Recent eLetters to the Editor are available at radiology.rsnajnls .org. eLetters that are no longer posted under ‘‘Recent eLet- ters’’ can be found as a link in the related article or by brows- ing through past Tables of Contents. Delayed Helical CT Acquisition in the Detection of Endoleak [letter] From: Jafar Golzarian, MD Department of Radiology, University of Iowa 200 Hawkins Drive, 3957 JPP, Iowa City, IA 52242 e-mail: jafar-golzarian@uiowa.edu Editor: I read with interest the article by Dr Rozenblit and colleagues in the May 2003 issue of Radiology (1), in which they dem- onstrate the role of delayed acquisition of computed tomo- graphic (CT) images in the evaluation of endoleak. Although the authors used a CT protocol that changes in relation to the technical evolution of CT scanners, their results confirm our findings when we first reported the importance of using delayed acquisition with biphasic helical CT (2). Very few articles are available on this topic. While it is generally ac- cepted that CT follow-up should include delayed acquisition in patients after endoluminal repair of aortic aneurysms, there is not a standardized CT technique that has been widely accepted. The initial CT technique at the time of publication of our article was based on our ongoing protocol used since 1994, with use of 5-mm section thickness. Like Dr Rozenblit and colleagues, we have changed our protocol to a 3-mm thick- ness and pitch of 2 (3). However, the delayed acquisition needs to be performed with the exact same parameters used for arterial phase acquisition. Moreover, the sections should be acquired with the same table position to be correctly comparable and allow detection of small leaks. It is for this reason that we acquire delayed images starting at the same proximal level (1 cm above the proximal end of the stent- graft) as that used for arterial phase images, with the same collimation. To reduce the radiation dose, the delayed acquisition cov- ers only the volume containing the stent-graft. With regard to unenhanced CT, we believe that with the above technique, there is no need to obtain systematically unenhanced acqui- sitions. If at the end of the delayed acquisition there is still a concern, another delayed section (obtained 1 minute after the end of the delayed phase) can still be acquired at the same table position to allow the distinction between calcifications and small leaks. I would like to congratulate the authors again for this interesting article. References 1. Rozenblit AM, Patlas M, Rosenbaum AT, et al. Detection of endoleaks after endovascular repair of abdominal aortic aneu- rysm: value of unenhanced and delayed helical CT acquisitions. Radiology 2003; 227:426 – 433. 2. Golzarian J, Dussaussois L, Abada HT, et al. Helical CT of the aorta after endoluminal stent-graft therapy: value of biphasic acquisition. AJR Am J Roentgenol 1998; 171:329 –331. 3. Golzarian J, Dussaussois L, Struyven J. Imaging of abdominal aortic aneurysms after endoluminal repair. Semin Ultrasound CT MR 1999; 20:16 –24. Dr Rozenblit and colleagues respond: We thank Dr Golzarian for his interest in our study (1) and appreciate his comments. We share his opinion that the craniocaudal extent of the delayed CT acquisition should be limited to the endovascular stent-graft for the reduction of radiation exposure; this is our technique, as well. Our study findings confirm those of Dr Golzarian and colleagues (2,3) regarding the value of delayed CT for detection of endoleaks. However, our delayed CT acquisitions are obtained with thicker sections than those used with arterial phase CT for two reasons. First, we use the delayed series specifically for the detection of low-flow endoleaks, which, we believe, do not require thin sections because of their relatively large size. Second, we do not need to use the delayed series to differen- tiate calcifications from endoleaks, since this is accomplished by comparing the arterial phase images with the unenhanced images. Dr Golzarian states, “the delayed acquisition needs to be performed with the exact same parameters used for arterial phase acquisition.” This may perhaps be necessary for bipha- sic CT images obtained without correlative unenhanced CT images. The use of identical section thicknesses and locations for the two series might then be of value to help avoid indeterminate results. Unfortunately, we have not found ref- erences on the evaluation of the specificity of this technique for the detection of endoleaks. In the biphasic arm of our study, which did not involve use of identical parameters for both series, we had a 20% indeterminate rate for endoleaks. Dr Golzarian believes that “there is no need to obtain systematically unenhanced acquisitions,” and that patients should be monitored prospectively during scanning to iden- tify those who need additional delayed imaging for differen- tiation between endoleak and calcifications. This is a valid approach. However, it cannot be implemented in our prac- tice, because the monitoring of routine outpatient CT scan- ning is not feasible. We also believe that unenhanced CT is an integral part of a CT angiographic study, analogous to a mask image obtained for digital subtraction angiography. Other authors report a similar approach to the role of the unenhanced CT scan (4,5) in CT angiography. Unfortu- nately, prospective randomized studies have not been per- formed to compare the effectiveness of different approaches to CT angiography of endografts. References 1. Rozenblit AM, Patlas M, Rosenbaum AT, et al. Detection of endoleaks after endovascular repair of abdominal aortic aneu- rysm: value of unenhanced and delayed helical CT acquisitions. Radiology 2003; 227:426 – 433. 2. Golzarian J, Dussaussois L, Abada HT, et al. Helical CT of the Letters to the Editor Volume 230  Number 1 Radiology  299 R adiology