Petri Rahkonen

Cortical somatosensory processing measured by magnetoencephalography predicts neurodevelopment in extremely low-gestational-age infants

Background:Higher cortical function during sensory processing can be examined by recording specific somatosensory-evoked magnetic fields (SEFs) with magnetoencephalography (MEG). We evaluated whether, in extremely low-gestational-age (ELGA) infants, abnormalities in MEG-recorded SEFs at term age are associated with adverse neurodevelopment at 2 y of corrected age.Methods:SEFs to tactile stimulation of the index finger were recorded at term age in 30 ELGA infants (26.5 ± 1.2 wk, birth weight: 884g ± 181 g). Neurodevelopment was evaluated at 2 y of corrected age. Controls were 11 healthy term infants.Results:In nine of the ELGA infants (30.0%), SEFs were categorized as abnormal on the basis of lack of response from secondary somatosensory cortex (SII). At 2 y, these infants had a significantly worse mean developmental quotient and locomotor subscale on the Griffiths Mental Development Scales than the ELGA infants with normal responses. Mild white matter abnormalities in magnetic resonance imaging at term age were detected in 21% of infants, but these abnormalities were not associated with adverse neurodevelopment.Conclusion:Abnormal SII responses at term predict adverse neuromotor development at 2 y of corrected age. This adverse development may not be foreseen with conventional neuroimaging methods, suggesting a role for evaluating SII responses in the developmental risk assessment of ELGA infants.

Evaluation of somatosensory cortical processing in extremely preterm infants at term with MEG and EEG

OBJECTIVE Prior studies on extremely preterm infants have reported long-term prognostic value of absent secondary somatosensory cortex (SII) responses in magnetoencephalography (MEG) at term. The present work (i) further examines the potential added value of SII responses in neonatal neurological evaluation of preterm infants, and (ii) tests whether SII responses are detectable in routine neonatal electroencephalogram complemented with median nerve stimulation (EEG-SEP). METHODS Altogether 29 infants born <28 gestational weeks underwent MEG, MRI, and neonatal neurological examination at term age, and Hempel neurological examination at 2-years corrected age. Term-age EEG-SEP was available for seven infants. RESULTS While in neonatal neurological examination severely abnormal finding predicted unfavorable outcome in 2/2 infants, outcome was unfavorable also in 3/9 (33%) moderately abnormal and in 5/18 (28%) mildly abnormal/normal infants. Of these eight infants four had unilaterally absent SII responses in MEG, compared with only two of the 24 infants with favorable outcome. Furthermore, SII responses (when present in MEG) were also usually detectable in EEG-SEP. CONCLUSIONS Complementing clinical EEG recording with SEP holds promise for valuable extension of neonatal neurophysiological assessment. SIGNIFICANCE Multimodal study of EEG and sensory evoked responses is informative, safe, and cheap, and it can be readily performed at bedside.

Mother-child interaction is associated with neurocognitive outcome in extremely low gestational age children

Early mother-child interaction is one of the factors suggested to have an impact on neurocognitive development of extremely low gestational age (ELGA) children. Our aim was to examine associations of mother-child interaction with neurocognitive outcome, neurological impairments and neonatal brain injuries in ELGA children. A prospective study of 48 ELGA children, born before 28 gestational weeks (26.3 ± 1.2 weeks, birth weight 876 g ± 194 g), and 16 term controls. Brain MRI was performed at term-equivalent age. At two years of corrected age, the mother-child interaction was assessed in a structured play situation using the Erickson Scales and Mutually Responsive Orientation Scales. Neurocognitive outcome was assessed with Griffiths Mental Developmental Scales (GMDS) and Bayley Scales of Infant and Toddler Development - Third Edition (BSID-III) and with Hempel neurological examination. Among ELGA children, higher quality of dyadic relationship and maternal sensitivity, responsiveness, and supportiveness were associated with positive neurocognitive outcome measured both with GMDS and BSID-III (adjusted p < 0.05). This association remained after adjusting for mothers educational level. Neurological impairments at two years, white matter or gray matter abnormalities in MRI at term-equivalent age, and grade III-IV intraventricular hemorrhage during the neonatal period were not associated with mother-child interaction. This study emphasizes the importance of the quality of mother-child interaction after extremely preterm birth for neurocognitive development. Neonatal brain injury and neurological impairments were not associated with worse parent-child interaction after two years.

Asphyxia, Neurologic Morbidity, and Perinatal Mortality in Early-Term and Postterm Birth.

BACKGROUND AND OBJECTIVES: Neonatal outcomes vary by gestational age. We evaluated the association of early-term, full-term, and postterm birth with asphyxia, neurologic morbidity, and perinatal mortality. METHODS: Our register-based study used retrospective data on 214 465 early-term (37+0–38+6 gestational weeks), 859 827 full-term (39+0–41+6), and 55 189 postterm (≥42+0) live-born singletons during 1989–2008 in Finland. Asphyxia parameters were umbilical cord pH and Apgar score at 1 and 5 minutes. Neurologic morbidity outcome measures were cerebral palsy (CP), epilepsy, intellectual disability, and sensorineural defects diagnosed by the age of 4 years. Newborns with major congenital anomalies were excluded from perinatal deaths. RESULTS: Multivariate analysis showed that, compared with full-term pregnancies, early-term birth increased the risk for low Apgar score (<4) at 1 and 5 minutes (odds ratio 1.03, 95% confidence interval 1.03–1.04 and 1.24, 1.04–1.49, respectively), CP (1.40, 1.27–1.55), epilepsy (1.14, 1.06–1.23), intellectual disability (1.39, 1.27–1.53), sensorineural defects (1.24, 1.17–1.31), and perinatal mortality (2.40, 2.14–2.69), but risk for low umbilical artery pH ≤7.10 was decreased (0.83, 0.79–0.87). Postterm birth increased the risk for low Apgar score (<4) at 1 minute (1.26, 1.26–1.26) and 5 minutes (1.80, 1.43–2.34), low umbilical artery pH ≤7.10 (1.26, 1.19–1.34), and intellectual disability (1.19, 1.00–1.43), whereas risks for CP (1.03, 0.84–1.26), epilepsy (1.00, 0.87–1.15), sensorineural defects (0.96, 0.86–1.07), and perinatal mortality (0.91, 0.69–1.22) were not increased. CONCLUSIONS: Early-term birth was associated with low Apgar score, increased neurologic morbidity, and perinatal mortality. Asphyxia and intellectual disability were more common among postterm births, but general neurologic morbidity and perinatal mortality were not increased.

Atypical sensory processing is common in extremely low gestational age children.

Atypical sensory processing is common in children born extremely prematurely. We investigated sensory processing abilities in extremely low gestational age (ELGA) children and analysed associated neonatal risk factors, neuroanatomical findings and neurodevelopmental outcome.

Comparison of Umbilical Serum Copeptin Relative to Erythropoietin and S100B as Asphyxia Biomarkers at Birth

Background: Birth asphyxia, estimated to account for a million neonatal deaths annually, can cause a wide variety of neurodevelopmental impairments. There is a need to develop new, swift methods to identify those neonates who would benefit from neuroprotective treatments such as hypothermia. Objectives: To examine the utility of cord serum copeptin, a stable byproduct of arginine vasopressin release, as a biomarker of birth asphyxia based on a comparison with 2 biomarkers of hypoxia and brain trauma: erythropoietin and S100B. Methods: The study population consisted of 140 singleton, term neonates: 113 controls and 27 with birth asphyxia (2/3 criteria met: umbilical artery pH <7.10, base excess ≤12 mmol/L, and 5-min Apgar score <7). All deliveries were planned vaginal, but 51 neonates were born by emergency cesarean section. Copeptin, S100B, and erythropoietin levels in umbilical artery samples were measured by immunoassays. Results: Copeptin correlated in the entire study population more strongly with umbilical artery base excess than S100B and erythropoietin, and only copeptin correlated with arterial pH. Furthermore, only copeptin levels were significantly higher in cases of birth asphyxia, and in vaginally born neonates they were found to increase as a function of labor duration. Copeptin was elevated in neonates born via vacuum extraction, whereas erythropoietin levels showed a slight increase after emergency cesarean section. Conclusions: In this study population, S100B and erythropoietin were not valid biomarkers of birth asphyxia. In contrast, our work suggests that copeptin has high potential to become a routinely used biomarker for acute birth asphyxia and neonatal distress.

Amniotic fluid erythropoietin and neonatal outcome in pregnancies complicated by intrauterine growth restriction before 34 gestational weeks

High amniotic fluid erythropoietin concentration reflects chronic fetal hypoxia. Our aim was to study amniotic fluid erythropoietin concentration in relation to neonatal outcome in pregnancies complicated by intrauterine growth restriction.

Low Apgar scores at both one and five minutes are associated with long-term neurological morbidity

This study evaluated the associations between low Apgar scores at one and five minutes and long‐term neurological impairments.

P814: Somatosensory stimulation during routine EEG enables evaluation of secondary somatosensory cortex activation in extremely preterm babies at term

tion and prediction in scalpand intracranial EEG for diagnostic purposes and novel closed-loop treatment options. Many patients with regular, intractable seizures, especially children with epilepsy syndromes and patients with cognitive impairments, have nocturnal seizures. Reliable seizure detection will provide a major step in patient safety, care, quality of life and disease management. In these patients invasive or uncomfortable EEG-sensors do not seem feasible, especially not in the near future. The use of other physiologic signals is investigated for automatic seizure detection. Heart rate and movement can be measured with simple and nonburdensome sensors making them very applicable for long-term monitoring of vulnerable patient groups. Methods: ECG, accelerometer (movement sensor, ACM) and video data was collected with simultaneous clinical video-EEG as a gold standard. All generalized tonic, tonic-clonic, hypermotor and clusters of tonic seizures were included for detection. Seizures were scored for clinical relevance by an expert panel. Four features were extracted: heart rate, waveformlength (from ACM) and normalized energy of frequency for 2-10 Hz (from AVM and video after optical flow analysis). Features were calculated for 2-second epochs and -2, -4, -6, -8, and -10 second values were added as extra features. A non-linear SVM with 5-fold cross-validation for optimisation of parameters was used to identify an optimal combination of thresholds. Results: Data was collected in 92 patients. 20 patients had major seizures and were included for interim analysis (191 hrs and 47 min of recordings). Initial analysis yielded a sensitivity of 61% and a false alarm rate of 1.3 per 24 hours for all included seizure categories. Only considering clinically relevant seizures yielded a sensitivity of 73%. Conclusion: A combination of non-EEG physiological signals movement and heart rate seems feasible for automatic seizure detection in a home setting. Higher sensitivity and lower false alarm rates are needed for algorithms to be implementable in a workable seizure detection device. Currently, survival-analysis is investigated as a tool to improve detection rate and usefulness of HRV-features to increase yield of our algorithms.

P507: Predicting neurodevelopmental outcome of infants born <28 gestational weeks with magnetoencephalography and somatosensory evoked fields

defect edge. The conductivity of the skull defect modulated the strength of change in the MEG & EEG. Dense spatial sampling revealed high spatial frequencies in MEG and EEG due to skull defects that are not detectable with current human helmet-type MEG devices and standard EEG setups. Conclusions: MEG and EEG changes due to a skull defect can be substantial and depend on the defect geometry and the relative orientation and position of the source. MEG forward modelling requires realistic volume conductor head model incorporating skull defects. This effect should be investigated further in humans.