Sara van den Berg

12 Chapter 1 in the host, that infects the majority of the population world-wide (40–100% of adults) and establishes latency in myeloid cells and epithelial cells [15]. Latent CMV infection was identified as part of the so-called ‘immune risk profile’ for all-cause mortality [13, 14]. Infection with CMV rarely causes clinical symptoms in healthy individuals, and viral reactivation from latency is thought to occur regularly during life. Therefore, CMV triggers frequent activity of the immune system leading to large expansions of differentiated T-cells. When the immune system is severely compromised, for example due to the use of immunosuppressive drugs after organ transplantation, CMV infection can no longer be controlled. The resulting reactivation of CMV leads to virus production, viral spread and severe CMV disease, and sometimes even mortality. In healthy individuals, controlling the attempts of the virus to reactivate leads to large expansions of differentiated T-cells which require large contribution of the immune system’s resources. Importantly, these changes in the T-cell profile, for example a decreased CD4 + /CD8 + T-cell ratio and increased numbers of late differentiated memory T-cells, were independently associated with the increased mortality risk [16]. The theory of life-long CMV infection gradually exhausting the immune system and accelerating immunosenescence was born. The CMV-enhanced immunosenescence theory relies on the hallmarks of the CMV-specific immune response and is supported by several lines of evidence. Namely 1) CMV establishes similar T-cell changes as ageing, 2) first studies suggested a negative association between CMV and the immune response to influenza and 3) epidemiological evidence links CMV to increased mortality ( Figure 1 ). The CMV-specific immune response with age and these three lines of evidence became subject to further investigation in subsequent years. THE CMV-SPECIFIC IMMUNE RESPONSE Seroprevalence of CMV increases gradually with age. A gradual pattern is not typical for a contagious pathogen in a primary infection phase, in which a quicker saturation of the whole population is expected. The gradual increase with age requires periodically infectious reactivation events, slowly affecting more and more individuals [19]. Indeed, CMV, just like its ‘sibling’ herpes-simplex virus-1 causing episodes of cold sores, is thought to reactivate periodically during life. Indeed, viral load in plasma and urine are sporadically detected in healthy adults [20], especially in older adults [21, 22]. CMV thus seems to establish a rather ‘dynamic’ state of latency instead of true silent latency. This characteristic of CMV is thought to be key in the immune response to CMV. Despite the overall decline in function of the immune system with age, CMV specific antibody levels correlate positively with age in large-scale cross-sectional studies [21, 30, 31]. Periodic CMV reactivation is thought to cause increases in the CMV-specific immune response over time [19, 21]. CMV-specific antibody levels are therefore often interpreted as a surrogate marker for experienced CMV reactivation, as well as to identify CMV-seropositive elderly with poor control of the virus who are at increased risk of adverse clinical outcomes [17, 32-37]. In line with this, CMV-specific antibody levels are positively associated with the frequency of occasional shedding of CMV in urine and saliva in older adults [21]. However, there is little