Adoptive therapy of malignant diseases with cytokine-induced killer (CIK) cells showed promise in a number of trials; the activation of CIK cells from cancer patients towards their autologous cancer cells still needs to be improved. Here, we generated CIK cells ex vivo from blood lymphocytes of colorectal cancer patients and engineered those cells with a chimeric antigen receptor (CAR) with an antibody-defined specificity for carcinoembryonic antigen (CEA). CIK cells thereby gained a new specificity as defined by the CAR and showed increase in activation towards CEA+ colon carcinoma cells, but less in presence of CEA− cells, indicated by increased secretion of proinflammatory cytokines. Redirected CIK activation was superior by CAR-mediated CD28-CD3ζ than CD3ζ signaling only. CAR-engineered CIK cells from colon carcinoma patients showed improved activation against their autologous, primary carcinoma cells from biopsies resulting in more efficient tumour cell lysis. We assume that adoptive therapy with CAR-modified CIK cells shows improved selectivity in targeting autologous tumour lesions.

No abstract

Summary In chronic myeloid leukaemia (CML), peptides from the fusion region of bcr3/abl2 are likely to play a role in anti‐leukaemic T cell immunity. We investigated whether T cells that recognize bcr/abl fusion peptides could be detected in healthy donors and CML patients. T cell responses against bcr3/abl2 fusion peptides were analysed by γ‐interferon enzyme‐linked immunospot assays after prestimulation of peripheral blood mononuclear cells in the presence of anti‐CD3‐antibodies and interleukin‐2. Our results suggest that the T cell repertoire contains bcr3/abl2‐reactive T cells in CML patients who are in cytogenetic remission, but also in some healthy individuals.

Summary In order to detect T cells against several chronic myeloid leukaemia (CML)‐associated antigens we used: (i) a novel T‐cell assay [cytometric bead array (CBA)]; (ii) γ‐interferon enzyme‐linked immunoSPOT (γ‐IFN‐ELISpot); and (iii) tetramer staining in peripheral blood from CML patients. Peptide‐specific cytokine release was detected by CBA in some patients, whereas standard γ‐IFN‐ELISpot and tetramer staining were negative in the vast majority of cases. In CBA, peptide‐specific cytokine release was predominantly tumour necrosis factor‐α, raising questions about the responding cells and their functional status. CBA appears to be a new useful tool for the detection of leukaemia‐reactive T cells.

T cells recognizing truly leukemia-specific antigens like bcr/abl or other leukemia-associated antigens such as proteinase-3 and WT-1 are present in the T cell repertoire of leukemia patients and several studies have suggested that anti-leukemic T cells might be of clinical relevance. Detection of these leukemia-reactive T cells in CML patients is generally hampered by their low frequency making in vitro prestimulation necessary in most cases. Furthermore, shaping of the T cell repertoire by deletion of high affinity leukemia-reactive T cells and functional unresponsiveness are likely to have additional influence on T cell detection in structural or functional assays. Aim of our study was the detection of T cells specific for different CML-associated antigens chronic phase CML patients, most of them under imatinib treatment. In peripheral blood of CML patients we used a novel T cell assay (cytometric bead array = CBA), g-interferon ELISpot and tetramer staining in order to detect CML-reactive T cells against several known and new epitopes from bcr3/abl2, proteinase-3, c-abl and SOCS-2. By the standard g-interferon ELISpot peptide-specific T cells were not detectable in the vast majority of patients (33/34). In contrast peptide-specific cytokine release was detected in some of the patients by CBA upon pulsing with peptides from bcr3/abl2, proteinase-3, c-abl and SOCS-2. Peptide-specific cytokine release was detected particulary against a HLA-B8-restricted peptide from the fusion region of bcr3/abl2 (3/9 cases), against a HLA-A2-restricted epitope from proteinase-3 (3/11 cases), against a HLA-A2-restricted peptide from SOCS-2 (4/11 cases) and against two HLA-A2 and -B8-restricted epitopes from c-abl (3/11 and 2/9 cases respectively). T cell reactivity against the HLA-class-I epitopes from c-abl and SOCS-2 is described here for the first time. Interestingly, peptide-specific cytokine release was predominantly TNF-a, a significant IFN-g secretion was detected only in a few cases raising questions about the responding cells and their functional status. By tetramer staining low frequency T cells recognizing the bcr3/abl2 fusion protein were detected only in 2/15 patients, but after prestimulation of PBMC bcr/abl-specific T cells could be detected in 4/8 HLA-B8+ patients. Low T cell frequeny and deletion of high avidity T cells is a general obstacle for immune monitoring in cancer patients which may explain negative results obtained by g-IFN-ELISpot or tetramer staining. Furthermore, in CML an altered cytokine secretion pattern of T cells might be an additional limitation of functional T cell assays. Summarizing, we have found T cells recognizing CML-associated antigens which display a TNF-a-dominated cytokine secretion profile. This would have been missed using a standard g-IFN-ELISpot assay both because of the cytokine pattern of the T cells and the sensitivity of the assay. Using CBA there is a higher chance to detect low frequency leukemia-reactive T cells and to identify new immunotherapeutic targets in CML.