Progestogen treatment options for early endometrial cancer
Dr TJ Cade, Royal Women’s Hospital, Parkville, Melbourne, Vic. 3052, Australia. Email firstname.lastname@example.org
Please cite this paper as: Cade T, Quinn M, Rome R, Neesham D. Progestogen treatment options for early endometrial cancer. BJOG 2010;117:879–884.
Objective Premenopausal women with early endometrial cancer may wish to maintain their fertility, and for some patients non-surgical treatment options may be attractive. We have examined our own experience with such patients, as there are limited published data so far to support clear guidelines in this area.
Design Retrospective analysis of a case series.
Setting Case series from a specialist gynaecological oncology unit in a major tertiary referral hospital.
Sample Sixteen patients receiving progestogen therapy for stage-1 endometrial cancer.
Methods We reviewed our experience of all patients receiving progestogen therapy for stage-1 endometrial cancer, and we particularly examined their cancer-free outcome and fertility potential.
Main outcome measures Response to treatment, duration of response, and subsequent pregnancies.
Results Of the 16 patients investigated, four received an oral progestogen, five received the levonorgestrel-releasing intrauterine system (Mirena), and seven received both forms of treatment. Ten patients (63%) responded to treatment, with a median time to response of 5.5 months. Six patients did not respond to treatment, but all were either early in treatment or opted for surgical management before the average time of response. No patient who responded had a later recurrence. The mean total follow-up time was 27 months (range 3–134 months), with no patient deaths. Three patients had successful pregnancies, with one patient having two children.
Conclusions This form of treatment appears to be a realistic treatment option in selected patients in the closely supervised environment of a specialist gynaecological oncology unit.
Endometrial cancer is currently the most common gynaecologic malignancy.1–4 In developed countries, it has an incidence of seven new cases per 100 000 population per year, and an overall mortality of nearly 20%.1 Current treatment is primarily surgical, typically with total abdominal hysterectomy and bilateral salpingo-oophorectomy, with additional lymph node dissection in high-risk cases.2,4–6
About 20% of cases occur in premenopausal women,1,4 for whom maintaining fertility can be a significant concern, as many may not have started or completed their expected family.4 In addition, surgical treatment may carry substantial risks in some patients, such as the morbidly obese or those with serious co-morbidities.
Progestogens have been used in the treatment of recurrent disease for over 50 years, with reported response rates as high as 80% where the recurrence is single, well differentiated and in a site not previously irradiated, such as the lungs.4,7–9 The presence of steroid receptors, however, is the single most important predictor for response. As well-differentiated tumours tend to retain their hormone receptors (for oestrogen and progesterone), because endometrial proliferation is oestrogen dependent, and because progestogens have an anti-oestrogenic effect on the endometrium,4,8–10 their use has also been assessed as a primary treatment for stage-I, grade-1 endometrial cancer11 in selected patients.4,7–9
There have been few studies published regarding this treatment approach, and although there appears to be a theme of generally promising results, as yet there are no large patient series following this form of treatment to support reliable guidelines.4,16–27
To provide some further perspective, we have evaluated the outcome of all patients recently exercising this non-surgical treatment option at two major tertiary, university-affiliated hospitals. In particular, we examined the effect of treatment on their disease and their ability to become pregnant.
A search of all patients with endometrial cancer was performed in the gynaecological oncology database of the Royal Women’s Hospital, and in the database of one of the hospital’s gynaecological oncologists from his private practice at Epworth Freemasons Hospital, Melbourne, Australia. Both hospitals are in Melbourne, Victoria, Australia, and are affiliated with the University of Melbourne. All patients with grade-1 endometrial cancer were selected,11 and their treatment details were examined. Those patients who had been treated hormonally with oral and/or intrauterine progestogens were then included in the study.
Patients were deemed suitable for this form of treatment provided they met strict inclusion criteria: firstly based on defined histological findings, secondly based on no myometrial invasion being identifiable on a magnetic resonance imaging (MRI) scan, and thirdly following detailed discussion with the patient explaining that this form of treatment is not current standard treatment, and therefore requires meticulous follow up. This included serial curettage with failure of disease regression warranting surgical treatment, and, for pre-menopausal patients, the recommendation of a hysterectomy once their expected family was complete.
The histology was reported by a single experienced pathologist in each institution using the same international criteria.12–15 Their findings, including a demonstration of the histological slides, were then presented for multidisciplinary discussion and decision at weekly clinicopathological meetings that the relevant clinicians, pathologists, and radiologists attended.
Background data were obtained from the patient databases and the hospital medical records. Age at diagnosis, gravidity/parity, previous diagnosis of polycystic ovary syndrome, other predisposing factors, duration and type of symptoms (menorrhagia, irregular bleeding, or neither), and co-morbidities were individually recorded.
The treatment for each patient was recorded as oral progesterone and dose, levonorgestrel-releasing intrauterine system (Mirena; Bayer HealthCare, Sydney, NSW, Australia), or both. Our current management protocol now combines both the intrauterine device with oral medroxyprogesterone acetate of 200 mg twice daily for 6 months. Recurettage is then undertaken and the device is reinserted. If the histopathology is benign, the device can be removed and oral treatment ceased if fertility is required. If the histopathology demonstrates atypia or persistent malignancy, this process is repeated at 12 months. If the abnormalities have not regressed, we recommend surgery at that time.
This protocol evolved with time, in that oral progestogen alone was initially used prior to the availability of the intrauterine device, and the combination regimen was used thereafter, except in those patients who could not tolerate the side-effects of oral progestogen. In addition, the first curettage was performed in some initial cases as early as 3 months after diagnosis, based on clinician or patient preference at the time.
For every patient, each subsequent dilatation and curettage was noted with the time in months after the start of treatment and the histology. A positive response was a normal curette or hyperplasia without atypia, whereas a negative response was hyperplasia with atypia or residual malignancy.
The total follow-up time for each patient was documented, together with the number of viable pregnancies, any recurrence of disease, and residual disease at hysterectomy if later performed.
There were 16 patients in the study. The mean age at diagnosis was 35 years (range 23–53 years). Fifteen patients were nulliparous, four had polycystic ovary syndrome, two had first-degree relatives with endometrial cancer, two had hereditary genetic syndromes (one with familial adenomatous polyposis and one with hereditary non-polyposis colorectal cancer), and one had previously been treated for clear cell carcinoma of the ovary. One patient was morbidly obese and suffering from a thromboembolic event.
Symptoms were irregular bleeding (five patients), menorrhagia (two patients), both (one patient), or neither (comprising eight patients who were investigated for primary infertility). Quantifying the duration of symptoms was problematic, as most patients were asymptomatic, whereas others were unsure of how long they had had symptoms, or answered with vague terms such as ‘months’. Of the eight patients who had an MRI scan, none had evidence of myometrial invasion, and all were therefore included in the study.
Treatment comprised an oral progestogen only (four patients), Mirena (three patients), or both modalities (nine patients). Treatment methods and follow up are summarised in Table 1.
Table 1. Details of patient treatment and response
| 1||31||Y||200 bid||N||N||–||Y||Y|
| 2||38||Y||200 bid||N||N||–||Y||Y|
| 3||42||Y||200 bid||N||Y||4||Y||N|
| 4||23||Y|| 60 od||N||Y||4||Y||Y|
| 5||34||N|| –||Y||Y||7||N||NA|
| 6||45||N|| –||Y||N||–||N||NA|
| 7||53||N|| –||Y||N||–||Y||Y|
| 8||33||Y||200 bid||Y||Y||6||N||N|
| 9||27||Y||100 bid||Y||Y||13||N||NA|
Ten of the 16 patients (63%) responded to treatment, and the mean time to response was 5.5 months. Three of these patients underwent a subsequent hysterectomy, one of whom had residual disease and two of whom were histopathologically normal. The patient with residual disease (patient 4 in Table 1) had a prophylactic hysterectomy 38 months after treatment (with previously normal curettes). The two patients reported as histopathologically normal (patients 3 and 10 in Table 1) both had subsequent curettes reported as hyperplasia (one with atypia and one without) after at least one normal curette.
Six of the 16 patients (patients 1, 2, 6, 7, 13, and 16) did not respond to treatment. All of these patients were in the early stages of treatment and below the mean time for response, and five underwent hysterectomy: two (patients 7 and 13 in Table 1) elected for a hysterectomy before a follow-up curette; three (patients 1, 2, and 16 in Table 1) had an abnormal first follow-up curette, and elected for a hysterectomy at that time; and the remaining patient (patient 6 in Table 1) was treated with radiotherapy before a follow-up curette, as she persistently expelled the intrauterine device, was unable to tolerate oral treatment, and was unsuitable for major surgery.
Successful pregnancy occurred in three patients, one of whom has now had two children, one of whom has had one child, and one of whom has had one child and two ectopic pregnancies.
The mean total follow-up time was 27 months (range 3–134 months), with no patient deaths and no patients lost to follow up.
Recent publications regarding progestogen treatment of endometrial cancer in recent years have chiefly comprised case reports and potential treatment guidelines.16,19–22,24,26,27 They suggest encouraging results, with remission rates as high as 75–85%. These clinical reports are summarised in Table 2.
Table 2. Summary of previously published clinical reports
|Susumu et al.25||14||MPA 600 mg od||86% response over 8 weeks|
14% no response (had myometrial invasion)
|Montz et al.22||12||Levonorgesterol IUD||64% biopsy negative at 6 months|
75% biopsy negative at 12 months
|Wang et al.16||9||Megesterol acetate, tamoxifen, gonadotrophin-releasing analogue||89% complete response|
50% subsequent recurrence
11% no response
|Ramirez et al.21||81 (literature review)||Review of all with oral or IM hormonal treatment||76% response with median time of 12 weeks|
24% initial response then recurrence
(median time 19 months)
23% no response
|Dhar et al.24||4||Levonorgesterol IUD||25% complete regression in 6 months|
75% no response in 6 months (one had myometrial invasion)
|Wang et al.26||6||Oral (dose not specified)||67% total response|
50% recurrence between 10 and 12 months
33% no response
|Yamazawa et al.19||9||MPA 400 mg od||78% complete response in 6 months|
22% partial response
22% recurrent disease between 10 and 22 months
|Chiva et al.20||133 (literature review)||Review of all hormonal treatment||76% complete response after an average of 12 weeks|
34% of these had a recurrence
24% no response
Four published deaths of patients conservatively treated
|Eftekhar et al.27||21||Megesterol acetate|
160–320 mg od
|86% response rate|
The present study also comprised relatively small numbers and a limited duration of follow up, but our initial assessment is that this form of treatment appears promising. Although we report a response rate of 63%, all patients who failed to respond were either early in treatment or opted for surgical management before the average time of response (5.5 months). It should also be noted that no patient had a curettage specimen reported as cancerous after receiving a previously normal result.
Our results are in accord with the recent report by Wheeler and colleagues of promising histological results in a series of patients with endometrial cancer or complex hyperplasia with atypia, given progestogens of various types, and by various routes.28 We concur with their recommendation that a minimum period of 6 months of treatment is required before a response can be accurately assessed.
There have been as yet no randomised studies in this clinical situation comparing types of progestogenic agents, and doses or routes of administration.4 However, the use of a combined oral and intrauterine approach seems attractive, particularly when there may be myometrial invasion undetectable on MRI scans. Although the numbers of patients were small, it appeared that the treatment regimen of combined oral and intrauterine progestogen was probably more effective than either modality alone. On the other hand, in some patients for whom this treatment option may seem particularly appropriate, such as the morbidly obese, the benefits need to be balanced against the increased risk of significant side effects from high-dose medroxyprogesterone, such as appetite stimulation, weight gain, dyslipidaemia, and thromboembolism.
We conclude that this form of treatment appears to be confirmed as a realistic treatment option in selected patients, with a safe cancer outcome for most in the short to medium term of the study, with rescue surgery being required in only about half the patients, and with successful pregnancy in some. However, we caution that this form of treatment cannot yet be considered a routine recommendation, but we suggest that it can be appropriate in the closely supervised environment of a specialist gynaecological oncology unit in a tertiary hospital. We stress that important further considerations are the need for pathological expertise, and an MRI scan negative for myometrial invasion or extension to the cervix. Finally, it is important that the patient is well informed and well motivated for informed consent, so that there is a clear understanding that this treatment program is not currently standard treatment, and that it is not likely to remain curative in the long term. Thus, the patient must not be lost to follow up, as hysterectomy is likely to be required eventually.
Disclosure of interests
The authors declare that they have no conflicts of interest with regard to any of the material presented in this paper.
Contribution to authorship
We confirm that each author has contributed materially to the paper, and that no individuals qualified for authorship have been omitted. The order of authorship is related to the relative individual contributions.
Details of ethics approval
The study was approved as an audit, and the need for individual patient consent was waived by the institutional ethics committee.
The study we report on here has been the subject of an advanced trainee project by T.J.C., and has received no financial support.
Commentary on ‘Progestogen treatment options for early endometrial cancer’
In developed countries the number of hysterectomies performed has been falling as a result of the successful conservative treatment of menstrual disorders. Could it be that the number performed for endometrial cancer could also show a similar decline? Cade et al. (BJOG 2010; doi: 10.1111/j.1471-0528.2010.02552.x) report on their experience of using progestogens for the treatment of early endometrial cancer. This subject is important as the number of younger women with endometrial cancer is rising because of increasing obesity. In recent years, the surgical management of the majority of endometrial cancers has become less extensive. Hysterectomy removes the primary source of the cancer and allows the assessment of the degree of local invasion. The ovaries are removed to eliminate a source of oestrogen, and to remove metastatic or synchronous tumours. The role of lymph node dissection is controversial (Creasman et al., Gynecol Oncol 2010;116:293–4).
Two recent randomised controlled trials have suggested that there is no survival benefit in performing a pelvic lymph node dissection (Kitchener et al., Lancet 2009;373:125–36; Benedetti Panici et al., J Natl Cancer Inst 2008;100:1707–16). However, the conclusions of these papers differed. The paper by Benedetti et al. concluded that pelvic lymph node dissection may be important in determining the need for adjuvant therapy.
The other paper by the A Study in the Treatment of Endometrial Cancer (ASTEC) group stated that pelvic node dissection cannot be recommended in early endometrial cancer (Kitchener et al., Lancet 2009;373:125–36).
Cade et al. report on their experience both with women unfit for a general anaesthetic and with premenopausal women who wish to conserve their fertility. The latter group includes women with hereditary cancer syndromes or chronic anovulation conditions, such as polycystic ovary syndrome. The paper reports on 16 women treated with various combinations of oral progestogens and the levonorgestrel-releasing intrauterine system, and monitored by repeat curettage at 6 months in a gynaecological oncology centre. Ten women responded to this regime and three had successful pregnancies.
These results are broadly similar to previous reports, the largest of which is a literature review including 133 young women (Chiva et al., Gynecol Oncol 2008;111:S101–4). This latter series recorded a lasting response of only 51% and, more worryingly, four deaths. Both studies have excluded the more aggressive histopathological types such as serous papillary, clear cell and carcinosarcomas, which are frequently associated with extrauterine dissemination even without myometrial invasion. Cade et al. used strict criteria before offering conservative treatment. The women had their histopathology reviewed by a single reference centre, and were deemed to have well-differentiated tumours without myometrial invasion detected on a magnetic resonance imaging (MRI) scan. Although this approach seems logical, there are some caveats. Even when the curettage histopathology specimen is reviewed in a gynaecological cancer referral centre, between 15% and 20% of women with grade-I cancer on a curettage specimen will be upgraded to grade-II or -III cancers on the final hysterectomy specimen (Obermair et al., Int J Gynecol Cancer 1999;9:383–86; Leitao et al., Gynecol Oncol 2008;111:244–8). In a similar manner, standard contrast-enhanced MRI has a limited ability to detect microscopic superficial myometrial invasion, rather than deep myometrial invasion, and is poor at detecting pelvic lymph node spread (Harry et al., Int J Gynecol Cancer 2009;19:186–93). It is therefore possible that a number of higher risk women will undergo conservative management. An additional concern is that many of these women will be subfertile and may require assisted conception with ovarian stimulation, leading to high oestrogen levels, with a possible effect on any remaining cancer cells. Cade et al.’s paper highlights the importance of fully informed consent and continued follow up, as the woman and clinician are taking a calculated risk. For women with endometrial cancer, a hysterectomy with a bilateral salpingo-oophorectomy (with or without a lymph node dissection) remains the standard treatment.
Disclosure of interest
Belfast city Hospital, Belfast, UK