During the past decade, our appreciation of the original experiments with myo-inositol supplementation in diabetic rats has greatly expanded. The effects of myo-inositol on nerve conduction are now explained by concepts that were largely unappreciated in 1976, including the fundamental role of phosphoinositide metabolism in cell regulation and the importance of the activity of sodium-potassium-ATPase in nerve conduction. Aldose reductase inhibitors firmly link defects in myo-inositol metabolism to activation of the polyol pathway in diabetes; the resulting "sorbitol-myo-inositol hypothesis" has been extended from its application to the lenses and peripheral nerves to most of the tissues involved with diabetic complications. These biochemical mechanisms provide a new framework within which to explore the complex interactions between hyperglycemia and the vascular, genetic, and environmental variables in the pathogenesis of diabetic complications. It is anticipated that these endeavors will result in the appearance of new classes of therapeutic agents, the first of which--the aldose reductase inhibitors--has emerged from the laboratory and is now undergoing extensive clinical testing. These efforts are very likely to result in the appearance of new treatment methods that may dramatically lighten the burden of chronic complications in patients with diabetes.

We review the rationale for, and progress in, the development of antiviral therapy for the acquired immunodeficiency syndrome (AIDS). A consideration of the replicative cycle of the human immunodeficiency virus (HIV) can lead to the identification of several steps that represent potential targets for antiretroviral therapy, and several substances that can inhibit the replication of HIV in vitro have already been identified. The 2',3'-dideoxynucleosides are a class of nucleoside analogues in which the 3'-hydroxy group is modified so that it cannot form phosphodiester linkages for nucleic acid chains. Some are potent in vitro inhibitors of HIV replication, possibly acting as chain terminators of viral DNA during reverse transcription. One of these dideoxynucleoside analogues, 3'-azido-2',3'-dideoxythymidine (AZT), has now been administered for up to 18 months to patients with AIDS. The drug has been shown to improve immunologic function, to reverse, at least partially, HIV-induced neurologic dysfunction in some patients, and to improve certain other clinical abnormalities associated with AIDS. The principal toxic effect associated with AZT is dose-dependent suppression of bone marrow, resulting particularly in anemia and leukopenia; however, most patients in whom this toxic effect occurs can subsequently tolerate a lower dose of the drug. The demonstration that AZT can be beneficial in patients with AIDS has removed some of the uncertainty about the rationale for an antiretroviral intervention in HIV infection.