Middle-aged Woman With an Excessively Prolonged INR

A 44-year-old woman is hospitalized for evaluation and management of an excessively elevated international normalized ratio (INR). She takes warfarin for a thromboembolic diathesis of 10 years’ duration. Her first episode of deep vein thrombosis (DVT) was in 1999 and was idiopathic. Evaluation at that time revealed factor V Leiden heterozygosity. She has had multiple episodes of DVT and a pulmonary embolism in the intervening years. A vena cava filter was placed 5 years ago.

Her usual warfarin dosage is 5.0 alternating with 7.5 mg daily, and there has been no change in recent months. She denies any pulmonary symptoms but is troubled by periodic pain and swelling in her legs, which she states is due to her “blood clots acting up again.” She reports no hemorrhagic symptoms.

HISTORY

She has a history of asthma and hypertension. Medications include albuterol, fluticasone/salmeterol, metoprolol, and occasionally alprazolam. She denies any recent changes in medications or dosage.

PHYSICAL EXAMINATION

This obese woman is in no apparent distress. Blood pressure is 158/94 mm Hg; heart rate, 88 beats per minute; and respiration rate, 16 breaths per minute. Temperature is normal. Results of the head, eye, ear, nose, and throat examination are normal. There are no enlarged lymph nodes. Chest is clear without wheezes. The heart is normal. There is bilateral (12) pedal edema with hemosiderin deposition in both ankles.

LABORATORY RESULTS

Results of the admission hemogram and chemistry panel are normal. Albumin level is 3.2 g/dL; alanine aminotransferase, 11 U/L; and aspartate aminotransferase, 14 U/L. Partial thromboplastin time (PTT) is 90 seconds; prothrombin time (PT), 65 seconds; and INR, 9.6. Vitamin K 5.0 mg is given orally, which results in a PT of 55 seconds and an INR of 6.0 the next morning. However, before planned discharge, a repeated PTT is 115 seconds and INR is 10.6. A mixing study is done and shows PTT corrected to 37 seconds and INR to 1.6. Over the following days, the results of coagulation tests are shown in the Table.

Which of the following is the most likely explanation for the findings manifested by this patient?

A. Enhanced warfarin sensitivity related to pharmacokinetic interactions.

B. Presence of a concordant hereditary coagulopathy such as von Willebrand disease.

C. Exogenous warfarin ingestion.

D. Enhanced warfarin sensitivity related to pharmacogenomics.

(Answer on next page)

Correct Answer: C

Although a variety of oral direct thrombin inhibitors are entering clinical use, warfarin remains the staple for chronic oral anticoagulation in many clinical areas. Experience with this agent now exceeds 5 decades, and we are more aware than ever of its complex pharmacology and clinical therapeutics.

Excellent reviews are available.¹ Although the efficacy of warfarin is proven, its complex pharmacology continues to make it somewhat difficult to use. We now know there are several major components to this complexity, including compliance, the use of biologic end points (eg, INR) in clinical therapeutics, very complex pharmacokinetics resulting in drug interactions (even more amplified now in the current era of multiple pharmaceutical use in so many patients) and, lately, the observation of the role of pharmacogenomics in warfarin metabolism. This patient presented with the all too common problem of erratic INR.

PHARMACOKINETICS OF WARFARIN

The most frequent cause of erratic INR remains the pharmacokinetic interaction of warfarin with other drugs (choice A). Mechanisms include alterations in plasma protein bindings and, more typically, up- or down-regulation of warfarin metabolism induced by other drugs the patient is using. These usually involve the P451 cytochromes in the liver.

Typical and classic examples of up-regulation of hepatic cytochromes causing enhanced warfarin metabolism and thus diminished warfarin effect include barbiturates and rifampin. Conversely, examples of drugs that compete for warfarin metabolic sites and thus potentiate warfarin effect include the macrolide antibiotics and amiodarone.¹ Such pharmacokinetic effects have been estimated to cause up to 90% of INR difficulties in clinical practice.

However, this patient, at least by history and review of clinic charts, has had no new drug introduced to her regimen, nor has she had dosage changes in
her long-term medications (none of which, incidentally, interact significantly with warfarin in any event). This makes choice A unlikely.

ROLE OF PHARMACOGENOMICS

Pharmacogenomic factors (choice D) have been a topic of recent investigation and review.^2-4 Genetic variations in two genes responsible for warfarin cytochrome metabolism—cytochrome P450 2C9 (CYP2C9) and vitamin K epoxide reductase (VKORC1)—have been shown to possibly contribute significantly to variability in patient dose requirements. Recent literature is mixed: some reviews emphasize testing for and using pharmacogenetic data to screen for sensitive and resistant patients,^3,4
while others believe much of the information provided is less expensively captured by the early INR responses.² Pharmacogenetic testing in warfarin therapeutics received FDA approval in 2007, but the authors agree with the investigators in the latter camp that this information adds significant expense with little additional real benefit in patients who are receiving warfarin.

In any event, this patient has had nearly a decade of stable, or at least reasonable, warfarin therapy with a dosage that does not indicate either genetic ultrasensitivity or resistance. It would not be reasonable to suddenly implicate this factor now.

WHAT COAGULATION TESTING REVEALED

Similarly, any genetic coagulopathy (choice B) would have been manifest at the onset of therapy long ago. Specifically, a factor VIII level was obtained during coagulation testing that excludes the von Willebrand disease presented in choice B.

Analysis of the history and laboratory findings thus excludes pharmacokinetic interactions, pharmacogenetic factors, and previously undiagnosed concomitant coagulation disorder. Analysis of coagulation testing shows she does respond to fresh frozen plasma, and her coagulation parameters correct when her plasma is mixed 1:1 with normal plasma, thus confirming factor deficiency rather than inhibitor (eg, anticardiolipin antibodies). Factor analysis confirmed a warfarin vitamin K lesion with profound lowering of vitamin
K–dependent factors VII, IX, and X. And, when she was given the antidote—oral vitamin K—her coagulation parameters responded.

This makes exogenous ingestion (and likely self-administration) of warfarin the best explanation here (choice C). Indeed, warfarin is one of the more common agents associated with that phenomenon, and the patient’s response to confrontation (see below) was quite typical.

OUTCOME OF THIS CASE

There continued to be a hectic INR/coagulation test curve with response to vitamin K (orally, observed by a nurse or a physician) followed by drifting upward INR values within a day or two. The patient’s room was searched during time off the floor for studies without finding medications. She eventually was confronted with the possibility of exogenous warfarin ingestion and left the hospital against medical advice (eloped) that evening.