2021 年 85 巻 10 号 p. 1842-1848
Background: The effect of in-hospital rapid cooling by intravenous ice-cold fluids for comatose survivors of out-of-hospital cardiac arrest (OHCA) is unclear.
Methods and Results: From the J-PULSE-HYPO study registry, data for 248 comatose survivors with return of spontaneous circulation (ROSC) who were treated with therapeutic hypothermia (34℃ for 12–72 h) after witnessed shockable OHCA were extracted. Patients were divided into 2 groups by the median collapse-to-ROSC interval (18 min), and then into 2 groups by cooling method (rapid cooling by intravenous ice-cold fluids vs. standard cooling). The primary endpoint was favorable neurological outcome (Cerebral Performance Category of 1 or 2) at 30 days after OHCA. In the whole cohort, the shorter collapse-to-ROSC interval group had significantly higher favorable neurological outcome than the longer collapse-to-ROSC interval group (78.2% vs. 46.8%, P<0.001). In the shorter collapse-to-ROSC interval group, no significant difference was observed in favorable neurological outcome between the 2 cooling groups (rapid cooling group: 79.4% vs. standard cooling group: 77.0%, P=0.75). In the longer collapse-to-ROSC interval group, however, favorable neurological outcome was significant higher in the rapid cooling group than in the standard cooling group (60.7% vs. 33.3%, P<0.01) and the adjusted odds ratio after rapid cooling was 3.069 (95% confidence interval 1.423–6.616, P=0.004).
Conclusions: In-hospital rapid cooling by intravenous ice-cold fluids improved neurologically intact survival in comatose survivors whose collapse-to-ROSC interval was delayed over 18 min after shockable OHCA.
Evidence-based international guidelines for cardiopulmonary resuscitation (CPR) and emergency cardiovascular care were first published in 2000, and have been revised every 5 years with the accumulation of additional evidence.1–4 In 2002, 2 randomized controlled trials reported that therapeutic hypothermia (TH) after return of spontaneous circulation (ROSC) improved neurological outcome.5,6 Evidence of the usefulness of TH has also accumulated since those reports, and the international consensus on CPR and emergency cardiovascular care science with treatment recommendations (CoSTR) 20103 increased the evidence level of TH to Class I in adult comatose survivors with ROSC who had out-of-hospital cardiac arrest (OHCA) due to shockable rhythm (ventricular fibrillation or pulseless ventricular tachycardia). The guidelines were revised in 2015 based on a paper by Nielsen et al7 and recommended that all adult comatose survivors with ROSC after OHCA be kept at a core temperature of 32–36℃.4,8 Although favorable neurological outcome at 30 days after cardiac arrest has been shown to be associated with the collapse-to-ROSC interval,1–4 there is no established consensus on the relationship of the duration from collapse to the beginning of TH, from ROSC to the beginning of TH or from the beginning of TH to the achievement of target temperature. Among the methods used to induce and maintain hypothermia are infusion of ice-cold fluids, body surface cooling with a cooling blanket, cooling of blood inside the body by blood purification therapy, or use of an extracorporeal circulatory cooling system.9–15 Among these, ice-cold fluid infusion for rapid cooling has been reported by many studies as a rapid, simple and inexpensive method.16–22 It has been previously reported that infusion of 30 mL/kg or 2L of cold fluid at 4℃ allowed TH to be rapidly induced without a high incidence of complications and allowed the core temperature to be more rapidly reduced in OHCA patients in whom ROSC was achieved prior to hospital arrival.19–21 However, many studies of prehospital induction of TH have reported that prehospital infusion of ice-cold fluids did not improve survival or neurological outcome as compared with induction of TH after hospital arrival.16,21–24 Furthermore, prehospital infusion of ice-cold fluids has been reported to cause recurrent cardiac arrest or in-hospital pulmonary edema in many cases, and current guidelines do not recommend prehospital induction of TH with ice-cold fluids.25
Editorial p 1849
We previously reported a strong correlation between the duration of cardiac arrest and the 30-day neurological outcome.26 We hypothesized that the effect of rapid induction of TH with ice-cold fluids could vary with the time interval from collapse to ROSC. Therefore, we investigated whether the effect of in-hospital-induced TH by ice-cold fluids depends on the interval from collapse to ROSC.
To investigate the effects of TH on patients with OHCA, we conducted a multicenter registry (retrospective and prospective cohort study) in Japan (Japanese Population based Utstein-style study27 with defibrillation and basic/advanced Life Support Education and implementation-Hypothermia [J-PULSE-HYPO study registry]28). An overview of the study was registered as a clinical trial (UMIN000001935, https://upload.umin.ac.jp/cgi-open-bin/ctr/ctr.cgi?function=brows&action=brows&type=summary&recptno=R000002348&language=J), and ClinicalTrials.gov (Identifier: NCT00901134; http://clinicaltrials.gov/), both accessed 30 January 2011. Because the primary purpose of the J-PULSE-HYPO study registry is quality improvement, participating hospitals were not required to obtain individual informed consent. The present study was conducted in accordance with the ethical guidelines for epidemiological studies and was approved by the Ethics Committee of the National Cerebral and Cardiovascular Center. Also, the relevant review boards in all 14 participating centers approved the study protocol.
We designed the J-PULSE-HYPO study data collection software in a standardized format to enter information about each case of TH and distributed this software to 14 hospitals that participated in this study; the data were accumulated and registered. The information was abstracted from hospital medical records. The database contains information about TH after ROSC. Each patient was assigned a unique code, and no specific patient identifiers were transmitted to the central database repository. The data could be submitted on diskette or by encrypted transmission over secure internet.
PatientsFrom the J-PULSE-HYPO study registry, we selected comatose survivors with prehospital ROSC who met the following inclusion criteria: adults (≥18 years old) treated with TH, witnessed cardiac arrest, initial cardiac arrest due to shockable rhythm, cardiac arrest due to cardiac origin, stable hemodynamics after ROSC (including stabilization by drugs or assisted circulation, such as intra-aortic balloon pumping and/or veno-arterial extracorporeal membrane oxygenation [VA-ECMO]), target core temperature of 34℃ and cooling duration of 12–72 h. The exclusion criteria were patients with non-shockable OHCA, patients with non-cardiac origin (aortic dissection, pulmonary embolism, subarachnoid hemorrhage etc), pregnant women, drug addicts, patients who failed ROSC prehospital with standard CPR, or patients whose families refused to provide consent for the study.
TreatmentIndications for TH, time to start TH, the target core temperature, duration and rewarming, and methods of induction and maintenance were according to the protocols in place at each of the participating institutions (Supplementary Table). Respirator settings, use of pressor and/or inotropic drugs, use of sedatives and/or muscle relaxants, nutritional control, etc. during TH were left to the discretion of the hospitals attending respective cases. Some of the patients with defective circulation after ROSC had the help of VA-ECMO and/or intra-aortic balloon pumping (IABP) to stabilize circulation. Also, patients suspected of acute coronary syndrome underwent emergency coronary angiography and percutaneous coronary intervention (PCI), as appropriate.
Study EndpointsThe primary endpoint was favorable neurological outcome at 30 days after OHCA, defined as a Cerebral Performance Category 1 (good performance) or 2 (moderate disability) on a 5-category scale.27 Unfavorable neurological outcomes was defined as Cerebral Performance Category 3 (severe disability), 4 (vegetative state), or 5 (death). Secondary endpoint was survival (Cerebral Performance Category 1–4) at 30 days after OHCA.
Statistical AnalysisStudy patients were divided into 2 groups based on the median collapse-to-ROSC interval (shorter collapse-to-ROSC interval: <18 min vs. longer collapse-to-ROSC interval: ≥18 min), and then into 2 groups based on induction of cooling (rapid cooling by intravenous ice-cold fluids vs. standard cooling without intravenous ice-cold fluids). Baseline characteristics and study outcomes were compared using the chi-square test for categorical variables, and the Mann-Whitney U test and Log-rank test for continuous variables, as appropriate. Multivariate logistic regression analyses were done for independent predictors of the primary endpoint, including age, sex, presence or absence of bystander CPR, presence or absence of rapid cooling with ice-cold fluids, time interval from collapse to cooling and time interval from cooling to 34℃, as appropriate. All hypothesis tests were two-sided, with a significance level of less than 0.05. All statistical analyses were performed with SPSS software (version 25J).
Of the 452 comatose adult survivors treated with post-ROSC cooling in the J-PULSE HYPO study registry, 248 met the criteria for this study. These patients were divided into 2 groups based on the median collapse-to-ROSC interval (124 [50%] cases in the shorter collapse-to-ROSC interval group [<18 min], 124 [50%] cases in the longer collapse-to-ROSC interval group [≥18 min]) (Figure 1). Generally, the 2 groups had similar baseline characteristics inclusive of use of ice-cold fluids (rapid induction of cooling). However, significant differences were exhibited between the 2 groups in collapse-to-ROSC interval, cause of shockable OHCA, emergency coronary angiography and PCI (Table 1).
Study subjects.
Shorter collapse-to-ROSC interval group |
Longer collapse-to-ROSC interval group |
P value | |
---|---|---|---|
Age*, years | 60 (52–68) | 63 (54–69) | 0.20 |
Male sex | 101 (80.2%) | 104 (82.5%) | 0.63 |
Bystander CPR | 62 (50.0%) | 70 (56.5%) | 0.31 |
Time from collapse to ROSC* (min) | 12 (8–14) | 33 (24–47) | <0.001 |
Core temperature at hospital arrival* (℃) |
35.9 (35.2–36.3) (n=109) |
36.0 (35.0–36.3) (n=101) |
0.57 |
Disease causing cardiac arrest due to cardiac origin |
0.001 | ||
Ischemic heart disease | 81 (65.3%) | 107 (86.3%) | |
Arrhythmia | 22 (17.8%) | 7 (5.7%) | |
Cardiomyopathy | 16 (12.9%) | 6 (4.8%) | |
Other | 5 (4.0%) | 4 (3.2%) | |
Use of IABP | 33 (26.6%) | 64 (51.6%) | 0.82 |
Use of VA-ECMO | 14 (11.3%) | 45 (36.3%) | 0.52 |
CAG during initial care | 99 (79.8%) | 112 (90.3%) | 0.02 |
PCI during initial care | 58 (46.8%) | 76 (61.3%) | 0.02 |
Use of ice-cold fluids | 63 (50.8%) | 61 (49.2%) | 0.80 |
Volume of ice-cold fluids* (L) | 2.0 (1.0–2.0) | 1.5 (1.0–2.0) | 0.44 |
Time from collapse to beginning cooling* (min) |
80 (43–173) | 70 (50–165) | 0.29 |
Time from collapse to 34℃* (min) | 315 (184–450) | 70 (130–480) | 0.80 |
Method of maintaining cooling | 0.14 | ||
Direct blood cooling | 63 (50.8%) | 72 (58.1%) | |
Surface cooling | 60 (48.4%) | 48 (38.7%) | |
Combination | 0 (0%) | 2 (1.6%) | |
Unknown | 1 (0.8%) | 2 (1.6%) | |
Cooling duration (h) | 0.21 | ||
<24 | 65 (52.4%) | 61 (49.2%) | |
24–47 | 52 (41.9%) | 45 (36.3%) | |
48–72 | 6 (4.8%) | 13 (10.5%) | |
Unknown | 1 (0.8%) | 5 (4.0%) | |
Rewarming duration (h) | 0.24 | ||
<24 | 22 (17.7%) | 32 (25.8%) | |
24–47 | 39 (31.5%) | 29 (23.4%) | |
48–71 | 40 (34.2%) | 31 (25.0%) | |
≥72 | 16 (12.9%) | 17 (13.7%) | |
Unknown | 7 (5.6%) | 5 (4.0%) |
*Median (interquartile range). CAG, coronary angiography; CPR, cardiopulmonary resuscitation; IABP, intra-aortic balloon pumping; PCI, percutaneous coronary intervention; ROSC, return of spontaneous circulation; VA-ECMO, veno-arterial extracorporeal membrane oxygenation.
Figure 2 shows the 30-day survival and favorable neurological outcomes for the 2 groups stratified by the median collapse-to-ROSC interval (18 min). The shorter collapse-to-ROSC interval group (<18 min) had significantly higher frequencies of 30-day survival and favorable neurological outcome than the longer collapse-to-ROSC interval group (≥18 min). Figure 3 shows 30-day favorable neurological outcome divided into 2 groups according to induction method of cooling (rapid cooling vs. standard cooling). In the shorter collapse-to-ROSC interval group, no significant difference was observed in favorable neurological outcome between the 2 cooling groups (rapid cooling group: 79.4% vs. standard cooling group: 77.0%, P=0.75). In the longer collapse-to-ROSC interval group, favorable neurological outcome was significant higher in the rapid cooling group than in the standard cooling group (60.7% vs. 33.3%, P<0.01) and multivariate logistic regression analysis (Table 2) showed that use of ice-cold fluids resulted in a higher proportion of favorable neurological outcome than non-use of ice-cold fluids (adjusted odds ratio [OR]: 3.069, 95% confidence interval [CI]: 1.423–6.616, P=0.004). Even if the collapse-to-induction of cooling interval was used instead of ice-cold fluids (Table 3), collapse-to-induction of cooling interval was an independent factor of favorable neurological outcome in the longer collapse-to-ROSC interval group (adjusted OR: 0.994, 95% CI: 0.990–0.999, P=0.013).
30-day survival and favorable neurological outcome for the 2 groups stratified by a median collapse-to-ROSC interval of 18 min. ROSC, return of spontaneous circulation.
30-day favorable neurological outcome divided into 2 groups according to induction method of cooling (rapid cooling vs. standard cooling). ROSC, return of spontaneous circulation.
Variable | OR | 95% CI | P value |
---|---|---|---|
A. Shorter collapse-to-ROSC interval group (<18 min) | |||
Age (years) | 0.936 | 0.895–0.979 | 0.0039 |
Male sex | 2.499 | 0.831–7.517 | 0.1031 |
Presence of bystander CPR | 2.071 | 0.796–5.389 | 0.1356 |
Use of ice-cold fluids | 0.783 | 0.308–1.991 | 0.6072 |
Time from start of cooling to 34℃ (min) | 1.000 | 0.998–1.002 | 0.8275 |
B. Longer collapse-to-ROSC interval group (≥18 min) | |||
Age (years) | 0.971 | 0.937–1.005 | 0.0975 |
Male sex | 0.932 | 0.316–2.743 | 0.8976 |
Presence of bystander CPR | 1.357 | 0.621–2.964 | 0.4439 |
Use of ice-cold fluids | 3.069 | 1.423–6.616 | 0.0042 |
Time from start of cooling to 34℃ (min) | 1.001 | 0.999–1.002 | 0.4457 |
CI, confidence interval; CPR, cardiopulmonary resuscitation; OR, odds ratio.
Variable | OR | 95% CI | P value |
---|---|---|---|
A. Shorter collapse-to-ROSC interval group (<18 min) | |||
Age (years) | 0.936 | 0.894–0.979 | 0.0042 |
Male sex | 2.461 | 0.822–7.372 | 0.1076 |
Presence of bystander CPR | 2.015 | 0.775–5.239 | 0.1505 |
Time from collapse to start of cooling (min) | 1.001 | 0.996–1.007 | 0.6353 |
Time from start of cooling to 34℃ (min) | 1.000 | 0.998–1.002 | 0.8236 |
B. Longer collapse-to-ROSC interval group (≥18 min) | |||
Age (years) | 0.969 | 0.936–1.004 | 0.0853 |
Male sex | 0.728 | 0.248–2.135 | 0.5627 |
Presence of bystander CPR | 1.357 | 0.545–2.573 | 0.6685 |
Time from collapse to start of cooling (min) | 0.994 | 0.990–0.999 | 0.0126 |
Time from start of cooling to 34℃ (min) | 1.001 | 0.999–1.003 | 0.2445 |
CI, confidence interval; CPR, cardiopulmonary resuscitation; OR, odds ratio.
This study showed that effect of in-hospital rapid induction of cooling by intravenous ice-cold fluids for comatose survivors after witnessed shockable OHCA varied with the time interval from collapse to ROSC (median [interquartile range], 18 min [12–33 min]). In comatose survivors whose collapse-to-ROSC interval was ≥18 min, in-hospital rapid cooling improved 30-day favorable neurological outcome compared with standard cooling without intravenous ice-cold fluids (Figure 3; Tables 2,3). On the other hands, the 2 cooling groups had similar 30-day favorable neurological outcomes when the collapse-to-ROSC interval was <18 min.
Some randomized studies have reported the effect of prehospital induction of cooling by intravenous ice-cold fluids. Bernard et al22 and Kim et al23 showed no significant difference in neurological outcome between comatose survivors after OHCA who were treated with prehospital cooling by cold fluids and those who were treated with in-hospital cooling. However, those studies did not focus on the collapse-to-ROSC interval. Soga et al of the J-PULSE-HYPO Investigators26 reported that the collapse-to-ROSC interval was deeply involved in the neurological outcome of TH. Furthermore, the Japanese Circulation Society with Resuscitation Science Study Group reported that prehospital CPR efforts should be continued for at least 40 min for 30-day favorable neurological outcome in all adults with witnessed shockable OHCA to achieve ≥99% sensitivity of favorable 30-day neurological outcome.29 The longer the cardiac arrest interval, the greater the degree of cell injury after ROSC. Experimental data suggest that targeted temperature management suppresses pathways leading to delayed cell death and decreases the cerebral metabolic rate, and consequently reduces the release of excitatory amino acids and free radicals. Up to 28℃, each 1° decrease in core body temperature reduces brain metabolism by 6%. Hypothermia protects and suppresses cell injury and suppresses brain metabolism.30,31 These findings suggested that as the cardiac arrest interval lengthens, the extent of cell injury after ROSC increases. In other words, we thought that rapid cooling would reduce the extent of this cell injury in comatose survivors with longer cardiac arrest interval. In the longer collapse-to-ROSC interval group of our study, the rapid cooling group had significant shorter intervals from collapse-to-induction of cooling (median [interquartile], 45 [30–86] min vs. 168 [85–217] min, P<0.001) and from collapse to attainment of 34℃ (median [interquartile], 270 [132–403] min vs. 355 [224–517] min, P=0.04) than the standard cooling group. However, other analysis revealed that time to attainment of 34℃ did not strongly contribute to favorable 30-day neurological outcome (Tables 2,3). It may be important that rapid induction of TH prevents core temperature exceeding 37℃ and/or quickly attain 35℃. Actually, in some patients with a favorable neurological outcome, the core temperature, once it was successfully lowered by approximately 1℃ by infusion of cold fluids, increased again when the patient entered the intensive care unit, because undergoing coronary angiography or PCI takes time in primary care. Thereafter, it often takes hours to achieve the target core temperature, even if body surface cooling or intravascular cooling is used. In addition, patients who later have favorable neurological outcomes often resist TH by developing shivering, thereby prolonging the time to obtain the target core temperature and making it difficult to maintain the core temperature.
Study LimitationsFirst, this study was not a randomized controlled trial. Second, the protocols (candidate, target core temperature, induction timing, cooling duration, rewarming duration, cooling methods, etc.) of TH were established at each institution, because the J-PULSE-HYPO study was designed to investigate the effect of TH on neurological dysfunction after OHCA due to cardiac causes through a large multicenter enrollment. Third, the CoSTR 20154 recommended that selecting and maintaining a constant target temperature between 32℃ and 36℃ for those patients in whom temperature control is used (strong recommendation, moderate-quality evidence). We selected a target core temperature of 34℃ because that was used in nearly 90% of this study’s patients. It is unclear whether 34℃ is the optimal target core temperature or not. The results may have been different if a target temperature lower than 34℃ or a target temperature of 35℃ or 36℃ had been used. Fourth, the dose of cold liquid was left to the protocol of each institution. Many studies used 30 mL/kg or 2L of cold fluids.16,21–24 In this study, the dose of cold fluids was similar (Table 1). Fifth, various cooling methods were used in this study, but the proportions of those methods were similar. In particular, the extracorporeal circulatory system of cooling method was developed in Japan to directly cool the blood.14,28 Finally, neurological outcomes were measured at 30 days after OHCA, but some patients might recover more gradually. A recent consensus statement acknowledged that optimal times for follow-up after OHCA have yet to be established. A 3-month post-discharge period would balance the opportunity for recovery with the number of patients lost to follow-up.32
Rapid cooling by intravenous ice-cold fluids after hospital arrival improved neurologically intact survival in comatose survivors whose collapse-to-ROSC interval was delayed more than 18 min after shockable OHCA. Further studies are needed for target temperature management.
We are grateful to the medical staff of the 14 institutions participating in the J-PULSE-HYPO study registry, citizens in the participating areas and the ambulance personnel.
This study was funded in part by a research grant for cardiovascular disease (H19-Sinkin-003) from the Ministry of Health, Labor and Welfare, Japan.
N. Matsumoto received research grants from FUJIFILM Toyama Chemical and lecture fees from Nihon Medi-Physics and FUJIFILM Toyama Chemical.
The Ethics Committees of Nihon University Hospital with Reference No. 100101.
The deidentified participant data will not be shared.
Please find supplementary file(s);
http://dx.doi.org/10.1253/circj.CJ-20-0793