who are to have adjuvant endocrine therapy.

102

O. Ortmann et al. / Cancer Treatment Reviews 37 (2011) 97104

optimal endocrine therapy in view of the recurrence risk associated with her node-positive disease. She therefore elected to undergo a laparoscopic oophorectomy and continue on the AI. At
24 months, her BMD had fallen and she was started on an oral bisphosphonate. She completed 5 years of adjuvant endocrine
treatment.
Discussion
Perimenopausal women with early breast cancer can be divided
into at least three separate categories:
1. Treatment-nave women undergoing a natural menopausal
transition.
2. Women with treatment-induced amenorrhoea (TIA), who can
be further subdivided into:
a) patients with chemotherapy-induced amenorrhoea (CIA);
b) patients with cessation of menses while taking tamoxifen.
The present paper is not intended to provide clear-cut clinical
recommendations on the use of AIs in perimenopausal women.
These are difcult to make and are still a matter of debate. In general, however, at least some of these women, i.e. those at increased
risk of relapse, may be candidates for AIs either up-front or as
sequential treatment following tamoxifen. Goss et al. have recently
presented (data not yet formally published) a subset analysis of the
MA17 trial relating to women who were premenopausal at diagnosis and became menopausal because of oophorectomy or during
adjuvant tamoxifen or chemotherapy.40 This retrospective analysis
suggests that these women beneted as much as patients who
were postmenopausal at diagnosis from being randomised to extended letrozole following completion of 5 years of tamoxifen.
However, there are certain caveats when recommending an AI to
perimenopausal women.
First, the reduced feedback of oestrogen to the hypothalamus
and pituitary can lead to increased secretion of gonadotrophins
which in turn can cause resumption of oestrogen production by
the ovaries. In both the natural and treatment-induced perimenopause (especially the latter), resumption of ovarian function can
occur even in women aged 50 years and above, and even after
many months of amenorrhoea.33
Secondly, tamoxifen has diverse effects on the ovary. One potential effect is stimulation of ovarian function with increases in
circulating E2. Hormone proles may therefore be an unreliable
reection of true ovarian function and should be interpreted with
caution.
Thirdly, single measurements of FSH, LH, E2 and inhibin-B reect ovarian function only at the time they are taken and therefore
are not useful for predicting the potential for ovarian recovery.
The combination of an AI plus a GnRH analogue (which is not a
standard treatment but is sometimes considered in selected cases)
suppresses plasma oestrogen levels more profoundly than either
GnRH analogues alone41 or the combination of GnRH analogue
and tamoxifen.42 Nevertheless, resumption of ovarian function
has been reported even when an AI is given in combination with
a GnRH analogue. GnRH analogues do not always lead to complete
suppression of ovarian function and combination treatment with
an AI is therefore safe only if it can be established (and periodically

conrmed) that hormone levels are postmenopausal. It should be

noted that to date the only trial to compare GnRH analogue with
tamoxifen or an AI (Austrian Breast and Colorectal Cancer Study
Group XII) showed no difference in disease-related outcomes.15

Box 1. Recommendations of the expert panel on the assessment of menopausal status in women with early breast cancer
who are to have adjuvant endocrine therapy (summarised in
Fig. 1).
(1) Treatment-nave women undergoing a natural menopausal transition.
In women whose postmenopausal status cannot reasonably be confirmed by available tools, treatment
with AIs alone is contraindicated and tamoxifen should
be given. E2 and gonadotrophin levels should be monitored serially (i.e. every 36 months) to allow an AI to
be started as soon as postmenopausal status is confirmed. The added value of ovarian suppression
remains to be defined.
(2) (a) Women with chemotherapy-induced amenorrhoea.
The effects of chemotherapy on ovarian function vary
with the agent(s) used, the duration of treatment, ovarian reserve, and the age of the patient. The likelihood of
resumed ovarian function diminishes as a woman
approaches the mean age of natural menopause
(51 years). Women with CIA who are younger than 40
are more likely to resume ovarian function and should
not receive an AI alone. In older women, provided that
validated and reliable testing is available, E2 and FSH
levels should be measured at baseline. If levels are
consistent with postmenopausal status, an AI can be
started. One reasonable approach is to undertake serial
monitoring at 3 and 6 months and then at intervals of
6 months during treatment. However, if reliable testing
is not available, AIs should be used alone only with
great caution.
(b) Women with cessation of menses while taking
tamoxifen.
In women who develop amenorrhoea while taking
tamoxifen and who are candidates for switching to an
AI, it is advisable first to check E2 and FSH levels. Measurements should be made at baseline and then serially (e.g. every 36 months) during treatment.
In summary:
When menopausal status is uncertain, starting endocrine
therapy with tamoxifen rather than an AI provides effective
treatment while allowing a period during which the patients
menopausal status may become clearer. If in doubt wait!
However, given the benefit certain women may gain from
receiving an AI rather than continuing on tamoxifen, it is
important regularly to monitor the hormonal status of perimenopausal patients so that treatment can be switched when
appropriate.

O. Ortmann et al. / Cancer Treatment Reviews 37 (2011) 97104

Box 2.Contraception.
 Following adjuvant chemotherapy or endocrine treatment
for breast cancer, menstrual cycles may become irregular
or cease. However, ovulation may occur.
 In women who were pre- or perimenopausal at diagnosis,
contraception should be offered even in the absence of
menses until postmenopausal status is confirmed. (It is
worth remembering that tamoxifen was originally
designed as a fertility drug.)
 Contraception should be non-hormonal.

Box 3.Bone health.

 Bone loss is most rapid in the 5 years post menopause.
 Risk factors for osteoporosis should be identified. Their
impact on fracture risk can be evaluated using the WHO
risk assessment available at www.FRAX.sheffield.ac.uk.
 While tamoxifen preserves bone once menopause is
established, it may not prevent bone loss following CIA
or during ovarian suppression therapy.
 A baseline measurement of bone mineral density (BMD) is
recommended to guide future management.
 Advice on exercise, cessation of smoking and moderation
of alcohol should be provided.
 Adequate calcium (1 g/day) and vitamin D (400800 i.u.) is
advised in all postmenopausal women.
 Women with BMD T score of < 2 or < 1 and annual bone
loss >4%/year should be considered for bisphosphonate
treatment.43,44

Funding
This project was made possible by an unrestricted grant from
Pzer Inc.
Acknowledgement
Rob Stepney PhD (medical writer, Charlbury, UK) assisted in the
drafting of this paper. Editorial assistance was provided by TRM
Oncology (The Hague, The Netherlands).
References
1. <http://seer.cancer.gov/statfacts/html/breast.html> accessed 02.03.09.
2. Tish Knobf M. The inuence of endocrine effects of adjuvant therapy on quality
of life outcomes in younger breast cancer survivors. The Oncologist
2006;11:96110.
3. <http://www.eurohealth.ie/newrep/trends.htm> accessed 08.09.09.
4. Ganz PA, Greendale GA, Petersen L, et al. Breast cancer in younger women:
reproductive and late health effects of treatment. J Clin Oncol 2003;21:418493.
5. Goldhirsch A, Ingle JN, Gelber RD, et al. Thresholds for therapies: highlights of
the St. Gallen international expert consensus on the primary therapy of early
breast cancer 2009. Ann Oncol 2009;20:131929.
6. Goodwin PJ. Options for preservation of fertility in women. N Engl J Med
2005;353:141820.
7. Jakesz R, Jonat W, Gnant M, et al. Switching of postmenopausal women with
endocrine-responsive early breast cancer to anastrozole after 2 years adjuvant
tamoxifen: combined results of ABCSG trial 8 and ARNO 95 trial. Lancet
2005;366:45562.

103

8. Coombes RC, Kilburn LS, Snowdon CF, et al. Survival and safety of exemestane
versus tamoxifen after 23 years tamoxifen treatment (Intergroup Exemestane
Study): a randomised controlled trial. Lancet 2007;369:55970.
9. The BIG-1 98 Collaborative Group. Letrozole therapy alone or in sequence with
tamoxifen. N Engl J Med 2009;361:76676.
10. Regan MM, Pagani O, Walley B, et al. Premenopausal endocrine-responsive
early breast cancer: who receives chemotherapy? Ann Oncol 2008;19:123141.
11. Peto R et al. Early Breast Cancer Trialists Collaborative Group. The worldwide
overview: new results for systemic adjuvant therapies. San Antonio breast
cancer symposium; 2007. p. A5.
12. Breast International Group (BIG) 1-98 Collaborative Group. A comparison of
letrozole and tamoxifen in postmenopausal women with early breast cancer. N
Engl J Med 2005;353:274757.
13. Ingle JN, Dowsett M, Cuzick J, Davies C. Aromatase inhibitors vs. tamoxifen for
postmenopausal women with ER positive breast cancer: meta-analyses of
randomized trials of monotherapy and switching strategies. San Antonio breast
cancer symposium; 2008 Abs. 12.
14. Winer EP, Hudis C, Burstein HJ, et al. American Society of Clinical Oncology
Technology Assessment on the use of aromatase inhibitors as adjuvant therapy
for postmenopausal women with hormone-receptor positive breast cancer:
status report 2004. J Clin Oncol 2005;23:61929.
15. Gnant M, Mlineritsch B, Schippinger W, et al. Endocrine therapy plus zoledronic
acid in premenopausal breast cancer. N Engl J Med 2009;360:67991.
16. Faddy MJ, Gosden RG. A mathematical model of follicle dynamics in the human
ovary. Hum Reprod 1995;10:7705.
17. Randolph JF, Crawford S, Dennerstein L, et al. The value of follicle-stimulating
hormone concentration and clinical ndings as markers of the late menopausal
transition. J Clin Endocrin Metab 2006;91:303440.
18. Burger HG, Hale GE, Robertson DM, et al. A review of hormonal changes during
the menopausal transition: focus on ndings from the Melbourne Womens
midlife health project. Hum Reprod Update 2007;13:55965.
19. Sowers MR, Zheng H, McConnell D, et al. Oestradiol rates of change in relation
to the nal menstrual period in a population-based cohort of women. J Clin
Endocrin Metab 2008;93:384752.
20. Soules MR, Sherman S, Parrott E, Rebar R, Santoro N, Utian W, et al. Stages of
reproductive aging workshop (STRAW). J Womens Health Gend Based Med
2001;10:8438.
21. Gracia CR, Sammel MD, Freeman EW, Lin H, Langan E, Kapoor S, et al. Dening
menopause status: creation of a new denition to identify the early changes of
the menopausal transition. Menopause 2005;12:12835.
22. Prior JC. Ovarian aging and the perimenopausal transition: the paradox of
endogenous ovarian hyperstimulation. Endocrine 2005;26:297300.
23. Welt CK, Jimenez Y, Sluss PM, et al. Control of estradiol secretion in
reproductive ageing. Hum Reprod 2006;21:218993.
24. Sowers MR, Eyvazzadeh AD, McConnell D, et al. Anti-mullerian hormone and
inhibin B in the denition of ovarian aging and the menopause transition. J Clin
Endocrinol Metab 2008;93:347883.
25. Breast cancer. NCCN practice guidelines in oncology v.2.; 2010.
26. Walshe JM, Denduluri N, Swain SM. Amenorrhea in premenopausal women
after adjuvant chemotherapy for breast cancer. J Clin Oncol 2006;24:576979.
27. Swain SM, Jeong J-H, Geyer CE, et al. NSABP B-30: denitive analysis of patient
outcome from a randomized trial evaluating different schedules and
combinations of adjuvant therapy containing doxorubicin, docetaxel and
cyclophosphamide in women with operable, node-positive breast cancer. San
Antonio breast cancer symposium; 2008 Abs. 75.
28. Speroff L, Glass RH, Kase NG. Clinical gynecologic endocrinology and infertility.
6th ed. Philadelphia, PA: Lippincott, Williams and Wilkins; 1999.
29. Shapiro CL, Recht A. Side effects of adjuvant treatment of breast cancer. N Engl J
Med 2001;344:19972008.
30. Davis AL, Klitus M, Mintzer DM. Chemotherapy-induced amenorrhea from
adjuvant breast cancer treatment: the effect of addition of taxanes. Clin Breast
Cancer 2005;6:4214.
31. Fornier MN, Modi S, Panageas KS, et al. Incidence of chemotherapy-induced,
long-term amenorrhea in patients with breast carcinoma aged 40 years and
younger after adjuvant anthracycline and taxane. Cancer 2005;104:15759.
32. Tham YL, Sexton K, Weiss H, et al. The rates of chemotherapy-induced
amenorrhea in patients treated with adjuvant doxorubicin and
cyclophosphamide followed by a taxane. Am J Clin Oncol 2007;30:12632.
33. Di Cosimo S, Alimonti A, Ferretti G, et al. Incidence of chemotherapy-induced
amenorrhea depending on the timing of treatment by menstrual cycle phase in
women with early breast cancer. Ann Ancol 2004;15:106571.
34. Smith IE, Dowsett M, Yap Y-S. Adjuvant aromatase inhibitors for early breast
cancer after chemotherapy-induced amenorrhoea: caution and suggested
guidelines. J Clin Oncol 2006;24:24447.
35. Hargis JB, Nakajima ST. Resumption of menses with initiation of letrozole after
ve years of amenorrhea on tamoxifen: caution needed when using tamoxifen
followed by aromatase inhibitors. Cancer Invest 2006;24:1747.
36. Requena A, Herrero J, Landeras J, et al. Use of letrozole in assisted reproduction:
a systematic review and meta-analysis. Hum Reprod Update 2008;14:57182.
37. Mourits MJ, de Vries EG, Willemse PH, et al. Ovarian cysts in women receiving
tamoxifen for breast cancer. Br J Cancer 1999;79:17614.
38. Mourits MJ, de Vries EG, ten Hoor KA, et al. Beware of amenorrhea during
tamoxifen: it may be a wolf in sheeps clothing. J Clin Oncol 2007;25:37878.

104

O. Ortmann et al. / Cancer Treatment Reviews 37 (2011) 97104

39. Ravdin PM, Fritz NF, Tormey DC, Jordan VC. Endocrine status of premenopausal
node-positive breast cancer patients following adjuvant chemotherapy and
long-term tamoxifen. Cancer Res 1988;48:10216.
40. Goss PE, Ingle JN, Martino S, et al. Outcomes of women who were
premenopausal at diagnosis of early stage breast cancer in the NCIC CTG
MA17 trial. Cancer Res 2009;69S Abs. 13.
41. Jannuzzo MG, Di Salle E, Spinelli R, et al. Estrogen suppression in
premenopausal women following 8 weeks of treatment with exemestane and
triptorelin versus triptorelin alone. Breast Cancer Res Treat 2009;113:4919.

42. Rossi E, Morabito A, De Maio E, et al. Endocrine effects of adjuvant

letrozole + triptorelin
compared
with
tamoxifen + triptorelin
in
premenopausal patients with early breast cancer. J Clin Oncol 2008;26:
26470.
43. Reid DM, Doughty J, Eastell R, et al. Guidance for the management of breast
cancer treatment-induced bone loss: A consensus position statement from a UK
expert group. Cancer Treatment Rev 2008;34(Suppl. 1):S3S18.
44. Hadji P, Body J-J, Aapro MS, et al. Practical guidance for the management of
aromatase inhibitor-associated bone loss. Ann Oncol 2008;19:140716.