The decline of breast cancer mortality in the western world can be partly attributed to the use of tamoxifen, which exerts an anti-estrogenic effect on the mammary tissue. A 5-year course of tamoxifen treatment has been shown to reduce the annual recurrence rate by almost half and the mortality rate by one-third.1
The incidences of endometrial polyps and hyperplasia had been reported to be 5–35% and 4.7–16%, respectively, among tamoxifen users.2,3 Furthermore, tamoxifen also doubles the risk of endometrial cancer.4 Nevertheless, the benefit of tamoxifen still outweighs its risk for women with breast cancer. Furthermore, regular endometrial surveillance, using hysteroscopy, of tamoxifen users appears to be neither cost-effective nor likely to reduce mortality from endometrial cancer.2
Levonorgestrel-releasing intrauterine system is approved for endometrial protection in women with a uterus who are receiving estrogen replacement therapy, and it was shown to induce regression of endometrial hyperplasia.5,6
However, its value in prophylactic use in women with breast cancer treated with tamoxifen is still uncertain. We previously reported an interim result of a randomized controlled trial confined to women who had completed 1 year of follow-up; women in the treatment group had a much lower incidence of endometrial polyps at 12-month follow-up.7 So far, there have been only two randomized controlled trials of the endometrial protective effects of levonorgestrel-releasing intrauterine system in women with breast cancer using adjuvant tamoxifen,8 but the follow-up was an average of 2 years at most.9 The objective of the present study was to estimate the long-term effect of levonorgestrel-releasing intrauterine system on the endometrium in tamoxifen users.
MATERIALS AND METHODS
The present article is the final report of the randomized controlled trial described.7 All participants were recruited from the gynecology clinic of the Prince of Wales Hospital. During the study period, Chinese women who had early-stage breast cancer requiring postoperative adjuvant tamoxifen therapy were referred from the oncology clinic of the same hospital on completion of adjuvant chemotherapy and radiotherapy, before the commencement of tamoxifen treatment.
Before their enrollment into the study, written consent was signed by the participants at the clinic after the provision of an information sheet and explanation of the study details by a research assistant. Women were then randomized into either the levonorgestrel-releasing intrauterine system treatment group or the control group according to a computer-generated random number in blocks of four placed in serially numbered sealed envelopes.
Exclusion criteria were contraindications for an intrauterine device such as pelvic inflammatory disease, congenital uterine anomaly, or uterine cavity length more than 10 cm measured on a transvaginal ultrasound examination.
At baseline, transvaginal ultrasound examination was performed to assess the uterine size, endometrial thickness, and any uterine and adnexal pathology. Endometrial thickness was measured at the widest point across the hyperechoeic area in a longitudinal section of the uterus along the sagittal plane.
Outpatient hysteroscopy was performed within 2 weeks after the transvaginal ultrasound examination using a 5-mm rigid hysteroscope using normal saline as distending medium. Guided endometrial biopsy was performed if there was a suspicious lesion, whereas endometrial sampling was performed universally through an aspirator after the hysteroscopy.
Any endometrial polyps and symptomatic submucosal fibroids detected were removed hysteroscopically under general anesthesia before the commencement of tamoxifen.7 Women randomized to the treatment group had levonorgestrel-releasing intrauterine system inserted immediately after the hysteroscopy at the outpatient clinic or after the hysteroscopic resection of endometrial pathology in the operating theater.
All participants were assessed at 12, 24, 45, and 60 months with transvaginal ultrasound examination, outpatient hysteroscopy, and aspiration. Symptoms of dysmenorrhea, pelvic pain, mastalgia, acne, and vaginal dryness were recorded through a standard questionnaire at each visit. All de novo endometrial polyps were removed during outpatient hysteroscopy (small size with diameter of 0.5 cm or less) or in the operating theater under general anesthesia (polyps of larger size or failure to remove in outpatient setting). All specimens were fixed in formalin for hematoxylin and eosin staining and reviewed by a histopathologist (M.-Y.Y.) who was blinded to the randomization.
Women were advised to observe and report any pelvic pain or any abnormal discharge after the procedures. A menstrual calendar was used to record any abnormal vaginal bleeding after levonorgestrel-releasing intrauterine system insertion. Ethics approval for the study was granted by the Institutional Ethics Committee (CRE-2001.053; CRE-2006.460).
Assuming a drop-out rate of 20%, 63 women were required in each group to detect reduction in the rate of endometrial pathology from 50% to 30% in women treated with levonorgestrel-releasing intrauterine system at 5 years after the tamoxifen treatment with a power of 80% and type I error of 5%.
All data are expressed as mean±standard deviation or proportion. Between-group differences were compared by using χ2 or Fisher exact tests for categorical variables, and by Student t test for continuous variables, as appropriate, on an intention-to-treat basis, except when otherwise specified. The endometrial thickness measured by transvaginal ultrasound examination was compared between the two groups at different periods of assessment by using repeated measures analysis of variance, whereas other categorical data, measured over multiple time points, were compared by Mantel-Haenszel test with or without continuity correction, as appropriate.10,11 The cumulative rates of de novo endometrial polyps also were compared by using the Kaplan-Meier survival analysis and log-rank test by GraphPad Prism 5.0. All other statistical analysis was performed by using PASW Statistics 18 software. P<.05 for two-tailed statistical tests was used to indicate significance.
A total of 145 women were recruited and 129 were randomized between July 2002 and April 2005. After randomization, 65 in the control group and 64 in the treatment group completed baseline assessment. Only those who completed baseline assessment were included in the analysis. In the end, 48 in the control group and 46 in the treatment group completed the 5-year follow-up assessment (Fig. 1). There were no statistically significant differences in the anthropometric parameters, stage of breast cancer, adjuvant therapy, or endometrial pathology at the baseline assessment (Table 1). Endometrial polyps occurred less commonly in the treatment group (4.3% compared with 32.7%; P<.001); three women had recurrence of endometrial polyps after a polypectomy. There was a significantly higher rate of abnormal bleeding among the women treated with levonorgestrel-releasing intrauterine system (27 of 58 [46.6%] compared with 5 of 60 [8.3%]; P<.001). However, the rate of vaginal spotting in the treatment group was not significantly higher than that of the control group by 24 months (6 of 55 [11%] compared with 3 of 57 [5.26%]; P=.45). No women reported abnormal vaginal bleeding beyond 24 months in either group. Breast tenderness was reported in two women in the treatment group in the first year of treatment, but the symptom subsided spontaneously thereafter. Only one woman in the treatment group reported mild acne.
Levonorgestrel-releasing intrauterine system was removed from five women (8.6%). One woman (1.6%) had development of acute pelvic inflammatory disease soon after the insertion, so it was removed 3 days after insertion. Levonorgestrel-releasing intrauterine system was not tolerated by three women (4.9%), who subsequently requested removal because of vaginal spotting at 1, 2, and 24 months, respectively; another woman (1.6%) requested removal at 24 months because she declined further participation.
The clinical conditions of participants who dropped -out from the study were retrieved through the computerized medical record system; all women received regular follow-up by the Department of Oncology at the Prince of Wales Hospital. There was no significant difference in the recurrence or mortality rate of breast cancer between the treatment and control groups (Table 2). There was no endometrial cancer in the study group; however, one woman who was initially randomized into the control group but dropped-out soon after the baseline assessment had endometrial hyperplasia diagnosed after 4 years of tamoxifen treatment.
There was no significant difference between the two groups in the endometrial thickness measured by transvaginal ultrasound examination. There remained no difference after categorization into those who were premenopausal and postmenopausal before the adjuvant therapy (Table 3).
Table 4 shows the endometrial histology obtained from aspiration. Women who were premenopausal before chemotherapy had a significantly higher rate of inactive endometrium when treated with levonorgestrel-releasing intrauterine system, whereas proliferative endometrium was not detected beyond 12 months in the treatment group.
There were, in total, 30 hysteroscopic endometrial polypectomy procedures during the study. Six (20%) endometrial polyps were removed in the operating theater under general anesthesia and the rest were small polyps removed in an outpatient setting. All polyps were confirmed histologically by a histopathologist (M.-Y.Y.) who was blinded to the grouping. Table 5 shows the rate of de novo endometrial polyps in both groups. There were only two women in the treatment group who had development of endometrial polyps at 12 and 45 months, respectively. One was a postmenopausal woman who had the levonorgestrel-releasing intrauterine system removed at 2 months after insertion because of vaginal spotting, so only one premenopausal woman (1.7%) had development of a small endometrial polyp with levonorgestrel-releasing intrauterine system in situ. The cumulative rate of de novo endometrial polyps is significantly lower in the treatment arm (hazard ratio 0.19, 95% confidence interval 0.07–0.48) (Fig. 2). The majority of these were asymptomatic; only three women (10%) presented with abnormal vaginal bleeding.
There also was no significant difference between the two groups in the occurrence of submucosal fibroids, which were detected at only 12 months follow-up in one woman in the treatment group (1.8%) and two women (3.4%) in the control group. All were resected and there was no de novo submucosal fibroid after the first year.
The present study is the final report of a 5-year randomized controlled trial of the prophylactic use of levonorgestrel-releasing intrauterine system for endometrial protection in tamoxifen users. The majority of women in this study accepted levonorgestrel-releasing intrauterine system as prophylactic treatment and were able to continue for 5 years throughout their tamoxifen treatment. The acceptance rate of levonorgestrel-releasing intrauterine system in this context was higher than that used therapeutically for menorrhagia in a previous Chinese population.12
Similar to the findings from previous studies,7,9 the present work shows that levonorgestrel-releasing intrauterine system is effective in preventing endometrial polyps in tamoxifen users over the course of 5 years. Although all lesions were benign, one-third of the women eventually had development of endometrial polyps in the control group. Because all de novo endometrial polyps were removed as soon as they were diagnosed, the study design could have prevented their progression into endometrial hyperplasia and endometrial cancer with time. Garuti et al13 have reported that up to 36% and 33% of the polyps could have hyperplasia and cancer, respectively. Therefore, all endometrial polyps warrant removal in tamoxifen-treated women despite being asymptomatic. The recent Cochrane Systematic Review also concluded that levonorgestrel-releasing intrauterine system significantly reduced the incidence of endometrial polyps in breast cancer in women treated with tamoxifen, but it also stated that there was no clear evidence that the levonorgestrel-releasing intrauterine system would prevent endometrial hyperplasia or adenocarcinoma in these patients.8 Our findings are also consistent with the literature in that, among women who use tamoxifen, endometrial polyps and atropic endometrium are the most common findings in hysteroscopic and histopathologic.14,15
Studies of the levonorgestrel-releasing intrauterine system have shown that a high dose of local progestogen induces endometrial epithelial atrophy, decidualization, and vascular change such that the endometrium becomes unresponsive to ovarian steroids.16,17 This may be the mechanism by which local progestogen prevents endometrial stimulation and polyp formation in tamoxifen-treated women. The recently identified G-protein-coupled estrogen receptor (formerly known as GPR30), which may assume a physiologic role in many parts of the body,18 also was found to have an important role in endometrial pathology after tamoxifen for breast cancer in both in vivo and in vitro studies.19 Tamoxifen exposure also was found to be associated with an overexpression of β-catenin oncoprotein, which may play a major role in the pathogenesis of endometrial adenocarcinoma.20
Because progestogen has been shown to increase mammary cell proliferation in both premenopausal and postmenopausal women, and because a low serum level of progestogen can be detected in levonorgestrel-releasing intrauterine system users,21,22 its effect on the risk of breast cancer recurrence is a matter of concern.8
A retrospective cohort study of women with breast cancer who continued to use levonorgestrel-releasing intrauterine system suggested a borderline increase in the risk of cancer recurrence.23 However, a large epidemiology study in a general population from Finland showed no such increase.24 Although the present study did not show a statistically significant increase in the breast cancer recurrence rate, this finding has to be interpreted with caution because the sample size was not of sufficient power to address this outcome. A post hoc analysis from our results suggests that at least 382 and 3,310 women are required in each group to detect, respectively, a genuine increase in the breast cancer recurrence rate and breast cancer--related deaths at a power of 80% and type I error of 5% with the prophylactic use of levonorgestrel-releasing intrauterine system for 5 years in women treated with tamoxifen.
Although the third-generation aromatase inhibitors that result in much fewer uterine pathologies compared with tamoxifen and that may even reverse tamoxifen-induced endometrial pathology are now part of the standard adjuvant endocrine therapy for postmenopausal women with endocrine-responsive breast cancer,25–28 tamoxifen remains part of the standard adjuvant endocrine therapy for premenopausal and postmenopausal breast cancer patients, and further clarification of the effect of progestogen on breast cancer is required.
An endometrial thickness of 5 mm has been generally used as a cut-off for further evaluation by endometrial sampling in postmenopausal women. Tamoxifen, however, induces subepithelial stromal hypertrophy, simulating endometrial hyperplasia at ultrasound examination, whereas atrophy is often the only finding at invasive assessment. A study of tamoxifen-treated women has shown that endometrial thickness more than 3 mm could allow a 100% sensitivity but could only reduce 25% of the invasive assessments.29 Our results did not show any difference in the measurement of endometrial thickness between the two groups, despite a significantly higher rate of endometrial polyp among the control group.
The present study showed that levonorgestrel-releasing intrauterine system significantly reduces the occurrence of endometrial polyp in women using tamoxifen throughout the 5 years of treatment. However, its role in the prevention of endometrial hyperplasia and adenocarcinoma as well as its effect on risk of breast cancer recurrence remain uncertain. Given the fact that most of the endometrial polyps were asymptomatic and of small size, the potential increase in breast cancer recurrence risk does not justify its clinical use in this population until there is further supportive evidence from a larger study.
1. Early Breast Cancer Trialists' Collaborative Group. Effects of chemotherapy and hormonal therapy for early breast cancer on recurrence and 15-year survival: an overview of the randomised trials. Lancet 2005;365:1687–717.
2. Neven P, De Muylder X, Van Belle Y, Van Hooff I, Vanderick G. Longitudinal hysteroscopic follow-up during tamoxifen treatment. Lancet 1998;351:36.
3. Ozsener S, Ozaran A, Itil I, Dikmen Y. Endometrial pathology of 104 postmenopausal breast cancer patients treated with tamoxifen. Eur J Gynaecol Oncol 1998;19:580–3.
4. van Leeuwen FE, Benraadt J, Coebergh JW, Kiemeney LA, Gimbrere CH, Otter R, et al.. Risk of endometrial cancer after tamoxifen treatment of breast cancer. Lancet 1994;343:448–52.
5. Varma R, Soneja H, Bhatia K, Ganesan R, Rollason T, Clark TJ, et al.. The effectiveness of a levonorgestrel-releasing intrauterine system (LNG-IUS) in the treatment of endometrial hyperplasia—a long-term follow-up study. Eur J Obstet Gynecol Reprod Biol 2008;139:169–75.
6. Wan YL, Holland C. The efficacy of levonorgestrel intrauterine systems for endometrial protection: a systematic review. Climacteric 2011;14:622–32.
7. Chan SS, Tam WH, Yeo W, Yu MM, Ng DP, Wong AW, et al.. A randomised controlled trial of prophylactic levonorgestrel intrauterine system in tamoxifen-treated women. BJOG 2007;114:1510–5.
8. Chin J, Konje JC, Hickey M. Levonorgestrel intrauterine system for endometrial protection in women with breast cancer on adjuvant tamoxifen. The Cochrane Database of Systematic Reviews 2009;4:CD007245. DOI: 10.1002/14651858.CD007245.pub2.
9. Gardner FJ, Konje JC, Bell SC, Abrams KR, Brown LJ, Taylor DJ, et al.. Prevention of tamoxifen induced endometrial polyps using a levonorgestrel releasing intrauterine system long-term follow-up of a randomised control trial. Gynecol Oncol 2009;114:452–6.
10. Tam WH, Yuen PM, Shan Ng DP, Leung PL, Lok IH, Rogers MS. Health status function after treatment with thermal balloon endometrial ablation and levonorgestrel intrauterine system for idiopathic menorrhagia: a randomized study. Gynecol Obstet Invest 2006;62:84–8.
11. Mantel N, Mocarelli P, Marocchi A, Brambilla P, Baretta R. Stratified analysis of multivariate clinical data: application of a Mantel-Haenszel approach. Statist Med 1983;2:259–66.
12. Begg MD, Panageas KS. Interval estimation of the common odds ratio from k(2×2) tables under cluster sampling. Statist Med 1999;18:1087–100.
13. Garuti G, Grossi F, Cellani F, Centinaio G, Colonnelli M, Luerti M. Hysteroscopic assessment of menopausal breast-cancer patients taking tamoxifen; there is a bias from the mode of endometrial sampling in estimating endometrial morbidity? Breast Cancer Res Treat 2002;72:245–53.
14. Dibi RP, Zettler CG, Pessini SA, Ayub AV, de Almeida SB, da Silveira GP. Tamoxifen use and endometrial lesions: hysteroscopic, histological, and immunohistochemical findings in postmenopausal women with breast cancer. Menopause 2009;16:293–300.
15. Zarbo G, Caruso G, Zammitti M, Caruso S, Zarbo R. The effects of tamoxifen therapy on the endometrium. Eur J Gynaecol Oncol 2000;21:86–8.
16. Pakarinen P, Toivonen J, Luukkainen T. Therapeutic use of the LNG IUS, and counseling. Semin Reprod Med 2001;19:365–72.
17. Luukkainen T, Pakarinen P, Toivonen J. Progestin-releasing intrauterine systems. Semin Reprod Med 2001;19:355–63.
18. Prossnitz ER, Barton M. The G-protein-coupled estrogen receptor GPER in health and disease. Nat Rev Endocrinol 2011;7:715–26.
19. Ignatov T, Eggemann H, Semczuk A, Smith B, Bischoff J, Roessner A, et al.. Role of GPR30 in endometrial pathology after tamoxifen for breast cancer. Am J Obstet Gynecol 2010;203:595 e9–16.
20. Turbiner J, Moreno-Bueno G, Dahiya S, Sanchez-Estevez C, Hardisson D, Prat J, et al.. Clinicopathological and molecular analysis of endometrial carcinoma associated with tamoxifen. Mod Pathol 2008;21:925–36.
21. Aupperlee M, Kariagina A, Osuch J, Haslam SZ. Progestins and breast cancer. Breast Dis 2005;24:37–57.
22. Raudaskoski T, Tapanainen J, Tomas E, Luotola H, Pekonen F, Ronni-Sivula H, et al.. Intrauterine 10 microg and 20 microg levonorgestrel systems in postmenopausal women receiving oral oestrogen replacement therapy: clinical, endometrial and metabolic response. BJOG 2002;109:136–44.
23. Trinh XB, Tjalma WA, Makar AP, Buytaert G, Weyler J, van Dam PA. Use of the levonorgestrel-releasing intrauterine system in breast cancer patients. Fertil Steril 2008;90:17–22.
24. Backman T, Rauramo I, Jaakkola K, Inki P, Vaahtera K, Launonen A, et al.. Use of the levonorgestrel-releasing intrauterine system and breast cancer. Obstet Gynecol 2005;106:813–7.
25. Morales L, Timmerman D, Neven P, Konstantinovic ML, Carbonez A, Van Huffel S, et al.. Third generation aromatase inhibitors may prevent endometrial growth and reverse tamoxifen-induced uterine changes in postmenopausal breast cancer patients. Ann Oncol 2005;16:70–4.
26. Bertelli G, Hall E, Ireland E, Snowdon CF, Jassem J, Drosik K, et al.. Long-term endometrial effects in postmenopausal women with early breast cancer participating in the Intergroup Exemestane Study (IES)—a randomised controlled trial of exemestane versus continued tamoxifen after 2-3 years tamoxifen. Ann Oncol 2010;21:498–505.
27. Josefsson ML, Leinster SJ. Aromatase inhibitors versus tamoxifen as adjuvant hormonal therapy for oestrogen sensitive early breast cancer in post-menopausal women: meta-analyses of monotherapy, sequenced therapy and extended therapy. Breast 2010;19:76–83.
28. van de Velde CJ, Rea D, Seynaeve C, Putter H, Hasenburg A, Vannetzel JM, et al.. Adjuvant tamoxifen and exemestane in early breast cancer (TEAM): a randomised phase 3 trial. Lancet 2011;377:321–31.
29. Ciatto S, Cecchini S, Gervasi G, Landini A, Zappa M, Crocetti E. Surveillance for endometrial cancer with transvaginal ultrasonography of breast cancer patients under tamoxifen treatment. Br J Cancer 2003;88:1175–9.