Uterine synechiae after intrauterine device use: a case series

Abstract

Purpose

The intrauterine device (IUD) is one of the most effective and safe contraceptive methods. Substantial literature suggests an overall return to normal fertility following IUD removal. However, there are no studies to date that evaluate endometrial function specifically in nulliparous women after levonorgestrel IUD use.

Methods

We present three nulliparous women with a history of levonorgestrel IUD use who were evaluated for uterine dysfunction at the University of California, San Francisco Center for Reproductive Health. These patients had no other known risk factors or history of uterine manipulation, including prior uterine surgery, pelvic radiation, intrauterine infection, hypothalamic amenorrhea, or uterine anomaly.

Results

Upon evaluation, these patients were found to have uterine synechiae concerning for Asherman syndrome. All three patients were eventually able to conceive through assisted reproductive technology or natural conception.

Conclusion

This case series is the first to suggest a possible effect of endometrial dysfunction on fertility resumption following levonorgestrel IUD removal in nulliparous patients. It is possible that a small subset of patients may be at risk for Asherman syndrome after IUD use. Larger prospective trials are needed to explore this possible association.

Introduction

The intrauterine device (IUD) is an effective long-acting reversible method of contraception. There are two types of IUDs available in the USA: the copper T 380A (Paragard) and levonorgestrel-containing hormonal IUDs (Mirena, Skyla, Liletta, and Kyleena). The primary mechanism of action of the copper IUD is prevention of fertilization through inhibition of sperm migration and viability, while the levonorgestrel IUD changes the amount and viscosity of cervical mucus [1, 2]. Pregnancy rates within the first year of insertion for both the copper and levonorgestrel IUDs range from 0.2 to 0.9% [3, 4]. IUD use has increased over the past decade in the USA, with rates as high as 10.6% in large, nationally representative studies [5]. Despite initial provider reluctance to recommend IUDs to nulliparous women due to perceived safety concerns, IUD use has increased in this population from 2.1% in 2009 to 5.9% in 2012 [1, 6]. Additionally, there has been an overall decrease in unintended pregnancy rates by 18%, and it has been shown that the levonorgestrel IUD in particular may lower the risk of pelvic inflammatory disease [1, 7–9].

Current evidence suggests that there is a normal return of fertility after discontinuation of the IUD, with pregnancy rates between 71 and 88% at 1 year following the removal of the device [4, 10–12]. This return to fertility has been shown for both types of IUDs; however, there is some evidence that the levonorgestrel IUD may affect endometrial functioning and fertility resumption [11]. In particular, the levonorgestrel IUD causes atrophy of the endometrial glands and decidualization of the endometrial stromal cells, which appear to normalize after levonorgestrel IUD removal [11, 13–17]. Subsequent studies, although limited, have shown an upregulation of decidualization markers IGFBP1 and prolactin and downregulation of estrogen and progesterone receptor expression with the levonorgestrel IUD [11, 16, 17]. It is unclear if changes in endometrial gene expression or function persist after removal of the inert IUD, or how prolonged IUD use, timing of IUD placement, or nulliparity influence these changes [11, 14].

As such, it is important to begin to study the possible effects of the levonorgestrel IUD on long-term endometrial function and fertility, particularly for women who desire future pregnancy. Infertile women represent a unique population who routinely undergo evaluation of the endometrium by ultrasound and hysteroscopy as part of their diagnostic workup and treatment. In this case series, we present three women with a history of levonorgestrel IUD use who presented to the University of California, San Francisco Center for Reproductive Health with infertility. These patients were identified by providers as unique because they had uterine adhesions or synechiae following IUD use without any other identifiable risk factors for endometrial dysfunction or scarring, including prior uterine surgery, pelvic radiation, intrauterine infection, hypothalamic amenorrhea, or uterine anomaly.

Cases

Patient 1

A 39-year-old G0 woman presented with 6 months of infertility (Table 1). Her history was notable for use of a levonorgestrel 52 mg (Mirena) IUD for 8 years, with removal 2 years prior to attempting conception. The patient reported having a painful IUD insertion. She denied any other contraceptive use. Her initial fertility evaluation showed regular ovulation, patent fallopian tubes bilaterally, and a normal semen analysis. Prior to her presentation, she was found to have a space-occupying lesion on hysterosalpingogram that was identified by her gynecologist.

Table 1.

Patient information regarding IUD use, infertility diagnosis and treatment, and uterine cavity evaluation

Patient number	Age diagnosis (years)	Gravity and parity	Duration IUD use (months)	Duration: IUD removal and conception attempt (months)	Duration of infertility (months)	Maximum endometrial lining thickness (mm)	Infertility diagnosis other than uterine factor	Infertility treatment	Cavity evaluation
1	39	G0	96	24	6	7.9	None	Super ovulation with intrauterine insemination (IUI)	Hysteroscopy ×2: circumferential adhesions mid-uterus
2	38	G0	12	96	18	7.0	None	IVF stimulation ×4, controlled frozen embryo transfer (FET) cycle ×3	Hysteroscopy: cornual adhesions
3	39	G2P0020	120	14	9	2.5	None	None to date	Hysteroscopy: cornual adhesions

She underwent an operative hysteroscopy, which was notable for circumferential scarring at the mid-uterus extending to the left cornua (Fig. 1). The adhesions were fully resected. Three months after resection, her endometrial thickness was noted to reach a maximum of 6.1 mm in the late follicular phase of her natural cycle. Repeat hysteroscopy was planned; however, she conceived naturally and had a first-trimester spontaneous miscarriage. A hysteroscopy was subsequently completed, with resection of residual scar tissue. She then underwent a single superovulation cycle with intrauterine insemination. During this cycle, her endometrium reached a maximal thickness of 7.9 mm with three dominant follicles and resulted in a single intrauterine pregnancy and full-term live birth.

Fig. 1.

Hysteroscopy, Patient 1: circumferential scarring at the mid-uterus, with more dense scarring toward the left and extending to the cornual region (shown with *)

Patient 2

A 38-year-old G0 woman presented with 18 months of infertility. Prior to her initial consultation at our clinic, she had completed two cycles of superovulation with intrauterine insemination and four in vitro fertilization (IVF) cycles, with a single euploid embryo cryopreserved and no prior embryo transfers. She had a history of a levonorgestrel 52 mg (Mirena) IUD for 12 months, which was removed 8 years prior to attempting conception. She reported using oral contraceptives (OCPs) before and after her IUD insertion. Her initial evaluation was notable for an endometrial thickness of 3.4 mm in the luteal phase.

At our clinic, she underwent two additional IVF cycles, with peak estradiol levels of 2266 ng/mL and 2160 ng/mL and maximum endometrial thickness of 6.8 and 7.0 mm. She then underwent a hysteroscopy, which showed filmy adhesions at the right cornua extending to the midline that were fully resected (Fig. 2). She underwent three programmed frozen embryo transfer (FET) cycles. During the first and second FET cycle, her endometrium reached a maximal thickness of 6.4 mm and 5.6 mm, respectively; a single euploid embryo was transferred both cycles but did not result in pregnancy. A third FET cycle was then attempted with a maximal endometrial thickness of 6.4 mm. Two euploid embryos were transferred, which resulted in a singleton pregnancy and a full-term live birth.

Fig. 2.

Hysteroscopy, Patient 2: filmy adhesions around the right cornua extending to the midline (shown with *)

Patient 3

A 39-year-old G2P0020 woman with a history of one prior pregnancy of unknown location treated with methotrexate and a biochemical pregnancy presented after actively trying to conceive for 9 months. The patient had a 10-year history of levonorgestrel 52 mg (Mirena) IUD use, with removal 14 months prior to infertility evaluation. She had a 4-year history of Depo Provera use prior to her IUD. Her initial evaluation was notable for a luteal phase ultrasound demonstrating an isoechoic endometrium with a thickness of 2.5 mm. Due to these ultrasound findings, she underwent hysteroscopy, which was notable for right cornual occlusion secondary to dense anterior wall adhesions (Fig. 3). The adhesions were fully resected. The patient was started on daily estrogen and developed an endometrial thickness of 6.4 mm. She attempted natural conception given improved endometrial thickness and is currently 7 weeks pregnant.

Fig. 3.

Hysteroscopy, Patient 3: right cornual occlusion secondary to dense anterior wall adhesions (shown with *)

Discussion

Since its introduction over 60 years ago, the IUD has become one of the most utilized forms of contraception in the USA, with approximately 4.4 million currently in use [18]. The safety and efficacy of the IUD have been rigorously documented in the scientific literature; however, there are limited data on endometrial function and resumption of normal fertility following levonorgestrel IUD use, particularly in nulliparous women. In this study, we present three cases of patients with infertility who were diagnosed with uterine synechiae without any identifiable risk factors other than the use of a levonorgestrel IUD. To our knowledge, this is the first case series of clinical endometrial dysfunction and particularly uterine synechiae following levonorgestrel IUD removal in an infertility setting. We believe this case report draws attention to a possible risk associated with levonorgestrel IUD use and provides the foundation to support further basic and translational studies in this area.

A recent publication described the potential for long-term impact on the return of fertility following the removal of the levonorgestrel IUD and highlighted the lack of research on this topic [11]. Current literature suggests that the inert IUD causes prolonged dysregulation of endometrial gene expression [19]. While the morphology of the endometrium appears to normalize within 1-2 months following levonorgestrel IUD removal, it is unknown if genetic expression and endometrial function normalize as well [11, 14]. It is therefore possible that, like the inert IUD, the levonorgestrel IUD may change gene expression patterns in the endometrium, including downregulation of estrogen and progesterone receptors, which could impact conception following IUD removal. At a translational level, there are no studies that assess the resumption of fertility of patients with prolonged levonorgestrel IUD use compared with those without prior IUD use, further limiting the knowledge on this subject.

Our case report outlines three unique cases of uterine synechiae consistent with Asherman syndrome following IUD use without any other history of uterine manipulation. It may be that a particular, and likely very small, subset of patients is at higher risk for scar tissue formation after IUD placement. The causes of this susceptibility are likely multifactorial; however, the risk of Asherman syndrome is known to be higher with hypoestrogenemia, as estrogen is required for endometrial regeneration [20]. It is plausible that IUD placement in the immediate postpartum period, while breastfeeding, or following extended progestin exposure like after long-term OCP use may make a susceptible patient more likely to develop scar tissue. During these times, the endometrial lining is thin because the normal proliferative and secretory endometrial changes seen during a menstrual cycle are inhibited, by either the absence of estrogen itself or the inhibition of its actions on the endometrium by progesterone. Similarly, the proliferation of the endometrial lining is inhibited with prolonged levonorgestrel IUD use. Traumatic insertion of the IUD is another plausible risk factor for formation of uterine adhesions, which may have played a role in the development of uterine synechiae for Patient 1. Finally, it is possible that a sub-clinical, undiagnosed intrauterine infection could promote scar formation. It is important to note, however, that all of the patients were ultimately able to conceive.

Our study has a number of limitations. Most importantly, case series are small and cannot determine causality, although these reports can be useful tools to identify rare outcomes. Additionally, we lack baseline data of our patients prior to IUD insertion; therefore, it is possible that these patients had evidence of Asherman syndrome before their IUD. The retrospective nature also presents the potential for both recall and selection bias. We cannot exclude the possibility of uterine manipulation that was not reported during clinical evaluation and interviewing; however, we believe that the risk of this possibility is unlikely in our patient population. Finally, our study does not include a control group of women with thin endometrial lining without any risk factors and/or without prior IUD use.

In conclusion, the IUD is tremendously safe and effective, and many women consider it to be the best method of contraception for them. Our case series raises the possibility that there may be a small risk of Asherman syndrome following IUD use. To ensure that we are providing the most evidence-based care to our patients, we believe that it is important to study this issue further. Ultimately, we hope that this case series serves as a foundation for larger prospective longitudinal cohort or case-control studies that can explore the relationship between IUD use, endometrial morphology, and fertility.

Author contribution

M.K.A. and K.W.’s roles included study design, data collection, data analysis, and manuscript writing; M.I.C.’s roles included study design and manuscript writing; M.N.’s roles included study design, data collection, data analysis, and manuscript writing.

Funding

This study was supported by departmental research funding within the University of California, San Francisco Department of Obstetrics, Gynecology and Reproductive Sciences.

Data availability

We are unable to provide our data due to its identified nature.

Declarations

Ethics approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the University of California, San Francisco (UCSF) Committee on Human Research and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Consent to participate

Informed consent was obtained from all individual participants included in the study.

Consent for publication

As part of the consent process, all participants gave written consent to use de-identified information for research purposes and publications.

Competing interests

The authors declare no competing interests.