Eur J Cardiothorac Surg 2001;20:527-532
© 2001 Elsevier Science NL
Doppler ultrasonographic identification of the critical segmental artery for spinal cord protection
Ko Shibataa,
Shinichi Takamotoa,
Yutaka Kotsukaa,
Takeshi Miyairia,
Tetsuro Morotaa,
Katsuhito Uenoa,
Hajime Satob
a Department of Cardiothoracic Surgery, University of Tokyo, 7-3-1 Hongo Bunkyo-ku, Tokyo 113-8655, Japan
b Department of Public Health and Occupational Medicine, University of Tokyo, 7-3-1 Hongo Bunkyo-ku, Tokyo 113-8655, Japan
Received 23 January 2001;
received in revised form 15 May 2001;
accepted 23 May 2001.
Corresponding author: Tel.: +81-3-5800-8654; fax: +81-3-5684-3989
e-mail: shibata-tho{at}h.u-tokyo.ac.jp
 |
Abstract
|
|---|
Objective: The purpose of this study is to evaluate the possibility of identifying critical segmental arteries (CSAs) based on Doppler ultrasonographic hemodynamics. Methods: In 18 mongrel dogs, the descending aorta was scanned directly with a 5-MHz linear probe through left thoracotomies and the flow velocities in segmental arteries were measured by pulsed Doppler. The aorta was cross-clamped between Th13 and L1, and flow velocity changes were recorded. According to flow increases, segmental arteries were divided into three groups: arteries with the largest flow increase (L-arteries), arteries with the smallest increase (S-arteries) and other arteries (O-arteries). Animals were divided into three groups. One aortic segment including an L-artery or an S-artery was perfused via a temporary shunt during 30-min aortic cross-clamping distal to the left subclavian artery (Group L or Group S) and neurological outcomes were compared with those of animals without shunting (Group N) after 24 and 48 h. Results: L-arteries had significantly larger flow increases than S- and O-arteries (74.3±33.8, 20.4±9.8 and 33.3±17.8 cm/s, P<0.01). In Group N, five of the six animals were completely paraplegic (Tarlov Grade 0) and the other was Grade 1. In Group S, four animals were Grade 4 and two were Grade 0 after 24 h. However, two animals showed delayed paraplegia. Therefore, four animals were Grade 0 and two were Grade 4 after 48 h. All animals in Group L were neurologically normal (Grade 4) at both after 24 h (vs. Group N, P=0.0013) and 48 h (vs. Group N, P=0.0013; vs. Group S, P=0.019). Conclusions: Flow responses to aortic cross-clamping differed among segmental arteries and selective perfusion of L-arteries completely prevented paraplegia. Therefore, L-arteries were considered to be CSAs. Hemodynamic measurement of segmental arterial flow using Doppler ultrasonography could be clinically useful for spinal cord protection during thoracoabdominal aortic surgery.
Key Words: Doppler ultrasonography • Aortic aneurysm • Paraplegia
|
1. Introduction
|
|---|
Paraplegia remains one of the most serious complications after thoracoabdominal aortic aneurysm repair. Spinal cord ischemia during the operation is thought to be the most important contributor to the development of this neurological deficit. Although much effort has been focused on preventing this dreaded complication, incidences are reportedly as high as 17.4% even in the most experienced institutes [1–4].
To minimize the degree and duration of spinal cord ischemia, various methods have been applied experimentally and clinically. These include distal aortic perfusion [1,3], cerebrospinal fluid (CSF) drainage [3,4], hypothermia [5,6], and pharmacological therapies. However, in order to prevent postoperative paraplegia, it is important to restore the spinal cord blood supply by revascularization of critical segmental arteries (CSAs) [7]. Various modalities are employed to identify CSAs, including preoperative selective angiography [8] and intraoperative spinal cord function monitoring [2,9,10,11], but none are widely used.
Doppler ultrasonography has been used to evaluate various cardiovascular pathologies. With improvement in its performance, indications have been expanded to new areas such as the assessment of coronary artery bypass grafts [12,13].
We hypothesized that CSAs have a larger flow velocity (i.e. flow volume) because they perfuse the spinal cord as well as the thoracic or abdominal wall. We thus evaluated the possibility of identifying CSAs by Doppler hemodynamic measurement of the flow in these vessels.
|
2. Materials and methods
|
|---|
All animals received humane care in compliance with the ‘Guide for the Care and Use of Laboratory Animals’ published by the National Institutes of Health (NIH publication 85–32, revised 1985).
2.1. Flow velocity measurement of segmental arteries by Doppler ultrasonography
Eighteen adult mongrel dogs weighing between 15 and 20 kg were premedicated with an intramuscular injection of ketamine (10 mg/kg), then intubated and ventilated throughout the procedure. Anesthesia was maintained with intermittent bolus injection of pentobarbital. Eighth and 11th left thoracotomies were performed to access the entire thoracic descending aorta.
All Doppler measurements were performed using a computerized system (LOGIQ500, GE Yokokawa Medical Systems, Japan) equipped with a 5 MHz linear probe for epiaortic scanning. The thoracic descending aorta was scanned directly and the origins of segmental arteries were visualized. Pulsed Doppler measurements of the flow velocity of segmental arteries were performed by placing a sample volume at their origins (Fig. 1
).
View larger version (106K):
[in this window]
[in a new window]
|
Fig. 1. Segmental arteries are clearly identified with a 5-MHz epiaortic scanning probe. The flow velocities of segmental arteries are measured with pulsed Doppler ultrasonography.
|
|
After evaluating segmental arteries of the lower thoracic descending aorta between Th7 and Th13, 100 U/kg heparin was administered and the thoracic descending aorta was cross-clamped between Th13 and L1 in order to induce ischemia in the lumbar spinal cord, and pulsed Doppler measurements of the segmental arteries were performed again (Fig. 2
). The flow velocity increase was calculated for each segmental artery. According to this increase, segmental arteries were divided into three groups:- L-artery, artery with the largest flow increase after aortic cross-clamping;
- S-artery, artery with the smallest flow increase after aortic cross-clamping;
- O-artery, other segmental arteries.
View larger version (35K):
[in this window]
[in a new window]
|
Fig. 2. The thoracic descending aorta was cross-clamped between Th13 and L1, and pulsed Doppler measurements of segmental arteries were performed again.
|
|
2.2. Selective perfusion of a single aortic segment with a temporary shunt
After pulsed Doppler measurement, animals were divided into three groups. In the first group (Group L, N=6), a temporary shunt was inserted into the left common carotid artery through a purse string suture. The distal limb of the shunt was inserted into the descending aorta, including an L-artery. Then, the descending aorta was cross-clamped just distal to the left subclavian artery for 30 min. The shunted segment was also isolated with two clamps and perfused via the shunt during cross-clamping (Fig. 3
). In the second group (Group S, N=6), one segment that included an S-artery was perfused via the temporary shunt. The other procedures were same as in the group L. In the third group (Group N, N=6), the descending aorta was cross-clamped just distal to the left subclavian artery for 30 min and no shunt was placed. In all the groups, proximal and distal blood pressures were measured with arterial lines in the right brachial artery and right femoral artery. No attempt was made to control proximal hypertension during aortic cross-clamping. After declamping, the wound was closed in anatomical fashion and the animals were allowed to recover.
View larger version (24K):
[in this window]
[in a new window]
|
Fig. 3. One aortic segment, including an L-artery, was selectively perfused via a temporary shunt during 30-min aortic cross-clamping just distal to the left subclavian artery.
|
|
Neurological outcome was evaluated 24 and 48 h after the procedure according to the modified Tarlov classification (Tarlov 0=no movement of hind limbs; 1=perceptible movement of the joints of the hind limbs; 2=good movement but unable to stand; 3=able to stand and walk; 4=complete recovery).
2.3. Statistical analysis
Each value was expressed as the mean±standard deviation. Statistical analysis for comparisons of the continuous variables was performed using analysis of variance (ANOVA) with post hoc pairwise comparisons. The differences of Tarlov scores were evaluated by Kruskal–Wallis test and Mann–Whitney U-test. A P value of <0.05 was considered statistically significant.
|
3. Results
|
|---|
3.1. Flow velocity measurement of segmental arteries by Doppler ultrasonography
Doppler measurement of segmental arteries was possible in all cases. Flow velocity increase was measured within a 10-min period of aortic cross-clamping between Th13 and L1. L-arteries were located at Th13 (N=12), Th12 (N=5), and Th9 (N=1). S-arteries were located at Th9 (N=9), Th8 (N=4), and Th7 (N=5). The mean flow velocities of segmental arteries before and during aortic cross-clamping were 12.1±7.4 and 37.3±25.3 cm/s, respectively. Before cross-clamping, the mean flow velocities of L-arteries, S-arteries, and O-arteries were 9.9±4.0, 11.8±4.9, and 12.6±8.2 cm/s, respectively, and the difference was not significant (P=0.43). During cross-clamping, the respective flows increased to 74.3±33.8, 20.4±9.8, and 33.3±17.8 cm/s, and the differences between the L- and S-arteries, and between the L- and O-arteries were significant (P<0.01) (Fig. 4
).
View larger version (14K):
[in this window]
[in a new window]
|
Fig. 4. The mean flow velocities of segmental arteries before and during aortic cross-clamping. During cross-clamping, the differences between L- and S-arteries, and between L- and O-arteries were statistically significant (*P<0.01).
|
|
3.2. Selective perfusion of a single aortic segment with a temporary shunt
3.2.1. Blood pressure measurement
The proximal pressure increased to 138±20% of the pre-cross-clamp value during aortic cross-clamping. The distal pressure during cross-clamping was 12±3% of the pre-cross-clamp value. Five minutes after declamping, the blood pressure decreased to 79±17% of the pre-cross-clamp value, then gradually recovered to 87±10% in 30 min. There were no significant differences among the three groups at any time point (Fig. 5
).
View larger version (19K):
[in this window]
[in a new window]
|
Fig. 5. Blood pressure changes during 30-min aortic cross-clamping and after declamping, expressed as the ratio to the pre-cross-clamp value.
|
|
3.2.2. Neurological outcomes
In Group N, five of the six animals were completely paraplegic (Tarlov Grade 0) and the other was Grade 1 at both 24 and 48 h. In Group S, four animals were Grade 4 and two were Grade 0 at 24 h. However, two animals showed delayed paraplegia. Therefore, four animals were Grade 0 and two were Grade 4 at 48 h. On the other hand, all Group L animals were neurologically normal (Tarlov Grade 4) at both 24 and 48 h. This outcome was significantly better than that of Group N (P=0.0013, both 24 and 48 h) and Group S (P=0.019, 48 h) (Table 1). Although Group S showed significantly better outcome than Group N at 24 h (P=0.045), the difference disappeared at 48 h.
|
4. Discussion
|
|---|
To prevent paraplegia after thoracoabdominal aneurysm repair, the ideal approach is to reattach as many patent intercostal or lumbar arteries as possible. However, in clinical situations, this approach is sometimes time-consuming and carries an increased risk of uncontrollable bleeding due to aortic wall fragility. To resolve this dilemma, many surgeons reattach the intercostal and lumbar arteries at the critical zone, usually between the lower thoracic and upper abdominal aorta. Safi reported a 14% incidence of neurological deficit in type 2 thoracoabdominal aortic aneurysm repairs even when as many intercostal arteries as possible between Th7 and Th12 were reattached [14]. The extreme anatomical divergence of spinal cord blood supply in humans makes this approach unreliable. Thus, there have been some reports focused on the identification of ‘critical’ segmental arteries either from an anatomical or a functional viewpoint. The anatomical approach includes preoperative selective angiography as a means of identifying the CSAs and sites of subsequent reattachment [8]. However, the possibility of identifying the CSAs is relatively low, and neurological deficit cannot always be prevented. The examination itself is time-consuming and carries a risk of complications, including paraparesis or paraplegia. Furthermore, in the presence of atherosclerotic plaques, mural thrombus, or dissection, segmental arteries are frequently occluded. Under these circumstances, the spinal cord is perfused through collateral vessels. However, anatomical evaluation of such collateral vessels is not yet feasible.
The functional approach includes intraoperative monitoring of somatosensory evoked potentials (SSEPs) and motor-evoked potentials (MEPs). SSEPs are used to detect spinal cord ischemia during aortic cross-clamping and to identify CSAs. The important problem associates with SSEP monitoring is false negative results because it evaluates the sensory conduction system, not motor pathways.
MEP monitoring is a relatively new technique. This technique is theoretically superior to SSEP monitoring because it evaluates the motor neuron system directly. Recently, Jacobs reported complete prevention of neurological deficits in thoracoabdominal aortic aneurysm repair based on MEP monitoring [11]. However, MEPs are highly susceptible to various factors and the control of anesthesia, especially neuromuscular blockage is difficult.
Moreover, both strategies require sequential aortic cross-clamping to identify CSAs. This procedure carries the risk of embolization and should be avoided in the case of a severely atherosclerotic aorta.
Intraoperative ultrasonography (IOUS) is now widely used in various surgical procedures. In cardiac surgery, Wareing and associates reported that they could reduce the frequency of stroke by modifying the cannulation and clamping techniques based on IOUS evaluation [15]. In aortic surgery, IOUS provides precise mapping not only of the deceptive anatomical relationship between the true and false lumens of aortic dissections, but also of aortic wall lesions such as calcification, atherosclerotic ulcerations or plaque, as well as mural thrombi in an aortic aneurysm [16]. With the improvement of instruments, indications are expected to expand even further.
Doppler ultrasonography has been used to evaluate various cardiac lesions and vascular diseases. At present, it can measure the flow velocity in small arteries, like coronary artery bypass grafts, transthoracically [12,13] and intraoperatively [17]. In this study, using a 5-MHz linear probe for epiaortic scanning, we measured the flow velocity of segmental arteries, the diameters of which averaged 1 mm.
A previous study performed with Doppler ultrasonography revealed that velocity patterns are determined by the vascular resistance offered by the organ supplied and its metabolic activity. For example, the internal carotid artery has a larger diastolic flow velocity than the external carotid artery. This is because the internal carotid artery perfuses the brain, which demands a high level of blood flow. On the other hand, the femoral artery shows a reverse flow in the early diastolic period because the lower extremities have high vascular resistance. Thus, we hypothesized that ‘critical’ segmental arteries have different flow velocities because they perfuse not only the thoracic wall but also the spinal cord. After aortic cross-clamping, the flow velocity of the intercostal arteries increased, because they perfused the ischemic lower body. Furthermore, the degree of the increase differed among intercostal arteries. The increase was considered to be a contributor to flow into ischemic tissues, including the spinal cord. Therefore, we speculated that the intercostal artery with the largest flow increase (i.e. L-artery) was the CSA. The selective perfusion of this segmental artery completely prevented spinal cord damage during aortic cross-clamping just distal to the left subclavian artery. During cross-clamping, neither proximal nor distal blood pressure significantly differed among groups. Therefore, we consider this segmental artery to directly perfuse the spinal cord.
Lowell and associates reported identification of the CSA by angiography and selective shunting in experimental models [18]. In that study, the CSA was identified by selective angiography and an isolated aortic segment that included the CSA was perfused via the selective shunt during aortic cross-clamping. However, the neurological outcomes did not improve and tissue blood flow of the lumbar cord was not increased by the shunt. In our study, on the other hand, the selective shunt completely protected the spinal cord, leading us to speculate that tissue blood flow of the lumbar cord increased due to shunting. This difference stems from the means by which CSAs were identified. The ‘anatomically’ critical artery identified by angiography is not always ‘functionally’ critical. Therefore, the functional evaluation of segmental arteries by Doppler ultrasonography has the potential to resolve this problem.
Even selective perfusion of the segmental artery with the smallest flow increase produced some degree of spinal cord protection, suggesting that non-critical arteries contribute to the blood supply of the spinal cord. In a porcine model, de Haan and associates demonstrated that clamping of non-CSAs significantly reduced the tolerance of the spinal cord to ischemia [19]. Our results are consistent with theirs. In the present study, O-arteries were not examined by selective perfusion. Each animal has five O-arteries, and it was impossible to examine all O-arteries because of the limited number of experimental animals. We think that the selective perfusion of O-arteries would produce an intermediate degree of spinal cord protection.
The major limitation of this study is the anatomical difference in spinal cord blood supply between dogs and humans. The mean number of radicular feeders of the anterior spinal artery is reported to be 8 in humans; in dogs, it is 8.1 in the thoracic region and 6.3 in the lumbar region [20]. This anatomical difference might alter the hemodynamics of the segmental arteries and produce different responses to the aortic cross-clamping. Before clinical applications of this method are developed, some problems must be discussed. The aortic cross-clamping required to measure the flow velocity increase might cause ischemic spinal cord injury. In this experimental study, however, the flow measurements were completed within a 10-min period of aortic cross-clamping, and no spinal cord injuries occurred using this procedure. Because of anatomical differences in humans and dogs, appropriate level of aortic cross-clamping for flow velocity measurement might also differ. We have no data on the hemodynamics of segmental arteries in humans. Further clinical investigations are needed to clarify these points.
In conclusion, flow responses to aortic cross-clamping differed among segmental arteries. Based on this difference, it was possible to identify CSAs by Doppler ultrasonography. This strategy could be clinically applicable to spinal cord protection during operations for thoracoabdominal aortic aneurysm.
|
References
|
|---|
-
Coselli J.S., LeMaire S.A. Left heart bypass reduces paraplegia rates after thoracoabdominal aortic aneurysm repair. Ann Thorac Surg 1999;67:1931-1934.[Abstract/Free Full Text]
-
Galla J.D., Ergin M.A., Lansman S.L., McCullough J.N., Nguyen K.H., Spielvogel D., Klein J.J., Griepp R.B. Use of somatosensory evoked potentials for thoracic and thoracoabdominal aortic resections. Ann Thorac Surg 1999;67:1947-1952.[Abstract/Free Full Text]
-
Safi H.J., Winnerkvist A., Miller C.C., 3rd, Iliopoulos D.C., Reardon M.J., Espada R., Baldwin J.C. Effect of extended cross-clamp time during thoracoabdominal aortic aneurysm repair. Ann Thorac Surg 1998;66:1204-1209.[Abstract/Free Full Text]
-
Svensson L.G., Hess K.R., D'Agostino R.S., Entrup M.H., Hreib K., Kimmel W.A., Nadolny E., Shahian D.M. Reduction of neurologic injury after high-risk thoracoabdominal aortic operation. Ann Thorac Surg 1998;66:132-138.[Abstract/Free Full Text]
-
Kouchoukos N.T., Rokkas C.K. Hypothermic cardiopulmonary bypass for spinal cord protection: rationale and clinical results. Ann Thorac Surg 1999;67:1940-1942.[Abstract/Free Full Text]
-
Kouchoukos N.T., Daily B.B., Rokkas C.K., Murphy S.F., Bauer S., Abboud N. Hypothermic bypass and circulatory arrest for operations on the descending thoracic and thoracoabdominal aorta. Ann Thorac Surg 1995;60:67-77.[Abstract/Free Full Text]
-
Svensson L.G., Hess K.R., Coselli J.S., Safi H.J. Influence of segmental arteries, extent, and atriofemoral bypass on postoperative paraplegia after thoracoabdominal aortic operations. J Vasc Surg 1994;20:255-262.[Medline]
-
Williams G.M., Perler B.A., Burdick J.F., Osterman F.A., Jr, Mitchell S., Merine D., Drenger B., Parker S.D., Beattie C., Reitz B.A. Angiographic localization of spinal cord blood supply and its relationship to postoperative paraplegia. J Vasc Surg 1991;13:23-33.[Medline]
-
Cunningham J.N., Jr, Laschinger J.C., Spencer F.C. Monitoring of somatosensory evoked potentials during surgical procedures on the thoracoabdominal aorta: IV. Clinical observations and results. J Thorac Cardiovasc Surg 1987;94:275-285.[Abstract]
-
de Haan P., Kalkman C.J., de Mol B.A., Ubags L.H., Veldman D.J., Jacobs M.J. Efficacy of transcranial motor-evoked myogenic potentials to detect spinal cord ischemia during operations for thoracoabdominal aneurysms. J Thorac Cardiovasc Surg 1997;113:87-100.[Abstract/Free Full Text]
-
Jacobs M.J.H.M., Meylaerts S.A., de Haan P., de Mol B.A., Kalkman C.J. Strategies to prevent neurologic deficit based on motor-evoked potentials in type 1 and 2 thoracoabdominal aortic aneurysm repair. J Vasc Surg 1999;29:48-59.[Medline]
-
de Bono D.P., Samani N.J., Spyt T.J., Hartshorne T., Thrush A.J., Evans D.H. Transcutaneous ultrasound measurement of blood flow in internal mammary artery to coronary artery grafts. Lancet 1992;339:379-381.[Medline]
-
van Son J.A., Skotnicki S.H., Peters M.B., Pijls N.H., Noyez L., van Asten W.N. Noninvasive hemodynamic assessment of the internal mammary artery in myocardial revascularization. Ann Thorac Surg 1993;55:404-409.[Abstract]
-
Safi H.J., Miller C.C., III, Carr C., Iliopoulos D.C., Dorsay D.A., Baldwin J.C. Importance of intercostal artery reattachment during thoracoabdominal aortic aneurysm repair. J Vasc Surg 1998;27:58-68.[Medline]
-
Wareing T.H., Davila-Roman V.G., Barzilai B., Murphy S.F., Kouchoukos N.T. Management of the severely atherosclerotic ascending aorta during cardiac operations. J Thorac Cardiovasc Surg 1992;103:453-462.[Abstract]
-
Takamoto S. In: Minami K., Korfer R., eds. Decision-making by transcutaneous and transesophageal Doppler color flow mapping followed by intraoperative direct scanning in dissecting aortic aneurysm. Amsterdam: Elsevier, 1992:103-114.
-
Oda K., Hirose K., Nishimori H., Sato K., Yamashiro T., Ogoshi S. Assessment of internal thoracic artery graft with intraoperative color Doppler ultrasonography. Ann Thorac Surg 1998;66:79-81.[Abstract/Free Full Text]
-
Lowell R.C., Gloviczki P., Bergman R.T., Stanson A.W., Dzsinich C., Bower T.C., Cherry K.J., Jr Failure of selective shunting to intercostal arteries to prevent spinal cord ischemia during experimental thoracoabdominal aortic occlusion. Int Angiol 1992;11:281-288.[Medline]
-
de Haan P., Kalkman C.J., Meylaerts S.A., Lips J., Jacobs M.J. Development of spinal cord ischemia after clamping of noncritical segmental arteries in the pig. Ann Thorac Surg 1999;68:1278-1284.[Abstract/Free Full Text]
-
Caulkins S.E., Purinton P.T., Oliver J.E., Jr Arterial supply to the spinal cord of dogs and cats. Am J Vet Res 1989;50:425-430.[Medline]
Top
Abstract
1. Introduction
