Abstract
Intrauterine contraceptives (IUC) are the most prevalent contraceptives worldwide, the two most popular models being the TCu380A intrauterine device and the levonorgestrel-releasing intrauterine system (LNG-IUS) [l]. These methods, together with subdermal implants, are considered long-acting reversible contraceptive methods or ‘forgettable contraceptives’ [2] since they provide contraceptive effectiveness for at least 3 years, with the TCu380A intrauterine device and the LNG-IUS offering contraceptive protection for 10 and 5 years, respectively [3,4]. Although both IUCs are highly effective and safe, with few adverse effects, some concerns remain, principally regarding their use by nulliparous and nulligravida women and with respect to expulsion rates [5]. Misgivings regarding their use by nulliparous and nulligravida women resulted from the myth that insertion could prove more difficult or painful in these women, while concern with respect to high expulsion rates is due to the concept that the uterine cavity of nulliparous and nulligravida women is smaller than that of parous women [6–8].
A difficult placement procedure could result in failure to insert the device or may limit its use, since many women may consider placement painful [9]. Higher expulsion rates are a concern, since unperceived expulsion could result in an unplanned or undesirable pregnancy. Nevertheless, the debate continues regarding which characteristics in women would increase the likelihood of expulsion or a difficult placement.
Only one LNG-IUS (Mirena®, Bayer, Turku, Finland) is currently available on the market. This device releases 20 µg/day of the steroid and has been approved in over 100 countries. Indications for its use vary from country to country but include contraception, the treatment of heavy menstrual bleeding (HMB) and endometrial protection during postmenopausal estrogen treatment [10]. Nevertheless, some professionals call for smaller IUCs [5], arguing that current devices, including the LNG-IUS, are inappropriate for nulligravidas. Furthermore, one report showed that the length of the endometrial cavity, as assessed by ultrasound, was <32 mm in approximately 33% of the women evaluated [11], and the cervical canal is generally narrower in nulligravida women compared with parous women, albeit it is on debate [12,13].
Gemzell-Danielsson et al. compared the efficacy, bleeding profile and safety of two low-dose LNG-IUSs and the currently available LNG-IUS, based on the concept that an LNG-IUS that releases a lower dose of the steroid could reduce systemic exposure to LNG, minimizing progestin-related side effects and rendering placement simpler and less painful for women [14].
Methods
Two low-dose LNG-IUSs were developed using the same T-frame. In vitro, these devices initially release 12 µg/day (LNG-IUS12) and 16 µg/day (LNG-IUS16) and can be used for up to 3 years. The current LNG-IUS measures 32 × 32 mm and its inserter measures 4.75 mm in diameter. However, the two new devices measure 28 × 28 mm, with an inserter of 3.80 mm in diameter. The authors conducted an open-label, randomized, three-arm study at 37 centers in Finland, Sweden, Norway, Hungary and the UK. The study was insufficiently powered to demonstrate noninferiority.
Sexually active nulliparous and parous women aged 21–40 years with an appropriately sized uterine cavity were included. Exclusion criteria consisted of: breastfeeding, <12 weeks since previous delivery or abortion, HMB, history of ectopic pregnancy, current/recurrent pelvic inflammatory disease and breast cancer. Although the women were blinded with respect to randomization, the investigators knew which device was being inserted because of their different dimensions and the size of the reservoir. The placement technique was the same for all three devices. The providers evaluated ease of placement and removal, and reported whether cervical dilatation, local anesthesia or pain medication were required. Return visits occurred at 1 month and every 6 months thereafter for up to 3 years. The volunteers completed vaginal bleeding diaries daily. The Pearl Index and 95% CIs were used to evaluate contraceptive efficacy, while bleeding patterns were analyzed using 90-day reference periods [15].
Results
Overall, 742 women were randomized and 738 (99.5%) received an LNG-IUS: LNG-IUS12, n = 239/240; LNG-IUS16, n = 245/246; and LNG-IUS20, n = 254/256. Approximately 70% of the women in the three groups completed the 3-year study, with approximately 600 women/years of exposure in each group. Pearl Indices (95% CI) of 0.17 (0.0–0.93), 0.82 (0.27–1.92) and 0.00 (0.0–0.59) reflect the contraceptive efficacy of the LNG-IUS12, LNG-IUS16 and LNG-IUS20, respectively. The cumulative 3-year pregnancy rates, calculated by Kaplan–Meier analysis, were 0.005, 0.025 and 0.000 for the LNG-IUS12, LNG-IUS16 and LNG-IUS20, respectively. Three of the six pregnancies that occurred were ectopic. Expulsions occurred in all three groups: LNG-IUS12 (n = 1/239; 0.4%), LNG-IUS16 (n = 5/245; 2.0%) and LNG-IUS20 (n = 4/256; 1.6%). No uterine perforations occurred.
The number of bleeding and spotting days decreased in all three groups, with no significant differences between them; however, the greatest reduction occurred in the first 90-day period following placement. Amenorrhea increased from 2.7, 6.1 and 5.9% in the second 90-day reference period to 12.7, 18.9 and 23.6% in the final 90-day reference period in the LNG-IUS12, LNG-IUS16 and LNG-IUS20 groups, respectively (p = 0.012 for LNG-IUS12 vs LNG-IUS20 in the final reference period). Adverse events occurred in a similar fashion in all three groups, the most prevalent being acne (~25%), breast discomfort (~20%), headache (~14%), abdominal swelling (~14%), mood changes (~12%) and ovarian cysts (~12%). One case of pelvic inflammatory disease associated with the placement procedure occurred in the LNG-IUS16 group. Dysmenorrhea was reported by approximately 50% of women at baseline, while at the end of the study only approximately 18% had this complaint, with no significant differences between the groups.
Placement was successful at first attempt in 98.5%. Cervical dilators were used in only 43 out of 738 women (5.8%), more commonly in the LNG-IUS20 group compared with the LNG-IUS12/LNG-IUS16 group (9.4 vs 3.9%; p = 0.004). Providers rated placement as ‘simple’ in 94.0% of cases when the two low-dose devices were used compared with 86.2% with the LNG-IUS20 (p < 0.001 for LNG-IUS12/LNG-IUS16 vs LNG-IUS20). Placement failed in one case each with the LNG-IUS12 and LNG-IUS16 and in two cases with the LNG-IUS20. The women described placement as significantly less painful with the new devices compared with the currently available device (p < 0.001 for LNG-IUS12 vs LNG-IUS20; p < 0.01 for LNG-IUS16 vs LNG-IUS20.)
Significance of the results
It is common knowledge in medicine that a reduced dose of any drug would be better than a higher dose if effectiveness were similar. Consequently, the two new LNG-IUS devices that release lower doses of LNG are to be welcomed, principally since a previous publication reported that an intrauterine LNG release of 10–30 µg/day induced similar endometrial changes [16]. The most important finding of this study was that contraceptive failure with the two new LNG-IUS devices was as low as that of the currently available LNG-IUS. Nevertheless, more pregnancies occurred with the LNG-IUS16 compared with the other two devices, although this may have occurred merely by chance. Otherwise, the clinical performance of the two new models was similar to that of the current device, including the reduction in menstrual bleeding, suggesting that the lower doses of LNG delivered by the two new devices may also be adequate for the treatment of HMB or as endometrial protection during postmenopausal estrogen therapy.
It is reasonable to speculate that placement may prove less difficult with smaller devices and narrower inserters compared with the currently available model. Nevertheless, these results show that cervical dilators were required in 5.8% of cases (9.4% with the LNG-IUS20 vs 3.9% with the LNG-IUS12/LNG-IUS16) and placement failed in 0.5% of cases. Unfortunately, the authors did not differentiate between nulligravida and parous women. A recent study conducted with the LNG-IUS20 reported that cervical dilators were required in only 7.7 and 3.1% of nulligravida and parous women, respectively, while placement failed in only 0.6 and 0.4%, respectively [13]. Nevertheless, it should be emphasized that women allocated to the new smaller devices reported less pain at placement compared with acceptors of the LNG-IUS20 device, probably owing to the smaller diameter of the inserter. This is an important finding because fear of pain at IUC insertion is a reason for some women to refuse IUCs [5,9] and there is no effective treatment for pain before, during or after IUC insertion [17].
Despite the excellent quality of this study, several limitations exist. The sample size was inadequate for a noninferiority or superiority trial comparing two new devices with the current one, probably because this was a Phase II, dose–response study. Furthermore, the reduction in menstrual blood loss could have been affected by the dropout rates of almost 28% in the three groups and, as the authors themselves recognize, the women who discontinued the study prematurely may have been those with breakthrough bleeding, a side effect described previously [18]. The occurrence of adverse events was higher than expected; however, it is important to recall that the women were asked specifically about each symptom at every visit, which introduces a bias regarding answers.
Another limitation refers to the few nulliparous women participating in the trial (20%). Furthermore, it is unclear how many of the women in the sample were nulligravida. The literature is conflicting in this respect with some studies using the term nulliparous as synonymous with nulligravida even when the woman's previous pregnancy ended in abortion or an elective cesarean section [19]. This difference is important, since the characteristics of the cervix may be different in nulligravida, nulliparous and parous women.
Future perspective
IUCs are used worldwide; however, their use varies between countries. In addition to the high contraceptive efficacy of the LNG-IUS, the device offers other benefits such as the treatment of HMB and postmenopausal endometrial protection [10,20]. Nevertheless, some misgivings persist among both women and providers with respect to the painful placement procedure and progestin-related side effects. The development of new LNG-IUSs that release lower doses of the steroid are smaller in length and width, and use a narrower inserter may help increase the use of this excellent contraceptive method. The final results of the large Phase III trial currently being conducted with the same three devices described in this article will provide further insight.
Executive summary
Intrauterine contraceptives are the most prevalent contraceptives worldwide, the two most popular models being the TCu380A intrauterine device and the levonorgestrel-releasing intrauterine system (LNG-IUS).
Two smaller, low-dose LNG-IUS that release 12 µg/day (LNG-IUS12) and 16 µg/day (LNG-IUS16) were developed and compared with the currently available LNG-IUS20.
Contraceptive efficacy was very high, expulsions were few and no differences were found in the number of bleeding and spotting days, amenorrhea or adverse events between the three groups. Dysmenorrhea decreased in a similar fashion in all three groups.
Placement was considered ‘simple’ in 94.0% of cases in the two low-dose LNG-IUS groups compared with 86.2% of cases in the LNG-IUS20 group. Placements failed in few cases. The women reported that placement was significantly less painful with the two new devices compared with the currently available device.
Footnotes
The authors declare that they have received financial support for research purposes from the WHO, Family Health International, The Population Council, Bayer and MSD. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.
No writing assistance was utilized in the production of this manuscript.