Pulmonary embolism (PE) is a serious condition responsible for significant morbidity and mortality. PE is currently the third leading cause of cardiovascular death in the United States [
MRI offers a potential alternative to CTPA in the evaluation of the pulmonary vasculature and the diagnosis of PE [
This HIPAA-compliant study was approved by our institutional review board. Eligible patients included those diagnosed with PE on CTPA. Exclusion criteria included any contraindication to MRI or gadolinium-based contrast media. All patients were already enrolled in a prospective study to evaluate right ventricular function by cardiac MR (CMR) in patients diagnosed with acute PE on CT pulmonary angiogram (CTPA), with informed written consent obtained at the time of study enrollment. The final study population comprised 11 patients (5 males, 6 females) with an average age (±SD) of
Four different CT scanners were used with the protocols outlined below.
Four studies were conducted on a 16-row CT scanner (Sensation 16, Siemens Medical Solutions, Erlangen, Germany). With the use of helical acquisition, patients were scanned supine in a caudocranial direction with a scan range based on the scout view from 1 cm below the lowest costophrenic angle to 1 cm above the lung apex. This range was intended to completely cover the lungs and to accommodate any slight changes in patient positioning or breath holding between the scout view and the CTPA. Scan parameters were as follows: in-plane field of view FOV, 38 cm; matrix,
Two studies were conducted on a 64-row CT scanner (Light Speed VCT General Electric Healthcare). With the use of helical acquisition, patients were scanned supine in a caudocranial direction from 1 cm below the lowest costophrenic angle to 1 cm above the lung apex. Scan parameters were as follows: FOV, 38 cm; matrix,
One study was conducted on a 16-row CT scanner (Light Speed 16, General Electric Healthcare). With the use of helical acquisition, the patient was scanned supine in a craniocaudal direction from 1 cm above the lung apex to 1 cm below the lowest costophrenic angle. Scan parameters were as follows: FOV, 38 cm; matrix,
Four studies were conducted on a 4-row CT scanner (Light Speed QX/I, General Electric Healthcare). With the use of helical acquisition, the patient was scanned supine in a craniocaudal direction from 1 cm above the lung apex to 1 cm below the lowest costophrenic angle. Scan parameters were as follows: FOV, 38 cm; matrix,
CT image acquisition took on average (±SD)
Images were reviewed at a PACS workstation picture archiving and communication system workstation (PACS, Centricity Radiology RA1000; General Electric Healthcare), where windows and levels could be adjusted by the reviewer.
MRI was obtained an average (±SD) of
Two independent observers reviewed each MRI examination. Each main and lobar pulmonary artery was scored as positive or negative for pulmonary embolism. After this analysis was completed, the reviewers jointly evaluated the CTPA studies, which were used as the gold standard.
Continuous variables are reported as means (±SD). Interobserver agreement was evaluated using the
PE was identified within a total of 48 pulmonary arterial branches by CTPA (Table
Sensitivity, specificity, positive and negative predictive values for detection of pulmonary emboli by pulmonary artery distribution.
Right | Left | |||||||
---|---|---|---|---|---|---|---|---|
Right pulmonary artery | Upper lobe | Middle lobe | Lower lobe | Left pulmonary artery | Upper lobe | Lingula | Lower lobe | |
Sensitivity | 100 (4/4) | 29 (2/7) | 67 (4/6) | 100 (8/8) | 100 (4/4) | 33 (2/6) | 50 (2/4) | 78 (7/9) |
Specificity | 100 (7/7) | 100 (4/4) | 100 (5/5) | 100 (3/3) | 100 (7/7) | 100 (5/5) | 100 (7/7) | 100 (2/2) |
Positive predictive value | 100 (4/4) | 100 (2/2) | 100 (4/4) | 100 (8/8) | 100 (4/4) | 100 (2/2) | 100 (2/2) | 100 (7/7) |
Negative predictive value | 100 (7/7) | 44 (4/9) | 71 (5/7) | 100 (3/3) | 100 (7/7) | 56 (5/9) | 78 (7/9) | 50 (2/4) |
Distribution of pulmonary emboli not identified on MRI.
Right upper | Right middle | Right lower | Left upper | Lingula | Left lower | Total | |
---|---|---|---|---|---|---|---|
Lobar | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
Segmental | 2 | 0 | 0 | 2 | 1 | 1 | 6 |
Subsegmental | 2 | 2 | 0 | 1 | 1 | 1 | 7 |
Not covered | 1 | 0 | 0 | 1 | 0 | 0 | 2 |
59-year-old male with a saddle pulmonary embolus.
Axial CT image demonstrates pulmonary embolism extending into the left and right pulmonary arteries (arrow)
Axial gradient echo cine image from the MRI obtained 42 hours later demonstrates the embolism as a linear hypointense filling defect within the right and left pulmonary arteries (arrow)
47-year-old female with left pulmonary artery and bilateral lower lobe segmental pulmonary artery emboli.
Axial CT image demonstrates pulmonary embolism within the left pulmonary artery (arrow)
Axial bright blood image from the MRI obtained 27 hours later demonstrates the embolism as a filling defect within left pulmonary artery (arrow)
Axial CT image demonstrates pulmonary emboli within the bilateral lower lobe segmental pulmonary artery branches (arrows)
Axial bright blood image from the MRI obtained 27 hours later fails to demonstrate the segmental pulmonary emboli seen on the CT
Of the 11 total patients, 9 had PE identified on MRI, for a per patient sensitivity of 82%. The two patients without identifiable PE by MRI had only segmental and sub-segmental emboli on CTPA.
This study demonstrates the feasibility of diagnosing pulmonary embolism without the use of radiation or intravenous contrast media. Although our MRI protocol was not optimized for the detection of PE, it was still 69% sensitive on a per vessel and 82% sensitive on a per patient basis.
Currently, CTPA is considered the gold standard for the diagnosis of PE [
CTPA also requires the use of iodinated contrast media and, therefore, has the possibility of contrast reactions [
Previous studies examining the use of MRI in the diagnosis of PE have evaluated noncontrast MRI only as part of a larger MRI protocol including contrast-enhanced MRI [
An additional potential diagnostic tool in the evaluation of patients with suspected pulmonary embolism is single photon emission computed tomography ventilation/perfusion lung scan (V/Q SPECT). Although this modality avoids the use of iodinated or gadolinium-based contrast agents, it does still subject the patient to ionizing radiation. One recent study examining the utility of V/Q SPECT found that the prevalence of pulmonary embolism was 88% in patients with a high-probability V/Q SPECT exam [
This study has several limitations. As mentioned above, the MRI protocol was not specifically designed for the evaluation of the pulmonary vasculature. An improved MRI protocol, using multiplanar imaging, decreased slice thickness and separation, increased field-of-view, and increased in-plane resolution would likely improve the sensitivity for detection of MRI. The study is also limited by the small number of patients.
This study was also limited by the timing of the MRI. As MRI was not performed immediately following CTPA, it is possible that by the time of the MRI the emboli seen on the CTPA had been lysed, particularly given that our patients were treated with anticoagulation between the CTPA and the MRI.
An additional potential limitation of MRI for the evaluation of PE is the substantially longer time required for image acquisition. Although in this study all patients were able to complete the MRI, it did take significantly longer than the CTPA (55.6 minutes for MRI versus 2.7 minutes for CTPA,
There is an inherent bias in our study design, in that all patients in the study we known to have PE on CTPA. Therefore when reviewing the MRIs the readers knew that PE should be present, this may have biased their interpretations of the exam.
We conclude that noncontrast MRI offers a means of diagnosing PE without intravenous contrast material or radiation, although further study is necessary before it can be considered a routine tool in the evaluation of the patient with suspected PE. It is our hope that with the increased availability of MRI in the emergency setting and growing concern over radiation exposure, this technique will be used in patients with contraindications to ionizing radiation or intravenous contrast.