Abstract
Background/Aim: Pembrolizumab, an immune checkpoint inhibitor used in cancer therapy, is associated with immune-related adverse effects, yet its impact on female reproductive health remains unclear. This study aimed to evaluate the effects of pembrolizumab on ovarian function and hormonal balance in a mouse model.
Materials and Methods: Twenty-four Swiss albino mice were divided into acute and chronic groups, each comprising a pembrolizumab-treated and a saline-control subgroup (n=6 per subgroup). All mice underwent a superovulation protocol, and serum levels of estradiol, follicle-stimulating hormone (FSH), luteinizing hormone (LH), and anti-Müllerian hormone (AMH) were measured using enzyme-linked immunosorbent assay kits. Ovarian tissues were evaluated histologically for follicular counts, including primordial, preantral, and secondary follicles.
Results: Estradiol levels were significantly reduced in pembrolizumab-treated mice compared with controls (p=0.007). Significantly higher levels of LH were observed in the chronic control group compared to the chronic pembrolizumab-treated group (p=0.023), while FSH and AMH levels did not differ significantly (p=0.461 and p=0.460, respectively). Primordial follicle counts were significantly higher in control groups than in pembrolizumab-treated groups (p=0.009).
Conclusion: Pembrolizumab administration in mice led to reduced estradiol levels and diminished primordial follicle counts, indicating potential adverse effects on ovarian reserve. These findings highlight the importance of evaluating reproductive risks in female patients receiving immunotherapy.
- Pembrolizumab
- immune checkpoint inhibitors
- ovarian reserve
- estradiol
- reproductive toxicity
- female fertility
- preclinical study
Introduction
Immunotherapies have emerged as fundamental elements in modern oncology treatment, dramatically altering the treatment landscape by increasing overall survival rates and generally inducing fewer side-effects than traditional chemotherapy does. These therapies employ monoclonal antibodies to target critical immune checkpoint regulators, such as programmed cell death protein 1 (PD1), its ligand PD-L1, and cytotoxic T-lymphocyte-associated antigen 4 (CTLA4). These checkpoints normally act to reduce immune responses and promote immune tolerance (1). By blocking PD1, PD-L1, or CTLA4, checkpoint inhibitors increase the activation and proliferation of tumor-specific T-cells, thereby strengthening the body’s antitumor immune response (1). Additionally, these therapies increase the levels of proinflammatory cytokines, including tumor necrosis factor-α (TNF-α), interferon-γ, and interleukin-1β, further intensifying the immune assault on tumors (2).
Initially employed predominantly for palliative care in older adults, immunotherapies are now being used for curative purposes in both neoadjuvant and adjuvant settings across diverse age groups and cancer stages (3, 4). This broadened application is particularly significant for adolescents and young adults with cancer, where long-term outcomes and quality of life are of utmost importance. Although the effects of immune checkpoint inhibitors (ICIs) on the male pubertal system are relatively well documented (5, 6), data on their impacts on the female reproductive system remain sparse and largely inconclusive.
Primary ovarian insufficiency may result from various etiologies, including metabolic, genetic, or presumed autoimmune causes (7). Given the immune-modulating effects of ICIs, there is growing concern about their potential to disrupt ovarian function. Preclinical data by Winship et al. demonstrated that ICI exposure in mice leads to immune cell infiltration, increased ovarian TNF-α expression, depletion of the follicular reserve, and impaired ovulation (8). In parallel, clinical findings from the North American Intergroup trial E1609, led by the ECOG-ACRIN Cancer Research Group, reported a significant decline in anti-Müllerian hormone (AMH) levels following ICI therapy, suggesting a possible impact on female fertility (9). These observations highlight the need for dedicated research to clarify and prevent the reproductive consequences of ICI treatment. To understand these effects, it is critical to examine key reproductive parameters, including serum estradiol, follicle-stimulating hormone (FSH), luteinizing hormone (LH), and AMH levels, along with follicle counts. Estradiol plays a vital role in the regulation of the menstrual cycle and ovulation. FSH and LH are essential for follicular development and ovulation, while AMH serves as a marker of ovarian reserve and reflects the remaining follicular pool (10).
The necessity for this study stems from the limited and inconsistent evidence regarding the reproductive effects of ICIs. This study specifically aimed to investigate the acute and chronic impacts of pembrolizumab on ovarian function using a mouse model. By evaluating histological changes in ovarian tissue and alterations in serum reproductive hormones, we aimed to elucidate the reproductive risks associated with immunotherapy in female patients with cancer and guide future fertility-preserving strategies.
Materials and Methods
This study received approval from the Gazi University Animal Experiments Local Ethics Committee (Approval number: G.Ü.ET-22.102, DATE: 26.10.2022), ensuring that all experimental procedures were conducted in accordance with established ethical standards for animal research.
In this study, we utilized 6- to 8-week-old Swiss albino female mice weighing 35-40 g. The subjects were divided into four groups to evaluate the acute and chronic side-effects associated with immunotherapy. Each group consisted of six mice, for a total of 24 mice for the study. The animals were housed in the Gazi University Animal Laboratory with ad libitum access to food and water. The environmental conditions were controlled, maintaining a temperature of 22±2°C, humidity of 50±10%, and a consistent 12/12-h light/dark cycle.
The mice were randomly assigned into four experimental groups to assess both acute and chronic effects of pembrolizumab on ovarian function:
Group 1 – Acute control group (n=6): This group served as the control for assessing acute effects. Mice received two intraperitoneal injections of sterile saline, administered on days 0 and 7. The injection volume was matched to that used for pembrolizumab administration in group 2.
Group 2 – Acute treatment group (n=6): Designed to evaluate the acute immunological effects of pembrolizumab, mice in this group were administered 5 mg/kg of pembrolizumab (Keytruda®; Merck & Co., Inc., Kenilworth, NJ, USA) intraperitoneally on day 0, followed by a second dose of the same quantity on day 7.
Group 3 – Chronic control group (n=6): This group functioned as the control for chronic exposure. Mice received intraperitoneal saline injections once weekly for four consecutive weeks (days 0, 7, 14, and 21), with volumes equivalent to those used in group 4.
Group 4 – Chronic treatment group (n=6): To assess the chronic impact of pembrolizumab, mice in this group received 5 mg/kg pembrolizumab intraperitoneally on day 0, followed by the same dose weekly for three additional doses on days 7, 14, and 21, totaling four doses over 4 weeks.
The mice underwent superovulation 5 days following the final treatment dose. For superovulation induction, the mice were administered 10 IU of pregnant mare serum gonadotropin (PMSG; Folligon®; Intervet International B.V., Boxmeer, the Netherlands) intraperitoneally (11, 12). Forty-eight hours after PMSG injection, 5 IU of human chorionic gonadotropin (hCG) were administered intraperitoneally. At 12-14 h after hCG injection, animals were deeply anesthetized with intraperitoneal ketamine (Alfamine® Vilsan Veteriner İlaçları, Ankara, Türkiye) and xylazine (Rompun®; Bayer Türk Kimya San. Ltd. Şti., Istanbul, Türkiye), and euthanasia was performed via intracardiac exsanguination under veterinary supervision, in accordance with approval by the local Ethics Committee (Figure 1). Additionally, ovaries were extracted for histopathological evaluation and preserved in formaldehyde solution. Levels of FSH, LH, estradiol and AMH were measured in the serum samples. Since all mice underwent superovulation with PMSG and hCG, estrous cycle staging was not monitored, in accordance with previous reports demonstrating that exogenous gonadotropins override endogenous cycle-dependent hormonal variations (13).
Schematic representation of the experimental design showing four groups of Swiss albino female mice: Acute control (group 1), acute immunotherapy (group 2), chronic control (group 3), and chronic immunotherapy (group 4). Mice received intraperitoneal injections of either saline or pembrolizumab on specified days (days 0 and 7 for acute groups; days 0, 7, 14, and 21 for chronic groups). Superovulation was induced using pregnant mare serum gonadotropin (PMSG; 10 IU) followed by human chorionic gonadotropin (hCG; 5 IU). Euthanasia was performed on day 15 (acute groups) or day 29 (chronic groups). IU: International unit.
Measurement of estradiol, LH, FSH and AMH levels in serum samples. Blood samples were collected from the study groups via yellow-capped gel separator tubes and centrifuged at 3500 rpm (approximately 1,715 ×g) for 10 min to separate the serum. The obtained serum samples were then stored at −80°C until the day of analysis. The levels of estradiol, LH, FSH, and AMH in the serum samples were measured using a commercial mouse sandwich enzyme-linked immunosorbent assay kit (CK-bio-16118; Coon Koon Biotech, Shanghai, PR China) according to the manufacturer’s instructions. Absorbance readings were performed using a BioTek ELx800 microplate reader and washer (BioTek Instruments, Winooski, VT, USA). Each sample was assayed in duplicate, and the mean of the two readings was used for statistical analysis. Standard curves were constructed based on the absorbance of standard samples, and hormone concentrations were calculated by matching the observed absorbance values with those of the curve.
Pathology. A total of 24 ovarian excisional biopsies were initially fixed in 10% formaldehyde solution, entirely sampled, processed via standard tissue processing protocols, embedded in paraffin, sectioned at a thickness of 4 μm, and stained with hematoxylin and eosin. The prepared slides were evaluated by two pathologists to assess their histological adequacy, and all the samples were deemed suitable for analysis.
Follicular classification was performed on the basis of morphological criteria previously described by Pedersen and Peters (14). Accordingly, the follicles were categorized as follows:
Primordial follicle: A single layer of flattened or a combination of flattened and cuboidal granulosa cells surrounding the oocyte.
Primary follicle: A complete layer of cuboidal granulosa cells encircling the oocyte.
Secondary follicle: Two or more layers of cuboidal granulosa cells with no evidence of antrum formation.
Antral follicle: Multiple layers of granulosa cells with a discernible antral space.
For each ovary, the follicles were classified and counted according to these criteria using a light microscope (Olympus CX31; Olympus Corporation, Tokyo, Japan) at ×100 and ×400 magnification, and the average follicle count was calculated for each mouse.
Statistical analysis. Statistical analyses were performed using IBM SPSS Statistics for Windows, Version 21.0 (IBM Corp., Armonk, NY, USA). The normality of the data distribution for each variable (estradiol, FSH, LH, AMH, and follicular counts) was assessed using the Shapiro-Wilk test. Since all data were found to be normally distributed, results are presented as the mean±standard deviation, and parametric statistical methods were applied.
Comparisons among the four experimental groups, namely the acute control, acute pembrolizumab-treated, chronic control, and chronic pembrolizumab-treated, were conducted using one-way analysis of variance. When significant differences were detected, Tukey’s honestly significant difference test was used for post hoc pairwise comparisons. All statistical tests were two-sided, and a p-value less than 0.05 was considered statistically significant.
Time-dependent differences were not evaluated within the same subjects, as each group was composed of independent animals exposed to different treatment durations (acute vs. chronic). Therefore, no repeated-measures or longitudinal analyses were performed. Between-group comparisons were used to assess the impact of treatment duration.
As there were no prior studies investigating ovarian toxicity associated with pembrolizumab or other ICIs, it was not possible to perform a power analysis. Furthermore, the local Ethics Committee restricted the use of animals to a maximum of six per group; therefore, the study was designed and completed with six rats in each experimental group.
Results
None of the mice died before the termination of the study, and the study was completed in accordance with the protocol.
Biochemical analysis. Figure 2 illustrates the distribution of serum hormone levels among the four study groups. Figure 2A demonstrates a visible reduction in estradiol levels in the pembrolizumab-treated groups compared to their respective controls. Figure 2B and C show that FSH levels appear relatively stable across the groups, LH levels noticeably increased in the chronic treatment group (group 4) compared to its control counterpart (group 3). Figure 2D suggests a modest reduction in AMH levels in the treatment groups.
Serum hormone levels across study groups. (A) Serum estradiol (E2). (B) Follicle-stimulating hormone (FSH). (C) Luteinizing hormone (LH). (D) Anti-Müllerian hormone (AMH). Data are presented as the mean±standard deviation. Statistical analysis was performed using one-way analysis of variance followed by Tukey’s post hoc test. Statistically significant comparisons are indicated.
The comparative analysis of serum hormone levels among the study groups is presented in Table I. Estradiol levels significantly differed across the four groups (p=0.007), with significantly lower estradiol levels in pembrolizumab-treated mice compared to their respective controls, particularly between the acute control (group 1) and acute treatment (group 2) groups (p=0.015). Post hoc analysis revealed a significant reduction in estradiol levels in group 2 compared to group 1 (p=0.015), while no significant difference was noted between the chronic treatment and control groups (group 4 vs. group 3; p=0.413). LH levels also differed significantly among the groups (p=0.038); pairwise comparisons showed a significant reduction in LH levels in group 4 compared to group 3 (p=0.023), whereas no significant difference was found between the acute groups (group 2 vs. group 1; p=0.962).
Comparison of serum levels of estradiol (E2), follicle-stimulating hormone (FSH), luteinizing hormone (LH), and anti-Müllerian hormone (AMH) between control and pembrolizumab-treated groups.
No significant group differences were observed in FSH and AMH levels (p=0.461 and p=0.460, respectively). Additionally, no statistically significant differences in estradiol, LH, FSH, or AMH levels were found between the acute and chronic pembrolizumab-treated groups (group 2 vs. group 4).
Histological analysis. Histological evaluation of ovarian tissue samples from all study groups was conducted to assess follicle counts, as illustrated in Figure 3.
Histopathological sections of excisional ovarian biopsies stained with hematoxylin and eosin. (A) High-magnification view of representative ovarian tissue showing various stages of follicular development. Arrow: oocyte; circles: primordial and primary follicles; stars: secondary follicles. (B) Low-magnification view of an ovary from an acute control mouse (group 1), showing preserved follicular architecture. (C) Low-magnification view of ovarian tissue from a chronically pembrolizumab-treated mouse (group 4), showing reduced follicle density. 4 All images were taken using a light microscope at ×100 (A) and ×40 (B, C) magnification. Scale bars represent 50 μm.
The control groups demonstrated a mean primordial follicle count of 6±1, whereas the pembrolizumab-treated groups exhibited a significantly lower count, averaging 3±2 (p =0.009). Further subgroup analyses revealed a statistically significant reduction in the number of primordial follicles in both the acute treatment group compared to its control (p =0.023) and the chronic treatment group compared to its respective control (p=0.002). However, no significant differences were detected among the groups regarding the number of preantral follicles, secondary follicles, or the total follicle count (p>0.05). The methodology and numerical findings related to follicle counts are detailed in Table II.
Counts of follicular stages in control and treatment groups.
Discussion
In our study, acute and chronic use of pembrolizumab caused a decrease in the number of primordial follicles and in estrogen level in this mouse model. However, other markers, such as FSH and AMH, did not significantly change. These findings suggest that while immunotherapy can affect certain reproductive and hormonal functions, its effects on other endocrine parameters remain unchanged. This differential impact underscores the need for targeted monitoring and management strategies in patients undergoing such treatments.
In this study, both acute and chronic administration of pembrolizumab resulted in a significant reduction in the number of primordial follicles. This finding aligns with prior preclinical research demonstrating the deleterious impact of immune checkpoint blockade on ovarian reserve (15). Notably, the observed reduction in primordial follicle counts occurred in the absence of significant alterations in FSH or AMH levels, suggesting an early-stage ovarian insult that may not yet be reflected in gonadotropin feedback mechanisms. Primordial follicles constitute the non-renewable oocyte pool, and their depletion is a critical event associated with diminished reproductive potential, irreversible infertility, and the risk of premature ovarian insufficiency (16). Our data imply that pembrolizumab-induced ovarian inflammation may trigger follicular apoptosis, consistent with previously described mechanisms involving proinflammatory cytokine upregulation and caspase activation following ICI therapy (8, 17). The concomitant decline in estradiol level observed in both treatment arms further supports the hypothesis of impaired steroidogenesis secondary to follicular damage.
FSH, secreted by the anterior pituitary, is essential for the growth and maturation of antral follicles, whereas the development of primordial and early preantral follicles is largely FSH-independent (16, 18). Consistent with this, our study demonstrated no significant differences in serum FSH levels among the study groups, and primordial follicle depletion occurred despite equivalent FSH exposure. The absence of a decrease in FSH levels also suggests that pembrolizumab did not induce hypophysitis or secondary hypopituitarism in the treated animals. Nevertheless, ovaries from pembrolizumab-exposed mice exhibited impaired estrogen production, most likely as a consequence of local inflammatory processes. Prior preclinical studies have reported similar outcomes, showing that immune checkpoint blockade can increase ovarian apoptosis and reduce ovulatory capacity, evidenced by reduced corpus luteum formation (8).
AMH is secreted exclusively by granulosa cells of preantral and small antral follicles, and its levels are considered a reliable marker of ovarian reserve (19, 20). In our study, AMH levels remained unchanged across all groups, which is consistent with the absence of significant differences in preantral follicle counts. These findings suggest that pembrolizumab did not impair the ovarian reserve as reflected by AMH in Swiss albino mice.
Interestingly, while there is limited preclinical evidence indicating a reduction in AMH in murine models, clinical data have shown AMH declines in patients receiving ICIs. Notably, the ECOG-ACRIN E1609 trial reported a significant reduction in AMH levels among female patients receiving adjuvant nivolumab and ipilimumab (9). Therefore, it may be useful to routinely monitor AMH in patients with cancer treated with immunotherapy who are expecting future fertility.
Immunotherapy increases systemic lymphocyte and cytokine release, which increases antitumor activity, as well as autoimmune side-effects (21). Previous study has shown that ovaries express PD-L1, and the hypothesis that immunotherapy agents may cause direct ovarian toxicity in addition to systemic side-effects has been proposed (22). In the absence of direct cytokine or apoptotic marker evaluation in our model, we refer to previous studies that have explored the immunopathogenic effects of checkpoint inhibitors on gonadal tissues. Winship et al. reported that anti-PD1 and anti-PD-L1 therapies increased ovarian TNF-α expression and immune cell infiltration, leading to heightened apoptotic activity in ovarian follicles via caspase-3 activation (8). Similarly, Türkmen et al. demonstrated that pembrolizumab exposure in rats led to increased oxidative stress, reduced antioxidant defense, and testicular damage, along with elevated apoptosis in spermatogenic cells (23). These findings suggest that the depletion of primordial follicles observed in our model may be mediated by localized inflammation and immune-induced apoptosis. While we did not perform immunohistochemical analysis for apoptosis or immune infiltrates, the histological loss of follicular structures and decline in estrogen may indirectly reflect these underlying mechanisms. Future studies including caspase-3 immunostaining and cytokine profiling (e.g. TNF-α, interleukin-6, interferon-γ) will be critical to confirm these mechanistic pathways.
Our study is one of the few investigations into the impacts of ICIs on the female reproductive system. However, there are several limitations to consider. One of the primary limitations of our study is the use of a humanized PD1 monoclonal antibody, pembrolizumab, in non-humanized mice. Given the structural differences in PD1 between humans and mice, the pharmacodynamic effects of pembrolizumab may not fully replicate its mechanism of action in clinical settings. Nonetheless, prior experimental studies have demonstrated that pembrolizumab can elicit both local and systemic immune-related toxicities in non-humanized rodents. For instance, Türkmen et al. reported significant testicular damage and immunological changes in Sprague-Dawley rats treated with pembrolizumab, highlighting the drug’s capacity to induce measurable biological effects in rodents despite species differences in PD1 structure (23). Similarly, El-Haroun et al. observed corneal toxicity in albino rats following systemic pembrolizumab administration, supporting the immunomodulatory impact of this compound in non-humanized models (24). These findings suggest that although murine PD1 may not be the optimal target for pembrolizumab, the observed histological and biochemical alterations still provide valuable insights into its potential off-target and systemic effects. Nevertheless, future studies using either humanized mouse models or murine-specific PD1 inhibitors would enhance the translational validity of our findings. One methodological consideration is the absence of estrous cycle monitoring in our experimental design. However, this was an intentional decision, as all mice were subjected to a standardized superovulation protocol using PMSG and hCG. This pharmacological approach induces a synchronized ovulatory response and functionally overrides the natural hormonal fluctuations of the estrous cycle. In alignment with previously published data, exogenous gonadotropins at appropriate doses have been shown to override the stage-specific effects of endogenous hormones, rendering estrous cycle staging unnecessary under superovulated conditions (13). Another important limitation of our study is the absence of a formal power analysis. Since as far as we are aware there are no previous studies evaluating the ovarian toxicity of pembrolizumab or other ICIs, we had no basis for reliably estimating effect sizes for endpoints such as hormone levels or follicular counts.
Future studies should explore the long-term reproductive consequences of immune checkpoint inhibitors, including their effects on fertility and hormonal regulation in different preclinical models and human subjects. Investigating the reversibility of ovarian damage and the potential protective strategies, such as fertility preservation techniques, will be crucial, especially for young female patients with cancer. Furthermore, mechanistic studies elucidating the role of local ovarian immune responses and cytokine-mediated pathways in checkpoint inhibitor-induced ovarian dysfunction are warranted. Establishing standardized reproductive monitoring protocols during immunotherapy may significantly contribute to oncofertility care.
This experimental study in a murine model demonstrates that pembrolizumab, an ICI, may adversely affect ovarian reserve by significantly reducing primordial follicle counts and estradiol levels. While levels of FSH and AMH remained stable, the observed hormonal and histological alterations suggest potential reproductive risks associated with immunotherapy. These findings emphasize the importance of reproductive health monitoring in female patients receiving ICIs. Our study contributes to a growing body of evidence that necessitates further preclinical and clinical investigations to guide fertility preservation strategies and individualized patient counseling in the era of immunotherapy.
Footnotes
Authors’ Contributions
Bengü Mutlu Sütcüoğlu: Conceptualization (equal), data curation, formal analysis, methodology, resources, software, validation, writing – original draft. Osman Sütcüoğlu: Conceptualization, data curation, methodology, project administration, resources, supervision, writing – review and editing. Betül Öğüt: Conceptualization, data curation, formal analysis, methodology, writing – original draft. Orhun Akdoğan: Conceptualization, data curation, formal analysis, methodology, writing – original draft. Cagnur Elpen Kodaz: Conceptualization, data curation, formal analysis, methodology, writing – original draft. Canan Yılmaz: Conceptualization, data curation, formal analysis, methodology, writing – original draft. Özlem Erdem: Conceptualization, data curation, formal analysis, methodology, writing – original draft. Elvan Anadol: Conceptualization, data curation, formal analysis, methodology, writing – original draft. Nuriye Özdemir: Conceptualization, data curation, formal analysis, methodology, writing – original draft. All Authors reviewed and approved the final version of the manuscript for submission.
Conflicts of Interest
None.
Funding
This study was carried out as part of the Project Support program of the Turkish Society of Medical Oncology.
Artificial Intelligence (AI) Disclosure
No artificial intelligence (AI) tools, including large language models or machine learning software, were used in the preparation, analysis, or presentation of this manuscript.
- Received August 24, 2025.
- Revision received September 28, 2025.
- Accepted October 8, 2025.
- Copyright © 2026 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).









