Neuroprotective body hypothermia among newborns with hypoxic ischemic encephalopathy: three-year experience in a tertiary university hospital. A retrospective observational study

ABSTRACT CONTEXT AND OBJECTIVE: Neonatal hypoxic-ischemic encephalopathy is associated with high morbidity and mortality. Studies have shown that therapeutic hypothermia decreases neurological sequelae and death. Our aim was therefore to report on a three-year experience of therapeutic hypothermia among asphyxiated newborns. DESIGN AND SETTING: Retrospective study, conducted in a university hospital. METHODS: Thirty-five patients with perinatal asphyxia undergoing body cooling between May 2009 and November 2012 were evaluated. RESULTS: Thirty-nine infants fulfilled the hypothermia protocol criteria. Four newborns were removed from study due to refractory septic shock, non-maintenance of temperature and severe coagulopathy. The median Apgar scores at 1 and 5 minutes were 2 and 5. The main complication was infection, diagnosed in seven mothers (20%) and 14 newborns (40%). Convulsions occurred in 15 infants (43%). Thirty-one patients (88.6%) required mechanical ventilation and 14 of them (45%) were extubated within 24 hours. The duration of mechanical ventilation among the others was 7.7 days. The cooling protocol was started 1.8 hours after birth. All patients showed elevated levels of creatine phosphokinase, creatine phosphokinase-MB and lactate dehydrogenase. There was no severe arrhythmia; one newborn (2.9%) presented controlled coagulopathy. Four patients (11.4%) presented controlled hypotension. Twenty-nine patients (82.9%) underwent cerebral ultrasonography and 10 of them (34.5%) presented white matter hyper-echogenicity. Brain magnetic resonance imaging was performed on 33 infants (94.3%) and 11 of them (33.3%) presented hypoxic-ischemic changes. The hospital stay was 23 days. All newborns were discharged. Two patients (5.8%) needed gastrostomy. CONCLUSION: Hypothermia as therapy for asphyxiated newborns was shown to be safe.


INTRODUCTION
Perinatal asphyxia consists of decreased metabolic and nutritional intake from mother to fetus, thereby causing low fetal tissue perfusion, hypoxia, hypercapnia and acidosis.2] The outcomes from HIE range from intact survival to death.The spectrum of long-term morbidity among survivors ranges from mild motor and cognitive deficits to cerebral palsy and severe cognitive deficits. 3It used to be believed that HIE was caused by intrapartum events and umbilical cord prolapse, breech presentation, forceps delivery and maternal fever. 4wever, intrapartum factors have been found to be the cause of HIE in only 4% of affected newborns, while 69% of such cases have shown evidence of antenatal risk factors such as severe preeclampsia, maternal thyroid disease, viral infection, moderate to severe vaginal bleeding during pregnancy and maternal hypertension. 5,6Restriction of primary and secondary intrauterine growth has also presented a strong association with HIE. 4 However, little evidence of antenatal damage has been observed on brain magnetic resonance imaging (MRI) performed at an early stage on neonates suffering from HIE. 7 Until recently, clinical HIE treatment consisted basically of neonatal intensive care support, correction of metabolic respiratory and hemodynamic disorders and use of anticonvulsants.
However, studies published over the last eight years have individually or collectively shown the effectiveness of using body hypothermia for HIE treatment, thereby promoting increased survival without neurological sequelae, with lower morbidity and mortality. 8,9nce the time of Hippocrates, therapeutic hypothermia has been applied to various clinical conditions.In a book written by Sir John Floyer, a physician and writer in the 17 th century, a procedure in which a stillborn infant was immersed in cold water to induce spontaneous breathing is described. 10The history of modern science has included periodic attempts to standardize the use of therapeutic hypothermia for a range of cerebral injury; this movement has been accelerated by advances in cardiopulmonary resuscitation. 10,11Randomized controlled trials on the effectiveness of therapeutic hypothermia in adult patients after cardiac arrest have shown improvements in survival and neurological outcomes. 12pothermia reduces brain injury through its impact on several biological processes.It reduces vasogenic edema, hemorrhage and neutrophil infiltration.It limits the release of excitatory neurotransmitters and the accumulation of intracellular calcium.The production of free radicals is restricted by hypothermia, and thus cells and organelles are protected from oxidative damage during reperfusion.Also, it reduces the activation of cytokine and coagulation cascades by increasing the concentration of interleukin-10, an anti-inflammatory cytokine, and reducing tumor necrosis factor-alpha.Furthermore, hypothermia helps to maintain cerebral metabolism during and after cerebral attacks by decreasing the metabolic rate of glucose and oxygen.Through reducing caspase-3 activity and increasing the expression of the anti-apoptotic protein BCL-2, hypothermia limits neuronal apoptotic death. 13ong newborns, this therapy consists of reducing body temperature by three to four degrees Celsius (moderate hypothermia), starting within six hours of birth and continuing for 72 hours.5][16] Our neonatology service has been using hypothermia as a routine clinical therapeutic practice since 2009 and was the pioneer in introducing a hypothermia protocol to Brazilian neonatal units, with effective participation by all the professionals involved in high-risk newborn care, from birth in the obstetric delivery room to procedures in the neonatal intensive care unit.

OBJECTIVE
Our aim in this study was therefore to report on our experience of three years of hypothermia therapy on asphyxiated newborns in a tertiary university hospital, demonstrating the characteristics of the newborn population undergoing this therapy, birth conditions, clinical complications, adverse effects, features of body temperature control and follow-up during the stay in the neonatal unit.

METHODS
This was a retrospective observational study on a cohort of newborns with HIE who fulfilled the criteria for inclusion in a total body cooling protocol at a neonatology service in a tertiary-level We performed the following imaging studies: cerebral ultrasonography (US) on first day of life and brain computed tomography (CT) and/or magnetic resonance imaging (MRI) between the 5 th and 21 st days.We took into consideration changes in the analyses that were consistent with hypoxic-ischemic injury.
We used the following assessments to determine the final outcome during the hospitalization period: length of hospital stay, mortality and need for gastrostomy.
For the statistical analysis, we firstly performed descriptive analysis on the data.Qualitative variables were represented as absolute and relative frequencies.Numerical variables were expressed as mean, median, standard deviation (SD) and minimum and maximum values.In cases in which we found great variability in the samples, we chose to use the median.We used the chi-square test to compare qualitative variables and the nonparametric Kruskal-Wallis test to compare the Apgar variable.P-values (P) ≤ 0.05 were considered statistically significant.We used the Minitab-15 statistical software.
The study was approved by our institution's Ethics Committee for Human Research.

RESULTS
Thirty-nine infants met the criteria for inclusion in the hypothermia protocol during the study period.Four of them had to be withdrawn from the study: two developed refractory septic shock, one failed to maintain a temperature below 34 °C and one evolved with severe coagulopathy.The perinatal characteristics of the 35 patients studied are shown in Table 1.
With regard to the type of delivery (vaginal or cesarean), we did not find any statistically significant difference between the sexes (P = 0.42).Comparing the type of delivery with the Apgar scores at 1 and 5 minutes, there were no statistical differences (P = 0.096 and 0.287).There were no statistical differences in the Apgar scores at 1 and 5 minutes in relation to sex (P = 0.847 and 0.296).Nine pregnant women (25.7%) had no complications during pregnancy.The main maternal complication was infection, diagnosed in seven patients (20%).Infection was also the most commonly diagnosed complication in the newborns, found in 14 of them (40%).Five of these (35.7%) had positive blood cultures.Seizures were diagnosed in 15 newborns (43%): 11 of them (73.3%) received monotherapy treatment (phenobarbital) and four (27%) received diphenylhydantoin in association with phenobarbital.
Thirty-one patients (88.6%) required mechanical ventilation and 14 (45%) of these patients were extubated within 24 hours of life.Among the other 17 infants, the duration of mechanical ventilation was 7.7 days (SD = 5.6), with a median of six days.
Body cooling began on average 1.8 hours (SD = 1.8) after birth (median of 1 hour).The target temperature was attained, on average, 1.2 hours after the beginning of the protocol, and 21 infants (60%) achieved this temperature within 1 hour.Brain MRI was performed on 33 infants, and 11 of them (33.3%) had signs suggestive of hypoxic-ischemic encephalopathy.
All the infants were discharged from the neonatal unit.Only two patients (5.8%) were discharged with gastrostomy.

DISCUSSION
This study reports on a three-year experience of administering hypothermia therapy to asphyxiated newborns in a tertiarylevel university hospital.The criteria for infant inclusion and exclusion were based on previous studies on safety.The gestational age for inclusion in the study (more than 35 weeks) made it possible to differentiate encephalopathy attributed to perinatal hypoxia from other problems relating to prematurity. 17Hypothermia is applied within the first six hours of life because this is the therapeutic window within which the neuroprotective effect relating to reduction of cerebral metabolism, reduction of excitatory neurotransmitter activity, suppression of free radical release, inhibition of the apoptotic process and reduction of the release of inflammatory mediators is most effective.In general, the efficacy of the neuroprotective effect diminishes if the cooling period starts after the therapeutic window, but evidence suggests that the neurological injury in HIE cases continues beyond this period. 18e fact that no significant complications attributed to therapeutic hypothermia were observed, with successful maintenance of body temperature within 72 hours, shows that this therapy is safe and easy to control and maintain when a multidisciplinary team is involved with the main aim of reducing neurological sequelae among infants with HIE.Regarding sedation, given that it is difficult to determine the degree of pain or discomfort in newborns with HIE who undergo hypothermia, and that opioids appear to boost the neuroprotective effect of this therapy, we considered that it would be beneficial to use systematic sedation with low doses of opioids in this group of infants. 9][10][11][12][13][14] Hypothermia could increase the already-present risk of occurrences of necrotizing enterocolitis secondary to intestinal ischemia and hypoxia.This approach contributed towards the absence of gastrointestinal complications in our patients, as also seen in the results from the abovementioned studies.Although we used cold packs rather than thermal mattresses, we did not record any difficulties in maintaining the temperature within the established goals.Use of gel packs has been reported to reduce death or developmental delays at six months of age among infants with HIE, without increasing the adverse events, 19 although use of servo control systems can avoid temperature fluctuations. 20Moreover, we had no difficulties during reheating.
We did not find any serious side effects or complications that were directly related to hypothermia.As in other studies, [9][10][11][12][13][14] we had cases of sinus bradycardia without hemodynamic effects.Cooling induces QTc prolongations in hypoxic fullterm newborns that become normal with rewarming. 21In our study, no patient had severe arrhythmia requiring antiarrhythmic intervention.
We had clinical diagnoses of seizures in 15 infants, representing 43% of the study group.3 There may be a selective neuroprotective effect on the cerebral cortex from hypothermia, with improved neurocognitive functioning. 24The ideal time to perform MRI is between the fifth and eleventh days of life, because the imaging findings have a better relationship to the neurological prognosis. 25We observed seizures in only 43% of the newborns, and there were no deaths, which may have demonstrated that the profile of the group studied was more similar to patients with moderate HIE.However, we cannot rule out the hypothesis that hypothermia is an effective form of therapy for reducing neurological injuries.
We succeeded in reducing the newborns' body temperature, at the beginning of and during hypothermia therapy, and successfully performed rewarming.We did not have any potentially serious complication relating to this therapy.The patients had favorable outcomes during the hospitalization period, given the high degree of neurological sequelae in this group of patients.
In most infants with HIE, feeding impairments are present from the neonatal period onwards or started within the first six months. 26Most of our patients were discharged with oral feeding, which provided hope, both for us and for the parents, that the neurological outcomes among these patients will be quite satisfactory, with a better quality of life.It is necessary to follow this group of newborns clinically and neurologically, in order to better assess the results from this therapy.
More studies involving hypothermia are needed, with longer exposure, lower temperatures and selective or whole-body hypothermia, and using other related therapies such as erythropoietin, xenon, melatonin, topiramate and stem cell therapy. 27,28

CONCLUSION
We concluded that the use of body hypothermia for neuroprotection among neonates with HIE within daily practice in our hospital did not lead to any problems when performed with a specific protocol and a trained multidisciplinary team.
Hypothermia as a form of therapy for asphyxiated newborns was shown to be safe.

university hospital (Chart 1 )
and were treated between May 2009 and November 2012.The laboratory tests indicative of perinatal asphyxia comprised gas analysis (pH and base excess, BE) on samples collected from cord blood or from the newborn within the first hour of life and assays on the enzymes creatine phosphokinase (CPK), creatine phosphokinase-MB (CK-MB) and lactate dehydrogenase (LDH) collected from the infant before the sixth hour of life.Furthermore, we analyzed the time required to reach the target temperature of 34 °C.The possible adverse effects evaluated were cardiac arrhythmias, hypotension, coagulopathy, infection and death.

Chart 1 .
Total body cooling protocol for infants with hypoxicischemic encephalopathy

Table 3 .
Adverse effects and outcome during hospitalization