Pregnancy and live birth after successful cancer treatment in young women: the need to improve fertility preservation and advice for female cancer patients

KEYWORDS:

• Fertility preservation
• survivorship
• late effects
• cancer

The past four decades have seen substantial improvements in the long-term survival of both children and adults with cancer, leading to the development of the services to optimize long-term health after successful treatment. The potential loss of fertility for some cancer survivors is a major concern of patients, and in the case of children, of their parents [1]. As a result, there has been a rapid development in the clinical application of cryopreservation services to preserve future fertility. These include cryostorage of oocytes, embryos, and sperm in adult and postpubertal adolescent women and men, and are regarded as established clinical practice. Storage of ovarian tissue is also an option for fertility preservation in postpubertal adolescent and adult women, but is the only option for prepubertal girls [2]. The analogous storage of testicular tissue provides an option for prepubertal boys although there remain substantial challenges in to how to use the tissue at a later date to restore fertility, and this has not yet been achieved in humans [3]. A key aspect of the appropriate and effective developments of such services is that decisions should be based on accurate information regarding the risk of loss of fertility following the proposed treatments [4]. However, an accurate fertility prognosis for the individual is often difficult because of a lack of long-term studies and evolving treatment strategies, which include the introduction of novel agents. An additional factor is interindividual variability, which in adult women is dependent on age at treatment, and her pretreatment ovarian reserve [5]. There is therefore a need for accurate and comprehensive information on the chances of pregnancy and its outcome following cancer and treatment, across different ages and diagnoses, to allow informed patient counselling and decision-making.

The US Childhood Cancer Survivor Study (USCCSS) has for many years provided some of the best information available on pregnancy and its outcome in survivors of childhood cancers [6,7]. The British Childhood Cancer Survivor Study (BCCSS) has also reported pregnancy outcomes in over 1000 adult survivors of childhood cancer [8]. The USCCSS has recently provided a very detailed analysis of the risks of different chemotherapeutic agents (in the absence of radiotherapy) on the chances of pregnancy and birth in both males and females [9], with a hazard ratio for live birth of 0.82 (95% confidence intervals 0.76–0.89) for female survivors and 0.63 (0.58–0.69) for males. Chemotherapy regimens that included lomustine and busulphan in girls and platinum compounds in boys were shown to be significantly detrimental to the chance of a live birth. These and other studies demonstrate the impact of childhood cancer and its treatment on subsequent fertility and pregnancy outcomes, particularly the adverse effects of radiotherapy to a field that includes the pelvis (with that also impacting uterine as well as ovarian function in girls) or to the brain. While most pregnancies are uncomplicated, cancer survivors are at increased risk of premature delivery and low birth weight, particularly with radiotherapy to the uterus [7], and pregnancy-related hypertension and diabetes associated with abdominal radiotherapy [10]. These large US and UK cohort studies however have their limitations, particularly that both focus on those with a cancer diagnosis before the ages of 21 and 15 years respectively and both rely on data collected by questionnaire. While the BCCSS is population based, the USCCSS is not, being based in specific centers. A recent analysis of Swedish registry-collected data provides a population based approach and showed an overall hazard ratio of 0.56 for pregnancy in girls diagnosed with cancer in childhood, but was limited to cancer survivors born between 1973 and 1977 [11]. A further informative population-based analysis is a German study of parenthood after Hodgkin lymphoma in girls treated under 18 years of age between 1978 and 1995 [12]. This demonstrated a similar chance of parenthood for most survivors to that of women in the general population. In this study, specific treatments associated with a reduced chance of parenthood included pelvic radiotherapy and alkylating agent chemotherapy at high doses.

These studies therefore provide invaluable information related to the likelihood of pregnancy in relation to both diagnosis and specific treatments, with the Swedish study also showing evidence of changing risk with time as treatment protocols have changed. There are however much more limited data for survivors of adult cancer, with most studies reporting surrogate markers of fertility such as menses in women which while important in indicating likely endocrine function of the ovary, do not reflect fertility, or indeed the likely duration of ovarian function, i.e. time to menopause/premature ovarian insufficiency (POI). These considerations also apply to the interpretation of the potential protective effect of treatments to reduce the impact of chemotherapy on gonadal function, largely the use of GnRH analogs in women with early breast cancer [13]. While the several randomized control trials now published show a consistent protective effect of GnRH analogs for protection against POI, they have only very limited data on subsequent pregnancy, and only one study provides data on an ovarian reserve marker, providing some indication of the amount of ovarian function that might be preserved [14]. A comprehensive analysis of Hodgkin Lymphoma survivors diagnosed in adulthood in the UK showed clear differences in risk of menopause before the age of 40 years by treatment regimen [15], highlighting the risks of alkylating agent therapy and pelvic radiotherapy, particularly in combination. As with the large-scale analyses in childhood cancer, there is the potential for bias as data were available on only approximately 50% of the women initially identified to have received relevant treatment, and age at menopause was self-reported. An additional challenge is the inevitable need for a significant period of time to elapse between treatment and final analysis when pregnancy is the outcome, generally amounting to decades, thus the data may not reflect current treatment protocols. This may have little impact for some conditions where treatments have not changed markedly over recent decades, such as the use of bone marrow transplantation for acute leukemia in young adults. In other conditions, however, the treatment landscape may have changed dramatically. One example of this is cervical cancer where the introduction of cytological screening in the late 1980s led to more cancers being diagnosed by screening with a fall in the incidence of invasive cervical cancer, and with a shift to less invasive, fertility compromising management. There is also an increasing trend for the use of molecular-targeted agents including monoclonal antibodies and kinase inhibitors, whose effects on fertility are largely unknown.

Despite the limitations of existing data, and the pressing need for more comprehensive analysis of pregnancy after cancer, particularly when diagnosed in adulthood, the literature very clearly shows the impact of some diagnoses and treatments. In particular, the harmful effect of pelvic radiotherapy on both fertility and pregnancy outcome in females is consistently reported. Alkylating agents are also associated with a reduced chance of a subsequent pregnancy, although in girls it appears that only the highest doses have marked effects. While we await more accurate data, the information that is currently available provides a basis for the counselling and assessment of girls and women facing a new diagnosis of cancer whose future fertility may be at risk, and who may wish to explore experimental or established strategies to preserve it.

Declaration of interest

The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed. Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Additional information

Funding

The authors work in this field has been supported by grants from the Medical Research Council (G1100357). The MRC Centre for Reproductive Health is funded by the MRC Centre grant MR/N022556/1.