A 67-year-old man with a history of relapsing heart failure (left ventricular ejection fraction 25%-30%), hypertension, and dyslipidemia presents to the emergency department with a 1-week history of spontaneous pain and swelling within the left calf. His chronic dyspnea on exertion is unchanged and he denies pleuritic chest pain or cough.

On physical examination the patient is alert and in no acute distress with normal vital signs. Pulse oximetry obtained on room air is 95%. On physical examination the patient appears euvolemic with a murmur of aortic sclerosis, an S₄ gallop, and no adventitial lung sounds. His left calf is swollen, tender, and slightly warmer than the right calf. The muscle compartments are soft and distal pulses are intact bilaterally.

Duplex ultrasonography of the left leg documents an acute deep venous thrombosis within the gastrocnemius and popliteal veins (Figure 18-1). A complete blood count, chemistry profile, as well as prothrombin time/international normalized ratio (INR) and partial thromboplastin times are within normal limits. Anticoagulation is begun with therapeutic doses of apixaban and he is discharged to home with instructions to see his primary care provider in 1 week.

Heart failure (HF) is a major risk factor for venous thromboembolism (VTE) that is independent of concurrent coronary events or atrial fibrillation.

Although classically listed as a risk factor for decompensated hospitalized patients, HF increases the risk of VTE in stable outpatients as well.

In the absence of thromboprophylaxis, the prevalence of deep venous thrombosis (DVT) in patients hospitalized with HF traditionally ranges from 4% to 26%.¹ However, in a more contemporary 2014 prospective study of Japanese patients hospitalized for HF, DVT was remarkably detected in 34% despite the use of mechanical prophylaxis and/or antiplatelet therapy.²

In the Worcester Venous thromboembolism study of 1822 patients, the concurrence of HF in subjects with VTE yielded a 3-fold higher risk of in-hospital death and approximately 2.5-fold increased risk of dying within 30 days of the thrombotic diagnosis.³

In the U.S. National Hospital Discharge Survey (NHDS), a patient with HF was 2.2 times more likely to incur a pulmonary embolism (PE) when compared to patients without HF.⁴ PE was documented in 9% of patients with decompensated HF admitted to an intensive care unit,⁵ and 10% of the symptomatic PE patients enrolled in the RIETE registry had underlying HF.⁶

HF is a potent risk factor for VTE in young patients. For instance, in the NHDS the relative risk (RR) for PE in octogenarian congestive heart failure (CHF) patients was only 1.3; however, in patients younger than 40 years of age, the RR considerably escalated to 11.7. Similarly, the RR for DVT in patients <40 years was significantly increased at 5.5.⁴

Gender-based and racial differences may influence the risk of VTE in HF patients. For instance, HF appears to be a more potent risk factor for PE in women (RR, 2.45; 95% CI, 2.44-2.46) than in men (RR, 180;95% CI, 1.79-1.80). Additionally, black patients are at greater risk for PE (RR, 2.82; 95% CI, 2.79-2.84) than white patients (RR, 2.10; 95% CI, 2.09-2.11).⁴

A meta-analysis of 32 trials of angiotensin-converting enzyme (ACE) inhibitors in HF patients identified PE as 1 of the 5 primary causes of death.⁷

Postmortem studies of patients with HF have documented PE incidence rates ranging from 0.4% to 50%. However, it can be difficult to ascertain if PE is directly responsible for immediate mortality. Prospective analysis suggests that PE is the predominant cause of mortality in 3% to 10% of patients with HF.

The prothrombotic tendency of HF is partially explained by Virchow classic triad. Stasis occurs due to depressed cardiac output as well as immobility. Endothelial dysfunction results from inhibition of endothelium-derived nitric oxide release as well as elevations in von Willebrand factor (vWF), thrombomodulin, and soluble E-selectin. Lastly, the thrombophilic component is reflected by elevated plasma viscosity, D-dimer, fibrinopeptide A, and fibrinogen levels in HF patients.

The link between thrombophilia and HF is also potentially related to an associated multifactorial chronic inflammatory state as proposed by Chong and Lip.⁸ Figure 18-2 displays the intricate relationship between a variety of HF-elaborated inflammatory proteins, molecules, and cells that can ultimately culminate in VTE.

A striking inverse relationship exists between ejection fraction and risk of VTE (Table 18-1).⁹

HF severity is proportional to VTE risk. In the placebo arm of the MEDENOX trial, VTE was documented in 12.3% of the patients with New York Heart Association (NYHA) class III HF. However, VTE occurred in 21.7% of the patients with class IV HF.¹⁰

Elevated biomarkers have been increasingly linked to VTE risk in HF. In a subanalysis of the MAGELLAN trial, multivariable analysis documented that the N-terminal probrain natriuretic peptide (NT-proBNP) level independently predicted short-term (up to 10 days) VTE risk whereas elevated D-dimer concentrations predicted both short- and mid-term (up to 35 days) VTE susceptibilty.¹¹

In both the Atherosclerosis Risk in Communities Study and Cardiovascular Health Study, a high sensitivity troponin (TnT) concentration was directly linked with the incidence of total and provoked VTE, but not with spontaneous VTE.¹²

CLINICAL FEATURES

Deep Venous Thrombosis

At least 50% of patients with an acute DVT are asymptomatic. Classic signs and symptoms include limb pain, tenderness, swelling, increased warmth, and/or increased venous collaterals (Figure 18-3). Unfortunately, these clinical manifestations (including Homan sign) are neither sensitive nor specific for the diagnosis of DVT.

The physical examination is relatively specific in 2 venous thromboembolic conditions: (1) Acute superficial venous thrombosis with a palpable tender, warm, erythematous cord (Figure 18-4), (2) Phlegmasia cerulea dolens manifested by a markedly turgid, blue, painful extremity (Figure 18-5), which can be both limb- and life-threatening.

Pulmonary Emboli

Acute PE frequently mimics other cardiopulmonary conditions (eg, HF) on account of its protean, nonspecific clinical manifestations. Consequently, it is imperative that a clinician maintains a high level of suspicion for PE, especially in the at-risk patient with HF.

Dyspnea is the most frequent symptom and tachypnea the most frequent sign. Other manifestations include pleuritic chest pain, cough, crackles, low-grade temperature, and/or an increased second heart sound.

Acute pulmonary infarction is associated with pleuritic chest pain, dyspnea, and hemoptysis. Physical findings include a pleural friction rub and possible decreased breath sounds due to a pleural effusion.

The presence of syncope, cyanosis, and/or cardiogenic shock suggests a massive PE.

LABORATORY

D-dimer

An elevated D-dimer via ELISA assay is highly sensitive for VTE; yet, the test is not specific. For example, various nonthrombotic disorders such as trauma, cancer, myocardial infarction, disseminated intravascular coagulation (DIC), pregnancy, and infection can increase D-dimer levels.

Consequently, elevated levels of D-dimer do not accurately predict the diagnosis of DVT/PE, yet normal or low levels have a very high negative predictive value (approaching 100%).

Arterial Blood Gas

Arterial blood gas (ABG) abnormalities in decreasing order of frequency include hypoxemia, widened alveolar-arterial gradient, and a respiratory alkalosis. None of these abnormalities are sensitive or specific for the PE, especially in the setting of coexistent HF.

The combination of cryptic hypoxemia and a normal chest x-ray should always raise the suspicion for PE. A pulse oximetry is extremely insensitive as it is normal in the majority of patients with PE.

Cardiac Biomarkers

Increased levels of brain natriuretic peptide (BNP) and high-sensitivity troponin T (TnT) can occur in response to significant right ventricular stretching and microinfarction, respectively.

BNP and TnT are both increased in the setting of moderate to large PE. Elevated levels of BNP and TnT are neither sensitive nor specific for diagnosing PE. These are best utilized for risk stratification and prognostication in patients with acute PE.

NONINVASIVE CARDIAC TESTING

Electrocardiography

Electrocardiographic (ECG) abnormalites are common in acute PE yet they lack sensitivity and specificity. The most common findings are tachycardia and nonspecific ST-segment and T-wave changes. In <10% of cases, ECG abnormalities frequently referenced in acute PE include the S1Q3T3 pattern (Figure 18-6), right ventricular strain, and/or a new incomplete right bundle branch block. Such abnormalities have been linked to a poor prognosis.

Additional ECG irregularities that predict a poor prognosis in acute PE include atrial arrhythmias (eg, atrial fibrillation), bradycardia or tachycardia, Q waves within the inferior leads (II, III, and aVF), and anterior ST-segment changes with T-wave inversion. An abnormal ECG is more likely to occur with a moderate to large PE.

Echocardiography

Echocardiography should not be routinely obtained when evaluating hemodynamically stable patients suspected of having a PE.

Echocardiographic findings are neither sensitive or specific for acute PE. Echo does yield valuable prognostic information by assessing the right ventricle in patients with acute PE. Approximately 30% to 40% of PE patients display echocardiographic abnormalities indicative of RV strain or overload, including increased RV diameter, impaired RV function, and/or acute tricuspid regurgitation. These findings are more common in large PE.

Rare yet highly morbid echocardiographic defects in the setting of PE include thrombus within the RV, thrombus within the pulmonary artery and its main branches, and/or regional wall hypokinesis that spares the right ventricular apex (ie, McConnell sign, Figure 18-7).

RADIOLOGICAL IMAGING

Chest Radiography

The chest x-ray is often abnormal yet nondiagnostic, with abnormalities including parenchymal abnormalities, cardiomegaly, diaphragmatic elevation, platelike atelectasis, and pleural effusion.

A chest x-ray can be useful in evaluating other conditions that can mimic PE, including HF. Classically described, yet rare and nonspecific radiographic findings include Hampton hump (pleural-based, wedge-shaped abnormality) and Westermark sign (focal decreased pulmonary vascularity, Figure 18-8).