Mari Falck

Treatment temperature and insult severity influence the neuroprotective effects of therapeutic hypothermia

Therapeutic hypothermia (HT) is standard care for moderate and severe neonatal hypoxic-ischaemic encephalopathy (HIE), the leading cause of permanent brain injury in term newborns. However, the optimal temperature for HT is still unknown, and few preclinical studies have compared multiple HT treatment temperatures. Additionally, HT may not benefit infants with severe encephalopathy. In a neonatal rat model of unilateral hypoxia-ischaemia (HI), the effect of five different HT temperatures was investigated after either moderate or severe injury. At postnatal-day seven, rat pups underwent moderate or severe HI followed by 5 h at normothermia (37 °C), or one of five HT temperatures: 33.5 °C, 32 °C, 30 °C, 26 °C, and 18 °C. One week after treatment, neuropathological analysis of hemispheric and hippocampal area loss, and CA1 hippocampal pyramidal neuron count, was performed. After moderate injury, a significant reduction in hemispheric and hippocampal loss on the injured side, and preservation of CA1 pyramidal neurons, was seen in the 33.5 °C, 32 °C, and 30 °C groups. Cooling below 33.5 °C did not provide additional neuroprotection. Regardless of treatment temperature, HT was not neuroprotective in the severe HI model. Based on these findings, and previous experience translating preclinical studies into clinical application, we propose that milder cooling should be considered for future clinical trials.

Hypothermia Does Not Reverse Cellular Responses Caused by Lipopolysaccharide in Neonatal Hypoxic-Ischaemic Brain Injury

Introduction: Bacterial lipopolysaccharide (LPS) injection prior to hypoxia-ischaemia significantly increases hypoxia-ischaemic brain injury in 7-day-old (P7) rats. In addition, therapeutic hypothermia (HT) is not neuroprotective in this setting. However, the mechanistic aspects of this therapeutic failure have yet to be elucidated. This study was designed to investigate the underlying cellular mechanisms in this double-hit model of infection-sensitised hypoxia-ischaemic brain injury. Material and Methods: P7 rat pups were injected with either vehicle or LPS, and after a 4-hour delay were exposed to left carotid ligation followed by global hypoxia inducing a unilateral stroke-like hypoxia-ischaemic injury. Pups were randomised to the following treatments: (1) vehicle-treated pups receiving normothermia treatment (NT) (Veh-NT; n = 40), (2) LPS-treated pups receiving NT treatment (LPS-NT; n = 40), (3) vehicle-treated pups receiving HT treatment (Veh-HT; n = 38) and (4) LPS-treated pups receiving HT treatment (LPS-HT; n = 35). On postnatal day 8 or 14, Western blot analysis or immunohistochemistry was performed to examine neuronal death, apoptosis, astrogliosis and microglial activation. Results: LPS sensitisation prior to hypoxia-ischaemia significantly exacerbated apoptotic neuronal loss. NeuN, a neuronal biomarker, was significantly reduced in the LPS-NT and LPS-HT groups (p = 0.008). Caspase-3 activation was significantly increased in the LPS-sensitised groups (p < 0.001). Additionally, a significant increase in astrogliosis (glial fibrillary acidic expression, p < 0.001) was seen, as well as a trend towards increased microglial activation (Iba 1 expression, p = 0.051) in LPS-sensitised animals. Treatment with HT did not counteract these changes. Conclusion: LPS-sensitised hypoxia-ischaemic brain injury in newborn rats is mediated through neuronal death, apoptosis, astrogliosis and microglial activation. In this double-hit model, treatment with HT does not ameliorate these changes.

Rectal temperature in the first five hours after hypoxia-ischaemia critically affects neuropathological outcomes in neonatal rats

BackgroundHyperthermia after hypoxia–ischemia (HI) in newborn infants is associated with worse neurological outcomes. Loss of thermoregulation may also be associated with greater injury.MethodsIn the postnatal-day 7 (P7) rat, the effect of 5 h of graded hyperthermia (38 °C or 39 °C) immediately after unilateral HI was compared with normothermia (NT, 37 °C) and therapeutic hypothermia (TH, 32 °C). Early (negative geotaxis) and late (staircase test) behavioral testing was performed, as well as neuropathology scoring in adulthood. Separately, P7 rats were exposed to HI, and individual nesting temperatures were monitored before analysis of neuropathology at P14.ResultsMortality increased as temperature was increased from 38 °C (0%) to 39 °C (50%) after HI. Hyperthermia also resulted in early behavioral deficits compared with NT. In adulthood, pathology scores in the thalamus, basal ganglia, cortex, and hippocampus increased as post-hypoxic temperature increased above NT. Significant global neuroprotection was seen in the TH group. However, no significant difference was seen between HI groups in the staircase test. One hour after HI, the core temperature of pups was inversely correlated with global pathology scores at P14.ConclusionEarly temperature is a significant determinant of injury after experimental HI. Spontaneous decreases in core temperature after HI may confound neuroprotection studies.

Monitoring of cerebral blood flow during hypoxia‐ischemia and resuscitation in the neonatal rat using laser speckle imaging

Neonatal hypoxic‐ischemic encephalopathy (HIE) is associated with alterations in cerebral blood flow (CBF) as a result of perinatal asphyxia. The extent to which CBF changes contribute to injury, and whether treatments that ameliorate these changes might be neuroprotective, is still unknown. Higher throughput techniques to monitor CBF changes in rodent models of HIE can help elucidate the underlying pathophysiology. We developed a laser speckle imaging (LSI) technique to continuously monitor CBF in six postnatal‐day 10 (P10) rats simultaneously before, during, and after unilateral hypoxia‐ischemia (HI, ligation of the left carotid artery followed by hypoxia in 8% oxygen). After ligation, CBF to the ligated side fell by 30% compared to the unligated side (P < 0.0001). Hypoxia induced a bilateral 55% reduction in CBF, which was partially restored by resuscitation. Compared to resuscitation in air, resuscitation in 100% oxygen increased CBF to the ligated side by 45% (P = 0.033). Individual variability in CBF response to hypoxia between animals accounted for up to 24% of the variability in hemispheric area loss to the ligated side. In both P10 and P7 models of unilateral HI, resuscitation in 100% oxygen did not affect hemispheric area loss, or hippocampal CA1 pyramidal neuron counts, after 1‐week survival. Continuous CBF monitoring using LSI in multiple rodents simultaneously can screen potential treatment modalities that affect CBF, and provide insight into the pathophysiology of HI.

Neonatal Systemic Inflammation Induces Inflammatory Reactions and Brain Apoptosis in a Pathogen-Specific Manner

Background: After neonatal asphyxia, therapeutic hypothermia (HT) is the only proven treatment option. Although established as a neuroprotective therapy, benefit from HT has been questioned when infection is a comorbidity to hypoxic-ischaemic (HI) brain injury. Gram-negative and gram-positive species activate the immune system through different pathogen recognition receptors and subsequent immunological systems. In rodent models, gram-negative (lipopolysaccharide [LPS]) and gram-positive (PAM3CSK4 [PAM]) inflammation similarly increase neuronal vulnerability to HI. Interestingly, while LPS pre-sensitisation negates the neuroprotective effect of HT, HT is highly beneficial after PAM-sensitised HI brain injury. Objective: We aimed to examine whether systemic gram-positive or gram-negative inflammatory sensitisation affects juvenile rat pups per se, without an HI insult. Methods: Neonatal 7-day-old rats (n = 215) received intraperitoneal injections of vehicle (0.9% NaCl), LPS (0.1 mg/kg), or PAM (1 mg/kg). Core temperature and weight gain were monitored. Brain cytokine expression (IL-6, IL-1β, TNF-α, and IL-10, via PCR), apoptosis (cleaved caspase 3, via Western blots), and microglial activation (Iba1, via immunohistochemistry) were examined. Results: LPS induced an immediate drop in core temperature followed by poor weight gain, none of which were seen after PAM. Furthermore, LPS induced brain apoptosis, while PAM did not. The magnitude and temporal profile of brain cytokine expression differed between LPS- and PAM-injected animals. Conclusion: These findings reveal sepsis-like conditions and neuroinflammation specific to the inflammatory stimulus (gram-positive vs. gram-negative) in the neonatal rat. They emphasise the importance of pre-clinical models being pathogen dependent, and should always be carefully tailored to their clinical scenario.

Hypothermia Is Neuroprotective after Severe Hypoxic-Ischaemic Brain Injury in Neonatal Rats Pre-Exposed to PAM 3 CSK 4

Background: Preclinical research on the neuroprotective effect of hypothermia (HT) after perinatal asphyxia has shown variable results, depending on comorbidities and insult severity. Exposure to inflammation increases vulnerability of the neonatal brain to hypoxic-ischaemic (HI) injury, and could be one explanation for those neonates whose injury is unexpectedly severe. Gram-negative type inflammatory exposure by lipopolysaccharide administration prior to a mild HI insult results in moderate brain injury, and hypothermic neuroprotection is negated. However, the neuroprotective effect of HT is fully maintained after gram-positive type inflammatory exposure by PAM3CSK4 (PAM) pre-administration in the same HI model. Whether HT is neuroprotective in severe brain injury with gram-positive inflammatory pre-exposure has not been investigated. Methods: 59 seven-day-old rat pups were subjected to a unilateral HI insult, with left carotid artery ligation followed by 90-min hypoxia (8% O2 at Trectal 36°C). An additional 196 pups received intraperitoneal 0.9% saline (control) or PAM1 mg/kg, 8 h before undergoing the same HI insult. After randomisation to 5 h normothermia (NT37°C) or HT32°C, pups survived 1 week before they were sacrificed by perfusion fixation. Brains were harvested for hemispheric and hippocampal area loss analyses at postnatal day 14, as well as immunostaining for neuron count in the HIP CA1 region. Results: Normothermic PAM animals (PAM-NT) had a comparable median area loss (hemispheric: 60% [95% CI 33–66]; hippocampal: 61% [95% CI 29–67]) to vehicle animals (Veh-NT) (hemispheric: 58% [95% CI 11–64]; hippocampal: 60% [95% CI 19–68]), which is defined as severe brain injury. Furthermore, mortality was low and similar in the two groups (Veh-NT 4.5% vs. PAM-NT 6.6%). HT reduced hemispheric and hippocampal injury in the Veh group by 13 and 28%, respectively (hemispheric: p = 0.048; hippocampal: p = 0.042). HT also provided neuroprotection in the PAM group, reducing hemispheric injury by 22% (p = 0.03) and hippocampal injury by 37% (p = 0.027). Conclusion: In these experiments with severe brain injury, Toll-like receptor-2 triggering prior to HI injury does not have an additive injurious effect, and there is a small but significant neuroprotective effect of HT. HT appears to be neuroprotective over a continuum of injury severity in this model, and the effect size tapers off with increasing area loss. Our results indicate that gram-positive inflammatory exposure prior to HI injury does not negate the neuroprotective effect of HT in severe brain injury.