Progestin Significantly Inhibits Carcinogenesis in the Mogp-TAg Transgenic Mouse Model of Fallopian Tube Cancer

Epithelial ovarian cancer (EOC) is the fifth leading cause of cancer-related deaths among women in the United States and causes over 140,000 deaths annually worldwide. The long-term survival of women with ovarian cancer has only improved modestly over the past few decades despite intensive research efforts directed towards early detection and treatment (1). Progress in the fight against ovarian cancer has been hampered by a number of factors, including late diagnosis, high molecular/genetic heterogeneity present in most ovarian cancers, and the lack of a known precursor lesion that can be effectively targeted for prevention.

Recent evidence suggests that the cell of origin for a majority of high-grade EOC resides in the distal, fimbriated end of the fallopian tube. The early events of ovarian carcinogenesis are now thought to involve accumulation of mutated p53 protein in secretory cells in the fallopian tube epithelium (FTE), leading to the p53 signature, the earliest putative cancer precursor lesion detected in the fimbria. Subsequent acquisition of secondary genomic alterations ultimately leads to the development of serous tubal intraepithelial carcinoma (STIC), which spreads via shedding/exfoliation directly to the adjacent ovary and into the abdominal cavity (2, 3). The STIC lesions likely evolve in an inflammatory, oxidative, and DNA-toxic microenvironment and are characterized by distinct features, including loss of polarity, epithelial tufting, nuclear pleomorphism, abnormal p53 expression, and a high proliferative index. The discovery that these changes in the FTE represent the earliest steps in ovarian carcinogenesis opens the door to the development of pharmacologic strategies that arrest and/or reverse the early transformative events in the fimbria, with the potential to decrease ovarian cancer incidence and mortality through prevention.

Extensive epidemiologic evidence has shown that routine use of the combination estrogen–progestin oral contraceptive confers a remarkable 30% to 50% reduction in the risk of developing subsequent ovarian cancer, suggesting that a robust pharmacologic approach for the prevention of ovarian cancer is feasible (4, 5). On the basis of our research findings in vitro and in primate models, we believe that the progestin component of the oral contraceptive functions as a chemopreventive agent in the gynecologic tract by activating molecular pathways, such as apoptosis, that serve to clear genetically damaged cells in the ovary and endometrium (6, 7). Although the biological mechanisms underlying the cancer-protective effect of progestins in the human fallopian tube are not yet well characterized, a recent randomized trial in women at increased genetic risk of ovarian cancer further supported the notion that progestins may clear genetically damaged cells from the fallopian tube. In this study, the karyometric signature of the fallopian tube was noted to deviate from normal in women at high risk of ovarian cancer, and became significantly less aberrant after treatment with the progestin levonorgestrel (8).

The murine oviduct–specific glycoprotein promotor-driven simian virus 40 large T-Antigen (mogp-TAg) transgenic mouse model of fallopian tube cancer recapitulates the cellular and molecular changes typical of human ovarian and fallopian tube cancers (9, 10). The T-Antigen affects both p53, which is the most common mutated gene in high-grade serous human ovarian cancer (96%), as well as the RB pathway, which has also been implicated in ovarian carcinogenesis and shown to be dysregulated in 67% of ovarian cancers (11, 12). The mogp-TAg mouse develops fallopian tube tumorigenesis starting at 7 weeks of age (10). Between 7 and 12 weeks of age, the mogp-TAg mouse exhibits tubal lesions that are dysplastic, characterized by epithelial stratification and marked nuclear atypia, including mitotic bodies, enlarged nuclei and irregular chromatin, and p53 alterations resembling STICs described in the human fallopian tube. The goal of this study was to gather further preclinical evidence in support of progestins as ovarian cancer preventives by directly testing the cancer-preventive effects of the progestin depo-medroxyprogesterone acetate (DMPA) in the fallopian tube of the mogp-TAg mouse. In addition, we sought to characterize biological mechanisms underlying the chemopreventive effect of DMPA in the fallopian tube.

The mogp-TAg mice were derived as described previously; a 2.2‐kb segment of the 5′‐flanking sequence of the mouse oviduct–specific glycoprotein gene is used to drive expression of the TAg, leading to development of tumors in the gynecologic tract (9, 10). The 5-week-old female mogp-TAg mice were obtained from The Jackson Laboratory. The mice were then housed at the NorthShore University HealthSystem Center for Comparative Medicine with ad libitum access to a phytoestrogen-free diet and to water in 12-hour light/dark cycles. To characterize the natural history of neoplastic transformation in the reproductive tract of the mogp-TAg mice, animals (four per group) were euthanized at 5, 8, and 12 weeks of age. For each case, 5-μm cross sections of the fallopian tube were examined for pathologic changes using hematoxylin and eosin (H&E) staining.

Forty female mice were injected subcutaneously on the left dorsal side between the shoulder and neck scruff at 5 weeks of age with either vehicle (DMPA solvent) or DMPA (1 mg/mouse; Amphastar Pharmaceuticals Stock # 5400) and divided into two cohorts of 20 mice each, to be euthanized at 8 and 12 weeks of age respectively. The 1 mg dose of DMPA per mouse was chosen based on published evidence demonstrating that this dose is well tolerated and results in serum levels of DMPA comparable with that observed clinically in humans (13). The weight of the mice in each cohort was obtained every week until the end of trial. An additional cohort of ten 7-week-old mice was treated with either vehicle (n = 5) or 1 mg DMPA (n = 5) and euthanized at 12 weeks of age. In all of the trials, euthanasia was performed by using isoflurane inhalation followed by blood collection and surgical removal of the reproductive tracts, which were fixed in 10% formalin, paraffin-embedded, and cut into 5-μm sections for histopathologic analysis. The animal protocol was approved by the NorthShore University HealthSystem Institute Animal Care and Use Committee. For gross measurements, a Vernier caliper was used to obtain the length and width of the uterine horn and the diameter of the fallopian tube. The volume of the fallopian tube was calculated using 4/3πr³ (r = radius). To examine the effects of DMPA on apoptosis, 5- and 7-week-old mice (n = 4–5 per age group) were treated with vehicle or 1 mg DMPA for 8 and 16 hours.

For morphologic analyses, 5 μm H&E sections of the reproductive tract in 5-, 8-, and 12-week-old mice were examined histologically on an Olympus BX43 at 10× to 40× magnification in a blinded fashion by a veterinary pathologist (J.M. Cline) for evidence of neoplasia or other abnormalities. Fiji (ImageJ) was used to measure the fallopian tube cross-sectional area (pixel per unit area) of H&E section images captured at 4× magnification by freehand drawing a border around each cross section.

IHC using 5 μm fallopian tube sections was accomplished by using the University of Chicago Human Tissue Resource Center and the NorthShore University HealthSystem Histology Core. For IHC and immunofluorescence (IF) staining on mouse tissue, standard protocols were used for p53 (Novacastra NCL-2-p53-CMp, 1:400), PAX8 (ProteinTech 10336–1-AP, 1:300), cleaved caspase-3 (Cell Signaling Technology #9664, 1:200), acetylated tubulin (Cell Signaling Technology 5355S, 1:1,000), and Ki67 (Cell Signaling Technology D3B5 12202S, 1:200). IHC images were captured at 2× to 10× magnification using MMI CellScan. A Leica DMi8 inverted confocal fluorescence microscope was used to capture IF photo-stitched images and images at 10× to 20× magnification. The 5 μm cross sections of a human fallopian tube containing a STIC were obtained from the NorthShore University HealthSystem tissue bank via an approved Institutional Review Board protocol to perform IF for p53 (DO-1: sc-126, 1:250) and PAX8 (ProteinTech 10336–1-AP, 1:300). Secondary goat anti-mouse, 488 (Thermo Fisher Scientific A28175) and donkey anti-rabbit, 555 (Thermo Fisher Scientific A31572) were used for IF staining at 1:10,000 dilution.

To obtain the percentage cross-sectional area of p53 3,3′-diaminobenzidine (DAB) immunoreactivity in the fallopian tube, Fiji (ImageJ) was used to calculate the p53 immuno-ratio (area of the fallopian tube cross section that is occupied by p53-positive cells; ref. 14). Histology slides of the fallopian tube from the vehicle- and DMPA-treated groups (one each per mouse, 10 mice per group) were examined in a blinded fashion. Because the fallopian tube is convoluted, each slide had approximately six fallopian tube cross sections. The total number of p53 signatures and STIC lesions on each slide were counted microscopically using p53 IHC-stained sections. To obtain the average number of p53 signatures and STIC lesions per fallopian tube cross section per treatment group, the total number of p53 signatures and STIC lesions per animal was divided by the number of fallopian tube cross sections. We then calculated the averages of these numbers to obtain the average number of p53 signatures and STIC lesions per fallopian tube cross section per treatment group.

Fiji (ImageJ) was used to count the number of Ki67⁺ (DAB staining) and Ki67⁻ (hematoxylin) cells in the epithelial layer of the fallopian tube at 10× magnification in vehicle- and DMPA-treated groups. Ki67⁺ cells were expressed as a percentage of total cells counted (Ki67⁺ and Ki67⁻). An Olympus BX43 light microscope was used to count the number of cleaved caspase-3–positive cells from IHC sections at 20×. The average number of caspase-3–positive cells per fallopian tube cross-sectional area was calculated.

Simian virus 40 large T-Antigen p53-inactivated human FTE cell lines FT190 and FT194 were obtained from Dr. Ronny Drapkin (Department of OBGYN, University of Pennsylvania, Biomedical Research, Philadelphia, PA) (15). The FT190 and FT194 cell lines were cultured according to Karst and colleagues (16), and treated with different concentrations of medroxyprogesterone acetate (MPA; Sigma; M1629), progesterone (Sigma; P8783), levonorgestrel (Sigma; N2260), and norethindrone (Sigma; N4128) for 72 hours on 6-cm cell culture plates. At the end of treatment, 20 to 40 μg of protein was used for standard Western blotting analysis for cleaved caspase-3. To measure protein fold changes, densitometry values of cleaved caspase-3 normalized to actin protein bands were quantified using LabWorks software for both untreated and progestin-treated FTE cells. FTE cell viability, cytotoxicity, and caspase activity were examined using ApoTox-Glo Triplex Assay (Promega, #G6321) according to the manufacturer's instructions using 5,000 cells per well on a 96-well white plate in phenol-free medium with different concentrations of MPA.

Mouse plasma estradiol (E2) concentrations at each treatment time point were determined using a Calbiotech Estradiol ELISA Kit (ES180S-100) according to the manufacturer's instructions. A linear standard curve was generated by plotting reference standards (0–30 ρg/mL) against absorbance values and calculating each specimen accordingly.

Data comparisons were performed using two-tailed unpaired t tests. P values of 0.05 or less were considered statistically significant. Statistical analysis was carried out using Windows, GraphPad Prism 9.0 Software (www.graphpad.com).

The data generated in this study are available upon request from the corresponding author.

With the limitation of not having a different species-specific primary antibody to perform IF staining on the mogp-TAg mouse fallopian tube with both p53 and PAX8 to show their colocalization in lesions, we used a human fallopian tube section with a known STIC lesion to probe with anti-mouse p53 and anti-rabbit PAX8. PAX8 is a nuclear protein specific to secretory cells (cell type in the FTE that accumulates p53 and transforms into STICs and then cancers). Notably, fallopian tube cancers are PAX8-positive, thus supporting a tubal secretory cell origin. Here, we show that p53 and PAX8 protein colocalized in the histologic STIC lesion (Supplementary Fig. S1A and S1B; triangle). Using an adjacent 5 μm tissue section to that in Supplementary Fig. S1A, we also show that the PAX8⁺ cells that were p53⁺ were not ciliated cells, as these ciliated cells were PAX8-negative and stained for the cilia protein marker acetylated tubulin (Supplementary Fig. S1C, arrowhead). We then used IF to examine secretory cells (PAX8-red), p53-positive cells (red), and ciliated cells [acetylated tubulin (ac-tub)-green; Supplementary Fig. S1D] in the mouse fallopian tube. In the mouse (Supplementary Fig. S1D), p53-positive cells were nonciliated and morphologically resembled human secretory cells (Supplementary Fig. S1A and S1B). On the basis of IF staining, p53-positive cells clustered morphologically into p53 signature (arrowhead), STIC (triangle), and invasive carcinoma (Supplementary Fig. S1D, arrow).

We then used H&E and IHC sections stained for p53 to assess morphologic changes and accumulation of p53-positive cells over time in the fallopian tube of 5-, 8-, and 12-week-old untreated mice (n = 4 mice per group). Similar to reports by Sherman–Baust, there was a significant increase in the size (volume) of the fallopian tube as the mice aged (Supplementary Fig. S2A and S2B; P < 0.00005). Overall, there was a significant increase in the incidence of diffuse epithelial hyperplasia, focal/multifocal epithelial hyperplasia, diffuse smooth muscle hyperplasia, segmental smooth muscle hyperplasia, and adenocarcinoma in the fallopian tube, most notable in 8- and 12-week-old mice (Supplementary Fig. S2A and S2C). In addition, there was a progressive increase in the incidence of p53-positive cells in the fallopian tube epithelial layer from 5 to 12 weeks (Supplementary Fig. S2C). At 5 weeks, there were small clusters of p53-positive cells in the epithelial layer (triangle); however, the majority of these cells were diffusely scattered. At 8 weeks, there were clusters of p53-positive cells in the epithelial layer with features of p53 signatures and, less commonly, STIC. By 12 weeks of age, the p53-positive cell clusters had features more similar to STIC and invasive carcinoma than p53 signature (Supplementary Fig. S2C, arrowhead).

Next, we tested the ability of DMPA to inhibit tumorigenesis in the fallopian tube of the mogp-TAg mice. The 5-week-old mice were injected with vehicle or 1 mg DMPA/mouse and then euthanized at 8 and 12 weeks of age (Fig. 1A). Grossly, at 8 weeks, the fallopian tube volume/cross-sectional area (Fig. 1B and C; Supplementary Fig. S3A and S3B; P < 0.0005) and the uterine horn length and width (Supplementary Fig. S3A and S3B; P < 0.0005) were significantly smaller in the DMPA group compared with the vehicle group. Similarly, at 12 weeks, the fallopian tube volume/cross-sectional area (Fig. 1B and C; Supplementary Fig. S3C and S3D; P < 0.0005) and the uterine horn length and width (P < 0.0005) were significantly smaller in the DMPA group compared with the vehicle group. Interestingly, there were no differences in the average weight of the mice in the vehicle- and DMPA-treated groups from 5 to 10 weeks; however, at 11 and 12 weeks, the vehicle-treated mice were significantly heavier than the DMPA group (Supplementary Fig. S3E and S3F; P < 0.05). Histologically at 8 weeks, the fallopian tube of the vehicle-treated mice developed invasive adenocarcinoma with moderate multifocal hyperplasia and atypia (n = 8) and multifocal hyperplasia and atypia (n = 1) while one had no significant lesion (NSL; Fig. 1D). In comparison, 9 of 10 mice in the DMPA-treated group had NSL while one developed a focal epithelial hyperplasia in the fallopian tube (P < 0.00005). The fallopian tube in the 12-week-old vehicle-treated mice developed invasive adenocarcinoma with severe multifocal epithelial hyperplasia in 80% of the cases, while the DMPA-treated group developed invasive adenocarcinoma with severe multifocal hyperplasia in 20% of the cases (Fig. 1D, P = 0.005; Supplementary Fig. S4). These observations demonstrate that DMPA treatment is preventing neoplastic transformation in the fallopian tube of the mogp-TAg mice.

At the end of both the 8 and 12 week trials, there were strikingly fewer acetylated tubulin–positive (ciliated) cells in the vehicle groups compared with the DMPA groups (Fig 2A and B). Using H&E sections and PAX8 IHC to stain the nuclei of the secretory cells in the epithelial layer, there was more epithelial tufting, loss of polarity, and pleomorphic nuclei in the vehicle-treated groups compared with the DMPA groups at 8 and 12 weeks (Fig 2C and D), respectively. Overall, the epithelial layer p53 immuno-ratio was significantly higher (P < 0.000005) in the vehicle-treated mice at 8 weeks compared with the DMPA-treated mice (Fig. 3A and B). At 12 weeks, the cross-sectional area of the FTE on average had more p53-positive cells in the vehicle-treated mice compared with the DMPA-treated mice (Fig. 3C). These data show that DMPA has an inhibitory effect on the number of p53-positive cells in the FTE.

To examine the effect of DMPA on proliferation in the FTE, Ki67 IHC was used to determine the proliferative index of the FTE in vehicle- and DMPA-treated animals. There was significantly higher epithelial proliferation (Ki67⁺ cells) in the vehicle-treated group compared with the DMPA-treated mice at 8 weeks (Fig. 4A and B; P = 0.001). At 12 weeks, the majority of the fallopian tube cross-sectional areas in the DMPA-treated group did not appear to have a large number of Ki67-positive cells when compared with the vehicle-treated group (Fig. 4C). Similar to what has been shown in the human FTE, Ki67-positive cells (arrowhead) clustered into features that were reminiscent of p53 signature and STIC lesions in the epithelial layer in the mogp-TAg mice FTE (Fig. 4B and C; refs. 17–19). PAX8 IHC staining and H&E sections showed that these regions exhibited epithelial stratification and marked nuclear atypia (Fig. 4B and C).

To further examine the inhibitory effects of DMPA on lesion accumulation in the fallopian tube, the number of p53 signatures and STIC lesions were quantified following treatments at 8 and 12 weeks. To ensure that there were no significant differences in the number of fallopian tube cross sections per histologic slide relative to treatment group, a similar number of fallopian tube cross sections were used for each treatment group (Supplementary Fig. S5A). There were significantly more p53 signatures and STICs in the vehicle groups compared with the DMPA groups at 8 and 12 weeks (Fig. 4D). For the average number of lesions per fallopian tube cross section per treatment group, there were five p53 signatures and three STIC lesions in the vehicle group at 8 weeks; in contrast, there was one p53 signature and no STIC lesions in DMPA group (Fig. 4B and D; P = 0.0004 and P < 0.0005, respectively). In 12-week-old mice, there were five p53 signatures and five STIC lesions per fallopian tube cross section per treatment group in the vehicle-treated mice, while there was one p53 signature and one STIC lesion in the DMPA-treated group (Fig. 4C and D; P < 0.0001 and P < 0.0005, respectively). These data further demonstrate that DMPA is inhibiting and/or clearing the genetically compromised cells from the FTE before they can lead to an invasive phenotype.

To determine whether DMPA can inhibit tumorigenesis in older mice, 7-week-old mogp-TAg mice were treated with vehicle or 1 mg DMPA/mouse and euthanized at 12 weeks (Fig. 5A). Similar to the results in studies initiated in 5-week-old mice, there was a significant reduction in the size of the fallopian tube in the DMPA-treated animals (Fig. 5B–D; P < 0.05). In addition, there were significantly more p53 signatures and STICs in the vehicle group compared with the DMPA group at 12 weeks (Fig. 5C and E). On average, there were five p53 signatures and five STIC lesions per cross section in the vehicle group at 12 weeks; in comparison, there were two p53 signatures and two STIC lesions in the DMPA group (Fig. 5E; P = 0.0007). These data suggest that DMPA can inhibit tumorigenesis in older mice; however, it is less efficacious than administering DMPA treatments at 5 weeks.

Because the mogp-TAg transgene expression is estrogen-dependent (9) and progestin may have an indirect or direct effect on estrogen signaling (20–22), we measured the plasma levels of estrogen in the vehicle-treated and DMPA-treated groups. There were no significant differences in the estradiol levels between DMPA- and vehicle-treated mice at 8 and 12 weeks (Supplementary Fig. S5B).

To examine whether apoptosis may be a potential mechanism underlying DMPA clearance of genetically damaged cells from the fallopian tube, 5- and 7-week-old mice were injected with vehicle and 1 mg DMPA, and then euthanized at 8 and 16 hours. There were no differences in the levels of cleaved caspase-3 in the vehicle (n = 3) group compared with DMPA-treated (n = 3) mice at 5 weeks of age at 8 hours (Supplementary Fig. S6A). However, there appeared to be more cleaved caspase-3 in 7-week-old mice treated with 1 mg DMPA for 8 hours (n = 3; Supplementary Fig. S6B) and 16 hours (n = 4) with DMPA compared with vehicle (n = 3; Fig. 6A and B; P < 0.005). Interestingly, the cleaved caspase-3 was observed in p53-positive cell clusters in the DMPA-treated mice (Fig. 6C; arrow and arrowhead); this observation was not seen in vehicle-treated mice (Supplementary Fig. S6C). These in vivo data suggest that DMPA is targeting FTE cells with alterations in p53. To further test whether progestins can induce apoptosis in vitro, p53-inactivated human FTE cell lines FT190 and FT194 were treated with the progestins MPA, progesterone, levonorgestrel, and norethindrone. The progestin MPA decreased cell viability while increasing cytotoxicity and cleaved caspase-3 in FT190 (Fig. 6D). In addition, all of the four progestins examined induced cleaved caspase-3. MPA and progesterone appeared more effective compared with levonorgestrel and norethindrone (Supplementary Fig. S7A–S7C). Our results suggest that progestin targets and clears precursor lesions in the FTE via apoptosis and prevents these cells from developing into a more carcinogenic phenotype (Supplementary Fig. S8).

In this study, we show for the first time marked inhibition of carcinogenesis in the fallopian tube in vivo by administration of the progestin DMPA. Using the mogp-TAg fallopian tube cancer mouse model, we confirmed the previously noted frequent development of precancerous and cancerous lesions in the fallopian tubes of mice, and observed almost complete eradication of these lesions in the DMPA-treated mice. The FTE in mice treated with DMPA also demonstrated preservation of normal architecture, with normal distribution of ciliated and secretory cell types. In 5-week-old untreated mice, we observed p53-positive cells scattered diffusely in the epithelium and stroma of the fallopian tube. At 7 to 8 weeks, the epithelial cells further evolved histologically into p53 signatures and STIC lesions, concomitant with increased proliferation. There was almost complete eradication of these p53-positive cells in the FTE in mice treated with DMPA, suggesting that progestins may be conferring a chemopreventive effect by targeting the earliest known transformative genetic alterations in the FTE. Interestingly, we also observed that progestin was able to inhibit proliferation and eradicate p53-positive cells in the stroma of the fallopian tube as well as in the uterine horn. This observation may suggest that progestin has a global effect in the reproductive tract.

Overall, our data support the hypothesis that progestins reduce the number of morphologically abnormal and genetically damaged FTE cells from the fallopian tube. The underlying biological mechanism(s) remain to be fully characterized. Wu and colleagues suggested that progesterone released during the latter half of the menstrual cycle may function to maintain a healthy FTE by clearing p53-deficient cells through necroptosis (23). In our study, we found that DMPA induced apoptosis, consistent with our prior observation in the primate ovary and endometrium (6, 7), as well as published data in the human endometrium (24). Notably, we did not observe any significant DMPA-induced apoptosis in the FTE of 5-week-old mice, where p53-inactivated cells are scattered in the FTE and fallopian tube stroma; in contrast, we observed marked DMPA-induced apoptosis in the FTE of 7-week-old mice, localizing to areas with p53 signatures and STIC lesion formation. Taken together, our results suggest that even though loss of p53 may be a critical early step in fallopian tube carcinogenesis, additional genetic or epigenetic events beyond p53 inactivation may be required both for neoplastic transformation as well as for genetically abnormal FTE cells to become increasingly vulnerable to progestin-induced apoptosis. Our study also suggests that the optimal timing window for administration of progestin chemoprevention in the fallopian tube may be early in the process of carcinogenesis. Although we found that DMPA inhibited lesion formation in the FTE in all mice studied, the results were less effective when DMPA was initiated at 7 weeks, when STICs had already developed. As the animals age and accumulate genetic damage in the FTE, the FTE in older mice would be more likely to harbor chemoresistant clones of cells that are harder to eradicate because of growing disruption in molecular pathways that are required for optimal progestin effect.

A recent study using a different in vivo model of fallopian tube carcinogenesis derived conclusions completely opposite to ours, suggesting that progesterone may both induce and promote high-grade serous carcinomas originating in the fallopian tube. Using the Dicer-Pten double knockout (DKO) mouse model of fallopian tube carcinogenesis, Kim and colleagues showed that ovariectomized DKO mice developed significantly fewer advanced cancers and had longer life expectancy (25). In the same mice, progesterone treatment completely restored the malignant phenotype seen in the nonovariectomized DKO mice, with high-grade serous cancers starting in the fallopian tubes, as well as marked peritoneal metastases and poorer survival. Blockade of progesterone signaling, either with the antiprogestin mifepristone or with genetic ablation of progesterone receptor, inhibited carcinogenesis in DKO mice similar to ovariectomy. Interestingly, in ovariectomized mice, the addition of estrogen abrogated the carcinogenic effect of progesterone. These data in the Dicer-Pten mouse model contrast with extensive epidemiologic and clinical evidence in humans that have consistently suggested that progestins lower ovarian and fallopian tube cancer risk across the full age spectrum: (i) in premenopausal women use of progestin-containing oral contraceptives is very protective; (ii) similarly, pregnancy, which represents a high progestin state markedly lowers ovarian cancer risk; (iii) in postmenopausal women estrogen-alone hormone replacement therapy has been shown to consistently increase ovarian cancer risk, whereas the addition of progestins to estrogen replacement mitigates the carcinogenic effect of estrogen (26). Although the Dicer-Pten mouse model develops fallopian tube cancers that resemble human serous ovarian cancer both clinically and at the molecular level, there are features of the model that may not mirror the earliest steps in human fallopian tube carcinogenesis. The tumors are driven by Pten, not thought to drive early carcinogenesis in human ovarian and fallopian tube cancers. In addition, the cancers in the Dicer-Pten mouse originate in the stroma of the fallopian tube, suggesting a mesenchymal lineage (nonepithelial). In contrast, the tumors in the mogp-TAg mouse model start in the FTE, and demonstrate early alterations in p53, thought to be the earliest genetic change underlying human fallopian tube carcinogenesis. Thus, the mogp-TAg mouse model may be the better model for studying chemopreventive pharmacologic approaches targeting the early steps in fallopian tube carcinogenesis.

Areas for future research could include in vivo and clinical investigations comparing the chemopreventive efficacy of different progestins and/or different schedules of progestin administration on fallopian tube carcinogenesis. Although our in vitro assays suggest that different progestins may have varying efficacy in inducing apoptosis, it is not known whether one progestin type may be more effective than another in vivo. In addition, if the mechanism underlying the chemopreventive effect of progestins on fallopian tube carcinogenesis primarily involves eradication of genetically abnormal FTE cells via apoptosis, it may be possible to achieve significant chemoprevention via periodic rather than continuous administration of progestin, thereby limiting possible long-term risks/side effects. The mogp-TAg and other recently developed fallopian tube mouse models as reviewed by Zakarya and colleagues (27) may provide an excellent opportunity to address these questions prior to testing the optimal chemopreventive approach in humans.

The optimal target population for an ovarian cancer chemoprevention trial evaluating progestins remains to be determined. Because most ovarian cancers are sporadic, that is unrelated to a hereditary predisposition, the greatest public health impact would be achieved via a pharmacologic preventive approach that can be applied in the general population. A clinical trial involving women at average risk, however, would require very large sample size and many years to complete. It would be more feasible to do a trial in a population whose ovarian cancer risk is enriched/increased, such as in women with one first degree relative with breast or ovarian cancer in whom lifetime ovarian cancer risk is as high as 5%. In addition, if the underlying preventive mechanism of action of progestins involves a cytocidal and not cytostatic effect on genetically damaged cells in the FTE, then a shortened course of administration of progestins may possibly be sufficient to achieve effective chemoprevention. Although progestin-containing regimens are well tolerated and nontoxic, limiting the time of progestin exposure would be expected to lessen risks related to chemoprevention. The ideal timing of initiation of chemoprevention of ovarian and fallopian tube cancers remains to be determined; however, our data suggest it will be most effective if started prior to the development of significant histologic lesions in the fallopian tube. It would be interesting to test a chemoprevention approach that is initiated several years prior to the onset of menopause (e.g., women ages 45–50 years). This would hopefully clear genetically damaged cells from the gynecologic tract just prior to the onset of menopause, thereby lowering subsequent risk ovarian cancer, which typically presents over a decade later.

In conclusion, our study adds to a growing and compelling body of preclinical human and animal data supportive of the notion that a progestin-based pharmacologic strategy has great potential for the chemoprevention of ovarian and fallopian tube cancers. These data include: (i) evidence that progestin-potent oral contraceptives confer greater protection against ovarian cancer than oral contraceptives containing weak progestin formulations (5); (ii) data from the WHO, demonstrating a 60% reduced risk of nonmucinous ovarian cancer in women who have ever used DMPA, a progestin-only contraceptive (28); (iii) in vivo evidence of a chemopreventive effect of progestins against reproductive tract (ovarian and oviductal) tumors in egg-laying hens (29, 30). Although it is not known if progestins directly target cells with aberrant p53 in the hen, it is interesting to note that reproductive tract cancers in the hen frequently harbor alterations in p53 (31), and thus may have a pathogenesis similar to fallopian tube cancers that develop in the mogp-TAg mouse model and in women; (iv) evidence that the use of the levonorgestrel intrauterine device lowers risk of subsequent ovarian cancer (32); (v) evidence that progestins activate surrogate endpoint biomarkers relevant to chemoprevention in the gynecologic tract (6, 7, 24), and (vi) human evidence revealing that progestins clear cells containing abnormal molecular signatures in the fallopian tube (8). Taken together, these data support further clinical investigation of progestins for the chemoprevention of ovarian cancer.

L.G. Thaete reports grants from NIH during the conduct of the study. J.M. Cline reports grants from Roche outside the submitted work. G.C. Rodriguez reports grants from NIH during the conduct of the study; in addition, G.C. Rodriguez has a patent 6028064 issued, a patent 6310054 issued, a patent 6319911 issued, a patent 6511970 issued, a patent 6977250 issued, a patent 6407082 issued, and a patent 7053074 issued. No disclosures were reported by the other authors.

The contents hereof are solely the responsibility of the authors and do not necessarily represent the official views of the funding sources. The funding sources had no role in the design and conduct of the study, nor in collection, management, analysis, and interpretation of the data, or in preparation, review, and approval of the manuscript.

O.L. Nelson: Conceptualization, resources, data curation, supervision, investigation, methodology, writing–original draft, project administration. R. Rosales: Data curation, investigation. J.M. Turbov: Data curation, investigation. L.G. Thaete: Data curation, investigation, methodology. J.M. Cline: Data curation, validation. G.C. Rodriguez: Conceptualization, resources, data curation, supervision, funding acquisition, investigation, methodology, writing–original draft, project administration.

G.C. Rodriguez's research reported in this publication is supported by the NCI-NIH under award number: 1R01CA214606-01A1. Additional financial support was also provided to G.C. Rodriguez by Bears Care and The Matthews Family Foundation.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.