Abstract
We describe a patient with severe accidental hypothermia (≤25.4°C) and prolonged refractory ventricular fibrillation, lasting at least 4 hours and 8 minutes, who underwent cardiopulmonary resuscitation with extracorporeal membrane oxygenation and survived without neurologic deficit.
Case Report
A 54-

Defibrillation attempt by prehospital EMS, resulting in only few beats of organized rhythm, which quickly deteriorated to ventricular fibrillation. EMS, emergency medical services.
ACLS, advanced cardiac life support; BLS, basic life support; EMS, emergency medical services; ROSC, return of spontaneous circulation; VA ECMO, veno-arterial extracorporeal membrane oxygenation; VF, ventricular fibrillation.
During ongoing chest compression, veno-arterial extracorporeal membrane oxygenation (ECMO) was placed percutaneously through the right femoral artery (15-Fr cannula) and vein (19-Fr cannula) using fluoroscopy and ultrasound guidance (Goslar et al., 2016). Total delay to ECMO, which was set to 2.5 L/min of retrograde flow, was 2 hours and 54 minutes. Using the ECMO circuit together with infusion of warm saline, the patient gradually rewarmed. Successful defibrillation to sinus rhythm with ST depression in several leads resulting in sustained ROSC was achieved 1 hour and 22 minutes after ECMO start at a core temperature of 30°C (Fig. 2). Because of relatively mild increase of cardiac troponin I ultra (peak 9.8, normal value <0.06 μg/L), normalization of electrocardiogram, and absence of echocardiographically obvious segmental akinesia/hypokinesia, coronary angiography was not performed. Early hospital stay was complicated by severe pulmonary edema, which was apparent already on thoracic computed tomography (CT) scan immediately after ECMO implantation. Mechanical ventilation with 100% oxygen and high positive end-expiratory pressure was required. Because of gradual recovery of cardiac function and improvement in pulmonary function, ECMO could slowly be discontinued and cannulas surgically removed. Metabolic acidosis resolved after 24 hours of treatment. After 2 days, the patient was successfully weaned off mechanical ventilation and extubated after confirming sensible contact with the patient. After 17 days of treatment in the ICU, he was discharged to the ward. Gradual neurological recovery was observed. Transient hyperactive delirium resolved spontaneously within a few days. Due to persistent symptoms of depression, sertraline was given. After 39 days, the patient was discharged from the hospital with normal cardiac function and without neurological and cognitive deficits. After discharge from hospital, a psychiatric review with cognitive testing was performed that confirmed a normal cognitive status of the patient.

12-Lead electrocardiogram recorded immediately after successful defibrillation, which resulted in sustained reestablishment of spontaneous circulation. Sinus rhythm with ST depression in several leads was documented.
Discussion
We reported a patient with severe accidental hypothermia (≤25.4°C) and refractory VF, lasting at least 4 hours and 8 minutes, who was treated with E-CPR and survived without neurological deficit. According to our knowledge, this is one of the longest VF durations with good clinical outcome. E-CPR was undoubtedly crucial because it provided both immediate and sufficient organ perfusion as well as effective rewarming until the temperature threshold for successful defibrillation was achieved. However, it has to be emphasized that the entire period of refractory VF, which may be divided into ≤8 min of no-flow state, 2 hours 46 minutes of low-flow state during chest compression, and 1 hour 22 minutes of ECMO-supported organ perfusion until sustained ROSC was achieved, was under profound hypothermia. Its protective effect was undoubtedly crucial to minimize cellular ischemic/reperfusion injury and prevent irreversible organ damage as it has been proven successful in several case reports (Gilbert et al., 2000; Wanscher et al., 2012).
Retrospectively, we believe that our patient developed profound hypothermia while immersed in the river for an unknown period of time before VF was triggered. This probably happened after lay witnesses pulled him out of the water and before EMS arrival. In submerged patients or patients found down in cold circumstances, deep hypothermia should always be considered as a cause of arrest; temperature should be measured and if the patient's core temperature is <28°C, attempts should be made to rewarm the patient to 30°C (but no higher than 32–33°C). Our indication for E-CPR was based on the resuscitation guidelines that encourage transfer of hypothermic cardiac arrest victims to an ECMO center (Truhlář et al., 2015). This recommendation was inspired by unexpected and excellent neurological recovery even after very prolonged cardiac arrest, especially if the reason was hypothermia itself. (Paal et al., 2016; Kosiński et al., 2017) We postulated that hypothermia itself caused cardiac arrest since the patient was pulled out of water soon after becoming unconscious and therefore decided for E-CPR after at least 2 hours and 3 minutes of refractory VF. The mechanical compression device in the field and during transport was probably one of the key factors for maintaining low-flow circulation that prevented further hypoxic injury. Since the catheterization laboratory was occupied, ECMO cannulas were implanted in the emergency department using ultrasound and fluoroscopy with a C-arm. We deliberately used a small arterial cannula to omit the need for distal limb perfusion. We postulated that VF will remain refractory until the patient is sufficiently rewarmed. Indeed, stable sinus rhythm with ROSC was achieved only after the core temperature reached 30°C. Of note, despite prolonged VF, there was no significant postresuscitation myocardial dysfunction. This may be best explained by cardioprotective effects of profound hypothermia (Hsu et al., 2009).
Beside resistant VF, a large amount of hemorrhagic frothy sputum was aspirated from the endotracheal tube already during ACLS. It is known that the most frequent complications after weaning from ECMO appeared to be adult respiratory distress syndrome and severe pulmonary edema (Vretenar et al., 1994). The reason is probably ischemia–reperfusion injury and severe diastolic dysfunction along with a large volume infusion given for rewarming (Lauri, 1996; Tveita et al., 1998). Since profound respiratory failure fortunately resolved relatively fast, conversion to veno-venous ECMO was not needed.
Conclusion
Our patient is an example of the well-known old phrase that »no one is dead until warm and dead«. This rule should be respected regardless of the duration of cardiac arrest if the patient is in profound hypothermia. However, to save such patients, E-CPR providing both temporary organ perfusion and effective rewarming seems to be essential.
Footnotes
Author Disclosure Statement
M.N. receives speaker fees from Getinge. Other authors have no competing financial interests.
