Abstract
Aim: To compare the risk of developing endometrial carcinoma (EC) in young women with atypical endometrial hyperplasia (AEH) undergoing fertility-sparing management compared to women treated by primary hysterectomy. Patients and Methods: In this multicentric retrospective study, 111 patients with a diagnosis of AEH by endometrial biopsy were included. EC incidence was compared in two groups: 32 patients treated with fertility-sparing management and 79 older patients treated with primary hysterectomy. Results: The rates of EC diagnosed by pathology of hysterectomy specimens were comparable between the groups. The probability of developing EC at 12, 24 and 36 months were 14%, 21% and 26%, respectively, in patients managed conservatively, and 29%, 37% and 37%, respectively, in patients treated with primary hysterectomy. Conclusion: Fertility-sparing management of AEH does not increase the risk of diagnosing EC from the hysterectomy specimen.
Atypical endometrial hyperplasia (AEH) is a precursor to endometrial carcinoma (EC), an estrogen-dependent cancer that presents itself along a continuum (1). The co-existence of EC or the progression to EC in patients diagnosed with AEH has been shown in several studies; EC was detected at the final histological analysis in 10% to 59% of AEH cases (2-4).
Approximately 5% of patients with EC are diagnosed before the age of 40 years (5). The standard treatment for AEH is hysterectomy, which is often unacceptable to women of childbearing age. Over the past 40 years, the fertility-sparing management of AEH has been reported in the literature (6, 7).
Oral progestin appears to be a good alternative to hysterectomy, with a regression of the lesions in 75 to 95% of patients with AEH. After a complete response, the recurrence rate varies from 20 to 38% of the cases, depending on the follow-up period (8-10). Other medical treatments [levonorgestrel-releasing intrauterine system, and gonadotropin-releasing hormone (GnRH) agonist] have been reported to produce similar results (11). The medical treatment should not be considered curative because recurrence is always possible. Obtaining remission with medical treatment allows one or several pregnancies before a more radical treatment is required. One of the most important factors that patients with AEH (and their practitioners) should know is whether they are at an increased risk of cancer during this period.
The aim of our study was to evaluate whether the rate of EC is increased in women with AEH who undergo fertility-sparing management compared with women who receiving non-conservative treatment.
Patients and Methods
Patients. We conducted a multicenter retrospective analysis that included two groups of patients diagnosed with AEH between 2001 and 2013. The first group consisted of patients extracted from the National Registry, which collects cases of AEH and EC managed conservatively. The patients with AEH were referred to our Center from 17 French Gynecological Units. The second group of patients did not have conservative disease management (primary hysterectomy group) and came from two French gynecological units (including the referral center for fertility preservation in cases of AEH). For the two groups, we included only patients who met the following inclusion criteria: Diagnosis of AEH was confirmed independently by two pathologists (including a reference pathologist); minimum follow-up period of six months or until a hysterectomy was performed; radiological examination was conducted with at least an ultrasound examination.
Patients who underwent fertility-sparing management were selected for this procedure if they were 42 years old [age limit for assisted reproductive technology (ART) reimbursement by the National Insurance system in France] or younger and had a strong desire to preserve fertility. We excluded all cases that received conservative management due to any concern other than fertility sparing (e.g. medical contraindication to surgery) from the study. Patients with conservatively treated EC have been reported in other studies (12,13). Patients who met the criteria were counseled extensively regarding the risk of recurrence or progression if they chose medical therapy and gave informed consent for the treatment.
Patients who underwent non-conservative treatment were included in the study if the first option of treatment was surgery for the treatment of AEH.
This study was approved by the Ethics Committee of the College National des Gynécologues Obstétriciens Français (Institutional Review Board number: 2012-GYN-07-02).
Patient management. The patients in the fertility-sparing management group were scheduled to receive the medical treatment for a minimum of three months. Given that a higher incidence of synchronous ovarian cancer has been reported in young patients with EC (14,15), a diagnostic laparoscopy was performed prior to medical treatment in the last 22 patients. In no case did the laparoscopic findings modify the disease staging.
All patients included in this study were assessed for infertility prior to conservative management. Treatment response was assessed using pathological specimens obtained by curettage or endometrial biopsy after 3-6 months of medical therapy. Remission was indicated if the last endometrial sample or hysterectomy specimen showed normal endometrium without hyperplasia. Persistence was defined if the last biopsy showed AEH. Recurrence was defined if a lesion that had initially regressed following treatment reappeared in the form of AEH or EC.
For patients showing no response to medical therapy, total hysterectomy with or without bilateral salpingo-oophorectomy was proposed. If the patient wished to continue the fertility-sparing management and if no progression was diagnosed, a secondary medical therapy different from the first was attempted. For example, if no remission occurred after 3-6 months of progestin therapy, a GnRH agonist could then be given. After the documentation of complete remission, women had a follow-up visit every 3-6 months with diagnostic hysteroscopy and endometrial biopsy. Patients were encouraged to conceive spontaneously. ART could be attempted after complete remission and was based on a couple's fertility parameters. Women who failed in their attempts to conceive or who successfully completed their desired childbearing were encouraged to undergo definitive surgery.
Patients were counseled to undergo hysterectomy if EC was found during the follow-up or if AEH persisted despite two different treatments.
Statistical analysis. Fisher's exact test, the Chi-square test and the two-sided t-test were employed for statistical analyses. Significance was set at the standard value of p<0.05. Kaplan–Meier curves were generated to examine the probability of finding EC in the hysterectomy specimens. The endpoint was the last known disease status, and patients who were lost to follow-up were censored. In cases for which hysterectomy was not performed, we considered there to be no progression if neither the endometrial biopsy nor the hysteroscopy showed AEH or EC, and the uterus was preserved. The log-rank test was used to compare the probability of a diagnosis of EC in the hysterectomy sample between the two groups.
When analyzing the outcomes of patients, we used the intention-to-treat (ITT) principle. Even when treated conservatively, patients sometimes undergo early hysterectomy (within three months) because they change their mind or because of persistent disease, 3 months after the beginning of hormone therapy. In contrast, even when primary surgery is decided upon in the management of AEH, it is sometimes delayed because of anesthetic risks to the patient related to age, obesity, the patient's treatment or because of the patient's wishes concerning the date of surgery. In such cases, the delay between the diagnosis and hysterectomy is similar to the delays usually seen in patients who undergo a fertility-sparing management. The discrepancy between the decisions made to plan a patient's treatment and what is actually done in real practice is the cornerstone of ITT.
The data were analyzed with the R software package, version 2.13.0, using the Design, Hmisc and Verification libraries (http://lib.stat.cmu.edu/R/CRAN/).
Results
Patients' characteristics. One hundred and eleven patients diagnosed with AEH on endometrial biopsy were included in the study, including 32 patients with fertility-sparing management and 79 with a primary hysterectomy. The patient age was significantly different between the groups, ranging from 23 to 42 years for the conservative treatment group and from 38 to 87 for the non-conservative treatment group (p<0.001). The diagnosis of AEH was made during infertility investigations in 20 vs. 0 cases and during an investigation of irregular bleeding in 11 vs. 70 cases in the fertility-sparing management and non-conservative treatment groups, respectively (Table I).
Patients' management and follow up. Among the 32 patients treated with conservative management, 23 patients received progestin, four received a GnRH agonist, and five did not receive any medical treatment (one became pregnant before starting treatment, and hysteroscopic resection was considered complete in four cases). Twenty-five (78%) out of the 32 patients managed conservatively had remission after a mean time of 5 months (2 to 12 months), 22 following the first treatment and three after the second-line treatment. Four out of these 25 patients (16%) had a recurrence, with a mean delay of 20 months between remission and recurrence. Twelve women (38%) achieved 14 pregnancies, four of which were spontaneous. A total of 11 live births were recorded.
Among the patients belonging to the primary surgery group, 60 (76%) had undergone an outpatient diagnostic hysteroscopy or an operative hysteroscopy before surgery. The mean delay to hysterectomy was 3.2 months, ranging from 1 to 23 months.
Outcome. The mean follow-up was 48 months for patients treated conservatively and 15 months for patients whose primary treatment was hysterectomy (p<0.001). Thirteen patients (41%) with fertility-sparing management ultimately underwent hysterectomy, with a mean delay of 24 months after the diagnosis of AEH. The risk of finding EC was comparable between the groups: 28% in women with fertility-sparing management and 24% in patients with initial hysterectomy (p=0.64). All patients who were managed conservatively and did not undergo hysterectomy had no evidence of AEH or EC on the last endometrial biopsy sample. Among the patients who underwent fertility-sparing management, the mean follow-up was 52 months for patients who ultimately underwent hysterectomy and 49 months for those who did not (p=0.79). The risk of finding stage IB or more EC was comparable between the groups: 9% (three cases) for women with fertility-sparing management and 5% (four cases) in the primary hysterectomy group (p=0.68) (Table II).
Of the 13 patients treated with conservative management who ultimately underwent a hysterectomy, two were performed after pregnancy, five were performed for persistence of AEH, four for recurrence, one for disease progression, and one underwent hysterectomy after ART failure.
The final histopathologyical evaluation obtained from the two hysterectomies performed after pregnancy showed non-atypical hyperplasia. Three stage IA, one stage IIIA and one stage IIIC ECs were diagnosed in patients who underwent hysterectomy for persistent AEH. One AEH, two stage IA and one stage IIIA ECs were diagnosed in the patients who underwent hysterectomy for disease recurrence. Stage IA EC was founded in the patient who underwent hysterectomy for progression, and AEH was found in the one who underwent hysterectomy after ART failure.
The probability of a diagnosis of EC in the hysterectomy sample over time is shown in Figure 1. For patients treated conservatively, the probability of EC at 12, 24 and 36 months was 14%, 21% and 26%, respectively. For the patients treated by primary hysterectomy, the probability of EC at 12, 24 and 36 months was 29%, 37%, and 37% (p=0.08).
Discussion
In the present study, we compared the risk of developing EC in women with AEH undergoing a fertility-sparing management and older women treated by primary hysterectomy. The oncological outcomes were comparable between the groups, and the only difference between the groups concerned the delay in diagnosis of EC. The highest probability of finding EC was observed later in women treated conservatively. This finding suggests that medical treatment permits a delay in the occurrence of EC, but does not prevent it from occurring.
Although this study is one of the largest to report the fertility-sparing management of AEH only, the number of patients included might seem too low to draw strong conclusions. However, AEH is rarely diagnosed in young women and in order to limit recruitment bias, all the women included in the present study were selected only if they met criteria that are not systematically checked when the diagnosis of AEH is made. Importantly, because we did not find any significant difference between the two groups in terms of cancer probability (analyzing the outcome of more than 100 well-selected women), it is unlikely that fertility-sparing management significantly increases EC incidence. Most studies of fertility preservation in women with endometrial lesions concern both AEH and early EC. We believe that AEH and EC outcomes should be analyzed separately. Recently, two meta-analyses of fertility-sparing treatment suggested that these diseases have different prognoses (16, 17).
Importantly, the current study is the first to include a control group of patients (who underwent primary surgery) treated at the same time and at the same institutions to compare the risk of EC in patients with conservatively managed AEH. The aim of the present study was not to compare oncological outcomes between young and older women treated for AEH, but rather to compare women who received fertility-sparing treatment versus women who elected to undergo primary surgery. Consequently, one of the underlying limits of this study is the significant difference in age between the groups. Most patients who elected to undergo primary surgery probably did not request a fertility-sparing management strategy because they were 45 years or older and were unable to achieve pregnancy or did not wish to. Despite this limitation, several patients in the primary surgery group still had childbearing potential.
The impact of age on the prognosis of AEH remains debated. In a study of 773 women who underwent hysterectomy for AEH, EC was found in 38% of pre-menopausal patients and 64% of the post-menopausal patients (p<0.001) (2). In another study of 824 patients diagnosed with AEH, the probability of finding EC in hysterectomy specimens was independently correlated with age (18). Based on these studies, one could expect that EC would be less likely to occur in the conservatively-treated group. However, in studies reporting a higher rate of EC among older women, the reported rates of EC in hysterectomy specimens are particularly elevated and are higher than the rates of EC observed in our population. A randomized study with two populations comparable in age would be impossible because it would involve fertility-sparing interventions. Even a retrospective study is unlikely because young women with AEH are often willing to preserve their fertility and most cases in young woman are diagnosed during explorations of infertility (19). Interestingly, infertility and AEH are both associated with the following conditions: estrogen exposure, obesity, chronic anovulation, and polycystic ovarian syndrome.
None of the patients who achieved pregnancy had a recurrence of AEH. This finding can be explained by the high levels of progesterone produced during pregnancy, that could have played a role in the treatment itself (20). Gynecologists treating such patients should advise women to conceive rapidly and resort to ART if necessary (21). Similarly, to improve the chances of pregnancy in an obese patient diagnosed with AEH, weight loss is part of the conservative management (22) but once pregnancy has been achieved, hysterectomy should be proposed. Postpartum EC cases have been described (23).
In both study groups, we recorded unfavorable outcomes concerning advanced EC. The underlying question is: Was this finding due to the conservative management or was this advanced-stage disease present from the time of diagnosis of AEH? Our study is a retrospective multicentric study reflecting real life rather than the best possible standard. Since unfavorable outcomes occurred in both groups, an optimal follow-up with histology is indispensable in patients managed conservatively. Similarly, in the pre-therapeutic evaluation of patients diagnosed with AEH, systematic magnetic resonance imaging is mandatory.
The principal obstacle to the fertility-sparing management of AEH relates to the difficulty of initially obtaining a specific diagnosis. There is neither a very reliable clinical parameter nor an imaging feature which distinguishes between AEH and EC. Endometrial biopsy has a low accuracy (55%) for differentiating AEH from EC compared with dilatation plus curettage (73%) (24) or hysteroscopy-directed biopsy (92%) (25). Moreover, there is poor reproducibility in the pathological diagnosis of AEH (26). Only the pathological analysis of the hysterectomy specimen is able to exclude the need for concurrent EC at the time of diagnosis.
Fertility-sparing management of women diagnosed with AEH is conceivable because this lesion has a slow evolution. In fact, 20% of patients are cured spontaneously (28).
This study confirms that hormonal treatment does not preclude the development of EC but instead delays the occurrence of EC without increasing its risk.
Acknowledgements
The Authors would like to thank Pr Benifla, Dr Marchand (Paris), Dr Challan Belval, Dr D'Argent (Paris), Pr Poncelet (Bondy), Pr Dauplat, Dr Bergzoll, Dr Bourdel (Clermont Ferrand), Pr Fernandez (Le Kremlin Bicêtre), Pr Deffieux (Clamart), Pr Raudrant, Pr Golfier (Lyon), Pr Morice, Dr Uzan, Dr Gouy (Villejuif), Dr Fourchotte (Paris), Dr Le Tohic, Dr Panel (Versailles), Pr Collinet, Dr Boyon, Dr Clouqueur, Dr Lucot (Lille), Dr Villefranque (Pontoise), Dr Debiolles (Beaumont), Pr Graesslin, Dr Derniaux (Reims), Dr Vincens (Montpellier), Pr Lecuru, Dr Hauser Douay, Dr Fournier (Paris), Pr Haddad, Dr Boujenah (Créteil), Dr Mabrouk (Amilly), Dr Ohl (Schiltigheim), Dr Faller, Dr Schindler, Pr Wattiez (Strasbourg), Dr Parmentier (Rang-du-Fliers), Dr Grynberg (Bondy), Dr Lambaudie (Marseille), Pr Brun, Dr Conri, Dr Legrand (Bordeaux) and Pr Agostini (Marseille) for providing us with details of patients under their care.
- Received July 29, 2015.
- Revision received October 19, 2015.
- Accepted October 23, 2015.
- Copyright© 2015 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved