Endometrial hyperplasia (EH) is a pre-cancerous proliferation of endometrium that results in increased volume and altered architecture of endometrial tissue with endometrial gland to stroma ratio of greater than 1:1 [1]. New 2014 World Health Organization (WHO) classification stratifies EH by the presence of nuclear atypia into atypical and non-atypical forms, which correspond to low and moderate-high risk of developing endometrial cancer [2]. Non-atypical hyperplasia occurs more commonly, and while undergoes spontaneous regression in 80% of cases, 3%–10% progress to atypical forms and 1%–5% to endometrial cancer [345].

EH occurs in 6.19–114.36 per 100,000 reproductive-aged women and these estimates are thought to be higher in view of considerable number of asymptomatic undiagnosed women [6]. EH is more common in women presenting with infertility, particularly those of advanced maternal age, and often coexists with polycystic ovary syndrome (PCOS), with a reported prevalence of 23%–36% [789].

Most clinical guidelines recommend progestins as a first-line treatment option for non-atypical EH and abstaining from pregnancy until normal endometrium is confirmed on histopathology [101112]. The disadvantages of progestin therapy include systemic side effects or vaginal bleeding, which are more pronounced with high-dose oral preparations. In addition, progestins can cause thinning of the endometrium and may require considerable time for endometrial recovery after treatment, which may not be acceptable to women who desire immediate pregnancy. There is, however, relative paucity of information about the effect of EH on fertility outcomes and most data are derived from small retrospective analyses in women with atypical EH or early endometrial cancer [131415].

Since first introduced in early 1960's, combined estrogen-progestin oral contraceptive pills (OCPs) have been used by millions of women worldwide, providing effective contraception and other non-contraception benefits [1617]. It has been demonstrated that OCPs reduce risk of endometrial cancer with long-lasting post treatment effect [1819]. Continuous estrogen-progestin replacement in postmenopausal women has been associated with complete regression of pre-existing complex EH and reduced long-term risk of endometrial cancer [2021]. Serious risks of OCP, such as thromboembolic events, stroke or gallbladder disease are infrequent among healthy women and common side effects are often self-limiting [22]. Thus, it is reasonable to infer that OCP can be effective management option for EH in young women, however, there are no clinical evaluations to support this assumption.

In our institution, OCPs have been initially introduced for managing EH in women who did not respond to progestins and became increasingly utilized as first-line choice due to favorable response. In this study, we aimed to summarize our experience and to evaluate the efficacy of short-term OCP versus oral progestin therapy for non-atypical EH in reproductive-aged women with infertility. We also assessed the reproductive outcomes following successful reversal of EH with either treatment approach.

Overall, 309 patients met the inclusion criteria: 216 women were treated with OCP and 93 women received progestin. Endometrial biopsy at initial diagnosis was obtained at hysteroscopy (233 women) or following in-office endometrial sampling (76 women). The baseline characteristics of the study population are presented in Table 1. Women in OCP group were younger than women in progestin group (31.12±4.45 vs. 33.17±4.29; p<0.001), but this difference was of marginal clinical significance. More women in OCP group were diagnosed with PCOS (43.06% vs. 30.11%; p=0.032) and had other uterine pathology, such as uterine septum, leiomyoma, adenomyosis and unicornuate uterus (55.56% vs. 43.01%; p=0.043). Duration of infertility was shorter in OCP than in progestin group (median 5, IQR 2–7 years vs. 6, IQR 3–8 years, respectively; p=0.036). Other demographic and clinical characteristics were comparable between the groups. Most participants were nulliparous and hormone treatment-naive. None of the participants reported serious side effects associated with hormone treatments such as venous thromboembolism, stroke or gallbladder disease.

The duration of treatment was similar between OCP and progestin groups (median 3, IQR 2–3 months vs. 3, IQR 3–3 months, respectively; p=0.063). Upon completion of the treatment course, 200/309 (64.73%) women underwent hysteroscopy and 109/309 (35.28%) women had in-office endometrial biopsy. Complete remission was confirmed in 282/309 (91.26%) women. Response to treatment is detailed in Table 2. Women treated with OCP were more likely to achieve complete remission than those treated with progestin (202/216, 93.52% vs. 80/93, 86.02%, respectively; p=0.032; aOR=2.35; 95% confidence interval [CI]=1.06–5.21). The rate of AUB was significantly lower in OCP (16/216, 7.41%) compared to progestin group (17/93, 18.30%); p=0.005. Cyclic OCP was associated with significantly higher response to treatment (175/184, 95.11%) than continuous administration of active pill (27/32, 84.38%); p=0.039; aOR 3.60; 95% CI=1.12–11.55. Both Marvelon and Diane-35 led to a comparable reversal rate of EH and had similar bleeding profile. Separate comparison between the three most used progestin regimes, namely dydrogesterone (46 women, 49.46%), medroxyprogesterone (28 women, 30.11%) and megestrol acetate (12 women, 12.90%) did not reveal significant difference in response rate (Table 3), although the sample size precluded meaningful conclusions.

The subgroup analyses revealed that non-obese women with BMI <27.5 kg/m² (a threshold for obesity in Asian women [24]) and women who were hormone treatment-naive had significantly higher remission rate by OCP than by progestin (Table 4). In contrast, OCP did not provide therapeutic benefit over progestin when women were stratified by age, in obese individuals (BMI ≥27.5 kg/m²), in women with PCOS or glucose intolerance and when gross findings of EH were observed at hysteroscopy (p>0.05).

No study participant was found with progression of non-atypical hyperplasia to more severe forms. Persistent disease was observed in 27 patients, of which 21 accepted an additional course of hormone therapy with either progestin (n=15), OCP (n=4), or combination of both (n=2). Among these patients, regression of EH was observed in 20 women and one woman was lost to follow-up. Six women declined second course of therapy: one woman opted for hysterectomy and five attempted pregnancy.

Overall, 298 study participants ultimately tried to conceive, of whom 114 (38.26%) achieved pregnancy: 19/298 (6.38%) women conceived spontaneously and in 95/298 (31.88%) women pregnancies resulted from assisted conception (seven following ovulation induction and 88 following in vitro fertilization-intracytoplasmic sperm injection [IVF-ICSI]). Eleven patients changed their fertility plans and decided not to pursue fertility any longer.

The reproductive outcomes for each treatment group are presented in Table 5. In OCP group, there were 87 cases of clinical pregnancy, 22 miscarriages (13 early and nine late) and 65 live births. In progestin group, there were 27 cases of clinical pregnancy, eight miscarriages (seven early and one late) and 19 live births. Most pregnancies were achieved through fertility treatments in either treatment group. The rate of spontaneous pregnancy was comparable between the groups: 11/208 (5.29%) in women from OCP and 8/90 (8.89%) in women from progestin group (p=0.243). CPR was higher in OCP than in progestin group (41.83% vs. 30.00%, respectively), but the difference only reached borderline significance (p=0.054). No significant difference was demonstrated after adjustment for age, PCOS, presence of uterine pathology and duration of infertility (data not shown). Sensitivity analysis performed with excluding the patients lost to follow-up or counting them as pregnant did not change the obtained results. MR was somehow lower in OCP than in progestin group (25.29% vs. 29.62%, respectively), although failed to achieve statistical significance (p=0.654). LBR was 31.25% in OCP and 21.11% in progestin group and was not significantly different between the groups (p=0.074).

To the best of our knowledge, this is the first study that evaluates use of OCP as first-line treatment option for EH and is the largest report on fertility outcomes in women with non-atypical EH. Our results demonstrate that short course of OCP is well tolerated and associated with good results. Remission of EH by OCP (93.52%) was significantly higher than by progestin (86.02%) after the initial course of treatment for two to six months. After adjustment for important demographic and clinical confounders, the odds of achieving normal endometrial histology were 2.35 higher than after similar length of treatment with progestin. Non-obese and hormone treatment-naive women had more favorable response to OCP than to progestin. We also showed improved bleeding profile with OCP treatment, which has important implication for patient compliance and satisfaction profile.

The mechanism by which OCP lead to reversal of EH is not entirely clear. It is supposedly related to the reduced exposure to unopposed estrogen and inhibition of estrogen-induced endometrial proliferation [25]. While OCP contain both an estrogen and a progestin component, there is an essentially progestin-dominant suppressive effect on the endometrium mediated via the progesterone receptor [26]. Full progestogenic effect has been observed on endometrium of healthy women on OCP [27]. Similarly, short-term treatment with OCP resulted in endometrial changes consistent with progestin effect in women with Lynch syndrome who are genetically predisposed for developing early-onset endometrial cancer [28]. Endometrial changes included prominent inhibition of proliferation (Ki67), significant decrease in expression of EIG121, IGF-1, sFRP-1, and sFRP4 implicated in endometrial proliferation signaling pathways, and downregulation of apoptosis inhibitor survivin. The authors proposed a molecular mechanism for hormone suppression of endometrial proliferation and showed a concordance between endometrial effect of progesterone and OCP, supporting progestin-dominant effect of the combined pill [28].

The effect of estrogen-progestin treatment on EH has been predominantly evaluated in postmenopausal women on hormone replacement therapy (HRT) who were followed-up with a serial histological assessment of endometrium. Continuous administration of combined HRT resulted in 64%–100% reversal of EH to normal endometrial pattern after at least three months of exposure and has not been associated with new cases of EH during five years of treatment [2029]. Anecdotal evidence is available on managing EH with OCP. Combined OCP-Metformin regime has been shown to cause remission of complex progestogen-resistant EH in two obese insulin-resistant individuals after three months of treatment, but this effect was partially attributed to metformin and there are no evaluations in large patient groups [30].

Indirect evidence on plausible effect of OCP on endometrium is derived from the epidemiological studies. Large population studies consistently report protective effect of OCP on endometrial cancer risk with 40%–60% reduced risk of developing endometrial cancer in ever-users of OCP compared with never-users [181931]. There is an apparent association between risk reduction and the duration of OCP treatment, an effect that persisted for more than 44 years [1819]. The risk-reduction effect of OCP does not seem to depend on personal characteristics such as parity, adiposity or menopausal status or on the dose of estrogen [1932]. There was no difference in risk reduction for those who used higher estrogen content OCP in the 1960's–1970's [1933].

In this study, we observed significantly higher response to cyclic OCP regime compared to continuous administration of active pill. This can be explained by the fact that effective sloughing off the endometrial lining during hormone-free interval may play a plausible role in therapeutic effect of OCP.

Our finding of 86.02% EH remission by oral progestin is in line with previous reports. A meta-analysis of 24 observational studies comprising 1,001 women reported that treatment with oral progestins resulted in 66%–89% regression of non-atypical hyperplasia [34]. In this study women received different types of progestins for varying periods of time as there is no consensus on the preferred progestin regime for managing EH with respect to type of progestin, dose and duration of treatment [11]. We did not observe any difference in response to two types of OCP that contained different types of progestin and showed comparable regression rate when the three progestin regimes were compared with each other. No difference between three different types of oral progestins in reversing simple non-atypical EH has been demonstrated in small randomized study, although direct comparative data in larger cohorts are lacking [35].

We demonstrated that among women who attempted conception, 38.3% (114/298) achieved clinical pregnancy and 28.2% (84/298) had live birth. Most pregnancies resulted from fertility treatments, although 6.38% of pregnancies in our cohort of infertile women were conceived naturally. There was no significant difference in fertility outcomes between women treated with OCP or with progestin, but we observed a trend towards higher pregnancy rate following OCP. It has been proposed that morphometric endometrial changes, altered intra-uterine inflammatory cascade and deranged molecular pathways observed in EH can affect endometrial receptivity, which can lead to implantation defects and decreased chance for pregnancy [3637]. In PCOS women with simple non-atypical hyperplasia, clinical pregnancy following IVF was 28%–46% with improved pregnancy rates and decreased odds of miscarriage after complete regression to normal endometrium [38]. Several systematic reviews reported 18%–40% pregnancy rate and 14%–35% LBR after complete remission of atypical complex EH or endometrial cancer [131415]. Overall, weighing our findings relative to existent literature is difficult, considering that most research focused on women with more advanced disease and unclear baseline fertility status. It has to be acknowledged that repeat endometrial curettage can increase risk of endometrial adhesions and affect fertility, therefore meticulous attention to endometrial sampling technique to minimize the endometrial trauma is crucial in women willing to retain fertility.

The importance of this study is that for the first time it provides the data on efficacy of OCP as first-line treatment option for EH in a relatively large cohort of 309 young women. We stratified the patients according to the most recent 2014 WHO classification and focused on low-risk non-atypical EH, the most common form of EH. Hence, our data are of relevance for majority of reproductive-aged women diagnosed with this condition. The presented information is important for patient counselling and can be useful for future comparisons of treatment endpoints as new studies using an updated WHO classification emerge. Despite compelling evidence from large epidemiological studies on endometrial protective effect of OCPs, they are rarely prescribed for prevention of endometrial cancer and are not used for managing EH. In addition to their potential therapeutic value in EH, OCP provide effective contraception and present a wide range of non-contraceptive benefits [17] and their role in managing young women with hyperplasia deserves exploration. Furthermore, expanding the range of effective treatment options allows to tailor management to patients' needs and offers personalized approach to patient care. We believe that our report will encourage further research efforts to evaluate the efficacy and safety of OCP for this indication.

The main limitations of this study are retrospective design, single-centre experience, arbitrary allocation of patients to the treatment groups and not uniform length of treatment. In addition, our results may be confounded by relatively short follow-up and inability to estimate relapse rate of EH. In postmenopausal women with complex non-atypical EH relapse was observed in 28.3% following regression by oral progestin at a median of 13.7 months [39], however the risk of relapse in young low-risk patients is less clear and should be considered in future studies. In addition, while none of the study participants reported serious side effects associated with hormone treatment, we were not able to evaluate the magnitude and prevalence of other side effects including gastro-intestinal symptoms, breast tenderness, mood changes, headaches, weight gain or decreased libido. Each of these symptoms can influence quality of life of women treated for EH and negatively affect their compliance with treatment, hence should be prospectively evaluated in future studies.

Taken together, our data suggest that OCP are effective in managing non-atypical EH in young women and are associated with favorable fertility prognosis. The findings of this study require confirmation in larger cohorts in prospective randomized setting with longer follow-up of gynecological and reproductive outcomes. The questions whether OCP management is associated with shorter time to pregnancy, and what is the optimal type of OCP and duration of treatment remain unanswered. It is important to note that OCP carry risks of adverse effects, thereby may be poorly tolerated or contraindicated for some patients. Treatment decision-making should include consideration of risks, benefits and patient preferences. Considering lack of observational data on OCP for managing non-atypical EH, we believe that this work will provide valuable evidence in the area where there is little information.