Magnetic resonance and near infrared spectroscopy for investigation of perinatal hypoxic-ischaemic brain injury.

Hypoxic-ischaemic injury to the brain is an important cause of perinatal death and seems to be the commonest cause of permanent neurodevelopmental disability in newborn infants who survive after intensive care. If this type of brain injury is to be prevented and treatment put on a rational basis, non-invasive methods are required for defining its mechanisms. This review has considered two such methods: magnetic resonance spectroscopy and near infrared spectroscopy. Magnetic resonance spectroscopy is used to measure, in brain tissue, the concentrations of the 'high energy' phosphorus metabolites that are dependent for their synthesis on the processes of oxidative phosphorylation. Intracellular pH can also be measured. Normal maturational changes in the brain have been defined and abnormalities detected in a range of conditions where hypoxic-ischaemic injury was suspected to have occurred. In laboratory animals the acute effects of curtailment of oxygen supply to the brain ('primary' energy failure) have been observed, and the effects of two commonly used treatments, infusions of sodium bicarbonate and glucose, have been tested. After resuscitation of newborn infants from severe intrapartum asphyxia, a latent period has often been noted before energy failure became detectable. This 'secondary' energy failure is due to a variety of damaging reactions initiated by the acute hypoxicischaemic episode and reperfusion of the brain. It is possible that in the future irreversible injury to brain cells following the episode may be prevented or ameliorated by the prompt use of cerebroprotective agents. The extent of abnormalities detected by magnetic resonance spectroscopy has prognostic implications: evidence of severe energy failure in the first days of life was regularly associated with subsequent death or with severe neurodevelopmental impairments. Many technical developments in magnetic resonance spectroscopy are under way, particularly employing proton (1H) spectroscopy, which will allow the intracerebral concentrations of a wide range of metabolites, including neurotransmitters, to be measured. The combination of spectroscopy with magnetic resonance imaging will permit quantitative data to be obtained from selected volumes within the brain. Near infrared spectroscopy is used to make observations at the cotside of the intracerebral concentrations of the chromophores oxyhaemoglobin, deoxyhaemoglobin, and oxidised cytochrome aa3, and it therefore provides information complementary to that obtained by magnetic resonance spectroscopy. Measurements can also be made of cerebral blood flow, cerebral blood volume, and other haemodynamic indices; in addition, the rea