| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Radiology University of Michigan Health System Ann Arbor, MI 48109-0326
We thank Drs. Kashif Ashraf and Omer Ashraf for their letter regarding our review of CT pulmonary angiography and venography for the evaluation of venous thromboembolic disease [1], in which they raise concern for the use of CT as a primary screening tool for the diagnosis of acute pulmonary embolism. Specifically, they refer to one report of greater sensitivity for ventilation–perfusion (V/Q) SPECT scintigraphy (97%) imaging with ultrafine aerosol, compared to CT pulmonary angiography (86%) [2]. In the referenced study of 83 patients, 37 patients or 44.6% had pulmonary embolism, which is much higher than seen in clinical practice, suggesting that the sample population was biased, which could artificially alter the accuracy data for any or all of the reported tests. Furthermore, there was no significant difference in overall diagnostic accuracy between the two techniques (94% for V/Q SPECT vs 93% for CT) and the specificity for the diagnosis of pulmonary embolism was higher for CT (98% vs 91%).
With respect to the greater sensitivity of SPECT for PE to which the authors refer, this was compared with an older generation of MDCT scanners (4-MDCT). The authors quote increased detection of subsegmental embolism as one of the strong points of V/Q scintigraphy; however, in that study [2] an intermediate probability V/Q result was seen in 26.5% with SPECT when compared to 28.9% with planar imaging, requiring an additional test to rule PE either in or out. While there are well-developed schemes to interpret V/Q scans that are associated with probabilities of PE, such as the Biello criteria [3, 4], there are no data to indicate what the probabilities are for V/Q SPECT findings. All of the existing probability schemes were developed from planar V/Q scans. It would be incorrect to infer the same probabilities seen with small-, medium-, and large-sized defects of various numbers from V/Q planar scans to V/Q SPECT scans.
Most of the published V/Q data to date have used conventional planar imaging rather than SPECT for the diagnosis of PE. We acknowledge that state-of-the-art SPECT imaging may be superior to planar V/Q imaging; however, this requires validation in larger samples of patients to draw conclusions. Currently, the use of V/Q SPECT is not universal; with most V/Q scans performed using the planar technique. Bajc et al. [5] reported a higher sensitivity for V/Q SPECT over V/Q planar imaging using pulmonary angiography as the reference standard in an experimental porcine model; however, the pulmonary arterial branching pattern in that model is substantially different from humans, and this type of evaluation requires validation in humans. When such data are available, the role of SPECT imaging in the diagnostic algorithm of PE diagnosis will be established.
As with newer V/Q technology, with newer generation MDCT scanners there may be increased sensitivity at both the segmental and subsegmental levels using 16-MDCT (Patel S, et al.,16-slice MDCT optimization of small pulmonary artery visualization for pulmonary embolism detection vs 4-slice MDCT, presented at SCBT/MR Annual Meeting, March 2003). The ability to perform CT venography at the same sitting increases the overall diagnostic yield for combined CT pulmonary angiography/venography (CTPA/CTV) over CTPA alone for venous thromboembolic disease, something not possible with V/Q scintigraphy alone. In our thoracic CT practice, CTPA/CTV is performed exclusively on 16-MDCT and greater scanners.
We welcome the correspondence from the authors and hope that this will lead to further research in which state-of-the-art V/Q SPECT and the latest generation of MDCT scanners are compared, to further improve our understanding of these techniques in the diagnosis of venous thromboembolic disease.
 Top References |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
