DOI:10.2214/AJR.05.1654
AJR 2007; 188:1273-1277
© American Roentgen Ray Society
CT Appearance of Thoracic Aortic Graft Complications
Baskaran Sundaram1,
Leslie Eisenbud Quint1,
Smita Patel1,
Himanshu J. Patel2 and
G. Michael Deeb2
1 Department of Radiology, University of Michigan Health System, 1500 E Medical
Center Dr., Box 0302, Ann Arbor, MI 48109-0302.
2 Division of Cardiothoracic Surgery, Department of Surgery, University of
Michigan Health System, Ann Arbor, MI.
Received September 21, 2005;
accepted after revision January 24, 2006.
Address correspondence to B. Sundaram
(sundbask{at}umich.edu).
Abstract
OBJECTIVE. The purpose of this study was to document the spectrum of
CT findings and the clinical outcome of thoracic aortic graft
complications.
CONCLUSION. Aortic graft complications detected with CT may or may
not be clinically apparent and/or relevant. CT characterization in combination
with clinical findings helps to determine patient treatment.
Keywords: aorta • cardiovascular imaging • CT • CT angiography • postoperative complications • thoracic
Introduction
Aortic interposition grafting is widely accepted in the management
of thoracic aortic aneurysm and dissection. Reported complications include
anastomotic dehiscence and graft infection
[1,
2]. Follow-up CT scans are
generally obtained on a routine basis to evaluate for potential complications
that may or may not be clinically evident. At our institution, these routine
scans are obtained approximately 3 and 12 months after surgery and annually
after that. The aim of our study was to document the spectrum of CT findings
and the clinical outcome among patients with thoracic aortic graft
complications detected with CT.
Materials and Methods
Institutional review board approval was obtained for this study. Patients
were retrospectively identified through a search of the institutional
radiology database for chest CT reports generated between 1998 and 2004 that
suggested the presence of a postoperative aortic graft complication on the
basis of imaging findings. The keywords used for the computerized search were
graft or surgery plus leak or extravasation or dehiscence or outpouching or
rupture. The imaging studies were retrieved and reviewed on a workstation by
two experienced chest radiologists, and a consensus interpretation was
generated.
Patients were included in the study if review of the CT examinations showed
any of the following findings: extravasation of contrast material, an
unusually large amount or increasing amount of low-attenuation material
adjacent to the graft, perigraft gas bubbles, and graft herniation into the
chest wall. Patients were excluded if CT showed a small amount of perigraft
low-attenuation material (< 10 mm thick) consistent with normal
postoperative hematoma. CT scans were performed with an IV bolus of contrast
material and thin (1.25- to 3-mm collimation) overlapped sections on a variety
of helical scanners. Most CT examinations were performed for routine
postsurgical follow-up, although a few were performed for evaluation of a
clinically suspected complication.
Axial images from the CT examinations were analyzed for the location of
perigraft low-attenuation collections and contrast collections and for change
in size of these collections if several examinations had been performed on an
individual patient. Scans also were evaluated for the abnormal presence of gas
adjacent to the graft, for evidence of fistula formation with adjacent
structures, and for abnormal graft position (e.g., chest wall herniation).
Clinical patient records were reviewed to obtain information regarding type of
surgery performed (Figs. 1A,
1B,
2A,
2B,
2C and
3), time interval between
surgery and imaging, and treatment and outcome after the CT examination.
View larger version (90K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A —Total aortic root technique (modified Bentall). Drawing shows
complete resection of native aortic root. Prosthetic valved conduit is sutured
to aortic valve annulus in proximal aspect. Native coronary arteries are
reimplanted as buttons (arrows) in prosthesis.
|
|
View larger version (74K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2A —Inclusion root technique. Drawing shows bioprosthetic valved conduit
within native aortic root. Graft is anchored in proximal aspect with series of
interrupted sutures to horizontal plane at nadir of aortic annulus (curved
arrows). Buttons are resected from valved conduit for subsequent native
coronary artery reimplantation (straight arrows). Asterisks indicate
space where small perigraft contrast pools occur in some patients. Dashed line
indicates aortic valve plane.
|
|
View larger version (88K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B —Inclusion root technique. Drawing shows running suture line
attaching distal aspect of valved conduit to aorta. Straight arrow indicates
site of coronary artery reimplantation; curved arrows, proximal graft
anastomosis.
|
|
View larger version (87K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2C —Inclusion root technique. Drawing shows completed repair. Unlike
total aortic root technique in Figure
1A,
1B, inclusion root technique
results in placement of valved conduit within native aortic root. Straight
arrow indicates site of coronary artery reimplantation; curved arrows,
proximal graft anastomosis.
|
|
View larger version (42K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3 —Drawing shows total aortic arch replacement with separate individual
great vessel anastomoses. Thick straight arrows indicate proximal and distal
aortic anastomoses. Anastomoses between graft side branches and native arch
vessels (thin straight arrows), graft cannulation site
(arrowhead), and sewn site of unused graft side branch site
(curved arrow) also are shown.
|
|
Results
Thirty-nine episodes of graft complications
(Table 1) were identified in 34
patients. Five patients had more than one separate episode of
complication.
Abnormal Perigraft Low-Attenuation Material
Twenty of 39 complications consisted of abnormal distribution of
low-attenuation material around the aortic graft (Figs.
4A,
4B,
4C and
5). The abnormal
low-attenuation material was adjacent to a graft in the aortic root in five
patients, ascending aorta in five patients, aortic arch in two patients, and
descending aorta in eight patients. These collections were detected 1 week-30
months after surgery. In one of these patients, the descending aortic graft
also had herniated into the posterolateral chest wall.
View larger version (82K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4A —56-year-old man 1 month after descending thoracic aortic graft
replacement for posttraumatic contained aortic rupture. Routine CT scan shows
minimal perigraft low-attenuation material (arrows). Arrowhead
indicates felt pledget at graft cannulation site.
|
|
View larger version (92K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4B —56-year-old man 1 month after descending thoracic aortic graft
replacement for posttraumatic contained aortic rupture. Axial (B) and
sagittal reformatted (C) CT scans 12 months after A show marked
increase in amount of perigraft low-attenuation material (arrows).
Patient, who had no symptoms and no clinical evidence of infection, was not
treated for this imaging finding. Arrowhead indicates felt pledget at graft
cannulation site.
|
|
View larger version (139K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4C —56-year-old man 1 month after descending thoracic aortic graft
replacement for posttraumatic contained aortic rupture. Axial (B) and
sagittal reformatted (C) CT scans 12 months after A show marked
increase in amount of perigraft low-attenuation material (arrows).
Patient, who had no symptoms and no clinical evidence of infection, was not
treated for this imaging finding. Arrowhead indicates felt pledget at graft
cannulation site.
|
|
View larger version (135K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5 —80-year-old woman 15 months after ascending aortic graft replacement
for aortic aneurysm. CT scan obtained because patient felt chest wall
heaviness reveals abnormal perigraft low-attenuation material
(arrows). Surgical drainage revealed purulent fluid.
|
|
Only one of 20 patients was receiving anticoagulation therapy. Eight of 20
patients had proven (three patients) or presumed (five patients) graft
infections. Two of these patients had gas bubbles in the mediastinum. Four
patients with infection underwent surgical exploration with drainage of a
mediastinal abscess. All four of the patients with infections who did not
undergo surgery were treated with antibiotics. Follow-up CT of three of these
patients showed a decrease in the amount of abnormal low-attenuation material.
The fourth patient did not undergo follow-up imaging.
Twelve of 20 patients had no clinical evidence of graft infection. One
patient underwent surgical exploration that revealed a large hematoma,
probably caused by anastomotic dehiscence. Another patient experienced acute
hypotension and died, presumably of anastomotic dehiscence and exsanguination.
The other 10 patients had no symptoms and were not treated for the imaging
finding initially detected 4-13 months after surgery. Follow-up CT
examinations of four of these patients 3-32 months later showed partial
resolution (two patients), stability (one patient), and progression then
regression (one patient) of the finding. During the surgical procedures on
seven of the 10 patients, fibrin glue had been applied at the graft
anastomoses, and seven of the 10 patients had bovine pericardium covering a
graft.
Pockets of Contrast Material Outside Inclusion Root Grafts
Six of 39 imaging complications consisted of small collections of contrast
material outside the aortic lumen, between the porcine inclusion root graft
and the surrounding native aortic wrap without extension into the mediastinum
(Fig. 6A,
6B,
6C). Three of six collections
appeared to arise from dehiscence at the right coronary artery button. In the
other three patients, the origin of the collection could not be determined.
These collections were detected 1-78 months after surgery, and all were
asymptomatic. All patients underwent echocardiography, and none of the
collections was detected with that technique. In three of the six patients,
follow-up CT examinations 8-51 months later showed stable findings. Two
additional patients did not undergo imaging follow-up. Three of these five
patients were taking aspirin, and none was treated for this imaging finding.
The sixth patient was taking aspirin. After detection of the abnormal contrast
collection on imaging, aspirin was suspended, and the imaging finding was not
present on CT 4 months later. Aspirin therapy was resumed, and follow-up CT 6
months later showed no recurrence of abnormal contrast collection.
View larger version (117K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6A —60-year-old man 1 month after porcine aortic root replacement for
type 1 aortic dissection. Routine follow-up axial (A and B) and
coronal reformatted (C) CT scans show small pockets of contrast
material (arrows) in space between inclusion root graft and
surrounding native aortic wrap, apparently arising from right coronary artery
button anastomosis. Appearance did not change on CT scans obtained over 5-year
period, and condition remained asymptomatic. Arrowhead (B) indicates
right coronary artery.
|
|
View larger version (125K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6B —60-year-old man 1 month after porcine aortic root replacement for
type 1 aortic dissection. Routine follow-up axial (A and B) and
coronal reformatted (C) CT scans show small pockets of contrast
material (arrows) in space between inclusion root graft and
surrounding native aortic wrap, apparently arising from right coronary artery
button anastomosis. Appearance did not change on CT scans obtained over 5-year
period, and condition remained asymptomatic. Arrowhead (B) indicates
right coronary artery.
|
|
View larger version (133K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6C —60-year-old man 1 month after porcine aortic root replacement for
type 1 aortic dissection. Routine follow-up axial (A and B) and
coronal reformatted (C) CT scans show small pockets of contrast
material (arrows) in space between inclusion root graft and
surrounding native aortic wrap, apparently arising from right coronary artery
button anastomosis. Appearance did not change on CT scans obtained over 5-year
period, and condition remained asymptomatic. Arrowhead (B) indicates
right coronary artery.
|
|
Mediastinal Contrast Extravasation
Thirteen of 39 complications consisted of mediastinal collections of
extravasated contrast material outside the aorta (Figs.
7A,
7B,
7C and
8A,
8B). A graft or patch was
present in the aortic root in six patients, in the ascending aorta in four
patients, in both root and ascending aorta in one patient, in the arch only in
one patient, and in both ascending aorta and arch in one patient.
Extravasation appeared to arise from a proximal or distal graft anastomosis in
nine cases, from a coronary artery button anastomosis in three cases, and from
a graft cannulation site in one case. These collections were detected 1
month-18 years after surgery.
View larger version (117K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 7A —47-year-old man taking anticoagulation therapy 27 months after
aortic arch replacement for chronic type A aortic dissection. Axial (A
and B) and sagittal reformatted (C) CT scans obtained because of
clinical suspicion of mediastinal hematoma and infection show active contrast
extravasation (white arrows) arising from cannulation site in graft
and extending into sternotomy defect in anterior chest wall. Leak and
infection at cannulation site were confirmed at subsequent surgical repair.
Straight black arrows indicate distal graft anastomosis; arrowhead (A),
felt pledget at cannulation site; curved arrows (A and B),
dissection flap in native aorta.
|
|
View larger version (113K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 7B —47-year-old man taking anticoagulation therapy 27 months after
aortic arch replacement for chronic type A aortic dissection. Axial (A
and B) and sagittal reformatted (C) CT scans obtained because of
clinical suspicion of mediastinal hematoma and infection show active contrast
extravasation (white arrows) arising from cannulation site in graft
and extending into sternotomy defect in anterior chest wall. Leak and
infection at cannulation site were confirmed at subsequent surgical repair.
Straight black arrows indicate distal graft anastomosis; arrowhead (A),
felt pledget at cannulation site; curved arrows (A and B),
dissection flap in native aorta.
|
|
View larger version (134K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 7C —47-year-old man taking anticoagulation therapy 27 months after
aortic arch replacement for chronic type A aortic dissection. Axial (A
and B) and sagittal reformatted (C) CT scans obtained because of
clinical suspicion of mediastinal hematoma and infection show active contrast
extravasation (white arrows) arising from cannulation site in graft
and extending into sternotomy defect in anterior chest wall. Leak and
infection at cannulation site were confirmed at subsequent surgical repair.
Straight black arrows indicate distal graft anastomosis; arrowhead (A),
felt pledget at cannulation site; curved arrows (A and B),
dissection flap in native aorta.
|
|
View larger version (104K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 8A —76-year-old woman with unremitting sternotomy wound infection and
mediastinitis 12 months after ascending aorta and aortic arch replacement for
aortic aneurysm. CT scan shows small pockets of contrast extravasation
(curved black arrow) at graft anastomosis (arrowhead)
consistent with dehiscence. Finding was confirmed at surgery.
|
|
View larger version (108K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 8B —76-year-old woman with unremitting sternotomy wound infection and
mediastinitis 12 months after ascending aorta and aortic arch replacement for
aortic aneurysm. CT scan shows tethering (curved arrow) of
midthoracic esophagus and mediastinal gas (straight arrow),
suggesting aortoesophageal fistula. Findings were confirmed at surgery.
Arrowheads indicate graft anastomosis
|
|
Two of 13 patients had gas bubbles in the mediastinum. One case of bubbles
was due to an aortoesophageal fistula (Fig.
8A,
8B), and the other was likely
due to infection. One patient without mediastinal gas also had an
aortopulmonary fistula. In seven of 13 patients with contrast extravasation,
the cause appeared to be graft infection (four proven cases and three
presumed). Eight of 13 patients were undergoing anticoagulation therapy, and
one of 13 was taking aspirin.
Six of 13 patients underwent follow-up CT examinations 7-20 months after
the initial CT. In two patients, the collection of contrast material resolved,
one after antibiotic therapy alone, and the other after antibiotic therapy and
deployment of embolization coils within the pseudoaneurysm. In the third
patient, the contrast collection first decreased and then remained stable
after discontinuation of anticoagulation therapy. In the fourth patient, the
extravasation waxed and waned on several subsequent examinations after
antibiotic treatment and discontinuation of anticoagulation. Operative
exploration eventually revealed infection and leak at the graft cannulation
site. In the fifth patient, who had no symptoms, the small collection of
extravasated contrast material increased minimally, and no treatment was
administered. The sixth patient was treated with antibiotics, and the leak was
stable on follow-up CT examinations.
Seven of 13 patients did not undergo follow-up CT examinations. Two of
these patients had proven and two had presumed graft infections causing the
leak, and three had no clinical evidence of infection. Of the four patients
with infection, three underwent surgical repair, and one died of
exsanguination. Of the three patients without infection, one underwent
surgical repair, one deferred treatment, and one transferred to another
hospital.
Discussion
Approximately 250 thoracic aortic graft replacement operations are
performed each year at our institution. Only 39 CT-evident complications were
identified over a 7-year period, suggesting a very low complication rate. The
most frequent type of complication was abnormal accumulation of
low-attenuation material around the graft, and the next most frequent was
collections of contrast material outside the graft.
Accumulation of low-attenuation material around the graft often was due to
infection, although bleeding due to anastomotic dehiscence without infection
appeared to be a less frequent cause. In one half of patients with
accumulation of low-attenuation material, the cause was not identified, and
the patients continued to have no symptoms. We hypothesize that these
perigraft low-attenuation collections might have been the result of
postoperative seroma and/or inflammatory edema developing as a result of an
allergic reaction to the aortic graft material or the surrounding bovine
pericardium [3,
4]. Yamamoto et al.
[3] noted that
collagen-impregnated vascular grafts may have contaminants containing
endotoxin and (1-3)b-D-glucan and causing a sterile inflammatory response
around the graft. A patient at our institution (not included in this study)
who did not have symptoms underwent CT-guided aspiration of such a perigraft
fluid collection that had increased in size over several months. The aspirated
fluid was sterile, clear, and yellow, and laboratory values were consistent
with seroma.
In a specific subset of patients with extravasation of contrast material
outside the graft, the contrast collections appeared to be confined between
the porcine aortic root graft and the surrounding native aortic wrap (i.e.,
inclusion root grafts). This imaging complication did not appear to have
clinical significance because the patients continued to be free of symptoms.
Subsequent imaging examinations showed that the collections had not increased
and that in some patients they had resolved. Using echocardiography, Oechslin
et al. [5] found that a
perfused echo-free space (pseudoaneurysm) between an aortic root homograft and
the native aortic wall was a common finding (22 of 30 patients). Willems et
al. [6] reported similar
echocardiographic findings in 15% (12/79) of patients who had undergone
subcoronary implantation of an aortic valve homograft. Those authors, however,
did not see this finding in patients who had received aortic root homografts.
Oechslin et al. suggested that the leaks were due to partial dehiscence at the
proximal suture line. We believe that these small contrast pools may be the
result of tiny leaks at the coronary artery or proximal graft anastomoses.
In our study, the leaking contrast material in three of six patients
appeared to come from dehiscence in the region of the right coronary artery
button. The origin of the leak in the other three patients was not obvious.
Rofsky et al. [7] found leakage
between an ascending aortic graft and the surrounding native aortic wrap on
MRI in 15% (5/34) of patients and on CT in 17% (4/24). Those grafts were
synthetic rather than tissue grafts, and three of the five patients with leaks
had grafts limited to the supravalvular region of the aorta. Therefore, the
findings are probably not comparable with those in our study.
Whereas pseudoaneurysms contained within a surrounding native aortic wall
in patients with inclusion root grafts appeared to be clinically
insignificant, extravasation of contrast material outside the aorta into the
mediastinum in patients with any type of graft was most often associated with
anastomotic infection (eight of 13 patients). Anticoagulation, however,
appeared to be a factor in a minority of cases. Surgical or angiographic
intervention was performed on most patients with this finding.
In our study, CT findings suggestive of complications occurred at any time
up to 18 years after surgery. A previous study
[8] also showed late occurrence
of graft dehiscence. This finding suggests that imaging surveillance should be
performed on a continuing basis for patients with aortic grafts. A potential
limitation of our study was the lack of standardization of imaging follow-up
of these complications. The follow-up evaluations were individually dictated
by each patient's complex and diverse clinical status and course.
Our data suggest that thoracic aortic graft complications are uncommonly
detected at CT. It is essential, however, to be aware of the spectrum of CT
findings of these complications so that life-threatening conditions can be
diagnosed when they do occur. Apparent complications on CT are of little or no
clinical significance, and knowledge of the surgical details and the patient's
clinical status is important for accurate interpretation of imaging
findings.
References
- Dossche KM, Tan ME, Schepens MA, Morshuis WJ, de la Riviere AB.
Twenty-four year experience with reoperations after ascending aortic or aortic
root replacement. Eur J Cardiothorac Surg1999; 16:607
-612[Abstract/Free Full Text]
- Coselli JS, Koksoy C, LeMaire SA. Management of thoracic aortic
graft infections. Ann Thorac Surg 1999;67
: 1990-1993[Abstract/Free Full Text]
- Yamamoto K, Noishiki Y, Mo M, Kondo J, Matsumoto A. Unusual
inflammatory responses around a collagen-impregnated vascular prosthesis.
Artif Organs 1993;17
: 1010-1016[Medline]
- Moore MA, Phillips RE. Biocompatibility and immunologic properties
of pericardial tissue stabilized by dye-mediated photo oxidation. J
Heart Valve Dis 1997; 6:307
-315[Medline]
- Oechslin E, Carrel T, Ritter M, et al. Pseudoaneurysm following
aortic homograft: clinical implications? Br Heart J1995; 74:645
-649[Abstract/Free Full Text]
- Willems TP, Van Herwerden LA, Taams MA, Kley-burg-Linker VE,
Roelandt JR, Bos E. Aortic allograft implantation techniques: pathomorphology
and regurgitant jet patterns by Doppler echocardiographic studies.
Ann Thorac Surg 1998;66
: 412-416[Abstract/Free Full Text]
- Rofsky NM, Weinreb JC, Grossi EA, et al. Aortic aneurysm and
dissection: normal MR imaging and CT findings after surgical repair with the
continuous-suture graft-inclusion technique. Radiology1993; 186:195
-201[Abstract/Free Full Text]
- Henriques JP, Brutel de la Riviere A, Schepens MA, Ernst JM.
Percutaneous occlusion of the entry to a leaking false aneurysm after
ascending aortic replacement for aortic dissection type A facilitating
surgical repair. Eur J Cardiothorac Surg1997; 11:381
-383[Abstract]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
|
|
 
J. K. Hoang, S. Martinez, and L. M. Hurwitz
MDCT Angiography After Open Thoracic Aortic Surgery: Pearls and Pitfalls
Am. J. Roentgenol.,
January 1, 2009;
192(1):
W20 - W27.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
B. Sundaram, L. E. Quint, H. J. Patel, and G. M. Deeb
CT Findings Following Thoracic Aortic Surgery
RadioGraphics,
November 1, 2007;
27(6):
1583 - 1594.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
