Abstracts

Rationale: Morphine is used to control shivering in neonates undergoing therapeutic hypothermia (TH) for hypoxic-ischemic encephalopathy. Side effects of opiates such as hyperalgesia, myoclonus and seizures, have been reported in animals as well as in human neonates. Elevated serum morphine concentrations have been demonstrated in TH-treated newborns compared to normothermic infants at similar morphine dosage. We aimed to investigate the presence of morphine-associated EEG and behavioral changes in newborns during TH. Methods: Continuous video-EEGs of 115 consecutive term newborns treated with TH between November 2007 and November 2011, who received morphine per protocol, were retrospectively reviewed. Infants with abnormal MRI, subdural hemorrhage or large scalp edema, AEDs prior to onset of video-EEG were excluded. Results: 45 infants were included. Video-EEG monitoring was initiated at a mean of 9.5 hours of life for 89 hours. 13/45 infants (30%) presented with symptoms ascribable to morphine: repetitive, intermittent, spontaneous myoclonic jerks(MJ) involving upper and lower extremities occurring during sleep and awake states, during several hours and despite escalating doses of morphine. All infants showed hyperalertness, hyperexcitability, limb hypertonia and hyperreflexia. MJ were observed 1 to 11 hours (mean 7 hours) after beginning of morphine administration, which included continuous infusion and boluses. EEG was normal or mildly discontinuous in 8 infants, low voltage in 2 infants, and showed a severe burst-suppression pattern with interburst intervals from 25 to 40 seconds in 3 infants. MJ did not have a clear EEG counterpart or were associated with a diffuse broad slow wave. However, in 3 patients with abnormal EEG (2 with burst-suppression and 1 with low voltage) MJ progressed into true ictal events characterized by prolonged focal or bilateral slow discharges of high voltage sharp waves lasting 2 to 30 minutes associated with clonic activity time-locked with the EEG discharges. Six patients were treated with lorazepam and/or Phenobarbital. Mean duration of symptoms was 12 hours (2 to 24 hours) with shorter duration in treated patients. When abnormal, EEG completely normalized within 12 hours and it was normal at 24 hours of life. Seizures never recurred. While the rate of morphine infusion did not significantly differed between symptomatic and asymptomatic infants, (20.4 microgram/kg/h vs 17.8 microgram/kg/h p =0.054), there was a significant difference in the amount of morphine received as boluses (0.22 mg/kg vs 0.12 mg/kg p=0.0051). Conclusions: Our findings suggest that morphine administration in newborns undergoing hypothermia treatment may paradoxically be associated to neuronal hyperexcitability and/or profound yet reversible EEG alterations. The spontaneous resolution of symptoms in some infants and the EEG rapid normalization could be explained by previous data showing that, differently from children and adults, morphine plasma concentrations in newborns tend to decrease with time despite unchanged dosage.