Heart disease is the primary cause of cardiovascular mortality (Virani et al., 2021) in the United States for men, women, and people of most racial and ethnic groups (Centers for Disease Control and Prevention, 2018). Coronary artery disease is the most common type of heart disease, killing 360,900 people in 2019 (Virani et al., 2021). Every year, more than 800,000 Americans suffer an acute myocardial infarction (AMI) (Fryar et al., 2012), more than 600,000 corresponding to a first episode (Fryar et al., 2012). ST-segment-elevation myocardial infarction (STEMI) is the most severe form of AMI, causing significant morbidity and mortality (Vogel et al., 2019), and imposing heavy clinical and financial burden on the American healthcare system. Primary percutaneous coronary intervention (PCI) and thrombolytic therapy are effective in treating STEMI (Welch et al., 2012; Frampton et al., 2020; Valgimigli and Gragnano, 2020), reducing hospitalizations and long-term mortality. Many patients, however, experience long-term cardiac reperfusion issues due to lethal arrhythmias, myocardial necrosis, microvascular dysfunction (no reflow phenomenon), stunning myocardium, and heart failure (Verma et al., 2002; Papageorgiou et al., 2018), as a result of ischemia reperfusion injury (RI) (Yellon and Hausenloy, 2007; Hausenloy and Yellon, 2013; Davidson et al., 2019) and distal embolization (De Maria et al., 2013). Hence, there is a need for practical modalities of treatment to limit the extent of RI without first requiring opening the arterial obstruction.

Unlike the obstructed coronary arterial system, the coronary venous system remains healthy and unobstructed, and thus has enormous potential for therapy delivery (Katti and Patil, 2012; Rodenberg et al., 2019), including retrograde delivery of arterial blood (Kassab et al., 1985). Although envisioned a century ago (Pratt, 1898; Beck and Stanton, 1948), therapeutic retroperfusion (a type of coronary sinus intervention), has not been adopted clinically because complicated equipment has been required to regulate perfusion to prevent damage to the coronary venous system when exposed to arterial pressures. Synchronized retrograde perfusion and pressure-controlled intermittent coronary sinus occlusion are retroperfusion methods used to treat myocardial ischemia. These techniques, combined with a balloon-tipped catheter, redirect coronary sinus blood or pump arterial blood during diastole to the ischemic myocardium. They reduce infarct size, myocardial hemorrhage, and no-reflow phenomenon (Mohl, 1988; Mohl et al., 1988; Wakida et al., 1993), and improve left ventricular (LV) function (Yamazaki et al., 1985). Their widespread use, however, has been limited due to safety concerns and the need for repeated occlusion of the coronary sinus (Aldea et al., 1996; Lazar, 2004). Our group has been investigating for several years the use of retroperfusion (selective autoretroperfusion, SARP) for ischemic myocardium without the need of complicated apparatuses to regulate pressure in the cardiac venous system (Choy et al., 2011; Choy et al., 2020). We have demonstrated that SARP, alone or in combination with mild-hypothermia was able to confer cardiac protection following occlusion of a coronary artery by reducing infarct size in a swine model of anterior left ventricular (LV) myocardial infarction (Choy et al., 2020). The present study demonstrates that SARP can be accomplished safely with a catheter-based, pressure-regulated (∼60 mmHg, using the animal’s own pulse pressure) method of selective (left ventricular anterior wall) delivery of arterial blood to the ischemic myocardium to substantially reduce the deleterious effects of reperfusion. This true preconditioning of the myocardium for a brief duration is hypothesized to substantially reduce RI. We demonstrated the beneficial effects of preconditioning STEMI with SARP for 10 min before opening the arterial obstruction to combat RI. The results presented here, make SARP a very appealing modality of treatment for mitigation of RI.

This study investigated the efficacy of short-term treatment of STEMI with selective autoretroperfusion (SARP) before coronary artery reperfusion. SARP mitigated the deleterious effects of reperfusion injury (RI), preserving cardiac function and reducing myocardial infarct size, suggesting a potential washout mode of action.

Table 1 summarizes the left ventricular diameter and wall thickness in end-systole and end-diastole, measured from echocardiographic images during baseline, ischemia, retroperfusion, and 4 weeks reperfusion. End-systolic wall thickness was significantly larger in the SARP-b group (0.84 ± 0.04 cm) than in the SARP-nb (0.71 ± 0.04 cm, p < 0.05) and control (0.70 ± 0.03, p < 0.05) groups during retroperfusion. After 4 weeks of reperfusion, end-systolic wall thickness was similar to baseline in the SARP-b group (0.79 ± 0.04) and larger than in the SARP-nb (0.67 ± 0.02, p < 0.05) and control (0.59 ± 0.03, p < 0.01) groups.

Left ventricular global and segmental longitudinal strains are summarized in Tables 2, 3, respectively. The global strain values in all three groups were similar at baseline (SARP-b = −21.4% ± 0.8%, SARP-nb = −22.6% ± 1.2%, Control = −21.7% ± 0.8%; p > 0.05) and ischemia (SARP-b = −13.5% ± 1.1%, SARP-nb = −13.6% ± 2.3%, Control = −12.7% ± 1.4%; p > 0.05). In the SARP-b group, the global strain decreased ∼40% during ischemia, p < 0.001, but recovered to similar baseline values during retroperfusion (−19.4% ± 0.9%, p > 0.05) and 4-weeks reperfusion (−19.1% ± 0.7%, p > 0.05). In the SARP-nb group, the global strain also decreased ∼40% during ischemia, p < 0.05 and recovered to baseline values during retroperfusion (−19.6% ± 1.7%, p > 0.05) and 4-weeks reperfusion (−20.6% ± 1.2%, p > 0.05). In the control group, however, the strain decreased ∼40% during the first half of ischemia, p < 0.001, continued decreasing during the second half (−11.7% ± 0.9%, p < 0.001), and remained almost at the same levels during 4-weeks retroperfusion (−13.1% ± 1.2%, p < 0.001). Table 3 shows the segmental longitudinal strain at basal, middle, and apical interventricular septum, and basal, middle, and apical anterolateral wall in all three groups.

Left ventricular function, assessed by measuring ejection fraction (LVEF) with echocardiography, is shown in Figure 1. There was significant interaction between Group (SARP-b, SARP-nb, and Control) and Time (baseline, ischemia, retroperfusion, and 4 weeks), (F_{(4.5, 45.4)} = 9.4, p < 0.001), therefore simple main effects at each time point were analyzed. There was no significant difference between the three groups at baseline (F_{(2, 20)} = 3.3, p > 0.05) and during ischemia (F_{(2, 20)} = 0.2, p = NS), but there were significant differences in LVEF during retroperfusion (F_{(2, 20)} = 26.7, p < 0.001) and at 4 weeks (F_{(2, 20)} = 11.5, p < 0.001). At retroperfusion, Bonferroni post hoc tests showed that both the SARP-b (50.5% ± 1.3%) and SARP-nb (47.7% ± 1.9%) groups had significantly higher LVEF than the control group (39.6% ± 0.8%, p < 0.001 for both comparisons), and there was no statistically significant difference in LVEF between the SARP-b and SARP-nb groups. At 4 weeks, Bonferroni post hoc tests showed that both the SARP-b (53.2% ± 2.2%) and the SARP-nb (54.9% ± 1.7%) groups had significantly higher LVEF than the control group (45.4% ± 1.1%, p < 0.01 for both comparisons), and there was no statistically significant difference in LVEF between the SARP-b and SARP-nb groups.

Figure 2 shows cardiac troponin (cTnI) values at different time points in all three cohorts. There were significant differences within each group over time. There was no significant interaction, however, between Group and Time. The levels of cTnI were similar (<0.03 ng/mL) in all three groups at baseline. They increased to 0.28 ± 0.09 ng/mL during ischemia-retroperfusion in the SARP-b group, 0.30 ± 0.12 ng/mL in the SARP-nb group, and 1.41 ± 0.79 ng/mL in the control group. Although cTnI in the control group was higher than in the SARP-b and SARP-nb groups, the values in the control group during ischemia-retroperfusion were not significantly different relative to baseline due to intragroup variability. During reperfusion, the values in all three groups continued to increase, 0.88 ± 0.31 ng/mL in the SARP-b group, 2.48 ± 0.76 ng/mL in the SARP-nb group, and 2.74 ± 0.78 ng/mL in the control.

Figure 3 shows the values of several analytes like pH (Figure 3A), glucose (Figure 3B), lactate (Figure 3C), sodium (Figure 3D), potassium (Figure 3E), and calcium (Figure 3F) measured every 2 min from effluent blood samples obtained through the lumen of the PTCA balloon catheter that was occluding the LAD artery. The values of the corresponding analytes were similar in both groups at baseline (p = NS). As shown in the plots (Figure 3), the values in the SARP group progressively got closer to baseline values as retroperfusion progressed from minute 62 to minute 70 while the LAD artery was still occluded. There was significant interaction between Group and Time for lactate (F_(1.6,16.1) = 6.2, p < 0.05), therefore simple main effects at each time point were analyzed. There was no significant difference between the SARP-b and SARP-nb groups at baseline (F_(1,10) = 0.1, p > 0.05), but there were significant differences in lactate levels at minutes 62 (F_(1,10) = 8, p < 0.05), 64 (F_(1,10) = 10, p < 0.01), 66 (F_(1,10) = 8, p < 0.05), 68 (F_(1,10) = 7, p < 0.05), and 70 (F_(1,10) = 8, p < 0.05). All the other analytes showed significant differences within each group over time but no interaction between the two groups.

The area of infarction, expressed as the percentage of the area at risk, is shown in Figure 4 (upper panel, A). The infarct area in the SARP-b group was significantly smaller than in the control group (10.4% ± 4.8% vs. 34.7% ± 10.5%, p < 0.05). Similarly, the infarct area in the SARP-nb group was significantly smaller compared to the control group (18.0% ± 3.8%, p < 0.05). There was no significant difference between the SARP-b and SARP-nb groups. The area at risk, expressed as the percentage of the total LV area, is also shown in Figure 4 (upper panel, B). The area at risk was similar in all three groups, SARP-b, SARP-nb, and control (26.0% ± 1.9%, 24.7% ± 2.4%, and 25.3% ± 0.7%, respectively, p > 0.05). The lower panel shows representative images of myocardial slices double-stained with Evans blue and triphenyl tetrazolium chloride, demonstrating the area of infarction (white), the area at risk (red), and the non-affected area (blue) in the A) SARP-b, B) SARP-nb, and C) control groups. The area of infarction extended throughout the full thickness of the myocardial wall in the control group, and to a lesser degree in the SARP-nb group, whereas in the SARP-b group it was scattered throughout the wall.

Figure 5 shows histological images taken from the area at risk in representative animals from the SARP-b (Figures 5A, D), SARP-nb (Figures 5B, E), and control (Figures 5C, F) groups. The blue area is the myocardial infarction stained with Masson’s Trichrome. The fibrotic area in the SARP-b group (9.4% ± 1.4%) was significantly lower than in the SARP-nb group (34.2% ± 5.3%, p < 0.05) and the control group (44.6% ± 1.7%, p < 0.01). There was no statistically significant difference in fibrotic area between the SARP-nb and the control groups.

Our model simulations indicate that SARP significantly enhances the clearance or “washout” — of metabolites from ischemic regions, channeling the tracers into the cardiac chambers via Thebesian veins (Supplementary Figures S2–S5). This washout mechanism helps to remove potentially toxic substances that can accumulate during ischemia, thereby mitigating cellular damage. Moreover, the model simulations show that oxygenated blood from the femoral artery can be redirected back to the capillaries within the ischemic region (Supplementary Figures S1, S2).

The data from the present study in a STEMI swine model demonstrates that selective autoretroperfusion (SARP) preserves cardiac function and reduces myocardial infarct size significantly, possibly through a washout effect. The LVEF in the SARP-b group was >50% during retroperfusion and at the end of the study, compared to the control group with LVEF ∼40%. The area of infarction in the SARP-b group was ∼10% compared to ∼35% in the control group. Furthermore, and perhaps the most interesting finding of the study, was the progressive reduction in the concentration of lactate and other analytes during 10 min of retroperfusion, which points out to a washout effect of SARP.

Novel modalities of treatment to limit ischemia reperfusion injury (RI) in patients with acute myocardial infarction are urgently needed to combat morbidity and mortality following STEMI (Fröhlich et al., 2013; Bulluck et al., 2016; Li et al., 2019). In the present study, we validated the efficacy of short-term selective autoretroperfusion (SARP), which was implemented using our novel SARP catheter via the great cardiac vein in a setting of impending myocardial infarction, as an intervention against RI prior to complete restoration of blood flow through the arterial system. This preconditioning technique, which uses the animal’s own pulse pressure without the need of synchronized pumps, resulted in significant reduction of myocardial infarct size possibly through a washout effect of metabolic byproducts like H+ and lactate accumulating in the ischemic tissue during the occlusion.

Ischemic preconditioning is a protective technique that provides the myocardium with short periods of nonlethal ischemia and reperfusion before the ischemic insult (Murry et al., 1986). This protective mechanism has been demonstrated in several animal species and also in humans and may be part of some naturally occurring ischemic syndromes (Schott et al., 1990). It also enhances tolerance to lethal cell injury, protects against other end points of ischemia-reperfusion injury, and decreases apoptosis, a mechanism that can contribute to cell death after myocardial ischemia and reperfusion (Iliodromitis et al., 2007). Although we applied traditional preconditioning to all three groups of animals, which may have conferred certain beneficial effects, the area of infarction in the control group was significantly larger than in the SARP-b and SARP-nb groups. More importantly, unlike SARP, traditional preconditioning cannot be applied to STEMI patients once the arterial occlusion has occurred.

Coronary venous retroperfusion has been used in the past (Pratt, 1898; Beck and Stanton, 1948) to treat the ischemic myocardium. Due to the significant harm, however, caused by persistent application of arterial pressure to the venous circulatory bed (Hammond et al., 1967), this method of therapy was largely abandoned (Syeda et al., 2004). SARP, our proposed method of retroperfusion, is differentiated by the use of pressure-controlled delivery of arterial blood to the ischemic myocardial region just prior to PCI and for a brief period of time. This study demonstrated the benefits of a catheter-based, pressure-regulated (∼60 mmHg) approach of SARP to supply arterial blood to the ischemic zone without the use of synchronized pumps. Furthermore, a short-term administration of SARP in a swine model of anterior LV myocardial infarction, was sufficient to significantly reduce infarct size. Although the area of infarction with 10 min of retroperfusion was larger than the area of infarction with 30 min of retroperfusion [data from our previous study (4.7% ± 1.6%)]²⁶, the values were not significantly different (p > 0.05).

Retroperfusion pressure (∼60 mmHg) as well as retroperfusion flow are two paramount factors that impact the reduction of myocardial edema and hemorrhage, and oxygen delivery to the ischemic territory, respectively. The efficacy of SARP relies in the ability to control the increase in pressure in the perfused venous territory preserving at the same time coronary artery inflow and cardiac venous drainage. SARP is possible because there exist many inter-venous connections (Kassab et al., 1985). In our study, pressure in the GCV during retroperfusion was 63.9 ± 4.6 mmHg in the SARP-b group. The LAD artery occlusion led to a significant decrease in arterial pressure, necessitating the use of norepinephrine for optimal retroperfusion pressure. In the SARP-nb group, the retroperfusion pressure was 39.2 ± 8.6 mmHg but since the balloon of the SARP catheter was deflated, the pressure was not transmitted towards the venous territory in the LV anterior wall. Additionally, controlled increased of pressure in the venous system increases collateral flow towards the ischemic region (Sato et al., 1996), due to the interconnections between the coronary sinus, the venous plexus and the Thebesian vessels (Kassab et al., 1985).

Diverse mechanisms of cardiac necrosis after ischemia-reperfusion injury that have been demonstrated include, reactive oxygen species, hemorrhage, and neutrophil accumulation (Kalogeris et al., 2012), microvascular spasm, edema, calcium overload, and mitochondrial dysfunction (Papageorgiou et al., 2018). The mechanisms behind the beneficial effects of retroperfusion, however, are not totally understood. Besides redistribution of flow as a mechanism of action for retroperfusion (De Maria et al., 2016), an increased washout of potentially hazardous byproducts from the microcirculation in the ischemic myocardium, seems to play a key role against reperfusion injury. Although the washout mechanism has been proposed previously (Chang et al., 1987; Mohl et al., 2018), to our knowledge, this is the first time that measurement of effluent blood samples corroborates the effect through SARP (Figure 3). The amount of several analytes measured during short-term retroperfusion when the LAD artery was still occluded, gradually returned to close-to-baseline levels, which prevented injury to the myocardium when reperfusion occurred. The ultimate beneficial effect of SARP was a substantial reduction in myocardial infarct size without myocardial injury from either reperfusion or SARP itself. Again, it is worth noting that effluent blood was not present in the control group, which supports the claim that blood from the femoral artery is redirected back to the capillaries and the occluded arteries within the ischemic territory during retroperfusion.

During acute myocardial infarction, dysfunction of myocardial cell membrane occurs, which leads to edema and leakage of cellular contents into the interstitial layer, with accumulation of toxic byproducts in the area affected by the lack of blood supply. Furthermore, the washout rate significantly decreases in the ischemic region (Chang et al., 1987). Ischemia leads to decrease in cell pH, excess hydrogen ions, and ATP depletion, causing calcium overload (Kalogeris et al., 2012). When blood supply is reestablished during reperfusion, the delivery of oxygen triggers a cascade of events that exacerbates myocardial injury. The antegrade delivery of oxygenated blood during reperfusion is higher in non-jeopardized areas of the myocardium (Chang et al., 1987) which may explain why, even with prompt reperfusion, salvage of the ischemic region is not always successful. On the other hand, when retroperfusion of oxygenated blood is established via the coronary veins, the delivery of oxygenated blood is more selective towards the ischemic myocardium (Chang et al., 1987), and at lower pressure than antegrade perfusion, more gently enhancing the washout of toxic metabolites in these regions. Furthermore, it has been hypothesized that the increase in pressure in the venous vessels triggers embryonic molecular pathways that may contribute to the salvage of the ischemic myocardium (Mohl et al., 2008).

The computer simulations offer some insights into the mechanisms of action of SARP in reducing infarct size. Specifically, retrograde perfusion not only facilitates the delivery of essential oxygen and nutrients but also ensures the removal of metabolic byproducts, thus creating a dual-effect mechanism of action of SARP in reducing infarct size. This interplay between the washout effect and the redistribution of oxygenated arterial blood is an important aspect of SARP. SARP significantly contributes to the reduction of infarct size as indicated by the experimental data, probably due to the combined effects of the removal of harmful metabolites and the supply of oxygen-rich blood. Simulation of the role of oxygen and nutrients in SARP remains for future studies.

Finally, deliverability is a common limitation for therapies targeting RI since they should be administered prior to reperfusion to truly precondition the coronary vasculature and myocardium. SARP provides an option to effectively deliver treatment since the device is relatively easy to navigate. The simplicity of the SARP catheter makes it attractive for utilization alone or in combination with other modalities of treatment like hypothermia, cell, or drug delivery.

PCI has significantly improved the outcome in acute myocardial infarction patients, but many still experience extensive myocardial damage due to reperfusion injury, leading to ventricular dysfunction and heart failure. Our study demonstrated that following coronary occlusion, a short (10 min) pretreatment with SARP prior to full restoration of antegrade flow greatly reduced myocardial infarct size in swine. The implications of this study suggest that myocardial injury due to reperfusion can be mitigated potentially by a washout mechanism by changing the manner of flow restoration and delaying the opening of the occluded artery for a brief period of time while SARP is implemented and the patient is prepared for PCI.