| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Perspective |
1 Department of Thoracic Radiology, Massachusetts General Hospital, Founders 202, 55 Fruit St., Boston, MA 02114.
Received August 18, 2006; accepted after revision November 8, 2006.
Address correspondence to C. Wittram.
OBJECTIVE. The purpose of this article is to describe the techniques to improve motion artifacts, vascular enhancement, flow artifacts, body habitus image noise, vascular opacification in parenchymal lung disease, streak artifacts, and the indeterminate CT pulmonary angiogram. In addition, this article will illustrate the diagnostic criteria of acute and chronic pulmonary emboli.
CONCLUSION. Pulmonary embolism is the third most common acute cardiovascular disease, after myocardial infarction and stroke, and it leads to thousands of deaths each year because it often goes undetected. For the more than 25 years that the direct signs of pulmonary embolism have been available to the radiologist on CT, this noninvasive technique has produced a paradigm shift that has raised the standard of care for patients with this disease.
Keywords: chest • CT arteriography • CT technique • embolism
Pulmonary embolism is the third most common acute cardiovascular disease, after myocardial infarction and stroke, and results in an estimated 200,000-300,000 hospitalizations and 37,000-44,000 deaths per year in the United States [1]. In 1980, Godwin et al. [2] were among the first to describe pulmonary embolism on contrast-enhanced CT. In 1990, the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) study results were published [3]. This large multicenter trial compared ventilation-perfusion (V/Q) scintigraphy with pulmonary angiography and established the diagnostic characteristics of pulmonary embolism on V/Q scintigraphy. The sensitivity of V/Q scintigraphy was found to be 98%, with a specificity of 10% [3]. The potential of the noninvasive technique, CT pulmonary angiography (CTPA), has now been realized at most institutions; it has become the test of choice and thus the de facto standard of care [4]. Recent studies have shown the sensitivity of thin-slice MDCTPA to be 90-100% and the specificity to be 89-94% for the detection of pulmonary emboli to the level of the subsegmental arteries, using pulmonary angiography as the gold standard [5, 6].
A much larger multicenter study has been recently published: The PIOPED II study, which used a composite gold standard, showed that CTPA has a sensitivity of 83% and specificity of 96% for the detection of pulmonary embolism and that combined CTPA and CT venography have a sensitivity of 90% and specificity of 95% for the detection of venous thromboembolic disease [7]. The PIOPED II study found that patients with a low or intermediate clinical probability of pulmonary embolism and normal results on CTPA had a high negative predictive value for PE (96% for patients with a low probability and 89% for patients with an intermediate probability); however, the negative predictive value was 60% in patients with a high probability before CTPA. The positive predictive value of abnormal findings on CTPA was high (92-96%) in patients with an intermediate or high clinical probability but much lower (58%) in patients with a low likelihood of pulmonary embolism. Therefore, additional testing is recommended when the clinical probability is inconsistent with the imaging results [7].
A limitation of the PIOPED II study was that the composite gold standard was not 100% accurate for the diagnosis of venous thromboembolic disease; it therefore follows that the performance of CT was likely better than the results indicate. In the PIOPED II study, among 824 patients with a reference diagnosis and a completed CT study, CTPA was inconclusive in 51 because of poor image quality [7]. A recent study that evaluated the causes of indeterminate CTPA findings found an indeterminate rate of 6.6% [8]. The most common cause was motion artifacts in 74% of the cases; other reasons included poor enhancement (40%), patient habitus (7%), parenchymal disease (12%), and streak artifacts (7%) [8]. The purpose of this article is to describe the techniques used to improve the quality of CT pulmonary angiography and to illustrate the diagnostic criteria of acute and chronic pulmonary emboli. Indirect CT venography will not be dealt with in detail in this article.
CT Technique
At the moment, at our institution, Light-speed (GE Healthcare) 16- and 64-MDCT scanners are used to acquire the images of the thorax in a caudal-cranial direction. The caudal-cranial direction is used because most emboli are located in the lower lobes and, if the patient breathes during image acquisition, there is more excursion of the lower lobes compared with the upper lobes. For IV access, the antecubital vein and an 18- or 20-gauge catheter is preferred. The CT parameters are given in Tables 1 and 2. Images are viewed on a PACS monitor using IMPAX version 4.1 (AGFA) because there is improved accuracy in viewing chest CT cases on a workstation compared with hard-copy film [9, 10]. The images are displayed with three different gray scales for interpretation of lung window (window width, 1,500 H; window level, -600 H), mediastinal window (window width, 350 H; window level, 40 H), and pulmonary embolism-specific (window width, 700 H; window level, 100 H) settings because pulmonary embolism can be missed when a case with very bright contrast is viewed only on mediastinal window settings [11]. The pulmonary embolism-specific settings also help to differentiate between a sharp margined embolus and an ill-defined artifact. However, modified window settings can also increase the conspicuity of artifacts caused by image noise and flow.
|
|
Multiplanar reformation images through the longitudinal axis of a vessel can be used to overcome some of the difficulties encountered with axial-orientated images of obliquely or axially orientated arteries [12]. Also, reformatted images can help to differentiate between some patient, technical, anatomic, and pathologic factors that mimic pulmonary embolism and true pulmonary embolism [11].
Contrast-enhanced helical CT of the veins of the lower extremities is performed using the same contrast bolus as used for chest CT. Images of the iliac, femoral, and popliteal veins are obtained 3 minutes after the onset of the initial contrast injection [13].
How to Reduce Motion Artifacts
Respiratory motion artifacts are the most common cause of an indeterminate CTPA and can be a cause of misdiagnosis of pulmonary embolism. They are best seen on lung window settings that show composite images of vessels [11]. A rapid change in position of vessels on contiguous images also confirms motion artifact. A low-density abnormality that simulates pulmonary embolism may result from partial voluming of vessel and lung [11]. Motion artifact renders the diagnosis of pulmonary embolism at the affected anatomic level indeterminate. The frequency of examinations devoid of motion artifacts is significantly higher for MDCT, which has a shorter breath-hold than single-detector CT [14, 15]. At the moment, the breath-hold required for 16-MDCT is approximately 10 seconds, and for 64-MDCT, less than 3 seconds. In dyspneic patients, oxygen supplementation can help the patient provide the desired period of apnea. The implementation of higher order MDCT scanners should lower the indeterminate CTPA rate due to respiratory motion.
Pulmonary Artery Enhancement
Theory
An increase in the attenuation of blood on CT may be obtained with
intravascular contrast material containing the atoms of iodine or gadolinium.
Previous work has defined the attenuation values of acute and chronic
pulmonary emboli [16].
Combining these values with experimental work by Meaney et al.
[17], it is possible to
calculate the minimum amount of IV attenuation required to perceive pulmonary
emboli on CT. Meaney et al. showed that the detection of a low-contrast
abnormality is not accurate when the SD of the mean of the abnormality exceeds
the difference in the means of the lesion and the surrounding region
[17]. For acute pulmonary
emboli, the mean attenuation value is 33 H (SD, 15 H)
[16].
Because it is important to detect all pulmonary emboli, we should calculate the highest possible attenuation of an acute pulmonary embolism to be the mean plus 3 SDs; this would include 99.75% of all acute emboli, which equates to 78 H. According to Meaney et al. [17], we need attenuation in the artery of at least one more SD; the final figure therefore equals 93 H. The mean attenuation and SD values for chronic pulmonary embolism are 87 and 31 H, respectively. Therefore, the highest possible attenuation value of chronic pulmonary emboli with 3 SDs is calculated to be 180 H. The minimum attenuation of adjacent opacified blood to identify this outlying chronic thrombus is 211 H. The theoretic minimum attenuations of blood required to see all acute and chronic pulmonary venous thromboemboli are 93 and 211 H, respectively.
To detect abnormalities with low differences in CT contrast, and to improve pulmonary embolism conspicuity, it is necessary to adjust the display window widths and levels [17-19]. Also, the decision of the reviewer to interpret a study as adequate or indeterminate will be affected by the interplay of factors that include the size of the suspected embolism, the anatomic level of the vessel being evaluated, and the amount of image noise.
|
Injection of contrast material can be considered in two components: first pass and recirculation. The first-pass effect is optimized by the use of contrast material with 370 mg I/mL. As the injection duration increases, the recirculation of contrast material causes a cumulative effect on enhancement over time [21], so that an increase in time increases the enhancement of the pulmonary arteries during the injection. This enhancement advantage is most optimally used with the empiric delay technique, whereas bolus tracking starts the CT scan earlier on the rise of the enhancement curve and results in worse pulmonary artery enhancement. Although no published data as yet can validate this statement, preliminary work appears to support this observation [22, 23]. One could argue that when the triggering threshold for bolus tracking is increased, CT would start later on the rise of the enhancement curve. However, in cases with poor function of the right side of the heart, the enhancement threshold might never be reached; this leaves the technologist uncertain as to when to start image acquisition.
Empiric scanning delay also has the advantage of reducing operator error and motion artifacts by removing the added complexity of when to start the study based on a threshold value. To comprehensively evaluate for venous thromboembolic disease, patients need to receive a large contrast material bolus to evaluate the lower-limb veins [7]. Using an empiric scanning delay on 16- and 64-MDCT scanners, one aims to be midscan at the peak of pulmonary artery enhancement; therefore, the start of the scanning is calculated to equal the injection time minus half the scanning time. If the size of the IV access catheter does not allow 4 mL/s, then the delay needs to increase, as illustrated in Table 3.
|
If an indeterminate scan occurs with standard delay due to poor enhancement, there is no extravasation of contrast material, and the timing is appropriate, then poor venous flow due to stenosis or obstruction may be a factor [8], in which case a different venous access site may be necessary. A repeat CTPA after hydration of the patient is recommended.
Flow Artifacts
A transient interruption of contrast material consists of a portion of the
pulmonary artery that shows relatively poor enhancement between areas of
higher attenuation both proximally and distally
[24,
25]
(Fig. 1). Comparing patients
with this artifact with age- and sex-matched controls, Wittram and Yoo
[25] showed that the artifact
results from an increase in flow of unopacified blood from the inferior vena
cava. What can be done to avoid this flow phenomenon? A review of the
literature shows that the transient interruption of contrast artifact was seen
in 3% of the study population in that study
[25], whereas in the study by
Gosselin et al. [24], it was
present in 37% of the study group. An interesting major difference between the
studies, and a possible explanation of the difference in frequency, is that
the patients in the study by Wittram and Yoo were instructed to "take a
breath in and hold it" before image acquisition. The patients in the
study by Gosselin et al. were instructed to have five respiratory cycles of
hyperventilation followed by a command of full inspiration 2 seconds before
initial images were obtained
[24]. The hyperventilation
before inspiration and the breath-hold is likely the exacerbating factor of
this artifact. Both studies used the same injection rate, but Gosselin et al.
used single-detector CT whereas Wittram and Yoo used MDCT. However, the number
of detectors should not affect the appearance of this artifact.
The solution to transient interruption of contrast flow of the pulmonary arteries is to reduce the volume of unopacified blood entering the right atrium from the inferior vena cava. Prescanning hyperventilation is likely the cause; with the implementation of faster scanners, prescan hyperventilation should be dropped. Because the venous return from the inferior vena cava to the right atrium is exaggerated with heightened respiratory movements [26], we verbally instruct our patients not to perform an exaggerated inspiration and the CT technologist prompts the patient to "hold your breath" before image acquisition. Further study is required to assess the possible benefits of these maneuvers.
|
|
Patient Habitus
Two major issues are related to imaging pulmonary arteries of large patients: image noise and the volume of IV contrast material. For patients weighing more than 250 lb (113 kg), it is necessary to increase the radiation dose to decrease the amount of image noise. In addition, the protocol is modified to help decrease display image noise and improve scan quality by increasing reconstruction width to 2.5 mm. However, the reconstruction width will decrease the sensitivity of pulmonary embolism detection [28]. In larger patients, for optimal pulmonary artery enhancement, the quantity of contrast material needs to be adapted to the patient's size [29]; to simplify the protocol, 110 mL of 370 mg I/mL contrast material is used for patients weighing 250 lb (113 kg) or less and 130 mL of 370 mg I/mL contrast material is used for those weighting more than 250 lb (113 kg) (Tables 1 and 2).
For pregnant patients, the volume of contrast material should be reduced to 70 mL and the timing adjusted accordingly (Table 4). The reason for this rationale is that the legs and pelvis are not imaged and that the quantity of iodine to the fetus is also reduced.
|
Parenchymal Disease
Consolidation can cause a focal increase in vascular resistance and focal poor vascular opacification [27]. However, the frequency of this artifact will be reduced with the use of empiric timing delay (Fig. 2A, 2B) because image acquisition is performed at the end of the injection. As for reviewing vessels surrounded by consolidation, as with all radiology interpretation, it is important to be systematic and review one vessel at a time and ignore the consolidation or any other pathology that might distract the attention of the reviewer. In this manner, any case with adequate enhancement and no or minimal motion can be confidently interpreted.
Streak Artifacts
Streak artifact that obscures pulmonary vessels because of metallic implants can make a study indeterminate, a repeat CT will not improve this problem, and additional imaging with V/Q scintigraphy or pulmonary angiography may be necessary. Streak artifact from high-density contrast material in the superior vena cava can obscure adjacent pulmonary arteries. The frequency of this artifact can be reduced by using a saline bolus immediately after the contrast material injection [30].
The Indeterminate CTPA
This article discusses the solutions to the common causes of an indeterminate CTPA. In practice, if a diagnosis of pulmonary embolism cannot be confidently confirmed or refuted and the study is indeterminate, it is recommended that the radiologist decide at which anatomic level the study is indeterminate; for example, if the radiologist can clear the vessels to the level of the segmental arteries, and the subsegmental arteries are indeterminate, the clinician might not require further imaging in cases with a low clinical pretest probability for pulmonary embolism. However, some patients with indeterminate CTPA findings will need further imaging, with ultrasound scan of the legs after hydration, a repeat CTPA, V/Q scintigraphy (if the lungs are clear on CT), or pulmonary angiography.
Direct Signs of Acute and Chronic Embolism
Both acute and chronic pulmonary emboli are identified as intraluminal filling defects that show a sharp interface with IV contrast material. The diagnostic criteria for acute pulmonary embolism include, first, complete arterial occlusion with failure to opacify the entire lumen; the artery may be enlarged in comparison with pulmonary arteries of the same order of branching [31-33] (Fig. 3); second, a central arterial filling defect surrounded by IV contrast material [31] (Fig. 4A, 4B); and third, a peripheral intraluminal filling defect that makes an acute angle with the arterial wall [32, 33] (Fig. 5).
|
|
|
|
The diagnostic criteria for chronic pulmonary embolism include complete occlusion of a vessel that is permanently smaller than pulmonary arteries of the same order of branching [32, 33] (Fig. 6), a peripheral eccentric filling defect that makes an obtuse angle with the vessel wall [32, 33] (Fig. 7), contrast material flowing through apparently thick-walled arteries that are smaller due to recanalization [32, 33] (Fig. 8), a band or web in a contrast-filled artery [32, 33] (Fig. 9), and an intraluminal filling defect with an acute pulmonary embolism morphology that has been present for more than 3 months [16].
|
|
|
|
These signs include nonuniform arterial perfusion for both acute and chronic pulmonary embolism; this radiologic sign is difficult to identify in cases of acute pulmonary embolism but manifests as mosaic attenuation in cases of chronic pulmonary embolism. A mosaic pattern of lung attenuation is identified on the lung window settings.
The three major causes of mosaic lung attenuation are airways disease, chronic pulmonary embolism (in which the abnormal region is more radiolucent), and interstitial lung disease (in which the abnormal lung is more opaque). Oligemia, or a decrease in the flow rate due to acute pulmonary embolism, is often identified on angiography [34, 35]. In my experience, this finding is more often seen on angiography than on CT; this discrepancy is thought to be related to the larger temporal window of IV contrast material for CT as compared with angiography. Occasionally, a large acute central pulmonary embolism can cause oligemia and a reversible decrease in vessel diameter; this CT equivalent of the Westermark sign has been previously illustrated [36].
Nonuniform arterial perfusion due to acute pulmonary embolism can uncommonly manifest as a mosaic pattern of attenuation on CT. Additional indirect signs seen in chronic pulmonary embolism include poststenotic dilatation, tortuous vessels, enlargement of the main pulmonary artery, and enlargement of the bronchial arteries [36]. For a long time we have been at a stage at which the direct radiologic signs, as shown on CT angiography, are required to make a diagnosis of acute or chronic pulmonary thromboembolic disease. Because the indirect signs have a differential diagnosis, they are helpful only as indicators of the sites of the direct radiologic signs of pulmonary embolism.
Severity of Acute Pulmonary Embolism
After the initial embolic event, the patient may be at risk for circulatory collapse secondary to right heart failure, and a subsequent embolism may be fatal. It has been suggested that the early detection of acute right ventricular failure allows the implementation of the most appropriate therapeutic strategy [37]. Right ventricular strain or failure is optimally monitored on echocardiography. However, some morphologic abnormalities that indicate right ventricular failure can be quantified by CTPA. The most robust CT sign is right ventricular dilation (in which the greatest right ventricle short-axis measurement is wider than the maximum left ventricle short-axis measurement) [38] (Fig. 10A, 10B). The greater the right ventricle-to-left ventricle short-axis ratio in acute pulmonary embolism, the greater the risk of death [39]. A ratio of 1.0 is associated with a 5% chance of death; 1.3, 10%; 1.7, 20%; 1.9, 30%; 2.1, 40%; and a ratio of 2.3 is associated with a 50% chance of death [39].
|
|
Conclusion
For more than 25 years, the direct signs of pulmonary embolism have been available to the radiologist on CT, and this noninvasive technique has produced a paradigm shift that has raised the standard of care for patients with this disease. This article outlines the approaches necessary to improve the quality of CT pulmonary angiography and summarizes the diagnostic criteria for acute and chronic pulmonary emboli.
References
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  J. D. Dodd, S. Kalva, A. Pena, F. Bamberg, M. D. Shapiro, S. Abbara, R. C. Cury, T. J. Brady, and U. Hoffmann Emergency Cardiac CT for Suspected Acute Coronary Syndrome: Qualitative and Quantitative Assessment of Coronary, Pulmonary, and Aortic Image Quality Am. J. Roentgenol., September 1, 2008; 191(3): 870 - 877. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
