Prognostic Factors for Re-Arrest with Shockable Rhythm during Target Temperature Management in Out-Of-Hospital Shockable Cardiac Arrest Patients

Re-arrest during post-cardiac arrest care after the return of spontaneous circulation is not uncommon. However, little is known about the risk factors associated with re-arrest. A previous study failed to show a benefit of prophylactic antiarrhythmic drug infusion in all kinds of out-of-hospital cardiac arrest (OHCA) survivors. This study evaluated high-risk OHCA survivors who may have re-arrest with shockable rhythm during targeted temperature management (TTM). Medical records of consecutive OHCA survivors treated with TTM at four tertiary referral university hospitals in the Republic of Korea between January 2010 and December 2016 were retrospectively reviewed. Patients who did not have any shockable rhythm during cardiopulmonary resuscitation (CPR) or unknown initial rhythm were excluded. The primary outcome of interest was the recurrence of shockable cardiac arrest during TTM. There were 289 cases of initial shockable arrest rhythm and 132 cases of shockable rhythm during CPR. Of the 421 included patients, 11.4% of patients had a shockable re-arrest during TTM. Survival to discharge and good neurologic outcomes did not differ between non-shockable and shockable re-arrest patients (78.3% vs. 72.9%, p = 0.401; 53.1% vs. 54.2% p = 0.887). Initial serum magnesium level, ST segment depression or ventricular premature complex (VPC) in initial electrocardiography (ECG), prophylactic amiodarone infusion, and dopamine and norepinephrine infusion during TTM were significantly higher and more frequent in the shockable re-arrest group (all p values < 0.05). Normal ST and T wave in initial ECG was common in the non-shockable re-arrest group (p = 0.038). However, in multivariate logistic regression analysis, only VPC was an independent prognostic factor for shockable re-arrest (OR 2.806 (95% CI 1.276–6.171), p = 0.010). Initial VPC may be a prognostic risk factor for shockable re-arrest in OHCA survivors with shockable rhythm.


Introduction
While restoration of spontaneous circulation (ROSC) after initial cardiac arrest resuscitation efforts is often successful, some patients subsequently develop circulatory instability and cardiovascular

Setting and Study Population
This retrospective, multicentre, observational study was performed at four urban academic emergency departments in the Republic of Korea. Data were extracted from a TTM registry, which prospectively collected data of consecutive patients with non-traumatic comatose OHCA who were treated with TTM between January 2010 and December 2016. The study was approved by the research ethics committees of each hospital. We included consecutive patients who survived from medical-cause OHCA and were treated with TTM. We excluded patients without an initial rhythm record and patients without any documented shockable rhythm. We divided patients by whether they had an experience of re-arrest with shockable rhythm (shockable re-arrest) or did not experience re-arrest or had re-arrest with non-shockable rhythm (non-shockable re-arrest), such as pulseless electrical activity (PEA) or asystole. Re-arrest was defined as loss of pulse following 20 min of sustained ROSC. The grade of VPC was assigned by the 'Lown grading system for ventricular arrhythmia' (Table 1) [17].

TTM Protocol
In all comatose OHCA survivors, TTM was induced with intravenous cold saline and cooling devices (Blanketrol II (Cincinnati Subzero Products, Cincinnati, OH, USA), Arctic Sun Energy Transfer Pad (Medivance Corp, Louisville, CO, USA), or COOLGARD3000 Thermal Regulation System (Alsius Corporation, Irvine, CA, USA)). The target temperature of 33 • C was maintained for 24 h and then patients were rewarmed at a rate of 0.25 • C/h to 37.0 • C. Normothermia was then maintained at 36.5-37.0 • C with the same device for 72 h from ROSC. Temperature was monitored with an esophageal temperature probe or rectal temperature probe. Propofol, benzodiazepine, and opioids were used for sedation and analgesia. If required, a neuromuscular blocker was administered to control shivering. All patients received standard intensive care according to institutional protocols.

Data Collection
Demographic and clinical data, including age, sex, previous medical history, initial vital signs, outcomes, and resuscitation profiles, such as cause of arrest, initial rhythm, duration of resuscitation, emergency department (ED) management drugs, and shock, were obtained. Laboratory values on admission were retrieved from the TTM registry. Initial ECG and coronary angiographic findings and vasopressor and inotrope infusion data were obtained from hospital electronic medical records. The primary outcome was the occurrence of repeated cardiac arrest with shockable rhythm during TTM. The secondary outcome was survival to discharge and good neurologic outcome at discharge, which is cerebral performance category 1 or 2 [18].

Statistical Analysis
This study was a secondary analysis of another observational study that compared outcomes of patients who did or did not receive prophylactic amiodarone continuous infusion during TTM. Continuous variables were expressed as means ± standard deviation (SD) or medians with the interquartile range (IQR) if the assumption of a normal distribution was violated. Categorical variables were expressed as numbers and percentages. To analyse baseline characteristics and laboratory values between shockable re-arrest patients and the patients with non-shockable re-arrest, the Student's t-test was used to compare the means of normally distributed continuous variables and the Mann-Whitney U-test was used to compare non-continuous variables. The Chi-squared or Fisher's exact test was used to compare categorical variables. We evaluated the association between each risk factor and shockable re-arrest during TTM by logistic regression. Risk factors where p < 0.1 in the univariate analysis were included in the multivariate analysis with the backward elimination method to find independently associated risk factors. We calculated the odds ratio (OR) with 95% confidential interval (CI) for each model. All tests in this study were two-sided, and a p-value < 0.05 was considered to be statistically significant. All statistical analyses were performed using SPSS for Windows version 20.0 (IBM Corp., Armonk, NY, USA).

Results
Among the 883 medical-cause OHCA survivors treated with TTM, we excluded 11 patients who did not have rhythm data from initial CPR, and 446 patients who did not have initial shockable rhythm and did not have any documented shockable rhythm during CPR. A total of 421 patients were included, 289 with initial shockable rhythm and 132 with shockable rhythm during CPR. Of the included patients, 48 patients (11.4%) developed shockable re-arrest during TTM (Figure 1).

Results
Among the 883 medical-cause OHCA survivors treated with TTM, we excluded 11 patients who did not have rhythm data from initial CPR, and 446 patients who did not have initial shockable rhythm and did not have any documented shockable rhythm during CPR. A total of 421 patients were included, 289 with initial shockable rhythm and 132 with shockable rhythm during CPR. Of the included patients, 48 patients (11.4%) developed shockable re-arrest during TTM ( Figure 1). The median time from ROSC to shockable re-arrest was 6.3 (interquartile range (IQR) 4.0-14.6) h. Re-arrest events occurred during the TTM induction period (54.0%), TTM maintenance (32.0%), rewarming (6.0%), and post-rewarming (8.0%).

Discussion
In this study we found that VPC in post-ROSC ECG was independently associated with shockable re-arrest during TTM. Patient survival to discharge and good neurologic outcome did not differ significantly between patients with shockable re-arrest or non-shockable re-arrest.
A recent observational study reported that 24.0% of restored cardiac arrest patients experienced recurrence of cardiac arrest and the median time from ROSC to re-arrest was 5.4 h (IQR 1.1-61.8) [9]. In their subgroup analysis of the 381 patients who experienced re-arrest, 108 (26%) patients had shockable rhythm during initial CPR; 60 (56%) patients had shockable re-arrest, 23 (38%) of whom survived to discharge [9]. Another study with a larger population, found that re-arrest occurred in 17.5% of ROSC patients, 66.4% of whom had a shockable rhythm, and that re-arrest was inversely associated with survival (OR 0.19, 95% CI 0.14-0.26) [14]. In our study the incidence of shockable re-arrest was 11.4%, which was less than the previous study, but the median time of re-arrest was similar (6.3 h (IQR 4.0-14.6)).
While previous studies have primarily focused on the incidence and outcome of re-arrest, the current study extends these findings by offering risk factors for shockable re-arrest in surviving OHCA patients during TTM. While shockable was not considered, Bhardwaj et al. reported that re-arrest was more associated with black race than white (OR 1.68, 95% CI 1.32-2.13), night time (19:00~07:00) than day time (OR 1.26, 95% CI 1.02-1.57) and more reverse associated in OHCA than in-hospital cardiac arrest (OR 0.46, 95% CI 0.33-0.62) [9]. In our study, VPC was significantly associated with shockable re-arrest by multivariate regression analysis. In Lown's grading system, the prevalence of grades 1 to 5 were 47.4%, 22.8%, 12.3%, 8.8% (4A), 7.0% (4B), and 1.8% respectively. However, the grade of VPC was not associated with shockable re-arrest. A previous study found that VPC within three months after cardiac arrest was prognostic of subsequent death; however, VPC patterns did not predict recurrent sudden cardiac death [19]. Different from our study, Weaver et al. reported that frequent multiform VPCs, bigeminal or trigeminal and ≥2 repetitive VPCs, were predictors of recurrent sudden cardiac death in OHCA survivors [20]. Because our study did not have enough VPC patients, we could not evaluate the role of frequent multiform VPCs in recurrent sudden cardiac death. In post-myocardial infarction populations, frequent VPC was associated with increased mortality, especially among those with ≥3 consecutive VPCs [21]. Therefore, guidelines recommended catheter ablation or anti-arrhythmia drugs such as amiodarone, beta blockers, and calcium channel blockers for patients with frequent VPCs [7].
This study was a first report to predict shockable re-arrest during TTM. Most of the previous studies conducted in prehospital or transporting settings [11,22]. During TTM, especially mild therapeutic hypothermia, one potential complication associated with cooling was electrographic changes including a prolonged PR interval, widening of the QRS complex, increased QT interval, and, sometimes, Osborne waves [23]. In our study we found electrographic changes in VPC were nearly three times higher in patients with shockable re-arrest (OR 2.806 (95% CI 1.276-6.171)). Moreover, post-cardiac arrest patients are often hemodynamically unstable due to the underlying etiology of the arrest, such as myocardial dysfunction and systemic ischemia reperfusion injury, sometimes requiring the administration of vasoactive adrenergic drugs [24]. Unfortunately, many adrenergic drugs increase cardiac arrhythmia [25], especially epinephrine, a non-selective catecholamine with potent alpha and beta adrenergic effects that may increase myocardial workload and disrupt the balance between oxygen supply and demand, increasing the risk of ventricular arrhythmias [26][27][28]. A recent paper reported that more frequent administration of epinephrine during cardiac arrest is associated with the development of secondary ventricular fibrillation or ventricular tachycardia [29]. However, in our study, numbers of epinephrine administration during initial CPR and also continuous infusion during TTM did not differ between patients with shockable re-arrest and non-shockable re-arrest (2.0 (0.0-5.8) vs. 2.0 (0.0-6.0), p = 0.410; OR 2.828 (95% CI 0.954-8.377), p = 0.061 in multivariate analysis). Moreover, continuous infusion of dopamine and norepinephrine during TTM also were not independently associated (OR 2.172 (95% CI 0.894-5.274), OR 1.304 (95% CI 0.606-2.807), respectively). In this study, amiodarone continuous infusion was more frequently used in patients with shockable re-arrest, likely because physicians use it in more electrically vulnerable patients. However, in the multivariate analysis amiodarone administration was not independently associated with shockable re-arrest.
This study had several limitations. First of all, it was a prospectively collected retrospective registry based study, so we did not have data associated with cardiac arrhythmia history, such as past history of inherited cardiomyopathy. While no significant associations were identified in the multivariate analysis, some patients received continuous amiodarone infusion, which can suppress shockable ventricular arrhythmia. This study was a secondary analysis of another study that compared continuous amiodarone infusion and not using prophylactic amiodarone for recurrent ventricular arrhythmia during TTM [16]. Therefore, we also only analysed patients who had confirmed shockable rhythm during initial CPR and we had not detail data of patients who had no shockable rhythm despite that they might have a shockable re-arrest. In this study we did not included intoxication and trauma patients because the number of these patients was very small and they rarely experience cardiac arrhythmia. In addition, while some intoxication is very closely associated with arrhythmia, the treatment is different from primary cardiac arrhythmia and most of these patients need sodium bicarbonate; therefore, our findings may not be generalizable.

Conclusions
In surviving OHCA patients with documented shockable ventricular arrhythmia during initial CPR, VPC in post-ROSC ECG was associated with shockable re-arrest during TTM. While the management of shockable re-arrest remains a challenge, the risk factors of re-arrest identified in this study will be helpful to better understand and address this phenomenon.