Work Up of Suspected Pulmonary Embolism

Catheter angiogram of the left pulmonary artery in the AP projection.

3D CTA depiction of the left pulmonary circulation.

Pulmonary embolism (PE) is a serious disease which is estimated to account for about 5% of all hospital deaths. It is an under-diagnosed condition because the clinical manifestations are so variable, and because accurate, noninvasive diagnostic tests have not been available. The chest x-ray is often abnormal in PE but is rarely diagnostic, and indeed may be completely normal. A normal ventilation perfusion lung scan is said to essentially exclude the presence of PE. A clinical review in Post Graduate Medicine, however, noted that the majority of VQ scans are non-diagnostic and that they regarded only normal and high probability VQ scans as truly diagnostic scans which required no further evaluation. Rarely, even large saddle emboli can present with a normal VQ scan, because they have not yet fragmented and caused lumenal occlusion. Additionally, patients with chronic lung diseases or COPD will typically have abnormal or indeterminate VQ scans even in the absence of PE, making it necessary to visualize the pulmonary arteries to rule out the presence of emboli.

Pulmonary catheter angiography is the current gold standard for visualizing the pulmonary arteries. This invasive examination has a reported morbidity range of less than 2-3%. Previously, the mortality was reported to range from 0.5 to 0.1 %; however, fatalities are now quite rare at high-volume centers. Catheter angiography is an expensive procedure which can be difficult to obtain emergently. Finally, there is significant interobserver variability in the diagnosis of subsegmental PE. For these reasons, pulmonary catheter angiography is an underutilized examination throughout the U.S. It will continue to be viable in the future, however, especially at centers that offer transcatheter therapy and suction embolectomy.

CT angiography (CTA), which is performed with intravenous contrast using a spiral CT scanner, offers a noninvasive alternative for the direct evaluation of the first to fourth generations of the pulmonary arterial branches. A multivariate computer analysis in the October issue of Radiology compared numerous diagnostic algorithms for PE, including CT angiography, catheter angiography, VQ scanning, ultrasound of the lower extremities, and D-dimer assay. Computer models were used to determine which strategies were optimal on the basis of patient mortality and the cost per life saved. With mortality as the primary outcome parameter, the best strategies all made use of spiral CTA, the best strategy being US followed by CTA. When preference was determined on the basis of cost per life saved, the best strategies again all included spiral CTA, with the best strategy being D-dimer blood assay followed by spiral CTA. Thus, the use of spiral CTA was not only found to be the most cost effective, but also to result in the lowest patient mortality. The reported accuracy of spiral CTA when compared directly to catheter angiography is 95.5% using the data from five published studies with a total of 224 patients.

CTA showing LLL PE.

Spiral CTA is less likely to detect tiny emboli trapped in the subsegmental pulmonary arteries. Only 61% of the subsegmental pulmonary vessels can be reliably imaged even with 2mm collimation. Catheter angiography is unquestionably better at detecting these tiny peripheral emboli, but even this test is not 100% sensitive. Solid data on the clinical importance of small peripheral pulmonary emboli does not exist. Nonetheless, statistical analysis of the PIOPED study suggests that those patients who likely had a small load of emboli appeared to do well clinically on long term follow-up, even in the absence of anti-coagulant therapy. The clinical significance of tiny undetected pulmonary emboli could also be lessened if lower extremity ultrasound excluded any large residual clot, which could serve as a source of recurrent emboli.

I suggest the use of lower extremity ultrasound as the first imaging test after the chest X-ray, when PE is suspected. If this simple, relatively inexpensive test is positive for DVT, then the therapy of choice is anti-coagulation, just as it would be for pulmonary embolism. Therefore, further imaging for PE would not be mandated. If the US is negative for DVT, I would then consider proceeding directly to CTA.

CTA 3D image with LLL PE.

In addition to diagnosing PE, the CT will also detect many other causes of chest pain and shortness of breath, such as aortic dissection, pneumomediastinum, and subtle infiltrates. Be aware that spiral CTA is a technically challenging examination that requires the radiologist to review the three dimensional data set on a dedicated computer workstation. This examination has a steep learning curve, so when a hospital chooses to use this technique, it may be helpful to initially compare the results with VQ scanning or catheter angiography until expertise is developed.

REFERENCES:

van Rossum AB, Pattynama PM, et al. Pulmonary embolism: Validation of spiral CT angiography in 149 patients. Radiology 1996; 201: 467-470

Goodman LR, Lipchik RJ. Diagnosis of acute pulmonary embolism: Time for a new approach. Radiology 1996;199:25-27

van Erkel AR, van Rossum AB, et al. Spiral CT angiography for pulmonary embolism: A cost-effective analysis. Radiology 1996;201: 29-36

Remy-Jardin M, Remy J, et al. Diagnosis of pulmonary embolism with spiral CT: comparison with pulmonary angiography and scintigraphy. Radiology 1996;200: 699-706

Remy-Jardin M, Remy J, et al. Peripheral pulmonary arteries: Optimization of the spiral CT acquisition protocol. Radiology 1997;204: 157-163

van Rossum AB, Treurniet FE, et al. Role of volumetric computed tomographic scanning in the assessment of patients with clinical suspicion of pulmonary embolism and an abnormal VQ lung scan. Thorax 1996;51: 23-28

Millar GH, Feied CF. Suspected pulmonary embolism. Postgraduate Medicine 1995; vol 97,1:51-58

Carswell H. Pulmonary embolism: CT and MRI assume higher profile. Diagnostic Imaging 1996:28-33

Winston CB, Wechsler RJ, et al. Incidental pulmonary emboli detected at helical CT: effect on patient care. Radiology 1996;201: 23-27