Evaluation of the CD8+ Tumor Associated vs. Non-Tumor Associated Immune Repertoire in Cancer Patients Following Induction of Lymphopenia and Following CD3/4-1BB Based Expansion.

Background: Most tumor antigens are self antigens, therefore an effective immune response to cancer must break self tolerance. Current models hold that lymphopenia results in proliferation toward self antigens. We sought to determine whether the frequency of T cells specific for self antigens, both tumor associated (TA) and non-tumor associated (NTA) were increased in cancer patients, before or after the induction of lymphopenia. Further, we sought to expand T cells specific for self antigens using artificial APCs which signal via CD3 and 4-1BB.

Methods: Using a broad panel of HLA-A2 tetramers, we used flow cytometry to enumerate circulating CD8+ T cells which recognize viral (CMV, EBV, flu), NTA self antigens (PR-1, WT1, MART-1 and CEA) and TA self antigens (NY-ESO1, PRAME, Ofa/iLRP, CYP239/190, Survivin and hTERT) in normal donors (ND) (n=13), cancer patients with Ewings sarcoma studied prior to chemotherapy (n=7) and following cytotoxic chemotherapy (n=5). To exclude non-specific tetramer binding, two tetramers with differing fluorochromes were included in each tube and cells simultaneously binding both tetramers were excluded as non-specific and binding frequency to control tetramers was subtracted.

Results: The frequency of T cells with specificity for viral antigens (range 0.0%–0.7%) were similar between ND and cancer patients, with no significant increase seen following lymphopenia. For NTA self antigens frequencies prior to chemotherapy were similar between ND and cancer patients respectively, with the lowest mean frequency seen in CEA tetramer specific cells (0.084% ND vs. 0.088% pts), then PR-1 (0.19% ND vs. 0.019% pts), then WT-1 (0.43% ND vs. 0.20% pts) then MART-1 CD8+ specific tetramer cells (0.82% ND vs. 0.83% pts). Remarkably, the frequency of MART-1 specific cells approximates that seen for CMV in both patients and normal donors. For TA self antigens, patients showed trends toward increased frequencies compared to ND but these were not statistically significant. Following lymphopenia, there was no consistent or statistically significant increase in the frequency of cells with specificity for either TA or NTA self antigens. In an attempt to increase the frequency of tumor reactive T cells for potential use in adoptive cellular immunotherapy, artificial APCs were used to deliver a signal via CD3 and 4-1BB and changes in antigen specificity via tetramers were monitored. This method consistently increased the frequency of viral and self antigen specific (TA and NTA) cells with some cultures demonstrating frequencies for individual TA self antigens which approached 10%.

Conclusions: 1. Both normal donors and cancer patients demonstrate sizable repertoires of self-reactive T cells, with the frequency of MART-1 specific cells similar to that seen toward CMV. 2. Patients with cancer retain cells which recognize tumor antigens. However, cells recognizing TA self antigens are only marginally increased compared to that found in normal individuals, and are found at frequencies similar to NTA self antigens implicating inefficient priming by the tumor. 3. Lymphopenia does not significantly augment the frequency of cells responding to TA or NTA self antigens. 4. Sizable expansions of T cells responding to TA self antigens can be induced using CD3/4-1BB based expansion.