8.1. Hormonal treatments for HMB
Heavy menstrual bleeding can occur for a variety of reasons. Many women with HMB will experience ovulatory (generally regular) cycles. In these women, their excessive bleeding may not be attributable directly to a hormonal imbalance but to a disturbance of the physiological pathway, such as increased fibrinolytic activity in the endometrium, increased prostaglandin levels or the presence of fibroids.
In women with HMB related to hormonal imbalance, there is often no recognisable pathology; their bleeding results from abnormalities in the hypothalamo-pituitary-ovarian-endometrial axis. This results in anovulatory (generally irregular) cycles, which are particularly common at the time of menarche and around the perimenopause. The failure of ovulation and progesterone-induced luteal phase secretory transformation of the endometrium results in bleeding that is often heavy, less clearly defined and irregular.
shows a schematic of all the RCT comparisons that have been undertaken for pharmaceutical treatments. What this figure highlights is the variable amount of RCT evidence available, comparing each treatment.
RCT evidence base for pharmaceutical interventions for HMB; trials comparing treatments are positioned along the line linking them; trials within the boxes are placebo controlled;,,,,,,, ,,,,,,, GnRH-a is not included here as it is an intervention for (more...)
Information from the individual RCTs included in the reviews can be found in evidence tables 8.1 to 8.9.104,105,260–270 A summary of all pharmaceutical treatment options for HMB is given in at the end of this chapter.
Pharmaceutical options for treating HMB; discuss the risks and benefits of each option with the woman and provide information and support to aid decision making.
8.2. Intrauterine levonorgestrel-releasing systems (LNG-IUS)
The levonorgestrel-releasing intrauterine system (LNG-IUS) is an intrauterine, long-term progestogen-only method of contraception licensed for 5 years of use. It has a T-shaped plastic frame with a rate-limiting membrane on the vertical stem that releases a daily dose of 20 micrograms of levonorgestrel. The effects of the LNG-IUS are local and hormonal, including prevention of endometrial proliferation and thickening of cervical mucus, and suppression of ovulation in a small minority of women. The system has to be fitted and removed by a qualified practitioner. As well as being licensed as a contraceptive device, the LNG-IUS is also licensed for the management of idiopathic menorrhagia and as the progestogen component of an HRT regimen.
8.2.1. Review on LNG-IUS
LNG-IUS
A systematic review from 2005 identified ten RCTs comparing LNG-IUS with surgery or pharmaceutical treatments.258 When comparing LNG-IUS against any pharmaceutical treatment (one RCT, n = 35), the review calculated the OR for amenorrhoea (> 3 months) as 8.67 (95% CI 1.52 to 49.35) in favour of LNG-IUS. The OR for proportion unwilling to continue with treatment (n = 91) was 0.27 (95% CI 0.10 to 0.67) in favour of LNG-IUS. The OR for proportion of women satisfied with treatment (one RCT, n = 40) was 2.13 (95% CI 0.62 to 7.33).
When comparing LNG-IUS with endometrial ablation, the review calculated the OR for the proportion of women satisfied with treatment (n = 136) as 0.61 (95% 0.26 to 1.46). The OR for amenorrhoea at up to 12 months (n = 223) was 0.75 (95% CI 0.36 to 1.54) in favour of surgery. The WMD for PBAC score at 12 months (one RCT, n = 66) was 33.2 ml (95% CI 27.2 ml to 39.2 ml) in favour of ablation.
When comparing LNG-IUS with hysterectomy, the study calculated the OR for satisfaction with treatment (n = 232) as being 1.17 (95% CI 0.41 to 3.34) in favour of hysterectomy.258 [EL = 1++]
In the second review ten studies met the inclusion criteria. This consisted of five RCTs and five case series. The review did not perform a meta-analysis but reported individual trial outcomes. The MBL reductions reported in the RCTs were between 71% and 96%.259 [EL = 1+]
Health economics
One trial conducted in Finland and reported in US dollars compared LNG-IUS with hysterectomy. The LNG-IUS was found to be cost-effective at 5 years when compared with hysterectomy. There was no statistically significant difference in quality of life scores at 5 years, as measured by the EQ-5D instrument, between the two treatment groups. Mean direct costs in the LNG-IUS arm remained significantly lower ($1,892) than the hysterectomy arm ($2,787), despite 40% of women in the LNG-IUS arm going on to have a hysterectomy. The trial did not compare the LNG-IUS with other pharmaceutical treatments (see Appendix A).104
No UK-based comparisons of LNG-IUS with any other medical or surgical treatment strategies were identified. In consultation with the GDG, a decision-analytic model was developed to examine the cost-effectiveness of pharmaceutical treatments as a first-line treatment for menorrhagia (for full results, see Appendix A). The results of the model showed that LNG-IUS generated more QALYs, at a lower cost, than any other pharmaceutical treatment strategy ().
Summary of cost–utility analysis for pharmaceutical treatments at 5 years for a cohort of 1000 women.
8.2.2. Evidence statements on LNG-IUS
The evidence from two systematic reviews and one subsequent publication shows that LNG-IUS produces a clinically relevant reduction in MBL in women complaining of HMB. RCTs showed this reduction to range between 71% and 96%. Evidence shows that the full benefit of treatment may not be seen for 6 months.
Economic modelling undertaken for this guideline shows that LNG-IUS is cost-effective when compared with both hormonal and non-hormonal treatments. It generates more QALYs at a lower cost than any other medical or surgical treatment strategy considered. When only those treatments that provide contraceptive benefits are compared, the combined oral contraceptive (COC) pill produces fewer QALYs at a higher cost than LNG-IUS. This analysis also considered surgery as a comparator treatment: the surgical strategy produced fewer QALYs at a higher cost than LNG-IUS.
The GDG is aware that an RCT (ECLIPSE) is currently underway comparing LNG-IUS with other pharmaceutical treatments. These direct comparison data may prove useful for determining the place of LNG-IUS in the treatment of HMB.
Recommendations on LNG-IUS
The recommendations for LNG-IUS can be found at the end of this chapter.
8.3. Combined oral contraceptives
Combined oral contraceptives (COCs) contain estrogen and progestogen in combination. Most brands are monophasic, being of the same strength throughout the 21 day treatment phase. Some vary to mimic the endogenous changes they replace. They act on the hypothalamo–pituitary axis to suppress ovulation and fertility. COCs are generally used in 21 day treatment cycles followed by a 7 day break, during which time endometrial breakdown and loss will occur. Such withdrawal bleeding is physiologically different from the bleeding that occurs after a natural ovulatory cycle. COCs have a number of general benefits and risks.271
8.3.1. Review on COCs
Overview of available evidence
Two systematic reviews272,273 and one primary study274 were included in the review of COCs and HMB. Further detail is provided in evidence table 8.3.
COCs
Two systematic reviews were identified,272,273 and these based their conclusions on the same RCT, which is described below.274
One RCT (n = 45) was identified on women with menorrhagia comparing COCs with naproxen, mefenamic acid and danazol. This comparative trial found that COCs reduced MBL by 43%, which was greater with than naproxen but less than with danazol or mefenamic acid. Adverse effects were not reported.274 [EL = 1+]
Health economics
No heath economic studies were identified on the use of COCs for HMB. Decision-analytic modelling was undertaken for this guideline to assess the cost-effectiveness of certain pharmaceutical treatments as the first-line treatment for menorrhagia. This analysis showed than when compared with all other pharmaceutical treatment strategies, the COC pill generated fewer QALYs at a greater cost. When compared with a strategy of no treatment, the COC pill generated an additional 1165.89 QALYs at an additional cost of £1,690,601, for a cohort of 1000 women. The incremental cost per additional QALY was £1,450: the use of the COC pill is cost-effective when compared with a strategy of no treatment. See Appendix A for the full model results.
8.3.2. Evidence statements on COCs
Evidence from one RCT of COCs (ethinyl estradiol 30 micrograms and levonorgestrel 150 micrograms for 21 days) on short-term outcomes found a reduction of MBL of 43%. The study did not report adverse effects.
8.3.3. GDG interpretation of evidence on COCs
The GDG highlighted that there is no available data on 20 microgram COC preparations. The GDG also highlighted that the COC pill has other non-contraceptive benefits, such as cycle control, reduced breast pain and reduced dysmenorrhoea.
Recommendations on COCs
The recommendations for COCs can be found at the end of this chapter.
8.4. Oral progestogens
Progesterone is a physiological hormone produced during the luteal phase of the menstrual cycle. It is responsible for secretory transformation of the endometrium and bleeding occurs after endogenous levels of estrogen and progesterone fall (fertilisation not having occurred). Progesterone is not available in oral formulation in the UK although vaginal preparations are available. A variety of oral synthetic progestogens are in clinical use. They vary in their potency and adverse effect profiles. The mechanisms by which oral progestogens reduce MBL are not fully understood.
8.4.1. Review on oral progestogens
Overview of available evidence
Two systematic reviews were identified examining the use of progestogens during the luteal phase of the menstrual cycle. Further detail is provided in evidence table 8.4.
Oral progestogens used in the luteal phase only
The first review from 1995 (four RCTs) showed norethisterone had no effect on MBL (MBL percentage change: 95% CI −6.1% to +1.1%).273 [EL = 1+]
The second review undertaken in 2003 (seven RCTs) showed that all other pharmaceuticals tested produced greater reductions in MBL than norethisterone (versus NSAIDs the change in MBL was −23.0 ml [95% CI −46.6 to 0.62] in favour of NSAIDs; versus danazol the change in MBL was −55.6 ml [95% CI −96.5 to −14.7] in favour of danazol; versus tranexamic acid the change in MBL was −111.0 ml [95% CI −178.5 to −43.5] in favour of tranexamic acid; and versus progesterone IUS the change in MBL was −51.0 ml [95% CI −83.6 to −18.4] in favour of IUS).275 [EL = 1++]
Detailed information from the individual RCTs included in the above reviews can be found in the evidence table.262,274,276–279
One cohort study (n = 16) of norethisterone and medroxyprogesterone acetate (MPA) was identified. The change in MBL associated with use of MPA was from a range of 104–107.5 ml prior to treatment to a range of 72–67 ml after treatment. However, this study only involved five women with menorrhagia and thus it is difficult to generalise the results.280 [EL = 2−]
Oral progestogens throughout both the follicular and luteal phases
An RCT (n = 44) on women with menorrhagia examined the use of oral progestogens cyclically for 21 days compared with LNG-IUS. The trial found an 83% reduction associated with longterm use of oral progestogens compared with a 94% reduction with LNG-IUS; the difference between groups was not statistically significant. However, 22% of women were satisfied with oral progestogens compared with 66% with LNG-IUS (no statistics provided).265 [EL = 1+]
Health economics
No health economic studies were found specifically on the use of oral progestogens for HMB.
8.4.2. Evidence statements on oral progestogens
The evidence from two reviews shows that oral progestogens taken during the luteal phase of the menstrual cycle (for 7–10 days) have no effect on MBL. Evidence from one small trial shows that oral progestogen (norethisterone 15 mg daily from day 5 to day 26 of the cycle) used long-term reduces MBL by 83%.
8.4.3. GDG interpretation of evidence on oral progestogens
The GDG discussion highlighted that:
other oral progestogens may be equally effective, but supportive data are not available
progestogen use for heavy menstrual loss requires a long-course regimen
while progestogens are effective, their clinical usefulness may be limited by tolerability.
Recommendations on oral progestogens
The recommendations for oral progestogens can be found at the end of this chapter.
8.5. Other hormonal treatments for HMB
Danazol is a synthetic androgenic steroid with anti-estrogenic and antiprogestogenic activity. It is antiproliferative with respect to the endometrium and is anovulatory by inhibiting the production of gonadotrophins by the pituitary gland.
Gestrinone has actions and adverse effects similar to those of danazol but is only required to be taken twice weekly as opposed to daily.
8.5.1. Review on other hormonal treatments for HMB
Overview of available evidence
Two systematic reviews273,281 and one primary study282 were identified. Further detail is provided in evidence table 8.5.
Other hormonal treatments
One review combined the results of five RCTs examining danazol and showed a weighted average reduction in MBL of 49.7% (95% CI 47.9% to 51.6%).273 [EL = 1+]
A second review included nine RCTs and found that danazol reduced MBL more than NSAIDs (one study; WMD −96.7 ml [95% CI −138.8 to −54.6]) or progestogens (one study; WMD −35.6 ml [95% CI −102.2 ml to +31 ml), but it also caused more adverse effects than NSAIDs (OR 7.0 [95% CI 1.7 to 28.2]) or progestogens (OR 4.05 [95% CI 1.6 to 10.2]).281 [EL = 1++]
Information from the individual RCTs included in the systematic reviews can be found in the evidence tables262,276,277,283–285
One small non-randomised trial (n = 37) compared gestrinone with a placebo. The study found that MBL was reduced in 15 of 19 women during the gestrinone treatment phase (P < 0.01) and that there was no change in MBL during placebo treatment. Adverse effects included dizziness, headaches, giddiness and tiredness in both groups.282 [EL = 2++]
Health economics
No heath economic studies were identified on the use of danazol for HMB.
8.5.2. Evidence statements on other hormonal treatments for HMB
Research shows that danazol is effective at reducing MBL, by approximately 50%, but is associated with significant androgenic adverse effects. One study shows that gestrinone, compared with placebo, reduces MBL. However, there is not enough evidence to make a recommendation about the use of gestrinone for the treatment of HMB.
8.5.3. GDG interpretation of evidence on other hormonal treatments for HMB
In their interpretation of the evidence for pharmaceutical treatments, the GDG placed a high value on reduction of MBL and minimising adverse effects. Based on these criteria, the risk–benefit analysis for danazol was balanced against recommending its use.
Recommendations on other hormonal treatments for HMB
The recommendations for other hormonal treatments can be found at the end of this chapter.
8.6. Injected/depot progestogens
Medroxyprogesterone acetate (Depo-Provera®; Pharmacia) can be injected intramuscularly to provide contraception for the following 12 weeks. A subdermal etonogestrel implant is also available, which achieves lower serum concentrations by using diffusion technology and is licensed as a contraceptive for 3 years. These preparations currently have no licence for the treatment of HMB.
8.6.1. Review on injected progestogens for HMB
Overview of available evidence
No studies were identified on the use of injected/depot progestogens on HMB. However, there are data for the impact on MBL, specifically amenorrhea rates, when used as a contraceptive (this is taken from the NICE Long-acting Reversible Contraception guideline).286
Injected progestogens
In one RCT (n = 3172) significantly more depot medroxyprogesterone acetate (DMPA) users reported amenorrhoea than norethisterone enantate (NET-EN) users (12% versus 7% and 24% versus 15% at 1 and 2 years, respectively).287 [EL = 1+]
One multinational RCT (n = 1216), undertaken mainly in developing countries, compared menstrual diaries in women given DMPA in 100 mg and 150 mg doses every 3 months. Amenorrhoea was experienced by 9–10% of women in the first 3 months and by 41–47% of women at 1 year.288 [EL = 1−]
In a study that assessed the effect of counselling on adherence in DMPA users, amenorrhoea was the major side effect reported, occurring in 34–35% of the women.289 [EL = 3]
Health economics
No health economic studies were identified examining the use of injected progestogens for HMB.
8.6.2. Evidence statements on injected progestogens for HMB
No evidence was found relating to the use of injected progestogens for the treatment of HMB. However, evidence from the NICE guideline on long-acting reversible contraception highlights that amenorrhoea is a side effect of injected progestogens.
Amenorrhoea is likely to occur during use of injectable contraceptives.
Recommendations on injected progestogens for HMB
The recommendations for injected progestogens can be found at the end of this chapter.
8.7. Hormone replacement therapy (HRT)
Estrogen replacement is used to relieve symptoms of menopause. In women with an intact uterus, progestogen opposition is added to reduce the risk of endometrial cancer that is associated with unopposed estrogen. Hormone replacement therapy (HRT) is not licensed for the treatment of HMB and may not contain a high enough dose of estrogen or progestogen to control an irregular cycle.
8.7.1. Review on HRT for treating HMB
Overview of available evidence
No studies were identified on the use of HRT alone to treat HMB.
Health economics
No health economic studies of the use of HRT to manage HMB were identified.
8.7.2. Evidence statements on HRT for treating HMB
No evidence was identified relating to the use of HRT to treat HMB. There is insufficient evidence to make any recommendation about the use of HRT in the treatment of HMB.
8.7.3. GDG interpretation of evidence on HRT for treating HMB
The GDG highlighted that HRT will theoretically redress the imbalance that results from anovulatory cycles, but to achieve this HRT with a higher dosage of progestogen may be required.
Recommendations on HRT for treating HMB
The recommendations for HRT can be found at the end of this chapter.
8.8. Gonadotrophin-releasing hormone analogue for treatment of HMB associated with uterine fibroids
A gonadotrophin-releasing hormone (GnRH) agonist is a synthetic peptide that acts like the natural GnRH secreted by the hypothalamus but which has a much longer biological half-life. As a result, there is an initial increase in follicle-stimulating hormone (FSH) and luteinising hormone (LH) secretion (the so-called flare effect). However, after about 10 days a profound hypogonadal effect is achieved through downregulation. Generally, this induced and reversible hypogonadism is the therapeutic goal. With no production of FSH or LH, there is no follicular development and estrogen production, no ovulation, no progesterone production and no menses. GnRH agonists (GnRH-a) are thus useful in the treatment of cancers that are hormonally sensitive, such as prostate cancer and breast cancer. GnRH agonists are also useful in the pharmaceutical treatment of estrogen-dependent lesions such as endometriosis and uterine leiomyoma. Current agonists used are given by subcutaneous or intramuscular injection or intranasally.
8.8.1. Review on GnRH-a for treating HMB
Overview of available evidence
Two primary studies were identified on GnRH-a alone. The primary aim of these studies was to examine the effect on uterine fibroids, but they did also report data on the effect on MBL. Further detail is provided in evidence table 8.6.
GnRH-a
An RCT (n = 128) on women with symptomatic uterine fibroids compared leuprolide acetate depot 3.75 mg with placebo for 24 weeks. The study found that leuprolide acetate produced superior outcomes on menorrhagia compared with placebo (leuprolide acetate group (n = 38): menorrhagia resolved or improved in 37 women, and no change or worse in one; placebo group (n = 37): menorrhagia resolved or improved in 26 women, and no change or worse in 11). However, all reported adverse effects were significantly higher in the leuprolide acetate group compared with placebo (hot flushes: leuprolide acetate 52 (83%) versus placebo 5 (8%) (P < 0.0001); vaginitis: 11 versus 0 (P < 0.0005); arthralgia: 9 versus 0 (P < 0.005); asthenia: 10 versus 3 (P < 0.05); peripheral oedema: 7 versus 1 (P < 0.05); insomnia: 6 versus 0 (P < 0.05); nausea: 6 versus 1 (P < 0.05); headaches: 18 versus 13; depression: 7 versus 2; emotional stability: 5 versus 1; decreased libido: 2 versus 0). The study concluded that treatment reduces MBL compared with placebo but with high levels of adverse effects. This shows that GnRH-a was more effective than placebo at improving subjective assessment of menorrhagia (RR 1.39 [95% CI 1.12 to 1.72]).290 [EL = 1+]
An RCT (n = 67) on women with symptomatic uterine fibroids compared buserelin MP 1.8 mg with depot leuprorelin 1.88 mg over 24 weeks. The study found that leuprorelin had a greater initial impact on menstrual bleeding, but by 24 weeks there was no difference between groups (buserelin: 8 weeks = 52.9% amenorrhea, 20 weeks = 88.9% amenorrhea; leuprorelin: 8 weeks = 84.4% amenorrhea, 20 weeks = 87% amenorrhea). The difference at 8 weeks was significant (P < 0.010), but at 20 weeks was non-significant. The study found that leuprorelin was associated with more hot flushes than buserelin (hot flushes at 12 weeks: buserelin = 5.9%, leuprorelin = 24.4%). The study shows that depot GnRH-a reduces menstrual bleeding. However, the study had high drop-out rates (11 of the buserelin and 15 of the leuprorelin group were lost to follow-up by 24 weeks).291 [EL = 1−]
8.8.2. Review on GnRH-a with HRT ‘add-back’ therapy
Some types of HRT are licensed for use as ‘add-back’ therapy. It is used in combination with GnRH-a to overcome hormone-related adverse effects associated with GnRH-a.
Overview of available evidence
Results from seven RCTs examining use on GnRH-a with ‘add-back’ therapy are shown below.
GnRH-a with HRT ‘add-back’
A crossover RCT (n = 16) on women with symptomatic uterine fibroids compared GnRH-a alone with GnRH-a plus medroxyprogesterone acetate (MPA) over 24 weeks. The study found that total uterine volume decreased to 73% of the baseline at 12 weeks in protocol B (monotherapy) (P < 0.04), but did not change in protocol A (combined therapy). After crossover at 12 weeks, the total uterine volume of women in protocol A decreased to 74% of the baseline (P < 0.02) at 24 weeks. A between-protocol comparison demonstrated a greater decline in total uterine volume in protocol B than A at 12 weeks but, after crossover, MPA addition was associated with a significant increase in total uterine volume (protocol B) compared with a decrease in protocol A at 24 weeks (P < 0.005). The study found that delayed use of ‘add-back’ allowed GnRH-a to shrink uterine fibroids. However, the study was small and the results can thus not be generalised. Furthermore, the study did not assess MBL outcomes.292 [EL = 1+]
An RCT (n = 16) on women with symptomatic uterine fibroids compared leuprolide acetate only with leuprolide acetate plus MPA over a 24 week period. Women in group A (monotherapy) had a significant reduction in uterine size from a pre-treatment volume of 601 ± 62 cm3 (mean ± standard error) to a mean uterine volume of 294 ± 46 cm3 at 24 weeks of therapy (P < 0.01). Women in group B (combined therapy) had a reduction in uterine volume, from 811 ± 174 cm3 to 688 ± 154 cm3, which was not statistically significant. However, only one woman in group B experienced hot flushes, whereas six women in group A had this symptom (P < 0.01). This study suggests that simultaneous use of GnRH-a and ‘add-back’ reduces the effect of GnRH-a. However, the study was small and the results can thus not be generalised.293 [EL= 1−]
An RCT (n = 51) on women with symptomatic uterine fibroids compared GnRH-a plus estrogen-progestin ‘add-back’ with GnRH-a plus progestin ‘add-back’. The study found that the symptoms of both groups improved (18 of 18 women in the estrogen-progestin group improved and 14 of 17 women in the progestin group improved), and that reduction in bone mineral density was similar in both groups (estrogen-progestin: pre-treatment = 1.102 g/cm3, 12 weeks = 1.074 g/cm3, 52 weeks = 1.053 g/cm3 (P < 0.05); progestin: pre-treatment = 1.081 g/cm3, 12 weeks = 1.045 g/cm3, 52 weeks = 1.047 g/cm3 (P < 0.05); control: pre-treatment = 1.081 g/cm3, 52 weeks = 1.078 g/cm3 (non-significant)). The study concluded that the regimens were equivalent and could be used as a long-term alternative to surgery for women with uterine fibroids. However, the high drop-out rate (16 of 51 women) was of concern.294 [EL = 1−]
An RCT (n = 12) on women with symptomatic uterine fibroids compared GnRH-a with GnRH-a plus estriol ‘add-back’ over 6 months. The study found a reduction in mean fibroid size of 53.6% by 2 months and a further 31.3% by 6 months in the non-‘add-back’ group, and in the ‘add-back’ group a reduction in mean fibroid size of 59.1% by 2 months and a marginal further reduction by 6 months. Bone mineral density reduced to 96.5% of its original density by 2 months and to 92.5% by 6 months in the non-‘add-back’ group, but did not change significantly in the ‘add-back’ group. The study concluded that GnRH-a plus estriol ‘add-back’ therapy might be considered for long-term treatment of uterine leiomyomas. However, the study was small and the results can thus not be generalised.295 [EL = 1−]
An RCT (n = 50) on women with symptomatic uterine fibroids compared GnRH-a plus placebo with GnRH-a plus tibolone over a 6 month period. The study found that menorrhagia improved in both groups (mean menorrhagia scores (0 to 10): baseline 8.2 versus 8.0, 6 months 0 versus 2.5 (both P < 0.01 from baseline)), but bone mineral density was reduced least in the tibolone group (baseline 1.056 g/cm3 versus 1.044 g/cm3, 6 months 1.002 g/cm3 versus 1.035 g/cm3 (P < 0.01 for placebo group versus baseline and versus treatment)). The study concluded that administering tibolone in association with GnRH-a reduces vasomotor symptoms and prevents bone loss, without compromising the therapeutic efficacy of GnRH-a alone.296 [EL = 1−]
An RCT (n = 100) on women with symptomatic uterine fibroids compared GnRH-a plus raloxifene with GnRH-a plus placebo. The study found that bone mineral density levels fell significantly in the placebo group in comparison with the baseline and in comparison with the treatment group (P < 0.05). The study concluded that raloxifene prevents GnRH-a related bone loss in premenopausal women with uterine leiomyomas. The study did not assess MBL outcomes.297 [EL = 1−]
Further results from this RCT examined the HRQoL impact of treatment compared with placebo and compared with healthy controls (GnRH-a plus raloxifene (n = 45) versus GnRH-a plus placebo (n = 46) versus normal population (n = 50)). The study found that cognitive functioning was adversely affected by treatment but that HRQoL was improved (Kupperman index (0 to 51): baseline = 2.6 (SD 1.2) versus 2.1 (SD 1.1) versus 2.1 (SD 1.2), sixth cycle = 22.8 (SD 3.9) versus 25.6 (SD 4.2) versus 2.5 (SD 1.3); SF-36: baseline = 50.4 (SD 14.1) versus 52.6 (SD 14.5) versus 84.2 (SD 10.4); sixth cycle = 80.3 (SD 11.5) versus 81.7 (SD 12.6) versus 83.4 (SD 10.2)). The paper concluded that GnRH-a causes a reduction in cognitive functioning in women with symptomatic fibroids but improves HRQoL to near-normal levels.298
An RCT (n = 12) on women with symptomatic uterine fibroids compared GnRH-a with placebo over 24 weeks. The study found that uterine volume and myoma volumes were improved in the treatment group but worsened in the placebo group (treatment versus placebo: uterine volume at baseline = 645 cm3 versus 457 cm3 (non-significant), post-treatment uterine volume = 467 cm3 versus 656 cm3 (P < 0.02); myoma volume at baseline = 402 cm3 versus 267 cm3, post-treatment myoma volume = 334 cm3 versus 417 cm3 (P = 0.06)). The study concluded that temporary hypoestrogenism induced by GnRH analogues can produce a significant, though temporary, reduction in uterine volumes, and that the non-myoma volume is responsible for much of the reduction and enlargement. However, the study was small and the results can thus not be generalised.299 [EL = 1−]
Health economics
No health economic studies on the use of GnRH-a with HRT ‘add-back’ to manage HMB were identified.
8.8.3. Evidence statements on GnRH-a
Evidence from two trials shows that GnRH-a reduces MBL, in the form of amenorrhea, with an RR of 1.39 [95% CI 1.12 to 1.72] for improvement in MBL, and amenorrhea rates of 89%. However, GnRH-a is associated with significant adverse effects, including perimenopausal symptoms, headaches and nausea.
Evidence from seven RCTs shows that GnRH-a causes fibroid shrinkage and a decrease in MBL. However, the effects do not continue after stopping treatment and adverse effects preclude longterm use. The addition of ‘add-back’ therapy should not alter efficacy and prevents the majority of the adverse effects.
Recommendations on GnRH-a
The recommendations for GnRH-a can be found at the end of this chapter.
8.9. Non-hormonal pharmaceutical treatments for HMB
Endometrial proliferation, secretory transformation and withdrawal bleeding are regulated by the hormonal cycle but the precise biochemical mechanisms are still not fully understood. Where HMB is a problem but hormonal therapy is declined or inappropriate, non-hormonal pharmaceuticals may be able to offer benefit through their effect on the physiology of the endometrium. These are not contraceptive and can be used in women seeking to become pregnant as they are initiated each cycle at the onset of bleeding.
A summary of all pharmaceutical treatment options for HMB is given in at the end of this chapter.
8.10. Tranexamic acid
Tranexamic acid is a competitive inhibitor of plasminogen activation, thus acting as an antifibrinolytic agent. Tranexamic acid inhibits factors associated with blood clotting but has no effect on coagulation within healthy blood vessels. There is no increase in the overall rate of thrombosis within those taking tranexamic acid compared with those not taking the drug when large communities are studied.
Tranexamic acid does not appear to affect platelet numbers or aggregation but acts to reduce the breakdown of fibrin in a pre-formed clot. As menstrual bleeding involves liquefaction of clotted blood from spiral endometrial arterioles, reduction in this process is believed to be the mechanism of reduced menstrual loss. Dosage for menorrhagia is 1 g (2 × 500 mg tablets) three to four times daily, from the onset of bleeding for up to 4 days.
8.10.1. Review on tranexamic acid for treating HMB
Overview of available evidence
This review includes three systematic reviews. No additional or subsequent primary studies were identified relating to those studies included in the systematic reviews. Further detail is provided in evidence table 8.7.
Tranexamic acid
The three reviews reported a range of MBL reduction depending on the studies included, but there was agreement that tranexamic acid produced a clinically important reduction in MBL in women with HMB.273,300,301 The main difference was between the inclusion and exclusion of studies involving women with or without IUD-induced* menorrhagia.
The first review pooled results from seven trials and found a reduction in MBL of 46.7% (95% CI 47.9% to 51.6%) with tranexamic acid.273 [EL = 1+]
The second review undertook a meta-analysis of two RCTs of tranexamic acid versus placebo and found a difference of −93.96 ml (95% CI −151.43 ml to −36.49 ml), P = 0.001, in favour of treatment.300 [EL = 1++]
A third review based on five trials concluded that oral tranexamic acid 2.0–4.5 g daily for 4–7 days per cycle reduced MBL by 34–59% over two to three cycles, and that 12% of women reported adverse events, such as nausea, vomiting, diarrhoea and dyspepsia.301 [EL = 1+]
There were no reports of deep vein thrombosis in any study in any of the reviews.299–301
Additional data from individual RCTs is available in evidence table 8.7.279,302–307
Health economics
No health economic studies were identified on the use of tranexamic acid to treat HMB. Decision-analytic modelling was undertaken for this guideline to assess the cost-effectiveness of certain pharmaceutical treatments as the first-line treatment for menorrhagia. This analysis showed that tranexamic acid generated fewer QALYs (3751.07) at a greater cost (£1,490,387) than the LNG-IUS (3818.89; £1,117,910). When compared with other non-hormonal treatments (NSAIDs), tranexamic acid generated more QALYs at a lower cost; NSAIDs generated 3699.38 QALYs at a cost of £1,529,051. When compared with a strategy of no treatment, tranexamic acid generated an additional 1306.25 QALYs at an additional cost of £1,466,387, giving an incremental cost per QALY of £1,122. Tranexamic acid is cost-effective when compared with either NSAIDs or no treatment, but not when compared with LNG-IUS. See Appendix A for the full model results.
8.10.2. Evidence statements on tranexamic acid for treating HMB
There is sufficient evidence based on RCT studies to make a recommendation on the use of tranexamic acid, but no evidence for other antifibrinolytics. Tranexamic acid at a dose of 2.0–4.5 g per day for 3–5 days from the onset of bleeding causes a clinically significant reduction in MBL for women with HMB, ranging from 29% to 58% in studies lasting up to 1 year. However, there are no long-term follow-up studies.
The results from the economic modelling show that when hormonal treatment methods are not considered acceptable, tranexamic acid generates more QALYs at a lower cost than NSAIDs, and more QALYs but at a greater cost than a strategy of no treatment. The cost per additional QALY when comparing tranexamic acid with no treatment is £1,122.
8.10.3. GDG interpretation of evidence on tranexamic acid for treating HMB
The GDG highlighted that tranexamic acid:
does not reduce dysmenorrhoea/pain associated with bleeding, so advice on suitable pain relief may be required
is not a contraceptive, so advice on suitable contraception may be required.
does not regulate cycles, so advice and suitable additional treatment should be given, if required.
Recommendations on tranexamic acid for treating HMB
The recommendations for tranexamic acid can be found at the end of this chapter.
8.11. Nonsteroidal anti-inflammatory drugs (NSAIDs)
A variety of NSAIDs have been used to treat HMB. These agents reduce prostaglandin synthesis by inhibition of cyclooxygenase. Prostaglandins affect local tissue reactivity and are implicated in inflammatory response, pain pathways, uterine bleeding and uterine cramps. HMB can be associated with increased prostaglandin levels. When NSAIDs are taken to reduce HMB, they should be taken regularly from the onset of bleeding, or just before, until heavy loss has abated.
8.11.1. Review on NSAIDs for treating HMB
Overview of available evidence
Two systematic reviews met the inclusion criteria. No additional or subsequent primary studies were identified relating to those studies included in the systematic reviews. Further detail is provided in evidence table 8.8.
NSAIDs
One systematic review undertook meta-analysis on individual NSAIDs and found a range of responses, the highest being for mefenamic acid and the lowest for ibuprofen (mefenamic acid (pooled results for ten studies) reduction in MBL = 29.0% [95% CI 27.9% to 30.2%]; diclofenac (two studies) reduction in MBL = 26.9% [95% CI 23.3% to 30.6%]; naproxen (five studies) reduction in MBL = 26.4% [95% CI 24.6% to 28.3%]; ibuprofen (three studies) reduction in MBL = 16.2% [95% CI 13.6% to 18.7%]).273 [EL = 1++]
A second review included only one placebo-controlled study but several comparative studies. The analysis showed that NSAIDs reduced MBL, but that tranexamic acid and danazol produced greater reductions (difference in reduction of MBL: NSAIDs versus placebo (one study, n = 11) −124 ml [95% CI −186 to −62]; NSAIDs versus tranexamic acid (one study, n = 48) +73 ml [95% CI 22 to 124]; NSAIDs versus etamsylate (two studies, n = 82) −43 ml [95% CI −86 to +0.5]; NSAIDS versus danazol (three studies, n = 79) +45 ml [95% CI 19 to 71]; NSAIDs versus oral progestogens (two studies, n = 48) −23 ml [95% CI −47 to +0.6]; NSAIDs versus IUD (one study, n = 16) −4 ml [95% CI −31 to +23]; NSAIDs versus oral contraceptive (one study, n = 26) +25 ml [95% CI −22 to +73]). However, NSAIDs had a better adverse effect profile than danazol, and a similar one to that of tranexamic acid.308 [EL = 1++]
Further information on individual RCTs can be found in the evidence table.262,274,284,305,306,309–318
Health economics
No health economic studies were identified on the use of NSAIDs to treat HMB. Decision-analytic modelling was undertaken for this guideline to assess the cost-effectiveness of certain pharmaceutical treatments as the first-line treatment for menorrhagia. This analysis showed that NSAIDs generated fewer QALYs (3699.38) at a greater cost (£1,529,051) than either the LNG-IUS (3818.89; £1,117,910) or tranexamic acid (3751.07; £1,490,387). The LNG-IUS and tranexamic acid are cost-effective alternatives when compared with NSAIDs. When compared with a strategy of no treatment, NSAIDs generated an additional 1254.56 QALYs at an additional cost of £1,505,051, giving an incremental cost per QALY of £1,199. NSAIDs are cost-effective when compared with a strategy of no treatment. See Appendix A for the full model results.
8.11.2. Evidence statements on NSAIDs for treating HMB
Overall, the evidence suggests that NSAIDs (mefenamic acid or naproxen) produce a clinically important reduction in MBL. Reported reductions in MBL ranged from 20% to 49%. NSAIDs were not as effective as danazol or tranexamic acid, but had a better adverse effect profile than danazol. The systematic reviews on which this statement is based are themselves based on an RCT. NSAIDs were found to be cost-effective when compared with no treatment, but generated fewer QALYs at a greater cost than LNG-IUS or tranexamic acid.
8.11.3. GDG interpretation of evidence on NSAIDs for treating HMB
In addition, the GDG highlighted that:
NSAIDs are not contraceptives, so advice on suitable contraceptives is recommended, if required
NSAIDs are additionally beneficial for the treatment of dysmenorrhoea
NSAIDs should not be used where it is thought that HMB is caused by bleeding disorders
owing to the cyclical nature of use, well-known adverse effects associated with long-term use of NSAIDs are reduced
there is no evidence regarding the effect of NSAIDs on HMB in the presence of uterine fibroids, as women with fibroids were excluded from the trials
NSAIDs should not be used where it is thought that HMB is caused by coagulation bleeding disorders.
Recommendations on NSAIDs for treating HMB
The recommendations for NSAIDs can be found at the end of this chapter.
8.12. Etamsylate
Etamsylate is believed to reduce bleeding from capillaries by correcting anomalies of platelet adhesion. It does not appear to affect the fibrin cascade. It is taken as 500 mg four times daily from, but not before, the onset of bleeding.
8.12.1. Review on etamsylate for treating HMB
Overview of available evidence
Three systematic reviews were identified that assessed the use of etamsylate.273,300,308 No additional or subsequent primary studies were identified with regard to those studies included in the systematic reviews. Further detail is provided in evidence table 8.9.
One review pooled results of four studies and found etamsylate reduced MBL in HMB by 13.1% (95% CI 10.9% to 15.3%), but this was less than most other interventions.273 [EL = 1+]
A review compared etamsylate with NSAIDs and found NSAIDs were more effective at reducing MBL (reduction in MBL for NSAIDs versus etamsylate = 43 ml [95% CI −0.5 to 86]).308 [EL = 1++]
A review compared antifibrinolytics with etamsylate and found the former reduced MBL more than the latter (reduction in MBL for antifibrinolytics versus etamsylate (one study) = 97 ml [95% CI 60 to 134] in favour of tranexamic acid).300 [EL = 1++]
Information on individual RCTs included in the reviews can be found in the evidence table305,310,319.
Health economics
No health economic studies were identified on the use of etamsylate for the treatment of HMB.
8.12.2. Evidence statements on etamsylate for treating HMB
The evidence on the MBL change for etamsylate is insufficient, with figures from one review reporting that etamsylate reduces MBL by an average of 13.1%, but that this is less than other pharmaceutical treatments.
Recommendations on etamsylate for treating HMB
The recommendations for etamsylate can be found at the end of this chapter.
GDG interpretation of evidence on pharmaceutical treatments for HMB
In their interpretation of the evidence for pharmaceutical treatments, the GDG placed a high value on reduction of MBL and minimising adverse effects.
The GDG based their assessment firstly on the clinical effectiveness of treatments and secondly on the cost-effectiveness of treatments. The results of the systematic review showed that LNG-IUS, NSAIDs, tranexamic acid and COCs could be considered equivalent in terms of effectiveness. Health economic modelling showed that the LNG-IUS was the preferred treatment option when long-term use of a treatment was required. Further details can be found in evidence tables 8.1 to 8.9.
Recommendations on pharmaceutical treatments for HMB
Pharmaceutical treatment should be considered where no structural or histological abnormality is present, or for fibroids less than 3 cm in diameter which are causing no distortion of the uterine cavity. [D(GPP)]
The healthcare professional should determine whether hormonal contraception is acceptable to the woman before recommending treatment (for example, she may wish to conceive). [D(GPP)]
If history and investigations indicate that pharmaceutical treatment is appropriate and either hormonal or non-hormonal treatments are acceptable, treatments should be considered in the following order:*
levonorgestrel-releasing intrauterine system (LNG-IUS) provided long-term (at least 12 months) use is anticipated
†‡ [A]
tranexamic acid [A] or nonsteroidal anti-inflammatory drugs (NSAIDs) [A] or combined oral contraceptives (COCs) [B]
norethisterone (15 mg) daily from days 5 to 26 of the menstrual cycle, or injected long-acting progestogens.
§ [A]
If hormonal treatments are not acceptable to the woman, then either tranexamic acid or NSAIDs can be used. [D(GPP)]
Women offered an LNG-IUS should be advised of anticipated changes in the bleeding pattern, particularly in the first few cycles and maybe lasting longer than 6 months. They should therefore be advised to persevere for at least 6 cycles to see the benefits of the treatment.¶[D(GPP)]
If pharmaceutical treatment is required while investigations and definitive treatment are being organised, either tranexamic acid or NSAIDs should be used. [D(GPP)]
When HMB coexists with dysmenorrhoea, NSAIDs should be preferred to tranexamic acid. [D(GPP)]
Ongoing use of NSAIDs and/or tranexamic acid is recommended for as long as they are found to be beneficial by the woman. [D(GPP)]
Use of NSAIDs and/or tranexamic acid should be stopped if it does not improve symptoms within three menstrual cycles. [D(GPP)]
When a first pharmaceutical treatment has proved ineffective, a second pharmaceutical treatment can be considered rather than immediate referral to surgery. [D]
Use of a gonadotrophin-releasing hormone analogue could be considered prior to surgery or when all other treatment options for uterine fibroids, including surgery or uterine artery embolisation (UAE), are contraindicated. If this treatment to be used for more than 6 months or if adverse effects are experienced then hormone replacement therapy (HRT) ‘add-back’ therapy is recommended.** [B]
Danazol should not be routinely used for the treatment of HMB. [A]
Oral progestogens given during the luteal phase only should not be used for the treatment of HMB. [A]
Etamsylate should not be used for the treatment of HMB. [A]
- *
World Health Organization ‘Pharmaceutical eligibility criteria for contraceptive use’ (WHOMEC) apply. These criteria can be used to assess the individual's suitability for particular contraceptives. This allows informed decision making by the woman prior to the start of treatment. [www.ffprhc.org.uk/admin/uploads/298_UKMEC_200506.pdf]
- †
Check the Summary of Product Characteristics for current licensed indications. Informed consent is needed when using outside the licensed indications. This should be discussed and documented in the notes.
- ‡
See ‘Long-acting reversible contraception’, NICE clinical guideline 30,www.nice.org.uk/CG030, for more detail.
- §
Check the Summary of Product Characteristics for current licensed indications. Informed consent is needed when using outside the licensed indications. This should be discussed and documented within the notes. In adolescents and women older than 40 years, refer to CSM advice issued in November 2004. Go to www.mhra.gov.uk and search for Depo Provera.
- ¶
See ‘Long-acting reversible contraception’, NICE clinical guideline 30, www.nice.org.uk/CG030, for more detail.
- **
Check the Summary of Product Characteristics for current licensed indications. Informed consent is needed when using outside the licensed indications. This should be discussed and documented in the notes.
Research recommendations on pharmaceutical treatments for HMB
- *
IUD here refers to copper products, not LNG-IUS.
