Pulmonary Embolism

James Allen, MD

I. Fat Embolism (Am J Orthop 2002; 31:507-12)

A. Usually seen in conjunction with long bone trauma or sickle cell anemia-associated acute chest syndrome

B. Diagnosis is clinical:

1. unexplained dyspnea, tachypnea, hypoxemia < 72 hours after traumatic fracture

2. unexplained diffuse alveolar infiltrates on chest x-ray; chest CT shows focal areas of consolidation, nodules, and/or ground glass infiltrates, the extent of which correlates with the severity of hypoxemia (J Comput Assist Tomogr 2000;24:24-9)

3. unexplained confusion

4. upper extremity petechiae, especially in the axilla & conjunctiva

5. although the finding of fat droplets in neutrophils and alveolar macrophages obtained from bronchoalveolar lavage was initially reported to be diagnostic for fat embolism syndrome, a recent study indicates that this finding is neither sensitive nor specific enough to be clinically useful (Chest 1992; 102:1323-7)

C. Treatment:

1. primarily supportive

II. Tumor Embolism (Am J Respir Crit Care Med 1997; 155:2089-95)

A. Typical presentation: dyspnea & hypoxemia with normal chest x-ray

B. Most common malignancies are:

1. breast 30%

2. lung 16%

3. prostate 13%

4. stomach 9%

5. liver 7%

C. Pulmonary angiograms typically normal

D. Diagnosis requires lung biopsy or intravascular cytology from a pulmonary artery catheter

1. the diagnosis is rarely made pre-mortum

E. Treatment usually unsatisfactory

III. Septic Embolism

A. Common causes:

1. right sided endocarditis

2. peri-pharyngeal infection (usually anaerobes)

3. infected intravenous catheters

4. infected dialysis shunts

5. osteomyelitis

B. Chest x-ray:

1. multiple ill-defined nodules

2. cavitation (frequent)

3. bronchopneumonia-like pattern

C. Diagnosis = blood cultures (sputum cultures non-diagnostic)

1. Staphylococcus aureus most common

D. Treatment:

1. antibiotics directed toward the specific pathogen

2. removal of infected shunts or other intra vascular appliances

3. pulmonary mycotic aneurysms can occur and should be suspected in patients with non-resolving radiographic densities or with hemoptysis. These can be verified by chest CT and require treatment with either surgery or catheter-based embolization (Ann Thorac Surg 2003; 75:593-6)

IV. Amniotic Fluid Embolism (Curr Opin Obstet Gynecol 2003; 15:119-22))

A. Cause:

1. sudden release of amniotic fluid or debris of fetal origin into maternal circulation

2. most commonly during labor (70%), vaginal delivery (11%), or C-section (19%) (Am J Obstet Gynecol, 1995; 172:1167-9)

3. reported conditions:

a) 1st trimester D & C

b) 2nd trimester abortion

c) spontaneous during 2nd trimester pregnancy

d) abdominal trauma

e) amniocentesis

B. Clinical presentation:

1. sudden dyspnea and hypoxemia

2. hypotension

3. acute pulmonary hypertension

4. cardiac arrest usually occurs within minutes of onset

5. seizures in 10-20%

6. DIC frequent

7. CXR = pulmonary edema in 70%

8. 80% of patients die, 50% within the first hour

a) 60% of fetuses die

C. Diagnosis:

1. usually made at autopsy by finding fetal hair or squamous cells in the pulmonary capillaries

2. occasionally made by finding hair or squamous cells in blood from a Swan-Ganz catheter (must be sure these are not maternal in origin from inserting the catheter)

D. Treatment:

1. high flow oxygen (usually intubate and give 100% oxygen with PEEP)

2. correct hypotension using fluids and dopamine

3. Swan-Ganz catheter to guide fluid administration

4. fresh frozen plasma to correct coagulation abnormalities

V. Chronic Thrombotic Large Vessel Pulmonary Artery Obstruction (Ann Int Med 1987; 107:560-5)

A. Pathophysiology:

1. chronic thrombotic obstruction of large pulmonary arteries

2. < 20% of patients will have a previous diagnosis of acute PE

3. anti-phospholipid antibodies are present in up to 50% in some series

B. Diagnosis:

1. Clinical findings non-specific:

a) dyspnea & fatigue early

b) pulmonary hypertension with cor pulmonale later

2. Chest x-ray often normal but may show asymmetry in pulmonary artery size

3. Pulmonary function tests not helpful

4. Resting or exertional hypoxemia invariably present

5. Ventilation/perfusion scan always abnormal with segmental or larger perfusion defects which are normally ventilated

6. Pulmonary angiogram necessary for diagnosis

C. Treatment:

1. Surgical pulmonary endarterectomy

2. Inferior vena cava filter to prevent recurrences

VI. Chronic Small Vessel Pulmonary Thrombosis (Chest 1988; 93:149S-54S)

A. Pathophysiology:

1. micro-thromboemboli of small pulmonary arterioles

2. often unclear if etiology is thrombosis in situ or showering of tiny emboli from a distant source

B. Diagnosis:

1. clinically present with unexplained pulmonary hypertension

2. chest x-ray usually normal but can show interstitial infiltrates

3. angiography usually not diagnostic

4. open lung biopsy required for diagnosis

C. Treatment:

1. anticoagulation occasionally reverses associated pulmonary hypertension

2. vasodilators are of unproven benefit

VII. Deep Venous Thrombosis (Am J Respir Crit Care Med 1999; 159:1-14)

A. Risk Factors:

1. venous stasis

a) prolonged lower extremity immobility

b) bed rest

(1) in one study of 100 consecutive patients in a MICU, 33% had DVT by duplex ultrasound (JAMA 1995; 274:335-7)

c) surgery

d) right sided congestive heart failure

e) pregnancy (NEJM 1996;335:108-114)

(1) occurs in 1/1,000 - 1/2,000 pregnancies

(2) DVT is considerably more likely in the left leg during pregnancy (Thromb Haemost 1992; 67:519-20)

(3) risk of DVT is 5X higher in pregnant women than controls

(4) in the third trimester of pregnancy, the pO2 is about 15 mm lower in the supine position rather than the upright position so ABGs should always be assessed upright if possible when evaluating the pregnant patient for pulmonary embolism

(5) patients known to have protein C or protein S deficiency have an 8-fold increase in pregnancy-associated thromboembolism, mainly in the post-partum period; prophylactic anticoagulation of these women may be warrented in the third trimester and in the post-partum period (Ann Intern Med 1996; 125:955-960)

2. endothelial injury

a) previous DVT

(1) risk of recurrent DVT is 17.5% in 2 years (Ann Intern Med 1996; 125:1-7)

(2) highest risk is in patients with malignancy and hypercoaguable states

b) femoral intravenous cathethers

(1) associated with DVT in 25% of patients in one study (Critical Care Medicine1995; 23:52-9)

c) trauma & lower extremity surgery (eg. hip surgery)

(1) calf vein thrombosis:

(a) 40-60% of patients with hip prothesis

(b) 60-70% of patients with knee prothesis

(2) proximal DVT

(a) 20% in either hip or knee prothesis

(b) 18% of patients with major trauma (N Engl J Med 1994; 331:1601-6)

d) strenuous muscle activity

(1) primarily a risk for upper extremity DVT

(a) in patients with isolated upper extremity DVT, inherited hypercoaguability is rarely present (Ann Intern Med 1997; 126:707-11)

3. hypercoagulability (Ann Intern Med 2003; 138:128-34)

a) primary (inherited) hypercoaguable states

(1) factor V Leiden mutation

(a) caused by a mutation in the factor V gene (factor V Leiden mutation) resulting in a resistance of factor V to activated protein C (protein C works by lysing factors Va and VIIIa)

i) this mutation appears to have originated in European populations and the incidence in racial groups varies considerably (JAMA 1997; 277:1305-7):

(1) Caucasian Americans - 5.27%

(2) Hispanic Americans - 2.21%

(3) African Americans - 1.23%

(4) Asian Americans - 0.45%

(5) Native Americans - 1.25%

ii) homozygous individuals are at extremely high risk of life-threatening thrombosis (approximately 80-fold increased risk)

(1) most homozygous individuals will experience at least 1 thrombotic event in their lifetime (Blood 1995; 85:1504-8)

(2) average age of thrombosis in homozygotes = 31

iii) heterozygous individuals are at about 5-10 fold increased risk of thrombosis

(1) average age of onset of thrombosis in heterozygotes = 44

(2) the incidence of thrombosis increases exponentially with age and factor V Leiden mutation should be considered in both young individuals who clot as well as elderly individuals who clot (Ann Intern Med 1997; 126:528-531)

iv) heterozygous individuals are at increased risk of thrombosis but not all affected individuals will clot

v) women who use oral contraceptives are at extremely high risk of thrombosis (about 30-fold increase) if they are heterozygous for the factor V Leiden mutation (Lancet 1994; 344:1453-7)

vi) there is a strong association with factor V leiden and pregnancy-associated DVT/PE - in a study of 407 obstetric patients, the incidence of pregnancy-associated thromboembolism in patients carrying the factor V Leiden gene was 28% whereas the incidence of thromboembolism in non-carriers was <1% (Am J Obstet Gynecol 1997;176:883-6, 1997); in another study, 43% of all pregnancy-associated thrombosis was attributable to factor V Leiden (N Engl J Med 2000; 342:374-80)

(b) factor V Leiden is the most common inherited cause of hypercoaguability (about 10 times more frequent than the other inherited hypercoaguable states

i) in 471 consecutive patients with DVT, 85 were found to be heterozygous and 7 homozygous (ie, 20% of all patients with DVT had the factor V Leiden mutation); by contrast, there were 14 heterozygous individuals amoung 474 control subjects (Blood 1995; 85:1504-8)

ii) this mutation appears to increase the risk of DVT/PE but not myocardial infarction or stroke (NEJM 1995; 332:912-7)

iii) functional resistance to activated protein C (which occurs with factor V Leiden mutation but can also occur independent of this genetic abnormality) is even more frequent and occurs in 33-40% of all patients with DVT (NEJM 1994; 330:517-22)

i) genetic testing using PCR (polymerase chain reaction) which tells you whether the individual is heterozygous (or homozygous) for the factor V Leiden mutation (performed in the molecular pathology laboratory)

ii) functional testing for resistance to activated protein C (performed in a coagulation laboratory)

iii) these two means of testing are complimentary and may both be required for complete analysis of hypercoaguability - about 10% of patients with functional resistance to protein C lack the factor V Leiden mutation and conversely, about 10% of patients with the factor V Leiden mutation have no demonstrable resistance to activated protein C

(d) the long term treatment for these individuals in not certain at present

i) some authorities anticoagulate the heterozygous patient indefinitely after 1 episode of thrombosis and other authorities give a standard course of coumadin for the first episode and then anticoagulate indefinitely only if there are further episodes of thrombosis

ii) homozygous individuals should probably be anticoagulated indefinitely, at least after they have had one thrombotic event

iii) the risk of recurrent thromboembolism is 40% in patients with factor V Leiden as compared to 18% in patients without factor V Leiden and long term anticoagulation may be considered in these patients (N Engl J Med 1997; 336:399-403). Other studies suggest that the risk of recurrent pulmonary embolism is the same in patients with and without factor V Leiden (Arch Intern Med 2002; 162:2357-60)

iv) the combination of factor V Leiden plus the prothrombin gene occurs in about 3% of all patients with deep venous thrombosis and portends an extremely high probability of recurrent thrombosis (65-88%) and these patients should probably all be anticoagulated indefinately after an initial thromboembolic event (N Engl J Med 1999; 341: 801-6)

(e) screening for factor V Leiden:

i) patients with otherwise unexplained thromboembolism should be screened because of the high likelihood of recurrent thromboembolism (see below)

ii) family members should be screened for the gene and provided appropriate counseling regarding risk avoidence and DVT prophylaxis (ie, during surgery, etc.)

(1) family members of patients with factor V Leiden who carry the gene have a 4.5-fold higher incidence of DVT/PE than family members without the factor V Leiden gene

(2) empiric anticoagulation of asymptomatic family members carring the gene is not warrented (Ann Intern Med 1998; 128:15-20)

iii) screening the general population does not appear cost effective at the present time (Ann Intern Med 1997; 127:895-903)

(2) hyperhomocysteinemia

(a) high plasma homocysteine levels are a risk for DVT (NEJM 1996; 334:759-62)

(b) genetic abnormalities of enzymes most commonly causing elevated homocysteine levels:

i) cystathionine B-synthetase

homozygotes have "homocystinuria"

ii) 5,10-methylenetetrahydrofolate reductase

i) methotrexate

ii) phenytoin

iii) theophylline

iv) carbamazepine

(d) medical conditions causing hyperhomocysteinemia:

i) renal insufficiency

ii) poor nutrition

iii) hypothyroidism

iv) carcinoma (especially ovarian, breast, and pancreatic)

v) diets high in animal fat

(e) co-existance of homocysteinuria and factor V Leiden mutation are at very high risk of thrombosis and should be screened for in patients with homocysteinuria (NEJM 1996; 334:763-8)

(f) diagnosis: fasting serum homocysteine level

i) 40% of patients with hyperhomocysteinemia will be missed and require serum homocysteine levels before and after oral methionine loading

(g) treatment:

i) folic acid 3-5 mg/day

ii) B6 50 mg/day

iii) B12 100 mg/day

(3) Prothrombin G-A20210 gene variant

(a) genetic variant which results in increased levels of prothrombin and in hypercoaguability

(b) incidence = 14% of all patients with DVT/PE

(d) homozygotes have extremely high incidence of thrombosis (all patients in some studies)

(e) the combination of heterozyous state for both factor V Leiden and prothrombin G-A20210 gene variant also gives extremely high probability of thrombosis and such patients may be candidates for life long anticoagulation (N Engl J Med 1999; 341:801-6)

(f) genetic testing available through the OSU Molecular Pathology Laboratory

(4) protein C deficiency

(a) heterozygote prevalence is about 1:200 - 1:300

(b) main, and in some studies, only risk is in patients with both heterozygous state and a family history of thrombosis; in these individuals, there is as much as a 50% incidence of thrombosis

(5) protein S deficiency

(a) heterozygous state incidence is about 1-2% of the population

(b) like protein C deficiency, the main risk is in individuals with a family history of thrombosis - in these patients, as many as 60% will develop thrombosis

(6) antithrombin III deficiency

(a) as with protein C and S, the main risk is in individuals with a family history of thrombosis - in these patients, as many as 50% will develop thrombosis

(b) patients with thrombosis and anti-thrombin III deficieny will have levels less than 70 or 80% of normal

(7) racial

(a) probably due to genetic differences in factor V Leiden, hyperhomocysteinemia, and prothrombin G-A20210 gene variant

(b) incidence of DVT/PE (Ann Intern Med 1998;128:737-40)

i) Caucasian = 230/million

ii) African American = 293/million

iii) Hispanic = 139

iv) Asian & Pacific Islanders = 60/million

(8) elevated levels of factor XI (NEJM 2000; 342:696-701)

(a) confers a 2-fold increased risk

(b) present in about 10% of the population of the Netherlands

(9) hypoplasminogenemia

(10) dysplasminogenemia

(11) tissue plasminogen activator release deficiency

(12) increased levels of plasminogen activator inhibitor

(13) dysfibrinogenemia

(14) heparin cofactor II deficiency

(15) fibrinolytic abnormalities

(16) elevated levels of factor VIII (NEJM 2000; 343:457-62)

(17) elevated levels of factor IX (Blood 2000; 95:3678-82)

b) secondary hypercoaguable states:

(1) anti-phospholipid antibody

(a) diseases causing anti-phospholipid antibodies:

i) systemic lupus erythematosus (50% of patients)

ii) rheumatoid arthritis

iii) Sjogren's syndrome

iv) systemic sclerosis

v) ankylosing spondylitis

vi) HIV

vii) syphilis

viii) Lyme disease

ix) drugs:

(1) phenothiazines

(2) procainamide

(3) hydralazine

(4) phenytoin

(5) oral contraceptives

(b) these patients have an unusually high incidence of recurrent thrombosis and some authors have recommended to undergo long term anticoagulation with an INR about 3 or above (NEJM 1995; 332:993-7)

i) chromogenic factor X levels and the prothrombin-proconvertin time are better tests of anticoagulation in these patients (Ann Intern Med 1997; 127:177-85). Many laboratories are now using reagents which are not susceptible to lupus anticoagulants

(2) certain malignancies

(a) malignancy will be diagnosed at the initial hospitalization of DVT/PE in 12% of patients (Ann Intern Med 1996; 125:785-93)

(b) most common primary malignancies associated with DVT/PE (N Engl J Med 1998; 338:1169-73):

i) pancreas

ii) ovary

iii) liver

iv) brain

i) history and physical examination

ii) chest x-ray

iii) stool quaiac

iv) mammography (in women > 40 without a mammogram in the past year)

v) CBC, chemistry profile, LFTs, urinalysis

(3) heparin-induced thrombocytopenia

(a) caused by an antibody directed against the heparin-platelet factor 4 complex which binds to Fc receptors on platelet surfaces (Blood 1994; 83:3232-9)

(b) occur in 3-4% of patients receiving unfractionated heparin for 7-10 days

(d) distinguish HIT from non-immune heparin-associated thrombocytopenia

i) HIT usually causes platelets to fall < 100,000 (as opposed to > 100,000 in non-immune heparin associated thrombocytopenia)

ii) HIT usually occurs 5-14 days after starting heparin in patients who have not previously received heparin (as opposed to 0-4 days for non-immune heparin-associated thrombocytopenia)

iii) HIT is associated with about a 50% incidence of subsequent thrombosis (usually venous) as opposed to non-immune heparin-associated thrombocytopenia which is benign (Am J Med 1996; 101:502-7)

(4) hypertension, obesity, and cigarette smoking were identified to be associated with pulmonary embolism in women in the Nurses' Health Study (JAMA 1997; 277:642-5) and smoking and obesity conferred a 3 and 4 fold risk for PE respectively in a 30 year follow up study of men from 50-80 years old (Arch Intern Med 1999;159:1886-1890)

(5) estrogen therapy

(a) in 2763 postmenopausal women randomly given conjugated estrogen or placebo, 34 women receiving estrogen and 13 women receiving placebo developed DVT/PE for a 3-fold increased risk (Ann Intern Med 2000; 132:689-96)

(b) the risk of estrogens was even higher in women with lower extremity fractures, cancer, surgery, or non-surgical hospitalization

(d) in New Zealand, the risk of venous thromboembolism was increased 9.6-fold in women using estrogens (Lancet 2000; 355:2088-90)

(6) polycythemia vera & other myeloproliferative diseases

(7) paroxysmal nocturnal hemoglobinuria

(8) nephrotic syndrome

(9) vasculitis

(10) increased factor VII and fibrinogen levels

(a) high factor VII levels confir a high incidence of recurrent venous thromboembolism (37% vs. 5% for the highest levels of factor VII vs. normal levels of factor VII) causing the authors of one paper to conclude that factor VII levels should be measured routinely in patients undergoing hypercoaguability work up (N Engl J Med 2000; 343:457-62)

(11) certain anti-cancer drugs

B. Diagnosis (Am J Respir Crit Care Med 1999; 160:1043-66):

1. Clinical:

a) very inaccurate (about50%)

b) frequent historical findings:

(1) pain

(2) swelling

(3) erythema

c) frequent physical findings:

(1) Homan's sign (present in 22% patients with DVT & 28% of patients without DVT)

(2) palpable cord (present in 30% patients with DVT & 27% of patients without DVT)

2. Duplex ultrasonography (Ann Int Med 1998; 129:1044-9)

a) 91% sensitive; 99% specific considering all DVT (NEJM 1989; 320:342-5).

(1) sensitivity approaches 100% for femoral DVT

b) will not detect isolated calf or iliac vein thrombosis

(1) sensitivity only 36% for isolated calf vein thrombosis (NEJM 1989; 320:342-5).

c) when used for serial non-invasive study to exclude DVT (eg. 3 studies over 8 days), duplex scan had a much better positive predictive value (94%) than IPG (83%) and was associated with a lower chance of diagnosis of thromboembolism in the subsequent 6 months if the study was negative (1.5% for duplex and 2.5% for IPG) (NEJM 1993; 329:1365-9).

d) for outpatients with suspected DVT, it is safe to withhold anticoagulation if an initial duplex ultrasound and a second duplex in 5-7 days are both negative (Ann Intern Med 1998; 128:1-7)

e) a positive duplex scan justifies treatment and does not require confirmation with venography (NEJM 329:1418-9).

f) duplex ultrasound should always be done in both legs with suspected DVT because of the high incidence of assymptomatic DVT in the contralateral leg (even when the ipsilateral leg has no DVT by duplex ultrasound) (Radiology 1996; 200:429:31)

3. Impedance plethysmography (IPG) (Am J Respir Crit Care Med 1994; 149:1044-7)

a) highly dependent on skill of technician

b) 95% sensitive; specificity variable depending on patient population

(1) recent studies suggest that the sensitivity may actually be lower than reported in the older literature.

(a) in one study, IPG was only 66% sensitive for proximal DVT in 384 patients (Ann Intern Med 1993; 118:25-30).

(b) in another study, IPG was only 65% sensitive for detection of proximal DVT in 132 patients (Arch Intern Med 1994; 154:1930-3)

c) not good for isolated calf vein thrombosis

d) a positive IPG justifies treatment and does not require confirmation with venography (NEJM 329:1418-9).

4. Venography

a) "gold standard"

b) inadequate visualization of external & common iliac veins in 18%

c) usually unnecessary if duplex scan or IPG is positive.

5. Helical leg CT

a) in a study of 52 patients with suspected DVT, helical CT of the leg was compared to conventional venography (Radiology 19967 200:423-8)

(1) sensitivity of helical CT = 100%

(2) specificity of helical CT = 96%

(3) helical CT used 80% less IV contrast dye

b) in a study of 116 patients with suspected pulmonary embolism undergoing both combined pulmonary artery/venous CT and venous ultrasound, CT detected all 15 cases of DVT identified on ultrasound PLUS 4 additional cases not identified by ultrasound (Radiology 2000; 216:744-51)

C. Treatment: refer to treatment of PE below

D. Calf Vein Thrombosis (Arch Int Med 1988; 148:2131-8):

1. 20% will propagate up the thigh to cause iliofemoral DVT

2. treat with heparin/coumadin like other DVTs

a) if anticoagulation risky, an alternative is to follow the patient with serial duplex scans

3. superficial vein thrombosis does not require treatment with anticoagulants but duplex ultrasonography should be performed to exclude concurrent calf vein thrombosis

VIII. Pulmonary Thromboembolism (N Engl J Med 1998; 339:93-104)

A. Background:

1. 5 million patients per year in U.S. have DVT - about 30% will have symptomatic and an additional 30% will have asymptomatic PE

2. PE accounts for about 50,000 deaths in the United States per year

a) fatality rates:

(1) if PE recognized and treated: mortality = 2.5% (NEJM 1992; 326:1240-5)

(2) if PE unrecognized and not treated: mortality = 30%

b) PE accounts for 15% of post-operative deaths & 10-20% of all in-hospital deaths

c) 20% of patients with PE will die of other medical conditions within 1 year of the diagnosis of PE with cancer & infection the most common causes of death (NEJM 1992; 326:1240-5)

B. Diagnosis (Am J Respir Crit Care Med 1999; 160:1043-66):

1. Clinical presentation

a) many PEs are asymptomatic

(1) in one study, 40% of patients with DVT had asymptomatic PE (JAMA; 271:223-5)

b) if symptomatic, most frequent symptoms (Arch Intern Med 1986; 146:961-7):

(1) chest pain (88%)

(2) dyspnea (84%)

(3) cough (53%)

(4) hemoptysis (30%)

c) physical exam:

(1) tachypnea & tachycardia most common

(a) 92% of patients with PE have respiratory rate > 14

(2) fever in 18-50%

d) in the PIOPED study, 97% of patients with PE had at least one of: pleuritic chest pain, dyspnea, or respiratory rate greater than 20/min

2. Chest x-ray

a) CXR often normal, especially in the acute phase

b) most common findings are:

(1) cardiomegaly - the most common acute chest x-ray abnormality (Chest 2000; 118:33-8)

(2) enlargement of main pulmonary artery [normal right PA = 16 mm (male), 15 mm (female)]

(3) elevated hemidiaphragm

(4) pulmonary opacities

(a) Hampton's hump: 3-5 cm opacity usually on right hemidiaphram

(b) atelectasis: usually 1-3 cm above diaphragm; horizontal, linear, & several cm long

(5) pleural effusions

(a) usually small & unilateral

(b) indicates pulmonary infarction

(6) regional oligemia

3. Plasma D-dimer levels:

a) can be elevated in patients with PE but sensitivity and specificity are relatively poor at 85% and 68% resepectively (Ann Intern Med 1998; 129:1006-11)

b) clinical utility remains unproven (Arch Intern Med 1996; 156:939-46)

c) in 173 patients undergoing pulmonary angiography for suspected PE (JAMA 1993; 270:2819-22):

(1) 3/35 patients with D-dimer < 500 ng/ml had PE by angiogram

(2) D-dimer levels were > 500 ng/ml in 42/45 patients with documented PE

(3) D-dimer levels were > 500 ng/ml in 96/128 patients without PE

d) in a study of outpatients, withholding anticoagulation in patients with D-dimer levels < 500 was associated with a < 1% chance of having a subsequent PE (Am J Resp Crit Care Med 1997; 156:492-6)

e) a negative D-dimer test in patients with cancer does not reliably exclude DVT and 21% of cancer patients with a normal D-dimer had a DVT compared to 3.5% of patients without cancer (Ann Int Med 1999; 131:417-23)

f) in a study of 314 patients with suspected PE, the D-dimer was much better for identifying patients with angiographically-proven segmental or larger pulmonary artery emboli (sensitivity = 93%) than subsegmental pulmonary artery emboli (sensitivity = 50%) (Am J Respir Crit Care Med 2002; 165:345-48)

g) use for diagnosis of DVT/PE not currently recommended (Chest 1998; 113:499-504) but if used, should only be used in conjunction as an adjunct to proven non-invasive venous studies

4. Arterial blood gases

a) A-a gradient increased in 95% of patients with proven PE (Arch Int Med 1988; 148:1617-9)

b) A-a gradient = [150 - (5/4 x pCO2)] - pO2

(1) normal = 4 + (age ÷ 4)

5. Ventilation/perfusion scan (Chest 1985; 88:819-28 & JAMA 1990; 263:2753-9)

a) definitions (JAMA 1987; 257:3257):

(1) normal: perfusion is normal

(2) low probability:

(a) small V/Q mismatches

(b) focal V/Q matches without x-ray abnormalities

(3) intermediate probability (indeterminate):

(a) diffuse severe airway obstruction

(b) matched V/Q and x-ray abnormalities

(4) high probability:

(a) perfusion defects much larger than x-ray abnormalities

(b) 2 or more segmental V/Q mismatches

b) ventilation defects can accompany pulmonary arterial embolization for up to 48 hours

c) ventilation/perfusion scans in patients with COPD (Chest 1992; 102:17-22)

(1) patients with COPD aremore likely to have indeterminate scans (60% of patients) and less likely to have normal (5% of patients) or high probability (5% of patients) scans compared to patients without cardiopulmonary disease

(2) A-a gradient, heart rate, respiratory rate, and symptoms do not help distinguish COPD patients with pulmonary embolus from those without pulmonary embolus

(3) a high probability V/Q scan or a normal V/Q scan has the same predictive value in patients with COPD as in those without COPD

(a) in the PIOPED study, all patients with COPD and a high probability V/Q scan had documented PE and none of those with a normal V/Q scan had a PE

d) predictive value:

(1) a high probability V/Q scan indicates true PE in approximately 86% of patients

(2) several prospective studies have indicated that a low probability V/Q scan does not rule out PE:

(a) Hull, et. al. found that 31% of patients with low probability V/Q scans had angiographic evidence of PE (Chest 1985; 88:819-28)

(b) in the PIOPED study, 14% of patients with low probability V/Q scans had angiographic evidence of PE (JAMA 1990; 263:2753-9)

(3) clinical decision making requires both the V/Q scan assessment and the clinical assessment of the likelihood of PE. The following table is the likelihood of angiographically verified PE given different V/Q scan probabilities coupled with different clinical probabilities from the PIOPED study:

	High Clinical Suspicion	Intermediate Clinical Suspicion	Low Clinical Suspicion
High Probability V/Q	96%	88%	56%
Intermediate Probability V/Q	66%	28%	16%
Low Probability V/Q	40%	16%	4%

e) decision making using the V/Q scan (refer to the flow diagram at the end of the handout) (Chest 1993; 103:1553-9)

(1) a normal V/Q scan effectively excludes clinically significant PE (Ann Int Med 1991; 114:300-6)

(2) a high probability V/Q scan plus high clinical suspicion justifies treatment without an angiogram

(3) a low probability V/Q scan plus low clinical suspicion justifies no treatment without an angiogram

(4) any other combination of V/Q scan probability and clinical suspicion requires further evaluation with either a duplex scan (or IPG) or a pulmonary angiogram

(a) a positive duplex scan or IPG justifies treatment without an angiogram

(b) a negative duplex scan or IPG indicates the need for an angiogram

(c) an important implication of this is that an indeterminate V/Q scan can never be a stopping point regardless of the clinical suspicion

(5) in approximately half of patients with suspected PE, angiogram can be avoided by using a combination of the V/Q scan probability, the clinical assessment, and (in certain cases) a duplex scan (or IPG)

6. Duplex ultrasound of the legs

a) in patients suspected of having PE, only 29% of duplex ultrasounds will be abnormal at the time of presentation (Ann Intern Med 1997; 126:775-81)

(1) it has been hypothesized that the clots in the deep veins of the legs have dislodged in patients with PE and thus are not present for identification by duplex ultrasound of the legs.

b) this test is best used serially to exclude clinically significant PE in patients with mild-moderate symptoms and a low or intermediate V/Q scan

7. Helical CT

a) initial reports indicate that it is highly specific and a positive is probably sufficient on which to base therapy (AJR 1995; 165:1131-8)

b) in a prospective study of 249 patients with suspected PE, the sensitivity of helical chest CT was 95% and the specificity was 97% (Thorax 1996; 51:23-8)

c) in a prospective study of 110 patients with suspected PE, helical CT had a sensitivity of 92% and specificity of 96% (Radiology 1999; 210:693-7)

d) in a prospective multicenter study of 1,041 patients in France, CT had a sensitivity of 82% but when combined with a clinical assessment of PE probability and with a duplex ultrasound of the lower extremities, then a combination of a negative CT + negative duplex + low clinical suspicion essentially excluded PE however, if the CT and the duplex were negative and the clinical suspicion was high, 5% of patients had PE (Lancet 2002; 360:1914-20)

e) the ultility of the helical CT is lower in critically ill patients is less - in a study of 22 patients in a surgical intensive care unit receiving both helical CT and traditional angiograms, the sensitivity of helical CT was found to be 55% overall (Arch Surg 2001; 136:505-11)

f) a major advantage of helical CT over ventilation-perfusion scanning is the ability to diagose other lung diseases, especially in patients with pre-existing cardiopulmonary disease

g) personal recommendations for current use:

(1) if you are at an institution where there are technicians who are experienced with the correct technique and there are radiologists experienced in interpreting the scans, CT can be used as a first line test in lieu of V/Q scans

(2) if you are at an institution where there are EITHER technicians who are not experienced in the proper technique OR radiologist lacking experience in interpreting the scans, V/Q is still the initial test of choice.

(3) if the clinical suspician is high, a negative helical CT does not exclude PE. In this situation, a second study (usually a duplex ultrasound of the legs) should be performed and if negative or indeterminate, then a pulmonary angiogram should be performed as a follow up test since false negative helical CT studies can occur

(4) if the clinical suspician is low, and the helical CT is positive, confirmation of PE with an alternative method (such as pulmonary angiogram) may be desirable since false positives studies can occur

(5) if the helical CT is "indeterminate" (usually becuase of improper dye injection timing or artifact), then additional tests are required (usually a pulmonary angiogram)

(6) if the patient has significant underlying lung disease and a baseline abnormal chest x-ray, then helical CT is probably preferable to V/Q scans as an initial technique since there is a high likelihood that V/Q will be "indeterminate" and require the patient to undergo additional testing to diagnose or rule out PE

(7) if the patient is at risk of dye injury (IV contrast dye allergy or renal insufficiency), then V/Q may be a better first study to avoid the large bolus of dye associated with helical CT, especially if a pulmonary angiogram becomes subsequently necessary

8. Magnetic resonance angiography

a) in one small preliminary study, gadolinium-enhanced MRI had a high sensitivity and specificity for PE when compared to traditional pulmonary angiography (N Engl J Med 1997; 336:1422-7)

(1) this study should be considered preliminary and clinical practice should be dictated by additional, double blinded studies.

b) MRI seems to be good for PE in the proximal pulmonary arteries but can miss up to 80% of thrombi in subsegmental pulmonary arteries (Radiology 1999 210:353-9)

9. Angiography

a) complication rate = 1.6% (0.3% mortality)

b) gold standard

c) false positives have been reported in:

(1) malignancy

(2) sarcoid

(3) Takayasu's arteritis

(4) angiosarcoma

C. Treatment (Chest 2001; 119:176S-193S):

1. heparinization strategies:

a) heparin should be begun when diagnosis is suspected, not when diagnosis is proven

(1) failure to achieve full anticoagulation within 24 hours increases risk of recurrent venous thrombosis by 15 times (N Engl J Med 1986; 315:1109-14)

(2) heparin does not cross the placenta and is safe during pregnancy

b) contraindicated in:

(1) intracranial bleeding

(2) intracranial lesions prone to bleeding

(a) in patients with primary & metastatic brain tumors, the risk of intracranial bleeding with anticoagulants equals risk of recurrent PE with Greenfield filter (Arch Int Med 1987; 147:2177-9)

(3) active internal bleeding

(4) treatment of heparin-induced thrombocytopenia

(a) lepirudin (Refludan)

i) inhibits free and clot bound thrombin

ii) 0.4 mg/kg (up to a maximum of 110 kg) IV over 20 seconds

iii) 0.15 mg/kg/hr (up to a maximum of 110 kg) x 2-10 days

iv) keep PTT 1.5-2.5 normal

v) adjust dose for renal failure

(b) Argatroban

i) thrombin inhibitor

ii) 2 ug/kg/min continuous infusion

iii) keep PTT 1.5 - 3.0 times normal

iv) adjust dose for hepatic failure (no dosage necessary for renal impairment)

i) mixture of anticoagulant glycosaminoglycans

ii) 2,250 u bolus then 400 u/hr x 4 hours then 300 u/hr x 4 hours then 150-200 u/hr

iii) dose adjust for wt less than 50 kg or more than 75 kg

iv) follwo anti Xa levels to keep between 0.5 u/ml an d0.8 u/ml

(d) do NOT give coumadin without one of the above anticoagulants because of the risk of venous limb gangrene

(e) plasmapheresis and/or IgG infusions may also be effective

c) treatment regimens:

(1) continuous intravenous infusion:

(a) adjust dose to achieve PTT = 1.5 - 2.5 times control (ie 60-105 seconds in the OSU lab)

(b) check PTT Q6 hr. until PTT is between 60 and 105 then check PTT QD; check PTT 6 hours after any dose change

(c) heparin drip adjustment based on patient's weight (Ann Intern Med 1993; 119:874-81 & Chest 2001; 119:64S-94S). Mix 25,00 units heparin in 250 ml D5W

i) initial dose: 80 units/kg then 16 units/kg/hour

ii) PTT < 45: additional 80 units/kg rebolus then increase drip by 3 units/kg/hour

iii) PTT 45-59: additional 40 units/kg rebolus than increase drip by 1 units/kg/hour

iv) PTT 60-105: no change

v) PTT 106-125: decrease drip by 2 units/kg/hour

vi) PTT 126-180: hold infusion for 1 hour, then decrease drip by 3 units/kg/hour

vii) PTT >180: hold infusion until dosing re-evaluated by physician

(d) check platelet count between day 3 & 5 and again betwen day 7 & 10 of heparin therapy and stop heparin if the platelet count falls below 100,000

(e) plasma heparin levels may be preferable to the PTT but are generally not widely available - therapeutic levels are 0.2-0.4 IU/mL

(f) continue heparin for minimum of 5 days (NEJM 1990; 322:1260-4) and the INR is therapeutic for 2 consecutive days

i) consider longer heparin administration in massive pulmonary emboli or iliofemoral thrombosis

(2) intermittent subcutaneous injection:

(a) as an alternative to 5 day administration of IV heparin in acute PE:

i) probably more effective and at least as safe as continuous infusion (Ann Int Med 1992; 116: 279-84)

ii) dose = 17,500 units SQ Q12 hours; adjust to keep the 6 hour post injection PTT between 1.5 and 2.5 times control (NEJM 1991; 324:1565-74)

iii) begin coumadin as you would for intravenous heparin

(b) as a long-term alternative to coumadin in patients in whom coumadin is contraindicated (eg., pregnancy) or ineffective (eg., Trousseau's syndrome of malignancy-associated DVT/PE)

(c) in most cases, low molecular weight heparin is preferred over subcutaneous unfractionated heparin because of ease of administration and predictability of anticoagulation effect

(3) low molecular weight heparin

(a) background: low molecular weight heparin offers longer half life, less hemorrhagic complications, greater ease of administration (once or twice per day subcutaneously), and perhaps less tendency toward anti-heparin antibody formation than standard heparin preparations (N Engl J Med 1997; 337:688-98)

1) enoxaparin (1 mg/kg SQ BID) given at home was equally effective as standard heparin given IV in the hospital for management of acute DVT (N Engl J Med 1996; 334:677-81)

2) nadroparin (200 IU/kg SQ BID) given at home was as effective as standard IV heparin given IV in the hospital for acute DVT ( N Engl J Med 1996; 334:682-7)

3) tinzaparin was a as effective as IV unfractionated heparin in patients being initially treated for PE (N Engl J Med 1997; 337:663-9)

4) in a large study of 1021 patients, reviparin was equivilent to unfractionated heparin in the initial treatement of DVT & PE with similar rates of recurrent DVT and major bleeding (N Engl J Med 1997;337:657-62).

5) enoxaparin 1.5 mg/kg SQ QD was equally effective as enoxaparin 1.0 mg/kg SQ BID or unfractionated heparin IV in 900 patients with DVT, including 273 with PE (Ann Intern Med 2001; 134:191-202)

6) dosing and administration costs for 5 day course (including infusion pumps, PTT assays, etc. - 1999 costs):

(i) standard heparin = $242

(ii) dalteparin (Fragmin) 200 u/kg SQ QD = $313

(iii) enoxaparin (Lovenox) 1 mg/kg SQ BID = $355

(b) current FDA-approved indications:

1) for inpatient treatment of DVT or PE

i) enoxaparin 1 mg/kg SQ BID

ii) or enoxaparin 1.5 mg/kg SQ QD

iii) tinzaparin 175 u/kg SQ QD

2) for outpatient treatment of DVT

i) enoxaparin 1 mg/kg SQ BID

ii) tinzaparin 175 u/kg SQ QD

3) precautions in use:

i) half life can be prolonged in renal insufficiency

ii) bleeding can occur in obese patients in whom correct dosing can be difficult to estimate

2. coumadin strategies (Chest 2001; 119:8S-38S):

a) contraindications:

(1) as per heparin

(2) additionally, coumadin contraindicated during pregnancy

(a) in the first trimester, coumadin causes congenital abnormalities

(b) coumadin crosses the placenta

b) begin same day heparin is begun

c) start with 10 mg/day for days #1 & #2, then administer at estimated daily dose (Ann Intern Med. 2003;138:714&endash;719)

(1) check PT/INR QD until stable

(a) most journals no longer report anticoagulation in terms of the PT (prothrombin time), nor are most of the consensus conferences making recommendations based on the PT. Instead, most now use the INR (international normalized ratio (NEJM 1993; 329:696-702).

(b) INR = (patient PT ÷ control PT)c

i) c = ISI (international sensitivity index) which is a measure of the responsiveness of a given thromboplastin (thromboplastin is used in the PT assay)

ii) the ISI varies from lot to lot of thromboplastin

iii) a typical ISI for the thromboplastin from the vendor used at OSU hospitals is 1.5

(2) adjust to achieve INR) of 2.0 - 3.0

(a) the INR is extremely inconvenient to calculate (must be recalculated every time the laboratory opens a new lot of thromboplastin) so many physicians simply try to keep the patient's PT 1.3 to 1.5 times the control PT if the individual laboratory reports values as PT rather than INR.

(b) the INR must be checked frequently (every 1-2 weeks) while patients are receiving coumadin

(c) patients must be cautioned about common foods and drugs which can alter coumadin's effect (often dramatically!) (Ann Intern Med 1994; 121:676-83)

d) need minimum of 4 - 5 days overlap with heparin

e) the duration of anti-coagulation is dependent on the patient population (N Engl J Med 1997; 336:433-4):

(1) although 4-6 weeks of anticoagulation is probably sufficient for DVT which occur as a consequence of surgery in patients with no additional risk factors; however, at least 3 months is recommended for non-post operative DVT (Lancet 1992; 340:873-6).

(2) 6 months of anticoagulation was associated with a lower incidence of DVT recurrance than 6 weeks of anticoagulation in a large multicenter study (NEJM 1995; 332:1661-5) and supported by a more recent double blind, placebo controlled study which concluded that 3 months was insufficient after a first episode of idiopathic venous thromboembolism (N Engl J Med 1999; 340:901-7)

(3) after a second episode of DVT or PE, the risk of a third thromboembolism is 21% and prophylactic indefinite anticoagulation may be considered after a second DVT/PE (N Engl J Med 1997; 336:393-8)

(4) patients with factor V Leiden have a 40% probability of recurrent thrombosis and may also be considered for indefinite anticoagulation (N Engl J Med 1997; 336:399-403) although many authorities recommend only a 6 month initial therapy and reserve indefinite anticoagulation for those patients with factor V Leiden with recurrent thrombosis

(5) patients with high levels of factor VIII have an increased incidence of recurrent thrombosis (NEJM 2000; 343:457-62)

(6) lifetime anticoagulation should be considered for patients with:

(a) cancer (until cured)

(b) homozygous factor V Leiden

(d) deficiency of antithrombin III, protein C, or protein S

(e) levels of factor VII above the 90th percentile (higher than 234 IU/dl)

(7) a preliminary study suggested that long-term, low dose anticoagulation may be beneficial. A recent study indicated that for patients with idiopathic pulmonary embolism treated with regular dose coumadin for 3 months and then coumadin to keep the INR between 1.5 - 2.0 thereafter, the incidence of recurrent PE was substantially reduced from 14% to less than 6% (N Engl J Med 2003; 348:1425-1434)

(8) however, a more recent study suggests that long term low dose coumadin (to keep the INR 1.5 - 2.0) is associated with twice as many episodes of recurrent venous thromboembolism as standard dose coumadin (to keep the INR 2.0 - 3.0). In this study, 738 patients who completed 3 months of standard dose heparin were randomized to receive long term low dose or standard dose heparin. the low dose group had 1.9 recurrences per 100 patient years whereas the standard dose group had 0.7 recurrences per 100 patient years. There was no differences in bleeding (N Engl J Med 2003; 349:631-9)

(9) in patients with cancer, low molecular weight heparin for 6 months was more effective than coumadin for 6 months for the treatment of venous thromboembolism (N Engl J Med 2003; 349:146-53)

(10) in a study of 326 patients with pulmonary embolism, the incidence of recurrent thromboembolism after completing a course of coumadin was the same whether patients received 3 or 12 months of coumadin indicating that indefinite treatment may be warrented in many patients (Ann Intern Med 2003; 139:19-25)

3. thrombolytic strategies (Chest 1999; 115:1695-1707):

a) indications:

(1) approved:

(a) angiographically-proven pulmonary embolus

i) proven to cause more rapid clot lysis than heparin

(1) angiograms done at 24 hours showed 3 times greater improvement with either urokinase or streptokinase compared to heparin (Circulation 1973; 47:1S-100S)

(2) V/Q scans done at 24 hours were twice as likely to show improvement in patients receiving thrombolytics compared to patients receiving heparin; however, at 7 days, V/Q scans were equally likely to show improvement in thrombolytic vs. heparin treated patients (Chest 1990; 98:1473-9)

ii) in a study of 66 patients with massive PE, there was no difference in outcome in patients treated with streptokinase vs. tPA (J Am Coll Cardiol 1998; 31:1057-63)

iii) no studies have proven advantage over heparin with respect to survival in unselected patients with PE but a small study of 8 patients demonstrated that patients with PE and shock were more likely to survive if they received thrombolytics than heparin (J Thromb Thrombolysis 1995; 2:227-9) and another study indicated that in patients with normal systemic blood pressure and RV dysfunction by cardiac echo, survival and the risk of recurrent PE is improved with thrombolytics compared to heparin alone (Circulation1997; 96:882-8). Patients with RV dysfunction by echo have a high in-hospital mortality rate (5%) despite having normal systemic blood pressure (Circulation 2000; 101:2817-22). The most recent study examining thrombolytics in PE associated with RV dysfunction foundthat t-PA reduced the need for "rescue thrombolysis" but did not affect the rate of recurrent PE or the mortality rate of PE (N Engl J Med 2002; 347:1143-50)

iv) unlike the case with myocardial infarction, thrombolytics do not result in complete clot lysis - probably because in myocardial infarction, the clot is very fresh but in PE, the clot is usually older

v) a recent review drew the following conclusions (Chest 1999; 115:1695-707)

(1) thrombolytics improve survival in patients with PE and shock and in patients with PE and RV dysfunction in the absence of shock

(2) all available thrombolytic agents are equally safe

(3) a 2 hour infusion of t-PA results in more rapid clot lysis compared to 12 or 24 hour infusions of urokinase or streptokinase

(4) intrapulmonary artery infusions of thrombolytics are not superior to systemic infusions

(5) thrombolytics are most effective when used immediately after PE but may have benefit up to 14 days post PE

(6) thrombolytics increase the chance of major bleeding and intracranial hemorrhage

vi) bleeding complications:

(1) in the early urokinase pulmonary embolism trial, the incidence of major bleeding from urokinase is twice that from heparin (45% vs. 27%) (JAMA 1970; 214:2163-72); other, more recent studies suggest that major bleeding occurs in about 20% of patients treated with either t-PA or urokinase (J Am Coll Cardiol 1992; 19:239-45) and with either t-PA, streptokinase, or urokinase, major bleeding is 3 times more likely than with heparin.

(2) in the ISIS-3 myocardial infarction trial (41,299 patients), major bleeding was slightly more likely with streptokinase than with t-PA (13,607 patients vs. 13,569 patients but intracranial bleeding was much less likely with streptokinase than with t-PA (32 patients vs. 89 patients) (Lancet 1992; 339:753-0).

(3) in the recent GUSTO myocardial infarction trial (2,431 patients), major bleeding, minor bleeding, and intracranial bleeding were all more likely (but not statistically significantly so) with streptokinase than with t-PA (NEJM 1993; 329:1615-22).

(4) bleeding complications in one study were higher if patients were diagnosed by angiography (14% - 18/129 patients) than if patients were diagnosed by non-invasive studies + V/Q scans (Ann Intern Med 1994; 121:313-7)

vii) in the absence of convincing prospective studies, use reserved for massive PE:

(1) PE with hemodynamic compromise (shock or respiratory failure)

(2) 2 or more lobar arteries occluded on pulmonary angiogram

(3) possibily in PE with hemodynamic stability but RV dysfunction by cardiac echo

(b) deep venous thrombosis

i) use in DVT is controversial (pro = Am J Med 1990; 88:389-95; con = Arch Int Med 1989; 149:1841-45)

ii) may result in better preservation of venous valves than heparin

iii) may reduce post-phlebitic syndrome compared to heparin

iv) pending better prospective studies, heparin is preferred (N Engl J Med 1994;330:1864-9)

v) catheter-directed thrombolysis may be preferable compared to heparin for massive ileo-femoral thrombosis

b) contraindications (Am J Med 1990; 88: 389-95):

(1) absolute:

(a) active internal bleeding

(b) CVA, intracranial surgery, or intraspinal surgery within the past 2 months

(2) relative (major):

(a) surgery, obstetrical delivery, organ biopsy, or puncture of non-compressible vessel within the previous 10 days

(b) trauma, GI bleeding, or CPR within the previous 10 days

(d) severe uncontrolled hypertension

(3) relative (minor):

(a) recent minor trauma

(b) diabetic hemorrhagic retinopathy

(d) coagulopathy

(e) pregnancy

(f) septic thrombophlebitis

(g) endocarditis

(h) age > 75

c) streptokinase

(1) load with 250,000 units IV over 30 minutes

(2) continuous drip of 100,000 units per hour for 24 hours

(3) additional contraindications:

(a) recent streptococcal infection

(b) receipt of streptokinase in previous 6 months

(4) begin heparin after 24 hour infusion complete

(5) 1999 OSU patient cost per course about $588

d) tissue plasminogen activator (t-PA)

(1) single dose of 100 mg IV over 2 hours

(2) begin heparin immediately following the end of t-PA infusion (without loading dose)

(3) 1999 OSU patient cost per course about $2,200

e) acylated plasminogen-streptokinase-activator complex (APSAC, anistreplase, Eminase)

(1) currently approved only for acute myocardial infarction and not for DVT/PE by the FDA

f) intrapulmonary artery thrombolytics

(1) urokinase given into the right atrium was effective in 16 patients with massive PE accompanied by shock at a lower dose of urokinase than is conventionally used with intravenous urokinase in 1 uncontrolled trial (Chest 1992; 102:341-6)

(a) 500,000 unit bolus then 85,000 units/hr. for 12 hours

(b) only 1/16 patients developed bleeding requiring transfusion compared to historical controls (20% incidence of bleeding requiring transfusion in patients receiving conventional dose urokinase intravenously)

(2) in a randomized, multicenter trial, equivalent doses of t-PA given intravenously or into the pulmonary artery were equally effective in lysing clot and reducing PA pressures; both methods of administration reduced fibrinogen levels equally, and had similar bleeding complications (Circulation 1988; 77:353-60)

(3) overall, there is no convincing benefit to local (pulmonary artery) as opposed to systemic administration through a peripheral vein (Chest 1998; 114:561-578S)

4. inferior venal caval interruption (Greenfield & birdnest filters [Chest 1988; 93:170-5])

a) indications

(1) absolute contraindication to anticoagulation present

(2) severe internal bleeding developing during anticoagulation

(3) recurrent emboli despite adequate anticoagulation

(4) possibly prophylactically in patients with a high likelihood of PE in whom preventative anticoagulation is not possible (eg. pelvic trauma) however their benefit as a prophylactic tool has been questioned (N Engl J Med 1998; 338:409-15)

(5) possibly as an adjunct to anticoagulation in patients with severe cardiopulmonary compromise and PE in whom recurrent PE would likely be fatal

(6) as an adjunct to surgical embolectomy in patients with chronic thrombotic large vessel pulmonary artery obstruction

b) Complications:

(1) long term recurrent DVT - incidence 2-fold greater than patients treated without filters (N Engl J Med 1998; 338:409-15)

(2) occlusion - 3.1%

(3) recurrent PE - 2.4%

(4) filter migration

(a) minor migration in about 10%

(b) only one case of significant proximal migration reported

(5) caval perforation - 2 reported cases

c) technique:

(1) introduced via catheter into either jugular or femoral veins

(2) struts engage caval wall to anchor filter

(3) can be placed in the superior vena cava also (J Vasc Surg 1996; 23:498-503)

5. emergency surgical embolectomy (Chest 1998; 114:561-578S):

a) angiographically-proven massive PE with hemodynamic compromise and,

b) availability of experienced thoracic surgical team and,

c) failure of or contraindication to thrombolysis

d) operative mortality = 35% (Thorac Cardiovasc Surg 1999 47:5-8)

6. catheter techniques:

a) transvenous extraction of emboli (Chest 1991; 100:655-8)

(1) results in an improvement in 61% of patients

(2) mortality = 28%

b) rotational catheter fragmentation (J Am Coll Cardiol 2000; 36:375-80)

(1) preliminary uncontrolled study suggeting an increase in hemodynamic profiles

D. Prophylaxis (Chest 2001; 119:132S-175S):

1. Justification:

a) incidence of DVT high in medical & surgical patients (range = 15% of medical patients to 40-70% of hip surgery patients [Chest 1989; 95:37S - 51S])

b) perioperative low dose heparin reduces fatal PE by 2/3 in general, orthopedic, & urologic surgery (NEJM 1988; 318:1162-72)

2. Recommendations:

a) general surgery patients:

(1) low risk, minor general surgery patients < 40 years: none

(2) moderate risk general surgery patients and those 40-60 years: elastic stockings, low dose SQ heparin, low molecular weight heparin, or intermittent pneumatic compression

(3) high risk general surgery patients and those over age 60: low dose SQ heparin or low molecular weight heparin (intermittent pneumatic compression if patient is prone to wound hematoma formation with low dose heparin)

(4) very high risk general surgery patients: low dose SQ heparin (or low molecular weight heparin) + intermittent pneumatic compression

b) gynecologic surgery patients:

(1) brief procedures for benign disease - early ambulation only

(2) major procedures for benign disease: low dose SQ heparin

(3) extensive procedures for malignancy: low dose SQ heparin TID or SQ heparin plus intermittant pneumatic compression

c) urologic surgery:

(1) TURP and other low risk procedures: early ambulation only

(2) major open procedures: low dose SQ heparin, elastic stockings, intermittant pneumatic compression, or low molecular weight heparin

(3) highest risk patients: elastic stockings and/or intermittant pneuatic compression plus either low dose SQ heparin or low molecular weight heparin

d) orthopedic surgery:

(1) total hip replacement: adjusted dose heparin (Q12 hour SQ injection sufficient to keep PTT in upper range of normal [eg. 31-36 seconds] when checked 6 hours after heparin injection), low molecular weight heparin, or coumadin (dosed to keep the INR 2.0 - 3.0)

(a) once daily SQ low molecular weight heparin (enoxaparin) is simpler and at least as effective in hip & knee replacement as coumadin (NEJM 1993; 329:1370-6) or standard heparin (Ann Intern Med 1993; 119:1105-12).

i) the patient cost of enoxaparin is about 10 times that of regular heparin ($62 per dose vs. $6 per dose)

ii) recommended dose is 30 mg SQ BID x 7-10 days, beginning as soon as clinically possible after surgery (but within 24 hours of surgery); 40 mg SQ QD was equally effective in one study (Ann Intern Med 1994;121:81-89)

iii) after discharge from the hospital, the incidence of DVT is 40% in the subsequent 1 month and this can be reduced to 20% if low molecular weight heparin is continued for 1 month post-op (N Engl J Med 1996; 335:696-700)

(b) fondaparinux (2.5 mg SQ QD) was significantly better than enoxaparin (30 mg SQ BID) at preventing post-operative DVT in patients undergoing knee surgery (N Engl J Med 2001; 345:1305-1310). Fondaparinux was also more effective than enoxaparin at preventing DVT after hip fracture surgery (N Engl J Med 2001; 345:1298-1304). Fondaparinux is a synthetic factor Xa inhibitor.

(c) ximelagatran, an oral direct thrombin inhibitor, was as effective as coumadin in preventing thromboembolism in patients undergoing knee replacement. Ximelagatran does not require coagulation monitoring, unlike coumadin (Ann Intern Med 2002; 137:648-55)

(d) the incidence of post-op DVT remains high even with prophylaxis

(2) hip fracture patients: coumadin (dosed to keep the INR 2.0 - 3.0) or low molecular weight heparin started pre-operatively

(3) total knee replacement surgery: low molecular weight heparin, coumadin, or intermittent pneumatic compression

(a) low molecular weight heparin is more effective in preventing post-knee arthroplasty DVT than coumadin (Ann Intern Med 1996; 124:619-26)

e) multiple trauma patients:

(a) low molecular weight heparin [superior to low dose standard heparin (N Engl J Med 1996; 335:701-7)]

(b) intermittant pneumatic compression devices if low molecular weight heparin is contraindicated

(c) screening duplex ultrasound if inadequate prophylaxis given; IVC filter only if duplex ultrasound is positive

f) intracranial neurosurgical patients: intermittent pneumatic compression with or without elastic stockings (low molecular weight heparin or low dose SQ heparin are alternatives)

g) acute spinal cord injury with paralysis patients: low molecular weight heparin

h) medical patients:

(1) Deep venous thrombosis occurred in 15% of acutely ill hospitalized patients with medical illnesses; this incidence was reduced to 5.5% in patients receiving 40 mg/day enoxaparin prophylactically (N Engl J Med 1999; 341:793-800)

(2) Recommendations by the 2001 ACCP consensus report:

(a) myocardial infarction: low dose SQ heparin or IV heparin

(b) ischemic stroke and lower extremity paralysis: low dose SQ heparin or low molecular weight heparin

(c) other medical patients requiring bedrest (eg., respiratory failure, severe congestive heart failure, etc.): low dose SQ heparin or low molecular weight heparin

2. drug dosing:

(1) SQ heparin: 5,000 u BID

(2) enoxaparin: 30 mg BID or 40 mg QD

(3) dalteparin: 2,500 to 5,000 u QD

(4) tinzaparin: 75 u/kg QD

(5) nadroparin: 2,850 u QD

3. special note regarding low molecular weight heparin for prophylaxis: lumbar puncture or epidural catheters should be placed with caution because of the possibility of spinal or epidural hematomas

4. aspirin: a recent multi-center study of 13,356 patients undergoing surgery indicated that aspirin (160 mg/day) decreased the incidence of DVT/PE by a third (Lancet 2000; 355:1288-9). Pending controlled studies comparing aspirin to heparin, it's use is probably best reserved for patients with contraindications to heparin prophylaxis

last updated 11/28/03

TOP OF THE PAGE

Return to the For Residents Page