Section 18: Vascular Medicine

WHAT IS THE RISK FOR VENOUS THROMBOEMBOLISM (VTE) IN HOSPITALIZED MEDICAL PATIENTS?

EPIDEMIOLOGY

Venous Thromboembolism (VTE), which comprises deep vein thrombosis (DVT) and pulmonary embolism (PE), is a common cause of morbidity and mortality in hospitalized medical patients. The baseline incidence of asymptomatic VTE in hospitalized medical patients without anticoagulant prophylaxis is 7% to 15%. Linked administrative database studies indicate that 1.7% of hospitalized medical patients develop symptomatic VTE within 3 months of hospitalization. This is lower than in surgical patients, who have a risk of 2% to 3%. However, because of the sheer number of hospitalized medical patients when compared to surgical patients, the burden of illness is high. Approximately 50% to 70% of symptomatic VTE and 70% to 80% of fatal PE occur in medical patients, and recent hospitalization for medical illness accounts for 25% of all VTE diagnosed in the community. The quoted risk of 1.7% is also based on the risk in all medical patients, some of whom have less severe illness. In prospective studies assessing medical patients who have at least one major risk factor for VTE such as severe cardiac or respiratory disease and do not receive VTE prophylaxis, the incidence of DVT as detected by venography is approximately 10% to 15%. In the absence of anticoagulant prophylaxis, the incidence of proximal DVT, which is the type of DVT most likely to embolize, is approximately 5% and the incidence of PE is 0.5%. VTE is associated with potentially serious long-term complications, including post-thrombotic syndrome, cardiorespiratory insufficiency, recurrent VTE, and bleeding associated with anticoagulant therapy. VTE is also a common cause of readmission to the hospital, and is associated with increased hospital costs and length of stay.

PATHOPHYSIOLOGY

Hospitalization for an acute medical illness is independently associated with about an eightfold increased risk for VTE. Chart audits have shown that nearly all hospitalized medical patients have at least one VTE risk factor, be it immobility, increased age, cancer (active or occult), or acute medical illness (eg, congestive heart failure, obstructive lung disease). Certain populations of hospitalized medical patients, such as those in the intensive care unit, have additional risk factors including central venous catheterization; these patients are considered to be at high risk for VTE, even after receiving routine prophylaxis (Table 252-1).

WHICH HOSPITALIZED MEDICAL PATIENTS NEED VTE PROPHYLAXIS?

VTE PROPHYLAXIS IN GENERAL MEDICAL PATIENTS

There is no formal, widely accepted, prospectively validated risk stratification algorithm to guide VTE prophylaxis in medical patients. As a general guide, medical patients presenting with ischemic stroke, acute exacerbations of chronic heart failure or chronic obstructive pulmonary disease, acute respiratory failure, cancer, history of prior VTE, sepsis, acute neurologic disease, or severe inflammatory disease should be given VTE prophylaxis. Immobility is considered a weaker risk factor, and is difficult to clearly define. However, patients who cannot ambulate without assistance still merit consideration for prophylaxis. A number of data-derived risk models to predict VTE in medical patients have been proposed, including the Padua prediction score, the IMPROVE risk score, and the Geneva risk score. They include many of the above criteria.

A recent meta-analysis has shown that pharmacologic prophylaxis is effective in reducing fatal PE, symptomatic PE, and symptomatic DVT by more than 50% with no increase in major bleeding compared with placebo in general medical patients. There is no effect on all-cause mortality, and the number needed to treat to prevent one symptomatic PE is high (more than 300). However, due to the large number of at-risk hospitalized medical patients, thromboprophylaxis still provides an opportunity to reduce morbidity in a significant number of patients. There does not appear to be a difference in bleeding rates or VTE rates between low-dose unfractionated heparin (LDUH), low-molecular-weight heparin (LMWH), and fondaparinux. Meta-analyses have suggested that LMWH is superior to LDUH (both twice- and thrice-daily regimens) in high-risk populations, though included studies included studies enrolled heterogeneous groups, and did not always differentiate symptomatic and asymptomatic VTE.

In patients with an increased risk for bleeding (Table 252-2), physicians commonly choose mechanical methods of thromboprophylaxis to increase venous flow and reduce stasis: graduated compression stockings (GCS), intermittent pneumatic compression devices (IPC), and the venous foot pump.

There are no randomized clinical trials evaluating mechanical thromboprophylaxis in general medical patients. In a placebo-controlled trial assessing GCS for DVT prophylaxis in patients with acute ischemic stroke, GCS did not provide additional therapeutic benefit for DVT prevention (10.0% vs 10.5%) but conferred a fourfold increased risk for skin breaks, ulceration, and blisters. Mechanical thromboprophylaxis should be considered inferior to pharmacologic prophylaxis in preventing VTE, and should only be used as an adjunct to pharmacologic prophylaxis, or as an option in patients with an unacceptably high-bleeding risk in whom anticoagulation is contraindicated (eg, gastrointestinal or intracranial hemorrhage). All patients with an increased risk for bleeding should be followed closely. If the bleeding risk decreases to an acceptable level, pharmacologic prophylaxis should be started as soon as possible.

VTE PROPHYLAXIS IN CANCER PATIENTS

Cancer confers a sixfold increased risk of VTE, and active cancer accounts for 20% of all new VTE events in the community. Conventional chemotherapies, erythropoietin-stimulating agents, angiogenesis inhibitors, and hormonal therapies (including selective estrogen receptor modulators and aromatase inhibitors) also increase the risk of VTE. Once patients with cancer develop VTE, it can be difficult to treat them due to intercurrent illnesses, the need for invasive procedures, thrombocytopenia, and high VTE recurrence rates. Therefore, it is essential to carefully risk stratify patients and consider thromboprophylaxis when patients with cancer are admitted to hospital. Cancer patients with an acute medical illness, either cancer-related or otherwise, or who are bedridden should receive thromboprophylaxis in accordance with the recommendations for general medical patients. VTE prophylaxis is not routinely recommended in cancer patients who are ambulatory, however it is indicated for many patients with multiple myeloma treated with thalidomide- or lenalidomide-containing regimens, and in individuals with a high score on the well-validated Khorana score. (This latter group includes individuals with stomach or pancreatic cancer, high prechemotherapy platelet or WBC counts, a high BMI, and individuals on erythropoeisis-stimulating agents.)

VTE PROPHYLAXIS IN CRITICALLY ILL PATIENTS?

Patients in the intensive care unit are particularly heterogeneous, both in terms of thrombosis risk and in terms of bleeding risk. The rates of asymptomatic DVT (detected by ultrasound or venography) reflect this heterogeneity, ranging from less than 10% to nearly 100%. There are only two randomized placebo-controlled trials of thromboprophylaxis in patients in the intensive care setting, one comparing LDUH to placebo, and the other comparing LMWH to placebo. Both significantly reduced the rate of DVT detected on routine screening, yielding a relative risk reduction of approximately 50%. More recently, the PROTECT trial randomized 3764 critically ill patients to LDUH or the LMWH dalteparin. Dalteparin was associated with a reduction in PE (1% vs 2%; hazard ratio 0.5) but had no effect on DVT, bleeding, mortality, or cost.

All patients admitted to the intensive care unit should have a routine assessment of risk for thrombosis and bleeding. Patients who have an acceptable bleeding risk should receive LMWH, while those considered at high risk for bleeding should receive GCS and/or IPC. This latter group should be reassessed daily, and if the bleeding risk decreases to an acceptable level, pharmacologic prophylaxis should commence as soon as possible.

THE RISKS ASSOCIATED WITH VENOUS THROMBOEMBOLISM PROPHYLAXIS

The major risk of pharmacologic VTE prophylaxis is bleeding. In medical patients, bleeding may occur at the anticoagulant injection site (eg, rectus sheath hematoma) or at a remote site (eg, from an occult peptic ulcer that is predisposed to bleeding). The risk for fatal bleeding associated with anticoagulant prophylaxis is 0.02% to 0.5%, or 32% higher than in patients who do not receive prophylaxis. Prospective trials have shown that this difference is not statistically significant, but this finding may be because the studies were underpowered to show a difference in bleeding risk. Bleeding complications can be minimized by carefully assessing and mitigating patients’ individual bleeding risks before starting anticoagulants (Table 252-2). Patients should also be assessed regularly for signs and symptoms of bleeding while they are on anticoagulants.

The other important risk associated with pharmacologic VTE prophylaxis is heparin-induced thrombocytopenia (HIT). This rare but serious complication of heparin-derived anticoagulants is caused by antibody formation to the heparin-derived anticoagulant and an antigen on the patient’s platelets. It is strongly associated with venous and to a lesser extent, arterial thrombosis, and can have devastating consequences. HIT is more common with LDUH compared to LMWH, and more common in surgical versus medical patients. In medical patients who received anticoagulant prophylaxis with LDUH, the risk for HIT is 1.4%. Data are lacking as to the risk for HIT in medical patients who are receiving LMWH but it is probably one-tenth the risk observed with LDUH. Irrespective of the anticoagulant administered, platelet counts should be monitored serially for all patients on anticoagulants and physicians should be watchful for signs and symptoms of arterial or venous thromboembolism that can herald the development of HIT.

Overall, the risks of major bleeding and HIT are small in hospitalized medical patients, and are far outweighed by the benefits of pharmacologic VTE prophylaxis.

PRACTICAL MANAGEMENT OF VTE PROPHYLAXIS

PHARMACOLOGIC PROPHYLAXIS IN THE SETTING OF RENAL INSUFFICIENCY

Renal impairment is a significant challenge in the administration of venous thromboprophylaxis. Drugs like LMWH and fondaparinux are primarily cleared by the kidney, and in the setting of reduced renal function, these drugs can accumulate and increase bleeding risk. Ongoing studies are evaluating the bioaccumulation of these anticoagulants and the clinical consequences. At this time, only prophylactic dose dalteparin has been shown not to bioaccumulate when the creatinine clearance is less than 30 mL/min. Before prescribing any renally cleared anticoagulant, physicians should measure a patient’s serum creatinine and formally calculate the creatinine clearance (which depends on age, weight, and sex). If renal function is found to be impaired, an anticoagulant that does not bioaccumulate should be chosen. If this is not possible, the dose should be lowered and/or the anti-Xa level should be monitored. In all cases, the manufacturer’s suggested dosage guidelines in the presence of renal impairment should be followed.

CONSIDERATIONS UPON DISCHARGE FROM HOSPITAL

There is no evidence that hospitalized medical patients benefit from routine DVT screening using venous ultrasound or venography. Therefore, routine DVT screening at the time of hospital discharge cannot be recommended. The risk-to-benefit ratio of extended out-of-hospital prophylaxis is also unclear. Medical patients with chronic illness or who are to be transferred to a long-term care facility may have an ongoing increased thrombosis risk, making extended-duration prophylaxis an appealing option. Three randomized trials have examined this issue: EXCLAIM (which compared 5 weeks vs 10 days of VTE prophylaxis with enoxaparin, 40 mg once daily), ADOPT (which compared 6-14 days of 40 mg enoxaparin daily vs 30 days of 2.5 mg apixaban twice daily), and MAGELLAN (which compared 10 days of VTE prophylaxis with 40 mg enoxaparin daily followed by a placebo vs extended prophylaxis for 35 ± 4 days with 10 mg rivaroxaban daily). EXCLAIM did show increased efficacy of extended-duration prophylaxis. However, benefits were restricted to only some groups (eg, those >75 years of age, those with recently reduced mobility). Further, the extended-duration arm in EXCLAIM had an increased risk of major bleeding. ADOPT found no difference between its study arms, though extended-duration prophylaxis increased the risk of major bleeding. MAGELLAN found that extended-duration prophylaxis reduced the risk of VTE (both symptomatic and asymptomatic), though it again conferred an increased risk of major bleeding. A meta-analysis of these three studies concluded that routine administration of postdischarge VTE prophylaxis is not likely to make a clinically meaningful impact for hospitalized medical patients, and could cause harm. At this time, extended-duration prophylaxis is not recommended in medical patients. However, all patients should be educated about the signs and symptoms of VTE at the time of discharge and be instructed to seek urgent medical care if thrombosis is suspected.

QUALITY IMPROVEMENT INITIATIVES TO OPTIMIZE VTE PROPHYLAXIS

A significant gap remains between evidence for VTE prophylaxis and clinical practice in hospitalized medical patients. A recent international registry demonstrated that in a population of 15,156 hospitalize medical patients only 50% received any form of prophylaxis. A multicenter Canadian chart audit determined that 90% of acutely ill medical patients were eligible for some form of prophylaxis, while only 23% received it. What is rather astonishing is that only 16% of patients received appropriate prophylaxis.

Medical patients have repeatedly been shown to have the poorest rates of VTE prophylaxis among all hospitalized patients. Yet since 1998, the American College of Chest Physicians has given anticoagulant prophylaxis in at-risk hospitalized medical patients a Grade 1A recommendation (their highest level). The American College of Physicians also strongly advocates for risk assessment and anticoagulant prophylaxis in this group. Why is VTE prophylaxis underused in hospitalized medical patients? It is likely because medical patients are more heterogeneous than their counterparts on surgical wards. Their need for prophylaxis is not driven by a standardized type of surgery, but by their underlying diseases, their mobility status, and their reason for hospitalization. Health care providers may be unclear about the indications and contraindications for anticoagulant prophylaxis in a given patient.

Many organizations have also identified VTE as a major patient safety concern, including the US Department of Health and Human Services, the World Health Organization, the World Alliance for Patient Safety, and the International Alliance of Patients’ Organizations. VTE prophylaxis has also been highlighted as an important feature of hospital accreditation commissions and quality improvement campaigns worldwide. Evidence supports a multicomponent strategy to optimize VTE prophylaxis in hospitalized medical patients, including formal hospital policy, standardized preprinted order sets, computer decision support systems, electronic and human alerts, and periodic audit and feedback. Hospitalists are in an ideal position to champion appropriate VTE prophylaxis in hospitalized medical patients, at a local, national, and international level.

SUGGESTED READINGS

INTRODUCTION

In 800 BC, Susruta, an Indian healer wrote about a patient with “a swollen and painful leg, which was difficult to treat.” Centuries later, Virchow, a Prussian physician, coined the term “embolism” after discovering the relationship between a blood clot that formed within a blood vessel (thrombus), and a blood clot that breaks loose and travels through the bloodstream to occlude the pulmonary vessels (embolus). The concept of venous thromboembolism was born from these early descriptions and today it remains one of the most important health problems in Europe and North America and is the third leading cause of vascular death after myocardial infarction and stroke.

The risk of venous thromboembolism (VTE) increases by approximately twofold per decade of age, rising from an annual incidence of 30/100,000 at 40 years of age, to 90/100,000 at 60 years, and 260/100,000 at 80 years. Approximately half of patients with untreated, symptomatic proximal deep vein thrombosis (DVT) will develop symptomatic pulmonary embolism (PE), and about 10% of symptomatic PE are fatal within an hour of onset. Left untreated, one-third of patients with initially nonfatal PE will have a fatal recurrence, generally within a few weeks of the original event. Even with optimal treatment, about 5% of patients with PE will die from fatal PE, and about 25% with proximal DVT will develop post-thrombotic syndrome, a chronic condition that is debilitating for patients.

Venous thromboembolism is now recognized as the leading cause of preventable death in hospitalized patients. Almost all hospitalized patients have one or more risk factors for VTE and 40% will have three or more risk factors. VTE prophylaxis (addressed in Chapters 56, 65, and 252) forms the cornerstone for preventing these deaths. In addition, although 75% of venous thromboembolic events are diagnosed in the outpatient setting, about half of all episodes of VTE are associated with recent surgery or hospitalization. These findings stress the importance of having a low threshold to perform diagnostic testing in patients who present with signs and symptoms compatible with VTE within 3 months of hospitalization.

Therefore, VTE is both an acute and a chronic disease that causes substantial patient morbidity and mortality, and it is a major burden on the health care system. Costs for VTE include not only the expense of initial diagnosis and treatment, but also the cost of the complications of VTE (ie, post-thrombotic syndrome, venous ulceration, chronic thromboembolic pulmonary hypertension, recurrent VTE) and its treatment (ie, bleeding). It is currently estimated that VTE costs the US health care system $1.5 billion/y.

NOMENCLATURE

Proximal DVT is defined as a DVT that involves the popliteal vein or more proximal veins of the leg (most also involve the calf veins).

Distal DVT is defined as a DVT that is confined to the calf veins.

PATHOPHYSIOLOGY

Virchow’s triad for the pathogenesis of thrombosis is as relevant today as it was when it was originally described in the 18th century: venous stasis, vessel wall damage, and hypercoagulability. A summary of common risk factors for VTE is given in Table 253-1.

VENOUS STASIS

Deep-vein thrombosis is more likely to occur in the paralyzed leg following stroke, and in the left leg during pregnancy because of extrinsic compression of the left iliac vein by the pregnant uterus and the right common iliac artery.

VESSEL WALL DAMAGE

Manipulation during surgery (eg, hip replacement), iatrogenic injury, and use of indwelling venous catheters all markedly increase the risk of DVT.

HYPERCOAGULABILITY

Inherited or acquired changes in the balance of naturally occurring coagulation and fibrinolytic factors and their inhibitors predispose to thrombosis. The inherited hypercoagulable conditions that are considered strong risk factors for thrombosis; antithrombin deficiency, protein C deficiency, and protein S deficiency, are rare (<1% prevalence). Conversely, the most common inherited hypercoagulable conditions, activated protein C resistance caused by the Factor V Leiden mutation (5% prevalence in Caucasians), and the prothrombin gene mutation that leads to a 25% increase in prothrombin levels (2% prevalence), are weak risk factors for thrombosis. If VTE is suspected or confirmed in inpatients or shortly after discharge from the hospital, it is important to consider the possibility of heparin-induced thrombocytopenia (HIT), another prothrombotic state.

DOES THIS PATIENT HAVE A DVT?

Edema, pain, tenderness, and erythema are signs and symptoms of DVT, but they are also commonly found in patients who do not have DVT (ie, nonspecific). Only 15% of ambulatory patients who are suspected of having DVT will have this diagnosis confirmed on objective testing. Alternate causes for these clinical features include recent leg surgery, muscle injury, Baker cyst, cellulitis, extrinsic compression of veins, anasarca or low albumin states, and venous insufficiency.

Hospitalized patients are more likely than ambulatory patients to have DVT confirmed when it is suspected (prevalence of 30%-40%). However, hospitalized patients are also more likely to have asymptomatic DVT (in a prophylaxis study of medical patients that routinely screened for asymptomatic DVT, only 6% of DVT were symptomatic). This finding stresses the importance of preventing VTE instead of relying on clinical surveillance to detect and treat it early.

Failure to diagnose DVT exposes patients to the risk of fatal PE; however, inappropriate use of anticoagulant therapy exposes patients to the risk of fatal bleeding. Because clinical assessment alone is unreliable, objective testing to confirm the diagnosis must always be performed when DVT is suspected.

CLINICAL ASSESSMENT

For the reasons outlined above, clinical assessment alone is an unreliable test for diagnosing DVT. However, clinical prediction rules have been developed that can help to stratify patients as having a low (5% prevalence), moderate (25% prevalence), or high (60% prevalence) probability of DVT (Table 253-2).

Hospitalized patients are less likely to have a low probability score. Classification of pretest probability is helpful when used in combination with other diagnostic tests (Figure 253-1).

If there is a high risk clinical suspicion for DVT (or PE) and patients are not at high risk for bleeding, it is recommended that anticoagulant therapy is started before diagnostic testing is performed. Treatment can be deferred if diagnostic testing will be performed within 4 hours in patients with a moderate and within 24 hours in patients with a low clinical suspicion for DVT (or PE).

IMAGING TESTS FOR DVT

Venography

Venography is the reference standard test for DVT, but it is rarely performed today because it is invasive (painful), technically difficult, and exposes the patient to the risks of contrast dye (eg, nephrotoxicity, anaphylaxis) and radiation. An intraluminal filling defect (ie, section of a vein that remains dark when surrounded by white contrast dye) seen on at least two views is considered diagnostic for DVT.

Compression venous ultrasound

Ultrasonography is the first-line imaging test used to diagnose or exclude DVT. When pressure is applied to the proximal veins with an ultrasound probe, the veins should fully compress. If they do not, acute or chronic DVT is present. In symptomatic patients, when the ultrasound probe is applied to the proximal veins, the sensitivity of ultrasound for proximal DVT is 97% and the specificity is 94% when compared with venography (for both inpatients and outpatients). The sensitivity and specificity of compression ultrasound is lower in the calf veins, and the clinical significance of isolated calf vein thrombosis is controversial. For these reasons, many centers do not examine the calf veins with ultrasound; instead, if the initial assessment (eg, low clinical suspicion and negative ultrasound of the proximal veins) cannot exclude the presence of DVT, a follow-up ultrasound of the proximal veins is done after a week to exclude the possibility of calf vein thrombosis with early extension into the proximal veins. If whole leg ultrasound is performed and is negative, repeat ultrasound scanning is not necessary. Compression venous ultrasound is frequently combined with assessment of Doppler flow; however, change in Doppler flow alone is not a sensitive or specific test for DVT.

Limitations of ultrasound include the factors listed below.

• Operator dependent in the calf veins

• Limbs with casts that cannot be removed

• Morbid obesity

• Massive edema

• Isolated pelvic DVT harder to detect

D-dimer blood testing

D-dimer is formed when cross-linked fibrin is broken down by plasmin. Normal levels can help to exclude DVT, but elevated D-dimer levels are common with other conditions (eg, malignancy, disseminated intravascular coagulation, acute infection, normal pregnancy, renal disease, cardiovascular disease). D-dimer levels are also elevated after surgery and may take up to 50 days to return to baseline, depending on the type of surgery. Consequently, a positive D-dimer is NEVER diagnostic for DVT! A negative D-dimer can still be used to exclude DVT in a hospitalized patient with a low clinical probability for DVT, but negative results in this patient population are so uncommon (<10%) that the utility of the test is very poor.

DOES THIS PATIENT HAVE RECURRENT DVT?

Approximately 30% of patients with DVT will experience recurrent DVT during the 10 years following the initial diagnosis. Unfortunately, diagnosing recurrent DVT is much more difficult than diagnosing a first DVT. To begin with, the signs and symptoms experienced by patients with recurrent DVT (ie, increased edema and/or pain) are also seen in patients with post-thrombotic syndrome, a chronic condition that affects approximately 25% of patients with previous DVT. Secondly, diagnostic tests for recurrent DVT are not as reliable as they are for first acute DVT (Figure 253-2).

CLINICAL ASSESSMENT

Clinical prediction rules for diagnosing recurrent DVT are less well developed and validated than for a first episode of DVT and they have not been widely used.

IMAGING TESTS FOR RECURRENT DVT

Venography

As with first acute DVT, this method is considered the reference standard test for recurrent DVT. However, veins previously affected by DVT may not fill with contrast dye, resulting in nondiagnostic findings.

Compression venous ultrasound

In patients with a recent ultrasound that demonstrated (1) limited extent of the previous DVT or (2) complete resolution of the previous DVT, the presence a new noncompressible segment of the proximal veins (eg, popliteal or common femoral vein) on ultrasound examination is diagnostic for recurrent DVT. However, it is important to remember that 50% of patients with first acute DVT will still have incomplete compressibility on ultrasound examination 1 year after diagnosis despite adequate anticoagulant therapy. Consequently, without clear evidence of a new noncompressible segment, it can be difficult to tell the difference between residual chronic DVT and new acute DVT. It can be useful, therefore, to perform an ultrasound when the decision is taken to stop anticoagulation in a patient who initially had an extensive DVT. This acts as a baseline scan to which future scans can be compared.

If a proximal US is abnormal but there is no new noncompressible segment, it has been proposed that an increase in residual vein diameter of greater than 4 mm (when a vein is compressed with an ultrasound probe) is diagnostic for recurrence. However, a recent ultrasound (prior to the episode of suspected recurrence) is required to make this comparison, and there is evidence that the reproducibility of the residual diameter measurement is only moderate. Other ultrasound features such as Doppler flow and thrombus echogenicity have not been validated as reliable methods for diagnosing recurrent DVT.

D-dimer blood testing

Although less studied than in patients with a suspected first DVT, recent clinical trials indicate that a negative D-dimer result can be used to help exclude recurrent DVT.

DOES THIS PATIENT HAVE PE?

The commonest symptoms of PE are shortness of breath and fatigue. Patients with PE may also present with pleuritic chest pain, palpitations, hemoptysis, and syncope. Signs of PE include tachycardia, tachypnea, accentuated pulmonic heart sound, and S1Q3T3 and evidence of right heart strain on ECG. As with DVT, clinical signs and symptoms are nonspecific, and objective diagnostic testing must always be performed when PE is suspected.

CLINICAL ASSESSMENT

As with suspected DVT, clinical assessment can be used in combination with other tests to diagnose PE (Table 253-3).

IMAGING TESTS FOR PE

Pulmonary angiography

This method is the reference standard test for diagnosing PE. However, it suffers from the same limitations as venography and it is now very rarely used.

Computed tomographic pulmonary angiography (CTPA)

Computed tomographic pulmonary angiography (also known as spiral or helical CT) is the current first-line imaging test for PE. Thrombus in the pulmonary arteries is outlined by radiologic contrast. CTPA has the additional major advantage of detecting alternate causes for symptoms in about one-third of patients with suspected PE (eg, pneumonia, lung mass). Less than 2% of patients with a negative CTPA who are not treated with anticoagulant therapy will return with symptomatic VTE during 3 months of follow-up.

Computed tomographic pulmonary angiography has been shown to have a sensitivity of 83% and a specificity of 96% for PE. However, accuracy varies according to the size of the largest pulmonary artery involved. The probability that an intraluminal filling defect is due to a PE has been reported as 97% for defects in the main or lobar artery, 68% for segmental arteries, and 25% for subsegmental arteries (which means that as many as 75% of abnormalities seen in subsegmental arteries are not due to PE). Accuracy of CTPA is also influenced by clinical assessment of PE; the higher the clinical probability of PE, the more likely the defect that is seen on CTPA is actually due to PE (Table 253-4).

Limitations include the factors listed below.

• Requires use of contrast dye

• Exposure to radiation

• Subject to technical difficulties (approximately 6% of CTPA are nondiagnostic)

• Expense

Ventilation-perfusion lung scanning (V/Q)

This was the first-line test used to diagnose PE before the advent of CTPA, and V/Q scanning is still used, particularly when CTPA is contraindicated (ie, patients with renal failure, young women due to a concern about breast cancer induced by radiation exposure). Perfusion defects are nonspecific; only about 33% of patients with perfusion defects have PE. The probability that a perfusion defect is due to PE increases with the size and number of defects, and if they are mismatched. A mismatch refers to a perfusion defect that is not associated with a corresponding defect on the ventilation scan. The V/Q scan probability of PE should be matched with the clinical probability of PE. A high probability lung scan can diagnose PE in patients with moderate or high clinical probability, and a low probability lung scan can exclude PE in a patient with low clinical probability of PE. Lung scan findings are highly age dependent with a relatively high proportion of normal scans in younger patients, including pregnant women, and a low proportion in those with other acute and chronic cardiorespiratory conditions (Table 253-5).

Limitations include the factors listed below:

• Large proportion of nondiagnostic scans.

• Inability to identify alternative causes for symptoms.

SPECT

In many hospitals, three-dimensional SPECT (single-photon emission CT) has replaced planar V/Q scanning. Unlike V/Q, there are no standardized criteria for reporting SPECT. SPECT scans are less likely to be reported as nondiagnostic, with the majority of scans reporting PE present or absent. As yet, there are no large diagnostic studies reporting on the safety of diagnosing and excluding PE with SPECT.

D-dimer blood testing

As with suspected DVT, a negative D-dimer result can exclude PE (on its own, if it is a very sensitive assay or in combination with other tests, if it is less sensitive) (Figure 253-3).

What if the tests I ordered to exclude PE were nondiagnostic?

If initial testing is nondiagnostic (eg, ventilation-perfusion scan), there is still a substantial chance that your patient could have a PE and further investigations are required. One approach to management of such patients is to perform ultrasonography and withhold anticoagulation if there is no DVT of the proximal veins. Ultrasonography of the proximal veins is then repeated after 1 and 2 weeks to exclude evolving VTE. See algorithm Figure 253-3.

TREATMENT OF VENOUS THROMBOEMBOLISM

The foundation of treatment of VTE is anticoagulant therapy. The objectives of anticoagulant therapy are: (1) to prevent extension and potentially fatal embolization of the initial thrombus and (2) to prevent recurrent VTE.

TRIAGE AND HOSPITAL ADMISSION

Patients who are hemodynamically stable with a low bleeding risk and normal renal function, and who are likely to be compliant with anticoagulant therapy, can be safely treated as outpatients. Patients with DVT and severe intractable pain or severe swelling and poor leg perfusion should be admitted to hospital for initiation of anticoagulant treatment, and assessment of the need for catheter-directed thrombolysis. Patients with PE and severe symptoms or abnormal vital signs should be admitted to the hospital, and those with signs of hemodynamic compromise (eg, systolic blood pressure <90 mm Hg) should be considered for thrombolytic therapy (discussed later). Validated prognostic scores, such as the Pulmonary Embolism Severity Index, are available to help physicians select which patients with PE can be treated as outpatients.

TREATMENT OF ACUTE VENOUS THROMBOEMBOLISM

The first step in treating acute VTE is preventing further thrombus formation by starting an agent that rapidly inhibits thrombin. The agents approved for initial treatment of acute VTE in Europe and North America are low-molecular-weight heparin (LMWH), fondaparinux, heparin, rivaroxaban, and apixaban (Table 253-6).

Patients with VTE who have heparin-induced thrombocytopenia (HIT) or a past history of HIT should receive danaparoid, argatroban, or fondaparinux (off-label) instead of heparin or low-molecular-weight heparin.

Rivaroxaban and apixaban have simplified the approach to anticoagulation in acute venous thrombosis. Both drugs are given in a higher dose for the initial treatment period, followed by dose reduction (see Table 253-6). They have the advantage of the oral route, simplified dosing, and freedom from dietary interactions (compared to warfarin). In contrast, dabigatran therapy is delayed until after 5 to 7 days of heparin therapy.

Not all patients are suitable for treatment with rivaroxaban, apixaban, or dabigatran. The group of patients who should not receive new oral anticoagulants includes those with renal impairment and a creatinine clearance <30 mL/min, hepatic impairment, and patients who are treated with strongly interacting drugs (see Table 253-6). Pregnant and breast feeding women should not be treated with the new oral anticoagulants.

When warfarin is the chosen oral anticoagulant, a heparin should be given for a minimum of 5 days and until the INR is greater than or equal to 2.0 on two consecutive measurements. A vitamin K antagonist (eg, warfarin) with a target INR of 2.0 to 3.0 is typically started at the same time as the parenteral antithrombotic agent. Dosing algorithms for VKAs are available. Lower VKA maintenance doses are required in older patients, women, those with impaired nutrition, and vitamin K deficiency. Optimal VKA management requires a systematic approach to obtaining INR measurements, adjusting VKA dose, and communicating these instructions to patients. Anticoagulation clinics, and use of computer programs to schedule appointments, adjust VKA dose and maintain records, can facilitate this process.

CATHETER-DIRECTED THROMBOLYTIC THERAPY FOR DVT

Thrombolytic therapy administered directly into the thrombus can achieve thrombus removal, improve acute symptoms, and reduce the risk of developing the post-thrombotic syndrome. Catheter-directed thrombolysis can be combined with mechanical thrombus disruption to further increase thrombus removal and shorten the procedure. It can be used for patients with extensive DVT (eg, involves the iliofemoral vein) of recent onset (eg, symptoms ≤14 days) that is associated with severe symptoms, provided the patient does not have risk factors for bleeding, there is available expertise, and it is the patient’s preference to receive such therapy.

SYSTEMIC AND LOCAL THROMBOLYSIS FOR PULMONARY EMBOLISM

Systemic thrombolysis is associated with a 2% risk of intracranial bleeding, and a 10% risk of major bleeding. Therefore it is generally reserved for patients with PE and a systolic blood pressure less than 90 mm Hg. Catheter-directed thrombolysis, which uses a lower dose of thrombolytic drug, may have a more rapid action and lower risk of bleeding; therefore, in centers with the expertise, catheter-directed thrombolysis may be preferred to systemic thrombolytic therapy in hypotensive patients with PE who are deteriorating or who have a high risk for bleeding.

BLEEDING IN PATIENTS WITH ACUTE VENOUS THROMBOEMBOLISM

Management of a patient with acute VTE and bleeding is difficult. The first step is to ensure the patient is hemodynamically stable. After that, the following questions can help to guide management: How severe is the bleeding? Which anticoagulant is the patient receiving (and when was the last dose given)? How acute is the VTE? Can the source of the bleeding be treated?

Moderate-to-severe bleeding

If the bleeding is moderate to severe, the anticoagulant should be reversed (see Table 235-7) (if possible), the patient resuscitated and the source of the bleeding treated. If the source of the bleeding cannot be treated, and the VTE is less than 1 month old, consideration should be given to inserting a permanent or removable inferior vena caval filter. If the source of the bleeding cannot be treated, and the VTE is older than 3 months, anticoagulants should be permanently discontinued. For reasons outlined later in this chapter, inserting an inferior vena caval filter is not recommended unless the patient has acute VTE (<1 month old).

Mild bleeding

If the bleeding is mild and likely to be transient (eg, hematuria in a patient with a urinary tract infection), the anticoagulant should be held for 24 to 48 hours and resumed once the bleeding has stopped. If the bleeding continues, consideration should be given to fully reversing the anticoagulant, and the source of the bleeding should be treated. If the bleeding resolves, anticoagulants should be reintroduced with careful monitoring and/or lower doses. If bleeding persists, and the VTE is older than 3 months, anticoagulants should be permanently discontinued.

Inferior vena caval filters

No randomized trial or prospective cohort study has evaluated inferior vena caval filters as sole therapy in patients with DVT (ie, without concurrent anticoagulant therapy). Permanent inferior vena caval filter insertion as an adjunct to anticoagulant therapy has been evaluated in a single large-randomized-controlled trial of patients with acute DVT who were considered to be at high risk for PE. The findings of this study, which were reported after 2 years and 8 years of follow-up, indicated that filters decreased the rate of PE, increased the rate of DVT, did not change the overall rate of VTE, and had no apparent influence on mortality. Indirectly, this study supports the use of vena caval filters to prevent PE in patients with acute DVT and/or PE who cannot be anticoagulated (ie, actively bleeding), but does not support more liberal use of filters. More recently, a large randomized trial found that placement of a retrievable inferior vena caval filter as an adjunct to anticoagulation in patients with acute PE and a high risk of recurrence did not reduce recurrent PE at 3 months. Patients who have a vena caval filter inserted should receive anticoagulant therapy if and when it becomes safe to do so, and for the same duration as for similar patients who do not have a filter.

DURATION OF ANTICOAGULANT TREATMENT OF VENOUS THROMBOEMBOLISM

Optimal duration of anticoagulant treatment for VTE is based on three factors: (1) the increase in the risk of recurrent VTE if anticoagulants are stopped, (2) the increase in the risk of bleeding on anticoagulant therapy, and (3) patient preference. The factors that determine the risk of recurrence are outlined below, while the risk of anticoagulant-related bleeding is addressed in the following section.

First, the risk of recurrence depends on whether the acute episode of VTE has been adequately treated. Studies evaluating different durations of anticoagulant therapy have shown that all patients with VTE should receive a minimum of 3 months of anticoagulant therapy. Patients who receive less than 3 months of anticoagulant therapy have double the risk of recurrent VTE during 1 to 2 years of follow-up.

Second, the risk of recurrence after 3 or more months of treatment is dependent on the patient’s intrinsic risk of having a new episode of VTE. The risk of recurrence is highest in patients who have no risk factors for VTE (ie, idiopathic or unprovoked) or have a VTE provoked by a strong persistent risk factor (eg, active malignancy), and lowest in patients with a VTE provoked by a strong transient risk factor (eg, surgery). The risk of recurrence in patients with VTE provoked by a transient risk factor is one-third of the risk of recurrence for patients with unprovoked VTE. Extending the duration of anticoagulant therapy for unprovoked VTE beyond 3 months (or 6 months if VTE was very severe) protects the patient from recurrence while they are taking the anticoagulant, but does not reduce the risk of recurrence once the anticoagulant is discontinued (Table 253-8).

Third, among patients with unprovoked VTE, the risk of recurrence after stopping anticoagulation is 10% in the first year, 30% in 5 years and 50% by 10 years. Recent studies have identified men as having a higher risk of recurrence (15% in the first year). Patients who have an elevated D-dimer level after stopping anticoagulation also have a risk of recurrence that is about twice that of patients with normal D-dimer levels.

Two other factors connected with a patient’s intrinsic risk of VTE that currently influence the duration of anticoagulant therapy include the number of previous episodes of VTE, and the presence of active malignancy. Patients with a second episode of unprovoked VTE have a 50% increased risk of recurrence and, therefore, should be considered for indefinite anticoagulant therapy. Patients with VTE and active cancer have about a threefold higher rate of recurrent VTE than patients with VTE who do not have cancer and, therefore, are usually maintained on anticoagulant therapy. The risk of recurrent VTE during and after anticoagulant therapy increases with more advanced stages of disease and with antineoplastic treatments such as chemotherapy.

Extended treatment with rivaroxaban, apixaban, or dabigatran is as effective as warfarin therapy at preventing recurrence, and is associated with less bleeding. LMWH is the preferred therapy for patients with VTE and because it is more effective at preventing VTE recurrence than warfarin. There are little data on using the newer anticoagulants in cancer patients, and chemotherapy may interact with these agents. Aspirin reduces recurrent VTE by up to a third, but it associated with bleeding and it is much less effective than anticoagulants (risk reduction of 80%-90%).

Inherited thrombophilias increase the risk of a first VTE, but their influence on the risk of recurrent VTE is minor or uncertain. The risk of recurrence is thought to be highest in patients with antithrombin deficiency, patients homozygous for the Factor V Leiden mutation, and patients who are double-heterozygotes for thrombophilic states (eg, patients heterozygous for both the Factor V Leiden mutation and the prothrombin gene mutation). However, these combinations are rare. The most common inherited hypercoaguable conditions, activated protein C-resistance caused by the Factor V Leiden mutation (5% prevalence), and the prothrombin gene mutation (2% prevalence), have not been shown to importantly increase the risk of recurrence. Overall, thrombophilia screens identify an abnormality in only about 30% of patients with unprovoked VTE. A negative thrombophilia screen does not exclude an inherited thrombophilia; it just excludes the thrombophilic abnormalities that we are able to test for at this time. Consequently, because the results of thrombophilia testing have very limited management implications, the pros and cons of thrombophilia screening should be carefully weighed before deciding to order these tests (Table 253-9).

Lupus anticoagulant and anticardiolipin antibodies (antiphospholipid antibodies) are acquired thrombophilic disorders, which may support indefinite anticoagulant therapy because of a higher risk of recurrent VTE and arterial events. However, testing for these disorders is not well standardized. It is important to be aware that antithrombin levels can be falsely low immediately following an acute VTE and that warfarin will interfere with testing for protein C deficiency, protein S deficiency and lupus anticoagulants. If thrombophilia testing is performed, it is generally better to defer this to the clinic rather than to test patients when they have acute VTE.

BLEEDING IN PATIENTS ON LONG-TERM ORAL ANTICOAGULANT THERAPY

Risk of bleeding on anticoagulants differs markedly among patients depending on the prevalence of risk factors (eg, age >75, previous gastrointestinal bleeding or stroke, renal failure, anemia, antiplatelet therapy, malignancy, poor anticoagulant control). Approximately 13% of episodes of major bleeding during the first 3 months of anticoagulant therapy are fatal and greater than 50% of intracranial bleeds are fatal. The risk of major bleeding in younger patients (eg, <60 years) who do not have risk factors for bleeding is about 1% per year. The risk of bleeding, and particularly intracranial bleeding, is lower with rivaroxaban, dabigatran, and apixaban than it is with warfarin. These new anticoagulants are contraindicated in patients with severe renal impairment. The risk of bleeding increases with the number and severity of risk factors for bleeding.

CONCLUSION

Given the challenges of the clinical diagnosis of VTE and the unpredictability of sudden fatal PE as the initial manifestation, appropriate VTE prophylaxis is required to minimize morbidity and mortality from hospital-acquired VTE. In contrast to the community setting, clinical prediction rules and D-dimer testing may be less likely to discriminate between hospitalized patients who have and those who do not have VTE because the majority of patients will have moderate or high pretest probability of VTE and will have other reasons for a positive D-dimer. Since clinical assessment alone is unreliable, objective testing to confirm the diagnosis must always be performed when VTE is suspected.

Almost all patients with VTE will require anticoagulation that extends beyond hospitalization. Clinicians need to ensure that care transitions include identification and communication with the receiving providers who will monitor anticoagulation. Discharge information includes drug, dosing (including last dose), timing of next dose and identification of who will advise patient of any dosage adjustment.

SUGGESTED READINGS

INTRODUCTION

Thrombosis is a major cause of morbidity, requires or prolongs hospital admission, and often necessitates long-term medical intervention to reduce the likelihood of recurrent events. Anticoagulants are the cornerstone for prevention and acute management of thrombosis. Despite their effectiveness in preventing thrombosis and reducing propagation, anticoagulants are often associated with adverse drug reactions and medication errors and remain a major target for health improvement efforts. While hospitalized patients are at risk for developing thrombosis, they are also at elevated risk for hemorrhagic complications resulting from invasive procedures, organ dysfunction, gastrointestinal stress-related mucosal damage, medication use, and concealed coagulopathies. Anticoagulation management requires thoughtful consideration in agent selection, route of administration, body weight, renal and hepatic function, and concomitant drug therapy to achieve optimal outcomes.

For over 50 years, unfractionated heparin (UFH) and vitamin K antagonists (VKAs) had been the primary anticoagulants prescribed in practice. A new group of oral anticoagulants has emerged in the marketplace with a wide range of indications, creating more options for clinicians, but also more challenges in drug and patient selection. This chapter focuses on the mechanisms of action, pharmacokinetics, pharmacodynamics, clinical indications, complications of therapy, and reversal options for antithrombotic pharmacotherapy in critically ill patients.

ANTICOAGULANT PHARMACOLOGY

Anticoagulant agents exert their therapeutic benefit by interfering with the coagulation cascade through receptor binding with circulating blood clotting factors. Of these clotting factors, thrombin (IIa) and activated factor X (Xa) are believed to play the most critical role in hemostasis. Anticoagulants ultimately attenuate thrombin generation, preventing the generation of insoluble fibrin, the final step in coagulation. Anticoagulants are administered in low doses for thromboprophylaxis and higher or therapeutic doses for acute thrombosis treatment.

Unfractionated heparin, low-molecular-weight heparin (LMWH), and fondaparinux contain a pentasaccharide sequence that binds to and potentiates the action of antithrombin (AT), an endogenous small protein molecule that inactivates several enzymes of the coagulation system (Figure 254-1). Fondaparinux is a synthetic analog of this naturally occurring pentasaccharide. The heparin-AT complex inactivates coagulation factors XIIa, IXa, XIa, Xa, and thrombin. Since these agents depend on the presence of AT for clotting factor inhibition, they are considered indirect anticoagulants. The active pentasaccharide sequence responsible for catalyzing AT is found on one-third and one-fifth of the chains of UFH and LMWH, respectively.

Vitamin K antagonists (eg, warfarin, acenocoumeral) inhibit the enzyme vitamin K epoxide reductase complex (VKORC), which converts vitamin K to an active form. In the absence of active vitamin K, hepatic production of factors thrombin, VII, IX, X, and the regulatory anticoagulant proteins C, S, and Z is reduced. Since warfarin does not directly inactivate functional clotting factors, it has a delayed onset of action.

The direct thrombin inhibitors (DTIs) (argatroban, bivalirudin, and dabigatran) and direct Xa inhibitors (rivaroxaban, apixaban, and edoxaban) have a specific mechanism of action. They bind directly to the surface of thrombin or Xa, and prevent further contributions to enzymatic activity and thrombus formation. The oral agents have collectively been referred to as nonvitamin K oral anticoagulants (NOACs), direct acting oral anticoagulants (DOACs), or target-­specific oral anticoagulants (TSOACs).

INDIRECT ACTING ANTICOAGULANTS—PARENTERAL AGENTS

UNFRACTIONATED HEPARIN

Pharmacodynamics and monitoring

Heparin is derived from porcine intestines and structurally consists of highly sulfated, linked disaccharide chains that vary in size and length. UFH is poorly absorbed orally; therefore intravenous (IV) or subcutaneous (SC) injections are the only options for administration. Due to erratic bioavailability, SC doses of UFH for therapeutic anticoagulation must be >30,000 units/d. Intravenously administered UFH readily binds to plasma proteins. These features contribute to UFH’s highly variable dose response and the need for frequent laboratory monitoring. UFH is eliminated from systemic circulation through two independent mechanisms: at low doses by binding to endothelial cells, macrophages, and circulating proteins ultimately leading to UFH depolymerization, and at higher doses UFH is eliminated through a slower, renal-mediated clearance. Large doses or prolonged UFH administration provides a disproportionate increase in both the intensity and duration of anticoagulant effect. The half-life of UFH with therapeutic IV doses is approximately 60 minutes.

The anticoagulant response to UFH is monitored using the activated partial thromboplastin time (aPTT), a measurement sensitive to the inhibitory effects of thrombin. The aPTT should be measured every 6 hours, and doses adjusted accordingly, until the patient maintains therapeutic levels. While the therapeutic range for the aPTT, 1.5 to 2.5 times the control value, has become widely accepted, the evidence supporting this range for predicting thrombotic or bleeding events is weak. Once two consecutive aPTT values are within therapeutic range, testing may be extended to once-daily. Weight-based dosing nomograms, comprised of a weight-based bolus dose and infusion rate with periodic monitoring utilizing aPTT, are recommended for treatment of thromboembolic disease (Table 254-1). UFH dosing nomograms differ between hospitals due to differences in thromboplastin reagents, calibration, and inter-laboratory standards in aPTT measurements. This has led to assessments of alternative monitoring strategies. The functional heparin assay, also known as the anti-Xa assay, has been promoted as a more favorable marker for monitoring a patient’s response to UFH. Unlike aPTT, this assay is insensitive to factors other than UFH (eg, concomitant warfarin use, sodium citrate in collection tubes, interference from the presence of lupus anticoagulants, elevated factor VIII activity, and liver disease). Weight-based UFH protocols utilizing anti-Xa levels to monitor UFH are now in use in many hospitals.

Medical uses

Medical indications for UFH include treatment of ACS, treatment or prevention of venous thromboembolism (VTE), and bridge therapy to chronic oral anticoagulation for atrial fibrillation, cardioversion, or invasive surgical procedure (Table 252-2). Due to UFH’s short half-life and reversibility, it remains the best anticoagulant option in patients with bleeding risk or organ dysfunction. When used for thromboprophylaxis in medical patients, three times daily UFH dosing provides better efficacy in reducing VTE events compared to twice daily dosing, but generates more major bleeding episodes.

Complications of therapy and reversal

The major adverse drug events associated with UFH therapy include bleeding (major bleeding, 0%-7%; fatal bleeding, 0%-3%), heparin-induced thrombocytopenia (HIT) (1%-5%), and osteoporosis (2%-3% risk of vertebral fracture with less than 1 month of treatment). Hemorrhagic episodes are associated with anticoagulation intensity, route of administration (continuous infusions are associated with lower rates), and concomitant use of P2Y12 and GP IIb/IIIa platelet inhibitors, aspirin, or fibrinolytic agents. Patient-specific risk factors for bleeding include age, gender, renal insufficiency, low body weight, and excessive alcohol consumption.

Treatment of UFH-related bleeding includes protamine sulfate administration, transfusion, and supportive care. Protamine binds to UFH to form a stable salt, rendering it pharmacologically inactive. Protamine dosing is dependent on timing of the last UFH dose. For immediate reversal (<30 minutes since the last heparin dose), 1 mg of protamine is administered for every 100 units of UFH and a follow-up aPTT can evaluate the reversal response. When UFH is given as a continuous IV infusion, only UFH delivered during the preceding 2 to 2.5 hours should be included in the calculation to determine the protamine dose. If the dose of UFH is unknown, the maximal tolerated protamine dose of 50 mg may be slowly administered followed by serial measurements of aPTT. Protamine is cleared rapidly from the plasma; therefore, repeated doses may be necessary to prevent rebound UFH anticoagulant effects. Adverse reactions, such as systemic hypotension, pulmonary hypertension, and bradycardia, are common. However, reaction frequency and severity may be reduced by slowly administering protamine over 1 to 3 minutes. Allergic responses to protamine are more common in patients who have been previously exposed to the drug, but patients may be pretreated with corticosteroids and antihistamines.

Heparin-induced thrombocytopenia (HIT) is an immune-­mediated, hypercoagulable adverse event resulting from antibodies formed against the heparin-platelet factor 4 complex. The incidence ranges from 0.3% to 0.5% of patients. It is associated with thrombocytopenia and life-threatening thrombosis in antibody-positive patients. HIT and heparin-induced thrombocytopenia with thrombosis syndrome (HITTS) typically occur in patients who have been exposed to UFH or LMWH for 5 to 7 days, or even sooner in patients with prior exposure within the past 100 days. A 50% decrease in platelet count occurring 4 to 10 days after the initiation of UFH or LMWH therapy, formation of a new thrombus during therapy, and skin lesions at the injection site may be indicative of HIT. Platelet counts should be measured prior to the initiation of UFH or LMWH and monitored every other day for the first 4 to 10 days of therapy. The use of an alternative anticoagulant, such as direct thrombin inhibitors (argatroban or bivalirudin) or fondaparinux, must be used in patients with HIT. If the patient is also receiving warfarin, this medication should be discontinued. Warfarin inhibits production of the endogenous anticoagulants proteins C and S. Protein C inhibition occurs rapidly and may result in a transient hypercoagulable state. When coupled with HIT mediated thrombin production, patients at high risk for microvascular arterial and venous thrombosis and resultant limb gangrene (Figure 254-2).

LOW-MOLECULAR-WEIGHT HEPARINS

Pharmacodynamics and monitoring

Low-molecular weight heparins are produced through chemical or enzymatic depolymerization of UFH. While the LMWH molecules contain the active pentasaccharides that catalyzes AT inhibition of factor Xa, their smaller size reduces affinity for plasma proteins and cellular binding sites. This results in increased bioavailability after SC injection, elimination through renal clearance, and a longer half-life (17-21 hours) when compared to UFH (Table 254-3). LMWHs are administered in fixed doses for thromboprophylaxis or in total body weight-adjusted doses for therapeutic anticoagulation. Peak anticoagulant activity is rapid, occurring within 3 to 5 hours after injection. While laboratory monitoring is usually not necessary, anti-Xa monitoring is often performed in high-risk patient populations, such as pregnancy, obesity, and renal dysfunction. In these cases, peak anti-Xa plasma levels are drawn 4 hours after administration, and subsequent doses are adjusted to a target range of 0.5 to 1.1 IU/mL. Trough levels can be assessed to determine patient compliance and direction for scheduled or urgent surgical procedures.

While standardized prophylaxis or treatment doses of LMWHs can be used in most patients without dose adjustment, the appropriate dosing of LMWHs in obese patients remains unclear. Patients with extremes of body weight are rarely enrolled in pharmacokinetic or clinical trials that shape drug dosing. Pharmacokinetic studies suggest that patients at extremes of body weight may not reach adequate anti-Xa levels. Retrospective analyses of LMWH-VTE prophylaxis trials have found a relationship between body mass index (BMI) and thrombosis. When evaluating subgroups of medically ill patients, LMWH administration in obese patients (BMI >40kg/m²) showed no reduction in VTE. In orthopedic surgery, obese patients treated with LMWH had higher thrombosis rates compared with nonobese patients. In bariatric surgery, a higher dose of LMWH was more effective in reducing VTE.

Pharmacokinetic studies with LMWH suggest that weight-based dosing of LMWH for treatment of VTE should be based on total body weight and that capping weight–based doses in obese patients is inappropriate. Registry data of LMWH-treated patients with acute VTE showed no significant differences between obese and nonobese patients in rates of recurrent VTE or major bleeding. Subgroup analysis of LMWH use in patients with acute coronary syndromes showed comparable rates of the composite endpoint (death, MI, urgent revascularization) in obese compared to nonobese patients. However, in another similar trial where enoxaparin doses were capped at 10,000 units every 12 hours patients weighing more than 76 kg had a higher incidence of the composite endpoint (death and MI) than those weighing less than 76 kg. Based on these findings, experts now recommend that LMWH dosing should be based on total body weight for the treatment of obese patients. Capping of LMWH doses does not appear to be needed in the treatment of VTE and clinicians should follow specific product recommendations when using LMWH in acute coronary syndromes.

Medical uses

Double blind randomized-controlled trials have shown the benefit of LMWH thromboprophylaxis over placebo in reducing VTE in hospitalized acutely ill medical patients (Table 254-4). While the duration of hospital-administered prophylaxis in these trials ranged from 6 to 14 days, the optimal duration remains unclear.

Low-molecular weight heparins require fewer injections and produce fewer adverse events compared to SC UFH. In hospitalized medical patients receiving thromboprophylaxis, LMWHs were associated with a lower risk of deep vein thrombosis (DVT), fewer injection site hematomas, and no difference in bleeding when compared with UFH. A recent study showed a reduction in pulmonary embolism (PE) rates with LMWH compared to UFH VTE prophylaxis in critically ill hospitalized patients. However, another study comparing LMWH plus graduated compression stockings as compared with graduated compression stockings alone was not associated with a reduction in the rate of death from any cause among hospitalized, acutely ill medical patients. Another analysis evaluating UFH and LMWH thromboprophylaxis in hospitalized medical patients and those with acute stroke showed no impact on mortality and more bleeding events. Professional societies and national guidelines advocate VTE and bleeding risk assessment prior to initiating anticoagulant thromboprophylaxis to facilitate identifying patients most likely to benefit.

In acute VTE, the outcomes of recurrent VTE and major bleeding episodes were similar with UFH and LMWH. LMWHs have largely replaced IV UFH in patients with acute VTE who are able to receive unmonitored, self-administered anticoagulation in the ambulatory setting. UFH remains the preferred option for ACS patients, those who may require an urgent surgical intervention, those with compromised renal function, or those requiring intensive monitoring for other reasons (see Chapter 253 [Diagnosis and Treatment of VTE]).

Complications of therapy and reversal

While major bleeding episodes are reported to occur in 0% to 3% of LMWH treated patients, data suggest hemorrhage rates are lower when compared to UFH. In the setting of overdose or hemorrhage, protamine only partially reverses LMWH anticoagulation, inhibiting approximately 60% of the anti-Xa activity. Protamine dose recommendations depend on the LMWH in use. For enoxaparin reversal within 8 hours of administration, 1 mg of protamine is administered IV to neutralize 1 mg of enoxaparin. If the enoxaparin was administered more than 8 hours ago, 0.5 mg of protamine per 1 mg of enoxaparin may be administered. A second dose of 0.5 mg of protamine per 1 mg of enoxaparin may be administered if bleeding continues or if the aPTT measured 2 to 4 hours after the initial dose remains prolonged. For dalteparin and tinzaparin reversal, protamine 1 mg is administered IV to neutralize every 100 anti-Xa units of the LMWH administered. If aPTT remains prolonged 2 to 4 hours after initial administration, a second dose of 0.5 mg of protamine per 100 anti-Xa units of LMWH may be administered.

Osteoporosis and HIT occur less frequently in patients treated with LMWHs as compared to UFH.

FONDAPARINUX

Pharmacodynamics and monitoring

Fondaparinux is a synthetic analog of the naturally occurring pentasaccharide found in UFH. Fondaparinux, after SC administration, has a long half-life, extending from 17 to 21 hours in patients with normal renal function (Table 254-3). Fondaparinux is renally excreted with elimination reduced in patients with renal impairment. While fondaparinux monitoring is not required on a routine basis the management of certain high-risk patients may benefit from measurement of anticoagulant activity. Because of differences in molecular size and affinity for factor Xa, anti-Xa assays established for LMWH are not accurate for monitoring fondaparinux therapy. Typically local anti-Xa assay, calibrated to fondaparinux, are required.

Medical uses

Fondaparinux has shown to be as effective and safe as LMWH and UFH for the treatment of acute DVT and PE (Table 254-4). It was superior to placebo in reducing VTE in older (>60 years) acutely ill medical patients with no increase in bleeding episodes. However, while fondaparinux showed superior efficacy in reducing VTE in patients undergoing knee arthroplasty, hip arthroplasty, and hip fracture surgery in a combined analysis, the overall incidence of major bleeding was statistically higher when compared with LMWH. Major bleeding in both hospitalized surgical and medical patients is a strong predictor of 30-day mortality. Fondaparinux has been used in the treatment of HIT, but evidence is limited and conclusive data to support routine practice are unavailable.

Complications of therapy and reversal

Fondaparinux is contraindicated in patients with severe renal impairment (calculated creatinine clearance <30 mL per minute) and should not be used for VTE thromboprophylaxis in patients weighing less than 50 kg. No antidote exists to reverse fondaparinux anticoagulant activity. Fondaparinux-related hemorrhage treatment is complicated by its prolonged half-life. Recombinant factor VIIa (rVIIa) reverses the coagulation parameters induced by fondaparinux, but the clinical benefit is unknown.

INDIRECT ACTING ANTICOAGULANTS-ORAL AGENTS

VITAMIN K ANTAGONISTS

Pharmacodynamics and monitoring

Warfarin is extensively metabolized by the hepatic cytochrome P450 isoenzyme system (CYP450) with multiple genetic polymorphisms influencing warfarin dose and duration of action. Patients with reduced enzymatic activity have slower warfarin metabolism and decreased warfarin clearance. They require lower warfarin maintenance doses and often require a longer period to reach steady-state. Likewise, mutations in the gene that code for VKORC lead to increased warfarin sensitivity or warfarin resistance. Altered vitamin K dietary intake impacts the anticoagulation effect of warfarin. Concurrent medication administration frequently impacts warfarin dosing. Warfarin-drug interactions may alter warfarin absorption, metabolism, and plasma protein binding, producing an enhanced or reduced anticoagulant effect. The International Normalized Ratio (INR) is used to measure warfarin’s anticoagulant effect. Intense INR monitoring during the early initiation of warfarin administration will typically will reveal a patient’s sensitivity or insensitivity to therapy. The INR target range will vary based on indication and the patient’s thromboembolic risk and bleeding risk (Table 254-5). The established ranges have been shown to maximize benefit (ie, reducing thrombosis) while minimizing risk (ie, risk of hemorrhage attributable to excessive anticoagulation). Warfarin dosing nomograms have been employed to reduce the time required to reach therapeutic range and to avoid critically high INR values. The time in therapeutic INR range (TTR) summarizes INR control over time and is used to evaluate the effectiveness of warfarin therapy. The greater the percentage of time spent in the therapeutic range, the higher the quality of warfarin dose management.

Parenteral anticoagulation is stopped and warfarin monotherapy continued when the INR is therapeutic for two consecutive measurements. A goal of warfarin therapy is minimize the time to a therapeutic INR. This reduces inconvenience of parenteral LMWH or UFH administration and the cost of therapy. Warfarin dosing nomograms facilitate safe and timely warfarin initiation and have been found to be superior to empiric clinician dosing. Nomograms using both warfarin 5- and 10-mg loading doses have been evaluated. In one study, patients who received a 10-mg loading dose, with INR assessments on days 3 and 5 during the first 8 days of therapy, achieved a therapeutic INR (>1.9) more rapidly, were more likely to have a therapeutic INR by day 5 of parenteral anticoagulant therapy.

Medical uses

Warfarin is effective for primary and secondary prevention of venous thromboembolism, for prevention of systemic embolism in patients with prosthetic heart valves or atrial fibrillation, and for prevention of stroke, recurrent infarction, or death in patients with acute myocardial infarction.

Complications and reversal of effect

The incidence of major bleeding episodes depends on indication for warfarin use. Major bleeding occurs in approximately 2% to 3% of patients per year during the treatment of venous thromboembolism. In contemporary atrial fibrillation trials of stroke prevention, patients taking oral VKAs had reported major bleeding rates ranging from 1.4% to 3.5% per year. Important risk factors for hemorrhage include anticoagulant intensity, time within therapeutic range, and patient age with the risk of anticoagulant-related bleeding highest at the beginning of therapy.

An elevated INR without active bleeding may be managed by withholding warfarin therapy, decreasing the dose, or extending the administration interval (Table 254-6). In patients with bleeding or at risk of bleeding, vitamin K administration will reverse the anticoagulant effects of warfarin. Vitamin K is given orally or parenterally. Correction of INR is fastest with the IV route compared to oral administration, with both faster and more complete than subcutaneous administration. In the setting of serious or life-threatening hemorrhage, warfarin should be held and vitamin K dosing up to 10 mg should be administered by slow IV infusion. Patients receiving vitamin K doses >5 mg may become resistant to warfarin for up to a week following administration, thereby prolonging the time to therapeutic INR in patients reinitiating therapy. The prothrombin complex concentrates (PCCs) are preferred to fresh frozen plasma (FFP) in cases of severe bleeding or where immediate reversal of the INR is necessary. PCC’s are faster to administer and require less fluid volume when compared to FFP. Anticoagulation has been reversed using rVIIa in a small number of patients with refractory bleeding; however, risk of prothrombotic events such as stroke and myocardial infarction limit its use to life threatening bleeds.

Other adverse events of warfarin include acute skin and tissue necrosis, calcification of the aortic valve and coronary arteries with long-term use and feelings of cold or chills.

DIRECT ACTING ANTICOAGULANTS

DIRECT THROMBIN INHIBITORS

Pharmacodynamics and monitoring

Direct thrombin inhibitors include IV argatroban and bivalirudin, and oral dabigatran. While bivalirudin is a peptide produced from leech salivary glands, argatroban is synthetic molecule derived from the amino acid arginine. Dabigatran etexilate is an oral small molecule prodrug with low bioavailability. The capsule formulation should not be opened or crushed because bioavailability is dramatically increased. After oral administration it is converted to its active form, dabigatran. The DTIs differ in their pharmacodynamic properties (Table 254-7). Bivalirudin has the shortest half-life, making it a particularly useful agent in the procedural or periprocedural period.

DTI prescribing is based on patient-specific characteristics such as hepatic function, and renal function. Critically ill patients typically require lower doses than those recommended by the manufacturer. Intravenous DTIs are monitored using aPTT measured every 4 to 6 hours until therapeutic levels are maintained, at which point the monitoring frequency may be extended. Intravenous and oral DTIs prolong the INR. This interaction is most pronounced with argatroban, and magnified when coadministered with warfarin. With concurrent warfarin administration, argatroban must be stopped and, for an accurate assessment, the INR measured within 4 to 6 hours. If the INR is within therapeutic range, warfarin monotherapy may be continued; otherwise argatroban therapy should be reinitiated. This process should be repeated daily until the desired therapeutic range on warfarin alone is reached.

Similar to the IV DTI’s, dabigatran prolongs the aPTT and the INR. While ecarin clotting time (ECT) or thrombin time (TT) may be useful laboratory tests for monitoring coagulation, no recommendations exist to guide therapy.

Medical uses

Argatroban significantly reduces thromboembolic complications (death, new thrombosis, and amputations) in patients with HIT. Bivalirudin has been safely used in critically ill HIT patients. DTI therapy is continued in HIT patients until platelet count returns to normal (platelets >150 x 10⁹/L) or to baseline values. Platelet recovery signifies a reduction in antibody mediated platelet activation and thrombin generation, creating the window for initiating oral anticoagulation.

Argatroban and bivalirudin are indicated for prophylaxis of thrombosis in patients with, or at risk for, HIT undergoing PCI. Bivalirudin is also indicated in the treatment of patients undergoing PCI as well as those with unstable angina/non-ST segment elevation myocardial infarction undergoing PCI (Table 254-8).

Dabigatran is indicated for the prevention of stroke and systemic embolism in patients with atrial fibrillation, for the acute treatment of VTE, and to reduce the risk of VTE recurrence.

Complications and reversal of effect

No specific antidotes for DTI reversal exist. rFVIIa and activated prothrombin complex concentrates (aPCC) may be options for returning coagulation parameters to normal values and reverse bleeding. Dabigatran’s low plasma protein binding allows for removal with hemodialysis and may be a strategy to reverse its effects. Idarucizumab, an antibody fragment, was developed to reverse the anticoagulant effects of dabigatran by selectively binding to and inhibiting its activity. An interim study suggests idarucizumab completely reverses the anticoagulant effect of dabigatran within minutes as measured by normalization of TT and ECT. Hemostasis was restored in approximately 12 hours.

DIRECT ACTING FACTOR XA-INHIBITORS

Pharmacodynamics and monitoring

Apixaban, edoxaban, and rivaroxaban are selective inhibitors of free and clot bound Xa. As a class these agents have predictable anticoagulant activity eliminating the need for routine laboratory monitoring and allowing for fixed dose administration. Despite these advantages, between the agents there are differences in their pharmacokinetic parameters (Table 254-9). They have varying degrees of renal clearance, and hepatic metabolism, elimination half-life, susceptibility to drug interactions, and dosing frequency that requires prescribing that is tailored to patient needs. While these agents alter aPTT and INR values, these tests cannot be used effectively to monitor therapy. Calibrated chromogenic anti-Xa assays provide a means to potentially monitor activity; however, exact therapeutic targets have yet to be described in practice.

Medical uses

Apixaban, edoxaban, and rivaroxaban have all established safety and efficacy in large, randomized trials compared with warfarin for risk reduction of stroke and systemic embolism in patients with nonvalvular atrial fibrillation. All patients included in these trials were at an increased risk of stroke due to one or more additional risk factors, such as previous stroke or transient ischemic attack (TIA), congestive heart failure, diabetes mellitus, hypertension, or an age ≥75 years (Table 254-10). Meta-analyses have shown these agents are associated with significant reductions in stroke, intracranial hemorrhage, and mortality. Major bleeding episodes were similar in occurrence to warfarin, but there was an increase in gastrointestinal bleeding. The relative efficacy and safety of new oral anticoagulants is consistent across a wide range of patients.

Similarly, meta-analyses have been performed to evaluate the efficacy and safety outcomes associated with UFH, LMWH, or fondaparinux in combination with VKA versus the direct acting oral agents for treatment of VTE. These agents have comparable efficacy to that of VKAs in preventing recurrent VTE. However, they are associated with a significantly lower risk of bleeding complications, including the risk of major bleeding, intracranial bleeding, fatal bleeding, and clinically relevant nonmajor bleeding. This efficacy and safety has also been seen consistently in sub groups of patients with PE, DVT, a body weight >100 kg, moderate renal insufficiency, an age >75 years, and cancer.

Clinical outcomes have been evaluated with new oral anticoagulants for prophylaxis against venous thromboembolism after total hip or knee replacement with meta-analysis showing a higher efficacy but with a higher bleeding tendency. Compared with enoxaparin, the risk of symptomatic VTE after surgery was lower with rivaroxaban and similar with dabigatran and apixaban. The risk of clinically relevant bleeding was higher with rivaroxaban, similar with dabigatran, and lower with apixaban.

Clinicians will be faced with routinely transitioning patients between parenteral therapy and oral therapy as well transitioning to and from warfarin. The strategies to do this are included in each products label (Table 254-11).

Complications and reversal of effect

Several antidotes to direct acting oral anticoagulants are in various stages of development. Andexanet alfa is a modified form of factor Xa that immediately reverses the anticoagulant effect of oral direct acting Xa inhibitors after administration (Table 254-12). Ciraparantag has a broader action, inhibiting all anticoagulants with the exception of intravenous DTIs and warfarin.

Until the doses of these agents are established and they reach the marketplace, the use of routine supportive care (stopping anticoagulant therapy, applying local hemostatic measures, administering fluid resuscitation, red blood cell transfusion, and initiating studies to identify the bleeding source) are recommended as initial steps. Several guidelines have emerged advocating treatment of life-threatening and major bleeding episodes with the administration of aPCC, PCC, and rFVIIa (Figure 254-3). Their inclusion is based on evidence from animal models and the reversal of coagulation parameters in healthy human volunteers. rFVIIa administration is associated with an increase in thrombosis in nonhemophiliac patients. The administration of PCC has the potential to increase the risk of thrombosis. Despite the limited efficacy and the increased likelihood of thrombosis, experts have deemed the administration to be a reasonable approach in extreme clinical situations of hemorrhage.

SUGGESTED READINGS

ONLINE RESOURCES

INTRODUCTION

The aorta is the largest artery in the body and connects the heart to the systemic vascular bed. It consists of both a thoracic and an abdominal portion, which are delineated by the ligamentum arteriosum. The thoracic aorta begins at the aortic valve and has four segments: the aortic root, ascending aorta, aortic arch, and descending aorta. These segments originate connections with several critical vessels including the coronary, innominate, subclavian, and carotid arteries. Similarly, the abdominal aorta, which begins as the vessel travels through the diaphragm, gives rise to the celiac, superior mesenteric, inferior mesenteric, and renal arteries and ultimately bifurcates into the common iliac arteries.

As with other arteries, the aortic wall consists of three distinct layers. The intima, or innermost wall, primarily contains endothelial cells. These cells function as a regulatory barrier between the blood and the rest of the vessel wall and modulate processes such as thrombosis, fibrinolysis, and inflammation. The middle layer, or media, consists of vascular smooth muscle cells, elastic fibers, and other extracellular matrix components. Finally, the outer layer, or adventitia, consists primarily of collagen, nerves, and small blood vessels known as the vasa vasorum, which help provide oxygen and nutrients to the vessel wall itself. Separating each vessel layer is an additional layer of elastic lamina. Functioning in conjunction with the medial layer, these elastic fibers confer both distensibility and elastic recoil of the aortic wall, which are essential properties for normal function.

The primary function of the aorta is transport of oxygenated blood from the left ventricle to the rest of the body. In order to accomplish this, the aorta must withstand the shear stress and transmural pressure generated by blood as it is ejected from the left ventricle. Rather than functioning as a passive conduit, the vessel wall components permit distension of the aorta during systole, thereby storing potential energy. The aorta undergoes compensatory recoil during diastole, which ensures ongoing blood flow throughout the cardiac cycle. Changes in the vessel wall, either from aging, mechanical trauma, or disease-associated pathology, may affect the elastic properties of the aorta and lead to abnormalities at any position along the vessel. According to data from the Centers for Disease Control and Prevention, diseases of the aorta were responsible for nearly 10,000 deaths in the United States in 2013. Given the significant morbidity and mortality of aortic pathology despite its relative rarity, clinicians must be able to recognize and manage the various disease processes that can affect the aorta.

THORACIC AORTIC ANEURYSM

An aneurysm is a pathologic, focal dilatation of a blood vessel that involves the intima, media, and adventitia. Most aneurysms affect the entire circumference of the vessel and are termed fusiforum. Saccular aneurysms involve only a small region of the vessel wall and cause diverticular outpouchings. The normal size of the thoracic aorta depends on the segment in question in addition to other factors, such as age, gender, and body surface area. However, in general, a segment that is at least 50% larger than its proximal normal segment is considered aneurysmal. Nearly 60% of thoracic aortic aneurysms involve the aortic root or ascending aorta with involvement of the descending aorta, arch, and thoracoabdominal segments occurring less frequently. In some cases, the aneurysm spans multiple segments, either contiguously or in isolation.

Numerous factors contribute to the development of thoracic aortic aneurysms. A lifetime of mechanical wall stress normally causes dilatation of the thoracic aorta with advancing age. Similarly, hypertension subjects the aorta to increased wall stress and increases the risk of aneurysm development. Aortic atherosclerosis is the primary risk factor for descending thoracic aortic aneurysms, and these aneurysms often extend into the abdominal segment. Conversely, atherosclerosis is rarely implicated as the primary cause of ascending aortic aneurysms. Other cardiovascular risk factors, such as a history of smoking and coronary artery disease, also increase the risk of aneurysm development. Histologically, the majority of ascending aortic aneurysms exhibit cystic medial degradation. This refers to the breakdown of collagen and elastic fibers within the media layer as well as the loss of vascular smooth muscle cells in this layer ultimately leading to the development of cystic pockets of extracellular matrix. Over time, the vessel wall becomes less distensible and more susceptible to wall stress, ultimately leading to aneurysm development and progressive dilatation. Cystic medial degradation, a pathological finding rather than a specific cause, is particularly prevalent in genetic syndromes of thoracic aortic aneurysm, and it occurs to some degree as a normal part of aging. Finally, cases of infectious and noninfectious aortitis (discussed later) can cause aneurysmal dilatation.

In recent years, researchers have identified numerous genetic disorders that are associated with thoracic aortic aneurysms. A connective tissue disorder inherited in an autosomal dominant pattern, Marfan syndrome is caused by mutations in fibrillin-1, which plays an important role in normal extracellular matrix function. This mutation not only causes structural weakening of the aortic wall itself but also leads to upregulation of TGF-β signaling. As a result, the aorta in these patients responds abnormally to mechanical stress and may become aneurysmal. Similarly, Loeys-Dietz syndrome, caused by mutations in one of the TGF-β receptors, can cause aortic tortuosity and aneurysm development along with craniofacial abnormalities. Other genetic disorders associated with thoracic aortic aneurysms include Ehlers-Danlos syndrome type IV and the familial thoracic aortic aneurysm syndrome. Turner syndrome is a chromosomal disorder affecting only females that is caused by a complete or partial absence of a second X chromosome. Up to 42% of affected individuals may develop aortic root aneurysms, often in conjunction with bicuspid aortic valves, and this group is at particularly high risk for aneurysm rupture and dissection.

In the United States, 1% to 2% of the general population has a bicuspid aortic valve, and up to half of these patients will also develop a thoracic aortic aneurysm during their lifetime, making it one of the most common causes of aneurysm development. The aneurysm typically involves the aortic root and ascending aorta and may occur with or without concomitant aortic stenosis. The hemodynamics of blood flow across bicuspid aortic valves is abnormal and may play a role in aneurysm development. Similar to Marfan syndrome, fibrillin-1 may also contribute to aortic pathology in patients with bicuspid aortic valves, and individuals can develop aneurysms even after surgical replacement of the bicuspid valve.

Most thoracic aortic aneurysms are clinically silent and detected incidentally during various thoracic imaging studies. Aneurysms involving the aortic root or ascending aorta can cause aortic insufficiency through distortion of the valve annulus. This may be detected on examination as an early diastolic murmur, particularly along the left sternal border. Symptoms attributable to thoracic aortic aneurysms are generally due to mass effects. Because of the aorta’s position within the mediastinum, aneurysms may compress the trachea, mainstem bronchi, esophagus, and recurrent laryngeal nerve. Thus, patients can experience shortness of breath, wheezing, dysphagia, and hoarseness. In some cases, patients may experience pain due to compression of other thoracic organs or even erosion into the spine, sternum, or ribs.

Once a thoracic aortic aneurysm has been diagnosed, clinicians should also look for other clues on the physical examination that may indicate an underlying genetic disorder in the appropriate clinical context. In individuals with a bicuspid aortic valve, heart auscultation may reveal a systolic ejection click after the first heart sound, which corresponds to abnormal opening of a nonstenotic valve. If the bicuspid valve has become stenotic due to early degeneration, clinicians may hear a systolic ejection murmur, often at the right upper sternal border, as one would with age-related calcific aortic stenosis. Common physical examination features of Marfan syndrome include tall stature, thin habitus with diminished muscle mass, elongated skulls, arachnodactyly, scoliosis, and chest wall abnormalities. Those with Loeys-Dietz syndrome typically exhibit hypertelorism, bifid uvula, or cleft palate. Patients with features concerning for an underlying genetic aortopathy should be referred to a clinician with expertise in the field to discuss further workup and family screening.

Chest x-rays in individuals with thoracic aortic aneurysm may show mediastinal widening and displacement of the airways. This is a low-sensitivity test for aneurysm and may not detect small aneurysms or those primarily involving the aortic root or ascending aorta. Concern for aneurysm on chest x-ray, particularly in the right clinical context, should prompt more definitive imaging. In many patients, particularly those who are not obese or barrel-chested (such as in COPD), transthoracic echocardiography can reliably image the aortic root and ascending aorta. This modality is particularly useful for screening and long-term monitoring in young patients with known risk factors for aneurysms in this region, such as those with Marfan syndrome and bicuspid aortic valves. However, there are several key limitations to transthoracic echocardiography in evaluating thoracic aortic aneurysms. Measurements of the aorta must be normalized to age and body surface area, and interobserver variability in measurements may cause confusion in the clinical setting if it is not possible to view the images oneself. Additionally, it is often important to image the entire aorta at the time of diagnosis to rule out aneurysmal dilatation of other segments, and detection of an aneurysm with echocardiography should not obviate the need for more comprehensive imaging. Transesophageal echocardiography is useful to view the aortic arch and descending aorta, but its invasive nature makes it more appropriate for acute aortic syndromes. Increasingly, CT and MR angiography have become important tools to image the entire thoracic aorta, and they are both highly sensitive and specific for aortic disease. In patients with prior aortic interventions, CT is the preferred modality. When using these modalities, it is important to note that aortic tortuosity may lead to inaccurate measurement of vessel diameter on axial images. Therefore, it is often necessary to use 3-D reconstruction to measure the diameter perpendicular to blood flow, and interpretation by a vascular imaging specialist may be necessary, particularly in complex cases.

The primary concern with thoracic aortic aneurysms is progressive dilatation and weakening leading to dissection, rupture, and death. Although aneurysm diameter is the most important risk factor for these clinical outcomes, the natural history of thoracic aortic aneurysms varies widely among individuals. In general, the annual growth rate is up to 0.2 cm/y, but this rate may accelerate with increasing size. Clinical registries show that the rate of dissection and rupture rapidly increases from 2% to 3% in aneurysms below 6 cm in diameter to ~7% in those 6 cm and larger. Additionally, descending aortic aneurysms grow more rapidly than those of the root and arch. Age, female sex, poorly controlled hypertension, and tobacco use are associated with more rapid growth. Those with strong family histories of aneurysm often progress at a faster rate. Aneurysms in patients with genetic causes of thoracic aortic aneurysm, such as Marfan syndrome, generally dilate more rapidly. Importantly, aneurysms in Marfan syndrome and individuals with bicuspid aortic valves are at particular risk for further dilation during pregnancy, likely due to both increased cardiac output and increased laxity of the vessel wall due to hormonal effects. Similarly, Marfan and other genetic syndromes are associated with an increased overall risk of dissection and rupture, particularly if there is a family history of these events, and this is more likely to occur at smaller diameters. Also, patients with bicuspid aortic valves exhibit a higher growth rate with increased risk of dissection and rupture at a smaller size.

Risk factor modification is an important component of thoracic aortic aneurysm management. All patients should be counseled on smoking cessation, and clinicians should initiate antihypertensive therapy for blood pressure control with a goal of <140/90 mm Hg. If the aorta exhibits atherosclerotic plaque, then high-intensity statin therapy is also reasonable, although there are insufficient data to support this for all causes of thoracic aortic aneurysm. Patients should be treated with β-blockers at maximally tolerated doses to reduce wall stress. Additionally, β-blocker therapy finds particular use in patients with Marfan syndrome, even in the absence of aneurysm. Animal data indicate that losartan may prevent aneurysm formation in Marfan syndrome due to its off target effect of TGF-β downregulation. Large clinical trials do not yet support the use of angiotensin receptor blockers over β-blockers, but it is reasonable to utilize them if additional blood pressure control is necessary. Lifestyle modification is important to prevent further aneurysm dilation. Patients should avoid strenuous physical activity and isometric exercises. Additionally, because of the dangers associated with pregnancy, individuals pursuing pregnancy should be referred to a high-risk obstetrician and a cardiologist to formulate a plan to minimize these risks during both pregnancy and delivery. If there is clinical concern for a genetic aortopathy, patients should be referred to a cardiologist or vascular medicine specialist for further evaluation and consideration of genetic testing as well as screening of first degree relatives.

Because thoracic aortic aneurysm is generally a progressive disease, long-term monitoring is essential. If the aneurysm is inadequately imaged on transthoracic echocardiogram, dedicated MR or CT imaging should be performed to fully characterize the dilatation. After diagnosis, the 2010 American College of ­Cardiology/­American Heart Association guidelines recommend follow-up imaging at 6 months to look for stability. Assuming the aneurysm is stable at 6 months, annual screening is reasonable; conversely, biannual imaging is recommended for rapidly enlarging aneurysms or those approaching indication for repair. It is important to recognize that interobserver variability may affect whether an aneurysm is interpreted as stable or enlarging, and detailed review of these imaging studies in person is often necessary. In syndromes that primarily affect the ascending aorta, such as bicuspid aortic valves and Marfan syndrome, transthoracic echocardiography is a reasonable modality for routine monitoring. However, there are increasing data characterizing the incidence of distal aorta disease in Marfan syndrome, and it is unclear how frequently the entire aorta should be monitored in such individuals.

All symptomatic aneurysms warrant surgical repair given the high rate of morbidity and mortality without intervention. Additionally, repair of asymptomatic aortic aneurysm is indicated for ascending aortic aneurysm ≥5.5 cm in the general population and ≥5.0 cm in those with bicuspid aortic valves or Marfan syndrome. In contrast, physicians may delay descending aortic aneurysms repair until ≥6.5 cm. The surgical approach to repairing thoracic aortic aneurysms varies based on the segments involved. Aneurysms of the ascending aorta are generally resected and replaced with a prosthetic graft. In some cases, the aortic valve is also replaced, such as in those with annular dilatation and aortic insufficiency or degenerative aortic valve disease in the setting of a bicuspid aortic valve. Such procedures require a median sternotomy as well as cardiopulmonary bypass. Arch aneurysms are particularly difficult to repair due to concerns regarding head perfusion during the procedure and reanastomosis of the branch vessels. These surgeries often involve some combination of open surgical repair, endovascular stent-graft placement, and branch vessel bypass construction. Descending aortic aneurysms are commonly repaired with prosthetic grafts via a thoracotomy. A significant risk of these procedures is paralysis due to interruption of spinal cord blood flow. For elective procedures, the rate of death is 3% to 5% at experienced centers, with the risk increasing to nearly 7% if arch repair is involved.

ABDOMINAL AORTIC ANEURYSMS

Abdominal aortic aneurysms (AAA) (Figure 255-1), classified as dilation ≥3.0 cm, are much more common than those in the thoracic segment. Incidence dramatically increases with age, particularly in men older than 50 and woman over 70. Up to 3% of men ≥50 and 5% of men ≥65 will develop abdominal aortic aneurysms. Men are 5 to 10 times more likely to develop aneurysms than women. Smoking is the primary risk factor for abdominal aortic aneurysms with a direct correlation between the number of years an individual smoked and aneurysm incidence. The overall risk is 3 to 5 times higher in smokers. Additionally, hypertension, dyslipidemia, and coronary atherosclerosis are associated with aneurysm formation. Caucasians are at higher risk than African Americans. In some cases, abdominal aortic aneurysms are also accompanied by aneurysms in the iliac, femoral, and popliteal arteries. Estrogen may play a protective role with respect to aneurysm formation, thus explaining the later onset in women. Unlike thoracic aortic aneurysms, the genetic underpinnings of abdominal aortic aneurysms are poorly understood. There does seem to be a heritable component, as individuals with a family history are at increased risk. Some data have shown that a common sequence variant on chromosome 9p21 is associated with abdominal aortic aneurysm, intracranial aneurysm, and coronary artery disease. However, more research is necessary to better understand the genetic basic of abdominal aortic aneurysms.

Histologic examination of abdominal aortic aneurysms shows chronic wall inflammation. In addition, various proteinases, particularly matrix metalloproteinases, cause breakdown of collagen and elastin within the vessel wall. Smooth muscle cells, which can help repair the vessel by releasing collagen and elastin, are often diminished in aneurysm walls. Therefore, over time, aneurysms are more susceptible to wall stress throughout the cardiac cycle and are less capable of repairing the damage caused by ongoing inflammation. This process leads to progressive aneurysmal dilatation and ultimately rupture.

As with thoracic aortic aneurysms, abdominal aneurysms are typically asymptomatic. Occasionally, patients present with a gnawing, constant pain within the abdomen, lower back, or flank. Rupture may cause severe onset of pain in these same regions. Physical examination may reveal a pulsatile mass along the midline of the abdomen. In individuals with a small body habitus (<40 in waistline), it may even be possible to estimate aneurysm diameter on examination.

Abdominal ultrasound is highly accurate for the assessment of the abdominal aorta, and it is the preferred modality for screening and follow-up in uncomplicated cases of abdominal aortic aneurysm. The US Preventive Services Task Force recommends a single ultrasound screening of all men age 65 to 75 with any prior smoking history at the Welcome to Medicare visit. Screening is also reasonable in individuals with a family history of aneurysm, although the age of screening and interval of follow-up are unclear. In some cases, particularly with large or tortuous aneurysms or in obese patients, CT imaging may be necessary. CT angiography permits 3-D reconstruction and measurement of the true aortic diameter, similar to cases of thoracic aortic aneurysm. It is an important modality prior to aneurysm repair as it permits imaging of the abdominal visceral branches, renal arteries, and iliac vessels. Abdominal aortic aneurysms often contain extensive atherosclerotic plaque and thrombus which can be important to identify prior to repair. MR angiography can also be used, particularly in young patient in whom radiation exposure is a concern. However, CT imaging generally permits more accurate measurements. Once an aneurysm is detected, the interval of follow-up depends on aneurysm size. In particular, aneurysm 3.5 to 4.4 cm in diameter should be followed annually with biannual follow-up for those ranging 4.5 to 5.4 cm. Those exhibiting rapid growth should also be followed more closely. Growth >1 cm/y is an indication for repair.

As with thoracic aortic aneurysms, abdominal aortic aneurysms dilate over time with an increased risk of rupture at larger sizes. The annual rate of growth is generally up to 0.4 cm/y, although this can vary significantly between individuals. Women are at higher risk for rupture than men. For men with aneurysms ≥5cm, the annual rate of rupture is 12% with a rate of 18% for women. The risk of rupture dramatically increases at larger size, with up to 50% of aneurysms greater than 7 cm culminating in rupture. Nearly 75% of individuals who suffer aneurysm rupture die before undergoing surgery, and the 30-day survival rate following rupture is only 11%, thus demonstrating the importance of early detection and close follow-up.

Similar to diseases within other vascular beds, smoking cessation is a critical component in abdominal aortic aneurysm therapy, as ongoing smoking is associated with more rapid aneurysm growth. There are no clear data to support statin therapy although the presence of significant atherosclerotic disease within the aorta should generally prompt guideline based statin therapy. Similarly, clinicians should target standard blood pressure goals in patients with abdominal aortic aneurysms to minimize their overall risk of adverse cardiovascular and cerebrovascular events. At present, there are no definitive data supporting β-blocker or angiotensin receptor blocker use in abdominal aortic aneurysms, although these are reasonable agents for achieving adequate blood pressure control.

The decision to repair abdominal aortic aneurysms is complex, particularly because these patients are typically older and have multiple medical comorbidities. Symptomatic aneurysms and those growing more than 1 cm/y should be repaired. Aortic aneurysm rupture should also prompt emergent surgical repair, although the operative mortality is nearly 50%. Outside of these situations, repair is generally considered once aneurysms reach 5.5 cm in diameter, although repair at smaller diameter may be warranted in woman or in individuals with a family history of aneurysm rupture. Repair can be achieved either through an open technique or via an endovascular approach. Open surgical repair typically requires a large midline incision followed by replacement of the aneurysmal segment with a prosthetic graft. Operative mortality ranges from 2% to 6% with lower rates at centers performing a high volume of repairs.

In recent years, endovascular abdominal aortic aneurysm repair (EVAR) has become increasingly popular, currently representing more than 50% of all AAA repairs (Figure 255-2). During this procedure, a prosthetic graft is typically delivered through a transfemoral catheter system. The graft spans the aneurysmal segment and forms a seal between the graft and a healthy segment of the vessel, thus prohibiting blood from leaking into the excluded aneurysm and causing further dilation or rupture. EVAR is associated with a lower morbidity and mortality rate than open surgical repair. However, this technique has several important limitations. Patients are more likely to require repeat interventions following EVAR. Many abdominal aortic aneurysms are not amenable to endovascular repair due to tortuosity or involvement of important branch vessels that cannot be protected via EVAR. Additionally, up to 20% of patients develop endoleaks, or ongoing blood flow between the aneurysm wall and endograft, following EVAR. In many cases, these endoleaks can cause further aneurysmal dilatation, and, ultimately, rupture. As a result, EVAR is typically reserved for individuals at high risk for perioperative complication of death with open surgical repair. Endoleaks can be visualized on CT, and given their high rate of recurrence, patients typically undergo frequent CT monitoring following EVAR. Additionally, clinicians should not assume an abdominal aortic aneurysm has been definitively managed following EVAR, and it is important to further investigate clinical scenarios concerning for ongoing aneurysm dilation or rupture. Finally, patients deemed “too sick” for open AAA repair do not derive any survival benefit from endovascular repair. Thus, EVAR has become a lower-morbidity method of repairing aneurysms that meet current recommendations.

AORTIC DISSECTION

Aortic dissection is typically caused by a rent within the aortic intima. Following this initial tear, ongoing shear stress within the vessel leads to further dissection into the vessel wall, and pulsatile blood flow into this false channel can lead to bleeding within the wall or propogation of the dissection along the length of the aorta, thus separating the intima and media layers. This often creates a dissection flap with a distinct false lumen separate from the true lumen of the aorta. In some cases, this is a blind-ended flap, although another intimal tear can develop at the distal end of the flap, thereby permitting blood flow back into the true lumen. Over time, stagnant blood within the false lumen may thrombose. Dissection is rare; according to the International Registry of Acute Aortic Dissection (IRAD) database, there are approximately 2000 new cases of acute aortic dissection annually in the United States, although other estimates place the figure as high as 10,000. Despite their relative infrequency, vigilance is required, for dissections are associated with an extremely high mortality. Classically, the early mortality rate is estimated at 1% per hour. Without treatment, ~25% of patients with dissections involving the ascending aorta die within 24 hours, and this rate increases to 50% at 48 hours and 75% at 2 weeks. Thus, it is critical for clinicians to maintain a high clinical suspicion for acute aortic dissection.

There are two primary classification symptoms for aortic dissection, the DeBakey and Stanford systems, although the Stanford classification is more commonly used clinically. Under the Stanford classification system, Type A dissections (Figure 255-3) involve the ascending and arch aorta proximal to the left subclavian artery and can involve any number of additional segments, while Type B dissections (Figure 255-4) do not involve the aorta proximal to the subclavian artery.

Many of the same risk factors that predispose individuals to aortic aneurysm development are also associated with aortic dissection. Hypertension is the most common risk factors associated with dissection and is seen in up to 75% of individuals. Atherosclerotic disease is also common in individuals with dissection. Men are twice as likely to be affected as women, and women typically develop dissections at an older age. Due to ongoing vessel wall changes, aneurysmal aortic segments are at higher risk for dissection. Additionally, dissections are more common with advanced age. There is a slightly higher risk of dissection during pregnancy, likely due to hormonal changes of the vessel wall. Vigorous exercise, particularly isometric activities like weight lifting, can precipitate dissections in susceptible patients. Infectious and autoimmune aortitis is associated with dissection as is cocaine use. Both bicuspid aortic valve and aortic coarctation can trigger aortic dissections. Finally, numerous connective tissue disorders increase one’s risk for acute aortic dissection, including Marfan syndrome, Loeys-Dietz syndrome, Ehlers-Danlos syndrome type IV, and the familial thoracic aortic aneurysm syndrome. Even without a clearly identifiable genetic aortopathy, nearly 20% of individuals with an acute aortic dissection have a family history of aneurysm or dissection.

The presentation of acute aortic syndromes ranges from the dramatic to the asymptomatic and depends on the segment involved and end-organs affected. Nearly all patients develop either pain that is both sudden in onset and severe or syncope. Large series indicate the pain is most often described as abrupt, severe, sharp, or stabbing with the classic description of tearing and ripping pain occurring less frequently. Within the IRAD database, most patients with Type A dissections presented with anterior chest pain, and Type B dissections more commonly caused back pain. Abdominal pain can occur in both scenarios. The pain can radiate to the neck, jaw, arms, or legs depending on branch vessel involvement. Occasionally, patients are asymptomatic, which is more common with advanced age. Symptoms often wax and wane, and this may correspond to interruption and restoration of blood flow by the dissection. Dissection flaps can limit flow due to direct occlusion of a branch vessel by the flap or extension of the dissection itself into the branch vessel. Expansion of the false lumen can cause branch vessel hypoperfusion, and thromboembolism of the false lumen can also lead to ischemia in various vascular beds.

A variety of other clinical scenarios are associated with acute dissection. Aortic valvular insufficiency occurs in up to 76% of Type A dissections and can result from acute root dilation, direct extension of the dissection into the aortic annulus, or prolapse of a dissection flap across the valve, thus interfering with valve closure. Dissection can permit fluid entry into the pericardial space, thus leading to cardiac tamponade in up to 10% of Type A dissections. Dissections can directly impair coronary perfusion, typically the right coronary artery, thus leading to an acute inferior myocardial infarction. Patients frequently experience syncope due to temporary interruption of cerebral flood flow. Hypertension is common in acute dissection, although hypotension can also occur in the presence of aortic rupture, tamponde, or heart failure. Pleural effusions, usually left-sided, and mesenteric ischemia are also common. Finally, arch vessel occlusion or distal embolization can cause stroke or transient ischemic attack.

On examination, up to 30% of patients will exhibit pulse deficits, and mortality increases with the number of limbs that exhibit diminished pulses. Cardiac auscultation may reveal the diastolic murmur of aortic insufficiency if the aortic valve is affected. Those with cardiac tamponade will exhibit pulsus paradoxus and Kussmaul’s sign. Obstruction of blood flow in the right coronary artery may lead to the classic signs of a right ventricular infarct, including elevated jugular venous pressure and a right ventricular heave. Neurologic deficits may be present. Chest x-ray often reveals a widened mediastinum. An enlarged cardiac silhouette may indicate the presence of an associated pericardial effusion, and a pleural effusion may also be present. ECG may reveal low-amplitude voltages or electrical alternans in the setting of a pericardial effusion or tamponade, and right coronary artery occlusion can lead to ST segment elevations in the inferior leads. On laboratory analysis, D-dimer is a useful diagnostic test for excluding dissection. Values <500 ng/mL within 24 hours of symptoms onset have a negative predictive value of 95%. Depending on which additional vascular beds are affected, patients may also have elevations of troponin isoforms, creatine kinase-MB, lactate, aspartate transaminase, and alanine transaminase, although none of these assays are specific for aortic dissection.

Definitive identification and characterization of aortic dissection is critical to ensure appropriate management. Special consideration must be given to body habitus, hemodynamic stability, availability of various imaging modalities, and the speed with which various studies can be obtained. Transthoracic echocardiography can image the aortic root and is commonly available as a rapid, point-of-care modality. However, it is limited in its ability to image the rest of the aorta, and it has extremely poor sensitivity for detecting aortic dissection. Time spent obtaining a transthoracic echocardiogram typically only delays definitive diagnosis and should not be performed as the initial screening study if there is an elevated index of suspicion. If follow-up imaging of the aortic valve is necessary, this can be quickly accomplished with a transthoracic echocardiogram. In contrast, transesophageal echocardiography is ~98% sensitive for aortic dissection and is also useful in imaging the aortic valve in patients with accompanying aortic insufficiency. Although the avoidance of radiation exposure is desirable, 24-hour availability of transesophageal echocardiography is not universal. Additionally, patients may be too hemodynamically unstable or in too much pain to tolerate the procedure without intubation.

Increasingly, contrast-enhanced CT has become the primary modality for aortic imaging in the acute setting because of its placement in emergency wards. It permits rapid visualization of the entire aorta as well as large branch vessels. Axial images typically demonstrate two lumens separated by the dissection flap with different degrees of contrast enhancement within each lumen. Particularly with experienced operators, sensitivity and specificity approach 100%. MRA with gadolinium enhancement is similarly effective at evaluating aortic dissection with the added benefit of avoiding radiation exposure and contrast-induced nephropathy. Unfortunately, image acquisition takes longer than CT, and it is less commonly available in the emergent setting. Moreover, the duration of the test and its limitations on patient surveillance diminish its value in the acute setting.

Once aortic dissection is diagnosed, the initial management should consist of ensuring hemodynamic stability and airway protection. Hemodynamically unstable patients should undergo emergent surgical evaluation. Among stable patients, it is not uncommon for them to require intravenous narcotics for adequate pain control. Tight blood pressure management is necessary to limit propagation of the dissection and to minimize the risk of wall rupture. The target systolic blood pressure should be less than 120 mm Hg. β-Blockers should be used to lower the heart rate as close to 60 beats/min as possible as initial therapy. Intravenous esmolol and labetalol infusions are often used to achieve both these targets. Esmolol in particular permits rapid titratability due to its short half-life. Diltiazem and verapamil are also options but are used less commonly. Many providers use sodium nitroprusside due to its rapid effects on systolic blood pressure. However, as a direct vasodilator, sodium nitroprusside alone can increase aortic wall stress and, therefore, must be administered in conjuction with β-blockers.

While medical management is ongoing, the clinical team should also begin the evaluation for definitive therapy. Ideally, this evaluation should involve cardiologists, cardiac surgeons, vascular surgeons, and vascular imaging experts. If necessary, stable patients should be transferred to high volume centers with experience in such cases. Patients with acute type A dissections should undergo urgent surgical repair, and surgery is associated with improved survival compared to medical therapy alone. Various studies have found mortality rates of 22% to 26% in patients undergoing surgical repair of Type A dissections compared to 58% for those managed medically. Surgical repair of Type A dissections typically requires sternotomy and excision of the dissected portion followed by graft replacement. Close examination of the aortic valve is necessary, and valve replacement should be performed as part of the same procedure if valve function has been disrupted. As in thoracic aorta aneurysm, dissections involving the aortic arch are more complex and also require grafting to the head and upper limb vessels either via open repair or through a hybrid procedure using additional endovascular repair. If the distal aorta is involved, it may be excluded with an endograft or simply left untouched if it is not causing end-organ malperfusion.

Type B dissections are associated with a lower-mortality rate of anywhere from 1% to 11% in individuals undergoing medical management. Typically, uncomplicated cases of Type B dissections are managed conservatively in the absence of end-organ malperfusion with in hospital mortality of up to 20%. However, if the dissection has led to compromise of limb, visceral, or renal blood flow, intervention may be necessary. Contained or impending rupture is also an indication for surgical repair. If surgery is warranted, options include open surgical excision with graft replacement or thoracic endovascular aortic repair (TEVAR). In the IRAD database, 10% of patients with type B dissections underwent open repair, whereas 12% underwent TEVAR. Mortality is nearly three times higher with open repair. Occasionally, patients with type B dissections and organ malperfusion must undergo percutaneous fenestration procedures. In these procedures, a hole is created in the distal portion of the false lumen, thereby permitting blood flow back into the true lumen and into the affected side branch.

Independent of the initial management, there are many long-term factors that must be addressed. Patients commonly require multiple antihypertensive agents to ensure adequate blood pressure control, and this should be tightly controlled for life to help prevent recurrence. β-Blockers should be part of the regimen for all patients following aortic dissection. Smoking cessation should be strongly counseled for all patients. If there is a clinical concern for an underlying connective tissue disorder, patients should be referred to a specialist for further evaluation. All patients should be counseled to avoid strenuous activity, particularly isometric activity, as this could provoke recurrence. Routine monitoring on CT or MR is necessary whether the patient was managed medically or with surgical repair. Residual dissection flaps can propagate further, and leaks may develop at graft anastamoses, occasionally leading to intrapericardial hemorrhage or hemothorax. Following type B dissection, up to 40% of cases will also lead to further aneurysmal dilatation over time. Many medically managed cases will ultimately require intervention, with a 41% intervention-free survival at 6 years for type B dissections. Therefore, clinicians must recognize that aortic dissection is truly a chronic disease that requires ongoing management and evaluation.

OTHER ACUTE AORTIC SYNDROMES

In addition to aortic dissection, aortic intramural hematoma (IMH) and penetrating atherosclerotic ulcer (PAU) are also considered acute aortic syndromes with similarly high risks of death. IMH is thought to be caused by bleeding into the vessel wall from the vasa vasorum. The blood collection can expand on its own and cause vessel wall weakness or it can lead to an intimal tear and aortic dissection. Overall, these represent ~5% to 10% of acute aortic syndromes. In general, patients with IMH tend to be older, with the mean age of detection being ~68. They can occur at any point along the length of the aorta and are classified using the Stanford system as in aortic dissection. Patients may present with chest or back pain, and both CT and MRI are excellent imaging modalities for detection and monitoring. IMHs are typically considered to be at similar risk and treated as an aortic dissection in the same region. Meta-analysis data revealed an overall mortality rate of 21% with Type A IMH being higher risk than Type B. Some IMHs may progress to aortic dissection and rupture, while others may remain stable or even involute over time. The additional presence of a PAU increases the risk of progression, whereas individuals with younger age and smaller aortic diameters are at lower risk. Most centers recommend urgent surgical intervention for Type A IMH and medical therapy with beta blockade and tight blood pressure control for Type B IMH. However, as with aortic dissection, the decision must incorporate numerous factors, and we recommend input from a variety of specialists as part of a multidisciplinary team to determine the best approach.

Penetrating atherosclerotic ulcers occur as sequelae of aortic atherosclerotic disease. Over time, atherosclerotic plaques can erode into the vessel wall, thus leading to focal weakness and, in some cases, intramural hematoma or wall rupture. Many PAUs are discovered incidentally, but some patients may present with chest or back pain as with aortic dissections. The natural history of PAU is unclear. Some centers report a high rate of rupture and recommend early surgical intervention, while others support a more conservative approach in most cases. Nonetheless, as with other types of aortic disease, progression varies considerably among individuals. PAUs should be regularly monitored on imaging, which is most easily achieved with contrast-enhanced CT or MRI. High-risk features, such as discovery coincident with pain, associated IMH, pseudoaneurysms, or true aneurysm, should prompt surgical consultation. Open surgical repair and endovascular grafts are both reasonable options, and there are no clear data supporting one approach over the other.

AORTITIS

Aortitis, or inflammation of the aorta, is broadly categorized as either autoimmune or infectious. Among the autoimmune causes, Takayasu’s arteritis is one of the most common. Patients with Takayasu arteritis more commonly develop inflammation of the abdominal aorta, although the ascending aorta and arch may also be involved. Histologically, there is infiltration of lymphocytes, macrophages, and giant cells into the media and adventitia layers. Eventually, this leads to extensive scarring and fibrosis of the vessel wall, which can ultimately cause stenosis of the affected segment. It more commonly affects individuals of Asian descent, and women are at ten times higher risk than men. Most cases present before age 40. Typically, there is both an acute and chronic phase of Takayasu’s arteritis. In the acute phase, individuals may experience fevers, chills, night sweats, and weight loss. Over time, ongoing inflammation leads to narrowing of the affected large vessel lumen. Thus, in the chronic phase, patients may exhibit diminished peripheral pulses, claudication, bruits, or discrepant blood pressures. Erythrocyte sedimentation rate and C-reactive protein values are often elevated but are nonspecific. Contrast-enhanced CT and MR imaging are useful in identifying the aortic segments and branch vessels involved. PET-CT may also be used to identify active inflammation as well as determine the success of therapy. Initial treatment typically consists of systemic steroids, although steroid-sparing regimens are often used as maintenance therapy. In some cases, surgical or endovascular therapy is necessary to correct narrowing of a large vessel.

Another commonly encountered cause of aortitis is giant cell arteritis. This involves not only the aorta but also large- and medium-sized branches. It is characterized by the presence of granulomas and giant cells throughout the aortic wall. This disease typically affects those over the age of 50 with a predilection for women and Caucasians. Patients often present with headache, visual disturbances, or jaw claudication. On examination, temporal artery tenderness or diminished pulse is common. Laboratory analysis is notable for a markedly elevated erythrocyte sedimentation rate. Imaging studies typically cannot differentiate this from other causes of aortitis, and temporal artery biopsy showing the presence of granulomas and giant cells often confirms the diagnosis. Treatment consists of long-term steroid therapy, and these individuals are at risk for subsequent aneurysm formation.

Behcet’s disease is a rarer cause of aortitis. This syndrome usually presents with oral ulcers, genital ulcers, uveitis, and skin lesions, and vascular involvement only occurs in approximately one-third of patients. The disease causes inflammation of the vasa vasorum with infiltration by lymphocytes, histiocytes, eosinophils, and giant cells. Although aortic inflammation and subsequent aneurysm formation can occur, Behcet’s can affect any number of arteries and veins, and patients may present with inflammation or aneurysm in multiple vessels. Treatment typically consists of corticosteroids. Depending on the vessel involved and the degree of aneurysmal dilatation, surgery or endovascular repair may be necessary. Other autoimmune disorders that can cause aortitis include rheumatoid arthritis, ankylosing spondylitis and other spondyloarthropathies, inflammatory bowel disease, and psoriatic arthritis.

Infectious aortitis is rare in the developed world but may have devastating consequences. It typically results from direct inoculation of the vessel wall through the vasa vasorum, although direct invasion and septic embolization are also possible. Infection typically occurs in a segment of the aorta with underlying pathology, such as aneurysm or atherosclerotic plaque. In addition to inflammation, infections may lead to abscesses, necrosis, mycotic aneurysm formation, or rupture. Patients may present with fever, chills, weight loss, or pain. CT and MRI are the primary means of identifying infective aortitis. Blood cultures are essential in identifying the causal organism. The most common bacterial causes of aortitis are Salmonella and Staphylococcus aureus, and fungal infections, such as Candida and Aspergillus are also possible, particularly in immunocompromised hosts. Mycobacterium tuberculae may cause aortitis that typically affects the distal aorta, likely through direct spread from adjacent thoracic structures. Although increasingly rare, syphilitic aortitis is a late complication of Treponema pallidum infection. It preferentially affects the ascending aorta and commonly leads to aneurysm formation several years after the initial infection. Prompt initiation of antibiotic therapy is essential. However, the active inflammation of infectious aortitis, particularly in cases of Gram-negative bacterial infection, leads to significant vessel wall weakening and an increased risk of rupture and death. In such cases, definitive repair also includes resection of the affected area and graft implantation.

SUGGESTED READINGS

EPIDEMIOLOGY

While specific estimates vary, it is clear that lower-extremity peripheral arterial disease (PAD) affects a substantial fraction of the US population. For example, a 1999 analysis of the National Health and Nutrition Examination Survey results demonstrated that 4.3% of Americans over the age of 40 years had PAD, defined as an ankle-brachial index (ABI) of less than 0.90, corresponding to over 5 million people. The PAD Awareness, Risk, and Treatment: New Resources for Survival (PARTNERS) program, a multicenter, cross-sectional US study, showed that 29% of its cohort of nearly 7000 patients at least 50 years old had PAD based on a history of revascularization or an identical ABI criterion.

Clearly, recognition of PAD is important to guide appropriate limb-based therapy. However, given its close association with coronary and cerebrovascular arterial disease, it is perhaps more critical as a reason to motivate aggressive management of cardiovascular risk, which is discussed below.

Nonmodifiable risk factors for PAD include increasing age and nonblack Hispanic race. Gender is controversial. Some studies have shown a male predominance with less severe disease and an equal gender ratio in more advanced disease while others have ranged from no difference at any severity to a female predominance.

Irrespective of the particular risk factors in any given patient an obstructive arterial lesion with diminished distal flow is the final common pathway of PAD. Typically, this is atherosclerotic disease in chronic ischemia but there is a large array of rare causes that require investigation when clinically appropriate (Table 256-1). Degree of obstruction, however, is not the sole determinant of patient symptomatology. Indeed, around 15% of asymptomatic men were found to have 50% stenosis in at least one leg artery at autopsy. In addition to severity, symptoms depend on the timescale of disease development. Rapid narrowing of a vessel produces more profound distal ischemia because of inadequate time for the development of collateral vessels. As such acute arterial occlusion, for example from cardioembolism, often leads to extreme ischemia requiring surgical intervention while slowly developing atherosclerosis may be asymptomatic even if extensive (Figure 256-1). Finally, symptoms will also depend upon the demands of the affected vascular bed. Lesions that are hemodynamically significant at rest may not be clinically apparent until exertion. For example, patients’ reports of cramping leg pain with walking that abates with rest—the so-called intermittent claudication—is described well by this mechanism. Once blood flow drops below a minimal threshold continuous ischemia will present as rest pain, ulceration, or gangrene.

RISK STRATIFICATION

All patients with acute limb threatening ischemia should be admitted. Patients with chronic limb ischemia do not universally require admission. Indications for admission include pain control, assessment, and therapy of ulcers and pedal infection including surgical debridement, and, expedited workup.

ACUTE LIMB ISCHEMIA

EVALUATION

Acute limb ischemia is caused by sudden compromise of extremity flow and comprises a variety of presentations and etiologies. Management is driven by the degree of ischemia and is commonly scored using the Rutherford scale for acute limb ischemia (Table 256-2). A focused and expeditious history and physical are essential. The history should focus on the duration of symptoms (prolonged time until revascularization favors the use fasciotomies, long-standing symptoms may suggest a chronic process), location and laterality of symptoms, conditions that would increase the risk of anticoagulation (eg, bleeding dyscrasias, trauma, recent surgery), arrhythmias, and endovascular instrumentation, including cardiac catheterization. The patient should receive a physical examination that, in addition to the typical cardiopulmonary examination, includes palpation of the abdomen for an aneurysm and examination of bilateral femoral, popliteal, pedal, and upper-extremity pulses. The pulse examination will suggest the location of the lesion. Furthermore comparison of the examinations between the two legs may suggest an etiology. For instance, a normal pulse examination on one leg with a normal contralateral femoral pulse and absent distal pulses in a patient with atrial fibrillation suggests embolization distal to the common femoral artery on the affected side. However, absent bilateral popliteal and pedal pulses with unilateral acute ischemia in the same patient may point toward in situ thrombosis in a background of chronic disease. A basic sensorimotor examination of the foot and leg provides a critical clue to the severity of the ischemia and the urgency of intervention. Sensory loss limited to the toes with intact motor function requires urgent intervention whereas more extensive sensory loss or any motor loss indicates more advanced ischemia and requires emergent intervention. The affected limb should be examined for compartment syndrome, including pain in the leg with passive or active ankle movement, pain with direct palpation of the leg, and numbness in the webspace between the first and second toes. Last, one should take into account the general appearance of the extremity. Common findings include pallor and mottling. Nonblanching erythema or ecchymosis over the leg calf is an ominous sign associated with advanced muscle ischemia. In rare cases of delayed presentation the patient may present with rigor mortis of the leg, which constitutes class III ischemia and is a contraindication to revascularization. The affected limb may also be cool to the touch below the level of obstruction. Despite the informal use of “cold leg” to refer to acute limb ischemia, physical coolness is nonspecific and not uniformly present.

Laboratory work is more directly related to preparation for intervention rather than diagnosis. For example, it is important to know the glomerular filtration rate, in case angiography is planned. Likewise, deranged electrolytes and coagulation parameters should be recognized and corrected as clinically appropriate. Creatine phosphokinase is sometimes ordered in the emergency room and a grossly elevated level would favor fasciotomies but this is uncommon with early presentation and the decision ultimately relies on clinical examination and judgment.

Vascular imaging is not mandatory prior to intervention and should not be obtained unless it could fundamentally alter management. In general, this decision should be made in concert with the consulting vascular surgery team. The most useful and commonly obtained noninvasive imaging is computed tomographic angiography (CTA) of the abdominal aorta with runoff views of the lower-extremity vasculature. The radiologist should be informed that the indication is acute limb ischemia so that delayed phase imaging is included in the protocol.

INPATIENT MANAGEMENT

The diagnosis of acute limb ischemia should immediately prompt vascular surgery consultation (Figure 256-2). In addition, to augment perfusion the affected limb should be placed in a dependent position and the patient started on supplemental oxygen, irrespective of oxygen saturation. In the absence of contraindications, intravenous heparin should be started to prevent thrombus propagation with a goal-activated partial thromboplastin time of two to three times above normal. Many institutions have established heparin protocols but in the absence of these the patient may be given a bolus of 80 to 100 m/kg followed by an infusion of 18 m/kg/h. Heparin administration will need to be individualized for patients at high risk for hemorrhagic complications: recent surgery, malignancy at risk for bleeding, and the like. Patients with known heparin induced thrombocytopenia will require direct thrombin inhibitor therapy, most commonly bivalirudin or argatroban.

Prompt vascular surgery consultation is important to determine the need for additional imaging as discussed above and surgical intervention. Patients with class I ischemia—no neurologic defect with Doppler-proven arterial flow—do not have an acutely threatened limb and warrant a workup in addition to anticoagulation. Patients with an irreversibly ischemic leg will benefit from heparin to prevent thrombus propagation but not from aggressive limb salvage. They should receive amputation when medically appropriate. All other patients—meaning those with sensorimotor loss and a viable limb—need revascularization immediately. Broadly, this may consist of open surgery (eg, thromboembolectomy, arterial bypass) or endovascular therapy (eg, percutaneous mechanical thrombectomy, thrombolysis), or both as determined by patient factors and local expertise.

History, examination, and findings at surgery should help guide an investigation into the etiology. Common possibilities include in situ thrombosis occurring in the context of advanced PAD. This will require little beyond aggressive PAD risk factor modification (discussed below). Patients with suspected embolization should have an embolic workup, including ECG for arrhythmia and echocardiography looking for cardiac thrombus or causal valvular disease. If these do not demonstrate a source, CTA or magnetic resonance imaging of the thoracic and abdominal aorta are indicated. Failing to identify any clear cause, patients should undergo investigation for a hypercoagulable condition. Examples include lupus anticoagulants, heparin-induced thrombocytopenic thrombosis, protein S deficiency, and prothrombin G20210A mutations. Guidance from vascular medicine or hematology should be obtained. Some of the salient laboratory tests may be confounded by ongoing heparin or warfarin use and should be drawn in advance of instituting these therapies if at all possible.

CHRONIC LIMB ISCHEMIA

EVALUATION

Chronic limb ischemia ranges widely in presentation from totally asymptomatic to claudication to critical limb ischemia, which includes rest pain, ischemic ulceration, and gangrene. Intermittent claudication classically consists of cramping muscle pain that is relieved within a few minutes of rest, though it should be noted that some reports have found that only a minority of patients describe all of these features. Claudication may occur in the buttocks, thigh, or calf and reflects ischemia one level below the area of disease, for example, a culprit superficial femoral artery will produce calf claudication. Rest pain is frequently described as a continuous burning pain in the foot, particularly the forefoot, that characteristically occurs at night when the foot is elevated in bed. Patients will also report relief with placement of the foot in the dependent position, which mechanically increases distal perfusion. Ulceration in the appropriate clinical context may represent PAD but caution is warranted. Patients with PAD often have a variety of potential causes and arterial insufficiency may be superimposed upon these or be unrelated. Examples include venous stasis ulcers, commonly associated with hemosiderin staining of the distal leg and foot and occurring on the medial distal leg and foot, neuropathic ulcers occurring in an area such as the head of the first metatarsal subject to trauma and often associated with a callus, or pressure ulceration as on the heels in bedbound patients. In addition to history and examination, the vascular testing discussed below will help clarify whether there is a component of arterial insufficiency.

As in acute limb ischemia a vascular examination should be performed along with tests of sensory and motor function. Areas of tissue loss and signs of infection (purulent exudate, periwound erythema, fluctuance) should be noted. Although a full discussion of diabetic foot infection is beyond the scope of this chapter, it should be noted that excision of necrotic tissue and drainage of sepsis are urgently indicated in infected diabetic ulcer (see Chapter 147 [Diabetic Foot Ulcers]).

There is no chronic ischemia-specific laboratory examination. Labs should be ordered as clinically indicated, for example, serum creatinine to assess baseline renal function if intervention or contrast-based imaging is contemplated or a CBC to look for a leukocytosis in cases of infection.

Whether PAD is responsible for the patient’s symptoms may not be apparent in all cases. Conditions such as neurogenic ­claudication—intermittent lower-extremity pain from, for example, spinal stenosis—may be viable alternative explanations in some cases. In all except the most unequivocal cases, it is appropriate to obtain confirmatory noninvasive vascular tests. The mainstays of testing are ankle-brachial index (ABI), segmental Doppler pressures, and pulse volume recordings (PVR).

Noninvasive tests serve the dual purposes of confirming the presence of arterial insufficiency as the etiology and elucidating the location and severity of the causative lesions. The most basic test is the ABI. Although the mean arterial pressure decreases as one progresses distally in the vasculature, the systolic pressure increases as a result of pressure waves reflected from the distal circulation. Thus a decrement in the ankle systolic pressure is suggestive of occlusive disease somewhere between the heart and the blood pressure cuff. There are, however, alternative reasons for a reduced ankle pressure unrelated to occlusive disease such as hypotension. To reduce the effect of this factor, the ankle pressure should be normalized by dividing it by the brachial pressure, which is a proxy for aortic pressure, as brachial pressures are much less commonly reduced when compared with lower-limb pressures. Practically, the ABI is derived by inflating a cuff placed above the ankle to above the systolic pressure and slowly deflating the cuff until the Doppler signal of the posterior tibial artery is again heard. The procedure is again repeated for the dorsalis pedis artery and the higher of the two measurements is used as the ankle pressure or the numerator. The systolic pressure is measured in both arms and the higher of the two values is again used as the denominator in the calculation of the ABI. It is important to note that the patient should be supine for these measurements to avoid the confounding effects of hydrostatic pressure. Normal ABI ranges from 1 to 1.30. An ABI between 0.41 and 0.90 is typical of patients with claudication while those with critical limb ischemia will often have values less than 0.41. Intermediate values of 0.91 to 0.99 are borderline while those greater than 1.30 suggest noncompressible blood vessels due to, for example, medial calcific disease often seen in diabetics. It is important to remember, however, that in patients with extensive calcification and occlusive disease, the ABI may be spuriously elevated and may appear normal despite the presence of severe occlusive disease. For this reason, it is essential to combine this pressure measurement with more extensive testing and a thorough examination. One should not, for example, accept as normal an ABI of 1.0 in the face of absent pedal pulses.

Patients with spuriously elevated ABIs should have alternative testing. The vasculature of the toes is often spared despite more proximal calcific disease, an observation that forms the basis of the toe brachial index. Toe pressures may be measured with small cuffs and again normalized using the brachial pressure. A normal toe brachial index is at least 0.65 and patients with severe limb ischemia will have toe pressures less than 30 mm Hg.

An ABI of less than 0.90 has a sensitivity and specificity of 95% and 100% respectively for a 50% stenosis. Nevertheless it is possible for patients with significant peripheral artery to have normal resting ankle pressures. Conceptually, such occult disease may be detected by increasing blood flow velocity, which increases turbulence around a stenosis and produces energy losses and potentially a new pressure decrement. Exercise testing is therefore used for cases with a high suspicion for arterial disease despite a normal ABI. Patients are walked on a treadmill at 2 miles per hour at an incline of 10° to 12° until claudication occurs or a maximum of 5 minutes after which ABIs are measured serially until return to baseline. The degree of ABI drop and the time to normalization roughly correlate with severity of disease.

Although ABI is effective at establishing the presence of arterial insufficiency it does not localize disease. Segmental limb pressures may provide anatomic information indirectly and are obtained by placing a series of cuffs on the leg, which are then inflated and deflated as for the ankle pressure while listening for a pedal signal. Standard cuff positions include high thigh, above-knee, below-knee, and ankle. A drop of 20 mm Hg or more compared to the immediately preceding segment suggests significant disease. There are several drawbacks to the use of segmental pressures. First, extra-axial vessel disease may not be detected, for example, severe profunda femoris stenosis. Second, collaterals may reduce a pressure drop and thus obscure disease. Last, distal disease may obscure the detection of more proximal disease by reducing flow and thus reducing a pressure decrement. This may occur, for example, in a patient with simultaneous common femoral and common iliac artery stenosis.

Air plethysmography, more commonly known as pulse volume recording in this context, determines changes in lower-extremity volumes with each cardiac cycle beneath the cuff to infer disease location and severity. Normal PVRs consist of a brisk upstroke, a dicrotic notch, and then a final downward stroke. Although a full discussion of PVRs is beyond the scope of this discussion, disease proximal to the cuff location will diminish the amplitude of the initial upstroke and blunt the waveform morphology (Figure 256-3). Thus an examination of a series of PVRs progressing from proximal to distal in an extremity should help clarify the locations and magnitude of disease. Unlike the ABI, PVRs should not be affected by calcific disease. However, they are subject to all of the disadvantages seen with segmental pressures (eg, insensitivity to extra-axial disease).

In practice, ABI, segmental Doppler pressures, and PVRs are often combined. For example, our vascular laboratory provides these as a single composite test. Additionally, interpretation of these studies is often part and parcel of the report and should be performed by an individual familiar with the intricacies of the tests and the evaluation of these studies. Noninvasive physiologic testing as outlined above is followed by additional imaging—duplex ultrasound, computed tomographic angiography, magnetic resonance angiography, conventional angiography—to produce a final treatment plan. A firm understanding of the principles, strengths, and disadvantages of each of these tests is crucial to formulating a rational diagnostic and therapeutic plan. For example, CTA would be reasonable to assess iliac in-stent disease while MRA would be grossly inadequate because of metal artifact. In a patient with renal insufficiency, however, invasive angiography with limited or no iodinated contrast and the ability to simultaneously intervene may be the best test. Vascular surgery consultation should be obtained in advance of ordering these additional tests to avoid unnecessary testing.

INPATIENT MANAGEMENT

Intermittent claudication has a benign natural history with respect to the affected limb: limb loss occurs in 1% to 3% of patients over 5 years. As such, these patients generally do not need inpatient admission and will benefit from outpatient evaluation. Initial therapy frequently consists of a supervised exercise plan, risk factor modification, and occasionally adjunctive cilostazol. In sharp contrast, patients with critical limb ischemia without revascularization may have up to a 40% limb loss rate at 6 months. As such, prompt evaluation and surgical or endovascular intervention is indicated (Figure 256-4). Even so, critical limb ischemia is not intrinsically an absolute indication for admission. Legitimate reasons for admission include pain control, as for intractable rest pain, wound care, and expedited workup and therapy for patients who are unable to comply with outpatient follow-up or who are at risk for impending limb loss.

Wound care options have recently proliferated and a complete treatment of wound care is beyond the scope of this work. In addition to debridement, common wound care strategies include wet-to-dry dressings with normal saline, sterile water, or dilute Dakin’s (sodium hypochlorite) solution for grossly infected wounds. The latter should not be used for more than 5 days to avoid injuring healthy tissue. Wound vacuum dressings have been shown to aid wound care and healing in diverse settings and may reduce the burden of care and time to resolution in vascular patients. Patients with extensive devitalized or infected tissue may need surgery to achieve a more expeditious aggressive debridement than is possible with wet-to-dry or enzymatic methods (eg, enzymes). Vascular surgery, plastic surgery, wound care services, or other local expertise will be able to provide additional input.

Although only lifestyle-limiting claudication and critical limb ischemia demand surgery, all patients with PAD need aggressive medical management. Consider as evidence the 5-year 20% combined myocardial infarction and stroke rate and 10% to 15% mortality in claudicants. Figures for critical limb ischemia are even more dismal with a commonly cited 25% 1-year mortality, again primarily from cardiovascular events.

Typical recommendations include smoking cessation, weight loss for overweight and obese patients, antihypertensive therapy, diabetes management, and lipid-lowering therapy. Significantly, statin therapy has been shown to improve a variety of outcomes, including cardiovascular adverse event rates and all-cause mortality, in PAD patients. Patients benefit irrespective of initial lipid levels and therefore all PAD patients should be placed on statin therapy. Although the evidence for routine antiplatelet therapy has been more ambiguous, general recommendations include daily aspirin therapy for all PAD patients. There is no clear benefit to 325 mg over 81 mg. Patients already on clopidogrel for other reasons do not derive cardiovascular benefit from the addition of aspirin although there is evidence of benefit in selected vascular patients (eg, prosthetic infrapopliteal bypass). In general clopidogrel monotherapy is adequate. Anticoagulation is not recommended for general PAD risk reduction but may be instituted for specific patients after intervention. Regrettably several studies have found widespread underuse of antiplatelet agents and statins despite their favorable risk profiles.

Finally, revascularization for patients with lifestyle-limiting claudication or critical limb ischemia may be performed via open surgery, endovascular therapy, or a hybrid approach. The intervention strategy will depend on patient anatomy, lesion characteristics, medical risk, patient preference, and local expertise.

POSTACUTE CARE

Postacute care will vary immensely from patient to patient but should include follow-up with vascular surgery, either for routine follow-up or additional workup, and primary care follow-up for cardiovascular risk factor modification.

DISCHARGE CHECKLIST

1. Have all of the patient’s medical risk factors for PAD been addressed? In particular, is the patient on an appropriate antiplatelet agent and statin?

2. If applicable, is there a wound care plan?

3. Is there follow up with a vascular surgeon?

SUGGESTED READINGS

INTRODUCTION

The diagnosis of vasculitis is considered far more often than these diseases are actually seen—appropriately, since these rare diseases have diverse presentations, are dangerous and often rapidly progressive, yet are treatable. All vasculitides feature inflammation of the walls of blood vessels, with symptoms and findings attributable to the size(s) and locations of the involved vessels. Vessel size, epidemiology, and/or a constellation of typical clinical features have proved useful for classifying the vasculitides, with key points summarized in (Table 257-1).

The classification of vasculitis remains imprecise with more than one system in use. “Primary” systemic vasculitides are further sorted by the sizes of the predominant arteries involved, leading to sets of small-, medium-, and large-vessel vasculitides. There are also several “secondary” forms of vasculitis associated with infections, drug exposure, or another form of systemic disease such as systemic lupus erythematosis or rheumatoid arthritis.

All of the primary, idiopathic vasculitides are rare diseases, each categorized as an “orphan” disease in the United States (prevalence of <200,000 persons). However, as a group, and especially among patients with initially unexplained systemic disease who are ill enough to be hospitalized, vasculitis is common enough that hospitalists will likely encounter cases.

PATHOPHYSIOLOGY

The pathophysiology of the primary vasculitides is poorly understood. They are all considered immune mediated, based on histologic and serologic studies and by analogy to the many cases of vasculitis that have occurred secondary to infections or drug exposure. Since no pathogens have been found, the primary vasculitides are suspected to be autoimmune, with the humoral and cellular branches of the adaptive immune system involved to different degrees in different diseases.

NOTES ON INDIVIDUAL FORMS OF VASCULITIS

It is impractical to describe each of the vasculitides in detail in a single chapter, but short summaries of most of the vasculitides are provided as reference points for the subsequent discussion.

Giant cell arteritis

Giant cell arteritis (GCA) only affects adults over the age of 50, and incidence markedly increases with age, to about 1:2000 per year among Caucasians of Northern European descent by age 80. Headache is the most common symptom (80%), with scalp tenderness, jaw claudication, constitutional symptoms, and polymyalgia rheumatica (pain and stiffness in the shoulders, hips, and/or neck) each seen in 40% to 70% of patients. The presentation of GCA is often insidious and diagnosed during outpatient care, but sometimes is acute and leads to hospitalization. GCA has also been reported to be an important cause (about 10%) of fever of unknown origin in the elderly, and since the aorta and its major branches are involved in more than 15% of cases, GCA is also an important consideration in cases of claudication, loss of pulse, and asymmetric blood pressure measurements in the elderly. (See the section on laboratory testing [acute phase reactants], imaging [angiography, ultrasound], and biopsy [temporal artery]).

Takayasu arteritis

Takayasu arteritis is a rare disease (prevalence <1:100,000), usually presenting in young adults, and 90% of patients are female. Patients often present with claudication, dizziness, abnormal blood pressure readings, and/or bruits. Diagnosis is confirmed by angiography. The types of presentation more likely to lead to hospitalization (eg, symptoms of coronary artery disease or cerebrovascular disease) carry a broad differential diagnosis, so suspicion for or against Takayasu may initially be gained by checking pulses and blood pressures and listening for bruits. (See the section on laboratory testing [acute phase reactants] and imaging [angiography, ultrasound]).

Granulomatosis with polyangiitis (Wegener’s) and microscopic polyangiitis

Granulomatosis with polyangiitis (Wegener’s) (GPA) and microscopic polyangiitis (MPA), often grouped together as antineutrophil cytoplasmic antibody (ANCA)-associated vasculitis (AAV), are small-vessel vasculitides that together affect ~1:20,000 persons. AAV is high in the differential diagnosis of patients presenting with pulmonary hemorrhage, acute renal disease, or acute peripheral neuropathy, and is also an important consideration in patients with palpable purpura, inflammatory eye disease (scleritis or episcleritis), acute sensorineural hearing loss, cutaneous ulcers, or digit ischemia. Renal disease in AAV often leads to elevated serum creatinine and presence of urinary RBC casts; however, in the early stages of renal disease with AAV patients may only have microscopic hematuria and modest proteinuria. GPA usually (70%-80% at initial presentation) also features granulomatous inflammation of the upper airway, with nasal discharge and crusting, epistaxis, facial pain, and/or hearing loss, and the presence of these symptoms often facilitates the diagnosis of GPA. GPA is an important consideration in cases of subglottic stenosis, orbital mass with proptosis, or pulmonary nodule(s). However, many patients with AAV, particularly MPA, present only with nonspecific symptoms of malaise, myalgias, and arthralgias. The role of testing for ANCA in diagnosing AAV is discussed in detail below. (See section on laboratory testing [autoimmune serologies, acute phase reactants, tests of renal function], imaging [chest imaging, sinus imaging], and biopsy [skin, nasal cavity and sinuses, kidney, lung, peripheral nerve and muscle]).

Eosinophilic granulomatosis with polyangiitis (Churg-Strauss)

Eosinophilic granulomatosis with polyangiitis (EGPA) is worth considering in a patient who is ill with what might be vasculitis, has a history of asthma (often severe or poorly controlled), and has eosinophilia. The presentations of EGPA most likely to lead to hospitalization are asthma exacerbation, pulmonary infiltrates, or acute peripheral neuropathy, but constitutional symptoms, nasal polyps, rash, and arthralgias are also common. Myocarditis, gastrointestinal inflammation or ischemia, stroke, and renal involvement (similar to AAV) are less common but important presentations. Positive tests for ANCA are present in 40% of patients and EGPA is considered part of the spectrum of ANCA-associated vasculitis. (See section on laboratory testing [autoimmune serologies, acute phase reactants, tests of renal function, and complete blood count (CBC)], imaging [chest imaging and sinus imaging], and biopsy [skin, nasal cavity, and sinuses, kidney, lung, peripheral nerve, and muscle]).

Antiglomerular-basement-membrane disease

In contrast to AAV, symptoms of antiglomerular-basement-­membrane (GBM) disease are limited to the kidneys and/or lungs (if both = Goodpasture syndrome), with rapidly progressive glomerulonephritis and/or pulmonary hemorrhage as manifestations. In the right setting, a positive test for serum antiglomerular-basement-membrane (anti-GBM) antibodies is diagnostic for this disorder, as is demonstration of the characteristic pattern of antibody deposition by immunofluorescence on kidney tissue obtained by biopsy. Testing for anti-GBM antibodies should be rapidly obtained for any patient for whom anti-GBM disease is possible, since the disease can lead to irreversible kidney damage or death from pulmonary hemorrhage. (See section on laboratory testing [autoimmune serologies, tests of renal function], and imaging [chest imaging], biopsy [kidney, lung]).

Polyarteritis nodosa

The most common features of polyarteritis nodosa (PAN) are livedo reticularis (a lacy pattern of cutaneous blood vessels on extremities that is more commonly due to benign conditions rather than vasculitis), painful cutaneous nodules or ulcers, hypertension, abdominal pain, myalgias, peripheral neuropathy, testicular pain, and digit ischemia. There is a strong association between hepatitis B virus (HBV) infection and PAN, but only in areas where HBV is endemic; the incidence of PAN has markedly declined in countries with high rates of HBV vaccination. PAN is probably the most challenging type of vasculitis to diagnose, because (1) it is particularly rare (in regions where HBV infection is also rare), (2) it features the least specific array of symptoms and findings among the vasculitides, and (3) there are no helpful diagnostic blood tests analogous to ANCA, cryoglobulins, or even eosinophil count. Biopsy or angiography is usually required for the diagnosis of PAN. For all of these reasons, PAN is contemplated much more often than it is diagnosed. (See section on laboratory testing [acute phase reactants, tests of renal function, selected testing for infectious diseases], imaging [angiography], and biopsy [skin, kidney, peripheral nerve and muscle, other organs]).

Behçet disease

The manifestations of Behçet disease (BD) that may lead to hospitalization are extraordinarily diverse: vision loss or a red and painful eye, deep vein thrombosis (including dural sinus thrombosis or Budd-Chiari syndrome), colitis resembling inflammatory bowel disease, meningoencephalitis, or pulmonary artery rupture leading to massive hemorrhage. Considering the diagnosis of BD in these settings requires recognition of the various milder but more common features of BD that may be concurrently present or reported by the patient, including recurrent oral ulcers (required for diagnosis), genital ulcers, erythema nodosum, acneiform skin lesions, inflammatory arthritis, and superficial thrombophlebitis. The diagnosis or exclusion of BD is always made on clinical grounds since there are no specific blood tests or biopsy features of the disease.

Primary angiitis of the central nervous system

This rare disease is limited to the central nervous system (CNS) and presents with symptoms of encephalopathy, multiple small strokes, and often headache. Angiography may be normal since the disease often affects only small vessels. Furthermore, angiography cannot easily distinguish vasculitis from atherosclerosis or from cerebral vasospasm (which usually presents with a severe headache of sudden onset). Thus, diagnosis of CNS vasculitis is ideally made by brain biopsy. (See section on imaging [angiography], and biopsy [brain]).

Cryoglobulinemic vasculitis

Cryoglobulinemia encompasses a variety of clinical syndromes associated with circulating immune complexes that precipitate at cold temperatures. The majority of cases of cryoglobulinemia (~70%) are associated with chronic hepatitis C virus infection, with plasma cell malignancies and systemic rheumatic diseases accounting for most other cases. Cryoglobulinemia associated with chronic hepatitis C virus infection or systemic rheumatic disease causes a small-vessel vasculitis that typically involves skin (palpable purpura and/or ulcers), commonly involves peripheral nerves or kidneys (glomerulonephritis), and less commonly involves a wide range of internal organs. (See section on laboratory testing [tests of renal function, paraproteins, selected testing for infectious diseases], and biopsy [skin, kidney, peripheral nerve and muscle]).

IgA vasculitis (Henoch-Schönlein)

IgA vasculitis (Henoch-Schönlein purpura, IgAV) is much more common among children than adults. It is usually self-limited and presents with palpable purpura, often in association with inflammatory arthritis and abdominal pain. A minority of patients has severe renal involvement, and the presence or absence of significant proteinuria or elevated creatinine is often used to determine whether to treat with immunosuppressive agents; however, the usefulness of such treatment has not been firmly established. (See section on laboratory testing [tests of renal function] and biopsy [skin, kidney]).

Kawasaki disease

Kawasaki disease (KD) is almost exclusively a disease of young children, and it is not rare in its characteristic age group (incidence 1:10,000 in the United States). The syndrome classically includes fever persisting for at least 5 days, erythematous rash with desquamation in the hands and feet, conjunctivitis, oral mucosal inflammation, and lymphadenopathy. However, many patients present without all of these features and, since a consequence of untreated KD is life-threatening coronary artery aneurysms, the diagnosis is considered in almost any young child with a fever lasting 5 days.

Vasculitis associated with other autoimmune diseases

Systemic lupus, rheumatoid arthritis, Sjögren syndrome, and inflammatory bowel disease have all been associated with vasculitis, mostly involving small vessels (palpable purpura, ulcers, neuropathy), but occasionally medium-sized vessels (analogous to PAN). Inflammatory bowel disease, relapsing polychondritis, and Cogan syndrome are each rarely associated with large-vessel vasculitis similar to Takayasu arteritis. (See section on laboratory testing [autoimmune serologies and tests of renal function]).

DIAGNOSIS: DOES THIS PATIENT HAVE VASCULITIS?

Vasculitis is commonly in the differential diagnosis for any patient with an unexplained systemic illness or any patient with one or more specific findings suggestive of vasculitis. It is important to realize that vasculitis is rare. Nonetheless, being open to the possibility of vasculitis is critical for reducing delays in diagnosis. There are many “red flags” that should alert clinicians that vasculitis must be considered (Table 257-2). Some of these findings merit urgent diagnostic testing and sometimes even empiric treatment for vasculitis, pending the results of testing.

Not shown in Table 257-2 are syndromes that could be caused by vasculitis but have other causes in the vast majority of cases such as fever of unknown origin, lower-extremity ulcers, abdominal pain, arthralgias, myalgias, and other nonspecific symptoms.

MIMICS OF VASCULITIS

Since vasculitis is rare, nonvasculitic causes of red-flag symptoms remain more common than vasculitis in many cases. A broad range of infectious, autoimmune, toxic, malignant, thromboembolic, vasospastic, and atherosclerotic disorders can mimic many of the characteristic features of vasculitis, as outlined in general terms in Table 257-2.

DIAGNOSTIC TESTING FOR SUSPECTED VASCULITIS

The method for diagnosing vasculitis generally depends on the size of vessels involved. Small-vessel vasculitis is usually diagnosed by biopsy, but sometimes the combination of a typical clinical syndrome and a laboratory test (see preceding section) is sufficient. Medium-vessel vasculitis (PAN) is diagnosed by biopsy or angiography. GCA is usually diagnosed by biopsy (temporal arteritis) but occasionally by angiography if the aorta or its major branches are involved. Takayasu is diagnosed by angiography. CNS vasculitis is usually diagnosed by biopsy but sometimes by angiography. Behçet disease and Kawasaki disease are diagnosed based on the clinical syndrome.

The uses of laboratory testing, diagnostic imaging, and biopsy for diagnosing vasculitis are outlined below.

LABORATORY TESTING FOR EVALUATION OF SUSPECTED VASCULITIS

Although individual laboratory tests on their own are almost never diagnostic for vasculitis, such tests are essential in the evaluation of a patient in whom vasculitis is being considered, for several reasons: (1) in the proper setting, selected serologic tests may confirm a diagnosis of vasculitis; (2) laboratory testing may identify organ systems involved in the disease process (especially for renal disease); (3) laboratory tests may establish a diagnosis other than vasculitis.

Autoimmune serologies

Testing for various autoantibodies plays a key role in the evaluation of patients suspected of having vasculitis, but over-reliance on such testing often leads to delayed or missed diagnoses.

Antineutrophil cytoplasmic antibodies

The widespread availability of testing for ANCA has had a striking impact on the evaluation of patients with suspected small-vessel vasculitis. The current standard of care for ANCA testing includes combining tests for ANCA by immunoflourescence and tests for specific ANCA autoantibodies by ELISA. Only two combinations of test results constitute a truly “positive” ANCA test: the cytoplasmic (C-ANCA) immunoflourescence pattern combined with antiproteinase 3 (PR3) antibodies by ELISA, or the perinuclear (P-ANCA) immunoflourescence pattern combined with antimyeloperoxidase (MPO) antibodies by ELISA. In the proper clinical settings, positive ANCA testing (C/anti-PR3 or P/anti-MPO) is highly specific for the diagnosis of AAV (>95%) and may preclude the need for biopsy or at least support initiation of treatment with glucocorticoids pending a biopsy. However, ANCA testing is presumed to have lower specificity in the setting of nonspecific symptoms such as malaise, myalgias, and arthralgias—a relatively common initial presentation of AAV. The sensitivity of ANCA testing is ~90% in patients with systemic GPA or MPA and is considerably lower (70%) in patients with GPA limited to the airway, so a patient may still have GPA even if the ANCA test is negative. Thus, ANCA testing is extremely useful diagnostically but must be used in the context of supportive clinical and other data.

Antiglomerular-basement-membrane antibodies

In the setting of renal insufficiency and/or pulmonary hemorrhage, anti-GBM antibodies are highly specific for the diagnosis of anti-GBM disease/Goodpasture disease as long as the initial anti-GBM antibody screening test is confirmed by Western blot. Because the treatment of choice for anti-GBM disease is the combination of high-dose glucocorticoids and plasma exchange, it is crucial to establish the diagnosis as early as possible. Antibody testing can play a central role in this process, but sometimes immunoflourescence staining of a kidney biopsy can make the diagnosis more rapidly.

Antinuclear antibodies

Testing for antinuclear antibodies (ANA) is quite useful when there is suspicion of systemic lupus erythematosus (SLE), either underlying vasculitis or as an alternative diagnosis for pulmonary hemorrhage and/or acute renal failure. ANA testing is >95% sensitive for the diagnosis of lupus; therefore, a negative test essentially rules out lupus as a diagnosis, unless a patient has clear pathognomonic features.

Acute phase reactants

Although the erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) are usually elevated in patients with untreated vasculitis, and ESR and CRP are often ordered when evaluating patients with suspected vasculitis, these tests have a surprisingly low value diagnostically. ESR or CRP are neither sensitive nor specific enough to be considered as screening tests to make or exclude a diagnosis of vasculitis, except, in a limited way, as part of a diagnostic algorithm for GCA. Even for new-onset GCA, the ESR or CRP may be normal in ~15% of cases, and both are normal in ~4% of cases. Therefore, a normal ESR is only helpful in ruling out GCA and avoiding the need for a temporal artery biopsy in a low-probability patient (atypical headache only), but not in a patient with additional cranial, arthritic, or constitutional symptoms. These tests may have selected roles in evaluation for persistent or recurrent disease but still must be interpreted cautiously. Furthermore, ESR and/or CRP are often elevated in the major mimics of vasculitis, especially infections (either acute or chronic) and malignancies.

Tests of renal function and other chemistry laboratory testing

Prompt assessment of renal function is an essential part of the evaluation of any patient suspected of having vasculitis. Renal disease can be either insidious or fulminant, and renal injury is usually asymptomatic until symptoms of end-stage renal disease arise. Serum creatinine with estimate of glomerular filtration rate should be done immediately in all patients and compared with prior measurements. Even subtle rises in creatinine levels within the “normal” range may be indicative of serious disease.

If urinary dipstick testing demonstrates the presence of any blood or protein, then a microscopic examination of the urine sediment on a freshly collected specimen must be performed for any patient suspected of having vasculitis. However, it is essential that the urinalysis be performed by someone experienced in evaluating patients with glomerulonephritis, usually a nephrologist or rheumatologist. It is the norm for hospital-based and reference laboratories to miss evidence of an “active” urinary sediment since casts are often destroyed by the time the specimens are processed, and laboratory personnel are simply not trained to detect these changes. The presence of red blood cell casts and, to a lesser extent, “dysmorphic” red cells in a urine specimen are strong indicators of glomerular disease. However, patients can still have highly active glomerulonephritis or other renal disease in vasculitis (eg, aneurysms in PAN) and not have urinary casts present.

Other routine chemistry panels, including measurements of electrolytes and liver function tests, are of limited value in diagnosing vasculitis per se and provide more useful information about nonvasculitic conditions.

Urine or serum toxicology screens for commonly used legal and illegal drugs of abuse may be appropriate for some clinical situations in which vasculitis is suspected. For example, both cocaine and methamphetamines have been associated with vasculitis and/or arterial vasospasm, and levamisole is now a common adulterant of cocaine and can cause a form of AAV.

Complete blood count

A complete blood count (CBC) should be conducted on all patients suspected of having vasculitis but is usually of limited diagnostic value. Many, but by no means all, patients with active vasculitis have anemia and/or thrombocytosis, but neither finding is sensitive or specific for the diagnosis. The one diagnostic use of the CBC in vasculitis concerns EGPA; although an elevated eosinophil count is nonspecific, a normal eosinophil count (in a patient who has not received glucocorticoids) usually rules out this diagnosis.

Paraproteins (abnormal immunoglobulins)

Abnormal immunoglobulins that can be associated with vasculitis include both those that precipitate at room temperature (cryoglobulins) and other clonal immunoglobulins that may cause small-vessel vasculitis. Cryoglobulins results in vasculitis are most closely associated with chronic infection with hepatitis C virus. Testing for cryoglobulins requires careful attention to specimen handling and processing with incorrect practice at any of several steps resulting in uninterpretable findings (ie, high false-negative rate due to poor processing). Similarly, standard serum protein electrophoresis testing (SPEP) may not pick up some immunoglobulin clones; immunofixation electrophoresis (IFE) is a more comprehensive screen for clonal immunoglobulins. Similar issues exist for urine protein electrophoresis (UPEP) versus urine immunofixation electrophoresis (UFE). The great majority of patients with cryoglobulinemic vasculitis tests positive for rheumatoid factor (RF) and/or has low levels of circulating complement proteins (C4 more so than C3), ~80% each for RF or low C4. RF has poor specificity, but results are obtained faster and more reliably than are tests for cryoglobulins.

Selected testing for infectious diseases

Hepatitis B, hepatitis C, and HIV virus infections are associated with vasculitis and tests for these viruses should be sent for all patients suspected of having a small- or medium-vessel vasculitis. Blood cultures are appropriate when bacteremia may be a possible cause of vasculitis. Several other infections may cause signs and symptoms mimicking vasculitis and serologic tests or cultures specific to those diseases are useful on a case-by-case basis.

DIAGNOSTIC IMAGING FOR EVALUATION OF SUSPECTED VASCULITIS

Chest imaging

The major forms of vasculitis that involve the lung are AAV (including GPA, MPA, and EGPA) and anti-GBM disease. A chest x-ray is therefore prudent in any patient with acute renal failure and an active urine sediment, and one should have a low threshold for performing a computed tomography (CT) test (noncontrast, but ideally with high-resolution cuts). However, the appearances of AAV on CT imaging—solitary or multiple nodules, multifocal or diffuse ­infiltrates—are not sufficiently specific to be diagnostic. For a patient with known vasculitis, comparison of abnormal findings on chest imaging to prior studies is essential.

Sinus imaging

A patient in whom GPA or EGPA is being considered due to prominent symptoms of nasal inflammation should have a CT (a noncontrast study is usually sufficient) performed to assess for sinus, mastoid, and retro-orbital involvement.

Angiography

Angiography, whether conventional or performed using CT (CTA) or magnetic resonance (MRA) techniques, is essential for the diagnosis of large-vessel vasculitis (Takayasu arteritis, GCA, and others) and many cases of medium-vessel vasculitis (PAN). For Takayasu and GCA involving the major branches of the aorta, either MRA or CTA is often sufficient. Angiography (CTA or conventional) is usually unable to differentiate between CNS vasculitis, vasospasm, and atherosclerosis.

Catheter-based conventional angiography is still considered the “gold standard” for arterial imaging and is considerably more sensitive for disease of smaller arteries. Furthermore, catheter-based angiography allows for measurement of blood pressures and selective sampling of specific arterial areas.

Ultrasound

Ultrasound of certain large arteries (particularly the carotids) can be helpful in diagnosing large-vessel vasculitis (GCA or Takayasu), since wall thickness as well as luminal diameter can be measured. Ultrasound of the temporal arteries for diagnosis of GCA has performed well in research studies at specialized centers but is not yet recommended for routine clinical use. Echocardiography, which allows visualization of the major coronary arteries in children, has an important role in the diagnosis and management of Kawasaki disease.

Positron emission tomography

Positron emission tomography (PET) scanning has been reported to identify areas of inflammation in large-vessel vasculitis in small studies, but the diagnostic usefulness of this modality for vasculitis has not been established.

BIOPSY FOR EVALUATION OF SUSPECTED VASCULITIS

Biopsy is required for the diagnosis of vasculitis in many cases, although there are important exceptions. Behçet disease and IgA vasculitis (Henoch-Schönlein) (in children) are diagnosed on clinical grounds. Kawasaki disease is diagnosed on clinical grounds, often with confirmation by imaging showing evidence of coronary artery aneurysm(s). Large-vessel vasculitis is diagnosed by angiography in patients with Takayasu and a subset of patients with GCA. Some patients with small-vessel vasculitides (GPA, MPA, EGPA, and cryoglobulinemia) can be diagnosed confidently on the basis of symptoms, signs, and laboratory and imaging findings but biopsy remains the best evidence for these diseases.

The usefulness of biopsy either in diagnosing vasculitis or in diagnosing a particular type of vasculitis depends on the organ system. Consideration of the need for biopsy is one of the main reasons for prompt consultation of the appropriate specialist.

Skin biopsy

Skin biopsies are simple, have low morbidity, and should be strongly considered in any patient with new-onset purpura or other lesions suspected to be due to vasculitis. Assuming lesions are due to vasculitis is a common clinical error. Cutaneous small-vessel vasculitis usually presents as palpable purpura, but other presentations include flat and angular (retiform) purpura, nodules, ulcers, and bullae. Vasculitis is probably the most common cause of palpable purpura, but there are others, and thus biopsy is usually indicated. The typical pathologic finding is leukocytoclastic vasculitis (LCV), which is quite helpful for confirming the presence of vasculitis but does nothing to address the many potential causes of this pathology. Immunofluorescence staining can be of additional value, especially if IgA vasculitis (Henoch-Schönlein) (associated with strong staining for IgA) is suspected.

It is important to realize that patients with LCV on skin biopsy may have one of the named primary vasculitides or any of several other conditions, including another named systemic rheumatic disease, a concurrent or recent infection, or drug/toxin exposure. Some patients will have vasculitis apparently limited to the skin and without any clear exposure. Thus, it is important to interpret the skin biopsy in the context of a compulsive search for other findings by history, examination, laboratory tests, and imaging.

A full-thickness skin biopsy that includes larger vessels and subdermal tissue may be necessary for some forms of vasculitis or related disease, such as PAN, panniculitis, and others.

Nasal cavity and sinuses biopsy

Although the sinonasal mucosa is frequently involved in either GPA or EGPA, biopsies of these readily accessible tissues are of limited diagnostic value since the three characteristic features of granulomatous inflammation, necrosis, and vasculitis are often not present; however, if seen in the right setting, then such biopsy findings are supportive of a diagnosis of vasculitis. Sinonasal biopsies and cultures may be useful to establish a diagnosis of infection or malignancy.

Temporal artery biopsy

Biopsy of the temporal artery is safe and ~85% sensitive for diagnosing GCA, so one should have a low threshold for having it performed in an elderly patient who has either a new and persistent headache and a high ESR or CRP, or a constellation of features suggestive of GCA regardless of the ESR and CRP. Performing bilateral biopsies improves sensitivity by ~5%. The specificity of finding vasculitis in temporal arteries is high for a diagnosis of GCA; however, other forms of vasculitis can rarely involve the temporal arteries.

Kidney biopsy

Kidney biopsy is quite valuable for diagnosing AAV and remains one gold standard for the diagnosis. However, in some settings of acute renal insufficiency with an active urine sediment, other features suggestive of GPA or MPA, and a positive test for serum anti-PR3 or anti-MPO antibodies, kidney biopsy may not be necessary. Kidney biopsy is also quite valuable for diagnosing anti-GBM disease. Kidney pathology in IgA vasculitis (Henoch-Schönlein) and cryoglobulinemia are both distinct from the other vasculitides and in many settings is therefore diagnostic. Kidney biopsy is rarely helpful in diagnosing PAN involving the kidneys since the vessels involved are often too large and/or too patchily distributed to be detected on biopsy, and PAN does not cause glomerular disease. Of course, kidney biopsies are extremely helpful in diagnosing nonvasculitic causes of renal disease or a coexisting second pathology in patients with vasculitis.

Brain biopsy

Brain biopsy is considered to be essential for the confident diagnosis of CNS vasculitis, since vasospastic disease has a similar angiographic appearance, and a variety of infectious, inflammatory, or malignant diseases can cause the type of multifocal small abnormalities that are usually seen on brain MRI in CNS vasculitis.

Lung biopsy

Lung biopsy has a high diagnostic yield for AAV for nodules or infiltrates (including eosinophilic infiltrates in EGPA). Similarly, small-vessel vasculitis with capillaritis can be diagnosed by lung biopsy, but the bronchoscopic evidence of diffuse alveolar hemorrhage in the right clinical and laboratory setting often makes biopsy unnecessary. Lung biopsy can also be important in a patient with known GPA who develops a new lung lesion to help differentiate vasculitis from infection or malignancy.

Peripheral nerve and muscle biopsy

Peripheral nerves are frequently involved in PAN, EGPA, GPA, and MPA, and biopsy of the sural nerve, particularly with simultaneous biopsy of the nearby muscle, has fairly good sensitivity for diagnosing vasculitis. Nerve biopsy is rarely performed on a clinically uninvolved sural nerve and probably has lower sensitivity. However, sural nerve biopsy often leaves the patient with a permanent sensory deficit in the innervated area, and occasionally with neuropathic pain.

Biopsy of other organs

Vasculitis has been reported to occur in almost every organ either as part of a named systemic vasculitis or in cases of “single organ” vasculitis. Thus, biopsy of organs other than those listed above may be useful. More common is the situation in which vasculitis is found in surgical specimens for patients thought to have other medical problems. Such findings then appropriately stimulate a full evaluation of possible systemic vasculitis.

ILLNESS IN A PATIENT WITH KNOWN VASCULITIS

A new illness in a patient already diagnosed with vasculitis still frequently presents diagnostic and therapeutic challenges. Many of the vasculitides involve diverse symptoms and findings, and the features during flare do not always resemble the features at initial diagnosis. Patients receiving immunosuppressive medications for vasculitis are at risk for both conventional and opportunistic infections. Noninfectious side effects of medications are also common. Importantly, permanent damage from prior episodes of vasculitis (eg, from neuropathy, nasal septal perforation, or stenosis of a large artery) can produce chronic symptoms that fluctuate in severity and mimic active vasculitis or infection.

An example of the complexity of evaluating new clinical problems in a patient with vasculitis is that of dyspnea in a patient with GPA. This problem could indicate pulmonary hemorrhage, subglottic stenosis, pulmonary nodules (with or without superinfection), pericarditis, pulmonary embolism, bacterial pneumonia, atypical pneumonia (particularly due to pneumocystis jeruvicii [carinii]), or glucocorticoid-induced myopathy, all problems for which a patient treated for GPA would be at relatively high risk. Distinguishing a manifestation of recurrent vasculitis from an infection or a medication side effect can be quite difficult, so the most appropriate general advice is to think broadly: resist the temptation to assume that vasculitis is the cause of symptoms, consider the possibility that infection may coexist with active vasculitis, consider the possibility that the original diagnosis of vasculitis may have been wrong, and have a low threshold for hospitalization and consultation.

TRIAGE AND CONSULTATION

The decision whether to hospitalize a patient who has or may have vasculitis depends on the severity of illness. Sometimes the need for admission is obvious: pulmonary hemorrhage, hematuria with elevated creatinine, or vision loss. Admission may also be indicated for other manifestations that may appear less serious but are often indicators of severe disease with the potential to cause permanent organ damage (peripheral neuropathy, hematuria, or new hypertension with normal creatinine), or to facilitate rapid evaluation. In patients who appear safe to return home, arrangement for close outpatient follow-up remains critical. For example, a patient with a moderate or high probability of having GCA may be discharged from the emergency room after receiving a first dose of glucocorticoids, but only with a prescription for prednisone (40-80 mg/d) and scheduling within 1 week for a temporal artery biopsy and follow-up with a primary care provider or rheumatologist.

Urgent rheumatologic consultation on either an inpatient or outpatient basis is usually indicated for patients suspected of having vasculitis and is influenced by the organ system(s) involved and the need for biopsy or advanced imaging. Consultations with specialists with expertise in particular clinical skills and diagnostic procedures (eg, ophthalmology, otolaryngology, pulmonary, nephrology, gastroenterology, dermatology, or neurology) may well also be indicated. Since syndromes that include vasculitis in the differential diagnosis often include infectious diseases and/or malignancies, it is common, and usually appropriate, for a patient with a severe illness of mysterious cause to require “pan-consults”. In such situations, the hospitalist has the challenge of facilitating and prioritizing the evaluations and communicating to the patient and concerned family members the recommendations of numerous specialists.

TREATMENT

Clinicians should expect their consultants, particularly rheumatologists or nephrologists, to guide treatment for vasculitis. Hospitalists may, however, be faced with the decision of whether to start high-dose glucocorticoids in advance of consultation, and play an important role in monitoring for the many potential side effects of treatment.

Glucocorticoids are the cornerstone of treatment for all severe forms of vasculitis and are often started in the hospital. If suspicion is high for organ-threatening vasculitis (eg, GCA or acute renal failure with active sediment), then it is often wise to start glucocorticoids while the diagnostic workup is in progress. For less urgent clinical situations in which vasculitis is one of several considerations, particularly if severe infection is a possibility, then it is advisable to forgo glucocorticoid treatment until the diagnosis is confirmed. Vasculitis causing compromise to vital organs (eg, pulmonary hemorrhage, glomerulonephritis, acute peripheral neuropathy, vision loss, angina, or digit ischemia) is often treated with high-dose IV methylprednisolone, 1000 mg daily for 1 to 3 days, followed by oral prednisone 1 mg/kg/d. Starting with oral prednisone is usually appropriate for disease that is potentially severe but not actively damaging organs such as GCA with headache, Takayasu with new arterial lesions, or GPA with nasal disease and a pulmonary nodule.

Another urgent inpatient treatment decision is whether a patient requires plasma exchange. This procedure is considered standard care in anti-GBM disease. Plasma exchange is also used in some centers for treatment of fulminant AAV, particularly with rapidly progressive renal failure, but the value of this approach is under active investigation. The decision for or against plasma exchange is always an urgent one and is a major reason why it is appropriate to push for renal biopsy with immunofluorescence staining and rapid serologic testing for anti-GBM, anti-PR3 ANCA, and anti-MPO ANCA.

Additional immunosuppressive drugs are widely used in vasculitis, with the choice of agent dependent on the disease and its severity. Cyclophosphamide, azathioprine, methotrexate, and rituximab are all used for various types and stages of vasculitis. Unlike glucocorticoids or plasma exchange, the effects of these other drugs are not rapid, and the decision to start them can await evaluation by a consultant with expertise in vasculitis and in the use of these medications.

Physicians caring for patients being treated for vasculitis should be attentive to common side effects of glucocorticoids including diabetes mellitus, osteoporosis, osteonecrosis, glaucoma, cataracts, psychosis and mania, myopathy, infection, and other problems. Prophylaxis against pneumocystis jeruvicii (carinii) is widely considered indicated for any patient receiving cyclophosphamide or for any patient receiving both high-dose glucocorticoids and another immunosuppressive agent. Trimethoprim-sulfamethoxazole, either a double-strength tablet three times per week or single-strength tablet daily, is considered the drug of choice for such prophylaxis for patients who are not sulfa-allergic; dapsone (first screen for G6PD deficiency) and atovaquone are alternatives.

DISCHARGE CONSIDERATIONS

Patients with vasculitis or those in whom results of key diagnostic tests such as temporal artery biopsy are pending need to be seen quickly after hospital discharge by a rheumatologist and/or another specialist, rather than relying on the patient’s primary care provider. Patients may leave the hospital with multiple new medications, often including high doses of prednisone that have a diverse range of side effects. One should also anticipate that the receipt of a new diagnosis of a dangerous and previously unheard-of disease will likely produce understandable psychological stress for both the patient and family members.

QUALITY IMPROVEMENT

Hospitals need systems to allow for rapid processing of selected tests that can establish a diagnosis of vasculitis and lead to initiation of time-critical therapies, including testing for ANCA and anti-GBM antibodies, measurement of cryoglobulins and partially surrogate tests for cryoglobulins (rheumatoid factor and complement components), temporal artery and renal biopsies, and angiography.

Hospitals should develop a multidisciplinary team comfortable with the many aspects of the evaluation and management of these challenging multisystem diseases.

CONCLUSION

The major factor that delays the correct diagnosis of vasculitis is a delay in even considering the diagnosis. Efforts must be made to educate clinicians that the presence of any of the red flags mentioned in this chapter should rapidly prompt evaluation for possible vasculitis and initiation of appropriate consultations.

SUGGESTED READINGS