Remote Ischemic Conditioning in Cardiovascular Surgery

Abstract

The existing clinical studies on remote ischemic preconditioning in patients undergoing cardiovascular surgery are critically reviewed, with a focus on infarct size reduction and clinical outcome as end points. Confounders, notably the use of propofol anesthesia are identified. The need for better designed trials with a more targeted approach is emphasized.

Remote conditioning is the systemic response to a variety of noxious stimuli (electrical, mechanical, chemical/pharmacological, ischemia/reperfusion) in a variety of different tissues and organs, which provides protection from ischemia/reperfusion injury to a variety of different target organs at a distance. ¹ The stimulus apparently always involves the activation of peripheral sensory nerves, the transfer of the protective signal to the target organs is through neuronal and humoral communication which also interact, and the protection of the target organs involves protective signaling pathways known from the local conditioning phenomena. ² More specifically, remote ischemic conditioning is elicited by ischemia/reperfusion of different tissues or organs and protects multiple target organs, including heart, brain, kidneys, etc from ischemia/reperfusion injury. Remote ischemic conditioning can be elicited before (preconditioning), during (perconditioning), or after (postconditioning) the index ischemia in the target organs. Remote ischemic conditioning has been observed in all species tested so far, including humans.

The original experimental study demonstrated in anesthetized dogs the protection of left anterior descending coronary artery–perfused anterior myocardium from infarction by prior brief cycles of ischemia/reperfusion in the left circumflex coronary artery–perfused posterior myocardium. ³ The original studies in humans demonstrated better preservation of endothelium-dependent vasodilation in the forearm by prior brief cycles of blood pressure cuff inflation/deflation on the contralateral arm in healthy volunteers. ^4,5 Myocardial protection by remote ischemic preconditioning was first evidenced by Hausenloy et al. in patients undergoing coronary artery bypass graft surgery; 3 cycles of 5-minute blood pressure cuff inflation/5-minute deflation after induction of anesthesia attenuated the cumulative postoperative release of troponin T. ⁶ Cardioprotection, that is, reduced biomarker release (creatine kinase-MB, troponin I or T) by remote ischemic conditioning in cardiosurgical patients was subsequently confirmed in a number of, but not in all, studies. ⁷ Of note, 2 studies even reported better clinical outcome as a secondary end point from cardiovascular surgery with remote ischemic preconditioning in the short- ⁸ or long-term, ⁹ whereas others found neither attenuated troponin release nor improved clinical outcome. ⁷

Two large phase III trials with clinical outcome as primary end point of protection by remote ischemic preconditioning were published last year in the New Engl J Med. In Effect of Remote Ischaemic preConditioning on clinical outcomes in patients undergoing Coronary Artery bypass graft surgery (ERICCA), 1612 patients undergoing coronary artery bypass graft surgery with or without additional valve surgery under cardiopulmonary bypass were randomized to sham or 4 cycles of 5-minute blood pressure cuff inflation/5-minute deflation on the upper arm after induction of anesthesia (involving propofol in >90% patients) and before skin incision. The combined primary end point of cardiovascular death, myocardial infarction, coronary revascularization, or stroke at 12 months after randomization was not significantly reduced. In patients with complete troponin data, the cumulative postoperative release of high-sensitivity troponin T was reduced, but in the entire cohort with imputed data for missing values it was not. The 6-minute walk test at 12 months was improved with remote ischemic preconditioning. ¹⁰

In the Remote Ischemic Preconditioning in Heart surgery (RIPHeart) trial, 1403 patients undergoing cardiovascular surgery under cardiopulmonary bypass were randomized to sham or remote ischemic preconditioning by 4 cycles of 5-minute blood pressure cuff inflation/5-minute deflation on the upper arm after induction of anesthesia (propofol per protocol). The primary composite end point of death, myocardial infarction, stroke, or acute renal failure up to the time of hospital discharge was not reduced. The postoperative troponin release (incomplete data sets for troponin T or I) was also not significantly reduced. Neither ERICCA nor RIPHeart reported any adverse events or raised safety concerns about remote ischemic preconditioning.

The two neutral phase III trials with a primary clinical end point came as a major disappointment to the remote ischemic conditioning field and also to the cardioprotection field at large. Accordingly, these 2 neutral phase III trials have stirred up intense discussion on the translation of cardioprotection from animal to human studies and from human proof-of-concept trials with surrogate end points to large clinical trials with outcome data. ^{7,11 ∓13}

In the following, I present my personal reflection of the current situation:

Cardiovascular Surgery—Clinical Need for Protection of Patients at High Risk Versus Surrogate Model for Acute Myocardial Infarction

As more and more patients of advanced age, with more comorbidities (ie, diabetes, heart failure, renal dysfunction, etc) undergo coronary artery bypass graft surgery without or with additional valve surgery, there is a clinical need for better protection, even beyond the excellent protection which current anesthetic protocols, notably opioids and volatile anesthetics which have also cardioprotective properties, and cardioplegia already provide. On the other hand, the controlled myocardial ischemia/reperfusion under cardiopulmonary bypass and cardioplegic arrest is taken as a surrogate for acute myocardial infarction where the timing and extent of coronary vascular occlusion and reperfusion are often not clear. It must be emphasized, however, that myocardial ischemia/reperfusion injury under cardioplegic arrest and in the presence of cardioprotective anesthesia is of far lesser magnitude than in acute myocardial infarction. Of note, all trials in patients with reperfused acute myocardial infarction have reported protection by remote ischemic perconditioning, as reflected by reduced release of biomarkers (creatine kinase-MB, troponin) or imaging (single photon emission computed tomography [SPECT], magnetic resonance imaging [MRI]), no matter whether reperfusion was by primary percutaneous coronary intervention (PCI) or thrombolysis. ⁷ Also, in the CONDI trial, clinical outcome on retrospective analysis was improved. ¹⁴ The dependence of evidence for protection on the magnitude of the underlying ischemia/reperfusion injury is supported by a retrospective analysis in which protection by remote ischemic preconditioning was more pronounced with longer aortic cross-clamp times in patients undergoing coronary artery bypass graft surgery. ¹⁵ In the future, it would be wise to direct research efforts and to establish the protective effects of remote ischemic perconditioning (RIPC) first in patients with acute myocardial infarction.

Anesthesia as a Confounder of Cardioprotection by Remote Ischemic Preconditioning

Both volatile anesthesia and propofol anesthesia are cardioprotective per se; however, volatile anesthesia is preferred in patients at greater risk of myocardial ischemia. ^16,17 In contrast, propofol appears to not impact on postoperative troponin I release in patients undergoing coronary artery bypass graft surgery without, but to specifically abrogate the attenuation of troponin I release in patients with remote ischemic preconditioning. ^{18 ∓20} Combined remote ischemic preconditioning with volatile anesthesia, but not with total intravenous anesthesia (propofol), reduced postoperative mortality in a meta-analysis. ¹⁷ It is mechanistically still unclear how propofol interferes with cardioprotection by remote ischemic conditioning, but it also abrogates the increase of myocardial STAT 5 phosphorylation ¹⁹ which is otherwise associated with cardioprotection by remote ischemic preconditioning. ²¹ Also, propofol may impair the sensory fiber activation ²² which is a requisite to elicit remote conditioning. Finally, propofol interferes with γ-aminobutyric acid–mediated central nervous control of cardiac vagal nerves, ^23,24 and cardiac vagal nerve activation may be important for the cardioprotection by remote ischemic conditioning. ^25,26

It is therefore unfortunate that both ERICCA and RIPHeart were performed with >90% of all patients under propofol anesthesia; both studies were designed before convincing data were available on the interference by propofol. Also, the use of propofol appears to be a major determinant of all cardiosurgical trials that reported no benefit from remote ischemic conditioning. ⁷ A large phase III trial without propofol, but with volatile anesthesia is therefore mandatory.

Trauma Versus Myocardial Ischemia/Reperfusion Injury During Cardiovascular Surgery

Remote conditioning protects from ischemia/reperfusion injury, but we lack evidence that it will protect against myocardial injury from direct surgical trauma. Obviously, additional valve surgery on top of coronary artery bypass graft surgery inflicts more traumatic injury. If the biomarker release which reflects such traumatic injury is not reduced by remote ischemic preconditioning, the attenuation of biomarker release from ischemia/reperfusion injury is diluted and more difficult to detect. The inclusion of many patients with additional valve surgery in ERICCA and RIPHeart, different from earlier smaller trials in patients undergoing only coronary artery bypass graft surgery, ^6,9 could contribute to the lack of remote ischemic preconditioning’s effect on postoperative troponin release.

Comorbidities and Comedications as Confounders of Protection by Remote Ischemic Conditioning

There is the awareness that experimental studies are mostly done in young and healthy animals and that lack of advanced age, comorbidities (hypertension, hypercholesterolemia, diabetes, chronic kidney disease, etc), and comedications (P2Y₁₂ inhibitors, statins, antidiabetics, beta-blockers, angiotensin converting enzyme [ACE] inhibitors, angiotensin receptor 1-antagonists, etc) may confound the translation of cardioprotection from experimental animal studies to clinical practice, apart from and in addition to species differences. ²⁷ These comorbidities and comedications can either interfere with the protective remote ischemic conditioning, for example, hypercholesterolemia ²⁸ or induce protection per se and leave little room for additional protection by remote ischemic conditioning. ²⁹ Indeed, our own trial on remote ischemic preconditioning had therefore excluded diabetics, ⁹ although on retrospective analysis we found no confounding effect of diabetes per se, but only of anti-diabetic sulfonylureas. ³⁰ In fact, there is currently no evidence of a major interference of risk factors and comedications with the efficacy of remote ischemic conditioning in patients undergoing coronary artery bypass graft surgery ¹⁵ or primary PCI for acute myocardial infarction ³¹ ; statins may even facilitate remote ischemic conditioning. ³¹

All-Comer Approach Versus Personalized Medicine

A large multicenter trial with an all-comer approach and wide inclusion criteria facilitates recruitment such that the trial is more quickly completed and results are available. However, if in fact the intervention under study targets only a specific cohort of individuals with more narrow inclusion criteria, then its effect is difficult to detect with an all-comer approach. In this context, the late Philip Poole-Wilson has coined the term “dilution to nonsense” and warned against an all-comer approach. Also, with the involvement of many centers the protocol adherence is more difficult to control, and the largely incomplete troponin data sets in ERICCA and RIPHeart are not encouraging in this respect.

In times of personalized medicine, a more tailored trial design is preferable: definitely avoidance of propofol, avoidance of concomitant valve surgery and its associated traumatic troponin release from which remote ischemic conditioning may not protect, and possibly also exclusion of diabetic patients under sulfonylurea treatment. If such more tailored trial design yielded positive results for more specified patients groups under volatile anesthesia, remote ischemic conditioning could still be used with an all-comer approach in clinical practice, since no adverse effects have been reported ^7,10,32 and the intervention is inexpensive and easily feasible. Even if only every tenth patient undergoing cardiovascular surgery had clinical benefit from remote ischemic conditioning, it would be worthwhile as long as the other patients had no disadvantage, and given the high number of cardiosurgical patients it would still reflect a major therapeutic progress.

Conclusion

From my perspective, it would be premature to give up on remote ischemic conditioning in cardiovascular surgery after ERICCA and RIPHeart. We need better, more focused trials that explicitly avoid propofol anesthesia and concomitant valve surgery and also observe all other established confounders, such as sulfonylureas ³⁰ and possibly nitrates. ^33,34 The potential benefit from remote ischemic conditioning is great as seen from all available studies in patients with acute myocardial infarction, ^7,35 and we must work hard to make such potential protection available for clinical practice in patients undergoing cardiovascular surgery. The protection from remote ischemic conditioning may relate not only to the heart but also to other organs suffering from cardiovascular surgery such as the kidneys, ^9,36 given that remote conditioning is a systemic response. However, eventually protection must be evident not only from surrogate end point but from clinical outcome. To provide firm evidence regarding improved clinical outcome, we will need at least 1 phase III trial that demonstrates reduced mortality (all-cause or at least cardiovascular) with RIPC, and this is most likely to be shown in patients with acute myocardial infarction; subsequently, studies with other, less robust end points, such as MACCE or hospitalization for heart failure, alone or in combination are certainly worthwhile in other patient cohorts, including patients undergoing cardiovascular surgery.