A ray of hope for a foreseeable therapy was provided recently by vaccination experiments in transgenic mice. Specifically, mice expressing mutant forms of human APP efficiently cleaved into A-beta _1-40/42 -peptides were shown to develop Alzheimer disease-like pathology. To the great surprise of the field, both plaque burden and dementia could be reduced in these mice by active immunization for antibody induction using A-beta _1-40/42 formulated in QS-21 as a vaccine (AN1792). Although the mechanism of plaque removal remains unclear, vaccination has proved to be efficient in a variety of APP-transgenic mouse models. Progress towards the clinic was rapid and ELAN pharmaceuticals, the front runner in the field, initiated a Phase II trial designed to demonstrate efficacy of the therapy also in humans. However, it was this particular study that brought hopes and progress to an abrupt halt, because a few treated patients developed signs of aseptic encephalitis/meningitis following the second administration of the vaccine. Although little information has been released publicly, the disease-symptoms were consistent with induction of autoimmunity, an inherent risk and concern when vaccinating against self-antigens such as A-beta _1-40/42 . Hence, this trial has been stopped.

A detailed understanding of the immune responses induced in the vaccinated individuals is now critical for further development of vaccination strategies that avoid these severe (albeit treatable) side-effects. It is of particular importance to understand the origin of the vaccine-induced disease, namely whether it was non-specifically caused by the adjuvants used, or the antibodies were precipitating inflammation or whether A-beta _1-40/42 -specific T cells were responsible for the side-effects.

Thankfully, the study by the group of Nitsch now offers the first insights into these questions. Of the 24 patients vaccinated in the study center in Zürich, almost all raised a significant antibody response. In addition, the antibodies recognized plaques of transgenic mice and patients, a requirement for plaque-removal in the mouse model. However, aseptic encephalitis/meningitis developed in one individual. Intriguingly, this particular patient had only raised a moderate antibody response. This observation indicates that specific antibodies alone might not be sufficient to cause disease in Alzheimer patients. Furthermore, the results are compatible with the view that T cells, rather than antibodies, are the major source of inflammation. This hypothesis is consistent with the fact that most, if not all, known inflammatory responses in the brain are caused by T cells rather than antibodies.

Although it is obviously impossible to draw firm conclusions from a single patient, the data nevertheless suggest that vaccines against Alzheimer disease should aim at circumventing the induction of inflammatory T cell responses. Indeed, avoiding the use of adjuvants and restricting the size of the epitopes used for vaccination could offer valuable solutions for the generation of a viable vaccine against one of the most devastating diseases afflicting mankind.