When properly diagnosed and treated, clinically apparent pulmonary embolism (PE) is an uncommon cause of death, and the PE recurs in only a small minority of patients. Most deaths are related to underlying diseases such as cancer, congestive heart failure, or chronic lung disease. In PE patients with such underlying disease(s), they show higher risk of death than did other patients with PE. And in such patients the deaths occur within one year after the attack of PE.1

It has been known that residual thrombi and microscopic vasculopathy after the attack(s) of acute PE lead to chronic thromboembolic pulmonary hypertension (CTEPH). However, the incidence of CTEPH after acute PE and associated risk factors for CTEPH developments are not well documented. The purpose of this pictorial essay was to demonstrate serial computed tomography pulmonary angiography (CTPA) features in patients with acute massive and submassive PE and to disclose acute CTPA and clinical features that are related to the prediction of oncoming CTEPH.

Currently, the most commonly used first-choice imaging examination in patients with suspected PE is CTPA as thrombi in the vessels can be visualized directly on CTPA (Fig. 1).234 However, clots in the small vessels (e.g., subsegmental branch) are sometimes difficult to visualize on CTPA. The CTPA can be a part of triple-rule-out (TRO) imaging technique where a cost-effective evaluation of the coronary arteries, aorta, pulmonary arteries and adjacent intrathoracic structures for the patient with acute chest pain can be accomplished (Fig. 2). TRO CT is most appropriate for the patient who is presumed to be at low-to-intermediate risk for acute coronary syndrome and whose symptoms may also be caused by acute abnormal conditions of the aorta or pulmonary artery.5 Technically, the heart is positioned at the center of the gantry to maximize resolution. Z-axis coverage starts 1–2 cm above the aortic arch and extends through heart base. Cranio-caudal direction acquisition is preferred. Standard scan parameters include mean effective mAs of 200 and 120 kVp.5 Artificial intelligence (AI) for the detection of PE endorses high degree of diagnostic accuracy for the detection of PE on CTPA.67 According to a study, the AI algorithm allows 93% sensitivity and 96% specificity on a per-exam basis for the identification of PE. On a per-lesion basis, 1,174 of 1,352 (89%) findings marked as embolus were true emboli.6 Thus, sensitivity and specificity are balanced and therefore the AI for automated PE detection could become a clinically useful diagnostic adjunct.

For the screening of CTEPH among patients with acute PE, radionuclide ventilation/perfusion scintigraphy (V/Q scintigraphy) is currently considered the technique of choice (Figs. 3 and 4).8910 Tunariu et al.9 reported that the sensitivity of V/Q scintigraphy, as well as the diagnostic specificity for CTEPH, is higher than that of CTPA.

The introduction of dual-energy CT (DECT) has allowed radiologists to generate functional images as well as traditional anatomic images.1112 With the recent advent of the DECT technique, the ability to obtain color-coded iodine perfusion map of the lung may allow us to evaluate perfusion changes in the lungs and further supports the role of pulmonary CTPA (Figs. 3 and 4).13141516 Hong et al.11 pieced together the diagnostic values of DECT for the detection of acute and chronic PE and the results are as follows: sensitivity and specificity for acute PE range from 60% to 90% and 88% to 99%, respectively, for acute PE; and those for chronic PE range from 96% to 100% and 76% to 96%, respectively, for chronic PE, when using SPECT or scintigraphy as a reference standard. It is also interesting that in terms of radiation dose, according to the same summary from Hong et al.,11 DECT does not add radiation dose, when compared with conventional CT. It has been reported in prior study that DECT is associated with higher radiation exposure than single-energy CT.17

Doppler and color Doppler ultrasonography has been regarded as a gold standard for the diagnosis of DVT (Figs. 1 and 5). Following the pulmonary examination, the inferior vena cava and the iliac, femoral, and popliteal veins can be studied with CT without additional intravenous contrast administration. Indirect CT venography (CTV) after CTPA simplifies and shortens venous thromboembolism work-up (Fig. 5). Initial studies indicate that CTV is comparable to ultrasound in the evaluation of femoral and/or popliteal DVT. CTV has the advantage of evaluating the iliac veins and inferior vena cava, namely vessels poorly seen on sonography and venography. Combining CTV with CTPA increases confidence in deferring treatment when results for both the pulmonary arteries and leg veins are negative, and increases the diagnosis of venous thromboembolism by 25% over CTPA alone.1819 It should be considered, however, that CT technique such as dual-energy spectral imaging and/or adaptive statistical iterative reconstruction could be adopted to mitigate patient radiation dose for combined CTV and CTPA.2021

According to a guidance document issued by the American Heart Association in 2011, the terms of massive (mortality of 25–65%), submassive (mortality of 3%), and low-risk (mortality of < 1%) PE were defined on patient mortality likelihood.22

Massive PE refers to the event in which the patient has sustained systemic hypotension (< 90 mmHg systolic, for > 15 minutes) and bradycardia with signs of shock or being pulseless (Fig. 2), whereas submassive PE refers to the condition where despite of the absence of systemic hypotension the patient has the evidence of myocardial necrosis or right ventricular (RV) dysfunction with the PE (Fig. 6).23 It should be noted that in patient prognosis (mortality) stratification, no mention has been present on clot burden or location (which can be determined on CTPA).23

Acute PE attack and embolic vascular obstruction of a significant extent of pulmonary circulation (usually more than 30%) leads to pulmonary vascular resistance and consequent acute pulmonary artery hypertension. This pulmonary vascular obstruction is further worsened by the release of vasoactive agents from plasma, platelets, or tissue and reflex pulmonary artery vasoconstriction, and thus resulting in systemic arterial hypotension.24

RV compensation and increased contractile performance are needed for RV outflow obstruction. The compensation requires further an increase in work transmission from the RV to its vascular bed; that transmission necessitates a greater myocardial oxygen requirement and a reduction of RV mechanical efficiency. RV dilatation and resultant pericardial constraint elicit bowing of the interventricular septum (IVS) toward the left ventricle (LV),25 resulting in a decreased LV preload; subsequently decreased LV output and decreased systemic arterial pressure may result in a downward vicious cycle of compromised coronary perfusion, increasing ventricular hypoxemia and RV dysfunction, and eventually cardiogenic shock and RV infarction and finally circulatory collapse (Fig. 7).26

In acute PE, circulatory failure and systemic hypotension are important for predicting poor prognosis. However, hemodynamically unstable PE accounts for only a minority of all PE presentations. Other clinical predictors such as age > 70 years, coexisting congestive heart failure, or chronic obstructive pulmonary disease are also regarded as bad prognosis factors. This is because such pre-existing cardiopulmonary dysfunctions impairs RV adaptability to abrupt afterload increase.2426

Echocardiography, CTPA, brain natriuretic peptide (BNP), N-terminal pro-brain natriuretic peptide (NT-proBNP), Troponin I, Troponin T, and heart-type fatty acid protein (H-FABP) are evaluated for predicting early mortality. By the use of test results, mortality risk can be classified into high-, intermediate- and low-risk groups. The main parameters and scores were determined by 1) the presence of shock or hypotension, 2) PE severity index,27 3) signs of RV dysfunction on an imaging test, and 4) cardiac laboratory biomarkers (Figs. 2, 4, and 6).28

Helical CT and DECT are more efficient in the detection of subsegmental PE.29 Patients with isolated subsegmental PE appear to have a more benign clinical course than those with segmental or more proximal PE; the patients with subsegmental PE have lesser associated DVT, less frequent dyspnea, and low clinical probability of PE. Therefore, the necessity to treat patients with isolated subsegmental PE is uncertain. However, small peripheral PE may prove clinically relevant in cases of diminished cardiorespiratory reserve. Currently no data are available on the long-term consequences of such events, especially on the occurrence of chronic pulmonary hypertension. As mentioned above, patients with isolated subsegmental PE are unlikely to have an associated DVT. Therefore, in patients with subsegmental PE, rather than directly commencing anticoagulation therapy, clinical surveillance might be a way of benefit in management. This approach is currently assessed in an ongoing prospective cohort study.3031

Acute PE patients with co-existing DVT are more serious clinically than those without a DVT. An acute picture may be present in the chronic background in a significant proportion of patients with PE with co-existing DVT. Moreover, in the presence of DVT, PE appears to be more central (Fig. 1).32

Acute PE usually resolve within months. Occasionally, the thrombi change into fibrous vascular obstruction and results in occlusion of the thrombosed veins. Alias et al.,33 in a mouse model, suggested that endothelial cell-specific deletion of vascular endothelial growth factor 2/kinase insert domain protein receptor (VEGF-R2/kdr) causes the mitigation of thrombus vessel formation and misguides thrombus resolution. Medical conditions compromising early thrombus angiogenesis may facilitate the development of CTEPH.

On imaging studies, parenchymal features such as mosaic attenuation, asymmetric pulmonary arterial size (Figs. 4 and 8), and pulmonary infarcts are commonly observed. Combined with right-sided heart chamber enlargement and particularly with evidence of PE on pulmonary arterial branches on enhanced scans, the CTPA features can be specific in the diagnosis of the CTEPH, In addition, bronchial artery hypertrophy and collateral vessel development can be seen.34 Unusual enlargement of bilateral pulmonary arteries can also be seen on chest radiograph and digital subtraction angiography (Fig. 8).

The RV is a thin-walled chamber having a smaller diameter than the LV. With increasing pulmonary artery pressure, the RV begins to be hypertrophied in order to maintain stroke volume and cardiac output. It becomes eventually dilated and fails to function properly. The increased RV muscle mass can be measured on CT at middle ventricular level. The normal thickness of the RV free wall is 4 mm or less. When the thickness is 6 mm or more (Fig. 4A-C), the sensitivity and specificity for the presence of pulmonary hypertension is 81% and 92%, respectively.3536

With increasing RV chamber related to RV strain and pulmonary hypertension, the RV/LV luminal diameter starts to increase. RV dilatation is regarded to be present when the RV/LV ratio is greater than 1:1 with a sensitivity and a specificity of shared 86%.35 Short axis of the RV is measured from inner wall to inner wall where tricuspid valve is scanned (typically in the basal third of the RV), whereas short axis of the LV is measured from inner wall to inner wall where mitral valve is scanned (typically in the basal third of the LV).

In healthy subjects, the IVS has a right-sided convexity. With increase in RV strain, the right-sided convexity becomes to be flattened or finally to be reversed with left-sided convexity. Once the leftwardly bowed IVS impedes on LV, cardiac output becomes compromised. Patients with a main pulmonary arterial pressure of > 30 mmHg appear to have leftward bowing of the IVS (Fig. 7).37

Initially devised as angiographic severity indices (Miller and Walsch scoring) for comparison of the PE resolution speed on heparin or fibrinolytic therapy, the modified Miller and Walsh scoring systems have been adopted for CTPA. The modified scoring systems provide excellent correlation between the two scoring systems and a good interobserver agreement.2638 Unlikely from the modified Miller and Walsh scoring system, Qanadli et al.39 and Mastora et al.,40 respectively, designed PE scoring system where partial and complete obstructions of segmental arteries with PE were scored differently from each other (Fig. 9).

In several studies, the clot load measured by the use of Qanadli and colleagues’ method showed positive correlation with patients’ death4142 or hemodynamic severity.43 In contrast, Ghaye et al.44 and Araoz et al.45 reported that pulmonary artery (PA) clot load scores are poor predictor of mortality in patients with acute PE. While PA clot load scores can be an indicator of the severity of the current episode of PE or treatment effectiveness, they may not be used directly as an indicator of RV failure or death of the patients. Therefore, pulmonary vascular resistance may not be determined only with mechanical obstruction of the pulmonary arteries and their branches by intravascular clot loads, but also with vasoactive amines (released from plasma, platelets or tissue), reflex PA vasoconstriction, and systemic arterial hypoxemia occurring during acute PE.444647

It has been known that residual thrombi and microscopic vasculopathy after the attack(s) of acute PE lead to CTEPH in 0.6% to 4.4% of acute PE patients (Fig. 10).4849 However, the real chronic CTEPH incidence after the episode of acute PE is not exactly known. According to a study, the incidence of CTEPH observed one year after the first episode of acute massive or submassive PE was approximately 6.9%.50 In other study, the cumulative two-year incidence of CTEPH was 8.6% (25/290 patients).51 In a Korean study, the incidence of CTEPH after acute PE was 6.1% (15/246 patients).52

Large RV diameter with RV/LV ratio greater than 1.0, high baseline pulmonary arterial pressure and RV dysfunction were independent predictors of oncoming CTEPH (Fig. 4).51

On a CTPA-based study, the presence of an occlusive clot on initial imaging is associated with an increased probability of CTEPH development (Fig. 4). The presence of mosaic attenuation or lung infarction may also help predict CTEPH development.53 Acute PE patients with co-existing DVT are more serious clinically than those without DVT.32 Recurrent DVT or unprovoked PE was significantly associated with a higher risk of CTEPH, with odds ratios of 3.2 (95% CI, 1.7–5.9) and 4.1 (95% CI, 2.1–8.2), respectively.54

More aggressive treatment including systemic thrombolysis, catheter-directed therapy, and/or surgical embolectomy is warranted to rapidly restore pulmonary blood flow and prevent death. The decision to escalate therapy for submassive PE is controversial; data thus far have been generated most prominently for systemic thrombolysis and to a lesser extent for catheter-directed thrombolysis.23

In addition to indefinite anticoagulation, pulmonary endarterectomy (PEA) is the treatment of choice for CTEPH in patients who are considered operative candidates by having obstructing lesions in the main, lobar, segmental, and subsegmental arteries (Figs. 4 and 11).55 In approximately half of patients with CTEPH whose pulmonary vascular resistance is disproportionate to degree of vascular obstruction on CTPA may disclose severe distal microvasculopathy and regarded as ineligible for the surgery (Fig. 11).49

Balloon pulmonary angioplasty (BPA) is an alternative therapy for patients with inoperable CTEPH.5657 By reviewing 500 consecutive procedures of BPA from 97 patients, the lesion distribution and characteristics evaluated were as follows: type A, ring-like stenosis lesion; type B, web lesion; type C, subtotal lesion; type D, total occlusion lesion; and type E, tortuous lesion.58 The success rate was higher, and the complication rate was lower, in ring-like stenosis and web lesions (Figs. 12 and 13). The total occlusion lesions had the lowest success rate. Tortuous lesions were associated with a high complication rate.

In acute PE, circulatory failure and systemic hypotension are clinically important for predicting poor prognosis. However, hemodynamically unstable PE accounts only a minority of all PE presentations. Even though PA clot load scores can be an indicator of the severity of the current episode of PE or treatment effectiveness, they may not be used directly as an indicator of RV failure or death of the patients. Other clinical predictors such as age > 70 years, coexisting congestive heart failure, or chronic obstructive pulmonary disease are also regarded as bad prognosis factors. Large RV diameter with RV/LV ratio greater than 1.0 and/or occlusive clot and pulmonary infarction on initial CTPA, and high baseline pulmonary arterial pressure and RV dysfunction are independent predictors of oncoming CTEPH.