Sudden death continues to be a great challenge for all physicians
around the world despite the remarkable advances in modern medicine.
In the United States, more than 350,000 people suffered from sudden
death outside the hospital every year(Gillum 1989). It is estimated that
sudden death occurs almost once in every minute in America, and
becomes the most common cause of death in adults aged 65 years old
and younger(Kuller, Cooper, and Perper 1972). For decades, efforts
have been launched toward primary prevention by searching and
modifying the risk factors associated with sudden death. In recent
years, this work has been extended further for optimizing the
resuscitation of the patients. By strengthening the concept of “chain of
survival”, researchers of both basic and clinical backgrounds have
made great efforts trying to understand the pathophysiology of cardiac
arrest and develop therapies aiming at mitigating the global injuries
resulting from circulatory arrest and cardiopulmonary resuscitation
(CPR).
The “chain of survival” consists of early access, early CPR, early
with the integrity of the emergency medical services system in the
community. For years, efforts have been launched trying to strengthen
the linkage of the “chain of survival”, and have proved important in
improving the resuscitation rate and overall prognosis of the patients.
Unfortunately, the final outcomes of the victims with out-of-hospital
cardiac arrest are still poor. About 50-70% of those who are initially
resuscitated eventually died before they are discharged from the
hospital(Kern 2002). One third of these died of post-resuscitation
cardiac dysfunction, one third died of neurological failure, and the
others died of the systemic inflammatory response syndrome, sepsis,
and multiple organ failure(Schoenenberger, von Planta, and von Planta
1994). Even in the patients who survive to hospital discharge, most
exhibit moderate to severe neurological deficits and need prolonged
nursing care, resulting in tremendous increases of the socioeconomic
burden. In general, only about 1-2% of the patients with out-of-hospital
sudden death are able to return to the society with adequate
neurological and functional recovery. Therefore, delicate goal-directed
post-resuscitation therapy and care are urgently mandated for
mitigating the global ischemia/reperfusion injuries and improving
outcomes in the post-resuscitation phase. Such extension of the “chain
of survival” has recently attracted significant attention, and becomes an
issue of great interest in both researches and clinical practice(Peberdy
and Ornato 2005).
Among all organ systems, heart and brain are of particular
importance since failure of these two organs is responsible for
two-thirds of the resuscitation mortality. For the former,
post-resuscitation myocardial dysfunction has been described in both animal
and human cardiac arrest(Tang et al. 1993; Gazmuri et al. 1996; Kern
et al. 1996; Laurent et al. 2002; Chang et al. 2007).Though it is
basically myocardial stunning due to transient ischemia and is usually
reversible, if the stunning is severe, the hemodynamics can be
significantly compromised in the critical first hours post-resuscitation,
resulting in severe circulatory failure that leads to global
hypoperfusion, multiple organ failure and mortality. As to brain, it is
least tolerable to ischemia. This can largely account for the poor
neurological outcomes usually seen in these patients. In the past
decades, efforts have been devoted trying to investigate the
pathophysiological mechanisms underlying the neurological injury,
based on which potential therapeutic interventions can be developed. A
number of pathogenic mechanisms have been proposed, such as
transient hyperemia in the immediate post-resuscitation period (15-30
min), microcirculatory disturbance thereafter leading to multiple
“no-reflow’ phenomenon (90 min to 12 hours), massive outpouring of
excitotoxic glutamate resulting in activation of NMDA and AMPA
receptors, and opening of calcium channels leading to calcium
overload(Maramattom and Wijdicks 2005). Therapeutic interventions
have also been developed aiming at these pathophysiological targets.
Examples include hypertensive bout of short duration in the
post-resuscitation period for resolving the cerebral “no-reflow
phenomenon”(Leonov et al. 1992; Sterz et al. 1992), nitric oxide
synthase inhibition for prevention of the damaging effects of
hyperemia(Schleien, Kuluz, and Gelman 1998), calcium entry blocker
for decreasing the calcium overload in the ischemic/hypoxic
neurons(Vaagenes et al. 1984), and free radical scavengers for
attenuating the oxidative stress during ischemia/reperfusion
injury(Traystman, Kirsch, and Koehler 1991; Cerchiari et al. 1987).
In current research, we aimed to develop a goal-directed therapy
to resolve the so-called post-cardiac arrest syndrome. Since the major
cause leading to post-resuscitation death was the neurological injury,
we set the neurological recovery as the primary outcome in the
analysis. We used three approaches to achieve our research goal. First,
we used systematic review and meta-analysis to examine the effect of
hyperoxia on post-resuscitation outcomes. Second, we analyzed the
cohort of in-hospital cardiac arrest (IHCA) in an attempt to identify the
optimal range of mean arterial pressure, partial pressure of arterial
oxygen and carbon dioxide, hemoglobin level and blood glucose level
for post-resuscitation patients. Finally, by directly observing the
influence of brain flow changes on the post-resuscitation rat models,
we tried to investigate the underlying mechanisms involved in the
pathogenesis of the neurological injuries and explored the possible
therapeutics.
First, in the systematic review and meta-analysis, we followed the
guidelines of the Preferred Reporting Items for Systematic reviews and
Meta-Analyses(Moher, Liberati, Tetzlaff, and Altman 2009) and the
Meta-analysis of Observational Studies in Epidemiology(Stroup,
Berlin, Morton, and et al. 2000). We searched PubMed and Embase
from the inception through October 2013. We did not set any
restrictions on publication year or language. We used 2 sets of search
terms to represent the primary variable and population of interest. The
search terms for the primary variable included “normox*,”
“hyperox*,” and “oxygen*.” Then, the search results were
cross-checked for the population of interest, using the terms “cardiac arrest”
and “cardiopulmonary resuscitation.” To ensure completeness, we also
reviewed the references of relevant articles. Studies that were eligible
for inclusion (1) compared different levels of partial pressure of arterial
oxygen (PaO2) in adult patients following return of spontaneous
circulation (ROSC); (2) included mortality or neurological status as
outcome; and (3) used an observational study design, either a cohort or
case-control study with an appropriate control group. Studies only
comparing hypoxia with normoxia were excluded. We defined
hyperoxia as a PaO2 higher than 300 mm Hg(Douzinas et al. 2001);
hypoxia, as a PaO2 lower than 60 mm Hg(Abraham et al. 2000); and
normoxia, as a PaO2 between 60 and 300 mm Hg. Odds ratio (OR) was
used as an effect estimate for the data synthesis. Data were combined
and expressed as a Mantel-Haenszel weighted mean of the ORs, with
their associated 95% CIs. Heterogeneity was quantified by the I2
statistics and tested with Cochran Q statistics (p < 0.05)(Higgins et al.
2003a; Ioannidis, Patsopoulos, and Evangelou 2007a). For values of I2
< 50% or p > 0.05, fixed-effects models were chosen; otherwise,
random-effects models were used(DerSimonian and Laird 1986). In
the literature search, 14 studies were identified from 2,982 references.
Meta-analysis indicated that hyperoxia appeared to be correlated with
increased in-hospital mortality (OR, 1.40; 95% Confidence
interval[CI], 1.02–1.93; I2, 69.27%; 8 studies) but not worsened
neurological outcome (OR, 1.62; 95% CI, 0.87–3.02; I2, 55.61%; 2
studies). However, the results were inconsistent in subgroup and
sensitivity analyses.
Second, we performed the retrospective cohort study at National
Taiwan University Hospital (NTUH), which is a tertiary medical center
with 2600 beds, including 220 beds in intensive care units. We
screened patients who suffered IHCA at NTUH between 2006 and
2014. We included patients who met the following criteria: (1) age 18
years or older, (2) documented absence of pulse with performance of
chest compression for at least 2 min, (3) no documentation of a
do-not-resuscitate order, and (4) achievement of sustained return of
spontaneous circulation (ROSC) (i.e., ROSC ≥ 20 min without
resumption of chest compression). If multiple cardiac arrest events
occurred in a single patient, only the first event of the same
hospitalization was recorded. We excluded patients without any
measurement of variables of interest within the first 24 h after
sustained ROSC, such as mean arterial pressure, partial pressure of
arterial oxygen or carbon dioxide, hemoglobin level or blood glucose
level. We also excluded patients who suffered a cardiac arrest related to
major trauma. We recorded the following information for each patient:
age, gender, comorbidities, variables derived from the Utstein template
(Jacobs et al. 2004), critical interventions implemented at the time of
cardiac arrest or after sustained ROSC, and the first, maximum and
minimum values of mean arterial pressure (MAP), partial pressure of
arterial oxygen (PaO2)or carbon dioxide(PaCO2), hemoglobin (Hb)
level and blood glucose (BG) level measured during the first 24 h after
sustained ROSC. The primary outcome was favourable neurological
outcome at hospital discharge, and the secondary outcome was survival
to hospital discharge. Favourable neurological outcome was defined as
a score of 1 or 2 on the Cerebral Performance Category (CPC) scale
(Becker et al. 2011). The CPC scale (Becker et al. 2011) is a validated
5-point scale of neurological disability (1, good cerebral performance;
2, moderate cerebral disability; 3, severe cerebral disability; 4,
coma/vegetative state; 5, death). Patients with a CPC score of 1 or 2
had sufficient cerebral function to live independently. We selected the
OR as the outcome measure and we performed multivariable logistic
regression analyses to examine the associations between independent
variables and outcomes. We considered all available independent
variables in the regression model, regardless of whether they were
significant by univariate analysis. We applied the stepwise variable
selection procedure (with iterations between the forward and backward
steps) to obtain the final regression model. Significance levels for entry
and to stay were set at 0.15 to avoid exclusion of potential candidate
variables. We calculated the final regression model by excluding
individual variables with a p-value greater than 0.05 until all regression
coefficients were statistically significant. We used generalized additive
models (Hastie TJ and Tibshirani RJ 1990) to examine the nonlinear
effects of continuous variables and, if necessary, to identify the
appropriate cut-off point(s) for dichotomizing a continuous variable
during the variable selection procedure.
The results were as follows: (1) MAP above 85 mm Hg was found
to correlate with a favorable neurological outcome (OR 4.12, 95% CI
1.47-14.39, p = 0.01). For patients without arterial hypertension, the optimal MAP was between 85 and 115 mm Hg (OR 8.80, 95% CI
3.13–28.55, p < 0.001); for patients with arterial hypertension, the threshold MAP for achieving a favorable neurological outcome was
above 88 mmHg (OR 4.04, 95% CI 1.41–13.03, p = 0.01). (2) PaO2
between 70 and 240 mmHg (OR 1.96, 95% CI 1.08–3.64) and PaCO2
levels (OR 0.98, 95% CI 0.95–0.99) were positively and inversely
associated with favorable neurological outcome, respectively. (3) The
product of Hb level × peripheral hemoglobin oxygen saturation (SpO2)
was correlated with a favorable neurological outcome (OR 1.003, 95%
CI 1.002-1.004). According to recommended SpO2 by resuscitation
guidelines [94% to 98%], we calculated the corresponding range of
minimum required Hb level to be 8.6 to 9.0 g/dL for a favorable
neurological outcome. (4) For diabetic patients, a mean BG level
between 183 and 307 mg/dL (10.2-17.1 mmol/L) was significantly
associated with favourable neurological outcome (OR: 2.71, 95% CI:
1.18-6.20; p-value = 0.02); a mean BG level between 147 and 317
mg/dL (8.2-17.6 mmol/L) was significantly associated with survival to
hospital discharge (OR: 2.38, 95% CI: 1.26-4.53; p-value = 0.008). For
non-diabetic patients, a mean BG level between 143 and 268 mg/dL
(7.9-14.9 mmol/L) was significantly associated with survival to
hospital discharge (OR: 2.93, 95% CI: 1.62-5.40; p-value < 0.001).
Finally, we used a rat model of asphyxia-induced cardiac arrest to
examine the effect of cerebral blood flow changes on post-resuscitation
outcomes. Male Wistar rats (450 to 550 g) were used for the
experiments. Animals were anesthetized with intra-peritoneal
pentobarbital (45 mg/ Kg). The trachea was intubated with a PE 200
catheter (Angiocath, Becton Dickinson). Mechanical ventilation
(Flexivent EC-VF-2, Scireq Scientific Respiratory Equipment Inc) was
initiated with a tidal volume of 2 ml, a respiratory rate of 100 breaths
per min and FiO2 of 21%. Arterial blood pressure was measured with
saline-filled PE-50 tube inserted through the left femoral artery. A
saline-filled PE-50 tube was inserted into the right jugular vein for
fluid and drug administration. Blood pressure and needle-probe ECG
monitoring data were recorded with the use of a PC-based data
acquisition system (Biobench, National Instruments, Inc). After
surgical preparation, animals were observed on the ventilator for 50
min. Cardiac arrest was induced by asphyxia (turning-off of the
ventilator with obstruction of the endotracheal tube in situ with clay).
Bradycardia and hypotension usually developed soon after asphyxia,
which soon progressed to asystole with complete loss of arterial
pressure about 3 to 4 min after asphyxia. After a total of 8 min of
asphyxia, ventilator was turned on, epinephrine (0.005 mg/100g) was
administrated via the central venous line, and chest compressions
instituted promptly by index and middle fingers at a rate 200-300 beats
per min. Chest compressions were adjusted to provide a uniform rate
seen on monitors and a target aortic diastolic pressure of > 20 mm Hg.
ROSC could usually be achieved within 1 min after start of CPR. If
ROSC could not be reached within 6 min, the animal was excluded
from the study. After 4 h of invasive monitoring, the endotracheal tube
and catheters were removed, and surgical wounds closed. The animals
achieving ROSC were monitored up to 24 hours for survival and
neurological assessment. In the experimental group, norepinephrine
was administered for one hour after ROSC to titrate the mean arterial
pressure in order to maintain the OxyFlo-measured local cerebral
perfusion close to the prearrest level. The result demonstrated that
there were no significant differences in 24-hour survival and
neurological outcomes between the control and the experimental
groups despite that the mean arterial pressure and cerebral perfusion
were significantly higher in the experimental group during the
norepinephrine infusion period.