|
|
||||||||
a The Interventional Centre, Rikshospitalet, University of Oslo, Oslo, Norway
b Department of Cardiothoracic and Vascular Surgery, Rikshospitalet, University of Oslo, Oslo, Norway
c Department of Cardiology, Rikshospitalet, University of Oslo, Oslo, Norway
d Department of Radiology, Rikshospitalet, University of Oslo, Oslo, Norway
e Department of Faculty of Medicine, University of Oslo, Oslo, Norway
Accepted for publication August 13, 2008.
* Address correspondence to Dr Bergsland, The Interventional Center, Rikshospitalet, Oslo, 0027, Norway (Email: nielsb{at}aol.com).
| Abstract |
|---|
|
|
|---|
Methods: Fifty-six patients undergoing off-pump coronary artery bypass using the left internal mammary artery to bypass the left anterior descending coronary artery were randomly assigned to have the bypass performed with intracoronary shunt or by occlusive snaring. Ischemia during grafting was monitored by tissue Doppler. Hemodynamic status and indicators of ischemia were monitored, and on-table and postoperative angiography were performed.
Results: In patients with retrograde filling of the left anterior descending coronary artery, ischemia did not develop, but occlusion of antegradely perfused vessels caused ischemia in 26 of 33 patients. Ischemia was reversed in 14 of 16 shunted patients, and in 3 of 17 nonshunted cases (p = 0.004). Angiography showed a trend toward improved on-table angiographic results in shunted patients. After 3 months, graft patency was 100%, but 1 patient treated without shunt required reintervention and 15 patients had new angiographic lesions, equally distributed between shunted and nonshunted patients.
Conclusions: Intracoronary shunt prevents ischemia during grafting of the left anterior descending coronary artery and provides satisfactory immediate- and short-term graft patency.
| Introduction |
|---|
|
|
|---|
The aim of this study was to evaluate risks and benefits of intracoronary shunt during grafting of LAD.
| Material and Methods |
|---|
|
|
|---|
Selection and Randomization
Fifty-six patients scheduled for OPCABG were randomly assigned to shunt or no-shunt groups by block randomization. Patients with significant stenosis or occlusion of LAD were included. Left ventricular function was above 25% in all cases, and the patients were clinically stable at the time of operation. Preoperative data are shown in Table 1.
|
Surgery
Cardiac surgeons experienced in OPCABG performed the operations. Premedication was 5 to 10 mg diazepam by mouth. Anesthesia was induced with 2 to 5 µg/kg fentanyl, thiopentone 2 to 5 mg/kg, and 0.15 mg/kg cisatracurium. Repeated doses of fentanyl and sevoflurane 1.0% to 2.5% maintained anesthesia. Heparin was given to maintain activated clotting time above 250 s.
Surgery was performed using sternotomy, and the heart positioned by pericardial sutures and a suction cup device (Starfish; Medtronic, Minneapolis, MN). Regional stabilization was obtained with an Octopus suction stabilizer (Medtronic). Patients in the shunt group had a shunt inserted through the arteriotomy after placing proximal and distal silicon snares around the LAD, 0.5 cm from the intended arteriotomy. Axius (Boston Scientific, Santa Clara, CA [n = 17]) or Chase (Chase Medical, Richardson, TX [n = 7]) coronary shunts were used. Size and type of shunt were determined by the surgeon. Oversizing was avoided. The proximal snare was tightened for at least 1 minute in the shunt group to ease shunt insertion. In the no-shunt group occlusion of LAD was maintained during construction of the anastomosis. The distal snare was loosely applied to minimize retrograde flow from LAD in no-shunt patients. To facilitate visualization, a Clear-View mist CO2 blower (Medtronic) [12] and irrigation with saline were used. Anastomoses were sutured with 7-0 polypropylene suture. The time used to construct the LIMA anastomosis was about 10 minutes in both groups. Saphenous vein grafts were performed after completing the LIMA graft. Graft flows were controlled using transit time flow measurements [13].
Hemodynamic Management and Measurements
Arterial and central venous pressures were monitored. Cardiac output was obtained using pulse contour measurements [14–16]. Measurements were recorded and compared before occlusion of LAD, during grafting, and at reperfusion. Fluid boluses, and injection of vasopressors, vasodilators, or inotropic agents were avoided during LAD grafting if possible, and otherwise administered as necessary to maintain stable hemodynamics.
Transesophageal Echocardiography
A transesophageal probe providing tissue imaging (6.7 MHz) and Doppler measurements (3.5 MHz) [17] was utilized to obtain two-dimensional recordings of color-coded tissue velocities in four-chamber view with a System FiVe echocardiograph (GE Vingmed Ultrasound, Horten, Norway). Pulse repetition frequency was set to 0.5 to 1.0 kHz, with a frame rate of 80 ± 2 s–1, which was considered acceptable for strain measurements [17–19]. All recordings were performed in the same position in each patient with the ultrasound beam aligned close to the direction of the interventricular septum. Elevation of the heart was minimized during LAD grafting to optimize echocardiographic recordings. Recordings were stored digitally as cine loops for offline analysis. A typical recording used for strain measurements is shown in Figure 1.
|
) represents regional myocardial contraction fraction. Strain of a myocardial segment is defined as a percent change from end-diastolic dimension (L0) to instantaneous length (L), defined by the equation
= (L – L0/L0) x 100%, and was derived from tissue velocity measurements. A strain curve for a given region was generated throughout the cardiac cycle, and peak systolic strain obtained. Negative values described segmental shortening while positive values signified segmental lengthening. Strain values less than –10% were considered as normal and strain values greater than –10% represented ischemia. Measurements were performed at four time points: (1) baseline before LAD occlusion; (2) 1 minute after LAD occlusion (LAD was still occluded in both groups); (3) 5 minutes after LAD occlusion (shunt was in place in the shunt group while the LAD remained occluded in the nonshunt group); and (4) after removal of the shunt or release of the occlusion with open LIMA. Strain was measured in the mid and apical segments of the septum [17]. If strain was greater than -10% in any of these segments, ischemia was considered to have appeared.
On-Table Angiography
On-table angiography was performed under anesthesia by femoral route after chest closure [20], enabling on-table graft revision. The LIMA to LAD anastomosis was imaged in two planes. Findings were graded as described by FitzGibbon and coworkers [21], where grade A is used for a graft with excellent flow without obstruction, grade B means a graft with significant obstruction, and grade O means an occluded graft.
Postoperative Course
Patients underwent standard postoperative management including daily recordings of ECG, creatine kinase with isoenzymes (creatine kinase–myocardial band), C-reactive protein, aspartate aminotransferase, and alanine aminotransferase. Troponin was not measured routinely.
Clinical and Angiographic Follow-Up
Patients were examined by a clinician after 3 months and asked to report on the presence of angina or other cardiac symptoms. Follow-up angiograms were performed by radial approach. Two independent angiographers described the LIMA to LAD anastomosis without knowledge of treatment group. If stenosis was demonstrated on 3-month follow-up, the previous angiograms were examined. If a lesion was not seen on earlier angiography, the lesion was assumed to be new and possibly related to operative manipulations.
Statistical Analyses
The number of patients included was based on power analysis. Variables were described using mean and standard deviations for continuous variables and by counts for categorical variables. Continuous variables were compared using t tests. Pearson correlation coefficient was calculated for normally distributed continuous variables. Crude associations between categorical variables were assessed with
2 tests. Adjusted associations were modeled using logistic regression. As 3 patients in the shunt group did not receive a shunt owing to technical difficulties, two sets of analyses were performed: one for the "intention to treat" group and one with the group for which shunt was actually used. Numbers presented in the text refers to grouping according to actual use if not otherwise indicated. A p value less than 0.05 was considered statistically significant. Analyses were performed using SPSS version 13 (SPSS, Chicago, IL).
| Results |
|---|
|
|
|---|
Clinical and Biochemical Outcomes
No patient had a perioperative myocardial infarction. One patient in the shunt group had Q waves in anterior ECG leads without a rise in enzymes and a patent LIMA graft. One no-shunt patient had creatine kinase–myocardial band level above 75 IU without ECG changes and patent grafts at 3 months. Two patients in the no-shunt group continued to have angina postoperatively. One required percutanous coronary intervention of a stenotic LAD and another of the right coronary artery. There were no significant differences in creatine kinase, creatine kinase-myocardial band, C-reactive protein, aspartate aminotransferase, alanine aminotransferase, or troponin between study groups.
Septal Ischemia by Transesophageal Echocardiography
No patient was ischemic before LAD snaring. No ischemia developed in 10 patients with retrograde filling of an occluded LAD, independent of whether shunt was utilized. Twenty-six of 33 patients with antegrade flow in the LAD had septal ischemia during snaring. Patients randomly allocated to the shunt group had less ischemia than patients in the no-shunt group (p = 0.03). When patients were grouped according to actual use of shunt (Table 2), the difference was more significant (p = 0.004). Typical recordings of strain patterns in shunted and not shunted patients are shown in Figure 2. Logistic regression demonstrated that presence of retrograde flow was a significant factor in avoiding ischemia, before and during grafting. Use of shunt was a highly significant independent factor contributing to avoidance of ischemia during grafting (p = 0.001). Prevention of ischemia was independent of shunt size.
|
|
|
Postoperative Angiography
Fifty-three patients underwent postoperative angiography. Twenty-three lesions were demonstrated in the area of the LIMA to LAD anastomosis or in the native vessel close to the anastomosis. Fifteen of the lesions were not present on preoperative angiography (Table 3). The lesions were more common proximal to the anastomosis than distal: proximal 9 of 53; distal 2 of 53 (p = 0.03). There was no difference between the two groups. Average severity of the new lesions was 44% (SD 17%). Both distal lesions were retrospectively seen to be present on the on-table angiogram, as were 4 of 11 proximally located new lesions. Five new proximal lesions were not seen on on-table angiogram; in 2 cases, comparison between on-table and postoperative angiogram was not possible.
|
| Comment |
|---|
|
|
|---|
Ischemia in the LAD distribution may be demonstrated by measuring strain in the interventricular septum [17]. This study included patients with significant stenosis or total occlusion of the LAD. None of the hearts with totally occluded LAD developed ischemia during snaring, probably because of well-established collaterals [24] and the absence of stress, which may challenge the adequacy of collateral blood supply [25].
Most patients with antegrade flow in the LAD had ischemia when the LAD was occluded. Insertion of a shunt reversed ischemia except in 2 patients, whereas most snared patients remained ischemic. Distal flow in a shunted artery is dependent on proximal perfusion pressure and the size of the shunt [26]. The 2 patients in whom shunt did not reverse ischemia may have had inadequate flow in the shunt due to technical factors or inadequate proximal perfusion pressure.
Hemodynamic collapse [23] requiring conversion to CPB did not occur during grafting of the LAD. Such events are infrequent, and this study was not powered to answer whether shunting could prevent conversion [2]. Other investigators have demonstrated that shunting during LAD grafting can prevent temporary wall motion abnormalities [27–29], and case reports have documented reversal of hemodynamic instability by shunting. It has also been demonstrated that shunt may reduce troponin release [30], although we did not demonstrate any difference in troponin or other markers of myocardial damage. In our study, shunt was only used routinely during LAD grafting, and any benefit on biochemical markers may have been diluted by temporary occlusion of other vessels.
Opinions have been divided on whether the use of shunt is helpful in the creation of the anastomosis. The presence of a shunt may prevent the surgeon from taking too large arterial bites during suturing [8] and may prevent bleeding and improve visibility [7]. This study demonstrated a trend toward improved on-table angiographic results in the shunt group. All shunt patients had Fitzgibbon grade A anastomosis, whereas there were 4 grade B anastomosis in the no-shunt group. Except in 1 of these 4 cases, angiograms were normal at 3 months, indicating that the on-table angiographic changes may frequently be due to spasm [20]. The use of shunt may, potentially, denude the endothelium and cause intimal hyperplasia and stenosis [11, 31]. Angiograms performed on 53 patients 3 months after surgery showed 15 new lesions in the anastomotic area, not seen on the preoperative angiogram. Most lesions were located proximal to the anastomosis corresponding to the proximal snare. Snaring of an arteriosclerotic artery may cause endothelial damage, plaque rupture, and microthrombosis [31]. In this study, the proximal LAD was exposed to snaring in both groups, although in the shunt group, potential endothelial denudation from the shunt could occur as well [11]. The area distal to the anastomosis was snared only in no-shunt patients, but the snare was applied with less force than the proximal. There was only 1 new distal lesion in each group, both present at on-table angiography. This finding indicates that light snaring and use of intracoronary shunt is relatively safe, with a low rate of permanent damage [11]. The increased incidence of new, proximal lesions, several of them absent on both preoperative and on-table angiography, could indicate that occlusive snaring of the LAD may damage the vessel and cause late obstructive lesions.
There was a trend toward more anastomotic lesions in no-shunt patients, both at on-table angiography and postoperatively in the distal LIMA, indicating that shunt may be helpful in avoiding technical mistakes. Other investigators have also demonstrated equal or improved angiographic results in shunted patients [32]. The avoidance of endothelial damage is important in any type of vascular surgery. A modified shunt design may decrease the potential for damage [33]. It seems logical that the shunt should not be oversized even though undersizing may cause increased bleeding and less shunt flow [34]. We avoid oversizing and if necessary allow some bleeding, which usually decreases rapidly, possibly because of mild coronary spasm. We saw no correlation between the presence of ischemia during grafting and the size of the shunt. The amount of blood flow necessary to prevent ischemia may be low in the anesthetized patient and dependent not only on the presence of a shunt, but also on the proximal intracoronary pressure as well as the collateral circulation [34].
To conclude, we have found that intracoronary shunts prevent ischemia in the antegradely perfused LAD during LIMA grafting. The quality of the anastomosis is at least as good as when a shunt is not utilized. The relatively frequent occurrence of new coronary lesions proximal to the anastomosis may indicate damage from occlusive snaring. We recommend the use of intracoronary shunt during OPCABG surgery. The use of snaring of the native vessel should preferably be avoided.
| Acknowledgments |
|---|
|
|
|---|
| References |
|---|
|
|
|---|
Related Article
This article has been cited by other articles:
![]() |
T. Kofidis, M. Y. Emmert, H. G. Paeschke, L. S. Emmert, R. Zhang, and A. Haverich Long-term follow-up after minimal invasive direct coronary artery bypass grafting procedure: a multi-factorial retrospective analysis at 1000 patient-years Interactive CardioVascular and Thoracic Surgery, December 1, 2009; 9(6): 990 - 994. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. P. Collison and Z. S. Meherwal Intraluminal shunts at off-pump surgery: friend or foe? Ann. Thorac. Surg., May 1, 2009; 87(5): 1650 - 1651. [Full Text] [PDF] |
||||
![]() |
L. Aklog and B. J. deGuzman Invited Commentary Ann. Thorac. Surg., January 1, 2009; 87(1): 60 - 61. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |