In our neonatal piglet model of hypoxic cardiac arrest, the use of epinephrine compared to placebo during resuscitation resulted in more than twice as many piglets with ROSC, and we found no difference in early biomarkers of brain damage between the groups six hours after resuscitation.
Our findings related to ROSC are consistent with some, but not all, previous experimental animal studies. Sobotka et al. [20] reported that chest compressions alone were insufficient to achieve ROSC and that epinephrine administration was critical to increase heart rate, carotid arterial pressure, and cerebral blood flow. Mendler et al. [21] and Solevåg et al. [22] also found that at least one dose of epinephrine was required for successful resuscitation. In contrast, Wagner et al. [23], Linner et al. [24], and McNamara et al. [25] concluded that epinephrine failed to increase the rate of ROSC. These conflicting results are likely explained by essential differences in the study protocols. The most striking difference between the three former studies and the three later studies is the chosen CA-criteria. In the study by Wagner et al. [23], CA was defined as HR < 25 % of baseline, and in the study by Linner et al. [24], CA was defined as HR < 50 bpm and MABP < 25 mmHg, however; if the criteria were not reached within 12 minutes of apnea, resuscitation was commenced. This indicates that the hypoxic insult was less severe in these studies. In the study by McNamara et al. [25], the CA criteria were similar to ours, yet the no-flow period was only 4 minutes compared to 5 minutes in our study. Furthermore, epinephrine was administered three minutes after resuscitation onset, compared to one minute in our study. Studies of pediatric CA have shown that the chance of achieving ROSC is decreased by every minute epinephrine administration is delayed [26, 27]. Additionally, only one dose of epinephrine was administered (two animals in our study required more than one dose epinephrine to achieve ROSC), and resuscitative efforts were discontinued if ROSC was not achieved within six minutes after chest compression onset (one animal in our study achieved ROSC beyond six minutes after chest compression onset).
Conversion from PEA or asystoli to VF was observed in 44% of the animals during the resuscitation period. Conversion from a non-shockable rhythm to a shockable rhythm (treated) is a good prognostic sign in pediatric CA [28]. As per the current standard for human neonatal resuscitation, defibrillation was not part of our procedure, and we observed no difference in the rate of ROSC between animals with and without VF in either experimental group (Supplementary S5). Swine are known to be more arrhythmogenic than humans [29]. High frequencies of VF were also observed by McNamara et al. [25]. These high frequencies emphasize the necessity of studies of arrest rhythms in human neonates.
In this study, very importantly, resuscitation with epinephrine did not result in worse markers of CNS outcome at six hours after ROSC. We found no difference in Lac/NAA ratio or NAA/Cr ratio between animals resuscitated with epinephrine compared to placebo. The Lac/NAA ratio is known to increase during the secondary phase of energy failure between 6 and 24 hours after cerebral hypoxia-ischemia [30]. Thus, it is possible that we failed to capture the differences, i.e., both beneficial and adverse effects of epinephrine, due to the relatively short interval between CA and MRS. However, Zheng et al. [31] showed that brain lactate levels peaked between 2 to 6 hours following hypoxia-ischemia, and during our pilot studies we performed both 6- and 12-hours MRS/MRI examinations with no additional information gained. Still, studies on long-term outcomes are warranted.
MRI-based DWI is a marker of cerebral edema [32]. In newborns with hypoxic ischemic encephalopathy (HIE) and in adults post CA, low ADC values have been associated with unfavorable neurologic outcome [33, 34]. The ADC values were low in animals with and without exposure to epinephrine (compared to healthy control piglets (n = 4) included in a previous study by our group; mean (SD) ADC = 974 (145) 10− 6 mm2/sec, unpublished data), indicating severe edema in both experimental groups. However, acute brain edema does not always persist; therefore, ADC measured six hours post ROSC, may not be a reliable predictor of irreversible tissue damage and later brain injury [35].
BOLD MRI measures regional differences in concentrations of oxy- and deoxyhemoglobin in response to neural activity via calculations of T2* values [36]. We found no difference between animals resuscitated with epinephrine compared to placebo in T2* measured in thalamus. The T2* values were low in both groups, (compared to healthy control piglets (n = 4), included in a previous study by our group; mean (SD) T2* = 63.3 (5.0) ms, unpublished data) possibly due to a compensatory increased oxygen extraction in the early hours after the hypoxic-ischemic insult. Furthermore, hypoperfusion may add to this finding. Early hypoperfusion (20 min to 12 hours after ROSC) has been demonstrated in animal models of adult/pediatric CA [37, 38] and has also been observed in neonates with HIE [39]. There is concern that epinephrine may impair cerebral perfusion and oxygenation [10]. However, in our BOLD measurements, epinephrine and placebo resulted in similar cerebral hemodynamics six hours post ROSC.
Our ROSC data and MRS/MRI data support that epinephrine increases short term survival without increasing brain injury. Our composite endpoint analysis of death or severe CNS outcome further supports this finding. However, our results were no longer significant due to a 25–30 % post-ROSC mortality in both experimental groups, and small absolute numbers of animals.
This study has a number of limitations; 1) we used anaesthetized and intubated animals. Anesthetics and drugs for pain relieve are necessary for ethical reasons, but may influence ROSC [40]. We did, however, pause all drugs during CA and resuscitation. 2) Newborns exhibit transitional cardiac- and lung physiology, while our study represents post-transitional neonatal hypoxic CA, and although the majority of CA in newborns are hypoxic in origin, co-morbidities such as infection or hypovolemia exist [41, 42]. 3) Although the anatomy of the newborn pig is very similar to that of the newborn human, it is possible that the beneficial effect of epinephrine on ROSC may be species specific. Interspecies differences in vascular sensitivity to catecholamines likely exist, and caution must be exercised when translating to human neonates. 4) All animals received TH, which might postpone the processes involved in the developing brain injury beyond the range of our observation period. This effect was, however, not evident in the study by Tang et al. [43] who investigated the effect of TH vs normothermia on brain injury by MRS/MRI at 6, 12, 24 and 72 hours post ROSC. 5) We used an MRS echo time of 135 ms for easy identification of lactate-peak due to peak inversion, and a small voxel to certify high spatial resolution. However, a small voxel decreases signal-to-noise ratio, which could challenge the detection of lactate peaks. A larger voxel and an echo time of 288 ms may help resolve this issue in future studies.