Epidemiology, Etiology, and Pathophysiology of Pulmonary Embolism

Abstract

Over the last 20 years, there has been a progressive increase in the incidence of pulmonary embolism (PE) diagnosis in the United States, Europe, and Australia. Increased use of computed tomography pulmonary angiography has likely contributed in part to this rising incidence. However, it is pertinent to note that the burden of comorbidities associated with PE, such as malignancy, obesity, and advanced age, has also increased over the past 20 years. Time-trend analysis in North American, European, and Asian populations suggests that mortality rates associated with PE have been declining. The reported improved survival rates in PE over the past 20 years are likely, at least in part, to be the result of better adherence to guidelines, improved risk stratification, and enhanced treatment. Factors contributing to the development of venous thromboembolism (VTE) include stasis of blood, hypercoagulability, endothelial injury, and inflammation. In 70 to 80% of cases of PE, the thrombi embolizes from the proximal deep veins of the lower extremities and pelvis. Strong risk factors for VTE include lower extremity fractures and surgeries, major trauma, and hospitalization within the previous 3 months for acute myocardial infarction or heart failure with atrial fibrillation. Acute PE causes several pathophysiological responses including hypoxemia and right ventricle (RV) failure. The latter is a result of pulmonary artery occlusion and associated vasoconstriction. Hemodynamic compromise from RV failure is the principal cause of poor outcome in patients with acute PE.

Pulmonary embolism (PE) is defined as an occlusion in the pulmonary arterial tree by material (e.g., thrombus, tumor, fat) that originates elsewhere in the body. ¹ The current review focuses on PE due to thrombus. We now discuss the epidemiology, etiology, and pathophysiology of PE.

Epidemiology

Over the last several decades, the incidence of PE diagnosis in the United States has risen sharply. ² The reported incidence in 1998 was 62 per 100,000; in 2006, it was 112 per 100,000; and in 2016, it was 120 per 100,000. In 2019, in the United States, 393,000 people were diagnosed with PE. ³ A similar temporal increase in the incidence of PE has been reported from various European countries (including Denmark, Spain, and Italy) and from Australia. ^{4 5 6 7} On the European front, particularly valuable insight with regard to temporal trends in the incidence of PE has been provided by a study by Sonne-Holm et al, published in 2022. ⁴ They identified all patients >18 years in Denmark with a first time in-hospital diagnosis of PE. They analyzed data on all such patients identified between 1999 and 2018 ( N  = 65,478). Sonne-Holm et al divided patients into seven age groups: 18 to 34; 35 to 44; 45 to 54; 55 to 64; 65 to 74; 75 to 84; and >85 years. Their study clearly showed an increasing incidence of PE over time in all age groups. ⁴ For example, the incidence of PE in the age group 55 to 64 was 92 in the calendar period 1999 to 2003; this increased to 229 in the calendar period 2014 to 2018. In addition, their data made it apparent that PE is predominantly a disease of the older population. For example, in the calendar period, 2014 to 2018, the incidence of PE in the age group > 85 was more than three times greater than that in the age group 55–64 (417 vs 123). ⁴ During the study period, the mean age of patients with PE increased from 66 in the first calendar period to 68 years in the final calendar period. In the last calendar period, there was an almost equal incidence of PE in women and men. ⁴ However, while this meticulously detailed and thoughtful epidemiological study yields a treasure trove of information, this report was confined to one northern European country (population in 2018 = 5.3 million) with a long-life expectancy (women = 83.1 years; men = 79.4 years). ⁸ Accordingly, the generalizability of these findings is somewhat limited.

It would appear likely that increased use of imaging, and, in particular, computed tomography pulmonary angiography (CTPA) has contributed to the increasing reported rates of PE over the last several decades. ^{9 10 11} The high resolution of CTPA makes it possible to detect filling defects in subsegmental arteries as small as 2 to 3 mm in diameter (such defects being of doubtful clinical significance). ⁹ It is of interest to note that, in the 25 years since CTPA was first introduced (in 1998), the reported incidence of PE in the United States has doubled. ¹¹ There has also been a general increase in the use of computed tomography in conditions other than suspected PE. ⁹ The incidental finding of PE is reported in 17% of patients >80 years and in 3.6% of oncology patients. ⁹

However, it is pertinent to note that the burden of comorbidities associated with PE (such as malignancy, obesity, and advanced age) has also increased over the past 20 years. ^{12 13 14 15 16} The thromboembolic risk associated with malignancies is well documented. There has been a steady increase in survival in many cancers over the last two decades. ¹⁴ In recent reports, an increasing proportion of patients with cancer was observed among patients diagnosed with PE. ⁴ Similarly, there has been a marked increase in the prevalence of obesity in many countries. ¹⁵ Specifically, the Center for Disease Control and Prevention estimated that the obesity prevalence in the United States increased from 31% in 2000 to 42% in 2020. ¹⁶ Of note, the data discussed previously were obtained in the pre-COVID (coronavirus disease) era, a time when the life expectancy of the population of many countries was progressively increasing.

It is estimated that, currently, in the United States, approximately 60,000 to 100,000 people die from PE each year. ^{1 2} In Europe, PE is responsible for 300,000 deaths annually. ¹ Accordingly, it is a leading cause of morbidity and mortality. ¹³

Time-trend analysis in North American, European, and Asian populations suggest that mortality rates associated with PE has been declining. ^{1 17} Barco et al reported age–sex-specific PE mortality in the United States between 2000 and 2018. ¹⁸ Age-standardized mortality rate decreased from 6.0 deaths per 100,000 population in 2000 to 4.4 deaths per 100,000 in 2006. The mortality rate in women decreased further to 4.1 deaths per 100,000 population in 2017. In contrast, the mortality rate in men plateaued at 4.5 deaths per 100,000 population in 2017. There was, during the study period, a decrease in median age at death from PE, from 73 to 68 years of age. PE-related mortality was >50% higher in Black patients compared with White patients. Mortality rates in white Americans were 50% higher than in Asian Americans. ¹⁸

Several European studies have demonstrated significant decreases in mortality rates from PE. In Denmark, between 1999 and 2018, 1-year mortality rate from PE continuously decreased in all patient age groups and in both men and women. ⁴ Strikingly, this rate halved in patients aged 65 to 74 years. ⁴ In a study from Northwestern Italy, Dentali et al collected data on hospitalizations for PE—drawn from a population of 13 million people—during the period of 2000 to 2012. ⁶ A total of 60,853 patients with PE were identified and analyzed. In-hospital case fatality rate significantly decreased throughout the study period both in women and in men. ⁶ Hobohm et al analyzed PE-related mortality and time trends for the DACH countries (Germany, Austria, and Switzerland). ¹⁹ In the 15-year period between 2000 and 2015, annual PE-related mortality decreased linearly from 15.6 to 7.8 per 1,000 deaths. ¹⁹

The reported improved survival rates in PE over the past 20 years are likely, at least in part, to be the result of better adherence to guidelines, improved risk stratification, and enhanced treatment. There is also the consideration that the fall in PE mortality rates may, as noted earlier in our review, be a result of increasing detection of small segmental or subsegmental PEs that do not adversely affect survival.

The great majority of data and analyses discussed so far relate to a time before the COVID pandemic. There is little doubt that COVID has resulted in an increase in the incidence of PE and in PE-related deaths. ^{20 21 22} As we accrue more data, it is likely that many of the pre-COVID epidemiological statements will have to be revised.

Pathogenesis and Causes of Pulmonary Embolism

The 19 ^th century German physician Rudolph Virchow identified three factors that contribute to the development of venous thrombosis. ²³ These factors—known as Virchow's triad—are stasis of blood flow, hypercoagulability, and endothelial injury. This triad continues to be clinically relevant today. ^{12 23 24 25} The majority of patients with venous thromboembolism (VTE) fulfill most or all of the Virchow triad. ²⁴ Although not included in the Virchow triad, inflammation is now also recognized as a vital component in the formation of thrombus. ²⁵ Inflammatory mediators induce endothelial damage together with platelet activation. In turn, the activated platelets release procoagulants and attach to neutrophils, triggering the release of neutrophil extracellular traps (NETs). ²⁵ The latter are copies of DNA, histones, and antimicrobial proteins. NETs provide a scaffold for red blood cells, platelets, and procoagulant molecules to promote thrombus formation. ^{25 26} The responsible thrombus dislodges from the vessel wall and propagates and embolizes into the pulmonary arteries. ^{26 27} In approximately 70 to 80% of cases of PE, the thrombus arises from the deep veins of the lower extremity or pelvis. ² Of patients with proximal vein deep venous thrombosis (DVT), more than 50% have concurrent PE at presentation. ¹ Calf vein DVTs rarely embolize to the lungs and two-thirds of such thrombi resolve spontaneously after detection. ¹ In approximately 6% of cases of PE, the source is the deep veins of the upper extremities. ^{2 12} Such emboli are typically associated with central venous lines and pacemakers leads. Lower extremity DVTs are much more likely than upper extremity DVTs to cause PE. ¹² Pulmonary emboli are usually multiple, with the lower lobes being involved in the majority of cases. ¹

Predisposing Factors

DVT and PE represent two expressions of a similar clinical pathological process, often referred to as VTE. ²⁸ The risk factors for PE largely overlap with those of DVT. ²⁹ The risk factors for VTE with their odds ratios are detailed in Table 1 . ³⁰ A notable exception to this overlap rule is what is known as the “factor V Leiden paradox.” The paradox is that for individuals with the factor V Leiden mutation—as opposed to individuals with the wild-type factor V—the relative risks are higher for DVT and lower for PE. ³¹

Table 1. Predisposing factors for venous thromboembolism.

Strong risk factors (odds ratio > 10)

Fracture (hip or leg)

Hip or knee replacement

Major general surgery

Major trauma

Spinal cord injury

Moderate risk factors (odds ratio 2–9)

Arthroscopic knee surgery

Central venous lines

Chemotherapy

Congestive heart or respiratory failure

Hormone replacement therapy

Malignancy

Oral contraceptive therapy

Paralytic stroke

Pregnancy, postpartum

Previous venous thromboembolism

Thrombophilia

Weak risk factors (odds ratio < 2)

Bed rest > 3 days

Immobility due to sitting (e.g., prolonged car or air travel)

Increasing age

Laparoscopic surgery (e.g., cholecystectomy)

Obesity

Pregnancy, antepartum

Varicose veins

Source: Reproduced with permission from Anderson and Spencer 2003. ³⁰

Strong provoking factors for VTE are lower limb fractures, hip and knee joint replacement, and spinal cord injury. ^{17 30} Strikingly, without prophylaxis, VTE develops in approximately 50% patients undergoing elective total hip or knee replacement. ³⁰ In patients with paralytic spinal cord injury PE occurs within 3 months in approximately 5%. Major general surgery (defined as abdominal or thoracic operations that require general anesthesia lasting ≥30 minutes) and major trauma are other strong risk factor for VTE. ³⁰ Additional important predisposing risk factors are the occurrence within the previous 3 months, of either myocardial infarction or hospitalization for heart failure with atrial fibrillation. ¹⁷ Previous VTE has a 25% rate of recurrence within 5 years of the initial episode. ²⁴

Active malignancy is a significant risk factor for VTE. The hypercoagulable state in patients with cancer results from production by the cancer cells of procoagulant substances, such as tissue factor. ²⁴ The tumor cells may invade a vessel inducing local stasis and endothelial damage, thus increasing the risk of VTE. Adenocarcinoma and metastatic disease confer a particularly high VTE risk. ²⁵

Inherited thrombophilia is a genetic tendency to VTE. The most frequent causes of an inherited hypercoagulable state are the factor V Leiden mutation and the prothrombin gene mutation. ²⁴ Together these two disorders account for 50 to 60% of cases associated with thrombophilia. Defects in protein S and protein C and antithrombin account for most of the remaining cases. ²⁴

Because of the widespread use, oral contraceptives (OC) are a notable risk factor for VTE in young women. The particular risk for VTE is determined by the composition of the OC. When a combined estrogen and a progestin is used, the relative risk for VTE is increased 3- to 4-fold. This risk is further increased when such OC are given to a woman with additional risk factors such as a thrombophilia or obesity. ^{24 25}

In pregnancy, as an adaptive response to reduce hemorrhage during delivery, coagulation is activated, and thrombolysis is inhibited. This results in a hypercoagulable state, associated with a 20-fold increased risk for VTE in the 3 months following delivery and a 4- to 5-fold increased risk during pregnancy. Indeed, VTE is the most common cause of maternal death during pregnancy and puerperium in North America and Europe. ³²

Infection is a common trigger for VTE. An increased risk of VTE has been noted in patients with COVID infection. VTE has been noted in 5 to 10% of patients with COVID in the intensive care unit (ICU) and in 5% of COVID patients hospitalized in a non-ICU setting. Rates may be lower with omicron and omicron subvariants. ²⁴

The increased risk for PE posed by obesity is thought to be the result of hypercoagulability and endothelial cell injury together with stasis of the lower extremities. Steffen et al found that abdominal obesity increased 2-fold—in both men and women—the risk for developing VTE. ³³

As noted earlier in our review, older age appears a risk factor for VTE. Postulated responsible factors for this association include hypercoagulability, endothelial senescence, and venous stasis. Other factors to consider are that increasing age is associated with decreasing mobility and higher rates of malignancy, obesity, and other comorbidities. ³⁴

In individual patients, there is often more than one risk factor present. In a population-based study (The Worcester Venous Thromboembolism study), 53% of VTE patients had three or more risk factors at the time of VTE. ³⁵ These risk factors, in order of occurrence, were limited mobility >48 hour in the last 30 days, recent hospitalization, recent surgical procedure, recent infection, active malignancy, and current hospitalization. ³⁵

Pathophysiology

Once thrombi lodge in the pulmonary arterial tree, a series of pathophysiological responses may occur:

• Infarction: pulmonary infarction may occur when small thrombi lodge distally in the segmental/subsegmental vessels. It has been reported to occur in approximately 10% of PE patients. ¹ This may, in turn, results in an intense inflammatory response in the lung and adjacent visceral and parietal pleura. Symptoms may include chest pain and hemoptysis.

• Hypoxemia: this is the most common physiological consequence of acute PE. ¹² The principal mechanisms of hypoxemia are ventilation/perfusion mismatch and intrapulmonary shunting. ^{1 12 27 29} There is redistribution of blood from obstructed regions to uninvolved areas of the pulmonary vascular bed. As a result, there is overperfusion of nonembolic, nonoccluded regions leading to abnormal perfusion in relation to ventilation (i.e., ventilation/perfusion mismatch). ^{1 12 27 29} Intrapulmonary shunting occurs when there are areas that retain blood flow but no ventilation, such as occurs in atelectasis due to loss of surfactant, or in areas of pulmonary infarction. ^{1 27 29} Finally, increased right atrial pressure—as occurs in larger PEs—may open a patent foramen ovale. This may result in intracardiac right-to-left shunting. ¹⁷ Although PE alters pulmonary gas exchange and can cause hypoxemia, hemodynamic compromise is the most significant contributor to worse prognosis. ²

• Cardiovascular compromise: in 1971, McIntyre and Sasahara, in a landmark clinical paper, reported their astute observations and insightful interpretations regarding the pathophysiological responses to acute PE. ³⁶ These observations remain relevant to the present day. They studied the hemodynamic and angiographic data of 20 patients, free of prior cardiopulmonary disease, presenting with acute PE. They found that mean pulmonary artery (PA) pressure was consistently increased when pulmonary arterial obstruction exceeded 30%. Good correlation was obtained between mean PA pressure and angiographic estimation of obstruction. They also made the important observation that the right ventricle (RV) was limited in its response to increasing afterload. Mean PA pressure never exceeded 40 mm Hg, despite massive obstruction in some patients. Accordingly, this level appears the maximal pressure response in the previously normal RV. ³⁶ Later, it was realized that, in addition to the degree of mechanical obstruction by the PE thrombi, other factors played a role in determining the abrupt increase in RV afterload. ³⁷ Chief among these other factors is pulmonary vasoconstriction caused by thromboxane A2 and serotonin. ³⁷ The latter are vasoactive mediators released by activated platelets. In acute PE, as a consequence of pulmonary vascular obstruction and, to a lesser extent, pulmonary vasoconstriction, there is an abrupt increase in pulmonary vascular resistance (PVR) ( Fig. 1 ). ²⁹ The latter is the main component of RV afterload. This leads to increased PA pressure, RV dilation, and a decrease in stroke volume. The increase in PA pressure and RV dilation will increase the stretch on the RV myocytes. The impact of this sudden increase on the RV myocardium is neurohumoral stimulation and increased oxygen demand on the myocardium. Furthermore, RV dilation leads to leftward bowing of the interventricular septum, causing decreased left ventricular (LV) diastolic filling and further decreased LV output. ¹⁷ As a consequence of these changes, RV failure will occur. This will be worsened if hypotension and low arterial oxygen saturation are present. Ventilation/perfusion mismatches may promote further arterial desaturation, thus worsening RV function and increasing RV dilation. ²⁹

Fig. 1.

The sequence of RV failure in acute pulmonary embolism. PE, pulmonary embolus; RV, right ventricle. Reproduced with permission from Huisman et al 2018. ²⁸

In PE patients free of prior cardiopulmonary disease, when the mean PA pressure required to maintain adequate pulmonary flow exceeds 40 mm Hg, the RV fails and hypotension ensues. In such patients, multiple large thrombi are generally responsible for hypotension. In contrast, in patients with underlying cardiopulmonary disease, especially in those with preexisting elevated PVR, hypotension can be induced by smaller emboli. ¹

Conclusion

Over the last 20 years, there has been a progressive increase in the incidence of PE diagnosis in the United States, Europe, and Australia. Increased use of CTPA has likely contributed in part to this rising incidence. However, it is pertinent to note that the burden of comorbidities associated with PE, such as malignancy, obesity, and advanced age, has also increased over the past 20 years. Time-trend analysis in North American, European, and Asian populations suggest that mortality rates associated with PE have been declining. The reported improved survival rates in PE over the past 20 years are likely, at least in part, to be the result of better adherence to guidelines, improved risk stratification, and enhanced treatment. Factors contributing to the development of VTE include stasis of blood, hypercoagulability, endothelial injury, and inflammation. In 70 to 80% of cases of PE, the thrombi embolizes from the proximal deep veins of the lower extremities and pelvis. Strong risk factors for VTE include lower extremity fractures and surgeries, major trauma, and hospitalization within the previous 3 months for acute myocardial infarction or heart failure with atrial fibrillation. Acute PE causes several pathophysiological responses including hypoxemia and RV failure. The latter is a result of PA occlusion and associated vasoconstriction. Hemodynamic compromise from RV failure is the principal cause of poor outcome in patients with acute PE.