Hematologic cancers in women: from fertility preservation to post-cancer fertility outcomes

Abstract

STUDY QUESTION

How do hematological characteristics affect ovarian reserve, ovarian response to ovarian stimulation, and fertility outcomes?

SUMMARY ANSWER

Although lymphoma characteristics impact serum AMH levels, they do not affect, per se, the response to ovarian stimulation and the number of mature oocytes recovered at the time of fertility preservation; in addition, fertility in survivors of hematologic malignancies is relatively conserved.

WHAT IS KNOWN ALREADY

Hematologic cancers can affect young women of reproductive age. While survival rates have improved over the years due to advances in treatment protocols, the treatments used can impact fertility. Fertility preservation methods, such as oocyte or ovarian tissue cryopreservation, are increasingly offered, but concerns remain about reduced ovarian reserve and response to ovarian stimulation in women with these cancers, which may influence the effectiveness of fertility preservation strategies. Moreover, fertility potential after hematologic cancers has been poorly studied.

STUDY DESIGN, SIZE, DURATION

This is a retrospective, observational bi-centric cohort study. All patients with hematologic cancer (lymphoma, leukemia, myeloma, and myelodysplastic syndrome) who underwent fertility preservation before gonadotoxic treatment (n = 286) from January 2013 to March 2023 were included. For fertility after cancer, and use of frozen oocytes/embryos, the endpoint date was 7 July 2023.

PARTICIPANTS/MATERIALS, SETTING, METHODS

Only patients with lymphoma were included for analysis of ovarian reserve (n = 238) and ovarian response to ovarian stimulation (n = 230). Low ovarian reserve and impaired ovarian response to ovarian stimulation were defined as AMH <1.2 ng/ml and ≤9 mature oocytes retrieved after ovarian stimulation, respectively, according to POSEIDON criteria. A Cox regression model was used to determine predictive factors of impaired response to ovarian stimulation, low ovarian reserve, and pregnancy after cancer. Cumulative incidence of pregnancy and cumulative use of frozen oocytes/embryos was calculated in all patients suffering from hematological malignancies.

MAIN RESULTS AND THE ROLE OF CHANCE

There was an impact of lymphoma characteristics on AMH levels independent of age. After adjustment based on POSEIDON Groups 3 and 4, no specific impact of lymphoma characteristics (e.g. stage, clinical, or biologic B signs) on ovarian response to ovarian stimulation was observed. Regarding post-cancer fertility in the whole population, among the women who tried to conceive, 62% achieved at least one pregnancy, and 85% of these occurred naturally. After adjustment, positive predictive factors for pregnancy were age <35 years, being in a relationship at the first oncofertility consultation, and absence of hematopoietic stem cell transplantation.

LIMITATIONS, REASONS FOR CAUTION

Limitations include potential biases due to the heterogeneity of hematological conditions and the retrospective design, which may lead to missing data. Additionally, the duration of follow-up may not be sufficient to evaluate long-term fertility outcomes.

WIDER IMPLICATIONS OF THE FINDINGS

Lymphoma characteristics did not affect the response to ovarian stimulation in terms of mature oocyte retrieval, although AMH levels were impaired. Reassuring post-cancer fertility data support informed decision-making regarding fertility preservation techniques. Larger prospective studies are needed to tailor oncofertility counseling, ensuring optimized care and reproductive outcomes.

STUDY FUNDING/COMPETING INTEREST(S)

Medical editorial support was provided by Peter Todd of Tajut Ltd (Kaiapoi, New Zealand) and was funded by AFPR (Advances in Fertility Preservation and Reproduction). The authors declare no conflicts of interest.

TRIAL REGISTRATION NUMBER

N/A.

Introduction

Approximately 45 000 new cases of hematologic cancers are reported yearly in France (Defossez et al., 2019). This incidence represents nearly 12% of all new cases of cancer. Hematologic cancers can affect children and young women of reproductive age. As treatment protocols and supportive care strategies continue to improve, the 5-year survival rate has increased over recent years (Shanbhag and Ambinder, 2018). Thus, cancer remission is not the only goal as another major issue is quality of life after treatment, including fertility potential. It is widely acknowledged that chemotherapy and radiotherapy protocols, particularly those involving high-dose alkylating agents, can lead to gonadal toxicity and diminished or a loss of fertility (Meirow and Nugent, 2001).

Fertility preservation is therefore recommended and offered to an increasing number of women diagnosed with hematologic cancer. Several fertility preservation techniques can be proposed: mature oocyte or embryo cryopreservation after ovarian stimulation or IVM and/or ovarian tissue cryopreservation (OTC) (Ashizawa and Kanda, 2020). Vitrification of oocytes or embryos following ovarian stimulation represents the most used method for pubertal women, with an efficiency depending on age, ovarian reserve, and ovarian response to ovarian stimulation by exogenous gonadotropins (Dolmans and Donnez, 2021). Several lines of evidence indicate a baseline decrease in markers of ovarian reserve, especially anti-Müllerian hormone (AMH) at the time of diagnosis of hematologic malignancies, both in adults (Wu et al., 2023) and in the pediatric population (Van Dorp et al., 2014). This finding raises significant concerns regarding the potential impact of treatment regimens on survivor fertility. Moreover, some authors support the controversial idea that ovarian response to ovarian stimulation in women with hematologic diseases may be lower in comparison to healthy controls or other women with solid tumors such as breast cancer (Lawrenz et al., 2012; Sonigo et al., 2018). An overall deterioration of the patient’s health status at the time of diagnosis is currently the main hypothesis to support the impact of hematologic malignancies on ovarian function (Friedler et al., 2012). Indeed, in men with lymphoid pathologies, significant sperm abnormalities were found, and were more severe with advanced disease stage, fever, and presence of biological parameters such as inflammatory syndrome or elevated lactate dehydrogenase (LDH) (Rueffer et al., 2001). Several clinical and biological characteristics such as fever, weight loss, night sweats, mediastinal masses, inflammatory syndrome, anemia, and elevated LDH levels are well-established predictors of lymphoma survival (Cheson et al., 2014). The objective of the present study was to investigate the impact of these characteristics on ovarian reserve tests, response to ovarian stimulation, and number of mature oocytes retrieved and cryopreserved in women included in a fertility preservation program before treatment. In addition, we analyzed the overall survival and fertility of these patients following fertility preservation.

Materials and methods

Study design and subjects

We conducted an observational, retrospective, bi-centric study at Jean Verdier and Antoine Béclère University Hospitals. All women with hematologic cancers (lymphoma, leukemia, myeloma, and myelodysplastic syndrome) having achieved fertility preservation in our centers between 1 January 2013 and 1 March 2023 were included. Patients who had not undergone fertility preservation, those who had already received gonadotoxic treatment prior to fertility preservation (chemotherapy, radiotherapy, or bone marrow transplantation), and those with misdiagnoses hematologic pathology were excluded.

Oncofertility consultation

During the initial oncofertility consultation (baseline = CS1), a blood sample including a hormonal profile (including AMH measurement) was collected irrespective of the day of the cycle. When possible, a transvaginal ultrasound using a 5- to 9-MHz multifrequency probe (Voluson E10, General Electric Medical Systems, Paris, France) was performed for antral follicle count (AFC) (i.e. follicles 2–9 mm). Following the oncofertility counseling, different fertility preservation strategies were offered, including oocyte or embryo cryopreservation after ovarian stimulation or IVM, and/or OTC, depending on patients’ clinical characteristics, type of malignancy and expected treatments, ovarian reserve parameters, and the possibility or not of general anesthesia. A combination of several techniques could also be proposed.

Fertility preservation techniques

Ovarian stimulation for oocytes/embryo vitrification was usually proposed as the first-line treatment after the agreement of the hematologic team. Ovarian stimulation was performed using a random start GnRH antagonist protocol. The gonadotropin starting dose was chosen according to patient’s age, weight, and ovarian reserve. Ovulation was induced by either subcutaneous (SC) choriogonadotropin alfa 250 µg (Ovitrelle^®, Merck Serono, Lyon, France), SC triptorelin 0.2 mg (Decapeptyl^®, Ipsen Boulogne-Billancourt, France), or a combination of both treatments. Transvaginal oocyte retrieval was performed 36 h after triggering. The collected oocytes were then denuded to ensure maturity. Only mature oocytes in metaphase II were vitrified or inseminated by ICSI.

IVM

When ovarian stimulation was not possible due to the urgent need to start chemotherapy, IVM could be proposed if the ovarian reserve was compatible. The IVM technique was performed independently of the ovarian cycle phase with no hormonal treatment and did not require any delay. Puncture was performed under general or local anesthesia under ultrasound guidance. Cumulus–oocyte complexes were recovered and cultured with the patient’s serum and a gonadotropin compound. Oocyte maturation was assessed at 24 and 48 h, and metaphase II oocytes could then be either vitrified or fertilized by ICSI. Oocytes remaining in the germinal vesicle stage were not cryopreserved.

Ovarian tissue cryopreservation

OTC could be proposed when ovarian stimulation was not possible due to an urgent need to start chemotherapy, if highly gonadotoxic treatment was planned before 35 years of age. After laparoscopy, a whole ovary was removed and dissected for OTC as previously described (Peigné et al., 2024).

Systematic patient follow-up

Our fertility preservation program includes a yearly visit for each woman to gather information about oncological treatment and prognosis (including treatments received and any recurrence of the condition), to evaluate ovarian reserve, and to discuss their gynecological monitoring and future plans for parenthood. Moreover, an agreement form was mailed yearly after cryopreservation to determine whether patients wished to keep storing their biological material, and women were asked to complete a survey to collect data on their cancer treatment, their cancer or general health status after treatment, their pregnancy desire, and fertility after treatment. For women who became pregnant, information was collected on the method of conception (natural, use of cryopreserved oocytes/embryos, use of a new ART protocol, or oocyte donation). Information on pregnancy outcome (e.g. live birth, pregnancy loss, ectopic pregnancy) was also collected when appropriate.

Data collection and outcome measures

Data were obtained retrospectively from the two centers after anonymization. For each woman included, characteristics at diagnosis, gynecological history, and ovarian reserve parameters, as well as disease characteristics were collected. Regarding the malignancy itself, the following data were retrieved: type of hemopathy, stage, presence of adverse clinical symptoms (fever, night sweats, weight loss of >10%, large mediastinal mass, and adenopathy), presence of adverse biological parameters (anemia, inflammatory syndrome, white blood cell abnormalities, increased LDH, and Eastern Cooperative Oncology Group Performance Status [ECOG-PS] which is a scale assessing the functional status of patients with cancer). Anemia was defined by hemoglobin <12 g/dl, inflammatory syndrome by C-reactive protein >5 ng/ml or erythrocyte sedimentation rate >20 mm/h, white blood cell abnormalities by hyperleukocytosis >10 000 mm³ or leukopenia <4000 mm³, increased LDH according to laboratory standards, and ECOG-PS 0 or 1. Cancer treatments were classified according to Levine et al. (2010) as either ‘high’ or ‘low’ for their level of gonadotoxicity, particularly for chemotherapy-induced gonadotoxicity. For example, ABVD regimens (Adriamycin, Bleomycin, Vinblastine, Dacarbazine) were classified as low risk of gonadotoxicity and BEACOPP regimens (Bleomycin, Etoposide, Adriamycin, Cyclophosphamide, Vincristine, Procarbazine, Prednisone) were classified as high risk of gonadotoxicity.

For all data collected concerning pregnancy, survival, and utilization of frozen oocytes/embryos, the endpoint date was 7 July 2023.

Response to ovarian stimulation was analyzed through the total number of mature oocytes (metaphase II) retrieved. Successful response to ovarian stimulation was considered if >9 mature oocytes were retrieved. This threshold was defined according to POSEIDON criteria (Alviggi et al., 2016; Humaidan et al., 2016).

Clinical pregnancy was defined as the detection of a fetal heartbeat during ultrasound examination between 6 and 11 gestational weeks.

Ethical approval

The study was approved by the local ethics committee (CLEA-2024-no. 358). A non-opposition agreement to the research was signed by the patients at the time of their initial treatment.

Statistical analysis

Qualitative variables were described as percentages. For quantitative variables, the normality of the distributions was assessed using the Shapiro–Wilk and Kolmogorov–Smirnov tests. Both tests indicated that none of the variables followed a normal distribution. Consequently, we chose to present the median and interquartile range (IQR) as descriptive statistics for the quantitative data. Comparative analysis was performed with Mann–Whitney test. A P-value <0.05 was considered statistically significant.

For patients suffering from lymphoma, univariate and multivariate analyses were performed via a logistic regression model to identify factors associated with diminished ovarian reserve and impaired ovarian response to ovarian stimulation. When patients had more than one ovarian stimulation cycle, only the first was considered for the study concerning ovarian stimulation outcomes. The analysis of the impact on ovarian reserve parameters was conducted only in patients with documented AMH levels. Factors associated with univariate analysis with a P ≤ 0.10 were selected for entry into the model, except in case of >20% missing data. Odds ratios (ORs) and 95% CIs were determined for each factor. A P-value <0.05 was considered statistically significant. A multivariate analysis of predictive factors of good response to stimulation (i.e. a number of preserved oocytes >9) was performed, adjusting for POSEIDON Groups 3 and 4, including as covariables type and stage of hemopathy, presence of clinical or biological parameters, and all parameters with P < 0.10 in the univariate analysis.

Regarding patient follow-up concerning pregnancy and reuse of frozen oocytes/embryos, the endpoint date was 7 July 2023 and involved the whole population of patients who underwent fertility preservation for hematological cancer. For those who had achieve pregnancy and live birth, the time to conception was calculated from the date of CS1 to the year of onset of pregnancy. For those who did not achieved pregnancy, follow-up was calculated from the date of CS1 to endpoint date or death. The cumulative incidence of pregnancies was estimated using the Kaplan–Meier method. A Cox logistic regression was calculated to analyze the hazard ratio of predictive factors of pregnancy after hematological cancer. All parameters with P < 0.20 in univariate analyses were included in the multivariate analysis.

Finally, an analysis of the overall survival and disease-free survival was conducted. Time to recurrence or death was calculated from the date CS1. Disease-free survival was estimated using the Kaplan–Meier method. All statistical tests were two-sided at the 0.05 statistical significance level. Statistical analyses were performed using NCSS (NCSS 2021 Statistical Software).

Results

Patient and hematologic disease characteristics at first oncofertility consultation and fertility preservation strategy

Between 1 January 2013 and 31 March 2023, 391 women with hematologic cancer were seen for oncofertility counseling at one of the two fertility preservation centers. Among these, 286 patients were included in the present study (Fig. 1). Patient characteristics at CS1, hematologic disease characteristics, and treatment strategy are reported in Supplementary Table S1. Median (IQR) age was 26.0 (22.0–30.0) years and most patients were nulliparous (81.4%). Median AMH level was 2.4 (1.3–3.9) ng/ml and median AFC was 21 (15–30). The majority of patients (96.1%) had lymphoid pathology and 56% received treatment with high gonadotoxicity levels.

Most patients (236/286, 82.5%) underwent a fertility preservation technique by oocyte/embryo cryopreservation after ovarian stimulation (Fig. 1). Three patients had two ovarian stimulation cycles. A total of 48 women had IVM and 18 had OTC. Among the 18 patients who underwent OTC, 15 had a combination of another fertility preservation technique (IVM [n = 9] or ovarian stimulation [n = 6]). One patient underwent ovarian stimulation followed by IVM. Patients who underwent fertility preservation through IVM were predominantly diagnosed with lymphoma (43/48), with the remaining five having acute leukemia. Their median age was 27.0 (22.0–32.7) years, AFC was 20.5 (15.0–30.1), and AMH was 1.8 (1.1–4.0) ng/ml. The respective numbers of retrieved and mature oocytes were 6.5 (3.3–12.8) and 4.5 (1.25–8).

Impact of hematological parameters on AMH levels and response to ovarian stimulation

As the majority (275/286, 96.2%) of patients had lymphoma, we only included patients with lymphoma for analyses of AMH levels and ovarian stimulation. Patient and disease characteristics for those with lymphoma are described in Supplementary Table S2.

AMH value and disease characteristics were available in 238 women diagnosed with lymphoma. The multivariate analysis found an independent effect of adverse biological parameters (OR 2.5, 95% CI 0.9–6.9; P  =  0.03), adverse clinical symptoms (OR 1.8, 95% CI 1.1–2.9; P  =  0.02), and lymphoma stage (OR 1.6, 95% CI 1.1–2.3; P  =  0.02) on the risk of low AMH levels at CS1 (Table 1).

Analysis of ovarian stimulation was performed in 230 patients with lymphoma. Ovarian stimulation characteristics and outcomes are presented in Table 2. Median (IQR) stimulation duration was 11 (10–12) days with a median starting FSH dose of 300 (225–300) IU/day and median total dose of 3000 (2025–3825) IU. The median number of oocytes retrieved was 14 (8–22), with a median maturation rate of 81.3% (67.9–90.9), resulting in 10.5 (6–17) mature oocytes per patient. According to POSEIDON criteria, 22.6% of patients were classified as poor responders (Group 3 or 4) before ovarian stimulation. After ovarian stimulation, 19.5% had an unexpected suboptimal response and were classified in POSEIDON Group 1 or 2.

We further examined whether the characteristics of all hematologic malignancies were related to a risk of impaired ovarian stimulation outcome defined as having ≤9 mature oocytes retrieved (Table 3). After adjustment for POSEIDON criteria before ovarian stimulation (Groups 3 and 4), we found that the type of hemopathy, lymphoma stage, presence of adverse clinical symptoms, presence of adverse biological parameters, or ECOG-PS stage were not associated with impaired ovarian response to ovarian stimulation in terms of number of mature oocytes retrieved (Table 4).

Survival and fertility after hematological cancer

Of the initial cohort of 286 patients with fertility preservation before hematologic cancer treatment, 38 were lost to follow-up immediately after completing their fertility preservation process. For the others (n = 248), the median (IQR) follow-up was 51 (27.4–75.0) months after CS1. Five patients from the ovarian stimulationort died during the study period. At 70 months after CS1, the estimated recurrence-free survival rate was 86.9% (95% CI 82.4%–91.4%) (Supplementary Fig. S1).

Among the 286 patients included, information about pregnancy desire was available for 237 women (Fig. 2A). A total of 95 (40.1%) women attempted pregnancy by the study endpoint date, 59 (62.1%) achieved at least one pregnancy and 54 (56.8%) had at least one live birth. Ten patients (10.5%) had two live births. The first pregnancy was mainly obtained without medical assistance (Fig. 2B). The first pregnancy ended with a live birth in 49 (83.0%) cases, an early miscarriage in six (10.2%) cases, and other termination in four (6.8%) cases (i.e. late miscarriage, ectopic pregnancy, therapeutic abortion, and intrauterine fetal death). In the case of pregnancy attempt, the cumulative incidence of pregnancy was 50% 48 months after CS1 (Fig. 2C).

Among women who attempted to conceive, we investigated predictive factors for achieving pregnancy. In multivariate analysis, factors that were positively associated with pregnancy after cancer were age <35 years at CS1, being in a relationship at the time of diagnosis, and not undergoing bone marrow transplant (Table 5). The level of gonadotoxicity of the treatment received per se was not found to be associated with pregnancy in the univariate analysis. The mode of conception was natural in 86.7% (19/22) of women who received low-risk gonadotoxic treatment and in 83.8% (31/37) of women who received high-risk gonadotoxic treatment. Pregnancies resulting from the use of frozen oocytes/embryos were all in patients who received high-risk gonadotoxic treatment.

Utilization of cryopreserved material

Most patients (259/276, 93.8%) opted to maintain cryopreservation of their materials at the time of their pregnancy attempt after cancer treatment for possible future use. Of the 95 women who attempted to become pregnant, 17 (17.9%) used their cryopreserved oocytes or embryos by the study endpoint (Supplementary Fig. S2). Six pregnancies and seven live births (1 twin pregnancy) were recorded after the use of frozen material, accounting for 10% of the total pregnancies. Among these, five resulted from the use of either oocytes or embryos retrieved from ovarian stimulation and one resulted from IVM. Two patients underwent ovarian cortex transplantation. At the end of data collection, there have been no pregnancies resulting from this procedure.

Discussion

In this large cohort of patients who had fertility preservation before lymphoma treatment, the specific stage of the lymphoma, adverse clinical symptoms, or adverse biological parameters were not found to influence ovarian stimulation results after adjustment for POSEIDON criteria. However, low AMH levels before gonadotoxic treatment were found to be associated with those parameters (i.e. adverse clinical and/or biological parameters, and/or higher lymphoma stage). For the entire cohort of patients with hematologic cancers who had fertility preservation, the survival rate was good, and the pregnancy rate was good in women who attempted to conceive, with pregnancies mainly obtained without medical assistance.

Ovarian stimulation in our cohort of patients with lymphoma demonstrated a relatively good outcome with a median of 14 oocytes retrieved in a young population (median age 26 years), which surpasses the cutoff of 9 when referring to POSEIDON criteria. The study by Cobo et al. (2018) reported an average of 12.5 oocytes retrieved in patients undergoing oncological fertility preservation with a median age of 32 years. Additionally, two other studies (Friedler et al., 2012; Fraison et al., 2023) found an average number of mature oocytes ranging from 6 to 9.4. Compared to other patients undergoing ovarian stimulation for male factor infertility (Quintero et al., 2010), elective oocyte cryopreservation (Cobo et al., 2018), oocyte donation (Brun et al., 2021), and other types of cancer (Domingo et al., 2012), some studies have found no significant difference in ovarian response in patients with lymphoma. However, other studies have suggested a reduced response to ovarian stimulation compared to either infertile patients (tubal or male infertility) undergoing IVF (Friedler et al., 2012) or patients with other types of cancer, primarily breast cancer (Sonigo et al., 2018), undergoing fertility preservation. In previous analyses (Quintero et al., 2010), there is often an increase in the total dose of gonadotropins used and increased duration of stimulation. Indeed, in our study, the median initial FSH dose was 300 IU/day, which is considered a high dose.

One hypothesis to explain this possible impaired response to ovarian stimulation is the impact of the lymphoma via catabolism, stress mechanisms, and pro-inflammatory cytokines produced by tumor tissue (Van Dorp et al., 2014; Paradisi et al., 2016). Regarding only the impact of hyperthermia on the outcomes of ovarian stimulation, studies in animals have consistently found a negative role of heat stress on ovarian steroidogenesis leading to impaired follicular growth (Awwad et al., 2012). We specifically focused on Ann Arbor staging and other prognostic factors commonly used in lymphoid hematologic malignancies, including clinical ‘B-symptoms’ such as weight loss, fever, or night sweats, bulky disease, or biological parameters such as inflammatory syndrome or elevated LDH. We did not observe an impact of any of these adverse clinical symptoms and biological parameters on the chances of obtaining a good number of mature oocytes after adjusting for POSEIDON Group 3 and 4 criteria. This is consistent with a previous work by our group which showed no influence of cancer stage on response to ovarian stimulation (Sonigo et al., 2018). As only fever at initial disease diagnosis was assessed in our ovarian stimulationort, a potential information bias regarding the presence or absence of fever during ovarian stimulation may exist.

Approximately 22% of patients with lymphoma in our cohort had POSEIDON poor response criteria (Group 3 or 4), which is higher than the 8.4% reported in the study by Esteves et al. (2021) for 13 146 patients undergoing ovarian stimulation for various indications. This difference in outcome might be explained by the hypothesis of a decrease in the initial ovarian reserve parameters within our cohort. Ovarian reserve can be evaluated by several measurements such as AFC, basal serum FSH, and AMH. We chose to focus on AMH levels to avoid measurement errors or inter-observer variability (Dewailly et al., 2014). After adjustment for age, disseminated disease and the presence of adverse clinical symptoms and biological parameters were significantly associated with impaired AMH levels at CS1. Several controversial studies have highlighted a decline in ovarian reserve in patients with lymphoma prior to treatment initiation. Some authors argue that there may be a difference in AMH levels among patients with oncologic disorders compared to healthy women (Wu et al., 2023), which is particularly pronounced in cases with fever and anemia in hematologic disease (Lekovich et al., 2016). The 2024 meta-analysis conducted by Katzir et al. (2024) included eight studies that investigated ovarian reserve parameters and response to ovarian stimulation among individuals with hematologic cancers undergoing fertility preservation prior to oncologic treatment, compared to healthy controls. Their findings revealed a diminished AMH level in the cancer cohort despite the latter being of a younger age. With respect to stimulation response, no discernible difference was observed in the quantity of retrieved oocytes between the two cohorts.

With respect to the fertility of the patients after all hematologic cancer treatment, after a median follow-up period of 51 months, 62% of women who attempted to conceive experienced at least one pregnancy, in most cases spontaneously, and 92% of them had at least one live birth. In our cohort, the cumulative incidence rate of pregnancy at 4 years following CS1 stands at 50%. These findings align with studies investigating the reproductive capacity of survivors of hematologic disease. In the meta-analysis by Drechsel et al. (2023) which examined the follow-up of patients treated for Hodgkin lymphoma, few studies investigated the likelihood of pregnancy compared to the general population. A German study found that the proportion of women Hodgkin lymphoma survivors having children was similar to the general population except for those aged from 40 to 44 years and those who had undergone pelvic irradiation (Behringer et al., 2012). Another study comparing the cumulative rate of live births among Danish women supported a comparable birth rate for patients who underwent treatment for Hodgkin lymphoma compared to those who did not (Øvlisen et al., 2021). It is necessary to consider the potential influence of selection bias for our cohort. Specifically, our study population exclusively comprised women who underwent fertility preservation before any cancer treatment, indicating a selective inclusion of generally younger individuals with favorable ovarian reserve parameters and promising initial pregnancy prospects. This inherent bias may result in an overestimation of pregnancy rates compared to the broader population of women diagnosed with hematologic cancer, as those with compromised ovarian reserve or delayed fertility preservation may exhibit divergent outcomes.

Predictive factors for pregnancy among patients with hematologic cancer encompass, in our cohort, age <35 years, being partnered at the time of diagnosis, and absence of bone marrow transplantation. As expected, a younger age and the absence of hematopoietic stem cell transplantation were positive predictive factors of pregnancy since these women have a better ovarian reserve and are not exposed to total body irradiation or myeloablative chemotherapy-based conditioning, both of which are known to significantly increase the risk of gonadal dysfunctional (Swerdlow et al., 2014). An explanation for the predictive significance of being in a relationship could be the presence of an established parenthood plan.

We observed a 6% total rate of gamete reutilization and 18% rate in patients attempting to become pregnant, which appears consistent with findings in the literature (Mayeur et al., 2021). However, caution is warranted as we included patients until March 2023 and this rate could be underestimated due to the limited follow-up duration for some patients. Regarding pregnancies resulting from the use of frozen oocytes or embryos, the achieved outcome shows a pregnancy rate of 35% (6 of 17 individuals). This is in line with the results from the meta-analysis by Fraison et al. (2023), where the percentage of women with at least one live birth was 32% after oocyte vitrification with a 4.6% reuse rate for preserved gametes across eight different studies.

The primary strengths of our study lie in its innovative approach, with a significant study population being the largest to date in assessing the characteristics of lymphoma and its effect on ovarian stimulation response. Furthermore, the included population is representative of patients affected by these oncologic conditions, characterized by a young age and good ovarian reserve. Thanks to regular patient follow-up, we were able to gather comprehensive information with minimal missing data. Only 38 patients were lost to follow-up out of 286, accounting for 13.3% of the total cohort.

Our study encompassed a broad spectrum of hematologic pathologies. However, these diseases vary in terms of origin, classification, prognosis, and treatments. This heterogeneity could introduce classification bias during their combined analysis. We opted to focus on studying ovarian stimulation response in patients with lymphoma. Indeed, other hematologic pathologies such as leukemias were often excluded due to previously administered gonadotoxic treatments, with their therapeutic management requiring a delay before initiating treatment. One limitation of using AMH level as the parameter for assessing ovarian reserve is the lack of consideration for potential factors that might induce fluctuations in AMH levels, such as the utilization of hormonal contraceptives. In addition, the retrospective design of our study introduces the possibility of missing data, particularly regarding the documentation of adverse clinical symptoms or biological parameters, which represents a potential limitation. Finally, even though our study had a median follow-up of more than 4 years, the duration was still too short to evaluate long-term fertility.

Conclusion

In conclusion, while our current findings indicate that lymphoma characteristics such as stage and the presence of adverse clinical symptoms or B biological signs do not appear to influence the response to ovarian stimulation in terms of number of mature oocytes retrieved, our study has revealed an independent impact of these disease characteristics on AMH levels at diagnosis. These results provide clearer information regarding the potential impact of their condition on ovarian response to ovarian stimulation and on ovarian reserve parameters.

At a median follow-up of 51 months, data regarding post-cancer fertility were reassuring with 40% of our cohort trying to conceive and 62% of these succeeding, mainly without medical assistance. These results improve our knowledge on how to provide better advice for women regarding various fertility preservation techniques, their effectiveness, and their safety.

However, conducting larger prospective studies holds the potential to refine the guidance offered during personalized oncofertility consultation. This would allow for a more nuanced approach to treatment planning, considering both the type of pathology and the prospects for natural pregnancy after treatment. By advancing our research efforts, we can better optimize personalized patient care and fertility preservation strategies, ensuring improved reproductive outcomes.

Data availability

The data underlying this article will be shared on reasonable request to the corresponding author.

Authors’ roles

E.G., C.So., S.R., and M.P. made substantial contributions to the study conception and design, or acquisition of data, or analysis and interpretation of data. C.V., F.E., V.P., C.W., J.L., C.Si., C.B., A.M., A.B., and M.G. drafted the article or revised it critically for important intellectual content. E.G., C.So., and M.P. gave final approval of the version to be published. E.G., C.So., and M.P. agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Funding

Medical editorial support was provided by Peter Todd of Tajut Ltd (Kaiapoi, New Zealand) and was funded by AFPR (Advances in Fertility Preservation and Reproduction).

Conflict of interest

The authors declare no conflicts of interest.

References