Ann Thorac Surg 2004;77:977-981
© 2004 The Society of Thoracic Surgeons
Original article: cardiovascular
Cardiac reoperation by Carpentier bicaval femoral venous cannula: GATA experience
Erkan Kuralay, MDa*,
Cengiz Bolcal, MDa,
Faruk Cingoz, MDa,
Celalettin Günay, MDa,
Vedat Yildirim, MDb,
Selim Kilic, MDc,
Ertugrul Özal, MDa,
Ufuk Demirkilic, MDa,
Mehmet Arslan, MDa,
Harun Tatar, MDa
a Department of Cardiovascular Surgery, Gülhane Military Medical Academy, Ankara, Turkey
b Department of Anesthesiology, Gülhane Military Medical Academy, Ankara, Turkey
c Department of Biostatistics and Public Health, Gülhane Military Medical Academy, Ankara, Turkey
Accepted for publication September 8, 2003.
* Address reprint requests to Dr Kuralay, Yazanlar sokak No = 31, 11, Asagi Ayranci, Ankara, Turkey 06540
e-mail: ekural{at}gata.edu.tr
 |
Abstract
|
|---|
BACKGROUND: Division of the sternum is primarily a blind procedure in reoperation and carries an increased risk of injury for major cardiac structures in the presence of adhesions between the posterior table and the heart.
METHODS: Two hundred patients were randomly divided into two groups. Cardiopulmonary bypass was established through the femoral artery and vein in group 1 (n = 100) patients before sternal reentry. Carpentier dual-stage femoral venous return cannula was used in all group 1 patients. Cardiopulmonary bypass was performed after sternal reentry in group 2 (n = 100) patients.
RESULTS: Six severe cardiac injuries developed in group 2. Cardiopulmonary bypass time was 93 ± 9 minutes in group 1 and 71 ± 11 minutes in group 2 (p = 0.011), and the operation time was 155 ± 23 minutes in group 1 and 185 ± 32 minutes in group 2 (p = 0.024). Inotropic therapy was required in 52 patients in group 1 and 76 patients in group 2 (p = 0.032). Average chest drainage was 450 ± 135 mL in group 1 and 850 ± 250 mL in group 2 (p < 0.001). Average fresh whole blood transfusion was 3.3 ± 1.2 U in group 1 and 5.8 ± 0.9 U in group 2 (p = 0.033). Average intensive care unit stay was 2.2 ± 1.3 days in group 1 and 4.5 ± 2.3 days in group 2 (p = 0.025). Average hospital stay was 7.3 ± 2.4 days in group 1 and 9.1 ± 3.1 days for group 2 (p = 0.011).
CONCLUSIONS: Cardiopulmonary bypass by bicaval Carpentier femoral venous cannula before resternotomy not only allows adequate cardiopulmonary bypass flow but also significantly reduces the risk of cardiac injury and catastrophic hemorrhage and allows safe reopening. Although this procedure increases cardiopulmonary bypass time, the operation time, bleeding, and blood transfusion requirement are significantly reduced.
|
Introduction
|
|---|
Since the median sternotomy incision was introduced by Julian and colleagues in the 1950s [1], it has evolved as the approach of choice for cardiac surgical procedures. Median sternotomy is the approach of choice for both congenital and acquired heart disease in the latest cardiac practice. As operative techniques evolve and survival after cardiac operations improves, the number of patients who have repeat sternotomy also continues to rise. Despite various advanced interventional cardiac innovations, the incidence of redo coronary artery bypass operation has been reported to be 8%. Similarly, the incidence of reoperations for aortic and mitral valve procedures has been reported as 11% and 20%, respectively [2]. Division of the sternum is primarily a blind procedure and carries an increased risk of injury for major cardiac structures in the presence of adhesions between the posterior table and the innominate vein, right ventricle, and extracardiac conduits and grafts. The risk of catastrophic hemorrhage on sternal reentry is estimated to be 0.5% to 1% with an associated mortality rate of 21% [2]. Dobell and Jain [3] obtained data from a questionnaire for the reoperation experience of 224 surgeons. They collected 144 reports of catastrophic hemorrhage on sternal redivision. The mortality rate was reported as 37% among these patients. Follis and colleagues [4] also carried out a questionnaire among the members of The Society of Thoracic Surgeons. They reported 2,046 catastrophic hemorrhages resulting in 392 deaths (19%). Several techniques have been reported for reducing sternal reentry complications and managing these complications [5–8]. The safest technique for eliminating and management of sternal reentry complications is the establishment of cardiopulmonary bypass (CPB). The full cardiac index cannot be established with classic venous cannula because of inadequate venous return. The recently designed dual-stage venous cannula (Carpentier), which is being used in ministernotomy approaches, allows adequate venous return. Full cardiac index can be established with this cannula. We have reported our experience with femorofemoral CPB with Carpentier venous cannula in reoperations.
|
Material and methods
|
|---|
Two hundred consecutive operations of repeat median sternotomy were performed in adult patients between June 1996 and June 2003 at Gülhane School of Medicine. The procedure of reoperation was considered only if the previous sternotomy was performed at least 4 weeks before the second procedure. Patients were randomly divided into two groups. This prospective randomized clinical study was done among the patients operated on for repeat valvular and congenital heart disease disregarding the type of previous operation.
Coronary reoperation was not included in this study, because most coronary reoperations were done in patients without using the left internal thoracic artery (LITA) in the first operation. We routinely harvested the LITA in these patients. If the LITA was not used in the first operation, we did not routinely establish CPB through the femoral artery and vein in coronary reoperation as harvesting of the LITA in coronary reoperation takes extra time.
A table of random digits was used for randomization of the patients [9]. Cardiopulmonary bypass was established through the femoral artery and vein in group 1 just before sternal reentry. A Carpentier dual-stage femoral venous return cannula was used in all group 1 patients. Cardiopulmonary bypass was performed in the classic way (ascending aorta and right atrium) after sternal reentry, and the heart was completely dissected in group 2. Each group included 100 patients. Patients' preoperative characteristics are summarized in Table 1. The same dosage of Trasylol protocol was administrated to both groups. (Two million units were added to the pump prime and 500,000 U/h were infused through the central venous route.)
Surgical technique
Group 1
The femoral artery and vein were prepared for cannulation, and simultaneously skin scar tissue on the previous sternotomy incision and sternal wires were removed. The incision line on the sternum was marked by electrocautery. Heparin was administrated at a dosage of 3.5 mg/kg. We tried to keep the activated clotting time more than 480 seconds. First, the venous cannula was inserted. The Carpentier femoral venous cannula was designed for minimally invasive cardiac procedures. First, we advanced guidewire through the femoral vein into the right atrium. Atrial arrhythmia, induced by the guidewire, was indicated when the guidewire had been placed. Then transverse venotomy was performed, and the Carpentier bicaval femoral venous cannula (Medtronic Inc, Minneapolis, MN) and its obturator were slipped over the guidewire without any difficulty into the right atrium and superior vena cava. The guidewire and cannula obturator were removed, and the Carpentier venous cannula was connected to the venous line of the pump. The prepared femoral artery was carefully palpated to recognize any atherosclerotic changes. The softest part of the femoral artery was chosen, and the guidewire was advanced into the femoral artery through a needle. When the guidewire was advanced into the arterial system without difficulty, then arterial clamps were placed and transverse arteriotomy was performed. Both the arterial cannula (DLP Femoral Arterial Cannula; Medtronic) and its obturator were advanced into the femoral artery. We were routinely using the guidewire both for advancing the cannula easily and avoiding retrograde dissection (Fig 1). When it was not possible to advance the guidewire into the femoral artery, we preferred the right brachial artery for arterial return as Ta
demir and colleagues had reported [10]. We used the brachial artery in 3 patients. Cardiopulmonary bypass was established without difficulty in all cases. Optimal flow could be pumped in most patients (n = 93). However, optimal flow could not be pumped in 7 patients because of arterial return resistance. Body temperature was reduced, and optimal flow was pumped according to the lowered body temperature. We generally did not cool patients to less than 32°C during the sternal reentry to avoid development of heart fibrillation, which might cause heart distention and compel safer heart dissection. An oscillating saw was used in all patients. After the resternotomy was completed, first the left part of the sternum was slightly elevated by the second assistant and then the adhesions beneath the sternum could be dissected by electrocautery. The left pleurotomy was routinely performed if it had not been done previously. Immediately after the completion of the dissection of the left part of the sternum, the right part was dissected similarly. We preferred electrocautery dissection in these patients instead of sharp dissection. Electrocautery dissection could easily be performed in these cases because the heart was decompressed by CPB. We did not intend to dissect all parts of the heart. We tried to dissect the aorta for placing both the cross-clamp and the antegrade cardioplegia cannula. Venting of the left ventricle was accomplished by a cannula placed in the right superior pulmonary vein extrapericardially. Air was removed from the heart with a long needle by advancing it from the right ventricle and the interventricular septum into the left ventricle.
View larger version (19K):
[in this window]
[in a new window]
|
Fig 1. Cannulation of femoral vessels. A Carpentier bicaval venous cannula is advanced into the superior vena cava. If tricuspid and right atrial procedures were planned, both cava-atrial junctions can be taped without any disturbance of venous drainage. The femoral artery is cannulated with a long arterial cannula. First, a guidewire is advanced into the femoral and iliac artery. If the guidewire is placed without difficulty, then the arterial cannula and its obturator are slipped over the guidewire into the arterial system. The risk of retrograde dissection can be eliminated with this cannula insertion technique.
|
|
There were 6 patients in this group with a patent LITA. The left side of the sternum, the heart, and the LITA were dissected using the technique described by Elami and colleagues [11] The LITA flow during the cross-clamp period was stopped from the supraclavicular approach as it has previously been described [12]. Myocardial preservation was performed using antegrade cardioplegia. Topical cooling was also used. A retrograde cardioplegia cannula was not routinely placed.
Group 2
A full sternotomy by oscillating saw was performed in all patients. Median sternotomy and mediastinal dissection were performed before full systemic heparinization and ascending aortic and right atrial cannulations were performed in 94 patients. Moderate hypothermia (28°C) was used. In this period the blood pressure was controlled by metaraminol. The left side of the heart was dissected after CPB was established in case of necessity for removing air from the heart. In case of excessive bleeding during the sternal reentry, dissection of the heart was stopped and CPB was established through the femoral artery and the vein. The femoral artery and vein cannulation was performed as described previously. Only in the 1 patient who had a patent LITA in the second heart operation were the heart and LITA dissected using the technique described by Elami and coworkers [11].
Statistical analysis
Statistical analysis was performed with the SPSS software, version 9.05 (SPSS Inc, Chicago, IL). The clinical data were expressed as the mean ± the standard deviation. The differences were analyzed with the Fisher's exact test, the independent Student's t test, and the chi-squared (
2) test.
|
Results
|
|---|
Six severe cardiac injuries developed during sternal reentry in group 2. Four of these injuries were localized to the right ventricle, one of them was localized to the right atrium, and one of them was localized to the ascending aorta. One patient who had extensive injury to the right ventricle and the patient who experienced an ascending aorta laceration died quickly. Cardiopulmonary bypass could not have been established because of the peripheral arterial disease in these 2 patients. Cardiopulmonary bypass was immediately established in the other 4 patients, and then the resternotomy was completed and the sternal retractor could be placed. Cardiac lacerations were repaired easily in these 4 patients.
There was not any cardiac injury in group 1 (p < 0.029). The average cross-clamp time was 57 ± 7 minutes in group 1 and 51 ± 8 minutes in group 2 (p = 0.234). The total CPB time was 93 ± 9 minutes in group 1 and 71 ± 11 minutes in group 2 (p = 0.011), whereas the operation time was 155 ± 23 minutes in group 1 and 185 ± 32 minutes in group 2 (p = 0.024). Intraaortic balloon counterpulsation was required in 9 patients during weaning from CPB in group 2, but in only 5 patients in group 1 (p = 0.268). Inotropic therapy was required in 52 patients in group 1 and 76 patients in group 2 (p = 0.032). The average chest drainage was 450 ± 135 mL in group 1 and 850 ± 250 mL in group 2 (p < 0.001). The average fresh whole blood transfusion was 3.3 ± 1.2 U in group 1 and 5.8 ± 0.9 U in group 2 (p = 0.033); the average fresh-frozen plasma transfusion was 2.1 ± 0.5 U in group 1 and 2.7 ± 0.5 U in group 2 (p = 0.344). The average intensive care unit stay was 2.2 ± 1.3 days in group 1 and 4.5 ± 2.3 days in group 2 (p = 0.025). The average hospital stay was 7.3 ± 2.4 days in group 1 and 9.1 ± 3.1 days for group 2 (p = 0.011). Five patients died in group 2 whereas only 2 patients died in group 1 (p = 0.445). Wound infection of the femoral incision developed in 4 obese patients. The data on the operative and postoperative variables of the patients are summarized in Table 2.
|
Comment
|
|---|
Resternotomy has become a common part of modern cardiac surgical practice (10% to 16% per year) [2]. Cardiac reoperation is associated with increased death and major morbidity as compared with the first operation. The resternotomy itself is the cause of some morbidity because of the adherence to structures of the heart. Despite several other reported reopening techniques [4–8], sternal reentry is the primary cause of morbidity and mortality in redo cardiac surgery [13]. O'Brien and colleagues [14] believe that the risks of cardiac injury and catastrophic hemorrhage at the time of resternotomy and the subsequent elevated morbidity in the postoperative period are completely avoidable. Pericardial closure at the first operation had been performed in 46% of their patients. They reported zero incidence of major cardiac injury in their study by using the technique of Grunwald [6]. Despite these outstanding results, cardiac injury during the sternal reentry remained a great challenge. The questionnaire by Follis and colleagues [4] among the members of The Society of Thoracic Surgeons revealed 2,046 catastrophic hemorrhages resulting in 392 deaths (19%). The rate of events per surgeon per 10 years was 2.09. The structures that were identified as injured (n = 1,557) were the right ventricle (609, 39%), the vein grafts (326, 20%), the mammary artery (181, 12%), and the innominate vein (99, 6%). Dobell and Jain [3] collected 144 reports of catastrophic hemorrhage caused by cardiac or ascending aorta injury during sternal reentry. The outcome in 2 patients was not indicated. Of the remaining 142 patients, 52 died, for a mortality of 37%. The incidence and mortality of the catastrophic hemorrhage-related cardiac injury were relatively lower than that of Dobell and Jain according to the questionnaire by Follis and associates [4]. They reported that the risk of catastrophic hemorrhage as 0.5% to 1% with an associated mortality rate of 21%. Today the precise incidence of cardiac injury during sternal reentry remains unknown. However, the experience gained with the large number of cases has decreased its frequency and the number of poor outcomes. We believe this entity still occurs sometimes in the career of a thoracic surgeon and it is not reported like any event considered secondary to the technical failure. Six catastrophic hemorrhages developed in our experience, and 2 of the patients in group 2 died.
Limited or focused surgical dissection has been emphasized by a large number of surgeons [13, 14]. As the dictum goes, "if you don't cut, it won't bleed" (avoiding unnecessary surgical dissection is the key to the modern surgical intervention). With this concept newer procedures have emerged in valvular and congenital surgery. A focal dissection is directed to the aorta or the right atrium. The front of the right ventricle is dissected to place the retractor, the left ventricle is not dissected, and the cava-atrial junctions are not dissected intentionally unless the right atrium is entered in the transseptal approach of mitral or tricuspid valvular surgery. But in the classic approach at least the ascending aorta and the right atrium dissections are obligatory for cannulation. We have routinely established CPB just before sternal reentry in group 1 in our study. Adequate venous return can be achieved by Carpentier bicaval femoral cannula. Adequate arterial flow can also be achieved by the recently developed femoral artery cannula. Sternal and substernal dissection can easily be performed on a nondistended heart. Satisfactory CPB flow by femoral cannulas allows practically every kind of cardiac operation. Right atrial and tricuspid valve operations can also be performed with our approach. Both cava-atrial junctions can be dissected and taped easily. The majority of cardiac surgeons use femorofemoral CPB before sternal reentry in patients with open LITA, conduit, ascending aorta aneurysm, enlarged right atrium or ventricle, or history of mediastinitis (classified as high-risk patients). They also avoid establishing CPB before the sternal reentry because of the adverse effect of extracorporeal circulation. Excessive tendency of hemorrhage is the most feared complication of relatively long CPB. In the classic approach, aspiration of bleeding materials into the oxygenator incites a significant consumptive coagulopathy and defibrillation syndrome, requiring several hours for correction after the cessation of CPB. Culliford and Spencer [5] advise irrigating the operative field to remove clotted blood and debris just before CPB. Blood loss (450 ± 135 mL versus 850 ± 250 mL; p < 0.001) and the blood transfusion (3.3 ± 1.2 U versus 5.8 ± 0.9 U; p = 0.034) rates were relatively less in group 1 patients. Decompressing the heart by CPB allows electrocautery dissection, which significantly reduces bleeding from vascularized dense adhesions. The blood in the operative field can be removed by a soft aspiration device. In our experience, these advantages of CPB have reduced the requirement of blood transfusion.
There was not any cardiac injury in group 1 (0 versus 6 in group 2; p = 0.029). The average CPB time was relatively longer in group 1 patients (93 ± 9 minutes versus 71 ± 11 minutes in group 2; p = 0.011). On the other hand, the total operation time was significantly reduced in group 1 (155 ± 23 minutes versus 185 ± 32 minutes; p = 0.024). The establishment of CPB just before sternotomy also has beneficial effects on the intensive care unit stay (2.2 ± 1.3 days versus 4.5 ± 2.3 days; p = 0.025), the hospital stay (7.3 ± 2.4 days versus 9.1 ± 3.1 days; p = 0.011), and the requirement of inotropic therapy (52 patients versus 76 patients; p = 0.032). The decompressed heart allows easy and rapid dissection so the total operation time is much less in group 1 patients. Myocardial function can also be preserved effectively in the decompressed heart.
The only limitation of femoral arterial cannulation is the arterial disease of the femoral and iliac arteries. We have used the brachial artery in 2 patients as Tademir and coworkers [10] described. The risk of retrograde dissection is extremely low because the arterial cannula and its obturator are slipped over the guidewire into the femoral arteries. Retrograde dissection of the aorta did not develop in our patients.
Cardiopulmonary bypass establishment by using the bicaval femoral venous cannula immediately before resternotomy not only allows adequate CPB flow but also significantly reduces the risk of cardiac injury and catastrophic hemorrhage and allows safe reopening. Although CPB time is increased, the operation time and bleeding and blood transfusion requirements are reduced by a statistically significant degree.
|
References
|
|---|
- Julian O.C., Lopez-Belio M., Dye W.S., Javid H., Grove W.J. The median sternal incision in intracardiac surgery with extracorporeal circulation: a general evaluation of its use in heart surgery. Surgery 1957;42:753-761.[Medline]
- Data Analysis of The Society of Thoracic Surgeons National Cardiac Surgery Database, January 1998
- Dobell A.R.C., Jain A.K. Catastrophic hemorrhage during redo sternotomy. Ann Thorac Surg 1984;37:273-278.[Abstract]
- Follis F.M., Pett S.B., Miller K.B., Wong R.S., Temes R.T., Wernly J.A. Catastrophic hemorrhage on sternal re-entry: still a dreaded complication?. Ann Thorac Surg 1999;68:2215-2219.[Abstract/Free Full Text]
- Culliford A.T., Spencer F.C. Guidelines for safety opening previous sternotomy incision. J Thorac Cardiovasc Surg 1979;78:633-638.[Abstract]
- Grunwald R.P. A technique for direct-vision sternal re-entry. Ann Thorac Surg 1985;40:521-522.[Abstract]
- Garret H.E., Jr, Matthews J. Reoperative median sternotomy. Ann Thorac Surg 1989;48:305.[Abstract]
- Akl B.F., Pett S.B., Wernly J.A. Use of sagittal oscillating saw for repeat sternotomy: a safer and simpler technique. Ann Thorac Surg 1984;38:646-647.[Abstract]
- Tull D., Albawn G.S. Survey research: a decisional approach. New York: Intext Press Inc, 1973:225-238.
- Tademir O., Sar
ta A., Küçüker 
., Özatik M.A., ener E. Aortic arch repair with right brachial artery perfusion. Ann Thorac Surg 2002;73:1837-1842.[Abstract/Free Full Text]
- Elami A., Laks H., Merin G. Technique for reoperative median sternotomy in the presence of a patent left internal mammary artery graft. J Card Surg 1994;9:123-127.[Medline]
- Kuralay E., Cingoz F., Gunay C., et al. Supraclavicular control of patent internal thoracic artery graft flow during aortic valve replacement. Ann Thorac Surg 2003;75:1422-1428.[Abstract/Free Full Text]
- Machiraju VR. Technical complications during redo cardiac surgery. The Third International Symposium on Redo Cardiac Surgery in Adults, New Orleans, LA, April 16–17. Symposium Program Abstracts 1999:17–8
- O'Brien M.F., Harrocks S., Clarke A., Garlick B., Barnett A.G. How to do safe sternal reentry and the risk factors of redo cardiac surgery: a 21-year review with zero major cardiac injury. J Card Surg 2002;17:3-13.
This article has been cited by other articles:
|
|
|
|
 
L. K. von Segesser, E. Ferrari, D. Delay, O. Maunz, J. Horisberger, and P. Tozzi
Routine use of self-expanding venous cannulas for cardiopulmonary bypass: benefits and pitfalls in 100 consecutive cases
Eur. J. Cardiothorac. Surg.,
September 1, 2008;
34(3):
635 - 640.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
N. Luciani, A. Anselmi, R. De Geest, L. Martinelli, M. Perisano, and G. Possati
Extracorporeal circulation by peripheral cannulation before redo sternotomy: Indications and results
J. Thorac. Cardiovasc. Surg.,
September 1, 2008;
136(3):
572 - 577.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
