has been shown that the consequence of these regulatory circuits is to divert the effector population from the dangerous Thl state to a protective Th2 direction.

Antibodies directed against cytokines or their receptors may appear to be just one step removed from the generic type of steroid therapy. Nevertheless, applied at the right time, an agent such as anti-TNF (tumor necrosis factor) antiserum can exert a remarkable effect on patients with rheumatoid arthritis. The success of this treatment has led to more efficacious approaches now in clinical trials that counteract the effects of TNF, such as ligands that bind preferentially to the TNFa receptor.

Gene therapy using viral vectors can be used to modulate autoimmune disease, not in an effort to permanently alter the recipient, but rather to provide a localized, potent, short-acting agent—more like a molecular medicine. For example, treatment with an adenovirus bearing IL-12 genes can serve as a protective vaccine for a Th2 mouse, which is unable to raise a strong, protective Thl response against a microorganism. Such animal experiments suggest that in treatments of the near future, combinations of agents provided through genetic alteration of a viral vector will be used as prime modifiers of disease.

Preventive vaccines for autoimmune diseases are also on the near horizon. Individuals who are genetically susceptible to type I diabetes can be identified now, and quite early in the disease course, when diagnostic antibodies appear to antigens of the islets of Langerhans, tolerance-inducing therapy to insulin or glutamic acid decarboxylase, two major candidate diabetogens, can be introduced. In other diseases in which the inciting antigens are known, immune deviation may be started at the very first signs of disease.