E

Pss

JXa

b

c

a

ARRA

KPMSII

D

(

h0

Resuscitation 85 (2014) 833–839

Contents lists available at ScienceDirect

Resuscitation

j ourna l h o me pa g e : www.elsev ier .com/ locate / resusc i ta t ion

xperimental paper

ost-resuscitation intestinal microcirculation: Its relationship withublingual microcirculation and the severity of post-resuscitationyndrome�,��

ie Qiana,c, Zhengfei Yanga, Jena Cahoona, Jiefeng Xua, Changqing Zhuc,∗, Min Yanga,ianwen Hua, Shijie Suna,b, Wanchun Tanga,b,∗∗

Weil Institute of Critical Care Medicine, Rancho Mirage, CA, United StatesKeck School of Medicine of the University of Southern California, Los Angeles, CA, United StatesRenji Hospital, Shanghai Jiaotong University School of Medicine, Shanghai, China

r t i c l e i n f o

rticle history:eceived 12 November 2013eceived in revised form 10 January 2014ccepted 20 February 2014

eywords:ost-resuscitation syndromeicrocirculation

ublingualntestinenflammatory response

a b s t r a c t

Objective: Post-resuscitation syndrome has been recognized as one of the major causes of the pooroutcomes of cardiopulmonary resuscitation. The aims of this study were to investigate the intestinalmicrocirculatory changes following cardiopulmonary resuscitation and relate those changes to sublingualmicrocirculation and the severity of post-resuscitation syndrome as measured by myocardial functionand serum inflammatory cytokine levels.Methods: Twenty-five rats were randomized into three groups: (1) short duration of cardiac arrest (n = 10):ventricular fibrillation (VF) was untreated for 4 min prior to 6 min of cardiopulmonary resuscitation (CPR);(2) long duration of cardiac arrest (n = 10): VF was untreated for 8 min followed by 8 min of CPR; (3) shamcontrol group (n = 5): a sham operation was performed without VF induction and CPR. Intestinal andsublingual microcirculatory blood flow was visualized by a sidestream dark-field (SDF) imaging deviceat baseline and 1, 2, 4, 6, 8 h post-resuscitation. Myocardial function was measured by echocardiographyand serum cytokine levels (TNF-� and IL-6) were measured by enzyme-linked immunosorbent assay(ELISA).Results: Both intestinal and sublingual microcirculatory blood flow decreased significantly with increasingduration of cardiac arrest and resuscitation. The decreases in intestinal microcirculatory blood flow wereclosely correlated with the reductions of sublingual microcirculatory blood flow (perfused small vesselsdensity: r = 0.772, p < 0.01; microcirculatory flow index: r = 0.821, p < 0.01). The decreased microcircula-tory blood flow was closely correlated with weakened myocardial function and elevated inflammatory

cytokine levels.Conclusions: The severity of post-resuscitation intestinal microcirculatory dysfunction is closely corre-lated with that of myocardial function and inflammatory cytokine levels. The measurement of sublingualmicrocirculation reflects changes of intestinal microcirculation and may therefore provide a new optionfor post-resuscitation monitoring.

© 2014 Elsevier Ireland Ltd. All rights reserved.

� A Spanish translated version of the abstract of this article appears as Appendix in the

�� Protocol number: R1304.∗ Corresponding author at: Department of Emergency Medicine, Renji Hospital, Shanistrict, Shanghai 200127, China.

∗∗ Corresponding author at: Weil Institute of Critical Care Medicine, 35100 Bob Hope DrE-mail addresses: Qjssmu@gmail.com (J. Qian), Yangzhengfei@vip.163.com (Z. Yang), Je

C. Zhu), 456ym@163.com (M. Yang), Xianwenhu001@sina.cn (X. Hu), Shijie.sun@aol.com

ttp://dx.doi.org/10.1016/j.resuscitation.2014.02.019300-9572/© 2014 Elsevier Ireland Ltd. All rights reserved.

final online version at http://dx.doi.org/10.1016/j.resuscitation.2014.02.019.

ghai JiaoTong University School of Medicine, 1630# Dongfang Road, Pudong New

ive, Rancho Mirage, CA 92270, United States.na.cahoon@weiliccm.org (J. Cahoon), Jeff.xu11@gmail.com (J. Xu), Zhucq@126.com

(S. Sun), Wanchun.tang@me.com (W. Tang).

8 itation

1

sdPm(fsdflibomto

tldtdscw

oircpagi

2

cfAaWwo

2

ww(h

(wwlTNtdt

34 J. Qian et al. / Resusc

. Introduction

Though approximately 50% of cardiac arrest (CA) victims areuccessfully resuscitated initially, only 5–15% survive to hospitalischarge.1,2 The prognosis after CA and resuscitation remains poor.ost-resuscitation syndrome is a complex state characterized byyocardial dysfunction, brain injury, global ischemia–reperfusion

I/R) injury and systemic inflammatory response. It accounts partlyor the poor outcome.3 The intestine is likely to be the most sen-itive to I/R injury among the internal organs. A previous studyemonstrated that CA caused a prolonged reduction of bloodow to the intestine.4 As a result, the intestinal permeability

ncreases and a systemic inflammatory response may subsequentlye triggered.5 This has been considered an important mechanismf sepsis.6 The elevated inflammatory cytokines after cardiopul-onary resuscitation mimics that of sepsis.7 However, whether

he intestinal circulatory dysfunction contributes to the severityf post-resuscitation syndrome remains unclear.

Current evidence indicated that microcirculation regulatesissue blood flow and is therefore more important than macrocircu-ation on tissue oxygen delivery and utilization.8,9 Microcirculatoryysfunction has been recognized as an important determinant ofhe outcome in circulatory shock.10,11 Microcirculatory changesiffer extensively among organs especially during the low-flowtates.4,9,12–14 The sublingual area is easy to access and has beenonsidered an ideal location to evaluate microcirculation. However,hether it fully reflects visceral microcirculation remains unclear.

In the present study, we investigated the concurrent changesf intestinal and sublingual microcirculation, cardiac function andnflammatory cytokine levels in a rat model of cardiopulmonaryesuscitation (CPR). Our hypotheses were that: (1) Intestinal micro-irculatory dysfunction is closely correlated with the severity ofost-resuscitation syndrome as measured by myocardial functionnd inflammatory cytokine levels. (2) The changes of sublin-ual microcirculation during post-resuscitation reflect changes ofntestinal microcirculation.

. Materials and methods

All animals received humane care in compliance with the Prin-iples of Laboratory Animal Care formulated by the National Societyor Medical Research and the Guide for the Care and Use of Laboratorynimals prepared by the Institute of Laboratory Animal Resourcesnd published by the National Institutes of Health (8th edition,ashington, DC, National Academies Press, 2011). The protocolas approved by the Institutional Animal Care and Use Committee

f the Weil Institute of Critical Care Medicine.

.1. Animal preparation

Twenty-five male Sprague–Dawley rats weighing 450–550 gere fasted overnight except for free access to water. The animalsere anesthetized by intraperitoneal injection of pentobarbital

45 mg/kg). Additional doses (10 mg/kg) were administered atourly intervals.

The trachea was orally intubated with a 14-gauge cannulaAbbocath-T; Abbott Hospital, North Chicago, IL). The animalsere breathing room-air spontaneously. End-tidal CO2 (ETCO2)as continuously monitored with a sidestream infrared CO2 ana-

yzer (model 200; Instrumentation Laboratories, Lexington, MA).wo 23-gauge PE-50 catheters (Becton-Dickinson, Franklin Lakes,

J) were advanced through the left external jugular vein into

he right atrium and through the left femoral artery into theescending aorta for measurement of right atrial pressure and aor-ic pressure with high-sensitivity transducers (model 42584-01;

85 (2014) 833–839

Abbott Critical Care Systems, North Chicago, IL). A thermocouplemicroprobe, 10 cm in length and 0.5 mm in diameter (9030-12-D-34; Columbus Instruments, Columbus, OH), was inserted intothe left femoral vein to measure blood temperature. A 3F PEcatheter (model C-PMS-301J; Cook Critical Care, Bloomington, IN)was advanced through the right external jugular vein into theright atrium. All catheters were flushed intermittently with salinecontaining 2.5 IU/mL crystalline bovine heparin. A conventionallead II ECG was continuously monitored. The blood temperaturefor all animals was maintained at 37 ◦C ± 0.5 ◦C with a heatinglamp.

A laparotomy was performed to expose the peritoneal cavitythrough the midline abdominal incision (∼1.5 cm). The wound wascovered by sterile 37 ◦C warmed normal saline-saturated gauze tominimize dehydration and loss of body heat.

2.2. Experimental procedures

The animals were randomly assigned to one of three groups:(1) short duration of CA (n = 10): ventricular fibrillation (VF) wasuntreated for 4 min prior to 6 min of cardiopulmonary resusci-tation (CPR); (2) long duration of CA (n = 10): VF was untreatedfor 8 min followed by 8 min of CPR; (3) sham control group(n = 5): a sham operation was performed without VF induction andCPR.

Ten minutes prior to inducing VF, baseline measurements wereobtained and mechanical ventilation was initiated at a tidal vol-ume of 0.60 ml/100 g body weight, a frequency of 100 breaths/minand an inspired O2 fraction (FiO2) of 0.21. VF was electricallyinduced with a progressive increase in 60-Hz current to a max-imum of 3.5 mA delivered to the right ventricular endocardium.The current flow was continued for 3 min to prevent sponta-neous defibrillation. Mechanical ventilation was discontinued afteronset of VF. Precordial compression (PC) was started after either4 or 8 min of VF. Coincident with the start of PC, the animalswere mechanically ventilated at a frequency of 100 breaths perminute and with an FiO2 of 1.0. PC was maintained at a rate of200/min and synchronized to provide a compression/ventilationratio of 2:1 with equal compression–relaxation. The depth of com-pressions was initially adjusted to maintain a coronary perfusionpressure (CPP: diastolic aortic pressure minus right atrial pres-sure) at 22 ± 2 mmHg with ETCO2 11 ± 2 mmHg. Resuscitation wasattempted with up to 3 two-joule countershocks. If return of spon-taneous circulation (ROSC) was not achieved, a 30-s interval of CPRwas performed before a subsequent sequence of up to 3 shockswas attempted. This procedure was repeated for a maximum of 3cycles. ROSC was defined as the return of supraventricular rhythmwith a mean aortic pressure above 50 mmHg for a minimum of5 min. Following ROSC, mechanical ventilation was continued with100% oxygen for the first hour, 50% for the second hour and21% thereafter. There was no pharmacology agent used duringCPR.

After the experiment, all animals were euthanized by anoverdose of pentobarbital (150 mg/kg via the femoral artery).Abdominal cavity culture was performed to prove the laparotomywas sterile. A necropsy was performed to document injuries tothoracic, abdominal vessels and viscera caused by surgical inter-vention.

2.3. Measurements

Aortic and right atrial pressures, electrocardiogram and ETCO2

were continuously recorded for 8 h on a personal computer-baseddata acquisition system (DATAQ Instruments, Akron, OH).

Myocardial function, including cardiac output (CO), ejectionfraction (EF) and myocardial performance index (MPI), were

itation 85 (2014) 833–839 835

art

4trbiwabbTf

8idwowiwAcaaicfaDlednwtwddwr

2

(odgvimuc

3

i

eb

Table 1Baseline characteristics.

Group Sham control Short durationof CA

Long durationof CA

Body weight, g 510 ± 24 506 ± 15 503 ± 14Heart rate, bpm 346 ± 21 356 ± 28 357 ± 24MAP, mmHg 135 ± 6 133 ± 10 136 ± 8End-tidal CO2, mmHg 41 ± 1 41 ± 1 41 ± 1Ejection fraction, % 70.0 ± 1.0 69.5 ± 1.5 69.5 ± 1.7Cardiac output, mL/min 107.0 ± 5.2 107.3 ± 2.6 107.6 ± 4.1MPI 0.71 ± 0.04 0.69 ± 0.04 0.68 ± 0.03pH 7.48 ± 0.03 7.51 ± 0.03 7.50 ± 0.04Lactate, mmol/L 0.9 ± 0.3 0.8 ± 0.1 0.9 ± 0.2

J. Qian et al. / Resusc

ssessed at baseline, 1, 2, 4, 6 and 8 h after ROSC using echocardiog-aphy (HD 11 XE, Philips Healthcare, Andover, MA) with a 12.5 MHzransducer.15

Artery blood samples (0.3 ml) were withdrawn at baseline, 2,, 6 and 8 h after ROSC for measuring arterial blood gases withhe aid of a Stat Profile pHOx analyzer (Nova Biomedical Corpo-ation, Waltham, MA). For the measurement of serum cytokine,lood samples (1.2 ml) were collected at the same time points and

mmediately centrifuged at 3000 rpm for 10 min. Then, the serumas stored until analysis. The experimental rat received an equiv-

lent volume (1.5 ml) of arterial blood from a donor rat after thelood withdrawal. TNF-� and IL-6 concentrations were determinedy using ELISA kits (TNF-�: R&D Systems, Minneapolis, MN; IL-6:hermo Scientific, Rockford, IL). The experiments were performedollowing the manufacturer’s instructions.

Microcirculations were visualized at baseline, 1, 2, 4, 6 and h after ROSC with the aid of a sidestream dark-field (SDF)

maging device (MicroScan; MicroVision Medical Inc., Amster-am, Netherlands) with a 5× optical probe. The sublingual regionas assessed near the base of the tongue. For the measurements

f intestinal microcirculation, a 2–3 cm segment of the jejunumas withdrawn with its neurovascular supply intact and cush-

oned with warm saline-soaked gauze. The jejunal microcirculationas assessed on the anti-mesenteric aspect of the serosal side.fter microcirculatory measurements were taken, the abdominalontents were then returned into the peritoneal cavity and thebdomen was closed in two layers. The muscular layer of thebdominal wall was sutured in a continuous pattern, and the skinncision was closed with wound clips. Three discrete fields wereaptured with precaution to minimize motion and pressure arti-acts. Microvascular images were recorded on a DVD disk using

DVD recorder (Model DMREZ47V; Panasonic AVC Networks,alian, China). Individual images were analyzed offline. Microcircu-

atory flow index (MFI) was quantitated by the method of Boermat al.16 The image was divided into four quadrants and the pre-ominant type of flow (absent = 0, intermittent = 1, sluggish = 2 andormal = 3) was assessed in the small vessels of each quadrant,hich were less than 20 �m in diameter. The MFI score represented

he average values of four quadrants. Perfused vessel density (PVD)as measured based on the method of De Backer et al.17 Vesselensity was calculated as the number of vessels crossing the linesivided by the total length of the lines. Vessel size was measuredith a micrometer scale superimposed in the video display. All

ecordings were analyzed by three independent observers.

.4. Statistical analyses

Measurements were reported as mean ± SD or medianinterquartile range) as indicated. Variables were compared withne-way ANOVA or the Kruskal–Wallis test for nonparametricata. Comparisons between time-based measurements within eachroup were performed with repeated-measurement analysis ofariance. If there was a significant difference in the overall compar-son of groups, comparisons between any other two groups were

ade by the Bonferroni test. Linear correlations were calculatedsing the Pearson correlation coefficient. A value of p < 0.05 wasonsidered significant.

. Results

Baseline hemodynamics, myocardial function and blood analyt-

cal measurements did not differ among groups (Table 1).

All animals were successfully resuscitated. There was no differ-nce in the amount of the electric current required to induce VFetween CPR groups. Body temperature did not differ significantly

CA, cardiac arrest; MAP, mean aortic pressure; MPI, myocardial performance index.Values are presented as mean ± SD.

among groups during the experiment. There was no significant dif-ference of CPP between the CPR groups during CPR. CPP at the endof CPR was 25.0 ± 3.2 mmHg in the short duration group and was24.5 ± 2.9 mmHg in the long duration group. The number of defib-rillations did not differ significantly (1.4 ± 0.5 vs. 2.1 ± 2.2) betweenthe short and the long duration groups.

The post-resuscitation PVD decreased in parallel with thereduced MFI in both intestinal and sublingual sites. In the long dura-tion group, intestinal PVD was significantly reduced from baselineof 19.8 ± 1.7 to 10.1 ± 1.8 n/mm at 8 h post-resuscitation and MFIfrom 3.0 ± 0.0 to 1.4 ± 0.7 (both p < 0.05 vs. baseline and sham con-trol). Similarly, sublingual PVD was significantly reduced from thebaseline of 5.6 ± 0.8 to 2.3 ± 0.6 n/mm at 8 h post-resuscitation andMFI from 3.0 ± 0.0 to 1.3 ± 0.6 (both p < 0.05 vs. baseline and shamcontrol). Contrary to a progressive recovery of both measurementsobserved in the short duration group, the significant reductionin both PVD and MFI persisted after resuscitation in the longduration group. At 8 h post-resuscitation, PVD and MFI were signif-icantly lower in the long duration group than in the short durationgroup: intestinal PVD (10.1 ± 1.8 vs. 17.7 ± 2.8 n/mm, p < 0.05);sublingual PVD (2.3 ± 0.6 vs. 4.8 ± 0.9 n/mm, p < 0.05); intestinalMFI (1.4 ± 0.7 vs. 2.4 ± 0.4, p < 0.05); sublingual MFI (1.3 ± 0.6 vs.2.4 ± 0.5, p < 0.05). The decreases in intestinal microcirculatoryblood flow were closely correlated with the reductions of sub-lingual microcirculatory blood flow (PVD: r = 0.772, p < 0.01; MFI:r = 0.821, p < 0.01) (Fig. 1).

Examples of the images of intestinal and sublingual microvas-culature, as obtained by the SDF video microscope at 8 hpost-resuscitation, are shown in Fig. 2.

Myocardial function was significantly impaired in the post-resuscitation period in both CPR groups compared with baselinevalues. Significantly worse post-resuscitation myocardial functionwas observed after long duration of CA when compared to the shortduration of CA group which gradually recovered over time (Fig. 3).

Serum cytokine levels in both CPR groups elevated after resus-citation. There was significantly elevated TNF-� levels in the longduration group 2 h after resuscitation. In the short duration group,there was a significant elevation only 8 h after resuscitation. How-ever, for IL-6, except in the long duration group at 6 and 8 hpost-resuscitation, none were statistically different from the shamcontrol group (Fig. 4).

There were significant correlations between MFI, PVD, myocar-dial function and serum cytokine levels (Table 2).

At 8 h post-resuscitation, ETCO2 values in the long durationgroup were significantly lower than those of the sham control andthe short duration group. Plasma lactate levels in the long dura-tion group remained significantly greater when compared with

the sham control and the short duration group (p < 0.05). How-ever, there was no significant difference in heart rate or mean aorticpressure (MAP) between the CPR groups (Table 3).

836 J. Qian et al. / Resuscitation 85 (2014) 833–839

F al andfl vs. sh

4

limi

TCa

PEf

ig. 1. Changes of microcirculatory variables and the correlation between sublinguow index; PVD, perfused vessel density. *p < 0.05, vs. sham control group. #p < 0.05

. Discussion

The present study demonstrated that both intestinal and sub-

ingual microcirculatory blood flow decreased significantly withncreasing duration of cardiac arrest and resuscitation. Intestinal

icrocirculatory dysfunction is closely correlated with the sever-ty of post-resuscitation syndrome as measured by myocardial

able 2orrelational analysis between myocardial function, microcirculatory parametersnd serum cytokine levels.

Sublingual Intestinal

PVD MFI PVD MFI

Cardiac functionCO 0.827* 0.846* 0.821* 0.839*

EF 0.846* 0.823* 0.767* 0.818*

MPI −0.794* −0.775* −0.750* −0.757*

Cytokine levelsTNF-� −0.332* −0.380* −0.420* −0.307*

IL-6 −0.415* −0.394* −0.406* −0.326*

VD, perfused vessel density; MFI, microcirculatory flow index; CO, cardiac output;F, ejection fraction; MPI, myocardial performance index; TNF-�, tumor necrosisactor-alpha; IL-6, interleukin-6.

* p < 0.01

intestinal microcirculation. BL, baseline; CA, cardiac arrest; MFI, microcirculatoryort duration of CA group.

function and inflammatory cytokine levels. Sublingual microcircu-latory changes were closely correlated with that of intestinal.

In humans, the easiest accessible site to detect microcirculationis the sublingual area. Sublingual microcirculation is considered asa surrogate measure for splanchnic blood flow. Under conditionsof circulatory shock, noninvasive sublingual capnometry yieldedmeasurements that were interchangeable with those of gastric

tonometry.18–20 Using a direct visualization approach, our find-ings demonstrated the similar behavior of sublingual and intestinalmicrocirculation at the early stage after ROSC. Our findings are

Table 3Characteristics at 8 h post-resuscitation.

Group Sham control Short durationof CA

Long durationof CA

Heart rate, bpm 337 ± 27 319 ± 51 342 ± 28MAP, mmHg 126 ± 4 104 ± 13* 97 ± 17*

End-tidal CO2, mmHg 40 ± 1 36 ± 2* 29 ± 6*,#

pH 7.54 ± 0.02 7.54 ± 0.04 7.51 ± 0.09Lactate, mmol/L 1.1 ± 0.2 1.3 ± 0.4 2.4 ± 0.9*,#

CA, cardiac arrest; MAP indicates mean aortic pressure; Values are presented asmean ± SD.

* p < 0.05 vs. sham control group.# p < 0.05 vs. short duration of CA group.

J. Qian et al. / Resuscitation 85 (2014) 833–839 837

y the s

caasstbtdtetm

otdsdvwmeomavolocisbt

Fig. 2. Images of sublingual and intestinal microcirculation obtained b

onsistent with earlier studies not only of circulatory shock butlso of endotoxic and septic shock, which indicates that the severitynd the time course of microcirculatory changes were similar in theublingual and gut region.21–24 Interestingly, two studies on sep-is demonstrated conflicting findings on the relationship betweenhese two sites. Though both revealed time-varying correlationsetween these two sites, a clinical study claimed that microcircula-ory changes were not parallel on day 1 while another animal studyemonstrated that the initial strong correlation disappeared overime.25,26 Highly heterogeneous nature of the sepsis victims mightxplain the conflicting results. Additionally, both studies appliedherapeutic interventions that may further influence the intestinal

icrocirculation.27,28

Although post-resuscitation abnormalities mimic the disordersbserved in sepsis, cardiac dysfunction is one of the primary dis-inguishing characteristics. The impairment of cardiac functionuring the post-resuscitation period can precipitate a downwardplanchnic blood flow. However, microcirculatory changes can beispatched from changes in regional blood flow and differ amongisceral organs especially in low-flow states.29 In the present study,e revealed a close correlation between cardiac function andicrocirculatory changes in sublingual and intestinal sites in the

arly stage following resuscitation. Noticeably, in the long durationf the CA group, cardiac function deteriorated over time, intestinalicrocirculatory parameters experienced a slightly upward trend

fter PR 1 h and then dropped to approximately 50% of baselinealue at PR 8 h. The autoregulation may account for the fluctuationf microcirculatory blood flow. In fact, the regulation of microcircu-atory blood flow in the intestine wall is extremely complicated. Notnly systemic factors but also local factors such as inflammatoryytokines, vasomotor function and metabolic products affect the

ntestinal microcirculation. Under this experimental condition, wepeculate that the intestinal microcirculatory dysfunction causedy systemic insults was beyond the limit of the autoregulatoryhreshold.

idestream dark-field (SDF) video microscope at 8 h post-resuscitation.

The severity of intestinal microcirculatory dysfunction wasmagnified when the duration of cardiac arrest and resuscitationwas prolonged. In the present study, intestinal microcirculationwas impaired in both CPR groups. A partially reversible injurywithin 8 h was observed in the short duration group. This pattern ofresponse was consistent with previous studies that demonstrated atransition from reversible to irreversible intestinal damage as dura-tion of ischemia prolonged.30 In Korth et al.’s study, 4 min of cardiacarrest caused a resembled reduction of intestinal blood flow dur-ing the following 60 min. Though they did not observe blood flowthereafter, this phenomenon that the intestinal metabolic changeswere gradually restored over time may indicate a resembled recov-ery of intestinal microcirculation.4

There may be several factors contributing to the fact thatprolonged cardiac arrest leads to more severe intestinal micro-circulatory derangement. First, as mentioned above, prolonged CAresults in more pronounced myocardial dysfunction, the intestinalmicrocirculation is therefore more likely to be injured as a result ofreduced cardiac output. Second, within 4–6 min of CA, high-energyphosphates become depleted, thus favor disruption of microvascu-lar endothelial cells.31 Third, longer duration of CA may result inlocal intestinal thrombosis and leukocyte plugging, thus causes thedecrease in capillary flow.32

Intestinal barrier integrity plays a pivotal role in the develop-ment of systemic inflammatory response syndrome, sepsis andmultiple organ failure. The intestine is not only a vulnerable tar-get to ischemia but a possible secondary source of inflammatorycytokines as well. We cannot rule out the possibility that post-resuscitation disturbed intestinal microcirculation may trigger aninflammatory cascade. IL-6 as well as TNF-� are key mediatorsof the systemic inflammatory response. Our results confirmed an

intense increase of such cytokines in the early hours after CA.7 Thedecreases of microcirculatory blood flow were correlated with theincreases of inflammatory cytokine levels. In our study, a discrepanttime course of TNF-� and IL-6 levels after different durations of CA

838 J. Qian et al. / Resuscitation 85 (2014) 833–839

Fig. 3. Myocardial function at baseline and post-resuscitation. EF, ejection fraction;CO, cardiac output; MPI, myocardial performance index; BL, baseline; CA, cardiacas

wRa�pgI

uciofslmlft

si

rrest; VF, ventricular fibrillation. Values are presented as mean ± SD. *p < 0.05, vs.ham control group. #p < 0.05 vs. short duration of CA group.

ere observed. The TNF-� level remained at high values 2 h afterOSC in the long duration of the CA group while there was a 6 h lagfter short duration. IL-6 expression, which can be induced by TNF-, elevated at 6 h post-resuscitation after 8 min of CA. Although theresent study was not designed to compare survival rates amongroups, several studies have confirmed high levels of TNF-� andL-6 could discriminate between survivors and nonsurvivors.7,33

Clinically, intestine injuries after cardiac arrest are oftennderestimated though it might be a potentially devastatingomplication even following successful resuscitation.34 The man-festations may be nonspecific and delayed. The direct in vivobservation of the intestinal microcirculation is not clinicallyeasible and making early detections is therefore difficult. Ourtudy demonstrated that the changes of sublingual microcircu-ation during post-resuscitation reflected changes of intestinal

icrocirculation, myocardial function and inflammatory cytokineevels. This suggests that sublingual microcirculation may there-ore provide a new option to monitor post-resuscitation patients at

he bedside.

To interpret the results of our experimental study, it is neces-ary to take several limitations into consideration. First, therapeuticntervention was not applied in the study, which is different from

Fig. 4. Serum TNF-� and IL-6 concentrations at baseline and post-resuscitation(mean ± SD). *p < 0.05, vs. sham control group. #p < 0.05 vs. short duration of CAgroup.

clinical situations. Microcirculation can be affected by several fac-tors. The use of epinephrine, hypothermia or fluid therapy mayhave reduced differences in microcirculation parameters betweengroups. Second, we investigated the microvasculature of serosalayer rather than the mucosa layer. It is likely that microcirculatorychanges in these two sites are different. Third, we didn’t investi-gate the neurologic function which is one important component ofpost-resuscitation syndrome. Fourth, using SDF imaging, PVD andMFI were evaluated in a semiquantitative way. A quantitative anal-ysis might generate a more accurate outcome. Fifth, movement andpressure artifacts hinder the analysis of the images though we triedto reduce the interference. Finally, we did not measure the abdomi-nal pressures which might affect the microcirculatory blood flow inthe intestinal wall.35 However, a sham control group was enrolledto rule out this factor.

5. Conclusion

The severity of post-resuscitation intestinal microcirculatorydysfunction is closely correlated with that of myocardial functionand inflammatory cytokine levels. The measurement of sublingualmicrocirculation reflects changes of intestinal microcirculation andmay therefore provide a new option for post-resuscitation moni-toring.

Conflict of interest

None of the authors have any conflict of interest to report.

Acknowledgement

Lisa Luna contributed to the editing of this manuscript.

Appendix A. Supplementary data

Supplementary material related to this article can be found,in the online version, at http://dx.doi.org/10.1016/j.resuscitation.2014.02.019.

itation

R

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

2

2

2

3

3

3

3

3

J. Qian et al. / Resusc

eferences

1. Goldberger ZD, Chan PS, Berg RA, et al. Duration of resuscitation effortsand survival after in-hospital cardiac arrest: an observational study. Lancet2012;380:1473–81.

2. Nichol G, Thomas E, Callaway CW, et al. Regional variation in out-of-hospitalcardiac arrest incidence and outcome. JAMA 2008;300:1423–31.

3. Neumar RW, Nolan JP, Adrie C, et al. Post-cardiac arrest syndrome: epidemiol-ogy, pathophysiology, treatment, and prognostication. A consensus statementfrom the international liaison committee on resuscitation (American heartassociation, Australian and new Zealand council on resuscitation, Europeanresuscitation council, heart and stroke foundation of Canada, InterAmericanheart foundation, resuscitation council of Asia, and the resuscitation councilof southern Africa); the American heart association emergency cardiovascularcare committee; the council on cardiovascular surgery and anesthesia; the coun-cil on cardiopulmonary, perioperative, and critical care; the council on clinicalcardiology; and the stroke council. Circulation 2008;118:2452–83.

4. Korth U, Krieter H, Denz C, et al. Intestinal ischaemia during cardiac arrest andresuscitation: comparative analysis of extracellular metabolites by microdialy-sis. Resuscitation 2003;58:209–17.

5. Stallion A, Kou TD, Latifi SQ, et al. Ischemia/reperfusion: a clinically rele-vant model of intestinal injury yielding systemic inflammation. J Pediatr Surg2005;40:470–7.

6. Coopersmith CM, Stromberg PE, Dunne WM, et al. Inhibition of intestinal epithe-lial apoptosis and survival in a murine model of pneumonia-induced sepsis.JAMA 2002;287:1716–21.

7. Adrie C, Adib-Conquy M, Laurent I, et al. Successful cardiopulmonary resus-citation after cardiac arrest as a sepsis-like syndrome. Circulation 2002;106:562–8.

8. van Genderen ME, Lima A, Akkerhuis M, Bakker J, van Bommel J. Persis-tent peripheral and microcirculatory perfusion alterations after out-of-hospitalcardiac arrest are associated with poor survival. Crit Care Med 2012;40:2287–94.

9. Fries M, Tang W, Chang YT, Wang J, Castillo C, Weil MH. Microvascular blood flowduring cardiopulmonary resuscitation is predictive of outcome. Resuscitation2006;71:248–53.

0. Top AP, Ince C, de Meij N, van Dijk M, Tibboel D. Persistent low microcirculatoryvessel density in nonsurvivors of sepsis in pediatric intensive care. Crit Care Med2011;39:8–13.

1. De Backer D, Donadello K, Sakr Y, et al. Microcirculatory alterations in patientswith severe sepsis: impact of time of assessment and relationship with outcome.Crit Care Med 2013;41:791–9.

2. Kern KB, Zuercher M, Cragun D, et al. Myocardial microcirculatory dysfunc-tion after prolonged ventricular fibrillation and resuscitation. Crit Care Med2008;36:S418–21.

3. Yu J, Ramadeen A, Tsui AK, et al. Quantitative assessment of brain microvas-cular and tissue oxygenation during cardiac arrest and resuscitation in pigs.Anaesthesia 2013;68:723–35.

4. Donadello K, Favory R, Salgado-Ribeiro D, et al. Sublingual and muscularmicrocirculatory alterations after cardiac arrest: a pilot study. Resuscitation2011;82:690–5.

5. Ye S, Weng Y, Sun S, et al. Comparison of the durations of mild therapeutic

hypothermia on outcome after cardiopulmonary resuscitation in the rat. Circu-lation 2012;125:123–9.

6. Boerma EC, Mathura KR, van der Voort PH, Spronk PE, Ince C. Quantifyingbedside-derived imaging of microcirculatory abnormalities in septic patients: aprospective validation study. Crit Care 2005;9:R601–6.

3

85 (2014) 833–839 839

7. De Backer D, Creteur J, Preiser JC, Dubois MJ, Vincent JL. Microvascular blood flowis altered in patients with sepsis. Am J Respir Crit Care Med 2002;166:98–104.

8. Povoas HPWM, Tang W, Moran B, Kamohara T, Bisera J. Comparisonsbetween sublingual and gastric tonometry during hemorrhagic shock. Chest2000;118:1127–32.

9. Weil MH, Nakagawa Y, Tang W, et al. Sublingual capnometry: a new noninvasivemeasurement for diagnosis and quantitation of severity of circulatory shock. CritCare Med 1999;27:1225–9.

0. Creteur J, De Backer D, Sakr Y, Koch M, Vincent JL. Sublingual capnome-try tracks microcirculatory changes in septic patients. Intensive Care Med2006;32:516–23.

1. Verdant CL, De Backer D, Bruhn A, et al. Evaluation of sublingual and gutmucosal microcirculation in sepsis: a quantitative analysis. Crit Care Med2009;37:2875–81.

2. Guzman JA, Dikin MS, Kruse JA. Lingual, splanchnic, and systemic hemodynamicand carbon dioxide tension changes during endotoxic shock and resuscitation.J Appl Physiol 2005;98:108–13.

3. Dubin A, Edul VS, Pozo MO, et al. Persistent villi hypoperfusion explains intra-mucosal acidosis in sheep endotoxemia. Crit Care Med 2008;36:535–42.

4. Fries M, Weil MH, Sun S, et al. Increases in tissue pco2 during circula-tory shock reflect selective decreases in capillary blood flow. Crit Care Med2006;34:446–52.

5. Boerma EC, vdVP, Spronk PE, Ince C. Relationship between sublingual and intesti-nal microcirculatory perfusion in patients with abdominal sepsis. Crit Care Med2007;35:1055–60.

6. Pranskunas A, Pilvinis V, Dambrauskas Z, et al. Early course of microcircula-tory perfusion in eye and digestive tract during hypodynamic sepsis. Crit Care2012;16:R83.

7. Oktar BK, Gulpinar MA, Bozkurt A, et al. Endothelin receptor blockers reduce i/r-induced intestinal mucosal injury: role of blood flow. Am J Physiol GastrointestLiver Physiol 2002;282:G647–55.

8. Nacul FE, Guia IL, Lessa MA, Tibirica E. The effects of vasoactive drugs onintestinal functional capillary density in endotoxemic rats: intravital video-microscopy analysis. Anesth Analg 2010;110:547–54.

9. Hiltebrand LB, Krejci V, Banic A, Erni D, Wheatley AM, Sigurdsson GH. Dynamicstudy of the distribution of microcirculatory blood flow in multiple splanchnicorgans in septic shock. Crit Care Med 2000;28:3233–41.

0. Guan Y, Worrell RT, Pritts TA, Montrose MH. Intestinal ischemia–reperfusioninjury: reversible and irreversible damage imaged in vivo. Am J Physiol Gas-trointest Liver Physiol 2009;297:G187–96.

1. Padosch SA, Teschendorf P, Fuchs A, et al. Effects of abciximab on postresus-citation microcirculatory dysfunction after experimental cardiac arrest in rats.Resuscitation 2010;81:255–9.

2. Schoots IG, Levi M, Roossink EH, Bijlsma PB, van Gulik TM. Local intravascularcoagulation and fibrin deposition on intestinal ischemia–reperfusion in rats.Surgery 2003;133:411–9.

3. Michael A, Basha G, Scott M, Steven LK, Robert MS, Emanuel PR, et al. Presenceof tumor necrosis factor in humans undergoing cardiopulmonary resuscitationwith return of spontaneous circulation. J Crit Care 1991;6:185–9.

4. Katsoulis IE, Balanika A, Sakalidou M, Gogoulou I, Stathoulopoulos A, DigalakisMK. Extensive colonic necrosis following cardiac arrest and successful car-diopulmonary resuscitation: report of a case and literature review. World J

Emerg Surg 2012;7:35.

5. Samel ST, Neufang T, Mueller A, Leister I, Becker H, Post S. A new abdominalcavity chamber to study the impact of increased intra-abdominal pressure onmicrocirculation of gut mucosa by using video microscopy in rats. Crit Care Med2002;30:1854–8.