European Journal of Cardio-thoracic Surgery 15 (1999) 180–185
Arch and descending aortic aneurysms: influence of perfusion
technique on neurological outcome1
Stephen Westaby*, Takahiro Katsumata, Giuseppe Vaccari
Department of Cardiac Surgery, Oxford Heart Centre,John Radcliffe Hospital,
Headley Way, Headington, Oxford OX3 9DU, UK
Received 5 August 1998; received in revised form 7 December 1998; accepted 16 December 1998
Abstract
Objective: Although cannulation of the femoral artery is used routinely for thoracic aortic operations with hypothermic circulatory arrest,
retrograde perfusion through the descending aorta carries the risk of cerebral malperfusion or embolism. We have, therefore, routinely used
a central cannulation technique for distal arch and descending aortic operations since 1995. In this study, we compared neurological
outcome in consecutive patients undergoing femoral versus ascending aortic perfusion for these aneurysms. Methods: Between 1987 and
1998, 61 patients underwent aortic resection with circulatory arrest, but without retrograde cerebral perfusion, for lesions of the aortic arch
and descending aorta. Thirty-one patients had fusiform true aneurysms, 19 had aortic dissection and 11 had extensive saccular or false
aneurysms. Thirty-two patients (52%) were perfused via the femoral artery (group A), and 29 patients (48%) from the ascending aorta
(group B). Operative mortality and morbidity, and neurological outcome, were reviewed. Results: There were no differences between the
groups in mean age, pathology, abdominal and peripheral vascular disease, net perfusion time, or circulatory arrest time. There were four
hospital deaths (three in group A and one in group B; P = 0.61), including one neurological death in group A. group A suffered a higher
incidence of neurological events (nine patients: 28%) than group B (two patients: 7%; P = 0.03). Temporary focal neurological deficits
occurred in both groups (two patients in group A, 6% and two patients in group B, 7%; P . 0.99), but permanent injury occurred
exclusively in group A (seven patients: four with monoplegia, one with hemiplegia, and two with diffuse cerebral injury with one
death; P = 0.01). Conclusions: Anterograde perfusion using a proximal aortic cannula provides a low risk of cerebral embolism and
allows extensive aortic resection with low morbidity.  1999 Published by Elsevier Science B.V. All rights reserved.
Keywords: Aorta; Arch; Cannulation; Complication; Operation; Technique
1. Introduction
Cerebral injury is a major source of morbidity during
surgery of the aortic arch [1–3]. Computerized tomographic
scans or autopsy in these patients frequently demonstrates
multifocal or hemispheric cerebral infarction presumably
caused by embolism.
We have previously speculated that, reversed aortic perfusion through a dissected or atheromatous aorta is a source
of this problem [4]. We now attempt to avoid femoral arterial perfusion by introducing the arterial cannula as close to
* Corresponding author. Tel.: +44-1865-220-269; fax: +44-1865-220268.
1
Presented at the 12th Annual Meeting of the European Association for
Cardio-thoracic Surgery, Brussels, Belgium, September 20–23, 1998.
1010-7940/99/$ - see front matter
PII: S10 10-7940(98)003 10-8
the aortic arch as possible. If necessary the aneurysm itself
is cannulated using ultrasonic guidance to avoid thrombus
pending hypothermic circulatory arrest [5]. The aim of central cannulation is to prevent scattering of debris by the
blood jet in an atheromatous thoraco-abdominal aorta or
malperfusion through a dissected aorta with partially thrombosed false lumen.
When we described our preferred perfusion method for
one stage resection of the aortic arch and descending thoracic aorta we found little evidence in the literature to support our negative impression of the conventional femoral
approach [4,6]. We, therefore, retrospectively studied aortic
arch operations from our aortic surgery database where
femoral arterial perfusion was employed. These were compared with the patients who had the central cannulation
technique in order to discern differences in outcome. The
 1999 Published by Elsevier Science B.V. All rights reserved.
S. Westaby et al. / European Journal of Cardio-thoracic Surgery 15 (1999) 180–185
findings have particular relevance given the increasing use
of the femoral artery in so-called ‘less invasive’ operations.
2. Materials and methods
Between April 1987 and April 1998, 64 patients underwent one stage resection and repair of aortic arch and descending thoracic aneurysms using hypothermic circulatory
arrest. Our study concentrates on 61 consecutive patients
who had hypothermic cerebral protection alone. In order
to provide an unambiguous comparison of antegrade versus
femoral perfusion in hypothermic circulatory arrest we
excluded those three patients who had retrograde cerebral
perfusion. There were 41 male and 20 female patients. Ages
ranged from 33 to 80 years with a mean age of 64 years.
Only one patient underwent a two staged (elephant trunk)
procedure and the first of which is included here.
Two patients (3%) had Marfan’s syndrome and 39 (64%)
were treated for hypertension. Thirty-six patients (59%) had
symptoms related to their aortic disease. Eighteen had a
hoarse voice, through left recurrent laryngeal nerve palsy.
Nineteen had persistent or sudden onset of backpain, three
had haemoptysis and one presented with orthopnea. The
remaining 25 patients (41%) either had radiologically identified aneurysms with a diameter greater than 6 cm, or evidence of progressive enlargement of the aorta by repeated
CT scans.
Three patients (5%) had undergone a previous primary
cardiac procedure. Four (7%) had undergone previous aortic
operations. These were: repair of acute type A dissection
(n = 2), repair of coarctation of the aorta, and graft replacement of an infrarenal abdominal aortic aneurysm.
The pathology and location of the diseased aortic segments are shown in Table 1. The aetiology of aortic disease
was degenerative or atherosclerotic in 34 patients (56%);
Table 1
Pathology and extent of aortic disease
Non-dissection Total
Proximal DTA Total
With distal arch
With whole arch
Whole DTA
Total
With distal arch
With whole arch
Dissection
Total
Proximal DTA Total
With distal arch
With whole arch
Whole DTA
Total
With distal arch
With whole arch
Group A
no. (%)
Group B
no. (%)
P-value*
23
17
15
2
6
4
2
9
2
1
1
7
3
4
19
13
13
0
6
5
1
10
3
0
3
7
4
3
(66)
(45)
0.59
0.52
(21)
0.85
(34)
(10)
0.59
0.66**
(24)
0.83
(72)
(53)
(19)
(28)
(6)
(22)
DTA, descending thoracic aorta; no., number of patients. *Chi-square test,
**Fisher’s exact test.
181
aortic dissection in 19 (31%); trauma in 7 (11%) and false
aneurysm formation 40 years after homograft conduit repair
of coarctation in one.
In all patients, the proximal two-thirds of the ascending
aorta was within normal size limits or had been replaced. A
fusiform aortic aneurysm involved the innominate artery in
one patient, and the celiac and superior mesenteric arteries
arose from the aneurysm in two patients. Eleven patients
had a saccular aneurysm of which seven gave a history of
deceleration blunt chest trauma, in road traffic accidents.
In 11 of the 19 patients with acute or chronic aortic dissection, the tear was clearly localised to the isthmus (type B)
with both retrograde and anterograde dissection. In three
patients the dissection extended beyond the renal arteries.
In one patient the left renal artery arose from the false
lumen.
2.1. Preoperative investigations
Computerized tomographic scanning was performed preoperatively in all but one patient (with massive haemoptysis) to image the entire thoraco-abdominal aorta. Twelve of
18 patients with chronic aortic dissection had preoperative
aortography to localize the site of the intimal tears. Two
patients with chronic type A dissection involving the entire
aortic arch and proximal descending aorta had preoperative
transoesophageal ultrasound investigation to locate the
proximal end of the false lumen. This information helped
determine the site of aortic cannulation. If there was a past
history of angina or myocardial infarction in elective cases,
coronary angiography was also carried out.
All 61 patients had a general physical examination for
their neurological status before and after operation (within
2 days of extubation). Six patients requiring emergency
operation were taken immediately to the anaesthetic preparation room, and preoperative assessment was as complete as time and conditions would permit. A past history
of cerebrovascular disease or atherosclerotic aortoiliac disease was noted.
2.2. The operation
Emergency surgery was required for a leaking aneurysm
in six patients (10%). Four of these had ruptured type B
dissection (one acute and three chronic), whilst one had a
ruptured atherosclerotic arch aneurysm. The remaining
patient underwent emergency surgery for a chronic traumatic aneurysm which ruptured into the left main bronchus
and presented with massive haemoptysis.
The patients were divided into two groups according to
the site of arterial cannulation at the beginning of cardiopulmonary bypass. Thirty-two patients underwent femoral
arterial perfusion during the cooling phase of cardiopulmonary bypass (group A). The remaining 29 patients had
ascending aortic or arch cannulation (group B). All but
one of these patients (97%) underwent operation after
182
S. Westaby et al. / European Journal of Cardio-thoracic Surgery 15 (1999) 180–185
June 1991 using the central cannulation technique as the
technique of choice for combined arch and descending aortic replacement. The extent of aortic resection for both
groups is summarized in Table 2. Thirty-one patients in
group A (97%) and 26 in group B (90%) had descending
aortic resection with or without arch replacement.
The aorta was approached through a median sternotomy
in four patients (7%), a left lateral thoracotomy in 56
patients (92%), and a sternotomy with left thoracotomy
extension in one patient (2%). The site of incision was
determined by ease of access to the distal end of the diseased aortic segment.
In the patients undergoing a left thoracotomy, a double
lumen endotracheal tube was used to collapse the left lung.
A combination of the neuroprotective agents nimodipine (1
mg/h), thiopental (6 mg/kg loading dose, then 6 mg/kg per
h), and mannitol (5 mg/kg) was administered to all patients
in the anaesthetic preparation room. Three patients who had
had ruptured aneurysms at presentation underwent urgent
femoral cannulation and femoro–femoral perfusion before
opening the chest. In the remaining 58 patients including
three ruptured cases, cardiopulmonary bypass was established after exposure of the diseased aortic segments.
The cannulation and hypothermic perfusion techniques
have been described previously [4]. In median sternotomy
patients, venous return was achieved using a single, twostage, right atrial cannula. For the thoracotomy approach the
patient was placed on the operating table in the left lateral
position. With the pelvis at 90° to the operating table, the
shoulders were tilted approximately 30° from vertical
toward the surgeon providing access to the sternal edge.
The incision was then performed through the fourth intercostal space. Usually it was not necessary to transect the
sternum (particularly with increasing experience). The pericardium was opened 2 cm anterior to the left phrenic nerve
and elevated towards the assistant. A tape was passed
around the aorta to assist cannulation. Arterial return was
then established with a short angled cannula just proximal to
the innominate artery. A right-angled 40 Fr. venous cannula
was inserted into the main pulmonary artery, through the
pulmonary valve and into the right ventricle for venous
return (Fig. 1).
Table 2
Surgical procedures
Procedure
Group A
no. (%)
Group B
no. (%)
P-value*
TAR
PAR
DAR
1
0
31
12
6
13
1
2
26
11
5
10
.0.99**
0.22**
0.34**
0.97
0.88
0.62
Total
Isolated
With TAR
With PAR
(3)
(0)
(97)
(38)
(19)
(41)
(3)
(7)
(90)
(38)
(17)
(34)
DAR, descending thoracic aortic replacement; no., number of patients;
PAR, partial arch replacement; TAR, total arch replacement. *Chi-square
test, **Fisher’s exact test.
Fig. 1. The central cannulation technique through extended left thoracotomy. (Reprinted with permission from the Society of Thoracic Surgeons
[6]).
Cardiopulmonary bypass was established with cooling
of the nasopharyngeal temperature to 20°C. In aortic dissection patients, care was taken to ensure that femoral perfusion cooled the head promptly. When coronary grafts
were required, the distal anastomoses were performed during the cooling period. Circulatory arrest and drainage of
venous blood into the oxygenator were undertaken at 20°C.
The patient was tilted head down during the arch repair.
The aortic arch was opened and for total arch replacement
the head vessels were mobilized collectively on a single
aortic patch. If the patient had a median sternotomy, the
distal aortic anastomosis was performed first. The aortic
arch was replaced by anastomosing the brachiocephalic
vessels to a collagen-coated vascular graft. For rewarming
the arterial cannula was relocated within this graft. The
head and body were then reperfused with a vascular
clamp placed on the graft proximal to the innominate
artery. For left thoracotomy patients, the proximal aortic
anastomosis and reimplantation of the head vessels were
accomplished first. When the cannulated segment of the
aorta was replaced, the arterial cannula was repositioned
within the vascular graft and the brachiocephalic and coronary vessels reperfused. In each case cardiopulmonary
bypass was re-established with antegrade aortic perfusion.
During the initial period of cerebral reperfusion (1 l/min),
the remaining aortic anastomosis was performed. Single
tube grafts were used for the arch and the descending
aorta in all cases. The ‘open-ended’ technique was used
routinely for distal aortic anastomoses. Rewarming was
initiated at the restoration of the pump flow; proximal coronary anastomoses were performed at a suitable site during
rewarming.
183
S. Westaby et al. / European Journal of Cardio-thoracic Surgery 15 (1999) 180–185
2.3. Statistical analysis
Table 4
Neurological complications
Statistical analysis was performed using StatView 4.51
(Abacus Concepts). Continuous data are presented as the
mean ± SD, and categoric data are presented as frequencies.
Continuous variables were compared using Mann–Whitney
U-test. Categoric data were compared using chi-square analysis, with Fisher’s exact test if appropriate.
3. Results
There were no differences between the groups in mean
age (group A 65 years ± 9; group B, 64 ± 11; P = 0.79),
preoperative history of cerebrovascular disease (group A,
three patients, 9%; group B, two patients, 7%; P . 0.99),
atherosclerotic aorto-iliac disease (group A, one patient,
3%; group B, two patients, 7%; P = 0.60), frequency of
emergent operation (group A, three patients, 9%; group B,
three patients, 10%; P . 0.99) and distribution of dissection/non-dissection patients (group A, nine patients, 28%;
group B, ten patients, 34%; P = 0.59).
Time data of extracorporeal circulation are summarised
in Table 3. Twenty-four (75%) group A patients had the
femoral arterial cannula repositioned to the dacron graft to
reperfuse the brain, upper extremities, and the myocardium.
There was no significant difference between the groups in
the periods of net perfusion, cerebral ischaemia and myocardial ischaemia. Lower body circulatory arrest time was
significantly longer in group B as no distal perfusion was
used during the distal aortic anastomosis.
There were four hospital deaths (30-day mortality 7%).
Three of four deaths occurred in elective cases of group A
(9%) and one in emergency cases of group B (3%;
P = 0.61). One patient in group A developed diffuse cerebral and brain stem infarction and died 12 days postoperatively. The other two patients in group A died from
bronchopneumonia and mediastinitis, respectively.
One patient in group B died. This was a 76-year-old
woman who presented with a ruptured atherosclerotic
aneurysm of the aortic arch, critical hypovolaemic shock,
and ventricular fibrillation. She was operated on in a moribund condition with little prospect for survival but after
Table 3
Duration of perfusion, circulatory arrest, and myocardial ischaemia
Group A
(n = 32)
Group B
(n = 29)
P-value*
Net perfusion
(min)
Cerebral
ischaemic
time (min)
Renal
ischaemic
time (min)
Myocardial
ischaemic
time (min)
90 ± 24
26 ± 9
18 ± 6
26 ± 10
94 ± 26
24 ± 10
30 ± 11
24 ± 10
0.61
0.42
,0.0001
0.37
*Mann–Whitney U-test; all figures are shown as the mean ± SD.
Temporary
Total
Monoplegia
Hemiplegia
Permanent
Total
Monoplegia
Hemiplegia
Diffuse injury
Total
Group A
no. (%)
Group B
no. (%)
P-value*
2 (6)
1
1
2 (7)
2
0
.0.99
7 (22)
4
1
2a
9 (28)
0 (0)
0
0
0
2 (7)
0.01
0.03
no., Number of patients. aOne patient died. *Chi-square analysis with
Fisher’s exact test.
total aortic arch replacement, she recovered without any
neurological deficit. She died on postoperative day 5 from
massive gastrointestinal bleeding.
Neurological events are summarised in Table 4. Group A
patients had a higher incidence of neurological events (nine
patients, 28%; P = 0.03) than group B (two patients; 7%).
Two patients in each group had a transient monoparesis (6%
in group A and 7% in group B; P . 0.99). One in group B
presented with rupture and had already been intubated. This
problem was apparent immediately upon recovery from
anaesthesia, but spontaneously receded within 2 weeks following operation. Only one of these patients who had total
arch and descending thoracic aortic replacement had any
abnormality on CT scan.
Permanent cerebral injury occurred only in seven group A
patients (22%; P = 0.01). Of these, one had a focal lesion on
CT scan, where six of them showed multifocal defects. Of
seven patients with permanent cerebral injury, six had an
atherosclerotic (non dissecting) aneurysm. No paraplegia or
temporary spinal cord dysfunction were observed in the
series. Reoperation for bleeding was required in three
patients (two in group A, one in group B; P . 0.99).
Renal failure requiring dialysis occurred in only one patient
in group A. This diabetic patient had undergone repair of
chronic type B dissection with only a short (12 min) renal
ischaemic time.
The mean extubation time was 16.5 ± 5.6 h for group A,
and 13.8 ± 5.1 h for group B (P = 0.47). Mean extubation
time for all survivors was 15.2 ± 5.7 h (median 15 h, range
4.3 to 29 h).
Other complications included respiratory failure in three
(one with tracheostomy), prolonged low output syndrome in
one, non-sustained ventricular tachycardia in one, and local
wound problems in two patients.
4. Discussion
In the elderly population with thoracic aneurysms, arter-
184
S. Westaby et al. / European Journal of Cardio-thoracic Surgery 15 (1999) 180–185
iosclerotic disease is commonly encountered in the aortoiliac and femoral vessels. Complications of femoral artery
cannulation besides embolism include local dissection, distal limb ischaemia, compartment syndrome, peripheral neurologic injury, and local wound complications such as
lymph leak. All can be eliminated by avoiding femoral cannulation [7,8].
The combination of arch and descending thoracic aortic
pathology is technically challenging even with hypothermic
circulatory arrest techniques. Though isolated aortic arch
replacement is straightforward via median sternotomy complex lesions in the distal arch and descending aorta are more
difficult and are often approached by a two stage approach
[9,10]. The use of extended left thoracotomy gives access to
ascending aorta, arch and whole of descending thoracic
aorta. We prefer one stage resection through this incision
to the staged ‘elephant trunk’ method but had concerns
about the risk of femoral perfusion when cerebral injury
occurred, despite the use of retrograde venous perfusion.
There are three principle causes of cerebral injury; prolonged global ischaemia, cerebral embolism and malperfusion through a dissected aorta. Global ischaemic injury can
be minimized by an ischaemic time within the safe duration
of total circulatory arrest or by continuous antegrade cerebral perfusion with blood [11–13]. It is questionable
whether retrograde cerebral perfusion increases the safe
duration of global ischaemia except by its continuous cooling effect [14]. It is in the prevention of cerebral embolism
that aspects of perfusion technique can make an important
difference, since debris may produce ischaemic injury
before hypothermia or retrograde cerebral perfusion are
established.
Our data show that the majority of cerebral events in this
consecutive series occurred in patients who underwent a
period of retrograde aortic perfusion through a diseased
segment of the aorta. Kouchoukos advocates retrograde
venous perfusion and antegrade reperfusion through the
arch graft during rewarming as strategies to avoid embolic
cerebral injury [15]. We would now go one step further and
avoid femoral perfusion altogether. Since 72% of group A
patients had cerebral reperfusion through the vascular graft
and all had a short period of retrograde venous flow to wash
the carotid arteries, ischaemic injury was probably initiated
during the cooling period. The proposed mechanism of
injury is as follows. At the beginning of perfusion the
high pressure jet impacts against the diseased aorta and
liberates atheromatous material or thrombus into the blood
stream. The debris then occludes parts of the cerebral microcirculation causing ischaemia before significant cooling
occurs. Cerebral infarction progresses in affected areas
before retrograde venous perfusion or antegrade reperfusion. This same mechanism may also affect the renal, and
visceral arteries causing other post operative problems in
these high risk patients.
Okita et al. [16] have also suggested that the majority of
intraoperative strokes are caused by embolism and not
through prolonged global ischaemia or inadequacy of cerebral protection. We now consider the site of arterial cannulation when commencing cardiopulmonary bypass to be an
important factor in prevention of cerebral embolic events.
The central cannulation method may also reduce the risk of
paraplegia by early restoration of cerebral blood flow, which
in turn perfuses the anterior spinal artery.
Svensson and Ergin independently have shown that
hypothermic circulatory arrest times greater than 45 min
increase neurological risk, and cerebral ischaemia exceeding 60 min increases both stroke rate and mortality [11,17].
However, none of our patients had circulatory arrest as long
as 45 min. Although there is some evidence to suggest that
retrograde cerebral perfusion can prolong the safe duration
of circulatory arrest our own experience has suggested that
cerebral edema may lead to delayed awakening and worse
neurological outcomes than our current methods [14]. Okita
and colleagues [18] analyzed 148 patients who underwent
arch operations with retrograde cerebral perfusion. They
suggested that risk factors for stroke were ruptured aneurysm and distal arch replacement but not prolonged (greater
than 60 min) circulatory arrest if retrograde cerebral perfusion was employed. However, 25% of their patients developed temporary neurological dysfunction, termed
‘delirium’, associated with extubation times of 38.5 ± 77.6
h even for the patients with less than 60 min of retrograde
cerebral perfusion. Extubation times for those perfused
more than 60 min were up to 77.0 ± 166.3 h. In contrast
our patients operated with deep hypothermia alone required
much shorter periods of intubation (mean 15.2 ± 5.7 h) and
did not experience delayed awakening or delirium. Neither
was there a significant difference in stroke rate between our
patients and Okita’s.
In summary, we suggest that perfusion through a diseased
thoraco-abdominal aorta via a femoral cannula predisposes
to embolic risk. Central cannulation close to the brachiocephalic vessels reduces the risk. Extended left thoracotomy
with proximal aortic and pulmonary arterial cannulation is a
safe alternative for one stage replacement of the arch and
descending thoracic aorta.
References
[1] Crawford ES, Svensson LG, Coselli JS, Safi HJ, Hess KR. Surgical
treatment of aneurysm and/or dissection of the ascending aorta,
transverse aortic arch, and ascending aorta and transverse aortic
arch: factors influencing survival in 717 patients. J Thorac Cardiovasc Surg 1989;98:659–674.
[2] Coselli JS, Buket S, Djukanovic B. Aortic arch operation: current
treatment and results. Ann Thorac Surg 1995;59:19–27.
[3] Griepp RP, Ergin MA, Lansman SL, Galla JD, Pogo G. The physiology of hypothermic circulatory arrest. Semin Thorac Cardiovasc
Surg 1991;3:188–193.
[4] Westaby S, Katsumata T. Proximal aortic perfusion for complex arch
and descending aortic disease. J Thorac Cardiovasc Surg 1998;115:
162–167.
[5] Marshall WG Jr, Barzilai B, Kouchoukos NT, Saffitz J. Intraopera-
S. Westaby et al. / European Journal of Cardio-thoracic Surgery 15 (1999) 180–185
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
tive ultrasonic imaging of the ascending aorta. Ann Thorac Surg
1989;48:339–344.
Katsumata T, Westaby S. Operation for chronic traumatic aortic
aneurysm: when and how. Ann Thorac Surg 1998;66:774–778.
Sabik JF, Lytle BW, McCarthy PM, Cosgrove DM. Axillary artery:
an alternative site of arterial cannulation for patients with extensive
aortic and peripheral vascular disease. J Thorac Cardiovasc Surg
1995;109:885–891.
Bichell DP, Balaguer JM, Aranki SF, Couper GS, Adams DH, Rizzo
RJ, Collins Jr JJ, Cohn LH. Axilloaxillary cardiopulmonary bypass: a
practical alternative to femorofemoral bypass. Ann Thorac Surg
1997;64:702–705.
Borst HG, Walterbusch G, Schaps D. Extensive aortic replacement
using elephant trunk prosthesis. Thorac Cardiovasc Surg 1983;31:
37–40.
Safi HJ, Miller CC III, Iliopoulos DC, Letsou GV, Baldwin JC.
Staged repair of extensive aortic aneurysm: improved neurologic
outcome. Ann Surg 1997;226:599–605.
Svensson LG, Crawford ES, Hess KR, Coselli JS, Raskin S, Shenaq
SA, Safi HJ. Deep hypothermia with circulatory arrest. Determinants
of stroke and early mortality in 656 patients. J Thorac Cardiovasc
Surg 1993;106:19–31.
Ergin MA, Griepp EB, Lansman SL, Galla JD, Levy M, Griepp RB.
Hypothermic circulatory arrest and other methods of cerebral protection during operations on the thoracic aorta. J Card Surg 1994;9:525–
537.
Kazui T, Kimura N, Komatsu S. Surgical treatment of aortic arch
aneurysms using selective cerebral perfusion. Experience with 100
patients. Eur J Cardio-thorac Surg 1995;9:491–495.
Bavaria JE, Pochettino A. Retrograde cerebral perfusion in aortic
arch surgery: efficacy and possible mechanisms of brain protection.
Semin Thorac Cardiovasc Surg 1997;9:222–232.
Kouchoukos NT. Adjuncts to reduce the incidence of embolic brain
injury during operations on the aortic arch. Ann Thorac Surg
1994;57:243–245.
Okita Y, Takamoto S, Ando M, Morota T, Yamaki F, Kawashima Y.
Predictive factors for postoperative cerebral complications in patients
with thoracic aortic aneurysm. Eur J Cardio-thorac Surg
1996;10:826–832.
Ergin MA, Galla JD, Lansman SL, Quintana C, Bodian C, Griepp
RB. Hypothermic circulatory arrest in operations on the thoracic
aorta: determinants of operative mortality and neurological outcome. J Thorac Cardiovasc Surg 1994;107:788–799.
Okita Y, Takamoto S, Ando M, Morota T, Matsukawa R,
Kawashima Y. Mortality and cerebral outcome in patients who
underwent aortic arch operations using deep hypothermic circulatory
arrest with retrograde cerebral perfusion: no relation of early death,
stroke, and delirium to the duration of circulatory arrest. J Thorac
Cardiovasc Surg 1998;115:129–138.
185
particulate matter emboli as opposed to air emboli? A difficulty of operating on the aortic arch through a lateral thoracotomy, in circulatory arrest, is
de-airing the aortic arch and proximal ascending aorta. Is it possible that
by cannulating the ascending aorta you are providing a better de-airing
process as opposed to preventing atheromatous retrograde emboli?
Dr Katsumata: We are not able to specify the direct cause of cerebral
injury in this study as we need the histologic section of the brain in each
case with a cerebral problem. I think, however, the air embolism is highly
unlikely in our series. We paid a great deal of attention on the position of
the patient when we de-air the arch. We routinely tip the patient’s head
down in Trendelenburg position, and also we tip the patient’s right side
down to locate their ascending aorta at the lowest position. When the arch
repair is completed we raise the blood level slowly till all the air is
evacuated from the distal end of the graft. It is a clue to avoid air embolism.
Dr von Oppell: Secondly, are you concerned when cannulating the
ascending aorta in patients with dissections, of causing a problem like
malperfusion?
Dr Katsumata: It is hazardous, obviously, to cannulate the dissected
ascending aorta. We always make sure that the ascending aorta is intact
and safely cannulable by transesophageal ultrasound.
I can see one great benefit from our central cannulation technique with
regard to aortic dissection. Since, as Dr. Laas in Hannover described sometime ago, in the patients with acute type A dissection, if you perfuse the
patient from one of the femoral arteries through the dissected aorta, you
may have so-called ‘blind pocket’ phenomenon where the vital aortic
branches could be occluded at any level of the aorta. It is quite unlikely
to happen with our antegrade aortic perfusion technique.
Discussion by Dimitrios Dougenis, M.D., FECTS,.
Dr D. Dougenis (Patras, Greece): My question concerns retrograde
perfusion of the brain. Obviously, in the left lateral position, this is quite
difficult. But there have been some reports where, by particular manipulation of the cardiopulmonary bypass system, you may be able to perfuse the
brain retrogradely. Have you considered that in order to minimize the
neurological deficits?
Dr Katsumata: In our initial series we put an extra cannula at the angle
between the innominate and the right common carotid vein. We have
recently integrated the technique originally developed by Dr. Takamoto,
in the University of Tokyo, into our central cannulation technique, where
the retrograde cerebral perfusion is easily achieved by raising central
venous pressure in Trendelenburg position by perfusing the lower torso
through the arterial cannula relocated within the distal aortic stump in the
left chest. The perfusate is venous, but highly oxygenated, so we use that
method as a means to avoid embolism but not for the prolonged cerebral
protection.
Dr V. Velebit (Geneva, Switzerland): Have you considered cannulating
the axillary artery?
Dr Katsumata: No.
Dr Velebit: Why not?
Dr Katsumata: It’s cumbersome.
Appendix A . Conference discussion
Dr U. von Oppell (Cape Town, South Africa): Have you got any evidence that the neurological complications were due to atheromatous or
Dr Velebit: It’s very elegant.
Dr Katsumata: Since we prefer the single incision of the body; we hate
the two incisions. The other thing is, now we know that the ascending aorta
is always accessible through left thoracotomy.