The levonorgestrel-releasing intrauterine system: Safety, efficacy, and patient acceptability

It is clear that in peri- and postmenopausal females who retain their uteri and desire estrogen replacement therapy secondary to climacteric side effects it is essential to provide progestin medication in order to counteract the proliferative effect of estrogen on the endometrium. The standard treatment has been to use a cyclic oral progestin preparation at the end of each cycle in combination with the estrogen. However, the oral regimen is associated with cyclic vaginal bleeding which is burdensome to some, and can also have a variety of androgenic side effects. In addition, multiple studies have shown possible links with systemic progesterone and increased cardiovascular risks in addition to increased breast cancer risks. 47 , 48 The LNG-IUS has been studied as an alternative method of providing locally acting progestin which can accomplish the endometrial protective effects with a minimal amount of adverse side effects secondary to its significantly decreased serum concentration. One study of 40 perimenopausal women complaining of climacteric symptoms evaluated two treatment regimens: three week cyclic treatments with 2 mg oral estradiol (E 2 ) valerate combined with 250 μg of oral levonorgestrel for the last 10 days of each cycle versus 2 mg of E 2 valerate continuously with the LNG-IUS. After one year the subjective symptoms in both groups were improved, and none of the women had endometrial proliferation. However, in the LNG-IUS group 15/18 women were amenorrheic, versus all women in the oral LNG group who continued with cyclic bleeding. 49 Another study reviewed the available literature (19 studies with a total of 826 subjects) regarding progestin intrauterine devices in combination with estrogen, with a duration of six months to five years, and determined that none of the subjects developed endometrial hyperplasia throughout the course of the study. 50 Hampton and colleagues evaluated 82 perimenopausal women treated with oral estrogen and the LNG-IUS for long term protection of the endometrium. Nonproliferative endometrium was present in 98.6% of the participants at the end of 12 months, 98.6% at 24 months, 95.5% at 36 months, 96.8% at 48 months, and 95.2% at 60 months. There were no cases of hyperplasia throughout the entire 60 month period. Amenorrhea was present in 54.4% of the women after 12 months, and 92.7% after 60 months. Overall the treatment was well tolerated, with 79.8% of the women still continuing with the estrogen and LNG-IUS after the full 60 months of the study and 92.4% of the participants rating the treatment as “good” or “very good”. 51 The LNG-IUS thus appears to be a highly effective and acceptable method of providing endometrial protection for peri- and postmenopausal women who are taking estrogen replacement therapy.

Worldwide, intrauterine devices (IUDs) are the most widely used and effective reversible contraceptive method and are safe, extremely “low maintenance” contraceptives. They are second only to female sterilization as the most prevalent method of family planning worldwide (13.6% vs 20.5%). 1 Over 150 million women worldwide use the IUD for contraception. However, for a variety of reasons, including the continued concern many women have about the morbidity and mortality associated with IUDs resulting (whether they know it or not) from the Dalkon shield experience of the 1970s, there is a large disparity between the numbers of users in developing versus developed countries. There is an especially low utilization rate within the United States. Nearly 15% of women who use contraception in less developed countries choose the IUD, compared with only 8% of women in developed countries and only 2% in the United States. Two thirds of worldwide IUD use (over 100 million women) is represented by China alone. 2 There is currently a wide variety of IUDs available worldwide; some are inert, some are copper containing, and some are medicated with levonorgestrel or indomethacin. They also come in a wide variety of sizes and shapes including T-shaped and “frameless” devices. This article will focus specifically on the LNG-IUS, exploring mechanism of action, safety, efficacy, variety of uses, and patient acceptability.

The LNG-IUS is a safe and effective device which has been shown to be useful not only as a contraceptive, but as a tool which can be used for a variety of gynecologic conditions or disorders. However, the IUS is not without its side effects. The most common side effect associated with the LNG-IUS is menstrual irregularity. In fact, nearly 25% of women discontinue using the LNG-IUS because of amenorrhea. 1 Approximately 20% of LNG-IUS users will be amenorrheic by the end of 12 months, 3 and 70% of users will be oligomenorrheic or amenorrheic by 24 months. 52 In addition, though the systemic absorption is much smaller than with other progestin-containing contraceptives, there is some absorption and it can occasionally be associated with a variety of undesirable side effects such as ovarian cysts, acne, weight gain, depression and decreased libido. However, the device seems to be well tolerated overall. A three-year study which specifically set out to evaluate the long-term acceptability of the LNG-IUS followed 165 women over 36 months. The study showed a three-year continuation rate of 90.3%. In total, 97% of participants reported alterations in their menstrual pattern, with 34% reporting a decrease in the amount of bleeding, 17% had persistent spotting or intermenstrual bleeding, and 56% reported at least a temporary period of amenorrhea. 81% of the women with amenorrhea viewed it as a positive change for them. There was also a significant improvement in the amount of pain associated with menstruation, with a decrease from 60% to 29% in women reporting dysmenorrhea after 36 months. The number of women who expressed that they were very satisfied with the LNG-IUS increased steadily with the duration of the treatment, with 29% after two weeks, 56% after two months, 69% after six months and 77% after 36 months. 53 Another study of 78 women who had the LNG-IUS inserted for menstrual disorders also evaluated patient satisfaction through questionnaires. 12% of the women had the IUS prematurely removed, with the major reason being pain and heavy bleeding. However, the majority of participants were satisfied with their results. 78% of the women reported improvement in their periods, and 84% reported improvement in menstrual discomfort. 72% of the women reported they would use the LNG-IUS again, 73% would recommend it to their peers, and the overall satisfaction rate was 76%. 54 The amount of satisfaction with the LNG-IUS, not surprisingly, appears to correlate with how well informed patients are about the possible side effects associated with it. A study that involved the evaluation of 17,914 questionnaires of current LNG-IUS users showed that 74% were very or fairly satisfied with it. User satisfaction correlated with the amount of information provided regarding different symptoms (menstrual irregularities, greasy hair/skin, pregnancy and PID) regardless of whether or not the patient actually experienced that specific symptom. In particular, the women who were warned of the possibility of amenorrhea were more satisfied than the women who were not. 55

The IUD, in general, is one of the most effective forms of contraception available today, with a global cumulative pregnancy rate of <2% at five years. The LNG-IUS, in particular, is possibly the most effective IUD available according to numerous studies that show its global cumulative pregnancy rate to be <0.5%. 22 A large randomized-controlled trial of 2,244 women over seven years showed a pregnancy rate of 1.1% with the LNG-IUS compared with 1.4% with the TCu380. 23 In two other seven-year follow up studies with 293 and 82 women respectively, there were no pregnancies in women using the LNG-IUS. 24 , 25 A Cochrane review in 2004 showed the LNG-IUS to be as effective as copper IUDs with a copper surface area of >250 mm 2 , and more effective than those with <250 mm. 22 , 26 The LNG-IUS is also an effective option for women to choose immediately post abortion. Multiple studies have shown that copper-releasing IUDs are safe and effective during this timeframe. Fewer studies have been done looking specifically at the LNG-IUS under this circumstance; however, they have shown it to be at least as effective as the copper IUD. One study followed 305 women who received the Mirena ® IUS immediately post-abortion for up to five years. There were a total of two pregnancies for a pregnancy rate of 0.8% at five years. There was a discontinuation rate secondary to expulsion of 7.1% at one year and 10.5% at five years 27 which is obviously higher than the interval expulsion rate of 2%–3% per year, 28 but not significantly different than that of the NovaT (8.6% at one year and 15.4% at five years). 28 A LNG-IUS may, therefore, be safely inserted immediately after either uncomplicated spontaneous or induced abortions. The levonorgestrel-releasing intrauterine system can also be safely inserted immediately after childbirth, but there is also an increased risk of expulsion compared with interval insertion. A study of 19 women who received a post placental LNG-IUS, defined as placement within 10 minutes of placental delivery, showed an expulsion rate of 10.5% and no infections. 27 There are no studies that specifically examine the timing of LNG-IUS insertion post partum, but one done with a copper-releasing IUD showed post-placental IUD insertion to have significantly lower expulsion rate (14.3% complete and 22.6% partial) than early post partum insertion, defined as taking place 10 min to 72 hours after delivery of the placenta, at the end of one year (18.6% complete and 51.2% partial). Both rates were much higher than interval insertion (3.8% complete and 3.1% partial). There were no perforations in either of the post partum groups, and 2.3% perforation rate in the interval groups. 29 Insertion of the LNG-IUS is therefore a reasonable, almost effortless, option for post placental insertion if the woman is properly counseled about the increased risk of expulsion versus interval insertion and receives proper counseling on checking the IUS strings and receives appropriate clinical follow-up. For many low-resource settings, since the cervix (or uterus in the case of insertions at Cesarean delivery) is open at delivery and the patient is commonly in lithotomy position, post-placental insertion may reduce the barriers to interval insertions such as lack of speculae, tenacula, sounds, exam tables, and already crowded clinics.

Although the intrauterine contraceptive device should ideally be inserted within the first seven days following a woman’s menstrual cycle, it can be inserted at any time during the cycle if it is reasonably certain that she is not pregnant and has not been at risk of pregnancy during that cycle. As discussed in more detail later (please see “Efficacy of the LNG-IUS”), an IUD may be safely inserted immediately after uncomplicated spontaneous or induced abortions, including those that are performed during the second trimester. Expulsion, pregnancy, and removal rates after first-trimester termination are comparable to those found in studies in which the device was inserted during or immediately after menses. The shorter the period of gestation at the time of spontaneous or induced abortion, the lower the event rates following IUD insertion. The LNG-IUS can also be safely inserted immediately after childbirth, but there are associated risks with insertion during the immediate postpartum period. If an IUD is inserted within the first 48 hours after childbirth, the risk of expulsion is greater than interval insertion, although the risk of expulsion is somewhat lower if the IUD is inserted within 10 minutes of the delivery of the placenta. If an IUD is inserted 48 hours and up to four weeks after childbirth, the risk of perforation is greater. According to the WHO medical eligibility criteria for contraceptive use, there are no restrictions for inserting an IUD four or more weeks after childbirth; however, no distinction is made between a vaginal or caesarean delivery. The type of delivery is important, because a cesarean section can distort the uterine anatomy. If the distortion is found to be significant during a bimanual examination, then inserting the IUD may be aided by uterine ultrasonography. 17 Adhering to the proper insertion method and using a sterile or “no-touch” technique are sufficient to decrease the risk of infection, uterine perforation, and expulsion. Data indicate that antibiotic prophylaxis does not prevent the rare occurrence of infections acquired at the time of insertion. A Cochrane review performed to assess the effectiveness of prophylactic antibiotic administration before IUD insertion in reducing IUD-related complications and discontinuations within three months of insertion showed little benefit from prophylactic antibiotics on early removal or occurrence of PID following insertion. 18 Prophylactic antibiotics are therefore not recommended prior to insertion of the IUD. Pain may occur during the insertion procedure that may result from a vasovagal reaction, the need for cervical dilation, difficulty with insertion, and uterine perforation. While some clinicians administer ibuprofen or another non-steroidal anti-inflammatory drug (NSAID) to control such pain, a recent study of 2,019 first-time IUD users found that ibuprofen had no significant impact on reducing pain. 19 In a comparative study conducted in the United Kingdom, 102 women presenting for IUD insertion were randomly assigned to 1 of 3 groups: no pretreatment, pretreatment with 2% lignocaine gel (Instillagel) applied to the cervical canal, and pretreatment with an inert gel. Pain scores for women in the nontreatment groups were significantly higher (p < 0.025) than those in the lignocaine group. These results suggest that a local lignocaine gel has the potential for reducing the pain and discomfort experienced during IUD insertion. 20 In an attempt to reduce cervical resistance, a randomized, controlled trial was undertaken among 80 nulliparous women randomly assigned to receive sublingually 400 μg misoprostol and 100 mg diclofenac or 100 mg diclofenac alone one hour prior to IUD insertion. Following treatment with misoprostol, insertion was significantly easier with fewer difficult and failed attempts at insertions than in the control group. However, pain scores, estimated using a visual analogue scale (VAS; 1–10), were not different between the groups. 21 Therefore, there may be some benefit to use of misoprostol for cervical ripening prior to LNG-IUS insertion as well as local anesthetic use for pain control, while NSAIDs appear to be of little benefit.

The levonorgestrel-releasing intrauterine system (LNG-IUS) has been available in Europe since 1990 and the United States since 2000. It is marketed under the name Mirena ® (Bayer Schering Pharma, Berlin, Germany). The LNG-IUS provides highly effective contraception for up to five years, with potential for approval for up to seven years in the near future. The mechanisms of action of the LNG-IUS are similar to that of levonorgestrel implants or levonorgestrel-containing mini-pills, although it accomplishes these effects with much lower peak serum levels than other progestin-containing contraceptives (0.1–0.4 ng/ml vs 1.7–15.2 ng/ml with combined and progestin-only oral contraceptives, respectively, and 5.4 ng/ml for combined vaginal preparations). 1 The LNG-IUS is a T-shaped device composed of a cylinder containing 52 mg of LNG covered by a rate-controlling membrane which serves to regulate the rate of hormonal release ( Figure 1 ). 3 Initially, 20 μg of levonorgestrel is released every 24 hours from this polymer cylinder. This decreases to 11 μg every 24 hours by the end of five years, with an average release rate of 14 μg per day over the life of the IUS. Levonorgestrel, a highly potent second generation progestin, thickens cervical mucus and suppresses endometrial proliferation (preventing decidualization of the stroma). This creates a hostile environment for sperm survival, inhibiting motility and capacitation with the net effect combining to prevent fertilization. 1 The LNG-IUS also produces endometrial thinning with fragile superficial vessels which, in the unlikely event of fertilization, may prevent implantation. The low serum levels of absorbed progestin are below the threshold for inhibition of ovulation, so that most women with the LNG-IUS continue to ovulate regularly. 1 As a result of these various contraceptive actions, the efficacy rate of the LNG-IUS is high, with only 0.1% of women experiencing an unintended pregnancy within the first year of typical use. In four clinical studies, representing more than 10,000 woman-years of use, the average Pearl Index (a statistical estimation of the number of unintended pregnancies in 100 woman-years of exposure to a contraceptive method) was 0.1. 4

Intrauterine contraceptive devices are appropriate contraceptive options for women who desire a very convenient, long-term contraceptive or who are considering sterilization. While there are contraindications to utilization of an IUD, they are relatively few compared to some other methods. Some contraindications, for example pregnancy, are absolute and undisputed. However, other contraindications, such as patients at increased risk of sexually transmitted infections or nulliparous women, are more controversial. The recommendations regarding these issues vary based on which specific organization is consulted ( Table 1 ). One of the most contentious contraindications to IUD use relates to patients who are perceived to be at increased risk for sexually transmitted infections (STIs). This concern likely dates back to the experience with the Dalkon Shield in the 1970s. The Dalkon Shield IUD was associated with an unacceptably high risk of pelvic inflammatory disease, tubal infertility and septic abortion, to a significant extent secondary to its braided polyfilament tail. 5 It was subsequently removed from the market, and is thought by many to be responsible for the initial significant decline and continued slow increase in numbers of patients interested in using IUDs for contraception. After the dismal experience with the Dalkon Shield physicians understandably were, and despite current reassuring data often continue to be, hesitant to prescribe this for unmarried women or patients with multiple sexual partners because of the persistent concern over serious pelvic infections and implications for future fertility. However, numerous studies on the currently available versions of the IUD have shown that, overall, IUD users are no more likely to suffer from PID than control women. One model estimated that the risk of clinical PID secondary to IUD use is only 0.15%. 6 Another recent study evaluating a high-risk population in West Africa showed the risk of PID related to IUD insertion to be only 0.075%. 7 One randomized trial (with data collection in the early 1980s) showed a decreased rate of PID among LNG-IUS users compared to users of a copper-containing IUD. 8 The IUD may, therefore, be a reasonable option for women who may be at risk for sexually transmitted infections. This is especially the case where prevalence in the community as a whole may be high but individual risk may be low (ie, no current or recent history of an STI or the patient uses condoms regularly if she has more than one partner). Morrison and colleagues developed a checklist to help with risk assessment for STIs in candidates for IUD insertion which may be helpful with management decisions (Appendix A). 9 Of note, all women must still be counseled on safe sex practices, and condom use must be encouraged. The appropriateness of IUD use in nulliparous women has also been controversial because of a variety of concerns specific to this group. First, because of the smaller size of the uterus it was thought that there might be an increased expulsion rate. In addition, because of the more stenotic nulliparous os, the device may cause increased pain with insertion. There has also been concern over a possible increased rate of infertility in these women, again secondary to the concern over increased rates of PID with the IUD and the thought that nulliparous women are less likely to be in stable, monogamous relationships. However, these concerns are not supported by recent studies. A retrospective cohort study of 129 nulliparous and 332 parous women showed that rates of expulsion for the levonorgestrel-releasing IUDs were 0% to 0.2% per year, and that nulliparous women did not show more complications than parous women. 10 A comparative cross-sectional study of 227 nulliparous women and 2,080 parous women showed similar reasons for removal between the two groups: 49.1% vs 48.2% because of device expiration, 21.7% vs 15.4% desired planned pregnancy, 4.8% vs 6.3% were secondary to accidental pregnancy, 1.6% vs 1.1% because of PID. There were actually decreased rates of expulsion in the nulliparous group (1.6% vs 5.1%), and slightly increased rates of pain/bleeding (14.5% vs 10.2%). 11 According to the medical eligibility criteria for contraceptive use developed by the World Health Organization (WHO), nulliparity is a condition for which the advantages of using intrauterine contraception generally outweigh the theoretical or proven risks (risk category 2. For the full definition of WHO risk categories please see Appendix B). 12 The intrauterine device is, therefore, generally considered to be a safe and effective contraceptive which should be offered to nulliparous women as part of their family planning counseling. There are no WHO restrictions for using IUDs among women who have uncomplicated valvular disease. When valvular heart disease is complicated by pulmonary hypertension, atrial fibrillation, or subacute bacterial endocarditis, prophylactic antibiotics should be administered to women with these conditions before an IUD is inserted to prevent endocarditis. The WHO provides highly useful guidelines regarding the appropriateness of the LNG-IUS for women with a variety of medical conditions (see Appendix B for the definition of WHO risk categories for eligibility for use of an intrauterine device). According to the WHO, women who have any degree of hypertension can use a LNG-IUS if no other risk factors for cardiovascular disease are present. When multiple risk factors do exist, the risk of cardiovascular disease may increase substantially. For this reason, because progestins have been shown in clinical trials to influence lipid metabolism, 13 there has been concern, albeit theoretical, about the effect of levonorgestrel-releasing IUDs on lipid levels in women with a history of heart disease or stroke. 12 However, these concerns about the effect of the LNG-IUS on lipid profiles have not been borne out. A recent study of 48 patients who had the LNG-IUS inserted for menorrhagia showed no significant change from baseline in mean total cholesterol, triglycerides, low-density lipoprotein (LDL) or very low density lipoprotein (VLDL) at the end of one year. 14 Another study of 92 women with the LNG-IUS inserted for menorrhagia showed a marginally significant reduction in total cholesterol from baseline, a reduction in HDL at 6 months which reverted to baseline by one year, while triglycerides, LDL, apolipoprotien A1 and apolipoprotein B remained stable over the course of 18 months. 15 Therefore, though the WHO recommendations warn of a theoretical effect of the LNG-IUS on lipid levels and caution its use in women with multiple cardiac risk factors, studies do not show this to be a significant problem associated with the LNG-IUS. Progestins are thought to increase the risk for deep vein thromboses (DVTs) and pulmonary emboli, although this increase is substantially less with a levonorgestrel-releasing IUS than with combined oral contraceptives. Since risks associated with the LNG-IUS are even greater for women who have a current DVT, the WHO states that the risks of using the LNG-IUS in such a situation usually outweigh the benefits (risk category 3) until they are established on anticoagulant therapy (risk category 2). 12 , 16 Among women who are infected with the human immunodeficiency virus (HIV) or are at risk of becoming infected with this pathogen, there is no known interaction between antiretroviral therapy and LNG-IUS use. The WHO, however, classifies insertion of a LNG-IUS in a woman with acquired immunodeficiency syndrome (AIDS) as a category 3 risk and continuation of the IUD as a category 2 risk unless the patient is clinically well and on antiretroviral therapy. In this case, insertion of an IUS is then considered to provide benefits that generally outweigh any theoretical or proven risk. Although IUD use is not associated with an increased risk of HIV transmission to sexual partners, these women should be counseled about safe sex practices and condom use must be encouraged. 12

The levonorgestrel-releasing intrauterine system is a safe, effective and acceptable form of contraception being used by over 150 million women worldwide. It also has a multitude of noncontraceptive benefits including improvement in menorrhagia, decreased pelvic pain associated with endometriosis and adenomyosis, as well as a treatment for endometrial hyperplasia, an alternative to hysterectomy and an adjunct to estrogen replacement therapy. Overall the device is very well tolerated and patient satisfaction is quite high when the proper education regarding possible side effects has been provided. Despite all of the obvious benefits of the LNG-IUS, utilization rates remain quite low in the developed countries, and especially the United States. This is largely attributed to the negative history of the Dalkon Shield intrauterine device in the 1970s, which continues to negatively influence the opinions of both patients and health care providers. In addition, cost remains a barrier in less developed countries where demand might be high but the device is currently unaffordable in the public sector. Recently, with a concerted effort to train residents and providers in IUD and IUS use, a reversal of trends has been noted and the IUS is rapidly gaining market share in the US. This safe, acceptable and highly effective method will likely contribute significantly to the contraceptive method mix in the coming years.

Like oral contraceptives, the levonorgestrel-releasing intrauterine system confers important noncontraceptive health benefits. The Mirena ® LNG-IUS can be effective in treating a variety of gynecological disorders including menorrhagia, dysmenorrhea, pain associated with endometriosis, anemia, endometrial hyperplasia and can be used as an alternative to hysterectomy for women with bleeding problems as well as an adjunct to estrogen replacement therapy.