Response to 'Influence of Age and HLA Alleles on the CMV-Specific Cell-Mediated Immunity Among CMV-Seropositive Kidney Transplant Candidates'

摘要

To the Editor: We would like to thank Fernandez-Ruiz et al (1) for their interest in our manuscript (2) and welcome the opportunity to address their comments on the factors influencing CMV-specific T cell response in transplant candidates. Whereas we show an association between HLA-A1 and/or HLA-A2 alleles and age older than 50 with CMV-specific CD8+ T cell response, their results did not show this association. Although the studies are apparently contradictory, they have considerable methodological differences which impact the results and their interpretation. Regarding the data analysis, we used logistic regression analysis and the results were given in terms of odds ratios. Thus, we reported that HLA-A1 and/or HLA-A2 candidates and those older than 50 have a higher probability of having CMV-specific CD8+T cell response (IFNg >0.2 ILJ/mL) than non-HLA-A1on-HLA-A2 candidates and those younger than 50. Fernandez-Ruiz and coworkers, however, analyze the association of HLA alleles and age with the magnitude of CMV-specific T cell response (absolute count of IFNg-producing T cells) using linear regression analysis. They also argue that the intracellular staining (ICS) method is the reference procedure for assessing CMV-specific T cell response, whereas the QuantiFERON-CMV (QF-CMV) assay used in our study introduces a bias because it primes the response against pp65 and IE-1 restricted by HLA-A1 and HLA-A2. However, the QF-CMV assay also includes pp50, gB, IE-2, and pp28 peptides. In fact, the immunodo-minant peptide restricted by HLA-A1 (VTEHDTLLY) derives from pp50, and not from pp65 or IE-1 (3-4). In contrast, they used peptide libraries of pp65 and IE-1 and not from other CMV proteins. Thus, each method uses different peptides for stimulation with its pros and cons. In addition, the QF-CMV assay has a well-established cut-off that classifies individuals as reactive (IFNg >0.2 ILJ/mL) and nonreactive in contrast to the ICS method, which does not have a standardized cut-off to define nonresponding individuals.