Helical Computed Tomographic Angiography
Technical advancements have led to the growing use of helical (spiral) CT angiograms for evaluating patients with suspected PE. The recent PIOPED-II study examined four-detector CT angiography in 824 patients who were suspected of having PE. Among 773 patients in this cohort with interpretable CT angiograms, the overall sensitivity and specificity of this technique were 83% and 96%, respectively. In addition to the technique's ability to directly visualize intravascular thromboemboli (Fig. 27.5), lung parenchymal abnormalities can be detected that might otherwise account for a patient's dyspnea and other cardiopulmonary symptoms if no thromboemboli are visualized.
A contrast CT angiogram of the chest showing multiple filing defects (arrowheads) in the left and the right pulmonary arteries secondary to obstruction by thromboembolic material.
Despite this diagnostic utility, there are drawbacks to the use of helical CT angiography in establishing or excluding PE. First, it requires the use of intravenous contrast material that is problematic to administer in patients with renal dysfunction or to those with contrast allergy. Second, the procedure carries a higher radiation exposure than a V/Q scan, and may miss subsegmental PE, whose incidence varies between 5% and 30% across different studies. Third, such a CT scan requires the patient to perform a breath-hold for 15-25 seconds, and any motion artifact due to a patient's inability to hold their breath causes suboptimal visualization of segmental arteries. Importantly, a negative helical CT angiographic study does not definitively exclude PE in those patients with high clinical probability (Table 27.4) since ~5% of this cohort will still have a PE. Likewise, there remains a 10% chance that high-probability patients have underlying PE when a "technically unsatisfactory" CT has been achieved.
Owing to the improving performance of routine spiral CT at detecting an acute PE, ventilation/perfusion (V/Q) lung scanning is being performed less often but remains a valuable diagnostic test. The perfusion scan utilizes human albumin macroaggregates or microspheres labeled with 99Tc. Once injected intravenously, the aggregates or microspheres lodge in pulmonary arterioles and capillaries and give a map of the pulmonary circulation. For the ventilation scan, a single deep inhalation of 133Xe to TLC is followed by a 20-second breath hold and then by 45 seconds of tidal breathing, followed by a washout period. Several images are taken during the V/Q scan, with the rationale that only the ventilation is normal in nonperfused areas, and thus a visual V/Q mismatch is diagnostic of PE (Fig. 27.6).
Perfusion images from a V/Q scan. Arrowheads point to mismatched perfusion defects that are consistent with subsegmental emboli.
In this sense, a normal V/Q scan virtually excludes the diagnosis of PE, while a high probability V/Q scan is diagnostic of PE (>90%) in the presence of an intermediate-to-high clinical pretest probability. Conversely, additional diagnostic workup is still indicated in the presence of a low clinical pretest probability, since the V/Q scan can be falsely positive (45%-66%) in this clinical scenario. In any case, the results of most V/Q scans are graded as low (16%) or intermediate (41%) probabilities, and consequently most patients will need further workup to make a definitive clinical diagnosis.
Duplex Venous Ultrasonographic Studies of Lower Extremities
The non-invasive assessment by duplex venous ultrasonography with Doppler flow studies for DVT may be helpful in evaluating the many patients who will remain with intermediate clinical and V/Q scan probabilities for PE. This imaging modality carries a sensitivity and specificity of 89% and 100%, respectively. However, it is less sensitive (38%) in high-risk but asymptomatic patients. The rationale for performing venous Doppler studies of the lower extremities is that if a proximal DVT is diagnosed, anticoagulation is indicated just as if the patient had a documented PE. Given this situation, a single negative leg duplex venous ultrasonographic study does not rule out the possibility of a subsequent DVT, and consequently a repeat study in 1 week is recommended for high-risk patients.
Pulmonary angiography remains the gold standard for the diagnosis of PE, and represents an important test for the evaluation of chronic thromboembolic disease. A properly performed pulmonary angiogram has a sensitivity and specificity for acute PE of >95%. The diagnosis is based on visually confirmed pulmonary artery occlusion or presence of intraluminal filing defects (Fig. 27.7). Other suggestive findings include asymmetrical blood flow, slow filling of the artery, and abrupt arterial cutoff.
Pulmonary angiogram of the right lung showing an acute pulmonary embolism in a 78-year-old woman (arrow). The arrowhead points to the oligemic lung fields. Adapted from Wittram et al: Acute and chronic pulmonary emboli: angiography-CT correlation Am J Roentgenol. Jun;186(6 Suppl 2):S421-9, 2006.
A negative pulmonary angiogram with magnification excludes clinically relevant PE. Properly performed, the mortality of the procedure is extremely low (<0.5%) and morbidity occurs in about 5% of patients. Those events relate primarily to complications from catheter insertion, arrhythmias, hypotension, and reaction to contrast materials, among other factors.