The Impact of Progesterone Receptor Expression on Platinum Sensitivity and Survival Outcomes in Patients With Ovarian Clear Cell Carcinoma

Abstract


Background
In the United States, approximately 22,530 patients were newly diagnosed with ovarian cancer and approximately 13,980 patients died from this disease in 2016, representing the fth leading cause of cancer-related deaths in women [1].According to the Taiwan Cancer Registry, there were 1,507 newly diagnosed ovarian cancer patients and 656 deaths in Taiwan in 2016, representing the seventh leading cause of cancer-related deaths in women.
Most ovarian malignancies are epithelial ovarian cancer (EOC), however differences in the incidence of EOC subtypes have been reported in different areas.Ovarian clear cell carcinoma (OCCC) is unique at cellular and molecular levels.Endometriosis is a precancerous lesion that has been associated with a 3fold increased risk of OCCC, and the incidence of OCCC ranges from 13-28% in Asia, which is much higher than in Western countries (5-10%) [2][3][4][5].According to a report from the Taiwan Cancer Registry, OCCC accounted for 14-18% of all cases of EOC from 2012-2016.OCCC is more commonly detected at an early stage and has a good prognosis.However, it is considered to be platinum resistant in the advanced stage, and the prognosis is worse than high-grade serous carcinoma [6][7][8][9].Therefore, novel treatments are urgently needed, and many molecular targeted therapies and immunotherapies are currently being studied.
The presence of sex steroid hormone receptors in EOC tissue suggests that hormones may play a role in the origin and promotion of this disease [10].Epidemiological evidence strongly suggests that progesterone-containing contraceptives and pregnancy appear to provide a protective effect against EOC, indicating that progesterone receptor (PR) may play a role in predicting the prognosis [11].Although previous studies have reported an association between a strong PR expression and improved diseasespeci c survival in patients with ovarian high-grade serous and endometrioid cancer, the results in OCCC have been con icting due to limited sample sizes and statistical power [12,13].However, differential expressions of PR from atypical endometriosis to OCCC have been reported, suggesting its role in predicting prognosis [14,15].As the association between PR status and chemosensitivity has not been investigated before, the aim of this study was to evaluate the associations among PR expression, platinum sensitivity, and survival in OCCC patients and cell line models.

Patients
We retrospectively reviewed patients with OCCC between January 2008 and December 2016 in Kaohsiung Chang Gung Memorial Hospital who underwent primary surgery followed by more than 4 courses of adjuvant chemotherapy.The stages were assigned according to the International Federation of Gynecology and Obstetrics (FIGO) 2014 system.The study cases were selected based on the availability of tissue for immunohistochemical (IHC) staining.Patients were excluded if su cient tissue samples were not available, they had not undergone chemotherapy or received fewer than 4 courses of chemotherapy, did not have regular follow-up data, or had double cancers.Clinical data including demographics, FIGO stage, carbohydrate antigen-125 (CA-125) level, concurrent endometriosis (de ned as endometriosis presenting at the same site of cancer), platinum sensitivity, PR expression, and survival time were collected from clinical records.Platinum sensitivity was classi ed as being either resistant or sensitive according to the time of relapse (6 months as the cut-off) since nishing rst-line chemotherapy.Clinical parameters including PR expression were compared between the platinumsensitive and platinum-resistant groups.The study protocols were approved by the Institutional Review Board of Chang Gung Memorial Hospital (approval number: 201601487B0C503).

Immunohistochemistry analysis
Para n-embedded ovarian tissues were sliced into 3-µm-thick sections, depara nized in xylene, and rehydrated through an alcohol gradient.Activity of endogenous peroxide was blocked with 3% H2O2 for 15 min.The slides were then treated with 1x citrate buffer pH 6.0 (SIGMA, C9999) for 17 min in a microwave oven for antigen retrieval, and then incubated with the primary PR antibody (1:200, ThermoFisher, MA1-411) at 4ºC overnight, which can detect both the A and B form of PR.IHC staining was performed using an UltraVision Quanto Detection System HRP kit (ThermoFisher, TL-060-QHL) and DAB Quanto reagent (ThermoFisher, TA-125-QHDX) according to the manufacturer's instructions.Finally, the sections were counterstained with hematoxylin and mounted with aqueous mounting media.The slides were visualized at 200X magni cation, and PR staining was scored using the H-score method (range 0-300) obtained by multiplying the tumor nuclei cell intensity (on a scale of 0 to 3) by the percentage of positive tumor nuclei cells (on a scale of 0 to 100) [16].One pathologist (Lan J) blindly scored all of the cases.

PR-overexpressing ovarian cancer cell lines
The plasmids pcDNA3-hPR-A (Cat.No.#89119) and pcDNA3-PRB (Cat.No.#89130) were purchased from Addgene.TOV21G cells were seeded in a 6-well plate at 3 x 10 5 cells per well for 24 h, and transfection was carried out using Lipofectamine 2000 reagent (Invitrogen, Cat.No.#11668019) according to the manufacturer's instructions.After transfection for 48 h, the cells were harvested for Western blot and cell viability analyses.

Cell viability assay
PR-overexpressing cells were seeded in a 96-well plate at 5000 cells per well, and incubated at 37ºC overnight before drug treatment.The cells were treated with different concentrations of cisplatin for 24 h, 48 h, 72 h, and 96 h.PrestoBlue Cell Viability Reagent (10:1) (Invitrogen, A13261) was then added and allowed to incubate at 37ºC for 2-3 h.The absorbance at a wavelength of 570-600 nm was measured using a SpectraMax ABS Absorbance Microplate Reader (Molecular Devices).The viability of cells was expressed as a percentage of absorbance in treated wells relative to that in untreated (control) wells.

Statistical analysis
Receiver operating characteristic (ROC) curve analysis was used to identify optimal cut-off values of PR H-score to predict platinum sensitivity and survival.Progression-free survival (PFS) and overall survival (OS) were de ned as the interval from the date of diagnosis to the rst evidence of progression and death, respectively.Comparisons of median and mean values were performed using the two-sample ttest.Frequency distributions between categorical variables were compared using the chi-square or Fisher's exact test.Multivariate Cox proportional hazards analysis was used to identify the most signi cant independent prognostic factors for PFS and OS.Actuarial rates of survival were estimated with the Kaplan-Meier method, and statistical differences between groups were examined using the log rank test.The Student's t-test was used to assess statistical signi cance between the control and PR-A/PR-B treatment groups and the PrestoBlue assay was used to assess differences between control and PR-A/PR-B groups in the difference concentration of cisplatin treatment using the cell line TOV-21G.Data management and analysis was performed using SPSS software for Windows version 22 (IBM, Armonk, NY, USA).A p value less than 0.05 was considered to indicate statistical signi cance.

Clinical data
A total of 592 ovarian cancer patients were identi ed during the study period.We excluded 481 patients who had a non-OCCC histology, 15 patients who did not undergo chemotherapy or had received fewer than 4 courses of chemotherapy, 9 patients due to insu cient tissue samples for IHC staining, and 7 patients with inadequate follow-up data or double cancers.Finally 80 patients met the eligibility criteria and constituted the study cohort.The mean age at diagnosis was 48.9 years (range, 28-61 years), and the median follow-up time was 50 months (range, 5-132 months).The basic characteristics of the patients are listed in Table 1.Among the 80 patients, 66 and 14 cases were de ned as having platinumsensitive and platinum-resistant disease, respectively.The platinum-sensitive patients were signi cantly associated with a higher rate of concurrent endometriosis and higher mean PR H-score (Table 2).Positive PR staining (H-score > 1) was noted in 30 patients (37.5%).Based on the ROC curve analysis, the optimal cut-off value of PR H-score to predict platinum sensitivity and survival was 50 (area under the curve 0.64, 95% [con dence interval] CI: 0.51-0.77).An H-score of 50 or higher was de ned as a strong PR expression (13 patients), and an H-score lower than 50 as a weak PR expression (67 patients).There were no signi cant differences in the distribution of clinicopathological variables between the weak and strong PR expression groups (Table 3).Typical examples of IHC staining of PR are shown in Fig. 1.Univariate analysis demonstrated signi cantly better 5-year PFS and OS for the patients with early FIGO stage (I/II), concurrent endometriosis, and pretreatment CA125 level lower than 35 U/mL, while only trends were noted in the strong PR expression group.Multivariate Cox regression analysis showed that only advanced FIGO stage was signi cantly associated with worse PFS (hazard ratio [HR] 2.427; 95% CI 1.004-5.869)and OS (HR 3.22; 95% CI 1.27-8.161)(Table 4).Because FIGO stage was the only independent factor predicting survival, we further investigated the correlation between PR expression and FIGO stage.We found a stepwise decrease in PR expression from FIGO stage I to IV, although without signi cance (Table 3).

Cell line data
The OCCC cell line TOV-21G was obtained from the American Type Culture Collection (Manassas, VA).According to a previous report, the TOV-21G cell line is highly recommended as a model for OCCC research because of similar molecular and immunologic pro les to typical primary tumors [16].Staining of PR-A and PR-B was very weak in the TOV-21G cells as shown in Fig. 2A.The TOV-21G cells were transfected with plasmids pcDNA3-PRA and pcDNA3-PRB (TOV-21G-PR-A and TOV-21G-PR-B).After transfection, we selected a stable clone with G418 200 ug/mL.Protein was extracted and subjected to Western blot, which showed an approximate 300% increase in PR-A protein expression in TOV-21G-PR-A compared to TOV-21G cells (Fig. 2A).TOV-21G-PR-B cells also demonstrated a signi cantly higher level of PR-B protein (210%) than TOV-21G cells (Fig. 2B).To determine the cytotoxic effects of cisplatin on TOV-21G cells with different PR-A and PR-B expressions, we used a PrestoBlue Cell Viability Assay.We found that compared with the control (TOV-21G) cells, the inhibition of viability by cisplatin was signi cantly greater in both TOV-21G-PR-A and TOV-21G-PR-BA cells (p < 0.01) at 48 and 72 h (Fig. 2A,  2B).

Discussion
Our results demonstrated that platinum-sensitive tumors were associated with a stronger PR expression both in clinical specimens and cell line models.However, PR status was not signi cantly associated with PFS or OS in the patients with OCCC.
Previous studies analyzing the expression rates of sex hormone receptors in patients with OCCC have reported a wide range of PR expressions from 0% [17] to 60% [18], with most demonstrating < 10% positivity [19][20][21][22].The percentage of PR-positive patients in the present study was relatively high at 37.5%.The differences among various studies may be due to differences in the PR antibodies used in IHC staining.The most common commercially available PR antibodies include clone 16, clone 636, clone 1A6, clone alpha PR6, and clone 1E2.Although all of these antibodies are thought to be able to recognize both PR-A and PR-B isoforms, a previous study demonstrated that some antibodies failed to detect PR-B in tissue sections when using IHC techniques despite their ability to do so with immunoblot analysis [23].
A possible reason is that some epitopes on PR-B may be masked during heat treatment of formalin-xed tissue, which results in a certain degree of protein tertiary structural change.Therefore, it is possible that PR expression will not be detected or at least will be underestimated in tissues where PR-B is predominant.
To date, no previous study has evaluated PR expression only in OCCC patients.The largest collaborative study from Sieh et al. reported that PR expression was associated with better high-grade serous and endometrioid carcinoma survival but not mucinous or OCCC [12].Another meta-analysis of 5685 EOC patients from 28 studies found that PR expression was related to PFS and OS, but that when focusing on studies of serous carcinoma the prognostic role of PR disappeared, which is in contrast to Sieh et al.'s study [13].This discrepancy may be related to differences in the methodologies of measuring PR and interpreting IHC.For example, we quanti ed PR status using H-score, which is quite different to Sieh et al.
Most previous studies have used a three-tier system: negative, weak, and strong, de ned as positive staining in < 1%, 1-50%, and ≥ 50% of tumor cell nuclei, respectively.In the present study, we found a stepwise reduction in PR expression from FIGO stage I to IV.Previous studies have reported different expressions of PR, ranging from 100% in atypical ovarian endometriosis to 35% in OCCC, suggesting that a loss of PR may be involved in the etiology and progression of OCCC [14,15].As the critical point of PR level change in relation to chemosensitivity or survival is unknown, we quanti ed PR status and identi ed an optimal cut-off value.However, even when using this cut-off value, PR was still not a signi cant prognosticator.There are some possible explanations for this.First, we enrolled many FIGO stage I/II patients.Early-stage OCCC has an excellent prognosis, and even with recurrence the time to relapse may be more than 6 months after completing chemotherapy.Whether these tumors were actually platinumsensitive remains unclear.Second, although we quanti ed PR status, the H-score values were not normally distributed.We attempted to stratify the patients by early stage versus advanced stage or with concurrent endometriosis versus without, and we even transformed the values to a certain exponent, however we still failed to overcome this problem.Third, the limited number of study patients after excluding others due to various reasons may also have biased the outcomes.
PR has two isoforms, A and B, and the role of each receptor isoform remains unclear with regards to hormone treatment responsiveness and prognosis.In breast cancer, PR-A-predominant breast tumors are more sensitive to anti-progestin treatment, while PR-B-predominant tumors are associated with advanced disease and poor prognosis [24].In endometrial cancer, both PR-A and PR-B are associated with less aggressive tumors, however only PR-B predicts a better prognosis [25].An interesting nding in our study was that the patients with either the A or B PR isoform appeared to be more sensitive to cisplatin, and that the patients with a high total PR expression had a trend towards better survival.The association between hormone receptor status and platinum sensitivity has seldom been investigated in EOC.In one preclinical study using OVCAR-3 cells, Preluso et al. reported that a high PR expression in ovarian cancer cells was associated with decreased PR membrane component-1 (PRMC-1) expression, which in turn increased the effectiveness of cisplatin [26].The same group later transplanted SKOV-3 cells into nude mice, resulting in the development of an OCCC animal model.They demonstrated that PGRMC1 regulated not only the growth and platinum sensitivity of cultured SKOV-3 cells, but also ovarian tumors derived from these cells [27].In an endometrial tumor model, they also found that PGRMC1-depleted tumors were more responsive to chemotherapeutic stress compared with PGRMC1-intact control tumors [28].Taken together, these data support the close relationship between PGRMC-1 expression and chemosensitivity.
Given the reported inverse relationship between PGRMC-1 and PR expression [26], our results are consistent with Preluso et al., although further studies investigating the relationship between PGRMC-1 and PR with different expressions of the isoforms are needed.
OCCC is thought to be a platinum-resistant malignant disease.Recent studies have identi ed speci c immune-related molecular pro les such as PD-L1 expression and mismatch repair protein defects in OCCC [29,30], which in turn has led to the development of a novel therapeutic approach focusing on immunotherapy.Although recent phase I/II results of using pembrolizumab were promising [31,32], the sample size was small and further phase II/III veri cation is needed.Until then, clinicians still need to identify which patients may really bene t from traditional chemotherapy.

Conclusions
In conclusion, we demonstrated that tumors with a weak PR expression were associated with greater platinum resistance and probably worse survival both in OCCC patients and cell lines.Ascertaining PR status may allow clinicians to more precisely treat OCCC patients with chemotherapy.Further animal experiments and clinical studies are needed to explore the mechanistic roles of PR in predicting chemosensitivity and disease prognosis in OCCC.

Declarations
Ethics approval and consent to participate Institutional Review Board of Chang Gung Memorial Hospital provided ethical approval for the current study (201601487B0C503).Informed consent for use of retrospective patient case data was waived.All the methods were carried out in accordance with relevant guidelines and regulations.

Figure 1 Representative
Figure 1

Figure 2 Western
Figure 2

Table 1
Clinicopathological characteristics of all patients (N = 80) CA Carbohydrate antigen; FIGO International Federation of Gynecology and Obstetrics; PR Progesterone receptor; PR weak PR H-score < 50; PR strong PR H-score ≥ 50; CA Carbohydrate antigen; CI Con dence interval; FIGO International Federation of Gynecology and Obstetrics; HR Hazard ratio; OS Overall survival; PFS Progression-free survival; PR Progesterone receptor