Immunosenescence and Cytomegalovirus: where do we stand after a decade?

A more difficult question to answer is whether age-associated immune changes believed to contribute to the immunodeficiency of ageing are likewise caused by CMV. This seems a priori unlikely because CMV infection is very widespread, yet very elderly CMV-seropositive individuals are easily identified (although formally it is impossible to exclude that they were only recently infected). Thus, an important question becomes whether the CMV-sensitive altered immune cell subset distributions in the elderly, particularly the decreased numbers and percentages of na？ve T cells and increased late-stage differentiated memory T cells (commonly used as biomarkers to assess "immunosenescence"), are in reality an epiphenomenon irrelevant for predicting health status in very old people. In order to discuss these and other relevant questions, a workshop was organised at the Center for Medical Research in Tübingen in December 2009, attended by virologists, immunologists and geriatricians whose brief was to elucidate the important questions on this issue, and make a first attempt at answering them. Questions focussed on the mechanisms of age-associated alterations to immunity, how far these alterations can be ascribed to CMV infection, and whether and to which degree other sources of "chronic antigenic stress" (eg. cancer, parasite infection, amyloid-β, other viruses, autoantigens) may be additive with CMV in this respect. The crucial question of how far we are able to assess the clinical impact of these factors using epidemiological approaches, longitudinal and cross-sectional studies was to be discussed, as well as the mechanisms responsible for such outcomes. The clinical relevance and predictive value of establishing immune signatures influenced by the above factors, in terms of longevity, cause of death, responsiveness to vaccination etc. must be established in order to make informed plans for translational work. Bearing this in mind, as well as the thus-far equivocal data on ass