Chemotherapy and radiation therapy destroy ovarian follicles, and predispose treated females to premature ovarian failure. The negative effect of chemotherapy on female fertility is dependent on the age of the patient at the time of treatment, the specific chemotherapeutic agent(s) used, and the cumulative dosing (Gracia and Woodruff 2012; Trudgen and Ayensu-Coker 2014). Similar to males, alkylating agents are the most gonadotoxic to female fertility. In contrast to males, significantly higher doses of alkylators are required to cause infertility in females. However, gonadotoxicity in females is more age dependent than males. Specifically, older age has a greater negative impact, as there is an overall smaller follicular pool during cancer therapy (Ginsberg 2011). The natural depletion of the number of follicles present in an individual female’s ovaries can be accelerated by cancer treatment. If the degree of depletion is near complete, then the result is acute ovarian failure defined by early menopause with consequent infertility that occurs during or shortly after treatment (Gracia and Woodruff 2012). If the degree of depletion is more moderate, then the individual is at risk for premature menopause, where females remain fertile following cancer therapy but have an overall shortened reproductive life span.

Radiation to the ovaries, either through direct pelvic radiation, abdominal or spine radiation, or scatter radiation, in doses as low as 1–2 Gy in girls and 4–6 Gy in adult females can have permanent negative effects on the ovaries by causing depletion of follicles (Gracia and Woodruff 2012). Irradiation may also impact fertility by causing damage to the uterine musculature and vascular structures, thereby limiting the ability of the survivor to carry a pregnancy to term. High dose cranial irradiation in the dose range of 35–40 Gy or greater, can cause hypogonadism through effects on the hypothalamus and pituitary. This exposure results in dysregulation of hormonal pathways responsible for menstruation and fertility.

Finally, the surgical resection of reproductive organs has obvious implications for infertility when hysterectomies or oophorectomies become necessary (See Table 5.2).

The primary exception from small studies is the Childhood Cancer Survivor Study (CCSS) which reported on multiple outcomes for cancer survivors including late mortality, subsequent malignant neoplasms, chronic health conditions, fertility, and health status using matched sibling controls (Ginsberg et al. 2010b). Childhood cancer survivors from the CCSS were half as likely as their matched control siblings to sire a pregnancy with an overall reported rate of premature ovarian failure of 6 % (Chemaitilly et al. 2006; Larsen et al. 2003; Longhi et al. 2000, 2003). High risk exposures identified in the study included radiation greater than 7.5 Gy to the testes and high doses of alkylating agents including cyclophosphamide dose greater than 7.5 g/m (Green et al. 2010). In fact, when controlled for these factors, survivors were equally likely as their matched control siblings to sire a pregnancy (Green et al. 2010).

Three hundred forty-one survivors of Ewing sarcoma (EWS) were separately analyzed within this cohort for their fertility status. Infertility was common among both male and female EWS survivors. Female survivors were 35 % less likely to report a pregnancy than siblings of childhood cancer survivors, even after excluding women who were surgically sterile (Ginsberg et al. 2010b). Similarly, male EWS survivors were 62 % less likely to report siring a child than male siblings (Ginsberg et al. 2010b). Contemporary therapy for EWS includes high dose cyclophosphamide and ifosfamide, so we can anticipate that the rate of infertility will not be much different for patients treated today. Another 58 patient analysis of sarcoma patients including 22 with EWS reviewed semen analyses before, during, and after treatment with chemotherapy with multi-agent chemotherapy including cyclophosphamide (Meistrich et al. 1992). While pretreatment levels were similar to the control subjects, azoospermia occurred within 4 months of treatment (Meistrich et al. 1992). The cumulative dose of cyclophosphamide was the most significant determinant of recovery to normospermic levels; approximately 70 % of those who had received doses less than 7.5 g/m² recovered, but only 10 % recovered when doses exceeded 7.5 g/m² (Meistrich et al. 1992). In a retrospective study of 17 male sarcoma survivors treated with high dose cyclophosphamide within a multi-agent chemotherapy regimen, a total dose of cyclophosphamide of 7.5 g/m² was also associated with a decreased risk of an abnormal sperm count with an incidence of azoospermia of 58.8 % (Kenney et al. 2001). This analyses also demonstrated that exposure prior to the onset of puberty did not appear to protect males from subsequent gonadal damage. All of the patients exposed to 25 g/m² of cyclophosphamide or higher were azoospermic irrespective of their pubertal status at the time of treatment (Kenney et al. 2001).

In patients with osteosarcoma, exposure to Ifosfamide has also been analyzed in several case series. Kenney et al. reviewed six patients with non-metastatic osteosarcoma who were treated with multi-agent chemotherapy (Kenney et al. 2001). They reported their results as consistent with previously reported studies that used semen analysis as their measure of spermatogenesis, demonstrating 52–89 % of survivors with abnormal sperm counts after treatment with Ifosfamide 24–45 g/m² (Kenney et al. 2001). The United Kingdom Children’s Cancer Study Group (UKCCSG) reviewed gonadal function of patients who were treated for Ewing or soft tissue sarcoma with regimens containing ifosfamide as the only potential gonadotoxic agent (Williams et al. 2008). Male patients who received high dose Ifosfamide (greater than 60 g/m²) showed 73 % to be subfertile and 27 % azoospermic (Williams et al. 2008). Hormone profiles were consistent with germ cell failure, 31 % had increased FSH, while Leydig cell function remained preserved (Williams et al. 2008). The female cohort for the study had different treatment regimens including ifosfamide (Longhi et al. 2003). Unfortunately, very few patients were willing to undergo semen analysis and/or hormone testing, so the study was only able to conclude that Ifosfamide seems to be a major cause of infertility in male osteosarcoma patients with a dose-dependent relationship between higher doses of ifosfamide with higher probability of becoming sterile (Longhi et al. 2003). They also noted that Cisplatin causes azoospermia or oligospermia with a gradual recovery of spermatogenesis in 50 % of patients after 2 years and 80 % after 5 years following cumulative dose of 600 mg/m² (Longhi et al. 2003). In contrast, the Rizzoli Institute also reviewed the reproductive function of female survivors of localized osteosarcoma. Twenty-four of their 92 patients were prepubertal at the time of chemotherapy and underwent menarche at a mean age of 13 (range 11–16), consistent with the mean age of menarche of the normal population in Italy (Longhi et al. 2000). Of the postpubertal patients, 69 % experienced amenorrhea during chemotherapy, 66/68 patients resumed menstrual activity by the end of chemotherapy, whereas the 2 patients age 39 and 43 at the start of treatment had permanent amenorrhea (Longhi et al. 2000). Of the 92 patients, only 2 patients reported a willingness to conceive without achieving a pregnancy and 20 others had successful pregnancies with healthy offspring (Longhi et al. 2000). A Finnish study demonstrated similar findings in a small cohort of female osteosarcoma survivors with all patients having amenorrhea at the completion of chemotherapy, but 90 % had some returned menses within 3–12 months of follow-up (Wikstrom et al. 2003). The three patients who had irregular menses cycles upon resumption of menstruation represented the highest cumulative dose of Ifosfamide (54.5 g/m², 63.5 g/m², and 60.4 g/m²) (Wikstrom et al. 2003). Importantly, resumption of menses does not directly correlate to fertility and pregnancy, and thus is a limitation of this small study.