Article Text
Relevant BNF section: BNF 7.3.5
Abstract
Until recently, women in the UK who wanted emergency contraception had two options: an oral hormonal method (levonorgestrel), which is licensed for use up to 3 days after unprotected sexual intercourse; or a copper-bearing intrauterine device (IUD), which can be inserted up to 5 days after unprotected intercourse or up to 5 days after the earliest likely calculated ovulation. Now ▼ulipristal acetate (ellaOne – HRA Pharma), a new oral hormonal emergency contraceptive, has been licensed in the European Union for use within 120 hours (5 days) of unprotected intercourse.1 Here we assess whether it is an advance for emergency contraception.
Statistics from Altmetric.com
Relevant BNF section: BNF 7.3.5
About emergency contraception
After a single act of unprotected sexual intercourse, the probability of consequent pregnancy can be as high as 30%, depending on (among other factors) when intercourse occurs in the menstrual cycle.2 A copper-bearing IUD has traditionally been considered the most effective method of emergency contraception, with a failure rate no higher than 1%,2 although such devices are not ‘labelled’ for such use.1 They work by inhibiting fertilisation and implantation, and can be used for ongoing contraception thereafter.2 However, they are associated with a risk of ectopic pregnancy and pelvic inflammatory disease.2 Also, because IUD insertion is invasive, the procedure needs to be performed by an appropriately skilled healthcare professional; so women may have difficulty in getting such a device fitted promptly.
Levonorgestrel is a progestogen, with a mode of action in emergency contraception that is not completely understood.2 However, its primary action is thought to be inhibition of ovulation by interfering with luteinising hormone, rather than inhibition of implantation of a fertilised ovum.3 Levonorgestrel emergency contraception has been licensed in the UK for around 10 years.4 When this product first became available, the dose of levonorgestrel was one 750μg tablet followed, 12 hours later, by a second 750μg tablet.4 Subsequently, a single 1.5mg tablet of levonorgestrel was found to be just as effective, with no increase in unwanted effects.5 Now, only the 1.5mg single dose is licensed in the UK. This dose is available free of charge either on prescription (e.g. from GPs, family planning services) or through supply via a Patient Group Direction as Levonelle 1500 (e.g. from certain community pharmacies and nurse-led services); it can also be purchased over the counter (OTC) from community pharmacies as Levonelle One Step. While this contraception is licensed for use up to 3 days (72 hours) after unprotected sexual intercourse, subgroup analysis of a randomised controlled trial undertaken by the World Health Organization involving 4,136 women suggested that the drug was probably effective for up to 5 days (an unlicensed use); however, this finding was not conclusive because the numbers were small (a total of 314 women in the 73–120-hour window).5
What is ulipristal?
Ulipristal (pronounced oo-li-pri-stal) is a prescription-only medicine described as a ‘synthetic selective progesterone receptor modulator’.1,6 Through binding with high affinity to progesterone receptors, ulipristal prevents the binding of progesterone itself,6 and it has both antagonistic and partial agonistic effects.1 It is thought to work as an emergency contraceptive by inhibiting or delaying ovulation, and also has some endometrial effects.1,6
EllaOne tablets contain 30mg of ulipristal. The licensed dose is one tablet, taken as soon as possible, within 120 hours (5 days) of unprotected sexual intercourse or contraceptive failure.6 It can be taken at any point during the menstrual cycle, with or without food. The dose should be repeated if vomiting occurs within 3 hours of taking the drug. The summary of product characteristics (SPC) states that, apart from this situation, repeated administration of ulipristal within the same menstrual cycle is not advisable because the safety and efficacy of this practice have not been investigated. It also states that pregnancy should be excluded before ulipristal is administered.
Clinical effectiveness
Evidence of the efficacy of ulipristal as emergency contraception comes from two published randomised controlled comparisons with levonorgestrel3,7 and one published uncontrolled study.8
Comparative studies with levonorgestrel
One of the comparisons with levonorgestrel was a double-blind non-inferiority trial* involving 1,672 women (aged at least 18 years) seeking emergency contraception within 72 hours of unprotected sexual intercourse.7 The women received either ulipristal 50mg (a higher dose than in the marketed product**) or two doses of levonorgestrel 750μg 12 hours apart, and returned for a follow-up visit 5–7 days after the expected onset of the next menstrual period. In the modified intention-to-treat (ITT) population (i.e. excluding 94 women lost to follow-up or with incomplete data), the proportion of women who became pregnant (the primary outcome measure) was similar in the two groups (1.5% with ulipristal vs. 1.8% with levonorgestrel). The difference between the treatments was −0.3%, with an upper 97.5% confidence limit of 1.42%, which was less than the predetermined non-inferiority margin of 2%. Data were also analysed for the “efficacy-evaluable” population of 1,549 women (i.e. those not pregnant at enrolment, who took the study drug without additional emergency contraception during the treatment cycle, and who met one of the following post-treatment criteria: had menses and a negative urine pregnancy test; verified amenorrhoea with a negative pregnancy test; or a verified pregnancy). In this group, the difference between the treatments on the primary outcome measure was −0.8%, with an upper 97.5% confidence limit of 0.77%, which was again less than the non-inferiority margin of 2%. Therefore, on the data for both the modified ITT and efficacy-evaluable populations, ulipristal was judged to be non-inferior to levonorgestrel.
The other comparison was a single-blind non-inferiority trial that randomised 2,221 women (aged 16 years or over) seeking emergency contraception to either ulipristal 30mg or levonorgestrel 1.5mg orally at up to 5 days after sexual intercourse.3 Only the participants were blinded to treatment. The primary outcome measure was the rate of pregnancy in women who received emergency contraception within 3 days of unprotected intercourse; the rate in women who received it within 5 days was a secondary outcome measure. The efficacy evaluable population comprised 1,899 women (i.e. it excluded those lost to follow-up; those aged over 35 years [as recommended by the US Food and Drug Administration, because of reduced fertility in this age group]; those of unknown follow-up pregnancy status; and those re-enrolled into the study). Also excluded were seven pregnancies judged to have occurred either before, or at least 10 days after, emergency contraception. In women receiving emergency contraception within 72 hours of intercourse in the efficacy evaluable population, there were 15 pregnancies with ulipristal (1.8% of women, 95% CI 1.0 to 3.0) and 22 with levonorgestrel (2.6%, 95% CI 1.7 to 3.9). The difference between the two treatments was less than 1% and, on this basis, ulipristal was judged non-inferior to levonorgestrel. Data were not given for the ITT population. There were no pregnancies in the 97 women given ulipristal and 3 pregnancies in the 106 women given levonorgestrel between 72 and 120 hours (p=0.037).
Data for the efficacy populations in the two published trials3,7 have been combined (i.e. giving a total of 3,445 women and 60 pregnancies) to calculate an overall rate of pregnancy; this was lower in the ulipristal group than in the levonorgestrel group when emergency contraception was taken within 24, 72 or 120 hours after sexual intercourse.3 However, this pooling of data is questionable, since the two studies used different ulipristal products and different treatment regimens.
Uncontrolled study evidence
The uncontrolled trial included 1,499 women aged 18 years and over who presented for emergency contraception 48–120 hours after unprotected sexual intercourse and were given ulipristal.8 In all, 1,241 women were included in the primary efficacy analysis (which excluded women lost to follow-up, aged over 35 years, of unknown pregnancy status at follow-up, with repeated treatment or with pregnancies not thought to be due to failure of emergency contraception). In this population, the overall pregnancy rate (the primary outcome measure) was 2.1% (95% CI 1.4 to 3.1%). This compared to a calculated expected pregnancy rate without emergency contraception of 5.5%. Ulipristal was deemed to be effective because the upper limit of the 95% CI was more than 1% lower than the expected rate. The rate in the ITT population was 1.9% (95% CI 1.3 to 2.8%) compared with an expected rate of 5.7%. When the pregnancy rates were analysed by the various 24-hour time intervals between unprotected intercourse and administration of ulipristal, there was no evidence of a change in efficacy over time, up to 120 hours.
The uncontrolled design of this study was considered an acceptable alternative to a randomised controlled trial by the European Medicines Agency because there is no comparator “registered for more than 72 hours after UPI [unprotected intercourse]”, placebo treatment would be unethical, and the outcome (pregnancy or no pregnancy) is an objective parameter.1 Nevertheless, the study included a highly selected subgroup, which must limit the generalisability of the findings.
Unwanted effects
In clinical trials, most unwanted effects of ulipristal have been mild or moderate and resolved spontaneously.3,6,7 Abdominal pain and menstrual disorder were very common (i.e. occurred in at least 1 in 10 women).3,6,7 Most of the women had their next menstrual period within 7 days or at the expected time, but in 19% the menstrual period was more than 7 days late, while in 6% it was more than 7 days early.6 A delay of more than 7 days might increase anxiety, and requests for pregnancy tests and consultation. The range and frequency of unwanted effects with ulipristal are similar to those with levonorgestrel, with the exception that onset of menstruation is more likely to be delayed following ulipristal, and to be early following levonorgestrel.3 The SPC recommends that, in case of doubt, pregnancy should be excluded by a pregnancy test if there is a delay of more than 7 days before onset of the next menstrual period; abnormal bleeding at the expected date of menses; or symptoms of pregnancy. Common unwanted effects (i.e. occurring in at least 1 in 100 but no more than 1 in 10 women) include nausea, vomiting, dyspepsia, muscle spasms, back pain, dysmenorrhoea, heavy menstrual bleeding, intermenstrual bleeding, headache, dizziness, fatigue, infections and mood disorders.6
Precautions and interactions
The use of ulipristal in women with severe asthma insufficiently controlled by oral corticosteroid therapy is not recommended in the SPC (which states that the drug has high affinity for the glucocorticoid receptor). This also advises avoiding use of the product in “severe hepatic impairment”, because of a lack of data. Ulipristal is contraindicated during an existing or suspected pregnancy.6 The relative lack of experience with ulipristal means that its effects on the fetus and any teratogenic potential, through inadvertent use during pregnancy, are not known6 (while levonorgestrel is not known to be harmful9). If pregnancy does occur, the possibility of ectopic pregnancy should be considered.6 It is not known whether ulipristal is excreted into breast milk, but the SPC advises avoiding breastfeeding for at least 36 hours after taking the drug. Safety of the product has been established only in women aged 18 years or over.6 In-vitro studies have shown that ulipristal is metabolised by the cytochrome P450 3A4 isoenzyme and its contraceptive effect might therefore be reduced by inducers of this isoenzyme (e.g. carbamazepine, rifampicin, ritonavir, St. John's wort) if they are used concurrently or have been used in the last 2–3 weeks (the time taken for enzyme induction to wear off).6
Because ulipristal binds to progesterone receptors with high affinity, it may interfere with the action of progestogen-containing medicinal products.6 The SPC therefore recommends that, after taking ulipristal, “subsequent acts of intercourse be protected by a reliable barrier method until the next menstrual period starts” in women taking progestogen-containing hormonal contraception. However, the advice in the SPC (and the patient information leaflet10) on this point could be clearer for women who have a prolonged menstrual cycle or amenorrhoea. Women should be counselled about the potential continuing risk of becoming pregnant if they have sexual intercourse again after using ulipristal, since they might think, wrongly, they are protected from pregnancy up to 72–120 hours or longer after taking it.11 Concurrent use of ulipristal and emergency contraception containing levonorgestrel is not recommended (again because of ulipristal's action on progesterone receptors).6
Cost
The cost to the NHS of one 30mg tablet of ulipristal is £16.95 compared with £5.20 for one tablet of 1.5mg Levonelle 1500. The official retail price of OTC Levonelle One Step is up to £22.35. We know of no published cost-effectiveness data for ulipristal. The cost to the NHS for a copper IUD (not including the cost of fitting and follow-up) is around £9 to £13. Of note, once fitted for emergency contraception, an IUD can continue to provide contraceptive cover for several years.
What guidelines say
The Scottish Medicines Consortium has accepted ulipristal acetate (ellaOne) for use within NHS Scotland for emergency contraception within 120 hours (5 days) of unprotected sexual intercourse or contraceptive failure.12 It concluded that, when administered within the licensed time frame for ulipristal or an active comparator for emergency hormonal contraception, contraceptive efficacy with ulipristal was non-inferior to that with the comparator in individual studies and statistically superior in a meta-analysis of two studies. However, it also noted that other treatments are available at lower acquisition cost than that for ulipristal.
Conclusion
▼Ulipristal acetate is a new hormonal emergency contraceptive licensed for use up to 5 days after unprotected sexual intercourse. Clinical trials have indicated that it is no less effective than levonorgestrel (another hormonal emergency contraceptive) when used up to 3 days after unprotected intercourse; and limited data suggest that it is effective when taken up to 5 days after unprotected intercourse. This is a potential advantage over levonorgestrel, which is only licensed for use up to 3 days after unprotected intercourse. Unwanted effects with ulipristal are similar to those with levonorgestrel, but the subsequent menstrual period is more likely to be delayed with ulipristal than with levonorgestrel. The effects of inadvertent fetal exposure to ulipristal are not known, while levonorgestrel is not known to be harmful in this setting. Ulipristal costs around three times more than levonorgestrel, and is only available on prescription, unlike levonorgestrel, which is also widely available at pharmacies via a Patient Group Direction or to buy over the counter.
Levonorgestrel is still the first choice for emergency hormonal contraception if administered within 72 hours of sexual intercourse. If available and suitable, emergency insertion of a copper intrauterine device (IUD) should also be considered up to 5 days after the earliest likely calculated ovulation. Further evidence is needed before there is a change in practice. In particular, a superiority trial would be needed to prove that ulipristal should be used rather than levonorgestrel, compared with which it is less accessible and more costly, and does not have the same safety data. However, ulipristal provides prescribers with a licensed option for use in women presenting 72–120 hours after unprotected intercourse, if insertion of a copper IUD is not feasible.
References
[R=randomised controlled trial; M=meta-analysis]