|
|
||||||||
Ann Thorac Surg 2003;75:S709-S714
© 2003 The Society of Thoracic Surgeons
a Department of Thoracic Surgery, Karolinska Hospital, Stockholm, Sweden,
b Crafoord Laboratory for Experimental Surgery, Karolinska Hospital, Stockholm, Sweden
* Address reprint requests to Dr Vaage, Department of Thoracic Surgery, Karolinska Hospital, 17176 Stockholm, Sweden.
e-mail: jarle.vaage{at}ks.se
Presented at the 3rd International Symposium on Myocardial Protection From Surgical Ischemic-Reperfusion Injury, Asheville, NC, June 26, 2002.
Abstract
Preconditioning is in experimental studies the most powerful mode of cardioprotection known. The signal transduction pathways involve a variety of trigger substances, mediators, receptors, and effectors. The studies of preconditioning in cardiac surgery provide conflicting results but the majority of studies show that ischemic preconditiong is an effective adjunct to myocardial protection. However, ischemic preconditioning with repeated clamping of the aorta will never get widespread use. If the "preconditioning reponse" is to be exploited in cardiac surgery, targeting the underlying molecular mechanisms must provide easily applicable techniques or drugs, which are shown in large scale clinical studies to be beneficial.
Short episodes of ischemia and reperfusion before a sustained ischemic event, termed ischemic preconditioning, reduces infarct size and improves cardiac function [1]. Preconditioning is protective either when the preconditioning takes place less than 2 hours before the sustained ischemia (termed classic preconditioning), or when the sustained ischemia is 24 to 72 hours after the preconditioning episode (termed delayed preconditioning). Preconditioning can be targeted at the heart itself or at other organs such as kidney or intestine to protect the heart or the organ itself (remote preconditioning). There is an abundance of literature, including reviews on classic preconditioning [24]. The literature is more scarce concerning delayed [5, 6] and remote [7] preconditioning.
Overview of possible mechanisms
Despite more than a decade of research trying to determine the mechanisms of preconditioning there is still a lot left to be discovered before the signal transduction pathways are unraveled. The mechanisms of preconditioning probably vary between acute and delayed models and between local versus remote. The initial phase of preconditioning may be similar between classic and delayed models and may start by release of a trigger substance during the brief episodes of ischemia and reperfusion that may bind to surface receptors coupled to Gi proteins [811]. Among these substances are adenosine, bradykinin, cathecholamines, opioids, and acetylcholine [811]. Other trigger substances have been identified as prostanoids, nitric oxide, and low doses of reactive oxygen intermediates [811]. Recently a possible role for innate immunity in triggering preconditioning has been suggested where specially tumor necrosis factor alpha is indicated as a trigger of the preconditioning response [12]. The trigger substances may cause activa-tion of kinase cascades where translocation of protein kinase C, especially the
isoform, from the cytosolic to the particulate fraction may be crucial to the response [13, 14]. Tyrosine kinases as well mitogen activated protein kinases appear involved in the signaling cascade in several species, although which kinase cascade is upstream or downstream is currently an issue of debate [15, 16]. Adenosine signaling has been linked to protein kinase C-dependent opening of mitochondrial ATP-sensitive potassium channels (KATP) [17]. The KATP channels are suggested to be the end-effectors of myocardial protection in classic models; however, their role is controversial as most data in the field are derived from the use of one pharmacologic blocker (glibenclamide), and the results are discrepant [16, 18]. Recent data indicate that activation of the KATP channel causes release of reactive oxygen species [11] and one may speculate that their role in preconditioning is linked to this. The next step in the signaling pathway is activation of transcription factors by protein kinases, reactive oxygen species, and nitric oxide, of which particularly nuclear factor
-B (NF
B) has been investigated in both classic and delayed models [19, 20]. NF
B is a redox sensitive transcription factor that regulates transcription of a battery of genes most of which are associated with proinflammatory effects such as cytokines, chemokines, and leukocyte adhesion molecules [21]. Some candidate genes for organ protection in preconditioning are also regulated by NF
B [21]. Pharmacologic blocking of NF
B translocation inhibits preconditioning in both classic and delayed models [19, 20].
The importance of transcription factors is easy to grasp in models of delayed preconditioning in which the time frame for translation of protective substances is ample. However, the fact that transcription factors appear important for protection in classic models is less easy to grasp. The cardioprotective effect of NF
B in immediate models as previously shown [18, 22] may be due to an upregulation of its own inhibitor I
B
[23] thus contributing to reducing a NF
B-dependent inflammatory response during sustained ischemia. However, a recent paper abolishing classic preconditioning effects through employing the transcriptional inhibitor actinomycin D indicates that the process is complex [24]. Other endogenous cardioprotective substances that are suggested as upregulated and protective are heat shock proteins of the 70 kDA [25] or 27 kDa [26] families. These have their own regulatory factors but do influence the activation of NF
B [21]. Antioxidants [27], inducible nitric oxide synthase [28], and inducible cyclooxygenase [29], all regulated by NFkB, have furthermore been indicated as organ effectors of protection. A last possible route of NF
B mediated cardioprotection is through the antiapoptotic effect of preconditioning. NF
B regulates several antiapoptotic molecules, such as Bcl-2 [19, 30], survivin [31], and inhibitor of apoptosis protein-1 [32], as well as X-linked inhibitor of apoptosis protein-1 [33]. A schematic presentation of the possible signaling pathways of preconditioning is presented in Figure 1.
|
The phenomenon of so-called remote preconditioning, that preconditioning of a region or an organ will protect neigboring tissues or even remote organs, was first suggested by Przyklenk and colleagues [34] when they found that a brief circumflex artery occlusion will precondition the area of the myocardium perfused by the left anterior descending artery. Subsequent animal studies have shown that ischemia of many different organs such as kidney, mesenterium, intestine and brain can also protect the heart [3538]. At the same time all organs tested demonstrate the presence of the preconditioning response. Furthermore, remote preconditioning of the brain protects not only the heart [38] but also preserves endothelial function of aortic rings in vitro [38]. Preconditioning hearts of patients with valvular disease undergoing open heart surgery protects lung morphology and function post bypass [39].
Unravelling of remote preconditioning effects has come shorter than cardiac models, but appears in a similar way to involve adenosine, bradykinin, nitric oxide, protein kinase C, and inducible nitric oxide synthase [3538, 40, 41]. The apparent general effect of ischemic preconditioning may be mediated by humoral or neurogenic factors or through substances in blood released from the preconditioned organ. Future research will determine transmission routes and components.
The perspective of preconditioning research is to find the underlying mechanism of action. Detailed knowledge of molecular mechanisms of cell protection provide the possibility to produce novel generations of cytoprotective drugs. The implications may be far beyond cardioprotection and cardiac surgery, for instance the endogenous celldefense may be increased during all kinds of operations.
Why precondition or not precondition in cardiac surgery
Preconditioning is in experimental studies the most powerful technique of myocardial protection available. However, the use and possible efficacy of preconditioning in patients during cardiac surgery has not been studied in details, although the first report is from 1993 [42]. There are many reasons for this: (1) Most surgeons who do not use intermittent cross-clamping as myocardial protection have a psychological antagonism against repeated episodes of cross-clamping the aorta, both due to the "unprotected" ischemia and the possible chances of embolism from the ascending aorta. (2) Ischemic preconditioning may prolong the surgical procedure by 15 to 30 minutes. (3) Because several of the drugs used for premedication and anesthesia may in themselves induce a preconditioning response, it is assumed that there is "nothing more to gain." (4) The cardiopulmonary bypass per se induces preconditioning [43], again, nothing more to gain. (5) Experimentally there are conflicting results about the possible protective effect of preconditioning when combined with hypothermia or cardioplegia or both [44, 45] so cold cardioplegia may leave nothing more to gain. (6) The ideal model of ischemic preconditioning in humans, namely the length and number of the preconditioning cycles, length of reperfusion, and so forth, is unknown. (7) There are conflicting animal studies whether the protection caused by preconditioning is reduced in the presence of arteriosclerosisnothing more to gain? (8) In an older patient population undergoing cardiac surgery, preconditioning may not be effective because the preconditioning response declines with increasing age [46]nothing more to gain? (9) There are many problems concerning endpoints and evaluation of clinical studies on myocardial protection. The primary clinical endpoints such as mortality and morbidity need large scale, preferably multicenter studies in order to ensure statistical power. Some studies have used biochemical parameters such as ATP levels in tissue or the early release of biochemical markers of cardiomyocyte injury. However, the clinical significance of such indicators are not clarified. Interestingly central hemodynamics have recently been shown by the group in Tampere, Finland, to be a reasonably sensitive indicator for the evaluation of preconditioning (see below). The classic endpoint in experimental studies is infarct size, which is not applicable to the cardiac surgery setting.
Consequently, taken together, ischemic preconditioning in the operating room is cumbersome and may potentially have little or no effect. At the same time there is increasing evidence that preconditioning is indeed effective in humans. The first reports came from patients undergoing angioplasty [47]. Later on the so-called "warm-up" phenomenon was described, that unstable angina preceding a myocardial infarction preconditions the myocardium causing reduced mortality, smaller infarct size, and less complicated in-hospital course than after infarctions without warm-up from unstable angina [48, 49].
Ischemic preconditioning during cardiac surgery
The initial studies by Yellon and associates [42] used intermittent crossclamping as myocardial protection and they found better maintained ATP levels in the preconditioned patients. Later the same group found less release of cardiac troponin T in this model [50]. However, the majority of surgeons use cardioplegic arrest as the cardioprotective principle. The first study during cardiac surgery with cardioplegia was performed with warm blood cardioplegia and the investigators found that preconditioning increased the early release of creatine kinase MB and the transmyocardial lactate gradient shifted toward production, thus preconditioning may increase the injury [51]. This study initiated several studies, some of which found a postive effect of preconditioning whereas others either found no effect [52] or even a worsened postoperative outcome indicated by increased inotropic support [53]. Recently it was again discussed that during coronary artery bypass surgery perhaps there is nothing more to gain beyond the preconditioning response triggered by the cardiopulmonary bypass per se [54]. Cardiopulmonary bypass seems to trigger preconditioning both through alpha adrenergic and adenosine-1 receptor stimulation [43]. Furthermore it also activates the kinase cascades, which is mechanistically linked to opening of potassium channels, to the same extent as the anesthetic sevoflurane, which has a known preconditioning effect [54]. Additionally, cardiopulmonary bypass induces a general inflammatory response involving generation of reactive oxygen species and increase of cytokines such as tumor necrosis factor alpha [55] that may be triggers of the preconditioning response [12].
However, most clinical studies have concluded that ischemic preconditioning does improve myocardial protection when combined with blood or crystalloid, warm or cold cardioplegia. Illes and Swoyer [56] found improved cardiac function and less inotropic support after preconditioning. In patients undergoing valve surgery preconditioning was protective [57, 58]. Szmagala and associates [59] found reduced release of cardiac troponin T in preconditioned patients undergoing coronary artery bypass grafting. Recently ischemic preconditioning was found to provide superior myocardial protection compared with pharmacologic preconditioning with an adenosine A1 receptor agonist and to standard intermittent cross-clamping [60].
It has been questioned whether diabetes, hypercholesterolemia, and arteriosclerosis make hearts with chronic ischemic disease difficult to precondition [7]. These states all cause deterioration of endothelial function. However, the endothelium is probably an underestimated target for the preconditioning response [61]. In experimental studies rapid pacing induced myocardial protection in hearts of rabbits fed high cholesterol for 8 weeks [62] but not in hearts of rats fed high cholesterol for 24 weeks [63]. In a recent study we explored whether ischemic preconditioning could protect hearts of mice with severe arteriosclerosis (apolipoprotein E/LDL receptor double knockouts fed an atherogenic diet for 7 to 9 months before the experiments) and found that the isolated hearts of these mice had a larger benefit from ischemic preconditioning than hearts of mice with normal coronary arteries [64]. Furthermore we found that hearts of severely atherosclerotic mice with spontaneous infarctions in vivo in hearts or brains were protected against induced global ischemia ex vivo, supporting the notion that adaptation to ischemia may be a naturally occurring phenomenon in species with arteriosclerosis causing serious ischemia [38]. Thus arteriosclerosis per se does not appear to be a factor abolishing the preconditioning effect.
The cardiac surgery group in Tampere, Finland, has performed several well-conducted, prospective, randomized studies on ischemic preconditioning in patients undergoing coronary bypass surgery with cold blood cardioplegia. They have found evidence that preconditioning improves both left and right ventricular function in the early postoperative period [65, 66]. However, preconditioning had no clear cut effect on the whole body inflammatory response to cardiopulmonary bypass as demonstrated by release of cytokines and free radical production [65, 67]. In a nicely conducted study they also showed that in older patients (> 68 years) the preconditioning response was not evident as in younger patients (< 68 years) [68]. The same group found that patients with unstable angina during the last 48 hours before surgery were preconditioned, they had improved hemodynamics in the early postoperative period [69]. However, in this study patients with low ejection fraction and those who had had a recent myocardial infarction were excluded. In another study this group discussed the possibility that free radicals act as triggers of preconditioning in the cardiac surgery scenario [70].
Several different preconditioning protocols have been employed. One cannot presume that the optimal preconditioning model in animal experiments can be directly transferred to cardiac surgery. Firstly there is no general agreement on the optimal preconditioning model in the experimental laboratory. Secondly species differences may play a role. Furthermore the preconditioning model used in regional ischemia with variable collateral flow at normothermia may have completely different effects in a situation with cardioplegic arrest and global ischemia where different temperatures and ways of administration (antegrade versus retrograde) may be used. Finally the anesthetic drugs used and the cardioplegic techniques used may profoundly influence the observed effects of preconditioning in cardiac surgery. So far the only model that has been used in a series of studies and consistently proved to be beneficial is the regimen used by the Finnish group in Tampere: to cross-clamp the aorta 2 minutes and then reperfuse for 3 minutes [6568]. This cycle is repeated once at normothermia before cardioplegic arrest is induced by intermittent cold blood cardioplegia without any enrichment. Before declamping a hotshot was given. Cardiopulmonary bypass was run at 32°C. In contrast two studies that either did not find any effect of preconditioning [52] or found that preconditioning worsened the injury [53] used completely different endpoints as compared with the Tampere group. Kaukoranta and associates [52] used continuous normothermic blood cardioplegia given retrogradely and biochemical endpoints such as lactate release, ATP levels in the myocardium, and release of biochemical markers of myocardial injury. The cardioplegia was enriched with glutamate and aspartate. The preconditioning model was 5 minutes of ischemia followed by 5 minutes of reperfusion. Cremer and associates [53] preconditioned by a protocol with 5 minutes of ischemia followed by 10 minutes of reperfusion. Cardioplegia was antegrade, cold intermittent blood cardioplegia according to Buckbergs protocol. The endpoints were release of markers of myocardial injury and the postoperative need of inotropic support. However, for the time being there is no robust explanation of the cause of the divergent findings concerning preconditioning in cardiac surgery. One cannot conclude that differences in myocardial protection or differences in preconditioning protocol may explain the inconsistant protection found in clinical studies. Other variables such as the anesthetic agents, differences in patient population, enrichment of the cardioplegia or insensitive endpoints may be additional factors.
Usable surgical preconditioning models
We are still far from "the pill the day before surgery" and while we work in our laboratories to unravel mechanisms to find the perfect agent, patients are still in need of treatment. The substance investigated in most detail in cardiac surgery is adenosine or adenosine A1 receptor agonists. The first clinical report was by Lee and associates [71] who found that pretreatment with adenosine improved myocardial protection. The beneficial effect of adenosine pretreatment in patients has subsequently been confirmed by others [7274]. Teoh and associates [60] found however that an adenosine A1 receptor agonists induced weaker myocardial protection than ischemic preconditioning in patients with intermittent cross clamping, while other investigators do not find convincing beneficial effects of adenosine in cardiac surgery [75]. All things considered adenosine A1 receptor stimulation appears a clinically acceptable way to achieve myocardial protection.
Recently we have performed animal experiments showing that a minor oxidative stress elicited by breathing hyperoxic gas for 30 to 60 minutes before harvesting and perfusing isolated rat and mouse hearts caused a preconditioning-like myocardial protection [76]. Pretreatment with hyperoxia induced both early and delayed myocardial protection but the extent of the response was species dependent [76, 77]. In rats exposure for 60 minutes increased the level of conjugated dienes and reduced the antioxidative capacity in serum, suggesting that oxidative stress had occurred [76]. Protection was evident both as a reduction of reperfusion arrhythmias and infarct size as well as functional improvement [76]. In animals pretreated with hyperoxia in vivo, in vitro reactivity of isolated vessels was influenced as well [78]. Short-term hyperoxia seemed to protect the heart by a mechanism dependent of the transcription factor NF
B [23]. In rats hyperoxia activated pulmonary and myocardial NF
B and pretreatment with NF
B inhibitors before hyperoxia abolished the functional and infarct-limiting protection of hyperoxia in isolated rat hearts. The signaling also involves phosphorylation of mitogen activated kinases and nitric oxide, as inducible nitric oxide synthase deficient mice could not be preconditioned by this method [79]. Regardless of the possible mechanisms, breathing gas with high oxygen content for an hour is a clinically acceptable, inexpensive, and potentially whole-body preconditioning concept.
Conclusions
The role of preconditioning in cardiac surgery is still uncertain. If preconditioning is to become an important and widespread technique of myocardial protection, it is necessary in large-scale studies to provide strong evidence of its beneficial effects. Preconditioning by aortic clamping will never gain widespread acceptance as a routine technique. It is necessary to find simple, pharmacologic ways of inducing preconditioning such as breathing hyperoxic gas for 1 hour or to have an effective drug without side effects. The concept of whole-body preconditioning has important implications far beyond myocardial protection and cardiac surgery; it is general organ protection in any serious stressful or traumatic situation.
Acknowledgments
Our research related to preconditioning has been supported by the Swedish Research Council (Grants 11235 and 12665), the Swedish Heart-Lung Foundation (several grants to Drs Vaage and Valen), the Foundations Fredrik o Ingrid Thuring, Tore Nilsson,
ke Wiberg, the Laerdahl Foundation for Acute Medicine, Sigurd and Elsa Goljes Memory, AGA Gas, Gösta Franckels Foundation, King Gustav Vs and Queen Victorias Foundation, Karolinska Institutet, and the Karolinska Hospital.
References
is responsible for the protection of preconditioning in rabbit cardiomyocytes. J Mol Cell Cardiol 1999;31:1937-1948.[Medline]
B-dependent mechanism in the rat heart. J Thorac Cardiovasc Surg 2002. In press
This article has been cited by other articles:
![]() |
R. A. Kloner and S. H. Rezkalla Preconditioning, postconditioning and their application to clinical cardiology Cardiovasc Res, May 1, 2006; 70(2): 297 - 307. [Abstract] [Full Text] [PDF] |
||||
![]() |
C.-H. Huang, J.-S. Wang, S.-C. Chiang, Y.-Y. Wang, S.-T. Lai, and Z.-C. Weng Brief pressure overload of the left ventricle preconditions rabbit myocardium against infarction Ann. Thorac. Surg., August 1, 2004; 78(2): 628 - 633. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Galagudza, D. Kurapeev, S. Minasian, G. Valen, and J. Vaage Ischemic postconditioning: brief ischemia during reperfusion converts persistent ventricular fibrillation into regular rhythm Eur. J. Cardiothorac. Surg., June 1, 2004; 25(6): 1006 - 1010. [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 |