METH ODSbolic perturbat ion imp osed by hypoxia, by seizures during normoxia, and by seizures during hypoxia. The se are topi cs of major clinical relevance since seizures in the neon ate are frequ entl y manifested by apnea and cyanosis (2).In measuring cerebral metabolic state , we employed high resolution 31P_NMRspectroscopy, a powerful technique for noninvasi ve, sequential measurement of brain energy change s in vivo (6-9). These in vivo measur ements were corro borated by in vitro determ inati on of brain metabolite concentratio ns.
Systemic changes. Experiments were carri ed out in neonatal(1-to 10-day old) mongrel dogs as previously described (8). Halothane (0.5-3.0%) in oxygen was administered to allow tracheostom y. Haloth ane was then discontinued and the animals were paral yzed with pan curonium (0.2 ml admi nistered intraperitoneally) and mechanically ventilated with a gas mixture of 30% O2/7 0% N20 . In add ition to the use of nitrou s oxide for analgesia, all sites of incision were infiltrated with local anesthetic (xylocaine, 1%). The femoral art eries were catheterized to continuously monitor blood pressure and to anaerobically withdraw sampl es for determination of blood gases. The femoral vein was cath eterized to allow infusion of dru gs. Wounds were periodically infiltrated with xylocaine.Carbon dioxide tensions were normalized (PC0 2, 30-40 mm Hg) by adju sting minute ventilation. Thereafter, ventilator settings were not changed. Animals who had abno rmal acid-base statu s (base deficit> 5 mm ol HC0 3 / liter) were not entered into the study. Following a 30-min normalization period, animals were randomly assigned to one of four stud y groups: control group , hypoxia group, seizure-normox ia or seizure-hypoxia group. Th e durat ion of hypoxia, seizure, or seizure and hypoxia was 45 min. Animals in the hypoxia group or seizure-hypoxia group were ventilated with a gas mixtu re containing 8% O 2 / 92% N2• Animals in the contro l group or seizure-norm oxia group were ventilated with a gas mixtu re consisting of 30% O 2 / 70% N20. Seizures were induced in the seizure-normo xia and seizurehypoxia group s with an intravenous infusion of the 'Y amino but yric acid blocking agent , bicuc ulline (Sigma Co.), 2 mg/kg intravenou sly. Animals in the contro l and hypoxia groups received an equ ivalent volume of 0.9% saline intravenously.Blood gases were measured with a Radiom eter microblood gas analyzer. Plasma glucose levels were measured using a Beckm an G lucose Analyzer II; plasma lactate levels were measured spectrophotometrically using standard enzymatic anal yses (10). Plasma catech olamine concentratio ns were determined by radioenzymati c analysis (1 1).Cerebral phys iology. Th e EEG was continuo usly recorded with scalp electrodes and a Beckman polygraph ic recorder. An auto-58 1 Abbreviations 3IP_NM R, phosphorus-31 nuclear magnetic resonance CBF, cerebral blood flow PCr, phosphocreatine ABSTRACf. Prolonged neonatal seizures are often accompanied or exacerbated by hypoxemia. To determine the e...