Hemodynamics and vasopressor support during targeted temperature management after cardiac arrest with non-shockable rhythm: A post hoc analysis of a randomized controlled trial

Background Patients admitted after cardiac arrest with non-shockable rhythm frequently experience hemodynamic instability. This study assessed the hemodynamic consequences of TTM in this sub population. Methods This is a post hoc analysis of the HYPERION trial (NCT01994772), that randomized patients to either hypothermia or normothermia after non-shockable rhythm related cardiac arrest. Patients with no, moderate or severe circulatory failure were identified with cardiovascular Sequential Organ Failure Assessment at randomization. Primary outcome was the number of patients at day 7 with resolution of shock, accounting for the risk of death (competing risk analysis). Secondary endpoint included neurological outcome and death at day-90. Results 584 patients were included in the analysis: 195 (34%), 46 (8%) and 340 (59%) had no, moderate and severe circulatory failure, respectively. Resolution of circulatory failure at day 7 was more frequently observed in the normothermia group than in the TTM group (60% [95 %CI 54–66] versus 53% [95 %CI 46–60], Gray-test: p = 0.016). The severity of circulatory failure at randomization was associated with its less frequent resolution at day 7 accounting for the risk of death (76 % [62–86] versus 54% [49–59] for patients with moderate versus severe circulatory failure, Gray test, p < 0.001, respectively). At day 90, the proportion of patients with Cerebral Performance Category score of 1 or 2 was lower in patients presenting severe circulatory failure (p = 0.038). Conclusion Circulatory failure is frequent after CA with non-shockable rhythm. Its severity at admission and TTM were associated with delayed resolution of circulatory failure.


Introduction
In consequence of whole body ischemia-reperfusion syndrome subsequent to cardiac arrest (CA), successfully resuscitated patients often develop severe post-resuscitation circulatory failure, 1,2 associating vasoplegia, myocardial dysfunction and systemic ischemic injury. [3][4][5] This circulatory failure, included in the well-known "postresuscitation disease", may be associated with different other organ failure and drastically impacts outcome. 6,7 Targeted temperature management (TTM) is so far recommended by international guideli-nes 8 to prevent or limit hypoxic-ischemic brain damage in CA patients remaining comatose after restoration of spontaneous circulation resuscitation (ROSC). 9 Circulatory failure may be potentially worsened by induced hypotermia, which is potentially a stress to the cardiovascular system with significant hemodynamic impact involving decreased cardiac index, lower heart rate and increased systemic vascular resistance. 3 While there is a lot of experimental and animal data [10][11][12][13] on the impact of temperature management and hemodynamics, [12][13][14] very few data have been published so far to describe the potential longitudinal impact of TTM on early hemo-dynamics of resuscitated CA patients. The HYPERION trial 15 recently provided data in favor of hypothermia in CA patients with initial non-shockable rhythm. Impact of such a strategy on hemodynamic parameters in patients presenting post-resuscitation circulatory failure has not been fully studied in this situation.
In the present study, using the Hyperion trial data, we aimed to investigate cardiovascular consequences of hypothermia in CA patients with non-shockable rhythm and its association with outcome.

Design
This was a post-hoc analysis of the Hyperion trial (NCT01994772), a multicentric French randomized controlled trial investigating TTM at 33°c versus normothermia during the first 24 hours in comatose (score 8 on the Glasgow Coma Scale) CA patients with nonshockable rhythm survivors. 15

Patients
In this post hoc analysis, all the patients included in the Hyperion trial were included. Severity of circulatory failure at admission was defined with the cardiovascular Sequential Organ Failure Assessment (cSOFA), 16 and patient were divided in 3 groups: no (cSOFA equal to 0 or 1, no vasopressor support), moderate (cSOFA equal to 2 or 3, inotrope support or low dose vasopressor support < 0.5 m g/kg/min) and severe circulatory failure (cSOFA equal to 4, high dose of vasopressor support > 0.5 mg/kg/min).
According to the French law, because the strategies used in both groups were considered to be components of standard care, informed consent for trial participation was not required. However, French data-protection authorities require that patients be given the opportunity to decline that their data to be used. Therefore, since the patients had coma, it was required that the closest available relatives received specific information about trial enrollment. Patients with no available relative were included in the trial, informed as soon as they regained competence, and were asked whether they wanted to remain in the trial; if the answer was negative, they were excluded from the analysis. Approval of the ethical comitee of the French Society of Intensive Care Medicine was obtained (CE-SRLF-11-335).

Intervention
Patients were randomized to intervention groups of moderate therapeutic hypothermia (33°C) or targeted normothermia (37°C). In the hypothermia group, hypothermia at 33°C was induced and maintained for 24 hours; an then slow rewarming with a maximum speed of 0.5°/ hour was performed to 36.5 to 37.5°C, and maintained for 24 hours. In the normothermia group, body temperature was maintained at 36.5 to 37.5°C for 48 hours. Hemodynamic evaluations were conducted to allow blood volume optimization. Hypovolemia was managed with crystalloid or colloid infusion, according to standard practice in each participating intensive care unit (ICU). The introduction of vasoactive drug treatment was at the discretion of the physicians, who followed international guidelines and local protocols. A mean arterial pressure of 65 mmHg and, if measured, central venous oxygen saturation (ScvO2) 70% were considered reasonable targets.

Endpoints
The primary endpoint of the present study was the number of patient with circulatory failure resolution at day 7. Resolution of shock was defined as the definite weaning of both vasopressors and inotropes for at least 24 hours, identified by a cSOFA equal to 0.
Secondary endpoints included the timing of circulator failure resolution, the association of circulatory failure with the Cerebral Performance Category (CPC) score on day 90 as well as day-90 mortality.

Statistical analysis
Continuous variables were expressed as median (1st-3rd quartiles) while categorical variables were expressed as frequencies (percentage). Baseline characteristics were compared across cSOFA categories using ordered Cochrane-Armitage test and Jonckheere's test for categorical and continuous variables, respectively.
The primary outcome was assessed using a competitive risk analysis and compared with a Gray Test. Indeed, the event of interest was the resolution of shock during the first 7 days. This outcome is impacted by the time of onset of the event of interest. Patients who have not experienced the event at the end of follow-up (day 7) were censored. To determine the risk of an event occurring at a certain time-point, a fundamental assumption is the outcome is not associated with an altered chance of the event occuring at any given moment. Death (from any causes) before resolution of shock is a competing event. For that puprpose, outcome was assessed using a competing risk model (Gray test, and cumulative incidence curves).
All statistical tests were two-sided, with p 0.05 considered significant. Statistical analysis was computed with R software (Version 3.6.3; R Foundation for Statistical Computing, Vienna, Austria) and the statistical package Crmpsk for the competing risk analysis.

Patients
All of the 581 patients included in the Hyperion trial were included in this post-hoc analysis: 284 allocated to TTM and 297 to normothermia. At inclusion, 195 (34%), 46 (8%) and 340 (59%) patients had no, moderate and severe circulatory failure, respectively. Baseline characteristics were not different between the 3 groups (Table1), except for a statistically longer no-flow time in patients who did not have circulatory failure (p = 0.045), a lower temperature at enrollment (p = 0.035) and a more frequent use of epinephrine (p = 0.012) in patients with severe circulatory failure. Proportions of patients receiving inotrope or vasopressor for the first 7 days after CA according to group of randomization are given in Fig. 1. Baseline characteristics and outcome of these patients are described in ESM Table 1. Hemodynamics variables and temperature during the first 68 hours of management are given in ESM Fig. 1. Twenty patients developed circulatory failure after the period of intervention (ESM Table 2).

Primary outcome
Among patients presenting circulatory failure at admission, resolution of shock at day-7 was more frequently observed in the normothermia than in the TTM group (

Secondary outcome
TTM interruption was not more frequent in patients presenting the most severe circulatory failure and was observed in 22 (13.4%), 4    Table 2). Proportion of patients with a CPC score of 1 or 2 at day 90 was lower in patients presenting severe circulatory failure at randomization (p = 0.038).

Discussion
In this post hoc analysis of the Hyperion trial involving 581 comatose patients after CA with non-shockable rhythm admitted to the ICU, we compared outcomes over severity of post-resuscitation circulatory failure within the first 7 days after ICU admission. To our knowledge, only few small studies have investigated hemodynamic during TTM in this subgroup of patients. We found that 1) while heart rate was significantly lower in patient with TTM, MAP was similar between the two groups of randomization 2) 66 % of patients evidenced post-resuscitation circulatory failure at randomization, and the severity of shock at admission and the intervention were associated with a delay in the timing of hemodynamic improvement 3) severity of shock at admission was associated with a worse functional outcome at day 90 evaluated with CPC score.
Post-CA syndrome is frequent in CA survivors resuscitated from initial non-shockable rhythm. 1,2,17 Its pathogenesis is complex, involving at different levels ischemia reperfusion syndrome 1 worsened by the common hypoxic etiology 15 as well as infectious phenomena and myocardial dysfunction. Consequent systemic inflammation induces vasodilatation and lower MAP 1, 17 motivating vasopressor infusion to avoid subsequent cerebral aggression, 5,17 which may drastically grave the functional outcome of these patients 7 . Precedent studies investigated the hemodynamic implication of TTM and found that several factors were associated with higher need of vasopressor, 3,6,19,20 as time to return to spontaneous circulation, age, or percutaneous coronary intervention. While the circulatory failure after ROSC is now well identified as a factor of poor prognosis, 18 the impact of TTM in these patients, potentially worsening the shock, is still unknown. Moreover, these moribund/instable patients are very frequently excluded from these studies, precluding us to drive conclusions. 9,21 In non-shockable rhythm CA survivors admitted to the ICU, we found that the severity of shock at admission was associated with a worse functional outcome at day-90. Furthermore, TTM delay the timing of shock resolution at day-7 although the rate of TTM interruption was similar in the three subgroups. Grand et al., 5 found similar findings in patient with prolonged TTM in their post-hoc analysis of the TT48 trial. Similar results have been described in a retrospective registry-based study of 412 CA patients. 6 In these situation, TTM may worsen the hemodynamic status, and consequently, could alter oxygen delivery to the brain. Indeed, hypothermia has negative hemodynamic effects by slowering metabolism and directly depressing myocardial function. 22 Patients can developped hypovolemia in consequence of an augmentation of urine output (also called û cold diuresisý). 23 Afterload is also raised through the release of catecholamins, leading to an increase of arterial resistance. 22 As a result of all these mechanisms, cardiac output is decreased. 22 Finally, it has been reported that, in patients with preexistent coronary artery disease, coronary vasconstriction may occur during hypothermia, 24 because of an endothelial dysfunction, and worsen oxygen delivery to heart. Then, whether to apply or not such a strategy in these patients remains a matter of debate, especially since the results of the TTM2 trial 9 in which the authors found discordant results in the subgroup of patients suffering cardiovascular failure at admission.
One strength of our study is the prospective and consecutive inclusion of patients and collection of data, in a multicentric randomized trial, increasing the validity of our result. However, we acknowledge that our study has several limitations. First, it was a nonprespecified post hoc analysis, and the study may have lack of power to identify association between variables and outcome. Second, some patients died during the period of analysis and hemodynamic data collection introducing a survival bias. However, the primary cause of death was withdrawal of care related to anoxic brain damage (64% of death), and we tried to limit this bias performing a competitive risk analysis, and expect the bias to be minimal. Third, the haemodynamic impact of TTM could have been more accurately evaluated using vasopressors and inotropic support doses over time as well as sedation doses. Unfortunately, these information were not collected in the HYPERION trial. While we believe this could have improved our findings, cardiovascular SOFA is a reliable tool used to evaluate one-point as well as longitudinal cardiovascular data. Finally, patients who developed circulatory failure after TTM were not analyzed as shocked patient, which can introduce a selection bias. However, we considered that this situation must not be considered as a post cardiac arrest syndrome but is due to other etiologies (sepsis, ventilator associated pneumonia, hemorrhage).

Conclusion
Circulatory failure is frequent after CA with non-shockable rhythm survivors admitted to the ICU. Severity of circulatory failure at admission and TTM strategy were associated with worse long-term neurological outcome and a delay in the timing of resolution of circulatory failure.

Conflict of Interest
Dr. Lascarrou reports consulting fees from BD and Zoll. None of the other authors has any conflicts of interest to disclose.