Other
Chest Imaging
February 2004

Risk of Pulmonary Embolism After Negative MDCT Pulmonary Angiography Findings

Abstract

OBJECTIVE. The purpose of our study was to determine the risk of pulmonary embolism in patients who have negative MDCT pulmonary angiography findings.
SUBJECTS AND METHODS. In this prospective study, one hundred two consecutive patients with suspected pulmonary embolism underwent MDCT pulmonary angiography. Scans were reviewed jointly by two observers and findings recorded by consensus. Observers noted whether pulmonary embolism or other disease was present. No pulmonary embolism was seen in 85 patients (52 men and 33 women; age range, 20–94 years; mean age, 60 years) who were followed up for a mean of 9 months (range, 4–13 months) for evidence of subsequent pulmonary embolism.
RESULTS. One patient had a diagnosis of pulmonary embolism made within 3 weeks of undergoing CT pulmonary angiography. MDCT pulmonary angiography showed additional potentially significant findings in 76% of patients; 47% of these findings were not suspected on chest radiography.
CONCLUSION. The risk of pulmonary embolism at a mean of 9 months after negative MDCT pulmonary angiography findings is 1%. In our study of patients without pulmonary embolism, MDCT pulmonary angiography revealed other causes for individual patients' signs or symptoms in most cases.

Introduction

Acute pulmonary embolism is a common and often fatal illness with more than 650,000 new cases each year in the United States [1]. Early diagnosis and anticoagulation therapy are essential because they reduce patient mortality from 30% to 2–10% [2]. In contrast to the diagnosis of many other illnesses, diagnosis of pulmonary embolism is almost entirely dependent on radiologic findings.
CT pulmonary angiography has increasingly become the tool of choice in the diagnosis of pulmonary embolism [3]. CT pulmonary angiography has a sensitivity of 60–100% and a specificity of 78–100% in the diagnosis of pulmonary embolism [4]. Patient treatment after a CT pulmonary angiogram that shows no signs of pulmonary embolism is controversial [5]. MDCT pulmonary angiography with reconstructed scans of 1.25-mm thickness allows improved visualization of subsegmental vessels [6]. MDCT technology is expected to improve diagnostic accuracy and therefore may lead to a lower rate of subsequent pulmonary embolism. The aim of our study was to determine the risk of subsequent clinical presentation with pulmonary embolism after an MDCT pulmonary angiogram that showed no signs of embolism.

Subjects and Methods

This prospective study was approved by the hospital ethics committee. Between June 2001 and March 2002, 290 patients presented to our institution with suspected acute pulmonary embolism. One hundred two of these patients were assessed with MDCT pulmonary angiography, and 188 were assessed with ventilation–perfusion scanning. Clinicians were free to order ventilation–perfusion scans or MDCT pulmonary angiograms as they thought appropriate.
Four-slice MDCT pulmonary angiography was performed using a Somatom Volume Zoom scanner (Siemens, Erlangen, Germany). Standard technique consisted of placement of an 18-gauge cannula into the right median cubital vein and bolus injection of 120 mL of 300 mg I/mL of nonionic iodinated contrast medium at 3 mL/sec. Scanning was performed during suspended full inspiration using a bolus-tracking technique with a region of interest centered on the main pulmonary artery [7]. A slice thickness of 1.25 mm was used with a pitch of less than 1.5. Data were reconstructed for viewing using 1.25-mm slice thickness with a slice interval of 1.2 mm. Images were viewed at a dedicated workstation (Magic-view, Siemens) by two radiologists.
Diagnosis of embolism was based on criteria previously described by Tillie-Leblond et al. [8]. The presence of an endoluminal filling defect on CT pulmonary angiogram was considered diagnostic of pulmonary embolism (Fig. 1A, 1B, 1C). Scan quality was also graded, with a grade of 1 indicating an excellent degree of pulmonary arterial enhancement permitting exclusion of pulmonary emboli to the subsegmental level. A grade of 2 indicated scan quality was adequate to exclude emboli to the segmental arterial level. A grade of 3 indicated that embolism could be excluded only at the main and lobar levels, and a grade of 4 indicated that the scan was nondiagnostic for pulmonary embolism. Patients with grades 3 and 4 scans were recommended to undergo further imaging to exclude pulmonary embolus as a cause of symptoms.
Fig. 1A. 50-year-old man with dyspnea and chest pain after coronary artery bypass surgery. MDCT pulmonary angiogram shows healthy right lower lobe pulmonary artery.
Fig. 1B. 50-year-old man with dyspnea and chest pain after coronary artery bypass surgery. Scan through right lower lobe shows subtle filling defects (arrow) in subsegmental lower lobe artery. This finding was initially missed.
Fig. 1C. 50-year-old man with dyspnea and chest pain after coronary artery bypass surgery. MDCT pulmonary angiogram obtained 20 days after A shows filling defect (arrow) at similar level to that seen in A in right lower lobe pulmonary artery, indicating progression of pulmonary emboli.
Additional abnormalities were also noted, including pleural effusion, consolidation, emphysema, bronchogenic neoplasm, atelectasis, and pulmonary fibrosis, that may have accounted for patient symptoms. Findings were recorded by consensus. The reports of chest radiography of patients referred for MDCT pulmonary angiography were reviewed. Chest radiography had been reported without awareness of subsequent CT findings. Referring clinicians were free to order additional examinations such as ventilation–perfusion scanning and pulmonary angiography as they thought appropriate.
Eighty-five patients (52 men and 33 women; age range, 20–94 years; mean age, 60 years) had no CT signs of pulmonary embolism and formed the study group. Patients were followed up for 4–13 months (mean, 9 months). Follow-up was performed using clinical interview by telephone (66 patients), interview with primary care or referring clinician (six patients), chart review (26 patients), review of autopsy reports (five patients), or review of death certificates (12 patients). A diagnosis of subsequent pulmonary embolism was based on findings on subsequent ventilation–perfusion scanning, CT pulmonary angiography, pulmonary angiography, or postmortem examination.

Results

Sixty-four patients (75%) had grade 1 CT pulmonary angiograms. Eighteen patients (21%) had grade 2 CT pulmonary angiograms. Three patients (4%) had grade 3 CT pulmonary angiograms. There were no grade 4 CT pulmonary angiograms. Additional potentially significant findings were made in 64 (75%) of the 85 patients. These included 10 cases of pleural effusion (12%), 15 cases of consolidation (18%), 18 cases of emphysema (21%), four cases of bronchogenic neoplasm (5%), seven cases of segmental or lobar atelectasis (8%), four cases of pulmonary fibrosis (5%), and two cases of florid mediastinal adenopathy (2%). Other diagnoses included one case of pulmonary tuberculosis, one case of an arteriovenous malformation (Fig. 2.), one case of pulmonary metastases (Fig. 3A, 3B, 3C), and one case of congestive cardiac failure. All patients who were referred for MDCT pulmonary angiography with no signs of pulmonary embolism and no additional potentially significant findings on CT had normal findings on chest radiography. Of the patients who had additional potentially significant findings on CT, 17 (27%) had normal findings on chest radiography, 34 (53%) had concordance between findings on chest radiography and CT, and 13 (20%) had abnormal findings on chest radiography but additional potentially significant findings on CT. In these 13 patients, findings seen only on CT included two cases of consolidation, two cases of bronchogenic neoplasm, six cases of emphysema, one case of pulmonary fibrosis, one case of pulmonary metastases, and one case of mediastinal adenopathy.
Fig. 2. 41-year-old woman with dyspnea and chest pain. MDCT pulmonary angiogram shows arteriovenous malformation with draining vein (arrow) at right lung base.
Fig. 3A. 74-year-old man with hemoptysis and chest pain. MDCT pulmonary angiogram obtained at level of left atrium shows left hilar mass (arrow).
Fig. 3B. 74-year-old man with hemoptysis and chest pain. Additional scan obtained at lower level shows metastasis with aortic and vertebral body (arrow) invasion.
Fig. 3C. 74-year-old man with hemoptysis and chest pain. Additional scan obtained of upper abdomen shows right adrenal metastasis (arrow). Bronchoscopic biopsy showed small cell carcinoma.
At follow-up, 17 (20%) of the 85 patients had died. No documented deaths occured as a result of pulmonary embolism. No patients were lost to follow-up. Of the 68 (80%) living patients, six were receiving anticoagulant therapy—two for cerebrovascular disease, two for deep venous thrombosis, and two for pulmonary embolism—despite negative findings on CT pulmonary angiography. One patient had an initial negative CT pulmonary angiogram followed by a high-probability ventilation–perfusion scan and subsequent positive findings on CT pulmonary angiogram within 3 weeks. This case was determined to be a false-negative initial CT pulmonary angiogram. On subsequent review of the patient's initial CT pulmonary angiogram, both radiologists thought that the examination was positive for subsegmental pulmonary emboli. This case was determined to be a false-negative examination because the diagnosis had been missed on the initial scan. This patient's initial CT pulmonary angiogram was considered a grade 1 scan. This patient was in the ICU after cardiac surgery and had multiple comorbid medical conditions.
The negative predictive value of MDCT pulmonary angiography for subsequent clinically significant pulmonary embolism was 99%. If we exclude the patients receiving anticoagulation at time of follow-up, the negative predictive value for subsequent clinically significant pulmonary embolism was 98%.
One hundred eighty-eight patients (104 women and 84 men; age range, 18–95 years; mean age, 53.3 years) were tested by ventilation–perfusion scanning for suspected pulmonary embolism. Ventilation—perfusion scans showed normal findings in 88 patients (47%), low-probability findings in 58 (31%), intermediate-probability findings in nine (5%), and high-probability findings in 33 (18%). These patients were not included in the follow-up group.

Discussion

Although pulmonary angiography has traditionally been regarded as the gold standard for diagnosis of pulmonary embolism, it is undoubtedly a flawed gold standard. In the Prospective Investigation of Pulmonary Embolism Diagnosis study, the rate of nondiagnostic pulmonary angiography was 3% [9]. Isolated subsegmental pulmonary emboli were seen in 6% of patients, and interobserver variability in the diagnosis of emboli at this level was 66% [10].
Clinical follow-up was chosen as the gold standard over pulmonary angiography because recent studies have questioned the accuracy of pulmonary angiography for the detection of isolated subsegmental pulmonary emboli [11, 12]. It is accepted that pulmonary embolism occurring within 3 months of a CT pulmonary angiogram that showed no signs of pulmonary embolism most likely indicates an initial false-negative examination [13, 14]. This approach may miss small emboli that do not present with patient symptoms, but such emboli are unlikely to be clinically significant if no evidence of recurrent thromboembolic disease has been noted 3 months or longer after initial presentation. By contrast, this approach may underestimate the negative predictive value of CT pulmonary angiography for pulmonary embolism because many of the patients studied have life-threatening comorbid medical illnesses at the time of initial examination and would be at increased risk of developing subsequent pulmonary embolism de novo.
This is one of the inherent weaknesses of follow-up studies. In the absence of any objective gold standard test for pulmonary embolism, follow-up studies remain important. This methodology is well established and has been used in several studies to assess the negative predictive value of helical single-detector CT pulmonary angiography for pulmonary embolism [8, 13].
The negative predictive value of MDCT pulmonary angiography for subsequent clinically significant pulmonary embolism in our study was 99%. Previous studies have shown the negative predictive value of helical CT pulmonary angiography to be 98–100% [8, 15]. Tillie-Leblond et al. [8] reported a 2% prevalence of clinically apparent pulmonary embolism at 1 year after a negative single-detector helical CT pulmonary angiography. In this study, the rate of indeterminate CT pulmonary angiography was 12%, and confident evaluation of subsegmental vessels was possible in only 23–30% of patients. Lomis et al. [15] followed up a group of 100 patients for 6 months after negative single-detector CT pulmonary angiography to the level of the segmental pulmonary arteries. These researchers documented no cases of subsequent pulmonary embolism at a mean of 9 months of clinical follow-up. In another study, Bourriot et al. [16] evaluated high-risk patients from cardiology and pneumology wards and found that after a minimum of 6 months of clinical follow-up, the risk of subsequent pulmonary embolism was 1.8–4.9%. Similarly, Goodman et al. [13] followed up patients after negative single-detector helical CT pulmonary angiography for 3 months and showed a rate of subsequent pulmonary embolism of 1%. In a large cohort study that followed up patients after negative electron beam CT pulmonary angiography, Swensen et al. [17] found that at 3 months the cumulative incidence of subsequent deep venous thrombosis or pulmonary embolism was 0.5% and that withholding anticoagulation therapy in patients with a negative CT pulmonary angiogram appears to be a safe therapeutic option. In our study, one patient had an initial false-negative MDCT pulmonary angiogram examination, followed by a high-probability ventilation–perfusion scan and a subsequent positive MDCT pulmonary angiogram within 3 weeks. This case was deemed an initial false-negative because subsequent review of the initial examination deemed the diagnosis of subsegmental pulmonary emboli to have been missed.
Previous studies have excluded from follow-up patients undergoing anticoagulation treatment for reasons other than pulmonary embolism [8, 13, 14]. We elected to include these patients in our clinical follow-up group because they represented 9% of the patient population with negative CT pulmonary angiography. We have also included the values for the remainder of the population with these excluded patients to allow comparison with prior studies.
Other researchers have found helical CT pulmonary angiography to be superior to ventilation–perfusion scanning when determining causes of patients' symptoms if both examinations were negative for pulmonary embolism [18]. Two other studies have found that helical CT pulmonary angiography revealed additional data that suggested or confirmed an alternative diagnosis in 65–67% of patients without pulmonary embolism [19, 20]. Sixty-one (72%) of 85 patients in our study had coexisting respiratory disease, which indicates the high percentage of patients with underlying respiratory disease who present for imaging to exclude pulmonary emboli in routine practice. The presence of underlying lung disease has previously been shown not to affect the negative predictive value of helical CT pulmonary angiography [8], leading to the conclusion that CT pulmonary angiography is the technique of choice in excluding pulmonary embolus in patients with underlying respiratory disease [21]. It is our current practice to offer CT pulmonary angiography as a first-line technique for patients with abnormal chest radiography and for all inpatients, and to reserve ventilation–perfusion scanning as a first-line technique for outpatients with normal chest radiography.
Most patients referred for CT pulmonary angiography with suspected pulmonary embolism do not have emboli seen on CT pulmonary angiography [22]. Unlike ventilation–perfusion scanning, CT pulmonary angiography showed a cause for symptoms in 76% of patients in our study; 47% of these findings were not suspected from findings on chest radiography. Results of this study also show that MDCT pulmonary angiography of patients with no signs of pulmonary embolism correlates with a very low risk of subsequent clinically significant pulmonary embolism. This finding is important in further treatment of these patients because unnecessary referral for pulmonary angiography or treatment with anticoagulation can result in serious complications. Anticoagulation treatment with warfarin is associated with fatal hemorrhage in 0.4% and nonfatal major hemorrhage in 6% of patients [23]. Pulmonary angiography is associated with major complications in 1% of patients and has a high rate of interobserver variability [24, 25]. We recommend that further imaging be reserved for the 4% of patients whose scans permit adequate analysis of only the main and lobar pulmonary arteries.
A limitation of our study was the small sample size and the fact that 17 of 85 patients had died at time of follow-up, 12 of them with no autopsy results. However, on review of the death certificates in these patients, there were no documented deaths caused by pulmonary embolism. The high overall death rate is a reflection of the patient population examined. Referrals for MDCT pulmonary angiography from all departments within the hospital were accepted including a large cohort of intensive care, cardiothoracic, and oncology patients. Patients in these subgroups typically have comorbid life-threatening medical conditions. Previous follow-up studies regarding negative helical CT pulmonary angiography have shown similar death rates: 12–17% at 3 months [13, 17] and 19% at a mean of 9 months' follow-up [15]. Similar death rates have been shown in follow-up studies regarding negative pulmonary angiography: 12% at a minimum of 6 months [26], and in follow-up studies regarding patients who had a low-probability ventilation–perfusion scan, from 5.7% at 3 months [13] to 14% at 5 months [27].
The rate of positive MDCT pulmonary angiography was 16.7%. The rates reported in other studies ranged from 15% to 33% [15, 17, 28]. One of the reasons for the low rate of positive MDCT pulmonary angiograms may be local practice. Referring clinicians did not routinely use d-dimer assays to stratify a patient's risk of pulmonary embolism. A recent study has shown that if a d-dimer assay were included in the diagnostic algorithm of these patients, a negative d-dimer would have rendered unnecessary CT pulmonary angiography in 36% of patients [28].
Currently, CT is undergoing dramatic evolution, and it is clear that CT has dramatically changed the diagnostic approach to suspected pulmonary embolism in the last decade [3, 29]. Since the advent of MDCT in 1998, four-channel systems have become widely available. In this study, four-channel MDCT pulmonary angiography permitted analysis of the subsegmental vessels in 75% of patients. The sole use of MDCT is one of the unique features of our study. We know of only one other study addressing the issue of negative MDCT pulmonary angiography [21]. That study showed that when compared with single-detector helical CT, four-channel MDCT pulmonary angiography gives better image quality and comparable negative predictive value [21]. It is certain that further improvements in image quality and spatial resolution will be seen with the introduction of 16-channel CT scanners [30].
In conclusion, an MDCT pulmonary angiogram with no signs of pulmonary embolism correlates with a low risk of subsequent clinically significant pulmonary embolism. In patients without pulmonary embolism, MDCT pulmonary angiography revealed other causes of their symptoms in most cases.

Footnote

Address correspondence to E. C. Kavanagh ([email protected]).

References

1.
Bell WR, Simon TL. Current status of pulmonary thromboembolic disease: pathophysiology, diagnosis, prevention and treatment. Am Heart J 1982; 103:239 –262
2.
Carson JL, Kelley MA, Duff A, et al. The clinical course of pulmonary embolism. N Engl J Med 1992; 326:1240 –1245
3.
Washington L, Goodman LR, Gonyo MB. CT for thromboembolic disease. Radiol Clin North Am 2002; 40:751 –771
4.
Remy-Jardin M, Remy J, Deschildre F, et al. Diagnosis of pulmonary embolism with spiral CT: comparison with pulmonary angiography and scintigraphy. Radiology 1996; 200:699 –706
5.
Rathbun SW, Raskob GE, Whitsett TL. Sensitivity and specificity of helical computed tomography in the diagnosis of pulmonary embolism: a systematic review. Ann Intern Med 2000; 132:227 –232
6.
Ghaye B, Szapiro D, Mastora I, et al. Peripheral pulmonary arteries: how far in the lung does multi-detector row spiral CT allow analysis? Radiology 2001; 219:629 –636
7.
Kircher J, Kickuth R, Laufer U, Noack M, Liermann D. Optimised enhancement in helical CT: experiences with a real-time bolus tracking system in 628 patients. Clin Radiol 2000; 55:368 –373
8.
Tillie-Leblond I, Masrota I, Radenne F, et al. Risk of pulmonary embolism after a negative spiral CT angiogram in patients with pulmonary disease: 1-year clinical follow-up study. Radiology 2002; 223:461 –467
9.
[No authors listed] Value of the ventilation/perfusion scan in acute pulmonary embolism; results of the prospective investigation of pulmonary embolism diagnosis (PIOPED)—the PIOPED investigators. JAMA 1990; 263:2753 –2759
10.
Stein PD, Henry JW. Prevalence of acute pulmonary embolism in central and subsegmental pulmonary arteries and relation to probability interpretation of ventilation/perfusion scans. Chest 1997; 111:1246 –1248
11.
Diffin DC, Leyendecker JR, Johnson SP, Zucker RJ, Grebe PJ. Effects of anatomic distribution of pulmonary emboli on interobserver agreement in the interpretation of pulmonary angiography. AJR 1998; 171:1085 –1089
12.
Baile EM, King GG, Muller NJ, et al. Spiral computed tomography is comparable to angiography for the diagnosis of pulmonary embolism. Am J Respir Crit Care Med 2000; 161:1010 –1015
13.
Goodman LR, Lipchik RJ, Kuzo RS, Liu Y, McAuliffe TL, O'Brien DJ. Subsequent pulmonary embolism: risk after a negative helical CT pulmonary angiogram—prospective comparison with scintigraphy. Radiology 2000; 215:535 –542
14.
Carson JL, Terrin ML, Duff A, Kelley MA. Pulmonary embolism and mortality in patients with COPD. Chest 1996; 110:1212 –1219
15.
Lomis NN, Yoon HC, Moran AG, Miller FJ. Clinical outcomes of patients after a negative spiral CT pulmonary arteriogram in evaluation of acute pulmonary embolism. J Vasc Interv Radiol 1999; 10:707 –712
16.
Bourriot K, Couffinhal T, Bernard V, Montaudon M, Bonnet J, Laurent F. Clinical outcome after a negative spiral CT pulmonary angiographic finding in an inpatient population from cardiology and pneumology wards. Chest 2003; 123:359 –365
17.
Swensen SJ, Sheedy PF 2nd, Ryu JH, et al. Outcomes after withholding anticoagulation from patients with suspected acute pulmonary embolism and negative computed tomographic findings: a cohort study. Mayo Clin Proc 2002; 77:130 –138
18.
Cross JJ, Kemp PM, Walsh CG, Flower CD, Dixon AK. A randomised trial of spiral CT and ventilation perfusion scintigraphy for the diagnosis of pulmonary embolism. Clin Radiol 1998; 53:177 –182
19.
Kim KI, Muller NJ, Mayo JR. Clinically suspected pulmonary embolism: utility of spiral CT. Radiology 1999; 210:693 –697
20.
Remy-Jardin M, Remy J, Baghaie F, Fribourg M, Artaud D, Duhamel A. Clinical value of thin collimation in the diagnostic workup of pulmonary embolism. AJR 2000; 175:407 –411
21.
Remy-Jardin M, Tillie-Leblond I, Szapiro D, et al. CT angiography of pulmonary embolism in patients with underlying respiratory disease: impact of multislice CT on image quality and negative predictive value. Eur Radiol 2002; 12:1971 –1978
22.
Ginsberg MS, Oh J, Welber A, Panicek DM. Clinical usefulness of imaging performed after CT angiography that was negative for pulmonary embolus in a high-risk oncologic population. AJR 2002; 179:1205 –1208
23.
Landefeld CS, Beyth J. Anticoagulant-related bleeding: clinical epidemiology, prediction and prevention. Am J Med 1993; 95:315 –328
24.
Zukerman DA, Sterling KM, Oser RF. Safety of pulmonary angiography in the 1990's. J Vasc Interv Radiol 1996; 7:199 –205
25.
Quinn MF, Lundell CJ, Klotz TA. Reliability of selective pulmonary arteriography in the diagnosis of pulmonary embolism. AJR 1987; 149:469 –471
26.
Novelline RA, Baltarowich OH, Althansaoulis CA, Waltman AC, Greenfield AJ, McKusick KA. The clinical course of patients with suspected pulmonary embolism and a negative pulmonary arteriogram. Radiology 1978; 126:561 –567
27.
Lee ME, Biello DR, Kumar B, Siegel BA. Low-probability ventilation-perfusion scintigrams: clinical outcomes in 99 patients. Radiology 1985; 156:497 –500
28.
Burkill GJ, Bell JR, Chinn RJ, et al. The use of a d-dimer assay in patients undergoing CT pulmonary angiography for suspected acute pulmonary embolism. Clin Radiol 2002; 57:41 –46
29.
Ghaye B, Remy J, Remy-Jardin M. Non-traumatic thoracic emergencies: CT diagnosis of acute pulmonary embolism—the first 10 years. Eur Radiol 2002; 12:1886 –1905
30.
Fleischmann D. Present and future trends in MDCT applications: CT angiography. Eur Radiol 2002; 12[suppl 2]:S11 –S16

Information & Authors

Information

Published In

American Journal of Roentgenology
Pages: 499 - 504
PubMed: 14736689

History

Submitted: March 20, 2003
Accepted: August 25, 2003

Authors

Affiliations

E. C. Kavanagh
All authors: Department of Radiology, Mater Misericordiae Hospital, Eccles St., Dublin 7, Ireland.
A. O'Hare
All authors: Department of Radiology, Mater Misericordiae Hospital, Eccles St., Dublin 7, Ireland.
G. Hargaden
All authors: Department of Radiology, Mater Misericordiae Hospital, Eccles St., Dublin 7, Ireland.
J. G. Murray
All authors: Department of Radiology, Mater Misericordiae Hospital, Eccles St., Dublin 7, Ireland.

Metrics & Citations

Metrics

Citations

Export Citations

To download the citation to this article, select your reference manager software.

Articles citing this article

Media

Figures

Other

Tables

Share

Share

Copy the content Link

Share on social media