Cleveland Clinic Journal of Medicine

A 50-year-old man developed severe chest pain and collapsed to the floor. The pain was sudden in onset, was burning in quality, and was located in the center of his chest. Emergency medical services arrived a few minutes later and found the patient diaphoretic and cyanotic, with an initial blood pressure of 74/54 mm Hg and a heart rate of 125 beats per minute. He was rushed to the hospital.

His medical history was unremarkable. He smoked one pack of cigarettes per day for 20 years. His father died of a “heart attack” at age 52.

In the emergency department he underwent echocardiography with a portable handheld unit, which showed a pericardial effusion and cardiac tamponade. He was sent for emergency computed tomography of the chest, which revealed an aneurysm of the aortic root and acute type A (Stanford classification) aortic dissection with hemopericardium.

He underwent emergency cardiac surgery. At the time of surgery, he was in cardiogenic shock from aortic dissection complicated by severe aortic regurgitation and cardiac tamponade with hemopericardium. The aortic valve was trileaflet. A 27-mm St. Jude composite valve graft root replacement was performed.

The patient did well and was discharged home 7 days after surgery. Pathologic study of the aorta revealed cystic medial degeneration. He did not have any features of Marfan syndrome or Loeys-Dietz syndrome. His three children underwent evaluation, and each had a normal physical examination and echocardiographic test results.

A HIGH INDEX OF SUSPICION IS CRITICAL

Acute aortic dissection is the most common aortic catastrophe, with an incidence estimated at 5 to 30 per 1 million people per year, amounting to nearly 10,000 cases per year in the United States.^1–4

The diagnosis of acute aortic dissection has many potential pitfalls.^2,3 Aortic dissection may mimic other more common conditions, such as coronary ischemia, pleurisy, heart failure, stroke, and acute abdominal illness. Because acute aortic dissection may be rapidly fatal, one must maintain a high index of suspicion.^2,3 Prompt diagnosis and emergency treatment are critical.

WHAT CAUSES AORTIC DISSECTION?

One hypothesis is that acute aortic dissection is caused by a primary tear in the aortic intima, with blood from the aortic lumen penetrating into the diseased media leading to dissection and creating a true and false lumen.² Another is that rupture of the vasa vasorum leads to hemorrhage in the aortic wall with subsequent intimal disruption, creating the intimal tear and aortic dissection.

Once a dissection starts, pulsatile flow of blood within the aortic wall causes it to extend. The dissection flap may be localized, but it often spirals the entire length of the aorta. Distention of the false lumen with blood may cause the intimal flap to compress the true lumen and potentially lead to malperfusion syndromes.

CLASSIFIED ACCORDING TO LOCATION

It is important to recognize the location of the dissection, as the prognosis and treatment depend on whether the ascending aorta is involved.^2,3 For classification purposes, the ascending aorta is the portion proximal to the brachiocephalic artery, while the descending aorta is the portion distal to the left subclavian artery.³

The DeBakey classification defines a type I aortic dissection as one that begins in the ascending aorta and extends at least to the aortic arch or beyond. Type II dissections involve the ascending aorta only, while type III dissections begin in the descending aorta, most often just distal to the left subclavian artery.

The Stanford classification scheme divides dissections into type A and type B. Type A dissections involve the ascending aorta, while type B dissections do not involve the ascending aorta.

Which classification scheme is used is not important. However, identifying patients with dissection of the ascending aorta (DeBakey type I or type II or Stanford type A) is critical, as emergency cardiac surgery is recommended for this type of dissection.^2,3 For the purposes of this paper, the Stanford classification scheme will be used.

Dissection that involves the ascending aorta most commonly occurs in people ages 50 to 60, whereas acute dissection of the descending aorta typically occurs in people 10 years older.^1,2

An acute aortic dissection is one that has occurred within 2 weeks of symptom onset. A chronic dissection is one that occurred more than 2 weeks after symptoms began.

DISEASES AND CONDITIONS ASSOCIATED WITH AORTIC DISSECTION

Hypertension and disorders leading to disruption of the normal structure and function of the aortic wall. About 75% of patients with acute aortic dissection have underlying hypertension.^1–3

Cystic medial degeneration is a common pathologic feature in many cases of aortic dissection.

Cocaine use and intense weight-lifting increase the shear stresses on the aorta.^2,3

Inflammatory aortic diseases such as giant cell arteritis.

Pregnancy can be complicated by aortic dissection, usually in the setting of an underlying aortopathy.⁵

Iatrogenic aortic dissection accounts for about 4% of cases, as a result of cardiac surgery, catheterization, stenting, or use of an intra-aortic balloon pump.¹

Aortic aneurysm. Patients with thoracic aortic aneurysm are at higher risk of aortic dissection, and the larger the aortic diameter, the higher the risk.^2,3,6 In the International Registry of Acute Aortic Dissection (IRAD), the average size of the aorta was about 5.3 cm at the time of acute dissection. Importantly, about 40% of acute dissections of the ascending aorta occur in patients with ascending aortic diameters less than 5.0 cm.^7,8

Thus, many factors are associated with acute dissection, and specific reasons leading to an individual’s susceptibility to sudden dissection are poorly understood.

CLINICAL FEATURES OF ACUTE AORTIC DISSECTION

Because the symptoms of acute dissection may mimic other, more common conditions, one of the most important factors in the diagnosis of aortic dissection is a high clinical suspicion.^1–3

What is the pretest risk of disease?

Recently, the American College of Cardiology (ACC) and the American Heart Association (AHA) released joint guidelines on thoracic aortic disease.³ These guidelines provide an approach to patients who have complaints that may represent acute thoracic aortic dissection, the intent being to establish a pretest risk of disease to be used to guide decision-making.³

The focused evaluation includes specific questions about underlying conditions, symptoms, and findings on examination that may greatly increase the likelihood of acute dissection. These include:

• High-risk conditions and historical features associated with aortic dissection, such as Marfan syndrome and other genetic disorders (Table 2), bicuspid aortic valve, family history of thoracic aortic aneurysm or dissection, known thoracic aortic aneurysm, and recent aortic manipulation
• Pain in the chest, back, or abdomen with high-risk features (eg, abrupt onset, severe intensity, or a ripping or tearing quality)
• High-risk findings on examination (eg, pulse deficits, new aortic regurgitation, hypotension, shock, or systolic blood pressure differences).

Using this information, expedited aortic imaging and treatment algorithms have been devised to improve the diagnosis.³

Using the IRAD database of more than 2,500 acute dissections, the diagnostic algorithm proposed in the ACC/AHA guidelines was shown to be highly sensitive (about 95%) for detecting acute aortic dissection.⁴ In addition, using this score may expedite evaluation by classifying certain patients as being at high risk of acute dissection.^3,4

Important to recognize is that almost two-thirds of patients who suffered dissection in this large database did not have one of the “high-risk conditions” associated with dissection.⁴ Additionally, the specificity of the ACC/AHA algorithm is unknown, and further testing is necessary.⁴

Acute onset of severe pain

More than 90% of acute dissections present with acute pain in the chest or the back, or both.^1–3 The pain is usually severe, of sudden onset, and often described as sharp or, occasionally, tearing, ripping, or stabbing. The pain usually differs from that of coronary ischemia, being most severe at its onset as opposed to the less intense, crescendo-like pain of angina or myocardial infarction. The pain may migrate as the dissection progresses along the length of the aorta or to branch vessels. It may abate, leading to a false sense of security in the patient and the physician.³ “Painless” dissection occurs in a minority, usually in those with syncope, neurologic symptoms, or heart failure.^1–3

The patient with acute dissection may be anxious and may feel a sense of doom.

Acute heart failure, related to severe aortic regurgitation, may be a predominant symptom in dissection of the ascending aorta.

Syncope may occur as a result of aortic rupture, hemopericardium with cardiac tamponade, or acute neurologic complications.

Vascular insufficiency may occur in any branch vessel, leading to clinical syndromes that include acute myocardial infarction, stroke, paraplegia, paraparesis, mesenteric ischemia, and limb ischemia.

PHYSICAL FINDINGS CAN VARY WIDELY

Findings on physical examination in acute aortic dissection vary widely depending on underlying conditions and on the specific complications of the dissection.

Although the classic presentation is acute, severe pain in the chest or back in a severely hypertensive patient with aortic regurgitation and pulse deficits, most patients do not have all these characteristics.⁴ Most patients with type B dissection are hypertensive on presentation, but many with type A dissection present with normal blood pressure or hypotension.¹ Pulse deficits (unequal or absent pulses) are reported in 10% to 30% of acute dissections and may be intermittent as the dynamic movement of the dissection flap interferes with branch vessel perfusion.^1–3

• Aortic leaflet prolapse or distortion of the leaflet alignment
• Malcoaptation of the aortic leaflets from dilation of the aortic root and annulus
• Prolapse of the intimal flap across the aortic valve, interfering with valve function
• Preexisting aortic regurgitation from underlying aortic root aneurysm or primary aortic valve disease (such as a bicuspid aortic valve).

Neurologic manifestations are most common in dissection of the ascending aorta and are particularly important to recognize, as they may dominate the clinical presentation and lead to delay in the diagnosis of dissection.^2,3 Neurologic syndromes include:

• Persistent or transient ischemic stroke
• Spinal cord ischemia
• Ischemic neuropathy
• Hypoxic encephalopathy.

These manifestations are related to malperfusion to branches supplying the brain, spinal cord, or peripheral nerves.⁹

Syncope is relatively common in aortic dissection and may be related to acute hypotension caused by cardiac tamponade or aortic rupture, cerebral vessel obstruction, or activation of cerebral baroreceptors.^2,9 It is important to consider aortic dissection in the differential diagnosis in cases of unexplained syncope.³

Aortic dissections may extend into the abdominal aorta, leading to vascular complications involving one or more branch vessels.¹⁰ The renal artery is involved in at least 5% to 10% of cases and may lead to renal ischemia, infarction, renal insufficiency, or refractory hypertension.²Mesenteric ischemia or infarction occurs in about 5% of dissections, may be difficult to diagnose, and is particularly dangerous.^2,8 Aortic dissection may extend into the iliac arteries and may cause acute lower extremity ischemia.

Acute myocardial infarction due to involvement of the dissection flap causing malperfusion of a coronary artery occurs in 1% to 7% of acute type A aortic dissections.^1–3 The right coronary artery (Figure 2) is most commonly involved, leading to acute inferior myocardial infarction. Acute myocardial ischemia and infarction in the setting of dissection may lead to a delay in the diagnosis of dissection and to bleeding complications from antiplatelet and anticoagulant drugs given to treat the acute coronary syndrome.

Cardiac tamponade, occurring in about 10% of acute type A dissections, portends a higher risk of death.^2,3

Additional clinical features of aortic dissection include a left-sided pleural effusion, usually related to an inflammatory response. An acute hemothorax may occur from rupture or leaking of a descending aortic dissection.

FINDINGS ON RADIOGRAPHY AND ELECTROCARDIOGRAPHY

Reproduced with permission from: Braverman AC, et al. Diseases of the aorta. In: Bonow RO, et al. Braunwald's Heart Disease, 9th edition. Elsevier: Philadelphia, PA; 2011.

Electrocardiography usually has normal or nonspecific findings, unless acute myocardial infarction complicates the dissection.

D-DIMER LEVELS

Biomarkers for the diagnosis of acute aortic dissection are of great interest.

D-dimer levels rise in acute aortic dissection as they do in pulmonary embolism.¹¹ A D-dimer level greater than 1,600 ng/mL within the first 6 hours has a very high positive likelihood ratio for dissection, so this test may be useful in identifying patients with a high probability for dissection. In the first 24 hours after symptom onset, a D-dimer level of less than 500 ng/mL has a negative predictive value of 95%. Thus, elevations in D-dimer may help decide which imaging to perform in a patient presenting with chest pain or suspicion of dissection.¹¹

However, D-dimer levels may not be elevated in dissection variants, such as aortic intramural hematoma or penetrating aortic ulcer. Additionally, once 24 hours have elapsed since the dissection started, D-dimer levels may no longer be elevated. The current ACC/AHA guidelines on thoracic aortic disease concluded that the D-dimer level cannot be used to rule out aortic dissection in high-risk individuals.³

Additional studies may clarify the appropriate role of the D-dimer assay in diagnosing aortic dissection.

DEFINITIVE IMAGING STUDIES: CT, MRI, TEE

Contrast-enhanced computed tomography (CT), magnetic resonance imaging (MRI), and transesophageal echocardiography (TEE) all have very high sensitivity and specificity for the diagnosis of aortic dissection.^2,3 The choice of imaging study often depends on the availability of these studies, with CT and TEE being the most commonly performed initial studies.

MRI is outstanding for detecting and following aortic dissection, but it is usually not the initial study performed because of the time required for image acquisition and because it is generally not available on an emergency basis.

Reproduced with permission from: Braverman AC, et al. Diseases of the aorta. In: Bonow RO, et al. Braunwald's Heart Disease, 9th edition. Elsevier: Philadelphia, PA; 2011.

While transthoracic echocardiography (TTE) can detect aortic dissection, its sensitivity is much lower than that of other imaging tests.^2,3 Therefore, negative findings on TTE do not exclude aortic dissection.

MANAGEMENT OF AORTIC DISSECTION

When acute aortic dissection is diagnosed, multidisciplinary evaluation and treatment are necessary. Time is of the essence, as the death rate in acute dissection may be as high as 1% per hour during the first 24 hours.^1–3 All patients with acute aortic dissection, whether type A or type B, should be transferred to a tertiary care center with a staff experienced in managing aortic dissection and its complications.³ Emergency surgery is recommended for type A aortic dissection, whereas type B dissection is generally treated medically unless complications occur.^2,3

The cornerstone of drug therapy is the prompt reduction in blood pressure with a beta-blocker to reduce shear stresses on the aorta. Intravenous agents such as esmolol (Brevibloc) or labetalol (Normodyne) are usually chosen. Sodium nitroprusside may be added to beta-blocker therapy for rapid blood pressure control in appropriate patients. The patient may require multiple antihypertensive medications. If hypertension is refractory, one must consider renal artery hypertension due to the dissection causing renal malperfusion.² Acute pain may also worsen hypertension, and appropriate analgesia should be used.

Definitive therapy in acute dissection

Patients with acute type A dissection require emergency surgery,^2,3 as they are at risk for life-threatening complications including cardiac tamponade from hemopericardium, aortic rupture, stroke, visceral ischemia, and heart failure due to severe aortic regurgitation. When aortic regurgitation complicates acute type A dissection, some patients are adequately treated by resuspension of the aortic valve leaflets, while others require valve-sparing root replacement or prosthetic aortic valve replacement.

Surgical therapy is associated with a survival benefit compared with medical therapy in acute type A dissection.¹ The 14-day mortality rate for acute type A dissection treated surgically is about 25%.¹ Patients with high-risk features such as heart failure, shock, tamponade, and mesenteric ischemia have a worse prognosis compared with those without these features.^2,12,13

Acute type B aortic dissection carries a lower rate of death than type A dissection.^1–3 In the IRAD cohort, the early mortality rate in those with type B dissection treated medically was about 10%.¹ However, when complications such as malperfusion, shock, or requirement for surgery occur in type B dissection, the mortality rate is much higher,^2,14 with rates of 25% to 50% reported.²

Thus, initial medical therapy is the preferred approach to acute type B dissection, and surgery or endovascular therapy is reserved for patients with acute complications.^2,3 Typical indications for surgery or endovascular therapy in type B dissection include visceral or limb ischemia, aortic rupture, refractory pain, and aneurysmal dilation (Table 3).²

Endovascular therapy in aortic dissection

The high mortality rate with open surgery in acute type B dissection has spurred tremendous interest in endovascular treatments for complications involving the descending aorta and branch vessels.²

Fenestration of the aorta and stenting of branch vessels were the earliest techniques used in complicated type B dissection. By fenestrating (ie, opening) the intimal flap, blood can flow from the false lumen into the true lumen, decompressing the distended false lumen.

Endovascular stenting is used for acute aortic rupture, for malperfusion syndromes, and for rapidly enlarging false lumens. Endovascular grafts may cover the area of a primary intimal tear and thus eliminate the flow into the false channel and promote false-lumen thrombosis. Many patients with complicated type B dissection are treated with a hybrid approach, in which one segment of the aorta, such as the aortic arch, is treated surgically, while the descending aorta receives an endovascular graft.²

Patients with a type B dissection treated medically are at risk for late complications, including aneurysmal enlargement and subsequent aortic rupture. The Investigation of Stent Grafts in Aortic Dissection (INSTEAD) trial included 140 patients with uncomplicated type B dissection and compared drug therapy with endovascular stent grafting.¹⁵ After 2 years of follow-up, there was no difference in the rate of death between the two treatment groups. Patients receiving endovascular grafts had a higher rate of false-lumen thrombosis.

More studies are under way to examine the role of endovascular therapy in uncomplicated type B dissection.

AORTIC DISSECTION VARIANTS

Aortic intramural hematoma

Aortic intramural hematoma is a form of acute aortic syndrome in which a hematoma develops in the aortic media and no intimal flap is visualized either by imaging or at surgery.^2,3,16 It is important to recognize this clinical entity in a patient presenting with acute chest or back pain, as sometimes it is mistaken for a “thrombus in a nonaneurysmal aorta.” Intramural hematoma accounts for 5% to 25% of acute aortic syndromes, depending on the study population (it is more common in Asian studies).^2,3,17 It may present with symptoms similar to classic aortic dissection and is classified as type A or type B, depending on whether the ascending aorta is involved.

Reproduced with permission from: Braverman AC, et al. Diseases of the aorta. In: Bonow RO, et al. Braunwald's Heart Disease, 9th edition. Elsevier: Philadelphia, PA; 2011.

Patients with an intramural hematoma may progress to having complications such as hemopericardium, classic aortic dissection, aortic rupture, or aneurysmal dilation.^2,3 However, many cases of type B aortic intramural hematoma result in complete resorption of the hematoma over time. In general, like classic aortic dissection, type A intramural hematoma is treated with emergency surgery and type B with initial medical therapy.^2,3

There are reports from Southeast Asia of successful initial medical therapy for type A intramural hematoma, with surgery used for acute complications.¹⁸ In the Western literature, improved outcomes are reported with initial surgical therapy.¹⁷ Given the unpredictable nature of type A intramural hematoma, most experts recommend surgical therapy for appropriate candidates with acute type A intramural hematoma.^2,3,19

Penetrating atherosclerotic ulcer of the aorta

Penetrating atherosclerotic ulcer of the aorta, another acute aortic syndrome, results from acute penetration of an atherosclerotic aortic lesion through the internal elastic lamina into the media.^2,3,20 It is often associated with bleeding into the media, or intramural hematoma. While the ulcer may be found incidentally on imaging studies, especially in patients with severe aortic atherosclerosis, the typical presentation is acute, severe chest or back pain. It occurs most often in the descending aorta and the abdominal aorta.

Penetrating atherosclerotic ulcer may lead to pseudoaneurysm formation, focal aortic dissection, aortic rupture, or late aortic aneurysm.²

Reproduced with permission from: Braverman AC, et al. Diseases of the aorta. In: Bonow RO, et al. Braunwald's Heart Disease, 9th edition. Elsevier: Philadelphia, PA; 2011.

LONG-TERM MANAGEMENT AFTER AORTIC DISSECTION

After hospital discharge, patients with aortic dissection require lifelong management. This includes blood pressure control, lifestyle modification, serial imaging of the aorta with CT or MRI, patient education about the condition, and, when appropriate, screening of family members for aortic disease.^5,21

Reported survival rates after hospitalization for type A dissection are 52% to 94% at 1 year and 45% to 88% at 5 years.^2,22 The 10-year actuarial survival rate for those with acute dissection who survive the acute hospitalization is reported as 30% to 60%. Long-term survival rates after acute type B dissection have been reported at 56% to 92% at 1 year and 48% to 82% at 5 years.²³ Survival rates depend on many factors, including the underlying condition, the age of the patient, and comorbidities.

It is important to treat hypertension after aortic dissection, with a goal blood pressure of 120/80 mm Hg or less for most patients. Older studies found higher mortality rates with poorly controlled hypertension. Beta-blockers are the drugs of first choice. Even in the absence of hypertension, long-term beta-blocker therapy should be used to lessen the aortic stress and force of ventricular contraction.

Genetically triggered causes of aortic dissection should be considered. In many circumstances, referral to a medical geneticist or other practitioner knowledgeable in these conditions is important when these disorders are being evaluated (Table 2).

Many of these disorders have an autosomal dominant inheritance, and the patient should be asked about a family history of aortic disease, aortic dissection, or unexplained sudden death. Features of Marfan syndrome, Loeys-Dietz syndrome, and familial thoracic aortic aneurysm syndromes should be sought. Through comprehensive family studies, it is now recognized that up to 20% of patients with thoracic aortic disease (such as aneurysm or dissection) have another first-degree relative with thoracic aortic disease.^2,3,24 Thus, first-degree relatives of patients with aortic aneurysm or dissection should be screened for thoracic aortic aneurysm disease.

Research into molecular genetics is providing a better understanding of the genetics of aortic dissection.³ New mutations associated with aortic dissection are being discovered in signaling pathways as well as elements critical for the integrity of the vascular wall.^2,3 However, at present, most patients with aortic dissection will not have a specific identifiable genetic defect.

Not only does genetic testing enable the accurate diagnosis of the affected individual, but also treatments are often based on this diagnosis.³ Importantly, the identification of a specific gene mutation (ie, in TGFBR1 or 2, FBN1, ACTA2, MYH11, and COL3A1) in an affected individual has the potential to identify other family members at risk.³

Follow-up imaging

It is important to continue to image the aorta after aortic dissection. Patients may develop progressive dilation or aneurysm formation of the dissected aorta, pseudoaneurysm formation after repair, or recurrent dissection. Many patients require additional surgery on the aorta because of late aneurysm formation.

CT or MRI is usually performed at least every 6 months in the first 2 years after dissection and at least annually thereafter. More centers are choosing MRI for long-term follow-up to avoid the repeated radiation exposure with serial CT.

Patient education

Besides receiving medical therapy and undergoing imaging, patients with aortic dissection should be educated about this condition.^5,21 The patient should be aware of symptoms suggesting dissection and should be instructed to seek attention for any concerning symptoms.

Lifestyle modifications are also important. The patient should be educated about safe activity levels and to avoid heavy isometric exercise, such as weight-lifting. Some patients will have to cease their current occupation because of activity restrictions.

A 50-year-old man developed severe chest pain and collapsed to the floor. The pain was sudden in onset, was burning in quality, and was located in the center of his chest. Emergency medical services arrived a few minutes later and found the patient diaphoretic and cyanotic, with an initial blood pressure of 74/54 mm Hg and a heart rate of 125 beats per minute. He was rushed to the hospital.

His medical history was unremarkable. He smoked one pack of cigarettes per day for 20 years. His father died of a “heart attack” at age 52.

In the emergency department he underwent echocardiography with a portable handheld unit, which showed a pericardial effusion and cardiac tamponade. He was sent for emergency computed tomography of the chest, which revealed an aneurysm of the aortic root and acute type A (Stanford classification) aortic dissection with hemopericardium.

He underwent emergency cardiac surgery. At the time of surgery, he was in cardiogenic shock from aortic dissection complicated by severe aortic regurgitation and cardiac tamponade with hemopericardium. The aortic valve was trileaflet. A 27-mm St. Jude composite valve graft root replacement was performed.

The patient did well and was discharged home 7 days after surgery. Pathologic study of the aorta revealed cystic medial degeneration. He did not have any features of Marfan syndrome or Loeys-Dietz syndrome. His three children underwent evaluation, and each had a normal physical examination and echocardiographic test results.

A HIGH INDEX OF SUSPICION IS CRITICAL

Acute aortic dissection is the most common aortic catastrophe, with an incidence estimated at 5 to 30 per 1 million people per year, amounting to nearly 10,000 cases per year in the United States.^1–4

The diagnosis of acute aortic dissection has many potential pitfalls.^2,3 Aortic dissection may mimic other more common conditions, such as coronary ischemia, pleurisy, heart failure, stroke, and acute abdominal illness. Because acute aortic dissection may be rapidly fatal, one must maintain a high index of suspicion.^2,3 Prompt diagnosis and emergency treatment are critical.

WHAT CAUSES AORTIC DISSECTION?

One hypothesis is that acute aortic dissection is caused by a primary tear in the aortic intima, with blood from the aortic lumen penetrating into the diseased media leading to dissection and creating a true and false lumen.² Another is that rupture of the vasa vasorum leads to hemorrhage in the aortic wall with subsequent intimal disruption, creating the intimal tear and aortic dissection.

Once a dissection starts, pulsatile flow of blood within the aortic wall causes it to extend. The dissection flap may be localized, but it often spirals the entire length of the aorta. Distention of the false lumen with blood may cause the intimal flap to compress the true lumen and potentially lead to malperfusion syndromes.

CLASSIFIED ACCORDING TO LOCATION

It is important to recognize the location of the dissection, as the prognosis and treatment depend on whether the ascending aorta is involved.^2,3 For classification purposes, the ascending aorta is the portion proximal to the brachiocephalic artery, while the descending aorta is the portion distal to the left subclavian artery.³

The DeBakey classification defines a type I aortic dissection as one that begins in the ascending aorta and extends at least to the aortic arch or beyond. Type II dissections involve the ascending aorta only, while type III dissections begin in the descending aorta, most often just distal to the left subclavian artery.

The Stanford classification scheme divides dissections into type A and type B. Type A dissections involve the ascending aorta, while type B dissections do not involve the ascending aorta.

Which classification scheme is used is not important. However, identifying patients with dissection of the ascending aorta (DeBakey type I or type II or Stanford type A) is critical, as emergency cardiac surgery is recommended for this type of dissection.^2,3 For the purposes of this paper, the Stanford classification scheme will be used.

Dissection that involves the ascending aorta most commonly occurs in people ages 50 to 60, whereas acute dissection of the descending aorta typically occurs in people 10 years older.^1,2

An acute aortic dissection is one that has occurred within 2 weeks of symptom onset. A chronic dissection is one that occurred more than 2 weeks after symptoms began.

DISEASES AND CONDITIONS ASSOCIATED WITH AORTIC DISSECTION

Hypertension and disorders leading to disruption of the normal structure and function of the aortic wall. About 75% of patients with acute aortic dissection have underlying hypertension.^1–3

Cystic medial degeneration is a common pathologic feature in many cases of aortic dissection.

Cocaine use and intense weight-lifting increase the shear stresses on the aorta.^2,3

Inflammatory aortic diseases such as giant cell arteritis.

Pregnancy can be complicated by aortic dissection, usually in the setting of an underlying aortopathy.⁵

Iatrogenic aortic dissection accounts for about 4% of cases, as a result of cardiac surgery, catheterization, stenting, or use of an intra-aortic balloon pump.¹

Aortic aneurysm. Patients with thoracic aortic aneurysm are at higher risk of aortic dissection, and the larger the aortic diameter, the higher the risk.^2,3,6 In the International Registry of Acute Aortic Dissection (IRAD), the average size of the aorta was about 5.3 cm at the time of acute dissection. Importantly, about 40% of acute dissections of the ascending aorta occur in patients with ascending aortic diameters less than 5.0 cm.^7,8

Thus, many factors are associated with acute dissection, and specific reasons leading to an individual’s susceptibility to sudden dissection are poorly understood.

CLINICAL FEATURES OF ACUTE AORTIC DISSECTION

Because the symptoms of acute dissection may mimic other, more common conditions, one of the most important factors in the diagnosis of aortic dissection is a high clinical suspicion.^1–3

What is the pretest risk of disease?

Recently, the American College of Cardiology (ACC) and the American Heart Association (AHA) released joint guidelines on thoracic aortic disease.³ These guidelines provide an approach to patients who have complaints that may represent acute thoracic aortic dissection, the intent being to establish a pretest risk of disease to be used to guide decision-making.³

The focused evaluation includes specific questions about underlying conditions, symptoms, and findings on examination that may greatly increase the likelihood of acute dissection. These include:

• High-risk conditions and historical features associated with aortic dissection, such as Marfan syndrome and other genetic disorders (Table 2), bicuspid aortic valve, family history of thoracic aortic aneurysm or dissection, known thoracic aortic aneurysm, and recent aortic manipulation
• Pain in the chest, back, or abdomen with high-risk features (eg, abrupt onset, severe intensity, or a ripping or tearing quality)
• High-risk findings on examination (eg, pulse deficits, new aortic regurgitation, hypotension, shock, or systolic blood pressure differences).

Using this information, expedited aortic imaging and treatment algorithms have been devised to improve the diagnosis.³

Using the IRAD database of more than 2,500 acute dissections, the diagnostic algorithm proposed in the ACC/AHA guidelines was shown to be highly sensitive (about 95%) for detecting acute aortic dissection.⁴ In addition, using this score may expedite evaluation by classifying certain patients as being at high risk of acute dissection.^3,4

Important to recognize is that almost two-thirds of patients who suffered dissection in this large database did not have one of the “high-risk conditions” associated with dissection.⁴ Additionally, the specificity of the ACC/AHA algorithm is unknown, and further testing is necessary.⁴

Acute onset of severe pain

More than 90% of acute dissections present with acute pain in the chest or the back, or both.^1–3 The pain is usually severe, of sudden onset, and often described as sharp or, occasionally, tearing, ripping, or stabbing. The pain usually differs from that of coronary ischemia, being most severe at its onset as opposed to the less intense, crescendo-like pain of angina or myocardial infarction. The pain may migrate as the dissection progresses along the length of the aorta or to branch vessels. It may abate, leading to a false sense of security in the patient and the physician.³ “Painless” dissection occurs in a minority, usually in those with syncope, neurologic symptoms, or heart failure.^1–3

The patient with acute dissection may be anxious and may feel a sense of doom.

Acute heart failure, related to severe aortic regurgitation, may be a predominant symptom in dissection of the ascending aorta.

Syncope may occur as a result of aortic rupture, hemopericardium with cardiac tamponade, or acute neurologic complications.

Vascular insufficiency may occur in any branch vessel, leading to clinical syndromes that include acute myocardial infarction, stroke, paraplegia, paraparesis, mesenteric ischemia, and limb ischemia.

PHYSICAL FINDINGS CAN VARY WIDELY

Findings on physical examination in acute aortic dissection vary widely depending on underlying conditions and on the specific complications of the dissection.

Although the classic presentation is acute, severe pain in the chest or back in a severely hypertensive patient with aortic regurgitation and pulse deficits, most patients do not have all these characteristics.⁴ Most patients with type B dissection are hypertensive on presentation, but many with type A dissection present with normal blood pressure or hypotension.¹ Pulse deficits (unequal or absent pulses) are reported in 10% to 30% of acute dissections and may be intermittent as the dynamic movement of the dissection flap interferes with branch vessel perfusion.^1–3

• Aortic leaflet prolapse or distortion of the leaflet alignment
• Malcoaptation of the aortic leaflets from dilation of the aortic root and annulus
• Prolapse of the intimal flap across the aortic valve, interfering with valve function
• Preexisting aortic regurgitation from underlying aortic root aneurysm or primary aortic valve disease (such as a bicuspid aortic valve).

Neurologic manifestations are most common in dissection of the ascending aorta and are particularly important to recognize, as they may dominate the clinical presentation and lead to delay in the diagnosis of dissection.^2,3 Neurologic syndromes include:

• Persistent or transient ischemic stroke
• Spinal cord ischemia
• Ischemic neuropathy
• Hypoxic encephalopathy.

These manifestations are related to malperfusion to branches supplying the brain, spinal cord, or peripheral nerves.⁹

Syncope is relatively common in aortic dissection and may be related to acute hypotension caused by cardiac tamponade or aortic rupture, cerebral vessel obstruction, or activation of cerebral baroreceptors.^2,9 It is important to consider aortic dissection in the differential diagnosis in cases of unexplained syncope.³

Aortic dissections may extend into the abdominal aorta, leading to vascular complications involving one or more branch vessels.¹⁰ The renal artery is involved in at least 5% to 10% of cases and may lead to renal ischemia, infarction, renal insufficiency, or refractory hypertension.²Mesenteric ischemia or infarction occurs in about 5% of dissections, may be difficult to diagnose, and is particularly dangerous.^2,8 Aortic dissection may extend into the iliac arteries and may cause acute lower extremity ischemia.

Acute myocardial infarction due to involvement of the dissection flap causing malperfusion of a coronary artery occurs in 1% to 7% of acute type A aortic dissections.^1–3 The right coronary artery (Figure 2) is most commonly involved, leading to acute inferior myocardial infarction. Acute myocardial ischemia and infarction in the setting of dissection may lead to a delay in the diagnosis of dissection and to bleeding complications from antiplatelet and anticoagulant drugs given to treat the acute coronary syndrome.

Cardiac tamponade, occurring in about 10% of acute type A dissections, portends a higher risk of death.^2,3

Additional clinical features of aortic dissection include a left-sided pleural effusion, usually related to an inflammatory response. An acute hemothorax may occur from rupture or leaking of a descending aortic dissection.

FINDINGS ON RADIOGRAPHY AND ELECTROCARDIOGRAPHY

Reproduced with permission from: Braverman AC, et al. Diseases of the aorta. In: Bonow RO, et al. Braunwald's Heart Disease, 9th edition. Elsevier: Philadelphia, PA; 2011.

Electrocardiography usually has normal or nonspecific findings, unless acute myocardial infarction complicates the dissection.

D-DIMER LEVELS

Biomarkers for the diagnosis of acute aortic dissection are of great interest.

D-dimer levels rise in acute aortic dissection as they do in pulmonary embolism.¹¹ A D-dimer level greater than 1,600 ng/mL within the first 6 hours has a very high positive likelihood ratio for dissection, so this test may be useful in identifying patients with a high probability for dissection. In the first 24 hours after symptom onset, a D-dimer level of less than 500 ng/mL has a negative predictive value of 95%. Thus, elevations in D-dimer may help decide which imaging to perform in a patient presenting with chest pain or suspicion of dissection.¹¹

However, D-dimer levels may not be elevated in dissection variants, such as aortic intramural hematoma or penetrating aortic ulcer. Additionally, once 24 hours have elapsed since the dissection started, D-dimer levels may no longer be elevated. The current ACC/AHA guidelines on thoracic aortic disease concluded that the D-dimer level cannot be used to rule out aortic dissection in high-risk individuals.³

Additional studies may clarify the appropriate role of the D-dimer assay in diagnosing aortic dissection.

DEFINITIVE IMAGING STUDIES: CT, MRI, TEE

Contrast-enhanced computed tomography (CT), magnetic resonance imaging (MRI), and transesophageal echocardiography (TEE) all have very high sensitivity and specificity for the diagnosis of aortic dissection.^2,3 The choice of imaging study often depends on the availability of these studies, with CT and TEE being the most commonly performed initial studies.

MRI is outstanding for detecting and following aortic dissection, but it is usually not the initial study performed because of the time required for image acquisition and because it is generally not available on an emergency basis.

Reproduced with permission from: Braverman AC, et al. Diseases of the aorta. In: Bonow RO, et al. Braunwald's Heart Disease, 9th edition. Elsevier: Philadelphia, PA; 2011.

While transthoracic echocardiography (TTE) can detect aortic dissection, its sensitivity is much lower than that of other imaging tests.^2,3 Therefore, negative findings on TTE do not exclude aortic dissection.

MANAGEMENT OF AORTIC DISSECTION

When acute aortic dissection is diagnosed, multidisciplinary evaluation and treatment are necessary. Time is of the essence, as the death rate in acute dissection may be as high as 1% per hour during the first 24 hours.^1–3 All patients with acute aortic dissection, whether type A or type B, should be transferred to a tertiary care center with a staff experienced in managing aortic dissection and its complications.³ Emergency surgery is recommended for type A aortic dissection, whereas type B dissection is generally treated medically unless complications occur.^2,3

The cornerstone of drug therapy is the prompt reduction in blood pressure with a beta-blocker to reduce shear stresses on the aorta. Intravenous agents such as esmolol (Brevibloc) or labetalol (Normodyne) are usually chosen. Sodium nitroprusside may be added to beta-blocker therapy for rapid blood pressure control in appropriate patients. The patient may require multiple antihypertensive medications. If hypertension is refractory, one must consider renal artery hypertension due to the dissection causing renal malperfusion.² Acute pain may also worsen hypertension, and appropriate analgesia should be used.

Definitive therapy in acute dissection

Patients with acute type A dissection require emergency surgery,^2,3 as they are at risk for life-threatening complications including cardiac tamponade from hemopericardium, aortic rupture, stroke, visceral ischemia, and heart failure due to severe aortic regurgitation. When aortic regurgitation complicates acute type A dissection, some patients are adequately treated by resuspension of the aortic valve leaflets, while others require valve-sparing root replacement or prosthetic aortic valve replacement.

Surgical therapy is associated with a survival benefit compared with medical therapy in acute type A dissection.¹ The 14-day mortality rate for acute type A dissection treated surgically is about 25%.¹ Patients with high-risk features such as heart failure, shock, tamponade, and mesenteric ischemia have a worse prognosis compared with those without these features.^2,12,13

Acute type B aortic dissection carries a lower rate of death than type A dissection.^1–3 In the IRAD cohort, the early mortality rate in those with type B dissection treated medically was about 10%.¹ However, when complications such as malperfusion, shock, or requirement for surgery occur in type B dissection, the mortality rate is much higher,^2,14 with rates of 25% to 50% reported.²

Thus, initial medical therapy is the preferred approach to acute type B dissection, and surgery or endovascular therapy is reserved for patients with acute complications.^2,3 Typical indications for surgery or endovascular therapy in type B dissection include visceral or limb ischemia, aortic rupture, refractory pain, and aneurysmal dilation (Table 3).²

Endovascular therapy in aortic dissection

The high mortality rate with open surgery in acute type B dissection has spurred tremendous interest in endovascular treatments for complications involving the descending aorta and branch vessels.²

Fenestration of the aorta and stenting of branch vessels were the earliest techniques used in complicated type B dissection. By fenestrating (ie, opening) the intimal flap, blood can flow from the false lumen into the true lumen, decompressing the distended false lumen.

Endovascular stenting is used for acute aortic rupture, for malperfusion syndromes, and for rapidly enlarging false lumens. Endovascular grafts may cover the area of a primary intimal tear and thus eliminate the flow into the false channel and promote false-lumen thrombosis. Many patients with complicated type B dissection are treated with a hybrid approach, in which one segment of the aorta, such as the aortic arch, is treated surgically, while the descending aorta receives an endovascular graft.²

Patients with a type B dissection treated medically are at risk for late complications, including aneurysmal enlargement and subsequent aortic rupture. The Investigation of Stent Grafts in Aortic Dissection (INSTEAD) trial included 140 patients with uncomplicated type B dissection and compared drug therapy with endovascular stent grafting.¹⁵ After 2 years of follow-up, there was no difference in the rate of death between the two treatment groups. Patients receiving endovascular grafts had a higher rate of false-lumen thrombosis.

More studies are under way to examine the role of endovascular therapy in uncomplicated type B dissection.

AORTIC DISSECTION VARIANTS

Aortic intramural hematoma

Aortic intramural hematoma is a form of acute aortic syndrome in which a hematoma develops in the aortic media and no intimal flap is visualized either by imaging or at surgery.^2,3,16 It is important to recognize this clinical entity in a patient presenting with acute chest or back pain, as sometimes it is mistaken for a “thrombus in a nonaneurysmal aorta.” Intramural hematoma accounts for 5% to 25% of acute aortic syndromes, depending on the study population (it is more common in Asian studies).^2,3,17 It may present with symptoms similar to classic aortic dissection and is classified as type A or type B, depending on whether the ascending aorta is involved.

Reproduced with permission from: Braverman AC, et al. Diseases of the aorta. In: Bonow RO, et al. Braunwald's Heart Disease, 9th edition. Elsevier: Philadelphia, PA; 2011.

Patients with an intramural hematoma may progress to having complications such as hemopericardium, classic aortic dissection, aortic rupture, or aneurysmal dilation.^2,3 However, many cases of type B aortic intramural hematoma result in complete resorption of the hematoma over time. In general, like classic aortic dissection, type A intramural hematoma is treated with emergency surgery and type B with initial medical therapy.^2,3

There are reports from Southeast Asia of successful initial medical therapy for type A intramural hematoma, with surgery used for acute complications.¹⁸ In the Western literature, improved outcomes are reported with initial surgical therapy.¹⁷ Given the unpredictable nature of type A intramural hematoma, most experts recommend surgical therapy for appropriate candidates with acute type A intramural hematoma.^2,3,19

Penetrating atherosclerotic ulcer of the aorta

Penetrating atherosclerotic ulcer of the aorta, another acute aortic syndrome, results from acute penetration of an atherosclerotic aortic lesion through the internal elastic lamina into the media.^2,3,20 It is often associated with bleeding into the media, or intramural hematoma. While the ulcer may be found incidentally on imaging studies, especially in patients with severe aortic atherosclerosis, the typical presentation is acute, severe chest or back pain. It occurs most often in the descending aorta and the abdominal aorta.

Penetrating atherosclerotic ulcer may lead to pseudoaneurysm formation, focal aortic dissection, aortic rupture, or late aortic aneurysm.²

Reproduced with permission from: Braverman AC, et al. Diseases of the aorta. In: Bonow RO, et al. Braunwald's Heart Disease, 9th edition. Elsevier: Philadelphia, PA; 2011.

LONG-TERM MANAGEMENT AFTER AORTIC DISSECTION

After hospital discharge, patients with aortic dissection require lifelong management. This includes blood pressure control, lifestyle modification, serial imaging of the aorta with CT or MRI, patient education about the condition, and, when appropriate, screening of family members for aortic disease.^5,21

Reported survival rates after hospitalization for type A dissection are 52% to 94% at 1 year and 45% to 88% at 5 years.^2,22 The 10-year actuarial survival rate for those with acute dissection who survive the acute hospitalization is reported as 30% to 60%. Long-term survival rates after acute type B dissection have been reported at 56% to 92% at 1 year and 48% to 82% at 5 years.²³ Survival rates depend on many factors, including the underlying condition, the age of the patient, and comorbidities.

It is important to treat hypertension after aortic dissection, with a goal blood pressure of 120/80 mm Hg or less for most patients. Older studies found higher mortality rates with poorly controlled hypertension. Beta-blockers are the drugs of first choice. Even in the absence of hypertension, long-term beta-blocker therapy should be used to lessen the aortic stress and force of ventricular contraction.

Genetically triggered causes of aortic dissection should be considered. In many circumstances, referral to a medical geneticist or other practitioner knowledgeable in these conditions is important when these disorders are being evaluated (Table 2).

Many of these disorders have an autosomal dominant inheritance, and the patient should be asked about a family history of aortic disease, aortic dissection, or unexplained sudden death. Features of Marfan syndrome, Loeys-Dietz syndrome, and familial thoracic aortic aneurysm syndromes should be sought. Through comprehensive family studies, it is now recognized that up to 20% of patients with thoracic aortic disease (such as aneurysm or dissection) have another first-degree relative with thoracic aortic disease.^2,3,24 Thus, first-degree relatives of patients with aortic aneurysm or dissection should be screened for thoracic aortic aneurysm disease.

Research into molecular genetics is providing a better understanding of the genetics of aortic dissection.³ New mutations associated with aortic dissection are being discovered in signaling pathways as well as elements critical for the integrity of the vascular wall.^2,3 However, at present, most patients with aortic dissection will not have a specific identifiable genetic defect.

Not only does genetic testing enable the accurate diagnosis of the affected individual, but also treatments are often based on this diagnosis.³ Importantly, the identification of a specific gene mutation (ie, in TGFBR1 or 2, FBN1, ACTA2, MYH11, and COL3A1) in an affected individual has the potential to identify other family members at risk.³

Follow-up imaging

It is important to continue to image the aorta after aortic dissection. Patients may develop progressive dilation or aneurysm formation of the dissected aorta, pseudoaneurysm formation after repair, or recurrent dissection. Many patients require additional surgery on the aorta because of late aneurysm formation.

CT or MRI is usually performed at least every 6 months in the first 2 years after dissection and at least annually thereafter. More centers are choosing MRI for long-term follow-up to avoid the repeated radiation exposure with serial CT.

Patient education

Besides receiving medical therapy and undergoing imaging, patients with aortic dissection should be educated about this condition.^5,21 The patient should be aware of symptoms suggesting dissection and should be instructed to seek attention for any concerning symptoms.

Lifestyle modifications are also important. The patient should be educated about safe activity levels and to avoid heavy isometric exercise, such as weight-lifting. Some patients will have to cease their current occupation because of activity restrictions.

A 50-year-old man developed severe chest pain and collapsed to the floor. The pain was sudden in onset, was burning in quality, and was located in the center of his chest. Emergency medical services arrived a few minutes later and found the patient diaphoretic and cyanotic, with an initial blood pressure of 74/54 mm Hg and a heart rate of 125 beats per minute. He was rushed to the hospital.

His medical history was unremarkable. He smoked one pack of cigarettes per day for 20 years. His father died of a “heart attack” at age 52.

In the emergency department he underwent echocardiography with a portable handheld unit, which showed a pericardial effusion and cardiac tamponade. He was sent for emergency computed tomography of the chest, which revealed an aneurysm of the aortic root and acute type A (Stanford classification) aortic dissection with hemopericardium.

He underwent emergency cardiac surgery. At the time of surgery, he was in cardiogenic shock from aortic dissection complicated by severe aortic regurgitation and cardiac tamponade with hemopericardium. The aortic valve was trileaflet. A 27-mm St. Jude composite valve graft root replacement was performed.

The patient did well and was discharged home 7 days after surgery. Pathologic study of the aorta revealed cystic medial degeneration. He did not have any features of Marfan syndrome or Loeys-Dietz syndrome. His three children underwent evaluation, and each had a normal physical examination and echocardiographic test results.

A HIGH INDEX OF SUSPICION IS CRITICAL

Acute aortic dissection is the most common aortic catastrophe, with an incidence estimated at 5 to 30 per 1 million people per year, amounting to nearly 10,000 cases per year in the United States.^1–4

The diagnosis of acute aortic dissection has many potential pitfalls.^2,3 Aortic dissection may mimic other more common conditions, such as coronary ischemia, pleurisy, heart failure, stroke, and acute abdominal illness. Because acute aortic dissection may be rapidly fatal, one must maintain a high index of suspicion.^2,3 Prompt diagnosis and emergency treatment are critical.

WHAT CAUSES AORTIC DISSECTION?

One hypothesis is that acute aortic dissection is caused by a primary tear in the aortic intima, with blood from the aortic lumen penetrating into the diseased media leading to dissection and creating a true and false lumen.² Another is that rupture of the vasa vasorum leads to hemorrhage in the aortic wall with subsequent intimal disruption, creating the intimal tear and aortic dissection.

Once a dissection starts, pulsatile flow of blood within the aortic wall causes it to extend. The dissection flap may be localized, but it often spirals the entire length of the aorta. Distention of the false lumen with blood may cause the intimal flap to compress the true lumen and potentially lead to malperfusion syndromes.

CLASSIFIED ACCORDING TO LOCATION

It is important to recognize the location of the dissection, as the prognosis and treatment depend on whether the ascending aorta is involved.^2,3 For classification purposes, the ascending aorta is the portion proximal to the brachiocephalic artery, while the descending aorta is the portion distal to the left subclavian artery.³

The DeBakey classification defines a type I aortic dissection as one that begins in the ascending aorta and extends at least to the aortic arch or beyond. Type II dissections involve the ascending aorta only, while type III dissections begin in the descending aorta, most often just distal to the left subclavian artery.

The Stanford classification scheme divides dissections into type A and type B. Type A dissections involve the ascending aorta, while type B dissections do not involve the ascending aorta.

Which classification scheme is used is not important. However, identifying patients with dissection of the ascending aorta (DeBakey type I or type II or Stanford type A) is critical, as emergency cardiac surgery is recommended for this type of dissection.^2,3 For the purposes of this paper, the Stanford classification scheme will be used.

Dissection that involves the ascending aorta most commonly occurs in people ages 50 to 60, whereas acute dissection of the descending aorta typically occurs in people 10 years older.^1,2

An acute aortic dissection is one that has occurred within 2 weeks of symptom onset. A chronic dissection is one that occurred more than 2 weeks after symptoms began.

DISEASES AND CONDITIONS ASSOCIATED WITH AORTIC DISSECTION

Hypertension and disorders leading to disruption of the normal structure and function of the aortic wall. About 75% of patients with acute aortic dissection have underlying hypertension.^1–3

Cystic medial degeneration is a common pathologic feature in many cases of aortic dissection.

Cocaine use and intense weight-lifting increase the shear stresses on the aorta.^2,3

Inflammatory aortic diseases such as giant cell arteritis.

Pregnancy can be complicated by aortic dissection, usually in the setting of an underlying aortopathy.⁵

Iatrogenic aortic dissection accounts for about 4% of cases, as a result of cardiac surgery, catheterization, stenting, or use of an intra-aortic balloon pump.¹

Aortic aneurysm. Patients with thoracic aortic aneurysm are at higher risk of aortic dissection, and the larger the aortic diameter, the higher the risk.^2,3,6 In the International Registry of Acute Aortic Dissection (IRAD), the average size of the aorta was about 5.3 cm at the time of acute dissection. Importantly, about 40% of acute dissections of the ascending aorta occur in patients with ascending aortic diameters less than 5.0 cm.^7,8

Thus, many factors are associated with acute dissection, and specific reasons leading to an individual’s susceptibility to sudden dissection are poorly understood.

CLINICAL FEATURES OF ACUTE AORTIC DISSECTION

Because the symptoms of acute dissection may mimic other, more common conditions, one of the most important factors in the diagnosis of aortic dissection is a high clinical suspicion.^1–3

What is the pretest risk of disease?

Recently, the American College of Cardiology (ACC) and the American Heart Association (AHA) released joint guidelines on thoracic aortic disease.³ These guidelines provide an approach to patients who have complaints that may represent acute thoracic aortic dissection, the intent being to establish a pretest risk of disease to be used to guide decision-making.³

The focused evaluation includes specific questions about underlying conditions, symptoms, and findings on examination that may greatly increase the likelihood of acute dissection. These include:

• High-risk conditions and historical features associated with aortic dissection, such as Marfan syndrome and other genetic disorders (Table 2), bicuspid aortic valve, family history of thoracic aortic aneurysm or dissection, known thoracic aortic aneurysm, and recent aortic manipulation
• Pain in the chest, back, or abdomen with high-risk features (eg, abrupt onset, severe intensity, or a ripping or tearing quality)
• High-risk findings on examination (eg, pulse deficits, new aortic regurgitation, hypotension, shock, or systolic blood pressure differences).

Using this information, expedited aortic imaging and treatment algorithms have been devised to improve the diagnosis.³

Using the IRAD database of more than 2,500 acute dissections, the diagnostic algorithm proposed in the ACC/AHA guidelines was shown to be highly sensitive (about 95%) for detecting acute aortic dissection.⁴ In addition, using this score may expedite evaluation by classifying certain patients as being at high risk of acute dissection.^3,4

Important to recognize is that almost two-thirds of patients who suffered dissection in this large database did not have one of the “high-risk conditions” associated with dissection.⁴ Additionally, the specificity of the ACC/AHA algorithm is unknown, and further testing is necessary.⁴

Acute onset of severe pain

More than 90% of acute dissections present with acute pain in the chest or the back, or both.^1–3 The pain is usually severe, of sudden onset, and often described as sharp or, occasionally, tearing, ripping, or stabbing. The pain usually differs from that of coronary ischemia, being most severe at its onset as opposed to the less intense, crescendo-like pain of angina or myocardial infarction. The pain may migrate as the dissection progresses along the length of the aorta or to branch vessels. It may abate, leading to a false sense of security in the patient and the physician.³ “Painless” dissection occurs in a minority, usually in those with syncope, neurologic symptoms, or heart failure.^1–3

The patient with acute dissection may be anxious and may feel a sense of doom.

Acute heart failure, related to severe aortic regurgitation, may be a predominant symptom in dissection of the ascending aorta.

Syncope may occur as a result of aortic rupture, hemopericardium with cardiac tamponade, or acute neurologic complications.

Vascular insufficiency may occur in any branch vessel, leading to clinical syndromes that include acute myocardial infarction, stroke, paraplegia, paraparesis, mesenteric ischemia, and limb ischemia.

PHYSICAL FINDINGS CAN VARY WIDELY

Findings on physical examination in acute aortic dissection vary widely depending on underlying conditions and on the specific complications of the dissection.

Although the classic presentation is acute, severe pain in the chest or back in a severely hypertensive patient with aortic regurgitation and pulse deficits, most patients do not have all these characteristics.⁴ Most patients with type B dissection are hypertensive on presentation, but many with type A dissection present with normal blood pressure or hypotension.¹ Pulse deficits (unequal or absent pulses) are reported in 10% to 30% of acute dissections and may be intermittent as the dynamic movement of the dissection flap interferes with branch vessel perfusion.^1–3

• Aortic leaflet prolapse or distortion of the leaflet alignment
• Malcoaptation of the aortic leaflets from dilation of the aortic root and annulus
• Prolapse of the intimal flap across the aortic valve, interfering with valve function
• Preexisting aortic regurgitation from underlying aortic root aneurysm or primary aortic valve disease (such as a bicuspid aortic valve).

Neurologic manifestations are most common in dissection of the ascending aorta and are particularly important to recognize, as they may dominate the clinical presentation and lead to delay in the diagnosis of dissection.^2,3 Neurologic syndromes include:

• Persistent or transient ischemic stroke
• Spinal cord ischemia
• Ischemic neuropathy
• Hypoxic encephalopathy.

These manifestations are related to malperfusion to branches supplying the brain, spinal cord, or peripheral nerves.⁹

Syncope is relatively common in aortic dissection and may be related to acute hypotension caused by cardiac tamponade or aortic rupture, cerebral vessel obstruction, or activation of cerebral baroreceptors.^2,9 It is important to consider aortic dissection in the differential diagnosis in cases of unexplained syncope.³

Aortic dissections may extend into the abdominal aorta, leading to vascular complications involving one or more branch vessels.¹⁰ The renal artery is involved in at least 5% to 10% of cases and may lead to renal ischemia, infarction, renal insufficiency, or refractory hypertension.²Mesenteric ischemia or infarction occurs in about 5% of dissections, may be difficult to diagnose, and is particularly dangerous.^2,8 Aortic dissection may extend into the iliac arteries and may cause acute lower extremity ischemia.

Acute myocardial infarction due to involvement of the dissection flap causing malperfusion of a coronary artery occurs in 1% to 7% of acute type A aortic dissections.^1–3 The right coronary artery (Figure 2) is most commonly involved, leading to acute inferior myocardial infarction. Acute myocardial ischemia and infarction in the setting of dissection may lead to a delay in the diagnosis of dissection and to bleeding complications from antiplatelet and anticoagulant drugs given to treat the acute coronary syndrome.

Cardiac tamponade, occurring in about 10% of acute type A dissections, portends a higher risk of death.^2,3

Additional clinical features of aortic dissection include a left-sided pleural effusion, usually related to an inflammatory response. An acute hemothorax may occur from rupture or leaking of a descending aortic dissection.

FINDINGS ON RADIOGRAPHY AND ELECTROCARDIOGRAPHY

Reproduced with permission from: Braverman AC, et al. Diseases of the aorta. In: Bonow RO, et al. Braunwald's Heart Disease, 9th edition. Elsevier: Philadelphia, PA; 2011.

Electrocardiography usually has normal or nonspecific findings, unless acute myocardial infarction complicates the dissection.

D-DIMER LEVELS

Biomarkers for the diagnosis of acute aortic dissection are of great interest.

D-dimer levels rise in acute aortic dissection as they do in pulmonary embolism.¹¹ A D-dimer level greater than 1,600 ng/mL within the first 6 hours has a very high positive likelihood ratio for dissection, so this test may be useful in identifying patients with a high probability for dissection. In the first 24 hours after symptom onset, a D-dimer level of less than 500 ng/mL has a negative predictive value of 95%. Thus, elevations in D-dimer may help decide which imaging to perform in a patient presenting with chest pain or suspicion of dissection.¹¹

However, D-dimer levels may not be elevated in dissection variants, such as aortic intramural hematoma or penetrating aortic ulcer. Additionally, once 24 hours have elapsed since the dissection started, D-dimer levels may no longer be elevated. The current ACC/AHA guidelines on thoracic aortic disease concluded that the D-dimer level cannot be used to rule out aortic dissection in high-risk individuals.³

Additional studies may clarify the appropriate role of the D-dimer assay in diagnosing aortic dissection.

DEFINITIVE IMAGING STUDIES: CT, MRI, TEE

Contrast-enhanced computed tomography (CT), magnetic resonance imaging (MRI), and transesophageal echocardiography (TEE) all have very high sensitivity and specificity for the diagnosis of aortic dissection.^2,3 The choice of imaging study often depends on the availability of these studies, with CT and TEE being the most commonly performed initial studies.

MRI is outstanding for detecting and following aortic dissection, but it is usually not the initial study performed because of the time required for image acquisition and because it is generally not available on an emergency basis.

Reproduced with permission from: Braverman AC, et al. Diseases of the aorta. In: Bonow RO, et al. Braunwald's Heart Disease, 9th edition. Elsevier: Philadelphia, PA; 2011.

While transthoracic echocardiography (TTE) can detect aortic dissection, its sensitivity is much lower than that of other imaging tests.^2,3 Therefore, negative findings on TTE do not exclude aortic dissection.

MANAGEMENT OF AORTIC DISSECTION

When acute aortic dissection is diagnosed, multidisciplinary evaluation and treatment are necessary. Time is of the essence, as the death rate in acute dissection may be as high as 1% per hour during the first 24 hours.^1–3 All patients with acute aortic dissection, whether type A or type B, should be transferred to a tertiary care center with a staff experienced in managing aortic dissection and its complications.³ Emergency surgery is recommended for type A aortic dissection, whereas type B dissection is generally treated medically unless complications occur.^2,3

The cornerstone of drug therapy is the prompt reduction in blood pressure with a beta-blocker to reduce shear stresses on the aorta. Intravenous agents such as esmolol (Brevibloc) or labetalol (Normodyne) are usually chosen. Sodium nitroprusside may be added to beta-blocker therapy for rapid blood pressure control in appropriate patients. The patient may require multiple antihypertensive medications. If hypertension is refractory, one must consider renal artery hypertension due to the dissection causing renal malperfusion.² Acute pain may also worsen hypertension, and appropriate analgesia should be used.

Definitive therapy in acute dissection

Patients with acute type A dissection require emergency surgery,^2,3 as they are at risk for life-threatening complications including cardiac tamponade from hemopericardium, aortic rupture, stroke, visceral ischemia, and heart failure due to severe aortic regurgitation. When aortic regurgitation complicates acute type A dissection, some patients are adequately treated by resuspension of the aortic valve leaflets, while others require valve-sparing root replacement or prosthetic aortic valve replacement.

Surgical therapy is associated with a survival benefit compared with medical therapy in acute type A dissection.¹ The 14-day mortality rate for acute type A dissection treated surgically is about 25%.¹ Patients with high-risk features such as heart failure, shock, tamponade, and mesenteric ischemia have a worse prognosis compared with those without these features.^2,12,13

Acute type B aortic dissection carries a lower rate of death than type A dissection.^1–3 In the IRAD cohort, the early mortality rate in those with type B dissection treated medically was about 10%.¹ However, when complications such as malperfusion, shock, or requirement for surgery occur in type B dissection, the mortality rate is much higher,^2,14 with rates of 25% to 50% reported.²

Thus, initial medical therapy is the preferred approach to acute type B dissection, and surgery or endovascular therapy is reserved for patients with acute complications.^2,3 Typical indications for surgery or endovascular therapy in type B dissection include visceral or limb ischemia, aortic rupture, refractory pain, and aneurysmal dilation (Table 3).²

Endovascular therapy in aortic dissection

The high mortality rate with open surgery in acute type B dissection has spurred tremendous interest in endovascular treatments for complications involving the descending aorta and branch vessels.²

Fenestration of the aorta and stenting of branch vessels were the earliest techniques used in complicated type B dissection. By fenestrating (ie, opening) the intimal flap, blood can flow from the false lumen into the true lumen, decompressing the distended false lumen.

Endovascular stenting is used for acute aortic rupture, for malperfusion syndromes, and for rapidly enlarging false lumens. Endovascular grafts may cover the area of a primary intimal tear and thus eliminate the flow into the false channel and promote false-lumen thrombosis. Many patients with complicated type B dissection are treated with a hybrid approach, in which one segment of the aorta, such as the aortic arch, is treated surgically, while the descending aorta receives an endovascular graft.²

Patients with a type B dissection treated medically are at risk for late complications, including aneurysmal enlargement and subsequent aortic rupture. The Investigation of Stent Grafts in Aortic Dissection (INSTEAD) trial included 140 patients with uncomplicated type B dissection and compared drug therapy with endovascular stent grafting.¹⁵ After 2 years of follow-up, there was no difference in the rate of death between the two treatment groups. Patients receiving endovascular grafts had a higher rate of false-lumen thrombosis.

More studies are under way to examine the role of endovascular therapy in uncomplicated type B dissection.

AORTIC DISSECTION VARIANTS

Aortic intramural hematoma

Aortic intramural hematoma is a form of acute aortic syndrome in which a hematoma develops in the aortic media and no intimal flap is visualized either by imaging or at surgery.^2,3,16 It is important to recognize this clinical entity in a patient presenting with acute chest or back pain, as sometimes it is mistaken for a “thrombus in a nonaneurysmal aorta.” Intramural hematoma accounts for 5% to 25% of acute aortic syndromes, depending on the study population (it is more common in Asian studies).^2,3,17 It may present with symptoms similar to classic aortic dissection and is classified as type A or type B, depending on whether the ascending aorta is involved.

Reproduced with permission from: Braverman AC, et al. Diseases of the aorta. In: Bonow RO, et al. Braunwald's Heart Disease, 9th edition. Elsevier: Philadelphia, PA; 2011.

Patients with an intramural hematoma may progress to having complications such as hemopericardium, classic aortic dissection, aortic rupture, or aneurysmal dilation.^2,3 However, many cases of type B aortic intramural hematoma result in complete resorption of the hematoma over time. In general, like classic aortic dissection, type A intramural hematoma is treated with emergency surgery and type B with initial medical therapy.^2,3

There are reports from Southeast Asia of successful initial medical therapy for type A intramural hematoma, with surgery used for acute complications.¹⁸ In the Western literature, improved outcomes are reported with initial surgical therapy.¹⁷ Given the unpredictable nature of type A intramural hematoma, most experts recommend surgical therapy for appropriate candidates with acute type A intramural hematoma.^2,3,19

Penetrating atherosclerotic ulcer of the aorta

Penetrating atherosclerotic ulcer of the aorta, another acute aortic syndrome, results from acute penetration of an atherosclerotic aortic lesion through the internal elastic lamina into the media.^2,3,20 It is often associated with bleeding into the media, or intramural hematoma. While the ulcer may be found incidentally on imaging studies, especially in patients with severe aortic atherosclerosis, the typical presentation is acute, severe chest or back pain. It occurs most often in the descending aorta and the abdominal aorta.

Penetrating atherosclerotic ulcer may lead to pseudoaneurysm formation, focal aortic dissection, aortic rupture, or late aortic aneurysm.²

Reproduced with permission from: Braverman AC, et al. Diseases of the aorta. In: Bonow RO, et al. Braunwald's Heart Disease, 9th edition. Elsevier: Philadelphia, PA; 2011.

LONG-TERM MANAGEMENT AFTER AORTIC DISSECTION

After hospital discharge, patients with aortic dissection require lifelong management. This includes blood pressure control, lifestyle modification, serial imaging of the aorta with CT or MRI, patient education about the condition, and, when appropriate, screening of family members for aortic disease.^5,21

Reported survival rates after hospitalization for type A dissection are 52% to 94% at 1 year and 45% to 88% at 5 years.^2,22 The 10-year actuarial survival rate for those with acute dissection who survive the acute hospitalization is reported as 30% to 60%. Long-term survival rates after acute type B dissection have been reported at 56% to 92% at 1 year and 48% to 82% at 5 years.²³ Survival rates depend on many factors, including the underlying condition, the age of the patient, and comorbidities.

It is important to treat hypertension after aortic dissection, with a goal blood pressure of 120/80 mm Hg or less for most patients. Older studies found higher mortality rates with poorly controlled hypertension. Beta-blockers are the drugs of first choice. Even in the absence of hypertension, long-term beta-blocker therapy should be used to lessen the aortic stress and force of ventricular contraction.

Genetically triggered causes of aortic dissection should be considered. In many circumstances, referral to a medical geneticist or other practitioner knowledgeable in these conditions is important when these disorders are being evaluated (Table 2).

Many of these disorders have an autosomal dominant inheritance, and the patient should be asked about a family history of aortic disease, aortic dissection, or unexplained sudden death. Features of Marfan syndrome, Loeys-Dietz syndrome, and familial thoracic aortic aneurysm syndromes should be sought. Through comprehensive family studies, it is now recognized that up to 20% of patients with thoracic aortic disease (such as aneurysm or dissection) have another first-degree relative with thoracic aortic disease.^2,3,24 Thus, first-degree relatives of patients with aortic aneurysm or dissection should be screened for thoracic aortic aneurysm disease.

Research into molecular genetics is providing a better understanding of the genetics of aortic dissection.³ New mutations associated with aortic dissection are being discovered in signaling pathways as well as elements critical for the integrity of the vascular wall.^2,3 However, at present, most patients with aortic dissection will not have a specific identifiable genetic defect.

Not only does genetic testing enable the accurate diagnosis of the affected individual, but also treatments are often based on this diagnosis.³ Importantly, the identification of a specific gene mutation (ie, in TGFBR1 or 2, FBN1, ACTA2, MYH11, and COL3A1) in an affected individual has the potential to identify other family members at risk.³

Follow-up imaging

It is important to continue to image the aorta after aortic dissection. Patients may develop progressive dilation or aneurysm formation of the dissected aorta, pseudoaneurysm formation after repair, or recurrent dissection. Many patients require additional surgery on the aorta because of late aneurysm formation.

CT or MRI is usually performed at least every 6 months in the first 2 years after dissection and at least annually thereafter. More centers are choosing MRI for long-term follow-up to avoid the repeated radiation exposure with serial CT.

Patient education

Besides receiving medical therapy and undergoing imaging, patients with aortic dissection should be educated about this condition.^5,21 The patient should be aware of symptoms suggesting dissection and should be instructed to seek attention for any concerning symptoms.

Lifestyle modifications are also important. The patient should be educated about safe activity levels and to avoid heavy isometric exercise, such as weight-lifting. Some patients will have to cease their current occupation because of activity restrictions.

Deep vein thrombosis and pulmonary embolism are collectively referred to as venous thromboembolic (VTE) disease. They affect approximately 100,000 to 300,000 patients per year in the United States.¹ Although patients with deep vein thrombosis can be treated as outpatients, many are admitted for the initiation of anticoagulation. Initial anticoagulation usually requires the overlap of a parenteral anticoagulant (unfractionated heparin, low-molecular-weight heparin [LMWH] or fondaparinux) with warfarin for a minimum of 5 days and until the international normalized ratio (INR) of the prothrombin time is above 2.0 for at least 24 hours.²

Three clinical issues need to be addressed after the initiation of anticoagulation for VTE:

• Determination of the length of anticoagulation with the correct anticoagulant
• Prevention of postthrombotic syndrome
• Appropriate screening for occult malignancy.

HOW LONG SHOULD VTE BE TREATED?

The duration of anticoagulation has been a matter of debate.

The risk of recurrent VTE appears related to clinical risk factors that a patient has at the time of the initial thrombotic event. An epidemiologic study³ found that patients with VTE treated for approximately 6 months had a low rate of recurrence (0% at 2 years of follow-up) if surgery was the risk factor. The risk climbed to 9% if the risk factor was nonsurgical and to 19% if there were no provoking risk factors.

The likelihood of VTE recurrence and therefore the recommended duration of treatment depend on whether the VTE event was provoked, cancer-related, recurrent, thrombophilia-related, or idiopathic. We address each of these scenarios below.

HOW LONG TO TREAT PROVOKED VTE

A VTE event is considered provoked if the patient had a clear inciting risk factor. As defined in various clinical trials, these risk factors include:

• Hospitalization with confinement to bed for 3 or more consecutive days in the last 3 months
• Surgery or general anesthesia in the last 3 months
• Immobilization for more than 7 days, regardless of the cause
• Trauma in the last 3 months
• Pregnancy
• Use of an oral contraceptive, regardless of which estrogen or progesterone analogue it contains
• Travel for more than 4 hours
• Recent childbirth.

However, the trials that tested different lengths of anticoagulation have varied markedly in how they defined provoked deep vein thrombosis.^4–7

Recommendation: Warfarin or equivalent for 3 months

The American College of Chest Physicians (ACCP) recommends 3 months of anticoagulation with warfarin or another vitamin K antagonist for patients with VTE secondary to a transient (reversible) risk factor,² and we agree.

HOW LONG TO TREAT CANCER-RELATED VTE

Patients with cancer are at higher risk of developing VTE. Furthermore, in one study,⁹ compared with other patients with VTE, patients with cancer were three times more likely to have another episode of VTE, with a cumulative rate of recurrence at 1 year of 21% vs 7%. Cancer patients were also twice as likely to suffer major bleeding complications while on anticoagulation.⁹

Warfarin is a difficult drug to manage because it has many interactions with foods, diseases, and other drugs. These difficulties are amplified in many cancer patients during chemotherapy.

Warfarin was compared with a LMWH in four randomized trials in cancer patients, and a meta-analysis¹⁰ found a 50% relative reduction in the rates of recurrent deep vein thrombosis and pulmonary embolism with LMWH treatment. These results were driven primarily by the CLOT trial (Comparison of Low-Molecular-Weight Heparin Versus Oral Anticoagulant Therapy for the Prevention of Recurrent Venous Thromboembolism in Patients With Cancer),¹¹ which showed an 8% absolute risk reduction (number needed to treat 13) without an increase in major bleeding when cancer-related VTE was treated with an LMWH—ie, dalteparin (Fragmin)—for 6 months compared with warfarin.

Current thinking suggests that VTE should be treated until the cancer is resolved. However, this hypothesis has not been adequately tested, and consequently, the ACCP gives it only a level 1C recommendation.² The largest of the four trials comparing warfarin and an LMWH lasted only 6 months. The safety of extending LMWH treatment beyond 6 months is currently unknown but is under investigation (clinicaltrials.gov identifier NCT00942968).

The ACCP guidelines recommend LMWH therapy for 3 to 6 months, followed by warfarin or another vitamin K antagonist or continued LMWH treatment until the cancer is resolved.²

The National Comprehensive Cancer Network guidelines recommend an LMWH for 6 months as monotherapy and indefinite anticoagulation if the cancer is still active.¹²

The American Society of Clinical Oncology guidelines recommend an LMWH for at least 6 months and indefinite anticoagulant therapy for selected patients with active cancer.¹³

We agree that patients with active cancer should receive an LMWH for at least 6 months and indefinite anticoagulation until the cancer is resolved.

In our experience, many patients are reluctant to give themselves the daily injections that LMWH therapy requires, and so they need to be well-informed about the marked decrease in VTE recurrence with this less-convenient and more-expensive therapy. Many patients face insurance barriers to cover the cost of LMWH therapy; however, careful attention to preauthorization can usually overcome this obstacle.

HOW LONG TO TREAT RECURRENT VTE

It makes clinical sense that patients who have a second VTE event should be treated indefinitely. This theory was tested in a randomized clinical trial¹⁴ in which patients with provoked or unprovoked VTE were randomized after their second event to receive anticoagulation for 6 months vs indefinitely.

After 4 years of follow-up, the recurrence rate was 21% in patients assigned to 6 months of treatment and only 3% in patients who continued anticoagulation throughout the trial. On the other hand, major hemorrhage occurred in 3% of patients treated for 6 months and in 9% in patients who continued anticoagulation indefinitely.

Of note, most of the patients in this trial had unprovoked (idiopathic) VTE, so the results should not be extrapolated to patients with provoked VTE, who accounted for only 20% of the study population.¹⁴

Recommendation: Long-term anticoagulation

We agree with the ACCP recommendation² that patients who have had a second episode of unprovoked VTE should receive long-term anticoagulation. Because of a lack of data, the duration of therapy for patients with a second episode of provoked VTE should be individualized.

HOW LONG TO TREAT THROMBOPHILIA-RELATED VTE

Inherited thrombophilias

Patients with VTE that is not related to a clear provoking risk factor or cancer frequently have testing to evaluate for a hypercoagulable state. This workup traditionally includes the most common inherited thrombophilias for gene mutations for factor V and prothrombin as well as for deficiencies in protein C, protein S, antithrombin and the acquired antiphospholipid syndrome.

The key questions that should be asked prior to embarking on this workup are:

• Will the results change the length of therapy for the patient?
• Will testing the patient help with genetic counseling and possible testing of family members?
• Will the results change the targeted INR range for warfarin or other vitamin K antagonist therapy?

Patients with inherited thrombophilia have a greater risk of developing an initial VTE event; however, these tests do not help predict the recurrence of VTE in patients with established disease more than clinical risk factors do. A prospective study demonstrated this by looking at the effect of thrombophilia and clinical factors on the recurrence of venous thrombosis and found that inherited prothrombotic abnormalities do not appear to play an important role in the risk of a recurrent event.¹⁵ On the other hand, clinical factors, such as whether the first event was idiopathic or provoked, appear more important in determining the duration of anticoagulation therapy.¹⁵ A systematic review of the common inherited thrombophilias showed the VTE recurrence rate of patients with factor V Leiden was higher than in patients without the mutation; however, the absolute rates of recurrence were not much different than what would be expected in patients with idiopathic VTE.¹⁶

A retrospective study involving a large cohort of families of patients who already had experienced a first episode of either idiopathic or provoked VTE showed high annual risks of recurrent VTE associated with hereditary deficiencies of protein S (8.4%), protein C (6.0%), and antithrombin (10%).¹⁷ However, for the more commonly occurring genetic thrombophilias, the factor V Leiden and prothrombin G20210A mutations, family members with either gene abnormality had low rates of VTE, suggesting that testing of relatives of probands is not clinically useful.¹⁶

Antiphospholipid syndrome

Antiphospholipid syndrome is an acquired thrombophilia. A patient has thrombotic antiphospholipid syndrome when there is a history of vascular thrombosis in the presence of persistently positive tests (at least 12 weeks apart) for lupus anticoagulants, anticardiolipin antibodies, or anti-beta-2 glycoprotein I. A prospective study of 412 patients with a first episode of VTE found that 15% were positive for anticardiolipin antibody at the end of 6 months of anticoagulation. The risk of recurrent VTE after 4 years was 29% in patients with antibodies and 14% in those without antibodies (relative risk 2.1; 95% confidence interval [CI] 1.3–3.3; P =.0013).¹⁸

Recent reviews advise indefinite warfarin anticoagulation in patients with VTE and persistence of antiphospholipid antibodies.¹⁹ However, the optimal duration of anticoagulation is uncertain. Until well-designed clinical trials are done, the current general consensus is to anticoagulate these patients indefinitely.^20,21 Retrospective studies had suggested that patients with antiphospholipid antibodies required a higher therapeutic INR range; however, this observation was tested in two trials that found no difference in thromboembolic rates when patients were randomized to an INR of 2.0–3.0 vs 3.1–4.0,²² or 2.0–3.0 vs 3.0–4.5.²³

No formal recommendations

In the absence of strong evidence, the ACCP guidelines do not include a recommendation on the duration of anticoagulation treatment specific to inherited thrombophilias. We believe that clinical factors are more important than inherited thrombophilias for deciding the duration of anticoagulation, and that testing is almost never indicated or useful. However, patients with antiphospholipid syndrome are at high risk of recurrence, and it is our practice to anticoagulate these patients indefinitely.

HOW LONG TO TREAT UNPROVOKED (IDIOPATHIC) VTE

A VTE event is thought to be idiopathic if it occurs without a clearly identified provoking factor.

In an observational study,³ patients with oral contraceptive use, transient illness, immobilization, or a history of travel had an 8.8% risk of recurrence vs 19.4% in patients with unprovoked VTE. The meta-analysis discussed above (Table 1)⁸ also shows that patients with these nonsurgical risk factors have a lower rate of recurrence than patients with idiopathic VTE.

The high rate of recurrence of idiopathic VTE (4% to 27% after 3 months of anticoagulation^24–26) suggests that a longer duration of treatment is reasonable. However, increasing the length of therapy from 3 to 12 months delays but does not prevent recurrence, the risk of which begins to accumulate once anticoagulation is stopped.^24,25

Three promising strategies to identify subgroups of patients with idiopathic VTE who are at highest risk of recurrence and who would benefit the most from prolonged anticoagulation are d-dimer testing, evaluation for residual vein thrombosis in patients who present with a deep vein thrombosis, and clinical prediction rules.

d-dimer testing

d-dimer is a degradation product of fibrin and is an indirect marker of residual thrombosis.³⁰

In a systematic review of patients with a first episode of unprovoked VTE,³¹ a normal d-dimer concentration at the end of at least 3 months of anticoagulation was associated with a 3.5% annual risk of recurrence, whereas an elevated d-dimer level at that time was associated with an annual risk of 8.9%. These results were confirmed in a systematic review of individual patient data.³²

In a randomized trial,²⁸ patients with an idiopathic VTE event who received anticoagulation for at least 3 months had their d-dimer level measured 1 month after cessation of treatment. Those with an elevated level were randomized to either resume anticoagulation or not. Patients who resumed anticoagulation had an annual recurrence rate of 2%; however, those who were allocated not to restart anticoagulation had a recurrence rate of 10.9% per year. There was no difference in the rate of major bleeding between the two groups. Patients in this clinical trial who had a normal d-dimer level did not restart anticoagulation and had an annual recurrence rate of 4.4%.

Evaluation for residual thrombosis

A randomized trial³⁴ in patients with both provoked and idiopathic deep vein thrombosis showed a reduction in recurrence when those who had residual vein thrombosis were given extended anticoagulation. In the subset of patients whose deep vein thrombosis was idiopathic, the recurrence rate was 17% per year when treatment lasted only 3 months and 10% when it was extended for up to 1 year.

Another trial³⁵ randomized patients with provoked and idiopathic deep vein thrombosis to receive anticoagulation for the usual duration or to continue treatment until recanalization of the residual thrombus was demonstrated on follow-up Doppler ultrasonography. Patients who received this ultrasonography-tailored treatment had a lower rate of recurrence of VTE; however, the absolute reductions in recurrence rates cannot be calculated from this report for patients with idiopathic deep vein thrombosis.

A prospective observational study³⁶ of the predictive value of d-dimer status and residual vein thrombus found that only d-dimer was an independent risk factor for recurrent VTE after vitamin K antagonist withdrawal.

A clinical prediction rule: ‘Men and HERDOO₂’

A promising tool for predicting if a patient is at low risk of recurrent VTE after the first episode of proximal deep vein thrombosis or pulmonary embolism is known by the mnemonic device “Men and HERDOO₂.” It is based on data prospectively derived by Rodger et al³⁷ to identify patients with less than a 3% annual risk of recurrent VTE after their first event of idiopathic proximal deep vein thrombosis or pulmonary embolism. Risk factors for recurrent VTE were male sex (the “men” of “Men and HERDOO₂”), signs of postthrombotic syndrome, including hyperpigmentation of the lower extremities, edema or redness of either leg, a d-dimer level > 250 μg/L, obesity (body mass index > 30 kg/m², and older age (> 65 years).

Overall, one-fourth of the population were women with no risk factors or one risk factor, and their risk of recurrence was 1.6% per year. Men and women who had two or more risk factors for postthrombotic syndrome (hyperpigmentation, edema, or redness), elevated d-dimer, obesity, or older age were predicted to be at higher risk of recurrent VTE. Patients such as this should be considered for indefinite anticoagulation.

Ideally, clinical prediction rules should be validated in a separate group of patients before they are used routinely in practice,³⁸ and this clinical prediction rule is currently being tested in the REVERSE II study. If the results are consistent, this will be an easy-to-use tool to help identify patients who likely can safely stop anticoagulation therapy after 3 to 6 months (clinicaltrials.gov Identifier: NCT00967304).

The location of the thrombosis also influences the likelihood of recurrence. Patients with isolated distal (calf) deep vein thrombosis are less likely to suffer recurrent VTE than those who present with proximal deep vein thrombosis. However, trials focusing specifically on the precise subset of idiopathic isolated distal deep vein thrombosis are lacking. In a randomized trial³⁹ comparing 6 vs 12 weeks of anticoagulation for isolated distal deep vein thrombosis and 12 vs 24 weeks for proximal deep vein thrombosis, the annual rates of recurrence after 12 weeks of treatment were approximately 3.4% for isolated distal and 8.1% for proximal deep vein thrombosis.³⁹

We agree with the ACCP recommendation² that patients with unprovoked VTE should receive at least 3 months of anticoagulation with a vitamin K antagonist.

If the patient has no risk factors for bleeding and good anticoagulant monitoring is achievable, we agree with long-term anticoagulation for proximal unprovoked deep vein thrombosis or pulmonary embolism, and 3 months of therapy for isolated distal unprovoked deep vein thrombosis.

Patient preferences and the risk of recurrence vs the risk of bleeding should be discussed with patients when contemplating indefinite anticoagulation.

If testing is being considered to assist in the decision to prescribe indefinite anticoagulation, we prefer using d-dimer levels rather than ultrasonography to detect residual venous thrombosis because of its ease of use and the strength of the current evidence.

PREVENTING POSTTHROMBOTIC SYNDROME

The postthrombotic (postphlebitic) syndrome is a chronic and burdensome consequence of deep vein thrombosis that occurs despite anticoagulation therapy. It is estimated to affect 23% to 60% of patients and typically manifests in the first 2 years.⁴⁰ It is not only costly in clinical terms, with decreased quality of life for the patient, but health care expenditures have been estimated to range from $400 per year in a Brazilian study to $7,000 per year in a US study.⁴⁰

Typical symptoms include leg pain, heaviness, swelling, and cramping. In severe cases, chronic venous ulcers can occur and are difficult to treat.⁴¹

The definition of postthrombotic syndrome has been unclear over the years, and six different scales that measure signs and symptoms have been reported.⁴²

The Villalta scale has been proposed by the International Society of Thrombosis and Hemostasis as a diagnostic standard to define postthrombotic syndrome.⁴² This validated scale is based on five clinical symptoms, six clinical signs, and the presence or absence of venous ulcers. Each clinical symptom and sign is scored as mild (1 point), moderate (2 points), or severe (3 points). Symptoms include pain, cramps, heaviness, paresthesia, and pruritus; the six clinical signs are pretibial edema, skin induration, hyperpigmentation, redness, venous ectasia, and pain on calf compression.

According to the International Society of Thrombosis and Hemostasis, postthrombotic syndrome is present if the Villalta score is 5 or greater or if a venous ulcer is present in a leg with previous deep vein thrombosis. Further, using the Villalta scale, postthrombotic syndrome can be categorized as mild (score 5–9), moderate (10–14), or severe (≥ 15).

A limitation of the Villalta scale is that the presence or absence of a venous ulcer has not been assigned a score. Since a venous ulcer requires more aggressive measures, the society defines postthrombotic syndrome as severe if venous ulcers are present.⁴²

Acute symptoms of deep vein thrombosis may take months to resolve and, indeed, acute symptoms may transition to chronic symptoms without a symptom-free interval. It is recommended that postthrombotic syndrome not be diagnosed before 3 months to avoid inappropriately attributing acute symptoms and signs of deep vein thrombosis to the postthrombotic syndrome.⁴²

Studies of stockings

A systematic review of three randomized trials⁴⁴ concluded that elastic compression stockings reduce the risk of postthrombotic syndrome (any severity) from 43% to 20% and severe postthrombotic syndrome from 15% to 7%.⁴³

The first of these trials⁴⁴ randomized patients soon after the diagnosis of deep vein thrombosis to receive made-to-order compression stockings that were rated at 30 to 40 mm Hg or to be in a control group that did not receive stockings. The second trial⁴⁵ randomized patients 1 year after the index event of deep vein thrombosis to receive 20- to 30-mm Hg stockings or stockings that were two sizes too large (the control group). The third study⁴⁶ randomly allocated patients to receive “off-the-shelf” stockings (30–40 mm Hg) or no stockings. Each study used its own definition of postthrombotic syndrome.

Although these studies strongly suggest compression stockings prevent postthrombotic syndrome, several methodologic issues remain:

• A standard definition of postthrombotic syndrome was not used
• The amount of compression varied between studies
• The studies were not blinded.

Lack of blinding becomes most significant when an outcome is based on subjective findings, like the symptoms that make up a large part of the diagnosis of postthrombotic syndrome.

The SOX trial, currently under way, is designed to address these methodologic issues and should be completed in 2012 (clinicaltrials.gov Identifier: NCT00143598).

Recommendation: Stockings for at least 2 years

We agree with the ACCP recommendation that a patient who has had a symptomatic proximal deep vein thrombosis should wear an elastic compression stocking with an ankle pressure gradient of 30 to 40 mm Hg as soon as possible after starting anticoagulant therapy and continuing for a minimum of 2 years.²

SCREENING FOR OCCULT MALIGNANCY

VTE can be the first manifestation of cancer.

French physician Armand Trousseau, in the 1860s, was the first to describe disseminated intravascular coagulation closely associated with adenocarcinoma. Ironically, several years later, after suffering for weeks from abdominal pain, he declared to one of his students that he had developed thrombosis, and he died of gastric cancer shortly thereafter.⁴⁷

Since cancer is a well-known risk factor for VTE, it is logical to screen for cancer as an explanation for an idiopathic VTE event.⁴⁸ To make an informed decision, one needs to understand the rate of occult cancer at the time VTE is diagnosed, the risk of future development of cancer, and the utility of extensive cancer screening.

The clinical efficacy, side effects, and cost-effectiveness of cancer screening in patients with idiopathic VTE are unknown. However, a systematic review⁴⁷ of 34 studies found that, in patients with idiopathic VTE, cancer was diagnosed within 1 month in 6.1%, within 6 months in 8.6%, and within 1 year in 10.0% (95% CI 8.6–11.3).

A subset of studies compared two strategies for screening soon after the diagnosis of idiopathic VTE: a strategy limited to the history, physical examination, basic blood work, and chest radiography vs an extensive screening strategy that also included serum tumor markers or abdominal ultrasonography or computed tomography. Limited screening detected 49% of the prevalent cancers; extensive screening increased this rate to 70%. Stated another way, the detection rate for prevalent cancers was 5% with limited screening and 7% with extensive screening soon after the diagnosis of idiopathic VTE.⁴⁷

Patients with idiopathic VTE had higher rates of cancer within 1 month of diagnosis than patients with provoked VTE (6.1% vs 1.9%), and this difference persisted at 1 year (10.0% vs 2.6%).⁴⁷

Recommendation: Individualized cancer screening

Patients with idiopathic VTE have a significant risk of occult cancer within the first year after diagnosis, and cancer screening should be considered. Our practice for patients with idiopathic VTE is to perform a history and physical examination and ensure that the patient is up to date on age- and sex-specific cancer screening.

The use of additional imaging or biomarkers should be discussed with patients so they can balance the risks (radiation and potential false-positive results with their downstream consequences), costs, and potential benefits, given the lack of proven survival benefit or cost-effectiveness.

ORAL ANTICOAGULANT MANAGEMENT

Warfarin’s multiple interactions, along with the need for INR monitoring, make it a difficult medication to manage.

The Joint Commission, the US organization for health service accreditation and certification, has defined National Patient Safety Goals and quality measures for the management of anticoagulation.⁴⁹ Organized anticoagulation management services, dosing algorithms, and patient self-testing using capillary INR meters or patient self-management of warfarin were recommended as tools to improve the time patients spend in the therapeutic INR range.⁵⁰

Two new oral anticoagulants

The limitations of warfarin have stimulated the search for newer oral anticoagulants that do not require laboratory monitoring or have as many diet and drug interactions.

Two trials have been published with experimental oral anticoagulants that had similar efficacy and safety as warfarin in the treatment of VTE.

The study of dabigatran (Pradaxa) vs warfarin in the treatment of acute VTE (the RECOVER trial)⁵¹ randomized 2,539 patients with acute VTE to receive the oral direct thrombin inhibitor dabigatran or warfarin for approximately 6 months. Of note, each treatment group received a median of 6 days of heparin, LMWH, or fondaparinux at the beginning of blinded therapy. The rates of recurrent VTE and major bleeding were similar between the treatment arms, and overall bleeding was less with dabigatran. Dabigatran was approved in the United States in October 2010 for stroke prevention in atrial fibrillation but has yet to be approved for the treatment of VTE pending further study (clinicaltrials.gov Identifier: NCT00680186).

A study of oral rivaroxaban (Xarelto) for symptomatic VTE (the EINSTEIN-DVT trial) ⁵² randomized 3,449 patients with acute deep vein thrombosis to rivaroxaban or enoxaparin (Lovenox) overlapped with warfarin or another vitamin K antagonist in the usual manner. No difference was noted between the treatments in the rate of recurrence of VTE or of major bleeding. Of note, patients randomized to rivaroxaban received 15 mg twice a day for the first 3 weeks of treatment and then 20 mg per day for the remainder of their therapy and did not require parenteral anticoagulant overlap.

The long-awaited promise of easier-to-use oral anticoagulants for the treatment of VTE is drawing near and has the potential to revolutionize the treatment of this common disorder. In the meantime, close monitoring of warfarin and careful patient education regarding its use are essential. And even with the development of new drugs in the future, it is still imperative that patients with acute VTE receive the correct length of anticoagulation treatment, are prescribed stockings to prevent postthrombotic syndrome, and are updated on routine cancer screening.

Deep vein thrombosis and pulmonary embolism are collectively referred to as venous thromboembolic (VTE) disease. They affect approximately 100,000 to 300,000 patients per year in the United States.¹ Although patients with deep vein thrombosis can be treated as outpatients, many are admitted for the initiation of anticoagulation. Initial anticoagulation usually requires the overlap of a parenteral anticoagulant (unfractionated heparin, low-molecular-weight heparin [LMWH] or fondaparinux) with warfarin for a minimum of 5 days and until the international normalized ratio (INR) of the prothrombin time is above 2.0 for at least 24 hours.²

Three clinical issues need to be addressed after the initiation of anticoagulation for VTE:

• Determination of the length of anticoagulation with the correct anticoagulant
• Prevention of postthrombotic syndrome
• Appropriate screening for occult malignancy.

HOW LONG SHOULD VTE BE TREATED?

The duration of anticoagulation has been a matter of debate.

The risk of recurrent VTE appears related to clinical risk factors that a patient has at the time of the initial thrombotic event. An epidemiologic study³ found that patients with VTE treated for approximately 6 months had a low rate of recurrence (0% at 2 years of follow-up) if surgery was the risk factor. The risk climbed to 9% if the risk factor was nonsurgical and to 19% if there were no provoking risk factors.

The likelihood of VTE recurrence and therefore the recommended duration of treatment depend on whether the VTE event was provoked, cancer-related, recurrent, thrombophilia-related, or idiopathic. We address each of these scenarios below.

HOW LONG TO TREAT PROVOKED VTE

A VTE event is considered provoked if the patient had a clear inciting risk factor. As defined in various clinical trials, these risk factors include:

• Hospitalization with confinement to bed for 3 or more consecutive days in the last 3 months
• Surgery or general anesthesia in the last 3 months
• Immobilization for more than 7 days, regardless of the cause
• Trauma in the last 3 months
• Pregnancy
• Use of an oral contraceptive, regardless of which estrogen or progesterone analogue it contains
• Travel for more than 4 hours
• Recent childbirth.

However, the trials that tested different lengths of anticoagulation have varied markedly in how they defined provoked deep vein thrombosis.^4–7

Recommendation: Warfarin or equivalent for 3 months

The American College of Chest Physicians (ACCP) recommends 3 months of anticoagulation with warfarin or another vitamin K antagonist for patients with VTE secondary to a transient (reversible) risk factor,² and we agree.

HOW LONG TO TREAT CANCER-RELATED VTE

Patients with cancer are at higher risk of developing VTE. Furthermore, in one study,⁹ compared with other patients with VTE, patients with cancer were three times more likely to have another episode of VTE, with a cumulative rate of recurrence at 1 year of 21% vs 7%. Cancer patients were also twice as likely to suffer major bleeding complications while on anticoagulation.⁹

Warfarin is a difficult drug to manage because it has many interactions with foods, diseases, and other drugs. These difficulties are amplified in many cancer patients during chemotherapy.

Warfarin was compared with a LMWH in four randomized trials in cancer patients, and a meta-analysis¹⁰ found a 50% relative reduction in the rates of recurrent deep vein thrombosis and pulmonary embolism with LMWH treatment. These results were driven primarily by the CLOT trial (Comparison of Low-Molecular-Weight Heparin Versus Oral Anticoagulant Therapy for the Prevention of Recurrent Venous Thromboembolism in Patients With Cancer),¹¹ which showed an 8% absolute risk reduction (number needed to treat 13) without an increase in major bleeding when cancer-related VTE was treated with an LMWH—ie, dalteparin (Fragmin)—for 6 months compared with warfarin.

Current thinking suggests that VTE should be treated until the cancer is resolved. However, this hypothesis has not been adequately tested, and consequently, the ACCP gives it only a level 1C recommendation.² The largest of the four trials comparing warfarin and an LMWH lasted only 6 months. The safety of extending LMWH treatment beyond 6 months is currently unknown but is under investigation (clinicaltrials.gov identifier NCT00942968).

The ACCP guidelines recommend LMWH therapy for 3 to 6 months, followed by warfarin or another vitamin K antagonist or continued LMWH treatment until the cancer is resolved.²

The National Comprehensive Cancer Network guidelines recommend an LMWH for 6 months as monotherapy and indefinite anticoagulation if the cancer is still active.¹²

The American Society of Clinical Oncology guidelines recommend an LMWH for at least 6 months and indefinite anticoagulant therapy for selected patients with active cancer.¹³

We agree that patients with active cancer should receive an LMWH for at least 6 months and indefinite anticoagulation until the cancer is resolved.

In our experience, many patients are reluctant to give themselves the daily injections that LMWH therapy requires, and so they need to be well-informed about the marked decrease in VTE recurrence with this less-convenient and more-expensive therapy. Many patients face insurance barriers to cover the cost of LMWH therapy; however, careful attention to preauthorization can usually overcome this obstacle.

HOW LONG TO TREAT RECURRENT VTE

It makes clinical sense that patients who have a second VTE event should be treated indefinitely. This theory was tested in a randomized clinical trial¹⁴ in which patients with provoked or unprovoked VTE were randomized after their second event to receive anticoagulation for 6 months vs indefinitely.

After 4 years of follow-up, the recurrence rate was 21% in patients assigned to 6 months of treatment and only 3% in patients who continued anticoagulation throughout the trial. On the other hand, major hemorrhage occurred in 3% of patients treated for 6 months and in 9% in patients who continued anticoagulation indefinitely.

Of note, most of the patients in this trial had unprovoked (idiopathic) VTE, so the results should not be extrapolated to patients with provoked VTE, who accounted for only 20% of the study population.¹⁴

Recommendation: Long-term anticoagulation

We agree with the ACCP recommendation² that patients who have had a second episode of unprovoked VTE should receive long-term anticoagulation. Because of a lack of data, the duration of therapy for patients with a second episode of provoked VTE should be individualized.

HOW LONG TO TREAT THROMBOPHILIA-RELATED VTE

Inherited thrombophilias

Patients with VTE that is not related to a clear provoking risk factor or cancer frequently have testing to evaluate for a hypercoagulable state. This workup traditionally includes the most common inherited thrombophilias for gene mutations for factor V and prothrombin as well as for deficiencies in protein C, protein S, antithrombin and the acquired antiphospholipid syndrome.

The key questions that should be asked prior to embarking on this workup are:

• Will the results change the length of therapy for the patient?
• Will testing the patient help with genetic counseling and possible testing of family members?
• Will the results change the targeted INR range for warfarin or other vitamin K antagonist therapy?

Patients with inherited thrombophilia have a greater risk of developing an initial VTE event; however, these tests do not help predict the recurrence of VTE in patients with established disease more than clinical risk factors do. A prospective study demonstrated this by looking at the effect of thrombophilia and clinical factors on the recurrence of venous thrombosis and found that inherited prothrombotic abnormalities do not appear to play an important role in the risk of a recurrent event.¹⁵ On the other hand, clinical factors, such as whether the first event was idiopathic or provoked, appear more important in determining the duration of anticoagulation therapy.¹⁵ A systematic review of the common inherited thrombophilias showed the VTE recurrence rate of patients with factor V Leiden was higher than in patients without the mutation; however, the absolute rates of recurrence were not much different than what would be expected in patients with idiopathic VTE.¹⁶

A retrospective study involving a large cohort of families of patients who already had experienced a first episode of either idiopathic or provoked VTE showed high annual risks of recurrent VTE associated with hereditary deficiencies of protein S (8.4%), protein C (6.0%), and antithrombin (10%).¹⁷ However, for the more commonly occurring genetic thrombophilias, the factor V Leiden and prothrombin G20210A mutations, family members with either gene abnormality had low rates of VTE, suggesting that testing of relatives of probands is not clinically useful.¹⁶

Antiphospholipid syndrome

Antiphospholipid syndrome is an acquired thrombophilia. A patient has thrombotic antiphospholipid syndrome when there is a history of vascular thrombosis in the presence of persistently positive tests (at least 12 weeks apart) for lupus anticoagulants, anticardiolipin antibodies, or anti-beta-2 glycoprotein I. A prospective study of 412 patients with a first episode of VTE found that 15% were positive for anticardiolipin antibody at the end of 6 months of anticoagulation. The risk of recurrent VTE after 4 years was 29% in patients with antibodies and 14% in those without antibodies (relative risk 2.1; 95% confidence interval [CI] 1.3–3.3; P =.0013).¹⁸

Recent reviews advise indefinite warfarin anticoagulation in patients with VTE and persistence of antiphospholipid antibodies.¹⁹ However, the optimal duration of anticoagulation is uncertain. Until well-designed clinical trials are done, the current general consensus is to anticoagulate these patients indefinitely.^20,21 Retrospective studies had suggested that patients with antiphospholipid antibodies required a higher therapeutic INR range; however, this observation was tested in two trials that found no difference in thromboembolic rates when patients were randomized to an INR of 2.0–3.0 vs 3.1–4.0,²² or 2.0–3.0 vs 3.0–4.5.²³

No formal recommendations

In the absence of strong evidence, the ACCP guidelines do not include a recommendation on the duration of anticoagulation treatment specific to inherited thrombophilias. We believe that clinical factors are more important than inherited thrombophilias for deciding the duration of anticoagulation, and that testing is almost never indicated or useful. However, patients with antiphospholipid syndrome are at high risk of recurrence, and it is our practice to anticoagulate these patients indefinitely.

HOW LONG TO TREAT UNPROVOKED (IDIOPATHIC) VTE

A VTE event is thought to be idiopathic if it occurs without a clearly identified provoking factor.

In an observational study,³ patients with oral contraceptive use, transient illness, immobilization, or a history of travel had an 8.8% risk of recurrence vs 19.4% in patients with unprovoked VTE. The meta-analysis discussed above (Table 1)⁸ also shows that patients with these nonsurgical risk factors have a lower rate of recurrence than patients with idiopathic VTE.

The high rate of recurrence of idiopathic VTE (4% to 27% after 3 months of anticoagulation^24–26) suggests that a longer duration of treatment is reasonable. However, increasing the length of therapy from 3 to 12 months delays but does not prevent recurrence, the risk of which begins to accumulate once anticoagulation is stopped.^24,25

Three promising strategies to identify subgroups of patients with idiopathic VTE who are at highest risk of recurrence and who would benefit the most from prolonged anticoagulation are d-dimer testing, evaluation for residual vein thrombosis in patients who present with a deep vein thrombosis, and clinical prediction rules.

d-dimer testing

d-dimer is a degradation product of fibrin and is an indirect marker of residual thrombosis.³⁰

In a systematic review of patients with a first episode of unprovoked VTE,³¹ a normal d-dimer concentration at the end of at least 3 months of anticoagulation was associated with a 3.5% annual risk of recurrence, whereas an elevated d-dimer level at that time was associated with an annual risk of 8.9%. These results were confirmed in a systematic review of individual patient data.³²

In a randomized trial,²⁸ patients with an idiopathic VTE event who received anticoagulation for at least 3 months had their d-dimer level measured 1 month after cessation of treatment. Those with an elevated level were randomized to either resume anticoagulation or not. Patients who resumed anticoagulation had an annual recurrence rate of 2%; however, those who were allocated not to restart anticoagulation had a recurrence rate of 10.9% per year. There was no difference in the rate of major bleeding between the two groups. Patients in this clinical trial who had a normal d-dimer level did not restart anticoagulation and had an annual recurrence rate of 4.4%.

Evaluation for residual thrombosis

A randomized trial³⁴ in patients with both provoked and idiopathic deep vein thrombosis showed a reduction in recurrence when those who had residual vein thrombosis were given extended anticoagulation. In the subset of patients whose deep vein thrombosis was idiopathic, the recurrence rate was 17% per year when treatment lasted only 3 months and 10% when it was extended for up to 1 year.

Another trial³⁵ randomized patients with provoked and idiopathic deep vein thrombosis to receive anticoagulation for the usual duration or to continue treatment until recanalization of the residual thrombus was demonstrated on follow-up Doppler ultrasonography. Patients who received this ultrasonography-tailored treatment had a lower rate of recurrence of VTE; however, the absolute reductions in recurrence rates cannot be calculated from this report for patients with idiopathic deep vein thrombosis.

A prospective observational study³⁶ of the predictive value of d-dimer status and residual vein thrombus found that only d-dimer was an independent risk factor for recurrent VTE after vitamin K antagonist withdrawal.

A clinical prediction rule: ‘Men and HERDOO₂’

A promising tool for predicting if a patient is at low risk of recurrent VTE after the first episode of proximal deep vein thrombosis or pulmonary embolism is known by the mnemonic device “Men and HERDOO₂.” It is based on data prospectively derived by Rodger et al³⁷ to identify patients with less than a 3% annual risk of recurrent VTE after their first event of idiopathic proximal deep vein thrombosis or pulmonary embolism. Risk factors for recurrent VTE were male sex (the “men” of “Men and HERDOO₂”), signs of postthrombotic syndrome, including hyperpigmentation of the lower extremities, edema or redness of either leg, a d-dimer level > 250 μg/L, obesity (body mass index > 30 kg/m², and older age (> 65 years).

Overall, one-fourth of the population were women with no risk factors or one risk factor, and their risk of recurrence was 1.6% per year. Men and women who had two or more risk factors for postthrombotic syndrome (hyperpigmentation, edema, or redness), elevated d-dimer, obesity, or older age were predicted to be at higher risk of recurrent VTE. Patients such as this should be considered for indefinite anticoagulation.

Ideally, clinical prediction rules should be validated in a separate group of patients before they are used routinely in practice,³⁸ and this clinical prediction rule is currently being tested in the REVERSE II study. If the results are consistent, this will be an easy-to-use tool to help identify patients who likely can safely stop anticoagulation therapy after 3 to 6 months (clinicaltrials.gov Identifier: NCT00967304).

The location of the thrombosis also influences the likelihood of recurrence. Patients with isolated distal (calf) deep vein thrombosis are less likely to suffer recurrent VTE than those who present with proximal deep vein thrombosis. However, trials focusing specifically on the precise subset of idiopathic isolated distal deep vein thrombosis are lacking. In a randomized trial³⁹ comparing 6 vs 12 weeks of anticoagulation for isolated distal deep vein thrombosis and 12 vs 24 weeks for proximal deep vein thrombosis, the annual rates of recurrence after 12 weeks of treatment were approximately 3.4% for isolated distal and 8.1% for proximal deep vein thrombosis.³⁹

We agree with the ACCP recommendation² that patients with unprovoked VTE should receive at least 3 months of anticoagulation with a vitamin K antagonist.

If the patient has no risk factors for bleeding and good anticoagulant monitoring is achievable, we agree with long-term anticoagulation for proximal unprovoked deep vein thrombosis or pulmonary embolism, and 3 months of therapy for isolated distal unprovoked deep vein thrombosis.

Patient preferences and the risk of recurrence vs the risk of bleeding should be discussed with patients when contemplating indefinite anticoagulation.

If testing is being considered to assist in the decision to prescribe indefinite anticoagulation, we prefer using d-dimer levels rather than ultrasonography to detect residual venous thrombosis because of its ease of use and the strength of the current evidence.

PREVENTING POSTTHROMBOTIC SYNDROME

The postthrombotic (postphlebitic) syndrome is a chronic and burdensome consequence of deep vein thrombosis that occurs despite anticoagulation therapy. It is estimated to affect 23% to 60% of patients and typically manifests in the first 2 years.⁴⁰ It is not only costly in clinical terms, with decreased quality of life for the patient, but health care expenditures have been estimated to range from $400 per year in a Brazilian study to $7,000 per year in a US study.⁴⁰

Typical symptoms include leg pain, heaviness, swelling, and cramping. In severe cases, chronic venous ulcers can occur and are difficult to treat.⁴¹

The definition of postthrombotic syndrome has been unclear over the years, and six different scales that measure signs and symptoms have been reported.⁴²

The Villalta scale has been proposed by the International Society of Thrombosis and Hemostasis as a diagnostic standard to define postthrombotic syndrome.⁴² This validated scale is based on five clinical symptoms, six clinical signs, and the presence or absence of venous ulcers. Each clinical symptom and sign is scored as mild (1 point), moderate (2 points), or severe (3 points). Symptoms include pain, cramps, heaviness, paresthesia, and pruritus; the six clinical signs are pretibial edema, skin induration, hyperpigmentation, redness, venous ectasia, and pain on calf compression.

According to the International Society of Thrombosis and Hemostasis, postthrombotic syndrome is present if the Villalta score is 5 or greater or if a venous ulcer is present in a leg with previous deep vein thrombosis. Further, using the Villalta scale, postthrombotic syndrome can be categorized as mild (score 5–9), moderate (10–14), or severe (≥ 15).

A limitation of the Villalta scale is that the presence or absence of a venous ulcer has not been assigned a score. Since a venous ulcer requires more aggressive measures, the society defines postthrombotic syndrome as severe if venous ulcers are present.⁴²

Acute symptoms of deep vein thrombosis may take months to resolve and, indeed, acute symptoms may transition to chronic symptoms without a symptom-free interval. It is recommended that postthrombotic syndrome not be diagnosed before 3 months to avoid inappropriately attributing acute symptoms and signs of deep vein thrombosis to the postthrombotic syndrome.⁴²

Studies of stockings

A systematic review of three randomized trials⁴⁴ concluded that elastic compression stockings reduce the risk of postthrombotic syndrome (any severity) from 43% to 20% and severe postthrombotic syndrome from 15% to 7%.⁴³

The first of these trials⁴⁴ randomized patients soon after the diagnosis of deep vein thrombosis to receive made-to-order compression stockings that were rated at 30 to 40 mm Hg or to be in a control group that did not receive stockings. The second trial⁴⁵ randomized patients 1 year after the index event of deep vein thrombosis to receive 20- to 30-mm Hg stockings or stockings that were two sizes too large (the control group). The third study⁴⁶ randomly allocated patients to receive “off-the-shelf” stockings (30–40 mm Hg) or no stockings. Each study used its own definition of postthrombotic syndrome.

Although these studies strongly suggest compression stockings prevent postthrombotic syndrome, several methodologic issues remain:

• A standard definition of postthrombotic syndrome was not used
• The amount of compression varied between studies
• The studies were not blinded.

Lack of blinding becomes most significant when an outcome is based on subjective findings, like the symptoms that make up a large part of the diagnosis of postthrombotic syndrome.

The SOX trial, currently under way, is designed to address these methodologic issues and should be completed in 2012 (clinicaltrials.gov Identifier: NCT00143598).

Recommendation: Stockings for at least 2 years

We agree with the ACCP recommendation that a patient who has had a symptomatic proximal deep vein thrombosis should wear an elastic compression stocking with an ankle pressure gradient of 30 to 40 mm Hg as soon as possible after starting anticoagulant therapy and continuing for a minimum of 2 years.²

SCREENING FOR OCCULT MALIGNANCY

VTE can be the first manifestation of cancer.

French physician Armand Trousseau, in the 1860s, was the first to describe disseminated intravascular coagulation closely associated with adenocarcinoma. Ironically, several years later, after suffering for weeks from abdominal pain, he declared to one of his students that he had developed thrombosis, and he died of gastric cancer shortly thereafter.⁴⁷

Since cancer is a well-known risk factor for VTE, it is logical to screen for cancer as an explanation for an idiopathic VTE event.⁴⁸ To make an informed decision, one needs to understand the rate of occult cancer at the time VTE is diagnosed, the risk of future development of cancer, and the utility of extensive cancer screening.

The clinical efficacy, side effects, and cost-effectiveness of cancer screening in patients with idiopathic VTE are unknown. However, a systematic review⁴⁷ of 34 studies found that, in patients with idiopathic VTE, cancer was diagnosed within 1 month in 6.1%, within 6 months in 8.6%, and within 1 year in 10.0% (95% CI 8.6–11.3).

A subset of studies compared two strategies for screening soon after the diagnosis of idiopathic VTE: a strategy limited to the history, physical examination, basic blood work, and chest radiography vs an extensive screening strategy that also included serum tumor markers or abdominal ultrasonography or computed tomography. Limited screening detected 49% of the prevalent cancers; extensive screening increased this rate to 70%. Stated another way, the detection rate for prevalent cancers was 5% with limited screening and 7% with extensive screening soon after the diagnosis of idiopathic VTE.⁴⁷

Patients with idiopathic VTE had higher rates of cancer within 1 month of diagnosis than patients with provoked VTE (6.1% vs 1.9%), and this difference persisted at 1 year (10.0% vs 2.6%).⁴⁷

Recommendation: Individualized cancer screening

Patients with idiopathic VTE have a significant risk of occult cancer within the first year after diagnosis, and cancer screening should be considered. Our practice for patients with idiopathic VTE is to perform a history and physical examination and ensure that the patient is up to date on age- and sex-specific cancer screening.

The use of additional imaging or biomarkers should be discussed with patients so they can balance the risks (radiation and potential false-positive results with their downstream consequences), costs, and potential benefits, given the lack of proven survival benefit or cost-effectiveness.

ORAL ANTICOAGULANT MANAGEMENT

Warfarin’s multiple interactions, along with the need for INR monitoring, make it a difficult medication to manage.

The Joint Commission, the US organization for health service accreditation and certification, has defined National Patient Safety Goals and quality measures for the management of anticoagulation.⁴⁹ Organized anticoagulation management services, dosing algorithms, and patient self-testing using capillary INR meters or patient self-management of warfarin were recommended as tools to improve the time patients spend in the therapeutic INR range.⁵⁰

Two new oral anticoagulants

The limitations of warfarin have stimulated the search for newer oral anticoagulants that do not require laboratory monitoring or have as many diet and drug interactions.

Two trials have been published with experimental oral anticoagulants that had similar efficacy and safety as warfarin in the treatment of VTE.

The study of dabigatran (Pradaxa) vs warfarin in the treatment of acute VTE (the RECOVER trial)⁵¹ randomized 2,539 patients with acute VTE to receive the oral direct thrombin inhibitor dabigatran or warfarin for approximately 6 months. Of note, each treatment group received a median of 6 days of heparin, LMWH, or fondaparinux at the beginning of blinded therapy. The rates of recurrent VTE and major bleeding were similar between the treatment arms, and overall bleeding was less with dabigatran. Dabigatran was approved in the United States in October 2010 for stroke prevention in atrial fibrillation but has yet to be approved for the treatment of VTE pending further study (clinicaltrials.gov Identifier: NCT00680186).

A study of oral rivaroxaban (Xarelto) for symptomatic VTE (the EINSTEIN-DVT trial) ⁵² randomized 3,449 patients with acute deep vein thrombosis to rivaroxaban or enoxaparin (Lovenox) overlapped with warfarin or another vitamin K antagonist in the usual manner. No difference was noted between the treatments in the rate of recurrence of VTE or of major bleeding. Of note, patients randomized to rivaroxaban received 15 mg twice a day for the first 3 weeks of treatment and then 20 mg per day for the remainder of their therapy and did not require parenteral anticoagulant overlap.

The long-awaited promise of easier-to-use oral anticoagulants for the treatment of VTE is drawing near and has the potential to revolutionize the treatment of this common disorder. In the meantime, close monitoring of warfarin and careful patient education regarding its use are essential. And even with the development of new drugs in the future, it is still imperative that patients with acute VTE receive the correct length of anticoagulation treatment, are prescribed stockings to prevent postthrombotic syndrome, and are updated on routine cancer screening.

Deep vein thrombosis and pulmonary embolism are collectively referred to as venous thromboembolic (VTE) disease. They affect approximately 100,000 to 300,000 patients per year in the United States.¹ Although patients with deep vein thrombosis can be treated as outpatients, many are admitted for the initiation of anticoagulation. Initial anticoagulation usually requires the overlap of a parenteral anticoagulant (unfractionated heparin, low-molecular-weight heparin [LMWH] or fondaparinux) with warfarin for a minimum of 5 days and until the international normalized ratio (INR) of the prothrombin time is above 2.0 for at least 24 hours.²

Three clinical issues need to be addressed after the initiation of anticoagulation for VTE:

• Determination of the length of anticoagulation with the correct anticoagulant
• Prevention of postthrombotic syndrome
• Appropriate screening for occult malignancy.

HOW LONG SHOULD VTE BE TREATED?

The duration of anticoagulation has been a matter of debate.

The risk of recurrent VTE appears related to clinical risk factors that a patient has at the time of the initial thrombotic event. An epidemiologic study³ found that patients with VTE treated for approximately 6 months had a low rate of recurrence (0% at 2 years of follow-up) if surgery was the risk factor. The risk climbed to 9% if the risk factor was nonsurgical and to 19% if there were no provoking risk factors.

The likelihood of VTE recurrence and therefore the recommended duration of treatment depend on whether the VTE event was provoked, cancer-related, recurrent, thrombophilia-related, or idiopathic. We address each of these scenarios below.

HOW LONG TO TREAT PROVOKED VTE

A VTE event is considered provoked if the patient had a clear inciting risk factor. As defined in various clinical trials, these risk factors include:

• Hospitalization with confinement to bed for 3 or more consecutive days in the last 3 months
• Surgery or general anesthesia in the last 3 months
• Immobilization for more than 7 days, regardless of the cause
• Trauma in the last 3 months
• Pregnancy
• Use of an oral contraceptive, regardless of which estrogen or progesterone analogue it contains
• Travel for more than 4 hours
• Recent childbirth.

However, the trials that tested different lengths of anticoagulation have varied markedly in how they defined provoked deep vein thrombosis.^4–7

Recommendation: Warfarin or equivalent for 3 months

The American College of Chest Physicians (ACCP) recommends 3 months of anticoagulation with warfarin or another vitamin K antagonist for patients with VTE secondary to a transient (reversible) risk factor,² and we agree.

HOW LONG TO TREAT CANCER-RELATED VTE

Patients with cancer are at higher risk of developing VTE. Furthermore, in one study,⁹ compared with other patients with VTE, patients with cancer were three times more likely to have another episode of VTE, with a cumulative rate of recurrence at 1 year of 21% vs 7%. Cancer patients were also twice as likely to suffer major bleeding complications while on anticoagulation.⁹

Warfarin is a difficult drug to manage because it has many interactions with foods, diseases, and other drugs. These difficulties are amplified in many cancer patients during chemotherapy.

Warfarin was compared with a LMWH in four randomized trials in cancer patients, and a meta-analysis¹⁰ found a 50% relative reduction in the rates of recurrent deep vein thrombosis and pulmonary embolism with LMWH treatment. These results were driven primarily by the CLOT trial (Comparison of Low-Molecular-Weight Heparin Versus Oral Anticoagulant Therapy for the Prevention of Recurrent Venous Thromboembolism in Patients With Cancer),¹¹ which showed an 8% absolute risk reduction (number needed to treat 13) without an increase in major bleeding when cancer-related VTE was treated with an LMWH—ie, dalteparin (Fragmin)—for 6 months compared with warfarin.

Current thinking suggests that VTE should be treated until the cancer is resolved. However, this hypothesis has not been adequately tested, and consequently, the ACCP gives it only a level 1C recommendation.² The largest of the four trials comparing warfarin and an LMWH lasted only 6 months. The safety of extending LMWH treatment beyond 6 months is currently unknown but is under investigation (clinicaltrials.gov identifier NCT00942968).

The ACCP guidelines recommend LMWH therapy for 3 to 6 months, followed by warfarin or another vitamin K antagonist or continued LMWH treatment until the cancer is resolved.²

The National Comprehensive Cancer Network guidelines recommend an LMWH for 6 months as monotherapy and indefinite anticoagulation if the cancer is still active.¹²

The American Society of Clinical Oncology guidelines recommend an LMWH for at least 6 months and indefinite anticoagulant therapy for selected patients with active cancer.¹³

We agree that patients with active cancer should receive an LMWH for at least 6 months and indefinite anticoagulation until the cancer is resolved.

In our experience, many patients are reluctant to give themselves the daily injections that LMWH therapy requires, and so they need to be well-informed about the marked decrease in VTE recurrence with this less-convenient and more-expensive therapy. Many patients face insurance barriers to cover the cost of LMWH therapy; however, careful attention to preauthorization can usually overcome this obstacle.

HOW LONG TO TREAT RECURRENT VTE

It makes clinical sense that patients who have a second VTE event should be treated indefinitely. This theory was tested in a randomized clinical trial¹⁴ in which patients with provoked or unprovoked VTE were randomized after their second event to receive anticoagulation for 6 months vs indefinitely.

After 4 years of follow-up, the recurrence rate was 21% in patients assigned to 6 months of treatment and only 3% in patients who continued anticoagulation throughout the trial. On the other hand, major hemorrhage occurred in 3% of patients treated for 6 months and in 9% in patients who continued anticoagulation indefinitely.

Of note, most of the patients in this trial had unprovoked (idiopathic) VTE, so the results should not be extrapolated to patients with provoked VTE, who accounted for only 20% of the study population.¹⁴

Recommendation: Long-term anticoagulation

We agree with the ACCP recommendation² that patients who have had a second episode of unprovoked VTE should receive long-term anticoagulation. Because of a lack of data, the duration of therapy for patients with a second episode of provoked VTE should be individualized.

HOW LONG TO TREAT THROMBOPHILIA-RELATED VTE

Inherited thrombophilias

Patients with VTE that is not related to a clear provoking risk factor or cancer frequently have testing to evaluate for a hypercoagulable state. This workup traditionally includes the most common inherited thrombophilias for gene mutations for factor V and prothrombin as well as for deficiencies in protein C, protein S, antithrombin and the acquired antiphospholipid syndrome.

The key questions that should be asked prior to embarking on this workup are:

• Will the results change the length of therapy for the patient?
• Will testing the patient help with genetic counseling and possible testing of family members?
• Will the results change the targeted INR range for warfarin or other vitamin K antagonist therapy?

Patients with inherited thrombophilia have a greater risk of developing an initial VTE event; however, these tests do not help predict the recurrence of VTE in patients with established disease more than clinical risk factors do. A prospective study demonstrated this by looking at the effect of thrombophilia and clinical factors on the recurrence of venous thrombosis and found that inherited prothrombotic abnormalities do not appear to play an important role in the risk of a recurrent event.¹⁵ On the other hand, clinical factors, such as whether the first event was idiopathic or provoked, appear more important in determining the duration of anticoagulation therapy.¹⁵ A systematic review of the common inherited thrombophilias showed the VTE recurrence rate of patients with factor V Leiden was higher than in patients without the mutation; however, the absolute rates of recurrence were not much different than what would be expected in patients with idiopathic VTE.¹⁶

A retrospective study involving a large cohort of families of patients who already had experienced a first episode of either idiopathic or provoked VTE showed high annual risks of recurrent VTE associated with hereditary deficiencies of protein S (8.4%), protein C (6.0%), and antithrombin (10%).¹⁷ However, for the more commonly occurring genetic thrombophilias, the factor V Leiden and prothrombin G20210A mutations, family members with either gene abnormality had low rates of VTE, suggesting that testing of relatives of probands is not clinically useful.¹⁶

Antiphospholipid syndrome

Antiphospholipid syndrome is an acquired thrombophilia. A patient has thrombotic antiphospholipid syndrome when there is a history of vascular thrombosis in the presence of persistently positive tests (at least 12 weeks apart) for lupus anticoagulants, anticardiolipin antibodies, or anti-beta-2 glycoprotein I. A prospective study of 412 patients with a first episode of VTE found that 15% were positive for anticardiolipin antibody at the end of 6 months of anticoagulation. The risk of recurrent VTE after 4 years was 29% in patients with antibodies and 14% in those without antibodies (relative risk 2.1; 95% confidence interval [CI] 1.3–3.3; P =.0013).¹⁸

Recent reviews advise indefinite warfarin anticoagulation in patients with VTE and persistence of antiphospholipid antibodies.¹⁹ However, the optimal duration of anticoagulation is uncertain. Until well-designed clinical trials are done, the current general consensus is to anticoagulate these patients indefinitely.^20,21 Retrospective studies had suggested that patients with antiphospholipid antibodies required a higher therapeutic INR range; however, this observation was tested in two trials that found no difference in thromboembolic rates when patients were randomized to an INR of 2.0–3.0 vs 3.1–4.0,²² or 2.0–3.0 vs 3.0–4.5.²³

No formal recommendations

In the absence of strong evidence, the ACCP guidelines do not include a recommendation on the duration of anticoagulation treatment specific to inherited thrombophilias. We believe that clinical factors are more important than inherited thrombophilias for deciding the duration of anticoagulation, and that testing is almost never indicated or useful. However, patients with antiphospholipid syndrome are at high risk of recurrence, and it is our practice to anticoagulate these patients indefinitely.

HOW LONG TO TREAT UNPROVOKED (IDIOPATHIC) VTE

A VTE event is thought to be idiopathic if it occurs without a clearly identified provoking factor.

In an observational study,³ patients with oral contraceptive use, transient illness, immobilization, or a history of travel had an 8.8% risk of recurrence vs 19.4% in patients with unprovoked VTE. The meta-analysis discussed above (Table 1)⁸ also shows that patients with these nonsurgical risk factors have a lower rate of recurrence than patients with idiopathic VTE.

The high rate of recurrence of idiopathic VTE (4% to 27% after 3 months of anticoagulation^24–26) suggests that a longer duration of treatment is reasonable. However, increasing the length of therapy from 3 to 12 months delays but does not prevent recurrence, the risk of which begins to accumulate once anticoagulation is stopped.^24,25

Three promising strategies to identify subgroups of patients with idiopathic VTE who are at highest risk of recurrence and who would benefit the most from prolonged anticoagulation are d-dimer testing, evaluation for residual vein thrombosis in patients who present with a deep vein thrombosis, and clinical prediction rules.

d-dimer testing

d-dimer is a degradation product of fibrin and is an indirect marker of residual thrombosis.³⁰

In a systematic review of patients with a first episode of unprovoked VTE,³¹ a normal d-dimer concentration at the end of at least 3 months of anticoagulation was associated with a 3.5% annual risk of recurrence, whereas an elevated d-dimer level at that time was associated with an annual risk of 8.9%. These results were confirmed in a systematic review of individual patient data.³²

In a randomized trial,²⁸ patients with an idiopathic VTE event who received anticoagulation for at least 3 months had their d-dimer level measured 1 month after cessation of treatment. Those with an elevated level were randomized to either resume anticoagulation or not. Patients who resumed anticoagulation had an annual recurrence rate of 2%; however, those who were allocated not to restart anticoagulation had a recurrence rate of 10.9% per year. There was no difference in the rate of major bleeding between the two groups. Patients in this clinical trial who had a normal d-dimer level did not restart anticoagulation and had an annual recurrence rate of 4.4%.

Evaluation for residual thrombosis

A randomized trial³⁴ in patients with both provoked and idiopathic deep vein thrombosis showed a reduction in recurrence when those who had residual vein thrombosis were given extended anticoagulation. In the subset of patients whose deep vein thrombosis was idiopathic, the recurrence rate was 17% per year when treatment lasted only 3 months and 10% when it was extended for up to 1 year.

Another trial³⁵ randomized patients with provoked and idiopathic deep vein thrombosis to receive anticoagulation for the usual duration or to continue treatment until recanalization of the residual thrombus was demonstrated on follow-up Doppler ultrasonography. Patients who received this ultrasonography-tailored treatment had a lower rate of recurrence of VTE; however, the absolute reductions in recurrence rates cannot be calculated from this report for patients with idiopathic deep vein thrombosis.

A prospective observational study³⁶ of the predictive value of d-dimer status and residual vein thrombus found that only d-dimer was an independent risk factor for recurrent VTE after vitamin K antagonist withdrawal.

A clinical prediction rule: ‘Men and HERDOO₂’

A promising tool for predicting if a patient is at low risk of recurrent VTE after the first episode of proximal deep vein thrombosis or pulmonary embolism is known by the mnemonic device “Men and HERDOO₂.” It is based on data prospectively derived by Rodger et al³⁷ to identify patients with less than a 3% annual risk of recurrent VTE after their first event of idiopathic proximal deep vein thrombosis or pulmonary embolism. Risk factors for recurrent VTE were male sex (the “men” of “Men and HERDOO₂”), signs of postthrombotic syndrome, including hyperpigmentation of the lower extremities, edema or redness of either leg, a d-dimer level > 250 μg/L, obesity (body mass index > 30 kg/m², and older age (> 65 years).

Overall, one-fourth of the population were women with no risk factors or one risk factor, and their risk of recurrence was 1.6% per year. Men and women who had two or more risk factors for postthrombotic syndrome (hyperpigmentation, edema, or redness), elevated d-dimer, obesity, or older age were predicted to be at higher risk of recurrent VTE. Patients such as this should be considered for indefinite anticoagulation.

Ideally, clinical prediction rules should be validated in a separate group of patients before they are used routinely in practice,³⁸ and this clinical prediction rule is currently being tested in the REVERSE II study. If the results are consistent, this will be an easy-to-use tool to help identify patients who likely can safely stop anticoagulation therapy after 3 to 6 months (clinicaltrials.gov Identifier: NCT00967304).

The location of the thrombosis also influences the likelihood of recurrence. Patients with isolated distal (calf) deep vein thrombosis are less likely to suffer recurrent VTE than those who present with proximal deep vein thrombosis. However, trials focusing specifically on the precise subset of idiopathic isolated distal deep vein thrombosis are lacking. In a randomized trial³⁹ comparing 6 vs 12 weeks of anticoagulation for isolated distal deep vein thrombosis and 12 vs 24 weeks for proximal deep vein thrombosis, the annual rates of recurrence after 12 weeks of treatment were approximately 3.4% for isolated distal and 8.1% for proximal deep vein thrombosis.³⁹

We agree with the ACCP recommendation² that patients with unprovoked VTE should receive at least 3 months of anticoagulation with a vitamin K antagonist.

If the patient has no risk factors for bleeding and good anticoagulant monitoring is achievable, we agree with long-term anticoagulation for proximal unprovoked deep vein thrombosis or pulmonary embolism, and 3 months of therapy for isolated distal unprovoked deep vein thrombosis.

Patient preferences and the risk of recurrence vs the risk of bleeding should be discussed with patients when contemplating indefinite anticoagulation.

If testing is being considered to assist in the decision to prescribe indefinite anticoagulation, we prefer using d-dimer levels rather than ultrasonography to detect residual venous thrombosis because of its ease of use and the strength of the current evidence.

PREVENTING POSTTHROMBOTIC SYNDROME

The postthrombotic (postphlebitic) syndrome is a chronic and burdensome consequence of deep vein thrombosis that occurs despite anticoagulation therapy. It is estimated to affect 23% to 60% of patients and typically manifests in the first 2 years.⁴⁰ It is not only costly in clinical terms, with decreased quality of life for the patient, but health care expenditures have been estimated to range from $400 per year in a Brazilian study to $7,000 per year in a US study.⁴⁰

Typical symptoms include leg pain, heaviness, swelling, and cramping. In severe cases, chronic venous ulcers can occur and are difficult to treat.⁴¹

The definition of postthrombotic syndrome has been unclear over the years, and six different scales that measure signs and symptoms have been reported.⁴²

The Villalta scale has been proposed by the International Society of Thrombosis and Hemostasis as a diagnostic standard to define postthrombotic syndrome.⁴² This validated scale is based on five clinical symptoms, six clinical signs, and the presence or absence of venous ulcers. Each clinical symptom and sign is scored as mild (1 point), moderate (2 points), or severe (3 points). Symptoms include pain, cramps, heaviness, paresthesia, and pruritus; the six clinical signs are pretibial edema, skin induration, hyperpigmentation, redness, venous ectasia, and pain on calf compression.

According to the International Society of Thrombosis and Hemostasis, postthrombotic syndrome is present if the Villalta score is 5 or greater or if a venous ulcer is present in a leg with previous deep vein thrombosis. Further, using the Villalta scale, postthrombotic syndrome can be categorized as mild (score 5–9), moderate (10–14), or severe (≥ 15).

A limitation of the Villalta scale is that the presence or absence of a venous ulcer has not been assigned a score. Since a venous ulcer requires more aggressive measures, the society defines postthrombotic syndrome as severe if venous ulcers are present.⁴²

Acute symptoms of deep vein thrombosis may take months to resolve and, indeed, acute symptoms may transition to chronic symptoms without a symptom-free interval. It is recommended that postthrombotic syndrome not be diagnosed before 3 months to avoid inappropriately attributing acute symptoms and signs of deep vein thrombosis to the postthrombotic syndrome.⁴²

Studies of stockings

A systematic review of three randomized trials⁴⁴ concluded that elastic compression stockings reduce the risk of postthrombotic syndrome (any severity) from 43% to 20% and severe postthrombotic syndrome from 15% to 7%.⁴³

The first of these trials⁴⁴ randomized patients soon after the diagnosis of deep vein thrombosis to receive made-to-order compression stockings that were rated at 30 to 40 mm Hg or to be in a control group that did not receive stockings. The second trial⁴⁵ randomized patients 1 year after the index event of deep vein thrombosis to receive 20- to 30-mm Hg stockings or stockings that were two sizes too large (the control group). The third study⁴⁶ randomly allocated patients to receive “off-the-shelf” stockings (30–40 mm Hg) or no stockings. Each study used its own definition of postthrombotic syndrome.

Although these studies strongly suggest compression stockings prevent postthrombotic syndrome, several methodologic issues remain: