Young patients with stage IA adenocarcinoma of the endometrium, who wish to preserve their reproductive functions, can select the conservative therapy by medroxyprogesterone acetate (MPA) and whole wall curettages [1] - [7] . Most of these patients are able to become free of disease and have child-bearing capacity by the above therapy. Some of them can conceive after infertility treatment. However, relapse of the disease or complex atypical hyperplasia (CAH) occurs during a long-term follow-up observation, and sometimes need hysterectomy [6] . Sensitivity to progestin is thought to be related with the presence of estrogen receptor (ER)-α and progesterone receptor (PgR). Another direct anti-tumor effects of progestin without mediation of ER-α and PgR are also known [8] . However, molecular mechanism of progestin’s anti-tumor effect on the endometrial carcinoma is not clearly understood.

Endometrial cancers have high rates of phosphatidylinotitol 3-kinase (PI3K)/Akt pathway alterations reported [9] [10] . The most common PI3K pathway alteration in endometrial cancer is phosphatase and tensin homolog (PTEN) loss, which occurs via mutation, methylation, or chromosomal loss, and is reported in over 50% of cases [11] [12] [13] [14] . PTEN protein can also lost through effects of microRNAs band and protein degradation [15] .

Studies evaluating the correlation between PI3K aberrations and clinical outcome have reported discrepant results. Loss of PTEN protein expression has been reported to show a better favorable [16] [17] [18] or less favorable prognosis [19] [20] [21] [22] .

Recently, there is a report that it should be divisible into obese and non-obese to consider clinical influence of PTEN loss/retain. PTEN loss in obese patients was associated with an improved outcome, while PTEN loss in non-obese those was associated with a worse one in endometrioid endometrial cancer [23] .

A well-known tumor suppressor gene, p53 is reported to regulate PTEN transcription [24] and PTEN inhibits degradation of p53 [25] . However, the effects of progestins on the expression of PTEN and p53 in endometrial cancer have not been clarified.

In the present study, we have tried to clarify the effects of MPA on the expression of PTEN, p53, ER-α and PgR in young patients with endometrial cancers, in conjunction with the pregnancy, delivery and prolapse by dividing into the non-obese and obese patents in a single institute.

From January 1988 to December 2009, twenty endometrial cancer patients FIGO IA (1988) (FIGO staging was essentially defined post-surgically), instead of the pathological specimen before surgery were recruited to the study at the Department of Obstetrics & Gynecology, Gifu University Hospital. Inclusion criteria were: (1) age younger than 40 years (at the diagnosis), (2) nulliparous, (3) confirmed endometrial adenocarcinoma with grade 1/2 differentiation, (4) presence of PgR, (5) normal serum 125 level (<35 U/ml), (6) absence of myometrial invasion or extra-uterine spread by vaginal ultrasound, and magnetic resonance imaging (MRI) and/or computed tomography (CT) and (7) having a strong desire to preserve fertility potential. Exclusion criteria were inadequate hepatic and renal function, and abnormality in blood coagulation tests and/or history of thrombosis. Using the WHO criteria and Kurman and Norris’ criteria [26] , gynecologic cytopathologist (KN) and pathologist (TM) independently reviewed the original histologic slides. PgR expressions of the tumor cells in the pathological specimens were examined immunohistochemically before starting therapy.

After approval from the Gifu University Hospital Ethical Committee and all patients provided informed consents about the risk for progression or relapse of the tumor and the research use of their samples, the conservative therapy started. Medroxyprogesterone acetate therapy (400 - 600 mg/day, po) was commenced with whole wall endometrial curettage was performed once every four weeks. Treatment continued for at least four months, and for at least two months after disappearance of pathological abnormalities. Response was assessed by the pathological specimens obtained at curettage. After the documentation of pathologically complete remission, the patients were closely followed-up, with pelvic examination, and serial serum tumor markers (CA125). The median follow-up period was 140.0 months, ranging from 72 to 254 months, after the beginning of the conservative therapy at our hospital.

A four μm section was cut from the paraffin blocks of the whole wall curettage specimens. Each section was mounted on a silane-coated glass slide, deparaffinized, and soaked for 15 min at room temperature in 0.3% H₂O₂/methanol to block endogenous peroxidase. Sections were incubated with antibodies against PTEN (mouse monoclonal, PTEN A2B1 with dilution 1:50; Santa Cruz Biotechnology, Santa Cruz, CA), p53 (mouse monoclonal, DO-7 with dilution 1:200; Novocastra, Newcastle, UK), ER (mouse monoclonal, 1D5 with 1:200; DAKO, Glosrup, Denmark) and PgR (mouse monoclonal, pgR 636 with dilution 1:200; DAKO) overnight at 4˚C. The primary antibody was visualized using the Hstofine Simple Stain PO(M) kit (Nichirei, Tokyo, Japan) according to the instruction manual. The slides were counterstained with hematoxylin. Normal endometrial epithelium provided an internal positive control, and negative controls without addition of primary antibody showed low background staining in all cases. The PTEN was scored positive if it showed cytoplasmic staining in the entire tumor or majority of the tumor cells, and negative if it showed no cytoplasmic staining in the tumor cells [27] . Tumors that were PTEN negative as classified as PTEN loss. PTEN heterogeneous and PTEN positive tumors were classified as PTEN retained. The intensity of staining for p53, ER-α and PgR in tumor cells was scored (IHC score: 0 through 2) by two investigators (KN and TM) independently as follows: 0, negative; 1+, weakly positive; and strongly positive 2+, when compared to the corresponding normal endometrial glandular epithelium.

In order to elucidate factors associated with responses to MPA or infertility treatment, correlation between the immunohistochemical findings and undergoing relapse or successful pregnancy was determined by dividing into non-obese (BMI < 30) and obese (BMI ≥ 30) patients, using chi-square test or Fisher’s exact test. Those findings were also compared prior to and during MPA treatment using Wilcoxon singned-rank test. P values less than 0.05 were considered statistically significant.

Twenty patients with endometrial cancer received conservative MPA treatment. Patients’ characteristics are summarized in Table 1. The patients’ ages ranged 23 to 37 years old (average, 30.7 ± 3.4). Pathological diagnoses were 18 with endometrioid adenocarcinoma and two with adenosquamous cell carcinoma, their grades being 19 with G1 and one with G2. The patient’s BMI ranged 17.1 to 42.2 (26.5 ± 7.5). Treated period ranged 4 to 11 (6.3 ± 1.5). In Figure 1, changes of the endometrial lesions on MRI (T2-weighted) in Case 19are shown. Pathological changes of the endometrial lesions in Case 1 were also shown in Figure 2.

Clinicopathological responses, relapse, marital status, pregnancy, delivery and hysterectomy are summarized in Table 1. The patients were initially treated with MPA at 400 - 600 mg/day at 4 - 11 months. All cases were completely pathological remission after conservative therapy with MPA and whole wall curettages. Relapses, including CAH were observed 12 among twenty patients. Six cases (Cases 2, 4, 5, 9, 14, 16) were strongly desired to preserve fertility-function and re-selected the MPA therapy after relapses; four (Cases 1, 3, 6 and 8) underwent hysterectomy after relapses, meanwhile three cases (Case 4, 14 and 16) were treated conservative therapy after relapse and finally received hysterectomy. Eighteen patients attempted to pregnant, ten became pregnant and delivered (56%). Only one (Case 4) refused all therapies after the second relapse, and finally died due to lung metastases. Other nineteen women showed no evidence of disease including after hysterectomy cases.

Immunohistochemical expressions of PTEN, p53, ER-α and PgR were performed. ER-α (Figure 3(a)), PgR (Figure 3(b)) and p53 (Figure 3(d)) expressions were observed in the nuclei, meanwhile PTEN expression was detected mainly in cytoplasm of endometrial glands (Figure 3(c)). Table 2 summarizes the immunohistochemical evaluation of pre- and under-MPA endometrial lesions in twenty cases. The expressions of the PTEN, p53, ER-α and PgR in the pre- and under MPA therapy pair samples were compared. MPA treatment significantly decreased the expressions of p53 (P < 0.01), ER-α (P < 0.05) and PgR (P < 0.01), respectively, by Wilcoxon’s test. PTEN positive reaction was shown in the pre-MPA endometrial lesions in twelve cases. PTEN loss was observed in the eight cases in the pre-MPA endometrial lesions. The presence of strong expressions of p53 (IHC score 2) in pre-MPA treatment endometrial samples were observed in two cases, and positive were shown in six cases. The correlations between outcomes (pregnancy, delivery and relapse) and PTEN loss/retain and p53 expressions before MPA-treatment are shown in Table 3. In the non-obese patients, the proportion of pregnancy and delivery in PTEN retain cases (8/10, 80%) was significantly higher than that with PTEN loss cases (1/5, 20%, P< 0.05). Other relapse rates and p53 expressions could not be found in a tendency.

The expressions of the PTEN (Figure 4(a)), p53 (Figure 4(b)), ER-α (Figure 4(c))