Pulmonary embolism is a potentially fatal condition and its diagnosis is a challenging task, both clinically and radiologically [12]. Clinical assessment of the patient is the first and most crucial step to reduce unnecessary imaging which has undesirable consequences as increased cost and ionizing radiation exposure [13]. However, in a good number of cases, clinical diagnosis may be problematic as symptoms may range from silent to hemodynamic instability [14]. Thus, the need for a ready available modality for timely diagnosis.
The introduction of the relatively non-invasive spiral CT angiography has shown to reliably exclude clinically important PE [15]. The development of multi-detector CT has led to improved visualization of peripheral pulmonary arteries and small sub-segmental emboli [16]. However in patients with allergies to iodinated contrast material or with elevated serum creatinine levels and those patients with nonspecific cardiopulmonary signs and symptoms, non-contrast CT chest may be the only ready accessible modality.
It is important to be aware of the hyperdense thrombus to help diagnose acute PE in patients undergoing non-contrast CT of the chest. Visualization of the clot is likely related to the age of the clot with increased density in the vessel either due to direct visualization of the thrombus itself or as a result of local slow intravascular blood flow due to intra-arterial thrombi [17].
This study included 80 adult patients clinically suspicious of PE. Age of the patients ranged from 35 to 72 years old with a mean age 54.18 years. Majority of our cases were females representing 44 patients (55% of the cases). A study done by Venkatesh et al. [18] also showed predominant female patients with 60%, 57%, and 67% incidence.
Symptoms of PE are typically sudden in onset and include dyspnea, tachypnea, chest pain of a “pleuritic” nature (worsened by breathing), cough, and hemoptysis.
The clinical presentation of patients was as follows: 50 patients presented by chest pain representing 62.5%, 48 patients had dyspnea representing 60%, 19 patients had tachypnea representing 23.8%, 22 patients presented by tachycardia representing 27.5%, and 14 patients presented by hemoptysis representing 17.5 %. Our study agreed with Tambe et al. [19] that showed that the most common clinical symptoms were sudden and/or unexplained chest pain, dyspnea, malaise, syncope, or shortness of breath. Another study by Crichlow et al. [20] showed that the most common presenting signs and symptoms were shortness of breath (77%), followed by chest pain (74.3%).
From the 80 patients, 47 were proved positive by CTPA. CTPA was the gold standard in our study. Visualization of complete or partial intraluminal filling defects surrounded by the contrast-enhanced blood pool in the central and subsegmental pulmonary arteries is a direct sign of PE. Distribution of the true positive cases of pulmonary thrombosis among included patients was as follows: 15 patients showed a thrombus within both main pulmonary arteries (18.7%), 15 patients showed a thrombus in the right main pulmonary artery (18.7%), 13 patients had a thrombus in the left main pulmonary artery (16.2%), and last 4 patients had thrombus in subsegmental peripheral pulmonary branch (5%).
Out of the 47 CTPA proved positive patients, 26 cases were positive by non-contrast CT chest depending on hyper dense lumen sign either in central main pulmonary branch or subsegmental peripheral branches (Figs. 1 and 2), with an overall sensitivity of 55.3%, specificity of 100% positive predictive value of 100%, and negative predictive value of 61.1% in the detection of emboli located within the main pulmonary arteries (central emboli). And 21 cases were false positive by non-contrast CT chest attributing to sluggish blood flow mimicking hyperdense sign; as regards false-negative cases, it is attributed to small sized embolus. There was a moderate degree of agreement according to Kappa method (0.505) with p value 0.000. However, chi-square detection rate of central pulmonary embolism for CTPA was significantly higher than that of non-contrast CT with a p value of 0.000.
Tatco and Piedad [21] reported an overall sensitivity of only 36% for detecting central PE, which is significantly lower than our study. On the other hand, Cobelli et al. [22] reported a 41.2% sensitivity and Kanne et al. [23] found that 46.1% of their unenhanced scans were positive for PE.
A number of indirect findings were also seen on the non-contrast CT chest including pleural effusion which was the most common finding seen in 26 patients (32.5%) of all cases (Fig. 3). Second most common finding was a peripheral wedge-shaped opacity in 12 patients (15%) of all cases, followed by pulmonary artery dilatation in 9 cases (11.3%) of all cases. A study done by Pfeil et al. [24] reported that wedge-shaped opacity was the most frequent indirect sign.
During the pandemic of COVID-19 which occurred at the same time of this study, the non-contrast CT was done to exclude viral infection followed by contrast CT after 48 h to exclude pulmonary embolism, in which retrospectively reviewing non-contrast CT revealing presence of hyperdense lumen sign, by which the importance of including this sign in reporting non-contrast CT is highlighted.
There were some limitations in this study; mainly, number of cases is still limited which hindered the possibility to study the usefulness of the hyperdense lumen sign in the segmental or subsegmental branches. Factors like motion artifacts, partial volume averaging, and image noise which almost always affect segmental and more peripheral pulmonary arteries are some of the possible causes why the hyper dense lumen sign is less useful in detecting peripherally located thrombi. Factors that may affect the visualization of a clot, such as the age of the clot, the patient’s hematocrit level at the time of imaging, and probably the patient’s hematocrit level at the time of formation of the clot in the venous system were not considered in this study.