Introduction

Cancer in childhood is rare. The most recent SEER Cancer Statistics Review, released in April 2018 and covering 2011–2015, shows that for ages 0–14 the incidence of all cancer sites was 16.4 per 100,000 children, with mortality 2.1 per 100,000. Five-year relative survival for the same age group was 83.8% for years 2008–2014 [1]. (Rates from the European Cancer Information System are similar but are currently available only to 2012.) As of 2015, 1 in 750 of the adult population of the United States was a long-term survivor of childhood cancer [1, 2]. Cancer is more common after puberty, with incidence rising from 12.9 for ages 5–9 to 23.3 for ages 15–19 in the United States for years 2011–2015 [1], and many of these patients will be cured by combination treatment, with surgery, chemotherapy, and radiotherapy. Long-term survivors are nevertheless at risk of developing a number of late sequelae [3], including impaired fertility, adverse pregnancy outcomes, and health problems in offspring [4–6]. Loss of fertility is one of the most devastating consequences of radiotherapy for these young patients, who, having overcome their disease, have expectations of a normal reproductive life.

Normal Ovarian Development and Follicular Depletion

Current understanding of the human ovarian reserve is that the ovary establishes several hundred thousand nongrowing follicles (NGFs) by followed by a decline in the population by menopause, when approximately 800 remain at an average age of 50–51 [7]. With approximately 450 ovulatory monthly cycles in the normal human reproductive life span, this progressive decline in NGF numbers is attributed to follicle death by apoptosis. The Wallace–Kelsey model of human ovarian reserve from conception to menopause best fits the combined histological evidence that has been described [8, 9] and validated using external data [7] (Figure 3.1). This model allows us to estimate the number of NGFs present in the ovary at any given age, and it suggests that 81% of the variance in NGF populations is due to age alone. Further analysis demonstrates that 95% of the NGF population variation is due to age alone for ages up to 25. The remaining 5% is due to factors other than age, e.g. smoking, body mass index (BMI), parity, and stress. We can speculate that as chronological age increases, factors other than age become more important in determining the rate at which NGFs are lost through apoptosis.

Figure 3.1 The Wallace–Kelsey model of NGF decline from birth to age 60. The average ovary contains 300,000 NGFs at birth, declining to 100,000 at age 20, with about 800 remaining at average age of menopause (50–51 years)

There is speculation that this widely held tenet of mammalian ovarian function may require revision. A report in 2004 suggested the presence of germ stem cells in the adult mouse ovary [10], with two subsequent reports by the same group suggesting the ability of bone marrow–derived cells to give rise to new immature oocytes [11, 12]. More recently, another group has identified the presence of proliferative and culturable female germline stem cells in newborn and adult mouse ovaries [13]. Strikingly, when these cells (transgenically labeled) were injected into a chemotherapy-treated ovary, they became enclosed within follicles and offspring bearing the transgene were produced. These data provide a basis for reevaluating the regenerative capacity of the mammalian ovary and new approaches for overcoming fertility loss. While the emerging evidence thus appears to provide evidence in support of the existence of germ stem cells within the adult mammalian ovary, there is no robust evidence to date that these cells contribute to follicle formation leading to viable eggs [14]. There is evidence that significant NGF population regeneration can occur in extreme situations in the human ovary [15], but the Wallace–Kelsey model of ovarian follicle decline provides no supporting evidence of neo-oogenesis in normal human physiological aging.

Radiation and the Hypothalamic–Pituitary–Ovarian Axis

The ovaries may be damaged following total body, abdominal, or pelvic irradiation [16], and the extent of the damage is related to the radiation dose, fractionation schedule, and age at the time of treatment [17–19]. The human oocyte is sensitive to radiation, with an estimated LD₅₀ of less than 2 Gy [20]. This is the lethal dose to destroy 50% of NGFs present in the ovary. The number of NGFs present at the time of treatment, together with the dose received by the ovaries, will determine the “fertile window” and influence the age of premature ovarian insufficiency (POI). Long-term ovarian failure has been reported in 90% of patients after total body irradiation (TBI) (10.00–15.75 Gy) and in 97% of females treated with total abdominal irradiation (20–30 Gy) during childhood [21–23]. Our understanding of the LD₅₀ of the human oocyte has made it possible to estimate the age at which POI may occur (Figure 3.2). Furthermore, we have estimated the sterilizing dose following any given dose of radiotherapy at any given age, based upon the Wallace–Kelsey model for NGF decline (Figure 3.3) [24, 25]. This will help clinicians provide accurate information when counseling women about fertility following treatment for childhood cancer. Gonadotropin deficiency following high-dose cranial irradiation (more than 24 Gy in the treatment of brain tumors) manifests as delayed puberty or absent menses and can be treated by hormone replacement therapy. Interestingly, early puberty is often reported in females with cranial radiation doses of less than 24 Gy [26]. However, we have shown a subtle decline in hypothalamic– pituitary–ovarian function following low-dose cranial radiotherapy (18–24 Gy). This is characterized by decreased luteinizing hormone (LH) secretion throughout the cycle, an attenuated LH surge, and short luteal phases [27], which may compromise reproductive function. Cranial radiotherapy is also associated with increased risk of oligomenorrhea, reduced estradiol, and increased prolactin in survivors [28, 29]. Recent data confirm lower fertility in childhood cancer survivors treated with either hypothalamic/pituitary irradiation more than 22 Gy or ovarian/uterine irradiation more than 5 Gy [30, 31]. Nowadays, the most commonly used fertility preservation measure is the surgical transposition of the ovaries outside the irradiation field before the initiation of pelvic radiation in adults with gynecological malignancies [24]. This approach may be considered for patients not planning to receive high-dose systemic chemotherapy. Surgery is effective at protecting the ovaries from direct irradiation damage, but fertility may be affected by scatter radiation, by damage to the ovarian vasculature during surgery, or by torsion of the transposed ovary. To avoid this, ovarian tissue cryopreservation may be considered at the same time as oophoropexy.

Figure 3.2 Predicted age at POI for known age and known dose to the ovary. A 5-year-old with average NGF population treated with 1 Gy has predicted POI at age 42, losing about 8 years of reproductive life span; the same child treated with 12 Gy has predicted POI at age 12, indicating consideration of the use of fertility preservation techniques. The predictions are modifications of existing predictions [25] adapted to use the Wallace–Kelsey model of NGF population in healthy females [9]

Figure 3.3 The mean and effective sterilizing doses of radiation to the human ovary. The blue line shows the radiation in Gy needed to sterilize 50% of subjects. The red line shows the radiation in Gy needed to sterilize 97.5% of subjects. The predictions are derived from the Wallace–Kelsey model of NGF population in healthy females [9], and the estimated radiosensitivity of the human NGFs [20]

Radiation and the Uterus

The uterus is at significant risk of damage following abdominal irradiation, pelvic irradiation, or TBI in a dose- and age-dependent manner [32]. Uterine function may be impaired following radiation doses of 14–30 Gy as a consequence of disruption to uterine vasculature and musculature elasticity [23]. Even lower doses of irradiation, as in TBI, have been reported to cause impaired growth and blood flow [33]. The clinical consequences are increased risk of miscarriage and premature delivery [22]. A uterine contribution to an inability to conceive following radiotherapy is not clear but seems likely.

A small number of studies have been reported that attempted to improve uterine function in survivors of cancer with POI. In young adult women previously treated with TBI, physiological sex steroid replacement therapy improves uterine function (blood flow and endometrial thickness) and may potentially allow them to benefit from assisted reproductive technologies [33]. Larsen et al. studied uterine volume in 100 childhood cancer survivors and assessed uterine response to high-dose estrogen replacement therapy in three patients with ovarian failure and reduced uterine volume following abdominal and/or pelvic irradiation [34]. There was no significant difference in uterine volume, endometrial thickness, or uterine artery blood flow following sex steroid treatment, suggesting that higher doses of pelvic radiation cause greater damage than lower doses (as in TBI), and that this damage may be irreversible. While data are sparse for the childhood case, there is good evidence that adult women treated with radiotherapy to the abdomen are at high risk of uterine dysfunction leading to miscarriage, premature delivery, low birth weight, and placental abnormalities [35]. There is no current evidence that the child uterus is inherently better protected from the effects of radiotherapy.