Purpose:

This surgical window of opportunity (window) study assessed the short-term effect of medroxyprogesterone acetate (MPA) alone versus MPA plus the histone deacetylase (HDAC) inhibitor entinostat on regulation of progesterone receptor (PR) in women with newly diagnosed endometrioid endometrial adenocarcinoma.

Patients and Methods:

This multisite, randomized, open-label surgical window study treated women intramuscularly on day 1 with 400 mg MPA. Entinostat given 5 mg by mouth on days 1, 8, and 15 was randomly assigned with equal probability. Surgery followed on days 21–24. Pretreatment and posttreatment tissue was assessed for PR H-scores, Ki-67 levels, and histologic response.

Results:

Fifty patients were accrued in 4 months; 22 and 20 participants had PR evaluable pretreatment and posttreatment slides in the MPA and MPA/entinostat arms, respectively. Median posttreatment PR H-scores were significantly lower than pretreatment H-scores in both arms but did not differ significantly (MPA: 247 vs. 27, MPA/entinostat 260 vs. 23, respectively, P = 0.87). Decreased Ki-67 was shown in 90% treated with MPA/entinostat compared with 68% treated with MPA alone (P = 0.13). Median PR H-score decreases were larger when Ki-67 was decreased (208) versus not decreased (45). The decrease in PR pretreatment versus posttreatment was associated with loss of Ki-67 nuclear staining, consistent with reduced cellular proliferation (P < 0.008).

Conclusions:

This surgical window trial rapidly accrued in a multisite setting and evaluated PR as its primary endpoint and Ki-67 as secondary endpoint. Despite no immediate effect of entinostat on PR in this short-term study, lessons learned can inform future window and treatment trials.

Translational Relevance

Hormonal therapy in endometrioid endometrial cancer remains an important and relatively nontoxic therapeutic modality, but novel hormonal approaches are critically needed because of loss of response to hormone over time. This two-arm surgical window trial evaluated the cellular response to adding the histone deacetylase inhibitor entinostat to standard medroxyprogesterone acetate (MPA). Progesterone receptor (PR) levels decreased significantly after treatment in both arms, suggesting rapid receptor turnover associated with enhanced PR transcriptional activity. Biopsies of pretreatment and posttreatment tumor cells from patients on both arms, entinostat+MPA and MPA alone, demonstrated a similar, significant reduction in cell proliferation as determined by loss of Ki-67, which correlated significantly with the downregulation of PR. This study sets the stage for future window trials to test novel drug combinations, and reports the validity of IHC for PR and Ki-67 as biomarkers.

Endometrial cancer is the most common gynecologic cancer in the United States, with 65,620 cases and 12,590 deaths estimated in 2020 (1). Unlike other solid tumors, the incidence of and death rates from endometrial cancer continue to rise. A better understanding of the biology of this disease and development of novel therapeutics represents a significant unmet need (2, 3).

Hormonal therapy has long been a mainstay of endometrioid endometrial cancer treatment due to its ease of administration and tolerability, both for advanced/recurrent disease and for the treatment of uterine confined disease for the purpose of fertility preservation. However, hormonal therapy with progestational agents loses effectiveness over time. The decrease in treatment activity may be related to the long-term depletion of the progesterone receptor (PR) over months and years (4). Therefore, a better understanding of the biologic effect of progesterone on endometrial cancer cells and the progesterone and estrogen receptors is needed, as well as development of strategies to prolong hormonal treatment efficacy.

Surgical window trials afford an ideal in vivo model to understand drug effect after short-term exposure and to identify potential response biomarkers (5). Endometrial cancer lends itself well to this paradigm, because most women with endometrial cancer are diagnosed with a pretreatment biopsy and treated with primary surgery. The “window” of time between the biopsy and the surgery allows for the introduction of a drug to be studied. In addition, and unique to endometrial cancer, the surgical specimen allows access to a large amount of tumor tissue and adjacent normal endometrium and the endometrial stroma (the microenvironment) for study of drug effect. The surgical window trial is not intended to demonstrate impact of therapy on clinical parameters or long-term outcome. Nevertheless, the surgical window model allows obtaining in depth understanding of the biology of the drug–tumor interaction for planning future therapeutic clinical trials and for identification of potential response biomarkers.

Surgical window trials, also called window studies, have been performed in many tumor types, including hormonally responsive tumors such as breast and endometrial cancer. In an early hormonal window breast cancer trial, Ki-67 expression levels were statistically decreased in breast tumors following preoperative tamoxifen (6). Since then, several window breast cancer studies have evaluated endocrine therapies, as well as chemotherapeutics, targeted agents, and other agents (7). Gynecology Oncology Group (GOG) trial GOG-0211 was the first nontherapeutic window trial in women with endometrial cancer; this study demonstrated a statistically significant decrease in PR and PRB following 21 days of progestin treatment (8). In addition, Ki-67 and Bcl-2 were significantly downregulated; more so in histologic responders than in nonresponders (8).

The expression of PR is required to ensure progesterone responsiveness in the endometrium (9, 10), and the long-term silencing of PR through epigenetic modification may contribute to endocrine resistance (4, 11–13). As illustrated in PTEN knockout mouse models, tumors with wild type PR expression in the stroma completely resolved in response to progesterone therapy whereas those without PR expression persisted (13). The deletion of stromal PR in an endometrial cancer animal model effectively converted hormone-sensitive endometrial cancer to hormone-refractory tumors (13). The critical role of PR in maintaining an endocrine-responsive phenotype is further supported by studies that demonstrate higher responses to endocrine therapy among women with PR-positive endometrial cancer compared with those with PR-negative disease (14–18). It is therefore possible that increasing the expression of PR with an epigenetic modifier may optimize responses to medroxyprogesterone acetate (MPA) among women with hormonally driven malignancies.

The ability of histone deacetylase (HDAC) inhibitors, including entionstat, to restore functional PR expression has been confirmed in endometrial cancer cell lines. In vitro studies have shown that either DNA methyltransferase (DNMT) or HDAC inihbitors can increase the expression of PR (19–23). Yang and colleagues have demonstrated that HDAC inhibition using LBH589 was more effective than DNMT inhibition using 5-aza-deoxycytidine in restoring PR expression in both Ishikawa and ECC1 cell lines (23). These finding suggest that HDAC inhibition may be the optimal epigenetic therapy to study.

The purpose of this study was to test the hypothesis that MPA-induced downregulation of PR observed in endometrial cancers in GOG-0211 is caused by HDAC-sensitive epigenetic silencing. This was done in a surgical window trial of women with newly diagnosed endometrioid endometrial adenocarcinoma. The primary endpoint was to determine whether the addition of entinostat to MPA prevented downregulation of PR compared with MPA alone. This endpoint was based upon the assumption that higher total PR posttreatment levels would lead to a better response to hormonal therapy, to be tested subsequently. Secondary objectives included assessing the activity of MPA plus or minus entinostat using Ki-67 as a marker for proliferation, and to explore therapeutic response as measured by cellular morphology and proliferation with the addition of entinostat at the time of hysterectomy.

Eligibility

All patients provided written informed consent prior to enrollment, and approval was provided by the Institutional Review Boards at all participating institutions. The study was conducted in accordance with good clinical practice guidelines, the Declaration of Helsinki, and the U.S. Common Rule. To enter the study, patients ≥18 years of age with primary endometrioid adenocarcinoma of the endometrium on diagnostic endometrial biopsy had to be surgical candidates and have a planned surgery to include removal of the uterine corpus via any surgical modality. Histologic diagnosis by endometrial curettage or biopsy was required within 8 weeks of registration. Endometrioid histology was established at the treating facility. Patients were required to have Eastern Cooperative Oncology Group (ECOG) performance status ≤3. Formalin-fixed paraffin-embedded (FFPE) tumor tissue from the preoperative tissue sampling as well as the hysterectomy specimen was required to be submitted. Patients were required to have normal hematologic (platelets≥100,000/μL, granulocytes≥1,500 μL), renal (creatinine ≤1.6 mg/dL), and liver function. Progesterone receptor expression was not required preenrollment.

Key exclusion criteria included nonendometrioid histology. Prior treatment with progestin or antiestrogen therapy within 3 months of diagnosis was not allowed (estrogen therapy alone was allowed). Patients with a history of thrombophlebitis within the past 2 years or ongoing thromboembolic disorders were not eligible.

Additional entry criteria are available in the study protocol. The trial was registered with clinicaltrials.gov (NCT03018249).

Study design and treatment

The study was a multisite, randomized, open-label surgical window trial of short-term MPA with or without entinostat. Patients were registered by the participating site through OPEN and randomization was carried out centrally by the NRG Statistics and Data Management Center. Treatment group assignment was determined using a procedure that tends to randomly allocate at a 1:1 ratio.

All patients were treated on day 1 with 400 mg MPA given intramuscularly. Patients randomized to the entinostat arm received entinostat 5 mg by mouth on days 1, 8, and 15. Surgery was required to be performed on days 21–24. Assessment for adverse events (AE) was required within 3 days of each entinostat administration, and could be done by telephone. A complete blood count was required ≤ 7 days preoperatively. A preoperative platelet value of less than 50,000/μL would delay surgery until resolution to greater than 50,000/μL.

Patients who received a treatment assignment that included entinostat were expected to complete all three doses of drug preoperatively. Patients who completed only two doses were allowed to stay on study and considered evaluable. Patients who completed ≤1 dose were followed per protocol. No dose escalations or reductions were allowed.

Biospecimen processing and analysis

This trial incorporated two biomarkers, PR (integral) and Ki-67 (integrated) as approved by the NCI Biomarker Review Committee.

PR and Ki-67 histology and IHC

All tissue samples were FFPE and freshly cut at 4 μm sections. All staining was performed within 48 hours of receipt at the University of Iowa Hospitals and Clinics (Iowa City, IA) in the Clinical Laboratory Improvement Amendments–certified Department of Pathology Histology and Immunohistochemistry Laboratory. Routine hematoxylin and eosin (H&E) staining was performed on each sample set, as well as PR and Ki-67 IHC using a polymer-based detection system on the Dako AutostainerLink 48 System [FDA 510(k)-cleared pharmDx assay Kit (Dako]. Antibodies used were as follows: PR, mouse mAb PgR636, Dako; Ki-67, clone MIB-1, Dako. Antibody dilutions were performed according to the manufacturer's recommendations.

PR analysis

IHC staining for PR was evaluated in a semiquantitative manner following the initial work of Carcangiu and colleagues and Chambers and colleagues, who demonstrated excellent correlation in the assessment of PR level using IHC compared with biochemical methods such as ligand binding (24, 25). Pretreatment and posttreatment slides stained for progesterone receptors were reviewed by three independent NRG pathologists without knowledge of treatment arm. The American Society of Clinical Oncology/College of American Pathologists guidelines for scoring PR were followed (26). Per the guidelines, at least 1% of the cells in the specimen must be immunoreactive to consider the tissue positive. The intensity of tumor staining (0 to 3+) was scored using the positive control tissue as 3+, and all samples were also assessed for presence of internal positive control staining (stroma, myometrium) to ensure tissue integrity. The reviewers also scored each slide with a percent of cells staining positive (0–100), which was multiplied by the intensity score (0, 1+, 2+, or 3+) to get a modified H-score. An initial planned analysis of concordance between reviewers in a subset of samples led to a decision to continue utilizing three reviewers for all samples. The values from each reviewer were averaged for each slide to define the primary endpoint of this study.

Ki-67 analysis

Because of marked intratumoral heterogeneity in Ki-67 IHC staining patterns in endometrial carcinomas, a qualitative method of analysis was utilized by three independent, experienced pathologists without knowledge of treatment arm. Matched paired pretreatment and posttreatment samples were sequentially reviewed, and Ki-67 IHC was scored as increased, decreased, or unchanged from the pretreatment to posttreatment tissue within each individual patient set.

Histologic response

The presence of a histologic response was subjectively determined by an experienced gynecologic pathologist as previously described in GOG-0211 (8), using matched H&E sections from pretreatment and posttreatment samples from each patient set.

Invalid test results

Per protocol, central pathology review was required at completion of accrual. Invalid test results included no tumor, wrong diagnosis, and lack of staining or nonspecific staining using the positive control tissue. If a specimen contained no tumor, the patient was excluded from analysis. If immunostaining results were nonspecific or negative using the positive control run with each cycle, the IHC was repeated.

Statistical design

A sample size of 40 patients (20 in each treatment group) with associated evaluable pretreatment and posttreatment specimens was targeted to test for a treatment difference in the distribution of posttreatment progesterone receptor scores using a Mann–Whitney test adhering to an intention-to-treat analysis plan. Type I error was set at 0.05 for a two-sided hypothesis test. With 40 evaluable participants 90% power was achieved to detect a treatment difference of −1.0 (−100 on H-score scale) assuming a SD of 0.8 in the reference group and 1.0 in the experimental treatment group. A total sample size of 50 accounted for a presumed 20% inevaluability rate for the primary endpoint. Acknowledging the potential bias, a sensitivity analysis removing patients on the experimental arm that did not receive at least two doses of entinostat was planned. However, patients who received less than two doses of entinostat withdrew consent prior to initiating treatment; no specimens were collected and inclusion in statistical endpoint analyses was not possible.

Secondary endpoints included histologic response when comparing pretreatment and posttreatment specimens (complete or partial response vs. other) and relative change in Ki-67 protein expression (decreased vs. no change or increase). The proportions with histologic complete or partial response and Ki-67 protein expression decrease were compared between the arms, and confidence intervals were constructed. In addition, the association between histologic response and Ki-67 protein expression response was explored. Evaluation of Ki-67 protein expression strictly by proportion of or number of cells staining positive proved to be a problematic endpoint due to marked intratumoral heterogeneity in its expression and was not carried out. A Kruskal–Wallis test was used to compare distributions of PR change by Ki-67 response group in an ad hoc analysis.

The frequency and maximum severity of acute drug emergent and postsurgical AEs was tabulated by treatment arm as graded by CTCAE v4.0.

Statistical analyses were generated using SAS/STAT software, Version 9.4 of the SAS System for Windows. Copyright 2016 SAS Institute Inc.

The first patient was enrolled in October 2017, with completion of planned accrual within 4 months. Fifty participants enrolled; 25 on each arm. Four patients withdrew consent prior to beginning protocol therapy; the reason given for consent withdrawal in all cases was patient refusal. All other patients initiated treatment. The CONSORT diagram is shown in Fig. 1. Ultimately, 22 and 20 patients in the MPA and MPA/entinostat arms, respectively, had paired pre-/posttreatment specimens.

Figure 1.

CONSORT diagram.

Figure 1.

CONSORT diagram.

Close modal

The patient and disease characteristics for all patients are summarized in Tables 1 and 2. The overall (and per arm) median age was 61 years (range, 37–92 years). Most patients were ages 60 to 69 years, white race, non-Hispanic or Latino ethnicity, and with ECOG performance status of 0. The histologic grade at biopsy was predominately grade 1. A focal clear cell change was reported in one tumor in the hysterectomy specimen.

Table 1.

Patient and pretreatment tumor characteristics for all enrolled patients.

Regimen
MPAMPA/EntinostatTotal
CharacteristicN (%)N (%)N (%)
Age (years) 
 30–49 2 (8.0) 2 (8.0) 4 (8.0) 
 50–59 7 (28.0) 7 (28.0) 14 (28.0) 
 60–69 13 (52.0) 13 (52.0) 26 (52.0) 
 70–99 3 (12.0) 3 (12.0) 6 (12.0) 
Ethnicity 
 Not Hispanic or Latino 25 (100.0) 24 (96.0) 49 (98.0) 
 Not reported 0 (0) 1 (4.0) 1 (2.0) 
Race 
 American Indian or Alaska Native 1 (4.0) 0 (0) 1 (2.0) 
 Black or African American 1 (4.0) 2 (8.0) 3 (6.0) 
 Not reported 0 (0) 1 (4.0) 1 (2.0) 
 White 23 (92.0) 22 (88.0) 45 (90.0) 
Performance status 
 0 23 (92.0) 21 (84.0) 44 (88.0) 
 1 1 (4.0) 3 (12.0) 4 (8.0) 
 2–3 1 (4.0) 1 (4.0) 1 (4.0) 
Endometrioid adenocarcinoma 
 Not reported 2 (8.0) 2 (8.0) 4 (8.0) 
 Yes 23 (92.0) 23 (92.0) 46 (92.0) 
Histologic grade 
 Not reported 2 (8.0) 2 (8.0) 4 (8.0) 
 1 13 (52.0) 15 (60.0) 28 (56.0) 
 2 5 (20.0) 6 (24.0) 11 (22.0) 
 3 4 (16.0) 2 (8.0) 6 (12.0) 
 Not graded 1 (4.0) 0 (0) 1 (2.0) 
Total 25 (50.0) 25 (50.0) 50 (100.0) 
Regimen
MPAMPA/EntinostatTotal
CharacteristicN (%)N (%)N (%)
Age (years) 
 30–49 2 (8.0) 2 (8.0) 4 (8.0) 
 50–59 7 (28.0) 7 (28.0) 14 (28.0) 
 60–69 13 (52.0) 13 (52.0) 26 (52.0) 
 70–99 3 (12.0) 3 (12.0) 6 (12.0) 
Ethnicity 
 Not Hispanic or Latino 25 (100.0) 24 (96.0) 49 (98.0) 
 Not reported 0 (0) 1 (4.0) 1 (2.0) 
Race 
 American Indian or Alaska Native 1 (4.0) 0 (0) 1 (2.0) 
 Black or African American 1 (4.0) 2 (8.0) 3 (6.0) 
 Not reported 0 (0) 1 (4.0) 1 (2.0) 
 White 23 (92.0) 22 (88.0) 45 (90.0) 
Performance status 
 0 23 (92.0) 21 (84.0) 44 (88.0) 
 1 1 (4.0) 3 (12.0) 4 (8.0) 
 2–3 1 (4.0) 1 (4.0) 1 (4.0) 
Endometrioid adenocarcinoma 
 Not reported 2 (8.0) 2 (8.0) 4 (8.0) 
 Yes 23 (92.0) 23 (92.0) 46 (92.0) 
Histologic grade 
 Not reported 2 (8.0) 2 (8.0) 4 (8.0) 
 1 13 (52.0) 15 (60.0) 28 (56.0) 
 2 5 (20.0) 6 (24.0) 11 (22.0) 
 3 4 (16.0) 2 (8.0) 6 (12.0) 
 Not graded 1 (4.0) 0 (0) 1 (2.0) 
Total 25 (50.0) 25 (50.0) 50 (100.0) 

Abbreviations: MPA, medroxyprogesterone acetate; N, number.

Table 2.

Postoperative disease characteristics for all enrolled patients.

Regimen
MPAMPA/EntinostatTotal
CharacteristicN (%)N (%)N (%)
FIGO stage 
 Not reported 2 (8.0) 2 (8.0) 4 (8.0) 
 I 1 (4.0) 1 (4.0) 2 (4.0) 
 IA 14 (56.0) 15 (60.0) 29 (58.0) 
 IB 1 (4.0) 3 (12.0) 4 (8.0) 
 II 3 (12.0) 3 (12.0) 6 (12.0) 
 III 2 (8.0) 0 (0) 2 (4.0) 
 IIIA 0 (0) 1 (4.0) 1 (2.0) 
 IIIC1 1 (4.0) 0 (0) 1 (2.0) 
 IIIC2 1 (4.0) 0 (0) 1 (2.0) 
Myometrial invasion 
 Not reported 3 (12.0) 3 (12.0) 6 (12.0) 
 None 7 (28.0) 6 (24.0) 13 (26.0) 
 <50% 9 (36.0) 11 (44.0) 20 (40.0) 
 ≥50% 6 (24.0) 3 (12.0) 9 (18.0) 
 Unknown 0 (0) 2 (8.0) 2 (4.0) 
Clear cell features 
 Yes 1 (4.0) 0 (0) 1 (2.0) 
Total 25 (50.0) 25 (50.0) 50 (100.0) 
Regimen
MPAMPA/EntinostatTotal
CharacteristicN (%)N (%)N (%)
FIGO stage 
 Not reported 2 (8.0) 2 (8.0) 4 (8.0) 
 I 1 (4.0) 1 (4.0) 2 (4.0) 
 IA 14 (56.0) 15 (60.0) 29 (58.0) 
 IB 1 (4.0) 3 (12.0) 4 (8.0) 
 II 3 (12.0) 3 (12.0) 6 (12.0) 
 III 2 (8.0) 0 (0) 2 (4.0) 
 IIIA 0 (0) 1 (4.0) 1 (2.0) 
 IIIC1 1 (4.0) 0 (0) 1 (2.0) 
 IIIC2 1 (4.0) 0 (0) 1 (2.0) 
Myometrial invasion 
 Not reported 3 (12.0) 3 (12.0) 6 (12.0) 
 None 7 (28.0) 6 (24.0) 13 (26.0) 
 <50% 9 (36.0) 11 (44.0) 20 (40.0) 
 ≥50% 6 (24.0) 3 (12.0) 9 (18.0) 
 Unknown 0 (0) 2 (8.0) 2 (4.0) 
Clear cell features 
 Yes 1 (4.0) 0 (0) 1 (2.0) 
Total 25 (50.0) 25 (50.0) 50 (100.0) 

Abbreviations: FIGO, International Federation of Gynecology and Obstetrics; MPA, medroxyprogesterone acetate; N, number.

Primary endpoint

Twenty-two and 20 participants had paired tissues for measurement of progesterone receptor in the MPA and MPA/Entinostat arms, respectively. Pretreatment, posttreatment, and the difference between posttreatment and pretreatment PR H-scores are shown in Table 3. The median posttreatment PR H-scores were substantially lower than pretreatment H-scores in both arms (247 vs. 27, 260 vs. 23, respectively) and did not differ between the arms (P = 0.87).

Table 3.

PR H-score descriptive statistics by arm.

RegimenEndpointNMeanMedianMaximumMinimumRangeSD
MPA Pretreatment PR 22 225.1 247.5 300.0 0.7 299.3 86.7 
 Posttreatment PR 22 53.6 26.7 210.0 210.0 64.8 
 PR H-score change 22 −171.5 −175.5 −0.7 −291.7 291.0 87.9 
MPA/Entinostat Pretreatment PR 20 217.6 260.0 300.0 300.0 90.7 
 Posttreatment PR 21 42.7 23.3 170.0 170.0 49.0 
 PR H-score change 20 −177.4 −193.3 3.3 −291.7 295.0 83.9 
RegimenEndpointNMeanMedianMaximumMinimumRangeSD
MPA Pretreatment PR 22 225.1 247.5 300.0 0.7 299.3 86.7 
 Posttreatment PR 22 53.6 26.7 210.0 210.0 64.8 
 PR H-score change 22 −171.5 −175.5 −0.7 −291.7 291.0 87.9 
MPA/Entinostat Pretreatment PR 20 217.6 260.0 300.0 300.0 90.7 
 Posttreatment PR 21 42.7 23.3 170.0 170.0 49.0 
 PR H-score change 20 −177.4 −193.3 3.3 −291.7 295.0 83.9 

Abbreviations: MPA, medroxyprogesterone acetate; N, number; PR, progesterone receptor; SD, standard deviation.

Secondary endpoints

Histologic response

Histologic scores for pretreatment and posttreatment blinded patient slide sets were determined by one reviewer. Overall, there were 30 patients with a partial response, 16 who received MPA and 14 who received MPA/entinostat. There were no complete responses. Among those who were treated and had evaluable specimens, there was no statistically significant difference in histologic response between the arms (73% vs. 70%; P = 1.0). Response could not be measured in 7 patients and 1 could not be evaluated.

Ki-67 protein response

Ki-67 protein expression was assessed in pretreatment and posttreatment paired blinded patient slide sets by three reviewers and defined as a dichotomous response. A positive response was defined as a decrease in Ki-67 protein expression from pretreatment to posttreatment. The results are shown in Table 4. Concordance between reviewers is described as a Fleiss Kappa of 0.91, and weighted Kappa scores for paired reviewer evaluation ranged between 0.85 and 0.95.

Table 4.

Ki-67 protein expression response in treated patients with evaluated specimens.

Regimen
MPAMPA/EntinostatTotal
N (%)N (%)N
Ki-67 Protein response 
 Response 15 (68.2) 18 (90.0) 33 
 No response 7 (31.8) 2 (10.0) 
Total 22 (52.4) 20 (47.6) 42 
Regimen
MPAMPA/EntinostatTotal
N (%)N (%)N
Ki-67 Protein response 
 Response 15 (68.2) 18 (90.0) 33 
 No response 7 (31.8) 2 (10.0) 
Total 22 (52.4) 20 (47.6) 42 

Abbreviations: MPA, medroxyprogesterone acetate; N, number.

A response was seen in tumors of 15 and 18 patients in the MPA and MPA/entinostat arms, respectively. Nearly all tumors (90%) from patients on the combination arm showed some reduction in Ki-67 score; whereas only 68% of tumors in patient receiving MPA alone showed a reduction in Ki-67 protein expression (P = 0.13). The difference in the proportion of responders is 0.22 (95% confidence interval, −0.17 to 0.45). The Ki-67 protein expression response and the posttreatment PR H-score category were related to each other where reduction in both was associated with a larger median drop in PR (208) than those who did not (45); P = 0.008; Fig. 2.

Figure 2.

Box plots for change in PR expression H-score from pretreatment to posttreatment.

Figure 2.

Box plots for change in PR expression H-score from pretreatment to posttreatment.

Close modal

Two patient samples, one on each arm, were found to have serous histologic features present in pretreatment research tumor specimens, including: papillary, micropapillary, and pseudoglandular architecture with irregular luminal borders and high-grade nuclei with pleomorphism and abundant mitoses. Serous morphology was confirmed by IHC, showing missense p53 staining pattern and strong/diffuse nuclear and cytoplasmic p16 expression; these cases were then further reviewed and confirmed to be serous by the two additional pathologists participating in the Ki-67 analysis. The case on the MPA only arm had pretreatment and posttreatment PR expression H-score of 255 and 210, respectively, a partial histologic response and an increase in Ki-67 protein expression. The other case, on the combination arm, had a pretreatment and posttreatment PR expression H-score of 150 and 50, respectively, no histologic response and decreased Ki-67 protein expression.

AEs

Overall grade 3 AEs were reported in 4 patients (8.7%); there were no grade 4 or 5 AEs. There was 1 patient on each arm who was reported to have experienced at least one grade 3 vascular thrombotic AE prior to surgery and one grade 3 skin infection was reported on the combination arm. During the entire treatment period including surgery, there were 4 patients reported to have experienced at least one grade 3 or higher AE; two on the MPA arm and two on the combination arm. On the MPA arm, in addition to the grade 3 preoperative thrombotic event there was one postoperative grade 3 hypertension event reported. On the MPA/entinostat arm in addition to the one grade 3 preoperative thrombotic event and skin infection, postoperative grade 3 events included anemia (n = 1), hypokalemia (n = 1), and fatigue (n = 1).

Surgical window trials offer a unique model for understanding drug interaction with tumor that ultimately and ideally results in the implementation of novel treatment modalities (5). Endometrial cancer, a surgically managed disease that is easily accessible to preoperative office biopsy, lends itself well to this model, and is also a cancer in need of novel therapies. First with GOG-0211 and now with NRG-GY011, we have demonstrated the ability to successfully complete surgical window trials in endometrial cancer in the NCI's National Clinical Trials Network setting. Moreover, the rapid accrual of NRG-GY011 confirms an enthusiasm of both investigators and patients for this type of research. Specifically, this study model allows important questions to be asked in endometrial cancer regarding how to modify known successful treatment modalities, and also offers scientific guidance regarding how to choose the most effective new investigational agent to apply in upcoming treatment trials.

Use of PR levels as an hormonal therapy trial endpoint requires consideration of the normal levels and the life cycle of PR over the short and long term in reproductive tissues. Progestin binding to its receptor results in its activation, dimerization, and binding to DNA to control the transcription of progestin-dependent genes. Ligand-bound receptors on DNA undergo rapid turnover to maintain transcriptional activity, and receptor levels are lower as a result of rapid turnover caused by ubiquitination of DNA-bound PR with proteasomal destruction (27). The resulting lower receptor levels have been interpreted to result in the attenuated response to progestin therapy that occurs over time. Low levels of PR at any point in time may be due to (i) the short-term rapid turnover during PR transcriptional activity in response to ligand, (ii) loss of transcription of the PR gene, or (iii) long-term epigenetic silencing of the PR gene. While the first mechanism signifies progestin sensitivity and PR activity, the latter two mechanisms relate to hormone insensitivity.

The previous window trial GOG-0211 demonstrated a significant downregulation of PR after treatment with MPA. One interpretation of this result was that lower PR levels signified loss of progestin sensitivity (8). However, an alternative hypothesis is that the downregulation of PR observed in GOG-211 actually represents hormonal response related to ligand-induced PR downregulation, a sign of PR transcriptional activation. If, on the other hand, the reduction in PR levels correlates with the previously reported epigenetic silencing of this gene, an HDAC inhibitor may reverse it and induce PR. NRG-GY011 was therefore designed to test the hypothesis that the addition of entinostat, an HDAC inhibitor, would prevent the epigenetic modification and presumed silencing of the PR, thereby enhancing the effect of progestin over longer periods of time. Our hypothesis in NRG-GY011 was that entinostat would induce higher PR levels if PR was epigenetically silenced in these cases by MPA exposure. In GY011, similar to GOG-0211, a decrease in expression of PR as measured by H-score following administration of MPA was demonstrated. Specific to GY011, there was no difference in decrease in PR seen between the two arms. This lack of difference in posttreatment PR levels only addresses whether there is a short-term epigenetic action of MPA. This design would not find a difference if epigenetic regulation of PR is a more gradual and long lasting event. It is also possible that HDAC inhibition has an effect on the tumor that is unrelated to regulation of PR. A potential trend toward fewer Ki-67 nonresponsive cases in the entinostat plus MPA arm was noted that will require further confirmation in future trials. In addition, the strong correlation between the posttreatment decrease in PR accompanied by loss of Ki-67 supports the alternative hypothesis that the decrease in PR after progestin therapy is due to ligand activity and inhibition of cellular proliferation, not resistance to MPA treatment.

Additional limitations to this study design include: the (relatively) short window, the powering of the study for the PR endpoint (integral) but not the Ki-67 endpoint (integrated), the loss of enrolled subjects (histology, lack of tumor in the final specimen), and the very high rate of PR expression in the majority of subjects, which may have negated the need for epigenetic desilencing of PR through the addition of entinostat. Lessons learned from this trial will inform subsequent studies.

We report other novel findings as a result of this study in addition to the negative primary endpoint. It is well known that Ki-67 measurements can be variable between and within laboratories and pathologists (28). We developed and demonstrated a reliable and reproducible method for measuring Ki-67 in these FFPE samples. This approach can be exported for local revalidation and potentially strengthen the applicability of Ki-67 in future integrated studies. This assay allowed us to observe a difference in Ki-67 expression between treatment arms. Nearly all tumors in the combination entinostat/MPA arm showed a decrease in Ki-67, compared with only 68% of tumors in the MPA group (P = 0.13). This finding suggests a greater response to the brief exposure to the combination therapy and can be used to support advancing the combination of entionstat/MPA to a treatment trial evaluation. This study also clarifies that an early drop in PR levels is most likely a marker for hormonal treatment responsiveness.

In addition to the results discussed above, we have access to additional FFPE tissue as well as fresh tissue from this study, and we look forward to the opportunity to complete the proposed exploratory endpoints of this trial; specifically, to evaluate progestin-dependent gene expression in pretreatment and posttreatment tissues. It is expected that the enhanced PR activity noted in NRG-GY011 and reflected in the drop in Ki-67 in both arms of this study will be reflected in the induction of progestin-dependent genes related to cellular differentiation.

L.R. Duska reports grants from NRG during the conduct of the study, as well as grants and personal fees from GOG Foundation, Morphotek, and Merck; grants from Tesaro, GSK, Ludwig, Harpoon, Seattle Genetics, AbbVie, Pfizer, LEAP, ImmunoGen, Genentech, Cerulean, Aduro Biotech, Eisai, Lycera, and Advaxis; and personal fees from Inovio outside the submitted work. V.L. Filiaci reports grants from NIH during the conduct of the study and other from GOG Foundation, Inc. outside the submitted work. R.G. Moore reports personal fees from Fujirebio Diagnostics and grants from Angle plc outside the submitted work. M.L. Pearl reports grants from NRG Oncology during the conduct of the study and other from Vita Tex, Inc outside the submitted work. C.Y. Muller reports grants from NCI MU-NCORP during the conduct of the study. K.K. Leslie reports grants from NIH during the conduct of the study. No disclosures were reported by the other authors.

L.R. Duska: Conceptualization, resources, data curation, supervision, writing–original draft, writing–review and editing. V.L. Filiaci: Data curation, writing–original draft. J.L. Walker: Data curation, writing–original draft. L.L. Holman: Writing–original draft, writing–review and editing. E.K. Hill: Writing–original draft, writing–review and editing. R.G. Moore: Resources, writing–review and editing. K.L. Ring: Data curation, methodology, writing–original draft, writing–review and editing. M.L. Pearl: Resources, writing–original draft, writing–review and editing. C.Y. Muller: Resources, data curation, writing–original draft, writing–review and editing. C.L. Kushnir: Resources, data curation, writing–original draft, writing–review and editing. H.A. Lankes: Resources, data curation, writing–original draft, writing–review and editing. M.I. Samuelson: Resources, data curation, writing–original draft, writing–review and editing. K.S. Carrick: Data curation, writing–original draft, writing–review and editing. A. Rajan: Resources, data curation, writing–original draft, writing–review and editing. W.H. Rodgers: Data curation, writing–original draft, writing–review and editing. E.C. Kohn: Resources, writing–original draft, writing–review and editing. R. Piekarz: Resources, writing–original draft, writing–review and editing. K.K. Leslie: Conceptualization, resources, supervision, writing–original draft, writing–review and editing.

This study was supported by grant no. NCT03018249. This study was supported by NIH grants to NRG Oncology (1 U10 CA180822), NRG Operations (U10CA180868), NCORP (UGICA 189867), and NRG Biospecimen Bank (U24CA196067). Support for the performance of the translational endpoints of this study was provided by NIH grant CA99908-17, to K.K. Leslie.

The following Gynecologic Oncology Group member institutions participated in the primary treatment studies: University of Oklahoma Health Sciences Center LAPS, University of Iowa/Holden Comprehensive Cancer Center, University of Rochester LAPS, University of Virginia Cancer Center, Stony Brook University Medical Center, New Mexico Minority Underserved NCORP, Women's Cancer Center of Nevada, Columbus NCI Community Oncology Research Program, University of Colorado Cancer Center LAPS, University of Chicago Comprehensive Cancer Center LAPS, Duke University–Duke Cancer Institute LAPS, Dartmouth College–Norris Cotton Cancer Center LAPS, and Montana Cancer Consortium NCORP.

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.

1.
Siegel
RL
,
Miller
KD
,
Jemal
A
. 
Cancer statistics, 2020
.
CA Cancer J Clin
2020
;
70
:
7
30
.
2.
Clarke
MA
,
Devesa
SS
,
Harvey
SV
,
Wentzensen
N
. 
Hysterectomy-corrected uterine corpus cancer incidence trends and differences in relative survival reveal racial disparities and rising rates of nonendometrioid cancers
.
J Clin Oncol
2019
;
37
:
1895
908
.
3.
Lortet-Tieulent
J
,
Ferlay
J
,
Bray
F
,
Jemal
A
. 
International patterns and trends in endometrial cancer incidence, 1978–2013
.
J Natl Cancer Inst
2018
;
110
:
354
61
.
4.
Yang
S
,
Jia
Y
,
Liu
X
,
Winters
C
,
Wang
X
,
Zhang
Y
, et al
Systematic dissection of the mechanisms underlying progesterone receptor downregulation in endometrial cancer
.
Oncotarget
2014
;
5
:
9783
97
.
5.
Lheureux
S
,
McCourt
C
,
Rimel
BJ
,
Duska
L
,
Fleming
G
,
Mackay
H
, et al
Moving forward with actionable therapeutic targets and opportunities in endometrial cancer: a NCI clinical trials planning meeting report
.
Gynecol Oncol
2018 Feb 22 [Epub ahead of print]
.
6.
Bajetta
E
,
Celio
L
,
Di Leo
A
,
Bartoli
C
,
Pilotti
S
,
Leutner
M
, et al
Effects of short-term pre-operative tamoxifen on steroid receptor and Ki-67 expression in primary breast cancer: an immunocytochemical study
.
Int J Oncol
1998
;
12
:
853
8
.
7.
Marous
M
,
Bièche
I
,
Paoletti
X
,
Alt
M
,
Razak
ARA
,
Stathis
A
, et al
Designs of preoperative biomarkers trials in oncology: a systematic review of the literature
.
Ann Oncol
2015
;
26
:
2419
28
.
8.
Zaino
RJ
,
Brady
WE
,
Todd
W
,
Leslie
K
,
Fischer
EG
,
Horowitz
NS
, et al
Histologic effects of medroxyprogesterone acetate on endometrioid endometrial adenocarcinoma: a Gynecologic Oncology Group study
.
Int J Gynecol Pathol
2014
;
33
:
543
53
.
9.
Graham
JD
,
Clarke
CL
. 
Physiological action of progesterone in target tissues
.
Endocr Rev
1997
;
18
:
502
19
.
10.
Shao
R
. 
Progesterone receptor isoforms A and B: new insights into the mechanism of progesterone resistance for the treatment of endometrial carcinoma
.
Ecancermedicalscience
2013
;
7
:
381
.
11.
Yang
S
,
Thiel
KW
,
Leslie
KK
. 
Progesterone: the ultimate endometrial tumor suppressor
.
Trends Endocrinol Metab
2011
;
22
:
145
52
.
12.
Yang
S
,
Thiel
KW
,
De Geest
K
,
Leslie
KK
. 
Endometrial cancer: reviving progesterone therapy in the molecular age
.
Discov Med
2011
;
12
:
205
12
.
13.
Janzen
DM
,
Rosales
MA
,
Paik
DY
,
Lee
DS
,
Smith
DA
,
Witte
ON
, et al
Progesterone receptor signaling in the microenvironment of endometrial cancer influences its response to hormonal therapy
.
Cancer Res
2013
;
73
:
4697
710
.
14.
Thigpen
JT
,
Brady
MF
,
Alvarez
RD
,
Adelson
MD
,
Homesley
HD
,
Manetta
A
, et al
Oral medroxyprogesterone acetate in the treatment of advanced or recurrent endometrial carcinoma: a dose-response study by the Gynecologic Oncology Group
.
J Clin Oncol
1999
;
17
:
1736
44
.
15.
Whitney
CW
,
Brunetto
VL
,
Zaino
RJ
,
Lentz
SS
,
Sorosky
J
,
Armstrong
DK
, et al
Phase II study of medroxyprogesterone acetate plus tamoxifen in advanced endometrial carcinoma: a Gynecologic Oncology Group study
.
Gynecol Oncol
2004
;
92
:
4
9
.
16.
Benraad
TJ
,
Friberg
LG
,
Koenders
AJ
,
Kullander
S
. 
Do estrogen and progesterone receptors (E2R and PR) in metastasizing endometrial cancers predict the response to gestagen therapy?
Acta Obstet Gynecol Scand
1980
;
59
:
155
9
.
17.
Lindemann
K
,
Malander
S
,
Christensen
RD
,
Mirza
MR
,
Kristensen
GB
,
Aavall-Lundqvist
E
, et al
Examestane in advanced or recurrent endometrial carcinoma: a prospective phase II study by the Nordic Society of Gynecologic Oncology (NSGO)
.
BMC Cancer
2014
;
14
:
68
.
18.
McCarty
KS
 Jr
,
Barton
TK
,
Fetter
BF
,
Creasman
WT
,
McCarty
KS
 Sr
. 
Correlation of estrogen and progesterone receptors with histologic differentiation in endometrial adenocarcinoma
.
Am J Pathol
1979
;
96
:
171
83
.
19.
Sasaki
M
,
Kaneuchi
M
,
Fujimoto
S
,
Tanaka
Y
,
Dahiya
R
. 
Hypermethylation can selectively silence multiple promoters of steroid receptors in cancers
.
Mol Cell Endocrinol
2003
;
202
:
201
7
.
20.
Yap
OW
,
Bhat
G
,
Liu
L
,
Tollefsbol
TO
. 
Epigenetic modifications of the Estrogen receptor beta gene in epithelial ovarian cancer cells
.
Anticancer Res
2009
;
29
:
139
44
.
21.
Ferguson
AT
,
Lapidus
RG
,
Baylin
SB
,
Davidson
NE
. 
Demethylation of the estrogen receptor gene in estrogen receptor-negative breast cancer cells can reactivate estrogen receptor gene expression
.
Cancer Res
1995
;
55
:
2279
83
.
22.
Yan
L
,
Yang
X
,
Davidson
NE
. 
Role of DNA methylation and histone acetylation in steroid receptor expression in breast cancer
.
J Mammary Gland Biol Neoplasia
2001
;
6
:
183
92
.
23.
Yang
S
,
Xiao
X
,
Jia
Y
,
Liu
X
,
Zhang
Y
,
Wang
X
, et al
Epigenetic modification restores functional PR expression in endometrial cancer cells
.
Curr Pharm Des
2014
;
20
:
1874
80
.
24.
Carcangiu
ML
,
Chambers
JT
,
Voynick
IM
,
Pirro
M
,
Schwartz
PE
. 
Immunohistochemical evaluation of estrogen and progesterone receptor content in 183 patients with endometrial carcinoma. Part I: clinical and histologic correlations
.
Am J Clin Pathol
1990
;
94
:
247
54
.
25.
Chambers
JT
,
Carcangiu
ML
,
Voynick
IM
,
Schwartz
PE
. 
Immunohistochemical evaluation of estrogen and progesterone receptor content in 183 patients with endometrial carcinoma. Part II: correlation between biochemical and immunohistochemical methods and survival
.
Am J Clin Pathol
1990
;
94
:
255
60
.
26.
Hammond
MEH
,
Hayes
DF
,
Dowsett
M
,
Allred
DC
,
Hagerty
KL
,
Badve
S
, et al
American Society of Clinical Oncology/College of American Pathologists guideline recommendations for immunohistochemical testing of estrogen and progesterone receptors in breast cancer
.
Arch Pathol Lab Med
2010
;
134
:
907
22
.
27.
Dennis
AP
,
Haq
RU
,
Nawaz
Z
. 
Importance of the regulation of nuclear receptor degradation
.
Front Biosci
2001
;
6
:
D954
9
.
28.
Polley
MY
,
Leung
SC
,
McShane
LM
,
Gao
D
,
Hugh
JC
,
Mastropasqua
MG
, et al
An international Ki67 reproducibility study
.
J Natl Cancer Inst
2013
;
105
:
1897
906
.