Warfarin

From Ganfyd

Warfarin is a coumarin oral anti-coagulant, named after the Wisconsin Alumni Research Foundation where Karl Link, its discoverer, worked. Investigations in 1922 into the cause of haemorrhage in cattle from sweet clover silage ingestion by Frank Schofield, a Canadian veterinarian, showed a chemical was involved. In 1940, Karl Link helped isolate the coumarin analogue responsible. In the USA, it is commonly known as Coumadin. It acts through inhibiting vitamin K dependent carboxylation of clotting factors II, VII, IX and X all of which are produced in the liver.

The response to warfarin is monitored by measuring the prothrombin time, a blood test that measures the speed of clotting under standard conditions. As different laboratories use different reagents, adjustments can be made to arrive at an International normalised ratio (INR), a standardised index of how long blood takes to clot after prothrombin activation.

Contents 1 Indications 2 Starting Warfarin 2.1 Genetic testing 3 Keeping People in Range With Warfarin 3.1 Monitoring 3.2 Interactions 3.3 Determining the time spent in the desired range 3.4 Adjusting Dose 4 Risks 4.1 And Aspirin? 4.2 Triple therapy 4.3 Anticoagulation/Warfarin and surgical or medical procedures 5 Overdose 6 Alternative anticoagulant drugs 6.1 Non-coumarin oral anticoagulants 7 External Links 8 References

Indications

• Atrial fibrillation
• Deep vein thrombosis
• Pulmonary embolism
• Antiphospholipid syndrome
• Mechanical prosthetic valves

Starting Warfarin

Several regimes for starting Warfarin exist. It is not possible to predict the eventual daily requirement of Warfarin from inspection of the patient and their metabolism. The only predictive factors for eventual dose in one study were the sex and the INR after 2 weeks on 2mg/day.(see Talk). A loading dose is commonly used on the first few days,^[2] but this very commonly produces overshoots and swings. In recent years, a more leisurely approach has been introduced. See also Warfarin tables

Warfarin increases coagulability in early stages of warfarin treatment due to disproportionate inhibition of protein C and S (both are anti-coagulant factors). This can result in adverse sequelae such as warfarin necrosis. Therefore, where there is an active clotting process such as a DVT, the first few days of warfarin treatment should be overlapped with another anti-coagulant such as heparin. Heparin is also more appropriate when an immediate anti-coagulant effect is required.

Genetic testing

• Never going to be economically viable in low risk of haemorrhage patients as clinically would prolong life of an otherwise fit 69 year old man in atrial fibrillation by no more than one day^[3]

Keeping People in Range With Warfarin

Monitoring

UKMI recommend [1] baseline measurement of prothrombin time (PT) and activated partial thromboplastin time (APTT). They also recommend platelet count and LFTs, if possible, but this should not delay treatment. Once the INR level is within therapeutic range, INR should be monitored weekly until stable, and then at longer intervals (up to every 12 weeks). However, if there are any changes in the patients condition (e.g. liver disease, illness) more frequent monitoring may be required.

Interactions

These are essentially predictable from the known pharmacology of the drug but to use this knowledge to predict you need to know a fair bit of pharmacology of not only warfarin but also of other drugs and even foods. The short term risk is with potentiation so this is the most important area with new or intermittent therapy. Many find it simpler to learn the common interactions and of course this should be done for a selected few drugs and food substances by every prescribing doctor. A white list approach is also useful, so the knowledge that a short course of amoxicillin does not significantly impact on the INR is useful in routine prescribing and deciding that monitoring not indicated.

Determining the time spent in the desired range

This is a non-trivial computation. Automated systems should include an algorithm to estimate this for comparison and audit purposes.

Adjusting Dose

Adjustments should not be made on the basis only of a current result and current dose. Adjustments should usually be rather small.

Measurement should occur when there has been time for the adjustment to accomplish a change in state - after an adjustment of half a mg per week, a measurement in 2 weeks time will have seen only one additional mg of drug given or withheld. Measurements should be frequent while the state is changing.

Deming's work on process control is highly relevant to considering the frequency of testing, and the size of response to the results.

Risks

The risk of bleeding increases exponentially with the INR. And it seems adult trauma patients using warfarin at the time of their injury are at increased risk of poor outcomes compared with those not taking the anticoagulant. The odds ratio of increased mortality is 1.72(95% CI, 1.63-1.81). This association is not really explained by pre-existing pathology^[4]

And Aspirin?

Warfarin and Aspirin do not do the same job.

Random aspirin for patients on warfarin is warned against in the patient-carried record booklet and is clearly a bad idea. However there remains space for discussion over combining aspirin and warfarin in certain high potential benefit patients. It is not and should not be a common occurrence, and every instance should have a clear rationale. See therapeutic bleeding risk for fuller context of the issues.

Aspirin combined with warfarin at therapeutic INR has a moderate harm ratio of one serious upper GI bleed per 184 patient years of treatment.^[5]. It is therefore likely to be beneficial relative to the particular harm of bleeding in selected indications such as after a myocardial infarction ^[6] and acute coronary syndromes.^[7] In practice, given other harms and as the evidence base after myocardial infarction is distinctly mixed, ^[8][9][10] unless your patient actually has characteristics of the recruitment population of the successful trials, a prescriber is likely to be doing net harm. Thus, few would consider using this proven therapeutic combination in the majority of elderly patients and, indeed, dangerous extrapolation from the clinical trials to subgroups excluded from the trials such as the very old, patients in renal failure and non-compliant is to be deprecated.

In patients with atrial fibrillation and ischaemic stroke the SPORTIF trials show that aspirin and warfarin is likely to be ineffective or harmful with no reduction in stroke, systemic embolism, or myocardial infarction compared to warfarin alone and an incremental rate of major bleeding of 1.6% per year over warfarin alone^[11] In peripheral vascular disease we also know the combination is likely to be ineffective or harmful^[12] being no more effective than antiplatelet therapy alone in preventing major cardiovascular complications and associated with an increase in life-threatening bleeding.

Triple therapy

Unhappily the safety of concomitant aspirin, clopidogrel, and warfarin therapy after percutaneous coronary intervention is a function of the warfarin therapy with a hazard ratio of 5 and number needed to harm of 7^[13]. Short term triple therapy in those with a strong indication to be on warfarin is accepted, with atrial fibrillation being one of these strong indications, as overall all cause mortality is reduced^[14]. ESC consensus guidelines on what to do in patients on prior antithrombotic therapy for atrial fibrillation, who present with acute coronary syndrome and/or undergo percutaneous coronary intervention/coronary stenting exist. ^[15] Essentially the antiplatelet agents are given as well as the anticoagulant for the period of highest thrombotic risk with some modification if the risk of haemorrhage is high. In acute coronary syndrome, standard therapy is used with oral anticoagulation generally being withheld while INR > 2.0. Triple therapy is now recommended for 3 to 6 months, possibly longer in those at highest coronary artery disease risk. However the INR control must be tight between 2 to 2.5. It is noted that those at highest risk of embolic event (eg metal heart valves) INR up to 3 does not contraindicate radial artery access PCI. For elective metal stents offer 1 month of triple therapy, elective 'olimus coated stents get 3 months triple therapy and to 12 months an antiplatelet and anticoagulant, while elective paclitaxel coated stents get 6 months triple therapy, then up to 12 months an antiplatelet and anticoagulant. This last regime also applies to all stents used in primary PCI. If risk of haemorrhage is high, use bare metal stents in all indications and up to 4 weeks of triple therapy, then in elective patients revert to warfarin but in primary PCI acute coronary syndrome patients give 6 months antiplatelet and anticoagulant before reverting to anticoagulation alone.

Anticoagulation/Warfarin and surgical or medical procedures

There is a balance between risk of the condition the anti-coagulation is to prevent, and of bleeding, which should be individualised for various conditions, procedures and patients. Conferring is often indicated and the problem should be presented to the patient as being an intelligent balancing of risks. See main article anticoagulation and procedures

Overdose

See Warfarin toxicity

Can be intentional (e.g. suicide attempt) or accidental (ingestion by child, or mistaking different coloured tablets). Often iatrogenic due to misjudged dosage, inadequate monitoring, or failure to appreciate interactions with other drugs.

Can be rapidly reversed with FFP or specific recombinant factors, e.g. Beriplex™, but this is a relatively short-lived effect as the half-life of many of these factors is in the order of several hours. More lasting reversal requires vitamin K, but this can take 8-12 hours to become effective.

If the level is moderately high, and the indication for anticoagulation remains good, using a small dose of vitamin K may be preferable to a large one.

Alternative anticoagulant drugs

In the UK phenindione is the usual second choice coumarin. It is suitable for use in the few patients with definite allergy to Warfarin. A third coumarin is nicoumalone.

In Holland phenprocoumon and acenocoumarol are used rather than warfarin. They differ subtly in pharmacokinetics which may affect the timing of monitoring. The usual drug in Germany is phenprocoumon.

Non-coumarin oral anticoagulants

Rivaroxaban, apixaban and dabigatran etexilate are under trial. The last was recently introduced for postoperative prophylaxis (studies in Lancet 2007) and is reported to have few interactions and not require routine monitoring. A previous drug of this type was withdrawn due to toxicity.

External Links

• SIGN guideline 36 (full guidance)
• SIGN guideline 36 (specific link)
• Coagulation algorithms at the MEDAL Project

References