Fertility preservation, especially in women of reproductive age undergoing gonadotoxic treatment, has become an important part of our practice. This is reflected in the increasing number of publications on the subject. A recent PubMed search yields 518 articles on fertility preservation published in the time period of 1980–2000 and 4,288 others between 2000 and 2017. Today, failure to discuss fertility preservation with young women scheduled to undergo radiotherapy or chemotherapy could be considered as malpractice.
Although cancer usually affects aged population, invasive gynecological tumors can occur during reproductive age. Long-term survival rates after gynecological cancer are improving and preservation of fertility has logically become a major issue.
In gynecologic oncologic surgery, there has been gradual development of fertility-saving surgery (FSS) with the aim of preserving the reproductive organs. Survival should not be compromised and thus indications are restricted to patients of a young age with a desire to preserve fertility and presenting with a well-differentiated cervical, ovarian, or endometrial low-grade tumor in its early stages or with low malignant potential.
In this chapter, we discuss indications for FSS in women with gynecological cancer, according to oncological and reproductive outcomes.
Cervical cancer is the fourth most common cancer among women worldwide, with many of them still in their reproductive age . In 2015, 12,900 patients had newly diagnosed cervical cancer in the United States . Of those, 39% in women were under the age of 45.
Standard surgery in early-stage disease is usually based on radical hysterectomy (RH) with paracervix and pelvic lymphadenectomy depriving women of their potential of pregnancy. FSS, by achieving equivalent oncological outcomes to hysterectomy, can only be offered to patients with early-stage disease, negative nodes, and nonaggressive histologic patterns .
In cases of FIGO stage IA1, cervical carcinoma can be treated with simple conization . Compared with this approach, hysterectomy does not offer better survival or recurrence rates if there is no sign of lymphovascular space invasion (LVSI) and negative margins are confirmed [4, 5]. This approach can be applied in micro-invasive squamous cell carcinoma as well as adenocarcinoma, with similar outcomes . However, FSS should not be recommended for rare histological subtypes of cervical cancer including neuroendocrine carcinomas and non-HPV-related adenocarcinomas (except for adenoid basal carcinoma), which exhibit aggressive behavior .
LVSI does not change FIGO staging but should be separately reported because it may affect treatment strategies. Guidelines for fertility preservation in stage IA1 patients with LVSI recommend radical trachelectomy (RT) with pelvic lymphadenectomy (National Comprehensive Cancer Network [NCCN] guidelines), and conization with pelvic lymphadenectomy as the alternative , while European Society for Medical Oncology (ESMO) guidelines recommend conization with pelvic lymphadenectomy . Indeed, stage IA1 patients who have LVSI but negative lymph node metastasis did not reported parametrial involvement . Conization or simple trachelectomy can then be a reasonable FSS for IA1 with LVSI in case of negative lymph node metastasis. However, due to the small number of nonrandomized studies, this specific issue should be addressed in further larger trials .
In patients affected by cervical cancer at FIGO stages IA2–IB1<2 cm who desire further pregnancy, RT with pelvic lymphadenectomy is the FSS of choice . Vaginal RT was described by Dargent already back in 1994  by using a laparoscopic approach to remove pelvic lymph nodes and a vaginal approach to remove the upper part of the vagina, the cervix, and proximal part of the vagina. Meta-analysis showed that there was no significant difference between the RT and RH in recurrence rate, five-year recurrence-free survival rate, five-year overall survival rate, postoperative mortality, intraoperative complications, postoperative complications, blood transfusion, and number of lymph nodes .
RT and RH offer similar oncologic outcomes: long-term survival rate of 98.4% and a recurrence rate of only 4.5% have been reported [14–16]. In a recent systematic review, Bentivegna et al. reported 4.2% recurrence rate after FSS for stage 1B cervical cancer (1.8% after conization, 3.8% vaginal RT, 3.8% after abdominal RT, 4.7% after minimally invasive RT, and 4.3% after NACT followed by conservative surgery) . The author did not find recurrence after resection of tumor 2 cm or smaller and without LVSI .
Similar outcomes can be achieved after FSS for adenocarcinoma . Neuroendocrine and anaplastic lesions are considered as contraindication for FSS because of the potential need for adjuvant therapy and the risk of recurrence even in small tumor .
Among FSS for FIGO stage 1 cervical cancer, six procedures are available: simple trachelectomy or conization, vaginal RT, abdominal or laparoscopic RT, and neoadjuvant chemotherapy (NACT) followed by conization/trachelectomy/RT [3, 15].
To determine whether fertility preservation could be offered, pelvic-node status and tumor size (≤2 cm vs >2 cm) are two main factors to determine the best fertility-sparing surgical technique .
Pelvic Node Status
In case of intraoperatively proven lymph node involvement, FSS should be abandoned, and the patient referred to adjuvant chemoradiation . Fertility cannot be safely preserved in this condition. Non-invasive nuclear methods are under investigation to assess the presence of metastases in early-stage cervical cancer  and thus improving patient selection for FSS in the future.
Pretrachelectomy magnetic resonance imaging (MRI) can help identify high-risk patients likely to need radical hysterectomy or confirm the absence of residual tumor in the cervix after a cone biopsy with negative margins. In experienced hands, MRI can show tumor size, stromal invasion, distance between the tumor, and the cervical internal os .
Indeed, after trachelectomy, over 60% of tissue samples have demonstrated absence of residual tumor after initial conization . Many women are then effectively being overtreated and that less radical surgery may be sufficient for oncological safety in selected patients. Therefore, conization in combination with laparoscopic lymphadenectomy can be considered as an appropriate procedure in patients presenting with early-stage and tumors <20 mm. However, it should be noticed that free margins of 8–10 mm are required when considering simple conization .
Depth of stromal invasion and LVSI are two prognostic factors for recurrence in early-stage disease .
In patients with stage IB1 disease >2 cm, vaginal RT should not be performed given the high risk of recurrence in such cases (17% vs. 4% in stage IB1 <2 cm) .
In case of stage 1B1 2–4 cm, two main strategies were proposed in the literature: NACT followed by RT [24, 25] or abdominal RT [26, 27]. Patients eligible for NACT should logically undergo first lymphadenectomy, but this approach is described in few series [28, 29]. However, even if necessitating two surgical procedure, lymphadenectomy before NACT avoid misdiagnosed small-sized nodal metastasis that might be sterilized after NACT when lymphadenectomy is performed during RT after NACT. However, oncologic outcomes are not different when compared with immediate RT without chemotherapy  but the risk of grade R3 postoperative morbidities is higher in patients undergoing abdominal RT (particularly with a laparotomic approach) .
However, because of few reported cases and no long-term follow-up outcomes, Fertility-sparing treatment in patients with tumours greater than 2 cm cannot be recommended and is considered as an experimental approach .
Bentivegna et al. identified 50 patients with stage IB2 and 17 with stage IIA (mainly IIA1) tumours, of whom 28 received NACT and 23 underwent abdominal RT . Because it is not possible to draw up definitive recommendations for conservative treatment in such cases, most teams consider such conditions to be absolute contraindications to conservative surgery.
Bentivegna et al recently published a systematic review on 2,777 patients submitted to FSS for cervical cancer stage IB and 944 ensuing pregnancies  The overall fertility, live birth, and prematurity rates after these procedures were, respectively, 55%, 70%, and 38%. The pregnancy rate was higher in patients submitted to a vaginal or minimally invasive RT compared with a laparotomic RT. The live birth rate was similar, whatever the FSS procedure.
The authors suggested that the lower pregnancy rate observed after laparotomic RT could potentially be due to adhesions related to the approach itself and/or to a higher frequency of septic morbidities (abcesses, local peritonitis), or potentially due to uterine artery ligation. However, uterine artery preservation or ligation had no impact on the fertility rate in this review. It seems that preservation of the ovarian vascularization in these young patients could allow them to achieve a pregnancy.
Although the fertility rates grow up to 55% , some patients become infertile with some of them requiring fertility treatment . The lack of cervical mucus and cervical stenosis could explain infertility . The most common after FSS is cervical stenosis and could be related to the surgical resection itself, the use of a cerclage, and the use or not of antistenosis tools .
According to the recent review of Li et al , the incidence rates of cervical stenosis ranged from 0% to 73.3% with an average rate of 10.5% after RT. The use of antistenosis tools such as catheter, intrauterine devices and Smit sleeves decreases the risk of stenosis from 12.7% to 4.6% .
In selected cases where radiotherapy or chemoradiation are mandatory, ovarian transposition allow ovarian removal from the radiation field with an efficacy d to be about 50% . Cryopreservation of oocytes or embryos before cancer treatment must be proposed in such cases  keeping in mind that irradiation of the uterus may cause irreversible damage.
RT is associated with severe obstetrical consequences , including second trimester miscarriage, preterm premature rupture of membranes (PPROM) and preterm delivery due to reduced mechanical support from a shortened cervix and ascending infection. Kasuga et al. reported that a mid-trimester residual cervical length of <13 mm is a good predictor of preterm delivery <34 weeks in women who have undergone abdominal RT . Moreover, the incidence of PROM and premature delivery is significantly higher, respectively 36.8% and 66.7%, when the residual cervix is less than 10 mm on post-operative MRI after vaginal RT .
The prematurity rate ranged from 39% and 57% according to the FSS used with significantly higher rates after RT by laparotomy but without any difference on live birth rate. Lower rate of premature delivery (and fetal loss) are observed following a conization or simple trachelectomy, because the impact on the length of the remaining cervix is reduced after less extensive surgery .
According to Bentivegna et al, several procedures or precautions aimed at decreasing this risk were reported . According to Dargent, the use of a prophylactic cerclage after vaginal RT reduced the risk of fetal loss from 50 to 22%  but this procedure is not systematic in other series mainly due to the risk of consecutive abcess or deep infection  and the absence of level A evidence that this procedure has an impact on prevention of fetal loss and prematurity. Other teams developed protocols with avoidance of digital examination, vaginal check of pH to detect ascending infection and adequate treatment of infection associated with closure of the cervical os .
Endometrial cancer is the fifth most common cancer in women (4.8% of cancers in women) . While most of endometrial cancer will occur in the post-menopausal period, 14% of cases are diagnosed in premenopausal women, with 5% of them before 40 years old . Patients at higher risk of presenting endometrial carcinoma are women exposed to endogenous and exogenous oestrogens associated with obesity, diabetes, early menarche, nulliparity, late-onset menopause, age ≥55, and use of tamoxifen .
FSS in Endometrial Cancer
Due to a higher proportion of well-differentiated tumor and limited myometrial invasion before the age of 45, patients have a better prognosis than older patients .
The standard treatment of endometrial cancer involves hysterectomy and bilateral salpingo-oophorectomy, due to the hormonal sensitivity of endometrial tumours, the risk of ovarian involvement lymph node evaluation and evaluation of extrauterine disease . In case of well-differentiated (grade I) endometrioid adenocarcinoma with disease limited to the endometrium, the NCCN guideline consider that conservative treatment can be offered to women who wish to maintain fertility [46, 47]. Very few patients with FIGO stage IA grade 2–3 with superficial invasion showed promising results (87.5% of complete response with 71.4% of recurrence rate after 48 months)  but these results need to be confirmed on larger series before being implemented in current practice.
Conservative management of patient with early endometrial cancer involves high dose progesterone either orally (medroxyprogesterone acetate or megestrol acetate) or delivered by an intrauterine device (levonorgestrel-releasing IUD). In case of nonobese patient, levonorgestrel-releasing IUD offers higher regression rate than medroxyprogesterone acetate but similar regression rate are observed in obese patients . After hormonal therapy, 77.7% of the patients showed a response with 48.2% complete response rate in women with carcinoma . However, in this systematic review, 35.4% recurrence rate was noted after initial treatment. In another prospective series giving 26 consecutive weeks of medroxyprogesterone acetate with aspirin, a complete response rate of 55% was achieved with 47% of patients presenting recurrence between 7 and 36 months . In cases of recurrence, a second cycle of progesterone treatment has been associated with response rates of up to 89% . Young patients with low-grade endometrial cancer appear to have excellent survival, regardless of the primary therapy chosen (hormonal treatment vs primary surgery) . Weight change during progestin therapy has little influence on complete response, recurrence, pregnancy, and live birth rates . However, pre and posttreatment Body Mass Index (BMI) of ≥25 kg/m2 were significant predictors for poor treatment response and high recurrence. Maintaining patients’ normal BMIs during progestin therapy
Hysteroscopic resection of localized carcinoma has been described and followed by hormonal therapy. In a review of the literature, including 4 studies and 36 patients treated with same procedure, the complete response rate for all patients was 88.9% with tumor recurrence in only 4 patients (11.1%) . Jadoul and Donnez reported 66.6% pregnancy rate after conservative treatment for grade I adenocarcinoma . Treatment consisted in hysteroscopic resection followed by 3 to 6 months of Gonadotrophin releasing hormone agonist and IVF attempt.
After hormonal treatment, between 24.7% and 46.9% of patients became pregnant [50, 57]. Pregnancy rate that ranged from 25% to 100% were reported after hysteroscopic approach but it should be noted that this was based on a total of 9 patients . Obesity and polycystic ovarian syndrome may also have a negative impact on pregnancy rates in young patients with endometrial cancer. The pregnancy rates are much lower in obese patients when compared with nonobese patients (13% in obese vs 48% in nonobese patients) .
As pregnancy should be achieved within the shortest period due to the risk of recurrence, assisted reproductive treatments may reduce time to conception. The use of fertility drugs was not associated with a higher incidence of cancer recurrence after successful fertility-sparing management . Levonorgestrel IUD in situ has also been found to minimize the effect of estrogenic stimulation on the endometrium . Moreover, disease-free survival was greater among patients who had achieved at least 1 pregnancy compared with those who did not . After completion of childbirth, patients should ideally undergo hysterectomy and bilateral salpingo-oophorectomy because of the persistent recurrence risk .
Ovarian cancer is the gynecological tumor with the highest mortality rate and constitutes the fifth most common cause of death from cancer in women . Although ovarian cancer is predominantly a disease of post-menopausal women, it is estimated that approximately 12% of cases occur in women with active childbearing potential .
In case of early ovary cancer, FSS preserves uterus and the contralateral ovary after an adequate staging including at least the examination of the entire abdominal cavity with histopathological analysis of peritoneal washing. Ovarian cancer type, stage/grade of the tumor are considered when selection of patients to be candidate for conservative treatment. Borderline Ovarian Tumors (BOT), epithelial ovarian carcinomas (EOC), malignant germ cell tumors and sex cord tumors are candidates to fertility-sparing surgery when limited to one ovary.
BOTs represent 10%–20% of all ovarian epithelial tumors . Survival rates are about 97–99%, with 70-month disease-free survival in cases of stage I tumours. The survival rate in cases of stage III tumors is about 89–90% [64, 65]. Conservative approach for BOT were proposed by Morice, Donnez and Zanetta since early 2000 [66–68]. Conservative surgery can then be performed in most BOT patients except in clear cell cancer due to a higher risk of recurrence compared to other BOT types . Recurrence rate are higher after ovarian cystectomy (10%–20%) than after adnexectomy (5%) . However, in case of recurrence on the spared ovary, a second FSS can be considered if the patient desires to preserve fertility. In a review including 5105 women with BOT, Vasconcelos et al. observed that cystectomy in unilateral serous BOT is significantly associated with a higher recurrence rate, albeit no impact on survival can be demonstrated . On the contrary, bilateral cystectomy should be definitively favored in bilateral BOT, which is almost always serous, because no significant difference is seen in terms of recurrence rate when compared to unilateral salpingo-oophorectomy and contralateral cystectomy .
In general, conservative treatment of BOTs is associated with higher recurrence rates compared with radical treatment . Patients treated by cystectomy were three times more likely to recur than those treated by oophorectomy . In this series, conservative management was associated with 11% recurrence rate with half of these recurrences successfully managed by repeat conservative surgery. Only 6% of women overall needing eventual complete removal of ovaries for recurrent disease. Malignant recurrences were rare, and while borderline recurrences often occurred more than three years after initial surgery, late malignant recurrences were not observed. These favorable long-term outcomes provide support for conservative surgery for these women. Fertility preservation was not found to be associated with an increased risk of relapse in young patients with advanced borderline tumors (FIGO stages IC–FIGO III) .
Histological subtypes of BOTs do not impact recurrence rates . However, a higher level of lethal recurrence has been reported in cases of micropapillary serous BOT and regarding mucinous BOTs, intraepithelial carcinoma and microinvasion do not play a role in the lethal recurrence rate .
Pregnancy rates of 30–50% can be achieved after FSS for BOT . After repeated conservative surgery or adnexectomy, patients should be advised about the risk of premature ovarian failure. Oocyte preservation can be an option for future pregnancy, but it is not clear whether ovarian stimulation affects relapse time [78, 79]. If FSS is technically not feasible owing to extensive tumor involvement of both ovaries, recent artificial reproductive technologies can be considered, including embryo, oocyte, and ovarian tissue freezing [35, 80].
Standard treatment for EOC consists of bilateral salpingo-oophorectomy, hysterectomy, omentectomy as well as pelvic and para-aortic lymphadenectomy . In women presenting with epithelial ovarian cancer diagnosed at an early stage (typically FIGO stage IA), except of clear-cell subtype, who wish to preserve fertility, unilateral salpingo-oophorectomy together with appropriate staging, omentectomy, pelvic and para-aortic lymphadenectomy can be performed to preserve the uterus and one healthy ovary . In a review of 1110 patients with stage I EOC treated by conservative surgery, Bentivegna et al reported 7% recurrence rate in grade 1 and 11% in grade 2 tumors . A higher recurrence rate (29%) was observed in grade 3 tumors. Recurrence site was the spared ovary in 37% cases. This meta-analysis also concluded that the rate of recurrence was higher in clear-cell tumors (22%) than in mucinous (10%), serous (15%) and endometrioid (13%) subtypes. These low recurrence rates confirm the absence of an increased risk of recurrences specifically due to the use of a conservative surgery at least in patients with stage IA and IC grade 1. The authors concluded to the safety of conservative surgery in stage IA/C grade1/2 EOC. Biopsy of the contralateral ovary is not recommended, if macroscopically normal, due to impairment of ovarian function . In case of bilateral ovarian involvement, FSS should be abandoned and conventional surgery for EOC must be performed.
The 3-year survival rate is about 95.4% after laparoscopic FSS for FIGO stage IA and disease-free survival is 84.6% . FSS study in high-risk ovarian cancer is limited due to the very small number of cases published. Data are insufficient about other tumor types such as clear-cell carcinoma. Overall five-year survival rates have been reported to be as high as 87%, with approximately 12% of patients suffering cancer recurrence after FSS, when combining both low- and high-risk cancers .
GCTs are rare malignancy (1–4% of all ovarian cancers), but frequently occurring in adolescent and young women . Ovarian GCTs are relatively heterogeneous with great variation in their management. They are overwhelmingly unilateral except for dysgerminoma. Most patients with malignant ovarian germ cell tumors are diagnosed with stage 1 disease . FSS for ovarian germ cell tumors usually consists of unilateral adnexectomy, peritoneal staging and omentectomy . Less invasive surgical procedures involving unilateral adnexectomy, cytology and peritoneal sampling have been described in cases of dysgerminoma and immature teratoma limited to the ovary .
In a series of 171 patients with early and advanced malignant ovarian GCT who underwent FSS, Park et al. reported 97% five-year overall survival rate and 86% five-year disease free survival . In this series, the five-year overall survival was 99% for stage I and 91% for stage II-IV. In cases of immature teratoma, five-year survival rates at stages I and II have been described as high as > 93%, with higher recurrence rates in cases of grade 2–3 tumors and advanced-stage tumors . After FSS and standard NACT in yolk-sac tumors, five-year survival has been found to be > 90% and a fertility-sparing approach has been suggested . In pure dysgerminoma, ten-year disease-free survival was > 90%, with overall survival around 100% .
In more advanced stages, debulking surgery without compromising fertility is advised and followed by chemotherapy . After FSS, chemotherapy with bleomycin, etoposide and cisplatin has been associated with improved disease-free survival for children with high-risk malignant GCT . However, close follow up after initial surgery has been suggested for 50% of patients with early stage I tumors .