What you need to know about hormonal contraception and drug interactions
Lin-Nam Wang presents a summary
Next time you dispense a course of amoxicillin to a woman and her patient medication record shows she is taking Microgynon, you do not need to give her the “additional contraceptive precautions” advice (PJ, 12 February 2011, p155).
The theory behind giving this advice in the past was that non-cytochrome P-450-inducing antibiotics could, temporarily, decrease numbers of bacteria in the colon, leading to reduced enterohepatic circulation of the oestrogen component of combined oral contraceptives.
(Bacteria in the colon would have broken down conjugates formed by first pass metabolism and the free drug that was reabsorbed might contribute to contraceptive efficacy.)
However, this theory was debatable. Previous guidance from the Faculty of Sexual and Reproductive Healthcare (FSRH; 2005) stated that trials investigating this interaction were limited by their size, short duration, inconsistent assessment of ovulation and failure to eliminate bias.
Supporting reports of pregnancies in women who had taken combined oral contraceptives and antibiotics were dismissed as no proof of causation. The British National Formulary also acknowledged that the risk of this interaction was “probably small”. Nevertheless, it was reasoned that the serious consequences of an unplanned pregnancy warranted a cautious approach.
This approach, however, was never universally adopted in other countries. The World Health Organization states that there is intermediate level evidence that contraceptive efficacy is not affected by most broad spectrum antibiotics.
Other supporting factors included that efficacy did not appear to be reduced in stoma patients with no enterohepatic circulation and that some of the reported pregnancies were in women on high-dose pills — if it is accepted that a low-dose pill provides effective contraception, failure of a high-dose pill due to a small reduction in enterohepatic circulation seems unlikely.
The WHO advises no restriction on use, and the UK, in guidance published in February 2011 by the FSRH, has now taken this approach.
But before you think “at last, some news that decreases pharmacists’ workloads”, note that it is still good practice to advise pill takers to use extra contraception in cases of diarrhoea or vomiting — common side effects of many antibiotics — lasting over 24 hours.
Continuing to provide the old advice is unlikely to harm patients but the new guidance also contains some changes with more significant consequences.
Key changes to guidance
Advising additional contraceptive precautions for those taking combined hormonal contraceptives (CHCs) while taking a non-enzyme inducing antibiotic is no longer advocated.
Recommendations on use of CHCs with enzyme-inducing drugs are modified.
Patients on lamotrigine monotherapy should be advised of potential toxicity in the hormone free interval and loss of seizure control with CHCs.
Ulipristal acetate (EllaOne) is included. This might reduce the efficacy of progestogen-containing contraceptives and could itself be affected by enzyme inducers and drugs altering gastric pH.
Information on interactions with antiretroviral drugs has been updated.
Griseofulvin is no longer considered a clinically important enzyme inducer and lansoprazole is no longer considered an enzyme inducer.
The interaction between hormonal contraception and coumarin anticoagulants has been removed due to lack of consistent evidence. (Monitoring may still be required but co-prescribing is unlikely due to contraindications.)
Prescribing with enzyme inducers
It is accepted that enzyme-inducing drugs (eg, carbamazepine, phenytoin, rifampicin) can reduce contraceptive efficacy of both combined and progestogen-only contraceptives although not of the progestogen injections (Depo-provera and Noristerat) or the levonorgestrel-releasing intrauterine system (Mirena).
The latest advice leading on from this has been modified. Women starting an enzyme-inducer should be advised to use a progestogen injection, Mirena or a copper-containing interuterine device for contraception while taking it and for 28 days after stopping.
For those who do not wish to switch contraceptive, the guidance now splits the use of an enzyme inducer into short- (two months or less) and long-term before listing alternatives. For example, with short-term therapy continued use of a combined hormonal contraceptive (CHC; pills containing at least 30µg of ethinylestradiol, patches or rings) is allowed but additional precautions (ie, condoms) are required.
In addition, to minimise risk of failure, the faculty recommends that the CHC regimen is extended (continuous [unlicensed] use until breakthrough bleeding) or “tricycled” (using three cycles with no break before a hormone-free interval of four days [unlicensed]). With pills, such regimens are only appropriate with monophasic 21-day products.
This option might suit women who are trying an enzyme inducer and do not wish to disrupt their CHC regimen, but many are unlikely to choose to have to use condoms as well as a CHC. For women on the pill, the guidance suggests prescribing two pills to give a daily dose of 50µg ethinylestradiol under an extended or tricycling regimen.
This last option is also suggested for women on long-term enzyme inducers (but not long-term potent enzyme-inducing antibiotics [ie, rifampicin and rifabutin]).
Prescribing with antiretroviral therapy
Interactions with antiretrovirals have been updated. In particular, Rhoda Lee, staff editor for ‘Stockley’s drug interactions’, who contributed to the development of the guidance, points out that ritonavir, one of the most potent inhibitors of CYP3A4 known, would have been predicted to increase ethinylestradiol levels but levels are decreased.
“Since publication of the previous guidelines, there have been at least five studies, which, with one exception (atazanavir), have all shown that ritonavir-boosted protease inhibitors (darunavir, fosamprenavir, lopinavir, tipranavir) modestly to markedly reduce ethinylestradiol levels from combined oral contraceptives.
“Since ethinylestradiol is also substantially metabolised by conjugation, it might be that ritonavir induces glucuronosyltransferase activity, thereby reducing ethinylestradiol levels,” she explained.
How contraceptives affect other drugs
The guidance continues to cover how contraceptives can affect other therapies. For example, it now recommends that women on lamotrigine monotherapy should be advised of the risk of loss of seizure control with CHCs, with possible toxicity in the hormone-free week, and that risks may outweigh the benefits of using this type of contraceptive. A higher dose of lamotrigine and extending the pill regimen can be considered.
The guidance includes new products such as ulipristal (EllaOne). This progesterone receptor modulator could also be affected by enzyme inducers but, unlike levonorgestrel (eg, Levonelle 1500), doubling the dose is not advised. Efficacy might also be affected by drugs that affect gastric pH (eg, antacids, proton pump inhibitors and H2 antagonists).
Moreover, the guidance highlights a concern that ulipristal could reduce efficacy of progestogen-containing contraceptives and additional precautions after taking this emergency contraceptive (for one week plus the time required for contraceptive efficacy to be established — see individual summaries of product characteristics) are advised.
The possibility that orlistat (ie, Alli) can reduce absorption of oral contraceptives by inducing diarrhoea is also mentioned.
Those supplying hormonal contraceptives should ask women about their current and previous medicines use (including herbal and dietary products) and tell them to seek advice before starting new drugs, the faculty says. Pharmacists might find the appendices in the new guidance useful, even if they do not read the full document (PDF 760K).
an FSRH statement on antiepileptic drugs and contraception (PDF 350K)
CPD articles on contraception:
Citation: The Pharmaceutical Journal URI: 11068907
Recommended from Pharmaceutical Press