Historical background

Radiation impact on gonadal function was recognized a century ago. Atomic Energy Commission studies of normal men in the 1960s confirmed exquisite sensitivity of spermatogonia to as little as 10 cGy of irradiation.¹ Oocytes were more resistant but also displayed a dose-dependent sensitivity, which resulted in sterility and POF with an age-related frequency. Research published in 1939 demonstrated that 500 cGy to human ovaries was associated with amenorrhea that persisted up to 18 months. All women over the age of 40 became permanently infertile.²

Initial reports suggesting detrimental effects of chemotherapy on reproductive function were confirmed by a pathologic study of testicular tissue of 30 men who received nitrogen mustard in the 1940s. Twenty-seven of these men had testicular atrophy and absent spermatogenesis.³

The first convincing report of menstrual irregularities in women undergoing chemotherapy appeared in 1956.⁴ Four women starting busulfan for chronic leukemia developed menopausal symptoms within 3 months. Ovarian and endometrial tissue showed findings consistent with POF. Gonadal toxicity from other drugs was soon recognized, and the list continues to grow (Table 1).

Assessment of gonadal function after gonadotoxic therapy

Assessment in males

Semen analysis has been the cornerstone of assessment of gonadal function in men, and measurement of gonadotropins (FSH, LH), anti-Müllerian hormone (AMH), and inhibin-B levels are used in children and prepubertal boys.

AMH is produced by Sertoli cells and impacts male sexual differentiation by causing regression of Müllerian ducts.⁵After age 9, AMH levels decline indicating androgen effect on Sertoli cells and early spermatogenesis.⁶ In prepubertal boys older than 9 years, Inhibin-B and basal testosterone levels assume more relevance. Leydig cells are more resistant to the effects of chemotherapy than germ cells, thus childhood survivors of cancer may have normal testosterone despite being azoospermic.⁶

Semen volume, sperm concentration, mobility, and morphology are markers of testicular function in adult men. Markers of spermatogenesis such as FSH and inhibin-B levels fluctuate widely and are not predictive of reproductive outcomes. Low or normal testosterone levels along with elevated LH levels are seen commonly in adult survivors of cancer,⁷ but are also subject to interindividual variability and lack sensitivity to detect small but meaningful changes in testicular function. Novel biomarkers including sperm messenger RNA, micro-RNAs, histone modifications, and DNA methylation patterns are under development.⁷ Genetic testing of sperm using fluorescent in situ hybridization (FISH) and DNA fragmentation identify chromosomal aneuploidy in the sex chromosomes and assess the extent of DNA damage after gonadotoxic therapy. A higher incidence of sperm aneuploidy was noted in men treated with BEP chemotherapy for testicular cancer as well as the ABVD regimen for HD.^{8, 9} Aneuploidy returned to baseline values within a few months, but some patients had persistent abnormalities for up to 2 years after therapy. Future impact of these measurements is potentially intriguing.

Assessment in females

Oocyte numbers are fixed at birth and are not replenished.¹⁰ Advancing age and anti-neoplastic therapy quantitatively and qualitatively decrease this pool and adversely affect fertility. Assessing ovarian reserve before fertility preservation is important, especially in women aged over 35.

Measurement of FSH, inhibin B, clomiphene citrate challenge test, antral follicular count (AFC), and AMH assess ovarian reserve and predict the oocyte yield with assisted reproduction. Elevated FSH (>20 mIU/mL) in the early follicular phase of the menstrual cycle indicates impaired ovarian reserve and predicts failure of assisted conception.¹¹ However, FSH level varies during the menstrual cycle, and neither FSH nor inhibin-B levels are reliable in prepubertal girls. The AFC can be quantified by transvaginal ultrasound to assess ovarian reserve but is not useful in determining oocyte quality or predicting pregnancy outcomes with IVF.¹²

The granulosa cells of ovarian antral follicles produce AMH, and serum levels are a surrogate for the number of developing ovarian follicles. AMH levels decline toward menopause. Unlike FSH and inhibin-B, serum AMH levels do not fluctuate through the menstrual cycle and are valid in children.¹⁰

Low serum AMH levels were first described in women with prior childhood cancer who still had regular menses indicating low ovarian reserve.¹³ Well-conducted studies have found low AMH levels in breast cancer and childhood HD survivors with a clear dose–response relationship between the number of chemotherapy cycles and serum AMH levels.^{14, 15} Therapy with alkylating agents and pelvic/total-body irradiation often results in low or undetectable AMH levels.¹⁶ Women with low pretreatment AMH levels are more likely to develop amenorrhea after chemotherapy for breast cancer.¹⁷ Nomograms incorporating age and AMH levels have been developed to predict postchemotherapy ovarian recovery in newly diagnosed breast cancer patients and accurately gauge the need for fertility preservation techniques. AMH levels do not predict spontaneous conception of pregnancy or pregnancy outcomes.

Effects of cytotoxic chemotherapy on gonadal function

Effects in boys

Early reports of therapy in prepubertal and pubertal boys suggested relative resistance of the less mature testicle to chemotherapy-induced effects. The frequency of testicular dysfunction, however, varies widely among studies. Three major factors determine the extent of testicular damage among prepubertal boys receiving cytotoxic chemotherapy: the specific drug, the cumulative dose of the drug, and the pubertal stage. The majority of boys progress normally through puberty without supplemental androgen.^{18, 19} Testicular volume may be reduced, however, and elevated LH levels indicate some degree of Leydig cell dysfunction.⁶

Years after treatment with single-agent cyclophosphamide, the prevalence of normal adult sperm counts has been reported in small, heterogeneous case series to range between 0% and 100%.^{20, 21} Oral cyclophosphamide cumulative doses of 0.7–52 g caused gonadal damage in 16% of prepubertal boys, but 67% of pubertal boys had evidence of gonadal dysfunction.²² Chlorambucil and azathioprine for renal disease in patients aged 6–15 years produced azoospermia in 17 of 21 patients for up to 11 years after cessation of treatment.²³

Most available data relate to multiagent chemotherapy. MOPP chemotherapy in boys with HD impairs subsequent spermatogenesis, a defect reported to last for years.^{19, 24}

A meta-analysis of 30 studies comprising 456 patients who received cyclophosphamide for renal disease, HD, or leukemia found that fewer than 10% of prepubertal boys receiving less than 400 mg/kg (total dose) of cyclophosphamide had gonadal dysfunction, whereas 30% of those over 400 mg/kg did.²⁵ Gonadal dysfunction ranged from 0% to 24% in prepubertal boys, but was 68–95% in sexually mature men. A recent analysis of 214 adult male survivors of childhood cancer treated with alkylating agents found the incidence of oligospermia and azoospermia to be 28% and 25% respectively.²⁶ Impaired spermatogenesis was unlikely when the cumulative cyclophosphamide equivalent dose (CED) was less than 4000 mg/SqM. CED was associated with a statistically significant increase in the risk per 1000 mg/SqM for azoospermia and oligospermia.²⁶ Unfortunately, poorly understood exceptions to these general trends and lack of reliable predictions for any given patient are problematic. Even small doses of alkylating drugs in prepubertal children can cause permanent sterility.

Whether Leydig cell function is affected in pubertal males is less clear. Gynecomastia with elevated FSH and LH has been reported in pubertal boys receiving MOPP treatment.¹⁹ Other studies find normal basal and stimulated gonadotropin tests,¹⁸ and abnormal testosterone responses to human chorionic gonadotropin (hCG) challenge are uncommon. Gonadotropin secretion was normal in 29 of the 32 patients studied in a cohort of 40 men treated in childhood for HD; 26 of 28 had elevated gonadotropin levels but normal serum testosterone and secondary sexual characteristics. Eleven of thirteen were azoospermic and remained so for up to 17 years.²⁷ Seventeen adult survivors of childhood sarcoma demonstrated azoospermia in 58%, oligospermia in another 30%, but normal testosterone in 94%. LH was elevated in 92% (40% of those with normal testosterone levels), suggesting Leydig cell insufficiency.⁶ Although levels of LH, FSH, and serum testosterone following chemotherapy in prepubertal boys may be normal, testicular biopsies after combination chemotherapy for acute lymphoblastic leukemia or HD commonly show seminiferous tubular damage and interstitial fibrosis.

Effects in men

Single-agent alkylating drugs induce permanent damage to adult seminiferous epithelium. Cyclophosphamide cumulative doses over 9 g result in universal azoospermia; with doses over 18 g that change is irreversible.²⁸ Multiple studies of HD patients receiving combination chemotherapy with or without procarbazine indicate that this drug is uniquely gonadotoxic. In one study of 19 patients treated with cyclophosphamide, vincristine, procarbazine, and prednisone (COPP) regimen, all remained oligospermic 11 years after therapy; 7 of 10 treated with COPP without procarbazine had return of spermatogenesis within 3 years.^{29, 30}

Methotrexate causes minimal long-term reproductive toxicity. That vincristine may be less toxic than vinblastine is inferred from the slightly lower incidence of infertility following MOPP therapy than MVPP (mechlorethamine, vinblastine, procarbazine, and prednisone). Although studies of single-agent daunorubicin are not available, it appears to have minimal long-term effect when used in combination therapy not containing cyclophosphamide. When used with cyclophosphamide, however, daunorubicin appears to potentiate gonadal toxicity. Long-term administration of azathioprine does not seem to affect semen quality.

Most data has been derived from studies of combination chemotherapy and indicate permanent infertility among HD and some nonseminomatous testicular cancer survivors. Complicating the interpretation of these studies is the observation that before therapy as many as 30% of men with HD and 50% with germ cell tumors are oligospermic; disorders of sperm motility and morphology are more common.^31–33 Multivariate analysis found that elevated erythrocyte sedimentation rate and advanced stage are predictors of pretherapy infertility among HD.³⁴ Pretreatment FSH levels may provide a prognostic marker for subsequent spermatogenesis in young men with germ cell cancer.³⁵ MOPP or MOPP-like regimens to treat HD render all men infertile during therapy, and recovery is unlikely (Table 2). In a prospective study of 37 men receiving MVPP, 12 had low sperm counts before treatment, but all were azoospermic after two cycles and remained so for the first 12 posttreatment months.³⁸ Longer term follow-up finds only 5–15% ever regain spermatogenesis. Studies comparing MOPP chemotherapy to ABVD (doxorubicin, bleomycin, vinblastine, and dacarbazine) conclude that the latter combination produces less gonadal toxicity.^{22, 36} For patients with advanced HD, BEACOPP (bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, procarbazine, and prednisone) is increasingly employed. The German Hodgkin study group (GHSG) found that the incidence of azoospermia was significantly higher among those treated with eight cycles of BEACOPP or four cycles of COPP/ABVD compared to two cycles of COPP/ABVD (93%, 91%, and 56%, respectively).⁴⁵ An update of the Stanford V regimen (vinblastine, doxorubicin, vincristine, bleomycin, mustard, etoposide, and prednisone) reported 19 conceptions in 13 male survivors.⁴³

Abbreviations: ABVD, Adriamycin (doxorubicin), bleomycin, vinblastine, dacarbazine; ChIVPP, chlorambucil, vinblastine, prednisone, procarbazine; CMF, cyclophosphamide, methotrexate, 5-fluorouracil; COPP, cyclophosphamide, vincristine, prednisone, procarbazine; 5-FU, 5-fluorouracil; MACOP-B, methotrexate, doxorubicin (Adriamycin), cyclophosphamide, vincristine (Oncovin), prednisone, bleomycin; MOPP, mechlorethamine, Oncovin (vincristine), prednisone, procarbazine; MVPP, mechlorethamine, vinblastine, prednisone, procarbazine; NOVP, mitoxantrone, vinblastine, vincristine, prednisone; PACE BOM, doxorubicin, cyclophosphamide, etoposide, bleomycin, vincristine, methotrexate, prednisolone; PEB, cisplatin (Platinol), etoposide, bleomycin; L-PAM, L-phenylalanine mustard; PVB, Platinol (cisplatin), vinblastine, bleomycin; PVB + dox, cisplatin, vinblastine, bleomycin, doxorubicin; VAPEC-B, vincristine, doxorubicin, prednisone, etoposide, cyclophosphamide, bleomycin.

Data on outcomes after treatment for non-Hodgkin lymphoma (NHL) are less robust, but evidence suggests that the cyclophosphamide, vinblastine, and prednisone regimen is less toxic than MOPP.³⁰ A report of 14 men treated with vincristine, doxorubicin, prednisone, etoposide, cyclophosphamide, and bleomycin suggests that this may be an effective, relatively nontoxic regimen for NHL.⁴⁷ Leukemia therapy appears less toxic,⁷¹ although both allogeneic and autologous stem cell transplant (SCT) increase the likelihood of long-term infertility. Kreuser found 100% recovery during maintenance therapy among 10 patients aged 14–38 years treated with combination chemotherapy.^{39, 52}

The majority of men presenting with testicular tumors are oligospermic. In 41 patients studied prospectively, Drasga reported that 77% were oligo-azoospermic and 17% were azoospermic; only 6% had adequate sperm counts for cryopreservation.⁴⁸ Abnormalities of sperm motility are at least as prevalent. Following 2 months of therapy with cisplatin, vinblastine, and bleomycin, with or without doxorubicin, 94% of men in Drasga’s study were azoospermic.

Recovery of spermatogenesis following chemotherapy for testis cancer is common. Most studies show a time-dependent recovery of spermatogenesis, with nearly 50% of patients recovering some sperm production after 2 years (Table 2).^{50, 51} Longer follow-up has shown no adverse effects on fertility or sexual function.⁵¹ Recovery seems to be partly related to the cumulative dose of cisplatin. In those who receive over 400 mg/m², permanent infertility should be anticipated. Limited data suggest that ifosfamide may cause less irreversible infertility than its similarity to cyclophosphamide might predict.^{70, 72}

As in boys, Leydig cell function is more resistant and is usually well compensated; despite frequently elevated gonadotropin levels, few men require androgen replacement.^{24, 30} Subclinical Leydig dysfunction may have underrecognized sequelae, including excess cardiovascular morbidity increased hypercholesterolemia and obesity, in conjunction with lower testosterone and elevated LH/FSH.⁷³

Effect in prepubertal girls

The ovarian effects of chemotherapy in prepubertal girls are variable and depend on the drug, dose, and duration of therapy. Single-agent cyclophosphamide for nonmalignant disorders rarely causes either a delay in puberty or permanent sterility.⁷⁴ Most girls treated with procarbazine or nitrosoureas for brain tumors show biochemical evidence of primary ovarian dysfunction, but progress normally through puberty. Ovarian function returns to normal over a period of years, and elevated gonadotropin levels decrease to baseline in most women. Eighty percent of girls treated for ALL also proceed normally through puberty.⁷⁵ In a large study of survivors of childhood cancer, the likelihood of premature menopause was 13-fold higher when compared with siblings: 8% by age 40.⁷⁶ This increase was associated with higher doses of alkylating agents, ovarian radiation, and a diagnosis of HD. In Ewing sarcoma survivors, 67% developed POF at a median follow-up of 5.7 years. All who underwent pelvic irradiation had POF.⁷⁷

Histologically, however, prepubertal ovaries are significantly damaged by cancer chemotherapy. Follicular maturation arrest, stromal fibrosis, and a partially depleted ova population have all been reported following single-agent cyclophosphamide as well as cytosine arabinoside (ara-C)-based antileukemic therapy.

Effects in women

The effects of antineoplastic agents on the female gonadal function have been inferred from the incidence of amenorrhea, gonadotropin levels, and long-term fertility rates and outcomes. One autopsy series of acute leukemia patients showed no difference in the number of primary follicles, but secondary follicles were markedly depleted.⁵¹ Clinically, women receiving these agents develop POF: vaginal dryness with dyspareunia, endometrial hypoplasia, decreased libido, hot flashes, oligomenorrhea evolving into amenorrhea, and low serum estrogen levels with compensatory elevations of serum FSH and LH levels.^{33, 78}

The frequency of amenorrhea and infertility depends on the drug, its total dose, concomitant radiation, and the patient age when treated. Single-agent alkylating drugs are those most consistently associated with POF. Small series report that 50–75% of women treated with cyclophosphamide develop amenorrhea within a month of starting therapy, although there is a strong age-related susceptibility. In one study, the total dose of cyclophosphamide received before the onset of amenorrhea was 5.2 g for patients over 40, 9.3 g for those 30–39, and 20.4 g for those 20–29. Menses returned in 50% of women under 40.⁶³ Return of menstrual function was correlated with the dose administered after the cessation of menses.

Old studies of adjuvant chemotherapy with L-phenylalanine mustard for breast cancer showed significant, age-related POF: 73% of the women aged 40–49, but only 22% of the women under 39, developed amenorrhea during therapy.⁵⁹ Single-agent treatment with busulfan or chlorambucil is associated with well-documented age and dose-related ovarian toxicity.^{79, 80}

Sarcoma patients treated with high-dose methotrexate rarely report amenorrhea and serum gonadotropin levels remain normal during and after therapy.⁵⁴ Lower dose methotrexate for gestational trophoblastic tumors appears to exert no significant toxicity, although one survey from England found that menopause occurred on average 3 years earlier in chemotherapy-treated women.⁸¹ Fluorouracil, daunorubicin, and bleomycin as single agents are also well tolerated.

Few data are available for etoposide, but ovarian dysfunction has been reported among women receiving the drug for gestational tumors.⁸²

Tamoxifen appears to exert a mild estrogenic effect associated with decreased gonadotropin levels in both premenopausal and postmenopausal women treated for breast cancer. Menstrual irregularities are common, but the incidence of persistent amenorrhea is unclear. Successful pregnancies have been reported among women on chronic interferon therapy for a variety of malignant and nonmalignant diseases.⁶⁵

Most reported outcomes come from studies of multiagent therapy. The incidence of amenorrhea in women treated with MOPP, MVPP, or COPP ranges from 15% to 80% (Table 2).^{29, 35, 44, 83–85} Two-thirds develop amenorrhea during therapy. A dose–response relationship is unclear. In one study, there appeared to be no difference between three and six cycles of MOPP.⁸⁶ Age at the time of treatment, however, is an important variable affecting the incidence and onset of permanent amenorrhea. Sixty to hundred percent of patients over age 25 develop permanent amenorrhea during therapy. POF occurs with initiation of therapy in 5–30% of women under 25 and in an additional percentage over the next months. Younger women experience at least a 50% likelihood of POF 5–10 years of therapy.⁸⁷ Preliminary reports suggest that HD regimens such as ABVD or doxorubicin, cyclophosphamide, etoposide, bleomycin, vincristine, methotrexate, and prednisolone may have lower rates of prolonged amenorrhea.^{38, 56} Horning has reported 24 conceptions among 19 women treated with the Stanford V regimen V.⁴³ Women receiving methotrexate, doxorubicin cyclophosphamide, vincristine, prednisone, and bleomycin for aggressive lymphomas appear, in small series, to maintain fertility.⁴¹ Women treated with four cycles of Mega-CHOP for NHL had recovery of ovarian function.⁸⁸ Eight of these patients conceived spontaneously. In a second study of women younger than age 40 receiving CHOP therapy for NHL, only 2 of 36 women developed POF. ⁸⁹ Fifty percent of these women conceived in first remission of their disease.

Women receiving cisplatin-containing therapy for germ cell tumors typically become amenorrheic during treatment, but over 90% resume menstruation within a few months of completing treatment.^{57, 90–92} Among women with breast cancer, who may already have age-related decreased reproductive potential, 80% receiving adjuvant therapy with cyclophosphamide, methotrexate, and 5-fluorouracil (CMF) become menopausal within 10 months of beginning therapy.^{61, 62} The reported ranges, however, vary between 20% and 100%.⁹³ Those given doxorubicin and cyclophosphamide usually become anovulatory within 3 months; sooner if they are perimenopausal. Few women under age 30 treated with doxorubicin-containing regimens experienced amenorrhea; about one-third of women aged 30–40 years, and nearly all those over 40 years of age do so. Epirubicin-containing regimens are similar. Adjuvant docetaxel with doxorubicin and cyclophosphamide resulted in a 61% rate of amenorrhea in a patient population whose average age was 49.⁹⁴

Fertility following high-dose chemotherapy

Case reports and small series have documented the recovery of fertility in very few patients treated with high-dose regimens. A retrospective survey of over 37,000 SCT patients found that only 0.6% subsequently conceived.⁹⁵ Follow-up of 187 young women previously treated with SCT for either aplastic anemia or leukemia found the anticipated age-dependent effect of cyclophosphamide on ovarian function.⁵³ Fifty-seven percent of female patients had fertility impairment at 2.3 years following SCT. Only 7% of 144 patients recovered fertility, and recovery took longer in women than in men. A longitudinal study of 217 childhood allogeneic SCT patients found that 56% of men had fertility impairment with elevated FSH/LH and low testosterone after a median of 2.6 years. Azoospermia was documented in 14 of 15. Only 3% recovered fertility 3.4 years after SCT.⁹⁶ Patients transplanted for aplastic anemia overall do somewhat better, particularly those under age 25.

Effects of radiation therapy on gonadal function

Effects in men

Therapeutic irradiation in boys with acute leukemia was a standard practice to prevent testicular relapse. Doses below 1200 cGy were insufficient to control disease, and early protocols were replaced by those delivering 2400 cGy to both testes. This dose induced permanent Leydig cell damage, puberty was delayed, testosterone levels were diminished, and gonadotropin levels were increased in most patients.⁹⁷

In adults, single 400–600 cGy doses of testicular radiation may produce azoospermia for 5 years or longer.¹ Berthelsen evaluated men undergoing prophylactic radiotherapy for seminoma and found that two-thirds became azoospermic from scatter dose of 201–130 cGy.⁹⁸ Shapiro has documented oligospermia/azoospermia lasting up to 24 months after as little as 27 cGy.⁹⁹ Adult Leydig cell dysfunction, with elevated LH values, occurs at radiation doses greater than 2000–3000 cGy, and can require hormone replacement. Fractionated radiation appears to produce tubular damage equivalent to that seen with single doses. Recovery occurs in the majority of those who are treated for germ cell tumors; series report 37–66% success rates in men wishing to conceive a pregnancy.¹⁰⁰

TBI for SCT conditioning is routinely associated with permanent azoospermia. Secondary infertility has been reported in association with radiation administered to the hypothalamus or pituitary in conjunction with chemotherapy for intracranial neoplasms.¹⁰¹

Whether there are permanent effects on the surviving germ cells of men receiving radiation remains uncertain. Most studies have been unable to document an increase in malignancies in offspring.¹⁰²

Effects in women

The radiation sensitivity of the human ovary has not been well defined. Small primordial oocytes are considerably more sensitive than large follicles, and ovarian sensitivity to radiation is dose and age dependent. In adult women, single doses of 500 cGy produce menstrual irregularities in women of all ages. For women over 40, 600 cGy reliably induces menopause. Women aged 20–30 can tolerate up to 3000 cGy if fractionated over 6 weeks.¹⁰³ Uterine radiation in childhood increases risk for nulliparity, spontaneous abortions, and intrauterine growth retardation, so fertility is not assured even if ovarian function is preserved. In a retrospective study of 162 patients with stage II/III colorectal cancer, the incidence of amenorrhea in patients was higher among women with rectal cancer compared to colon cancer (94% vs 4.2% respectively), possibly related to pelvic irradiation in rectal cancer patients.¹⁰⁴

Effects of targeted therapies on gonadal function

In the era of molecularly targeted therapy, several new drug classes have entered clinical use, but there is a paucity of data regarding their effects on fertility. Furthermore, cytotoxic agents are often used concurrently, and the effect of combination therapy on gonadal function has not been studied.

Effect of TKIs

Imatinib inhibits c-KIT, crucial for Leydig cell function, and platelet-derived growth factor receptor (PDGFR) essential for gonocyte migration. c-KIT and PDGFR are also expressed in oocytes and have important roles in fulliculogenesis and female fertility. Oligozoospermia, gynecomastia, and testicular failure have been described in prepubertal boys.¹⁰⁵ Case reports document oligospermia in men, but most men father normal offspring.¹⁰⁶ Women treated with imatinib appear to conceive normally but may have an increased risk of congenital malformations, hence discontinuation of therapy before pregnancy is recommended.¹⁰⁷

Rapid onset of hypogonadism and low testosterone has been described in men treated with crizotinib.¹⁰⁸ A follow-up study documented low testosterone levels with reductions in sex-hormone binding globulin, LH, FSH, and free testosterone, suggesting central hypogonadism.¹⁰⁹ Dasatinib and sunitinib have been associated with gynecomastia, but their effects on fertility are unknown.

Effects of M-TOR inhibitors

Sirolimus and everolimus also inhibit c-KIT. The reproductive side effects of m-TOR inhibitors are inferred from organ-transplant patients. Men treated with sirolimus and everolimus have low testosterone and oligospermia, with elevated FSH and LH; effects are reversible after discontinuation of therapy.¹¹⁰ In human ovarian cortical strips treated in vitro with rapamycin, oocyte destruction by granulosa cells was discovered, although previous reports had implicated m-TOR inhibitor-mediated apoptosis as the mechanism for oocyte loss.¹¹¹

Effects of monoclonal antibodies and immunotherapy

Alemtuzumab, a monoclonal antibody against CD52, causes immobilization and agglutination of sperm because CD52 is also expressed on the surface of the sperm.¹¹² Cytotoxic T-lymphocyte-associated antigen 4 (CTLA-4) blockade with ipilimumab and programmed cell death-1 (PD-1) receptor agents for the treatment of melanoma can cause hypophysitis and hypopituitarism. In the sentinel trial of ipilimumab in melanoma, the incidence of serious hypopituitarism was low at 1.8%.¹¹³ The incidence of reproductive side effects with CTLA-4 and PD-1 blockade is not known.

Protective measures

Protection for men

It has long been speculated that halting spermatogenesis through hormonal manipulation might ameliorate testicular damage, as most chemotherapy agents are selectively toxic to dividing cells. In clinical trials of men receiving chemotherapy for HD, two attempts using GnRH analogs have been unsuccessful.^{114, 115} For men who desire fertility following combination chemotherapy for advanced HD, ABVD is clearly preferable to MOPP.³⁸ Use of GnRH analogs does not shorten the recovery time to spermatogenesis. Masala et al. documented some protection with the use of testosterone in patients treated with cyclophosphamide. ASCO Guidelines indicate this procedure to be experimental.

Men anticipating cancer therapy should consider semen cryopreservation. The process of sperm cryopreservation is no more detrimental to sperm quality in cancer patients than in noncancer controls.¹¹⁶ Because of the high prevalence of abnormal pretherapy semen analyses, many patients have been considered poor candidates, but successful impregnation has been achieved following artificial insemination using semen with quite low sperm counts and poor sperm motility.¹¹⁷ With meticulous reproductive assistance, fertilization may be effective in up to 45% of cases. In addition, in vitro fertilization and subsequent implantation has been successful in cases of even lower sperm counts and motility.¹¹⁸ With the advent of intracytoplasmic sperm injection (ICSI), where only a single live spermatozoon is selected and injected into an oocyte, the chance of conception can be greatly enhanced despite exceedingly low sperm counts. Garcia compared ICSI in patients with malignancies with noncancer patients who underwent the same assisted reproductive procedure. The cumulative pregnancy and cumulative live birth rates in cancer patients were 69% and 62%, respectively, comparable to noncancer patients.¹¹⁹ Fertilization has been reported using spermatozoa retrieved by testicular biopsy and with sperm extracted from the vas deferens at the time of orchiectomy in several azoospermic testis cancer survivors.¹²⁰ Many centers nonetheless report that the overall success rate among men who elect to preserve semen may be somewhat limited, and perhaps is influenced by factors other than semen quality. One series from Memorial Sloan-Kettering Cancer Center reported locating 48 of 69 men who had banked sperm, but at a median of 27 months posttreatment only 11 had attempted to use their sperm for artificial insemination. Of these, only three achieved successful pregnancies.^{117, 121}

Penile vibratory stimulation and electro-ejaculation may provide an option for pubertal boys.¹²² Very preliminary studies of testicular circulatory isolation suggest that this mechanical procedure is protective in a rat model and might be feasible in human clinical trials.¹²³

Testis sperm extraction is reported to recover spermatozoa in 55–85% of men with nonobstructive azoospermia of various etiologies. Damani reported a series of men azoospermic after chemotherapy. Spermatozoa were found in 65%. A total of 26 ICSI cycles were performed in 12 couples. Fertilization rate was 65%. All babies born to date have had normal neonatal examinations.¹²⁴

Testicular tissue biopsy and cryopreservation is experimental, as is isolation of germ cells from testicular tissue for storage. Gonadal shielding remains the mainstay of protection from therapeutic radiation. No convincing studies to suggest benefit from scrotal cooling are currently available.

Protection for women

Reversible suppression of ovarian function by oral contraceptives or GnRH analog might offer gonadal protection to cycling women about to undergo potentially sterilizing radiation or chemotherapy. Some animal models have validated the hypothesis, both in protection from radiation and in chemotherapy-induced ovarian damage.¹²⁵ Results of such trials in humans have been conflicting, however, and the number of patients studied remains small. Most promising of these reports is that of Chapman and Sutcliffe, who administered oral contraceptives to women anticipating receiving MVPP therapy for HD. Five of six women had resumption of normal menses at a mean follow-up of 26 months.¹²⁶

Pacheco administered leuprolide acetate to 12 patients aged 15–20 anticipating treatment for lymphoma.¹²⁷ Suppression was continued monthly with depot administration until 1 month after cessation of chemotherapy. All 12 treated patients resumed normal menstrual cycling within 6 months, compared to none of the four women in this study who did not receive leuprolide suppression. Three pregnancies were reported. In a meta-analysis of six randomized controlled trials (RCTs) that examined reproductive outcomes among women with HD and ovarian and breast cancer, administration of GnRH analogs with chemotherapy was associated with a statistically significant increase in resumption of spontaneous menstruation and ovulation (OR 3.46 and 5.70, respectively).¹²⁸ Subsequent meta-analyses of RCTs exclusively among patients with breast cancer have yielded conflicting results regarding prevention of POF and spontaneous resumption of menses with GnRH agonists.^129–131 In a large trial evaluating the effect of adjuvant chemotherapy in premenopausal women with node positive breast cancer (NSABP-B-30 trial), those with prolonged amenorrhea had better disease-free survival and overall survival.¹³² Thus, ovarian suppression in breast cancer patients for fertility preservation is controversial.

The highest live birth rates are reported following ovarian stimulation and embryo cryopreservation, and are therefore the best current option for women in stable relationships. Superovulation is required before fertilization, and neither the time required nor the hormone manipulation necessary may be acceptable. The implantation rates range from 8% to 30%. The cumulative pregnancy rate can be more than 60%.¹³³

Cryopreservation of ovarian tissue remains experimental but could be considered in women who require urgent cancer treatment. Ovarian cortical tissue also has the advantage of containing large numbers of follicles and thereby increasing the potential for successful future pregnancies. The ovarian tissue can ultimately be transplanted back in to the patient, thus restoring ongoing fertility. Transplanted tissue has been reported to restore normal menstrual cycling¹³⁴ and live births have been reported following orthotopic transplantation of cortical strips to the remaining ovary. In 114 Belgian women treated for various malignancies who underwent cryopreservation, a total of 49 spontaneous pregnancies in 33 patients and two induced pregnancies were reported after a mean follow-up of 50 months.¹³⁵ This technique may ultimately offer an option for young girls as well as sexually mature young women.

Recently, the Edinburgh selection criteria for identifying patients at risk for POF after gonadotoxic therapy was validated and may be useful in selecting patents for intervention.¹³⁶ Unresolved medical and ethical issues have fueled debate regarding these techniques.¹³⁷

For patients undergoing pelvic radiotherapy, oophoropexy can be considered. At the time of exploratory or staging laparotomy, the ovaries are moved either medially behind the uterine fundus or laterally out of the radiation port. Radiation exposure is decreased 90%, and hormonal function is preserved in 55–95% of patients; however, fertility is still compromised, possibly owing to the abnormal tubo-ovarian anatomy or radiation scatter. A small trial of laparoscopic propriosacral ovariopexy offers the possibility of a less invasive approach to fertility preservation.¹³⁸ One option in this setting is the transposition of only one ovary and removal of the second for cryopreservation.¹³⁹ Specific ovarian shielding may be useful in some cases.

Protection for children

No proven methods for protection of future fertility in children are available at this time. Some centers offer ovarian tissue or testicular tissue cryopreservation as an experimental approach, but research in this area is fraught with ethical problems. Many excellent reviews of the technical and ethical issues of fertility preservation are available (Table 3).^{127, 135}

Abbreviations: GnRH analog, gonadotropin-releasing hormone analog; OCP, oral contraceptives.

Outcomes of pregnancy

Chemotherapy

Case reports document successful conception and delivery of normal infants to patients who have received even the most aggressive of chemotherapy regimens; neither male nor female permanent infertility can be presumed following chemotherapy for cancer. Several retrospective series have evaluated the outcome of pregnancy in women treated with chemotherapeutic agents as children or young adults who completed therapy and then became pregnant.

One large study evaluating offspring of children treated for a variety of cancers found that in a total of 286 subsequent pregnancies there was no increase in congenital anomalies, and chromosomal analysis was normal in 23 of 24 children tested.¹⁴⁰ Pregnancies in women previously treated for trophoblastic tumors also appear to have no associated increased risk of congenital anomalies, spontaneous abortions, or neonatal mortality.¹⁴¹

Holmes evaluated women treated for HD and compared the 93 pregnancies in their chemotherapy-treated patients to 288 sibling-control pregnancies. Overall, there was no difference between the groups, although when the subgroup that received both radiation therapy and chemotherapy was analyzed separately, it appeared that combined treatment produced more spontaneous abortions in wives of male patients, and those female patients were slightly more likely to produce abnormal offspring than control women.¹⁴²

Offspring of fathers treated with prior chemotherapy likewise appear to be normal. When large series are combined, nearly 1400 live-born children have been reported to have a congenital defect incidence of about 4%, not significantly different from the general population. Most of these anomalies represent common, nongenetic abnormalities.¹⁴²

Further follow-up suggests that offspring growth, development, and school performance are probably normal. A National Cancer Institute study to address the question of cancer in offspring of treated patients found a slight and statistically insignificant excess of cancers in these children when compared to offspring of sibling-matched controls (0.3% vs 0.23%), numbers not different from those expected in the general population. When analyzed by age and sex, however, it appeared that there was an excess of cancers diagnosed in male offspring under age 5.¹⁴³

Risk to the fetus exposed in utero to chemotherapy agents depends on gestational age and the drug and dose administered. Folate antagonists should not be administered during the first trimester. Other antimetabolites have rarely been associated with congenital abnormalities. First trimester exposure to 5-fluorouracil, cyclophosphamide, busulfan, and chlorambucil has been associated with low birth weight in infants and other abnormalities on rare occasion.¹⁴⁴ Fetal myocardial necrosis has been reported following maternal administration of anthracyclines.¹⁴⁵ Imatinib has demonstrated teratogenicity in animal models, but case reports have documented successful pregnancies in women who conceive during treatment.¹⁴⁶ Rituximab given unintentionally to a pregnant woman resulted in an uncomplicated pregnancy delivering an apparently healthy infant. Interferon-a has been given safely during pregnancy in a small number of women.¹⁴⁷

Whether the risk to the fetus is further increased with drug combinations is uncertain. Case reports and small series indicate that exposure in the second and third trimesters is associated with minimal risk to the fetus and that long-term development of these offspring is normal.^{148, 149} Nonteratogenic effects including low birth weight, intrauterine growth retardation, and more subtle developmental abnormalities remain to be defined. In utero exposure to diethylstilbestrol has been linked to the development of genital clear cell carcinomas in the female offspring of these women, but other clear documentation of carcinogenesis from in utero exposure to chemotherapy is lacking. No information is available on the reproductive potential of these children.

Radiation therapy

Most of what is known about the genetic effects of radiation therapy is inferred from data on survivors of atomic bomb exposure. The increase in untoward outcomes of pregnancies (major congenital defects, stillbirth, and death during the first week of life) is small, estimated at 0.00182/gonadal rem (roentgen-equivalent–man)—the quantity of any ionizing radiation equivalent to the biologic effect of 1 rad (cGy).

Among women treated with radiation therapy below the diaphragm, preterm delivery in up to 20% of pregnancies and an excess of low birth-weight infants have been reported. That these adverse outcomes are often clustered in the first posttreatment year suggests they may result from local uterine or hormonal factors and may not be because of genetic defects.^{46, 150}

In utero exposure to irradiation produces the greatest risk of teratogenesis during the period of organogenesis from the second to the eighth week, with growth retardation, eye problems, and microcephaly appearing as the predominant abnormalities. A safe dose has not yet been defined, but generally, a therapeutic abortion is recommended for any uterine dose of 10 cGy during the first trimester. Supradiaphragmatic irradiation is associated with considerable scatter to the fetus, much of which can probably be prevented with abdominal shielding. Local irradiation of the neck and axilla may be safe during the first trimester.¹⁴⁴

Psychosocial issues

The psychosocial issues of disfigurement, loss of fertility, anxiety about birth defects, sexual performance, and recurrence of tumor all have important impacts on the unmarried patient facing dating and mate-selection issues, as well as on married patients in a stable relationship, for whom the separation rate may be fourfold of that of the general population.^{46, 149, 150} Detailed discussion of these important issues is beyond the scope of this publication, but excellent reviews are available for the interested reader seeking further information.^{42, 149, 151–155}

For physician assessment of sexual functioning, Andersen has proposed a model that helps the provider address many issues before they arise.⁴²

Summary

• Gonadal dysfunction is related to drug-dose and is age-related
• Younger patients are more likely to preserve fertility
• Alkylating drug therapy causes damage in all age groups
• Children of both genders enter puberty normally after chemotherapy
• Traditional endocrine measures of gonadal toxicity are insensitive
• Alkylating drugs appear to be the most toxic; procarbazine particularly
• Many cytotoxic drugs are not well-studied
• ABVD for HD should replace MOPP if possible for young patients with HD
• Perimenopausal women with breast cancer are likely to become menopausal with chemotherapy
• New agents have theoretical risks but are not well-studied
• Risks to fertility should be part of chemotherapy consent discussion
• Proven options to preserve fertility in men are limited to semen cryopreservation
• Many men with HD or testicular tumors are oligospermic at presentation
• Oligospermia should not prevent cryopreservation
• New and effective options to preserve fertility in women are under development
• Fetal exposure to most chemotherapy agents is safe after the first trimester
• Outcomes of pregnancy during chemotherapy are good
• Outcomes of pregnancy after childhood chemotherapy are normal
• Early fetal exposure to folate antagonists must be avoided

---

Buy Membership for Hematology, Oncology and Palliative Medicine Category to continue reading. Learn more here

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

Herpesviruses Epigenetic contributions to human cancer MicroRNA expression in cancer Cardinal manifestations of cancer Endocrine complications and paraneoplastic syndromes Cancer genomics and evolution

Holland-Frei Cancer Medicine