| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, University of Chicago, Chicago, IL.
2 Mallinckrodt Institute of Radiology, Washington University School of Medicine,
St. Louis, MO.
3 Department of Surgery, Division of Vascular Surgery, Washington University
School of Medicine, St. Louis, MO.
4 Department of Radiology, University of Pittsburgh School of Medicine, 3362
Fifth Ave., Pittsburgh, PA 15213.
Received January 12, 2008;
accepted after revision March 19, 2008.
Address correspondence to K. T. Bae
(baek{at}upmc.edu).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. One hundred forty-four patients underwent periodic follow-up CT examinations 12–72 months after abdominal aortic aneurysm repair. The CT protocol consisted of an unenhanced scan and contrast-enhanced scans in the arterial and 90-second delayed phases. The endoleaks detected on dual-phase CT scans were evaluated in association with the outcome (therapeutic intervention or endoleak resolution).
RESULTS. The 144 patients underwent 728 CT examinations with a mean follow-up period of 35.5 ± 14.5 months. Fifty endoleaks were detected in 50 (34.7%) of the patients. Eight endoleaks were detected in the arterial phase only, eight in the delayed phase only, and 34 in both phases. Intervention was performed to manage 16 endoleaks detected in both phases. CT showed that three endoleaks were stable (two in the arterial phase only and one in both phases) and that 31 had resolved completely (six in the arterial phase only, eight in the delayed phase only, and 17 in both phases). This finding represents a higher frequency of resolution of endoleaks detected in one phase only than in both phases (Fisher's exact test, p = 0.006).
CONCLUSION. Endoleaks detected only in the delayed phase of CT had resolved spontaneously without intervention. Therefore, we can consider eliminating the delayed phase of acquisition to minimize radiation exposure.
Keywords: aneurysm • angiography • aortic aneurysm • arteries • CT • graft • prosthesis
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CT is the imaging technique of choice for routine follow-up after EVAR [5, 10–12]. A common CT protocol for this purpose includes unenhanced, arterial, and delayed phase contrast-enhanced scans [10–13]. Scanning during both the arterial and the delayed phases has been recommended because it improves the sensitivity of endoleak detection compared with scanning in one phase [10, 11]. However, the advantage of two-phase CT must be weighed against repeated radiation exposure of patients [14], inconvenience to patients and radiologists, and the cost involved.
Although CT during the delayed phase in addition to the arterial phase may show more endoleaks [15], it is not clear whether the additional endoleaks detected with delayed acquisition have the same clinical sig nificance and treatment implications as those detected on arterial phase CT scans. We postulated that endoleaks detected only on delayed CT scans would have a better prognosis than those detected on arterial phase CT scans. The purpose of this study was to evaluate the clinical significance of endoleaks detected at combined arterial and delayed contrast-enhanced follow-up CT exam inations of patients who have undergone endovascular repair of AAA.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CT Protocol and Analysis
Before EVAR, all patients underwent contrast-enhanced CT, the findings of
which served as the baseline for follow-up examinations. After EVAR, patients
underwent CT at intervals of 0, 1, 6, and 12 months and yearly thereafter. The
interval for follow-up CT was shortened to 3 or 6 months in cases of
persistent endoleak lasting 6 months or longer, late occurrence of a new
endoleak, or enlargement of the aneurysm. All patients underwent follow-up CT
(Somatom Plus S scanner, Siemens Medical Solutions). The CT protocol consisted
of unenhanced, arterial, and delayed phase contrast-enhanced scans. No oral
contrast agent was administered. The gantry rotation time, tube voltage, and
current for all three acquisitions were 0.5 second, 120 kVp, and 180–250
mAs. First, unenhanced CT covering the entire volume of the abdomen and pelvis
was performed at 2.5-mm collimation and 5-mm slice thickness. Then
125–160 mL of nonionic contrast material (ioversol, Optiray 350,
Mallinckrodt Imaging) was administered into an antecubital vein with a power
injector at a rate of 4 mL/s. With bolus tracking, the arterial phase CT scans
were acquired from above the origin of the superior mesenteric artery to the
symphysis pubis with 1-mm collimation and 1.25-mm slice thickness. Ninety
seconds after the start of administration of the contrast bolus, delayed phase
CT acquisition at 2.5-mm collimation and 3-mm slice thickness was performed
covering the same range as the arterial phase acquisition.
All preoperative and follow-up CT scans were reviewed by attending radiologists, and the CT reports were used for data analysis. Multiplanar reformations were generated from the axial images for quantitative evaluation. CT images acquired during the arterial and delayed phases were assessed to determine the presence and types of endoleaks. The endoleaks were defined as type I, endoleak at the graft attachment site; type II, endoleak originating from the lumbar artery or inferior mesenteric artery; type III, endoleak due to graft separation; type IV, endoleak resulting from graft porosity; and type V, endoleak with indeterminate source. Information on the progression, regression, or resolution of endoleaks over time was noted. The feeding vessels, an eurysm size (maximal diameter perpendicular to the longitudinal axis of the aorta measured on sagittal and coronal images), trend in size change, and graft-related complications (migration, dil a tation, extrusion, erosion) were noted.
Interventions for Endoleak
Our institutional protocol called for aggressive management of type I and
III endoleaks. Type II endoleaks were managed conservatively. Our protocol
included embolization of type II endoleaks that persisted for more than 6
months and were associated with aneurysmal sac growth of 5 mm or more
[16]. Before management of
presumed endoleaks, detailed transfemoral arteriography was performed with
selective injection of the hypogastric arteries and the superior mesenteric
artery to evaluate the contribution of the lumbar and inferior mesenteric
arteries to the endoleak and to rule out an attachment site leak (type I) or
leak arising from the endograft itself (type III). The interventions needed
for the endoleaks were performed 20 hours after arteriography. Type I or III
endoleaks were managed with endovascular placement of an extension graft or
cuff. Type II endoleaks were sealed either by injection of N-butyl
cyanoacrylate glue (Trufill, Cordis Endovascular) into the aneurysmal sac
through a translumbar approach or by transluminal coil embolization of one or
more lumbar arteries or the inferior mesenteric artery through the collateral
vessels at their origin from the aneurysmal sac.
Statistical Analysis
The yearly number of follow-up CT examinations of patients with and without
endoleaks was compared by calculation of means, and differences were tested
for statistical sig nificance with an independent samples Student's t
test. Aneurysmal size change over time was studied by calculation of the
change from baseline at subsequent follow-up intervals. Change versus time was
plotted in scatterplots, and patterns were tested for statistical significance
with analysis of variance. The association between CT findings (endoleaks
detected in a single phase and in combined phases) and outcome (resolution)
was examined with contingency tables. Patterns were tested for statistical
significance with Fisher's exact tests. Results were expressed as mean
± SD, and p < 0.05 was considered significant. All analyses
were performed with JMP 4.0 software (SAS Institute).
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Type of Endoleak
Among the 50 patients with endoleaks, CT depicted type I endoleak in nine
(18%), type II endoleak in 32 (64%), type III endoleak in two (4%), and type V
endoleak in seven (14%) of the cases. Nine type I endoleaks were detected at
the proximal (n = 5), distal (n = 3), and junctional
(n = 1) stent attachment sites. Thirty-two type II endoleaks were
identified with evidence of collateral flow through a patent lumbar artery
(n = 17), inferior mesenteric artery (n = 7), or both
(n = 8). The two type III endoleaks were caused by separation of the
proximal cuff and main body of the stents. Eighteen patients with endoleaks
underwent catheter arteriography to confirm the sources of endoleak or assess
for endoleak treatment. Table 1
compares CT and catheter arteriography in the diagnosis of endoleak. The other
graft-related complications in these 50 cases were two cases of stent
migration and one case of dilation.
|
Follow-Up CT of Endoleak
Table 2 shows the number and
type of endoleaks detected on the arterial and delayed phase CT scans.
Thirty-four of the 50 endoleaks were visualized on both arterial and delayed
phase CT scans; the other 16 endoleaks were detected either on arterial CT
scans only (n = 8) or on delayed CT scans only (n = 8). All
type I and III endoleaks were visualized on arterial phase CT scans; five
(16%) of 32 type II endoleaks were missed on arterial phase CT scans but were
detected on delayed phase CT scans only. With combined arterial and delayed
phase scans as a reference standard and detection of an endoleak on either
phase of scan considered a positive finding, the sensitivity of arterial phase
CT scans alone in the detection of endoleaks was 84% (42 of 50 cases).
|
Outcome of Endoleak
Of the 34 endoleaks detected on both arterial and delayed phase CT scans,
16 were managed after the initial detection of the leak. In seven of these
cases, treatment was endovascular placement of an extension graft or cuff for
type I or III endoleak. In nine of the 16 cases, 6–24 months after the
initial detection of the endoleak and aneurysmal sac growth, glue or coil
embolization was performed for type II endoleaks. All 16 endoleaks were
successfully excluded with no evidence of leak recurrence (Fig.
1A,
1B,
1C,
1D). In 17 (50%) of the 34
cases of endoleak detected in both phases, the leak completely resolved
2–36 months after the initial detection of endoleak. In the last of the
34 cases, the endoleak was stable with no aneurysmal sac growth at the end of
the follow-up period. Six (75%) of the eight endoleaks detected only on
arterial phase CT scans resolved, and the other two remained stable. All eight
of the endoleaks detected only on delayed phase CT scans resolved and
necessitated no intervention (Fig.
2A,
2B,
2C,
2D). The difference in the
frequency of resolution between the endoleaks detected in the arterial phase
only and in the delayed phase only was not statistically significant (Fisher's
exact test, p = 0.47). The frequency of resolution, however, was much
higher for endoleaks detected only in one phase than for those detected in
both phases (Fisher's exact test, p = 0.006).
Figure 3 shows the outcomes of
endoleaks detected on CT scans in the various phase conditions.
|
|
|
|
|
|
|
|
|
Changes in Aneurysm Size Associated with Endoleak
The mean maximal aneurysmal diameter among the 50 patients with endoleaks
was 54.0 ± 8.4 mm at baseline (range, 32–73 mm). The 50 patients
underwent 325 CT exam inations in the follow-up period eval uated. The
measurements of aneurysmal dia meter showed an overall small decrease of 0.2
mm from baseline, almost evenly bal anced be tween increases and decreases.
Although it was statistically significant (anal ysis of var i ance, p
= 0.0015), this change had questionable validity because there were probably
two patterns from multiple counting. Therefore, we examined the trend in
change in aneurysmal diameter by categorizing the endoleaks by detection
approach (arterial, delayed, and both phases) and outcome (resolution and
intervention) and made the following findings.
The pattern of diameter change in aneurysms associated with the 34 endoleaks detected in both phases was similar to the overall pattern described earlier; most of the diameters were stable, and a few increased over time. An overall diameter increase of 0.4 mm was statistically significant (analysis of variance, p = 0.0003) but of questionable validity. The diameter changes of the aneurysms associated with eight endoleaks de tected only in the arterial phase appeared to have two patterns: one of stable size and the other a large increase in size over time. Although the numbers of aneurysms with these two patterns appeared equal, the large increase outweighed stability for an overall diameter increase of 0.7 mm. Again, re gression analysis was not a valid approach to a pattern such as this, and the pattern was not statistically significant (analysis of variance, p = 0.11). The diameter of the eight aneurysms detected only in the delayed phase had almost entirely small changes, predominantly negative (Fig. 4). The overall trend to a 2.4-mm decrease in diameter was statistically significant (analysis of variance, p = 0.0045). The 16 patients who underwent treatment of endoleaks had a clear increase in most diameter measurements after approximately 1 year of stable measurements (Fig. 5). An overall diameter increase of 1.8 mm was statistically significant (analysis of variance, p < 0.0001). The 31 patients in whom endo leaks resolved accounted for most of the changes in aneurysmal diameter. The tendency for most of the measurements to decrease slightly balanced the tendency of a few to increase greatly, resulting in an almost level overall trend (Fig. 6). An overall diameter decrease of 1.4 mm was not statistically significant (analysis of variance, p = 1.00).
|
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Several risk factors for endoleak have been identified [16–19], including type and choice of device, angulation of the neck of the aneurysm, and branch vessel patency in the aneurysmal sac. Among the devices currently approved by the U.S. Food and Drug Administration, the Excluder graft (W. L. Gore) is associated with a higher rate of type II endoleaks [16, 17]. In follow-up of 184 patients who had undergone AAA repair, Albertini et al. [18] found that patients in whom proximal perigraft endoleak developed had a high incidence of angulation of the neck of the aneurysm. Fan et al. [19] found that total endoleak rate increased with the number of patent sac branch vessels in a group of 158 patients who had undergone AAA repair. Nevertheless, onset and outcome of endoleaks in individual AAAs after EVAR are unpredictable [11, 20].
Endoleak occurs frequently after endovascular AAA repair, the reported range being 15–52% of cases [6–9]. However, approximately one half of type II endoleaks seal within the first 6 months [21–23]. The endoleak rate of 34.7% in our group of 144 patients is consistent with previous findings. We found that 62% of endoleaks (31 of 50) completely resolved. The most common endoleak was type II, due to collateral flow through a patent lumbar or inferior mesentery artery. In our study, a larger percentage of endoleaks detected in only a single phase of imaging, in particular on delayed phase CT scans (100% rate of resolution), had resolved than had leaks detected in both phases. We assume that an endoleak detectable in only a single imaging phase may have a smaller volume of leakage and thus be more likely to resolve. Iezzi et al. [15] found that images in the delayed phase of enhancement depicted low-flow endoleaks not seen during the arterial phase. That study, however, did not address the clinical significance of depicting additional low-flow endoleaks. Our study showed that this type of endoleak did not progress, resolved in all cases, and did not necessitate intervention.
Most clinicians would agree that patients with type I and III endoleaks should be treated, because the aneurysm is persistently exposed to central arterial pressure, which has been associated with rupture [4]. There is no consensus on the optimal regimen for type II endoleaks. Although type II endoleak should have a relatively small risk of rupture owing to the altered flow dynamics and lower pressure of retrograde flow, risk exists and may depend on the magnitude of the col lateral vessels. Some investigators [24, 25] have suggested early intervention for type II endoleak to reduce continued risk of rupture with persistent leakage. Other authors [16, 26, 27] recommend a conservative treatment strategy because of spontaneous sealing of type II endoleaks in as many as 50% of cases. In our study, 16 endoleaks, including nine type II endoleaks (detected on both arterial and delayed CT), were managed with add itional interventions after they persisted more than 6 months or were associated with aneu rysmal sac growth of 5 mm or more. The other endoleaks resolved or remained stable without rupture.
It is generally accepted that long-term surveillance of AAA managed with EVAR is needed because late-onset type II endoleaks are as common as early-onset leaks [28]. Periodic follow-up CT is performed because of concerns related to new development of endoleak or interval progression of a previously stable endoleak. Follow-up CT after AAA repair, however, results in substantial radiation exposure of patients. In our patient population, endoleaks detected on delayed CT images alone resolved and did not necessitate intervention. Thus delayed CT acquisition appears to be of less clinical importance than arterial phase acquisition. In our study, all type I and III endoleaks were visualized during both the arterial and delayed phases of CT. Thus it is unlikely that a type I or III endoleak, which warrant prompt treatment, would be missed with arterial phase imaging alone. Furthermore, arterial CT is sufficient and reliable for evaluating changes in AAA size. Even though some endoleaks are missed on arterial phase CT scans, decisions regarding man agement of endoleaks can be based on changes in the size of the aneurysm. On the basis of results of this and previous studies, we propose that patients who have undergone EVAR for AAA undergo follow-up with only arterial phase CT.
To reduce CT radiation dose further, one can consider acquiring only contrast-enhanced CT scans without unenhanced images. This approach, however, is not commonly used because it hampers detection and dif fer entiation of endoleaks (i.e., a focal region of contrast enhancement) from calcified mural thrombus, perigraft hematoma, or, in some cases, the high-attenuation glue material used to repair endoleak. In our study, 125–160 mL of contrast material was used to enhance the aorta and abdominopelvic vessels. The amount of contrast medium can be reduced further with the use of more advanced, faster MDCT.
Our study had limitations. First, it was a retrospective study based on analysis of data from CT reports. In our clinical practice, we use a standardized EVAR measurement report form completed by technologists perform ing 3D image analysis. CT reports were dictated with these measurements and findings. When the standardized CT report findings were unclear, CT images were retrieved and reviewed. Second, because patients were evaluated over a relatively long period, data were incomplete in terms of follow-up time point. Third, type I and III endoleaks in this study group were managed aggressively and thus were not given a chance to resolve. Because the probability of spontaneous resolution of these types of endoleaks is low, we regarded them as un resolved for this study. Although there were only 11 type I and III cases and no definite clinical significance or outcome available in these cases, we were compelled to include these endoleaks in our study because none of them was detected only on delayed phase images, indicating that they are unlikely to be missed if delayed phase images are elim inated. Fourth, the size of the sample of endoleaks detected only in the delayed phase was rather small, eight cases. Larger pro spective studies are needed for validation of our findings.
In our study, endoleaks detected only during the delayed phase of CT resolved spontaneously and did not necessitate intervention. Thus the delayed phase of CT can be eliminated on follow-up CT examinations after EVAR of AAA.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  T. A. Bley, P. J. Chase, S. B. Reeder, C. J. Francois, K. Shinki, G. Tefera, F. N. Ranallo, T. M. Grist, and M. Pozniak Endovascular Abdominal Aortic Aneurysm Repair: Nonenhanced Volumetric CT for Follow-up Radiology, October 1, 2009; 253(1): 253 - 262. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
