CLL Is Associated with Development of Subclinical Cytomegalovirus Viraemia and Accumulation of Large Populations of “exhausted” CMV-Specific CD4+ T Cells

Abstract

Chronic Lymphocytic Leukaemia (CLL) is associated with T cell dysfunction and increased expression of markers of T cell exhaustion. Cytomegalovirus is a common herpesvirus infection and is associated with development of accelerated immune senescence in older adults. Within patients with CLL, CMV leads to marked expansion of virus-specific CD4 and CD8 T cells and development of oligoclonal T cell populations. However these features are not known to be associated with clinical symptoms and CMV viral load remains essentially undetectable by conventional qPCR. In this study we used digital PCR to determine CMV viral load in patients with CLL and correlated this with the magnitude and phenotype of the CMV-specific T cell immune response. In particular, we utilised HLA class II tetramers for the first time, in order to assess the contribution of CMV-specific T cell populations to the profile of T cell exhaustion seen in this disease. 68 CMV-seropositive CLL patients and 19 age-matched healthy donors (HD) were recruited for study. All patients were either untreated or had not received chemotherapy for at least 6 months. CMV viral load was determined within purified monocyte populations using a digital droplet PCR method. Viral load per monocyte was increased in CLL patients compared to HD (p=0.04), with the highest viral loads detected in stage C patients. In order to investigate the CMV-specific immune response, nine HLA class I tetramers were generated, containing viral epitopes from pp65, pp50 and IE-1, and two class II tetramers were also available, with epitopes from glycoprotein B and pp65. CMV-specific CD4 T cell responses were increased in CLL patients (n=14) compared to HD (n=11) (4.1 % vs. 0.9 %; p=0.049). Remarkably, in one patient more than half (50.9%) of all CD4 T cells were directed against a single CMV epitope. CD4 CMV-specific T cells were nearly always effector memory in phenotype (78%), somewhat in contrast to CD8 populations where the memory phenotype is split between effector memory (CCR7-, CD45RA-) and terminally differentiated memory cells (CCR7-, CD45RA+). PD1 is an important marker of T cell exhaustion and expression was increased on both CD8 T cells (16.9% Vs 9.6%; p=0.003); and CD4 T cells (16.2 % Vs 8.7 %; p=0.0007) in CLL patients compared to HD. Interestingly, PD1 expression was much higher on CMV-specific CD4 T cells compared to the total CD4 T cell repertoire (50.6% Vs 21%; p=0.01), whereas the opposite profile was observed in relation to CD8 populations. CMV-specific CD4 T cells demonstrated a Th1 cytotoxic phenotype with production of TNF-alpha, IFN-y and Granzyme B. Production of IFN-y and TNF-alpha was reduced in PD1+ populations, consistent with an exhausted phenotype. No CD25+FoxP3 + regulatory T cells were observed within the CMV-specific CD4 T cell population which is of interest given the well documented expansion of this population in patients with CLL. In order to investigate the replicative history of CMV-specific T cells we investigated the telomere length of antigen-specific cells using single cell telomere length analysis. CMV-specific cells had markedly shortened telomere lengths compared to background CD4 and CD8 T cells, with the mean difference of 0.911 kb (maximum 1.836 kb) indicating that they have undergone extensive proliferation in vivo. This work is the first to use digital PCR to measure the subclinical CMV load within peripheral blood and also to examine CMV-specific CD4 T cells using HLA class II tetramers. Our results indicate that immune control of CMV viral load is impaired during the clinical progression of CLL and that a proportion of virus-specific CD4 T cells show signs of exhaustion. These data may reflect ‘cross presentation’ of viral protein by B-CLL tumour cells, which are known to have poor capacity for antigen presentation. The potential clinical importance of these observations is now being addressed. Disclosures No relevant conflicts of interest to declare.