Levonorgestrel Intrauterine System in Adjuvant Tamoxifen Treatment

Introduction

The breast cancer is the most common cancer worldwide and the second cause of cancer death in the female population. ¹ Breast mortality rate seems to decrease in the last decades in developed countries due to the increased rate of screening and advances in adjuvant treatment. ² Randomized clinical trials demonstrated survival benefits associated with the use of adjuvant therapies, with estimated reductions in the annual odds of death ranging from 8% to 28%, depending on the type and duration of therapy, age of patients, and tumor characteristics. ³

The American Society of Clinical Oncology (ASCO) Clinical Practice Guidelines Committee (CPGC) and the College of American Pathologists (CAP) Council on Scientific Affairs (CSA) drafted an expert panel recommending that estrogen receptor (ER), progesterone receptor (PgR), and Human Epidermal Growth Factor Receptor 2 (HER2) receptor status should be determined in all case of invasive breast cancers and breast cancer recurrences. Hormonal receptor investigation is useful in selecting patients who can benefit of adjuvant hormonal therapy after surgical treatment. ^3,4

In 75% to 80% of patients with early ER-positive breast cancer, 5 years of treatment with tamoxifen (TAM) substantially reduces local, contralateral, and distant recurrence rates and reduces 15-year breast cancer mortality. So tumors with high or uncertain hormone responsiveness (ER >1%) should be treated with endocrine therapy. ⁵ Despite its benefits, many previous studies reported that TAM treatment is associated with various endometrial pathologies, ranging from polyps to endometrial cancer. ⁶

Nowadays, it is universally accepted that TAM-treated women should be closely monitored for endometrial hyperplasia or cancer, but there is no consensus on the most proper management. ⁷

Many authors proposed that the endometrial estrogen-like effect of TAM could be balanced by Pg therapy, either orally (noretisterone, megestrol, and medroxyprogesterone acetate) or locally (levonorgestrel intrauterine system [LNG-IUS]). ^6,7

The LNG-IUS, often referred to by its brand name (Mirena; Bayer Schering Pharma-AG, Berlin, Germany) releases locally constant continuous dose of Pg over 5 years with endometrial concentrations of up to 100 times higher than that of oral tablet containing 30 mg of LNG, with minimal systemic absorption. ⁸

Despite LNG-IUS has been used in patients with breast cancer treated with TAM as a way of preventing endometrial proliferation, nowadays it does not define the relation between benefits (on endometrial tissue) and side effects (on breast tissue) and the cohort of patients most suitable to undergo such treatment.

Aim of this systematic review is to define breast risks and endometrial benefits in TAM-treated women using Mirena, and the second aim is to define the cohort of patients who could benefit from the LNG-IUS use.

Data Sources

An English literature research was conducted in the electronic databases MEDLINE, EMBASE Sciencedirect, and the Cochrane Library in the last two decades. The search terms used included “breast cancer,” “hormone receptors,” “progesterone,” “adjuvant therapy,” “LNG-IUS,” “Mirena,” “tamoxifen,” “endometrial hyperplasia,” “endometrial carcinoma,” “endometrial protection.”

Population considered was breast cancer survivor treated with TAM. Intervention considered was Pg treatment through LNG-IUS. Studies were selected if the participants had an histological or cytological evaluation of endometrial status previous and after treatment, for studies on LNG-IUS endometrial effects; or certain diagnosis of breast cancer recurrence for studies on LNG-IUS breast effects.

Original descriptions, retrospective evaluations, and review articles were analyzed. A manual search of reference lists of included studies and review articles was also performed. References from the articles were searched to identify any articles initially excluded.

Eligible articles included descriptions of population study, hormone treatment, side effects, complications, and oncologic outcomes. We excluded studies that provided ambiguous or insufficient data.

Studies were selected in a 2-stage process. First, titles and abstracts from electronic searches were scrutinized by 2 reviewers independently and full manuscripts of all citations that joined predefined selection criteria were obtained. Second, inclusion and exclusion criteria were performed on the full manuscript examination.

Available Methods

Tamoxifen: Breast and Endometrial Mechanism of Action

Tamoxifen is a selective ER modulator (SERM). It requires metabolic activation by cytochrome P450 enzymes producing active metabolites with 30- to 100-fold greater ER affinity. After hepatic phase I and phase II metabolism, TAM is excreted predominantly in the feces. ^9,10

Serum TAM concentrations vary widely among patients. In long-term treatment, serum steady state concentrations of TAM and its metabolites remain constant for 10 years. ¹¹

Tamoxifen, as all SERMs, acts as an estrogen (E2) agonist or antagonist in different tissues. This feature is related to specific actions on at least 2 distinct ERs, whose proportions differ depending on the tissue type. ³

The ER is expressed as 2 different isoforms: ER-α (predominantly activating) and ER-β (predominantly inhibiting). The ER-α is expressed more in vagina, uterus, and ovary, in the breast, hypothalamus, and in the endothelial cells; while ER-β is expressed more in prostate and ovary. ¹²

The ER-β inhibits the action of ER-α by forming an heterodimer. On this basis, expression of the relative levels of the 2 isoforms will affect the cellular and tissue responsiveness to E2s. ^1,12

Most of the available studies on TAM effects point out to its antiestrogenic activity on breast tissue in opposition to the estrogenic activity on endometrium. In breast tissue, TAM inhibits both tumor growth, by competitive inhibition of E2 binding to ERs, and expression of E2-regulated genes, including growth and angiogenic factors secreted by the tumor. The result is a G1-phase block of cell cycle and a slowing of cell proliferation: tumors may regress because of this altered balance between cell proliferation and ongoing cell loss. Tamoxifen may also directly induce apoptosis. ¹¹

Endometrial effects are mediated by the same mechanism of endogenous E2s, accounting for endometrial stimulation that represents the basis of endometrial benign or malignant disease onset, particularly in the case of unopposed Pg activity. ¹¹

Postmenopausal women have very low serum E2 and Pg levels; on the contrary, they have high serum luteinizing hormone (LH) and follicle-stimulating hormone (FSH) concentrations; while in these women TAM reduces gonadotropin secretion, in premenopausal women it slightly increases gonadotropin and E2 secretion. An increase in endogenous E2 secretion might displace TAM from ERs limiting its therapeutic efficacy. ^9,11

OPEN IN VIEWER

Table 1.

Endometrial Safety, Benefits, and Side Effects of Mirena in Tamoxifen-Treated Breast Cancer Survivors.

Author (Year)	Study Design	Study Population	Exclusion Criteria	Methods (Pts)	Uterine Cavity Surveillance (Techniques and Timing)	Results (Cases vs Controls)
Gardner 200017	RCT	113 women in at least 1 year TAM therapy after breast cancer – all postmenopausal	– Suspected pelvic inflammatory disease – active liver disease – History of malignant disease other than breast cancer – Presence of G2 submucous fibroids or endometrial polyps – Refused to receive the LNG-IUS	Treatment group (57) endometrial surveillance with insertion of LNG-IUS Control group (56) endometrial surveillance	Transvaginal Ultrasonography Hysteroscopy and endometrial biopsy – at recrutiment – 12 months after TAM	Endometrial polyp 2% vs 8% Submucosal fibroid 2% vs 6% Complex hyperplasia 0% vs 2%
Chan 200718	RCT	113 women in TAM therapy after breast cancer – 66 premenopausal – 47 postmenopausal	– Pelvic inflammatory disease – Congenital uterine anomaly – Uterine cavity length >10 cm	Treatment group (55) prophylactic LNG-IUS insertion before TAM Control group (58) endometrial surveillance	Transvaginal Ultrasonography Hysteroscopy and endometrial biopsy – before TAM – 12 months after TAM	Endometrial polyp 1.8% vs 15.5% Submucosal fibroid 1.8% vs 3.4% Cystic endometrium 7.3% vs 3.4%
Kesim 200819	RCT	142 women in at least 1 year TAM therapy after breast cancer – all postmenopausal	– History of malignant disease other than breast cancer	Treatment group (70) endometrial surveillance with insertion of LNG-IUS Control group (72) endometrial surveillance	Transvaginal Ultrasonography Hysteroscopy and endometrial biopsy – at entry – 36 months after TAM	Endometrial polyp 5.7% vs 19% Submucosal fibroid 2% vs 6% Endometrial hyperplasia without atypia 0% vs 5.5%

Abbreviations: RCT, randomized controlled trial; pts, patients; LNG-IUS, levonorgestrel intrauterine system; TAM, tamoxifen.

Results

In the interval time considered, we found more than 1800 articles in scientific database literature, of which only 7 satisfied our selection criteria.

Eligible studies evaluating LNG-IUS effect in TAM-treated patients were three ^{17 –19} for endometrial effects and four ^{20 –23} for breast effect.

Mirena Endometrial Effect

Through a randomized controlled trial, Gardner et al compared endometrial surveillance alone versus endometrial surveillance before and after the insertion of Mirena for 12 months in 113 postmenopausal women who had been taking adjuvant TAM treatment for at least 1 year and who were undergoing regular follow-up for breast cancer. Women with known endometrial pathology or contraindications to LNG-IUS were excluded. Exclusion criteria were suspected pelvic inflammatory disease, active liver disease, history of malignant disease other than breast cancer, presence of grade III submucous fibroids or endometrial polyps, or reluctance to receive the LNG-IUS. Patients underwent transvaginal ultrasonography for monitoring endometrial thickness and hysteroscopy with endometrial biopsy before and 12 months after TAM treatment. The authors discovered an increased incidence of endometrial polyps (2% vs 8%), submucosal fibroid (2% vs 6%), and complex hyperplasia (0% vs 2%) in the control group when compared to the study group. ¹⁷

In another randomized controlled trial, Chan et al compared endometrial surveillance alone versus prophylactic Mirena insertion in pre- and postmenopausal women treated with TAM for 1 year before TAM treatment and after completion of postoperative radiotherapy and chemotherapy. The authors considered 113 TAM-treated women after breast cancer: 66 premenopausal and 47 postmenopausal. Exclusion criteria were pelvic inflammatory disease, congenital uterine anomaly, and uterine cavity length >10 cm. Treatment group was constituted by 55 patients who received a prophylactic LNG-IUS insertion before TAM compared to the control group of 58 patients who underwent endometrial surveillance. Patients underwent transvaginal ultrasonography for monitoring endometrial thickness and hysteroscopic endometrial biopsy before and 12 months after TAM treatment. Increased incidence of endometrial pathologies was reported in the control group, endometrial polyps accounting for 1.8% vs 15.5% (study vs control), submucosal fibroid 1.8% versus 3.4%, and cystic endometrium 7.3% versus 3.4%. ¹⁸

Kesim et al conducted a cohort study on 142 postmenopausal women taking adjuvant TAM treatment for at least 1 year. In all, 70 patients with LNG-IUS insertion constituted the treatment group undergoing endometrial surveillance, while 72 patients represented the control group.

Endometrial surveillance by transvaginal ultrasound and hysteroscopic biopsy after 36 months of follow-up was performed. Analyzing incidence of endometrial polyps, submucosal fibroids, and endometrial hyperplasia without atypia, they concluded that LNG-IUS was the better option for endometrial protection. In fact, the incidence of endometrial polyp was 5.7% versus 19% (treatment vs control), submucosal fibroid 2% versus 6%, and endometrial hyperplasia without atypia 0% versus 5.5%. ¹⁹

OPEN IN VIEWER

Table 2.

Breast Safety, Benefits, and Side Effects of Mirena in Breast Cancer Survivors.

Author (Years)	Design	Population	Exclusion Criteria	Methods (Pts)	Data Collection	Results (Cases vs Controls)
Backman (2005) 20	EPID	23 885 women filled with LNG-IUS	Not reported	Epidemiological data sourced (17 360)	Questionnaire to collect data on exposure, prognostic factors and diagnosis of breast cancer	Age group 30-34 ys 27.2 vs 25.5 per 100 000 (LNG-IUS vs overall population) Age group 35-39 ys 74.0 vs 49.2 per 100 000 (LNG-IUS vs overall population) Age group 40-44 ys 120.3 vs 122.4 per 100 000 (LNG-IUS vs overall population) Age group 45-49 ys 203.6 vs 232.5 per 100 000 (LNG-IUS vs overall population) Age group 50-54 ys 258.5 vs 272.6 per 100 000 (LNG-IUS vs overall population)
Trinh (2008) 22	RETa	Breast cancer patients who were using the LNG-IUS – 97.4% premenopausal – 2.6% postmenopausal	– Insertion of LNG IUS after recurrence – Distant metastasis at the time of LNG IUS insertion. – In situ breast carcinoma	Case group (79) – subgroup A (38) LNG IUS insertion before breast cancer diagnosis – subgroup B,(41 pts) LNG IUS insertion after breast cancer diagnosis Control group (120) (women with no history of LNG IUS use)	Oncologic follow-up	Breast cancer recurrence 21.5% vs 16.6% (LNG-IUS vs controls) % of chemotherapy 84% vs 51% (subgroup A vs subgroup B) % of axillary LN involvement 47.4% vs 29.3% (subgroup A vs subgroup B)
Lyytinen (2010) 23	RET	All Finnish women diagnosed with their first invasive breast cancer between 50 and 62 years (1995-2007)	Not reported	Case group (9956) (patients from the Finnish Cancer Registry) Control group (29 868; women, 3 for each cancer case free of breast cancer randomly selected from the Finnish National Population Register)	History of HT use	Invasive breast cancer 329 vs 708
Dinger (2011) 21	RET	Women with diagnoses of breast cancer (invasive carcinoma or carcinoma in situ) younger than 50 years old at diagnosis	– Prior history of Malignancy – Currently living abroad or were unable to provide the required study information in the respective national language	Case group (5113) (breast cancer cases) 1085 used LNG-IUS (22%) 1130 used Copper IUS (22%) Control group (20 452) 4261 used LNG-IUS (21%) 4298 used copper IUS (21%)	Filling questionnaires to collect data on exposure, prognostic factors and diagnosis of breast cancer	Known risk factors were more frequent in cases In situ breast cancer 6.0% vs 6.1% (LNG-IUS vs copper IUS) Invasive breast cancer 94% vs 93.9% (LNG-IUS vs copper IUS) no significant increased risk of cancer for LNG-IUS Vs hormonal contraceptives, progestin-only pills and injections, nonuse hormonal contraceptives.

Abbreviations: RCT, perspective randomized controlled trial; EPID, epidemiologic study; RET, retrospective controlled cohort study; HT, hormone therapy; pts, patients; LNG-IUS, levonorgestrel intrauterine system; TAM, tamoxifen.

^a Multicentric.

Discussion

Tamoxifen has been approved by US Food and Drug Administration with the following indications: adjuvant therapy of breast cancer, metastatic breast cancer, and reduction in the incidence of breast cancer in women at high risk. ¹

Tamoxifen is well tolerated by most patients with breast cancer: less than 5% of patients withdrew from therapy because of toxicity. The most common adverse effects are menopausal symptoms, especially before menopause. Most serious adverse effect is the potential endometrial tumor-promoting activity. Endometrial hyperplasia, development of polyps, increase in endometrial thickness, and ovarian cysts have been attributed to TAM. It is important to notice that incidence of endometrial cancer in TAM users is similar to the one in EPT users. Nearly all reported endometrial cancers have been diagnosed in postmenopausal women. ¹¹

Even if TAM increases the risk of preneoplastic and neoplastic endometrial disease, several randomized clinical trials have shown that therapeutic benefits of TAM as adjuvant treatment of breast cancer exceed the risks related to endometrium stimulation. ^{24 –26}

In an early study of the National Surgical Adjuvant Breast and Bowel Project (NSABP), the rate of endometrial cancer occurrence among TAM users who were administered 20 mg/die was 1.6/1000 patient-years compared to 0.2/1000 patient-years among patients taking a placebo. The 5-year disease-free survival rate from breast cancer was 38% higher in the TAM than in the placebo group, suggesting that the small risk of developing endometrial cancer is outweighed by significant survival benefits provided by TAM therapy. Frequently, TAM-related endometrial cancer shows low stage and grade, similar to those associated with EPT. ²⁷ In a more recent update of all NSABP trials on patients with breast cancer, the rate of endometrial cancer was 1.26/1000 patient-years in TAM users versus 0.58/1000 patient-years in the placebo. ²⁶ This result is similar to that of Wickerham et al. ²⁸ The increased incidence has been shown even in the Early Breast Cancer Trialists' Collaborative Group (EBCTCG) meta-analysis, considering 20 clinical trials and involving 20 000 women. Overall, the risk was increased (relative risk [RR]: 2.40), in particular in women aged more than 55 years (RR: 2.96 versus 1.75) compared to women aged 45 to 54 years. The RR was 4.1 for women aged less than 45 years. ²⁴ So the TAM estrogenic effects appear to be largely confined to postmenopausal women, while premenopausal women do not appear to have an increased risk of endometrial cancer onset. ⁷

Nevertheless, last ACOG Committee Opinion on TAM and endometrial cancer found the prospective trials on proper follow-up of asymptomatic postmenopausal patients with breast cancer treated with TAM as insufficient to give definitive guidelines. ^7,29

Progesterone has been used in the treatment of recurrent endometrial cancer for over 50 years, with reported response rates 80% high if the recurrence is single, well differentiated and not in a previously irradiated site. The presence of steroid receptors, however, is the most important predictor of response.

Well-differentiated tumors (stage 1, grade 1) would be first treated by Pg since endometrial proliferation is E2 dependent, and Pg has an antiestrogenic effect on the endometrium. ^{30 –33} Failure is linked to the loss of PgR with prolonged Pg treatment: gradual loss of PgR begets resistance to Pg-based therapy, leading to the need for alternative approaches to pharmacologically modulate the downstream differentiating pathways controlled by Pg when PgR is robustly expressed. ¹³

More recently, LNG-IUS has also been successfully used to treat endometrial hyperplasia. Recently Gallos et al concluded that oral Pg appear to induce lower disease regression rate than LNG-IUS in the treatment of endometrial hyperplasia. The authors considered these findings linked to Pg delivery since the concentrations acting directly on the uterine mucosa seem to exceed that of the oral treatment, reducing the treatment dropout. ³⁴

Since hyperestrogenic milieu seems the major risk factor for endometrial hyperplasia, patients treated with TAM patients can be compared to patients with unbalanced E2 therapy: these patients had an RR till 7.5 for endometrial cancer and they may benefit of Pg treatment. ³⁵

From our analysis, as in a recent Cochrane considering the role of LNG-IUS for endometrial protection in women with breast cancer treated with TAM, only few studies are available in literature to guide the physicians in endometrial surveillance. ⁶ In all eligible studies, the LNG-IUS in TAM users led to a significant incidence reduction in endometrial polyps, but they were not sufficiently powered to detect whether LNG-IUS leads to significant changes in the incidence of endometrial hyperplasia or adenocarcinoma.

The LNG-IUS is easy to introduce, well tolerated, despite spotting or menorrhagia being the most frequent side effect. All authors investigated the endometrial uterine cavity before and after LNG-IUS placement in order to guarantee the feasibility of LNG-IUS insertion and the detection of endometrial disease onset.

In none of the studies, cases of endometrial cancer or atypical endometrial hyperplasia were reported, perhaps because of efficacy of LNG-IUS or because of the dimension of the samples reached, too little to gain a sufficient power for discovering precancerous or cancerous lesions.

Nevertheless, none of the studies define the safety of LNG-IUS in relation to breast cancer recurrence. In fact, the LNG effect on breast tissue is still unclear. Mirkin et al reported that LNG increases vascular endothelial growth factor (VEGF) messenger RNA (mRNA) in T47D (P receptor rich) breast cancer cell lines. In vitro studies have shown LNG stimulating cell growth, even if subclones of cell lines may respond differently to high pharmacological levels of LNG. The LNG stimulatory effect involves the ER. Administration of E2 or LNG separately stimulates cell growth in MCF-7 and T47-A breast cancer cell lines. However, with combined administration, the E2-induced cell growth could be significantly inhibited. So LNG-only administration can stimulate breast cancer cell growth, and this effect may play a role in cancer angiogenesis by increasing VEGF mRNA. ³⁶

The hypothesis of an LNG-IUS stimulation on breast cancer cell depends on the possible significant LNG-IUS systemic exposure. Serum concentrations of LNG-IUS in premenopausal endometriosic patients reached 459.2 ± 100 pg/mL after 1, 368.2 ± 51.8 pg/mL after 3, and 357.3 ± 53.0 pg/mL after 6 months. These concentrations are comparable to those reached among users of 30-mg LNG-only pill. Occasionally, LNG-IUS may release more LNG depending on the environment or if systemic uptake of LNG in the peritoneal cavity is higher. In LNG-IUS users, there is a significant systemic uptake of LNG comparable to those in women taking LNG-only pills and higher only in certain subgroups of women. ³⁷

The studies on LNG-IUS breast effects in TAM treatment do not agree on the real results on the cancer recurrence: there is no evidence in efficacy and safety to support its use in TAM-treated breast cancer survivors. None of these contemporary studies evaluated side effects and safety of LNG-IUS both on breast tissue and endometrium.

Progesterone systemic bioavailability during local therapy may play a key role in the balance between benefits and side effects on different tissues, since plasmatic levels are responsible for breast tissue proliferation.

Despite the suspicious that LNG-IUS acts not only locally but also systematically, the very recent long-term follow-up study (Hong Kong Randomized Controlled Trial Study) conducted by Wong et al ³⁸ reported that after 5 years of LNG-IUS treatment in TAM users, no statistically significant increase in the breast cancer recurrence rate and breast cancer–related deaths was found when compared to controls. However, the authors recommended to interpret the finding with caution because the sample size was not of sufficient power to address this outcomes.

At the same time, Wong et al ³⁸ recommended further larger study in this field concluding that LNG-IUS significantly reduces the occurrence of endometrial polyp in women using TAM, but its role in the prevention of endometrial hyperplasia and adenocarcinoma as well as its effect on risk of breast cancer recurrence remain uncertain.

Certainly, the actual evidences confirm that endometrial status before TAM treatment should be investigated in order to identify precancerous lesions already present that could regress through a progestin therapy balancing TAM-E2-like effects. ³⁹

Nowadays, the breast antiblastic effects of raloxifene (a second-generation SERM) have been proven. ⁴⁰ In the next future, if the raloxifene antiproliferative effects will be clarified and confirmed also in endometrial tissue, this SERM could give the possibility to shift the TAM user patients into raloxifene treatment, removing the concern about Pg endometrial protection.

Conclusion

The considered studies did not solve the doubts concerning LNG-IUS use for endometrial protection in TAM users, since its breast effects are still unclear and published literature is insufficient to guide clinical practice. Nowadays, there is no plausible evidence concerning efficacy and safety supporting LNG-IUS use in TAM-treated breast cancer survivors nor clear contraindications to its application in endometrial protection.

It is mandatory assessing hormonal status before TAM treatment since premenopausal women have unknown increased risk for endometrial cancer and require no further than routine gynecological check.

Postmenopausal women should routinely receive an hysteroscopic follow-up, that is the only procedure that could directly evaluate endometrial status before TAM therapy, since ultrasound evaluation alone is hindered by the LNG-IUS itself. This permits to individuate high-risk patients developing endometrial pathology.

Nevertheless in absence of endometrial atypia, postmenopausal patients demonstrate a hyperestrogenic environment due to physiologic progestin deficiency resulting in high-risk endometrial pathology: such a risk could be reduced by Mirena, making equal postmenopausal women’s risk to premenopausal one. It has more relevance in case of endometrial hyperplasia, where Mirena plays a therapeutic role.

It is mandatory to define the hormonal status before TAM treatment. In accurately selected patients, Mirena has been proven to protect endometrium. Perspective clinical trials on pharmacodynamics and pharmacokinetics of Mirena are necessary in order to assess whether systemic levels of Pg could increase breast cancer recurrence rate in such patients.