- Email: [email protected]
Bleeding patterns with the levonorgestrel-releasing intrauterine system when used for heavy menstrual bleeding in women without structural pelvic pathology: a pooled analysis of randomized controlled studies
Contraception 87 (2013) 107 – 112
Original research article
Bleeding patterns with the levonorgestrel-releasing intrauterine system when used for heavy menstrual bleeding in women without structural pelvic pathology: a pooled analysis of randomized controlled studies☆ Jeffrey Jensen a,⁎, Diana Mansour b , Eeva Lukkari-Lax c , Pirjo Inki d , Karin Burock e , Ian S. Fraser f a
Department of Obstetrics and Gynecology, Oregon Health and Science University, Portland, 97239 OR, USA b New Croft Centre, Newcastle upon Tyne, NE1 6ND, UK c Bayer Oy, 02151 Espoo, Finland d Bayer HealthCare, 20101 Turku, Finland e Bayer HealthCare, 42096 Wuppertal, Germany f University of Sydney, Sydney, New South Wales 2006, Australia Received 14 June 2012; revised 13 August 2012; accepted 20 August 2012
Abstract Background: The study was conducted to characterize the changes in bleeding pattern over time in women receiving the levonorgestrelreleasing intrauterine system (LNG-IUS) for heavy menstrual bleeding (HMB). The reduction in menstrual blood loss volume has been well documented elsewhere. Study Design: Post hoc pooled analysis of the impact of the LNG-IUS on bleeding patterns in four comparator studies of medical and surgical treatment options for HMB. We enrolled women aged ≥ 18 years with HMB without organic pathology. The change in the number of bleeding and spotting (B/S) days and bleeding patterns was assessed over the duration of the studies pooled. Results: One hundred and sixty-three women received the LNG-IUS in randomized trials. Relative to pretreatment baseline, there was a transient increase in the mean number of bleeding days in the first month of treatment, which returned to baseline by the second month and declined thereafter. Although the number of spotting days also increased during the first month of treatment, these declined with continued use but remained elevated relative to baseline during the first year of treatment. Conclusion: In women with HMB, the LNG-IUS is associated with an initial increase in number of B/S days that improve over time. © 2013 Elsevier Inc. All rights reserved. Keywords: Amenorrhea; Contraception; Heavy menstrual bleeding; Levonorgestrel-releasing intrauterine system (LNG-IUS)
☆
Conflict of Interest: Jeffrey Jensen has received payments for consulting and giving talks for Bayer HealthCare, a company that may have a commercial interest in the results of this research and technology. This potential conflict of interest has been reviewed and managed by Oregon Health & Science University. He is also a consultant for Merck, Agile Pharmaceuticals and HRA Pharma and has received research funding from Bayer HealthCare, Medicines360, Abbott, Warner Chilcott, the Population Council and the National Institutes of Health (NIH). Diana Mansour has received support to give talks, undertake research and attend clinical meetings and scientific advisory boards for Bayer HealthCare, HRA Pharma, Consilient Healthcare, Astellas, Vifor Pharmaceuticals and Merck. Ian Fraser is a consultant and speaker for Bayer HealthCare, Merck/MSD, Daiichi Sankyo Pharmaceuticals and Vifor Pharmaceuticals and has received research support from the NIH, the Australian National Health and Medical Research Council, the Population Council, Bayer HealthCare and Schering–Plough. Eeva Lukkari-Lax is a current employee of Bayer Oy, Espoo, Finland. Pirjo Inki is a current employee of Bayer HealthCare, Turku, Finland. Karin Burock was an employee of Bayer HealthCare, Wuppertal, Germany. ⁎ Corresponding author. Tel.: + 1 503 494 5113; fax: + 1 503 494 5083. E-mail address: [email protected] (J. Jensen). 0010-7824/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.contraception.2012.08.022
108
J. Jensen et al. / Contraception 87 (2013) 107–112
1. Introduction The contraceptive reliability of the levonorgestrel-releasing intrauterine system (LNG-IUS; Mirena®, Bayer HealthCare, Berlin, Germany) is well established, with efficacy comparable to female sterilization [1–3], but with the added advantage of rapid return of fertility upon removal. Women using the LNG-IUS for contraception experience a change in their menstrual bleeding pattern [1,4]. This change is characterized by an initial increase in the number of bleeding and spotting (B/S) days during the first few months following insertion, after which a reduction in the number of B/S days occurs with the proportion of women with amenorrhea or infrequent bleeding increasing to 16% and 57%, respectively, over the first year of LNG-IUS use [5]. Nearly 30% of women using the LNG-IUS have amenorrhea by the fourth year, a rate that is further increased with subsequent consecutive LNG-IUS use [6]. The profound reduction in blood loss achieved with the LNG-IUS makes it an effective option for the treatment of heavy menstrual bleeding (HMB). HMB, objectively defined, for research purposes, as blood loss ≥ 80 mL per menstrual cycle [7], has a prevalence of up to 14% in women of reproductive age from the general population in studies where menstrual blood loss (MBL) was measured using objective assessments such as the alkaline hematin method (National Institute for Clinical Excellence 2007 guidelines) [8]. In studies based solely on women's perceptions of the heaviness of their bleeding episodes, prevalence rates as high as 50% have been reported [9]. The efficacy of the LNG-IUS in reducing MBL in women with HMB without recognizable structural pelvic pathology is well established; median reductions of MBL have ranged from 83% to 97% over 3–12 months of treatment in prospective studies using objective assessments of MBL [10–12]. Similar effects are seen with less rigorous methodology; mean reductions of 71% to 97% in bleeding
scores have been observed in randomized controlled studies using the subjective pictorial blood assessment chart (PBAC) to quantify MBL [13]. Collectively, these studies demonstrate the efficacy and tolerability of the LNG-IUS in reducing the volume of uterine bleeding in women with HMB [14], but in contrast to women using the LNG-IUS for contraception, details of the number of bleeding and/or spotting days observed with long-term treatment have not been reported. The purpose of this current study, therefore, was to characterize the bleeding pattern in women using the LNGIUS solely for the treatment of HMB. We pooled data from four randomized, comparative studies [including one unpublished study (data on file; Bayer HealthCare)] where similar inclusion and exclusion criteria were applied and bleeding diary data were collected [11,12,15,16] with the aim to assess, for the first time, the number of bleeding and/ or spotting days experienced by women receiving treatment with the LNG-IUS for HMB and to better define the bleeding pattern among a larger population. These studies focused primarily on the changing volume of MBL over the first 12 months of LNG-IUS use. 2. Methods 2.1. Study design and participants In this post hoc pooled analysis, we focused on analyzing the number of bleeding and spotting days and bleeding patterns from four randomized studies that compared the efficacy of the LNG-IUS with that of medroxyprogesterone acetate (MPA) [11], tranexamic acid (unpublished observations), mefenamic acid [12] and transcervical endometrial resection [15,16] for the treatment of HMB without organic pathology (Table 1). The primary focus of our analysis was on bleeding pattern (which is not adequately addressed in the published literature in women with HMB).
Table 1 Characteristics of randomized controlled trials pooled Study
Location/Date
Treatments
Duration
Kaunitz et al. 2010 (ClinicalTrials.gov: NCT00360490) [11]
Fifty-five centers in the United States, Canada and Brazil between July 2006 and June 2008
6 cycles (30-day cycles) of treatment
Unpublished study
Single center in Sweden between 1995 and 1998
Reid et al. 2005 [12]
Two centers in the UK between 1996 and 2003
Istre et al. 2001 (for year 1 data) [15] Rauramo et al. 2004 (for data through year 3) [16]
Single center in Norway between 1993 and 1998
LNG-IUS (n= 82) Oral MPA acetate (10 mg once daily for 10 days beginning on Day 16 of each cycle) (n= 83) LNG-IUS (n= 29) Oral tranexamic acid (2× 500 mg four time daily for a maximum of 5 days from the onset of menstruation) (n= 30) LNG-IUS (n= 25) Oral mefenamic acid (500 mg three times daily for the first 4 days of each menstrual cycle) (n= 26) LNG-IUS (n= 30) Transcervical endometrial resection (n= 29)
12 cycles of treatment
Six cycles of treatment
Initially planned for 12 months of treatment but extended to 36 months before first patient was enrolled
J. Jensen et al. / Contraception 87 (2013) 107–112
All four studies recruited women aged ≥ 18 years with HMB without organic pathology. The main criteria for inclusion were as follows: confirmed HMB at baseline, determined either by a PBAC score of ≥ 100 [17] (confirmed HMB at baseline was determined by a PBAC score of N 75 in one study; however, only two of the patients with baseline PBAC scores b 100 contributed to the MBL data during treatment) or the alkaline hematin method (MBL ≥ 80 mL) per cycle [7]; follicle-stimulating hormone level ≤ 30 IU/L (two studies) [11,15,16]; regular menstrual cycles of 21–35 days with no significant intermenstrual B/S and normal or moderately enlarged uterine cavities (a sound measure including the cervical canal of ≤ 10 cm in length). Although nulliparous women could be recruited in three of the pooled studies, one study specifically enrolled only parous women [11]. The main exclusion criteria for all four studies included undiagnosed abnormal uterine bleeding (changes in menstrual regularity), submucous fibroids distorting the uterine cavity, uterine fibroids ≥ 4 cm in diameter (or total volume ≥ 5 cm 3; small intramural and subserous fibroids were permitted), pelvic inflammatory disease in 6 months prior to screening, genital infection at the time of screening and known anomalies of the uterine cavity or cervix. A body mass index (BMI) of N 35 kg/m 2 was an exclusion criterion in the study comparing the LNG-IUS with MPA [11]. 2.2. Assessments and analyses Only data pertaining to women using the LNG-IUS were assessed in the current analysis. The primary efficacy variable of all four studies was the reduction in MBL. MBL was objectively measured using the alkaline hematin method in three studies [11,12] (including one unpublished study), whereas the fourth study [15,16] used the PBAC to subjectively quantify blood loss. Full study details of the three individual studies pooled can be found elsewhere [11,12,15,16]. Briefly, in the unpublished study included in the current pooled analysis, a total of 92 women were screened, of whom 59 were randomized to the LNG-IUS (n= 29) or oral tranexamic acid (2×500 mg four times daily for a maximum of 5 days from the onset of menstruation) (n= 30) for 12 cycles of treatment. One woman randomized to receive the LNG-IUS did not receive treatment. In both treatment groups, median MBL decreased from baseline to Cycle 12; from 168.5 to 4.5 mL and from 140.0 to 71.2 mL in the LNG-IUS and oral tranexamic acid treatment groups, respectively. Overall, 20 women treated with the LNG-IUS completed the study, and a total of 9 women discontinued the study due to adverse events. In all four studies, the women were provided with identical diary cards to record their uterine bleeding as either bleeding (bloody vaginal discharge that required the use of pads or tampons) or spotting (vaginal discharge that required only light sanitary protection such as panty liners) on a daily basis. B/S days were analyzed for the intention-to-treat (ITT)
109
population using a 30-day reference period, and bleeding patterns were assessed as proposed by the World Health Organization [18]. Baseline diary data before LNG-IUS placement were collected for 30–90 days. In addition, the women were classified by bleeding pattern — prolonged (i.e., length of a single B/S episode 14 days or more per 90-day reference period), frequent (more than five B/S episodes per 90-day reference period), infrequent (i.e., less than three B/S episodes per 90-day reference period) or amenorrhea (i.e., absence of B/S for at least 90 consecutive days) [18] — based on diary card recording over up to 3 years of treatment. All B/S variables and bleeding patterns were summarized using descriptive statistics.
3. Results A total of 163 women formed the ITT population on whom this pooled analysis is based. Overall, 37 women (22.7%) prematurely discontinued treatment with the LNGIUS across their respective studies. Of these, 10 (6.1%) and 12 (7.4%) prematurely discontinued their planned treatment duration due to bleeding problems and expulsion (either partial or full), respectively. Because only a small number of women (n= 22) were followed up beyond 1 year, the analysis was restricted to this time period. The baseline characteristics of this pooled population of women are summarized in Table 2. Parity data were not collected for 32.5% of women; of the remaining women, however, almost all (99.1%; 109/110) of the women had given birth one or more times. The proportion of women with fibroids at baseline was 2.4% (2/83) — fibroid data were not collected at baseline for the study comparing the efficacy of the LNG-IUS with that of MPA [11], but the presence of substantial fibroids or a fibroid that involved the uterine cavity was part of the exclusion criteria of all the individual studies. From the data available for three of the four studies, 11.6% (16/138) of women had received prior treatment for HMB, which included the use of antifibrinolytics, progestins and combined oral contraceptives.
Table 2 Baseline characteristics of women who received the LNG-IUS (n= 163) (pooled ITT population) Age (years) Weight (kg) BMI (kg/m 2) Current smoker (%) Yes No Missing Length of menstrual cycle (days)
39.1 ± 5.3 73.7 ± 13.2 a 26.8 ± 4.5 b 36 (22.1) 99 (60.7) 28 (17.2) 27.4 ± 2.9
All data are presented as mean±standard deviation unless otherwise stated. a n= 161. b n= 160.
110
J. Jensen et al. / Contraception 87 (2013) 107–112
Fig. 1. Change in the mean number of bleeding/spotting (bleeding only and spotting only) days per 30-day reference periods experienced by women with HMB treated with the LNG-IUS over 12 months.
Following LNG-IUS insertion, there was a transient increase in the mean of bleeding days during the first month of treatment (from 7.2 to 9.9 days), which returned to baseline levels by the second month and continued to decline thereafter (Fig. 1). Although the number of mean spotting days also increased during the first month of treatment (from 1.8 to 10.0 days) before gradually declining, this number remained elevated relative to baseline throughout the first year of treatment. At baseline, spotting represented 20% (1.8/9.0) of the total number of B/S days increasing to 50% (10.0/19.9) of B/S days during the first 30-day reference period. Spotting appeared to account for about half of the total number of B/S days throughout the first year of treatment. The bleeding pattern associated with LNG-IUS use for HMB over 12 months of treatment is summarized in Table 3. Consistent with the decrease in the mean number of B/S days with continued use of the LNG-IUS, the proportion of women with amenorrhea increased over the first year of use.
4. Discussion To the authors' knowledge, this is the first study to report the change in the number of bleeding and/or spotting days in women with HMB (in women without structural pelvic pathology) treated with the LNG-IUS. Although a couple of earlier studies have also reported bleeding patterns after Table 3 Bleeding pattern associated with LNG-IUS treatment in women with HMB (expressed as percentage of women experiencing these patterns in 90 days up to 6 and 12 months, respectively) Bleeding pattern (%)
Amenorrhea Infrequent bleeding/spotting Frequent bleeding/spotting Prolonged bleeding/spotting
Duration of LNG-IUS treatment 6 months (n= 143)
1 year (n= 47)
1.7 21.6 11.2 55.6
8.8 23.8 6.3 25.0
insertion of the LNG-IUS in women with HMB [19,20], these used different definitions for the various bleeding patterns making direct comparability among studies difficult. Broadly, these earlier studies in women with HMB are suggestive of a pattern of reduced bleeding days with LNGIUS treatment (i.e., increased amenorrhea rates) up to 3 years [19,20], a pattern that is confirmed in our current analysis. Our results provide guidance to clinicians and patients as they demonstrate that the changes in bleeding and/or spotting days observed following insertion of the LNG-IUS in women with HMB follows a similar pattern to that observed in women using the intrauterine system for contraception [1,4,5,21], except that women with HMB may have more B/S days over the first year of use (see below). A systematic review and meta-analysis of bleeding/ spotting data and bleeding patterns were not possible because there are no universally agreed definitions for these terms, and as such, the many disparate definitions hinder direct comparisons across studies from different groups. The studies included in the current analysis were selected as they were all conducted by a single sponsor (Bayer HealthCare) using the same methodology to assess number of bleeding/spotting days and bleeding patterns (data not previously published) such that complete data sets were available for analysis. Pooling of the measured MBL data to assess the overall reduction in MBL volume was not done as two different MBL measurement techniques were used in the four studies, and the effects of the LNG-IUS on reducing measured MBL in women with HMB are well established in the literature [14]. We have shown in women with HMB that there is an initial increase in mean bleeding days during the first month following LNG-IUS placement that subsides back to baseline levels by the second month and continues to subside further over the first year of treatment. The number of mean spotting days also increases during the first month of treatment, but although this number declines with continued use, it remains elevated relative to pretreatment during the first year of treatment. “Spotting” is usually very limited in
J. Jensen et al. / Contraception 87 (2013) 107–112
normal menstruations. The reduction in the number of bleeding days is accompanied by an increase in the proportion of women with amenorrhea and a concomitant decrease in the proportion of women with prolonged bleeding. Although this study data does not provide information on duration of use more than 1 year, a small number of women (n= 22) were followed for up to 3 years, and a trend toward a further reduction in the mean number of B/S days with continuing use was observed (data not shown). Hence, “bleeding” is converted into spotting or “absence of bleeding” in most of these women. The rate of amenorrhea at 1 year in the current pooled analysis (8.8%; Table 3) appears lower than that observed in women using the LNG-IUS for contraception following postmenstrual (16%) and postabortal (11%) LNG-IUS insertion [5]. Andersson et al. [1] reported that 17% of women using the LNG-IUS for contraception experienced a period of at least 90 days of no B/S during the first year. Over 3 years of follow-up of contraceptive users, Wang et al. [22] reported that approximately 17% of women experienced amenorrhea during the fourth 90-day reference period. In contrast, the earlier studies that reported bleeding patterns after insertion of the LNG-IUS in women with idiopathic HMB suggested much higher amenorrhea rates at 1 year (60% and 63%, respectively) [19,20]. It is unclear why such high amenorrhea rates were reported in these earlier studies; moreover, in one of the studies, the amenorrhea rate was higher than in a comparator group using the LNG-IUS for contraception (63% vs. 44%) [19]. The changes in bleeding pattern achieved with the LNGIUS in women with HMB in the current study appear to be well tolerated, with only 6% of women discontinuing their planned treatment duration due to bleeding problems. In women using the LNG-IUS for contraception, Andersson et al. [1] reported net cumulative discontinuation rates for bleeding problems of 6 per 100 women during the first year of contraceptive use, increasing to 10 per 100 women during the third year. Therefore, despite an apparent higher number of mean B/S days in women using the LNG-IUS for HMB compared with those using it for contraception [1,21], this does not result in an increased rate of discontinuation due to bleeding problems. A likely explanation for this clinically important finding is that the substantial and highly significant reduction in the volume of bleeding observed during the use of the LNG-IUS in women with HMB more than compensates for the small increase in the mean number of spotting days. Indeed, with median reductions in MBL of 95% [11,12] and 97% (unpublished observations) reported at 6 and 12 months, respectively, the four studies show that the LNG-IUS effectively reduces MBL in women with HMB to very low levels [11,12,15,16], with much greater reductions than those achieved with the comparator treatments MPA [11], tranexamic acid (unpublished observations) and mefenamic acid [12]. Another explanation for the similar discontinuation rate due to bleeding problems may be counseling. It has been shown
111
that counseling women about the possibility of bleeding problems increases satisfaction with the LNG-IUS compared with women who are less informed [23]. The current pooled analysis in women with HMB has a number of limitations that need to be considered when assessing its relative generalizability. The number of women who used the LNG-IUS across the four randomized studies pooled was relatively small (n= 163). As such, more longterm studies (N 1 year) assessing the changes in bleeding pattern in women using the LNG-IUS for the treatment of HMB would be required to confirm our observations. Although most of the same inclusion and exclusion criteria applied to all four studies, the exclusion criterion of a BMI of N 35 kg/m 2 applied only to the study by Kaunitz et al. [11]. A BMI of N 35 kg/m 2 was not deemed to be a potential confounding factor for the interpretation of the data because only four women in all four studies fell into this category. In conclusion, the LNG-IUS is an effective and welltolerated treatment option for HMB. Although women using the LNG-IUS for HMB may have more B/S days than those using it for contraception, overall, the changes in the bleeding pattern follow a similar and favorable pattern in both populations, and the discontinuation rates due to bleeding problems do not appear to be higher in the HMB versus the contraceptive population. These data may be useful for counseling women with HMB about the use of the LNG-IUS and the bleeding patterns they may experience. Acknowledgments This study was funded by Bayer HealthCare, Berlin, Germany, the manufacturer of the LNG-IUS. Thomas Schmelter (Bayer HealthCare) provided assistance with programming and statistical analysis. Medical writing services during the preparation of this manuscript were provided by Amy Evans and Richard Glover of inScience Communications, Springer Healthcare. Funding for this editorial assistance was provided by Bayer HealthCare. References [1] Andersson K, Odlind V, Rybo G. Levonorgestrel-releasing and copper-releasing (Nova T) IUDs during five years of use: a randomized comparative trial. Contraception 1994;49:56–72. [2] Backman T, Rauramo I, Huhtala S, Koskenvuo M. Pregnancy during the use of levonorgestrel intrauterine system. Am J Obstet Gynecol 2004;190:50–4. [3] Cox M, Tripp J, Blacksell S. Clinical performance of the levonorgestrel intrauterine system in routine use by the UK Family Planning and Reproductive Health Research Network: 5-year report. J Fam Plann Reprod Health Care 2002;28:73–7. [4] Jensen JT, Nelson AL, Costales AC. Subject and clinician experience with the levonorgestrel-releasing intrauterine system. Contraception 2008;77:22–9. [5] Suvisaari J, Lahteenmaki P. Detailed analysis of menstrual bleeding patterns after postmenstrual and postabortal insertion of a copper IUD or a levonorgestrel-releasing intrauterine system. Contraception 1996; 54:201–8.
112
J. Jensen et al. / Contraception 87 (2013) 107–112
[6] Gemzell-Danielsson K, Inki P, Boubli L, O'Flynn M, Kunz M, Heikinheimo O. Bleeding pattern and safety of consecutive use of the levonorgestrel-releasing intrauterine system (LNG-IUS) — a multicentre prospective study. Hum Reprod 2010;25:354–9. [7] Hallberg L, Nilsson L. Determination of menstrual blood loss. Scand J Clin Lab Invest 1964;16:244–8. [8] National Collaborating Centre for Women's and Children's Health (NCCWCH). National Institute for Clinical Excellence (NICE) guidelines. CG44 Heavy Menstrual Bleeding: full guideline. 24 January; 2007. [9] Fraser IS, Langham S, Uhl-Hochgraeber K. Health-related quality of life and economic burden of abnormal uterine bleeding. Expert Rev Obstet Gynecol 2009;4:179–89. [10] Andersson JK, Rybo G. Levonorgestrel-releasing intrauterine device in the treatment of menorrhagia. Br J Obstet Gynaecol 1990;97:690–4. [11] Kaunitz AM, Bissonnette F, Monteiro I, Lukkari-Lax E, Muysers C, Jensen JT. Levonorgestrel-releasing intrauterine system or medroxyprogesterone for heavy menstrual bleeding: a randomized controlled trial. Obstet Gynecol 2010;116:625–32. [12] Reid PC, Virtanen-Kari S. Randomised comparative trial of the levonorgestrel intrauterine system and mefenamic acid for the treatment of idiopathic menorrhagia: a multiple analysis using total menstrual fluid loss, menstrual blood loss and pictorial blood loss assessment charts. Br J Obstet Gynaecol 2005;112:1121–5. [13] Kaunitz AM, Meredith S, Inki P, Kubba A, Sanchez-Ramos L. Levonorgestrel-releasing intrauterine system and endometrial ablation in heavy menstrual bleeding: a systematic review and meta-analysis. Obstet Gynecol 2009;113:1104–16. [14] Kaunitz AM, Inki P. The levonorgestrel-releasing intrauterine system in heavy menstrual bleeding: a benefit–risk review. Drugs 2012;72: 193–215.
[15] Istre O, Trolle B. Treatment of menorrhagia with the levonorgestrel intrauterine system versus endometrial resection. Fertil Steril 2001;76: 304–9. [16] Rauramo I, Elo I, Istre O. Long-term treatment of menorrhagia with levonorgestrel intrauterine system versus endometrial resection. Obstet Gynecol 2004;104:1314–21. [17] Higham JM, O'Brien PM, Shaw RW. Assessment of menstrual blood loss using a pictorial chart. Br J Obstet Gynaecol 1990;97:734–9. [18] Belsey EM, Machin D, d'Arcangues C. The analysis of vaginal bleeding patterns induced by fertility regulating methods. World Health Organization Special Programme of Research, Development and Research Training in Human Reproduction. Contraception 1986; 34:253–60. [19] Magalhães J, Aldrighi JM, de Lima GR. Uterine volume and menstrual patterns in users of the levonorgestrel-releasing intrauterine system with idiopathic menorrhagia or menorrhagia due to leiomyomas. Contraception 2007;75:193–8. [20] Monteiro I, Bahamondes L, Diaz J, Perrotti M, Petta C. Therapeutic use of levonorgestrel-releasing intrauterine system in women with menorrhagia: a pilot study. Contraception 2002;65:325–8. [21] Gemzell-Danielsson K, Schellschmidt I, Apter D. A randomized, phase II study describing the efficacy, bleeding profile, and safety of two low-dose levonorgestrel-releasing intrauterine contraceptive systems with Mirena. Fertil Steril 2012;97:616–22. [22] Wang SL, Wu SC, Xin XM, Chen JH, Gao J. Three years' experience with levonorgestrel-releasing intrauterine device and Norplant-2 implants: a randomized comparative study. Adv Contracept 1992;8: 105–14. [23] Backman T, Huhtala S, Luoto R, Tuominen J, Rauramo I, Koskenvuo M. Advance information improves user satisfaction with the levonorgestrel intrauterine system. Obstet Gynecol 2002;99:608–13.
Original research article
Bleeding patterns with the levonorgestrel-releasing intrauterine system when used for heavy menstrual bleeding in women without structural pelvic pathology: a pooled analysis of randomized controlled studies☆ Jeffrey Jensen a,⁎, Diana Mansour b , Eeva Lukkari-Lax c , Pirjo Inki d , Karin Burock e , Ian S. Fraser f a
Department of Obstetrics and Gynecology, Oregon Health and Science University, Portland, 97239 OR, USA b New Croft Centre, Newcastle upon Tyne, NE1 6ND, UK c Bayer Oy, 02151 Espoo, Finland d Bayer HealthCare, 20101 Turku, Finland e Bayer HealthCare, 42096 Wuppertal, Germany f University of Sydney, Sydney, New South Wales 2006, Australia Received 14 June 2012; revised 13 August 2012; accepted 20 August 2012
Abstract Background: The study was conducted to characterize the changes in bleeding pattern over time in women receiving the levonorgestrelreleasing intrauterine system (LNG-IUS) for heavy menstrual bleeding (HMB). The reduction in menstrual blood loss volume has been well documented elsewhere. Study Design: Post hoc pooled analysis of the impact of the LNG-IUS on bleeding patterns in four comparator studies of medical and surgical treatment options for HMB. We enrolled women aged ≥ 18 years with HMB without organic pathology. The change in the number of bleeding and spotting (B/S) days and bleeding patterns was assessed over the duration of the studies pooled. Results: One hundred and sixty-three women received the LNG-IUS in randomized trials. Relative to pretreatment baseline, there was a transient increase in the mean number of bleeding days in the first month of treatment, which returned to baseline by the second month and declined thereafter. Although the number of spotting days also increased during the first month of treatment, these declined with continued use but remained elevated relative to baseline during the first year of treatment. Conclusion: In women with HMB, the LNG-IUS is associated with an initial increase in number of B/S days that improve over time. © 2013 Elsevier Inc. All rights reserved. Keywords: Amenorrhea; Contraception; Heavy menstrual bleeding; Levonorgestrel-releasing intrauterine system (LNG-IUS)
☆
Conflict of Interest: Jeffrey Jensen has received payments for consulting and giving talks for Bayer HealthCare, a company that may have a commercial interest in the results of this research and technology. This potential conflict of interest has been reviewed and managed by Oregon Health & Science University. He is also a consultant for Merck, Agile Pharmaceuticals and HRA Pharma and has received research funding from Bayer HealthCare, Medicines360, Abbott, Warner Chilcott, the Population Council and the National Institutes of Health (NIH). Diana Mansour has received support to give talks, undertake research and attend clinical meetings and scientific advisory boards for Bayer HealthCare, HRA Pharma, Consilient Healthcare, Astellas, Vifor Pharmaceuticals and Merck. Ian Fraser is a consultant and speaker for Bayer HealthCare, Merck/MSD, Daiichi Sankyo Pharmaceuticals and Vifor Pharmaceuticals and has received research support from the NIH, the Australian National Health and Medical Research Council, the Population Council, Bayer HealthCare and Schering–Plough. Eeva Lukkari-Lax is a current employee of Bayer Oy, Espoo, Finland. Pirjo Inki is a current employee of Bayer HealthCare, Turku, Finland. Karin Burock was an employee of Bayer HealthCare, Wuppertal, Germany. ⁎ Corresponding author. Tel.: + 1 503 494 5113; fax: + 1 503 494 5083. E-mail address: [email protected] (J. Jensen). 0010-7824/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.contraception.2012.08.022
108
J. Jensen et al. / Contraception 87 (2013) 107–112
1. Introduction The contraceptive reliability of the levonorgestrel-releasing intrauterine system (LNG-IUS; Mirena®, Bayer HealthCare, Berlin, Germany) is well established, with efficacy comparable to female sterilization [1–3], but with the added advantage of rapid return of fertility upon removal. Women using the LNG-IUS for contraception experience a change in their menstrual bleeding pattern [1,4]. This change is characterized by an initial increase in the number of bleeding and spotting (B/S) days during the first few months following insertion, after which a reduction in the number of B/S days occurs with the proportion of women with amenorrhea or infrequent bleeding increasing to 16% and 57%, respectively, over the first year of LNG-IUS use [5]. Nearly 30% of women using the LNG-IUS have amenorrhea by the fourth year, a rate that is further increased with subsequent consecutive LNG-IUS use [6]. The profound reduction in blood loss achieved with the LNG-IUS makes it an effective option for the treatment of heavy menstrual bleeding (HMB). HMB, objectively defined, for research purposes, as blood loss ≥ 80 mL per menstrual cycle [7], has a prevalence of up to 14% in women of reproductive age from the general population in studies where menstrual blood loss (MBL) was measured using objective assessments such as the alkaline hematin method (National Institute for Clinical Excellence 2007 guidelines) [8]. In studies based solely on women's perceptions of the heaviness of their bleeding episodes, prevalence rates as high as 50% have been reported [9]. The efficacy of the LNG-IUS in reducing MBL in women with HMB without recognizable structural pelvic pathology is well established; median reductions of MBL have ranged from 83% to 97% over 3–12 months of treatment in prospective studies using objective assessments of MBL [10–12]. Similar effects are seen with less rigorous methodology; mean reductions of 71% to 97% in bleeding
scores have been observed in randomized controlled studies using the subjective pictorial blood assessment chart (PBAC) to quantify MBL [13]. Collectively, these studies demonstrate the efficacy and tolerability of the LNG-IUS in reducing the volume of uterine bleeding in women with HMB [14], but in contrast to women using the LNG-IUS for contraception, details of the number of bleeding and/or spotting days observed with long-term treatment have not been reported. The purpose of this current study, therefore, was to characterize the bleeding pattern in women using the LNGIUS solely for the treatment of HMB. We pooled data from four randomized, comparative studies [including one unpublished study (data on file; Bayer HealthCare)] where similar inclusion and exclusion criteria were applied and bleeding diary data were collected [11,12,15,16] with the aim to assess, for the first time, the number of bleeding and/ or spotting days experienced by women receiving treatment with the LNG-IUS for HMB and to better define the bleeding pattern among a larger population. These studies focused primarily on the changing volume of MBL over the first 12 months of LNG-IUS use. 2. Methods 2.1. Study design and participants In this post hoc pooled analysis, we focused on analyzing the number of bleeding and spotting days and bleeding patterns from four randomized studies that compared the efficacy of the LNG-IUS with that of medroxyprogesterone acetate (MPA) [11], tranexamic acid (unpublished observations), mefenamic acid [12] and transcervical endometrial resection [15,16] for the treatment of HMB without organic pathology (Table 1). The primary focus of our analysis was on bleeding pattern (which is not adequately addressed in the published literature in women with HMB).
Table 1 Characteristics of randomized controlled trials pooled Study
Location/Date
Treatments
Duration
Kaunitz et al. 2010 (ClinicalTrials.gov: NCT00360490) [11]
Fifty-five centers in the United States, Canada and Brazil between July 2006 and June 2008
6 cycles (30-day cycles) of treatment
Unpublished study
Single center in Sweden between 1995 and 1998
Reid et al. 2005 [12]
Two centers in the UK between 1996 and 2003
Istre et al. 2001 (for year 1 data) [15] Rauramo et al. 2004 (for data through year 3) [16]
Single center in Norway between 1993 and 1998
LNG-IUS (n= 82) Oral MPA acetate (10 mg once daily for 10 days beginning on Day 16 of each cycle) (n= 83) LNG-IUS (n= 29) Oral tranexamic acid (2× 500 mg four time daily for a maximum of 5 days from the onset of menstruation) (n= 30) LNG-IUS (n= 25) Oral mefenamic acid (500 mg three times daily for the first 4 days of each menstrual cycle) (n= 26) LNG-IUS (n= 30) Transcervical endometrial resection (n= 29)
12 cycles of treatment
Six cycles of treatment
Initially planned for 12 months of treatment but extended to 36 months before first patient was enrolled
J. Jensen et al. / Contraception 87 (2013) 107–112
All four studies recruited women aged ≥ 18 years with HMB without organic pathology. The main criteria for inclusion were as follows: confirmed HMB at baseline, determined either by a PBAC score of ≥ 100 [17] (confirmed HMB at baseline was determined by a PBAC score of N 75 in one study; however, only two of the patients with baseline PBAC scores b 100 contributed to the MBL data during treatment) or the alkaline hematin method (MBL ≥ 80 mL) per cycle [7]; follicle-stimulating hormone level ≤ 30 IU/L (two studies) [11,15,16]; regular menstrual cycles of 21–35 days with no significant intermenstrual B/S and normal or moderately enlarged uterine cavities (a sound measure including the cervical canal of ≤ 10 cm in length). Although nulliparous women could be recruited in three of the pooled studies, one study specifically enrolled only parous women [11]. The main exclusion criteria for all four studies included undiagnosed abnormal uterine bleeding (changes in menstrual regularity), submucous fibroids distorting the uterine cavity, uterine fibroids ≥ 4 cm in diameter (or total volume ≥ 5 cm 3; small intramural and subserous fibroids were permitted), pelvic inflammatory disease in 6 months prior to screening, genital infection at the time of screening and known anomalies of the uterine cavity or cervix. A body mass index (BMI) of N 35 kg/m 2 was an exclusion criterion in the study comparing the LNG-IUS with MPA [11]. 2.2. Assessments and analyses Only data pertaining to women using the LNG-IUS were assessed in the current analysis. The primary efficacy variable of all four studies was the reduction in MBL. MBL was objectively measured using the alkaline hematin method in three studies [11,12] (including one unpublished study), whereas the fourth study [15,16] used the PBAC to subjectively quantify blood loss. Full study details of the three individual studies pooled can be found elsewhere [11,12,15,16]. Briefly, in the unpublished study included in the current pooled analysis, a total of 92 women were screened, of whom 59 were randomized to the LNG-IUS (n= 29) or oral tranexamic acid (2×500 mg four times daily for a maximum of 5 days from the onset of menstruation) (n= 30) for 12 cycles of treatment. One woman randomized to receive the LNG-IUS did not receive treatment. In both treatment groups, median MBL decreased from baseline to Cycle 12; from 168.5 to 4.5 mL and from 140.0 to 71.2 mL in the LNG-IUS and oral tranexamic acid treatment groups, respectively. Overall, 20 women treated with the LNG-IUS completed the study, and a total of 9 women discontinued the study due to adverse events. In all four studies, the women were provided with identical diary cards to record their uterine bleeding as either bleeding (bloody vaginal discharge that required the use of pads or tampons) or spotting (vaginal discharge that required only light sanitary protection such as panty liners) on a daily basis. B/S days were analyzed for the intention-to-treat (ITT)
109
population using a 30-day reference period, and bleeding patterns were assessed as proposed by the World Health Organization [18]. Baseline diary data before LNG-IUS placement were collected for 30–90 days. In addition, the women were classified by bleeding pattern — prolonged (i.e., length of a single B/S episode 14 days or more per 90-day reference period), frequent (more than five B/S episodes per 90-day reference period), infrequent (i.e., less than three B/S episodes per 90-day reference period) or amenorrhea (i.e., absence of B/S for at least 90 consecutive days) [18] — based on diary card recording over up to 3 years of treatment. All B/S variables and bleeding patterns were summarized using descriptive statistics.
3. Results A total of 163 women formed the ITT population on whom this pooled analysis is based. Overall, 37 women (22.7%) prematurely discontinued treatment with the LNGIUS across their respective studies. Of these, 10 (6.1%) and 12 (7.4%) prematurely discontinued their planned treatment duration due to bleeding problems and expulsion (either partial or full), respectively. Because only a small number of women (n= 22) were followed up beyond 1 year, the analysis was restricted to this time period. The baseline characteristics of this pooled population of women are summarized in Table 2. Parity data were not collected for 32.5% of women; of the remaining women, however, almost all (99.1%; 109/110) of the women had given birth one or more times. The proportion of women with fibroids at baseline was 2.4% (2/83) — fibroid data were not collected at baseline for the study comparing the efficacy of the LNG-IUS with that of MPA [11], but the presence of substantial fibroids or a fibroid that involved the uterine cavity was part of the exclusion criteria of all the individual studies. From the data available for three of the four studies, 11.6% (16/138) of women had received prior treatment for HMB, which included the use of antifibrinolytics, progestins and combined oral contraceptives.
Table 2 Baseline characteristics of women who received the LNG-IUS (n= 163) (pooled ITT population) Age (years) Weight (kg) BMI (kg/m 2) Current smoker (%) Yes No Missing Length of menstrual cycle (days)
39.1 ± 5.3 73.7 ± 13.2 a 26.8 ± 4.5 b 36 (22.1) 99 (60.7) 28 (17.2) 27.4 ± 2.9
All data are presented as mean±standard deviation unless otherwise stated. a n= 161. b n= 160.
110
J. Jensen et al. / Contraception 87 (2013) 107–112
Fig. 1. Change in the mean number of bleeding/spotting (bleeding only and spotting only) days per 30-day reference periods experienced by women with HMB treated with the LNG-IUS over 12 months.
Following LNG-IUS insertion, there was a transient increase in the mean of bleeding days during the first month of treatment (from 7.2 to 9.9 days), which returned to baseline levels by the second month and continued to decline thereafter (Fig. 1). Although the number of mean spotting days also increased during the first month of treatment (from 1.8 to 10.0 days) before gradually declining, this number remained elevated relative to baseline throughout the first year of treatment. At baseline, spotting represented 20% (1.8/9.0) of the total number of B/S days increasing to 50% (10.0/19.9) of B/S days during the first 30-day reference period. Spotting appeared to account for about half of the total number of B/S days throughout the first year of treatment. The bleeding pattern associated with LNG-IUS use for HMB over 12 months of treatment is summarized in Table 3. Consistent with the decrease in the mean number of B/S days with continued use of the LNG-IUS, the proportion of women with amenorrhea increased over the first year of use.
4. Discussion To the authors' knowledge, this is the first study to report the change in the number of bleeding and/or spotting days in women with HMB (in women without structural pelvic pathology) treated with the LNG-IUS. Although a couple of earlier studies have also reported bleeding patterns after Table 3 Bleeding pattern associated with LNG-IUS treatment in women with HMB (expressed as percentage of women experiencing these patterns in 90 days up to 6 and 12 months, respectively) Bleeding pattern (%)
Amenorrhea Infrequent bleeding/spotting Frequent bleeding/spotting Prolonged bleeding/spotting
Duration of LNG-IUS treatment 6 months (n= 143)
1 year (n= 47)
1.7 21.6 11.2 55.6
8.8 23.8 6.3 25.0
insertion of the LNG-IUS in women with HMB [19,20], these used different definitions for the various bleeding patterns making direct comparability among studies difficult. Broadly, these earlier studies in women with HMB are suggestive of a pattern of reduced bleeding days with LNGIUS treatment (i.e., increased amenorrhea rates) up to 3 years [19,20], a pattern that is confirmed in our current analysis. Our results provide guidance to clinicians and patients as they demonstrate that the changes in bleeding and/or spotting days observed following insertion of the LNG-IUS in women with HMB follows a similar pattern to that observed in women using the intrauterine system for contraception [1,4,5,21], except that women with HMB may have more B/S days over the first year of use (see below). A systematic review and meta-analysis of bleeding/ spotting data and bleeding patterns were not possible because there are no universally agreed definitions for these terms, and as such, the many disparate definitions hinder direct comparisons across studies from different groups. The studies included in the current analysis were selected as they were all conducted by a single sponsor (Bayer HealthCare) using the same methodology to assess number of bleeding/spotting days and bleeding patterns (data not previously published) such that complete data sets were available for analysis. Pooling of the measured MBL data to assess the overall reduction in MBL volume was not done as two different MBL measurement techniques were used in the four studies, and the effects of the LNG-IUS on reducing measured MBL in women with HMB are well established in the literature [14]. We have shown in women with HMB that there is an initial increase in mean bleeding days during the first month following LNG-IUS placement that subsides back to baseline levels by the second month and continues to subside further over the first year of treatment. The number of mean spotting days also increases during the first month of treatment, but although this number declines with continued use, it remains elevated relative to pretreatment during the first year of treatment. “Spotting” is usually very limited in
J. Jensen et al. / Contraception 87 (2013) 107–112
normal menstruations. The reduction in the number of bleeding days is accompanied by an increase in the proportion of women with amenorrhea and a concomitant decrease in the proportion of women with prolonged bleeding. Although this study data does not provide information on duration of use more than 1 year, a small number of women (n= 22) were followed for up to 3 years, and a trend toward a further reduction in the mean number of B/S days with continuing use was observed (data not shown). Hence, “bleeding” is converted into spotting or “absence of bleeding” in most of these women. The rate of amenorrhea at 1 year in the current pooled analysis (8.8%; Table 3) appears lower than that observed in women using the LNG-IUS for contraception following postmenstrual (16%) and postabortal (11%) LNG-IUS insertion [5]. Andersson et al. [1] reported that 17% of women using the LNG-IUS for contraception experienced a period of at least 90 days of no B/S during the first year. Over 3 years of follow-up of contraceptive users, Wang et al. [22] reported that approximately 17% of women experienced amenorrhea during the fourth 90-day reference period. In contrast, the earlier studies that reported bleeding patterns after insertion of the LNG-IUS in women with idiopathic HMB suggested much higher amenorrhea rates at 1 year (60% and 63%, respectively) [19,20]. It is unclear why such high amenorrhea rates were reported in these earlier studies; moreover, in one of the studies, the amenorrhea rate was higher than in a comparator group using the LNG-IUS for contraception (63% vs. 44%) [19]. The changes in bleeding pattern achieved with the LNGIUS in women with HMB in the current study appear to be well tolerated, with only 6% of women discontinuing their planned treatment duration due to bleeding problems. In women using the LNG-IUS for contraception, Andersson et al. [1] reported net cumulative discontinuation rates for bleeding problems of 6 per 100 women during the first year of contraceptive use, increasing to 10 per 100 women during the third year. Therefore, despite an apparent higher number of mean B/S days in women using the LNG-IUS for HMB compared with those using it for contraception [1,21], this does not result in an increased rate of discontinuation due to bleeding problems. A likely explanation for this clinically important finding is that the substantial and highly significant reduction in the volume of bleeding observed during the use of the LNG-IUS in women with HMB more than compensates for the small increase in the mean number of spotting days. Indeed, with median reductions in MBL of 95% [11,12] and 97% (unpublished observations) reported at 6 and 12 months, respectively, the four studies show that the LNG-IUS effectively reduces MBL in women with HMB to very low levels [11,12,15,16], with much greater reductions than those achieved with the comparator treatments MPA [11], tranexamic acid (unpublished observations) and mefenamic acid [12]. Another explanation for the similar discontinuation rate due to bleeding problems may be counseling. It has been shown
111
that counseling women about the possibility of bleeding problems increases satisfaction with the LNG-IUS compared with women who are less informed [23]. The current pooled analysis in women with HMB has a number of limitations that need to be considered when assessing its relative generalizability. The number of women who used the LNG-IUS across the four randomized studies pooled was relatively small (n= 163). As such, more longterm studies (N 1 year) assessing the changes in bleeding pattern in women using the LNG-IUS for the treatment of HMB would be required to confirm our observations. Although most of the same inclusion and exclusion criteria applied to all four studies, the exclusion criterion of a BMI of N 35 kg/m 2 applied only to the study by Kaunitz et al. [11]. A BMI of N 35 kg/m 2 was not deemed to be a potential confounding factor for the interpretation of the data because only four women in all four studies fell into this category. In conclusion, the LNG-IUS is an effective and welltolerated treatment option for HMB. Although women using the LNG-IUS for HMB may have more B/S days than those using it for contraception, overall, the changes in the bleeding pattern follow a similar and favorable pattern in both populations, and the discontinuation rates due to bleeding problems do not appear to be higher in the HMB versus the contraceptive population. These data may be useful for counseling women with HMB about the use of the LNG-IUS and the bleeding patterns they may experience. Acknowledgments This study was funded by Bayer HealthCare, Berlin, Germany, the manufacturer of the LNG-IUS. Thomas Schmelter (Bayer HealthCare) provided assistance with programming and statistical analysis. Medical writing services during the preparation of this manuscript were provided by Amy Evans and Richard Glover of inScience Communications, Springer Healthcare. Funding for this editorial assistance was provided by Bayer HealthCare. References [1] Andersson K, Odlind V, Rybo G. Levonorgestrel-releasing and copper-releasing (Nova T) IUDs during five years of use: a randomized comparative trial. Contraception 1994;49:56–72. [2] Backman T, Rauramo I, Huhtala S, Koskenvuo M. Pregnancy during the use of levonorgestrel intrauterine system. Am J Obstet Gynecol 2004;190:50–4. [3] Cox M, Tripp J, Blacksell S. Clinical performance of the levonorgestrel intrauterine system in routine use by the UK Family Planning and Reproductive Health Research Network: 5-year report. J Fam Plann Reprod Health Care 2002;28:73–7. [4] Jensen JT, Nelson AL, Costales AC. Subject and clinician experience with the levonorgestrel-releasing intrauterine system. Contraception 2008;77:22–9. [5] Suvisaari J, Lahteenmaki P. Detailed analysis of menstrual bleeding patterns after postmenstrual and postabortal insertion of a copper IUD or a levonorgestrel-releasing intrauterine system. Contraception 1996; 54:201–8.
112
J. Jensen et al. / Contraception 87 (2013) 107–112
[6] Gemzell-Danielsson K, Inki P, Boubli L, O'Flynn M, Kunz M, Heikinheimo O. Bleeding pattern and safety of consecutive use of the levonorgestrel-releasing intrauterine system (LNG-IUS) — a multicentre prospective study. Hum Reprod 2010;25:354–9. [7] Hallberg L, Nilsson L. Determination of menstrual blood loss. Scand J Clin Lab Invest 1964;16:244–8. [8] National Collaborating Centre for Women's and Children's Health (NCCWCH). National Institute for Clinical Excellence (NICE) guidelines. CG44 Heavy Menstrual Bleeding: full guideline. 24 January; 2007. [9] Fraser IS, Langham S, Uhl-Hochgraeber K. Health-related quality of life and economic burden of abnormal uterine bleeding. Expert Rev Obstet Gynecol 2009;4:179–89. [10] Andersson JK, Rybo G. Levonorgestrel-releasing intrauterine device in the treatment of menorrhagia. Br J Obstet Gynaecol 1990;97:690–4. [11] Kaunitz AM, Bissonnette F, Monteiro I, Lukkari-Lax E, Muysers C, Jensen JT. Levonorgestrel-releasing intrauterine system or medroxyprogesterone for heavy menstrual bleeding: a randomized controlled trial. Obstet Gynecol 2010;116:625–32. [12] Reid PC, Virtanen-Kari S. Randomised comparative trial of the levonorgestrel intrauterine system and mefenamic acid for the treatment of idiopathic menorrhagia: a multiple analysis using total menstrual fluid loss, menstrual blood loss and pictorial blood loss assessment charts. Br J Obstet Gynaecol 2005;112:1121–5. [13] Kaunitz AM, Meredith S, Inki P, Kubba A, Sanchez-Ramos L. Levonorgestrel-releasing intrauterine system and endometrial ablation in heavy menstrual bleeding: a systematic review and meta-analysis. Obstet Gynecol 2009;113:1104–16. [14] Kaunitz AM, Inki P. The levonorgestrel-releasing intrauterine system in heavy menstrual bleeding: a benefit–risk review. Drugs 2012;72: 193–215.
[15] Istre O, Trolle B. Treatment of menorrhagia with the levonorgestrel intrauterine system versus endometrial resection. Fertil Steril 2001;76: 304–9. [16] Rauramo I, Elo I, Istre O. Long-term treatment of menorrhagia with levonorgestrel intrauterine system versus endometrial resection. Obstet Gynecol 2004;104:1314–21. [17] Higham JM, O'Brien PM, Shaw RW. Assessment of menstrual blood loss using a pictorial chart. Br J Obstet Gynaecol 1990;97:734–9. [18] Belsey EM, Machin D, d'Arcangues C. The analysis of vaginal bleeding patterns induced by fertility regulating methods. World Health Organization Special Programme of Research, Development and Research Training in Human Reproduction. Contraception 1986; 34:253–60. [19] Magalhães J, Aldrighi JM, de Lima GR. Uterine volume and menstrual patterns in users of the levonorgestrel-releasing intrauterine system with idiopathic menorrhagia or menorrhagia due to leiomyomas. Contraception 2007;75:193–8. [20] Monteiro I, Bahamondes L, Diaz J, Perrotti M, Petta C. Therapeutic use of levonorgestrel-releasing intrauterine system in women with menorrhagia: a pilot study. Contraception 2002;65:325–8. [21] Gemzell-Danielsson K, Schellschmidt I, Apter D. A randomized, phase II study describing the efficacy, bleeding profile, and safety of two low-dose levonorgestrel-releasing intrauterine contraceptive systems with Mirena. Fertil Steril 2012;97:616–22. [22] Wang SL, Wu SC, Xin XM, Chen JH, Gao J. Three years' experience with levonorgestrel-releasing intrauterine device and Norplant-2 implants: a randomized comparative study. Adv Contracept 1992;8: 105–14. [23] Backman T, Huhtala S, Luoto R, Tuominen J, Rauramo I, Koskenvuo M. Advance information improves user satisfaction with the levonorgestrel intrauterine system. Obstet Gynecol 2002;99:608–13.