| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters to the Editor |
,
Marc R. Sapoval, MD,*,
Jean-François Paul, MD,
and
Jean-Claude Gaux, MD*
Department of Cardiovascular Radiology, Hôpital Européen Georges Pompidou, Paris, France*
Department of Imaging and Interventional Radiology, René Dubos Hospital, 6 avenue de l’ile de France, 95303 Cergy-Pontoise, France. e-mail: htabada@yahoo.com
Department of Radiology, Hôpital Marielannelongue, Le Plessis Robinson, France
Editor:
We read with great interest the article of Dr Rozenblit and colleagues in the May 2003 issue of Radiology (1) in which the authors explained the use of unenhanced computed tomography (CT) with delayed acquisition to depict endoleak in patients treated with aortic stent-graft implantation.
We believe that the problem of detecting leaks appropriately is not solved in this study. The authors made their best efforts to assess the ability of different CT protocols to depict endoleaks. This would suggest that CT findings are compared against a reference standard. In their article, this is not the case because in the group with positive findings, most were considered positive only at CT (27 of 33 patients). Thus, the diagnosis method to be evaluated is assessed against itself and not against another diagnosis method accepted as the standard of reference.
In addition, the group with negative findings is also flawed. Findings are considered negative according to CT by means of selecting a group of patients with evidence of absence of endoleak when correlated with aneurysm evolution after 12 months of follow-up. This demonstrated a stable aneurysm or an aneurysm that decreased in size. The authors found, in this group, a high percentage of patients (20%) in which CT was unable to depict the presence or absence of leakage on the basis of the biphasic protocol considered solely. Again, in this group, CT is compared against CT, and no sound conclusion can be drawn.
Furthermore, the authors did not mention if they used diameter measurements or volume calculation: This is crucial in the determination of aneurysm evolution. We demonstrated (2) that the method used to differentiate most accurately between stable increasing or decreasing aneurysms was volume calculation. In our study, we found that an aneurysm that appears stable at diameter measurement could increase or decrease when volume was calculated, leading to a change in the classification of such aneurysms. Consideration of volume calculation might reduce the number of undetermined cases.
We thank Dr Rozenblit and colleagues again for their interesting study and for their efforts to look for the optimal way to follow up patients after endovascular repair, which still represents a challenge in our clinical practice.
REFERENCES
Department of Radiology, Montefiore Medical Center, 111 East 210th Street, Bronx, NY 10467
We read with interest the critique of our study by Dr Abada and colleagues.
Dr Abada and colleagues claim that in our study, "the diagnosis method to be evaluated [CT] is assessed against itself and not against another diagnosis method accepted as the standard of reference." The diagnostic value of CT for detection of endoleaks has been assessed previously and confirmed by numerous investigators in the past (1–6). It was not our goal to compare the effectiveness of CT with another standard of reference for detection of endoleaks. Rather, the purpose of our study was to determine the best combination of CT acquisitions that is accurate for both detection and exclusion of endoleaks. Therefore, our standard of reference was the combination of biphasic and unenhanced CT, against which two other sets of CT images were compared. The diagnosis of endoleak in our positive group was based on the detection of contrast medium outside the graft but within the aneurysm sac. This criterion is specific for endoleaks and has been used by other investigators (5). Moreover, all endoleaks in our study were confirmed by means of either multiple follow-up CT examinations or angiography.
According to Dr Abada and colleagues, our "group with negative findings is also flawed. Findings are considered negative according to CT by means of selecting a group of patients with evidence of absence of endoleak when correlated with aneurysm evolution after 12 months of follow-up." In this statement, the follow-up period is misquoted. The actual mean follow-up period for our negative group was 27 months (range, 12–64 months), during which a decrease in size of the aneurysm was found in 92% of patients. The criteria we applied to select patients for the negative group were more strict than those suggested by the Society for Vascular Surgery reporting standards for the definition of primary clinical success of the endovascular procedure (7).
Dr Abada and colleagues pointed out that we did not specify if we used diameter measurements or volume calculation and that "this is crucial in the determination of aneurysm evolution." In fact, it is stated in the Materials and Methods section that we measured the largest diameter of the aneurysm perpendicular to the aortic axis on the transverse images by using an electronic cursor.
We did not attempt to address the issue of interobserver variability for diameter measurements, which is discussed in the recent article by Dr Abada and colleagues (8). Aortic volume calculation advocated by Dr Abada and colleagues may prove to be more accurate. With our current equipment, however, volume calculation is cumbersome, time-consuming, and prone to serious errors. Alternatively, our measurement approach is standard practice that is appropriate to the current level of radiology software in wide use worldwide.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
