The intracarotid amobarbital sodium, or Wada, test has been used to localize speech and memory function prior to surgical treatment of temporal lobe seizures. The authors mixed technetium-99m hexamethyl-propyleneamine oxime (HMPAO) with amobarbital sodium and injected the mixture in 25 patients with epilepsy. Single photon emission computed tomography (SPECT) of the brain was then performed to determine intracerebral distribution of the amobarbital sodium. Results of SPECT were compared with those of conventional and digital subtraction angiography (DSA). The distribution of Tc-99m HMPAO and, presumably, amobarbital sodium varied from patient to patient. SPECT revealed a statistically different distribution from that predicted with conventional angiography. The distribution also often differed from that of DSA, although the difference was not significant. SPECT revealed infrequent delivery to mesial temporal lobe structures. This emphasizes the need for caution in the use of the intracarotid amobarbital sodium test to predict the outcome of removal of these areas.

Regional cerebral blood flow was measured with single photon emission computed tomography in 10 obsessive-compulsive patients and eight comparison subjects. The patients had a significantly higher ratio of medial-frontal to whole cortex blood flow; this was unrelated to symptom severity but was correlated negatively with anxiety. No differences in orbital-frontal blood flow were found.

The authors administered 48 mg of intravenous cocaine or placebo to eight abstinent cocaine users in a double-blind, crossover design and examined blood flow using single photon emission computed tomography. Cocaine produced significant decreases in frontal cortical and basal ganglia blood flow; these latter correlated negatively with increases in self-ratings of "rush" and "high." The authors conclude that these local effects are compatible with dopaminergic system involvement.

We report a case of massive cerebrospinal fluid (CSF) leakage where the tracer injected intra-thecally for radionuclide cisternography was later visualized in the bowel as well as the nasopharynx. We discuss the potential implications of this finding in patients with CSF leaks. A brief review of the diagnosis of CSF leaks is included.

SUMMARY Positron emission tomography (PET) has been used extensively as a research tool in the investigation of human physiology and pathology for over a decade. By labelling suitable compounds (for example, glucose, amino acids, ammonia, DOPAor drugs) with positron emitting isotopes which are then administered in tracer amounts, the blood flow, metabolism and even the cell receptor or neurotransmitter distributions may be assessed in vivo. Advances in PET technology and experience now make PET a powerful clinical diagnostic tool, enabling investigation of disease at a molecular level, even in the absence of anatomical abnormalities apparent on computerised tomography (CT) or magnetic resonance imaging (MRI). Clinical PET is already utilised in the management of patients with epilepsy, cerebrovascular and cardiovascular disease, dementia and a wide variety of oncological applications. PET will become more widely available shortly in the UK, with the opening of centres such as the Guy's and St Thomas's Clinical PET Centre in 1992. It will therefore become increasingly important that clinicians are aware of those specific areas in which PET may be the investigation of choice to optimise patient diagnosis, treatment and/or follow‐up. This review will endeavour to explain briefly the principles of the PET technique, and then outline those areas where PET has already had an impact on patient management in comparison with the more widely available diagnostic tests, finally outlining promising areas where PET may become more clinically useful in the future.