Enumeration and characterization of circulating anti–MAGE-A3 T cells in patient CIP-26 and analysis of anti–MAGE-A3 T cells infiltrating tumor lesions. (A) Clinical evolution of patient CIP-26 and timing of treatments. PD indicates progressive disease; CR, complete response; and PR, partial response. (B) Frequency of circulating anti–MAGE-A3 T cells estimated before, during, and after treatment. The frequency of total anti–MAGE-A3 T cells was measured as described in Figure 2. P indicates after infusion; P30*, this blood sample was withdrawn in 2007, 1 year after the 30th vaccination. (C) HLA restriction of a representative MAGE-A3–specific CD8⁺ microculture. The microculture released IFN-γ against autologous M3-GMLs preincubated with 4E mAb (recognizing HLA-B and -C alleles) or with anti-NGFr mAb (control mAb). Release of IFN-γ was instead inhibited in the presence of W6/32 mAb (anti–HLA-I mAb). Error bars represent SD of experimental replicates. (D) Relationship between the increase of MAGE-A3–specific effectors and the development of MAGE-A3–specific DTH reaction. (E) Ex vivo IFN-γ release assay performed on anti–MAGE-A3 T cells from a regressing tumor lesion and a regional lymph node collected in 2005. T cells recognized autologous M3-GMLs but not mock-transduced lymphocytes (UT-GMLs). (F) Only CD8⁺ purified TILs specifically released IFN-γ when challenged with M3-GMLs.