Author + information
- Received June 1, 2017
- Revision received July 28, 2017
- Accepted August 9, 2017
- Published online January 15, 2018.
- Rajesh Dash, MD, PhDa,b,∗ (, )
- Yoshiaki Mitsutake, MDa,
- Wook Bum Pyun, MD, PhDa,c,
- Fady Dawoud, PhDd,
- Jennifer Lyons, RVTa,
- Atsushi Tachibana, RTa,
- Kazuyuki Yahagi, MDe,
- Yuka Matsuura, MSa,
- Frank D. Kolodgie, PhDe,
- Renu Virmani, MDe,
- Michael V. McConnell, MD, MSEEa,b,f,g,
- Uday Illindala, MSd,
- Fumiaki Ikeno, MDa,b and
- Alan Yeung, MDa,b
- aDivision of Cardiovascular Medicine, Stanford University, Stanford, California
- bCardiovascular Institute, Stanford University, Stanford, California
- cSchool of Medicine, Ewha Womans University, Seoul, South Korea
- dZOLL Circulation, Inc., San Jose, California
- eCVPath Institute, Inc., Gaithersburg, Maryland
- fDepartment of Electrical Engineering; Stanford University, Stanford, California
- gVerily Life Sciences, Mountain View, California
- ↵∗Address for correspondence:
Dr. Rajesh Dash, Department of Medicine, Division of Cardiovascular Medicine, Stanford Medical Center, 300 Pasteur Drive, H2157, Stanford, California 94305-5233.
Objectives The study investigated whether a dose response exists between myocardial salvage and the depth of therapeutic hypothermia.
Background Cardiac protection from mild hypothermia during acute myocardial infarction (AMI) has yielded equivocal clinical trial results. Rapid, deeper hypothermia may improve myocardial salvage.
Methods Swine (n = 24) undergoing AMI were assigned to 3 reperfusion groups: normothermia (38°C) and mild (35°C) and moderate (32°C) hypothermia. One-hour anterior myocardial ischemia was followed by rapid endovascular cooling to target reperfusion temperature. Cooling began 30 min before reperfusion. Target temperature was reached before reperfusion and was maintained for 60 min. Infarct size (IS) was assessed on day 6 using cardiac magnetic resonance, triphenyl tetrazolium chloride, and histopathology.
Results Triphenyl tetrazolium chloride area at risk (AAR) was equivalent in all groups (p = 0.2), but 32°C exhibited 77% and 91% reductions in IS size per AAR compared with 35°C and 38°C, respectively (AAR: 38°C, 45 ± 12%; 35°C, 17 ± 10%; 32°C, 4 ± 4%; p < 0.001) and comparable reductions per LV mass (LV mass: 38°C, 14 ± 5%; 35°C, 5 ± 3%; 32°C 1 ± 1%; p < 0.001). Importantly, 32°C showed a lower IS AAR (p = 0.013) and increased immunohistochemical granulation tissue versus 35°C, indicating higher tissue salvage. Delayed-enhancement cardiac magnetic resonance IS LV also showed marked reduction at 32°C (38°C: 10 ± 4%, p < 0.001; 35°C: 8 ± 3%; 32°C: 3 ± 2%, p < 0.001). Cardiac output on day 6 was only preserved at 32°C (reduction in cardiac output: 38°C, –29 ± 19%, p = 0.041; 35°C: –17 ± 33%; 32°C: –1 ± 28%, p = 0.041). Using linear regression, the predicted IS reduction was 6.7% (AAR) and 2.1% (LV) per every 1°C reperfusion temperature decrease.
Conclusions Moderate (32°C) therapeutic hypothermia demonstrated superior and near-complete cardioprotection compared with 35°C and control, warranting further investigation into clinical applications.
Therapeutic hypothermia (TH) confers a cardioprotective benefit by increasing the tolerance to injury from myocardial ischemia and reperfusion. Due to the broad effect of temperature on most cellular processes, TH may protect from ischemia through a variety of interrelated mechanisms, which are not completely understood. Direct detriments to ATP consumption trigger intracellular acidosis, Na+/Ca2+ overload, and mitochondrial oxygen radical formation, leading to both necrosis and apoptosis (1).
Cardioprotection from TH was previously demonstrated in small and large animal ischemia models (2–7). An earlier observation of TH dose-response was made by Miki et al. (2) in rabbits, where significant cardioprotection was noted with cooling to 32°C, but less protection with cooling to 35°C when TH was administered before reperfusion. Later animal studies confirmed the need to initiate TH before reperfusion to retain myocardial salvage of the therapy (1,8).
Clinically, hypothermia has been shown to be safe and feasible; however, the outcomes of clinical trials, such as the COOL-MI (Cooling As An Adjunctive Therapy To Percutaneous Intervention In Patients With Acute Myocardial Infarction) (9), ICE-IT (Intravascular Cooling Adjunctive To Primary Coronary Intervention) (10), RAPID MI ICE (Rapid Intravascular Cooling in Myocardial Infarction as Adjunctive to Percutaneous Coronary Intervention) (11), and CHILL MI (Efficacy of Endovascular Catheter Cooling Combined With Cold Saline for the Treatment of Acute Myocardial Infarction) (12) trials, have not been consistent with the majority of the experimental literature. The recent CHILL MI (12) trial showed a nonsignificant trend of infarct size (IS) reduction by 13% with hypothermia (ice-cold intravenous saline plus endovascular cooling) (13) imaging. Following the trend of previous trials, the subgroup of early (<4 h) anterior ST-segment myocardial infarction (STEMI) showed a significant 33% reduction in IS. Notably, only 76% of patients receiving TH reached the goal temperature <35°C (∼9-min increase in door-to-balloon time).
Thus, timing (pre-reperfusion temperature) and depth of cooling have been identified as important factors modulating the efficacy of TH. Moreover, rapid induction of TH is important in the setting of clinical AMI in which minimizing door-to-balloon time and reaching target temperature before reperfusion are crucial for salvaging cardiac muscle. Thus, it is critical to understand the optimal target reperfusion temperature to maximally attenuate MI damage in a clinically relevant model of MI.
In this study, we evaluated the effect of rapid endovascular TH on salvaging myocardial tissue in AMI in a porcine model. We hypothesized that a lower rapid TH target temperature would further reduce the size of infarction resulting in increased cardioprotection.
AMI animal model
Animal care and interventions were in accordance with the Laboratory Animal Welfare Act and all animals received humane care and treatment in accordance with the Guide for Care and Use of Laboratory Animals (www.nap.edu/catalog/5140.html). Yorkshire female swine (n = 24) weighing 46 ± 3 kg were randomized into 3 groups: normothermia, hypothermia to 35°C (mild), and hypothermia to 32°C (moderate). Sample size determined using a power analysis based on prior experience with this animal model to see a 30% reduction in IS. Animals were pre-anesthetized with intramuscular Telazol (6 mg/kg) prior to tracheal intubation. Animals were maintained in a surgical plane of anesthesia with inhaled anesthetic isoflurane (1% to 4%) through a volume-controlled ventilator (tidal volume 10 to 15 ml/kg and ventilation rate 8 to 15 breaths/min).
Tidal volume, respiration rate, and FiO2 were adjusted to maintain normocapnia (end-tidal PaCO2 [EtCO2] 35 to 45 mm Hg) as measured by a capnometer (NPB-75, Nellcor-Puritan-Bennett, Boulder, Colorado) placed on the intubation tube. Electrocardiogram, EtCO2, temperatures (left ventricular [LV], rectal, esophageal), and blood pressures (aortic, central venous) were monitored and recorded throughout using a multichannel data acquisition system (ADInstruments, PowerLab, Colorado Springs, Colorado).
Occlusion ischemia was induced using a coronary balloon advanced to a mid-left anterior descending artery (LAD) location typically proximal to the first diagonal branch and inflated to 3 to 6 atm. Vessel occlusion and ischemia were confirmed with contrast dye injection and ST-segment elevation on electrocardiogram. For animals receiving TH, an endovascular temperature management balloon catheter (Proteus Intravascular Temperature Management System, ZOLL, San Jose, California) was introduced via 12-F femoral venous sheath. The cooling balloon of the catheter was positioned in the inferior vena cava below the right atrium. The Proteus catheter and console controlled core body temperature using a probe passing through a balloon lumen at the user-specified temperature set point.
Myocardial ischemia was maintained for 60 min. Hypothermia was initiated 30 min after balloon occlusion and maintained for 30 min after reperfusion (Figure 1A). Upon balloon deflation or reperfusion, the deflated balloon catheter was left in place, and flow to the distal LAD was evaluated by contrast injection. At the end of cooling, swine were actively rewarmed at the rate of 1.2°C/h until reaching 36.5°C and rewarmed passively. Blood samples were collected at baseline, 15, and 60 min after reperfusion to measure cardiac troponin I (cTnI) serum levels.
Cardiac magnetic resonance image acquisition and analysis
Cardiac magnetic resonance (CMR) was performed at baseline day 0 as well as day 6 after MI on a 3T scanner (Signa HDx 3.0T, GE Healthcare, Fairfield, Connecticut) using an 8-channel flexible cardiac coil. Cardiac function was assessed using short-axis stack, 2-chamber and 4-chamber cine images, 10 mm slice thickness with no gap, 30 phases, gradient-echo balanced steady-state free precession (fast imaging employing steady-state acquisition), flip angle 45°, repetition time ∼3.8 ms, echo time minimum-full, matrix 224 × 224, field of view 35 cm. Myocardial edema was assessed by T2 using free-breathing fast spin-echo double inversion recovery sequence (typical parameters: echo times ∼10, 20, 30, 40 ms, 3 short-axis slices, flip angle 90°, repetition time >2,500 ms). Late gadolinium delayed-enhancement CMR (DE-CMR) was acquired to assess IS using short-axis stack (10 mm thickness, no gap), 2- and 4-chamber fast gradient echo steady-state breath-hold sequence 15 to 25 min post-contrast injection, which provided an optimal blood pool–myocardial contrast for infarct quantification. Typical parameters: flip angle 15°, repetition time ∼4.6 ms, matrix 256 × 256, field of view 35 cm, number of averages 2 to 4, inversion time adjusted to null normal myocardium.
CMR images were analyzed using cmr42 (Circle Cardiovascular Imaging Inc., Calgary, Canada) with semiautomatic LV volume, mass, and injury or infarct tracings. DE-CMR infarct regions were quantified using the 5 SD, 6 SD, and full width at half maximum methods. Hypointense microvascular obstruction regions were included as infarct. T2 edema analysis was similarly analyzed using a threshold of 3 SD above normal tissue mean signal. Figure 1A shows the complete experimental protocol. Notably, T2 edema was not used for area-at-risk (AAR) estimation due to the impact of hypothermia on T2 edema signal.
Ex vivo pathology examination
Following day 6 CMR, animals were injected with triphenyl tetrazolium chloride (TTC) and Evans Blue dye to characterize infarct region and AAR, respectively, by advancing a coronary balloon to the same mid-LAD location followed by dye injection as previously described (14). Animals were euthanized by injection of pentobarbital and potassium chloride with confirmation of asystole. The heart was removed surgically by thoracotomy and rinsed in cold saline while maintaining the vessel occlusion using a surgical clip. After 12 min of flash freezing, the heart was sliced in 10 mm increments parallel to the atrioventricular sulcus from apex to base. Slices were submerged into 37°C 2% TTC solution and incubated for 15 min. All slices were scanned over a white background using a high-resolution scanner (Epson Workforce DS-7500, Shiojiri, Japan). Images were manually traced to calculate IS, AAR, and LV mass using ImageJ (version 1.50d) (15).
Additional subcellular histopathological analysis was performed on 16 heart specimens at an independent external specialized laboratory (CVPath, Gaithersburg, Maryland) blinded to the treatment group. Heart slices placed in VWR Mega Cassettes (VWR Inc., Radnor, Pennsylvania) were processed through a graded series of alcohols and xylenes for dehydration and clearing and embedded in paraffin. Digital macro-photographs of each paraffin block were acquired to document the Evans Blue dye staining delineating the AAR. Histologic sections were prepared using a rotary microtome at 4 to 6 μm, mounted on charged slides (51 × 75 mm), and stained with hematoxylin and eosin and Masson’s trichrome.
Confirmation of granulation tissue and overall IS was investigated by immunohistochemical staining. Briefly, histologic sections were deparaffinized and endogenous peroxidase activity limited by exposure to 0.3% H2O2 while nonspecific binding of primary antibody was minimized by blocking with a species-appropriate normal serum. Histologic sections were incubated with primary polyclonal antibodies directed against vimentin (fibroblast marker) (∼dilution 1:40, Cat#M0725, Dako, Carpinteria, California) or swine macrophage marker (∼dilution 1:50, clone CD107a/4E9/11, Cat#MCA2315GA, AbD Serotec, Raleigh, North Carolina) overnight at 4°C. Prior to incubating with primary antibody antigen unmasking was performed with steam heat in ethylenediaminetetraacetic acid buffer. Antibody binding was detected using an EnVision + Dual-Link System HRP (Dako) with a NovaRed chromogen (Vector, Burlingame, California) where Gill’s hematoxylin was used as a counterstain. Negative controls were treated with an identical dilution of nonimmune rabbit serum. A biotin-conjugated Lectin from Dolichos biflorus (dilution 1:5, Cat#L6533, Sigma-Aldrich, St. Louis, Missouri) was used as a neoangiogenesis marker. After overnight incubation, the lectin staining was visualized using a LSAB labeling kit (Dako) with a Nova Red chromogen (Vector). Dako hematoxylin was used as counterstain.
Whole-heart sections stained by hematoxylin and eosin and Masson’s trichrome were digitized using the Axio Scan.Z1 slide scanner (Carl Zeiss Inc., Oberkochen, Germany) equipped with Zen 2012 software (blue edition) (Carl Zeiss Inc.) while offline area measurements were performed using Zen desk 2012 (Carl Zeiss Inc.). LV area, AAR, viable area, and infarct area were manually traced. The area of granulation tissue was calculated as AAR – necrotic area – viable area.
Continuous data were represented as mean ± SD. One-way analysis of variance was used to compare group mean differences. Holm-Sidak’s test was used for multiple group comparisons and calculating adjusted p values. Pearson’s r coefficient was used to test cross-modality correlation. Cross-modality agreement was assessed using mean difference ± 1 SD and was graphically represented by Bland-Altman plots. Linear regression was used to test for dose-response trend. A 2-sided p value <0.05 was considered statistically significant. Histopathology variables with non-normal distributions were expressed as median with 25th and 75th percentiles and compared by the Kruskal-Wallis test and Dunn’s test for multiple groups comparisons.
Baseline weight and LV temperature were similar among all groups (p > 0.05). Figure 1B shows the cooling timeline for both hypothermia groups. The 35°C group reached target temperature in 9 ± 5 min, and the 32°C group in 29 ± 8 min. All animals achieved their target temperature before reperfusion. cTnI levels increased significantly from baseline to post-reperfusion returning to baseline levels at day 6 confirming AMI. Both hypothermia treatment groups showed significantly less increase in cTnI levels at 15 min, 60 min, and 6 days post-ischemia compared with the control group (15 min: p = 0.040 and 0.040; 60 min: p < 0.001 for both; day 6: p = 0.004 for both; 38°C vs. 35°C and 38°C vs. 32°C, respectively) (Figure 2D). Hemodynamic changes during ischemia and reperfusion are detailed in the Online Appendix, Online Figures 1 and 2, which demonstrate a favorable response to reperfusion in the 32°C group. Briefly, 32°C animals showed favorable results at reperfusion compared with normothermia with regard to rate-pressure product (reduction of 17% and 4%, 32°C vs. 38°C and 35°C, p = 0.046 and 0.019, respectively) and CO2 production (19% reduction for both 38°C vs. 35°C or 32°C; p < 0.001).
TTC infarct quantification
The AAR was similar in all treatment groups (analysis of variance p = 0.2). Both the 32°C and 35°C groups showed strong significant reduction (62% and 91%, respectively, vs. LV mass, p < 0.001 vs. 38°C) in IS compared with the normothermic group, whether normalized to AAR or to LV mass. Furthermore, the 32°C group showed an additional significant 29% IS reduction compared with the 35°C group (%IS/AAR: p = 0.013) indicating more efficacy of the deeper cooling on salvaging tissue. The 32°C group also had borderline significant reduction in IS compared with 35°C group relative to LV mass (p = 0.052). Target temperature was a significant predictor of IS reduction in the linear regression model both relative to AAR and LV mass (goodness-of-fit residual R2 = 0.746 and 0.685, respectively; p < 0.001). The predicted IS reduction using this model is 6.7% of AAR and 2.1% of LV mass per 1°C drop of core temperature at reperfusion. Figure 2 and Table 1 show TTC quantification of infarct tissue and AAR by percent LV mass in the study groups.
CMR function, infarct and edema quantification
Although ejection fraction and chamber sizes showed a nonsignificant trend of preserved function in both TH groups (p > 0.05) (Table 2 and Online Appendix), cardiac output change at day 6 was only preserved in the 32°C group (38°C vs. 32°C, p = 0.041) (Figure 3C). CMR images evaluated for IS using 5-SD and 6-SD image thresholding methods showed a significant reduction in LV IS for the 32°C group compared with normothermia (LV reduction 53% and 70% for 5 SD and 6 SD, respectively; p ≤ 0.001) (Figure 3A). Interestingly, full width at half maximum analysis failed to demonstrate a reduction in IS compared with control, unlike the aforementioned 5-SD and 6-SD CMR analysis and direct TTC histopathological scar measurements. Furthermore, a significant reduction was observed between the 35°C and 32°C groups for both 5 SD and 6 SD (38% and 63% LV reduction of 32°C vs. 35°C; p = 0.023 and 0.014, respectively). Temperature was a significant predictor of IS reduction in the linear regression model (goodness-of-fit residual R2 = 0.474 and 0.483, respectively; p < 0.001 for 5 SD and 6 SD). The predicted IS reduction from this model was 1.5% and 2.1% of LV per 1°C drop of reperfusion temperature for 5 SD and 6 SD, respectively.
Evaluation of tissue edema using T2-weighted imaging showed decreased edema (39% reduction; p = 0.017 in both groups) in the hypothermia groups compared with normothermia, making T2 unsuitable as a surrogate for AAR. Figure 3 and Tables 1 and 2 show data for CMR function, DE-CMR, and edema.
Significant reduction of irreversibly damaged infarct tissue was seen between the 32°C and 38°C groups relative to TTC AAR (69%, p = 0.004) and similar reductions were observed relative to total LV area (80%, p = 0.008). TTC-AAR was similar across all groups (analysis of variance p = 0.554). Non-necrotic granulation tissue was significantly larger in the 32°C group compared with the normothermic groups (52%, p = 0.011). Temperature was a significant predictor of IS reduction in the linear regression model both relative to AAR and LV (goodness-of-fit residual R2 = 0.69 and 0.58, respectively; both p < 0.001). The predicted IS reduction was 5.8% of AAR and 1.8% of LV per 1°C drop of cooing temperature at reperfusion. Figure 4 and Table 1 depict histopathology data.
CMR and TTC comparisons
Comparison between IS measured with the DE-CMR 6-SD method and TTC method showed overall good agreement (systematic bias 1%, 1-SD bias limit 5%). Bland-Altman plots showed good visual match (Figure 5A) and strong positive correlation (r = 0.59, p = 0.003) (Figure 5B). The 6-SD threshold method had the least systematic bias (6 SD: 1 ± 5%; 5 SD: 6 ± 5%) compared with TTC. Although injury regions measured by CMR T2-weighted images showed significant reduction with decreasing temperature, the AAR measured by TTC was similar in all groups (Figure 5C).
The major findings of this study are that: 1) pre-reperfusion moderate hypothermia (32°C) demonstrated near-complete cardioprotection compared with both normothermia and mild hypothermia (35°C); and 2) moderate hypothermia also preserved cardiac output compared with either normothermia or mild hypothermia.
The primary objective of the present study was to investigate the dose-response relationship between TH temperature before reperfusion and IS in a large animal model of AMI using a venous catheter-based temperature management systems mimicking the clinical setting of acute ischemia. Previous experimental studies have suggested incremental benefit of colder hypothermia in the context of AMI (1,16); however, to our knowledge, this is the first attempt to directly compare rapid (under 30 min) moderate (32°C) versus mild (35°C) hypothermia before reperfusion in a clinically relevant in vivo large animal model of AMI. Moreover, the current study tests a hypothermia dose-response using an intravascular cooling method that achieves target temperature within an acceptable time frame and technical capacity for a catheterization laboratory, such that hypothermia can be achieved with no significant delay to reperfusion. In this study, IS was measured using 3 techniques: in vivo DE-CMR, ex vivo gross tissue staining with TTC, and cellular histopathology tissue staining. Each method independently demonstrated substantial reduction of IS in the hypothermia groups. Additionally, the 32°C group showed improved myocardial salvage when compared to 35°C using each of the 3 measurement techniques. Finally, both DE-CMR and TTC quantification techniques showed good correlation and agreement.
Mechanism of TH
TH has been shown to have cardioprotective benefit by increasing tolerance to injury due to myocardial ischemia and reperfusion. One of the many mechanisms underlying the protective effects of TH is decreased metabolism. Because many enzyme systems in cell membranes including adenosine triphosphatases are temperature sensitive, as the core temperature drops, cardiac and systemic metabolic rates decrease, oxygen and glucose consumption and carbon dioxide production also decrease, which help hypothermic hearts endure ischemia for longer periods of time than normothermic hearts (17). The cardioprotective effect of hypothermia includes a reduction in metabolic demand with both preserved adenosine triphosphate and glycogen stores (18,19) as well as reduced tissue-level accumulation of lactate (20). TH was also shown to reduce calcium and sodium overload during ischemia reperfusion injury and the release of reactive oxygen species (21). Following ischemia with subsequent reperfusion, cells may recover to varying degrees, become necrotic, or enter a pathway to programmed cell death (apoptosis). TH inhibits proapoptotic calcium-induced mitochondrial permeability transition pore (22) and promotes antiapoptotic pathways (23). Recently, hypothermia was shown to mediate cardioprotection through the reperfusion injury salvage kinase signaling pathways (PI3, Akt, NO, and ERK) (24). Further research is needed to elucidate exactly how the temperature dose modulates interplay between these mechanisms.
Endovascular TH has been shown to be safe and feasible. However, the profound cardioprotective effect seen in experimental animal studies has yet to be demonstrated in a prospective randomized clinical trial. Factors that were suggested to address shortcomings of previous studies include time to induction after injury, target temperature, cooling duration, rewarming rate, and controlling TH complications. Using an endovascular cooling catheter that can achieve rapid, deep, and precisely controlled hypothermia has the potential to mitigate many of these shortcomings. The current study showed that using a high-power catheter and heat exchange system can achieve 3°C drop in <10 min and 5°C drop in <30 min. Minimizing induction time of TH translates to minimizing door-to-balloon time, which is critical to salvaging myocardial tissue. Also, previous experimental studies have shown that cardioprotection is lost when cooling is delivered after reperfusion (6) indicating that delaying reperfusion to reach target temperature might improve TH cardioprotection (8). Notably, neuromuscular blockade was not required to achieve target pre-reperfusion temperature, making TH more feasible within current clinical STEMI protocols.
The current study focused on evaluating parameters showing trends of TH cardioprotective benefit seen in post hoc subgroups analysis of previous clinical trials. In the COOL-MI study (9), although there was no overall difference of IS reduction reported at 30 days post-MI as measured by nuclear imaging, anterior STEMI treated patients reaching core temperature of <35°C before reperfusion had a 49% relative reduction in IS compared with normothermic patients (9.3% IS treated vs. 18.2% control; p = 0.05). Similarly, in the ICE-IT study (10), a trend for benefit was observed in the subgroup with anterior infarction and body temperature of <35°C at reperfusion (43% reduction, p = 0.09). In the current study, a treatment effect of TH on IS reduction was observed at 35°C, and the treatment effect was profound at 32°C.
Infarct reduction dose-response
The current study confirms the trend of beneficial IS reduction of both mild (35°C) and moderate (32°C) pre-reperfusion TH. It further shows profound reduction of IS relative to AAR in the 32°C TH group (91% for moderate TH and 62% for mild TH) which is an additional 29% reduction over mild TH. The dramatic reduction of IS with endovascular, moderate TH is among the highest reported in the literature for preclinical studies. For comparison, 90% reduction was reported following rapid perfluorocarbon total liquid ventilation TH to 32°C in rabbits, Tissier et al. (22). Also, an 80% reduction was reported following endovascular TH to 34°C in human-sized pigs, Dae et al. (4). Of note is the percent cooling time was less in the present study. This study initiated achieved target TH temperature by 30 min into the 60-min ischemic period, compared with the 25-min mark of the 30-min ischemic period, and 40–min mark of the 60-min ischemic period for Tissier et al. (22) and Dae et al. (4) studies, respectively. This perhaps indicates that delayed initiation but deeper target rapid cooling from ischemia onset before reperfusion can salvage substantial myocardium similar to very early mild TH.
An intriguingly high infiltration of immune cells and mediators for repair or remodeling (myofibroblasts, macrophages, endothelial cells) was observed in 32°C hearts as indicated by more granulation tissue on the infarct periphery. The 35°C group did not demonstrate the same effect. Dae et al. (4) noticed a similar effect of moderate hypothermia creating a smaller, patchy-appearing infarct. This may indicate more rapid transition from the inflammatory phase (debris and dead cell cleaning) to the proliferation phase (rebuilding extracellular matrix and angiogenesis) perhaps due to less acute insult to cardiac myocytes. A longer-term follow up is required to confirm if this day 6 acute TH benefit translates into long-term improvement on remodeling such as LV size and function.
Although this validated experimental model of ischemia-reperfusion demonstrated dramatic cardioprotection, these results must be tested in a clinical AMI population. However, this cooling protocol can be initiated with venous access, potentially prior to transport to a cath lab, making it more feasible to achieve target temperature. Shivering was not a major confounder in this study, but clinical trials may experience variable shivering responses to cooling. Importantly, post-reperfusion day 6 was chosen in this study based upon evidence that edema patterns would be similar to immediate post-reperfusion injury edema levels (25). Nevertheless, a significant T2 edema suppression was observed in both hypothermia groups at day 6, compared with normothermia, and consequently, only TTC AAR measurements were used to ensure consistency. Swine are also known to have less collateral coronary circulation than humans, and consequently humans may exhibit less dramatic IS reduction. Last, the rate or rewarming (1.2°C/h) was much faster than ideal rewarming rates (0.4°C/h). Indeed, a slower rate of rewarming may achieve even greater IS reduction.
Pre-reperfusion rapid TH shows a dose-dependent reduction in IS and preserved cardiac function for moderate hypothermia (32°C) compared with mild hypothermia (35°C). The magnitude of protection at 32°C was dramatic and warrants further preclinical and clinical studies to evaluate cardiovascular prognosis and the effect of temperature dose on chronic infarct remodeling and cardiac function.
WHAT IS KNOWN? Recent clinical trials of TH have failed to reduce myocardial injury consistently following ischemia-reperfusion, in part due to variable achievement of pre-reperfusion core temperature. In addition, preclinical and neuroprotection data suggest that a deeper hypothermic target of 32°C may confer more consistent cardioprotection.
WHAT IS NEW? In this study, we achieved 32°C pre-reperfusion TH in a swine model of ischemia-reperfusion that conferred a 91% relative IS reduction compared with normothermia, preserved cardiac output, and increased granulation tissue in the peri-infarct zone.
WHAT IS NEXT? These data support performing additional clinical trials that target a lower, more consistent, and more rapid induction of TH to maximize cardioprotection in STEMI patients.
The authors thank Alfredo Green and Jorge Reifart of Stanford University and Brian Holt of ZOLL for assistance with the study.
The study was partially sponsored by ZOLL Circulation, Inc., who also played a role in study design, data collection, and review of technical details within manuscript. CV Path, Inc., played a role in data collection, production of immunohistochemistry results, and review of technical details within the manuscript. Neither ZOLL nor CVPath played a role in data interpretation, decision to publish, or manuscript writing. Dr. Dash was supported by the National Institutes of Health/National Heart, Lung, and Blood Institute. Drs. Dawoud and Kieno are employees of ZOLL Circulation. Drs. Kolodgie, Virmani, and Yeung are employees of CVPath. Dr. McConnell is currently on partial leave of absence from Stanford; and is an employee at Verily Life Sciences, Inc. Dr. Dash has received research grant support from ZOLL Circulation. Dr. Illindala is an employee of Shockwave Medical. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- area at risk
- acute myocardial infarction
- cardiac magnetic resonance
- cardiac troponin I
- delayed-enhancement cardiac magnetic resonance
- infarct size
- left anterior descending artery
- left ventricular
- ST-segment elevation myocardial infarction
- therapeutic hypothermia
- triphenyl tetrazolium chloride
- Received June 1, 2017.
- Revision received July 28, 2017.
- Accepted August 9, 2017.
- 2018 American College of Cardiology Foundation
- Kloner R.A.
- O’Neill W.W.,
- Dixon S.R.
- ↵Grines CL. Intravascular cooling adjuctive to percutaneous coronary intervention (Part 1). TCT 2004. Available at: http://www.scribd.com/doc/40117148/ICE-IT-Presentation-TCT2004. Accessed January 1, 2015.
- Götberg M.,
- Olivecrona G.K.,
- Koul S.,
- et al.
- Erlinge D.,
- Götberg M.,
- Lang I.,
- et al.
- Dash R.,
- Chung J.,
- Ikeno F.,
- et al.
- ↵ImageJ. Available at: https://imagej.nih.gov/ij/. Accessed December 21, 2017.
- Erlinge D.
- Badeer H.
- Ichihara K.,
- Robishaw J.D.,
- Vary T.C.,
- Neely J.R.
- Fernández-Jiménez R.,
- Sánchez-González J.,
- Agüero J.,
- et al.