Warfarin and oral miconazole: a major interaction overlooked in practice
A patient, MC, takes warfarin for atrial fibrillation. His target international normalised ratio range is 2.0–3.0 and his control was stable. At his regular INR appointment he mentioned that he had seen his GP the week before, for haematuria (who referred him for investigation on a 2WW [two-week wait] form, but the results were negative).
The patient was reminded to telephone the anticoagulant service if he experienced any bleeding in the future. A near patient INR test (see Panel 1) gave a reading over 8.
Panel 1: International normalised ratio
The INR is a measure of how long it takes for blood to clot. In near patient testing, a drop of blood taken from the patient’s finger-tip is dropped onto a strip placed in the INR monitor. Intrahealth clinics use CoaguChek XS Plus meters, which determine the INR by an electrochemical method after coagulation of the blood sample is activated with human recombinant thromboplastin.
On questioning, MC revealed that he had used a tube of miconazole oral gel two weeks previously, which he had bought in a community pharmacy.
An urgent venous INR was performed and found to be 15 — an extremely dangerous level. The patient was admitted to hospital for vitamin K administration to reverse the over-anticoagulation. His INR had probably been raised for over a week.
MC eventually returned to his previous warfarin maintenance dose. The haematuria resolved once his INR was back in range.
Miconazole oral gel (Daktarin) is frequently prescribed or sold over the counter for the treatment of oral candidal infection. Systemic miconazole can potentiate the anticoagulant effect of warfarin, because it is a strong inhibitor of the P450 isoenzyme CYP2C9, one of the main enzymes involved in warfarin metabolism.1
Clinically significant increases in INR can occur when miconazole oral gel is given to patients taking warfarin because miconazole can be absorbed through the oral mucosa, or from the bowel after the gel has been swallowed.
Although the major interaction between warfarin and miconazole oral gel is widely reported, time and time again, we are finding patients with clinically significant raised INRs due to this. In my 10 years of running anticoagulant clinics I have seen at least eight such interactions. Three further cases are described in Panel 2.
Panel 2: Three more reports from the warfarin clinic
Case 2 GG takes warfarin for atrial fibrillation. Her target INR range is 2.0–3.0 and her INR control was usually reasonably stable. She attended the clinic once every four weeks. At one appointment she presented as being well with no recent medication changes, but her near patient INR test gave a reading over 8. It was found that the patient had asked her community pharmacist for treatment for a sore mouth, and had been sold miconazole oral gel. She had not thought that this was a “medication change”, so had not informed the anticoagulant service. She had no bruising or bleeding. An urgent venous INR was sent to the haematology laboratory, which later telephoned me out of hours to say that the blood sample was failing to clot on its machine, so that it was going to record the INR as greater than 20.
There appears to be a perception among the pharmacists who were contacted following these interactions that, because miconazole oral gel is a topical application, its risks are minimal.
Points for practice
Miconazole oral gel should never be sold to patients taking warfarin. There are other, much safer options to treat oral thrush in patients taking oral anticoagulants. In our experience, nystatin oral suspension rarely appears to affect INR compared with miconazole oral gel, although there may be some potential for an interaction [see Case comment, below].
Patients should be told to inform their anticoagulant clinic of any new medicine — including topical medicines — because their appointment for an INR check may need to be brought forward.
It is important to make sure that patients know that interactions can also occur with herbal and homoeopathic products and vitamins, and that all changes should be reported to their anticoagulant monitoring clinic.
If a warfarin patient has been prescribed miconazole oral gel, community pharmacists should alert the prescriber and recommend an alternative.
Any patient taking warfarin who presents with bleeding should be immediately referred to their GP and anticoagulant monitoring clinic.
Since November 2011, our clinics, which currently look after over 5,000 patients, have encountered at least three patients with substantial increases in their INR levels following use of miconazole oral gel. As a result, we have produced a warning bulletin that has been distributed to local pharmacies and to the GP practices to which we provide anticoagulant monitoring. We have also alerted patients to this interaction via our patient information bulletins.
If more prescribers, including pharmacists, dentists and nurses, and pharmacy staff are made aware of the seriousness of this interaction, potential harm to patients in future will be reduced.
1 Miki A, Ohtani H, Sawada Y. Warfarin and miconazole oral gel interactions: analysis and therapy recommendations based on clinical data and a pharmacokinetic model. Journal of Clinical Pharmacy and Therapeutics 2011;36:642–50.
About the author
Sonia Filmer is a warfarin pharmacist, at Intrahealth Ltd, a community oral anticoagulant monitoring service in north east England
Case comment: Jennifer Sharp, staff editor, ‘Stockley’s drug interactions’
These case reports highlight the clinical importance of the pharmacokinetic interaction between miconazole and warfarin. Although there are no controlled studies, there are at least 15 published case reports of an interaction between miconazole oral gel and warfarin resulting in raised INRs, with bleeding in many cases. There have also been case reports in patients taking other coumarins, such as acenocoumarol, and just one report involving an indanedione (fluindione). Any interaction appears to occur within a few days but in some cases it has taken up to 15 days for bleeding to develop. It would seem prudent to avoid miconazole oral gel in patients taking coumarin anticoagulants but if concurrent use is considered essential, patients should be closely monitored, with more frequent INR checks. Dose reduction should be considered, if appropriate. Patients should be advised to report signs of bruising or bleeding to their anticoagulant clinic or GP.
Options for alternative treatments are limited. Nystatin can also be used for candidal infections. It is not known to affect cytochrome P450, so it would not be expected to affect warfarin metabolism. However, data from a recently published retrospective case study, spanning eight years, reported raised INRs and bleeding, or both, in four out of eight patients given nystatin. This is the only report of a possible interaction and, given its nature, an interaction is by no means established. At present, there are insufficient data to recommend that nystatin be avoided in patients taking coumarins but, until more is known, it would seem prudent to bear the possibility of an interaction in mind.
The risk of an interaction with other non-systemic routes of miconazole use seems low because systemic absorption of intravaginal or cutaneous miconazole is minimal. Nevertheless, there have been isolated cases of interactions with warfarin and some degree of caution is probably prudent. Increased monitoring may be appropriate, particularly in patients who might be at greater risk of systemic absorption (eg, those with severe inflammation or using large amounts).
The Stockley monograph “Coumarins and related drugs + azoles; Miconazole” is due to be updated next month and is available at www.medicines complete.com.
Citation: The Pharmaceutical Journal URI: 11099329
Recommended from Pharmaceutical Press