Platelet-Mediated Transfer of Cardioprotection by Remote Ischemic Conditioning and Its Abrogation by Aspirin But Not by Ticagrelor

Despite substantial improvements of pharmacological and interventional strategies to treat patients with acute myocardial infarction, mortality remains high, with about 15% 1-year mortality in a recent large Scandinavian registry [1]. Thus, even in developed countries with optimal medical treatment and rapid initiation of reperfusion, there is still a need for cardioprotection. Although reperfusion is the only way to rescue myocardium at risk from myocardial infarction, reperfusion per se adds specific irreversible damage to the ischemic injury [2]. So far, cardioprotective strategies beyond that of rapid reperfusion have been developed in preclinical studies but not successfully translated to clinical benefit of patients [3].

Platelet activation exacerbates ischemia/reperfusion (I/R) injury: activated platelets reduce myocardial perfusion through thrombus formation, release vasoconstrictive factors, promote endothelial dysfunction, and trigger inflammation [4]. Pharmacological dual anti-platelet treatment, i.e., the combination of aspirin and a P2Y₁₂ inhibitor, is therefore a cornerstone of current therapy of patients undergoing elective or primary percutaneous coronary intervention (PCI) [5], particularly with the aim to prevent thrombosis at the implanted stent surface. However, apart from the deleterious role of activated platelets during myocardial I/R, platelets also exert cardioprotection. Platelets carry and release multiple cardioprotective factors, which can activate intracellular cardioprotective pathways [4, 6, 7]. Indeed, infusion of washed platelets in a subphysiological concentration into an isolated perfused rodent heart improved its functional recovery during reperfusion and reduced infarct size [8, 9]. The potential protective role of platelets during myocardial I/R is supported by a reverse experimental approach: platelet depletion, a most radical anti-platelet intervention, did not reduce infarct size in anesthetized dogs with coronary occlusion/reperfusion [10], possibly because both damage and protection were abrogated. The ability of platelets to store and release cardioprotective factors renders them a potential target for novel cardioprotective strategies [6].

One strategy to induce cardioprotection in acute myocardial infarction consists of brief cycles of I/R in tissues/organs remote from the heart, which then protect the myocardium from sustained I/R injury [3]. Such remote ischemic conditioning (RIC) is operative in all species tested so far and improved patients’ outcome in several but not in all studies [3, 11, 12]. Problems of RIC’s translation into the clinic have been attributed to an incomplete understanding of the complex underlying signal transfer [13] and confounders such as, age, sex, anesthetic regimen, comorbidities, and comedications [14, 15]. RIC is a systemic phenomenon, which requires a stimulus, e.g., I/R cycles in peripheral tissues/organs and a signal transfer to the target organ. The signal transfer of RIC from the periphery to the heart involves humoral and neuronal pathways [16]. Evidence for RIC’s humoral signal transfer was derived from preclinical experiments, where RIC’s cardioprotection was transferred with plasma or plasma-dialysate from conditioned donors to another individual’s isolated heart, which was subjected to ex vivo I/R [17‐20]. As a systemic phenomenon, RIC impacts also on circulating blood cells [6]. RIC attenuated platelet activation in patients with coronary artery disease after treadmill exercise [21] or after coronary angiography [22]. In patients undergoing PCI [23] or interventional ablation for atrial fibrillation [24], RIC attenuated platelet-monocyte aggregation [23, 24] and in vitro platelet aggregation in response to adenosine diphosphate (ADP) [24]. RIC increased occlusion time in an in vitro thrombosis test of blood samples taken from ST-segment elevation myocardial infarction patients 48 h after primary PCI [25], reflecting attenuated platelet aggregation. However, while RIC attenuates platelet activation, it is unclear whether or not RIC’s cardioprotective signal transfer involves platelets.

We therefore now studied whether or not platelets serve as transmitters of RIC’s cardioprotective signal. Venous blood was taken from healthy volunteers before and after RIC, respectively. Washed platelets or plasma-dialysates were infused into isolated perfused rat hearts before global I/R. To study the impact of commonly used platelet aggregation inhibitors, the same volunteers were again subjected to RIC after pretreatment with an oral loading dose of either aspirin or ticagrelor, respectively. In the isolated perfused rat hearts, infarct size served as the most robust endpoint of cardioprotection [26, 27].

Healthy volunteers were recruited, and experiments were performed between October 2020 and December 2021. The study was approved by the institutional ethical review board (No. 18–8279-BO) of the University of Essen Medical School and conforms to the Declaration of Helsinki. The experimental protocols conform to the guidelines from the Directive 2010/63/EU of the European Parliament on the protection of animals used for scientific purposes. We followed the ARIVE guidelines 2.0. Male Lewis rats (200–380 g, 2.0–3.5 months, Central Animal Laboratory, University of Duisburg-Essen, Essen, Germany) were used in the present study. The experimental protocols in isolated buffer-perfused rat hearts, the plasma-dialysate preparation [19], and the methods for the measurement of hemodynamics and quantification of infarct size were standard [26] and have been described in detail previously [20]. The preparation of venous blood samples and washed platelets was modified from a previous study [28]. Unless otherwise specified, materials were obtained from Sigma-Aldrich (Deisenhofen, Germany).

Our study supports the notion that platelets carry and release cardioprotective factor(s) which reduce rather than promote myocardial I/R injury [4, 6, 7]. Platelets contributed to the humoral signal transfer of RIC’s cardioprotection. RIC’s cardioprotection was transferred via washed platelets from healthy volunteers with RIC to isolated perfused rat hearts (Fig. 5). Different from previous studies, where infusion of platelets in a subphysiological concentration induced cardioprotection [8, 9], infusion of washed platelets in our study did not impact on infarct size per se. Subtle differences in the timing of administration and concentration of platelets may account for such difference. There is one prior study, indicating a potential role of platelets for RIC’s signal transfer and protection of the liver. In a mouse model with hepatic I/R, thrombocytopenia through infusion of anti-cluster of differentiation 41 immunoglobulin abrogated RIC’s protection [33], supporting a potential role for platelets in RIC’s signal transfer, but that study was possibly also confounded by other effects of thrombocytopenia [34]. As in our previous studies [19, 20, 35], RIC’s cardioprotection was also transferable with plasma-dialysate (Fig. 5). The observation that aspirin pretreatment abrogated RIC’s transfer of cardioprotection through platelets but not that through plasma-dialysate supports a platelet-specific component of RIC’s humoral transfer of cardioprotection (Fig. 5). However, we looked at the inhibition of cardioprotection by aspirin only in a qualitative/binary (yes/no) fashion. Although 500 mg aspirin effectively abrogated arachidonic acid-induced platelet aggregation in all volunteers, in six volunteers, only 1000 mg aspirin sufficiently abrogated RIC’s transfer of cardioprotection with platelets. Thus, it appears that there are quantitative differences between aspirin’s inhibition of platelet aggregation and inhibition of cardioprotection by platelets. Abrogation of RIC’s platelet-mediated transfer of cardioprotection by aspirin pretreatment in our study suggests a cyclooxygenase (Cox)-dependent signaling. Doses > 300 mg aspirin (we used 500/1000 mg) inhibit Cox-1 [36], which is predominant in platelets [37], but also Cox-2, which is induced in response to inflammatory stimuli in various tissues but also constitutively expressed in neuronal tissue. In platelets, prostaglandin G₂ formation is Cox-1 dependent, and prostaglandin G₂ serves as a common precursor for the potent platelet agonist thromboxane A₂ and for additional prostaglandins and prostacyclins. There is evidence for several prostaglandin species to mediate cardioprotection. Endogenous prostaglandins I₂ [38, 39], D₂ [40], and E₂ [41‐43] contributed to cardioprotection through the activation of prostaglandin receptors 3 and 4 in different species with I/R. To dissect which specific prostaglandin/prostaglandin receptor is involved in our setup would require use of selective pharmacological antagonists. The possible nonspecific nature of such agents [44] and their off-target effects [45], however, made such approach not feasible within the framework of our current study.

The constitutively expressed Cox-2 is relevant for spinal nociceptive signaling [46]. Of note, RIC’s signal transfer from the periphery to the target organ heart involves activation of peripheral sensory nerves, which project into autonomic centers of the central nervous system and consequently the vagal nerves [16, 20]. Activation of the vagal nervous system results in a release of humoral cardioprotective factors from the spleen [20] and other abdominal organs [16]. Thus, not only the platelet-dependent Cox signaling but also that in the periphery may have been affected by aspirin. However, again, in our study, aspirin pretreatment only abrogated the transfer of cardioprotection by platelets, while plasma-dialysate still mediated cardioprotection, and, consequently, a platelet-specific Cox-sensitive component of RIC’s cardioprotection is more likely. Also, aspirin in prior studies had abolished the cardioprotection by local ischemic preconditioning [47] and postconditioning [48] when given in vivo in rodent models with regional I/R. The attenuation of ischemic conditioning’s cardioprotection in these studies was attributed to the inhibition of myocardial Cox-2 activity [49]. In our study, any direct impact of aspirin on the rat myocardium and thus on infarct size seems unlikely, since platelets were repeatedly washed, resulting in an estimated dilution of hypothetical contaminating plasma by 10⁸ and since infusion of aspirin in saline control experiments had no impact on infarct size. Nevertheless, in vivo aspirin may not only impact on the platelet-mediated transfer of cardioprotection but also on the myocardial responsiveness to cardioprotective signaling. Taken together, it is virtually impossible to attribute the aspirin-mediated abrogation of cardioprotection by RIC exclusively to the inhibition of platelet-dependent Cox-1 or to inhibition of Cox-2 signaling in the periphery or the myocardium in our and the above studies.

Ticagrelor exerts cardioprotective properties per se [50, 51]. Our data support the cardioprotective potential of ticagrelor: the cardioprotective factor(s) recruited by ticagrelor pretreatment in vivo reduced infarct size in the isolated perfused heart per se. Platelets, and interestingly, also plasma-dialysate, prepared from the same volunteers after ticagrelor pretreatment, reduced infarct size in the isolated perfused rat heart. Whether or not platelets were the origin of such plasmatic cardioprotective factor(s), however, remains unclear. Nevertheless, a systemic administration of ticagrelor appears to be mandatory to exert its cardioprotective effect. Not only in the present study, but also in prior studies, ticagrelor did not impact on infarct size in isolated perfused rat hearts [51, 52]. In vivo, ticagrelor increases the systemic circulating adenosine levels [53, 54], which can activate cardioprotective pathways [3]. Such ticagrelor-mediated increase in systemic adenosine was attributed to the inhibition of the equilibrative nucleoside transporter-1, which results in a decreased adenosine uptake by erythrocytes [55] and thus increased local adenosine levels in ischemic tissues [50]. Such prolongation of adenosine’s half-life may subsequently enhance adenosine receptor-mediated cardioprotection [50]. Therefore, platelet-mediated — but also plasma-mediated — cardioprotection after in vivo ticagrelor pretreatment in our study could be related to a yet unknown effect of ticagrelor on circulating platelets, hypothetically impacting on adenosine metabolism within platelets and erythrocytes [56, 57]. In line with this notion, platelet-derived adenosine reduced infarct size in isolated perfused rat hearts [56]. However, ticagrelor pretreatment obviously did not impact on protection by RIC, since the level of infarct size reduction achieved with platelets or plasma-dialysate from blood after RIC was not different in the absence or presence of ticagrelor (Fig. 5). Interestingly, in an in vivo rat model of myocardial infarction, ticagrelor pretreatment was also not additive to the cardioprotection afforded by ischemic postconditioning — although cardioprotective per se [58]. The lack of an additive action of ticagrelor may reflect the use of the same cardioprotective signal transduction cascade by ticagrelor and RIC. Apart from and in addition to the above platelet-mediated cardioprotective pathways, platelets may also contribute to the humoral signal transfer of cardioprotection via released extracellular vesicles, which contain cardioprotective microRNAs [59, 60]. In response to RIC specifically, microRNA144-3p and microRNA451a were increased in the circulating extracellular vesicles of healthy humans [60].

Anti-platelet drugs have repeatedly been discussed as potential confounders of RIC’s cardioprotection in patients with acute myocardial infarction [15]. However, there is currently no study analyzing whether or not anti-platelet drugs interfere with RIC’s cardioprotection. In studies on RIC’s impact on the outcome of patients undergoing PCI for the treatment of acute ST-segment elevation myocardial infarction, virtually, all patients received a loading dose of aspirin in temporal relation to PCI [61]. Among those studies, however, there are some reporting cardioprotection by RIC, whereas others reported neutral results [61]. The largest multicenter study on RIC’s cardioprotection in patients with ST-segment elevation myocardial infarction, the CONDI-2/ERIC-PPCI trial, failed to demonstrate a reduction of infarct size or an improved outcome [62]. Recently, Ye et al. [63] hypothetically proposed that aspirin may attenuate cardioprotection by RIC. Our data now clearly demonstrate that indeed aspirin abrogates the platelet-mediated part of RIC’s cardioprotective signal. Thus, in clinical studies, differences between the placebo and RIC groups may have been obscured by aspirin. As outlined above, ticagrelor has pleiotropic cardioprotective effects beyond that of platelet inhibition and also reduces infarct size in patients [64]. Approximately 70% of all patients in both, the placebo and the RIC arm of the CONDI-2/ERIC-PPCI trial, received ticagrelor in temporal relation to PCI, and indeed, in a post hoc subgroup analysis of the CONDI-2 trial, ticagrelor-pretreated patients had an improved clinical outcome compared to those pretreated with clopidogrel or prasugrel [65]. Of note, in a previous study with a similar setting conducted in the pre-ticagrelor era, where 95–97% of patients still had received clopidogrel in temporal relation to PCI, RIC did reduce infarct size [66, 67]. The use of dual platelet inhibition with aspirin and ticagrelor in clinical trials on novel cardioprotective interventions may then impair cardioprotection in the treatment group by aspirin and recruit some cardioprotection per se in the placebo group by ticagrelor, such that the difference between the treatment and placebo group induced by the cardioprotective intervention under study is minimized — exactly that was probably true in the ERIC-PPCI/ CONDI 2 trial.

Platelets are not only a target of RIC, but also transmit RIC’s cardioprotective signal through a Cox-dependent mechanism. Despite the unequivocal detrimental impact of platelet activation during myocardial I/R injury, patients could benefit from a more targeted approach of anti-platelet drugs. Aspirin interferes with the transfer of cardioprotective factors through platelets. In contrast, ticagrelor induces the formation of cardioprotective factors which are carried by platelets and plasma. Therefore, in clinical trials, the use of dual platelet inhibition with aspirin and ticagrelor may obscure the effect of the cardioprotective intervention under study. Finally, to reduce the discrepancy between preclinical and clinical studies (i.e., the influence of comedications and/or comorbidities), the present study further highlights the need to develop animal models which more closely resemble the pathophysiological and pharmacological background of patients[69].