Postgraduate Medical Journal (1989) 65, 611 - 612 Postgrad Med J: first published as 10.1136/pgmj.65.767.611 on 1 September 1989. Downloaded from

Leading Article Control ofwarfarin therapy D.K. Scott

Pharmaceutical Services, West Birmingham Health Authority, Dudley Road Hospital, Dudley Road, Birmingham B18 7QH, UK.

Warfarin has been established as the oral anti- regime would call for 9-10 mg daily for 3 days and coagulant of choice for many years. It is preferred to then adjust the dose according to INR.6 Patients with other coumarin-type drugs because of its greater thyrotoxicosis, liver impairment, heart failure, pre- reliability in clinical practice, and it benefits many existing high INR, or low body-weight should all be thousands of patients each year. Nonetheless, it is not given lower doses (say, 9,6,6 mg) as should those over without hazard and there must be very few doctors 80 years. who have not experienced difficulties in controlling a Careful control in an anticoagulant clinic is essential patient's coagulation state. to good management and this can be achieved by Warfarin is indicated for prophylaxis of thrombi in practised doctors, nurses or pharmacists.7 Patient patients with existing thrombi, such as deep venous compliance is a major determinant of success, as are thrombosis (DVT) or pulmonary embolism, or in the numerous potential interactions with food and those at high risk, such as atrial fibrillation, prolonged drugs. Some clinics try to improve compliance by bed-rest, post-surgery.' The degree of anticoagulation using only one strength of warfarin tablet, thus Protected by copyright. required depends upon the condition being treated.2 simplifying the dosage instructions and reducing the Prophylaxis for non-hip surgery requires an internat- risk of incorrect doses. Interactions may be divided ional normalized ratio (INR) of 2-2.5, hip surgery into those that affect absorption, those that affect and treatment of DVT, pulmonary embolus or tran- metabolism, those that affect protein-binding and sient ischaemic attacks 2.0-3.0 and recurrent throm- those that affect coagulation directly or cause bosis, valvular or arterial disease 3.0-4.5. The stan- bleeding. Only the most important examples are given dardization of assay procedures is vital but has been here, there are several good texts which may be well covered elsewhere.3 Many minor surgical proce consulted for details of others.8-" dures can be adequately covered by the use of Absorption takes 2-4 hours, with a peak serum subcutaneous heparin and warfarin is in any case level at about 2 hours, and is generally faster after contraindicated in the first and third trimesters of food. It is probably unaffected by antacids, despite pregnancy because of the high rate of fetal malform- some in vitro evidence to the contrary, but may be ation and bleeding. It is also contraindicated in decreased by the use oflaxatives. Cholestyramine resin

patients with bleeding diatheses, peptic ulceration, or liquid paraffin can have variable effects on clotting http://pmj.bmj.com/ severe hypertension or endocarditis, although the because of effects on both warfarin and vitamin K. presence of an artificial valve may over-ride the last Special diets containing large amounts of vitamin K, contra-indication. In rare cases of allergy or intrinsic or even a high intake of green vegetables, such as resistance to warfarin, phenindione can be used, as broccoli, can alter coagulation. Patients should be may intravenous or subcutaneous heparin. counselled to take their warfarin at the same time each The choice ofan initial dose, and its conversion to a day (the evening is usually convenient because it is maintenance dose, is hampered by the delay in res- easier to make dosage adjustments in the clinic with ponse (1-2 days), the wide variation between indiv- immediate effect) and either with or without food, but on September 30, 2021 by guest. iduals and the lack of a close relationship between not sometimes with and sometimes without. To avoid serum warfarin concentration and response. Several problems associated with tablets sticking in the authors have published predictive schedules for dosing oesophagus, patients should swallow their tablets with based on INR measurements in the first 2 days of a drink of water whilst standing or sitting upright.'2 treatment,'6 but these have proved to be of only Problems that may occur include oesophageal ulcera- limited benefit and the most common method is still to tion, a well-known complication for some other drugs 'try it and see'. In an otherwise healthy adult, a typical but recently reported for the first time for warfarin by Loft et al.'3 An unexpected difficulty caused by a Correspondence: D.K. Scott, Ph.D., M.R.Pharm.S. pharyngeal pouch is reported by Ong and Slater in this Received: 14 March 1989 issue'4 and serves to illustrate the great vigilance C The Fellowship of Postgraduate Medicine, 1989 612 LEADING ARTICLE Postgrad Med J: first published as 10.1136/pgmj.65.767.611 on 1 September 1989. Downloaded from and detective work needed to control warfarin in total serum level. The opposite happens when the therapy. interfering drug is stopped, there is a period ofreduced Hepatic metabolism of warfarin by the mixed- anticoagulation followed by a return to the status quo. function oxidases is enhanced by smoking, alcohol The time taken for full induction/inhibition of abuse, rifampicin, carbamazepine and barbiturates. enzymes varies from 3-10 days but the time taken for Cimetidine, erythromycin, metronidazole and the the effect to wear off, once the interfering drug is quinolone antibiotics, such as ciprofloxacin and stopped, is less well-documented. In the author's norfloxacin, inhibit hepatic enzyme function and experience, it has occurred after 6-7 days. potentiate the anticoagulant effect of warfarin. Drugs which affect platelet function (dipyridamole, Isoniazid also potentiates warfarin but the mechanism aspirin, sulphinpyrazone) should be used with caution is not clear. Drugs which decrease hepatic blood flow, with warfarin, or avoided completely. Patients should such as propranolol, decrease warfarin metabolism, be counselled to avoid aspirin in home remedies for but to a lesser extent. Modest alcohol consumption colds, headache etc. They should also avoid drugs does not appear to affect anticoagulation.'5-'6 which may cause bleeding by damaging the gastric Warfarin is extensively protein-bound in serum mucosa. All non-steroidal anti-inflammatory agents, (about 99% to albumin) and is susceptible to displace- including ibuprofen which is available for purchase in ment by other more-strongly bound drugs, such as pharmacies in UK, should be avoided or used under phenylbutazone and sulphamethoxazole (in co- close medical supervision. Other drugs affect coagu- trimoxazole). The principal problem is not the lation by a variety of mechanisms, warfarin being existence of concomitant therapy, but the changes potentiated by amiodarone, quinine, quinidine, which occur when the other treatment is started or thyroxine and anabolic steroids. Spironolactone has stopped. Only un-bound drug is active and when been reported to antagonize warfarin, possibly by another protein-bound drug is introduced, the level of concentrating clotting factors. free warfarin rises as it is displaced from albumin. The There are many potential pitfalls in warfarin anticoagulant effect is temporarily enhanced but the therapy, but most can be avoided by close monitoring Protected by copyright. liver quickly metabolizes the free drug and the patient with a reliable prothrombin time assay and intensive returns to a steady-state with a concentration of free patient counselling. Warfarin remains a valuable drug warfarin similar to the former level, despite a decrease but should be treated with respect.

References 1. British National Formulary No. 16, London. British 9. Hansten, P.D. Drug Interactions. Lea and Febiger, Medical Association and Royal Pharmaceutical Society Philadelphia, 1985. of Great Britain, 1988. 10. Griffin, J.P., D'Arcy, P.F. & Speirs, C.J. A Manual of 2. Poller, L. Therapeutic ranges in anticoagulant adminis- Adverse Drug Interactions. 4th Edition. John Wright, tration. Br Med J 1985, 290: 1683- 1686. London, 1988. 3. Poller, L. Laboratory control of oral anticoagulants. Br 11. Standing Advisory Committee for Haematology of the Med J 1987, 294: 1184. Royal College of Pathologists. Drug interaction with

4. Carter, B.L., Taylor, J.W. & Becker, A. Evaluation of coumarin derivative anticoagulants. Br MedJ 1982, 285: http://pmj.bmj.com/ three dosage-prediction methods for initial in-hospital 274-275. stabilisation of warfarin therapy. Clin Pharm 1987, 6: 12. Channer, K.S. & Virjee, J. Effect of posture and drink 37-45. volume on the swallowing of capsules. Br Med J 1982, 5. Sawyer, W.T., Poe, T.E., Canaday, B.R. et al. Multi- 285: 1702. center evaluation of six methods for predicting warfarin 13. Loft, D.E., Stubington, S., Clark, C. & Rees, W.D.W. maintenance-dose requirements from initial response. Oesophageal ulcer caused by warfarin. Postgrad Med J Clin Pharm 1985, 4: 440-446. 1989, 65: 258-259. 6. Fennerty, A., Campbell, I.A. & Routledge, P.A. 14. Ong, A. & Slater, J.D.H. Intermittent absorption of Anticoagulants in venous thromboembolism. Br Med J warfarin caused by an unrecognized pharyngeal pouch. on September 30, 2021 by guest. 1988, 297: 1285-1288. Postgrad Med J 1989, 65: 660-661. 7. Bourne, J.G. & Pegg, M. Pharmacy contribution to 15. Udall, J.A. Drug interference with warfarin therapy. Clin outpatient management oforal anticoagulation. Pharm J Med 1970, 77: 20. 1987, 238: 733-735. 16. O'Reilly, R.A. Lack of effect of fortified wine ingested 8. Stockley, I. Drug Interactions. Blackwell Scientific Publi- during fasting and anticoagulant therapy. Arch Intern cations, Oxford, 1981. Med 1981, 141: 458.