Fully human IgG and IgM antibodies directed against the carcinoembryonic antigen (CEA) Gold 4 epitope and designed for radioimmunotherapy (RIT) of colorectal cancers

BACKGROUND: Human monoclonal antibodies (MAbs) are needed for colon cancer radioimmunotherapy (RIT) to allow for repeated injections. Carcinoembryonic antigen (CEA) being the reference antigen for immunotargeting of these tumors, we developed human anti-CEA MAbs. METHODS: XenoMouse(®)-G2 animals were immunized with CEA. Among all the antibodies produced, two of them, VG-IgG2κ and VG-IgM, were selected for characterization in vitro in comparison with the human-mouse chimeric anti-CEA MAb X4 using flow cytometry, surface plasmon resonance, and binding to radiolabeled soluble CEA and in vivo in human colon carcinoma LS174T bearing nude mice. RESULTS: Flow cytometry analysis demonstrated binding of MAbs on CEA-expressing cells without any binding on NCA-expressing human granulocytes. In a competitive binding assay using five reference MAbs, directed against the five Gold CEA epitopes, VG-IgG2κ and VG-IgM were shown to be directed against the Gold 4 epitope. The affinities of purified VG-IgG2κ and VG-IgM were determined to be 0.19 ± 0.06 × 10(8 )M(-1 )and 1.30 ± 0.06 × 10(8 )M(-1), respectively, as compared with 0.61 ± 0.05 × 10(8 )M(-1 )for the reference MAb X4. In a soluble phase assay, the binding capacities of VG-IgG2κ and VG-IgM to soluble CEA were clearly lower than that of the control chimeric MAb X4. A human MAb concentration of about 10(-7 )M was needed to precipitate approximatively 1 ng (125)I-rhCEA as compared with 10(-9 )M for MAb X4, suggesting a preferential binding of the human MAbs to solid phase CEA. In vivo, 24 h post-injection, (125)I-VG-IgG2κ demonstrated a high tumor uptake (25.4 ± 7.3%ID/g), close to that of (131)I-X4 (21.7 ± 7.2%ID/g). At 72 h post-injection, (125)I-VG-IgG2κ was still concentrated in the tumor (28.4 ± 11.0%ID/g) whereas the tumor concentration of (131)I-X4 was significantly reduced (12.5 ± 4.8%ID/g). At no time after injection was there any accumulation of the radiolabeled MAbs in normal tissues. A pertinent analysis of VG-IgM biodistribution was not possible in this mouse model in which IgM displays a very short half-life due to poly-Ig receptor expression in the liver. CONCLUSION: Our human anti-CEA IgG2κ is a promising candidate for radioimmunotherapy in intact form, as F(ab')(2 )fragments, or as a bispecific antibody.