Itraconazole should prove to be a useful replacement for ketoconazole on hospital formularies. This recommendation is based on itraconazole's greater apparent safety and efficacy. Reevaluation of this agent will be necessary upon the release of newer imidazoles and triazoles.

SUMMARY Two patients with subacute arsenic neuropathy are described and the results of serial conduction velocity determinations from the very early stages of the illness to partial recovery are reported. Sensory and motor deficits were maximal within four weeks of the estimated time of exposure. Recovery was slow, with only partial improvement of the neurological deficits two years after onset of the illness. Progressive slowing of motor conduction velocity was observed in the first three months followed by a gradual increase in velocity thereafter. The possible mechanisms underlying the temporal profile of the electrophysiological changes are considered. The need for initiating chelation therapy before the development of the neuropathy is emphasised. Case reportsCase I A 52-year-old man was in good health until two weeks before admission when he developed malaise, fever, mild non-productive cough and occipital headaches. These symptoms resolved within one week and were followed by numbness and tingling sensation in the toes and fingers and weakness of the legs. The patient had used alcohol in excess for several years, but otherwise the past history was unremarkable. On examination he was thin and appeared ill. The blood pressure was 98/50 mm Hg and the pulse was 84 per minute and regular. A diffuse confluent maculopapular scaly rash was present over the trunk, shoulders and anterior aspect of the thighs. The liver edge was palpable 2 cm below the right costal margin. There was moderate weakness of distal muscles of both upper and lower extremities. Touch and pain sensation were impaired in a glove-stocking distribution and proprioception and pallaesthesia were absent in toes and ankles. Muscle stretch reflexes were normal in the arms, but were absent in the legs; the plantar responses were flexor. Gait was broad-based and ataxic, tandem walking could not be performed and Romberg sign was positive.The following laboratory abnormalities were found:

No abstract