Fast Hypothermia Induced by Continuous Renal Replacement Therapy Alleviates Renal and Intestinal Injury After Cardiac Arrest in Swine

Background: Renaland intestinal damagelead tomultiple organ dysfunction and death after cardiopulmonary resuscitation (CPR), and can be partly mitigated by therapeutic hypothermia. Currently, continuous renal replacement therapy (CRRT) was demonstrated to be an effectiveway to induce hypothermia. In the present study, we aimed to investigate the inuence of CRRT cooling on renal and intestinal damage after CPR based on a porcine model.


Background
In resuscitated patients undergone cardiac arrest, thesystemic ischemia-reperfusion (I/R) injury often inducesa sepsis-like syndrome orpost-cardiac arrest syndrome (PCAS), leading to multiple organ injuries with brain, heart, kidneyandintestine, etc, involved in [1].Several attempts have been made to mitigate the brain and heart injury after resuscitation, which have been considered as main causes of morbidity and mortality for cardiac arrest (CA) survivors, yet renal and intestinal injury have received less attention. In fact, almost 50% of CA survivors suffer acute kidney injury (AKI), and nearly16% are treated with renal replacement therapy [2].AKI occurs owing to a combination of I/R injury and hypoperfusion due to circulatory shock following resuscitation [2]. Post-arrest AKIisstronglyrelated with worse neurological outcome and survival [2][3][4].Besides, nearly 60% of survivors of CA are presented with intestinal injury [5], which has been proved via endoscopy and autopsy after CPR [5,6].Due to the susceptibility to I/R, mesenteric ischemia often occurs, causing an increase in intestinal permeability and subsequent bacterial translocation [7],which may beexceedinglydetrimental [8]. Additionally, the elevation of biomarkers of intestinal injury such as fatty-acid binding protein (IFABP) after CAarerelated with endotoxemia, further worsening post-resuscitation organ failure and shock [9].
Therapeutic hypothermia (TH)has been a main therapeutic intervention ofpost-resuscitation care. The protective role of TH oncardiac dysfunction and brain injury has been studied previously, but its in uence onrenaland intestinal damage was not intensively investigated yet. A previous study demonstratedthat24hr of TH after CPR was associated with a delayed improvement in renal function [10].Hasslacher et al [11] showed that TH had a protective effect against the development and recovery of AKI, accompanied by lower levels of creatinine (Cr) and cystatin C. With regard to the role of TH on intestinal injury, a study showed TH alone or in the combination with sevo urane could affect small intestinal protein expression and activity, and the proteins might be involved in intestinal I/R injury following resuscitation [12]. Another study showed that ulinastatincombined with TH mightexert a protective effect in the small intestinal mucosa by decreasing oxidative stress in a rat model [13].
However,despite TH has been proved to exert potent protection for CA survivors [14,15], more than 50% of them die or have poor neurological outcome [16][17][18].A review demonstrated that rapid cooling exerted a higher rate of good neurological outcome than slower cooling methods [19]. A cooling rate of > 3℃/h appear to be bene cial to arrive at a targeted temperature below 34℃ within 3.5 h after return of spontaneous circulation (ROSC) [19]. Therefore, the undesirable outcome of TH may be deprived from the unsatisfactory cooling effects of traditional methods such as surface blanket cooling, rapid infusion of ice lactated Ringer's solution, etc. Previous studies noted that conventional cooling methods usually take several hours to decrease the body temperature [20].Hence,rapid cooling may bene tresuscitated patients more than slower cooling methods. Previous research has attempted to apply several techniquesto induce fast hypothermia, including liquid ventilation [21,22], peritoneal lavage system [23] and intravenous cooling [24], yet it's hard to incorporate them into the clinical setting due to the required speci c equipment.
Recently, continuous renal replacement therapy (CRRT) can also be used as another type of cooling method.Karacan et al [25] and Ma et al [26]have reported case reports on usingCRRT to induce and maintain TH.Du et al [27] showed the combination of CRRT and TH enabled acute heart failure to achieve recovery of cardiac function. CRRT, as a common and well-developed treatment in critical ill patients, will be a potent cooling method for CA survivors.Recently, we have demonstrated the fast hypothermia induced by CRRT had protective effects in systemic in ammation, heartand brain damageafter resuscitation in a porcine model [28].
However, the in uence of hypothermia induced byCRRT on renal and intestinal dysfunction haven't been determined.The aim of thisresearch was to investigate the in uence of CRRT cooling on renal and intestinal injury following CPR in swine.

Methods
This study was arandomized, controlled laboratory experiment based on a pig model of cardiac arrest.All the experimental procedures were performed based on the methods of our one previous study, in which the animal model has been well established [29]. Ethics committee approval was obtained from the Second A liated Hospital, Zhejiang University School of Medicine(No. 2019004). The study included 32healthy male domestic pigs purchased from same vendor (Shanghai Jiagan Biotechnology Inc., Shanghai, China), weighting 36 ± 2 kg. The pigs were housed under controlled pressure, temperature, humidity and lighting conditions. They were given water and food regularly, washed regularly and disinfected in closed cages. Allpigsreceived care based on the Principles of Laboratory Animal Care and the Guide for the Care and Use of Laboratory Animals.

Animal preparation
Thenight before the experiment food was withdrawn from all animals, but they were available to water intake. Induction of anesthesia in pigs was achieved by a combination of intramuscular ketamine injection at 20 mg/kg and sodium pentobarbital injection at 30 mg/kg in an ear vein. Then, to maintain theanesthesia, sodium pentobarbital at 8 mg/kg/h as well as fentanyl at 2mg/kg/h were givenintravenously. Ventilation was maintained using a volume-controlled ventilator (SynoVent E5, Mindray, Shenzhen, China) with the following setting: tidal volume, 12 mL/kg; peak ow, 40 L/min and FiO 2 , 21%.End-tidal carbon dioxide (ETCO2) was measured using an ETCO 2 /SPO 2 monitor (PMSH-300, SunLife Science Inc., Shanghai, China), maintaining at 35mmHg ~ 40mmHg by respiratory frequency.The standard lead II electrocardiogram surface electrode was secured.
A double-lumen catheter (11 F, Gambro KathetertechnikHechingen, Hechingen, Germany) was placed into the left femoral vein to establish vascular access of CRRT.A uid-lled catheter (8 Fr, C.R. Bard Inc., Salt Lake, UT) was placed via the right femoral artery to the thoracic artery to measure aortic pressure. A thermodilution catheter (7 Fr, Abbott Critical Care # 41216, Chicago, IL) was placed via right femoral vein to the right atrium to monitor right atrial pressure and blood temperature. Intermittent heparinised saline ushes weremade to avoid clogging of the catheters. A pacing catheter (5 F, EP Technologies Inc., Mountain View, CA) was placed via the right external jugular vein to the right ventricle to induceventricular brillation (VF). Animals were placed supine on a heating blanket to maintain the normal temperature at 38.0 ± 0.5℃.

Experimental protocol
Baseline characteristics were recordedafter a 10-min stabilization. Pigs were randomly allocated to 1 of 4 groups:normothermia (NT, n=9), surface cooling (SC, n=9), CRRT cooling (CRRT, n=9), or sham control (Control, n = 5).Pigs in the NT and Control groups had the body temperature maintained at 38.0 ± 0.5℃using the Blanketrol III (Cincinnati Sub-Zero, Cincinnati, OH).For pigs in the other 2 groups, TH was startedat 5 min after CPR,and then maintained at the temperature of 33 ± 0.5℃. The hypothermiainduced in the CRRT group was achieved by 8-hr CRRT, using an AN69ST hemo lter (Gambro Industries Inc., Meyzieu, France), followed by 16-hr SC; while the cooling inthe SC group was achieved by 24-hr SC with the Blanketrol III.Then,a 1℃/h rate of rewarming were followed.
For the CRRT group,a 180 ml/minrate of the blood ow was determined initially, immersing the circuit in 4℃ice water until the target temperature of 33℃ arrived. Then, the temperaturewas maintained and the blood ow reducedby 60 mL/min. The rates of liquid replacementand ultra ltrationwere 30 mL/kg/h and 20 mL/kg/h, respectively. Immediately when CRRT started, a load dose of 1000-IU heparin was given for anticoagulation, followed by a dose of 150 IU, 300 IU, 450 IU for the rst three hours, respectively, and 600 IU/h for the rest 5 hours.
For all pigs, we induced VF by deliveringalternating current of 1 mA.Anesthesia and ventilation were disconnected, and the animals underwent an8-min period of untreated VF. Then, cardiac pulmonary resuscitation (CPR) was started, with a ratio of 30:2 of compression to ventilation. The chest compressions were achieved at a rate of 100 ~ 120/min (reaching 50 ~ 60mm deep) bya monitor de brillator (ZOLL Medical Inc., Chelmsford, MA). A dose of 20mg/kg epinephrine was administered at 2.5 min during resuscitation. Biphasic de brillation at 150 J wasperformed at 5min of CPR. ROSC was determined as an organized rhythm and mean arterial pressure of >50 mmHgsustaining for > 5 min.If not achieved, CPR was held on for another 2 min prior to the next de brillation. This cycle wasduplicatedevery 2 min, and administration of epinephrine was carried out every 3 min until successful ROSC or 15 min had elapsed.If achieved, 30-hr mechanical ventilation and infusion of normal salinewere subsequently continued to keep uid balance. After completionofthe study,euthanasia and a following necropsy wereexecuted to con rm potential injuries of thoracic or abdominal viscera due to experimental intervention.The experimental ow diagram was shown in Fig. 1A.

Measurements
Blood samples of veins and arteries were collected at 1, 3, 6, 12, 24, and 30h post resuscitation. Then the researchers separated serums from venous blood samples and stored them at -80℃ prior tofurther analyzing the levels of creatinine (Cr), blood urea nitrogen (BUN), intestinal fatty acid binding protein (IFABP) and diamine oxidase (DAO).Toevaluatethe in ammatory response and oxidative stress,kidney and intestine tissues from renal parenchyma and middle part of small intestine were harvested immediatelyaftereuthanasia, andsubsequentlyfrozen in liquid nitrogen prior to analysis. Levels of tumor necrosis factor-α (TNF-α), interleukin-6 (IL-6)were analyzed using enzyme-linked immunosorbent assay kits (ELISA, Meixuan Biotechnology Inc., Shanghai, China).The contents of malondialdehyde (MDA) were measured bythiobarbituricacid reactive substances assay, and activities of superoxide dismutase (SOD)were measured by xanthine oxide assay [30](Nanjing Jiancheng Bioengineering Institute, Nanjing, China).
The extent of apoptosis of the kidney and intestine were measured using terminal deoxynucleotidyl transferase-mediated dUTP-biotin nick end labeling (TUNEL) assay. The proportion of apoptotic cells was determined as the percentage of TUNEL-positive cells/total cells, and the cleaved caspase-3 protein was detected using immunohistochemistry. The staining intensity of cleaved caspase-3-positive undergone semiquantitative analysis through integrated optical density based on aprevious study [31] with Image-Pro Plus 6.0 software (Media Cybernetics, Silver Spring, MD).

Statistical analysis
Continuous variables were described as mean ± standard deviation (SD)or median (25th~ 75th percentiles) for data normally distributed or not.For comparisons amongmultiple groups,one-way analysis of variance was used for data normally distributed, Kruskal-Wallis test for data notnormally distributed.Bonferroni test was used to account for any two group comparisons when the overall comparison was signi cant.Categorical data were analyzed using Fisher exact test.A two-sided P value of <0.05 was considered as statistically signi cant.

Results
A total of 32pigs went through experiments completely,baseline characteristics and chemistries among all groups were mathematically the same (Fig. 1B). During CPR period,animals that experienced CA and resuscitation had an even level of CPP. In all three groups, 8/9 animals achieved ROSC, respectively.There was no signi cant difference as to coronary perfusion pressure (CPP),CPRduration, timesof de brillation, epinephrine dosage among 3 groups ( Fig. 2A). During post resuscitation period, the temperature in the CRRT group decreased signi cantly faster than in the SC group (9.8 ± 1.6 vs. 1.5 ±0.4 ℃/h,P< 0.01, Fig.  2B).
During post resuscitation period, the values ofcreatinine (Cr)showed tendency to descend in two hypothermic groups. At 6h, 12h and 24hafter resuscitation, compared to the NT group, the serum levels of Cr were signi cantly lower in two hypothermic groups, with the lowest in the CRRT group (all P< 0.05). In addition, the levels of blood urea nitrogen (BUN) in all groups increased in the rst 24h-post resuscitation but decreased in the remaining 6h-rewarming period.Compared to the NT group, the levels of BUN were signi cantly lower in the two hypothermic groups, with the lowest in the CRRT group at 6h,12h, 24h and 30h post resuscitation (all P<0.05).Conclusively, in the 24h-post resuscitation period, compared with the SC and NT group, renal injury in the CRRT group wasalleviated (all P< 0.05, Fig. 3).
In the tissues of kidney, compared with the NT group, the levels of TNF-α, IL-6were much lowerin two hypothermic groups, with the lowest in the CRRT group (all P< 0.05). With regard to the effect on oxidative stress,the lowest level of MDA and the highest level of SOD were observed in the CRRT group among three experiment groups (P < 0.05). In the same way, the proportions of apoptotic cells and the expressions of cleaved caspase-3 in the kidneyofthe two hypothermic groups were lower than the NT group, with the lowest in the CRRT group (P< 0.05). Conclusively, the coolinginduced by CRRT alleviated the tissue in ammation, oxidative stress as well as theextent of apoptosis in the kidney compared with SC (P< 0.05, Fig. 4).
Intestinal status was assessed using the levels of IFABP and DAO.After resuscitation, the values of IFABP increased in the rst 24h and decreased in the remaining 6h-rewarming period. At 24h after resuscitation, the serum levels of IFABP were the lowest in CRRT group, followed by the SC group and the NT group (P< 0.05).In addition, the levels of DAO in all groups showed tendency to ascend in the rst12h post resuscitation, followed to descendin the remaining 18hpost resuscitation.Compared to the NT group, the levels of DAO were signi cantly lower in two hypothermic groups, with the lowest in the CRRT group at 12h, 24h and 30h post resuscitation (all P<0.05).Conclusively, in the 24h-post resuscitation period, compared with the SC and NT group, intestinal injury was signi cantly alleviated in the CRRT group (all P < 0.05, Fig. 5).
In the tissues of intestine, compared with the NT group, the levels of TNF-α, IL-6 were much lower in the two hypothermic groups, with the lowest in the CRRT group (all P< 0.05). With regard to the effect on oxidative stress, the lowest level of MDA and the highest level of SOD were observed in the CRRT group among the three experiment groups (P< 0.05). In the same way, the proportions of apoptotic cells and the expression of cleaved caspase-3 in the intestine in the two hypothermic groups were lower than in the NT group, with the lowest in the CRRT group (P< 0.05). Conclusively,compared with SC, cooling induced by CRRT mitigated the tissue in ammation, oxidative stress and apoptosis of cells in the intestine (P< 0.05, Fig. 6).

Discussion
This present study focused on the potential organ protection of CRRT cooling against PCASbased on a swine model. We foundthe protective effect of CRRTcooling was signi cantly superior to SC from the following aspects: 1) hypothermia induced by CRRT was achieved signi cantly faster than SC;2) CRRT cooling signi cantly alleviated the renal and intestinal injury post resuscitation compared to SC; 3) with pathologic evidence, the protective effect of CRRT cooling on kidney and intestine were exertedby decreasing the in ammationof tissues, suppressing oxidative stressand decreasing the extent of apoptosis compared with SC. Laurent et al [32] found that the combination of CRRT with TH was feasible in CA patients, and might improve the prognosis of PCAS. Two cases reports demonstrated that CRRT cooling yielded better neurological recovery for CA and acute heart failure [26,27]. Therefore, cooling induced by CRRTmight offer an e cient way to produce THin the clinical practice. Based on a previous study,8-hr CRRT cooling followed by 16-hr SC was determined in our study to reduce the possibility of adverse events [32].And no adverse events were observed in the CRRTgroup. Thus, CRRT cooling might be a safe way to induce TH with high e ciency. Additionally, a review [19] demonstrated that the achievement of temperature below 34℃ within 3.5h after ROSC seemed to be bene cial. In the present study, the blood temperature was initially decreased rapidly by submerging the circuit in ice water to release heat, followed by the maintenance of TH by wrapping the circuit within an adjustable heating device.
Most importantly, the present study found that rapid therapeutic hypothermia induced by CRRT signi cantly mitigated the renal and intestinal injury after resuscitation with comparation with SC.Cr and BUN are two well-recognized markers of renal function. Previous review showed that higher Cr levels on admission was a signi cant predictorof AKI occurrence [2].The IFABP is a cytosolic protein, which will be released into the circulation when intestinal permeability increases, making it an e cient marker for intestinaldamage [33]. Evidence has shown a linear relation between max endotoxin level and max IFABP (p = 0.01), suggesting that the higher the intestinal injury, the higher the level of endotoxin [9].And DAO is an intracellular enzyme mainly located in intestinal mucosa, catalyzing oxidative deamination of diamines [34]. Recent evidence suggests that the risein plasma DAO level is strongly associated with the severity of intestinal tissue damage [35].Therefore, the activity of plasma DAO can manifest the intestinal status and will increase in case of intestinal ischemia [34]. In our study,with a smaller increase of levels of Cr, BUN, IFABP and DAO, potential renal and intestinal protection by CRRT cooling was proved.
Evidence from pathologic processes, including decreasing the in ammationof tissues, suppressing oxidative stress and decreasing the level of apoptosis,was observed in tissues of kidney and intestine in the CRRT group.According to a growing body of literature, TNF-α and IL-6 are typical indexes to manifest the in ammatory response of the whole body or certain organs [36]. With a smaller elevation of TNF-α and IL-6, CRRT cooling mighthave a protective in uence by decreasing tissue in ammation in the kidney and small intestine. And the levels of MDA and SOD in tissues were also determined in our study. MDA, a substance produced by reactive oxygen after degrading polyunsaturated lipids and may lead to intracellular toxic stress, is an indicator of oxidative stress, while SOD is a pivotal antioxidant in cells [13]. With a smaller elevation of MDA and larger elevation of SOD, CRRT cooling may alleviate the renal and intestinal injury by suppressing oxidative stress. Previous studiesfound that CRRT could not decrease systemic in ammation after resuscitation in both CA patients [32] and rat models [37]. Therefore, the organ protective effects in our study might be mostly ascribed to the fast induction of hypothermiaby CRRT cooling.
Inthe kidney and intestine among 3 experiment groups, the lowestproportions of apoptotic cells and lowest expression of cleaved caspase-3 were observed in CRRT group (P< 0.05), which contribute to strengthening the protective effect by hypothermia induced by CRRT on renal and intestinal injury.Caspase-3 is required for most of the proteolysis during apoptosis, and the cleaved caspase-3 has been used to represent the level of cell apoptosis [38]. Therefore, we can conclude that CRRT cooling might exert the organ protective effect by inhibitingthe process of apoptosis.
There were several limitations that should be stated. Firstly, a1℃/h rate of rewarming was used in our study based on previous results [39], whilerewarming shouldbe achieved at rate of about 0.25 ~ 0.5℃ per hourbased on the current guideline [40]. Second, the apparatus for cooling induced by CRRT was ready in advance, so the induction of TH wasstarted immediately ROSC was obtained. However, a longer duration of more than 15 min are required to perform CRRT in the clinical setting. Third, heparin was used as the anticoagulant in CRRT in our study, while the use of citrate is preferred as recommended in the Kidney Disease Improving Global (KDIGO) clinical practice because of its high e cacy and safety [41].A higher risk of bleeding exists when heparin is used [42].

Conclusions
Fast hypothermia induced by continuous renal replacement therapy was superior to surface cooling in mitigating renal and intestinal injurypostresuscitation.

Consent for publication
Not applicable.

Availability of data and materials
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.    The changes of Cr and BUN in different groups(note: except the Control group containing 5 swine, the other groups have 9 swine each). BL indicates baseline; BUN, blood urea nitrogen; Cr, creatinine; CRRT, continuous renal replacement therapy; DF, de brillation; NT, normothermia; PC, precordial compression; SC, surface cooling; VF, ventricular brillation. aP< 0.05 versus Control group; bP< 0.05 versus NT group; cP< 0.05 versus SC group.