Combination Hormonal Drug Therapy Inhibits In Vitro Growth of Adult Ovarian Granulosa Tumor Cells

This study investigated the mechanism of action and in vitro efficacy of an anti-hormonal combination drug regimen in treating adult-type granulosa cell tumors of the ovary (AGCT). Using patient-derived and commercial AGCT cell lines, the IC50 values of bicalutamide, anastrozole, and leuprolide acetate were established alone and in combination via cell proliferation assays. Averaged across cell lines, IC50 was determined for bicalutamide (IC50: 7.3 µM) and anastrozole (IC50: 6.2 µM). Minimal effect was seen with leuprolide acetate alone. Suppression of proliferation was increased with the combination of an antiandrogen and an aromatase inhibitor. RNA was extracted from treated and untreated cells. RNA sequencing of cells treated with both anastrozole and darolutamide suggested downregulation of pathways involved in DNA synthesis, cell cycle regulation, and chromosomal organization. MHC I activity and DNA damage response were upregulated. This anti-hormonal drug regimen for AGCT merits prospective investigation.

Introduction Achieving targeted therapies for cancer treatment requires dedicated investigations of cancer subtypes to develop specific, sustainable treatment options with decreased toxicity. Recent breakthroughs in targeted treatment for epithelial ovarian, cervical, and endometrial cancers have generated increased enthusiasm for targeted therapies across gynecologic malignancies. Though, at initial diagnosis, outcomes are generally favorable for early-stage adult-type granulosa cell tumors of the ovary (AGCT), 10-30% of cases recur which leads to death in 50-80% of patients (Citation1–3). Single-agent blockade of hormone production or hormone receptors, a targeted treatment for AGCT, has thus far shown limited activity (Citation4,Citation5). AGCT’s pathognomonic mutation, FOXL2C402G, and its associated protein FOXL2C134W, have many intracellular effects that remain incompletely understood (Citation6–8). This mutation does exhibit effects related to hallmarks of cancer development, including apoptosis inhibition (Citation9). For example, via binding to SMAD4, a FOXL2C134W/SMAD4/SMAD2/3 complex is created, leading to enhanced gene transcription and possibly contributing to epithelial-to-mesenchymal transition (EMT) (Citation8). FOXL2C134W has also been shown to affect hormone production and signaling pathways in AGCT, as illustrated by Farkkila et al. (Citation2) Blockade of these pathways continues to be an attractive target for treatment. The physiologic function of FOXL2 in hormonal pathways has been postulated given its role in embryonic sex differentiation and premature ovarian insufficiency (Citation10–12). Our current knowledge regarding the effects of FOXL2C134W on hormonal production in AGCT is illustrated in . The FOXL2C134W mutation upregulates aromatase activation, and therefore estrogen production, by stimulating expression of CYP19A1 and estrogen receptor β (Erβ) compared to wild type FOXL2 (Citation13,Citation14). The increased estrogen levels continue the cascade of increased estrogen production via FOXL2 and CYP19A1 activation (Citation10). In the presence of a FOXL2 mutation, estrogen no longer causes apoptosis or inhibits proliferation (as it would with wild type FOXL2) (Citation15). AGCT is known to express higher levels of Erβ than Erα, so this signaling may be mediated more dominantly by Erβ signaling pathways (Citation16). AGCT is also known to express androgen receptor (AR) and, in addition to increasing estrogen production, the FOXL2 mutation also increases androgen production. When FOXL2 is mutated, CYP17A1 (key enzyme involved in androgen production) is no longer repressed by the FOXL2-SF1 complex, leading to increased androgen expression (Citation14). AR inhibitors in conjunction with gonadotropin-releasing hormone (GnRH) blockade are used in prostate cancer to prevent cells from circumventing AR inhibition, termed “complete androgen blockade” (Citation17). AR inhibition has previously been attempted in AGCT with ketoconazole, ortonorel, and enzalutamide (Citation18–20). GREKO I, which evaluated ketoconazole due to its antiandrogen properties, included 6 patients and reported a 14.06 month median progression-free survival (PFS). Stable disease was seen in 5/6 patients for 12 months but no objective responses were achieved (Citation18). GREKO II aimed to investigate the efficacy of orteronel (CYP17A1 inhibitor), but the study closed early due to low patient accrual (Citation20). GREKO III investigated the antiandrogen enzalutamide and stable disease was noted to be the best response with median PFS of 3.8 months (Citation19). Darolutamide has not been previously studied in AGCT, nor have antihormonal combination treatments been formally evaluated in a clinical trial. Based on the above information, we devised a triplet anti-hormonal therapy regimen for AGCT. We previously reported an institutional case series of patients who received this triplet therapy of an AR antagonist, GnRH agonist, and aromatase inhibitor (AI) with favorable outcomes and minimal toxicity (Citation21). The first objective of this study was to investigate the efficacy and synergy of this drug combination using in vitro models. We hypothesized improved in vitro efficacy of this combination treatment used to treat AGCT compared to each drug alone. Our secondary objective was to elucidate hormonal pathways affected by FOXL2C134W.

Clinical samples KGN (cell line A) is a commercially developed cell line derived from a 69-year-old patient with a primary stage III AGCT. Patient samples were collected through the Obstetrics and Gynecology (OB/GYN) Tissue Bank at the Medical College of Wisconsin (MCW). This study was approved by the MCW Institutional Review Board. Additional cell lines were derived from a 52-year-old patient with Stage IIA AGCT (cell line B), a 37-year-old patient with recurrent AGCT (cell line C), and a 63-year-old with Stage IA AGCT (cell line D). Previous immunohistochemistry (IHC) staining through Wisconsin Diagnostic Lab (WDL) showed that cell lines B and C were positive for estrogen, progesterone, and androgen receptors. Cell line D was derived from an AGCT with a prominent theca cell component which was strongly positive for estrogen, progesterone, and androgen receptors. The granulosa cell component was negative for estrogen and progesterone receptors but displayed weak staining for androgen receptors. Cell culture Cell culture conditions Cell line A was obtained from Creative Bioarray, New York, USA (CSC-C9202W). All cell cultures were maintained in DMEM/F12 (Gibco) culture media supplemented with 10% FBS and 1% Pen/Strep (Supplemental Table 1). Cells were incubated at 37 °C in a humidified atmosphere containing 5% CO2. Culture medium was changed every 2–3 days, and cells were passaged at 70-90% confluency. All cell lines were tested for mycoplasma when cultures were established or when cell lines were thawed from frozen using the ATCC Universal Mycoplasma Detection Kit (Virginia, USA) and InvivoGen MycoStripTM Mycoplasma Detection Kit. Primary culture protocol Cell lines B and C were developed from the above banked tumor samples. An additional cell line, D, was used for spheroid experiments as specified in the spheroid portion of the Methods. Tumors were minced with a scalpel to 5 mm or less, rinsed with 1X PBS, then placed in culture media supplemented with dispase 10 units, 3X gentamicin, and 1X fungizone. The subsequent day, tumors were minced mechanically with a scalpel and then chemically via Trypsin-EDTA (0.25%). Tumor fragments were further cultured in culture media described above and 2D cell cultures developed once cells adhered to the culture plate. Cell lines B and C were immortalized; experiments were performed at passages ranging from 15 to 43. Immortalized cell line development was attempted from other tumors, including recurrent tumor samples, but was not successfully achieved. FOXL2 mutation sequencing and STR analysis Cell line DNA was extracted using the Lucigen® BuccalAmp™ DNA extraction kit. FOXL2 was amplified and reverse transcribed to 10 ug total cDNA using the following primer sequences: forward (5′-GGCTGGCAAAATAGCATCCG-3′) and reverse (5′-TGTACGAGTTCACTACGCCG-3′). Primers were obtained from Integrated DNA Technologies (IDT, IA, USA). Amplified DNA was run on a 2.5% agarose gel and the band excised using a UV light. Amplified DNA from the band was extracted and concentrated using the Invitrogen PureLink® Quick Gel Extraction Kit and the Zymo Research DNA Clean and Concentrator Kit, respectively. Sanger sequencing of FOXL2 was performed by Functional Biosciences to confirm presence of the mutation (Citation22). KGN and the two primary, patient-derived AGCT cell lines (B and C) were authenticated via short tandem repeat testing (ATCC) to confirm that they were distinct from KGN and each other. Cell line D did not develop well in two-dimensional culture and thus could not be authenticated. Western blots AR and GnRHR protein expression were evaluated with Western blot in each cell line. Cultured cells were rinsed with 1X PBS and incubated overnight in RIPA buffer at 20 °C. Cells were collected with a cell scraper, homogenized with a 21-gauge needle, and centrifuged for 30 min at 4 °C. The supernatant was collected, and protein concentration measured using the BCA assay (Pierce, Illinois, USA) per manufacturer instructions. Protein lysates (30 µg) and ladder were loaded on a 10% Tris-HCl gel (BioRad) and run for approximately 1 hr at 150 V. Protein was then transferred to a PVDF membrane via running for 30 min at 100 V. Membranes were washed with 1X TBS and incubated in block solution (10% nonfat milk/TBS) for 1 hr. Membranes were then incubated with AR (Cell Signaling; #5153S; 1:2000) or GnRHR (Invitrogen; #MA1-35383; 1:5000) primary antibodies diluted in 1X TBS with 0.1% Tween (TBST) and 5% nonfat milk or bovine serum albumin (BSA), respectively. Incubation was performed at 4 °C overnight. Secondary horseradish peroxidase (HRP)-linked goat anti-rabbit and goat anti-mouse antibodies (Cell Signaling) were used at concentrations of 1:2000 and 1:5000, respectively, in 5% nonfat milk/TBST for 1 hr at room temperature. Membranes were imaged with the Invitrogen iBright FL1500 following 3–5 min incubation with SuperSignal West Pico PLUS Chemoluminescent substrate (ThermoScientific, #34580). For normalization of expression data, membranes were stripped and re-probed for β-actin (Cell Signaling; #3700; 1:5000) using an HRP-linked goat anti-rabbit secondary (1:5000) as above. Viability assays with single drugs Cell lines A, B, C were treated with the drugs utilized in our triple antihormonal therapy regimen sequentially, using various culture conditions to optimize results. Conditions included varying dilutions of cells; media supplemented with normal fetal bovine serum (FBS) or charcoal-depleted FBS to ensure hormone depletion in the media; or media supplemented with hydrocortisone, insulin, transferrin, estradiol, sodium selenite (HITES) (Citation23,Citation24). All subsequent experiments were carried out using 3000 cells/100 µL (0.03 x 106 cells/mL) with DMEM/F12 media containing 5% charcoal-depleted FBS and 1% Pen/Strep, without HITES. Experiments were performed in 96-well plates (Genesee Scientific, CA, USA). Drugs (leuprolide acetate, bicalutamide or darolutamide, and anastrozole) were obtained from Selleck Chemicals (Texas, USA). Bicalutamide was used initially to calculate AR blocker IC50s. Darolutamide, a second-generation androgen receptor pathway inhibitor with a higher binding affinity for AR than bicalutamide, was used in subsequent experiments given its clinical use in an activated AGCT clinical trial (Citation25). The single-agent IC50 of darolutamide was confirmed to be similar to that of bicalutamide in all three cell lines. Stock solutions (10 mM) of each drug were formulated in DMSO. As no IC50 was yet known for these medications in AGCT, each of the three cell lines were initially treated with varying concentrations (100 nM to 100 µM), which were selected based on cell line research in breast and prostate cancer (Citation26–28). Media containing anastrozole or bicalutamide was applied on day one and remained for the entirety of experiment. Media containing leuprolide acetate was exchanged daily due to its short half-life. Cellular proliferation was assessed with the CellTiter 96® Aqueous One Solution Cell Proliferation Assay (MTS), Promega (Wisconsin, USA) at 72 to 96 hrs using the Tecan Spark Multimode Plate Reader. The percentage of cellular proliferation compared to control was calculated for each well. Scatter plots were constructed using GraphPad Prism™. A 4-parameter dose-response curve was constructed for each medication in each cell line. The R2 was used to assess the goodness of fit. Concentrations were altered as needed to optimize the dose-response curves. Viability assays with combination therapies Subsequent dose-response assays were completed with a combination of anastrozole, darolutamide, and leuprolide acetate using the same culture conditions. Final dose ranges for treatment were: anastrozole (1.5–24 µM); leuprolide acetate (10–160 µM); and darolutamide (5–80 µM). Proliferation assays were performed as above. Dose-response scatter plots were created using SynergyFinderPlus (Citation29). Four-parameter dose-response curves were fit for individual drugs and drug combinations. To fully investigate drug synergy or antagonism, four models were used for synergy calculations: zero interaction potency (ZIP), Bliss, Loewe, and HAS models. Synergy scores and combined sensitivity scores were calculated for each combination of two drugs, as well as all three together. The Bliss model was chosen for reporting due to its simplicity. This model reports the difference between the efficacy of combination treatment and the maximum efficacy of a single treatment. Dose-response curves were constructed for anastrozole with varying concentrations of darolutamide. Each curve was fit with a four-parameter dose-response curve. Bar plots were constructed in R, version 4.2.1, using ggplot2 3.4.0. Spheroid formation Spheroid formation was initially attempted with the three primary cell lines (B, C, and D) in DMEM/F12 media. 5000 cells/well were plated in 200 µL media in a 96-well low-attachment, U-bottom plate (Falcon). Cell line B and cell line D were selected for further experiments based on successful spheroid formation. Culture media is specified in Supplemental Table 1. Plates were centrifuged at a low speed for 5–10 minutes to aid in spheroid formation. Spheroids were examined and imaged daily with the Nikon Eclipse Ts2. 3D spheroid treatment, imaging, viability and apoptosis assays On days 4–5 of spheroid formation, spheroids were imaged and measured for consistency in size. As previously described, if within 30% of the average spheroid diameter of 176 µm, spheroids were approved for medication assays (Citation30). Spheroids were treated with anastrozole (1.5-24 µM) and darolutamide (2.5-40 µM) in phenol-free media with 5% charcoal-depleted FBS. Three spheroids were treated with each concentration combination. Diameter and volume measurements were taken using the Nikon Eclipse Ts2. Spheroids were lysed by pipetting and half of the 200 µL media was transferred to an opaque 96-well plate where 100 µL of Celltiter Glo-Viability or 100 µL Celltiter Glo Caspase 3/7 (Promega) was then added to assess viability or apoptosis, respectively. The luminometer was calibrated and luminescence was measured on the Tecan Spark. RNA extraction Cell lines A, B, C were cultured in media containing 5% charcoal-depleted FBS without medication, with anastrozole (7 µM) or with anastrozole (7 µM), darolutamide (40 µM) and leuprolide acetate (50 µM). Cells were lysed and RNA was extracted 72 hours after treatment using the Invitrogen PureLink RNA Mini Kit. On-column DNase treatment was performed with the PureLink DNase Set. Five experimental replicates were performed using sequential passages of cells. In an effort to confirm that results would not vary by time of RNA extraction, the same was then performed with RNA extraction 24 and 48 hours after treatment, using cell line A, with two replicates each. RNA was then extracted from banked tumors as above, but with the following modifications. Tumor fragments were either homogenized with a rotor-stator in lysis buffer or with a mortar and pestle in liquid nitrogen. Lysis buffer was added to the mortar and pestle. Samples were homogenized with Invitrogen homogenizers. Granulosa cells were collected from patients undergoing oocyte cryopreservation. These cells were collected in media, oocytes were removed, then granulosa cells were centrifuged, and washed with 1X PBS twice. Cells were then treated with ThermoScientific Red Blood Cell Lysis Buffer, per manufacturer instructions. This was completed on the day of collection. Cells were washed once more with PBS before RNA extraction was performed as detailed above. RNA sequencing Following extraction, all RNA samples were stored at −80 °C. Extracted total RNA was shipped overnight to Novogene (California, USA) on dry ice for library preparation and paired-end sequencing. RNA integrity number was confirmed to be at least 4.0 with a smooth baseline. Sequencing was performed on the Illumina NovaSeq 6000 platform to a depth of 20 million reads per sample. Analysis of RNA sequencing data Data quality was confirmed with fastqc 0.11.9 and summarized with multiqc 1.11 (Citation31,Citation32). Reads were pseudoaligned to the Human Genome (hg38) with kallisto 0.46.1, and the percent of aligned reads was checked with multiqc (Citation33,Citation34). The transcriptome FASTA file was downloaded from Ensembl 9/1/22, and index and count files were created with kallisto. The remainder of the analyses were performed in R 4.2.1. Genes from experimental replicates were filtered to include those with ≥ 5 counts per million and expressed in at least 30% of any of the treatment conditions, then TMM normalized with edgeR 3.14.0 (Citation35). Principal component analysis (PCA) plots were constructed by grouping samples according to experimental condition. Outliers identified using principal components 1 and 2 were excluded. Heatmaps of top variable genes were used to confirm group assignments. Log2 fold change and p-values adjusted for multiple testing (adjusted p-values) by Benjamini-Hochberg false discovery rate method were calculated using a contrast matrix, subtracting control expression from anastrozole and combination replicates with limma (version 3.56.2) (Citation36). Differentially expressed genes (DEGs) are defined as those with adjusted p-values less than 0.05. Experimental replicate number was used as a co-variate. The same analysis was repeated with each cell line alone. Pathway analysis of DEGs was performed using GSEABase 1.62.0, clusterProfiler 4.8.1, and Pathview 1.40.0 (Citation37,Citation38). WGCNA 1.72-1 was used to assess module co-expression, and to confirm similarity of these themes with the otherwise identified pathways (Citation39). A soft power of 5 was used for signed network analyses within WGCNA. Cytoscape was used to illustrate WikiPathways. Normal ovary sample data, along with linked age data, was obtained from the University of California Santa Cruz TOIL Recompute (Kallisto quantification data) of the Genotype-Tissue Expression Project (GTEx). Data were downloaded 6/5/23 (Citation40) to the MCW Research Computing Center cluster. Tumor gene expression was then compared to that of normal granulosa cells and normal ovarian GTEx samples with PCA analysis. Tumor outliers were excluded and normal GTEx samples were selected as the control group for further analysis. Genes with at least five counts per million in at least five samples were selected, and expression was normalized as above. Log2 fold change of tumor to normal ovary samples and adjusted p-values were calculated use limma to identify DEGs. Z-scores for gene expression across tumor samples were calculated and averaged to determine within-sample gene importance. Pathway analysis was performed as above, then queried for molecules relevant to reproductive hormone signaling, including ESR1, ESR2, AR, CYP17, CYP19, FSH, LH, GNRHR, PR, and PRLH. Validation of DEGs by qPCR A set of genes with the highest and lowest log-fold change (combination treatment compared with control cells) observed in RNAseq studies, was selected for validation. Each of these was suspected to be physiologically relevant. RNA samples for validation had not been previously sent for RNAseq. Primers were selected as previously published (Citation41–50) and are listed in Supplementary Table 2. Complementary DNA (cDNA) was reverse transcribed from total RNA using the High-Capacity RNA-to-cDNA Kit (Applied Biosystems). Gene expression was assessed using iTaq Universal SYBR Green Supermix (Bio-Rad, Hercules, CA) on a StepOnePlus Real-Time PCR System (Applied Biosystems). The thermal cycling program included an enzyme activation step at 94 °C for 2 min, followed by 40 cycles of a 15 sec denaturing step at 94 °C, and a 1 min annealing/extension step at 55 °C. Fluorescent intensity was measured at 65 °C at the end of each cycle. Expression of target genes was normalized using the housekeeping gene, RPS18. Relative gene expression was calculated using the 2-ΔΔCt method. Data availability The AGCT data utilized in this study will publicly available in the SRA at the time of publication. The code used for RNA sequencing data analysis is available at https://github.com/rsummey/AGCT.

Combination anti-hormonal drug therapy, but not single-drug therapy, significantly decreased AGCT cell growth Sanger sequencing of the three cell lines utilized in our experiments (A, B, C) validated presence of the expected FOXL2C402G mutation and STR verification confirmed their uniqueness. Western blotting confirmed AR and GnRHR protein expression (Figure S1). Dose-response curves were created for bicalutamide and anastrozole in all cell lines. Dose-response curve R2, IC50, and confidence intervals for all medications and cell lines are displayed along with dose-response curves in Figure S1. A dose-response curve could be created for leuprolide acetate only in cell line C. We confirmed the dose-response curves of bicalutamide and darolutamide were similar. Effective treatment with a combination regimen required higher doses than those used for any medication in isolation but exhibited augmented suppression of proliferation (). Cells treated with the combination of darolutamide and anastrozole with leuprolide acetate did not exhibit impaired proliferation in vitro as compared with darolutamide and anastrozole alone (data not shown). Synergy scores varied by cell line and drug concentration. The overall synergy score from the HSA model (excess response over that of a single drug), was favorable (HSA = 5.72, p = 4.65 e-13). In each cell line, a nadir in proliferation suppression was seen with additive combination of anastrozole and darolutamide. Doses higher than this nadir commonly resulted in an increase in cell proliferation (data not shown). Spheroids were effectively created with cell lines B and D, and showed improved inhibition of volume and viability with increasing doses of combination treatment, above that of either drug alone (). Pathway analysis of RNAseq highlighted mechanisms for treatment efficacy including chromosome structure alteration, altered DNA replication, cell cycle regulation, and antigen presentation After exclusion of outliers, our final analytical cohort included 40 cell line samples (13 treated with anastrozole and darolutamide, 14 treated with anastrozole alone, and 13 controls). PCA plot is shown in Figure S6. Combination treatment, when adjusted for cell line and experiment number as covariates, resulted in 4,108 DEGs. No DEGs were identified in the anastrozole-treated group compared to control. The top 40 terms in the cellular components, biological processes and metabolic functions gene ontology terms of the gene set enrichment analysis are illustrated in and S3C-D. Terms featuring chromosome structure, DNA replication, cell cycle regulation, and major histochemical class I molecules (MHC I) were prominently enriched. shows genes that are significantly up or downregulated and are found in multiple known pathways. Genes linking chromosome segregation, cell cycle regulation and DNA replication processes () may be of particular importance (MMP, Gi/o, HB-EGF, Akt, EGFR, Shc, PI3K, MEK, ERK1/2, pCREB, TF, CoA, KKRT19, CTSD, RAR-alpha, FKBP52, Hsp70, GLI1, DBF4B, DTL, E2F8, RRM1, CDC7, GEN1, FBXO5, CDC6, CDT1, RMI1, MRE11, DSCC1, CHTF8, CCNE2, FANCM, RMI2). Log fold change in the treatment condition, significant or not, is highlighted in the WikiPathways Androgen Receptor Signaling pathway in Figure S3B. WGCNA analysis identified 33 modules of coordinately expressed genes. When pathway analysis was performed on genes with a statistically significant difference in expression between treatment condition and control, genetic structure, cell cycle regulation, and sexual development were prominently featured. Hormone receptor pathways affecting chromatin structure, genetic material transcription alteration and apoptotic signaling were enriched in AGCT tumors Our final analytical cohort included nine AGCT tumors from MCW and 88 normal ovary samples from GTEx. A total of 15,992 DEGs were identified (74 downregulated and 15,918 upregulated; FDR < 0.05). GSEA analysis was performed with gene ontology terms. Statistically significant DEGs including ESR1, ESR2, AR, CYP17, CYP19, FSH, LH, GnRHR, PR, and PRLH were identified. In the GO cellular components terms, one included AR. In GO biological pathways terms, 88 pathways included AR and 49 included PR. In the GO metabolic functions terms, 9 pathways included AR and 5 included PR. All ontology terms including AR are illustrated in . The highest enrichment score was seen in “cellular response to testosterone stimulus.” In addition to changes in hormone signaling pathways, ESR1-containing pathways affecting chromatin structure, DNA and RNA transcription were increased in tumor samples compared to normals. Apoptotic signaling pathways containing AR were altered as well. Quantitative PCR validated RNA sequencing results in known cancer-related genes We selected and validated 3 upregulated genes (MT3, GPR146, and GAL3ST1) and 3 downregulated genes (CNN1, FAM20C, and TCF19) from RNAseq experiments. These were selected from the DEG list, amongst those with the highest and lowest logFC list for their possible physiologic relevancy. Relative mRNA expression of these genes in control and combination experimental samples is shown in Supplementary Figure 2. Analysis of qPCR data confirmed higher average expression of MT3, GPR146, and GAL3ST1 in experimental samples versus control, while average expression of CNN1, FAM20C, and TCF19 was lower in experimental samples compared to control.

The combination of leuprolide acetate, anastrozole, and darolutamide for recurrent AGCT was shown by our group to be well tolerated and effective in a case series (Citation21). In this study, we sought to generate hypotheses for the mechanism of action of this combination. We identified limited suppression of tumor cell proliferation when anastrozole or darolutamide was utilized as a single agent, and no effect with leuprolide acetate alone. When used in combination, anastrozole and darolutamide demonstrated enhanced suppression of cell proliferation. Although limited, darolutamide did suppress growth on its own better than anastrozole and was most effective in combination with anastrozole. Although leuprolide acetate did not add additional benefit in vitro, these in-vitro experiments were not able to simulate an intact hypothalamic-pituitary-ovarian (HPO) axis, as would be seen in a patient with AGCT. The immune microenvironment was also not able to be evaluated with these experiments. While most AGCT patients are menopausal (naturally or following oophorectomy), tumor estrogen production remains active. AGCT patients have been shown to have low follicle-stimulating hormone levels and high luteinizing hormone levels (Citation51,Citation52). This is commonly attributed to inhibin presence, but may be similar to the HPO axis function of polycystic ovarian syndrome patients, which is another hyperandrogenic and high-estrogen state. Thus in order to provide complete hormonal blockade, HP”O” inhibition (with the tumor cells acting similar to an ovary) is likely required. Notably, mice overexpressing luteinizing hormone have an increased propensity to develop granulosa cell tumors (Citation53). GnRHR expression was previously identified in KGN cells, and its presence was confirmed in our cell lines (Citation23). While GnRHR and GnRH signaling have been found to modulate apoptosis in FOXL2-wildtype cells, it did not have the same activity in those with FOXL2 overexpression (Citation54). In addition to the lack of HPO axis and immune microenvironment in our in vitro experiments, this altered GnRHR effect may be relevant to our findings showing limited in vitro effect with leuprolide acetate alone. Further investigation is needed. Past attempts have been made to study androgen blockade in AGCT with limited results. These clinical trials were complicated by the difficulties inherent in studying a rare and slow-growing tumor. While anastrozole alone has been shown to exhibit clinical benefit in the treatment of AGCT, the clinical benefit consists mainly of stable disease with a low rate of objective tumor response. The reason for this limited benefit was previously unknown (Citation4). A possible mechanism may be the increased androgen concentration in anastrozole-treated cells, which has previously been shown in in vitro breast cancer studies (Citation55). While we did not see a significant difference in gene expression in anastrozole-treated cells, androgen signaling pathways were reduced in combination treatment cells. Pathways involving AR also appeared important in tumor analyses; thus, further investigation is warranted. Based on the pathway analysis of DEGs detected in the combination treatment cells, we hypothesize that combination treatment may increase tumor antigen peptide production as well as MHC I expression on tumor cells. Our results cannot verify this mechanism directly and this would require further investigation. Using preclinical prostate cancer models, Chesner et al. demonstrated that AR downregulated MHC 1 expression and that inhibition of AR improved T cell response. AR blockade could potentially allow increased antigen presentation and subsequent recognition by CD8+ killer T cells (Citation56). Steroid hormones, including androgens, are known to reduce CD8+ T cell activation (Citation57,Citation58). The immune microenvironment cannot be replicated in our in vitro experiments, so the totality of this effect would not be seen here. Presence of CD8+ T cell activation in recurrent and primary AGCT has been previously studied; CYP19A1 (aromatase) was upregulated in recurrent tumors as compared to primary, but no difference was seen in the proportions of T cells present (Citation59). This relationship has not been studied in AGCT in the setting of hormone suppression. However, Guan et al. (Citation58) demonstrated androgen suppression of CD8+ T cell function and also showed tumor regression in mice implanted with immunotherapy-resistant prostate cancers upon treatment with enzalutamide and an anti-PDL1 antibody. When CD8+ T cells were depleted, benefit of combination with enzalutamide and anti-PDL1 antibody was not seen (Citation58). This strategy of antihormonal therapy followed by immune checkpoint blockade has been reported effective in three castration-resistant prostate cancer patients (Citation60). The phase II KEYNOTE-199 investigated the addition of pembrolizumab to ongoing enzalutamide therapy in metastatic castration-resistant prostate cancer and found modest response with a tolerable safety profile but the randomized double-blind phase 3 KEYNOTE-641 trial of pembrolizumab and enzalutamide in metastatic castration-resistant prostate cancer did not demonstrate improvement in overall survival or progression-free survival (Citation61,Citation62). Biomarker-directed immunotherapy may be possible in prostate cancer as well, since CDK12 expression has been associated with genomic instability, increased T-cell infiltration and CDK4/6 inhibitor sensitivity (Citation63). Similar biomarkers may merit investigation in AGCT. Five GCT patients were included in a basket trial of pembrolizumab in advanced solid tumors, and while there were no objective responses, durable disease control greater than 12 months was seen in two of five patients (Citation64). An additional hypothesized mechanism of action of the drug combination treatment is supported by the chromatin remodeling ontologies seen in treated cell lines. This is consistent with published literature in the breast cancer space. Mourad et al. (Citation65) showed that chromatin remodeling and differential loci interactions allow for increased transcription upon estrogen treatment of breast cancer cells. Estrogen was separately shown to be required for large-scale chromatin unfolding, exposing coordinately-regulated breast cancer genes (Citation66). Strengths of this study include the utilization of primary, patient-derived AGCT cell lines and a novel drug combination with medications that are already clinically available. We established efficacy in 2D monolayer cultures, as well as in more physiologically-relevant 3D spheroid models. RNA sequencing analysis allowed for identification of possible underlying mechanisms of the combination treatment and highlighted a role for future exploration of the immune microenvironment in AGCT. DEGs selected for qPCR validation were also relevant to our pathway analysis. MT3, upregulated with combination treatment, has previously been found to be decreased with testosterone treatment and increased with bicalutamide treatment in prostate cancer cells (Citation67). MT3 up- and down-regulation are associated with carcinogenesis in various cancer types (Citation68). GPR146, upregulated in treated cells, is associated with cholesterol regulation, a steroid hormone precursor (Citation69). GAL3ST1, upregulated with treatment, has been implicated in renal cell carcinoma immune evasion (Citation70). CNN1, downregulated in treatment conditions, is implicated in oncogenesis in multiple cancers including gastric cancer (Citation71). It is also thought to confer aromatase inhibitor sensitivity in estrogen-positive breast cancer (Citation72). FAM20C, downregulated in treated cells, has been found to contribute to proliferation and migration in glioma cells. TCF19, downregulated in treatment cells, has been found to contribute to proliferation of lung and colorectal cancers (Citation41,Citation73,Citation74). Therefore, the up and downregulation of each of these genes is also concordant with oncologic benefit.

include the fact that bulk RNA sequencing does not provide temporality or definitive evidence of this mechanism. Additionally, we were unable to replicate the HPO axis or the tumor microenvironment in vitro, and thus were unable to investigate the effect of these important in vivo variables. This would not be possible without extraction of hypothalamus, pituitary and ovarian tissue and development of an appropriate in vitro system, or in vivo model (Citation75). We were unable to immortalize cell line D in 2D culture, and thus unable to perform STR authentication. This does introduce the possibility that it was contaminated by one of our other cell lines. It was only used for spheroid experiments, however, and thus involved in only a small portion of the data presented here. Regarding the lack of detectable DEGs in cells treated with anastrozole alone, this may be due to a small effect size that was not detectable with the power of our experiments. As the in vitro half life of anastrozole is not well-delineated, we did perform extractions at multiple time points to exclude this as a source of technical error (Citation76). Extractions had initially been performed at the time point at which dose response curves had been optimized; extractions were then trialed at shorter time points as RNA is unstable and its expression may be seen sooner than its biologic effect (Citation77,Citation78). RNA sequencing is also not able to identify important protein changes caused by treatment. Future experiments in proteomic alterations induced by these drugs will be invaluable to our understanding of drug mechanisms. Evaluation of androgen levels after aromatase inhibitor treatment may also warrant further investigation to delineate mechanism. A prospective investigation of these drugs in recurrent AGCT patients is currently being conducted in NRG-GY033 A Phase II Study of Androgen Receptor (AR) Inhibition by Darolutamide in Combination with Leuprolide Acetate and Exemsetane in Recurrent Adult-type Ovarian Granulosa Cell Tumor, clinicaltrials.gov identifier NCT06169124). Biomarkers predictive of clinical response including ER, PR, and AR will be investigated as part of this trial. Doses already known to be physiologically relevant were used in this trial, rather than attempting to translate doses from this study (Citation21,Citation79). Vasomotor symptoms of menopause were the only dose-limiting toxicity in a prior case series (Citation21). The toxicity profile of complete hormonal blockade is well-described in men but not women, with toxicities including decreased bone mineral density, metabolic changes, sexual dysfunction, hot flashes, and fatigue (Citation80). Mitigation strategies are available (Citation80).

Hormonal blockade, an effective treatment for prostate cancer, may be a promising option for treating adult-type granulosa cell tumors of the ovary. Our study has established a rationale for promoting androgen receptor inhibitor-based combination drug therapy. CRediT roles Rebekah M Summey: data curation, formal analysis, investigation, methodology, project administration, validation, visualization, writing – original draft, writing – review & editing Marissa Iden: formal analysis, investigation, methodology, resources, supervision, validation, writing – review & editing Shirng-Wern Tsaih: formal analysis, methodology, resources, supervision, writing – review & editing Rachel Schmidt: formal analysis, methodology, writing – review & editing Deepak Parashar: formal analysis, investigation, methodology, supervision, writing – review & editing Shunping Wang: data curation, resources, writing – review & editing Janet S Rader: conceptualization, investigation, methodology, resources, supervision, writing – review & editing Elizabeth Hopp: conceptualization, data curation, funding acquisition, investigation, methodology, resources, supervision, writing – review & editing. Supplemental material Supplemental Material Download PDF (851.4 KB)Supplemental MaterialAcknowledgements We would like to acknowledge Ling Wang, PhD for her support. This research was supported by the Medical College of Wisconsin Women’s Health Research Program (WHRP) fund and an anonymous donor. This research was partly completed with computational resources and technical support provided by the Research Computing Center at the Medical College of Wisconsin. Disclosure statement No potential conflict of interest was reported by the author(s). This research was presented as a poster at the Society of Gynecologic Oncology Annual Meeting, March 2023, Tampa, FL, USA. Additional information Funding

- Schumer ST, Cannistra SA. Granulosa cell tumor of the ovary. J Clin Oncol. 2003;21(6):1180–1189. doi:10.1200/JCO.2003.10.019. - Färkkilä A, Haltia U-M, Tapper J, McConechy MK, Huntsman DG, Heikinheimo M. Pathogenesis and treatment of adult-type granulosa cell tumor of the ovary. Ann Med. 2017;49(5):435–447. doi:10.1080/07853890.2017.1294760. - Plett H, Ricciardi E, Vacaru V, Ramspott JP, Colombo N, Sehouli J, et al. Adult ovarian granulosa cell tumors: analysis of outcomes and risk factors for recurrence. Int J Gynecol Cancer. 2023;33(5):734–740. doi:10.1136/ijgc-2022-003854. - Banerjee SN, Tang M, O’Connell RL, Sjoquist K, Clamp AR, Millan D, PARAGON investigators, et al. A phase 2 study of anastrozole in patients with oestrogen receptor and/progesterone receptor positive recurrent/metastatic granulosa cell tumours/sex-cord stromal tumours of the ovary: the PARAGON/ANZGOG 0903 Trial. Gynecol Oncol. 2021;163(1):72–78. doi:10.1016/j.ygyno.2021.07.024. - van Meurs HS, van Lonkhuijzen LRCW, Limpens J, van der Velden J, Buist MR. Hormone therapy in ovarian granulosa cell tumors: a systematic review. Gynecol Oncol. 2014;134(1):196–205. doi:10.1016/j.ygyno.2014.03.573. - Shah SP, Köbel M, Senz J, Morin RD, Clarke BA, Wiegand KC, et al. Mutation of FOXL2 in granulosa-cell tumors of the ovary. N Engl J Med. 2009;360(26):2719–2729. doi:10.1056/NEJMoa0902542. - Jamieson S, Butzow R, Andersson N, Alexiadis M, Unkila-Kallio L, Heikinheimo M, et al. The FOXL2 C134W mutation is characteristic of adult granulosa cell tumors of the ovary. Mod Pathol. 2010;23(11):1477–1485. doi:10.1038/modpathol.2010.145. - Weis-Banke SE, Lerdrup M, Kleine-Kohlbrecher D, Mohammad F, Sidoli S, Jensen ON, et al. Mutant FOXL2C134W hijacks SMAD4 and SMAD2/3 to drive adult granulosa cell tumors. Cancer Res. 2020;80(17):3466–3479. doi:10.1158/0008-5472.CAN-20-0259. - Kim JH, Yoon S, Park M, Park HO, Ko JJ, Lee K, Bae J. Differential apoptotic activities of wild-type FOXL2 and the adult-type granulosa cell tumor-associated mutant FOXL2 (C134W). Oncogene. 2011;30(14):1653–1663. doi:10.1038/onc.2010.541. - Georges A, Auguste A, Bessière L, Vanet A, Todeschini AL, Veitia RA. FOXL2: a central transcription factor of the ovary. J Mol Endocrinol. 2014;52(1):R17–33. doi:10.1530/JME-13-0159. - Pannetier M, Fabre S, Batista F, Kocer A, Renault L, Jolivet G, et al. FOXL2 activates P450 aromatase gene transcription: towards a better characterization of the early steps of mammalian ovarian development. J Mol Endocrinol. 2006;36(3):399–413. doi:10.1677/jme.1.01947. - Georges A, L’Hôte D, Todeschini AL, Auguste A, Legois B, Zider A, Veitia RA. The transcription factor FOXL2 mobilizes estrogen signaling to maintain the identity of ovarian granulosa cells. Elife. 2014;3:e04207. doi:10.7554/eLife.04207. - Fleming NI, Knower KC, Lazarus KA, Fuller PJ, Simpson ER, Clyne CD. Aromatase is a direct target of FOXL2: C134W in granulosa cell tumors via a single highly conserved binding site in the ovarian specific promoter. PLoS One. 2010;5(12):e14389. doi:10.1371/journal.pone.0014389. - Leung DTH, Fuller PJ, Chu S. Impact of FOXL2 mutations on signaling in ovarian granulosa cell tumors. Int J Biochem Cell Biol. 2016;72:51–54. doi:10.1016/j.biocel.2016.01.003. - Wu J, Miao C, Lv X, Zhang Y, Li Y, Wang D. Estrogen regulates forkhead transcription factor 2 to promote apoptosis of human ovarian granulosa-like tumor cells. J Steroid Biochem Mol Biol. 2019;194:105418. doi:10.1016/j.jsbmb.2019.105418. - Ciucci A, Ferrandina G, Mascilini F, Filippetti F, Scambia G, Zannoni GF, Gallo D. Estrogen receptor β: potential target for therapy in adult granulosa cell tumors? Gynecol Oncol. 2018;150(1):158–165. doi:10.1016/j.ygyno.2018.05.013. - Definition of complete androgen blockade - NCI Dictionary of Cancer Terms - NCI. https://www.cancer.gov/publications/dictionaries/cancer-terms/def/complete-androgen-blockade(open in a new window). (accessed 2022-09-26). - Garcia-Donas J, Hurtado A, Garrigos L, Santaballa A, Redondo A, Vidal L, Spanish Group for Research in Orphan, Infrequent Tumors (GETHI), et al. Open-label phase II clinical trial of ketoconazole as CYP17 inhibitor in metastatic or advanced non-resectable granulosa cell ovarian tumors: The GREKO (GRanulosa Et KetOconazole) Trial, GETHI 2011-03. Clin Transl Oncol. 2023;25(7):2090–2098. doi:10.1007/s12094-023-03085-w. - García-Donas J, Redondo A, Santaballa A, et al. Phase II clinical trial assessing the efficacy of enzalutamide in advanced non-resectable granulosa cell ovarian tumors: The GREKO III study (GETHI2016-01). Gynecol Oncol. 2024;191:233–239. doi:10.1016/j.ygyno.2024.10.019. - Rodriguez-Moreno J, García-Donas J, Garrigos L, Laínez N, Santaballa A, Redondo A, et al. Open label phase II clinical trial of orteronel (TAK-700) in metastatic or advanced non-resectable granulosa cell ovarian tumors: The Greko II Study—GETHI 2013-01. JCO. 2015;33(15_suppl):TPS5612–TPS5612. doi:10.1200/jco.2015.33.15_suppl.tps5612. - Summey RM, Rader JS, Moh M, Bradley W, Uyar D, Bishop E, et al. A case series of triplet anti-hormonal therapy in androgen receptor-positive recurrent adult ovarian granulosa cell tumor. Gynecol Oncol Rep. 2022;44:101118. doi:10.1016/j.gore.2022.101118. - Kent WJ, Sugnet CW, Furey TS, Roskin KM, Pringle TH, Zahler AM, Haussler D. The human genome browser at UCSC. Genome Res. 2002;12(6):996–1006. doi:10.1101/gr.229102. - Nishi Y, Yanase T, Mu Y, Oba K, Ichino I, Saito M, et al. Establishment and characterization of a steroidogenic human granulosa-like tumor cell line, KGN, that expresses functional follicle-stimulating hormone receptor. Endocrinology. 2001;142(1):437–445. doi:10.1210/endo.142.1.7862. - Carney DN, Bunn PA, Gazdar AF, Pagan JA, Minna JD. Selective growth in serum-free hormone-supplemented medium of tumor cells obtained by biopsy from patients with small cell carcinoma of the lung. Proc Natl Acad Sci U S A. 1981;78(5):3185–3189. doi:10.1073/pnas.78.5.3185. - National Cancer Institute (NCI). A Phase II Study of Androgen Receptor (AR) Inhibition by Darolutamide in Combination With Leuprolide Acetate and Exemestane in Recurrent Adult-Type Ovarian Granulosa Cell Tumor. Clinical trial registration NCT06169124; clinicaltrials.gov, 2024. https://clinicaltrials.gov/study/NCT06169124(open in a new window). (accessed 2023-12-31). - Festuccia C, Gravina GL, Muzi P, Pomante R, Ventura L, Ricevuto E, et al. In vitro and in vivo effects of bicalutamide on the expression of TrkA and P75 neurotrophin receptors in prostate carcinoma. Prostate. 2007;67(12):1255–1264. doi:10.1002/pros.20616. - Pather K, Dix-Peek T, Duarte R, Chetty N, Augustine TN. Breast cancer cell-induced platelet activation is compounded by tamoxifen and anastrozole in vitro. Thromb Res. 2019;177:51–58. doi:10.1016/j.thromres.2019.02.027. - Connor JP, Buller RE, Conn PM. Effects of GnRH analogs on six ovarian cancer cell lines in culture. Gynecol Oncol. 1994;54(1):80–86. doi:10.1006/gyno.1994.1170. - Zheng S, Wang W, Aldahdooh J, Malyutina A, Shadbahr T, Tanoli Z, et al. SynergyFinder Plus: toward better interpretation and annotation of drug combination screening datasets. Genomics Proteomics Bioinformatics. 2022;20(3):587–596. doi:10.1016/j.gpb.2022.01.004. - Roper SJ, Coyle B. Establishing an in vitro 3D spheroid model to study medulloblastoma drug response and tumor dissemination. Curr Protoc. 2022;2(1):e357. doi:10.1002/cpz1.357. - Andrews. FastQC: A Quality Control Tool for High Throughput Sequence Data. Online]2010. http://www.bioinformatics.babraham.ac.uk/projects/fastqc/(open in a new window). - Ewels P, Magnusson M, Lundin S, Käller M. MultiQC: summarize analysis results for multiple tools and samples in a single report. Bioinformatics. 2016;32(19):3047–3048. doi:10.1093/bioinformatics/btw354. - Schneider VA, Graves-Lindsay T, Howe K, Bouk N, Chen HC, Kitts PA, et al. Evaluation of GRCh38 and de novo haploid genome assemblies demonstrates the enduring quality of the reference assembly. Genome Res. 2017;27(5):849–864. doi:10.1101/gr.213611.116. - Bray NL, Pimentel H, Melsted P, Pachter L. Near-optimal probabilistic RNA-seq quantification. Nat Biotechnol. 2016;34(5):525–527. doi:10.1038/nbt.3519. - Robinson MD, McCarthy DJ, Smyth GK. edgeR: a bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics. 2010;26(1):139–140. doi:10.1093/bioinformatics/btp616. - Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, Smyth GK. Limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015;43(7):e47–e47. doi:10.1093/nar/gkv007. - Wu T, Hu E, Xu S, Chen M, Guo P, Dai Z, et al. clusterProfiler 4.0: a universal enrichment tool for interpreting omics data. Innovation. 2021;2(3):100141. doi:10.1016/j.xinn.2021.100141. - Luo W, Brouwer C. Pathview: an R/bioconductor package for pathway-based data integration and visualization. Bioinformatics. 2013;29(14):1830–1831. doi:10.1093/bioinformatics/btt285. - Langfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinf. 2008;9(1):559. doi:10.1186/1471-2105-9-559. - Vivian J, Rao AA, Nothaft FA, Ketchum C, Armstrong J, Novak A, et al. Toil enables reproducible, open source, big biomedical data analyses. Nat Biotechnol. 2017;35(4):314–316. doi:10.1038/nbt.3772. - Xu G, Zhu Y, Liu H, Liu Y, Zhang X. LncRNA MIR194-2HG promotes cell proliferation and metastasis via regulation of miR-1207-5p/TCF19/Wnt/β-catenin signaling in liver cancer. Onco Targets Ther. 2020;13:9887–9899. doi:10.2147/OTT.S264614. - Liu Z, Zhang R, Sun Z, Yao J, Yao P, Chen X, et al. Identification of hub genes and small-molecule compounds in medulloblastoma by integrated bioinformatic analyses. PeerJ. 2020;8:e8670. doi:10.7717/peerj.8670. - Juang H-H, Chung L-C, Sung H-C, Feng T-H, Lee Y-H, Chang P-L, Tsui K-H. Metallothionein 3: an androgen-upregulated gene enhances cell invasion and tumorigenesis of prostate carcinoma cells. Prostate. 2013;73(14):1495–1506. doi:10.1002/pros.22697. - Huang H, Zhang N, Xiong Q, Chen R, Zhang C, Wang N, et al. Elimination of GPR146-mediated antiviral function through IRF3/HES1-signalling pathway. Immunology. 2017;152(1):102–114. doi:10.1111/imm.12752. - Suchanski J, Grzegrzolka J, Owczarek T, Pasikowski P, Piotrowska A, Kocbach B, et al. Sulfatide decreases the resistance to stress-induced apoptosis and increases p-selectin-mediated adhesion: a two-edged sword in breast cancer progression. Breast Cancer Res. 2018;20(1):133. doi:10.1186/s13058-018-1058-z. - Zeisel A, Yitzhaky A, Bossel Ben-Moshe N, Domany E. An accessible database for mouse and human whole transcriptome qPCR primers. Bioinformatics. 2013;29(10):1355–1356. doi:10.1093/bioinformatics/btt145. - Xu X, Nie J, Lu L, Du C, Meng F, Song D. LINC00337 promotes tumor angiogenesis in colorectal cancer by recruiting DNMT1, which suppresses the expression of CNN1. Cancer Gene Ther. 2021;28(12):1285–1297. doi:10.1038/s41417-020-00277-2. - Liu W, Fu X, Li R. CNN1 regulates the DKK1/Wnt/β-Catenin/c-Myc signaling pathway by activating TIMP2 to inhibit the invasion, migration and EMT of lung squamous cell carcinoma cells. Exp Ther Med. 2021;22(2):855. doi:10.3892/etm.2021.10287. - Gong B, Liang Y, Zhang Q, Li H, Xiao J, Wang L, et al. Epigenetic and transcriptional activation of the secretory kinase FAM20C as an oncogene in glioma. J Genet Genomics. 2023;50(6):422–433. doi:10.1016/j.jgg.2023.01.008. - Barbosa-Sampaio HC, Liu B, Drynda R, Rodriguez de Ledesma AM, King AJ, Bowe JE, et al. Nupr1 deletion protects against glucose intolerance by increasing beta cell mass. Diabetologia. 2013;56(11):2477–2486. doi:10.1007/s00125-013-3006-x. - Kitamura S, Abiko K, Matsumura N, Nakai H, Akimoto Y, Tanimoto H, Konishi I. Adult Granulosa cell tumors of the ovary: a retrospective study of 30 cases with respect to the expression of steroid synthesis enzymes. J Gynecol Oncol. 2017;28(4):e31. doi:10.3802/jgo.2017.28.e31. - Ran S, Yu Q, Deng S, Xu L. Luteinizing hormone elevation in ovarian granulosa cell tumor: a case report and review of the literature. J Ovarian Res. 2017;10(1):30. doi:10.1186/s13048-017-0327-2. - Keri RA, Lozada KL, Abdul-Karim FW, Nadeau JH, Nilson JH. Luteinizing hormone induction of ovarian tumors: oligogenic differences between mouse strains dictates tumor disposition. Proc Natl Acad Sci U S A. 2000;97(1):383–387. doi:10.1073/pnas.97.1.383. - Cheng J-C, Klausen C, Leung PCK. Overexpression of wild-type but not C134W mutant FOXL2 enhances GnRH-induced cell apoptosis by increasing GnRH receptor expression in human granulosa cell tumors. PLoS One. 2013;8(1):e55099. doi:10.1371/journal.pone.0055099. - Hole S, Pedersen AM, Hansen SK, Lundqvist J, Yde CW, Lykkesfeldt AE. New cell culture model for aromatase inhibitor-resistant breast cancer shows sensitivity to fulvestrant treatment and cross-resistance between letrozole and exemestane. Int J Oncol. 2015;46(4):1481–1490. doi:10.3892/ijo.2015.2850. - Dersh D, Hollý J, Yewdell JW. A few good peptides: MHC class I-based cancer immunosurveillance and immunoevasion. Nat Rev Immunol. 2021;21(2):116–128. doi:10.1038/s41577-020-0390-6. - Chakraborty S, Pramanik J, Mahata B. Revisiting steroidogenesis and its role in immune regulation with the advanced tools and technologies. Genes Immun. 2021;22(3):125–140. doi:10.1038/s41435-021-00139-3. - Guan X, Polesso F, Wang C, Sehrawat A, Hawkins RM, Murray SE, et al. Androgen receptor activity in T cells limits checkpoint blockade efficacy. Nature. 2022;606(7915):791–796. doi:10.1038/s41586-022-04522-6. - Khlebus E, Vuttaradhi VK, Welte T, Khurana N, Celestino J, Beird HC, et al. Comparative tumor microenvironment analysis of primary and recurrent ovarian granulosa cell tumors. Mol Cancer Res. 2023;21(5):483–494. doi:10.1158/1541-7786.MCR-22-0623. - Markowski MC, Shenderov E, Eisenberger MA, Kachhap S, Pardoll DM, Denmeade SR, Antonarakis ES. Extreme responses to immune checkpoint blockade following bipolar androgen therapy and enzalutamide in patients with metastatic castration resistant prostate cancer. Prostate. 2020;80(5):407–411. doi:10.1002/pros.23955. - Graff JN, Tagawa ST, Hoimes CJ, Gerritsen WR, Vaishampayan UN, Elliott T, et al. Pembrolizumab plus enzalutamide for enzalutamide-resistant metastatic castration-resistant prostate cancer (mCRPC): updated analyses after one additional year of follow-up from cohorts 4 and 5 of the KEYNOTE-199 Study. JCO. 2021;39(15_suppl):5042–5042. doi:10.1200/JCO.2021.39.15_suppl.5042. - Graff JN, Liang LW, Kim J, Stenzl A. KEYNOTE-641: a phase III study of pembrolizumab plus enzalutamide for metastatic castration-resistant prostate cancer. Future Oncol. 2021;17(23):3017–3026. doi:10.2217/fon-2020-1008. - Ghose A, Moschetta M, Pappas-Gogos G, Sheriff M, Boussios S. Genetic aberrations of DNA repair pathways in prostate cancer: translation to the clinic. Int J Mol Sci. 2021;22(18):9783. doi:10.3390/ijms22189783. - How JA, Jazaeri A, Westin SN, Sood AK, Ramondetta LM, Xu M, et al. The clinical efficacy and safety of single-agent pembrolizumab in patients with recurrent granulosa cell tumors of the ovary: a case series from a phase II Basket trial. Invest New Drugs. 2021;39(3):829–835. doi:10.1007/s10637-020-01043-9. - Mourad R, Hsu P-Y, Juan L, Shen C, Koneru P, Lin H, et al. Estrogen induces global reorganization of chromatin structure in human breast cancer cells. PLoS One. 2014;9(12):e113354. doi:10.1371/journal.pone.0113354. - Rafique S, Thomas JS, Sproul D, Bickmore WA. Estrogen-induced chromatin decondensation and nuclear re-organization linked to regional epigenetic regulation in breast cancer. Genome Biol. 2015;16(1):145. doi:10.1186/s13059-015-0719-9. - Otsuka T, Hamada A, Iguchi K, Usui S, Hirano K. Suppression of metallothionein 3 gene expression by androgen in LNCaP prostate cancer cells. Biomed Rep. 2013;1(4):614–618. doi:10.3892/br.2013.107. - Si M, Lang J. The roles of metallothioneins in carcinogenesis. J Hematol Oncol. 2018;11(1):107. doi:10.1186/s13045-018-0645-x. - Wilkins BP, Finch AM, Wang Y, Smith NJ. Orphan GPR146: an alternative therapeutic pathway to achieve cholesterol homeostasis? Trends Endocrinol Metab. 2022;33(7):481–492. doi:10.1016/j.tem.2022.04.008. - Robinson CM, Poon BPK, Kano Y, Pluthero FG, Kahr WHA, Ohh M. A hypoxia-inducible HIF1-GAL3ST1-sulfatide axis enhances ccRCC immune evasion via increased tumor cell-platelet binding. Mol Cancer Res. 2019;17(11):2306–2317. doi:10.1158/1541-7786.MCR-19-0461. - Zhou H, Ke J, Liu C, Zhu M, Xiao B, Wang Q, et al. Potential prognostic and immunotherapeutic value of calponin 1: a pan-cancer analysis. Front Pharmacol. 2023;14:1184250. doi:10.3389/fphar.2023.1184250. - Jia X, Liu G, Cheng J, Shen Z, Shao Z. CYR61 confers the sensitivity to aromatase inhibitor letrozole in ER positive breast carcinoma. Curr Cancer Drug Targets. 2017;17(2):191–197. doi:10.2174/1568009616666160426125058. - Du W-B, Huang Z, Luo L, Tong S-P, Li H-Q, Li X, et al. TCF19 aggravates the malignant progression of colorectal cancer by negatively regulating WWC1. Eur Rev Med Pharmacol Sci. 2020;24(2):655–663. doi:10.26355/eurrev_202001_20042. - Zhou ZH, Chen G, Deng C, Tang JM, Xie L, Zhou HY, et al. TCF19 contributes to cell proliferation of non-small cell lung cancer by inhibiting FOXO1. Cell Biol Int. 2019;43(12):1416–1424. doi:10.1002/cbin.11189. - Moidel MA, Belz EE, Czambel RK, Rubin RT, Rhodes ME. Novel in vitro perfusion system for the determination of hypothalamic-pituitary-adrenal axis responses. J Pharmacol Toxicol Methods. 2006;53(3):264–271. doi:10.1016/j.vascn.2005.10.001. - Kamdem LK, Liu Y, Stearns V, Kadlubar SA, Ramirez J, Jeter S, et al. In vitro and in vivo oxidative metabolism and glucuronidation of anastrozole. Br J Clin Pharmacol. 2010;70(6):854–869. doi:10.1111/j.1365-2125.2010.03791.x. - Changes in RNA in Relation to Growth of the Fibroblast. IV Alterations in theproduction and processing of mRNA and rRNA in resting and growing cells. J Cell Biol. 1976;71(3):933–938. doi:10.1083/jcb.71.3.933. - Wilusz CJ, Wilusz J. Bringing the role of mRNA decay in the control of gene expression into focus. Trends Genet. 2004;20(10):491–497. doi:10.1016/j.tig.2004.07.011. - Fishman A, Kudelka AP, Tresukosol D, Edwards CL, Freedman RS, Kaplan AL, et al. Leuprolide acetate for treating refractory or persistent ovarian granulosa cell tumor. J Reprod Med. 1996;41(6):393–396. - Nguyen PL, Alibhai SMH, Basaria S, D’Amico AV, Kantoff PW, Keating NL, et al. Adverse effects of androgen deprivation therapy and strategies to mitigate them. Eur Urol. 2015;67(5):825–836. doi:10.1016/j.eururo.2014.07.010.