The Danish Out-of-Hospital Cardiac Arrest Study

BACKGROUND

Initially resuscitated Out-of-Hospital Cardiac Arrest (OHCA) patients admitted to a hospital are often in an unstable condition necessitating both circulatory and respiratory support. Most of these patients experiences the Post Cardiac Arrest Syndrome (PCAS) immediately after resuscitation due to a whole-body ischemia and reperfusion injury. It consists of four elements: 1) a universal systemic inflammatory response, 2) cerebral dysfunction, 3) myocardial dysfunction, and 4) the precipitation condition that led to the cardiac arrest - often a coronary occlusion.

The majority of OHCA patients have a presumed cardiac cause for their cardiac arrest. These patients are all evaluated for immediate coronary revascularization, stabilized with vasopressors, inotropes or mechanical circulatory support, and if comatose further sedated, connected to a ventilator and given targeted temperature management (TTM) for at least 24 hours. Half of the admitted resuscitated OHCA patients still die during the hospitalization despite all these interventions. The main cause of death is withdrawal of life sustaining therapy due to irreversible brain injury.

Several targets can be approached in order to potentially improve the outcome of OHCA patients. During the initial phase early after hospital admittance, interventions to dampen the systemic inflammatory response and efforts to secure optimum cerebral perfusion pressure seem intuitive important. Deep sedation has been required for TTM but may also have brain protective effects. After discontinuation of the sedation when the TTM is over, most patients have some degree of cerebral dysfunction. This may make them more prone to develop delirium in the days after awakening which is an independent risk factor in critical ill patients.

The DANOHCA trial will therefore evaluate new potential interventions to each of these different targets: systemic inflammation, cerebral perfusion, duration of sedation, and delirium.

INTERVENTIONS

High-dose steroids to dampen systemic inflammation in resuscitated OHCA patients:

By randomization, the patients will be allocated 1:1 to either dexamethasone (administered as dexamethasonephosphate) or placebo for three days. As soon as possible after admittance 20 mg of dexamethasonephosphate (Dexavit, Vital Pharma Nordic ApS, Denmark) or placebo will be given intravenously (i.v.) over 15 minutes - followed by 20 mg dexamethasonephosphate (or placebo) i.v. administered daily at 0600 (or at least 8 hours after initial dose) for two days, for a total of 3 doses.

Backrest elevation in post cardiac arrest care and outcome for OHCA patients:

By randomization, the patients will be allocated to have their headrest positioned at 5 degrees backrest vs 35 degrees straight elevation of backrest in Semi-Fowler's position (elevated lower limp position) during the initial 72 hours or until extubated. The adherence to the assigned stratum will be checked every 8 hours and cuff pressure will be assessed and corrected if needed at the same time during the intervention period. The backrest position intervention may be temporarily canceled by the treating physician if needed for procedures or mobilization but will return to the assigned position if invasive ventilator treatment with orotracheal intubation is continued. The intervention will be terminated if the patient is extubated, or a tracheostomy is performed, during the intervention period of 72 hours.

Early wake up and extubation in OHCA patients:

By randomization, the patients will be allocated 1:1 to early wakeup call and potential extubation after ≤6 hours or late wakeup call and extubation between 28-36 hours after admission to the ICU. Definition for "ready for extubation" will be: GCS≥12, RASS 0 to -1, able to raise arm or voluntary hand shake on command, spontaneous breathing trial and low ventilator settings (pressure support≤14, PEEP≤8 (10 if obese), and FiO2≤40%). A wakeup call may for both groups be aborted for the following reasons: seizures, respiratory distress, shock, or "other cause" with specification. Sedation prior to the scheduled wakeup calls is permitted in the early wakeup call group as needed for clinical care, while it is mandatory for the late wakeup call group. For both groups, sedation as needed for clinical care will be permitted after the scheduled wakeup call.

Prophylactic use of olanzapine to prevent delirium in patients resuscitated from OHCA:

By randomization, the patients will be allocated 1:1 to olanzapine 10 mg (Olanzapin, Accord Healthcare B.V., The Netherlands) administered by feeding tube (or orally in awake patients) or matching placebo at admission and again the following two evenings until discharge from ICU or a maximum of 3 doses. As soon as possible after arriving at the ICU: olanzapine/placebo is given by feeding tube. Thereafter 10 mg olanzapine or matching placebo will be given by feeding tube (or orally in awake patients) the following two evenings at 1800 (with a minimum of 12 hours between the initial doses) for a total of 3 doses. Due to the potential QT-prolonging effect, and accompanying concern for arrythmia, patients will be excluded prior to randomization if Long QT Syndrome (LQTS) has been confirmed or suspected, and all patients will be monitored by mandatory telemetry of heart rhythm for 96 hours or until life sustaining therapies are withdrawn. In case of delirium, patients are treated according to standard care most often including dexmedetomidine, haloperidol or midazolam.

STUDY DESIGN

The study is an investigator-initiated, randomized, controlled, multicenter clinical trial with co-enrollment into 4 different interventions. The pharmacological interventions are placebo-controlled, double-blind interventions and physiological interventions of early wakeup and extubation, as well as the back-rest position interventions are open label. Following successful completion of screening procedures, patients will be randomized in a 1:1 fashion to receive either the active intervention or placebo/control intervention for each intervention. The randomization will be revealed by opening a sealed box (referred to as the "DANOHCA trial kit" below) containing the study drug vials/tablets, instructions for wakeup and extubation and backrest positioning.

The trial aims to include 1000 patients from 5 sites. Approximately 350 patients are eligible for inclusion at the five hospitals combined each year, and therefore inclusion is expected to be concluded after 3-4 years.

Sub-studies further investigating the effect of the interventions will be performed. A research biobank will be established. Further co-enrollment into additional trials must be permitted by the steering group after careful assessment of risk of interaction with the four per-protocol interventions.

STATISTICAL ANALYSES

The data will be analyzed and reported as 4 individual trials, hence no adjustment for multiple comparison is planned.

Should more than one of the interventions be non-neutral a post hoc evaluation of interaction will be performed and highlighted in the reporting of the results. Also, as an investigation of potential safety aspects related to possible interactions between the four interventions, interactions related to mortality and occurrence of other SAE' will be reported.

All analyses will be conducted in the modified intention-to-treat (ITT) population with a set two-level significance level of 0.05. Modified ITT in this regard means that patients included without fulfilling the inclusion criteria will be excluded from the analyses.

Throughout, categorical variables will be presented as numbers (frequencies), whereas continuous variables will be presented as mean ±SD if normally distributed, and as median (25th percentile-75th percentile) if non-normally distributed. The primary analysis of primary endpoint will include the modified ITT population. Further the investigators plan a sensitivity analysis in the per protocol population to assess the effect. The per protocol cohort will be defined as only those trial participants who have received ALL scheduled doses of the pharmacologic interventions, have adhered to the assigned back-rest position for ≥80% of the time, and have had a wake-up call attempted within the assigned time frames.

Continuous endpoints being assessed at multiple time points will be analyzed by application of linear mixed models of covariance. The main result of these analyses will be the treatment-by-time interaction as a marker of whether the individual endpoint changes differently over time in the active intervention vs. the placebo arm, and group differences will be reported as relative differences in %. Logarithmic transformation will be applied to approximate normal distribution as appropriate.

Categorical endpoints by treatment allocation will be analyzed by the Chi-Squared or the Fisher's exact test, as appropriate.

For analysis of survival data, Kaplan-Meier curves for each allocation group will be estimated, graphically displayed, and compared by the log-rank test. Further Cox proportional hazard models will be applied to assess differences in time to death between treatment groups. These models will sequentially be adjusted for the interaction between treatment allocation, and each of the following variables: sex, age, time to ROSC, lactate level upon admission, shockable primary rhythm, witnessed cardiac arrest, bystander performed cardiopulmonary resuscitation, presence of STEMI in the admission ECG. Further, models stratified by cause of death (cardiovascular, neurological, multi-organ failure) will be applied as hypothesis generating. Any changes from the pre-specified analysis plan will be reported.