Impact of immune checkpoint inhibitors on pregnancy: current evidence and future directions

Abstract

Cancer treatment has dramatically improved with the introduction of immune checkpoint inhibitors (ICIs). They are now common treatments for individuals of reproductive age, an important consideration when treating pregnant patients. Cancer in pregnancy is rare and guidelines available for ICI use during gestation are based on pre-clinical studies, single human case reports and spontaneous report databases. The immune pathways including PD-1 and PD-L1/PD-L2 and CTLA-4 provide critical immune inhibitory checkpoints at the maternal–fetal interface. These pathways support immune tolerance during implantation and the development of the placenta. Experimental animal studies demonstrated that inhibiting these immune pathways will result in increased frequency of adverse events such as spontaneous abortions and premature births. Human placental transfer of ICIs across the placenta is minimal in early gestation and increases in later gestation, coinciding with less exposure to developing organs during organogenesis and more exposure closer to term. There are now seventeen documented cases in the literature describing pregnant patients treated with ICIs. Most neonates developed normally in infancy, while there are a few documented cases of suspected neonatal immune-mediated complications. Experts including regulatory agencies and oncology professional organisations advise against the use of ICIs in pregnancy unless their use is critical to the survival of the patient and there is no safer alternative. Patient counselling regarding the risks of using ICIs during pregnancy, postpartum or both, involves a multidisciplinary team and requires shared decision making regarding maternal health, infant follow-up and standardised documentation for understanding potential delayed effects on neonate immunity as well as providing accurate information for future counselling.

Search strategy

References were identified using PubMed through searches using combinations of the terms “pregnancy,” “immune checkpoint inhibitors,” “PD-1,” “PD-L1,” “CTLA-4,” “immunotherapy” and “cancer in pregnancy.” Searches were limited to articles published in English and between 1 January 1999 and 20 November 2025. Additional sources, including case reports, case series, pharmacovigilance analyses and regulatory documents, were identified through citation tracking and review of reference lists. The final selection of articles was based on relevance to the mechanisms, clinical outcomes and management considerations discussed in this review. Formal study selection processes were not applied, as this was not a systematic review.

Patient and public involvement

Patients and/or members of the public were not involved in the design, conduct, reporting or dissemination plans of this research. This study is a review of previously published literature and did not involve direct patient or public participation.

Introduction

Immune checkpoint inhibitors (ICIs) have transformed the treatment of many solid and haematological malignancies. Since the approval of ipilimumab in 2011, their use has expanded quickly, including among individuals of reproductive age. As a result, unintended exposure in pregnancy to ICIs may occur much more commonly. Cancer in pregnancy is rare, occurring in about 0.02% to 0.1% of cases.¹ The most frequently diagnosed malignancies during pregnancy include breast and cervical cancer, haematological malignancies and melanoma.^2,3 As current treatment options for several malignancies increasingly include immunotherapies, safety information on human exposures to ICIs during pregnancy is based on limited data from preclinical animal studies and a small number of case reports documenting human exposures to ICIs in utero.

During pregnancy, an equilibrium exists in the maternal immune system that allows tolerance of the semiallogeneic fetus and placenta, which expresses both maternal and paternally derived antigens, while simultaneously maintaining protection against infection.^4,5 This immune response changes throughout pregnancy. During early pregnancy, there is a pro-inflammatory environment that supports implantation and the development of the placenta.⁶ This process occurs at the maternal–fetal interface within the decidua, a specialised maternal tissue containing diverse immune cell populations that regulate trophoblast invasion and promote immune tolerance.⁶ As gestation advances, the immune system becomes more anti-inflammatory and promotes fetal growth, and returns to a more pro-inflammatory state near term to stimulate labour.^7,8 Immune checkpoints that control these interactions, such as PD-1 and its ligands PD-L1 and PD-L2, as well as CTLA-4, function to facilitate maternal tolerance at the maternal–fetal interface.⁹ Blocking these pathways can enhance anti-tumour immunity but may interfere with immune mechanisms that are important for maintaining a healthy pregnancy.

The increasing use of ICIs in individuals of reproductive age and the rare but clinically important occurrence of cancer in pregnancy highlights an urgent need to understand the current body of evidence. Therefore, this review examines the immunology underlying checkpoint signalling during pregnancy, provides an overview of the pharmacokinetics and mechanisms that contribute to fetal exposure to ICIs, and compiles all available peer-reviewed literature that describes in utero exposure to ICIs. Additionally, the clinical and ethical implications will be discussed, and recommendations for future research will be made.

Pharmacology and transplacental transfer of ICIs

As IgG monoclonal antibodies, ICIs do not cross the placenta by passive diffusion due to their large molecular size. They depend on active transport through the neonatal Fc receptor (FcRn). FcRn expression is low in the first trimester, begins to rise at around 14 weeks, and reaches peak activity in the third trimester.¹⁰ As a result, fetal exposure during organogenesis is expected to be minimal, with transfer increasing later in pregnancy. Transport efficiency also varies by IgG subclass: IgG₁ is transported most efficiently, followed by IgG₄.¹⁰ Therefore, ICIs based on an IgG₁ framework, such as ipilimumab, might theoretically result in greater fetal exposure compared with IgG₄-based PD-1 agents. However, clinical data directly comparing fetal exposure between ICI subclasses are currently limited.

Immune checkpoint pathways at the maternal–fetal interface

PD-1 pathway

Programmed death protein 1 (PD-1) is a key regulator of maternal–fetal immune tolerance. It is present on activated T cells, B cells, natural killer cells and antigen-presenting cells.^11,12 In early gestation, PD-1 levels fall in peripheral T cells, while expression stays high in decidual immune cells and remains elevated across pregnancy.¹³

When PD-1 binds with receptors PD-L1 or PD-L2, an inhibitory signal is sent that limits T-cell proliferation, cytokine release and cytotoxic function.¹⁴ This pathway is most active at the maternal–fetal interface, where decidual CD4+T cells, CD8+T cells and regulatory T cells show much higher PD-1 expression than their peripheral counterparts.¹⁵ These features suggest that PD-1 helps control maternal effector responses to fetal alloantigens.¹⁶

PD-L1, the receptor for PD-1, is highly expressed by extravillous trophoblasts, decidual stromal cells and decidual macrophages. Its expression increases between 10 and 12 weeks of gestation.^17,18 Early in pregnancy, PD-L1 on decidual macrophages, together with major histocompatibility complex molecules, aids in implantation and early placental development.^18,19 Interferon-gamma levels continue to decline throughout the second and third trimesters.²⁰ Disruption of PD-1 signalling can elevate local interferon-gamma activity, which experimental models associate with impaired implantation, fetal stress and pregnancy loss.²¹

Preclinical studies further highlight the role of PD-1 and PD-L1 in supporting pregnancy. In an allogenic (genetically different) murine model, PD-L1 blockade raised fetal resorption from 18% to 86%, while syngeneic (genetically identical) matings remained unaffected.²² Placental tissue from these pregnancies exhibited T-cell clusters around resorption sites, indicating T-cell-mediated rejection.²²

The PD-1 pathway also closely interacts with regulatory T-cell-mediated tolerance.¹⁶ Removing CD25+regulatory T cells increased fetal resorption from 19% in controls to 32% in treated murine models.²³ In the abortion-prone strain, adoptive transfer of regulatory T cells restored normal pregnancy outcomes.

Data from non-human primates align with these findings. Pregnant cynomolgus monkeys exposed to nivolumab showed a dose-independent increase in spontaneous abortions and early neonatal deaths. In contrast, surviving infants appeared structurally normal and demonstrated typical neurobehavioural and immunological development during 6 months of postnatal observation.¹⁶

CTLA-4 pathway

Cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) is an inhibitory receptor that supports maternal–fetal immune tolerance. It is expressed on regulatory T cells and is upregulated on activated T cells.^24,25 CTLA-4 reduces effector T-cell activation, limits interleukin-2 secretion and restricts proliferation, thereby helping to prevent excessive maternal immune responses following T-cell activation.²⁶

During pregnancy, CTLA-4 expression is higher in decidual CD4+T cells, CD8+T cells and regulatory T cells compared with matched non-uterine peripheral cells.²⁷ Its ligands, CD80 and CD86, are found on decidual macrophages and stromal cells, with peak levels in early and mid-gestation.²⁸ CTLA-4 engagement at the maternal–fetal interface promotes a T helper 2-dominant cytokine profile and suppresses T helper 1-driven inflammation.^29,30 These changes support implantation, placental growth and the maintenance of immune tolerance.³¹

CTLA-4 expression is also found in placental fibroblasts.³² These cells do not directly present antigens, but the finding indicates a local regulatory function in placental tissue.

Preclinical data further highlight the role of CTLA-4 in maintaining pregnancy. Murine models show that CTLA-4 blockade or treatment with anti-CTLA-4 antibodies increases embryo resorption, decreases fetal size and reduces litter viability.³³ Conversely, CTLA4Ig, which binds to CD80 and CD86 and prevents CD28-mediated activation, expands peripheral CD4+CD25+ regulatory T cells and improves outcomes in mouse strains prone to miscarriage, underscoring CTLA-4 signalling at the maternal–fetal interface.³⁴

Findings in non-human primates align with these observations. Ipilimumab, a CTLA-4 monoclonal antibody, caused rare congenital anomalies, including unilateral renal agenesis and an imperforate urethra, at the highest experimental dose of 30 mg/kg.^35,36 When given to pregnant cynomolgus monkeys during the first two trimesters at doses 2.6 to 7.2 times the human clinical dose, ipilimumab caused no significant maternal or fetal toxicity.^35,36 However, high-dose ipilimumab in the third trimester led to maternal weight loss, stillbirths, premature delivery and increased neonatal mortality.^35,36

Beyond these well-characterised immune checkpoint pathways, newer targets such as lymphocyte activation gene-3 (LAG-3) have entered clinical practice. Dual inhibition of LAG-3 and PD-1 has demonstrated improved efficacy in advanced melanoma.³⁷ However, there are currently no published data on LAG-3 inhibitor exposure in human pregnancy, and its role in maternal–fetal immune tolerance remains undefined.

Maternal and fetal safety evidence

Seventeen peer-reviewed reports now describe pregnancies with in utero exposure to ICIs, including two twin gestations.^38,39 Most cases involved metastatic melanoma.^38–48 Other reported malignancies included renal cell carcinoma,⁴⁷ gastric carcinoma,⁴⁹ relapsed Hodgkin lymphoma,^50,51 placental-site trophoblastic tumour⁵² and Lynch-associated endometrial and colorectal cancer.⁵³ Findings from each study are summarised in online supplemental table S1.

Anti-PD-1 monotherapy remained the most frequent exposure.^{39–41,43,45–53} Combination PD-1 and CTLA-4 blockade was used mainly for melanoma.^38,42,44,47

Fetal complications included intrauterine growth restriction,^38,43,49 umbilical Doppler abnormalities suggesting placental dysfunction⁴⁹ and abnormal fetal heart tracing.⁴⁷ Preterm birth was common across reports, with gestational age at delivery ranging from extreme prematurity at 24 weeks to term. Arrest of cervical dilation was also reported.⁴⁷

Maternal immune-related adverse events (irAEs) included diarrhoea,⁴¹ immune hepatitis,^42,47 gestational hypothyroidism,⁴⁹ suspected hypophysitis or adrenalitis⁴⁸ and severe thrombocytopaenia.⁴⁹ Obstetric complications included intrahepatic cholestasis⁴⁸ and Hemolysis, Elevated Liver enzymes, and Low Platelet count. syndrome,³⁹ one case of which was attributed to obstetric complications rather than immunotherapy.³⁹ No maternal deaths were linked to irAEs; one death resulted from aggressive malignancy 1 day post partum.⁴⁴

Where reported, most post-delivery placental pathology findings were negative for metastatic involvement.^{38–40,43,45,47–49,51,52} One report identified maternal-side placental melanoma deposits without fetal involvement.⁴⁴

Neonatal outcomes varied widely across reported cases. One structural abnormality, unilateral limb absence from amniotic band sequence, was identified but unrelated to gestational ICI exposure as noted by the authors.³⁹ A limited number of reports described neonatal findings that the original authors interpreted as possible or suspected immune-related effects. These included congenital hypothyroidism following first-trimester nivolumab exposure, which has been reported in the literature as a potential immune-related effect⁴³; severe immune-mediated gastroenterocolitis requiring prolonged immunosuppression⁴⁸; and a complex neonatal course with features the authors described as possibly suggestive of an irAE.⁴⁹ Prenatal imaging in the latter case also showed fetal bowel dilatation, which the authors suggested could be a gastrointestinal irAE linked to in utero pembrolizumab exposure.⁴⁹ Most other neonates in the literature, including those born very preterm, showed normal early development.

In addition to individual case reports, pharmacovigilance databases have provided some insight into the safety of ICIs in pregnancy. Noseda et al conducted an initial evaluation of VigiBase, the global pharmacovigilance database developed and maintained by the WHO-Uppsala Monitoring Centre (WHO-UMC), examining reports of spontaneous abortion, intrauterine growth restriction and preterm birth following ICI exposure.⁵⁴ Although the analysis lacked multivariable adjustment and detailed clinical information, it did not identify meaningful differences between ICI-exposed pregnancies and those exposed to non-ICI anticancer therapies.⁵⁴

A more comprehensive analysis of VigiBase by Gougis et al assessed 3558 pregnancy-related reports, including 91 involving ICIs, across 45 predefined maternal, fetal and neonatal outcomes.⁵⁵ The reported rates of fetal loss, congenital anomalies, intrauterine growth restriction, neonatal complications and maternal adverse outcomes in the ICI group were similar to those seen with other anticancer drugs.⁵⁵ The only clear signal was a higher reporting rate of preterm birth in cases that used combined PD-1 or PD-L1 blockade with CTLA-4 inhibition, while there was no pattern with monotherapy.⁵⁵ A limited number of reports described neonatal immune-related effects. These included a maternal stage III melanoma case treated with adjuvant pembrolizumab in which maternal immune complications (antiphospholipid syndrome, pneumonitis and thyroiditis) resulted in a spontaneous abortion, along with a neonatal pneumonitis report following combined in utero exposure to pembrolizumab, carboplatin and pemetrexed during maternal treatment for non-small cell lung cancer.⁵⁵

Such pharmacovigilance work has notable strengths; since ICI use during pregnancy is rare, no single centre or national registry can gather enough cases to identify significant trends. A global system like VigiBase relies on reports from numerous countries and records exposures that might otherwise go unnoticed.⁵⁶ The limitations of spontaneous-reporting systems must, however, be acknowledged; the clinical detail in these reports varies, and essential information is often missing, including gestational age at exposure, timing of delivery, placental findings and neonatal follow-up.⁵⁶ Additionally, pharmacovigilance databases record only cases in which an adverse event was reported and do not indicate how many exposed patients had no complications.⁵⁷ Without this denominator, measures such as the reporting OR cannot estimate risk or incidence. They indicate that an event is reported more often than expected, which reflects reporting patterns rather than true rates of harm.

Clinical translation, care and patient considerations

Given the complexities of managing pregnant individuals with cancer who are candidates for ICIs, a coordinated, multidisciplinary team should be assembled, involving oncology, maternal–fetal medicine, neonatology, gynaecology, genetic counselling, pharmacy and psychosocial support. The patient can be supported in decision making about their medical care, pregnancy management, giving and understanding maternal and fetal benefits and risks.

ICI treatment during pregnancy

Due to limited human safety data, most current guidelines advise against using ICIs during pregnancy unless maternal benefit clearly outweighs fetal risk. The American Society of Clinical Oncology (ASCO) notes that, although isolated case reports show positive neonatal outcomes after first-trimester or second-trimester exposure to ipilimumab or nivolumab, these cases should not be seen as proof of safety.⁵⁸ Therefore, ASCO recommends that PD-1/PD-L1 or CTLA-4 inhibitors be used during pregnancy only when no safer options are available and when withholding treatment would be expected to significantly compromise maternal outcomes. Product labelling by the US Food and Drug Administration (FDA) reflects the same level of caution. All FDA-approved ICIs are categorised as pregnancy risk agents based on non-human primate studies indicating increased rates of fetal loss and neonatal death following maternal exposure to nivolumab or ipilimumab.^16,35,36 As a result, FDA labelling advises against ICI therapy during pregnancy, recommends effective contraception and suggests delaying pregnancy for at least several months after the final dose (table 1).

Table 1. ICIs: IgG subtype, drug half-life and FDA recommendations for lactation and contraception.

Immune checkpoint	Agent	IgG subclass	Half-life	Breastfeeding avoidance (post-treatment duration)	Contraception (post-treatment duration)
PD-1	Nivolumab69	IgG4	25 days	5 months	5 months
	Pembrolizumab70	IgG4	22 days	4 months	4 months
	Cemiplimab71	IgG4	22 days	4 months	4 months
PD-L1	Atezolizumab72	IgG1	27 days	5 months	5 months
	Durvalumab73	IgG1	21 days	3 months	3 months
	Avelumab74	IgG1	6 days	1 month	1 month
CTLA-4	Ipilimumab75	IgG1	15 days	3 months	3 months

FDA, Food and Drug Administration; ICIs, immune checkpoint inhibitors.

Once pregnancy is established, management should focus on maternal safety, fetal growth and the early detection of placental dysfunction. Management should also consider the timing of cancer diagnosis in pregnancy, as this may influence decisions regarding continuation of pregnancy, including termination or induction of delivery in cases of advanced maternal disease or poor prognosis, balanced against maternal–fetal risk. In cases of favourable treatment response, temporary interruption of ICI therapy with close maternal and fetal monitoring may be considered to reduce fetal exposure. Prophylactic maternal anticoagulation may be appropriate as pregnancy and malignancy increase the risk of venous thromboembolism.⁵⁹ When cancer treatment continues during pregnancy, obstetrical care ideally will take place in a centre with expertise in managing complex pregnancies.⁶⁰ Fetal ultrasound can track fetal growth, amniotic fluid volume, placental features and markers of fetal well-being, helping to detect intrauterine growth restriction, placental dysfunction or other complications.⁶⁰ Postpartum histological assessment of the placenta is clinically important as placental involvement has been documented in prior reports. When placental metastases are identified, it is recommended that the neonate undergo a complete skin examination, abdominal ultrasound and blood testing to assess for possible neonatal disease.^61,62

An additional consideration is the development of maternal irAEs during ICI therapy in pregnancy, which may require systemic immunosuppression, most commonly corticosteroids. These interventions, alongside the irAEs themselves, may carry potential maternal and fetal risks, although data in pregnancy remain limited.

ICI exposure during lactation and reproductive safety guidance

Evidence regarding ICI use during breastfeeding remains limited. Although full-length IgG molecules are generally present in breast milk at low levels, the extent of their transfer and the amount absorbed by the infant’s gastrointestinal tract are not well understood.¹⁰ One published report found ipilimumab present in breast milk without immediate adverse effects on the neonate; in this case, breastfeeding was initiated 3 weeks after the final dose, earlier than currently recommended guidelines.⁶³ However, due to the long half-lives of ICIs and FcRn-mediated recycling, functional antibody activity may continue for months after completing therapy. Receptor occupancy could also vary with tumour burden and FcRn polymorphisms, potentially prolonging biological effects even as serum levels decline.⁶⁴ In view of these uncertainties, major oncology guidelines, including those from the National Comprehensive Cancer Network, advise against breastfeeding during ICI treatment and for a minimum of 5 months after the final dose of immunotherapy.⁶⁵ Furthermore, recommendations from the European Society for Medical Oncology advise that individuals of reproductive potential use dual contraceptive methods for 5 months following cessation of therapy with ICIs.⁶⁶

Although these recommendations are broadly consistent in their precautionary approach to ICI exposure in pregnancy and lactation, the specific recommended duration of contraception and breastfeeding avoidance varies slightly between agencies, reflecting differences in regional regulatory frameworks and interpretation of limited reproductive safety data.

Pregnancies planned after completion of ICI therapy

Whenever possible, individuals planning pregnancy after ICI therapy should consult maternal–fetal medicine and pharmacy specialists for preconception advice. These discussions should cover the limited human safety data and the appropriate timing for conception. Prenatal genetic counselling may be offered when needed.

Ethical considerations

The care of pregnant patients with cancer also raises ethical considerations. Across the ethics literature, patient autonomy is the primary guiding principle, placing the patient’s values, priorities and understanding of risk at the centre of decision-making.⁶⁷ Shared decision-making requires open communication that reflects the patient’s perspective rather than the clinician’s assumptions. Beneficence towards both the pregnant individual and the fetus remains a core professional obligation.⁶⁸ Fetal well-being is closely linked to maternal well-being, which is important when treatment delays or changes may harm maternal prognosis. Ethical tension may arise when a patient’s choices differ from what clinicians consider optimal for their health or the health of the fetus. Access to psychosocial and spiritual care can help patients navigate these choices and support decisions that are both ethically sound and consistent with their values.

Future directions and conclusions

The evidence supporting the use of ICIs in pregnancy is still very limited. There is a lack of prospective studies, standardised reporting and long-term follow-up restricting our understanding of true maternal, obstetric and neonatal risks. Much of what has been reported is based on a limited number of case reports, small case series, as well as retrospective pharmacovigilance data. Pregnant individuals remain excluded from clinical trials, and current knowledge is based on post hoc observations rather than systematic investigation.

Strengthening the evidence base will require coordinated international efforts. Pregnancy exposure registries with harmonised data elements, timely case reporting and structured neonatal follow-up would improve the reliability of safety assessments. Detailed documentation of treatment timing, placental pathology, neonatal immune function and long-term developmental outcomes is essential, especially given the risk of delayed immune effects after in utero exposure.

To date, most case reports and case series describing gestational exposure to ICIs have documented favourable short-term outcomes. However, some case reports also describe adverse outcomes including intrauterine growth restriction, preterm birth and rare instances of neonatal irAEs. These findings highlight the importance of adopting a cautious and individually tailored approach when considering the administration of ICIs to this population. Until more robust evidence is available to guide such decisions, use of ICIs in should be limited to those cases in which the anticipated benefits to the pregnant patient clearly outweigh the potential risks to the fetus. Decisions regarding whether to administer ICIs to pregnant patients should be made by a multidisciplinary team that includes input from the patient and incorporates their values and clinical context.

Data availability statement

Data sharing not applicable as no datasets generated and/or analysed for this study.