TABLE 16-17

Key Elements of Patient Education Regarding Warfarin

Identification of generic and brand names

Purpose of therapy

Expected duration of therapy

Dosing and administration

Visual recognition of drug and tablet strength

What to do if a dose is missed

Importance of prothrombin time/INR monitoring

Recognition of signs and symptoms of bleeding

Recognition of signs and symptoms of thromboembolism

What to do if bleeding or thromboembolism occurs

Recognition of signs and symptoms of disease states that influence

warfarin dosing requirements

Potential for interactions with prescription and over-the-counter

medications and natural/herbal products

Dietary considerations and use of alcohol

Avoidance of pregnancy

Significance of informing other health care providers that warfarin has

been prescribed

When, where, and with whom follow-up will be provided

INR, international normalized ratio.

are available from the manufacturers of warfarin, and from other

non-commercial sources.

B.H. should receive extensive education about warfarin therapy in an individual teaching session or an organized education

program. A wallet card, medical bracelet, or alternative method

of identifying her as a patient treated with warfarin should be

provided. The health care provider who assumes responsibility

for her outpatient warfarin therapy will need to provide continuing reinforcement of the essential elements of medication

information at each follow-up visit.

FACTORS THAT INFLUENCE WARFARIN DOSING

CASE 16-6, QUESTION 3: After receiving 6 days of

dalteparin therapy and six doses of warfarin 4 mg/day PO,

B.H.’s INR is 2.4. Dalteparin is discontinued, and B.H. is

instructed to continue her current dosage of warfarin. She

is scheduled to return to the anticoagulation clinic in 1 week

for re-evaluation. At that time, her INR is 1.7. What factors

might account for this change in the intensity of anticoagulation?

Patients should always be questioned about their understanding of the prescribed dose and their adherence to the prescribed

regimen. Questions might include, “What dose of the medication have you been taking?”, “What time of the day do you take

your medication?”, and “How many times in the last week did you

miss a dose of your medication?” If there is no evidence of misunderstanding of the correct dose or of noncompliance, numerous

other factors should be considered that are known to influence

warfarin dosing requirements in individual patients during both

initiation and maintenance phases of therapy. Changes in dietary

vitamin K intake, underlying disease states and clinical condition,

alcohol ingestions, genetic factors, and concurrent medications

can significantly change the intensity of therapy, resulting in the

need for dosing adjustments to maintain the INR within the

therapeutic range.

Dietary Vitamin K Intake

The two primary sources of vitamin K in humans are the biosynthesis of vitamin K2 (menaquinone) by intestinal bacteria and

dietary intake of vitamin K1 (phytonadione). The US recommended daily allowance for vitamin K is 70 to 140 mcg/day,

and the typical Western diet provides approximately 300 to 500

mcg/day.81 Vitamin K is found in high concentrations in certain foods, including green leafy vegetables (asparagus, broccoli,

Brussels sprouts, cabbage, cauliflower, chickpeas, collard greens,

endive, kale, lettuce, parsley, spinach, and turnip greens), soy

milk, certain oils, certain nutritional supplements, and multiple

vitamin products. Green tea and chewing tobacco are other significant sources of vitamin K.

Variations in vitamin K intake have been linked to INR fluctuations in patients taking warfarin.82,83 In addition, diets high in

vitamin K content have been associated with acquired warfarin

resistance, defined as excessive warfarin dosing requirements to

reach a therapeutic INR range.84 Numerous cases have also been

reported in which patients previously stabilized with warfarin

experienced elevations in INR with or without hemorrhagic

complications when dietary sources of vitamin K were eliminated. Conversely, reductions in INR with or without thromboembolic complications have been reported in patients in whom

dietary sources of vitamin K have been added.

These data illustrate the potential clinical significance of

dietary changes in patients taking warfarin. To minimize these

potential effects, B.H. should be counseled to maintain a consistent intake of dietary vitamin K.85 Her final warfarin maintenance

dose will be partially influenced by her typical diet. However,

restriction of dietary vitamin K intake is unnecessary, except in

cases of significant resistance to the anticoagulant effect of warfarin. B.H. should be aware of the types of foods and supplements that contain large quantities of vitamin K, and should be

counseled to maintain a consistent diet, to avoid bingeing with

foods high in vitamin K content, and to report significant dietary

changes to her health care provider. Appropriate assessment and

follow-up are essential to prevent hemorrhagic or thromboembolic complications that may arise from changes in INR resulting

from dietary alterations.

Underlying Disease States and Clinical Conditions

The presence or exacerbation of various medical conditions can

also influence anticoagulation status61 (Table 16-18). Diarrheaassociated alterations in intestinal flora can reduce vitamin K

absorption, resulting in elevations in INR. Fever enhances the

catabolism of clotting factors and can increase INR. Heart failure,

hepatic congestion, and liver disease can also cause significant

elevations in INR because of a reduction in warfarin metabolism.

End-stage renal disease is associated with decreased CYP2C9

activity, resulting in lower warfarin dose requirements.

Thyroid function can influence warfarin therapy significantly.

Hypothyroidism decreases the catabolism of certain clotting factors, increasing their availability and producing a relative refractoriness to warfarin therapy. This results in the need for increased

dosages to reach a therapeutic INR. The addition of thyroid supplementation in these patients reverses the influence of hypothyroidism and can lead to significant elevations in INR unless the

warfarin dose is reduced. Conversely, hyperthyroidism increases

the catabolism of clotting factors, leading to an increased sensitivity to warfarin. Frequent monitoring of and adjustments in

warfarin therapy are necessary in patients with changing thyroid

function.

Acute physical or psychological stress has been reported to

increase INR. Increased physical activity has also been reported

to increase the warfarin dosing requirement. Smoking can induce

CYP1A2, which may increase warfarin metabolism in certain

patients, resulting in increased dose requirements. Due to its high

vitamin K content, chewing smokeless tobacco can suppress the

INR response.

365Thrombosis Chapter 16

TABLE 16-18

Warfarin Interactions With Disease States and Clinical Conditions

Clinical Condition Effect on Warfarin Therapy

Advanced age Increased sensitivity to warfarin due to reduced vitamin K stores and/or lower plasma concentrations of vitamin

K–dependent clotting factors

Pregnancy Teratogenic; avoid exposure during pregnancy

Lactation Not excreted in breast milk; can be used postpartum by nursing mothers

Alcoholism  Acute ingestion: inhibits warfarin metabolism, with acute elevation in INR

 Chronic ingestion: induces warfarin metabolism, with higher dose requirements

Liver disease  May induce coagulopathy by decreased production of clotting factors, with baseline elevation in INR

 May reduce clearance of warfarin

Renal disease Reduced activity of CYP2C9, with lower warfarin dose requirements

Heart failure Reduced warfarin metabolism due to hepatic congestion

Cardiac valve replacement Enhanced sensitivity to warfarin postoperatively due to hypoalbuminemia, lower oral intake, decreased physical

activity, and reduced clotting factor concentrations after cardiopulmonary bypass

Nutritional status Changes in dietary vitamin K intake (intentional or as the result of disease, surgery, etc.) alter response to warfarin

Use of tube feedings Decreased sensitivity to wafarin, possibly caused by changes in absorption or vitamin K content of nutritional

supplements

Thyroid disease  Hypothyroidism: decreased catabolism of clotting factors requiring increased dosing requirements

 Hyperthyroidism: increased catabolism of clotting factors causing increased sensitivity to warfarin

Smoking and tobacco use  Smoking: may induce CYP1A2, increasing warfarin dosing requirements.

 Chewing tobacco: may contain vitamin K, increasing warfarin dosing requirements

Fever Increased catabolism of clotting factors, causing acute increase in INR

Diarrhea Reduction in secretion of vitamin K by gut flora, causing acute increase in INR

Acute infection/inflammation Increased sensitivity to warfarin

Malignancy Increased sensitivity to warfarin by multiple factors

INR, international normalized ratio.

Source: Wittkowsky AK. Warfarin. In: Murphy J (ed). Clinical Pharmacokinetics (5th ed). Bethesda, MD: American Society of Health System Pharmacists; 2011:345.

Thorough education of patients taking warfarin should

include detailed attention to recognizing the signs and symptoms of changes in underlying disease states and clinical conditions that can influence warfarin dosing requirements. They

should be instructed to contact their anticoagulation management program whenever changes occur that might influence

INR and warfarin dose requirement.

Alcohol Ingestion

Chronic alcohol ingestion has been associated with induction of

the hepatic enzyme systems that metabolize warfarin. Therefore,

warfarin dosing requirements are sometimes higher in alcoholic

patients. Conversely, acute ingestion of large amounts of alcohol

can slow warfarin metabolism through competitive inhibition

of metabolizing enzymes, leading to elevations in INR and an

increased risk of bleeding complications.85 Despite some reports

linking low amounts of alcohol to an elevated INR,86 in general

it is believed that moderate intake of alcoholic beverages is not

associated with alterations in the metabolism or the therapeutic

effect of warfarin as measured by INR. Patients taking warfarin

should be educated to limit their alcohol consumption to less

than one to two alcoholic beverages per day. Chronic drinkers

should be counseled to limit their drinking and maintain a regular

pattern to avoid fluctuations in INR.85 B.H. does not need to

abstain from drinking alcoholic beverages in moderation, but

she should be counseled to avoid the sporadic ingestion of large

amounts of alcohol.

Conversely, alcoholic liver disease (i.e., cirrhosis) can alter

multiple hemostatic mechanisms and reduces production of hepatic clotting factors. Decreased production and clearance of vitamin K–dependent clotting factors accounts for prolonged PT and

INR often seen in these patients. Therefore, an increased response

to warfarin would be expected in patients with liver impairment.

Worsening liver function is also a predictor for bleeding complications and patients with end-stage liver disease are at increased risk

of bleeding. Before instituting warfarin therapy in these patients,

the risks of bleeding associated with both the underlying liver disease and warfarin therapy must be weighed against the benefit

of preventing thromboembolic events. If warfarin is indicated,

the best approach would be to use a cautious initiation and dose

titration approach by starting with lower doses and titrate up

slowly to goal. Small increases in dosage should be made, if indicated, recognizing that the full effect of any dose adjustment may

be delayed in patients with severe liver dysfunction. Monitoring

for bleeding complications is essential, even at goal INR ranges,

when warfarin is used in patients with liver dysfunction.

Genetic Factors

CYP2C9 genotype12,87 and VKORC1 haplotype12,88 have been

shown to correlate with the dose of warfarin required for effective

anticoagulation. Dosing algorithms that incorporate CYP2C9

genotype and VKORC1 haplotype along with other patient characteristics to predict warfarin maintenance doses are being tested.

A comparison of various dose prediction methods (empiric initiation, regression equation accounting for various clinical factors, suggested dosing from the warfarin package insert based

on genomic information, mean dosing requirements based on

genotype, and regression equation based on genomic and clinical

factors) found that the percentage of patients whose predicted

doses were within 20% of their stable therapeutic doses were

37%, 39%, 43%, 44%, and 52%, respectively.89 The long-term

utility of genetic-based warfarin dosing prediction methods is not

yet clear. Importantly, these methods do not replace the need for

routine coagulation monitoring with INR, and dose adjustments

based on INR results.

CASE 16-6, QUESTION 4: How should B.H. be assessed and

evaluated at this clinic appointment?

At each clinic visit, B.H. should be assessed for signs and symptoms of bleeding, and for signs and symptoms of clot progression

and/or recurrence. Regardless of the INR result, all factors that

itrate combination.158,294 The outcome is difficult to interpret

because of the absence of a placebo group. The all-cause annual

mortality rate for African American patients (n=215) was identical in the two drug groups (12.8% with enalapril and 12.9% with

hydralazine–isosorbide). In non–African American patients (n =

574), the corresponding mortality rates were 11% with enalapril

and 14.9% with hydralazine–isosorbide. These data could be

interpreted as either superior response to hydralazine–isosorbide

in African American subjects or inferior activity of ACE inhibitors

in African American patients. The latter interpretation is consistent with the hypothesis that ACE inhibitors might have a lesser

BP-lowering effect in African American patients with HTN compared with non–African American patients. A similar reanalysis of the SOLVD Prevention and Treatment trials,147 both of

which compared enalapril with placebo in patients with recent

MI, concluded that enalapril therapy is associated with a significant reduction in the risk for hospitalization for HF among white

patients (44% reduction) with LV dysfunction, but not among

similar African American patients. Confounding variables contributing to all of the analyses presented include disproportionately low numbers of African American subjects in the trials, and

possibly more underlying risk factors (e.g., HTN) in the African

American subjects.

To further address the effect of race on response to ACE

inhibitors, a meta-analysis of the seven major ACE inhibitor studies, representing a total of 14,752 patients, was conducted.295 The

conclusion was that the relative risk for mortality when taking

an ACE inhibitor compared with placebo was identical (0.89)

for both African American and white patients. The authors of

the meta-analysis urged that ACE inhibitors not be withheld

from African American patients. One other consideration with

ACE inhibitors is an observed higher rate, although still rare, of

angioedema in black patients than in white patients. Until more

data are available, ACE inhibitors should not be withheld from

African American patients, but careful monitoring is required to

assess response.

The African American Heart Failure Trial (AHeFT) was

designed to determine possible superiority of combination

hydralazine plus isosorbide dinitrate in African American

patients.72 AHeFT was a randomized comparison trial (n =

1,050) of hydralazine–isosorbide dinitrate and placebo in African

American patients with NYHA class III or IV HF who were

receiving standard HF therapy (94% diuretics, 87% β-blockers,

93% ACE inhibitors or ARB, 62% digoxin, and 39% aldosterone

antagonists). The primary end point was a composite of all-cause

death, first hospitalization for HF, and quality of life scores at

6 months with secondary end points being individual components of the primary end point. Reduction in the composite

primary end point events was statistically significant in favor

of the active drug combination. Importantly, all-cause mortality

declined 43% in the hydralazine–isosorbide dinitrate arm versus

that for the placebo group (p = 0.012). The study also reported a

39% reduction in first hospitalization for HF in the hydralazine–

isosorbide dinitrate group versus placebo (p <0.001). As a result,

the 2009 ACC/AHA guideline recommendations state that the

addition of hydralazine and isosorbide dinitrate to HF patients

receiving ACE inhibitors and β-blockers is effective in African

American patients with NYHA functional class III or IV HF.1,21

The results of AHeFT were also the primary factor leading the

FDA to approve the combination product of hydralazine and

isosorbide dinitrate (BiDil) for adjunctive treatment of HF in selfidentified African American patients already receiving standard

therapy. Advantages of using BiDil in clinical practice include

use of the same product studied in the clinical trial and potential

improved compliance by using a combination tablet. Cost, however, may be lower when using generic hydralazine and isosorbide

as separate drugs.

β-BLOCKERS

A possible racial difference in response to β-blocker drugs has

also been hypothesized based on differential effects observed in

patients with HTN.293,295,296 A post hoc analysis of the various

US Carvedilol Heart Failure trials214–218 concluded, however, that

the benefit of carvedilol was apparently of similar magnitude in

both black (n=217) and not black (n=877) patients.296 Using the

combined end point of the risk of death as a result of any cause or

hospitalization, the risk reduction of β-blockers compared with

placebo was 48% in black patients and 30% in not black patients.

Because fewer black patients were studied, these differences did

not reach statistical significance. Also seen were a significant

improvement in NYHA functional class, EF, and patient global

symptom assessment with carvedilol in both black and not black

patients.

Contradictory evidence comes from the BEST.227 In this trial

(also discussed in Case 19-1 Question 20), 2,708 patients with

NYHA class III (92%) or IV (8%) HF were randomly assigned

to either bucindolol or placebo. Bucindolol is a nonselective

β-blocker with partial agonist activity that imparts weak vasodilation. A unique characteristic of this study was that a subgroup analysis for racial differences was planned from the start.

Although there was a trend toward reductions in CV mortality

and hospitalization with the active drug, the trial was terminated after 2 years when it was determined that there was no

mortality benefit of active drug compared with placebo (33%

mortality in placebo group vs. 30% with bucindolol). A subgroup analysis showed a mortality benefit in not black subjects,

but none in black subjects. Subsequently, a meta-analysis of the

five major β-blocker in HF studies was conducted, representing

a total of 12,727 patients.295 When the BEST was included in the

meta-analysis, the relative risk for mortality when taking an ACE

inhibitor compared with placebo was 0.69 in white subjects, but

only 0.97 for black patients. When this trial was excluded, the

relative risks were reduced in both groups to 0.63 and 0.67 in

whites and blacks, respectively. The difference between the two

groups was not statistically significant, although the 95% confidence interval for black patients was broader and included 1.0

(0.38–1.16). On the basis of all of these factors, it is likely that

black patients will derive similar benefit from β-blockers as do

white patients when given carvedilol, metoprolol, or bisoprolol.

Bucindolol should be avoided depending on the pharmacogenomics of the patient, which happen to differ in black versus not

black patients.

Critical Care Management of

Heart Failure

CASE 19-4

QUESTION 1: L.M., a 62-year-old black man, was admitted