CMV infected or not CMV infected: That is the question


Full correspondence: Prof. Vincent C. Emery, Department of Microbial and Cellular Sciences, University of Surrey, Guildford, Surrey GU2 7XH, UK

Fax: +44-0-1483-689043


See accompanying article by Sester et al.


Cytomegalovirus (CMV) infection is widespread in the human population. Normally, in adolescents and adults, prior exposure to CMV can readily be determined by IgG assays. However, in individuals where antibody production is impaired, such as patients with common variable immunodeficiency disease or in the very young where maternal antibodies are present, diagnosis of CMV infection is problematic using such assays. In this issue of the European Journal of Immunology, a study by Sester and colleagues [Eur J Immunol 2013. 43: 1099-1108] using CD4+ T-cell immunity as a marker of infection clearly differentiates young children with prior exposure to CMV from those who only have passive maternal antibody. This information will quickly find application in the pretransplant screening of young children for CMV infection and help with the stratification of these children to identify those who are truly CMV negative and are therefore at risk for future CMV infection and disease if receiving an organ from a CMV-positive donor as discussed in this Commentary.

Human cytomegalovirus (CMV) remains an important pathogen in the setting of transplantation [1, 2]. CMV replication is highly dynamic [3] and has the ability to elicit a very strong and multi-faceted B-cell and T-cell response in the human host [4], as well as having an impact on the NK-cell compartment [5]. Consequently, it is not unusual in healthy humans to find up to 10% of T cells recognising single CMV antigens or, in the case of some CD8+ T cells, specific class I HLA-restricted epitopes [4, 6]. The B-cell and T-cell responses against CMV are directed against the entire CMV proteome [4]. However, in the case of neutralising antibodies, those directed against glycoproteins B, M/N, H and the recently described pentameric complex comprising gH, gL, UL128, UL130 and UL131 are particularly noteworthy [7, 8] with the majority of immunoglobulins deployed for adjunct therapy in renal transplantation and for CMV pneumonitis containing antibodies against the pentameric complex [9]. In the context of the CD4+ and CD8+ T-cell response, specific antigens that are targeted include the surface glycoprotein gB(UL55), the pp50 protein (UL44) and pp65 (UL83) and the immediate early 1 protein (UL122) [4].

To assess whether an individual has been infected with CMV, a number of commercial assays based on IgG detection are available. These assays provide a robust way of identifying infected individuals with high specificity and sensitivity. However, in certain circumstances IgG assays can either be unhelpful or not able to provide appropriate information. For example, in the case of patients with common variable immunodeficiency (CVID), an adolescent/early adult-onset disease that results in the inability of B cells to produce secretory antibodies, the patients do not have sufficient circulating antibodies to provide meaningful results with respect to IgG assays for determining CMV status. In addition, such patients would normally receive immunoglobulin replacement therapy to guard against significant infections and so passive antibody may be detected in a conventional IgG assay and therefore not be reflective of the infection state of the CVID patient for CMV. Previous work in our group has addressed this issue through the use of T-cell responses, especially CD4+ T-cell responses against a lysate of CMV-infected cells [10-12]. Using this approach we were able to show that a number of CVID patients had previously been infected with CMV and that these patients showed an enhanced frequency of inflammatory diseases affecting the gut, lungs and kidney, whereas patients who had no evidence of CMV infection as determined by T-cell reactivity did not express these pathological inflammatory complications [11, 12].

Another group of individuals in whom the presence of antibodies may not be truly reflective of their infectious status is young children. In this case, immunoglobulins present in the maternal circulation will pass through the placenta and reside in the newborn for up to 18 months postpartum. Clearly, in order to determine whether these children have been infected with CMV requires an alternative approach. The direct detection of CMV by PCR in these infants is unlikely to be informative since replication may have occurred only transiently during their initial infection and so CMV would be unlikely to persist in blood or urine [13]. In this issue of the European Journal of Immunology, an article by Sester et al. [14] builds upon our knowledge of the T-cell response against CMV to determine whether the deployment of CD4+ T-cell assays in this population of young children can provide important data on their prior exposure to CMV. Initially, the authors show that CMV-specific CD4+ T cells can be readily detected in all CMV-seropositive mothers and children above the age of 18 months [13]. In the children, there was a 100% agreement between the serology results and the CD4+ T-cell results. In the mothers, the results had a high kappa value at 0.93 indicating an excellent correlation between the T-cell results and serology but two individuals were positive by T-cell responses although negative by serology. This observation is consistent with other studies indicating that occasionally patients appear to be CMV seronegative but have evidence of prior exposure through their T-cell profile [15]. The authors went on to apply this assay to determine the presence of CMV-specific T cells in children younger than 18 months of age, showing that none of the 17 infants analysed who were IgG negative had evidence of CMV-specific T-cell immunity [14]. In 13 infants who were identified as CMV IgG positive, only eight (61.5%) had detectable CMV-specific CD4+ T cells.

In recent years, there has been increasing interest in the ability of CD4+ and CD8+ T cells to produce multiple cytokines in response to stimulation [16]. In particular, the co-production of IFN-γ, IL-2 and TNF-α is regarded as a surrogate for the quality of the T-cell response. In the present study [14], the authors investigated IFN-γ, IL-2 and TNF-α production in the CMV-specific CD4+ T cells to reveal that IFN-γ was the most reliable readout for cytokine production for CMV-specific T-cell immunity in all the age groups studied (Fig. 1). In addition, Sester et al. [14] show that there was an increase in the frequency of cytokine-producing CMV-specific CD4+ T cells with increasing age (although the proportion of IFN-γ-secreting cells was similar in all age groups). Interestingly, the proportion of triple-cytokine-producing T cells appeared to be lower in infants compared with that in adults. The observation that CMV infection changes the immune profile of an individual was evidenced through analysis of polyclonal IFN-γ, IL-2 and TNF-α, IL-17 and IL-4 responses following SEB stimulation. CD4+ T cells from infants with pre-existing CMV immunity had an elevated frequency of SEB-reactive T cells secreting IFN-γ, TNF-α, IL-2 and IL-17. In contrast, infants who had not yet experienced CMV showed a T-cell profile with a high capacity for TNF-α and IL-2 production but relatively weak IFN-γ production.

Figure 1.

Concept behind the use of CMV-specific CD4+ T-cell responses to determine the infection status of an infant less than 18 months of age. The blood vessel (brown structure) is shown containing anti-CMV antibodies, which are of either maternal IgG (blue) or fetal (red) origin, and CD4+ T cells (green oval structures); black cross structures represent CMV. (A) In blood isolated from CMV-infected infants, the presence of maternal (blue) and infant (red) CMV-specific antibodies leads to a positive IgG assay. In addition, the presence of CMV-specific T cells leads to the production of cytokines (yellow, blue, and orange dots representing IFN-γ, IL-2, and TNF-α, respectively) after stimulation. (B) An infant blood sample that contains maternal antibodies (blue) but no CMV-specific T cells, as occurs in uninfected infants, leads to an IgG-positive result but a negative T-cell assay.

The value of the study by Sester et al. [14] can be directly translated into the setting of solid organ and stem cell transplantation in young infants. Currently, the presence of maternal antibodies serves to confuse the picture with respect to risk stratification for CMV — does a positive CMV IgG result in a child younger than 18 months really mean past CMV infection, or not? With the advent of a robust assay based on CD4+ T-cell reactivity, we move a step closer to providing a solution to this problem. These T-cell assays have now become routine and can be performed in approximately 1 day, although they still require some standardisation. Consequently, this approach may become widely applicable in infant and child transplant centres whereby the true infection status for CMV of an infant younger than 18 months can be determined and their post-transplant risk of acquiring CMV infection from the donor assessed appropriately and managed satisfactorily.


The author is grateful to his colleagues who have been involved in the work on CVID.

Conflict of interest

The author has declared no conflict of interest.




common variable immunodeficiency