Immune checkpoint inhibitor treatment does not impair ovarian or endocrine function in a mouse model of triple negative breast cancer

27

Representing 15 -20% of all breast cancer cases, triple negative breast cancer 28 (TNBC) is diagnosed more frequently in reproductive -age women and exhibits higher rates of 29 disease metastasis and recurrence when compared with other subtypes. Few targeted treatments 30 exist for TNBC, and many patients experience infertility and endocrine disruption as a result of 31 frontline chemotherapy treatment. While they are a promising option for less toxic therapeutic 32 approaches, little is known about the effects of immune checkpoint inhibitors on reproductive and 33 endocrine function. 34

Our findings in a syngeneic TNBC mouse model revealed that therapeutically relevant 35 immunotherapies targeting PD-1, LAG-3, and TIM-3 had no effect on the quality and abundance 36 of ovarian follicles, estrus cyclicity, or hormonal homeostasis. Similarly, in a tumor -free mouse 37 model, we found that ovarian architecture, follicle abundance, estrus cyclicity, and ovulatory 38 efficiency remain unchanged by PD-1 blockade. 39

Taken together, our results suggest that immunotherapy may be a promising 40 component of fertility -sparing therapeutic regimens for patients that wish to retain ovarian and 41 endocrine function after cancer treatment. 42 43

44 breast cancer, TNBC, PD-1 inhibitor, immune checkpoint inhibition, oncofertility, ovarian reserve 45 46 47 48 49 50 51 52 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint 3

53 Representing 15-20% of all breast cancer cases, triple negative breast cancer (TNBC) is 54 an aggressive form of breast cancer that is diagnosed more frequently in reproductive-age women 55 than other sub -types1,2. Because TNBC lacks expression of the estrogen receptor (ER), 56 progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2), it is 57 unresponsive to many of the targeted treatments that are used for other subtypes of breast 58 cancer3,4. Thus, cytotoxic chemotherapies, which are associated with several severe systemic 59 side effects, are often key components of frontline treatments 5. Indeed, these side effects can 60 impact the reproductive system, and many as half of women who receive conventional 61 chemotherapies experience reproductive and endocrine dysfunction as a result6. However, recent 62 advances in cancer immunotherapy have led to the rapid integration of Pembrolizumab, a 63 programmed cell death protein 1 (PD -1) inhibitor, into standard -of-care treatment regimens for 64 TNBC, which may allow for reduced toxicity during treatment7,8. 65 The development of immunotherapies has shown great promise for the targeted treatment 66 of a variety of cancers, and they lack many of the side effects observed with standard of care 67 chemotherapeutics9. One such class of immunotherapies that have seen clinical success is 68 immune checkpoint inhibitors, which target immune checkpoint regulators such as PD-1 and its 69 ligand PD-L15. These ligands act as modulators in normal tissue to promote self-tolerance but are 70 overexpressed in tumor cells as a mode of evading immunosuppression 10. Immune checkpoint 71 inhibitors such Pembrolizumab employ monoclonal antibodies to block immune checkpoint 72 interactions with the goal of activating tumor-specific cytotoxic T cells and promoting immune cell-73 mediated killing4. 74 Though the introduction of immune checkpoint inhibitors has revolutionized cancer 75 treatment paradigms and patient quality of life, the systemic blocking of immune checkpoint 76 interactions can cause immune-related adverse events (irAEs) associated with a lack of immune 77 tolerance of self -tissues11. While these effects are not as common as those associated with 78 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint 4 cytotoxic chemotherapy and are usually mild, they can occasionally be severe and may affect a 79 variety of organs systems 9. Indeed, endocrine irAEs are some of the most commonly reported 80 irAEs in the clinic and include hyper- and hypothyroidism, hypophysitis, hypogonadism, and Type 81 I diabetes 12,13. In a 2022 study by Winship et al., anti -CTLA-4 and anti -PD-L1 monoclonal 82 antibodies were associated with an increase in intra -ovarian immune cells and tumor necrosis 83 factor-a (TNF-a) expression, as well as a decrease in ovarian follicle quality and abundance 14. 84 However, no studies have evaluated the ovarian and endocrine effects of standard-of-care PD-1 85 inhibitors or blockade of other exploratory targets such as lymphocyte-activation gene 3 (LAG-3) 86 or T cell immunoglobulin and mucin domain 3 (TIM-3). 87 In the mammalian ovary, immature oocytes are stored in a quiescent state as primordial 88 follicles in a finite population known as the ovarian reserve 15,16. After puberty, these primordial 89 follicles are continuously activated to begin folliculogenesis, the process of transforming into 90 larger, mature follicles that produce steroid hormones and may eventually be ovulated 17,18. This 91 process occurs continuously through the entire reproductive lifespan until the ovarian reserve is 92 depleted, at which point menopause begins19,20. Because new oocytes cannot be generated after 93 birth, it is critical that primordial follicles must be available in sufficient amounts and be maintained 94 through the reproductive lifespan of the host to ensure fertility and endocrine function 21,22. 95 Cytotoxic chemotherapies are among the foremost causes of ovarian reserve damage, often 96 resulting in the condition of Primary Ovarian Insufficiency (POI) 6,23. Caused by the depletion of 97 the ovarian reserve, POI can lead to infertility and impaired endocrine function, and increases the 98 risk of conditions associated with aging such as heart attack, stroke, and osteoporosis24. 99 In all mammals, ovarian follicles undergo highly -conserved processes of growth, 100 maturation, ovulation, and death, although the timeline of these events is species-specific16,17. In 101 humans, the time for an activated primordial follicle to mature into a fully -grown pre-ovulatory 102 follicle takes about 12 months, while in mice, this process takes around 21 days 25,26. The 103 menstrual cycle in humans takes an average of 28 days and typically results in the ovulation of a 104 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint 5 single oocyte per cycle, while the mouse estrous cycle takes only 4 -5 days and results in the 105 ovulation of several oocytes per cycle 19,26. The mouse reproductive cycle begins around 6 -8 106 weeks of age and ends approximately after 12 months26,27. In humans, the menstrual cycle begins 107 at the onset of puberty and continues until menopause 16,19. Though mice do not undergo a true 108 “menopause”, they experience similarities to human females in the processes of ovarian reserve 109 depletion, loss of fertility, and endocrine dysfunction with aging, and are therefore tremendously 110 useful models for study of mammalian reproductive function28. 111 Immune regulation plays an important role in reproductive and endocrine health. The 112 ovary is subject to immune cell infiltration as it is a highly vascularized, non -immunoprivileged 113 organ29. In the ovary, immune cells are critical in processes related to granulosa cell turnover, 114 ovulation, clearing atretic follicles, and the development and breakdown of the corpus luteum30,31. 115 In addition, signaling by cytokines such as TNF-a and transforming growth factor-b (TGF-b) are 116 crucial in follicle maturation and ovulation 20. However, it is likely that dysregulation or over -117 activation of these immune factors could cause damage to the ovary 32–34. A few epidemiologic 118 studies have reported a higher incidence of POI and unexplained infertility in women with chronic 119 inflammatory or autoimmune conditions such as Crohn’s Disease and psoriasis35,36. In addition, 120 there is some evidence linking inflammation and autoimmunity on ovarian aging and follicle 121 depletion in animal models34. However, the specific mechanisms of autoimmune depletion of the 122 ovarian reserve remain unknown, and it is unclear whether immune checkpoint inhibitor treatment 123 creates a sufficiently heightened systemic immune response to elicit ovarian damage and 124 endocrine disruption. 125 The ovarian toxicity of many of the frontline chemotherapeutics for TNBC has been well -126 characterized, with the non-renewable population of oocytes being some of the most vulnerable 127 cells to damage 6,23. As immune checkpoint inhibitors continue to be incorporated into clinical 128 practice, more studies on their reproductive and endocrine effects are necessary, especially if 129 they are to be used as a component of fertility -sparing treatment regimen. Moreover, given the 130 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint 6 fact that they are still relatively new to the cancer therapeutic repertoire, data on long-term fertility 131 outcomes in human TNBC patients treated with immune checkpoint inhibitors is not yet available 132 and thus preclinical models must be used to evaluate the ir ovarian impacts37. We hypothesized 133 that ICIs targeting PD -1, LAG-3 and TIM -3 would be relatively benign to ovarian reserve and 134 endocrine function. 135 136

137 Animals 138 Wild-type C57Bl/6 mice were obtained from the Jackson Laboratory (strain # 000664). All 139 animal protocols were approved by Brown University Institutional Animal Care and Use 140 Committee and were performed in accordance with the National Institutes of Health Guide for the 141 Care and Use of Laboratory Animals (# 22 -09-0002). All animal protocols were reviewed and 142 acknowledged by the Lifespan University Institutional Care and Use Committee (# 1987412-1). 143 144 E0771 tumor-bearing mouse model and tissue collection 145 Mouse E0771 cells were obtained from ATCC and cultured in DMEM, 10% FBS, and 146 penicillin/streptomycin. Cells were found to be free of pathogens and mycoplasma (per Charles 147 River pathogen testing). Eight-week-old female C57Bl/6J mice were injected with 100 μL of 1x105 148 E0771 cell suspension in Matrigel or saline control into the bilateral 4 th mammary pads under 149 isoflurane sedation. Once palpable 14 days later, a group of pre -treatment mice were collected, 150 and remaining mice were randomly allocated into study groups to begin treatment regimen of 151 immune checkpoint inhibitor monotherapy or control. Mice received 200 μg doses of mouse anti-152 PD-1 (clone: 29F.1A12), anti-LAG-3 (clone: C9B7W), anti-TIM-3 (clone: RMT3-23), or rat IgG2a 153 isotype control (clone: 2A3) every 4 days via intraperitoneal injection, with treatments stopping 154 after the third dose. All antibodies used for in-vivo treatments were purchased from BioXcell. 155 Doses were based on previously described tumor-reducing regimens45. Mice were monitored for 156 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint 7 14 days and then collected once they reached proestrus stage. Upon collection, tumors, ovaries, 157 and serum were obtained for further analysis. Tumor burden was determined by quantifying tumor 158 weight as a proportion of total mouse weight. 159 160 Non-tumor-bearing mouse model and tissue collection 161 Eight-week-old female C57Bl/6J mice were mock -injected with 100 μL saline in the 4 th 162 mammary pad under isoflurane sedation. 14 days later, mice were randomly allocated into study 163 groups to begin treatment with anti-PD-1 monoclonal antibody, IgG isotype control, or saline via 164 intraperitoneal injection. Mice in the anti -PD-1 and IgG isotype control group received 200 μg 165 doses of monoclonal antibodies every 4 days via intraperitoneal injection, with treatments 166 stopping after the third dose. Mice were monitored for 14 days and then collected once they 167 reached proestrus stage. Upon collection, ovaries and serum were obtained for further analysis. 168 169 Vaginal cytology and estrus cycle analysis 170 After the final dose of immunotherapy or control, estrus cycle stage was monitored daily 171 over the course of 14 days via vaginal smear cytology as previously described 46. Briefly, the 172 vagina of each mouse was flushed with saline and then mixed with toluidine blue O dye on a glass 173 slide, then classified into the different sub -stages of estrus based on the cell types visualized in 174 the sample. The percentage of time spent in each sub -stage of estrus was calculated for each 175 mouse and results were compared between treatment groups. 176 177 Ovarian histology and follicle quantification 178 Ovaries from tumor-bearing and non-tumor-bearing mice were stained and analyzed as 179 previously described 46. Briefly, ovaries were fixed in formalin and embedded in paraffin for 180 sectioning at 5 μm by the Brown Molecular Pathology Core, and every fourth slide was 181 deparaffinized and stained with hematoxylin and eosin (H&E). Five slides per ovary were 182 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint 8 quantified. Follicles were staged and counted on one section per every H&E -stained slide, and 183 these counts were normalized to section area to yield follicle density. 184 185 TUNEL staining 186 Prior to staining, ovarian FFPE slides were deparaffinized as previously described. Slides 187 were then washed for 3 minutes in Phosphate Buffered Saline (PBS), then permeabilized by 188 applying freshly prepared 20 μg/mL Proteinase K in 10mM Tris -HCl solution and incubating for 189 15 minutes in a humid chamber at room temperature. Slides were then washed 2x, 3 minutes in 190 PBS, 1x 3 minutes in PBS + 0.1% Triton (PBST), then again 2x, 3 minutes in PBS. Slides were 191 then stained with the In Situ Cell Death Detection Kit, Fluorescein (Roche) according to 192 manufacturer’s instructions. Briefly, slides were incubated with TUNEL reaction mixture for 1 hour 193 in a humid chamber at 37°C protected from light. Slides were then washed 3x for 3 minutes in 194 PBS, once for 3 minutes in DAPI/PBS solution, and then once for 5 minutes in PBS. Slides were 195 then mounted and analyzed on an EVOS M5000 Fluorescence Imaging System and images of 196 all fields of a single section were captured. 197 198 Serum hormone analysis 199 Whole blood from mice was collected via post -mortem cardiac puncture and serum 200 separated out as previously described 46. Serum was sent to the University of Virginia Ligand 201 Assay & Analysis Core for the Center for Research in Reproduction for quantification of serum 202 concentrations of AMH, LH, and FSH. 203 204 Superovulation of tumor-bearing and non-tumor-bearing mice 205 For superovulation experiments in tumor-bearing mice, eight-week-old wild-type C57Bl/6J 206 mice were orthotopically injected with E0771 cells into the right 4 th mammary pad as previously 207 described. Once palpable 14 days later, mice were randomly allocated into groups to receive 208 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint 9 intraperitoneal injections of anti-200 μg doses of PD-1 monoclonal antibody or IgG isotype control, 209 or 100 μL of saline control every 4 days with treatments stopping after the third dose. Eight days 210 after receiving their final dose, mice underwent ovarian hyperstimulation as previously 211 described46. Briefly, mice were injected with 5 -IU pregnant mare goat serum (PMSG; Prospec 212 Bio), and then 48 hours later, injected with 5 -IU human chorionic gonadotropin (HCG; Prospec 213 Bio). Twelve hours after HCG injection, mice were euthanized, and ovulated oocytes were 214 collected from the ampullae. Number of oocytes ovulated were counted per mouse and results 215 were compared between treatment groups. 216 217 For superovulation experiments in non -tumor-bearing mice, six -week-old wild -type 218 C57Bl/6J mice were mock-injected with saline into the mammary pad as described. As with the 219 tumor-bearing superovulation study, mice were randomly allocated into groups 14 days later to 220 receive intraperitoneal injections of anti-200 μg doses of PD-1 monoclonal antibody or IgG isotype 221 control, or 100 μL of saline control every 4 days with treatments stopping after the third dose. 222 Eight days after receiving their final dose, mice underwent ovarian hyperstimulation and ovulated 223 oocytes were quantified as described. 224 225 Statistical analysis 226 Ovarian area and follicle density were quantified from H&E -stained ovarian sections. 227 Follicles were counted by stage, and these counts were then normalized to the section area to 228 calculate oocyte density and account for size differences between different ovaries. For estrus 229 cycling analyses, the percentage of time spent in each substage was then averaged between all 230 animals in a treatment group. For superovulation studies, recovered oocytes were quantified and 231 averaged per treatment group, and mice were classified as “successfully stimulated” if the number 232 of retrieved oocytes met an age -adjusted threshold based on average litter sizes for wild -type 233 C57Bl/6J mice in our lab (7 oocytes for 8-week-old mice, 10 oocytes for 6-week-old mice). One-234 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint 10 way ANOVA with post -hoc Tukey’s tests for multiple comparisons were performed to evaluate 235 differences in tumor burden, follicle abundance, serum hormone levels, percentage of time spent 236 in estrus between treatment groups. 237 238

239 Inhibition of PD-1, LAG-3, and TIM-3 does not impact ovarian follicle abundance or quality in a 240 mouse model of triple negative breast cancer 241 To assess the effects of immune checkpoint inhibition on ovarian health and endocrine 242 function in a clinically-relevant model of TNBC, we collected ovaries and serum of C57Bl/6J mice 243 who had been orthotopically injected with syngeneic E0771 cells into the mammary pad, then 244 treated with intraperitoneal injections of anti-PD-1, anti-LAG-3, or anti-TIM-3 monotherapy. These 245 immunotherapies represent an array of immune checkpoint inhibitor candidates that have 246 demonstrated varying efficacy in clinical trials for solid cancers, with anti -PD-1 being the most 247 effective in TNBC populations 38–40. By choosing these specific targets, we were able to 248 comprehensively evaluate the effects of differential immune activation on the ovarian reserve and 249 hormonal homeostasis. Tumor-bearing control group mice received intraperitoneal injections of 250 IgG isotype at the same timepoints as immunotherapy-treated mice, and healthy control mice only 251 received a mock-injection of saline into the mammary pad at the time of orthotopic injections. To 252 control for cycle stage -dependent fluctuations in folliculogenesis, ovaries were collected during 253 the proestrus stage 14 -16 days following the final immunotherapy treatment . This collection 254 timepoint allowed us to capture any effects that may have a pattern of delayed onset seen in 255 some other adverse immune effects. 256 Morphologically, ovarian architecture was similar among all treatment groups (Fig 1a -f). 257 Follicles were classified by stage, including the immature, quiescent primordial follicles that make 258 up the ovarian reserve, the developing primary, secondary, and preantral follicles that have been 259 recruited and activated to undergo maturation, and the antral follicles that are nearly ready for 260 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint 11 ovulation. Degenerative follicles, which are undergoing atresia and dying, were also quantified 261 per section. At the post -treatment timepoint, anti -PD-1-treated mice showed almost complete 262 tumor regression, which was statistically significant when compared to the IgG isotype control 263 group (p=0.0125) (Fig 1f). Mice treated with anti-LAG-3 or anti-TIM-3 exhibited a more variable 264 response to treatment, with the anti-TIM-3-treated group achieving a significant reduction in tumor 265 burden (p=0.0492). Though the anti -LAG-3 group had a lower mean tumor burden than IgG 266 isotype controls, this difference was not statistically significant (p=0.8006). Ovarian area, as well 267 as overall and stage -specific oocyte densities were not significantly different between immune 268 checkpoint inhibitor-treated groups and IgG isotype-treated or saline mock-injected controls (Fig 269 1g-i). There were also no appreciable levels of oocyte or granulosa cell apoptosis found via 270 TUNEL staining (Supp Fig 1). These findings indicate that inhibition of a variety of immune 271 checkpoint interactions has no long -term effect on ovarian morphology or folliculogenesis in a 272 tumor-bearing system. 273 274 Inhibition of PD -1, LAG -3, or TIM-3 does not perturb endocrine homeostasis or reproductive 275 cyclicity 276 Given that long -term endocrine dysfunction and disruption of reproductive cycling are 277 common side effects of cancer therapy that are at times unrelated to ovarian reserve function, we 278 assessed these outcomes via serum hormone and estrus cycling analysis in the E0771 tumor -279 bearing mouse model. As with the ovarian collections, we collected serum during the proestrus 280 stage 14 -16 days following the final immunotherapy treatment to control for cycle stage -281 dependent fluctuations in hormone levels. Serum levels of luteinizi ng hormone (LH) and follicle 282 stimulating hormone (FSH) were quantified to evaluate possible disturbances in the hypothalamic-283 pituitary-gonadal axis and hormonal cycling. Serum levels of anti -Mullerian hormone (AMH), 284 produced by granulosa cells of maturing follicles and current clinical gold standard for evaluating 285 follicle abundance, was also quantified 41. Serum levels of LH and FSH levels did not differ 286 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint 12 significantly between any of the immunotherapy treatment and control groups (Fig 2a -b). 287 Likewise, there were no significant differences observed in LH:FSH ratio (Fig 2c), a common 288 clinical metric in which high values correlate with a lack of ovulation in polycystic ovarian 289 syndrome42. Consistent with our ovarian follicle density results, we found that AMH levels were 290 unchanged between immunotherapy-treated and control group animals. 291 Estrus cyclicity was analyzed by monitoring vaginal cytology daily for two weeks starting 292 the day after the final immunotherapy or IgG isotype control treatment. We found that there were 293 no significant differences in the amount of time spent in each of the cycle stages between any of 294 the immunotherapy treatment and control groups (Figure 2e). Taken together with the serum 295 hormone data, these findings suggest that immune checkpoint blockade does not impair 296 endocrine function or hormonal cyclicity in tumor-bearing mice. 297 298 Treatment with anti-PD-1 immunotherapy does not impact ovarian follicle density or reproductive 299 cyclicity in a non-tumor-bearing mouse model 300 To further investigate ovarian and endocrine function after treatment with anti -PD-1 301 immunotherapy and to disentangle any observed effects from tumor burden-specific phenomena, 302 we sought to validate our results in a tumor-free mouse model. To simulate orthotopic implantation 303 of tumor cells, young adult female C57Bl/6J mice age-matched to the E0771 cohort were mock-304 injected with saline into the mammary pad . Narrowing down our treatment focus to answer 305 questions of key clinical relevancy, we then administered anti-PD-1 mAb, IgG isotype control, or 306 saline control intraperitoneal injections, all at the same timepoints as the cohort of E0771 tumor-307 bearing mice. As with all tumor-bearing studies, ovaries and serum collected 14-16 days following 308 the final treatment when mice entered the proestrus stage. 309 Ovaries from all groups of non -tumor-bearing mice appeared morphologically similar to 310 each other (Fig 3a-c) and showed no differences in total area (Fig 3d). Upon quantifying ovarian 311 follicle abundance, anti-PD-1-treated mice did not differ significantly in oocyte and follicle stage 312 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint 13 densities compared to IgG isotype and saline -treated controls (Fig 3e -f). These findings 313 recapitulate our results from the tumor -bearing model and suggest that PD -1 blockade has no 314 negative effect on follicle abundance or folliculogenesis. 315 To evaluate any effects of anti-PD-1 immunotherapy on hormonal cyclicity in the absence 316 of tumor burden, we monitored estrus cycling via vaginal cytology for two weeks following 317 administration of the final immunotherapy or control treatment. Unsurprisingly, we found that the 318 percentage of time spent in each substage of estrus was unchanged between anti -PD-1-treated 319 animals and control animals treated with IgG isotype or saline (Fig 3g). These results are 320 consistent with the tumor-bearing model, suggesting that hormonal cyclicity is unaffected by PD-321 1 inhibition in both a tumor-bearing and non-tumor-bearing system. 322 Interestingly, the only differences observed in ovarian composition are found when 323 comparing treatment-matched groups from the tumor -bearing and non -tumor-bearing cohorts. 324 IgG isotype-treated animals from the tumor-bearing group have a statistically significant decrease 325 in ovarian area compared to their non-tumor-bearing counterparts (Supp Fig 2). Likewise, tumor-326 bearing anti-PD-1-treated animals had higher levels of preantral follicle density than the non -327 tumor-bearing anti-PD-1-treated group (Supp Fig 2). These findings demonstrate that the effect 328 of tumor burden may play a larger role in determining ovarian and endocrine health than any 329 immune-related changes resulting from PD-1 blockade. 330 331 Treatment with anti-PD-1 immunotherapy does not impair ovulatory capacity 332 To investigate whether PD -1 inhibition could affect ovulatory efficiency, we conducted 333 superovulation studies to assess responses to ovarian hyperstimulation. One week following 334 treatment with anti-PD-1 immunotherapy, IgG isotype, or saline control, animals of each group 335 were hormonally stimulated with PMSG, then given a trigger shot of HCG 48 hours later. Cumulus-336 oocyte complexes and ovaries were collected from the ampullas of all animals 12 hours after the 337 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint 14 HCG injection. Ovaries from stimulated mice displayed corpora lutea formation consistent with 338 recent ovulation (Fig 4 a-c). 339 There were no significant differences in the numbers of successfully stimulated mice or 340 number of oocytes recovered per stimulated mouse between the anti -PD-1-treated group and 341 control groups (Fig 4d -e). These results indicate that PD -1 inhibition does not impact ovulatory 342 efficiency or the ability to respond to hormonal stimulation. 343 344

345 As cancer therapeutics and survivorship outcomes improve, preserving the ovarian 346 reserve during cancer treatment has become critical. Not only does the ovarian reserve comprise 347 an individual’s entire reproductive potential, but it also directly contributes to endocrine balance, 348 making it critical for female health -span and lifespan. Moreover, though immune checkpoint 349 inhibitors may potentially be a less-toxic approach to treating cancer, they are still associated with 350 a suite of adverse immune effects and thus require characterization of their impact on 351 reproductive and hormonal function 2,13. Based on our rigorous studies in both a TNBC tumor -352 bearing model and a tumor-free model, we found no evidence that immune checkpoint inhibition 353 negatively affects the ovarian reserve or endocrine homeostasis. 354 In this work, we demonstrate that ovarian architecture and follicle density remain 355 unchanged by immune checkpoint blockade. Ovaries from tumor -bearing mice treated with 356 immunotherapies targeting PD -1, LAG -3, and TIM -3 appeared healthy and bore remarkable 357 resemblance to ovaries from mice treated with IgG isotype control. While we were interested in 358 understanding any ovarian or endocrine effects of anti-PD-1 therapy as it is commonly being used 359 to treat TNBC in the clinic, we chose to also investigate anti -LAG-3 and anti-TIM-3 because of 360 their possible future uses in treatment regimens for TNBC or other solid cancers. However, due 361 to their variable efficacy in TNBC clinical trials, we narrowed our scope to focus solely on anti -362 PD-1 therapy in non -tumor-bearing and subsequent studies 38. As seen in the tumor -bearing 363 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint 15 cohort, ovaries from non -tumor-bearing mice treated with anti -PD-1 were equally as healthy as 364 the control group ovaries. Critically, this lack of ovarian and endocrine damage observed after 365 immune checkpoint inhibitor treatment coincides with tumor regression in the tumor -bearing 366 cohort, particularly in anti -PD-1-treated mice. This anti -tumor efficacy indicates that our 367 monotherapy doses are therapeutically relevant, and thus, are appropriate doses for evaluating 368 reproductive toxicity. 369 Interestingly, the primary differences were found only when comparing ovaries from tumor-370 bearing mice to those from non-tumor-bearing mice. Among mice treated with anti-PD-1 therapy, 371 preantral follicles were significantly higher in the tumor -bearing group, while non -significant 372 increases could also be seen in degenerative and primary follicles of the tumor -bearing group. 373 One possible reason for this could be a delay in follicle maturation in tumor -bearing animals, 374 causing an accumulation of maturing follicles and thus a higher percentage of degeneration. 375 These disruptions in folliculogenesis may be attributable to non -specific inflammation related to 376 tumor burden, especially considering the anatomical proximity of the mammary tumor to the 377 ovaries. Moreover, among mice treated with IgG isotype, ovarian area was reduced in the tumor-378 bearing group while oocyte density remained comparable. This finding could potentially be due to 379 cancer-related inflammatory effects on the somatic compartment of the ovary, a mechanism 380 proposed by Chaqour et al 43. These results may possibly indicate a trend that local cancer -381 associated inflammation may have a larger role in determining ovarian health than immune 382 checkpoint blockade itself. 383 Our studies also show that hormonal cyclicity and serum hormone levels are similar 384 between immunotherapy and control group mice. These results were not particularly surprising 385 given that we found that immunotherapy did not diminish the ovarian reserve. However, 386 considering that autoimmune disorders affecting endocrine function are among the most common 387 of the adverse immune effects associated with immune checkpoint inhibitor treatment, it was 388 important for our research to investigate possible extra -ovarian effects 13. Finally, our results 389 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint 16 suggest that anti-PD-1 immunotherapy has no effect on responsiveness to hormonal stimulation 390 or ovulatory capacity. Taken together, we can conclude that immune checkpoint inhibitors, namely 391 anti-PD-1-based immunotherapies, could be a critical component of ovary -sparing cancer 392 therapeutic regimens. 393 Though there is little literature available on the effects of immunotherapy on the ovarian 394 reserve, a 2022 study from Winship et al. found that PD -L1 and CTLA -4 inhibition causes 395 depletion of the ovarian reserve, disruption of estrus cyclicity, reduced ovulatory capacity, and an 396 increase in intra-ovarian immune activity14. Similar to our studies, Winship et al. evaluated these 397 immunotherapies in both a tumor -bearing and non -tumor-bearing model. Given that we 398 investigated different immune checkpoint targets, our results are not necessarily inconsistent with 399 those reported by Winship et al. However, our conclusions are strikingly different when 400 generalizing about the effects of immune checkpoint blockade as a whole. While the Winship 401 study posits that immune checkpoint inhibitors are harmful to the ovarian reserve, our study is the 402 first to test the effects of anti -PD-1 therapy, the only standard-of-care immunotherapy approved 403 by the FDA for the treatment of TNBC 7. In addition to the difference in selection of immune 404 checkpoint targets, some of our contrasting results may also be attributable to the difference in 405 mouse models used. For example, the Winship study employed the use of C57Bl/6J mice 406 orthotopically injected with AT3OVA cells for their tumor -bearing experiments 14. While this is 407 indeed a syngeneic model for mammary carcinoma, the cell line is not as commonly used as our 408 E0771 model and could have critical differences in immunogenicity, receptor expression, or 409 clinical relevance 44. Ultimately, our study provides critical context to the current body of 410 knowledge on the reproductive and endocrine impact of immune checkpoint inhibitors. 411 Our mouse studies sought to model a clinically-relevant dosing schedule equivalent to one 412 round of immunotherapy treatment widely used in previous research. By allowing 14 days to 413 elapse between treatment and collection, we were able to monitor and record daily estrus cycling 414 to assess immediate effects before eventually capturing delayed effects on ovarian health and 415 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint 17 serum hormone balance. Future studies will target long -term outcomes such as fecundity and 416 premature ovarian aging after immunotherapy exposure. 417 Though we aimed to design and execute comprehensive studies, this work is not without 418 its limitations. Our studies do not investigate subtle immunotherapy -induced changes that may 419 not be apparent in histological analyses. However, due to the overwhelming lack of negative 420 functional effects observed, any subtle phenotypes will be subject of future studies and are 421 beyond the purview of this current study. In addition, we recognize that the number of mice who 422 were responsive to hormonal stimulation is low, making it more difficult to draw conclusions about 423 ovulatory capacity. However, considering the need to balance a clinically-relevant dosing schema 424 with the inherent challenges of hormonally stimulating mice older than 4 weeks of age, we chose 425 to perform our study as described and make statistical adjustments to account for reasonable 426 oocyte recoveries. 427 428

429 Our research adds novel information to a burgeoning field studying the effects of 430 immunotherapy on reproductive and endocrine function, and provides reassuring data to support 431 that anti-PD-1 therapy for TNBC, which is now standard-of-care, does not detrimentally affect 432 ovarian function. Given the rising number of young women being diagnosed with TNBC, there is 433 a critical need to design treatment approaches that can maximize therapeutic efficacy while 434 minimizing damage to the ovarian reserve, thereby improving treatment outcomes and quality of 435 life for survivors. Though population-level clinical data would provide the most clarity on the fertility 436 outcomes after immune checkpoint inhibitor treatment, our studies present a human -relevant 437 animal model to assess ovarian and endocrine effects of novel immunotherapies to inform clinical 438 recommendations. 439 440 DECLARATIONS 441 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint 18 Ethics approval and consent to participate 442 Not applicable. 443 444 Consent for publication 445 Not applicable. 446 447 Availability of data and materials 448 The data generated from this study are available from the corresponding author upon reasonable 449 request. 450 451 Competing interests 452 The authors declare no competing interests. 453 454 Funding 455 Primary sources of funding for personnel, reagents, and supplies was provided by Swim Across 456 America and the Foundation for Women's Wellness. Support for key instruments used for this 457 research was provided by the Kilguss Research Core of Women and Infants Hospital and the 458 Brown University Genomics Core. 459 460 Authors’ contributions 461 PDLC, MFW-S, JNM , ES, LH, MMA, and KJG performed the experiments. PDLC and KJG 462 analyzed and interpreted the data. PDLC and KJG wrote and revised the manuscript. 463 464

465 The authors would like to thank the Program in Women’s Oncology of Women and Infants Hospital 466 and the generous funding support of Swim Across America and the Foundation for Women’s 467 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint 19 Wellness. Technical support for this work was provided by the University of Virginia Center for 468 Research in Reproduction Ligand Assay and Analysis Core, which is supported by the Eunice 469 Kennedy Shriver NICHD Grant R24 HD102061. The authors thank the Freiman and James 470 laboratories for providing valuable feedback on this project. 471 472 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint 20 FIGURE LEGENDS 473 Figure 1- Inhibition of PD -1, LAG-3, and TIM -3 does not impact ovarian follicle abundance or 474 quality in a mouse model of triple negative breast cancer 475 Ovaries from E0771 tumor-bearing mice treated with monoclonal antibodies targeting PD-1, LAG-476 3, TIM-3 (A-C), IgG isotype (D), and a healthy control mock-injected with saline (E) were collected, 477 formalin-fixed, paraffin -embedded, and then stained with hematoxylin and eosin via standard 478 protocols (n=3 per group). Degenerative follicles are denoted with white arrows. Tumor burden 479 analysis (F) at study endpoint reveals near-complete tumor regression in anti-PD-1-treated mice, 480 along with varying reductions in tumor size in the anti -LAG-3 and anti -TIM-3 groups (n=3 per 481 treatment group). Ovarian size, oocyte density, and follicle stage density was not significantly 482 different between any of the treatment or control groups ( G-I). Ovarian follicle counts were 483 quantified using one section on every fourth slide of sectioned ovary to capture all follicle stages 484 without over-representing larger follicles. To account for size differences between ovaries, section 485 area was used to normalize follicle counts. Data for follicle counts, follicle density, and ovarian 486 area can be found in Supplementary File 1. 487 *p<0.05, as indicated, 488 489 Figure 2 - Inhibition of PD -1, LAG -3, and TIM -3 does not perturb endocrine homeostasis or 490 reproductive cyclicity 491 Serum concentrations of LH (A) and FSH (B), as well as the LH:FSH ratio (C), and AMH (D) were 492 not significantly different between any of the treatment and control groups (n=3 per group). Estrus 493 cyclicity was monitored daily via vaginal cytology for two weeks starting the day following the final 494 immunotherapy treatment, and the percentage of time spent in each substage was averaged 495 between mice in each treatment group (n=3 per group). The amount of time spent in each 496 substage of the estrus cycle was unchanged between all treatment and control groups (E). 497 498 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint 21 Figure 3- Treatment with anti -PD-1 immunotherapy does not impact ovarian follicle density or 499 reproductive cyclicity in a non-tumor-bearing mouse model 500 Ovaries were collected for FFPE from healthy, non -tumor-bearing mice treated with anti -PD-1 501 monoclonal antibodies, IgG isotype control, or saline control ( A-C), and then stained with 502 hematoxylin and eosin (n= 3 per group). There were no significant differences between ovarian 503 size, oocyte density, follicle abundance, or the percentage of time spent in each substage of 504 estrus between treatment groups (D-G). Data for follicle counts, follicle density, and ovarian area 505 can be found in Supplementary File 2. 506 507 Figure 4- Treatment with anti-PD-1 immunotherapy does not impair ovulatory capacity 508 One week following treatment with anti-PD-1 monoclonal antibodies, IgG isotype control, or saline 509 control, mice were super-ovulated with PMSG and HCG using standard protocols. Ovaries from 510 superovulated mice were collected for FFPE and then stained with hematoxylin and eosin (A-C). 511 There were no significant differences in the number of mice who responded to hormonal 512 stimulation (D) or the number of oocytes retrieved from successfully superovulated mice ( E). 513 Retrieved oocytes were collected from each animal, then counted and averaged per treatment 514 group (n=6 for anti-PD-1 and IgG isotype control, n=5 for saline control). Mice were classified as 515 “successfully stimulated” if their number of retrieved oocytes reached an age-adjusted threshold 516 for hyperstimulation (10 oocytes for 7-week-old mice, 7 oocytes for 10-week-old mice). Counts of 517 retrieved oocytes from “successfully stimulated” mice were averaged and compared between 518 treatment groups (n=5 for anti-PD-1, n=2 for IgG isotype control, n=4 for saline control). 519 520 Supplemental figure 1 521 TUNEL staining of ovaries from E0771 tumor-bearing mice treated with monoclonal antibodies 522 targeting PD-1, LAG-3, TIM-3, and IgG isotype control showed no appreciable levels of apoptosis 523 in follicles. 524 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint 22 525 Supplemental figure 2 526 Differences in ovarian area, follicle density, and estrus cyclicity in tumor -bearing vs. non-tumor-527 bearing mice . Among mice treated with IgG isotype control, ovarian area was significantly 528 increased in non -tumor-bearing mice, which may indicate increased volume in the ovarian 529 somatic compartment of healthy mice. Among mice receiving anti -PD-1 treatment, preantral 530 follicle counts were significantly higher in the tumor -bearing group, perhaps indicating an 531 accumulation of maturing follicles in TNBC mice. 532 *p<0.05, as indicated 533 534 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint 23

535 1. American Cancer Society. About Breast Cancer. Am Cancer Soc Cancer Facts Fig Atlanta, 536 Ga Am Cancer Soc . Published online 2017:1 -19. http://www.cancer.org/cancer/breast -537 cancer/about/what-is-breast-cancer.html 538 2. Duma N, Lambertini M. It Is Time to Talk About Fertility and Immunotherapy. Oncologist. 539 2020;25(4):277-278. doi:10.1634/theoncologist.2019-0837 540 3. Yin L, Duan JJ, Bian XW, Yu SC. Triple -negative breast cancer molecular subtyping and 541 treatment progress. Breast Cancer Res. 2020;22(1):1-13. doi:10.1186/s13058-020-01296-542 5 543 4. Kalimutho M, Parsons K, Mittal D, López JA, Srihari S, Khanna KK. Targeted Therapies for 544 Triple-Negative Breast Cancer: Combating a Stubborn Disease. Trends Pharmacol Sci . 545 2015;36(12):822-846. doi:10.1016/j.tips.2015.08.009 546 5. Singh S, Numan A, Maddiboyina B, et al. The emerging role of immune checkpoint 547 inhibitors in the treatment of triple -negative breast cancer. Drug Discov Today . 548 2021;26(7):1721-1727. doi:10.1016/j.drudis.2021.03.011 549 6. Morgan S, Anderson RA, Gourley C, Wallace WH, Spears N. How do chemotherapeutic 550 agents damage the ovary? Hum Reprod Update . 2012;18(5):525 -535. 551 doi:10.1093/humupd/dms022 552 7. U.S. Food and Drug Administration. FDA approves pembrolizumab for high-risk early-stage 553 triple-negative breast cancer. Published 2021. https://www.fda.gov/drugs/resources -554 information-approved-drugs/fda-approves-pembrolizumab-high-risk-early-stage-triple-555 negative-breast-cancer 556 8. Volckmar X, Vallejo M, Bertoldo MJ, et al. Oncofertility Information Available for Recently 557 Approved Novel Non Cytotoxic and Immunotherapy Oncology Drugs. Clin Pharmacol Ther. 558 2021;0(0):1-9. doi:10.1002/cpt.2254 559 9. Schirrmacher V. From chemotherapy to biological therapy: A review of novel concepts to 560 reduce the side effects of systemic cancer treatment (Review). Int J Oncol. 2019;54(2):407-561 419. doi:10.3892/ijo.2018.4661 562 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint 24 10. Darvin P, Toor SM, Sasidharan Nair V, Elkord E. Immune checkpoint inhibitors: recent 563 progress and potential biomarkers. Exp Mol Med. 2018;50(12):1-11. doi:10.1038/s12276-564 018-0191-1 565 11. Brown TJ, Mamtani R, Bange EM. Immunotherapy Adverse Effects. JAMA Oncol . 566 2021;7(12):1908. doi:10.1001/jamaoncol.2021.5009 567 12. Wright JJ, Powers AC, Johnson DB. Endocrine toxicities of immune checkpoint inhibitors. 568 Nat Rev Endocrinol. 2021;17(7):389-399. doi:10.1038/s41574-021-00484-3 569 13. Husebye ES, Castinetti F, Criseno S, et al. Endocrine-related adverse conditions in patients 570 receiving immune checkpoint inhibition: an ESE clinical practice guideline. Eur J 571 Endocrinol. 2022;187(6):G1-G21. doi:10.1530/EJE-22-0689 572 14. Winship AL, Alesi LR, Sant S, et al. Checkpoint inhibitor immunotherapy diminishes oocyte 573 number and quality in mice. Nat Cancer. 2022;3(8):1-13. doi:10.1038/s43018-022-00413-574 x 575 15. Kerr JB, Myers M, Anderson RA. The dynamics of the primordial follicle reserve. 576 Reproduction. 2013;146(6). doi:10.1530/REP-13-0181 577 16. Grive KJ, Freiman RN. The developmental origins of the mammalian ovarian reserve. Dev. 578 2015;142(15):2554-2563. doi:10.1242/dev.125211 579 17. Pepling ME. Follicular assembly: Mechanisms of action. Reproduction. 2012;143(2):139-580 149. doi:10.1530/REP-11-0299 581 18. Monniaux D, Clément F, Dalbiès -Tran R, et al. The ovarian reserve of primordial follicles 582 and the dynamic reserve of antral growing follicles: what is the link? Biol Reprod . 583 2014;90(4):85. doi:10.1095/biolreprod.113.117077 584 19. Richards JS, Pangas SA, Richards JS, Pangas SA. The ovary : basic biology and clinical 585 implications Find the latest version : Review series The ovary : basic biology and clinical 586 implications. 2010;120(4):963-972. doi:10.1172/JCI41350.critical 587 20. Edson MA, Nagaraja AK, Matzuk MM. The mammalian ovary from genesis to revelation. 588 Endocr Rev. 2009;30(6):624-712. doi:10.1210/er.2009-0012 589 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint 25 21. Kerr JB, Brogan L, Myers M, et al. The primordial follicle reserve is not renewed after 590 chemical or γ -irradiation mediated depletion. Reproduction. 2012;143(4):469 -476. 591 doi:10.1530/REP-11-0430 592 22. Grive KJ. Pathways coordinating oocyte attrition and abundance during mammalian 593 ovarian reserve establishment. Mol Reprod Dev . 2020;87(8):843 -856. 594 doi:10.1002/mrd.23401 595 23. Bedoschi G, Navarro PA, Oktay K. Chemotherapy-induced damage to ovary: Mechanisms 596 and clinical impact. Futur Oncol. Published online 2016. doi:10.2217/fon-2016-0176 597 24. Wesevich V, Kellen AN, Pal L. Recent advances in understanding primary ovarian 598 insufficiency. F1000Research. Published online 2020. 599 doi:10.12688/f1000research.26423.1 600 25. Gougeon A. Dynamics of follicular growth in the human: a model from preliminary results. 601 Hum Reprod. 1986;1(2):81-87. doi:10.1093/oxfordjournals.humrep.a136365 602 26. Chaffin CL, VandeVoort CA. Follicle growth, ovulation, and luteal formation in primates and 603 rodents: A comparative perspective. Exp Biol Med . 2013;238(5):539 -548. 604 doi:10.1177/1535370213489437 605 27. Byers SL, Wiles M V., Dunn SL, Taft RA. Mouse estrous cycle identification tool and 606 images. PLoS One. 2012;7(4):2-6. doi:10.1371/journal.pone.0035538 607 28. Coxworth JE, Hawkes K. Ovarian follicle loss in humans and mice: lessons from statistical 608 model comparison. Hum Reprod. 2010;25(7):1796-1805. doi:10.1093/humrep/deq136 609 29. Best CL, Pudney J, Welch WR, Burger N, Hill JA. Localization and characterization of white 610 blood cell populations within the human ovary throughout the menstrual cycle and 611 menopause. Hum Reprod. 1996;11(4):790-797. 612 doi:10.1093/oxfordjournals.humrep.a019256 613 30. Pate JL. Involvement of immune cells in regulation of ovarian function. J Reprod Fertil 614 Suppl. 1995;49:365-377. doi:10.1530/biosciprocs.3.028 615 31. Wu R, Van der Hoek KH, Ryan NK, Norman RJ, Robker RL. Macrophage contributions to 616 ovarian function. Hum Reprod Update. 2004;10(2):119-133. doi:10.1093/humupd/dmh011 617 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint 26 32. Sharif K, Watad A, Bridgewood C, Kanduc D, Amital H, Shoenfeld Y. Insights into the 618 autoimmune aspect of premature ovarian insufficiency. Best Pract Res Clin Endocrinol 619 Metab. 2019;33(6):101323. doi:https://doi.org/10.1016/j.beem.2019.101323 620 33. Zimon A, Erat A, Reindollar R, Usheva A. NFkB and other markers of chronic inflammation 621 in the prediction of ovarian aging and infertility. Fertil Steril. 2004;82(September):S318. 622 doi:10.1016/j.fertnstert.2004.07.857 623 34. Lliberos C, Liew SH, Zareie P, La Gruta NL, Mansell A, Hutt K. Evaluation of inflammation 624 and follicle depletion during ovarian ageing in mice. Sci Rep . 2021;11(1):1 -15. 625 doi:10.1038/s41598-020-79488-4 626 35. Tuğrul Ayanoğlu B, Özdemir ED, Türkoğlu O, Alhan A. Diminished ovarian reserve in 627 patients with psoriasis. Taiwan J Obstet Gynecol . 2018;57(2):227 -230. 628 doi:10.1016/J.TJOG.2018.02.010 629 36. Zhao Y, Chen B, He Y, et al. Risk Factors Associated with Impaired Ovarian Reserve in 630 Young Women of Reproductive Age with Crohn’s Disease. ir. 2020;18(2):200 -209. 631 doi:10.5217/ir.2019.00103 632 37. Alesi LR, Winship AL, Hutt KJ. Evaluating the impacts of emerging cancer therapies on 633 ovarian function. Curr Opin Endocr Metab Res . 2021;18:15 -28. 634 doi:10.1016/j.coemr.2020.12.004 635 38. Cai L, Li Y, Tan J, Xu L, Li Y. Targeting LAG -3, TIM -3, and TIGIT for cancer 636 immunotherapy. J Hematol Oncol. 2023;16(1):1-34. doi:10.1186/s13045-023-01499-1 637 39. Schmid P, Cortes J, Pusztai L, et al. Pembrolizumab for Early Triple -Negative Breast 638 Cancer. N Engl J Med. 2020;382(9):810-821. doi:10.1056/nejmoa1910549 639 40. Takahashi M, Cortés J, Dent R, et al. Pembrolizumab Plus Chemotherapy Followed by 640 Pembrolizumab in Patients With Early Triple-Negative Breast Cancer. JAMA Netw Open. 641 2023;6(11):e2342107. doi:10.1001/jamanetworkopen.2023.42107 642 41. Kedem-Dickman A, Maman E, Yung Y, et al. Anti -Müllerian hormone is highly expressed 643 and secreted from cumulus granulosa cells of stimulated preovulatory immature and atretic 644 oocytes. Reprod Biomed Online. 2012;24(5):540-546. doi:10.1016/j.rbmo.2012.01.023 645 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint 27 42. Saadia Z. Follicle Stimulating Hormone (LH: FSH) Ratio in Polycystic Ovary Syndrome 646 (PCOS) - Obese vs. Non - Obese Women. Med Arch (Sarajevo, Bosnia Herzegovina) . 647 2020;74(4):289-293. doi:10.5455/medarh.2020.74.289-293 648 43. Chaqour J, Ozcan MCH, De La Cruz P, Woodman -Sousa MF, McAdams JN, Grive KJ. 649 Effects of Maternal Taxane Chemotherapy Exposure on Daughters’ Ovarian Reserve and 650 Fertility Potential. F&S Sci. Published online 2023. doi:10.1016/j.xfss.2023.10.003 651 44. Le Naour A, Koffi Y, Diab M, et al. EO771, the first luminal B mammary cancer cell line 652 from C57BL/6 mice. Cancer Cell Int. 2020;20(1):328. doi:10.1186/s12935-020-01418-1 653 45. Ueha S, Yokochi S, Ishiwata Y, et al. Robust antitumor effects of combined anti -CD4-654 depleting antibody and anti -PD-1/PD-L1 immune checkpoint antibody treatment in mice. 655 Cancer Immunol Res. 2015;3(6):631-640. doi:10.1158/2326-6066.CIR-14-0190 656 46. Woodman MF, Ozcan MCH, Gura MA, De La Cruz P, Gadson AK, Grive KJ. The 657 requirement of ubiquitin C -terminal hydrolase L1 in mouse ovarian development and 658 fertility. Biol Reprod. Published online May 3, 2022:ioac086. doi:10.1093/biolre/ioac086 659 660 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint Figure 1 1 mm anti-PD-1 a anti-LAG-3 b f c anti-TIM-3 IgG isotype control d Healthy control (Saline mock-injected) e i g h .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint Figure 2 dc ba e .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint Figure 3 IgG isotype control b Saline control c d e anti-PD-1 1 mm a g f .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint Figure 4 1 mm anti-PD-1 a ed Saline control c IgG isotype control b .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 19, 2024. ; https://doi.org/10.1101/2024.08.14.607933doi: bioRxiv preprint