Pharmacy Blog – Dr Vibore

Clinical evidence,mechanism, importance and management

A woman taking warfarin for atrial fibrillation whose INR was normally within range 2 to 3 had a modest rise in her INR to 3.4, apparently due to use of 10 drops of boldo after meals and one capsule of fenugreek before meals. A week after stopping these two herbal medicines her INR had fallen to 2.6. When she restarted them,her INR rose to 3.1 after a week,and to 3.4 after 2 weeks. Her INR was later restabilised in her normal range in presence of these two medicines by reducing warfarin dosage by 15 % (See reference number 1). The mechanism of this apparent interaction remains unknown,and it is not known whether both herbs or just one was responsible for what happened. Boldo comes from Peumus boldus, and might have antiplatelet activity,and fenugreek comes from Trigonella foenumgraecum, which might contain coumarins (but see Coumarins + Herbal medicines interaction, above).

This patient had no undesirable reactions (e.g. bruising or bleeding), but this case serves to draw attention to possibility of an interaction in other patients taking anticoagulants if these herbal medicines are also taken.

1. Lambert JP,Cormier A. Potential interaction between warfarin and boldo-fenugreek. Pharmacotherapy (2001) 21, 509–12.

The anticoagulant effects of warfarin were modestly and markedly reduced,respectively, in 2 patients stabilised on warfarin when they were given griseofulvin 1 g daily in divided doses. Griseofulvin 1 g daily had no effect on prothrombin time in one healthy subject given warfarin, whereas 2 g daily caused a marked reduction in prothrombin time. Another healthy subject showed no interaction, even when griseofulvin dosage was raised to 4 g daily for 2 weeks (See reference number 1).

In another study there was no change in mean prothrombin time in 10 patients stabilised on warfarin when they were given griseofulvin 1 g daily in divided doses for 2 weeks. Four of patients had an equivocal average reduction in prothrombin time of 4.2 seconds (See reference number 2). One case report describes decreased anticoagulant effects in a man stabilised on warfarin when he took griseofulvin 250mg twice daily,which took 12 weeks to develop fully (See reference number 3). He eventually needed a 41 % increase in his daily dose of warfarin. Another report very briefly mentions a case of a coagulation defect in a patient taking warfarin and griseofulvin (See reference number 4).

Not understood. It has been suggested that griseofulvin acts as a liver enzyme inducer, which increases metabolism of warfarin, thereby reducing its effects (See reference number 1,3).

This interaction is poorly documented and not well established. It possibly affects some patients. Because of uncertainty, prothrombin times of all patients taking warfarin who are given griseofulvin should be monitored, and suitable warfarin dosage increases made as necessary.

Cullen SI,Catalano PM. Griseofulvin-warfarin antagonism. JAMA (1967) 199, 582–3.

Udall JA. Drug interference with warfarin therapy. Clin Med (1970) 77,20–5.

Okino K,Weibert RT. Warfarin-griseofulvin interaction. Drug Intell Clin Pharm (1986) 20, 291–3.

McQueen EG. New Zealand Committee on Adverse Drug Reactions: 14th Annual Report1979. N Z Med J (1980) 91,226–9.

No clinically significant interaction occurred between ezetimibeand warfarin in one study. However,raised INRs have been seenin patients taking warfarin after they were also given ezetimibe.

Clinical evidence,mechanism, importance and management

In a two-way,crossover study, 12 healthy subjects were given ezetimibe 10mg or placebo daily for 11 days, with a single 25mg dose of warfarin on day 7. The pharmacokinetics and pharmacodynamics (prothrombin time) of warfarin were not significantly altered by ezetimibe. In addition, pharmacokinetics of ezetimibe were similar to those previously seen with drug alone (See reference number 1). However, manufacturers of ezetimibe state that raised INRs have been seen in patients taking warfarin after they were also given ezetimibe. They therefore advise that INR should be monitored if ezetimibe is given with any coumarin or fluindione,(See reference number 2) (this is probably a prudent precaution for any indanedione).

Bauer KS,Kosoglou T, Statkevich P, Calzetta A, Maxwell SE, Patrick JE, Batra V. Ezetimibedoes not affect the pharmacokinetics or pharmacodynamics of warfarin. Clin Pharmacol Ther (2001) 69, P5.

Ezetrol (Ezetimibe). MSD-SP Ltd. UK Summary of product characteristics,December 2006.

In two studies in patients anticoagulant effects of warfarinwere unchanged by small to large doses of vitamin E, althoughthere is an isolated case of bleeding attributed to concurrent use.In three healthy subjects, effects of dicoumarol were slightlyincreased by vitamin E.

A study in 3 healthy subjects found that 42 units of vitamin E daily for a month increased response to a single dose of dicoumarol after 36 hrs (decrease in prothrombin activity from 52 to 33%) (See reference number 1)

In a double-blind,placebo-controlled study in 25 patients stabilised on warfarin, moderate to large daily doses of vitamin E (800 or 1200 units) for a month caused no clinically relevant changes in prothrombin times and INRs (See reference number 2). Similarly, in another study in 12 patients taking warfarin, anticoagulant effects of warfarin were unchanged by smaller daily doses of 100 or 400 units of vitamin E given for 4 weeks (See reference number 3).

However,in one case, a patient taking warfarin (and multiple other drugs) developed ecchymoses and haematuria, which was attributed to him taking 1200 units of vitamin E daily over a 2-month period. His pro-thrombin time was found to be 36 seconds. A later study in this patient showed that 800 units of vitamin E daily for 6 weeks reduced his blood clotting factor levels, increased prothrombin time from about 21 to 29 seconds, and caused ecchymoses (See reference number 4).

The suggested explanation is that vitamin E interferes with activity of vitamin K in producing blood clotting factors,(See reference number 4,5)and increases in dietary requirements of vitamin K (See reference number 6,7)

Information is limited but evidence suggests that most patients taking warfarin are unlikely to have problems if given even quite large daily doses (up to 1200 units) of vitamin E. Nevertheless isolated case cited here shows that occasionally and unpredictably warfarin effects can be changed. It has been recommended that prothrombin times should be monitored when vitamin E is first given (within 1 to 2 weeks has been rec-ommended) (See reference number 2). The same precautions could be applied to dicoumarol as well. However,as only one case of bleeding has been reported this does seem somewhat over-cautious. Information about other oral anticoagulants is lacking.

Schrogie JJ. Coagulopathy and fat-soluble vitamins. JAMA (1975) 232,19.

Kim JM,White RH. Effect of vitamin E on the anticoagulant response to warfarin. Am J Cardiol (1996) 77, 545–6.

Corrigan JJ,Ulfers LL. Effect of vitamin E on prothrombin levels in warfarin-induced vitaminK deficiency. Am J Clin Nutr (1981) 34, 1701–5.

Corrigan JJ,Marcus FI. Coagulopathy associated with vitamin E ingestion. JAMA (1974) 230, 1300–1.

Booth SL,Golly I, Sacheck JM, Roubenoff R, Dallal GE, Hamada K, Blumberg JB. Effect ofvitamin E supplementation on vitamin K status in adults with normal coagulation status. Am J Clin Nutr (2004) 80, 143–8.

Anon. Vitamin K,vitamin E and the coumarin drugs. Nutr Rev (1982) 40, 180–2.

Anon. Megavitamin E supplementation and vitamin K-dependent carboxylation. Nutr Rev (1983) 41,268–70.

A few reports suggest that glucosamine with or without chondroitin may increase INR in patients taking warfarin

Clinical evidence,mechanism, importance and management

A 69-year-old man stabilised on warfarin 47.5mg weekly had an increase in his INR from 2.58 to 4.52 four weeks after starting to take 6 capsules of Cosamin DS (glucosamine hydrochloride 500 mg,sodium chondroitin sulfate 400 mg, manganese ascorbate per capsule) daily. His warfarin dose was reduced to 40mg weekly, and his INR returned to target range of 2 to 3 (INR 2.15) with continued Cosamin DS therapy (See reference number 1). A comment on this report noted that this is twice usual dose of glucosamine (See reference number 2). The Canadian Adverse Drug Reaction Monitoring Program briefly reported that an increase in INR had been noted when glucosamine was given to patients on warfarin,and that INR values decreased when glucosamine was stopped (See reference number 3). Moreover, in 2006 CHM in UK reported that they had received 7 reports of an increase in INR in patients taking warfarin after they started taking glucosamine supplements (See reference number 4).

In contrast,a 71-year-old man stabilised on acenocoumarol 15mg weekly had a decrease in his INR to 1.6 after taking glucosamine sulfate (Xicil) 1.5 g daily for 10 days. The glucosamine was stopped and INR reached 2.1. When glucosamine was restarted, with an increase in acenocoumarol dose to 17mg weekly, INR only reached 1.9. The glucosamine was eventually stopped (See reference number 5).

There do not appear to have been any controlled studies of effects of glucosamine supplements on pharmacodynamics or pharmacokinetics of oral anticoagulants. The cases described suggest it would be prudent to monitor INR more closely if glucosamine is started. If a patient shows an unexpected change in INR, bear in mind possibility of self-medication with supplements such as glucosamine. Note that CHM in UK recommend that patients taking warfarin do not take glucosamine (See reference number 4).

Rozenfeld V,Crain JL, Callahan AK. Possible augmentation of warfarin effect by glucosamine–chondroitin. Am J Health-Syst Pharm (2004) 61, 306–7.

Scott GN. Interaction of warfarin with glucosamine—chondroitin. Am J Health-Syst Pharm (2004) 61,1186.

Canadian Adverse Drug Reaction Monitoring Program. Communiqué. Warfarin and glucosamine: interaction. Can Adverse Drug React News (2001) 11 (Apr),8.

Commission on Human Medicines/Medicines and Healthcare Products Regulatory Agency.Glucosamine adverse reactions and interactions. Current Problems (2006) 31,8.

Garrote García M,Iglesias Piñeiro MJ, Martín Álvarez R, Pérez González J. Interacción farmacológica del sulfato de glucosamina con acenocumarol. Aten Primaria (2004) 33, 162–4.

Bosentan modestly enhanced metabolism of warfarin and reduced its anticoagulant effects in one study and in one case a patient needed a 64 % increase in her warfarin dose after takingbosentan

In a double-blind,randomised, placebo-controlled, crossover study, 12 healthy subjects were given bosentan 500mg twice daily or placebo for 10 days, with a single 26mg dose of warfarin on day 6. Bosentan reduced AUC of R-warfarin by 38 % and of S-warfarin by 29%. A significant decrease in anticoagulant effects of warfarin was also noted, with a 23 % reduction in prothrombin time occurring with bosentan (See reference number 1).

One case highlights clinical significance of this interaction. A 35year-old woman taking warfarin with a stable INR of 2 to 3 over three months started taking bosentan 62.5mg twice daily. After 10 days her INR was 1.7, and remained at this level over next 4 weeks, despite an increase in her weekly warfarin dose from 27.5 to 40 mg. The bosentan dose was then increased to maintenance dose of 125mg twice daily, and two further weekly increases in warfarin dose were made. The INR was then high (3.2 to 4.1) for 3 weeks,before she was finally stabilised on warfarin 45mg each week (See reference number 2).

However, manufacturer of bosentan notes that, in clinical experience, use of bosentan with warfarin did not result in clinically relevant changes in INR or warfarin dose. There was no difference in frequency of warfarin dose changes (due to INR changes or adverse effects) between bosentan or placebo recipients (See reference number 3).

It has been suggested that bosentan induces both cytochrome P450 isoenzymes CYP3A4 and CYP2C9, which are involved in metabolism of R-warfarin and S-warfarin, respectively (See reference number 1).

Both reports suggest that a clinically significant interaction between warfarin and bosentan is possible, although exactly how frequently this may occur is unclear, since it was not detected in clinical studies. However, INR should be closely monitored in any patient taking warfarin during period that bosentan is started or stopped, or if dose is altered (See reference number 3).

Weber C,Banken L, Birnboeck H, Schulz R. Effect of the endothelin-receptor antagonistbosentan on the pharmacokinetics and pharmacodynamics of warfarin. J Clin Pharmacol (1999) 39, 847–54.

Murphey LM,Hood EH. Bosentan and warfarin interaction. Ann Pharmacother (2003) 37, 1028–31.

Tracleer (Bosentan monohydrate). Actelion Pharmaceuticals UK. UK Summary of productcharacteristics,October 2006.

An isolated report describes a very marked increase in anticoagulant effects of warfarin, accompanied by bruising and bleeding, in a patient given itraconazole. Limited evidence suggeststhat itraconazole may increase risk of over-anticoagulationwith acenocoumarol or phenprocoumon.

A woman stabilised on warfarin 5mg daily and also taking ipratropium bromide, salbutamol, budesonide, quinine sulfate and omeprazole, was given itraconazole 200mg twice daily for oral candidiasis caused by inhaled steroid. Within 4 days she developed generalised bleeding and recurrent nosebleeds. Her INR had risen to more than 8. The warfarin and itraconazole were stopped,but next day she had to be admitted to hospital for intractable bleeding and increased bruising, for which she was treated with fresh frozen plasma. Two days later when bleeding had stopped, and her INR had returned to 2.4,she was restarted on warfarin and later restabilised on her original dosage (See reference number 1).

In one cohort study in patients taking acenocoumarol or phenprocoumon, itraconazole significantly increased risk of over-anticoagulation (INR greater than 6: relative risk of 13.9,range 1.7 to 115). However, authors say this figure should be interpreted cautiously since it was based on just one case (See reference number 2).

Itraconazole is a known potent inhibitor of cytochrome P450 isoenzyme CYP3A4, but this isoenzyme is involved only in metabolism of less potent R-warfarin, and therefore inhibition would not be expected to have a marked effect on warfarin metabolism, . However,there appear to be no pharmacological studies to confirm this. Omeprazole

may also have had some minor part to play in case described (See reference number 1)

A minor to modest pharmacokinetic interaction would be predicted,but as yet there appear to be no studies to confirm this. The case report and cohort study suggest that this interaction might be clinically important in some individuals. Therefore,it would be prudent to increase monitoring of anticoagulant control when any patient on a coumarin anticoagulant is given itraconazole. Further study is needed.

Yeh J,Soo S-C, Summerton C, Richardson C. Potentiation of action of warfarin by itraconazole. BMJ (1990) 301, 669.

Visser LE,Penning-van Beest FJA, Kasbergen AAH, De Smet PAGM, Vulto AG, Hofman A,Stricker BHC. Overanticoagulation associated with combined use of antifungal agents andcoumarin anticoagulants. Clin Pharmacol Ther (2002) 71, 496–502.

The anticoagulant effects of acenocoumarol,and warfarin can bemarkedly increased if miconazole is given orally as an oral (buccal) gel, and bleeding can occur. Oral miconazole has also been reported to interact with ethyl biscoumacetate,fluindione, phenindione and tioclomarol in a few reports. The interaction hasalso rarely been seen in some women using intravaginal miconazole,and in those using a miconazole cream on skin. In one cohortstudy, use of oral miconazole markedly increased risk of overanticoagulation with acenocoumarol and phenprocoumon,whereas intravaginal miconazole caused a non-statistically significant minor increase, and cutaneous miconazole barely increasedthe risk.

In one early report,a patient with a prosthetic heart valve and stabilised on warfarin developed blood blisters and bruised easily 12 days after starting miconazole gel 250 g four times a day for a presumed fungal mouth infection. Her prothrombin time ratio had risen from less than 3 to about

She was subsequently restabilised in absence of miconazole on her

Numerous other cases of this interaction with warfarin have been reported, and, where stated, often involved use of 5 mL (125 mg) of gel four times daily for oral candidiasis (See reference number 2-10). One case of an increase in INR to 11.4 with frank haematuria and spontaneous bruising was reported in a women who had used 30 g of non-prescription miconazole (Daktarin) over 8 days (estimated daily dose of 75 mg) (See reference number 7). In 1996, New Zealand Centre for Adverse Reactions Monitoring reported 5 patients taking warfarin whose INRs rose from normal values to between 7.5 and 18 within 7 to 15 days of starting to use miconazole oral gel (See reference number 5). In 2002, Australian Adverse Drug Reactions Advisory Committee (ADRAC) stated that they had received 18 reports of this interaction. In 17 cases for which it was documented, INR was above 7.5. Eight of cases had bleeding complications, 9 required vitamin K, and 5 fresh frozen plasma (See reference number 11).

A few similar cases have also been reported for acenocoumarol(See reference number 12-14) or fluindione(See reference number 15) with miconazole oral gel. In addition, in one cohort study in patients taking acenocoumarol or phenprocoumon, use of oral miconazole (form and doses not stated) markedly increased risk of over-anticoagulation (INR greater than 6: adjusted relative risk 36.6; range 12.4 to 108). When analysed separately, adjusted relative risk was higher for acenocoumarol than phenprocoumon (35.1 versus 16.5) (See reference number 16).

In a study in 6 healthy subjects, miconazole 125mg daily for 18 days (in form of tablets) caused a very marked fivefold increase in prothrombin time response to a single dose of warfarin given on day 3. In addition, there was a threefold increase in AUC of warfarin, with S-warfarin most affected (fourfold), and R-warfarin increased 1.7-fold (See reference number 17). In one early case report with warfarin,one patient with a prosthetic heart valve and stabilised on warfarin was found to have a prothrombin time ratio of 23.4 within 10 days of starting miconazole tablets 250mg four times a day for a suspected fungal diarrhoea. He developed two haematomas soon after both drugs were withdrawn, and was subsequently restabilised, in absence of miconazole, on his former dose of warfarin (See reference number 1).

The Centres de Pharmacovigilance Hospitalière in Bordeaux have on record 5 cases where miconazole (oral doses of 500mg daily,where stated; form not mentioned) was responsible for a marked increase in pro-thrombin times and/or bleeding (haematomas, haematuria, gastrointestinal bleeding) in patients taking acenocoumarol (2 cases), ethyl biscoumacetate (1 case), tioclomarol (1 case) and phenindione (1 case) (See reference number 18). Other cases and reports of this interaction involving acenocoumarol have been described elsewhere (See reference number 19-21).

An 80-year-old man stabilised on warfarin with an INR of 2.2 to 3.1 was found to have an INR of 21.4 at a routine check 2 weeks after starting to use miconazole cream for a fungal infection in his groin. He showed no evidence of bruising or bleeding (See reference number 22). In 2001,Health Canada reported that they had on record a case of an 80-year-old man taking warfarin and using topical miconazole who had a cerebral vascular accident, although this case was complicated by multiple medical conditions and medications (See reference number 23). In 2002, Australian Adverse Drug Reactions Advisory Committee stated that they had received one report of an interaction involving topical miconazole cream (See reference number 11).

In one cohort study in patients taking acenocoumarol or phenprocoumon,use of topical miconazole was associated with a small increased risk of over-anticoagulation (INR greater than 6: adjusted relative risk 1.4) but this was not statistically significant. Note that this was markedly less than increased risk seen with oral miconazole (relative risk 36.6) (See reference number 16).

In 1999, Netherlands Pharmacovigilance Foundation LAREB reported 2 elderly women patients taking acenocoumarol whose INRs rose sharply and rapidly when they were given a 3-day course of 400mg miconazole pessaries (See reference number 24). Another report describes development of bruising and an INR of about 9.78 in a 55-year-old woman taking warfarin on third day of using 200mg miconazole pessaries. For a subsequent course of intravaginal miconazole 100mg daily for 7 days, dose of warfarin was decreased by 28%, and her INR was 3.27 (See reference number 25). Yet another report describes haemorrhage of kidney in a 52-year old woman taking warfarin after she used vaginal miconazole for 12 days (See reference number 23). In one cohort study in patients taking acenocoumarol or phenprocoumon,use of vaginal mico-

Note that this was markedly less than increased risk seen with oral miconazole (relative risk 36

There is evidence that miconazole is a very potent inhibitor of metabolism of S-warfarin by cytochrome P450 isoenzyme CYP2C9, and that it also inhibits metabolism of R-warfarin to a lesser extent (See reference number 17). Even low oral doses of miconazole (125 mg daily) markedly inhibit warfarin metabolism, so it is not surprising that prescription doses of miconazole oral gel (480 to 960mg daily) interact, since this is swallowed after retaining in mouth. Very unusually, absorption of miconazole from vagina (see also comments below) and even exceptionally through skin, can result in increased anticoagulant effects.

The interaction of miconazole oral gel and miconazole tablets with and coumarin anticoagulants is a very well established and potentially serious interaction. Most of reports are about warfarin or acenocoumarol, but many other oral anticoagulants have been implicated. In some cases bleeding has taken 7 to 15 days to develop,(See reference number 1,3,18) whereas others have bled within only 3 days (See reference number 20,25). Raised INRs have been seen even sooner. Usual prescription doses of miconazole oral gel [5 to 10 mL (120 to 240 mg) four times daily] should therefore not be given to patients taking any oral anticoagulant unless prothrombin times can be closely monitored and suitable dosage reductions made. Given very large increased relative risk of over-anticoagulation seen in one cohort study, authors suggest that concurrent use of oral miconazole and coumarins should be discouraged (See reference number 16). The interaction has been seen with a lower oral dose of about 75mg daily (one 30 g tube given over 8 days), which is not surprising in context of pharmacokinetic study, and suggests that patients taking oral anticoagulants should also avoid using non-prescription miconazole. Nevertheless, UK patient information leaflet for non-prescription Daktarin oral gel contains no specific cautions regarding anticoagulants (See reference number 26). Nystatin and amphotericin are possible alternative antifungals to miconazole for mouth infections.

An interaction with intravaginal miconazole would not normally be expected because its systemic absorption is usually very low (less than 2%) in healthy women of child-bearing age (See reference number 27). However, reports cited above show that significant absorption could apparently occur in a few patients with particular conditions (possibly in postmenopausal women with inflamed vaginal tissue), which allows an interaction to occur. Appropriate monitoring is therefore needed even with this route of administration in potentially at-risk women.

Topical (cutaneous) miconazole would also not be expected to interact, but few reports cited shows that some caution might be warranted.

Watson PG,Lochan RG, Redding VJ. Drug interaction with coumarin derivative anticoagulants. BMJ (1982) 285, 1044–5

Colquhoun MC,Daly M, Stewart P, Beeley L. Interaction between warfarin and miconazoleoral gel. Lancet (1987) i, 695–6.

Bailey GM,Magee P, Hickey FM, Beeley L. Miconazole and warfarin interaction. Pharm J (1989) 242, 183.

Shenfield GM,Page M. Potentiation of warfarin action by miconazole oral gel. Aust N Z J Med (1991) 21, 928.

Pillans P,Woods DJ. Interaction between miconazole oral gel (Daktarin) and warfarin. N Z Med J (1996) 109, 346.

Ariyaratnam S,Thakker NS, Sloan P, Thornhill MH. Potentiation of warfarin anticoagulantactivity by miconazole oral gel. BMJ (1997) 314, 349.

Evans J,Orme DS, Sedgwick ML, Youngs GR. Treating oral candidiasis: potentially fatal.Br Dent J (1997) 182, 452.

Marco M,Guy AJ. Retroperitoneal haematoma and small bowel intramural haematomacaused by warfarin and miconazole interaction. Int J Oral Maxillofac Surg (1997) 27, 485.

Pemberton MN,Sloan P, Ariyaratnam S, Thakker NS, Thornhill MH. Derangement of warfarin anticoagulation by miconazole oral gel. Br Dent J (1998) 184, 68–9.

Øgard CG,Vestergaad H. Interaktion mellem warfarin og oral miconazol-gel. Ugeskr Laeger (2000) 162, 5511.

ADRAC. Miconazole oral gel elevates INR –a reminder. Aust Adverse Drug React Bull (2002) 21; 14–15.

Marotel C,Cerisay D, Vasseur P, Rouvier B, Chabanne JP. Potentialisation des effets del’acénocoumarol par le gel buccal de miconazole. Presse Med (1986) 15, 1684–5.

Ducroix JP,Smail A, Sevenet F, Andrejak M, Baillet J. Hématome oesophagien secondaireà une potentialisation des effets de l’acénocoumarol par le gel buccal de miconazole. Rev Med Interne (1989) 10, 557–9.

Ortín M,Olalla JI, Muruzábal MJ, Peralta FG, Gutiérrez MA. Miconazole oral gel enhancesacenocoumarol anticoagulant activity: a report of three cases. Ann Pharmacother (1999) 33, 175–77.

Ponge T,Rapp MJ, Fruneau P, Ponge A, Wassen-Hove L, Larousse C, Cottin S. Interactionmédicamenteuse impliquant le miconazole en gel et la fluindione. Therapie (1987) 42, 412–

13.

Visser LE,Penning-van Beest FJA, Kasbergen AAH, De Smet PAGM, Vulto AG, HofmanA, Stricker BHC. Overanticoagulation associated with combined use of antifungal agents andcoumarin anticoagulants. Clin Pharmacol Ther (2002) 71, 496–502.

O’Reilly RA,Goulart DA, Kunze KL, Neal J, Gibaldi M, Eddy AC, Trager WF. Mechanismsof the stereoselective interaction between miconazole and racemic warfarin in human subjects. Clin Pharmacol Ther (1992) 51, 656–67.

Loupi E,Descotes J, Lery N, Evreux J C. Interactions médicamenteuses et miconazole. Therapie (1982) 37, 437–41

Anon. New possibilities in the treatment of systemic mycoses. Reports on the experimentaland clinical evaluation of miconazole. Round table discussion and Chairman’s summing up.Proc R Soc Med (1977),70 (Suppl l), 51–4.

Ponge T,Barrier J, Spreux A, Guillou B, Larousse C, Grolleau JY. Potentialisation des effetsde l’acénocoumarol par le miconazole. Therapie (1982) 37, 221–2.

Goenen M,Reynaert M, Jaumin P, Chalant CH, Tremouroux J. A case of candida albicansendocarditis three years after an aortic valve replacement. J Cardiovasc Surg (1977) 18, 391–

6.

Devaraj A,O’Beirne JPO, Veasey R, Dunk AA. Interaction between warfarin and topical miconazole cream. BMJ (2002) 325, 77.

Anon. Miconazole-warfarin interaction: increased INR. Can Med Assoc J (2001) 165,81.

Lansdorp D,Bressers HPHM, Dekens-Konter JAM, Meyboom RHB. Potentiation of acenocoumarol during vaginal administration of miconazole. Br J Clin Pharmacol (1999) 47, 225–

6.

Thirion DJ,Zanetti LAF. Potentiation of warfarin’s hypoprothrombinemic effect with miconazole vaginal suppositories. Pharmacotherapy (2000) 20, 98–9.

Daktarin Oral Gel (Miconazole). Janssen-Cilag Ltd. UK Patient information leaflet,December 2003.

Daneshmend TK. Systemic absorption of miconazole from the vagina. J Antimicrob Chemother (1986) 18,507–11.

Aprepitant modestly reduces warfarin levels and slightly decreases INR in healthy subjects

Clinical evidence,mechanism, importance and management

In a double-blind study,healthy subjects were stabilised on warfarin and then given either aprepitant (125 mg on day one, then 80mg daily on days 2 and 3) or placebo. On day 3,there was no change in warfarin levels. However, by day 8 (5 days after stopping aprepitant) there was a 34 % decrease in trough S-warfarin levels, and a 14 % decrease in INR in aprepitant group (See reference number 1).

Aprepitant is an inducer of cytochrome P450 isoenzyme CYP2C9, by which S-warfarin is metabolised. The manufacturer recommends that, in patients taking warfarin, INR should be monitored closely for 2 weeks, particularly at 7 to 10 days,(See reference number 2) after each 3-day course of aprepitant,(See reference number 2,3) and this seems a prudent precaution. They similarly recommend caution with acenocoumarol,which is also metabolised by CYP2C9 (See reference number 3).

1. Depré M,Van Hecken A, Oeyen M, De Lepeleire I, Laethem T, Rothenberg P, Petty KJ, Majumdar A, Crumley T, Panebianco D, Bergman A, de Hoon JN. Effect of aprepitant on thepharmacokinetics and pharmacodynamics of warfarin. Eur J Clin Pharmacol (2005) 61, 341–

6.

Emend (Aprepitant). Merck & Co.,Inc. US Prescribing information, June 2006.

Emend (Aprepitant). Merck Sharp & Dohme Ltd. UK Summary of product characteristics,February 2007.

A number of case reports describe an increase in effects ofwarfarin, accompanied by bleeding in some cases, caused by antineoplastic regimens containing carboplatin, chlormethine, cyclophosphamide, doxorubicin, etoposide, gefitinib, gemcitabine,ifosfamide with mesna, methotrexate, procarbazine, trastuzumab, vincristine or vindesine. A decrease in effects of warfarin has been seen with azathioprine, cyclophosphamide, mercaptopurine and mitotane, a decrease in effects of acenocoumarol has been seen with mercaptopurine, and a decrease in effectsof phenprocoumon have been seen with azathioprine.

A survey of 103 patients with antiphospholipid syndrome found that azathioprine appeared to increase warfarin requirements.(See reference number 1) A woman who was resistant to warfarin, needing 14 to 17mg daily while taking azathioprine, began to bleed (epistaxes, haematemesis) when azathioprine was stopped. She was restabilised on warfarin 5mg daily (See reference number 2). Reduced warfarin effects were seen in 2 other patients taking azathioprine,(See reference number 3,4) one of whom had a marked fall in serum warfarin levels during azathioprine treatment (See reference number 4). Two women with systemic lupus erythematosus taking phenprocoumon(See reference number 5) and a third taking warfarin(See reference number 6) had marked falls in their INRs during treatment with azathioprine, and another woman needed an almost fourfold increase in dose of warfarin when she was given azathioprine (See reference number 7).

A woman stabilised on warfarin required a gradual decrease in dose from 4mg to 2.5mg daily after starting gefitinib 250 mg/m(See reference number 2) daily for lung cancer. In contrast,a second patient did not have an increase in warfarin effect while taking gefitinib (See reference number 11).

A 63-year-old man needed a reduction in his weekly warfarin dosage from 59.23mg to 50.75mg in order to keep his INR at about 2.5 during 2 cycles of gemcitabine. When gemcitabine was stopped his warfarin dosage had to be increased again (See reference number 12). The manufacturers have information on 4 cases of suspected interactions between gemcitabine and warfarin,and one with phenprocoumon (reported by December 2000) (See reference number 13). Based on 724 reports of concurrent use of gemcitabine and anticoagulants,(See reference number 13)they suggest that incidence of suspected interaction is 0.8%.

A man well stabilised on warfarin had a marked reduction in his anticoagulant response on two occasions while taking mercaptopurine,but no changes occurred when he took busulfan, cyclophosphamide, cytarabine, hydroxycarbamide, mitobronitol, demecolcine or melphalan (See reference number 15). A woman needed a marked increase in her dosage of acenocoumarol,from 21 to 70mg weekly, when she was given mercaptopurine 100mg daily (See reference number 16). Another patient required about a 25 % increase in warfarin dose while taking mercaptopurine 100mg daily (See reference number 17).

The prothrombin times of an elderly man given warfarin increased by 50 to 100 % in middle of three cycles of treatment with ProMace-Mopp (cyclophosphamide, doxorubicin, etoposide, chlormethine, vincristine, procarbazine, methotrexate and prednisone), and he developed a subconjunctival haemorrhage during first cycle (See reference number 19).

Two women stabilised on warfarin developed nosebleeds after 10 and 8 weekly doses of trastuzumab,respectively, and were found to have INRs of 6 and 5.8,respectively (See reference number 20). However, manufacturer notes that in an analysis of clinical study data rate of bleeding events was similar for patients receiving or not receiving trastuzumab, with or without anticoagulants (See reference number 21).

Not well understood. Mercaptopurine possibly increases synthesis or activation of prothrombin (See reference number 22). Azathioprine is metabolised to mercaptopurine,and would therefore be expected to interact similarly.

The absence of problems in early small studies using warfarin as an adjunct to chemotherapy,(See reference number 23,24) and small number of reports describing difficulties, suggest that many of these interactions may be uncommon events. The concurrent use of these drugs need not be avoided but there is clearly a need to be aware that any antineoplastic regimen might increase response to anticoagulants. It would also be prudent to note that mercaptopurine and azathioprine may decrease anticoagulant response. The anticoagulant dosages may need adjustment. Note that,from a disease perspective, when treating venous thromboembolic disease in patients with cancer, warfarin is generally inferior (higher risk of major bleeds and recurrent thrombosis) to low-molecular-weight heparins (See reference number 25).

Khamashta MA,Cuadrado MJ, Mujic F, Taub N, Hunt BJ, Hughes GRV. Effect of azathioprine on the anticoagulant activity of warfarin in patients with the antiphospholipid syndrome. Lupus (1998) 7 (Suppl 2), S227.

Singleton JD,Conyers L. Warfarin and azathioprine: an important drug interaction. Am J Med (1992) 92, 217.

Rivier G,Khamashta MA, Hughes GRV. Warfarin and azathioprine: a drug interaction doesexist. Am J Med (1993) 95, 342.

Rotenberg M,Levy Y, Shoenfeld Y, Almog S, Ezra D. Effect of azathioprine on the anticoagulant activity of warfarin. Ann Pharmacother (2000) 34, 120–2.

Jeppesen U,Rasmussen JM, Brøsen K. Clinically important interaction between azathioprine(Imurel) and phenprocoumon (Marcoumar). Eur J Clin Pharmacol (1997) 52, 503–4.

Walker J,Mendelson H, McClure A Smith MD. Warfarin and azathioprine: clinically significant drug interaction. J Rheumatol (2002) 29, 398–9.

Havrda DE,Rathbun S, Scheid D. A case report of warfarin resistance due to azathioprineand review of the literature. Pharmacotherapy (2001) 21, 355–7.

Le AT,Hasson NK, Lum BL. Enhancement of warfarin response in a patient receiving etoposide and carboplatin chemotherapy. Ann Pharmacother (1997) 31, 1006–8.

Tashima CK. Cyclophosphamide effect on coumarin anticoagulation. South Med J (1979) 72,633–4.

Ward K,Bitran JD. Warfarin, etoposide, and vindesine interactions. Cancer Treat Rep (1984)68, 817–18.

Onoda S,Mitsufuji H, Yanase N, Ryuge S, Kato E, Wada M, Ishii K, Hagiri S, YamamotoM, Yokoba M, Yanaihara T, Kuboto M, Takada N, Katagiri M, Abe T, Tanaka M, KobayashiH, Masuda N. Drug interaction between gefitinib and warfarin. Jpn J Clin Oncol (2005) 35, 478–82.

Kinikar SA,Kolesar JM. Identification of a gemcitabine-warfarin interaction. Pharmacotherapy (1999) 19, 1331–3.

Kilgour-Christie J,Czarnecki A. Gemcitabine and the interaction with anticoagulants. Lancet Oncol (2002) 3, 460.

Hall G,Lind MJ, Huang M, Moore A, Gane A, Roberts JT, Cantwell BMJ. Intravenous infusions of ifosfamide/mesna and perturbation of warfarin anticoagulant control. Postgrad Med J (1990) 66, 860–1.

Spiers ASD,Mibashan RS. Increased warfarin requirement during mercaptopurine therapy:a new drug interaction. Lancet (1974) ii, 221–2.

Fernádez MA,Regadera A, Aznar J. Acenocoumarol and 6-mercaptopurine: an importantdrug interaction. Haematologica (1999) 84, 664–5.

Martin LA,Mehta SD. Diminished anticoagulant effects of warfarin with concomitant mercaptopurine therapy. Pharmacotherapy (2003) 23, 260–4.

Cuddy PG,Loftus LS. Influence of mitotane on the hypoprothrombinemic effect of warfarin.South Med J (1986) 79, 387–8.

Seifter EJ,Brooks BJ, Urba WJ. Possible interactions between warfarin and antineoplasticdrugs. Cancer Treat Rep (1985) 69, 244–5.

Nissenblatt MJ,Karp GI. Bleeding risk with trastuzumab (Herceptin) treatment. JAMA (1999) 282, 2299–2300.

Stewart SJ. Bleeding risk with trastuzumab (Herceptin) treatment. In reply. JAMA (1999)282,2300–1.

Martini A,Jähnchen E. Studies in rats on the mechanisms by which 6-mercaptopurine inhibitsthe anticoagulant effect of warfarin. J Pharmacol Exp Ther (1977) 201, 547–53.

Zacharski LR,Henderson WG, Rickles FR, Forman WB, Cornell CJ, Forcier RJ, Edwards RL, Headley E, Kim S-H, O’Donnell JR, O’Dell R, Tornyos K, Kwaan HC. Effect of warfarin anticoagulation on survival in carcinoma of the lung, colon, head and neck, and prostate.Cancer (1984) 53, 2046–52.

Zacharski LR,Henderson WG, Rickles FR, Forman WB, Cornell CJ, Forcier RJ, Edwards R, Headley E, Kim S-H, O’Donnell JR, O’Dell R, Tornyos K, Kwaan HC. Effect of warfarin onsurvival in small cell carcinoma of the lung. JAMA (1981) 245, 831–5.

Baglin TP,Keeling DM, Watson HG, for the British Committee for Standards in Haematology. Guidelines on oral anticoagulation (warfarin): third edition – 2005 update. Br J Haematol (2006) 132, 277–85.