Eur J Cardiothorac Surg 1999;15:571-578
© 1999 Elsevier Science NL
Comparative clinical study between retrograde cerebral perfusion and selective cerebral perfusion in surgery for acute type A aortic dissection
Akihiko Usui,
Kenzo Yasuura,
Takashi Watanabe,
Takashi Maseki
Department of Thoracic Surgery, Nagoya University School of Medicine,65 Tsurumai, Showa-ku, Nagoya, Japan 466
Received 9 June 1998;
received in revised form 16 February 1999;
accepted 10 March 1999.
Corresponding author. Tel.: +81-52-7442376; fax: +81-52-7442383
 |
Abstract
|
|---|
Objective: Selection of a brain protection method is a primary concern for aortic arch surgery. We performed a retrospective study to compare the respective advantages and disadvantages of retrograde cerebral perfusion (RCP) and selective cerebral perfusion (SCP) in patients who underwent surgery for acute type A aortic dissection. Methods: The study reviewed 166 patients who underwent surgery at Nagoya University or its eight branch hospitals between January 1990 and August 1996. There were 91 patients who received SCP and 75 patients who underwent RCP. Results for these two groups were compared. Results: There were no significant differences in age, gender, Marfan syndrome rate, DeBakey classification, or emergency operation rate. Rates of various preoperative complications were similar except for aortic valve regurgitation. Arch replacement was performed more often in SCP than in RCP patients (49% vs. 27%, P=0.0028). There were no significant differences between groups in cardiac ischemic time or visceral organ ischemic time. However, RCP group showed shorter cardio-pulmonary bypass time (297±99 vs. 269±112 min, P=0.013) and lower the lowest core temperature (21.6±3.1°C vs. 18.7±2.1°C, P=0.0001). SCP duration was longer than RCP duration (103±56 vs. 54±24 min, P<0.0001). Despite these differences, RCP patients were not significantly different from SCP patients with regard to any postoperative complication, neurological dysfunction (16 vs. 19%), or operative mortality (all deaths within the hospitalization; 24 vs. 21%). Regarding neurologic dysfunction, there were six cases of coma, six of motor paralysis, two of paraplegia and one of visual loss among SCP patients, and eight cases of coma, three of motor paralysis, and three of convulsion in the RCP group. The incidence of motor paralysis was higher in the SCP group, while the incidence of coma was higher in the RCP group. Conclusions: RCP can be performed without clamping or cannulation of the cervical arteries, which is an advantage in reducing the chances of arterial injury or cerebral embolization. RCP is comparable to SCP in terms of clinical outcome.
Key Words: Stanford type A • Aortic dissection • Aortic aneurysm • Brain protection • Surgical results
|
1. Introduction
|
|---|
Protection of the brain is a primary concern when performing surgery on the aortic arch. Retrograde cerebral perfusion (RCP) is an alternate method of brain protection during deep hypothermic circulatory arrest. RCP was reported as a technique for protection of the brain during aortic arch surgery by Lemole et al. in 1982 [1], and by Ueda et al. in 1990 [2], and has become generally accepted as an adjunct in cerebral protection. We have employed RCP for brain protection during aortic arch surgery since January, 1990. Several animal experimental studies have also been performed to determine the optimum manner of RCP management [3–6]. We have applied RCP via the SVC cannulae while clamping the IVC cannulae and maintaining the SVC pressure under 25 mmHg. Although our experience has shown the proven advantages of RCP, it remains to determine whether it should be chosen over conventional selective cerebral perfusion (SCP). We have thus performed a retrospective study among RCP and SCP in patients who underwent surgery for acute type A aortic dissection at Nagoya University or its eight branch hospitals. Ours was not a randomized clinical study but a retrospective and multicenter study; therefore, there were certain limitations in interpreting and evaluating the results. However, this study may be the first clinical comparison of RCP and SCP, which attempts to bring into clearer focus the several advantages and disadvantages of RCP.
|
2. Materials and methods
|
|---|
The study group consisted of 166 patients who underwent surgery for acute type A aortic dissecting aneurysm at Nagoya University or its eight branch hospitals between January, 1990 and August, 1996. SCP was used with 91 patients and RCP with 75. Operative indications, operative procedures and choice of brain protection method were decided by each surgeon or institute. The maximum series were 41 cases and the minimum ones were 12 cases. The present study is a clinical comparison between SCP and RCP in order to assess the usefulness of RCP for brain protection with surgery for acute type A aortic dissection. Acute aortic dissection was defined within 14 days since onset in this study.
RCP was performed as follows. The patient was generally placed on cardio-pulmonary bypass (CPB) with bicaval venous cannulation and groin arterial cannulation through a midsternotomy, and his or her temperature lowered by core cooling until nasopharyngeal temperature was approximately 20°C. CPB was then stopped and RCP started while maintaining SVC pressure at 25 mmHg or less. CPB was resumed as soon as RCP was terminated, and the patient was rewarmed. SCP was performed with elastic cannulae placed directly into the brachiocepharic and left carotid and/or left subclavian artery via aortotomy. Blood flow was managed 500–1000 ml/min per each cannulae.
Clinical profiles were based on age, gender, DeBakey classification, presence of Marfan syndrome, and emergency situation. Preoperative complications were determined by the presence of hemodynamic compromise, cardiac arrest, neurologic symptoms, free rupture, cardiac tamponade, greater than moderate aortic valve regurgitation, renal dysfunction or leg ischemia. Surgical factors included the approach (midsternotomy or additional left thoracotomy), operative procedure (ascending aorta and proximal aortic arch repair or aortic arch replacement) and combination aortic root reconstruction or elephant trunk procedure. Cardiopulmonary bypass time, cardiac ischemic time, visceral organ ischemic time and lowest nasopharyngeal temperature were used to characterize the surgical conditions. Clinical results were based on timing of awakening and duration of endotracheal intubation. Surgical outcomes were classified according to operative complications (pulmonary, bleeding, cardiac, renal, hepatic), and visceral organ ischemia or infection. Cardiac complications were defined as severe low output syndrome or conditions requiring an assist circulatory device (LVAD, PCPS or IABP). Pulmonary complications were defined as conditions requiring over 5 days of respirator use. Hepatic complication was limited to severe jaundice (over 10 mg/dl of total bilirubin). Renal complication was a condition which required artificial dialysis. Any neurologic symptoms persisting after operation was defined as a neurologic dysfunction. Operative mortality included all deaths occurring during the hospitalization in which the operation was performed.
Data were evaluated using the Statistical Analysis System (SAS). Results were expressed as mean±SD. Statistical significance between groups was determined using the Fisher's exact test and the Mann–Whitney test. All preoperative and operative factors were further subjected to multi-variant analysis using forward stepwise logistic regression to evaluate the significant predictors for neurologic outcome. A P value less than 0.05 was considered significant.
|
3. Results
|
|---|
3.1. General
The mean age of all patients was 60.6±12.8 years (26–86 years). The male-female ratio was 93:73. According to DeBakey classifications, 74% of patients were type I, 13% of patients, type II and 13% type III retrograde. There were eight patients diagnosed as typical Marfan syndrome. Seventy-five percent of all patients were operated on within 48 h of the onset of symptoms. Preoperative complications included hemodynamic compromise (21%), cardiac arrest (6%), neurologic disorders (11%), free rupture (7%), cardiac tamponade (23%), aortic regurgitation (26%), renal failure (8%) and leg ischemia (4%). Ascending aorta and proximal arch repair were performed in 61% of patients and aortic arch replacement in 39%. Combination aortic root reconstruction or elephant trunk was done in 7% and 11% of cases, respectively.
Eighty-four percent of patients awoke within 24 h after surgery. Thirteen patients, however, never awoke. Twenty-nine patients (17%) presented various neurological disorders, including coma (14), motor paralysis (9), convulsion (3), paraplegia (2) and visual loss (1). Seven patients recovered from their symptoms completely, but, 13 died within 30 days of operation (Table 1). The various other postoperative complications were pulmonary (44%), bleeding (10%), cardiac (8%), renal (11%) and hepatic (4%) complications, visceral organ ischemia (2%) and infection (10%). Operative mortality was 23%, including 15% of mortality within 30 days of surgery.
View this table:
[in this window]
[in a new window]
|
Table 6. Causes of death in patients who have undergone selective cerebral perfusion (SCP) or retrograde cerebral perfusion (RCP)a
|
|
View this table:
[in this window]
[in a new window]
|
Table 1. Neurologic symptoms in patients who have undergone selective cerebral perfusion (SCP) or retrograde cerebral perfusion (RCP)a
|
|
3.2. Comparison of clinical and operative factors between the SCP and RCP groups
There were no significant differences in mean age, male-female ratio or presence of Marfan syndrome between the groups. Distribution of DeBakey classification also showed no significant differences (Table 2). The ratio of operations within 48 h of the onset was slightly higher in the RCP group (83%) than in the SCP group (69%); however, this was not significant. The presence ratio of various preoperative complications was similar between the groups, except for aortic valve regurgitation (34% in SCP and 16% in RCP, P=0.0122) (Table 2).
View this table:
[in this window]
[in a new window]
|
Table 2. Clinical characteristics and preoperative complications of patients who have undergone selective cerebral perfusion (SCP) or retrograde cerebral perfusion (RCP)a
|
|
More than half the patients (51%) in the SCP group underwent aortic arch replacement. In contrast, replacements in 73% of the RCP group were done only in the ascending and proximal arch (P=0.0039). Combination elephant trunk procedure was also more frequent in the SCP group (15% vs. 7%, P=0.0895). The rate of combination ratio of aortic root reconstruction was the same in both groups. There were no significant differences between groups in cardiac ischemic time or visceral organ ischemic time. However, cardio-pulmonary bypass time of SCP group was longer than that of RCP group (297±99 vs. 269±112 min, P=0.013) and patients of the RCP group were cooled down more deeply than SCP group patients (21.6±3.1°C vs. 18.7±2.1°C, P=0.0001). SCP duration was not comparable to RCP duration, however, there were many differences (SCP; 103±56 vs. RCP; 54±24 min, P<0.0001) (Table 3).
View this table:
[in this window]
[in a new window]
|
Table 3. Surgical factors in patients who have undergone selective cerebral perfusion (SCP) or retrograde cerebral perfusion (RCP)a
|
|
3.3. Comparison of operative outcome between SCP and RCP group
Eighty-eight percent of the SCP group and 79% of the RCP group awoke within 24 h after operation. Four percent of the SCP group and 12% of the RCP group never awoke. The percentage of patients who never awoke was higher in the RCP group than the SCP group, but this difference was not significant. Sixteen percent of the SCP group and 19% of the RCP group suffered from various neurologic disorders after operation, compared to preoperative neurologic symptoms in 12% of SCP and 9% of RCP patients. There were six cases of coma, six of motor paralysis, two of paraplegia and one of visual loss in the SCP group and eight cases of coma, three of motor paralysis, and three of convulsion in the RCP group. Thirteen of these 29 patients who suffered from neurologic disorders died in its hospitalization due to neurologic disorders in eight, multi-organ failure (MOF) in two, sepsis in two and respiratory failure in one. Only seven patients recovered without adverse effects. Ten of 14 patients who fell into a coma died early, while only one of nine patients with motor paralysis died in its hospitalization (Table 1).
The findings of head CT examination in patients who suffered from neurologic dysfunction were the following: infarction in seven, bleeding in one and edema in two patients of the SCP group and infarction in five, edema or atrophy in five patients of the RCP group. Two patients from each group presented no particular findings. Eight of nine patients with motor paralysis revealed brain infarction, whereas half of patients who fell into a coma revealed brain edema or atrophy and the other half did brain infarction (Table 4).
View this table:
[in this window]
[in a new window]
|
Table 4. Head CT findings in patients who have undergone selective cerebral perfusion (SCP) or retrograde cerebral perfusion (RCP)a
|
|
Pulmonary complication was another major postoperative complication. Forty-two percent of the SCP group and 47% of the RCP group required more than 5 days of respiratory support and tracheotomy was performed 3% of patients from each group. However, these ratios were not significantly different between the groups. The rates of the following postoperative complications were also not significantly different between the groups: bleeding, cardiac, renal or hepatic complication, visceral organ ischemia or infection (Table 5).
View this table:
[in this window]
[in a new window]
|
Table 5. Postoperative complications and mortality in patients who have undergone selective cerebral perfusion (SCP) or retrograde cerebral perfusion (RCP)
|
|
Operative mortality were 24% in the SCP group and 21% in the RCP group. There were not significant differences between groups. The causes of death in the SCP group were MOF in seven patients, cardiac failure in four, neurologic dysfunction in four, sepsis in three, bleeding in one, re-rupture in one, myonephropathic metabolic syndrome (MNMS) in one and respiratory failure in one. In the RCP group, death was caused by MOF in four patients, cardiac failure in five, neurologic dysfunction in four, sepsis in one, bleeding in one, and re-rupture in one (Table 6).
3.4. Comparison of operative outcome in patients underwent ascending aortic replacement between SCP and RCP group
Considering only patients who underwent ascending aortic replacement, there were 46 patients in the SCP group and 55 patients in the RCP group. When these patients are compared, there were no significant differences in age, gender, Marfan syndrome rate, DeBakey classification, or emergency situation. Rates of various preoperative complications showed no significant differences. There were no significant differences between groups in cardio-pulmonary bypass time, cardiac ischemic time or visceral organ ischemic time, except in the lowest core temperature. The SCP duration was significant longer than RCP duration. Despite of these differences, postoperative complication rates including neurologic dysfunction showed no significant differences. Operative mortality was somewhat higher in the SCP group than in RCP group, although this difference was not significant (Tables 7 and 8).
View this table:
[in this window]
[in a new window]
|
Table 7. Clinical characteristics and preoperative complications of patients who have undergone ascending aortic replacement by using selective cerebral perfusion (SCP) or retrograde cerebral perfusion (RCP)a
|
|
View this table:
[in this window]
[in a new window]
|
Table 8. Postoperative complications and mortality in patients who have undergone ascending aortic replacement by using selective cerebral perfusion (SCP) or retrograde cerebral perfusion (RCP)
|
|
3.5. Relationship between neurologic dysfunction and RCP management
RCP was generally performed by maintaining the SVC pressure at less than 25 mmHg. In this condition, the RCP blood flow rate and RCP duration are important factors in controlling the RCP. Furthermore, only these two factors were remained to be significantly associated with the occurrence of neurologic dysfunction by the forward stepwise logistic regression (RCP duration; odds ratio 1.0555, P=0.0020, RCP blood flow rate; odds ratio 0.9875, P=0.0074). The figure shows the relationship among neurologic outcome, RCP duration and RCP blood flow rate. There were 14 patients who suffered from neurological dysfunctions with coma in eight, motor paralysis in three and convulsion in three. Twelve of them were grouped in the region of RCP blood flow rates below 350 ml/min. The incidence of neurologic dysfunction was 32% (12/38) in this region, reaching 44% (8/18) when considering only those cases in which RCP was continued for more than 60 min. On the other hand, in the region of RCP blood flow rates above 350 ml/min and duration under 60 min, there was only one patient with a neurologic disorder for a complication rate of 4% (1/28) (Fig. 1)
.
View larger version (21K):
[in this window]
[in a new window]
|
Fig. 1. Prevalence of neurologic dysfunction related to duration of retrograde cerebral perfusion (RCP) and RCP blood flow rate. X-Axis indicates RCP blood flow rate and Y-axis shows RCP duration. Open circles shows patients with no neurological disorders, shaded triangles shows patients with motor paralysis, closed squares shows patients with coma and shaded circles shows patients with convulsion, respectively. Parenthesis shows percentage with neurologic dysfunction in each area.
|
|
|
4. Discussion
|
|---|
Retrograde cerebral perfusion (RCP) is an alternate method of brain protection. It was first reported in the successful management of a massive air embolism during cardio-pulmonary bypass [7]. As a technique for protecting the brain during aortic arch surgery, intermittent RCP was reported by Lemole et al. in 1982 [1], and continuous RCP was reported by Ueda et al. in 1990 [2]. Experimental studies showed that RCP supplied oxygen to the brain during interruption of physiological cerebral circulation and that it minimized ischemic damage to the brain [3–6]. RCP may therefore be a useful adjunct procedure in cases of hypothermic circulatory arrest because it can extend the period of time during which cerebral circulation can be safely interrupted. However, there have been no comparative clinical studies between RCP and SCP. Since RCP requires no additional cannulation or clamps, which might cause arterial injury and lead to embolization, and no additional extracorporeal circuits or complicated techniques such as those used in SCP, it is presumably a better, safer procedure. But since it can not assure a sufficient supply of blood and oxygen to the brain for long periods, it is less reliable than SCP for brain protection; it is a non-physiological form of perfusion and thus can not be performed uniformly in every case. The precise advantages and disadvantages of RCP compared with SCP require much further clarification.
The present study may be the first clinical comparison between RCP and SCP methods. This is not a randomized clinical study but a retrospective and also multicenter study. Despite these limitations in result evaluation, the study at least serves to clarify the advantages and disadvantages of the respective approaches.
Patients in the SCP group have complications with preoperative aortic valve regurgitation more frequently than do those in the RCP group. There are four obvious surgical differences: operative procedure, cardio-pulmonary bypass time, lowest body temperature, and duration of SCP or RCP. Half of patients in the SCP group underwent aortic arch replacement, while this was done in only a quarter of the RCP group. Patients in the RCP group were cooled down more deeply, to under 20°C, than those in the SCP group. Despite these differences, however, there were no significant differences between the groups with regard to any postoperative complication, neurological outcome or operative mortality. Each group tended to have certain characteristic neurological dysfunctions. While half of patients had some motor paralysis and there was no convulsion in the SCP group, there were eight patients with coma and three with convulsion of the 14 patients in the RCP group. Motor paralysis may be caused by focal brain infarction, while coma or convulsion are more likely due to ischemic brain damage. Actually, most of patients with motor paralysis revealed cerebral infarction on head CT finding. It might be expected, therefore, that neurological dysfunctions in the SCP group would be more closely related to embolic episodes, arterial injuries or unbalanced perfusion and ones in the RCP group may result chiefly from brain ischemia due to prolonged RCP.
There were much differences in operative methods between groups. Considering only patients who underwent ascending aortic replacement, therefore, there are 46 patients in the SCP group and 55 patients in the RCP group. Even these groups have some problems to be compared, however, the incidence of postoperative neurologic dysfunction is the same (15% in both groups). The RCP group has somewhat lower mortality than the SCP group, although the difference is not significant. We believe that RCP is comparable to SCP in terms of clinical outcome.
Operative mortality and postoperative neurological dysfunction in the present study were 24% and 16% in the SCP group, and 21% and 19% in the RCP group, respectively. These incidences may be relatively high. However, 69% of the SCP group and 83% of the RCP group were operated within 48 h of the onset of symptoms, and 49% of the SCP group and 27% of the RCP group had aortic arch replacement. Furthermore, there was preoperative cardiac arrest in 5% of SCP and 7% of RCP patients, and preoperative neurological symptoms in 12% of SCP and 9% of RCP patients. The mortality of patients undergoing emergency aortic arch replacement for acute type A aortic dissection has been reported to be high [8–11] even in recent reports; 23% by Kazui et al. [12] and 15% by Tabayashi et al. [13]. Our results do not differ markedly from rates reported in other reports of surgery for acute type A aortic dissection under emergency situations with high incidences of preoperative complications.
The present study also indicates a relationship among neurologic symptoms, RCP duration and RCP blood flow rate. We maintain RCP by monitoring SVC pressure at less than 25 mmHg. Under this condition, RCP blood flow rate is a major factor in RCP management. When an RCP blood flow of more than 350 ml/min could not be obtained, neurological dysfunction occurred in 32% of patients. However, when RCP is limited to less than 60 min and RCP blood flow rate is maintained at more than 350 ml/min, the incidence of neurologic disorder decreases to be only 4%. RCP is a non-physiological perfusion and cannot be performed uniformly. Therefore, when insufficient, such as when a perfusion flow rate of more than 350 ml/min can not be obtained, RCP should be limited to 60 min and normograde circulation resumed as early as possible. Embolic strokes with atheromata, thrombi, or air occur even in operations for dissecting aneurysm and these episodes may be confused with RCP-related neurological outcomes. However, the neurological outcome is expected to improve when RCP is controlled so that a perfusion rate of more than 350 ml/min is obtained and RCP is limited to 60 min.
Retrograde cerebral perfusion can be performed without clamping or cannulation of the cervical arteries. This is an advantage in that it reduces the chances of arterial injury or cerebral thrombosis, while not obscuring the operative field. Retrograde cerebral perfusion can extend the duration of safe cerebral circulatory arrest. Despite its time limitations, RCP may be comparable to SCP in terms of clinical outcome.
|
Acknowledgments
|
|---|
The authors wish to express their deepest gratitude to all of the following contributors who participated in this survey: Hiroshi Ina, Atsukata Kobayashi, Yutaka Ogawa, Hiroshi Okamoto, Akira Seki, Kazuki Tajima, Shuji Tamaki, Masanobu Maeda, Yoshiya Miyata, Hoshito Moriya, Hiroshi Yano, Katsuhiko Yoshida.
|
References
|
|---|
- Lemole G.M., Strong M.D., Spagna P.M., Karmilowicz N.P. Improved results for dissecting aneurysms: intraluminal sutureless prosthesis. J Thorac Cardiovasc Surg 1982;83:249-255.[Abstract]
- Ueda Y., Miki S., Kusuhara K., Okita Y., Tahara T., Yamanaka K. Surgical treatment of aneurysm of dissection involving the ascending aorta and aortic arch, utilizing circulatory arrest and retrograde cerebral perfusion. J Cardiovasc Surg 1990;31:553-558.[Medline]
- Usui A., Hotta T., Hiroura M., Murase M., Maeda M., Koyama T., Tanaka M., Takeuchi E., Yasuura K., Watanabe T., Abe T. Retrograde cerebral perfusion through a superior vena caval cannulae protects the brain. Ann Thorac Surg 1992;53:47-53.[Abstract]
- Usui A., Hotta T., Hiroura M., Murase M., Maeda M., Koyama T., Tanaka M., Takeuchi E., Abe T. Cerebral metabolism and function during normothermic retrograde cerebral perfusion. J Cardiovasc Surg 1993;1:107-112.
- Usui A., Oohara K., Tong-lin L., Murase M., Tanaka M., Takeuchi E., Abe T. Comparative experimental study between retrograde cerebral perfusion and circulatory arrest. J Thorac Cardiovasc Surg 1994;107:1228-1236.[Abstract/Free Full Text]
- Usui A., Oohara K., Tong-lin L., Murase M., Tanaka M., Takeuchi E., Abe T. Determination of optimum retrograde cerebral perfusion conditions. J Thorac Cardiovasc Surg 1994;107:300-308.[Abstract/Free Full Text]
- Mills N.L., Ochsner J.L. Massive air embolism during cardiopulmonary bypass: causes, prevention, and management. J Thorac Cardiovasc Surg 1980;80:708-717.[Abstract]
- Boulafendis D., Bastounis E., Panayiotopoulos Y.P., Papalambros E.L. Acute type A dissecting aortic aneurysm requiring emergency arch replacement. Cardiovasc Surg 1993;1:414-418.[Medline]
- Dottori V., Spagnolo S., Passerone G., Lijoi A., Barberis L., Agostini M., De Gaetano G., Parodi E., Maccario M., Fumagalli C. Ten years of surgery of aortic dissections and aneurysms: clinical experience and original contributions. Minerva Cardioangiologica 1992;40:431-436.[Medline]
- Bachet J., Teodori G., Goudot B., Diaz F., el Kardany A., Dubois C., Brodaty D., de Lentdecker P., Guilmet D. Replacement of the transverse aortic arch during emergency operations for type A acute aortic dissection: report of 26 cases. J Thorac Cardiovasc Surg 1988;96:878-886.[Abstract]
- Yun K.L., Glower D.D., Miller D.C., Fann J.I., Mitchell R.S., White W.D., Rankin J.S., Wolfe W.G., Shumway N.E. Aortic dissection resulting from tear of transverse arch: is concomitant arch repair warranted?. J Thorac Cardiovasc Surg 1991;102:355-370.[Abstract]
- Kazui T., Kimura N., Yamada O., Komatsu S. Total arch graft replacement in patients with acute type A aortic dissection. Ann Thorac Surg 1994;58:1462-1468.[Abstract]
- Tabayashi K., Niibori K., Iguchi A., Shoji Y., Ohmi M., Mohri H. Replacement of the transverse aortic arch for type A acute aortic dissection. Ann Thorac Surg 1993;55:864-867.[Abstract]
Top
Abstract
1. Introduction
