Fewer Mature Eggs, Similar Birth Chances: New Insight Into IVF Outcomes After Cancer Treatment

Cancer treatment can save a life while leaving difficult questions about fertility years after treatment has ended. A new study suggests that chemotherapy and radiotherapy may continue to affect egg maturation and early embryo development even among women whose ovarian reserve appears comparable with women without cancer. Still, those differences may not translate into poorer pregnancy or live birth outcomes once suitable frozen embryos are available for transfer.

Published in Frontiers in Endocrinology, the study, 'IVF outcomes in cancer survivors: lower ovarian response and early embryo development but comparable success in frozen embryo transfer cycles — a retrospective cohort study,' examined assisted reproductive technology (ART) outcomes among women previously treated with chemotherapy or radiotherapy. Researchers Xiaoqian Zhou and colleagues wanted to separate the effects of cancer treatment from the well-known loss of ovarian reserve, while also exploring whether the amount of time between cancer treatment and IVF or intracytoplasmic sperm injection (ICSI) influenced reproductive outcomes.

Cancer Treatment's Impact May Extend Beyond Ovarian Reserve

Chemotherapy and radiotherapy can damage the ovaries, reduce the pool of remaining follicles and contribute to premature ovarian failure, making fertility an increasingly important part of survivorship care as more young women live for many years after cancer treatment. Earlier research has struggled to determine whether poorer ART results are caused mainly by having fewer available eggs or whether cancer treatment can also affect the developmental quality of surviving oocytes. Laboratory and animal research has raised concerns about DNA damage, mitochondrial dysfunction and epigenetic changes, but evidence from women undergoing fertility treatment has remained limited.

The researchers reviewed women undergoing their first IVF or ICSI cycle at Shanghai Ninth People's Hospital between 2013 and 2024. From 6,616 initially identified cycles, 4,976 remained eligible after exclusions. The final comparison involved 32 cancer survivors previously exposed to chemotherapy and/or radiotherapy and 160 women without malignancy, selected at a 1:5 ratio. Women with ovarian cancer, previous ovarian surgery or endometrial cancer were excluded to reduce the influence of ovarian conditions unrelated to systemic cancer treatment.

A statistical technique called inverse probability of treatment weighting, or IPTW, was used to balance important characteristics between the groups. After weighting, the survivors and controls were comparable in age, body mass index, antral follicle count, basal FSH and estradiol levels, infertility duration, previous pregnancies and doses of fertility medications. This matching was important because it allowed the researchers to examine whether treatment history was associated with reproductive differences even when ovarian function was broadly comparable.

Gestational trophoblastic malignancies were the largest category with 11 patients, followed by breast cancer with six, nasopharyngeal cancer with five, brain cancer with two, thyroid cancer with one and cervical cancer with one; another six women had cancers including osteosarcoma, lung, liver and gallbladder cancers and lymphoma. Ovarian stimulation protocols were comparable between survivors and controls.

Cancer Type Was Not Linked to Major Differences in IVF Outcomes

The clearest differences appeared during ovarian stimulation and early embryo development. After adjustment, cancer survivors produced significantly fewer metaphase II (MII) oocytes, the mature eggs generally capable of being fertilized. The adjusted difference was approximately three fewer MII oocytes among survivors (β = −3.06, 95% CI −5.568 to −0.555, P = 0.02).

The proportion of high-quality embryos was also significantly lower. Controls had a high-quality embryo rate of about 80%, compared with approximately 55% among survivors, producing an adjusted difference of −0.25 (95% CI −0.375 to −0.128). The absolute number of high-quality embryos was also lower after adjustment.

Researchers found no clear difference in the number of retrieved oocytes overall, oocyte retrieval rate, fertilization rate, total number of embryos or number of blastocysts. Cancer type itself was not significantly associated with the IVF outcomes examined.

These findings suggest that chemotherapy or radiotherapy may have reproductive effects extending beyond simply reducing the number of follicles remaining in the ovaries. Since antral follicle counts were comparable after adjustment, the lower number of mature eggs raises the possibility that previous treatment could influence oocyte maturation or developmental competence independently of ovarian reserve. The researchers caution that treatment history alone cannot reliably predict embryo quality, particularly because individual susceptibility and specific treatment exposures may differ considerably.

Frozen Embryo Transfer Brings a More Reassuring Picture

A different picture emerged after embryos reached the frozen embryo transfer (FET) stage. Despite survivors having fewer mature oocytes and a smaller proportion of high-quality embryos earlier in treatment, their chances of pregnancy and live birth after FET were not significantly different from those of controls.

The cumulative pregnancy rate showed no significant difference between groups (HR 1.18, 95% CI 0.59–1.38, P = 0.62), while the cumulative live birth rate was also comparable (HR 0.96, 95% CI 0.43–2.16, P = 0.91). Implantation and biochemical pregnancy rates did not differ significantly either. Spontaneous abortion occurred in 6.25% of survivor pregnancies compared with 4.23% among controls, a difference that was not statistically significant.

The results offer an important distinction for fertility counselling. Cancer treatment may create challenges during egg maturation and embryo generation, but once a viable embryo has been produced and selected for transfer, its capacity to implant and progress toward a live birth may be broadly comparable with that seen in women without a cancer history. The researchers note that FET success also depends on endometrial receptivity, and their results do not prove that uterine function is completely unaffected by cancer treatment because the number of FET cycles was relatively small and confidence intervals were wide.

Does Waiting Longer After Cancer Treatment Matter?

The researchers also investigated a question that matters greatly to survivors planning a family: whether waiting for a longer period after chemotherapy or radiotherapy improves ART outcomes. Within this study population, the interval between cancer treatment and beginning IVF or ICSI was not significantly associated with any measured IVF outcome.

That finding should not be interpreted as evidence that timing never matters. Ovarian function may recover to some extent after treatment while natural age-related fertility decline continues at the same time, creating a complicated relationship between recovery and waiting. The study's small survivor population may also have been unable to detect subtle, non-linear or treatment-specific effects, so the researchers limit their conclusion to saying that no significant relationship was detected within this particular cohort.

The study has useful strengths, including exclusion of women with previous ovarian surgery, careful balancing of ovarian reserve and other characteristics through IPTW, analysis of factors associated with embryo quality, and separate examination of IVF/ICSI and FET outcomes. Its limitations remain substantial: only 32 cancer survivors were included, the research was retrospective and conducted at a single centre, and survivors had different cancers and treatment regimens.

The sample was too small to reliably separate the effects of individual chemotherapy drugs, radiation characteristics or specific cancer types. Smoking, alcohol use, diet, socioeconomic circumstances, genetic background and detailed treatment intensity were also unavailable as potential confounding factors.