A 10 min “no-touch” time – is it enough in DCD? A DCD Animal Study

Authors


  • Conflicts of Interest
    None.

Vanessa Stadlbauer, Department of Internal Medicine, Medical University of Graz, Auenbruggerplatz 29, 8036 Graz, Austria. Tel.: 0316 385 82282; fax: +43 316 385 17560; e-mail: vanessa.stadlbauer@medunigraz.at

Abstract

Summary Donation after cardiac death (DCD) is under investigation because of the lack of human donor organs. Required times of cardiac arrest vary between 75 s and 27 min until the declaration of the patients’ death worldwide. The aim of this study was to investigate brain death in pigs after different times of cardiac arrest with subsequent cardiopulmonary resuscitation (CPR) as a DCD paradigm. DCD was simulated in 20 pigs after direct electrical induction of ventricular fibrillation. The “no-touch” time varied from 2 min up to 10 min; then 30 min of CPR were performed. Brain death was determined by established clinical and electrophysiological criteria. In all animals with cardiac arrest of at least 6 min, a persistent loss of brainstem reflexes and no reappearance of bioelectric brain activity occurred. Reappearance of EEG activity was found until 4.5 min of cardiac arrest and subsequent CPR. Brainstem reflexes were detectable until 5 min of cardiac arrest and subsequent CPR. According to our experiments, the suggestion of 10 min of cardiac arrest being equivalent to brain death exceeds the minimum time after which clinical and electrophysiological criteria of brain death are fulfilled. Therefore shorter “no-touch” times might be ethically acceptable to reduce warm ischemia time.

Introduction

In order to meet the demands of increasing waiting lists, transplant programs have expanded the range of acceptable donor organs [1,2]. Generally, organ procurement is only permitted when the donor is already dead and the so-called dead donor rule (DDR) is respected, which states that vital organs can only be taken from dead patients and, correlatively, living patients must not be killed by organ retrieval [3–5]. Brain dead donors currently comprise the major part of the organ pool; upon determining irreversible loss of brain function, the patient is declared legally dead and organ donation can be performed [6].

In contrast, organ donation after cardiac death (DCD) is defined as the surgical recovery of organs after the declaration of death based on cessation of cardiopulmonary function [7–9]. While it was the initial form of organ donation prior to the definition of brain death criteria, DCD was excelled by donation from brain dead patients because of improved graft and recipient outcomes resulting from the shorter time of warm ischemia [10]. However, DCD is prohibited by law in some countries such as Germany [11,12]. DCD is considered to be a promising way to increase the number of organs available for transplantation [1,2,13–16]. However, DCD evokes a number of ethical issues that have to be solved in order to be accepted by the society [17,18]. For organ transplantation to be successful, the arrest of circulation and resulting warm ischemic injury must be minimized [20,21]. This time pressure forced the identification of a precise waiting period, which is long enough to ensure the person has died in order to fulfill the DDR but short enough to maintain organ viability for transplantation [8,20].

Several DCD protocols are currently used in different countries. As there exist no common guidelines for “no-touch” times which even vary within some countries ranging from 2 to 10 min, recently published data of a survey on DCD activities of some European countries [9] as well as the United States of America and Australia are compiled in Table 1. The British Transplant Society claims an interval of 5 min “hands-off” [22] the Maastricht experience 5 min [23] and the Pittsburgh Protocol allows death to be declared 2 min after loss of cardiopulmonary function; [17,24,25] the shortest “no touch” time reported in literature was 75 s [26]. However, there is evidence that the time required for irreversible loss of brain function after cessation of circulation is longer than 5 min of cardiorespiratory arrest [27–30]. Coronary and cerebral reperfusion after cardiac arrest can lead to the return of cardiac and brain function during the procurement process hereby not respecting the DDR [4,31,32]. Therefore, it has been claimed to wait until the patient fulfils the brain death criteria prior to organ donation [10]. Taking in consideration the wide range of “no-touch” times, international guidelines for DCD would be desirable.

Table 1.   Results of a recently published survey on different “no-touch” times for some European countries reporting DCD activity [1,16] as well as Australia [59], Canada [60] and the United States of America [61–63] where recommendations for “no-touch” times vary between 2 and 10 min. However, there are no guidelines currently available for European countries and therefore “no-touch” times vary within countries ranging from 2 to 10 min.
Country“no-touch” period (min)
Austria10
Australia 2
Belgium 5
Canada 5
Czech Republic10
France 5
Italy20
Latvia15
The Netherlands 5
Spain 5
Switzerland10
United Kingdom 5
United States of America 2–10

The aim of this study was to determine brain death by clinical and electrophysiological criteria after cardiac arrest with varying “no-touch” times and subsequent cardiopulmonary resuscitation (CPR) in an animal model in order to explore the scientific basis of the wide range of differing “no-touch” times.

Materials and methods

Donation after Cardiac Death (DCD) Model

Animals

Twenty pigs (German Large White, with a body weight of 35 ± 3.2 kg) were used and the experimental design (Fig. 1) was approved by the Ethical Committee for Animal Studies of the University of Veterinary Medicine Vienna and the Austrian Federal Ministry for Science and Research (GZ: BMWF 66.010/0053-II/10b/2009). All experiments were performed in accordance with European and Austrian laws on animal experimentation and the principles stated in the “Guide for the Care and Use of Laboratory Animals” published by the National Institute of Health [33]. Pigs were judged to be healthy on the basis of physical examination and were acclimatized at the Medical University of Graz (Biomedical Research) for at least two weeks before each trial. The animals were housed in groups of two to four in solid floor pens on straw bedding and were allowed free access to drinking water and a standard pig diet (PorkoCidKorn F, Garant, Graz, Austria). Environmental temperature was held at 20–26 °C at ambient humidity. Lighting was both natural and artificial with a 12-h on and off cycle (06:00–18:00 h).

Figure 1.

 Design of the DCD paradigm in pigs in this study. After premedication and induction of anesthesia, maintenance anesthesia was performed using 2% etSEVO and a remifentanil. 60 min after premedication the animals were randomized into different groups differing in the “no-touch” time ranging from 2 to 10 min. Ventricular fibrillation was induced using a 9-V direct current. An isoelectric EEG was detected within the first minute without cardiac output. Then, dependent on the “no-touch” time CPR was performed for 30 min according to the standard guidelines achieving a MAP of 50 mmHg. Maintenance anesthesia was reintroduced immediately in all animals where a reappearance of brain bioelectric activity was detected. In animals without brain bioelectric activity, anesthesia remained discontinued and brain death diagnostics including brain stem reflex testing and apnea testing was performed after a total time of CPR of 30 min. Then, all animals were sacrificed and multiorgan donation was performed as described elsewhere. DCD, donation after cardiac death; EEG, electroencephalogram; CPR, cardiopulmonary resuscitation; MAP, mean arterial pressure.

Anesthesia

Preanesthetic medication was intramuscular 0.4 mg/kg midazolam (Midazolam “ERWO” 5 mg/ml; ERWO Pharma GmbH, Brunn am Gebirge, Austria), 2 mg/kg azaperone (Stresnil® 40 mg/ml; Janssen Pharmaceutica NV, Belgium) and 14 mg/kg ketamine (Ketasol® 10%; Gräub AG, Berne, Switzerland) injected in one syringe 20 minutes before induction of anesthesia with intravenous (IV) 3 mg/kg propofol (Propofol Fresenius 1%; Fresenius Kabi Austria GmbH, Graz, Austria) via an IV cannula placed in the marginal auricular vein. After endotracheal intubation and connection to a circle system (Sulla 808V anesthetic machine with Ventilog; oxygen flow rate 2 l/min, air flow rate 0.5 l/min), pigs were mechanically ventilated. Standardized ventilator settings for intermittent positive pressure ventilation (IPPV) were used to maintain eucapnia (etCO2 35–45 mmHg; tidal volume 8–10 ml/kg; respiratory rate 15–20 breaths/min; volume controlled ventilation mode; positive end-expiratory pressure (PEEP) 2–4 cm H2O). General anesthesia was maintained with 2% end-tidal sevoflurane (etSEVO) (Sevorane® Abbott Ges.m.b.H, Vienna, Austria) and a continuous rate infusion (CRI) of 0.08–0.1 μg/kg/h remifentanil (Ultiva® 2g; GlaxoSmithKline Pharma GmbH, Vienna, Austria). ELO-MEL isoton solution (Fresenius Kabi Austria GmbH, Graz, Austria) was infused continuously at 10 ml/kg/h IV (Heska Vet-IV Infusion Pump; Heska USA Corporation, Fort Collins, CO, USA). The animals were placed on a heating blanket to maintain normothermia (37.0–39.5 °C). Body temperature was measured continuously using a rectal thermometer. Pulse oximetry was performed at the tail and electrocardiogram (ECG) monitoring was used to observe cardiac function. Arterial blood pressure was measured invasively via a cannula placed in the left femoral artery. A central venous catheter was placed into the left jugular vein to monitor central venous pressure and to have a second venous access for additional drug and infusion therapy. Throughout the experiments arterial blood gas checks were performed in the routine laboratory; according to these results, animals were treated following standard anesthesiologic guidelines [34]. Instrumentation and stabilization phase was finished approximately 60 min after premedication; then, the animals were randomized into experimental groups differing in the “no-touch” time as compiled in Table 2. For randomization, sealed envelopes containing the treatment assignments were drawn out of a bowl prior to the induction of ventricular fibrillation for each animal.

Table 2.   The “no-touch” time (indicated in minutes) and time to isoelectric electroencephalogram (EEG) from the beginning of ventricular fibrillation in each of the animals (indicated in seconds). When no EEG activity reappeared throughout 30 min of continuous recording, brainstem reflexes were tested and painful stimuli applied (n.a. – not applicable). Brain death was confirmed by apnea testing when no EEG activity reappeared and no brainstem reflexes and reaction to painful stimuli were found. Animals in which spontaneous circulation (SC) after CPR reoccurred can be distinguished among the animals which underwent CPR throughout the whole experiment.
Animal“No-touch” time (min)Spontaneous circulation/CPRTime to isoelectric EEG (s)Reappearance of EEG activityBrainstem reflexesBrain death
  1. DCD, donation after circulatory death; CPR, cardiopulmonary resuscitation.

DCD IV 2SC40Yesn.a.No
DCD VIII 4SC22NoYesNo
DCD XVI 4SC35NoYesNo
DCD XVII 4SC35Yesn.a.No
DCD XIII 4.5SC28Yesn.a.No
DCD XIV 4.5CPR32NoYesNo
DCD XV 4.5CPR22NoNoYes
DCD IX 5CPR22NoNoYes
DCD X 5CPR34NoYesNo
DCD XI 5SC32NoYesNo
DCD XII 5SC24NoYesNo
DCD V 6CPR74NoNoYes
DCD VI 6SC35NoNoYes
DCD VII 6SC74NoNoYes
DCD I 9SC24NoNoYes
DCD XVIII 9CPR27NoNoYes
DCD II10CPR46NoNoYes
DCD III10SC27NoNoYes
DCD XIX10CPR54NoNoYes
DCD XX10CPR32NoNoYes

Surgical procedure

A subcostal thoracotomy was performed and the pericardium was opened. Maintenance anesthesia was discontinued one minute before ventricular fibrillation was induced by a 9-V direct current. The etSEVO levels were monitored until a decrease to 0.1%; then, mechanical ventilation was stopped. Pigs underwent a “no-touch” period ranging from 2 to 10 min (Table 2). After the defined “no-touch” period, mechanical and medical resuscitation (CPR) was performed for 30 min according to standard guidelines [35,36]. The aim was to achieve sufficient cardiac output (mean arterial pressure (MAP) 50 mmHg) to enable brain perfusion. When cardiac activity reoccurred during CPR, animals were treated by the anesthesiologist according to standard guidelines for a total time of 30 min [34]. Arterial blood pressure, oxygenation, ECG, body temperature, capnometry, blood glucose and central venous pressure were monitored continuously. Pigs were kept normothermic (37.0–39.5 °C) throughout the experiment and intensive care medication was provided when indicated in order to avoid acidosis, keep electrolytes within normal limits and avoid any metabolic disturbances which would impact on neurologic examinations.

EEG monitoring and brain death diagnostics

Electroencephalograms (EEGs) were obtained by a clinical EEG system (alpha-trace, Vienna, Austria) with needle electrodes positioned as described previously [37]. Briefly, a pair of electrodes (FP1, FP2) was inserted 2.0 cm in front of a reference line connecting both medial eye borders and 1.0 cm left and right of the midline. Another electrode pair (F7, F7) was positioned 1.0 cm behind the reference line and 3.0 cm left and right of the midline. A third (T7, T8) and fourth (P7, P8) pair of electrodes were inserted 4.0 cm to the left and right of the midline and 2.5 cm and 4.0 cm behind the reference line, respectively. A common average montage was used and ECGs were co-registered. Baseline EEGs were obtained under general anesthesia. EEG recordings were continued throughout interruption of anesthesia, induction of ventricular fibrillation, the “no-touch” time, and CPR. With maximum signal amplification, the disappearance of EEG activity (isoelectric EEG) and the eventual reappearance of brain bioelectric activity were noted. With the appearance of an isoelectric EEG, recordings were continued for at least 30 min. When no bioelectric activity reappeared during this time, brainstem reflexes were tested and painful stimuli were applied. In animals with a loss of brainstem reflexes and a lack of reaction to painful stimuli, apnea testing was performed. Animals were disconnected from the ventilator until a pCO2 >60 mmHg was recorded using arterial blood gas analysis as described above. A lack of spontaneous respiration was regarded confirmatory for brain death [38].

Biochemistry

Blood samples were taken after induction of anesthesia, prior to the induction of ventricular fibrillation, after the “no-touch” time as well as after 30 min of CPR. Blood gas analysis was performed every 5 min during CPR. Full blood count, electrolytes, renal and liver function tests were immediately analyzed in the central laboratory. Serum was stored at −80 °C for batch analysis of midazolam with a reversed phase HPLC method [39,40]. The within-day coefficients of variation (CVs) for midazolam were 2.0% and 1.1%, the between-day CVs were 7% and 5.7% at 40 and 200 ng/ml, respectively. According to current national brain death diagnosis guidelines fully reflecting American Academy of Neurology (AAN) practice parameters, midazolam levels have to be below 50 ng/ml in order to fulfill brain dead criteria [38,41,42].

Results

CPR and vital parameters

All animals were declared healthy and did not differ significantly in terms of blood pressure, heart rate, temperature as well as routine laboratory values prior to the experiments (Table 3). Vital parameters were within normal limits; mean arterial blood pressure (MAP) was 73 ± 15 mmHg, mean heart rate (HR) 79 ± 20 beats per minute (bpm) and mean temperature was 39.1 ± 0.3 °C prior to the induction of cardiac fibrillation.

Table 3.   Baseline data of the animals used in these experiments.
AnimalMAP (mmHg)HR (beats/min)Temperature (°C)Leuco (G/L)Ery (G/L)Hb (g/dl)Na+ (mmol/l)K+ (mmol/l)Crea (mg/dl)BUN (mg/dl)tot. Bili (mg/dl)AP (U/l)gGT (Ul)AST (U/l)ALT (U/l)LDH (U/l)Midazolam (ng/ml)
  1. Normal values of the pigs used for these experiments are indicated in braces according to the routine laboratory of the University for Veterinary Medicine, Vienna.

  2. DCD, donation after cardiac death; MAP, mean arterial blood pressure (mmHg) (60–70 mmHg in anesthesia); HR, heart rate (beats per minute/bpm) (60–90 bpm in anesthesia); temperature: (27,0–39,5 °C); Leuco, leucocytes (G/L) (11–22 G/L); Ery: erythrocytes (G/L) (5–8 G/L); Hb, hemoglobin (g/dl) [10–16 g/dl); Na+: sodium (mmol/l) (135–150 mmol/L); K+: potassium (mmol/L) (7.8–10.9 mmol/L); Crea, creatinine (mg/dl) (1–3 mg/dl); BUN: blood urea nitrogen (mg/dl) (8–24 mg/dl); tot. Bili., total Bilirubin (mg/dl) (0–0.7 mg/dl); AP: alkaline phosphatase (U/l) (9–20 U/l); gGT, gamma-glutamyl transferase (U/l) [10–27 U/l); AST, aspartate aminotransferase (U/l) (0–35 U/l); ALT, alanine-aminotransferase (U/l) (0–40 U/l); LDH, lactate dehydrogenase (U/l) [0–700 U/l); midazolam (ng/ml) [<50 ng/ml to allow brain death diagnostics].

DCD I655238.713.495.589.31383.91.11110.1110867223856139
DCD II634839.227.44.998.21413.71.57200.0710476134348426
DCD III687539.822.775.1991393.71.08110.191142331499< 20
DCD IV10912039.215.835.69.91404.31.18100.099254294361122
DCD V838739.511.784.417.61453.61.03200.0910845175143036
DCD VI677038.913.424.678.114640.79110.115444374949742
DCD VII75663912.55.038.11414.30.77160.0911945317245540
DCD VIII10510439.521.215.879.61443.81.84130.137665183038926
DCD IX757839.312.946.0210.81433.81.16120.2712632325152842
DCD X797039.224.045.789.81464.62.18310.0895391513252343
DCD XI616839.316.135.1691444.41.57130.1311830283746246
DCD XII6512038.818.396.1810.21423.71.42210.114356214249120
DCD XIII797439.212.565.9110.214441.4200.0411369335565044
DCD XIV527338.719.246.4610.91463.71.52170.0521869365759439
DCD XV607039.215.266.2710.41423.41.42200.15157452137479<20
DCD XVI8010038.79.065.469.214141.04300.3616757384158226
DCD XVII859739.315.935.078.614041.41230.0982322945502<20
DCD XVIII799338.915.786.6511.31414.21.27170.181181193338584<20
DCD XIX536038.913.745.9210.41424.21.51250.0999604343501<20
DCD XX756839.217.636.29.81443.51.2170.0989652239470<20

CPR was performed successfully for the whole experimental period according to the standard guidelines in all 20 animals. Eleven animals regained spontaneous circulation (SC) after CPR, whereas nine animals had to be resuscitated throughout the experimental period.

After the different “no-touch” times and 5 min of CPR, MAP was 60 ± 43 mmHg, HR 117 ± 45 bpm and the mean temperature was 39.0 ± 0.3 °C. Then, 10 min after the beginning of CPR, pigs showed a MAP of 61 ± 41 mmHg, a HR of 113 ± 41 bpm as well as a body temperature of 38.9 ± 0.23 °C; MAP was 54 ± 36 mmHg, HR 106 ± 32 bpm and mean body temperature 39 ± 0.23 °C after 20 min of CPR respectively. Prior to sacrification, 30 min after the different “no-touch” times, animals showed a MAP of 46 ± 16 mmHg, a HR of 104 ± 37 bpm as well as a mean body temperature of 38.8 ± 0.36 °C. Detailed values for vital parameters of all animals during CPR are compiled in Table 4.

Table 4.   Vital parameters of all pigs are documented throughout the experiments. Mean arterial pressure (MAP), heart rate (HR) as well as temperature were documented 5 min, 10 min, 20 min as well as 30 min after the beginning of CPR after the different “no-touch” times. All animals lacked massive hypotensive periods interfering with brain death diagnosis throughout the experiments; Body temperature was kept normothermic (37.0–39.5 °C) during the whole experiment.
AnimalMAP 5 min CPRMAP 10 min CPRMAP 20 min CPRMAP 30 min CPRHR 5 min CPRHR 10 min CPRHR 20 min CPRHR 30 min CPRTemp 5 min CPRTemp 10 min CPRTemp 20 min CPRTemp 30 min CPR
  1. DCD, donation after circulatory death; CPR, cardiopulmonary resuscitation.

DCD I3740444811212712813738.338.238.238
DCD II303534408679606239.138.939.238.4
DCD III3244383718017816417039.539.239.138.8
DCD IV8665546712014012013039.239.139.239.2
DCD V5038454320090887039.339.239.339
DCD VI566365601501601009838.938.83938.9
DCD VII504547468278798338.738.838.938.7
DCD VIII5055404845122919539.339.239.339.4
DCD IX383532318381786839.239.239.138.9
DCD X32304439907810713039.239.139.239
DCD XI1221531387017013615714339.239.239.339.5
DCD XII15772444613914015015538.638.638.538.6
DCD XII4061564714012012010239.13939.239.1
DCD XIV373737368882836538.638.738.538.5
DCD XV263935327275807739.139.239.139.1
DCD XVI17016363462001801401603938.93938.7
DCD XVII911401709613018414013339.239.339.239.1
DCD XVIII293634378793908838.738.638.638.5
DCD XIX343437298250815238.738.738.838.7
DCD XX383832398171656739.139.239.239

Serum levels of midazolam

Mean midazolam levels prior to induction of ventricular fibrillation are 33 ± 10.1 ng/ml ranging from values below 20 ng/ml up to 46 ng/ml. In all animals, midazolam levels were below the threshold of 50 ng/ml. Therefore midazolam levels did not interfere with brain death diagnosis according to our national guidelines fully reflecting AAN practice parameters (Table 2) [38,41,42].

EEG monitoring and brain death diagnostics

Brain death diagnostic criteria are given in Table 3 for all animals. An isoelectric EEG appeared after a mean of 36.0 s (range 22–74 s) following induction of ventricular fibrillation. One animal (DCD IV) underwent a 2-min “no-touch” time. During 30 min of CPR, this animal showed reappearance of EEG activity (Fig 2a–c). Among the animals with 4 min “no-touch” time (= 3), two did not show reappearance of EEG activity but brainstem reflexes after 30 min of CPR. With 4.5 min of “no-touch” time (= 3), one animal did not show reappearance of EEG activity and brainstem reflexes were absent. In the group which underwent 5 min of “no-touch” time (= 4), one animal showed neither reappearance of EEG activity nor brainstem reflexes. With longer “no-touch” times, 6 min (= 3), 9 min (= 2) and 10 min (= 4), respectively, all animals neither showed reappearance of bioelectric brain activity nor brainstem reflexes (Fig. 2d–f). In all animals with 30 min of isoelectric EEG, loss of brainstem reflexes and lack of reaction to painful stimuli, apnea testing confirmed brain death.

Figure 2.

 Electroencephalogram (EEG) recording in animal DCD IV 1 min before the induction of ventricular fibrillation (a) and at the beginning of isoelectric EEG at 40 s of cardiac arrest (b). Following 2 min of “no-touch” time and subsequent cardiopulmonary resuscitation (CRP), EEG activity reappeared as shown here at 3 min of CPR (c). EEG recording in animal DCD V 1 min before the induction of ventricular fibrillation (d) and at the beginning of isoelectric EEG at 74 s of cardiac arrest (e). Following 6 min of “no-touch” time and subsequent CPR, no reappearance of EEG activity was found throughout 30 min of continuous recording (f). See Materials and Methods for electrode positions (FP1–P8). Electrocardiogram (ECG) is co-registered. Minimum time interval on x-axis is 0.2 s. DCD, donation after cardiac death.

Discussion

DCD is increasingly recognized for organ donation, [16,43] but still discussed controversially. Moreover, DCD is prohibited by law in Germany, the largest country of EUROTRANSPLANT and some other European countries [11,12,16]. Moreover, international guidelines are missing what is reflected by the wide range of “no-touch” times currently used in the different countries practicing DCD [40]. This might be because of the lack of exact data on the required “no-touch” time as well as ethical and legal issues which have to be taken in consideration [17,19].

In the previous literature, several animal models are described focusing on different DCD protocols. However, none of these studies has included brain death diagnosis [44–46].

Therefore, the aim of this study was to determine brain death by clinical and electrophysiological criteria after cardiac arrest with varying “no-touch” times and subsequent cardiopulmonary resuscitation (CPR) in a large animal study to examine the scientific basis of different “no-touch” times.

Several DCD protocols [9,16,19,43,47] are currently used in different countries including the protocol of The British Transplant Society (5 min “hands-off”) [22] the Maastricht experience (5 min “no-touch” time) [23] and the Pittsburgh Protocol (2 min “no-touch” time). As there exist no common guidelines and the “no-touch” times even vary within countries, recently published data on DCD programs are compiled in Table 1 [17,24,25]. The shortest “no touch” time reported in literature was 75 s [26]. However, there is still a lack of knowledge about reappearance of bioelectric brain activity when declaring a patient dead according to DCD criteria. There is evidence that the time required for irreversible loss of brain function after cessation of circulation is longer than 5 min of cardiorespiratory arrest [27–30].

Among the four animals with cardiac arrest for 2 and 4 min, respectively, two pigs showed no reappearance of bioelectric brain activity but brain stem reflexes. Therefore, these animals could not be declared brain dead. In animals with a “no-touch” time of 4.5 min, only one pig could be declared brain dead because of an isoelectric EEG as well as the loss of brainstem reflexes and confirmatory apnea testing. Moreover, 5 min of ventricular fibrillation without CPR led to an isoelectric EEG in all animals. However, only one of these animals showed a loss of brainstem reflexes and a lack of reaction to painful stimuli and was declared brain dead upon apnea testing. Animals which suffered from “no-touch” times of 6 min or more followed by CPR could be declared brain dead after 30 min of CPR.

According to our study, DCD organ donors according to the Pittsburgh Protocol [17,25] supposedly do not fulfill clinical and electrophysiological criteria of brain death at the time of donation. Regarding the recommendations of the British Transplant Society (5 minutes “hands-off”) [22], we found that in pigs which suffered from ventricular fibrillation without cardiac output for 5 minutes, all pigs showed an isoelectric EEG for at least 30 min. However, only one of these was declared brain dead upon the loss of brainstem reflexes the lack of reaction to painful stimuli, and confirmatory apnea testing. All animals which were treated 10 min “no-touch” time, as suggested by Kootstra and Jacobs as equivalent to brain death [48] fulfilled clinical and electrophysiological brain death criteria.

Since the warm ischemia time should also be as short as possible to avoid hypoxic damage of the organs that should be transplanted, the ideal “no-touch” time in our model seems to be somewhere between 5 and 10 min. As humans suffering from cardiac death are usually not completely healthy at the outset and cardiac output also does not cease immediately but they rather suffer from a prolonged period of low cardiac output before cardiac arrest, it can be hypothesized that in humans the duration until brain death occurs, might be even shorter. However, since there are no data available, this cannot be taken for granted. Therefore using the 10 min “no-touch” time seems to be safe from an ethical perspective to ensure brain death in all DCD donors before organ retrieval; however, from an organ quality perspective every minute of warm ischemia that can be safely cut down will improve the success of transplantation. This is an important result which can be used in ethical and legal discussions concerning “no-touch” time in DCD.

Factors such as hypothermia, lack of brain perfusion because of low cardiac output during CPR, as well as metabolic influences and impact of anesthesia on bioelectric brain activity monitoring were excluded throughout the experiments. While the precise mechanism that inhalant anesthetics exert their general anesthetic effects is not precisely known, they may interfere with functioning of nerve cells in the brain by acting at the lipid matrix of the membrane. Sevoflurane has a very low blood gas partition coefficient (0.6) allowing very rapid anesthesia induction and recovery. This low solubility in blood means that sevoflurane is rapidly removed from the lungs. It is unknown to which proportion sevoflurane is bound to plasma proteins. The majority of sevoflurane is excreted via the lungs, but about 3% is metabolized in the liver [49]. As we observed the etSEVO going down from 2% to 0.1% on the capnograph display after switching off the sevoflurane vaporizer, the vast majority of sevoflurane was considered to be removed from the lungs and, therefore, from the circulation as well.

Remifentanyl is known not to accumulate in the human circulation and the time required to achieve a 50% decrease in plasma concentration after termination of the infusion is independent of cardiovascular circulation [50]; the low doses used as maintenance anesthesia during these experiments prior to induction of ventricular fibrillation therefore may not influence recording of bioelectric brain activity after the “no-touch” time and during CPR.

Massive hypotension (MAP <30 mmHg) is known to negatively impact on EEG activity because of the disruption of the autoregulation of blood flow to the brain [51]. However, during our experiments, MAP during CPR as well as MAP of animals which showed a spontaneous circulation after the “no-touch” time and short CPR was kept over a mean value of 30 mmHg throughout the all procedures [52]. This is in accordance with a recent study published by Liao et al. [53].

Of course the major limitation of this study is the fact that pigs and not humans were used. Any animal model reproduces at best a very limited component of the pathophysiologic spectrum of the human disease state studied. We used healthy pigs as a surrogate for end of life humans. However, this might not be representative for most DCD subjects. In our model, the animals were under controlled ventilation and cardiac output was suddenly stopped by the initiation of ventricular fibrillation. In contrast, human DCD have variable respiratory drive and cardiac output before approaching cardiac death. Presumably most humans approaching cardiac death have a lower physiological reserve as cardiac output slows some time before cardiac arrest. Pig models are regularly used to simulate brain death [54] or cardiovascular disease [55]. The pig model is the preferred large animal model of heart damage because it reflects the pathophysiology of human best. To our knowledge, there are no reports of animal models reflecting an “end-of-life” state with a low physiological reserve in which complete clinical and electrophysiological criteria for brain death diagnosis were applied.

Keeping in mind the limitations, we have chosen our model to match the human situation best. Since our pigs were healthy at the beginning of the experiment, this would most likely bias the result towards a better functional brain reserve.

Previous studies in “end-of-life” states in animal models have applied only limited EEG recordings by devices used for measuring depth of anesthesia. In our study, we applied an EEG montage covering the whole porcine brain, and EEGs were recorded by equipment used in clinical routine for brain death diagnosis. Criteria for brain death in our animal model are equivalent to those required for brain death diagnosis in human patients in Austria, fully reflecting AAN practice parameters [38,41,42]. The strength of our animal study is to apply continuous whole brain EEG recordings and subsequent clinical brain death diagnosis in a DCD setting to obtain information about the chronology of brain destruction after complete cessation of circulation.

In summary, 10 min of “no-touch” time guarantee that clinical and electrophysiological criteria of brain death are fulfilled in a pig model, as it has also been suggested by [48]. However, after 5 min of “no-touch” time also no evidence of electrophysiological brain activity could be found any more, what would ethically allow to shorten the “no-touch” time in order to minimize warm ischemia time and therefore probably improve transplantation outcome.

The results of our study suggest that it would be necessary to evaluate the time course of brain damage in human DCD donors in order to establish evidence based guidelines for the management of DCD. To avoid ethical concerns against DCD, the declaration of death has to be based on scientific facts and not on a personal opinion [7–9,21,47,56,57]. Otherwise legitimated opposition will rise in society, especially in religious communities; therefore limiting the acceptance of DCD resulting in a loss of organs being available for transplantation [47,58]. Of course the major limitation of this study is the fact that pigs and not humans were studied; however from our point of view it helps to support the “no-touch” period suggested by Kootsra and Jacobs [48] regarding 10 min “no-touch” time being equivalent to brain death in DCD organ donation and even to allow shorter “hands-off” times as practiced in some other countries.

Authorship

VS, PS and MS: wrote the article, planned the experiments and performed the animal experiments with AP. TS-H and GZ: performed EEG readings. IW and WM: performed anesthesia on the animals. MZ: involved in the planning of the experiments and fund raising. TS and AM: performed routine laboratory analysis. TS-H, VS, MZ and KT: reviewed the article.

Funding

The authors have declared no funding.

Acknowledgements

This work was sponsored by the “Christine-Vranitzky Stiftung” Austria.

Ancillary