Age
Annual incidence
<20
0.1/100,000
20–24
1.4/100,000
25–29
8.1/100,000
30–34
24.8/100,000
35–39
58.4/100,000
40–44
116.1/100,000
Young age, even after adjustment for socio-demographic and tumor characteristics, is generally considered an independent predictor of poorer survival after BC [4]. In a series of 873 patients aged ≤45 years from 20 public data sets, proliferation gene signatures showed no significant interaction with age in estrogen receptor positive/human epidermal growth factor receptor negative (ER+/HER2−) tumors but an inferior relapse-free survival was suggested in this subgroup as compared to women >40 years at diagnosis [2]. On the contrary, the outcome of 315 very young patients (<35 years at BC diagnosis) with Luminal A subtype who received adjuvant endocrine therapy (ET) was similar to that of older women [5].
The increased number of long-term young BC survivors has focused the attention of healthcare professionals to long-term adverse effects of cancer therapies. In addition to the risk of heart failure and secondary neoplasms, oncologists must consider the impact of antineoplastic treatments on premature ovarian failure and, thus, on fertility.
Young women with BC often face dilemmas about fertility, pregnancy, breastfeeding, and contraception. In the last decades, a trend toward delaying childbearing has been observed and the number of childless women at BC diagnosis is still likely to increase. A significant number of premenopausal BC survivors in the USA (approximately 20,000 women) were estimated to be at risk for infertility [7], and about half of them might want children and could benefit from fertility counseling and preservation. Preliminary data of the Helping Ourselves, Helping Others (HOHO) study, a prospective observational study conducted in the USA in women with BC diagnosed <40 years to address disease and psychosocial outcomes at diagnosis and during long-term (10 years) follow-up, show that 68 % of women discussed fertility issues with their physicians before starting therapy and 51 % were concerned about becoming infertile after treatment [8]. Despite these worries, only 10 % of patients took special steps to lessen the chance of infertility. Eleven percent of the studied population also considered receiving ET for <5 years. Unpublished preliminary data from the cohort of women followed outside the USA, within the International Breast Cancer Group (IBCSG) HOHO study (IBCSG 43–09), show that 20 % of patients desire children after BC and are willing to take <5 years of adjuvant tamoxifen. A prospective survey in 212 evaluable patients with ER+ early BC, less than 37 years at diagnosis, from 5 regions (Europe/US/Canada/Middle-East/Australia) showed almost 40% of patients were interested in a study of ET interruption to allow pregnancy [9].
The impact of anticancer treatments on reproductive organs may be direct (e.g., pelvic surgery or irradiation) or by influence of the hormonal milieu (e.g., ET, chemotherapy, targeted therapies, alteration of the pituitary axis subsequent to cranial irradiation).
3.2 Cytotoxic Chemotherapy and Targeted Therapies
Ovaries contain a fixed pool of oocytes that do not proliferate and are not replaced and whose number declines with age. Cytotoxic chemotherapy affects primordial follicles, oocytes, and granulosa cells. The most relevant toxic effect is the loss of follicles under maturation which results in ovulatory dysfunction and subsequent amenorrhea. Follicular atrophy can be reversed according to the number of active follicles remaining after the end of chemotherapy. As a consequence, the probability of chemotherapy-induced amenorrhea (CIA) depends on the age at the time of treatment, the type of chemotherapy received, its duration, and drug/s cumulative doses. Overall, the rate of CIA is between 20 % and 70 % in women <40 years but can approach 100 % in women >40 years (Table 3.2).
Table 3.2
Risk of chemotherapy-induced amenorrhea
Age | Regimen | Degree of risk |
|---|---|---|
<30 | AC × 4 and docetaxel × 4 | 6 % |
CMF, CEF, or CAF × 6 | <20 % | |
30–39 | AC × 4 and docetaxel × 4 | 12 % |
AC, EC × 4 | <20 % | |
CMF, CEF, or CAF × 6 | 30–70 % | |
≥40 | AC × 4 and docetaxel × 4 | 35 % |
AC, EC × 4 | 30–70 % | |
CMF, CEF, or CAF × 6 | >80 % | |
All age | Methotrexate + fluorouracil | Very low |
Monoclonal antibodies | Little evidence | |
Taxanes | Little evidence |
The two main mechanisms of chemotherapy-induced ovarian toxicity are direct follicle and oocyte apoptosis [10] and vascular damage [11]. Compared to untreated women, patients having received chemotherapy show a significantly lower follicle count [10]. Another important mechanism of ovarian injury is focal damage of the ovarian cortex through hyalinization of cortical vessels and intimal fibrosis. Ovaries exposed to chemotherapy show several areas of focal cortical subcapsular fibrosis [12].
CIA can either be “temporary” or “permanent.” Temporary amenorrhea is mainly related to insufficient follicle development and alteration of hypothalamic function by disrupted estrogen metabolism. Permanent amenorrhea is more closely related to the direct toxicity of chemotherapy on the ovarian reserve. Transient menstrual irregularity or amenorrhea is common during chemotherapy, but a proportion of patients will resume menses within 6–12 months from treatment completion [13], the time required for damaged developing follicles to be replaced by new follicles from the remaining primordial follicle pool.
The most frequently used drugs in BC are alkylating agents (cyclophosphamide), anthracyclines, taxanes, and antimetabolites (methotrexate, 5-fluorouracil, capecitabine, gemcitabine). Alkylating agents are associated with the highest ovarian toxicity. The median dose of cyclophosphamide required to induce amenorrhea increases with decreasing age (5 g in 40-year-old patients, 9 g in 30-year-old patients, 20.4 g in patients <30 years). No data are available on the incidence of amenorrhea after single-agent anthracyclines. On the contrary, the anthracycline-containing combination regimens are highly toxic on ovarian function: about 34 % of women receiving the AC regimen (doxorubicin + cyclophosphamide) develop amenorrhea; CEF (cyclophosphamide-epirubicin-5-fluorouracil) and CMF (cyclophosphamide-methotrexate-5-fluorouracil) are associated with greater ovarian toxicity, with reported amenorrhea rates of 51 % and 43 %, respectively. The impact of taxanes on primordial follicles is still unclear. The addition of paclitaxel to the AC regimen, concurrently or sequentially, does not apparently increase the rate of amenorrhea [14]. Antimetabolites are less toxic for ovaries [15].
Fertility can be compromised even if women continue or resume menses after chemotherapy [16], and they can undergo early menopause due to the loss of a significant proportion of their primordial follicle pool [17]. Anthracycline- and/or cyclophosphamide-containing regimens may determine a loss of ovarian reserve of about 10 years, i.e., the amount of primordial follicles of women aged 26–27 years after this type of combination chemotherapy is similar to that of women aged 36–37 years. The reduction in the antimüllerian hormone (AMH), the best biochemical marker of ovarian reserve currently available, was shown to be associated with a loss of ovarian reserve of about 10 years [18]. Low pretreatment AMH was also found to be an independent predictor of CIA at 2 years after the end of chemotherapy (P = 0.005; odds ratio 0.013), independently from age, in 59 premenopausal women with early BC, as compared with other markers of ovarian function (FSH and inhibin B) [19].
Patients with triple-negative BCs and BRCA mutation carriers often have multiple deficits in DNA repair pathways and may selectively benefit from platinum derivatives and Poly-(ADP-ribose) polymerase (PARP) inhibitors [20]. Trials are ongoing, but PARP inhibitors are likely to be less gonadotoxic than cyclophosphamide-based regimens. In a study of 168 cancer patients (38 with BC), the odds ratio (OR) of platinum-related ovarian failure in exposed versus unexposed patients was 1.77, second only to alkylating agents (OR 3.98) [21]. The available data with trastuzumab [16] report no increase in the likelihood of CIA. Data on bevacizumab are limited to patients with colorectal cancer: ovarian failure occurred in 34 % of women receiving a bevacizumab-containing regimen compared with 2 % of women receiving the same regimen without bevacizumab. Only approximately one fifth of these women recovered ovarian function and the US Food and Drug Administration (FDA) issued a warning in 2011 in order to properly inform women before starting treatment [22]. Information with newer chemotherapy agents or other targeted drugs (e.g., epothilones, lapatinib, pertuzumab) is missing.
A recent meta-analysis of 15,916 premenopausal BC patients from 46 studies showed that cyclophosphamide- taxane- and anthracyclines-based regimens significantly increased the incidence of CIA with pooled ORs of 2.25 (95 % CI 1.26–4.03, p = 0.006), 1.26 (95 % CI 1.11–1.43, p = 0.0003), and 1.39 (95 % CI 1.15–1.70, p = 0.0008), respectively. The three-drug combination regimens of cyclophosphamide, anthracyclines, and taxanes caused the highest rate of CIA compared with other three-drug combinations (OR 1.41, 95 % CI 1.16–1.73, P = 0.0008). The addition of tamoxifen was also associated with a higher incidence of CIA, with an OR of 1.48 [23].
The impact of cytotoxic therapy on ovarian function and fertility possibly depends not only on the type of drug and its cumulative doses, but also on the type of schedule chosen. It is well known that for some drugs (i.e., paclitaxel) the use of different schedules is correlated with variations of their pharmacodynamics [24]. It is therefore intuitive to think that a metronomic administration (often used in the metastatic setting) may be less toxic on the ovary, despite no data to confirm this hypothesis.
3.3 Endocrine Therapy
Being estrogen promoters, and possibly initiators, of most BCs, blocking their synthesis/function represents a logic therapeutic target in women with ER+ BC. In premenopausal women estrogen synthesis primarily occurs in the ovaries, but also in fat tissue, muscles, skin, stromal breast cells, adrenal glands, and in the neoplastic tissue itself.
The most commonly prescribed endocrine pharmacological treatments are selective estrogen receptor modulators (SERMs) (i.e., tamoxifen and toremifen), gonadotropin-releasing-hormone agonists (GnRHa), and aromatase inhibitors (AIs). In premenopausal women, tamoxifen for at least 5 years is a standard of care [25].
GnRHa have been introduced in the treatment of BC in premenopausal women at the end of the 1980s. GnRHa have a biphasic effect on the pituitary gland. Initially, they stimulate the secretion of both follicle-stimulating-(FSH) and luteinizing hormone (LH), while with long-term continuous administration, pituitary cells become resistant. The final result is a reversible inhibition of FSH and LH secretion and a fall in circulating levels of sex hormones similar to that produced by irreversible surgical- or radiation-induced castration. Serum levels of 17 beta-estradiol and progesterone fall within the 3rd–4th week after therapy start and the levels of LH and FSH remain suppressed. If a GnRHa is given, estradiol levels should be checked on a regular basis (at least every 6 months) because in some patients ovarian suppression is not achieved [26].
The recently published results of the Suppression of Ovarian Function Trial (SOFT) showed that, after a median follow-up of 67 months, the addition of ovarian function suppression/ablation (OFS/OA) to tamoxifen did not result in a significant benefit in terms of disease-free-survival (DFS) in the overall study population. However, for women who were considered at sufficient risk for recurrence to warrant adjuvant chemotherapy and who remained premenopausal after a median of 8 months after its completion, the addition of OFS significantly improved disease outcomes, especially if younger than 35 years at diagnosis [27].
Recent data from the randomized trials Tamoxifen and Exemestane Trial (TEXT) and SOFT also showed that the AI exemestane plus OFS/OA significantly reduces recurrences as compared with tamoxifen plus OFS/OA [28]. On the other hand, the ATLAS and aTToM trials represent the first evidence of a beneficial effect of extended adjuvant tamoxifen (10 years) in premenopausal women [29, 30]. As a consequence of these data supporting complete estrogen deprivation and/or longer duration of ET, larger numbers of women will be older and thus at higher risk for infertility at the time their therapy is completed.
Oophorectomy remains a viable ovarian suppression option both in the adjuvant and metastatic setting. Ovarian ablation by radiation therapy (RT) (15–20 Gy in 10–18 fractions to a modified pelvic treatment volume) is generally as effective as surgical oophorectomy or GnRHa administration but may take some months to be complete [31, 32]. These last two techniques, which are associated with permanent ovarian suppression, might be more indicated in older premenopausal patients and represent a valid alternative especially in country with limited resources [33].
The impact of single-agent tamoxifen on ovarian function is not well understood, since the number of premenopausal patients not receiving also adjuvant chemotherapy in the literature is very small. Tamoxifen may interfere with normal negative pituitary feedback mechanisms resulting in increased secretion of gonadotropins and, hence, in increased ovarian estrogen production [34]. The consequent hyper-estrogenism can be associated with ovarian cysts and oligo-amenorrhea [35, 36]. Overall, the incidence of tamoxifen-induced amenorrhea ranges between 16 and 38 % [37]. Amenorrhea while on tamoxifen did not, in itself, translate in definitive menopause in 65 patients receiving single-agent adjuvant tamoxifen as compared to 68 patients treated with tamoxifen after adjuvant chemotherapy [38]. Women should therefore be informed of the possibility of getting pregnant while on tamoxifen, despite having amenorrhea, and of the need for adequate non-hormonal contraception [26].
As compared to permanent CIA, tamoxifen-induced oligo-amenorrhea and OFS induced by GnRHa are reversible and ovarian function usually recovers after 3–6 months from their discontinuation, as the ovarian reserve is not permanently damaged. This can avoid the burden of premature menopause and be particularly appealing in women who did not complete their childbearing before BC diagnosis. Within the Breast International Group (BIG)-North American Breast Cancer Group (NABCG) collaboration the IBCSG has just launched a trial (Pregnancy Outcome and Safety of Interrupting Therapy for women with endocrine responsIVE breast cancer – POSITIVE – IBCSG 48–14 – clintrials.gov NCT02308085) which will assess the pattern of fertility recovery and pregnancy in women with maternity desire under different ETs.
AIs interfere with the enzyme aromatase, which, in highly estrogen-sensitive tissues, such as the breast, uterus, vagina, bone, brain, heart, and blood vessels, is responsible for the final step of estrogen synthesis from androgens (androstenedione and testosterone). Premenopausal women have a large amount of aromatase substrate in the ovary: AIs induce, by a pituitary loop effect, a dramatic increase of gonadotrophins and a subsequent increase in hormone levels. AIs should therefore not be given to premenopausal women without the addition of a GnRHa. The initial surge in FSH and LH induced by GnRHa before pituitary suppression together with the ovarian stimulation triggered by AIs are also used for embryo/oocyte cryopreservation. Women should be counseled to use non-hormonal contraception both during the first weeks of treatment and afterwards as sustained ovarian suppression is not achieved in some patients. The preliminary results of the SOFT-EST prospective substudy, measuring serial serum estrogens in 116 patients under ET (either Tamoxifen or Exemestane) and GnRHa in the SOFT trial show that about 20 % of patients had suboptimal estrogen suppression [39].
3.4 Radiation Therapy
RT is recommended to all women who underwent conservative breast surgery because significantly reduces the rate of local recurrence and improves overall survival [40]. A boost on the tumor bed is of particular benefit in young women [41]. RT is also recommended to women who underwent mastectomy and have a high risk of loco-regional relapse [42].
Ovarian follicles are sensible to radiation damage. Adjuvant loco-regional RT for early BC is not associated with significant ovarian toxicity, although internal scatter radiation can reach the pelvis and ovaries (2.1–7.6 Gy). In the palliative setting, it is sometimes necessary to irradiate the pelvis, for the treatment of bone or visceral metastases. The dose tolerance of the ovary is dependent on several factors (the volume irradiated, the total radiation dose, the fractionation schedule, and the patient’s age at the time of treatment). Radiation doses to the pelvis exceeding 24 Gy will likely produce permanent ovarian ablation, but lower doses can be associated with premature ovarian failure with increasing age [43]. Cranial irradiation, affecting the hypothalamic-pituitary axis, may also impair fertility [44].
3.5 Surgical Treatment
Breast surgery does not have an impact on fertility, except for oophorectomy, mainly performed in developing countries as a substitute of costly GnRHa therapy, or as prophylactic surgery in patients harboring BRCA1/2 mutations.
3.6 Fertility Assessment After Breast Cancer Treatment
When assessing the impact of different chemotherapies on fertility, many clinical trials used amenorrhea or menstrual irregularities, which are not reliable markers of infertility, as endpoints. Despite maintenance or resumption of regular menses after BC treatment, fertility may be compromised due to the poor quality of surviving oocytes [45]. Women with decreased ovarian reserve often have shorter, more regular cycles due to accelerated follicle development. Permanent amenorrhea is also not uniformly defined across studies. The most used definitions range from irregular menses in the 1st year after treatment completion to continuous cessation of menses for more than 1 year [46].
Ovarian function recovery (OFR) can be measured by monitoring circulating levels of FSH, LH, estradiol, inhibin B, and AMH. Decline in the ovarian reserve leads to low levels of estradiol, inhibin B, and AMH, normally produced by the granulosa cells of the ovarian follicles. AMH is the most sensitive of all these tests, as it is consistent throughout the menstrual cycle and more closely approximates the number of ovarian primordial follicles [47]. Low pre-chemotherapy AMH and older age were both statistically significant predictors of CIA (p = 0.04 and p = 0.008, respectively) in 124 patients with early BC participating in the multicenter randomized controlled trial ECOG5103 (doxorubicin-cyclophosphamide followed by paclitaxel with either placebo or one of two durations of bevacizumab therapy) [48].
The widespread use of hormone level’s monitoring has been limited by costs, lack of sensitivity, and cross reproducibility of available assays. Single measurements reflect ovarian function only at that specific time point and therefore do not predict the potential for sustained ovarian recovery. In addition, the validity of these tests following chemotherapy for BC is not yet been validated, and not all authors agree on their reliability in this particular setting of patients. A new highly sensitive AMH assay, prospectively used and validated in a cohort of 98 women with early BC, showed a tenfold increased sensitivity as compared to AMH, inhibin B, FSH, and estradiol measured with standard methods. In particular, the study showed that, even in women with regular menses, AMH measured 2 years after chemotherapy was generally very low for their age (mean age 35 years) and similar to older women who did not receive chemotherapy (mean age 45 years) [49].
In addition to AMH, an ultrasound-guided estimation of ovarian volume and antral follicle count, a simple and relatively inexpensive technique [50], can be used to estimate ovarian reserve. The sensitivity and specificity of this method are still being evaluated.
3.7 Psychosocial Impact of Fertility Impairment
The impact of cancer-related infertility on long-term distress and quality of life (QoL) has been poorly investigated in BC patients.
In the above mentioned prospective HOHO study, 37 % of patients wished future biologic children before BC diagnosis as compared to 26 % at the time of survey and 9 % did not want more children because they were afraid that pregnancy would increase their risk of recurrence [8]. A cross-sectional survey was conducted by phone interview in 240 US young women (mean age at the time of survey 43.7 years) diagnosed with cancer (130 with BC), unselected for their desire for children at diagnosis, 5–10 years post-treatment. A significant higher distress, more intrusive thoughts, and avoidance strategies were experienced by childless women as compared to women with adopted/stepchildren and women with at least one biological child, irrespective of cancer type [51].
One-hundred and thirty-one women with early BC diagnosed ≤40 years, originally participating in the Women’s Healthy Eating and Living (WHEL) Study evaluating the influence of diet on BC outcome, participated in the WHEL Survivorship Study to investigate whether the level of reproductive concerns after treatment was associated with long-term depressive symptoms. Data were collected an average of 12 years post-diagnosis: reproductive concerns were a significant contributor to persistent depressive symptoms (p = 0.0002) as were not having children, being nulliparous at diagnosis and treatment-related ovarian damage (all with p < 0.01) [52].
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
