Pharmacy Blog – Dr Vibore

In healthy subjects, tolterodine did not alter pharmacokinetics or pharmacodynamics of warfarin, but isolated cases of raisedINRs have been reported.

Clinical evidence,mechanism, importance and management

In a placebo-controlled study in healthy subjects, tolterodine 2mg twice daily for 7 days did not affect pharmacokinetics of R-or S-warfarin after a single 25mg dose of warfarin given on day 4, nor were pharmacokinetics of tolterodine altered by warfarin. In addition, prothrombin time and factor VII activity were not altered by tolterodine (See reference number 1).

However,a report describes two patients on stable warfarin doses who had elevated INRs shortly after tolterodine 2mg daily was started and stopped: one had an INR of 6.1 one day after stopping 13 days of tolterodine, and other had an INR of 7.4 two days after stopping 8 days of tolterodine. In both cases no bleeding occurred and warfarin was withheld for three doses and successfully reinstated at original dose (See reference number 2). Another case has been reported of a patient who required a 15 % reduction in warfarin dose while taking tolterodine 4mg daily. Her maximum INR had been 3.9. However, she had undergone several warfarin dose increases over previous 2 months and her INR had been fluctuating (See reference number 3).

The controlled study suggests that no warfarin dose adjustment would expected to be needed when tolterodine is added to warfarin therapy. However, 3 cases introduce a note of caution. Although additional monitoring would seem over-cautious on basis of these cases, bear them in mind in case of an unexpected response to warfarin.

Rahimy M,Hallén B, Narang P. Effect of tolterodine on the anticoagulant actions and pharmacokinetics of single-dose warfarin in healthy volunteers. Arzneimittelforschung (2002) 52, 890–5.

Colucci VJ,Rivey MP. Tolterodine-warfarin drug interaction. Ann Pharmacother (1999) 33, 1173–6.

Taylor JR. Probable interaction between tolterodine and warfarin. Pharmacotherapy (2006)26,719–21.

The activated clotting time (ACT) may not be a reliable methodto monitor heparin therapy when aprotinin is used concurrently.This is because aprotinin increases ACT monitored by somemethods, without actually increasing anticoagulation.

Clinical evidence,mechanism, importance and management

Aprotinin prolongs activated clotting time (ACT) as measured by a celite surface activation method, although kaolin ACT is much less affected (See reference number 1). Therefore, if ACT is used to monitor effectiveness of heparin anticoagulation during cardiopulmonary bypass incorporating aprotinin, this may lead to an overestimation of degree of anticoagulation. This may result in patients not receiving additional necessary heparin during extended extracorporeal circulation, or receiving excess protamine to reverse effects of heparin at end of procedure. The UK manufacturer of aprotinin notes that it is not necessary to adjust usual heparin/protamine regimen used in cardiopulmonary bypass procedures when aprotinin is also used (See reference number 2). The US manufacturer provides additional detailed information on appropriate methods to monitor heparin anticoagulation in presence of aprotinin (See reference number 1).

Trasylol (Aprotinin). Bayer HealthCare. US Prescribing information,December 2003.

Trasylol (Aprotinin). Bayer plc. UK Summary of product characteristics,September 2006.

A woman taking warfarin with a stable INR (mean 2.2 over previous year) developed petechiae and purpura 55 days after starting duloxetine 30mg daily, and was found to have an INR of 5. Warfarin was stopped on day 58, but INR continued to rise to greater than 19 on day 85, and she was given vitamin K. On day 94 duloxetine was stopped. Warfarin was restarted on day 110 and by day 140 INR was 2.2 with thewarfarin dose stabilised at original level (See reference number 1).

The possible interactions of warfarin or other anticoagulants with venlafaxine do not appear to have been studied, but, as of May 2000, manufacturers had on record 6 case reports of increased prothrombin times, raised INRs and bleeding (haematuria, gastrointestinal bleeding, melaena, haemarthrosis) in patients taking warfarin with venlafaxine (See reference number 2).

Just why these adverse interactions should have occurred is not understood,especially as no pharmacokinetic interaction is thought likely. Venlafaxine alone may uncommonly cause ecchymosis and mucosal bleeding and,rarely, prolonged bleeding time and haemorrhage (See reference number 3). However, given many other factors that can influence anticoagulant control, reports of possible interactions could just represent idiosyncratic cases.

No interaction is established, and general relevance of these unpublished cases is uncertain. If one subscribes to view that increased monitoring is necessary when any drug is started or stopped in a patient on warfarin or related drugs, then it would be prudent to monitor prothrombin times with venlafaxine. More study is needed.

Glueck CJ,Khalil Q, Winiarska M, Wang P. Interaction of duloxetine and warfarin causing severe elevation of international normalized ratio. JAMA (2006) 295, 1517–18.

Wyeth Laboratories. Personal communication,May 2000.

Efexor (Venlafaxine hydrochloride). Wyeth Pharmaceuticals. UK Summary of product characteristics,May 2006.

Case reports describe reduced warfarin effects in a patient givenetretinate,and in a patient given isotretinoin. Acitretin did notsignificantly alter anticoagulant effects of phenprocoumon inhealthy subjects.

Acitretin 50mg daily for 10 days slightly increased Quick test of 10 healthy subjects stabilised on phenprocoumon, from 22 to 24%, and corresponding INR value decreased from 2.91 to 2.71. However,these changes were not considered to be significant (See reference number 1).

Etretinate. A man with T-cell lymphoma who had recently been given chemotherapy (cyclophosphamide,doxorubicin, vincristine and prednisolone) was anticoagulated with warfarin after developing a pulmonary embolism. About three weeks later,he started etretinate 40mg daily and it was found necessary to increase his warfarin dosage from 7 to 10mg daily. His liver function tests were normal (See reference number 2). This patient had also recently started taking co-proxamol,, tolbutamide, and cimetidine, ’, but all of these have been reported to only rarely increase the effect of warfarin.

Isotretinoin. A 61-year-old man stabilised on warfarin 2.5mg daily for 2 to 3 years had a decrease in his INR to below 2.5 after starting oral cefpodoxime proxetil 200mg twice daily and oral isotretinoin 30mg daily for inflammatory lesions of face. He required an increase in warfarin dose to 3.75mg daily. The cefpodoxime was stopped after 10 days without a further change in warfarin requirement. However, when isotretinoin was discontinued after 40 days, INR progressively increased and warfarin dose was eventually reduced to pretreatment dose of 2.5mg daily (See reference number 3).

It has been suggested that etretinate or isotretinoin may increase rate of metabolism of warfarin (See reference number 2,3)

The clinical relevance of two case reports of a modest increase in warfarin requirements on

starting etretinate or isotretinoin is uncertain, but, until more is known, consideration could be given to monitoring INR if patients are given warfarin and these retinoids. The study with acitretin suggests that no phenprocoumon dose adjustments are expected to be needed on starting acitretin.

Hartmann D,Mosberg H, Weber W. Lack of effect of acitretin on the hypoprothrombinemicaction of phenprocoumon in healthy volunteers. Dermatologica (1989) 178, 33–6.

Ostlere LS,Langtry JAA, Jones S, Staughton RCD. Reduced therapeutic effect of warfarincaused by etretinate. Br J Dermatol (1991) 124, 505–10.

Fiallo P. Reduced therapeutic activity of warfarin during treatment with oral isotretinoin. Br J Dermatol (2004) 150,164.

Orlistat had no effect on pharmacodynamics or pharmacokinetics of single-dose warfarin in healthy subjects. However,orlistat reduces fat absorption, and might therefore reduce vitamin Kabsorption. There is a published report of a patient taking warfarin who developed a modest increase in INR after taking orlistat.Similar cases have been reported to regulatory authorities.

In a placebo-controlled,randomised, crossover study, 12 healthy subjects were given orlistat 120mg three times daily for 16 days, with a single 30mg dose of warfarin on day 11. The pharmacokinetics and pharmacodynamics of warfarin were not altered by orlistat, and markers of vitamin K nutritional status were not affected.(See reference number 1) However, regarding this study, manufacturers US prescribing information states that vitamin K levels did tend to decline in subjects taking orlistat.(See reference number 2) It is also noted that reports of “decreased prothrombin,increased INR and unbalanced anticoagulant treatment resulting in change of haemostatic parameters” have been reported in patients taking orlistat and anticoagulants.(See reference number 2) In addition, in 2001 Canadian regulatory authorities reported that unexpected increases in INR were noted after orlistat was given to patients taking either warfarin or acenocoumarol. These were managed by dosage adjustments of coumarin or dis-continuation of orlistat (See reference number 3).

In a published report,a 66-year-old man stabilised on warfarin for

2.5 years who started taking orlistat 120mg three times daily for weight reduction had a modest increase in his INR,from less than 3, to 4.7 within 18 days. Warfarin was withheld and he was later restabilised on approximately two-thirds of previous dose while continuing orlistat (See reference number 4).

Orlistat may reduce absorption of fat soluble vitamins including vitamin K,(See reference number 4,5) and a change to a lower fat diet associated with use of orlistat may also contribute to changes in balance between vitamin K and war

The manufacturers say that patients stabilised on anticoagulants and given orlistat should be closely monitored for changes in coagulation parameters (See reference number 2,5). Given reports of changes in INRs, and fact that changes in dietary vitamin K, are known to affect warfarin efficacy, this seems prudent in patients taking a coumarin or an indanedione.

Zhi J,Melia AT, Guerciolini R, Koss-Twardy SG, Passe SM, Rakhit A, Sadowski JA. The effect of orlistat on the pharmacokinetics and pharmacodynamics of warfarin in healthy volunteers. J Clin Pharmacol (1996) 36, 659–666.

Xenical (Orlistat). Roche Pharmaceuticals. US Prescribing information,January 2007.

Canadian Adverse Drug Reaction Monitoring Program. Communiqué. Orlistat (Xenical) interaction with coumarin derivatives: increased INR. Can Adverse Drug React News (2001) 11 (Jul),7.

MacWalter RS,Fraser HW, Armstrong KM. Orlistat enhances warfarin effect. Ann Pharmacother (2003) 37, 510–12.

Xenical (Orlistat). Roche Products Ltd. UK Summary of product characteristics,May 2006.

In pharmacological studies,sulindac did not significantly alterthe anticoagulant effect of warfarin or phenprocoumon. Isolatedcases of a modest to marked increase in anticoagulant effectsof warfarin have been reported with sulindac. Note that all NSAIDs increase risk of gastrointestinal bleeding, and anincreased risk is seen when they are combined with anticoagulants.

In a study in healthy subjects stabilised on warfarin, sulindac 200mg twice daily for 7 days did not significantly alter prothrombin time, when compared with placebo, although prothrombin time was slightly higher in sulindac group (See reference number 1). Similarly,in 20 patients stabilised on phenprocoumon, sulindac 200 to 400mg daily for 4 weeks did not alter measures of coagulation or bleeding time (See reference number 2).

However,a patient stabilised on warfarin, ferrous sulfate, phenobarbital and sulfasalazine had a marked increase in his prothrombin time ratio from about 3.2 to 10 after taking sulindac 100mg twice daily for 5 days (See reference number 3-5). There are 4 similar cases of this interaction on record (See reference number 1,5-7). One of patients had a gastrointestinal bleed after taking only three 100mg doses of sulindac, although this patient was also taking flurbiprofen (See reference number 5). Another patient was stabilised on about a 40 % lower dose of warfarin with continuation of sulindac (See reference number 6). Another patient had a potassium-losing renal tubular defect, which was thought to contribute to interaction (See reference number 1).

Not understood. In one patient,renal impairment may have caused sulindac accumulation, which in turn may have affected warfarin pharmacokinetics (See reference number 1). See also Coumarins and related drugs + NSAIDs interaction.

The pharmacological studies cited suggest that usually no coumarin dose adjustment would be needed in patients given sulindac. However, isolated cases of an interaction suggest that, rarely, some patients may be affected. Also note that all NSAIDs, can irritate gastric mucosa, affect platelet activity and cause gastrointestinal bleeding, which will be more severe in anticoagulated patients.

Loftin JP,Vesell ES. Interaction between sulindac and warfarin: different results in normal subjects and in an unusual patient with a potassium-losing renal tubular defect. J Clin Pharmacol (1979) 19, 733–42.

Schenk H,Klein G, Haralambus J, Goebel R. Coumarintherapie unter dem antirheumaticumsulindac. Z Rheumatol (1980) 39, 102–8.

Beeley L. Bulletin of the West Midlands Centre for Adverse Drug Reaction Reporting (1978) No. 6.

Beeley L,Baker S. Personal communication, 1978.

Ross JRY,Beeley L. Sulindac, prothrombin time, and anticoagulants. Lancet (1979) ii, 1075.

Carter SA. Potential effect of sulindac on response of prothrombin-time to oral anticoagulants.Lancet (1979) ii,698–9.

McQueen EG. New Zealand Committee on Adverse Drug Reactions. 17th Annual Report1982. N Z Med J (1983) 96,95–9.

Methaqualone may cause a very small and clinically unimportantreduction in anticoagulant effects of warfarin

Clinical evidence,mechanism, importance and management

20.9 seconds before,20.4 seconds during,and 19.6 seconds after taking methaqualone 300mg at bedtime for 4 weeks (See reference number 1). The plasma warfarin levels of another patient were unaffected by methaqualone,although there was some evidence that enzyme induction had occurred (See reference number 2). Methaqualone has some enzyme-inducing effects so that any small changes in pro-thrombin times reflect a limited increase in metabolism and clearance of warfarin, but these appear to be too small to be of clinical significance (See reference number 2,3). No special precautions seem to be necessary.

Udall JA. Clinical implications of warfarin interactions with five sedatives. Am J Cardiol (1975) 35,67–71.

Whitfield JB,Moss DW, Neale G, Orme M, Breckenridge A. Changes in plasma :5.5pt; font-weight:normal; color:#000000″>γ-glutamyl transpeptidase activity associated with alterations in drug metabolism in man. BMJ (1973) 1, 316–18.

Nayak RK,Smyth RD, Chamberlain AP, Polk A, DeLong AF, Herczeg T, Chemburkar PB,Joslin RS, Reavey-Cantwell NH. Methaqualone pharmacokinetics after single- and multiple-dose administration in man. J Pharmacokinet Biopharm (1974) 2, 107–21.

In a study,lasofoxifene caused a minor 16 % decrease in pro-thrombin time without changing warfarin pharmacokinetics.

Clinical evidence,mechanism, importance and management

In 12 healthy postmenopausal women, lasofoxifene 4mg on day one then 500 micrograms daily for 13 days had no effect on pharmacokinetics of R-or S-warfarin when a single 20mg dose of warfarin was given on day 8. Conversely, maximum prothrombin time was decreased by 16%, with a similar decrease in maximum INR from 1.9 to 1.6 (See reference number 1).

Because this slight decrease in warfarin effects has been seen with raloxifene, , authors suggested that it might be because oestrogenic compounds increase plasma concentrations of vitamin K-dependent clotting factors (See reference number 1).

The authors suggest that small decrease in warfarin effect with lasofoxifene may not be clinically relevant. Nevertheless,they say that until more data are available on longer-term concurrent use, it is recommended that prothrombin times should be monitored when starting or stopping lasofoxifene (See reference number 1). This seems a sensible precaution.

1. Quellet D,Bramson C, Carvajal-Gonzalez S, Roman D, Randinitis E, Remmers A, GardnerMJ. Effects of lasofoxifene on the pharmacokinetics and pharmacodynamics of single-dosewarfarin. Br J Clin Pharmacol (2006) 61, 741–5.

St John’s wort can cause a moderate reduction in anticoagulant effects of phenprocoumon and warfarin

In a randomised, placebo-controlled crossover study in 10 healthy men,(See reference number 1)St John’s wort extract (LI 160, Lichtwer Pharma) 900mg daily for 11 days reduced AUC of a single 12mg dose of phenprocoumon by a modest 17.4%. There is also a case report of a 75-year-old woman taking phenprocoumon who had a reduced anticoagulant response (a rise in Quick value) 2 months after starting to take St John’s wort (See reference number 2).

In a randomised, crossover study in 12 healthy subjects, one tablet of St John’s wort three times daily for 3 weeks modestly decreased AUC of both R- and S-warfarin by about 25 % after a single 25mg dose of warfarin taken on day 14. In this study, brand of St John’s wort used was Bioglan tablets, each tablet containing an extract equivalent to 1 g of Hypericum perforatum flowering herb top containing 825 micrograms of hypericin and 12.5mg of hyperforin (See reference number 3).

The Swedish Medical Products Agency received 7 case reports over 1998 to 1999 period of patients stabilised on warfarin who showed decreased INRs when St John’s wort was added. Their INRs fell from normal therapeutic range of about 2 to 4 to about 1. The INRs of 4 of patients returned to their former values when St John’s wort was stopped (See reference number 4)

Uncertain, but it is suggested that St John’s wort increases metabolism and clearance of anticoagulants(See reference number 1,3,4) possibly by induction of cytochrome P450 isoenzymes. It affected both R– and S-warfarin (See reference number 3). Oral bioavailability was not altered (See reference number 3).

Information seems to be limited to these reports,but a modest pharmacokinetic interaction is established, which might be clinically important. It would be prudent to monitor INRs of patients taking phenprocoumon, warfarin or any other coumarin if they start taking St John’s wort, being alert for need to slightly raise anticoagulant dosage. However, note that advice of CSM in UK is that St John’s wort should not be used with warfarin. They note that degree of induction of warfarin metabolism is likely to vary because levels of active ingredients may vary between St John’s wort preparations. If a patient is already taking combination, they advise checking INR, stopping St John’s wort and then monitoring INR closely and adjusting anticoagulant dosage as necessary (See reference number 5).

1. Maurer A,Johne A, Bauer S, Brockmöller J, Donath F, Roots I, Langheinrich M, Hübner W

D. Interaction of St John’s wort extract with phenprocoumon. Eur J Clin Pharmacol (1999) 55,A22.

Bon S,Hartmann, Kuhn M. Johanniskraut: Ein Enzyminduktor? Schweiz Apothekerzeitung (1999) 16, 535–6.

Jiang X,Williams KM, Liauw WS, Ammit AJ, Roufogalis BD, Duke CC, Day RO, McLachlanAJ. Effect of St John’s wort and ginseng on the pharmacokinetics and pharmacodynamics ofwarfarin in healthy subjects. Br J Clin Pharmacol (2004) 57, 592–9.

Yue Q-Y,Bergquist C, Gerdén B. Safety of St John’s wort (Hypericum perforatum). Lancet (2000) 355, 576–7.

Committee on Safety of Medicines. Message from Professor A Breckenridge,Chairman, Committee on Safety of Medicines, and Fact Sheet for Health Care Professionals, February 2000.Available at http://www.mhra.gov.uk/home/idcplg?IdcService=SS_GET_PAGE&useSecondary=true&ssDocName=CON2015756&ssTargetNodeId=221 (accessed 17/08/07).

One pharmacological study found that American ginseng (Panax quinquefolius) modestly decreased effect of warfarin, whereasanother study found that Panax ginseng did not alter effect of warfarin. Two case reports describe decreased warfarin effects,one with thrombosis, attributed to use of ginseng (probablyPanax ginseng).

In a placebo-controlled study,20 healthy subjects were given warfarin 5 mg daily for 3 days alone then again on days 15 to 17 of a 3-week course of American ginseng 1 g twice daily. In 12 subjects given ginseng, peak INR was modesty reduced by 0.16,compared with a non-significant reduction of 0.02 in 8 subjects given placebo. There was also a modest reduction in AUC of warfarin. In this study,American ginseng (Panax quinquefolius) root was ground and capsulated (See reference number 1).

Evidence from two earlier case reports supports a reduction in warfarin effect. A man taking warfarin long-term,and also diltiazem, glyceryl trinitrate and salsalate, had a fall in his INR from 3.1 to 1.5 within 2 weeks of starting to take ginseng capsules (Ginsana) three times daily. This preparation contains 100mg of standardised concentrated ginseng [probably Panax ginseng] in each capsule. Within 2 weeks of stopping ginseng his INR had risen again to 3.3 (See reference number 2). Another patient taking warfarin was found to have thrombosis of a prosthetic aortic valve,with a subtherapeutic INR of 1.4. Three months prior to this episode his INR had become persistently subtherapeutic,requiring a progressive increment in his warfarin dose. It was suggested that this might have been because he had begun using a ginseng product (not identified) (See reference number 3). In contrast, in a randomised, crossover study in 12 healthy subjects, ginseng capsules 1 g three times daily for 2 weeks did not affect either pharmacokinetics or pharmacodynamics (INR) of a single 25mg dose of warfarin taken on day 7. The brand of ginseng used was Golden Glow,each capsule containing an extract equivalent to 0.5 g of Panax ginseng root (See reference number 4).

Ginseng can come from root of Panax ginseng (known in US as Asian ginseng) or Panax quinquefolius (American ginseng), which differ in concentrations and specific ginsenosides (See reference number 5). The study showing a reduction in effect of warfarin used American ginseng, and study showing no effect used Panax ginseng. A study in rats also failed to find any evidence of an interaction between warfarin and an extract from Panax ginseng (See reference number 6). Nevertheless, two case reports of reduced warfarin effect are probably Panax ginseng.

In contrast there have been handful of reports of spontaneous bleeding in patients using ginseng preparations (unspecified) in absence of an anticoagulant,(See reference number 7,8) and, in vitro, Panax ginseng has been found to contain antiplatelet components (See reference number 9). See also Coumarins + Herbal medicines interaction.

The available evidence suggests that ginseng might decrease effect of warfarin. It is possible that effect is greater with, or specific to, American ginseng (Panax quinquefolius), since this interacted in one study whereas Panax ginseng did not. Although ginseng dose was higher in Panax ginseng study, treatment duration was not as long, which may have obscured an effect. Moreover, two case reports of decreased warfarin effects attributed to use of ginseng were probably Panax ginseng.

Until further information becomes available it would seem prudent to be alert for decreased effects of coumarins in patients using ginseng,particularly American ginseng. However, possibility of an increased risk of bleeding due to antiplatelet component of Panax ginseng cannot entirely be ruled out, but see also Coumarins + Herbal medicines interaction.

Yuan C-S,Wei G, Dey L, Karrison T, Nahlik L, Maleckar S, Kasza K, Ang-Lee M, Moss J.Brief communication: American ginseng reduces warfarin’s effect in healthy patients. Ann Intern Med (2004) 141, 23–27.

Janetzky K,Morreale AP. Probable interaction between warfarin and ginseng. Am J Health-Syst Pharm (1997) 54, 692–3.

Rosado MF. Thrombosis of a prosthetic aortic valve disclosing a hazardous interaction between warfarin and a commercial ginseng product. Cardiology (2003) 99,111.

Jiang X,Williams KM, Liauw WS, Ammit AJ, Roufogalis BD, Duke CC, Day RO, McLachlanAJ. Effect of St John’s wort and ginseng on the pharmacokinetics and pharmacodynamics ofwarfarin in healthy subjects. Br J Clin Pharmacol (2004) 57, 592–99.

Plotnikoff GA,McKenna D, Watanabe K, Blumenthal M. Ann Intern Med (2004) 141, 893–4.

Zhu M,Chan KW, Ng LS, Chang Q, Chang S, Li RC. Possible influences of ginseng on thepharmacokinetics and pharmacodynamics of warfarin in rats. J Pharm Pharmacol (1999) 51, 175–80.

Hopkins MP,Androff L, Benninghoff AS. Ginseng face cream and unexplained vaginal bleeding. Am J Obstet Gynecol (1988) 159, 1121–2.

Greenspan EM. Ginseng and vaginal bleeding. JAMA (1983) 249,2018.

Kuo S-C,Teng C-M, Leed J-C, Ko F-N, Chen S-C, Wu T-S. Antiplatelet components in Panaxginseng. Planta Med (1990) 56, 164–7.