Amplitude-integrated EEG: a useful electrophysiological tool in prognosis of asphyctic babies

Background: Perinatal asphyxia remains one of the most important causes for neonatal mortality and morbidities such as cerebral palsy, epilepsy and cognitive impairment. Amplitude-integrated EEG (cerebral function monitoring CFM) has become a standard diagnostic tool in asphyctic babies providing an easy bedside method for early brain injury assessment and for prognosis. Methods: We reviewed all cases of neonatal asphyxia (pH<7.0 and/or Apgar <3at 5' and or hypoxic-ischemic encephalopathy, HIE, according to Sarnat) between 32–42 weeks gestation in our Level III Neonatology Unit from December 2005 to December 2007 and assessed patterns and recovery of moderate and severely abnormal initial aEEG patterns in comparison to standard EEG, early and late MR-imaging, Sarnat classification and neurological follow up (up to 18 months). There was no hypothermia protocol used at that time. Each diagnostic tool was reviewed independently by 5 different experts. Results: 56 patients presented over the study period with the inclusion criteria mentioned above, 9 patients had to be excluded presenting other primary pathologies mainly malfor-mations and/or syndromes. 47 were finally taken in account, 34% were inborn, with gestational age ranging from 32 to 42 weeks of gestation, mean GA was 38±2.3wk. Mortality was 17%, all with Sarnat III HIE, male: female ratio was 26:21. Clinical evaluation showed no encephalopathy at all for 44% (n=21) of patients, Sarnat I for 17%, Sarnat II for 21% and Sarnat III for 17%. CFM was used in 73% (n=34) of all patients in the first 6h of life and monitored for 12h to 72h range, 96% CFM for the Sarnat II and III patients. 20/34 CFM patterns (58%) were abnormal in the first 6h and (n=14) 41% remained abnormal over 24 to 48h with a poor outcome (death for n=8 or major handicap n=4 and n=2 for abnormal development at 6 months). Correlation with standard EEG was high (96%; 24/25). In the Sarnat II group, CFM severely abnormal pattern and persistence of the pattern over 48h was significantly correlated with major handicap at 18 months follow up. Conclusions:our data confirms the data from the literature with a good sensitivity and specificity of the CFM specially for Sarnat II and III patients. It is an easy and bedside technique for assessing rapidly severity of brain injury, diagnosing seizures before standard EEG and permitting therefore adequate therapy. It will be a useful tool, specially in combination with other neurodiagnostic tools. In view of the cooling protocols which are discussed as neuroprotective therapy, it will allow selection of the most targeted patients which are according to the literature the Sarnat II patients.