| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J Thorac Cardiovasc Surg 2004;127:1703-1712
© 2004 The American Association for Thoracic Surgery
Cardiopulmonary support and physiology |
a Department of Cardiology, Internal Medicine II, University of Vienna, Vienna, Austria
b Department of Medical Computer Sciences, University of Vienna, Vienna, Austria
c Department of Cardiothoracic Surgery, University of Vienna, Vienna, Austria
Received for publication November 20, 2003; revisions received December 22, 2003; accepted for publication January 28, 2004.
* Address for reprints: Werner Mohl, MD, PhD, Department of Cardiothoracic Surgery, University of Vienna, Waehringer Guertel 18-20, A-1090 Vienna, Austria
werner.mohl{at}univie.ac.at
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionReferences |
|---|
METHODS: A Medline search was performed to review the published literature on intermittent coronary sinus occlusion. The study inclusion criterion was a randomized, placebo-controlled trial with area of infarction (expressed as a percentage of the area at risk) as the primary end point.
RESULTS: Seven experimental trials comprising 125 test animals were found that analyzed the effects of intermittent coronary sinus occlusion on ischemic damage during coronary occlusion. A further 5 studies comprising 88 animals were designed to evaluate the effect of intermittent coronary sinus occlusion in combination with retroperfusion of arterial blood on the infarct size. A meta-analysis of the 7 studies analyzing the effect of intermittent coronary sinus occlusion revealed a significant reduction in infarct size of 29.3% in the treatment group compared with that in the placebo group (P < .001; 95% confidence interval, –40.9 to –17.7). A meta-analysis of the 5 trials analyzing the effect of intermittent coronary sinus occlusion in combination with retroperfusion revealed a reduction in infarct size of 39.4% in the treatment group compared with that in the placebo group (P < .001; 95% confidence interval, –48.9 to –29.9). Comparison between intermittent coronary sinus occlusion and intermittent coronary sinus occlusion in combination with retroperfusion of arterial blood showed no statistical difference (P = .19). An inverse relationship between achieved coronary sinus pressure increase per minute and infarct size could be found in the intermittent coronary sinus occlusion group (r = –0.92; P < .007), whereas in combination with retroperfusion, there was a negative correlation both between achieved coronary sinus pressure and the amount of the retroperfusate and myocardial salvage (r = –0.97; P < .004).
CONCLUSIONS: The use of intermittent coronary sinus occlusion and intermittent coronary sinus occlusion in combination with retroperfusion of arterial blood significantly decreases ischemic damage during coronary occlusions. Intermittent coronary sinus occlusion in combination with retroperfusion exhibits no significant profit in salvaging the ischemic myocardium in comparison with that provided by intermittent coronary sinus occlusion alone.
| See related editorial on page 1549.
|
The concept of arterialization of the cardiac venous system to increase the flow of oxygenated blood to ischemic myocardial tissue was suggested more than 100 years ago1 and has since been modified with the development of new technologies. Initial attempts of permanent retroperfusion of the coronary sinus by means of anastomosis to an arterial vessel were soon truncated because of ultrastructural damage of the wall of the coronary sinus caused by drainage disruption.2,3 Moreover, with the birth of the coronary artery bypass grafting era and the further development of percutaneous transluminal coronary angioplasty, the concept of selective retroperfusion was discarded. Attempts to introduce coronary sinus interventions clinically by interventional cardiologists also have been abandoned.
The advent and current praxis of off-pump revascularization and its potential benefits through avoidance of extracorporeal circuits again draws interest to myocardial protection through the coronary sinus. It is well understood that occlusion of a coronary vessel might create regional ischemia and subsequently cause an inflammatory response.4 This might be one of the reasons why the clinical benefits of this procedure are still difficult to detect.5
Several retroperfusion methods have been advocated to ameliorate myocardial ischemia during off-pump surgery.6,7 Lazar and colleagues8,9 also suggested that the simple method of transient coronary venous pressure increase caused by coronary sinus occlusion might at least be equivalent to that seen with retroperfusion techniques. There is ample evidence that coronary sinus intervention salvages ischemic myocardium during myocardial infarction.10-12
Although intermittent coronary sinus occlusion (ICSO), pressure-controlled intermittent coronary sinus occlusion (PICSO), synchronized retrograde perfusion (SRP), and simplified retroperfusion (SR) have all been described to be effective in salvaging ischemic myocardium in several experimental models of coronary artery occlusion,13,14 the mode of action of these interventions remained speculative. The common denominator of all coronary sinus interventions is the temporal increase in coronary venous pressure through temporal occlusion of the outflow, retroperfusion of arterial blood, or both. ICSO is a simple method that occludes the coronary sinus intermittently through inflation and deflation of a balloon-tipped catheter that is positioned in the orifice of the coronary sinus. This results in a redistribution of coronary sinus blood flow within the venous compartment to the ischemic myocardium through changes in pressure gradients throughout the coronary venous system.15 The balloon catheter is connected to a pneumatic pump that automatically increases and decreases pressure in the coronary venous system according to a preset cycle. The SRP and SR systems are modified ICSO techniques that actively pump arterial (SRP) or venous (SR) blood into the coronary sinus and have been used clinically for myocardial protection during high-risk angioplasty.16,17 Systems using selective retroperfusion and synchronized suction of the oxygenated blood represent an extended version of the SRP technique.18 Further developments include the left ventricle to coronary sinus shunt, which is a left ventricle–powered system that provides a graded systolic retroperfusion of the coronary sinus.7 Experimental and clinical studies have previously demonstrated all these techniques to significantly increase blood flow to the ischemic myocardium and thus reduce infarct size.7,13,19,20
The aim of this study was to systemically review published evidence of coronary sinus interventions in the medical literature and to perform a meta-analysis of available randomized trials to assess the efficacy of ICSO-PICSO and ICSO in combination with retroperfusion of arterial blood as myocardial salvage in terms of reduced infarct size. Furthermore, we sought to perform a comparison between both procedures and to draw pathophysiologic conclusions. This evidence-based analysis could be the basis for matching an old and almost forgotten but attractive concept of myocardial protection and the new techniques of off-pump surgery, thus clarifying the clinical benefit of both procedures.
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Two authors independently evaluated all 100 relevant reports identified by the searches. The study inclusion criterion was a randomized, placebo-controlled trial with area of infarction (expressed as a percentage of the area at risk) as the primary end point. Studies with application of ICSO without other techniques except for retroperfusion of arterial blood were included in the analysis. Reports of ICSO and ICSO in combination with retroperfusion of arterial blood were handled separately for a separate analysis. All information retrieved from the reports were tabulated by one reviewer and subsequently confirmed by a second reviewer.
Data extraction
The following information was assembled for each study: (1) publication data, first author's last name, and year of publication; (2) study design; (3) characteristics of treated versus placebo group (including number of subjects in each group); and (4) target parameter as area of infarction (expressed as a percentage of the area at risk), including mean values and either SD or SEM. All reported SEMs were converted to SDs as follows:
 | (1) |
Statistical analysis
All similar studies were pooled by using weighted mean differences, with 95% confidence intervals on the basis of normal distribution. Heterogeneity of study results was tested by estimating the between-study variance (BSV) by using restricted maximum likelihood and testing the null hypothesis of a BSV of 0; a P value of less than .05 was considered as indicating the use of random-effects models. Comparison between the weighted mean differences of ICSO and ICSO in combination with retroperfusion of arterial blood was performed by using a z test.
The linear regression test of Egger and associates21 was used to test the null hypothesis of the absence of publication bias. This test performs a linear regression of the standardized effect estimates (mean difference/SE) on SEs of all trials. The estimated intercept parameter is then used to measure asymmetry, with a significant difference from 0 being assumed to indicate publication bias. A negative intercept is considered to indicate smaller studies to be associated with bigger effects.
To determine the range of results that would be obtained with varying plausible assumptions, we evaluated the sensitivity of the weighted results on the basis of (1) omission of particular studies and (2) varying the assumptions on BSV. By using the jackknife technique,22 the analysis was repeated multiple times, with the data from a single study removed each time from the data for the entire group of studies. This allowed us to determine the influence of each single study on the pooled result. In a second sensitivity analysis, we used the Thompson plot,23 which indicates the effect of the choice of statistical method on the analysis. To construct a Thompson plot, the analysis is repeated by changing the percentage of weight allocated to each trial from allocating equal weight to each trial, which corresponds to assuming infinite BSV, to weighting each study with the inverse of its variance, corresponding to assuming zero BSV. Although the y-axis of this plot corresponds to the pooled mean difference, the x-axis represents the following: Assumed BSV/(Assumed BSV + Estimated BSV), and thus the value 0 of the x-axis corresponds to a fixed effects analysis, and the value 1 correlates to equal weighting of all studies. The Spearman nonparametric correlation coefficient was computed from the study means. SAS System Version 8.2 (SAS Institute Inc, Cary, NC) was used for statistical analysis.
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
|
|
The meta-analysis of the 7 studies analyzing the effect of ICSO on the infarct size revealed a significant reduction in infarct size of 29.2% in the treatment group compared with that in the placebo group (weighted mean difference between the ICSO and control groups, –29.3; 95% confidence interval, –40.9 to –17.7; P < .0001; Table 3). The test of heterogeneity indicated a significant difference of BSV from zero (P < .0001), thus implying the use of a random-effects model. The linear regression test of Egger and associates21 revealed a negative yet not significant intercept term in the linear regression test for publication bias (intercept, –13.2; P = .16). Thus the linear regression test shows a tendency of publication bias and overestimation of the pooled mean difference; however, this trend is not significant. The range of weighted mean differences assessed by using the jackknife sensitivity analysis is calculated as –33.2 to –26.2 (weighted mean difference between the ICSO and control groups remained significant for all jackknife calculations, P < .001). The Thompson plot showed some dependence of the weighted mean difference on the used weighting scheme (smaller effect estimated in a fixed-effects analysis with a weighted mean difference of –24.8; however, the significance between the ICSO and control groups remains, P < .001; Figure 1).
|
|
|
A correlation between achieved coronary sinus pressure per minute (ratio between coronary sinus occlusion and release) and infarct size showed a negative association in the ICSO group (r = –0.70; P < .7). Significance could be reached after eliminating the results reported by Zalewski and colleagues,27 which apparently do not fit into the overall reported spectrum of evidence (r = –0.92; P < .007).
In contrast to the relationship of pressure increase and salvage potential in the ICSO group, ICSO in combination with retroperfusion resulted both in a negative correlation between achieved coronary sinus pressure and the amount of the retroperfusate and myocardial salvage (r = –0.97; P < .004).
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Coronary sinus interventions are known to salvage ischemic myocardium, especially by limiting the infarct from its borders. The primary goal of all these approaches is not only to restore blood flow in the occluded vessels but also to salvage the damaged myocardium. In this study we systematically reviewed available data on coronary sinus interventions and performed 2 separate meta-analyses on the efficacy of ICSO alone and ICSO in combination with retroperfusion of arterial blood in reducing the area of infarction during acute ischemia assessed as a percentage of area at risk. We found an overall positive treatment response of ICSO in both analyses, with a reduction of 29.3% in animals undergoing ICSO versus a reduction of 39.4% in animals undergoing ICSO in combination with retroperfusion, both in comparison with the control group. Although both interventions significantly reduce infarct size and the fact that no further significance can be reached between the two methods, the available statistical data do not allow a further discrimination and are not helpful in determining which of the methods is superior.
The efficacy of ICSO in salvaging the ischemic myocardium during acute ischemia has been a matter of controversy for many years. Whereas the mechanism of retrograde coronary sinus perfusion of arterial blood and its positive effect on reducing the infarct size generally seem to be easily understood, doubts still remain regarding how a pressure increase in the coronary venous bed induced by ICSO alone might reduce the area of necrosis without any additional oxygen supply. In addition, ICSO summarizes a cohort of different methods, including variations in the timing of occlusions and the achieved pressures.
On the other hand, in interventions with retroperfusion of arterial blood, one would expect a diffusion of oxygen with subsequent salvage in a Krogh-like pattern from large veins throughout the infarct zone. This, however, does not seem to be the case because in both types of intervention, the pattern of myocardial salvage seems to be the reduction of the infarct zone from its borders. It is known that occlusion of the coronary sinus increases the coronary sinus pressure to levels greater than the distal coronary artery pressure, thus establishing a gradient for venous flow from the coronary sinus to the ischemic region when the coronary sinus is being occluded. It is well known that a desaturation of blood in the ligated coronary artery occurs according to a venoarterial pressure gradient and reversal of flow at the microcirculatory level. Pressure forces within the ischemic vascular bed caused by alternating flow over the ischemic territory also influence perfusion of the ischemic region from nonischemic myocardium and result in delivery of small amounts of oxygen through collaterals and further enhance the removal of toxic metabolites in the deflation interval.28 Bayer and associates29 showed that the coronary sinus pressure was an important determinant of desaturation in the occluded artery, supporting the hypothesis that the degree of perfusion achieved with ICSO depends on coronary sinus pressure. Thus because of the pressure gradient, the blood is squeezed by the healthy myocardium into the relatively empty coronary venous system in the ischemic zone. The sudden release of the coronary sinus occlusion results in a rapid drainage of the excessive volume not only from the well-perfused zones but also from the microvasculature recruited during the balloon occlusion phase, allowing an optimal washout of metabolites.
Zalewski and colleagues27 reported on the effects of various modalities of myocardial protection devices on the extent of myocardial salvage in an experimental model, with 7 different groups undergoing different protection techniques, including retrograde perfusion and ICSO. They concluded that only arterialization of the cardiac venous system, and not retroperfusion with venous blood or pressure-controlled intermittent occlusion of the great cardiac vein, was capable of reducing the size of the infarction. However, their analysis also revealed higher rates of complications in terms of myocardial hemorrhage in animals undergoing high-flow retroperfusion, which was explained as a consequence of the impaired drainage from the retroperfused region of the myocardium. This potential side effect of retrograde perfusion has also been previously described by other authors.3,30,31 Furthermore, on the basis of xenon-133 injections in the great cardiac veins of dogs subjected to arterial blood retroperfusion in the presence of left anterior descending coronary artery occlusion, Cohen and coworkers32 describe that only 10% of the injected isotope actually reaches the myocardium distal to the coronary occlusion site, suggesting that the delivery of oxygenated blood might be insufficient as an additional nutritional support to the ischemic myocardium. Thus the question arises of whether the beneficial effects found in several trials with active retroperfusion might have been related to the high coronary sinus pressures achieved as a result of retroperfusion and therefore were caused by an increase of outflow impedance and not necessarily by the substantial delivery of oxygen into the ischemic bed.
In our meta-analysis we found that although retroperfusion with arterial blood resulted in 10.2% greater salvage than intermittent occlusion of the coronary sinus, this difference was not significant. Although, as stated above, the small sample size does not allow a definite answer, pathophysiologic considerations might imply that an arterial blood source might not be necessary to protect jeopardized myocardium, especially in consideration of the fact that active retroperfusion is associated with higher rates of complications caused by uncontrolled impedance to venous outflow.
Mohl and coworkers20 describe the use of coronary sinus pressure as an important parameter to optimize occlusion-release cycles of the venous effluence. The coronary artery flow during coronary sinus occlusion is reduced according to the pressure increase in the venous circulation, while the plateau of the systolic pressure increase is reached. This reduction of flow is overpaid by a hyperemic response–like flow increment. Mohl and coworkers20 have documented that an occlusion of the coronary sinus through inflation of the balloon for 15 ± 4 seconds generates a plateau in coronary sinus pressure. An inflation time of 10 seconds to obtain this plateau has also been reported previously.33 In contrast, Feindel and associates34 describe an occlusion of the coronary sinus for only 5 seconds, which by no means is sufficient to reach a pressure plateau and might thus explain why their experiments revealed no favorable results for ICSO (achieved coronary sinus pressure during coronary sinus occlusion in trials of Feindel and associates34: 17.1 ± 6.4 mm Hg; achieved coronary sinus pressure during coronary sinus occlusion in trials of Mohl and coworkers20: 44.0 ± 7.0 mm Hg). This is about the same pressure Feindel and associates34 reached in their retroperfusion studies resulting in significant salvage. In contrast, Aldea and colleagues,13 who achieved about the same coronary sinus pressure with both methods, also reported similar salvage in both groups. The smaller amount of salvage in some of the ICSO studies therefore might be explained by the insufficient filling of the coronary venous system to allow redistribution of blood in significant quantities from the nonischemic myocardium to the ischemic zone, thus preventing an optimal recruitment of collaterals.
The negative association between remaining infarct size and the achieved coronary sinus pressure in the ICSO group suggest the potential of the transient pressurization of the venous compartment in inducing myocardial salvage, although results have to be interpreted with caution because of the lack of raw data. Furthermore, the results of the negative correlation between achieved coronary sinus pressure and myocardial salvage during the addition of active retroperfusion can be interpreted as a potential hazard of this method. This can be seen also in the relationship to the amount of retroperfusate.
The seeming contradiction of the relationship between pressure and salvage potential comes as a surprise only at a first glance because the pressure in the ICSO group is only the consequence of myocardial contraction on the venous compartment during coronary sinus occlusion, whereas the pressure observed in the ICSO plus retroperfusion group is the consequence of summing the intramyocardial pressure acting on the venous compartment and the external pumping force of the retroperfusate. It is therefore no surprise that the potential hazard of myocardial hemorrhage can be seen in active retroperfusion.27 It seems that 2 factors act in opposition: first, the ability of the pressure increase to induce myocardial salvage, and second, the improper timing of outflow impedance leading to the engorgement of the coronary microcirculation, thus reducing positive results by inducing severe changes of coronary artery flow, myocardial hemorrhage, edema formation, and subsequent decrease of myocardial performance.
Our own experience with coronary sinus interventions in a clinical setting applying PICSO during the early reperfusion period showed improvements also in regional myocardial function.35 PICSO was applied in 15 patients with 3-vessel disease undergoing on-pump coronary artery bypass grafting for about 60 minutes during the reperfusion period, and the results were compared with those of 15 control subjects. A significant improvement of regional function in hypokinetic segments (P < .04 treated vs control subjects) was found. In addition, severely hypokinetic segments significantly improved during PICSO therapy from prebypass values (P < .01), whereas in control subjects these changes were not significant. With this relatively simple method using a balloon catheter in the coronary sinus without additional blood supply, one can expect a reduction in myocardial jeopardy during beating-heart surgery.
In conclusion, this meta-analysis demonstrates that myocardial protection through the coronary sinus has an efficacy in reducing the size of experimental myocardial infarction irrespective of the time of ischemia and subsequent revascularization procedures. Active retroperfusion, with its known difficulties, resulted in an insignificant further reduction of the experimental infarct size in comparison with intermittent occlusion of the coronary sinus alone. In addition, these results show the importance of a physiologic adaptation of the coronary venous pressure increase, as well as the amount of the retroperfusate, to optimize the salvage potential and to minimize the consequences of the potential hazard of myocardial engorgement.
Evaluation of the effects on transient ventricularization of the venous vasculature as a consequence of all known coronary sinus interventions seems to be the logical next step to improve our understanding on the salvage potential of these methods.
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  G. S. Kassab, J. A. Navia, K. March, and J. S. Choy Coronary venous retroperfusion: an old concept, a new approach J Appl Physiol, May 1, 2008; 104(5): 1266 - 1272. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Weigel, I. Kajgana, H. Bergmeister, G. Riedl, H.-D. Glogar, M. Gyongyosi, S. Blasnig, G. Heinze, and W. Mohl Beck and back: A paradigm change in coronary sinus interventions--pulsatile stretch on intact coronary venous endothelium J. Thorac. Cardiovasc. Surg., June 1, 2007; 133(6): 1581 - 1587. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Banai, S. Ben Muvhar, K. H. Parikh, A. Medina, H. Sievert, A. Seth, J. Tsehori, Y. Paz, A. Sheinfeld, and G. Keren Coronary sinus reducer stent for the treatment of chronic refractory angina pectoris: a prospective, open-label, multicenter, safety feasibility first-in-man study. J. Am. Coll. Cardiol., May 1, 2007; 49(17): 1783 - 1789. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Mohl, I. Kajgana, H. Bergmeister, and F. Rattay Intermittent pressure elevation of the coronary venous system as a method to protect ischemic myocardium Interactive CardioVascular and Thoracic Surgery, February 1, 2005; 4(1): 66 - 69. [Abstract] [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 |
