Characteristics of endometrial cancer progressed to extrauterine lesions following fertility preserving medroxyprogesterone acetate therapy for young endometrial cancer patients.

Recently, the number of patients with endometrial cancer (EC) has been increasing because of lifestyle changes, even among young patients. Since EC is hormone-dependent, there are wide ranges of the risks of carcinogenesis including infertility, obesity, and so on [ 1 ]. In addition, women in Japan are marrying later in life and may wish to have a baby when diagnosed with EC. The standard treatment for EC is surgery including hysterectomy. Atypical endometrial hyperplasia (AEH), which is a precancerous lesion, is also treated by hysterectomy. Given the above facts, high-dose medroxyprogesterone acetate (MPA) is an attractive option as a fertility-preserving therapy for young patients with early-stage EC and AEH who desire to have a baby. According to previous reports, the complete remission (CR) rates for AEH and endometrioid carcinoma grade 1 (EMG1) were 55%–98.5%, indicating that MPA is highly effective [ 1 2 3 4 ]. Studies have reported overall survival values with 5-year survival rates of 95% or better. Thus, although surgery is the standard treatment for EC, MPA therapy is considered acceptable for young patients with AEH or EC that remains within the endometrium under careful management. However, there have also been a few reports of deaths after MPA therapy [ 1 2 3 5 ] since extrauterine lesions (Els) that occur during or after MPA therapy could be fatal [ 6 7 8 9 10 11 12 13 14 15 ]. In this study, we aimed to characterize patients with ELs during or after MPA therapy.

Between 1998 and 2021, 319 patients underwent MPA therapy at our hospital. Patient characteristics are shown in Table 1 . The number of patients with AEH and EMG1 at the time of initial treatment was 142 and 170, respectively. The median age was 35 (19–47) and 35 (19–44) years for AEH and EMG1, respectively; BMI was 21.9 kg/m 2 (20–43.4) and 21.8 kg/m 2 (15.8–37.9), respectively; and the median observation period was 66 months (4–244) and 76.5 months (4–269), respectively. AEH, atypical endometrial hyperplasia; BMI, body mass index; EMG1, endometrioid carcinoma grade1; PCO, polycystic ovary. The CR rates for AEH and EMG1 were 94.4% and 90.6%, respectively, and recurrence rates for AEH and EMG1 were 44.0% and 70.8%, respectively, for the initial treatment. The median time to recurrence was 11 and 9 months, respectively ( Table 2 ). AEH, atypical endometrial hyperplasia; CR, complete remission; EMG1, endometrioid carcinoma grade1; MPA, medroxyprogesterone acetate. During and after MPA therapy, cases of patients who developed PD were investigated irrespective of the presence of ELs. There were 15 cases where PD occurred without achieving CR after initial treatment, and of these 2 cases had ELs. The breakdown of PD cases was as follows: 10 cases with myometrial invasion, 5 cases with aggravated histology during treatment (duplicated). Two cases exhibited ELs during repeated treatment after relapse ( Table S1 ). Among the 319 patients who received MPA therapy between 1998 and 2021, 7 patients displayed ELs during MPA therapy or in the follow-up period after MPA therapy. These 7 cases were not limited to initial treatment, but also included cases where ELs were observed after recurrent treatment. ELs were detected during the first treatment in 2 patients, during the second treatment in 2 patients, during the follow-up period after 6 months of initial treatment in 1 patient, during the follow-up period after second treatments in 1 patient, and during the follow-up period after a further 4 treatments in 1 patient. Characteristics of the 7 patients are shown in Table 3 . Among these 7 patients, BMI ranged from 19.5–32.7. As stated, case 1, 2 and 7 underwent MPA therapy only once, while cases 3, 4, 5, and 6 were treated with MPA multiple times. Case 6 was treated 4 times and was diagnosed as AEH at the initial treatment but was subsequently found to have EMG1 after repeated MPA therapy. AEH, atypical endometrial hyperplasia; BMI, body mass index; DM, diabetes mellitus; EMG1, endometrioid carcinoma grade 1; GF, grand father; GM, grand mother; M, mother; MPA, medroxyprogesterone acetate; PCO, polycystic ovary. Cases 1, 2, 4, 6, and 7 had a history of uterine adenomyosis. Uterine leiomyoma was noted in case 4. PCO-like multifocal ovaries were seen in cases 1 and 5, but hormone levels were not examined, and a diagnosis of polycystic ovarian syndrome was not made. Endometriotic cysts were not seen in any case. Case 6 showed a history of multiple malignancies, gallbladder polyps, esophageal polyps, and a family history of dense malignancy. Table 4 shows how the diagnosis of ELs was made in the 7 cases. Cases 5, 6 and 7 were diagnosed during follow-up after MPA therapy, whereas the other cases were diagnosed during MPA therapy. The reason for CT scans in cases 1, 3 and 7 was an increase in the tumor marker CA125, and in cases 2 and 4 was the failure to achieve CR despite long-term MPA therapy, i.e., prolongation of time to CR (TTCR). On the other hand, an MRI scan was performed for case 5 due to elevated CA125 and CA19-9 levels and suspected myometrial invasion based on MRI results, leading to the diagnosis of para-aortic lymph node metastasis after surgery. EL was diagnosed by CT for case 6 incidentally during follow-up for colon cancer. The tumor marker CA19-9 showed mildly elevated levels. Since the CA19-9 level was elevated, the attending surgeon decided to perform positron emission tomography (PET) that showed fluorine-18-deoxyglucose accumulation on the left side of the uterus. Although a recurrence of colon cancer was also considered, myometrial invasion was clearly indicated by the results of an additional MRI scan, and the standard surgical procedure was implemented for treating a metastasis of EC. In case 7, which occurred more than 20 years ago, CA125 was not assessed because CA602 was the preferred marker in our institution at that time. However, CT was conducted due to elevated tumor marker levels. The results of the CT scan indicated an enlarged para-aortic lymph node and MRI indicated myometrial invasion, leading to a change to standard therapy. CR, complete remission; CT, computed tomography; EL, extrauterine lesion; MPA, medroxyprogesterone acetate; MRI, magnetic resonance imaging; PAN, paraaortic lymph node; PET-CT, positron emission tomography-computed tomography; PLN, pelvic lymph node; TTCR, time to complete remission. All 7 patients who had ELs underwent surgery. The results of the final diagnosis made by surgery are shown in Table 5 . Myometrial invasion was observed in all cases, and in cases 3, 4, 5, 6, and 7 had progressed to greater than one-half of the thickness of the myometrium. Lymph node metastasis was the most common site of extrauterine metastasis, and regional lymph node metastasis was observed in 5 cases (cases 1, 3, 5, 6, and 7), while case 2 was negative for lympho-vascular invasion and showed only peritoneal dissemination on the liver surface. Distant lymph node metastases were found only in case 1, and most were multiple metastases in remote areas such as Virchow and mediastinum. A cervical stromal invasion was observed in cases 4, 5, and 6, and adnexal metastasis was observed in cases 1, 6 and 7. Peritoneal cytology was positive in case 4, which showed a disseminated peritoneal lesion, but negative in the other cases. EL, extrauterine lesion; EMG1, endometrioid carcinoma grade 1; LVSI, lymphovascular invasion; PAN, paraaortic lymph node; PLN, pelvic lymph node; TNM, tumor-node-metastasis. As shown in Table 6 , case 1 was administered 4 cycles of doxorubicin + cisplatin (AP) therapy and 2 cycles of paclitaxel + cisplatin (TC) therapy after surgery, but lymph node metastasis increased, and the patient died because of EC 15 months after MPA therapy was initiated. Case 2 was diagnosed as PD after 8 cycles of TC therapy and 10 cycles of docetaxel + cisplatin therapy, and is now receiving lenvatinib + pembrolizumab therapy. Case 6 is currently receiving AP therapy because she was allergic to paclitaxel, and CR has been obtained at this point. In case 7, the disease recurred repeatedly, and the patient succumbed after undergoing approximately 9 years of chemotherapy. AP, doxorubicin + cisplatin; CAP, cyclophosphamide-doxorubicin-cisplatin; DP, docetaxel + cisplatin; 5-FU, fluorouracil; LP, lenvatinib + pembrolizumab; MPA, medroxyprogesterone acetate; TC, paclitaxel + cisplatin. The results of the immunohistochemical (IHC) tests are shown in Table S2 . For case 1, only the hematoxylin and eosin (HE) specimen from the D&C before surgery was available as it was an old case. As for case7, it was challenging to reevaluate pathologically because it was a case from more than 20 years ago, and neither the HE specimen nor tissue blocks were available. The remaining cases were examined using specimens obtained at the time of hysterectomy. In cases 2, 4, 5, and 6, GATA3, TTF-1, and p53 were all negative. In cases 3, the tumor exhibited a moderately differentiated adenocarcinoma pattern with areas showing p53 mutation pattern staining, leading to a final diagnosis of EMG2 after review.

MPA has a high response rate and is a feasible treatment for young patients who wish to preserve their fertility [ 1 2 3 4 ]. It is also known to have a high recurrence rate, but the reported number of cases of PD during MPA therapy is quite small. We speculate that this may be because MPA therapy is only adapted for treating AEH or early EC, or it may be that EC progresses relatively slowly. Interestingly, there were no AEH cases diagnosed with ELs if a patient had not been diagnosed with G1. In the 319 cases treated at our hospital over the past 25 years, 15 patients (4.7%) developed PD during initial MPA therapy. Among them, one patient was referred to another hospital, one died of a rapidly PD (case1), one died after recurrence and subsequent multiple treatments while the other patients are still alive after hysterectomy. This is because our protocols are managed in such a way that PD is diagnosed promptly. As mentioned, we follow up patients after CR every 3–4 months within 2 years and every 6 months after 2 years, and imaging studies (MRI, CT, and hysteroscopy) were conducted if recurrence was suspected. Japanese EC guidelines recommend a follow-up every 3 to 6 months (ver. 2023), and every 6 months in a guideline from the National Comprehensive Cancer Network [ 16 17 ]. Despite our strict follow-up protocols, there were 7 cases (2.2%) showing a presence of ELs during or after MPA therapy. Five of the 7 cases with ELs in this study showed lymph node metastasis. The risk of lymph node metastasis in EC is reported to be 4% in EMG1 or EMG2 cases and 1% in cases without myometrial invasion [ 18 19 20 21 ]. As mentioned, our protocol involves administering MPA to patients who have a strong desire for a baby and have been diagnosed with EMG1 without myometrial invasion. However, myometrial invasion had been detected histologically in all cases, suggesting the possibility of false negatives in the results of the imaging tests or the disease progression during or after MPA therapy. As for lymph nodes metastasis, some patients were found to have lymph node metastasis even though imaging tests had not been detected any. It ought to be noted that there are many different patterns of ELs development regardless of imaging test results. It should be further emphasized that all 7 cases diagnosed with extrauterine lesions (ELs) in this study also had intrauterine recurrent lesions, and there were no cases of ELs without accompanying intrauterine lesions. Therefore, the search for intrauterine lesions such as EMB is considered the most critical aspect of follow-up during observation. Furthermore, even after CR is achieved with MPA therapy, the possibility of recurrence remains, and the cancer may have progressed to EL at the time of detection. Thus, during the follow-up period, patients should be monitored using a combination of tumor marker quantification, TVUS, endometrial cytology, EMB, and imaging tests such as CT and MRI. Imaging tests should be particularly considered in cases diagnosed with EMG1, in patients who do not follow a typical clinical course, or in those who have not achieved CR for an extended period of time. CA125 is used as a tumor marker for EC [ 21 22 23 24 25 ]. Kadija et al. [ 26 ] suggested that CA125 concentrations ≥222 U/mL indicate the possibility of an EC stage III/IV or higher cancer, and have also reported that the positive rate of CA125 tends to increase with extrauterine extension or lymph node metastasis. However, patients for whom MPA therapy is indicated sometimes have endometriosis or benign ovarian tumors: therefore, we must consider false-positive CA125 levels. In addition, it has also been reported that the diagnosis rate increases when CA19-9 and CA125 blood tests are combined [ 27 ]. Case 6 was followed up with the value of CA125 and CA19-9 levels because she also had a history of colon cancer; however, only CA19-9 levels were elevated, and therefore, imaging was performed. Since PET-CT findings showed an ovarian mass suspected of metastasis, the patient was transferred to standard treatment due to the possibility of ELs. In fact, the patient had myometrial invasion of more than one-half of the thickness of the myometrium (21/23 mm), pelvic lymph node metastasis, and even left ovarian metastasis. Although tumor markers are just indicators, ELs should be considered in cases of atypical elevation of tumor markers. Among the 6 cases in which CA125 were investigated, 3 patients exhibited elevated CA125 levels, and only one out of 7 cases showed elevated CA19-9 levels. There is no consensus in the literature regarding long-term administration of MPA therapy. In Europe and the U.S., it is recommended to transition to the standard therapy if lesions do not disappear for 6–12 months [ 17 ], but in our institution, there was a surprising case where CR was achieved even after an initial MPA treatment of up to 34 months (this case was not included in cases 1–7). Cases 1, 2, and 5 showed ELs during MPA therapy, with a TTCR of more than 6 months. Although some researchers consider long-term administration of MPA therapy to be acceptable, the possibility that long-term administration may cause resistance to MPA therapy should also be considered. If MPA therapy is not effective for more than 6 months, the patient should be monitored carefully. Sometimes double cancers are reported during MPA therapy, especially with ovarian cancer [ 3 27 ]. It is possible that ovarian metastasis of EC may occur, or ovarian cancer may develop independently during MPA therapy. Ushijima et al. [ 3 ] reported a multicenter study where 3.6% of patients presented with simultaneous ovarian cancer. Of these, 4%–13.3% had metastasis from EC. In the present study, ovarian metastasis was observed in case 6 and 7. Case 6 was suspected of ovarian metastasis by PET-CT. Although she also had a history of colon cancer, it was difficult to determine whether the ovarian mass was metastasis from colon cancer or EC. Since it has been reported that multiple primary ovarian metastases can occur during MPA therapy, tumor marker levels and TVUS should be conducted during and after MPA therapy, while always keeping in mind the possibility of double cancer. Among 7 patients who experienced ELs, case 6 was suspected of having hereditary tumor syndrome because she had history of multiple cancers. According to ESGO/ESHRE/ESGE guidelines [ 28 ], hereditary tumor syndrome (especially Lynch Syndrome), is not a contraindication for MPA therapy due to the lack of evidence. It has also been reported that patients with deficient mismatch repair may have a poor prognosis compared to those with proficient mismatch repair [ 29 ]. Therefore, we should keep patients’ genetic condition in mind. As genetic assessment and analysis continue to advance, the way of using MPA for young patients who desire to have a baby may also evolve in the future. Three of the 7 patients were diagnosed with extrauterine metastases during and after MPA therapy but were transitioned to standard treatment without being endangered to a fatal status and are currently under observation without recurrence. On the other hand, one patient had a dramatic course of disease and passed away. Another patient had recurrence once, then recurred multiple times after standard treatment and passed away despite repeated bouts of chemotherapy. Two of our patients are still undergoing treatment for recurrent disease. Although MPA therapy is a fertility-sparing treatment, the risk of disease progression and death must be considered [ 6 7 8 9 10 ]. The most important aspect of oncologic monitoring is survival rather than fertility preservation, and early detection of extrauterine metastases that may lead to a fatal status is very important. As gynecologic oncologists, we should be aware that detecting ELs is significantly important. Pathological diagnoses, not limited to EC, can vary depending on the pathologist and the gynecologic oncologist reviewing the case. In this study, some cases date back more than 15 years, making it important to reassess the pathology to ensure diagnostic reproducibility. Therefore, a reevaluation of the pathological findings was conducted. The recent revision of the World Health Organization classification in 2020 has highlighted the diagnostic challenge of differentiating between endometrioid carcinoma and mesonephric-like adenocarcinoma, a malignancy associated with poor prognosis [ 30 ]. To aid in this distinction, IHC markers such as GATA-3, TTF-1, and p53 are employed. GATA-3 and TTF-1 play a crucial role in the differential diagnosis, particularly in distinguishing between advanced cancers like mesonephric-like adenocarcinoma and endometrial carcinoma. p53 is a widely known tumor suppressor protein, and its mutation is involved in various cancers including EC. Abnormal diffused p53 expression is strongly linked to high-grade aggressive tumor behavior and poor clinical outcomes. Since diffused p53 pattern were observed in case 3, findings suggested EMG2, but the final diagnosis did not change significantly. Given that the preoperative diagnosis in case 3 was EMG1, the result suggested potential disease progression during treatment. Nonetheless, since there are cases where the progression leads to ELs, caution is still required even for the patients with MPA therapy, which is basically expected to have a favorable prognosis. For both case 1 and 7 were old cases, we could not evaluate IHC tests but reevaluate only HE staining in case 1. However, as both showed poorer prognosis among those 7 cases, there were of interest for further analysis. As a limitation, the possibility of the presence of high-grade tumors including mesonephric-like adenocarcinoma cannot be ruled out by HE staining. It is important to note that these possibilities should be considered in the cases with unfavorable clinical prognosis or outcomes. MPA therapy for EC is intended for use during early-stage cancer, and the development of ELs during and after MPA therapy is very rare. The limitation of our study is that prognosis is difficult to determine because this was a single-center study and extrauterine involvement is rare. Although there have been reports of extrauterine metastases during MPA therapy, only 7 out of 319 cases have been reported in our hospital. Since there have been 2 deaths, MPA therapy for EC should be carefully administered by a well-experienced gynecologic oncologist under strict surveillance. In conclusion, in this study, we described 7 patients of extrauterine metastasis during and after MPA therapy, where 2 resulted in death. We should always keep in mind that MPA therapy is not a standard treatment, and it is essential to conduct imaging tests and screen for tumor markers during and after MPA treatment regularly and also when cancer progression is suspected.

The study included 319 patients treated with initial MPA therapy at our institution between 1998 and 2021. At our institution, the main indications for MPA therapy included patients with a current or future desire for children, those with a histological diagnosis of AEH or EMG1, and those with imaging tests showing lesions confined to the endometrium of the uterus. With approval from the ethical committee of Keio University School of Medicine (ethical approval number: 20110237), the characteristics of the cases with ELs were reviewed based on medical records. We collected clinicopathological data of all patients such as age at the time of initial diagnosis, body mass index (BMI), history of polycystic ovary (PCO), diabetes, and pregnancy, and examined CR rate, recurrence rate, and prognosis. The detailed indications for MPA therapy at our institution were patients diagnosed with AEH or EMG1 by endometrial dilation and curettage (D&C), magnetic resonance imaging (MRI) or a computed tomography (CT) scan showing lesions confined to the endometrium of uterus, and no evidence of myometrial invasion. Minor criteria included: 1) age 42 or younger, 2) no severe obesity (BMI <35), 3) no liver function abnormalities, 4) no coagulopathy, 5) no history of thrombosis, 6) non-smoker, and 7) willing to accept close follow-up and come to the hospital regularly and frequently. If these criteria were not met, patients were treated surgically with hysterectomy as a general rule. In principle, MPA (Hysron®-H Tablets 200 mg) was administered orally at a dose of 600 mg/day. After the start of treatment, a medical interview, internal examination, transvaginal ultrasonography (TVUS), endometrial cytology, endometrial biopsy (EMB), and tumor marker detection such as CA125 were performed every month. After 4 months, D&C under intravenous anesthesia was performed to evaluate the efficacy of MPA, with the disease location having been confirmed beforehand via hysteroscopy. Treatment was continued until histological examination showed an absence of endometrial hyperplasia: this is defined as CR. If the lesions did not disappear, D&C was performed every 2 months unless the lesions worsened or were diagnosed as progressive disease (PD). PD was defined as myometrial invasion diagnosed by MRI and/or an aggravated histological diagnosis, while ELs were diagnosed using imaging tests. In cases of non-CR or non-PD, patients continued MPA therapy. Even if a patient was not diagnosed as PD, the patient underwent a hysterectomy unless CR was achieved within a year as a general rule. After MPA therapy, endometrial cytology, biopsy, and TVUS were performed every 3–4 months within 2 years and every 6 months after 2 years, and imaging studies (MRI, CT, and hysteroscopy) were performed if recurrence was suspected. Additionally, patients with EMG1 underwent CT scans annually as follow-up. The presence of the tumor marker, CA125, was tested at each hospital visit. Patients were performed the standard treatment when there was no longer any desire to have a baby, when the disease progressed beyond the criteria of MPA therapy, such as when the histological type worsened to G2 or more, when a myometrial invasion was observed upon MRI, or when ELs were suspected via imaging tests. Based on these criteria, we clarified the characteristics of patients undergoing MPA therapy and those with ELs at our institution. The cases presenting ELs were reviewed again by a gynecologic pathologist and a gynecologic oncologist. GATA-binding protein 3 (GATA3), thyroid transcription factor-1 (TTF-1), and p53 staining were examined, and the possibilities of high-grade carcinoma and mesonephric-like adenocarcinoma were discussed. With uterus specimens with abundant tissues, additional immunostaining exam was performed to evaluate the pathology.